Alternative Splicing: Role in Cancer Development and Progression

International Journal of Cell Biology

Guest Editors: Claudia Ghigna, Michael Ladomery, and Claudio Sette Alternative Splicing: Role in Cancer Development and Progression International Journal of Cell Biology

Alternative Splicing: Role in Cancer Development and Progression

This is a special issue published in “International Journal of Celllogy.” Bio All articles are open access articles distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Editorial Board

Paul N. Adler, USA Wiljan Hendriks, The Netherlands Liza Pon, USA Emad Alnemri, USA Paul J. Higgins, USA Jerome Rattner, Canada Avri Ben-Ze’ev, Israel Michael Hortsch, USA Maria Roubelakis, Greece Jeannette Chloe Bulinski, USA Pavel Hozak, Czech Republic Afshin Samali, Ireland Michael Bustin, USA Jeremy Hyams, France Michael Peter Sarras, USA John Cooper, USA Anton M. Jetten, USA Hirofumi Sawai, Japan Adrienne D. Cox, USA Edward M. Johnson, USA R. Seger, Israel J. R. Davie, Canada Daniel P. Kiehart, USA Barry D. Shur, USA Govindan Dayanithi, France Sharad Kumar, Australia Arnoud Sonnenberg, The Netherlands Arun M. Dharmarajan, Australia Paul Marks, USA Gary S. Stein, USA Dara Dunican, Ireland Seamus J. Martin, Ireland Tung Tien Sun, USA William Dunn, USA Manuela Martins-Green, USA Ming Tan, USA Victor Faundez, USA Takeshi Noda, Japan Guido Tarone, Italy Roland Foisner, Austria Helga Ogmundsd¨ ottir,´ Iceland Jean-Pierre Tassan, France Hans Hermann Gerdes, Norway Shoichiro Ono, USA Richard Tucker, USA Richard Gomer, USA Howard Beverley Osborne, France Andre Van Wijnen, USA Hinrich Gronemeyer, France Markus Paulmichl, Austria Gerhard Wiche, Austria Mehran Haidari, USA H. Benjamin Peng, Hong Kong Steve Winder, UK Thomas Hays, USA Craig Pikaard, USA Timothy J. Yen, USA Contents

Alternative Splicing: Role in Cancer Development and Progression, Claudio Sette, Michael Ladomery, and Claudia Ghigna Volume 2013, Article ID 421606, 2 pages

Oncogenic Alternative Splicing Switches: Role in Cancer Progression and Prospects for Therapy, Serena Bonomi, Stefania Gallo, Morena Catillo, Daniela Pignataro, Giuseppe Biamonti, and Claudia Ghigna Volume 2013, Article ID 962038, 17 pages

U1 snRNP-Dependent Suppression of Polyadenylation: Physiological Role and Therapeutic Opportunities in Cancer, Lee Spraggon and Luca Cartegni Volume 2013, Article ID 846510, 10 pages

Role of Pseudoexons and Pseudointrons in Human Cancer, Maurizio Romano, Emanuele Buratti, and Diana Baralle Volume 2013, Article ID 810572, 16 pages

RNA Splicing: A New Player in the DNA Damage Response, Silvia C. Lenzken, Alessia Loffreda, and Silvia M. L. Barabino Volume 2013, Article ID 153634, 12 pages

Aberrant Alternative Splicing Is Another Hallmark of Cancer,MichaelLadomery Volume 2013, Article ID 463786, 6 pages

Ewing Sarcoma Protein: A Key Player in Human Cancer, Maria Paola Paronetto Volume 2013, Article ID 642853, 12 pages

Regulation of the Ras-MAPK and PI3K-mTOR Signalling Pathways by Alternative Splicing in Cancer, Zahava Siegfried, Serena Bonomi, Claudia Ghigna, and Rotem Karni Volume 2013, Article ID 568931, 9 pages

Phosphorylation-Mediated Regulation of Alternative Splicing in Cancer, Chiara Naro and Claudio Sette Volume 2013, Article ID 151839, 15 pages

Alternative Splicing Regulation of Cancer-Related Pathways in Caenorhabditis elegans:AnInVivo Model System with a Powerful Reverse Genetics Toolbox,SergioBarberan-Soler´ and James Matthew Ragle Volume 2013, Article ID 636050, 10 pages

Alternative Splicing Programs in Prostate Cancer,ClaudioSette Volume 2013, Article ID 458727, 10 pages

Expression of Tra2𝛽 in Cancer Cells as a Potential Contributory Factor to Neoplasia and Metastasis, Andrew Best, Caroline Dagliesh, Ingrid Ehrmann, Mahsa Kheirollahi-Kouhestani, Alison Tyson-Capper, and David J. Elliott Volume 2013, Article ID 843781, 9 pages

Clinical Significance of HER-2 Splice Variants in Breast Cancer Progression and Drug Resistance, Claire Jackson, David Browell, Hannah Gautrey, and Alison Tyson-Capper Volume 2013, Article ID 973584, 8 pages Hindawi Publishing Corporation International Journal of Cell Biology Volume 2013, Article ID 421606, 2 pages http://dx.doi.org/10.1155/2013/421606

Editorial Alternative Splicing: Role in Cancer Development and Progression

Claudio Sette,1,2 Michael Ladomery,3 and Claudia Ghigna4

1 Dipartimento di Biomedicina e Prevenzione, Universita` di Roma Tor Vergata, Via Montpellier 1, 00133 Roma, Italy 2 Fondazione Santa Lucia, Laboratorio di Neuroembriologia, Via del Fosso di Fiorano 64, 00143 Roma, Italy 3 Faculty of Health and Life Sciences, University of the West of England, Coldharbour Lane, Bristol BS16 1QY, UK 4 Istituto di Genetica Molecolare, Consiglio Nazionale delle Ricerche (IGM-CNR), Via Abbiategrasso 207, 27100 Pavia, Italy

Correspondence should be addressed to Claudia Ghigna; [email protected]

Received 22 September 2013; Accepted 22 September 2013

Copyright © 2013 Claudio Sette et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Alternative splicing of precursor messenger RNAs (pre- the role of alternative splicing in cancer still remain to be mRNAs) is a fundamental step in the regulation of gene addressed. expression. This processing step of the nascent messenger Themainfocusofthisspecialissueistoemphasizekey amplifies the coding potential of eukaryotic genomes by mechanisms involved in oncogenic splicing changes, their allowing the production of multiple protein isoforms with connection with other steps of gene expression, and the distinct structural and functional properties. The advent therapeutic potential of cancer-associated alternative splicing of high-throughput sequencing techniques has recently isoforms. revealed that alternative splicing of exons and introns repre- More specifically, M. Ladomery discusses alternative sents a major source of proteomic diversity in complex organ- splicing in the context of the so-called hallmarks of cancer, isms characterized by a limited number of protein-coding originally proposed by Hanahan and Weinberg in 2000. genes. Nevertheless, the evolutionary advantage provided by The list of hallmarks was originally six; recently itwas alternative splicing can also turn into a source of deleterious augmented to ten. M. Ladomery proposes that a compre- problems for the organism. Indeed, the extreme flexibility hensive dysregulation of alternative splicing could, in itself, of its regulation, which relies on the combinatorial action be considered yet another hallmark of cancer. The idea is of multiple non stringent factors, is subject to errors and that the aberrant expression and activity of key oncogenic the aberrant splicing of key genes can result in the onset of splicing factors and/or their regulatory kinases could lead many human genetic and sporadic diseases. In this regard, to a systematic change in gene expression by favouring the mounting evidence illustrates how changes in alternative concurrent production of several oncogenic splice variants of splicing patterns of specific genes is an important tool used genes involved in critical biological aspects of tumour cells. by cancer cells to produce protein isoforms involved in all S. C. Lenzken et al. review our current knowledge of the areas of cancer cell biology, including numerous aspects of role of alternative splicing in the multiple and various aspects tumor establishment, progression, and resistance to thera- of the DNA damage response (DDR) and the control of peutic treatments. Importantly, cancer-specific splice variants genome stability. This review illustrates several mechanisms have the potential to become suitable therapeutic targets for through which pre-mRNA splicing and genomic stability can human cancer, as novel tools to correct splicing defects are influence each other and contribute to tumorigenesis. being developed and, in some cases, have entered clinical M.Romanoandcolleaguesdrawattentiontothefunction trials for other human diseases, such as spinal muscular that pseudoexons and pseudointrons can play directly in atrophy. Nevertheless, these findings are likely to represent cancer pathology. These sequences can be found in genes just the tip of the iceberg and important questions regarding that have well-established roles in cancer, including BRCA1, 2 International Journal of Cell Biology

BRCA2, NF-1, and ATM. They describe the mechanisms the regulation of alternative splicing is likely to present through which pseudoexons and pseudointrons can be acti- novelopportunitiesinthediagnosis,prognosis,andtreat- vated or repressed. In addition, they discuss their potential ment of prostate cancer. S. Bonomi et al. deal with novel use as tumour biomarkers to provide a more detailed staging information on how alternative splice variants of many and grading of cancer. cancer-related genes can directly contribute to the oncogenic C. Naro and C. Sette discuss the key role that reversible phenotype, focusing on a number of processes involved in phosphorylation plays in the regulation of alternative splic- cancer progression, such as response to hypoxia, migra- ing. Both splice factors and core components of the spliceo- tion, invasion, and metastasis. Furthermore, they discuss some are affected by phosphorylation. The review focuses some significant examples of alternative splicing isoforms ontheroleofproteinkinasesandphosphataseswhose selectively expressed by tumors and not by normal tissues, activity has specifically been linked to aberrant alternative which may not only represent diagnostic and prognostic splicing associated with neoplastic transformation. Moreover, tumor biomarkers, but also provide potential targets for the it illustrates the fact that signal transduction routes that are development of new therapeutic strategies. frequently altered in cancer cells, such as the RAS/ERK and In their article, L. Spraggon and L. Cartegni focus on the the PI3K/AKT pathways, can modulate alternative splicing role of U1 snRNP, an essential component of the splicing events through the direct or indirect phosphorylation of machinery, in the regulation of alternative polyadenylation splicing regulatory proteins. Thus, protein kinases and phos- and they strongly emphasize its implications in cancer patho- phatases involved in this step of gene expression regulation genesis. Moreover, this review underlines the interesting may provide exciting opportunities for novel drug design. possibility of manipulating this U1 snRNP function for A. Best et al. describe the emerging role of Tra2𝛽,an anticancer therapeutic purposes. SR-related protein, in human cancer. The gene encoding this Lastly,S.Barberan-SolerandJ.M.Raglegiveanoverview splicing factor is amplified in various types of cancer and of the advantages of using the nematode Caenorhabditis the increased expression of Tra2𝛽 is associated with cancer elegans to study splicing regulation in vivo.Importantly, cell survival. Interestingly, the Tra2𝛽 gene is a transcriptional a large percentage of genes undergo alternative splicing target of the proto-oncogene ETS-1, whereas known Tra2𝛽 also in this simple and genetically useful organism. A big splicing targets play key roles in cancer cells, where they affect proportion of these events are functional, conserved, and metastasis, proliferation, and cell survival. These observa- under strict regulation across development, suggesting that tions point to regulation of Tra2𝛽 expression in cancer cells their investigation is likely to provide general mechanisms of as an important step in tumorigenesis. regulation that can be applied also to human genes. Notably, The review by Z. Siegfried et al. gives a series of specific the review illustrates several examples of alternatively spliced examples that cover misregulated alternative splicing events genes that have human homologues implicated in cancer affecting both the Ras-MAPK and PI3K-mTOR signalling biology. pathways during carcinogenesis. These pathways show exten- We hope that this special issue will attract the attention of sive crosstalk and are commonly altered in many cancers by researchers on new progresses in the fields of alternative splic- genetic and epigenetic aberrations. This article also addresses ing and cancer. In particular, the articles presented herein how these signalling pathways play key roles in the transmis- might highlight how this posttranscriptional mechanism of sion of extracellular signals to the splicing machinery and to gene expression plays important roles in the generation specific RNA-binding proteins that ultimately regulate exon of oncogenes and tumor suppressors, describe its interplay definition events. with signaling pathways, and suggest how our knowledge C. Jackson et al. give an overview on a topic of significant of these processes is leading to a better comprehension clinical interest: the roles (often opposed) of the HER2 splice of malignant transformation, thus helping develop novel isoforms in breast cancer progression and drug resistance. therapeutic strategies for the treatment of cancers. M. P. Paronetto describes the function of the Ewing Sarcoma protein (EWS) in cancer biology. EWS is best Claudio Sette known for its involvement in translocations associated with Michael Ladomery sarcomas. Recent evidence has implicated EWS in the regu- Claudia Ghigna lation of DNA damage response (DDR) in cancer. EWS is a multifunctional protein thought to help coordinate multiple steps in the synthesis and processing of pre-mRNAs. This review illustrates in detail the biochemical features and the physiological roles of this RNA binding protein and provides some hints on its possible contribution to human cancer. Other two reviews give a series of specific examples of cancer-associated splicing variants. C. Sette discusses the growing evidence that dysregulated alternative splicing is a major factor in the remarkable biological heterogeneity of prostate cancer. Key genes, including the androgen receptor itself, are alternatively spliced, thereby expressing isoforms with opposing functions. The review also illustrates how Hindawi Publishing Corporation International Journal of Cell Biology Volume 2013, Article ID 962038, 17 pages http://dx.doi.org/10.1155/2013/962038

Review Article Oncogenic Alternative Splicing Switches: Role in Cancer Progression and Prospects for Therapy

Serena Bonomi, Stefania Gallo, Morena Catillo, Daniela Pignataro, Giuseppe Biamonti, and Claudia Ghigna Istituto di Genetica Molecolare, Consiglio Nazionale delle Ricerche (IGM-CNR), Via Abbiategrasso 207, 27100 Pavia, Italy

Correspondence should be addressed to Claudia Ghigna; [email protected]

Received 30 May 2013; Accepted 12 August 2013

Academic Editor: Claudio Sette

Copyright © 2013 Serena Bonomi et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Alterations in the abundance or activities of alternative splicing regulators generate alternatively spliced variants that contribute to multiple aspects of tumor establishment, progression and resistance to therapeutic treatments. Notably, many cancer-associated genes are regulated through alternative splicing suggesting a significant role of this post-transcriptional regulatory mechanism in the production of oncogenes and tumor suppressors. Thus, the study of alternative splicing in cancer might provide a better understanding of the malignant transformation and identify novel pathways that are uniquely relevant to tumorigenesis. Understanding the molecular underpinnings of cancer-associated alternative splicing isoforms will not only help to explain many fundamental hallmarks of cancer, but will also offer unprecedented opportunities to improve the efficacy of anti-cancer treatments.

1. Introduction intronic splicing silencers (ESSs and ISEs) inhibiting splice sites recognition. The regulation of alternative splicing has Alternative splicing is the process by which splice sites in been discussed in several excellent reviews [6–8]. precursor messenger RNAs (pre-mRNAs) are differentially Changes in alternative splicing patterns have an essential selected and paired to produce multiple mature mRNAs role in normal development, differentiation, and in response and protein isoforms with distinct structural and functional to physiological stimuli, but aberrant splicing generates vari- properties. The first example of alternative splicing was ants that contribute to multiple aspects of tumor establish- discovered almost 30 years ago, when membrane-bound ment and progression and in the resistance to therapeutic andsecretedantibodiesweredemonstratedtobeencoded treatments [5, 9, 10]. Many cancer-associated splicing iso- bythesamegene[1, 2]. Now, we know that alternative forms are expressed during embryonic development, but not splicing is a very accurate, efficient, and extraordinarily in normal adult tissues, whereas others are entirely novel flexible process that regulates all major aspects of eukaryotic transcripts [11]. Central to the splicing oncogenic switch cell biology. Affecting approximately 94% of human genes are changes in the expression, activity, or post-translational [3, 4], it represents the major source of the human proteomic modification of splicing regulatory factors, such as SR and diversity. hnRNP proteins [5, 9]. Thus, modification of alternative splic- Regulation of alternative splicing decisions involves the ing profiles contemporaneously affects multiple key aspects recognition of target sequences in the pre-mRNA by a num- of cancer cell biology, including control of cell proliferation, ber of splicing regulatory factors with antagonistic functions cancer metabolism, angiogenesis, evasion from apoptosis, such as SR (serine-arginine-rich) and hnRNP (heterogeneous invasiveness, and metastasis [5, 9, 10]. nuclear ribonucleoprotein) protein families [5]. Generally, Here, we discuss aberrant alternative splicing networks SR proteins promote exon recognition by binding to exonic that contribute to the oncogenic phenotype and have a promi- or intronic splicing enhancer sequences (ESEs and ISEs, nent role in important aspects of tumorigenesis process, resp.), while hnRNP factors typically interact with exonic or including response to hypoxia and cancer cell invasion and 2 International Journal of Cell Biology metastasis. In addition, we also discuss important questions enzymes involved in glycolysis (Glut-1 and PDK-1), angiogen- connected to the role of alternative splicing in cancer: what esis, iron metabolism (transferrin), cell adhesion molecules are the relevant splicing switches that are critical to malignant (MIC2), fibronectin and matrix metalloproteinase (MMP-2), transformation? How the amounts/activity of the splicing extracellular matrix (ECM) modifying enzymes such as lysyl regulatory factors modulate these splicing switches? What are oxidase, and pluripotency factors including OCT4, NANOG, the main functions of cancer-associated alternatively spliced and SOX2 [19–21]. variants? By illustrating specific examples, it will be clear how One of the primary targets of HIF-1 is VEGFA (Vascular the production of cancer-related isoforms offers the potential Endothelial Growth Factor A), a cytokine that promotes todevelopnoveldiagnostic,prognostic,andmorespecific blood vessel growth and stimulates angiogenesis [22]. VEGF, anticancer therapies. secreted by hypoxic cancer cells, diffuses through the tumor mass to neighbouring host normal vessels, where, upon binding to its receptor VEGFR2, it stimulates endothelial 2. Alternative Splicing Changes of Cancer Cells cells to proliferate, migrate toward the tumor, and form in Response to Hypoxia new capillaries [23]. Even though the angiogenesis process does not initiate malignancy, it promotes tumor growth by Through the activation of oncogenes and inactivation of allowing oxygen and nutrients to reach proliferating cancer tumor suppressor genes, cancer cells become able to pro- cells [23]. Alternative splicing occurs extensively within liferate, survive, and resist to apoptosis. Nevertheless, also VEGF pre-mRNA, generating various isoforms with different microenvironmental signaling plays a crucial role in con- C-terminal domains and distinct affinity for its receptors trolling cancer cell homeostasis, metabolism, growth, and and with a non-redundant role in angiogenesis [24, 25]. differentiation12 [ ]. The microenvironment in solid tumors VEGF isoforms are classified in two families, called VEGFxxx is very distinct from that in normal tissues and the cross-talk (pro-angiogenic) and VEGFxxxb(anti-angiogenic),where between cancer and stromal cells contributes to the formation xxx denotes the position of the amino acid residue in the of a clinically relevant tumor and to response to antitumor mature protein. Pro- or anti-angiogenic isoforms are gener- therapy [13, 14]. Modifications of the microenvironment ated through alternative splicing of the mutually exclusive 󸀠 (mostofthesestartearlyduringtumorprogression)result terminal exons 8a and 8b. Selection of the proximal 5 splice 󸀠 from metabolic alterations in cancer cells and from recruit- site (PSS) in exon 8a generates VEGFxxx whereas the distal 5 ment or activating of nontumoral cells, including blood and splice site (DSS) in exon 8b results in VEGFxxxbisoforms.The lymphatic endothelial cells, pericytes, carcinoma-associated two types of isoforms bind VEGFR2 (Vascular Endothelial fibroblasts, bone marrow-derived cells, and immune and Growth Factor Receptor 2) with equal affinity. However, inflammatory cells [15, 16]. In this altered microenvironment VEGFxxx activates the downstream signaling pathways and cancer cells are exposed to pro-proliferative growth factors. In induces angiogenesis, while VEGFxxxbblocksthisprocess addition, transformed cells often hijack the signaling circuits [26]. Given these characteristics, it is not surprising that acting on normal cells in order to become independent VEGFxxxb is preferentially expressed in normal tissues and it from external stimulation to grow and proliferate [12, 13]. is downregulated in cancer [25]. The choice between PSS or Due to deregulated cancer cell metabolism (the consequence DSS depends upon the activity of the splicing factors of the of uncontrolled and rapid proliferation) and to an altered SR family SRSF1 and SRSF6: SRSF6 preferentially selects the structure and functionality of tumor blood vessels, the tumor DSS and promotes the production of anti-angiogenic VEGF microenvironment is characterized by hypoxia and acidosis isoforms, whereas SRSF1 mainly activates the proximal PSS [15, 17, 18]. Hypoxic tumor microenvironments are now [27]. Thus, the altered expression of SRSF1 and SRSF6 in many recognized as a selective pressures that promote tumor types of tumors could impact malignant transformation by aggressiveness, inducing cancer cells to metastasize and ensuring the proper balance of pro- and anti-angiogenic making them refractory to radiotherapy and chemotherapy. isoforms in cancer cells [28, 29]. Importantly, microenviro- Cells cope with hypoxia by activating transcription factor mental factors can influence alternative splicing of VEGF. HIF-1 (hypoxia-inducible factor-1), a heterodimer formed For example, treatment with TGF-𝛽1increasestheanti- 𝛽 𝛽 by the constitutively expressed subunit (HIF-1 )andthe angiogenic VEGFxxxb isoforms via p38 MAPK signaling [27]. inducible 𝛼 subunit (HIF-1𝛼)[19]. The regulation of HIF-1 Conversely, Insulin-like Growth Factor 1 (IGF-1) decreases activity is mainly at the protein level [19]. Under non-hypoxic the expression of VEGFxxxb isoforms in retinal pigmented conditions, the HIF-1𝛼 subunit is rapidly ubiquitinylated and epithelial cells through PKC and SRPK1 [30]. SRPK1 is degraded by the pVHL (von Hippel-Lindau tumor suppressor a protein kinase that specifically phosphorylates proteins protein) and the proteasome, thus preventing the dimeriza- containing SR domains and is upregulated in various types tion with HIF-1𝛽 [19]. Under hypoxia, HIF-1𝛼 degradation is of cancer, like pancreatic, breast, and colon carcinoma [31]. suppressed and its level increases rapidly. The HIF-1𝛼/HIF- Recently, a genome-wide analysis of hypoxia-induced 1𝛽 heterodimer translocates to the nucleus where it binds to alternative splicing changes in endothelial cells has identified hypoxia response elements (HRE) in the promoters of target target genes implicated in angiogenesis-mediated cytoskele- genes.Throughtheactivationofmorethan150genes,HIF-1 ton remodeling (cask, itsn1, larp6, sptan1, tpm1,androbo1), affectsimportantbiologicalprocessessuchasangiogenesis, in the synthesis of membrane anchors (pign)andinthe glucose metabolism, cell proliferation, survival, apoptosis, regulation of gene expression (cugbp1 and max)[32]. These and invasion/metastasis [19]. Target genes include several changes are likely part of the adaptation of endothelial cells International Journal of Cell Biology 3 to stressing conditions. The impact of hypoxia on post- observed that some variants are over-expressed in various transcriptionaleventsisprovenalsobyLAMA3-A (Laminin tumors and implicated in cancer cell invasion and metastasis alpha 3),asplicingvariantoftheLAMA3 gene induced by [41, 43, 44]. In particular, CD44v6 and CD44v8 variants are reduction of oxygen supply which promotes cell invasion and associated with tumor progression and poor diagnosis in is associated with a poor prognosis in head and neck cancer several types of carcinoma including breast and colorectal [33, 34]. cancers [45–47]. Interestingly, HIF-1𝛼 is able to increase the As mentioned before, HIF is a heterodimer that acts as a expression of the CD44 mRNA and to upregulate CD44 dominant regulator of adaptive cellular responses to hypoxia. variants containing exons v6 and v7/8 [48]. In line with There are three principal isoforms of the HIF-𝛼 subunit this, Krishnamachary and collaborators have reported that encoded by distinct genes: HIF-1𝛼, HIF-2𝛼 (EPAS1), and HIF- hypoxic regions of breast cancer specimens contain cells 3𝛼. HIF-3𝛼 gene consists of 19 exons and it is subjected to with elevated expression of CD44 [48]. Additional studies extensive alternative splicing leading to the production of at are necessary to identify the molecular mechanisms and least six isoforms [35]. HIF-3𝛼4isoneoftheHIF-3𝛼 splicing signaling pathways regulating CD44 alternative splicing in variants whose functions have recently been linked the response to hypoxia. development of hypervascular malignant meningiomas [36]. Hypoxia-induced alternative splicing changes were also Indeed, HIF-3𝛼4 directly binds to HIF-1𝛼 and suppresses recently investigated in the study of Hirschfeld and colleagues HRE dependent transcription of VEGF.Importantly,HIF- [49]. The YT521 (YTH domain containing 1) is a ubiquitously 3𝛼4 is able to inhibit proliferation and invasion, to reduce expressed nuclear splicing factor containing a novel RNA- neovascularization and glucose metabolism in hypervascular binding domain (YTH domain) necessary for YT521 to meningiomas [36]. directly influence splice site selection. Importantly, low YT521 Another gene induced by hypoxia is Cyr61 (cysteine rich expression was associated with clinical outcome in patients 61)encodingforasecreteproteinthatfunctionsasaproan- with type I endometrial cancer, suggesting its potential role giogenic factor promoting adhesion, migration, and survival as a tumor suppressor [50]. Furthermore, YT521 alternative of vascular endothelial cells [37]. Cyr61 is supposed to be splicing targets are well-known cancer-associated genes such a promoter of tumor progression since its high expression as BRCA2, ESR1, MDM2, VEGF,andCD44 [49, 51–55]. levels were detected in various cancer types [38, 39]. Notably, Similar to other proteins involved in mRNA processing, in addition to transcriptional level, the expression of Cyr61 the expression of YT521 is regulated by alternative splicing. can be also regulated through an alternative splicing event Interestingly, it was reported that under hypoxic conditions that stimulates the retention of intron 3 in Cyr61 mRNA lead- the alternative splicing of YT521 pre-mRNA containing exons ing to the production of a nonsense-mediated decay (NMD) 8 and 9 generates two transcripts that are subjected to sensible transcript [40]. Importantly, this alternative splicing degradation through NMD, suggesting that these AS-NMD event coupled with NMD pathway (called AS-NMD) was events of YT521 could simultaneously affect the processing of reported to be altered in breast cancer and associated with several cancer-associated genes. a shift from an intron 3 retained transcript (IR) toward an Further functional investigations on hypoxia and its intron 3 spliced mRNA (IS) encoding for the biological active impact in regard to alternative splicing of target genes may Cyr61 protein. Moreover, in several breast cancer cell lines contribute to better understanding of a key regulatory epiphe- under hypoxic conditions AS-NMD of Cyr61 was reported nomenon in tumor growth involved in the development of an to enhance the expression of the IS transcript, suggesting aggressive cancer phenotype. that hypoxia-mediated alternative splicing changes could be a central mechanism regulating the Cyr61 expression and its tumor-promoting activity [40]. 3. Invasion and Metastases The CD44 glycoprotein provides another attractive example of hypoxia-induced alternative splicing changes. CD44 The term metastasization is currently used to indicate the is a transmembrane molecule able to bind several ligands, ability of tumor cells to invade adjacent tissues and dissem- including important components of ECM, such as hyaluronic inate toward distant organs [56]. This is a multistep process acid, collagen, fibronectin, laminin, and matrix metallopro- that involves (i) local invasion of tumor cells through the teinases (MMPs), and involved in cell-cell and cell-matrix basement membrane and endothelial walls into the host interactions, migration, and invasion [41]. CD44 pre-mRNA stroma, (ii) intravasation within the blood and the lymphatic consists of 20 exons, 10 of which (v1–v10) undergo alternative circulatory systems; (iii) extravasation into distant tissues, splicing events, thus generating multiple CD44 isoforms and (iv) proliferation of tumor cells allowing growth and with different molecular sizes and with diverse extracel- efficient metastatic colonization [56]. lular domains [41, 42]. The predominant CD44 isoform, As observed for other hallmarks of cancer, gene expres- the low-molecular-weight CD44s (90 kDa standard form), sion programs implicated in the metastatic process are is expressed by several tissues, including hematopoietic, thesamethatparticipateinembryonicdevelopment,mor- fibroblast, and glial cells, whereas high-molecular-weight phogenesis, and wound healing [57, 58]. One of these CD44v variants (140–230 kDa) are restricted to epithelial programs that physiologically pertain to embryogenesis is cells and abundant in epithelial-type carcinomas [41, 42]. the epithelial-to-mesenchymal transition (EMT). The EMT The physiological/pathological functions of most CD44v process involves dedifferentiation steps in which cells lose variants remain still poorly understood. However, it has been their epithelial phenotype to acquire mesenchymal traits 4 International Journal of Cell Biology

[58, 59]. The epithelial cell layer consists of polarized cells dynamically controlled in epithelial and mesenchymal cells with cohesive cell-cell junctions. Through EMT cells undergo through an AS-NMD event [65]. AS-NMD of SRSF1,which 󸀠 an extensive reorganization of the cytoskeletal architecture involves an intron in the 3 UTR region of the gene, decreases with the loss of intercellular junctions and cell polarity and mRNA stability and SRSF1 protein levels and, notably, it is the acquisition of an elongated, fibroblast-like shape. During alteredincoloncancer.Thisscenarioisfurthercomplicated tumor progression, EMT is one of the major routes through by the involvement of another splicing regulator, Sam68. which cancer cells acquire invasive capabilities and spread Sam68, the 68 kD Src-associated protein in mitosis, is a throughout the body as single cells [58, 59]. Importantly, member of the STAR (signal transduction and activation of tumor EMT is a transient process that occurs in a subset RNA) family of RNA-binding proteins [66]. It contains a of cells at the invasive front of the metastasizing primary single KH-type RNA-binding domain and several protein- carcinoma and is reversed at the final metastatic sites, where protein interaction motifs including potential binding sites cells undergo the reverse process, namely, the mesenchymal- for SH2, SH3, and WW domains, which are characteristic to-epithelial transition (MET) [58, 59]. This plasticity and the of signaling transducers [67]. Sam68 has been recognized as redifferentiation of metastatic cells to an epithelial identity a substrate of several kinases, such as members of the Src help in organ colonization, ensuring metastasis formation. family, ERK1/2 and BRK (Breast Tumor Kinase) [65, 68–70]. Moreover, it clearly indicates that EMT is not driven by As such, Sam68 is the first identified “hub factor” able to stable genetic mutations but by activation of gene expression communicate extracellular stimuli to specific RNA splicing programs in response to external cues in the tumor microen- decisions. In particular, directed by ERK1/2 signaling, Sam68 vironment. controls AS-NMD of SRSF1 transcript, thus modulating its Several relevant players involved in EMT/MET cycles protein level. Notably, epithelial cell-derived soluble factors have been identified including transcription factors, growth areabletoinhibitERK1/2signalingtherebyrepressingSam68 factors, cytokines and chemokines, pro-angiogenic factors, phosphorylation, which increases the production of the cell adhesion molecules, modifiers of cytoskeletal organiza- NMD-sensitive transcript of SRSF1. tion, and extracellular matrix-remodeling enzymes [60]. The prevailing models propose that these regulators act largely 3.2. Rac1. The matrix metalloproteinases (MMPs) are the attranscriptionallevel.Indeed,severalpathwaysactivatea most important family of proteinases of the tumor microen- network of transcription factors that promote the expression vironment that degrade structural components of the extra- of mesenchymal markers, such as N-cadherin and vimentin, cellular matrix (ECM), thus regulating proliferation, cell-cell while inhibiting E-cadherin production, not only a key adhesion, angiogenesis, invasion, and metastases [71]. In line component of adherent junctions but also a tumor suppressor with this, MMPs are upregulated in almost all types of human frequently repressed, mutated or degraded during tumor cancer and associated with poor survival [71]. Notably, over- transformation [58, 59]. However, in the last few years, several expression of MMP-3 in mammary and lung epithelial cells studies have highlighted additional layers of EMT con- triggers a cascade of events that determine activation of trol, including epigenetic reprogramming, small noncoding EMT process [72, 73]. Notably, these events, which induce RNAs, translational and post-translational regulations, and tumorigenesis process in transgenic mice, are dependent on alternative splicing changes [10, 57, 61, 62]. In particular, an the expression of a constitutively active alternatively spliced increasing body of evidence indicates that splicing regulation isoform of the Rac1 gene, encoding for a small GTPase of alonecandrivecriticalaspectsofEMT-associatedphenotypic the mammalian Rho family involved in actin cytoskeleton changes [10]. Below, we discuss some interesting examples of organization, cell growth, cell-cell adhesion, and migration specific alternative splicing events that are implicated not just [74]. This splice variant (known as Rac1b) is produced by in the EMT/MET program but also in the different stages of inclusion of the exon 3b that contains the encoding region the metastatic cascade. for a 19-amino acid domain involved in the interaction with regulator and effector molecules [75]. Since Rac1b shows increased expression during progression of several cancer 3.1. Ron. Alternative splicing of the Ron proto-oncogene types [72, 75, 76], it is tempting to speculate that the exon 3b (also known as MST1R) has provided the first example of an (or the 19-amino acid insertion) could offer the opportunity alternative splicing variant linked to the activation of tumor for a selective targeting to develop anticancer therapies that EMT. Ron encodes for a tyrosine kinase receptor involved in block EMT-associated progression towards advanced tumor control of cell dissociation, migration, and matrix invasion. stages. The activity of the Ron receptor is regulated through alternative splicing. In particular, a constitutively active isoform (called ΔRon), which confers increased motility to expressing 3.3. KLF6. The complexity of networks regulating the EMT cellsandaccumulatesduringtumorprogressionofepithelial program makes it very difficult to identify early inducers cancers [63, 64], is generated through skipping of exon 11. The and a unifying molecular basis for this transition. Many oncoprotein SRSF1 deeply impacts cell physiology since its transcription factors have been extensively studied for their upregulation stimulates skipping of exon 11, thus promoting involvement in activation of EMT. Among these are Slug the production of ΔRon isoform that in turn triggers activa- (referred also to as SNAI2) and Twist, two repressors of E- tion of the EMT program increasing the invasive properties cadherin promoter activity [57]. Recent studies have revealed of the cells [64]. Interestingly, SRSF1 expression levels are important roles for specific alternatively spliced variants of International Journal of Cell Biology 5 upstream regulators of Slug and Twist activity. This is the interaction with F-actin [95]. As a consequence of their case of oncogenic splice variant 1 of KLF6 (KLF6-SV1), a reduced capabilities in F-acting binding and polymerization, tumor suppressor gene belonging to the Kruppel-like¨ family over-expression of SV1- or SV2-cortactin shows reduced cell of transcription factors, able to act as functional driver of migration when compared with FL-cortactin over-expressing the entire metastatic cascade through Twist induction [77, cells [95]. One of the factors that control alternative splicing 78]. Notably, KLF6-SV1 antagonizes the tumor suppressive of cortactin transcripts is SPF45 [96]. Initially character- activity of the full-length KLF6 protein and sustains tumor ized as a component of the spliceosome, SPF45 has been growth and dissemination in ovarian and prostate cancer also implicated in the regulation of Fas and fibronectin models [79–81]. Interestingly, increased expression of KLF6- alternative splicing [97, 98]. In their interesting article, Liu SV1 occurs in many tumors and is associated with poor and coworkers have demonstrated that SPF45 mediates cell prognosis in prostate, lung, and ovarian cancers [82]. KLF6- migration and invasion in ovarian cancer cells by promoting SV1hasbeenrecentlyshowntoinduceEMTandtodrive the formation of the FL-cortactin isoform. SPF45 activity aggressive multiorgan metastasis formation in both sub- appears to be controlled via phosphorylation by Clk1 (Cdc2- cutaneous and orthotopic mouse models of breast cancer like kinase). Finally, SPF45 over-expression correlates also [78]. In this process, it acts through a Twist-dependent with increased cortactin phosphorylation by ERK, which mechanism and Twist downregulation reverts the phenotype enhances cortactin-mediated actin polymerization [96, 99]. of KLF6-SV1 over-expressing cells, restoring the expression of epithelial markers [78]. Finally, high levels of KLF6-SV1 were foundassociatedwithincreaseinEMTmarkersinalarge 3.6. MENA. As for invadopodia, filopodia nucleation and cohort of primary breast cancer patients [78]. extension require actin cytoskeleton remodeling and involve several regulators of actin dynamics, such as the Ena/VASP protein family that in mammals includes three members: 3.4. FAM3B. FAM3B isamemberofthenovelFAM3 family MENA (also called ENAH), VASP, and EVL [88]. Up- of cytokine-like genes, predicted to produce at least 7 alter- regulation of MENA has been detected in several human can- natively spliced isoforms [83, 84]. The most studied variant is cers, including breast cancer and melanoma and correlates the secreted form PANDER, so called for its robust expression with invasiveness of breast tumors [100, 101]. MENA pre- and activity in the pancreatic cells [85, 86]. In addition, mRNA is alternatively spliced to generate different isoforms, lowerPANDERlevelshavebeenobservedinhumangastric expressed in a tissue-specific manner [102]. Importantly, cancers in respect to the corresponding normal tissues [87]. the alternative splicing profile of MENA in invasive tumor Recently, Li and colleagues have identified FAM3B-258, a cells is different from non-migratory resident cancer cells. 258-amino acids non-secretory protein, as a novel splicing Non-invasive tumor cells as well as poorly invasive breast 11a variant of FAM3B up-regulated in colon adenocarcinoma cell cancer cells with epithelial morphology express MENA , linesaswellasinhumancolorectaltumors[84]. FAM3B- the epithelial-associated variant generated from inclusion of 258 is able to induce changes typical of an activated EMT exon 11a [102, 103]. This exon is inserted within the EVH2 program, stimulating cell migration and invasion in vitro domain, very close to the F-actin binding motif and the and promoting metastases formation in nude mice. The tetramerization site. This insertion has been predicted to oncogenic abilities of FAM3B-258 require a Slug-mediated affect the ability of MENA tetramers to interact with F- transcriptional repression of E-cadherin and JAM (Junctional actin and thus to drive filopodia and lamellipodia maturation 11a Adhesion Molecule). Knock-down of Slug in FAM3B-258 [103]. In agreement with this, the increment of MENA over-expressing cells restores higher levels of E-cadherin and reduces both the number and the length of filopodia in a JAM and prevents the FAM3B-258-dependent cell invasion 3D culture assay [104]. On the contrary, invasive cancer cells 11a INV [84]. Whether Slug is a direct target of FAM3B-258 or lack MENA and express MENA ,anisoformcontaining whether other signaling effectors are involved remains to be an additional exon, also referred to as exon +++ [103, 105]. INV 11a investigated. The relative abundance of MENA and MENA seems to be important to regulate key stages of the metastatic INV 3.5. Cortactin. To become migratory and invasive, cells must cascade in breast cancer cells. Thus, high levels of MENA extend plasma membrane protrusions (such as lamellipodia enhance coordinated motility, transendothelial migration, and filopodia) forward and overcome the epithelial basement and intravasation of tumor cells, promoting spontaneous layer as a first barrier56 [ , 88]. The formation of invadopodia lung metastases in a murine model of breast cancer [105, 11a has been recently characterized as another important step 106]. In contrast, increased expression of MENA correlates of the EMT program [89–91]. Invadopodia are enriched with decreased invasion, intravasation, and dissemination with a variety of proteins, including actin and actin reg- of cancer cells [106]. Recently, Di Modugno and coworkers ulatory proteins [92, 93]. The filamentous actin (F-actin) have identified another splice variant of human MENA binding protein cortactin is one of the regulators of actin lacking exon 6 and called hMENAΔv6 [104]. Contrary to 11a polymerization/branching involved in invadopodia assembly MENA ,hMENAΔv6 is expressed selectively in invasive and maturation [92, 94]. Three splicing products of cortactin cancer cells with mesenchymal phenotype and is able to have been characterized until now [95]. Unlike the full-length enhance invasiveness of breast cancer cells but only in 11a protein (FL), the SV1 and SV2 variants lack, respectively, the absence of MENA [104]. This evidence suggests that 11a one and two of the six “cortactin repeats” that mediate the MENA behaves as dominant anti-invasive player, making 6 International Journal of Cell Biology alternative splicing of this gene a potential target for anti- to what observed after TNF𝛼 treatment, several splicing cancer therapies. Furthermore, MENA splicing occurs also in isoforms of SVEP1, such as the full-length isoform a and 11a primary breast tumors and in particular MENA -negative isoform e, are up-regulated in both cell lines upon co- tumors display lower level of E-cadherin when compared culture conditions. In parallel embryonic variants g and f are 11a to MENA -positive samples [104], supporting the anti- silenced in adenocarcinoma DA3 cells, whereas no effect is migratory functions of this splicing isoform [104]. observed in pre-osteoblastoma cells [116]. The same authors Splicing of MENA exon 11a is regulated by the expression also observed that the ratio between splicing isoforms of levels of epithelial-specific alternative splicing factors ESRP1 SVEP1 is perturbed after treatment with epigenetic drugs and 2 (Epithelial Splicing Regulatory Proteins 1 and 2), such as DNA demethylating or histone deacetylase inhibitors, two important regulators of the mesenchymal and epithelial supporting a link between the epigenetic organization and splicing signatures [107]. In particular, ESRP proteins, by splicing of SVEP1 pre-mRNA. However, further studies are promoting inclusion of exon 11a and the production of needed to establish the pathological role of the different 11a MENA isoform, cause a drastic reorganization of actin SVEP1 isoforms in the metastatic process. cytoskeleton as well as cell morphology and a reduction of invasive properties [107]. In addition to MENA, ESRPs 3.9. CD44. Several microenvironmental stressors, including control the alternative splicing of several genes involved in nutrient deficiency, low pH, mediators of inflammatory different aspects of the metastatic cascade, such as cell-cell responses, and reactive oxygen species (ROS), can affect and cell-matrix adhesion, actin cytoskeleton organization, successful colonization by disseminating cells at the final cell polarity, and migration [108, 109]. In line with this, metastatic tissues [117]. Disseminating cancer cells can take modification of ESRPs expression levels results in alternative advantage of antioxidant systems to counteract the exposure splicing changes of CD44 and p120-Catenin, a protein found to oxidative stress. The synthesis of reduced glutathione at cell-cell junctions and also involved in signal transduction (GSH), a reducing thiol peptide, protects cancer cells against [108, 109]. ROS-mediated damage and confers resistance to anticancer therapies [118, 119]. A rate-limiting factor for GSH synthesis is 3.7. L1CAM. The penetration of migrating cancer cells into the availability of cysteine and the cystine transporter system tissue barriers, including the basement membrane, is sup- xc- (composed by two subunits, xCT and CD98hc) is essential ported by the proteolytic degradation of the extracellular in the GSH antioxidant mechanism [119, 120]. Recently, a matrix (ECM) through the activity of secreted enzymes, link between the GSH-dependent evasion from oxidative such as matrix metalloproteinases MMP-2 and MMP-9 stress and alternative splicing of CD44 has been identified [110]. The expression and activity of MMPs are regulated [121, 122]. As mentioned before, CD44 has an important through several signals, mainly induced by growth factors role in cell-cell and cell-matrix interactions, migration, and and chemokines, as well as through integrin and extracel- invasion [41]. Through alternative splicing, CD44 pre-mRNA lular matrix-mediated signals [110]. Recently, the alternative generates multiple CD44 high-molecular-weight isoforms splicingofthecelladhesionmoleculeL1 (L1CAM)hasbeen with different extracellular domains [41, 42]. Intriguingly, foundtocontroltheinvasivecapabilitiesoftumorcells CD44v8-10 has been demonstrated to interact with the by regulating MMP-2 and MMP-9 expression and activity cystine transporter xCT, increasing the levels of GSH and, [111]. More specifically, although initially the splicing variant as a consequence, the ability of cancer cells to avoid ROS considered as cancer-associated was SV-L1CAM (lacking of damage [122]. In line with this, CD44v-positive 4T1 mouse exons 2 and 27), only the full-length FL-L1CAM has been breast cancer cells display high levels of GSH and xCT and found up-regulated upon exposure of tumor cells to the pro- enhanced ROS defense compared to CD44v-negative 4T1 metastatic factors TGF-𝛽 and HGF. Importantly, the over- cells [123]. Thus, CD44v-positive 4T1 are able to establish expressionofFL-L1CAMbutnotoftheSVisoformisable lung metastatic lesions in mouse models with an incidence to induce metastasis formation in mice [111]. higher than CD44v-negative cells [123]. Interestingly, downregulation of the splicing regulator ESRP1 in metastatic 4T1 cells shifts the splicing pattern toward the production of 3.8. SVEP1. Invasion and colonization of a secondary organ CD44s and results in suppression of lung metastasis [123]. by disseminating cancer cells are influenced by the microen- On the contrary, forced expression of CD44v8-10 in ESRP1- vironment and the cross-talk between cancer populations depleted cells is sufficient to restore high content of GSH and andcellsinthenicheofthereceivingtissue[56, 112]. The lung colonization potential [123]. cell adhesion molecule SVEP1 has been recently involved in the interactive network that affects breast cancer cells homing to bone niches [113]. SVEP1 expression is stimulated by 4. Diagnostic, Prognostic, and Anticancer TNF𝛼, a pro-inflammatory cytokine able to affect adhesion Therapy Potentials of Alternative Splicing and migration, and to induce EMT [114, 115]. Recently, Glait-Santar and colleagues have investigated the alternative Cancer chemotherapy relies on the expectation that anti- splicing of SVEP1 transcripts in a co-culture model of pre- cancer drugs will preferentially kill rapidly dividing tumor osteoblastic MDA-15 and mammary adenocarcinoma DA3 cells, rather than normal cells. Unfortunately, most pharma- cells, which mimic the molecular interactions in the bone cological approaches for the treatment of solid tumors suffer niche after invasion of breast carcinoma cells [116]. Similar from poor selectivity, which limits the overall dose of drug International Journal of Cell Biology 7

Table 1: Examples of genes that encode cancer-associated alternative splicing variants.

Gene Splice variant Cancer type Function Reference ABCC1 Various internal deletions Ovarian cancer Drug resistance [127] Mdm2 Various internal deletions Ovarian cancer Loss of p53 binding [127] Fibronectin Exclusion of EDB exon Ovarian cancer Tumor angiogenesis [127] MENA Exclusionofexon6 Breastcancer Increase invasiveness and drug [104] resistance MENA Skipping of exon 11A Breast cancer Enhance EMT [103, 106] Ron Skipping of exon 11 Colon and gastric breast cancer Increase motility and invasion [64] HDMX Exclusionofexon6 Soft-tissuesarcoma Increase tumor aggressiveness [132] p73 Inclusionofexon13 Prostatecancer Prostate hyperplasia and malignancy [133] CASP9 Exclusion of a four-exon cassette Non–small cell lung cancer Susceptibility to chemotherapy [134] CASP8 Retained intron Breast cancer Inhibition of caspase [135] APC Skipping of exon 4 Colon cancer Cause familial adenomatous polyposis [136] (FAP) BRCA1 Skipping of exon 18 Breast cancer Breast cancer susceptibility [137] PTEN Retained introns 3 and 5 Breast cancer Pathogenesis of sporadic breast [137] cancers with p53 p53 Retained intron Breast cancer Tumorigenesis [137] 󸀠 KLF6 Alternative 5 ss Prostate cancer Tumor cell proliferation [82] NF1 Inclusion of exon 23a Neurofibromatosis type I Inactive tumor suppressor [138] 󸀠 ASIP Alternative 3 ss Hepatocellular carcinoma Blocks Fas-mediated apoptosis [139] 󸀠 Bcl-X Alternative 5 ss Hepatocellular carcinoma Regulation of apoptosis [140] TACC1 Exon inclusion Gastric cancer Altered centrosome functions [141] 󸀠 TERT Alternative 3 ss Astrocytic gliomas Rescue of telomerase activity [142] CDH17 Exclusionofexon13 Hepatocellularcarcinoma Incidence of tumor recurrence [143]

that can be administered because of unacceptable toxicities Along this line, a SpliceDisease database (http://cmbi to normal tissues. .bjmu.edu.cn/sdisease), which provides information for rela- As shown in previous sections, alternative splicing vari- tionships among gene mutations, splicing defects, and dis- ants of many cancer-related genes can directly contribute to ease, has been recently developed [131].Alistofcancer- the oncogenic phenotype and to the acquisition of resistance specific alternative splicing isoforms is shown in Table 1. to therapeutic treatments [5, 9, 10]. Alternative splicing Cancer-specific splice variants may not only serve as isoforms selectively expressed by tumors and not by normal diagnostic and prognostic tumor biomarkers but also provide tissues may represent suitable targets for new therapeutical potential targets for the development of new therapeutic approaches [11]. In this section, we discuss some significant strategies. Promising avenues towards the development of examples to illustrate how cancer-specific splicing events can more selective anticancer drugs are (i) antibodies against be a powerful source of new diagnostic, prognostic, and tumor-associated markers, (ii) small molecules targeting the therapeutic tools. spliceosome or trans-acting splicing regulatory factors, and Several highly sensitive methods allowed the identifica- (iii) antisense oligonucleotides that prevent the production of tion of cancer-specific splicing isoforms [124–128]. For exam- specific aberrant alternative splicing variants. ple, the splicing profile of ABCC1, Mdm2, and fibronectin transcripts has been used to distinguish normal ovary from 4.1. Monoclonal Antibodies Targeting Splicing Isoforms. Alter- epithelial ovarian cancer [127], whereas altered splicing of native splicing in cancer can generate unique epitopes in MED24, MFI2, SRRT, CD44,andCLK1 has been associated the extracellular domain of cell membrane proteins. Indeed, with metastatic phenotype in breast cancer and poor progno- many receptors involved in cell-cell and cell-matrix inter- sisinpatients[128]. Notably, splicing of hMENA may improve actions undergo alternative splicing and specific splicing the early diagnosis of breast cancer and clinical decision [104], isoforms are associated with human malignancies [152]. whereas the balance between splicing variants of KLF6 and These novel, or embryo-restricted, epitopes seem to be ideally caspase-9 genes could be useful to predict the susceptibility suited for tumor-targeting strategies consisting in the delivery of cancer cells to chemotherapy [129, 130]. of bioactive compounds, for example, monoclonal antibodies 8 International Journal of Cell Biology

Prostimulatory Proinﬂammatory growth factors cytokines (a) Antibodies recognizing Hypoxia cancer-specific splicing isoforms Signaling cascades

(b) N N SRFs O O N

N Drugs targeting spliceosome components or SRFs effectors Proangiogenic 7 m GpppG isoforms of VEGF Promigratory isoforms of (c) AAAAAAAAAAAAAAAAAAA Antiapoptotic isoforms: Ron, fibronectin, CD44, Bcl-XL, Mcl-1L, Bim𝛾1, Bim cortactin, MENA 𝛾2, caspase-9b, IG20 MADD SRF Enhanced translation of oncoproteins Alternative splicing Splicing targeting antisense of oncogenes oligonucleotides (ASOs and TOES)

Enhanced cell proliferation, survival, angiogenesis, migration, and invasion

Figure 1: SRFs (splicing regulatory factors) at the cross-road between oncogenic signaling pathways and targets for anticancer treatments. During tumorigenesis, cancer cells are exposed to stressing conditions such as hypoxia and acidosis. In this altered tumor microenvironment, growth factors and cytokines, provided by either cancer or non-tumoral cells, activate signaling cascades affecting both the activity and/or the expression levels of splicing regulatory factors (SRFs). In the cytoplasm, SRSFs can enhance the translation of oncogenic variants involved in key aspects of cancer cell biology. In the nucleus, SRFs are mainly involved in the regulation of alternative splicing of pre-mRNAs relevant to cancer progression mechanisms, namely, pro-liferation, angiogenesis, survival, invasion, and metastasis. Alternative splicing variants of cancer-related genes represent powerful targets for new therapeutic approaches. (a) Alternative splicing can generate unique epitopes in cell surface proteins that can be targeted by monoclonal antibodies (mAbs), able to lead to down-regulation or neutralization of the specific isoforms. Moreover, mAbs can be also used to selectively deliver bioactive molecules to cancer cells without affecting normal tissues. (b) Small molecules, by interfering with the spliceosome assembly or with the phosphorylation status of SRFs (i.e., SR proteins), can in turn affect the balance of alternative splicing products, preventing the generation of cancer-associated variants. (c) Standard ASOs (antisense oligonucleotides) block the interaction between the splicing machinery and the cognate splicing sequences (splice sites, enhancer or silencer elements), whereas TOES (targeted oligonucleotide enhancers of splicing) oligonucleotides contain a “tail” of ESE sequences to recruit SRFs on a specific alternative exon. By inhibiting or activating specific splicing events, TOES can be used to shift the ratio between biologically functional splice variants toward the production of non-pathological isoforms.

(mAbs). Binding of mAbs to tumor-associated biomarkers of this is the EGFRvIII variant, and two mAbs (Ch806 can determine down-regulation or inhibit the function of and CH12), targeting the unique extramembrane epitope of the target (Figure 1(a)). Moreover, when conjugated with EGFRvIII, are used in clinical trial [155, 156]. Another cancer- radioemitters or chemotherapics, the mAbs can efficiently specific EGFR isoform, de4 EGFR, is also recognized by ensure in situ delivery of the bioactive molecule to cancer mAb CH12 [157, 158]. Treatment with CH12, but not with cells, sparing normal tissues. cetuximab, of mice over-expressing the de4 EGFR variant sig- A very well studied target for the mAbs-mediated therapy nificantly suppresses tumor proliferation and angiogenesis, is the Epidermal Growth Factor Receptor (EGFR), which leading to tumor apoptosis [158]. is over-expressed in several tumors [153]. Two antibodies An important correlation between aberrant alternative directed against the EGFR, cetuximab (C225) and panitu- splicing and tumor progression has been shown for CD44. mumab, are currently used in therapy [154]. Unfortunately, In particular, CD44 isoforms containing the variant exon v6 since EGFR is expressed also in normal tissue, this therapy and v8 (CD44v6 or CD44v8) are commonly over-expressed may have severe side effects. Interestingly, several tumor- in epithelial tumors [159]. Unfortunately, the expression of specific splice variants of EGFR have been identified. One these isoforms is not confined to cancer cells, but it occurs International Journal of Cell Biology 9 also in normal tissues, as skin keratinocytes [160]. Various [173], the most commonly mutated gene in human cancers. mAbs targeting CD44v6 have been evaluated in clinical trials Activation of p53 is promoted by down-regulation of MdmX [152]. Anti-v6 mAbs are effective in treatment of patients and decreased stability of Mdm2, two key repressors of p53. with head and neck squamous cell carcinoma but they Interestingly, TG003 seems to contribute to their degradation show severe skin toxicity [161, 162]. Recently, Masuko and rather than to change their pre-mRNA splicing. Further collaborators have developed a mAb (GV5) against CD44R1, studies are required to elucidate the role of p53 pathway as a CD44 isoform containing exons v8, v9, and v10 [163]. GV5 a sensor of alterations in the splicing machinery. exhibited therapeutic effects in xenografts models, probably As mentioned before, SRSF1 and SRSF6 control the by inducing antibody-dependent cellular cytotoxicity, with choice between VEGFxxx (angiogenic) and VEGFxxxb undetectable reactivity with skin keratinocytes. However, (anti-angiogenic) isoforms. Interestingly, phosphorylation of despite these promising developments, solid tumors are SRSF1 by SRPK1 promotes the use of the proximal splice site frequently resistant to antibody-based therapies probably for within exon 8 of VEGF pre-mRNA and thus the production the poor penetration of antibodies into the tumor tissue. of the angiogenic isoform, while phosphorylation of SRSF6 Neo-vascularization (or angiogenesis) is needed for growth activates the distal splice site in exon 8 [30]. Along this of cancer cells and for the metastatic process [22]. Tumor line, inhibitors of SRPK1 and Clk functions, as SRPIN340 endothelial cells have a central role in this process because or TG003, have been shown to block SRSF1 activation and they are readily accessible to drugs via the blood circulation. to inhibit angiogenesis process both in vitro and in vivo Exploiting this feature, new cancer therapies have been devel- [30, 174]. oped to target the tumor vasculature with the aim to deprive Indole derivatives represent a new class of strong splicing thetumorofoxygenandnutrientsandinduceitsregression inhibitors able to interact with SR proteins and prevent their [164]. Endothelial cells of tumor vessels express splicing phosphorylation [175]. Some of them show anti-proliferative isoforms of matrix proteins such as the fibronectin (FN) [165, activity, with an acceptable toxicity [175]. In our recent study, 166]. For example, the oncofetal isoform containing the extra we have used indole derivatives to modulate the splicing domain EDB of FN is exclusively expressed around newly event that generates the cancer-associated ΔRon variant developing tumor vasculature, whereas it is absent in adult [176]. Binding of SRSF1 to an ESE sequence within exon 12 tissue [165]. Notably, EDB-specific radiolabeled antibodies leads to skipping of exon 11 and to the production of the are used in clinical trial for antiangiogenic cancer treatment constitutively active ΔRon isoform [64]. Interestingly, three [167]. indole derivatives can reverse aberrant ΔRon splicing. Among these, only IDC92 is able to revert the invasive phenotype of cancer cells without affecting the splicing profile of other 4.2. Small Molecules Targeting Splicing Components. The first SRSF1 targets, suggesting that this small molecule is suitable drugsusedtotargetthespliceosomemachinery,FR901464 for further in vivo studies [176]. and herboxidiene, are natural compounds extracted from Because of their lack of specificity in modulating pre- bacteria [168, 169]. Subsequently, synthetic analogues, with mRNA splicing, these compounds are expected to cause less complex structure, therefore with minor synthesis costs, deleterious undesired events in normal cells as well. Surpris- but with higher stability, solubility, and activity, have been ingly, however, most of them have been found to possess obtained [168]. All these molecules have been demonstrated selective tumor cytotoxicity. One hypothesis is that tumor to have selective toxicity and anti-cancer properties in human cells respond differently from normal cells to changes in tumor xenografts [168]. Their molecular mechanism has mRNA splicing. Alternatively, transformed cells may differ been recently elucidated [170]. They bind to the splicing from normal counterparts for the expression of modified factor 3b (SF3b), destabilizing the U2 snRNP-pre-mRNA version of tumor suppressors originated by aberrant splicing complex and altering the conformation of the branch site 󸀠 and drug treatment may reverse this defect [168]. However, sequence. As a consequence, the correct selection of the 3 molecular mechanisms underlying this specificity toward splice acceptor site fails to occur and alternatively spliced cancerous cells remain elusive and additional studies are mRNAs are generated [170]. One more promising antitu- necessary to characterize the effects of these drugs in both mor agent characterized for its anti-proliferative activity is tumor and normal cells. the biflavonoid isoginkgetin, a natural product found ina Several works have recently demonstrated that also variety of plants [171]. Interestingly, isoginkgetin inhibits the inhibitors of oncogenic pathway components can indi- transition from pre-spliceosomal complex A to complex B, rectly target splicing reactions. For example, treatment of probably by preventing stable recruitment of the U4/U5/U6 melanoma cells, harbouring B-Raf (V600E) mutation with nuclear ribonucleoproteins [171]. B-RAF inhibitors, determines over-expression of SRSF6 that SR proteins are targets of extensive phosphorylation that in turn regulates alternative splicing of the Bim gene, a influences protein interactions and regulates their activity member of the Bcl-2 family, promoting the production of the and sub-cellular localization [5](Figure 1(b)). The benzoth- proapoptotic short isoform BimS [177, 178]. iazole compound TG003 has been described as a potent inhibitory of Clk1/Sty able to affect SFSR1-depending alternative splicing events [172]. Interestingly, the exposure of 4.3. Oligonucleotides-Mediated Therapies. Antisense oligonu- human colon carcinoma cells to TG003 has been described cleotides (ASOs) are short oligonucleotides, usually 15–25 to determine accumulation of the tumor suppressor p53 basesinlength,thataredesignedtoannealtoaspecific 10 International Journal of Cell Biology

Table 2: Examples of ASO treatments in cancer cell lines.

Gene Function Reference Retentionofintron1altersthereadingframeandoccursinbreast SRA1, Steroid Receptor RNA Activator gene [144] tumors with high progesterone receptor contents Antiapoptotic protein of the Bcl-2 family overexpressed in many Mcl-1, myeloid cell leukemia-1 [145] tumors erbB-2, Her-2 receptor Skipping of exon 19 produces a dominant-negative protein isoform [146] IG20, death-domain adaptor protein Antiapoptotic alternative splicing isoform expressed in gliomas [147] Insulinoma-Glucagonoma 20 Two isoforms generated by alternative splicing: a proapoptotic and CASP9,caspase9 [148] aprosurvivalvariant The PKM2 isoform is crucial for aerobic glycolysis (the Warburg PKM,pyruvatekinaseM [149, 150] effect) and tumor growth Alternative splicing generates many nonfunctional products. ASOs hTERT,telomerase treatment increases nonfunctional telomerase products in cancer [151] cells

target region on a pre-mRNA molecule, thus interfering with Another example is the exploitation of ASOs to efficiently the splicing reaction (Figure 1(c)). ASO targeting an exon- modify splicing of STAT3, another gene involved in apoptosis. intron junction may sterically block the access to the splic- The usage of an alternative acceptor site within exon 23 of ing machinery, redirecting splicing reaction to an adjacent STAT3 pre-mRNA leads to the production of the truncated splicing site. Alternatively, ASO can bind to splicing enhancer STAT3𝛽 isoform that promotes apoptosis and cell-cycle arrest or silencer elements, masking the sequence to trans-acting [183]. Interestingly, by using a modified ASO, targeting a regulatory factors and determining inclusion or skipping of splicing enhancer element that regulates alternative splicing specific exons. of exon 23, it was possible to promote a shift from STAT3𝛼 to The latest generations of ASOs contain chemical modi- STAT3𝛽 leading to tumor regression in a xenograft model of fications and appear more stable compared to conventional cancer [184]. oligonucleotides. All ASOs share the following characteris- Recently Cartegni’s group showed that the antago- tics: (i) they bind tightly to RNA through Watson-Crick base- nism/association between intronic polyadenylation and pre- pairs; (ii) they are specific for RNA molecule, and (iii) they mRNA splicing can produce truncated soluble receptor do not alter the genomic sequence. In addition, other features tyrosine kinases (RTKs). These isoforms can act as dominant- make them appreciable therapeutic tools. Indeed, the delivery negative regulators since they are deficient of the anchoring technology, usually nanoparticle, is noninvasive, nontoxic, transmembrane and the intracellular kinase domains [185]. efficient, and very stable. Notably, these secreted “decoy receptors” are able to inactivate Clinical trials have been already started that exploit ASOs the associated tumorigenic signaling pathways as a conse- fortreatmentsofhumangeneticdisorders[179]. Even though quence of their ability to titrating the ligand or by block- the use of these molecules in anti-cancer therapy is still at ing the wild-type receptors in non-functional heterodimers early stages [180], several recent works report therapeutically [185]. Interestingly, in the case of the Vascular Endothelial relevant and encouraging results. Below, we describe several Growth Factor Receptor 2 (VEGFR2/KDR), the key molecule studies, some of them performed on xenograft models of involved in the control of the VEGF signaling, morpholino human tumors, while the more recent preliminary results on antisense oligonucleotides (ASOs extremely stable within cancer cell lines are listed in Table 2. biological systems because they are resistant to a wide range of The first demonstration of in vivo anti-tumor efficacy of nucleases) were used to induce the expression of dominant- ASOswasreportedbyBaumanandcolleagues[181]. The negative secreted VEGFR2/KDR and more importantly to authors challenged a modified ASO, targeting the down- inhibit the angiogenesis process [185]. 󸀠 stream 5 alternative splice site of exon 2 in Bcl-X pre-mRNA A new generation of ASOs called TOES (Targeted (Bcl-X ASO), in a mouse model of metastatic melanoma, Oligonucleotide Enhancers of Splicing) have been developed an aggressive malignancy that shows poor prognosis when to induce the inclusion of otherwise skipped exons. TOES are associated with increased expression Bcl-XL splice variant complex modified antisense RNA oligonucleotides formed 󸀠 [182]. The oligo efficiently redirected splicing machinery to by two functionally distinct regions: the 5 half of the oligo 󸀠 the upstream 5 splice site, decreasing the anti-apoptotic is complementary to a sequence within an exon of interest Bcl-XL isoform, while increasing the pro-apoptotic Bcl-XS and is followed by a non-complementary RNA tail, designed variant [181]. Importantly, the administration of the oligo to mimic an ESE sequence (Figure 1(c)). In this manner the coupled to nanoparticles produced a significant reduction oligo recruits specific trans-acting regulatory factors (such of tumor burden in rapidly growing and highly tumorigenic as SR proteins) and provides a sort of enhancer in trans lung metastases [181]. that promotes exon inclusion [180, 186]. TOES have been International Journal of Cell Biology 11 first tested for their ability to induce the inclusion of SMN2 evaluated for clinical trial (Phase II) in combination with exon7inspinalmuscularatrophy(SMA)patientfibroblasts docetaxel for the treatment of prostate cancer, acute myeloid [187]. The TOES technology has been so far applied only leukemia, and non-small cell lung cancer [196, 197]. once to correct splicing in cancer cells [176]. As described before, SRSF1 over-expression produces skipping of Ron exon 5. Conclusions 11 and the production of the oncogenic ΔRon isoform [176]. In order to correct pathological ΔRon splicing,wehavedesigned The most important concept opened by the results reviewed a TOES complementary to the first region of exon 11 and here is that the RNA-binding proteins are at the centre of containing a tail of GGA repeats, known to function as a the oncogenic alternative splicing switch that controls all the strong enhancer. This treatment was able to revert ΔRon major aspects of cancer cell biology (Figure 1). Understand- splicingandtoincreaseexon11inclusion[176], suggesting the ing the molecular basis and the effects of the splicing regula- exciting possibility to consider splicing of exon 11 as a possible tion on the transcriptome of cancer cells promises to identify target of new anti-metastatic therapeutic approaches. keycircuitsthathaveafundamentalroleincellproliferation, Another well-established method to down-regulate a spe- apoptosis, and other aspects of tumor progression. Moreover, cific splicing isoform is through RNA interference (RNAi). despite the progress, significant challenges remain towards Small interfering RNAs (siRNA) are a class of double- the rational design of more specific and selective approaches stranded RNA molecules interfering with the expression able to modulate alternative splicing events in order to control of specific genes with complementary nucleotide sequence cancer growth. through endonucleases-mediated degradation mechanism In the era of the personalized medicine, each therapy [188]. Among the most recent studies on siRNA technology would have to fit the combination of markers specific for each performed in xenograft tumor models, it is worth mentioning patient. Powerful and cost-effective methods are required the article of Sangodkar and collaborators [130]. KLF6-SV1, to evaluate cancer markers, including those generated by an oncogenic splice variant of the tumor suppressor KLF6 alternative splicing, not only to provide a diagnosis and a gene, is significantly up-regulated in several human cancers prognosis but also to suggest the right personalized therapy. [81, 189] and its over-expression is associated with decreased survival in prostate and lung cancers [79, 190]. Sangodkar Disclosure and colleagues demonstrated that knock-down of this variant via RNAi restored chemotherapy sensitivity and induction Serena Bonomi and Stefania Gallo should be considered first of apoptosis in lung cancer cells both in vitro and in vivo authors. [130]. Conversely, over-expression of KLF6-SV1 resulted in a marked reduction in chemotherapy sensitivity in a tumor Acknowledgments xenograft model. Another target for siRNA-mediated anticancer therapy is This work was supported by grants from the Associazione hnRNP L. Like to SRSF1 [191], hnRNP L binds to a splicing ItalianaperlaRicercasulCancro(AIRC),theAssociation regulatory element and regulates the splicing profile of for International Cancer Research (AICR, UK), to Claudia caspase-9 gene [129], which is altered in a large percentage of Ghigna and grants from the Associazione Italiana per la human lung cancer [192]. Recently, it has been reported that Ricerca sul Cancro (AIRC) and Progetto d’Interesse Invec- RNAi-mediated down-regulation of hnRNP L is sufficient to chiamento (CNR/MIUR) to Giuseppe Biamonti. increase the caspase-9a/9b ratio and, more importantly, to cause a complete loss of tumorigenic capacity in xenograft model [192]. References Finally, some of ASOs prevent ribosomal assembly and [1] P.Early, J. Rogers, and M. Davis, “Two mRNAs can be produced hencemRNAtranslationandseemtobewelltoleratedin from a single immunoglobulin 𝜇 gene by alternative RNA patients [193]. An example of this approach is provided by processing pathways,” Cell,vol.20,no.2,pp.313–319,1980. survivin, an inhibitor of apoptotic proteins [194]. Survivin is [2]F.W.Alt,A.L.M.Bothwell,andM.Knapp,“Synthesisof expressed in several human cancers and its over-expression secreted and membrane-bound immunoglobulin Mu heavy is associated with a poor prognosis [195]. These features chains is directed by mRNAs that differ at their 3’ ends,” Cell, make survivin an attractive target for anti-cancer therapy vol. 20, no. 2, pp. 293–301, 1980. and several efforts, so far unsuccessful, have been made [3] E. T. Wang, R. Sandberg, S. Luo et al., “Alternative isoform along this line. Recently, down-regulation of survivin has regulation in human tissue transcriptomes,” Nature,vol.456,no. been achieved by using LY2181308, an ASO targeting the 7221, pp. 470–476, 2008. translation initiation codon of survivin mRNA and inducing [4]Q.Pan,O.Shai,L.J.Lee,B.J.Frey,andB.J.Blencowe,“Deep its RNase H-mediated degradation [196]. LY2181308 treat- surveying of alternative splicing complexity in the human tran- ment in multiple cancer cell lines caused apoptosis through scriptome by high-throughput sequencing,” Nature Genetics, activation of caspase-3. Most importantly, in a xenograft vol.40,no.12,pp.1413–1415,2008. tumor model, LY2181308 produced significant anti-tumor [5] C. Ghigna, C. Valacca, and G. Biamonti, “Alternative splicing activity and sensitized tumor cells to chemotherapeutic- and tumor progression,” Current Genomics,vol.9,no.8,pp. induced apoptosis. All these findings led LY2181308 to be 556–570, 2008. 12 International Journal of Cell Biology

[6] T. W. Nilsen and B. R. Graveley, “Expansion of the eukaryotic [25] Y. Qiu, C. Hoareau-Aveilla, S. Oltean, S. J. Harper, and D. O. proteome by alternative splicing,” Nature,vol.463,no.7280,pp. Bates, “The anti-angiogenic isoforms of VEGF in health and 457–463, 2010. disease,” Biochemical Society Transactions,vol.37,no.6,pp. [7] D. D. Licatalosi and R. B. Darnell, “RNA processing and its 1207–1213, 2009. regulation: global insights into biological networks,” Nature [26] C. Hilmi, M. Guyot, and G. Pages,` “VEGF spliced variants: Reviews Genetics, vol. 11, no. 1, pp. 75–87, 2010. possible role of anti-angiogenesis therapy,” Journal of Nucleic [8] U. Braunschweig, S. Gueroussov, A. M. Plocik, B. R. Graveley, Acids, vol. 2012, Article ID 162692, 7 pages, 2012. and B. J. Blencowe, “Dynamic integration of splicing within [27] D. G. Nowak, J. Woolard, E. M. Amin et al., “Expression of gene regulatory pathways,” Cell,vol.152,no.6,pp.1252–1269, pro- and anti-angiogenic isoforms of VEGF is differentially 2013. regulated by splicing and growth factors,” Journal of Cell Science, [9]C.J.DavidandJ.L.Manley,“Alternativepre-mRNAsplicing vol.121,no.20,pp.3487–3495,2008. regulation in cancer: pathways and programs unhinged,” Genes [28]R.Karni,E.DeStanchina,S.W.Lowe,R.Sinha,D.Mu,andA. and Development,vol.24,no.21,pp.2343–2364,2010. R. Krainer, “The gene encoding the splicing factor SF2/ASF isa proto-oncogene,” Nature Structural and Molecular Biology,vol. [10] G. Biamonti, S. Bonomi, S. Gallo, and C. Ghigna, “Making 14,no.3,pp.185–193,2007. alternative splicing decisions during epithelial-to-mesenchymal transition (EMT),” Cellular and Molecular Life Sciences,vol.69, [29] M. Cohen-Eliav, R. Golan-Gerstl, Z. Siegfried et al., “The no. 15, pp. 2515–2526, 2012. splicing factor SRSF6 is amplified and is an oncoprotein in lung and colon cancers,” The Journal of Pathology,vol.229,no.4,pp. [11] C. He, F. Zhou, Z. Zuo, H. Cheng, and R. Zhou, “A global 630–639, 2013. view of cancer-specific transcript variants by subtractive [30] D. G. Nowak, E. M. Amin, E. S. Rennel et al., “Regulation transcriptome-wide analysis,” PLoS ONE,vol.4,no.3,Article of Vascular Endothelial Growth Factor (VEGF) splicing from ID e4732, 2009. pro-angiogenic to anti-angiogenic isoforms: a novel therapeutic [12] M. Jewer, S. D. Findlay, and L. Postovit, “Post-transcriptional strategy for angiogenesis,” Journal of Biological Chemistry,vol. regulation in cancer progression: microenvironmental control 285, no. 8, pp. 5532–5540, 2010. of alternative splicing and translation,” Journal of Cell Commu- [31] G. M. Hayes, P. E. Carrigan, and L. J. Miller, “Serine-arginine nication and Signaling,vol.6,no.4,pp.233–248,2012. protein kinase 1 overexpression is associated with tumorigenic [13] D. Hanahan and R. A. Weinberg, “Hallmarks of cancer: the next imbalance in mitogen-activated protein kinase pathways in generation,” Cell,vol.144,no.5,pp.646–674,2011. breast, colonic, and pancreatic carcinomas,” Cancer Research, [14] D. Hanahan and R. A. Weinberg, “The hallmarks of cancer,” Cell, vol.67,no.5,pp.2072–2080,2007. vol. 100, no. 1, pp. 57–70, 2000. [32] J. E. Weigand, J. Boeckel, P.Gellert, and S. Dimmeler, “Hypoxia- [15] P. Vaupel, “The role of hypoxia-induced factors in tumor induced alternative splicing in endothelial cells,” PLoS ONE,vol. progression,” Oncologist,vol.9,no.5,pp.10–17,2004. 7, no. 8, Article ID e42697, 2012. [16] G. Lorusso and C. Ruegg,¨ “The tumor microenvironment [33]C.S.Moller-Levet,G.N.J.Betts,A.L.Harris,J.J.Homer,C.M. and its contribution to tumor evolution toward metastasis,” L. West, and C. J. Miller, “Exon array analysis of head and neck Histochemistry and Cell Biology,vol.130,no.6,pp.1091–1103, cancers identifies a hypoxia related splice variant of LAMA3 2008. associated with a poor prognosis,” PLoS Computational Biology, [17] R. A. Gatenby, E. T. Gawlinski, A. F. Gmitro, B. Kaylor, and R. vol. 5, no. 11, Article ID e1000571, 2009. J. Gillies, “Acid-mediated tumor invasion: a multidisciplinary [34] M. P. Marinkovich, “Tumour microenvironment: laminin 332 study,” Cancer Research, vol. 66, no. 10, pp. 5216–5223, 2006. in squamous-cell carcinoma,” Nature Reviews Cancer,vol.7,no. 5,pp.370–380,2007. [18] R. Cairns, I. Papandreou, and N. Denko, “Overcoming physio- logic barriers to cancer treatment by molecularly targeting the [35] M. A. Maynard, H. Qi, J. Chung et al., “Multiple splice variants 𝛼 tumor microenvironment,” Molecular Cancer Research,vol.4, of the human HIF-3 locusaretargetsofthevonHippel-Lindau no. 2, pp. 61–70, 2006. E3 ubiquitin ligase complex,” JournalofBiologicalChemistry, vol.278,no.13,pp.11032–11040,2003. [19] G. L. Semenza, “Targeting HIF-1 for cancer therapy,” Nature Reviews Cancer,vol.3,no.10,pp.721–732,2003. [36]H.Ando,A.Natsume,K.Iwamietal.etal.,“Ahypoxia- inducible factor (HIF)-3𝛼 splicing variant, HIF-3𝛼4impairs [20] C.E.Forristal,K.L.Wright,N.A.Hanley,R.O.C.Oreffo,andF. angiogenesis in hypervascular malignant meningiomas with D. Houghton, “Hypoxia inducible factors regulate pluripotency epigenetically silenced HIF-3𝛼4,” Biochemical and Biophysical and proliferation in human embryonic stem cells cultured at Research Communications,vol.433,no.1,pp.139–144,2013. reduced oxygen tensions,” Reproduction,vol.139,no.1,pp.85– [37] J. A. Menendez,I.Mehmi,D.W.Griggs,andR.Lupu,“The´ 97, 2010. angiogenic factor CYR61 in breast cancer: molecular pathology [21] J. T. Erler, K. L. Bennewith, M. Nicolau et al., “Lysyl oxidase is and therapeutic perspectives,” Endocrine-Related Cancer,vol. essential for hypoxia-induced metastasis,” Nature,vol.440,no. 10,no.2,pp.141–152,2003. 7088, pp. 1222–1226, 2006. [38] D. Xie, D. Yin, H.-J. Wang et al., “Levels of expression of CYR61 [22] H. Gerhardt, “VEGF and endothelial guidance in angiogenic and CTGF are prognostic for tumor progression and survival of sprouting,” Organogenesis,vol.4,no.4,pp.241–246,2008. individuals with gliomas,” Clinical Cancer Research,vol.10,no. [23] M. Potente, H. Gerhardt, and P. Carmeliet, “Basic and thera- 6, pp. 2072–2081, 2004. peutic aspects of angiogenesis,” Cell,vol.146,no.6,pp.873–887, [39] A.-M. Bleau, N. Planque, and B. Perbal, “CCN proteins and 2011. cancer: two to tango,” Frontiers in Bioscience, vol. 10, pp. 998– [24] S. D. Finley and A. S. Popel, “Predicting the effects of anti- 1009, 2005. angiogenic agents targeting specific VEGF isoforms,” BMC [40] M. Hirschfeld, A. Z. Hausen, H. Bettendorf, M. Jager,¨ and E. Biology,vol.14,no.3,pp.500–509,2012. Stickeier, “Alternative splicing of Cyr61 Is regulated by hypoxia International Journal of Cell Biology 13

and significantly changed in breast cancer,” Cancer Research, [57] B. De Craene and G. Berx, “Regulatory networks defining EMT vol. 69, no. 5, pp. 2082–2090, 2009. during cancer initiation and progression,” Nature Reviews,vol. [41]D.Naor,S.Nedvetzki,I.Golan,L.Melnik,andY.Faitelson, 13, no. 2, pp. 97–110, 2013. “CD44 in cancer,” Critical Reviews in Clinical Laboratory Sci- [58] J. Lim and J. P. Thiery, “Epithelial-mesenchymal transitions: ences,vol.39,no.6,pp.527–579,2002. insights from development,” Development,vol.139,no.19,pp. [42] H. Ponta, L. Sherman, and P.A. Herrlich, “CD44: from adhesion 3471–3486, 2012. molecules to signalling regulators,” Nature Reviews Molecular [59] K. Polyak and R. A. Weinberg, “Transitions between epithelial Cell Biology,vol.4,no.1,pp.33–45,2003. and mesenchymal states: acquisition of malignant and stem cell [43] V. Orian-Rousseau, “CD44, a therapeutic target for metastasis- traits,” Nature Reviews Cancer,vol.9,no.4,pp.265–273,2009. ing tumours,” European Journal of Cancer,vol.46,no.7,pp. [60]J.P.Thiery,H.Acloque,R.Y.J.Huang,andM.A.Nieto, 1271–1277, 2010. “Epithelial-mesenchymal transitions in development and dis- [44] P. Klingbeil, R. Natrajan, G. Everitt et al., “CD44 is overex- ease,” Cell,vol.139,no.5,pp.871–890,2009. pressed in basal-like breast cancers but is not a driver of 11p13 [61] M. D. Bullock, A. E. Sayan, G. K. Packham, and A. H. amplification,” Breast Cancer Research and Treatment,vol.120, Mirnezami, “MicroRNAs: critical regulators of epithelial to no.1,pp.95–109,2010. mesenchymal (EMT) and mesenchymal to epithelial transition [45] P. Auvinen, R. Tammi, M. Tammi, R. Johansson, and V.-M. (MET) in cancer progression,” Biology of the Cell/Under the Kosma, “Expression of CD44s, CD44v3 and CD44v6 in benign Auspices of the European Cell Biology Organization,vol.104,no. and malignant breast lesions: correlation and colocalization 1, pp. 3–12, 2012. with hyaluronan,” Histopathology,vol.47,no.4,pp.420–428, 2005. [62]C.Wu,Y.Tsai,M.Wu,S.Teng,andK.Wu,“Epigenetic [46] S. Saito, H. Okabe, M. Watanabe et al. et al., “CD44v6 expression reprogramming and post-transcriptional regulation during the is related to mesenchymal phenotype and poor prognosis in epithelial-mesenchymal transition,” Trends in Genetics,vol.28, patients with colorectal cancer,” Oncology Reports,vol.29,no. no.9,pp.454–463,2012. 4, pp. 1570–1578, 2013. [63]Y.-Q.Zhou,C.He,Y.-Q.Chen,D.Wang,andM.-H.Wang, [47]R.Kopp,M.Fichter,G.Schalhorn,J.Danescu,andS.Classen, “Altered expression of the RON receptor tyrosine kinase in “Frequent expression of the high molecular, 673-bp CD44v3,v8- primary human colorectal adenocarcinomas: generation of 10 variant in colorectal adenomas and carcinomas,” Interna- different splicing RON variants and their oncogenic potential,” tional Journal of Molecular Medicine,vol.24,no.5,pp.677–683, Oncogene,vol.22,no.2,pp.186–197,2003. 2009. [64] C. Ghigna, S. Giordano, H. Shen et al., “Cell motility is [48] B. Krishnamachary, M. Penet, S. Nimmagadda et al., “Hypoxia controlled by SF2/ASF through alternative splicing of the Ron regulates CD44 and its variant isoforms through HIF-1𝛼 in protooncogene,” Molecular Cell,vol.20,no.6,pp.881–890,2005. triple negative breast cancer,” PLoS ONE,vol.7,no.8,Article [65] C. Valacca, S. Bonomi, E. Buratti et al., “Sam68 regulates ID e44078, 2012. EMT through alternative splicing-activated nonsense-mediated [49] M. Hirschfeld, B. Zhang, M. Jaeger et al., “Hypoxia-dependent mRNA decay of the SF2/ASF proto-oncogene,” Journal of Cell mRNA expression pattern of splicing factor YT521 and its Biology,vol.191,no.1,pp.87–99,2010. impact on oncological important target gene expression,” [66] P. Bielli, R. Busa,` M. P. Paronetto, and C. Sette, “The RNA- Molecular Carcinogenesis,2013. binding protein Sam68 is a multifunctional player in human [50] B. Zhang, A. Z. Hausen, M. Orlowska-Volk et al., “Alternative cancer,” Endocrine-Related Cancer,vol.18,no.4,pp.R91–R102, splicing-related factor YT521: an independent prognostic factor 2011. in endometrial cancer,” International Journal of Gynecological [67] K. E. Lukong and S. Richard, “Sam68, the KH domain- Cancer,vol.20,no.4,pp.492–499,2010. containing superSTAR,” Biochimica et Biophysica Acta,vol.1653, [51] S. C. Jeyaraj, D. M. O’Brien, and D. S. Chandler, “MDM2 no. 2, pp. 73–86, 2003. and MDM4 splicing: an integral part of the cancer spliceome,” [68] C. Brignatz, M. P. Paronetto, S. Opi et al., “Alternative splicing Frontiers in Bioscience,vol.14,no.7,pp.2647–2656,2009. modulates autoinhibition and SH3 accessibility in the Src kinase [52] J. M. Garc´ıa Pedrero, P. Zuazua, C. Mart´ınez-Campa, P. S. Fyn,” Molecular and Cellular Biology,vol.29,no.24,pp.6438– Lazo, and S. Ramos, “The naturally occurring variant of 6448, 2009. estrogen receptor (ER) ERΔE7 suppresses estrogen-dependent transcriptional activation by both wild-type ER𝛼 and ER𝛽,” [69] K. E. Lukong, D. Larocque, A. L. Tyner, and S. Richard, Endocrinology,vol.144,no.7,pp.2967–2976,2003. “Tyrosine phosphorylation of Sam68 by breast tumor kinase [53]D.J.Farrugia,M.K.Agarwal,V.S.Pankratzetal.,“Functional regulates intranuclear localization and cell cycle progression,” assays for classification of BRCA2 variants of uncertain signifi- JournalofBiologicalChemistry,vol.280,no.46,pp.38639– cance,” Cancer Research,vol.68,no.9,pp.3523–3531,2008. 38647, 2005. [54] M. Matsuyama, T. Chijiwa, Y. Inoue et al., “Alternative splicing [70] N.Matter,P.Herrlich,andH.Konig,¨ “Signal-dependent regula- variantofvascularendothelialgrowthfactor-Aisacritical tion of splicing via phosphorylation of Sam68,” Nature,vol.420, prognostic factor in non-small cell lung cancer,” Oncology no.6916,pp.691–695,2002. Reports,vol.22,no.6,pp.1407–1413,2009. [71] M. Egeblad and Z. Werb, “New functions for the matrix met- [55]D.M.W.Cork,T.W.J.Lennard,andA.J.Tyson-Capper, alloproteinases in cancer progression,” Nature Reviews Cancer, “Alternative splicing and the progesterone receptor in breast vol. 2, no. 3, pp. 161–174, 2002. cancer,” Breast Cancer Research,vol.10,no.3,article207,2008. [72] M. L. Stallings-Mann, J. Waldmann, Y. Zhang et al. et al., [56] S. Valastyan and R. A. Weinberg, “Tumor metastasis: molecular “Matrix metalloproteinase induction of Rac1b, a key effector of insights and evolving paradigms,” Cell,vol.147,no.2,pp.275– lung cancer progression,” Science Translational Medicine,vol.4, 292, 2011. no. 142, Article ID 142ra95, 2012. 14 International Journal of Cell Biology

[73] D. C. Radisky, D. D. Levy, L. E. Littlepage et al., “Rac1b and [90] B. Gligorijevic, J. Wyckoff, H. Yamaguchi, Y. Wang, E. T. reactive oxygen species mediate MMP-3-induced EMT and Roussos, and J. Condeelis, “N-WASP-mediated invadopodium genomic instability,” Nature,vol.436,no.7047,pp.123–127, formation is involved in intravasation and lung metastasis of 2005. mammary tumors,” Journal of Cell Science,vol.125,no.3,pp. [74] E. E. Bosco, J. C. Mulloy, and Y. Zheng, “Rac1 GTPase: a “Rac” 724–734, 2012. of all trades,” Cellular and Molecular Life Sciences,vol.66,no.3, [91] M. A. Eckert, T. M. Lwin, A. T. Chang et al., “Twist1-induced pp. 370–374, 2009. invadopodia formation promotes tumor metastasis,” Cancer [75] P.Jordan,R.Brazao,M.G.Boavida,C.Gespach,andE.Chastre,˜ Cell,vol.19,no.3,pp.372–386,2011. “Cloning of a novel human Rac1b splice variant with increased [92] M. Oser, H. Yamaguchi, C. C. Mader et al., “Cortactin reg- expression in colorectal tumors,” Oncogene,vol.18,no.48,pp. ulates cofilin and N-WASp activities to control the stages 6835–6839, 1999. of invadopodium assembly and maturation,” Journal of Cell [76] A. Schnelzer, D. Prechtel, U. Knaus et al., “Rac1 in human breast Biology,vol.186,no.4,pp.571–587,2009. cancer: overexpression, mutation analysis, and characterization [93] R. Buccione, G. Caldieri, and I. Ayala, “Invadopodia: specialized of a new isoform, Rac1b,” Oncogene,vol.19,no.26,pp.3013– tumor cell structures for the focal degradation of the extracellu- 3020, 2000. lar matrix,” Cancer and Metastasis Reviews,vol.28,no.1-2,pp. [77] B. B. McConnell and V. W. Yang, “Mammalian Kruppel-Like¨ 137–149, 2009. factors in health and diseases,” Physiological Reviews,vol.90,no. [94] C. C. Mader, M. Oser, M. A. O. Magalhaes et al., “An EGFR- 4, pp. 1337–1381, 2010. Src-Arg-cortactin pathway mediates functional maturation of [78]R.Hatami,A.M.Sieuwerts,S.Izadmehretal.etal.,“KLF6- invadopodia and breast cancer cell invasion,” Cancer Research, SV1drivesbreastcancermetastasisandisassociatedwithpoor vol.71,no.5,pp.1730–1741,2011. survival,” Science Translational Medicine,vol.5,no.169,Article [95] A. G. S. H. Van Rossum, J. H. De Graaf, E. Schuuring-Scholtes et ID 169ra12, 2013. al., “Alternative splicing of the actin binding domain of human [79] G. Narla, A. DiFeo, Y. Fernandez et al., “KLF6-SV1 overexpres- cortactin affects cell migration,” JournalofBiologicalChemistry, sion accelerates human and mouse prostate cancer progression vol. 278, no. 46, pp. 45672–45679, 2003. and metastasis,” Journal of Clinical Investigation,vol.118,no.8, [96] Y. Liu, L. Conaway, J. Rutherford Bethard et al., “Phosphoryla- pp.2711–2721,2008. tion of the alternative mRNA splicing factor 45 (SPF45) by Clk1 [80] G. Narla, A. Difeo, S. Yao et al., “Targeted inhibition of the regulates its splice site utilization, cell migration and invasion,” KLF6 splice variant, KLF6 SV1, suppresses prostate cancer cell Nucleic Acids Research,vol.41,no.9,pp.4949–4962,2013. growth and spread,” Cancer Research, vol. 65, no. 13, pp. 5761– [97] A. M. Al-Ayoubi, H. Zheng, Y. Liu, T. Bai, and S. T. Eblen, 5768, 2005. “Mitogen-activated protein kinase phosphorylation of splicing [81] A. DiFeo, G. Narla, J. Hirshfeld et al., “Roles of KLF6 and factor 45 (SPF45) regulates SPF45 alternative splicing site KLF6-SV1 in ovarian cancer progression and intraperitoneal utilization, proliferation, and cell adhesion,” Molecular and dissemination,” Clinical Cancer Research,vol.12,no.12,pp. Cellular Biology,vol.32,no.14,pp.2880–2893,2012. 3730–3739, 2006. [98] L. Corsini, S. Bonna, J. Basquin et al., “U2AF-homology motif [82] A. DiFeo, J. A. Martignetti, and G. Narla, “The role of KLF6 and interactions are required for alternative splicing regulation by its splice variants in cancer therapy,” Drug Resistance Updates, SPF45,” Nature Structural & Molecular Biology,vol.14,no.7,pp. vol.12,no.1-2,pp.1–7,2009. 620–629, 2007. [83] Y. Zhu, G. Xu, A. Patel et al., “Cloning, expression, and initial [99]N.Martinez-Quiles,H.-Y.H.Ho,M.W.Kirschner,N.Ramesh, characterization of a novel cytokine-like gene family,” Genomics, andR.S.Geha,“Erk/Srcphosphorylationofcortactinactsas vol.80,no.2,pp.144–150,2002. a switch on-switch off mechanism that controls its ability to [84]Z.Li,H.Mou,T.Wangetal.etal.,“Anon-secretoryform activate N-WASP,” Molecular and Cellular Biology,vol.24,no. of FAM3B promotes invasion and metastasis of human colon 12, pp. 5269–5280, 2004. cancer cells by upregulating Slug expression,” Cancer Letters, [100]F.DiModugno,M.Mottolese,A.DiBenedettoetal.,“The vol. 328, no. 2, pp. 278–284, 2013. cytoskeleton regulatory protein hMena (ENAH) is overex- [85] C. E. Robert-Cooperman, J. R. Carnegie, C. G. Wilson et al., pressed in human benign breast lesions with high risk of “Targeted disruption of pancreatic-derived factor (PANDER, transformation and human epidermal growth factor receptor- FAM3B) impairs pancreatic 𝛽-cell function,” Diabetes,vol.59, 2-positive/hormonal receptor-negative tumors,” Clinical Cancer no. 9, pp. 2209–2218, 2010. Research,vol.12,no.5,pp.1470–1478,2006. [86] J. R. Carnegie, C. E. Robert-Cooperman, J. Wu, R. A. Young, [101] F. Di Modugno, G. Bronzi, M. J. Scanlan et al., “Human Mena B. A. Wolf, and B. R. Burkhardt, “Characterization of the protein, a serex-defined antigen overexpressed in breast cancer + expression, localization, and secretion of PANDER in 𝛼-cells,” eliciting both humoral and CD8 T-cell immune response,” Molecular and Cellular Endocrinology,vol.325,no.1-2,pp.36– International Journal of Cancer,vol.109,no.6,pp.909–918, 45, 2010. 2004. [87] H.-L. Huang, B.-Y. Wu, S.-B. Yang et al., “Down-regulation of [102]F.DiModugno,L.DeMonte,M.Balsamoetal.,“Molecular PANDER mRNA in gastric cancer,” World Chinese Journal of cloning of hMena (ENAH) and its splice variant hMena +11a: Digestology, vol. 16, no. 14, pp. 1513–1518, 2008. epidermal growth factor increases their expression and stim- [88] A. J. Ridley, “Life at the leading edge,” Cell,vol.145,no.7,pp. ulates hMena+11a phosphorylation in breast cancer cell lines,” 1012–1022, 2011. Cancer Research, vol. 67, no. 6, pp. 2657–2665, 2007. [89] M. Yilmaz and G. Christofori, “EMT, the cytoskeleton, and [103] S. Goswami, U. Philippar, D. Sun et al., “Identification of cancer cell invasion,” Cancer and Metastasis Reviews,vol.28,no. invasion specific splice variants of the cytoskeletal protein Mena 1-2, pp. 15–33, 2009. present in mammary tumor cells during invasion in vivo,” International Journal of Cell Biology 15

Clinical and Experimental Metastasis,vol.26,no.2,pp.153–159, [121]T.Yae,K.Tsuchihashi,T.Ishimotoetal.etal.,“Alternative 2009. splicing of CD44 mRNA by ESRP1 enhances lung colonization [104] F. Di Modugno, P. Iapicca, A. Boudreau et al. et al., “Splicing of metastatic cancer cell,” Nature Communications,vol.3,article program of human MENA produces a previously undescribed 883, 2012. isoform associated with invasive, mesenchymal-like breast [122] T. Ishimoto, O. Nagano, T. Yae et al., “CD44 variant regulates tumors,” Proceedings of the National Academy of Sciences of the redox status in cancer cells by stabilizing the xCT subunit of United States of America,vol.109,no.47,pp.19280–19285,2012. system xc- and thereby promotes tumor growth,” Cancer Cell, [105]U.Philippar,E.T.Roussos,M.Oseretal.,“AMenainvasion vol. 19, no. 3, pp. 387–400, 2011. isoform potentiates EGF-induced carcinoma cell invasion and [123] M. Yoshikawa, K. Tsuchihashi, T. Ishimoto et al. et al., “xCT metastasis,” Developmental Cell,vol.15,no.6,pp.813–828,2008. inhibition depletes CD44v-expressing tumor cells that are [106]E.T.Roussos,M.Balsamo,S.K.Alfordetal.,“Menainvasive resistant to EGFR-targeted therapy in head and neck squamous (MenaINV) promotes multicellular streaming motility and cell carcinoma,” Cancer Research,vol.73,no.6,pp.1855–1866, transendothelial migration in a mouse model of breast cancer,” 2013. Journal of Cell Science, vol. 124, no. 13, pp. 2120–2131, 2011. [124]J.P.Venables,J.Brosseau,G.Gadeaetal.etal.,“RBFOX2isan [107]C.C.WarzechaandR.P.Carstens,“Complexchangesinalter- important regulator of mesenchymal tissue-specific splicing in native pre-mRNA splicing play a central role in the epithelial- both normal and cancer tissues,” Molecular and Cellular Biology, to-mesenchymal transition (EMT),” Seminars in Cancer Biol- vol.33,no.2,pp.396–405,2013. ogy,vol.22,no.5-6,pp.417–427,2012. [125]C.M.Misquitta-Ali,E.Cheng,D.O’Hanlonetal.,“Global [108]C.C.Warzecha,P.Jiang,K.Amirikianetal.,“AnESRP- profiling and molecular characterization of alternative splicing regulated splicing programme is abrogated during the events misregulated in lung cancer,” Molecular and Cellular epithelial-mesenchymal transition,” EMBO Journal,vol.29,no. Biology, vol. 31, no. 1, pp. 138–150, 2011. 19, pp. 3286–3300, 2010. [126] R. Menon and G. S. Omenn, “Identification of alternatively [109] C. C. Warzecha, S. Shen, Y. Xing, and R. P. Carstens, “The spliced transcripts using a proteomic informatics approach,” epithelial splicing factors ESRP1 and ESRP2 positively and Methods in Molecular Biology,vol.696,pp.319–326,2011. negatively regulate diverse types of alternative splicing events,” [127] R. Klinck, A. Bramard, L. Inkel et al., “Multiple alternative RNA Biology,vol.6,no.5,pp.546–562,2009. splicing markers for ovarian cancer,” Cancer Research,vol.68, [110] M. Bjorklund¨ and E. Koivunen, “Gelatinase-mediated migra- no.3,pp.657–663,2008. tion and invasion of cancer cells,” Biochimica et Biophysica Acta, vol.1755,no.1,pp.37–69,2005. [128]A.Bemmo,C.Dias,A.A.N.R.Rose,C.Russo,P.Siegel, and J. Majewski, “Exon-Level transcriptome profiling in murine [111] S. Hauser, L. Bickel, D. Weinspach et al., “Full-length L1CAM breast cancer reveals splicing changes specific to tumors with and not its 𝛿2𝛿27 splice variant promotes metastasis through different metastatic abilities,” PLoS ONE,vol.5,no.8,Article induction of gelatinase expression,” PLoS ONE,vol.6,no.4, ID e11981, 2010. Article ID e18989, 2011. [129] J. C. Shultz, R. W. Goehe, D. S. Wijesinghe et al., “Alternative [112] J. E. Talmadge and I. J. Fidler, “AACR centennial series: the splicing of caspase 9 is modulated by the phosphoinositide 3- biology of cancer metastasis: historical perspective,” Cancer kinase/Akt pathway via phosphorylation of SRp30a,” Cancer Research, vol. 70, no. 14, pp. 5649–5669, 2010. Research,vol.70,no.22,pp.9185–9196,2010. [113] I. Shur, E. Zemer-Tov, R. Socher, and D. Benayahu, “SVEP1 expression is regulated in estrogen-dependent manner,” Journal [130] J. Sangodkar, A. DiFeo, L. Feld et al., “Targeted reduction of of Cellular Physiology,vol.210,no.3,pp.732–739,2007. KLF6-SV1 restores chemotherapy sensitivity in resistant lung adenocarcinoma,” Lung Cancer,vol.66,no.3,pp.292–297,2009. [114] C.-W. Li, W. Xia, L. Huo et al., “Epithelial-mesenchymal transition induced by TNF-𝛼 requires NF-𝜅B-mediated transcrip- [131] J. Wang, J. Zhang, K. Li, W. Zhao, and Q. Cui, “SpliceDisease tional upregulation of Twist1,” Cancer Research,vol.72,no.5, database: linking RNA splicing and disease,” Nucleic Acids pp.1290–1300,2012. Research,vol.40,pp.D1055–D1059,2012. [115] C. Glait-Santar and D. Benayahu, “Regulation of SVEP1 gene [132]F.Bartel,J.Schulz,A.Bohnke¨ et al., “Significance of HDMX-S expression by 17𝛽-estradiol and TNF𝛼 in pre-osteoblastic and (or MDM4) mRNA splice variant overexpression and HDMX mammary adenocarcinoma cells,” Journal of Steroid Biochem- gene amplification on primary soft tissue sarcoma prognosis,” istry and Molecular Biology,vol.130,no.1-2,pp.36–44,2012. International Journal of Cancer,vol.117,no.3,pp.469–475,2005. [116] C. Glait-Santar, M. Pasmanik-Chor, and D. Benayahu, “Expres- [133]D.A.Arvanitis,E.Lianos,N.Soulitzis,D.Delakas,andD.A. sion pattern of SVEP1 alternatively-spliced forms,” Gene,vol. Spandidos, “Deregulation of p73 isoform equilibrium in benign 505, no. 1, pp. 137–145, 2012. prostate hyperplasia and prostate cancer,” Oncology Reports,vol. [117] G. P. Gupta and J. Massague,´ “Cancer metastasis: building a 12, no. 5, pp. 1131–1137, 2004. framework,” Cell,vol.127,no.4,pp.679–695,2006. [134] J.C.Shultz,R.W.Goehe,C.S.Murudkaretal.,“SRSF1regulates [118] D. Trachootham, J. Alexandre, and P. Huang, “Targeting can- the alternative splicing of caspase 9 via a novel intronic splicing cer cells by ROS-mediated mechanisms: a radical therapeutic enhancer affecting the chemotherapeutic sensitivity of non- approach?” Nature Reviews Drug Discovery,vol.8,no.7,pp. small cell lung cancer cells,” Molecular Cancer Research,vol.9, 579–591, 2009. no. 7, pp. 889–900, 2011. [119] O. Nagano, S. Okazaki, and H. Saya, “Redox regulation in stem- [135] A. Mohr, R. M. Zwacka, G. Jarmy et al., “Caspase-8L expression + like cancer cells by CD44 variant isoforms,” Oncogene,2013. protects CD34 hematopoietic progenitor cells and leukemic [120] M. Lo, Y.-Z. Wang, and P. W. Gout, “The xc-cystine/glutamate cells from CD95-mediated apoptosis,” Oncogene,vol.24,no.14, antiporter: a potential target for therapy of cancer and other pp. 2421–2429, 2005. diseases,” Journal of Cellular Physiology,vol.215,no.3,pp.593– [136] D. W. Neklason, C. H. Solomon, A. L. Dalton, S. K. Kuwada, 602, 2008. and R. W.Burt, “Intron 4 mutation in APC gene results in splice 16 International Journal of Cell Biology

defect and attenuated FAP phenotype,” Familial Cancer,vol.3, receptor and extracellular matrix-related targets for antibody- no. 1, pp. 35–40, 2004. based therapy of cancer,” Cancer Genomics and Proteomics,vol. [137] N. Okumura, H. Yoshida, Y. Kitagishi, Y. Nishimura, and S. 8, no. 5, pp. 211–226, 2011. Matsuda, “Alternative splicings on p53, BRCA1 and PTEN genes [153] I. Okamoto, L. C. Kenyon, D. R. Emlet et al., “Expression of involved in breast cancer,” Biochemical and Biophysical Research constitutively activated EGFRvIII in non-small cell lung cancer,” Communications, vol. 413, no. 3, pp. 395–399, 2011. Cancer Science,vol.94,no.1,pp.50–56,2003. [138] V.A. Barron and H. Lou, “Alternative splicing of the neurofibro- [154] C. Gridelli, P. Maione, M. L. Ferrara, and A. Rossi, “Cetuximab matosis type I pre-mRNA,” Bioscience Reports,vol.32,no.2,pp. and other anti-epidermal growth factor receptor monoclonal 131–138, 2012. antibodies in the treatment of non-small cell lung cancer,” [139] Y. Hu, C. Fang, and Y. Xu, “The effect of isoforms of the cell Oncologist,vol.14,no.6,pp.601–611,2009. polarity protein, human ASIP, on the cell cycle and Fas/FasL- [155] A. M. Scott, F.-T. Lee, N. Tebbutt et al., “A phase I clinical trial mediated apoptosis in human hepatoma cells,” Cellular and with monoclonal antibody ch806 targeting transitional state Molecular Life Sciences,vol.62,no.17,pp.1974–1983,2005. and mutant epidermal growth factor receptors,” Proceedings of [140] T. Takehara, X. Liu, J. Fujimoto, S. L. Friedman, and H. Taka- the National Academy of Sciences of the United States of America, hashi, “Expression and role of Bcl-xL in human hepatocellular vol. 104, no. 40, pp. 4071–4076, 2007. carcinomas,” Hepatology,vol.34,no.1,pp.55–61,2001. [156] K. Klausz, S. Berger, J. J. Lammerts van Bueren et al., [141] A.Line,Z.Slucka,A.Stengrevics,G.Li,andR.C.Rees,“Altered “Complement-mediated tumor-specific cell lysis by antibody splicing pattern of TACC1 mRNA in gastric cancer,” Cancer combinations targeting epidermal growth factor receptor Genetics and Cytogenetics,vol.139,no.1,pp.78–83,2002. (EGFR) and its variant III (EGFRvIII),” Cancer Science,vol.102, [142]V.Kotoula,S.Barbanis,E.Nikolakaki,D.Koufoyannis,C. no.10,pp.1761–1768,2011. S. Papadimitriou, and G. Karkavelas, “Relative expression of [157] H. Wang, M. Zhou, B. Shi et al., “Identification of an exon human telomerase catalytic subunit (hTERT) transcripts in 4-deletion variant of epidermal growth factor receptor with astrocytic gliomas,” Acta Neuropathologica,vol.107,no.5,pp. increased metastasis-promoting capacity,” Neoplasia,vol.13,no. 443–451, 2004. 5, pp. 461–471, 2011. [143] X. Q. Wang, J. M. Luk, P. P. Leung, B. W. Wong, E. J. [158]H.Wang,B.Shi,Q.Zhangetal.,“Growthandmetastasis Stanbridge, and S. T. Fan, “Alternative mRNA splicing of suppression of glioma xenografts expressing exon 4-deletion liver intestine-cadherin in hepatocellular carcinoma,” Clinical variant of epidermal growth factor receptor by monoclonal anti- Cancer Research, vol. 11, no. 2, pp. 483–489, 2005. body CH12-mediated receptor degradation,” FASEB Journal, [144] C. Cooper, J. Guo, Y. Yan et al., “Increasing the relative vol.26,no.1,pp.73–80,2012. expression of endogenous non-coding Steroid Receptor RNA [159] S. Seiter, W. Tilgen, K. Herrmann et al., “Expression of CD44 Activator (SRA) in human breast cancer cells using modified splice variants in human skin and epidermal tumours,” Virchows oligonucleotides,” Nucleic Acids Research,vol.37,no.13,pp. Archiv,vol.428,no.3,pp.141–149,1996. 4518–4531, 2009. [160]B.MackandO.Gires,“CD44sandCD44v6expressioninhead [145] J.-J. Shieh, K.-T. Liu, S.-W. Huang, Y.-J. Chen, and T.-Y. Hsieh, and neck epithelia,” PLoS ONE,vol.3,no.10,ArticleIDe3360, “Modification of alternative splicing of Mcl-1 pre-mRNA using 2008. antisense morpholino oligonucleotides induces apoptosis in [161] H. Riechelmann, A. Sauter, W.Golze et al., “Phase I trial with the basal cell carcinoma cells,” Journal of Investigative Dermatology, CD44v6-targeting immunoconjugate bivatuzumab mertansine vol. 129, no. 10, pp. 2497–2506, 2009. in head and neck squamous cell carcinoma,” Oral Oncology,vol. [146] S. Pankratova, B. N. Nielsen, T. Shiraishi, and P. E. Nielsen, 44,no.9,pp.823–829,2008. “PNA-mediated modulation and redirection of Her-2 pre- [162]P.K.E.Borjesson,E.J.Postema,J.C.Roosetal.,“Phase¨ mRNA splicing: specific skipping of erbB-2 exon 19 coding for I therapy study with 186Re-labeled humanized monoclonal the ATP catalytic domain,” International Journal of Oncology, antibody BIWA4 (Bivatuzumab) in patients with head and neck vol.36,no.1,pp.29–38,2010. squamous cell carcinoma,” Clinical Cancer Research,vol.9,no. [147]C.V.Lefave,M.Squatrito,S.Vorlovaetal.,“Splicingfactor 10, 2003. hnRNPH drives an oncogenic splicing switch in gliomas,” [163] K. Masuko, S. Okazaki, M. Satoh et al., “Anti-tumor effect EMBO Journal, vol. 30, no. 19, pp. 4084–4097, 2011. against human cancer xenografts by a fully human monoclonal [148] J. C. Shultz and C. E. Chalfant, “Caspase 9b: a new target antibody to a variant 8-epitope of CD44R1 expressed on cancer for therapy in non-small-cell lung cancer,” Expert Review of stem cells,” PLoS ONE,vol.7,no.1,ArticleIDe29728,2012. Anticancer Therapy,vol.11,no.4,pp.499–502,2011. [164] P. Carmeliet, “Angiogenesis in life, disease and medicine,” [149] M. S. Goldberg and P. A. Sharp, “Pyruvate kinase M2-specific Nature,vol.438,no.7070,pp.932–936,2005. siRNA induces apoptosis and tumor regression,” Journal of [165] P.Castellani, V.G. Viale g., A. Dorcaratto et al., “The fibronectin Experimental Medicine,vol.209,no.2,pp.217–224,2012. isoform containing the ED-B oncofetal domain: a marker of [150]Z.Wang,H.Y.Jeon,F.Rigo,C.F.Bennett,andA.R.Krainer, angiogenesis,” International Journal of Cancer,vol.59,no.5,pp. “Manipulation of PK-M mutually exclusive alternative splicing 612–618, 1994. by antisense oligonucleotides,” Open Biology,vol.2,no.10,p. [166]E.Ventura,F.Sassi,A.Parodietal.,“Alternativesplicingofthe 120133, 2012. angiogenesis associated extra-domain B of fibronectin regulates [151] M. S. Wong, L. Chen, C. Foster, R. Kainthla, J. W.Shay, and W.E. the accessibility of the B-C loop of the type III repeat 8,” PLoS Wright, “Regulation of telomerase alternative splicing: a target ONE,vol.5,no.2,ArticleIDe9145,2010. for chemotherapy,” Cell Reports,vol.3,no.4,pp.1028–1035. [167] M. Czabanka, G. Parmaksiz, S. H. Bayerl et al., “Microvascular [152] U. H. Weidle, D. Maisel, S. Klostermann, E. H. Weiss, and M. biodistribution of L19-SIP in angiogenesis targeting strategies,” Schmitt, “Differential splicing generates new transmembrane European Journal of Cancer,vol.47,no.8,pp.1276–1284,2011. International Journal of Cell Biology 17

[168] T. R. Webb, A. S. Joyner, and P.M. Potter, “The development and [184] F. Zammarchi, E. De Stanchina, E. Bournazou et al., “Antitu- application of small molecule modulators of SF3b as therapeutic morigenic potential of STAT3 alternative splicing modulation,” agents for cancer,” Drug Discovery Today,vol.18,no.1-2,pp.43– Proceedings of the National Academy of Sciences of the United 49, 2013. States of America,vol.108,no.43,pp.17779–17784,2011. [169] S. Bonnal, L. Vigevani, and J. Valcarcel,´ “The spliceosome as a [185] S. Vorlova,´ G. Rocco, C. V. LeFave et al., “Induction of antago- target of novel antitumour drugs,” Nature Reviews, vol. 11, no. 11, nistic soluble decoy receptor tyrosine kinases by intronic poly pp. 847–859, 2012. A activation,” Molecular Cell,vol.43,no.6,pp.927–939,2011. [170] A. Corrionero, B. Minana,˜ and J. Valcarcel,´ “Reduced fidelity of [186] J. A. Bauman and R. Kole, “Modulation of RNA splicing as a branch point recognition and alternative splicing induced by potential treatment for cancer,” Bioengineered Bugs,vol.2,no.3, the anti-tumor drug spliceostatin A,” Genes and Development, pp.125–128,2011. vol. 25, no. 5, pp. 445–459, 2011. [187]L.A.Skordis,M.G.Dunckley,B.Yue,I.C.Eperon,and [171]K.O’Brien,A.J.Matlin,A.M.Lowell,andM.J.Moore,“The F. Muntoni, “Bifunctional antisense oligonucleotides provide biflavonoid isoginkgetin is a general inhibitor of pre-mRNA a trans-acting splicing enhancer that stimulates SMN2 gene splicing,” Journal of Biological Chemistry,vol.283,no.48,pp. expression in patient fibroblasts,” Proceedings of the National 33147–33154, 2008. Academy of Sciences of the United States of America,vol.100,no. 7, pp. 4114–4119, 2003. [172]M.Muraki,B.Ohkawara,T.Hosoyaetal.,“Manipulationof alternative splicing by a newly developed inhibitor of Clks,” [188] S. M. Elbashir, J. Harborth, W. Lendeckel, A. Yalcin, K. Weber, JournalofBiologicalChemistry,vol.279,no.23,pp.24246– and T. Tuschl, “Duplexes of 21-nucleotide RNAs mediate RNA 24254, 2004. interference in cultured mammalian cells,” Nature,vol.411,no. 6836, pp. 494–498, 2001. [173] N. Allende-Vega, S. Dayal, U. Agarwala, A. Sparks, J.-C. Bourdon, and M. K. Saville, “p53 is activated in response to [189] O.Camacho-Vanegas,G.Narla,M.S.Teixeiraetal.,“Functional disruption of the pre-mRNA splicing machinery,” Oncogene, inactivation of the KLF6 tumor suppressor gene by loss of het- vol.32,pp.1–14,2012. erozygosity and increased alternative splicing in glioblastoma,” International Journal of Cancer,vol.121,no.6,pp.1390–1395, [174] E. M. Amin, S. Oltean, J. Hua et al., “WT1 mutants reveal 2007. SRPK1 to be a downstream angiogenesis target by altering [190] A. DiFeo, L. Feld, E. Rodriguez et al., “A functional role for VEGF splicing,” Cancer Cell, vol. 20, no. 6, pp. 768–780, 2011. KLF6-SV1 in lung adenocarcinoma prognosis and chemother- [175] J. Soret, N. Bakkour, S. Maire et al., “Selective modifica- apy response,” Cancer Research,vol.68,no.4,pp.965–970,2008. tion of alternative splicing by indole derivatives that target [191] A. Massiello and C. E. Chalfant, “SRp30a (ASF/SF2) regulates serine-arginine-rich protein splicing factors,” Proceedings of the the alternative splicing of caspase-9 pre-mRNA and is required National Academy of Sciences of the United States of America, for ceramide-responsiveness,” Journal of Lipid Research,vol.47, vol. 102, no. 24, pp. 8764–8769, 2005. no.5,pp.892–897,2006. [176] C. Ghigna, M. De Toledo, S. Bonomi et al., “Pro-metastatic [192]R.W.Goehe,J.C.Shultz,C.Murudkaretal.,“hnRNPL splicing of Ron proto-oncogene mRNA can be reversed: ther- regulates the tumorigenic capacity of lung cancer xenografts in apeutic potential of bifunctional oligonucleotides and indole mice via caspase-9 pre-mRNA processing,” Journal of Clinical derivatives,” RNA Biology,vol.7,no.4,pp.495–503,2010. Investigation,vol.120,no.11,pp.3923–3939,2010. [177] F. Lai, C. C. Jiang, M. L. Farrelly, X. D. Zhang, and P. Hersey, [193]G.R.Devi,T.M.Beer,C.L.Corless,V.Arora,D.L.Weller,and “Evidence for upregulation of Bim and the splicing factor SRp55 P. L. Iversen, “In vivo bioavailability and pharmacokinetics of in melanoma cells from patients treated with selective BRAF a c-MYC antisense phosphorodiamidate morpholino oligomer, inhibitors,” Melanoma Research,vol.22,no.3,pp.244–251,2012. AVI-4126, in solid tumors,” Clinical Cancer Research, vol. 11, no. [178]C.C.Jiang,F.Lai,K.H.Tayetal.,“Apoptosisofhuman 10, pp. 3930–3938, 2005. melanoma cells induced by inhibition of B-RAFV600E involves [194] D. C. Altieri, “Survivin, cancer networks and pathway-directed preferential splicing of bimS,” Cell Death and Disease,vol.1,no. drug discovery,” Nature Reviews Cancer,vol.8,no.1,pp.61–70, 9, article e69, 2010. 2008. [179]F.MuntoniandM.J.A.Wood,“TargetingRNAtotreat [195] D. C. Altieri, “Validating survivin as a cancer therapeutic target,” neuromuscular disease,” Nature Reviews Drug Discovery,vol.10, Nature Reviews Cancer,vol.3,no.1,pp.46–54,2003. no. 8, pp. 621–637, 2011. [196]R.A.Carrasco,N.B.Stamm,E.Marcusson,G.Sandusky,P. [180] M.A.Havens,D.M.Duelli,andM.L.Hastings,“TargetingRNA Iversen, and B. K. R. Patel, “Antisense inhibition of survivin splicing for disease therapy,” Wiley Interdisciplinary Reviews, expression as a cancer therapeutic,” Molecular Cancer Therapeu- vol. 4, no. 3, pp. 247–266, 2013. tics,vol.10,no.2,pp.221–232,2011. [181]J.A.Bauman,S.-D.Li,A.Yang,L.Huang,andR.Kole,“Anti- [197] D. C. Talbot, M. Ranson, J. Davies et al., “Tumor survivin is tumor activity of splice-switching oligonucleotides,” Nucleic downregulated by the antisense oligonucleotide LY2181308: a Acids Research, vol. 38, no. 22, pp. 8348–8356, 2010. proof-of-concept, first-in-human dose study,” Clinical Cancer [182] R.Liu,C.Page,D.R.Beidler,M.S.Wicha,andG.Nu´nez,˜ “Over- Research, vol. 16, no. 24, pp. 6150–6158, 2010. expression of Bcl-X(L) promotes chemotherapy resistance of mammary tumors in a syngeneic mouse model,” American Journal of Pathology,vol.155,no.6,pp.1861–1867,1999. [183] G. Xu, C. Zhang, and J. Zhang, “Dominant negative STAT3 suppresses the growth and invasion capability of human lung cancer cells,” Molecular Medicine Reports,vol.2,no.5,pp.819– 824, 2009. Hindawi Publishing Corporation International Journal of Cell Biology Volume 2013, Article ID 846510, 10 pages http://dx.doi.org/10.1155/2013/846510

Review Article U1 snRNP-Dependent Suppression of Polyadenylation: Physiological Role and Therapeutic Opportunities in Cancer

Lee Spraggon1 and Luca Cartegni1,2

1 Molecular Pharmacology and Chemistry Program, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA 2 Department of Chemical Biology, Ernest Mario School of Pharmacy, Rutgers, The State University of New Jersey, Piscataway, NJ 08854, USA

Correspondence should be addressed to Luca Cartegni; [email protected]

Received 5 July 2013; Accepted 5 September 2013

Academic Editor: Claudia Ghigna

Copyright © 2013 L. Spraggon and L. Cartegni. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Pre-mRNA splicing and polyadenylation are critical steps in the maturation of eukaryotic mRNA. U1 snRNP is an essential component of the splicing machinery and participates in splice-site selection and spliceosome assembly by base-pairing to the 󸀠 󸀠 5 splice site. U1 snRNP also plays an additional, nonsplicing global function in 3 end mRNA processing; it actively suppresses the polyadenylation machinery from using early, mostly intronic polyadenylation signals which would lead to aberrant, truncated mRNAs. Thus, U1 snRNP safeguards pre-mRNA transcripts against premature polyadenylation and contributes to the regulation 󸀠 of alternative polyadenylation. Here, we review the role of U1 snRNP in 3 end mRNA processing, outline the evidence that led to the recognition of its physiological, general role in inhibiting polyadenylation, and finally highlight the possibility of manipulating this U1 snRNP function for therapeutic purposes in cancer.

1. Introduction AS and APA are deregulated and exploited by cancer cells to promote their growth and survival [5–7]. The generation of translationally competent messenger RNAs This review will focus on the recently described splicing- (mRNAs)isacomplexmolecularprocessthatinvolvesdis- independent functions of U1 small ribonucleoprotein particle tinctive enzymatic reactions and dedicated cellular machiner- (snRNP) in pre-mRNA processing, with emphasis on its ies that result in the splicing, capping, editing, and polyadeny- role in the regulation of APA site selection and in the lation of a pre-mRNA transcript. During this process, the suppression of intronic polyadenylation (IPA). Furthermore, choice and usage of splice sites (alternative splicing, AS) we will address innovative approaches to leverage U1 snRNP and of polyadenylation signals (alternative polyadenylation, functions as therapeutic avenues in cancer treatment. APA) within a common pre-mRNA can be differentially regulated depending on the developmental state, tissue, and 2. U1 snRNP Canonical Role in Splicing: cell type or in response to a variety of physiological stimuli or A Harbinger of Spliceosome Assembly pathological conditions [1, 2]. Collectively, alternative splicing and polyadenylation are key molecular mechanisms for The established function of U1 snRNP, which includes the increasing the functional diversity of the human proteome, 164 nt U1 snRNA, seven Sm proteins, and three U1-specific allowingtherelativelysmallhumangenome(<25,000 genes) proteins (U1-70K, U1-A, and U1-C), is its role in the early steps to generate an excess of 100,000 different protein isoforms3 [ ]. of pre-mRNA splicing as a key component of the spliceosome, However, because of the pervasiveness and essential role of the large ribonucleoprotein complex responsible for the AS in all physiological processes, aberrant RNA processing is removal of intronic sequences and subsequent rejoining of also frequently associated with many diseases [4], and both exons, to form a mature mRNA [8]. 2 International Journal of Cell Biology

󸀠 The spliceosome assembles through the sequential bind- within the same 3 UTRs (tandem PAS). The choice between ing of the five snRNPs (U1, U2, U4, U5, and U6) and multiple tandem PAS can directly affect the levels of gene expression. auxiliary RNA-binding proteins to form the large, active For example, it can lead to the exclusion in the shorter spliceosome [8]. U1 snRNP plays an essential role in this UTRs of miRNA-binding sites that affect translation and/or process by driving the initial steps of spliceosome assembly of other posttranscriptional regulatory sequences embedded 󸀠 onto the pre-mRNA at the exon-intron boundary, through in the 3 UTR that might influence stability or localiza- 󸀠 definition of the 5 splice site (ss) by RNA-RNA base-pairing tion. 󸀠 with the 5 end of U1 snRNA (Figure 1(a)), which can occur in Recent transcriptome-wide studies of APA have also 󸀠 multiple registers [9]. The 3 ss and the branch point (BP) are highlighted the common occurrence of IPA, with active recognized by the U2 complex to form the prespliceosomal intronic PAS that are used in up to 20% of human genes, either complex, and the tri-snRNP complex—containing U4, U5, by read-through into the intron or by usage of an alternative 󸀠 and U6—is then recruited, with U6 replacing U1 at the 5 terminal exon within introns [17, 18]. Whereas tandem APA ss. Following the release of U4 and further remodeling in only affects expression levels, usage of these IPA sites leads to the activated spliceosome, two subsequent transesterification shortened mRNAs containing truncated open reading frames reactions occur, to produce the final spliced exons and the (ORFs) and hence expressing qualitatively different protein released lariat [8]. Most splicing events, as well as polyadeny- isoforms, with unique C-terminal domains. The function of lation, occur cotranscriptionally, and the integration of the these isoforms can be vastly different to their full-length three processes, which mutually affect each other, is in counterparts and often encode for variants with dominant- large part mediated by the C-terminal domain (CTD) of negative properties [7, 19]. 󸀠 the elongating RNA Polymerase II (RNAPII) [2]. Multiple Both APA phenomena—IPA and 3 UTR variability—add splicing and cleavage/polyadenylation factors, including U1 an extra layer of complexity to the genome, by increasing the snRNP, are directly associated with the RNAPII CTD from potential number of transcripts encoded by a single gene and the onset of transcription and are then deposited on their controlling their expression. The mechanisms governing APA cognate-binding sites along the pre-mRNA, determining have only recently begun to be understood. Variations in the splice-site and poly(A) signal (PAS) selection. levels of the core polyadenylation factors or abundance of a Variants to this main pathway exist and in some cases broad range of transacting RNA-binding proteins have been U1 snRNP might not be strictly required [8, 9], whereas in showntomodulatePASselection(reviewedin[7]). In addi- other situations multiple U1 snRNPs can bind to competing tion, emerging evidence suggests that APA is also influenced 󸀠 alternative 5 splice sites [9]. However, the mechanics of by chromatin organization and epigenetic modifications at the canonical splicing reaction mandate that all snRNPs the DNA level [7, 20]. participate in 1 : 1 stoichiometric ratios in the actual removal Specific and global modulation of APA are important of each intron. Hence, it has been a long-standing puzzling aspects of many physiological processes, and their deregu- observation that the cellular levels of U1 snRNP exceed those lation can contribute to the etiology of numerous diseases, of other snRNPs by 2-3-fold [10, 11], suggesting that U1 may including cancer. APA is associated with increased cellular carry out additional roles besides its canonical one in splicing. proliferation and potentially with oncogenic transforma- 󸀠 󸀠 Indeed, its involvement in different aspects of 3 pre-mRNA tion, with the switch toward isoforms with a shorter 3 end formation has been long proposed [12, 13]. UTR in proliferating/cancer cells [21–24]. This shift typically removes negative posttranscriptional regulatory motifs (such 󸀠 as miRNA-binding sites), resulting in higher expression 3. Alternative 3 End Processing: levels, for example, of oncogenes. 󸀠 Physiological Regulation and In addition to global 3 UTR shortening, activation of Deregulation in Cancer specific IPA sites can also be modulated in cancer cells, with the generation of truncated variants possessing oncogenic 󸀠 RNAPII mRNAs in eukaryotic cells undergo 3 end process- properties or the suppression in tumors of antitumorigenic ing, which typically involves the cotranscriptional endonu- variants. For example, the usage of an IPA site in cyclin D1, cleolytic cleavage of the pre-mRNA, followed by the addition which normally controls progression through the cell cycle, of a polyadenylate (poly(A)) tail (reviewed in [7, 14]) by results in the transforming, constitutively active truncated a multisubunit complex that includes CPSF (cleavage and cyclin D1b isoform [25]. Conversely, a soluble extracellular polyadenylation specificity factor), CstF (cleavage stimula- variantofVEGFR2,whichactsasanaturalinhibitorof tion factor), CFI and CFII (cleavage factors I and II), and VEGF signaling in lymphangiogenesis [26], is significantly poly(A)polymerase(PAP).CPSFrecognizesthecanonical downregulated in neuroblastoma [27]. hexanucleotide PAS located upstream of the cleavage site, Yet, despite a greater appreciation of APA-mediated whilst CstF binds to a less well-defined downstream U/GU- events in cancer, the underlying mechanisms regulating this rich region. Following cleavage, PAP promotes polyadenyla- process remain relatively obscure. One simple possibility tion of the mRNA, adding the ∼250 nt poly(A) tail [14]. is that the increased proliferative rate and transcriptional Global analysis indicates that approximately 50% of status of cancer cells might be associated with the general human genes use different PAS, a process known as alter- upregulation of the cleavage and polyadenylation machinery. native polyadenylation (APA) [15, 16]. The study of APA This would lead to a more effective recognition of all PAS, hasbeentypicallyfocusedtowardsthemultiplePASlocated which would then by default result in usage of the more International Journal of Cell Biology 3

U1 snRNP U1A + Splicing

snRNA + U1-70K − 󳰀 3 − U1C Sm 󳰀 GUCCAΨΨCAUA-5 ...CAGGUAAGU ... 󳰀 ISS ISE ESE ESS 5 splice site (a)

PAP-binding PAP-binding pocket pocket CstF Basic-residue motif IgM U1A U1A PAP BPV 70K PAP G/U 󳰀 U1A-binding sites Secretory PAS 5 ss PAS (b) (c)

U1 MS2 loop MS2-70K

PAP Luciferase PAP EXO U1 PAS binding site U1 adaptor PAS (d) (e)

󸀠 󸀠 Figure 1: Modes of U1 snRNP activity in splicing and suppression of 3 end processing. (a) Role of U1 snRNP in splicing. The 5 end of U1 󸀠 snRNA base-pairs to the 5 splice site cotranscriptionally, to define the functional splice donor site. The process is positively and negatively modulated by splicing factors binding to exonic and intronic splicing enhancer and silencers (ESE, ISE, ESS, and ISS, resp.). (b) Role of U1A protein in suppression of IgM cleavage and polyadenylation. U1 snRNP component U1A binds to multiple motifs (blue boxes) upstream or downstream of the intronic PAS (purple box) in IgM C𝜇4 = M1 intron. These sites contain the A(U/G)GCN1−3C consensus motif and are similar to U1A-binding site on U1 snRNA. From the upstream sites, U1A multimers interact directly with the C-terminal domain of PAP through a binding pocket formed by the basic-residue motifs (green triangles), blocking polyadenylation. When U1A binds downstream, it prevents CstF binding to the G/U element (light purple box) and thus interferes with cleavage. (c) Role of U1 snRNP in suppression of BPV 󸀠 polyadenylation. U1 binds to a 5 ss a few nucleotides upstream of the alternative PAS. Like U1A multimers, U1-70K interacts directly with the C-terminal domain of PAP through a binding pocket formed by the basic-residue motifs, blocking polyadenylation. (d) Tethered U1-70 K suppresses polyadenylation. A modified U1-MS2 snRNA is engineered to contain an MS2-binding loop in place of the U1-70 K binding loop. A U1-binding site represses polyadenylation and expression from a luciferase reporter vector when the wt U1 snRNA is expressed. The mutant U1-MS2 snRNA recapitulates PAP inhibition only when a fusion MS2-70 K protein is also coexpressed. (e) U1 adaptors recruit U1 snRNP to suppress polyadenylation. Single-stranded, bifunctional modified ASOs are used to recruit endogenous U1 to target sites within ∼1 Kb region upstream of a PAS. The targeting moiety (red) base-pairs to a unique, transcript-specific sequence, while the recruiting moiety contains a 󸀠 consensus 5 ss sequence to bind U1 snRNP with high affinity. The tethered U1 snRNP suppresses polyadenylation and the mRNA is then 󸀠 󸀠 degraded by the 3 -5 exosome. Large blue boxes represent exons/coding regions.

proximal sites because of the directionality of transcription. 4. U1 snRNP and Its Role in the Suppression 󸀠 However, as mentioned, multiple transacting RNA-binding of 3 End Processing proteins such as CPEB1 [28], hnRNP H [29], and PTB [30, 31] 󸀠 can also affect APA, and their activity might be differentially A putative role for U1 snRNP in mRNA 3 end processing affected in cancer cells (reviewed in7 [ , 18]). While future hadbeenfirstrecognizednearly30yearsagowhenitwas work will be required to delineate the contribution of each of observed that antibodies directed against U1 snRNP were these processes to APA regulation, an intriguing possibility capable of interfering with in vitro polyadenylation reactions is that the modulation of the levels of U1 snRNP would [32–34]. differentially impact APA in cancer cells and thus contribute These initial observations were followed by reports that to UTR shortening (see below). the human U1 snRNP-associated protein U1A was capable 4 International Journal of Cell Biology of inhibiting polyadenylation of its own pre-mRNA by this mechanism can be harnessed as an antisense tool for binding two highly conserved U1-snRNA-like motifs in close gene-silencing, with relevant therapeutic perspectives, which proximity to the U1A mRNA PAS, via direct inhibition of will be discussed below. poly(A) polymerase activity (PAP) [13, 35, 36]. U1A has also been implicated in the regulation of an APA switch from membrane-bound to secretory forms of immunoglobulin M 5. U1 snRNP-Dependent IPA Suppression: (IgM) heavy chain mRNAs [37]. The transition to secretory A Safeguard of Transcript Integrity and an B cells is associated with the usage of an upstream IPA NMD Companion site, resulting in the truncation of the IgM pre-mRNA, with loss of the membrane-anchor encoding terminal exons. Two Overall, the studies illustrated above highlighted a clear separate mechanisms appear to be involved in the regulation splicing-independent role for U1 snRNP in suppression of 󸀠 of the secretory IPA site (Figure 1(b)). On one hand, U1A 3 end processing in the context of PAS contained within a 󸀠 binds to multiple U1-snRNA-like A(U/G)GCN1−3C-binding canonical 3 UTR. sites upstream of the PAS and suppresses polyadenylation However, PAS contain limited amount of information, as initiation, by direct inhibitory interactions with PAP, much well as sequences that could potentially work as PASs, but like the self-regulation of U1A own pre-mRNA [37]. However, are never or seldom used, are very abundant in pre-mRNAs, U1A also binds to similar tandem U1-snRNA-like motifs in particular within introns [15, 17]. Usage of these putative downstream of the secretory IPA site, positioned between two PAS and would lead to significantly shortened mRNAs either GU-rich domains normally utilized by the cleavage compo- unable to express functional proteins or expressing truncated nent CstF64. This impedes CstF64 binding and thus prevents variants, with potentially deleterious effects. It stands to cleavage of the pre-mRNA at the secretory IPA site [38]. reason that a suppressive mechanism must have evolved to In both cases, it is the free, non-snRNP-bound U1A that minimize aberrant widespread IPA. suppresses PAS usage by either inhibiting formation of the In its canonical splicing function, U1 snRNP is located 󸀠 cleavage complex from a downstream position, or by directly at the 5 ss of each intron, from where it could also inhibit repressing PAP activity from the upstream sites. As such, downstream activation of IPA sites, much in the way by which the physiologically decreasing levels of free U1A during B- tethered U1 snRNP blocks polyadenylation within a proper 󸀠 cell differentiation release the suppression of the IPA site, 3 UTR (Figure 2(a)). Two recent studies directly tested this allowing expression of the secreted IgM ([39]). Importantly, hypothesis and proposed an expanded role for U1 snRNP in 󸀠 free U1A needs to be present at least in dimer form in order 3 end processing inhibition, which would put U1 at the center for its single basic-residue motif to form a binding pocket that of a novel RNA surveillance mechanism that safeguards the can interact with and inhibit PAP,which implies that snRNP- integrity of pre-mRNAs from improper usage of premature bound U1A would not be inhibitory. PAS and that modulates the usage of legitimate IPA sites 󸀠 However, previous work on the 3 end processing of [19, 48]. bovine and human papillomavirus (BPV and HPV) tran- Functional knockdown of U1 activity can be achieved 󸀠 scripts had shown that U1 snRNP itself was capable of with decoy RNA oligonucleotides directed against the 5 󸀠 󸀠 suppressing 3 end processing by binding to a 5 ss and end of the U1 snRNA, which in high concentration lead inhibiting polyadenylation [40]. In this case, though direct to the block of the splicing reaction [51]. When the effects PAP inhibition was mediated by a separate U1 snRNP com- of functional U1 depletion using high concentration of ponent, U1-70 K (Figure 1(c)), through a domain containing morpholino antisense compounds (ASOs) were assayed on related basic-residue motifs, suggesting a similar mechanism a genomic tiling array [48], the generation of shorter stable is involved [41]. Unlike U1A, U1-70 K contains multiple pre-mRNA transcripts was observed in addition to splicing basic-residue motifs and thus is self-sufficient in repressing inhibition (Figure 2(c)). In these transcripts, cleavage and PAP. Like in the case of free U1A, U1 snRNP can also polyadenylation had occurred prematurely, typically within in some contexts (e.g., in HIV RNA regulation) inhibit the first intron. Importantly, they were not activated when cleavage from a downstream position [42, 43]. Overall, these splicing was pharmacologically inhibited, or when U2 snRNP key observations implied that U1snRNP itself might play activity was abrogated [48]. In a parallel study [19], similar a more general inhibitory role in polyadenylation, via U1- decoy RNAs were used to impair U1 snRNP functionality in 70 K. Indeed, natural or mutated U1 snRNP targeted at conditions where splicing was preserved (Figure 2(b)). Under 󸀠 the 3 UTR of reporter [44] or endogenous genes [45] these conditions, a general activation of natural alternative show strong inhibitory effects on polyadenylation, leading IPA sites occurred and was recapitulated by siRNA knock- 󸀠 to robust downregulation of target genes. Tethering the down of U1-70 K [19]. When ASOs targeted at specific 5 ss PAP-interacting domain of U1-70 K (Figure 1(d))issuffi- upstream of the natural alternative IPA were used to prevent cient to mediate inhibition [46], whereas nearby tethering U1 snRNP binding there (rather than targeting U1 snRNA of SR proteins (which would engage U1-70 K) interferes and preventing U1 from binding anywhere), only that specific with the inhibition, supporting the proposed model. More IPAsitewasactivated[19](Figure 2(d)). Furthermore, inhibi- 󸀠 recently, the powerful effect of tethering U1 snRNP to block tion of the same splicing event by blocking the downstream 3 polyadenylation was also demonstrated in trans (Figure 1(e)), ss with a separate ASO does not activate IPA, while blocking by the use of adaptor bifunctional modified oligonucleotides the PAS directly with another ASO impedes activation of the 󸀠 [47]. Importantly, this strategy introduced the concept that truncated transcript [19], demonstrating that both the 5 ss International Journal of Cell Biology 5

Normal conditions

DNA Promoter

U1 RNA AAAAA AAAAA PA S U1 sites AAAA APA APA ATG Regulated IPA STOP

RNA expression (a)

Limited/regulated AAAA IPA activation AAAA APA APA ATG IPA STOP

U1-binding ASO (b)

Global IPA activation AAAAA AAAA ATG

(c)

Specific IPA activation AAAAA AAAA ATG STOPIPA

Transcript-specific ASO (d)

Figure 2: U1 snRNP suppression of cleavage and polyadenylation safeguards transcriptome integrity. (a) In normal conditions, transcription starts bidirectionally from a RNAPII promoter. Relative enrichment of PAS and depletion of U1 sites leads to short nonproductive mRNAs on the antisense strand, thus enforcing promoter directionality (described in Almada et al. 2013 [49]). In the sense strand, the presence of U1-binding motifs at splice sites and along introns recruits U1 snRNP and suppresses IPA, allowing full-length expression. Regulated IPA sites can still be expressed depending on cellular context. (b) Reduction of U1 snRNP activity to levels still compatible with efficient splicing by partial functional knockdown using U1-binding ASOs (described in Vorlovaetal.[´ 19]) or following changes in endogenous levels [50], leads to selective APA. PASs in stronger context (or suppressed by weaker U1 sites) are activated first. These likely include regulated IPA sites and tandem APA sites, a situation that reflects what may occur in proliferative/cancer cells, with the appearance of shorter average mRNAs in part due to relative shortage of available U1 snRNP. (c) Complete disruption of U1 activity by sequestering ASOs leads to loss of splicing and release of global IPA activation. Because of transcriptional directionality, earlier IPA sites get used first, resulting in massive shortening 󸀠 of mRNAs (described in Kaida et al. [48]). (d) When ASOs targeted to a specific 5 ss are used, U1 binding is disrupted in that particular location but still functions normally elsewhere. The result is the selective activation of the targeted IPA site (highlighted), with expression of a truncated variant (described in Vorlovaetal.[´ 19]). Expression of other genes is not disrupted. U1 snRNP is indicated, as well as U1-targeting ASOs (red) and transcript-specific ASOs, (blue). Large blue and light blue boxes indicate exons and UTRs, respectively, while lines depict introns. PAS are depicted as purple ovals, U1 sites as red bars. Blue arrows indicate mRNA species generated in a specific context. 6 International Journal of Cell Biology and the intronic PAS are essential. Together, these two studies underestimated, as they would not be predicted based on 󸀠 show that U1 snRNP ability to inhibit 3 -end processing is not DNA sequencing data, nor would they be detected by routine 󸀠 limited to the 3 UTR but it extends to the entire pre-mRNA PCR-based analysis. transcript, indicating a broad surveillance role for U1. The range of U1’s inhibitory effect appears to be limited to ∼1Kb[50], suggesting that the positioning of U1 at its natural 6. Modulation of U1 snRNP Functions as 󸀠 5 position would not be sufficient to silence IPA events along a Potential Cancer Therapy 󸀠 large introns. Rather, its binding to the abundant pseudo 5 The role of U1 snRNP in ensuring the integrity of press that “litter” introns [52] might be functionally important 󸀠 mRNA through suppression of polyadenylation can also be to extend the protection from improper 3 end processing harnessed in novel antisense-based strategies to control gene to more distal areas (Figure 2(a)). Indeed, evidence of such expression. The use of next-generation antisense oligonu- binding of U1 to pseudosites was recently provided, as func- cleotides is emerging as effective approaches to treat many tional down-titration of cellular U1 snRNP levels with ASOs 󸀠 󸀠 conditions, including genetic diseases and cancer (reviewed resulted in the directional 3 →5release of suppression of in [55]) and can be similarly used to modulate U1 snRNP APA, whereas the opposite was observed following U1 snRNA function to develop therapeutic approaches to target such overexpression [50]. diseases. The role of U1 snRNP as a safeguard of transcript integrity is also reflected in its recently described essential role in ensuring promoter directionality [49]. In fact, while RNAPII 7. U1 Adaptors to Knockdown Tumorigenic transcription is inherently bidirectional, transcripts in the Pre-mRNA Transcripts antisense direction are shortly terminated by the presence of 󸀠 immediate PAS. On the contrary, the asymmetric enrichment The unique features of U1 snRNP in suppression of3 of U1 snRNP-binding sites in the sense strand ensures that end processing described above constitute the foundation early PAS are suppressed and transcription is allowed to of a novel powerful gene-silencing technology, U1 small proceed productively. nuclear interference (U1i). This technology uses bifunctional In summary, U1 snRNP, in addition to its role in AS oligonucleotide “U1 adaptors” to recruit U1 to the UTR regulation, appears to play a nonsplicing key role in ensuring of endogenous pre-mRNAs (Figure 1(e))tosuppresstheir 󸀠 󸀠 proper gene expression, by acting as an essential safeguard polyadenylation and target them for degradation by the 3 -5 of the integrity of all transcripts. Moreover, it also com- exosome [47]. 󸀠 plements the mRNA surveillance functions carried out by 2-O-Methyl (2 OMe) or locked nucleic acids (LNA) 󸀠 the conserved nonsense-mediated decay (NMD) machinery, oligonucleotides are designed to contain a 5 targeting moiety, 󸀠 which monitors and tags for degradation mRNAs harboring specific to the 3 UTR of the target transcript, coupled to 󸀠 󸀠 premature termination codons (PTC) within their ORF and a3 U1-binding moiety, which is similar to a 5 ss and that potentially encode for deleterious truncated proteins base-pairs to the U1 snRNA [47]. Hybridization of the U1 (reviewed in [53]). Typically, PTCs are generated by direct adaptor oligonucleotide to the target transcript results in nonsense mutations or by frameshifts due to insertions, recruitment of U1 snRNP, which in turn inhibits nuclear deletions, and aberrant splicing. In mammalian cells, NMD poly(A) polymerase activity through U1-70K, leading to RNA recognizes PTCs by their position upstream of the last degradation. exon-exon junction during translation and targets them for Recently, in an important proof-of-concept study, the degradation. However, this leaves a significant gap in the U1 adaptor approach was shown to be highly effective in mRNA surveillance process, as equally deleterious mRNA suppression of melanoma growth in xenografts models, by products could be generated from activation of cryptic targeting metabotropic glutamate receptor 1 (GRM1) and B- intronic PAS. Such aberrant mRNAs would be effectively cell lymphoma 2 (BCL2) transcripts [56]. Aberrant expres- immune to NMD and evade degradation, since an activated sion of GRM1, a transmembrane domain G protein-coupled intronic PAS would generate a novel “terminal exon”, with receptor that mediates glutamate signaling, plays a crucial likely a new STOP codon located beyond the “new last” role in the onset of melanoma in mouse models [57], and exon-exon boundary (if no in-frame new termination codons dysregulated glutamatergic signaling leads to transformation were present, the mRNA would be degraded by the nonstop and tumorigenesis in multiple cancer types [57, 58]. The mRNAdecaypathway,reviewedin[54]). The U1 snRNP- antiapoptotic gene BCL2 is a frequent player in a variety of mediated RNA surveillance function described above, would cancers, providing an escape mechanism to avoid the effects therefore, also protect from potential damage derived from of apoptosis-inducing chemotherapeutic compounds [59]. mutations/lesions that introduce de novo PAS in introns. U1 adaptors were delivered to tumors by a dendrimer RGD Finally, these observations have additional far-reaching delivery system which binds with high affinity to an integrin implications for the interpretation of disease-associated variant overexpressed on the surface of many solid tumors 󸀠 mutations. For example, mutations within 5 ss, typically pre- [60]. Systemic delivery of the U1 adaptors targeting BCL2 and dicted to affect splicing and/or induce exon skipping, could GRM1 suppresses tumor growth in melanoma xenografts by in fact activate downstream intronic poly (A) sites, with the up to 60–70% [56]. consequent generation of stable and potentially pathogenic In a separate study, U1 adaptors were targeted to truncated variants. Importantly, such events would likely be PIM1 kinase (proviral integration site for Moloney murine International Journal of Cell Biology 7 leukemia virus 1), a constitutively active serine/threonine IPA variants in receptor tyrosine kinase genes, to induce the kinase that regulates a diverse array of cellular responses, expression of secreted decoy RTK (sdRTK) isoforms that including apoptosis and cell signal transduction [61]. PIM1 antagonize RTK signaling [19]. overexpression in multiple tumor types is linked to poor Deregulation and constitutive activation of RTK signaling prognosis. In vitro inhibition of PIM-1 by siRNAs was previ- represents a key aspect of tumorigenesis in a broad range of ously shown to mediate antitumorigenic effects in colorectal human cancers [66, 67]. Targeting these oncogenic pathways andprostatecarcinomacells[62]. U1i adaptors targeted to has provided the basis for targeted therapies, with the devel- PIM1 effectively and specifically silence PIM1 in GBM cell opment of effective tyrosine kinase inhibitors (TKI) or anti- lines, with antiproliferative and proapoptotic effects61 [ ]. bodies directed at the extracellular domains (ECD) [68]. An Similarly, PIM1-U1i adaptors, encapsulated within nanoparti- alternative approach to inhibit oncogenic RTK signaling has cles and injected intratumorally into glioblastoma xenografts, been the delivery of recombinant sdRTKs variants, composed reduced tumor growth by 40–50%, compared to controls. of the ligand-binding ECD [69]. In this context, signaling Off-target and nonspecific effects, observed in other is blocked via ligand sequestration and/or the engagement antisense approaches such as RNAi, could also represent an of endogenous full-length RTKs in nonproductive dimers. issuewithU1itechnology.Inparticular,apossiblepitfallof ThisisthebasisforhowAflibercept,aVEGFtrap,functions U1i is its potential to nonspecifically sequester endogenous to effectively suppress neovascularization and tumor growth U1 snRNP, an occurrence that can in fact induce global [70]. changes in splicing and processing of nontargeted transcripts UsageofIPAsitesupstreamoftheexonsencodingfor [63]. However, this should become evident only at very high the transmembrane domain of RTKs, to generate sdRTK, concentration of U1i adaptors, whereas the high-potency was recently shown to commonly occur in most RTK compounds developed by Goraczniak and colleagues are mRNAs [19]. Treatment with ASOs coupled to a dendrimer 󸀠 used at concentrations that should not significantly affect delivery moiety, to target the 5 ss upstream of the IPA global U1 snRNP functions. site, resulted in its activation and expression of specific Collectively, these studies demonstrate that U1i technol- endogenous sdRTKs isoforms for multiple RTKs, including 󸀠 ogy, by exploiting the ability of U1 snRNP to inhibit 3 EGFR, MET, HER2, VEGFR1,andVEGFR2 [19]. VEGFR2 is polyadenylation and hence targeting a transcript for degra- the major mediator of VEGF signaling and a central player 󸀠 󸀠 dation by the 3 –5 exosome, provides a viable approach for in tumor vascularization [71]. As such, anti-VEGF treatment effective knockdown of tumorigenic transcripts in vivo. is a cornerstone of cancer therapy, with drugs targeting both theVEGFligand(bevacizumabandaflibercept)aswellas its receptor (sunitinib and axitinib). As mentioned, soluble 8. Activation of Antitumorigenic Isoforms by VEGFR2 is a powerful natural inhibitor of angiogenesis [26] Release of U1 snRNP-Mediated Suppression and is underrepresented in tumors [27]. Its induction by of Intronic PAS activation of a PAS in intron 13 of the VEGFR2/KDR pre- mRNA (Figure 3)resultedinthegenerationofasoluble As described above, treatment of cells with decoy ASOs protein isoform that potently inhibited angiogenesis in a 󸀠 mimicking 5 ss inhibits U1 snRNP functions globally and paracrine and autocrine fashion [19]andalsoshowedactivity nonspecifically, and leads to global activation of normally in vivo [72]. suppressed PAS [19, 48]. However, ASOs that instead com- Given the central role of aberrant RTK signaling in 󸀠 pete with endogenous U1 snRNP for specific 5 ss within cancer and the existence of sdRTK variants for most RTKs, target transcripts can be used to induce the activation of their induction by the specific, U1 snRNP-competing, ASO- specific natural IPA events [19]. Often, the truncated proteins mediated activation of IPA has a tremendous potential as a generated by the activation of the APA event lack impor- broad therapeutic approach in cancer therapy. tant functional C-terminal regions and possess dominant- negative properties. In absence of an actionable downstream IPA site, splicing 9. Conclusion and Future Directions redirection ASOs would interfere with the splicing reaction, typically leading to exon skipping [55]. Antisense modulation Our understanding of the role of U1 snRNP in pre-mRNA of splicing events with this class of compounds is emerging processing has gradually expanded from its initial splicing as a viable therapeutic approach to induce more desirable functions in splice-site selection and exon definition to its splicing variants in genetic diseases such as Duchenne moonlighter act in the suppression of polyadenylation in a musculardystrophyorspinalmuscularatrophyandhas very limited, gene-specific manner to the currently proposed reached the clinical trial stage (reviewed in [64]). In cancer, role as a key player in RNA surveillance and global safeguard a leading approach has been to induce antagonist/dominant of mRNA integrity (and possibly long noncoding RNAs) negative variants of oncogenic proteins (reviewed in [55]), against spurious cleavage and polyadenylation, with a job for example, in the case of signal transducer and activator description that has come to include also the control of of transcription 3 (STAT3), where redirection of its splic- promoter directionality in transcription. ing to the antitumorigenic Stat3 beta variant leads to full Much still remains to be uncovered about the specific tumorregressioninbreastcancermousemodels[65]. A mechanisms underlying how U1 snRNP manages to effec- similar approach was adapted to activate antitumorigenic tively wear its many hats. However, the compendium of 8 International Journal of Cell Biology

Fl VEGFR2 VEGF binding, ECD KD TM dimerization, and activation U1 Sustained VEGF signaling and angiogenesis

VEGFR2/KDR ASO AAAA AAAA 1 12 13 14 15 30 STOP STOP Ligand trapping, nonproductive dimerization

Dominant negative ECD inhibition of VEGF signaling sVEGFR2 and angiogenesis

Figure 3: Therapeutic potential of IPA activation: induction of secreted decoy VEGFR2. An IPA site in intron 13 of VEGFR2 can be specifically 󸀠 and effectively activated using ASOs targeted to the 5 ss immediately upstream, preventing U1 from binding and thus releasing suppression [19]. This leads to the expression of a variant secreted decoy VEGFR2 encoding the sole ECD. This variant can still bind VEGF or other ligands and can still dimerize with VEGF receptors, but it cannot signal. On the contrary, it leads to a blockade of VEGF signaling in targeted and surrounding cells, with dominant negative properties. evidence outlined above points to a basic model where U1 Overall, the possibility to harness U1 snRNP-mediated snRNP associates with the CTD of RNAPII and is then suppression of polyadenylation has created an attractive and recruited cotranscriptionally to the nascent transcript at still mostly unexplored opportunity to reshape the transcrip- multiple sites throughout the length of the pre-mRNA. These tome for therapeutic purposes, in cancer and other diseases. 󸀠 correspond to either genuine or pseudo 5 ss and collectively U1i technology provides an effective alternative approach result in the suppression of cleavage and polyadenylation to siRNA in vivo,andatthesametimetheactivationof along the pre-mRNA, enforcing its full-length expression sdRTKs by IPA derepression could serve as a blueprint (Figure 2(a)). for the induction of therapeutically relevant, endogenous, Genuine PAS utilization is achieved by the evolutionary potent dominant-negative IPA variants of oncogenes. A depletion of U1-binding sites in the UTR and/or by the better molecular understanding of the role of U1 snRNP in presence of dominant cis-acting regulatory elements that APA is an essential step in the design of rationally targeted potentiate PAS or inhibit U1 activity through specific factors. antisense strategies as effective therapies. For example, any factor that promotes recruitment of U1 snRNP to a splice site would strengthen IPA inhibition and vice versa. Therefore, in certain contexts, like in the case of Acknowledgments VEGFR2 and other RTKs, the system has evolved past the The authors thank Elisa de Stanchina, Pedro Miura, and Peter default suppressive status, in order to allow specific APA Smibert for their helpful comments on the paper. This work events to occur. was supported by awards BC112622 by the Department of In general, proliferative/cancer cells demonstrate a global 󸀠 Defense, 1R21CA173303-01 by National Institute of Health, shortening of 3 UTRs, typically resulting in enhanced gene andI5-A561bytheSTARRFoundation. expression. U1 snRNP might be indirectly contributing to this phenomenon if the increase in transcription—and the related increase in associated machinery, including cleavage and References polyadenylation factors—is not paralleled by an equivalent increase in U1 snRNP levels. The relative depletion of U1 [1] A. Kalsotra and T. A. Cooper, “Functional consequences of snRNP would release PAS and could in part explain the developmentally regulated alternative splicing,” Nature Reviews observed directional effect, as shown in activated neurons Genetics,vol.12,no.10,pp.715–729,2011. [50]. This selective switch from distal to proximal PAS can [2] A. R. Kornblihtt, I. E. Schor, M. Allo, G. Dujardin, E. Petrillo be countered in vitro via overexpression of U1 snRNA [50], et al., “Alternative splicing: a pivotal step between eukaryotic anditis,thus,possibletoenvisageanapproachtopromote transcription and translation,” Nature Reviews Molecular Cell 󸀠 3 UTR re-lengthening by ectopic expression of U1 snRNA Biology,vol.14,pp.153–165,2013. 󸀠 by gene therapy technologies. Alternatively, since 3 end [3]T.W.NilsenandB.R.Graveley,“Expansionoftheeukaryotic processing can be specifically blocked by ASO directed at proteome by alternative splicing,” Nature,vol.463,no.7280,pp. the PAS [19], transcript-specific ASO could be employed to 457–463, 2010. selectively block the usage of proximal PAS while reactivating [4] L. Cartegni, S. L. Chew, and A. R. Krainer, “Listening to distal ones. silence and understanding nonsense: exonic mutations that International Journal of Cell Biology 9

affect splicing,” Nature Reviews Genetics,vol.3,no.4,pp.285– [23] C. Mayr and D. P.Bartel, “Widespread shortening of 3’ UTRs by 298, 2002. alternative cleavage and polyadenylation activates oncogenes in [5] R. K. Singh and T. A. Cooper, “Pre-mRNA splicing in disease cancer cells,” Cell,vol.138,no.4,pp.673–684,2009. and therapeutics,” Trends in Molecular Medicine,vol.18,no.8, [24]R.Sandberg,J.R.Neilson,A.Sarma,P.A.Sharp,andC.B. pp.472–482,2012. Burge, “Proliferating cells express mRNAs with shortened 3’ [6] C. Ghigna, C. Valacca, and G. Biamonti, “Alternative splicing untranslated regions and fewer microRNA target sites,” Science, and tumor progression,” Current Genomics,vol.9,no.8,pp. vol.320,no.5883,pp.1643–1647,2008. 556–570, 2008. [25] K. E. Knudsen, “The cyclin D1b splice variant: an old oncogene [7]R.Elkon,A.P.Ugalde,andR.Agami,“Alternativecleavage learns new tricks,” Cell Division,vol.1,article15,2006. and polyadenylation: extent, regulation and function,” Nature [26] R. J. C. Albuquerque, T. Hayashi, W.G. Cho et al., “Alternatively Reviews Genetics,vol.14,pp.496–506,2013. spliced vascular endothelial growth factor receptor-2 is an [8] M. C. Wahl, C. L. Will, and R. Luhrmann,¨ “The spliceosome: essential endogenous inhibitor of lymphatic vessel growth,” design principles of a dynamic RNP machine,” Cell,vol.136,no. Nature Medicine,vol.15,no.9,pp.1023–1030,2009. 4, pp. 701–718, 2009. [27] J. Becker, H. Pavlakovic, F. Ludewig et al., “Neuroblastoma [9] X.Roca,A.R.Krainer,andI.C.Eperon,“Pickone,butbequick: progression correlates with downregulation of the lymphangio- 5’ splice sites and the problems of too many choices,” Genes & genesis inhibitor sVEGFR-2,” Clinical Cancer Research,vol.16, Development,vol.27,pp.129–144,2013. no.5,pp.1431–1441,2010. [10] S. J. Beserga and J. A. Steitz, “The diverse world of small ribonucleoproteins,” in The RNA World. Cold Spring Harbor,R. [28] F. A. Bava, C. Eliscovich, P. G. Ferreira, B. Minana, C. Ben-Dov F. Gesteland and J. F. Atkins, Eds., pp. 359–381, CSHL Press, et al., “CPEB1 coordinates alternative 3’-UTR formation with 1993. translational regulation,” Nature,vol.495,no.7439,pp.121–125, 2013. [11] R. Reddy and H. Busch, “RNA sequences, structure, and modifications,” in Structure and Function of Major and Minor [29] S. Danckwardt, I. Kaufmann, M. Gentzel et al., “Splicing factors Small Nuclear Ribonucleoprotein Particles,M.Birnstiel,Ed.,pp. stimulate polyadenylation via USEs at non-canonical 3’ end 1–37, Springer, New York, NY, USA, 1988. formation signals,” EMBO Journal,vol.26,no.11,pp.2658–2669, [12] K. M. Wassarman and J. A. Steitz, “Association with terminal 2007. exonsinpre-mRNAs:anewrolefortheU1snRNP?”Genes and [30] A. Moreira, Y. Takagaki, S. Brackenridge, M. Wollerton, J. L. Development,vol.7,no.4,pp.647–659,1993. Manley, and N. J. Proudfoot, “The upstream sequence element [13]W.C.Boelens,E.J.R.Jansen,W.J.VanVenrooij,R.Stripecke,I. of the C2 complement poly(A) signal activates mRNA 3’ end W.Mattal, and S. I. Gunderson, “The human U1 snRNP-specific formation by two distinct mechanisms,” Genes and Develop- U1A protein inhibits polyadenylation of its own pre-mRNA,” ment,vol.12,no.16,pp.2522–2534,1998. Cell,vol.72,no.6,pp.881–892,1993. [31] P. Castelo-Branco, A. Furger, M. Wollerton, C. Smith, A. Mor- [14] N. J. Proudfoot, “Ending the message: poly(A) signals then and eira, and N. Proudfoot, “Erratum: polypyrimidine tract binding now,” Genes and Development,vol.25,no.17,pp.1770–1782,2011. protein modulates efficiency of polyadenylation,” Molecular and [15]B.Tian,J.Hu,H.Zhang,andC.S.Lutz,“Alarge-scaleanalysis Cellular Biology, vol. 24, no. 15, pp. 4174–4183, 2004. of mRNA polyadenylation of human and mouse genes,” Nucleic [32] C. L. Moore and P. A. Sharp, “Site-specific polyadenylation in a Acids Research,vol.33,no.1,pp.201–212,2005. cell-free reaction,” Cell,vol.36,no.3,pp.581–591,1984. [16] F. Ozsolak, P. Kapranov, S. Foissac et al., “Comprehensive [33] C. L. Moore and P.A. Sharp, “Accurate cleavage and polyadeny- polyadenylation site maps in yeast and human reveal pervasive lation of exogenous RNA substrate,” Cell,vol.41,no.3,pp.845– alternative polyadenylation,” Cell,vol.143,no.6,pp.1018–1029, 855, 1985. 2010. [34] C. Hashimoto and J. A. Steitz, “A small nuclear ribonucleo- [17] B. Tian, Z. Pan, and Y. L. Ju, “Widespread mRNA polyadeny- protein associates with the AAUAAA polyadenylation signal in lation events in introns indicate dynamic interplay between vitro,” Cell,vol.45,no.4,pp.581–591,1986. polyadenylation and splicing,” Genome Research,vol.17,no.2, pp. 156–165, 2007. [35] C. W. G. Van Gelder, S. I. Gunderson, E. J. R. Jansen et al., [18] B. Tian and J. L. Manley, “Alternative cleavage and polyadeny- “A complex secondary structure in U1A pre-mRNA that binds lation: the long and short of it,” Trends in Biochemical Sciences, two molecules of U1A protein is required for regulation of vol. 38, no. 6, pp. 312–320, 2013. polyadenylation,” EMBO Journal,vol.12,no.13,pp.5191–5200, 1993. [19] S. Vorlova,´ G. Rocco, C. V. LeFave et al., “Induction of antagonistic soluble decoy receptor tyrosine kinases by intronic polya [36] S. I. Gunderson, K. Beyer, G. Martin, W. Keller, W. C. Boelens, activation,” Molecular Cell,vol.43,no.6,pp.927–939,2011. and I. W. Mattaj, “The human U1A snRNP protein regulates [20] N. Spies, C. B. Nielsen, R. A. Padgett, and C. B. Burge, “Biased polyadenylation via a direct interaction with poly(A) poly- chromatin signatures around polyadenylation sites and exons,” merase,” Cell,vol.76,no.3,pp.531–541,1994. Molecular Cell,vol.36,no.2,pp.245–254,2009. [37] C. Phillips, S. Jung, and S. I. Gunderson, “Regulation of nuclear [21] A. R. Morris, A. Bos, B. Diosdado, K. Rooijers, R. Elkon et poly(A) addition controls the expression of immunoglobulin al., “Alternative cleavage and polyadenylation during colorectal MsecretorymRNA,”EMBO Journal,vol.20,no.22,pp.6443– cancer development,” Clinical Cancer Research,vol.18,pp. 6452, 2001. 5256–5266, 2012. [38]C.Phillips,N.Pachikara,andS.I.Gunderson,“U1Ainhibits [22] Y. Fu, Y. Sun, Y. Li et al., “Differential genome-wide profiling of cleavage at the immunoglobulin M heavy-chain secretory tandem 3’ UTRs among human breast cancer and normal cells poly(A) site by binding between the two downstream GU-rich by high-throughput sequencing,” Genome Research,vol.21,no. regions,” Molecular and Cellular Biology,vol.24,no.14,pp. 5, pp. 741–747, 2011. 6162–6171, 2004. 10 International Journal of Cell Biology

[39] J. Ma, S. I. Gunderson, and C. Phillips, “Non-snRNP U1A levels [56] R. Goraczniak, B. A. Wall, M. A. Behlke, K. A. Lennox, E. decrease during mammalian B-cell differentiation and release S. Ho et al., “U1 adaptor oligonucleotides targeting BCL2 and the IgM secretory poly(A) site from repression,” RNA,vol.12, GRM1 suppress growth of human melanoma xenografts in no. 1, pp. 122–132, 2006. vivo,” Molecular Therapy,vol.2,p.e92,2013. [40] P. A. Furth, W.-T. Choe, J. H. Rex, J. C. Byrne, and C. C. [57] J. L. F. Teh and S. Chen, “Glutamatergic signaling in cellular Baker, “Sequences homologous to 5’ splice sites are required transformation,” Pigment Cell and Melanoma Research,vol.25, for the inhibitory activity of papillomavirus late 3’ untranslated pp. 331–342, 2012. regions,” Molecular and Cellular Biology,vol.14,no.8,pp.5278– [58] T. D. Prickett and Y. Samuels, “Molecular pathways: dysreg- 5289, 1994. ulated glutamatergic signaling pathways in cancer,” Clinical [41] S. I. Gunderson, M. Polycarpou-Schwarz, and I. W. Mattaj, Cancer Research,vol.18,pp.4240–4246,2012. “U1 snRNP inhibits pre-mRNA polyadenylation through a [59] R. J. Youle and A. Strasser, “The BCL-2 protein family: opposing direct interaction between U1 70K and poly(A) polymerase,” activities that mediate cell death,” Nature Reviews Molecular Cell Molecular Cell,vol.1,no.2,pp.255–264,1998. Biology,vol.9,no.1,pp.47–59,2008. [42] M. P. Ashe, A. Furger, and N. J. Proudfoot, “Stem-loop 1 of [60] E. Ruoslahti, “RGD and other recognition sequences for inte- the U1 snRNP plays a critical role in the suppression of HIV- grins,” Annual Review of Cell and Developmental Biology,vol. 1 polyadenylation,” RNA,vol.6,no.2,pp.170–177,2000. 12, pp. 697–715, 1996. [43] M. P. Ashe, L. H. Pearson, and N. J. Proudfoot, “The HIV-1 5’ [61] U. Weirauch, A. Grunweller,¨ L. Cuellar, R. K. Hartmann, LTRpoly(A)siteisinactivatedbyU1snRNPinteractionwith and A. Aigner, “U1 adaptors for the therapeutic knockdown the downstream major splice donor site,” EMBO Journal,vol.16, of the oncogene pim-1 kinase in glioblastoma,” Nucleic Acid no. 18, pp. 5752–5763, 1997. Therapeutics, vol. 23, no. 4, pp. 264–272, 2013. [44] S.A.Beckley,P.Liu,M.L.Stover,S.I.Gunderson,A.C.Lichtler, [62] T. Zhang, X. Zhang, K. Ding, K. Yang, Z. Zhang, and Y. and D. W. Rowe, “Reduction of target gene expression by a Xu, “PIM-1 gene RNA interference induces growth inhibition modified U1 snRNA,” Molecular and Cellular Biology,vol.21,no. and apoptosis of prostate cancer cells and suppresses tumor 8, pp. 2815–2825, 2001. progressioninvivo,”Journal of Surgical Oncology,vol.101,no. [45] P. Fortes, Y. Cuevas, F. Guan et al., “Inhibiting expression of 6, pp. 513–519, 2010. 󸀠 specific genes in mammalian cells with 5 end-mutated U1 [63] T. A. Vickers, M. Sabripour, and S. T. Crooke, “U1 adaptors small nuclear RNAs targeted to terminal exons of pre-MRNA,” result in reduction of multiple pre-mRNA species principally Proceedings of the National Academy of Sciences of the United by sequestering U1snRNP,” Nucleic Acids Research,vol.39,no. States of America,vol.100,no.14,pp.8264–8269,2003. 10,p.e71,2011. [46] X. Abad, M. Vera, S. P. Jung et al., “Requirements for gene [64] M.A.Havens,D.M.Duelli,andM.L.Hastings,“TargetingRNA silencing mediated by U1 snRNA binding to a target sequence,” splicing for disease therapy,” RNA,vol.4,no.3,pp.247–266, Nucleic Acids Research,vol.36,no.7,pp.2338–2352,2008. 2013. [47] R. Goraczniak, M. A. Behlke, and S. I. Gunderson, “Gene [65] F. Zammarchi, E. De Stanchina, E. Bournazou et al., “Antitu- silencingbysyntheticU1Adaptors,”Nature Biotechnology,vol. morigenic potential of STAT3 alternative splicing modulation,” 27,no.3,pp.257–263,2009. Proceedings of the National Academy of Sciences of the United [48] D. Kaida, M. G. Berg, I. Younis et al., “U1 snRNP protects pre- States of America,vol.108,no.43,pp.17779–17784,2011. mRNAs from premature cleavage and polyadenylation,” Nature, [66] A. Gschwind, O. M. Fischer, and A. Ullrich, “The discovery vol. 468, no. 7324, pp. 664–668, 2010. of receptor tyrosine kinases: targets for cancer therapy,” Nature [49]A.E.Almada,X.Wu,A.J.Kriz,C.B.Burge,andP.A. Reviews Cancer,vol.4,no.5,pp.361–370,2004. Sharp, “Promoter directionality is controlled by U1 snRNP and [67] M. A. Lemmon and J. Schlessinger, “Cell signaling by receptor polyadenylation signals,” Nature,vol.2013,2013. tyrosine kinases,” Cell,vol.141,no.7,pp.1117–1134,2010. [50]M.G.Berg,L.N.Singh,I.Younis,Q.Liu,A.M.Pintoetal., [68] Y. Yarden and G. Pines, “The ERBB network: at last, cancer “U1 snRNP determines mRNA length and regulates isoform therapy meets systems biology,” Nature Reviews Cancer,vol.12, expression,” Cell,vol.150,pp.53–64,2012. pp. 553–563, 2012. [51] X. Roca and A. R. Krainer, “Recognition of atypical 5’ splice [69] Z. Tiran, A. Oren, C. Hermesh et al., “A novel recombinant sites by shifted base-pairing to U1 snRNA,” Nature Structural soluble splice variant of met is a potent antagonist of the and Molecular Biology,vol.16,no.2,pp.176–182,2009. hepatocyte growth factor/scatter factor-met pathway,” Clinical [52] X. H.-F. Zhang, K. A. Heller, I. Hefter, C. S. Leslie, and L. Cancer Research, vol. 14, no. 14, pp. 4612–4621, 2008. A. Chasin, “Sequence information for the splicing of human [70] K. K. Ciombor, J. Berlin, and E. Chan, “Aflibercept,” Clinical pre-mRNA identified by support vector machine classification,” Cancer Research,vol.19,pp.1920–1925,2013. Genome Research, vol. 13, no. 12, pp. 2637–2650, 2003. [71] L. Claesson-Welsh and M. Welsh, “VEGFA and tumour angio- [53] F. Lejeune and L. E. Maquat, “Mechanistic links between genesis,” Journal of Internal Medicine,vol.273,no.2,pp.114–127, nonsense-mediated mRNA decay and pre-mRNA splicing in 2013. mammalian cells,” Current Opinion in Cell Biology,vol.17,no. [72]L.A.Owen,H.Uehara,J.Cahoon,W.Huang,J.Simonis,and 3, pp. 309–315, 2005. B. K. Ambati, “Morpholino-mediated increase in soluble flt-1 [54] M. Graille and B. Seraphin, “Surveillance pathways rescu- expression results in decreased ocular and tumor neovascular- ing eukaryotic ribosomes lost in translation,” Nature Reviews ization,” PLoS ONE,vol.7,no.3,ArticleIDe33576,2012. Molecular Cell Biology,vol.13,pp.727–735,2012. [55] L. Spraggon and L. Cartegni, “Antisense modulation of RNA processing as a therapeutic approach in cancer therapy,” Drug Discovery Today,2013. Hindawi Publishing Corporation International Journal of Cell Biology Volume 2013, Article ID 810572, 16 pages http://dx.doi.org/10.1155/2013/810572

Review Article Role of Pseudoexons and Pseudointrons in Human Cancer

Maurizio Romano,1 Emanuele Buratti,2 and Diana Baralle3,4

1 Department of Life Sciences, University of Trieste, Via A. Valerio 28, 34127 Trieste, Italy 2 International Centre for Genetic Engineering and Biotechnology (ICGEB), 34149 Trieste, Italy 3 Human Development and Health, University of Southampton, Duthie Building, Mailpoint 808, Southampton General Hospital, TremonaRoad,SouthamptonSO166YD,UK 4 Human Genetics Division, University of Southampton, Duthie Building, Mailpoint 808, Southampton General Hospital, Tremona Road, Southampton SO16 6YD, UK

Correspondence should be addressed to Diana Baralle; [email protected]

Received 17 June 2013; Accepted 9 August 2013

Academic Editor: Claudia Ghigna

Copyright © 2013 Maurizio Romano et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

In all eukaryotic organisms, pre-mRNA splicing and alternative splicing processes play an essential role in regulating the flow of information required to drive complex developmental and metabolic pathways. As a result, eukaryotic cells have developed a very efficient macromolecular machinery, called the spliceosome, to correctly recognize the pre-mRNA sequences that need tobe insertedinamaturemRNA(exons)fromthosethatshouldberemoved(introns).Inhealthyindividuals,alternativeandconstitutive splicing processes function with a high degree of precision and fidelity in order to ensure the correct working of this machinery. In recent years, however, medical research has shown that alterations at the splicing level play an increasingly important role in many human hereditary diseases, neurodegenerative processes, and especially in cancer origin and progression. In this minireview, we will focus on several genes whose association with cancer has been well established in previous studies, such as ATM, BRCA1/A2, and NF1. In particular, our objective will be to provide an overview of the known mechanisms underlying activation/repression of pseudoexons and pseudointrons; the possible utilization of these events as biomarkers of tumor staging/grading; and finally, the treatment options for reversing pathologic splicing events.

1. Introduction the mature transcript. Therefore, after the basic elements that define introns and exons and are composed by donor and Starting from the first description of alternative splicing and acceptor splice sites plus the branch point sequence, there constitutive splicing processes in 1977 [1–3], the importance was the discovery of enhancer and silencer elements that of this process that guarantees the correct flow of information from transcription to translation in eukaryotic cells has can affect either positively or negatively the way these basic continued to grow exponentially. elements are recognized by the spliceosome [7–9]. Usually, In particular, one major branch of research in this area enhancer and silencer elements are bound by members of the has the aim to investigate and characterize the cellular macro- SR and hnRNP protein families, respectively. molecular machine (i.e., the spliceosome) that is physically The recruitment of these proteins to specific sites is crucial responsible for the cutting and joining of intronsexons by for their activity as depending on their position with respect catalyzing two transesterification reactions4 [ , 5]andthe to the basic elements their roles can be antagonistic or, in mechanisms that ensure its fidelity6 [ ]. As a result, research in some case, agonistic [10, 11]. Global analyses of the ways these spliceosome composition and functioning has been comple- factors cooperate and influence each other in shaping the mented by studies aiming to understand the sequences and splicing process has only recently begun to shed light on how molecules that determine under which conditions a partic- they promote or hinder exon recognition in a co-ordinated ular exon or intron is selectively recognized and included in manner [12–14]. 2 International Journal of Cell Biology

Rather unexpectedly, these studies have shown that SR revealed many new splicing alterations that were not previ- proteins interact not only with alternatively spliced exons but ously described. In particular, for example, it was reported also with constitutively spliced exons, and that their major that 423 primary transcripts (derived from 377 genes) were role consists in recruiting the splicing machinery to splice differentially spliced in triple-negative breast cancer samples sites [15, 16]. (generating 496 novel isoforms) [40]. By analyzing non- On the other hand, hnRNP proteins bind to nascent pre- triple-negative breast cancer samples, the same study found mRNA and influence splicing decision through a complex that 270 and 460 primary transcripts (derived from 242 and and finely tuned network of protein/protein interactions. The 387 differentially spliced genes, resp.) generated 331 and 550 mechanisms of hnRNPs actions on splicing is generally less novel isoforms that were not present in normal breast tissue defined than those of SR proteins [17]. [40]. In addition to these elements, we now also know that splicing choices are affected by a myriad of other factors, ranging from chromatin modifications [18, 19], transcrip- 2. Aberrant Splicing Events in Cancer Genes tional factors [20], RNA secondary structure [21], short noncoding RNAs [22, 23], and various cellular stresses [24]. It has become clear that in several human pathologies, In parallel to these mechanistic studies, another branch including cancer, the alternative splicing profile is aberrantly of splicing research has also addressed the functional impor- modified in a specific manner and in ways that can favor the tance of alternative splicing processes in biological pathways growth and survival of cancer cells [41–44]. The generation of and in particular the way that alternative splicing isoforms of these aberrant splicing profiles can occur in many ways, such proteins can acquire different or even antagonistic biological as through the re-expression of developmentally regulated properties [25, 26]. Because of this ability to expand the isoforms that had previously been shut off following early proteome of cells, alternative splicing has represented a developmental stages [45, 46], by affecting the splicing very useful and powerful tool that allows cells to execute profiles of genes that are implicated in tumor progression the various expression programs which underlie many fun- [47, 48], or through other mechanisms that generally have damental needs of higher organisms: from general needs anti-apoptotic/metastatic consequences and that can affect such as controlling normal development and tissue-specific response to therapies [49, 50]. expression of proteins, to highly specialized processes such In addition, several studies have defined that cancer- as DNA damage response or microRNA biogenesis [27–33]. associated splicing alterations arise not only from mutations Moreover, the rearrangements that a pre-mRNA undergoes in the cancer-related genes but derive also from variations in during the splicing process are also advantageous in terms of the expression and/or activity of splicing regulatory factors providing a longer half-life and better translational capacity, [51]. For example, it has been well established that the levels something that has recently begun to be exploited by the of SR- and hnRNP-proteins undergo changes associated with biotechnology industry [34]. transformation and progression of cancers [44, 52]. Strik- Considering all these necessities and advantages, it is ingly, it has been recently reported that some members of the therefore not surprising that the number of genes that are SR protein family of splicing factors and other components subjecttoalternativesplicingineukaryoticorganismshas of the spliceosomal cellular machinery can actually act as been steadily growing. Indeed, it was recently estimated that oncoproteins and play a direct role in promoting tumor origin more than 90% of mammalian protein coding genes can and progression [53–55] and that external stimuli such as produce at least one splicing-derived isoform expressed at hypoxia can cause the aberrant redistribution of important potentially significant biological levels35 [ ]. splicing factors such as Tra2 and promote the expression of The complexity of this system, however, also puts spliceo- tumor-promoting splicing isoforms [56]. some functioning at risk of being impaired by the occur- The case of Bcl-X (BCL2L1) gene, belonging to the rence of single-point mutations in splicing regulatory ele- BclII family and implicated in the control of mitochondrial ments, deletions/insertions, genomic rearrangements, and breakdown during apoptosis, is emblematic of this concept. alterations at the splicing factor expression level [36]. Any For example, two splicing Bcl-X isoforms can arise from 󸀠 of these alterations can result in a variety of aberrant splic- the use of two alternative 5 splice sites within exon 2 ing outcomes that usually include aberrant exon skipping, and lead to the synthesis of a short apoptosis-promoting cryptic splice site selection, intron retention, and pseudoexon protein (Bcl-XS) and to a long antiapoptotic form (Bcl-XL) activation [37, 38]. As expected, many of these changes can [57]. Different splicing factors, including Sam68, hnRNPA1, lead directly to the occurrence of disease in humans, and it SF2/ASF, hnRNP F/H, hnRNP K, SAP155, and SRp30c have has now been estimated that a sizable proportion of all gene beenfoundtobeinvolvedintheselectionofthetwo 󸀠 mutations leading to disease can be directly connected with competing alternative 5 splice sites that give rise to these two thepresenceofasplicingdefect[7, 39]. isoforms [58–61]. In addition, recent studies have shown that Importantly, the introduction of novel technologies that the elongation and splicing-related factor TCERG1 can bind allow fast profiling of the transcriptome seems promising for to the Bcl-X pre-mRNA and promote the proapoptotic Bcl- 󸀠 simplifying investigation of which genomic variability and XS 5 splice site in a promoter-dependent manner [62]. plasticity events allow cancer cells to tailor specific functional Finally, the production of the proapoptotic Bcl-XS splice units from the available exons of a gene. For example, recent variant seems to be improved by the core (Y14 and eIF4A3) RNA sequencing of the breast cancer transcriptome has and auxiliary (RNPS1, Acinus, and SAP18) components of International Journal of Cell Biology 3 the exon junction complex (EJC) [63], suggesting that EJC- similar to normal genes but nonfunctional, although some associated components can regulate apoptosis at the alterna- can still be transcribed). In this case, inspection of 14775 tive splicing level and represent a further level of vulnerability pseudogenes returned a list of alternative splicing hits from of cancer. which it is not easy to distinguish real events in expressed Therefore, one of the most interesting research areas pseudogenes (Table 1). in this field consists in the identification of cancer-specific More recently, some bioinformatic studies have tried to splice variants or the aberrant expression of splicing-affecting extrapolate the presence of pseudoexons in cancer tissues by proteinsthatcouldleadtotheirgeneration.Examplesof developing new analysis methods of GeneChip gene expres- both these events, in fact, have already been shown to sion array data [74, 75]. In this way, by comparing normal occur in some individual types of cancer, such as breast and cerebellum and medulloblastoma tumors, it was possible ovarian cancer [64, 65]. Another interesting research area to predict 811 significantly different expressed pseudoexons in aberrant splicing events connected with tumors is the (derived from 577 genes). In addition, when nonmetastatic occurrence of particular types of splicing defects that involve and metastatic medulloblastomas were compared, 13 pseu- the inclusion of “new” sequences (known as pseudoexons) doexonic sequences were significantly expressed in a differen- in the mature mRNA of cancer-related genes. Rather more tial manner (derived from 8319 strong candidates). However, rarely, the opposite has also been shown to occur: the aberrant it should be noted that no experimental validation was carried recognition of intronic sequences (pseudointrons) within out to support these predictions. Therefore, presently, manual normalexons.Inthisreview,wehavealsodecidedtoprovide annotations remain the most reliable system for identifying particular attention to these events as they are probably more real pseudoexons. common than previously considered and have not yet been Fortunately, the scientific community has recently iden- the subject of particular attention. tified a certain number of cases where pseudoexon inclusion Oneofthereasonswhythesetwoeventsareparticularly has been validated in detail. For this reason, Table 2 reports all interesting is that these types of defects are ideally suited for the cryptic exons described in the literature that are localized novel therapeutic effector molecules that are based in RNA within genes whose expression was altered in cancer. biology.Inthecaseofpseudoexonsandpseudointrons,in First of all, regarding the mechanisms underlying pseu- fact, the major advantage of targeting this type of inclusion doexon activation, it is interesting to note that these phe- events is that the antisense oligonucleotides would be targeted nomena are caused by inherited mutations resulting in the against normal intronic sequences and thus would not remain insertion of intronic sequences in the mature mRNA [38]. bound to the mature mRNA (possibly to interfere with later In this respect, it is interesting to note that among the genes stages of RNA processing such as export/translation). presenting pseudoexon “awakening” directly associated with cancer origin (Table 2), the creation of new splicing donor 3. Pseudoexon Activation in Cancer (12 events) and acceptor (7 events) sites represents the more frequent occurrence (48% and 28% of the listed events, resp.). In the pre-mRNA splicing field, the term “pseudoexon” Some of the most paradigmatic examples of cancer genes such has been introduced to describe exonic-like sequences that as BRCA1, BRCA2, NF1,andATM are included in these two are present within intronic regions but are ignored by the sets. spliceosomal machinery. A closer look at these sequences has On the other hand, among pseudoexons described in 󸀠 often provided a reason for their inability to be recognized cancer-related genes (Table 3),thecreationof5 splice sites 󸀠 as normal exons: the presence of intrinsic defects in their and of 3 splice sites comprise 55% and 15% of the listed apparently viable donor and acceptor sites [66]orofsilencer records, respectively. Moreover, pseudoexon insertion can elements [67–69] and the formation of inhibiting RNA also be triggered by the deletion of nearby donor or acceptors secondary structures [70–72]. splice sites (4 events in both lists), highlighting the impor- From a functional point of view, in most cases of pseudo- tance of the genetic milieu in splicing decisions. exon insertion, the presence of an extraneous exon within the Adeeperlookintothemechanismsunderlyingpseu- mature mRNA causes either the disruption of the transla- doexon activation has revealed the involvement of hnRNPs in tional reading frame or the insertion of novel amino acid regulation of these events. In particular, the partial inclusion sequences following translation. As a result, the normal of a pseudoexon of NF1 gene has been found to arise from biological properties of the resulting protein are very likely a novel intronic mutation c.31−279A>Gintron30,creatinga 󸀠 disrupted, and this can be associated with the development new acceptor splice site and activation of a cryptic 5 splice of disease. site [76]. It has been found that both PTB and nPTB play Unfortunately, it is still quite hard to identify reliable an active role in the pseudoexon splicing by repressing its pseudoexon insertion events in human genes implicated inclusion [76]. This finding is particularly interesting since it in cancers by just performing a general interrogation of further supports the hypothesis that PTB/nPTB might have a databases (matching 22719 ENSEMBL protein coding genes general role in repressing weak exons and in particular most versus 31057 entries of CanGEM Gene list) [73]. In fact, at of pseudoexons [77]. present, it is only possible to retrieve strong candidates for Another example of mutation causing pseudoexon acti- alternative splicing events in protein coding genes (Table 1). vation consists in the creation of novel branch site (Table 2, As expected, a similar situation was seen when pseudo- one event). This latter case is associated with a peculiar genes were investigated (defined as genomic DNA sequences mutation (IVS5 ds +232 G−A) that leads to pseudoexon 4 International Journal of Cell Biology

Table 1: Alternative splicing events detected in protein coding genes and pseudogenes implicated in cancer.

Alternative splicing event Protein coding genes CanGEM gene list (31057 entries) % cancer protein coding genes Cassette exon (CE) 13506 11771 87 Intron retention (IR) 8190 6775 83 󸀠 Alternative 3 ss (A3SS) 6701 5685 85 󸀠 Alternative 5 ss (A5SS) 6624 5640 85 Alternative first exons (AFE) 8050 7136 89 Alternative last exons (ALE) 3631 3233 89 Mutually exclusive exons (MXE) 3376 3006 89 Alternative splicing event Pseudogenes CanGEM gene LIST (31057 entries) % cancer pseudogenes Cassette exon (CE) 434 139 32 Intron retention (IR) 465 187 40 󸀠 Alternative 3 ss (A3SS) 241 91 38 󸀠 Alternative 5 ss (A5SS) 218 82 38 Alternative first exons (AFE) 91 27 30 Alternative last exons (ALE) 104 32 31 Mutually exclusive exons (MXE) 50 16 32 Homo sapiens genes (GRCh37.p10) dataset (62252 total genes) was used to filter alternative splicing events in protein coding genes (22719 entries)d an pseudogenes (14775 entries) and verified their presence in the Cancer GEnome Mine database (31057 entries—CanGEM, http://www.cangem.org/).

Table 2: Pseudoexon insertion events directly involved in cancer pathology.

Size (bp) Gene name Description Entrez gene Activating mutation Reference pseudoexon 󸀠 APC Adenomatosis polyposis coli 324 167 5 ss creation [78] 󸀠 ATM Ataxia telangiectasia mutated 470 58 3 ss creation [79] ATM 65 SRE deletion [80] 󸀠 ATM 137 5 ss creation [81] 󸀠 ATM 212 5 ss creation [82] Breakpoint cluster region 613 BCR-ABL V-ablAbelsonmurineleukemiaviraloncogene 42 Genomic rearrangement [83] 25 homolog 1 fusion BCR-ABL 35 Unknown [84, 85] 󸀠 BRCA1 Breast cancer 1, early onset 672 66 3 ss creation [86] 󸀠 BRCA2 Breast cancer 2, early onset 675 93 Downstream 3 ss deletion [87] 󸀠 BRCA2 95 5 ss creation [88] 󸀠 ESR1 Estrogen receptor 1 3467 69 5 ss creation [89] MutS homolog 2, colon cancer, nonpolyposis MSH2 4436 75 5’ss creation [90] type 1 (E. coli) Neurofibromin 1 (neurofibromatosis, von 󸀠 NF1 4763 67/99 3 ss creation [91] Recklinghausen disease, Watson disease) 󸀠 NF1 70 5 ss creation [92] 󸀠 NF1 107 5 ss creation [92] 󸀠 NF1 172 3 ss creation [93] 󸀠 NF1 58 3 ss creation [94] 󸀠 NF1 76 5 ss creation [94] 󸀠 NF1 54 5 ss creation [95] 󸀠 [92, 96, NF1 177 5 ss creation 97] NF2 Neurofibromin 2 (bilateral acoustic neuroma) 4771 106 BP creation98 [ ] 󸀠 RB1 Retinoblastoma 1 (including osteosarcoma) 5925 103 3 ss creation [99] 󸀠 TSC2 Tuberous sclerosis 2 7249 89 5 ss creation [100] 󸀠 WRN Werner syndrome 7486 106 5 ss creation [101] 󸀠 WRN 69 3 ss creation [102] International Journal of Cell Biology 5

Table 3: Pseudoexons described in cancer-related genes.

Entrez Size (bp) Gene name Description Activating mutation Reference gene pseudoexon ATP-binding cassette, sub-family C (CFTR MRP), 󸀠 ABCC8 6833 76 5 ss creation [103] member 8 󸀠 ALDH7A1 Aldehyde dehydrogenase 7 family, member A1 501 36 5 ss mutation [104] CD40 ligand (TNF superfamily, member 5, 󸀠 CD40LG 959 59 5 ss creation [105] hyper-IgM syndrome) 󸀠 CEP290 Centrosomal protein 290 kDa 80184 128 5 ss creation [106, 107] 󸀠 CHM Choroideremia (Rab escort protein 1) 1121 98 3 ss creation [108] 󸀠 COL4A3 Collagen, type IV, alpha 3 (Goodpasture antigen) 1285 74 3 ss creation [109] 󸀠 COL11A1 Collagen, type XI, alpha 1 1301 50 5 ss creation [110] CTD (carboxy-terminal domain, RNA 󸀠 CTDP1 polymerase II, polypeptide A) phosphatase, 9150 95 5 ss creation [111] subunit 1 Cytochrome b-245, beta polypeptide (chronic 󸀠 CYBB 1536 56 5 ss creation [112] granulomatous disease) 󸀠 CYBB 61 5 ss creation [113] Cytochrome P450, family 17, subfamily A, 󸀠 CYP17A1 1586 94 Upstream 5 ss mutation [114] polypeptide 1 󸀠 QDPR Quinoid dihydropteridine reductase 5860 152 5 ss creation [115] 󸀠 DPYD Dihydropyrimidine dehydrogenase 1806 44 5 ss creation [116] 󸀠 FBN1 Fibrillin 1 2200 93 5 ss creation [117] GHR Growth hormone receptor 2690 102 SRE deletion [118, 119] 󸀠 GUSB Glucuronidase, beta 2990 68 5 ss creation [120] 󸀠 HADH Hydroxyacyl-coenzyme A dehydrogenase 3033 141 5 ss creation [103] Hydroxyacyl-coenzyme A dehydrogenase 3-ketoacyl-coenzyme A thiolase 56 󸀠 HADHB 3032 5 ss creation [121] enoyl-coenzyme A hydratase (trifunctional 106 protein), beta subunit 󸀠 HSPG2 Heparan sulfate proteoglycan 2 3339 141 5 ss creation [103] SWI 󸀠 SMARCB1 SNF related, matrix associated, actin dependent 6598 72 5 ss creation [122] regulator of chromatin, subfamily b, member 1 86 󸀠 ISCU Iron-sulfur cluster scaffold homolog (E. coli)23479 3 ss creation [123–125] 100 SLC14A1 Solute carrier family 14 (urea transporter), 136 Internal 7 kb deletion [126] (JK) member 1 (Kidd blood group) MCCC2 Methylcrotonyl-coenzyme A carboxylase 2 (beta) 64087 64 SRE deletion [127] MFGE8 Milk fat globule-EGF factor 8 protein 4240 102 SRE creation [128] Megalencephalic leukoencephalopathy with 󸀠 MLC1 23209 246 5 ss creation [129] subcortical cysts 1 5-methyltetrahydrofolate-homocysteine MTRR 4552 140 SRE creation [130] methyltransferase reductase 󸀠 GPR143 G protein-coupled receptor 143 4935 165 3 ss creation [131] 󸀠 OAT Ornithine aminotransferase (gyrate atrophy) 4942 142 5 ss creation [132] 󸀠 OFD1 Oral-facial-digital syndrome 1 8481 62 5 ss creation [133] 󸀠 OTC Ornithine carbamoyltransferase 5009 135 3 ss creation [134] Propionyl-coenzyme A carboxylase, alpha PCCA 5095 84 SRE creation [135] polypeptide Propionyl-coenzyme A carboxylase, beta 󸀠 PCCB 5096 72 5 ss creation [135] polypeptide Phosphate regulating endopeptidase homolog, 50 󸀠 PHEX X-linked (hypophosphatemia, vitamin D resistant 5251 100 5 ss creation [136] rickets) 170 6 International Journal of Cell Biology

Table 3: Continued. Entrez Size (bp) Gene name Description Activating mutation Reference gene pseudoexon Polycystic kidney and hepatic disease 1 󸀠 PKHD1 5314 116 5 ss creation [137] (autosomal recessive) 󸀠 PMM2 Phosphomannomutase 2 5373 66 3 ss creation [138] 󸀠 PMM2 123 5 ss creation [138, 139] PRP31 pre-mRNA processing factor 31 homolog 󸀠 PRPF31 26121 175 5 ss creation [140] (S. cerevisiae) 󸀠 Upstream 3 ss deletion RHD Rh blood group, D antigen 6007 170 [141] and SNP in int7 󸀠 RYR1 Ryanodine receptor 1 (skeletal) 6261 119 5 ss creation [142] Solute carrier family 12 (sodium chloride 󸀠 SLC12A3 6559 238 5 ss creation [143] transporters), member 3 󸀠 USH2A Usher syndrome 2A (autosomal recessive, mild) 7399 152 5 ss creation [144] inclusion in NF2 gene by creating a consensus branch point. to efficient inclusion of the pseudoexon in the mature ATM In particular, it has been found that the G>A transition, mRNA [72]. occurring at position −18 from the acceptor site of NF2 IVS5, This example clearly shows the complexity of pseudoexon works in combination with existing consensus splice acceptor activation events that relies on a variety of factors well beyond 󸀠 󸀠 and donor sites and causes the formation of an extra exon 5a thesimpleinitialmutation(i.e.,5ss or 3 ss creation), and that in the NF2 gene that introduces a premature stop codon [98]. can include the eventual presence of enhancer and silencer Finally, in spite of being lowly represented, gross genomic elements within the cryptic exon or intro, RNA secondary rearrangements can bring together or expose splice site structures, and probably several other mechanisms that have sequences that would normally be very distant from each not yet been described in detail. other or absent form the original sequence (one event). Pseudoexon activation can also be caused by mutations that cause the creation/deletion of splicing regulatory ele- 4. Pseudointron Activation in Cancer ments (SREs, 6 events). Among these events, one of the most Pseudointrons (PSIs) represent an intriguing set of intron- representative examples of the complex regulatory networks like sequences localized within exons that can undergo that can underlie pseudoexon insertion is represented by the alternative splicing and that therefore can be included in or identification of an unusual splicing defect directly related excluded from the ORF within the final mRNA species. with neoplasia (Figure 1). Although the examples of pseudointrons reported in the In the case of this particular event, a patient affected review are not associated with activating mutations, we have by ataxia-telangiectasia was found to carry a 4 nt-deletion includedthembecauseofthepeculiarityofeventsinsofar (GTAA) within intron 20 of ATM gene [80]. This deletion, that they do not represent simple intron retention events, termed intron-splicing processing element (ISPE), caused the but rather alternatively spliced intraexonic sequences whose inclusion of a 65 nucleotide “cryptic exon” in the ATM mRNA behaviour can resemble that of introns under particular (Figure 1(a)). The mechanism through which this occurs has circumstances (e.g., occurrence of previous splicing events been well characterized in recent studies. 󸀠 in the processed transcript). It is this regulated use of the In normal conditions, the repression of the 3 ss in the intraexonic splice sites that fits well with the idea that these wild-type pseudoexon sequence (ATM WT) was shown to be intraexonic splice site represent “false” introns, even in the a direct consequence of U1snRNP binding in correspondence absence of genetic alterations. to the ISPE sequence. The importance of this U1snRNP 󸀠 binding was twofold: first of all, to sterically hinder 3 ss recog- In general, however, pseudointrons are less well described nition by U2snRNP and secondly, to also inhibit binding of compared withpseudoexons probably because of the diffi- the SRSF1 splice factor to the stem-loop sequence formed culty in detecting them using available technology (a situa- by this pseudoexon (Figure 1(b))[71, 72]. As a result, this tion that may well soon change for the better because of the internal U1snRNP binding event to the ISPE sequence causes introduction of more sensitive and comprehensive technical 󸀠 an unproductive U2snRNP association with the 3 ss that approaches such as RNA sequencing). At the moment, only results in the complete inhibition of pseudoexon inclusion three examples of PSI have been characterized, and for at least in normal conditions (Figure 1(b)). Following the deletion of two of these pseudointrons there are data supporting their the GUAA motif observed in the patient, U1snRNP binding relevance in the pathogenesis of cancer. to the ISPE is relieved, and SRSF1 is also free to bind to the internal enhancer site that becomes available (Figure 1(c)). 4.1. Fibronectin: IIICS-C5. Fibronectin (FN) is an extra- Taken together, these two events result in efficient recruit- cellular matrix protein whose functions range from cell ment of U2snRNP to the branchsite and a better recruitment adhesion and migration to wound healing and oncogenic 󸀠 of U1snRNP to the rather inefficient 5 gc splice site that lead transformation [145]. The functional complexity of FN is International Journal of Cell Biology 7

IVS20 Ex20 Ex21 1873 nt 2454 nt 581 nt

uuuucucuaauccucacagUUAUCUGGCCAGGUAAGUGAUAUAUCUUCACUCUACUGAUGAGGGUACGAAGGCCCUAGAUGACAUAAGgcaaguuuuu ISPE (a) AC U U C–G U–A AC C U U A–U C–G C–G U–A U–A WT ATM C ·· A–U U G C–G C–G U–A ISPE U··G ·· A–U U G ΔGUAA ATM AU–A C–G GAU U ·· G U C SRSF1 A–U G G U– A A A A C A A U G UG G A GACC–G A A SRSF1 U1snRNP G G G–G U ·· G G–CC G–C U U–A AG–C C–G C C U–A G–C A–U G ··U U··G U1snRNP U–A C–G U2snRNP U–A U–A G–C A–U a CA AUAAGC ·· U1snRNP 󳰀 󳰀 U2snRNP U G 5 ss U–A 3 ss G–C a CA AUAAGGC 󳰀 󳰀 3 ss 5 ss

No PE inclusion PE inclusion (b) (c)

Figure 1: ATM pseudoexon activated in cancer. (a) Scheme of the ATM pseudoexon insertion caused by a 4 nt deletion (GUAA) occurring 1873 nt downstream from the donor splice site of ATM exon 20. The solid lines indicate introns. Dotted lines include the 65 nt-ATM pseudoexon (gray box). The intron-splicing processing element (ISPE) complementary to U1 snRNA, the RNA component of the U1small nuclear ribonucleoprotein (snRNP), is underlined. (b) Schematic model of U1snRNP mediated inhibition of pseudoexon insertion in normal 󸀠 conditions. In this model, U1snRNP binding to the ISPE blocks pseudoexon inclusion by inhibiting recruitment of U2snRNP to the 3 splice site region and SRSF1 binding to the stem loop of the pseudoexon (shown as dotted lines). (c) In the case of the disease associated ATM ΔGUAA deletion, the internal U1snRNP binding site is no longer present due to the deletion of the ISPE. This leads to a more efficient 󸀠 binding of SRSF1 to the enhancer site and stabilization of the U2snRNP interaction with the branch site and U1snRNP with the 5 gc donor site. related to the structural diversity arising from cell type- interact with the integrin 𝛼4𝛽1 with different affinity149 [ , specific alternative splicing [146]. The fibronectin pre-mRNA 150]. Indeed, the CS1 site has approximately 20-fold higher undergoes alternative splicing primarily at three sites: two affinity for integrins than the CS5 site [151]. extra domain exons encoding extra structural repeats and Because of its intraexonic localization and splicing behav- a region of nonhomologous sequence called the type-III ior, the 93 bp segment can be considered a pseudointron connecting segment (IIICS). (Figure 2(a)). Different tissue-specific and disease-associated The IIICS domain is divided into three subdomains of changes in the variants of the IIICS region have been reported 75 bp, 192 bp, and 93 bp, respectively, that can be alternatively [152–154], and the CS5 pseudointron is upregulated in foetal spliced, thus generating up to fivevariant mRNAs in humans tissue and in adult liver [155]. [147, 148]. The subdomains of 75 bp and 93 bp encode for two For the purpose of this review, it is important to note cell-specific binding sites, CS1 (residues 1–25 of the IIICS), that alternative splicing of the IIICS-CS5 subdomain can and CS5 (residues 90–109 of the IIICS), respectively, that influence the onset and progression of cancers by affecting 8 International Journal of Cell Biology

THPO IIICS 6 7 6 −1 CS1 CS5 +1 75 192 93 THPO-1 −12nt IIICS-A (CS1+, CS5+) −1 5 +1 CS AchE THPO-2 192 93 −116 nt 23456 󳰀 IIICS-B (CS1−, CS5+) 4 3 −1 CS5 +1 THPO- T −197nt 75 192 AchE-S/T

󳰀 IIICS-C (CS1+, CS5−) 23454 −1 +1 THPO-4 R AchE-R −12nt −116 nt 192

IIICS-D (CS1−, CS5−) THPO-5 −1 +1 2345 +60 −116 󳰀 nt nt 4 AchE-E/H IIICS-E (CS1−, CS5−) H THPO-6 (a) (b) (c)

Figure 2: Pseudointrons whose alternative splicing can be altered in tumors. Schemes of pseudointron alternative splicing events shown to occur within the Fibronectin (IIICS, (a)), Acetylcholinesterase (AchE, (b)), and Thrombopoietin (THPO, (c)) genes. Pseudointrons are shown 󸀠 as white boxes in IIICS (CS5-93 nt), AchE (4 -R), and THPO (116 nt). Solid lines indicate introns and dotted lines distinguish the different splicing events. The C-terminal sequence of each splicing variant for each gene is also shown. For IIICS and THPO, additional exonic elements undergoing alternative splicing are shown as gray boxes. In the AchE gene, gray triangles indicate stop codon within different ORFs. See text for definition of each splice variant.

Fibronectin activities related to tissue organization and cell- the resulting protein will be prematurely terminated by a C- 󸀠 ECM interactions. In keeping with this hypothesis, in vitro terminus translated from the 5 region of I4. This variant is studies have shown that CS5 can modulate the spreading of expected to remain monomeric [161]. melanoma cells [151]. In general, the AChE-R transcript can be coexpressed with the AChE-T or AChE-H transcripts, and variations in their relative levels seem to be important for modulation 4.2. Acetylcholinesterase: AChE-R. Acetylcholinesterase is a of AChE functions, as well as for the pathogenesis of neu- serine specific hydrolase that hydrolyzes acetylcholine, and rological and autoimmune disorders [161, 162]. Intriguingly, its main function is related to the clearance of the neuro- recent studies have found that AChE-R mRNA accumulates transmitter from the synaptic cleft [156]. Nonetheless, the in primary human astrocytomas, and that its presence is observation that the expression of AChE is not limited to correlated with their grade of aggressiveness [163]. In human cholinergic tissues has suggested the presence of additional U87MG glioblastoma cells, it was found that AChE-R protein functions [157]. variant can support proliferation of glioblastoma tumors by forming a complex with the scaffold protein RACK1 and From the splicing process point of view, mammal AChE protein kinase Ce [164]. has been shown to undergo alternative splicing resulting In addition, increased levels of an N-terminally extended in expression of 3 isoforms, called R (“readthrough”), H N-AChE-R isoform were detected in human testicular (“hydrophobic”), and T/S (“tailed” or “synaptic”), that differ tumors indicating that the generation of the AChE-R variant in their subcellular localization, tissue distribution, and canbeassociatedwiththeutilizationofanalternative developmental pattern of expression [158]. Whereas, AChE- promoter in the transformed cells. H or AChE-T is the most abundant and common isoforms, Finally, related functional studies have suggested that the AChE-R expression is low and restricted in specific cell both AChE-R variants (AChE-R and N-AChE-R) might be lines, cell differentiation stage, or muscle development159 [ , crucial for increasing the cellular ATP levels and might 160]. supportselectivemetabolicadvantagesaswellasgenotoxic Interestingly, the AChE-R mRNA arises by retention of resistance by altering p73 gene expression [165]. Although it is pseudointron4, that leads to generation of the E1-E2-E3-E4- not yet clear whether the AChE-R (as well as the other AChE I4-E5 mRNA transcript (Figure 2(b)). Since pseudointron 4 splicing variants) can be related directly to tumorigenesis, includes a stop codon when it is included in mature RNA, the observed downregulation of all these splicing isoforms International Journal of Cell Biology 9 in colorectal carcinoma suggests a crucial role for tumor recognition by their cellular binding factors. Therefore, they development [166]. are ideally suited to block unwanted splice sites that may become activated following their creation or activation. It is interesting that the role of using antisense strategies to reg- 4.3. Thrombopoietin: THPO-3,THPO-5, and THPO-6. A ulate exon inclusion also occurs in physiological conditions. third example of pseudointron possibly implicated in cancer An example of this is snoRNA HBII-52 in the regulation of pathogenesis is found within the Thrombopoietin (THPO) exon Vb inclusion in the serotonin receptor 2C [182]. gene. THPO is the most important cytokine for regulation of Tothis moment, a few pilot studies aimed at inhibiting the platelet production, and its expression has also been reported inclusion of a pseudoexon in patient cell lines have already in skeletal muscle, ovary, testis, and central nervous system been reported. In particular, antisense oligonucleotides have (CNS) [167–169]. 󸀠 󸀠 been used to target newly created 5 or 3 splice sites The THPO gene shows a complex pattern of alternative deep within intronic regions of the NF1 gene to restore the splicing [170]; among the six splice variants that can be gener- normal splicing profile [91, 92]. Similar approaches have ated, three of them (THPO-3, THPO-5, and THPO-6) show also been successfully used to target deep intronic muta- the peculiar alternative splicing of 116 nucleotides within tions causing pseudoexon activation within BRCA2 intron exon 6 (that can therefore be classified as a pseudointron) 12 (c.6937+594T>G) [88], ATM intron 11 (c.1236−405C>T) (Figure 2(c))[171]. [82], and ATM IVS19 (c.2639−384A>G) [79]. Contrary to the Fibronectin IIICS-C5 and AchE-R pseu- Recent advances in antisense-mediated exon skipping for dointrons, the involvement in cancerogenesis of the vari- DMD suggest that not only the splice sites but also exonic ants deriving from the alternative splicing of the 116 nt splicing enhancer sites or branch points might represent pseudointron is less characterized. Nonetheless, functional potential targets for inducing pseudoexon skipping [183, studies have demonstrated that the protein translated from 184], and that the action of ASOs might be potentiated by mouse THPO-3 is not secreted efficiently [172–174]. These coadministration of drugs, as shown for Duchenne muscular results have suggested that THPO-3 might be implicated in dystrophy [185]. the fine regulation of THPO full-length levels and in the An intriguing variation on the theme of the antisense modulation of several biological functions beyond throm- 󸀠 oligonucleotides, successfully tested for inhibiting the 5 bopoiesis. Interestingly, parallel studies have also shown an splice site of Bcl-XL, might consist in the use of tailed altered expression pattern of the THPO-3, THPO-5, and oligonucleotides containing a portion annealing to sequences THPO-6 splicing variants in human carcinomas [175]. As a immediately upstream of the target donor splice site joined result, these observations have led researchers to hypothesize 󸀠 to a nonhybridizing 5 tail that includes binding sites for the that alterations in alternative splicing of the 116 nt-THPO hnRNP A1/A2 proteins [186]. pseudointron might be employed as biomarker of tumor In spite of being prominent, antisense technologies are formation or progression [170]. not the only possibility available for targeting these aberrant In conclusion, although currently there are few examples splicing events. An alternative to antisense usage is the of pseudointron removal, at least three of these events have development of siRNA molecules capable of targeting the been associated with different levels of cancerogenesis, rang- newly created system. In this manner, only the mRNAs ing from transformation to cancer progression. Therefore, it containing the pseudoexon might be selectively degraded is advisable that future high throughput screening analysis leaving any residual normal mRNA to be properly processed should be able to better associate the alternative splicing [187]. The disadvantage of simply degrading the pseudoexon- process of these pseudointrons with tumoral events. In bearing transcript (rather than acting on the splicing event particular, with regard to their use as possible biomarkers of itself) is that everything depends on the level of pseudoexon transformation or of cancer staging/grading. inclusion (less than 100%) and on the status of the other allele (i.e., whether it is normally expressed or not). 5. Therapeutic Outlooks Finally, small molecule compounds might be also considered as possible therapeutic options in order to prevent Regarding therapies, it is likely that all this information on the recognition of pseudoexons or modulate the recognition the importance of alternative splicing for tumor development of pseudointrons. For example, it has been shown that and progression may soon become extremely important for sodium butyrate, an histone deacetylase inhibitor known the development of novel therapeutic effectors in the fight to upregulate the expression of Htra2-beta1 and SC35, pro- against cancer [176, 177]. Indeed, one of the reasons why these motes skipping of the pseudoexon activated by the CFTR defects are particularly interesting is the opportunity they 3849+10kbC>T mutation and can restore functional CFTR provide for RNA therapies and the chance to actually make channels [188]. In addition, the splicing of different NF1 the jump from the bench to the bedside. skipped exons as a result of mutations in cis-acting sequences Currently, the field of RNA therapy is rapidly growing has been restored with administration of the small molecule especially in the inhibition of undesirable splice site choices kinetin [189]. by the use of antisense oligos [178–180]. This technique, More recently, a growing body of research indicates that pioneered by Dominski and Kole [181], is based on the use TALE nucleases (TALENs) have been used with great success of suitably modified antisense oligonucleotides that target in a number of organisms to generate site-specific DNA specific sequences within introns and exons and block the variations [190]. As a result, this approach is another good 10 International Journal of Cell Biology candidate for an innovative therapeutic anticancer strategy [8] K. J. Hertel, “Combinatorial control of exon recognition,” The forcorrectionofdeepintronicmutationsthatcreatenovel JournalofBiologicalChemistry,vol.283,no.3,pp.1211–1215, splice sites. 2008. Of course, crossing the difficult gap between the bench [9] A. R. Kornblihtt, I. E. Schor, M. Allo, G. Dujardin, E. Petrillo, to the bedside will not be immediate even in the presence and M. J. Munoz, “Alternative splicing: a pivotal step between of highly efficient therapeutic molecules, whether antisense eukaryotic transcription and translation,” Nature Reviews oligos, siRNAs, or others. Indeed, before any of these can Molecular Cell Biology,vol.14,pp.153–165,2013. be used in humans to treat cancer, several additional factors [10] B. Chabot, “Directing alternative splicing: cast and scenarios,” need to be considered. First of all, there is the development Trends in Genetics,vol.12,no.11,pp.472–478,1996. of an efficient and safe carrier system to deliver these [11] D. L. Black, “Mechanisms of alternative pre-messenger RNA compounds into the human body. Secondly, these systems splicing,” Annual Review of Biochemistry,vol.72,pp.291–336, will need to be optimized in order to achieve the best balance 2003. between successful delivery, intrinsic toxicity (if any), and [12]Y.Wang,X.Xiao,J.Zhangetal.,“Acomplexnetworkoffactors avoidance of undesired immune responses (in the case of with overlapping affinities represses splicing through intronic antisense oligos). Finally, even when all these achievements elements,” Nature Structural & Molecular Biology,vol.20,pp. are met, there will still be the need to optimize recurrent- 36–45, 2013. administration protocols to determine the uptake levels, [13]S.C.Huelga,A.Q.Vu,J.D.Arnoldetal.,“Integrativegenome- clearance, and accumulation in various tissues (this is an wide analysis reveals cooperative regulation of alternative splic- often overlooked, since none of these methods will cause ing by hnRNP proteins,” Cell Reports,vol.1,no.2,pp.167–178, 2012. permanent correction of mRNA splicing defects). Nonetheless, the first results of this exciting news strategy [14] S. Pandit, Y.Zhou, L. Shiue et al., “Genome-wide analysis reveals are available, and there is the distinct possibility that RNA- SR protein cooperation and competition in regulated splicing,” Molecular Cell,vol.50,pp.223–235,2013. based treatment of cancer may soon enter the application stage. [15] T. D. Schaal and T. Maniatis, “Multiple distinct splicing enhancers in the protein-coding sequences of a constitutively spliced pre-mRNA,” Molecular and Cellular Biology,vol.19,no. Acknowledgments 1, pp. 261–273, 1999. [16] B. R. Graveley, “Sorting out the complexity of SR protein func- This work was supported by the University of Trieste- tions,” RNA,vol.6,no.9,pp.1197–1211,2000. FinanziamentoperRicercatoridiAteneotoMaurizio [17] R. Martinez-Contreras, P. Cloutier, L. Shkreta, J. F. Fisette, T. Romano and the Higher Education Founding Council for Revil, and B. Chabot, “hnRNP proteins and splicing control,” England (HEFCE) and Cancer research UK (CRUK) to Diana Advances in Experimental Medicine and Biology,vol.623,pp. Baralle. 123–147, 2007. [18] J. HnilicovaandD.Stan´ ek,ˇ “Where splicing joins chromatin,” Nucleus,vol.2,no.3,pp.182–188,2011. References [19] L. I. G. Acuna,˜ A. Fiszbein, M. Allo,´ I. E. Schor, and A. R. [1]L.T.Chow,R.E.Gelinas,T.R.Broker,andR.J.Roberts,“An Kornblihtt, “Connections between chromatin signatures and 󸀠 amazing sequence arrangement at the 5 ends of adenovirus 2 splicing,” Wiley Interdisciplinary Reviews,vol.4,no.1,pp.77– messenger RNA,” Cell,vol.12,no.1,pp.1–8,1977. 91, 2013. [2]S.M.Berget,C.Moore,andP.A.Sharp,“Splicedsegmentsat [20] G. Dujardin, C. Lafaille, E. Petrillo et al., “Transcriptional elon- 󸀠 the 5 terminus of adenovirus 2 late mRNA,” Proceedings of the gation and alternative splicing,” Biochimica et Biophysica Acta, National Academy of Sciences of the United States of America, vol.1829,no.1,pp.134–140,2013. vol. 74, no. 8, pp. 3171–3175, 1977. [21] E. Buratti and F. E. Baralle, “Influence of RNA secondary structure on the pre-mRNA splicing process,” Molecular and Cellular [3] J. E. Darnell Jr., “Reflections on the history of pre-mRNA proce- Biology, vol. 24, no. 24, pp. 10505–10514, 2004. ssing and highlights of current knowledge: a unified picture,” RNA,vol.19,no.4,pp.443–460,2013. [22] A. Khanna and S. Stamm, “Regulation of alternative splicing by short non-coding nuclear RNAs,” RNA Biology,vol.7,no.4,pp. [4] T. W.Nilsen, “The spliceosome: the most complex macromolec- 480–485, 2010. ular machine in the cell?” Bioessays,vol.25,no.12,pp.1147–1149, [23] R. F. Luco and T. Misteli, “More than a splicing code: integrating 2003. the role of RNA, chromatin and non-coding RNA in alternative [5]C.L.WillandR.Luhrmann,¨ “Spliceosome structure and func- splicing regulation,” Current Opinion in Genetics and Develop- tion,” Cold Spring Harbor Perspectives in Biology,vol.3,no.7, ment, vol. 21, no. 4, pp. 366–372, 2011. 2011. [24] G. Biamonti and J. F. Caceres, “Cellular stress and RNA splic- [6] D. R. Semlow and J. P. Staley, “Staying on message: ensuring ing,” Trends in Biochemical Sciences,vol.34,no.3,pp.146–153, fidelity in pre-mRNA splicing,” Trends in Biochemical Sciences, 2009. vol. 37, no. 7, pp. 263–273, 2012. [25] O. Kelemen, P. Convertini, Z. Zhang et al., “Function of alter- [7] L. Cartegni, S. L. Chew, and A. R. Krainer, “Listening to native splicing,” Gene,vol.514,pp.1–30,2013. silence and understanding nonsense: exonic mutations that [26] T. L. Johnson and J. Vilardell, “Regulated pre-mRNA splicing: affect splicing,” Nature Reviews Genetics,vol.3,no.4,pp.285– the ghostwriter of the eukaryotic genome,” Biochimica et Bio- 298, 2002. physica Acta,vol.1819,no.6,pp.538–545,2012. International Journal of Cell Biology 11

[27] S. Zheng and D. L. Black, “Alternative pre-mRNA splicing in [46] J. P.Venables, “Aberrant and alternative splicing in cancer,” Can- neurons: growing up and extending its reach,” Trends in Genet- cer Research,vol.64,no.21,pp.7647–7654,2004. ics,vol.29,no.8,pp.442–448,2013. [47] Z. Kalnin¸a, P.Zayakin, K. Silin¸a, and A. Line, “Alterations of pre- [28] A. D. Norris and J. A. Calarco, “Emerging roles of alternative mRNA splicing in cancer,” Genes Chromosomes and Cancer,vol. pre-mRNA splicing regulation in neuronal development and 42,no.4,pp.342–357,2005. function,” Frontiers in Neuroscience, vol. 6, no. 122, 2012. [48] A. Srebrow and A. R. Kornblihtt, “The connection between [29] J. S. Mattick, “The central role of RNA in human development splicing and cancer,” JournalofCellScience,vol.119,no.13,pp. and cognition,” FEBS Letters,vol.585,no.11,pp.1600–1616,2011. 2635–2641, 2006. [30] P. Grabowski, “Alternative splicing takes shape during neuronal [49] C. A. Pettigrew and M. A. Brown, “Pre-mRNA splicing aberra- development,” Current Opinion in Genetics and Development, tions and cancer,” Frontiers in Bioscience,vol.13,no.3,pp.1090– vol. 21, no. 4, pp. 388–394, 2011. 1105, 2008. [31] K. Yap and E. V. Makeyev, “Regulation of gene expression in [50]A.R.Grosso,S.Martins,andM.Carmo-Fonseca,“Theemerg- mammalian nervous system through alternative pre-mRNA ing role of splicing factors in cancer,” EMBO Reports,vol.9,no. splicing coupled with RNA quality control mechanisms,” Molec- 11,pp.1087–1093,2008. ular and Cellular Neuroscience,2013. [51] C. Ghigna, M. Moroni, C. Porta, S. Riva, and G. Biamonti, [32] A. Montecucco and G. Biamonti, “Pre-mRNA processing fac- “Altered expression of heterogeneous nuclear ribonucleopro- tors meet the DNA damage response,” Frontiers in Genetics,vol. teins and SR factors in human colon adenocarcinomas,” Cancer 4, no. 102, 2013. Research, vol. 58, no. 24, pp. 5818–5824, 1998. [33] Z. Melamed, A. Levy, R. Ashwal-Fluss et al., “Alternative splic- [52] L. K. Zerbe, I. Pino, R. Pio et al., “Relative amounts of antagon- ing regulates biogenesis of miRNAs located across exon-intron istic splicing factors, hnRNP A1 and ASF/SF2, change during junctions,” Molecular Cell,vol.50,no.6,pp.869–881,2013. neoplastic lung growth: implications for pre-mRNA processing,” Molecular Carcinogenesis,vol.41,no.4,pp.187–196,2004. [34] N. Skoko, M. Baralle, S. Tisminetzky, and E. Buratti, “InTRONs in biotech,” Molecular Biotechnology,vol.48,no.3,pp.290–297, [53]R.Karni,E.DeStanchina,S.W.Lowe,R.Sinha,D.Mu,andA. 2011. R. Krainer, “The gene encoding the splicing factor SF2/ASF isa proto-oncogene,” Nature Structural and Molecular Biology,vol. [35] Q.Pan,O.Shai,L.J.Lee,B.J.Frey,andB.J.Blencowe,“Deepsur- 14,no.3,pp.185–193,2007. veying of alternative splicing complexity in the human transcriptome by high-throughput sequencing,” Nature Genetics, [54] R. A. Padgett, “New connections between splicing and human vol. 40, pp. 1413–1415, 2008. disease,” Trends in Genetics,vol.28,no.4,pp.147–154,2012. [36]E.Buratti,M.Baralle,andF.E.Baralle,“Defectivesplicing, [55] Y. Fu, B. Huang, Z. Shi et al., “SRSF1 and SRSF9 RNA bind- disease and therapy: searching for master checkpoints in exon ing proteins promote Wnt signalling-mediated tumorigenesis definition,” Nucleic Acids Research,vol.34,no.12,pp.3494–3510, by enhancing beta-catenin biosynthesis,” EMBO Molecular 2006. Medicine,vol.5,pp.737–750,2013. [56] M. Hirschfeld, M. Jaeger, E. Buratti et al., “Expression of tumor- [37] M. Krawczak, N. S. T.Thomas, B. Hundrieser et al., “Single base- promoting Cyr61 is regulated by hTRA2-𝛽1 and acidosis,” pair substitutions in exon-intron junctions of human genes: Human Molecular Genetics,vol.20,no.12,pp.2356–2365,2011. nature, distribution, and consequences for mRNA splicing,” Human Mutation, vol. 28, no. 2, pp. 150–158, 2007. [57] L. H. Boise, M. Gonzalez-Garcia, C. E. Postema et al., “bcl-x, a bcl-2-related gene that functions as a dominant regulator of [38] A. Dhir and E. Buratti, “Alternative splicing: role of pseu- apoptotic cell death,” Cell,vol.74,no.4,pp.597–608,1993. doexons in human disease and potential therapeutic strategies: minireview,” FEBS Journal, vol. 277, no. 4, pp. 841–855, 2010. [58] D. Garneau, T.Revil, J. F.Fisette, and B. Chabot, “Heterogeneous nuclear ribonucleoprotein F/H proteins modulate the alterna- [39] D. Baralle, A. Lucassen, and E. Buratti, “Missed threads: the tive splicing of the apoptotic mediator Bcl-x,” The Journal of impact of pre-mRNA splicing defects on clinical practice,” Biological Chemistry,vol.280,no.24,pp.22641–22650,2005. EMBO Reports, vol. 10, no. 8, pp. 810–816, 2009. [59] A. Massiello, J. R. Roesser, and C. E. Chalfant, “SAP155 binds [40] J. Eswaran, A. Horvath, S. Godbole et al., “RNA sequencing of to ceramide-responsive RNA cis-element 1 and regulates the 󸀠 cancer reveals novel splicing alterations,” Scientific Reports,vol. alternative 5 splice site selection of Bcl-x pre-mRNA,” The 3, article 1689, 2013. FASEB Journal,vol.20,no.10,pp.1680–1682,2006. [41] J. P. Venables, “Unbalanced alternative splicing and its signifi- [60]P.Cloutier,J.Toutant,L.Shkreta,S.Goekjian,T.Revil,andB. cance in cancer,” Bioessays,vol.28,no.4,pp.378–386,2006. Chabot, “Antagonistic effects of the SRp30c protein and cryptic 󸀠 [42] A. Relogio,´ C. Ben-Dov, M. Baum et al., “Alternative splicing 5 splice sites on the alternative splicing of the apoptotic microarrays reveal functional expression of neuron-specific regulator Bcl-x,” TheJournalofBiologicalChemistry,vol.283, regulators in Hodgkin lymphoma cells,” The Journal of Biological no. 31, pp. 21315–21324, 2008. Chemistry,vol.280,no.6,pp.4779–4784,2005. [61] T. Revil, J. Toutant, L. Shkreta, D. Garneau, P. Cloutier, and B. [43]R.Klinck,A.Bramard,L.Inkeletal.,“Multiplealternative Chabot, “Protein kinase C-dependent control of Bcl-x alterna- splicing markers for ovarian cancer,” Cancer Research,vol.68, tive splicing,” Molecular and Cellular Biology,vol.27,no.24,pp. no. 3, pp. 657–663, 2008. 8431–8441, 2007. [44] C. Ghigna, C. Valacca, and G. Biamonti, “Alternative splicing [62] M. Montes, A. Cloutier, N. Sanchez-Hern´ andez´ et al., “TCERG1 and tumor progression,” Current Genomics,vol.9,no.8,pp. regulates alternative splicing of the Bcl-x gene by modulating 556–570, 2008. the rate of RNA polymerase II transcription,” Molecular and [45] C. J. David and J. L. Manley, “Alternative pre-mRNA splicing Cellular Biology,vol.32,no.4,pp.751–762,2012. regulation in cancer: pathways and programs unhinged,” Genes [63] L. Michelle, A. Cloutier, J. Toutant et al., “Proteins associated and Development,vol.24,no.21,pp.2343–2364,2010. with the exon junction complex also control the alternative 12 International Journal of Cell Biology

splicing of apoptotic regulators,” Molecular and Cellular Biology, [80] F. Pagani, E. Buratti, C. Stuani, R. Bendix, T. Dork,¨ and F. E. vol. 32, no. 5, pp. 954–967, 2012. Baralle, “A new type of mutation causes a splicing defect in [64]E.Stickeler,F.Kittrell,D.Medina,andS.M.Berget,“Stage- ATM,” Nature Genetics,vol.30,no.4,pp.426–429,2002. specific changes in SR splicing factors and alternative splicing in [81] C. M. McConville, T. Stankovic, P. J. Byrd et al., “Mutations ass- mammary tumorigenesis,” Oncogene,vol.18,no.24,pp.3574– ociated with variant phenotypes in ataxia-telangiectasia,” Amer- 3582, 1999. icanJournalofHumanGenetics,vol.59,no.2,pp.320–330,1996. [65] A.Scorilas,L.Kyriakopoulou,D.Katsaros,andE.P.Diamandis, [82] S. Cavalieri, E. Pozzi, R. A. Gatti, and A. Brusco, “Deep- “Cloning of a gene (SR-A1), encoding for a new member of intronic ATM mutation detected by genomic resequencing and the human Ser/Arg-rich family of pre-mRNA splicing factors: corrected in vitro by antisense morpholino oligonucleotide overexpression in aggressive ovarian cancer,” British Journal of (AMO),” European Journal of Human Genetics,vol.21,no.7,pp. Cancer, vol. 85, no. 2, pp. 190–198, 2001. 774–778, 2013. [66] H. Sun and L. A. Chasin, “Multiple splicing defects in an [83] N. Sorel, C. Mayeur-Rousse, S. Deverriere` et al., “Comprehen- intronic false exon,” Molecular and Cellular Biology,vol.20,no. sive characterization of a novel intronic pseudo-exon inserted 17, pp. 6414–6425, 2000. within an e14/a2 BCR-ABL rearrangement in a patient with [67] M. Sironi, G. Menozzi, L. Riva et al., “Silencer elements chronic myeloid leukemia,” Journal of Molecular Diagnostics, as possible inhibitors of pseudoexon splicing,” Nucleic Acids vol.12,no.4,pp.520–524,2010. Research,vol.32,no.5,pp.1783–1791,2004. [84] J. Laudadio, M. W. N. Deininger, M. J. Mauro, B. J. Druker, and [68] X. H. F. Zhang and L. A. Chasin, “Computational definition of R. D. Press, “An intron-derived insertion/truncation mutation sequence motifs governing constitutive exon splicing,” Genes in the BCR-ABL kinase domain in chronic myeloid leukemia and Development,vol.18,no.11,pp.1241–1250,2004. patients undergoing kinase inhibitor therapy,” Journal of Molec- [69] W. G. Fairbrother and L. A. Chasin, “Human genomic sequen- ular Diagnostics,vol.10,no.2,pp.177–180,2008. ces that inhibit splicing,” Molecular and Cellular Biology,vol.20, [85] N. B. Quigley, D. C. Henley, R. A. Hubbard, J. Laudadio, no. 18, pp. 6816–6825, 2000. and R. D. Press, “ABL kinase domain pseudoexon insertion is [70] X. H. F. Zhang, C. S. Leslie, and L. A. Chasin, “Dichotomous not uncommon in BCR-ABL transcripts,” Journal of Molecular splicingsignalsinexonflanks,”Genome Research,vol.15,no.6, Diagnostics,vol.10,no.5,pp.475–476,2008. pp. 768–779, 2005. [86] V.Balz,H.B.Prisack,H.Bier,andH.Bojar,“AnalysisofBRCA1, [71] E. Buratti, A. Dhir, M. A. Lewandowska, and F. E. Baralle, “RNA TP53, and TSG101 germline mutations in German breast and/or structure is a key regulatory element in pathology ATM and ovarian cancer families,” Cancer Genetics and Cytogenetics,vol. CFTR pseudoexon inclusion events,” Nucleic Acids Research, 138, no. 2, pp. 120–127, 2002. vol.35,no.13,pp.4369–4383,2007. [87] X. Chen, T. T. N. Truong, J. Weaver et al., “Intronic alterations [72]A.Dhir,E.Buratti,M.A.VanSanten,R.Luhrmann,¨ and F. E. in BRCA1 and BRCA2: effect on mRNA splicing fidelity and Baralle, “The intronic splicing code: multiple factors involved in expression,” Human Mutation,vol.27,no.5,pp.427–435,2006. ATM pseudoexon definition,” The EMBO Journal,vol.29,no.4, pp.749–760,2010. [88] O. Anczukow, M. Buisson, M. Leone et al., “BRCA2 deep intronic mutation causing activation of a cryptic exon: opening [73] I. Scheinin, S. Myllykangas, I. Borze, T.Bohling,¨ S. Knuutila, and toward a new preventive therapeutic strategy,” Clinical Cancer J. Saharinen, “CanGEM: mining gene copy number changes in Research,vol.18,pp.4903–4909,2012. cancer,” Nucleic Acids Research,vol.36,no.1,pp.D830–D835, 2008. [89]M.Wang,H.Dotzlaw,S.A.W.Fuqua,andL.C.Murphy,“A pointmutationinthehumanestrogenreceptorgeneisasso- [74] G. K. Hu, S. J. Madore, B. Moldover et al., “Predicting splice ciatedwiththeexpressionofanabnormalestrogenreceptor variant from DNA chip expression data,” Genome Research,vol. mRNA containing a 69 novel nucleotide insertion,” Breast 11, no. 7, pp. 1237–1345, 2001. Cancer Research and Treatment,vol.44,no.2,pp.145–151,1997. [75] W. Fan, N. Khalid, A. R. Hallahan, J. M. Olson, and P. Z. Lue, “Astatistical method for predicting splice variants between [90] M. Clendenning, D. D. Buchanan, M. D. Walsh et al., “Mutation two groups of samples using GeneChip expression array data,” deep within an intron of MSH2 causes Lynch syndrome,” Theoretical Biology and Medical Modelling,vol.3,article19, Familial Cancer, vol. 10, no. 2, pp. 297–301, 2011. 2006. [91] J. Fernandez-Rodr´ ´ıguez, J. Castellsague,´ L. Benito et al., “Amild [76]M.Raponi,E.Buratti,M.Llorian,C.Stuani,C.W.J.Smith,and neurofibromatosis type 1 phenotype produced by the combina- D. Baralle, “Polypyrimidine tract binding protein regulates tion of the benign nature of a leaky NF1-splice mutation and the alternative splicing of an aberrant pseudoexon in NF1,” FEBS presence of a complex mosaicism,” Human Mutation,vol.32,no. Journal,vol.275,no.24,pp.6101–6108,2008. 7, pp. 705–709, 2011. [77] E. J. Wagner and M. A. Garcia-Blanco, “Minireview: polypyrim- [92] E. Pros, J. Fernandez-Rodr´ ´ıguez, B. Canet et al., “Antisense ther- idine tract binding protein antagonizes exon definition,” Molec- apeutics for neurofibromatosis type 1 caused by deep intronic ular and Cellular Biology, vol. 21, no. 10, pp. 3281–3288, 2001. mutations,” Human Mutation,vol.30,no.3,pp.454–462,2009. [78] I. Spier, S. Horpaopan, S. Vogt et al., “Deep intronic APC muta- [93] M. Raponi, M. Upadhyaya, and D. Baralle, “Functional splicing tions explain a substantial proportion of patients with familial assay shows a pathogenic intronic mutation in neurofibromato- or early-onset adenomatous polyposis,” Human Mutation,vol. sis type 1 (NF1) due to intronic sequence exonization,” Human 33,no.7,pp.1045–1050,2012. mutation,vol.27,no.3,pp.294–295,2006. [79] K. Nakamura, L. Du, R. Tunuguntla et al., “Functional charac- [94] E. Pros, C. Gomez,´ T. Mart´ın, P. Fabregas,´ E. Serra, and C. terization and targeted correction of atm mutations identified in Lazaro,´ “Nature and mRNA effect of 282 different NF1 point japanese patients with ataxia-telangiectasia,” Human Mutation, mutations: focus on splicing alterations,” Human mutation,vol. vol. 33, no. 1, pp. 198–208, 2012. 29,no.9,pp.E173–E193,2008. International Journal of Cell Biology 13

[95] K. Wimmer, X. Roca, H. Beiglbocketal.,“Extensiveinsilico¨ [110] N. Akawi, L. Al-Gazali, and B. R. Ali, “Clinical and molecular analysis of NF1 splicing defects uncovers determinants for splic- analysis of UAE fibrochondrogenesis patients expands the phe- 󸀠 ing outcome upon 5 splice-site disruption,” Human Mutation, notype and reveals two COL11A1 homozygous null mutations,” vol.28,no.6,pp.599–612,2007. Clinical Genetics,vol.82,pp.147–156,2012. [96] G. Perrin, M. A. Morris, S. E. Antonarakis, E. Boltshauser, and [111] R. Varon, R. Gooding, C. Steglich et al., “Partial deficiency P.Hutter, “Two novel mutations affecting mRNA splicing of the of the C-terminal-domain phosphatase of RNA polymerase neurofibromatosis type 1 (NF1) gene,” Human Mutation,vol.7, II is associated with congenital cataracts facial dysmorphism pp.172–175,1996. neuropathy syndrome,” Nature Genetics,vol.35,no.2,pp.185– [97] E. Ars, E. Serra, J. Garcia et al., “Mutations affecting mRNA 189, 2003. splicing are the most common molecular defects in patients [112] A. Rump, A. Rosen-Wolff,¨ M. Gahr et al., “A splice-supporting with neurofibromatosis type 1,” Human Molecular Genetics,vol. intronic mutation in the last bp position of a cryptic exon within 9, no. 2, pp. 237–247, 2000. intron 6 of the CYBB gene induces its incorporation into the [98] A. de Klein, P.H. J. Riegman, E. K. Bijlsma et al., “AG → Atran- mRNA causing chronic granulomatous disease (CGD),” Gene, sition creates a branch point sequence and activation of a cryptic vol. 371, no. 2, pp. 174–181, 2006. exon, resulting in the hereditary disorder neurofibromatosis 2,” [113] D. Noack, P.G. Heyworth, P.E. Newburger, and A. R. Cross, “An Human Molecular Genetics,vol.7,no.3,pp.393–398,1998. unusual intronic mutation in the CYBB gene giving rise to [99] C. Dehainault, D. Michaux, S. Pages-Berhouet` et al., “A deep chronic granulomatous disease,” Biochimica et Biophysica Acta, intronic mutation in the RB1 gene leads to intronic sequence vol. 1537, no. 2, pp. 125–131, 2001. exonisation,” European Journal of Human Genetics,vol.15,no. [114] D. Y. Hwang, C. C. Hung, F. G. Riepe et al., “CYP17A1 4, pp. 473–477, 2007. intron mutation causing cryptic splicing in 17𝛼-hydroxylase [100]K.Mayer,W.Ballhausen,W.Leistner,andH.D.Rott,“Three deficiency,” PloS ONE,vol.6,articlee25492,no.9,2011. novel types of splicing aberrations in the tuberous sclerosis [115] H. Ikeda, Y. Matsubara, H. Mikami et al., “Molecular analysis TSC2 gene caused by mutations apart from splice consensus of dihydropteridine reductase deficiency: identification of two sequences,” Biochimica et Biophysica Acta,vol.1502,no.3,pp. novel mutations in Japanese patients,” Human Genetics,vol.100, 495–507, 2000. no. 5-6, pp. 637–642, 1997. [101] M. V. Masala, S. Scapaticci, C. Olivieri et al., “Epidemiology [116]A.B.P.vanKuilenburg,J.Meijer,A.N.P.M.Muletal., and clinical aspects of Werner’s syndrome in North Sardinia: “Intragenic deletions and a deep intronic mutation affecting description of a cluster,” European Journal of Dermatology,vol. pre-mRNA splicing in the dihydropyrimidine dehydrogenase 17,no.3,pp.213–216,2007. gene as novel mechanisms causing 5-fluorouracil toxicity,” [102]K.Friedrich,L.Lee,D.F.Leistritzetal.,“WRNmutationsin Human Genetics,vol.128,no.5,pp.529–538,2010. Werner syndrome patients: genomic rearrangements, unusual [117] D. C. Guo, P.Gupta, V.Tran-Fadulu et al., “AnFBN1 pseudoexon intronic mutations and ethnic-specific alterations,” Human mutation in a patient with Marfan syndrome: confirmation Genetics, vol. 128, no. 1, pp. 103–111, 2010. of cryptic mutations leading to disease,” JournalofHuman [103] S. E. Flanagan, W.Xie, R. Caswell et al., “Next-generation sequ- Genetics,vol.53,no.11-12,pp.1007–1011,2008. encing reveals deep intronic cryptic ABCC8 and HADH splic- [118] L. A. Metherell, S. A. Akker, P. B. Munroe et al., “Pseudoexon ing founder mutations causing hyperinsulinism by pseudoexon activation as a novel mechanism for disease resulting in atypical activation,” The American Journal of Human Genetics,vol.92, growth-hormone insensitivity,” American Journal of Human pp. 131–136, 2013. Genetics,vol.69,no.3,pp.641–646,2001. [104]M.Milh,A.Pop,W.Kanhaietal.,“Atypicalpyridoxine-depen- [119] S. A. Akker, S. Misra, S. Aslam et al., “Pre-spliceosomal binding dent epilepsy due to a pseudoexon in ALDH7A1,” Molecular of U1 small nuclear ribonucleoprotein (RNP) and heterogenous Genetics and Metabolism,vol.105,no.4,pp.684–686,2012. nuclear RNP E1 is associated with suppression of a growth [105]W.I.Lee,T.R.Torgerson,M.J.Schumacher,L.Yel,Q.Zhu,and hormone receptor pseudoexon,” Molecular Endocrinology,vol. H.D.Ochs,“Molecularanalysisofalargecohortofpatientswith 21, no. 10, pp. 2529–2540, 2007. the hyper immunoglobulin M (IgM) syndrome,” Blood,vol.105, [120] R. Vervoort, R. Gitzelmann, W. Lissens, and I. Liebaers, “Amu- no. 5, pp. 1881–1890, 2005. tation (IVS8+0.6kbdelTC) creating a new donor splice site [106] A. I. den Hollander, R. K. Koenekoop, S. Yzer et al., “Mutations activates a cryptic exon in an Alu-element in intron 8 of the in the CEP290 (NPHP6) gene are a frequent cause of Leber human 𝛽-glucuronidase gene,” Human Genetics,vol.103,no.6, congenital amaurosis,” AmericanJournalofHumanGenetics, pp. 686–693, 1998. vol. 79, pp. 556–561, 2006. [121] J. Purevsuren, T. Fukao, Y. Hasegawa, S. Fukuda, H. Kobayashi, [107] R. W. Collin, A. I. den Hollander, S. D. van der Velde-Visser, J. and S. Yamaguchi, “Study of deep intronic sequence exoniza- Bennicelli, J. Bennett, and F. P. Cremers, “Antisense Oligonu- tion in a Japanese neonate with a mitochondrial trifunctional cleotide (AON)-based Therapy for Leber Congenital Amaurosis protein deficiency,” Molecular Genetics and Metabolism,vol.95, caused by a Frequent Mutation in CEP290,” Molecular Therapy, no. 1-2, pp. 46–51, 2008. vol. 1, article e14, 2012. [122] N. Sevenet,´ A. Lellouch-Tubiana, D. Schofield et al., “Spectrum [108]J.A.J.M.vandenHurk,D.J.R.vandePol,B.Wissingeretal., of hSNF5/INI1 somatic mutations in human cancer and “Novel types of mutation in the choroideremia (CHM) gene: genotype-phenotype correlations,” Human Molecular Genetics, a full-length L1 insertion and an intronic mutation activating a vol.8,no.13,pp.2359–2368,1999. cryptic exon,” Human Genetics,vol.113,no.3,pp.268–275,2003. [123] A.Olsson,L.Lind,L.E.Thornell,andM.Holmberg,“Myopathy [109] B. Knebelmann, L. Forestier, L. Drouot et al., “Splice-mediated with lactic acidosis is linked to chromosome 12q23.3-24.11 and insertion of an Alu sequence in the COL4A3 mRNA caus- caused by an intron mutation in the ISCU gene resulting in a ing autosomal recessive Alport syndrome,” Human Molecular splicing defect,” Human Molecular Genetics,vol.17,no.11,pp. Genetics,vol.4,no.4,pp.675–679,1995. 1666–1672, 2008. 14 International Journal of Cell Biology

[124]F.Mochel,M.A.Knight,W.H.Tongetal.,“Splicemutationin polycystic kidney disease,” Clinical Genetics,vol.75,no.2,pp. the Iron-Sulfur cluster scaffold protein ISCU causes myopathy 203–206, 2009. with exercise intolerance,” American Journal of Human Genetics, [138] A. I. Vega, C. Perez-Cerda, L. R. Desviat, G. Matthijs, M. Ugarte, vol. 82, no. 3, pp. 652–660, 2008. and B. Perez,´ “Functional analysis of three splicing mutations [125] G. Kollberg, M. Tulinius, A. Melberg et al., “Clinical manifesta- identified in the PMM2 gene: toward a new therapy for tion and a new ISCU mutation in ironsulphur cluster deficiency congenital disorder of glycosylation type IA,” Human Mutation, myopathy,” Brain,vol.132,no.8,pp.2170–2179,2009. vol.30,no.5,pp.795–803,2009. [126] N. Lucien, J. Chiaroni, J. P. Cartron, and P. Bailly, “Partial dele- [139] E.Schollen,L.Keldermans,F.Foulquieretal.,“Characterization tionintheJKlocuscausingaJK(null)phenotype,”Blood,vol. of two unusual truncating PMM2 mutations in two CDG-Ia 99,no.3,pp.1079–1081,2002. patients,” Molecular Genetics and Metabolism,vol.90,no.4,pp. 408–413, 2007. [127]M.Stucki,T.Suormala,B.Fowler,D.Valle,andM.R.Baum- gartner, “Cryptic exon activation by disruption of exon splice [140] T. R. Frio, T. L. McGee, N. M. Wade et al., “A single-base sub- enhancer: a novel mechanism causing 3-methyl crotonyl-CoA stitution within an intronic repetitive element causes dominant carboxylase deficiency,” The Journal of Biological Chemistry,vol. retinitis pigmentosa with reduced penetrance,” Human Muta- 284,no.42,pp.28953–28957,2009. tion,vol.30,no.9,pp.1340–1347,2009. [128] H. Yamaguchi, T. Fujimoto, S. Nakamura et al., “Aberrant [141]H.C.Liu,H.L.Eng,Y.F.Yangetal.,“AberrantRNAsplicing splicing of the milk fat globule-EGF factor 8 (MFG-E8) gene in RHD 7-9 exons of DEL individuals in Taiwan: A Mechanism in human systemic lupus erythematosus,” European Journal of Study,” Biochimica et Biophysica Acta,vol.1800,no.6,pp.565– Immunology,vol.40,no.6,pp.1778–1785,2010. 573, 2010. [142] N. Monnier, A. Ferreiro, I. Marty, A. Labarre-Villa, P. Mezin, [129] C. Mancini, G. Vaula, L. Scalzitti et al., “Megalencephalic leu- and J. Lunardi, “A homozygous splicing mutation causing a koencephalopathy with subcortical cysts type 1 (MLC1) due depletion of skeletal muscle RYR1 is associated with multi- to a homozygous deep intronic splicing mutation (c.895- minicore disease congenital myopathy with ophthalmoplegia,” 226T>G) abrogated in vitro using an antisense morpholino Human Molecular Genetics,vol.12,no.10,pp.1171–1178,2003. oligonucleotide,” Neurogenetics,vol.13,no.3,pp.205–214,2012. [143] K. Nozu, K. Iijima, Y. Nozu et al., “A deep intronic mutation [130] K. Homolova, P. Zavadakova, T. K. Doktor, L. D. Schroeder, V. > in the SLC12A3 gene leads to gitelman syndrome,” Pediatric Kozich, and B. S. Andresen, “The deep intronic c.903+469T C Research,vol.66,no.5,pp.590–593,2009. mutation in the MTRR gene creates an SF2/ASF binding exonic [144] C. Vache,´ T. Besnard, P.le Berre et al., “Usher syndrome type 2 splicing enhancer, which leads to pseudoexon activation and caused by activation of an USH2A pseudoexon: implications for causesthecblEtypeofhomocystinuria,”Human Mutation,vol. diagnosis and therapy,” Human Mutation,vol.33,no.1,pp.104– 31,no.4,pp.437–444,2010. 108, 2012. [131] F. Vetrini, R. Tammaro, S. Bondanza et al., “Aberrant splicing [145] E. Ruoslahti, “Fibronectin and its receptors,” Annual Review of in the ocular albinism type 1 gene (OA1/GPR143) is corrected Biochemistry, vol. 57, pp. 375–413, 1988. in vitro by morpholino antisense oligonucleotides,” Human Mutation,vol.27,no.5,pp.420–426,2006. [146]E.S.White,F.E.Baralle,andA.F.Muro,“Newinsightsinto form and function of fibronectin splice variants,” Journal of [132] G. A. Mitchell, D. Labuda, G. Fontaine et al., “Splice-medi- Pathology,vol.216,no.1,pp.1–14,2008. ated insertion of an Alu sequence inactivates ornithine 𝛿- [147] K. Vibe-Pedersen, S. Magnusson, and F. E. Baralle, “Donor and aminotransferase: a role for Alu elements in human mutation,” acceptor splice signals within an exon of the human fibronectin Proceedings of the National Academy of Sciences of the United gene: a new type of differential splicing,” FEBS Letters,vol.207, States of America, vol. 88, no. 3, pp. 815–819, 1991. no. 2, pp. 287–291, 1986. [133] T.R.Webb,D.A.Parfitt,J.C.Gardneretal.,“Deepintronicmu- [148] R. P.Hershberger and L. A. Culp, “Cell-type-specific expression tation in OFD1, identified by targeted genomic next-generation of alternatively spliced human fibronectin IIICS mRNAs,” sequencing, causes a severe form of X-linked retinitis pigmen- Molecular and Cellular Biology,vol.10,no.2,pp.662–671,1990. tosa (RP23),” Human Molecular Genetics,vol.21,pp.3647–3654, [149] A. P. Mould and M. J. Humphries, “Identification of a novel 2012. recognition sequence for the integrin 𝛼4𝛽1intheCOOH- [134] W.Ogino, Y. Takeshima, A. Nishiyama et al., “Mutation analysis terminal heparin-binding domain of fibronectin,” The EMBO oftheornithinetranscarbamylase(OTC)geneinfiveJapanese Journal,vol.10,no.13,pp.4089–4095,1991. OTC deficiency patients revealed two known and three novel [150] M. J. Humphries, A. Komoriya, and S. K. Akiyama, “Identifica- mutations including a deep intronic mutation,” Kobe Journal of tion of two distinct regions of the type III connecting segment Medical Sciences, vol. 53, no. 5, pp. 229–240, 2007. of human plasma fibronectin that promote cell type-specific [135] A. Rincon,´ C. Aguado, L. R. Desviat, R. Sanchez-Alcudia,´ M. adhesion,” TheJournalofBiologicalChemistry,vol.262,no.14, Ugarte, and B. Perez,´ “Propionic and methylmalonic acidemia: pp. 6886–6892, 1987. antisense therapeutics for intronic variations causing aberrantly [151]A.P.Mould,J.A.Askari,S.E.Craig,A.N.Garratt,J.Clements, spliced messenger RNA,” American Journal of Human Genetics, andM.J.Humphries,“Integrin𝛼4𝛽1-mediated melanoma cell vol. 81, no. 6, pp. 1262–1270, 2007. adhesion and migration on vascular cell adhesion molecule- [136]P.T.Christie,B.Harding,M.A.Nesbit,M.P.Whyte,andR. 1 (VCAM-1) and the alternatively spliced IIICS region of V. Thakker, “X-linked hypophosphatemia attributable to pseu- fibronectin,” The Journal of Biological Chemistry,vol.269,no. doexons of the PHEX gene,” Journal of Clinical Endocrinology 44, pp. 27224–27230, 1994. and Metabolism,vol.86,no.8,pp.3840–3844,2001. [152] T. Kumazaki, Y. Mitsui, K. Hamada, H. Sumida, and M. Nish- [137] L. Michel-Calemard, F. Dijoud, M. Till et al., “Pseudoexon iyama, “Detection of alternative splicing of fibronectin mRNA activation in the PKHD1 gene: a French founder intronic in a single cell,” JournalofCellScience,vol.112,part10,no.10, mutation IVS46+653A>G causing severe autosomal recessive pp.1449–1453,1999. International Journal of Cell Biology 15

[153] K. P. Schofield and M. J. Humphries, “Identification of fibrone- [169] C. Dame, E. M. Wolber, P. Freitag, D. Hofmann, P. Bartmann, ctin IIICS variants in human bone marrow stroma,” Blood,vol. and J. Fandrey, “Thrombopoietin gene expression in the devel- 93, no. 1, pp. 410–411, 1999. oping human central nervous system,” Developmental Brain [154] U. Trefzer, Y. Chen, G. Herberth et al., “The monoclonal Research,vol.143,no.2,pp.217–223,2003. antibody SM5-1 recognizes a fibronectin variant which is widely [170] R. Marcucci and M. Romano, “Thrombopoietin and its splic- expressed in melanoma,” BMC Cancer, vol. 6, article 8, 2006. ing variants: structure and functions in thrombopoiesis and beyond,” Biochimica et Biophysica Acta,vol.1782,no.7-8,pp. [155] H. J. Mardon and G. Sebastio, “Regulation of alternative splicing 427–432, 2008. in the IIICS region of human fibronectin pre-mRNA encoding cell binding sites CS1 and CS5,” Journal of Cell Science,vol.103, [171] M. Romano, R. Marcucci, and F. E. Baralle, “Splicing of con- part 2, no. 2, pp. 423–433, 1992. stitutive upstream introns is essential for the recognition of intra-exonic suboptimal splice sites in the thrombopoietin [156] I. Silman and J. L. Sussman, “Acetylcholinesterase: “Classi- gene,” Nucleic Acids Research, vol. 29, no. 4, pp. 886–894, 2001. cal” and “non-classical” functions and pharmacology,” Current [172] T. Wada, Y. Nagata, H. Nagahisa et al., “Characterization of Opinion in Pharmacology,vol.5,no.3,pp.293–302,2005. the truncated thrombopoietin variants,” Biochemical and Bio- [157] H. Soreq and S. Seidman, “Acetylcholinesterase—new roles for physical Research Communications,vol.213,no.3,pp.1091–1098, an old actor,” Nature Reviews Neuroscience,vol.2,no.4,pp.294– 1995. 302, 2001. [173] R. C. Hoffman, H. Andersen, K. Walker et al., “Peptide, [158] J. Massoulie,´ “The origin of the molecular diversity and func- disulfide, and glycosylation mapping of recombinant human tional anchoring of cholinesterases,” NeuroSignals, vol. 11, no. 3, thrombopoietin from Ser1 to Arg246,” Biochemistry,vol.35,no. pp. 130–143, 2002. 47, pp. 14849–14861, 1996. [159] R. Y. Y. Chan, F. A. Adatia, A. M. Krupa, and B. J. Jasmin, [174] X. L. Wu, M. Nakayama, and J. W. Adamson, “Identification of “Increased expression of acetylcholinesterase T and R tran- five new isoforms of murine thrombopoietin mRNA,” Biochem- scripts during hematopoietic differentiation is accompanied by ical and Biophysical Research Communications,vol.276,no.1, parallel elevations in the levels of their respective molecular pp.137–143,2000. forms,” The Journal of Biological Chemistry,vol.273,no.16,pp. [175] Y. Sasaki, T. Takahashi, H. Miyazaki et al., “Production of thr- 9727–9733, 1998. ombopoietin by human carcinomas and its novel isoforms,” Blood,vol.94,no.6,pp.1952–1960,1999. [160] N. A. Perrier, M. Salani, C. Falasca, S. Bon, G. Augusti-Tocco, [176] S. Bonnal, L. Vigevani, and J. Valcarcel, “The spliceosome as a and J. Massoulie,´ “The readthrough variant of acetylcholine- target of novel antitumour drugs,” Nature Reviews Drug Discov- sterase remains very minor after heat shock, organophosphate ery,vol.11,pp.847–859,2012. inhibition and stress, in cell culture and in vivo,” Journal of Neurochemistry, vol. 94, no. 3, pp. 629–638, 2005. [177] E.Zaharieva,J.K.Chipman,andM.Soller,“Alternativesplicing interference by xenobiotics,” Toxicology,vol.296,no.1–3,pp.1– [161]D.Grisaru,M.Sternfeld,A.Eldor,D.Glick,andH.Soreq, 12, 2012. “Structural roles of acetylcholinesterase variants in biology and [178] F. Muntoni and M. J. A. Wood, “Targeting RNA to treat neuro- pathology,” European Journal of Biochemistry,vol.264,no.3,pp. muscular disease,” Nature Reviews Drug Discovery,vol.10,no. 672–686, 1999. 8, pp. 621–637, 2011. [162] E. Meshorer and H. Soreq, “Virtues and woes of AChE alter- [179] R. Kole, A. R. Krainer, and S. Altman, “RNA therapeutics: native splicing in stress-related neuropathologies,” Trends in beyond RNA interference and antisense oligonucleotides,” Neurosciences,vol.29,no.4,pp.216–224,2006. Nature Reviews Drug Discovery,vol.11,no.2,pp.125–140,2012. [163]C.Perry,E.H.Sklan,K.Birikhetal.,“Complexregulationof [180]A.Aartsma-RusandG.J.B.VanOmmen,“Antisense-mediated acetylcholinesterase gene expression in human brain tumors,” exon skipping: a versatile tool with therapeutic and research Oncogene,vol.21,no.55,pp.8428–8441,2002. applications,” RNA,vol.13,no.10,pp.1609–1624,2007. [164]C.Perry,E.H.Sklan,andH.Soreq,“CREBregulatesAChE-R- [181] Z. Dominski and R. Kole, “Restoration of correct splicing in induced proliferation of human glioblastoma cells,” Neoplasia, thalassemic pre-mRNA by antisense oligonucleotides,” Proceed- vol. 6, no. 3, pp. 279–286, 2004. ings of the National Academy of Sciences of the United States of America, vol. 90, no. 18, pp. 8673–8677, 1993. [165] I.Mor,T.Bruck,D.Greenbergetal.,“AlternateAChE-Rvariants facilitate cellular metabolic activity and resistance to genotoxic [182] S. Kishore and S. Stamm, “The snoRNA HBII-52 regulates alter- stress through enolase and RACK1 interactions,” Chemico- native splicing of the serotonin receptor 2C,” Science,vol.311,no. Biological Interactions,vol.175,no.1–3,pp.11–21,2008. 5758, pp. 230–232, 2006. [183] A. Aartsma-Rus, L. van Vliet, M. Hirschi et al., “Guidelines for [166] M. F. Montenegro, F. Ruiz-Espejo, F. J. Campoy et al., “Choline- antisense oligonucleotide design and insight into splice- sterases are down-expressed in human colorectal carcinoma,” modulating mechanisms,” Molecular Therapy,vol.17,no.3,pp. Cellular and Molecular Life Sciences,vol.63,no.18,pp.2175– 548–553, 2009. 2182, 2006. [184] H.Heemskerk,C.L.DeWinter,G.J.B.VanOmmen,J.C.T.Van [167] B. Li, H. Pan, J. C. Winkelmann, and W. Dai, “Thrombopoietin Deutekom, and A. Aartsma-Rus, “Development of antisense- and its alternatively spliced form are expressed in human mediated exon skipping as a treatment for Duchenne muscular amygdala and hippocampus,” Blood,vol.87,no.12,pp.5382– dystrophy,” Annals of the New York Academy of Sciences,vol. 5384, 1996. 1175, pp. 71–79, 2009. [168] R. Sungaran, B. Markovic, and B. H. Chong, “Localization and [185]G.C.Kendall,E.I.Mokhonova,M.Moranetal.,“Dantrolene regulation of thrombopoietin mRNA expression in human kid- enhances antisense-mediated exon skipping in human and ney, liver, bone marrow, and spleen using in situ hybridization,” mouse models of Duchenne muscular dystrophy,” Science Blood,vol.89,no.1,pp.101–107,1997. Translational Medicine, vol. 4, no. 164, article 164ra160, 2012. 16 International Journal of Cell Biology

󸀠 [186] D. Gendron, S. Carriero, D. Garneau et al., “Modulation of 5 splice site selection using tailed oligonucleotides carrying splicing signals,” BMC Biotechnology,vol.6,article5,2006. [187] M. A. Garcia-Blanco, A. P. Baraniak, and E. L. Lasda, “Alterna- tive splicing in disease and therapy,” Nature Biotechnology,vol. 22,no.5,pp.535–546,2004. [188] M. Nissim-Rafinia, M. Aviram, S. H. Randell et al., “Restoration ofthecysticfibrosistransmembraneconductanceregulator function by splicing modulation,” EMBO Reports,vol.5,no.11, pp.1071–1077,2004. [189] E. Pros, J. Fernandez-Rodr´ ´ıguez, L. Benito et al., “Modulation of aberrant NF1 pre-mRNA splicing by kinetin treatment,” Euro- pean Journal of Human Genetics,vol.18,no.5,pp.614–617,2010. [190] T. Cermak, E. L. Doyle, M. Christian et al., “Efficient design and assembly of custom TALEN and other TAL effector-based constructs for DNA targeting,” Nucleic Acids Research,vol.39, no.12,articlee82,2011. Hindawi Publishing Corporation International Journal of Cell Biology Volume 2013, Article ID 153634, 12 pages http://dx.doi.org/10.1155/2013/153634

Review Article RNA Splicing: A New Player in the DNA Damage Response

Silvia C. Lenzken, Alessia Loffreda, and Silvia M. L. Barabino

Department of Biotechnology and Biosciences, University of Milano-Bicocca, Piazza della Scienza 2, I-20126 Milan, Italy

Correspondence should be addressed to Silvia M. L. Barabino; [email protected]

Received 8 May 2013; Revised 13 August 2013; Accepted 14 August 2013

Academic Editor: Claudio Sette

Copyright © 2013 Silvia C. Lenzken et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

It is widely accepted that tumorigenesis is a multistep process characterized by the sequential accumulation of genetic alterations. However, the molecular basis of genomic instability in cancer is still partially understood. The observation that hereditary cancers are often characterized by mutations in DNA repair and checkpoint genes suggests that accumulation of DNA damage is a major contributor to the oncogenic transformation. It is therefore of great interest to identify all the cellular pathways that contribute to the response to DNA damage. Recently, RNA processing has emerged as a novel pathway that may contribute to the maintenance of genome stability. In this review, we illustrate several different mechanisms through which pre-mRNA splicing and genomic stability can influence each other. We specifically focus on the role of splicing factors in the DNA damage response and describe how,in turn, activation of the DDR can influence the activity of splicing factors.

1. Overview of the DNA Damage Response and effectors. Recognition of DNA damage is the first step in the activation of the signaling cascade that controls the DNA Genomic instability is one of the most common characteris- damage checkpoints. DNA lesions are recognized by various tics of tumor cells and is probably due to the combined effect sensor proteins: the MRN (MRE11-RAD50-NBS1) complex of DNA damage, tumor-specific DNA repair defects, and a that signals double-strand DNA breaks (DBSs), and by RPA failure to arrest the cell cycle before the damaged DNA is that binds single-strand DNA at sites of DNA damage. Sub- passed on to daughter cells. Genomic instability is recognized sequently, recruitment and activation of the highly conserved as a characteristic of most solid tumors and adult-onset apical DDR kinases ataxia-telangiectasia mutated (ATM) leukaemias and is manifested as alterations in chromosome and ataxia-telangiectasia and RAD3-related (ATR) occur. In number and structure (chromosomal instability) and as both yeast and mammals Tel1/ATM recognizes DSBs, while changes to the structure of DNA, such as nucleotide substitu- Mec1/ATR is activated in the presence of long single-stranded tions, insertions, and deletions. To maintain genomic stability DNA tracts. Once activated, ATR and ATM transduce the and to counteract DNA damage, cells have evolved a complex DDR signal by promoting the phosphorylation/activation of cellular response, called DNA damage response (DDR), downstream kinases, such as CHK1 and CHK2, which in which is coordinated by the DNA damage checkpoints [1, 2]. turn regulate downstream checkpoint effectors. Checkpoint Somatic mutations in DDR genes have been found in activation elicits a multifaceted cellular response that coor- several cancer types [3]. Indeed, on one hand, inactivation dinatescellcycleprogressionwithDNArepairactivity,thus of the DDR favors the accumulation of mutations in proto- allowing cells to block the cell cycle until the damage is oncogenes increasing the risk of tumor development. On the repaired. In addition, ATM phosphorylates the histone 2A other, since the anticancer activity of most chemotherapeutic variant 𝛾H2AX that marks chromatin regions flanking DSBs. drugs relies on the induction of DNA damage, alterations in These phosphorylation events promote the recruitment of the DDR also affect the tumor’s sensitivity to chemotherapy several mediator proteins that facilitate ATM/ATR signaling [4]. (see [5]). Conceptually the molecules that orchestrate the DDR can Recently, a novel layer of complexity in the cellular be functionally organized in sensors, mediators, transducers response to DNA damage has emerged with the involvement 2 International Journal of Cell Biology of RNA metabolism. A first link between mRNA biogenesis the expression level of specific splicing factors. SR proteins and genome stability has been provided by the observation and hnRNP proteins were the first splicing factors identi- that transcription is inhibited in response to DNA damage, fied [15, 16]. These proteins are components of the basal both generally and locally at DNA damage sites [6, 7]. splicing machinery. However, since their concentration can Changes in the pre-mRNA splicing pattern of crucial genes influence splice site selection, they contribute to alternative in the DDR have long been observed (for review see [8]), and splicing. Indeed, deregulation of several members of the SR splicingfactorshavebeenobservedtochangetheirintracel- protein and hnRNP families of splicing regulators has been lular distribution following genotoxic damage. Finally, DNA observed following various genotoxic treatments [17, 18]. One damageisknowntoaffectmRNAstabilitybothpositivelyand additional recent example is provided by the upregulation of negatively (for review see [9]). This review is focused on the SC35 by E2F1, a transcription factor that has a key function interplay between AS and DDR. We initially discuss how acti- during S phase progression and apoptosis in response to vation of the DDR signaling cascade can influence the activity DNA-damaging agents [19]. of splicing factors and how this can affect cell fate. Then, we examine how, in turn, alternative splicing factors can con- 2.2. Posttranslational Modification of Splicing Factors. As tribute to the DDR. Finally, we will discuss how alternative mentioned above the presence of phosphorylated 𝛾H2AX is splicing regulators can represent novel targets for cancer a hallmark of sites of DSBs. However, phosphorylation of therapies. H2AX is only the tip of the iceberg: a host of posttranslational modifications of both the histones and components of the 2. How DDR Can Affect Alternative DDR machinery have been reported at sites of DNA damage, Pre-mRNA Splicing (AS) including acetylation, ubiquitination and SUMOylation, and arginine methylation (for review see [20]). These modifica- In recent years several reports have uncovered how DNA tions play a central role in coordinating cell cycle progression damage can induce splicing changes that give rise to mRNA and DNA repair [21]. Therefore, it is not surprising that acti- variants encoding different protein isoforms with the poten- vation of the damage signaling cascade can lead to the post- tial to affect the cellular response and the cell fate. A first translational modification of splicing factors that can mod- indication that the DDR can alter AS came from the obser- ify their intracellular localization and/or activity. Here, we vation that Drosophila S2 cells treated with camptothecin, will discuss examples of posttranslational modifications of a topoisomerase I inhibitor that leads to replication forks splicing regulatory proteins in response to DNA damage that arrest, or exposed to ionizing radiation (IR) express specific have recently been reported. alternative splicing variants of the transcription factor TAF1 that was shown to control the G2/M transition [10]. Alter- 2.2.1. Phosphorylation. The activity and the intracellular native splicing of TAF1 in response to DNA damage was distribution of the serine/arginine (SR) proteins are tightly shown to depend on ATM/CHK2 and ATR/CHK1 signaling regulated by phosphorylation. SR proteins are RNA-binding and to induce degradation of the splicing factor Tra2 [10, 11]. proteins that are characterized by at least one domain Another example is provided by the effect of UV irradia- enriched in RS dipeptides. This region, often located at the C- tion and cisplatin on the splicing of MDM2 and MDM4 terminus of the protein, is termed RS domain and is subjected transcripts [12]. The MDM2 gene encodes an E3 ubiquitin to reversible phosphorylation. The assembly of the splicing ligase that targets p53 for proteasome-mediated degradation. complex and the catalysis of the splicing reaction require MDM2 expression is positively controlled by p53 at the dephosphorylation/phosphorylation cycles. While phospho- transcription level, generating a feedback loop. The proteins rylated SR proteins favor spliceosome assembly, intron encoded by the mRNA variants induced by DNA damage removal is associated with dephosphorylation of SR proteins. lack the p53 interaction domain so that they may favor p53 After splicing, a subset of SR proteins remains associated with activation [12]. A detailed description of the regulation of the the mature mRNA and is exported to the cytoplasm. Here, alternative splicing of the MDM2 transcript in response to their reimport into the nucleus requires phosphorylation by various genotoxic treatments can be found in [8]. SRPK1 and SRPK2, two SR protein kinases that are predo- How genotoxic stress can influence the activity of the minantly localized in the cytoplasm. It was recently shown splicing machinery is to date largely unknown. Two major that genotoxic stress can induce the phosphorylation and mechanisms are known to control the activity of splicing fac- the relocalization of these kinases to the nucleus where they tors in response to external and internal stimuli: (i) changes in turn hyperphosphorylate SR proteins leading to changes in in expression level and (ii) posttranslational modifications. pre-mRNA splicing [22, 23]. Moreover, chronic replication- In addition, it has emerged in recent years that pre-mRNA dependent DNA damage was shown to induce the hyper- splicing occurs largely cotranscriptionally and that also the phosphorylation of ASF/SF2 (SRSF1) [24]. processivity of RNA polymerase II (RNAPII) can influence Several studies underline the importance for catalysis of the recognition of alternative exons ([13, 14], Figure 1). dephosphorylation within the assembled spliceosome [25– 27]. Interestingly, the protein phosphatase PPM1G, which 2.1. The Expression Level of Splicing Factors Changes in promotes pre-mRNA splicing, is phosphorylated in response Response to DNA Damage. The simplest way by which DNA to DNA damage and is rapidly recruited to DNA damage sites damage can affect the splicing machinery is by modifying [28]. International Journal of Cell Biology 3

Genotoxic Genomic stress instability

DNA damage/DDR activation

Expression of RNAPII SFs processivity ? ? Posttranslational Localization of modifications of SFs SFs

Alternative splicing modulation

Figure 1: How the DDR can affect alternative splicing. The activation of the DDR can modify alternative splicing by affecting the expression, or by inducing posttranslational modifications of splicing factors (SFs), that may alter their intracellular localization and/or their activity. Moreover, also the elongation activity of RNA polymerase II (RNAPII) can be influenced by genotoxic stress, modifying in turn pre-mRNA splicing.

2.2.2. Acetylation. The first evidence of the role of lysine Many different E3 ubiquitin ligases participate in the acetylation in the DDR came with the observation that over- DDR. For example, RNF2 catalyzes the monoubiquitination expression of a dominant-negative allele of the acetyltrans- of H2AX contributing to ATM recruitment [32, 33], while ferase TIP60 reduced the efficiency of DSB repair29 [ ]. Now, depletion of RNF4 impairs RAP80, BRCA1, and RAD51 we know several TIP60 targets among crucial DDR factors, recruitment to sites of DNA damage [34–36]. including H2AX, H4, and ATM [30]. A recent quanti- Indirect evidence suggests that ubiquitination of splicing tative mass spectrometry analysis revealed that splicing factors may also be involved in the DDR. Ubiquitin regulates factors, including SR proteins and hnRNPs, have numer- spliceosome assembly [37]. Moreover, the essential yeast ous acetylation sites [31], often located within their RNA- splicing factor Prp19 and its human ortholog have E3 ligase binding domain. Since acetylation of lysine neutralizes the activity in vitro [38]. A role of PRP19 in the mammalian DDR positive charge of the amino acid, this reversible modi- first emerged with the report that it was strongly upregulated fication could contribute to the regulation of their activ- by DNA damage in human cells and that its depletion by ity in splicing. Consistent with this report, Edmond and siRNA resulted in an accumulation of DSBs and apoptosis colleagues recently identified the SR protein ASF/SF2 as and reduced survival after exposure to ionizing radiation a novel substrate of TIP60 acetyltransferase activity in [39]. Prp19 is part of a large complex that contains >30 response to genotoxic treatments [23]. In the case of proteins [40]. PRP19/Pso4 itself is modified by ubiquitination ASF/SF2, however, acetylation does not influence the splicing in response to DNA damage, and this modification reduces its activity but rather controls its protein turnover by pro- affinity for other members of the Pso4 complex41 [ ]. moting degradation. Although these observations clearly Besides ubiquitin, vertebrate cells encode several other demonstrate that acetylation can contribute to the regu- ubiquitin-like proteins that are structurally related to ubiq- lation of the DDR, a recent genome-wide characteriza- uitin. The best characterized one is small ubiquitin modi- tion of the DDR-regulated acetylome revealed a rather fier (SUMO). Similar to ubiquitination, SUMOylation is a weak overall increase in site-specific acetylation compared reversible posttranslational modification that plays a crucial to phosphorylation suggesting that acetylation may be a role in the control of the DDR and in DNA repair pathways by more selective/subtle modification [28]. modulating protein:protein interactions (for review see [42]). ASF/SF2, the best characterized member of the SR protein 2.2.3. Ubiquitination/SUMOylation. In addition to targeting family of splicing factors, can act as a cofactor stimulating proteins to proteasome-mediated degradation, ubiquitina- SUMO conjugation by the SUMO E2 conjugating enzyme tion has emerged as an important regulatory signal that func- Ubc9 [43]. ASF/SF2 also interacts with PIAS1 and regulates tions in many cellular processes. Ubiquitination involves the its SUMO E3 ligase activity in response to DNA damage [43]. covalent attachment of a 76 amino acid ubiquitin chain to a HnRNP K, as several other hnRNPs, is SUMOylated [44]. lysine of the modified protein by an E3 ubiquitin ligase. Mul- HnRNP K is a multifunctional protein involved in many tiple lysines can be ubiquitinated in the target protein, and steps of mRNA biogenesis [45]. Various genotoxic treatments following addition of the first ubiquitin additional ubiquitin stimulate SUMOylation of hnRNP K by the Polycomb E3 molecules can be added, yielding a polyubiquitin chain. ligase Pc2 that in turn is activated by DNA damage via 4 International Journal of Cell Biology phosphorylation by HIPK2 kinase [46]. HnRNP K cooperates Several splicing-sensitive microarray studies have exam- with p53 in the transcriptional activation of cell cycle arrest ined the effect of CPT on cotranscriptional AS54 [ –56]. genes such as 14-3-3𝜎, GADD45, and p21 in response to DNA CPT appears to reduce RNA polymerase elongation rate damage [47], and its SUMOylation stimulates p53 transcrip- promoting predominantly exon inclusion [54, 55]. Interest- tional activity [46]. ingly, many AS events leading to exon inclusion result in the production of mRNAs containing premature stop codons (PTCs) that will undergo nonsense-mediated decay. Gene 2.3. Transcriptional Effects on AS. Many chemotherapeutic Ontology analysis of the functional categories associated drugs are potent inducers of DNA damage that interferes with the AS indicated that CPT treatment appeared to with transcription. Platinum derivatives such as cisplatin and affect transcription and splicing of RNA-binding proteins carboplatin induce DNA adducts and intra- and interstrands [54, 55]. These observations suggest that this may represent cross-links between purine bases. Platinated adducts distort a mechanism that allows the cell to respond to genotoxic the DNA helix impairing replication and transcription elon- damage by rapidly adjusting the level of RNA processing gation,whichinturncanleadtotheformationofDSBs[48]. factors to the level of transcription. Camptothecin (CPT) is a topoisomerase I inhibitor that leads A specific example of AS event that is affected by DNA to a block of transcription elongation and DNA replication. damage is provided by the MDM2 gene. As mentioned previ- Doxorubicin is an inhibitor of DNA topoisomerase II that ously, the MDM2 gene produces several different mRNA vari- induces structural alterations in promoter DNA [49]. ants due to AS. The biological significance of these variants is Although pre-mRNA splicing can occur independently of presently unknown since only few of them are translated into transcription, many different studies have provided evidence proteins. The best characterized alternative MDM2 mRNA that AS, as other RNA processing events required for the isoform is the ALT1 transcript that lacks 8 of 12 exons synthesis of the mature mRNA, is mostly cotranscriptional duetoexonskipping.IthasbeenreportedthattheALT1 (for review see [50]). Several RNA processing factors are variant is upregulated by UV treatment [12, 57]. Dutertre recruited on the C-terminal domain of RNAPII and are and colleagues demonstrated that production of ALT1 and deposited on the nascent pre-mRNA molecule during tran- other variants due to exon skipping is regulated cotranscrip- scription elongation. Moreover, AS is influenced by the rate of tionally [56]. Specifically, different genotoxic treatments such transcription elongation; slowing down the polymerase may as camptothecin, doxorubicin, and cisplatin induce MDM2 favour the use of weak splice sites by delaying the synthesis exon skipping by disrupting the interaction between EWS, a of downstream splice sites, thus facilitating the recognition transcriptional coregulator [58], and the splicing factor YB- of suboptimal exons (for review see [51]).Inthelightofthese 1[56]. These results suggest that DNA damage may interfere considerations it is not surprising that genotoxic treatments with the coupling between transcription and splicing leading have been reported to induce transcription-dependent splic- to the production of alternative or aberrant mRNA variants. ing alterations. Recent work that examined the response to IR using Recent studies have uncovered how the processivity of exon sensitive microarrays in lymphoblastoid cells and in RNAPII can influence the recognition of alternative exons fibroblasts confirms the genome-wide effects on transcription [52]. Munoz˜ and colleagues showed that DNA damage and splicing induced by genotoxic stress. Exon-level analysis induced by UV can directly modulate the activity of RNAPII revealed a general increase in internal exon skipping in during transcript elongation thereby affecting the selection of response to radiation. The affected genes are involved in cell alternative exons [53]. The effect of UV and cisplatin treat- cycle regulation, chromatin dynamics, p53 regulation, and mentonASwasinitiallycharacterizedontheEDIalternative cell growth. In addition this study revealed an increase in cassette exon of the fibronectin gene. Both genotoxic treat- the use of alternative promoters. These promoters have p53 ments despite eliciting quite different types of DNA damage bindingelementsatornearthestartsitesuggestingthatthe strongly stimulated the inclusion of the EDI exon. The effects protein isoforms encoded by these mRNA variants may have induced by UV were independent of p53 since they were an active role in regulating the response to IR [59]. observed also in Hep3B cells that are considered to be p53 Collectively, these reports suggest that genotoxic damage null. Moreover, they were not due to changes in the expres- can interfere with RNA polymerase II activity and may influ- sion or intracellular localization of splicing factors known to ence cotranscriptional AS by distinct mechanisms. However, regulate this splicing event. Interestingly, however, both UV considering that the functional categories associated with the and cisplatin induced the hyperphosphorylation of RNAPII’s affected genes are related to crucial cellular programs one can CTD leading to the inhibition of transcription elongation. speculate that these changes might be functionally relevant Therefore, in this case genotoxic damage affects the kinetic for the response to DNA damage. coupling between transcription and splicing, thereby affecting cotranscriptional AS. Interestingly, UV does not generally affect the level of either gene expression or AS, but its effects 3. How AS Can Affect the DDR are restricted to a subset of responsive genes. In addition, the effects on AS induced by DNA damage may depend onthe Regulation of AS depends, on one hand, on the presence specific type of damage-inducing treatment. Indeed, when of specific sequence elements in the pre-mRNA and, on the doxorubicin was used in this same study it did not induce the other hand, on trans-acting protein factors. In this section same splicing changes observed upon UV treatment [53]. we first report the most recent findings on how mutations in International Journal of Cell Biology 5

DDR genes Splicing factors Changes in ∙ expression Novel functions Mutations Depletion ∙ activity in DDR? ∙ localization

R-loops ? Splicing alterations of DDR-related genes Genomic instability

DNA damage/DDR activation

Figure 2: Alternative splicing alterations can activate the DDR. Mutations in splicing regulatory signals can inactivate the function of genes directly involved in the DDR, resulting in the accumulation of DNA damage. However, also the inactivation of canonical splicing factors can have similar effects, either by inducing aberrant splicing of DDR genes or by perturbing cotranscriptional splicing and inducing R-loop formation. However, it cannot be ruled out that AS- and RNA-binding proteins may play novel roles in the DDR and the control of genome stability. genes involved in the DDR affect splicing of the transcript validation of the functional consequences of a specific muta- thereby altering protein function. Then, we review how tion. Hereafter, we describe examples of these types of muta- alteration of the expression of splicing factors can contribute tions affecting critical genes in the DDR (see also Table 1). to genomic instability and cancer (Figure 2). First, we will discuss mutations in the well-characterized BRCA1 and BRCA2 genes. Second, we will describe an 3.1. Mutations in Genes Involved in the DDR Can Disrupt example of intronic mutation affecting AS of the ATM gene. Pre-mRNA Splicing. The splicing machinery assembles on Then, we will illustrate how different AS isoforms of key DDR conserved sequence elements that define the intron-exon regulator can have different functions. Finally, we will review junctions, the so-called splice sites, and on the branch point how the loss of splicing factors can cause DNA damage. sequence (BPS), a poorly conserved sequence located near 󸀠 the 3 end of the intron. In addition to these core splicing signals, splicing is influenced by other regulatory elements 3.1.1. BRCA1 and BRCA2: The Reclassification Issue. BRCA1 [60]. These elements are conventionally classified as exonic (OMIM 113705) and BRCA2 (OMIM 600185), the two most splicing enhancers (ESEs) or silencers (ESSs) depending important breast cancer susceptibility genes [87], are key whether they function to promote or inhibit inclusion of the players in DDR. They are involved in homologous recombi- exon they reside in and as intronic splicing enhancers (ISEs) nation (HR) and DNA repair (or review see [1]). or silencers (ISSs) if they enhance or inhibit usage of adjacent More than 3500 mutations have been reported that splice sites or exons from an intronic location. These regula- affect the BRCA1/2 genes, about one-third of which are tory elements function by recruiting factors that activate or unclassified variants (UVs) (Breast Cancer Information Core inhibit splice site recognition or spliceosome assembly. Database (BIC), French UMD-BRCA1/2 mutation database: Considering that about the 95% of our genes produce at http://www.umd.be/BRCA2/, http://www.umd.be/BRCA1/) least two isoforms [61]thereisahighprobabilitythatmuta- that may induce aberrant splicing. In 1998 Mazoyer et al. tions affecting these cis-acting elements could induce aber- described a missense mutation within the exon 18 of the rant splicing with deleterious consequences. Accordingly, BRCA1 gene, which leads to exon skipping and consequently it has widely been suggested that most of the unclassified to the disruption of the BRCT domain, producing a non- mutations (missense or silent) could affect splicing [62–65]. functional protein [62]. Using an in vitro splicing system Furthermore, synonymous mutations, that do not change Liu and colleagues subsequently showed that exon skipping the protein sequence and therefore have traditionally been resulted from the disruption of a splicing enhancer in exon considered innocuous polymorphisms, could induce splicing 18 [88]. Subsequently, Fackenthal et al. detected a missense aberrations by modifying splice sites (either canonical or mutation in BRCA2 affecting an ESE that caused the skipping cryptic) or splicing regulatory sequences [63, 64]. Splicing of exon 18 and the out-of-frame fusion of exons 17 and 19 affecting mutations could lead to transcript instability by [68]. Later on, several other mutations that affect splicing nonsense mediate decay (NMD) or to the synthesis of were described in BRCA1 and BRCA2 contributing to a better truncated or dysfunctional protein products. Despite their understanding of the mechanism underlying the role of these potential functional relevance, characterization of splicing- proteins in tumorigenesis [65, 66]. More recently, Gaildrat affecting mutations, in general, and in DDR genes, in particu- et al. used a BRCA2 minigene reporter system to study the lar, is not extensive. This may be in part due to historical rea- effect of predicted splice-site mutation and some unclassified sons and also to technical issues related to the experimental mutations occurring at a distance from the splicing sites. 6 International Journal of Cell Biology

Table 1: List of DDR-related genes found to be aberrantly spliced in several cancer types.

Gene Function Cancer References An E3 ubiquitin ligase contained in several cellular BRCA1 complexes, involved in DNA repair, genome stability Breast and ovarian cancer [62, 64–67] maintenance, and cell cycle checkpoint control. Breast and ovarian cancer BRCA2 Involved in HR, it associates with RAD51 [64, 67–70] Familial pancreatic cancer ∗ Ataxia-telangiectasia Hereditary breast and ovarian cancer Apical kinase of DDR response, ATM Mantle cell lymphoma [71–74] mainly involved in HR Colon tumor derived cell lines Leukemia-lymphoma-derived cell lines Mismatch repair deficient colorectal cancer MRE11 Component of DNA damage sensor complex MRN Leukemia-lymphoma and colorectal [73, 75] cancer-derived cell lines ∗ Seckel syndrome ATR Apical kinase of DDR response, mainly involved in HR Hodgkin’s lymphoma [76–79] Breast and ovarian cancer ∗ XPA Nucleotide excision repair Xeroderma pigmentosum [80] Apical kinase of DDR response, mainly involved in ∗ DNAPK Xeroderma pigmentosum [81] NHEJ MSH2 and MLH1 Mismatch repair Hereditary nonpolyposis colorectal cancer [82–84] Breast cancer CHEK2 DNA damage checkpoint kinase [85, 86] Li-Fraumeni syndrome ∗ HR: homologous recombination; NHEJ: nonhomologous end joining. ataxia-telangiectasia, xeroderma pigmentosum, and Seckel syndrome were included because they display strong predisposition to malignancies.

The study identified a group of mutations that induced the The aberrant inclusion of a pseudoexon of 65 nt located aberrant splicing of exon 7 [69]. in intron 20 (termed here exon 20A) was reported more In recent years, several groups have proposed a combined than 10 years ago in a patient affected by AT [92]. The approach to study the potential effects of BRCA1/2 genetic inclusion of exon 20A was associated with a deletion of four variants on splicing efficiency64 [ , 67]. These strategies nucleotides (GTAA) in intron 20. The authors identified a exploit splicing prediction programs to detect potential novel regulatory element within intron 20, termed “intron- splicing alterations followed by functional assays to analyze splicing processing element” (ISPE), which acts as an intronic 󸀠 BRCA1/2 unclassified mutations. Using this approach Sanz silencer of a cryptic 3 splice site [93–95]. The ISPE is and colleagues identified 57 putative splicing-affecting muta- recognized by the U1 snRNP,a core component of the splicing 󸀠 tions. However, an effect on splicing could be confirmed by apparatus that normally binds the 5 splice site. Sequestration functional analysis only for half of the tested mutations [64]. of U1snRNP by the ISPE prevents inclusion of exon 20A. This low rate of correlation could be explained by the high However, in the AT patient the four nucleotides deletion falsepositiveoutputsofthealgorithmpredictingESEandESS impairs U1 snRNP binding to the regulatory element, causing mutations. Therefore, the results of this study and of other the inclusion of the cryptic exon [94]. similar reports underlie the importance of a functional validation of bioinformatic predictions [64, 67]. Nevertheless, 3.2. Same Gene, Two Isoforms with Quite Different Functions. these studies support the reclassification of many UVs as Cyclin D1 was first characterized as a cell cycle regulator. splicing affecting mutations. Cyclin D1 associates with CDK4/6 promoting the phosphorylation of Rb; this last event leads to the derepression of E2F, a transcription factor that controls the expression of 3.1.2. ATM: The Intronic Case. The ataxia-telangiectasia DNA replication genes, allowing cell cycle progression (for mutated (ATM) (OMIM 607585) gene, a key player in the review see [96]).Inaddition,cyclinD1hasatranscriptional DDR, is mutated in the autosomal recessive disorder ataxia- regulatory activity that is independent of its association with telangiectasia (AT, OMIN number 208900). The gene was CDK4/6 [96]. identified in 1995 by positional cloning; it encodes 66 exons TwoASvariantsofcyclinD1havebeendescribed,called spanning ∼160 kb of genomic DNA [89]. It soon became clear D1a and D1b [97]. The respective protein isoforms display that a significant proportion of the known mutations in the different intracellular localization: while D1a shuttles between ATM gene causes splicing defects [89–91]. Here, we will focus thenucleusandthecytoplasm,D1bisconstitutivelynuclear our attention on a specific aberrant splicing event that illus- [98]. Both transcript variants are expressed in normal tissues trates an additional layer of complexity in splicing regulation. but D1b often appears upregulated in several forms of cancer, International Journal of Cell Biology 7 including breast cancer [99, 100]. The oncogenic properties embryonic fibroblasts, loss of SC35 (SFSR2) resulted in G2/M of D1b isoform have been extensively described [98, 101]. cell-cycle arrest and genomic instability [111]. Indeed, not Instead, cyclin D1a has directly been related to the DDR only the siRNA-mediated silencing of splicing factors but also [102]. Recruitment of cyclin D1a, but not of D1b, to chromatin the genetic depletion of a transcriptional coactivator, such as is sufficient to activate the DDR promoting H2AX phos- SKIIP, besides affecting splicing, induced genomic instability phorylation. Moreover, after genotoxic stress D1a enhances [112]. Interestingly, overexpression of RNAse H, which cleaves the recruitment of repair factors contributing to checkpoint the RNA molecule in a RNA-DNA hybrid, reduced the for- activation and G2/M arrest [102]. A recent report by Myk- mation of H2AX foci [112]. Although the precise mechanism lebust and colleagues identifies high expression of cyclin through which the formation of R-loops results in genomic D1a protein as a positive predictive factor for the benefit of instability is still unclear, these structures have recently been adjuvant chemotherapy with levamisole and 5-fluorouracil demonstrated to impair the replication fork progression [113]. (5-FU), which leads to replication stress, due to the depletion Finally, in the last few years numerous mutations affecting of the intracellular deoxythymidine triphosphate (dTTP) components of the splicing machinery have been reported pool in colorectal cancer [103]. This last report underlies the in several types of cancer [114–119]. The functional con- importance to link expression patterns of splicing isoforms sequences of these mutations, and how they contribute to with therapeutic approaches. tumorigenesis not fully understood. However considering Two recent reports show that the splicing factors Sam68 the genotoxic consequences of splicing factor depletion, and ASF/SF2 regulate cyclin D1 splicing favoring the D1b discussed previously, it is possible that at least some of these isoform [104, 105]. In addition, transcriptional regulation mutations may result in the functional inactivation of the also affects the cyclin D1a/D1b ratio. Sanchez and colleagues splicing factors thus affecting genomic stability. reported that the EWS-FL11 fusion protein, a well characterized transcription factor, is able to favor the expression of the D1b isoform by decreasing the rate of polymerase 3.4. Splicing-Related Proteins as Novel Factors in the DDR. II elongation [106]. Additionally, a polymorphism at the The results of several recent genome-wide studies strongly 󸀠 3 splice site of intron 4 may influence splice site choice suggest an overlap between the pathways leading to mRNA favoring the production of the D1b isoform [97]. All these biogenesis and the cellular response to DNA damage. Some mechanisms show that alternative splicing of cyclin D1 is years ago, Matsuoka and colleagues performed a large-scale under a tight control making it a very interesting target for proteomic analysis that identified more than 700 proteins the development of new therapeutic strategies. phosphorylated by ATM and ATR in response to DNA dam- Another quite interesting example of AS variants was age. Interestingly, RNA processing was among the enriched recently described by Pabla and colleagues [107]. CHK1-S is a Gene Ontology categories that had not been previously linked novel splice variant of CHK1, which functions as endogenous to the DDR [120]. The involvement of RNA maturation in regulatorofCHK1inbothphysiologicalconditionsandafter the cellular response to DNA damage has been subsequently DNA damage. The authors demonstrated that in normal supported by various siRNA-based screens to identify genes conditions CHK1-S is able to interact with CHK1 inhibiting its involvedinDNAdamagesensitivityandgenomestabilization activity and promoting the S to G2/M transition. Upon DDR [112, 121]. In particular, two genome-wide shRNA screens activation, CHK1 becomes phosphorylated and the CHK1- identified the Ewing sarcoma (EWS) protein, a member of the S/CHK1 interaction is disrupted, so that CHK1 can induce cell TET (TLS/EWS/TAF15) family of RNA- and DNA-binding cycle arrest facilitating DNA repair. Interestingly, the authors proteins, as necessary for resistance to camptothecin [122] reported a deregulated expression of CHK1-S in testicular andIR[121]. Consistent with a role in the DDR, EWS knock- and ovarian cancer. How the CHK1-S expression is regulated out mice show hypersensitivity to ionizing radiation [123]. remains to be explored. Recently, Paronetto et al. showed that the Ewing sarcoma (EWS) protein, a member of the TET (TLS/EWS/TAF15) family of RNA- and DNA-binding proteins, regulates AS in 3.3. Loss of Splicing Factors and Genomic Instability. A response to DNA damage [124]. growing body of evidence suggests that the depletion of Using a similar approach to identify novel genes con- splicing factors may induce genomic instability. It is by now tributing to telomeres protection Lackner and colleagues well established that transcription and RNA processing are observed that silencing of several splicing-related genes tightly coupled processes [13, 14]. This, on one hand, favors (including SKIIP and SF3A1) induced a general activation accurate and efficient mRNA processing, and on one hand, of the DDR, possibly because of R-loop formation while protects the genome from the likely disastrous effects of havingaweakeffectontelomerestability[125]. More recently, the nascent transcripts themselves [108]. Accordingly, in S. RBMX, an hnRNP that associates with the spliceosome and cerevisiae, when genes involved in mRNA processing are influences alternative splicing [126], was found in a genome- mutated, defects occur in the packaging of nascent mRNAs wide siRNA-based screen to detect regulators of homologous [109]; as a result the nascent pre-mRNA hybridizes with the recombination (HR) regulators. RBMX regulates HR in transcribed strand generating an RNA-DNA duplex, known a positive manner, accumulates at sites of DNA damage as R-loop, causing genomic instability. Analogous effects in a poly(ADP-ribose) polymerase 1- (PARP1-) dependent have been observed in chicken DT40 cells upon silencing manner and promotes resistance to several DNA damaging of the splicing factor ASF/SF2 [108, 110]. Similarly, in mouse agents [127]. As discussed in Section 2.2,Beliandcolleagues 8 International Journal of Cell Biology identified the protein phosphatase PPM1G, that regulates been demonstrated to regulate chemoresistance [103], it is spliceosome activity, using high resolution mass spectrome- reasonable that splicing modulation has been proposed as an try combined with stable isotope labeling with amino acids appealing therapeutic target [135, 136]. Strategies to modulate in cell culture (SILAC) to quantify regulated changes in AS by antisense oligonucleotides are already in advanced phosphorylation and acetylation induced by different DNA- clinical trial phases for some neuromuscular disorders, such damaging agents [28]. In addition, they detected the phos- asDuchennemusculardystrophyorspinalmuscularatrophy, phorylation of THRAP3, a protein involved in RNA process- and oligonucleotides are being developed to target specific ing and stability [128, 129]. Interestingly, while many DDR mRNA variants to enhance the efficacy of conventional factors are recruited to H2AX foci, THRAP3 is excluded from chemotherapy [137]. In addition, in recent years several sites of DNA damage in a manner that parallels transcrip- bacterial compounds, and other small molecules have been tion inhibition. Two THRAP3-binding proteins BCLAF1 identifiedthattargetspliceosomalcomponents[138]. Inter- and PNN [130], that have also been implicated in mRNA estingly, some of these compounds display strong cytostatic metabolism, behaved in a similar manner suggesting that effects and significant antitumor activity in animal models. they are part of a novel protein complex, which may link RNA Further understanding of how AS regulation and the DDR splicing to the DDR. A comprehensive review, that appeared are interconnected and linked to different signal transduction while this paper was under revision, describes in detail some pathways should help us to better understand tumor proof these novel RNA-binding proteins involved in DDR [9]. gression and provide a basis for innovative splicing-targeted cancer therapies.

4. Summary and Future Perspectives References

Genomic instability is a hallmark of cancer cells, and the [1] A. Ciccia and S. J. Elledge, “The DNA damage response: making understanding of the mechanisms able to limit or counteract it safe to play with knives,” Molecular Cell,vol.40,no.2,pp.179– it can positively impact on the therapy of tumors. Genome- 204, 2010. wide approaches have revealed that genes involved in RNA [2] M. P. Longhese, D. Bonetti, N. Manfrini, and M. Clerici, processing are often deregulated in response to genotoxic “Mechanisms and regulation of DNA end resection,” EMBO treatments. Since many chemotherapeutic compounds are Journal,vol.29,no.17,pp.2864–2874,2010. DNA-damaging agents, AS can be an important determinant [3] P. Bouwman and J. Jonkers, “The effects of deregulated DNA of how tumor cells respond to therapy. damage signalling on cancer chemotherapy response and resis- Pre-mRNA splicing is a crucial step in the control of gene tance,” Nature Reviews Cancer,vol.12,no.9,pp.587–598,2012. expression. The activity of splicing factors must therefore [4] C. J. Lord and A. Ashworth, “The DNA damage response and cancer therapy,” Nature,vol.481,no.7381,pp.287–294,2012. be tightly regulated since both their depletion and their [5] S. E. Polo and S. P. Jackson, “Dynamics of DNA damage upregulation can have harmful consequences. On one hand, response proteins at DNA breaks: a focus on protein modifica- deregulation of splicing factors may affect AS leading to the tions,” Genes and Development,vol.25,no.5,pp.409–433,2011. generation of cancer driving transcripts [131, 132]. On the [6] D. A. P. Rockx, R. Mason, A. Van Hoffen et al., “UV-induced other hand, depletion of splicing factors may induce aberrant inhibition of transcription involves repression of transcription splicing of critical DDR effectors altering indirectly the initiation and phosphorylation of RNA polymerase II,” Proceed- cellular response to DNA damage [133]. Moreover, splicing ings of the National Academy of Sciences of the United States of factors’ depletion may slow down intron removal favoring America,vol.97,no.19,pp.10503–10508,2000. the formation of DNA/RNA hybrid thereby leading to the [7] N. M. Shanbhag, I. U. Rafalska-Metcalf, C. Balane-Bolivar, S. collapse of replication forks and to the generation of DSBs M. Janicki, and R. A. Greenberg, “ATM-Dependent chromatin [108, 112, 113]. Finally, the activation of the DDR can promote changes silence transcription in cis to dna double-strand the posttranslational modification of splicing factors altering breaks,” Cell,vol.141,no.6,pp.970–981,2010. their intracellular localization and/or their activity [22, 23, [8] M. Dutertre, G. Sanchez, J. Barbier, L. Corcos, and D. Auboeuf, 41, 46, 134]. Although some aspects of the relationship “The emerging role of pre-messenger RNA splicing in stress between DDR and mRNA processing have been clarified, responses: sending alternative messages and silent messengers,” certain observations are still to be explained. Why are some RNA Biology, vol. 8, no. 5, 2011. splicing factors recruited to sites of DNA damage? What [9] A. Montecucco and G. Biamonti, “Pre-mRNA processing fac- are their functions there? Moreover, when splicing changes tors meet the DNA damage response,” Frontiers in Genetics,vol. 4, p. 102, 2013. are considered, a causal link between splicing alteration and [10] R. J. Katzenberger, M. S. Marengo, and D. A. Wassarman, “ATM disease occurrence can only be established if the stability of and ATR pathways signal alternative splicing of Drosophila the mRNA variant and the function of the encoded protein TAF1 pre-mRNA in response to DNA damage,” Molecular and have been determined. Finally, the possibility should be Cellular Biology, vol. 26, no. 24, pp. 9256–9267, 2006. considered that an involvement of splicing factors in the [11] R. J. Katzenberger, M. S. Marengo, and D. A. Wassarman, “Con- DDR may not necessarily imply an involvement of splicing trol of alternative splicing by signal-dependent degradation of regulation in the DDR but may reflect other functions of splicing-regulatory proteins,” Journal of Biological Chemistry, these proteins. Nevertheless, considering that AS plays a vol. 284, no. 16, pp. 10737–10746, 2009. major role in the regulation of the apoptotic response (see [12]D.S.Chandler,R.K.Singh,L.C.Caldwell,J.L.Bitler,and C.Sette’sreviewinthisissue)andthatASvariantshave G. Lozano, “Genotoxic stress induces coordinately regulated International Journal of Cell Biology 9

alternative splicing of the p53 modulators MDM2 and MDM4,” [29] T. Ikura, V. V. Ogryzko, M. Grigoriev et al., “Involvement of the Cancer Research,vol.66,no.19,pp.9502–9508,2006. TIP60 histone acetylase complex in DNA repair and apoptosis,” [13] A. R. Kornblihtt, M. De La Mata, J. P. Fededa, M. J. Munoz,˜ and Cell,vol.102,no.4,pp.463–473,2000. G. Nogues,´ “Multiple links between transcription and splicing,” [30] Y. Sun, X. Jiang, and B. D. Price, “Tip60: connecting chromatin RNA, vol. 10, no. 10, pp. 1489–1498, 2004. to DNA damage signaling,” Cell Cycle,vol.9,no.5,pp.930–936, [14] A. R. Kornblihtt, “Coupling transcription and alternative splic- 2010. ing,” Advances in Experimental Medicine and Biology,vol.623, [31] C. Choudhary, C. Kumar, F. Gnad et al., “Lysine acetylation pp.175–189,2007. targets protein complexes and co-regulates major cellular func- [15] Z. Zhou and X. D. Fu, “Regulation of splicing by SR proteins tions,” Science,vol.325,no.5942,pp.834–840,2009. and SR protein-specific kinases,” Chromosoma,vol.122,no.3, [32] G. Buchwald, P. Van Der Stoop, O. Weichenrieder, A. Perrakis, pp.191–207,2013. M. Van Lohuizen, and T. K. Sixma, “Structure and E3-ligase [16] R. Martinez-Contreras, P.Cloutier, L. Shkreta, J. Fisette, T.Revil, activity of the Ring-Ring complex of Polycomb proteins Bmi1 and B. Chabot, “hnRNP proteins and splicing control,” Advances and Ring1b,” EMBO Journal, vol. 25, no. 11, pp. 2465–2474, 2006. in Experimental Medicine and Biology,vol.623,pp.123–147, [33] R. Cao, Y. Tsukada, and Y. Zhang, “Role of Bmi-1 and Ring1A 2007. in H2A ubiquitylation and hox gene silencing,” Molecular Cell, [17] B. Haley, T. Paunesku, M. Protic, and G. E. Woloschak, vol. 20, no. 6, pp. 845–854, 2005. “Response of heterogeneous ribonuclear proteins (hnRNP) [34] C. M. Guzzo, C. E. Berndsen, J. Zhu et al., “RNF4-dependent to ionising radiation and their involvement in DNA damage hybrid SUMO-ubiquitin chains are signals for RAP80 and repair,” International Journal of Radiation Biology,vol.85,no. thereby mediate the recruitment of BRCA1 to sites of DNA 8, pp. 643–655, 2009. damage,” Science Signaling, vol. 5, no. 253, Article ID ra88, 2012. [18] V.Filippov, M. Filippova, and P.J. Duerksen-Hughes, “The early [35] R. Vyas, R. Kumar, F. Clermont et al., “RNF4 is required response to DNA damage can lead to activation of alternative for DNA double-strand break repair in vivo,” Cell Death & splicing activity resulting in CD44 splice pattern changes,” Differentiation,vol.20,no.3,pp.490–502,2013. Cancer Research,vol.67,no.16,pp.7621–7630,2007. [36] Z. Yin, D. Menendez, M. A. Resnick, J. E. French, K. S. [19] G. Merdzhanova, V. Edmond, S. De Seranno et al., “E2F1 con- Janardhan, and A. M. Jetten, “RAP80 is critical in maintaining trols alternative splicing pattern of genes involved in apoptosis genomic stability and suppressing tumor development,” Cancer through upregulation of the splicing factor SC35,” Cell Death Research,vol.72,no.19,pp.5080–5090,2012. and Differentiation, vol. 15, no. 12, pp. 1815–1823, 2008. [37] P. Bellare, E. C. Small, X. Huang, J. A. Wohlschlegel, J. P. Staley, [20] Y. Auclair and S. Richard, “The role of arginine methylation in and E. J. Sontheimer, “A role for ubiquitin in the spliceosome the DNA damage response,” DNA Repair,vol.12,no.7,pp.459– assembly pathway,” Nature Structural & Molecular Biology,vol. 465, 2013. 15,no.5,pp.444–451,2008. [21] L. Shi and P. Oberdoerffer, “Chromatin dynamics in DNA [38]M.D.Ohi,C.W.VanderKooi,J.A.Rosenberg,W.J.Chazin, double-strand break repair,” Biochimica et Biophysica Acta,vol. and K. L. Gould, “Structural insights into the U-box, a domain 1819,no.7,pp.811–819,2012. associated with multi-ubiquitination,” Nature Structural Biol- [22] S. Vivarelli, S. C. Lenzken, M. D. Ruepp MD et al., “Paraquat ogy,vol.10,no.4,pp.250–255,2003. modulates alternative pre-mRNA splicing by modifying the [39] K. N. Mahajan and B. S. Mitchell, “Role of human Pso4 in mam- intracellular distribution of SRPK2,” PLoS One,vol.8,no.4, malian DNA repair and association with terminal deoxynu- Article ID e61980, 2013. cleotidyl transferase,” Proceedings of the National Academy of [23] V. Edmond, E. Moysan, S. Khochbin et al., “Acetylation and Sciences of the United States of America,vol.100,no.19,pp. phosphorylation of SRSF2 control cell fate decision in response 10746–10751, 2003. to cisplatin,” EMBO Journal,vol.30,no.3,pp.510–523,2011. [40]M.C.Wahl,C.L.Will,andR.Luhrmann,¨ “The spliceosome: [24]V.Leva,S.Giuliano,A.Bardonietal.,“Phosphorylationof design principles of a dynamic RNP machine,” Cell,vol.136,no. SRSF1 is modulated by replicational stress,” Nucleic Acids 4, pp. 701–718, 2009. Research, vol. 40, no. 3, pp. 1106–1117, 2012. [41]X.LuandR.J.Legerski,“ThePrp19/Pso4corecomplexunder- [25] J. E. Mermoud, P. Cohen, and A. I. Lamond, “Ser/Thr-specific goes ubiquitylation and structural alterations in response to proteinphosphatasesarerequiredforbothcatalyticstepsof DNA damage,” Biochemical and Biophysical Research Commu- pre-mRNA splicing,” Nucleic Acids Research, vol. 20, no. 20, pp. nications,vol.354,no.4,pp.968–974,2007. 5263–5269, 1992. [42] S. P. Jackson and D. Durocher, “Regulation of DNA damage [26] M. V. Murray, R. Kobayashi, and A. R. Krainer, “The type 2C responses by ubiquitin and SUMO,” Molecular Cell,vol.49,no. Ser/Thr phosphatase PP2C𝛾 is a pre-mRNA splicing factor,” 5, pp. 795–807, 2013. Genes and Development,vol.13,no.1,pp.87–97,1999. [43] F. Pelisch, J. Gerez, J. Druker et al., “The serine/arginine-rich [27]E.Allemand,M.L.Hastings,M.V.Murray,M.P.Myers,and protein SF2/ASF regulates protein sumoylation,” Proceedings of A. R. Krainer, “Alternative splicing regulation by interaction the National Academy of Sciences of the United States of America, of phosphatase PP2Cgamma with nucleic acid-binding protein vol. 107, no. 37, pp. 16119–16124, 2010. YB-1,” Nature Structural & Molecular Biology,vol.14,no.7,pp. [44] M. T. Vassileva and M. J. Matunis, “SUMO modification 630–638, 2007. of heterogeneous nuclear ribonucleoproteins,” Molecular and [28] P. Beli, N. Lukashchuk, S. A. Wagner et al., “Proteomic investi- Cellular Biology,vol.24,no.9,pp.3623–3632,2004. gations reveal a role for RNA processing factor THRAP3 in the [45] K. Bomsztyk, O. Denisenko, and J. Ostrowski, “hnRNP K: one DNAdamageresponse,”Molecular Cell,vol.46,no.2,pp.212– protein multiple processes,” BioEssays,vol.26,no.6,pp.629– 225, 2012. 638, 2004. 10 International Journal of Cell Biology

[46]F.Pelisch,B.Pozzi,G.Risso,M.J.Munoz,andA.Srebrow, [63] L. Cartegni, S. L. Chew, and A. R. Krainer, “Listening to “DNA damage-induced heterogeneous nuclear ribonucleopro- silence and understanding nonsense: exonic mutations that tein K sumoylation regulates p53 transcriptional activation,” affect splicing,” Nature Reviews Genetics,vol.3,no.4,pp.285– Journal of Biological Chemistry,vol.287,no.36,pp.30789– 298, 2002. 30799, 2012. [64] D. J. Sanz, A. Acedo, M. Infante et al., “A high proportion [47]A.Moumen,P.Masterson,M.J.O’Connor,andS.P.Jackson, of DNA variants of BRCA1 and BRCA2 is associated with “hnRNP K: an HDM2 target and transcriptional coactivator of aberrant splicing in breast/ovarian cancer patients,” Clinical p53 in response to DNA damage,” Cell,vol.123,no.6,pp.1065– Cancer Research,vol.16,no.6,pp.1957–1967,2010. 1078, 2005. [65] E. Rouleau, C. Lefol, V. Moncoutier et al., “A missense variant [48]D.WangandS.J.Lippard,“Cellularprocessingofplatinum within BRCA1 exon 23 causing exon skipping,” Cancer Genetics anticancer drugs,” Nature Reviews Drug Discovery,vol.4,no.4, and Cytogenetics,vol.202,no.2,pp.144–146,2010. pp. 307–320, 2005. [49] I. Collins, A. Weber, and D. Levens, “Transcriptional conse- [66] M. Raponi, J. Kralovicova, E. Copson et al., “Prediction of quences of topoisomerase inhibition,” Molecular and Cellular single-nucleotide substitutions that result in exon skipping: Biology,vol.21,no.24,pp.8437–8451,2001. identification of a splicing silencer in BRCA1 exon 6,” Human Mutation,vol.32,no.4,pp.436–444,2011. [50]R.F.Luco,M.Allo,I.E.Schor,A.R.Kornblihtt,andT.Misteli, “Epigenetics in alternative pre-mRNA splicing,” Cell,vol.144, [67] J. C. Thery,´ S. Krieger, P. Gaildrat et al., “Contribution of no. 1, pp. 16–26, 2011. bioinformatics predictions and functional splicing assays to [51] E. Allemand, E. Batsche,´ and C. Muchardt, “Splicing, transcrip- theinterpretationofunclassifiedvariantsoftheBRCAgenes,” tion, and chromatin: a menage a trois,” Current Opinion in European Journal of Human Genetics,vol.19,no.10,pp.1052– Genetics and Development,vol.18,no.2,pp.145–151,2008. 1058, 2011. [52] M. de la Mata and A. R. Kornblihtt, “RNA polymerase II C- [68] J. D. Fackenthal, L. Cartegni, A. R. Krainer, and O. I. Olopade, terminal domain mediates regulation of alternative splicing by “BRCA2 T2722R is a deleterious allele that causes exon skip- SRp20,” Nature Structural and Molecular Biology,vol.13,no.11, ping,” American Journal of Human Genetics,vol.71,no.3,pp. pp.973–980,2006. 625–631, 2002. [53] M. J. Munoz,˜ M. S. P. Santangelo, M. P. Paronetto et al., “DNA [69] P. Gaildrat, S. Krieger, D. Di Giacomo et al., “Multiple sequence damage regulates alternative splicing through inhibition of variants of BRCA2 exon 7 alter splicing regulation,” Journal of RNA polymerase II elongation,” Cell,vol.137,no.4,pp.708–720, Medical Genetics,vol.49,no.10,pp.609–617,2012. 2009. [70]K.M.Murphy,K.A.Brune,C.Griffinetal.,“Evaluationof [54] J. Y. Ip, D. Schmidt, Q. Pan et al., “Global impact of RNA candidate genes MAP2K4, MADH4, ACVR1B, and BRCA2 in polymerase II elongation inhibition on alternative splicing familial pancreatic cancer: deleterious BRCA2 mutations in regulation,” Genome Research, vol. 21, no. 3, pp. 390–401, 2011. 17%,” Cancer Research, vol. 62, no. 13, pp. 3789–3793, 2002. [55] S. Solier, J. Barb, B. R. Zeeberg et al., “Genome-wide analysis of novel splice variants induced by topoisomerase I poisoning [71]N.Y.Fang,T.C.Greiner,D.D.Weisenburgeretal.,“Oligonu- shows preferential occurrence in genes encoding splicing fac- cleotide microarrays demonstrate the highest frequency of tors,” Cancer Research, vol. 70, no. 20, pp. 8055–8065, 2010. ATM mutations in the mantle cell subtype of lymphoma,” [56]M.Dutertre,G.Sanchez,M.C.DeCianetal.,“Cotranscrip- Proceedings of the National Academy of Sciences of the United tional exon skipping in the genotoxic stress response,” Nature States of America,vol.100,no.9,pp.5372–5377,2003. Structural & Molecular Biology,vol.17,no.11,pp.1358–1366, [72] Y. Ejima, L. Yang, and M. S. Sasaki, “Aberrant splicing of 2010. the ATM gene associated with shortening of the intronic [57]C.S.Dias,Y.Liu,A.Yau,L.Westrick,andS.C.Evans, mononucleotide tract in human colon tumor cell lines: a alt1 “Regulation of hdm2 by stress-induced hdm2 in tumor and novel mutation target of microsatellite instability,” International nontumorigenic cell lines correlating with p53 stability,” Cancer Journal of Cancer, vol. 86, no. 2, pp. 262–268, 2000. Research,vol.66,no.19,pp.9467–9473,2006. [73]M.F.Ham,T.Takakuwa,W.Luo,A.Liu,A.Horii,and [58] H.A.Chansky,M.Hu,D.D.Hickstein,andL.Yang,“Oncogenic K. Aozasa, “Impairment of double-strand breaks repair and TLS/ERG and EWS/Fli-1 fusion proteins inhibit RNA splicing aberrant splicing of ATM and MRE11 in leukemia-lymphoma mediated by YB-1 protein,” Cancer Research,vol.61,no.9,pp. cell lines with microsatellite instability,” Cancer Science,vol.97, 3586–3590, 2001. no.3,pp.226–234,2006. [59]C.N.Sprung,J.Li,D.Hovan,M.J.McKay,andH.B.Forrester, [74]Y.R.Thorstenson,A.Roxas,R.Kroissetal.,“Contributionsof “Alternative transcript initiation and splicing as a response to ATM mutations to familial breast and ovarian cancer,” Cancer DNA damage,” PloS One,vol.6,no.10,ArticleIDe25758,2011. Research, vol. 63, no. 12, pp. 3325–3333, 2003. [60]Z.WangandC.B.Burge,“Splicingregulation:fromapartslist [75] G. Giannini, E. Ristori, F. Cerignoli et al., “Human MRE11 of regulatory elements to an integrated splicing code,” RNA,vol. is inactivated in mismatch repair-deficient cancers,” EMBO 14,no.5,pp.802–813,2008. Reports,vol.3,no.3,pp.248–254,2002. [61]Q.Pan,O.Shai,L.J.Lee,B.J.Frey,andB.J.Blencowe,“Deep surveying of alternative splicing complexity in the human tran- [76]F.Durocher,Y.Labrie,P.Soucyetal.,“Mutationanalysisand scriptome by high-throughput sequencing,” Nature Genetics, characterization of ATR sequence variants in breast cancer vol. 40, no. 12, pp. 1413–1415, 2008. cases from high-risk French Canadian breast/ovarian cancer [62] S. Mazoyer, N. Puget, L. Perrin-Vidoz, H. T. Lynch, O. M. families,” BMC Cancer,vol.6,p.230,2006. Serova-Sinilnikova, and G. M. Lenoir, “A BRCA1 nonsense [77] A. Liu, T. Takakuwa, S. Fujita et al., “ATR alterations in mutation causes exon skipping,” American Journal of Human Hodgkin’s lymphoma,” Oncology Reports,vol.19,no.4,pp.999– Genetics, vol. 62, no. 3, pp. 713–715, 1998. 1005, 2008. International Journal of Cell Biology 11

[78] H. Mokrani-Benhelli, L. Gaillard, P. Biasutto et al., “Primary [94]A.Dhir,E.Buratti,M.A.VanSanten,R.Luhrmann,¨ and F. E. microcephaly, impaired DNA replication, and genomic instabil- Baralle, “The intronic splicing code: multiple factors involved in itycausedbycompoundheterozygousATRmutations,”Human ATM pseudoexon definition,” EMBO Journal,vol.29,no.4,pp. Mutation, vol. 34, no. 2, pp. 374–384, 2013. 749–760, 2010. [79]M.O’Driscoll,V.L.Ruiz-Perez,C.G.Woods,P.A.Jeggo,and [95] M. A. Lewandowska, C. Stuani, A. Parvizpur, F.E. Baralle, and F. J. A. Goodship, “A splicing mutation affecting expression of Pagani, “Functional studies on the ATM intronic splicing proc- ataxia-telangiectasia and Rad3-related protein (ATR) results in essing element,” Nucleic Acids Research,vol.33,no.13,pp.4007– Seckel syndrome,” Nature Genetics,vol.33,no.4,pp.497–501, 4015, 2005. 2003. [96] K. E. Knudsen, J. Alan Diehl, C. A. Haiman, and E. S. Knudsen, [80] J. E. Cleaver, L. H. Thompson, A. S. Richardson, and J. C. States, “Cyclin D1: polymorphism, aberrant splicing and cancer risk,” “A summary of mutations in the UV-sensitive disorders: xero- Oncogene,vol.25,no.11,pp.1620–1628,2006. derma pigmentosum, Cockayne syndrome, and trichothiodys- [97] D. C. Betticher, N. Thatcher, H. J. Altermatt, P. Hoban, W. D. trophy,” Human Mutation,vol.14,no.1,pp.9–22,1999. J. Ryder, and J. Heighway, “Alternate splicing produces a novel [81] F. Abbaszadeh, P. H. Clingen, C. F. Arlett et al., “A novel splice cyclin D1 transcript,” Oncogene,vol.11,no.5,pp.1005–1011,1995. variant of the DNA-PKcs gene is associated with clinical and [98] F. Lu, A. B. Gladden, and J. A. Diehl, “An Alternatively Spliced cellular radiosensitivity in a patient with xeroderma pigmen- Cyclin D1 Isoform, Cyclin D1b, Is a Nuclear Oncogene,” Cancer tosum,” Journal of Medical Genetics,vol.47,no.3,pp.176–181, Research,vol.63,no.21,pp.7056–7061,2003. 2010. [99]Y.Wang,J.L.Dean,E.K.A.Millaretal.,“CyclinD1bis [82] B. Betz, S. Theiss, M. Aktas et al., “Comparative in silico analyses aberrantly regulated in response to therapeutic challenge and and experimental validation of novel splice site and missense promotes resistance to estrogen antagonists,” Cancer Research, mutations in the genes MLH1 and MSH2,” Journal of Cancer vol.68,no.14,pp.5628–5638,2008. Research and Clinical Oncology,vol.136,no.1,pp.123–134,2010. [100] C. J. Burd, C. E. Petre, L. M. Morey et al., “Cyclin D1b variant [83] C. Pagenstecher, M. Wehner, W. Friedl et al., “Aberrant splicing influences prostate cancer growth through aberrant androgen in MLH1 and MSH2 due to exonic and intronic variants,” receptor regulation,” Proceedings of the National Academy of Human Genetics,vol.119,no.1-2,pp.9–22,2006. Sciences of the United States of America,vol.103,no.7,pp.2190– [84]E.Renkonen,H.Lohi,H.J.Jarvinen,¨ J. Mecklin, and P. 2195, 2006. Peltomaki,¨ “Novel splicing associations of hereditary colon cancer related DNA mismatch repair gene mutations,” Journal [101]D.A.Solomon,Y.Wang,S.R.Foxetal.,“CyclinD1splicevari- of Medical Genetics,vol.41,no.7,p.e95,2004. ants: differential effects on localization, RB phosphorylation, and cellular transformation,” Journal of Biological Chemistry, [85] V. Staalesen, J. Falck, S. Geisler et al., “Alternative splicing and vol. 278, no. 32, pp. 30339–30347, 2003. mutation status of CHEK2 in stage III breast cancer,” Oncogene, vol.23,no.52,pp.8535–8544,2004. [102]Z.Li,X.Jiao,C.Wangetal.,“AlternativecyclinD1splice forms differentially regulate the DNA damage response,” Cancer [86]P.Bellare,A.K.Kutach,A.K.Rines,C.Guthrie,andE.J. Research,vol.70,no.21,pp.8802–8811,2010. Sontheimer, “Ubiquitin binding by a variant Jab1/MPN domain in the essential pre-mRNA splicing factor Prp8p,” RNA,vol.12, [103] M.P.Myklebust,Z.Li,T.H.Tranetal.,“ExpressionofcyclinD1a no. 2, pp. 292–302, 2006. and D1b as predictive factors for treatment response in colo- [87] M. R. Stratton and N. Rahman, “The emerging landscape of rectal cancer,” British Journal of Cancer,vol.107,no.10,pp.1684– breast cancer susceptibility,” Nature Genetics,vol.40,no.1,pp. 1691, 2012. 17–22, 2008. [104]N.A.Olshavsky,C.E.S.Comstock,M.J.Schieweretal., [88]H.X.Liu,L.Cartegni,M.Q.Zhang,andA.R.Krainer,“A “Identification of ASF/SF2 as a critical, allele-specific effector of mechanism for exon skipping caused by nonsense or missense the cyclin D1b oncogene,” Cancer Research,vol.70,no.10,pp. mutations in BRCA1 and other genes,” Nature Genetics,vol.27, 3975–3984, 2010. no. 1, pp. 55–58, 2001. [105]M.P.Paronetto,M.Cappellari,andR.Busa,` “Alternative [89] P. Concannon and R. A. Gatti, “Diversity of ATM gene muta- splicing of the cyclin D1 proto-oncogene is regulated by the tions detected in patients with ataxia-telangiectasia,” Human RNA-binding protein Sam68,” Cancer Research,vol.70,no.1, Mutation, vol. 10, no. 2, pp. 100–107, 1997. pp. 229–239, 2010. [90] S. N. Teraoka, M. Telatar, S. Becker-Catania et al., “Splicing [106]G.Sanchez,D.Bittencourt,K.Laudetal.,“Alterationof defects the ataxia-telangiectasia gene, ATM: underlying muta- cyclin D1 transcript elongation by a mutated transcription tions and consequences,” AmericanJournalofHumanGenetics, factor up-regulates the oncogenic D1b splice isoform in cancer,” vol.64,no.6,pp.1617–1631,1999. Proceedings of the National Academy of Sciences of the United [91] L. Eng, G. Coutinho, S. Nahas et al., “Nonclassical splicing States of America, vol. 105, no. 16, pp. 6004–6009, 2008. mutations in the coding and noncoding regions of the ATM [107]N.Pabla,K.Bhatt,andZ.Dong,“Checkpointkinase1(Chk1)- gene: maximum entropy estimates of splice junction strengths,” short is a splice variant and endogenous inhibitor of Chk1 that Human Mutation,vol.23,no.1,pp.67–76,2004. regulatescellcycleandDNAdamagecheckpoints,”Proceedings [92]F.Paganil,E.Burattil,C.Stuani,R.Bendix,T.Dork,¨ and F. E. of the National Academy of Sciences of the United States of Baralle, “A new type of mutation causes a splicing defect in America,vol.109,no.1,pp.197–202,2012. ATM,” Nature Genetics,vol.30,no.4,pp.426–429,2002. [108] X. Li and J. L. Manley, “Inactivation of the SR protein splicing [93]T.Pastor,G.Talotti,M.A.Lewandowska,andF.Pagani, factor ASF/SF2 results in genomic instability,” Cell,vol.122,no. “An Alu-derived intronic splicing enhancer facilitates intronic 3, pp. 365–378, 2005. processing and modulates aberrant splicing in ATM,” Nucleic [109] A. Aguilera and B. Gomez-Gonz´ alez,´ “Genome instability: Acids Research,vol.37,no.21,ArticleIDgkp778,pp.7258–7267, a mechanistic view of its causes and consequences,” Nature 2009. Reviews Genetics,vol.9,no.3,pp.204–217,2008. 12 International Journal of Cell Biology

[110] X. Li and J. L. Manley, “Cotranscriptional processes and their [127]B.Adamson,A.Smogorzewska,F.D.Sigoillot,R.W.King, influence on genome stability,” Genes and Development,vol.20, and S. J. Elledge, “A genome-wide homologous recombination no. 14, pp. 1838–1847, 2006. screen identifies the RNA-binding protein RBMX as a compo- [111] R. Xiao, Y.Sun, J. Ding et al., “Splicing regulator SC35 is essential nent of the DNA-damage response,” Nature Cell Biology,vol.14, for genomic stability and cell proliferation during mammalian no.3,pp.318–328,2012. organogenesis,” Molecular and Cellular Biology,vol.27,no.15, [128]K.M.Lee,I.W.Hsu,andW.Y.Tarn,“TRAP150activates pp. 5393–5402, 2007. pre-mRNA splicing and promotes nuclear mRNA degradation,” [112] R. D. Paulsen, D. V. Soni, R. Wollman et al., “A genome-wide Nucleic Acids Research,vol.38,no.10,ArticleIDgkq017,pp. siRNA screen reveals diverse cellular processes and pathways 3340–3350, 2010. that mediate genome stability,” Molecular Cell,vol.35,no.2,pp. [129]C.P.Bracken,S.J.Wall,B.Barre,´ K. I. Panov, P. M. Ajuh, and 228–239, 2009. N. D. Perkins, “Regulation of cyclin D1 RNA stability by SNIP1,” [113] W. Gan, Z. Guan, J. Liu et al., “R-loop-mediated genomic insta- Cancer Research, vol. 68, no. 18, pp. 7621–7628, 2008. bility is caused by impairment of replication fork progression,” [130] C.Li,R.Lin,M.Lai,P.Ouyang,andW.Tarn,“NuclearPnn/DRS Genes and Development,vol.25,no.19,pp.2041–2056,2011. protein binds to spliced mRNPs and participates in mRNA [114]E.M.Je,N.J.Yoo,Y.J.Kim,M.S.Kim,andS.H.Lee, processing and export via interaction with RNPS1,” Molecular “Mutational analysis of splicing machinery genes SF3B1, U2AF1 and Cellular Biology, vol. 23, no. 20, pp. 7363–7376, 2003. and SRSF2 in myelodysplasia and other common tumors,” [131] J. P. Venables, R. Klinck, C. Koh et al., “Cancer-associated reg- International Journal of Cancer,vol.133,no.1,pp.260–265,2013. ulation of alternative splicing,” Nature Structural and Molecular [115] J. W.Harbour, E. D. Roberson, H. Anbunathan, M. D. Onken, L. Biology, vol. 16, no. 6, pp. 670–676, 2009. A. Worley, and A. M. Bowcock, “Recurrent mutations at codon [132]R.Karni,E.DeStanchina,S.W.Lowe,R.Sinha,D.Mu,andA. 625 of the splicing factor SF3B1 in uveal melanoma,” Nature R. Krainer, “The gene encoding the splicing factor SF2/ASF isa Genetics, vol. 45, no. 2, pp. 133–135, 2013. proto-oncogene,” Nature Structural and Molecular Biology,vol. [116] O. Abdel-Wahab, T. Manshouri, J. Patel et al., “Genetic analysis 14,no.3,pp.185–193,2007. of transforming events that convert chronic myeloproliferative [133] B. Adamson, A. Smogorzewska, F. D. Sigoillot, R. W. King, neoplasms to leukemias,” Cancer Research,vol.70,no.2,pp. and S. J. Elledge, “A genome-wide homologous recombination 447–452, 2010. screen identifies the RNA-binding protein RBMX as a compo- [117] V. Quesada, L. Conde, N. Villamor et al., “Exome sequencing nent of the DNA-damage response,” Nature Cell Biology,vol.14, identifies recurrent mutations of the splicing factor SF3B1 gene no.3,pp.318–328,2012. in chronic lymphocytic leukemia,” Nature Genetics,vol.44,no. [134] R. Busa,` R. Geremia, and C. Sette, “Genotoxic stress causes the 1,pp.47–52,2012. accumulation of the splicing regulator Sam68 in nuclear foci of [118] K. Yoshida, M. Sanada, Y. Shiraishi et al., “Frequent pathway transcriptionally active chromatin,” Nucleic Acids Research,vol. mutations of splicing machinery in myelodysplasia,” Nature, 38, no. 9, Article ID gkq004, pp. 3005–3018, 2010. vol. 478, no. 7367, pp. 64–69, 2011. [135]L.DuandR.A.Gatti,“Progresstowardtherapywithantisense- [119] E. Papaemmanuil, M. Cazzola, J. Boultwood et al., “Somatic mediated splicing modulation,” Current Opinion in Molecular SF3B1 mutation in myelodysplasia with ring sideroblasts,” New Therapeutics, vol. 11, no. 2, pp. 116–123, 2009. England Journal of Medicine,vol.365,no.15,pp.1384–1395,2011. [136] F. Zammarchi, E. De Stanchina, E. Bournazou et al., “Antitu- [120]S.Matsuoka,B.A.Ballif,A.Smogorzewskaetal.,“ATM morigenic potential of STAT3 alternative splicing modulation,” and ATR substrate analysis reveals extensive protein networks Proceedings of the National Academy of Sciences of the United responsive to DNA damage,” Science,vol.316,no.5828,pp. States of America,vol.108,no.43,pp.17779–17784,2011. 1160–1166, 2007. [137] J. A. Bauman and R. Kole, “Modulation of RNA splicing as a [121]K.E.Hurov,C.Cotta-Ramusino,andS.J.Elledge,“Agenetic potential treatment for cancer,” Bioengineered Bugs,vol.2,no.3, screen identifies the Triple T complex required for DNA damage pp.125–128,2011. signaling and ATM and ATR stability,” Genes and Development, [138] S. Bonnal, L. Vigevani, and J. Valcarcel, “The spliceosome vol.24,no.17,pp.1939–1950,2010. as a target of novel antitumour drugs,” Nature Reviews Drug [122]B.C.O’Connell,B.Adamson,J.R.Lydeardetal.,“Agenome- Discovery, vol. 11, no. 11, pp. 847–859, 2012. wide camptothecin sensitivity screen identifies a mammalian MMS22L-NFKBIL2 complex required for genomic stability,” Molecular Cell,vol.40,no.4,pp.645–657,2010. [123] H. Li, W. Watford, C. Li et al., “Ewing sarcoma gene EWS is essential for meiosis and B lymphocyte development,” Journal of Clinical Investigation,vol.117,no.5,pp.1314–1323,2007. [124]M.P.Paronetto,B.Minana,˜ and J. Valcarcel,´ “The ewing sarcoma protein regulates DNA damage-induced alternative splicing,” Molecular Cell,vol.43,no.3,pp.353–368,2011. [125] D. H. Lackner, D. Durocher, and J. Karlseder, “A siRNA-based screen for genes involved in chromosome end protection,” PLoS ONE,vol.6,no.6,ArticleIDe21407,2011. [126] B. Heinrich, Z. Zhang, O. Raitskin et al., “Heterogeneous nuclear ribonucleoprotein G regulates splice site selection by binding to CC(A/C)-rich regionsin pre-mRNA,” Journal of Biological Chemistry,vol.284,no.21,pp.14303–14315,2009. Hindawi Publishing Corporation International Journal of Cell Biology Volume 2013, Article ID 463786, 6 pages http://dx.doi.org/10.1155/2013/463786

Review Article Aberrant Alternative Splicing Is Another Hallmark of Cancer

Michael Ladomery

Faculty of Health and Life Sciences, University of the West of England, Coldharbour Lane, Bristol BS16 1QY, UK

Correspondence should be addressed to Michael Ladomery; [email protected]

Received 31 May 2013; Accepted 1 August 2013

Academic Editor: Claudia Ghigna

Copyright © 2013 Michael Ladomery. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The vast majority of human genes are alternatively spliced. Not surprisingly, aberrant alternative splicing is increasingly linked to cancer. Splice isoforms often encode proteins that have distinct and even antagonistic properties. The abnormal expression of splice factors and splice factor kinases in cancer changes the alternative splicing of critically important pre-mRNAs. Aberrant alternative splicing should be added to the growing list of cancer hallmarks.

1. The Growing List of Cancer Hallmarks significant amount of intersection between the ten hallmarks. Specific cancer-associated genes can also be involved in more In the year 2000, Doug Hanahan and Bob Weinberg pub- than one hallmark. There is therefore a tension between the lished a paper in which they suggested that all cancers need to think systematically about cancer and the reality that share six common features, or hallmarks [1]. They were cancer is a remarkably complex and heterogenous disease. self-sufficiency in relation to growth signals; insensitivity to growth inhibitory signals; limitless replicative potential; the 2. Dysregulated Alternative Splicing Is ability to evade apoptosis; the ability to sustain angiogenesis; Another Key Feature of Cancer and lastly, the ability to invade tissues and metastasize. These hallmarks provided a useful framework with which In cancer, genetic lesions arise in several forms including to conceptualise cancer. The paper has been cited several chromosomal rearrangements, point mutations, and gene thousands of times as a result. Despite the benefit of the six amplifications. Genetic lesions often cause the activation of hallmarks concept, it became clear that other processes in a proto-oncogene or the inactivation of a tumour suppressor cancer are also consistently altered. This led Hanahan and gene. However, the very definition of oncogenes and tumour Weinberg to publish a follow-up review in 2011 in which suppressors is not necessarily straightforward. Several pro- they extended the cancer hallmarks to a list of ten. The teins, in different contexts, can exhibit properties of both four new hallmarks were the ability to evade the immune oncogenes and tumour suppressors. A classical example is system, the presence of inflammation, the tendency towards the Wilms tumour suppressor gene WT1.TheWT1 gene was genomic instability, and dysregulated metabolism [2]. The discovered in the early 1990s due to its association with a latter hallmark resonates with an observation made in the chromosome 11 deletion linked to WAGR syndrome (the early 20th century by Otto Warburg, namely, that cancer Wilms tumour, aniridia, genitourinary problems, and mental cells are characterised by abnormal respiration and unusually retardation). Soon after its discovery, mouse knockout studies high anaerobic metabolism [3]. This was called the “Warburg confirmed its involvement in urogenital development, and it effect” and is generally thought to be linked to the fact that was found to be inactivated in over 10% of Wilms tumours tumour cells need to adapt to hypoxic environments [4–6]. (nephroblastoma) consistent with being a classic tumour Itisundoubtedlyusefultothinkofcommonprocesses suppressor gene. However, over the years, it became apparent that apply to all cancers. The ten hallmarkssuggest theoretical that WT1 is involved in the development of several other frameworks for research and therapy. However, several addi- organ systems and that it can also be overexpressed in many tional hallmarks could be added to the list, and there is also a different types of cancer consistent with the properties ofa 2 International Journal of Cell Biology proto-oncogene [7]. WT1 function is affected by alternative allowing them to form sarcomas in nude mice [14]. SRSF1 also splicing altering its C-terminal zinc-finger domain, radically promotes the transformation of mammary epithelial cells by changing its DNA-binding properties. Thus, alternative splic- favouring the expression of antiapoptotic splice isoforms of ing complicates the biological and biochemical activities of BIM and BIN1 [15]. In fact, SRSF1 has several RNA targets WT1. Not only must its expression be examined in cancer but in pre-mRNAs that are transcribed from genes implicated the balance of its splice isoforms must also be measured. The in cancer. Thus, the overexpression of SRSF1 could lead to same principle applies to many, perhaps most, widely stud- a change in alternative splicing of several pre-mRNAs that, ied cancer-associated genes—their function is significantly together, result in a phenotype that gives the tumour a growth affected by alternative splicing. advantage. Inthe1970sand1980s,itwasthoughtthatgeneexpression SRSF1 function and intracellular localization are regu- is regulated mainly at a transcriptional level. However, it is lated by protein kinases and phosphatases. The protein kinase now clear that epigenetics and cotranscriptional and post- SRPK1 phosphorylates SRSF1 in the cytoplasm favouring its transcriptional processes are equally important. The discov- nuclear accumulation. SRPK1 overexpression has also been ery of splicing in the late 1970s was only the beginning of observed in several cancers [16, 17], and so it is conceivable whatwastobecomeaveryprominentfieldofresearch. that its activation could also synergistically potentiate the The vast majority of human genes, perhaps over 94%, are oncogenicpropertiesofSRSF1.Thispointisillustratedbythe alternatively spliced [8]. A cancer-associated gene can express VEGF-A example. VEGF-A pre-mRNA is one of the targets splice isoforms that either favour or counteract the growth of SRSF1. SRSF1 drives the expression of pro-angiogenic of cancer cells. For example, several regulators of apoptosis VEGF-A. Abnormally, high levels of SRPK1 cause the nuclear can express isoforms that are proapoptotic or antiapoptotic accumulation of SRSF1 in glomerular podocytes, leading to [9]. A frequently quoted example is Bcl-x, a member of the upregulation of pro-angiogenic VEGF-A splice isoforms [11]. Bcl-2 family of proteins that regulates the permeability of the outermembraneofmitochondria.WhereastheBcl-xSsplice isoform is proapoptotic, the Bcl-xL isoform is antiapoptotic 3. Regulation of the Expression of the as it prevents the release of mitochondrial components that Oncogenic Splice Factor SRSF1 wouldleadtoapoptosis. Sometimes quite unexpected splice isoforms are dis- Thus,itisclearthattheexpressionofsplicefactorsandof covered.Formanyyears,VEGF-Awasthoughttobean their protein kinases varies significantly in different tissues exclusively pro-angiogenic growth factor. However, VEGF- and in cancer. Yet surprisingly little is known about the factors A actually expresses splice isoforms that are anti-angiogenic. that regulate the expression of splice factors and splice factor An alteration in the balance of VEGF-A splice isoforms can kinases, perhaps because their involvement in pathological either promote or inhibit angiogenesis. The overexpression processes has only become apparent relatively recently. If of the pro-angiogenic splice isoform of VEGF-A is consis- SRSF1 is indeed an oncogenic splice factor, it is then very tently observed in solid tumours. By shifting the balance of important to understand how its expression is regulated. expression in favour of the anti-angiogenic isoform of VEGF- Several groups have begun to address this question and A, angiogenesis and tumour growth can even be inhibited. growing evidence suggests that SRSF1 expression is regulated Thus, the manipulation of VEGF-A alternative splicing might at multiple levels. provide opportunities for novel therapeutic targets [10, 11]. Myc is one of the best studied oncogenes; it encodes a Aberrant alternative splicing could also affect, system- transcription factor that binds to DNA elements known as E- atically, an entire cancer-associated process, such as the boxes. A ChIP on ChIP analysis using CpG arrays suggested a epithelial to mesenchymal transition (EMT) [12]. In other decade ago that Myc might interact with the SRSF1 promoter words, the systematic and coordinated alteration of alter- [18]. More recently, Das et al. found that Myc activates the native splicing of several functionally linked pre-mRNAs transcription of SRSF1 through two E-boxes. They show that could underpin specific processes in carcinogenesis. If this is Myc-dependent SRSF1 upregulation resulted in changes in the case, then dysregulated alternative splicing could surely alternative splicing of known SRSF1 targets favouring the consider itself a hallmark of cancer (Figure 1). But how might expression of splice isoforms that are consistent with an this work? Systematic changes in alternative splicing might oncogenic phenotype. In contrast, the knockdown of SRSF1 be due to the inappropriate activation (or inactivation) of reduces the oncogenic activity of Myc [19]. Furthermore, critically important splice factors or of the protein kinases and SRPK1, the protein kinase that promotes the entry of SRSF1 phosphatases that regulate their activity. into the nucleus, is transcriptionally repressed by the Wilms The fact that the expression of several RNA-binding tumour suppressor WT1 [11]. Thus, both a known oncogene proteins and splice factors is altered in cancer has been (Myc) and tumour suppressor transcription factor (WT1) can known for several years [13],butitisalsoclearthatspecific regulate the expression of the oncogenic splice factor SRSF1 splice factors are particularly important in cancer. The splice and of SRPK1, a protein kinase that regulates its localization. factor SRSF1 (previously known as ASF/SF2), a member Several splice factors are themselves alternatively spliced of the SR protein family (members of this family include to express functionally distinct isoforms. Some splice factors RNA Recognition Motifs and serine-arginine rich domains), even bind to their own pre-mRNAs to autoregulate their was described by Karni and colleagues as a proto-oncogene expression—for example, the splice factor SRSF2 (previously in 2007. Its overexpression transforms rodent fibroblasts known as SC35) favours the expression of splice isoforms International Journal of Cell Biology 3

Original six Self-sufficiency hallmarks Even more to growth (2000) Resistance to hallmarks? signals Others? antigrowth signals Unlimited Aberrant cell division alternative splicing The growing Evasion of list of cancer apoptosis hallmarks Abnormal metabolism Sustained angiogenesis Evasion of the immune Invasion and system metastasis Genomic Four additional instability Inﬂammation hallmarks (2011)

Figure 1: The original six hallmarks of cancer were proposed by Hanahan and Weinberg in 2000. Eleven years later, their list hadgrown to ten; but it could conceivably grow even further. An additional hallmark could be aberrant alternative splicing, in which the regulation of alternative splicing has gone astray systematically causing the inappropriate expression of multiple oncogenic splice isoforms.

that are unstable at the mRNA level [20]. SRSF1 also appears also occur through a failure of its autoregulation or through to bind to its own pre-mRNA in a similar autoregulatory the inactivation of specific microRNAs. Undoubtedly, the loop. There are at least six splice isoforms of SRSF1; one regulation of other splice factors is also likely to be complex expresses the full length splice factor, but the other five and multilayered. There are clearly many ways in which to are unproductive. Unproductive transcripts are targeted by perturbsplicefactorlevelsincancer. nonsense-mediated decay (NMD) due to the presence of premature stop codons. However, SRSF1 also has cytoplasmic 4. Alternative Splicing and Multistage activities—like many splice factors, it is multifunctional. The Carcinogenesis ability of SRSF1 to downregulate its own expression is to a large extent mediated at the level of the regulation of Histopathologists have known for a long time that there are translation [21]. distinct phases in the development of cancer. In other words, There is also evidence that SRSF1 expression is regulated the onset of cancer is generally thought to be a multistep by microRNAs. Leukemia/lymphoma-related factor (LRF) is process. The classical example to illustrate this phenomenon an oncogenic transcription factor. A recent study shows that is the adenoma-carcinoma sequence in the development of LRF represses the microRNAs miR-28 and miR-505 [22]. colorectal cancer. Intestinal epithelial cells form a thin layer 󸀠 These two microRNAs target the 3 UTR of SRSF1 mRNA. that is constantly being replaced. These epithelial cells are Thus, a reduction in LRF results in higher levels of the associated with the formation of carcinomas. microRNAs and therefore reduced SRSF1 expression. In turn, The sequence of events that leads to malignant disease the reduction in SRSF1 levels can result in genomic instability, is thought to be as follows: normal epithelium changes to cell cycle arrest, and apoptosis. Furthermore, there is an hyperplastic epithelium; this change is often associated with alternative intron in SRSF1 in between two highly conserved loss of function of the APC (adenomatous polyposis coli) 󸀠 elements in its 3 UTR. Skipping of this alternative intron gene. APC encodes a complex multidomain, multifunctional appears to increase transcript stability, perhaps by removing tumour suppressor. It is involved in the regulation of the amicroRNAbindingsite[23]. cell-cycle, apoptosis, intercellular adhesion, and cytoskeletal In summary, SRSF1 expression can be controlled at mul- architecture. Further changes alter the hyperplastic epithe- tiple levels; transcriptionally, cotranscriptionally (through lium to early, intermediate, and late adenomas. These changes alternative splicing), and posttranscriptionally (through the canbedrivenbyseveralgenes,includingK-Ras. Ras genes regulation of translation and mRNA stability). In other encode GTPases that are involved in cell signalling associated words, there are several ways to fine tune the expression of with cellular proliferation. Adenomas can then develop into oncogenic SRSF1. The ability of Myc to regulate SRSF1 and of carcinomas that eventually acquire the ability to invade and WT1 to regulate SRPK1 expression is particularly interesting metastasize.Thislatterchangeisoftenassociatedwiththe as these two transcription factors are associated with a wide loss of function of the TP53 gene. TP53 is the widely studied range of cancers. The upregulation of SRSF1 in cancer could “guardian of the genome” tumour suppressor transcription 4 International Journal of Cell Biology

Further changes to Therapy alternative splicing resistant Metastasis disease Therapy Further Therapy changes to alternative splicing Normal Hyperplastic Adenoma Carcinoma epithelium epithelium

Aberrant Aberrant Aberrant APC K-Ras TP53 splicing splicing splicing

Splice factor Splice factor 1 Splice factor 2 kinase activation activation activation

Figure 2: The adenoma-carcinoma sequence classically illustrates the multistage aetiology of colorectal cancer. Three genes frequently involved are APC, K-Ras,andTP53. This theoretical model suggests that the genetic lesions that drive the stages include changes that cause the inappropriate activity of oncogenic splice factors or splice factor kinases. The result is a significant change in the ratio of splice isoforms that drastically alter APC, TP53,andK-Ras function. Conventional treatments might cause selective pressures that drive further changes in the alternative splicing of key genes, leading to resistance to therapy. factor that regulates a multitude of processes including the exon skipping is what gives rise to the nonfunctional APC, cell cycle, apoptosis, and the response to DNA damage. All not the missense mutation per se. These studies illustrate three genes (APC, K-Ras,andTP53) express splice isoforms two principles. The first is the need to consider the potential whose functional properties are distinct and even antagonis- effects of intronic mutations on alternative splicing (these are tic. Several other genes have been shown to be involved in the often ignored); and the second is the need to examine the aetiology of colorectal cancer; for the purposes of this review, potential consequences of exonic mutations on alternative onlythesethreearediscussed.However,itisimportantto splicing because exons contain splicing enhancer and silencer note that other genes involved in colorectal cancer are also sequences. alternative spliced in malignancy, giving rise to functionally Two splice variants of K-Ras have been described, K- distinct isoforms. Notable examples are the mismatch repair Ras 4A and 4B. They arise from two alternative versions genes MLH1 and MSH2 [24]. of exon 4. Whereas the inclusion of exon 4A results in Several alternative splice isoforms of APC have been a proapoptotic K-Ras, the inclusion of exon 4B results in described that result in proteins with molecular weights rang- an antiapoptotic protein. Both are coexpressed in many ing from 90 to 300 kDa. De Rosa and colleagues examined tissues, but their ratio is altered in sporadic colorectal cancer a cohort of 24 patients suffering from familial adenomatous favouring the antiapoptotic 4B isoform [27]. However, in polyposis (FAP), each with a germline mutation in the APC mice, the expression of the 4A splice variant is not required gene, comparing them to 17 FAP patients without apparent for normal development despite its expression in a wide range 󸀠 󸀠 APC mutations and 9 controls [25]. Using nested 5 and 3 of tissues [28]. It is nonetheless conceivable that mutations primers, they identified nine novel transcripts. Three of these that favour the expression of 4B over 4A might augment the preserved the open reading frame. One of the transcripts oncogenic properties of activated K-Ras. contained an additional exon termed exon 1A; its inclusion The TP53 gene was thought for a long time to encode leads to a premature stop codon in exon 2. The inclusion of a single protein, but it is now abundantly clear that it is exon 1A was found to be 3.5-fold higher in colorectal cancer extensively alternatively spliced [29]. A remarkably complex versus normal mucosa. Transcripts that harbour premature series of splice isoforms have been described and several of stop codons are generally degraded by nonsense mediated these can regulate the transcriptional activity of p53 [30, 31]. decay (NMD)—this was shown to be the case when exon P53 is also thought to suppress tumorigenesis by inducing 1A is included. The potential significance is that greater senescence; this is facilitated by the splice isoform p53𝛽.The inclusion of exon 1A could effectively result in less APC expression of p53𝛽 is induced by the splice factor SRSF3 protein being expressed because less productive mRNAs are (previously known as SRp20). The knockdown of SRSF3 synthesised overall. Another group took a different approach induces senescence in fibroblasts through a p53-mediated andanalysedanunusualFAPpatient.Thispatienthada mechanism [32]. SRSF3 binds to a consensus binding site in missense mutation in codon 640. Rather than necessarily theexonthatisuniquetop53𝛽. To add to these findings, altering the functional properties of the protein, the mutation another study published earlier in the year in the same journal was shown to interfere with a splicing enhancer motif bound demonstrated that the downregulation of SRSF3 induces G1 by the oncogenic splice factor SRSF1 [26]. The consequences cell cycle arrest and apoptosis in colon cancer cells [33]. How- of the mutation were found to be the skipping of exon 14. This ever, in this case, the G1 arrest and proapoptotic mechanism International Journal of Cell Biology 5 wereshowntobethroughachangeinthealternativesplicing [2] D. Hanahan and R. A. Weinberg, “Hallmarks of cancer: the next of homeodomain-interacting protein kinase 2 (HIPK-2); the generation,” Cell,vol.144,no.5,pp.646–674,2011. consequences of SRSF3 downregulation on p53 alternative [3] O. Warburg, “On the origin of cancer cells,” Science,vol.123,no. splicing were not examined in detail. However, HIPK2 3191, pp. 309–314, 1956. phosphorylates p53 at serine 46, favouring its proapoptotic [4] P. Vaupel and A. Mayer, “Hypoxia in cancer: significance and properties [34]. Thus, SRSF3, like most splice factors, appears impact on clinical outcome,” Cancer and Metastasis Reviews, to be working in a coordinated way perhaps on groups of vol. 26, no. 2, pp. 225–239, 2007. transcripts that are involved in related pathways. [5]R.F.JohnsonandN.D.Perkins,“Nuclearfactor-𝜅B, p53, Together, the observations on the alternative splicing of and mitochondria: regulation of cellular metabolism and the the APC, K-Ras,andTP53 genes suggest that in the context of Warburg effect,” Trends in Biochemical Sciences,vol.37,no.8, multistep carcinogenesis of colorectal cancer, genetic lesions pp. 317–324, 2012. that affect alternative splicing are likely to contribute signif- [6] B. Faubert, G. Boily, S. Izreig et al., “AMPK is a negative icantly to the aetiology of disease (Figure 2). It is therefore regulator of the Warburg effect and suppresses tumor growth notsurprisingthatspliceisoformsofthesegeneshavebeen in vivo,” Cell Metabolism,vol.17,no.1,pp.113–124,2013. indicated as potential therapeutic targets in colorectal cancer [7] V. Huff, “Wilms’ tumours: about tumour suppressor genes, an [35]. oncogene and a chameleon gene,” Nature Reviews Cancer,vol. 11, no. 2, pp. 111–121, 2011. 5. Conclusion and Future Directions [8]Q.Pan,O.Shai,L.J.Lee,B.J.Frey,andB.J.Blencowe,“Deep surveying of alternative splicing complexity in the human tran- The emergence of high-throughput next generation sequenc- scriptome by high-throughput sequencing,” Nature Genetics, ing will make it possible to examine the transcriptomes of vol.40,no.12,pp.1413–1415,2008. tumours in detail [36]. It will be possible to understand [9] J. P. Venables, “Unbalanced alternative splicing and its signifi- how aberrant alternative splicing can contribute to carcino- cance in cancer,” BioEssays, vol. 28, no. 4, pp. 378–386, 2006. genesis and to the progression of disease. This is illustrated [10] S. J. Harper and D. O. Bates, “VEGF-A splicing: the key to anti- by a recent study in which normal epithelial cells were angiogenic therapeutics?” Nature Reviews Cancer, vol. 8, no. 11, stressed with nicotine. The expression of 54,699 transcripts pp. 880–887, 2008. was examined; 173 were alternatively spliced in response [11] E. M. Amin, S. Oltean, J. Hua et al., “WT1 mutants reveal to nicotine exposure. These transcripts encoded proteins SRPK1 to be a downstream angiogenesis target by altering associated with DNA recombination, replication, and repair VEGF splicing,” Cancer Cell, vol. 20, no. 6, pp. 768–780, 2011. [37]. It is therefore a good time to apply next generation [12] B. de Craene and G. Berx, “Regulatory networks defining sequencing technology to the study of alternative splicing in EMT during cancer initiation and progression,” Nature Reviews cancer systematically. However, a caveat to bear in mind is Cancer,vol.13,no.2,pp.97–110,2013. that alternative splicing patterns within primary, secondary [13] C. Ghigna, M. Moroni, C. Porta, S. Riva, and G. Biamonti, tumour, and metastases are heterogenous, perhaps reflecting “Altered expression of heterogeneous nuclear ribonucleopro- the differential expression and activity of splice factors in teins and SR factors in human colon adenocarcinomas,” Cancer different parts of the tumour. Nonetheless, the power of next Research, vol. 58, no. 24, pp. 5818–5824, 1998. generation sequencing means that these complexities can also [14] R. Karni, E. de Stanchina, S. W. Lowe, R. Sinha, D. Mu, and A. be addressed. R. Krainer, “The gene encoding the splicing factor SF2/ASF isa The number of publications that associate changes in proto-oncogene,” Nature Structural and Molecular Biology,vol. 14,no.3,pp.185–193,2007. alternative splicing with specific cancers has risen rapidly. It seems highly likely that alterations in the expression [15] O. Anczukow,A.Z.Rosenberg,M.Akermanetal.,“Thesplicing´ factor SRSF1 regulates apoptosis and proliferation to promote andactivityofcriticalsplicefactorsoroftheirmodifiers mammary epithelial cell transformation,” Nature Structural and (factor kinases and phosphatases) could result in a series Molecular Biology,vol.19,no.2,pp.220–228,2012. of changes to the alternative splicing of several genes that [16] G. M. Hayes, P. E. Carrigan, and L. J. Miller, “Serine-arginine together provide a growth or survival advantage to cancer protein kinase 1 overexpression is associated with tumorigenic cells. A systematic dysregulation of alternative splicing should imbalance in mitogen-activated protein kinase pathways in therefore be considered yet another hallmark of cancer. breast, colonic, and pancreatic carcinomas,” Cancer Research, Despite the growing evidence that alternative splicing vol.67,no.5,pp.2072–2080,2007. plays an important role in cancer, the pharmaceutical sector [17] S. Gout, E. Brambilla, A. Boudria et al., “Abnormal expression has yet to exploit fully the therapeutic potential of manip- of the pre-mRNA splicing regulators SRSF1, SRSF2, SRPK1 and ulating alternative splicing. It is therefore arguably a very SRPK2innonsmallcelllungcarcinoma,”PLoS ONE,vol.7,no. good time for funders to invest even more resources in 10, Article ID e46539, 2012. basic research into the dysregulation of alternative splicing in [18]D.Y.L.Mao,J.D.Watson,P.S.Yanetal.,“AnalysisofMYC cancer. bound loci identified by CpG island arrays shows that Max is essential for MYC-dependent repression,” Current Biology,vol. References 13, no. 10, pp. 882–886, 2003. [19] S. Das, O. Anczukow, M. Akerman, and A. R. Krainer, “Onco- [1] D. Hanahan and R. A. Weinberg, “The hallmarks of cancer,” Cell, genic splicing factor SRSF1 is a critical transcriptional target of vol. 100, no. 1, pp. 57–70, 2000. MYC,” Cell Reports,vol.1,no.2,pp.110–117,2012. 6 International Journal of Cell Biology

[20] A.Sureau,R.Gattoni,Y.Dooghe,J.Stevenin,´ and J. Soret, “SC35 [37] J. H. Bavarva, H. Tae, R. E. Settlage, and H. R. Garner, “Char- autoregulates its expression by promoting splicing events that acterizing the genetic basis for nicotine induced cancer devel- destabilize its mRNAs,” EMBO Journal,vol.20,no.7,pp.1785– opment: a transcriptome sequencing study,” PLoS ONE,vol.18, 1796, 2001. no. 6, Article ID e67252. [21] S. Sun, Z. Zhang, R. Sinha, R. Karni, and A. R. Krainer, “SF2/ASF autoregulation involves multiple layers of posttranscriptional and translational control,” Nature Structural and Molecular Biology,vol.17,no.3,pp.306–312,2010. [22] L. Verduci, M. Simili, M. Rizzo et al., “MicroRNA (miRNA)- mediated interaction between leukemia/lymphoma-related factor (LRF) and alternative splicing factor/splicing factor 2 (ASF/SF2) affects mouse embryonic fibroblast senescence and apoptosis,” The Journal of Biological Chemistry,vol.285,no.50, pp. 39551–39563, 2010. [23] Y. Akaike, K. Kurokawa, K. Kajita et al., “Skipping of an 󸀠 alternative intron in the srsf1 3 untranslated region increases transcript stability,” Journal of Medical Investigation,vol.58,no. 3-4, pp. 180–187, 2011. [24] K. Miura, W. Fujibuchi, and I. Sasaki, “Alternative pre-mRNA splicing in digestive tract malignancy,” Cancer Science,vol.102, no. 2, pp. 309–316, 2011. [25] M. de Rosa, G. Morelli, E. Cesaro et al., “Alternative splicing and nonsense-mediated mRNA decay in the regulation of a new adenomatous polyposis coli transcript,” Gene,vol.395,no.1-2, pp. 8–14, 2007. [26] V. Gonc¸alves, P. Theisen, O. Antunes et al., “A missense mutation in the APC tumor suppressor gene disrupts an ASF/SF2 splicing enhancer motif and causes pathogenic skipping of exon 14,” Mutation Research,vol.662,no.1-2,pp.33–36,2009. [27] S. J. Plowman, R. L. Berry, S. A. Bader et al., “K-ras 4A and 4B are co-expressed widely in human tissues, and their ratio is altered in sporadic colorectal cancer,” Journal of Experimental and Clinical Cancer Research,vol.25,no.2,pp.259–267,2006. [28]S.J.Plowman,M.J.Arends,D.G.Brownsteinetal.,“The K-Ras 4A isoform promotes apoptosis but does not affect either lifespan or spontaneous tumor incidence in aging mice,” Experimental Cell Research,vol.312,no.1,pp.16–26,2006. [29] V. Marcel and P. Hainaut, “p53 isoforms—a conspiracy to kidnap p53 tumor suppressor activity?” Cellular and Molecular Life Sciences,vol.66,no.3,pp.391–406,2009. [30] J. C. Bourdon, K. Fernandes, F. Murray-Zmijewski et al., “p53 isoforms can regulate p53 transcriptional activity,” Genes and Development,vol.19,no.18,pp.2122–2137,2005. [31] J. C. Bourdon, “p53 and its isoforms in cancer,” The British Journal of Cancer,vol.97,no.3,pp.277–282,2007. [32] Y. Tang, I. Horikawa, M. Ajiro et al., “Downregulation of splicing factor SRSF3 induces p53b, an alternatively spliced isoform of p53 that promotes cellular senescence,” Oncogene, vol.32,no.22,pp.2792–2798,2013. [33] K. Kurokawa, Y. Akaike, K. Masuda et al., “Downregulation of serina.arginine-rich splicing factor 3 induces G1 cell cycle arrest and apoptosis in colon cancer cells,” Oncogene,2013. [34] G. D’Orazi, B. Cecchinelli, T. Bruno et al., “Homeodomain- interacting protein kinase-2 phosphorylates p53 at Ser 46 and mediates apoptosis,” Nature Cell Biology,vol.4,no.1,pp.11–19, 2002. [35] K. Miura, W. Fujibuchi, and M. Unno, “Splice isoforms as therapeutic targets for colorectal cancer,” Carcinogenesis,vol.33, no. 12, pp. 2311–2319, 2012. [36] H. Feng, Z. Qin, and X. Zhang, “Opportunities and methods for studying alternative splicing in cancer with RNA-Seq,” Cancer Letters,2012. Hindawi Publishing Corporation International Journal of Cell Biology Volume 2013, Article ID 642853, 12 pages http://dx.doi.org/10.1155/2013/642853

Review Article Ewing Sarcoma Protein: A Key Player in Human Cancer

Maria Paola Paronetto1,2 1 Department of Health Sciences, University of Rome “Foro Italico”, 00135 Rome, Italy 2 Laboratory of Cellular and Molecular Neurobiology, Fondazione Santa Lucia IRCSS, 00143 Rome, Italy

Correspondence should be addressed to Maria Paola Paronetto; [email protected]

Received 1 June 2013; Accepted 26 July 2013

Academic Editor: Michael Ladomery

Copyright © 2013 Maria Paola Paronetto. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The Ewing sarcoma protein (EWS) is a well-known player in cancer biology for the specific translocations occurring in sarcomas. The EWS-FLI1 gene fusion is the prototypical translocation that encodes the aberrant, chimeric transcription factor, which is a landmark of Ewing tumors. In all described Ewing sarcoma oncogenes, the EWS RNA binding domains are completely missing; thus RNA binding properties are not retained in the hybrid proteins. However, it is currently unknown whether the absence of EWS function in RNA metabolism plays a role in oncogenic transformation or if EWS plays a role by itself in cancer development besides its contribution to the translocation. In this regard, recent reports have highlighted an essential role for EWS in the regulation of DNA damage response (DDR), a process that counteracts genome stability and is often deregulated in cancer cells. The first partof this review will describe the structural features of EWS and its multiple roles in the regulation of gene expression, which are exerted by coordinating different steps in the synthesis and processing of pre-mRNAs. The second part will examine the role ofEWSin the regulation of DDR- and cancer-related genes, with potential implications in cancer therapies. Finally, recent advances on the involvement of EWS in neuromuscular disorders will be discussed. Collectively, the information reviewed herein highlights the broad role of EWS in bridging different cellular processes and underlines the contribution of EWS to genome stability and proper cell-cycle progression in higher eukaryotic cells.

1. Introduction chromosome 21 [4]. Thus, the oncogenic conversion of EWS follows a common scheme of activation, by exchanging its EWS was originally identified because of the t(ll;22) (q24;q12) RNA binding domain with different DNA binding domains, chromosome translocation, characteristic of Ewing sarcoma thus generating tumor-specific fusions proteins. The onco- and related subtypes of primitive neuroectodermal tumors genes generated in Ewing sarcoma have been extensively [1]. In these types of cancers, a hybrid transcript is generated studied and the main aspects of Ewing sarcoma have been by the genomic regions where the breakpoints of chromo- covered by several excellent recent reviews [5–8]. On the some22and11occur.Thetranslocationalterstheopen contrary, only limited information is available on the function reading frame of the EWSR1 gene on chromosome 22 by of the EWS protein itself and whether or not it plays a role in substituting a sequence encoding a putative RNA-binding oncogenesis or in other pathological situations. domain with that of the DNA-binding domain of FLI-1,the ThisreviewwillfocusontheroleplayedbyEWSprotein human homologue of murine Fli-1 [2]. Another translocation in the regulation of RNA metabolism, in normal conditions involving the EWSR1 gene was subsequently identified in as well as in pathological situations. malignant melanoma of soft parts: the deduced chimaeric protein consisted of the N-terminal domain of EWS linked to the bZIP domain of ATF-1, a transcription factor regulated 2. EWS Family by cAMP that was not previously implicated in oncogenesis [3]. In addition to the EWS-FLI-1 and EWS-ATF-1 fusions, EWS belongs to the TET family of DNA and RNA- the EWSR1 genecanbefusedtoERG, which encodes a binding proteins, which consists of translocated in liposar- transcription factor closely related to FLI-1 but located on coma/fused in sarcoma protein (FUS/TLS), Ewing sarcoma 2 International Journal of Cell Biology protein (EWS), and TATA-binding protein-associated fac- (AS) could be target of PTB/nPTB regulation in neurons. tor 15 (TAF15). TET proteins bind RNA as well as DNA Notably, whereas the level of isoN expression is stable from and are implicated in the regulation of gene expression embryonic stage E11.5 to birth, the level of the main isoform andincellsignaling.InadditiontomRNAsplicingand shows a considerable decrease. This suggests that the two iso- processing, TET proteins also directly interact with tran- forms play distinct roles during brain development and that scription factors and can function as transcriptional cofac- the isoN variant is associated with neural differentiation20 [ ]. tors. Moreover, like EWS, all the members of the TET Additional isoforms lacking exon 6 and/or including exon family also contribute to human pathologies, as they are 8A have been sequenced from a pool of 107 human cDNAs involved in sarcoma translocations [2, 8] and neurological and annotated [21]. Notably, this triplet encodes Serine 325 diseases [9, 10]. They contain several conserved domains: a [22], which is a putative casein kinase II phosphorylation serine-tyrosine-glycine-glutamine (SYGQ) domain, an RNA- site and may represent a posttranslational modification that recognition motif (RRM), a zinc finger motif, and three RNA distinguishes between the two protein isoforms of EWS. binding Arginine-Glycine-Glycine- (RGG-) rich domains (Figure 1)[11]. The TET proteins are ubiquitously expressed in most human tissues; they have a nuclear localization 4. Physiological Functions of the EWS Protein and predominantly reside in the nucleus [12, 13]butcan translocate to the cytoplasm upon stress conditions [13, 14] The functions of the EWS protein have been partially revealed by the generation of a knockout mouse model through or in pathological situations [15, 16]. −/− agenetargetingstrategy[23]. Ews mice were born at the expected mendelian ratio, but displayed smaller size 3. Genomic Structure of EWSR1 Gene than their littermates and have a high rate (about 90%) of postnatal mortality prior to weaning. Analysis of these mice The EWSR1 gene spans about 40 kb on chromosome 22 and also demonstrated that the Ewsr1 gene is essential for pre-B is encoded by 17 exons [17]. The first 7 exons encode the lymphocyte development and meiosis. Interestingly, Fus/Tls- N-terminal domain of EWS, which consists of a repeated null mice showed a quite similar phenotype [24, 25]; however, degenerated polypeptide of 7 to 12 residues rich in tyro- while FUS/TLS is required for pairing of the autosomes, EWS sine, serine, threonine, glycine, and glutamine (SYGQ). The appearstobeinvolvedinpairingtheXYsexchromosomes degenerated SYGQ repeat [19]isencodedbysequencesthat during meiosis [23]. Indeed, Ews-null spermatocytes show a stretch from exon 3 to the end of exon 7. Exons 8 to 17 of high frequency of unsynapsed XY chromosomes but display EWS encode regions associated with RNA binding. The three normal formation of autosomal bivalents; in contrast, defects RGG motifs are mainly encoded by exons 8, 9, 14, and 16, in the formation of autosomal bivalents were observed while exons 11, 12, and 13 encode the RRM (Figure 1). The in Fus/Tls-deficient spermatocytes. Moreover, while EWS 󸀠 DNA sequence in the 5 region of the EWSR1 gene lacks expression is diffuse throughout spermatocytes, FUS/TLS canonical promoter elements, such as TATA and CCAAT expression is excluded from the sex body region [23, 25]. consensus sequences [17], but contains G/C rich stretches Nevertheless, as a result of these defects in chromosome [18], suggesting a housekeeping role for TET proteins. pairing, both knockout mice are sterile. TAF15, FUS/TLS, and EWSR1 genes display extensive Interestingly, inactivation of Ews in mouse embryonic similarities, suggesting that they are closely related and may fibroblasts resulted in hypersensitivity to ionizing radiation have originated from a common ancestor gene (Figure 1)[18]. and caused premature cellular senescence, possibly due to Although their total genomic size differs, TET genes share lack of the reported interaction of the EWS protein with a similar genomic organization with exon/intron junctions nuclear lamin A/C [23]. Moreover, loss of Ews brought about at the same sites, reflecting the conserved protein structure dramatic changes in the dynamics of hematopoietic stem consisting of an N-terminal transcription activation domain cells (HSCs). In particular, deletion of Ews promoted stem and the RRM, zinc finger domain, and several RGG repeat cell exit from a quiescent state and entrance into the cell boxes at the C-terminus (Figure 1). The first exon of these cycle, which is one of the common features found in HSCs 󸀠 three genes encodes a 5 untranslated region and the trans- from aged wild-type mice. Cellular senescence is associated lation initiation codon. The RRM region and flanking parts with a progressive impairment of the immune system or show extensive homologies in amino acid composition and “immunosenescence,” which is characterized by a defective exon/intron structure. On the other hand, the parts of the Th1 (T helper 1) response. T cells can be separated depending genes located between FUS/TLS exons 4, 5, and 6, TAF15 upon the specific cytokines they secrete in response to exons 5, 6 and 7, and EWSR1 exons5,6,7,and8displaylower antigenic stimulation. Th1 cells primarily produce interferon- degree of homology. (IFN-) 𝛾 and interleukin- (IL-) 2 [26]. In line with this notion, −/− Different isoforms of the EWS protein have been Ews mice displayed a specific decrease in IFN-𝛾 and IL-2 described. A splicing isoform containing the exon 4A is production [27]. Thus, the phenotype of Ews ablation mouse specifically expressed in the central nervous system (isoN), models is consistent with a physiological function for EWS in both mice and humans, and is not present in the tumor- in response to the occurrence of genomic alterations, as specific EWS fusions [20]. Interestingly, the regions flanking suggested by the hypersensitivity to IR in the absence of EWS this variable exon contain many putative PTB binding sites and the defects in cell types where physiological DNA breaks [20], highlighting the possibility that EWS alternative splicing occur, including B cells engaged in VDJ recombination and International Journal of Cell Biology 3

hsRPB7 (RNAPII) [28] YB-1 [63; 64] Protein U1C [58] SR proteins [62] interactions ZFM1 [60] PRMT1 [35] CBP [35] TFIID [28] PYK2 [14] p68 [102]

EWS DNA activation domain Nucleic acid binding 1 656AA

Domains SYGQ richRGG RRMRGG ZF RGG

Exons 12 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

FUS/TLS DNA activation domain Nucleic acid binding 1 526AA

Domains SYGQ richRGG RRMRGG ZF RGG

Exons 12 3 4 5 6 7 8 9 10 11 12 13 14 15

TAF15 DNA activation domain Nucleic acid binding 1 592AA Domains SYGQ richRGG RRMRGG ZF RGG

Exons 12 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Figure 1: Domain structure of TET proteins: the N-terminal activation domain and the C-terminal nucleic acid binding domain of TET proteins are schematized: the SYGQ rich domain, RGG boxes, and the RNA binding domain (RBD), Cys2-Cys2 zinc finger (ZF). Below the scheme, the exons encoding each domain are indicated for EWS, FUS/TLS, and TAF15, respectively [17, 18]. In the upper part of the figure the proteins that have been described to interact with either the DNA activation domain or the nucleic acid binding domain of EWS are listed. meiotic germ cells engaged in homologous recombination. and the p300 transcriptional activator, thus cotransactivating Collectively these results highlight a role for EWS in the in vitro several promoters in a cell-type specific manner [34, response to DNA damage. 35]. Collectively these studies indicate that EWS may regulate transcription both positively and negatively. 5. EWS and Transcription The amino terminus of EWS protein was shown to act as a transcriptional activator when fused to a DNA-binding Several observations document the involvement of EWS in domain [36]. However, the transcription activation potential transcriptional regulation of gene expression. EWS is able to isdecreasedorevenabolishedinthefull-lengthprotein, associate with different subunits of the transcription factor indicating that the RNA or DNA binding domains may TFIID [28]. Notably, this feature is not maintained by the regulatetheactivationdomain[37, 38].Deletionsofeitherthe oncogenic fusion protein EWS-FLI-1, suggesting that EWS zinc finger domain or the RRM play no role in repression. In and EWS-FLI-1 behave differently in this respect. Moreover, contrast, deletion of each RGG domain results in substantial EWS associates directly with RNA Polymerase II (RNAP II) loss of repression, indicating that the RGG regions are neces- [28, 29] and with its subunits Rpb3, Rpb4, and Rpb7 [28, sary for repression [37, 38]. Thus, the transcription repression 30]. In addition to directly contacting general transcription might result from the C-terminal RGG domain that blocks factors and RNAPII, EWS protein regulates transcription the interaction between the glutamine-rich domain in the through the interaction with activators and repressors. For N-terminus and hsRPB7 (human homologue of the seventh instance, EWS binds various transcription factors containing largest subunit of yeast RNAPII (RPB7) (Figure 1,upper thePOU(fromthemammalianPit-1andOct-1,andthe panel). C. elegans Unc86) homeodomain, a specific DNA-binding domain conserved throughout evolution that appears to exert critical developmental functions. EWS binds OCT4, 6. EWS and Nucleic Acid Binding a transcriptional activator that is essential to maintain an undifferentiated totipotent state of embryonic stem and germ As mentioned above, EWS protein contains several domains cells [31], and Brn3a, the brain-specific homeobox/POU capable to bind independently nucleic acid sequences. The domain protein 3A, which is expressed in the developing and RRM, a classical RNA Binding Domain (RBD), is the most adult nervous systems and promotes neuronal differentiation conserved region within the TET protein family. The RRM [32, 33]. Additionally, EWS has been shown to interact with is composed of approximately 90 amino acids and contains a the histone acetyl-transferase CREB-binding protein (CBP) central sequence of eight conserved residues that are mainly 4 International Journal of Cell Biology aromatic and positively charged. The RRM folds into a sand- factors [50]. Indeed, mobility shift experiments revealed that wichstructurecomposedofonefour-strandedantiparallel𝛽- EWS binds to single-stranded (but not to double-stranded) sheet and two 𝛼-helices (𝛼1and𝛼2) packed against the 𝛽- DNA, with preference for its target exon sequences [51], open- sheet [39]. Within this domain, two short ribonucleoprotein ing the possibility that EWS is recruited during transcription motifs, RNP-1 (in 𝛽3-strand) and RNP-2 (in 𝛽1-strand) through specific recognition of DNA and/or RNA sequences. usually separated by 25–35 residues, directly contact RNA via Lastly, posttranslational modifications can affect bind- hydrogen bonds and ring stacking [39]. Unlike other proteins ing of EWS to nucleic acids. It has been shown that with this type of RBD, TET proteins have an acidic residue EWS IQ domain, a 25-amino-acid domain located at the at the second position and a threonine residue in the fourth end of the activation domain, interacts with calmodulin position of RNP-1, as well as an unusually long loop after the and is phosphorylated by protein kinase C [52]. The IQ first 𝛼-helix [40], which may affect the structure of the protein domain forms an amphiphilic seven-turn 𝛼-helix capable 2+ since this region contributes to the hydrophobic core of the of binding calmodulin in a Ca -independent manner, a domain and to RNA-binding specificity or affinity. well-characterized feature of neuronal proteins such as The (DW) C4 zinc finger motif has several highly con- neuromodulin and neurogranin [52]. Interestingly, PKC served features, distinct from those present in other known phosphorylation of EWS within the IQ domain inhibits C5-C5 zinc-fingers [41]. It contains an aspartic acid (D) its binding to RNA homopolymers, and, conversely, RNA followed by tryptophan (W), preceding the first cysteine binding to EWS interferes with PKC phosphorylation [53]. (C) residue of the finger. Within the zinc-finger loop, the Moreover, enzymatic methylation of arginine residues by fourth residue after the second cysteine is an asparagine protein arginine N-methyltransferase 3 (PRMT3) reduces the andisfollowedbyanaromaticresidue.Secondarystructure affinity of EWS RGG (RGG3) domain toward G-quadruplex, prediction algorithms suggest that the C-terminal region of while it increases its binding to single-strand DNA and the motif may give rise to an alpha helix, a general feature of RNA. Conversely, replacement of the arginine methylated many zinc fingers motif [42]. by PRMT3 with lysine residues within the RGG domains The carboxy-terminus of EWS contains also three RGG decreases the binding affinity of EWS toward G-quadruplex motifs that may increase RNA affinity of the RRM or DNA and RNA [53]. Furthermore, EWS methylation by zinc finger and may also be the site of posttranslational PRMT1 (a homologue of PRMT3) reduces CBP-dependent modifications that regulate RNA binding or protein-protein EWS transcriptional activity and induces EWS translocation interactions [39]. Sequence-specific RNA binding by EWS into the cytoplasm [54]. Thus, these observations suggest has been examined by various groups. It was shown that that several posttranslational modifications can fine-tune the EWS binds polyU and polyG sequences through its carboxy- affinity of EWS for its nucleic acid targets. terminal RGG domain [43]. Moreover, in vitro systematic evolution of ligands by exponential enrichment (SELEX) [44] experiments suggested that the RRM and RGG motifs of 7. EWS and Splicing TET proteins cooperate to bind GGUG RNA, although with relatively low affinity (𝐾𝑑 = 250 nM) [45]. Many reports have highlighted a potential role for EWS in RGG domains are present in several RNA and DNA splicing regulation. Most human genes contain introns that binding proteins. The RGG domain of FMRP, an RNA are removed by splicing, a process orchestrated and catalyzed binding protein involved in neural cell differentiation, inter- by a large multiprotein/RNA complex named spliceosome. acts with RNA forming G-quartets, a unique structural The spliceosome is composed of five small nuclear RNAs arrangement intrinsic to guanine-rich DNA and RNA [46]. (snRNAs)—U1, U2, U4, U5, and U6 snRNA—as well as In addition, the RRM and the RGG domains of nucleolin, many protein factors [55]. Some of these proteins are tightly a DNA-binding protein contributing to the transcription of associated with the snRNAs, forming small nuclear ribonu- ribosomal RNA, bind to ribosomal DNA forming G-quartets cleoproteins (snRNPs) assembled in a stepwise manner onto [47]. EWS protein specifically targets DNA and RNA form- the pre-mRNA [55]. Work over the last two decades has ing G-quadruplex in vitro, the four-stranded nucleic acid elucidated the temporal sequence of recognition of the splice structures folded from G-rich repeat sequences stabilized sites by the respective snRNPs and protein factors and the by the stacking of planar G-quartets. The specificity of G- identity of former proteins of the spliceosome. EWS and quadruplex recognition depends on the guanidinium group FUS/TLS were identified by enhanced mass spectrometric of the arginine in the RGG domain of the C-terminus of EWS tools and improved databases as proteins copurifying with [48]. splicing complexes assembled on pre-mRNAs [56]. More- It has been reported that overexpression of EWS leads to over, in vitro snRNP reconstitution and snRNP immuno- the activation of the c-fos, Xvent-2, and ErbB2 promoters [12], precipitation experiments have demonstrated, at least for but whether the role of EWS in transcription is determined FUS/TLS, the association with the Sm core domain of the by its ability to target a specific DNA and RNA structure or various spliceosomal snRNPs [57]. only by binding to the RNAPII is not clear. Notably, the c- Regarding EWS, a yeast two-hybrid screen revealed that it fos and ErbB2 promoters contain G-rich sequences that could interacts with U1C, one of the three U1 small nuclear protein 󸀠 potentially form G-quadruplex structures [49]. component of the U1snRNP [58], which binds to the 5 splice Nevertheless, DNA binding is likely to occur also through site on pre-mRNA to form a stable complex identified as the the zinc finger, a domain frequently found in transcription early (E) complex in mammalian splicing extracts [59]. On International Journal of Cell Biology 5 the other hand, EWS has also been shown to interact with CPT inhibits the splicing activity of these two RNA binding ZFM1, a transcriptional repressor identical to the branch- proteins (RBPs). Similarly, UV irradiation causes EWS dis- point binding protein BBP/SF1 [60],butwhetherornotitis sociation from genomic and transcription sites, concomitant 󸀠 involved in the recognition of 3 splice site, as demonstrated with increased translocation of the protein in nucleoli [51]. for FUS/TLS during the second step of splicing [61], is still This mechanism contributes, at least in part, to the AS unknown. changesdetecteduponEWSknockdownorUVirradiation As described above, EWS protein binds to hyperphos- (Figure 2(b))[51]. These reports document that EWS binds phorylated RNAPII through the N-terminal domain and directly to the alternatively spliced regions of its target genes, recruits serine-arginine (SR) splicing factors through the suggesting a direct role for EWS on the regulation of these C-terminal domain, thus coupling transcription to RNA events (Figures 2(a) and 2(b)). The reported correlation splicing [62](Figure 1, upper panel). The C-terminal domain between CLIP (cross-linked and immunoprecipitation) and is also involved in the interaction with YB-1, a shuttling ChIP (chromatin immunoprecipitation) signals further sug- splicing activator also playing a role in mRNA stability and gests a direct association of EWS with DNA, as also confirmed translation [63, 64]. It has been shown that the recruitment by EMSA experiments [51]. EWS binding was also observed of processing factors and the maturation of pre-mRNAs in promoter regions, although not specifically for genes occur, at least in part, cotranscriptionally and is enhanced displaying AS regulated by EWS [51]. Moreover, the partial by RNAPII and by its phosphorylation [65]. On one hand, decrease in ChIP signals upon RNase treatment suggests that transcriptional coregulators are recruited by transcription EWS associates with its target transcripts cotranscriptionally factors to their target genes thus affecting splicing decisions; (Figure 2(a))[51]. on the other hand, transcriptional coregulators can modulate Collectively these observations confirm the dual function the rate of transcription elongation, which in turn affects AS for EWS in transcription and RNA processing and suggest a decisions [65]. Thus, recruitment of splicing factors through potential role for this protein in the coupling between the two interactions with the transcription machinery, as for EWS, is processes. However, the specific mechanism by which EWS one mechanism by which coupling between transcription and exerts splicing regulation has not been unraveled yet. AS occurs. Consistent with this potential dual role, EWS has been showntoregulatecyclinD1transcriptsbothtranscriptionally 8. EWS and Noncoding RNAs andatthelevelofsplicing[66]. Human cyclin D1 (CCND1 gene) is expressed as two isoforms derived by AS, termed EWSprotein,aswellastheothermembersofTETfamily,has D1a and D1b, which differ for the inclusion of intron 4 in the been identified as part of a multiprotein complex associated D1b mRNA [67]. Both isoforms are frequently upregulated with DROSHA [70], the RNase III endonuclease assigned to in human cancers, but cyclin D1b displays relatively higher the cleavage of the primary microRNA (pri-miRNA) [71]. oncogenic potential [68]. It has been shown that EWS mod- DROSHA is a component of two multiprotein complexes. The ulates AS by altering RNAPII dynamics (phosphorylation larger complex contains multiple classes of RBPs including and speed) over the CCND1 gene [69]. EWS increases the RNA helicases, proteins that bind double-stranded RNA, speed of elongating RNAPII over the CCND1 gene, and several hnRNPs, and, as mentioned above, the TET family EWS depletion decreases RNAPII occupancy, but not Serine- of RBPs. The smaller complex, composed by DROSHA and 󸀠 5 phosphorylation at the 5 end of the CCND1 gene, thus the double-stranded-RNA-binding protein DGCR8, forms decreasing cyclin D1a but not D1b mRNA levels [66]. the microprocessor complex, necessary and sufficient for the Changes in RNAPII elongation rates modulate the timing genesis of miRNAs from the pri-miRNA transcripts [70]. at which competing splice sites become available, and it is Since DROSHA is required for miRNA biogenesis and its another coupling mechanism that could possibly be used activity can be modulated by regulatory proteins, TET pro- by EWS. The interaction of EWS with some of the earliest teins might play a role in miRNA expression by modulating 󸀠 󸀠 factors involved in 5 and 3 splice site recognition (i.e., U1 the activity of this processing enzyme. Interestingly, some snRNP and BBP/SF1 [58, 60]) opens the possibility that EWS evidence has recently linked EWS to microRNA processing. establishes initial links between splice sites across introns Depletion of EWS in osteosarcoma U2OS cells results in or exons that contribute to splice site selection and exon the accumulation of precursor let-7g and downregulation of definitionFigure ( 2(a)). mature let-7g [72].Thesefindingsleadtothehypothesisthat More recently, two studies have directly linked EWS EWS might play a role in miRNA biogenesis and maturation. function to AS of specific genes in cancer cells. EWS was Nevertheless, wide spectrum analyses of the impact of EWS showntoregulateASofthep53repressor(MouseDou- on miRNA expression are still lacking; thus its contribution ble Minute 2, Human Homolog) MDM2 [64]andseveral to this process is still largely unknown. other genes involved in the DNA damage response (DDR) In addition, TET proteins have been recently involved [51]. Camptothecin (CPT) treatment inhibits the interac- in the regulation of other classes of noncoding RNAs. It has tion between EWS (associated with RNAPII) and YB-1 (a been shown that the interaction of FUS/TLS with CBP/p300 spliceosome-associated factor), thus resulting in general exon is induced by ncRNAs transcribed from the promoter regions skipping both in MDM2 and in other genes. These AS of CCND1 gene. These ncRNAs bind to FUS/TLS and events parallel the events induced either by EWS or YB-1 inhibit CBP/p300 histone acetyltransferase (HAT) activity knockdown (Figure 2(b),leftscheme)[64], suggesting that [73]. Similarly, EWS and TAF15 were found to bind to CBP 6 International Journal of Cell Biology

CPT treatment UV irradiation RNAPII

EWS EWS RNAPII TFIID YB-1 EWS Spliceosomeeosome

RNAPII

EWS YB-1 SpliceosomeSpliceos

EWS and YB-1 regulated AS events

(a) (b)

Figure 2: Schematic representation of the cross-talk between the transcriptional and splicing machineries mediated by RNAPII, EWS, and YB-1. (a) In normal condition EWS associates with the preinitiation complex TFIID, with the RNAPII and with its target transcripts and associated genomic regions; once the transcription has started, EWS jumps on the nascent RNA promoting specific AS events. (b) Upon CPT treatment EWS dissociates from YB-1 and from the spliceosome [64]. Similarly, upon low intensity UV light irradiation EWS dissociates from active transcription sites and relocalizes to nucleoli [51]. As a result, EWS regulated AS events cannot occur [51, 64].

and to exert inhibitory effects on CBP/p300 HAT activities recombination and/or repair process. Homologous recombi- [73]. The binding of EWS and FUS/TLS to G-quadruplex nation during meiosis is a physiological process that requires structures is the key of this regulation. Upon binding to DNA double-strand breaks and repair. It is essential for the GGUG RNA oligonucleotides in ncRNAs, the inhibitory proper segregation of chromosomes and it is regulated by activity of FUS/TLS against the HAT activity of CBP/p300 is most of the genes that are also involved in DNA repair enhanced. Biochemical experiments showed that the carboxy after genotoxic stress, as revealed by the gametogenesis terminus of FUS/TLS binds to GGUG RNA, whereas the defects observed in several knockout mouse models of genes N-terminus interacts with CBP. The interaction of FUS/TLS involved in the DDR [78]. It was reported that in the absence with CBP/p300 is stimulated when FUS/TLS is bound to G- of EWS the meiotic crossovers are conspicuously reduced or quadruplex, probably due to a conformational change, and even absent in some of the bivalents. This finding demon- this in turn results in inhibition of CBP/p300 HAT activity strated that EWS is critically important for the homologous [73]. recombination during meiosis and suggested its possible Similar to FUS/TLS, EWS can bind to TERRA [48, 53], a involvement in the DDR pathway [23]. Indeed, consistent large noncoding RNA component of telomeric heterochro- with the hypothesis, Ews-null mice and mouse embryonic matin, which inhibits the telomerase activity [74]. Also in fibroblasts (MEFs) showed increased sensitivity to ionizing this case, the interaction occurs through the G-quadruplex radiation [23]. in a structure-specific manner [53]. Thus, although not A physiological function for EWS in response to genomic investigated in detail yet, binding of EWS to ncRNAs appears alterations is consistent with two recent genetic screens that to contribute to its function in gene expression regulation. identified EWSR1 as a gene required for resistance to ionizing radiations (IR) [79], which release intermediary ions and free radicals that cause DNA lesions, and to CPT, a topoisomerase 9. EWS and Genotoxic Stress I inhibitor which forms a tight complex with topoisomerase I-DNA adducts and prevents DNA religation, thus leading Genomic integrity is critical to cell survival and is controlled to formation of single-strand breaks (SSBs) [80]. Further by the DDR network, an elaborate signal transduction system evidence in support of its involvement in the DDR is the that senses DNA damage and recruits appropriate repair reported phosphorylation of EWS on threonine 79 upon factors [75]. DNA damage [81]. This posttranslational modification occurs As mentioned above, Ews deficiency in mice leads to in response to both mitogens and DNA alkylating agents and developmental defects in meiosis and pre-B cell forma- is catalyzed by mitogen-activated protein kinases (MAPKs); tion [23]. Interestingly, defects in lymphocyte development, however, while the phosphorylation upon mitogen signals meiosis, and cellular senescence (or aging) are features is operated by the ERK1/2 (extracellular signal-regulated also observed in mice deficient in Atm and c-abl [76, 77], kinases 1 and 2), DNA damage induced phosphorylation two genes that are critical for the DDR pathway. These of EWS is catalyzed by the JNK1/2 (extracellular signal- observations implicate a possible role of EWS in the DNA regulated kinases 1 and 2) [81]. Interestingly, phosphorylation International Journal of Cell Biology 7 on threonine 79 occurs also in the oncogenic EWS-FLI1 possible that it plays a direct role in DNA recombination and protein upon treatment with DNA alkylating agents and repair. is operated by p38𝛼/p38𝛽 MAPK [81]; nevertheless, the functional relevance of this phosphorylation has not been unraveled yet. 10. EWS and Neuromuscular Disorders Moreover, EWS protein has also been described to inter- actwithBARD1,aputativetumorsuppressorcomponentof Regulation of RNA metabolism in the nervous system plays a the BRCA1/BARD1 complex that performs essential DNA crucial role, as highlighted by the development of devastating repair and recombination functions [82]. Altogether, the neurological diseases, including spinal muscular atrophy and observations reported above suggest that, in addition to certain trinucleotide repeat expansions, upon mutation or its involvement in endogenous processes of homologous misregulation of key RBPs. Mutations in FUS/TLS, as well recombination, EWS plays also a role in DDR triggered by as in the RBP TAR DNA-binding protein 43 (TDP-43), were genotoxic stresses. found to be causative of familial cases of amyotrophic lateral In addition to the well-described activation of a signal sclerosis (ALS) [15, 16, 86], further implicating altered RNA transduction pathway that regulates cell cycle progression metabolism as a central feature of these diseases. FUS/TLS and apoptosis, low-intensity UV light induces also extensive protein was also reported to be the major component of poly- changesinASdriven,atleastinpart,byhyperphospho- Q aggregates in cellular models of spinal cerebellar ataxia rylation of RNAPII carboxyterminal domain (CTD) and and in intracellular inclusions in neurons of patients with reduced transcription elongation [83]. Moreover, we have Huntington’s disease [87]. Indeed, TDP-43 and FUS/TLS recently documented that UV light influences AS of a subset [88], as well as TAF15 and EWS [86], are recruited in of genes involved in the DDR, such as BRCA1, CHEK2, and cytoplasmic aggregates in frontotemporal lobar degeneration ABL1, through regulation of the activity of EWS [51]. DNA (FTLD). Thus, aggregation of RBPs is emerging as a crucial repair and checkpoint responses are critically important to event in clinicopathological conditions of neurodegenerative ensure the integrity of the genome. Checkpoint pathways diseases. Since all these RBPs are directly involved in RNA are composed of sensors that detect damaged or unrepli- processing and utilization, these observations highlight RNA cated DNA, transducers that convey the signal, and effector dysmetabolism as potential key step underlying neurode- proteins that act on the ultimate targets of the checkpoint. generative diseases. Intriguingly, many of these RBPs that CHK2 (the protein product of the CHEK2 gene) is an effector aggregate in pathological inclusions are expressed broadly, protein conserved in eukaryotes and phosphorylated by the some even ubiquitously, but they seem to only aggregate in sensor kinases ATM/ATR in response to DNA damage; upon certain neurons and not others, remarking cell type specificity phosphorylation by ATM/ATR, CHK2 phosphorylates and [89]. thus modifies the function of key targets of the check- Most disease-causing mutations in TDP-43 and some point response [84]. Interestingly, upon EWS knockdown an disease-causing mutations in FUS/TLS are found within mRNA isoform of CHEK2 is generated that lacks the exon the glycine-rich domain, which is rich in uncharged, polar 2 containing the translation initiation codon, thus leading amino acids, similar to that observed in the nucleation to a decrease of the levels of CHK2 protein [81]. These domains found in many yeast prions [90]. This has led to findings suggest that proper levels and activity of EWS are speculation that a prion-like aggregation property of some important to maintain CHK2 expression and, given the key RBPs may contribute to onset and/or progression of ALS role of this kinase in DNA damage response, for the initial and related neurodegenerative diseases [91]. The majority cellular response to genotoxic stress. EWS knockdown could of yeast prion proteins contain a distinctive prion domain therefore lead to defective CHK2 response and explain the that is enriched in uncharged polar amino acids (particularly higher sensitivity to UV irradiation. asparagine, glutamine, and tyrosine) and glycine [92, 93]. An Interestingly, UV light induces dissociation of EWS from algorithm designed to detect prion domains has identified 19 sites of active transcription, in particular from alternatively domains that can confer prion behavior [90]. Scouring the spliced regions regulated by this protein, and its translocation human genome with this algorithm enriches a select group to the nucleoli [51]. Similar translocation to the nucleolar of RBPs harboring a canonical RRM and a putative prion compartment has been described also for FUS/TLS upon domain. Indeed, of the 210 human RRM-bearing proteins, inhibition of transcription [85]. As mentioned above, EWS is 29 have a putative prion domain and these prion RBPs also involved in other genotoxic stress-mediated changes in are emerging, one by one, in the pathology of devastating AS regulation. It was shown that treatment with CPT leads to neurodegenerative disorders [91, 94]. substantialexonskippingandthatasubsetofthetriggered Similar to TDP-43, FUS/TLS, and TAF15, expression of AS events could be recapitulated by knockdown of EWS or of EWSR1 cDNA in yeast resulted in cytoplasmic aggregation its interacting protein YB-1 [64]. and toxicity in a functional screen [94]. Furthermore, like Collectively, these observations highlight the role of EWS TDP-43, FUS/TLS, and TAF15, EWSR1 also harbors a bioin- in the DDR and suggest that this protein acts, at least in formatically predicted prion-like domain. Since most of the part,throughtheregulationofASofgenesinvolvedinthe pathogenic mutations in FUS/TLS are located in the C- activation of the DNA damage checkpoint and for the repair terminal domain of the protein, the last four exons of the of DNA lesions. Nevertheless, since EWS binds DNA and EWSR1 gene (exons 15–18) were sequenced in 817 individuals it is found associated with the chromatin [23, 51], it is also diagnosedwithALSandin1082healthypopulationcontrol 8 International Journal of Cell Biology individuals. This approach identified two patient-specific EWSR1 knockdown in HeLa cells resulted also in a signif- missense variants in exon 16 (1532G>C, G511A, correspond- icantly high incidence of abnormal spindles and mislocaliza- ingtothefirstRGGdomain)andinexon17(1655C>T, P552L, tion of Aurora B kinase, the enzymatic core of the chromo- corresponding to the second RGG domain) of EWSR1 in two somal passenger complex (CPC), which orchestrates in space unrelated ALS patients with sporadic disease. Interestingly, and time chromosome alignment, histone modification, and these variants alter EWS localization in motor neurons and cytokinesis [98]. Proper localization of the CPC complex is exhibit enhanced aggregation propensity [95]. Nevertheless, key to the precise control of chromosome segregation over theroleplayedbyEWSinALSandinotherneurological mitosis, and dysregulation of Aurora kinase activity has been diseases still remains unveiled, and further studies are needed linked to tumorigenesis [99]. Indeed, the absence of Aurora to identify the molecular targets affected in these pathological B results in increased numbers of aneuploid cells, genetic conditions. Notably, whether EWS RNA binding affinity and instability, and oncogenic transformation [99]. pre-mRNA regulation are affected in the ALS-related mutants On the other hand, the oncogenic potential of EWS in the is also still unknown. chimeric protein is driven by its N-terminal transactivation domain, which is partially inhibited by the RGG domains in the C-terminal part of the full-length protein. So the C-terminal part of EWS protein, missing in the fusion 11. EWS and Cancer oncogenes, could be strategically relevant in protecting cells from cancer either by playing a key role in the regulation of Much knowledge about the function of EWS is derived AS of DDR genes or by inhibiting the DNA activation domain from studies of oncogenic fusion proteins involved in sev- or through both mechanisms. eral malignancies and arising from in-frame chromosomal Interestingly, it has been shown that in the nucleus fusions with multiple partners. However, besides the onco- EWS interacts with STRAP (serine-threonine kinase genic translocations, an independent role for EWS in cancer receptor-associated protein), a WD40 domain-containing has not been unraveled yet. protein. EWS-STRAP interaction is strategically important Gene expression in cancer cells is deregulated at both for STRAP-induced inhibition of EWS transactivation thetranscriptionandsplicinglevels,andtherearemany function. Strikingly, STRAP is upregulated in colon and examples of oncogenic and cancer-associated splice variants lung carcinomas [100]andSTRAPupregulationincolon [96]. As described above, the regulation of specific splicing tumors correlates with EWS expression [101]. Furthermore, events and transcriptional events by EWS might be involved STRAP blocks EWS-induced p300-mediated expression of in the altered transcriptome of cancer cells [51, 64, 66, 69]. c-fos by interfering with EWS-p300 interaction, inhibiting Cancer progression is thought to be dependent on the in this way EWS-mediated transactivation [101]. Thus, the accumulation of mutations that change the transcriptional cooperation between EWS and STRAP could be involved profile of the cell to support its escape from the tight in tumor progression and interfering with EWS-mediated regulation of cell cycle progression. Improper responses to transactivation could be a mechanism of regulation of its different types of DNA lesions may lead to accumulation transcriptional properties in human cancers. of mutations in the genome, which then accelerate the progression of the disease. Thus, given the reported role of 12. Conclusive Remarks EWSingenomestability,itispossiblethatthisRBPisalso essential to preserve the genome integrity and protect cells EWS protein is involved in the regulation of a wide range from neoplastic transformation. of cellular processes to ensure genome integrity and proper Accordingly, EWS plays a critical role in the DDR [51, pursuance of cellular functions. Early studies suggested a 64], and EWSR1 gene has been shown to be essential for a role for EWS in transcription and splicing, possibly coupling proper response to genotoxic agents [23, 79, 80]. In particular, these two processes. Since then, many studies have unraveled EWS regulates AS of genes playing a key role in oncogenesis these functions and unveiled new roles for EWS protein in [51]likeCHEK2 (described above), BRCA1 (breast cancer 1 other aspects of the regulation of gene expression and RNA gene), and ABL1. BRCA1 is a tumor suppressor and germline metabolism. Importantly, some RNA targets and new inter- mutations in BRCA1 gene predispose individuals to breast acting proteins have now been identified [102]. Moreover, and ovarian cancers [97]. Interestingly, BRCA1 gene generates new insights have been provided on the regulation of the several splice variants that play an important role in tumor amino terminus of EWS, which is involved in oncogenic development [97]. Mutations have been described in BRCA1 fusion proteins and responsible for the inappropriate tran- gene that result in skipping of more than one exons leading scriptional activation of target genes. either to shorter transcripts or to frameshift and premature More recently, novel potential functions have been sug- stop codons, thus rendering mis-spliced mRNAs subject to gested for EWS protein in cancer and in neuromuscular non-mediated decay (NMD). Thus, these mutations yield disorders, disclosing novel roles for the protein. Nevertheless, either a truncated BRCA1 protein or loss of a BRCA1 there are still many important unsolved questions concerning transcript. As a result the functionality of the resulting the molecular and cellular biology of EWS and further BRCA1 protein is severely compromised. Remarkably, EWS knowledge is needed to understand if EWS represents a knockdowninHeLacellsresultsinaberrantisoformsof suitable target for the development of new approaches to BRCA1 gene lacking the exons from 12 to 17 [51]. cancer therapy. International Journal of Cell Biology 9

Acknowledgment [16] D. A. Bosco, N. Lemay, H. K. Ko et al., “Mutant FUS proteins that cause amyotrophic lateral sclerosis incorporate into stress Work in the laboratory of M. P. Paronetto was supported by granules,” Human Molecular Genetics,vol.19,no.21,pp.4160– Grants from the “Associazione Italiana Ricerca sul Cancro” 4175, 2010. (AIRC) (MFAG 11658). [17] B. Plougastel, J. Zucman, M. Peter, G. Thomas, and O. Delattre, “Genomic structure of the EWS gene and its relationship to EWSR1, a site of tumor-associated chromosome translocation,” References Genomics,vol.18,no.3,pp.609–615,1993. ˚ [1] A. Aurias, C. Rimbaut, and D. Buffe, “Translocation involving [18] P. Aman,I.Panagopoulos,C.Lassenetal.,“Expressionpatterns chromosome 22 in Ewing’s sarcoma. A cytogenetic study of four of the human sarcoma-associated genes FUS and EWS and the fresh tumors,” Cancer Genetics and Cytogenetics,vol.12,no.1, genomic structure of FUS,” Genomics,vol.37,no.1,pp.1–8,1996. pp.21–25,1984. [19] A. Crozat, P. Aman, N. Mandahl, and D. Ron, “Fusion of CHOP to a novel RNA-binding protein in human myxoid [2]O.Delattre,J.Zucman,B.Plougasteletal.,“Genefusion liposarcoma,” Nature,vol.363,no.6430,pp.640–644,1993. with an ETS DNA-binding domain caused by chromosome translocation in human tumours,” Nature, vol. 359, no. 6391, pp. [20] T.Melot, L. Dauphinot, N. Sevenet,´ F.Radvanyi, and O. Delattre, 162–165, 1992. “Characterization of a new brain-specific isoform of the EWS oncoprotein,” EuropeanJournalofBiochemistry,vol.268,no.12, [3] J. Zucman, O. Delattre, C. Desmaze et al., “EWS and ATF- pp. 3483–3489, 2001. 1 gene fusion induced by t(12;22) translocation in malignant [21] T. Ota, Y.Suzuki, T. Nishikawa et al., “Complete sequencing and melanoma of soft parts,” Nature Genetics,vol.4,no.4,pp.341– characterization of 21, 243 full-length human cDNAs,” Nature 345, 1993. Genetics,vol.36,no.1,pp.40–45,2004. [4] J. Zucman, T. Melot, C. Desmaze et al., “Combinatorial genera- [22] H. Kovar, G. Jug, C. Hattinger et al., “The EWS protein is tion of variable fusion proteins in the Ewing family of tumours,” dispensable for Ewing tumor growth,” Cancer Research,vol.61, EMBO Journal, vol. 12, no. 12, pp. 4481–4487, 1993. no.16,pp.5992–5997,2001. [5] N. J. Balamuth and R. B. Womer, “Ewing’s sarcoma,” The Lancet [23]H.Li,W.Watford,C.Lietal.,“EwingsarcomageneEWSis Oncology, vol. 11, no. 2, pp. 184–192, 2010. essential for meiosis and B lymphocyte development,” Journal [6] H. Kovar, “Downstream EWS/FLI1—upstream Ewing’s sar- of Clinical Investigation,vol.117,no.5,pp.1314–1323,2007. coma,” Genome Medicine,vol.2,no.1,article8,2010. [24] G. G. Hicks, N. Singh, A. Nashabi et al., “Fus deficiency in mice [7] J. L. Ordo´nez,˜ D. Osuna, D. Herrero, E. De Alava,´ and J. Madoz- results in defective B-lymphocyte development and activation, Gurpide,´ “Advances in Ewing’s sarcoma research: where are we high levels of chromosomal instability and perinatal death,” now and what lies ahead?” Cancer Research,vol.69,no.18,pp. Nature Genetics,vol.24,no.2,pp.175–179,2000. 7140–7150, 2009. [25]M.Kuroda,J.Sok,L.Webbetal.,“Malesterilityandenhanced [8] S. L. Lessnick and M. Ladanyi, “Molecular pathogenesis of radiation sensitivity in TLS(-/-) mice,” EMBO Journal,vol.19, Ewing sarcoma: new therapeutic and transcriptional targets,” no.3,pp.453–462,2000. Annual Review of Pathology,vol.7,pp.145–159,2012. [26] T. R. Mosmann and R. L. Coffman, “TH1 and TH2 cells: [9]I.R.A.Mackenzie,R.Rademakers,andM.Neumann,“TDP- different patterns of lymphokine secretion lead to different 43 and FUS in amyotrophic lateral sclerosis and frontotemporal functional properties,” Annual Review of Immunology,vol.7,pp. dementia,” The Lancet Neurology,vol.9,no.10,pp.995–1007, 145–173, 1989. 2010. [27] J. Cho, H. Shen, H. Yu et al., “Ewing sarcoma gene Ews regulates [10] C. Vance, B. Rogelj, T. Hortobagyi´ et al., “Mutations in FUS, hematopoietic stem cell senescence,” Blood,vol.117,no.4,pp. an RNA processing protein, cause familial amyotrophic lateral 1156–1166, 2011. sclerosis type 6,” Science,vol.323,no.5918,pp.1208–1211,2009. [28] A.Bertolotti,T.Melot,J.Acker,M.Vigneron,O.Delattre,andL. Tora, “EWS, but not EWS-FLI-1, is associated with both TFIID [11] F. Morohoshi, Y. Ootsuka, K. Arai et al., “Genomic structure of and RNA polymerase II: interactions between two members of the human RBP56/hTAF(II)68 and FUS/TLS genes,” Gene,vol. the tet family, EWS and HTAF(II)68, and subunits of TFIID and 221, no. 2, pp. 191–198, 1998. RNA polymerase II complexes,” Molecular and Cellular Biology, [12] K. L. Rossow and R. Janknecht, “The Ewing’s sarcoma vol. 18, no. 3, pp. 1489–1497, 1998. gene product functions as a transcriptional activator,” Cancer [29] A. Bertolotti, Y. Lutz, D. J. Heard, P. Chambon, and L. Tora, Research,vol.61,no.6,pp.2690–2695,2001. “hTAF(II)68, a novel RNA/ssDNA-binding protein with homol- [13]M.K.Andersson,A.Stahlberg,˚ Y. Arvidsson et al., “The mul- ogy to the pro-oncoproteins TLS/FUS and EWS is associated tifunctional FUS, EWS and TAF15 proto-oncoproteins show with both TFIID and RNA polymerase II,” EMBO Journal,vol. cell type-specific expression patterns and involvement in cell 15,no.18,pp.5022–5031,1996. spreading and stress response,” BMC Cell Biology,vol.9,article [30] H. Zhou and K. A. W.Lee, “AnhsRPB4/7-dependent yeast assay 37, 2008. for trans-activation by the EWS oncogene,” Oncogene,vol.20, [14] J. S. Felsch, W.S. Lane, and E. G. Peralta, “Tyrosine kinase Pyk2 no.12,pp.1519–1524,2001. mediates G-protein-coupled receptor regulation of the Ewing [31] J. Lee, B. K. Rhee, G.-Y.Bae, Y.-M. Han, and J. Kim, “Stimulation sarcoma RNA-binding protein EWS,” Current Biology,vol.9,no. of Oct-4 activity by Ewing’s sarcoma protein,” Stem Cells,vol.23, 9, pp. 485–488, 1999. no. 6, pp. 738–751, 2005. [15] T. J. Kwiatkowski Jr., D. A. Bosco, A. L. LeClerc et al., “Mutations [32] D. M. Gascoyne, G. R. Thomas, and D. S. Latchman, “The in the FUS/TLS gene on chromosome 16 cause familial amy- effects of Brn-3a on neuronal differentiation and apoptosis are otrophic lateral sclerosis,” Science,vol.323,no.5918,pp.1205– differentially modulated by EWS and its oncogenic derivative 1208, 2009. EWS/Fli-1,” Oncogene,vol.23,no.21,pp.3830–3840,2004. 10 International Journal of Cell Biology

[33] G. R. Thomas and D. S. Latchman, “The pro-oncoprotein [49] S. Ishii, F. Imamoto, and Y. Yamanashi, “Characterization of EWS (Ewing’s sarcoma protein) interacts with the Brn-3a the promoter region of the human c-erbB-2 protooncogene,” POU transcription factor and inhibits its ability to activate Proceedings of the National Academy of Sciences of the United transcription,” Cancer Biology and Therapy,vol.1,no.4,pp.428– States of America, vol. 84, no. 13, pp. 4374–4378, 1987. 432, 2002. [50] P. Pieler and O. Theunissen, “TFIIA: nine fingers—three [34] Y. Fujimura, H. Siddique, L. Lee, V. N. Rao, and E. S. P. Reddy, hands?” Trends in Biochemical Sciences,vol.18,no.6,pp.226– “EWS-ATF-1 chimeric protein in soft tissue clear cell sarcoma 230, 1993. associates with CREB-binding protein and interferes with p53- [51] M. P. Paronetto, B. Minana,˜ and J. Valcarcel,´ “The Ewing mediated trans-activation function,” Oncogene, vol. 20, no. 46, sarcoma protein regulates DNA damage-induced alternative pp. 6653–6659, 2001. splicing,” Molecular Cell,vol.43,no.3,pp.353–368,2011. [35] N. Araya, K. Hirota, Y. Shimamoto et al., “Cooperative inter- [52]J.C.Deloulme,L.Prichard,O.Delattre,andD.R.Storm,“The action of EWS with CREB-binding protein selectively activates prooncoprotein EWS binds calmodulin and is phosphorylated hepatocyte nuclear factor 4-mediated transcription,” Journal of by protein kinase C through an IQ domain,” Journal of Biological Biological Chemistry,vol.278,no.7,pp.5427–5432,2003. Chemistry,vol.272,no.43,pp.27369–27377,1997. [36]R.-A.Bailly,R.Bosselut,J.Zucmanetal.,“DNA-bindingand [53]K.Takahama,K.Kino,S.Arai,R.Kurokawa,andT.Oyoshi, transcriptional activation properties of the EWS-FLI-1 fusion “Identification of Ewing’s sarcoma protein as a G-quadruplex protein resulting from the t(11;22) translocation in Ewing DNA- and RNA-binding protein,” FEBS Journal,vol.278,no.6, sarcoma,” Molecular and Cellular Biology,vol.14,no.5,pp. pp. 988–998, 2011. 3230–3241, 1994. [54] N.Araya,H.Hiraga,K.Kako,Y.Arao,S.Kato,andA.Fukamizu, [37]K.K.C.LiandK.A.W.Lee,“Transcriptionalactivationbythe “Transcriptional down-regulation through nuclear exclusion Ewing’s sarcoma (EWS) oncogene can be cis-repressed by the of EWS methylated by PRMT1,” Biochemical and Biophysical EWS RNA-binding domain,” Journal of Biological Chemistry, Research Communications,vol.329,no.2,pp.653–660,2005. vol.275,no.30,pp.23053–23058,2000. [38]D.AlexandK.A.W.Lee,“RGG-boxesoftheEWSoncoprotein [55] M. L. Hastings and A. R. Krainer, “Pre-mRNA splicing in the repress a range of transcriptional activation domains,” Nucleic new millennium,” Current Opinion in Cell Biology,vol.13,no.3, Acids Research,vol.33,no.4,pp.1323–1331,2005. pp. 302–309, 2001. [39] C. G. Burd and G. Dreyfuss, “Conserved structures and diver- [56] J. Rappsilber, U. Ryder, A. I. Lamond, and M. Mann, “Large- sity of functions of RNA-binding proteins,” Science,vol.265,no. scale proteomic analysis of the human spliceosome,” Genome 5172, pp. 615–621, 1994. Research,vol.12,no.8,pp.1231–1245,2002. [40] A. Bertolotti, B. Bell, and L. Tora, “The N-terminal domain [57] W.Hackl and R. Luhrmann,¨ “Molecular cloning and subcellular of human TAF(II)68 displays transactivation and oncogenic localisation of the snRNP-associated protein 69KD, a structural properties,” Oncogene,vol.18,no.56,pp.8000–8010,1999. homologue of the proto-oncoproteins TLS and EWS with RNA and RNA-binding properties,” Journal of Molecular Biology,vol. [41] M. Ladomery and G. Dellaire, “Multifunctional zinc finger pro- 264, no. 5, pp. 843–851, 1996. teins in development and disease,” Annals of Human Genetics, vol.66,no.5-6,pp.331–342,2002. [58] L. L. Knoop and S. J. Baker, “The splicing factor U1C [42] J. W. R. Schwabe and D. Rhodes, “Beyond zinc fingers: steroid represses EWS/FLI-mediated transactivation,” Journal of Bio- hormonereceptorshaveanovelstructuralmotifforDNA logical Chemistry,vol.275,no.32,pp.24865–24871,2000. [59] S. Michaud and R. Reed, “A functional association between the recognition,” Trends in Biochemical Sciences,vol.16,no.8,pp. 󸀠 󸀠 291–296, 1991. 5 and 3 splice sites is established in the earliest prespliceosome [43] T. Ohno, M. Ouchida, L. Lee, Z. Gatalica, V. N. Rao, and complex (E) in mammals,” Genes and Development,vol.7,no.6, E. S. P. Reddy, “The EWS gene, involved in Ewing family of pp.1008–1020,1993. tumors, malignant melanoma of soft parts and desmoplastic [60] D. Zhang, A. J. Paley, and G. Childs, “The transcriptional small round cell tumors, codes for an RNA binding protein with repressor, ZFM1 interacts with and modulates the ability of EWS novel regulatory domains,” Oncogene,vol.9,no.10,pp.3087– to activate transcription,” JournalofBiologicalChemistry,vol. 3097, 1994. 273, no. 29, pp. 18086–18091, 1998. [44] C. Tuerk and L. Gold, “Systemic evolution of ligands by [61] S. Wu and M. R. Green, “Identification of a human protein that exponential enrichment: RNA ligands to bacteriophage T4 recognizes the 3’ splice site during the second step of pre-mRNA DNA polymerase,” Science,vol.249,no.4968,pp.505–510,1990. splicing,” EMBO Journal,vol.16,no.14,pp.4421–4432,1997. [45] A. Lerga, M. Hallier, L. Delva et al., “Identification of an RNA [62] L. Yang, H. A. Chansky, and D. D. Hickstein, “EWS⋅Fli-1 fusion binding specificity for the potential splicing factor TLS,” Journal protein interacts with hyperphosphorylated RNA polymerase of Biological Chemistry,vol.276,no.9,pp.6807–6816,2001. II and interferes with serine-arginine protein-mediated RNA [46] A. Ramos, D. Hollingworth, and A. Pastore, “G-quartet- splicing,” JournalofBiologicalChemistry,vol.275,no.48,pp. dependent recognition between the FMRP RGG box and RNA,” 37612–37618, 2000. RNA,vol.9,no.10,pp.1198–1207,2003. [63] H.A.Chansky,M.Hu,D.D.Hickstein,andL.Yang,“Oncogenic [47] L. A. Hanakahi, H. Sun, and N. Maizels, “High affinity interac- TLS/ERG and EWS/Fli-1 fusion proteins inhibit RNA splicing tions of nucleolin with G-G-paired rDNA,” Journal of Biological mediated by YB-1 protein,” Cancer Research,vol.61,no.9,pp. Chemistry,vol.274,no.22,pp.15908–15912,1999. 3586–3590, 2001. [48] T. Oyoshi and R. Kurokawa, “Structure of noncoding RNA [64] M. Dutertre, G. Sanchez, M.-C. De Cian et al., “Cotranscrip- is a determinant of function of RNA binding proteins in tional exon skipping in the genotoxic stress response,” Nature transcriptional regulation,” Cell & Bioscience,vol.2,no.1,article Structural and Molecular Biology, vol. 17, no. 11, pp. 1358–1366, 1, 2012. 2010. International Journal of Cell Biology 11

[65] D. L. Bentley, “Rules of engagement: co-transcriptional recruit- BARD1 links the Ewing’s sarcoma gene to a common tumor ment of pre-mRNA processing factors,” Current Opinion in Cell suppressor pathway,” Cancer Research,vol.62,no.16,pp.4583– Biology,vol.17,no.3,pp.251–256,2005. 4587, 2002. [66] G. Sanchez, D. Bittencourt, K. Laud et al., “Alteration of [83] M. J. Munoz,˜ M. S. P. Santangelo, M. P. Paronetto et al., “DNA cyclin D1 transcript elongation by a mutated transcription damage regulates alternative splicing through inhibition of factor up-regulates the oncogenic D1b splice isoform in cancer,” RNA polymerase II elongation,” Cell,vol.137,no.4,pp.708–720, Proceedings of the National Academy of Sciences of the United 2009. States of America, vol. 105, no. 16, pp. 6004–6009, 2008. [84] J. Falck, N. Mailand, R. G. Syljuasen,˚ J. Bartek, and J. Lukas, [67] D. C. Betticher, N. Thatcher, H. J. Altermatt, P. Hoban, W. D. “The ATM-Chk2-Cdc25A checkpoint pathway guards against J. Ryder, and J. Heighway, “Alternate splicing produces a novel radioresistant DNA synthesis,” Nature,vol.410,no.6830,pp. cyclin D1 transcript,” Oncogene,vol.11,no.5,pp.1005–1011,1995. 842–847, 2001. [68] K. E. Knudsen, J. Alan Diehl, C. A. Haiman, and E. S. Knudsen, [85] H. Zinszner, D. Immanuel, Y. Yin, F.-X. Liang, and D. Ron, “A “Cyclin D1: polymorphism, aberrant splicing and cancer risk,” topogenic role for the oncogenic N-terminus of TLS: nucleolar Oncogene,vol.25,no.11,pp.1620–1628,2006. localization when transcription is inhibited,” Oncogene,vol.14, [69] G. Sanchez, O. Delattre, D. Auboeuf, and M. Dutertre, “Coupled no. 4, pp. 451–461, 1997. alteration of transcription and splicing by a single oncogene: [86] M. Neumann, E. Bentmann, D. Dormann et al., “FET proteins boostingtheeffectoncyclinD1activity,”Cell Cycle,vol.7,no. TAF15 and EWS are selective markers that distinguish FTLD 15, pp. 2299–2305, 2008. with FUS pathology from amyotrophic lateral sclerosis with [70] R. I. Gregory, K.-P.Yan, G. Amuthan et al., “The Microprocessor FUS mutations,” Brain,vol.134,no.9,pp.2595–2609,2011. complex mediates the genesis of microRNAs,” Nature,vol.432, [87] H. Doi, K. Okamura, P. O. Bauer et al., “RNA-binding protein no.7014,pp.235–240,2004. TLS is a major nuclear aggregate-interacting protein in Hunt- [71] Y. Lee, K. Jeon, J.-T. Lee, S. Kim, and V. N. Kim, “MicroRNA ingtin exon 1 with expanded polyglutamine-expressing cells,” maturation: stepwise processing and subcellular localization,” Journal of Biological Chemistry,vol.283,no.10,pp.6489–6500, EMBO Journal,vol.21,no.17,pp.4663–4670,2002. 2008. [72] E.J.Sohn,J.Park,S.I.Kang,andY.P.Wu,“Accumulationofpre- [88]E.D.Huey,R.Ferrari,J.H.Morenoetal.,“FUSandTDP43 let-7g and downregulation of mature let-7g with the depletion genetic variability in FTD and CBS,” Neurobiology of Aging,vol. of EWS,” Biochemical and Biophysical Research Communica- 33, no. 5, pp. 1016.e9–1016.e17, 2012. tions,vol.426,no.1,pp.89–93,2012. [89] H. Ilieva, M. Polymenidou, and D. W. Cleveland, “Non-cell [73] X. Wang, S. Arai, X. Song et al., “Induced ncRNAs allosterically autonomous toxicity in neurodegenerative disorders: ALS and modify RNA-binding proteins in cis to inhibit transcription,” beyond,” JournalofCellBiology,vol.187,no.6,pp.761–772,2009. Nature,vol.454,no.7200,pp.126–130,2008. [90] M. Cushman, B. S. Johnson, O. D. King, A. D. Gitler, and [74] C. M. Azzalin, P. Reichenbach, L. Khoriauli, E. Giulotto, J. Shorter, “Prion-like disorders: blurring the divide between and J. Lingner, “Telomeric repeat-containing RNA and RNA transmissibility and infectivity,” Journal of Cell Science,vol.123, surveillance factors at mammalian chromosome ends,” Science, no. 8, pp. 1191–1201, 2010. vol. 318, no. 5851, pp. 798–801, 2007. [91] A. D. Gitler and J. Shorter, “RNA-binding proteins with prion- [75] J. W. Harper and S. J. Elledge, “The DNA damage response: ten like domains in ALS and FTLD-U,” Prion,vol.5,no.3,pp.179– years after,” Molecular Cell,vol.28,no.5,pp.739–745,2007. 187, 2011. [76]V.L.J.Tybulewicz,C.E.Crawford,P.K.Jackson,R.T.Bronson, [92] S. Alberti, R. Halfmann, O. King, A. Kapila, and S. Lindquist, and R. C. Mulligan, “Neonatal lethality and lymphopenia “Asystematic survey identifies prions and illuminates sequence in mice with a homozygous disruption of the c-abl proto- features of prionogenic proteins,” Cell,vol.137,no.1,pp.146–158, oncogene,” Cell,vol.65,no.7,pp.1153–1163,1991. 2009. [77]Y.XuandD.Baltimore,“DualrolesofATMinthecellular response to radiation and in cell growth control,” Genes and [93]J.A.Toombs,B.R.McCarty,andE.D.Ross,“Compositional Development,vol.10,no.19,pp.2401–2410,1996. determinants of prion formation in yeast,” Molecular and Cellular Biology,vol.30,no.1,pp.319–332,2010. [78] W. M. Baarends, R. van der Laan, and J. A. Grootegoed, “DNA repair mechanisms and gametogenesis,” Reproduction,vol.121, [94] J. Couthouisa, M. P. Harta, J. Shorter et al., “A yeast functional no. 1, pp. 31–39, 2001. screen predicts new candidate ALS disease genes,” Proceedings of the National Academy of Sciences of the United States of [79] K. E. Hurov, C. Cotta-Ramusino, and S. J. Elledge, “A genetic America,vol.108,no.52,pp.20881–20890,2011. screen identifies the Triple T complex required for DNA damage signaling and ATM and ATR stability,” Genes and Development, [95] J. Couthouis, M. P. Hart, R. Erion et al., “Evaluating the role vol.24,no.17,pp.1939–1950,2010. of the FUS/TLS-related gene EWSR1 in amyotrophic lateral [80] B. C. O’Connell, B. Adamson, J. R. Lydeard et al., “A genome- sclerosis,” Human Molecular Genetics,vol.21,no.13,pp.2899– wide camptothecin sensitivity screen identifies a mammalian 2911, 2012. MMS22L-NFKBIL2 complex required for genomic stability,” [96] J. P. Venables, “Unbalanced alternative splicing and its signifi- Molecular Cell,vol.40,no.4,pp.645–657,2010. cance in cancer,” BioEssays, vol. 28, no. 4, pp. 378–386, 2006. [81] I. V. Klevernic, S. Morton, R. J. Davis, and P. Cohen, “Phos- [97] T. I. Orban and E. Olah, “Emerging roles of BRCA1 alternative phorylation of Ewing’s sarcoma protein (EWS) and EWS-Fli1 in splicing,” Molecular Pathology,vol.56,no.4,pp.191–197,2003. response to DNA damage,” Biochemical Journal,vol.418,no.3, [98] M. Azuma, L. J. Embree, H. Sabaawy, and D. D. Hickstein, pp.625–634,2009. “Ewing sarcoma protein Ewsr1 maintains mitotic integrity and [82] L. Spahn, R. Petermann, C. Siligan, J. A. Schmid, D. N. T. Aryee, proneural cell survival in the zebrafish embryo,” PLoS ONE,vol. and H. Kovar, “Interaction of the EWS NH2 terminus with 2, no. 10, article e979, 2007. 12 International Journal of Cell Biology

[99] B. A. A. Weaver and D. W. Cleveland, “Decoding the links between mitosis, cancer, and chemotherapy: the mitotic checkpoint, adaptation, and cell death,” Cancer Cell,vol.8,no.1,pp. 7–12, 2005. [100] S. K. Halder, G. Anumanthan, R. Maddula et al., “Oncogenic function of a novel WD-domain protein, STRAP, in human carcinogenesis,” Cancer Research,vol.66,no.12,pp.6156–6166, 2006. [101]G.Anumanthan,S.K.Halder,D.B.Friedman,andP.K. Datta, “Oncogenic serine-threonine kinase receptor-associated protein modulates the function of ewing sarcoma protein through a novel mechanism,” Cancer Research,vol.66,no.22, pp. 10824–10832, 2006. [102]S.Pahlich,L.Quero,B.Roschitzki,R.P.Leemann-Zakaryan, and H. Gehring, “Analysis of Ewing Sarcoma (EWS)-binding proteins: interaction with hnRNP M, U, and RNA-helicases p68/72 within protein-RNA complexes,” Journal of Proteome Research,vol.8,no.10,pp.4455–4465,2009. Hindawi Publishing Corporation International Journal of Cell Biology Volume 2013, Article ID 568931, 9 pages http://dx.doi.org/10.1155/2013/568931

Review Article Regulation of the Ras-MAPK and PI3K-mTOR Signalling Pathways by Alternative Splicing in Cancer

Zahava Siegfried,1 Serena Bonomi,2 Claudia Ghigna,2 and Rotem Karni1

1 Department of Biochemistry and Molecular Biology, The Institute for Medical Research Israel-Canada, Hebrew University-Hadassah Medical School Ein Kerem, 91120 Jerusalem, Israel 2 Institute of Molecular Genetics, National Research Council (IGM-CNR), Via Abbiategrasso 207, 27100 Pavia, Italy

Correspondence should be addressed to Rotem Karni; [email protected]

Received 20 June 2013; Accepted 26 July 2013

Academic Editor: Claudio Sette

Copyright © 2013 Zahava Siegfried et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Alternativesplicingisafundamentalstepinregulationofgeneexpressionofmanytumorsuppressorsandoncogenesincancer. Signalling through the Ras-MAPK and PI3K-mTOR pathways is misregulated and hyperactivated in most types of cancer. However, the regulation of the Ras-MAPK and PI3K-mTOR signalling pathways by alternative splicing is less well established. Recent studies have shown the contribution of alternative splicing regulation of these signalling pathways which can lead to cellular transformation, cancer development, and tumor maintenance. This review will discuss findings in the literature which describe new modes of regulation of components of the Ras-MAPK and PI3K-mTOR signalling pathways by alternative splicing. We will also describe the mechanisms by which signals from extracellular stimuli can be communicated to the splicing machinery and to specific RNA- binding proteins that ultimately control exon definition events.

1. Introduction The Ras-MAPK and PI3K-mTOR signalling pathways are deregulated in many cancers by genetic and epigenetic In the past several decades cancer research has focused on aberrations [16–18]. Several key components in these path- genetic alterations such as mutations, copy number varia- ways, such as Ras, B-RAF, C-RAF, MEK1, PI3K, and Akt, are tions, and translocations that are detected in malignant tis- activated by mutations or gene amplifications, while other sues and contribute to the initiation and progression of can- components that inhibit these pathways, such as PTEN, LKB1, cer. In recent years it is becoming clear that epigenetic chan- and TSC1/2, are inactivated by genomic deletions and muta- ges, including transcriptional and posttranscriptional alter- tions [16–20]. Pharmacological inhibitors of enzymes in these ations, also play a major role in cancer development and thus pathways, such as BRAF inhibitors and mTOR inhibitors, should be the direction of future research [1–4]. Mutations arealreadybeingusedinclinicalsettingstotreatcancer, and copy number variations in splicing regulators have been whileothers(PI3KandMEK1inhibitors)areinadvanced identified in several types of cancer, supporting the notion stages of clinical trials [21–26]. Although the Ras-MAPK and that changes in splicing fidelity contribute to cancer develop- PI3K-mTOR pathways are at the center of intensive research, ment [5–9]. and many genetic alterations that activate or inactivate these Alternative splicing plays a major role in cancer devel- pathways have been discovered, much less is known about opment and progression as many tumor suppressors and the epigenetic and posttranscriptional regulation of these sig- oncogenes are modulated by alternative splicing [10, 11]. nalling pathways. Recent studies have revealed how these However, the role of alternative splicing regulators in cancer pathwayscanberegulatedbyalternativesplicingandbysplic- development is mostly unknown, and only recently the first ingregulatorsandarethefocusofthisreview. direct evidence for an oncogenic role of a splicing factor has Here, we discuss the intricate relationship between alter- been shown [9, 12–15]. native splicing and signalling at different levels: (i) how the 2 International Journal of Cell Biology

Table 1: Alternative splicing of Ras-MAPK and PI3K-mTOR signaling components.

Signalling component Gene name Splicing isoform activity Type of cancer Reference number Constitutively active receptor/soluble decoy Glioblastoma, RTK EGFR isoform, enhanced signalling, survival, and [28, 47–49, 51, 56–58] lung tumorigenicity. Glioblastoma, Constitutively active receptor, enhanced RTK RON colon, breast, [34, 35] signalling, invasion, and motility. and gastric Constitutively active receptor/soluble decoy Ovarian, lung, RTK MET isoform, enhanced/reduced signalling, [38, 39] and HCC invasion, and motility. Prostate, RTK FGFR Induction of EMT, invasion, and motility. pancreatic, and [28, 29, 36, 61] breast RTK INSR Differential ligand binding (IGF-II) and HCC, thyroid, [31, 37] oncogenic activity. and ovarian Soluble decoy isoform, enhanced/reduced RTK VEGFR Lung, breast [28, 32] angiogenesis, and survival. Enhanced/reduced kinase activity, survival Cytosolic kinase Fyn Unknown [43] of epithelial cells. Constitutively active kinase, oncogenic Cytosolic kinase mTOR HCC [75] activity. Tumor suppressor/oncogenic isoforms, Cytosolic kinase S6K1 Breast, lung [14, 78] activates/inhibits mTORC1. Cytosolic kinase A-Raf Enhanced/reduced binding to Ras and HCC, head, and [45, 46] activation of the MAPK pathway. neck Enhanced/reduced kinase activity, activation Colon, Cytosolic kinase B-Raf of the MAPK pathway, and resistance to [20, 21, 40] melanoma BRAF kinase inhibitors. Alternative pathway with a different Cytosolic kinase MEK1 Unknown [66, 67] substrate. Alternative pathway with different Cytosolic kinase ERK1 Unknown [66, 67] substrates. Oncogenic isoform that enhances eIF4E Lung, breast, Cytosolic kinase MKNK2 phosphorylation and a tumor-suppressive colon, and [14, 68] isoform. pancreas Phospholipid phosphatase PTEN Active/inactive tumor suppressor. Unknown [71] Constitutively active kinase, enhanced Phospholipid kinase PI3K Unknown [69, 70] downstream signalling. Enhanced/reduced binding to Raf and Rin Small GTPase Ras Unknown [32, 44] and activation of the MAPK pathway. Tuberous GTPase activator (GAP) TSC2 Inactivation of a tumor suppressor. [72–74] sclerosis activity of components in the Ras-MAPK signalling pathway members of non-receptor cytosolic protein kinases, such as is regulated by alternative splicing in cancer cells; (ii) how the Src, Ras, and Raf and on non-kinase cytosolic receptors, activity of components in the PI3K-mTOR pathway is regu- including androgen and estrogen receptors [20, 40–43] lated by alternative splicing in cancer cells; (iii) mechanisms (Table 1). by which extracellular stimuli can be communicated to the splicing machinery and to specific RNA-binding proteins that ultimately control exon definition events. 2. Regulation of the Ras-MAPK Pathway by Alternative splicing can affect the activity of signalling Alternative Splicing effectors contributing to their constitutive (or improper) function. The most well-characterized examples are repre- Downstream to RTK activation, the small GTPase Ras is sented by members of the receptor tyrosine kinase (RTK) loaded with GTP and activated. Of the three genes encoding family; EGFR, FGFR, INSR, VEGFR, MET, and Ron [2, for Ras proteins (K-Ras, H-Ras, and N-Ras), K-Ras and H- 19, 27–39]. In addition, recent studies have also focused on Ras can include or exclude an exon termed IDX and generate International Journal of Cell Biology 3

Growth factors, cytokines GTPase activator [45, 46](Figure 1). A-Rafshort which arises from retention of introns 2 and 4 is a 171-amino acid protein Dominant negative ⇌ RTKs ⇌ Constitutively active (soluble/kinase dead) lacking two-thirds of the C-terminal region including the Ras ⇌ p19-Ras kinase domain [46]. In human head and neck carcinomas c- myc can induce high levels of the splicing regulator hnRNP ⇌⇌ Short Raf (RBD) Raf Active Raf (kinase only) H, which in turn increases expression of the full-length A-Raf P protein, while decreasing the expression of A-Rafshort [46]. hnRNP H MAPKK ⇌ MAPKK1b Another truncated isoform is DA-Raf1, which contains a P premature termination codon, due to intron 6 retention. ERK ⇌ MAPKERK1c MAPK This isoform comprises of only the N-terminal Ras-binding P domain of A-Raf [45]. Its overexpression impairs the tumori- MNK2a ⇌⇌MKNK2 MNK2b genic potential of mutant K-Ras transformed mouse NIH3T3 P SRSF1 eIF4E SRSF1 fibroblasts in nude mice [45].Althoughidentifiedinmurine cells, it can also affect the activity of human A-Raf. In a similar ⇌ Alternative splicing Parallel signalling pathway fashion, B-Raf truncated splicing variants devoid of kinase Tumor suppressor activity Oncogenic activity activity have been recently identified in colorectal cancer cells Figure 1: Alternative splicing regulation of the Ras-MAPK pathway. [40]. These truncated B-Raf splicing variants are generated Growth factors and cytokines activate receptor tyrosine kinases through several alternative splicing events, such as skipping (RTKs)whichinturnleadtoactivationofRas.RTKscanbealterna- of exons 14-15 or inclusion of one of the additional exons 15b, tively spliced to generate soluble truncated isoforms which act in a 16b, and 16c, resulting in premature stop codons [40]. dominant-negative manner or constitutively active isoforms which The Ras/Raf/MAPK cascade can be activated by the epi- are active regardless of ligand binding. Ras can be alternatively dermal growth factor receptor (EGFR/ErbB1), a member of spliced to generate p19-Ras which cannot activate its downstream the ErbB receptor tyrosine kinase family, which is frequently effector Raf. GTP-bound p21-Ras activates A-, B-, and C-Raf which can be alternatively spliced to generate inactive dominant- mutated and overexpressed in different human cancers, negative isoforms containing only the Ras binding domain (RBD) including glioma, non-small-cell lung carcinoma, ovarian or constitutively active isoforms containing only the kinase domain. carcinoma, and prostate carcinoma [47]. The most studied Oncogenic splicing factor hnRNP H inhibits the production of EGFR variant is the type III epidermal growth factor recep- dominant-negative A-Raf isoforms. Raf phosphorylates MAPKK tor mutant EGFRvIII (also referred to as ΔEGFR or de2- (MEK) which in turn phosphorylates MAPK-ERK. MAPKK1 and 7 EGFR), containing an inframe deletion of exons 2–7 ERK1 can generate MAPKK1b and ERK1c, respectively, by alter- that can be generated either by gene rearrangement or altered native splicing to generate a parallel signalling pathway. MAPK- pre-mRNA processing [48, 49]. EGFRvIII lacks a portion ERK can phosphorylate MNK2, which is alternatively spliced and of the extracellular ligand-binding domain, is constitutively regulated by the oncogenic splicing factor SRSF1. SRSF1 upregulates active in a ligand-independent manner, and confers growth a prooncogenic Mnk2b isoform and reduces Mnk2a isoform of this advantage to cancer cells [49, 50](Figure 1). The selective kinase. Blue: tumor suppressors, red: oncogenes, and green: parallel expression of EGFRvIII in several tumors, but not in normal pathway. tissues, makes it an extremely attractive target for anticancer therapy [50]. Another splicing isoform of EGFR, called de4EGFR,isproducedbyskippingofexon4.Similarto p19- or p21-Ras, respectively [44]. p19-Ras cannot interact EGFRvIII, de4 EGFR undergoes ligand-independent activa- with A-Raf or Rin1, but binds RACK1 and may act in an tion and self-dimerization and displays transformation capa- antagonistic manner to p21-Ras [42]. Alternative splicing bilities as well as metastasis-promoting potential [51]. Not of the cytosolic kinase B-Raf, a downstream effector of the only is de4 EGFR detectable in several human tumors, includ- small GTPase Ras in the mitogen-activated protein (MAP) ing glioma, prostate, and ovarian, but also its expression kinase pathway, gives rise to several isoforms devoid of the correlates with the malignant degree of glioblastomas [51]. N-terminal autoinhibitory domain [20]. These constitutively A fascinating example of the link between alternative active protein isoforms are able to activate the MAP kinase splicing and signalling cascades has been recently provided by signalling pathway and to induce tumor formation in nude studying the metabolic effects, increased glucose uptake and mice [20](Figure 1, Table 1). Furthermore, alterations in the lactate production, of EGFR activation in human cancer cells. splicing profile of signalling components can also play a role The ability of EGFR activation to modulate the metabolism in the acquisition of drug resistance. As an example, the of cancer cells requires the expression of the PKM2 splicing mutant B-Raf(V600E) allele generates splicing isoforms that isoform of pyruvate kinase M (PKM) [52], the enzyme that lack the Ras-binding domain and are able to dimerize in a catalyzes the final step of glycolysis. There are two variants Ras-independent manner [21]. Importantly, B-Raf(V600E) of PKM which are generated through alternative splicing of splicetranscriptshavebeendetectedinseveralmelanoma two mutually exclusive exons: exon 9 included in PKM1 tran- patients with acquired resistance to vemurafenib, an inhibitor scripts and exon 10 in PKM2. The choice between exon 9 and of Raf activity used in clinical practice [21]. In addition, two 10 is controlled by polypyrimidine tract binding protein truncated splicing isoforms of the cytosolic kinase A-Raf, (PTB, also known as PTBP1 or hnRNP I) and hnRNP A1/A2 another member of the Raf kinase family, negatively regulate that bind to sequences flanking exon 9, thus inhibiting the A-Raf (Raf1) kinase activity by sequestering the upstream Ras selection of exon 9 and promoting exon 10 inclusion [53, 54]. 4 International Journal of Cell Biology

EGFR activation acts at different levels in the expression Growth factors, cytokines of PKM2 by increasing transcription of both PTB and RTKs PKM genes. Yang and collaborators have recently reported Inactive PTEN ⇌ PTEN that EGFR upregulation of PKM2, but not PKM1, requires PI3K ⇌ Active PI3K NF-𝜅B activation, which is mediated by PLC𝛾1andPKC𝜀 P monoubiquitylation-dependent IKKbeta activation [52, 55]. LKB1 Akt Moreover, RNAi-mediated knock-down experiments indi- AMPK P cate that PTB mediates the effect of EGFR on splicing of TSC2 the PKM gene but not on transcription. Thus, a coordinated Rheb transcription-splicing program controlled by EGFR activa- SRSF1 tion is responsible for the expression of the PKM2 isoform mTOR ⇌ mTOR𝛽 and for the distinctive metabolic features of cancer cells. P P ⇌ Another interesting example of regulation of the Ras- S6K1 (h6A, h6C) S6K1 (Iso-1) 4E-BP1 MAPK pathway by alternative splicing is the observation eIF4E made by Cartegni’s group that intronic polyadenylation, con- ⇌ comitantly (and in competition) with pre-mRNA splicing, Alternative splicing Tumor suppressor activity can generate truncated soluble receptor tyrosine kinases Oncogenic activity (RTKs). These isoforms lack the anchoring transmembrane domain and the intracellular kinase domain and can act as Figure 2: Alternative splicing regulation of the PI3K-mTOR dominant-negative regulators [28](Figure 1). These secreted pathway. Growth factors and cytokines activate receptor tyrosine decoy receptors can shut down the relevant tumorigenic kinases (RTKs) which in turn lead to activation of PI3K. PI3K signalling pathways by titrating out the ligand or by trapping phosphorylates phospholipids inducing the recruitment of Akt to the wild-type receptors in nonfunctional heterodimers [28]. the plasma membrane and its activation by PDK1. A splicing variant of the catalytic subunit of PI3K (p37 delta) is an active form that Inparticular,forthevascularendothelialgrowthfactorrecep- enhances PI3K activity. Akt phosphorylates and inactivates the tor 2 (VEGFR2/KDR), the pivotal molecule in controlling tumor suppressor TSC2, which inhibits the small GTPase Rheb. VEGF-dependent functions, the expression of the dominant- GTP-bound Rheb can activate mTOR. mTOR𝛽 is an active splicing negative sKDR strongly inhibited the angiogenesis process in isoform of mTOR. mTOR phosphorylates S6K1 and 4E-PB1. 4E- both primary HUVEC endothelial cells and in the same cells BP1 phosphorylation induces its release from eIF4E, enhancing cap- exposed to conditioned media and simultaneously treated dependent translation and malignant transformation. Oncogenic with VEGF-A. Thus, artificially increasing production of splicing factor SRSF1 can affect the alternative splicing of S6K1 thesetruncatedsolublereceptorscouldbeavalidapproach inducing oncogenic short isoforms of this kinase (h6A, h6C) to interfere with angiogenic paracrine and autocrine loops. which bind mTOR and enhance 4E-BP1 phosphorylation and cap- Soluble isoforms produced by alternative splicing and dependent translation. Blue: tumor suppressors, red: oncogenes. containing only the extracellular domain of the protein have also been detected for the EGFR/ErbB1 tyrosine kinase receptor [47, 56]. These truncated soluble EGFR variants prostate cancer cells to chemotherapy [68]. Altogether, these have been detected in several cancer types, and their levels, examplesofcomponentsoftheRas-MAPKpathwaywhich circulatingaswellasintumortissues,havebeenusedas are regulated by alternative splicing open up the exciting prognostic and predictive markers for ovarian, cervical, lung, possibility of exploiting dominant-negative splicing isoforms and breast cancers [57–59]. Alternative splicing of other for therapeutic purposes, as antioncogenic agents. tyrosine kinase receptors such as the insulin, IGF-1, FGF, ERBB2, and ERBB4 receptors can also modulate the activation of their downstream signalling pathways [2, 19, 29, 36, 3. Regulation of the PI3K-mTOR Pathway by 60–65]. An additional example of alternative splicing in the Alternative Splicing Ras-MAPK pathway is the splicing of MEK1b and ERK1c [66, 67]. MEK1 (MAPKK1) phosphorylates and activates Upon RTK activation, both the Ras-MAPK and the PI3K- ERK1/ERK2 [66, 67]. MEK1b is a unique splicing isoform of mTOR pathways are activated [16, 17](Figure 2). Kinase MEK1 which specifically phosphorylates ERK1c, an isoform PI3K, which phosphorylates 3,5, phosphoinositides, consists of ERK1 [66, 67]. Thus, by alternative splicing a parallel of a regulatory subunit, p85, and a catalytic subunit, p110. pathway is generated, and this confers higher substrate Both p85 and p110 undergo alternative splicing which can specificity to this branch of the pathway66 [ , 67](Figure 1). modulate the activity of PI3K [69, 70](Figure 2). The tumor Another component in the MAPK pathway which is regu- suppressor phosphatase PTEN, which counteracts PI3K lated by alternative splicing is the kinase MKNK2,whichis activity, is also regulated by alternative splicing, and splicing regulated by the oncogenic splicing factor SRSF1. Upregu- isoforms of PTEN that are inactive and act in a dominant- lation of SRSF1 downregulates the expression of the Mnk2a negative fashion have been detected in Cowden syndrome isoform and induces the expression of the Mnk2b isoform andinbreastcancer[71]. Upon production of phospho- [14](Figure 1). Mnk2b is a prooncogenic isoform that acti- inositide by PI3K, the kinase Akt is recruited to the plasma vates eIF4E independently of MAPK activation [14]. Recently, membrane and is activated by PDK1 [17]. Active Akt phos- this splicing switch was shown to regulate the resistance of phorylates and inactivates the tumor suppressor complex of International Journal of Cell Biology 5

TSC1/TSC2. Both TSC1 and TSC2 tumor suppressors are Pathways for transformation by SRSF1 alternatively spliced although inactivation by alternative Akt P splicing has not been demonstrated [72–74]. TSC1/2 inacti- CLK vates the GTPase Rheb, a small GTPase from the Ras protein c-Myc SRPK family which binds and activates mTOR [18]. Alternative P splicing of Rheb is not known. mTOR can undergo alternative SRSF1 mTOR splicing to generate an activated form called mTORbeta which is oncogenic, although its regulation and role in human cancer have yet to be demonstrated [75]. Recently it was BIM 𝛾1 BIN1 12A S6K16a/6c Mnk2a MAPK ⇌ ⇌ shown that the splicing factor Sam68 controls alternative ⇌ ⇌ BIM BIN1 S6K1 Iso-1 Mnk2b splicing of mTOR, reducing the retention of intron 5. Cells P from Sam68 knockout mice display reduced levels of mTOR 4EBP1 mRNA due to nonsense-mediated decay degradation of the Apoptosis P intron-retained transcript and reduced activity of the mTOR Transformation eIF4E pathway [76]. mTOR phosphorylates and activates several substrates. Among the best-characterized mTOR substrates ⇌ Alternative splicing are S6K1, which phosphorylates the ribosomal protein S6 and Alternative protein domain regulates the translation process [18], and eIF4E-BP1, which Figure 3: Pathways for transformation by SRSF1. SRSF1 is a is involved in the formation of an active translation initiation transcriptional target of the c-myc protooncogene and can be complex [14, 18, 77]. Alternative splicing of S6K1 transcripts is phosphorylated by SRPK or CLK downstream to Akt. SRSF1 alters controlled by the oncogenic splicing factor SRSF1 [14]. Specif- the splicing of BIM, BIN1, S6K1, and Mnk2 regulating the mTOR and ically, SRSF1 promotes the production of short S6K1 isoforms, MAPK pathways, increasing translation, and inhibiting apoptosis. frequently upregulated in breast cancer cell lines and tumors [14, 78]. These variants display oncogenic properties as they are able to enhance cell transformation, motility, and anchorage-independent growth of breast epithelial cells [78]. Signals from the EGF receptor, which activate the PI3K- The short splicing variants of S6K1 are not substrates of the Aktpathway,inducephosphorylationoftheSRproteinkinase signalling cascade (Akt/mTOR) but generate a signal loop SRPK1 which in turn phosphorylates SR proteins, such as to activate the mTOR pathway in the absence of external SRSF1. Phosphorylation of SRSF1 leads to its activation and stimuli. Indeed, S6K1 short isoforms are able to bind and modulation of the cell’s alternative splicing program [86]. The increase mTORC1 activity, leading to 4E-BP1 inactivation and complete landscape of alternative splicing changes by such enhancing translation of several oncogenes and antiapoptotic signalshasyettobeelucidated. genes [78](Figure 2). Stress signals emanating from osmotic shock and other modes of stress which activate the MEK4-/MEK7-p38- 4. Extracellular Stimuli Targeting Components MAPK pathway control cell localization of the splicing factor of the Splicing Machinery hnRNP A1 [87–89]. Activation of the p38-MAPK pathway induces hnRNP A1 phosphorylation and export from the Although for a long time transcription factors have been con- nucleus into the cytoplasm [87, 89]. Reducing nuclear hnRNP sidered the ultimate effectors of the signalling cascade path- A1 levels in response to stress is expected to affect many ways, in the last few years splicing machinery components alternative splicing events which have yet to be determined. and alternative splicing regulatory proteins have also been A very complex and intriguing story has emerged from recognized as important targets [79]. The activity of splicing the study of the multiple functions of SRSF1. SRSF1 is a factors is normally regulated through posttranscriptional splicing factor with an oncogenic activity that derives from modifications, mainly phosphorylation80 [ ], and therefore it its involvement in the Akt-mTOR pathway [14, 90](Figure 3). is not surprising that signalling cascades are involved in mis- The Akt-dependent phosphorylation of SRSF1 affects its func- regulation of splicing factors. tion in splicing regulation and translation. Thus, for instance, The PI3K/Akt/mTOR signalling cascade, frequently upon phosphorylation by Akt, SRSF1 enhances the produc- alteredinmalignancies[81], has been found to affect the tion of EDA-FN, a splicing isoform of fibronectin 1 (FN1) alternative splicing profile of cancer-relevant genes via expressed in various malignancies but undetectable in nor- Akt-dependent phosphorylation of several SR proteins, such mal tissues [82, 91]. SRSF1 phosphorylation by Clk (CLK1-4) as SRSF1 and SRSF5 [82–84]. Alternative splicing factor or SRPK kinases (SRPK1-2) has the opposite effect on EDA- 45 (SPF45) has also been identified as a target of multiple FN splicing, pointing to SRSF1 as a sort of integrator compo- MAP kinases, such as ERKs (extracellular signal-regulated nent that can switch between splicing decisions in response kinases), JNKs (Jun N-terminal kinases), and p38 MAPK to various external cues [91].Thisviewisfurthersupportedby [85]. Interestingly, phosphorylation of SPF45 impacts cell the observation that the effect of SRSF1 on EDA-FN splicing proliferation and cell adhesion programs, through down- is abrogated by rapamycin, an inhibitor of mTORC1 (mTOR regulation of the human epidermal growth factor recep- complex-1) [82]. SRSF1 is not only a target but also an tor (ErbB2) and regulation of the alternative splicing of the effectorofsignallingcascades.Indeed,itisabletoactivate fibronectin 1 gene (FN1), respectively [85]. mTORC1 leading to downstream phosphorylation of two 6 International Journal of Cell Biology of its substrates; S6K1 kinase and 4E-BP1 [77, 78, 90]. The [8]V.Quesada,L.Conde,N.Villamoretal.,“Exomesequencing mechanism by which SRSF1 activates mTORC1 still needs to identifies recurrent mutations of the splicing factor SF3B1 gene be clarified. However, mTORC1 activation by SRSF1 bypasses in chronic lymphocytic leukemia,” Nature Genetics,vol.44,no. upstream PI3K/Akt signalling and is essential for SRSF1- 1, pp. 47–52, 2012. mediated transformation in nude mice, suggesting that [9] M. Cohen-Eliav, R. Golan-Gerstl, Z. Siegfried et al., “The mTOR inhibitors may be useful for treatment of tumors with splicing factor SRSF6 is amplified and is an oncoprotein in lung SRSF1 upregulation [78, 90]. and colon cancers,” The Journal of Pathology,vol.229,no.4,pp. 630–639, 2013. 5. Conclusions [10] J. P. Venables, “Unbalanced alternative splicing and its significance in cancer,” BioEssays, vol. 28, no. 4, pp. 378–386, 2006. Our current knowledge is only the tip of the iceberg in term [11] J. P. Venables, R. Klinck, C. Koh et al., “Cancer-associated reg- s of the many mechanisms by which alternative splicing ulation of alternative splicing,” Nature Structural & Molecular regulates the Ras-MAPK and the PI3K-mTOR signalling Biology, vol. 16, no. 6, pp. 670–676, 2009. pathways. This is also true for the mechanisms by which [12]R.Jia,C.Li,J.P.McCoy,C.-X.Deng,andZ.-M.Zheng,“SRp20 these signalling cascades regulate the activity of splicing is a proto-oncogene critical for cell proliferation and tumor regulators such as SR and hnRNP proteins leading to mod- induction and maintenance,” International Journal of Biological ulation of the cell’s splicing landscape. The diagnostic and Sciences,vol.6,no.7,pp.806–826,2010. therapeutic potential of manipulating the Ras-MAPK and [13] C. V. Lefave, M. Squatrito, S. Vorlova et al., “Splicing factor the PI3K-mTOR pathways, by modulating alternative splicing hnRNPH drives an oncogenic splicing switch in gliomas,” of components of these pathways, is immense. Detection of EMBO Journal,vol.30,no.19,pp.4084–4097,2011. alternatively spliced isoforms of key substrates, such as S6K1, [14]R.Karni,E.DeStanchina,S.W.Lowe,R.Sinha,D.Mu,andA. mightbeusedtopredictthesensitivityoftumorstomTOR R. Krainer, “The gene encoding the splicing factor SF2/ASF isa inhibitors, enabling better choices of drug treatment in the proto-oncogene,” Nature Structural and Molecular Biology,vol. clinic [78]. Another clinical application concerns resistance 14,no.3,pp.185–193,2007. of tumors to drugs which, as in the case of B-Raf splicing, can [15] R. Golan-Gerstl, M. Cohen, A. Shilo et al., “Splicing factor arise by alterations in alternative splicing. Elimination of a hnRNP A2/B1 regulates tumor suppressor gene splicing and is drug-binding domain, for example, does not require a an oncogenic driver in glioblastoma,” Cancer Research,vol.71, genomicchangebutrathercanoccurbyskippingoftherele- no.13,pp.4464–4472,2011. vant exon, resulting in a new splicing isoform which is drug [16] T. L. Yuan and L. C. Cantley, “PI3K pathway alterations in resistant. Even with all these new findings there still remains cancer: variations on a theme,” Oncogene,vol.27,no.41,pp. a lot to be learned about how alternative splicing regulates 5497–5510, 2008. signalling and how signalling regulates splicing. This field [17] R. J. Shaw and L. C. Cantley, “Ras, PI(3)K and mTOR signalling holds great promise for advancing our understanding of the controls tumour cell growth,” Nature,vol.441,no.7092,pp. cancer process and for use in the clinic. 424–430, 2006. [18] M. Laplante and D. M. Sabatini, “MTOR signaling in growth References control and disease,” Cell,vol.149,no.2,pp.274–293,2012. [19] S. Druillennec, C. Dorard, and A. Eychene, “Alternative splicing [1]S.Bonnal,L.Vigevani,andJ.Valcarcel,“Thespliceosomeasa in oncogenic kinases: from physiological functions to cancer,” target of novel antitumour drugs,” Nature Reviews Drug Discov- Journal of Nucleic Acids, vol. 2012, Article ID 639062, 14 pages, ery, vol. 11, no. 11, pp. 847–859, 2012. 2012. [2]C.C.WarzechaandR.P.Carstens,“Complexchangesinalter- [20]E.Y.Baitei,M.Zou,F.Al-Mohannaetal.,“AberrantBRAF native pre-mRNA splicing play a central role in the epithelial- splicing as an alternative mechanism for oncogenic B-Raf acti- to-mesenchymal transition (EMT),” Seminars in Cancer Biol- vation in thyroid carcinoma,” Journal of Pathology,vol.217,no. ogy,vol.22,no.5-6,pp.417–427,2012. 5,pp.707–715,2009. [3] E. Kim, A. Goren, and G. Ast, “Insights into the connection [21] P. I. Poulikakos, Y. Persaud, M. Janakiraman et al., “RAF between cancer and alternative splicing,” Trends in Genetics,vol. inhibitor resistance is mediated by dimerization of aberrantly 24,no.1,pp.7–10,2008. spliced BRAF(V600E),” Nature,vol.480,no.7377,pp.387–390, [4] R. K. Singh and T. A. Cooper, “Pre-mRNA splicing in disease 2011. and therapeutics,” Trends in Molecular Medicine,vol.18,no.8, [22] Y. Heakal, M. Kester, and S. Savage, “Vemurafenib (PLX4032): pp.472–482,2012. an orally available inhibitor of mutated BRAF for the treatment [5] E. Papaemmanuil, M. Cazzola, J. Boultwood et al., “Somatic of metastatic melanoma,” Annals of Pharmacotherapy,vol.45, SF3B1 mutation in myelodysplasia with ring sideroblasts,” New no. 11, pp. 1399–1405, 2011. England Journal of Medicine,vol.365,no.15,pp.1384–1395,2011. [23] C. M. Barnett, “Everolimus: targeted therapy on the horizon for [6] J.W.Harbour,E.D.O.Roberson,H.Anbunathan,M.D.Onken, the treatment of breast cancer,” Pharmacotherapy,vol.32,no.4, L. A. Worley, and A. M. Bowcock, “Recurrent mutations at pp. 383–396, 2012. codon 625 of the splicing factor SF3B1 in uveal melanoma,” [24] K. B. Kim, R. Kefford, A. C. Pavlick et al., “Phase II study of the Nature Genetics, vol. 45, no. 2, pp. 133–135, 2013. MEK1/MEK2 inhibitor Trametinib in patients with metastatic [7] V. Quesada, A. J. Ramsay, and C. Lopez-Otin, “Chronic lym- BRAF-mutant cutaneous melanoma previously treated with or phocytic leukemia with SF3B1 mutation,” New England Journal without a BRAF inhibitor,” Journal of Clinical Oncology,vol.31, of Medicine,vol.366,no.26,p.2530,2012. no. 4, pp. 482–489, 2013. International Journal of Cell Biology 7

[25]K.T.Flaherty,J.R.Infante,A.Daudetal.,“CombinedBRAF [40] B. Hirschi and F. T. Kolligs, “Alternative splicing of BRAF tran- and MEK inhibition in melanoma with BRAF V600 mutations,” scripts and characterization of C-terminally truncated B-Raf New England Journal of Medicine,vol.367,no.18,pp.1694–1703, isoforms in colorectal cancer,” International Journal of Cancer, 2012. vol. 133, no. 3, pp. 590–596, 2013. [26] S. P. Patel, A. J. Lazar, N. E. Papadopoulos et al., “Clini- [41] X. Yang, Z. Guo, F. Sun et al., “Novel membrane-associated cal responses to selumetinib (AZD6244; ARRY-142886)-based androgen receptor splice variant potentiates proliferative and combination therapy stratified by gene mutations in patients survival responses in prostate cancer cells,” Journal of Biological with metastatic melanoma,” Cancer,vol.119,no.4,pp.799–805, Chemistry,vol.286,no.41,pp.36152–36160,2011. 2013. [42] M. Camats, M. Kokolo, K. J. Heesom, M. Ladomery, and M. [27] M. W. Pedersen, V. Tkach, N. Pedersen, V. Berezin, and H. Bach-Elias, “P19 H-ras induces G1/S phase delay maintaining S. Poulsen, “Expression of a naturally occurring constitutively cells in a reversible quiescence state,” PLoS ONE,vol.4,no.12, active variant of the epidermal growth factor receptor in mouse Article ID e8513, 2009. fibroblasts increases motility,” International Journal of Cancer, [43] C. Brignatz, M. P. Paronetto, S. Opi et al., “Alternative splicing vol. 108, no. 5, pp. 643–653, 2004. modulates autoinhibition and SH3 accessibility in the Src kinase [28] S. Vorlova,´ G. Rocco, C. V. LeFave et al., “Induction of antago- Fyn,” Molecular and Cellular Biology,vol.29,no.24,pp.6438– nistic soluble decoy receptor tyrosine kinases by intronic polyA 6448, 2009. activation,” Molecular Cell,vol.43,no.6,pp.927–939,2011. [44] S. Guil, N. De La Iglesia, J. Fernandez-Larrea´ et al., “Alternative [29] C. C. Warzecha, T. K. Sato, B. Nabet, J. B. Hogenesch, and R. splicing of the human proto-oncogene c-H-ras renders a new P. Carstens, “ESRP1 and ESRP2 are epithelial cell-type-specific Ras family protein that trafficks to cytoplasm and nucleus,” regulators of FGFR2 splicing,” Molecular Cell,vol.33,no.5,pp. Cancer Research,vol.63,no.17,pp.5178–5187,2003. 591–601, 2009. [45] T. Yokoyama, K. Takano, A. Yoshida et al., “DA-Raf1, a com- [30]T.H.Ho,R.S.Savkur,M.G.Poulos,M.A.Mancini,M.S. petent intrinsic dominant-negative antagonist of the Ras-ERK Swanson, and T.A. Cooper, “Colocalization of muscleblind with pathway, is required for myogenic differentiation,” Journal of RNA foci is separable from mis-regulation of alternative splic- Cell Biology,vol.177,no.5,pp.781–793,2007. ing in myotonic dystrophy,” JournalofCellScience,vol.118,no. [46] J. Rauch, K. Moran-Jones, V. Albrecht et al., “c-Myc regulates 13, pp. 2923–2933, 2005. RNA splicing of the A-Raf kinase and its activation of the ERK [31]R.S.Savkur,A.V.Philips,andT.A.Cooper,“Aberrantregu- pathway,” Cancer Research, vol. 71, no. 13, pp. 4664–4674, 2011. lation of insulin receptor alternative splicing is associated with [47] E. K. Rowinsky, “The erbB family: targets for therapeutic devel- insulin resistance in myotonic dystrophy,” Nature Genetics,vol. opment against cancer and therapeutic strategies using mono- 29, no. 1, pp. 40–47, 2001. clonal antibodies and tyrosine kinase inhibitors,” Annual [32] B. Mezquita, J. Mezquita, M. Pau, and C. Mezquita, “A novel Review of Medicine,vol.55,pp.433–457,2004. intracellular isoform of VEGFR-1 activates Src and promotes [48]N.Sugawa,A.J.Ekstrand,C.D.James,andV.P.Collins, cell invasion in MDA-MB-231 breast cancer cells,” Journal of “Identical splicing of aberrant epidermal growth factor receptor Cellular Biochemistry,vol.110,no.3,pp.732–742,2010. transcripts from amplified rearranged genes in human glioblas- [33] C. P. Thomas, N. S. Raikwar, E. A. Kelley, and K. Z. Liu, “Alter- tomas,” Proceedings of the National Academy of Sciences of the nate processing of Flt1 transcripts is directed by conserved cis- United States of America,vol.87,no.21,pp.8602–8606,1990. elements within an intronic region of FLT1 that reciprocally [49] R. Nishikawa, X.-D. Ji, R. C. Harmon et al., “Amutant epidermal regulates splicing and polyadenylation,” Nucleic Acids Research, growth factor receptor common in human glioma confers vol. 38, no. 15, Article ID gkq198, pp. 5130–5140, 2010. enhanced tumorigenicity,” Proceedings of the National Academy [34] C. Ghigna, M. De Toledo, S. Bonomi et al., “Pro-metastatic of Sciences of the United States of America,vol.91,no.16,pp. splicing of Ron proto-oncogene mRNA can be reversed: ther- 7727–7731, 1994. apeutic potential of bifunctional oligonucleotides and indole [50] U. H. Weidle, D. Maisel, S. Klostermann, E. H. Weiss, and M. derivatives,” RNA Biology,vol.7,no.4,pp.495–503,2010. Schmitt, “Differential splicing generates new transmembrane [35] C. Ghigna, S. Giordano, H. Shen et al., “Cell motility is con- receptor and extracellular matrix-related targets for antibody- trolled by SF2/ASF through alternative splicing of the Ron based therapy of cancer,” Cancer Genomics and Proteomics,vol. protooncogene,” Molecular Cell,vol.20,no.6,pp.881–890,2005. 8, no. 5, pp. 211–226, 2011. [36] P.Zhang,J.S.Greendorfer,J.Jiao,S.C.Kelpke,andJ.A.Thomp- [51] H. Wang, M. Zhou, B. Shi et al., “Identification of an exon son, “Alternatively spliced FGFR-1 isoforms differentially mod- 4-deletion variant of epidermal growth factor receptor with ulate endothelial cell activation of c-YES,” Archives of Bioc- increased metastasis-promoting capacity,” Neoplasia,vol.13,no. hemistry and Biophysics,vol.450,no.1,pp.50–62,2006. 5, pp. 461–471, 2011. [37]H.Chettouh,L.Fartoux,L.Aoudjehaneetal.,“Mitogenic [52]W.Yang,Y.Xia,Y.Caoetal.,“EGFR-inducedandPKCepsilon insulin receptor-A is overexpressed in human hepatocellular monoubiquitylation-dependent NF-kappaB activation upregu- carcinoma due to EGFR-mediated dysregulation of RNA splic- lates PKM2 expression and promotes tumorigenesis,” Molecular ing factors,” Cancer Research,vol.73,no.13,pp.3974–3986,2013. Cell,vol.48,no.5,pp.771–784,2012. [38]J.-H.Lee,F.G.Chong,C.L.Chong,D.K.Myung,andG.F. [53]C.V.Clower,D.Chatterjee,Z.Wang,L.C.Cantley,M.G.V. V. Woude, “An alternatively spliced form of Met receptor is Heidena, and A. R. Krainer, “The alternative splicing repressors tumorigenic,” Experimental and Molecular Medicine,vol.38,no. hnRNP A1/A2 and PTB influence pyruvate kinase isoform 5, pp. 565–573, 2006. expression and cell metabolism,” Proceedings of the National [39] Z. Tiran, A. Oren, C. Hermesh et al., “A novel recombinant Academy of Sciences of the United States of America,vol.107,no. soluble splice variant of met is a potent antagonist of the hepato- 5,pp.1894–1899,2010. cyte growth factor/scatter factor-met pathway,” Clinical Cancer [54]C.J.David,M.Chen,M.Assanah,P.Canoll,andJ.L.Manley, Research, vol. 14, no. 14, pp. 4612–4621, 2008. “HnRNP proteins controlled by c-Myc deregulate pyruvate 8 International Journal of Cell Biology

kinase mRNA splicing in cancer,” Nature,vol.463,no.7279,pp. [70] E. Lefai, M. Roques, N. Vega, M. Laville, and H. Vidal, 364–368, 2010. “Expressionofthesplicevariantsofthep85𝛼 regulatory subunit [55]W.Yang,Y.Zheng,Y.Xiaetal.,“ERK1/2-dependentphos- of phosphoinositide 3-kinase in muscle and adipose tissue phorylation and nuclear translocation of PKM2 promotes the of healthy subjects and type 2 diabetic patients,” Biochemical Warburg effect,” Nature Cell Biology,vol.14,no.12,pp.1295– Journal,vol.360,no.1,pp.117–125,2001. 1304, 2012. [71] S. Agrawal and C. Eng, “Differential expression of novel [56]J.L.Reiter,D.W.Threadgill,G.D.Eleyetal.,“Comparative naturally occurring splice variants of PTEN and their functional genomic sequence analysis and isolation of human and mouse consequences in Cowden syndrome and sporadic breast can- alternative EGFR transcripts encoding truncated receptor iso- cer,” Human Molecular Genetics,vol.15,no.5,pp.777–787,2006. forms,” Genomics,vol.71,no.1,pp.1–20,2001. [72] M. Hoogeveen-Westerveld, R. Ekong, S. Povey et al., “Func- [57] A. T. Baron, J. A. Wilken, D. E. Haggstrom, S. T. Goodrich, tional assessment of TSC2 variants identified in individuals with and N. J. Maihle, “Clinical implementation of soluble EGFR tuberous sclerosis complex,” Human Mutation,vol.34,no.1,pp. (sEGFR) as a theragnostic serum biomarker of breast, lung and 167–175, 2013. ovarian cancer,” IDrugs,vol.12,no.5,pp.302–308,2009. [73] K. Mayer, W. Ballhausen, W. Leistner, and H.-D. Rott, “Three [58] S. Maramotti, M. Paci, F. Micciche` et al., “Soluble epidermal novel types of splicing aberrations in the tuberous sclerosis growth factor receptor isoforms in non-small cell lung cancer TSC2 gene caused by mutations apart from splice consensus tissue and in blood,” Lung Cancer,vol.76,no.3,pp.332–338, sequences,” Biochimica et Biophysica Acta,vol.1502,no.3,pp. 2012. 495–507, 2000. [59] C. Halle, M. Lando, D. H. Svendsrud et al., “Membranous [74] L. Xu, C. Sterner, M. M. Maheshwar et al., “Alternative splicing expression of ectodomain isoforms of the epidermal growth of the tuberous sclerosis 2 (TSC2) gene in human and mouse factor receptor predicts outcome after chemoradiotherapy of tissues,” Genomics,vol.27,no.3,pp.475–480,1995. lymph node-negative cervical cancer,” Clinical Cancer Research, [75] G. Panasyuk, I. Nemazanyy, A. Zhyvoloup et al., “mTOR𝛽 vol. 17, no. 16, pp. 5501–5512, 2011. splicing isoform promotes cell proliferation and tumorigenesis,” [60] A. Philippou, A. Armakolas, and M. Koutsilieris, “Evidence for JournalofBiologicalChemistry,vol.284,no.45,pp.30807– the possible biological significance of the igf-1 gene alternative 30814, 2009. splicing in prostate cancer,” Frontiers in Endocrinology,vol.4, [76] M.-E.Huot,G.Vogel,A.Zabarauskasetal.,“TheSam68STAR´ article 31, 2013. RNA-binding protein regulates mTOR alternative splicing dur- [61] G. Chen, J. Wang, Z. Liu, and M. Kornmann, “Exon III splicing ing adipogenesis,” Molecular Cell,vol.46,no.2,pp.187–199, of fibroblast growth factor receptor 1 is modulated by growth 2012. factors and cyclin D1,” Pancreas,vol.37,no.2,pp.159–164,2008. [77] G. Michlewski, J. R. Sanford, and J. F. Caceres,´ “The splicing [62] G. Condorelli, R. Bueno, and R. J. Smith, “Two alternatively factor SF2/ASF regulates translation initiation by enhancing spliced forms of the human insulin-like growth factor I receptor phosphorylation of 4E-BP1,” Molecular Cell,vol.30,no.2,pp. have distinct biological activities and internalization kinetics,” 179–189, 2008. JournalofBiologicalChemistry,vol.269,no.11,pp.8510–8516, [78] V. Ben-Hur, P. Denichenko, Z. Siegfried et al., “S6K1 alter- 1994. native splicing modulates its oncogenic activity and regulates [63] C. Marchini, F. Gabrielli, M. Iezzi et al., “The human splice mTORC1,” Cell Reports,vol.3,no.1,pp.103–115,2013. variant 𝛿16HER2 induces rapid tumor onset in a reporter [79] K. W.Lynch, “Regulation of alternative splicing by signal trans- transgenic mouse,” PLoS ONE,vol.6,no.4,ArticleIDe18727, duction pathways,” Advances in Experimental Medicine and 2011. Biology,vol.623,pp.161–174,2007. [64] V. Veikkolainen, K. Vaparanta, K. Halkilahti, K. Iljin, M. Sund- [80] S. Stamm, “Regulation of alternative splicing by reversible vall, and K. Elenius, “Function of ERBB4 is determined by alter- protein phosphorylation,” Journal of Biological Chemistry,vol. native splicing,” Cell Cycle,vol.10,no.16,pp.2647–2657,2011. 283, no. 3, pp. 1223–1227, 2008. [65] M. Sundvall, V. Veikkolainen, K. Kurppa et al., “Cell death or [81]J.A.McCubrey,L.S.Steelman,W.H.Chappelletal.,“Ras/Raf/ survival promoted by alternative isoforms of ErbB4,” Molecular MEK/ERK and PI3K/PTEN/Akt/mTOR cascade inhibitors: Biology of the Cell,vol.21,no.23,pp.4275–4286,2010. howmutationscanresultintherapyresistanceandhowto [66] D. M. Aebersold, Y. D. Shaul, Y. Yung et al., “Extracellular overcome resistance,” Oncotarget, vol. 3, no. 10, pp. 1068–1111, Signal-Regulated Kinase 1c (ERK1c), a novel 42-kilodalton ERK, 2012. demonstrates unique modes of regulation, localization, and [82] E. S. White, R. L. Sagana, A. J. Booth et al., “Control of fibroblast function,” Molecular and Cellular Biology,vol.24,no.22,pp. fibronectin expression and alternative splicing via the PI3K/ 10000–10015, 2004. Akt/mTOR pathway,” Experimental Cell Research,vol.316,no. [67] Y. D. Shaul, G. Gibor, A. Plotnikov, and R. Seger, “Specific 16, pp. 2644–2653, 2010. phosphorylation and activation of ERK1c by MEK1b: a unique [83]J.C.Shultz,R.W.Goehe,D.S.Wijesingheetal.,“Alternative route in the ERK cascade,” Genes and Development,vol.23,no. splicing of caspase 9 is modulated by the phosphoinositide 3- 15, pp. 1779–1790, 2009. kinase/Akt pathway via phosphorylation of SRp30a,” Cancer [68] L. Adesso, S. Calabretta, F. Barbagallo et al., “Gemcitabine trig- Research,vol.70,no.22,pp.9185–9196,2010. gers a pro-survival response in pancreatic cancer cells through [84] N. A. Patel, S. Kaneko, H. S. Apostolatos et al., “Molecular activation of the MNK2/eIF4E pathway,” Oncogene,vol.32,no. and genetic studies imply Akt-mediated signaling promotes 23, pp. 2848–2857, 2013. protein kinase C𝛽II alternative splicing via phosphorylation of [69] S. Fransson, A. Uv, H. Eriksson et al., “p37delta is a new serine/arginine-rich splicing factor SRp40,” Journal of Biological isoform of PI3K p110delta that increases cell proliferation and Chemistry,vol.280,no.14,pp.14302–14309,2005. is overexpressed in tumors,” Oncogene,vol.31,no.27,pp.3277– [85] A. M. Al-Ayoubi, H. Zheng, Y. Liu, T. Bai, and S. T. Eblen, 3286, 2012. “Mitogen-activated protein kinase phosphorylation of splicing International Journal of Cell Biology 9

factor 45 (SPF45) regulates SPF45 alternative splicing site utilization, proliferation, and cell adhesion,” Molecular and Cel- lular Biology, vol. 32, no. 14, pp. 2880–2893, 2012. [86] Z.Zhou,J.Qiu,W.Liuetal.,“TheAkt-SRPK-SRaxisconstitutes a major pathway in transducing EGF signaling to regulate alternative splicing in the nucleus,” Molecular Cell,vol.47,no. 3, pp. 422–433, 2012. [87]E.Allemand,S.Guil,M.Myers,J.Moscat,J.F.Caceres,´ and A. R. Krainer, “Regulation of heterogenous nuclear ribonucleoprotein A1 transport by phosphorylation in cells stressed by osmotic shock,” Proceedings of the National Academy of Sciences of the United States of America,vol.102,no.10,pp.3605–3610, 2005. [88] W. V. D. H. Van Oordt, M. T. Diaz-Meco, J. Lozano, A. R. Krainer, J. Moscat, and J. F. Caceres,´ “The MKK(3/6)-p38- signaling cascade alters the subcellular distribution of hnRNP A1 and modulates alternative splicing regulation,” Journal of Cell Biology,vol.149,no.2,pp.307–316,2000. [89]S.Guil,J.C.Long,andJ.F.Caceres,´ “hnRNP A1 relocalization to the stress granules reflects a role in the stress response,” Molecular and Cellular Biology,vol.26,no.15,pp.5744–5758, 2006. [90] R. Karni, Y.Hippo, S. W.Lowe, and A. R. Krainer, “The splicingfactor oncoprotein SF2/ASF activates mTORC1,” Proceedings of the National Academy of Sciences of the United States of America, vol. 105, no. 40, pp. 15323–15327, 2008. [91] M. Blaustein, F. Pelisch, T. Tanos et al., “Concerted regulation of nuclear and cytoplasmic activities of SR proteins by AKT,” Nature Structural & Molecular Biology,vol.12,no.12,pp.1037– 1044, 2005. Hindawi Publishing Corporation International Journal of Cell Biology Volume 2013, Article ID 151839, 15 pages http://dx.doi.org/10.1155/2013/151839

Review Article Phosphorylation-Mediated Regulation of Alternative Splicing in Cancer

Chiara Naro1,2 and Claudio Sette1,2 1 Department of Biomedicine and Prevention, University of Rome “Tor Vergata”, 00133 Rome, Italy 2 Laboratories of Neuroembryology and of Cellular and Molecular Neurobiology, Fondazione Santa Lucia IRCCS, 00143 Rome, Italy

Correspondence should be addressed to Claudio Sette; [email protected]

Received 31 May 2013; Accepted 26 July 2013

Academic Editor: Michael Ladomery

Copyright © 2013 C. Naro and C. Sette. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Alternative splicing (AS) is one of the key processes involved in the regulation of gene expression in eukaryotic cells. AS catalyzes the removal of intronic sequences and the joining of selected exons, thus ensuring the correct processing of the primary transcript into the mature mRNA. The combinatorial nature of AS allows a great expansion of the genome coding potential, as multiple splice- variants encoding for different proteins may arise from a single gene. Splicing is mediated by a large macromolecular complex, the spliceosome, whose activity needs a fine regulation exerted by cis-acting RNA sequence elements and trans-acting RNA binding proteins (RBP). The activity of both core spliceosomal components and accessory splicing factors is modulated by their reversible phosphorylation. The kinases and phosphatases involved in these posttranslational modifications significantly contribute toAS regulation and to its integration in the complex regulative network that controls gene expression in eukaryotic cells. Herein, we will review the major canonical and noncanonical splicing factor kinases and phosphatases, focusing on those whose activity has been implicated in the aberrant splicing events that characterize neoplastic transformation.

1. Introduction regulative mechanism for both developmentally regulated and pathological cellular processes is now well recognized In eukaryotic cells, the expression of each gene is finely tuned (reviewed in [3]). by a complex network of regulative processes affecting all The splicing process is carried out by the spliceosome, a steps of transcript maturation, from nuclear transcription complex macromolecular machinery composed of five small to cytosolic export and utilization of the mRNA. A crucial nuclear ribonucleoprotein particles (U1, U2, U4, U5, and step in this regulative network is represented by pre-mRNA U6 snRNPs) and more than 200 auxiliary proteins. The splicing, the molecular process that mediates the removal of spliceosome mediates the recognition of the short consensus 󸀠 󸀠 intronic sequences and the joining of exons. What makes sequences surrounding the 5 -(GU) and the 3 -(AG) splice splicing an outstanding player in controlling gene expression site and catalyzes the two transesterification reactions neces- is its flexibility, which allows a remarkable increase of the sary for the removal of the intron and ligation of the selected coding potential of the genome through alternative selection exons (reviewed in [4]). Due to the degenerate nature of the of exons. Indeed, alternative splicing (AS) allows each gene sequence elements recognized by the spliceosome, its recruit- to encode for several coding and noncoding mRNA variants, ment to the maturing pre-mRNA requires the action of both which often display different activities and/or patterns of cis-acting RNA sequence elements and trans-acting RNA expression. AS is, therefore, one of the principal mechanisms binding proteins (RBPs), such as Ser/Arg (SR) rich proteins, underlying the well-known discrepancy between increasing heterogeneous nuclear ribonucleoproteins (hnRNPs), and organismal complexity and content of genes contained in splicing factors belonging to other RBP families. In addition, thegenome[1]. In line with its central contribution to AS is also regulated by mechanisms acting both co- and genome complexity, it is estimated that up to 90% of human posttranscriptionally, through epigenetic modifications of the multiexon genes undergo AS [2], and the importance of this chromatin, regulation of the RNA polymerase II (RNAPII) 2 International Journal of Cell Biology transcription rate, and posttranslational modifications of (previously known as SRp38). This SR protein acts as a both spliceosome components and auxiliary splicing factors, specific splicing activator in its RS-phosphorylated form16 [ ], among which reversible phosphorylation acts as a major whereas dephosphorylation converts it into a potent splicing player. repressor [17]. Notably, dephosphorylation of SRSF10 occurs duringtheMphaseofthecellcycle[17] or under stress condition [18], when general inhibition of splicing occurs. 2. Impact of Phosphorylation on In particular, it was demonstrated that, under normal condi- the Catalysis of Splicing tions, phosphorylated SRSF10 is a sequence-specific splicing activator, which promotes U1 and U2 snRNP assembly on Aproperregulationofthephosphorylationstatusofthe target pre-mRNAs endowed with SRSF10-dependent exonic spliceosomal proteins and of accessory splicing factors is splicing enhancer (ESE) sequences [16]. Conversely, under crucial for the correct regulation of both constitutive and AS stressful cellular conditions, as during heat shock, SRSF10 events. Early studies already described the importance of a is rapidly dephosphorylated by PP1, while other SR proteins correct balance between phosphorylation and dephosphory- are maintained in phosphorylated state by SR protein kinases lation events in the splicing process by showing that both (SRPKs) [19]. Interestingly, during the stress response all SR activation [5]andinhibition[6] of the PP1 and PP2A phos- proteins are similarly dephosphorylated by PP1. However, phatases are required for splicing catalysis. Several reports they are rapidly rephosphorylated by SRPKs, while SRSF10, have then further highlighted the importance of regulated which is a poor substrate for SRPKs, remains dephospho- phosphorylation events for the correct assembly and catalytic rylated. Under this condition, SRSF10 can still interact with activation of spliceosomal components, such as PRP28 [7], the U1 snRNP, but in this case the interaction impairs its 󸀠 PRP6, or PRP31 [8]. Equally, dephosphorylation events, such ability to recognize the 5 splice site, thus resulting in splicing as those regarding the U5 and U2 snRNP component, U5- inhibition [18]. 156 kDa and SAP155 [9, 10],wereshowntobeessentialfor Phosphorylation of the RS domain can also dictate spliceosome activity, proving the importance of subsequent SR protein subcellular localization, by affecting both their rounds of phosphorylation and dephosphorylation events in intranuclear localization and their nucleocytoplasmic shut- the regulation of the splicing process. tling. In interphase cells, SR proteins are enriched in Regulative phosphorylation and dephosphorylation interchromatin granules called nuclear speckles, which are events concern not only the spliceosomal components enriched in factors involved in pre-mRNA processing and but also some accessory RBPs that cooperate with the RNA transport (reviewed in [20]). The recruitment of SR spliceosome in the selection of splice sites. For example, proteins to nascent pre-mRNAs from these sites of storage phosphorylation of the splicing factors SF1 and SRSF1 (pro- is regulated by their phosphorylation (Figure 1); indeed, it totypic SR protein previously known as ASF/SF2) modulates has been shown that phosphorylation of the RS domain is their interaction with U2AF65 and U1snRNP, respectively, a prerequisite for their recruitment to transcription sites in thus modulating spliceosome assembly [11, 12]. The dynamic vivo [21]. This modification plays also an important role in phosphorylation/dephosphorylation of SR proteins is the regulation of nucleocytoplasmic shuttling of SR proteins. particularly relevant for the regulation of their functions, as For instance, phosphorylation of SR proteins in the cytoplasm both hypo- and hyperphosphorylation can inhibit splicing is required for their nuclear import [22]. On the other hand, [13]. For instance, phosphorylation of SRSF1 promotes dephosphorylation of the RS domain is essential for their spliceosome assembly, whereas its dephosphorylation is translocation to the cytoplasm during mRNP maturation necessary for the catalysis of the first transesterification (Figure 1)[23]. Interestingly, dephosphorylation of SRSF1 and reaction [14]. SRSF7 (previously known as 9G8) enhances their interaction with the export receptor TAP, thereby favoring also the 3. Phosphorylation and Splicing Factors export of their target mRNAs [24, 25]. Furthermore, SRSF1 translational activity is increased by dephosphorylation of its SR proteins are a family of nuclear RBPs involved in the reg- RS domain (Figure 1)[26]. Phosphorylation has therefore a ulation of both constitutive and AS, whose activity is greatly greatimpactalsoonthesplicingunrelatedfunctionsinwhich modulated by reversible phosphorylation. Their structure is many SR proteins are involved (reviewed in [27]). generally characterized by two N-terminal RNA recogni- Ser/Thr phosphorylation represents an important regulation motif (RRM) a C-terminal region enriched in Arg-Ser tive process not only for SR proteins but also for hnRNPs and residues (RS domain), which are the main targets of regulative other splicing factors. For instance, hnRNP A1 is phosphory- phosphorylation. Phosphorylation of the RS domain of SR lated by the mitogen-activated protein kinase (MAPK) p38 in proteins has a great impact on their functionality, as it may response to stress conditions (Figure 2), thus causing its cyto- affect their binding to target mRNAs, their interaction with plasmic translocation and consequent modulation of hnRNP other proteins and their intracellular localization. As an A1-sensitive AS events [28, 29]. Similarly, phosphorylation of example, binding of SRSF5 (previously reported as SRp40) the SR-like protein TRA2-𝛽 by CLK2 induces its relocaliza- to its high-affinity RNA-binding site is strictly dependent tion into the cytoplasm, thus reducing its ability to bind its on the phosphorylation of its RS domain [15]. One of the own mRNA and regulate its splicing [30]. For other splicing most significant examples of how phosphorylation may affect factors, instead, Ser/Thr phosphorylation affects the splicing the splicing activity of SR proteins is represented by SRSF10 activity by modulating their interaction with other proteins. International Journal of Cell Biology 3

Cytoplasm Nucleus P SRps P CLKs P SRps SRps P SRps Nuclear speckles P P P PPase P SRps SRps P P P SRps SRps SRps SRps P Nascent pre-mRNA SRPKs P SRps SRps

P SRps P P P P SRps P SRps P mRNA export pre-mRNA SRPKs SRps

SRps SRps P PPase SRps mRNA translation SRps Spliced mRNA

SRps SRps

P RNA pol II Spliceosome SRps Hypophosphorylated SR proteins

P P P SRps Ribosomes Unphosphorylated SR proteins SRps Hyperphosphorylated SR proteins

Figure 1: Phosphorylation-mediated regulation of SR proteins activity. Reversible phosphorylation of their RS domain profoundly affect SR protein (SRps) activity and subcellular localization. Newly synthesized SRps need SRPK-mediated phosphorylation in order to enter the nucleus and assemble in nuclear speckles. CLKs mediate SRps hyperphosphorylation and induce their release from nuclear speckles and their recruitment to transcription sites. Dephosphorylation of SRps is successively required for proper splicing catalysis. Moreover, dephosphorylated SRps facilitate export of spliced mRNA in the cytosol, where they enhance protein translation.

For instance, phosphorylation of hnRNP L reduces its inter- atargetpre-mRNA(Figure 2). This was formally shown by action with the U2AF65 subunit of the U2 auxiliary factor [31, studying its effect on the CCND1 gene. Increased expression 32]. The function of some splicing factors can be influenced of SAM68 promotes splicing of the cyclin D1b variant of also by Tyr phosphorylation. A well-documented example in the CCND1 gene in prostate cancer cells. This activity is this sense is SAM68, a member of the signal transduction furtherenhancedbyactivationoftheRAS/ERKpathway and activation of RNA (STAR) family of RBPs (reviewed andcounteractedbySFKs[38]. In both cases, the effect in [33]). Tyr phosphorylation by SRC-family kinases (SFKs) was due to modulation of the affinity for RNA, as ERK- caused the accumulation of SAM68 in nuclear granules, dependent phosphorylation increased binding of SAM68 to named SAM68 nuclear bodies (SNBs) [34, 35]. Moreover, it intron 4, whereas SFK-dependent phosphorylation abolished was shown that this posttranslational modification negatively it (Figure 2)[38]. Thus, activation of signaling pathways affected the interaction of SAM68 with hnRNP A1 and can indirectly modulate AS events through posttranslational with the BCL-X pre-mRNA,thusimpairingitsabilityto modification of selected splicing factors (see also later). promote splicing of this target gene [35]. On the other hand, Ser/Thr phosphorylation of SAM68 by the MAPKs ERK1/2 4. Splicing Factor Kinases was reported to increase splicing of the variable exons in the CD44 gene [36, 37]. Notably, SAM68 represents an interesting Phosphorylation of spliceosomal components and splicing example of how Ser/Thr and Tyr-phosphorylation may have factors is mediated by numerous protein kinases. Some of opposite impact on the splicing activity of an RBP toward these kinases, such as the SRPK and CLK families, are 4 International Journal of Cell Biology

Stress stimuli Growth factors Other ligands

PI3K

p38 RAS nRTKs AKT

hnRNPA1 RAF HSP70 ERK HSP90 SRPKs MEK HSP70 MNK1/2 JNK

ERK HSP90 SRPKs CLKs

hnRNPA1 SRPKs hnRNPL SRps SPF45 SAM68

Figure 2: Signaling-activated kinases regulate splicing factor activity. Various extracellular cues, like growth factors or stress stimuli, activate different signal-transduction cascades impinging on protein kinases that in turn phosphorylate RBPs, thereby modulating their splicing activity. SAM68 splicing activity is inversely regulated by ERKs and nRTKs, which, respectively, activate and inhibit its splicing activity. The PI3 K-AKT pathway regulates the activity of several SR proteins both directly or by phosphorylating and modulating the activity and localization of CLKs and SRPKs. Stress signal-activated kinases, like JNK or p38, can both modulate splicing factor localization, like for hnRNPA1, or activity, like for SPF45 (see text for details). specifically devoted to this function, whereas others also domain [41] and of their interaction with the molecular participate to signal transduction pathways or phosphorylate chaperones HSP70 and HSP90, which in complex seem distinct primary substrates in addition to the splicing factors. to favor the folding of SRPKs into an active state [42]. Herein, we will review the kinases whose ability to influence However, SRPKs can translocate into the nucleus of cells splicing decisions has been better characterized. For conve- under several conditions, such as during the G2/M phase nience, we will classify them as SR-protein specific kinases; of the cell cycle [39], or after osmotic stress42 [ ], or as a signaling-activated splicing kinases, and “atypical” splicing consequence of activation of the epidermal growth factor factor kinases. (EGF) signal transduction pathway [43]. Due to this dual localization, SRPKs can phosphorylate SR proteins both in 5. SR-Protein Specific Kinases the nucleus and in the cytoplasm, thus affecting several aspects of their function. SRPK-mediated phosphorylation 5.1. SR-Protein Kinases (SRPKs). The first SR protein kinase of SR proteins in the cytoplasm is necessary to ensure identified was SRPK1, which was isolated from mitotic cells, SR proteins nuclear import (Figure 1)[44], as it enhances and it was described to phosphorylate SR proteins and to their interaction with the specific transportin SR222 [ , 45]. promote their release from nuclear speckles during the G2/M SRPK nuclear activity promotes release of SR proteins in the phase of the cell cycle [39]. However, SRPK1 is present and nucleoplasm from the nuclear speckles [46]. For instance, active also in interphase cells. SRPK1 is the prototype of several reports suggest that SRPK-mediated phosphorylation the SRPK family, which also includes the two homologous of SRSF1 is essential for its nuclear localization and the SRPK2 and SRPK3 proteins. SRPKs are characterized by resulting splicing activity triggered by stimulation of specific a bipartite catalytic domain separated by a unique spacer signaling pathways (i.e., IGF-1 and EGF treatments) [43, 47]. sequence(reviewedin[40]) and are mainly localized in the However, under conditions that strongly increase nuclear cytoplasm of mammalian cells. This is due to the presence of localization of SRPKs, such as under cellular stress, they can a strong cytoplasmic retention signal localized in the spacer also induce nuclear speckles enlargement [42, 48]. Indeed, International Journal of Cell Biology 5

Zhong and colleagues showed that osmotic stress induced by 6. Signaling-Activated Splicing Factor Kinases sorbitol treatment can lead to a massive nuclear translocation of SRPK1, which causes hyperphosphorylation of SR proteins AS represents a crucial step in the regulation of gene expres- and inhibits their splicing activity toward the reporter E1A sion in eukaryotic cells. Therefore, its regulation needs to minigene [42]. These studies indicated that SRPK-mediated be finely integrated in the complex network of regulative phosphorylation of SR proteins can finely tune their splicing mechanisms that allows the cell to modulate gene expression activity in response to external and internal cues. in response to the different physiological and pathological stimuli that are received from both the internal and external 5.2. Cyclin-Dependent Like Kinases (CLKs). The cyclin- environment. In support of this notion, activation of signal dependent like kinases (CLK1-4) represent the other proto- transduction pathways has been shown to modulate AS in typicalfamilyofSRproteinkinases.Theyarecharacterized a large number of situations. However, while in some cases by a C-terminal kinase domain, with dual specificity, and the mechanism(s) has been described, in other cases the an N-terminal RS domain, which allows their interaction transacting factors mediating the response are unknown. with the SR proteins. CLKs colocalize with SR proteins in Here we will review signaling-activated kinases that can nuclear speckles, and their overexpression leads to hyper- modulate AS by directly phosphorylating splicing factors or phosphorylation of SR proteins and induces speckles dis- their regulators, such as the SRPKs or CLKs. assembly [49]. Several studies reported the ability of CLKs to influence splicing events by regulating the subnuclear 6.1. AKT. The Ser/Thr kinase AKT, also known as PKB, is localization of SR proteins (Figure 1). In particular, the release the hinge molecule of the phosphoinositide-3-kinase-protein of SR proteins from nuclear speckles induced by CLKs kinase (PI3 K) signaling pathway, which transduces the signal overexpression has been reported to modulate splicing of the of several growth factors and cytokines. Through the phos- E1A reporter minigene [50]andoftheexon10oftheTAU phorylation of its many nuclear and cytosolic targets, AKT gene [51], whose aberrant regulation has been implicated can regulate a multitude of cellular processes, such as cell in several neurodegenerative diseases. Recently it has been metabolism, proliferation, and survival (reviewed in [56]). shown that CLKs also modulate the activity of splicing To exert its multiple functions, AKT regulates different steps factors not related to the SR-protein family, such as SPF45. of the gene expression network, from transcription, to AS CLK-mediated phosphorylation of SPF45 interferes with its and translation. Indeed, several reports have highlighted the proteasomal degradation and enhances exon 6 inclusion of ability of AKT to directly and indirectly modulate the func- FAS by promoting binding of this splicing factor to the tion of many RBPs. AKT phosphorylates both hnRNPs and FAS pre-mRNA [52]. The nuclear localization of CLKs is SR proteins, which contain within their RS domain multiple one of the major differences between them and SRPKs, AKT phosphorylation consensus sequences: RXRXX(S/T) which are instead mainly cytosolic. Because of their different [57]. By modulating their phosphorylation status, AKT localization, CLKs and SRPKs can cooperate in regulating SR regulates both splicing and splicing independent functions proteins subcellular localization. Indeed, it has been shown of hnRNPs and SR proteins. For example, AS of a four- that SRPK1 interacts with SRSF1 and phosphorylates the N- exon cassette in the CASPASE-9 gene allows expression of a terminal part (RS1) of its RS domain, a posttranslational mod- proapoptotic splice variant (exon inclusion, CASPASE-9a)or ification that is essential for its assembly into nuclear speckles, an antiapoptotic splice variant, (exon skipped, CASPASE-9b). whereas CLKs phosphorylate the C-terminal part (RS2) of AKT-dependent phosphorylation of hnRNP L increases its itsRSdomain,therebycausingreleaseofSRSF1fromthe affinity for exon 3 and induces expression of the antiapoptotic speckles [53]. Moreover, SRPKs and CLKs have also distinct variant. Indeed, AKT-phosphorylated hnRNP L competes substrate specificity, as SRPKs preferentially phosphorylate with hnRNP U for the binding to the mRNA and impairs Ser-Arg sites, while CLKs have a broader specificity and can its ability to promote the pro-apoptotic CASPASE-9a [58] phosphorylate also Ser-Lys or Ser-Pro sites [54]. Therefore, (Figure 2).ThisisaclearexampleofhowAKTmaypromote even if apparently redundant, the coordinated activity of cell survival by regulating a key AS event. On the other hand, SRPKs and CLKs is crucial for correct splicing regulation. phosphorylation of hnRNP A1 by AKT has no effects on This was well illustrated by Nowak and colleagues, whose splicing, but it modulates the translational activity of this work highlighted how SR-proteins phosphorylation induced RBP.Following phosphorylation, hnRNP A1 loses its ability to by these two families of kinases may differently control a promote IRES dependent translation of the CCND1 and the c- single splicing event [55]. The vascular endothelial growth MYC mRNAs [59].InthecaseofSRproteins,AKTwasshown factor A (VEGFA) gene, a key regulator of angiogenesis, to modulate both splicing and translational activity through produces several isoforms by alternative splice-site selection phosphorylation. Growth factor-induced phosphorylation of in the terminal exon 8: proximal splice-site selection results SRSF1 and SRSF7 by AKT enhanced their ability to promote in proangiogenic VEGFxxx isoforms, whereas distal splice- the inclusion of the EDA exon in the fibronectin mRNA site selection results in antiangiogenic isoforms VEGFxxxb. and translation of the spliced mRNA [60]. One of the most Different growth factors inversely influence these splicing characterized AS events regulated by AKT is affecting PKC𝛽 events by inducing in both cases phosphorylation of SR pre-mRNA after insulin stimulation. This hormone induces proteins. However, IGF-1 and TNF-𝛼 induced production splicing of the PKC𝛽 II isoform, which enhances insulin- of VEGFxxx through activation of SRPKs, whereas TGF-𝛽1 stimulated glucose transport better than the PKC𝛽 I variant, enhanced VEGFxxxb production by activating CLKs [55]. even if they differ only for two residues in their C-terminus 6 International Journal of Cell Biology

[61]. Insulin stimulation induces PI3K-dependent activation other spliceosomal components [69]. SAM68 is not the only of AKT, which phosphorylates SRSF5 [62, 63], thus pro- RBP regulated by ERK1/2. Furthermore, other MAPKs, like moting PKC𝛽 II splicing. Furthermore, AKT phosphorylates p38 or JNKs, are known to phosphorylate and modulate the CLK1 and enhances its activity (Figure 2). In turn, CLK1 activity of splicing factors. For example, it has been recently phosphorylates SRSF4 (previously named SRp75) and SRSF6 demonstrated that phosphorylation of the splicing factor (previously named SRp55), thus contributing to the PKC𝛽 SPF45 can be mediated by all the three families of MAPKs II splicing regulation [64]. Recently, it has been suggested in response to different stimuli (e.g., oxidative stress activates thatAKTandCLKmayalsoregulateSRSF5splicingactivity ERK1/2 and JNK mediated phosphorylation, whereas UV- by affecting its nuclear localization, which was impaired light induces p38 and JNK activity) [70](Figure 2). These when a CLK mutant that cannot be phosphorylated by AKT kinases phosphorylate SPF45 on two residues, Thr 71 and is expressed [65]. Importantly, this concerted regulation of Ser 222; these posttranslational modifications inhibit SPF45- SRSF5-dependent PKC𝛽 II splicing by AKT and CLK was dependent exon 6 inclusion in the FAS gene, thus leading to essential for adypogenetic differentiation, thus providing the production of a dominant negative isoform of this death physiological relevance for this signaling route. receptor [70]. AKT was also shown to regulate the activity of SRPKs. Modulation of hnRNP A1 activity by p38 is another well A recent work documented that EGF signaling induces a characterized regulative phosphorylation event operated by massive reprogramming of AS that depends on AKT-induced a MAPK. Environmental stresses, such as osmotic stress or nuclear translocation of SRPKs [43]. In fact, AKT binding to UV irradiation, induce p38 activation and phosphorylation SRPKs induces their autophosphorylation and dissociation of hnRNP A1, leading to the relocalization of this nuclear RBP from the HSP70 chaperone, which normally holds SRPKs into the cytoplasm, where it concentrates into discrete phase- into the cytoplasm, thus favoring their nuclear translocation dense particles, called stress granules (SGs) [29](Figure 2). guided by HSP90 (Figure 2). Once in the nucleus, SRPKs The nuclear exclusion of hnRNP A1 is the result of its reduced can phosphorylate SR proteins and modulate the splicing interaction with the transportin Trn1, which under normal pattern of several genes. Thus, given its ability to modulate the conditions mediates its nuclear translocation [71], and leads activity of both regulators (SRPKs and CLKs) and effectors to consequent modulation of hnRNP A1-dependent splicing (SRproteinsandhnRNPs)ofAS,AKTstandsupasacrucial events, which were tested using the E1A minigene reporter player in the modulation of splicing in response to external [28]. HnRNP A1 phosphorylation is mediated by the p38 cues, and this activity might represent a primary function of effectors MAP kinase signal-integrating kinases (MNK1/2) AKT in the regulation of multiple cellular processes. [28], which can also regulate the translational activity of this splicing factor. It was observed that the increase in TNF- 6.2. Mitogen-Activated Protein Kinases (MAPKs). MAPKs are 𝛼 protein production following T-cell activation relies on a family of Ser/Thr kinases that transduce external signals MNK-mediated phosphorylation of hnRNP A1. However, in into the cell and regulate many different cellular processes, this cellular context, phosphorylation of hnRNP A1 does not such as metabolism, proliferation, survival, differentiation, affect its localization, but it rather lowers its affinity for the 󸀠 and motility (reviewed in [66]). The MAPK family includes AU-rich element (ARE) in the 3 UTR of the TNF-𝛼 mRNA, the extracellular regulated kinases (ERK 1/2), the c-Jun amino thus probably relieving a translation repressive control and terminal kinases (JNK 1–3), p38 (𝛼, 𝛽, 𝛾,and𝛿), and ERK5 allowing enhanced TNF-𝛼 production [72]. family.TheroleofMAPKsinthesecellularprocessesis Thus, MAPKs can regulate different steps of mRNA pro- mediated by regulation of protein activity and stability and cessing through phosphorylation of several splicing factors, by modulation of gene expression, which also occurs through integrating in this way this complex regulative step of gene AS. The first evidence of MAPK-modulated splicing came expression with the response of the cell to external cues. from studies on the regulation of AS of the CD44 gene, which encodes for the extracellular receptor for hyaluronic 6.3. Tyrosine Kinases. Protein tyrosine kinases (PTKs) cat- acid, a key component of the extracellular matrix. The CD44 alyze the transfer of a phosphate group from ATP to a gene is characterized by a block of variable exons (v2–v10) tyrosine residue of their target proteins. PTKs may be clas- embedded between ten constant exons; the inclusion of the sified in two different classes: the transmembrane receptors variable exons into the mature transcript modulates CD44 tyrosine kinases (RTKs) and the nonreceptor tyrosine kinases protein interaction with its substrate, thus significantly affect- (nRTKs). PTKs mediate the phosphorylation of several pro- ing cell adhesion, migration, and proliferation (reviewed in teins in response to both internal and external cues, leading to [67]).Theinclusionofthevariableexonv5inthemature the modification of their activity or affecting their interaction mRNA of CD44 upon T-cell activation is dependent by the with other proteins. Transduction pathways triggered by PTK RAS-RAF-MEK-ERK signaling cascade [68]. The target of activation affect gene expression, also at the level of AS, even this pathway is SAM68, whose ability to promote exon v5 though only a small number of splicing factors have been inclusion is increased by ERKs-mediated phosphorylation shown to be regulated by Tyr-phosphorylation. Among these [36]. SAM68 interacts with the splicing factor U2AF65, fewRBPs,themembersoftheSTARproteinsfamily,andin and this interaction seems to enhance the recognition of particular SAM68, stand out (Figure 2). In many STAR family 󸀠 the 3 splice-site. Phosphorylation by ERKs reduces the members, the RNA binding domain is flanked by regulatory affinity of the SAM68/U2AF65 complex to the CD44 pre- regions, like proline-rich or tyrosine-rich sequences, which mRNA, probably favoring the subsequent recruitment of mediate their interaction with the Src Homology 3 (SH3) International Journal of Cell Biology 7 and SH2 domains of other proteins, including PTKs. For demonstrated that forskolin, which stimulates the synthesis instance, the breast tumor kinase (BRK) is a nRTKs that of cAMP, modulates AS of exon 10 of the TAU gene [80, interacts in the nucleus with a proline rich region of SAM68 81]. Notably, activated PKA affects the activity of two SR through its SH3 domain. BRK-dependent phosphorylation of proteins, SRSF1 and SRSF7, which inversely modulate exon 10 SAM68 reduces its RNA binding affinity73 [ ]. Analogously, splicing: SRSF1 promotes exon 10 inclusion, whereas SRSF7 BRK phosphorylates also the SAM68 homologous proteins prevents it. However, PKA-dependent phosphorylation of SLM-1 and SLM-2, reducing their affinity to the RNA also SRSF1 enhances its activity [80] whereas it inhibits SRSF7 [81], in this case [74]. SAM68 is also substrate of FYN, another thus globally favoring exon 10 inclusion. soluble nRTKs. In this case, it was also shown that Tyr- PKAisalsoabletomodulateASofgenesthatarecrucial phosphorylation interfered with SAM68-dependent splicing for neuronal differentiation, through the phosphorylation of of the BCL-X and CCND1 genes [35, 38]. FYN-dependent hnRNP K. After phosphorylation by PKA, hnRNP K shows phosphorylation reduced the affinity of SAM68 for these tar- higher binding activity to its target mRNAs with respect to its get RNAs and affected its interaction with different proteins, competitor U2AF65; this mechanism impairs the recognition 󸀠 such as hnRNP A1, thereby affecting the outcome of AS events of the 3 splice site and leads to the skipping of its target exons [35, 38]. Tyr-phosphorylation also influences the splicing [82]. On a broader view, hnRNP K target motifs are found activity of the nuclear RBP YT521-B, which can also be elicited in many genes involved in neuronal differentiation and in by several nRTKs such as FYN, SRC, or c-ABL [75, 76]. neurological diseases [82]. These pieces of evidence suggest This posttranslational modification induces translocation of that PKA mediated regulation of hnRNPs and SR proteins YT521-BfromthenuclearYTbodies,whereitnormally activity may be an important player in the complex network resides, to the nucleoplasm. Phosphorylated YT521-B shows of regulative mechanisms that finely control AS events during reduced ability to modulate splice-site selection of different neuronal development (reviewed in [83]). Although cAMP targets, in association with a reduced binding to their mRNA, and PKA are usually involved in cell differentiation, their possibly because the nucleoplasmic translocation distances contribution to cancer has also been demonstrated. It will YT521-B from the effective sites of pre-mRNA processing be interesting to investigate whether PKA-dependent mod- [76]. ulationofASalsooccursingeneswithrelevancetohuman cancer. 6.4. cAMP-Dependent Protein Kinase (PKA). Increased intracellular levels of the second messenger cyclic adenosine 󸀠 󸀠 7. Other Kinases 3 ,5 -monophosphate (cAMP) lead to the activation of the cAMP-dependent protein kinase (PKA), which In this section, we will describe the regulative activity of transduces the signals of many hormones, growth factors, some proteins that showed an unexpected kinase activity and neurotransmitters [77]. PKA is a tetrameric protein, towards splicing factors, so that they cannot be included in composed of two regulatory subunits (R) and two catalytic any of the classes described previously. Some of these kinases subunits (C): binding of cAMP to the R subunits leads to were known to have other specific substrates different from theirdissociationfromtheCsubunitandactivationofthe splicing factors, for others, instead, the kinase activity was kinase [77]. Activated PKA phosphorylates several effectors, totally unpredicted. including transcription factors, ion channels, and metabolic enzymes, thus influencing multiple cellular functions. PKA activity is also regulated by interaction of the R subunits with 7.1. DNA Topoisomerase I. The first of these atypical kinases the PKA-anchoring proteins (AKAPs); AKAPs maintain to be described was the DNA topoisomerase I, whose best PKA in specific subcellular compartments and in proximity knownfunctionistorelievebothpositiveandnegative of its substrates, thus retaining PKA activity where it is DNA supercoils ensuring correct DNA topology during needed. The first evidence of a possible involvement of PKA transcription, DNA replication and repair (reviewed in [84]). in the regulation of AS came from the observation that Despite the absence of a canonical ATP binding site, DNA a fraction of the C subunit translocates into the nucleus, topoisomerase I was shown to phosphorylate SR proteins, colocalizes with SRSF2 (previously reported as SC35) in in particular the prototypic SRSF1, within their RS domain splicing speckles, and phosphorylates several SR proteins, at [85]. This phosphorylation event can significantly affect SR least in vitro [78]. Localization of the C subunit in nuclear proteins modulatory activity on AS events. Indeed, it has speckles seems to be related to its interaction with the been demonstrated that cells deficient for this enzyme show C-subunit binding protein HA95 [78] and to the SR protein a general status of hypophosphorylation for the SR proteins, SRSF17A, which was shown to be a novel AKAP required to which correlates with an impaired regulation of several AS anchor PKA C subunit in splicing speckles [79]. Importantly, events, whereas constitutive splicing results unaffected [86]. modulation of E1A reporter minigene splicing by SRSF17A Moreover, treatment with a selective inhibitor of the kinase is dependent on its interaction with PKA [79]. Moreover, activity of DNA topoisomerase results in reduced phospho- nuclear PKA itself is able to modulate AS of the E1A reporter rylation levels for SR proteins, which in turn leads to a minigene,evenintheabsenceofthecAMPstimulation[78]. defective spliceosome assembly and alterations in the splicing Several stimuli that increase the cAMP intracellular levels pattern of several genes [87]. As it is now well established that were shown to affect AS events through phosphorylation of pre-mRNA splicing occurs cotranscriptionally, it has been both SR proteins and hnRNPs by PKA. For example, it was suggested that this double activity of DNA topoisomerase I 8 International Journal of Cell Biology could be one of the mechanisms ensuring the correct coordi- be interesting to determine how many targets are influenced nation between DNA transcription and splicing [88]. Indeed, by this kinase in T cells. DNA topoisomerase I activity is fundamental to solve DNA supercoils generated by RNA pol II progression along the 7.4. Aurora Kinase A (AURKA). AURKA was identified in DNA template and might simultaneously ensure a regulated a high-throughput siRNA screening for factors involved in splicing factor activity through their phosphorylation. the regulation of AS of two apoptotic genes: BCL-X and MCL1 [98]. Among several regulators identified by the screen, 7.2. Dual-Specificity Tyrosine-(Y)-Phosphorylation Regulated authors noticed a peculiar enrichment for proteins involved Kinase 1A (DIRK1A). Another protein kinase able to mod- in the regulation of the cell cycle. They focused their study on ulate the splicing activity of SR proteins is DIRK1A (dual- AURKA,akinaseinvolvedintheregulationofcentrosomal specificity tyrosine-(Y)-phosphorylation regulated kinase splitting that is frequently upregulated in cancers, where 1A). This dual-specificity protein kinase autophosphorylates it is supposed to promote aneuploidy [99]. AURKA was on Tyr, Ser, and Thr residues but phosphorylates substrates demonstrated to positively regulate splicing of the antiapop- only on Ser or Thr residues (reviewed in [89]). The human totic variant 𝐵𝐶𝐿-𝑋𝐿 through stabilization of SRSF1. Cells DYRK1A gene maps to chromosome 21, and it is ubiquitously depleted of AURKA showed reduced levels of SRSF1, which expressed in adult and fetal tissues, with high levels of expres- then resulted in increased levels of the 𝐵𝐶𝐿-𝑋𝑆 pro-apoptotic sion in the brain. DYRK1A is supposed to play a major role variant. Moreover, since AURKA is activated at the G2/M during neuronal development, through its interaction with phase of the cell cycle, the authors suggested that this kinase several cytoskeletal, synaptic, and nuclear proteins (reviewed links BCL-X splicing regulation to cell cycle progression. in [89]). Several SR proteins were shown to interact with These observations suggest that, in addition to the effects on DYRK1A. Indeed, DYRK1A has been reported to colocal- centrosome duplication, upregulation of AURKA can favor ize with SRSF2 in nuclear speckles, and its overexpression neoplastic transformation also by promoting antiapoptotic induces the disassembly of this subnuclear structures [90]. splice variants. Alteration of subcellular localization of the SR proteins phosphorylated by DYRK1A seems to be the main mechanism by 8. Splicing Factor Kinases in Cancer and which this kinase regulates the splicing activity of its target Other Human Diseases factors. For instance, phosphorylation of SRSF1 and SRSF7 by DYRK1A induces their cytoplasmic translocation [91, 92], Due to the important role played by the AS process in the whereas phosphorylation of SRSF2 and SRSF6 causes their control of gene expression, any alterations of its regulation dissociation from nuclear speckles [93, 94]. For each of these can profoundly modify important cellular processes, thus splicing factors the mislocalization induced by DYRK1A resulting in a potential cause of disease (reviewed in [100]). impaired their ability to modulate the inclusion of exon 10 of Altered expression, activity, or subcellular localization of the TAU gene, thus shifting the splicing balance toward the splicing factor kinases can be among the causes of the aber- exclusion of this exon. rant splicing events associated to several diseases, particularly neurodegenerative pathologies and cancer. 7.3. Fas-Activated Serine/Threonine Kinase (FAST). Fas-acti- Aberrant inclusion of exon 10 of the TAU gene is a vated serine/threonine kinase (FAST) is a constitutively phos- well-known example of pathogenetic splicing event, caused phorylated Ser/Thr kinase, which undergoes rapid dephos- by the deregulated activity or expression of splicing factor phorylation after the binding of Fas ligand to its receptor kinases. TAU protein is a microtubule associated protein, Fas, an interaction that triggers T-cell apoptosis. It was which controls assembly and stability of microtubules. Exon known that dephosphorylated FAST was able to interact with 10 of the TAU gene encodes for one of the four microtubule and phosphorylate the RBP TIA1 [95], but the functional binding domain repeats (R) of the TAU protein and regulates relevance of this interaction in the regulation of the splicing its affinity for microtubules and, consequently, its ability to process remained unknown for a long time. It was later induce their polymerization. Alternative inclusion of exon discovered that phosphorylation of TIA1 by FAST regulates 10 leads to the production of either 4R-tau (inclusion) or its ability to promote the inclusion of exon 6 of the FAS gene 3R-tau (exclusion), and equal levels of these two isoforms [96]. Phosphorylated TIA1 enhances U1 snRNP recruitment seem to be essential for normal function of the human to FAS pre-mRNA, thus favoring the recognition of this brain. Alteration of the normal ratio 1 : 1 between the 4R variable exon. Inclusion of exon 6 into the FAS mRNA and the 3R isoform, in both directions, has been observed favors the production of a proapoptotic isoform of this in several cases of Alzheimer’s disease (AD); moreover, gene, suggesting that FAST and TIA1 take part to a positive nearly half of the mutations in the TAU gene associated regulative circuitry that enhances Fas-dependent apoptosis with FTDP-17 (frontotemporal dementia with parkinsonism once activated. Furthermore FAST is also endowed with an linked to chromosome 17) affects exon 10 splicing, both intrinsic splicing activity, independent from TIA1 [97]. It was inhibiting or promoting its inclusion, strongly suggesting observed that FAST can modulate the splicing of the FGFR2 that a proper regulation of this splicing event is essential for reporter gene in the same direction of TIA1, favoring the themaintenanceofthehealthybalancebetween4Rand3R inclusion of exon III b but independently from this RBP.Thus, isoforms (reviewed in [101]).Severalreportshavehighlighted FAST can directly and indirectly affect splicing, and it would or suggested a strong correlation between aberrant splicing of International Journal of Cell Biology 9 the exon 10 of the TAU gene in tauopathies and deregulated their upstream regulator could be a potential tool to target activity of the kinases regulating this splicing event. Stamm’s pathological angiogenesis in cancer [48]. group, for example, observed an increased production of Several signal transduction kinases, whose activity is an inactive isoform of the CLK2 kinase in the brain of often deregulated by neoplastic transformation, exert their AD patients, which correlated with increased inclusion of oncogenic activity in part through the aberrant regulation TAU exon 10. This observation suggested that the CLK2- of splicing events. For instance the MAPK pathway, which dependent phosphorylation of SR proteins and the SR-like is frequently hyperactivated in tumors, can promote the protein TRA2-𝛽 is required for the correct regulation of this acquisition of an invasive and migratory phenotype by splicing event [102].Anotherkinasesupposedtobeinvolved modulatingtheASpatternofthecelladhesionmolecule in the altered regulation of TAU splicing is PKA, which also CD44. In fact, it has been shown that hepatocyte growth promotes the inclusion of the exon 10 of this gene through factor (HGF) can induce cell migration of cancer cells by the phosphorylation of different SR proteins [80]. As reduced promoting this splicing event, as a consequence of induced levels of PKA-C𝛼 have been observed in AD brains [103], it ERK1/2-mediated phosphorylation of SAM68, induced by has been speculated that the lack of its activity may participate the MET receptor signaling pathway [110]. Also epithelial- in the alteration of the normal balance between the 3R and 4R to-mesenchymal transition (EMT), which is crucial for the splice variants of TAU [81]. As described in previous section, invasiveness of cancer cells, is regulated by AS events that the kinase DYRK1A exerts an important regulation on the AS are sensitive to activation of the MAPK pathway. Indeed, of the TAU gene, thus strongly suggesting that its increased production of the constitutively active ΔRON splice variant of dosage due to the trisomy of chromosome 21 could be the the RON oncogene, the extracellular receptor for HGF, leads main cause of the early onset of tauopathies in patients with to EMT in colorectal cancer cells [111]. This splicing event is Down syndrome [103]. Modulation of TAU gene splicing is promoted by the upregulation of SRSF1. Remarkably, under a very attractive potential therapeutic target for treatment conditions that favor EMT, epithelial cells release soluble of tauopathies (reviewed in [104]); since protein kinases factors that activate the ERK1/2 pathway. This in turn causes regulate this splicing event and are involved in tauopathy phosphorylation and activation of SAM68, which causes 󸀠 pathogenesis, targeting the activity of these kinases should be retention of a cryptic intron in the 3 UTR of the SRSF1 certainly considered in the development of future approaches mRNA, reducing the amount of the nonsense-mediated- for the treatment of these pathologies. decay (NMD) targeted splice variant and enhancing expres- Upregulation and/or misregulated activity of splicing kin- sion of SRSF1 [112]. Thus, activation of the ERK1/2 pathway asesareoftenassociatedtocancerdevelopment.Thishas triggers a cascade of splicing events that culminate in a been widely reported for SRPK1, which is overexpressed in cellular response favoring cancer cell invasion. several cancer types, such as pancreatic carcinomas [105], Activation of the AKT pathway has also been suggested breast and colon carcinomas [106], and lung cancer [107]. to promote cancer cell survival through the regulation of Moreover, increased SRPK1 levels positively correlate with specific splicing events. For example, it has been observed tumor grade [106]andareassociatedwithhigherresis- that hyperactivation of AKT through the RAS signaling tance to chemotherapeutic treatments [105, 108]. Through pathwayisimplicatedintheproductionofpro-survivalsplice variants of the KLF-6 and CASPASE-9 genes in nonsmall- modulation of selective splicing events, SRPK1 may allow cell lung cancer and hepatocellular carcinoma, respectively, cancer cells to enhance their proliferative, invasive, and [113, 114]. In both cases, AKT induces SRSF1 phosphorylation, angiogenetic potential. For example, in pancreatic, breast, enhancing its ability to promote KLF-6SV1 and CASPASE-9b and colon cancer cells SRPK1 promotes the generation of isoforms. These observations strongly suggest a primary role specific splice variants of the MAP2K2 gene, which sustained for this splicing regulatory activity in the oncogenic potential higher activity of the MAPK pathway [106]. Recently, SRSF1 of AKT. mediated splicing of the MNK2b isoform of the MKNK2 gene has been correlated with resistance to gemcitabine treatment in pancreatic cancer cells [109]; since SRPK1 is upregulated 9. Protein Phosphatase Regulating in this cancer type and promotes cell survival, it would be Splicing Factors interesting to evaluate whether this kinase contributes to the SRSF1-induced prosurvival pathway. A similar regulation has In the previous paragraphs, we have broadly described the been described in Wilms Tumor, wherein SRPK1 promotes importance of a proper balance between phosphorylation the production of the proangiogenic isoform VEGF165 of and dephosphorylation events in the regulation of the pre- the VEGFA gene through the phosphorylation and nuclear mRNA splicing process. Therefore, even if this review focuses translocation of SRSF1 [48]. In these nephroblastomas tumors primarily on the activity of the numerous kinases involved in SRPK1 transcriptional upregulation is driven by the mutated this regulation, a brief description of the protein phosphatases transcription factor WT1, and its splicing activity is fun- counteracting their activity is also required for a comprehen- damental for the high levels of vascularization required by sive overview of the phosphorylative regulation of splicing. these tumors [48]. Importantly, the physiological relevance of PP1 and PP2A were the first Ser/Thr phosphatases SRPK1 for angiogenesis has been demonstrated, as injection whose activity was demonstrated to be necessary for splic- of an SRPK1/2 inhibitor reduced it in a mouse model of retinal ing catalysis [5, 6]. PP1 and PP2A are required for the neovascularization, suggesting that targeting AS through later steps of the splicing reaction, in particular for the 10 International Journal of Cell Biology second transesterification reaction, whose accomplishment which could be used for the therapeutic correction of the is favored by dephosphorylation of U5 (U5-156 kDa) and splicing defects occurring in several human diseases. U2 (SAP155) snRNP components [9]. In particular, PP1- mediated dephosphorylation of SAP155 is favored by the 10. Concluding Remarks nuclear inhibitor of PP1 (NIPP1). NIPP1 is a nuclear regulatory subunit of PP1, enriched in nuclear speckles [115], known Increasing evidence points out to a key role of misregula- to interact with several splicing regulators, like CDC5L or tion of AS in the cellular transformation process. Cancer- the same SAP155 [116, 117]. In particular, NIPP1 stimulates specific splice variants can potentially be used as accu- PP1-mediated dephosphorylation of SAP155 by facilitating rate diagnostic and prognostic markers, as it was recently the interaction between the phosphatase and its substrate highlighted by genome-wide studies [126, 127]. Targeting [10]. In subsequent studies, PP2C𝛾 was also demonstrated to the splicing process represents, therefore, an attractive ther- be important for the splicing process, as it was shown to be apeutic target for cancer treatment, and it is currently physically associated with the spliceosome, and its enzymatic under intense investigation. Therapeutic modulation of AS is activity was necessary for the early steps of spliceosome mainly realized through RNA-based technologies (reviewed assembly [118]. in [128]) or through chemical reagents inhibiting spliceosome Ser/Thr phosphatases are important regulators of both activity (reviewed in [129]). The RNA-based technologies constitutive and AS events, as it was suggested by pioneering exploit antisense oligonucleotide masking specific sequence 󸀠 studies showing alternative 5 splice selection after addition elements to splicing factors and/or the spliceosome [130], of PP1 in splicing assay in vitro [119]. Furthermore, PP2C𝛾 whereas chemical approaches make use of drugs that directly was shown to interact with the RBP YB-1 and to modulate target the activity of spliceosomal components, as for exam- AS of the CD44 gene [120], while PP1 was demonstrated ples spliceostatin A, which inhibits the SF3b subunit of the U2 to interact with a short motif RVXF motif within the RRM snRNP,thereby modulating the AS of genes important for cell of several splicing factors, like SRSF1, SRSF9 (previously cycle control [131]. known as SRp30C), and the SR-like protein TRA2-𝛽 [121]. Considering the important control exerted by protein Dephosphorylation of TRA2-𝛽 by PP1 positively modulates kinases on AS, modulation of their activity represents a its dimerization and its interaction with partner proteins, like potential approach for the development of new drugs target- SRSF1. Moreover, PP1 regulates alternative splice selection in ing RNA splicing in cancer therapy. These suggestions are TRA2-𝛽 target mRNAs like the SMN2 gene [121]. Exclusion supported by recent reports highlighting the high efficacy of the exon 7 of SMN2 gene, combined with the primary of SRPK1/2 inhibitors in reducing angiogenesis through the deletion of SMN1 gene, is the cause of the spinal muscular negative modulation of the AS of the proangiogenic splice atrophy (SMA) [122]. TRA2-𝛽 promotes the inclusion of the variant VEGFxxx gene [48]. Considering the great impact exon 7 of SMN2 favoring the production of a functional that SRPKs have on the splicing activity of SR proteins and the full length protein. TRA2-𝛽 splicing activity is enhanced by largenumberofASeventsthattheyregulate,modulationof inhibition of PP1 activity [121] and, surprisingly, by activation SRPKactivitycouldbeapowerfultoolintheemergingfieldof of PP2A [123]. Indeed, the Stamm’s group found that a splicing-modulating therapies. It is also important to mention class of compounds derivative from cantharidin (a well- that SRSF1, a well-known target of SRPKs, is upregulated in known phosphatase inhibitors) activates PP2A, which in turn human cancers and functions as an oncogene [132]. For the dephosphorylates TRA2-𝛽 on Thr33, favoring inclusion of same reasons, CLKs are a fascinating chemotherapeutic target exon 7 [123]. These observations suggest the possibility to too, and important efforts are being made for the realization develop new protein phosphatase inhibitors that could be of selective and efficient CLKs inhibitors [133]. used for the therapeutic correction of the splicing defects Signal-transduction pathways able to modulate the phos- occurring in neurodegenerative diseases like SMA. phorylationstatusofSRproteinsortheactivityofotherRBPs Modulating protein phosphatases’ activity in order to represent another potential druggable target for RNA splicing manipulate pathogenetic splicing events has been suggested modulation. For example, it has been recently shown that as a potential therapeutic tool also for cancer treatment. amiloride,awell-knowndiuretic,canreduceproliferative and invasive properties of both hepatocellular carcinoma and Indeed, it has been shown that genotoxic agents inducing leukemia cancer cells by inducing hypophosphorylation of apoptosis in cancer cells act through the generation of SR-proteins [134, 135]. Genome-wide exon array analysis has ceramide and activation of PP1, which in turn promotes 𝐵𝐶𝐿 𝑋 demonstrated that amiloride treatment induces the modula- the formation of the proapoptotic - 𝑆 and CASPASE-9b tion of a large number of AS events, and, in particular, it neg- splice variants [124]. On the other hand, it has been shown atively regulates protumoral splice variants of several genes, that the proapoptotic activity of synthetic ceramides, like C6 such as the antiapoptotic 𝐵𝐶𝐿-𝑋𝐿 or proinvasive ΔRON. pyridinium ceramide, is instead associated with activation Reduced phosphorylation levels of AKT and ERKs were of PP1 and the consequent reduced phosphorylation of observed after amiloride treatment, suggesting that this drug several splicing factors and modulation of several splicing reduces SR protein phosphorylation through inactivation of events [125]. These observations underline the importance these kinases [135]. of the regulated activity of protein phosphatases for proper Deregulation of signal-transduction pathways in cancer regulation of the splicing process and strongly suggest the cells is a general feature, and much effort has been made possibility to develop new molecules targeting their activity, in order to develop chemotherapeutic agents that efficiently International Journal of Cell Biology 11 inhibit the activity of the kinases mediating the intracellular [8]M.Schneider,H.-H.Hsiao,C.L.Will,R.Giet,H.Urlaub,and transduction of these signals, such as AKT or kinases of the R. Luhrmann,¨ “Human PRP4 kinase is required for stable tri- MAPK and SRC families (reviewed in [136–138]). As many snRNP association during spliceosomal B complex formation,” of these inhibitors are already in clinical practice, and many Nature Structural and Molecular Biology,vol.17,no.2,pp.216– of them are undergoing promising clinical trials, it would be 221, 2010. very interesting to understand whether their antiproliferative [9]Y.Shi,B.Reddy,andJ.L.Manley,“PP1/PP2Aphosphatasesare and cytotoxic effects could be partly due to their ability to required for the second step of Pre-mRNA splicing and target interfere with AS events regulated by these kinases. Even specific snRNP proteins,” Molecular Cell,vol.23,no.6,pp.819– 829, 2006. more attractive is the possibility to exploit protein kinase [10] N. Tanuma, S.-E. Kim, M. Beullens et al., “Nuclear inhibitor inhibitors to selectively affect splicing decisions in order to of protein phosphatase-1 (NIPP1) directs protein phosphatase- restoreincancercellsanormal,nonpathologicalASpattern. 1 (PP1) to dephosphorylate the U2 small nuclear ribonucleo- Shedding light on the expression, structure, and functions protein particle (snRNP) component, spliceosome-associated of the kinases regulating the activity of splicing factors is protein 155 (Sap155),” JournalofBiologicalChemistry,vol.283, therefore an important step for a comprehensive under- no.51,pp.35805–35814,2008. standing of the molecular mechanisms regulating pre-mRNA [11] V. Manceau, M. Swenson, J.-P. Le Caer, A. Sobel, C. L. Kielkopf, processing, which is essential for the rational design of future and A. Maucuer, “Major phosphorylation of SF1 on adjacent therapies targeting the aberrant AS process in cancer and Ser-Pro motifs enhances interaction with U2AF65,” FEBS Jour- other human diseases. nal,vol.273,no.3,pp.577–587,2006. [12] S. Cho, A. Hoang, R. Sinha et al., “Interaction between the RNA binding domains of Ser-Arg splicing factor 1 and U1-70K snRNP Conflict of Interests protein determines early spliceosome assembly,” Proceedings of the National Academy of Sciences of the United States of America, The authors declare no conflict of interests. vol. 108, no. 20, pp. 8233–8238, 2011. [13]J.Prasad,K.Colwill,T.Pawson,andJ.L.Manley,“Theprotein Acknowledgments kinase Clk/Sty directly modulates SR protein activity: both hyper- and hypophosphorylation inhibit splicing,” Molecular The work in the laboratory of C. Sette was supported by and Cellular Biology, vol. 19, no. 10, pp. 6991–7000, 1999. Association for International Cancer Research (AICR Grant [14] W. Cao, S. F. Jamison, and M. A. Garcia-Blanco, “Both phos- no.12-0150),theAssociazioneItalianaRicercasulCancro phorylation and dephosphorylation of ASF/SF2 are required for (AIRC), and Istituto Superiore della Sanita(ISSGrantCon-` pre-mRNA splicing in vitro,” RNA,vol.3,no.12,pp.1456–1467, venzione 11US/6A). 1997. [15]R.Tacke,Y.Chen,andJ.L.Manley,“Sequence-specificRNA binding by an SR protein requires RS domain phosphorylation: References creation of an SRp40-specific splicing enhancer,” Proceedings of the National Academy of Sciences of the United States of America, [1] N. L. Barbosa-Morais, M. Irimia, Q. Pan et al., “The evolu- vol. 94, no. 4, pp. 1148–1153, 1997. tionary landscape of alternative splicing in vertebrate species,” [16]Y.Feng,M.Chen,andJ.L.Manley,“Phosphorylationswitches Science,vol.338,no.6114,pp.1587–1593,2012. the general splicing repressor SRp38 to a sequence-specific [2] M. Sultan, M. H. Schulz, H. Richard et al., “A global view of activator,” Nature Structural and Molecular Biology,vol.15,no. gene activity and alternative splicing by deep sequencing of the 10, pp. 1040–1048, 2008. human transcriptome,” Science,vol.321,no.5891,pp.956–960, [17] C. Shin and J. L. Manley, “The SR protein SRp38 represses 2008. splicing in M phase cells,” Cell, vol. 111, no. 3, pp. 407–417, 2002. [3] M. Irimia and B. J. Blencowe, “Alternative splicing: decoding an [18] C. Shin, Y. Feng, and J. L. Manley, “Dephosphorylated SRp38 expansive regulatory layer,” Current Opinion in Cell Biology,vol. acts as a splicing repressor in response to heat shock,” Nature, 24,no.3,pp.323–332,2012. vol. 427, no. 6974, pp. 553–558, 2004. [4]C.L.WillandR.Luhrmann,¨ “Spliceosome structure and [19] Y. Shi and J. L. Manley, “A complex signaling pathway regulates function,” Cold Spring Harbor Perspectives in Biology,vol.3,no. SRp38 phosphorylation and pre-mRNA splicing in response to 7, 2011. heat shock,” Molecular Cell, vol. 28, no. 1, pp. 79–90, 2007. [5]J.E.Mermoud,P.Cohen,andA.I.Lamond,“Ser/Thr-specific [20] D. L. Spector and A. I. Lamond, “Nuclear speckles,” Cold Spring proteinphosphatasesarerequiredforbothcatalyticstepsof Harbor Perspectives in Biology,vol.3,no.2,2011. pre-mRNA splicing,” Nucleic Acids Research, vol. 20, no. 20, pp. [21] T. Misteli, J. F. Caceres,´ J. Q. Clement, A. R. Krainer, M. F. 5263–5269, 1992. Wilkinson, and D. L. Spector, “Serine phosphorylation of SR [6] J. E. Mermoud, P. T. W. Cohen, and A. I. Lamond, “Regulation proteins is required for their recruitment to sites of transcrip- of mammalian spliceosome assembly by a protein phosphorylation in vivo,” JournalofCellBiology,vol.143,no.2,pp.297–307, tion mechanism,” EMBO Journal, vol. 13, no. 23, pp. 5679–5688, 1998. 1994. [22] M.-C. Lai, R.-I. Lin, and W.-Y.Tarn, “Transportin-SR2 mediates [7]R.Mathew,K.Hartmuth,S.Mohlmann,¨ H. Urlaub, R. Ficner, nuclear import of phosphorylated SR proteins,” Proceedings of and R. Luhrmann,¨ “Phosphorylation of human PRP28 by the National Academy of Sciences of the United States of America, SRPK2 is required for integration of the U4/U6-U5 tri-snRNP vol. 98, no. 18, pp. 10154–10159, 2001. into the spliceosome,” Nature Structural and Molecular Biology, [23] S. Lin, R. Xiao, P. Sun, X. Xu, and X.-D. Fu, vol. 15, no. 5, pp. 435–443, 2008. “Dephosphorylation-dependent sorting of SR splicing factors 12 International Journal of Cell Biology

during mRNP maturation,” Molecular Cell,vol.20,no.3,pp. [39]J.-F.Gui,W.S.Lane,andX.-D.Fu,“Aserinekinaseregulates 413–425, 2005. intracellular localization of splicing factors in the cell cycle,” [24] Y.Huang, T. A. Yario, and J. A. Steitz, “Amolecular link between Nature,vol.369,no.6482,pp.678–682,1994. SR protein dephosphorylation and mRNA export,” Proceedings [40] G. Ghosh and J. A. Adams, “Phosphorylation mechanism and of the National Academy of Sciences of the United States of structure of serine-arginine protein kinases,” FEBS Journal,vol. America, vol. 101, no. 26, pp. 9666–9670, 2004. 278,no.4,pp.587–597,2011. [25] M.-C. Lai and W.-Y.Tarn, “Hypophosphorylated ASF/SF2 binds [41] J.-H. Ding, X.-Y. Zhong, J. C. Hagopian et al., “Regulated cel- TAP and is present in messenger ribonucleoproteins,” Journal of lular partitioning of SR protein-specific kinases in mammalian Biological Chemistry,vol.279,no.30,pp.31745–31749,2004. cells,” Molecular Biology of the Cell,vol.17,no.2,pp.876–885, [26] J. R. Sanford, J. D. Ellis, D. Cazalla, and J. F. Caceres,´ “Reversible 2006. phosphorylation differentially affects nuclear and cytoplasmic [42]X.-Y.Zhong,J.-H.Ding,J.A.Adams,G.Ghosh,andX.-D. functions of splicing factor 2/alternative splicing factor,” Pro- Fu, “Regulation of SR protein phosphorylation and alternative ceedings of the National Academy of Sciences of the United States splicing by modulating kinetic interactions of SRPK1 with of America,vol.102,no.42,pp.15042–15047,2005. molecular chaperones,” Genes and Development,vol.23,no.4, [27] X.-Y. Zhong, P. Wang, J. Han, M. G. Rosenfeld, and X.-D. Fu, pp. 482–495, 2009. “SR proteins in vertical integration of gene expression from [43] Z.Zhou,J.Qiu,W.Liuetal.,“TheAkt-SRPK-SRaxisconstitutes transcription to RNA processing to translation,” Molecular Cell, a major pathway in transducing EGF signaling to regulate vol.35,no.1,pp.1–10,2009. alternative splicing in the nucleus,” Molecular Cell,vol.47,no. [28] W. V. D. H. Van Oordt, M. T. Diaz-Meco, J. Lozano, A. R. 3,pp.422–433,2012. Krainer, J. Moscat, and J. F. Caceres,´ “The MKK(3/6)-p38- [44] C. Y. Yun and X.-D. Fu, “Conserved SR protein kinase functions signaling cascade alters the subcellular distribution of hnRNP in nuclear import and its action is counteracted by arginine A1 and modulates alternative splicing regulation,” Journal of Cell methylation in Saccharomyces cerevisiae,” Journal of Cell Biol- Biology,vol.149,no.2,pp.307–316,2000. ogy,vol.150,no.4,pp.707–717,2000. [29]S.Guil,J.C.Long,andJ.F.Caceres,´ “hnRNP A1 relocalization [45] C. Y. Yun, A. L. Velazquez-Dones, S. K. Lyman, and X.-D. Fu, to the stress granules reflects a role in the stress response,” “Phosphorylation-dependent and -independent nuclear import Molecular and Cellular Biology,vol.26,no.15,pp.5744–5758, of RS domain-containing splicing factors and regulators,” Jour- 2006. nal of Biological Chemistry, vol. 278, no. 20, pp. 18050–18055, [30] P. Stoilov, R. Dauod, O. Nayler, and S. Stamm, “Human tra2- 2003. beta1 autoregulates its protein concentration by infuencing [46] H.-Y. Wang, W. Lin, J. A. Dyck et al., “SRPK2: a differentially alternative splicing of its pre-mRNA,” Human Molecular Genet- expressed SR protein-specific kinase involved in mediating the ics,vol.13,no.5,pp.509–524,2004. interaction and localization of pre-mRNA splicing factors in [31] R. W. Goehe, J. C. Shultz, C. Murudkar et al., “hnRNP L mammalian cells,” JournalofCellBiology,vol.140,no.4,pp.737– regulates the tumorigenic capacity of lung cancer xenografts in 750, 1998. mice via caspase-9 pre-mRNA processing,” Journal of Clinical Investigation,vol.120,no.11,pp.3923–3939,2010. [47] D. G. Nowak, E. M. Amin, E. S. Rennel et al., “Regulation of Vascular Endothelial Growth Factor (VEGF) splicing from [32] G. Liu, A. Razanau, Y. Hai et al., “A conserved serine of pro-angiogenic to anti-angiogenic isoforms: a novel therapeutic heterogeneous nuclear ribonucleoprotein L (hnRNP L) medi- strategy for angiogenesis,” Journal of Biological Chemistry,vol. ates depolarization-regulated alternative splicing of potassium 285, no. 8, pp. 5532–5540, 2010. channels,” TheJournalofBiologicalChemistry,vol.287,no.27, pp. 22709–22716, 2012. [48] E. M. Amin, S. Oltean, J. Hua et al., “WT1 mutants reveal [33] C. Sette, “Post-translational regulation of star proteins and SRPK1 to be a downstream angiogenesis target by altering effects on their biological functions,” Advances in Experimental VEGF splicing,” Cancer Cell, vol. 20, no. 6, pp. 768–780, 2011. Medicine and Biology,vol.693,pp.54–66,2010. [49] K. Colwill, T. Pawson, B. Andrews et al., “The Clk/Sty protein [34] K. E. Lukong, D. Larocque, A. L. Tyner, and S. Richard, kinase phosphorylates SR splicing factors and regulates their “Tyrosine phosphorylation of Sam68 by breast tumor kinase intranuclear distribution,” EMBO Journal,vol.15,no.2,pp.265– regulates intranuclear localization and cell cycle progression,” 275, 1996. Journal of Biological Chemistry,vol.280,no.46,pp.38639– [50]P.I.Duncan,D.F.Stojdl,R.M.Marius,andJ.C.Bell,“Invivo 38647, 2005. regulation of alternative pre-mRNA splicing by the Clk1 protein [35] M. P. Paronetto, T. Achsel, A. Massiello, C. E. Chalfant, and kinase,” Molecular and Cellular Biology,vol.17,no.10,pp.5996– C. Sette, “The RNA-binding protein Sam68 modulates the 6001, 1997. alternative splicing of Bcl-x,” Journal of Cell Biology,vol.176,no. [51] A. M. Hartmann, D. Rujescu, T. Giannakouros et al., “Regula- 7, pp. 929–939, 2007. tion of alternative splicing of human tau exon 10 by phosphory- [36] N.Matter,P.Herrlich,andH.Konig,¨ “Signal-dependent regula- lation of splicing factors,” Molecular and Cellular Neuroscience, tion of splicing via phosphorylation of Sam68,” Nature,vol.420, vol.18,no.1,pp.80–90,2001. no.6916,pp.691–695,2002. [52] Y. Liu, L. Conaway, J. Rutherford Bethard et al., “Phosphoryla- [37] C. Cheng and P. A. Sharp, “Regulation of CD44 alternative tion of the alternative mRNA splicing factor 45 (SPF45) by Clk1 splicingbySRm160anditspotentialroleintumorcellinvasion,” regulates its splice site utilization, cell migration and invasion,” Molecular and Cellular Biology,vol.26,no.1,pp.362–370,2006. Nucleic Acids Research,vol.41,no.9,pp.4949–4962,2013. [38] M. P. Paronetto, M. Cappellari, R. Busaetal.,“Alternative` [53] J.C.K.Ngo,S.Chakrabarti,J.-H.Dingetal.,“Interplaybetween splicing of the cyclin D1 proto-oncogene is regulated by the SRPK and Clk/Sty kinases in phosphorylation of the splicing RNA-binding protein Sam68,” Cancer Research,vol.70,no.1, factor ASF/SF2 is regulated by a docking motif in ASF/SF2,” pp. 229–239, 2010. Molecular Cell,vol.20,no.1,pp.77–89,2005. International Journal of Cell Biology 13

[54]K.Colwill,L.L.Feng,J.M.Yeakleyetal.,“SRPK1and [69] A. Tisserant and H. Konig,¨ “Signal-regulated pre-mRNA occu- Clk/Sty protein kinases show distinct substrate specificities pancy by the general splicing factor U2AF,” PLoS ONE,vol.3, for serine/arginine-rich splicing factors,” Journal of Biological no. 1, Article ID e1418, 2008. Chemistry,vol.271,no.40,pp.24569–24575,1996. [70] A. M. Al-Ayoubi, H. Zheng, Y. Liu, T. Bai, and S. T. Eblen, [55] D. G. Nowak, J. Woolard, E. M. Amin et al., “Expression of “Mitogen-activated protein kinase phosphorylation of splicing pro- and anti-angiogenic isoforms of VEGF is differentially factor 45 (SPF45) regulates SPF45 alternative splicing site regulated by splicing and growth factors,” Journal of Cell Science, utilization, proliferation, and cell adhesion,” Molecular and vol. 121, no. 20, pp. 3487–3495, 2008. Cellular Biology,vol.32,no.14,pp.2880–2893,2012. [56] T. F. Franke, “PI3K/Akt: getting it right matters,” Oncogene,vol. [71] E. Allemand, S. Guil, M. Myers, J. Moscat, J. F. Caceres,´ and 27,no.50,pp.6473–6488,2008. A. R. Krainer, “Regulation of heterogenous nuclear ribonucleoprotein A1 transport by phosphorylation in cells stressed by [57]T.Obata,M.B.Yaffe,G.G.Leparcetal.,“Peptideand osmotic shock,” Proceedings of the National Academy of Sciences protein library screening defines optimal substrate motifs for of the United States of America,vol.102,no.10,pp.3605–3610, AKT/PKB,” JournalofBiologicalChemistry, vol. 275, no. 46, pp. 2005. 36108–36115, 2000. [72] M. Buxade,´ J. L. Parra, S. Rousseau et al., “The Mnks are [58] N. T. Vu, M. A. Park, J. C. Shultz et al., “hnRNP U enhances novel components in the control of TNF𝛼 biosynthesis and caspase-9 splicing and is modulated by AKT-dependent phos- phosphorylate and regulate hnRNP A1,” Immunity,vol.23,no. phorylation of hnRNP L,” The Journal of Biological Chemistry, 2,pp.177–189,2005. vol. 288, no. 12, pp. 8575–8584, 2013. [73] J. J. Derry, S. Richard, H. V. Carvajal et al., “Sik (BRK) [59] O. D. Jo, J. Martin, A. Bernath, J. Masri, A. Lichtenstein, and J. phosphorylates sam68 in the nucleus and negatively regulates Gera, “Heterogeneous nuclear ribonucleoprotein A1 regulates Its RNA binding ability,” Molecular and Cellular Biology,vol.20, cyclin D1 and c-myc internal ribosome entry site function no.16,pp.6114–6126,2000. through Akt signaling,” JournalofBiologicalChemistry,vol.283, [74] A. Haegebarth, D. Heap, W. Bie, J. J. Derry, S. Richard, and A. no.34,pp.23274–23287,2008. L. Tyner, “The nuclear tyrosine kinase BRK/Sik phosphorylates [60] M. Blaustein, F. Pelisch, T. Tanos et al., “Concerted regulation and inhibits the RNA-binding activities of the Sam68-like of nuclear and cytoplasmic activities of SR proteins by AKT,” mammalian proteins SLM-1 and SLM-2,” Journal of Biological Nature Structural and Molecular Biology,vol.12,no.12,pp.1037– Chemistry, vol. 279, no. 52, pp. 54398–54404, 2004. 1044, 2005. [75] A. M. Hartmann, O. Nayler, F. W. Schwaiger, A. Obermeier, [61]C.E.Chalfant,H.Mischak,J.E.Watsonetal.,“Regulationof and S. Stamm, “The interaction and colocalization of Sam68 alternative splicing of protein kinase C𝛽 by insulin,” Journal of with the splicing-associated factor YT521-B in nuclear dots is Biological Chemistry,vol.270,no.22,pp.13326–13332,1995. regulated by the Src family kinase p59(fyn),” Molecular Biology of the Cell,vol.10,no.11,pp.3909–3926,1999. [62]N.A.Patel,C.E.Chalfant,J.E.Watsonetal.,“Insulin regulates alternative splicing of protein kinase C beta II through [76] I. Rafalska, Z. Zhang, N. Benderska et al., “The intranuclear a phosphatidylinositol 3-kinase-dependent pathway involving localization and function of YT521-B is regulated by tyrosine the nuclear serine/arginine-rich splicing factor, SRp40, in skele- phosphorylation,” Human Molecular Genetics, vol. 13, no. 15, pp. tal muscle cells,” Journal of Biological Chemistry,vol.276,no.25, 1535–1549, 2004. pp. 22648–22654, 2001. [77] P. Sassone-Corsi, “The cyclic AMP pathway,” Cold Spring Harbor Perspectives in Biology,vol.4,no.12,2012. [63]N.A.Patel,S.Kaneko,H.S.Apostolatosetal.,“Molecular and genetic studies imply Akt-mediated signaling promotes [78] A.-K. Kvissel, S. Ørstavik, S. Eikvar et al., “Involvement of the protein kinase C𝛽II alternative splicing via phosphorylation of catalytic subunit of protein kinase A and of HA95 in pre-mRNA serine/arginine-rich splicing factor SRp40,” Journal of Biological splicing,” Experimental Cell Research,vol.313,no.13,pp.2795– Chemistry, vol. 280, no. 14, pp. 14302–14309, 2005. 2809, 2007. [64] K. Jiang, N. A. Patel, J. E. Watson et al., “Akt2 regulation of [79] E. Jarnæss, A. J. Stokka, A.-K. Kvissel et al., “Splicing factor arginine/serine-rich 17A (SFRS17A) is an A-kinase anchoring Cdc2-like kinases (Clk/Sty), serine/arginine-rich (SR) protein phosphorylation, and insulin-induced alternative splicing of protein that targets protein kinase A to splicing factor com- PKC𝛽JII messenger ribonucleic acid,” Endocrinology,vol.150, partments,” JournalofBiologicalChemistry,vol.284,no.50,pp. 35154–35164, 2009. no. 5, pp. 2087–2097, 2009. [80] J. Shi, W. Qian, X. Yin et al., “Cyclic AMP-dependent protein [65] P. Li, G. Carter, J. Romero et al., “Clk/STY (cdc2-like kinase 1) kinase regulates the alternative splicing of tau exon 10: a and Akt regulate alternative splicing and adipogenesis in 3T3- mechanism involved in tau pathology of Alzheimer disease,” L1 pre-adipocytes,” PLoS ONE,vol.8,no.1,ArticleIDe53268, JournalofBiologicalChemistry,vol.286,no.16,pp.14639– 2013. 14648, 2011. [66] M. Cargnello and P. P. Roux, “Activation and function of [81] J.Gu,J.Shi,S.Wuetal.,“CyclicAMP-dependentproteinkinase the MAPKs and their substrates, the MAPK-activated protein regulates 9G8-mediated alternative splicing of tau exon 10,” kinases,” Microbiology and Molecular Biology Reviews,vol.75, FEBS Letters,vol.586,no.16,pp.2239–2244,2012. no. 1, pp. 50–83, 2011. [82]W.Cao,A.Razanau,D.Feng,V.G.Lobo,andJ.Xie,“Control [67] H. Ponta, L. Sherman, and P.A. Herrlich, “CD44: from adhesion of alternative splicing by forskolin through hnRNP K during molecules to signalling regulators,” Nature Reviews Molecular neuronal differentiation,” Nucleic Acids Research,vol.40,no.16, Cell Biology,vol.4,no.1,pp.33–45,2003. pp. 8059–8071, 2012. [68]S.Weg-Remers,H.Ponta,P.Herrlich,andH.Konig,¨ “Regula- [83] S. Zheng and D. L. Black, “Alternative pre-mRNA splicing tion of alternative pre-mRNA splicing by the ERK MAP-kinase in neurons: growing up and extending its reach,” Trends in pathway,” EMBO Journal,vol.20,no.15,pp.4194–4203,2001. Genetics,vol.29,no.8,pp.442–448,2013. 14 International Journal of Cell Biology

[84] J. C. Wang, “Cellular roles of DNA topoisomerases: a molecular shortened survival in head and neck squamous cell carcinoma,” perspective,” Nature Reviews Molecular Cell Biology,vol.3,no. Clinical Cancer Research,vol.12,no.17,pp.5136–5141,2006. 6, pp. 430–440, 2002. [100] T. A. Cooper, L. Wan, and G. Dreyfuss, “RNA and disease,” Cell, [85] E. Labourier, F. Rossi, I.-E. Gallouzi, E. Allemand, G. Divita, vol. 136, no. 4, pp. 777–793, 2009. and J. Tazi, “Interaction between the N-terminal domain of [101] A. Andreadis, “Tau splicing and the intricacies of dementia,” human DNA topoisomerase I and the arginine-serine domain Journal of Cellular Physiology,vol.227,no.3,pp.1220–1225,2012. of its substrate determines phosphorylation of SF2/ASF splicing [102]D.C.Glatz,D.Rujescu,Y.Tangetal.,“Thealternativesplicing factor,” Nucleic Acids Research, vol. 26, no. 12, pp. 2955–2962, of tau exon 10 and its regulatory proteins CLK2 and TRA2- 1998. BETA1 changes in sporadic Alzheimer’s disease,” Journal of [86] J. Soret, M. Gabut, C. Dupon et al., “Altered serine/arginine- Neurochemistry,vol.96,no.3,pp.635–644,2006. rich protein phosphorylation and exonic enhancer-dependent [103]Z.Liang,F.Liu,I.Grundke-Iqbal,K.Iqbal,andC.-X.Gong, splicing in Mammalian cells lacking topoisomerase I,” Cancer “Down-regulation of cAMP-dependent protein kinase by over- Research, vol. 63, no. 23, pp. 8203–8211, 2003. activated calpain in Alzheimer disease brain,” Journal of Neuro- [87] B. Pilch, E. Allemand, M. Facompre´ et al., “Specific inhibition chemistry,vol.103,no.6,pp.2462–2470,2007. of serine- and arginine-rich splicing factors phosphorylation, [104] J. Zhou, Q. Yu, and T. Zou, “Alternative splicing of exon 10 spliceosome assembly, and splicing by the antitumor drug NB- in the tau gene as a target for treatment of tauopathies,” BMC 506,” Cancer Research,vol.61,no.18,pp.6876–6884,2001. Neuroscience,vol.9,no.2,articleS10,2008. [88] M. Malanga, A. Czubaty, A. Girstun, K. Staron, and F. [105]G.M.Hayes,P.E.Carrigan,A.M.Beck,andL.J.Miller, R. Althaus, “Poly(ADP-ribose) binds to the splicing factor “Targeting the RNA splicing machinery as a novel treatment ASF/SF2 and regulates its phosphorylation by DNA topoiso- strategy for pancreatic carcinoma,” Cancer Research,vol.66,no. merase I,” JournalofBiologicalChemistry,vol.283,no.29,pp. 7, pp. 3819–3827, 2006. 19991–19998, 2008. [106] G. M. Hayes, P. E. Carrigan, and L. J. Miller, “Serine-arginine [89] F. J. Tejedor and B. Hammerle,¨ “MNB/DYRK1A as a multiple protein kinase 1 overexpression is associated with tumorigenic regulator of neuronal development,” FEBS Journal,vol.278,no. imbalance in mitogen-activated protein kinase pathways in 2, pp. 223–235, 2011. breast, colonic, and pancreatic carcinomas,” Cancer Research, [90] M. Alvarez,´ X. Estivill, and S. de la Luna, “DYRK1A accumulates vol.67,no.5,pp.2072–2080,2007. in splicing speckles through a novel targeting signal and induces [107] S. Gout, E. Brambilla, A. Boudria et al., “Abnormal expression speckle disassembly,” JournalofCellScience,vol.116,no.15,pp. of the pre-mRNA splicing regulators SRSF1, SRSF2, SRPK1 and 3099–3107, 2003. SRPK2 in non-small cell lung carcinoma,” PLoS ONE,vol.7,no. [91] J. Shi, T. Zhang, C. Zhou et al., “Increased dosage of Dyrk1A 10, Article ID e46539, 2012. alters alternative splicing factor (ASF)-regulated alternative [108] K. Odunsi, P. Mhawech-Fauceglia, C. Andrews et al., “Elevated splicingoftauindownsyndrome,”Journal of Biological Chem- expression of the serine-arginine protein kinase 1 gene in istry,vol.283,no.42,pp.28660–28669,2008. ovarian cancer and its role in Cisplatin cytotoxicity in vitro,” [92] S. Ding, J. Shi, W.Qian et al., “Regulation of alternative splicing PLoS ONE,vol.7,no.12,2012. of tau exon 10 by 9G8 and Dyrk1A,” Neurobiology of Aging, 2011. [109]L.Adesso,S.Calabretta,F.Barbagalloetal.,“Gemcitabine [93]W.Qian,H.Liang,J.Shietal.,“Regulationofthealternative triggers a pro-survival response in pancreatic cancer cells splicingoftauexon10bySC35andDyrk1A,”Nucleic Acids through activation of the MNK2/eIF4E pathway,” Oncogene, Research,vol.39,no.14,pp.6161–6171,2011. vol. 32, no. 23, pp. 2848–2857, 2013. [94] X. Yin, N. Jin, J. Gu et al., “Dual-specificity tyrosine [110] A. Locatelli and C. A. Lange, “Met receptors induce Sam68- phosphorylation-regulated kinase 1A (Dyrk1A) modulates dependent cell migration by activation of alternate extracellular serine/arginine-rich protein 55 (SRp55)-promoted Tau exon 10 signal-regulated kinase family members,” Journal of Biological inclusion,” TheJournalofBiologicalChemistry,vol.287,no.36, Chemistry,vol.286,no.24,pp.21062–21072,2011. pp. 30497–30506, 2012. [111] C. Ghigna, S. Giordano, H. Shen et al., “Cell motility is [95] Q. Tian, J.-L. Taupin, S. Elledge, M. Robertson, and P.Anderson, controlled by SF2/ASF through alternative splicing of the Ron “Fas-activated serine/threonine kinase (FAST) phosphorylates protooncogene,” Molecular Cell,vol.20,no.6,pp.881–890,2005. TIA-1 during Fas-mediated apoptosis,” JournalofExperimental [112] C. Valacca, S. Bonomi, E. Buratti et al., “Sam68 regulates Medicine,vol.182,no.3,pp.865–874,1995. EMT through alternative splicing-activated nonsense-mediated [96] J. M. Izquierdo and J. Valcarcel,´ “Fas-activated mRNA decay of the SF2/ASF proto-oncogene,” Journal of Cell Serine/Threonine Kinase (FAST K) synergizes with TIA- Biology,vol.191,no.1,pp.87–99,2010. 1/TIAR proteins to regulate Fas alternative splicing,” Journal of [113] S. Yea, G. Narla, X. Zhao et al., “Ras promotes growth by Biological Chemistry,vol.282,no.3,pp.1539–1543,2007. alternative splicing-mediated inactivation of the KLF6 tumor [97] M. Simarro, D. Mauger, K. Rhee et al., “Fas-activated ser- suppressor in hepatocellular carcinoma,” Gastroenterology,vol. ine/threonine phosphoprotein (FAST) is a regulator of alterna- 134, no. 5, pp. 1521–1531, 2008. tive splicing,” Proceedings of the National Academy of Sciences [114]J.C.Shultz,R.W.Goehe,D.S.Wijesingheetal.,“Alternative of the United States of America, vol. 104, no. 27, pp. 11370–11375, splicing of caspase 9 is modulated by the phosphoinositide 3- 2007. kinase/Akt pathway via phosphorylation of SRp30a,” Cancer [98] M. J. Moore, Q. Wang, C. J. Kennedy, and P. A. Silver, “An alter- Research,vol.70,no.22,pp.9185–9196,2010. native splicing network links cell-cycle control to apoptosis,” [115] L. Trinkle-Mulcahy, P. Ajuh, A. Prescott et al., “Nuclear organ- Cell,vol.142,no.4,pp.625–636,2010. isation of NIPP1, a regulatory subunit of protein phosphatase 1 [99] R. Reiter, P. Gais, U. Jutting et al., “Aurora kinase A messenger that associates with pre-mRNA splicing factors,” Journal of Cell RNA overexpression is correlated with tumor progression and Science,vol.112,no.2,pp.157–168,1999. International Journal of Cell Biology 15

[116] A. Boudrez, M. Beullens, P. Groenen et al., “NIPP1-mediated the anti-tumor drug spliceostatin A,” Genes and Development, interaction of protein phosphatase-1 with CDC5L, a regulator vol.25,no.5,pp.445–459,2011. of pre-mRNA splicing and mitotic entry,” Journal of Biological [132]R.Karni,E.DeStanchina,S.W.Lowe,R.Sinha,D.Mu,andA. Chemistry,vol.275,no.33,pp.25411–25417,2000. R. Krainer, “The gene encoding the splicing factor SF2/ASF isa [117] A. Boudrez, M. Beullens, E. Waelkens, W. Stalmans, and proto-oncogene,” Nature Structural and Molecular Biology,vol. M. Bollen, “Phosphorylation-dependent interaction between 14,no.3,pp.185–193,2007. the splicing factors SAP155 and NIPP1,” Journal of Biological [133] O. Fedorov, K. Huber, A. Eisenreich et al., “Specific CLK Chemistry,vol.277,no.35,pp.31834–31841,2002. inhibitors from a novel chemotype for regulation of alternative [118] M. V. Murray, R. Kobayashi, and A. R. Krainer, “The type 2C splicing,” Chemistry and Biology,vol.18,no.1,pp.67–76,2011. 𝛾 Ser/Thr phosphatase PP2C is a pre-mRNA splicing factor,” [134] W.-H.Chang,T.-C.Liu,W.-K.Yangetal.,“Amiloridemodulates Genes and Development,vol.13,no.1,pp.87–97,1999. alternative splicing in leukemic cells and resensitizes Bcr- [119] B. Cardinali, P. T. W. Cohen, and A. I. Lamond, “Protein 󸀠 AblT315I mutant cells to imatinib,” Cancer Research,vol.71,no. phosphatase 1 can modulate alternative 5 splice site selection in 2, pp. 383–392, 2011. a HeLa splicing extract,” FEBS Letters,vol.352,no.3,pp.276– [135] J.-G. Chang, D.-M. Yang, W.-H. Chang et al., “Small molecule 280, 1994. Amiloride modulates oncogenic RNA alternative splicing to [120] E. Allemand, M. L. Hastings, M. V. Murray, M. P. Myers, and devitalize human cancer cells,” PLoS ONE,vol.6,no.6,Article A. R. Krainer, “Alternative splicing regulation by interaction of ID e18643, 2011. 𝛾 phosphatase PP2C with nucleic acid-binding protein YB-1,” [136] S. K. Pal, K. Reckamp, H. Yu, and R. A. Figlin, “Akt inhibitors Nature Structural and Molecular Biology,vol.14,no.7,pp.630– in clinical development for the treatment of cancer,” Expert 638, 2007. Opinion on Investigational Drugs, vol. 19, no. 11, pp. 1355–1366, [121] T. Novoyatleva, B. Heinrich, Y. Tang et al., “Protein phosphatase 2010. 1 binds to the RNA recognition motif of several splicing fac- [137] L. Santarpia, S. M. Lippman, and A. K. El-Naggar, “Targeting the tors and regulates alternative pre-mRNA processing,” Human MAPKRASRAF signaling pathway in cancer therapy,” Expert Molecular Genetics,vol.17,no.1,pp.52–70,2008. Opinion on Therapeutic Targets,vol.16,no.1,pp.103–119,2012. [122]C.L.Lorson,E.Hahnen,E.J.Androphy,andB.Wirth,“A [138] S. Zhang and D. Yu, “Targeting Src family kinases in anti- single nucleotide in the SMN gene regulates splicing and is cancer therapies: turning promise into triumph,” Trends in responsible for spinal muscular atrophy,” Proceedings of the Pharmacological Sciences,vol.33,no.3,pp.122–128,2012. National Academy of Sciences of the United States of America, vol. 96, no. 11, pp. 6307–6311, 1999. [123] Z. Zhang, O. Kelemen, M. A. Van Santen et al., “Synthesis and characterization of pseudocantharidins, novel phosphatase modulators that promote the inclusion of exon 7 into the SMN (survival of motoneuron) pre-mRNA,” Journal of Biological Chemistry, vol. 286, no. 12, pp. 10126–10136, 2011. [124] C. E. Chalfant, K. Rathman, R. L. Pinkerman et al., “De novo ceramide regulates the alternative splicing of caspase 9 and Bcl- x in A549 lung adenocarcinoma cells. Dependence on protein phosphatase-1,” Journal of Biological Chemistry,vol.277,no.15, pp. 12587–12595, 2002. [125] C. Sumanasekera, O. Kelemen, M. Beullens et al., “C6 pyridinium ceramide influences alternative pre-mRNA splicing by inhibiting protein phosphatase-1,” Nucleic Acids Research,vol. 40, no. 9, pp. 4025–4039, 2012. [126] J. P. Venables, R. Klinck, A. Bramard et al., “Identification of alternative splicing markers for breast cancer,” Cancer Research, vol.68,no.22,pp.9525–9531,2008. [127]K.Lenos,A.M.Grawenda,K.Lodderetal.,“Alternatesplic- ing of the p53 inhibitor HDMX offers a superior prognostic biomarker than p53 mutation in human cancer,” Cancer Research, vol. 72, no. 16, pp. 4074–4084, 2012. [128] R. Kole, A. R. Krainer, and S. Altman, “RNA therapeutics: beyond RNA interference and antisense oligonucleotides,” Nature Reviews Drug Discovery,vol.11,no.2,pp.125–140,2012. [129] S. Bonnal, L. Vigevani, and J. Valcarcel,´ “The spliceosome as a target of novel antitumour drugs,” Nature Reviews. Drug Discovery, vol. 11, no. 11, pp. 847–859, 2012. [130] F. Zammarchi, E. De Stanchina, E. Bournazou et al., “Antitu- morigenic potential of STAT3 alternative splicing modulation,” Proceedings of the National Academy of Sciences of the United States of America,vol.108,no.43,pp.17779–17784,2011. [131] A. Corrionero, B. Minana,˜ and J. Valcarcel,´ “Reduced fidelity of branch point recognition and alternative splicing induced by Hindawi Publishing Corporation International Journal of Cell Biology Volume 2013, Article ID 636050, 10 pages http://dx.doi.org/10.1155/2013/636050

Review Article Alternative Splicing Regulation of Cancer-Related Pathways in Caenorhabditis elegans:AnInVivoModel System with a Powerful Reverse Genetics Toolbox

Sergio Barberán-Soler1 and James Matthew Ragle2

1 SomaGenics, Inc., Santa Cruz, CA 95060, USA 2 Department of Molecular, Cell and Developmental Biology, The Center for Molecular Biology of RNA, UniversityofCalifornia,SantaCruz,SantaCruz,CA95064,USA

Correspondence should be addressed to Sergio Barberan-Soler;´ [email protected]

Received 17 June 2013; Accepted 29 July 2013

Academic Editor: Claudio Sette

Copyright © 2013 S. Barberan-Soler´ and J. M. Ragle. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Alternative splicing allows for the generation of protein diversity and fine-tunes gene expression. Several model systems have been used for the in vivo study of alternative splicing. Here we review the use of the nematode Caenorhabditis elegans to study splicing regulation in vivo. Recent studies have shown that close to 25% of genes in the worm genome undergo alternative splicing. A big proportion of these events are functional, conserved, and under strict regulation either across development or other conditions. Several techniques like genome-wide RNAi screens and bichromatic reporters are available for the study of alternative splicing in worms. In this review, we focus, first, on the main studies that have been performed to dissect alternative splicing in this system and later on examples from genes that have human homologs that are implicated in cancer. The significant advancement towards understanding the regulation of alternative splicing and cancer that the C. elegans system has offered is discussed.

1. Introduction pathwaysofthisspermtooocytetransitionhavebeenstudied extensively [2]. Its complete genome, sequenced in 1998, was Since the early 1960s with the efforts of Sydney Brenner, the the first sequenced genome from a multicellular organism. It nematode Caenorhabditis elegans has been established as a has a genome size of 97 megabases containing close to 19,000 popular model organism in developmental biology and neu- protein coding genes [3]. Genome-wide alignments with robiology. There are many biological advantages that make it other five related nematodes are now available for any com- an attractive system for several fields of research. The adult parative genomics approach [4]. The modENCODE project worm contains 959 somatic cells, making its anatomy rela- systematically generated genome-wide data from transcrip- tively simple. Experiments in the late 1970s showed that it has tome profiling, transcription factor-binging sites, and maps an invariant cell lineage during establishment of the somatic ofchromatinorganizationtoimprovegenomeannotation[5]. tissues [1]. It has two sexes, a self-fertilizing hermaphrodite When comparing the C. elegans genome to higher eukaryotes, and males, allowing genetic crosses to be performed. C. it was found that close to 40% of the genes that have been elegans has a short life cycle of less than three days and associated with diseases in humans have worm orthologues, each hermaphrodite produces about 300 progeny by self- and cancer is not the exception [6]. fertilization or up to 1000 progeny from cross progeny with males. The gonad is a relatively large organ in this animal, 2. Alternative Splicing Prevalence allowing for studies of organogenesis, cell proliferation, meiosis, and embryogenesis. During its development, the her- In 1990, the first report of an alternative splicing event in maphrodite worms produce sperm at one stage of their life the C. elegans genome was published. Interestingly the event cycle before switching to produce oocytes. The molecular corresponds to the PKA mRNA, a kinase implicated in the 2 International Journal of Cell Biology

C. elegans 40% of genes with alternative genome 19,000 genes splicing are regulated during development 25% of genes are alternatively spliced 35% of genes with alternative splicing are putative targets of NMD

(a) Human APAF1, median intron size 2746 bp, median exon size 132 bp

Scale 20 kb hg19 chr12: 99,045,000 99,050,000 99,055,000 99,060,000 99,065,000 99,070,000 99,075,000 99,080,000 99,085,000 99,090,000 99,095,000 99,100,000 99,105,000 99,110,000 99,115,000 99,120,000 99,125,000 APAF1 RefSeq genes APAF1 APAF1 APAF1 APAF1

C. elegans ced-4, median intron size 50 bp, median exon size 181 bp 1 kb ce6 Scale chrIII: 4,852,500 4,853,000 4,853,500 4,854,000 4,854,500 RefSeq genes ced-4 ced-4 (b) C. elegans Y48C3A.5 10 kb Scale ce6 chrII: 13,275,000 13,280,000 13,285,000 13,290,000 13,295,000 13,300,000 13,305,000 RefSeq genes Y48C3A.5

Intron 4 size 19,927 bp (c) C. elegans unc-52 generates 12 isoforms by alternative splicing

Scale 5 kb ce6 chrII: 14,653,000 14,654,000 14,655,000 14,656,000 14,657,000 14,658,000 14,659,000 14,660,00014,661,000 14,662,000 14,663,000 14,664,000 14,665,000 14,666,000 RefSeq genes unc-52

unc-52 (d)

Figure 1: Caenorhabditis elegans alternative splicing events. (a) Genome-wide analysis of alternative splicing in C. elegans;(b)comparisonof the human APAF1 and the C. elegans homolog ced-4 gene models shows significant differences in intron size between species for genes with important alternative splicing events; (c) Y48C3A.5 intron 4 (19,927 bp) is one example of the 144 introns in the C. elegans genome that are more than 10 kb in length; (d) C. elegans unc-52 gene undergoes complex alternative splicing that generates at least 12 different isoforms by the use of nine different cassette exons.

onset and progression of several cancers [7]. Since then, C. elegans is not an uncommon mechanism to generate pro- several groups have used different approaches to predict the tein diversity. percentage of genes in the C. elegans genome that undergo alternative splicing. Initial estimates using a limited number 3. Effects of Genome Compaction on AS of Expressed Sequence Tags (ESTs) predicted fewer than 1000genestobealternativelyspliced[8]. By normalizing the The worm genome appears to be under selective pressure to occurrence of alternative splicing, taking into account the promoteareductioningenomesize[11]. This natural selec- coverageofESTs,itwasestimatedthatcloseto10%ofthe tion towards a small genome can be seen in features like short C. elegans genome undergoes alternative splicing [9]. More intergenic regions, short UTRs, and small introns [8, 12, 13]. recent analysis using next-generation sequencing identified For example, in humans, the median size of introns in the 8,651 putative novel splicing events, suggesting that up to coding sequences is 1,334 bases, while in worms the average 25% of genes have an alternative splicing event (Figure 1(a)) intron is just 65 bases (Figure 1(b))[4, 14]. More than 20 years [10].Whilethispercentageisfarfromthe>90% reported for ago, the small size of C. elegans introns was already the subject the human genome, it does show that alternative splicing in of study. It was demonstrated that a short 53 nucleotide worm International Journal of Cell Biology 3 intron could not be efficiently spliced in human extracts, demonstrated that up to 40% of the events detected are regu- while an expansion of this intron with 31 extra nucleotides lated (>2 fold) between different stages of worms (Figure 1(a)) allowed for efficient splicing [15]. In spite of this reduction in [23]. This result was further validated with tiling arrays and intron size, the worm spliceosome is still capable of removing next-generation sequencing [24]. Several examples of tissue- big introns (144 introns in the C. elegans genome are bigger specific splicing are known. The alternative splicing regula- than 10 kb (Figure 1(c))[4]). Larger intron size has been cor- tion of a neuron-specific exon of unc-32,theworma subunit + related with alternative splicing in several species [16]. Worms of V0 complex of vacuolar-type H -ATPases, was recently also have complex patterns of alternative splicing where characterized [25]. A male-specific isoform of unc-55,atrans- multiple exons in the same gene are alternatively spliced to cription factor, has also been reported [26]. While individual generate multiple isoforms (Figure 1(d)). This means that the examples of tissue-specific splicing are known, a genome- information content of a worm intron is on average greater wide analysis of tissue-specific splicing is still missing. With than in higher eukaryotes (higher density of functional the increased sensitivity of next-generation sequencing tech- elements in introns). This makes the molecular dissection niques, together with the availability of manual or molecular of worm introns easier to achieve. By using genomic align- dissections that allow the isolation of tissue-specific mRNA ments between two Caenorhabditis species, the identification (mRNA-tagging [27]) a complete catalog of tissue-specific of novel intronic elements important for alternative splicing splicing should be possible in the near future. regulation has been described [17]. Other groups have also While a big set of splicing events produce two protein used comparative genomics together with UV cross-linking isoforms, another important group introduces premature and Electrophoretic Mobility Shift Assays (EMSA) to identify termination codons (PTC) to one of the isoforms. It is known cis-elements important for alternative splicing regulation that the introduction of a PTC to one isoform targets it to [18]. nonsense-mediated decay (NMD) (reviewed in [28]). NMD mutants were first described in C. elegans almost 25 years ago [29]. Contrary to other systems where mutations in NMD 4. Evidence That AS in Nematodes Is under factors are lethal, in worms, null mutations for all the seven Stabilizing Selection core factors of NMD are viable (smg-1 to smg-7). This has facilitated the study of AS coupled to NMD in worms. Some Several studies comparing alternative splicing events in mam- of the first events of wild-type transcripts that were shown malsandinsectsconcludedthatahighpercentageofthe to be regulated by NMD are splicing factors in the C. elegans events are not conserved and are species specific [19, 20]. genome [30]. An important discovery concerning the regula- This high variability in alternative splicing makes it necessary tion of alternative splicing in worms is that between 20–35% to validate the functionality of the events studied. To test oftheeventsinthegenomeappeartobetargetsofNMD(Fig- whether the smaller percentage of genes with alternative ure 1(a))[24, 31]. The conclusion from these studies is, then, splicing in worms also follows these patterns of high variabil- that alternative splicing in C. elegans is not just a generator ity during evolution, several groups have measured the levels of protein diversity but also an important regulator that fine- of conservation between different C. elegans populations or tunes gene expression levels by targeting specific isoforms for between related species [21, 22]. In comparison to the findings degradation. Furthermore, it has been proposed that NMD in higher eukaryotes, the regulation of alternative splicing in in worms can also be regulated with the potential to stabilize natural populations of nematodes appears to be under strong particular NMD targets allowing them to be translated into stabilizing selection with low intra- an interspecies variability. truncatedproteinswithputativedominantnegativefunctions These results point to an essential intrinsic characteristic of [31]. alternative splicing in worms: its functionality. The detection of alternative splicing in a C. elegans transcript has a higher 6. Powerful Reverse Genetics to Dissect probability of being a functional and regulated event than in other systems. Alternative Splicing One of the advantages of C. elegans as a model system is 5. Alternative Splicing Regulation the availability of powerful tools for reverse genetics. Any laboratory can obtain stable mutants for many genes from the A proxy for the functionality of an alternative splicing event Caenorhabditis Genetics Center (CGC) at The University of that has been used for other systems is its regulation. If a Minnesota. The National Bioresource Project in Japan runs a particular event is detected as regulated across different con- program where mutants for a gene of interest can be requested ditions or during development then the possibility that the and they are obtained at the facility by a protocol involving isoforms have specific functions is higher. In the last five years random mutagenesis with TMP/UV. These two centers allow C. elegans joined other species in terms of the detection of for any group to obtain mutant strains for the gene of interest changes in alternative splicing at genome-wide levels. Initially in an inexpensive and expedited manner. Recently, a more with the use of splicing-sensitive microarrays and later with ambitious project to create a million different mutants has the use of next-generation sequencing the regulation of been performed by the Moerman and Waterston labs [32]. alternative splicing in worms has been studied in detail The aim of this project was to use next-generation sequencing across different conditions and mutations [10, 23, 24]. Initial and a collection of 2000 mutagenized strains to identify mul- measurement of changes in splicing during development tiple mutations in virtually all the genes in the worm genome. 4 International Journal of Cell Biology

Ras pathway

Retinoblastoma Proliferation enhanced Proliferation pathway versus by splicing factors regulation

differentiation Isoform-specific Synmuv splicing factor interactions

Alternative Functional isoform splicing diversity

SR kinase activity? Opposingfunctions isoform Protein Programmed kinase A cell death expression

Differential isoform

Topoisomerase-1

Figure 2: Connections between alternative splicing in C. elegans and pathways homologous to those that cause excessive cell proliferation or apoptosis in humans, as reviewed in this paper.

This project will likely allow the study of mutant worms alternative splicing event can be performed in vivo. Different that have mutations in isoform-specific regions. All these events that have been studied with this technology are egl- resourcesallowfortheuseofstablemutantstocharacterize 15, let-2, unc-60,andunc-32 [25, 37–40]. This technology the roles of either putative splicing factors or more specifically has allowed the identification of splicing factors that regulate of particular isoforms. This approach to screen the effects that these events, as well as the cis-elements that are necessary for mutations in different splicing factors have on a particular its regulation. splicing event has been used before [33]. The conclusion from this work was that the coregulation of alternative splicing by 8. C. elegans AS Regulation of diverse factors is a common phenomenon in worms. Cancer-Related Pathways Another great resource for the worm community is the availability of genome-wide RNAi libraries [34]. Researchers Given the diversity of tools available to perform studies in C. aiming to characterize genetic pathways have used these elegans, researchers have set out to understand details about libraries for genome-wide screens. Some of them have found cancer that were long a mystery. Questions surrounding that several splicing-related components have interesting cell proliferation and cell-cell communication as well as the phenotypes when targeted by RNAi [35]. More focused cellular and molecular components that make up a stem cell screens with a subset of clones from these libraries are used nicheoftenmustbeansweredinamulticellularcontext.Fur- to characterize a specific group of genes. For example, Kerins thermore, the fact that biological processes and factors active etal.usedanRNAiscreenofallthepredictedC. elegans in splicing and human cancers are almost wholly homologous splicing factors and found that many of them showed an over- with those in C. elegans makes the move to worm studies proliferation phenotype in a sensitized germline background rewarding. Disrupted cancer pathways in C. elegans,forthe [36]. most part, have clear, simple, and observable phenotypes that extend beyond just cell death or proliferation. Using these 7. In Vivo Alternative Splicing Reporters pathways as tools and read-outs, splicing factors and splicing events have been identified, in C. elegans,tointeractwith, Oneofthemostadvantageoustoolsforthestudyofalterna- regulate, and cooperate with homologous pathways that lead tive splicing in C. elegans is the use of bichromatic alternative to cancer in humans (Figure 2). splicing reporters [37].Thistransgenicreportersystemallows the visualization of splicing events by tagging each one of 8.1. Extensive Splicing of Ras/let-60 Pathway Components Con- the isoforms with a different fluorescent protein. A worm tributes to Cell Fate Determination and Proliferation. Over- population with differences in splicing regulation can then whelming evidence links cell signaling and gene expression be sorted using FACS cytometry and worms with a partic- changes in the promotion of human cancers. Alternative ular splicing ratio obtained. This together with the use of splicing plays a major role in defining the activity of signal mutagenesisorRNAiallowsperforminganinvivoscreen transduction pathways such as Ras/let-60.Strikingly,ofthe for regulators of any splicing event of interest. The advantage 50–60 genes that are known components of, regulate or, are of these reporters in worms is that the characterization of an directly regulated by the Ras/let-60 pathway in C. elegans International Journal of Cell Biology 5

Table1:Raspathwaycomponents,regulators,interactors,and while the activity of LIN-3S is not [46]. LIN-3S is expressed targets in C. elegans.Thosemarkedwitha√ produce 2 or more in a specialized Anchor Cell (AC) responsible for inducing a isoforms differing by at least 1 alternative exon (according to[4, 39]). primary set of VPCs during larval development, and LIN- 3L expression and possible secretion by primary VPCs Ras pathway component/interactor themselves act as a ligand stimulant of the Ras pathway in ark-1 lin-10 secondary VPCs [42, 43]. This stimulation may cause the √ cdf-1 lin-25 secondary VPCs to switch their cell fate from hypodermis cnk-1 lin-31 to vulva [43]. It is unknown if coordinated expression of dab-1 √ lin-39 √ these isoforms from distinct cell types establishes a gradient dpy-22 lin-45 √ through Ras signaling that induces nearby VPCs to proceed dpy-23 √ lip-1 through necessary cell divisions to develop into a proper egl-5 √ lrp-1 vulva. egl-18 lst-1 √ RasGAPs are GTPase-activating proteins that specifi- egl-15 √ lst-2 cally promote the hydrolysis of Ras-bound GTP molecules, thereby inactivating the Ras molecule and its signaling [47]. egl-17 lst-3 √ √ √ gap-2,inC. elegans, is similar to the p120 Ras-GAP family egl-19 lst-4 [48]. The gene contains 25 exons that use alternative splicing √ egl-30 mek-2 and transcription start sites to produce 9 mRNA products √ elt-6 mpk-1/sur-1 [49]. These isoforms were identified by a rapid amplification 󸀠 eor-1 par-1 √ of cDNA end (5 RACE) technique using primers specific to eor-2 √ ptp-2 a sequence common to many mature mRNAs in C. elegans, gap-1 rom-1 the SL1 trans-splice leader. This 22nt sequence is naturally 󸀠 √ spliced onto 5 ends of transcripts from ∼70% of genes gap-2 sem-4 󸀠 gpa-5 sem-5 [50]. In this case, they served as 5 primer binding sites in reverse transcription reactions followed by PCR to identify ksr-1 sli-1 √ 󸀠 ksr-2 √ soc-1 isoforms with alternative 5 ends and promoters [46]. Dif- ferential expression of gap-2 isoforms in several tissue types let-23 √ sos-1/let-341 √ was revealed by expression of transgenes containing these let-60 sra-13 alternative gap-2 promoters fused to GFP in C. elegans. The let-92 sur-2 function of RasGAP proteins from various genes in other let-756 sur-5 organisms [51] may be relegated to multiple isoforms from lin-1 sur-6 a small number of RasGAP genes producing many isoforms lin-2 √ sur-7 √ in C. elegans. While the effects of this dynamic expression lin-3 √ sur-8/soc-2 √ patternofisoformsonmolecularsignalingpathwayssuchas lin-7 unc-101 √ the Ras pathway are difficult to tease apart, target sites for PKC, PKA, and protein tyrosine kinases present or absent in each isoform [49] may provide clues to individual isoform activities. [41], approximately half of them produce multiple isoforms differing by at least one alternative exon4 [ ]. This highlights the importance of proper gene expression through post- 8.2. Redundant Retinoblastoma/lin-35 Pathways Are Popu- transcriptional regulation (Table 1). latedbySplicingFactors.Homozygous mutations of the The ligand-dependent Ras pathway in C. elegans stimu- retinoblastoma (Rb) gene have been shown to promote retinal lates the induction of vulval precursor cells (VPCs) in the cancer, small-cell lung carcinomas, and osteosarcomas [52]. hypodermis to divide and differentiate into a functional vulva In worms, a single homolog of Rb, lin-35, represses the multi- during larval development [42, 43]. Increased signaling from vulval phenotype observed in Ras pathway hyper-signaling the Ras pathway leads to a multivulval (Muv) phenotype, mutants [53]. Two redundant pathways, termed synthetic while decreased signaling causes a vulvaless (Vul) phenotype multivulval (synmuv) class A and class B, are active in the [44]. These simple and observable phenotypes provide an VPCs and induce changes in expression of genes that lead to a excellent system to ascertain the contributions of alternative multivulval fate through histone modification and chromatin splicing to Ras signaling activity. The C. elegans homolog of remodeling [53, 54]. To induce a phenotype, disruptive the Ras-MAPK pathway stimulant Epidermal Growth Factor, mutations in single genes within each class must be present. lin-3, produces several isoforms with unique characteristics. Interestingly, mutations in the lin-15 locus created alleles LIN-3 isoform specialization of function has previously been with Muv, class A synmuv, and class B synmuv phenotypes, observed in pathways aside from Ras-mediated cell signaling leading to speculation that two genes existed at the locus with and has been found to differentially mediate growth rates, distinct functions in each pathway. In RNAi screens for feeding behavior, and cellular quiescence [45]. In VPCs, the interactors of the class B synmuv pathway in C. elegans,57 activity of the LIN-3L isoform is dependent on interaction genes were found to enhance the Muv phenotype observed with the C. elegans Rhomboid protease homolog, ROM-1, in lin-35 or other synmuv class B mutant lines [35, 55]. Ten 6 International Journal of Cell Biology of these potential interactors (rsr-2, lsm-2/gut-2, lsm-4, snr-1– localization differences between the isoforms lead to spec- 7) have roles in splicing. Their synmuv B mutant phenotypes ulation they may function in different processes. The TOP- are dependent on Ras activity and lin-3 ligand binding. While 1𝛽 isoform skips exon 2 and is more ubiquitously expressed the activities, interactions, and mutant phenotypes of most of throughout C. elegans development, being found in multiple these splicing-related Rb enhancers are not well understood, cell types and all stages from embryo through adulthood. RNAi of some of the snr genes (snr-1, snr-2, snr-4, snr-5, snr- Conversely, the inclusion isoform, TOP-1𝛼, is detectably 6,andsnr-7) led to embryonic lethality and nuclear pore present in embryos and, interestingly, in neurons at the organization disruption [56]. The question of whether these comma stage, then decreases in abundance as worms enter phenotypes arise from disrupted functions that have no role larval stages [63]. Isoform-specific immuno-histochemical in splicing regulation is still unanswered, but evidence is localization assays in germline cells identified TOP-1𝛽 con- mounting supporting the idea that constitutive splicing and centration in nucleoli and TOP-1𝛼 concentration at centro- alternative splicing are major contributors in C. elegans to somes and on chromosomes [64, 65]. Expression of a GFP pathways that, when altered, lead to cancer in human tissues. transgene driven by the topoisomerase promoter was also Undoubtedly, future studies in C. elegans will enrich current detected strongly in the distal tip cell of L3/L4 worms indicat- knowledge concerning the interwoven activities of splicing ing that somatic gonad sheath development and/or germline and the Rb/lin-35 pathway. stem cell proliferation may be regulated by TOP-1 activity. RNAi of C. elegans top-1 reduced the number of germ cells 8.3. Protein Kinase A Isoform Diversity Replaces the Need for inthematuregermlinebyanywherefrom50–100%[65]. Multiple Genetic Loci. ProteinkinaseA(PK-A)isinvolved It is currently unknown how top-1 isoforms contribute to inmanycellularprocessesandhasbeenimplicatedinseveral the maintenance of stem cell proliferation or exactly which cancers [57]. The mammalian PK-A is composed of two isoform would be performing such functions. Localization of distinct subunits, the regulatory and catalytic, that are each the inclusion isoform, TOP-1𝛼, to centrosomes and chromo- interchangeable with protein subunits from several genetic somes suggests it could be involved in chromosome segre- loci. The two subunits of PK-A are conserved in C. elegans gation, in regulation of transcription, or possibly in posttran- [7] but derived from possibly only three genes, kin-1, kin- scriptional regulation. DNA topoisomerase I was identified as 2, and F47F2.1. Interestingly, the modularity of the subunits an SR protein kinase in HeLa cell extracts [66]. It was found that make up PK-A in C. elegans has been conserved through to phosphorylate splicing factors involved in cell cycle regu- alternative usage of exons. The catalytic subunit (C-subunit) lation, such as SF2/ASF [67, 68].Itmaybepossiblethatthese alone is thought to undergo alternative splicing at both N- isoforms are working more closely than originally thought if and C-termini to create at least 12 different isoforms [58]. one or both of the top-1 isoforms affect the cell cycle and carry These isoforms are conserved in C. briggsae,arelativeof thesameSRkinaseactivityasshowninHeLacells.Theactiv- C. elegans, and show differential expression during develop- ity of one or both of the TOP-1 isoforms as SR protein kinases ment,suggestingthatthediversityofpotentialsubunitsitself suggests that top-1 alternative splicing may be autoregulated. has functional importance [59]. An isoform of the C-subunit Kinase activity of TOP-1 and its autoregulation have yet to be 󸀠 containing the N 1 exon harbors a myristoylation site and determined in worms. 󸀠 is highly expressed in eggs, where an alternative exon N 4 lacking the myristoylation site is highly expressed in adult 8.5.SplicingFactorsasRegulatorsofCellProliferationand 󸀠 worms [59]. The presence of N 1 protein isoform expression Differentiation. Several genes involved in splicing regulation and myristoylation was found to affect substrate targeting have been implicated in the proliferation/differentiation deci- ofPK-Abutnotthecatalyticactivityoftheenzymeitself sion in the C. elegans germline. glp-1(ar202gf) worms were [60]. Similarly, isoforms of the regulatory (R-subunit) have used to perform an RNAi screen searching for genes that led recently been identified and found to contain or lack domains to increased cell proliferation in the germline. The resulting necessary for docking to A-kinase anchor proteins and the C- genes included factors involved in every step of the splicing subunit of the PK-A holoenzyme [61]. A second gene on the X process (spliceosome construction initiation, reorganization chromosome of C. elegans, F47F2.1, has homology to murine of the snRNP complex and removal of the lariat; [36]). prp- 󸀠 C𝛼 subunits [58]. Like the kin-1 N 4 isoform, expression of 17, for example, encodes an ortholog of yeast CDC40 and thelongerF47F2.1bisoformislowineggsbuthighin humanprp17andisinvolvedinthe2ndcatalyticstepof mixed populations of worms [59]. A truncated isoform, intron removal in the spliceosome [36, 69]. RNAi of prp- F47F2.1a, lacks amino acids important for ATP-binding, and 17 in the rrf-1(pk1417);glp-1(oz264) background enhanced 󸀠 knockdown of N 4 kin-1 isoform has no obvious phenotype germline tumors, suggesting that it is involved in the prolifer- leading to speculation that some of these isoforms may be ation versus differentiation decision. Similarly, the mog genes redundant or unnecessary [59, 62]. represent C. elegans homologs of core yeast splicing factors that regulate the pathway downstream of glp-1 to promote 8.4. Topoisomerase-1 Isoforms Have Differential Expression differentiation and the oocyte fate of maturing germ cells and Potential SR Kinase Activity. Isoforms from the same [70–72]. The U2 snRNP-associated splicing complex SF3b is genetic locus often have distinct expression patterns and essential for splicing [73, 74]. TEG-4 is the worm homolog of functions that are seemingly unrelated. C.elegans DNA topoi- human SF3b subunit 3 (aka SAP130) that increases excessive somerase I (top-1) produces two isoforms that vary in tempo- cell proliferation in the glp-1(ar202gf) background [75]. ral and spatial expression patterns. Cellular and subcellular Interestingly, epistasis experiments attempting to determine International Journal of Cell Biology 7 the genetic relationship between teg-4 and the glp-1/Notch 8.7. Research Focusing on the Links between Splicing and signaling pathway in various cell types were inconclusive. Cancer in C. elegans Is Bright. Great strides have been taken Human CD2BP2 was suggested to regulate splicing via in research to understand the involvement of splicing factors U4 U5 U6 tri-snRNP formation [76]. TEG-1, the C. elegans in biological processes such as cell proliferation, differenti- homologofCD2BP2hasbeenfoundtobindUAF-1,theworm ation, migration, communication, and death (as discussed U2AF large subunit homolog, suggesting that it could work in previoussly). RNAi screens have identified splicing factors two important steps in the mechanism of pre-mRNA splicing that promote or inhibit cell proliferation. Transgenic assays [77]. teg-1(oz230), in combination with a glp-1 mild gain-of- have revealed cellular and subcellular localization of specific function mutant background, produces a tumorous germline isoforms that may be involved in apoptotic or cell cycle phenotype as well. To what extent splicing factors, them- regulation. Transparent cuticles and eggs make worms ideal selves, have individual genetic interactions within the prolif- specimens for the use of fluorescent reporters. Lineage trac- eration versus differentiation pathways or if global splicing ing experiments have not only identified every cell in an adult regulation is more indirectly influential is still unknown. worm but also precursors and cell behavior throughout development. Furthermore, only in a multicellular organism, such as C. elegans, can the effects of cell-to-cell signaling on 8.6. CED-4 Isoforms Promote and Inhibit Systematic Apoptosis processes such as differentiation or cell proliferation become during Development. Disruption of alternative splicing pat- apparent. High-throughput sequencing and bioinformatic terns may affect the onset of programmed cell death so intri- analyses are opening a new chapter on comparisons in iso- cately regulated in C. elegans. The worm homolog of APAF1, form usage between genetic backgrounds. The modEncode ced-4, was first identified as a core factor in the developmental project seeks to identify functional elements within the induction of neuronal programmed cell death [78]. ced-4 Drosophila and Caenorhabditis genomes by providing access physically links ced-3, a member of a caspase family of pro- to high-quality gene expression datasets. Furthermore, the teases to ced-9, a cell death suppressor [79–81]. As a regulatory million-mutation project in C. elegans will allow researchers switch, ced-4 is a core determinant of the decision of a cell to obtain gene and even isoform-specific mutant strains for to be systematically culled or not. ced-4 encodes two protein further study. Given the speed and depth of discoveries using isoforms, CED-4L and CED-4S, differing by 24 amino acids methodssuchasRNAiinworms,thisprojectshouldmakeC. 󸀠 at the 5 end of exon 4. While CED-4S normally pro- elegans a core model to study alternative splicing and its link motes programmed cell death, overexpression of the CED-4L to genes homologous to those tied to cancer in humans. isoform led to ectopic cell growth and rescue of lethal phenotypes seen in ced-9 loss-of-function mutants [82, 83]. It Acknowledgments is CED-4L association with CED-3 that surprisingly inhibits cell death in a dominant negative manner. The presence of the James Matthew Ragle is funded by a Grant from the National longer P-loop in the long isoform still allows for association Science Foundation to Alan Zahler, University of California, of CED-4L with the protease domain of CED-3 but disrupts Santa Cruz (MCB1121290). Funding for open access is pro- the association of CED-4L with the prodomain of CED-3, vided by the UC Santa Cruz Open Access Fund. which normally contributes to its activation [84]. Studies in mammalian cells have similarly identified ced-3 and References ced-4 homologous genes that produce isoforms with oppos- ingcelldeathfunctions[85, 86]. An imbalance in the [1] J. E. Sulston and H. R. Horvitz, “Post embryonic cell lineages of concentrationsoftheseisoformsorintheabundanceor the nematode. Caenorhabditis elegans,” Developmental Biology, vol.56,no.1,pp.110–156,1977. activities of specific SR or hnRNP proteins often leads to one developmental cell fate over the other [87]. Complex reg- [2]S.Zanetti,S.Grinschgl,M.Meolaetal.,“Thesperm-oocyte switch in the C. elegans hermaphrodite is controlled through ulation of alternative splicing factors, therefore, may direct steady-state levels of the fem-3 mRNA,” RNA,vol.18,no.7,pp. developmental apoptosis programs. Loss-of-function of an 1385–1394, 2012. SR protein kinase, spk-1,inaced-4 partial loss-of-function [3] C elegans Sequencing Consortium, “Genome sequence of the background, as well as mutant alleles of several SR proteins nematode C. elegans: a platform for investigating biology,” themselves, leads to an increase in apoptosis [88], suggesting Science, vol. 282, no. 5396, pp. 2012–2018, 1998. that control of alternative splicing factor activity may play a [4]K.Yook,T.W.Harris,T.Bierietal.,“WormBase2012:more direct or indirect role in the regulation of developmentally genomes, more data, new website,” Nucleic Acids Research,vol. programmed apoptosis. In vitro studies have shown that SPK- 40, pp. D735–D741, 2012. 1 can bind and phosphorylate alternative splicing factors like [5]M.B.Gerstein,Z.J.Lu,E.L.VanNostrandetal.,“Integrative SF2 and RNAi-mediated knockdown of spk-1 in C. elegans analysis of the Caenorhabditis elegans genome by the modEN- causes embryonic lethality and germline development defects CODE Project,” Science,vol.330,no.6012,pp.1775–1787,2011. [89, 90]. Experiments aimed at understanding the complex [6] E.CulettoandD.B.Sattelle,“AroleforCaenorhabditis elegans in code that dictates what genetic material is included or understanding the function and interactions of human disease skipped in mature mRNA and subsequent protein products genes,” Human Molecular Genetics,vol.9,no.6,pp.869–877, will undoubtedly clarify further the impact that alternative 2000. splicing has in the mediation of programmed cell death [7]R.E.Gross,S.Bagchi,X.Lu,andC.S.Rubin,“Cloning,charac- through factors such as ced-4. terization, and expression of the gene for the catalytic subunit 8 International Journal of Cell Biology

of cAMP-dependent protein kinase in Caenorhabditis elegans. [24]A.K.Ramani,A.C.Nelson,P.Kapranov,I.Bell,T.R.Gingeras, Identification of highly conserved and unique isoforms gener- and A. G. Fraser, “High resolution transcriptome maps for wild- ated by alternative splicing,” Journal of Biological Chemistry,vol. type and nonsense-mediated decay-defective Caenorhabditis 265, no. 12, pp. 6896–6907, 1990. elegans,” Genome Biology,vol.10,no.9,articleR101,2009. [8] W. J. Kent and A. M. Zahler, “The intronerator: exploring [25] H. Kuroyanagi, Y. Watanabe, and M. Hagiwara, “CELF family introns and alternative splicing in Caenorhabditis elegans,” RNA-binding protein UNC-75 regulates two sets of mutually Nucleic Acids Research,vol.28,no.1,pp.91–93,2000. exclusive exons of the unc-32 gene in neuron-specific manners [9]D.Brett,H.Pospisil,J.Valcarcel,´ J. Reich, and P. Bork, “Alterna- in Caenorhabditis elegans,” PLoS Genetics,vol.9,no.2,Article tive splicing and genome complexity,” Nature Genetics,vol.30, ID e1003337, 2013. no. 1, pp. 29–30, 2002. [26] G. Shan and W. W. Walthall, “Copulation in C. elegans males [10]A.K.Ramani,J.A.Calarco,Q.Panetal.,“Genome-wideanal- requires a nuclear hormone receptor,” Developmental Biology, ysis of alternative splicing in Caenorhabditis elegans,” Genome vol. 322, no. 1, pp. 11–20, 2008. Research, vol. 21, no. 2, pp. 342–348, 2011. [27] F. Pauli, Y. Liu, Y. A. Kim, P.-J. Chen, and S. K. Kim, “Chro- [11] D. R. Denver, K. Morris, M. Lynch, and W. K. Thomas, “High mosomal clustering and GATA transcriptional regulation of mutation rate and predominance of insertions in the Caeno- intestine-expressed genes in C. elegans,” Development,vol.133, habditis elegans nuclear genome,” Nature,vol.430,no.7000,pp. no. 2, pp. 287–295, 2006. 679–682, 2004. [28] S. Kervestin and A. Jacobson, “NMD: a multifaceted response to [12] C. H. Jan, R. C. Friedman, J. G. Ruby, and D. P. Bartel, premature translational termination,” Nature Reviews Molecular “Formation, regulation and evolution of Caenorhabditis elegans Cell Biology,vol.13,no.11,pp.700–712,2012. 3’UTRs,” Nature,vol.469,no.7328,pp.97–103,2011. [29] J. Hodgkin, A. Papp, R. Pulak, V. Ambros, and P. Anderson, “A new kind of informational suppression in the nematode [13]L.W.Hillier,A.Coulson,J.I.Murray,Z.Bao,J.E.Sulston,and Caenorhabditis elegans,” Genetics,vol.123,no.2,pp.301–313, R. H. Waterston, “Genomics in C. elegans: so many genes, such 1989. alittleworm,”Genome Research,vol.15,no.12,pp.1651–1660, 2005. [30] M. Morrison, K. S. Harris, and M. B. Roth, “smg mutants affect the expression of alternatively spliced SR protein mRNAs in [14] X. Hong, D. G. Scofield, and M. Lynch, “Intron size, abundance, Caenorhabditis elegans,” Proceedings of the National Academy of and distribution within untranslated regions of genes,” Molecu- Sciences of the United States of America,vol.94,no.18,pp.9782– lar Biology and Evolution,vol.23,no.12,pp.2392–2404,2006. 9785, 1997. [15]S.C.Ogg,P.Anderson,andM.P.Wickens,“SplicingofaC. [31]S.Barberan-Soler,N.J.Lambert,andA.M.Zahler,“Global elegans myosin pre-mRNA in a human nuclear extract,” Nucleic analysis of alternative splicing uncovers developmental regula- Acids Research,vol.18,no.1,pp.143–149,1990. tion of nonsense-mediated decay in C. elegans,” RNA,vol.15,no. [16] N. P. Kandul and M. A. F. Noor, “Large introns in relation to 9,pp.1652–1660,2009. alternative splicing and gene evolution: a case study of Dro- [32] O. Thompson, M. Edgley, P. Strasbourger et al., “The Million sophila bruno-3,” BMC Genetics, vol. 10, article 1471, p. 67, 2009. Mutation Project: a new approach to genetics in Caenorhabditis [17] J. L. Kabat, S. Barberan-Soler, P. McKenna, H. Clawson, T. elegans,” Genome Research,2013. Farrer, and A. M. Zahler, “Intronic alternative splicing regula- [33] S. Barberan-Soler, P. Medina, J. Estella, J. Williams, and A. M. tors identified by comparative genomics in nematodes,” PLoS Zahler, “Co-regulation of alternative splicing by diverse splicing Computational Biology,vol.2,no.7,articlee86,2006. factors in Caenorhabditis elegans,” Nucleic Acids Research,vol. [18] H. Kuroyanagi, G. Ohno, S. Mitani, and M. Hagiwara, “The 39,no.2,pp.666–674,2011. Fox-1 family and SUP-12 coordinately regulate tissue-specific [34] R. S. Kamath, A. G. Fraser, Y.Dong et al., “Systematic functional alternative splicing in vivo,” Molecular and Cellular Biology,vol. analysis of the Caenorhabditis elegans genome using RNAi,” 27,no.24,pp.8612–8621,2007. Nature,vol.421,no.6920,pp.231–237,2003. [19] B. Modrek and C. J. Lee, “Alternative splicing in the human, [35] J. Ceron, J.-F. Rual, A. Chandra, D. Dupuy, M. Vidal, and S. Van mouse and rat genomes is associated with an increased fre- Den Heuvel, “Large-scale RNAi screens identify novel genes quency of exon creation and/or loss,” Nature Genetics,vol.34, that interact with the C. elegans retinoblastoma pathway as well no. 2, pp. 177–180, 2003. as splicing-related components with synMuv B activity,” BMC [20]D.B.Malko,V.J.Makeev,A.A.Mironov,andM.S.Gelfand, Developmental Biology,vol.7,article30,2007. “Evolution of exon-intron structure and alternative splicing in [36] J. A. Kerins, M. Hanazawa, M. Dorsett, and T. Schedl, “PRP-17 fruit flies and malarial mosquito genomes,” Genome Research, and the pre-mRNA splicing pathway are preferentially required vol. 16, no. 4, pp. 505–509, 2006. for the proliferation versus meiotic development decision and [21] S. Barberan-Soler and A. M. Zahler, “Alternative splicing and germline sex determination in Caenorhabditis elegans,” Devel- the steady-state ratios of mRNA isoforms generated by it are opmental Dynamics, vol. 239, no. 5, pp. 1555–1572, 2010. under strong stabilizing selection in Caenorhabditis elegans,” [37] H. Kuroyanagi, T. Kobayashi, S. Mitani, and M. Hagiwara, Molecular Biology and Evolution,vol.25,no.11,pp.2431–2437, “Transgenic alternative-splicing reporters reveal tissue-specific 2008. expression profiles and regulation mechanisms in vivo,” Nature [22] M. Irimia, J. L. Rukov, D. Penny, J. Garcia-Fernandez, J. Vinther, Methods,vol.3,no.11,pp.909–915,2006. and S. W. Roy, “Widespread evolutionary conservation of alter- [38] G. Ohno, M. Hagiwara, and H. Kuroyanagi, “STAR family RNA- natively spliced exons in Caenorhabditis,” Molecular Biology binding protein ASD-2 regulates developmental switching of and Evolution,vol.25,no.2,pp.375–382,2008. mutually exclusive alternative splicing in vivo,” Genes and Devel- [23] S. Barberan-Soler and A. M. Zahler, “Alternative splicing regula- opment,vol.22,no.3,pp.360–374,2008. tion during C. elegans development: splicing factors as regulated [39]G.Ohno,K.Ono,M.Togoetal.,“Muscle-specificsplicing targets,” PLoS Genetics, vol. 4, no. 2, Article ID e1000001, 2008. factors ASD-2 and SUP-12 cooperatively switch alternative International Journal of Cell Biology 9

pre-mRNA processing patterns of the ADF/cofilin gene in [57] D. Merkle and R. Hoffmann, “Roles of cAMP and cAMP- Caenorhabditis elegans,” PLoS Genetics,vol.8,no.10,ArticleID dependent protein kinase in the progression of prostate cancer: e1002991, 2012. cross-talk with the androgen receptor,” Cellular Signalling,vol. [40] H. Kuroyanagi, H. Kuroyanagi, Y. Watanabe, Y. Suzuki, and M. 23, no. 3, pp. 507–515, 2011. Hagiwara, “Position-dependent and neuron-specific splicing [58]M.Tabish,R.A.Clegg,H.H.Rees,andM.J.Fisher,“Orga- regulation by the CELF family RNA-binding protein UNC-75 in nization and alternative splicing of the Caenorhabditis elegans Caenorhabditis elegans,” Nucleic Acids Research,vol.41,no.7,pp. cAMP-dependent protein kinase catalytic-subunit gene (kin- 4015–4025, 2013. 1),” Biochemical Journal,vol.339,no.1,pp.209–216,1999. [41] M. V. Sundaram, “RTK/Ras/MAPK signaling,” WormBook,pp. [59] L. C. Bowen, A. V. Bicknell, M. Tabish, R. A. Clegg, H. H. Rees, 1–19, 2006. andM.J.Fisher,“ExpressionofmultipleisoformsofthecAMP- dependent protein kinase (PK-A) catalytic subunit in the [42] J. Kimble, “Alterations in cell lineage following laser ablation of nematode, Caenorhabditis elegans,” Cellular Signalling,vol.18, cells in the somatic gonad of Caenorhabditis elegans,” Develop- no.12,pp.2230–2237,2006. mental Biology,vol.87,no.2,pp.286–300,1981. [60] R. A. Clegg, L. C. Bowen, A. V.Bicknell et al., “Characterisation [43] P. W. Sternberg and H. R. Horvitz, “Pattern formation during of the N’1 isoform of the cyclic AMP-dependent protein kinase vulval development in C. elegans,” Cell,vol.44,no.5,pp.761– (PK-A) catalytic subunit in the nematode, Caenorhabditis ele- 772, 1986. gans,” Archives of Biochemistry and Biophysics,vol.519,no.1,pp. [44] M. Han and P. W. Sternberg, “let-60, a gene that specifies cell 38–45, 2012. fates during C. elegans vulval induction, encodes a ras protein,” [61] M. W.Pastok, M. C. Prescott, C. Dart et al., “Structural diversity Cell,vol.63,no.5,pp.921–931,1990. of the cAMP-dependent protein kinase regulatory subunit in [45]C.VanBuskirkandP.W.Sternberg,“Epidermalgrowthfactor Caenorhabditis elegans,” Cell Signaling,vol.25,no.1,pp.168–177, signaling induces behavioral quiescence in Caenorhabditis ele- 2013. gans,” Nature Neuroscience, vol. 10, no. 10, pp. 1300–1307, 2007. [62]P.Murray,R.A.Clegg,H.H.Rees,andM.J.Fisher,“siRNA- [46]A.Dutt,S.Canevascini,E.Froehli-Hoier,andA.Hajnal,“EGF mediatedknockdownofasplicevariantofthePK-Acatalytic signal propagation during C. elegans vulval development medi- subunit gene causes adult-onset paralysis in C. elegans,” Gene, ated by ROM-1 Rhomboid,” PLoS Biology,vol.2,no.11,article vol. 408, no. 1-2, pp. 157–163, 2008. e334, 2004. [63] M. H. Lee, Y. J. Jang, and H.-S. Koo, “Alternative splicing in the [47] M. Trahey and F.McCormick, “Acytoplasmic protein stimulates Caenorhabditis elegans DNA topoisomerase I gene,” Biochimica normal N-ras p21 GTPase, but does not affect oncogenic et Biophysica Acta, vol. 1396, no. 2, pp. 207–214, 1998. mutants,” Science,vol.238,no.4826,pp.542–545,1987. [64] M. H. Lee, H. Park, G. Shim, J. Lee, and H.-S. Koo, “Regulation [48] M. Trahey, G. Wong, R. Halenbeck et al., “Molecular cloning of of gene expression, cellular localization, and in vivo function of two types of GAP complementary DNA from human placenta,” Caenorhabditis elegans DNA topoisomerase I,” Genes to Cells, Science, vol. 242, no. 4886, pp. 1697–1700, 1988. vol. 6, no. 4, pp. 303–312, 2001. [49] S. Hayashizaki, Y. Iino, and M. Yamamoto, “Characterization [65] D. S. Cha, S. E. Hollis, U. S. Datla et al., “Differential subcellular of the C. elegans gap-2 gene encoding a novel Ras-GTpase localization of DNA topoisomerase-1 isoforms and their roles activating protein and its possible role in larval development,” during Caenorhabditis elegans development,” Gene Expression Genes to Cells,vol.3,no.3,pp.189–202,1998. Patterns,vol.12,no.5,pp.189–195,2012. [50] M. A. Allen, L. W. Hillier, R. H. Waterston, and T. Blumen- [66] F. Rossi, E. Labourier, T. Forne´ et al., “Specific phosphorylation thal, “A global analysis of C. elegans trans-splicing,” Genome of SR proteins by mammalian DNA topoisomerase I,” Nature, Research, vol. 21, no. 2, pp. 255–264, 2011. vol.381,no.6577,pp.80–82,1996. [67] E. Labourier, F.Rossi, I.-E. Gallouzi, E. Allemand, G. Divita, and [51] S. Iwashita and S.-Y. Song, “RasGAPs: a crucial regulator J. Tazi, “Interaction between the N-terminal domain of human of extracellular stimuli for homeostasis of cellular functions,” DNA topoisomerase I and the arginine-serine domain of its Molecular BioSystems,vol.4,no.3,pp.213–222,2008. substrate determines phosphorylation of SF2/ASF splicing fac- [52] A. G. Knudson Jr., “Retinoblastoma: a prototypic hereditary tor,” Nucleic Acids Research, vol. 26, no. 12, pp. 2955–2962, 1998. neoplasm,” Seminars in Oncology,vol.5,no.1,pp.57–60,1978. [68] T.D.Dubatolova,E.I.Volkova,andL.V.Omelyanchuk,“Droso- [53] X. Lu and H. R. Horvitz, “Lin-35 and lin-53, two genes that phila splicing factor SF2 knock-down mutant shows altered cell- antagonize a C. elegans Ras pathway, encode proteins similar cycle in vivo,” Cell Biology International,vol.37,no.2,pp.187– to Rb and its binding protein RbAp48,” Cell,vol.95,no.7,pp. 190, 2013. 981–991, 1998. [69] Y. Kassir and G. Simchen, “Meiotic recombination and DNA [54]C.J.CeolandH.R.Horvitz,“dpl-1DPandefl-1E2Fact synthesis in a new cell cycle mutant of Saccharomyces cerevisiae,” with lin-35 Rb to antagonize Ras signaling in C. elegans vulval Genetics,vol.90,no.1,pp.49–68,1978. development,” Molecular Cell,vol.7,no.3,pp.461–473,2001. [70] P. L. Graham, T. Schedl, and J. Kimble, “More mog genes that [55] E. L. Ferguson and H. R. Horvitz, “The multivulva phenotype influence the switch from spermatogenesis to oogenesis in the of certain Caenorhabditis elegans mutants results from defects hermaphrodite germ line of Caenorhabditis elegans,” Develop- in two functionally redundant pathways,” Genetics,vol.123,no. mental Genetics,vol.14,no.6,pp.471–484,1993. 1, pp. 109–121, 1989. [71] A. Puoti and J. Kimble, “The Caenorhabditis elegans sex deter- [56] D. Joseph-Strauss, M. Gorjan´ acz,´ R. Santarella-Mellwig, E. mination gene mog-1 encodes a member of the DEAH-box Voronina, A. Audhya, and O. Cohen-Fix, “Sm protein down- protein family,” Molecular and Cellular Biology,vol.19,no.3,pp. regulation leads to defects in nuclear pore complex disassembly 2189–2197, 1999. and distribution in C. elegans embryos,” Developmental Biology, [72] A. Puoti and J. Kimble, “The hermaphrodite sperm/oocyte vol. 365, no. 2, pp. 445–457, 2012. switch requires the Caenorhabditis elegans homologs of PRP2 10 International Journal of Cell Biology

and PRP22,” Proceedings of the National Academy of Sciences of [88] B. D. Galvin, D. P. Denning, and H. R. Horvitz, “SPK-1, an SR the United States of America,vol.97,no.7,pp.3276–3281,2000. protein kinase, inhibits programmed cell death in Caenorhab- [73] F. Caspary, A. Shevchenko, M. Wilm, and B. Seraphin,´ “Partial ditis elegans,” Proceedings of the National Academy of Sciences of purification of the yeast U2 snRNP reveals a novel yeast pre- the United States of America,vol.108,no.5,pp.1998–2003,2011. mRNA splicing factor required for pre-spliceosome assembly,” [89] H. Kuroyanagi, T. Kimura, K. Wada, N. Hisamoto, K. Mat- EMBO Journal,vol.18,no.12,pp.3463–3474,1999. sumoto, and M. Hagiwara, “SPK-1, a C. elegans SR protein [74] A. Dziembowski, A.-P. Ventura, B. Rutz et al., “Proteomic anal- kinase homologue, is essential for embryogenesis and required ysis identifies a new complex required for nuclear pre-mRNA for germline development,” Mechanisms of Development,vol.99, retention and splicing,” EMBO Journal,vol.23,no.24,pp.4847– no. 1-2, pp. 51–64, 2000. 4856, 2004. [90] D. Longman, I. L. Johnstone, and J. F. Caceres,´ “Functional [75] P. Mantina, L. MacDonald, A. Kulaga, L. Zhao, and D. Hansen, characterization of SR and SR-related genes in Caenorhabditis “A mutation in teg-4, which encodes a protein homologous to elegans,” EMBO Journal,vol.19,no.7,pp.1625–1637,2000. the SAP130 pre-mRNA splicing factor, disrupts the balance between proliferation and differentiation in the C. elegans germ line,” Mechanisms of Development,vol.126,no.5-6,pp.417–429, 2009. [76] B. Laggerbauer, S. Liu, E. Makarov et al., “The human U5 snRNP 52K protein (CD2BP2) interacts with U5-102K (hPrp6), a U4/ U6.U5 tri-snRNP bridging protein, but dissociates upon tri- snRNP formation,” RNA, vol. 11, no. 5, pp. 598–608, 2005. [77] C. Wang, L. Wilson-Berry, T. Schedl, and D. Hansen, “TEG-1 CD2BP2 regulates stem cell proliferation and sex determination in the C. elegans germ line and physically interacts with the UAF-1 U2AF65 splicing factor,” Developmental Dynamics,vol. 241, no. 3, pp. 505–521, 2012. [78] H. M. Ellis and H. R. Horvitz, “Genetic control of programmed cell death in the nematode C. elegans,” Cell,vol.44,no.6,pp. 817–829, 1986. [79]D.Wu,H.D.Wallen,N.Inohara,andG.Nunez,˜ “Interaction andregulationoftheCaenorhabditis elegans death protease CED-3 by CED-4 and CED-9,” JournalofBiologicalChemistry, vol.272,no.34,pp.21449–21454,1997. [80] A. M. Chinnaiyan, D. Chaudhary, K. O’Rourke, E. V. Koonin, and V. M. Dixit, “Role of CED-4 in the activation of CED-3,” Nature,vol.388,no.6644,pp.728–729,1997. [81] S. Seshagiri and L. K. Miller, “Caenorhabditis elegans CED-4 stimulates CED-3 processing and CED-3-induced apoptosis,” Current Biology,vol.7,no.7,pp.455–460,1997. [82] S.ShahamandH.R.Horvitz,“AnalternativelysplicedC. elegans ced-4 RNA encodes a novel cell death inhibitor,” Cell,vol.86,no. 2, pp. 201–208, 1996. [83]M.O.Hengartner,R.E.Ellis,andR.Horvitz,“Caenorhabditis elegans gene ced-9 protects cells from programmed cell death,” Nature,vol.356,no.6369,pp.494–499,1992. [84] D. Chaudhary, K. O’Rourke, A. M. Chinnaiyan, and V. M. Dixit, “The death inhibitory molecules CED-9 and CED-4L use a common mechanism to inhibit the CED-3 death protease,” JournalofBiologicalChemistry,vol.273,no.28,pp.17708–17712, 1998. [85]L.Wang,M.Miura,L.Bergeron,H.Zhu,andJ.Yuan,“Ich-1, an ice/ced-3-related gene, encodes both positive and negative regulators of programmed cell death,” Cell,vol.78,no.5,pp. 739–750, 1994. [86]Z.-H.Jiang,W.-J.Zhang,R.Yi,andJ.Y.Wu,“RegulationofIch- 1 pre-mRNA alternative splicing and apoptosis by mammalian splicing factors,” Proceedings of the National Academy of Sciences of the United States of America,vol.95,no.16,pp.9155–9160, 1998. [87] M. J. Moore, Q. Wang, C. J. Kennedy, and P. A. Silver, “An alternative splicing network links cell-cycle control to apoptosis,” Cell,vol.142,no.4,pp.625–636,2010. Hindawi Publishing Corporation International Journal of Cell Biology Volume 2013, Article ID 458727, 10 pages http://dx.doi.org/10.1155/2013/458727

Review Article Alternative Splicing Programs in Prostate Cancer

Claudio Sette1,2 1 Department of Biomedicine and Prevention, University of Rome “Tor Vergata,” 00133 Rome, Italy 2 Laboratory of Neuroembryology, Fondazione Santa Lucia IRCCS, 00143 Rome, Italy

Correspondence should be addressed to Claudio Sette; [email protected]

Received 31 May 2013; Accepted 11 July 2013

Academic Editor: Michael Ladomery

Copyright © 2013 Claudio Sette. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Prostate cancer (PCa) remains one of the most frequent causes of death for cancer in the male population. Although the initial antiandrogenic therapies are efficacious, PCa often evolves into a hormone-resistant, incurable disease. The genetic and phenotypic heterogeneity of this type of cancer renders its diagnosis and cure particularly challenging. Mounting evidence indicates that alternative splicing, the process that allows production of multiple mRNA variants from each gene, contributes to the heterogeneity of the disease. Key genes for the biology of normal and neoplastic prostate cells, such as those encoding for the androgen receptor and cyclin D1, are alternatively spliced to yield protein isoforms with different or even opposing functions. This review illustrates some examples of genes whose alternative splicing regulation is relevant to PCa biology and discusses the possibility to exploit alternative splicing regulation as a novel tool for prognosis, diagnosis, and therapeutic approaches to PCa.

1. Introduction PCa cells rely on androgens and on the androgen receptor (AR) for proliferation [1]. Under normal conditions, the AR Cancer cells are characterized by uncontrolled growth and is localized in the cytoplasm; upon binding to androgens, the ability to migrate from the primary lesion and to establish receptor dimerizes, translocates to the nucleus, and trans- metastases in distant tissues. Standard therapies involve sur- activates genes containing androgen-responsive elements in gical removal of the tumor mass, radiation, and chemother- their promoter regions. Clinical treatments are currently apy, which exploit the increased growth rate of cancer cells basedonandrogenablationtherapies,obtainedbychemical with respect to surrounding cells. More targeted approaches castration with drugs that block secretion of the hormone have also been developed in the last decades by directly or by directly targeting the AR with androgen antagonists inhibiting the function of the oncoproteins responsible for [1, 2]. However, after initial remission, many patients develop the neoplastic transformation. Nevertheless, although many a hormone-resistant or castration-resistant form of PCa human cancers initially respond to therapies, and in some (CRPCa), for which no cure is available [1, 2]. Notably, cases patients are cured, most of them are characterized by in most cases CRPCa cells still require the AR, but they disease relapse that often occurs in more aggressive and can bypass activation by androgens or be stimulated by the incurable forms. In this regard, a clear example of aggressive lowandrogenlevelspresentduringthetherapyorbythe relapsing tumor is represented by prostate carcinoma (PCa), antagonists used for the therapy [3]. Several mechanisms for which remains one of the main causes of death for cancer in the development of androgen insensitivity of the AR have themalepopulation[1, 2]. Understanding the mechanisms been documented [1]. Among these, recent evidence points that lead to the acquisition of resistance to therapies in PCa to alternative splicing (AS) of AR as a key resource utilized patients might offer new molecular markers for earlier and byPCacellstoevadethenormalrouteofactivationofthis more accurate diagnoses. Furthermore, identification of the pathway [4]. key players involved in the transition to therapy-refractory AS is emerging as a key step in the regulation of crucial stages may shed light on new targets for pharmaceutical cellular and developmental pathways in higher eukaryotes intervention and open the path for the development of novel [5]. With regard to PCa, it has been proposed that the and more efficacious therapies. “splicing signature” represents a more accurate parameter 2 International Journal of Cell Biology to stratify patients than the “transcriptome signature,” which from the inclusion of cryptic exons located in intron 2 and 3 is typically analysed by conventional microarray analyses [6]. of the AR gene. The heterogeneity of the AR variants reported Thus, understanding the regulation of splicing in normal and in various studies is also due to the frequent amplification pathological prostate cells may help identify novel markers oftheexon2-exon3genomicregionofAR [4, 9]. In all and targets for future therapeutic approaches to this neoplas- reported variants, however, the splicing of these cryptic exons tic disease. invariably introduces premature stop codons and termination sites, thereby yielding shorter AR proteins of 75–80 kDa, which lack the androgen-binding domain [4]. In some cases, 2. Alternative Splicing and Cancer these truncated AR variants can function independently of the full-length AR, and their selective knockdown was The recent advent of high-throughput RNA sequencing has shown to block the androgen-independent growth of CRPCa unveiled new layers of regulation of gene expression and cells while maintaining responsiveness to the hormone [10]. highlighted the extreme complexity and versatility of the Importantly, expression of an AR variant containing a cryptic genome. The majority of human genes encode multiple tran- exon located in intron 3 (CE3) in clinical PCa specimens scripts through the use of alternative promoters, AS and positively correlated with poor prognosis after surgery11 [ ]. alternative polyadenylation [5, 7, 8]. AS is a combinatorial This variant was also expressed at higher levels in CRPCa mechanism that expands the coding potential of the genome withrespecttoPCapatients[12]. Furthermore, constitutively by allowing the production of protein isoforms with different active truncated AR splice variants were recently shown to or even antagonistic functions from a single gene [5, 7, 8]. confer resistance also to the next generation of AR inhibitors, Splicing is orchestrated by a ribonucleoprotein complex thus limiting their therapeutic efficacy for many patients [13]. called “spliceosome,” which recognizes exon-intron junc- However, since the expression of these shorter AR variants tions, excises introns, and ligates exons. The lack of stringent was also observed in normal prostate tissues, it is unlikely consensus sequences at exon-intron junctions in higher that they drive the initial steps of neoplastic transformation, eukaryotes allows flexibility in recognition by the spliceo- openingthepossibilitythatASoftheAR gene plays also a some. Numerous RNA binding proteins (RBPs) interact with physiological role in the gland [4]. components of the spliceosome and reinforce or weaken rec- The mechanisms that lead to increased expression of ognition of exon-intron junctions. The interplay between aberrant AR splice variants in PCa are still largely unknown. these splicing factors determines the choice of variable exons One possible cause of defective splicing is the alteration of by the spliceosome and causes heterogeneity in pre-mRNA the genomic AR locus, which often occurs in CRPCa. For processing [5, 7]. As a consequence, changes in the expression instance, disruption of the AR splicing pattern in the 22Rv1 levels or in the activity of splicing factors can selectively PCa cell line was linked to duplication of the genomic influence AS of many genes [5, 7]. Although the flexibility of region containing exon 3 and some of the cryptic exons [9]. AS regulation has represented an evolutionary advantage for Alternatively, aberrant expression of specific splicing factors higher eukaryotes, it also represents a risk factor. In particu- in PCa cells may also contribute to unbalanced splicing and lar, mounting evidence illustrates how defective regulation of aberrant recognition of cryptic exons in the AR gene. Thus, AS correlates with onset and progression of human cancers given the strong relevance of these constitutively active AR [7, 8]. variants for CRPCa progression, further studies elucidating Herein, the literature describing the impact of AS in the regulation of their expression are strongly encouraged. the onset and progression of PCa will be reviewed. High- throughput analyses of specimens from PCa patients have highlighted more than 200 genes whose AS is differentially regulatedintheneoplastictissue[6], indicating that this 4. KLF6 mechanism can substantially contribute to the heterogeneity Kruppel-like factor 6 (KLF6) is a zinc finger transcription of gene expression in cancer cells. Nevertheless, the physi- factor that is mutated in a subset of human PCas [14, 15]. ological consequences of the majority of these aberrant AS KLF6isknowntoregulatecellproliferationbyinducing events are still unknown and will require direct investigation. the expression of the cell cycle inhibitor p21 (WAF1/CIP1). This review will focus on the regulation of genes and splicing Notably, this effect of KLF6 does not require p53, suggesting regulators whose relevance for PCa has been firmly docu- that KLF6 is a tumor suppressor gene that functions as a p53- mented. alternative brake for cell cycle progression in normal cells [14]. One of the mutations found in PCa patients consists 3. The Androgen Receptor of a single nucleotide change that creates a binding site for the splicing factor SRSF5 (SRp40) and enhances splicing of Several reports have documented the expression of alter- three alternative mRNA variants encoding for truncated KLF natively spliced AR variants lacking the C-terminal ligand- proteins,namedKLF6-SV1,SV2,andSV3[16]. These splice binding domain of the canonical receptor (reviewed in [4]). variants are upregulated in tumor versus normal prostatic Many of these AR splice variants are constitutively nuclear tissue. A single G > Amutationinintron1wasshownto andactiveevenintheabsenceofandrogens(Figure 1), thus recapitulate the altered splicing pattern of KLF6 when it indicating their potential role in the acquisition of the CRPCa was introduced in a minigene, and it was found to correlate phenotype [4]. Expression of most of these variants arises with worse prognosis in patients [16]. The KLF-SV1 variant International Journal of Cell Biology 3

1 2 3 4 5 6 7 8

AR gene Androgen-dependent activation

Cryptic exons 3 CE3 4

AR3 1 2 3 CE3

Constitutive activation Androgen-independent growth

CYD1 A

1 2 3 4 5 CCND1gene

pA sites

4 Intron 4 5

CYD1 B

1 2 3 4

Cellular transformation Enhancement of AR-dependent transcription

BCL-XS 1 2 3 BCL-X gene Proapoptotic 󳰀 󳰀 󳰀 5 5 3 1 2 3

BCL-XL 1 2 3

Antiapoptotic chemotherapy resistance

Figure 1: Representative examples of genes whose alternative splicing affects prostate cancer cell biology. The left side of the figure illustrates the genomic structure of the alternatively spliced regions of the AR, CCND1, and BCL-X genes. Solid and dashed lines show the alternative splicing events reported in the literature. On the right side, the alternative variants produced by splicing are shown. The specific features of the protein isoforms produced by alternative splicing are summarized under the scheme of each variant. was characterized further and shown to function as a knockdown of the full length KLF6 promoted tumor forma- dominant-negative protein, which antagonizes the function tion in nude mice, selective silencing of the KLF6-SV1 variant of full length KLF6, leading to decreased p21 expression inhibited it [18]. Conversely, PCa cells overexpressing KLF6- and enhanced cell growth [16]. Increased expression of SV1 are more prone to develop metastases in various organs this splice variant in PCa patients predicted poorer out- of the mouse models used in the study [17]. Thus, a mutation come after surgery and was associated with development of affecting KLF6 AS represents a critical mechanism for the hormone-refractory metastatic PCa [17]. Furthermore, while inactivation of a tumor suppressor gene in PCa, suggesting 4 International Journal of Cell Biology

that interfering with this splicing event in PCa cells might oneattheendoftheexonyieldsthelongBCL-XL variant, restore the growth-inhibitory activity of this transcription whereas selection of the distal one located 220 bp upstream factor. in the exon produces the short BCL-XS variant. Notably, these two splice variants have opposite effects in the cell, with BCL-XL being prosurvival whereas BCL-XS is proapoptotic 5. Cyclin D1 (Figure 1)[28]. Thus, regulation of BCL-X AS can finely modulate cell viability, illustrating the biological importance CCND1 is a protooncogene that encodes for cyclin D1, which of this splicing event. In most cancer cells, including PCa associates with the cyclin-dependent kinase 4 (CDK4) to cells, the anti-apoptotic BCL-X variant is overexpressed and drive progression through the G1 phase of the cell cycle. L confers resistance to chemotherapeutic treatments [29, 30]. Importantly, CCND1 expression is often deregulated in cancer It is predictable that a full understanding of the mechanisms cells [19, 20]. This gene encodes for two alternative tran- of regulation of BCL-X splicing will help develop tools to scripts: the common cyclin D1a isoform, containing all five switch it toward the proapoptotic BCL-X variant, thereby exons, and cyclin D1b, which derives from retention of intron S offering a therapeutic opportunity to sensitize cancer cells 4 and premature termination of the transcript (Figure 1)[19, to treatments. In line with this notion, treatment of PCa 20]. Unlike cyclin D1a, cyclin D1b alone can promote cellular cells with an antisense oligonucleotide (ASO) masking the transformation [21], and its expression has been associated BCL-X splice site effectively switched BCL-X splicing and with PCa progression and poor prognosis [22]. Interestingly, L induced apoptosis [29]. Interestingly, the proapoptotic effect recent evidence indicated that cyclin D1b promotes AR- of the ASO was more pronounced in cancer cells, which dependent transcription of genes involved in PCa metastatic display high levels of expression of BCL-X ,anditalso potential, such as the transcription factor SLUG [23]. Cyclin L enhanced their response to chemotherapeutic treatments D1a was instead reported to repress the transcriptional [30]. Thus, ASOs targeting BCL-X splicing may have the activity of AR (Figure 1)[19, 20]. Thus, it is conceivable that advantage of being selective for cancer cells with respect to a change in the ratio between the cyclin D1 variants will normal cells, which is a positive feature for an antineoplastic potently enhance hormone-dependent growth of PCa cells. drug. Unfortunately, the delivery of ASOs to cancer cells is Given the relevance for PCa cell biology, understanding still not efficient, thus limiting their application in the clinic, the regulation of CCND1 splicing is of crucial importance. even though development of vehicles favoring their delivery, It was observed that a polymorphism (G870/A) at the such as lipid nanoparticles [31], may aid in this direction. exon 4-intron 4 boundary predisposes cells to cyclin D1b Although regulation of BCL-X splicing is highly relevant splicing [19, 20]. The splicing factor SRSF1 was shown to toPCacellbiology,notmuchisknownonthemechanism(s) bind the exon4/intron4 junction in the nascent CCND1 pre- of its regulation in prostate cells. A possible regulator is the mRNA, thereby promoting intron 4 retention and cyclin D1b protein phosphatase 1 (PP1), whose activity is required for the expression [24]. SRSF1 was hypothesized to favour intron 4 regulatory effect of ceramide on BCL-X splicing [32]. Indeed, retention by altering exon 4 definition and limiting assembly PP1 activity was required also for induction of BCL-X of the spliceosome at the exon-intron junction [24]. Another S splicing by emetine, a protein synthesis inhibitor, and other splicing factor promoting cyclin D1b expression in PCa cells proapoptotic drugs in PCa cells [33, 34]. Nevertheless, the is SAM68 [25]. In this case, the binding site was identified mechanism by which PP1 modulates splicing of BCL-X is still within intron 4, in proximity of the termination site utilized unknown. PP1 is known to regulate splicing by modulating for the cyclin D1b mRNA. The binding of SAM68 to this the activity of splicing factors, either by direct binding to region of the pre-mRNA was shown to compete with that them and regulation of their phosphorylation status [35]or of the U1 snRNP [25]. Since deposition of U1 snRNP near indirectly by regulating kinases involved in their post- cryptic polyadenylation sites located in introns is known to translational modifications36 [ ]. Thus, activation of pathways prevent premature termination of transcripts in a genome- impinging on PP1 may affect BCL-X splicing and cell viability wide fashion [26], it is possible that up-regulation of SAM68 through the regulation of the activity of specific splicing unmasks the cyclin D1b termination site by interfering with factors in PCa cells. U1 snRNP binding in intron 4. Notably, both SRSF1 and Several splicing factors have been shown to modulate SAM68 display oncogenic features in several cell types and BCL-X splicing. Studies performed in a variety of cell models tissues [7, 27], and their expression positively correlates with indicated that the heterogeneous nuclear ribonucleoprotein that of cyclin D1b in clinical specimens of PCa patients [24, (hnRNP) H and F [37] and the splicing regulators SAM68 25]. Thus, it is possible that interfering with the activity of [38], RBM25 [39], and RBM11 [40]promotesplicingofthe these splicing factors will exert positive effects in therapeutic proapoptotic BCL-X variant. By contrast, hnRNP K [41], the treatments of PCa through modulation of CCND1 splicing S serine-arginine (SR) rich proteins SRSF1 [38, 42] and SRSF9 and expression. [43], and the splicing factor SAP155 [44] enhance splicing of the anti-apoptotic BCL-XL. Which of these factors contribute 6. BCL-X to the regulation of BCL-X splicing in PCa cells is still largely unknown. SRSF1 [24]andSAM68[45]wereshown The BCL-X (BCL2L1) gene contains 3 exons and encodes for to be upregulated in PCa and might represent strong candi- 󸀠 two splice variants [28]. Two alternative 5 splice sites are dates for the regulation of this splicing event. Intriguingly, present in exon 2 of the gene: selection of the canonical these splicing factors normally modulate BCL-X splicing in International Journal of Cell Biology 5 opposite directions [38]; while the up-regulation of SRSF1 is to be another suitable target for an AS-directed therapeutic in line with the high levels of BCL-XL in PCa cells, SAM68 approach that would spare normal cells not expressing the should favour the proapoptotic short variant. However, the chimeric proteins. splicing activity of SAM68 is finely tuned by phosphorylation [46],anditwasshownthattyrosinephosphorylation by the Src-related kinase FYN switched SAM68-dependent 8. Splicing Programs in Prostate Cancer splicing of BCL-X toward the anti-apoptotic variant [38, 47]. Since tyrosine phosphorylation of SAM68 is increased in Cancer cells express a number of splice variants that confer specimens of PCa patients [48], it is likely that this RBP them higher resistance to chemotherapeutic drugs and sur- can also contribute to the upregulation of BCL-XL in PCa vival advantages. When it was investigated in detail, the spe- cells. In line with this hypothesis, BCL-XL expression was cific signature of splice variants expressed by cancer cells has decreased, and sensitivity to genotoxic agents was increased, been recognized as a powerful diagnostic and prognostic tool afterknockdownofSAM68intheandrogen-sensitiveLNCaP [7, 8]. Even more importantly, the existence of cancer-specific cell line [45]. splicing variants of key genes, such as the AR or CCND1 Thus, based on the observations reported previously, in PCa, might offer a therapeutic opportunity for targeting it is predictable that exogenous modulation of BCL-X AS proteins that are not expressed in healthy cells. For instance, through administration of ASOs, or by interfering with the developing tools that specifically modulate the expression activity of the splicing factors that promote the anti-apoptotic of transcript variants preferentially or uniquely produced by BCL-XL variant, will enhance the efficacy of chemotherapy in cancer cells might slow down tumor growth and/or promote advanced PCa, as suggested by preclinical studies in PCa cell cell death during therapy, while sparing the healthy tissues. lines [30, 31, 45]. Thus, understanding AS regulation at the genome-wide level in PCa cells may not only lead to the identification of novel diagnostic or prognostic biomarkers, but it could also help 7. TMPRSS2:ERG find tools for novel therapeutic approaches to this neoplastic disease. ERG is a member of the ETS transcription factor family that A few studies have directly investigated the genome-wide is expressed at very low levels in benign prostate epithelial regulation of AS in PCa cell lines and primary tumor tissues. cells. However, PCa patients often carry a fusion of the Using a splicing-sensitive microarray, comprising a selected androgen-responsive TMPRSS2 gene with ERG,whichcauses subset of genes and splice variants, it was shown that splicing aberrantly high expression levels of the transcription factor signatures could efficiently segregate PCa cells lines from can- intheneoplasticcells.Adetailedsequencinganalysisofthe cer cell lines derived from other organs or tissues [6]. Among TMPRSS2:ERG transcripts isolated from PCa tissues revealed the alternatively spliced genes, the majority also showed that fusion-derived transcripts underwent profound AS reg- variation in expression levels [6], suggesting that regulation ulation, which yielded mRNA variants encoding both full of splicing and transcription were coupled, as also observed in length ERG proteins and isoforms lacking the ETS domain. cells exposed to DNA damage [52]. Using the same splicing- Notably, an increase in the abundance of transcripts encoding sensitive platform, it was also possible to identify splicing full length ERG correlated with less favorable outcome in signatures that were specific for normal or neoplastic prostate patients [49]. These results support a possible functional role tissues obtained from biopsies [6]. Although this approach for this transcription factor in PCa pathology and suggest that was limited to the genes and the splice variants selected for the modulation of AS events promoting less pathogenic variants platform, it provided a first indication that specific changes in may produce beneficial effects. splicing occur during prostate tumorigenesis and suggested Thishypothesisisalsosupportedbyanotherstudythat that splicing variants can represent accurate biomarkers for tested the effects exerted by the expression of TMPRSS2:ERG PCa. Nevertheless, how and when these changes occur, as well alternatively spliced transcripts in an immortalized prostate as to what extent they contribute to the acquisition of the cell line [50]. It was found that these TMPRSS2:ERG splice transformed phenotype, are still open questions. Given the variants had different oncogenic activities, in terms of pro- tight association between transcription and splicing, a spe- moting proliferation, invasion, and motility. Notably, coex- cific splicing program could result from the different activity pression of different variants produced stronger effects than of transcription factors, splicing factors, or both. Mounting either variant alone, suggesting that the presence of several evidence indicates that all these events contribute to some TMPRSS2:ERG isoforms, as it normally occurs in PCa cells, extent to the acquisition of specific splicing signatures in PCa. might confer a more malignant phenotype [50]. A further The most relevant transcription factor involved in PCa contribution of AS to the heterogeneity of TMPRSS2:ERG is the AR. Several observations suggest that in addition to 󸀠 expression is provided by the extensive variability of the 5 regulating the expression levels of target genes, AR can also untranslated region (UTR) in the splice variants observed influence the transcript variants encoded by them. Using 󸀠 in patients [51]. Indeed, AS of the 5 UTR affects the onco- comprehensive splicing-sensitive arrays, it was demonstrated genic potential of the encoded proteins by regulating their that stimulation of LNCaP cells with androgens caused qual- translation and activity. Thus, although a functional link itative changes in expression of splice variants [53]. Many of between TMPRSS2:ERG expression and PCa pathology has the events altered by treatment with androgens were due to not been firmly established yet, this fusion gene appears usage of alternative promoters within the transcription unit 6 International Journal of Cell Biology of the target gene. Some of these alternative transcripts were that AS is also affected by this interaction. Another splicing predicted to influence the function of proteins with relevance factor that may participate to AR-dependent splicing regula- to PCa, such as the mTOR regulator TSC2. Following andro- tion is SAM68 [60], which is frequently upregulated in PCa genic stimulation, AR was recruited to a cryptic promoter [25, 45]. SAM68 interacts with AR and is recruited to the PSA upstream of exon 33 in the TSC2 gene, thereby leading to promoter [60], like DDX5 [55]. Interestingly, however, the 󸀠 expression of a truncated transcript lacking the 5 exons of the interaction between SAM68 and AR exerted different effects gene. This alternative TSC2 variant would encode a truncated on transcription and splicing, as the two proteins cooperated protein lacking the domain required for the interaction with in transcriptional activation of AR-target genes but opposed TSC1, which is needed to exert negative regulation of mTOR. each other in splicing of the CD44 variable exons from a Thus, androgens may lead to activation of mTOR by relieving reporter minigene [60]. Unfortunately, the direct effects of the repressive function of the TSC1/TSC2 complex through all these RBPs on AR-dependent splicing of endogenous AR-dependent induction of a defective variant. It is worthy transcripts have not been addressed yet. Nevertheless, it is of notice that activation of the mTOR pathway has been likely that, depending on the specific complex formed, AR can linked to both tumorigenesis and resistance to therapy in PCa differentially influence splicing of its target genes in PCa cells. [54]. Thus, AR might contribute to prostate tumorigenesis An additional layer of regulation of the aberrant splicing also by causing mTOR activation through expression of this program in PCa might rely on the up-regulation of specific alternative mRNA variant of TSC2. splicing factors. Beside the already mentioned SAM68 [25, 45], one likely candidate is SRSF1, a splicing factor that is upregulated in many human cancers and was shown to 9. Splicing Regulators Contributing to Altered behave as an oncogene in mice and humans [61]. In cancer Gene Expression in Prostate Cancer cells of other tissues, SRSF1 modulates the expression of splice variants of the BIN1 and BIM genes that lack proapoptotic In addition to affecting recruitment of AR to alternative functions [61, 62]. Moreover, SRSF1 promotes splicing of promoters, androgens also affected a number of AS events in MNK2b [61], a splice variant of the eIF4E kinase MNK2 that several genes [53]. Although the mechanism(s) involved in was shown to confer chemoresistance in pancreatic adenocar- these events and their potential relevance to PCa biology was cinoma cells [63]. Importantly, MNK-dependent phospho- not investigated, it might involve the ability of AR to interact rylation of eIF4E strongly contributes to PCa tumorigenesis with cofactors that modulate the transcriptional elongation both in vitro and in vivo [64, 65], and a tight balance between rate and/or the recognition of splicing enhancers or silencers the MNK/eIF4E and the mTOR pathways is required to in the pre-mRNA (Figure 2). For instance, AR interacts with maintain efficient protein synthesis in PCa cells, thereby the cofactor of BRCA1 (COBRA1), and this interaction was enhancing their proliferation rate [64]. Thus, it will be inter- shown to influence splicing of the nascent transcripts pro- esting to determine whether SRSF1 contributes to fine-tuning duced from an androgen-dependent promoter [55]. A similar the activation of these pathways in PCa cells through the regulation of AR activity was also documented for the DEAD regulation of MNK2 AS. box RNA helicase p68 (DDX5) in the LNCaP cell line. AR and Other splicing factors may also contribute to the altered DDX5 interact and are recruited to the promoter region of the splicing program of PCa cells. Indeed, the activity of several of androgen-responsive prostate-specific antigen (PSA) gene. these RBPs is modulated by signal transduction pathways that Thisinteractionwasfunctionallyrelevant,asDDX5enhanced arefrequentlyturnedonincancer,suchasthePI3K/AKTand AR-dependent PSA expression. In addition, by using an AR- the RAS/ERK pathways (see also [66]). For instance, it was dependent minigene reporter, it was shown that DDX5 and shown that activation of AKT downstream of the epidermal AR cooperated in repressing the splicing of variable exons in growth factor (EGF) receptor modulated the activity of the CD44 gene [56]. DDX5 is involved in several steps of co- the SR protein-specific kinases and phosphorylation ofSR and posttranscriptional RNA processing, including splicing proteins, thereby affecting a large spectrum of AS events [57], and some genes appear to be particularly sensitive to the [67]. Similarly, the RAS/ERK pathway modulates a number intracellular levels of DDX5 [57, 58]. Hence, since this RNA of splicing factors involved in cancer, such as SAM68 [68] helicase is upregulated in PCa [56], it will be interesting to and the alternative splicing factor 45 (SPF45) [69], which determine to what extent it contributes to RNA processing of in turn affect expression of splice variants that regulate cell AR target genes in PCa cells. motility, proliferation, and survival. Thus, it is likely that the AR is also known to interact with several splicing fac- examples reported above represent only a small picture of the tors, suggesting a direct link between androgen-regulated overall contribution of AS and splicing factors to the wide transcription initiation and pre-mRNA splicing in PCa cells heterogeneity in gene expression observed in PCa cells and (Figure 2). The PTB-associated splicing factor (PSF) and its patients. cofactor p54nrb participate to androgen-dependent protein complexes containing the AR. PSF and p54nrb inhibit the transcriptional activity of AR by interfering with its binding 10. Conclusions and Perspectives to androgen response element and by recruiting a histone deacetylase to AR-responsive promoters [59]. Although AS is widely recognized as a powerful tool that eukaryotic direct investigation of the effect of these splicing factors on cells employ to expand the coding potential and the plasticity AR-dependent splicing events was not addressed, it is likely of their genomes. The flexibility in the recognition of exons International Journal of Cell Biology 7

AR coactivators AR corepressor

COBRA1 SAM68 PSF p54nrb

DDX5

AR AR AR transcriptional-responsive gene AR transcriptional-responsive gene RNAPII RNAPII

AR splicing-responsive mRNAs

Figure 2: Regulation of cotranscriptional splicing by proteins interacting with the androgen receptor. Coregulators of the androgen receptor (AR) can affect splicing of target genes by direct interaction with AR and modulation of its activity. COBRA1, SAM68, and DDX5 appear to promote the transcriptional activity of AR but differentially act on splicing of variable exons (red box in the left side of the figure); PSF and its interacting protein p54 (right side of the figure) repress the transcriptional activity of AR, but their effect on splicing is currently unknown (see text for more details). and introns within the transcription unit of the majority of have been provided. For instance, SAM68 can bind to RNA human genes offers the possibility to compose many mRNA only as a dimer. By exploiting this requirement, it was shown variants from each gene. Subtle changes in the cellular envi- that an RNA binding-defective SAM68 mutant exerted dom- ronment, or in external cues conveyed from the surrounding inant negative effects on SAM68-mediated SMN2 splicing by environment, may result in global changes in the tran- associating with the endogenous protein and preventing its scriptome, which in part rely on the regulation of AS. An binding to the pre-mRNA [71]. This experiment suggests that interesting observation is that apparently homogenous cell small molecules interfering with SAM68 function might dis- populations actually display large differences in gene expres- play therapeutic potential. As homodimerization is a prereq- sion. This was recently exemplified by studies that applied uisite for RNA binding, one possibility is to target the SAM68 global RNA sequencing techniques to the analysis of single dimerization domain, which was restricted to a small region cell transcriptomes. After treatment of bone-marrow-derived within its Gld1-Sam68-Grp33 (GSG) homology domain [72]. dendritic cells (BMDCs) with an inflammatory cue, it was The potential value of targeting specific components of the found that hundreds of key immune genes were differentially splicing machinery in cancer cells is also suggested by the expressed by single cells. The heterogeneity in the response antioncogenic properties of natural compounds, such as was particularly remarkable with regard to the splicing pat- spliceostatin A (SSA), in a variety of cancer cell models. terns expressed by these cells [70], suggesting that fine-tuning SSA targets the splicing factor 3B subunit 1 (SF3B1) of the of AS regulation strongly contributes to the heterogeneity of spliceosome, thus affecting a large number of splicing events a cell population. This aspect might be particularly relevant concomitantly [73]. Perhaps, more specific drugs targeting in the context of PCa, which is a neoplastic disease charac- splicing factors involved in subsets of oncogenic splicing terized by extreme heterogeneity and unpredicted response events in cancer cells, as those described above, might repre- of patients to the therapy [1, 2]. The improvements in cell sent more specific therapeutic approaches in the next future. isolation techniques coupled to the higher sensitivity of the Although the extreme flexibility of AS regulation is prone next-generation sequencing techniques may soon allow a to errors that may concur to neoplastic transformation [7, 8], highly detailed description of the transcriptome of patients, it can also be exploited therapeutically. Indeed, examples of which might result in more personalized treatments. AS modulation in selected genes by administering splicing- The studies illustrated previously suggest that the upregu- correcting ASOs to cells have been reported. In some cases, lation of selected splicing regulators in PCa, such as SAM68, this approach has also been challenged with a therapeutic SRSF1, or DDX5, directly contributes to the phenotype by application. One of the most remarkable examples is repre- altering the splicing profile of key genes. Thus, these RBPs sented by the recovery of the phenotype observed in mouse might represent potential therapeutic targets for intervention. models of Spinal Muscular Atrophy (SMA). This neurode- Although blocking the activity of a given splicing factor is generative disease is caused by inactivation of the SMN1 gene not necessarily an easy task, some examples in this direction and skipping of exon 7 in the highly homologous SMN2 gene 8 International Journal of Cell Biology

[74]. It was recently demonstrated that systemic injection [5] A. Kalsotra and T. A. Cooper, “Functional consequences of of a chemically modified ASO restored SMN2 splicing in developmentally regulated alternative splicing,” Nature Reviews vitro and in vivo and profoundly ameliorated the viability Genetics,vol.12,no.10,pp.715–729,2011. and phenotypic features of mice affected by a severe form of [6]C.Zhang,H.-R.Li,J.-B.Fanetal.,“Profilingalternatively SMA [75]. Although cancer is caused by multiple alterations, spliced mRNA isoforms for prostate cancer classification,” BMC thus limiting the application of gene-specific ASOs, it is Bioinformatics,vol.7,article202,2006. conceivable that these tools could be used in combination [7]C.J.DavidandJ.L.Manley,“Alternativepre-mRNAsplicing with standard therapies to improve the clinical response of regulation in cancer: pathways and programs unhinged,” Genes patients. For instance, an ASO that switched BCL-X splicing and Development,vol.24,no.21,pp.2343–2364,2010. toward the proapoptotic variant was effective in sensitizing [8] S. Pal, R. Gupta, and R. V. Davuluri, “Alternative transcription cancer cells to drug-induced apoptosis and to reduce growth and alternative splicing in cancer,” Pharmacology & Therapeu- of tumors in nude mice [30, 31]. A similar effect was obtained tics,vol.136,no.3,pp.283–294,2012. by switching expression of the 𝛼 to the 𝛽 variant of the signal [9]Y.Li,M.Alsagabi,D.Fan,G.S.Bova,A.H.Tewfik,andS.M. transducer and activator of transcription 3 (STAT3)gene, Dehm, “Intragenic rearrangement and altered RNA splicing of which modulates multiple oncogenic pathways [76]. In this the androgen receptor in a cell-based model of prostate cancer case, administration of a modified ASO targeted to a splicing progression,” Cancer Research,vol.71,no.6,pp.2108–2117,2011. enhancer induced expression of the endogenous STAT3𝛽 and [10] S. M. Dehm, L. J. Schmidt, H. V.Heemers, R. L. Vessella, and D. an anti-oncogenic response in vitro and in vivo [76]. These J. Tindall, “Splicing of a novel androgen receptor exon generates studies suggest that modulation of AS with synthetic drugs a constitutively active androgen receptor that mediates prostate is possible and has entered a therapeutic perspective. ASOs cancer therapy resistance,” Cancer Research,vol.68,no.13,pp. are particularly appealing in terms of high specificity and 5469–5477, 2008. reduced side effects, as they may exploit their ability to anneal [11] R. Hu, T. A. Dunn, S. Wei et al., “Ligand-independent androgen with specific sequences in the genome without affecting other receptor variants derived from splicing of cryptic exons signify features or target genes of the splicing factors involved in the hormone-refractory prostate cancer,” Cancer Research,vol.69, no. 1, pp. 16–22, 2009. oncogenicASevent.Thus,itislikelythatthesemethodscould be applied soon to the development of novel therapies aimed [12] Z. Guo, X. Yang, F. Sun et al., “A novel androgen receptor splice at fighting human cancers in which expression of specific variant is up-regulated during prostate cancer progression and promotes androgen depletion-resistant growth,” Cancer oncogenic splice variants has been firmly confirmed. Research,vol.69,no.6,pp.2305–2313,2009. [13]Y.Li,S.C.Chan,L.J.Brand,T.H.Hwang,K.A.Silverstein,and Conflict of Interests S. M. Dehm, “Androgen receptor splice variants mediate enza- lutamide resistance in castration-resistant prostate cancer cell The author declares no conflict of interests. lines,” Cancer Research,vol.73,no.2,pp.483–489,2013. [14] G. Narla, K. E. Heath, H. L. Reeves et al., “KLF6, a candidate tumor suppressor gene mutated in prostate cancer,” Science,vol. Acknowledgments 294, no. 5551, pp. 2563–2566, 2001. The author wishes to thank Dr. Chiara Naro for help with the [15] X. Liu, A. Gomez-Pinillos, C. Loder et al., “KLF6 loss of func- preparation of the figures and critical reading of the paper. tion in human prostate cancer progression is implicated in The work in the laboratory of C. Sette was supported by resistance to androgen deprivation,” The American Journal of Pathology,vol.181,no.3,pp.1007–1016,2012. Association for International Cancer Research (AICR Grant no.12-0150),theAssociazioneItalianaRicercasulCancro [16] G. Narla, A. DiFeo, H. L. Reeves et al., “A germline DNA (AIRC), and Istituto Superiore della Sanita(ISSGrantCon-` polymorphism enhances alternative splicing of the KLF6 tumor suppressor gene and is associated with increased prostate cancer venzione 11US/6A). risk,” Cancer Research,vol.65,no.4,pp.1213–1222,2005. [17] G. Narla, A. DiFeo, Y. Fernandez et al., “KLF6-SV1 overexpres- References sion accelerates human and mouse prostate cancer progression and metastasis,” Journal of Clinical Investigation,vol.118,no.8, [1] B. J. Feldman and D. Feldman, “The development of androgen- pp.2711–2721,2008. independent prostate cancer,” Nature Reviews Cancer,vol.1,no. [18] G. Narla, A. Difeo, S. Yao et al., “Targeted inhibition of the 1,pp.34–45,2001. KLF6 splice variant, KLF6 SV1, suppresses prostate cancer cell [2]M.KohliandD.J.Tindall,“Newdevelopmentsinthemedical growth and spread,” Cancer Research, vol. 65, no. 13, pp. 5761– management of prostate cancer,” Mayo Clinic Proceedings,vol. 5768, 2005. 85, no. 1, pp. 77–86, 2010. [19] K. E. Knudsen, “The cyclin D1b splice variant: an old oncogene [3] K. E. Knudsen and H. I. Scher, “Starving the addiction: new learns new tricks,” Cell Division,vol.1,article15,2006. opportunities for durable suppression of AR signaling in pros- [20] K. E. Knudsen, J. A. Diehl, C. A. Haiman, and E. S. Knudsen, tate cancer,” Clinical Cancer Research, vol. 15, no. 15, pp. 4792– “Cyclin D1: polymorphism, aberrant splicing and cancer risk,” 4798, 2009. Oncogene,vol.25,no.11,pp.1620–1628,2006. [4] S. M. Dehm and D. J. Tindall, “Alternatively spliced androgen [21]F.Lu,A.B.Gladden,andJ.A.Diehl,“Analternativelyspliced receptor variants,” Endocrine-Related Cancer,vol.18,no.5,pp. cyclin D1 isoform, cyclin D1b, is a nuclear oncogene,” Cancer R183–R196, 2011. Research,vol.63,no.21,pp.7056–7061,2003. International Journal of Cell Biology 9

[22]C.E.S.Comstock,M.A.Augello,R.PeBenitoetal.,“Cyclin the alternative splicing of the apoptotic mediator Bcl-x,” Journal D1 splice variants: polymorphism, risk, and isoform-specific of Biological Chemistry, vol. 280, no. 24, pp. 22641–22650, 2005. regulation in prostate cancer,” Clinical Cancer Research,vol.15, [38] M. P. Paronetto, T. Achsel, A. Massiello, C. E. Chalfant, and no.17,pp.5338–5349,2009. C. Sette, “The RNA-binding protein Sam68 modulates the [23] M. A. Augello, C. J. Burd, R. Birbe et al., “Convergence of onco- alternative splicing of Bcl-x,” JournalofCellBiology,vol.176,no. genic and hormone receptor pathways promotes metastatic 7, pp. 929–939, 2007. phenotypes,” The Journal of Clinical Investigation,vol.123,no. [39] A. Zhou, A. C. Ou, A. Cho, E. J. Benz Jr., and S.-C. Huang, 1,pp.493–508,2013. “Novel splicing factor RBM25 modulates Bcl-x Pre-mRNA 5󸀠 [24]N.A.Olshavsky,C.E.S.Comstock,M.J.Schieweretal.,“Iden- splice site selection,” Molecular and Cellular Biology,vol.28,no. tification of ASF/SF2 as a critical, allele-specific effector ofthe 19, pp. 5924–5936, 2008. cyclin D1b oncogene,” Cancer Research, vol. 70, no. 10, pp. 3975– [40] S. Pedrotti, R. Busa,` C. Compagnucci, and C. Sette, “The RNA 3984, 2010. recognition motif protein RBM11 is a novel tissue-specific [25] M. P. Paronetto, M. Cappellari, R. Busaetal.,“Alternativesplic-` splicing regulator,” Nucleic Acids Research,vol.40,no.3,pp. ing of the cyclin D1 proto-oncogene is regulated by the RNA- 1021–1032, 2012. binding protein Sam68,” Cancer Research,vol.70,no.1,pp.229– [41] T. Revil, J. Pelletier, J. Toutant, A. Cloutier, and B. Chabot, 239, 2010. “Heterogeneous nuclear ribonucleoprotein K represses the [26] D. Kaida, M. G. Berg, I. Younis et al., “U1 snRNP protects pre- production of pro-apoptotic Bcl-xS splice isoform,” Journal of mRNAs from premature cleavage and polyadenylation,” Nature, Biological Chemistry,vol.284,no.32,pp.21458–21467,2009. vol. 468, no. 7324, pp. 664–668, 2010. [42] M. J. Moore, Q. Wang, C. J. Kennedy, and P. A. Silver, “Analter- [27]P.Bielli,R.Busa,` M. P. Paronetto, and C. Sette, “The RNA- native splicing network links cell-cycle control to apoptosis,” binding protein Sam68 is a multifunctional player in human Cell,vol.142,no.4,pp.625–636,2010. cancer,” Endocrine-Related Cancer,vol.18,no.4,pp.R91–R102, [43]P.Cloutier,J.Toutant,L.Shkreta,S.Goekjian,T.Revil,andB. 2011. Chabot, “Antagonistic effects of the SRp30c protein and cryptic 󸀠 [28] L. H. Boise, M. Gonzalez-Garcia, C. E. Postema et al., “bcl-x, 5 splice sites on the alternative splicing of the apoptotic A bcl-2-related gene that functions as a dominant regulator of regulator Bcl-x,” Journal of Biological Chemistry,vol.283,no.31, apoptotic cell death,” Cell,vol.74,no.4,pp.597–608,1993. pp. 21315–21324, 2008. [29]D.R.Mercatante,C.D.Bortner,J.A.Cidlowski,andR.Kole, [44] A. Massiello, J. R. Roesser, and C. E. Chalfant, “SAP155 Binds “Modification of alternative splicing of Bcl-x Pre-mRNA in to ceramide-responsive RNA cis-element 1 and regulates the 󸀠 prostate and breast cancer cells: analysis of apoptosis and cell alternative 5 splice site selection of Bcl-x pre-mRNA,” The death,” JournalofBiologicalChemistry,vol.276,no.19,pp.16411– FASEB Journal,vol.20,no.10,pp.1680–1682,2006. 16417, 2001. [45] R. Busa,` M. P.Paronetto, D. Farini et al., “The RNA-binding pro- [30]D.R.Mercatante,J.L.Mohler,andR.Kole,“Cellularresponse tein Sam68 contributes to proliferation and survival of human to an antisense-mediated shift of Bcl-x pre-mRNA splicing and prostate cancer cells,” Oncogene,vol.26,no.30,pp.4372–4382, antineoplastic agents,” JournalofBiologicalChemistry,vol.277, 2007. no.51,pp.49374–49382,2002. [46] C. Sette, “Post-translational regulation of star proteins and [31] J. A. Bauman, S.-D. Li, A. Yang, L. Huang, and R. Kole, “Anti- effects on their biological functions,” Advances in Experimental tumor activity of splice-switching oligonucleotides,” Nucleic Medicine and Biology,vol.693,pp.54–66,2010. Acids Research, vol. 38, no. 22, pp. 8348–8356, 2010. [47] C. Brignatz, M. P. Paronetto, S. Opi et al., “Alternative splicing [32]C.E.Chalfant,K.Rathman,R.L.Pinkermanetal.,“Denovo modulates autoinhibition and SH3 accessibility in the Src kinase ceramide regulates the alternative splicing of caspase 9 and Bcl- Fyn,” Molecular and Cellular Biology,vol.29,no.24,pp.6438– x in A549 lung adenocarcinoma cells. Dependence on protein 6448, 2009. phosphatase-1,” Journal of Biological Chemistry,vol.277,no.15, [48] M. P. Paronetto, D. Farini, I. Sammarco et al., “Expression of a pp. 12587–12595, 2002. truncated form of the c-Kit tyrosine kinase receptor and activa- [33]K.Boon-Unge,Q.Yu,T.Zou,A.Zhou,P.Govitrapong,and tion of Src kinase in human prostatic cancer,” American Journal J. Zhou, “Emetine regulates the alternative splicing of Bcl-x of Pathology,vol.164,no.4,pp.1243–1251,2004. through a protein phosphatase 1-dependent mechanism,” [49] Y. Hu, A. Dobi, T. Sreenath et al., “Delineation of TMPRSS2- Chemistry and Biology,vol.14,no.12,pp.1386–1392,2007. ERGsplicevariantsinprostatecancer,”Clinical Cancer [34] M. H. Kim, “Protein phosphatase 1 activation and alternative Research, vol. 14, no. 15, pp. 4719–4725, 2008. splicing of Bcl-X and Mcl-1 by EGCG + ibuprofen,” Journal of [50] J. Wang, Y. Cai, W. Yu, C. Ren, D. M. Spencer, and M. Itt- Cellular Biochemistry,vol.104,no.4,pp.1491–1499,2008. mann, “Pleiotropic biological activities of alternatively spliced [35] T. Novoyatleva, B. Heinrich, Y. Tang et al., “Protein phosphatase TMPRSS2/ERG fusion gene transcripts,” Cancer Research,vol. 1 binds to the RNA recognition motif of several splicing fac- 68, no. 20, pp. 8516–8524, 2008. 󸀠 tors and regulates alternative pre-mRNA processing,” Human [51] F. Zammarchi, G. Boutsalis, and L. Cartegni, “5 UTR control Molecular Genetics,vol.17,no.1,pp.52–70,2008. of native ERG and of Tmprss2: ERG variants activity in prostate [36] N. Ghosh, N. Patel, K. Jiang et al., “Ceramide-activated procancer,” PLoS One,vol.8,no.3,ArticleIDe49721,2013. tein phosphatase involvement in insulin resistance via Akt, [52] M. J. Munoz,˜ M. S. P. Santangelo, M. P. Paronetto et al., “DNA serine/arginine-rich protein 40, and ribonucleic acid splicing in damage regulates alternative splicing through inhibition of L6 skeletal muscle cells,” Endocrinology,vol.148,no.3,pp.1359– RNA polymerase II elongation,” Cell,vol.137,no.4,pp.708–720, 1366, 2007. 2009. [37]D.Garneau,T.Revil,J.-F.Fisette,andB.Chabot,“Hetero- [53] P.Rajan, C. Dalgliesh, P.J. Carling et al., “Identification of novel geneous nuclear ribonucleoprotein F/H proteins modulate androgen-regulated pathways and mrna isoforms through 10 International Journal of Cell Biology

genome-wide exon-specific profiling of the LNCaP transcrip- [69] A. M. Al-Ayoubi, H. Zheng, Y. Liu, T. Bai, and S. T. Eblen, tome,” PLoS One, vol. 6, no. 12, Article ID e29088, 2011. “Mitogen-activated protein kinase phosphorylation of splicing [54]T.M.Morgan,T.D.Koreckij,andE.Corey,“Targetedtherapy factor 45 (SPF45) regulates SPF45 alternative splicing site for advanced prostate cancer: inhibition of the PI3K/Akt/mTOR utilization, proliferation, and cell adhesion,” Molecular and Cel- pathway,” Current Cancer Drug Targets,vol.9,no.2,pp.237–249, lular Biology, vol. 32, no. 14, pp. 2880–2893, 2012. 2009. [70] A. K. Shalek, R. Satija, X. Adiconis et al., “Single-cell transcrip- [55] J. Sun, A. L. Blair, S. E. Aiyar, and R. Li, “Cofactor of BRCA1 tomics reveals bimodality in expression and splicing in immune modulates androgen-dependent transcription and alternative cells,” Nature,2013. splicing,” JournalofSteroidBiochemistryandMolecularBiology, [71] S. Pedrotti, P.Bielli, M. P.Paronetto et al., “The splicing regulator vol. 107, no. 3–5, pp. 131–139, 2007. Sam68 binds to a novel exonic splicing silencer and functions in [56] E. L. Clark, A. Coulson, C. Dalgliesh et al., “The RNA helicase SMN2 alternative splicing in spinal muscular atrophy,” EMBO p68 is a novel androgen receptor coactivator involved in splicing Journal,vol.29,no.7,pp.1235–1247,2010. and is overexpressed in prostate cancer,” Cancer Research,vol. [72] N. H. Meyer, K. Tripsianes, M. Vincendeau et al., “Structural 68,no.19,pp.7938–7946,2008. basis for homodimerization of the Src-associated during mito- [57] E. Zonta, D. Bittencourt, S. Samaan, S. Germann, M. Dutertre, sis, 68-kDa protein (Sam68) Qua1 domain,” Journal of Biological and D. Auboeuf, “The RNA helicase DDX5/p68 is a key factor Chemistry,vol.285,no.37,pp.28893–28901,2010. promoting c-fos expression at different levels from transcription [73] S. Bonnal, L. Vigevani, and J. Valcarcel,´ “The spliceosome as to mRNA export,” Nucleic Acids Research,vol.41,no.1,pp.554– a target of novel antitumour drugs,” Nature Reviews Drug Dis- 564, 2013. covery, vol. 11, no. 11, pp. 847–859, 2012. [58] A. Honig,D.Auboeuf,M.M.Parker,B.W.O’Malley,and¨ [74] S. Pedrotti and C. Sette, “Spinal muscular atrophy: a new player S. M. Berget, “Regulation of alternative splicing by the ATP- joinsthebattleforSMN2exon7splicing,”Cell Cycle,vol.9,no. dependent DEAD-box RNA helicase p72,” Molecular and Cel- 19, pp. 3874–3879, 2010. lular Biology, vol. 22, no. 16, pp. 5698–5707, 2002. [75] Y. Hua, K. Sahashi, F. Rigo et al., “Peripheral SMN restoration is [59] X.Dong,J.Sweet,J.R.G.Challis,T.Brown,andS.J.Lye,“Tran- essential for long-term rescue of a severe spinal muscular atro- scriptional activity of androgen receptor is modulated by two phy mouse model,” Nature, vol. 478, no. 7367, pp. 123–126, 2011. RNA splicing factors, PSF and p54nrb,” Molecular and Cellular [76] F. Zammarchi, E. De Stanchina, E. Bournazou et al., “Antitu- Biology,vol.27,no.13,pp.4863–4875,2007. morigenic potential of STAT3 alternative splicing modulation,” [60] P.Rajan, L. Gaughan, C. Dalgliesh et al., “The RNA-binding and Proceedings of the National Academy of Sciences of the United adaptor protein Sam68 modulates signal-dependent splicing States of America,vol.108,no.43,pp.17779–17784,2011. and transcriptional activity of the androgen receptor,” Journal of Pathology,vol.215,no.1,pp.67–77,2008. [61]R.Karni,E.DeStanchina,S.W.Lowe,R.Sinha,D.Mu,andA. R. Krainer, “The gene encoding the splicing factor SF2/ASF isa proto-oncogene,” Nature Structural and Molecular Biology,vol. 14,no.3,pp.185–193,2007. [62] O. Anczukow,A.Z.Rosenberg,M.Akermanetal.,“Thesplicing´ factor SRSF1 regulates apoptosis and proliferation to promote mammary epithelial cell transformation,” Nature Structural and Molecular Biology,vol.19,no.2,pp.220–228,2012. [63] L. Adesso, S. Calabretta, F. Barbagallo et al., “Gemcitabine triggers a pro-survival response in pancreatic cancer cells through activation of the MNK2/eIF4E pathway,” Oncogene,vol.32,pp. 2848–2857, 2013. [64] A. Bianchini, M. Loiarro, P. Bielli et al., “Phosphorylation of eIF4E by MNKs supports protein synthesis, cell cycle progression and proliferation in prostate cancer cells,” Carcinogenesis, vol.29,no.12,pp.2279–2288,2008. [65] L. Furic, L. Rong, O. Larsson et al., “EIF4E phosphorylation promotes tumorigenesis and is associated with prostate cancer progression,” Proceedings of the National Academy of Sciences of the United States of America,vol.107,no.32,pp.14134–14139,2010. [66] C. Naro and C. Sette, “Phosphorylation-mediated regulation of alternative splicing in cancer,” International Journal of Cell Bio- logy.Inpress. [67] Z.Zhou,J.Qiu,W.Liuetal.,“TheAkt-SRPK-SRaxisconstitutes a major pathway in transducing EGF signaling to regulate alternative splicing in the nucleus,” Molecular Cell,vol.47,no. 3, pp. 422–433, 2012. [68] N.Matter,P.Herrlich,andH.Konig,¨ “Signal-dependent regulation of splicing via phosphorylation of Sam68,” Nature,vol.420, no.6916,pp.691–695,2002. Hindawi Publishing Corporation International Journal of Cell Biology Volume 2013, Article ID 843781, 9 pages http://dx.doi.org/10.1155/2013/843781

Review Article Expression of Tra2𝛽 in Cancer Cells as a Potential Contributory Factor to Neoplasia and Metastasis

Andrew Best,1 Caroline Dagliesh,1 Ingrid Ehrmann,1 Mahsa Kheirollahi-Kouhestani,1 Alison Tyson-Capper,2 and David J. Elliott1

1 Institute of Genetic Medicine, Newcastle University, Central Parkway, Newcastle upon Tyne NE1 3BZ, UK 2 Institute of Cellular Medicine, Newcastle University, Framlington Place, Newcastle upon Tyne NE2 4HH, UK

Correspondence should be addressed to David J. Elliott; [email protected]

Received 2 May 2013; Accepted 9 June 2013

Academic Editor: Claudia Ghigna

Copyright © 2013 Andrew Best et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The splicing regulator proteins SRSF1 (also known as ASF/SF2) and SRSF3 (also known as SRP20) belong to the SR family of proteins and can be upregulated in cancer. The SRSF1 gene itself is amplified in some cancer cells, and cancer-associated changes in the expression of MYC also increase SRSF1 gene expression. Increased concentrations of SRSF1 protein promote prooncogenic splicing patterns of a number of key regulators of cell growth. Here, we review the evidence that upregulation of the SR-related Tra2𝛽 protein might have a similar role in cancer cells. The TRA2B gene encoding Tra2𝛽 is amplified in particular tumours including those of the lung, ovary, cervix, stomach, head, and neck. Both TRA2B RNA and Tra2𝛽 protein levels are upregulated in breast, cervical, ovarian, and colon cancer, and Tra2𝛽 expression is associated with cancer cell survival. The TRA2B gene is a transcriptional target of the protooncogene ETS-1 which might cause higher levels of expression in some cancer cells which express this transcription factor. Known Tra2𝛽 splicing targets have important roles in cancer cells, where they affect metastasis, proliferation, and cell survival. Tra2𝛽 protein is also known to interact directly with the RBMY protein which is implicated in liver cancer.

1. Introduction implications. Drugs which target the spliceosome are also being developed as potential therapies for treating cancer Cancer is associated with a number of distinctive disease patients [7]. hallmarks [1]. These hallmarks include the ability of cancer In this review, we particularly examine the potential cells to continuously divide by maintaining proliferative role of the splicing regulator Tra2𝛽 as a modulator of gene signalling pathways and to evade growth suppressors, to 𝛽 resist cell death; to induce angiogenesis to ensure a supply of function in cancer cells. Tra2 is part of a larger protein oxygen and nutrition, and to invade other parts of the body family which contains RNA recognition motifs (RRMs) (metastasis). These hallmarks of cancer cells occur against and extended regions of serine and arginine residues (RS other changes including decreasing genome stability and domains, named following the standard 1 letter amino acid inflammation1 [ ]. code for serine and arginine) [8–10]. Core SR proteins include Changes in splicing patterns in cancer cells compared to SRSF1 (previously known as ASF/SF2) and SRSF3 (previously normal cells can contribute to each of these cancer hallmarks known as SRP20) (Figure 1). Tra2𝛽 is considered an SR-like through effects on the expression patterns of important proteinratherthanacoreSRfamilymemberbecauseoftwo protein isoforms which regulate cell behaviour [2–4]. The features.Firstly,Tra2𝛽 contains both an N- and C-terminal splicing alterations which occur in cancer cells are partially RS domains (each of the core members of the SR family has due to changes in the activity and expression of core spliceo- just a single C-terminal RS domain, with the RRM at the some components [5] and in the RNA binding proteins N-terminus). Secondly, the core group of SR proteins but which regulate alternative exon inclusion [6]. Changes in the not Tra2𝛽 can restore splicing activity to S100 extracts [11]. splicing environment in cancer cells might have therapeutic S100 extracts are made from lysed HeLa cells by high-speed 2 International Journal of Cell Biology

Modular structures of two core SR family proteins development and is essential for normal mouse embry- which have been implicated in cancer onic and brain development (TRA2B knockout mice fail to develop normally) [15, 19]. TRA2B has a paralog gene called SRSF3 N- RRM RS -C TRA2A on the long arm of human chromosome 7, and this paralog encodes Tra2𝛼 protein [20]. Paralogs are additional copies of a gene derived by duplication. TRA2A derived by PP1 gene duplication from TRA2B early in the vertebrate lineage Ψ and so is found in all vertebrates. SRSF1 N- RRM RRM RS -C AnumberoftheSRproteinshavebeenfoundtohave Modular structure of Tra2𝛽 protein roles in cancer, amongst them, SRSF1 and SRSF3 (Figures 1 and 2). The mechanism of SRSF1 upregulation in cancer cells PP1 has been explained at a mechanistic level, and the effects of this upregulation in terms of gene expression control have 𝛽 N- RS RRM RS -C Tra2 been mapped onto the pathway of oncogenesis. Here, we Glycine-rich domain review these important principles for SRSF1 and then apply these principles to gauge the likely effect of the Tra2𝛽 protein RS Region enriched in serine and arginine on cancer-specific gene expression. RRM RNA recognition motif (RRM)

Figure 1: Modular structure of the core SR family proteins SRSF1 (also known as ASF/SF2) and SRSF3 (also known as SRp20) and the 2. SRSF1 Is Upregulated in Cancer and Is a SR-like protein Tra2𝛽. The RNA recognition motif (RRM) binds to Target for the Prooncogenic Transcription target RNAs, and the RS region is responsible for protein-protein Factor Myc interactions. SRSF1 has a second RRM, annotated 𝜓RRM. SRSF1 and Tra2𝛽 have a PP1 docking site. SRSF1 upregulation in cancer cells can occur through two distinct mechanisms. Firstly, the SRSF1 gene itself can become amplified in cancer. The SRSF1 gene is on a region of chromosome 17q23 which is amplified in some breast cancers, ultracentrifugation to remove nuclei but contain most of the including in tumours with a poor prognostic outlook and core spliceosome components necessary for splicing with the in the MCF7 breast cancer cell line [21]. Analysis of the important exception of SR proteins which are insoluble in the SRSF1 gene on the cBio Cancer Genomics Portal shows magnesium concentrations used [12]. Addition of any single amplification of SRSF1 mainly in breast cancers (Figure 2) SR protein is sufficient to restore splicing activity to these S100 [22, 23]. Secondly, SRSF1 gene transcription is activated by extracts [13]. the prooncogenic transcription factor Myc which is itself Tra2𝛽 protein functions as a splicing regulator in the cell activated in some cancers. Myc upregulation in cancer leads nucleus, where it activates the inclusion of alternative exons to downstream increases in both SRSF1 mRNA and SRSF1 [14, 15]. Tra2𝛽 protein is able to interact with two types of protein expression [24]. RNA targets through its RRM. Firstly, the major RNA binding Protein expression analysis using a highly specific mon- site for Tra2𝛽 is an AGAA-rich sequence [11, 16, 17]. Although oclonal antibody showed that a number of tumours have an AGAA RNA sequence works best for Tra2𝛽 protein, an increased SRSF1 protein compared to normal tissue [21]. As NGAA sequence is actually sufficient for binding. However, well as being upregulated in some cancer cells, SRSF1 operates substituting the first A with either C, G, or T nucleotides in as a bona fide oncogene. Increased SRSF1 gene expression the NGAA target sequence decreases binding efficiency (the can transform rodent fibroblasts in an NIH3T3 assay, and the Kd value increases 2-fold between AGAA and NGAA) [16]. resulting transformed cells form tumours in nude mice [21]. Secondly, the RRM of Tra2𝛽 is able to switch to a second Tumour formation by these transformed fibroblasts is directly mode of RNA binding, in which it interacts with single- dependent on SRSF1 expression, since it is blocked by parallel stranded CAA-rich sequences within a stem loop structure shRNA inhibition of SRSF1 [21]. Together, these data suggest [17]. that upregulation of SRSF1 gene expression can be one of the When Tra2𝛽 bindstotargetRNAsiteswithinanexon,it initial steps in oncogenesis. activates splicing inclusion of these bound exons into mRNA Experiments support an important function for SRSF1 [11, 15–17]. Splicing activation by Tra2𝛽 protein is concen- protein in breast cancer cells. Mouse COMMA1-D mammary tration dependent: increased Tra2𝛽 protein concentration epithelial cells form tumours more efficiently in mice after leads to increased levels of target exon splicing inclusion transduction with SRSF1, and transduction of MF10A cells [14, 15]. The RRMs of Tra2𝛽 and SRSF1 proteins both with SRSF1 results in increased acinar size and decreased contain a docking site for protein phosphatase 1 (PP1), and apoptosis in a 3D culture model [25].Anumberofsplicing dephosphorylation of these proteins by PP1 affects alternative targetshavebeenidentifiedwhichrespondtoincreasedlevels splicing regulation [18]. of SFRS1 expressionincancercells(Table 1). These SRSF1- Tra2𝛽 protein is encoded by the TRA2B gene (also called driven splicing changes produce prooncogenic mRNA splice SFRS10) on human chromosome 3. As well as any potential isoforms, which encode proteins which decrease apoptosis role in cancer cells, Tra2𝛽 hasimportantrolesinnormal and increase cellular survival and proliferation. International Journal of Cell Biology 3

TRA2B gene SRSF1gene (encodes Tra2𝛽) 10 30 7.5

20 5

10 2.5

0 0 study with mutation data (%) data mutation with study study with mutation data (%) data mutation with study Sample alteration for each cancer cancer each for alteration Sample Sample alteration for each cancer cancer each for alteration Sample Sarcoma Sarcoma Glioblastoma Glioblastoma Bladder cancer Bladder Bladder cancer Bladder Lung adenocarcinoma Lung Lung adenocarcinoma Lung Acute myeloid leukemia myeloid Acute Brain lower grade glioma lower Brain Prostate adenocarcinoma Prostate Breast invasive carcinoma invasive Breast carcinoma invasive Breast Breast invasive carcinoma invasive Breast carcinoma invasive Breast Stomach adenocarcinoma Stomach Stomach adenocarcinoma Stomach Skin cutaneous melanoma Skin cutaneous melanoma Cancer cell line encyclopedia Bladder urothelial carcinoma urothelial Bladder Bladder urothelial carcinoma urothelial Bladder Lung, squamous cell carcinoma squamous Lung, Lung, squamous cell carcinoma squamous Lung, cell carcinoma squamous Lung, cell carcinoma squamous Lung, Kidney renal clear cell clear carcinoma Kidney renal cell clear carcinoma Kidney renal Kidney renal clear cell clear carcinoma Kidney renal cell clear carcinoma Kidney renal Cervical squamous cell carcinoma cell Cervical squamous Cervical squamous cell carcinoma cell Cervical squamous and endocervicaland adenocarcinoma endocervicaland adenocarcinoma Colon and rectum adenocarcinoma rectum and Colon adenocarcinoma rectum and Colon Colon and rectum adenocarcinoma rectum and Colon Ovarian cystadenocarcinoma serous Ovarian cystadenocarcinoma serous Ovarian cystadenocarcinoma serous Ovarian cystadenocarcinoma serous Kidney renal papillary cell carcinoma cell papillary renal Kidney Uterine corpus endometrioid carcinoma endometrioid corpus Uterine carcinoma endometrioid corpus Uterine Uterine corpus endometrioid carcinoma endometrioid corpus Uterine carcinoma endometrioid corpus Uterine Head and neck squamous cell carcinoma squamous neck and Head Head and neck squamous cell carcinoma squamous neck and Head (a) (b) SRSF3 gene 7.5

2.5

study with mutation data (%) data mutation with study 0 Sample alteration for each cancer cancer each for alteration Sample Sarcoma Glioblastoma Glioblastoma Bladder cancer Bladder Lung adenocarcinoma Lung Breast invasive carcinoma invasive Breast Breast invasive carcinoma invasive Breast Stomach adenocarcinoma Stomach Skin cutaneous melanoma Cancer cell line encyclopedia Lung, squamous cell carcinoma squamous Lung, cell carcinoma squamous Lung, Colon and rectum adenocarcinoma rectum and Colon adenocarcinoma rectum and Colon Ovarian cystadenocarcinoma serous Ovarian cystadenocarcinoma serous Kidney renal papillary cell carcinoma papillary renal Kidney Uterine corpus endometrioid carcinoma endometrioid corpus Uterine carcinoma endometrioid corpus Uterine Head and neck squamous cell carcinoma squamous neck and Head

Multiple alterations Deletion Mutation Amplification (c)

Figure 2: The (a) TRA2B,(b)SRSF1 (also known as ASF/SF2), and (c) SRSF3 (also known as SRP20) genes are amplified or otherwise mutated in several cancer types. For each of the three genes, data for genetic changes in all cancers were obtained using the cBioPortal database, filtering for percentage of altered cases (studies using mutation data) [22, 23]. The percentage, of cancer samples which showed genetic alterations in large cancer studies are shown on the Y axis and the respective type of cancer on the X axis. Full details of this kind of analysis are given on the cBioPortal website http://www.cbioportal.org/public-portal/index.do.

3. Increased SRSF3 Expression Is Also Increased SRSF3 expressionlevelshavebeenassociated Associated with Cancer with an increased tumour grade in ovarian cancer [27]. Intracellular levels of SRSF3 mRNA are important for cancer Increased expression of the SR protein SRSF3 is also asso- cells: siRNA-mediated downregulation of SRSF3 leads to cell ciated with cancer. The SRSF3 gene is amplified in some cycle arrest at G1 in colon cancer cells, and their increased cancers (Figure 2)[22, 23]. Loss of SRSF3 expression in a death through apoptosis. The mechanism of increased apop- number of cancer cell lines increases apoptosis and decreases tosis in response to higher levels of SRSF3 protein might proliferation, and increased expression of SRSF3 leads to include aberrant splicing of the HIPK2 pre-mRNA (which transformation of rodent fibroblasts and enables them to encodes an important apoptotic regulator related to HIPK3, form tumours in nude mice [26]. which is a known splicing target of Tra2𝛽), such that 4 International Journal of Cell Biology

Table 1: Known prooncogenic splicing targets of SFRS1 (previously whichisakeydriverofestrogenreceptorpositivebreast known as ASF/SF2). cancer development. Taken together, these observations suggest that the pathological mechanism of Tra2𝛽 upregulation Splicing Possible role in cancer cells Reference target in cancer cells might result from underlying changes in transcription factors in cancer cells. Other positive regulators Δexon11 splice isoform increases RON [21, 25] of cell growth might also stimulate Tra2𝛽 expression, since cell motility and metastasis expression of Tra2𝛽 isupregulatedinresponsetogrowth BIN12a splice isoform encodes factors in normal smooth muscle cells [38]. BIN protein no longer able to bind Myc [21, 25] andactsastumoursuppressor Reactive oxygen species made during inflammation provide a further potential mechanism for Tra2𝛽 upregulation MNK2 13b splice isoform makes 𝛽 kinase which can phosphorylate in cancer cells. Tra2 expression is activated in response to MNK2 [21, 25] EIF4E independent of MAP kinase reoxygenation of astrocytes following a period of oxygen activation deprivation and by ischaemia in rat brains [39]. Expression of 𝛽 Promotes oncogenic isoform of Tra2 in smooth muscle cells is similarly induced following S6K S6kinase which phosphorylates [21, 25] reoxygenation of hypoxic cells [38], and is upregulated in smallsubunitofribosome response to oxidative stress in human colorectal carcinoma Promotes production of cell line HCT116 [32]. Ischaemia has also been reported MCL-1/BCL- antiapoptotic mRNAs to result in [63–65] to induce cytoplasmic accumulation of Tra2𝛽 along with X/CASPASE9 cell survival accompanying changes in splice site use [40]. Tra2𝛽 translo- catesintothecytoplasmingastriccancercellsinresponseto cell stress induced by sodium arsenate [32], and changes in 𝛽 a proteasome-resistant form of HIPK2 protein is made after the nuclear concentration of Tra2 might have downstream SRSF3 depletion [28]. effects on the splicing inclusion of target exons. The increased levels of Tra2𝛽 observed in cancer cells mean that the TRA2B gene must be able to bypass the normal 4. Tra2𝛽 Is Amplified in Particular feedback expression control mechanisms which exist to keep Cancers and Is a Target of the Oncogenic Tra2𝛽 protein levels under tight control. An important feed- Transcription Factor ETS-1 back control mechanism uses an alternatively spliced “poison exon” in the TRA2B gene. Poison exons introduce premature The TRA2B gene which encodes Tra2𝛽 becomes amplified stop codons when they are spliced into mRNAs, preventing in several cancers (Figure 2) and particularly in cancers of translation of full-length proteins and often targeting mRNAs thelung,cervix,headandneck,ovary,stomach,anduterus for nonsense-mediated decay [41]. Poison exon splicing into [22, 23]. Upregulation of Tra2𝛽 protein expression has also the TRA2B mRNA is activated by binding of Tra2𝛽 itself. been observed in several cancers, including breast, cervical Splicing inclusion of this poison exon acts as a brake on and ovarian [29–31], and colon [32]. Tra2𝛽 upregulation is production of more Tra2𝛽 protein. The predicted outcome associated with invasive breast cancer [30], and medium to is that increased expression of Tra2𝛽 protein should lead to high Tra2𝛽 expression correlates with a poorer prognosis in increased TRA2B poison exon inclusion and so correspond- cervical cancer compared to patients with lower expression ingly less newly translated Tra2𝛽 protein through a negative levels [29]. feedback loop [42]. Tra2𝛽 protein expression has been demonstrated to be Similarly, the levels of SRSF1 and the other SR proteins important for cancer cell biology. Downregulation of Tra2𝛽 are thought to be normally autoregulated through poison inhibits cell growth of a gastric cancer cell line, measured exon inclusion [43]; so these other SR proteins must similarly by a corresponding decrease in BrdU incorporation which bypass these mechanisms in cancer cells to enable their monitors cells which have entered S phase [33]. Knockdown higher levels of expression to be established. of Tra2𝛽 in colon cancer cells reduced cell viability and increased the level of apoptosis monitored using a TUNEL assay and through measurement of levels of cleaved PARP 5. Tra2𝛽 Protein Regulates Splicing Patterns [32]. Which Are Important to Cancer Cells As well as TRA2B gene amplification, the expression levels of the ETS-1 transcription factor provide a possible How might upregulation of Tra2𝛽 affect the biology of cancer mechanism through which Tra2𝛽 might be upregulated in cells? Three Tra2𝛽-target exons have been identified in genes cancer cells. Regulated transcription of the TRA2B gene in knowntohaveimportantrolesincancercells(Table 2). For human colon cells is positively controlled by binding of two of these target exons, the actual regulated isoforms have the HSF1 and ETS-1 transcription factors to its promoter also been demonstrated in cancer cells. proximal region [32]. The ETS-1 protein is itself encoded bya Firstly, strong Tra2𝛽 binding to a cancer-associated protooncogene. ETS1 expression in metastatic breast cancer exon in the nuclear autoantigenic sperm protein (abbrevi- correlates with a poor prognosis [34, 35]andisassociated ated NASP) gene has been detected using HITS-CLIP of with an invasive phenotype [36]. Expression of both ETS-1 endogenous Tra2𝛽 protein in the mouse testis [14, 15]. This [35]andTra2𝛽 [37] might also be under control of estrogen, Tra2𝛽-target exon is abbreviated NASP-T.Whilstthesomatic International Journal of Cell Biology 5

Table 2: Known pro-oncogenic splicing targets of Tra2𝛽. chaperones which import histone H1 into the nucleus [49]. NASP protein isoforms also stably maintain the soluble pools Splicing target Possible role in cancer cells Reference of H3 and H4 histones needed for assembly of chromatin Affects cancer cell mobility and CD44 [30] at times of high replication activity and are part of the metastasis complexes which load these into chromatin [45]. The NASP Homeodomain- HIPK3 increases gene is critical for cell cycle progression in cultured cells and interacting kinase 3 phosphorylation of cJun and cell [57] for mouse embryogenesis [50]. (HipK3) proliferation Why might NASP protein be important for cancer cells? Histone chaperone important for NASP belongs to a network of genes important for cell efficient replication Nasp-T [15] survival [51], and NASP protein is a tumour-associated Implicated in DNA repair antigeninovariancancer[52]. NASP is highly expressed in processes Sphaseofthecellcycle[49], when chromatin needs to be reassembled after replication. Higher levels of NASP protein expression might be needed by cancer cells to enable their NASP splice isoform is expressed ubiquitously, the NASP-T higher rates of replication to be achieved. NASP protein splicing isoform has a much tighter anatomic distribution alsohasotherrolesrelatedtochromatinstability.NASP and its splicing is associated particularly with cancer cells and protein is phosphorylated by the ATM and ATR kinases in embryonic development. While most normal adult tissues do response to ionising radiation and implicated in the repair of not splice the NASP-T exons into their mRNAs, high levels of DNAdoublestrandbreaks[53]. One of the protein partners splicing inclusion are seen in the testis and to a lesser extent of NASP protein is the DNA repair protein Ku, and the the heart, gut, and ovary [15]. yeast homologue of NASP is present at double strand breaks Splicing inclusion of the NASP-T exon is strongly acti- suggesting an important role in DNA repair (reviewed in vatedintransfectedcellsinresponsetocoexpressionof [44]). Tra2𝛽,andNASP-T splicing also decreases in TRA2B knock- The second known splicing target of Tra2𝛽 with likely out mouse brains compared to wild type, confirming that the important functions in cancer cells is within the CD44 pre- NASP-T exon is a bona fide regulated target exon of Tra2𝛽 mRNA. CD44 encodes an important transmembrane protein [14, 15]. Tra2𝛽 is currently the only known splicing regulator partly displayed on the cell surface as the CD44 antigen of the NASP-T exon. The NASP-T exon is unusually long (a (Figure 3(b)).CD44proteinactsasareceptorforhyaluronic 975nucleotidelongcassetteexon,whilethetypicalsizefor acid and possibly other molecules and controls interactions a human exon is more like 120 nucleotides), with at least 37 with other cells, the extracellular matrix, and cellular motility Tra2𝛽 protein binding sites within its sequence, making a through modulation of intracellular signalling cascades [54]. very responsive target for Tra2𝛽 expression. Splicing inclu- The N- and C-termini of the CD44 protein are encoded sion of the NASP-T exon into the NASP mRNA introduces by constitutive exons, but the CD44 gene also contains an the coding information for an extra 375 amino acids into the internal block of 10 consecutive internal alternative exons encoded NASP protein (Figure 3). which are differentially regulated during development and The NASP protein has a strongly biased peptide sequence in cancer [55]. These alternative exons encode portions of which contains a high frequency of glutamic acid residues. the extracellular domain of the protein (Figure 3(b)). CD44 The negative charges of the glutamic acid residues facilitate variable exons show variant splicing inclusion in breast interactions with the positively charged histone partner cancer cells [30]. In particular, two CD44 internal variable proteins that NASP protein interacts with. NASP proteins exons, CD44v4 and CD44v5, increase their splicing inclusion also use tetratricopeptide repeats (TPRs) and histone binding in transfected HeLa cells in response to increased Tra2𝛽 motifs to facilitate interactions with protein partners includ- protein expression [30], suggesting that Tra2𝛽 might also ing histones [44]. Both the somatic (sNASP) and NASP- increase their inclusion in breast tumours with elevated T isoforms of the NASP protein contain the same TPRs Tra2𝛽 expression. Although expression of variant CD44 involved in protein-protein interactions and seem to be func- exons has historically been associated with cancer metastasis, tionally interchangeable in cells [45]. However, the longer the picture regarding CD44 alternative splicing in cancer is NASP-T protein isoform has an additional histone binding complex. Very recent data suggest that the standard isoform motif and a longer stretch of the glutamic-acid-enriched of CD44 mRNA (without splicing inclusion of its variable sequence, suggesting that it might more efficiently interact exons)mightinfactplayakeyroleinmetastaticbreastcancer, with histones (Figure 3(a)). The NASP-T peptide cassette particularly in enabling an epithelial-mesenchyme transition also adds a number of potentially phosphorylated serine and of breast cancer cells [56]. threonine residues to the NASP protein [44, 46]. Splicing The third known Tra2𝛽-target exon which might be inclusion of the NASP-T exon is likely to be important in potentially relevant in cancer cells is in the HIPK3 gene, which cancer cells. The specific siRNA-mediated downregulation of encodes a serine/threonine kinase involved in transcriptional NASP mRNAs containing the NASP-T exon leads to a block in regulation and negative control of apoptosis. High cellular proliferation and increased levels of apoptosis in cancer cells levels of Tra2𝛽 stimulate splicing inclusion of a poison [47, 48]. exon called HIPK3-T into the HIPK3 mRNA [57]. Normal Isoforms of the NASP protein with and without the HIPK3 protein is concentrated in subnuclear structures called peptide cassette inserted by the NASP-T exon are molecular promyelocytic leukemia bodies (PML bodies). The shorter 6 International Journal of Cell Biology

NASP protein domain structure

N C Amino Amino acid 1 acid 788 Peptide insert from Tra2𝛽-responsive NASP-T cassette exon amino acids 138–476 Protein domains TPR motif Glutamic-acid-rich region of protein (negative charge)

Histone binding motif Nuclear localisation sequence (NLS)

(a) CD44 protein domain structure CD44 antigen (amino acids 21–742) on cell surface ∗ ∗∗ ∗ N C Amino Amino acid 1 Peptide encoded by acid 742 Tra2𝛽-responsive exons v4 and v5 Block of alternatively spliced exons amino acids 224–614 Protein domains

Signal peptide Transmembrane span

Extracellular domain Cytoplasmic domain ∗ Hyaluronan binding site (b)

Figure 3: Protein domain architecture of known Tra2𝛽 splicing targets which are expressed in cancer cells. (a) Modular structure of NASP protein assembled from the UniProt database (http://www.uniprot.org/uniprot/P49321)[46], showing the position of the peptide insert encoded by the Tra2𝛽-target exon NASP-T. (b) Modular structure of CD44 protein assembled using information from the UniProt database (http://www.uniprot.org/uniprot/P16070#P16070-6)[46], showing the position of peptide sequences encoded by the Tra2𝛽-target exons CD44 v4 and v5. The CD44 antigen is displayed on the cell surface, and the protein is anchored on the cell surface by a single trans-membrane domain. Alternative isoforms are made through alternative splicing of 10 exons out of 19 encoding amino acids in the extracellular domain and also 2 exons which encode peptide sequence in the cytoplasmic domain. The two exons reported CD44 v4 and v5 exons correspond to amino acids 386–428 and 429–472, respectively, in the encoded protein. The protein domain structures are not drawn to scale.

HIPK3 protein isoform made under control of Tra2𝛽 fails Of these retrogene-derived proteins, one called HNRNP G- to localise in PML bodies and lacks regions of the protein T is both highly conserved in mammals and specifically predictedtobindtheandrogenreceptor,homeodomains,Fas, expressed in meiosis. The interaction between 𝛽Tra2 and and p53 [57]. HIPK3-T is not confirmed as a splicing target hnRNP G family members likely buffers the splicing activity of Tra2𝛽 in cancer, since splicing of the HIPK3-T exon has of Tra2𝛽 [58, 59], although they might also coregulate some only been observed thus far in human testis and has not been target exons [60]. Expression of the RBMY protein has been directly reported from cancer cells [57]. directly implicated in liver cancer biology, where it may contribute to the male specificity of this cancer61 [ ]. RBMY proteinalsointeractswithSRSF3protein[62]. 6. Tra2𝛽 Is Involved in Protein Interaction Networks with Partner Proteins Involved 7. Summary in Cancer The splicing regulator Tra2𝛽 is upregulated in some human Some of the proteins which are known to interact either cancers. Possible mechanisms for this upregulation include directly or indirectly with Tra2𝛽 have themselves been changes in oncogenic transcription factor expression and implicated with roles in cancer cells. Tra2𝛽 directly interacts oxygen free radical concentrations in neoplastic tissue, both with members of the hnRNP G family of proteins which of which affect TRA2B gene expression (Figure 4). We do not includes the prototypic member hnRNP G (encoded by currently know whether the TRA2B gene can function as an the RBMX gene located on the X chromosome); RBMY oncogene in its own right until experiments to test transfor- protein (which is encoded by a multigene family on the Y mation of NIH3T3 cells are performed or the behaviour of chromosome); and a number of retrogene-derived proteins. such transformed cells in nude mice is tested. However, we International Journal of Cell Biology 7

Changes in the cellular environment in cancer cells [7] S. Bonnal, L. Vigevani, and J. Valcarcel, “The spliceosome Increasing intracellular Increasing intracellular as a target of novel antitumour drugs,” Nature Reviews Drug concentration of free concentration of ETS-1 Discovery,vol.11,pp.847–859,2012. oxygen ions [8] J. C. Long and J. F. Caceres, “The SR protein family of splicing factors: master regulators of gene expression,” Biochemical Journal,vol.417,no.1,pp.15–27,2009. Increases in Tra2𝛽 expression [9]P.J.ShepardandK.J.Hertel,“TheSRproteinfamily,”Genome change splicing profile of cells Biology, vol. 10, no. 10, p. 242, 2009. [10] Z. Zhou and X. D. Fu, “Regulation of splicing by SR proteins and SR protein-specific kinases,” Chromosoma,vol.122,no.3, pp.191–207,2013. [11] R. Tacke, M. Tohyama, S. Ogawa, and J. L. Manley, “Human Tra2 proteins are sequence-specific activators of pre-mRNA Downstream protein isoform changes in cancer cells NASP CD44 HIPK3 splicing,” Cell, vol. 93, no. 1, pp. 139–148, 1998. Changes efficiency Affects cell movement Decreases apoptosis [12] A. R. Krainer and T. Maniatis, “Multiple factors including the of chromatin packaging and metastasis small nuclear ribonucleoproteins U1 and U2 are necessary for Pre-mRNA splicing in vitro,” Cell,vol.42,no.3,pp.725–736, Figure 4: Hypothetical model suggesting how changes in the 1985. 𝛽 cellular environment may influence the expression of Tra2 and lead [13]A.M.Zahler,K.M.Neugebauer,W.S.Lane,andM.B.Roth, to downstream changes in mRNA splice isoform production. “Distinct functions of SR proteins in alternative pre-mRNA splicing,” Science,vol.260,no.5105,pp.219–222,1993. [14] D. J. Elliott, A. Best, C. Dalgliesh, I. Ehrmann, and S. 𝛽 Grellscheid, “How does Tra2beta protein regulate tissue-specific do know that some of the known splicing targets of Tra2 RNA splicing?” Biochemical Society Transactions,vol.40,pp. identified in normal tissues are important for cancer cell 784–788, 2012. biology and are particularly implicated in cell division and [15]S.Grellscheid,C.Dalgliesh,M.Storbecketal.,“Identification 𝛽 motility. Tra2 is essential during embryonic development, of evolutionarily conserved exons as regulated targets for the and many embryonic developmental pathways involved in splicing activator Tra2𝛽 in development,” PLoS Genetics,vol.7, cellgrowthandmotilitywhichareturnedoffinadultcells no. 12, Article ID e1002390, 2011. often become reactivated in cancer cells. Future analysis of [16] A. Clery,´ S. Jayne, N. Benderska, C. Dominguez, S. Stamm, the role of Tra2𝛽 in cancer cells will require the detailed and F. H.-T. Allain, “Molecular basis of purine-rich RNA identification of its endogenous splicing targets in cancer cells recognition by the human SR-like protein Tra2-𝛽1,” Nature and the elucidation of their physiological roles. Structural and Molecular Biology,vol.18,no.4,pp.443–451, 2011. [17] K.Tsuda,T.Someya,K.Kuwasakoetal.,“Structuralbasisforthe Acknowledgments dual RNA-recognition modes of human Tra2-𝛽 RRM,” Nucleic Acids Research, vol. 39, no. 4, pp. 1538–1553, 2011. This work was funded by the Breast Cancer Campaign, [18] T. Novoyatleva, B. Heinrich, Y. Tang et al., “Protein phosphatase the Wellcome Trust (Grant numbers WT080368MA and 1 binds to the RNA recognition motif of several splicing fac- WT089225/Z/09/Z), and the BBSRC (grant numbers tors and regulates alternative pre-mRNA processing,” Human BB/D013917/1 and BB/I006923/1). Molecular Genetics,vol.17,no.1,pp.52–70,2008. [19] Y. Mende, M. Jakubik, M. Riessland et al., “Deficiency of the splicing factor Sfrs10 results in early embryonic lethality in mice References and has no impact on full-length SMN /Smn splicing,” Human Molecular Genetics,vol.19,no.11,ArticleIDddq094,pp.2154– [1] D. Hanahan and R. A. Weinberg, “Hallmarks of cancer: the next 2167, 2010. generation,” Cell,vol.144,no.5,pp.646–674,2011. [20] L. R. Meyer, A. S. Zweig, A. S. Hinrichs et al., “The UCSC [2] P.Rajan,D.J.Elliott,C.N.Robson,andH.Y.Leung,“Alternative Genome Browser database: extensions and updates 2013,” splicing and biological heterogeneity in prostate cancer,” Nature Nucleic Acids Research,vol.41,pp.D64–D69,2013. Reviews Urology,vol.6,no.8,pp.454–460,2009. [21]R.Karni,E.DeStanchina,S.W.Lowe,R.Sinha,D.Mu,andA. [3] J. P. Venables, “Aberrant and alternative splicing in cancer,” R. Krainer, “The gene encoding the splicing factor SF2/ASF isa Cancer Research,vol.64,no.21,pp.7647–7654,2004. proto-oncogene,” Nature Structural and Molecular Biology,vol. 14,no.3,pp.185–193,2007. [4] J. P. Venables, “Unbalanced alternative splicing and its significance in cancer,” BioEssays,vol.28,no.4,pp.378–386,2006. [22] E.Cerami,J.Gao,U.Dogrusozetal.,“ThecBiocancergenomics portal: an open platform for exploring multidimensional cancer [5]V.Quesada,L.Conde,N.Villamoretal.,“Exomesequencing genomics data,” Cancer Discovery,vol.2,pp.401–404,2012. identifies recurrent mutations of the splicing factor SF3B1 gene [23] J. Gao, B. A. Aksoy, U. Dogrusoz et al., “Integrative analysis in chronic lymphocytic leukemia,” Nature Genetics,vol.44,no. of complex cancer genomics and clinical profiles using the 1,pp.47–52,2012. cBioPortal,” Science Signaling, vol. 6, no. 269, p. pl1, 2013. [6]A.R.Grosso,S.Martins,andM.Carmo-Fonseca,“Theemerg- [24] S. Das, O. Anczukow, M. Akerman, and A. R. Krainer, “Onco- ing role of splicing factors in cancer,” EMBO Reports,vol.9,no. genic splicing factor SRSF1 is a critical transcriptional target of 11, pp. 1087–1093, 2008. MYC,” Cell Reports, vol. 1, pp. 110–117, 2012. 8 International Journal of Cell Biology

[25] O. Anczukow,A.Z.Rosenberg,M.Akermanetal.,“Thesplicing´ [42] P. Stoilov, R. Dauod, O. Nayler, and S. Stamm, “Human tra2- factor SRSF1 regulates apoptosis and proliferation to promote beta1 autoregulates its protein concentration by infuencing mammary epithelial cell transformation,” Nature Structural and alternative splicing of its pre-mRNA,” Human Molecular Genet- Molecular Biology,vol.19,no.2,pp.220–228,2012. ics,vol.13,no.5,pp.509–524,2004. [26]R.Jia,C.Li,J.P.McCoy,C.-X.Deng,andZ.-M.Zheng,“SRp20 [43]L.F.Lareau,M.Inada,R.E.Green,J.C.Wengrod,andS.E. is a proto-oncogene critical for cell proliferation and tumor Brenner, “Unproductive splicing of SR genes associated with induction and maintenance,” International Journal of Biological highly conserved and ultraconserved DNA elements,” Nature, Sciences,vol.6,no.7,pp.806–826,2010. vol. 446, no. 7138, pp. 926–929, 2007. [27]X.He,A.D.Arslan,M.D.Pooletal.,“Knockdownofsplicing [44] R. M. Finn, K. Ellard, J. M. Eirin-Lopez, and J. Ausio, “Verte- factor SRp20 causes apoptosis in ovarian cancer cells and its brate nucleoplasmin and NASP: egg histone storage proteins expression is associated with malignancy of epithelial ovarian with multiple chaperone activities,” The FASEB Journal,vol.26, cancer,” Oncogene,vol.30,no.3,pp.356–365,2011. pp. 4788–4804, 2012. [28] K. Kurokawa, Y. Akaike, K. Masuda et al., “Downregulation of [45] A. J. L. Cook, Z. A. Gurard-Levin, I. Vassias, and G. Almouzni, serine/arginine-rich splicing factor 3 induces G1 cell cycle arrest “A specific function for the histone chaperone NASP to fine- and apoptosis in colon cancer cells,” Oncogene,2013. tune a reservoir of soluble H3-H4 in the histone supply chain,” [29] B. Gabriel, A. Z. Hausen, J. Bouda et al., “Significance of nuclear Molecular Cell,vol.44,no.6,pp.918–927,2011. hTra2-beta1 expression in cervical cancer,” Acta Obstetricia et [46] UniProt Consortium, “Reorganizing the protein space at the Gynecologica Scandinavica,vol.88,no.2,pp.216–221,2009. Universal Protein Resource (UniProt),” Nucleic Acids Research, [30]D.O.Watermann,Y.Tang,A.Z.Hausen,M.Jager,¨ S. Stamm, vol. 40, pp. D71–D75, 2012. and E. Stickeler, “Splicing factor Tra2-𝛽1 is specifically induced [47] O. M. Alekseev, R. T. Richardson, J. K. Tsuruta, and M. G. in breast cancer and regulates alternative splicing of the CD44 O’Rand, “Depletion of the histone chaperone tNASP inhibits gene,” Cancer Research,vol.66,no.9,pp.4774–4780,2006. proliferation and induces apoptosis in prostate cancer PC-3 [31] D.-C. Fischer, K. Noack, I. B. Runnebaum et al., “Expression cells,” Reproductive Biology and Endocrinology,vol.9,p.50,2011. of splicing factors in human ovarian cancer,” Oncology Reports, [48] W. Ma, S. Xie, M. Ni et al., “MicroRNA-29a inhibited epididy- vol. 11, no. 5, pp. 1085–1090, 2004. mal epithelial cell proliferation by targeting nuclear autoanti- [32] K. Kajita, Y. Kuwano, N. Kitamura et al., “Ets1 and heat shock genic sperm protein (NASP),” JournalofBiologicalChemistry, factor 1 regulate transcription of the Transformer 2𝛽 gene in vol. 287, no. 13, pp. 10189–10199, 2012. human colon cancer cells,” Journal of Gastroenterology,2013. [49]R.T.Richardson,I.N.Batova,E.E.Widgrenetal.,“Character- [33] K. Takeo, T. Kawai, K. Nishida et al., “Oxidative stress-induced ization of the histone H1-binding protein, NASP, as a cell cycle- alternative splicing of transformer 2𝛽 (SFRS10) and CD44 regulated somatic protein,” JournalofBiologicalChemistry,vol. pre-mRNAs in gastric epithelial cells,” American Journal of 275, no. 39, pp. 30378–30386, 2000. Physiology,vol.297,no.2,pp.C330–C338,2009. [50] R. T. Richardson, O. M. Alekseev, G. Grossman et al., “Nuclear [34] D. W. Lincoln II and K. Bove, “The transcription factor Ets-1 in autoantigenic sperm protein (NASP), a linker histone chaper- breast cancer,” Frontiers in Bioscience,vol.10,pp.506–511,2005. one that is required for cell proliferation,” JournalofBiological [35] D. W. Lincoln II, P. G. Phillips, and K. Bove, “Estrogen- Chemistry,vol.281,no.30,pp.21526–21534,2006. induced Ets-1 promotes capillary formation in an in vitro tumor [51] O. M. Alekseev, R. T. Richardson, O. Alekseev, and M. G. angiogenesis model,” Breast Cancer Research and Treatment,vol. O’Rand, “Analysis of gene expression profiles in HeLa cells in 78,no.2,pp.167–178,2003. response to overexpression or siRNA-mediated depletion of [36] J. Dittmer, “The biology of the Ets1 proto-oncogene,” Molecular NASP,” Reproductive Biology and Endocrinology,vol.7,article Cancer,vol.2,article29,2003. 45, 2009. [37]X.Zhang,A.N.Moor,K.A.Merkler,Q.Liu,andM.P.McLean, [52] R. Ali-Fehmi, M. Chatterjee, A. Ionan et al., “Analysis of the “Regulation of alternative splicing of liver scavenger receptor expression of human tumor antigens in ovarian cancer tissues,” class B gene by estrogen and the involved regulatory splicing Cancer Biomarkers,vol.6,no.1,pp.33–48,2009. factors,” Endocrinology,vol.148,no.11,pp.5295–5304,2007. [53] S. Matsuoka, B. A. Ballif, A. Smogorzewska et al., “ATM [38] Y.Tsukamoto, N. Matsuo, K. Ozawa et al., “Expression of a novel and ATR substrate analysis reveals extensive protein networks RNA-splicing factor, RA301/Tra2𝛽,invascularlesionsandits responsive to DNA damage,” Science,vol.316,no.5828,pp. role in smooth muscle cell proliferation,” American Journal of 1160–1166, 2007. Pathology,vol.158,no.5,pp.1685–1694,2001. [54] H. Ponta, L. Sherman, and P.A. Herrlich, “CD44: from adhesion [39] N. Matsuo, S. Ogawa, Y. Imai et al., “Cloning of a novel RNA molecules to signalling regulators,” Nature Reviews Molecular binding polypeptide (RA301) induced by hypoxia/reoxygena- Cell Biology,vol.4,no.1,pp.33–45,2003. tion,” Journal of Biological Chemistry,vol.270,no.47,pp.28216– [55] G.R.Screaton,M.V.Bell,D.G.Jackson,F.B.Cornelis,U.Gerth, 28222, 1995. and J. I. Bell, “Genomic structure of DNA encoding the lym- [40] R. Daoud, G. Mies, A. Smialowska, L. Olah,´ K.-A. Hossmann, phocyte homing receptor CD44 reveals at least 12 alternatively and S. Stamm, “Ischemia induces a translocation of the splicing spliced exons,” Proceedings of the National Academy of Sciences factor tra2-𝛽1 and changes alternative splicing patterns in the of the United States of America,vol.89,no.24,pp.12160–12164, brain,” Journal of Neuroscience,vol.22,no.14,pp.5889–5899, 1992. 2002. [56] R. L. Brown, L. M. Reinke, M. S. Damerow et al., “CD44 [41] N. J. McGlincy and C. W.J. Smith, “Alternative splicing resulting splice isoform switching in human and mouse epithelium in nonsense-mediated mRNA decay: what is the meaning of is essential for epithelial-mesenchymal transition and breast nonsense?” Trends in Biochemical Sciences,vol.33,no.8,pp. cancer progression,” Journal of Clinical Investigation,vol.121,no. 385–393, 2008. 3,pp.1064–1074,2011. International Journal of Cell Biology 9

[57] J. P.Venables, C. F.Bourgeois, C. Dalgliesh, L. Kister, J. Stevenin, and D. J. Elliott, “Up-regulation of the ubiquitous alternative splicing factor Tra2𝛽 causes inclusion of a germ cell-specific exon,” Human Molecular Genetics, vol. 14, no. 16, pp. 2289–2303, 2005. [58] Y. Liu, C. F. Bourgeois, S. Pang et al., “The germ cell nuclear proteins hnRNP G-T and RBMY activate a testis-specific exon,” PLoS Genetics, vol. 5, no. 11, Article ID e1000707, 2009. [59]M.T.Nasim,T.K.Chernova,H.M.Chowdhury,B.-G.Yue,and I. C. Eperon, “HnRNP G and Tra2𝛽: opposite effects on splicing matched by antagonism in RNA binding,” Human Molecular Genetics,vol.12,no.11,pp.1337–1348,2003. [60] Y. Hofmann, C. L. Lorson, S. Stamm, E. J. Androphy, and B. Wirth, “Htra2-𝛽1 stimulates an exonic splicing enhancer and can restore full-length SMN expression to survival motor neuron 2 (SMN2),” Proceedings of the National Academy of Sciences of the United States of America, vol. 97, no. 17, pp. 9618– 9623, 2000. [61] D.-J. Tsuei, P.-H. Lee, H.-Y. Peng et al., “Male germ cell-specific RNA binding protein RBMY: a new oncogene explaining male predominance in liver cancer,” PLoS ONE,vol.6,no.11,Article ID e26948, 2011. [62] D. J. Elliott, C. F. Bourgeois, A. Klink, J. Stevenin,´ and H. J. Cooke, “Amammalian germ cell-specific RNA-binding protein interacts with ubiquitously expressed proteins involved in splice site selection,” Proceedings of the National Academy of Sciences of the United States of America, vol. 97, no. 11, pp. 5717–5722, 2000. [63] M. J. Moore, Q. Wang, C. J. Kennedy, and P. A. Silver, “An alternative splicing network links cell-cycle control to apoptosis,” Cell,vol.142,no.4,pp.625–636,2010. [64] M. P. Paronetto, T. Achsel, A. Massiello, C. E. Chalfant, and C. Sette, “The RNA-binding protein Sam68 modulates the alternative splicing of Bcl-x,” Journal of Cell Biology,vol.176,no. 7, pp. 929–939, 2007. [65] H. L. Gautrey and A. J. Tyson-Capper, “Regulation of Mcl-1 by SRSF1 and SRSF5 in cancer cells,” PLoS ONE,vol.7,ArticleID e51497, 2012. Hindawi Publishing Corporation International Journal of Cell Biology Volume 2013, Article ID 973584, 8 pages http://dx.doi.org/10.1155/2013/973584

Review Article Clinical Significance of HER-2 Splice Variants in Breast Cancer Progression and Drug Resistance

Claire Jackson,1 David Browell,2 Hannah Gautrey,1 and Alison Tyson-Capper1

1 Institute of Cellular Medicine, Faculty of Medical Sciences, Newcastle University, Newcastle upon Tyne NE2 4HH, UK 2 QueenElizabethHospital,Gateshead,TyneandWearNE96SX,UK

Correspondence should be addressed to Alison Tyson-Capper; [email protected]

Received 6 June 2013; Accepted 13 June 2013

Academic Editor: Claudia Ghigna

Copyright © 2013 Claire Jackson et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Overexpression of human epidermal growth factor receptor (HER-2) occurs in 20–30% of breast cancers and confers survival and proliferative advantages on the tumour cells making HER-2 an ideal therapeutic target for drugs like Herceptin. Continued delineation of tumour biology has identified splice variants of HER-2, with contrasting roles in tumour cell biology. For example, thesplicevariantΔ16HER-2 (results from exon 16 skipping) increases transformation of cancer cells and is associated with treatment resistance; conversely, Herstatin (results from intron 8 retention) and p100 (results from intron 15 retention) inhibit tumour cell proliferation. This review focuses on the potential clinical implications of the expression and coexistence of HER-2 splice variants in cancer cells in relation to breast cancer progression and drug resistance. “Individualised” strategies currently guide breast cancer management; in accordance, HER-2 splice variants may prove valuable as future prognostic and predictive factors, as well as potential therapeutic targets.

1. Introduction and the use of adjuvant therapy to eliminate “micrometas- tases.” This approach improved survival, reduced the risk of Breast cancer is a heterogeneous disease comprising subtypes recurrence, and minimised the impact of treatment on quality of varied morphology, prognostic profiles, and clinical out- of life thus emphasising a need for more directed treatment comes [1, 2]. Tumours arise from malignant transformation strategies [4]. of hyperplasic epithelia within the breast [3], and numer- Consequently, there has been a subsequent shift in more ous mutagenic changes contribute to the transformation recent years to “individualized” treatment with better ther- process which abnormally alters the cellular environment. apeutic targeting. The advent of the humanised monoclonal Atypical hyperplasic cells may progress to carcinoma in situ, antibody trastuzumab (commonly referred to as Herceptin) categorised as ductal carcinoma in situ (DCIS) or lobular which targets human epidermal growth factor-2 (HER- carcinoma in situ (LCIS) [3](Figure 1). These terms denote 2) transformed management of breast cancer patients [5]. malignant cells restricted to ducts or acini of lobules. Car- Patients whose tumours are shown to overexpress HER-2 cinoma becomes invasive when atypical cells penetrate the now undergo more rigorous treatment, with Herceptin and basement membrane and spread into the surrounding stroma chemotherapy. This modernised approach of “targeted” treat- [3](Figure 1). Cancer cells then have the potential to spread ment now guides cancer management with attempts to tailor to surrounding skin or muscles or to metastasise to axillary therapeutics to specific tumours4 [ ]. lymph nodes or distant sites such as bone, liver, and brain where new tumours may form [3]. 2. HER-2: Structure and Function In recent decades, there has been a paradigm shift from increasingly extensive and invasive surgery to “cure” and HER-2 is a 185 kDa transmembrane cell surface receptor of prevent relapse to conservation surgery with lower morbidity the human epidermal growth factor (EGF) family [6]. There 2 International Journal of Cell Biology

(a) (b)

Figure 1: Histological images of breast carcinoma. Images of (a) ductal carcinoma no special type (NST), (b) ductal carcinoma in situ (DCIS), and (c) lobular carcinoma. (d) HercepTest positive staining: immunocytochemical staining indicates HER-2 overexpression in invasive breast cancer (Images courtesy of Dr. D. Hemming, Queen Elizabeth Hospital, Gateshead).

are four receptor members of this family: HER-1 (EGFR, 3. HER-2: Insights into Tumour Biology ErbB-1), HER-2 (ErbB-2), HER-3 (ErbB-3), and HER-4 (ErbB-4). EGF receptors have a highly conserved extracellu- Each subtype of invasive breast cancer is associated with cer- lar domain, a transmembrane domain, and an intracellular tain clinical characteristics and treatment options. Although domain with tyrosine kinase activity [7](Figure 2). Ligand- the umbrella term of “breast cancer” remains, the discovery receptor binding induces conformational changes and recep- of new biomarkers and gene expression profiling prompted tor dimerisation via interaction at both extracellular cysteine- a move to consider subtypes of breast cancer as different rich regions [7, 8]. This results in autophosphorylation and diseases within their own right [11]. HER-2-positive breast kinase activation [8]. EGF receptor signalling has important cancer is typically more aggressive with a poorer prognostic roles in cell proliferation, differentiation, and survival9 [ ] outlook [12]. HER-2 is routinely measured in clinical prac- (Figure 2). tice, and patients whose tumours score 3+ on HercepTest Thirteen ligands interact with EGF receptors. HER-1 and immunocytochemical staining (Figure 1(d))andtestpositive HER-4 may actively homodimerise. HER-2 and HER-3 are foramplificationoftheHER-2 gene using fluorescence in nonautonomous as HER-2 has no known ligand, and HER-3 situ hybridization (FISH) will be offered treatment with lacks tyrosine kinase activity [8]. HER-2 and HER-3 therefore Herceptin in combination with chemotherapy. form heterodimers with other EGF receptors to promote HER-2 has been acknowledged as a protooncogene since signal induction (Figure 2). amutatedform,theNEU oncogene, was isolated using cell HER-2 was first identified in 1984 by Schechter et al. [6] transformation studies in the rat that used tumour DNA and has since been recognised as the “preferred” dimerisation [13]. Moreover, amplification of the HER-2 gene occurs in partner [10]. Whilst it lacks the “typical” ligand-binding a number of different cancers and is particularly prevalent structure, HER-2 sustains an active conformation acting as in invasive carcinoma of the breast (Figure 3)[14–16]. HER- a potent coreceptor for other EGF receptors [8]. Prolonged 2 protein is overexpressed in many human cancers and dimer interaction consequently sustains downstream sur- associated with 20–30% of breast cancers [7, 17]. High levels vival and proliferative signalling [10]. of the receptor result in enhancement of oncogenic signalling International Journal of Cell Biology 3

EGF receptors HER-2 receptor

L I C I Extracellular domain L II C II

Transmembrane domain

Tyrosine Intracellular kinase domain

(a) Receptor structure HER-2 receptor EGF receptors

P P

PI3K RAS RAF

AKT MEK Survival Apoptosis MAPK

Proliferation (b) Activation and signalling

Figure 2: Schematic of HER-2 structure, activation, and signalling. (a) HER-2 is a single transmembrane cell surface receptor with extracellular, transmembrane, and intracellular regions. The extracellular region comprises of two ligand-binding domains (L I and L II) and two cysteine-rich domains (C I and C II) [8]. Intracellularly, HER-2 receptors have intrinsic tyrosine kinase activity (TK). (b) HER-2 does not bind ligands but is activated by forming heterodimers with other ErbB receptors via interaction at the cysteine-rich domains. This results in autophosphorylation of the tyrosine kinase domains and induction of downstream signalling. Normal signalling includes stimulation of the PI3K/AKT pathway which induces survival mechanisms and inhibits apoptosis, whilst the RAS/RAF/MEK/MAPK pathway stimulates cell proliferation [8]. pathways [12]. Consequently, HER-2-positive tumours are responders regress within 6 months [12, 22]. There are also associated with increased metastatic potential, poor progno- some HER-2-positive patients who relapse early, and their sis, and recurrence [18, 19]. more common pattern of metastatic disease involves spread As HER-2 is expressed at much higher levels in certain to the bone, liver, and lungs, whilst there are also long- tumours (Figure 3)thaninnormaltissueandplaysakeyrole term responders who can relapse with the less commonly in mitogenic and antiapoptotic signalling [7, 12, 20], it was seen metastases to the brain. Further exploration of HER- recognised as an ideal target for anticancer drugs. Current 2 biology, signaling, and resistance mechanisms is therefore approved therapies include the aforementioned monoclonal essential to develop and implement new strategies of thera- antibody Herceptin and tyrosine kinase inhibitor lapatinib. peutic intervention. Such agents fostered improved survival rates with five-year survival now at 84% for women in England [21]. However, 4. HER-2 Splice Variants and Cancer Biology whilst these drugs improve breast cancer treatment, they are stillnotfullyunderstoodandacontinuingchallenge[5, 11, 12]. Many cancer-related changes in alternative splicing have For example, it is still unclear why some patients do not been identified to distinguish splicing patterns in “normal” respond to Herceptin as a single agent, and also why initial breast compared to cancer samples [23–26]. Cancer-specific 4 International Journal of Cell Biology

10 Altered (%) Altered

0 Sarcoma Glioblastoma Bladder cancer Bladder Lung adenocarcinoma Lung Acute myeloid leukemia myeloid Acute leukemia myeloid Acute Prostate adenocarcinoma Prostate Breast invasive carcinoma invasive Breast carcinoma invasive Breast Stomach adenocarcinoma Stomach Skin cutaneous melanoma Bladder urothelial carcinoma urothelial Bladder Cancer cell line encyclopedia Lung squamous cell carcinoma squamous Lung cell carcinoma squamous Lung Uterine endometroid carcinoma endometroid Uterine carcinoma endometroid Uterine Kidney renal clear cell carcinoma clear Kidney renal cell carcinoma clear Kidney renal Cervical cell carcinoma squamous Colon and rectum adenocarcinoma rectum and Colon adenocarcinoma rectum and Colon Ovarian serous cystadenocarcinomaOvarian serous cystadenocarcinomaOvarian serous Kidney renal capillary cell carcinoma Head and neck squamous cell carcinoma squamous neck and Head

Multiple alterations Deletion Mutation Amplification

Figure 3: Patterns of HER-2 in different cancers. Bar chart shows genetic changes in a wide range of different tumours and cancer types, including mutation, deletion, and amplification. Note that amplification is particularly prevalent in invasive carcinoma of the breast. Data was generated using the cBIO Cancer Genomics Portal [14, 15].

Table 1: HER-2 spliced variants and their role in cancer.

HER-2 splice variant Alternatively spliced event Function in cancer Reference Exon 16 skipped Able to homodimerise to activate oncogenic Δ16HER2 pathways. [32, 33, 36] Increased transforming capacity. 15 16 17 15 16 Associated with treatment resistance. Intron 15 retained p100 Truncated inhibitor of tumour cell proliferation [34] and oncogenic signalling. 15 16 17 15 16 Intron 8 retention Inhibitor of HER-2 which interferes with Herstatin dimerisation and autophosphorylation. [30, 31] Inhibits growth of transformed cells which 8 8 9109 overexpress HER2.

events can result in proteins with “procancer” properties investigation centred on the different variants of HER-2 that which may promote malignant transformation or confer a can be produced by alternative splicing [30–33]. survival advantage on cancer cells, such as resistance to To date, three naturally occurring HER-2 spliced treatment [27–29]. In recent years, focus has been directed variants in breast cancer have been reported (Table 1), at the level of the transcriptome with one area of continued namely, Δ16HER-2, Herstatin, and p100. As new therapeutic International Journal of Cell Biology 5 strategies are devised in efforts to tackle current problems that retain intron 8 [42]. This secreted HER-2 variant, like with treatment resistance, attention has been directed to p100, contains only the extracellular domain of the full-length further unravel the impact of HER-2 with particular focus protein and has a novel C-terminus of 79 amino acids [42]. on these spliced variants [30]. Studies have investigated the Several lines of evidence demonstrate that Herstatin can act transforming, oncogenic and drug-resistant activities of as an inhibitor of full-length HER-2, since it is able to interfere these isoforms [30, 33–35]. with dimerization, decrease tyrosine phosphorylation, and consequently inhibit the growth of transformed cells which 5. Δ16HER-2 overexpress HER-2 [43]. Interestingly, the autoinhibitory properties of Herstatin can also impede HER-2 activity Δ 16HER-2 arises from the in-frame deletion of exon 16; a by preventing transactivation of its hetesrodimeric partner 48 bp cassette exon which encodes a small region of the HER-3; Herstatin does this by specifically disrupting HER- extracellular domain of HER-2 [32]. Resultant loss of cysteine 2/HER-3 and also HER-2/EGFR dimer phosphorylation [31, residues in the extracellular domain of HER-2 induces a 43]. Since Herstatin has been perceived to be a “protec- conformational change, promoting homodimerisation via tive” HER-2 variant rather than an “oncogenic” protein, its intermolecular disulfide bonds32 [ , 33, 36]. Castiglioni et al. expression profile has been assessed in normal versus tumour Δ propose a causal role of 16HER-2 in cancer development tissues [35]. Findings from this study, not surprisingly, show suggesting that malignant transformation occurs once the that Herstatin levels are significantly higher in noncancerous Δ proportion of 16HER-2 expressed reaches a specific thresh- breast cells compared to carcinoma cells. old [32]. Conversely, wild-type HER-2, whilst relevant, is not considered sufficient to induce transformation [32, 37]. 8. Clinical Implications of Δ In addition, numerous studies have linked 16HER-2 with HER-2 Splice Variants resistance to trastuzumab advocating the use of tyrosine kinase inhibitors as an alternative [32, 33]. Prior to the advent of specific markers, such as HER-2, and Δ16HER-2 appears to constitute a more aggressive vari- drugs like Herceptin, cancer management was directed by antcomparedtowild-typeHER-2.Notonlyhasitbeen tumour grade and status alone. Treatments were not specif- purported to be important in malignant transformation, but ically targeted, for example, the use of chemotherapeutic research also suggests a role in disease progression. Mitra et agents which target cell division. Today, finding ways to al.reportedthat89%ofpatientswithHER-2-positivebreast further exploit tumour biology is central to overcoming chal- tumours, in whom disease progresses to local lymph nodes, lenges to current diagnostics and managing as well as devel- expressed Δ16HER2 [33]. This suggests that patients express- oping new “individualized” interventions with improved ing Δ16HER-2 may benefit from more aggressive therapeutic stratification to treatment. intervention. Δ16HER-2 has also been implicated in resistance of HER- 2-positive tumours to anti-HER-2 therapies [33]. Therefore, 6. P100 measurement of this variant may also be especially infor- mative in predicting response to treatment with anti-HER-2 Scott et al. first described an HER-2 mRNA variant encoding therapies. a protein constituting only the extracellular domain of the This is somewhat intriguing considering the aggres- full-length protein [38]. Termed p100, this splice variant sive nature of HER-2-positive tumours, and p100 has been interferes with the oncogenic activity of wild-type HER-2 reported to decrease with increasingly aggressive tumours and arises via an in-frame stop codon as a result of intron [39]. In view of this, it would be of value to compare the pro- 15 retention [34].Studiedincelllinesandtumoursderived portions of p100 (and Herstatin mRNA) and protein between from breast cancer and gastric cancer, p100 has the capacity to tumour samples to accurately determine how expression inhibittumourcellproliferation[34, 38]. Further exploration varies with the “aggressiveness” of a tumour. As these HER-2 reported a decrease in downstream signal induction such as splice variants secrete proteins [34], in future studies it may the MAP kinase pathways [30]. be of value to obtain corresponding patient serum samples Several studies have provided evidence that this secreted alongwithtumoursamplestogainamoreaccuraterepre- truncated form of HER-2 may serve as a serum biomarker sentation of their protein levels. Additional variables which particularly in informing treatment decisions [30, 39, 40]. contribute to clinical outcome, such as hormone receptor Leyland-Jones et al. demonstrated reduced levels of p100 status or lymph node involvement, would also need to be expression in more aggressive tumours [39]. Further studies considered. in breast cancer have continued to evaluate its role as a biomarker, and its value remains an issue for debate [30, Potential presence of other truncated HER-2 proteins 41]. Some reports suggest that this variant may compete for should also be considered when interpreting HER-2 protein selection by monoclonal antibodies such as trastuzumab expression. Truncated proteins arise not only from alternative thereby interfering with its treatment activity [38, 41]. splicing but also via proteolytic cleavage or alternative initiation of translation [44]. HER-2 proteins encoding only the 7. Herstatin extracellular domain (ECD) are produced, ranging from 95 to 105 kDa [44]; therefore, it cannot be assumed that all 100 kDa Herstatin is another naturally occurring truncated HER-2 HER-2 proteins are p100 as they may constitute other HER-2 protein generated from alternative HER-2 mRNA transcripts ECD-derived proteins. 6 International Journal of Cell Biology

9. What Is a ‘‘True’’ HER-2 Status? SSO-induced skipping of exon 15 produced a novel protein, Δ15HER2, which acted to downregulate wild-type HER-2 and Hormone receptor status can be predictive of the efficacy of induce apoptosis of HER-2 overexpressing tumour cells [52]. endocrine therapies, but we now know that current screening Such strategies could be adapted to manipulate production strategies, using immunochemistry and N-terminal antibod- of Δ16HER-2 or p100. Whilst research is ongoing to improve ies, may overlook the “true” hormonal receptor status since delivery methods of splicing-targeted therapies [53], they the procedure does not take into account truncated splice do appear as a promising strategy for future anticancer variants of either the oestrogen receptor or progesterone therapeutic intervention. receptor [26, 45, 46]. The same principle can be applied to Detecting the proportion and relevance of HER-2 spliced HER-2 status. As HER-2 positive status is determined when variants, as described, could potentially “redefine” HER- immunocytochemical staining exceeds a specified threshold 2 status. These spliced variants could consequently impact (Figure 1(d)), tumours deemed that HER-2 negative may not treatment routes in HER-2-positive tumours and also HER- be wholly negative but do not exceed this “threshold of 2-negative tumours and DCIS. For example, tumours pre- positivity.”Previous reports by Castiglioni et al. demonstrated viously deemed HER-2 negative which express that these Δ that the proportion of 16HER-2 expressed was central to variants above a specified threshold may in fact benefit from malignant transformation [32]. In DCIS samples where exon therapies targeting the HER-2 spliced variant thereby improv- Δ 16 skipping occurs, 16HER-2 may have been a trigger ing stratification of patients to “individualized” treatments. to transformation. Although HER-2 status is not routinely Additionally, proportions of splice variants in patients who do measured in DCIS, Harada et al. reported that HER-2 pos- not respond to, or regress on, anti-HER-2 drugs may indicate itivity in DCIS patients was associated with increased risk of treatment with alternative drugs as a superior alternative. developing invasive carcinoma [47].Thisisespeciallyrelevant This could potentially have implications regarding the when previous reports regarding the cancer-related and valueofHER-2splicedvariantsinaclinicalcontext.The Δ treatment resistance properties of 16HER-2 are considered presence or absence of HER-2 spliced variants may influence Δ [30, 33]. The proportion of 16HER-2 has already been shown prognosis or response to treatment. Further investigation to be important in breast cancer progression [33, 34]. If a could reveal a clinical use for Δ16HER-2, Herstatin, or p100, Δ DCIS sample was shown to express high levels of 16HER- for example, in making treatment decisions or as a potential 2, this patient may be at greater risk of disease progression therapeutic target. Further exploration of HER-2 biology, and therefore may benefit from more rigorous treatment or signaling,andresistancemechanismsisthereforeessentialto followup. develop and implement new strategies of therapeutic inter- Previous studies report that p100 expression decreases vention. in more aggressive tumours [39]; DCIS is considered a less aggressive form of breast cancer as it is preinvasive therefore Acknowledgment unable to metastasise. Such results align with expectations that p100 expression is higher in less aggressive tumours. The authors would like to thank Clinical Pathologist Dr. These spliced variants may play a role in determining the Dianne Hemming for providing the histological images of nature and clinical outcome of breast tumours in which they breast carcinoma and the HercepTest. are expressed. It remains to be fully explored as to whether coexpression References of the mRNA of the three HER-2 spliced variants has any impact on subsequent translation, or indeed how the proteins [1] P. T. Simpson, J. S. Reis-Filho, T. Gale, and S. R. Lakhani, collectively might interact when coexpressed. It would be of “Molecular evolution of breat cancer,” Journal of Pathology,vol. 205, no. 2, pp. 248–254, 2005. value to determine the proportions of all three HER-2 splice variants in the same tumour cells and to evaluate their impact [2] D. Hanahan and R. A. Weinberg, “Hallmarks of cancer: the next generation,” Cell,vol.144,no.5,pp.646–674,2011. on cell growth and drug resistance. One study has evaluated full-length HER-2 status using qPCR [48]andadvocatedits [3]A.BombonatiandD.C.Sgroi,“Themolecularpathologyof breast cancer progression,” Journal of Pathology,vol.223,no.2, use in concordance with immunohistochemistry; however, it pp. 307–317, 2011. did not consider quantification of HER-2 splice variants. [4] G. Mountzios, D. Sanoudou, and K. N. Syrigos, “Clinical phar- macogenetics in oncology: the paradigm of molecular targeted 10. HER-2 Variants as Clinical Targets? therapies,” Current Pharmaceutical Design,vol.16,no.20,pp. 2184–2193, 2010. One potential line of development for targeted anticancer [5]C.L.Arteaga,M.X.Sliwkowski,C.K.Osborne,E.A.Perez,F. therapeutics is the manipulation of HER-2 spliced variants Puglisi, and L. Gianni, “Treatment of HER2-positive breast can- [49, 50]. Splicing-targeted therapeutics has already shown cer: current status and future perspectives,” Nature Reviews a promise in treatment of disease. For example, induced Clinical Oncology,vol.9,no.1,pp.16–32,2012. exon skipping in Duchenne muscular dystrophy produces [6]A.L.Schechter,D.F.Stern,andL.Vaidyanathan,“Theneu a “Becker muscular dystrophy-like dystrophin isoform,” oncogene: an erb-B-related gene encoding a 185,000-M(r) successfully reducing disease severity [51]. One study has tumour antigen,” Nature,vol.312,no.5994,pp.513–516,1984. also demonstrated success using splice-switching oligonu- [7] I. Rubin and Y. Yarden, “The basic biology of HER2,” Annals of cleotides (SSO) to target HER-2 [52]. Wan et al. reported that Oncology,vol.12,no.1,supplement,pp.S3–S8,2001. International Journal of Cell Biology 7

[8] A. Citri and Y. Yarden, “EGF-ERBB signalling: towards the sys- [26] M. A. J. Van Dijk, A. A. M. Hart, and L. J. Van’t Veer, “Differences tems level,” Nature Reviews Molecular Cell Biology,vol.7,no.7, in estrogen receptor 𝛼 variant messenger RNAs between normal pp.505–516,2006. human breast tissue and primary breast carcinomas,” Cancer [9]P.Casalini,M.V.Iorio,E.Galmozzi,andS.Menard,´ “Role of Research,vol.60,no.3,pp.530–533,2000. HER receptors family in development and differentiation,” Jour- [27] J. P.Venables, “Aberrant and alternative splicing in cancer,” Can- nal of Cellular Physiology,vol.200,no.3,pp.343–350,2004. cer Research,vol.64,no.21,pp.7647–7654,2004. [10]D.Graus-Porta,R.R.Beerli,J.M.Daly,andN.E.Hynes, [28] R.Klinck,A.Bramard,L.Inkeletal.,“Multiplealternativesplic- “ErbB-2, the preferred heterodimerization partner of all ErbB ing markers for ovarian cancer,” Cancer Research,vol.68,no.3, receptors, is a mediator of lateral signaling,” EMBO Journal,vol. pp. 657–663, 2008. 16,no.7,pp.1647–1655,1997. [29] H. L. Gautrey and A. J. Tyson-Capper, “Regulation of Mcl-1 by [11] C. Curtis, S.P. Shah, S. F. Chin et al., “The genomic and tran- SRSF1 and SRSF5 in cancer cells,” PLoS ONE,vol.7,no.12, scriptomic architecture of 2,000 breast tumours reveals novel Article ID e51497, 2012. subgroups,” Nature,vol.486,no.7403,pp.346–352,2012. [30] M. Sasso, F. Bianchi, V.Ciravolo, E. Tagliabue, and M. Campigl- [12] F. F. T. Barros, D. G. Powe, I. O. Ellis, and A. R. Green, “Under- io, “HER2 splice variants and their relevance in breast cancer,” standing the HER family in breast cancer: interaction with Journal of Nucleic Acids Investigation,vol.2,no.1,pp.52–59, ligands, dimerization and treatments,” Histopathology,vol.56, 2011. no. 5, pp. 560–572, 2010. [31]N.G.Azios,F.J.Romero,M.C.Denton,J.K.Doherty,andG. M. Clinton, “Expression of herstatin, an autoinhibitor of HER- [13]A.L.Schechter,M.C.Hung,andL.Vaidyanathan,“Theneu 2/neu, inhibits transactivation of HER-3 by HER-2 and blocks gene: an erbB-homologous gene distinct from and unlinked to EGF activation of the EGF receptor,” Oncogene,vol.20,no.37, the gene encoding the EGF receptor,” Science,vol.229,no.4717, pp. 5199–5209, 2001. pp.976–978,1985. [32] F. Castiglioni, E. Tagliabue, M. Campiglio, S. M. Pupa, A. [14] E.Cerami,J.Gao,U.Dogrusozetal.,“ThecBiocancergenomics Balsari, and S. Menard,´ “Role of exon-16-deleted HER2 in breast portal: an open platform for exploring multidimensional cancer carcinomas,” Endocrine-Related Cancer,vol.13,no.1,pp.221– genomics data,” Cancer Discovery,vol.2,no.5,pp.401–404, 232, 2006. 2012. [33] D. Mitra, M. J. Brumlik, S. U. Okamgba et al., “An oncogenic [15]J.Gao,B.A.Aksoy,U.Dogrusozetal.,“Integrativeanalysisof isoform of HER2 associated with locally disseminated breast complex cancer genomics and clinical profiles using the cBio- cancer and trastuzumab resistance,” Molecular Cancer Thera- Portal,” Science Signaling, vol. 6, no. 269, p. pl1, 2013. peutics,vol.8,no.8,pp.2152–2162,2009. [16] A. Valent, F.Penault-Llorca, A. Cayre, and G. Kroemer, “Change [34] A. Aigner, H. Juhl, C. Malerczyk, A. Tkybusch, C. C. Benz, in HER2 (ERBB2) gene status after taxane-based chemotherapy and F. Czubayko, “Expression of a truncated 100 kDa HER2 for breast cancer: polyploidization can lead to diagnostic pitfalls splice variant acts as an endogenous inhibitor of tumour cell with potential impact for clinical management,” Cancer Genet- proliferation,” Oncogene, vol. 20, no. 17, pp. 2101–2111, 2001. ics,vol.206,no.1-2,pp.37–41,2013. [35] T. Koletsa, I. Kostopoulos, E. Charalambous, B. Christoforidou, [17]D.J.Slamon,G.M.Clark,andS.G.Wong,“Humanbreast E. Nenopoulou, and V. Kotoula, “A splice variant of HER2 cor- cancer: correlation of relapse and survival with amplification of responding to herstatin is expressed in the noncancerous breast the HER-2/neu oncogene,” Science,vol.235,no.4785,pp.177– and in breast carcinomas,” Neoplasia,vol.10,no.7,pp.687–696, 182, 1987. 2008. [18] R. Seshadri, F. A. Firgaira, D. J. Horsfall, K. McCaul, V. Setlur, [36] C. Marchini, F. Gabrielli, M. Iezzi et al., “The human splice and P. Kitchen, “Clinical significance of HER-2/neu oncogene variant Δ16HER2 induces rapid tumor onset in a reporter trans- amplification in primary breast cancer,” Journal of Clinical genic mouse,” PLoS ONE,vol.6,no.4,ArticleIDe18727,2011. Oncology,vol.11,no.10,pp.1936–1942,1993. [37] K. Pedersen, P.-D. Angelini, S. Laos et al., “Anaturally occurring [19] N. Harbeck, M. D. Pegram, J. Ruschoff,¨ and V.Mobus,¨ “Targeted HER2 carboxy-terminal fragment promotes mammary tumor therapy in metastatic breast cancer: the HER2/neu oncogene,” growth and metastasis,” Molecular and Cellular Biology,vol.29, Breast Care,vol.5,no.1,pp.3–7,2010. no.12,pp.3319–3331,2009. [20] Y. Yarden, “Biology of HER2 and its importance in breast [38] G.K.Scott,R.Robles,J.W.Parketal.,“Atruncatedintracellular cancer,” Oncology,vol.61,no.2,pp.1–13,2001. HER2/neu receptor produced by alternative RNA processing affects growth of human carcinoma cells,” Molecular and Cel- [21] Office for National Statistics, “Summary: Breast Cancer in Eng- lular Biology,vol.13,no.4,pp.2247–2257,1993. land,” 2009, http://www.ons.gov.uk/ons/rel/cancer-unit/breastcancer-in-england/2009/breast-cancer.html. [39] B. Leyland-Jones, S. Lennon, C. Barton et al., “Utility of serum HER2 extracellular domain assessment in clinical decision [22] H. Mukai, “Treatment strategy for HER2-positive breast can- making: pooled analysis of four trials of trastuzumab in cer,” International Journal of Clinical Oncology,vol.15,no.4,pp. metastatic breast cancer,” Journal of Clinical Oncology,vol.27, 335–340, 2010. no.10,pp.1685–1693,2009. [23] A. J. Ward and T. A. Cooper, “The pathobiology of splicing,” [40] H. Makino, M. Iraha, N. Manba, H. Hashidate, and M. Shiotani, Journal of Pathology,vol.220,no.2,pp.152–163,2010. “Utility of serum human epidermal growth factor receptor- [24] E. Kim, A. Goren, and G. Ast, “Insights into the connection 2 extracellular domain (HER2 ECD) assessment in patients between cancer and alternative splicing,” Trends in Genetics,vol. with advanced or recurrent breast cancer and those who 24,no.1,pp.7–10,2008. received neoadjuvant therapy,” Japanese Journal of Cancer and [25] J. P. Venables, R. Klinck, A. Bramard et al., “Identification of Chemotherapy,vol.39,no.2,pp.237–240,2012. alternative splicing markers for breast cancer,” Cancer Research, [41] A. F. Leary, W. M. Hanna, M. J. Van De Vijver et al., “Value vol.68,no.22,pp.9525–9531,2008. and limitations of measuring HER-2 extracellular domain in the 8 International Journal of Cell Biology

serum of breast cancer patients,” Journal of Clinical Oncology, vol. 27, no. 10, pp. 1694–1705, 2009. [42] J. K. Doherty, C. Bond, A. Jardim, J. P.Adelman, and G. M. Clin- ton, “The HER-2/neu receptor tyrosine kinase gene encodes a secreted autoinhibitor,” Proceedings of the National Academy of Sciences of the United States of America,vol.96,no.19,pp. 10869–10874, 1999. [43] Q. A. Justman and G. M. Clinton, “Herstatin, an autoinhibitor of the human epidermal growth factor receptor 2 tyrosine kinase, modulates epidermal growth factor signaling pathways resulting in growth arrest,” JournalofBiologicalChemistry,vol. 277,no.23,pp.20618–20624,2002. [44] R. Zagozdzon, W. M. Gallagher, and J. Crown, “Truncated HER2: implications for HER2-targeted therapeutics,” Drug Dis- covery Today,vol.16,no.17-18,pp.810–816,2011. [45]D.M.Cork,T.W.Lennard,andA.J.Tyson-Capper,“Proges- terone receptor (PR) variants exist in breast cancer cells characterised as PR negative,” Tumor Biology,vol.33,no.6,pp.2329– 2340, 2012. [46]D.M.W.Cork,T.W.J.Lennard,andA.J.Tyson-Capper,“Alter- native splicing and the progesterone receptor in breast cancer,” Breast Cancer Research,vol.10,no.3,article207,2008. [47] S. Harada, R. Mick, R. E. Roses et al., “The significance of HER- 2/neu receptor positivity and immunophenotype in ductal carcinoma in situ with early invasive disease,” JournalofSurgical Oncology,vol.104,no.5,pp.458–465,2011. [48] J. Lehmann-Che, F.Amira-Bouhidel, E. Turpin et al., “Immuno- histochemical and molecular analyses of HER2 status in breast cancers are highly concordant and complementary approaches,” British Journal of Cancer, vol. 104, no. 11, pp. 1739–1746, 2011. [49] M. A. Garcia-Blanco, “Alternative splicing: therapeutic target and tool,” Progress in Molecular and Subcellular Biology,vol.44, pp.47–64,2006. [50] J. A. Bauman, S.-D. Li, A. Yang, L. Huang, and R. Kole, “Anti- tumor activity of splice-switching oligonucleotides,” Nucleic Acids Research, vol. 38, no. 22, pp. 8348–8356, 2010. [51] C. T. Fragall, A. M. Adams, R. D. Johnsen, R. Kole, S. Fletcher, and S. D. Wilton, “Mismatched single stranded antisense oligonucleotides can induce efficient dystrophin splice switching,” BMC Medical Genetics, vol. 12, article 141, 2011. [52] J. Wan, P. Sazani, and R. Kole, “Modification of HER2 pre- mRNA alternative splicing and its effects on breast cancer cells,” International Journal of Cancer,vol.124,no.4,pp.772–777, 2009. [53] S. A. El Andaloussi, S. M. Hammond, I. Mager,¨ and M. J. Wood, “Use of cell-penetrating-peptides in Oligonucleotide splice switching therapy,” Current Gene Therapy,vol.12,no.3,pp.161– 178, 2012.