The Role of DIS3L2 in the Degradation of the Uridylated RNA Species in Humans

MASARYK UNIVERSITY Faculty of Science

National Center for Biomedical Research

Dmytro Ustianenko

The role of DIS3L2 in the degradation of the uridylated RNA species in humans

Ph.D. THESIS

SUPERVISOR doc. Mgr. ŠTĚPÁNKA VAŇÁČOVÁ, Ph.D.

BRNO, 2014

______

Cover: A schematic representation of the RNA degradation process which occurs in the cytoplasm.

______

Bibliographic entry

Author: Mgr. Dmytro Ustianenko

Faculty of Science, Masaryk University National Centre for Biomolecular Research Central European Institute of Technology

Title of Dissertation: The role of DIS3L2 in the degradation of the uridylated RNA species in humans.

Degree Programme: Biochemistry

Field of Study: Biomolecular chemistry

Supervisor: Doc. Mgr. Štěpánka Vaňáčová, Ph.D.

Academic Year: 2014

Number of Pages: 131

Keywords: RNA degradation, DIS3L2, uridylation, humans, miRNA, let-7

Bibliografický záznam

Autor: Mgr. Dmytro Ustianenko

Přírodovědecká fakulta, Masarykova univerzita Národní centrum pro výzkum biomolekul Středoevropský technologický institut

Název práce: The role of DIS3L2 in the degradation of the uridylated RNA species in humans.

Studijní program: Biochemie

Studijní obor: Biomolekulární chemie

Školitel: doc. Mgr. Štěpánka Vaňáčová, Ph.D.

Akademický rok: 2014

Počet stran: 131

Klíčová slova: RNA degradace, DIS3L2, uridylace, mikro RNA, let-7

Acknowledgements

I would like to thank to all the people who have contributed to this work. Foremost I would like to thank my supervisor Stepanka Vanacova for her enthusiasm, support and patience though all of my studies. I would like to acknowledge the members of the RNA processing and degradation group for their support and help in achieving this goal.

Besides, I would like to thank to all those people who were standing next by me for all this long years. Those who were supporting me, loved me and inspired me. Those who have not gave up on me, those who gave me energy and courage to go on. Without you this work would be impossible to accomplish.

To those who died defendending Ukrainian independence in winter 2014

Подяка

Я хочу подякувати всім хто сприяв цій роботі. Перш за все я хочу подякувати моєму керівнику Степанце Ванячовій за її ентузіазм, підтримку та терпіння протягом мого навчання. Також хочу подякувати всім членам моєї лабораторії за їх підтримку в досягненні мети.

Особливо хочу відзначити всіх хто був поруч зі мною всі ці довгі роки. Тим хто підтримував мене, кохав та надихав. Тим хто не зрадив, тим хто давав натхнення та сили йти далі. Без вас це було б не можливо.

Тим хто загинув захищаючи незалежність України зимою 2014 року

В целях природы обуздания, В целях рассеять неученья Тьму Берём картину мироздания – да! И тупо смотрим, что к чему…

Братья Стругацкие, “Понедельник начинается в субботу”

Abstract

The process of the RNA decay is an essential mechanism required for the function and maintenance of the cell homeostasis. It is closely connected with the RNA processing, metabolism and quality control. Here I characterized both in vivo and in vitro the Perlman syndrome associated human 3' to 5' exonuclease DIS3L2. I report that unlike its other human homologs, DIS3L2 is acting independently of the core exosome complex in the cytoplasm of human cells. I show that DIS3L2 is involved in targeting uridylated species of the pre-miRNA from the let-7 family which play a critical role during the cell differentiation. Besides this, I report on a novel repertoire of the RNA substrates which are targeted by DIS3L2. I show that DIS3L2 mediated RNA degradation is triggered by the 3' end uridylation of the RNA in vivo. We show that various classes of the RNA can be uridylated and recognized by DIS3L2. These include mRNAs, snRNAs, miRNA’s and previously unreported uridylated tRNA’s. Suggesting, that exonuclease DIS3L2 is an essential component of the RNA turnover and potential processing.

Abstrakt

Process degradace RNA představuje mechanismus nezbytný pro funk- ci a udržení vnitřní rovnováhy buněk. Tento proces je spojen s procesováním, metabolismem a kontrolou kvality RNA. V této části práce charakterizujeme in vivo a in vitro lidskou 3' - 5' exonukleáza u DIS3L2, která je spojena s tzv. Perlmanovým syndromem. Narozdíl od ostatních lidských homologních proteinů, DIS3L2 funguje nezávisle na jádru exosomu v cytoplazmě lidských buněk. Ukazujeme zde také, že DIS3L2 je zahrnuta v rozpoznávání uridylo- vaných typů pre-mRNA z rodiny let-7, která hraje důležitou roli během difer- enciace lidských buněk. Kromě toho bylo identifikováno nové spektrum RNA molekul, které reprezentují cílovou skupinu proteinu DIS3L2. Ukazujeme, že zcela nová skupina molekul RNA může být rozpoznána a uridylována pro- teinem DIS3L2. Tato skupina zahrnuje mRNA, snRNA miRNA a uridylované molekuly tRNA, které nebyly zatím publikovány. Výsledky ukazují, ze exo- nukleáza DIS3L2 představuje nezbytný prvek v metabolismu a procesování RNA.

Table of contents Bibliographic entry...... 7 Bibliografický záznam...... 9 Acknowledgements...... 11 Подяка...... 11 Abstract...... 15 Abstrakt...... 17 Table of contents...... 19 Introduction...... 23 Nuclear RNA quality control machinery...... 23 Noncanonical polymerases in human...... 25 Mechanisms of the UTP specificity of terminal uridyl transferases...... 28 Uridylation of mammalian RNAs...... 30 The role of the uridylation in histone mRNA degradation...... 31 The role of uridylation in miRNA processing and degradation.... 32 Uridylation of the miRNA...... 35 RNA degradation at the 3' ends...... 38 Exosome complex in eukaryotes...... 38 DIS3 homologs in humans...... 40 Human DIS3 Like exonuclease 2...... 41 Other known 3' to 5' exonucleases...... 42 Genome-wide approaches to study protein-RNA interactions.43 References...... 46 Mammalian DIS3L2 exoribonuclease targets the uridylated precursors of let-7 miRNA’s...... 64 Supplementary data...... 74 DIS3L2 is involved in the degradation of the uridylated tRNA fragments...... 84 Abstract...... 85 Introduction...... 86 Results...... 89 Identification of DIS3L2 RNA targets by PAR-CLIP...... 89 DIS3L2 predominantly crosslinks to tRNA 5' ends...... 92 DIS3L2 and 5' tRFs associate with active polysomes...... 95 TUTase 4 uridylates 5’ tRF’s...... 99 Discussion...... 101 Material and Methods...... 105 Summary...... 124 Curriculum vitae...... 126 List of conferences...... 129 List of publications...... 130

Dmytro Ustianenko Introduction

Introduction

Nuclear RNA quality control machinery

The processes of the RNA degradation, turnover and quality control are absolutely essential for the survival and homeostasis of the eukaryotic cell. Each step of the RNA metabolism is tightly controlled by the quality control machinery to ensure the immediate recognition and removal of the aberrant molecule from the cell. This process is crucial as the presence of misfunc- tional RNAs could lead to the severe disbalance of the cell or even whole organism. Eukaryotes have evolved several mechanisms to monitor RNA processing, maturation as well as proper functioning of the RNA though so called quality control mechanisms. This occurs in both nucleus and cytoplasm of the cell. In the cytoplasm, nonsense-mediated mRNA decay is targeting the mRNA’s that possesses the prematurely located stop codon. It is important for cells to eliminate mRNAs that prematurely terminate translation since the resulting truncated proteins have the potential to be nonfunctional or acquire dominant-negative or gain-of-function activities (Pulak & Anderson, 1993). Premature termination codons can be a result of the random mutation, ineffi- cient splicing (He et al, 1993), using non-frame start codons (Welch & Jacob- son, 1999) etc. Interestingly, that mRNA’s with normal stop codon positioning by extended 3’ UTR due to the mutation in the polyadenylation site are also targets of the NMD (Muhlrad & Parker, 1999). Another cytoplasmic mechanism, a nonstop mRNA decay (NSD) is targeting the mRNA which do not possess a stop codon due to the premature polyadenylation of the messenger (Graber et al, 1999; Sparks & Dieckmann, 1998) or transcription abortion (Cui & Denis, 2003). During this process the ribosome that reaches the very 3′ terminus of the message is recruiting Ski7 protein that can be incorporated into the ribosome A-site (Frischmeyer et al, 2002). Ski7 in return is recruiting the exosome, 3' - 5' degradation complex (van Hoof et al, 2002). NSD can occur also from 5' - 3' end in the absence of the Ski7 component (Inada & Aiba, 2005). The least studied process of the

23 Introduction Dmytro Ustianenko mRNA decay in the cytoplasm is the no-go decay which was reported only in yeasts by using artificial systems. It involves an endonucleolytic cleavage of the mRNA next to the ribosome that is stalled on the complex secondary RNA structure (Doma & Parker, 2006). Nuclear RNA quality control pathway was originally discovered in budding yeast, Saccharomyces cerevisiae. The key component of this pathway is the TRAMP complex which consist of three factors: poly(A) polymerase Trf4p/Trf5p, RNA helicase Mtr4p and zinc knuckle RNA binding protein Air1p/ Air2p (Kadaba et al, 2004; LaCava et al, 2005; Vanacova et al, 2005). The TRAMP complex selectively recognizes aberrant, misfolded or incorrectly processed RNAs and modifies them with a stretch of adenines at their 3’ ends. The oligo(A) tail then serves as a mark for the RNA degradation complex, the exosome which targets such RNAs thus eliminates them from the cellular environment. The mechanism of TRAMP-exosome coordinated action is depicted in Figure 1.

Target recognition Oligoadenylation by Trf4p Exosome recruitment and activation Mtr4p Mtr4p Trf4p Air2p Air2p Air2p Trf4p Trf4p AA Mtr4p A A AA AA AA AA

Deadenylation RNP substrate by exosome Oligoadenylation of intermetiates? Trf4p Trf4p Air2p Air2p Mtr4p Mtr4p

Complete Further degradation degradation of the RNA

Figure 1. The proposed mechanism of TRAMP-exosome coordinated action. TRAMP is recognizing and labeling aberrant RNA on the 3' end with the stretch of ad- enins. This serves as a signal for the RNA degradation machinery (nuclear exosome) for further elimination of the transcript. (modified from La Cava et al, 2005)

Trf4p belongs to the superfamily of RNA-specific nucleotidyl transferases, a diverse category of template-independent polymerases that add

24 Dmytro Ustianenko Introduction

ribonucleotides to the 3′-ends of RNA molecules. Due to its similarities in the catalytic site architecture Trf4 belongs to the family of DNA polymerase β family of enzymes. The canonical family of the RNA nucleotidyl transferases contains three major nuclear polymerases (polymerase α, β, and γ) (Lingner et al, 1991; Raabe et al, 1991) that are involved in the RNA polyadenylation. Due to the late discovery and their functional differences from the canonical polyA polymerases (PAPs) proteins such as Trf4, and GLD-2 were designated as non-canonical polyA polymerases. (Read et al, 2002; Saitoh et al, 2002; Wang et al, 2002b). The activity of Trf4p is essential for the survival of the cell as a double deletion of the Trf4 and Trf5 is lethal (Castano et al, 1996). Single deletion of Trf4 gene causes alteration in the expression of almost all types of RNA, leading to severe defects in telomere length shorten- ing and retrotransposon upregulation (San Paolo et al, 2009).

Noncanonical polymerases in human

Noncanonical PAPs are present in all eukaryotes. The best characterized are those from budding and fission yeast. The first member of the noncanonical PAP family that was identified is the Cid1 protein in yeast Schizo- saccharomyces pombe (Wang et al, 2000) and proposed to be a putative nucleotidyl transferase. It was shown that the overexpression of Cid1 can cre- ate a certain level of resistance to hydroxyurea and caffeine that are known suppressors of DNA replication and S to M phase checkpoint. Further works have established a Cid1 as a novel cytoplasmic poly(A) polymerase that is together with its homolog Cid13 are involved in the poly(A) tail extension of certain mRNA transcripts. Proposing, that in this way they can mediate the cellular response to the S phase arrest (Read et al, 2002). The Cid14 homolog of Cid1 in S. pombe was implicated in the nuclear adenylation of the rRNA (Win et al, 2006). To date, there are six reported Cid-like proteins in fission yeasts (Wang et al, 2000) that are involved in various processes of replication check point, heterochromatin formation, quality control etc. (reviewed in (Stevenson & Norbury, 2006). In contrast, mammalian homologs have been characterized only to a certain extent. Human and mouse genome encodes at least seven nonca-

25 Introduction Dmytro Ustianenko

TUTase1 (mtPAP/ PAPD1/ Hs4) 582 a.a TUTase2 (GLD2/PAPD4) 484 a.a

TUTase3 (PAPD5/ TRF-2) 489 a.a

TUTase 4 (ZCCHC11/ PAPD3/ Hs3) 1640 a.a

TUTase5 (POLS/ TRF4-1) 542 a.a

TUTase6 (TUT1/ PAPD2/ Hs5) 874 a.a

TUTase7 (ZCCHC6/ PAPD6/ Hs2) 1495 a.a

Figure 2. Domain organization of the human non canonical polymerases. Domains are indicated as followed: red, nucleotidyl transferase domain; orange, PAP-associated domain; hatched red, inactive nucleotidyl transferase domain; blue, C2H2 zinc finger domain; green, CCHC zinc finger domain; yellow, RNA recognition motif ( adopted from Heo et al, 2012). nonical PAPs (Figure 2) which possess either adenylation and/or uridylation activities (Rammelt et al, 2011). All these proteins share similar domain architecture including the nucleotidyl transferase domain (NTD) and poly(A) polymerase (PAP)-associated domain (Figure 2) (Heo et al, 2012; Scott & Norbury, 2013).PolS/PAPD7 and PAPD5 (TRF4-2, TUT3) are the human homologs of the yeast Trf4p. Similarly to Trf4p and Trf5p, they have been shown to assemble into a TRAMP-like complex (Fasken et al, 2011; Lubas et al, 2011) which localizes in the nucleus (and nucleolus) and targets pre-rRNA (Lubas et al, 2011; Rammelt et al, 2011; Shcherbik et al, 2010). Interest- ingly, unlike yTrf4p, PAPD5 does not require any additional cofactors for the RNA binding and nucleotidyl transferase activity in vitro (Rammelt et al, 2011). PAPD5 was shown to be involved in the processing of snoRNA (Berndt et al, 2012) and in adenylation of aberrant ribosomal RNA (Shcherbik et al, 2010). POLS/TRF4-1 is poorly studied and to date there are no reports available about the function and role of this protein in the cell. The function of human TUTase 2 (GLD2, PAPD4) is the readenylation of mRNA in the cytoplasm. Process of readenylation depends on several trans acting factors (e.g. CPEB1 protein (Hake et al, 1998; Richter, 2007)) that recognizes a specific signal (cytoplasmic polyadenylation element, CPE) with a conserved sequence (UUUUA/UAU) located on the mRNA 3' UTR (Ver- rotti et al, 1996). This primarily serves for the stabilization of the mRNA. As an example, TUTase 2 is recruited to the p53 mRNA and promotes its cy-

26 Dmytro Ustianenko Introduction

toplasmic polyadenylation, although there is a certain controversy between reports (Burns et al, 2011; Glahder & Norrild, 2011). It is also able to affect the miRNA-122 by stabilizing it through a 3' adenylation. Interestingly, TUT- ase 2 recruiting factor CPEB1 is the primary target of miR-122 (D'Ambrogio et al, 2012; Katoh et al, 2009). This indicates the presence of regulatory loop, between miR-122 that targets mRNA of the CPEB (Burns et al, 2011) and affects the TUTase 2 mediated adenylation of p53 mRNA influencing the cellular survival. Genome-wide studies identified more than hundreds of mRNA in the neuronal dendrites that can be targets of the TUTase 2 and undergo a cytoplasmic readenylation (Udagawa et al, 2012) that has a pronounced effect on the synaptic plasticity. Besides this fact, it is essential for the mouse oocyte development, its expression levels are strictly controlled and it can function in both nucleus and the cytoplasm by targeting mouse latent mRNA’s (Nakanishi et al, 2006). Similar TUT2 function is observed in Caenorhabditis elegans, where it is associated with P granules and is implicated in readenylation of the certain oocyte specific mRNA transcripts leading to their translational enhancement (Wang et al, 2002a). The knock down of TUTase 2 in flies causes the male sterility by influencing spermatid formation (Sartain et al, 2011) but more importantly, it affects the long term memory formation by readenylation of certain transcripts in the dendrites of neurons (Kwak et al, 2008). PAPD1 mitochondrial poly(A) polymerase (mtPAP, TUT1, PAPD1), contains a mitochondria targeting sequence (Nagaike et al, 2005; Tomecki et al, 2004). This enzyme is able to act without the help of cofactors but requires formation of a homodimer for its proper function. (Bai et al, 2011). PAPD1 is responsible for the polyadenylation of the mitochondrial messenger RNAs (mt-mRNA) which are excised from the transcribed polycistronic RNA’s (Mer- cer et al, 2011). Such a polyadenylation serves for the completion of stop codon of mt-mRNA as it often ends only with U or UA instead of full stop codon UAA (Ojala et al, 1981). The rest of it roles remain unclear. Like in fission yeast, several noncanonical PAPs in mammals are terminal uridyl transferases (TUTs). U6 TUTase (TUTase 6, TUT1, PAPD2, Hs5) is the first enzyme from this family that was discovered in humans. It was isolated from HELA cell extract and it’s shown to be responsible for the specific 3’ end uridylation of U6 sn-

27 Introduction Dmytro Ustianenko

RNA in human cells (Trippe et al, 2003; Trippe et al, 1998). This modification is important for U6 snRNA recycling, which facilitates the formation of the U4- U6 splicing complex (Trippe et al, 2006; Vankan et al, 1992). One of the best studied human noncanonical poly(U) polymerases are TUTase 4 (ZCCHC11, PAPD3, Hs3) and TUTase7 (ZCCHC6, PAPD6, Hs2). Both belonging to the same phylogenetic group (Scott & Norbury, 2013) and are characterized by the presence of two catalytic domains (NTDs), one of which contains the mutation leading to the inability to catalyze the polymerase reaction. Both polymerases also possess the central domain (PAP associated domain) that is represented in many RNA modifying enzymes (Marchler- Bauer et al, 2007). They also feature zinc-coordinating domains. One zinc finger of the CCHH type (CX2CX12HX5H) located at the N-terminal part of the protein and three CCHC zinc knuckle motifs (CX2CX12HX5H) surrounding the NTD domain. The zinc finger motif in TUT4 mediates interaction with the LIN28A cofactor (Thornton et al, 2012). Zinc knuckles are generally responsible for the interaction and recognition of the nucleic acid (Thornton et al, 2012) they are potentially involved in the protein-protein interaction (Krishna et al, 2003).

Mechanisms of the UTP specificity of terminal uridyl transferases

The catalytic domain of the non-canonical RNA polymerases is highly conserved and contains similarities which were transferred even through the species barrier (Martin & Keller, 2007). Despite this, different functional di- versity within the group of the polymerases is observed. The homologs were separated on two categories based on their function. Some polymerases like PAPD5, TUTase 1 are specialized on the addition of the adenines to 3' end of the RNA substrate, while the other group was selectively incorporating the uridines (TUTase 4, TUTase 7). Such a preference for the particular nu- cleoside triphosphate as an original substrate was not clear. The detailed understanding of the nucleotide selection by the terminal uridyl transferases was explained with a recently solved crystal structures that are available for and 3' terminal TUTase from Trypanosoma brucei (Aphasizhev et al, 2002)

28 Dmytro Ustianenko Introduction and yeast Cid1 (Lunde et al, 2012; Munoz-Tello et al, 2012; Munoz-Tello et al, 2013; Yates et al, 2012) (Figure 3). These works explained certain aspects of the RNA binding and UTP preference which are crucial for the family of TUTases. Cid1 crystal structure showed that the preference for UTP substrate can be explained by the active site architecture that allows the selection of UTP at the sugar-base edge, by the interaction between Asn171 and the pyrimidine carbonyl and the ribose 2′ OH (Figure 4, top panel). Unexpected was the involvement of the conserved His336 residue in the recognition of the UTP substrate (Figure 4). It occurs that the additional uracil selection process takes advantage of the flipped His336 conformer as a detector of the uracil- specific carbonyl. It appears to be an evolutionary novel mechanism as the

Figure 3. Crystal structure of the CID-1 protein. C-terminal domain is highlighted in green, N-terminal domain in dark blue, catalytic aspartate triad is orange (Yates et al, 2012)

mutation by a single nucleotide substitution of His336 for alanine was able to convert the enzyme specificity to ATP. Lunde et al, 2012 has reported the structure that was solved after soaking the crystals with all four nucleotides.

29 Introduction Dmytro Ustianenko

It indicated that Cid1 nucleotide binding pocket can potentially incorporate all of the nucleotides (Read et al, 2002) by applying conformational changes that lead to a different mode of recognition. Importantly, the amino acid motif that is responsible for the UTP recognition is highly conserved between Cid1 and two human TUTases: TUTase 4 and TUTase 7 (Figure 4 left panel).

Figure 4. The detailed view on the catalytic center of the CID-1 enzyme. Catalytically required aspartates and residues involved in the RNA binding are indicated. Bottom panel shows the alignment of several TUTase family members. Histi- dine336 that is involved in the UTP binding is red framed (Yates et al, 2012).

Uridylation of mammalian RNAs

Till now our understanding of the RNA uridylation in human cells is limited to several reports and only some classes of RNA. The most well studied examples are described below. Recent work by (Choi et al, 2012) has shed some light on the overall variety of the uridylated RNA in human cell. Deep sequencing analysis of the RNA ~200 nt length revealed that a variety of the miRNA, certain tRNA, snRNA and snoRNA can possess at least one untemplated uridine at their 3′ end. A class of transcription start site associated RNA (TSS-RNA) and spliced introns were also reported to contain untemplated uridines that are potentially involved in the quick turnover of these transcripts in human nucleus. Several known examples of the uridylated RNA in humans and the protein factors responsible for this are summarized in the Table 1.

30 Dmytro Ustianenko Introduction

Table 1. Reported uridylated RNA in humans. Only several RNA in human cells were reported to be uridylated. Among those well- studied is the mRNA of histone gene and precursors of miRNA. The enzymes responsible for the 3’ uridylation are listed.

RNA type Enzyme References Histone mRNA, Pre-miRNA, TUTase4, TUTase 7 (Heo et al, 2012; Heo et al, 2009; Schmidt miRNA et al, 2011; Thornton et al, 2012)

U6 snRNA TUT1 (Mellman & Anderson, 2009; Trippe et al, 2006; Trippe et al, 2003) TSS-RNA, Spliced Introns ??? (Choi et al, 2012)

The role of the uridylation in histone mRNA degradation

The majority of the cellular mRNA possesses a canonical polyA tail that is essential for their cytoplasmic export, stability and translation. The exception is the mRNA of the histone genes which undergo a unique and complex 3' end processing mechanism that involves specific structural and sequence elements of the mRNA molecule. The U7 snRNA recognizes a specific sequence (histone downstream element, HDE) in the histone pre-mRNA and facilitates the 3′ - end cleavage (Birchmeier et al, 1984; Galli et al, 1983; Mowry & Steitz, 1987). The interaction of the U7 snRNA with HDE is stabilized by the protein factor SLBP (Melin et al, 1992; Spycher et al, 1994). Mature histone mRNA accompanied by the bound SLBP is exported to the sites of translation. The histone expression is cell cycle dependent and occurs during the S phase and rapidly decreased after the transition to G2 phase (Borun et al, 1967). The rapid removal of the histone coding transcript is required for the cell cycle progression or during DNA damage response (Kaygun & Marzluff, 2005a; Kaygun & Marzluff, 2005b). Histone mRNA is getting 3′ terminally uridylated by TUTase 4 (ZCCHC11) (Schmidt et al, 2011) in human and by Cid1 in yeast Schizosaccharomyces pombe (Rissland et al, 2007) prior to its degradation. This facilitates the consequent decapping and exonucleolytic cleavage of the mRNA by Xrn1 from the 5' end and by a 3' to 5' exoribonuclease Eri1 (Mullen & Marzluff, 2008; Rissland & Norbury, 2009) (Figure 5).

31 Introduction Dmytro Ustianenko

Besides mammalian histones, uridylation of the mRNA has been observed only in yeasts and plants but was not reported for other mammalian mRNAs. In yeast, the polymerase Cid1 can add stretches of uridines after the poly(A) tail (Rissland et al, 2007). This modification of the mRNA leads to the recruitment of the PAT1-Lsm1-7 complex which mediates subsequent decapping leading to the translation inhibition of the transcript and destabili- zation with consequent 5' – 3' degradation of the mRNA (Rissland & Norbury, 2009). Malecki et al, 2013 has reported that the uridylated mRNA in yeast are targeted by a novel 3′ – 5′ exoribonuclease Dis3l2 that has a high substrate preference towards the polyuridine tract.

cis-regulatory loop

Histone mRNA

Cell cycle progression, Figure 5. Process of DNA damage the uridylation medi- ZCCHC11 ated histone mRNA degradation. Prior to mitosis or after the DNA damage in Lsm 1-7 Recruiting human histone mRNA 3’ Exo is 3' uridylated by TUT- Decapping ase 4. U-tail is recruiting the Lsm complex that stimulates the de- Xrn1 capping and Eri1 (3′ 3’ Exo Exo) nuclease. mRNA Degradation is degraded probably from both 3' to 5' and 5′ 5’ to 3’ 3’ to 5’ to 3′ end.

The role of uridylation in miRNA processing and degradation

Currently, there are more than 25000 miRNA annotated in 206 species from worms, to humans (Kozomara & Griffiths-Jones, 2014). Micro RNAs

32 Dmytro Ustianenko Introduction

are small RNA molecules which undergo a complicated, multistep processing mechanism. This class of functional RNA’s is transcribed by the RNA polymerase II (Lee et al, 2004). In certain cases, when miRNA is located in the area of the Alu element, transcription can be performed by the RNA polymerase III (Borchert et al, 2006). miRNA genes are often located in clusters (approx. 50% of all mammalian miRNA’s (Lee et al, 2002)) that can be in both intronic as well as exonic areas of coding and non-coding transcripts (Calin et al, 2002; Tam, 2001). The primary miRNA precursor is transcribed as a longer transcript. Primary miRNAs are processed by DGCR8/Drosha complex in the nucleus of all eukaryotes (Denli et al, 2004; Lee et al, 2003). The obtained product of the endonucleolytic reaction is approximately 70nt long precursor miRNA molecule (pre-miRNA) which is characterized not only by its length but more importantly by a typical stem loop secondary structure. Pre-miRNA is exported to the cytoplasm by the exportin V protein (Lund et al, 2004) where it undergoes the final maturation step. This is mediated by the endonuclease DICER which possesses PAZ/PIWI domains and results in the production of mature 20 - 25nt long, double stranded miRNA (Hutvagner et al, 2001; Ketting et al, 2001). Mature miRNA duplexes are recognized, unwound and incorporated into the functional RNA induced silencing complex (RISC) by Argo- naut proteins (Hammond et al, 2001) (Figure 6).Hu- man cells contain four pri-miRNA homologs of the Ago pro- Drosha/DGCR8

Nucleus EXP5 Figure 6. Schematic path- pre-miRNA Cytoplasm way of the miRNA maturation in mammals. miRNAs are transcribed by Dicer RNA Pol II and undergo a two steps of the endonucleolytic cleavage first in the nucleus mature miRNA and then in the cytoplasm. Mature form is incorporated into the functional RISC com- AGO/RISC plex.

33 Introduction Dmytro Ustianenko tein (Ago1, 2, 3 and 4), and several Ago related proteins, thus at least eight RISC complexes exist in humans. Besides Argonaut proteins, RISC complex contain additional components such as nucleases (Meister et al, 2005), helicases (Robb & Rana, 2007) and RNA binding proteins (Chendrimada et al, 2005; Lee et al, 2006). Only Ago2 is catalytically active endonuclease and possess a slicer activity (ability to endonucleolytically “slice” the targeted mRNA). Recognition and binding of miRNA by Ago proteins strongly depends on their secondary structure. miRNA duplexes with central mismatches are preferentially sorted into AGO1 complex, whereas perfectly matching short interfering (siRNA) duplexes are incorporated into AGO2 (Forstemann et al, 2007; Tomari et al, 2007) RISC complex silencing activity depends on the ability to complementary recognize mRNA molecule. Once it is recognized and bound it is cleaved (sliced) or translationary repressed. Thus, such an activity of the RISC complex is mainly affected by the miRNA component of RISC complex. The 5' part of the miRNA’s is more conserved then the 3' part, this area is called the “seed sequence” of the miRNA. Disruption of this area by mutation relived the miRNA target repression suggesting that this seed paring is important for the target recognition (Brennecke et al, 2005; Doench & Sharp, 2004). Translational repression is a hallmark of the miRNA mediated gene repression while mRNA cleavage is more typical for the siRNA mediated gene repression. It was shown that miRNA Lin4 of C.elegans is able to bind 3’ UTR of the lin-14 transcript and repress its expression (Lee et al, 1993; Olsen & Ambros, 1999; Wightman et al, 1993). A whole range of the mRNA targets of miRNA’s was identified during genome wide studies using modern CLIP tech- nique (crosslinking and immunoprecipitation) (Hafner et al, 2010; Jaskiewicz et al, 2012a).After the primary recognition of the mRNA by RISC complex, the GW182 complex is recruited. The direct interaction between RISC and GW182 is essential for miRNA mediated repression mechanism (Eulalio et al, 2008). This cytoplasmic complex mediates the mRNA deadenylation by CCR4-Not1 or PAN2 (Figure 7) complexes which also trigger the decapping mechanisms (Behm-Ansmant et al, 2006) and deposit the mRNA transcript in the “hibernating” state to the P-bodies (Franks & Lykke-Andersen, 2008). Thus Argonaut and GW182 proteins are enriched in the P-bodies (Eulalio et al, 2007). Depletion of any of these proteins causes the reduction or even

34 Dmytro Ustianenko Introduction

Figure 7. Mechanisms of miRNA RISC complex functioning. Upper panel: initiation of translation repression by recruitment of the GW182 complex with consequent engagement of the CCR4 and CAF complex for the deadenylation. Lower panel: miRNA guided mRNA degradation initiated by AGO2 slicer activity (adapted from Krol et al, 2010)

disappearance of the P-bodies (Chu & Rana, 2006; Yang et al, 2004). P-bodies are enriched with mRNA decay factors and pools of stored mRNAs. The messenger that is stored in the P-bodies can be further polyadenylated and recruited to the sites of the protein translation (Bhattacharyya et al, 2006).

Uridylation of the miRNA

Several terminal uridine transferases were implicated in the addition of a single or up to two uridine nucleotides to the pre-miRNA molecules. This modification was shown to be mediated by TUTase 7, TUTase 4 and TUTase 2 enzymes in the cytoplasm of human cells (Heo et al, 2012). As the effect, the increased DICER cleavage was observed, suggesting that such a modification can stimulate the DICER endonucleolytic activity and pre-miRNA production efficiency. The proposed mechanism is based on the enhanced loading of pre-miRNA to the functional DICER effector complex.

35 Introduction Dmytro Ustianenko

Mechanisms of the miRNA mediated gene expression regulation are essential for the cell during each step of the life cycle. Interesting phenomena was observed in the nematode C. elegans and later projected on other eukaryotes from Drosophila to mouse and human cells (Pasquinelli et al, 2003; Pasquinelli et al, 2000; Reinhart et al, 2000). During the embryonic cell differentiation the high level of one particular class of miRNA was detected (let-7 miRNA family members). Surprisingly the negative correlation was observed with Lin28 protein component which is one of the stem cell pluripotency factors (detailed scheme in Figure 8). Lin28 protein is essential for maintaining the self-renewing state of the cell. The regulatory loop was established between Lin28 protein and miRNAs from the Let-7 family which can directly target mRNA of the Lin28. LIN28 overexpression or let-7 inhibition promotes reprogramming of human and mouse fibroblasts to induced pluripotent stem cells (Martinez & Gregory, 2010; Yu et al, 2007). Further investigation of this mechanism identified Lin28 to be an RNA binding protein which can specifically recognize the let-7 pre-miRNAs. The molecular mechanism of the pre- let7 recognition was explained with the crystal structure published by (Nam et al, 2011). Both LIN28 RNA binding domains are involved in the recognition of the let-7 precursor. The cold shock (CSD) domain of the Lin28 was proposed to recognize the GNGAYN motif (where N – any nucleotide, Y – pyrimidine) while two zinc finger domain of the protein are recognizing the GGAG motif on the stem loop part of the miRNA (Figure 9). Recent in vivo studies using CLIP technology in both human and mouse ES

Figure 8. Expression profiles of lin28 and let7 during differentiation. Expression of lin-28 and let-7 miRNA during transition from ESC to terminally dedifferentiated cell line in worms (top panel) and vertebrates (bottom panel). (modified from (Nimmo & Slack, 2009).

36 Dmytro Ustianenko Introduction

cells confirmed the recognition mode of aaGxGG in the in vivo Lin28 substrates (Cho et al, 2012; Wilbert et al, 2012). Several laboratories have simultaneously reported on their identification of the TUTase 4 being responsible for further modification of the pre-miRNA on its 3’ end with the stretches of polyuridines (poly(U)s) (Hagan et al, 2009; Heo et al, 2009). This modification of the pre-miRNA abolishes the loading and further processing of the molecule by DICER protein. It was proposed that 3’ end uridylation of the pre-miRNA serves as a signal for the cytoplasmic degradation machinery.

Figure 9. Schematic model of pre-miRNA let-7 recognition by the LIN28 protein. RNA binding domains of the LIN28: CSD – in blue, CCHC zink fingers in light green. Yellow RNA bases indicate the position of the mature miRNA sequence start. DICER 5' and 3' cleavage sites are indicated (Nam et al, 2011).

TUTase 4 is also involved in the cytokine expression in human cells. It is able to directly uridylate the miR-26, adding up to three nucleotides to the miRNA duplex in this way triggering the degradation of miRNA. This results in the increased expression of the miR-26 targeted mRNA, interleukin-6 (IL-6) (Jones et al, 2009). The depletion of TUTase 4 destabilizes the IL-6 mRNA by affecting its poly(A) tail length that is classical/typical example of the miRNA mediated gene silencing. This represents the fine tuning of miRNA function expression, since the miR-26b isoform of the miR-26 has a 30x times less 3' terminal uridylation (Jones et al, 2009) and is not capable of targeting the IL-6 mRNA. The studies performed on TUTase 4 deficient mice indicated the- in volvement of the gene in the growth and survival of the offspring’s. Depletion

37 Introduction Dmytro Ustianenko of the gene results in the global reduction of the mature miRNA uridylation levels. This leads to the increased stability of the miRNA and as a result to the decreased expression of their targets (i.e. insulin-like growth factor (IGF-1)) in both liver and blood of the mutant mice. IGF-1 depleted mice recapitulated symptoms observed by the depletion of the TUTase 4 gene (Jones et al, 2012).

RNA degradation at the 3' ends

Exosome complex in eukaryotes

Exosome is a major 3' end degradation complex which is present both in the nucleus and cytoplasm of eukaryotic cells, and is highly conserved from yeasts to humans. It is a barrel shaped multiprotein complex which consists of ten subunits (Rrp4p, Csl4p, Mtr3p, Rrp40p, Rrp41, pRrp42p, Rrp43p, Rrp45p and Rrp46p) (Allmang et al, 1999b) (Figure 10). Nine out of ten subunits form the scaffold and in higher eukaryotes are catalytically inactive while the active component of the complex is the tenth subunit Rrp44 (Dis3) or Rrp6 (Burkard & Butler, 2000; Mitchell et al, 1997). Dis3 protein possesses a distinct 3' exoribonucleolytic activity (Dziembowski et al, 2007). Besides the 3' exo activity, it was recently shown that the C' terminal domain, the PIN domain is also able to catalyze RNA in endoribonucleolytic manner broaden- ing the RNAse activity of the protein and the exosome (Lebreton et al, 2008; Schaeffer et al, 2009; Schneider et al, 2009). The exosome participates in the maturation of the ribosomal and small nuclear RNA (Allmang et al, 1999a; Mitchell et al, 1996). It is also involved in the nonsense mediated decay and the mRNA turnover (Mitchell & Tollervey, 2003). Exosome has a wide spectrum of substrates. That is possible due to the association of the exosome with various co-factors. Nuclear exosome interacts with the TRAMP complex (LaCava et al, 2005; Vanacova et al, 2005) and degrades the aberrant RNAs that are recognized by the TRAMP. Exosome can also be directly recruited to the sites of the transcription (Vasiljeva & Buratowski, 2006) where it interacts with the NRD1 transcription termination complex that consists of several proteins

38 Dmytro Ustianenko Introduction

(Nrd1p, Nab3p, CTD kinase CTDK-I) and is assembled directly on the C’ terminal domain of the RNA polymerase II (Conrad et al, 2000). In human cells, nuclear exosome is also associated with the novel NEXT (nuclear exosome targeting complex) complex that consist of hMTR4, the Zn-knuckle protein ZCCHC8, and the RNA binding protein RBM7 (Lubas et al, 2011). This complex targets promoter upstream transcripts (PROMTS) that were detected at the first time upon the depletion of the human exosome components (Preker et al, 2008). PROMTS are transcripts that are similar to the cryptic unstable transcripts (CUT’s) in yeasts; their stability is also dependent on the activity of the nuclear exosome (Wyers et al, 2005). Thus, human NEXT complex shares certain function similarity with the yeast TRAMP and NRD1 complex that is absent in human cells.

Figure 10. Crystal structure of the yeast exosome. Structure is represented in to orientations rotated 90○ around vertical axis. RNA is indicated in black and positioned in the major exosome channel. Rrp44 (Dis3) is located in the bottom of the barrel. Rrp6 subunit interacting with the Csl4 and PH subunits of the exosome (Makino et al, 2013).

39 Introduction Dmytro Ustianenko

DIS3 homologs in humans

Unlike in yeasts, human cells contain three homologs of the Rrp44 the DIS3 gene (Staals et al, 2010). These homologs contain identical domain organization (Figure 11, top panel). DIS3 protein family members possess several cold shock domains (CSD1 and CSD2) and N terminal S1 domain. (Lorentzen et al, 2008) The crystal structure of the yeast Rrp44 explains the mode of the RNA binding of the active exosome subunit. From which is clear that CSD1, 2 and S1 domains are responsible for the RNA binding while the very 3' end of the RNA is hidden inside the RNB domain that is responsible

Figure 11. Schematic alignment of DIS3 proteins from yeast and human. PIN domain in brown, inactive PIN domain in hash brown, CSD1 and CSD2 in green and dark green, RNB in orange, S1 RNA binding domain in yellow. Bottom panel, the crystal structure of the yeast Rrp44 and E.coli RNase II. Showing similar structural domain positioning, interaction with the RNA (black) through CSD domains and S1 domain (Lorentzen et al, 2008).

40 Dmytro Ustianenko Introduction

for the catalytic function of the protein (Figure 11, bottom) PIN domain of the Rrp44 harbors the endonucleolityc activity (Schaeffer et al, 2009) and allows the interaction of the Rrp44 with the exosome (Schnei- der et al, 2009). Strains that had a point mutation in the PIN domain showed the defective processing of the 5’-ETS and ITS2 regions of the rRNA. Abolish- ing both endo- and exonucleolityc abilities of the protein is lethal (Schaeffer et al, 2009). DIS3L protein just like its yeast homolog contain the C' terminal PIN domain, but the endonucleolytic function is abolished due to the mutation of the catalytically required aspartate. Surprisingly, the third mammalian homolog of DIS3, DIS3L2 is missing the PIN domain completely. Due to the absence of the PIN domain in DIS3L2, its interaction with the exosome in vivo was not observed, suggesting that this previously uncharacterized exoribonuclease is acting exosome independently in the cytoplasm of the human cells. The role of DIS3L protein was shown to be connected to the degradation of the ribosomal RNA (Staals et al, 2010). Polyadenilated degradation intermediates were increased in the cytoplasm of HELA cells after the knock down of DIS3L (Tomecki et al, 2010).

Human DIS3 Like exonuclease 2

Recent work by Astutti showed that DIS3L2 is a cellular component responsible for the formation of the Perlman syndrome of overgrowth and increases susceptibility to Wilms tumor formation (Astuti et al, 2012). Patients that developed Perlman syndrome which often accompanies the progression of Wilms tumor were diagnosed with the chromosomal breakage in the area of the DIS3L2 gene (2q37.1). Such kind of gene structure interference resulted in the production of functionally incapable DIS3L2 protein, with completely absent or greatly suppressed nucleolytic potential. siRNA mediated knock down of DIS3L2 gene in the population of HELA cells leads to aneuploidy and severe chromosomal abnormalities during the cell division. Problems with chromosomal segregation, cells with enlarged nucleus and also polynuclear cells were detected after DIS3L2 depletion (Astuti et al, 2012). Besides that an increased amount of apoptotic cells were observed after the DIS3L2 knock down. This all resulted in the decrease of the culture growth, interestingly that

41 Introduction Dmytro Ustianenko cells with reduced amount of the DIS3L2 protein were also more suscepti- ble to the colony formation during the soft agar assay that could be rescued by the gene overexpression pointing towards the potential tumor suppressor function of the DIS3L2. Nevertheless, downregulation of the DIS3L2 also results in the reduced expression of the AURKB, TTK while stabilizing Cycline B1 and Rad21. That provides a solid evidence for the direct role of DIS3L2 in the pathogenesis and origin of this pathology (Astuti et al, 2012). Studies of DIS3L2 in yeasts Schizosaccharomyces pombe which unlike S. cerevisiae contain DIS3L2 homolog showed that this protein depletion sta- bilizes the mRNA in the cytoplasm of the cell, as well as the uridylation of the mRNA, proposing that DIS3L2 is the essential cellular factor responsible for the mRNA degradation (Malecki et al, 2013). Parallel work by Lubas showed similar observations of the mRNA stabilization and an increased number of P-bodies after human DIS3L2 depletion. Unlike its other homologs DIS3 and DIS3L, DIS3L2 is able to degrade a structural substrates and need a rather short “landing platform”, only a 2 nt overhang is sufficient for the degradation initiation (Lubas et al, 2013).

Other known 3' to 5' exonucleases

Besides exosome and DIS3 family members there are other protein factors that are capable of degrading RNA in 3' to 5' direction. The exonuclease Eri1 (3’ Exo) that was characterized by (Dominski et al, 2003) is implicated in the degradation of the histone mRNA (Mullen & Marzluff, 2008; Yang et al, 2006). Eri1 homologs are also present in worms where it is involved in the regulation of the RNA silencing mechanism (Kennedy et al, 2004). Inter- estingly, Sniper (Snp) the Drosophila homolog of Eri1 is able to degrade both structured DNA and RNA substrates with short overhangs. It is not involved in the degradation of the histone mRNA and do not participate in the RNA interference pathway as the mutant flies carrying a gene deletion has no phenotypes observed in yeast or C. elegans (Kupsco et al, 2006). Another set of proteins that possess distinct 3’ to 5 exoribonucleolytic activity is involved in the deadenylation of the mRNA transcripts in the cytoplasm. They were purified and characterized in the early 90’s by analyzing

42 Dmytro Ustianenko Introduction

the HELA cell extracts (Astrom et al, 1991; Astrom et al, 1992) and later on identified in yeasts (Boeck et al, 1996). The yeasts complex consists of at least two components Pan2p and Pan3p (Boeck et al, 1996; Brown et al, 1996). Pan3p is the catalytic subunit of the complex, and belongs to the family of the RNase T 3′ to 5′ exoribonucleases (Moser et al, 1997). Cells that are deficient ofPAN2 or PAN3 contain mRNA molecules with abnormally long poly(A) tails. Protein complex with similar activity was also reported in yeast (Tuck- er et al, 2001). Ccr4p and Caf1p are components of the major cytoplasmic poly(A)-specific complex. The same authors reported that even after the Ccr4p/Caf1p gene deletion the strains still possess the modest deadenylation activity that was explained by the involvement of Pan2p/Pan3p deadenylation complex (Tucker et al, 2001) suggesting a partial redundancy in the their activity. Currently there are more than eleven deadenylases characterized in a different organism. Here we will not describe them all, for details please see (Goldstrohm & Wickens, 2008)

Genome-wide approaches to study protein-RNA interactions

The cross-linking and immunoprecipitation (CLIP) (Ule et al, 2005) is originally developed as a modification of the chromatin immune precipitation technology (ChIP) (Ule et al, 2005) . DNA ChIP was successfully used for the identification of the DNA-protein interaction. The main principle of the method is the formaldehyde crosslinking of the DNA to the protein located in the close proximity with the consequent purification of the protein using tag or protein specific antibodies. Protein-DNA complexes are broken and DNA sequence is analyzed. Similar principle is used in the CLIP protocol. Protein-RNA complexes are stabilized and crosslinked with a UV light to form a covalent bond between amino acids and RNA nucleotides. CLIP protocol is taking advantage of the revolutionary large scale sequencing technology. For this, the protein bound RNA is trimmed using unspecific bacterial RNAses (RNAse I, T1 RNAse etc.) to shorten up the RNA fragment. Further the RNA is modified with the 5' and 3'

43 Introduction Dmytro Ustianenko linkers which are necessary for the deep sequencing library preparation and analysis. Obtained RNA fragments are converted to the cDNA and analyzed using currently available genome wide sequencing technology.In the recent years this method became one of the essential tools for the substrate identification of the RNA binding proteins in vivo. The CLIP method allows us to investigate in vivo assembled and physiologically existent protein-RNA complexes, helps to understand the RNA metabolism and function on the molecular level. CLIP protocol modifications (PAR-CLIP (Hafner et al, 2010), iCLIP (Konig et al, 2010)) allows for the exact binding position identification taking advantage of the single nucleotide mutations occurring during the material preparation. This gives an opportunity for a complete positional reconstitution of the RNA binding complexes on their substrates for example during the mRNA maturation (Martin et al, 2012). Such approaches have helped to obtain a deeper understanding of the mechanisms of splicing (Konig et al, 2010; Tollervey et al, 2011), alternative splicing (Licatalosi et al, 2008) and RNA maturation (Tuck & Tollervey, 2013). Besides this, a novel types and classes of the RNA’s were identified with the help of the CLIP protocol (Tuck & Tollervey, 2013). The existing miRNA databases are currently relying not just on the computational prediction of the miRNA gene and analysis of the RNAome, but also on the functional incorporation of the miRNA substrates in to Ago RISC complex (Chi

UV irradiation et al, 2009; Jaskiewicz et al, 2012a).

5’ 3’ RBP SOLEXA sequencing

RNAse treatment 5’ RBP 3’ PCR amplifcation Alkaline phosphatase RNA ligase

5’ RBP 3’ L5 Linker

RNA ligase 32 5’ 5’-γ- P RBP L3 Linker

Polynucleotide Proteinase K treatment kinase Figure 12. The work-

32 5’-γ- P RBP flow scheme of the CLIP Excision of the protein RNA complexes analysis. Schematic representation SDS-PAGE electrophoresis of the consequent steps for the library preparation.

References Dmytro Ustianenko

References

Allmang C, Kufel J, Chanfreau G, Mitchell P, Petfalski E, Tollervey D (1999a) Func- tions of the exosome in rRNA, snoRNA and snRNA synthesis. EMBO J 18(19): 5399- 5410

Allmang C, Petfalski E, Podtelejnikov A, Mann M, Tollervey D, Mitchell P (1999b) The yeast exosome and human PM-Scl are related complexes of 3' --> 5' exonucleases. Genes Dev 13(16): 2148-2158

Aphasizhev R, Sbicego S, Peris M, Jang SH, Aphasizheva I, Simpson AM, Rivlin A, Simpson L (2002) Trypanosome mitochondrial 3' terminal uridylyl transferase (TUT- ase): the key enzyme in U-insertion/deletion RNA editing. Cell 108(5): 637-648

Astrom J, Astrom A, Virtanen A (1991) In vitro deadenylation of mammalian mRNA by a HeLa cell 3' exonuclease. EMBO J 10(10): 3067-3071

Astrom J, Astrom A, Virtanen A (1992) Properties of a HeLa cell 3' exonuclease specific for degrading poly(A) tails of mammalian mRNA. J Biol Chem 267(25): 18154- 18159

Astuti D, Morris MR, Cooper WN, Staals RH, Wake NC, Fews GA, Gill H, Gentle D, Shuib S, Ricketts CJ, Cole T, van Essen AJ, van Lingen RA, Neri G, Opitz JM, Rump P, Stolte-Dijkstra I, Muller F, Pruijn GJ, Latif F, Maher ER (2012) Germline mutations in DIS3L2 cause the Perlman syndrome of overgrowth and Wilms tumor susceptibility. Nat Genet 44(3): 277-284

Bai Y, Srivastava SK, Chang JH, Manley JL, Tong L (2011) Structural basis for dimeri- zation and activity of human PAPD1, a noncanonical poly(A) polymerase. Mol Cell 41(3): 311-320

Behm-Ansmant I, Rehwinkel J, Doerks T, Stark A, Bork P, Izaurralde E (2006) mRNA degradation by miRNAs and GW182 requires both CCR4:NOT deadenylase and DCP1:DCP2 decapping complexes. Genes Dev 20(14): 1885-1898

46 Dmytro Ustianenko References

Berndt H, Harnisch C, Rammelt C, Stohr N, Zirkel A, Dohm JC, Himmelbauer H, Tavanez JP, Huttelmaier S, Wahle E (2012) Maturation of mammalian H/ACA box snoRNAs: PAPD5-dependent adenylation and PARN-dependent trimming. RNA 18(5): 958-972

Bhattacharyya SN, Habermacher R, Martine U, Closs EI, Filipowicz W (2006) Relief of microRNA-mediated translational repression in human cells subjected to stress. Cell 125(6): 1111-1124

Birchmeier C, Schumperli D, Sconzo G, Birnstiel ML (1984) 3' editing of mRNAs: sequence requirements and involvement of a 60-nucleotide RNA in maturation of histone mRNA precursors. Proc Natl Acad Sci U S A 81(4): 1057-1061

Boeck R, Tarun S, Jr., Rieger M, Deardorff JA, Muller-Auer S, Sachs AB (1996) The yeast Pan2 protein is required for poly(A)-binding protein-stimulated poly(A)-nuclease activity. J Biol Chem 271(1): 432-438

Borchert GM, Lanier W, Davidson BL (2006) RNA polymerase III transcribes human microRNAs. Nat Struct Mol Biol 13(12): 1097-1101

Borun TW, Scharff MD, Robbins E (1967) Rapidly labeled, polyribosome-associated RNA having the properties of histone messenger. Proc Natl Acad Sci U S A 58(5): 1977-1983

Brennecke J, Stark A, Russell RB, Cohen SM (2005) Principles of microRNA-target recognition. PLoS Biol 3(3): e85

Brown CE, Tarun SZ, Jr., Boeck R, Sachs AB (1996) PAN3 encodes a subunit of the Pab1p-dependent poly(A) nuclease in Saccharomyces cerevisiae. Mol Cell Biol 16(10): 5744-5753

Burkard KT, Butler JS (2000) A nuclear 3'-5' exonuclease involved in mRNA degradation interacts with Poly(A) polymerase and the hnRNA protein Npl3p. Mol Cell Biol 20(2): 604-616

47 References Dmytro Ustianenko

Burns DM, D'Ambrogio A, Nottrott S, Richter JD (2011) CPEB and two poly(A) polymerases control miR-122 stability and p53 mRNA translation. Nature 473(7345): 105- 108

Calin GA, Dumitru CD, Shimizu M, Bichi R, Zupo S, Noch E, Aldler H, Rattan S, Keat- ing M, Rai K, Rassenti L, Kipps T, Negrini M, Bullrich F, Croce CM (2002) Frequent deletions and down-regulation of micro- RNA genes miR15 and miR16 at 13q14 in chronic lymphocytic leukemia. Proc Natl Acad Sci U S A 99(24): 15524-15529

Castano IB, Heath-Pagliuso S, Sadoff BU, Fitzhugh DJ, Christman MF (1996) A novel family of TRF (DNA topoisomerase I-related function) genes required for proper nuclear segregation. Nucleic Acids Res 24(12): 2404-2410

Chendrimada TP, Gregory RI, Kumaraswamy E, Norman J, Cooch N, Nishikura K, Shiekhattar R (2005) TRBP recruits the Dicer complex to Ago2 for microRNA processing and gene silencing. Nature 436(7051): 740-744

Chi SW, Zang JB, Mele A, Darnell RB (2009) Argonaute HITS-CLIP decodes micro- RNA-mRNA interaction maps. Nature 460(7254): 479-486

Cho J, Chang H, Kwon SC, Kim B, Kim Y, Choe J, Ha M, Kim YK, Kim VN (2012) LIN28A is a suppressor of ER-associated translation in embryonic stem cells. Cell 151(4): 765-777

Choi YS, Patena W, Leavitt AD, McManus MT (2012) Widespread RNA 3'-end oligouridylation in mammals. RNA 18(3): 394-401

Chu CY, Rana TM (2006) Translation repression in human cells by microRNA-induced gene silencing requires RCK/p54. PLoS Biol 4(7): e210

Conrad NK, Wilson SM, Steinmetz EJ, Patturajan M, Brow DA, Swanson MS, Corden JL (2000) A yeast heterogeneous nuclear ribonucleoprotein complex associated with RNA polymerase II. Genetics 154(2): 557-571

Cui Y, Denis CL (2003) In vivo evidence that defects in the transcriptional elongation

48 Dmytro Ustianenko References factors RPB2, TFIIS, and SPT5 enhance upstream poly(A) site utilization. Mol Cell Biol 23(21): 7887-7901

D'Ambrogio A, Gu W, Udagawa T, Mello CC, Richter JD (2012) Specific miRNA stabilization by Gld2-catalyzed monoadenylation. Cell Rep 2(6): 1537-1545

Denli AM, Tops BB, Plasterk RH, Ketting RF, Hannon GJ (2004) Processing of primary microRNAs by the Microprocessor complex. Nature 432(7014): 231-235

Doench JG, Sharp PA (2004) Specificity of microRNA target selection in translational repression. Genes Dev 18(5): 504-511

Doma MK, Parker R (2006) Endonucleolytic cleavage of eukaryotic mRNAs with stalls in translation elongation. Nature 440(7083): 561-564

Dominski Z, Yang XC, Kaygun H, Dadlez M, Marzluff WF (2003) A 3' exonuclease that specifically interacts with the 3' end of histone mRNA. Mol Cell 12(2): 295-305

Dziembowski A, Lorentzen E, Conti E, Seraphin B (2007) A single subunit, Dis3, is essentially responsible for yeast exosome core activity. Nat Struct Mol Biol 14(1): 15-22

Eulalio A, Behm-Ansmant I, Izaurralde E (2007) P bodies: at the crossroads of post- transcriptional pathways. Nat Rev Mol Cell Biol 8(1): 9-22

Eulalio A, Huntzinger E, Izaurralde E (2008) GW182 interaction with Argonaute is essential for miRNA-mediated translational repression and mRNA decay. Nat Struct Mol Biol 15(4): 346-353

Fasken MB, Leung SW, Banerjee A, Kodani MO, Chavez R, Bowman EA, Purohit MK, Rubinson ME, Rubinson EH, Corbett AH (2011) Air1 zinc knuckles 4 and 5 and a conserved IWRXY motif are critical for the function and integrity of the Trf4/5-Air1/2- Mtr4 polyadenylation (TRAMP) RNA quality control complex. J Biol Chem 286(43): 37429-37445

49 References Dmytro Ustianenko

Forstemann K, Horwich MD, Wee L, Tomari Y, Zamore PD (2007) Drosophila micro- RNAs are sorted into functionally distinct argonaute complexes after production by dicer-1. Cell 130(2): 287-297

Franks TM, Lykke-Andersen J (2008) The control of mRNA decapping and P-body formation. Mol Cell 32(5): 605-615

Frischmeyer PA, van Hoof A, O'Donnell K, Guerrerio AL, Parker R, Dietz HC (2002) An mRNA surveillance mechanism that eliminates transcripts lacking termination codons. Science 295(5563): 2258-2261

Galli G, Hofstetter H, Stunnenberg HG, Birnstiel ML (1983) Biochemical complemen- tation with RNA in the Xenopus oocyte: a small RNA is required for the generation of 3' histone mRNA termini. Cell 34(3): 823-828

Glahder JA, Norrild B (2011) Involvement of hGLD-2 in cytoplasmic polyadenylation of human p53 mRNA. APMIS 119(11): 769-775

Goldstrohm AC, Wickens M (2008) Multifunctional deadenylase complexes diversify mRNA control. Nat Rev Mol Cell Biol 9(4): 337-344

Graber JH, Cantor CR, Mohr SC, Smith TF (1999) Genomic detection of new yeast pre-mRNA 3'-end-processing signals. Nucleic Acids Res 27(3): 888-894

Hafner M, Landthaler M, Burger L, Khorshid M, Hausser J, Berninger P, Rothballer A, Ascano M, Jr., Jungkamp AC, Munschauer M, Ulrich A, Wardle GS, Dewell S, Zavolan M, Tuschl T (2010) Transcriptome-wide identification of RNA-binding protein and microRNA target sites by PAR-CLIP. Cell 141(1): 129-141

Hagan JP, Piskounova E, Gregory RI (2009) Lin28 recruits the TUTase Zcchc11 to inhibit let-7 maturation in mouse embryonic stem cells. Nat Struct Mol Biol 16(10): 1021-1025

Hake LE, Mendez R, Richter JD (1998) Specificity of RNA binding by CPEB: requirement for RNA recognition motifs and a novel zinc finger. Mol Cell Biol 18(2): 685-693

50 Dmytro Ustianenko References

Hammond SM, Boettcher S, Caudy AA, Kobayashi R, Hannon GJ (2001) Argonaute2, a link between genetic and biochemical analyses of RNAi. Science 293(5532): 1146- 1150

He F, Peltz SW, Donahue JL, Rosbash M, Jacobson A (1993) Stabilization and ribosome association of unspliced pre-mRNAs in a yeast upf1- mutant. Proc Natl Acad Sci U S A 90(15): 7034-7038

Heo I, Ha M, Lim J, Yoon MJ, Park JE, Kwon SC, Chang H, Kim VN (2012) Mono- uridylation of pre-microRNA as a key step in the biogenesis of group II let-7 microR- NAs. Cell 151(3): 521-532

Heo I, Joo C, Kim YK, Ha M, Yoon MJ, Cho J, Yeom KH, Han J, Kim VN (2009) TUT4 in concert with Lin28 suppresses microRNA biogenesis through pre-microRNA uridylation. Cell 138(4): 696-708

Hutvagner G, McLachlan J, Pasquinelli AE, Balint E, Tuschl T, Zamore PD (2001) A cellular function for the RNA-interference enzyme Dicer in the maturation of the let-7 small temporal RNA. Science 293(5531): 834-838

Inada T, Aiba H (2005) Translation of aberrant mRNAs lacking a termination codon or with a shortened 3'-UTR is repressed after initiation in yeast. EMBO J 24(8): 1584- 1595

Jaskiewicz L, Bilen B, Hausser J, Zavolan M (2012) Argonaute CLIP--a method to identify in vivo targets of miRNAs. Methods 58(2): 106-112

Jones MR, Blahna MT, Kozlowski E, Matsuura KY, Ferrari JD, Morris SA, Powers JT, Daley GQ, Quinton LJ, Mizgerd JP (2012) Zcchc11 uridylates mature miRNAs to enhance neonatal IGF-1 expression, growth, and survival. PLoS Genet 8(11): e1003105

Jones MR, Quinton LJ, Blahna MT, Neilson JR, Fu S, Ivanov AR, Wolf DA, Mizgerd JP (2009) Zcchc11-dependent uridylation of microRNA directs cytokine expression. Nat Cell Biol 11(9): 1157-1163

51 References Dmytro Ustianenko

Kadaba S, Krueger A, Trice T, Krecic AM, Hinnebusch AG, Anderson J (2004) Nucle- ar surveillance and degradation of hypomodified initiator tRNAMet in S. cerevisiae. Genes Dev 18(11): 1227-1240

Katoh T, Sakaguchi Y, Miyauchi K, Suzuki T, Kashiwabara S, Baba T (2009) Selective stabilization of mammalian microRNAs by 3' adenylation mediated by the cytoplasmic poly(A) polymerase GLD-2. Genes Dev 23(4): 433-438

Kaygun H, Marzluff WF (2005a) Regulated degradation of replication-dependent histone mRNAs requires both ATR and Upf1. Nat Struct Mol Biol 12(9): 794-800

Kaygun H, Marzluff WF (2005b) Translation termination is involved in histone mRNA degradation when DNA replication is inhibited. Mol Cell Biol 25(16): 6879-6888

Kennedy S, Wang D, Ruvkun G (2004) A conserved siRNA-degrading RNase nega- tively regulates RNA interference in C. elegans. Nature 427(6975): 645-649

Ketting RF, Fischer SE, Bernstein E, Sijen T, Hannon GJ, Plasterk RH (2001) Dicer functions in RNA interference and in synthesis of small RNA involved in developmental timing in C. elegans. Genes Dev 15(20): 2654-2659

Konig J, Zarnack K, Rot G, Curk T, Kayikci M, Zupan B, Turner DJ, Luscombe NM, Ule J (2010) iCLIP reveals the function of hnRNP particles in splicing at individual nucleotide resolution. Nat Struct Mol Biol 17(7): 909-915

Kozomara A, Griffiths-Jones S (2014) miRBase: annotating high confidence microR- NAs using deep sequencing data. Nucleic Acids Res 42(1): D68-73

Krishna SS, Majumdar I, Grishin NV (2003) Structural classification of zinc fingers: survey and summary. Nucleic Acids Res 31(2): 532-550

Krol J, Loedige I, Filipowicz W (2010) The widespread regulation of microRNA biogenesis, function and decay. Nat Rev Genet 11(9): 597-610

Kupsco JM, Wu MJ, Marzluff WF, Thapar R, Duronio RJ (2006) Genetic and biochem-

52 Dmytro Ustianenko References ical characterization of Drosophila Snipper: A promiscuous member of the metazoan 3'hExo/ERI-1 family of 3' to 5' exonucleases. RNA 12(12): 2103-2117

Kwak JE, Drier E, Barbee SA, Ramaswami M, Yin JC, Wickens M (2008) GLD2 poly(A) polymerase is required for long-term memory. Proc Natl Acad Sci U S A 105(38): 14644-14649

LaCava J, Houseley J, Saveanu C, Petfalski E, Thompson E, Jacquier A, Tollervey D (2005) RNA degradation by the exosome is promoted by a nuclear polyadenylation complex. Cell 121(5): 713-724

Lebreton A, Tomecki R, Dziembowski A, Seraphin B (2008) Endonucleolytic RNA cleavage by a eukaryotic exosome. Nature 456(7224): 993-996

Lee RC, Feinbaum RL, Ambros V (1993) The C. elegans heterochronic gene lin-4 encodes small RNAs with antisense complementarity to lin-14. Cell 75(5): 843-854

Lee Y, Ahn C, Han J, Choi H, Kim J, Yim J, Lee J, Provost P, Radmark O, Kim S, Kim VN (2003) The nuclear RNase III Drosha initiates microRNA processing. Nature 425(6956): 415-419

Lee Y, Hur I, Park SY, Kim YK, Suh MR, Kim VN (2006) The role of PACT in the RNA silencing pathway. EMBO J 25(3): 522-532

Lee Y, Jeon K, Lee JT, Kim S, Kim VN (2002) MicroRNA maturation: stepwise processing and subcellular localization. EMBO J 21(17): 4663-4670

Lee Y, Kim M, Han J, Yeom KH, Lee S, Baek SH, Kim VN (2004) MicroRNA genes are transcribed by RNA polymerase II. EMBO J 23(20): 4051-4060

Licatalosi DD, Mele A, Fak JJ, Ule J, Kayikci M, Chi SW, Clark TA, Schweitzer AC, Blume JE, Wang X, Darnell JC, Darnell RB (2008) HITS-CLIP yields genome-wide insights into brain alternative RNA processing. Nature 456(7221): 464-469

Lingner J, Kellermann J, Keller W (1991) Cloning and expression of the essential

53 References Dmytro Ustianenko gene for poly(A) polymerase from S. cerevisiae. Nature 354(6353): 496-498

Lorentzen E, Basquin J, Tomecki R, Dziembowski A, Conti E (2008) Structure of the active subunit of the yeast exosome core, Rrp44: diverse modes of substrate recruitment in the RNase II nuclease family. Mol Cell 29(6): 717-728

Lubas M, Christensen MS, Kristiansen MS, Domanski M, Falkenby LG, Lykke-An- dersen S, Andersen JS, Dziembowski A, Jensen TH (2011) Interaction profiling identifies the human nuclear exosome targeting complex. Mol Cell 43(4): 624-637

Lubas M, Damgaard CK, Tomecki R, Cysewski D, Jensen TH, Dziembowski A (2013) Exonuclease hDIS3L2 specifies an exosome-independent 3'-5' degradation pathway of human cytoplasmic mRNA. EMBO J 32(13): 1855-1868

Lund E, Guttinger S, Calado A, Dahlberg JE, Kutay U (2004) Nuclear export of micro- RNA precursors. Science 303(5654): 95-98

Lunde BM, Magler I, Meinhart A (2012) Crystal structures of the Cid1 poly (U) polymerase reveal the mechanism for UTP selectivity. Nucleic Acids Res 40(19): 9815- 9824

Makino DL, Baumgartner M, Conti E (2013) Crystal structure of an RNA-bound 11-subunit eukaryotic exosome complex. Nature 495(7439): 70-75

Malecki M, Viegas SC, Carneiro T, Golik P, Dressaire C, Ferreira MG, Arraiano CM (2013) The exoribonuclease Dis3L2 defines a novel eukaryotic RNA degradation pathway. EMBO J

Marchler-Bauer A, Anderson JB, Derbyshire MK, DeWeese-Scott C, Gonzales NR, Gwadz M, Hao L, He S, Hurwitz DI, Jackson JD, Ke Z, Krylov D, Lanczycki CJ, Liebert CA, Liu C, Lu F, Lu S, Marchler GH, Mullokandov M, Song JS, Thanki N, Yamashita RA, Yin JJ, Zhang D, Bryant SH (2007) CDD: a conserved domain database for in- teractive domain family analysis. Nucleic Acids Res 35(Database issue): D237-240

Martin G, Gruber AR, Keller W, Zavolan M (2012) Genome-wide analysis of pre-

54 Dmytro Ustianenko References mRNA 3' end processing reveals a decisive role of human cleavage factor I in the regulation of 3' UTR length. Cell Rep 1(6): 753-763

Martin G, Keller W (2007) RNA-specific ribonucleotidyl transferases. RNA 13(11): 1834-1849

Martinez NJ, Gregory RI (2010) MicroRNA gene regulatory pathways in the establishment and maintenance of ESC identity. Cell Stem Cell 7(1): 31-35

Meister G, Landthaler M, Peters L, Chen PY, Urlaub H, Luhrmann R, Tuschl T (2005) Identification of novel argonaute-associated proteins. Curr Biol 15(23): 2149-2155

Melin L, Soldati D, Mital R, Streit A, Schumperli D (1992) Biochemical demonstration of complex formation of histone pre-mRNA with U7 small nuclear ribonucleoprotein and hairpin binding factors. EMBO J 11(2): 691-697

Mellman DL, Anderson RA (2009) A novel gene expression pathway regulated by nuclear phosphoinositides. Adv Enzyme Regul 49(1): 11-28

Mercer TR, Neph S, Dinger ME, Crawford J, Smith MA, Shearwood AM, Haugen E, Bracken CP, Rackham O, Stamatoyannopoulos JA, Filipovska A, Mattick JS (2011) The human mitochondrial transcriptome. Cell 146(4): 645-658

Mitchell P, Petfalski E, Shevchenko A, Mann M, Tollervey D (1997) The exosome: a conserved eukaryotic RNA processing complex containing multiple 3'-->5' exoribonucleases. Cell 91(4): 457-466

Mitchell P, Petfalski E, Tollervey D (1996) The 3' end of yeast 5.8S rRNA is generated by an exonuclease processing mechanism. Genes Dev 10(4): 502-513

Mitchell P, Tollervey D (2003) An NMD pathway in yeast involving accelerated deadenylation and exosome-mediated 3'-->5' degradation. Mol Cell 11(5): 1405-1413

Moser MJ, Holley WR, Chatterjee A, Mian IS (1997) The proofreading domain of Es- cherichia coli DNA polymerase I and other DNA and/or RNA exonuclease domains.

55 References Dmytro Ustianenko

Nucleic Acids Res 25(24): 5110-5118

Mowry KL, Steitz JA (1987) Both conserved signals on mammalian histone pre-mR- NAs associate with small nuclear ribonucleoproteins during 3' end formation in vitro. Mol Cell Biol 7(5): 1663-1672

Muhlrad D, Parker R (1999) Aberrant mRNAs with extended 3' UTRs are substrates for rapid degradation by mRNA surveillance. RNA 5(10): 1299-1307

Mullen TE, Marzluff WF (2008) Degradation of histone mRNA requires oligouridylation followed by decapping and simultaneous degradation of the mRNA both 5' to 3' and 3' to 5'. Genes Dev 22(1): 50-65

Munoz-Tello P, Gabus C, Thore S (2012) Functional implications from the Cid1 poly(U) polymerase crystal structure. Structure 20(6): 977-986

Munoz-Tello P, Gabus C, Thore S (2013) A critical switch in the enzymatic properties of the Cid1 protein deciphered from its product-bound crystal structure. Nucleic Acids Res

Nagaike T, Suzuki T, Katoh T, Ueda T (2005) Human mitochondrial mRNAs are stabilized with polyadenylation regulated by mitochondria-specific poly(A) polymerase and polynucleotide phosphorylase. J Biol Chem 280(20): 19721-19727

Nakanishi T, Kubota H, Ishibashi N, Kumagai S, Watanabe H, Yamashita M, Kashi- wabara S, Miyado K, Baba T (2006) Possible role of mouse poly(A) polymerase mGLD-2 during oocyte maturation. Dev Biol 289(1): 115-126

Nam Y, Chen C, Gregory RI, Chou JJ, Sliz P (2011) Molecular basis for interaction of let-7 microRNAs with Lin28. Cell 147(5): 1080-1091

Nimmo RA, Slack FJ (2009) An elegant miRror: microRNAs in stem cells, developmental timing and cancer. Chromosoma 118(4): 405-418

Ojala D, Montoya J, Attardi G (1981) tRNA punctuation model of RNA processing in

56 Dmytro Ustianenko References human mitochondria. Nature 290(5806): 470-474

Olsen PH, Ambros V (1999) The lin-4 regulatory RNA controls developmental timing in Caenorhabditis elegans by blocking LIN-14 protein synthesis after the initiation of translation. Dev Biol 216(2): 671-680

Pasquinelli AE, McCoy A, Jimenez E, Salo E, Ruvkun G, Martindale MQ, Baguna J (2003) Expression of the 22 nucleotide let-7 heterochronic RNA throughout the Meta- zoa: a role in life history evolution? Evol Dev 5(4): 372-378

Pasquinelli AE, Reinhart BJ, Slack F, Martindale MQ, Kuroda MI, Maller B, Hayward DC, Ball EE, Degnan B, Muller P, Spring J, Srinivasan A, Fishman M, Finnerty J, Corbo J, Levine M, Leahy P, Davidson E, Ruvkun G (2000) Conservation of the sequence and temporal expression of let-7 heterochronic regulatory RNA. Nature 408(6808): 86-89

Preker P, Nielsen J, Kammler S, Lykke-Andersen S, Christensen MS, Mapendano CK, Schierup MH, Jensen TH (2008) RNA exosome depletion reveals transcription upstream of active human promoters. Science 322(5909): 1851-1854

Pulak R, Anderson P (1993) mRNA surveillance by the Caenorhabditis elegans smg genes. Genes Dev 7(10): 1885-1897

Raabe T, Bollum FJ, Manley JL (1991) Primary structure and expression of bovine poly(A) polymerase. Nature 353(6341): 229-234

Rammelt C, Bilen B, Zavolan M, Keller W (2011) PAPD5, a noncanonical poly(A) polymerase with an unusual RNA-binding motif. RNA 17(9): 1737-1746

Read RL, Martinho RG, Wang SW, Carr AM, Norbury CJ (2002) Cytoplasmic poly(A) polymerases mediate cellular responses to S phase arrest. Proc Natl Acad Sci U S A 99(19): 12079-12084

Reinhart BJ, Slack FJ, Basson M, Pasquinelli AE, Bettinger JC, Rougvie AE, Horvitz HR, Ruvkun G (2000) The 21-nucleotide let-7 RNA regulates developmental timing

57 References Dmytro Ustianenko in Caenorhabditis elegans. Nature 403(6772): 901-906

Richter JD (2007) CPEB: a life in translation. Trends Biochem Sci 32(6): 279-285

Rissland OS, Mikulasova A, Norbury CJ (2007) Efficient RNA polyuridylation by noncanonical poly(A) polymerases. Mol Cell Biol 27(10): 3612-3624

Rissland OS, Norbury CJ (2009) Decapping is preceded by 3' uridylation in a novel pathway of bulk mRNA turnover. Nat Struct Mol Biol 16(6): 616-623

Robb GB, Rana TM (2007) RNA helicase A interacts with RISC in human cells and functions in RISC loading. Mol Cell 26(4): 523-537 Saitoh S, Chabes A, McDonald WH, Thelander L, Yates JR, Russell P (2002) Cid13 is a cytoplasmic poly(A) polymerase that regulates ribonucleotide reductase mRNA. Cell 109(5): 563-573

San Paolo S, Vanacova S, Schenk L, Scherrer T, Blank D, Keller W, Gerber AP (2009) Distinct roles of non-canonical poly(A) polymerases in RNA metabolism. PLoS Genet 5(7): e1000555

Sartain CV, Cui J, Meisel RP, Wolfner MF (2011) The poly(A) polymerase GLD2 is required for spermatogenesis in Drosophila melanogaster. Development 138(8): 1619- 1629

Schaeffer D, Tsanova B, Barbas A, Reis FP, Dastidar EG, Sanchez-Rotunno M, Ar- raiano CM, van Hoof A (2009) The exosome contains domains with specific endori- bonuclease, exoribonuclease and cytoplasmic mRNA decay activities. Nat Struct Mol Biol 16(1): 56-62

Schmidt MJ, West S, Norbury CJ (2011) The human cytoplasmic RNA terminal U- transferase ZCCHC11 targets histone mRNAs for degradation. RNA 17(1): 39-44

Schneider C, Leung E, Brown J, Tollervey D (2009) The N-terminal PIN domain of the exosome subunit Rrp44 harbors endonuclease activity and tethers Rrp44 to the yeast core exosome. Nucleic Acids Res 37(4): 1127-1140

58 Dmytro Ustianenko References

Scott DD, Norbury CJ (2013) RNA decay via 3' uridylation. Biochim Biophys Acta 1829(6-7): 654-665

Shcherbik N, Wang M, Lapik YR, Srivastava L, Pestov DG (2010) Polyadenylation and degradation of incomplete RNA polymerase I transcripts in mammalian cells. EMBO Rep 11(2): 106-111

Sparks KA, Dieckmann CL (1998) Regulation of poly(A) site choice of several yeast mRNAs. Nucleic Acids Res 26(20): 4676-4687

Spycher C, Streit A, Stefanovic B, Albrecht D, Koning TH, Schumperli D (1994) 3' end processing of mouse histone pre-mRNA: evidence for additional base-pairing between U7 snRNA and pre-mRNA. Nucleic Acids Res 22(20): 4023-4030

Staals RH, Bronkhorst AW, Schilders G, Slomovic S, Schuster G, Heck AJ, Raijmak- ers R, Pruijn GJ (2010) Dis3-like 1: a novel exoribonuclease associated with the human exosome. EMBO J 29(14): 2358-2367

Stevenson AL, Norbury CJ (2006) The Cid1 family of non-canonical poly(A) polymerases. Yeast 23(13): 991-1000

Tam W (2001) Identification and characterization of human BIC, a gene on chromo- some 21 that encodes a noncoding RNA. Gene 274(1-2): 157-167

Thornton JE, Chang HM, Piskounova E, Gregory RI (2012) Lin28-mediated control of let-7 microRNA expression by alternative TUTases Zcchc11 (TUT4) and Zcchc6 (TUT7). RNA 18(10): 1875-1885

Tollervey JR, Curk T, Rogelj B, Briese M, Cereda M, Kayikci M, Konig J, Hortobagyi T, Nishimura AL, Zupunski V, Patani R, Chandran S, Rot G, Zupan B, Shaw CE, Ule J (2011) Characterizing the RNA targets and position-dependent splicing regulation by TDP-43. Nat Neurosci 14(4): 452-458

Tomari Y, Du T, Zamore PD (2007) Sorting of Drosophila small silencing RNAs. Cell 130(2): 299-308

59 References Dmytro Ustianenko

Tomecki R, Dmochowska A, Gewartowski K, Dziembowski A, Stepien PP (2004) Identification of a novel human nuclear-encoded mitochondrial poly(A) polymerase. Nucleic Acids Res 32(20): 6001-6014

Tomecki R, Kristiansen MS, Lykke-Andersen S, Chlebowski A, Larsen KM, Szczesny RJ, Drazkowska K, Pastula A, Andersen JS, Stepien PP, Dziembowski A, Jensen TH (2010) The human core exosome interacts with differentially localized processive RNases: hDIS3 and hDIS3L. EMBO J 29(14): 2342-2357

Trippe R, Guschina E, Hossbach M, Urlaub H, Luhrmann R, Benecke BJ (2006) Iden- tification, cloning, and functional analysis of the human U6 snRNA-specific terminal uridylyl transferase. RNA 12(8): 1494-1504

Trippe R, Richly H, Benecke BJ (2003) Biochemical characterization of a U6 small nuclear RNA-specific terminal uridylyltransferase. Eur J Biochem 270(5): 971-980

Trippe R, Sandrock B, Benecke BJ (1998) A highly specific terminal uridylyl transferase modifies the 3'-end of U6 small nuclear RNA. Nucleic Acids Res 26(13): 3119- 3126

Tuck AC, Tollervey D (2013) A transcriptome-wide atlas of RNP composition reveals diverse classes of mRNAs and lncRNAs. Cell 154(5): 996-1009

Tucker M, Valencia-Sanchez MA, Staples RR, Chen J, Denis CL, Parker R (2001) The transcription factor associated Ccr4 and Caf1 proteins are components of the major cytoplasmic mRNA deadenylase in Saccharomyces cerevisiae. Cell 104(3): 377-386

Udagawa T, Swanger SA, Takeuchi K, Kim JH, Nalavadi V, Shin J, Lorenz LJ, Zukin RS, Bassell GJ, Richter JD (2012) Bidirectional control of mRNA translation and synaptic plasticity by the cytoplasmic polyadenylation complex. Mol Cell 47(2): 253-266

Ule J, Jensen K, Mele A, Darnell RB (2005) CLIP: a method for identifying protein- RNA interaction sites in living cells. Methods 37(4): 376-386 van Hoof A, Frischmeyer PA, Dietz HC, Parker R (2002) Exosome-mediated recog-

60 Dmytro Ustianenko References nition and degradation of mRNAs lacking a termination codon. Science 295(5563): 2262-2264

Vanacova S, Wolf J, Martin G, Blank D, Dettwiler S, Friedlein A, Langen H, Keith G, Keller W (2005) A new yeast poly(A) polymerase complex involved in RNA quality control. PLoS Biol 3(6): e189

Vankan P, McGuigan C, Mattaj IW (1992) Roles of U4 and U6 snRNAs in the assembly of splicing complexes. EMBO J 11(1): 335-343

Vasiljeva L, Buratowski S (2006) Nrd1 interacts with the nuclear exosome for 3' processing of RNA polymerase II transcripts. Mol Cell 21(2): 239-248

Verrotti AC, Thompson SR, Wreden C, Strickland S, Wickens M (1996) Evolutionary conservation of sequence elements controlling cytoplasmic polyadenylylation. Proc Natl Acad Sci U S A 93(17): 9027-9032

Wang L, Eckmann CR, Kadyk LC, Wickens M, Kimble J (2002a) A regulatory cytoplasmic poly(A) polymerase in Caenorhabditis elegans. Nature 419(6904): 312-316

Wang SW, Toda T, MacCallum R, Harris AL, Norbury C (2000) Cid1, a fission yeast protein required for S-M checkpoint control when DNA polymerase delta or epsilon is inactivated. Mol Cell Biol 20(9): 3234-3244

Wang Z, Castano IB, Adams C, Vu C, Fitzhugh D, Christman MF (2002b) Structure/ function analysis of the Saccharomyces cerevisiae Trf4/Pol sigma DNA polymerase. Genetics 160(2): 381-391

Welch EM, Jacobson A (1999) An internal open reading frame triggers nonsense- mediated decay of the yeast SPT10 mRNA. EMBO J 18(21): 6134-6145 Wightman B, Ha I, Ruvkun G (1993) Posttranscriptional regulation of the heterochronic gene lin-14 by lin-4 mediates temporal pattern formation in C. elegans. Cell 75(5): 855-862

Wilbert ML, Huelga SC, Kapeli K, Stark TJ, Liang TY, Chen SX, Yan BY, Nathanson

61 References Dmytro Ustianenko

JL, Hutt KR, Lovci MT, Kazan H, Vu AQ, Massirer KB, Morris Q, Hoon S, Yeo GW (2012) LIN28 binds messenger RNAs at GGAGA motifs and regulates splicing factor abundance. Mol Cell 48(2): 195-206

Win TZ, Draper S, Read RL, Pearce J, Norbury CJ, Wang SW (2006) Requirement of fission yeast Cid14 in polyadenylation of rRNAs. Mol Cell Biol 26(5): 1710-1721

Wyers F, Rougemaille M, Badis G, Rousselle JC, Dufour ME, Boulay J, Regnault B, Devaux F, Namane A, Seraphin B, Libri D, Jacquier A (2005) Cryptic pol II transcripts are degraded by a nuclear quality control pathway involving a new poly(A) polymerase. Cell 121(5): 725-737

Yang XC, Purdy M, Marzluff WF, Dominski Z (2006) Characterization of 3'hExo, a 3' exonuclease specifically interacting with the 3' end of histone mRNA. J Biol Chem 281(41): 30447-30454

Yang Z, Jakymiw A, Wood MR, Eystathioy T, Rubin RL, Fritzler MJ, Chan EK (2004) GW182 is critical for the stability of GW bodies expressed during the cell cycle and cell proliferation. J Cell Sci 117(Pt 23): 5567-5578

Yates LA, Fleurdepine S, Rissland OS, De Colibus L, Harlos K, Norbury CJ, Gilbert RJ (2012) Structural basis for the activity of a cytoplasmic RNA terminal uridylyl transferase. Nat Struct Mol Biol 19(8): 782-787

Yu J, Vodyanik MA, Smuga-Otto K, Antosiewicz-Bourget J, Frane JL, Tian S, Nie J, Jonsdottir GA, Ruotti V, Stewart R, Slukvin, II, Thomson JA (2007) Induced pluripotent stem cell lines derived from human somatic cells. Science 318(5858): 1917-1920

Mammalian DIS3L2 exoribonuclease targets the uridylated precursors of let-7 miRNA’s.

Ustianenko D, Hrossova D, Potesil D, Chalupnikova K, Hrazdilova K, Pachernik J, Cet- kovska K, Uldrijan S, Zdrahal Z, Vanacova S

RNA, 2013 Dec;19(12):1632-8.

Author’s contribution: manuscript preparation, manuscript revision, experimental work connected with Figures 2, 3, 4, S1, S2, S3, S4.

Downloaded from rnajournal.cshlp.org on October 20, 2013 - Published by Cold Spring Harbor Laboratory Press

Mammalian DIS3L2 exoribonuclease targets the uridylated precursors of let-7 miRNAs

Dmytro Ustianenko, Dominika Hrossova, David Potesil, et al.

RNA published online October 18, 2013

Supplemental http://rnajournal.cshlp.org/content/suppl/2013/09/30/rna.040055.113.DC1.html Material

Open Access Freely available online through the RNA Open Access option.

Creative This article, published in RNA, is available under a Creative Commons License Commons (Attribution-NonCommercial 3.0 Unported), as described at License http://creativecommons.org/licenses/by-nc/3.0/.

Email Alerting Receive free email alerts when new articles cite this article - sign up in the box at the top Service right corner of the article or click here.

To subscribe to RNA go to: http://rnajournal.cshlp.org/subscriptions

Mammalian DIS3L2 exoribonuclease targets the uridylated precursors of let-7 miRNAs

DMYTRO USTIANENKO,1 DOMINIKA HROSSOVA,1 DAVID POTESIL,1 KATERINA CHALUPNIKOVA,1 KRISTYNA HRAZDILOVA,1 JIRI PACHERNIK,2 KATERINA CETKOVSKA,3 STJEPAN ULDRIJAN,3 ZBYNEK ZDRAHAL,1 and STEPANKA VANACOVA1,4 1CEITEC-Central European Institute of Technology, Masaryk University, 625 00, Brno, Czech Republic 2Department of Experimental Biology, Faculty of Science, Masaryk University, 611 37, Brno, Czech Republic 3Department of Biology, Faculty of Medicine, Masaryk University, 625 00, Brno, Czech Republic

ABSTRACT The mechanisms of gene expression regulation by miRNAs have been extensively studied. However, the regulation of miRNA function and decay has long remained enigmatic. Only recently, 3′ uridylation via LIN28A-TUT4/7 has been recognized as an essential component controlling the biogenesis of let-7 miRNAs in stem cells. Although uridylation has been generally implicated in miRNA degradation, the nuclease responsible has remained unknown. Here, we identify the Perlman syndrome- associated protein DIS3L2 as an oligo(U)-binding and processing exoribonuclease that specifically targets uridylated pre-let-7 in vivo. This study establishes DIS3L2 as the missing component of the LIN28-TUT4/7-DIS3L2 pathway required for the repression of let-7 in pluripotent cells. Keywords: DIS3L2; RNA degradation; RNA uridylation; let-7 miRNA

INTRODUCTION to take place predominantly on the post-transcriptional level: The RNA binding protein LIN28A specifically binds the pre- RNA processing and stability play key roles in the regulation let-7 family of miRNAs and, in an RNA-dependent manner, of gene expression and have impacts on complex cellular pro- recruits the terminal uridyltransferases TUT4 and TUT7, cesses including cell growth, proliferation, and differentia- which add a stretch of uridines (Heo et al. 2008, 2009; tion. RNA stability is regulated by a dynamic interplay of Piskounova et al. 2008; Hagan et al. 2009; Thornton et al. post-transcriptional modifications and trans-acting protein 2012). In turn, this modification inhibits Dicer-mediated or RNA molecules that can be stabilizing as well as destabiliz- processing of pre-let-7 miRNAs. In contrast to its role in em- ing factors. MicroRNAs (miRNAs) were identified as post- bryonic stem cells, the monouridylation of pre-let-7 cata- transcriptional regulators of gene expression more than two lyzed by TUT2/4/7 proteins enhances Dicer processing of decades ago. MicroRNAs are essential for normal develop- certain types of miRNAs in somatic cells (Heo et al. 2012). ment and overall cellular physiology, among other functions. Single UMP addition repairs the 3 ends of some pre- Dysregulation of miRNA expression can be detrimental ′ miRNAs, forming a dinucleotide overhang that is preferred and is often associated with human disease (Croce 2009). Al- by Dicer (Zhang et al. 2004). though miRNA transcription, processing, and function have In mammals, nontemplated uridylation also plays a role in been studied in great detail, the mechanisms of regulation the processing of other small RNAs, such as U6 small nuclear and turnover of miRNAs remain unknown. RNA (snRNA) (Trippe et al. 2006), and in the degradation of Recently, the post-transcriptional control of let-7 biogen- replication-dependent histone mRNAs (Schmidt et al. 2011; esis in early embryogenesis was revealed. Let-7 miRNA is Minasaki and Eckmann 2012; Su et al. 2013). Intriguingly, re- one of the key regulators of embryonic cell differentiation, cent transcriptome-wide studies revealed extensive uridyla- and its expression must be tightly controlled because it tar- tion of a wide spectrum of RNAs (Rissland et al. 2007; gets mRNAs encoding factors required for the maintenance Schmidt and Norbury 2010; Choi et al. 2012), suggesting a of pluripotency (Bussing et al. 2008). This regulation appears general role for (U)-tailing in RNA metabolism and stability.

4Corresponding author E-mail [email protected] Article published online ahead of print. Article and publication date are at © 2013 Ustianenko et al. This article, published in RNA, is available under a http://www.rnajournal.org/cgi/doi/10.1261/rna.040055.113. Freely available Creative Commons License (Attribution-NonCommercial 3.0 Unported), as online through the RNA Open Access option. described at http://creativecommons.org/licenses/by-nc/3.0/.

RNA 19:1–7; Published by Cold Spring Harbor Laboratory Press for the RNA Society 1 Ustianenko et al.

Although the molecular mechanism of LIN28A-TUT-me- Skiv2l2/mMtr4 RNA helicase, which is a component of the diated oligouridylation has been studied in great detail (Yeom yeast and human exosome targeting complexes TRAMP et al. 2011), the fate of the uridylated molecules remains enig- and NEXT (LaCava et al. 2005; Vanacova et al. 2005; Lubas matic. Oligouridylation of pre-let-7 has been generally as- et al. 2011), specifically coprecipitated with U30 RNA from sumed to trigger its degradation (Hagan et al. 2009; Heo both cellular compartments (Fig. 1B,C; Supplemental Tables et al. 2009); however, the downstream-acting factors, such S1, S2). However, the most intriguing finding was the identi- as the specific nucleases responsible for this activity, remain fication of a putative exoribonuclease, Dis3l2, in the U30 sam- unknown. In this work, we have searched for factors involved ple from the cytoplasmic fraction (Fig. 1C; Supplemental in the downstream processing or degradation of uridylated Table S1). To further test the oligo(U)-specific binding of RNAs, and we have identified mammalian DIS3L2 as an DIS3L2, we have performed filter-aided sample preparation oligo(U)-binding exonuclease that specifically targets uridy- of IP eluates obtained with random and U30 RNAs, and we lated let-7 miRNA precursors in vivo. This finding establishes analyzed the resulting peptide mixtures using LC-MS/MS. DIS3L2 as the missing component of the LIN28-TUT4/ The quantification of label-free DIS3L2 showed more than 7-DIS3L2 pathway, which is required for the regulation of sevenfold enrichment of DIS3L2 (two biological replicates) let-7 expression in pluripotent cells. over the random RNA sample (Supplemental Fig. S1A,B; Supplemental Table S2). DIS3L2 protein has gained attention only recently due to its association with the Perlman syn- RESULTS AND DISCUSSION drome in humans (Astuti et al. 2012). Although it has been reported that disruption of the DIS3L2 gene can lead to an- Identification of Dis3l2 as an oligo(U)-binding euploidy, mitotic errors, and expression changes in several nuclease in mouse embryonic stem cells mitosis-related proteins (Astuti et al. 2012), the molecular mechanism of DIS3L2 function has remained elusive. To identify factors involved in the regulation of uridylated RNAs, we carried out affinity purifications of nuclear and cytoplasmic fractions from mouse embryonic stem cells Human DIS3L2 is an oligo(U)-binding, cytoplasmic, (mESC) with biotin-labeled U RNA and a nonspecific 30 Mg++-dependent exoribonuclease that does not 30-nt RNA control (Fig. 1A). Analyses of both the nuclear associate with exosomes and cytoplasmic U30-precipitated proteins revealed several proteins that were previously reported to specifically inter- DIS3L2 is a member of the RNase II family of enzymes act with (mostly internal) (U)-rich motifs, such as ELAV1 and is a sequence homolog of the main catalytic subunit (Kim et al. 2012), TIAR (Kim et al. 2012), hnRNP C of yeast and mammalian exosomes (Rrp44p in yeast, DIS3 (Soltaninassab et al. 1998), and La protein (Stefano 1984; and DIS3L in humans) (Fig. 2A; Dziembowski et al. 2007; Fig. 1B,C; Supplemental Tables S1, S2). When searching Schneider et al. 2007; Tomecki et al. 2010). In mice, DIS3L2 for factors involved in RNA degradation, we found that the is expressed in at least two isoforms. We have identified

ABNucleus C Cytoplasm

CtrlkDa U30 Ctrl kDa U30 mESC nuclear/cytoplasmic pyruvate carboxylase fractionation 250 250 Skiv2l2 Skiv2l2 GTP-binding protein SYF1 100 Dis3l2 matrin-3 100 hnRNP U + biotinylated RNA 75 PABP-1 FUSE 75 nucleolin on streptavidin beads Puf60 50 target of EGR1 protein 1 lupus La prot homol U2af65 50 lupus La prot homol TIAR 37 ELAV1 300 mM KCL hnRNP D Y box protein 1 0.01% NP40 myelin regulatory factor 37 hnRNP C1/C2 ELAV2 Raly annexin A2 hnRNP A/B ELAV1 25 AUF1 hnRNP C1/C2 SDS buffer elution Raly 25 20 Rbm7 hnRNP A/B 20 U2af1 SDS-PAGE AUF1 Mass spectrometry

FIGURE 1. Identification of Dis3l2 as an oligo(U)-binding nuclease in mouse embryonic stem cells (mESCs). (A) Schematic overview of the protocol. (B) Proteins identified in control and U30 RNA-bound nuclear extracts. (C) Proteins identified in control and U30 RNA-bound cytoplasmic extracts. Equal amounts of eluates from control RNA (Ctrl) and U30 RNA bait fractions were separated on 12% SDS-PAGE gels and silver stained. Proteins identified only in the U30 RNA sample are indicated on the right.

2 RNA, Vol. 19, No. 12 DIS3L2 in pre-let-7 degradation

we demonstrate that DIS3L2 is localized A PIN CSD1 CSD2 RNB S1 yRrp44 in the cytoplasm of mouse ESCs (Fig. 1C) as well as HeLa and HEK293 cells hDIS3 (Fig. 2B; Supplemental Fig. S2A). This hDIS3L finding is in agreement with recent re- 1 49 137 228 329 775 885 hDIS3L2 ports indicating the cytoplasmic localiza- CSD1 CSD2 RNB S1 tion of human and fission yeast DIS3L2 (Astuti et al. 2012; Lubas et al. 2013; B α-GFP α-DIS3L2 Merge + DAPI C Malecki et al. 2013). In contrast to DIS3 HeLa Input IP and DIS3L, DIS3L2 lacks the PIN domain (Fig. 2A), which is essential for Control

CTRL DIS3 DIS3L DIS3L2 CTRL DIS3 DIS3L DIS3L2 interaction with the exosome core (Dzie- HEK293T EGFP DIS3L2-EGFP α-FLAG mbowski et al. 2007; Lebreton et al. 2008;

α-RRP40 Schneider et al. 2009). Here, we showed DIS3L2-EGFP GFP-DIS3L2

HeLa DIS3L2 α-RRP41 that DIS3L2 did not coprecipitate with exosome subunits from HEK293T cells, E whereas these subunits were detected in MgCl2 MnCl2 purified control samples of DIS3 and np EDTA 0.25 0.5 1 5 10 0.25 0.5 1 5 10 mM DIS3L (Fig. 2C). This finding indicates U30 RNA the functional independence of DIS3L2 D rDIS3L2,nM from the exosome core complex. To fur- np 0.5 3200

30 ther explore the role of DIS3L2 in RNA uridylation, we evaluated the binding and activity of wild-type (WT) and cata- DIS3L2-U

dp lytically inactive (D391N) recombinant

RNA DIS3L2 (rDIS3L2) (Supplemental Fig. 30 U S2B) toward U30 RNA. The recombinant protein showed nanomolar affinity for U30 RNA in vitro (Fig. 2D), required divalent metal ions for exonuclease activ- FIGURE 2. DIS3L2 is a cytoplasmic, oligo(U)-binding, Mg++-dependent exoribonuclease that ity (Fig. 2E), and degraded U30 RNA to 2- does not associate with exosomes. (A) Schematic representation of the domain organization of DIS3L2 homologs from Saccharomyces cerevisiae and Homo sapiens. The PIN domain is shown to 4-nt end-products (Fig. 2E; Supple- in green, the CSD1 and CSD2 RNA binding domains are in orange, the RNB ribonuclease domain mental Fig. S2C,D). No degradation is in blue, and the S1 domain is in pink. (B) Immunofluorescence staining of HEK293T and HeLa was observed with the D391N mutant cells transfected with either DIS3L2 C-terminally fused to an EGFP (DIS3L2-EGFP) or an empty (Supplemental Fig. S2C). EGFP-expressing plasmid (EGFP). DAPI was used to visualize nuclei. The scale bar corresponds to 10 μm. The Western blot on the right shows the relative abundance of endogenous and EGFP- tagged DIS3L2 in transfected HeLa cells as detected with a DIS3L2-specific antibody. (C) DIS3L2 does not interact with core exosome components. FLAG-tagged DIS3, DIS3L, and DIS3L2 were Human DIS3L2 targets 3′-uridylated immunoprecipitated from stable cell lines inducibly overexpressing the individual proteins. The precursors of let-7 miRNA composition of the IP samples was analyzed by Western blot with the indicated antibodies. (D) Because we initially pulled down DIS3L2 DIS3L2 binds U30 RNA with nanomolar affinity. Electromobility shift assay with recombinant 32 DIS3L2 and 5′ end- P-labeled U30 RNA. The migration patterns of free RNA and protein- from embryonic stem cells, we next asked RNA complexes are indicated. (E) The catalytic activity of DIS3L2 requires divalent metal cations. whether DIS3L2 could be the missing A degradation assay using recombinant DIS3L2 with U30 RNA as a substrate was performed in the component of the pre-let-7 uridylation/ presence of different divalent metal ions as indicated. (EDTA) Reaction mixture containing 5 mM EDTA; (np) control reaction without the addition of any protein. degradation pathway. To test the ability of DIS3L2 to bind pre-let-7 miRNA, we ectopically expressed pri-let-7 miRNA isoform 2 in our U30 RNA samples (Supplemental Tables S1, and LIN28A in HEK293T-Rex cell lines stably expressing ei- S2). Interestingly, isoform 2 was previously identified in the ther the WT or D391N forms of FLAG-DIS3L2 to promote mouse embryo transcriptome (Diez-Roux et al. 2011). To pre-let-7 uridylation according to Heo et al. 2008. The in further study the biochemistry and function of this protein, vivo association between pre-let-7 and DIS3L2 was moni- we have subcloned the human DIS3L2 isoform 1, which is tored by RNA immunoprecipitation (RIP) followed by the closest homolog of mouse DIS3L2 isoform 2. Northern blot analyses. Mature let-7 miRNA coprecipitated Human DIS3 and DIS3L show distinct subcellular locali- only with AGO2, which was used as a positive control for zation patterns; DIS3 is primarily nuclear, whereas DIS3L is let-7 interaction (Fig. 3A). Intriguingly, the D391N mutant cytoplasmic (Staals et al. 2010; Tomecki et al. 2010). Here, of DIS3L2 coprecipitated with slower migrating forms of

www.rnajournal.org 3 Ustianenko et al.

A IP-FLAG RNA input B DIS3L2 DIS3L2

let-7 Empty Ago2 D391N Empty Ago2 D391N WT nt WT 150 TTT TT 12 TTT TT 80 10 pre-let-7a+U TTT TT 50 pre-let7a TTT TT 8 TTT TTT TTT TTT TT 6 TTT TTT TT mature let-7a TTT TTT TT 4 TTT TTT TT 2 EtBr TTT TTT TT No of clones sequenced miR-30 TTT TTT TT 0 pre-let-7a GTT TTT TC 5 6 8 9 1112141517 150 TCT TTT TT U-tail length 80 TTT TTT TT TTT TTT TT 50 TTT TTT TT TTT TTT TCT TTT TTT ATT TT mature miR-30 TTT TTT TTT AT TTC TTT TTT TT EtBr TTT CTT TTT TT miR-16 TTT TTT TTC TT TTT TTT TCT TTT 150 TTT TCT TTT TTT 80 pre-miR-16 TTC TTT ATT TTT 50 TTT TTT TTA TTT TT TTT TTT TTA TTT TT TTT TTT TTT TTT TTC mature miR-16 TTT TTT TTC TTT TTT TT TTT TTT TT pre-let-7i EtBr TTT TTT TTA TTT TT

DIS3L2 DIS3L2 Empty D391N Ago2 Empty D391N Ago2 WT WT α-FLAG LIN28-Myc Elution Input C D WT FLAG-DIS3L2 D391N WT FLAG-DIS3L2

labelled RNA U2 U4 U8 A8 U8 A8 U30 A30 competitor RNA Time, min. 0 5 15 30 0 5 15 30 0 5 15 30 0 5 15 30 0 5 15 30 30 0 5 15 30 0 5 15 30 0 5 15 30 0 5 15 30

FIGURE 3. DIS3L2 targets 3′ end-uridylated precursors of let-7 miRNA. (A) Mutant DIS3L2 (D391N) specifically coprecipitates with extended forms of pre-let-7a miRNA. Northern blot analyses of RNAs coimmunoprecipitated with FLAG-tagged proteins as indicated on the upper side of the autoradiograph (IP-FLAG). The RNA input was total RNA isolated from whole cell lysates. The ethidium bromide (EtBr)-stained gels represent input loading controls. (Empty) Control RIP performed using cells that were not expressing any FLAG-tagged protein. The lower panel shows Western blot analysis of Myc-tagged LIN28A expression and the efficiency of FLAG tag-mediated immunoprecipitation of proteins used for RIP analyses. “Input” shows the whole cell lysates used for the IPs. The ectopic expression of Myc-tagged LIN28A was monitored with specific anti-Myc antibodies. DIS3L2 was detected with anti-FLAG antibodies. (B) Sequencing analysis of the 3′ termini of pre-let-7 RNAs coprecipitated with the D391N DIS3L2 mutant in HEK293T-Rex cells ectopically overexpressing LIN28A and pri-let-7a-1. The analysis suggests that the DIS3L2 mutant binds to uridylated pre-let-7a-1 and pre-let-7i because 31 out of 35 sequenced clones contained nontemplate oligo(U)-tails. The gray box schematically represents the body of pre-let-7 miRNAs. A graphical representation of the (U)-extension statistics is shown at the right.(C) Oligouridylation stimulates DIS3L2- mediated degradation of pre-let-7 miRNA. An in vitro degradation assay with FLAG-DIS3L2 purified from HEK293T-Rex cells and pre-let-7 RNA (represented by the schematic stem–loop) containing either no 3′ extension or 3′ oligo(U) and oligo(A) extensions of increasing lengths (indicated at the top) as substrates. D391N indicates a reaction with purified catalytically inactive DIS3L2. (D) DIS3L2 exhibits a preference for RNA substrates with oligo(U) extensions. A 5′ end-labeled pre-let-7-U8 RNA was incubated with purified DIS3L2 in the presence of a 10-fold excess of different unlabeled competitor RNAs (indicated at the top) for the time periods shown. pre-let-7 miRNA (pre-let-7+U) (Fig. 3A). This binding was tion pattern of the extended form of pre-let-7 miRNA resem- specific to pre-let-7 because no signal was observed with bled that of the uridylated pre-let7 miRNAs induced in probes specific for miR-30 or miR-16 (Fig. 3A). The migra- HEK293 cells upon LIN28A overexpression (Heo et al.

4 RNA, Vol. 19, No. 12 DIS3L2 in pre-let-7 degradation

2008). To uncover the identity of D391N-bound fragments, In summary, we have demonstrated that the cytoplasmic we modified the immunoprecipitated RNAs with 3′ RNA exoribonuclease DIS3L2 specifically recognizes uridylated linkers to allow for cDNA synthesis using linker-specific pre-let-7 miRNAs in vivo and that pre-let-7 uridylation is im- primers. The subsequent PCR amplification with pre-let-7- portant for DIS3L2-mediated degradation. This establishes specific forward and linker-specific reverse primers revealed DIS3L2 as a strong candidate for the sought-after nuclease a product of ∼70 nt in length in D391N-bound RNAs but targeting uridylated pre-let-7 in embryonic stem cells (Fig. not in background controls from untransfected HEK293T- 4). Because the TUT-DIS3L2 mechanism strongly resembles Rex cells (Supplemental Fig. S3A). The PCR products were the nuclear polyadenylation-mediated RNA degradation subcloned, and 35 clones were sequenced. Sequencing anal- pathway (Kadaba et al. 2004; LaCava et al. 2005; Vanacova ysis confirmed that the bound fragments corresponded to et al. 2005), future studies should examine to what extent it pre-let-7 miRNAs (Supplemental Fig. S3B). Most important- operates as a general cytoplasmic (mi)RNA surveillance ly, 31 out of the 35 sequenced clones displayed nontemplate and decay pathway. uridine stretches at their 3′ ends (Fig. 3B). These oligo(U) ex- Over the course of preparing this manuscript, two other tensions ranged between five and 17 nt in length (Fig. 3B), groups reported their findings on the role of DIS3L2 in the which strongly correlates with the length distribution of degradation of uridylated RNAs. Malecki et al. identified oligo(U) tails previously identified on pre-let-7a-1 in DIS3L2 as an oligo(U)-specific enzyme targeting uridylated HEK293T cells ectopically expressing LIN28A and pri-let- mRNAs in Schizosaccharomyces pombe (Malecki et al. 7a-1 (Heo et al. 2008). To test the importance of oligouridy- 2013). Chang et al. revealed the role of mouse Dis3l2 in the lation for pre-let-7 degradation, we performed in vitro degra- LIN28A-TUT4/7 pathway (Chang et al. 2013). These authors dation assays with purified FLAG-DIS3L2. We observed showed the in vitro activity of Dis3l2 toward uridylated that 4- to 8-nt oligo(U) extensions significantly enhanced pre-let-7 miRNAs and demonstrated that down-regulation pre-let-7 miRNA degradation by DIS3L2 in vitro (Fig. 3C; of Dis3l2 in mESCs can lead to elevated levels of uridylated Supplemental Fig. S3C). Because the oligo(A8) modification pre-let-7 without affecting the level of mature let-7 also activated degradation by DIS3L2, we examined DIS3L2 miRNA. Because we observed down-regulation of mature nucleotide preference by using unlabeled RNA competitors. let-7 upon DIS3L2 knockdown, it is possible that DIS3L2 We demonstrated that DIS3L2 prefers oligo(U) over oligo (A), as only uridylated pre-let-7 and oligo(U) RNAs were able to inhibit the degradation of labeled uridylated pre-let- 7 RNA substrates (Fig. 3D; Supplemental Fig. S3D). The lack of pre-let-7+U coprecipitation with WT DIS3L2 pri-miRNA was likely due to high turnover of the bound substrate in vivo. Drosha/DGCR8 Importantly, overexpression of WT DIS3L2 caused a reduction in the uridylated pre-let-7 miRNA level compared with that in untransfected cells or cells overexpressing the EXP 5 Nucleus D391N mutant (Fig. 3A, pre-let-7a+U, RNA input). Interest- ingly, D391N was also able to pull down pre-let-7U+ frag- Cytoplasm ments without the ectopic expression of LIN28A or pri-let- pre-miRNA 7a-1 (Supplemental Fig. S4A). This result suggests that either Dicer Lin28A / TUT4/TUT7 low endogenous levels of LIN28A (Heo et al. 2012) or other UUUUU TUT cofactor(s) are sufficient to promote efficient pre-let-7 DIS3L2 oligouridylation in epithelial cells. We, therefore, examined mature miRNA whether DIS3L2 targets pre-let-7 in HeLa cells that exhibit UUU higher endogenous levels of let-7 miRNA expression. Northern blot analysis of RNAs isolated from cells treated AGO/RISC with two different sets of siRNAs did not indicate pre-let-7 UC U U A A A G stabilization. On the contrary, Northern blot and Q-PCR A UC G C UG G analyses revealed reduced levels of mature let-7 and, less U C significantly, miR-30 miRNAs (Supplemental Fig. S4B,C). This miRNA down-regulation was not due to altered levels FIGURE 4. A model for the regulation of let-7 miRNA biogenesis via of Dicer (Supplemental Fig. S4D). Currently, the connection the LIN28A-TUT4/7-DIS3L2 pathway. The primary transcripts (pri- between DIS3L2 down-regulation and let-7 reduction in let-7) are processed by the microprocessor complex in the nucleus to somatic cells is unknown. Future studies will reveal whether pre-let-7, which is exported to the cytoplasm. In the cytoplasm, pre- miRNA dysregulation may contribute to the development of let-7 is either processed by Dicer to mature let-7 miRNA, or it is oligouridylated in the presence of LIN28A. The oligo(U) tails are then recog- Perlman syndrome or increased tumor incidence in individ- nized by the DIS3L2 exonuclease, which initiates the degradation of uals bearing the mutant DIS3L2 allele (Astuti et al. 2012). RNA in the 3′ to 5′ direction.

www.rnajournal.org 5 Ustianenko et al. has two distinct cell type-specific mechanisms operating dur- probe, the membrane was prehybridized at 42°C for 2 h. The radio- ing let-7 biogenesis. active signal was monitored using an FLA-9000 phosphorimager (FujiFilm).

MATERIALS AND METHODS SUPPLEMENTAL MATERIAL LC-MS/MS analysis Supplemental material is available for this article. Thin slices excised from a stained SDS-PAGE gel were de-stained, washed, and incubated with trypsin. Liquid chromatography–tandem mass spectrometry (LC–MS/MS) analysis was performed using ACKNOWLEDGMENTS the EASY-nLC system (Thermo Fisher Scientific) coupled with an We thank Ger Pruijn and Petr Svoboda for the hRRP41, hRRP40, and HCTultra PTM Discovery System ion trap mass spectrometer Dicer antibodies; Torben Jensen for the HEK293T-Rex cell lines ex- (Bruker Daltonik). The MASCOT 2.3.02 (MatrixScience) search en- pressing FLAG-DIS3 and FLAG-DIS3L; and Narry Kim for the pri- gine was used for processing the MS and MS/MS data. Database miRNA DNA constructs. We also thank Leona Svajdova for excellent searches were performed against the NCBI database (nonredundant, technical support and Ivana Horvathova for providing the LIN28A taxonomy: Mus musculus). (For details, see Supplemental Material.) construct. This work was supported by the Wellcome Trust (084316/Z/07/Z to S.V.), the Czech Science Foundation (305/11/ 1095 to S.V., D.H. was supported by P305/12/G034, and P206-12- siRNA-mediated knock down of DIS3L2 G151 to D.P. and Z.Z.), and the CEITEC-Central European On the day prior to transfection, 1.7 × 105 cells were seeded. siRNAs Institute of Technology (CZ.1.05/1.1.00/02.0068) from the were transfected using INTERFERin transfection reagent (Polyplus European Regional Development Fund. transfections) at a 20-nM final concentration, following the manu- Author contributions: D.U. designed, performed, and analyzed facturer’s instructions. The siRNA treatment was repeated after 24 h, most of the experiments; D.H. identified DIS3L2 using RNA- and the cells were collected for further analysis on the following day. based protein precipitation; J.P. provided the mESC material; K. (See Supplemental Table S4 for the list of siRNAs used.) Chalupnikova performed immunofluorescence studies; K.H. prepared DIS3L2 constructs and specific antibodies; K. Cetkovska and S.U. assisted with the initial HEK293T manipulations; and D.P. In vitro degradation assay and Z.Z. performed the MS analysis. D.U. and S.U. were also involved in the preparation of the manuscript. S.V. designed the project and In vitro degradation assays were performed in 10-μL reaction vol- experiments, provided funding, supervised students, and prepared umes containing 10 mM Tris (pH 8.0), 50 mM KCl, 5 mM the manuscript. MgCl2, and 10 mM DTT (modified from Lorentzen et al. 2008; Staals et al. 2010; Tomecki et al. 2010). Typically, 150 nM of purified Received May 8, 2013; accepted August 27, 2013. recombinant protein and 20 pmol of 5′ end-labeled RNA substrate were incubated at 37°C for the time periods indicated. The reactions were terminated with one volume of formamide loading buffer REFERENCES (80% formamide, 0.1% bromphenol blue, 0.1% xylene cyanol, 5 Astuti D, Morris MR, Cooper WN, Staals RH, Wake NC, Fews GA, mM EDTA). The reactions were resolved on denaturing 20% poly- Gill H, Gentle D, Shuib S, Ricketts CJ, et al. 2012. Germline muta- acrylamide gels containing 8 M urea. The radioactively labeled sub- tions in DIS3L2 cause the Perlman syndrome of overgrowth and strate was detected using a phosphorimaging screen and an FLA- Wilms tumor susceptibility. Nat Genet 44: 277–284. 9000 phosphorimager (FujiFilm). Bussing I, Slack FJ, Grosshans H. 2008. let-7 microRNAs in development, stem cells and cancer. Trends Mol Med 14: 400–409. Chang HM, Triboulet R, Thornton JE, Gregory RI. 2013. A role for the RNA binding assay Perlman syndrome exonuclease Dis3l2 in the Lin28-let-7 pathway. Nature 497: 244–248. The binding reactions were performed in 10-μL volumes. Choi YS, Patena W, Leavitt AD, McManus MT. 2012. Widespread RNA 3 -end oligouridylation in mammals. RNA 18: 394–401. Recombinant DIS3L2 was incubated with 5 end-labeled U RNA ′ ′ 30 Croce CM. 2009. Causes and consequences of microRNA dysregulation for 20 min at room temperature in binding buffer (50 mM KCl, in cancer. Nat Rev Genet 10: 704–714. 10 mM Tris pH 8.0, 10 mM DTT, 10% glycerol, 0.1 μg of BSA). Diez-Roux G, Banfi S, Sultan M, Geffers L, Anand S, Rozado D, Proteins and RNAs were resolved on native 8% polyacrylamide Magen A, Canidio E, Pagani M, Peluso I, et al. 2011. A high-resolu- gels at 100 V for ∼8 h and visualized by phosphorimaging. tion anatomical atlas of the transcriptome in the mouse embryo. PLoS Biol 9: e1000582. Dziembowski A, Lorentzen E, Conti E, Seraphin B. 2007. A single sub- Northern blot analysis unit, Dis3, is essentially responsible for yeast exosome core activity. Nat Struct Mol Biol 14: 15–22. Total RNA was resolved on a 15% denaturing polyacrylamide gel Hagan JP, Piskounova E, Gregory RI. 2009. Lin28 recruits the TUTase Zcchc11 to inhibit let-7 maturation in mouse embryonic stem cells. and transferred to a Hybond-N+ membrane (GE Healthcare) Nat Struct Mol Biol 16: 1021–1025. by electroblotting (BioRad). Hybridization with radioactively la- Heo I, Joo C, Cho J, Ha M, Han J, Kim VN. 2008. Lin28 mediates the beled oligonucleotides was performed in ULTRAhyb-oligo hybridi- terminal uridylation of let-7 precursor MicroRNA. Mol Cell 32: zation buffer (Ambion) at 42°C. Prior to the addition of the labeled 276–284.

6 RNA, Vol. 19, No. 12 DIS3L2 in pre-let-7 degradation

Heo I, Joo C, Kim YK, Ha M, Yoon MJ, Cho J, Yeom KH, Han J, Schmidt MJ, West S, Norbury CJ. 2011. The human cytoplasmic RNA Kim VN. 2009. TUT4 in concert with Lin28 suppresses microRNA terminal U-transferase ZCCHC11 targets histone mRNAs for degra- biogenesis through pre-microRNA uridylation. Cell 138: 696–708. dation. RNA 17: 39–44. Heo I, Ha M, Lim J, Yoon MJ, Park JE, Kwon SC, Chang H, Kim VN. Schneider C, Anderson JT, Tollervey D. 2007. The exosome subunit 2012. Mono-uridylation of pre-microRNA as a key step in the bio- Rrp44 plays a direct role in RNA substrate recognition. Mol Cell genesis of group II let-7 microRNAs. Cell 151: 521–532. 27: 324–331. Kadaba S, Krueger A, Trice T, Krecic AM, Hinnebusch AG, Anderson J. Schneider C, Leung E, Brown J, Tollervey D. 2009. The N-terminal PIN 2004. Nuclear surveillance and degradation of hypomodified initia- domain of the exosome subunit Rrp44 harbors endonuclease activity tor tRNAMet in S. cerevisiae. Genes Dev 18: 1227–1240. and tethers Rrp44 to the yeast core exosome. Nucleic Acids Res 37: Kim HS, Wilce MC, Yoga YM, Pendini NR, Gunzburg MJ, 1127–1140. Cowieson NP, Wilson GM, Williams BR, Gorospe M, Wilce JA. Soltaninassab SR, McAfee JG, Shahied-Milam L, LeStourgeon WM. 2012. Different modes of interaction by TIAR and HuR with target 1998. Oligonucleotide binding specificities of the hnRNP C protein RNA and DNA. Nucleic Acids Res 39: 1117–1130. tetramer. Nucleic Acids Res 26: 3410–3417. LaCava J, Houseley J, Saveanu C, Petfalski E, Thompson E, Jacquier A, Staals RH, Bronkhorst AW, Schilders G, Slomovic S, Schuster G, Tollervey D. 2005. RNA degradation by the exosome is promoted by Heck AJ, Raijmakers R, Pruijn GJ. 2010. Dis3-like 1: A novel exori- a nuclear polyadenylation complex. Cell 121: 713–724. bonuclease associated with the human exosome. EMBO J 29: Lebreton A, Tomecki R, Dziembowski A, Seraphin B. 2008. 2358–2367. Endonucleolytic RNA cleavage by a eukaryotic exosome. Nature Stefano JE. 1984. Purified lupus antigen La recognizes an oligouridylate 456: 993–996. stretch common to the 3′ termini of RNA polymerase III transcripts. Lorentzen E, Basquin J, Tomecki R, Dziembowski A, Conti E. 2008. Cell 36: 145–154. Structure of the active subunit of the yeast exosome core, Rrp44: Su W, Slepenkov SV, Slevin MK, Lyons SM, Ziemniak M, Kowalska J, Diverse modes of substrate recruitment in the RNase II nuclease Darzynkiewicz E, Jemielity J, Marzluff WF, Rhoads RE. 2013. family. Mol Cell 29: 717–728. mRNAs containing the histone 3′ stem-loop are degraded primarily Lubas M, Christensen MS, Kristiansen MS, Domanski M, Falkenby LG, by decapping mediated by oligouridylation of the 3′ end. RNA 19: Lykke-Andersen S, Andersen JS, Dziembowski A, Jensen TH. 2011. 1–16. Interaction profiling identifies the human nuclear exosome targeting Thornton JE, Chang HM, Piskounova E, Gregory RI. 2012. Lin28-me- complex. Mol Cell 43: 624–637. diated control of let-7 microRNA expression by alternative TUTases Lubas M, Damgaard CK, Tomecki R, Cysewski D, Jensen TH, Zcchc11 (TUT4) and Zcchc6 (TUT7). RNA 18: 1875–1885. Dziembowski A. 2013. Exonuclease hDIS3L2 specifies an exo- Tomecki R, Kristiansen MS, Lykke-Andersen S, Chlebowski A, some-independent 3′-5′ degradation pathway of human cytoplasmic Larsen KM, Szczesny RJ, Drazkowska K, Pastula A, Andersen JS, mRNA. EMBO J 32: 1855–1868. Stepien PP, et al. 2010. The human core exosome interacts with dif- Malecki M, Viegas SC, Carneiro T, Golik P, Dressaire C, Ferreira MG, ferentially localized processive RNases: hDIS3 and hDIS3L. EMBO J Arraiano CM. 2013. The exoribonuclease Dis3L2 defines a novel eu- 29: 2342–2357. karyotic RNA degradation pathway. EMBO J 32: 1842–1854. Trippe R, Guschina E, Hossbach M, Urlaub H, Luhrmann R, Minasaki R, Eckmann CR. 2012. Subcellular specialization of multifac- Benecke BJ. 2006. Identification, cloning, and functional analysis eted 3′end modifying nucleotidyltransferases. Curr Opin Cell Biol 24: of the human U6 snRNA-specific terminal uridylyl transferase. 314–322. RNA 12: 1494–1504. Piskounova E, Viswanathan SR, Janas M, LaPierre RJ, Daley GQ, Sliz P, Vanacova S, Wolf J, Martin G, Blank D, Dettwiler S, Friedlein A, Gregory RI. 2008. Determinants of microRNA processing inhibition Langen H, Keith G, Keller W. 2005. A new yeast poly(A) polymerase by the developmentally regulated RNA-binding protein Lin28. J Biol complex involved in RNA quality control. PLoS Biol 3: e189. Chem 283: 21310–21314. Yeom KH, Heo I, Lee J, Hohng S, Kim VN, Joo C. 2011. Single-molecule Rissland OS, Mikulasova A, Norbury CJ. 2007. Efficient polyuridylation approach to immunoprecipitated protein complexes: Insights into by noncanonical poly(A) polymerases. Mol Cell Biol 27: 3612–3624. miRNA uridylation. EMBO Rep 12: 690–696. Schmidt MJ, Norbury CJ. 2010. Polyadenylation and beyond: Emerging Zhang H, Kolb FA, Jaskiewicz L, Westhof E, Filipowicz W. 2004. Single roles for noncanonical poly(A) polymerases. Wiley Interdiscip Rev processing center models for human Dicer and bacterial RNase III. RNA 1: 142–151. Cell 118: 57–68.

www.rnajournal.org 7

Supplementary data

Mammalian DIS3L2 exoribonuclease targets the uridylated precursors of let-7 miRNAs.

Ustianenko D, Hrossova D, Potesil D, Chalupnikova K, Hrazdilova K, Pachernik J, Cetkovska K, Uldrijan S, Zdrahal Z, Vanacova S

RNA, 2013 Dec;19(12):1632-8. Ustianenko et al. DIS3L2 in pre-let-7 degradation

SUPPLEMENTAL MATERIAL AND METHODS Cell culture and manipulation Undifferentiated mouse ES cells (mESC D3 cells; (Doetschman et al., 1985)) were adapted to feeder-free culture and grown in Dulbecco’s modified Eagle’s medium (DMEM; Gibco-Invitrogen, Carlsbad, CA) supplemented with 15% fetal calf serum (PAA Laboratories GmbH, Pasching, Austria), 100 mM non-essential amino acids (Gibco-Invitrogen), 0.05 mM β-mercaptoethanol (Sigma-Aldrich, St. Louis, MO), 100 U/ml penicillin, 0.1 mg/ml streptomycin (Gibco-Invitrogen), and 1000 U/ml recombinant leukemia inhibitory factor (LIF; Chemicon International, Temecula, CA). Mammalian cells (HEK293T-Rex, HeLa) were maintained in DMEM supplemented with 10% fetal calf serum at 37°C in the presence of 5% CO2. For doxycycline- inducible protein expression; doxycycline was added to culture media at 10 - 50 ng/ml. For transient transfections, cells were grown to 70% confluency, and plasmid DNA was transfected using TURBOFECT (Fermentas) following the manufacturer’s instructions. Protein extracts preparation and cell fractionation Mouse embryonic stem cells (mESC) were washed, re-suspended in buffer containing

10 mM HEPES pH 7.9, 1.5 mM MgCl2, 10 mM KCl, 2 mM DTT and 0.2 mM PMSF and incubated 10 minutes on ice. Cells were broken in a Dounce homogenizer and nuclei were pelleted by centrifugation at 2000 rpm for 15 minutes at 4°C. The supernatant represented the cytoplasmic fraction. Nuclear pellet was then incubated with buffer containing 20 mM HEPES pH 7.9, 25% glycerol, 20 mM KCl, 1.5 mM

MgCl2, 0.2 mM EDTA, 2 mM DTT and 0.2 mM PMSF. Next, the concentration of KCl was increased up to 1.2 M and extracts were incubated for additional 30 minutes at 4°C. The insoluble fraction was removed by centrifugation at 10000 rpm for 1 hour at 4°C. The cytoplasmic and nuclear extracts were subsequently dialyzed to the buffer containing 20 mM HEPES pH 7.9, 20% glycerol, 20 mM KCl, 1.5 mM MgCl2, 0.2 mM EDTA, 2 mM DTT and 0.2 mM PMSF. After dialysis extracts were centrifuged for 30 minutes at 10000 rpm at 4°C, supernatant was aliquoted and frozen in liquid nitrogen. RNA-based protein precipitation

Affinity purifications were performed with biotin-labeled U30 RNA and a random 30 nt RNA. As a control for unspecific binding to RNA we used a random 30nt RNA sequence (5' GAACAUAUUUCACCAACAUUAUACUGUGUC 3'), the expression of which has not been detected in mouse and human cells (our BLAST search). To remove unspecific binders, mESC protein extracts were first pre-cleared using 100 µl of packed Streptavidin agarose (SAg) resins (Thermo Scientific), washed twice with Low salt buffer (LSB) (20 mM HEPES pH 8.0, 100 mM KCl, 10 mM MgCl2, 0.01% NP40, 1 mM DTT). Fresh SAg was pre-blocked in blocking buffer (LSB containing 1 mg/ml RNase-free BSA, 20 µg glycogen, 50 µg yeast total RNA) for 1 hour at 4°C with slow rotation. One milliliter of blocking buffer was used per 100 µl of packed SAg beads. SAg were then washed twice with High salt buffer (HSB) (20 mM HEPES pH8.0, 300 mM KCl,

10 mM MgCl2, 0.01% NP40, 1 mM DTT) and stored as 1:1 slurry (SAg beads : HSB). To prepare SAg-RNA matrix, 40 µl of pre-blocked slurry of SAg beads in HSB was mixed with 5 volumes of HSB containing 10 µg biotinylated U30 RNA oligo (Sigma) and 50U of RNasin Plus RNase inhibitor (Promega) in HSB. The mixture was incubated for 5 hours at 4°C with rotation. SAg beads were collected by 1 minute centrifugation at 1500g and washed 3 times with 1 ml of HSB. For protein precipitation, 150 µl of pre-cleared protein extract was added and incubated for 1 hour at 30°C with rotation. SAg beads were briefly collected by centrifugation at 1500g and washed 3 times with 1 ml of HSB. Bound proteins were eluted with 20 µl of 1x SDS loading buffer. Proteins were separated on 12% polyacrylamide gels. Preparation of stable cell lines Plasmids (pcDNA5 FRT/TO FLAG-DIS3L2, pcDNA5/FRT/TO FLAG-D391N and pcDNA FRT/TO FLAG-AGO2) were transfected to Flp-InTM T-RExTM (Invitrogen) cell line with TURBOFECT reagent and selected for stable insertions according to the manufacturer’s protocol. LC-MS/MS analysis and database searching

The profiles of U30 precipitates from nuclei and cytoplasm and of proteins bound to control RNA were compared by SDS-PAGE, and bands unique to U30 RNA samples were identified by tandem mass spectrometry analysis (MS-MS). 1D gel areas to be analyzed were excised from the corresponding 1D gel lines. After destaining and washing procedures, each gel band was incubated with trypsin. Liquid

chromatography–tandem mass spectrometry (LC–MS/MS) analysis was performed using EASY-nLC system (Proxeon) on-line coupled with an HCTultra PTM Discovery System ion trap mass spectrometer (Bruker Daltonik). Sample volume was 10 µl. Prior to LC separation, tryptic digests were concentrated and desalted using trapping column (100 µm × 30 mm) filled with 4-µm Jupiter Proteo sorbent (Phenomenex, Torrance, CA). After washing with 0.1% formic acid, the peptides were eluted from the trapping column using an acetonitrile/water gradient (350 nL/min) onto a fused-silica capillary column (100 µm x 100 mm), on which peptides were separated. The column was filled with 3.5-µm X-Bridge BEH 130 C18 sorbent (Waters). The mobile phase A consisted of 0.1% formic acid in water and the mobile phase B consisted of 0.1% formic acid in acetonitrile. The gradient elution started at 5% of mobile phase B and increased linearly from 5% to 35% during the first 20 minutes. The gradient linearly increased to 90% of mobile phase B in the next two minutes and remained at this state for next 8 minutes. The analytical column outlet was directly connected to the nanoelectrospray ion source. Nitrogen was used as nebulizing as well as drying gas. The pressure of nebulizing gas was 8 psi. The temperature and flow rate of drying gas were set to 250ºC and 6 L/min, respectively, and the capillary voltage was 4.0 kV. The mass spectrometer was operated in the positive ion mode in m/z range of 300 – 1500 for MS and 100-2500 for MS/MS scans. Two precursor ions per MS spectrum were selected in data dependent manner for further fragmentation. Extraction of the mass spectra from the chromatograms, mass annotation and deconvolution of the mass spectra were performed using DataAnalysis 4.0 software (Bruker Daltonik). MASCOT 2.3.02 (MatrixScience, London, UK) search engine was used for processing the MS and MS/MS data. Database searches were done against the NCBI database, taxonomy Mus musculus (non redundant; downloaded from ftp://ftp.ncbi.nih.gov/blast/db/FASTA/; database version 20120407; 146,059 protein sequences for taxonomy Mus musculus; 17,751,536 protein sequences in total). Mass tolerances of peptides and MS/MS fragments for MS/MS ion searches were 0.5 Da. Oxidation of methionine and carbamidomethylation of cysteine as optional modifications, one enzyme miscleavage and correction for one 13C atom were set for all searches. Peptides with statistically significant peptide score (p < 0.05) were considered. Manual MS/MS spectra assignment validation was done with

identification verification using database search against the whole NCBI database without taxonomy restriction. Cell fractionation for whole-IP LC-MS-MS analyses D3 mESCs (108) were washed twice with 5 ml of ice-cold 1x PBS with protease inhibitor (PI) cocktail (Roche), harvested from dishes and resuspended in 2 ml ice- cold 1x PBS with PI. The buffer was removed by brief centrifugation and cell pellets were gently lyzed in 1 ml of ice-cold lysis buffer (0.1% NP-40, 1x PBS, 1X PI). The cytoplasmic and nuclear fractions were crudely separated by brief centrifugation 10000g for 10s in a table-top microcentrifuge. The supernatants representing the cytoplasmic fraction were collected and pellets representing the nuclear fraction were washed with 1 ml of ice-cold lysis buffer. Nuclei were broken in 1 ml of lysis buffer by sonication. The fractionation efficiency was subsequently monitored by western blot analysis with anti-Histon H1 antibodies for nuclear and anti-Tubulin alpha antibodies for cytoplasmic fractions. The extracts were the used for RNA precipitations as described above. FASP processing Protein pull-downs were processed by filter-aided sample preparation (FASP) method {Wisniewski, 2009 #421}{Wisniewski, 2011 #422}. The whole samples were mixed with 8 M UA buffer (8 M urea in 100 mM Tris-HCl, pH 8.5), loaded onto the Vivacon 500 device with MWCO 10 kDa (Sartorius Stedim Biotech) and centrifuged at 14,000 × g for 30 min at 20°C. The retained proteins were washed with 400 µL UA buffer. The final protein concentrates kept in the Vivacon 500 device were mixed with 100 µL of UA buffer containing 50 mM dithiothreitol and incubated for 30 min. After additional centrifugation, the samples were mixed with 100 µL of UA buffer containing 50 mM iodoacetamide and incubated in the dark for 30 min. After the next centrifugation step, the samples were washed three times with 400 µL UA buffer and three times with 200 µL of 50 mM NaHCO3. Trypsin (sequencing grade, Promega) was added onto the filter and the mixture was incubated for 14 h at 37°C. The tryptic peptides were finally eluted by centrifugation followed by two additional elutions with 50 µL of 50 mM NaHCO3. LC-MS/MS analysis of peptides from FASP LC-MS/MS analyses of peptide mixture were done using RSLCnano system connected to Orbitrap Elite hybrid spectrometer (Thermo Fisher Scientific, Waltham, MA, USA). Prior to LC separation, tryptic digests were online concentrated and

desalted using trapping column (100 µm × 30 mm) filled with 3.5-µm X-Bridge BEH 130 C18 sorbent (Waters, Milford, MA, USA). After washing of trapping column with 0.1% FA, the peptides were eluted (flow 300 nl/min) from the trapping column onto a Acclaim Pepmap100 C18 column (2 µm particles, 75 µm × 250 mm; Thermo Fisher Scientific, Waltham, MA, USA) by the following gradient program (mobile phase A: 0.1% FA in water; mobile phase B: 0.1% FA in acetonitrile): the gradient elution started at 1% of mobile phase B and increased from 1% to 45% during the first 40 min (28% in the 30th and 45% in 40th min), then increased linearly to 95% of mobile phase B in the next 2 min and remained at this state for the next 13 min. Equilibration of the trapping column and the column was done prior to sample injection to the sample loop. The analytical column outlet was directly connected to the Nanospray Flex Ion Source (Thermo Fisher Scientific, Waltham, MA, USA). MS data were acquired in a data-dependent strategy selecting up to top 20 precursors based on the precursor abundance in the survey scan (350-1700 m/z). The resolution of the survey scan was 120 000 (400 m/z) with a target value of 1×106 ions, one microscan and maximum injection time of 200 ms. Low resolution CID MS/MS spectra were acquired with a target value of 10 000 in rapid CID scan mode with m/z range adjusted according to actual precursor mass and charge. MS/MS acquisition in the linear ion trap was carried out in parallel to the survey scan in the Orbitrap analyser by using the preview mode. The maximum injection time for MS/MS was 150 ms. Dynamic exclusion was enabled for 45 s after one MS/MS spectra acquisition and early expiration was disabled. The isolation window for MS/MS fragmentation was set to 2 m/z. The analysis of the mass spectrometric RAW data files was carried out using the Proteome Discoverer software (Thermo Fisher Scientific; version 1.3) with in-house Mascot (Matrixscience, London, UK; version 2.3.1) and Sequest search engines utilisation. Mascot MS/MS ion searches were done against UniProt protein database for mouse (downloaded from ftp://ftp.uniprot.org/pub/databases/uniprot/current_release/knowledgebase/proteomes/ ; version 20130626; 50,818 sequences; 24,421,122 residues) with additional sequences from cRAP database (downloaded from http://www.thegpm.org/crap/). Mass tolerance for peptides and MS/MS fragments were 5 ppm and 0.5 Da, respectively. Oxidation of methionine and deamidation (N, Q) as optional modification, carbamidomethylation of C as fixed modification and two enzyme miss

cleavages were set for all searches. Percolator was used for post-processing of Mascot search results. Peptides with false discovery rate (FDR; q-value) < 1%, rank 1 and with at least 6 amino acids were considered. Label-free quantification using protein area calculation in Proteome Discoverer was used (“top 3 protein quantification” {Silva, 2006 #423}). Two LC-MS/MS analyses in total were done for each sample with the same sample volume. The second LC-MS/MS analysis was performed with exclusion of m/z masses already assigned to peptide from target database (FDR < 1%) based on the first LC-MS/MS analysis. Mass tolerance for m/z exclusion was set to 10 ppm and retention time window to two minutes. The two resulting raw files for each sample were searched as a single data set. The results for U30 RNA and random RNA IP samples were compared and at least five-fold enrichment in U30 RNA sample over random RNA sample was set as a trash hold (labelled as oligo(U)-specific interactor in this experimental set up).

DIS3L2 constructs Constructs for bacterial expression. The coding sequence of human DIS3L2 protein (isoform 1, 885 aminoacids, NP_689596.4) was cloned into pET28b between NheI and BamHI sites allowing expression of N-terminaly fused 6xHis-Smt3 tag. DIS3L2 was amplified by PCR in two parts (because of XhoI internal cleavage site) using cDNA prepared from HEK293 RNA as a template. First, the 3ʹ′ end of CDS (1664 – 2658 nt) was amplified with primers (Forward 5ʹ′ AGCAGCGAGGAGGTACACCAG3ʹ′, Reverse 5ʹ′ ACCTCGAGTCAGCTGGTGCTTGAGTCCTCG 3ʹ′) and subcloned into pET28b using HindIII and XhoI sites, than the 5ʹ′ part (1 – 1663 nt) was amplified with primers (Forward 5ʹ′ CGGGATCCATGAGCCATCCTGACTACAG 3ʹ′, Reverse 5ʹ′ ATGACATCCTTGAGGCAATCC 3ʹ′) and ligated to the 3ʹ′ end via BamHI and HindIII sites. The sequence of the final clone was verified by sequencing. To produce the mutant form, we introduced a point mutation D391N (Figure S1A), in a residue that corresponds to the catalytically essential aspartate 209 in RNase II (Frazao et al., 2006) and D551 in Rrp44p (Frazao et al., 2006; Schneider et al., 2007). D209N mutation within RNB motif allows RNA binding but prevents cleavage

(Amblar & Arraiano, 2005). The mutation was introduced by site-directed mutagenesis (Stratagene). For the expression of the protein in human cells, the coding sequence of DIS3L2 and AGO2 were cloned in pcDNA5/FRT/TO (Invitrogen). For immunofluorescence, the coding sequence of DIS3L2 was subcloned into pEGFP-N1 (GenBank Accession #U55762) and pEGFP-C3 vectors (GenBank Accession #: U57607) to obtain N- terminal and C-terminal EGFP fused tags, respectively. DNA construct for episomal let-7 expression was a kind gift of Prof. Narry Kim. LIN28A coding region was amplified from the HEK293 cDNA library and the PCR product was subcloned to pcDNA4 vector to produce C-terminal Myc-6xHis fusion protein. Purification of FLAG-DIS3L2 from human cells FLAG-DIS3L2 was purified from stable HEK293T-Rex cell line expressing the fusion protein. Cells were lyzed in 4 ml of ice cold lysis buffer (50 mM Tris pH 8.0, 150 mM NaCl, 0.5% Triton X100 and Complete Protease Inhibitor Cocktail (Roche) and incubated rocking at 4°C for 15 min. Lysates were cleared by centrifugation (14000 rpm, 30 minutes, 4°C). For purification of FLAG-DIS3L2, 100 µl of anti- FLAG M2 beads (Sigma-Aldrich) washed with lysis buffer were incubated with cell extract for 1 hour in a cold room rotating. Beads were extensively washed with 10 volumes of wash buffer (50 mM Tris pH 8.0, 300 mM NaCl, 0.1% Triton X100). Protein elution was done with 1 volume of 3x FLAG peptide (Sigma-Aldrich) resuspended in lysis buffer or by boiling with SDS loading buffer for 5 min. Expression and purification of recombinant wild type and mutant proteins Recombinant DIS3L2 was expressed and purified from BL21-DE3 RIPL strain of E. coli. Bacterial cells were grown at 37°C and protein expression was induced at OD600 0.7 with 0.5 M IPTG at 27°C for 2 hours. Cells were harvested by centrifugation and lysed by sonication in buffer containing 50 mM Tris pH 7.9, 500 mM NaCl, 10% glycerol, 2 mM 2-mercaptoethanol, and 0.1% NP-40. Lysate was cleared by centrifugation (14000 rpm, 30 minutes, 4°C). Protein was purified by Ni-NTA chromatography. SMT3 tag was removed by proteolysis with Ulp1 protease. Resulting recombinant DIS3L2 (rDIS3L2) was further purified by gel filtration on Superdex 200 column (GE Healthcare) in buffer containing 20 mM Tris pH 7.9, 300 mM NaCl, 10% glycerol, 0.01 NP-40, 2 mM 2-mercaptoethanol, 10 mM imidazole.

Immunofluorescence and image processing Cells expressing GFP-fusion proteins were plated in dishes with cover slips coated with 0.2% gelatin. Paraformaldehyde fixed cells were permeabilized with 0.2% Triton-X100 in PBS in the presence of DAPI to stain the nucleus. Coverslips with FlouroMount reagent (Invitrogen) were mount to glass slides and fluorescent images were captured with a Leica DM 6000 B microscope. To visualize DAPI and GFP, 405 nm diode and Argon 488nm lasers were applied, respectively. Transfected HeLa cells were fixed with 3.7% PFA in PBS for 30 min, permeabilized with 0.2% Triton X-100 in PBS for 30 min and blocked with 5% horse serum in PBS for 1 h. All steps were performed at RT. Samples were incubated with primary antibodies (rabbit anti-DIS3L2 or anti-GFP) diluted 1:200 in the blocking buffer overnight at 4°C. The next day, fixed cells were washed three times with 0.2% Triton X-100 in PBS (5 min) and incubated in the dark with the secondary antibody for 40 min at RT. Cy3-, or Cy5-conjugated donkey secondary antibodies (Jackson ImmunoResearch Laboratories, West Grove, PA) were used at dilutions of 1:2,000 or 1:200 with 2.5% horse serum in PBS buffer, respectively. The cells were mounted in FluoroMount reagent (SouthernBiotech, Birmingham, AL) and pictures were captured with a confocal microscope Leica TSC SP5. RNA immunoprecipitation (RIP) Empty HEK293T-REX and HEK293T-REX FLAG DIS3L2 cells were grown to 80% confluence, washed with ice cold PBS. In order to prevent unspecific protein-RNA reassociations during cell lysis (as reported in (Riley et al., 2012)), we stabilized RNA-protein contacts by UV-crosslinking (400mJ, 254nm). Cells were lysed in buffer (LB) containing 150 mM NaCl, 50 mM Tris pH 7.6, 0.5% Triton X-100, supplemented with protease inhibitors (EDTA-free Complete Protease Inhibitor Cocktail, Roche), 0.5 mM EDTA, 1 mM DTT, RNase In (Promega). Lysates were cleared by centrifugation, supernatants applied on FLAG M2 Magnetic beads (Sigma) and incubated for 60 minutes. Beads were washed twice with LB, two times with LB containing 300 mM NaCl and RNA was eluted by treating the beads with 2 mg/ml Proteinase K (New England Biolabs) for 120 min at 37 ºC. Eluted RNA was extracted using phenol/chlorophorm and precipitated with ethanol. After DNase treatment (Turbo DNase, Fermentas), equal amounts of RNA were taken for the synthesis of cDNA by Superscript III reverse transcriptase (Invitrogen). Obtained cDNA was used for PCR amplification and PCR products were resolved on 2% agarose gel.

RNA isolation, cDNA synthesis and quantitative PCR Total RNA was isolated with TRIzol (Invitrogen) according to manufacturer’s instructions, followed by RNase-free DNase (TURBO DNase, Fermentas) treatment. The total RNA concentration was measured in a Beckman Coulter DU 730. 1 ug of purified RNA was reverse transcribed using specific primers or random hexamers (as indicated for individual experiments) and SuperscriptRT III (Invitrogen) according to the manufacturer’s instructions. Real-time quantitative PCR (RT-qPCR) was performed using FastStart Universal SYBR Green Master (Roche) and gene-specific primer pairs (Table S3) on Q-PCR Light Cycler 7500 (Applied Biosystems). Each experiment was performed at least in triplicate. Transcript abundance was calculated by the ΔΔCt (delta delta Ct) method. Data were normalized to an internal control of the housekeeping gene HPRT mRNA or RNU44 for small RNA detection. Results are expressed as means and standard errors of the mean. P-values were calculated by Student’s t-test; P- values < 0.05 were considered significant.

Manuscript

Working title:

DIS3L2 is involved in the degradation of the uridylated tRNA fragments

Authors: Dmytro Ustianenko1, Biter Bilen2, Zuzana Feketova1, Georges Martin2, Mihaela Zavolan2, Stepanka Vanacova1*

Affiliations: 1. CEITEC-Central European Institute of Technology, Masaryk Univer- sity, Kamenice 5, 625 00, Brno, Czech Republic 2. Biozentrum, University of Basel, Klingelbergstrasse 50 / 70 CH - 4056 Basel, Switzerland

*Corresponding author

Authors contribution: manuscript preparation, performance of the CLIP experiment, Figure 4, 5, 7, 8, 10, 11, 12, S1, S4

Keywords: DIS3L2/exosome/tRNA-derived fragments/tRF/small RNAs/CLIP Dmytro Ustianenko Manuscript

Abstract

DIS3L2 is a mammalian homolog of DIS3/Rrp44p the main active subunit of yeast and mammalian exosomes. Mutations in DIS3L2 have been linked to Perlman syndrome and Wilms tumor development. In contrast to DIS3, DIS3L2 do not associate with exosomes. It localizes to cytoplasm where it targets uridylated precursors let-7 miRNA in embryonic stem cells. Altered DIS3L2 expression causes cell cycle and mitotic defects. However, the molecular mechanism causing these phenotypes remains unclear. We have searched for physiological RNA targets of DIS3L2 using in vivo crosslinking and high-throughput sequencing (PAR-CLIP-Seq). We demonstrate that DIS3L2 targets uridylated forms of misprocessed tRNAs. Moreover, our data suggest, that DIS3L2 is involved in the formation of tRNA-derived small RNAs (tRFs). A fraction of DIS3L2 along with tRFs associates with ribosomes and DIS3L2 overexpression alters the polysome/monosome ratio, suggesting that DIS3L2 is involved in translational regulation. Our results therefore establish human DIS3L2 as an exonuclease that specifically targets uridylated RNAs in cytoplasm of embryonic as well as somatic cells.

85 Manuscript Dmytro Ustianenko

Introduction

Every minute, living cells produces millions of RNA molecules. This rapid RNA production is balanced by equally rapid trimming and degradation, and it is this interplay that maintains a proper RNA homeostasis and allows cells to respond to new stimuli. The amount of protein that is produced from an mRNA is proportional to the lifetime of the mRNA, which is directly related to its turnover rate. Thus, mRNA degradation is a key step in gene expression that is subjected to extensive regulation, particularly through Argonaute protein-bound miRNAs (reviewed in (Czech & Hannon, 2010; Lee et al, 1993; Wightman et al, 1993). miRNAs are only one class of small RNAs that have been uncovered in studies employing deep sequencing. Other small RNAs derive from well-known non-coding RNAs (ncRNAs) such as tRNAs, snoR- NAs and/or rRNAs (Cole et al, 2009; Lee et al, 2009; Li et al, 2012b; Liao et al, 2010) These small RNAs do not appear to be exclusively unstable degradation intermediates. tRNAs in particular, give rise to fragments that are stably present in various cell types, and this leads to the emergence of the paradigm of tRNA-derived small RNAs which have additional biological functions (Hey- er et al, 2012; Jochl et al, 2008; Thompson et al, 2008; Thompson & Parker, 2009; Yamasaki et al, 2009). In various studies and depending on the processing pattern, tRNA fragments have acquired different names; the shorter, 20-30 nucleotide (nt) long stable forms originating from 5' or 3' tRNA ends or pre-tRNA trailers were called tRFs, or tsRNAs (Haussecker et al, 2010; Heyer et al, 2012; Lee et al, 2009), while the longer, 30-50 nt stress-induced fragments were called tiRNAs (Emara et al, 2010; Yamasaki et al, 2009), sitRNAs (Li et al, 2008) and tRNA halves (Fu et al, 2009; Jochl et al, 2008; Thompson et al, 2008). Little is known however about the physiological significance of this diverse set of tRNA-derived fragments. The tiRNAs appear able to dis- place translation initiation factors eIF4G/eIF4A, eIF4F from capped and un- capped mRNAs (Ivanov et al, 2011; Yamasaki et al, 2009), thereby inhibiting translation. Some tRNA (but also other non-coding RNA)-derived fragments were found to associate with Argonaute 2 (AGO2), the effector protein of the RNA-induced Silencing Complex (RISC), indicating that they could have

86 Dmytro Ustianenko Manuscript miRNA-like or siRNA-like functions (Haussecker et al, 2010; Li et al, 2012b). Mechanisms responsible for generating short RNA fragments have also started to emerge. For example, Angiogenin has been identified as the key enzyme responsible for the cleavage of tRNAs (Fu et al, 2009; Saxena et al, 1992) upon the exposure of cells to various stress conditions (Fu et al, 2009; Li et al, 2012b; Thompson et al, 2008). Enzymes such as DICER (Cole et al, 2009), the cytoplasmic form of RNase Z (Elbarbary et al, 2009a; Elbarbary et al, 2009b; Lee et al, 2009; Takaku et al, 2003), and RNase A (Li et al, 2012b) have also been linked to the production of short tRNA-derived fragments. Through endonucleolytic cleavage, these enzymes often generate extended products that need to be further digested exonucleolytically to yield the tRFs that are typically detected through RNA sequencing. Processing of non-coding RNAs in eukaryotes involves 3' to 5' exoribonucleolytic trimming catalyzed by the exosome complex that contains a type II nuclease (RNase II) (Mitchell et al, 1997). In yeast, the exosome consists of ten core subunits, all of which are essential. Nine of the subunits are catalytically inactive (reviewed in (Vanacova & Stefl, 2007)) and form a barrel- shaped structure, while the tenth subunit, Rrp44 (also known as Dis3), carries the nuclease activity (Dziembowski et al, 2007; Liu et al, 2006; Schneider et al, 2007). Exosomes operate in both nucleus and cytoplasm, being involved in RNA interference, processing, maturation and turnover as well as in RNA surveillance pathways that recognize and degrade aberrant RNAs (reviewed (Houseley & Tollervey, 2009)). In the cytoplasm, the exosome is responsible for degradation of unstable mRNAs containing AU-rich sequence elements and of defective mRNAs that are detected by quality control pathways such as nonsense- or nonstop-mediated mRNA decay (reviewed in (Houseley et al, 2006; Houseley & Tollervey, 2009)). The nuclear exosomes of S. cerevisiae additionally contain a non-essential 3' to 5' exonuclease, termed Rrp6 (Vanacova & Stefl, 2007). The human genome encodes three Rrp44/Dis3 homologs: DIS3, DIS3L and DIS3L2, which slightly differ in domain composition. The DIS3 and DIS3L proteins associate with the nuclear and cytoplasmic exosomes, respectively (Staals et al, 2010; Tomecki et al, 2010) and have been implicated in nuclear rRNA processing, degradation of promoter upstream transcripts (PROMPTS) and mRNAs such as c-MYC and c-FOS (Tomecki et al, 2010). DIS3L2 is the

87 Manuscript Dmytro Ustianenko most distant Rrp44 homolog. Lacking the PilT N-terminus (PIN) domain that mediates DIS3 association with the exosome core in yeast (Lebreton et al, 2008; Schneider et al, 2009), DIS3L2 may not associate with the exosome, but rather act independently or in association with other proteins. Recent works have linked DIS3L2 mutations with the Perlman syndrome, a genetic overgrowth disorder appearing at birth (Perlman, 1986), and revealed that DIS3L2 plays a role in cell cycle regulation and cell division (Astuti et al, 2012; Neumann et al, 2010). Interestingly, disruptions of the DIS3L2 gene lead to aneuploidy, mitotic errors and changes in expression of mitosis-related proteins such as Aurora-B kinase, Cyclin B1 and p21 (Astuti et al, 2012). How- ever, the molecular mechanism of DIS3L2 function remained elusive. In this work we aimed to uncover the biochemical properties and physiological RNA targets of DIS3L2. By using in vivo crosslinking and immunoprecipitation followed by next generation sequencing (CLIP) we show that DIS3L2 is bound to relatively short (18-34 nt) 5' tRNA-derived fragments (5’ tRFs). Importantly, we found that DIS3L2 is targeting a portion of the uridylated tRNA molecules. Together with 5’ tRFs, a fraction of DIS3L2 associates with ribosomes and DIS3L2 overexpression shifts the polysome/monosome ratio. These results suggest an involvement of DIS3L2 in translational regulation. These findings point to a role of tRFs and DIS3L2 in the regulation of gene expression in human cells via an RNAi-like pathway.

88 Dmytro Ustianenko Manuscript

Results

Identification of DIS3L2 RNA targets by PAR-CLIP

To identify physiological targets of DIS3L2 in vivo, we applied the PAR- CLIP (Photoactivatable-Ribonucleoside-Enhanced Crosslinking and Immu- noprecipitation) method combined with high-throughput Illumina-based sequencing (Martin et al, 2012; Ule et al, 2005; Wang et al, 2009) in a HEK293 cell line overexpressing FLAG-tagged DIS3L2 (Ustianenko et al. 2013). We performed two independent experiments and deposited the sequence data in the Gene Expression Omnibus (Edgar et al, 2002) with accession number (GSExxxx) (Edgar et al, 2002). The experiments resulted in 18,113,339 and 16,151,628 good quality reads, respectively, which we mapped to the human genome assembly (hg19 from the http://genome.cse.ucsc.edu) (Kent et al, 2002; Rosenbloom et al, 2012) using the CLIPZ server (http://www.clipz.unibas.ch/, (Khorshid et al, 2011)). 51% and 21% of the reads, respectively, mapped to the genome and the 3,632,533 (20%) and 1,192,858 (6%) reads that mapped uniquely to the genome were used for our subsequent analyses. For both replicates, more than half of the uniquely mapping reads originated from mRNAs (Figure 1).

Figure 1. Functional annotation of reads uniquely mapping to the genome in two DIS3L2 PAR-CLIP replicates. Categories that accumulate less than 1% of the reads are not represented individu- ally, but are grouped in the “other” category and reads mapping to genomic regions without a functional annotation are represented as none category.

89 Manuscript Dmytro Ustianenko

Repeat elements and tRNAs together represented approximately 30% of the reads, whereas rRNAs, miscRNA (various other known ncRNAs), and reads mapping to un-annotated regions accounted for approximately 15% of the reads (Figure 1). To identify the sites that were most significantly crosslinked to DIS3L2, we used a method that takes advantage of the T-to- C mutations that are introduced during the PAR-CLIP sample preparation (Jaskiewicz et al, 2012b). Obtained DIS3L2 samples, displayed a high fre- quency of T-to-C mutations that are indicative of crosslinking that is a unique feature of the PAR-CLIP protocol (Hafner et al, 2010; Kishore et al, 2011). Both PAR-CLIP samples showed substantial reproducibility at the level of crosslinked genes, especially when considering a limited number (1,000) of top most significantly crosslinked positions. These positions mapped to 225 and 160 distinct genes, respectively, with 106 genes identified in both replicates (Fisher’s Exact test p-value=1.3e-160, odds-ratio=225,Table S1). Fur- thermore, around half of the 1,000 most significantly crosslinked positions from one sample were within 10 nucleotides distance from at least one of the 1,000 most significantly crosslinked nucleotides in the second sample further showing the reproducibility of the crosslinking sites across the replicates. To identify the main in vivo substrates bound by DIS3L2, we tested whether specific categories of genes were over-represented among the most significantly crosslinked 1,000 positions compared to crosslinked positions recovered with much more permissive crosslink score cut-offs (30,318 and 25,811 positions, respectively, taking half of the positions that we would con- sider significantly crosslinked if they were supported by only one read that had a T-to-C mutation). We found that tRNAs were the most enriched (ratio between relative abundance in top 1,000 most significantly crosslinked positions and relative abundance among the 30,318 and 25,811 positions from the two samples, respectively) category of genes represented among the most significantly crosslinked positions: DIS3L2 bound 109 and 107 distinct tRNA genes in the first and second replicates, respectively. 81 tRNA transcripts are common between two replicates (Figure 2). The number of significantly crosslinked positions on individual tRNA genes that were crosslinked in both replicates had a modest linear correlation (R=0.47) further supporting the reproducibility of DIS3L2 preference for these substrates. The most crosslinked tRNAs were Leu(CAA), Val(TAC), Arg(CCG), Ser(GCT), Lys(TTT), Val(AAC),

90 Dmytro Ustianenko Manuscript

SeC(e)(TCA), Arg(TCT), and Ala(AGC). Although more than half of the uniquely mapping reads were annotated as mRNAs, only a small number of mRNAs (N=7) contained positions with a crosslink score in the top 1,000 in both libraries. These transcripts corresponded to HSPA1B, MAZ, BIRC5, ACTG1, USP22, and HIST1H4B. Gene Ontology enrichment analysis with the Ontologizer software (Bauer et al, 2008), see also Methods) of an extended set of 89 mRNA genes that gave rise to significantly crosslinked sites in either of the replicates revealed endo- plasmic reticulum (adjusted p-value=0.001, N=21) and the endomembrane system (adjusted p-value=0.004, N=24) cellular components (Table S2). The abundantly and ubiquitously expressed (Landgraf et al, 2007) mature miRNAs from the miR-17-92 locus (miR-17, mir-18a, mir-20a) were the only miRNAs that were identified in this experiment. Among rRNA, we obtained crosslinks to RN5S9 and RN5S215 5S rRNA transcripts, which are part of the large ribosomal subunit (Table S1). Although snoRNAs are predominantly nuclear and DIS3L2 predominantly cytoplasmic, we identified 35 snoRNAs, among which U60, U14A, U8, mgU6-77, HBII-99B, U33, and U56 were re- producibly identified in both replicates. With the exception of U1 and RNU12, almost all crosslinked snRNAs (RNU2-2, RNU2-6P, RNU4-2, RNU5E-1, U2, and RNU11) were identified in both replicates.

Figure 2. tRNA genes are enriched among 1000 significantly crosslinked positions. Categories shown on the x- axis are in descending order of the number of distinct genes represented in the set of the 1000 most significantly crosslinked positions. Y-axis is the p-value (-log10) of Fish- ers Exact test.

91 Manuscript Dmytro Ustianenko

DIS3L2 predominantly crosslinks to tRNA 5' ends

To further investigate the DIS3L2 binding to the tRNA, we used 44 CLIPped tRNAs with mature forms that are precisely 72-nucleotide-long to calculate the profile of the average crosslink score along the tRNAs. We found that the highest binding scores were in the 5' region of the tRNAs (Figure 3).

Figure 3. DIS3L2 predominantly crosslinks to tRNA 5' ends of the tRNA. On the right, profile of average crosslink score (log of the probability that crosslink- diagnostic T-to-C mutations are more frequent than expected by chance, see Meth- ods) along the 72-nucleotide long tRNAs that were crosslinked in the two replicate samples. The empirical 90% confidence intervals of the average crosslink scores obtained by shuffling the crosslinked positions along the T positions of the crosslinked tRNA species 2,000 times is shown with dashed lines. The crosslink position on specific regions of tRNAs are indicated below the x-axis. On the left, secondary structure of tRNA AGC as a representative tRNA species of length 72 (predicted by tRNAscan- SE (Lowe & Eddy, 1997) and rendered by VARNA (Darty et al, 2009)).

This positional preference was not due to the preferential loss of the 3' end reads that end in the non-templated CCA tri-nucleotides by our processing pipeline, and also was not dependent on whether we used uniquely or multi-mapping reads in our analysis. Hence the DIS3L2-crosslinked tRNA fragments closely resemble the 5' tRNA fragments designated as 5' tRFs (Haussecker et al, 2010; Lee et al, 2009), we decided to use this term for the

92 Dmytro Ustianenko Manuscript

DIS3L2-bound 5' tRNA fragments. To further validate the CLIP results we performed RNA immunoprecipitation (RIP) analysis from a new cell line with tetracycline-inducible expression of stably integrated 3xFLAG-DIS3L2. The northern blot analysis with probes directed to the 5' and 3' ends, of tRNAVal(AAC) revealed 24-30 nt long fragments, which corresponds to the tRF length distribution in the CLIP data (Figure 4, DIS3L2 samples).

Figure 4. Northern blot analysis of the tRNA Val AAC immunoprecipitated with FLAG-DIS3L2. RNAs obtained by RIP were separated on 10 % denaturing polyacrylamide gel, transferred to nylon membrane and hybridized with 5'-end labeled DNA oligos corresponding to 3' and 5' part of Homo sapiens chr6.trna136-ValAAC. Ago2 and DIS3 proteins used as a negative control. While MUT DIS3L2 is binding to the longer form of the tRNA, wild type protein is binding to “mature” form of the tRNA derived fragments, indicating its potential involvement in their processing.

These fragments were not immunoprecipitated with the DIS3 protein control. Due to an active exonuclease activity of DIS3L2 it was not clear whether the observed tRNA fragments represent the starting or end point of DIS3L2 mediated degradation/processing activity. We have previously shown, that by using D391N catalytic mutant it is possible to track the physi-

93 Manuscript Dmytro Ustianenko ological DIS3L2 targets more efficiently (Ustianenko et al, 2013). Interest- ingly, D391N mutant co-precipitated longer forms of tRNAs, which appeared as a smeary pattern of fragments migrating slower than mature tRNAs (Fig- ure 4, D391N lane). Similar result was also obtained with probes directed to RNU-11 snRNA (Figure S1). To uncover the identity of these tRNAs, we added a DNA linker at the 3' end of the RIPed RNAs that were used for both, cDNA synthesis and subsequently for PCR amplification in combination with a tRNA-specific forward primer (Figure 5).

Figure 5. D391N DIS3L2 mutant binds uridylated forms of tRNAs. Left panel: Scheme of the experiment. Right panel: Sequencing analysis of 3' termini of tRNAs coprecipitated with D391N DIS3L2 mutant from HEK293T-Rex cells. The genome-encoded tRNA sequence is represented in black, whereas the nontemplated nucleotides are in red. The tRNA region corresponding to the mature and precursor form is depicted on top. Shortest fragments are representing 5’ tRF. Untemplated 3’ uridines are in red.

The PCR products were subcloned and 20 clones of each tRNA were sequenced. Five out of 20 clones represented precursor tRNAs Leu that contain encoded stretch uridines. The rest, 15 for tRNA Leu and 10 for tRNAVal corresponded to tRNAs, truncated from their 3' ends possessing untemplated stretches of up to 12 uridines (Figure 5). This highly resembled the DIS3L2 targeting uridylated precursors of let-7 miRNA (Chang et al, 2013; Ustianenko et al, 2013). We conclude that DIS3L2 exonuclease targets uridylated tRNA fragments.

94 Dmytro Ustianenko Manuscript

DIS3L2 and 5' tRFs associate with active polysomes

The fact that DIS3L2 has been shown to associate with active polysomes (Lubas et al, 2013) and tRFs have been known to interact with proteins involved translation regulation, such as YB1, TIA1, PABP1, AGO2 (Ivanov et al, 2011), we asked whether DIS3L2 targets tRNAs on polysomes. In agreement with the recent report, we observed that DIS3L2 specifically co- sedimented with ribosomal fractions (Figure 6). Interestingly, the overexpression of WT DIS3L2 resulted in changed polysome to monosome ratio, which is indicative of translational inhibition (Figure 6A, Figure S2).

Figure 6. Polysomal profiles of sucrose gradients from HEK293T-Rex cells. Top panel: polysome enriched fraction of HEK293 T-Rex and HEK293 T-Rex DIS3L2 was centrifuged through a linear 10–50% (wt/vol) sucrose gradient. The top portion of the gradient is on the left. The peaks correspond to the particular ribosomal fractions are indicated above. Bottom panel: A fraction of DIS3L2 copurifies with HEK293 T-Rex ribosomes on sucrose gradients. The distribution of DIS3L2 and ribosomal protein L8 (RPL8) in the fractions of the gradient were analyzed by immunoblotting using anti-DIS3L2 and anti-RPL8 antibodies, respectively.

95 Manuscript Dmytro Ustianenko

The polysomal profiles showed an increased signal for 80S monosomes in FLAG-DIS3L2 overexpressing samples. Comparing the relative abundance of polysome and monosome peaks of control and FLAG-DIS3L2 overexpressing cells, we found that the monosomal fraction increased three times upon DIS3L2 overexpression (Figure 7). Interestingly, overexpression of the catalytically inactive D391N mutant did not reveal as significant shift in P/M ratio, indicating that the nuclease activity of DIS3L2 is responsible for

Figure 7. Ratio of the polysomes versus monosomes extracted from the profiles shown in figure 6 and S2. Wild type protein overexpression significantly decreases the amount of the translat- ing comparing to the empty cells. Slight reduction in the P/M ration is also observed in the mutant DIS3L2 overexpressing cells. this phenotype (Figure 6 80S peak on the right graph, Figure 7 last column). Because the yeast Dis3 and exosome are essential for ribosomal RNA (rRNA) processing and degradation, we checked whether DIS3L2 overexpression leads to changes in the rRNA profile. We did not observe any significant changes in the rRNA profile in cells with altered DIS3L2 expression (Figure 8). Moreover, Astuti et al. ruled out the involvement of DIS3L2 in 5.8S rRNA maturation.Reasoning that 5' tRFs mediate at least some of DIS3L2 activity on ribosomes, we isolated RNA from the individual fractions of the sucrose gradient (see above Figure 6) and used it for stem-loop RT-PCR analysis with specific primers for two 5' Leu tRFs. We found that thetRF signal paralleled DIS3L2 distribution along the whole sucrose gradient, with

96 Dmytro Ustianenko Manuscript

Figure 8. The ribosomal RNA profile is not affected in cells with altered expression of DIS3L2. Total RNA was isolated from HEK293T-Rex cells overexpressing wild-type (WT) or D391N DIS3L2 mutant (MUT), and control cells (empty) as well as HeLa cells treated with control (CTRL) and Dis3L2 siRNAs (K.D.). The total RNA was separated on denaturing agarose gel and stained with SYBER green II RNA gel Stain (Invitrogen). The migration position of large subunit ribosomal RNA (LSU rRNA), small subunit rRNA (SSU rRNA) and tRNAs is indicated on the right side. a stronger signal over the 60 and 80S monosomes (Figure 9), indicating that they may have a role during translation initiation. To understand better the effect of DIS3L2 and tRF overproduction we monitored this effect on the growth of the HEK293 T-Rex cell culture with overexpressed DIS3L2. The delay in the growth of the cells overexpressing wild type of DIS3L2 was up to 40% compared to the empty and mutant cell lines. The change in the ribosome profiles and reduced translation efficiency (unpublished data from our laboratory) could cause this slow growth phenotype (Figure 10).

97 Manuscript Dmytro Ustianenko

Figure 9. tRFs and miR-30 associate with ribosomal fractions. PCR analysis of cDNAs prepared from RNAs associated with individual fractions from the sucrose gradient shown in (A). The RT-PCR protocol is designed to amplify 5' tRFs, not full length tRNAs (see methods). No RT is a control where no reverse transcriptase was used.

Figure 10. Growth analysis of control and cells overexpressing WT and D391N form of DIS3L2.

98 Dmytro Ustianenko Manuscript

TUTase 4 uridylates 5’ tRF’s

We next asked which enzyme is responsible for tRNA uridylation. Sev- eral terminal uridine transferases are known to be acting in the cytoplasm of the human cells (Scott & Norbury, 2013). In order to identify which of these can be responsible for the 3' uridylation of the tRNA derived fragments we have expressed and purified the potential candidates; TUTase 4, TUTase 6 and TUTase 7 (TUTase 7 expressing constructs are a kind gift from Dr. Nor- bury) in human HEK293T-Rex cells (Figure S3). Proteins were used for the in vitro uridylation assay using 37nt 5' tRF Leu as a target. TUTase 4 and TUTase 7 revealed uridylation activity of tRF as a substrate. Reaction resulted in the extensive uridylation of the substrate by TUTase 7 protein (Figure 11), closely resembling the phenotype reported previously by (Rissland et al, 2007). On the other hand, TUTase 4 extended the 5' tRF Leu RNA by approximately 8-10 UMPs, a length observed at tRNAs precipitated with mutant DIS3L2 (Figure 5). Therefore we propose that TUTase 4 is the enzyme acting on tRNAs. In order to investigate how the 3' end modification of the 37nt 5' tRF Leu with untemplated uridines can influence the activity of the DIS3L2 protein. We have performed a coupled in vitro uridylation - degradation

Figure 11. In vitro uridylation assay of 5’ tRF Leu. Human TUTase4, TUTase6 and TUTase 7 were expressed in human HEK293T cells and purified using FLAG-tag specific antibodies. In vitro uridylation assay performed on the 37nt long 5’ tRNA fragment identified in the RIP experiment.

99 Manuscript Dmytro Ustianenko assay. The degradation activity of DIS3L2 was significantly enhanced in the presence TUTase 4 (Figure 12). Moreover this activation was dependent on TUTase 4 uridylation activity because the catalytically inactive form of TUT- ase 4 (DADA) did not stimulate tRF degradation by DIS3L2 (Figure 12, last three lines).

Figure 12. TUT4 activity stimulates tRNA degradation by DIS3L2. Left panel: in vitro uridyltaion-degradation assay. The RNA substrate was incubated with TUT4 proteins for 40 min prior to DIS3L2 addition. The reactions were subsequently incubated for time indicated on top. TUT4 WT is the wild type form, TUT4 DADA is catalytically inactive form of TUT4, DIS3L2 MUT is the D391N inactive mutant of DIS3L2. Right panel: Western blot analysis of proteins purified from HEK293T- Rex cells. The proteins were detected with FLAG-tag antibodies (both TUTase 4 and DIS3L2).

100 Dmytro Ustianenko Manuscript

Discussion

The application of the high-throughput sequencing of various RNA size fractions have recently led to the identification of new types of small and long RNAs, some of which play role in the regulation of gene expression. In this work we have established DIS3L2 as a factor that is potentially involved in the generation of tRNA-derived fragments with previously reported regulatory potential (Figure 13). DIS3L2 is a human homolog of yRrp44p, the key exonuclease of yeast exosomes. In contrast to its yeast and human homologs DIS3L2 does not associate with exosome. To uncover the DIS3L2-specific RNA targets in vivo, we applied photo-crosslinking followed by DIS3L2 immunoprecipitation and next generation sequencing, (DIS3L2 PAR-CLIP) in HEK293 Flp-In cells. Analysis of CLIP data combined with further experimental validation revealed a strong enrichment in binding of DIS3L2 to tRNA

Figure 12. Proposed mechanism of the DIS3L2 mediated tRNA degradation and potential tRF formation through uridylation of the endonucleolitically cleaved tRNA molecules and tRNA precursors.

101 Manuscript Dmytro Ustianenko molecules, particularly to fragments originating from tRNA 5'-ends. We designated these molecules tRFs, akin to the tRNA-derived fragments recently reported by (Haussecker et al, 2010). A number of studies reported that tRNA halves are produced in response to certain stress stimuli, but also persist in a stable form in the cell (Fu et al, 2009; Lee et al, 2009; Li et al, 2012a). Up to date, several endonu- cleases such as DICER1, Angiogenin or RNase Z, have been proposed to generate different types of tRNA fragments (Cole et al, 2009; Lee et al, 2009; Takaku et al, 2003). In many cases, such cleavage would generate fragments longer than those identified by RNA sequencing, which presumably captures mature/functional fragments. This indicates an additional exonuclease activity is involved in the processing of tRFs. Till now it was not clear whether DIS3L2 targets tRNA halves result from an initial endonucleolytic cleavage or whether it targets full length tRNAs. We uncovered, that the 5'tRFs are produced from longer uridylated forms of tRNAs. We have identified two types of DIS3L2 bound tRNAs with 3' terminal stretches of Us. One we believe, corresponds to tRNA precursor forms that for some reason escaped the processing by tRNaseZ and were exported to cytoplasm. The other type is tRNAs, that are partially trimmed/cleaved from the 3' end and uridylated. These fragments appear to be restricted in the single stranded area of the tRNA in the antico- don and T-loop and consequently extended on their 3' end with stretches of uridines. Similarly to the pre-let-7 uridylatuon (Chang et al, 2013; Ustianenko et al, 2013), we have identified TUTase 4 as the probable candidate for this activity. Interestingly, Xue et al. have recently reported that Dis3 and the exosome can produce, together with Argonaute, mature miRNAs in Neurospora crassa (Xue et al, 2012). It is currently unclear how DIS3L2 selects its targets. Apart from 5' tRFs, DIS3L2 CLIP identified other small RNAs originating from RNA Polymerase III transcripts, such as 5S rRNA and a miR-1975 originating from the Y-RNA (Canella et al, 2010) (Table S1). On the other hand, it is possible that DIS3L2 associates with other RNA-binding cofactors that provide RNA substrate specificity. The key question is whether DIS3L2-linked tRFs possess any biological function. Mostly due to tRF association with Argonaute proteins, some recent studies focused on the function of tRFs in the RNAi pathway. It was shown that certain 3' tRFs in association with AGO2 are able

102 Dmytro Ustianenko Manuscript to induce cleavage of a complementary RNA in vitro (Li et al, 2012a), yet in vivo their trans-silencing potential was rather small (Haussecker et al. 2010). On the other hand, in Tetrahymena, nuclear tRNA derived 3' RNAs associate with a Piwi-like protein and, in complex with the nuclear 5' to 3' exonuclease Xrn2, are responsible for the processing of the ribosomal RNAs (Couvillion et al, 2012). In fission yeast, tRNAGlu-derived small 5' tRFGlu represent a highly abundant class of sRNAs that have been proposed to interfere with the RNAi pathway (Buhler et al, 2008). Last, but not least, in mammalian cells, tRNA fragments have been shown to induce translational arrest (Yamasaki et al, 2009). In this context, we show that a fraction of DIS3L2 and the respective 5' tRFs associate with ribosomes and that DIS3L2 overexpression changes ribosomal profiles towards monosomes, which is generally understood as a sign of translational inhibition. The uridylation-mediated degradation/processing of the aberrant species of tRNAs highly resembles the nuclear RNA surveillance pathway that involves a noncanonical poly(A) polymerase and the exosome complex (Kadaba et al, 2004; LaCava et al, 2005; Vanacova et al, 2005). Therefore, we propose that TUT-DIS3L2 activities represent cytoplasmic form of RNA quality control. In this respect, the noncanonical polyadenylation would be a hallmark of the nuclear and uridylation of the cytoplasmic RNA surveillance.

103

Dmytro Ustianenko Manuscript

Material and Methods

Mammalian cell culture and transfection

Mammalian cells (HEK Flip In, HEK Flip In FLAG-DIS3L2, HEK293T, HeLa) were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM), supplemented with 10% fetal calf serum at 37°C in the presence of 5% CO2. Inducible conditions: 10-50ng/ml of doxycycline. For transient expressions, cells were grown to 70% confluency, plasmid DNA was transfected using TURBOFECT (Fermentas) following the manufacturer manual.

Preparation of stable cell lines

Plasmids (pcDNA5 FRT/TO FLAG DIS3L2, pcDNA5/FRT FLAG DIS3L2, pcDNA5/FRT/TO DIS3l2 D391N mutant) were transfected to Flp-InTM or to Flp-InTM T-RExTM (Invitrogen) cell line with TURBOFECT reagent. The stable inducible cell lines overexpressing different versions hDIS3L2 were generated with the use of the HEK293 Flp-In™ and Flp-In T-REx™ system (Invitrogen) according to the protocol of the manufacturer.

Purification of FLAG-DIS3L2 from human cells

Flag-DIS3L2 was purified from stable HEK293 Flp-In cell line expressing the fusion protein. Cells were resuspended in 4 ml of ice cold lysis buffer (50 mM Tris pH 8.0, 150 mM NaCl, 0.5% Triton X100 and Complete Protease Inhibitor Cocktail (Roche) and incubated rocking at 4 °C for 15 min. Lysate was cleared by centrifugation (14000 rpm, 30 minutes, 4 °C). For purification of FLAG DIS3L2, 100 μl of anti- FLAG M2 beads (Sigma-Aldrich) washed with lysis buffer were incubated with cell extract for 1 hour in a cold room rotating. Beads were extensively washed with 10 volumes of wash buffer (50 mM Tris pH 8.0, 300 mM NaCl, 0.1% Triton X100). Protein elution was done with 1 volume of 3x FLAG peptide (Sigma-Aldrich) resuspended in lysis buffer or by boiling with SDS loading buffer for 5 min.

105 Manuscript Dmytro Ustianenko

Purification of human TUTases and in vitro uridylation assay

DNA constructs expressing human TUTase 4, 6 and 7 were transfected to HEK293T-Rex cells. 48 hours after transfection cells were lysed in a buffer (50 mM Tris-HCl (pH 8), 150 mM KCL, 0.5% TritonX100, 1 mM dithiothreitol, 1 mM PMSF, and 1 Complete Mini, EDTA-free protease inhibitor cocktail tab- let (Roche). Cell lysates (TUTase 4 and TUTase 6) were applied on FLAG- magnetic beads (Sigma-Aldrich) and IgG FastFlow Sepharose (Amersham) (for TUTase 7). After one hour incubation with beads, bound complexes were extensively washed with Lysis buffer and Wash buffer (Lysis buffer with 300 mM KCL). In vitro uridylation reaction was performed in a total volume of 30 ml in 4 mM MgCl2, 1 mM DTT, 0.25 mM UTP and 15 ul of immunopurified proteins on beads in Lysis buffer. The reaction mixture was incubated at 37C for 40 min.

RNA isolation, cDNA synthesis and quantitative PCR

Total RNA was isolated with TRIzol (Invitrogen) according to manufacturer instructions followed by RNase-free DNase (TURBO DNase, Fermen- tas) treatment. The RNA concentration was measured in a Beckman Coulter DU 730. The RNA for in vitro tests was either prepared by in vitro transcription (tRNA Ala) from DNA templates containing T7 promotor and T7 RNA polymerase or purchased as synthetic molecules from Dharmacon or Sigma. For the degradation assays, the RNAs were radioactively labeled at the tRNA 5’ end by T4 polynucleotide kinase (New England Biolabs) and 32P gamma-ATP (company, MGP-Zlin).

Northern blot analysis

Total RNA was resolved on 10% denaturing polyacrylamide gel and transferred to Hybond-N+ membrane (GE Healthcare) by electroblotting (Bio- Rad). The hybridization with radioactively labeled oligonucleotides was performed in ULTRAhyb-oligo hybridizatioin buffer (Ambion) at 38 °C. Prior to adding the labeled probe membrane was prehybridized at 42°C for 2 hours. The radioactive signal was monitored by phosphorimager FLA-9000 (FUJI-

106 Dmytro Ustianenko Manuscript

FILM). Quantification of signals was done using Multi Gauge software v3.2 (FUJIFILM).

Photoactivatable-Ribonucleoside-Enhanced Crosslinking and Immuno- precipitation (PAR-CLIP)

PAR-CLIP was performed essentially as described (Martin et al, 2012). Human embryonic kidney cells (HEK293 FlpIn, Invitrogen) stably expressing human N-terminal FLAG-DIS3L2 fusion protein were grown overnight in a medium supplemented with 100 uM 4-thiouracyl (4-SU). Cells were washed with 1xPBS and exposed to 150 mJ of 365 nm UV light in a Stratalinker 2400 device (Stratagene). Cells were collected, frozen in liquid nitrogen and stored at -80°C. PAR-CLIP was performed essentially as described (Martin et al, 2012). Whole cell lysate was prepared in PNDS buffer (1x PBS, 0.5 NP-40, 0.25% deoxycholate, 0.05% SDS supplemented with 1 mM DTT, protease inhibitor cocktail (Roche) and RNase inhibitor RNAsin (Promega). Lysates were cleared by centrifugation and split in two aliquots. One aliquot was treated with 4 Units/ml and the other aliquot with 12 Units/ml of RNase I (Ambion, AM2294) for 10 min at 32°C and pooled on ice. FLAG-tagged DIS3L2 was immunoprecipitated using anti-FLAG M2 monoclonal antibody (Sigma) bound to Protein G Dynabeads (Invitrogen). Beads were treated with alkaline phosphatase (Fast-AP, Fermentas). 3' adaptor (P-UCGUAUGCCGUCUUCUGC- UUGU-Pur) was ligated to the bound RNA with T4 RNA ligase (Fermentas) in buffer containing 25% PEG 8000 at 16°C over night. The crosslinked RNAs were radiolabeled with polynucleotide kinase (T4 PNK, NEB) and γ-32P ATP. Protein-RNA complexes were resolved on a 4-12% gradient SDS-PAGE (Nu- PAGE, Invitrogen), the band corresponding to tagged DIS3L2 was cut out from the gel and eluted with proteinase K containing elution buffer (50mM Tris pH 7.5, 50 mM NaCl, 10 mM EDTA, 2 M urea, 2 mg/ml proteinase K) at 50°C for 2 hrs. RNAs were then ligated to 5' adaptor (OH-GUUCAGAGUUCUACA- GUCCGACGAUC-OH). RNA was size fractionated on 8% polyacrylamide- 8 M urea gel and 70-110 nt long RNA fragments were eluted. Reverse transcription was done with a 3' primer (5' CAAGCAGAAGACGGCATACGA 3’) and with Superscript III reverse transcriptase (Invitrogen). Resulting cDNAs were used as templates for PCR amplification (5' Primer: 5' AATGATACGGC-

107 Manuscript Dmytro Ustianenko

GAC-CACCGACAGGTTCAG-AGTTCTACAGTCCGA 3' and the lowest possible number of PCR cycles were used. The PCR products were sequenced on an Illumina Genome Analyzer IIx.

Sucrose gradient fractionation for polysomal profiling

The polysomal profiling was performed with minor modifications according to (Damgaard & Lykke-Andersen, 2011). Polysomes were isolated from one 15 cm dish of the cell lines HEK293T-Rex empty control, HEK293T-Rex Flag-DIS3L2 and HEK293T-Rex Flag-DIS3L2 D391N. Cells were cultured in DMEM supplemented with 10% fetal bovine serum without antibiotics. When indicated overproduction of both wt and mutant forms of the DIS3L2 protein was achieved by adding doxycycline (100 ng/ml) directly to media. Actively growing cells (confluence around 85-90%) were treated with 10 ug/ml cycloheximide for 5 minutes at 37°C, all subsequent steps were performed in a cold room. Cells were washed twice with cold PBS supplemented with 10 ug/ml cycloheximide and then lysed on plate using PB (100 mM KCl, 5 mM MgCl2, 10 mM HEPES pH 7, 10 ug/ml cycloheximide, 0.5 % NP-40, 1 mM DTT, 1 mM PMSF, 200 ug/ml Heparin, 100 U/ml RNAsin Plus, 2 mM vanadyl ribonucleoside complex, protease and phosphatase inhibitors). Lysates were incubated on ice for 15 min. and cleared by centrifugation at 15 000 g for 15 min. Supernatants (24 Units) were separated in 5-45% sucrose gradients by ultracentrifugation (35 000 rpm, 105 min. in Beckman SW41 rotor at 4 ° C). Polysome profiles were fractionated (1 ml fractions) using an ISCO UA-5 gradient analyzer connected to a Clarity data acquisition station (DataApex). Proteins from individual fractions were ethanol precipitated and separated on 9% SDS-PAGE gels and proteins were transferred and immobilized onto PVDF membranes. Proteins were detected by western blot analyses with antibodies indicated (α-DIS3L2 rabbit polyclonal serum 1:800, and α-RPL8 (Ab- cam) goat polyclonal serum 1:500, Abcam). Both α-goat (Sigma) and α-rabbit (Promega) HRP secondary antibodies were diluted 1:5000. The polysome to monosome ratios were acquired using the Image SXM 193 software (NIH).

108 Dmytro Ustianenko Manuscript

RNA immunoprecipitation

Empty HEK293T-REX cells and HEK293T-REX FLAG DIS3L2 were grown to 80% confluence. Washed with ice cold PBS and UV cross-linked (400mJ, 254nm). Cells lysed in buffer containing 150mM NaCl, 50mM Tris pH 7.6, 0.5% Triton X 100, supplemented with protease inhibitors (EDTA- free Complete Protease Inhibitor Cocktail, Roche), 0.5mM EDTA, 1mM DTT, RNase In (Promega). Lysate cleared by centrifugation. Supernatant applied on FLAG M2 Magnetic beads (Sigma) and incubated for 60 min. Washed two times with Lysis buffer, two times with Lysis buffer containing 300mM NaCl. RNA was eluted by treating whole beads with 2mg/ml Protease K (New England Biolabs) for 120 min at 37C. RNA was phenol/chlorophorm cleaned and ethanol precipitated. After DNase treatment (Turbo DNase, Fermentas) equal amount of RNA was taken for the cDNA synthesis by Superscript III (Invitrogen). Obtained cDNA was used for semi quantitative PCR. Reaction resolved on 2% agarose gel.

In silico analysis of high-throughput sequencing data

We extensively used custom Perl (version 5.8.1) and R (version 2.15.1) scripts as well as the Bedtools (version 2.16.1) package (Quinlan & Hall, 2010) for many steps of the analysis.

Read-to-genome mapping

The high-throughput sequencing resulted in an average of 4 x 107 reads per sample. After filtering out reads with nucleotides that could not be unambiguously called, we mapped the reads to the human genome 19 assembly downloaded from the University of California at Santa Cruz (hg19) on the CLIPZ server (www.clipz.unibas.ch and (Khorshid et al, 2011). The copy numbers of reads that mapped to multiple locations (multi-mappers) were distributed equally among the possible mapping locations. Reads that did not map in the first step were processed to remove the intact and truncated CCAs at the 3' end and then remapped with the CLIPZ server. We merged the results of these two mapping steps and calculated the genome-wide crosslink

109 Manuscript Dmytro Ustianenko scores as described in (Jaskiewicz et al, 2012b). We defined the crosslink score as a measure of the significance of the position being a crosslink position. More specifically, we defined as log(1-pc), where pc is the posterior probability that the position is a crosslink position. Low values indicate high significance, similar to p-values. In the calculation of the crosslink score we used only uniquely mapping reads that were not annotated by the CLIPZ server as being of bacterial, fungal, vector, marker/adaptor or viral origin.

Annotation of the genomic crosslinked sites

We used the RefSeq, snomir, and tRNA gene tracks (Dreszer et al, 2011) and the wgEncodeGencodeBasicV12 tracks for snRNA and rRNA genes (Rosenbloom et al, 2012) from University of California at Santa Cruz (http://genome.cse.ucsc.edu). We further annotated the RefSeq genes with their biotype information and used one representative transcript from long- noncoding RNAs or mRNAs. The selection of the representative transcripts of single-locus genes were done based on a hierarchy of several criteria or- dered by biotype (mRNA > non-coding RNA), RefSeq gene status (Reviewed > Validated > Inferred > Provisional > Predicted), length of genomic locus, and length of the exonic region with a higher precedence for the longer loci and transcripts.

110

Manuscript Dmytro Ustianenko

Figure S1.

Northern blot analysis of the RNU11 immunoprecipitated with FLAG-DIS3L2. RNAs obtained by RIP were separated on 10 % denaturing polyacrylamide gel, transferred to nylon membrane and hybridized with 5ʹ - end labeled DNA oligos corresponding to the sequence of Homo sapiens U11 small nuclear RNA gene (RefSeq number: NR_004407.1). Ago2 and DIS3 proteins used as a negative control. While in both control samples and wild type DIS3L2 are co-purifying the background levels of the RNU11 RNA, the mutant DIS3L2 is bound to the extended and potentially uridylated RNU11 transcript. Experiment was performed multiple times, the most represented image is provided. Thus the possibility of the smear signal detected in the DIS3L2 MUT sample being an artifact is eliminated.

112 Dmytro Ustianenko Manuscript

Figure S2.

Uper panel. Polysomal profiles of sucrose gradients from HEK293T-Rex cells overexpressing D391N mutant form of the protein. A polysome-enriched fraction of HEK293T was centrifuged through a linear 10–50% (wt/vol) sucrose gradient. The top of the gradient is on the left. The peaks corresponding to particular ribosomal fractions are indicated above the peaks. The distributions of DIS3L2 and ribosomal protein L8 (RPL8) in the fractions of the gradient were analyzed by immunoblotting using anti-DIS3L2 and anti-RPL8 antibodies, respectively (below each graph). The overexpression of the tagged proteins was either not induced (leaky expression) or induced with doxycyclin (+Dox). Lower panel the control ribosome profiles treated with EDTA (middle) and puromycin that abolishes the formation of polysomes. Signal for the DIS3L2 protein has disappeared (westen blot below the gels) indicating that the presence of the protein is dependent on the ribosome assembly.

113 Manuscript Dmytro Ustianenko

Figure S3.

Immunoprecipitation of the human TUTase 4, TUTase 6 and TUTase 7. All proteins were expressed in the HEK293 cell lines. TUT- ase 7 construct (pcDNA-Hs2-TAP) is a kind gift from Dr. Norbury. TUTase 7 was purified using TAP tag, TUTase 3 and TUTase 6 were purified using FLAG tag antibodies.

114 Dmytro Ustianenko Manuscript

Table S1.

Gene Symbol BioType CrosslinkCount_1 CrosslinkCount_2 hsa-mir-1308 miRNA 5 1 hsa-mir-1975 miRNA 3 2 hsa-mir-17 miRNA 1 0 hsa-mir-18a miRNA 0 1 hsa-mir-20a miRNA 0 1 DIS3L2 mRNA 23 24 HSPA1B mRNA 3 16 MAZ mRNA 3 10 BIRC5 mRNA 7 2 ACTG1 mRNA 6 2 SCD mRNA 5 0 USP22 mRNA 1 2 CD164 mRNA 2 0 HIST1H4B mRNA 1 1 LAPTM4B mRNA 2 0 PTMA mRNA 0 2 RAB14 mRNA 2 0 ADCY6 mRNA 1 0 ARF1 mRNA 1 0 ARHGAP32 mRNA 1 0 ARHGDIA mRNA 1 0 ARL6IP1 mRNA 1 0 ATP2A2 mRNA 1 0 ATP2B4 mRNA 1 0 ATP5G2 mRNA 1 0 C9orf40 mRNA 1 0 CANX mRNA 1 0 CCT4 mRNA 1 0 CD63 mRNA 1 0 CDC25B mRNA 1 0 CDCA5 mRNA 1 0 CLDN12 mRNA 1 0 CLINT1 mRNA 1 0 CLPTM1L mRNA 1 0 DDX6 mRNA 1 0 DHCR7 mRNA 1 0 DIAPH1 mRNA 1 0 DOLK mRNA 1 0 DSC3 mRNA 1 0 EDC4 mRNA 1 0

115 Manuscript Dmytro Ustianenko

Gene Symbol BioType CrosslinkCount_1 CrosslinkCount_2 EEF1A1 mRNA 1 0 EIF5A mRNA 1 0 ERGIC1 mRNA 1 0 ERP29 mRNA 1 0 FAM96A mRNA 1 0 GMPS mRNA 1 0 H3F3B mRNA 1 0 HMGA1 mRNA 1 0 HN1L mRNA 1 0 HSD17B10 mRNA 1 0 HSPA1A mRNA 0 1 HSPA5 mRNA 1 0 ILF2 mRNA 1 0 IVNS1ABP mRNA 1 0 KCTD5 mRNA 1 0 KDELR1 mRNA 1 0 KDELR2 mRNA 1 0 KLHL23 mRNA 1 0 KPNA6 mRNA 1 0 LDHA mRNA 1 0 LPHN2 mRNA 1 0 MANEA mRNA 1 0 MATR3 mRNA 1 0 MZT2B mRNA 1 0 NMT1 mRNA 1 0 NUFIP2 mRNA 1 0 PKM mRNA 1 0 PLAGL2 mRNA 1 0 PRC1 mRNA 1 0 PRRC2A mRNA 1 0 PSME2 mRNA 1 0 RAB11A mRNA 1 0 RCC2 mRNA 1 0 RHOA mRNA 1 0 RNF145 mRNA 1 0 RPLP0 mRNA 1 0 RPS16 mRNA 1 0 SERINC1 mRNA 1 0 SNTB2 mRNA 1 0 SP1 mRNA 1 0 SPTLC2 mRNA 1 0 STRN3 mRNA 1 0

116 Dmytro Ustianenko Manuscript

Gene Symbol BioType CrosslinkCount_1 CrosslinkCount_2 TAP2 mRNA 0 1 TMBIM6 mRNA 1 0 TMED2 mRNA 1 0 TPBG mRNA 1 0 TUBA1B mRNA 1 0 UNC119B mRNA 1 0 ZC3H7B mRNA 1 0 ZCCHC2 mRNA 1 0 ZNF207 mRNA 1 0 ZNF362 mRNA 1 0 ZNF703 mRNA 1 0 ZSCAN18 mRNA 1 0 XIST ncRNA 84 228 RN5S9 rRNA 5 2 RN5S215 rRNA 5 1 RN5S145 rRNA 2 0 RN5S149 rRNA 1 0 RN5S202 rRNA 0 1 RN5S248 rRNA 1 0 U60 snoRNA 5 7 U14A snoRNA 1 9 U8 snoRNA 3 7 HBII-295 snoRNA 0 6 U14B snoRNA 0 6 mgU6-77 snoRNA 1 3 U82 snoRNA 0 4 HBII-99B snoRNA 1 2 U33 snoRNA 1 2 U35B snoRNA 0 3 U56 snoRNA 1 2 HBII-234 snoRNA 0 2 snR38B snoRNA 0 2 U16 snoRNA 0 2 U34 snoRNA 0 2 U42B snoRNA 0 2 U45A snoRNA 0 2 U80 snoRNA 0 2 U83 snoRNA 0 2 HBII-180C snoRNA 1 0 HBII-316 snoRNA 0 1 HBII-336 snoRNA 1 0 mgU6-47 snoRNA 0 1

117 Manuscript Dmytro Ustianenko

Gene Symbol BioType CrosslinkCount_1 CrosslinkCount_2

SNORD119 snoRNA 0 1 snR38A snoRNA 1 0 U13 snoRNA 0 1 U15A snoRNA 1 0 U36A snoRNA 1 0 U37 snoRNA 1 0 U43 snoRNA 1 0 U46 snoRNA 0 1 U48 snoRNA 0 1 U50 snoRNA 0 1 U58C snoRNA 0 1 U61 snoRNA 0 1 RNU2-2 snRNA 2 9 RNU2-6P snRNA 2 5 RNU4-2 snRNA 3 4 RNU5E-1 snRNA 4 3 U2 snRNA 1 3 RNU11 snRNA 1 2 RNU12 snRNA 0 3 U1 snRNA 1 0 chr6.tRNA74-LeuCAA tRNA 11 7 chr10.tRNA6-ValTAC tRNA 8 7 chr17.tRNA23-ArgCCG tRNA 7 8 chr6.tRNA141-LeuCAA tRNA 6 9 chr11.tRNA8-SerGCT tRNA 6 6 chr6.tRNA149-LysTTT tRNA 7 5 chr5.tRNA15-ValAAC tRNA 7 4 chr19.tRNA8-SeC(e)TCA tRNA 7 3 chr1.tRNA9-ArgTCT tRNA 5 5 chr6.tRNA105-AlaAGC tRNA 5 5 chr11.tRNA16-ValTAC tRNA 6 3 chr12.tRNA5-AspGTC tRNA 5 4 chr14.tRNA3-ProTGG tRNA 4 5 chr6.tRNA83-LeuTAA tRNA 3 6 chr12.tRNA2-SerCGA tRNA 5 3 chr14.tRNA4-ThrTGT tRNA 4 4 chr17.tRNA29-CysGCA tRNA 6 2 chr1.tRNA56-ThrTGT tRNA 4 4 chr6.tRNA113-AlaTGC tRNA 3 5 chr6.tRNA131-GlnCTG tRNA 4 4 chr6.tRNA143-LysTTT tRNA 5 3 chr12.tRNA12-AspGTC tRNA 4 3

118 Dmytro Ustianenko Manuscript

Gene Symbol BioType CrosslinkCount_1 CrosslinkCount_2 chr14.tRNA13-LysCTT tRNA 3 4 chr17.tRNA18-ArgCCT tRNA 6 1 chr19.tRNA13-ValCAC tRNA 6 1 chr3.tRNA7-CysGCA tRNA 4 3 chr6.tRNA139-ValAAC tRNA 4 3 chr6.tRNA152-ValCAC tRNA 5 2 chr6.tRNA69-ThrAGT tRNA 4 3 chr7.tRNA3-ArgCCT tRNA 4 3 chr10.tRNA2-SerTGA tRNA 3 3 chr15.tRNA4-ArgTCG tRNA 3 3 chr16.tRNA27-LeuTAG tRNA 2 4 chr1.tRNA47-AsnGTT tRNA 0 6 chr5.tRNA9-LysCTT tRNA 5 1 chr6.tRNA108-AlaAGC tRNA 3 3 chr6.tRNA120-AlaAGC tRNA 4 2 chr6.tRNA14-TyrGTA tRNA 3 3 chr6.tRNA64-GlnTTG tRNA 4 2 chr6.tRNA70-AlaCGC tRNA 3 3 chr11.tRNA12-ProTGG tRNA 2 3 chr12.tRNA8-AlaTGC tRNA 3 2 chr14.tRNA2-LeuTAG tRNA 3 2 chr16.tRNA15-ThrCGT tRNA 3 2 chr16.tRNA2-ArgCCT tRNA 3 2 chr2.tRNA13-AlaCGC tRNA 3 2 chr5.tRNA13-ThrTGT tRNA 3 2 chr6.tRNA100-LeuCAA tRNA 2 3 chr6.tRNA101-AlaAGC tRNA 2 3 chr6.tRNA65-AlaAGC tRNA 1 4 chr6.tRNA73-ArgCCG tRNA 3 2 chr6.tRNA85-PseudoTTC tRNA 2 3 chr14.tRNA12-PseudoTTT tRNA 2 2 chr14.tRNA16-TyrGTA tRNA 1 3 chr14.tRNA21-ThrTGT tRNA 3 1 chr16.tRNA11-ProAGG tRNA 1 3 chr17.tRNA12-TrpCCA tRNA 2 2 chr17.tRNA21-ArgCCT tRNA 3 1 chr2.tRNA3-AlaAGC tRNA 3 1 chr6.tRNA134-LeuTAA tRNA 1 3 chr6.tRNA166-AlaAGC tRNA 1 3 chr12.tRNA10-AspGTC tRNA 2 1 chr14.tRNA19-TyrGTA tRNA 0 3 chr15.tRNA9-HisGTG tRNA 3 0

119 Manuscript Dmytro Ustianenko

Gene Symbol BioType CrosslinkCount_1 CrosslinkCount_2 chr16.tRNA22-MetCAT tRNA 1 2 chr16.tRNA32-LysCTT tRNA 1 2 chr16.tRNA34-GlyCCC tRNA 2 1 chr17.tRNA10-GlyTCC tRNA 1 2 chr17.tRNA19-ArgTCG tRNA 3 0 chr17.tRNA2-LysTTT tRNA 1 2 chr17.tRNA41-SerCGA tRNA 2 1 chr17.tRNA5-GlyGCC tRNA 2 1 chr1.tRNA119-LysCTT tRNA 0 3 chr1.tRNA85-ValCAC tRNA 0 3 chr2.tRNA20-GluTTC tRNA 0 3 chr5.tRNA22-AspGTC tRNA 2 1 chr6.tRNA115-ValAAC tRNA 2 1 chr6.tRNA119-AlaCGC tRNA 2 1 chr6.tRNA127-ThrTGT tRNA 3 0 chr6.tRNA146-GlnCTG tRNA 1 2 chr6.tRNA57-IleAAT tRNA 1 2 chr6.tRNA5-SerAGA tRNA 1 2 chr16.tRNA23-LysTTT tRNA 0 2 chr17.tRNA28-CysGCA tRNA 0 2 chr17.tRNA42-LeuTAG tRNA 1 1 chr1.tRNA67-LeuCAG tRNA 2 0 chr1.tRNA86-ArgTCT tRNA 1 1 chr2.tRNA25-PseudoCTC tRNA 1 1 chr2.tRNA2-TyrGTA tRNA 0 2 chr5.tRNA12-ValAAC tRNA 1 1 chr6.tRNA125-ThrCGT tRNA 2 0 chr6.tRNA140-LeuCAA tRNA 0 2 chr6.tRNA40-ValTAC tRNA 1 1 chr6.tRNA46-SerAGA tRNA 1 1 chr6.tRNA80-IleAAT tRNA 0 2 chr6.tRNA87-GluCTC tRNA 0 2 chr8.tRNA4-TyrGTA tRNA 1 1 chr11.tRNA4-LeuTAA tRNA 0 1 chr12.tRNA4-AspGTC tRNA 0 1 chr14.tRNA9-AlaAGC tRNA 0 1 chr15.tRNA3-CysGCA tRNA 1 0 chr15.tRNA7-GlnCTG tRNA 1 0 chr16.tRNA31-PseudoTGG tRNA 1 0 chr16.tRNA7-LysCTT tRNA 1 0 chr17.tRNA14-ThrCGT tRNA 1 0 chr17.tRNA16-GlnTTG tRNA 1 0

120 Dmytro Ustianenko Manuscript

Gene Symbol BioType CrosslinkCount_1 CrosslinkCount_2 chr17.tRNA35-SerAGA tRNA 0 1 chr17.tRNA39-TrpCCA tRNA 1 0 chr17.tRNA40-ThrAGT tRNA 0 1 chr17.tRNA8-ThrAGT tRNA 1 0 chr19.tRNA10-IleTAT tRNA 1 0 chr1.tRNA116-GluCTC tRNA 1 0 chr1.tRNA7-AsnGTT tRNA 1 0 chr2.tRNA5-IleTAT tRNA 1 0 chr3.tRNA11-ArgACG tRNA 0 1 chr6.tRNA106-PheGAA tRNA 1 0 chr6.tRNA110-AlaTGC tRNA 1 0 chr6.tRNA117-AlaCGC tRNA 1 0 chr6.tRNA137-SerCGA tRNA 1 0 chr6.tRNA145-SerAGA tRNA 0 1 chr6.tRNA148-SerTGA tRNA 0 1 chr6.tRNA155-LeuTAA tRNA 0 1 chr6.tRNA15-TyrGTA tRNA 0 1 chr6.tRNA172-SerTGA tRNA 0 1 chr6.tRNA30-ProCGG tRNA 1 0 chr6.tRNA34-ThrAGT tRNA 1 0 chr6.tRNA55-IleTAT tRNA 0 1 chr6.tRNA59-IleAAT tRNA 0 1 chr6.tRNA61-MetCAT tRNA 1 0 chr6.tRNA63-IleTAT tRNA 1 0 chr6.tRNA67-AlaAGC tRNA 1 0 chr7.tRNA1-TrpCCA tRNA 1 0 chr7.tRNA5-CysGCA tRNA 1 0 chr8.tRNA10-MetCAT tRNA 0 1 chr8.tRNA11-SerAGA tRNA 0 1

121 Manuscript Dmytro Ustianenko

Table S2

Gene Symbol GOterm ARF1 GO:0012505 ARHGAP32 GO:0005783,GO:0012505 ARL6IP1 GO:0005783,GO:0012505 ATP2A2 GO:0005783,GO:0012505 CANX GO:0005783,GO:0012505 CD63 GO:0012505 DHCR7 GO:0005783,GO:0012505 DOLK GO:0005783,GO:0012505 EIF5A GO:0005783,GO:0012505 ERGIC1 GO:0005783,GO:0012505 ERP29 GO:0005783 HSD17B10 GO:0005783 HSPA1A GO:0005783 HSPA5 GO:0005783,GO:0012505 KDELR1 GO:0005783,GO:0012505 KDELR2 GO:0005783,GO:0012505 KPNA6 GO:0012505 LAPTM4B GO:0012505 MANEA GO:0012505 MATR3 GO:0012505 RAB14 GO:0005783,GO:0012505 SCD GO:0005783,GO:0012505 SERINC1 GO:0005783,GO:0012505 SNTB2 GO:0012505 SPTLC2 GO:0005783,GO:0012505 TAP2 GO:0005783,GO:0012505 TMBIM6 GO:0005783 TMED2 GO:0005783,GO:0012505

122

Summary Dmytro Ustianenko

Summary

In the current work I have used a genome wide and biochemical in vivo and in vitro methods to characterize a third mammalian homolog of yeast Dis3 exonuclease. We have identified DIS3L2 during our search of proteins that binds to oligouridylated RNA. We used RNA that consists of 30 uridines (U30) and random sequence RNA as bait. Whole cell lysates were prepared from the mouse embryonic stem cells and applied to the RNA. Consequent MS/MS analysis has identified a DIS3L2 being highly enriched on the U30 RNA together with already known poly(U) RNA binders (HUR, hnRNP U, AUF1). Prior to this study, DIS3L2 function and its role in the cell were poorly inves- tigated. Our immunolocalization studies have shown that the protein has a cytoplasmic residence. DIS3L2 is the only human homolog that does not possess a PIN domain which is necessary for the interaction with the major exosome core complex. We performed a series of the co-immunoprecipitation (IP) experiments and confirmed that DIS3L2 is not associated with exosome complex. The recombinant protein, expressed and purified from the cells of the bacteria E.coli had a high RNA binding affinity and preference for the uridine enriched RNA that was able to degrade effectively from 3' to 5' end. Process of the miRNA uridylation is well characterized and was shown to be essential for maintaining the pluripotent state of the cells. Using the RNA-IP studies we have shown that DIS3L2 specifically recognizes the uridylated precursors of the let-7 miRNA family members. Using the in vitro degradation system we were able to show, that pre-miRNA uridylation stimulates the activity of DIS3L2 which results in the higher degradation rate of this particular miRNA. In this way we have contributed and completed the uridylation mediated miRNA degradation pathway in vertebrates. By applying a CLIP technology to study the array of in vivo targets of DIS3L2 we have identified a previously unknown mechanism of the tRNA degradation and potentially a quality control in the cytoplasm of the human cells. This mechanism involves the 3' terminal uridine transferase that post- transcriptionally modifies the precursor or endonucleolitycaly cleaved tRNAs.

124 Dmytro Ustianenko Summary

Such a modification serves as a hallmark for DIS3L2 exonuclease that eliminates these transcripts. Certain experiments also indicate that DIS3L2 is potentially involved in the production of tRNA derived fragments that are involved in the translation regulation. All together, the novel CLIP analysis of DIS3L2 and biochemical studies reveals its global role in the degradation of various RNA transcripts. All obtained data are bringing new insights to the mechanisms of cytoplasmic RNA degradation and uridylation processes, which are determining RNA stability, translation potential and differentiation of human cells.

125 Curriculum vitae Dmytro Ustianenko

Curriculum vitae

Dmytro Ustianenko Masaryk University CEITEC - Central European Institute of Technology Kamenice 5/A4 625 00 Brno Czech Republic Personal e-mail: [email protected] University e-mail: [email protected] +420 605 56 3996

General:

Name: Dmytro Surname: Ustianenko Place of birth: Kiev, Ukraine Date of birth: October 05, 1985 Citizenship: Ukraine

Education:

2008 – present Ph.D. in Biomolecular Chemistry Central European Institute of Technology, Brno, Czech Republic Thesis Topic: “The role of DIS3L2 in the degradation of the uridylated RNA species in humans.” Supervisor: Assoc. Prof. Mgr. Štěpánka Vaňáčová, Ph.D. Expected graduation March 2014 2006 – 2008 Master Degree National Chung Hsing University, Taichung, Taiwan Thesis Topic: “Transgenic tobacco with resistance to potyvirus generated by artificial miRNAs targeting at highly conserved regions of potyviral genomes Graduated with Honours Supervisor: Prof. Shyi-Dong Yeh, Ph.D.

2006 – 2008 Master Degree Kiev National Taras Shevchenko University, Kiev, Ukraine Specialization: Virology, Microbiology, Immunology Thesis Topic:“Transgenic tobacco with resistance to potyvirus generated by artificial miRNAs targeting at highly conserved regions of potyviral genomes.” Supervisor: Prof. Polishuk V.P. Ph.D.

126 Dmytro Ustianenko Curriculum vitae

2002 – 2006 Bachelor of Biological Science and Virology Kiev National Taras Shevchenko University, Kiev, Ukraine Specialization: General biology, Virology Thesis Topic: “Discovery of antiviral activity heterometallic coordinate compounds on the TMV-plants of tobacco model system.” Graduated with Honours Supervisor: Prof. Polishuk V.P. Ph.D.

Research:

2008 - present Laboratory of RNA Processing and Degradation CEITEC - Central European Institute of Technology (Brno, Czech Republic) „The role of DIS3L2 in the degradation of the uridylated RNA species in humans.” Functional characterisation of DIS3L2, human homolog of yeast exosome component DIS3p.

2006 – 2008 Laboratory of molecular Plant Virology, National Chung Hsing University (Taichung, Taiwan) “Using the artificial miRNA for achieving a viral resistance in plants.” Devel- opment and design of amiRNA against conserved regions of CP, CI and Nib genes of Potyviruses. Demonstration of transgenic plants triple amiRNAs generation from pre-amiRNAs—amiR- CI159 , amiR-Nib159, and amiR- CP159 with sequences complementary to CI, Nib and CP conserved coding sequence. Transgenic plant production.

2003 - 2004 D. K. Zabolotny Institute of Microbiology and Virology, (Kiev, Ukraine) Identification of novel microorganisms in the arctic soil and water samples.

2000 - 2002 The Institute of Hydrobiology of NAS of the Ukraine, (Kiev, Ukraine) Involved in project of investigating the water pollution of Dnieper river with oil products and their derivatives.

127 Curriculum vitae Dmytro Ustianenko

Practical skills:

• RNA techniques: RNA protein interaction using CLIP, PAR-CLIP, RNA immunoprecipitation. Consequent sample preparation for the deep sequencing. Northern blotting. • Gene expression techniques: Quantitative PCR analysis of mRNA, miRNA and long non-coding RNAs. • In vivo protein methods: Microscopy and protein immunostaining, protein- protein interaction (co-immunoprecipitation), Western blotting, enzymatic assays, preparation of the sample for MS analysis. • Recombinant protein techniques: expression and purification of recombinant proteins from E.coli. Size exclusion and affinity chromatography. Sample preparation for the NMR spectroscopy. Functional in vitro assays with recombinant proteins (EMSA, degradation assays, modification of nucleic acids, complex reconstitution). Antibody purification. • Molecular cloning techniques: DNA cloning, isolation, purification, cDNA library preparation. RNA isolation, fractionation, RNA in vitro transcription. • Tissue culture and cell lines: FACS analysis, luciferase assay, siRNA and miRNA mediated gene down regulation, plasmid transfection. Cells synchro- nization. Establishment of the stable cell lines.

Awards and scholarships:

2012 Czech RNA Club Poster Competition: Best poster

2008 - 2012 Scholarship of Masaryk University for Doctoral Students and Scholarship for International Students

2006 - 2008 National Chung Hsing University Scholarship for Foreign students

2001 Kiev Junior Scientist Research Project Competition: Gold Medal

2000 Kiev Junior Scientist Research Project Competition: Silver Medal

Teaching and administration:

2008 - 2013 Junior lab member’s supervision 2009 RNA Club Conference, Brno, Czech Republic: part of organiza- tional team

Language skills:

Ukrainian, Russian (native), English (fluent), Czech, Slovak (good), Chinese man- darin, German (Basic)

128 Dmytro Ustianenko Curriculum vitae

List of conferences

2013 Vienna RNA Meeting 2013, Vienna, Austria Talk: “DIS3L2 is the key component of the mammalian cytoplasmic uridylation mediated RNA degradation pathway”

2013 RNA Society Meeting, Davos, Switzerland Poster: “Human DIS3L2 exonuclease is invoved in the processing of tRNA- derived small RNA’s”

2013 Eukaryotic RNA Turnover: From Structural Insights to Diseases, Stras- bourg, France Poster: “Human DIS3L2 exonuclease is invoved in the processing of tRNA- derived small RNA’s”

2012 RNA Club, Prague, Czech Republic Talk: “Human DIS3L2 exonuclease is involved in processing of tRNA derived small RNA’s”

2012 Life and Death of mRNA in the Cytoplasm, Riva del Garda, Italy Poster: “Functional and biochemical characterization of DIS3L2, the third mammalian homolog of yeast exosomal nuclease Dis3p“

2011 RNA Society Meeting, Kyoto, Japan Poster: “Functional and biochemical characterization of DIS3L2, the third mammalian homolog of yeast exosomal nuclease Dis3p“

2010 RNA Club, Ceske Budejovice, Czech Republic Talk: “Functional and biochemical characterization of DIS3L2, the third mammalian homolog of yeast exosomal nuclease Dis3p”

2008 RNA Club, Prague, Czech Republic Talk: “Transgenic tobacco with resistance to potyvirus generated by artificial miRNA targeting at highly conserved regions of potyviral genomes”

129 Curriculum vitae Dmytro Ustianenko

List of publications

1. Ustianenko D., Hrossova D., Potesil D., Chalupnikova K., Pachernik J., Cetkovska K., Uldrijan S. , Zdrahal Z., Vanacova S. “Mammalian DIS3L2 exoribonuclease is targeting uridylated precursors of let-7 miRNAs”. RNA, 2013.

2. Ustianenko D. , Bilen B., Chalupnikova K., Feketova Z., Martin G., Hrazdilo- va K., Zavolan M., Vanacova S. “Human DIS3L2 exonuclease is involved in the processing of tRNA-derived small RNAs” in preparation

3. Krulko I., Ustyanenko D., Polischuk V. “Role of siRNAs and miRNAs in the processes of RNA-mediated gene silencing during viral infection”. Cytology and Genetics 2009, v 43, number 1, 78-90

4. Korniichuk I., Ustyanenko D., Kot T., Kharina A., Polishuk V. “Influence of heteropolynuclear coordinate compounds of metals and salts of heavy metals on the development of the viral infection in plants of Nicotiana tabacum”.Annals of Mechnicov Inst., No 3, 2007

130