CELL CYCLE TRANSCRIPTION CONTROL BY UBIQUITIN SIGNALING
Anthony P. Arceci
A dissertation submitted to the faculty at the University of North Carolina at Chapel Hill in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Curriculum in Genetics and Molecular Biology in the School of Medicine
Chapel Hill 2019
Approved by:
Ian J. Davis
Channing J. Der
Robert J. Duronio
Brian Kuhlman
Michael J. Emanuele
© 2019 Anthony P. Arceci ALL RIGHTS RESERVED
ii
ABSTRACT
Anthony P. Arceci: Cell Cycle Transcription Control by Ubiquitin Signaling (Under the direction of Michael J. Emanuele)
The cell cycle is a tightly regulated series of molecular events which dictates proliferation. Both the timely activation of genes through transcription and destruction of proteins through the ubiquitin-proteasome system are integral to normal cell cycles.
Dysregulation of these networks often underlie a variety of malignant diseases such as cancer.
Forkhead box protein M1 (FOXM1) is an essential cell cycle transcription factor.
FOXM1 regulates a transcriptional network that controls the G2/M transition and G1/S transition.
Additionally, aberrant upregulation of the FOXM1 transcriptional network is linked to a variety of cancers. The kinases which activate FOXM1 are well explored, but the influence of the ubiquitin-proteasome system on FOXM1 remains unclear. Here, I described the role that two such enzymes, the E3 ubiquitin ligase CUL4-VPRBP and the deubiquitinating enzyme (DUB)
USP21, have on the stability and activity of FOXM1 in both normal and dysregulated cell cycles.
First, I demonstrate that FOXM1 degradation is enhanced by association with CUL4-
VPRBP. Depletion of VPRBP enhances FOXM1 stability and causes mitotic entry defects.
Interestingly, overexpression of VPRBP enhances both FOXM1 ubiquitination and transcriptional activity by a process that occurs independent of CUL4. Finally, VPRBP and
FOXM1 levels are assessed in high-grade serous ovarian cancer (HGSOC) patient tumors, demonstrating a plausible mechanism for FOXM1 activation.
iii Second, I demonstrate that FOXM1 is protected from degradation through association with the DUB USP21. Knockdown or overexpression of USP21 is able to destabilize or stabilize
FOXM1, respectively, through deubiquitination of FOXM1. USP21 is able to influence mitotic entry and proliferation through regulating the FOXM1 transcriptional network. Furthermore,
USP21 and FOXM1 are both significantly amplified in basal-like breast cancer with the knockdown of both sensitizing cells to the chemotherapy paclitaxel thus describing a novel combination treatment for this disease.
Taken together, these results contribute to our understanding of how the ubiquitin- proteasome system positively and negatively regulates the abundance and activity of FOXM1.
The research presented here further extends our understanding of the network of interactions regulating normal cell cycle dynamics and provide mechanistic and novel therapeutic insights into the promotion and treatment of cancer.
iv
To what great things await.
v
ACKNOWLEDGEMENTS
Looking back on my journey to this point, it is clear to me that I did not arrive here without help and support along the way. Starting from the beginning of my academic journey, I would like to acknowledge my mother, Darlene, and friend, Ron, for believing in me from the start and helping me establish myself as I relocated from Winchendon, Massachusetts to Tucson,
Arizona to take my first step in what would become a 12 year journey of discovery and learning.
I would like to acknowledge my network of friends, Anthony, Armando, Brandon, Craig,
Greg, Patrick and Wayne who became my band of brothers, helping me navigate through the journey.
I would also like to acknowledge Dr. Kathleen Dixon, who fostered my passion for scientific inquiry and set me on the path to becoming a PhD.
Dr. Angela Wandinger-Ness, Dr. Richard Cripps and Antonio Banuelos deserve recognition for providing me with the launching pad and experience to transition from the undergraduate level to the post-graduate level during my year at the University of New Mexico.
I would like to acknowledge my colleagues in the lab Allie, Jen, Raj, Taylor, Thomas and
Xianxi who were not only important participants in the research to follow, but also a great group of people to work with over the years.
My partner, Amanda, deserves an incredible amount of recognition and has been with me through my entire graduate career. Without your love and support over the years I’m not sure if I would have made it.
vi Finally, I would be remiss without recognizing my advisor, Dr. Michael Emanuele and the role that he played in guiding me to this point. Without your insight and knowledge to help me guide my research and humor and grace to make working in the lab fun, I surely would not have succeeded.
I thank you all for the roles that you have played in my life.
vii
PREFACE
In each of the following chapters portions of the work described were completed by other talented scientists and exceptional colleagues. Additionally, most chapters describe work that has been published and that I have permission to use as part of this dissertation.
Chapter 2 describes work from a project which I made significant contributions to during my rotation and first year in the Emanuele lab and was ultimately completed by my colleague in the lab, Xianxi Wang. I contributed all experiments utilizing cell cycle analysis, RT-qPCR, and chromatin immunoprecipitation (ChIP). Additionally, I cloned some of the plasmids used in many of the cell biology assays described and provided critical revisions and suggestions to the manuscript throughout the peer-review process. Permission to include this article in its entirety is granted by the journal Molecular and Cellular Biology as described in their editorial policy
(https://mcb.asm.org/content/editorial-policy).
Chapter 3 describes work from a project in which I designed, conducted and analyzed the vast majority of experiments, wrote the manuscript and made most critical revisions throughout the peer-review process. The work described in Chapter 3 was published as an open-access article in Cell Reports. As such, I retain the copyright to this article as an author to include it in this dissertation in its entirety per editorial guidelines
(https://www.elsevier.com/about/policies/copyright#Author-rights).
viii
TABLE OF CONTENTS
LIST OF TABLES ...... xii
LIST OF FIGURES ...... xiii
LIST OF ABBREVIATIONS ...... xiv
CHAPTER 1: INTRODUCTION ...... 1
The Ubiquitin-Proteasome System ...... 1
Background ...... 1
Enzymes of the Ubiquitin-Proteasome System ...... 2
E3 Ubiquitin Ligases ...... 3
CRL4A and CRL4B ...... 5
DUBs ...... 6
USP21 ...... 8
The Cell Cycle ...... 9
Historical Background ...... 9
Mechanisms of Control ...... 10
Cell Cycle Transcription by FOXM1 ...... 14
Background ...... 14
Overview of FOXM1 Biology ...... 15
Post-Translational Regulation of FOXM1 Through the Cell Cycle ...... 16
Ubiquitin-dependent Regulation of FOXM1 ...... 18
FOXM1 in Cancer ...... 19
ix CHAPTER 2: VPRBP/DCAF1 REGULATES THE DEGRADATION AND NONPROTEOLYTIC ACTIVATION OF THE CELL CYCLE TRANSCRIPTION FACTOR FOXM1 ...... 25
Introduction ...... 25
FOXM1 Stability is Regulated by CRL4VPRBP ...... 28
FOXM1 Interaction and Ubiquitylation by VPRBP ...... 31
VPRBP is a FOXM1 Activator ...... 33
VPRBP Protein is Upregulated in High-Grade Serous Ovarian Tumors ...... 37
Discussion ...... 37
Materials and Methods ...... 41
Immunoblot Analysis ...... 41
Immunoprecipitation ...... 42
RT-qPCR ...... 42
Luciferase Reporter Assays ...... 42
Flow Cytometry ...... 43
In Vivo Ubiquitination Assay ...... 43
Expression and Binding Assay for VPRBP and FOXM1c ...... 44
CHAPTER 3: FOXM1 DEUBIQUITINATION BY USP21 REGULATES CELL CYCLE PROGRESSION AND PACLITAXEL SENSITIVITY IN BASAL-LIKE BREAST CANCER ...... 56
Introduction ...... 56
Results ...... 58
USP21 Binds and Alters FOXM1 Abundance ...... 58
FOXM1 is a USP21 Substrate ...... 60
USP21 Affects the FOXM1 Transcriptional Network and Proliferation ...... 62
USP21 Amplification Correlates with FOXM1 Protein Levels and Tumor Proliferation in BLBC ...... 64
x USP21 Affects Paclitaxel Response in Basal-like Breast Cancer ...... 66
Discussion ...... 68
Materials and Methods ...... 70
Cell Culture ...... 70
Transfections and Treatments ...... 71
Molecular Cloning ...... 71
Lentivirus Production ...... 72
Immunoblotting ...... 72
Immunoprecipitations ...... 73
FOXM1 and USP21 In Vitro Binding Assay ...... 73
FOXM1 Transcriptional Reporter Assay ...... 74
Cell Cycle Progression into Mitosis ...... 74
In Vitro Paclitaxel Sensitivity Assays ...... 75
Tumor Xenograft ...... 75
Proliferation Assays ...... 76
In Vivo Deubiquitination Assay ...... 76
In Vitro Deubiquitination Assay ...... 77
RT-qPCR ...... 77
TCGA Dataset Analysis ...... 78
Other Statistical Analysis ...... 79
CHAPTER 4: CONCLUSIONS AND FUTURE DIRECTIONS ...... 97
REFERENCES ...... 104
xi LIST OF TABLES
Table 2.1 Antibodies used in experiments described in Chapter 2...... 54
Table 2.2 Sequences of primers used in experiments described in Chapter 2...... 55
Table 3.1 FOXM1 target gene enrichment correlated to USP21 amplification...... 93
Table 3.2 siRNA and shRNA used in Chapter 3 experiments...... 94
Table 3.3 Primers used in Chapter 3 experiments...... 95
Table 3.4 Antibodies used in Chapter 3 experiments...... 96
xii LIST OF FIGURES
Figure 1.1 Overview of the ubiquitin-proteasome system...... 23
Figure 1.2 FOXM1 post-translational modification landscape...... 24
Figure 2.1 FOXM1 is regulated by CUL4, DDB1 and VPRBP...... 45
Figure 2.2 FOXM1 is regulated by VPRBP...... 46
Figure 2.3 VPRBP binds and ubiquitylates FOXM1 in vivo...... 47
Figure 2.4 Examining the FOXM1-VPRBP interaction...... 48
Figure 2.5 VPRBP activates the FOXM1 6xDB transcriptional reporter...... 49
Figure 2.6 VPRBP regulates FOXM1 independent of binding to the CRL4 complex...... 50
Figure 2.7 VPRBP depletion reduces FOXM1 target gene expression and the number of mitotic cells...... 51
Figure 2.8 VPRBP associates with FOXM1 promoters and activates FOXM1 independent of CUL4...... 52
Figure 2.9 Implications and cell cycle control and malignancy...... 53
Figure 3.1 USP21 binds and regulates FOXM1 abundance...... 80
Figure 3.2 Analysis of USP21-FOXM1 binding...... 81
Figure 3.3 USP21 protects FOXM1 from proteasomal degradation...... 82
Figure 3.4 USP21 controls FOXM1 ubiquitination and half-life...... 83
Figure 3.5 USP21 affects growth through modulation of the FOXM1 transcriptional network. . 84
Figure 3.6 USP21 regulation of cell cycle and proliferation...... 85
Figure 3.7 Amplification of FOXM1 and USP21 is linked to proliferation in BLBC...... 86
Figure 3.8 USP21 knockdown sensitizes BLBC cells to paclitaxel...... 87
Figure 3.9 Analysis of USP21 and proliferation in breast cancer...... 88
Figure 3.10 Illustration of proposed FOXM1-USP21 axis in cell cycle and disease...... 89
xiii LIST OF ABBREVIATIONS
AEBSF 4-(2-aminoethyl)-benzenesulfonly fluoride
APC/C Anaphase promoting complex/cyclosome
ATP Adenosine triphosphate
ATRA All-trans retinoic acid
BLBC Basal-like breast cancer
BSA Bovine serum albumin
CAK CDK-activating kinase
CDI Cyclin-dependent kinase inhibitor
CDK Cyclin-dependent kinase
ChIP Chromatin immunoprecipitation
CHX Cycloheximide
CNA Copy number alteration
Co-IP Co-immunoprecipitation
CRL Cullin-RING ligase
CRL4 Cullin4-based ligase
CSN COP9 signalsome
DCAF DDB1 and CUL4-associated factor
DiGly Diglycine
DMEM Dulbecco’s modified Eagle’s medium
DMSO Dimethyl sulfoxide
DNA Deoxyribonucleic acid
xiv dsRed Discosoma red fluorescent protein
DTT Dithithreitol
DUB Deubiquitinating enzyme
DWD DDB1 binding WD40 protein
EGFP Enhanced green fluorescent protein
ER Estrogen receptor
FACS Fluorescence activated cell sorting gDNA Genomic deoxyribonucleic acid
GISTIC Genome identification of significant targets in cancer
GSEA Gene set enrichment analysis
HECT Homologous to the E6AP carboxyl terminus
HGSOC High-grade serous ovarian cancer
HIV Human immunodeficiency virus
HRP Horseradish peroxidase
IB Immunoblot
IP-MS/MS Immunoprecipitation-tandem mass spectrometry
IPTG Isopropyl-b-D-thiogalactopyranoside
JAMM/MPN+ Jab1/Mov34/Mpr1 Pad1 N-terminal+
MCC Mitotic checkpoint complex
MINDY Motif interacting with Ub-containing novel DUB family
MJD Machado-Josephin domain protease mRNA Messenger ribonucleic acid
NOD/SCID Non-obese diabetic/severe combined immunodeficiency
xv ORF Open reading frame
OTU Ovarian tumor protease
P-H3 Phosphorylation on histone H3 serine 10
PBS Phosphate buffered saline
PCR Polymerase chain reaction
PIP PCNA-interacting peptide motif
PMSF Phenylmethylsulfonyl fluoride
PR Progesterone receptor
PTM Post-translational modification
RING Really Interesting New Gene
RNA Ribonucleic acid
RPPA Reverse phase protein array
RT-qPCR Quantitative reverse transcription polymerase chain reaction
SAC Spindle assembly checkpoint
SCF Skp, Cullin, F-box containing complex siRNA Small interfering ribonucleic acid
SUMO Small ubiquitin-like modification
TAD Transcriptional activation domain
TCEP Tris(2-carboxyethyl)phosphine
TCGA The Cancer Genome Atlas
Ub Ubiquitin
UCH Ubiquitin C-terminal hydrolase
USP Ubiquitin-specific peptidase
xvi
CHAPTER 1: INTRODUCTION
The Ubiquitin-Proteasome System
Background
The ubiquitin-proteasome system (UPS) will play an important role in the work described in the subsequent chapters. As such, I would like to give a cursory overview of the history, mechanisms and general functions of the UPS in this section (Figure 1.1). A more thorough explanation of E3 ubiquitin ligases and cullin-RING ligase 4 will be described as they are importantly involved in the studies described in Chapter 2. Similarly, deubiquitinating enzymes and ubiquitin specific protease 21 (USP21) will be described in more detail as they are central to the studies described in Chapter 3.
The UPS describes a network of proteases which influence the abundance, stability and function of the vast majority of the proteins within the cell. It was long believed that most cellular proteins were stable and long-lived. It wasn’t until work done in the late 1970’s and early 1980’s that this paradigm began to shift. Hershko, Ciechanover and Rose elegantly demonstrated that proteins added to a reticulocyte extract can become covalently conjugated to a small protein called ubiquitin (1, 2). Furthermore, they discovered that these proteins conjugated to ubiquitin are subsequently destroyed by an ATP-dependent protease, later described as the
26S proteasome, in the extract (3–5). Due to the broad influence of the UPS on nearly all of the cells major signaling pathways, efforts continue to fully identify and understand links between enzymes of the UPS, specific substrates and how these interactions regulate essential cell
1 processes, how their dysregulation can lead to disease and how we may be able to leverage the
UPS for therapeutic effect (6–9).
Enzymes of the Ubiquitin-Proteasome System
The signal by which this system communicates occurs through the conjugation of ubiquitin, a small 8.6 kDa regulatory protein, to the lysine’s (or methionine in N-terminal ubiquitin-ubiquitin linkages) of other proteins. This process is orchestrated by a cascade of enzymes described as E1 (ubiquitin-activating enzymes), E2 (ubiquitin-conjugating enzymes) and E3 (ubiquitin ligases) (10). The process of conjugating ubiquitin onto the lysine of a protein begins when the E1 enzyme generates a high-energy thiol ester intermediate with ubiquitin in a reaction requiring ATP (E1-S-ubiquitin) (11). From here, one of several E2 enzymes participates in a reaction with the E1-S-ubiquitin intermediate in which the ubiquitin is transferred in a thiol ester reaction to the E2 enzyme (E2-S-ubiquitin) (12). At this point, E2 enzymes bind into a complex with E3 ubiquitin ligase enzymes which also engage the substrate that is to be ubiquitinated (10, 13). By keeping the substrate and the E2 enzyme conjugated to ubiquitin close to one another, the E3 ubiquitin ligases facilitate the transfer of ubiquitin onto the substrate. In this reaction, the C-terminal glycine of ubiquitin forms an amide linkage with the side chain of a lysine residue on the protein being ubiquitinated (14). This reaction can occur once, where only one ubiquitin is placed onto a substrate (monoubiquitination) (15), multiple times where a ubiquitin is added to a different lysine than the first (multi-monoubiquitination) or multiple times where ubiquitin is conjugated on to the lysines of the first ubiquitin conjugated onto the protein
(polyubiquitination) (10).
2 Additionally, there is a class on enzymes known as deubiquitinating enzymes (DUBs) which antagonize this reaction and are able to bind substrates and remove ubiquitin directly from the protein itself or remove or edit polyubiquitin linkages (16). The specific ubiquitin code remaining on substrate proteins following interaction with the enzymes of the ubiquitin- proteasome system can have a variety of functional or signaling outcomes (10, 15).
One outcome of ubiquitin signaling, which is closely linked with polyubiquitin chain formation, is degradation. This occurs when ubiquitinated proteins interact with the 26S proteasome where they are digested into their component amino acids and the ubiquitin is recycled (10). Finally, there are a number of ubiquitin binding proteins (UBPs) within the cell which are able to bind and recognize the ubiquitin modifications on target proteins and determine downstream signaling outcomes for these ubiquitinated proteins (17).
E3 Ubiquitin Ligases
With around six hundred E3s encoded in the human genome (compared to two E1s and around fifty E2s), the E3 ubiquitin ligases represent the largest and most diverse set of enzymes within the broader family of enzymes comprising the ubiquitin-proteasome system (18). Due to this large number and the fact that E3 ubiquitin ligases are the enzymes which directly recognize unique substrates, most of the specificity in ubiquitin signaling is conferred through E3 ubiquitin ligase-substrate interactions. As such, the linking of specific substrates to their cognate E3 ubiquitin ligases remains of great interest for understanding how and which cell signaling pathways are influenced by the ubiquitin-proteasome system.
Much is known about the molecular mechanisms regulating E3 ubiquitin ligase formation and activity. Broadly, E3 ubiquitin ligases are separated into two major classes based on
3 sequence and function. These are the RING (really interesting new gene)-domain E3s and the
HECT (homologous to the E6AP carboxyl terminus)-domain E3s. The RING-domain E3 ubiquitin ligases are the largest of the two groups with around 300 RING-containing genes encoded by the human genome (18). In some RING-domain E3 ligases the substrate recognition site and the catalytic site is found within the same polypeptide, but for the vast majority of
RING-domain E3 ligases, the substrate recognition module and the catalytic subunit are separated into distinct polypeptides which come together to form a multi-subunit complex. More specifically, there are over two hundred modular cullin-RING ligases (CRLs) which can be formed from the combination of RING-domain containing proteins with a rigid cullin protein scaffold (19). The human genome encodes eight different cullins (CUL1, 2, 3, 4A, 4B, 5, 7 and
9), each of which engage a distinct set of subunits which are often referred to by various names
(19). When fully assembled, E2-E3 CRLs are able to engage with substrates and participate in reactions that both initiate the addition of ubiquitin on a target substrate or elongate ubiquitin chains on a target substrate which result in substrates which are competent for subsequent degradation by the proteasome (20).
In addition to assembly dynamics, CRLs are regulated by the addition of a small, ubiquitin-like modification on the cullin scaffold known as NEDD8. The addition of NEDD8 to the cullin changes the conformation of the complex and allows for the E2 to be brought into closer proximity to the substrate being held by the E3 (21). This allows for the transfer of ubiquitin on to the substrate. Much like ubiquitin conjugation, the conjugation of NEDD8 is a reversible reaction and can be removed by the COP9 signalsome (CSN) (22). CSN deconjugation of NEDD8 is inhibited when a CRL complex is fully assembled with substrate attached but is able to remove NEDD8 following CRL disengagement with substrate. In this way, CRLs are
4 able to be reset to a deNEDDylated state to engage new substrates but their activity is not compromised when they are indeed bound to substrate.
Additionally, there is another regulator of CRL complex formation known as CAND1.
When CRLs are deNEDDylated, CAND1 can bind to the cullin scaffold and compete for binding directly with the E3 (23). When CAND1 predominates, this can block E3 binding and inhibits the cullin scaffold from engaging with other E3 partners. However, the addition of NEDD8 blocks the binding of CAND1 to the cullin scaffold and shifts the balance towards productive, fully assembled CRL complexes. This interplay between CAND1, E3s and NEDD8 is essential for the dynamic switching of E3s (24).
CRL4A and CRL4B
Of all the cullins, CUL4 is the only one which exists as a subfamily including 2 members,
CUL4A and CUL4B which share over 80% sequence identity and functional redundancy (25).
CUL4A and CUL4B engage with DNA damage binding protein 1 (DDB1) and CUL4-associated factors (DCAFs) to form active complexes (26, 27). CUL4A and CUL4B display differential expression based on tissue type with high expression observed in the thymus, testis, ovaries, T cells, multiple blood types as well as skeletal muscle (28). CRL4 promotes the ubiquitination of a number of substrates, both known (29) and predicted (27) involved in a range of physiological processes including DNA damage repair, chromatin remodeling, cell cycle progression, embryogenesis, hematopoiesis and spermatogenesis (29). CRL4 has been closely linked to a number of diseases. In cancer, it was discovered that CUL4A overexpression is a common feature of many breast cancers with its genomic loci, 13q34, being amplified in many p53- deficient, BRCA1 deficient and basal-like breast cancers (30–32). Hepatitis B virus (33, 34),
5 Simian Virus 5 (35) and human immunodeficiency virus (HIV) (36) encode accessory proteins which are able to bind and engage CRL4 to degrade host proteins such as STAT1 (37), UNG2
(38, 39) and SAMHD1 (40–42). Intriguingly, HIV Vpr is able to engage CRL4 through interaction with DCAF1/VPRBP which is sufficient to promote a G2 arrest and apoptosis (43–47) though the exact mechanism which promotes this arrest phenotype remains elusive.
Understanding the exact targets which exert this G2 arrest remains of great interesting in the field. Additionally, it has been discovered through genetic analysis of families who suffer from
X-linked intellectual disabilities that mutations in the CUL4B gene are highly prevalent (48).
More work remains to thoroughly validate and categorize specific CUL4 substrates and will significantly contribute to our understanding of cellular process which depend on CUL4 and how
CUL4 dysregulation contributes to diseases.
DUBs
Currently, there are 99 identified DUBs which can be further categorized into two broader functional groups and seven separate subfamilies based on their mechanism of action with respect to catalytic activity and other features (16). The larger of the two groups are the cysteine protease DUBs, which depend on a catalytically active triad or diad of amino acids, including cysteine, to catalyze the removal of ubiquitin from cognate substrates. Six of the seven families of DUBs (ubiquitin-specific proteases (USPs), ubiquitin C-terminal hydrolase (UCHs), ovarian tumor proteases (OTUs), Machado-Josephin domain proteases (MJDs), motif interacting with Ub-containing novel DUB family (MINDYs) and ZUP1) are cysteine protease DUBs. The smaller of the two functional groups are the metalloprotease DUBs, in which the catalytically active residues coordinate a zinc ion which is important for ubiquitin cleavage (49). The seventh
6 family, the Jab1/Mov34/Mpr1 Pad1 N-terminal+ (JAMM/MPN+) domain proteases (49–51), fall under this functional category.
DUBs can cleave the isopeptide bonds between ubiquitin molecules and modified proteins or the isopeptide bond which links ubiquitin molecules together in chains. There are a wide variety of other preferences which influence the specificity of DUB activity (52). Some examples of these specific preferences include: chain linkage type (i.e. K11 versus K48 versus
K63 linkage) (53), end processing (distal-end versus inter-chain) (16, 53), chain length (54–56), number of ubiquitin molecules removed (monoubiquitin versus en bloc chain removal) (49, 57) and substrate recognition (58). Some DUBs can also remove small ubiquitin-like modifications such as SUMO (59), ISG15 (60) and NEDD8 (61). Abundance and localization of DUBs varies greatly and heavily influences the cellular processes which they regulate (62–64).
Much like E3 ubiquitin ligases, pairing DUBs with their substrates and the process with which they exert influence is important to informing our understanding of the biology of DUB behavior (58). Also, DUBs have the potential to make excellent therapeutic targets due to their influence over specific cell signaling pathways and their biochemical properties (16). Linking
DUBs to specific disease conditions and generating potent and selective DUB inhibitors has great therapeutic potential and should be the focus of future studies (65). In contrast to the ubiquitin ligases, the study of DUBs is far less mature. Comparatively less is known about the molecular mechanisms which regulate their activity or substrate binding preferences. These are open questions which will be addressed as the field continues to mature over the coming years.
7 USP21
Ubiquitin specific peptidase 21 (USP21) is a 565 amino acid member of the USP family of DUBs. The USP21 gene is mapped to chromosome 1q23 (66). USP21 was first cloned from a human placenta cDNA library and was demonstrated to have the ability to remove both ubiquitin and NEDD8 conjugates (67), though reports on NEDD8 deconjugation remain mixed (68, 69).
Like many USP family DUBs, USP21 utilizes a catalytically active triad of cysteine, histidine and asparagine to cleave ubiquitin isopeptide bonds. USP21 has been crystalized in complex with a linear di-ubiquitin probe where it adopts a “finger, palm, thumb” conformation around ubiquitin chains (68). The C-terminal domain extends out and interacts with the distal end of the ubiquitin molecule and the catalytically active site wraps around the ubiquitin isopeptide bond in a conformation similar to the region between the palm and thumb of a hand when gripping a handle (68).
USP21 primarily displays a nuclear localization, but does contain a nuclear export sequence (70) and has been demonstrated to regulate centrosome and microtubule-associated functions (71) as well as the MARK family of kinases (58, 72). However, most of the identified
USP21 substrates exist within the nucleus. USP21 influences general transcription through the removal of ubiquitin on histone H2A K119 (73) and EZH2 (74). USP21 also influences transcription more acutely through influence the stability of a variety of transcription factors such as GATA3 (75), GLI1 (76), NANOG (77). It has also been demonstrated that USP21 can bind to the promoter region of IL-8 and aid in its transcriptional initiation (78). USP21 has also been implicated in host immune response through interaction and deubiquitination of RIP1 (79),
RIG-I (80) and IL-33 (81) together with the stimulation of interferon genes following viral infection (82) and through the preventing the generation of FOXP3+ regulatory T cells through
8 stabilizing FOXP3 (83). Recently, USP21 has also been implicated in DNA damage repair through the stabilization of BRCA2 (84). USP21 expression is linked to cell proliferation with knockdown of USP21 slowing cell growth and tumor formation with USP21 overexpression enhancing growth and tumor formation (67, 74).
Intriguingly, the genomic loci of USP21 (1q23) is often amplified in a variety of cancers, though the exact oncogenes or reasons why this region seems to be preferentially selected for in cancers remains unclear (85). The exact mechanisms which regulate the activity of USP21 remain elusive. Future work describing how and when USP21 is activated and additional substrates it engages will greatly add to our knowledge of how USP21 broadly influences cell signaling pathways and how USP21 may contribute to the progression of disease.
The Cell Cycle
Historical Background
The cell cycle describes a series of events which regulate the process by which cells reproduce. This is an evolutionarily conserved process which is conserved in all living organisms from unicellular microorganisms to plants to man which ensures growth, differentiation and the transmission of genetic material from one cell to the next generation. The earliest description of what could be considered the theory of the cell cycle dates back to the late nineteenth century when a number of biologists studying early developmental processes realized that all of the cells which later create tissues, organs and organisms arise from the differentiation of a single cell following fertilization (86). Later improvements to microscopy lead to further characterizations of what happens to the chromosomes during mitosis and along with the rediscovery of
9 rediscovery of Mendel’s Laws on Inheritance, tightly linked hereditary transmission of genetic material to the cell growth and division (86). Following Watson and Crick’s discovery of the
DNA double-helix structure and further work describing the time at which DNA replication occurs during the cell cycle, the cell cycle was broadly divided into the four phases which we use to describe the cell cycle today, S-phase, when DNA replication occurs, M phase or mitosis, when cells and their genetic material are split and divided, and the gap phases (G1 before S-phase and G2 following S-phase, but before mitosis) (86–88). Cell fusion experiments performed by
Rao and Johnson in 1970 determined that there is a one-way directionality in the cell cycle where cells proceed from G1 to S to G2 to M (89). During the late 1970’s through the 1980’s the essential molecular determinants of cell cycle progression were described (90–99). For their contributions in describing the molecular determinants of cell cycle control, Leland Hartwell,
Timothy Hunt and Paul Nurse were awarded the 2001 Nobel Prize in Physiology or Medicine.
Our understanding of the molecular determinants of cell cycle control continues to expand to this day as we determine how the integration of myriad factors and signaling pathways combine with the essential cell cycle control machinery to promote cell growth and division and how these processes, when aberrant, contribute to disease.
Mechanisms of Control
The decision to commit to the cell cycle begins during G1 phase where the decision to proceed forward is regulated by sensing external growth factor signals from the environment and from internal signaling pathways which orchestrate the creation of various proteins and ready the necessary nutrients and macromolecules needed to ensure a successful cell division. Until the necessary factors are in place, the start of S-phase is restrained by a checkpoint known as the
“restriction point” whereby cells either fully commit to completing the cell cycle, remain as they
10 are, or leave the cell cycle and enter a state of quiescence (G0) (100, 101). At a molecular level, it is conventionally believed that the liberation of the E2F family of transcription factors from the retinoblastoma protein (Rb) is essential for the transcription of a variety of genes necessary to both accelerate and maintain the liberation of E2F from Rb and trigger the processes needed for
DNA synthesis to occur. In order for E2Fs to become liberated from Rb, Rb must be phosphorylated by cyclin D1 bound to cyclin-dependent kinases 4 and 6 (CDK4-6) (102). This phosphorylation event causes a conformational change in the Rb protein which frees E2F. This process is further maintained and accelerated by the increase in transcription of cyclin E1 (a transcriptional target of E2F), which when bound to CDK2 can promote the hyperphosphorylation of Rb. In order for this process of Rb phosphorylation to begin, cyclin D1 levels must rise above the levels of the cyclin-dependent kinase inhibitors (CDIs) p16INK4A and
WAF1 p21 (103–107). Cyclin D1 levels typically peak during mid-G1, thus satisfying this requirement. Similarly, p27KIP1, functions to restrain cyclin E1-CDK2 activity (107, 108).
CDH1 Enzymes of the UPS also modulate the G1/S transition in a variety of ways. APC/C restrains the G1/S transition by ubiquitinating a number of substrates involved in the G1/S transition (109).
CyclinF Loss of CDH1 accelerates the progression from G1/S. Similarly, SCF promotes the transition from G1/S through targeting CDH1 for degradation, thus committing the cells to the S-
SKP2 phase transition (110, 111). Additionally, SCF promotes the G1/S transition through the degradation of the previously mention CDIs, p27KIP1 and p21WAF1 (112–114).
Once triggered, DNA synthesis occurs as multiple origins are fired which results in the unwinding of DNA and the recruitment and activation of replicative polymerases which generate a high-fidelity duplication of the cells genetic material (115). Any DNA damage caused during replication of the DNA will trigger an Intra-S-phase checkpoint which halts further progression
11 until the DNA damage is recognized, repaired and fixed. There are a variety of DNA damage sensing and repair mechanisms activated under a variety of conditions at this point and also in G2 before the transition to mitosis occurs (116).
Following S-phase, the cells enter G2 where similar to G1, many transcriptional, translational, and post-translational process connecting a variety of signaling networks come together to prepare the cell to divide into two daughter cells. The critical cyclin regulating the
G2/M transition is cyclin B1. Cyclin B1 binds to its cognate kinase partner CDK1 in the cytoplasm (117). At this point, a number of molecular events must occur as the cell proceeds towards mitosis to fully activate the cyclin B1-CDK1 complex. First, CDK1 must be phosphorylated on T161 by CDK-activating kinase (CAK) and inhibitory phosphorylation on sites T14 and T15, orchestrated by the kinases WEE1 and Myt1 (118–121), must be removed by
CDC25 (122–125). Second, active cyclin B1-CDK1 must be translocated from the cytoplasm to the nucleus in order to activate the processes necessary to complete mitosis (117). This largely occurs through phosphorylation events on cyclin B1 by kinases which include PLK1 (126, 127).
These phosphorylation events block nuclear export factors, such as CRM1, from binding and exporting cyclin B1 and promotes the association with cyclin B1 with nuclear import factors to promote its import into the nucleus (117, 128–130). While in the nucleus, cyclin B1-CDK1 activate further activate transcription factors such as B-MYB and FOXM1 (131–133) to facilitate the creation of many of the necessary proteins to begin mitosis and also induce changes in the microtubule network, actin microfilaments and the nuclear lamin (131, 134–136).
Mitosis, the final step in the cell cycle, can be separated into five distinct sub-phases
(prophase, prometaphase, metaphase, anaphase and telophase) based on the physical state of chromosomes and spindle attachments. During prophase, chromosome condensation begins and
12 is largely orchestrated by the localization and activity of a protein called condensin (137).
Cohesin is mostly removed from the arms of the chromosomes and the mitotic spindle beings to form as centrioles polarize within the cells and microtubules begin to polymerize from the centrosomes.
Prometaphase is marked by nuclear envelope breakdown. This allows chromosomes to be accessed by the microtubules which connect to chromosomes at the kinetochores and ensure that sister chromatids have microtubule connections that will allow them to segregate to opposite poles. The activity of CDK1 and PLK1 are crucial for regulating centrosome and microtubule dynamics needed during this stage as well as in prophase (102, 138).
During metaphase, the chromosomes align on the metaphase plate, the equator between spindle attachments. The spindle assembly checkpoint (SAC) is active during this period and ensures that chromosome microtubule and spindle attachments have been made and that the chromosomes are properly aligned. Mechanistically, unattached kinetochore promotes the creation of the mitotic checkpoint complex (MCC). This complex composed of the SAC proteins
MAD2, BUBR1 and BUB3 bind the anaphase-promoting complex/cyclosome (APC/C) substrate recognition protein CDC20. APC/CCDC20 triggers the proteolytic degradation of cyclin B1 and securin, which are essential for the transition into anaphase (139, 140). The degradation of cyclin
B1 inactivates CDK1, thus turning off the signal promoting the earlier stages of mitosis (141).
The degradation of securin releases a protease called separase which degrades the cohesin ring holding sister chromatids together, thus allowing for their separation during anaphase (139, 142,
143). When kinetochores are both correctly attached and under tension, a process regulated by the Aurora B kinase, the SAC signal is downregulated and the previously described proteolytic events are set in motion (139, 144).
13 During anaphase, the sister chromatids are pulled apart due to the degradation of cohesin, which previously kept sister chromatids attached, the shortening of kinetochore microtubules, which causes the chromatids to move towards the spindle poles and the movement of non- kinetochore microtubules which further accelerate movement apart. Proteolytic destruction of many of the proteins essential for the G2/M transition continues during this time as the
APC/CCDC20 gives way to APC/CCDH1 thus leading to the destruction of additional mitotic proteins such as FOXM1, PLK1 and CDC20 itself (145).
During telophase, the chromosomes have reached their respective poles and are fully separated. The nuclear membrane then reforms around the chromosomes and the chromosomes begin to expand. Mitosis ends with cytokinesis which sees the cytoplasm separated from one cell to two cells.
Cell Cycle Transcription by FOXM1
Background
The cell cycle transcription factor forkhead box protein M1 (FOXM1) will be a focus of the studies described in subsequent chapters. Here, I will give a general overview of FOXM1 biology, how FOXM1 is regulated and the implications that this has on the cell cycle. Additional detail specifically describing the regulation of FOXM1 by the UPS and the role of FOXM1 in disease will be described as this information, as well as the open questions remaining, are critical to understanding how the work described in the subsequent chapters fits in to the broader context of the field.
14 Overview of FOXM1 Biology
As previously mentioned, changes in transcriptional activity are central to establishing and ordering the distinct phases of the cell cycle as cells proceed from G1 to mitosis. One of these critical transcription factors is forkhead box protein M1 (FOXM1). FOXM1 was first cloned and characterized 25 years ago and is also known as MPP2, HFH-11, FKHL16, WIN and
Trident as it was uniquely named following independent cloning of various groups in the mid
1990’s (146–149). Further analysis of FOXM1 placed it within the forkhead box (FOX) family of evolutionarily conserved transcription factors (150). FOX family transcription factors are characterized by a common DNA binding motif with a preference for recognizing the
“TAAACA” nucleic acid sequence, though this is not as predictive for FOXM1 promoter binding than it is with other FOX family proteins (151). Expression of FOXM1 oscillates throughout the cell cycle where levels are very low through most of G1, begin to rise at the G1/S transition, continue to rise through G2 and peak during the G2/M transition where levels precipitously drop due to rapid proteolytic degradation caused by the APC/CCDH1 (152, 153). As such, FOXM1 is only expressed in proliferating cells where it functions as a proliferation- promoting transcription factor regulating the expression of a variety of genes associated with the
G2/M transition and to a lesser extent, the G1/S transition (133, 154–156). Due to its central role in orchestrating the expression of a transcriptional network of genes needed for cell cycle progression, knockdown of FOXM1 is sufficient to slow proliferation whereas overexpression of
FOXM1 is sufficient to drive proliferation (157–159).
15 Post-Translational Regulation of FOXM1 Through the Cell Cycle
FOXM1 activity is tightly regulated throughout the cell cycle by a variety of post- translational modifications including phosphorylation, ubiquitination, SUMOylation and acetylation (Figure 1.2). During G1, FOXM1 levels are low due to proteolytic degradation caused
CDH1 by the APC/C (152, 153). As cells proceed closer to the G1/S transition, FOXM1 levels begin to rise, likely due to the inactivation of APC/CCDH1 at this point (110). Before becoming activated, the inhibitory N-terminal domain of FOXM1 must be phosphorylated, which releases it from association with its C-terminal transactivation domain (160). Though not entirely clear which kinases specifically phosphorylate the N-terminal domain to promote FOXM1 activation, it has been reported that FOXM1 is phosphorylated by cyclin D1-CDK4/6 and cyclin E1-CDK2 during the G1/S transition (161, 162). This phosphorylation accomplishes a number of goals.
First, it liberates FOXM1 from association with Rb (161, 163). Second, it allows for the recruitment of the p300/CBP transcriptional coactivator complex to FOXM1 which promotes
FOXM1 acetylation (161, 164). This enhances the ability of FOXM1 to transcribe a number of genes essential for the G1/S transition and DNA damage repair (when further phosphorylated by
CHK2) (162, 165, 166). Third, the phosphorylation of FOXM1 seems to promote its stabilization, though it remains unclear what factors may be modulating its stability during this time (162). FOXM1 dephosphorylation through binding to the phosphatase PP2A during this time restrains premature FOXM1 activation during this time (167). Interestingly, FOXM1 seems to promote the ubiquitination of proteins necessary to promote the G1/S transition though the enhancing the transcription of SKP2 and CKS1, two subunits of the SCF ubiquitin ligase complex (155).
16 During S-phase, FOXM1 levels continue to rise and nuclear translocation is enhanced through phosphorylation from the Raf/MEK/MAPK signaling pathway (168). Additionally,
FOXM1 is a downstream target of the Wnt-signaling pathway where Wnt signaling promotes
FOXM1 association with β-catenin and subsequent nuclear translocation and activation of Wnt- target gene expression (169). Also during late S-phase, FOXM1 is bound by the B-Myb-MuvB complex which targets FOXM1 to the promoters of mitotic genes containing the CHR sequence and may also increase the affinity of FOXM1 for its consensus “TAAACA” binding sequence which it has been demonstrated to have low affinity for (151, 170).
FOXM1 is at its peak in terms of both expression and activity throughout G2 through the prometaphase stage of mitosis. FOXM1 activity is heavily regulated by phosphorylation at this point. The C-terminal transactivation domain (TAD) of FOXM1 contains a multitude of phosphorylation sites and is phosphorylated by cyclin A1-CDK1, cyclin B1-CDK1, PLK1 and
MELK (171–174). Hyperphosphorylation of FOXM1 by these kinases is essential for FOXM1 transcriptional activation during this time and the targets of FOXM1 transcription are essential to executing the necessary processes to conduct mitosis and ensure that chromosomes segregation occurs without error (133, 155, 156, 171). The phosphatase CDC25B has also been shown to interact with FOXM1 at this time (161). FOXM1 is also SUMOylated during this phase of the cell cycle, though it is unclear if this modification activates or inhibits FOXM1 transcriptional activity (175–177). Following the transition from prometaphase to anaphase, FOXM1 and many of its transcriptional targets are rapidly ubiquitinated by APC/CCDH1 (152, 153). Recent work has demonstrated that the APC/CCDH1-mediated destruction of FOXM1 can be further enhanced through O-GlcNAcylation in a manner that depends on the activity of the deacetylase sirtuin 1
(SIRT1), though the direct mechanism of action with respect to FOXM1 remains unclear (178).
17 Ubiquitin-dependent Regulation of FOXM1
Far less is known about the regulation of FOXM1 by ubiquitin when compared to what is known about phosphorylation-dependent regulation of FOXM1. As mentioned previously, ubiquitin-dependent degradation of FOXM1 orchestrated by interaction with the APC/CCDH1 regulates the destruction of FOXM1 following the onset of anaphase and continues throughout
CDH1 G1 when APC/C is active (152, 153). The N-terminus of FOXM1 contains the KEN box
(amino acids K-E-N) and D-box (amino acids R-X-X-L) sequences which are recognized by
APC/C and lead to K11 polyubiquitination on FOXM1 (179). Additionally, FOXM1 is ubiquitinated by two SCF E3 ubiquitin ligases, SCFFBXL2 and SCFFBXO31, the latter of which seems to primarily be active during G2/M to restrain FOXM1 levels (180, 181). SUMOylated
FOXM1 has been demonstrated to be ubiquitinated by the E3 ubiquitin ligase RNF168 following epirubicin-induced DNA damage (182). FOXM1 is also ubiquitinated by SCFFBXW7 when
FOXM1 is phosphorylated by the GSK3-Axin complex. This phosphorylation and ubiquitination event is inhibited by Wnt signaling (183).
At the time the studies described in Chapter 3 began, there were no DUBs linked to
FOXM1 despite the fact that it is well established that FOXM1 levels are heavily regulated by ubiquitin. OTUB1 has been shown to promote the stability of FOXM1 by catalyzing the removal of K48-polyubiquitin chains (184, 185). USP5 has been shown to interact with FOXM1 following Wnt signaling activation, thus promoting its stability through deubiquitination (183).
Intriguingly, FOXM1 seems to be targeted by naturally occurring thiopeptides which also happen to be general proteasome inhibitors (186). In two independent studies, thiostrepton and siomycin A were demonstrated to reduce the transcriptional activity of FOXM1 (187, 188).
18 Work remains to be done to determine the exact mechanism of binding of these thiopeptides to
FOXM1 and to separate the effects on FOXM1 from the general proteotoxic stress induced from the general proteasomal inhibition that these molecules induce.
As evidenced by what has been discovered about FOXM1 ubiquitination thus far, these processes are quite complex and involve the interplay of multiple signaling pathways. Much work remains to identify additional E3 ubiquitin ligases regulating FOXM1 abundance through the cell cycle and what signaling pathways are involved in modulating these effects.
FOXM1 in Cancer
Due to its role as a proliferation-promoting transcription factor, the activity and overexpression of FOXM1 has been linked to a variety of cancers. FOXM1 is commonly identified as one of the most commonly upregulated genes in a variety of cancers. The FOXM1 gene is located on 12p13 and copy number amplification at this site is common to a number of cancers (171). Additionally, FOXM1 expression levels typically correlate with poor prognosis and survival across most solid tumors (189). Importantly, in a multi-platform, high-throughput analysis of the genes most amplified in basal-like breast cancers and ovarian cancers carried out by The Cancer Genome Atlas Network, it was uncovered that FOXM1 is one of two master regulators which broadly promotes the amplification of most of the differentially expressed genes in those cancer subtypes (190). However, the correlation with FOXM1 and cancer is more than just a by-product of its general association with proliferative cells. Forced FOXM1 overexpression or knockdown has been demonstrated to drive or retard cancer progression, respectively. Sustained expression of FOXM1 in murine models of colorectal, prostate, and lung cancer as well as hepatocellular carcinoma was enough to significantly accelerate the
19 development, proliferation and size of tumors formed and the knockdown or knockout of
FOXM1 had the reciprocal effect in these same models (191–194).
In regard to cancer treatment, FOXM1 has also been demonstrated to modulate the response to chemotherapy. This is particularly concerning because chemotherapy remains the primary means of treatment for the management of many cancers which lack more targeted treatment options and many of these cancers are likely to overexpress FOXM1. FOXM1 overexpression has been demonstrated to confer resistance to paclitaxel, a microtubule- stabilizing drug which induces cell death through disrupting mitosis (195). FOXM1 knockdown increases sensitivity to paclitaxel. A FOXM1 transcriptional target, stathmin, a tubulin- destabilizing protein, was suggested to be involved in the mechanism of paclitaxel sensitivity
(195). FOXM1 has also been demonstrated to promote resistance to platinum-based chemotherapy. Platinum-based chemotherapies induce cell death by crosslinking with DNA to form adducts that result in DNA damage. FOXM1 expression was noted to be higher in MCF7 breast cancer cell lines which had been made resistant to the platinum-based chemotherapy, cisplatin (196). Knockdown of FOXM1 in these cisplatin-resistant cells resensitized the cells to cisplatin. Mechanistically, FOXM1 seems to promote resistance to cisplatin through the activation of DNA damage repair mechanisms (196). Additionally, FOXM1 expression promotes resistance to the DNA intercalating chemotherapy epirubicin through a mechanism that hinges on FOXM1 regulation by ATM and p53 (197).
Due to importance of FOXM1 in cancer initiation, progression and altering treatment outcomes, a great deal of attention has surrounded the promise of direct FOXM1 inhibition in treating cancer (158, 198). Unfortunately, this strategy is fraught with difficulties as transcription factors typically make poor drug targets. Most of the pharmacological inhibitors that have been
20 demonstrated to have an effect on FOXM1 bind to FOXM1 in a way that impairs the ability of
FOXM1 to bind DNA (189, 199–202). These molecules tend to be large peptides which are typically difficult to deliver in a clinical setting. Additionally, many of these compounds are validated against the forkhead box DNA binding motif. Not only is this common to all FOX family transcription factors, FOXM1 has been shown to partner with other transcriptional coactivators to bind promoters that would not contain consensus FOXM1 DNA binding sites
(169, 203, 204). Other strategies which have a more indirect effect on FOXM1 stability may be more pharmacologically viable. Small molecules which inhibit the transcription of FOXM1 mRNA or the nuclear translocation of FOXM1 have recently been demonstrated (189, 205–207).
In closing this introduction, I would like to emphasize that the path to realizing the promise of targeting FOXM1 as a therapeutic target lies in understanding the network of effects which regulate FOXM1. Not only will understanding the networks of proteins and signaling pathways that conspire together to activate FOXM1 throughout the cell cycle, inform our understanding of how FOXM1 is aberrantly activated or overexpressed in cancers where there is not a simple genetic explanation available to explain FOXM1 upregulation, but many of these proteins and signaling pathways may make excellent targets for small-molecule inhibitors.
Notably, little is known about the ubiquitin-dependent regulation of FOXM1 relative to what is known about the phosphorylation-dependent regulation of FOXM1. In Chapter 2, I will describe the ubiquitin-dependent regulation of FOXM1 by CUL4-VPRBP. Not only does this E3 ubiquitin ligase regulate the stability of FOXM1, but it also seems to activate FOXM1 transcriptional activity, thus providing additional context to FOXM1 regulatory mechanisms that have never been previously described. In Chapter 3, I will describe the identification and characterization of FOXM1 by the DUB USP21. Not only does USP21 stabilize FOXM1
21 through the removal of polyubiquitin chains, but the amplification of the genomic region containing the USP21 gene may partly explain why FOXM1 is elevated in cancers with a 1q21-
23 copy number amplification as is commonly observed in basal-like breast cancers. Moreover, it will be demonstrated that USP21 knockdown can slow proliferation in a FOXM1-dependent manner and that the targeting of USP21 can sensitize breast cancer cells to paclitaxel and may prove a viable therapeutic strategy to use in conjunction with chemotherapy to target cancers driven by FOXM1 amplification.
22
Figure 1.1 Overview of the ubiquitin-proteasome system. A graphical depiction of substrate ubiquitination by a cullin-RING ubiquitin ligase. The cascade begins when ubiquitin (U) is conjugated onto an E1 ubiquitin-activating enzyme (E1) in an ATP-dependent reaction. This ubiquitin molecule is then transferred onto an E2 ubiquitin- conjugating enzyme (E2). This enzyme is able to be incorporated into a culling-RING ligase complex (RING, CUL, ROC1/2) which also contains an E3 ubiquitin ligase substrate adaptor (E3). When activated by NEDD8 (N8), substrates are able to be ubiquitinated as ubiquitin moves from the E2 to the substrate attached to the E3. These proteins can interact with ubiquitin-binding proteins (UBPs), be degraded by the 26S proteasome (26S) or interact with deubiquitinating enzymes which remove conjugated ubiquitin. Cullin-RING ligases can also be inactivated by the COP9 signalsome (CSN), which deNEDDylates the complex. CAND1 can also inactivate the cullin scaffold by competing for substrate adaptor binding. NEDD8 and substrate binding inhibit the activity of CAND1 and CSN, respectively.
23
Figure 1.2 FOXM1 post-translational modification landscape. A graphical overview of the known proteins and sites which post-translationally modify FOXM1. E3 ubiquitin ligases and deubiquitinating enzymes in dark green were added as a result of the work described in the subsequent chapters. Matching colors between kinases and phosphorylation sites denotes sites that are uniquely modified by that kinase. Black color for phosphorylation sites denotes site can be modified by multiple kinases or kinase responsible for phosphorylation is unknown.
24
CHAPTER 2: VPRBP/DCAF1 REGULATES THE DEGRADATION AND NONPROTEOLYTIC ACTIVATION OF THE CELL CYCLE TRANSCRIPTION FACTOR FOXM11
Introduction
Changes in gene expression combined with targeted protein degradation dynamically shape the protein landscape. Gene expression is coordinated by transcription factors that specify genes for activation and cofactors that modulate transcription factor activity or alter the local chromatin environment. Post-translational modifications (PTMs) play a crucial role in transcriptional dynamics. Phosphorylation, acetylation, methylation and ubiquitylation of histone proteins is well-studied, and contributes significantly to gene expression dynamics (208).
Similarly, post-translational modification of transcription factors plays an important role in regulating genome output.
FOXM1 is an oncogenic, cell cycle regulated transcription factor that was discovered as both a marker and key mediator of cell proliferation (147, 157, 209). Subsequent work clarified the importance of FOXM1 in proliferation through its role in cell cycle progression (210).
FOXM1 controls the mitotic transcriptional program and its depletion significantly impairs normal mitotic entry and progression (156, 211–213). In addition, FOXM1 and its transcriptional network have been associated with numerous cancers (158, 210). Notably, FOXM1 is the key regulator of a proliferative gene expression signature found in high-grade serous ovarian cancer
1 This chapter previously appeared as an article in Molecular and Cellular Biology. The original citation is as follows: X. Wang, A. Arceci, K. Bird, C. A. Mills, R. Choudhury, J. L. Kernan, C. Zhou, V. Bae-Jump, A. Bowers, M. J. Emanuele, VprBP/DCAF1 regulates the degradation and nonproteolytic activation of the cell cycle transcription factor FoxM1. MCB. 37, 13, e00609-16 (2017).
25 (HGSOC), basal-like breast cancers and uterine serous carcinomas (190, 214, 215). In HGSOC the FOXM1 signature is found in ~90% of patient tumors (214).
FOXM1 activity peaks in G2/M phase, consistent with its role in dictating mitotic gene expression, and several kinases have been implicated in its activation (155, 170, 204, 211, 216,
217). FOXM1 is repressed by an intra-molecular interaction with its amino-terminal domain and this inhibition is relieved by cyclin-dependent kinase (CDK) phosphorylation (160). In addition, the cell cycle kinases MELK and PLK1 can activate FOXM1 (161, 172–174, 177).
In addition to phosphorylation, post-translational addition of ubiquitin is utilized to control gene expression. The oncogenic transcription factor c-Myc highlights the complex role of ubiquitin in transcriptional regulation (218). Myc is targeted for proteolysis by several E3 ubiquitin ligases, including a SCF-type Cullin Ring Ligase (CRL) and the substrate receptor
SKP2 (SCFSKP2) (219, 220). Whereas protein ubiquitylation and degradation are most often considered inactivating events, unexpectedly, SKP2 activates Myc dependent transcription (219,
220). Ubiquitylation dependent activation of Myc is further borne out by studies using a lysine- less version that cannot be ubiquitylated and is deficient in activating transcription (221). The role of ubiquitylation in transcriptional activation builds on pioneering studies on the VP16 transcription activation domain whose activation in yeast requires an SCF ligase together with its substrate receptor Met30 (222). Furthermore, the ability of the ubiquitin machinery to activate transcription is corroborated by regulation of the human estrogen receptor (ERa) and its coactivator, SRC-3/AIB1, whose degradation is coupled to activation (222–225). Together, these studies highlight the complex role ubiquitin plays in transcriptional control.
The role of ubiquitin ligases in activating FOXM1 has not been studied. We recovered
FOXM1 in a global screen for substrates of the modular CRLs, which represent the largest E3
26 ligase family in humans (6). CRL assembly is based on a common molecular scaffold and relies on a cullin backbone that simultaneously engages substrates and E2 ubiquitin-conjugating enzymes. Cullin4-based ligases (CRL4) use either of two highly related cullin proteins, CUL4A or CUL4B, that bind to the triple-β-propeller protein DDB1 and simultaneously recruit specific proteins to the enzyme complex for ubiquitylation (Figure 2.1A) (226). Human CUL4A and
CUL4B are highly similar (75% amino acid similarity), however CUL4B has an amino-terminal extension and localizes exclusively to the nucleus, whereas CUL4A is both nuclear and cytoplasmic (227). Importantly, CRL4 function has been linked to chromatin regulation, cell cycle, viral infection, and the DNA damage response (226).
More than 50 CRL4 substrate receptors, termed DCAFs or DWD proteins (DDB1 and
CUL4 Associated Factors; DDB1 binding WD40 proteins), have been identified (26, 228, 229).
VPRBP/DCAF1 is a nucleus-localized CRL4 substrate receptor named for its ability to bind the
HIV accessory protein Vpr (and Vpx) following viral infection (36). Ectopic Vpr expression in
VPRBP human cells triggers a G2 arrest that is dependent on CRL4 (47). Significantly, VPRBP associates with chromatin only during G2/M phase of the cell cycle (230). Knockout of VPRBP in mice causes embryonic death prior to embryonic day 7.5 (E7.5), and conditional inactivation of VPRBP in mouse cells or depletion using RNA interference (RNAi) in human cells, produces cell cycle defects (230). Despite its importance in cell cycle control and development,
CRL4VPRBP has few known substrates and it remains unclear how it contributes to cell cycle progression. Endogenous CRL4VPRBP substrates include the methylcytosine dioxygenase Tet2
(231) and the replication regulator Mcm10 (232). Other proteins are targeted for ubiquitylation only in response to HIV infection and include the phosphohydrolase SAMHD1 (40). VPRBP has been linked to the NF2 tumor suppressor and YAP dependent transcription, suggesting a role in
27 transcriptional regulation and cancer (233). Here, we describe a role for VPRBP in controlling both the degradation and activation of FOXM1.
Results
FOXM1 Stability is Regulated by CRL4VPRBP
Using a fluorescence-based genetic reporter system termed global protein stability
Profiling (GPS) we previously searched for substrates of the CUL4-based Cullin-RING ligase
(CRL4) (6). The GPS expression system relies on a viral vector that expresses a bicistronic mRNA encoding both DsRed and an enhanced green fluorescent protein-open reading frame
(EGFP-ORF) fusion protein. Using this system, we infer relative changes in the stability of
EGFP-ORF fusions by examining the ratio between EGFP and DsRed fluorescence using flow cytometry. DsRed normalizes for expression of the reporter cassette on a single cell basis. Using this system, we screened a pooled library of 293T cell expressing more than 13,000 individual
EGFP-ORF fusion proteins (one ORF per cell). A schematic overview of the GPS screening system is described in detail elsewhere (6, 234).
To identify proteins whose stability are controlled by CRL4, we turned off the ligase by introducing a dominant-negative construct targeting CUL4 and compared this condition to a negative-control vector (6). GPS-library cells from these two conditions were sorted into bins based on their EGFP/DsRed ratio using fluorescence activated cell sorting (FACS). Genomic
DNA (gDNA) was isolated from sorted cells in each bin, and ORFs were amplified from gDNA using PCR primers designed against the viral backbone. Amplified ORF DNA was labeled and hybridized to custom-designed DNA microarrays containing multiple independent probes per gene. Quantifying the distribution of probe signals across bins allows us to infer changes in the
28 stability for EGFP-ORFs (6). For example, a shift in the probe distribution to higher-numbered bins suggests that cells expressing a particular EGFP-ORF showed an increase in their
EGFP/DsRed ratio, indicative of an increase in the stability of the EGFP-ORF fusion protein.
Three of the four probes corresponding to FOXM1 showed a shifted distribution after dominant-negative CUL4 treatment, suggesting that EGFP-FOXM1 was stabilized by CRL4 inactivation (Figure 2.1B). These three probes showed a highly consistent distribution across bins suggesting that they accurately report the distribution of EGFP-FOXM1 expressing cells and that
FOXM1 stability is regulated by CRL4.
To validate endogenous FOXM1 as a CRL substrate we treated cells with MLN4924, a pharmacological small-molecule inhibitor that impairs CRL activation by interfering with the neddylation cascade (235). FOXM1 abundance was increased after a 4 h MLN4924 treatment in both 293T and U2OS cells (Figure 2.1C). The slower migrating, neddylated form of CUL4 was undetectable after MLN4924 treatment, showing that neddylation was impaired. We next determined if FOXM1 is regulated by CUL4. We treated cells with small interfering RNA (siRNA) targeting either CUL4A, CUL4B, DDB1 or with control oligonucleotides targeting firefly luciferase (FF). Seventy-two hours after transfection cells were harvested for immunoblot.
Depletion of DDB1 and CUL4B lead to a reproducible increase in the abundance of endogenous
FOXM1, whereas depletion of CUL4A had a weaker and less consistent effect (Figure 2.1D).
Consistently, depletion of DDB1 and CUL4B increases the abundance of Myc-FOXM1, which is stably expressed form a heterologous promoter, and again, CUL4A had no effect (Figure 2.1E; the underline corresponds to the antigen used for blotting (e.g. Myc-FOXM1)). Taken together, these data suggest FOXM1 stability is regulated by CRL4.
29 The two best-studied CRL4 substrate receptors implicated in cell cycle control are CDT2 and VPRBP. CDT2 engages substrates through a degron embedded within a PCNA-interacting peptide motif (PIP-box) (236). However, FOXM1 lacks a PIP-box. We therefore depleted VPRBP using siRNA and measured FOXM1 abundance by immunoblot. Depletion of VPRBP increased endogenous FOXM1, as well as the known substrate MCM10, in both HeLa and U2OS cells
(Figure 2.1F).
FOXM1 is degraded in G1 phase by the anaphase-promoting complex/cyclosome-CDH1
(APC/CCDH1) ubiquitin ligase (153, 237). To determine if VPRBP regulates FOXM1 at other times during the cell cycle and to rule out the possibility that differences in FOXM1 abundance where due to changes in cell cycle progression, we treated U2OS cells with siRNA targeting VPRBP and then synchronized cells using thymidine in accordance with a previously established protocol
(238). Cells were harvested 0, 2, 4 and 6 h after release and analyzed by immunoblot. FOXM1 was consistently increased in the VPRBP-depleted cells (Figure 2.2B), suggesting that VPRBP regulates FOXM1 degradation and that the increase in FOXM1 abundance is independent of gross cell cycle changes. Two independent siRNA oligonucleotides targeting VPRBP produced a similar increase in FOXM1 levels in S-phase synchronized cells (Figure 2.2C). Similarly, we synchronized HCT116 cells in early S-phase, late S-phase/G2 and G2/M using thymidine, thymidine block and release, and nocodazole, respectively. FOXM1 was increased in synchronized, VPRBP-depleted cells, indicating that VPRBP controls FOXM1 independent of its effect on the cell cycle (Figure 2.2D).
Next, we measured FOXM1 stability/half-life in control and VPRBP-depleted cells that were synchronized in S-phase, 2 h after thymidine release. VPRBP depletion increased FOXM1 stability after treatment with the protein synthesis inhibitor cycloheximide (CHX) (Figure 2.1G).
30 This is a reproducible result observed across several experiments. The increase in FOXM1 is evident in otherwise untreated cells after the addition of CHX (Figure 2.2E). While the mechanism underlying this phenomenon remains unclear, it has long been appreciated that some proteins can increase following treatment with CHX (239).
FOXM1 Interaction and Ubiquitylation by VPRBP
We determined if VPRBP binds to FOXM1 in 293T cells transiently transfected with
Myc-FOXM1 and hemagglutinin-tagged VPRBP (HA-VPRBP) to examine the possibility that
FOXM1 is a direct CRL4VPRBP substrate. Cells were treated with the proteasome inhibitor
MG132 for 4 h prior to lysis and co-immunoprecipitation (co-IP) to promote the interaction between E3 ligase and substrate (240, 241). Following anti-HA-VPRBP IP, we detected an interaction with FOXM1, and this interaction was enhanced by treatment with MG132 (Figure
2.3A and Figure 2.4A). The interaction was also detectable when Myc-FOXM1 was immunoprecipitated (Figure 2.4C). In addition, we immunoprecipitated endogenous FOXM1 from 293T cell lysates and detected endogenous VPRBP (Figure 2.3B). Finally, we determined if
FOXM1 and VPRBP can directly interact. We immobilized bacterially purified hexahistidine- tagged FOXM1 (FOXM1-6HIS) on nickel-agarose beads and mixed them with purified,
[35S]methionine-labeled VPRBP produced in vitro using purified transcription and translation machinery. VPRBP bound to beads that had immobilized FOXM1-6HIS but was observed only in the flow through of control beads, strongly suggestive of a direct interaction (Figure 2.3C).
To determine if VPRBP can regulate FOXM1 ubiquitylation we ectopically expressed
6HIS-ubiquitin in 293T cells with and without HA-VPRBP. Cells were treated with MG132 prior to lysis in strong denaturing buffer (6M Guanidine-HCl) and 6HIS-ubiquitin conjugates were
31 isolated on nickel-agarose. VPRBP expression enhanced the ubiquitylation of endogenous
FOXM1, measured by immunoblotting for FOXM1 (Figure 2.3D). Likewise, VPRBP significantly increased the ubiquitylation of ectopically expressed Myc-FOXM1 (Figure 2.3E). Therefore,
VPRBP regulates the abundance, stability and ubiquitylation of FOXM1.
To identify the domain in FOXM1 that interacts with VPRBP we synthesized a series of constructs encoding 160 amino acid (aa) fragments of the FOXM1 protein spanning the length of its largest known open reading frame. We tested their ability to interact by expressing full-length
HA-VPRBP and Myc-FOXM1 fragments in 293T cells and analyzing precipitates after anti-HA
IP. A fragment of FOXM1 spanning amino acids 321-480 (FOXM1321-480) interacted most strongly with VPRBP (Figure 2.4B). The same Myc-FOXM1321-480 fragment bound to VPRBP when Myc was precipitated (Figure 2.4C). We conclude that the interaction between VPRBP and
FOXM1 is dependent on the region of FOXM1 spanning amino acids 321-480.
To identify the degron sequence motif in FOXM1 we examined molecular and structural data of a known VPRBP substrate. SAMHD1 is targeted by CRL4VPRBP after HIV infection, and a ternary complex between VPRBP, SAMHD1 and the viral accessory protein Vpx has been crystalized (Figure 2.3F) (42). Importantly, the amino acids in SAMHD1 that mediate binding to
VPRBP and degradation have been mapped (41, 42). A segment between amino acids 615 and
625 in SAMHD1 contributes significantly to VPRBP binding and several residues in this region are critical for its degradation. We looked for matching sequences in FOXM1 between residues
321 and 480 and identified a region of similarity between residues 414 and 422 (Figure 2.3F).
We synthesized three mutant versions of the same 321-480 region, making amino acid substitutions that correspond to residues which we predicted would affect recognition by
VPRBP. These substitutions in FOXM1 included changed of RV to AA (aa 416 and 417), RI to
32 AA (aa 418 and 419), and K to A (aa 422). We tested the ability of each version of the fragment to bind HA-VPRBP by co-IP following treatment with proteasome inhibitors to normalize protein levels across IPs. We found that alanine substitutions at residues 418 and 419 (RI to AA) and 422 (K to A) impaired binding to VPRBP (Figure 2.3G). Notably, K622 in SAMHD1, corresponding to K422 in FOXM1, directly contacts VPRBP in the crystal structure and is required for SAMHD1 degradation (42).
We next tested the stability of these fragments by CHX chase. Significantly, fragments that showed reduced binding also had higher basal expression (note these experiments were not done in proteasome inhibitors) and increased stability relative to the wild type (WT) fragment when expressed in 293T cells (Figure 2.3H). We observed a similar increase in the stability of full-length FOXM1 when amino acids in the putative degron motif were changed (RVRIAPK to
AAAAAPA; the half-life increased from 1.0 to 1.5 h). Minor changes, specifically K422A, similarly extended the half-life (Figure 2.2F and 2.2G). These data are suggestive that this motif sequence, R-I/V-X-X-(X)-K, represents a putative VPRBP degron. Similar sequences were found in established substrates TET2 and MCM10 (Figure 2.3F) (231, 232).
VPRBP is a FOXM1 Activator
To address the contribution of VPRBP to FOXM1 activity we utilized a well-characterized
FOXM1 luciferase reporter construct (6xDB) (161). We measured reporter activity following transient expression of FOXM1 and VPRBP in 293T cells. Expression of FOXM1 increased reporter activity as expected (Figure 2.5A). Remarkably, we also observed a strong, dose- dependent increase in reporter activity in response to VPRBP when co-expressed with FOXM1
(Figure 2.5A). We compared the ability of VPRBP to activate FOXM1 activity to the known
33 coactivator MELK (174). VPRBP activated the reporter to a greater extent than MELK when transfected in equal amounts (Figure 2.5B). Furthermore, VPRBP activates the 6xDB reporter in the absence of ectopically expressed FOXM1 (Figure 2.5B). We also measured the effect of
VPRBP on an unrelated luciferase reporter controlled by the oxidative stress response transcription factor Nrf2. FOXM1 and VPRBP expression did not affect the Nrf2 luciferase reporter, whereas strong activation was achieved by Nrf2 expression (Figure 2.6A).
We next expressed the 6xDB reporter with increasing amounts of VPRBP and analyzed luciferase activity, cell cycle dynamics and protein abundance all in the same experiment (Figure
2.5C). VPRBP activated reporter activity at both 24 and 48 h post-transfection in a dose-dependent manner (Figure 2.5C.1). Importantly, the abundance of FOXM1 remained unchanged at both time points and concentrations of VPRBP (Figure 2.5C.2). This demonstrates that the change in reporter activity is not due to changes in the overall level of FOXM1. It also suggests that VPRBP does not activate FOXM1 by triggering its degradation, as is the case for Myc activation by SKP2 (219,
220). We examined the cell cycle using propidium iodide staining and analysis by flow cytometry and found no significant changes at either time point relative to controls (Figure 2.5C.3). We therefore conclude that VPRBP activates FOXM1, and that activation is independent of FOXM1 abundance, and is not due to gross changes in cell cycle dynamics.
To directly interrogate the consequence of VPRBP depletion on FOXM1 activity, we measured the transcript levels of FOXM1 target genes in synchronized U2OS cells (to alleviate cell cycle effects) that were depleted of VPRBP. First, U2OS cells were synchronized in G2/M following an 8 h release from thymidine block. The mRNA from control and VPRBP-depleted cells was isolated and FOXM1 target gene expression was measured using quantitative reverse transcription PCR (RT-qPCR). The expression of several FOXM1 target genes was reduced in
34 VPRBP depleted cells, including CDC25B, CCNB1, PLK1 and CENPF (Figure 2.7A). The expression of VPRBP was reduced by siRNA treatment as expected; however, the level of FOXM1 mRNA was unchanged. We performed a similar experiment by treating cells with siRNA targeting
FF, FOXM1, and VPRBP, and then blocking cells in mitosis with nocodazole. Mitotic cells were specifically isolated by shake-off and we found that the expression of CDC25B, PLK1 and CCNB1 were reduced to a similar extent in both FOXM1 and VPRBP-depleted cells relative to controls
(Figure 2.6B). We therefore conclude that VPRBP activates the transcription of FOXM1 target genes during G2/M phase of the cell cycle.
FOXM1 has been implicated in mitotic entry and progression (211, 212). We next examined the role of VPRBP in mitotic entry since its depletion reduces the expression of several
FOXM1 target genes. U2OS cells were treated with either control or VPRBP siRNAs and after 72 h were fixed and processed for phospho-histone H3 (P-H3) immunostaining to mark mitotic cells.
Imaging was performed and the percent of P-H3 positive cells was determined. 8% of cells were in mitosis in control depleted populations (Figure 2.7B). Depletion with two independent VPRBP siRNAs significantly reduced the percentage of P-H3 positive mitotic cells, to 2.8% and 0.5%
(Figure 2.7B). Thus, VPRBP depletion blocks the accumulation of mitotic cell cycle genes and prevents M-phase entry, consistent with a role for VPRBP in activating FOXM1.
We determined if VPRBP localizes to the promoters of FOXM1 target genes since it binds
FOXM1 and regulates FOXM1 target gene expression. U2OS cells were synchronized in G2/M phase and processed for chromatin immunoprecipitation (ChIP). Immunoprecipitation was performed in parallel with antibodies to control IgG, FOXM1 and VPRBP. We detected an enrichment of VPRBP and FOXM1 at the promoters of previously characterized FOXM1 promoters, including PTMS, BRCA2 and FZR1 (217). In addition, VPRBP localized to a lesser
35 extent to the FOXM1 targets CENPF and CDK1 (Figure 2.8A). Thus, FOXM1 and VPRBP co- localize to FOXM1 target promoters.
Myc and ERα are activated by ubiquitylation via their respective ligases (219, 220, 242).
We sought to gain similar mechanistic insight into the regulation of FOXM1 by VPRBP. We determined if two independent versions of VPRBP that cannot bind DDB1 (N909 and RARA) are able to activate FOXM1 reporter activity (231). The 6xDB reporter activity was increased using an amino-terminal fragment of VPRBP (N909) which cannot bind to the DDB1/CUL4 complex
(Figure 2.8B). Similar results were obtained with a mutant version of VPRBP (RARA) that is also impaired in DDB1/CUL4 binding (Figure 2.8B) (231). Further, the response of 6xDB to
VPRBPWT, VPRBPN909 and VPRBPRARA is dose dependent (Figure 2.6C). We conclude that
VPRBP activation of FOXM1 is independent of the CRL4 complex. These data also suggest that
FOXM1 activation by VPRBP is not due to a change in the abundance or stability of a secondary, unknown CRL4VPRBP substrate.
We next asked if VPRBP is still bound to CUL4B and DDB1 at a time in the cell cycle when FOXM1 is active. We introduced HA-VPRBP into 293T cells that were then treated with either nocodazole or dimethyl sulfoxide (DMSO). Cells were lysed and HA immunoprecipitates were analyzed by immunoblot. We found that the association of HA-VPRBP with endogenous
CUL4B, and to a lesser extent DDB1, was reduced in mitotic cells (Figure 2.8C, lane 6 vs 8).
However, there was no change in the interaction between VPRBP and FOXM1 in mitosis (Figure
2.8C and 2.8D). Thus, VPRBP association with the CRL4 complex is cell cycle regulated, and its dissociation coincides with the time at which FOXM1 is activated.
36 VPRBP Protein is Upregulated in High-Grade Serous Ovarian Tumors
The FOXM1 gene expression signature is upregulated in ~90% of high-grade serous ovarian cancers (HGSOC) (214). However, FOXM1 mRNA is only overexpressed in ~12% of tumors. To analyze the expression of VPRBP in HGSOC tumors we obtained surgically resected ovaries from HGSOC patients that had been histologically confirmed as serous ovarian cancer.
As controls we examined ovaries from women who underwent oophorectomy for reasons other than gynecological malignancy. FOXM1 levels were not elevated in HGSOC tumors relative to controls. Remarkably, VPRBP protein was upregulated in seven out of nine HGSOC tumors tested relative to control ovaries where VPRBP levels were low or undetectable (Figure 2.9A).
Further, the well-established FOXM1 target cyclin B was expressed in six of the seven tumors where VPRBP was increased. VPRBP mRNA is not overexpressed in HGSOC based on genomic analysis, suggesting that post-transcriptional mechanisms account for its overexpression (214).
Discussion
Ubiquitylation has long been implicated as a key regulator of transcription and chromatin regulation in human cells. Ubiquitin was identified due to its conjugation to the core histone
H2A, and only later was identified by Hershko and colleagues as part of intracellular protein degradation system (243). The yeast α-2 transcriptional repressor was one of the first identified in vivo targets of ubiquitin dependent protein degradation (244). Ubiquitin has been most well described for its role in inactivating target proteins through proteolysis. However, the emergence of complex ubiquitin chains of varying topologies and the identification of their roles in various
37 aspects of cellular physiology, beside protein degradation, illustrates the complex and sometimes paradoxical role the ubiquitin machinery can play in signal transduction.
These multiple functions are evident in the role of ubiquitin signaling in transcriptional regulation, where transcription factors and coactivators can be activated by the ubiquitin machinery. For example, the transcriptional activation domain (TAD) of VP16 is activated in yeast by the SCF substrate receptor F-box protein, MET30 (222). Moreover, fusion of ubiquitin directly to VP16 restores its activity in the absence of MET30 without effecting VP16-TAD stability, providing an example of ubiquitin-dependent, degradation-independent transcriptional activation (222). There are also well-established examples of ubiquitin- and proteasome- dependent transcriptional activation in human cells. First is the Myc transcription factor. Myc degradation is controlled by the SCFSKP2 E3 ubiquitin ligase and SKP2 also promotes Myc transcriptional activation (219, 220). Consistently, a lysine-less version of Myc, which cannot be ubiquitylated and degraded with normal kinetics, still binds to the coactivator Max and localizes to target gene promoters but is unable to fully activate gene expression (221). Similarly, ERa is ubiquitylated and degraded in response to ligand (estrogen) and this is required for its full activation (223).
Here we show a FOXM1 activation mechanism that is both ubiquitin- and degradation- independent yet involves a component of the ubiquitin system that can also regulate its destruction.
We found that FOXM1 is a substrate of the CRL4VPRBP E3 ubiquitin ligase. Analogous to Myc activation by SKP2, we discovered that VPRBP potently activates FOXM1. However, the FOXM1 reporter was activated by two independent versions of VPRBP that are impaired in binding to
CRL4. Moreover, we demonstrate biochemically that VPRBP is dissociated from the CRL4 complex at the time during the cell cycle when FOXM1 activity peaks. Together, these data
38 suggest that FOXM1 activation by VPRBP is ubiquitin- and degradation- independent. The ability of VPRBP to activate FOXM1 independently of CRL4 binding provides a clear demonstration of
E3 ligase substrate receptor repurposing. Moreover, it suggests the possibility that the activity of other CRL substrate receptors can be context dependent and dynamically altered by controlling their association with ubiquitin machinery.
We hypothesize that VPRBP acts as a rheostat to control FOXM1 degradation and activation (Figure 2.9B). FOXM1 abundance is tightly controlled throughout the cell cycle by ubiquitination. By tempering FOXM1 accumulation, VPRBP could impair its activation. Late in the cell cycle, before mitosis, FOXM1 activity increases, contributing significantly to the expression of genes involved in mitotic entry and progression (156). At this time, the partial dissociation of VPRBP from DDB1 and CUL4 (Figure 2.8) would enhance its ability to activate
FOXM1 (216). Consistent with this prediction, VPRBP partially dissociates from CRL4 in mitosis and a mutant version of VPRBP that is impaired for DDB1 binding can still activate FOXM1.
Importantly, since VPRBP depletion increases FOXM1 protein levels in mitotic cells, it is unlikely that VPRBP acts through a switch-like mechanism, whereby it is either targeting FOXM1 for degradation of promoting its activation. In a rheostat model, VPRBP could tune the output of
FOXM1 by controlling both its activation and its degradation with the relative role of each in a given cell determined by the extent of CRL4VPRBP disassembly. It is interesting that VPRBP is unique among CRL4 substrate receptors in that it is likely stoichiometrically bound to CRL4 (230,
245). In addition, VPRBP association with chromatin is tightly cell cycle regulated and occurs only at the time when FOXM1 is activated (G2/M).
The mechanism by which VPRBP dissociates from CRL4 remains unknown. We predict that post-translational modification of VPRBP could alter its chromatin association and/or binding
39 to DDB1/CUL4. Alternatively, modification of DDB1 or CUL4 could regulate their association with VPRBP. Modifications that affect FOXM1-VPRBP binding cannot be ruled out, although they did not change in mitosis in our experiments. It is interesting that CHK2 phosphorylates
FOXM1 in the same region as the one that we mapped as being important for VPRBP binding
(165) and CHK2 has been linked to mitotic progression (246). Dissecting the signaling pathways that control the relationship and interactions between VPRBP, CRL4 and FOXM1 is an important area of future study.
FOXM1 is activated in variety of human malignancies. Specifically, its transcriptional signature is upregulated in HGSOC, serous uterine cancer and basal-like breast cancer. (190, 214,
215). However, it is mechanistically unclear how FOXM1 is activated in each of these disease subtypes. Defects in ubiquitin signaling have been linked to changes in transcription factor stability in malignancies. This is illustrated by the regulation of p53 by MDM2, as well as mutations
FBXW7 that effect Myc. Unlike these two, prominent examples, VPRBP is neither significantly mutated nor transcriptionally altered in cancers in which the FOXM1 signature is upregulated.
However, we show that VPRBP is overexpressed at the protein level in HGSOC patient tumors.
This suggests that VPRBP might, in part, contribute to the FOXM1 expression signature observed in these cancers. It is important to note that cyclin B gene and other G2/M genes can be controlled by multiple factors and FOXM1 itself is controlled by myriad mechanisms. Since VPRBP is not overexpressed at the mRNA level based on transcriptomic analysis, it is presumably due to regulation of either its translation or its degradation. Cross-talk between cell cycle E3 ligases is an emerging theme in cell cycle control (110) and perhaps VPRBP stability is controlled by a second ubiquitin ligase.
40 Materials and Methods
Immunoblot Analysis
Cell extracts were prepared by lysis in ice-cold NETN buffer (20 mM Tris pH 8.0, 100 mM NaCl, 0.5 mM EDTA, 0.5% NP40, 1 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride
(AEBSF), 10 μg/ml leupeptin, 2 μg/ml aprotonin, 2 μg/ml pepstatin A). Cell were lysed on ice for 15 minutes and then centrifuged at 14,000 rpm for 15 minutes at 4°C. Supernatant was collected and protein concentrations were determined by Bradford assay.
Tumor specimens were sampled from patients undergoing surgery for HGSOC at the
University of North Carolina at Chapel Hill. The protocol was reviewed, and exemption granted by the Institutional Review Board at the University. To extract protein from tissues, normal ovarian tissues and ovarian tumors were homogenized in ice-cold tissue homogenizing buffer
(10 mM HEPES (pH 7.4), 50 mM β-glycerophosphate, 1% Triton X-100, 10% glycerol, 2 mM
EDTA, 2 mM EGTA, 1 mM dithiothreitol (DTT), 10 mM NaF, 1 mM Na3VO4, 10 μg/ml leupeptin, 2 μg/ml aprotonin, 2 μg/ml pepstatin A, and 1 mM AEBSF) using TissueLyser II
(Qiagen). The homogenates were placed on ice for 15 minutes and then centrifuged at 14,000 rpm for 15 minutes at 4°C. Supernatant was collected, and protein concentrations were determined by Bradford assay.
For immunoblotting, samples were heated for 5 minutes at 95°C before electrophoresis.
Samples were analyzed by SDS-PAGE using either homemade or Bio-Rad TGX gels which were subsequently transferred onto nitrocellulose membranes (Bio-Rad). Membranes were blocked with
5% bovine serum albumin (BSA). Antibodies were incubated overnight, and signals were detected using Pierce ECL (Thermo Scientific). The primary antibodies used for immunoblotting analysis, their source and the concentrations used are described in detail in Table 2.1. Secondary antibodies
41 conjugated to horseradish peroxidase (HRP) were purchased from Jackson ImmunoResearch
Laboratories.
Immunoprecipitation
Briefly, soluble protein extracts were prepared from 293T cells transiently transfected with
HA-VPRBP and Myc-FOXM1 (wild type or mutants). Cell were lysed in NETN as described above. Precipitation was performed by rotating 50 μl of EZview™ Red Anti-HA Affinity Gel or
EZview™ Red Anti-c-Myc Affinity Gel (Sigma) with 2 mg of soluble, clarified lysate overnight at 4 °C. The affinity resin was recovered by centrifugation at 1,000 × g for 1 minute and washed 3 times with ice-cold lysis buffer. Precipitates were eluted in SDS-PAGE buffer and analyzed by immunoblot.
RT-qPCR
Total RNA was isolated with an RNeasy Plus mini Kit (Qiagen) and then reverse transcribed with a SuperScript® III First-Strand Synthesis System (Invitrogen). The resulting complementary DNA was used for quantitative PCR with iTaq™ Universal SYBR® Green
Supermix (Bio-Rad). Data was normalized to GAPDH. Real-time PCR and data collection were done with a QuantStudio 6 Flex Real-Time PCR System and Software (Applied Biosystems). All of the primers sequences used for RT-qPCR analysis, ChIP and site directed mutagenesis are described in Table 2.2.
Luciferase Reporter Assays
The FOXM1 6xDB luciferase reporter was a kind gift from Michael Whitfield (Dartmouth
University). 293T cells grown on 12-well plates were transiently transfected with 6xDB reporter
42 in combination with different constructs (FOXM1, MELK and VPRBP) using PolyJet Plus transfection reagent (SignaGen Laboratories). The MELK expression vector was a kind gift from
Lee Graves (University of North Carolina). The Nrf2 expression vector and Nrf2 luciferase reporter were gifts from Ben Major (University of North Carolina). VPRBP expression vectors were a kind gift from Yue Xiong (University of North Carolina). Cells were routinely collected at
48 h post-transfection. Luciferase activity was measured using the luciferase reporter assay system
(Promega). Experiments were performed in three technical replicates each. Statistical differences were determined using Student's t test.
Flow Cytometry
For cell cycle profiling, trypsinized cells were washed with phosphate-buffered saline
(PBS), fixed in cold, 70% ethanol and stored overnight at −20°C. DNA was stained for 30 minutes in 25 μg/ml propidium iodide and 100 μg/ml RNase A. Samples were analyzed using CyAn flow cytometer (Beckman Coulter) and FlowJo X software.
In Vivo Ubiquitination Assay
293T cells were co-transfected with 6xHIS-Ubiquitin and HA-VPRBP, with or without
Myc-FOXM1. 48 h after transfection, cells were lysed in denaturing buffer, mixed and sonicated.
Samples were incubated with equilibrated HisPur™ Ni-NTA Resin (ThermoFisher) overnight at
4˚C on a rotator. After repeated washing, the resin was eluted with SDS-PAGE sample buffer.
Detailed experimental procedures describing the wash and lysis buffers are described in (247).
43 Expression and Binding Assay for VPRBP and FOXM1c
VPRBP was expressed from pcDNA3 using the PURExpress® system supplemented with
35S-methionine according to the manufacturer protocol. Full-length FOXM1c (1-763) was subcloned into pET-28b, producing a c-terminally 6xHIS-tagged clone, and expressed in
Escherichia Coli BL21(DE3) cells. Protein expression was induced with isopropyl-b-D- thiogalactopyranoside (IPTG) (250 uM) at 17°C for 18 h. The pellet was sonicated in 40 mL of
Lysis Buffer (20 mM KH2PO4 pH 7.5, 500 mM NaCl, 10 mM Imidazole, 1 mM Tris(2- carboxyethyl)phosphine (TCEP), 10% glycerol, 2 mM phenylmethylsulfonyl fluoride (PMSF), 1 protease tablet, 5 mM MgCl2, 40 units of DNase I). Lysis supernatant was loaded on a 5 mL
HisTrapä HP column, washed with Buffer A (20 mM KH2PO4 pH 7.5, 500 mM NaCl, 10 mM
Imidazole, 1 mM TCEP, 10% glycerol), and eluted with Buffer B (20 mM KH2PO4 pH 7.5, 500 mM NaCl, 500 mM Imidazole, 1 mM TCEP, 10% glycerol). Protein fractions were pooled and desalted with a 200 mL HiPrepä 26/10 column into Buffer C (20 mM KH2PO4 pH 7.5, 200 mM
NaCl, 10 mM Imidazole, 1 mM TCEP).
Nickle-charged nitrilotriacetic acid (Ni-NTA) resin was prepared with wash buffer (20 mM
KH2PO4 pH 7.5, 200 mM NaCl, 10 mM Imidazole, 1 mM TCEP, 0.002% Tween-20) and loaded with FOXM1c-6xHIS. The column was washed with 3 columns volumes of wash buffer and purified VPRBP was added to the resin for 1 h at 4°C. The resin was washed and eluted (using 20 mM KH2PO4 pH 7.5, 200 mM NaCl, 500 mM Imidazole, 1 mM TCEP, 0.002% Tween-20) at
75°C and samples visualized by autoradiography.
44 Figure 2.1 FOXM1 is regulated by CUL4, DDB1 and VPRBP. A. Illustration of CRL4 complex. DDB1 bridges either CUL4A or CUL4B to substrate adaptor subunits termed DCAF or DWD proteins. Substrates are ubiquitylated by E2 enzymes that are recruited to the cullin backbone by RBX1 or RBX2. B. FOXM1 scored in a GPS screen for CRL4 substrates. Three independent microarray probes were shifted in response to dominant-negative CUL4 treatment. C. U2OS cells treated with MLN4924 for 4 h and analyzed by IB. D. U2OS cells were treated with siRNAs targeting FF (control), DDB1, CUL4A or CUL4B and analyzed by IB after 72 h. E. Same as in D. but performed in U2OS cells stably expressing Myc-FOXM1. F. U2OS and HeLa cells treated with siRNAs targeting FF or VPRBP and analyzed by IB after 72 h. G. U2OS cells treated with siRNAs targeting FF or VPRBP were synchronized in S phase using thymidine. Cycloheximide (CHX) was added 4 h after release into the cell cycle and FOXM1 stability was assessed by IB.
45 Figure 2.2 FOXM1 is regulated by VPRBP. A. 293T cells treated with MLN4924 and analyzed by IB. B. U2OS cells treated with siRNAs targeting FF or VPRBP were synchronized in S phase using thymidine. Following release into the cell cycle for indicated times, FOXM1 abundance was analyzed by IB. C. Same as B. Two independent oligonucleotides were used to deplete VPRBP. FOXM1 abundance was analyzed by IB 4 h after thymidine release. D. HCT116 cells were synchronized throughout the cell cycle after VPRBP depletion. FOXM1 abundance was analyzed by IB. E. U2OS cells were treated with cycloheximide (CHX) and analyzed by IB at the indicated times. F. The stability of a mutant Myc-FOXM1 with five amino acid substitutions in the putative degron sequence was analyze by IB. G. The stability of mutant Myc-FOXM1 with indicated amino acid substitutions was analyzed by IB following CHX treatment.
46
Figure 2.3 VPRBP binds and ubiquitylates FOXM1 in vivo. A. HA-VPRBP and Myc-FOXM1 were transiently expressed in 293T cells. VPRBP was recovered on anti-HA-agarose and analyzed by IB. Cells were treated with MG132 prior to lysis. B. Endogenous FOXM1 was precipitated from a 293T cells extract and analyzed by IB. C. Bacterially purified 6HIS-FOXM1 was incubated with in vitro produced, [35S]methionine-labeled VPRBP. Flowthrough and eluates after pulldown were analyzed by autoradiography. D and E. 6HIS-ubiquitin was transiently expressed in 293T cells with and without HA-VPRBP. 6HIS-ubiquitin was recovered after lysis under strong denaturing conditions and analyzed by IB. Analysis of either endogenous (D) or exogenously expressed (E) Myc-FOXM1 was done by IB. (F) Crystal structure from Schwefel et al. depicting the interaction between VPRBP and SAMHD1. Alignment of similar residues between SAMHD1, FoxM1, and two other VPRBP substrates (TET2 and MCM10) is shown below. G. Amino acid substitutions were made in a FOXM1 fragment (aa 321 to 480) corresponding to the residues found in the VPRBP-SAMHD1 interface. Binding was analyzed as described for (A) between Myc-FOXM1321–480 and HA-VPRBP. H. The stability of Myc-FOXM1321–480 fragments that were impaired in their binding to VPRBP was analyzed by CHX chase.
47
Figure 2.4 Examining the FOXM1-VPRBP interaction. A. HA-VPRBP and Myc-FOXM1 were transiently expressed in 293T cells. VPRBP was recovered on HA-agarose following lysis and analyzed by IB. Cells were treated with or without MG132 prior to lysis. B. Fragments of FOXM1 were transiently expressed with a Myc tag in combination with HA-VPRBP in 293T cells. VPRBP was recovered on HA-agarose following lysis and analyzed by IB. Cells were treated with MG132 prior to lysis. C. Fragments of FOXM1 were transiently expressed with a Myc- tag in combination with HA-VPRBP in 293T cells. FOXM1 and the indicated fragments were recovered on Myc- agarose following lysis and analyzed by IB. Cells were treated with MG132 prior to lysis.
48 Figure 2.5 VPRBP activates the FOXM1 6xDB transcriptional reporter. A. FOXM1 and VPRBP were expressed in 293T cells in combination with the 6xDB reporter. Luciferase activity (RLA) was measured 48h after transfection. Error bars represent SD of triplicates. B. Combinations of MELK, FOXM1 and VPRBP were expressed in 293T cells with the 6xDB reporter. Luciferase activity was measured 48h after transfection. Error bars represent SD of triplicates. C. High and low concentrations of VPRBP were expressed in 293T cells with the 6xDB reporter. At 24 and 48h after transfection identical populations of cells were measured for luciferase activity (C1) and harvested and split for analysis by IB (C2) and cell cycle analysis by propidium iodide staining (C3). Luciferase activity was measured. Error bars show SD of triplicates.
49
Figure 2.6 VPRBP regulates FOXM1 independent of binding to the CRL4 complex. A. A NRF2 luciferase reporter was expressed in 293T cells in combination with either FOXM1, VPRBP or NRF2. B. U2OS cells treated with siRNAs targeting FF, FOXM1, VPRBP were synchronized in mitosis by treatment with nocodazole. Mitotic cells were isolated by shake-off and the expression of FOXM1 target genes was analyzed using RT-qPCR. Errors bars show SD of triplicates. C. VPRBPWT and mutant versions that cannot bind DDB1 (RARA and N909) were analyzed for their ability to activate the 6xDB FOXM1 transcriptional reporter.
50
Figure 2.7 VPRBP depletion reduces FOXM1 target gene expression and the number of mitotic cells. A. U2OS cells treated with siRNAs targeting FF or VPRBP were synchronized in G2/M using a thymidine block and release. The FOXM1 target genes were analyzed using RT-qPCR. Error bars represent SD of triplicates. B. U2OS cells treated with siRNAs targeting FF or VPRBP were fixed and immunostained for phosphor-histone H3 serine 10 (P-H3S10). DNA was counterstained with Hoechst. Images were captured using an automated imaging system and the percentage of P-H3S10 positive cells was calculated and averaged from six, wide-field images for each condition. Representative images are shown.
51
Figure 2.8 VPRBP associates with FOXM1 promoters and activates FOXM1 independent of CUL4. A. U2OS cells were synchronized in G2/M phase and fixed for ChIP analysis with control IgG, FOXM1 and VPRBP antibodies. VPRBP associated with the promoters of some FOXM1 target genes. B. VPRBPWT and mutant versions that cannot bind DDB1 (RARA and N909) were expressed in 293T cells in combination with the 6xDB reporter. Luciferase activity was measured 48 h after transfection. Error bars represent SD of triplicates. C. HA-VPRBP was transiently expressed in 293T cells which were subsequently treated with nocodazole or vehicle control (DMSO). VPRBP was recovered or HA-agarose following lysis and analyzed by IB. D. HA-VPRBP and Myc-FOXM1 were transiently expressed in 293T cells which were subsequently treated with nocodazole or vehicle control (DMSO). VPRBP was recovered on HA-agarose following lysis and analyzed by IB.
52
Figure 2.9 Implications and cell cycle control and malignancy. A. Ovaries resected from patients with HGSOC were blotted for VPRBP and compared to normal ovaries. Dotted line indicates splicing together of samples run on the same gel. B. Graphical depiction of model described by data. Our data suggests that VPRBP contributes to the degradation and activation of FOXM1. VPRBP binds to the CRL4 ligase complex and targets FOXM1 for degradation from S-phase. During G2/M phases, when FOXM1 activity peaks, VPRBP disengages from CRL4 and CDH1 acts as a coactivator. FOXM1 is ubiquitinated and degraded by APC/C during G1 phase.
53 Antigen Source Animal Company Cat. # Dilution Use FOXM1 rabbit Santa Cruz sc-502 1:2000 Immunoblot VPRBP rabbit Proteintech 11612-1-AP 1:1000 Immunoblot CUL4A rabbit Abcam ab72548 1:1000 Immunoblot CUL4B rabbit Sigma- HPA011880 1:1000 Immunoblot Aldrich DDB1 rabbit Bethyl A300-462A 1:1000 Immunoblot GAPDH rabbit Santa Cruz sc-25778 1:5000 Immunoblot Ran mouse BD 610340 1:10000 Immunoblot Biosciences Myc mouse Santa Cruz sc-40 1:2000 Immunoblot HA mouse Covance MMS-100P 1:1000 Immunoblot Phospho- Rabbit Cell 3377 1:1600 Immunofluoresce Histone Signaling nce H3 Technologies (Ser10) (D2C8) Mcm10 Rabbit gift from Anja NA 1:1000 Immunoblot Bielinsky (University of Minnesota)
Table 2.1 Antibodies used in experiments described in Chapter 2.
54 Sequence Description AAACCGCGTCTAGCATTAGC CENPF Fwd Primer for RT-qPCR (FOXM1 ChIP) CTCCACGCCTATTGGTCACT CENPF Rev Primer for RT-qPCR (FOXM1 ChIP) AGTCTACGGGCTACCCGATT CDK1 Fwd Primer for RT-qPCR (FOXM1 ChIP) AGCCACTGTACCCGGCTTAT CDK1 Rev Primer for RT-qPCR (FOXM1 ChIP) AACCACGCCCTCCATCTAC FZR1 Fwd Primer for RT-qPCR (FOXM1 ChIP) CCCGCGATTCAAATCAATAC FZR1 Rev Primer for RT-qPCR (FOXM1 ChIP) GGCAAGCAGTTAGGCTTCTG PTMS Fwd Primer for RT-qPCR (FOXM1 ChIP) CGGTCACCTCCACTCCTCT PTMS Rev Primer for RT-qPCR (FOXM1 ChIP) TGGGAAGGCAGAACCAGAATC XRCC1 Rev Primer for RT-qPCR (FOXM1 ChIP) GGGAAGGTGAGAAAACAGGGTG XRCC1 Rev Primer for RT-qPCR (FOXM1 ChIP) CATAAGGGGGCAGAATAAGAGTTG BRCA2 Fwd Primer for RT-qPCR (FOXM1 ChIP) TGATAGAAGGTGGAAATGAGGTCC BRCA2 Rev Primer for RT-qPCR (FOXM1 ChIP) TTCACCTGCTGTGCATTCTC VPRBP forward primer for real-time PCR CAAAGGCTGGCTCCAAGTAG VPRBP reverse primer for real-time PCR TTCCAGTTATGCAGCACCT CCNB1 Fwd Primer for real-time PCR ACTTGTTCTTGACAGTCATGTG CCNB1 Rev Primer for real-time PCR GGCCTCCAAGGAGTAAGACC GAPDH forward primer for real-time PCR AGGGGTCTACATGGCAACTG GAPDH reverse primer for real-time PCR GCAGCAAAGATCAATTGAAG CENPF Fwd Primer for real-time PCR TTTCCACCTGGTCTTTGTG CENPF Rev Primer for real-time PCR CTCACCAGAAACGATGGTG CDC25B Fwd Primer for real-time PCR GGTATCTGCAGTCTACAATCAC CDC25B Rev Primer for real-time PCR GACAAGTACGGCCTTGGGTA PLK1 forward primer for real-time PCR GTGCCGTCACGCTCTATGTA PLK1 reverse primer for real-time PCR CATAGCAAGCGAGTCGCGGCCGCCCCCAAGGTTTTT forward primer for cloning RI418AA into full length FOXM1C AAAAACCTTGGGGGCGGCCGCGACTCGCTTGCTATG reverse primer for cloning RI418AA into full length FOXM1C CGAGTCCGCATTGCCCCCGCTGTGCTGCTAGCTGAGGAG forward primer for cloning K422 into full length FOXM1C CTCCTCAGCTAGCAGCACAGCGGGGGCAATGCGGACTCG reverse primer for cloning K422 into full length FOXM1C GCAGCTAGATCGTTGGGAAG MYC ECR1 forward primer for real-time PCR (ChIP) GCTGGTGATTTCAGTGCAGA MYC ECR1 reverse primer for real-time PCR (ChIP)
Table 2.2 Sequences of primers used in experiments described in Chapter 2.
55
CHAPTER 3: FOXM1 DEUBIQUITINATION BY USP21 REGULATES CELL CYCLE PROGRESSION AND PACLITAXEL SENSITIVITY IN BASAL-LIKE BREAST CANCER2
Introduction
FOXM1 is a member of the forkhead box family of transcription factors, comprised of 44
members in humans, all of which share a conserved forkhead DNA-binding motif (248, 249).
FOX family transcription factors are involved in numerous processes, and play particularly
important roles in stem cell differentiation, proliferation, cell cycle, and metabolism (150).
FOXM1 is best known for its role in regulating the cell cycle, primarily through the transcription
of a set of targets involved in establishing a timely entry and progression through mitosis (156,
170, 211, 217), and ensuring chromosome stability (211). In addition to its important role in
normal cell cycles, the periodically expressed cell cycle genes controlled by FOXM1 are
significantly upregulated in specific cancers and cancer subtypes (190, 214, 250).
Breast cancers are stratified into five subtypes based on molecular features and
transcriptome-based analysis which predict outcomes and guide treatment paradigms. The
molecularly-defined basal-like breast cancer (BLBC) subtype largely overlaps with the triple-
negative classification, named due to a lack of overexpression of oestrogen receptor (ER),
progesterone receptor (PR), and HER2 genes (251, 252). Outside of mutations to TP53,
2 This chapter previously appeared as an article in Cell Reports. The original citation is as follows: A. Arceci, T. Bonacci, X. Wang, K. Stewart, J.S. Dramrauer, K.A. Hoadley, M.J. Emanuele. FOXM1 Deubiquitination by USP21 Regulates Cell Cycle Progression and Paclitaxel Sensitivity in Basal-like Breast Cancer. Cell Rep. 26, 3076–3086.e6 (2019).
56 occurring in 80% of patient tumors, BLBC lacks recurrent somatic mutations to known oncogenes, and instead exhibits a high degree of chromosome instability in the form of copy- number gains and losses as well as severe aneuploidy (190, 250). Upregulation of the FOXM1 transcriptional signature represents a defining feature of the BLBC subtype, as well as other related malignancies, such as high-grade serous ovarian cancer (214). Despite the upregulation of
FOXM1 in BLBC, it is largely unknown how FOXM1 is activated in patient tumors. Due to its roles in the promotion of both cancer and chemotherapy resistance (195, 253), FOXM1 represents an attractive anti-cancer treatment target (254).
FOXM1 is cell cycle regulated, with peak expression and activity occurring during G2/M phase (147). FOXM1 is controlled by a variety of post-translational modifications, including phosphorylation (161, 172, 173) and SUMOylation (175, 176), however, it is unclear how these might account for its upregulation in cancer. Ubiquitination also plays an important role in FOXM1 regulation. FOXM1, and many of its transcriptional targets are ubiquitinated by the anaphase- promoting complex/cyclosome (APC/C), thereby triggering their degradation following the onset of mitosis (153, 237). Ubiquitination of FOXM1 by CUL4VPRBP (255), RNF168 (182), SCFFBXO31
(181) and SCFFBXW7 (256) has also been reported. Together, this points to a pivotal role for ubiquitination dependent degradation in regulating FOXM1.
A potential strategy for inactivating FOXM1 may exist through targeting enzymes which prevent FOXM1 destruction. To that end, we sought to identify deubiquitinating enzymes (DUBs) that regulate FOXM1 stability through the removal of ubiquitin, thus protecting FOXM1 from degradation. DUBs are catalytic proteases which remove ubiquitin from substrates with high in vivo selectivity. The human genome encodes approximately 100 DUBs, which are grouped into a
57 growing number of subfamilies (53). Notably, several selective small-molecule DUB inhibitors have been developed, pointing to their potential druggability (257, 258).
By deploying an RNAi-based screen against nucleus-localized DUBs, Ubiquitin-Specific
Protease 21 (USP21) was identified as a potential regulator of FOXM1 stability. USP21 is a member of the Ubiquitin Specific Protease (USP) family of DUBs which contains 56 members unified by a highly conserved USP domain, featuring a catalytic triad essential for activity (53).
USP21 has been linked to transcriptional regulation through interaction with the transcription factors NANOG (77), GATA3 (75), GLI1 (76) as well as histone H2A (73).
Here, we demonstrate a substrate-enzyme relationship between FOXM1 and USP21.
USP21 regulates FOXM1 abundance and USP21 binds and removes polyubiquitin chains from
FOXM1, thus protecting it from proteasomal degradation. We also show that USP21 expression can alter the FOXM1 transcriptional network which has consequences in regulating mitotic timing and proliferation. Furthermore, we show that FOXM1 and USP21 are specifically upregulated in
BLBC and that depletion of USP21 can improve sensitivity to paclitaxel, primarily through its relationship with FOXM1. These findings demonstrate that USP21, through the maintenance of
FOXM1 stability, regulates cell cycle progression and that inhibiting USP21 has therapeutic potential in treating BLBC with a FOXM1-high, USP21-high expression signature.
Results
USP21 Binds and Alters FOXM1 Abundance
To determine if FOXM1 abundance is regulated by deubiquitinating enzyme (DUB) activity, HeLa cells were treated with PR-619, a small-molecule, non-specific pan-DUB inhibitor for 8 hours. Immunoblot (IB) analysis revealed that FOXM1 abundance significantly decreased
58 with increasing concentrations of PR-619 (Figure 3.1A). This suggested that the degradation of
FOXM1 could be actively prevented by DUBs.
To discover specific USP-family DUBs which directly affect FOXM1 abundance, HeLa cells were transfected with pooled siRNAs against a subset of USP-family DUBs which show nuclear localization. IB analysis of cell lysates 48 hours after transfection revealed that USP21 knockdown reproducibly reduced the level of endogenous FOXM1 (Figure 3.1B). Deconvolution of the siRNA pool revealed that multiple, independent siRNAs reagents targeting USP21 reduced the protein levels of FOXM1 (Figure 3.2A). Furthermore, FOXM1 abundance was not significantly reduced in cells stably expressing a FLAG-HA-tagged USP21 variant made resistant to USP21 siRNA (Figure 3.2B), demonstrating that the reduction in FOXM1 abundance is specifically linked to an on-target effect of USP21 knockdown. Correspondingly, ectopic expression of USP21 significantly increased FOXM1 abundance in 293T cells (Figure 3.1C).
These results demonstrate that FOXM1 abundance is regulated by USP21.
To determine if the effects on FOXM1 stability resulting from changes of USP21 expression were due to an interaction between the two proteins, FLAG- or HA-tagged FOXM1b and Myc-USP21 plasmids were ectopically expressed in 293T cells. An interaction between
FOXM1 and USP21 was detected by co-immunoprecipitation (co-IP), regardless of whether the
IP was directed against Myc-USP21 (Figure 3.1D) or HA-FOXM1b (Figure 3.1E). Additionally, an interaction between endogenous FOXM1 and USP21 was detected using antibodies directed against USP21 (Figure 3.1F). To determine the domain on FOXM1 recognized by USP21, a series of Myc-FOXM1b fragments (Figure 3.2C) were co-transfected with FLAG-HA-USP21 into 293T cells and interactions were assessed by co-IP. Results from IP directed against both Myc-FOXM1b
59 (Figure 3.2D) and FLAG-HA-USP21 (Figure 3.2E) demonstrate that USP21 binds to a region encompassing FOXM1 amino acids 321-400.
Since exogenous and endogenous FOXM1 and USP21 can interact in vivo, we assessed their binding in vitro by incubating recombinant, bacterially-produced hexa-histidine (6xHIS) tagged FOXM1b with recombinant, bacterially-produced GST-tagged USP21. 6xHIS-FOXM1b was detected in complex with GST-USP21 following pulldown on glutathione resin (Figure 3.1G).
Taken together, these data demonstrate that USP21 and FOXM1 directly interact, strongly suggesting an enzyme-substrate relationship between the two proteins.
FOXM1 is a USP21 Substrate
FOXM1 is a known substrate of several E3-ubiquitin ligases, including APC/C (153,
237) and CUL4VPRBP (255), which can conjugate polyubiquitin chains onto FOXM1, triggering its subsequent proteasomal degradation. To determine if the effects that USP21 knockdown or overexpression has on FOXM1 are mediated through proteasomal degradation, HeLa cells transfected with siFF (control) or siUSP21 were treated with the proteasome inhibitor MG132.
IB analysis confirms that FOXM1 destabilization resulting from USP21 knockdown is partially rescued through proteasome inhibition (Figure 3.3A).
DUBs oppose proteasomal degradation through removal of polyubiquitin chains, particularly K48 and K11 topologies and USP21 has been shown to cleave both of these polyubiquitin linkages in vitro (68). To determine if USP21 influences polyubiquitin chain conjugation on FOXM1, we analyzed FOXM1 ubiquitination in vivo. 293T cells were transfected with HA-FOXM1b and 6xHIS-FLAG-ubiquitin in combination with either wild-type
USP21 (USP21WT) or a catalytically-dead variant of USP21 (cysteine at amino acid position 221
60 changed to alanine; USP21C221A). 24 hours post-transfection, cells were harvested and lysed under denaturing conditions, subjected to pulldown on Ni-NTA resin to enrich for ubiquitinated proteins and analyzed by IB. Polyubiquitinated FOXM1 (FOXM1 ≥100kDa on top panel of
Figure 3.3B) was readily detectable under conditions with endogenous levels of USP21 expression (Figure 3.3B, lane 2), but was greatly diminished by overexpression of USP21WT
(Figure 3.3B, lane 3). However, levels of polyubiquitinated FOXM1 were restored close to levels observed in control conditions when USP21C221A was overexpressed (Figure 3.3B, lane 4).
Similarly, USP21WT, but not USP21C221A, immunopurified from 293T cells reduced polyubiquitination of FOXM1 isolated from proteasome inhibitor treated 293T cells, in an in vitro deubiquitination assay (Figure 3.4A).
Since USP21 can directly bind and deubiquitinate FOXM1, we next evaluated the impact of USP21 on FOXM1 stability. To assess the effect of USP21 on FOXM1 half-life, FOXM1 levels were assessed by IB following overexpression of either USP21WT or USP21C221A in 293T cells (Figure 3.3C) and depletion of USP21 in BLBC MDA-MB-231 cells (Figure 3.4B) treated with cycloheximide to block protein translation. Interestingly, FOXM1 abundance decreases with similar kinetics over the first 4 hours following cycloheximide treatment in all conditions, but USP21WT overexpression can protect the remaining population from degradation for at least another 12 hours whereas FOXM1 abundance falls even more rapidly in USP21C221A cells than control after this point (Figure 3.3C). Half-life quantified from the best-fit curve for FOXM1 in control cells is approximately 7.6 hours. The expression of USP21WT increased FOXM1 stability, and its half-life was approximately 12.4 hours. In contrast, ectopic expression of USP21C221A shortened FOXM1 half-life to approximately 3.7 hours, consistent with it having a dominant- negative effect on FOXM1 stability as well as controlling its own ubiquitination (Figure 3.4B).
61 Similarly, depletion of USP21 in MDA-MB-231 reduced FOXM1 half-life from 11.5 hours to
4.2 hours (Figure 3.4C). Taken together, these results indicate that USP21 controls FOXM1 ubiquitination and proteasomal-mediated degradation.
USP21 Affects the FOXM1 Transcriptional Network and Proliferation
FOXM1 is the master transcriptional regulator of a set of genes which are involved in establishing and ensuring a successful G2/M transition and progression through mitosis.
Dysregulation of this transcriptional network has been shown to induce cell cycle delays, mitotic defects, and mis-segregation of chromosomes (211).
We hypothesized that modulating USP21 levels would have corresponding effects on
FOXM1 transcriptional output due to its effects on stabilizing FOXM1. To determine the effect that USP21 overexpression has on FOXM1 activity, 293T cells were co-transfected with USP21 and a FOXM1 luciferase-based transcriptional activity reporter (6X-DBE) (259). USP21 overexpression significantly enhanced FOXM1-dependent transcriptional activity compared to control (Figure 3.5A). This is consistent with our previous data demonstrating that FOXM1 abundance is positively enhanced by USP21 overexpression. Correspondingly, a panel of
FOXM1 transcriptional targets was assessed by RT-qPCR following siRNA transfection to knockdown either FOXM1 or USP21 in MDA-MB-231 cells. Knockdown of USP21 leads to a reduction in all FOXM1 transcriptional targets assessed compared to controls (Figure 3.5B, compare red bars to black bars). USP21 knockdown did not lead to the same decrease in FOXM1 transcriptional target abundance as FOXM1 knockdown (Figure 3.5B, compare red bars to orange bars), however this is consistent with biochemical data demonstrating that USP21 knockdown reduces FOXM1 levels to an amount between what is observed in control and
62 FOXM1 knockdown conditions (Figure 3.5C). Additionally, USP21 overexpression did not significantly alter FOXM1 mRNA levels in MDA-MB-231 cells transduced with a pINDUCER20 lentivirus to introduce a doxycycline-inducible USP21 transgene (260) (Figure
3.6A). Taken together, these data demonstrate that changes in FOXM1 levels due to USP21 have downstream effects on the FOXM1 transcriptional network.
Tight regulation of the FOXM1 transcriptional network is required for a timely and high- fidelity transition from G2 to mitosis. Having demonstrated that USP21 levels can affect the
FOXM1 transcriptional network, we next sought to determine if the kinetics of mitotic entry are disrupted by USP21 loss. To observe effects on mitotic timing, the accumulation of phospho- histone H3 S10 (P-H3) following release of cells from synchronization at the G1/S boundary with aphidicolin was monitored in MDA-MB-231 cells depleted of FOXM1 or USP21. In parallel, we also analyzed cells where FOXM1 was ectopically expressed in USP21-depleted cells (Figure
3.6B). Following release from synchronization, cells were treated with the microtubule poison nocodazole so that they would get trapped in mitosis, allowing us to determine the percent of cells that had entered mitosis throughout the duration of the experiment. Depletion of either
FOXM1 or USP21 significantly impaired cell cycle progression into mitosis. 22 hours after release only 1.6% and 10.9% of cells were P-H3 positive, in FOXM1 and USP21-depleted populations, respectively (Figure 3.5D). In contrast, 40.7% of control cells had entered mitosis by that time. Interestingly, mitotic entry was significantly rescued (27.4% P-H3 positive) in the
USP21-depleted/FOXM1 overexpression populations. These results confirm prior reports that
FOXM1 knockdown impairs mitotic entry (211). Furthermore, they demonstrate that USP21 promotes progression through mitosis since its depletion severely impairs the accumulation of mitotic cells. Since this defect is largely rescued by forced FOXM1 expression, this indicates that
63 the cell cycle phenotype in USP21-depleted cells occurs largely due to the reduction of FOXM1 abundance.
Finally, we hypothesized that slowed mitotic entry resulting from FOXM1 and USP21 knockdown would cause a corresponding reduction in proliferation. Indeed, proliferation was significantly impaired in the BLBC cell lines MDA-MB-468 and MDA-MB-231 (Figure 3.5E).
In addition, ectopic, doxycycline-inducible expression of FOXM1b in MDA-MB-231 cells partially rescued the impaired proliferation caused by USP21 depletion. Similar results were observed in HeLa cells over a four-day period (Figure 3.6C). Thus, downregulation of USP21 can affect overall proliferation by downregulating both FOXM1 stability and the downstream effects of the FOXM1 transcriptional network.
USP21 Amplification Correlates with FOXM1 Protein Levels and Tumor Proliferation in BLBC
FOXM1 upregulation has been reported in a variety of cancers (159, 190, 214, 261, 262).
Notably, in breast cancer, we found that FOXM1 mRNA expression is significantly correlated with breast cancer subtypes, showing the highest expression in the BLBC subtype (t-test basal versus non-basal, p ≤ 2.2 x 10-16) (Figure 3.7A, top panel). This is consistent with prior genomic studies which found the FOXM1 transcriptional signature upregulated in BLBC (190). We next analyzed the proliferative capacity of tumors where FOXM1 is activated. Remarkably, the proliferation score of individual tumors was highly and significantly correlated with FOXM1 expression, both across all breast cancer subtypes, as well as within individual subtypes. That is, a tumor proliferation score increases as a function of FOXM1 mRNA expression among all patient breast tumors even within only the luminal A breast cancers (Figure 3.7A, top panel, blue dots) which generally proliferate more slowly than more aggressive HER2-enriched and BLBC
64 subtypes. As expected, expression analysis of well-established FOXM1 transcriptional targets revealed a corresponding upregulation in FOXM1-high, BLBC patient samples (Figure 3.7A, bottom panel). We conclude that FOXM1 expression and activity exists on a sliding scale, with increasing FOXM1 levels alone being a strong indicator of tumor proliferation.
Next, we analyzed copy number changes and mRNA expression for all human DUBs among breast tumors in the TCGA dataset. USP21 is the fourth-most frequently CNA-amplified and the most frequently mRNA-overexpressed DUB across all patient samples (Figure 3.7B) with USP21 amplification being significantly correlated with both proliferation score and the
BLBC subtype (Figure 3.9A). The genomic location of USP21 is 1q23. Strikingly, this region is the most frequently amplified region in BLBC patient tumors (85, 263). However, this amplicon lacks a well-described oncogene which contributes to disease. Correspondingly, IB analysis of
FOXM1 and USP21 protein expression in cell lines representing the normal-like, luminal, and
BLBC subtypes of breast cancer revealed that FOXM1 and USP21 expression are uniquely high in the BLBC subtypes (Figure 3.7C, Figure 3.9B).
FOXM1 is among the 227 proteins/antigens profiled by reverse phase protein array
(RPPA) in TCGA. This provided us the opportunity to examine the abundance of FOXM1 protein levels in breast cancer, and its relationship to USP21 mRNA expression in tumors. Since our results demonstrate that the FOXM1 protein is post-translationally stabilized by USP21, we examined FOXM1 protein levels as a function of USP21 expression. Remarkably, FOXM1 protein is the most significantly upregulated protein/antigen in USP21-amplified tumors (Figure
3.7D). Among the other most statistically significant increasing proteins were the FOXM1 targets cyclin B1, the cell cycle regulated gene MSH6 (264), adenosine deaminase ADAR1
(which is proximal to the genetic location of USP21), and the asparagine deaminase, ASNS,
65 which has been previously implicated in breast cancer proliferation and cell cycle progression
(265). Since FOXM1 is upregulated by USP21, we asked if a curated list of 114 FOXM1 target genes are correlated with USP21 amplification using gene set enrichment analysis (GSEA). This analysis revealed a strong enrichment for the FOXM1 transcriptional network in USP21 amplified breast cancers (Table 3.1).
USP21 Affects Paclitaxel Response in Basal-like Breast Cancer
In addition to its role in cancer proliferation, high FOXM1 expression has been attributed to resistance to platinum-based drugs (253) and taxanes (195, 266) in breast cancer and correlated with metastasis and poor patient outcomes. Having identified FOXM1 as a USP21 substrate, we examined if USP21 knockdown could increase sensitivity to paclitaxel in a
FOXM1-dependent manner. The BLBC cell lines MDA-MB-231 and MDA-MB-468 and
SUM149 cells were used to assess viability following 72 hours of exposure to DMSO or paclitaxel in control (siFF), FOXM1-depleted (siFOXM1) and USP21-depleted (siUSP21) cells.
In addition, in MDA-MB-231 cells we utilized a cells line containing a stably integrated, doxycycline inducible version of FOXM1, allowing us to examine viability in USP21-depleted /
FOXM1-induced (siUSP21 + FOXM1b) conditions. Across all cell lines assessed, there is a significant reduction in viability between FOXM1 and USP21-depleted conditions treated with paclitaxel compared to the same depletion condition where cells were not treated with paclitaxel
(DMSO control; Figure 3.8A). Therefore, while depletion of USP21 (and FOXM1) impairs proliferation in each of these cell lines under control conditions, there is a statistically significant decrease in viability when USP21 depletion is combined with paclitaxel treatment. Specifically, in MDA-MB-231 cells exposed to paclitaxel, cell viability was reduced by 40.9% and 38.3% compared to control in FOXM1 and USP21-depleted conditions, respectively (Figure 3.8A, top
66 panel). However, cell viability is only reduced by 16.7% in cells where USP21 was depleted but
FOXM1 levels are restored with doxycycline induction. In MDA-MB-468 cells treated with paclitaxel, viability was more strongly reduced by 81.3% and 41.5% compared to control in
FOXM1 and USP21-depleted conditions, respectively (Figure 3.8A, middle panel). Finally, in
SUM149 cells treated with paclitaxel, viability was significantly reduced by 43.6% and 20.2% in
FOXM1 and USP21-depleted conditions respectively (Figure 3.8A, bottom panel). This is particularly striking considering that SUM149 cells do not show a significant reduction in viability when treated with paclitaxel alone. The significant difference in cell viability resulting from FOXM1 depletion and USP21 depletion combined with a significant rescue of viability when FOXM1 is overexpressed in the USP21-depleted conditions suggests that FOXM1 significantly contributes to paclitaxel sensitivity and is a critical USP21 substrate, although perhaps not the only one, contributing to the differences in paclitaxel sensitivity observed in
USP21-depleted cells.
Based on the realization that USP21 depletion sensitizes cells to paclitaxel, we next determined how BLBC cells would respond to paclitaxel following USP21 depletion in a mouse xenograft model. SUM149 is a BLBC cell line that is paclitaxel insensitive in mouse xenografts
(Figure 3.8B, red line). SUM149 were transduced with lentivirus expressing scrambled shRNA
(shControl) or shUSP21 and implanted in the mammary fat pads of NOD/SCID gamma mice following selection for infected cells (Figure 3.9C). Mice were then administered either paclitaxel (10mg/kg of bodyweight) or vehicle twice weekly and tumor volume was monitored twice weekly for six weeks. The results revealed that average tumor volume was reduced from
1101.75mm3 in the control condition to 414mm3 (a 62.4% reduction) over six weeks due to depletion of USP21 and exposure to paclitaxel (Figure 3.8B, blue line). These results suggest
67 that inhibition of USP21 may provide therapeutic benefit in BLBC by antagonizing proliferation and enhancing chemotherapy resistance effects driven by FOXM1 activation.
Discussion
Here, we identify the transcription factor FOXM1 as a substrate of the deubiquitinating enzyme USP21. USP21 stabilizes FOXM1, and suppressing USP21 reduces FOXM1 abundance, which subsequently downregulates the FOXM1 transcriptional network. This leads to a mitotic entry delay, slowed proliferation and sensitivity to paclitaxel, both in culture and in animal xenografts (Figure 3.10). Interestingly, phenotypes linked to USP21 depletion were partially, but not completely, mitigated by forced expression of FOXM1. This suggests that FOXM1 is a key
USP21 substrate with respect to these phenotypes, but likely is not the only target determining the role of USP21 in cell cycle, cancer proliferation and paclitaxel sensitivity. Nevertheless, we predict that targeting USP21 could have therapeutic potential in aggressive cancers marked by high FOXM1 expression and with concordant genome instability.
To date, few USP21 substrates have been identified. Many validated substrates, including
GATA3 (75), GLI1 (76), NANOG (77), and ubiquitinated histone H2A K119 (73) are either transcription factors or involved in the regulation of transcription. Our identification of FOXM1, a transcription factor that regulates cell cycle progression, as a USP21 substrate supports this theme. Given these roles, it is interesting to consider USP21 as a general transcriptional regulator through its dual function in stabilizing transcription factors and creating a permissive landscape for transcription to occur through the removal of ubiquitin at H2A K119, a post-translational modification associated with transcriptional repression. Future work integrating these aspects of
68 transcription under the control of USP21 may underscore its importance in the promotion of transcriptional activity.
BLBC is characterized by having high genomic instability which contributes to a distinctive pattern of CNA gains at 1q, 6p, 8q, and 10p along with frequent losses at 4p, 5q, 14q, and 15q (267). Outside of TP53 loss, which occurs in more than 80% of cases, there are few mutations in well-characterized oncogenes or additional tumor suppressors linked to BLBC. The
USP21 genomic location is 1q23, an amplified region recently suggested to contain putative driver genes for the BLBC subtype (85). The FOXM1 genomic location is 12p13, another amplified region associated with the BLBC subtype. Indeed, our analysis supports a link between
FOXM1 expression, the BLBC subtype, and proliferation as well as a correlation between
USP21 expression at the genomic and transcriptional level (Figure 3.7B) and a correlation between USP21 and FOXM1 levels in BLBC. Notably, we show that USP21 amplification correlates with proliferation score of the tumor. While this result is partially complicated by the fact that amplification of 1q, and particularly the USP21 genomic locus, is enriched in aggressive
BLBC subtyped tumors (85), this further points to its potential role in disease. The evidence presented here suggests that USP21 could act as a proliferative driver by promoting FOXM1 stability and transcriptional function which ultimately promotes proliferation in BLBC with amplified USP21 and FOXM1 expression. We speculate that increased gene dosage of USP21 contributes to the proliferative features of BLBC, and that it likely acts coordinately with additional proliferative drivers on 1q, as well as with other genes located within recurrently altered genomic regions.
Due to its myriad roles in the promotion and maintenance of cancer, FOXM1 has garnered significant attention as a target for therapeutic intervention (158). Since FOXM1 is
69 recurrently activated in specific disease subtypes, including the majority of BLBC and high- grade serous ovarian cancers, there exists the rationale, therapeutic window, and patient numbers to support its therapeutic candidacy. Moreover, while current therapies that target specific transcription factors are clinically effective, including hormone therapy in breast cancer and all- trans retinoic acid (ATRA) in acute promyelocytic leukemia (268), in most cases, transcription factors have defied efforts aimed at chemical inhibition (e.g. c-Myc).
Here, we provide the rationale for an indirect method of downregulating FOXM1 activity through the inhibition of USP21. In contrast to transcription factors, DUBs have favorable properties which make them attractive therapeutic targets. More specifically, USP-family DUBs have deep, active pockets relying on a catalytic triad (53) which is generally tractable with small- molecule inhibitors (269). Recently, the identification of potent and highly selective USP7 inhibitors has brought to the forefront the chemical tractability of DUBs as a therapeutic class
(257, 258). The crystal structure of the USP21 catalytic core has been solved (68) which facilitates the identification and development of potent and highly selective small-molecule inhibitors through structure-activity relationship studies. Future work describing additional substrates and processes influenced by USP21 may uncover additional examples where USP21 inhibition may demonstrate therapeutic benefit.
Materials and Methods
Cell Culture
Cell lines were all obtained from the American Type Culture Collection, or BioIVT. 293T,
HeLa, MDA-MB-231, MDA-MB-468, HMEC, MCF7, MCF10A, SUM159, and BT474 cells were
70 cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco) supplemented with 10% FBS
(Atlanta Biologicals) and 1% Pen/Strep (Gibco). SUM149 cells were obtained from BioIVT and cultured in either F-12 medium (Gibco) supplemented with 5% FBS (Atlanta Biologicals), 10mM
HEPES pH7.5, 1µg/mL hydrocortisone, 5µg/mL insulin (Gibco) and 1% Pen/Strep (Gibco) or in
HuMEC Ready Medium (Gibco). All cells were incubated at 37°C and 5% CO2.
Transfections and Treatments
All siRNA transfections were performed using Lipofectamine RNAiMAX (ThermoFisher) following instructions in the manufacturer’s instructions. Sequence of siRNAs used in this study are described in Table 3.2. All plasmid transfections into 293T cells were performed using Mirus
TransIT-293 reagent (Mirus Bio) following instructions in the manufacturer’s protocol. Plasmid transfections into other cell lines were performed using Lipofectamine 2000 (ThermoFisher) according to manufacturer’s instructions. PR-619 and MG132 (Selleck Chemicals) were used at a concentration of 10µM for 8 hours and 6 hours, respectively. Cycloheximide (MilliporeSigma) was used at a concentration of 100ng/mL for the indicated time.
Molecular Cloning
pDEST-HA-FOXM1b, pDEST-Myc-USP21, pDEST-FLAG-FOXM1b, pINDUCER20-
FOXM1b and pINDUCER20 USP21 were all generated using Gateway Cloning Technology
(ThermoFisher) from pDONR223-FOXM1b and pDONR223-USP21 plasmids. All FOXM1 fragments described in Figure 3.2 were ordered as synthetic gene fragments (IDT) and subsequently cloned into pDONR223 and pDEST-Myc plasmids for expression using Gateway
Cloning Technology. The FOXM1 6X-DBE luciferase reporter plasmid was a gift from Michael
Whitfield (Dartmouth University). FLAG-HA-USP21 plasmid was a gift from Wade Harper
71 (Addgene plasmid # 22574) (58). pINDUCER20 was a gift from Stephen Elledge (Addgene plasmid # 44012) (260). 6-HIS-FLAG Ubiquitin plasmid was a gift from Philippe Soubeyran.
Site-directed mutagenesis with QuickChange Lightning Kit (Agilent) was used to generate
C221A MSCV-FLAG-HA-USP21 following manufacturer’s instructions. Primers used are described in Table 3.3.
Lentivirus Production
293T cells were transfected with pINDUCER20 plasmids mixed with VSV-G, Gag-Pol,
Tat and Rev lentivirus packaging helper plasmids in a 1:4 ratio. Media containing viral particles was harvested 24 and 48 hours later, passed through a 0.45µM filter and stored frozen at -80°C.
MDA-MB-231 cells were transduced with lentiviral particles and selected with G418 to isolate successfully transduced cells.
Immunoblotting
Samples were lysed in RIPA buffer (150mM NaCl, 50mM Tris pH 7.5, 5mM EDTA, 1%
NP-40, 0.5% sodium deoxycholate, 0.1% SDS) supplemented with 1µg/mL aprotinin and leupeptin, 10µg/mL leupeptin, 1mM sodium orthovanadate, 1mM NaF, and 1mM AEBSF. Protein concentration was quantified by Pierce BCA assay (ThermoFisher) and samples were prepared by boiling in Laemmli buffer for 5 minutes. Samples were separated by gel electrophoresis using homemade or TGX (Bio-Rad) SDS-PAGE gels and transferred to nitrocellulose membrane. All blocking and antibody washing steps were performed in 5% nonfat dried milk (Bio-Rad) diluted in TBS-T (137mM NaCl, 2.7mM KCl, 25mM Tris pH 7.4, 1% Tween-20). All primary antibody incubations were performed shaking at 4°C for 16 hours. All secondary antibody incubations were performed shaking at room temperature for 1 hour. All washing steps were performed using TBS-
72 T. Protein abundance was visualized by chemiluminescence using Pierce ECL (ThermoFisher), or
Clarity ECL (Bio-Rad). A list of all antibodies used in this study are described is Table 3.4.
Immunoprecipitations
293T cells were transfected with the indicated plasmids and lysed in NETN buffer (100mM
NaCl, 20mM Tris pH 8, 0.5mM EDTA, 0.5% NP-40 supplemented with 1µg/mL aprotinin and leupeptin, 10µg/mL leupeptin, 1mM sodium orthovanadate, 1mM NaF, and 1mM AEBSF) 48 hours after transfection. 1µg of the indicated antibody was conjugated to a 50µl mix of magnetic
Protein A and Protein G beads (Bio-Rad) at room temperature for 1 hour. 1mg of lysate from each sample was pre-cleared to remove non-specific interactions with a similar volume of Protein A/G beads rotating at 4°C for 1 hour. Following antibody conjugation and pre-clearing, lysate and beads were combined rotating at 4°C for 4 hours. Beads were subjected to multiple rounds of washing with NETN. Proteins were eluted by boiling beads in Laemmli buffer for 10 minutes and visualized by IB.
FOXM1 and USP21 In Vitro Binding Assay
FOXM1b was subcloned into a pET-28b backbone to create a C-terminal 6X-Histidine tagged clone, transformed into Escherichia coli BL21(DE3) cells, grown to an OD600 of 0.5, and induced to express with 250µM of IPTG shaking at 18°C for 16 hours. USP21 was cloned into a pDEST15 backbone using Invitrogen Gateway cloning technology (ThermoFisher) to create a N- terminal GST-tagged clone, transformed into Escherichia coli BL21(DE3) cells, grown to an
OD600 of 0.5, and induced to express with 100µM of IPTG shaking at 25°C for 5 hours. Bacterial pellets were resuspended in 40mL lysis buffer (20mM KH2PO4 pH 7.5, 500mM NaCl, 10% glycerol, 2µg/mL aprotinin, 10µg/mL leupeptin, 2µg/mL pepstatin, 1mM sodium orthovanadate,
73 1mM NaF 1mM AEBSF, 10mg/mL lysozyme, 40U DNase I; leupeptin and AEBSF not included in GST-USP21 lysis). 6HIS-FOXM1b and GST-USP21 were batch purified from crude lysate on
His-Pur Ni-NTA resin (ThermoFisher) and glutathione agarose resin (GoldBio) following manufacturer’s protocols, respectively. Following elution, proteins were buffer exchanged (25mM
Tris pH 7.4, 200mM NaCl, 2mM DTT) using Zeba desalting columns (ThermoFisher) following manufacturer’s protocols. For interaction studies, 1µg of GST or GST-USP21 was conjugated to
GSH agarose resin rotating at room temperature for 1 hour. Then, 1µg 6HIS-FOXM1b was added and incubation continued with rotation at room temperature for 1 hour. Following three rounds of washing with PBS, protein complexes were eluted (50mM Tris pH 8.0, 10mM glutathione), resuspended in Laemmli buffer and assessed by IB.
FOXM1 Transcriptional Reporter Assay
293T cells were transfected with FOXM1 6X-DBE luciferase reporter and FLAG-HA-
USP21 plasmids as indicated and collected 24 hours after transfection. Lysates were prepared according to the protocol described in the Luciferase Assay System (Promega). Measurements were made in a 96-well plate and results represent the average of triplicates for each experiment condition described.
Cell Cycle Progression into Mitosis
MDA-MB-231 cells were transfected with the indicated siRNA using RNAiMAX. 48 hours following transfection, media on cells was refreshed with media containing 2µg/mL aphidicolin. 24 hours later, media on cells was refreshed with media containing 200ng/mL nocodazole. Cells were harvested at the indicated time points and samples were prepared for IB or for flow cytometry by fixation in 70% ethanol. For flow cytometry, cells were blocked for 30
74 minutes, incubated with rabbit anti-phospho-Histone H3 (S10) antibody for 1 hour, then incubated with goat anti-rabbit Alexa Fluor 488 for 1 hour, and finally stained with 25µg/mL propidium iodide, 100µg/mL RNase A for 1 hour. All blocking, washing, and staining steps were done with
1% BSA in PBS. All samples were run on a CyAn ADP analyzer (Beckman Coulter) and data was analyzed using FlowJo X software.
In Vitro Paclitaxel Sensitivity Assays
MDA-MB-231 pINDUCER20-FOXM1b, MDA-MB-468 and SUM149 cells were transfected with the indicated siRNAs. 24 hours following transfection, cells were seeded in triplicates at a density of 7,500 cells / well in a 96-well plate. 12 hours after seeding, media supplemented with paclitaxel, and in one set of siUSP21 triplicates (MDA-MB-231 pINDUCER20
FOXM1b only), paclitaxel with 10ng/mL doxycycline was added to cells. After a 72 hour incubation, cell viability was assessed using PrestoBlue reagent according to manufacturer’s instructions.
Tumor Xenograft
SUM149 cells were transduced with lentivirus expressing shRNA against USP21 or a non- targeting control and selected of positive expression with puromycin. Immediately following selection, orthotopic xenografts were established for all conditions by injecting subcutaneously 5 x 106 cells mixed 50:50 in serum-free media:matrigel into the right mammary fat pad of athymic, female nude mice. Tumor volume was measured with calipers and calculated using the equation
Volume = Length x Width2 x 0.563. Upon tumors reaching a volume 100mm3, mice were administered a bi-weekly intraperitoneal injection of 10mg/kg (body weight) paclitaxel or
75 equivalent volume of vehicle. Tumor volume measurements were recorded twice weekly for 6 weeks.
Proliferation Assays
MDA-MB-231 pINDUCER20 FOXM1b and MDA-MB-468 cells were transfected with the indicated siRNA, then 12 hours later, 30,000 cells/well were seeded into 24-well plates in triplicates for each condition. Growth was monitored through phase-contrast imaging using the
IncuCYTE Zoom (Essen Bioscience) live-cell imaging device. Images were taken every 6 hours for 120 total hours and growth was calculated for each timepoint as the average percent confluence of 4 images taken per well. Values were normalized for each respective condition so initial confluence starts at 0%.
HeLa cells were transfected with the indicated siRNA, then 24 hours later, transfected with the indicated plasmids. Growth was monitored using PrestoBlue cell viability reagent
(ThermoFisher) following the manufacturers protocol. Measurements were taken in triplicate for each timepoint with each point described as a percentage of the first measurement taken following seeding.
In Vivo Deubiquitination Assay
The in vivo ubiquitination assay was performed as described previously (247). Briefly,
293T cells were transfected with the indicated plasmids and harvested in PBS 24 hours following transfection. 80% of the cell suspension was lysed in 6M guanidine-HCL buffer and 6HIS- ubiquitinated proteins were captured on HisPur Ni-NTA resin (ThermoFisher) while the remaining 20% of sample was used to prepare inputs. Pull-down eluates and inputs were separated on SDS-PAGE gels and analyzed by IB.
76 In Vitro Deubiquitination Assay
293T cells were transfected with HA-FOXM1b and 6xHIS-FLAG-ubiquitin and separately with FLAG-HA-USP21WT, FLAG-HA-USP21C221A and empty vector, respectively. 48 hours following transfection, cells transfected with HA-FOXM1b/6xHIS-FLAG-ubiquitin were treated with complete DMEM supplemented with 25µM MG132 for 1 hour prior to harvesting. HA-
FOXM1b/6xHIS-FLAG-ubiquitin cells were lysed in RIPA buffer (150mM NaCl, 50mM Tris pH
7.5, 5mM EDTA, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) supplemented with 1µg/mL aprotinin and leupeptin, 10µg/mL leupeptin, 1mM sodium orthovanadate, 1mM NaF, 1mM
AEBSF and 20mM NEM. FLAG-HA-USP21WT, FLAG-HA-USP21C221A and empty vector cells were lysed in RIPA buffer without additional inhibitors. HA-tagged proteins were captured from clarified cell lysate on 50µL EZview Red Anti-HA Affinity Gel (MilliporeSigma) according to manufacturer’s instructions. Following capture and washing, beads containing captured HA-
FOXM1b/6xHIS-FLAG-ubiquitin were resuspended in DUB buffer (50mM Tris pH 7.5, 50mM
NaCl, 5mM DTT, 5mM MgCl2) and mixed equally with beads resuspended in DUB buffer containing captured FLAG-HA-USP21WT, FLAG-HA-USP21C221A and empty vector cell lysate.
Beads were incubated with rotation at 37°C for 4 hours and proteins were eluted directly from beads with the addition of Laemlli sample buffer and boiling. Ubiquitination status of HA-
FOXM1b was assessed by IB as described above.
RT-qPCR
MDA-MB-231 cells were transfected with the indicated siRNA and harvested 72 hours after transfection. RNA was extracted using RNeasy Mini Kit (Qiagen) following manufacturer’s instructions. 1µg of RNA was used to generate cDNA libraries with random primers using
77 SuperScript III Reverse Transcriptase (ThermoFisher) following manufacturer’s instructions.
Samples were diluted 1:10 and transcript abundance was quantified using SsoAdvanced SYBR
Green (Bio-Rad) on a QuantStudio 7 Flex Real-Time PCR system (ThermoFisher) normalized to
GAPDH, with each condition run in triplicates. Data displayed as the relative quantity of transcript quantified using the 2−ΔΔCt method (270). All primers used are described in Table
3.3.
TCGA Dataset Analysis
Upper quartile normalized RSEM gene-level data for 1095 invasive breast cancers were downloaded from the legacy TCGA-BRCA project using the NIH Genome Data Commons
(https://portal.gdc.cancer.gov). Proliferation score was calculated for each sample analyzed as previously described (271) and compared to log2-transformed, median-centered FOXM1 expression data for each sample analyzed. FOXM1 transcriptional targets were log2- transformed, median-centered and compared to log2-transformed, median-centered FOXM1 expression. Gene Set Enrichment Analysis (GSEA) (272, 273) was performed for a previously derived FOXM1 target gene list (217) comparing USP21-amplified and USP21-non-amplified patient tumors. RPPA normalized data was downloaded from the TCGA PanCanAtlas
Publication page (http://api.gdc.cancer.gov/data/fcbb373e-28d4-4818-92f3-601ede3da5e1) and data for 879 breast samples were extracted. GISTIC values were downloaded from Broad’s
Firehouse (http://gdac.broadinstitute.org/runs/analyses__2016_01_28/data/BRCA-
TP/20160128/gdac.broadinstitute.org_BRCA-
TP.CopyNumber_Gistic2.Level_4.2016012800.0.0.tar.gz). There were 857 samples with both
RPPA and copy number data. GISTIC values of 2 were considered highly amplified for USP21 and compared to those without high amplification of USP21 for volcano plots and to compare
78 proliferation score. An FDR-corrected Wilcoxon rank sum test was used to calculate proteins associated with USP21 copy number status. Data from 505 breast invasive carcinoma samples from TCGA (190) were downloaded from cBioPortal (http://www.cbioportal.org/) (274, 275).
Percent of samples with CNA amplification and mRNA overexpression (Z-score threshold +/- 2) of 92 DUBs was aggregated.
Other Statistical Analysis
Analysis of FOXM1 half-life stability was assessed by densitometry of bands on immunoblots using ImageJ. Band intensity was calculated for FOXM1 and the respective loading control for each lane/timepoint. Non-specific background was subtracted from the signal of each band. FOXM1 intensity at a given timepoint was normalized to the amount of loading control in that lane. For calculating half-life, all timepoints were calculated as a percentage of initial FOXM1 abundance for each respective condition. The best-fit nonlinear regression curve was drawn through points and half-life was assed as time when y = 50% using GraphPad Prism
6. Key details of the statistical analysis pertaining to RT-qPCR, proliferation assays, paclitaxel sensitivity assays and tumor xenograft assays are described in the corresponding method details for those experiments and in the corresponding figure legends.
79
Figure 3.1 USP21 binds and regulates FOXM1 abundance. A. HeLa cells treated with vehicle, 2.5, 5 or 10µM PR-619 for 8h were analyzed by immunoblot (IB). B. HeLa cells were transfected with a pool of four independent small-interfering RNAs (siRNAs) targeting respective DUBs. FOXM1 stability was assessed by IB 72h after transfection. C. FOXM1 levels were assessed by IB following transfection of Myc-FOXM1b and FLAG-HA-USP21 in 293T cells 48h after transfection. D. FLAG-FOXM1b and Myc-USP21 were co-expressed in 293T cells. Protein complexes were immunopurified (IP) with anti-Myc and analyzed by IB. E. HA-FOXM1b and Myc-USP21 were co-expressed in 293T cells. Lysates were IP with anti-mouse immunoglobulin G (IgG) or anti-HA and analyzed by IB. F. Endogenous USP21 was IP from HeLa whole-cell lysates and analyzed by IB. G. Recombinant 6xHIS- FOXM1b was incubated with recombinant GST-USP21. Complexes were captured on glutathione (GSH) agarose beads and analyzed by IB.
80
Figure 3.2 Analysis of USP21-FOXM1 binding. A. IB confirming USP21 knockdown and FOXM1 expression in HeLa cells transfected with multiple individual USP21 siRNAs for 48h. B. HeLa cells stably expressing FLAG-HA- USP21 WT or FLAG-HA-USP21 mutated to resist siUSP21 were transfected with indicated siRNAs. FOXM1 levels were assessed by IB 48h after transfection. C. Diagram of FOXM1b fragments assessed for binding with USP21. Putative USP21 binding domain indicated by red line. (Fkh = forkhead domain). D. 293T cells were transfected with FLAG-HA-USP21 and the indicated Myc-FOXM1b fragments. Cells were lysed and subjected to Myc IP. USP21 binding to FOXM1b fragments was assessed by IB. E. 293T cells were transfected with FLAG-HA- USP21 and the indicated Myc-FOXM1b fragments. Cells were lysed and subjected to FLAG IP. FOXM1b fragments bound to USP21 were assessed by IB.
81
Figure 3.3 USP21 protects FOXM1 from proteasomal degradation. A. HeLa cells were transfected with the indicated siRNAs. After 48h, HeLa cells were treated with 10µM MG132 or vehicle for 6h. B. 293T cells transfected with the indicated plasmids were lysed under denaturing conditions and subjected to Ni-NTA pulldown. The ubiquitination status of FOXM1 in each condition was assessed by IB. C. Top: 293T cells were transfected with the indicated plasmids. After 48h, 100ng/mL of cycloheximide was added to the cells and samples were taken every 4h for 16h. Bottom: densitometry analysis performed on corresponding IBs to assess FOXM1 half-life in the indicated conditions.
82
Figure 3.4 USP21 controls FOXM1 ubiquitination and half-life. A. Ubiquitinated HA-FOXM1b was IP from 293T cells that were treated with MG132 prior to lysis. The precipitated HA-FOXM1b was combined with FLAG- HA-USP21WT or FLAG-HA-USP21C221A IP from separately transfected 293T cells (not treated with MG132). Proteins were resuspended in DUB buffer and incubated at 37°C for 4h. Ubiquitination status of FOXM1 in each condition was assessed by IB. B. 293T cells transfected with the indicated plasmids were lysed under denaturing conditions and subjected to Ni-NTA pulldown. Ubiquitination status of USP21 in each condition was assessed by IB. C. Top: IB of MDA-MB-231 treated with 100ng/mL cycloheximide for the indicated times 48h following transfection with the indicated siRNAs. Bottom: densitometry analysis performed on corresponding IBs to assess FOXM1 half-life in the indicated conditions.
83 Figure 3.5 USP21 affects growth through modulation of the FOXM1 transcriptional network. A. FOXM1 transcriptional activity was measured by quantifying luciferase activity in 293T cells expressing the 6x-DBE FOXM1 luciferase reporter with and without USP21 overexpression. Error bars represent SEM. B. Relative abundance of FOXM1 target transcripts was assessed by RT-qPCR following RNA extraction from MDA-MB-231 cells transfected with the indicated siRNA for 72h. Each condition represents means of triplicates. Error bars represent min-max. * = p £ 0.05 based on Student’s t-test performed on DCt values. C. IB of cells used in B. showing protein levels of various FOXM1 targets. D. Mitosis was scored by measuring P-H3-positive cells by flow cytometry in MDA-MB-231 cells transfected with the indicated siRNA or plasmids following release from aphidicolin synchronization at the indicated times. E. Growth was measured as the percentage of confluence in the dish every 6h over 120h in MDA-MB-468 (left) and MDA-MB-231 pINDUCER20 FOXM1b cells (right) treated with the indicated siRNAs (or 10ng/mL of doxycycline) using the IncuCyte live-cell imaging system. Each point represents mean of triplicates. Error bars represent SEM. ** = p £ 0.01 based on linear regression analysis.
84
Figure 3.6 USP21 regulation of cell cycle and proliferation. A. Left: RT-qPCR analysis of FOXM1 and USP21 mRNA levels in corresponding samples. Right: IB of MDA-MB-231 pINDUCER20 USP21 cells treated with vehicle or 10ng/mL doxycycline for 48h. B. IB of samples from the experiment described in Figure 3.5D. C. Left: Proliferation measured with PrestoBlue reagent in HeLa cells transfected with indicated siRNA or plasmids over 4 days. Each measurement is of triplicates. Error bars represent SEM. * = p £ 0.05 and ** = p £ 0.01 based on paired t-test. Right: IB of samples used in the experiment.
85
Figure 3.7 Amplification of FOXM1 and USP21 is linked to proliferation in BLBC. A. Top: Log2 median FOXM1 expression in TCGA breast cancers compared with proliferation score and subtype. Bottom: Log2 median FOXM1 expression and 7 FOXM1 target genes ordered by FOXM1 expression. PAM50 breast cancer subtypes: basal-like, red; HER2-enriched, hot pink; luminal A, dark blue; luminal B, light blue; normal-like, green. B. Analysis comparing mRNA overexpression (Z score threshold ± 2) and CNA amplification of 92 DUBs in 482 tumor samples from the TCGA breast invasive carcinoma dataset. C. FOXM1 and USP21 abundance assessed by IB in a panel of breast cancer cell lines. Breast cancer subtype denoted by color: normal-like, green; luminal, blue; HER2, pink; BLBC, red. D. Volcano plot showing protein expression levels in USP21-amplified patient tumors based on RPPA data from the TCGA breast cancer dataset.
86
Figure 3.8 USP21 knockdown sensitizes BLBC cells to paclitaxel. A. Relative viability of MDA-MB-231 cells (pINDUCER20 FOXM1b transduced) (top), MDA-MB-468 cells (middle) and SUM149 cells (bottom) treated with the indicated siRNAs and exposed to the indicated concentration of paclitaxel (and 10ng/mL of doxycycline [Dox] in siUSP21 + Dox) for 72h. Each condition represents mean of triplicates. Error bars represent SEM. * = p £ 0.05, ** = p £ 0.01, *** = p £ 0.001 and **** = p £ 0.0001 based on two-way ANOVA with post-hoc pairwise analysis. IB confirming expression of FOXM1 and USP21 is shown to the right for each corresponding experiment. B. Quantification of mean tumor volume of SUM149 mouse xenograft for the indicated conditions. Error bars represent SEM. ** = p £ 0.01 and *** = p £ 0.001 based on paired t-test of respective measurements.
87
Figure 3.9 Analysis of USP21 and proliferation in breast cancer. A. Proliferation score of TCGA breast cancers sorted by USP21. GISTIC scores and PAM50 breast cancer subtypes: basal-like, red; HER2-enriched, hot pink; luminal A, dark blue; luminal B, light blue; normal-like, green. B. Comparing relative levels of FOXM1 and USP21 between BLBC cell lines used throughout the study. C. IB of shRNA transduced SUM149 cells used in mouse xenograft experiment described in Figure 3.8B. * denotes non-specific band.
88
Figure 3.10 Illustration of proposed FOXM1-USP21 axis in cell cycle and disease. Under conditions where USP21 is overexpressed (right), FOXM1 is more protected from ubiquitin-dependent proteasomal degradation. This enhanced protection leads to enhanced FOXM1 stability which contributes to cell cycle progression which drives tumor growth and resistance to paclitaxel, based on mechanisms which are regulated by FOXM1. Conversely, knockdown of USP21 enhances the ubiquitin-dependent proteasomal degradation of FOXM1. Enhanced FOXM1 degradation leads to a delayed cell cycle (with cells proceeding slowly through the G2/M cell cycle transition), slowed proliferation and enhances response to paclitaxel in breast cancer.
89 Enrichment Gene Symbol Description Score Enriched? Gene Symbol null null 0.020818 Yes null glucosidase, beta; acid (includes GBA glucosylceramidase) 0.043756 Yes GBA acidic (leucine-rich) nuclear phosphoprotein ANP32E 32 family, member E 0.061532 Yes ANP32E cell division cycle CDCA2 associated 2 0.076739 Yes CDCA2 KIF14 kinesin family member 14 0.086463 Yes KIF14 family with sequence FAM64A similarity 64, member A 0.101637 Yes FAM64A NIMA (never in mitosis NEK2 gene a)-related kinase 2 0.112635 Yes NEK2 centromere protein F, CENPF 350/400ka (mitosin) 0.127031 Yes CENPF null null 0.137598 Yes null TAF5 RNA polymerase II, TATA box binding protein (TBP)-associated factor, TAF5 100kDa 0.150960 Yes TAF5 null null 0.163863 Yes null cornichon homolog 4 CNIH4 (Drosophila) 0.176389 Yes CNIH4 TTK TTK protein kinase 0.189483 Yes TTK ring finger and WD repeat RFWD3 domain 3 0.200674 Yes RFWD3 AURKB aurora kinase B 0.213545 Yes AURKB KIF23 kinesin family member 23 0.224606 Yes KIF23 kinesin family member KIF20A 20A 0.237301 Yes KIF20A TPX2, microtubule- associated, homolog TPX2 (Xenopus laevis) 0.247021 Yes TPX2 polo-like kinase 1 PLK1 (Drosophila) 0.258741 Yes PLK1 CCNA2 cyclin A2 0.270580 Yes CCNA2 CENPA centromere protein A 0.282854 Yes CENPA asp (abnormal spindle) homolog, microcephaly ASPM associated (Drosophila) 0.292733 Yes ASPM ubiquitin-conjugating UBE2C enzyme E2C 0.304951 Yes UBE2C non-SMC condensin I NCAPD2 complex, subunit D2 0.314267 Yes NCAPD2 family with sequence FAM83D similarity 83, member D 0.325644 Yes FAM83D cell division cycle CDCA8 associated 8 0.337109 Yes CDCA8 CDC20 cell division cycle 20 homolog (S. CDC20 cerevisiae) 0.342095 Yes CDC20
90 nuclear casein kinase and cyclin-dependent kinase NUCKS1 substrate 1 0.351519 Yes NUCKS1 CCNB1 cyclin B1 0.361443 Yes CCNB1 null null 0.369896 Yes null cytoskeleton associated CKAP2L protein 2-like 0.380350 Yes CKAP2L CHK2 checkpoint CHEK2 homolog (S. pombe) 0.390651 Yes CHEK2 CCNB2 cyclin B2 0.400656 Yes CCNB2 null null 0.410612 Yes null null null 0.419783 Yes null KIF11 kinesin family member 11 0.430068 Yes KIF11 KIF2C kinesin family member 2C 0.440508 Yes KIF2C pituitary tumor- PTTG1 transforming 1 0.447744 Yes PTTG1 splicing factor proline/glutamine-rich (polypyrimidine tract binding protein SFPQ associated) 0.452136 Yes SFPQ LMNB1 lamin B1 0.461743 Yes LMNB1 hematological and HN1 neurological expressed 1 0.471860 Yes HN1 null null 0.479986 Yes null calmodulin 2 (phosphorylase kinase, CALM2 delta) 0.489990 Yes CALM2 zinc finger and BTB ZBTB5 domain containing 5 0.496343 Yes ZBTB5 DEPDC1 DEP domain containing 1 0.506175 Yes DEPDC1 H2A histone family, H2AFZ member Z 0.510321 Yes H2AFZ NAV2 neuron navigator 2 0.518510 Yes NAV2 sperm associated antigen SPAG5 5 0.523995 Yes SPAG5 AURKA aurora kinase A 0.525815 Yes AURKA Rac GTPase activating RACGAP1 protein 1 0.534634 Yes RACGAP1 null null 0.539924 Yes null Rho GTPase activating ARHGAP11A protein 11A 0.543528 Yes ARHGAP11A extra spindle poles like 1 ESPL1 (S. cerevisiae) 0.551075 Yes ESPL1 inner centromere protein INCENP antigens 135/155kDa 0.554538 Yes INCENP ubiquitin-conjugating UBE2S enzyme E2S 0.559898 Yes UBE2S karyopherin alpha 2 (RAG cohort 1, importin alpha KPNA2 1) 0.567097 Yes KPNA2 cytoskeleton associated CKAP2 protein 2 0.572269 Yes CKAP2
91 nucleolar and spindle NUSAP1 associated protein 1 0.579700 Yes NUSAP1 hyaluronan-mediated motility receptor HMMR (RHAMM) 0.585622 Yes HMMR chromosome 15 open C15ORF23 reading frame 23 0.592526 Yes C15ORF23 HMGB2 high-mobility group box 2 0.600037 Yes HMGB2 null null 0.605212 Yes null protein regulator of PRC1 cytokinesis 1 0.612895 Yes PRC1 kinesin family member KIF18A 18A 0.617277 Yes KIF18A cyclin-dependent kinase inhibitor 3 (CDK2- associated dual specificity CDKN3 phosphatase) 0.622019 Yes CDKN3 ubiquitin specific peptidase 13 USP13 (isopeptidase T-3) 0.627647 Yes USP13 FBXO5 F-box protein 5 0.633751 Yes FBXO5 PHF19 PHD finger protein 19 0.639432 Yes PHF19 family with sequence FAM100B similarity 100, member B 0.639173 Yes FAM100B citron (rho-interacting, serine/threonine kinase CIT 21) 0.641250 Yes CIT Rho GTPase activating ARHGAP11B protein 11B 0.645701 Yes ARHGAP11B ZNF165 zinc finger protein 165 0.652198 Yes ZNF165 null null 0.658336 Yes null butyrophilin, subfamily 2, BTN2A1 member A1 0.657704 Yes BTN2A1 ATPase family, AAA ATAD2 domain containing 2 0.654522 Yes ATAD2 null null 0.656237 Yes null epithelial cell transforming ECT2 sequence 2 oncogene 0.655710 Yes ECT2 cytoskeleton associated CKAP5 protein 5 0.659738 Yes CKAP5 additional sex combs like ASXL1 1 (Drosophila) 0.663116 Yes ASXL1 cyclin-dependent kinase inhibitor 2D (p19, inhibits CDKN2D CDK4) 0.668286 Yes CDKN2D TUBB2C tubulin, beta 2C 0.673261 Yes TUBB2C topoisomerase (DNA) II TOP2A alpha 170kDa 0.676790 Yes TOP2A glutamine and serine rich QSER1 1 0.668105 No QSER1 SVIL supervillin 0.613905 No SVIL centrosomal protein CEP192 192kDa 0.612476 No CEP192 protein phosphatase 1, PPP1R10 regulatory subunit 10 0.596670 No PPP1R10
92 FNBP4 formin binding protein 4 0.589752 No FNBP4 PTMS parathymosin 0.592817 No PTMS MAD2L1BP MAD2L1 binding protein 0.590704 No MAD2L1BP transcription termination TTF2 factor, RNA polymerase II 0.573532 No TTF2 H2A histone family, H2AFX member X 0.564126 No H2AFX fizzy/cell division cycle 20 FZR1 related 1 (Drosophila) 0.531594 No FZR1 null null 0.500078 No null centrosomal protein CEP110 110kDa 0.432021 No CEP110 SMAD, mothers against DPP homolog 6 SMAD6 (Drosophila) 0.379289 No SMAD6 v-Ki-ras2 Kirsten rat sarcoma viral oncogene KRAS homolog 0.379367 No KRAS null null 0.337533 No null null null 0.322027 No null RNF26 ring finger protein 26 0.309723 No RNF26 WEE1 homolog (S. WEE1 pombe) 0.307540 No WEE1 serine/threonine kinase STK17A 17a (apoptosis-inducing) 0.302677 No STK17A protein phosphatase 1B (formerly 2C), magnesium-dependent, PPM1B beta isoform 0.300841 No PPM1B kelch repeat and BTB (POZ) domain containing KBTBD2 2 0.300771 No KBTBD2 HIST1H2AG histone cluster 1, H2ag 0.288100 No HIST1H2AG apolipoprotein L domain APOLD1 containing 1 0.282930 No APOLD1 null null 0.219432 No null null null 0.137236 No null null null 0.121622 No null ZNF785 zinc finger protein 785 0.074998 No ZNF785 DZIP3 - 0.071994 No DZIP3 dehydrogenase/reductase DHRS7B (SDR family) member 7B 0.057230 No DHRS7B SMAD, mothers against DPP homolog 3 SMAD3 (Drosophila) 0.010142 No SMAD3 null null -0.009193 No null null null 0.000685 No null
Table 3.1 FOXM1 target gene enrichment correlated to USP21 amplification. Gene Set Enrichment Analysis (GSEA) was performed for a previously derived FOXM1 target gene list comparing USP21-amplified and USP21- non-amplified patient tumors.
93 Primer Name Sequence (5' - 3') Note siFF CGUACGCGGAAUACUUCGA siUSP21 1 GAGCAGCACUCGACCUCUU siUSP21 2 GACAAGAUGGCUCAUCACA siUSP21 3 GAGCUGUCUUCCAGAAAUA siFOXM1 1 CAACAGGAGUCUAAUCAAG siFOXM1 2 GUAGUGGGCCCAACAAAUU GAGCUGACGUGUACUCAUA GGACAACCCUGCUCGAAUC siUSP5 GGAGAGACAUUUCAAUAAG GAUCUACAAUGACCAGAAA GAAGAUGGGUGAUUUACAA GCACUGGAUUGGAUCUUUA siUSP13 GCACGAAACUGAAGCCAAU UGAUUGAGAUGGAGAAUAA AGAGCGAGCUGGCGGGUUU GCUCGGAGUAUCUGAAGUA siUSP35 CAACAUCCUUUACCUACAG GGGCUUUGAUGAAGACAAG GCACACCACUGAAGAGAUU GAAAGCAGAUGUCCUGAGU siUSP36 GGACUCGGCUGAUGAUGGA CCGUAUAUGUCCCAGAAUA shNonT CAACAAGATGAAGAGCACCAA pLKO.1 backbone shUSP21 CCTGCGAAGCTGTGAATCCTA pLKO.1 backbone Note: All siRNA have [dT][dT] overhang. All siRNA were used pooled unless otherwise noted. Table 3.2 siRNA and shRNA used in Chapter 3 experiments.
94 Primer Name Sequence (5' - 3') Notes AURKA qPCR GAGGTCCAAAACGTGTTCTCG Fwd ACAGGATGAGGTACACTGGTTG Rev CCNB1 qPCR TTCCAGTTATGCAGCACCT Fwd ACTTGTTCTTGACAGTCATGTG Rev CDK1 qPCR GGTTCCTAGTACTGCAATTCG Fwd AGCACATCCTGAAGACTGAC Rev FOXM1 qPCR CGTGGATTGAGGACCACTTT Fwd GGCTTAAACACCTGGTCCAA Rev GAPDH qPCR GGCCTCCAAGGAGTAAGACC Fwd AGGGGTCTACATGGCAACTG Rev PLK1 qPCR GACAAGTACGGCCTTGGGTA Fwd GTGCCGTCACGCTCTATGTA Rev USP21 qPCR GATTGTGGACCTGTTTGTGG Fwd GAAACAATCCCGCAGAGAC Rev CCTGGGAAACACGGCCTTCCTGAATGCTG Fwd C221A USP21 Catalytically Dead SDM CAGCATTCAGGAAGGCCGTGTTTCCCAGG Rev C221A GCAGTGTCTGAGCTCCACTCGACCTCTTC Fwd AG694TC GAAGAGGTCGAGTGGAGCTCAGACACTGC Rev AG694TC TCTGAGCTCCACTCGGCCTCTTCGGGACTT Fwd A702G AAGTCCCGAAGAGGCCGAGTGGAGCTCAGA Rev A702G USP21 siRNA Resistant SDM CCCGAAGAGGCCGTGTGGAGCTCAGAC Fwd T699A GTCTGAGCTCCACACGGCCTCTTCGGG Rev T699A GCTGCAGTGTCTAAGCTCCACTCGG Fwd G60A CCGAGTGGAGCTTAGACACTGCAGC Rev G60A Table 3.3 Primers used in Chapter 3 experiments.
95 Antibody Company/Product # Species Aurora A CST / 3092 Rabbit Anti-Mouse IgG (H+L) HRP Jackson ImmunoResearch / 115-035-003 Goat Anti-Mouse IgG (L) HRP Jackson ImmunoResearch / 115-035-174 Goat Anti-Rabbit IgG (H+L) Alexa Fluor 488 Thermo Fisher / A-11034 Goat Anti-Rabbit IgG (H+L) HRP Jackson ImmunoResearch / 111-035-003 Goat Anti-Rabbit IgG (L) HRP Jackson ImmunoResearch / 211-032-171 Mouse CCP110 Bethyl / A301-343A Rabbit CDK1 Bethyl / A303-663A Rabbit Cyclin B1 Abcam / ab32053 Rabbit Cyclin E1 CST / 4129 Mouse FLAG M2 Sigma-Aldrich / F1804 Mouse FOXM1 Santa Cruz / sc-502 Rabbit GAPDH Santa Cruz / sc-25778 Rabbit GST Gene Tex / GTX114099 Rabbit HA BioLegend/Covance / MMS-101P Mouse Myc Santa Cruz / sc-40 Mouse Normal Mouse IgG Santa Cruz / sc-2025 Mouse Phospho-Histone H3 S10 CST / 3377 Rabbit Tubulin Santa Cruz / sc-32293 Mouse USP21 Rabbit Table 3.4 Antibodies used in Chapter 3 experiments.
96 CHAPTER 4: CONCLUSIONS AND FUTURE DIRECTIONS
At its core, the cell cycle is orchestrated by a series of molecular steps that must be carried out in an ordered and timely fashion. The transcription of genes necessary to create the proteins that will allow the cell to proceed from one step of the growth cycle to the next at the right time is essential. Also necessary is the destruction of proteins that either serve to restrain the cell cycle when the time is right to push forward or that were once needed to satisfy the growth requirements for one stage but are no longer needed for the next phase. The UPS oversees this fundamental process and also serves to activate and fine-tune myriad signaling pathways in a multitude of ways that we are still uncovering. Taken broadly, these two pathways, transcriptional activation and ubiquitin-dependent destruction, are intimately linked together. By understanding all of the connections between these two pathways we can shed light on the molecular network of events that allow us to develop and grow, and also better understand the diseases that result when these networks become dysregulated.
The previous chapters focus on FOXM1, one transcription factor that is a master transcriptional regulator of the cell cycle. Its importance in both normal cell cycle regulation and the promotion of a variety of cancers is well documented. Though a great deal is known about the interactions with various kinases and phosphatases that regulate FOXM1 through the addition and removal of phosphorylation, respectively, much less was known about its regulation through ubiquitin-dependent mechanisms. The work presented here extends our knowledge of how the
UPS regulates FOXM1 and as a result, regulates cell cycle transcription.
In Chapter 2, experiments are described that identify FOXM1 as a substrate of the E3 ubiquitin ligase CUL4VPRBP. CUL4VPRBP is able to bind to FOXM1, adding polyubiquitin chains that lead to FOXM1 proteasomal degradation primarily during S-phase. Knockdown of VPRBP
97 is sufficient to increase the stability of FOXM1. Interestingly, ectopic overexpression of VPRPB seems to activate the transcriptional activity of FOXM1, a process which seems to occur independent of the association of VPRBP with the CUL4 scaffold. Furthermore, The FOXM1-
CUL4VPRBP axis seems to be upregulated in high-grade serous ovarian cancer, where the FOXM1 network of genes is highly activated. The activation of FOXM1 by VPRBP may in part explain why the FOXM1 transcriptional network is so amplified in ovarian cancer despite the fact that
FOXM1 protein and mRNA is not highly expressed in this type of cancer. Taken together, these studies unveil a new link between the UPS and cell cycle transcription which demonstrates the complex role that ubiquitin can play as both a degradative and an activating post-translational modification and how looking at the influence of the UPS on transcriptional control may explain how transcriptional networks can be upregulated in spite of obvious mechanistic explanation.
In Chapter 3, the experiments identify and characterize FOXM1 as a substrate of the
DUB USP21. The experiments demonstrate that USP21 is able to directly bind and remove polyubiquitin chains from FOXM1. As such, USP21 overexpression is sufficient to enhance the stability of FOXM1 by protecting it from proteasomal degradation and the knockdown of USP21 is sufficient to destabilize FOXM1 by promoting its proteasomal degradation. Through this action, USP21 levels are able to influence the broader FOXM1 transcriptional network through its effect on FOXM1. The FOXM1 is able to influence mitotic progression, which more broadly influences cell growth and proliferation. To that point, USP21 knockdown is able to slow both cell cycle progression through mitosis and cell growth in a FOXM1-dependent manner.
Intriguingly, both the expression of USP21, FOXM1 and numerous FOXM1 transcriptional targets are significantly higher in basal-like breast cancer, a particularly aggressive subtype of breast cancer with few treatment options and through which the upregulation of the FOXM1
98 transcriptional network is believed to contribute significantly to disease progression.
Intriguingly, the USP21 gene is located on chromosome 1q23, a region that is focally amplified in the vast majority of basal-like breast cancers, but to which a clear oncogene has yet to be attributed (85). Finally, the clinical implications that USP21 knockdown may have on sensitizing basal-like breast cancer to the frontline chemotherapy paclitaxel is investigated. No targeted treatment options are available for aggressive basal-like breast cancers and furthermore, FOXM1 overexpression is linked to paclitaxel resistance (195). In both cell culture and mouse xenograft models, it was demonstrated that USP21 knockdown is able to sensitize cells to paclitaxel in a
FOXM1-dependent manner. Furthermore, this was demonstrated in a cell line SUM149, that is resistant to clinically-relevant concentrations of paclitaxel. Taken together, this study not only describes one mechanism which the UPS works to promote the stability of FOXM through the activity of USP21, but also sheds insight into a possible interaction which may be relevant to the promotion of basal-like breast cancer and might also be an avenue of therapeutic intervention to treat the disease clinically.
There are a number of implications and future directions branching off from this research.
One area of inquiry that should be explored is the molecular nature of the Vpr-induced G2/M arrest that occurs following HIV infection. VPRBP is named for its known association with the
HIV viral accessory protein Vpr, which engages with CUL4VPRBP to degrade a number of host proteins (36–38, 40). Though it is well established that a G2/M arrest follows a productive HIV infection, and that this arrest hinges on Vpr association with VPRBP, the molecular reason for this remains incomplete (43–47). The research described in Chapter 2 demonstrates that
FOXM1, the master transcriptional regulator of mitosis, is bound and ubiquitylated by
CUL4VPRBP. It is intriguing to think that enhanced degradation of FOXM1 through Vpr
99 VPRBP associating with CUL4 is the key molecular event that arrests cells in G2/M during HIV infection. It is well established and confirmed by studies in Chapter 3 that FOXM1 knockdown is sufficient to arrest cells in G2. Future work should explore this explanation within the context of
HIV infection in relevant T-cell lineages.
Another intriguing line of inquiry lies in understanding how changes to the epigenetic landscape mediated by VPRBP and USP21 change FOXM1-dependent transcriptional activity.
In addition to its role as a ubiquitin ligase, it was discovered that the N-terminal of VPRBP contains a cryptic kinase domain which is able to regulate phosphorylation on histone H2AT120
(276). This post-translational modification is tied to regulating transcriptional activity.
Additionally, recent work demonstrated that H2AT120 phosphorylation and H2AK119 ubiquitination are antagonistic post-translational modifications that compete to influence transcriptional activity of genes such as CCND1, which intriguingly, is a well-characterized
FOXM1 transcriptional target (277). Even more intriguing is that the removal of transcriptionally repressive ubiquitin from H2AK119 is mediated by USP21 (73, 277). As such, there is a lot of connectedness between FOXM1, VPRBP and USP21 that converge at this level. The model suggested by Aihara et al. proposes that phosphorylation on H2AT120 helps to activate genes whereas ubiquitination on H2AK119 is associated with the repression of active transcription.
The research presented here is compatible with this model in which both VPRBP has been shown to activate FOXM1 and localize to FOXM1 target promoters and that USP21 also promotes the transcription of FOXM1 target genes. Perhaps both USP21 and VPRBP are necessary to sculpt the histone landscape to promote active transcription of a number of targets in addition to their activity in stabilizing FOXM1 and promoting its transcriptional activity respectively. Future work should study what is happening with these histone modifications when
100 VPRBP and USP21 levels are altered and what result this has on localizing FOXM1 to target promoters and on FOXM1 target transcription. Additionally, it should be explored if VPRBP- dependent phosphorylation is in any way needed for the activation of FOXM1 as it is well established that phosphorylation is needed to fully activate FOXM1. Furthermore, it should be explored if USP21 is also targeted to FOXM1 target promoters much like VPRBP is. This would be compatible with the proposed model from Aihara et al. and also compatible with research that suggests that FOXM1 binding to transcriptional coactivators is very important in dictating its localization to various target gene promoters.
Future work should better characterize the specific amino acids recognized by USP21 on
FOXM1. While we do have recognition motifs for many classes of enzymes which confer post- translational modifications on substrates, there are few rules dictating substrate recognition by
DUBs. The interaction domain on FOXM1 recognized by USP21 was narrowed down to a 79 amino acids region on FOXM1 which contained a large portion of the Forkhead box motif.
Considering VPRBP binding to FOXM1 was dependent on this region as well and was shown to be highly dependent on residues 418, 419 and 422, it would be interesting to determine if these regions also affected USP21 binding or influence on FOXM1 ubiquitination status. This would tie USP21 and VPRBP closer together in the control of FOXM1 stability.
There are still many specific questions remaining surrounding how the interaction between FOXM1 and USP21 is regulated. It is still unclear if there is a particular time in the cell cycle where the USP21-FOXM1 interaction is more productive. It is also unclear what role phosphorylation has to play in this interaction. The results from the ectopic co-IP studies are suggestive that phosphorylated FOXM1 is a stronger interactor with USP21 (Figure 3.1D), but it remains true that these proteins are able to interact without phosphorylation modifying either
101 protein (Figure 3.1G). USP21 is known to be phosphorylated itself, but it is unclear what role this plays in enhancing or repressing USP21 activity.
In light of the findings which implicate FOXM1 as an essential piece of the Wnt/b- catenin signaling cascade, it is interesting to speculate how USP21 fits into this model (169).
Indeed, a genome wide screen of DUBs involved in potentiating Wnt/b-catenin signaling demonstrated that USP21 is the second most potent activator or inhibitor of signaling with overexpression and knockdown, respectively (278). The molecular details of this observation were not interrogated more deeply. It is interesting to speculate that the stabilization of FOXM1 in a USP21-dependent manner is at work here. It is also interesting to speculate that the FOXM1-
USP21 interaction is enhanced by Wnt signaling, as has been demonstrated by the interaction between FOXM1 and USP5 (183).
Additional experiments should be done to address the role that USP21 has in mediating the effects of paclitaxel through direct association with centrosomes and microtubules. In
Chapter 3 it is demonstrated that USP21 knockdown can sensitize cells to the microtubule poison paclitaxel. Though this effect is mostly rescued by forced FOXM1 overexpression, it is not entirely rescued. Various groups have reported that USP21 is an important regulator of centrosome and microtubule-associated functions (71). The functions that USP21 has directly on the centrosome or on microtubules combined with paclitaxel remain completely unexplored and may prove insightful in fully understanding the observed sensitization that USP21 knockdown confers.
The finding that the FOXM1-USP21 interaction seems to enhance basal-like breast cancer growth and treatment sensitivity should be explored in additional cancer models. Bladder
102 cancer is particularly interesting due to the molecular similarities to basal-like breast cancer and the high degree of USP21 expression in the disease. Additionally, there are two common mutations in the USP domain of USP21 at R441 and S476 that occur across cancer types (279).
It would be interesting to dissect what effect these mutations had on USP21 substrate ubiquitination given that USP21 amplification seems to be associated with cancer.
Finally, a more complete catalog of USP21 substrates identified by mass spectrometry would be beneficial. Studies have been performed to broadly identify substrates of USP21 by IP-
MS/MS (58), but I believe the lysis techniques utilized to analyze substrates by IP miss a large portion of the available substrates as many nuclear and chromatin interacting substrates are likely removed from analysis. The application of diGly (280) proteomics with DUB knockdown and overexpression to capture broad changes to protein ubiquitination has the potential to discover novel USP21 substrates or proteins which are directly or indirectly affect by USP21 knockdown or overexpression.
103 REFERENCES
1. A. Ciechanover, Y. Hod, A. Hershko, 8,1978. Biochem. Biophys. Res. Commun. 81, 1100–1105 (1978). 2. A. Hershko, A. Ciechanover, H. Heller, A. L. Haas, I. A. Rose, Proposed role of ATP in protein breakdown: Conjugation of proteins with multiple chains of the polypeptide of ATP-dependent proteolysis (protein/turnover/energy dependence/isopeptide linkage). Biochemistry. 77, 1783–1786 (1980). 3. L. Waxman, J. M. Fagan, A. L. Goldberg, Demonstration of two distinct high molecular weight proteases in rabbit reticulocytes, one of which degrades ubiquitin conjugates. J. Biol. Chem. 262, 2451–2457 (1987). 4. R. Hough, G. Pratt, M. Rechsteiner, Purification of two high molecular weight proteases from rabbit reticulocyte lysate. J. Biol. Chem. 262, 8303–8313 (1987). 5. J. Driscoll, A. L. Goldberg, The proteasome (multicatalytic protease) is a component of the 1500-kDa proteolytic complex which degrades ubiquitin-conjugated proteins. J. Biol. Chem. 265, 4789–4792 (1990). 6. M. J. Emanuele et al., Global Identification of Modular Cullin-RING Ligase Substrates. Cell. 147, 459–474 (2011). 7. D. Hoeller, I. Dikic, Targeting the ubiquitin system in cancer therapy. Nature. 458, 438– 444 (2009). 8. D. Popovic, D. Vucic, I. Dikic, Ubiquitination in disease pathogenesis and treatment. Nat. Med. 20, 1242–1253 (2014). 9. F. Bassermann, R. Eichner, M. Pagano, The ubiquitin proteasome system - Implications for cell cycle control and the targeted treatment of cancer. Biochim. Biophys. Acta - Mol. Cell Res. 1843, 150–162 (2014). 10. A. Hershko, A. Ciechanover, The Ubiquitin System. Annu. Rev. Biochem. 67, 425–79 (1988). 11. A. L. Haas, I. A. Rose, The Mechanism of Ubiquitin Activating Enzyme. 257, 10329– 10337 (1982). 12. A. Hershko et al., Components of Ubiquitin-Protein Ligase System. J Biol Chem. 258, 8206–8214 (1983). 13. C. M. Pickart, Mechanisms Underlying Ubiquitination. Annu. Rev. Biochem. 70, 503–533 (2001). 14. M. Scheffner, U. Nuber, J. M. Huibregtse, Protein ubiquitination involving an E1–E2–E3 enzyme ubiquitin thioester cascade. Nature. 373, 81–83 (1995). 15. L. Hicke, Protein regulation by monoubiquitin. Nat. Rev. Mol. Cell Biol. 2, 195–201 (2001).
104 16. M. J. Clague, S. Urbé, D. Komander, Breaking the chains: deubiquitylating enzyme specificity begets function. Nat. Rev. Mol. Cell Biol. (2019), doi:10.1038/s41580-019- 0099-1. 17. J. H. Hurley, S. Lee, G. Prag, Ubiquitin-binding domains. Biochem. J. 399, 361–372 (2006). 18. R. J. Deshaies, C. A. P. Joazeiro, RING Domain E3 Ubiquitin Ligases. Annu. Rev. Biochem. 78, 399–434 (2009). 19. M. D. Petroski, R. J. Deshaies, Function and regulation of cullin-RING ubiquitin ligases. Nat. Rev. Mol. Cell Biol. 6, 9–20 (2005). 20. D. C. Scott et al., Two Distinct Types of E3 Ligases Work in Unison to Regulate Substrate Ubiquitylation. Cell. 166, 1198-1214.e24 (2016). 21. D. M. Duda, L. A. Borg, D. C. Scott, H. W. Hunt, B. A. Schulman, Structural Insights into NEDD8 Activation of Cullin-RING Ligases : Conformational Control of Conjugation. 134, 995–1006 (2008). 22. S. Lyapina et al., Promotion of NEDD8-CUL1 conjugate cleavage by COP9 signalosome. Science (80-. ). 292, 1382–1385 (2001). 23. J. Zheng et al., CAND1 Binds to Unneddylated CUL1 and Regulates the Formation of SCF Ubiquitin E3 Ligase Complex Despite the importance of cullins in controlling many essential biological processes, the mechanism that reg-ulates the cullin-containing ubiquitin E3 ligases . Mol. Cell. 10, 1519–1526 (2002). 24. N. W. Pierce et al., Cand1 promotes assembly of new SCF complexes through dynamic exchange of F box proteins. Cell. 153, 206–215 (2013). 25. Y. Sang, F. Yan, X. Ren, The role and mechanism of CRL4 E3 ubiquitin ligase in cancer and its potential therapy implications. Oncotarget. 6 (2015), doi:10.18632/oncotarget.6052. 26. Y. J. He, C. M. McCall, J. Hu, Y. Zeng, Y. Xiong, DDB1 functions as a linker to recruit receptor WD40 proteins to CUL4-ROC1 ubiquitin ligases. Genes Dev. 20, 2949–2954 (2006). 27. J. Lee, P. Zhou, DCAFs, the Missing Link of the CUL4-DDB1 Ubiquitin Ligase. Mol. Cell. 26, 775–780 (2007). 28. A. I. Su et al., A gene atlas of the mouse and human.pdf. Proc. Natl. Acad. Sci. 101, 6062–6067 (2004). 29. J. Hannah, P. Zhou, Distinct and overlapping functions of the cullin E3 ligase scaffolding proteins CUL4A and CUL4B. Gene. 573, 33–45 (2015). 30. L.-C. Chen et al., Cancer Res., in press (available at http://cancerres.aacrjournals.org/content/58/16/3677.abstract). 31. M. C. Abba et al., Identification of novel amplification gene targets in mouse and human breast cancer at a syntenic cluster mapping to mouse ch8A1 and human ch13q34. Cancer
105 Res. 67, 4104–4112 (2007). 32. L. Melchor et al., Comprehensive characterization of the DNA amplification at 13q34 in human breast cancer reveals TFDP1 and CUL4A as likely candidate target genes. Breast Cancer Res. 11, 1–14 (2009). 33. T. H. Lee, S. J. Elledge, J. S. Butel, Hepatitis B virus X protein interacts with a probable cellular DNA repair protein. J Virol. 69, 1107–1114 (1995). 34. S. Bontron, N. Lin-Marq, M. Strubin, Hepatitis B virus X protein associated with UV- DDB1 induces cell death in the nucleus and is functionally antagonized by UV-DDB2. J. Biol. Chem. 277, 38847–38854 (2002). 35. G. Y. Lin, R. G. Paterson, C. D. Richardson, R. A. Lamb, The V protein of the paramyxovirus SV5 interacts with damage-specific DNA binding protein. Virology. 249, 189–200 (1998). 36. S. Zhang, Y. Feng, O. Narayan, L. J. Zhao, Cytoplasmic retention of HIV-1 regulatory protein V pr by protein-protein interaction with a novel human cytoplasmic protein V prBP. Gene. 263, 131–140 (2001). 37. J. Andrejeva, E. Poole, D. F. Young, S. Goodbourn, R. E. Randall, The p127 Subunit (DDB1) of the UV-DNA Damage Repair Binding Protein Is Essential for the Targeted Degradation of STAT1 by the V Protein of the Paramyxovirus Simian Virus 5. J. Virol. 76, 11379–11386 (2002). 38. B. Schrofelbauer, Q. Yu, S. G. Zeitlin, N. R. Landau, Human Immunodeficiency Virus Type 1 Vpr Induces the Degradation of the UNG and SMUG Uracil-DNA Glycosylases. J. Virol. 79, 10978–10987 (2005). 39. B. Schrofelbauer, Y. Hakata, N. R. Landau, HIV-1 Vpr function is mediated by interaction with the damage-specific DNA-binding protein DDB1. Proc. Natl. Acad. Sci. 104, 4130– 4135 (2007). 40. K. Hrecka et al., Vpx relieves inhibition of HIV-1 infection of macrophages mediated by the SAMHD1 protein. Nature. 474, 658–661 (2011). 41. J. Ahn et al., HIV/Simian Immunodeficiency Virus (SIV) Accessory Virulence Factor Vpx Loads the Host Cell Restriction Factor SAMHD1 onto the E3 Ubiquitin Ligase Complex CRL4 DCAF1. J. Biol. Chem. 287, 12550–12558 (2012). 42. D. Schwefel et al., Structural basis of lentiviral subversion of a cellular protein degradation pathway. Nature. 505, 234–238 (2014). 43. J. P. Belzile et al., HIV-1 Vpr-mediated G2 arrest involves the DDB1-CUL4AVPRBP E3 ubiquitin ligase. PLoS Pathog. 3, 0882–0893 (2007). 44. K. Hrecka et al., Lentiviral Vpr Usurps Cul4-DDB1[VprBP] E3 Ubiquitin Ligase to Modulate Cell Cycle. Proc. Natl. Acad. Sci. 104, 11778–11783 (2007). 45. E. Le Rouzic et al., HIV1 Vpr arrests the cell cycle by recruiting DCAF1/VprBP, a receptor of the Cul4-DDB1 ubiquitin ligase. Cell Cycle. 6, 182–188 (2007).
106 46. X. Wen, K. M. Duus, T. D. Friedrich, C. M. C. De Noronha, The HIV1 protein Vpr acts to promote G2 cell cycle arrest by engaging a DDB1 and cullin4A-containing ubiquitin ligase complex using VprBP/DCAF1 as an adaptor. J. Biol. Chem. 282, 27046–27051 (2007). 47. J. L. Andersen, E. Le Rouzic, V. Planelles, HIV-1 Vpr: Mechanisms of G2 arrest and apoptosis. Exp. Mol. Pathol. 85, 2–10 (2008). 48. P. S. Tarpey et al., Mutations in CUL4B, Which Encodes a Ubiquitin E3 Ligase Subunit, Cause an X-linked Mental Retardation Syndrome Associated with Aggressive Outbursts, Seizures, Relative Macrocephaly, Central Obesity, Hypogonadism, Pes Cavus, and Tremor. Am. J. Hum. Genet. 80, 345–352 (2007). 49. E. J. Worden, C. Padovani, A. Martin, Structure of the Rpn11-Rpn8 dimer reveals mechanisms of substrate deubiquitination during proteasomal degradation. Nat. Struct. Mol. Biol. 21, 220–227 (2014). 50. R. Verma et al., Role of Rpn11 Metalloprotease in Deubiquitination and Degradation by the 26S Proteasome. Science (80-. ). 298, 611–615 (2002). 51. T. Yao, R. E. Cohen, A cryptic protease couples deubiquitination and degradation by the proteasome. Nature. 419, 403–407 (2002). 52. T. E. T. Mevissen, D. Komander, Mechanisms of Deubiquitinase Specificity and Regulation. Annu. Rev. Biochem. 86, 159–192 (2017). 53. D. Komander, M. J. Clague, S. Urbé, Breaking the chains: structure and function of the deubiquitinases. Nat. Rev. Mol. Cell Biol. 10, 550–563 (2009). 54. B. J. Winborn et al., The deubiquitinating enzyme ataxin-3, a polyglutamine disease protein, edits Lys63 linkages in mixed linkage ubiquitin chains. J. Biol. Chem. 283, 26436–26443 (2008). 55. T. E. T. Mevissen et al., OTU deubiquitinases reveal mechanisms of linkage specificity and enable ubiquitin chain restriction analysis. Cell. 154, 169 (2013). 56. S. A. Abdul Rehman et al., MINDY-1 Is a Member of an Evolutionarily Conserved and Structurally Distinct New Family of Deubiquitinating Enzymes. Mol. Cell. 63, 146–155 (2016). 57. S. A. H. de Poot, G. Tian, D. Finley, Meddling with Fate: The Proteasomal Deubiquitinating Enzymes. J. Mol. Biol. 429, 3525–3545 (2017). 58. M. E. Sowa, E. J. Bennett, S. P. Gygi, J. W. Harper, Defining the Human Deubiquitinating Enzyme Interaction Landscape. Cell. 138, 389–403 (2009). 59. S. Schulz et al., Ubiquitin-specific protease-like 1 (USPL1) is a SUMO isopeptidase with essential, non-catalytic functions. EMBO Rep. 13, 930–938 (2012). 60. M. P. Malakhov, O. A. Malakhova, K. Il Kim, K. J. Ritchie, D. E. Zhang, UBP43 (USP18) specifically removes ISG15 from conjugated proteins. J. Biol. Chem. 277, 9976– 9981 (2002).
107 61. S. Cavadini et al., Cullin-RING ubiquitin E3 ligase regulation by the COP9 signalosome. Nature. 531, 598–603 (2016). 62. S. Urbe et al., Systematic survey of deubiquitinase localization identifies USP21 as a regulator of centrosome- and microtubule-associated functions. Mol. Biol. Cell. 23, 1095– 1103 (2012). 63. M. J. Clague, C. Heride, S. Urbé, The demographics of the ubiquitin system. Trends Cell Biol. 25, 417–426 (2015). 64. D. N. Itzhak, S. Tyanova, J. Cox, G. H. Borner, Global, quantitative and dynamic mapping of protein subcellular localization. Elife. 5, 1–36 (2016). 65. J. A. Harrigan, X. Jacq, N. M. Martin, S. P. Jackson, Deubiquitylating enzymes and drug discovery: Emerging opportunities. Nat. Rev. Drug Discov. 17, 57–77 (2018). 66. T. S. Smith, C. Southan, Sequencing, tissue distribution and chromosomal assignment of a novel ubiquitin-specific protease USP23. Biochim. Biophys. Acta - Gene Struct. Expr. 1490, 184–188 (2000). 67. L. Gong, T. Kamitani, S. Millas, E. T. H. Yeh, Identification of a novel isopeptidase with dual specificity for ubiquitin- and NEDD8-conjugated proteins. J. Biol. Chem. 275, 14212–14216 (2000). 68. Y. Ye et al., Polyubiquitin binding and cross-reactivity in the USP domain deubiquitinase USP21. EMBO Rep. 12, 350–357 (2011). 69. Y. C. Shin, J. H. Chen, S. C. Chang, The molecular determinants for distinguishing between ubiquitin and NEDD8 by USP2. Sci. Rep. 7, 2–11 (2017). 70. I. García-Santisteban, S. Bañuelos, J. A. Rodríguez, A global survey of CRM1-dependent nuclear export sequences in the human deubiquitinase family. Biochem. J. 441, 209–217 (2011). 71. S. Urbe et al., Systematic survey of deubiquitinase localization identifies USP21 as a regulator of centrosome- and microtubule-associated functions. Mol. Biol. Cell. 23, 1095– 1103 (2012). 72. H. T. Nguyen, J.-M. Kugler, A. C. Loya, S. M. Cohen, USP21 regulates Hippo pathway activity by mediating MARK protein turnover. Oncotarget. 8, 64095–64105 (2017). 73. T. Nakagawa et al., Deubiquitylation of histone H2A activates transcriptional initiation via trans-histone cross-talk with H3K4 di- and trimethylation. Genes Dev. 22, 37–49 (2008). 74. Y. Chen, B. Zhou, D. Chen, USP21 promotes cell proliferation and metastasis through suppressing EZH2 ubiquitination in bladder carcinoma. Onco. Targets. Ther. 10, 681–689 (2017). 75. J. Zhang et al., Identification of the E3 deubiquitinase ubiquitin-specific peptidase 21 (USP21) as a positive regulator of the transcription factor GATA3. J. Biol. Chem. 288, 9373–9382 (2013).
108 76. C. Heride et al., The centrosomal Deubiquitylase USP21 regulates Gli1 transcriptional activity and stability. J. Cell Sci. 129, 4001–4013 (2016). 77. J. Jin et al., The deubiquitinase USP21 maintains the stemness of mouse embryonic stem cells via stabilization of Nanog. Nat. Commun. 7, 13594 (2016). 78. L. Peng, Y. Hu, D. Chen, S. Jiao, S. Sun, Ubiquitin specific peptidase 21 regulates interleukin-8 expression, stem-cell like property of human renal cell carcinoma. Oncotarget. 7 (2016), doi:10.18632/oncotarget.9751. 79. G. Xu et al., Ubiquitin-specific peptidase 21 inhibits tumor necrosis factor α-induced nuclear factor κB activation via binding to and deubiquitinating receptor-interacting protein 1. J. Biol. Chem. 285, 969–978 (2010). 80. Y. Fan et al., USP21 negatively regulates antiviral response by acting as a RIG-I deubiquitinase. J. Exp. Med. 211, 313–28 (2014). 81. L. Tao et al., Deubiquitination and stabilization of IL-33 by USP21. Int. J. Clin. Exp. Pathol. 7, 4930–4937 (2014). 82. Y. Chen et al., p38 inhibition provides anti–DNA virus immunity by regulation of USP21 phosphorylation and STING activation. J. Exp. Med. 214, 991–1010 (2017). 83. Y. Li et al., USP21 prevents the generation of T-helper-1-like Treg cells. Nat. Commun. 7, 1–10 (2016). 84. J. Liu et al., Ubiquitin-specific protease 21 stabilizes BRCA2 to control DNA repair and tumor growth. Nat. Commun. 8 (2017), doi:10.1038/s41467-017-00206-2. 85. G. O. Silva et al., Cross-species DNA copy number analyses identifies multiple 1q21-q23 subtype-specific driver genes for breast cancer. Breast Cancer Res. Treat. 152, 347–56 (2015). 86. P. Nurse, A long twentieth century of the cell cycle and beyond. Cell. 100, 71–78 (2000). 87. J. D. Watson, F. H. Crick, Molecular structure of nucleic acids: a structure for deoxyribonucleic acid. Nature. 171, 737–738 (1953). 88. H. Swift, The constancy of desoxyribose nucleic acid in plant nuclei. Proc. Natl. Acad. Sci. U. S. A. 36, 643–54 (1950). 89. P. N. Rao, R. T. Johnson, Mammalian Cell Fusion: Studies on the Regulation of DNA Synthesis and Mitosis. Nature. 225, 159–164 (1970). 90. L. H. Hartwell, J. Culotti, B. Reid, Genetic control of the cell-division cycle in yeast. I. Detection of mutants. Proc. Natl. Acad. Sci. U. S. A. 66, 352–9 (1970). 91. L. H. Hartwell, Genetic control of the cell division cycle in yeast. II. Genes controlling DNA replication and its initiation. J. Mol. Biol. 59, 183–194 (1971). 92. Paul Nurse, Genetic control of cell size at cell division in yeast. Nature. 256, 547–551 (1975). 93. P. Nurse, P. Thuriaux, Regulatory genes controlling mitosis in the fission yeast
109 Schizosaccharomyces pombe. Genetics. 96, 627–637 (1980). 94. D. Beach, B. Durkacz, P. Nurse, Functionally homologous cell cycle control genes in budding and fission yeast. Nature. 300, 706–709 (1982). 95. T. Evans, E. T. Rosenthal, J. Youngblom, D. Distel, T. Hunt, Annotated Classic The Discovery of Cyclins Cyclin : A Protein Specified by Maternal mRNA in Sea Urchin Eggs That Is Destroyed at Each Cleavage Division. Cell. 33, 389–396 (1983). 96. P. Russell, P. Nurse, Cdc25+ Functions As an Inducer in the Mitotic Control of Fission Yeast. Cell. 45, 145–153 (1986). 97. M. G. Lee, P. Nurse, Complementation used to clone a human homologue of the fission yeast cell cycle control gene cdc2. Nature. 327, 31–35 (1987). 98. P. Russell, P. Nurse, Negative regulation of mitosis by wee1+, a gene encoding a protein kinase homolog. Cell. 49, 559–567 (1987). 99. T. A. Weinert, L. H. Hartwell, The RAD9 gene controls the cell cycle response to DNA damage in saccharomyces cerevisiae. Science (80-. ). 241, 317–322 (1988). 100. A. B. Pardee, A restriction point for control of normal animal cell proliferation. Proc. Natl. Acad. Sci. U. S. A. 71, 1286–90 (1974). 101. A. Zetterberg, O. Larsson, K. G. Wiman, What is the restriction point? Curr. Opin. Cell Biol. 7, 835–842 (1995). 102. D. O. Morgan, Principles of CDK regulation. Nature. 374, 131–134 (1995). 103. Y. Xiong, H. Zhang, D. Beach, Subunit rearrangement of the cyclin-dependent kinases is associated with cellular transformation. Genes Dev. 7, 1572–1583 (1993). 104. Y. Xiong et al., P21 is a universal inhibitor of cyclin kinases. Nature. 366, 701–704 (1993). 105. J. W. Harper, G. R. Adami, N. Wei, K. Keyomarsi, S. J. Elledge, The p21 Cdk-Interacting Protein Cip1 is a Potent Inhibitor of G1 Cyclin-Dependent Kinases. Cell. 75, 805–816 (1993). 106. O. Aprelikova, Y. Xiong, E. T. Liu, Both p16 and p21 Families of Cyclin-dependent Kinase (CDK) Inhibitors Block the Phosphorylation of Cyclin-dependent Kinases by the CDK-activating Kinase. J. Biol. Chem. 270, 18195–18197 (1995). 107. C. J. Sherr, J. M. Roberts, Inhibitors of mammalian G1 cyclin-dependent kinases. Genes Dev. 9, 1149–1163 (1995). 108. K. Polyak et al., Cloning of p27Kip1, a cyclin-dependent kinase inhibitor and a potential mediator of extracellular antimitogenic signals. Cell. 78, 59–66 (1994). 109. J. Kernan, T. Bonacci, M. J. Emanuele, Who guards the guardian? Mechanisms that restrain APC/C during the cell cycle. Biochim. Biophys. Acta - Mol. Cell Res. 1865, 1924– 1933 (2018). 110. R. Choudhury et al., APC/C and SCFcyclin F Constitute a Reciprocal Feedback Circuit
110 Controlling S-Phase Entry. Cell Rep. 16, 3359–3372 (2016). 111. S. D. Cappell, M. Chung, A. Jaimovich, S. L. Spencer, T. Meyer, Irreversible APCCdh1 Inactivation Underlies the Point of No Return for Cell-Cycle Entry. Cell. 166, 167–180 (2016). 112. M. Pagano et al., Role of the Ubiquitin-Proteasome Pathway in Regulating Abundance of the Cyclin- Dependent Kinase Inhibitor p27 Del Sal , Vincent Chau , P . Renée Yew , Giulio F . Draetta and Mark Rolfe Published by : American Association for the Advancement of Science St. Science (80-. ). 269, 682–685 (1995). 113. A. C. Carrano, E. Eytan, A. Hershko, M. Pagano, SKP2 is required for ubiquitin-mediated degradation of the CDK inhibitor p27. Nat. Cell Biol. 1, 193–199 (1999). 114. G. Bornstein et al., Role of the SCF Skp2 Ubiquitin Ligase in the Degradation of p21 Cip1 in S Phase. J. Biol. Chem. 278, 25752–25757 (2003). 115. D. Y. Takeda, A. Dutta, DNA replication and progression through S phase. Oncogene. 24, 2827–2843 (2005). 116. J. H. Houtgraaf, J. Versmissen, W. J. van der Giessen, A concise review of DNA damage checkpoints and repair in mammalian cells. Cardiovasc. Revascularization Med. 7, 165– 172 (2006). 117. C. G. Takizawa, D. O. Morgan, Control of mitosis by changes in the subcellular location of cyclin-B1-Cdk1 and Cdc25C. Curr. Opin. Cell Biol. 12, 658–665 (2000). 118. L. L. Parker, P. J. Sylvestre, M. J. Byrnes, F. Liu, H. Piwnica-Worms, Identification of a 95-kDa WEE1-like tyrosine kinase in HeLa cells. Proc. Natl. Acad. Sci. 92, 9638–9642 (1995). 119. N. Watanabe, M. Broome, T. Hunter, Regulation of the human WEE1Hu CDK tyrosine 15-kinase during the cell cycle. EMBO J. 14, 1878–91 (1995). 120. R. N. Booher, P. S. Holman, A. Fattaey, Human Myt1 is a cell cycle-regulated kinase that inhibits Cdc2 but not Cdk2 activity. J. Biol. Chem. 272, 22300–22306 (1997). 121. F. Liu, J. J. Stanton, Z. Wu, H. Piwnica-Worms, The human Myt1 kinase preferentially phosphorylates Cdc2 on threonine 14 and localizes to the endoplasmic reticulum and Golgi complex. Mol. Cell. Biol. 17, 571–583 (1997). 122. M. J. Solomon, T. Lee, M. W. Kirschner, Role of phosphorylation in p34cdc2 activation: identification of an activating kinase. Mol. Biol. Cell. 3, 13–27 (1992). 123. A. Kumagai, W. G. Dunphy, Regulation of the cdc25 protein during the cell cycle in Xenopus extracts. Cell. 70, 139–151 (1992). 124. I. Hoffmann, P. R. Clarke, M. J. Marcote, E. Karsenti, G. Draetta, Phosphorylation and activation of human cdc25-C by cdc2--cyclin B and its involvement in the self- amplification of MPF at mitosis. EMBO J. 12, 53–63 (1993). 125. I. Nilsson, I. Hoffmann, Cell cycle regulation by the Cdc25 phosphatase family. Prog. Cell Cycle Res. 4, 107–114 (2000).
111 126. J. Li, A. N. Meyer, D. J. Donoghue, Nuclear localization of cyclin B1 mediates its biological activity and is regulated by phosphorylation. Proc. Natl. Acad. Sci. 94, 502–507 (1997). 127. F. Toyoshima-Morimoto, E. Taniguchi, N. Shinya, Labelling of purified sGC cyclin B1 and targets it to the nucleus during prophase. Nature. 410, 215–220 (2001). 128. J. D. Moore, J. Yang, R. Truant, S. Kornbluth, Nuclear import of Cdk/cyclin complexes: Identification of distinct mechanisms for import of Cdk2/cyclin E and Cdc2/cyclin B1. J. Cell Biol. 144, 213–224 (1999). 129. J. Yang et al., Control of Cyclin B1 localization through regulated binding of the nuclear export factor CRM1. Genes Dev. 12, 2131–2143 (1998). 130. C. G. Takizawa, K. Weis, D. O. Morgan, Ran-independent nuclear import of cyclin B1- Cdc2 by importin. Proc. Natl. Acad. Sci. 96, 7938–7943 (1999). 131. J. A. Ubersax, E. L. Woodbury, P. N. Quang, Targets of the cyclin-dependent kinase CDK1. Nature. 425, 859–864 (2003). 132. T. K. Fung, R. Y. C. Poon, A roller coaster ride with the mitotic cyclins. Semin. Cell Dev. Biol. 16, 335–342 (2005). 133. J. Laoukili et al., FoxM1 is required for execution of the mitotic programme and chromosome stability. Nat. Cell Biol. 7, 126–136 (2005). 134. M. Peter, J. Nakagawa, M. Doree, C. Labbe, A. Nigg, In Vitro Disassembly of the Nuclear Lamina and M Phase-Specific Phosphorylation of Lamins by cdc2 Kinase. Cell. 61, 591– 602 (1990). 135. S. Yamashiro, Y. Yamakits, H. Hosoya, F. Matsumura, Phosphorylation of non-muscle caldesmon by p34cdc2 kinase during mitosis. Nature. 349, 169–172 (1991). 136. A. Blangy et al., Phosphorylation by p34cdc2 regulates spindle association of human Eg5, a kinesin-related motor essential for bipolar spindle formation in vivo. Cell. 83, 1159– 1169 (1995). 137. K. A. Hagstrom, B. J. Meyer, Condensin and cohesin: More than chromosome compactor and glue. Nat. Rev. Genet. 4, 520–534 (2003). 138. H. A. Lane, E. A. Nigg, Cell-cycle control: POLO-like kinases join the outer circle. Trends Cell Biol. 7, 63–68 (1997). 139. A. Musacchio, E. D. Salmon, The spindle-assembly checkpoint in space and time. Nat. Rev. Mol. Cell Biol. 8, 379–393 (2007). 140. G. Fang, H. Yu, M. W. Kirschner, The checkpoint protein MAD2 and the mitotic regulator CDC20 form a ternary complex with the anaphase-promoting complex to control anaphase initiation. Genes Dev. 12, 1871–1883 (1998). 141. P. Clute, J. Pines, Temporal and spatial control of cyclin B1 destruction in metaphase. Nat. Cell Biol. 1, 82–87 (1999).
112 142. S. Hauf, I. C. Waizenegger, J. M. Peters, Cohesin cleavage by separase required for anaphase and cytokinesis in human cells. Science (80-. ). 293, 1320–1323 (2001). 143. F. Uhlmann, D. Wernic, E. V Koonin, K. Nasmyth, Cleavage of Cohesin by the CD Clan Protease Separin Triggers Anaphase in Yeast. Cell. 103, 375–386 (2000). 144. C. Ditchfield et al., Aurora B couples chromosome alignment with anaphase by targeting BubR1, Mad2, and Cenp-E to kinetochores. J. Cell Biol. 161, 267–280 (2003). 145. J. M. Peters, The anaphase promoting complex/cyclosome: A machine designed to destroy. Nat. Rev. Mol. Cell Biol. 7, 644–656 (2006). 146. J. Westendorf, P. N. Rao, L. Gerace, Cloning of cDNAs for M-Phase Phosphoprotein Recognized by the MPM2 Monoclonal Antibody and Determination of the Phosphorylated Epitope. Proc. Natl. Acad. Sci. 91, 714–718 (1994). 147. W. Korver, J. Roose, H. Clevers, The winged-helix transcription factor trident is expressed in cycling cells. Nucleic Acids Res. 25, 1715–1719 (1997). 148. K.-M. Yao, M. Sha, Z. Lu, G. G. Wong, Molecular Analysis of a Novel Winged Helix Protein, WIN. J. Biol. Chem. 272, 19827–19836 (1997). 149. H. Ye et al., Hepatocyte Nuclear Factor 3 / fork head Homolog 11 Is Expressed in Proliferating Epithelial and Mesenchymal Cells of Embryonic and Adult Tissues. 17, 1626–1641 (1997). 150. S. S. Myatt, E. W.-F. Lam, The emerging roles of forkhead box (Fox) proteins in cancer. Nat. Rev. Cancer. 7, 847–59 (2007). 151. D. R. Littler et al., Structure of the FoxM1 DNA-recognition domain bound to a promoter sequence. Nucleic Acids Res. 38, 4527–4538 (2010). 152. J. Laoukili, M. Alvarez-Fernandez, M. Stahl, R. H. Medema, FoxM1 is degraded at mitotic exit in a Cdh1-dependent manner. Cell Cycle. 7, 2720–2726 (2008). 153. H. J. Park, R. H. Costa, L. F. Lau, A. L. Tyner, P. Raychaudhuri, Anaphase-promoting complex/cyclosome-CDH1-mediated proteolysis of the forkhead box M1 transcription factor is critical for regulated entry into S phase. Mol. Cell. Biol. 28, 5162–71 (2008). 154. R. H. Costa, FoxM1 dances with mitosis. Nat. Cell Biol. 7, 108–110 (2005). 155. I.-C. Wang et al., Forkhead box M1 regulates the transcriptional network of genes essential for mitotic progression and genes encoding the SCF (Skp2-Cks1) ubiquitin ligase. Mol. Cell. Biol. 25, 10875–94 (2005). 156. G. D. Grant et al., Identification of cell cycle-regulated genes periodically expressed in U2OS cells and their regulation by FOXM1 and E2F transcription factors. Mol. Biol. Cell. 24, 3634–50 (2013). 157. H. Ye, A. X. Holterman, K. W. Yoo, R. R. Franks, R. H. Costa, Premature Expression of the Winged Helix Transcription Factor HFH-11B in Regenerating Mouse Liver Accelerates Hepatocyte Entry into S Phase. Mol. Cell. Biol. 19, 8570–8580 (1999).
113 158. M. Halasi, A. L. Gartel, Targeting FOXM1 in cancer. Biochem. Pharmacol. 85, 644–52 (2013). 159. T. V. Kalin et al., Increased levels of the FoxM1 transcription factor accelerate development and progression of prostate carcinomas in both TRAMP and LADY transgenic mice. Cancer Res. 66, 1712–1720 (2006). 160. H. J. Park et al., An N-terminal inhibitory domain modulates activity of FoxM1 during cell cycle. Oncogene. 27, 1696–1704 (2008). 161. M. L. Major, R. Lepe, R. H. Costa, Forkhead Box M1B Transcriptional Activity Requires Binding of Cdk-Cyclin Complexes for Phosphorylation-Dependent Recruitment of p300/CBP Coactivators. Mol. Cell. Biol. 24, 2649–2661 (2004). 162. L. Anders et al., A Systematic Screen for CDK4/6 Substrates Links FOXM1 Phosphorylation to Senescence Suppression in Cancer Cells. Cancer Cell. 20, 620–634 (2011). 163. I. Wierstra, J. Alves, Transcription factor FOXM1c is repressed by RB and activated by cyclin D1/Cdk4. Biol. Chem. 387, 949–962 (2006). 164. C. Lv et al., Acetylation of FOXM1 is essential for its transactivation and tumor growth stimulation. Oncotarget. 7 (2016), doi:10.18632/oncotarget.11332. 165. Y. Tan, P. Raychaudhuri, R. H. Costa, Chk2 Mediates Stabilization of the FoxM1 Transcription Factor To Stimulate Expression of DNA Repair Genes. Mol. Cell. Biol. 27, 1007–1016 (2006). 166. I. C. Wang et al., FoxM1 regulates transcription of JNK1 to promote the G1/S transition and tumor cell invasiveness. J. Biol. Chem. 283, 20770–20778 (2008). 167. M. Alvarez-Fernández, V. A. Halim, M. Aprelia, S. Mohammed, R. H. Medema, Protein phosphatase 2A (B55α) prevents premature activation of forkhead transcription factor FoxM1 by antagonizing cyclin A/cyclin-dependent kinase-mediated phosphorylation. J. Biol. Chem. 286, 33029–33036 (2011). 168. R. Y. M. Ma et al., Raf/MEK/MAPK signaling stimulates the nuclear translocation and transactivating activity of FOXM1. J. Cell Sci. 118, 795–806 (2004). 169. N. Zhang et al., FoxM1 Promotes β-Catenin Nuclear Localization and Controls Wnt Target-Gene Expression and Glioma Tumorigenesis. Cancer Cell. 20, 427–442 (2011). 170. S. Sadasivam, S. Duan, J. a DeCaprio, The MuvB complex sequentially recruits B-Myb and FoxM1 to promote mitotic gene expression. Genes Dev. 26, 474–89 (2012). 171. J. Laoukili, M. Stahl, R. H. Medema, FoxM1: at the crossroads of ageing and cancer. Biochim. Biophys. Acta. 1775, 92–102 (2007). 172. Z. Fu et al., Plk1-dependent phosphorylation of FoxM1 regulates a transcriptional programme required for mitotic progression. Nat. Cell Biol. 10, 1076–82 (2008). 173. J. Laoukili et al., Activation of FoxM1 during G2 requires cyclin A/Cdk-dependent relief of autorepression by the FoxM1 N-terminal domain. Mol. Cell. Biol. 28, 3076–87 (2008).
114 174. K. Joshi et al., MELK-Dependent FOXM1 Phosphorylation is Essential for Proliferation of Glioma Stem Cells. Stem Cells. 31, 1051–1063 (2013). 175. S. S. Myatt et al., SUMOylation inhibits FOXM1 activity and delays mitotic transition. Oncogene. 33, 1–14 (2013). 176. J. Schimmel et al., Uncovering SUMOylation Dynamics during Cell-Cycle Progression Reveals FoxM1 as a Key Mitotic SUMO Target Protein. Mol. Cell, 1053–1066 (2014). 177. J. Zhang, C. Yuan, J. Wu, Z. Elsayed, Z. Fu, Polo-like Kinase 1-mediated Phosphorylation of Forkhead Box Protein M1b Antagonizes Its SUMOylation and Facilitates Its Mitotic Function *. 290, 3708–3719 (2015). 178. C. M. Ferrer et al., O-GlcNAcylation regulates breast cancer metastasis via SIRT1 modulation of FOXM1 pathway. Oncogene. 36, 559–569 (2017). 179. C. M. Pfleger, M. W. Kirschner, The KEN box: An APC recognition signal distinct from the D box targeted by Cdh1. Genes Dev. 14, 655–665 (2000). 180. L. Q. Li, D. Pan, H. Chen, L. Zhang, W. J. Xie, F-box protein FBXL2 inhibits gastric cancer proliferation by ubiquitin-mediated degradation of forkhead box M1. FEBS Lett. 590, 445–452 (2016). 181. J. M. Jeffery et al., FBXO31 protects against genomic instability by capping FOXM1 levels at the G2/M transition. Oncogene. 36, 1012–1022 (2017). 182. M. Kongsema et al., RNF168 cooperates with RNF8 to mediate FOXM1 ubiquitination and degradation in breast cancer epirubicin treatment. Oncogenesis. 5, e252 (2016). 183. Y. Chen et al., Wnt‐induced deubiquitination FoxM1 ensures nucleus β‐catenin transactivation. EMBO J. 35, 668–684 (2016). 184. U. Karunarathna et al., OTUB1 inhibits the ubiquitination and degradation of FOXM1 in breast cancer and epirubicin resistance. Oncogene. 35, 1433–1444 (2016). 185. Y. Wang et al., OTUB1-catalyzed deubiquitination of FOXM1 facilitates tumor progression and predicts a poor prognosis in ovarian cancer. Oncotarget. 7, 36681–36697 (2016). 186. U. G. Bhat, M. Halasi, A. L. Gartel, FoxM1 is a general target for proteasome inhibitors. PLoS One. 4, e6593 (2009). 187. S. K. Radhakrishnan et al., Identification of a Chemical Inhibitor of the Oncogenic Transcription Factor Forkhead Box M1. Cancer Res. 66, 9731–9735 (2006). 188. N. S. Hegde, D. A. Sanders, R. Rodriguez, S. Balasubramanian, The transcription factor FOXM1 is a cellular target of the natural product thiostrepton. Nat. Chem. 3, 725–731 (2011). 189. G. Bin Liao et al., Regulation of the master regulator FOXM1 in cancer. Cell Commun. Signal. 16, 1–15 (2018). 190. Cancer Genome Atlas Network, Comprehensive molecular portraits of human breast
115 tumours. Nature. 490, 61–70 (2012). 191. I. M. Kim et al., The Forkhead Box m1 transcription factor stimulates the proliferation of tumor cells during development of lung cancer. Cancer Res. 66, 2153–2161 (2006). 192. T. V. Kalin et al., Increased levels of the FoxM1 transcription factor accelerate development and progression of prostate carcinomas in both TRAMP and LADY transgenic mice. Cancer Res. 66, 1712–1720 (2006). 193. O. A. Kalinina et al., Sustained hepatic expression of FoxM1B in transgenic mice has minimal effects on hepatocellular carcinoma development but increases cell proliferation rates in preneoplastic and early neoplastic lesions. Oncogene. 22, 6266–6276 (2003). 194. Y. Yoshida, I. C. Wang, H. M. Yoder, N. O. Davidson, R. H. Costa, The Forkhead Box M1 Transcription Factor Contributes to the Development and Growth of Mouse Colorectal Cancer. Gastroenterology. 132, 1420–1431 (2007). 195. J. R. Carr, H. J. Park, Z. Wang, M. M. Kiefer, P. Raychaudhuri, FoxM1 mediates resistance to herceptin and paclitaxel. Cancer Res. 70, 5054–5063 (2010). 196. J. M.-M. Kwok et al., FOXM1 Confers Acquired Cisplatin Resistance in Breast Cancer Cells. Mol. Cancer Res. 8, 24–34 (2010). 197. J. Millour et al., ATM and p53 Regulate FOXM1 Expression via E2F in Breast Cancer Epirubicin Treatment and Resistance. Mol. Cancer Ther. 10, 1046–1058 (2011). 198. I. Wierstra, FOXM1 (Forkhead box M1) in Tumorigenesis: Overexpression in Human Cancer, Implication in Tumorgenesis, Oncogenic Functions, Tumor-Suppressive Properties and Target of Anticancer Therapy (Elsevier Inc., ed. 1, 2013; http://dx.doi.org/10.1016/B978-0-12-407190-2.00016-2), vol. 119. 199. M. V. Gormally et al., Suppression of the FOXM1 transcriptional programme via novel small molecule inhibition. Nat. Commun. 5, 1–11 (2014). 200. Z. Bi et al., A novel peptide, 9R-P201, strongly inhibits the viability, proliferation and migration of liver cancer HepG2 cells and induces apoptosis by down-regulation of FoxM1 expression. Eur. J. Pharmacol. 796, 175–189 (2017). 201. Q. Xiang et al., Suppression of FOXM1 Transcriptional Activities via a Single-Stranded DNA Aptamer Generated by SELEX. Sci. Rep. 7, 1–12 (2017). 202. M. Halasi et al., Honokiol is a FOXM1 antagonist article. Cell Death Dis. 9, 1–8 (2018). 203. S. Sadasivam, S. Duan, J. A. DeCaprio, The MuvB complex sequentially recruits B-Myb and FoxM1 to promote mitotic gene expression. Genes Dev. 26, 474–489 (2012). 204. D. A. Sanders et al., FOXM1 binds directly to non-consensus sequences in the human genome. Genome Biol. 16, 1–23 (2015). 205. V. Petrovic, R. H. Costa, L. F. Lau, P. Raychaudhuri, A. L. Tyner, Negative regulation of the oncogenic transcription factor FoxM1 by thiazolidinediones and mithramycin. Cancer Biol. Ther. 9, 1008–1016 (2010).
116 206. G. zhi Dong et al., Diarylheptanoids suppress proliferation of pancreatic cancer PANC-1 cells through modulating shh-Gli-FoxM1 pathway. Arch. Pharm. Res. 40, 509–517 (2017). 207. L. Sun et al., The FOXM1 inhibitor RCM-1 suppresses goblet cell metaplasia and prevents IL-13 and STAT6 signaling in allergen-exposed mice. Sci. Signal. 10, 1–10 (2017). 208. C. D. Allis, T. Jenuwein, The molecular hallmarks of epigenetic control. Nat. Rev. Genet. 17, 487–500 (2016). 209. H. Ye et al., Hepatocyte nuclear factor 3/fork head homolog 11 is expressed in proliferating epithelial and mesenchymal cells of embryonic and adult tissues. Mol. Cell. Biol. 17, 1626–41 (1997). 210. L. Bella, S. Zona, G. N. De Moraes, E. W.-F. Lam, FOXM1: A key oncofoetal transcription factor in health and disease. Semin. Cancer Biol., 1–8 (2014). 211. J. Laoukili et al., FoxM1 is required for execution of the mitotic programme and chromosome stability. Nat. Cell Biol. 7, 126–36 (2005). 212. U. Schuller et al., Forkhead Transcription Factor FoxM1 Regulates Mitotic Entry and Prevents Spindle Defects in Cerebellar Granule Neuron Precursors. Mol. Cell. Biol. 27, 8259–8270 (2007). 213. K. Krupczak-Hollis et al., The mouse Forkhead Box m1 transcription factor is essential for hepatoblast mitosis and development of intrahepatic bile ducts and vessels during liver morphogenesis. Dev. Biol. 276, 74–88 (2004). 214. Cancer Genome Atlas Research Network, Integrated genomic analyses of ovarian carcinoma. Nature. 474, 609–15 (2011). 215. T. Cancer, G. Atlas, Integrated genomic characterization of endometrial carcinoma. Nature. 497, 67–73 (2013). 216. G. D. Grant et al., Live-cell monitoring of periodic gene expression in synchronous human cells identifies Forkhead genes involved in cell cycle control. Mol. Biol. Cell. 23, 3079–3093 (2012). 217. X. Chen et al., The Forkhead Transcription Factor FOXM1 Controls Cell Cycle- Dependent Gene Expression through an Atypical Chromatin Binding Mechanism. Mol. Cell. Biol. 33, 227–236 (2013). 218. F. Geng, S. Wenzel, W. P. Tansey, Ubiquitin and Proteasomes in Transcription. Annu. Rev. Biochem. 81, 177–201 (2012). 219. S. Y. Kim, A. Herbst, K. A. Tworkowski, S. E. Salghetti, W. P. Tansey, Skp2 regulates Myc protein stability and activity. Mol. Cell. 11, 1177–1188 (2003). 220. N. Von Der Lehr et al., The F-box protein Skp2 participates in c-Myc proteosomal degradation and acts as a cofactor for c-Myc-regulated transcription. Mol. Cell. 11, 1189– 1200 (2003).
117 221. L. A. Jaenicke et al., Ubiquitin-Dependent Turnover of MYC Antagonizes MYC/PAF1C Complex Accumulation to Drive Transcriptional Elongation. Mol. Cell. 61, 54–67 (2016). 222. S. E. Salghetti, A. A. Caudy, J. G. Chenoweth, W. P. Tansey, Regulation of transcriptional activation domain function by ubiquitin. Science (80-. ). 293, 1651–1653 (2001). 223. G. Reid et al., Cyclic, Proteasome-Mediated Turnover of Unliganded and Liganded ER alpha on Responsive Promoters Is an Integral Feature of Estrogen Signaling. Mol. Cell. 11, 695–707 (2003). 224. D. M. Lonard, Z. Nawaz, C. L. Smith, B. W. O’Malley, The 26S proteasome is required for estrogen receptor-α and coactivator turnover and for efficient estrogen receptor-α transactivation. Mol. Cell. 5, 939–948 (2000). 225. R. C. Wu, Q. Feng, D. M. Lonard, B. W. O’Malley, SRC-3 Coactivator Functional Lifetime Is Regulated by a Phospho-Dependent Ubiquitin Time Clock. Cell. 129, 1125– 1140 (2007). 226. S. Jackson, Y. Xiong, CRL4s: the CUL4-RING E3 ubiquitin ligases. Trends Biochem. Sci. 34, 562–570 (2009). 227. J. Guerrero-Santoro et al., The Cullin 4B-Based UV-Damaged DNA-Binding Protein Ligase Binds to UV-Damaged Chromatin and Ubiquitinates Histone H2A. Cancer Res. 68, 5014–5022 (2008). 228. J. Jin, E. E. Arias, J. Chen, J. W. Harper, J. C. Walter, A Family of Diverse Cul4-Ddb1- Interacting Proteins Includes Cdt2, which Is Required for S Phase Destruction of the Replication Factor Cdt1. Mol. Cell. 23, 709–721 (2006). 229. S. Angers et al., Molecular architecture and assembly of the DDB1-CUL4A ubiquitin ligase machinery. Nature. 443, 590–593 (2006). 230. C. M. McCall et al., Human Immunodeficiency Virus Type 1 Vpr-Binding Protein VprBP, a WD40 Protein Associated with the DDB1-CUL4 E3 Ubiquitin Ligase, Is Essential for DNA Replication and Embryonic Development. Mol. Cell. Biol. 28, 5621– 5633 (2008). 231. T. Nakagawa et al., CRL4VprBP E3 Ligase Promotes Monoubiquitylation and Chromatin Binding of TET Dioxygenases. Mol. Cell. 57, 247–260 (2015). 232. M. Kaur, M. M. Khan, A. Kar, A. Sharma, S. Saxena, CRL4-DDB1-VPRBP ubiquitin ligase mediates the stress triggered proteolysis of Mcm10. Nucleic Acids Res. 40, 7332– 7346 (2012). 233. W. Li et al., Merlin/NF2 Suppresses Tumorigenesis by Inhibiting the E3 Ubiquitin Ligase CRL4DCAF1 in the Nucleus. Cell. 140, 477–490 (2010). 234. H. S. Yen, Global Protein Stability Profiling in. 918, 918–924 (2010). 235. T. A. Soucy et al., An inhibitor of NEDD8-activating enzyme as a new approach to treat cancer. Nature. 458, 732–736 (2009). 236. C. G. Havens, J. C. Walter, Docking of a specialized PIP Box onto chromatin-bound
118 PCNA creates a degron for the ubiquitin ligase CRL4Cdt2. Mol. Cell. 35, 93–104 (2009). 237. J. Laoukili, M. Alvarez-Fernandez, M. Stahl, R. H. Medema, FoxM1 is degraded at mitotic exit in a Cdh1-dependent manner. Cell Cycle. 7, 2720–2726 (2008). 238. F. Kruiswijk et al., Coupled Activation and Degradation of eEF2K Regulates Protein Synthesis in Response to Genotoxic Stress. Sci. Signal. 5, ra40–ra40 (2012). 239. A. L. Goldberg, A. C. S. John, Intracellular Protein Degradation in Mammalian and Bacterial Cells: Part 2. Annu. Rev. Biochem. 45, 747–804 (1976). 240. M. K. M. Tan, H. J. Lim, E. J. Bennett, Y. Shi, J. W. Harper, Parallel SCF Adaptor Capture Proteomics Reveals a Role for SCFFBXL17 in NRF2 Activation via BACH1 Repressor Turnover. Mol. Cell. 52, 9–24 (2013). 241. T. Y. Kim et al., Substrate Trapping Proteomics Reveals Targets of the βTrCP2/FBXW11 Ubiquitin Ligase. Mol. Cell. Biol. 35, 167–181 (2014). 242. Z. Nawaz, C. L. Smith, D. M. Lonard, A. P. Dennis, B. W. O’Malley, Proteasome- dependent degradation of the human estrogen receptor. Proc. Natl. Acad. Sci. 96, 1858– 1862 (2002). 243. A. Varshavsky, The early history of the ubiquitin field. Protein Sci. 15, 647–654 (2006). 244. M. Hochstrasser, A. Varshavsky, In vivo degradation of a transcriptional regulator: The yeast α2 repressor. Cell. 61, 697–708 (1990). 245. E. J. Bennett, J. Rush, S. P. Gygi, J. W. Harper, Dynamics of cullin-RING ubiquitin ligase network revealed by systematic quantitative proteomics. Cell. 143, 951–965 (2010). 246. A. Stolz et al., The CHK2-BRCA1 tumour suppressor pathway ensures chromosomal stability in human somatic cells. Nat. Cell Biol. 12, 492–499 (2010). 247. T. Bonacci et al., Identification of new mechanisms of cellular response to chemotherapy by tracking changes in post-translational modifications by ubiquitin and ubiquitin-like proteins. J. Proteome Res. 13, 2478–2494 (2014). 248. M. L. Golson, K. H. Kaestner, Fox transcription factors: from development to disease. Development. 143, 4558–4570 (2017). 249. S. Hannenhalli, K. H. Kaestner, The evolution of Fox genes and their role in development and disease. Nat. Rev. Genet. 10, 233–240 (2009). 250. K. a. Hoadley et al., Multiplatform Analysis of 12 Cancer Types Reveals Molecular Classification within and across Tissues of Origin. Cell. 158, 929–944 (2014). 251. C. M. Perou et al., Molecular portraits of human breast tumours. Nature. 406, 747–52 (2000). 252. T. Sorlie et al., Repeated observation of breast tumor subtypes in independent gene expression data sets. Proc. Natl. Acad. Sci. 100, 8418–8423 (2003). 253. J. Zhou et al., FOXM1 modulates cisplatin sensitivity by regulating EXO1 in ovarian cancer. PLoS One. 9, e96989 (2014).
119 254. M. Halasi, A. L. Gartel, FOX(M1) news--it is cancer. Mol. Cancer Ther. 12, 245–54 (2013). 255. X. Wang et al., VprBP/DCAF1 Regulates the Degradation and Nonproteolytic Activation of the Cell Cycle Transcription Factor FoxM1. Mol. Cell. Biol. 37, 1–15 (2017). 256. Y. Chen et al., Wnt‐induced deubiquitination FoxM1 ensures nucleus β‐catenin transactivation. EMBO J. 35, 668–684 (2016). 257. L. Kategaya et al., USP7 small-molecule inhibitors interfere with ubiquitin binding (Nature Publishing Group, 2017; http://dx.doi.org/10.1038/nature24006), vol. 550. 258. A. P. Turnbull et al., Molecular basis of USP7 inhibition by selective small-molecule inhibitors. Nature. 550, 481–486 (2017). 259. T. Furuyama, T. Nakazawa, I. Nakano, N. Mori, Identification of the differential distribution patterns of mRNAs and consensus binding sequences for mouse DAF-16 homologues. Biochem. J. 349, 629–34 (2000). 260. K. L. Meerbrey et al., The pINDUCER lentiviral toolkit for inducible RNA interference in vitro and in vivo. Proc. Natl. Acad. Sci. 108, 3665–3670 (2011). 261. I.-M. Kim et al., The Forkhead Box m1 Transcription Factor Stimulates the Proliferation of Tumor Cells during Development of Lung Cancer. Cancer Res. 66, 2153–2161 (2006). 262. C. E. M. Liu, B. Dai, S.H. Kang, K. Ban, F.J. Huang, F.F. Lang, K.D. Aldape, T.X. Xie, S. H. Pelloski, K. Xie, R. Sawaya, FoxM1B is overexpressed in human glioblastomas and critically regulates the tumorigenicity of glioma cells. Cancer Res. 66, 3593–3602 (2006). 263. V. J. Weigman et al., Basal-like Breast cancer DNA copy number losses identify genes involved in genomic instability, response to therapy, and patient survival. Breast Cancer Res. Treat. 133, 865–880 (2012). 264. M. Fischer, P. Grossmann, M. Padi, J. A. DeCaprio, Integration of TP53, DREAM, MMB-FOXM1 and RB-E2F target gene analyses identifies cell cycle gene regulatory networks. Nucleic Acids Res. 44, 6070–86 (2016). 265. H. Yang et al., Down-regulation of asparagine synthetase induces cell cycle arrest and inhibits cell proliferation of breast cancer. Chem. Biol. Drug Des. 84, 578–584 (2014). 266. F. Zhao et al., Overexpression of forkhead box protein M1 (FOXM1) in ovarian cancer correlates with poor patient survival and contributes to paclitaxel resistance. PLoS One. 9, e113478 (2014). 267. J. Adélaïde et al., Integrated profiling of basal and luminal breast cancers. Cancer Res. 67, 11565–11575 (2007). 268. J. E. Bradner, D. Hnisz, R. A. Young, Transcriptional Addiction in Cancer. Cell. 168, 629–643 (2017). 269. D. Shi, S. R. Grossman, Ubiquitin becomes ubiquitous in cancer. Cancer Biol. Ther. 10, 737–747 (2010).
120 270. T. D. Schmittgen, K. J. Livak, Analyzing real-time PCR data by the comparative CT method. Nat. Protoc. 3, 1101–1108 (2008). 271. T. O. Nielsen et al., A comparison of PAM50 intrinsic subtyping with immunohistochemistry and clinical prognostic factors in tamoxifen-treated estrogen receptor-positive breast cancer. Clin. Cancer Res. 16, 5222–5232 (2010). 272. V. K. Mootha et al., PGC-1 alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes PGC-1 α -responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat. Genet. 34, 267–273 (2003). 273. A. Subramanian et al., Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. 102, 15545–15550 (2005). 274. E. Cerami et al., The cBio Cancer Genomics Portal: An open platform for exploring multidimensional cancer genomics data. Cancer Discov. 2, 401–404 (2012). 275. J. Gao et al., Integrative Analysis of Complex Cancer Genomics and Clinical Profiles Using the cBioPortal. Sci. Signal. 6, p11 (2013). 276. K. Kim et al., VprBP Has Intrinsic Kinase Activity Targeting Histone H2A and Represses Gene Transcription. Mol. Cell. 52, 247–253 (2013). 277. H. Aihara et al., Histone H2A T120 Phosphorylation Promotes Oncogenic Transformation via Upregulation of Cyclin D1. Mol. Cell. 64, 176–188 (2016). 278. B. Madan et al., USP6 oncogene promotes Wnt signaling by deubiquitylating Frizzleds. Proc. Natl. Acad. Sci. 113, E2945–E2954 (2016). 279. J. Liu et al., An Integrated TCGA Pan-Cancer Clinical Data Resource to Drive High- Quality Survival Outcome Analytics. Cell. 173, 400-416.e11 (2018). 280. W. Kim et al., Systematic and quantitative assessment of the ubiquitin modified proteome. Mol. Cell. 44, 325–340 (2011).
121