Hyperphosphorylation Repurposes the CRL4B E3 Ligase to Coordinate Mitotic Entry and Exit

Research Collection

Doctoral Thesis

Identification and Characterization of Novel Cullin 4-based E3 Ligases in Somatic Cell Division

Author(s): da Graça Gilberto, Samuel Filipe

Publication Date: 2017

Permanent Link: https://doi.org/10.3929/ethz-b-000219014

Rights / License: In Copyright - Non-Commercial Use Permitted

This page was generated automatically upon download from the ETH Zurich Research Collection. For more information please consult the Terms of use.

ETH Library

DISS. ETH NO. 24583

Identification and characterization of novel Cullin 4-based E3 ligases in somatic cell division

A thesis submitted to attain the degree of

DOCTOR OF SCIENCES of ETH ZURICH (Dr. sc. ETH Zurich)

Presented by SAMUEL FILIPE DA GRAÇA GILBERTO

MSc in Biochemistry, University of Lisbon Born on 21.02.1988

Citizen of Portugal

Accepted on the recommendation of Prof. Dr. Matthias Peter Prof. Dr. Anton Wutz

2017

Table of contents

1. General introduction ...... 1 1.1. The ubiquitylation machinery ...... 1 1.2. Principles of cell cycle regulation: a focus on CRLs and the APC/C ...... 3 1.3. Cullin-4 RING E3 ligases: cell-cycle regulation and beyond ...... 14 1.4. Functional distinctions between CUL4A and CUL4B ...... 22 1.5. Open questions and objectives ...... 29 1.6. References ...... 31 2. Article: CRL4RBBP7 is required for efficient CENP-A deposition at centromeres ...... 46 2.1. Supplementary figures ...... 61 3. Manuscript in preparation: Hyperphosphorylation repurposes the CRL4B E3 ligase to coordinate mitotic entry and exit ...... 66 3.1. Supplementary Figures ...... 100 4. Investigation of the role of CRL4WDTC1 in cell cycle progression ...... 117 4.1. Introduction ...... 117 4.2. Results ...... 118 4.3. Discussion ...... 135 4.4. Materials and Methods ...... 137 4.5. References ...... 138 5. Extended discussion and future perspectives ...... 141 5.1. Mitotic function of CUL4B: a molecular explanation for XLID? ...... 142 5.2. CUL4B loss-of-function in mouse embryonic lethality, syndromic XLID and mitosis: how do they come together? ...... 143 5.3. References ...... 146 6. Appendices...... 149 7. Publications ...... 168 7.1. CRL4RBBP7 is required for efficient CENP-A deposition at centromeres ...... 168 7.2. Dynamic ubiquitin signaling in cell cycle regulation ...... 168 8. Acknowledgements ...... 182 9. Curriculum vitae ...... 184

Cover: crystal structure of CUL4B (front cover) and CUL4A (back), PDB ID: 2HYE and 4A0L. ii

Abstract

Cell proliferation requires tightly coordinated alternating phases of DNA synthesis and separation, collectively termed the cell division cycle. The cell cycle makes use of timely regulation of protein activity to coordinate its processes and transitions. A pivotal post-translational modification central in this regulation is ubiquitylation. Specificity in ubiquitylation reactions depends on an E3 ligase, such as the Cullin-4 RING E3 ligase (CRL4) family, which use cullin-4A (CUL4A) or CUL4B as a scaffolding subunit. These cullins are regarded as acting redundantly in the regulation of multiple cellular processes, in particular DNA-related functions such as chromatin remodeling and DNA repair. CRL4 can assemble a multiplicity of different complexes that function as independent E3 ligases, by recruiting an array of substrate adaptor subunits. Evidence suggests that CRL4 regulates the cell cycle, but through mechanisms that remain mostly unexplored. It is additionally possible that CUL4A and CUL4B do not always assemble identical complexes. CUL4B loss-of-function mutations in humans result in the development of syndromic X-linked intellectual disability, and in mice the deletion of the Cul4b gene is embryonically lethal. Remarkably, CUL4A has not yet been associated with human disorders and its deletion is not lethal in mice. Therefore, the identity of CRL4 complexes which function in development, and in the cell cycle in particular, remains largely obscure. We investigated regulatory functions of CRL4 complexes on the cell cycle of HeLa cells, a cancer cell line widely used in cell cycle studies. By making use of siRNA-mediated mRNA depletion and gene deletion by CRISPR/Cas9 to modulate the activity of the respective E3 ligases, we discovered that indeed previously undescribed CRL4 complexes regulate important proliferative functions. In particular, we identified that the substrate adaptor Retinoblastoma-binding protein 7 (RBBP7) works together with CRL4 in the determination of the centromere. Specifically, CRL4RBBP7 participates in the process of loading the Histone H3 variant Centromere protein-A (CENP-A) at centromeric nucleosomes. As a consequence, cells have problems in aligning and separating their chromosomes during mitosis. While functions of CRL4 with RBBP7 are expected to depend on both CUL4A and CUL4B, we noticed that CUL4B operates in mitosis in a manner distinct from CUL4A. Indeed, knockout of CUL4B alone delays mitotic entry by approximately 2h, even in the presence of CUL4A, indicating that CRL4B is necessary for early mitotic stages. Because these cullins share extensive sequence identity that only significantly differs in an extended N-terminal sequence in CUL4B, we examined the possibility that this sequence is responsible for the functional distinction. Indeed, we found that the N-terminus of CUL4B is phosphorylated during mitosis and that this modification alters the subcellular localization of this cullin, which is no longer recruited to the chromatin. Our data supports a model where CUL4B is redirected from chromatin to ubiquitylate a yet unidentified

iii

Abstract substrate. Importantly, this mitotic function cannot be maintained outside of the context of mitosis and results in lethal phenotype in cells. A third independent involvement of CRL4 complexes in the regulation of the cell cycle comes from the identification of another substrate adaptor, WDTC1, which interacts with a mitotic regulator during in the late cell cycle stages, although the functional consequences of WDTC1 loss-of-function remain to be identified. Together, these studies establish newly-identified roles for CRL4 complexes in regulating cell proliferation. By placing special focus in the description of mechanisms of cell division, we clarified how CRL4-type E3 ligases participate in this extended functional network. Finally, our studies determine for the first time the mechanisms of a functional distinction between CUL4B and CUL4A, based on an alternative cullin regulatory mode. We aim at describing which novel substrates CRL4 can ubiquitylate at different cell cycle stages. Besides, we believe that our observations set the stage to uncover why CUL4A cannot compensate for CUL4B loss in human XLID patients.

Abstract

Sommario

La proliferazione cellulare si basa sulla precisa coordinazione dell’alternarsi di fasi di sintesi del DNA e di divisione, collettivamente chiamate il ciclo di divisione cellulare. Il ciclo cellulare utilizza la tempestiva regolazione dell’attività delle proteine per coordinare i suoi processi e le sue transizioni. Una delle modificazioni post-traduzionali più importanti per questa regolazione è l’ubiquitinazione. La specificità delle reazioni di ubiquitinazione dipende da delle ligasi E3, come quelle della famiglia delle Cullin-4 RING E3 ligase (CRL4), che usano cullin-4A (CUL4A) o CUL4B come scaffolding subunit. Si pensa che queste cullins agiscano in modo ridondante nella regolazione di diversi processi cellulari, in particolare nelle funzioni legate al DNA, come il rimodellamento della cromatina o la riparazione del DNA. CRL4 può formare una multitudine di complessi diversi che funzionano come ligasi E3 indipendenti, tramite il reclutamento di specifiche subunità che fungono da adattatore per il substrato. Alcune indicazioni suggeriscono che CRL4 regoli il ciclo cellulare tramite meccanismi che restano però per lo più inesplorati. È inoltre possibile che CUL4A e CUL4B non formino sempre complessi identici. Infatti, mutazioni che causano la perdita di funzionalità di CUL4B negli umani risultano nella disabilità intellettiva sindromica legata al cromosoma X, e nei topi la rimozione del gene Cul4b è letale già a livello embrionale; sorprendentemente invece, CUL4A non è ancora stata associata a nessuna malattia umana e la sua rimozione nei topi non è letale. Quindi, l’identità dei complessi formati da CRL4 che funzionano nello sviluppo, e in particolare nel ciclo cellulare, rimane per lo più oscura. Abbiamo studiato le funzioni regolatorie dei complessi di CRL4 sul ciclo cellulare delle cellule Hela, una linea cellulare cancerosa molto utilizzata degli studi sul ciclo cellulare. Utilizzando la deplezione di mRNA mediata da siRNA e una delezione genetica tramite CRISPR/Cas9 per modulare l’attività delle rispettive ligasi E3, abbiamo scoperto che, come ipotizzato, alcuni complessi di CRL4 mai descritti prima regolano importanti funzioni legate alla proliferazione cellulare. In particolare, abbiamo identificato che l’adattatore per il substrato retinoblastoma-binding protein 7 (RBBP7) collabora con CRL4 per determinare il centromero. In particolare, abbiamo mostrato che CRL4RBBP7 participa nel processo di caricare una variante dell’istone 3, la proteina centromere protein-A (CENP-A), sui nucleosomi centromerici. Di conseguenza, queste cellule hanno problemi ad allineare e separare i loro cromosomi durante la mitosi. Mentre ci si aspetterebbe che le funzioni di CRL4 in collaborazione con RBBP7 dipendano sia da CUL4A che CUL4B, abbiamo notato che durante la mitosi CUL4B opera in modo diverso da CUL4A. Infatti, il knockout di CUL4B ritarda l’entrata in mitosi di circa 2 ore, anche in presenza di CUL4A, indicando che CRL4B è necessaria per le prime fasi della mitosi. Siccome queste cullins hanno una sequenza molto simile

Abstract che differisce significativamente sono nell’estesa sequenza al termine N di CUL4B, abbiamo esaminato la possibilità che questa sequenza sia responsabile per la differenza funzionale. Abbiamo osservato che il termine N di CUL4B è fosforilato durante la mitosi e che questa modificazione altera la localizzazione subcellulare della proteina, che non viene più reclutata alla cromatina. I nostri dati sostengono quindi un modello in cui CUL4B viene ridirezionata dalla cromatina a un altro subtrato tuttora non identificato perché venga ubiquitinato. È importante notare che questa funzione mitotica non può essere mantenuta al di fuori del contesto della mitosi e risulta in un fenotipo letale nelle cellule. Un terzo coinvolgimento indipendente dei complessi formati da CRL4 nella regolazione del ciclo cellulare è stato notato grazie alla scoperta di un altro adattatore per il substrato, WDTC1, che interagisce con un regolatore mitotico durante gli ultimi stadi del ciclo celuulare. Tuttavia, le conseguenze funzionali della perdita di funzionalità di WDTC1 devono ancora essere identificate. In sintesi, il nuostro studio stabilisce nuovi ruoli nella regolazione del ciclo cellulare dei complessi formati da CRL4. Concentrandoci sulla descrizione dei meccanismi della divisione cellulare, abbiamo compreso come le ligasi E3 del tipo di CRL4 partecipino in questo vasto network funzionale. Infine, il nostro lavoro mostra per la prima volta i meccanismi alla base della distinzione funzionale tra CUL4B e CUL4A, basati su una diversa regolazione delle cullins. Il nostro obbiettivo è ora di descrivere quali nuovi substrati possano venire ubiquitinati da CRL4 nei diversi stadi del ciclo cellulare. Inoltre, crediamo che le nostre osservazioni aprano la via per comprendere perché CUL4A non può compensare la perdita di CUL4B nei pazienti che soffrono di XLID.

1. General introduction

1.1. The ubiquitylation machinery

Ubiquitin is a small 76-residue protein modifier that can be covalently attached lysine residues of target proteins, thereby altering their stability, function or localization. Ubiquitin itself contains seven lysine residues, plus its modifiable N-terminus, and therefore can be itself conjugated to other ubiquitin moieties, thereby constituting ubiquitin oligomers or chains (the latter is referred to as polyubiquitylation). The fact that ubiquitin itself is modified can give rise to several types of chain topologies that serve as a scaffold for the recruitment of ubiquitin-binding proteins, which we refer to as “readers”. The most widely studied ubiquitin reader is the proteasome, traditionally known to read K48- linked ubiquitin chains, as well as K11-linked. The 26S proteasome is a proteolytic machinery and ubiquitin-guided binding to its regulatory 19S particle results in the ubiquitylated protein’s degradation (Livneh et al., 2016). Current research determined that reading ubiquitin signal goes far beyond the proteasome. For example, a monoubiquitin moiety or K63-linked chain can recruit factors that allow for a specific localized response, for example the recruitment of a DNA damage-tolerant polymerase to a site of replication stress (García-Rodríguez et al., 2016). In many cases, ubiquitylated proteins first need to be extracted from interacting partners or chromatin, a function typically attributed to the ATPase Valosin- Containing Protein (VCP)/p97 (Cdc48 in yeast) (Meyer et al., 2012; Franz et al., 2016). Importantly, specific proteases termed deubiquitylating enzymes (DUBs) can cleave off ubiquitin moieties and reverse the signal (Lim et al., 2016). Ubiquitylation depends on a cascade of enzymes: an E1 ubiquitin-activating enzyme first binds ubiquitin, which is transferred to an E2 ubiquitin-conjugating enzyme (Figure I-1). The ubiquitin-loaded E2 enzyme binds an E3 ubiquitin ligase for transfer of ubiquitin to the substrate. Because it is the E3 ligase who binds the substrate directly, it provides specificity to this reaction. Three families of E3 ligases have been described, HECT, RING and RBR E3 ligases (Ye and Rape, 2009; Wenzel and Klevit, 2012). The bulk of cell cycle regulation known to date is performed by RING-E3 ligases, especially its major family Cullin-RING E3 ligases (CRLs).

1. General introduction

Figure I-1. The concerted action of E1, E2 and E3 enzymes drive substrate ubiquitylation. The first step is the adenylation of ubiquitin catalyzed by the E1 ubiquitin-activating enzyme, followed by covalent binding of a cysteine residue of the E1 to ubiquitin. Ubiquitin is then transferred to the catalytic cysteine of an E2 ubiquitin-conjugating enzyme. In the case of cullin-RING ligases (depicted), the E3 acts as a scaffold to bring together the E2 and the substrate and thus promote the transfer of ubiquitin to a substrate lysine residue. Of note, the E3 ligase may add additional ubiquitin moieties to a previously attached moiety, thereby establishing a polyubiquitin chain.

1.1.1. Cullin-RING E3 ligases

The modular Cullin-RING ligases (CRLs) comprise the largest group of E3 ligases in eukaryotes and are estimated to assemble up to 600 different complexes (Petroski and Deshaies, 2005). CRLs use one of the six cullin proteins encoded by the human genome as a scaffolding subunit that brings together the ubiquitin-loaded E2 enzyme and the substrate. The E2 enzyme is recruited by the C-terminally bound RING subunit (RBX1 or RBX2). Substrates associate to CRLs via an N-terminal receptor module composed of a variable substrate-specific adaptor and a cullin-bound linker subunit, except for the case of CRL3. CRLs are activated by modification with NEDD8, termed neddylation, and they associate dynamically with regulators that modulate the neddylation state and block substrate access (the COP9 signalosome), or promote substrate receptor release and exchange (CAND1) (Lydeard et al., 2013). CRLs are thus regarded as modular, dynamic assemblies with substrate-specific adaptors that associate and dissociate in a regulated manner to ensure timely and specific substrate ubiquitylation (Craney and Rape, 2013). Within the N-terminal receptor module, specific adaptors are linked to individual cullins (Figure I- 2). SCF E3 ligases contain an F-box protein, CRL3 contains a Broad-complex, Tramtrack, Bric-a-brac (BTB) domain-containing protein, and CRL4 has a DDB1- and CUL4-associated factor (DCAF) protein (Lydeard et al., 2013). Subdivided into CRL2 and CRL5, the Elongin B-C-CUL2/CUL5-SOCS box protein (ECS) E3 ligases recruit BC-box-containing adaptors, in particular VHL-box and SOCS-box proteins (Cai and Yang, 2016).

1. General introduction

Figure I-2. Cullin-RING E3 ligases (CRLs) and the related anaphase-promoting complex/cyclosome (APC/C) E3 ligase. Left, general structure of CRLs and the APC/C. Right, structural distinctions of the substrate adaptor module in different CRLs.

Unlike the other known classes of E3 ligases, RING E3 ligases work by facilitating the direct transfer of ubiquitin from the E2 to the substrate lysine residue. A different E2 enzyme may be used to initiate and elongate a polyubiquitin chain (Deshaies and Joazeiro, 2009; Ye and Rape, 2009). Alternatively, as for a subset of CRLs, an independent E3 ligase may be recruited to catalyze the initiation step (Scott et al., 2016). The E2 enzyme used for chain elongation is the major determinant of ubiquitin chain topology (Deshaies and Joazeiro, 2009; Ye and Rape, 2009). In the case of CRLs, UBCH5 E2 enzymes allow for mono or multi-monoubiquitylation, while CDC34 drives chain extension, forming canonical K48-linked polyubiquitin chains (Lydeard et al., 2013; Grice and Nathan, 2016).

1.2. Principles of cell cycle regulation: a focus on CRLs and the APC/C

The cell cycle is essentially succeeding phases of DNA synthesis (S phase) and separation (mitosis), culminating in the process of cell division. Biochemically, one can view the cell cycle simply as an oscillator of cyclin-dependent kinase 1 (CDK1)-cyclin B kinase activity, brought about by fluctuations in the levels of cyclin B, the positive regulatory subunit. Cyclin B synthesis drives cells into mitosis, whereas its destruction brings cells back to interphase – referred to as mitotic entry and exit, respectively (Morgan and Roberts, 2002). This was a first glimpse of the importance of ubiquitin-mediated proteolysis in cell cycle regulation (Glotzer et al., 1991; Hershko et al., 1991).

1. General introduction

The E3 ligase responsible for cyclin B degradation was termed the anaphase-promoting complex or Cyclosome (APC/C). The APC/C can bind two activators that function as substrate adaptors: CDC20 and CDH1 (Zhou et al., 2016). Because CDK1 activates APC/CCDC20, its own inhibitor, this generates a simple feedback loop that is the basis for the CDK1-cyclin B activity oscillations and hence cell cycle progression (Morgan and Roberts, 2002). Multiple cell cycle transitions critically depend on CRL1 or SKP1-CUL1-F-box (SCF) E3 ligases (Figure I-3), in particular for targeting degradation of CDK inhibitors such as p27 and WEE1 at G1/S and G2/M, respectively. CRL4 complexes have been described for their functions in preventing DNA re- replication, while CRL3 is probably the most emergent CRL in cell cycle control, in particular by regulating mitosis. Several non-proteolytic functions of CRL3 and CRL4 complexes are now attributed (Bassermann et al., 2014; Teixeira and Reed, 2013). DNA replication is one of the few cell cycle functions currently attributed to ECS E3 ligases (Dewar et al., 2017). The APC/C is related to the CRL class, because it contains the cullin-homology subunit APC2 (Yu et al., 1998).

Figure I-3. Functions of CRLs and the APC/C in the cell cycle. On the cell cycle core is illustrated the oscillations of cyclin A and B, which together with CDK2 and CDK1 are the main mediators of the G1/S an G2/M transitions, respectively. Green, functions attributed to SCF/CRL1; Yellow, functions of CRL3; Blue, functions of CRL4; Red, functions of the APC/C.

1. General introduction

1.2.1. Driving the cell cycle: ubiquitin in cell cycle transitions

The directionality of the cell cycle is defined by ordered and irreversible transitions between cycle stages, brought about by all-or-nothing bistable activity switches of major cell cycle players, especially that of CDKs – a topic of past and current research (Barr et al., 2016; Hégarat et al., 2016). This behavior depends greatly on CRLs and the APC/C.

Growth decisions and the G1/S transition. If growth conditions are not suitable (lack of nutrients or growth signals or after genotoxic stress), cells enter a temporary non-proliferative state referred to as quiescence or G0. In mammals, new data argues that the decision to commit to the cell cycle in presence of mitogens is made during mitosis or at the G0/G1 transition, the restriction window and restriction point, respectively, in line with new understandings (Spencer et al., 2013; Cappell et al., 2016). The maintenance of a quiescent state relies in part on the inactivation of growth signaling pathways by degradation of their components (Teixeira and Reed, 2013; Bassermann et al., 2014). The restriction point is switch-like (Yao et al., 2008) and requires the inhibition of activator E2F transcription factors to be lifted, thereby activating S phase gene transcription. The retinoblastoma protein RB is the E2F inhibitor, inactivated by the so-called G1 cyclins (D- and E-type) when bound to certain CDK family members (Figure I-4) (Bertoli et al., 2013; Matson and Cook, 2016), but a growing body of evidence shows that RB family members are subject to degradation to facilitate the transition to S phase (Sengupta and Henry, 2015; Wang et al., 2015). DNA replication is initiated by CDKs, with CDK2-cyclin A (human cyclin A2 gene) as the likely main driver of its commencement (Pagano et al., 1992). This kinase is kept inhibited during G1, brought about by the upregulation of CDK inhibitors (Toyoshima and Hunter, 1994; Wade Harper et al., 1993) and the activation of the E3 ligase APC/CCDH1. This E3 ligase takes over APC/CCDC20 during mitotic exit to continue promoting the degradation of cyclins throughout G1, but is inactivated in S phase (Yeong et al., 2000). Two aspects are now considered as drivers of the G1/S transition: 1) the inactivation of APC/CCDH1 (Cappell et al., 2016) and 2) the degradation of the CDK inhibitor p27KIP1 (Figure I-4)(Barr et al., 2016) – brought about by SCF, with SKP2 functioning as the SCF substrate adaptor (Tsvetkov et al., 1999; Carrano et al., 1999). Both events relieve CDK2-cyclin A inhibition, but they are intrinsically connected because APC/CCDH1 targets SKP2 for degradation (Bashir et al., 2004; Wei et al., 2004). Consequently, APC/CCDH1 inactivation defines the point-of-no-return in entry into S phase (Cappell et al., 2016). The inhibition of APC/CCDH1 to drive the G1/S transition is further consolidated by a number of mechanisms, including the direct APC/C inhibitor EMI1 and that CDK2-cyclin E and CDK2-cyclin A inhibit this E3 ligase (Lukas et al., 1999; Hsu et al., 2002; Cappell et al., 2016). Further, cyclin A is stabilized by the

1. General introduction

USP37 and Ubp15 deubiquitylating enzymes in mammals and yeast, respectively (Huang et al., 2011). Ubiquitin-mediated proteolysis also takes an important part in inactivating APC/CCDH1, with the action of SCF complexes, namely SCFβTRCP and SCFcyclin F (Fukushima et al., 2013; Choudhury et al., 2016), which is clear evidence that APC/CCDH1 must be swiftly and permanently inhibited. Remarkably, because during G1 APC/CCDH1 targets cyclin A and cyclin F for degradation, these networks define several double-negative feedback loops that are a characteristic of switch-like irreversible transitions.

The G2/M transition. During S and G2 phases, the inactivity of the APC/C allows the accumulation of cyclin B in preparation for the next mitosis (Lukas et al., 1999), but CDK1-cyclin B is kept inhibited by two kinases, WEE1 and MYT1. Entry into mitosis requires the dephosphorylation of their inhibitory sites on CDK by CDC25 phosphatases and the destruction of WEE1 (Figure I-4) (Teixeira and Reed, 2013; Bassermann et al., 2014; Wieser and Pines, 2015). Both processes depend on CDK1 itself, but also on other kinases, such as Polo-like kinase 1 (PLK1) (Toyoshima-Morimoto et al., 2002; Lobjois et al., 2009; Watanabe et al., 2005): a major kinase not only in mitotic entry, but for several subsequent mitotic processes (Zitouni et al., 2014). Incomplete DNA replication or the presence of DNA damage prevent CDK1-cyclin B activation. When the damage is resolved, PLK1 is required to undermine the checkpoint signal (Zitouni et al., 2014). An important concept is that PLK1 inactivation maintains the G2/M checkpoint, achieved through 1) the untimely re-activation of APC/CCDH1 to induce its degradation (Bassermann et al., 2008; de Boer et al., 2016) and 2) destruction of the PLK1 activator Bora, mediated by SCFβTRCP (Qin et al., 2013). Also monitored during G2 is DNA catenation (entangled or interlinked chromatids). Decatenation is performed before and during mitosis by Topoisomerase IIα (TOP2A), an enzyme that can hydrolyze dsDNA (Broderick and Niedzwiedz, 2015). The decatenation checkpoint monitors the entanglement of DNA and acts by inhibiting PLK1 (Deming et al., 2002), and is itself orchestrated by TOP2A (Luo et al., 2009). Important in the checkpoint function of TOP2A is its polyubiquitylation (K63-linked chains) by the E3 ligase RNF168 – counteracted by USP10 (Guturi et al., 2016). Not only RNF168, but also another E3 ligase, BRCA1, promote the decatenation activity of TOP2A (Guturi et al., 2016; Lou et al., 2005), and TOP2A is stabilized by the OTUD3 DUB (Kang et al., 2015) – though the E3 ligase responsible for TOP2A degradation hasn’t been identified.

1. General introduction

Figure I-4. Overview of the mechanisms of cell cycle transitions and influence by the SCF (green) and APC/C (red) in these processes. Please refer to the main text for details.

The metaphase-to-anaphase transition. Unlike the cell cycle transitions mentioned above, where CDK activation is the driver, anaphase onset is triggered by the activation of the APC/CCDC20 E3 ligase. The degradation of cyclin B (human cyclin B1) results in the inhibition of CDK1, and it is this step that is responsible for the switch-like transition into anaphase and back to interphase (Figure I-4). Inefficient degradation of cyclin B1 renders the metaphase-to-anaphase transition reversible (Potapova et al., 2006; Vázquez-Novelle et al., 2014; Rattani et al., 2014; Clijsters et al., 2014). Paradoxically, cyclin B1 degradation does not strictly require its polyubiquitylation, and multi-monoubiquitylation appears sufficient to target it to the proteasome (Garnett et al., 2009; Dimova et al., 2012) – evidence that the signals that promote recruitment to the proteasome are more diverse than previously thought. The efficiency of cyclin B1 degradation is enhanced by prior binding to APC/C (Voets and Wolthuis, 2015) and by the concomitant degradation of USP22, a DUB that shields cyclin B1 in earlier mitotic stages (Lin et al., 2015). Although cyclin B1 is only degraded as cells enter anaphase, APC/CCDC20 is already activated by CDK1-cyclin B itself and PLK1 as cells enter mitosis, by mechanisms recently clarified (Kramer et al., 2000; Golan et al., 2002; Qiao et al., 2016; Zhang et al., 2016; Fujimitsu et al., 2016). However, APC/CCDC20 is kept

1. General introduction in a mostly inactive state due to the activation of the spindle assembly checkpoint (SAC), reviewed in (Lischetti and Nilsson, 2015). This checkpoint ensures that during metaphase all kinetochores (structures which assemble at the centromere of each chromatid (Pesenti et al., 2016)) are attached to microtubules emanating from the two opposing mitotic spindle poles – a mechanism necessary for faithful chromatid separation. Unattached kinetochores originate the APC/CCDC20 inhibitor, termed mitotic checkpoint complex (MCC). The nature of the inhibitor and how the kinetochores generate a diffusible signal that can globally inhibit APC/CCDC20 is still not entirely clear and is subject to controversy (Han et al., 2013b; Izawa and Pines, 2014; Kaisari et al., 2016). Nevertheless, not only is APC/CCDC20 inhibited by MCC, but also in a concerted action of PLK1 and the checkpoint kinase BUB1 (Jia et al., 2016). The transition to anaphase requires the disassembly of MCC, promoted by the joint action of the AAA-ATPase TRIP13 and the checkpoint antagonist p31comet (Figure I-4)(Westhorpe et al., 2011; Miniowitz- Shemtov et al., 2015). Regardless, the docking of MCC to the APC/C during metaphase is more dynamic than what we previously believed, because CDC20 (as an integral component of MCC) is polyubiquitylated by APC/C itself due to a peculiar structural arrangement (Nilsson et al., 2008; Ge et al., 2009; Yamaguchi et al., 2016). This polyubiquitylation allows a constant turnover of MCC to generate uninhibited APC/C that can be activated when the SAC is turned off, but it also contributes to keeping APC/C activity low before the SAC is silenced (Varetti et al., 2011; Mansfeld et al., 2011; Foster and Morgan, 2012). CUEDC2, an ubiquitin-binding domain-containing protein that associates with the APC/C, has been implicated in MCC disassembly in humans, but its mechanism of action remains to be clarified (Gao et al., 2011)

1.2.2. Regulation of cell cycle events by ubiquitin

DNA replication: ubiquitin in elongation and termination. CDKs initiate the firing of replication forks, i.e. DNA replication, because together with the DDK kinase they activate the replicative helicase (termed CMG complex). In turn, the CMG helicase recruits several other replication factors, including that of the sliding clamp PCNA (Fragkos et al., 2015; Labib, 2010). PCNA encircles DNA and serves as an interaction platform to tether DNA polymerases and repair factors to chromatin (Moldovan et al., 2007). Ubiquitylation regulates DNA replication itself – including the termination of replication (Bell, 2014) – as well as nucleosome re-formation (García-Rodríguez et al., 2016; Moreno and Gambus, 2015). A concept that is attracting more attention is the importance of ubiquitylation in lagging strand synthesis. This process involves a constant shift in factors, orchestrated by PCNA: Polα/primase synthesizes the RNA- DNA primer, after which PCNA is recruited to promote extension by Polδ (mammalian Polδ4), followed by recruitment of the FEN1 nuclease and finally of DNA ligase I act to connect the Okazaki fragments

1. General introduction

(Moldovan et al., 2007). Work of about a decade ago suggested that the switch between yeast Polα and PCNA, thus Polδ, is dependent on non-degradative ubiquitylation of a replication fork factor involved in this process, MCM10 (Das-Bradoo et al., 2006). Recently, work in human cells demonstrated that the p12 subunit of the replicative polymerase Polδ4 is ubiquitylated by PCNA/CRL4CDT2 and degraded in an unperturbed S phase (Zhang et al., 2013). This process generates Polδ3, which has altered properties and had been associated with gap filling during DNA repair. Hence, a possible view is that its specialized properties (e.g. enhanced proofreading activity) might facilitate lagging strand replication (Figure I-5) (Lee et al., 2014), but this remains to be confirmed. Similar to lagging strand synthesis, in the context of DNA damage or replication stress PCNA acts as a platform for a possibly ordered recruitment of repair factors (Moldovan et al., 2007; Mailand et al., 2013). At least upon UV damage, CRL4CDT2 – recruited by PCNA – is emerging as key in repair factor exchange by inducing the degradation of PCNA-associated factors (Oda et al., 2010; Han et al., 2015). The parallel to the ordered PCNA recruitment of lagging strand synthesis factors is remarkable, and thus perhaps CRL4CDT2 is also involved in local lagging strand factor exchange (even though global protein levels are not affected for PIP-box containing proteins, as is the case for DNA ligase I (Michishita et al., 2011).

Figure I-5. CRLs in DNA replication-related processes: Lagging strand synthesis, nucleosome remodeling and the termination of replication. Red circles, ubiquitin. Sc, Saccharomyces cerevisiae.

1. General introduction

Nucleosomes undergo a dramatic change as the replication fork progresses: they must be evicted from DNA and re-formed as the fork transits (Alabert and Groth, 2012). An important role depends on the Rtt101-containing E3 ligase – the human CUL4 homolog (Figure I-5). Because Rtt101 has been recently determined to be tethered to the replication fork via the substrate adaptor Mms22 (Buser et al., 2016), one exciting hypothesis is that budding yeast uses this resident replication fork E3 ligase to mediate nucleosome remodeling. It would be of high interest to further explore this hypothesis, and to identify a human E3 ligase that behaves similarly. Nucleosome re-formation requires an adequate histone supply (Alabert and Groth, 2012). In humans, a recent report provides evidence of a distinct mechanism CRL4WDR23 promotes core histone mRNA processing – thereby generating mature mRNAs that can be translated (Figure I-5). It appears to do so through multi-monoubiquitylation of the histone mRNA processing protein SLBP, promoting its function instead of inducing its proteolysis. It was hypothesized that SLBP multi-monoubiquitylation might influence the binding to its interacting partners (Brodersen et al., 2016), though additional clarification is required.

Timely shutdown of undesired processes. The function of proteins with cell cycle roles is often restricted to precise time windows. This is because the untimely activation of these proteins can be deleterious to the cell. For example, G1 cyclins are degraded in S phase and the disruption of these mechanisms is often associated with cancer (Teixeira and Reed, 2013; Bassermann et al., 2014). Replication licensing (loading of the pre-replicative complex) takes place during G1, but is undesired during S phase – otherwise, the cell could re-replicate its DNA. CDK2-cyclin A (Coverley et al., 2002) and two E3 ligases, SCFSKP2 and CRL4CDT2, are central in the prevention of origin re-licensing during S phase (Figure I-6) (Siddiqui et al., 2013; Moreno and Gambus, 2015). These E3 ligases share several common substrates, but differ dramatically on the mechanism that provides specificity for S phase: SCFSKP2 is itself regulated by APC/CCDH1 during G1, and SCFSKP2 ubiquitylates its licensing substrates after their phosphorylation by CDK2. CRL4CDT2 substrate targeting requires chromatin-loaded PCNA, because substrates are first recruited to PCNA via the PIP-degron – a specialized PIP-box (Havens and Walter, 2009; Havens et al., 2012). In other words, CRL4CDT2 will only polyubiquitylate these targets in case DNA replication is ongoing. An additional mechanism to prevent licensing factor degradation is through the action of DUBs, as recently unveiled for USP37 acting on the licensing factor CDT1 (Hernández-Pérez et al., 2016). But why the usage of two different systems in S phase? Several reasons might account for this, such as enhanced degradation or the targeting of independent pools of these factors. Moreover, because CRL4CDT2-based replication-coupled destruction is actively inactivated in G2 by CDK1 (Rizzardi et al., 2015),

1. General introduction

Figure I-6. Inhibitory roles of CRLs and the APC/C in diverse cell cycle-related processes.

it is likely that SCFSKP2 has additional importance in this phase (as suggested by (Takeda et al., 2005)), and current evidence also suggests that the cell compensates by G2-specific targeting of the licensing factors CDT1 and CDC6 by other SCF-based complexes (Johansson et al., 2014; Walter et al., 2016). Factors involved in replication itself are downregulated in late cell cycle stages: for example, the lagging strand nuclease FEN1 is destroyed, possibly to avoid its potentially deleterious nuclease activity (Guo et al., 2012). Core histone gene transcription is also reduced by proteolysis of their transcription factor, at least in fission yeast (Moreno and Gambus, 2015). Importantly, recent research is unveiling SCFcyclin F as an important factor in the downregulation of S phase processes (Figure I-6), for instance by triggering the degradation of RRM2 – an enzyme responsible for dNTP synthesis (D’Angiolella et al., 2012), as well as downregulating SLBP – the protein behind histone mRNA synthesis and therefore histone production (Dankert et al., 2016). Not only is DNA duplicated in S phase, but also the centriole – the organizing structure of the centrosome. Importantly, ubiquitin-mediated proteolysis regulates both the timing (S phase) and the extent of duplication (only once). SCF-based complexes (including SCFcyclin F) also play a key role in the regulation of this event (Zhang and Galardy, 2016), with the DUB USP33 allowing normal centriole duplication to take place (Li et al., 2013).

1. General introduction

Another process regulated by ubiquitylation to avoid undesired effects is the response to double- strand breaks (DSBs): the repair mechanism of homologous recombination (HR) is suppressed during G1, restricting HR to when two sister chromatids are present (S and G2 phases). This inhibition depends on the non-degradative ubiquitylation of the HR factor PALB2 by CRL3KEAP1, which averts the recruitment of repair factors. Interestingly, the DUB USP11 counteracts this response, but it is itself targeted for degradation in G1 in a CRL4-dependent manner (Orthwein et al., 2015) – though further clarification on this CRL4 complex composition and regulation is required. Anaphase onset and mitotic exit are also a key examples of how the cell uses ubiquitylation to turn off specific events: APC/CCDH1 is activated after cells enter anaphase to degrade multiple factors involved in mitosis, such as the mitotic kinases Aurora B and PLK1 and spindle assembly factors, alongside the continued degradation of Cyclin B1 (Figure I-6) (Zhou et al., 2016). It is interesting that spindle assembly factors are usually shielded from APC/C-dependent degradation by microtubule binding and are therefore protected in mitosis (Song et al., 2014). Nevertheless, the activation of APC/CCDH1 is preceded by APC/CCDC20, and together these E3 ligases work by removing the central machinery that orchestrates mitosis and thus that would keep cells endlessly in a mitotic state. APC/CCDC20 additionally promotes the separation of sister chromatin cohesion, by targeting securin for degradation. Securin is an inhibitor of the protease separase, who can hydrolyze cohesin – the complex that holds sister chromatids together. Because CDK1-cyclin B1 also inhibits separase (Gorr et al., 2005; Boos et al., 2008; Hellmuth et al., 2015), the degradation of both cyclin B and securin accounts for the segregation of sister chromatids to the two opposing mitotic spindle poles in anaphase and the generation of two daughter interphase cells (Hirano, 2015).

1.2.3. Ubiquitin in the regulation of protein-chromatin association

Polyubiquitylation and degradation of chromatin-associated proteins is a way of limiting their function. However, polyubiquitylation per se doesn’t directly result in the delivery of a protein to the proteasome. First, it is required that the substrate is extracted from the chromatin by the segregase VCP/p97 (yeast CDC48), after which it can be degraded (Franz et al., 2016). This is the case for various chromatin-associated proteins, for example, of the replication licensing factors CDT1 and PR-SET7 (Figure I-7) (Raman et al., 2011; Franz et al., 2011). The termination of replication also depends on this segregase (Figure I-5) (Maric et al., 2014; Moreno et al., 2014). p97 is also important for the correct localization of Aurora B exclusively at the kinetochore, because it removes this kinase from chromosomal arms, very

1. General introduction

Figure I-7. Regulation of protein localization to the chromatin by ubiquitin. Left and middle: upon ubiquitylation by CRL4 and SCF E3 ligases, the segregase p97 is required to remove ubiquitylated chromatin-bound proteins. Right: recruitment of topoisomerase IIα to the chromatin and histone deposition are dependent on ubiquitin. likely following its polyubiquitylation by CRL3KLHL9-KLHL13 (Sumara et al., 2007; Ramadan et al., 2007; Dobrynin et al., 2011). Not only ubiquitylation promotes the removal of a chromatin-bound proteins, but it can rather promote their association (Figure I-7). For instance, TOP2A recruitment to the chromatin is promoted by K63-linked polyubiquitylation by the RNF168 E3 ligase, thereby allowing its function in DNA decatenation and in the decatenation checkpoint – a process counteracted by USP10 (Guturi et al., 2016). Another example is in the deposition of histones on nucleosomes, performed by histone chaperones. During DNA replication, non-proteolytic ubiquitylation of histone chaperones or histones themselves facilitates their deposition, at least in budding yeast (Moreno and Gambus, 2015). Another process that involves histone deposition is the determination of the centromere, where the Histone H3 variant CENP-A (also called CenH3) is incorporated into centromeric nucleosomes. Recently, it was described that the deposition of human CENP-A is itself promoted by ubiquitin, and two recent studies involve CRL4 in this process, though different mechanisms were suggested (Mouysset et al., 2015; Niikura et al., 2015). For most of these processes, it is still unclear how their ubiquitylation induces chromatin recruitment. In the case of histone deposition, it seems to regulate the association with histone chaperones (Moreno et al., 2014; Niikura et al., 2015), but for example how ubiquitylated TOP2A is recruited to the chromatin remains elusive and possibly requires a ubiquitin-binding adaptor chromatin-bound protein. Whether non-degradative ubiquitylation can be a widespread mechanism for timely and local recruitment of proteins to the chromatin remains a highly speculative hypothesis. Nevertheless, it would certainly be of primary interest to identify chromatin-binding proteins that allow the recruitment of ubiquitylated factors.

1. General introduction

1.3. Cullin-4 RING E3 ligases: cell-cycle regulation and beyond

CRL4-type E3 ligases use as the scaffold cullin one of the CUL4A or CUL4B paralogs. Both cullins exist in vertebrates and share extensive sequence homology. CRL4 (CUL4-RBX1-DDB1-substrate adaptor) adopts a U-shaped conformation which brings the substrate and the ubiquitin-loaded E2 together (Angers et al., 2006; Fischer et al., 2011). The cullin contains an arc-shaped helical amino-terminal domain (NTD) which binds DDB1, the protein that works as the linker between the cullin and the substrate adaptor. The cullin C-terminal region is globular and integrates RBX1, the RING-domain-containing protein which recruits the E2 enzyme (Figure I-8) (Lydeard et al., 2013). Neddylation, which is activatory to the function of CRLs, occurs within the globular C-terminal domain. DDB1 is instrumental in the function of CRL4. It is composed of three β-propeller folds utilized in protein binding, where β-propeller B (BPB) binds the cullin. The tightly coupled double propeller BPA-BPC is responsible for substrate presentation and is thus positioned in the complex facing the E2 binding site (Angers et al., 2006; Li et al., 2006). BPA-BPC mediates binding to substrate adaptor proteins, collectively termed DCAFs (DDB1 and CUL4-associated factors). Interestingly, this BPA-BPC double propeller is flexibly connected to BPB, which results in the adoption of various domain configurations by the CRL, thereby facilitating transfer of ubiquitin to a specific lysine residue on the substrate. Hence, it is understood that the DDB1 conformational versatility can modulate the E3 ligase activity towards its substrates and create a defined ubiquitylation zone (Angers et al., 2006; Fischer et al., 2011).

1.3.1. DCAF proteins: the substrate adaptors

More than 60 proteins are predicted to function as substrate adaptors for CRL4 (Figure I-9), but most are yet to be described as taking up an ubiquitylation function as part of a CRL4 complex. A common feature that characterizes DCAFs is the presence of multiple WD40 motifs, which may fold together to assemble a circularized beta-propeller structure termed WD40 domain. These domains are generally known to be involved in establishing protein-protein interactions, including binding to DDB1 (Angers et al., 2006; Jin et al., 2006). DCAF proteins are recruited to the CRL4 by binding the BPA-BPC double propeller in DDB1, in particular BPC. Different DCAFs have been observed to bind to different surfaces in DDB1, including distinct regions within BPC (Angers et al., 2006; Jin et al., 2006; Fischer et al., 2011).

1. General introduction

Figure I-8. Structure of CRL4 complexes. (A) Scheme of a CRL4A/B-DDB1-RBX1 E3 ligase, bound to a DCAF substrate adaptor protein that contains a HLH box (such as DDB2). DCAF binding is mediated by the BPA-BPC double propeller on DDB1. The BPB propeller binds DDB1 to the cullin. “?” denotes the remaining cullin N-terminal sequence for which no electronic density was obtained. (B,C) Crystal structures or CRL4ASV5-V and CRL4BDDB2, respectively. Note the structural resemblance of the CUL4-DDB1 core. Two different substrate adaptors are presented as examples. These adaptors are thought to bind CRL4A and CRL4B interchangeably. PDB IDs: 2HYE and 4A0L. BP, β-propeller.

DCAF binding to DDB1. Two main binding modes for DCAF binding to DDB1 have been described: via the DCAF’s WD40 propeller or via a characteristic alpha-helical domain. Early studies reported that an Aspartate-X-Arginine motif (DxR; X denotes any amino acid) present next to two consecutive blades of the WD40 domain, which are exposed on the surface of the WD40 β-propeller, are critical for DDB1 binding (Angers et al., 2006; Jin et al., 2006). Moreover, despite the strict conservation of this motif, other close residues were found to be conserved to some extent, and this region is thus referred to as the DWD

1. General introduction

(DDB1-binding and WD40 repeat) or CDW box (CUL4 and DDB1-associated WD40-repeat proteins) (Angers et al., 2006; Jin et al., 2006; He et al., 2006; Higa et al., 2006a). Importantly, these sequence features are not conserved in unrelated WD40 domain-containing proteins. Another binding mode of DCAFs to DDB1 was first identified as a short alpha-helical motif that can be inserted into the large pocket of the DDB1 double propeller fold. This motif was originally found in viral proteins that can hijack CRL4 (Angers et al., 2006; Li et al., 2006, 2010), but subsequent analysis identified that these proteins in fact mimic a second binding mode of canonical host DCAFs, which was then identified (Li et al., 2010). The reason why this motif had not been recognized in earlier studies was the high sequence divergence, with no strictly conserved amino acid, though there are overall conserved side chain properties (Li et al., 2010). Soon after, a second helix was detected for the DCAF CSA and shown to be present in several other putative CRL4 adaptors, together defining a helix-loop-helix motif (referred to as HLH box). The presence of the HLH box appears to indeed be a general feature in DCAF binding to DDB1 (Fischer et al., 2011). Interestingly, this study confirms an earlier report that two helices in the N- terminus of the DCAF DDB2 are very important for DDB1 binding (Jin et al., 2006), though in this case the authors did not recognize that such region represents a more general binding mode, present in a subset of DCAF proteins.

1.3.2. Functions of CRL4

Substrate adaptor DCAF proteins are responsible for the function of a particular CRL4 complex, because they can potentially recruit a very specific set of substrates. DCAFs are very divergent and, together with CRL4 core components, participate in a variety of cellular processes. The adaptor is also responsible to recruit the E3 ligase to the sites of action. We summarized the known functions of CRL4 complexes in table I-1. CRL4 is prominent in the regulation of the cell cycle (mostly S phase regulation) and in orchestrating the response to UV DNA damage. A common feature of several DCAF proteins is that they localize and perform functions as CRL4 complexes at the chromatin (Figure I-9 and table I-1). Hence, out of many functions that CRL4 can perform, some that stand out are epigenetic control of gene expression or chromatin remodeling, including in the context of the DNA damage response. DNA repair following UV damage is also much dependent on the action of CRL4, so that it transiently recruits repair factors to DNA.

1. General introduction

(figure legend in the next page)

1. General introduction

Figure I-9. Putative human CRL4 substrate adaptors, their binding mode to DDB1 and subcellular localization. Additionally, we indicate disease implications for mutations in their respective gene. DCAF-DDB1 binding mode is according to (Angers et al., 2006; Jin et al., 2006; He et al., 2006; Higa et al., 2006a; Li et al., 2010; Bennett et al., 2010; Fischer et al., 2011). DWD/CDW boxes are specialized regions in one or two consecutive WD40 blades that determine binding to DDB1. Subcellular localization and involvement in disease is according to Human Protein Atlas (proteinatlas.org) and UniProtKB (uniprot.org). The phrasing “organelles” refers to both membrane-enclosed or non- membrane organelles. Disease acronyms: IBD10, Inflammatory bowel disease 10; MRT2A, Mental retardation, autosomal recessive 2A; CSA, Cockayne syndrome A; UVSS2, UV-sensitive syndrome 2; WoSaS, Woodhouse-Sakati syndrome; XP-E, Xeroderma pigmentosum complementation group E; WVS, Weaver syndrome; LIS6, Lissencephaly 6, with microcephaly; LIS1, Lissencephaly 1; SBH, Subcortical band heterotopia; MDLS, Miller-Dieker lissencephaly syndrome; GAN2, Giant axonal neuropathy 2, autosomal dominant; CORD20, Cone-rod dystrophy 20; MCPH2, Microcephaly 2, primary, autosomal recessive. Other acronyms: HLH, Helix-Loop-Helix; Cyt, cytoplasm; Nuc, nuclear; Chr, known to function at the chromatin; Specification of nuclear localization: B, nuclear bodies or foci; E, nuclear envelope; N, nucleoli. Specification of cytoplasmic localization: C, centrosome; E, Endoplasmic reticulum; L, lysosome; M, mitochondria; P, plasma membrane; T, microtubules; V, vesicles. *CRBN has a unique binding mode to DDB1 (Fischer et al., 2014). “?”, not known.

Substrate E3 (co-factors) E Role of ubiquitylation References

Cell cycle

DNA Ligase I CRL4DCAF7 D Maintain low levels in non-proliferative cells (Peng et al., 2016) G0

p150-SAL2 CRL4RBBP7 D Probably promotes cellular transition to an (Sung et al., 2011)

active growing state

G0 G0 G1

USP11 CRL4/? D Allows CRL3KEAP1 ubiquitylation of PALB2 and (Orthwein et al., 2015)

HR inhibition G1 CDT2

CDT1, CDC6, CRL4 (PCNA) D (Clijsters and Wolthuis, 2014; PR-SET7, Prevents DNA re-replication by inhibiting the Senga et al., 2006; Higa et al., MMSET formation of new origins of replication 2006a; Oda et al., 2010; Evans et

al., 2016)

S: licensing S: inhibition p21/CIP1 CRL4CDT2 (PCNA) D Thought to maintain CDK2 activation (Nishitani et al., 2008; Abbas et

throughout S phase; also important to prevent al., 2008; Kim et al., 2008;

replication re-licensing Havens and Walter, 2009; Galanos et al., 2016) P21/CIP1 CRL4DCAF11 D Promotes S phase progression (Chen et al., 2017) P27 CRL4/? D Promotes S phase progression (Higa et al., 2006b; Miranda- Carboni et al., 2008)

Cyclin E CRL4B/? D Promotes S phase progression and prevents (Higa et al., 2006b; Zou et al., phase:CDK activity

- genomic instability 2009) S p12 CRL4CDT2 (PCNA) D Promote the switch between Polδ4 and Polδ3 (Darzynkiewicz et al., 2015; Lee for gap filling during lagging strand synthesis et al., 2014) SLBP CRL4WDR23 N Promote histone mRNA expression (Brodersen et al., 2016) Promote H3 deposition in newly replicated (Han et al., 2013a; Zhang et al., Rtt101Mms22 [Sc]; Histone H3 N DNA and aids in the establishment of sister 2017) CRL4?

chromatid cohesion phase:replication

- Spt16 Rtt101/? N Stabilizes of the FACT complex at the fork (Han et al., 2010)

S andnucleosomes CHK1 CRL4CDT2 (ATR; D Promotes G2/M progression (Van Leung-Pineda et al., 2009;

Claspin) Huh and Piwnica-Worms, 2013)

CDC25A CRL4BWDR42A D Prevent premature mitotic entry? (Wu et al., 2016) G2

RASSF1A CRL4/? D Possibly promotes mitotic progression (Jiang et al., 2011)

RBBP7 / CENP-A? CRL4 N? Promote CENP-A deposition at centromeres (Mouysset et al., 2015) CENP-A CRL4COPS8 N Promote interaction with HJURP and CENP-A (Niikura et al., 2015)

LateM earlyG1

Table I-1 (continues in the next page) 18

1. General introduction

Response to DNA damage and replication stress p21/CIP1, CRL4CDT2 (PCNA) D As tightly-bound PCNA binding proteins, p21 (Nishitani et al., 2008; Abbas et CDT1 and CDT1 degradation upon UV irradiation is al., 2008; Mansilla et al., 2013; thought to allow recruitment of repair factors Tsanov et al., 2014; Jin et al., by PCNA 2006; Higa et al., 2006a; Senga et al., 2006) p12 CRL4CDT2 (PCNA) D Regulates Polδ activity in the event of DNA (Zhang et al., 2013; Terai et al., damage to halt replication fork progression 2013)

and to promote gap-filling repair

PR-SET7 CRL4CDT2 (PCNA) D Double function: recruit PR-SET7 to promote (Oda et al., 2010; Tsanov et al., H4K20me1-mediated 53BP1 recruitment, 2014) followed by degradation of PR-SET7 to allow recruitment of repair factors by PCNA MCM10 CRL4VPRBP D Inhibits DNA replication in the event of UV (Kaur et al., 2012) DNA damage REV7 CRL4/? D Possibly prevents repair of UV lesions with the (Bhat et al., 2017)

error-prone Polζ Responseto damage UV Histone H2A CRL4B-RING1BDDB2 N Early line of action after damage detection by (Guerrero-Santoro et al., 2008;

NER [1][3] DDB2

- DDB1-DDB2. Promotes subsequent CRL4A Gracheva et al., 2016) assembly at lesion sites. Histones CRL4ADDB2 [1] N Thought to remodel the nucleosome structure (Kapetanaki et al., 2006; Wang H2A, H3, H4 to allow access of repair factors et al., 2006) XPC CRL4ADDB2 [1] N Recruits XPC to the site of DNA damage (Sugasawa et al., 2005)

DDB2 CRL4A? [2] D Damage handover from DDB2 to XPC (Sugasawa et al., 2005) UV damage: UV GG

CSB CRL4CSA D Allows recruitment of a subset of repair factors (Fousteri et al., 2006; Groisman -

and post-repair recovery of transcription et al., 2006)

TC NER WDR70 H2B CRL4 N Facilitates DNA end resection for repair by (Zeng et al., 2016)

homologous recombination DSB CHK1 CRL4ACDT2 (ATR; D Might facilitate checkpoint recovery after (Van Leung-Pineda et al., 2009; Claspin) replication stress Huh and Piwnica-Worms, 2013) Topoisomer CRL4/? D Aids in the resolution of topoisomerase I – (Kerzendorfer et al., 2010) ase I DNA covalent complexes ? Rtt101-Mms1Mms22 ? Promote homologous recombination- (Zaidi et al., 2008; Buser et al., [Sc] dependent repair and restart of stalled 2016)

replication forks

Responseto replicationstress Regulation of gene expression (non-cell cycle functions) ? CRL4WDR5 ? Promote H3K4me1 and H3K4me3 (Higa et al., 2006a) WDR5 CRL4B/? [2] D Regulates neuronal gene expression by (Nakagawa and Xiong, 2011) reducing H3K4me3; promotes neurite outgrowth of PC12 neuroendocrine cells Histone H2A CRL4BRBBP7? N Epigenetic silencing of negative regulators of (Hu et al., 2012) (PRC2) cell proliferation

Histone H2A CRL4WDTC1 N Suppress transcription of genes involved in (Groh et al., 2016) adipogenesis ? CRL4RBBP7? (SIN3A- ? Repression of p21 gene transcription (Ji et al., 2014) HDAC) Histone H3 CRL4DCAF8 N Induces H3K9 methylation and gene repression (Li et al., 2017)

in the liver Chromatinremodeling

Table I-1 (continues in the next page) 19

1. General introduction

c-Jun CRL4COP1 (DET1) D Reduces c-Jun-activated transcription (Wertz et al., 2004) c-myc CRL4TRPC4AP D Inhibits cell growth by transcriptional (Choi et al., 2010) N-myc repression of target genes HOXA9 CRL4/? D Promotes hematopoiesis (granulopoiesis) (Zhang et al., 2003) HOXB4 CRL4/? D Inhibition of hematopoietic stem cell (Lee et al., 2013) proliferation LATS1, CRL4VPRBP D Promotes YAP- and TEAD-dependent (Li et al., 2014) LATS2 transcription of genes involved in cell survival

and proliferation FOXM1 CRL4VPRBP D Non-CRL4 VPRBP promotes mitotic (Wang et al., 2017) progression, but CRL4VPRBP leads to degradation of FOXM1 ERα CUL4BAhR [3] D Negatively regulates sex hormone-dependent (Ohtake et al., 2007) AR (Dioxins) signaling ERα CRL4VPRBP D Controls human breast cell fate (Britschgi et al., 2017) (LATS1,2) RORα CRL4VPRBP (EZH2) D Requires a monomethyl degron generated by (Lee et al., 2012) EZH2; significance unknown, but its deregulation promotes tumor development

MEIS2 CRL4CRBN D Possibly promotes normal development (Fischer et al., 2014) Regulationof transcription factors Growth signaling pathways and metabolism TSC2 CRL4FBW5 [Dm] D Promotes growth by mediating the PI3K– (Hu et al., 2008)

mTOR pathway (demonstrated for Drosophila

larvae) ? CRL4B D Regulates mTORC1 by interacting with Raptor (Ghosh et al., 2008)

(Raptor,mLST8) and mLST8

Regulationof mTORC1 - GS CRL4CRBN D (Matyskiela et al., 2016) Degradation of glutamine synthetase in (Glutamine; response to glutamine to regulate metabolism

p300/CBP)

Metabo lism

Table I-1. Published CRL4 substrates and its respective E3 ligase, including functions suggested to be specific to CRL4A and CRL4B. Dm, Drosophila melanogaster; Sc, Saccharomyces cerevisiae; E: Effect of ubiquitylation (D, degradation; N-non-proteolytic). “?”, unknown. [1]Both cullins bind the same substrate adaptor, but execute partially different functions [2]Autoubiquitylation [3]Atypical E3 ligase

Remarkably, CRL4 was found to be hijacked by viral proteins that function as substrate adaptors, to induce the degradation of proteins that are otherwise deleterious for the viral infection. Viruses that hijack CRL4B are for example HIV-1, Paramyxovirus simian virus 5 or Hepatitis B virus (Ahn et al., 2010; Ulane and Horvath, 2002; Angers et al., 2006; Li et al., 2010). Another form of hijacking of CRL4 to degrade non-canonical substrates is by usage of immune-modulatory drugs (iMiDs). iMiDs such as thalidomide and its derivatives lenalidomide and pomalidomide are being used as effective drugs in the treatment of haematologic malignancies such as multiple myeloma and myelodysplastic syndrome (List et al., 2005). In the past, these drugs were used as sedatives until the discovery that they cause teratogenic side effects

1. General introduction

in the developing embryo. iMiDs have a dual role in the modulation of CRL4CRBN activity, as they 1) prevent the ubiquitylation of some endogenous substrates and 2) induce the binding and ubiquitylation of non- canonical CRL4CRBN substrates (termed neosubstrates) (Kronke et al., 2014; Lu et al., 2014; Krönke et al., 2015; An et al., 2017). Neosubstrates include the B cell transcription factors IKZF1 and IKZF3, which are essential in multiple myeloma cells (Kronke et al., 2014). Hence, iMiDs are in use to treat multiple myeloma. Likewise, another neosubstrate is CKIα, which is exploited to treat myelodysplastic syndrome (Krönke et al., 2015). Importantly, degradation of a number of CRL4CRBN substrates, including the neosubstrates, requires the action of p97 (Nguyen et al., 2017).

CRL4 in chromatin remodeling. CRL4 is a dedicated chromatin-related E3 ligase that ubiquitylates DNA- bound proteins in multiple cellular contexts (table I-1). For example, CRL4CDT2 works during DNA replication to ensure that replication licensing is not re-established, and that CDKs remain active (see section 1.2.2). Other mechanisms include the ubiquitylation of CSB, which is important for the recruitment of a specific subset of repair factors in transcription-coupled nucleotide excision repair (TC-NER). Early studies in fission yeast revealed that CRL4 is important in heterochromatin formation, by directly or indirectly regulating the Clr4 methyltransferase and hence promoting H3K9me (Jia et al., 2005). Indeed, human CRL4 (specially CRL4B) appear to take up several functions in gene repression. CRL4B has been associated with gene repression via targeting for degradation the methyltransferase component WDR5. WDR5 is a core subunit of multiple methyltransferase complexes, namely of the MLL1 family, NIF- 1 complex, ATAC, RING1b, CHD8 and H1.2 transcriptional repressive complex. It was suggested that as a consequence of WDR5 degradation, CUL4B negatively regulates H3K4me3, a histone mark associated with active gene transcription (Nakagawa and Xiong, 2011). Several studies associate CUL4B with the monoubiquitylation of histone H2A in lysine 119, a histone mark associated with gene silencing (Baarends et al., 2005). It has been reported that CRL4B directly associates with the DNMT3A/SUV39H1/HP1 complex, thereby facilitating H3K9me3 and DNA methylation, likely in a manner that depends on H2A ubiquitylation by CRL4B (Yang et al., 2013). CRL4B also associates with Polycomb-repressive complex 2 (PRC2) to mediate transcription repression. Target genes include p16 and PTEN, which could indicate that CRL4B promotes cell proliferation by silencing negative regulators of cell growth (Hu et al., 2012). CUL4B and the PRC2 subunit EZH2 co-occupy target gene loci, promoting H2AK119 monoubiquitylation and H3K27me3, both histone marks associated with gene repression, in a manner independent of PCR1 (a PRC2 interactor also involved in H2AK119ub1). Another regulatory feature of CRL4B is its association with the SIN3A-HDAC complex, observed in mouse embryonic fibroblasts. As a consequence, acetylation of histones H3 and H4 in specific loci is

1. General introduction reduced, which was observed to be the case for the p21 gene promoter (Ji et al., 2014). As opposed to the well-established role of CRL4CDT2 in p21 degradation (Table I-1), the authors report that CRL4B does not regulate p21 levels by proteasomal targeting, but instead regulates p21 expression independent of p53 (Ji et al., 2014). However, the implications of CUL4B in p21 protein level regulation are controversial, as a very recent study clearly demonstrates that CRL4B does induce p21 degradation, but using WDR23 (and not CDT2) as the substrate adaptor (Chen et al., 2017). CRL4 has essential functions in DNA repair, mostly the nucleotide excision repair (NER) pathway, where it performs additional chromatin remodeling roles. In global genome (GG) NER, CRL4DDB2 is known to recruit repair factors to the site of lesion (Sugasawa et al., 2005). The DDB1-DDB2 dimer acts as a damage surveillance protein complex that specifically binds products of UV damage: 6-4 photoproducts (6–4PPs) and cyclobutane pyrimidine dimers (CPDs) (Wittschieben et al., 2005; Scrima et al., 2008; Fischer et al., 2011). Upon damage detection, CRL4DDB2 ubiquitylates the repair factor XPC, which alters its chromatin-binding properties and is recruited to the damage site (Sugasawa et al., 2005). CRL4DDB2 additionally ubiquitylates surrounding histones while DNA-bound due to the flexibility of the BPA-BPC double propeller tether to BPC on DDB1 (Wang et al., 2006; Kapetanaki et al., 2006; Guerrero-Santoro et al., 2008; Gracheva et al., 2016; Fischer et al., 2011). The ensuing H2A, H3 and H4 histone ubiquitylation is thought to render damaged DNA more accessible to repair factors. DDB2 is also itself ubiquitylated and degraded, thought to allow the transfer of damaged DNA to XPC and the ensuing repair factors (Sugasawa et al., 2005). Surprisingly, a recent report suggested that CRL4ADDB2 and CRL4BDDB2 complexes do not act redundantly, but rather at different stages. It was found that the first line of response to UV damage by CRL4 involves the assembly of a non-canonical complex composed of CUL4B-DDB1-DDB2 and another E3 ligase instead of RBX1, RING1B, to ubiquitylate histone H2A (Gracheva et al., 2016). The subsequent recruitment of the H2A-ubiquitin binding protein ZRF1 promotes the assembly of a canonical CRL4ADDB2 E3 ligase that performs the above-mentioned functions in histone and XPC ubiquitylation (Gracheva et al., 2016).

1.4. Functional distinctions between CUL4A and CUL4B

CRL4A and CRL4B are widely regarded as redundant. These cullins share 83% sequence identity, and bind essentially the same interacting partners: DDB1 and RBX1. Because DDB1 is the linker to the DCAF protein, it is not easily conceivable that CUL4A and CUL4B assemble distinct E3 ligases. It has been

1. General introduction

suggested that CUL4A is mostly cytoplasmic and CUL4B nuclear, which could justify such differences (Hannah and Zhou, 2015; Nakagawa and Xiong, 2011). However, enough evidence argues that endogenous CUL4A is abundant in the nucleus and the localization of both cullins is almost indistinguishable (Olma et al. 2009; see the Human Protein Atlas project entries for CUL4A and CUL4B), suggestive of cell-type specific localization. Nevertheless, there have been reported functional differences between the cullins (Table I-1), even though the rationale for such distinctions is usually absent. One notable exception is the action of CUL4B together with the nuclear dioxin receptor (AhR, arylhydrocarbon receptor). CUL4B-AhR assemble an atypical E3 ligase independent on DDB1 that regulates sex hormone signaling (Ohtake et al., 2007). In the presence of its agonists such as dioxins (environmental toxins), AhR recruits CRL4B by interacting directly with the unique N-terminus of CUL4B to target for degradation the sex hormone receptors estrogen receptor α (ERα) and androgen receptor (AR) in mice uterus or prostate cells, respectively. While the significance of CRL4BAhR-mediated sex hormone receptor regulation remains unexplored, it reveals a mechanism how dioxins can affect human health.

1.4.1. CUL4A and CUL4B in human disease

Diseases where CRL4 complexes may be implicated can be classified depending on mutations in DCAF proteins, or on the cullins themselves. Known DCAF genes whose mutations are associated with non-cancer diseases are indicated in Figure I-9, although in the majority of cases the direct involvement of the CRL4 complex remains to be explored. Here we display established genetic links between CUL4A and CUL4B function and disease development.

CUL4B in XLID. While no human disorder has been described for mutations in the CUL4A gene, mutations in CUL4B were associated with syndromic X-linked intellectual disability (XLID, Cabezas type; Figure I-10). CUL4B mutations account for about 3% of all XLID cases (Vulto-van Silfhout et al., 2015). Loss-of-function CUL4B mutations impairs cognition and physical development that is not restricted to the brain. Besides intellectual disability, patients show neurological symptoms such as gait ataxia and behavioral problems (e.g. aggressive outbursts), as well as motor and speech delays during childhood. In the brain, macrocephaly and severe ventricular enlargement is often observed. Besides, patients are characterized by short stature and craniofacial dysmorphism and other physical features, such as hypogonadism and underdeveloped limbs (see Figure I-11) (Cabezas et al., 2000; Tarpey et al., 2007; Zou et al., 2007; Badura-

1. General introduction

Figure I-10. Reported mutations in CUL4B that have been associated with XLID (Cabezas et al., 2000; Tarpey et al., 2007; Zou et al., 2007; Badura-Stronka et al., 2010; Isidor et al., 2010; Londin et al., 2014; Vulto-van Silfhout et al., 2015; Grozeva et al., 2015; Okamoto et al., 2017). Top, binding regions of DDB1 and RBX1, and site of neddylation.

Stronka et al., 2010; Isidor et al., 2010; Londin et al., 2014; Vulto-van Silfhout et al., 2015; Grozeva et al., 2015; Okamoto et al., 2017; Kerzendorfer et al., 2011). All described mutations that link CUL4B and XLID are thought to be loss-of-function mutations. In particular, truncating mutations (nonsense, frame-shifting indels or splice variants) disrupt the functional domains of CUL4B, its interaction with critical subunits (in particular RBX1) or the neddylation domain, thereby inactivating the E3 ligase. At least the R388 truncation also leads to nonsense-mediated mRNA decay (Zou et al., 2007). Missense mutations are thought to destabilize the protein, which has been demonstrated for T213A, R572C, and V745A (Vulto-van Silfhout et al., 2015; Nakagawa and Xiong, 2011). Regarding in-frame indels, L785del destabilizes the protein, as well as A621dup to some extent (Vulto-van Silfhout et al., 2015). P50L appears to have no impact on CUL4B stability. Despite much speculation (Kerzendorfer 2011), the reasoning behind CUL4B mutations and disease development is not yet established, which suggests that a yet unknown function of CUL4B is not discovered. Female heterozygous carriers have no signs of this disease, which correlates with the observed skewed X inactivation pattern towards the mutated CUL4B-containing X chromosome (Zou et al., 2007).

Suggested specific CRL4B functions involved in XLID. Despite that to this date a link between a specific CUL4B function and syndromic XLID is lacking, all described specific CUL4B functions (i.e. that cannot be compensated by CUL4A) could potentially be involved in development of this disease. It is also not known which stage of development of the embryo/fetus is affected. However, some attempts to connect CUL4B functions and disease have been made (Kerzendorfer et al., 2011).

1. General introduction

Figure I-11. Frequency of observed neurological and physical traits in syndromic XLID patients (Cabezas type), which harbor CUL4B mutations. Patient information refers to adulthood unless specified. Central nervous system (CNS) abnormalities include malformations of cortical development, colpocephaly, ventriculomegaly, diminished white matter volume. Other physical phenotypes that are not represented include pes cavus, Gynecomastia, Kyphosis (observed in 1/3 to half of the patients). HSR, hyperplastic supraorbital ridges.

One of the connections made between CRL4B-mediated gene expression regulation and XLID was via its role in targeting for degradation a H3K4 methyltransferase complex subunit, WDR5 (Nakagawa and Xiong, 2011). The authors demonstrated that CUL4B depletion increases H3K4me3 on neuronal gene promoters, thereby inducing their expression. Additionally, experiments where the differentiation of PC12

1. General introduction cells is induced with nerve growth factor indicate that CUL4B is important for this process, by counteracting the function of WDR5 (Nakagawa and Xiong, 2011). Another report suggested that CUL4B may impact mouse neural progenitor cell (NPC) differentiation by negatively regulating the expression of glial fibrillary acidic protein (GFAP) together with PRC2 (Zhao et al., 2015). It was also reported that CUL4B depletion impacts the proliferation of rat neural progenitor cells and human NT-2 cells, which exhibit G2/M delays (Liu et al., 2012a). Finally, studies with patient-derived lymphoblastoid cell lines revealed increased sensitivity to camptothecin (CPT) due to apparent slight misregulation of topoisomerase degradation (Kerzendorfer et al., 2010).

CUL4A and CUL4B involvement in cancer. Amplifications or gain-of-function of CUL4A is often implicated in cancer development (Lee and Zhou, 2010). CUL4A amplification has been reported in several types of carcinomas, childhood medulloblastoma and in the majority of primary malignant pleural mesothelioma. It also accounts for ~5% of familial and sporadic breast cancers (Lee and Zhou, 2010). Upregulation of CUL4B has also been noted in various human cancers (Hu et al., 2012). Interestingly, one of the most likely protein-protein interfaces affected by cancer mutations is the CUL4B-CAND1 interface (note that CUL4A is not excluded, but a structure of CUL4A bound to CAND1 is not available). Because CAND1 is a known inhibitor of cullins (Lydeard et al., 2013), this data once again demonstrates that deregulated activation of CUL4 is correlated with cancer. Moreover, the observation that accumulation of cyclin D1 during S phase drives both B cell lymphomas and mammary carcinomas in mice was correlated to CDK4-cyclin D1- dependent transcriptional regulation of CUL4A and CUL4B expression. Indeed, it has been proposed that CUL4A and CUL4B expression is negatively regulated by CDK4-cyclin D1 before S phase, so that CDT1 can be stabilized (Aggarwal et al., 2010). Constitutive CDK4-cyclin D1 activation observed in cancer thus result in S-phase CDT1 stabilization that results in genomic instability due to DNA re-replication (Aggarwal et al., 2007, 2010). CUL4 was also indirectly associated with breast and prostate cancer. It was proposed that the methyltransferase PRC2 component EZH2, which commonly works by mediating gene repression, can methylate non-histone proteins. Via this mechanism, EZH2 generates a monomethyldegron recognized by the chromo domain of CRL4VPRBP (Lee et al., 2012). One of the proteins that can be targeted for degradation by this EZH2-CRL4VPRBP mechanism is RORα, suggested to be a tumor suppressor. Hence, it has been proposed that the upregulation of EZH2 leads CRL4-dependent downregulation of RORα, thereby supporting tumorigenesis (Lee et al., 2012).

1. General introduction

1.4.2. Mouse knockout models.

Homozygous deletion of Ddb1 is embryonically lethal, with extensively degenerated embryos observed at E12.5 (Cang et al., 2006), whereas heterozygous Ddb1–/+ mice showed no gross phenotype. Conditional CNS-specific deletion of DDB1 lead to neonathal death due to neuronal and lens degradation and brain hemorrhages. Proliferating cells were largely absent throughout development of the brain, where extensive apoptosis was observed. These tissues showed markedly increased DNA damage, and the upregulation of CRL4 substrates was verified (Cang et al., 2006), which is evidence that general CUL4 function is essential for cell proliferation and viability. Surprisingly, the knockout of Cul4a leads to viable mice that survive to adulthood without overt physical or behavioral phenotypes (Kopanja et al., 2011; Yin et al., 2011). Male mice, however, are infertile and exhibit deficiencies in spermatogenesis. Besides extensive apoptosis observed in the testis, spermatocytes were shown to have impaired development past meiotic prophase I (pachytene stage). These exhibit double stranded DNA breaks, suggesting failed homologous recombination at prophase I (Kopanja et al., 2011). However, another study reported defects rather in later meiotic stages (Yin et al., 2011). Importantly, these cells are not expected to express CUL4B due to meiotic sex chromosome inactivation at the pachytene-diplotene stages, leading to the conclusion that CUL4A is essential in spermatogenesis because CUL4B is absent (Kopanja et al., 2011; Yin et al., 2011). This implies that CUL4B can compensate for all CUL4A functions as long as it is present. Interestingly, three independent groups reported that CUL4B deletion in mice is embryonically lethal, which dramatically contrasts with the rather innocuous CUL4A knockout. Cre-lox-mediated deletion of CUL4B in the zygote (actin CAG promoter) or at E2.5-4.5 (EIIa promoter) was lethal, with no remaining male Cul4b–/Y embryos by embryonic day E9.5 (Jiang et al., 2012; Liu et al., 2012b; Chen et al., 2012). Analysis of earlier stages revealed that at E7.5 (WT gastrula stage), hemizygous knockout embryos are dramatically smaller, with lack of expansion of the embryonic part of the egg cylinder and without discernible germ layers. These embryos showed increased apoptosis and markedly decreased proliferation in both embryonic and extraembryonic tissues (measured by KI67 and BrdU stainings) (Jiang et al., 2012; Liu et al., 2012b). Interestingly, a considerable population of cells was positive for phosphorylated histone H3 (S10) staining, indicative of a mitotic arrest (Liu et al., 2012b). Surprisingly, epiblast-specific deletion of Cul4b by E6.5 prevented embryonic lethality in Cul4b–/Y mice, which survived to adulthood without overt phenotypes (Liu et al., 2012b; Chen et al., 2012). Notably, syndromic XLID-like phenotypes were not observed, except impaired learning and memory and increased epileptic susceptibility (Chen et al., 2012). Because the Sox2 promoter employed is not active in extraembryonic

1. General introduction tissues, the authors concluded that CUL4B is dispensable in the embryo proper. Accordingly, CUL4B downregulation in an extraembryonic cell line showed G2/M cell cycle arrest (Liu et al., 2012b). Analysis of female heterozygous knockout embryos also revealed embryogenesis phenotypes, though not as severe as the hemizygous knockout in males. Liu and colleagues describe that these embryos equally perish, but later by E12.5-E13.5, and observed that vitelline vessels were underdeveloped (Liu et al., 2012b). These authors demonstrated that this is a direct cause of paternal- specific X imprinted inactivation in mice extraembryonic tissues, as the non-imprinted maternal X chromosome is deleted for Cul4b. It also further supports that Cul4b deletion is very deleterious in extraembryonic tissue development. Jiang and colleagues similarly report underdeveloped extraembryonic compartment and impaired vascularization (Jiang et al., 2012). However, in contrast to the observations by Liu et al., the heterozygous deletion was not lethal despite an observed growth delay. In these mice, Cul4b null cells in the embryo and the adult are selected against (skewed X-inactivation), indicative of the importance of CUL4B expression for development (Jiang et al., 2012). The fact that Cul4b knockout was lethal prompt the search as to why CUL4A cannot compensate for its loss. By performing immunostainings, it was determined that while CUL4B is ubiquitously expressed in the embryo and the extraembryonic tissues, CUL4A protein levels are low, but it is not absent (Liu et al., 2012b; Jiang et al., 2012). It was thus speculated that CUL4A cannot compensate for the loss of CUL4B in extraembryonic tissues simply due to CUL4A’s low expression (Liu et al., 2012b). Further analysis of mouse models also determined that CUL4B is pivotal for mouse spermatogenesis, but in a manner distinct from CUL4A. Conditional germ-cell specific or global knockout of Cul4b in mice led to complete male infertility (Yin et al., 2016; Lin et al., 2016). The germ-cell conditional knockout yielded comparable numbers of spermatozoa and normal testicular morphology, but more than half of mutant spermatozoa had a markedly reduced mitochondrial content (Yin et al., 2016). Curiously, the other spermatozoa appeared to have normal mitochondrial staining. Global knockout revealed a more striking phenotype, with highly impaired sperm production and adults lost their germ cell population in an age-dependent manner, also displaying shrunken testis (Yin et al., 2016; Lin et al., 2016). This is reminiscent of the often-observed human hypogonadism phenotypes in patients with CUL4B deletions (Figure I-11), and evidence for a developmental phenotype that is not dependent on the germ cells themselves. In fact, it appeared that global Cul4b knockout mice failed to maintain the spermatogonial stem cell niche, for reasons that were not determined (Yin et al., 2016). Importantly, it was commonly observed that spermatids underwent apoptosis during spermiogenesis and displayed aberrant acrosomes and nuclear morphology, resulting in massive spermatid loss (Lin et al.,

1. General introduction

2016). The few detected sperm cells also showed these defects, and in some the acrosome, nucleus or mitochondria were hardly discernible. In line with these observations, it was suggested that CUL4B takes part in acrosome formation, nuclear morphogenesis and post-mitotic histone turnover, which occur during spermiogenesis (Lin et al., 2016). Importantly, although CUL4A is present during meiosis itself, it appears absent during spermiogenesis (Kopanja et al., 2011; Yin et al., 2011). Taken together, these studies in mice argue towards a full redundancy between CUL4A and CUL4B, with a consequent arrest in developing cells at stages when neither is present, or only one is expressed but at low levels.

1.5. Open questions and objectives

The maintenance of living organisms requires cell proliferation. This process inevitably involves DNA segregation so that genetic information is preserved. In eukaryotes, very complex mechanisms ensure that this genetic information in faithfully preserved, so that all “daughter” cells are genetically identical to their “mother”. Because cell proliferation depends on the activity of thousands of proteins (in fact, evolution dictates that the purpose of all proteins in our body is to promote propagation of the species), their function must be orchestrated, directly or indirectly dependent on post-translational modifications. Here, ubiquitin provides the means to modulate the function of a protein, or “kill” its activity if undesired: ubiquitin-mediated degradation ensures for example that there’s always a phase of DNA duplication before separation and prevents that a protein is at the wrong place at the wrong time. The existence of a vast number of proteins calls for detailed individual regulation. Keeping the ubiquitin view, the E3 ligase CRL4 is one of the E3 ligases known to orchestrate the cell cycle and drive cell proliferation by ubiquitylating specific protein targets. However, we believe that its functions have remained mostly obscure, and out of more than 60 predicted distinct complexes that it can assemble, well-described functions are only available for a handful. In line with our interest to investigate how the function of cell cycle proteins is timely and accurately modulated to ensure proper cell division, we inquire as to how exactly CRL4 orchestrates cell cycle processes. We aim to identify which substrates does CRL4 ubiquitylate, and the precise identity of the CRL4-type complex that executes this function. My second main focus was to establish whether the scaffold subunits of CRL4, the paralogs CUL4A and CUL4B, work fully redundantly or can also perform separate functions, specifically in the regulation of cell division. This hypothesis follows several observations that these cullins might act distinctively at

1. General introduction times, but few studies provide an explanation on how can these cullins foster a specific, non-redundant, function. One main motivation is the specific association of CUL4B mutations in syndromic X-linked intellectual disability, which signifies that indeed we know very little about the function of this cullin. How are mutations in CUL4B involved in XLID? Which specific process does CRL4B execute, and what determines that CUL4A cannot perform such function? We thus inquire whether CUL4A and CUL4B can assemble different complexes, and if so, how does this come about. Put together, another objective was to identify and characterize functional differences between cullin-4 paralogs. This thesis has been divided in three independent result sections where I present the work I performed during my post-graduate studies at ETH. Chapter 2 contains a collective effort to unveil how CRL4 (via the RBBP7 substrate adaptor) is essential in the determination of the centromere in early cell cycle stages, and therefore how it ensures chromosome separation in mitosis. In Chapter 3, I present my efforts in establishing that the two CRL4 cullin scaffolds can indeed mediate distinct functions. CUL4B, but not CUL4A, is repurposed via a mechanism that involves another post-translational modification, phosphorylation. The specificity of this modification resides within mitosis, and indeed I show that CUL4B regulates this process via a yet-to-identify role that must be restricted to the time of cell division in order to support cell survival. Finally, in chapter 4 I present evidence for the action of another CRL4 complex, CRL4WDTC1 similarly in late cell cycle stages. I demonstrate the recruitment of a cell cycle-specific regulator, and provide detailed insights on our efforts to establish new tools for the study of protein functions and interactions. In particular, I demonstrate the usage of a CRISPR-based system for inducible protein degradation. Hence, my efforts during my post-graduate studies have resided within the investigation of ubiquitin-mediated regulation of the process of cell division itself. I believe that we are one step forward to understand how this stunning protein network machinery works to flawlessly and perpetually propagate genetic information across eukaryotes.

1. General introduction

1.6. References

Abbas, T., U. Sivaprasad, K. Terai, V. Amador, M. Pagano, and A. Dutta. 2008. PCNA-dependent regulation of p21 ubiquitylation and degradation via the CRL4Cdt2 ubiquitin ligase complex. Genes Dev. 22:2496–2506. doi:10.1101/gad.1676108. Aggarwal, P., M.D. Lessie, D.I. Lin, L. Pontano, A.B. Gladden, B. Nuskey, A. Goradia, M.A. Wasik, A.J.P. Klein- Szanto, A.K. Rustgi, C.H. Bassing, and J.A. Diehl. 2007. Nuclear accumulation of cyclin D1 during S phase inhibits Cul4-dependent Cdt1 proteolysis and triggers p53-dependent DNA rereplication. Genes Dev. 21:2908–2922. doi:10.1101/gad.1586007. Aggarwal, P., L.P. Vaites, J.K. Kim, H. Mellert, B. Gurung, H. Nakagawa, M. Herlyn, X. Hua, A.K. Rustgi, S.B. McMahon, and J.A. Diehl. 2010. Nuclear cyclin D1/CDK4 kinase regulates CUL4 expression and triggers neoplastic growth via activation of the PRMT5 methyltransferase. Cancer Cell. 18:329–340. doi:10.1016/j.ccr.2010.08.012. Ahn, J., T. Vu, Z. Novince, J. Guerrero-Santoro, V. Rapic-Otrin, and A.M. Gronenborn. 2010. HIV-1 Vpr loads uracil DNA glycosylase-2 onto DCAF1, a substrate recognition subunit of a cullin 4A-RING E3 ubiquitin ligase for proteasome-dependent degradation. J. Biol. Chem. 285:37333–37341. doi:10.1074/jbc.M110.133181. Alabert, C., and A. Groth. 2012. Chromatin replication and epigenome maintenance. Nat. Rev. Mol. Cell Biol. 13:153–167. doi:10.1038/nrm3288. An, J., C.M. Ponthier, R. Sack, J. Seebacher, M.B. Stadler, K.A. Donovan, and E.S. Fischer. 2017. pSILAC mass spectrometry reveals ZFP91 as IMiD-dependent substrate of the CRL4CRBN ubiquitin ligase. Nat. Commun. 8:15398. doi:10.1038/ncomms15398. Angers, S., T. Li, X. Yi, M.J. MacCoss, R.T. Moon, and N. Zheng. 2006. Molecular architecture and assembly of the DDB1-CUL4A ubiquitin ligase machinery. Nature. 443:590–3. doi:10.1038/nature05175. Baarends, W.M., E. Wassenaar, R. van der Laan, J. Hoogerbrugge, E. Sleddens-Linkels, J.H.J. Hoeijmakers, P. de Boer, and J.A. Grootegoed. 2005. Silencing of unpaired chromatin and histone H2A ubiquitination in mammalian meiosis. Mol. Cell. Biol. 25:1041–53. doi:10.1128/MCB.25.3.1041- 1053.2005. Badura-Stronka, M., A. Jamsheer, A. Materna-Kiryluk, A. Sowińska, K. Kiryluk, B. Budny, and A. Latos- Bieleńska. 2010. A novel nonsense mutation in CUL4B gene in three brothers with X-linked mental retardation syndrome. Clin. Genet. 77:141–144. doi:10.1111/j.1399-0004.2009.01331.x. Barr, A.R., F.S. Heldt, T. Zhang, C. Bakal, and B. Novák. 2016. A Dynamical Framework for the All-or-None G1/S Transition. Cell Syst. 2:27–37. doi:10.1016/j.cels.2016.01.001. Bashir, T., N.V. Dorrello, V. Amador, D. Guardavaccaro, and M. Pagano. 2004. Control of the SCFSkp2–Cks1 ubiquitin ligase by the APC/CCdh1 ubiquitin ligase. Nature. 428:190–193. doi:10.1038/nature02330. Bassermann, F., R. Eichner, and M. Pagano. 2014. The ubiquitin proteasome system - implications for cell cycle control and the targeted treatment of cancer. Biochim. Biophys. Acta. 1843:150–162. doi:10.1016/j.bbamcr.2013.02.028. Bassermann, F., D. Frescas, D. Guardavaccaro, L. Busino, A. Peschiaroli, and M. Pagano. 2008. The Cdc14B- Cdh1-Plk1 axis controls the G2 DNA-damage-response checkpoint. Cell. 134:256–67. doi:10.1016/j.cell.2008.05.043. Bell, S.P. 2014. Terminating the replisome. Science. 346. Bennett, E.J., J. Rush, S.P. Gygi, and J.W. Harper. 2010. Dynamics of cullin-RING ubiquitin ligase network revealed by systematic quantitative proteomics. Cell. 143:951–965. doi:10.1016/j.cell.2010.11.017.

1. General introduction

Bertoli, C., J.M. Skotheim, and R.A.M. de Bruin. 2013. Control of cell cycle transcription during G1 and S phases. Nat. Rev. Mol. Cell Biol. 14:518–28. doi:10.1038/nrm3629. Bhat, A., Z. Qin, G. Wang, W. Chen, and W. Xiao. 2017. Rev7, the regulatory subunit of Polζ, undergoes UV-induced and Cul4-dependent degradation. FEBS J. 284:1790–1803. doi:10.1111/febs.14088. de Boer, H.R., S. Guerrero Llobet, and M.A.T.M. van Vugt. 2016. Controlling the response to DNA damage by the APC/C-Cdh1. Cell. Mol. Life Sci. 73:949–60. doi:10.1007/s00018-015-2096-7. Boos, D., C. Kuffer, R. Lenobel, R. Korner, and O. Stemmann. 2008. Phosphorylation-dependent Binding of Cyclin B1 to a Cdc6-like Domain of Human Separase. J. Biol. Chem. 283:816–823. doi:10.1074/jbc.M706748200. Britschgi, A., S. Duss, S. Kim, J.P. Couto, H. Brinkhaus, S. Koren, D. De Silva, K.D. Mertz, D. Kaup, Z. Varga, H. Voshol, A. Vissieres, C. Leroy, T. Roloff, M.B. Stadler, C.H. Scheel, L.J. Miraglia, A.P. Orth, G.M.C. Bonamy, V.A. Reddy, and M. Bentires-Alj. 2017. The Hippo kinases LATS1 and 2 control human breast cell fate via crosstalk with ERα. Nature. 541:541–545. doi:10.1038/nature20829. Broderick, R., and W. Niedzwiedz. 2015. Sister chromatid decatenation: bridging the gaps in our knowledge. Cell Cycle. 14:3040–3044. doi:10.1080/15384101.2015.1078039. Brodersen, M.M.L., F. Lampert, C.A. Barnes, M. Soste, W. Piwko, and M. Peter. 2016. CRL4WDR23- Mediated SLBP Ubiquitylation Ensures Histone Supply during DNA Replication. Mol. Cell. 62:627– 635. doi:10.1016/j.molcel.2016.04.017. Buser, R., V. Kellner, A. Melnik, C. Wilson-Zbinden, R. Schellhaas, L. Kastner, W. Piwko, M. Dees, P. Picotti, M. Maric, K. Labib, B. Luke, and M. Peter. 2016. The Replisome-Coupled E3 Ubiquitin Ligase Rtt101Mms22 Counteracts Mrc1 Function to Tolerate Genotoxic Stress. PLoS Genet. 12:e1005843. doi:10.1371/journal.pgen.1005843. Cabezas, D.A., R. Slaugh, F. Abidi, J.F. Arena, R.E. Stevenson, C.E. Schwartz, and H.A. Lubs. 2000. A new X linked mental retardation (XLMR) syndrome with short stature, small testes, muscle wasting, and tremor localises to Xq24-q25. J. Med. Genet. 37:663–8. Cang, Y., J. Zhang, S.A. Nicholas, J. Bastien, B. Li, P. Zhou, and S.P. Goff. 2006. Deletion of DDB1 in Mouse Brain and Lens Leads to p53-Dependent Elimination of Proliferating Cells. Cell. 127:929–940. doi:10.1016/j.cell.2006.09.045. Cappell, S.D., M. Chung, A. Jaimovich, S.L. Spencer, and T. Meyer. 2016. Irreversible APCCdh1 Inactivation Underlies the Point of No Return for Cell-Cycle Entry. Cell. 166:167–180. doi:10.1016/j.cell.2016.05.077. Carrano, A.C., E. Eytan, A. Hershko, and M. Pagano. 1999. SKP2 is required for ubiquitin-mediated degradation of the CDK inhibitor p27. Nat. Cell Biol. 1:193–9. Chen, C.-Y., M.-S. Tsai, C.-Y. Lin, I.-S. Yu, Y.-T. Chen, S.-R. Lin, L.-W. Juan, Y.-T. Chen, H.-M. Hsu, L.-J. Lee, and S.-W. Lin. 2012. Rescue of the genetically engineered Cul4b mutant mouse as a potential model for human X-linked mental retardation. Hum. Mol. Genet. 21:4270–85. doi:10.1093/hmg/dds261. Chen, Z., K. Wang, C. Hou, K. Jiang, B. Chen, J. Chen, L. Lao, L. Qian, G. Zhong, Z. Liu, C. Zhang, and H. Shen. 2017. CRL4BDCAF11 E3 ligase targets p21 for degradation to control cell cycle progression in human osteosarcoma cells. Sci. Rep. 7:1175. doi:10.1038/s41598-017-01344-9. Choi, S.H., J.B. Wright, S.A. Gerber, and M.D. Cole. 2010. Myc protein is stabilized by suppression of a novel E3 ligase complex in cancer cells. Genes Dev. 24:1236–1241. doi:10.1101/gad.1920310. Choudhury, R., T. Bonacci, A. Arceci, D. Lahiri, C.A. Mills, J.L. Kernan, T.B. Branigan, J.A. DeCaprio, D.J. Burke, and M.J. Emanuele. 2016. APC/C and SCF(cyclin F) Constitute a Reciprocal Feedback Circuit Controlling S-Phase Entry. Cell Rep. 16:3359–72. doi:10.1016/j.celrep.2016.08.058.

1. General introduction

Clijsters, L., and R. Wolthuis. 2014. PIP-box-mediated degradation prohibits re-accumulation of Cdc6 during S phase. J. Cell Sci. 127:1336–45. doi:10.1242/jcs.145862. Clijsters, L., W. Van Zon, B. Ter Riet, E. Voets, M. Boekhout, J. Ogink, C. Rumpf-Kienzl, and R.M.F. Wolthuis. 2014. Inefficient degradation of cyclin B1 re-activates the spindle checkpoint right after sister chromatid disjunction. Cell Cycle. 13:2370–2378. doi:10.4161/cc.29336. Coverley, D., H. Laman, and R.A. Laskey. 2002. Distinct roles for cyclins E and A during DNA replication complex assembly and activation. Nat. Cell Biol. 4:523–528. doi:10.1038/ncb813. D’Angiolella, V., V. Donato, F.M. Forrester, Y.T. Jeong, C. Pellacani, Y. Kudo, A. Saraf, L. Florens, M.P. Washburn, and M. Pagano. 2012. Cyclin F-mediated degradation of ribonucleotide reductase M2 controls genome integrity and DNA repair. Cell. 149:1023–1034. doi:10.1016/j.cell.2012.03.043. Dankert, J.F., G. Rona, L. Clijsters, P. Geter, J.R. Skaar, K. Bermudez-Hernandez, E. Sassani, D. Fenyö, B. Ueberheide, R. Schneider, and M. Pagano. 2016. Cyclin F-Mediated Degradation of SLBP Limits H2A.X Accumulation and Apoptosis upon Genotoxic Stress in G2. Mol. Cell. 64:507–519. doi:10.1016/j.molcel.2016.09.010. Darzynkiewicz, Z., H. Zhao, S. Zhang, M.Y.W.T. Lee, E.Y.C. Lee, and Z. Zhang. 2015. Initiation and termination of DNA replication during S phase in relation to cyclins D1, E and A, p21WAF1, Cdt1 and the p12 subunit of DNA polymerase δ revealed in individual cells by cytometry. Oncotarget. 6:11735– 50. doi:10.18632/oncotarget.4149. Das-Bradoo, S., R.M. Ricke, and A.-K. Bielinsky. 2006. Interaction between PCNA and diubiquitinated Mcm10 is essential for cell growth in budding yeast. Mol. Cell. Biol. 26:4806–4817. doi:10.1128/MCB.02062-05. Deming, P.B., K.G. Flores, C.S. Downes, R.S. Paules, and W.K. Kaufmann. 2002. ATR Enforces the Topoisomerase II-dependent G2 Checkpoint through Inhibition of Plk1 Kinase. J. Biol. Chem. 277:36832–36838. doi:10.1074/jbc.M206109200. Dewar, J.M., E. Low, M. Mann, M. Räschle, and J.C. Walter. 2017. CRL2 Lrr1 promotes unloading of the vertebrate replisome from chromatin during replication termination. Genes Dev. 31:275–290. doi:10.1101/gad.291799.116. Dimova, N. V., N.A. Hathaway, B.-H. Lee, D.S. Kirkpatrick, M.L. Berkowitz, S.P. Gygi, D. Finley, and R.W. King. 2012. APC/C-mediated multiple monoubiquitylation provides an alternative degradation signal for cyclin B1. Nat. Cell Biol. 14:168–176. doi:10.1038/ncb2425. Dobrynin, G., O. Popp, T. Romer, S. Bremer, M.H.A. Schmitz, D.W. Gerlich, and H. Meyer. 2011. Cdc48/p97- Ufd1-Npl4 antagonizes Aurora B during chromosome segregation in HeLa cells. J. Cell. Biochem. 124:1571–1580. doi:10.1242/jcs.069500. Evans, D.L., H. Zhang, H. Ham, H. Pei, S. Lee, J. Kim, D.D. Billadeau, and Z. Lou. 2016. MMSET is dynamically regulated during cell-cycle progression and promotes normal DNA replication. Cell Cycle. 15:95–105. doi:10.1080/15384101.2015.1121323. Fischer, E.S., K. Böhm, J.R. Lydeard, H. Yang, M.B. Stadler, S. Cavadini, J. Nagel, F. Serluca, V. Acker, G.M. Lingaraju, R.B. Tichkule, M. Schebesta, W.C. Forrester, M. Schirle, U. Hassiepen, J. Ottl, M. Hild, R.E.J. Beckwith, J.W. Harper, J.L. Jenkins, and N.H. Thomä. 2014. Structure of the DDB1-CRBN E3 ubiquitin ligase in complex with thalidomide. Nature. 512:49–53. doi:10.1038/nature13527. Fischer, E.S., A. Scrima, K. Böhm, S. Matsumoto, G.M. Lingaraju, M. Faty, T. Yasuda, S. Cavadini, M. Wakasugi, F. Hanaoka, S. Iwai, H. Gut, K. Sugasawa, and N.H. Thomä. 2011. The molecular basis of CRL4 DDB2/CSA ubiquitin ligase architecture, targeting, and activation. Cell. 147:1024–1039. doi:10.1016/j.cell.2011.10.035. Foster, S.A., and D.O. Morgan. 2012. The APC/C Subunit Mnd2/Apc15 Promotes Cdc20 Autoubiquitination

1. General introduction

and Spindle Assembly Checkpoint Inactivation. Mol. Cell. 47:921–932. doi:10.1016/j.molcel.2012.07.031. Fousteri, M., W. Vermeulen, A.A. van Zeeland, and L.H.F. Mullenders. 2006. Cockayne syndrome A and B proteins differentially regulate recruitment of chromatin remodeling and repair factors to stalled RNA polymerase II in vivo. Mol. Cell. 23:471–82. doi:10.1016/j.molcel.2006.06.029. Fragkos, M., O. Ganier, P. Coulombe, and M. Méchali. 2015. DNA replication origin activation in space and time. Nat. Rev. Mol. Cell Biol. 16:360–374. doi:10.1038/nrm4002. Franz, A., L. Ackermann, and T. Hoppe. 2016. Ring of change: CDC48/p97 drives protein dynamics at chromatin. Front. Genet. 7. doi:10.3389/fgene.2016.00073. Franz, A., M. Orth, P.A. Pirson, R. Sonneville, J.J. Blow, A. Gartner, O. Stemmann, and T. Hoppe. 2011. CDC- 48/p97 Coordinates CDT-1 Degradation with GINS Chromatin Dissociation to Ensure Faithful DNA Replication. Mol. Cell. 44:85–96. doi:10.1016/j.molcel.2011.08.028. Fujimitsu, K., M. Grimaldi, and H. Yamano. 2016. Cyclin-dependent kinase 1-dependent activation of APC/C ubiquitin ligase. Science. 352:1121–4. doi:10.1126/science.aad3925. Fukushima, H., K. Ogura, L. Wan, Y. Lu, V. Li, D. Gao, P. Liu, A.W. Lau, T. Wu, M.W. Kirschner, H. Inuzuka, and W. Wei. 2013. SCF-mediated Cdh1 degradation defines a negative feedback system that coordinates cell-cycle progression. Cell Rep. 4:803–16. doi:10.1016/j.celrep.2013.07.031. Galanos, P., K. Vougas, D. Walter, A. Polyzos, A. Maya-Mendoza, E.J. Haagensen, A. Kokkalis, F.-M. Roumelioti, S. Gagos, M. Tzetis, B. Canovas, A. Igea, A.K. Ahuja, R. Zellweger, S. Havaki, E. Kanavakis, D. Kletsas, I.B. Roninson, S.D. Garbis, M. Lopes, A. Nebreda, D. Thanos, J.J. Blow, P. Townsend, C.S. Sørensen, J. Bartek, and V.G. Gorgoulis. 2016. Chronic p53-independent p21 expression causes genomic instability by deregulating replication licensing. Nat. Cell Biol. 18:777–89. doi:10.1038/ncb3378. Gao, Y.-F., T. Li, Y. Chang, Y.-B. Wang, W.-N. Zhang, W.-H. Li, K. He, R. Mu, C. Zhen, J.-H. Man, X. Pan, T. Li, L. Chen, M. Yu, B. Liang, Y. Chen, Q. Xia, T. Zhou, W.-L. Gong, A.-L. Li, H.-Y. Li, and X.-M. Zhang. 2011. Cdk1-phosphorylated CUEDC2 promotes spindle checkpoint inactivation and chromosomal instability. Nat. Cell Biol. 13:924–933. doi:10.1038/ncb2287. García-Rodríguez, N., R.P. Wong, and H.D. Ulrich. 2016. Functions of ubiquitin and SUMO in DNA replication and replication stress. Front. Genet. 7:87. doi:10.3389/fgene.2016.00087. Garnett, M.J., J. Mansfeld, C. Godwin, T. Matsusaka, J. Wu, P. Russell, J. Pines, and A.R. Venkitaraman. 2009. UBE2S elongates ubiquitin chains on APC/C substrates to promote mitotic exit. Nat. Cell Biol. 11:1363–1369. doi:10.1038/ncb1983. Ge, S., J.R. Skaar, and M. Pagano. 2009. APC/C- and Mad2-mediated degradation of Cdc20 during spindle checkpoint activation. Cell Cycle. 8:167–171. doi:10.4161/cc.8.1.7606. Ghosh, P., M. Wu, H. Zhang, and H. Sun. 2008. mTORC1 signaling requires proteasomal function and the involvement of CUL4-DDB1 ubiquitin E3 ligase. Cell Cycle. 7:373–381. doi:10.4161/cc.7.3.5267. Glotzer, M., A.W. Murray, and M.W. Kirschner. 1991. Cyclin is degraded by the ubiquitin pathway. Nature. 10:132–138. doi:10.1038/350055a0. Golan, A., Y. Yudkovsky, and A. Hershko. 2002. The Cyclin-Ubiquitin Ligase Activity of Cyclosome/APC Is Jointly Activated by Protein Kinases Cdk1-Cyclin B and Plk. J. Biol. Chem. 277:15552–15557. doi:10.1074/jbc.M111476200. Gorr, I.H., D. Boos, and O. Stemmann. 2005. Mutual inhibition of separase and Cdk1 by two-step complex formation. Mol. Cell. 19:135–141. doi:10.1016/j.molcel.2005.05.022. Gracheva, E., S. Chitale, T. Wilhelm, A. Rapp, J. Byrne, J. Stadler, R. Medina, M.C. Cardoso, and H. Richly.

1. General introduction

2016. ZRF1 mediates Remodeling of E3 ligases at DNA lesion sites during Nucleotide Excision Repair. J. Cell Biol. 213:185–200. doi:10.1083/jcb.201506099. Groh, B.S., F. Yan, M.D. Smith, Y. Yu, X. Chen, and Y. Xiong. 2016. The antiobesity factor WDTC1 suppresses adipogenesis via the CRL4WDTC1 E3 ligase. EMBO Rep. 17:e201540500. doi:10.15252/embr.201540500. Groisman, R., I. Kuraoka, O. Chevallier, N. Gaye, T. Magnaldo, K. Tanaka, A.F. Kisselev, A. Harel-Bellan, and Y. Nakatani. 2006. CSA-dependent degradation of CSB by the ubiquitin-proteasome pathway establishes a link between complementation factors of the Cockayne syndrome. Genes Dev. 20:1429–34. doi:10.1101/gad.378206. Grozeva, D., K. Carss, O. Spasic-Boskovic, M.-I. Tejada, J. Gecz, M. Shaw, M. Corbett, E. Haan, E. Thompson, K. Friend, Z. Hussain, A. Hackett, M. Field, A. Renieri, R. Stevenson, C. Schwartz, J.A.B. Floyd, J. Bentham, C. Cosgrove, B. Keavney, S. Bhattacharya, M. Hurles, and F.L. Raymond. 2015. Targeted Next-Generation Sequencing Analysis of 1,000 Individuals with Intellectual Disability. Hum. Mutat. 36:1197–204. doi:10.1002/humu.22901. Guerrero-Santoro, J., M.G. Kapetanaki, C.L. Hsieh, I. Gorbachinsky, A.S. Levine, and V. Rapić-Otrin. 2008. The cullin 4B-based UV-damaged DNA-binding protein ligase binds to UV-damaged chromatin and ubiquitinates histone H2A. Cancer Res. 68:5014–22. doi:10.1158/0008-5472.CAN-07-6162. Guo, Z., J. Kanjanapangka, N. Liu, S. Liu, C. Liu, Z. Wu, Y. Wang, T. Loh, C. Kowolik, J. Jamsen, M. Zhou, K. Truong, Y. Chen, L. Zheng, and B. Shen. 2012. Sequential Posttranslational Modifications Program FEN1 Degradation during Cell-Cycle Progression. Mol. Cell. 47:444–456. doi:10.1016/j.molcel.2012.05.042. Guturi, K.K.N., M. Bohgaki, T. Bohgaki, T. Srikumar, D. Ng, R. Kumareswaran, S. El Ghamrasni, J. Jeon, P. Patel, M.S. Eldin, R. Bristow, P. Cheung, G.S. Stewart, B. Raught, A. Hakem, and R. Hakem. 2016. RNF168 and USP10 regulate topoisomerase IIα function via opposing effects on its ubiquitylation. Nat. Commun. 7:12638. doi:10.1038/ncomms12638. Han, C., G. Wani, R. Zhao, J. Qian, N. Sharma, J. He, Q. Zhu, Q.-E. Wang, and A.A. Wani. 2015. Cdt2- mediated XPG degradation promotes gap-filling DNA synthesis in nucleotide excision repair. Cell Cycle. 14:1103–15. doi:10.4161/15384101.2014.973740. Han, J., Q. Li, L. McCullough, C. Kettelkamp, T. Formosa, and Z. Zhang. 2010. Ubiquitylation of FACT by the Cullin-E3 ligase Rtt101 connects FACT to DNA replication. Genes Dev. 24:1485–1490. doi:10.1101/gad.1887310. Han, J., H. Zhang, H. Zhang, Z. Wang, H. Zhou, and Z. Zhang. 2013a. A Cul4 E3 ubiquitin ligase regulates histone hand-off during nucleosome assembly. Cell. 155:817–829. doi:10.1016/j.cell.2013.10.014. Han, J.S., A.J. Holland, D. Fachinetti, A. Kulukian, B. Cetin, and D.W. Cleveland. 2013b. Catalytic Assembly of the Mitotic Checkpoint Inhibitor BubR1-Cdc20 by a Mad2-Induced Functional Switch in Cdc20. Mol. Cell. 51:92–104. doi:10.1016/j.molcel.2013.05.019. Hannah, J., and P. Zhou. 2015. Distinct and overlapping functions of the cullin E3 ligase scaffolding proteins CUL4A and CUL4B. Gene. 573:33–45. doi:10.1016/j.gene.2015.08.064. Havens, C.G., N. Shobnam, E. Guarino, R.C. Centore, L. Zou, S.E. Kearsey, and J.C. Walter. 2012. Direct role for proliferating cell nuclear antigen in substrate recognition by the E3 ubiquitin ligase CRL4Cdt2. J. Biol. Chem. 287:11410–11421. doi:10.1074/jbc.M111.337683. Havens, C.G., and J.C. Walter. 2009. Docking of a specialized PIP Box onto chromatin-bound PCNA creates a degron for the ubiquitin ligase CRL4Cdt2. Mol. Cell. 35:93–104. doi:10.1016/j.molcel.2009.05.012. He, Y.J., C.M. McCall, J. Hu, Y. Zeng, and Y. Xiong. 2006. DDB1 functions as a linker to recruit receptor WD40 proteins to CUL4-ROC1 ubiquitin ligases. Genes Dev. 20:2949–2954.

1. General introduction

doi:10.1101/gad.1483206. Hégarat, N., S. Rata, and H. Hochegger. 2016. Bistability of mitotic entry and exit switches during open mitosis in mammalian cells. BioEssays. 38:627–643. doi:10.1002/bies.201600057. Hellmuth, S., C. Pöhlmann, A. Brown, F. Böttger, M. Sprinzl, and O. Stemmann. 2015. Positive and negative regulation of vertebrate separase by Cdk1-cyclin B1 may explain why securin is dispensable. J. Biol. Chem. 290:8002–8010. doi:10.1074/jbc.M114.615310. Hernández-Pérez, S., E. Cabrera, H. Amoedo, S. Rodríguez-Acebes, S. Koundrioukoff, M. Debatisse, J. Méndez, and R. Freire. 2016. USP37 deubiquitinates Cdt1 and contributes to regulate DNA replication. Mol. Oncol. 1–11. doi:10.1016/j.molonc.2016.05.008. Hershko, A., D. Ganoth, J. Pehrson, R.E. Palazzo, and L.H. Cohen. 1991. Methylated ubiquitin inhibits cyclin degradation in clam embryo extracts. J. Biol. Chem. 266:16376–9. Higa, L.A., M. Wu, T. Ye, R. Kobayashi, H. Sun, and H. Zhang. 2006a. CUL4-DDB1 ubiquitin ligase interacts with multiple WD40-repeat proteins and regulates histone methylation. Nat. Cell Biol. 8:1277–83. doi:10.1038/ncb1490. Higa, L.A., X. Yang, J. Zheng, D. Banks, M. Wu, P. Ghosh, H. Sun, and H. Zhang. 2006b. Involvement of CUL4 ubiquitin E3 ligases in regulating CDK inhibitors Dacapo/p27Kip1 and cyclin E degradation. Cell Cycle. 5:71–77. doi:10.4161/cc.5.1.2266. Hirano, T. 2015. Chromosome Dynamics during Mitosis. Cold Spring Harb. Perspect Biol. 7:1–14. doi:10.1101/cshperspect.a015792. Hsu, J.Y., J.D.R. Reimann, C.S. Sørensen, J. Lukas, and P.K. Jackson. 2002. E2F-dependent accumulation of hEmi1 regulates S phase entry by inhibiting APC(Cdh1). Nat. Cell Biol. 4:358–366. doi:10.1038/ncb785. Hu, H., Y. Yang, Q. Ji, W. Zhao, B. Jiang, R. Liu, J. Yuan, Q. Liu, X. Li, Y. Zou, C. Shao, Y. Shang, Y. Wang, and Y. Gong. 2012. CRL4B Catalyzes H2AK119 Monoubiquitination and Coordinates with PRC2 to Promote Tumorigenesis. Cancer Cell. 22:781–795. doi:10.1016/j.ccr.2012.10.024. Hu, J., S. Zacharek, Y.J. He, H. Lee, S. Shumway, R.J. Duronio, and Y. Xiong. 2008. WD40 protein FBW5 promotes ubiquitination of tumor suppressor TSC2 by DDB1-CUL4-ROC1 ligase. Genes Dev. 22:866– 71. doi:10.1101/gad.1624008. Huang, X., M.K. Summers, V. Pham, J.R. Lill, J. Liu, G. Lee, D.S. Kirkpatrick, P.K. Jackson, G. Fang, and V.M. Dixit. 2011. Deubiquitinase USP37 Is Activated by CDK2 to Antagonize APCCDH1 and Promote S Phase Entry. Mol. Cell. 42:511–523. doi:10.1016/j.molcel.2011.03.027. Huh, J., and H. Piwnica-Worms. 2013. CRL4(CDT2) targets CHK1 for PCNA-independent destruction. Mol. Cell. Biol. 33:213–26. doi:10.1128/MCB.00847-12. Isidor, B., O. Pichon, S. Baron, A. David, and C. Le Caignec. 2010. Deletion of the CUL4B gene in a boy with mental retardation, minor facial anomalies, short stature, hypogonadism, and ataxia. Am. J. Med. Genet. Part A. 152:175–180. doi:10.1002/ajmg.a.33152. Izawa, D., and J. Pines. 2014. The mitotic checkpoint complex binds a second CDC20 to inhibit active APC/C. Nature. 517:631–634. doi:10.1038/nature13911. Ji, Q., H. Hu, F. Yang, J. Yuan, Y. Yang, L. Jiang, Y. Qian, B. Jiang, Y. Zou, Y. Wang, C. Shao, and Y. Gong. 2014. CRL4B interacts with and coordinates the SIN3A-HDAC complex to repress CDKN1A and drive cell cycle progression. J. Cell Sci. 127:4679–91. doi:10.1242/jcs.154245. Jia, L., B. Li, and H. Yu. 2016. The Bub1-Plk1 kinase complex promotes spindle checkpoint signalling through Cdc20 phosphorylation. Nat. Commun. 7:10818. doi:10.1038/ncomms10818. Jia, S., R. Kobayashi, and S.I.S. Grewal. 2005. Ubiquitin ligase component Cul4 associates with Clr4 histone

1. General introduction

methyltransferase to assemble heterochromatin. Nat. Cell Biol. 7:1007–1013. doi:10.1038/ncb1300. Jiang, B., W. Zhao, J. Yuan, Y. Qian, W. Sun, Y. Zou, C. Guo, B. Chen, C. Shao, and Y. Gong. 2012. Lack of Cul4b, an E3 ubiquitin ligase component, leads to embryonic lethality and abnormal placental development. PLoS One. 7. doi:10.1371/journal.pone.0037070. Jiang, L., R. Rong, M.S. Sheikh, and Y. Huang. 2011. Cullin-4A·DNA damage-binding protein 1 E3 ligase complex targets tumor suppressor RASSF1A for degradation during mitosis. J. Biol. Chem. 286:6971– 8. doi:10.1074/jbc.M110.186494. Jin, J., E.E. Arias, J. Chen, J.W. Harper, and J.C. Walter. 2006. A Family of Diverse Cul4-Ddb1-Interacting Proteins Includes Cdt2, which Is Required for S Phase Destruction of the Replication Factor Cdt1. Mol. Cell. 23:709–721. doi:10.1016/j.molcel.2006.08.010. Johansson, P., J. Jeffery, F. Al-Ejeh, R.B. Schulz, D.F. Callen, R. Kumar, and K.K. Khanna. 2014. SCF-FBXO31 E3 ligase targets DNA replication factor Cdt1 for proteolysis in the G2 phase of cell cycle to prevent re-replication. J. Biol. Chem. 289:18514–25. doi:10.1074/jbc.M114.559930. Kaisari, S., D. Sitry-Shevah, S. Miniowitz-Shemtov, and A. Hershko. 2016. Intermediates in the assembly of mitotic checkpoint complexes and their role in the regulation of the anaphase-promoting complex. Proc. Natl. Acad. Sci. U. S. A. 113:966–971. doi:10.1073/pnas.1524551113. Kang, X., C. Song, X. Du, C. Zhang, Y. Liu, L. Liang, J. He, K. Lamb, W.H. Shen, and Y. Yin. 2015. PTEN stabilizes TOP2A and regulates the DNA decatenation. Sci Rep. 5:17873. doi:10.1038/srep17873. Kapetanaki, M.G., J. Guerrero-Santoro, D.C. Bisi, C.L. Hsieh, V. Rapic-Otrin, and A.S. Levine. 2006. The DDB1-CUL4ADDB2 ubiquitin ligase is deficient in xeroderma pigmentosum group E and targets histone H2A at UV-damaged DNA sites. Proc. Natl. Acad. Sci. 103:2588–2593. doi:10.1073/pnas.0511160103. Kaur, M., M.M. Khan, A. Kar, A. Sharma, and S. Saxena. 2012. CRL4-DDB1-VPRBP ubiquitin ligase mediates the stress triggered proteolysis of Mcm10. Nucleic Acids Res. 40:7332–46. doi:10.1093/nar/gks366. Kerzendorfer, C., L. Hart, R. Colnaghi, G. Carpenter, D. Alcantara, E. Outwin, A.M. Carr, and M. O’Driscoll. 2011. CUL4B-deficiency in humans: Understanding the clinical consequences of impaired Cullin 4- RING E3 ubiquitin ligase function. Mech. Ageing Dev. 132:366–373. doi:10.1016/j.mad.2011.02.003. Kerzendorfer, C., A. Whibley, G. Carpenter, E. Outwin, S.C. Chiang, G. Turner, C. Schwartz, S. El-Khamisy, F.L. Raymond, and M. O’Driscoll. 2010. Mutations in Cullin 4B result in a human syndrome associated with increased camptothecin-induced topoisomerase I-dependent DNA breaks. Hum. Mol. Genet. 19:1324–1334. doi:10.1093/hmg/ddq008. Kim, Y., N.G. Starostina, and E.T. Kipreos. 2008. The CRL4Cdt2 ubiquitin ligase targets the degradation of p21Cip1 to control replication licensing. Genes Dev. 22:2507–2519. doi:10.1101/gad.1703708. Kopanja, D., N. Roy, T. Stoyanova, R.A. Hess, S. Bagchi, and P. Raychaudhuri. 2011. Cul4A is essential for spermatogenesis and male fertility. Dev. Biol. 352:278–287. doi:10.1016/j.ydbio.2011.01.028. Kramer, E.R., N. Scheuringer, A. V Podtelejnikov, M. Mann, and J.M. Peters. 2000. Mitotic regulation of the APC activator proteins CDC20 and CDH1. Mol. Biol. Cell. 11:1555–69. doi:10.1091/mbc.11.5.1555. Krönke, J., E.C. Fink, P.W. Hollenbach, K.J. MacBeth, S.N. Hurst, N.D. Udeshi, P.P. Chamberlain, D.R. Mani, H.W. Man, A.K. Gandhi, T. Svinkina, R.K. Schneider, M. McConkey, M. Järås, E. Griffiths, M. Wetzler, L. Bullinger, B.E. Cathers, S.A. Carr, R. Chopra, and B.L. Ebert. 2015. Lenalidomide induces ubiquitination and degradation of CK1α in del(5q) MDS. Nature. 523:183–188. doi:10.1038/nature14610. Kronke, J., N.D. Udeshi, A. Narla, P. Grauman, S.N. Hurst, M. McConkey, T. Svinkina, D. Heckl, E. Comer, X.

1. General introduction

Li, C. Ciarlo, E. Hartman, N. Munshi, M. Schenone, S.L. Schreiber, S.A. Carr, and B.L. Ebert. 2014. Lenalidomide Causes Selective Degradation of IKZF1 and IKZF3 in Multiple Myeloma Cells. Science. 343:301–305. doi:10.1126/science.1244851. Labib, K. 2010. How do Cdc7 and cyclin-dependent kinases trigger the initiation of chromosome replication in eukaryotic cells? Genes Dev. 24:1208–19. doi:10.1101/gad.1933010. Lee, J., J.-H. Shieh, J. Zhang, L. Liu, Y. Zhang, J.Y. Eom, G. Morrone, M.A.S. Moore, and P. Zhou. 2013. Improved ex vivo expansion of adult hematopoietic stem cells by overcoming CUL4-mediated degradation of HOXB4. Blood. 121. Lee, J., and P. Zhou. 2010. Cullins and Cancer. Genes Cancer. 1:690–699. doi:10.1177/1947601910382899. Lee, J.M., J.S. Lee, H. Kim, K. Kim, H. Park, J.Y. Kim, S.H. Lee, I.S. Kim, J. Kim, M. Lee, C.H. Chung, S.B. Seo, J.B. Yoon, E. Ko, D.Y. Noh, K. Il Kim, K.K. Kim, and S.H. Baek. 2012. EZH2 Generates a Methyl Degron that Is Recognized by the DCAF1/DDB1/CUL4 E3 Ubiquitin Ligase Complex. Mol. Cell. 48:572–586. doi:10.1016/j.molcel.2012.09.004. Lee, M.Y.W.T., S. Zhang, S.H.S. Lin, X. Wang, Z. Darzynkiewicz, Z. Zhang, and E.Y.C. Lee. 2014. The tail that wags the dog: P12, the smallest subunit of DNA polymerase δ, is degraded by ubiquitin ligases in response to DNA damage and during cell cycle progression. Cell Cycle. 13:23–31. doi:10.4161/cc.27407. Van Leung-Pineda, J. Huh, and H. Piwnica-Worms. 2009. DDB1 targets Chk1 to the Cul4 E3 ligase complex in normal cycling cells and in cells experiencing replication stress. Cancer Res. 69:2630–2637. doi:10.1158/0008-5472.CAN-08-3382. Li, G., T. Ji, J. Chen, Y. Fu, L. Hou, Y. Feng, T. Zhang, T. Song, J. Zhao, Y. Endo, H. Lin, X. Cai, and Y. Cang. 2017. CRL4 DCAF8 Ubiquitin Ligase Targets Histone H3K79 and Promotes H3K9 Methylation in the Liver. Cell Rep. 18:1499–1511. doi:10.1016/j.celrep.2017.01.039. Li, J., V. D’Angiolella, E.S. Seeley, S. Kim, T. Kobayashi, W. Fu, E.I. Campos, M. Pagano, and B.D. Dynlacht. 2013. USP33 regulates centrosome biogenesis via deubiquitination of the centriolar protein CP110. Nature. 495:255–9. doi:10.1038/nature11941. Li, T., X. Chen, K.C. Garbutt, P. Zhou, and N. Zheng. 2006. Structure of DDB1 in complex with a paramyxovirus V protein: Viral Hijack of a propeller cluster in ubiquitin ligase. Cell. 124:105–117. doi:10.1016/j.cell.2005.10.033. Li, T., E.I. Robert, P.C. van Breugel, M. Strubin, and N. Zheng. 2010. A promiscuous alpha-helical motif anchors viral hijackers and substrate receptors to the CUL4-DDB1 ubiquitin ligase machinery. Nat. Struct. Mol. Biol. 17:105–111. doi:10.1038/nsmb.1719. Li, W., J. Cooper, L. Zhou, C. Yang, H. Erdjument-Bromage, D. Zagzag, M. Snuderl, M. Ladanyi, C.O. Hanemann, P. Zhou, M.A. Karajannis, and F.G. Giancotti. 2014. Merlin/NF2 loss-driven tumorigenesis linked to CRL4(DCAF1)-mediated inhibition of the hippo pathway kinases Lats1 and 2 in the nucleus. Cancer Cell. 26:48–60. doi:10.1016/j.ccr.2014.05.001. Lin, C.-Y., C.-Y. Chen, C.-H. Yu, I.-S. Yu, S.-R. Lin, J.-T. Wu, Y.-H. Lin, P.-L. Kuo, J.-C. Wu, and S.-W. Lin. 2016. Human X-linked Intellectual Disability Factor CUL4B Is Required for Post-meiotic Sperm Development and Male Fertility. Sci. Rep. 6:20227. doi:10.1038/srep20227. Lin, Z., C. Tan, Q. Qiu, S. Kong, H. Yang, F. Zhao, Z. Liu, J. Li, Q. Kong, B. Gao, T. Barrett, G.-Y. Yang, J. Zhang, and D. Fang. 2015. Ubiquitin-specific protease 22 is a deubiquitinase of CCNB1. Cell Discov. 1:15028. doi:10.1038/celldisc.2015.28. Lischetti, T., and J. Nilsson. 2015. Regulation of mitotic progression by the spindle assembly checkpoint. Mol. Cell. Oncol. 2:e970484. doi:10.4161/23723548.2014.970484.

1. General introduction

List, A., S. Kurtin, D.J. Roe, A. Buresh, D. Mahadevan, D. Fuchs, L. Rimsza, R. Heaton, R. Knight, and J.B. Zeldis. 2005. Efficacy of Lenalidomide in Myelodysplastic Syndromes. N. Engl. J. Med. 352:549–557. doi:10.1056/NEJMoa041668. Liu, H.C., G. Enikolopov, and Y. Chen. 2012a. Cul4B regulates neural progenitor cell growth. BMC Neurosci. 13:112. doi:10.1186/1471-2202-13-112. Liu, L., Y. Yin, Y. Li, L. Prevedel, E.H. Lacy, L. Ma, and P. Zhou. 2012b. Essential role of the CUL4B ubiquitin ligase in extra-embryonic tissue development during mouse embryogenesis. Cell Res. 22:1258–1269. doi:10.1038/cr.2012.48. Livneh, I., V. Cohen-Kaplan, C. Cohen-Rosenzweig, N. Avni, and A. Ciechanover. 2016. The life cycle of the 26S proteasome: from birth, through regulation and function, and onto its death. Cell Res. 26:869– 885. doi:10.1038/cr.2016.86. Lobjois, V., D. Jullien, J.-P. Bouché, and B. Ducommun. 2009. The polo-like kinase 1 regulates CDC25B- dependent mitosis entry. Biochim. Biophys. Acta - Mol. Cell Res. 1793:462–468. doi:10.1016/j.bbamcr.2008.12.015. Londin, E.R., J. Adijanto, N. Philp, A. Novelli, E. Vitale, C. Perria, G. Serra, V. Alesi, S. Surrey, and P. Fortina. 2014. Donor splice-site mutation in CUL4B is likely cause of X-linked intellectual disability. Am. J. Med. Genet. Part A. 164:2294–2299. doi:10.1002/ajmg.a.36629. Lou, Z., K. Minter-Dykhouse, and J. Chen. 2005. BRCA1 participates in DNA decatenation. Nat. Struct. Mol. Biol. 12:589–593. doi:10.1038/nsmb953. Lu, G., R.E. Middleton, H. Sun, M. Naniong, C.J. Ott, C.S. Mitsiades, K.-K. Wong, J.E. Bradner, and W.G. Kaelin. 2014. The myeloma drug lenalidomide promotes the cereblon-dependent destruction of Ikaros proteins. Science. 343:305–9. doi:10.1126/science.1244917. Lukas, C., C.S. Sørensen, E. Kramer, E. Santoni-Rugiu, C. Lindeneg, J.M. Peters, J. Bartek, and J. Lukas. 1999. Accumulation of cyclin B1 requires E2F and cyclin-A-dependent rearrangement of the anaphase- promoting complex. Nature. 401:815–818. doi:10.1038/44611. Luo, K., J. Yuan, J. Chen, and Z. Lou. 2009. Topoisomerase IIα controls the decatenation checkpoint. Nat. Cell Biol. 11:204–210. doi:10.1038/ncb1828. Lydeard, J.R., B.A. Schulman, and J.W. Harper. 2013. Building and remodelling Cullin-RING E3 ubiquitin ligases. EMBO Rep. 14:1050–1061. doi:10.1038/embor.2013.173. Mailand, N., I. Gibbs-Seymour, and S. Bekker-Jensen. 2013. Regulation of PCNA–protein interactions for genome stability. Nat. Rev. Mol. Cell Biol. 14:269–282. doi:10.1038/nrm3562. Mansfeld, J., P. Collin, M.O. Collins, J.S. Choudhary, and J. Pines. 2011. APC15 drives the turnover of MCC- CDC20 to make the spindle assembly checkpoint responsive to kinetochore attachment. Nat. Cell Biol. 13:1234–1243. doi:10.1038/ncb2347. Mansilla, S.F., G. Soria, M.B. Vallerga, M. Habif, W. Martínez-López, C. Prives, and V. Gottifredi. 2013. UV- triggered p21 degradation facilitates damaged-DNA replication and preserves genomic stability. Nucleic Acids Res. 41:6942–6951. doi:10.1093/nar/gkt475. Maric, M., T. Maculins, G. De Piccoli, and K. Labib. 2014. Cdc48 and a ubiquitin ligase drive disassembly of the CMG helicase at the end of DNA replication. Science. 346:1253596. doi:10.1126/science.1253596. Matson, J.P., and J.G. Cook. 2016. Cell cycle proliferation decisions: the impact of single cell analyses. FEBS J. doi:10.1111/febs.13898. Matyskiela, M.E., G. Lu, T. Ito, B. Pagarigan, C.-C. Lu, K. Miller, W. Fang, N.-Y. Wang, D. Nguyen, J. Houston, G. Carmel, T. Tran, M. Riley, L. Nosaka, G.C. Lander, S. Gaidarova, S. Xu, A.L. Ruchelman, H. Handa, J.

1. General introduction

Carmichael, T.O. Daniel, B.E. Cathers, A. Lopez-Girona, and P.P. Chamberlain. 2016. A novel cereblon modulator recruits GSPT1 to the CRL4CRBN ubiquitin ligase. Nature. 535:252–257. doi:10.1038/nature18611. Michishita, M., A. Morimoto, T. Ishii, H. Komori, Y. Shiomi, Y. Higuchi, and H. Nishitani. 2011. Positively charged residues located downstream of PIP box, together with TD amino acids within PIP box, are important for CRL4Cdt2-mediated proteolysis. Genes to Cells. 16:12–22. doi:10.1111/j.1365- 2443.2010.01464.x. Miniowitz-Shemtov, S., E. Eytan, S. Kaisari, D. Sitry-Shevah, and A. Hershko. 2015. Mode of interaction of TRIP13 AAA-ATPase with the Mad2-binding protein p31comet and with mitotic checkpoint complexes. Proc. Natl. Acad. Sci. U. S. A. 112:11536–11540. doi:10.1073/pnas.1515358112. Miranda-Carboni, G.A., S.A. Krum, K. Yee, M. Nava, Q.E. Deng, S. Pervin, A. Collado-Hidalgo, Z. Galic, J.A. Zack, K. Nakayama, K.I. Nakayama, and T.F. Lane. 2008. A functional link between Wnt signaling and SKP2-independent p27 turnover in mammary tumors. Genes Dev. 22:3121–3134. doi:10.1101/gad.1692808. Moldovan, G.L., B. Pfander, and S. Jentsch. 2007. PCNA, the Maestro of the Replication Fork. Cell. 129:665– 679. doi:10.1016/j.cell.2007.05.003. Moreno, S.P., R. Bailey, N. Campion, S. Herron, and A. Gambus. 2014. Polyubiquitylation drives replisome disassembly at the termination of DNA replication. Science. 346:477–481. doi:10.1126/science.1253585. Moreno, S.P., and A. Gambus. 2015. Regulation of Unperturbed DNA Replication by Ubiquitylation. Genes (Basel). 6:451–468. doi:10.3390/genes6030451. Morgan, D.O., and J.M. Roberts. 2002. Cell cycle: Oscillation sensation. Nature. 418:495–496. doi:10.1038/418495a. Mouysset, J., S. Gilberto, M.G. Meier, F. Lampert, M. Belwal, P. Meraldi, and M. Peter. 2015. CRL4RBBP7 is required for efficient CENP-A deposition at centromeres. J. Cell Sci. 7:1732–1745. doi:10.1242/jcs.162305. Nakagawa, T., and Y. Xiong. 2011. X-Linked Mental Retardation Gene CUL4B Targets Ubiquitylation of H3K4 Methyltransferase Component WDR5 and Regulates Neuronal Gene Expression. Mol. Cell. 43:381–391. doi:10.1016/j.molcel.2011.05.033. Nguyen, T. Van, J. Li, C.-C. (Jean) Lu, J.L. Mamrosh, G. Lu, B.E. Cathers, and R.J. Deshaies. 2017. p97/VCP promotes degradation of CRBN substrate glutamine synthetase and neosubstrates. Proc. Natl. Acad. Sci. 114:3565–3571. doi:10.1073/pnas.1700949114. Niikura, Y., R. Kitagawa, H. Ogi, R. Abdulle, V. Pagala, and K. Kitagawa. 2015. CENP-A K124 Ubiquitylation Is Required for CENP-A Deposition at the Centromere. Dev. Cell. 32:589–603. doi:10.1016/j.devcel.2015.01.024. Nilsson, J., M. Yekezare, J. Minshull, and J. Pines. 2008. The APC/C maintains the spindle assembly checkpoint by targeting Cdc20 for destruction. Nat Cell Biol. 10:1411–1420. doi:10.1038/ncb1799. Nishitani, H., Y. Shiomi, H. Iida, M. Michishita, T. Takami, and T. Tsurimoto. 2008. CDK inhibitor p21 is degraded by a proliferating cell nuclear antigen-coupled Cul4-DDB1Cdt2 pathway during s phase and after UV irradiation. J. Biol. Chem. 283:29045–29052. doi:10.1074/jbc.M806045200. Oda, H., M.R. Hübner, D.B. Beck, M. Vermeulen, J. Hurwitz, D.L. Spector, and D. Reinberg. 2010. Regulation of the Histone H4 Monomethylase PR-Set7 by CRL4Cdt2-Mediated PCNA-Dependent Degradation during DNA Damage. Mol. Cell. 40:364–376. doi:10.1016/j.molcel.2010.10.011. Ohtake, F., A. Baba, I. Takada, M. Okada, K. Iwasaki, H. Miki, S. Takahashi, A. Kouzmenko, K. Nohara, T.

1. General introduction

Chiba, Y. Fujii-Kuriyama, and S. Kato. 2007. Dioxin receptor is a ligand-dependent E3 ubiquitin ligase. Nature. 446:562–566. doi:10.1038/nature05683. Okamoto, N., M. Watanabe, T. Naruto, K. Matsuda, T. Kohmoto, M. Saito, K. Masuda, and I. Imoto. 2017. Genome-first approach diagnosed Cabezas syndrome via novel CUL4B mutation detection. Hum. genome Var. 4:16045. doi:10.1038/hgv.2016.45. Olma, M.H., M. Roy, T. Le Bihan, I. Sumara, S. Maerki, B. Larsen, M. Quadroni, M. Peter, M. Tyers, and L. Pintard. 2009. An interaction network of the mammalian COP9 signalosome identifies Dda1 as a core subunit of multiple Cul4-based E3 ligases. J. Cell Sci. 122:1035–44. doi:10.1242/jcs.043539. Orthwein, A., S.M. Noordermeer, M.D. Wilson, S. Landry, R.I. Enchev, A. Sherker, M. Munro, J. Pinder, J. Salsman, G. Dellaire, B. Xia, M. Peter, and D. Durocher. 2015. A mechanism for the suppression of homologous recombination in G1 cells. Nature. 528:422–426. doi:10.1038/nature16142. Pagano, M., R. Pepperkok, F. Verde, W. Ansorge, and G. Draetta. 1992. Cyclin A is required at two points in the human cell cycle. EMBO J. 11:961–71. Peng, Z., Z. Liao, Y. Matsumoto, A. Yang, and A.E. Tomkinson. 2016. Human DNA Ligase I Interacts with and Is Targeted for Degradation by the DCAF7 Specificity Factor of the Cul4-DDB1 Ubiquitin Ligase Complex. J. Biol. Chem. 291:21893–21902. doi:10.1074/jbc.M116.746198. Pesenti, M.E., J.R. Weir, and A. Musacchio. 2016. Progress in the structural and functional characterization of kinetochores. Curr. Opin. Struct. Biol. 37:152–63. doi:10.1016/j.sbi.2016.03.003. Petroski, M.D., and R.J. Deshaies. 2005. Function and regulation of cullin-RING ubiquitin ligases. Nat. Rev. Mol. Cell Biol. 6:9–20. doi:10.1038/nrm1547. Potapova, T.A., J.R. Daum, B.D. Pittman, J.R. Hudson, T.N. Jones, D.L. Satinover, P.T. Stukenberg, and G.J. Gorbsky. 2006. The reversibility of mitotic exit in vertebrate cells. Nature. 440:954–958. doi:10.1038/nature04652. Qiao, R., F. Weissmann, M. Yamaguchi, N.G. Brown, R. VanderLinden, R. Imre, M.A. Jarvis, M.R. Brunner, I.F. Davidson, G. Litos, D. Haselbach, K. Mechtler, H. Stark, B.A. Schulman, and J.-M. Peters. 2016. Mechanism of APC/CCDC20 activation by mitotic phosphorylation. Proc. Natl. Acad. Sci. U. S. A. 113:E2570–E2578. doi:10.1073/pnas.1604929113. Qin, B., B. Gao, J. Yu, J. Yuan, and Z. Lou. 2013. Ataxia telangiectasia-mutated- and Rad3-related protein regulates the DNA damage-induced G2/M checkpoint through the Aurora A cofactor Bora protein. J. Biol. Chem. 288:16139–16144. doi:10.1074/jbc.M113.456780. Ramadan, K., R. Bruderer, F.M. Spiga, O. Popp, T. Baur, M. Gotta, and H.H. Meyer. 2007. Cdc48/p97 promotes reformation of the nucleus by extracting the kinase Aurora B from chromatin. Nature. 450:1258–1262. doi:10.1038/nature06388. Raman, M., C.G. Havens, J.C. Walter, and J.W. Harper. 2011. A genome-wide screen identifies p97 as an essential regulator of DNA damage-dependent CDT1 destruction. Mol. Cell. 44:72–84. doi:10.1016/j.molcel.2011.06.036. Rattani, A., P.K. Vinod, J. Godwin, K. Tachibana-Konwalski, M. Wolna, M. Malumbres, B. Novák, and K. Nasmyth. 2014. Dependency of the spindle assembly checkpoint on Cdk1 renders the anaphase transition irreversible. Curr. Biol. 24:630–637. doi:10.1016/j.cub.2014.01.033. Rizzardi, L.F., K.E. Coleman, D. Varma, J.P. Matson, S. Oh, and J.G. Cook. 2015. CDK1-dependent Inhibition of the E3 Ubiquitin Ligase CRL4CDT2 Ensures Robust Transition from S Phase to Mitosis. J. Biol. Chem. 290:556–567. doi:10.1074/JBC.M114.614701. Scrima, A., R. Koníčková, B.K. Czyzewski, Y. Kawasaki, P.D. Jeffrey, R. Groisman, Y. Nakatani, S. Iwai, N.P. Pavletich, and N.H. Thomä. 2008. Structural Basis of UV DNA-Damage Recognition by the DDB1-

1. General introduction

DDB2 Complex. Cell. 135:1213–1223. doi:10.1016/j.cell.2008.10.045. Senga, T., U. Sivaprasad, W. Zhu, J.P. Jong, E.E. Arias, J.C. Walter, A. Dutta, J.H. Park, E.E. Arias, J.C. Walter, and A. Dutta. 2006. PCNA is a cofactor for CDT1 degradation by CUL4/DDB1-mediated N-terminal ubiquitination. J. Biol. Chem. 281:6246–6252. doi:10.1074/jbc.M512705200. Sengupta, S., and R.W. Henry. 2015. Regulation of the retinoblastoma-E2F pathway by the ubiquitin- proteasome system. Biochim. Biophys. Acta. 1849:1289–97. doi:10.1016/j.bbagrm.2015.08.008. Siddiqui, K., K.F. On, and J.F.X. Diffley. 2013. Regulating DNA Replication in Eukarya. Cold Spring Harb. Perspect. Biol. 5:a012930–a012930. doi:10.1101/cshperspect.a012930. Song, L., A. Craney, and M. Rape. 2014. Microtubule-Dependent Regulation of Mitotic Protein Degradation. Mol. Cell. 53:179–192. doi:10.1016/j.molcel.2013.12.022. Spencer, S.L., S.D. Cappell, F.-C.C. Tsai, K.W. Overton, C.L. Wang, and T. Meyer. 2013. The proliferation- quiescence decision is controlled by a bifurcation in CDK2 activity at mitotic exit. Cell. 155:369–383. doi:10.1016/j.cell.2013.08.062. Sugasawa, K., Y. Okuda, M. Saijo, R. Nishi, N. Matsuda, G. Chu, T. Mori, S. Iwai, K. Tanaka, K. Tanaka, and F. Hanaoka. 2005. UV-Induced Ubiquitylation of XPC Protein Mediated by UV-DDB-Ubiquitin Ligase Complex. Cell. 121:387–400. doi:10.1016/j.cell.2005.02.035. Sumara, I., M. Quadroni, C. Frei, M.H. Olma, G. Sumara, R. Ricci, and M. Peter. 2007. A Cul3-Based E3 Ligase Removes Aurora B from Mitotic Chromosomes, Regulating Mitotic Progression and Completion of Cytokinesis in Human Cells. Dev. Cell. 12:887–900. doi:10.1016/j.devcel.2007.03.019. Sung, C.K., J. Dahl, H. Yim, S. Rodig, and T.L. Benjamin. 2011. Transcriptional and post-translational regulation of the quiescence factor and putative tumor suppressor p150(Sal2). FASEB J. 25:1275–83. doi:10.1096/fj.10-173674. Takeda, D.Y., J.D. Parvin, and A. Dutta. 2005. Degradation of Cdt1 during S phase is Skp2-independent and is required for efficient progression of mammalian cells through S phase. J. Biol. Chem. 280:23416– 23. doi:10.1074/jbc.M501208200. Tarpey, P.S., F.L. Raymond, S. O’Meara, S. Edkins, J. Teague, A. Butler, E. Dicks, C. Stevens, C. Tofts, T. Avis, S. Barthorpe, G. Buck, J. Cole, K. Gray, K. Halliday, R. Harrison, K. Hills, A. Jenkinson, D. Jones, A. Menzies, T. Mironenko, J. Perry, K. Raine, D. Richardson, R. Shepherd, A. Small, J. Varian, S. West, S. Widaa, U. Mallya, J. Moon, Y. Luo, S. Holder, S.F. Smithson, J. a Hurst, J. Clayton-Smith, B. Kerr, J. Boyle, M. Shaw, L. Vandeleur, J. Rodriguez, R. Slaugh, D.F. Easton, R. Wooster, M. Bobrow, A.K. Srivastava, R.E. Stevenson, C.E. Schwartz, G. Turner, J. Gecz, P.A. Futreal, M.R. Stratton, and M. Partington. 2007. Mutations in CUL4B, which encodes a ubiquitin E3 ligase subunit, cause an X-linked mental retardation syndrome associated with aggressive outbursts, seizures, relative macrocephaly, central obesity, hypogonadism, pes cavus, and tremor. Am. J. Hum. Genet. 80:345–352. doi:10.1086/511134. Teixeira, L.K., and S.I. Reed. 2013. Ubiquitin ligases and cell cycle control. Annu. Rev. Biochem. 82:387– 414. doi:10.1146/annurev-biochem-060410-105307. Terai, K., E. Shibata, T. Abbas, and A. Dutta. 2013. Degradation of p12 subunit by CRL4Cdt2 E3 ligase inhibits fork progression after DNA damage. J. Biol. Chem. 288:30509–14. doi:10.1074/jbc.C113.505586. Toyoshima-Morimoto, F., E. Taniguchi, and E. Nishida. 2002. Plk1 promotes nuclear translocation of human Cdc25C during prophase. EMBO Rep. 3:341–348. doi:10.1093/embo-reports/kvf069. Toyoshima, H., and T. Hunter. 1994. p27, a novel inhibitor of G1 cyclin-Cdk protein kinase activity, is related to p21. Cell. 78:67–74. doi:10.1016/0092-8674(94)90573-8.

1. General introduction

Tsanov, N., C. Kermi, P. Coulombe, S. Van der Laan, D. Hodroj, and D. Maiorano. 2014. PIP degron proteins, substrates of CRL4Cdt2, and not PIP boxes, interfere with DNA polymerase η and κ focus formation on UV damage. Nucleic Acids Res. 42:3692–706. doi:10.1093/nar/gkt1400. Tsvetkov, L.M., K.-H. Yeh, S.-J. Lee, H. Sun, and H. Zhang. 1999. p27Kip1 ubiquitination and degradation is regulated by the SCFSkp2 complex through phosphorylated Thr187 in p27. Curr. Biol. 9:661-S2. doi:10.1016/S0960-9822(99)80290-5. Ulane, C.M., and C.M. Horvath. 2002. Paramyxoviruses SV5 and HPIV2 assemble STAT protein ubiquitin ligase complexes from cellular components. Virology. 304:160–6. Varetti, G., C. Guida, S. Santaguida, E. Chiroli, and A. Musacchio. 2011. Homeostatic control of mitotic arrest. Mol. Cell. 44:710–720. doi:10.1016/j.molcel.2011.11.014. Vázquez-Novelle, M.D., L. Sansregret, A.E. Dick, C.A. Smith, A.D. Mcainsh, D.W. Gerlich, and M. Petronczki. 2014. Cdk1 inactivation terminates mitotic checkpoint surveillance and stabilizes kinetochore attachments in anaphase. Curr. Biol. 24:638–645. doi:10.1016/j.cub.2014.01.034. Voets, E., and R. Wolthuis. 2015. MASTL promotes cyclin B1 destruction by enforcing Cdc20-independent binding of cyclin B1 to the APC/C. Biol. Open. 4:1–12. doi:10.1242/bio.201410793. Vulto-van Silfhout, A.T., T. Nakagawa, N. Bahi-Buisson, S.A. Haas, H. Hu, M. Bienek, L.E.L.M. Vissers, C. Gilissen, A. Tzschach, A. Busche, J. Müsebeck, P. Rump, I.B. Mathijssen, K. Avela, M. Somer, F. Doagu, A.K. Philips, A. Rauch, A. Baumer, K. Voesenek, K. Poirier, J. Vigneron, D. Amram, S. Odent, M. Nawara, E. Obersztyn, J. Lenart, A. Charzewska, N. Lebrun, U. Fischer, W.M. Nillesen, H.G. Yntema, I. Järvelä, H.H. Ropers, B.B.A. de Vries, H.G. Brunner, H. van Bokhoven, F.L. Raymond, M.A.A.P. Willemsen, J. Chelly, Y. Xiong, A.J. Barkovich, V.M. Kalscheuer, T. Kleefstra, and A.P.M. de Brouwer. 2015. Variants in CUL4B are associated with cerebral malformations. Hum. Mutat. 36:106–117. doi:10.1002/humu.22718. Wade Harper, J., G.R. Adami, N. Wei, K. Keyomarsi, and S.J. Elledge. 1993. The p21 Cdk-interacting protein Cip1 is a potent inhibitor of G1 cyclin-dependent kinases. Cell. 75:805–816. doi:10.1016/0092- 8674(93)90499-G. Walter, D., S. Hoffmann, E.-S. Komseli, J. Rappsilber, V. Gorgoulis, and C.S. Sørensen. 2016. SCFCyclin F- dependent degradation of CDC6 suppresses DNA re-replication. Nat. Commun. 7:10530. doi:10.1038/ncomms10530. Wang, H., L. Zhai, J. Xu, H.-Y. Joo, S. Jackson, H. Erdjument-bromage, P. Tempst, C. Hill, N. Carolina, Y. Xiong, and Y. Zhang. 2006. Histone H3 and H4 ubiquitylation by the CUL4-DDB-ROC1 ubiquitin ligase facilitates cellular response to DNA damage. Mol. Cell. 22:383–94. doi:10.1016/j.molcel.2006.03.035. Wang, X., A. Arceci, K. Bird, C.A. Mills, R. Choudhury, J.L. Kernan, C. Zhou, V. Bae-Jump, A. Bowers, and M.J. Emanuele. 2017. VprBP/DCAF1 Regulates the Degradation and Nonproteolytic Activation of the Cell Cycle Transcription Factor FoxM1. Mol. Cell. Biol. 37:e00609-16. doi:10.1128/MCB.00609-16. Wang, Y., Z. Zheng, J. Zhang, Y. Wang, R. Kong, J. Liu, Y. Zhang, H. Deng, X. Du, and Y. Ke. 2015. A Novel Retinoblastoma Protein (RB) E3 Ubiquitin Ligase (NRBE3) Promotes RB Degradation and Is Transcriptionally Regulated by E2F1 Transcription Factor. J. Biol. Chem. 290:28200–13. doi:10.1074/jbc.M115.655597. Watanabe, N., H. Arai, J.-I. Iwasaki, M. Shiina, K. Ogata, T. Hunter, and H. Osada. 2005. Cyclin-dependent kinase (CDK) phosphorylation destabilizes somatic Wee1 via multiple pathways. Proc. Natl. Acad. Sci. U. S. A. 102:11663–11668. doi:10.1073/pnas.0500410102. Wei, W., N.G. Ayad, Y. Wan, G.-J. Zhang, M.W. Kirschner, and W.G. Kaelin. 2004. Degradation of the SCF component Skp2 in cell-cycle phase G1 by the anaphase-promoting complex. Nature. 428:194–8.

1. General introduction

doi:10.1038/nature02381. Wenzel, D.M., and R.E. Klevit. 2012. Following Ariadne’s thread: a new perspective on RBR ubiquitin ligases. BMC Biol. 10:24. doi:10.1186/1741-7007-10-24. Wertz, I.E., K.M. O’Rourke, Z. Zhang, D. Dornan, D. Arnott, R.J. Deshaies, and V.M. Dixit. 2004. Human De- Etiolated-1 Regulates c-Jun by Assembling a CUL4A Ubiquitin Ligase. Science. 303:1371–1374. doi:10.1126/science.1093549. Westhorpe, F.G., a. Tighe, P. Lara-Gonzalez, and S.S. Taylor. 2011. p31comet-mediated extraction of Mad2 from the MCC promotes efficient mitotic exit. J. Cell Sci. 124:3905–3916. doi:10.1242/jcs.093286. Wieser, S., and J. Pines. 2015. The Biochemistry of Mitosis. Cold Spring Harb. Perspect. Biol. 7:a015776. doi:10.1101/cshperspect.a015776. Wittschieben, B.Ø., S. Iwai, and R.D. Wood. 2005. DDB1-DDB2 (xeroderma pigmentosum group E) protein complex recognizes a cyclobutane pyrimidine dimer, mismatches, apurinic/apyrimidinic sites, and compound lesions in DNA. J. Biol. Chem. 280:39982–9. doi:10.1074/jbc.M507854200. Wu, Y., L. Zhou, X. Wang, J. Lu, R. Zhang, X. Liang, L. Wang, W. Deng, Y.-X. Zeng, H. Huang, and T. Kang. 2016. A genome-scale CRISPR-Cas9 screening method for protein stability reveals novel regulators of Cdc25A. Cell Discov. 2:16014. doi:10.1038/celldisc.2016.14. Yamaguchi, M., R. VanderLinden, F. Weissmann, R. Qiao, P. Dube, N.G. Brown, D. Haselbach, W. Zhang, S.S. Sidhu, J.M. Peters, H. Stark, and B.A. Schulman. 2016. Cryo-EM of Mitotic Checkpoint Complex- Bound APC/C Reveals Reciprocal and Conformational Regulation of Ubiquitin Ligation. Mol. Cell. 63:593–607. doi:10.1016/j.molcel.2016.07.003. Yang, Y., R. Liu, R. Qiu, Y. Zheng, W. Huang, H. Hu, Q. Ji, H. He, Y. Shang, Y. Gong, and Y. Wang. 2013. CRL4B promotes tumorigenesis by coordinating with SUV39H1/HP1/DNMT3A in DNA methylation-based epigenetic silencing. Oncogene. 1–15. doi:10.1038/onc.2013.522. Yao, G., T.J. Lee, S. Mori, J.R. Nevins, and L. You. 2008. A bistable Rb-E2F switch underlies the restriction point. Nat. Cell Biol. 10:476–482. doi:10.1038/ncb1711. Ye, Y., and M. Rape. 2009. Building ubiquitin chains: E2 enzymes at work. Nat. Rev. Mol. Cell Biol. 10:755– 764. doi:10.1038/nrm2780. Yeong, F.M., H.H. Lim, C.G. Padmashree, and U. Surana. 2000. Exit from Mitosis in Budding Yeast: Biphasic Inactivation of the Cdc28-Clb2 Mitotic Kinase and the Role of Cdc20. Mol. Cell. 5:501–511. doi:10.1016/S1097-2765(00)80444-X. Yin, Y., C. Lin, S.T. Kim, I. Roig, H. Chen, L. Liu, G.M. Veith, R.U. Jin, S. Keeney, M. Jasin, K. Moley, P. Zhou, and L. Ma. 2011. The E3 ubiquitin ligase Cullin 4A regulates meiotic progression in mouse spermatogenesis. Dev. Biol. 356:51–62. doi:10.1016/j.ydbio.2011.05.661. Yin, Y., L. Liu, C. Yang, C. Lin, G.M. Veith, C. Wang, P. Sutovsky, P. Zhou, and L. Ma. 2016. Cell autonomous and non-autonomous function of CUL4B in Mouse Spermatogenesis. 291. jbc. M115.699660 pp. Yu, H., J.M. Peters, R.W. King, A.M. Page, P. Hieter, and M.W. Kirschner. 1998. Identification of a cullin homology region in a subunit of the anaphase-promoting complex. Science. 279:1219–1222. doi:10.1126/science.279.5354.1219. Zaidi, I.W., G. Rabut, A. Poveda, H. Scheel, J. Malmström, H. Ulrich, K. Hofmann, P. Pasero, M. Peter, and B. Luke. 2008. Rtt101 and Mms1 in budding yeast form a CUL4(DDB1)-like ubiquitin ligase that promotes replication through damaged DNA. EMBO Rep. 9:1034–1040. doi:10.1038/embor.2008.155. Zeng, M., L. Ren, K. Mizuno, K. Nestoras, H. Wang, Z. Tang, L. Guo, D. Kong, Q. Hu, Q. He, L. Du, A.M. Carr,

1. General introduction

and C. Liu. 2016. CRL4Wdr70 regulates H2B monoubiquitination and facilitates Exo1-dependent resection. Nat. Commun. 7:11364. doi:10.1038/ncomms11364. Zhang, J., D. Shi, X. Li, L. Ding, J. Tang, C. Liu, K. Shirahige, Q. Cao, and H. Lou. 2017. Rtt101-Mms1-Mms22 coordinates replication-coupled sister chromatid cohesion and nucleosome assembly. EMBO Rep. e201643807. doi:10.15252/embr.201643807. Zhang, S., L. Chang, C. Alfieri, Z. Zhang, J. Yang, S. Maslen, M. Skehel, and D. Barford. 2016. Molecular mechanism of APC/C activation by mitotic phosphorylation. Nature. 533:260–264. doi:10.1038/nature17973. Zhang, S., H. Zhao, Z. Darzynkiewicz, P. Zhou, Z. Zhang, E.Y.C. Lee, and M.Y.W.T. Lee. 2013. A novel function of CRL4(Cdt2): regulation of the subunit structure of DNA polymerase δ in response to DNA damage and during the S phase. J. Biol. Chem. 288:29550–29561. doi:10.1074/jbc.M113.490466. Zhang, Y., and P.J. Galardy. 2016. Ubiquitin, the centrosome, and chromosome segregation. Chromosome Res. 24:77–91. doi:10.1007/s10577-015-9511-7. Zhang, Y., G. Morrone, J. Zhang, X. Chen, X. Lu, L. Ma, M. Moore, and P. Zhou. 2003. CUL-4A stimulates ubiquitylation and degradation of the HOXA9 homeodomain protein. EMBO J. 22:6057–67. doi:10.1093/emboj/cdg577. Zhao, W., B. Jiang, H. Hu, S. Zhang, S. Lv, J. Yuan, Y. Qian, Y. Zou, X. Li, H. Jiang, F. Liu, C. Shao, and Y. Gong. 2015. Lack of CUL4B leads to increased abundance of GFAP-positive cells that is mediated by PTGDS in mouse brain. Hum. Mol. Genet. 24:4686–4697. doi:10.1093/hmg/ddv200. Zhou, Z., M. He, A.A. Shah, and Y. Wan. 2016. Insights into APC/C: from cellular function to diseases and therapeutics. Cell Div. 11:9. doi:10.1186/s13008-016-0021-6. Zitouni, S., C. Nabais, S.C. Jana, A. Guerrero, and M. Bettencourt-Dias. 2014. Polo-like kinases: structural variations lead to multiple functions. Nat. Rev. Mol. Cell Biol. 15:433–452. doi:10.1038/nrm3819. Zou, Y., Q. Liu, B. Chen, X. Zhang, C. Guo, H. Zhou, J. Li, G. Gao, Y. Guo, C. Yan, J. Wei, C. Shao, and Y. Gong. 2007. Mutation in CUL4B, which encodes a member of cullin-RING ubiquitin ligase complex, causes X-linked mental retardation. Am. J. Hum. Genet. 80:561–6. doi:10.1086/512489. Zou, Y., J. Mi, J. Cui, D. Lu, X. Zhang, C. Guo, G. Gao, Q. Liu, B. Chen, C. Shao, and Y. Gong. 2009. Characterization of nuclear localization signal in the N terminus of CUL4B and its essential role in cyclin E degradation and cell cycle progression. J. Biol. Chem. 284:33320–33332. doi:10.1074/jbc.M109.050427.

2. Article: CRL4RBBP7 is required for efficient CENP-A deposition at centromeres

Research article: Journal of Cell Science, 2015

Full citation Mouysset, J., S. Gilberto, M.G. Meier, F. Lampert, M. Belwal, P. Meraldi, and M. Peter. 2015. CRL4RBBP7 is required for efficient CENP-A deposition at centromeres. J. Cell Sci. 7:1732–1745. doi:10.1242/jcs.162305

Contribution (co-first author) o Writing the manuscript and assembling the figures together with JM and MP o Planning and executing the following experiments: - Optimization and execution of co-immunoprecipitation assays to evaluate the interaction between the endogenous proteins (Figure 3) - Determination of the effects of CUL4RBBP7 depletion in centromeric CENP-A maintenance (Figure 6) - Evaluation of the effect of the different mRNA depletions in cell cycle progression by different methods (Figure S4)

Additional information The supplementary figures (S1 to S4) are included after the references section within this chapter. The following additional information can be found in the Appendix A: A.1 - Optimization of immunoprecipitation assays

2015. Published by The Company of Biologists Ltd | Journal of Cell Science (2015) 128, 1732–1745 doi:10.1242/jcs.162305

RESEARCH ARTICLE CRL4RBBP7 is required for efficient CENP-A deposition at centromeres

Julien Mouysset1,*, Samuel Gilberto1,*, Michelle G. Meier1,2, Fabienne Lampert1, Mukta Belwal1, Patrick Meraldi1,2 and Matthias Peter1,`

ABSTRACT (Westhorpe and Straight, 2013). Loss of CENP-A results in errors during chromosome congression and segregation, which are a The mitotic spindle drives chromosome movement during mitosis source of genomic instability (Re´gnier et al., 2005). Beside its and attaches to chromosomes at dedicated genomic loci named structural function, CENP-A is strictly and constitutively centromeres. Centromeres are epigenetically specified by their associated with the centromere, thus serving as an epigenetic histone composition, namely the presence of the histone H3 variant marker of these loci (Allshire and Karpen, 2008; Fachinetti et al., CENP-A, which is regulated during the cell cycle by its dynamic 2013; Jansen et al., 2007; Schuh et al., 2007). expression and localization. Here, we combined biochemical CENP-A protein levels and localization are stringently cell- methods and quantitative imaging approaches to investigate a new cycle-regulated by mechanisms that are uncoupled from DNA function of CUL4-RING E3 ubiquitin ligases (CRL4) in regulating replication (Jansen et al., 2007; Schuh et al., 2007), during which CENP-A dynamics. We found that the core components CUL4 and time centromeric CENP-A is diluted twofold (Hemmerich et al., DDB1 are required for centromeric loading of CENP-A, but do not 2008; Jansen et al., 2007; Mellone et al., 2011). CENP-A influence CENP-A maintenance or pre-nucleosomal CENP-A levels. expression peaks during the G2 phase and its protein level is Interestingly, we identified RBBP7 as a substrate-specific CRL4 tightly regulated, possibly to avoid non-centromeric incorporation adaptor required for this process, in addition to its role in binding and (Fachinetti et al., 2013; Lacoste et al., 2014). stabilizing soluble CENP-A. Our data thus suggest that the CRL4 The CENP-A-specific histone chaperone Holliday junction complex containing RBBP7 (CRL4RBBP7) might regulate mitosis by recognition protein (HJURP) is expressed concomitantly with promoting ubiquitin-dependent loading of newly synthesized CENP- CENP-A and is found associated with soluble pre-nucleosomal A during the G1 phase of the cell cycle. CENP-A (Bodor et al., 2013; Dunleavy et al., 2009; Foltz et al., 2009; Shuaib et al., 2010). HJURP is known to function in pre- KEY WORDS: Centromere, Cullin, CENP-A, RBBP7 nucleosomal CENP-A stabilization and in centromere deposition, by a mechanism that requires its dimerization (Zasadzin´ska et al., INTRODUCTION 2013), its binding to Mis18b (also known as OIP5) (Wang et al., During mitosis, replicated genetic material is faithfully and 2014) and its DNA-binding activities (Mu¨ller et al., 2014). equally segregated between the two daughter cells. In order to HJURP specifically recognizes CENP-A by interacting with the achieve chromosomal segregation, microtubules emanating from centromere-targeting domain (CATD; Black et al., 2007) of the spindle are bound to chromosomes at specific genomic loci CENP-A to protect it from proteolysis (Foltz et al., 2009). HJURP called centromeres (Westhorpe and Straight, 2013). Spindle is also responsible for the replenishment of the CENP-A pool dynamics then drive chromosome movement and separation (Dunleavy et al., 2011; Foltz et al., 2009; Kato et al., 2007) and (Walczak et al., 2010). for targeting the histone variant to centromeres during G1 (Jansen Except for in budding yeast, eukaryotic centromeres are not et al., 2007; Lagana et al., 2010; Moree et al., 2011). defined by a particular DNA sequence, but by an array of specific In the early stage of CENP-A deposition, CENP-C, which nucleosomes containing the histone H3 variant centromere connects centromeres to kinetochores, recruits Mis18BP1 and protein-A (CENP-A, also known as CenH3) that replaces the HJURP to centromeres and localizes itself at these genomic loci in canonical histone H3.1 variant (Ekwall, 2007). CENP-A a CENP-A-dependent manner (Barnhart et al., 2011; Carroll et al., nucleosomes are interspersed with H3.1-containing nucleosomes 2010; Dambacher et al., 2012; Moree et al., 2011). Subsequently, at the inner region of centromeres (Blower et al., 2002; Ribeiro in a positive-feedback loop, Mis18BP1 recruits additional CENP-A et al., 2010; Sullivan and Karpen, 2004). CENP-A is required for to centromeres (Barnhart et al., 2011). CDK1 inactivation in late the assembly of a large multi-protein complex, the kinetochore, mitosis and during G1 allows HJURP to associate with the which forms the interface between chromatin and the spindle centromere and to incorporate CENP-A (Mu¨ller et al., 2014; Silva et al., 2012). In early G1, CENP-A incorporation is facilitated by the remodelling and spacing factor complex, which associates with 1Department of Biology, Institute of Biochemistry, Swiss Federal Institute of centromeres and is required for CENP-A retention (Perpelescu Technology, 8093 Zurich, Switzerland. 2Department of Physiology and Metabolism, Faculty of Medicine, University of Geneva, 1211 Geneva, et al., 2009). In late G1, the small GTPase MgcRacGAP associates Switzerland. with centromeres and, through a GTP switch, stabilises freshly *These authors contributed equally to this work incorporated CENP-A (Lagana et al., 2010). `Author for correspondence ([email protected]) Additional factors seem important for chaperoning CENP-A. Retinoblastoma-binding protein 7 (RBBP7) and retinoblastoma- Received 24 August 2014; Accepted 13 March 2015 binding protein 4 (RBBP4) (also known as RbAp46 and RbAp48,

1732 RESEARCH ARTICLE Journal of Cell Science (2015) 128, 1732–1745 doi:10.1242/jcs.162305 respectively) are histone chaperones that are part of different mitotic cells as visualized by staining for phosphorylated histone chromatin modifier complexes such as the nucleosome remodelling H3. In contrast, we did not observe such mitotic defects in and deacetylase complex, the Sin3-histone deacetylase complex, RBBP4-downregulated cells (denoted siRBBP4) (supplementary the polycomb repressive complex 2 or the nucleosome remodelling material Fig. S1), implying that RBBP7 is specifically required factor complex (Loyola and Almouzni, 2004). RBBP7 and RBBP4 for timely progression through mitosis. interact with the N-terminal part of histone H4 (Murzina et al., RBBP7 and RBBP4 are WD40-containing proteins that share 2008; Saade et al., 2009). It has been shown in fission yeast that the almost 90% protein sequence identity, and are considered to single RBBP7 and RBBP4 ortholog, Mis16, localizes to centromeres represent the human orthologs of fission yeast Mis16, which is and is essential for CENP-A/Cnp1 centromeric localization and essential for CENP-A protein stability and centromeric association influences the acetylation status of centromeric regions (Hayashi (Hayashi et al., 2004). We thus examined CENP-A protein levels in et al., 2004). Experiments in human cells have revealed that RBBP7 whole-cell extracts prepared from asynchronous HeLa cells after and RBBP4 are required for CENP-A pre-nucleosomal stability downregulation of RBBP4 and RBBP7 by siRNA (Fig. 1B). (Dunleavy et al., 2009), therefore their depletion leads to a reduction Indeed, CENP-A levels were reduced in HeLa cells lacking RBBP7 in levels of CENP-A at centromeres (Hayashi et al., 2004). and both RBBP7 and RBBP4, similar to cells downregulated Excess soluble CENP-A that is not associated with HJURP is for the CENP-A-specific histone chaperone HJURP (Dunleavy degraded by the ubiquitin-proteasome system (UPS). The UPS et al., 2009). Because CENP-A levels fluctuate through the cell consists of the sequential action of multiple enzymes that cycle (Stellfox et al., 2013), we performed a similar experiment in ultimately catalyse the transfer and covalent attachment of cells synchronized at the G1/S transition by double thymidine block ubiquitin moieties to substrate proteins, terminating with the release (DTBR) treatment. In contrast to siRBBP4, siRBBP7 action of an E3 ubiquitin ligase (Pickart, 2004). In budding yeast, treatment was sufficient to reduce CENP-A and HJURP protein Psh1 has been identified as an E3 ubiquitin ligase targeting CENP- levels (Fig. 1C,D), implying that RBBP7 stabilizes CENP-A A/Cse4 for degradation by competing with HJURP for interaction through a cell-cycle-independent mechanism, possibly by with the CATD domain of CENP-A/Cse4 (Hewawasam et al., regulating HJURP levels. 2010; Ranjitkar et al., 2010). In fruit flies, a Skp1-cullin-F-box To address whether both cytosolic and chromatin-bound CENP- (SCF) E3 ubiquitin ligase containing Ppa (SCFPpa) has been A pools were destabilized following RBBP7 depletion, whole-cell suggested to regulate CENP-A/CID levels by a similar mechanism extracts were separated into cytosolic and chromatin fractions involving CATD recognition (Moreno-Moreno et al., 2011). SCF (Fig. 1E). As expected, CENP-A was enriched in chromatin ubiquitin ligases belong to the large cullin–RING ligases (CRL) fractions (siControl lanes). RBBP7 and RBBP4 were found in the family, in which distinct cullin subunits function as scaffolding cytosol and associated with chromatin, which is characteristic for elements to assemble multi-protein complexes (Hotton and Callis, components of chromatin modification complexes and histone 2008). The core CUL4-based complex (CRL4) is CUL4–DDB1– chaperones (Loyola and Almouzni, 2004). Unfortunately, we could RBX1 (with either CUL4A or CUL4B), where DDB1 bridges the not address the role of RBBP4 in CENP-A chromatin association, association between CUL4A or CUL4B and substrate receptors. as we failed to efficiently downregulate chromatin-bound RBBP4. Substrate receptors for CRL4 complexes usually contain specific We therefore observed that siRBBP7 decreased CENP-A levels in WD40 repeats and have been termed DDB1- and CUL4-associated both the cytosol and chromatin fractions (Fig. 1E), whereas factors (DCAF) (Lee and Zhou, 2007). CRL4 complexes are siRBBP4 reduced cytosolic CENP-A. We conclude that RBBP7 involved in different cellular processes, including DNA repair, is required for CENP-A stabilization and chromatin association, DNA replication and chromatin remodelling (O’Connell and possibly in cooperation with RBBP4. As expected, CENP-A levels Harper, 2007). were also decreased in whole-cell extracts prepared from G1/S Here, we propose a new role for CRL4, together with RBBP7 as synchronized cells downregulated for HJURP (Fig. 1F) (Dunleavy a substrate specific receptor, in CENP-A centromeric deposition. et al., 2009; Foltz et al., 2009). It is noteworthy that chromatin- Using an automated quantitative image analysis pipeline and associated HJURP seems to be mildly affected after 72 h of siRNA biochemical approaches, we found that CUL4–DDB1 and RBBP7 treatment. Based on these cell fractionation experiments, we form a stable complex (denoted CRL4RBBP7) that is required for the conclude that RBBP7, RBBP4 and HJURP are required for deposition of CENP-A at centromeres. Moreover, RBBP7, but not stabilizing soluble CENP-A, and RBBP7 and HJURP also affect the CRL4 complex, was necessary to protect the pre-nucleosomal efficient CENP-A chromatin deposition. However, because these fraction of CENP-A from proteolysis. Taken together, our data CENP-A pools are linked, we cannot exclude that the latter suggest that RBBP7 not only stabilizes soluble CENP-A, but, in observation might be an indirect consequence of the reduced complex with CRL4, also promotes loading of newly synthesized cytoplasmic CENP-A levels. CENP-A at centromeres. Association of CENP-A with centromeres requires RBBP7 RESULTS and HJURP RBBP7 is required to stabilize CENP-A protein levels As CENP-A chromatin association is restricted to centromeric loci To investigate the functional importance of CUL4 substrate (Allshire and Karpen, 2008), we next developed a microscopy- receptors for mitosis, the duration of distinct mitotic stages was based experimental pipeline that allowed automated analysis of quantified by automated time-lapse imaging of HeLa cells thousands of centromeres (Fig. 2A,B). Briefly, frames from fixed expressing the chromatin marker histone-H2B–mCherry in cells stained with appropriate antibodies were analysed using which DCAFs were targeted by small interfering RNA (siRNA) CellProfiler software (Carpenter et al., 2006), which allows the (Held et al., 2010; Piwko et al., 2010). Interestingly, this analysis segmentation of cell nuclei and the quantification of pixel intensity revealed that downregulation of RBBP7 (denoted siRBBP7) values from centromeres with a high level of confidence (Fig. 2B; causes a prolonged metaphase-to-anaphase transition (Fig. 1A; supplementary material Fig. S2A–D). To assess the cell cycle stage supplementary material Fig. S1), resulting in an enrichment of of individual cells, we used antibodies against cyclin A as a marker

1733 RESEARCH ARTICLE Journal of Cell Science (2015) 128, 1732–1745 doi:10.1242/jcs.162305

Fig. 1. RBBP7 regulates mitosis and is required to stabilize CENP-A protein levels. (A) Representative stills from time-lapse microscopy taken from a HeLa cell line expressing the chromatin marker histone-H2B–mCherry after treatment with the indicated siRNA. Time 0 min corresponds to the first still image displaying a metaphase plate morphology. Qiagen AllStar negative control (siControl_AS) is included, as well as controls for extended (siCDC20) and reduced (siMAD2L1) metaphase timing. (B–D) Western blot analyses from whole-cell extracts (WCE) from HeLa cells treated with the indicated siRNA that are in asynchronous culture (asyn.) (B) or had been synchronized at the G1/S transition by DTBR (C,D). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as a loading control. (E,F) Western blot analyses from cell fractionation experiments of HeLa cells synchronized at the G1/S transition by DTBR treatment. GAPDH and a non-specific anti-GAPDH band were used as loading controls for the cytoplasmic and the chromatin fractions, respectively. siRBBP7/4 indicates treatment with siRNA against both RBBP7 and RBBP4. of the G1/S transition (Fig. 2B). To validate this in silico maximum pixel intensities, centromere coordinates, nuclear cyclin synchronization method, we compared single-cell cyclin A A pixel intensity, nucleus coordinates and nucleus–centromere values from asynchronous and synchronized cell populations, affiliations. The efficiency of centromere detection was compared and observed that the mean nuclei pixel intensity of cyclin A in by quantifying the same image dataset with our automated pipeline DTBR-synchronized cells changed the distribution of the measured and with a semi-automated quantification protocol using Imaris values from a bimodal to a Gaussian distribution (supplementary software (Bitplane). The overall number of detected centromeres material Fig. S2A,B). Based on the cyclin A staining, the total as well as the maximum pixel intensity distribution between number of G1/S cells in an asynchronous population was ,14% the two methods was similar with 3369 spots detected with Imaris (supplementary material Fig. S2A). To facilitate the analysis, we software versus 3473 spots detected with our analysis pipeline developed a library of in-house MATLAB scripts to extract various (supplementary material Fig. S2C). Moreover, manual validation parameters from the CellProfiler data files including CENP-A of centromere detection revealed that the analysis pipeline

1734 RESEARCH ARTICLE Journal of Cell Science (2015) 128, 1732–1745 doi:10.1242/jcs.162305

Fig. 2. RBBP7 is required for CENP-A centromeric association. (A) Schematic of the automated imaging pipeline workflow to quantify centromeric fluorescence. (B) Image analysis showing representative examples of CellProfiler image outputs. An example of each channel with use description is shown. (C) Timeline for cell handling prior to imaging to quantify endogenous CENP-A. (D) Quantification of centromeric CENP-A maximum pixel levels in DTBR- synchronized HeLa cells treated with the indicated siRNA normalized to those in siControl. In silico synchronization ensures that only cells at the G1/S transition were quantified, thus excluding cells that arrested in the cell cycle due to siRNA treatment or experimental conditions. ‘N’ represents the number of replicates, and ‘n’ the number of measured centromeres. Results are mean6s.d. P-values from a Wilcoxon rank sum test are indicated for each condition. siRBBP7/4 indicates treatment with siRNA against both RBBP7 and RBBP4. (E) Representative images for centromeric CENP-A protein levels from the quantitative analysis shown in D. A single nucleus and a representative centromere (inset) is shown per condition.

1735 RESEARCH ARTICLE Journal of Cell Science (2015) 128, 1732–1745 doi:10.1242/jcs.162305 correctly assigned over 92% of the centromeres that were identified performed Strep- or His-pull downs, respectively. Interestingly, by manual counting (supplementary material Fig. S2D). Taken we observed that RBBP7 and RBBP4 were able to co-purify with together, these data demonstrate that the established image-based DDB1, indicating that there is a direct interaction (Fig. 3D,E). centromere detection protocol is reliable and efficient, thus Thus, we conclude that RBBP7 and RBBP4 assemble in vivo and allowing quantification of a large number of centromeric in vitro in a manner analogous to that shown by known substrate- fluorescent signals in an automated and unbiased fashion. specific receptors of CRL4 complexes (Fig. 3F). To examine whether RBBP7, RBBP4 and HJURP are required to regulate CENP-A levels at centromeres, HeLa cells were CRL4 is required for centromeric association but not stability synchronized by DTBR at the G1/S transition (Fig. 2C; of CENP-A supplementary material Fig. S2E), which corresponds to the We next investigated the requirement of CUL4 and DDB1 in cell stage when most of the CENP-A molecules have been loaded CENP-A protein stability and centromeric association. In contrast at centromeres (Jansen et al., 2007), but have not yet been diluted to RBBP7 and RBBP4, total CENP-A protein levels were during DNA replication (Stellfox et al., 2013). The cells were unaffected after simultaneous downregulation of both CUL4A fixed 72 h after siRNA treatment to downregulate RBBP7, and CUL4B using two distinct siRNAs (referred to hereafter as RBBP4 or HJURP (Fig. 2C), and centromeric CENP-A levels siCUL4A/B) or siRNA against DDB1 (denoted siDDB1) either in were quantified using the image-based analysis platform asynchronous cell populations (Fig. 4A) or in cells synchronized described above. For each experiment, DTBR synchronization at the G1/S transition by a DTBR protocol (Fig. 4C). Likewise, and the siRNA-mediated knockdown efficiency were additionally siCUL4A/B and siDDB1 treatment did not alter HJURP levels. monitored in parallel with the immunofluorescence staining Cell fractionation experiments confirmed that, like CENP-A, a (supplementary material Fig. S2E,F). As expected (Dunleavy fraction of CUL4A, CUL4B and DDB1 associated with et al., 2009; Foltz et al., 2009), siRNA against HJURP and CENP- chromatin in cells synchronized at the G1/S transition (Fig. 4B; A treatments significantly reduced centromeric CENP-A levels Obuse et al., 2004), consistent with the notion that CRL4 (Fig. 2D,E). Importantly, we also observed that siRBBP7, but not regulates chromatin-associated processes (O’Connell and Harper, siRBBP4, led to a decrease of centromeric CENP-A (Fig. 2D,E). 2007). Interestingly, CENP-A levels were decreased in the Treatment with siRNA against both RBBP7 and RBBP4 led to an chromatin fraction of cells lacking DDB1 and to a lesser extent even stronger reduction of centromeric CENP-A compared to the also in cells lacking CUL4A/B (Fig. 4C), suggesting that CRL4 single RBBP7 downregulation, consistent with previous results modulates chromatin association and/or maintenance of CENP-A (Hayashi et al., 2004). Taken together, these results confirm our without affecting the stability of soluble CENP-A. cell fractionation experiments, and suggest that RBBP7 and To support these data, we measured centromeric CENP-A RBBP4 cooperate to load and/or maintain CENP-A at protein levels using our quantitative automated imaging-based centromeres. assay (Fig. 2A). Cells were treated as outlined in Fig. 2C, and DTBR synchronization and siRNA knockdown efficiency were RBBP7 and RBBP4 directly interact with DDB1 monitored (supplementary material Fig. S2G,H). Consistently, we Although RBBP7 and RBBP4 lack the characteristic helix-loop- observed that centromeric CENP-A levels were reduced in DDB1- helix motif mediating the interaction with DDB1 (Fischer et al., downregulated cells (67%68, mean6s.d.) and also in CUL4A/B- 2011), they have been previously proposed to function as downregulated cells, although not in a statistically significant substrate receptors in CRL4 E3 ligase complexes (He et al., manner (87%612) (Fig. 4D,E). The reduction of centromeric 2006). To confirm this observation, we immunoprecipitated CENP-A in siDDB1-treated cells was less pronounced compared to endogenous CUL4A from extracts prepared from HeLa cells cells lacking siRBBP7 (30% as opposed to 50%, respectively), using specific anti-CUL4A antibodies. Like DDB1, RBBP7 was possibly because RBBP7, but not DDB1, also plays a role in readily detected in CUL4A immunoprecipitates (Fig. 3A), stabilizing soluble CENP-A. demonstrating that the two proteins interact under physiological conditions. Importantly, binding between CUL4A and RBBP7 CRL4RBBP7 is required to deposit newly synthesized CENP-A was reduced upon DDB1 depletion, supporting the notion that at centromeres DDB1 links the two proteins in vivo. Conversely, N-terminally To examine whether RBBP7 in a complex with CRL4 is specifically HA–26Strep-tagged RBBP7 ectopically overexpressed in HeLa required to load newly synthesized CENP-A molecules onto cells was able to co-precipitate endogenous DDB1 and RBBP4 centromeres during G1, we performed CENP-A-loading (Fig. 3B). Thus, it appears that RBBP7 and RBBP4 stably experiments using the SNAP-tag labelling technology (Bodor associate with the known CRL4 core components, and are likely et al., 2012; Jansen et al., 2007). To this end, we synchronized to function as genuine subunits of the complex. To substantiate SNAP–CENP-A-expressing HeLa cells at the G1/S transition by these findings, we co-expressed RBBP7, DDB1 and RBX1 with DTBR prior to quenching of the SNAP-tag by the addition of either FLAG-tagged CUL4A or FLAG–CUL4B in insect cells. bromothenylpteridine (Fig. 5A). The cells were released to allow Importantly, we obtained a stable heterotetrameric complex after translation of new SNAP–CENP-A protein, and thymidine was specific elution with 36FLAG peptide (Fig. 3C), demonstrating added again during this chase period to block the next G1/S that RBBP7 assembles into stable CRL4 complexes in vitro. transition. Before fixation, cells were fluorescently pulse-labelled These experiments also indicate that RBBP4 and RBBP7 with the SNAP-reactive TMR-Star reagent, thus ensuring that only interact with each other, possibly forming homodimers or SNAP–CENP-A protein produced after quenching is fluorescent heterodimers, as it has been shown for multiple CRL-substrate- (referred to as ‘new CENP-A’). Although the efficiency of the specific receptors. To corroborate these findings and further quenching, chasing and labelling steps were carefully tested and determine whether both RBBP7 and RBBP4 directly interact with optimized (supplementary material Fig. S3A,B), the dynamic range DDB1, we co-expressed recombinant His-tagged DDB1 with of the SNAP signal was too weak to allow a reliable automated either Strep-tagged RBBP7 or RBBP4 in insect cells, and analysis. We thus quantified manually the proportion of cells that

1736 RESEARCH ARTICLE Journal of Cell Science (2015) 128, 1732–1745 doi:10.1242/jcs.162305

Fig. 3. RBBP7 and RBBP4 interact with DDB1 in vivo and in vitro. (A) Immunoprecipitates (IP) of endogenous CUL4A from HeLa cell extracts using a rabbit anti-CUL4A antibody were analysed by western blotting, with or without depletion of DDB1 by siRNA (siDDB1). A non-specific immunoglobulin G (IgG) was also used as a negative control. (B) HeLa cells were transiently transfected with a HA–Strep–Strep–RBBP7 construct. Cell extracts were used for HA pulldown experiments and analysed by western blotting. Control conditions correspond to cells treated with transfection reagents without DNA. (C) DDB1, RBX1and RBBP7 were co-expressed with either FLAG–CUL4A or FLAG–CUL4B in a single multibac vector in Sf-9 insect cells. CUL4 complexes were purified on a FLAG-antibody column, and eluted with 36FLAG peptide. Co-precipitating proteins with CUL4A (left panels) and CUL4B (right panels) were visualized by Ponceau S staining (right lanes) and western blotting for RBBP7 (left lanes). (D,E) High Five insect cells were infected with His–DDB1, Strep–RBBP7 and Strep–RBBP4. Strep (D) or His pulldowns (E) were performed and analysed by western blotting. (F) Model for the CRL4RBBP7–RBBP4 E3 ubiquitin ligase complex. Ub, ubiquitin; N8, Nedd8; E2, E2 ubiquitin-conjugating enzyme. The dotted arrow indicates ubiquitin transfer from the E2 enzyme to the bound protein substrate.

1737 RESEARCH ARTICLE Journal of Cell Science (2015) 128, 1732–1745 doi:10.1242/jcs.162305

Fig. 4. Depletion of DDB1 reduces centromeric CENP-A association. (A,C) Western blot analyses of whole-cell extracts (WCE) prepared from non-synchronized (A) or DTBR- synchronized HeLa cells accumulating at the G1/S transition that had been treated with the indicated siRNA. In C, GAPDH and RPA was used as a loading control for WCE and the chromatin fraction, respectively. The intensity of the bands was quantified using Fiji software and normalized to the loading control. ‘+’ marks the neddylated form of CUL4. (B) HeLa cells treated with siControl were synchronized at the G1/S transition by DTBR. Cells were lysed, separated into cytoplasmic, nuclear (chromatin excluded), and chromatin-bound fractions and analysed by western blotting with the indicated antibodies. ‘+’ marks neddylated CUL4; ‘*’ marks a nonspecific band. (D) Quantification of centromeric CENP-A maximum pixel levels in DTBR- synchronized HeLa cells treated with the indicated siRNA normalized to siControl. In silico synchronization based on cyclin A staining excludes the analysis of cells that arrested in the cell cycle due to siRNA treatment or experimental conditions. ‘N’ represents number of replicates, and ‘n’ the number of centromeres measured. Results are mean6s.d. P-values from a Wilcoxon rank sum test are indicated accordingly for each condition. (E) Representative images for centromeric CENP-A protein levels from the quantitative analysis as in D are shown. A single nucleus and a representative centromere (inset) is shown per condition. siCUL4A/B indicates treatment with siRNA against both CUL4A and CUL4B.

displayed fluorescent centromeric TMR-Star signals. As expected, cells lacking CUL4A/B (supplementary material Fig. S4A,B), we observed that siHJURP led to a complete absence of cells although we observed a mild cell cycle delay in DDB1-depleted positive for centromeric fluorescent signal (Fig. 5B,C) (Foltz et al., cells. These results thus uncover an important role of RBBP7 and 2009). Strikingly, siCUL4A/B, siDDB1 and siRBBP7 drastically the CRL4 core components in centromeric loading of CENP-A, and impaired the loading of new CENP-A at centromeres (Fig. 5B,C), suggest that a CRL4RBBP7-dependent ubiquitylation step is required whereas siRBBP4 had no effect. Cell cycle analysis by flow- for this process. cytometry and immunofluorescence excluded the possibility that CENP-A is a variant belonging to the histone H3 family that is these strong defects were indirectly caused by cell cycle delays in specifically assembled at centromeres (Allshire and Karpen,

1738 RESEARCH ARTICLE Journal of Cell Science (2015) 128, 1732–1745 doi:10.1242/jcs.162305

Fig. 5. CRL4RBBP7 is required to deposit CENP-A at centromeres. (A) Timeline of cell handling prior to imaging of the SNAP-histone loading assay. The SNAP-tag is quenched by the addition of bromothenylpteridine and TMR-Star allows SNAP–CENP-A fluorescence pulse-labelling. (B) Fluorescent centromeric TMR-Star-labelled SNAP-CENP-A nuclei were counted manually in cells depleted for the indicated proteins by siRNA, and normalized to control conditions. ‘N’ represents number of replicates, and ‘n’ is the number of nuclei counted. Results are mean6s.d. (C) Representative images of centromeric fluorescently labelled SNAP–CENP-A protein as in B are shown. A single nucleus and a representative centromere (inset) is shown per condition. (D,E) Representative images from HeLa cell lines expressing H3.1–SNAP (D) and H3.3–SNAP (E) from the quantitative analysis shown in F and G, respectively. Cells were treated as described in A. (F,G) Quantification of chromatin H3.1–SNAP (F) and H3.3–SNAP (G) maximum pixel levels in in silico synchronized HeLa cells. ‘n’ represents number of nuclei measured. Results are mean6s.e.m. P-values from a Wilcoxon rank sum test are indicated accordingly for each condition. siRBBP7/4 indicates treatment with siRNA against both RBBP7 and RBBP4; siCUL4A/B indicates treatment with siRNA against both CUL4A and CUL4B.

1739 RESEARCH ARTICLE Journal of Cell Science (2015) 128, 1732–1745 doi:10.1242/jcs.162305

2008). Histone H3.1 represents the canonical histone H3, and its centromeres during the G1 phase, but is not involved in expression is dependent on DNA replication. Histone H3.3 maintaining CENP-A on centromeres throughout the cell cycle. denotes a non-DNA replication-dependent histone H3, which is incorporated into chromatin throughout the cell cycle at active DISCUSSION transcription promoters, telomeres and centromeres (Biterge and CENP-A centromeric deposition has been described as a multi-step Schneider, 2014; Tagami et al., 2004). We thus tested whether, event that requires a priming phase, in which centromeres recruit like CENP-A, chromatin deposition of histones H3.1 and H3.3 specific factors, a deposition phase and a maintenance phase in also requires CRL4RBBP7 activity. As described for CENP-A, we order to stabilize freshly incorporated histones (Mu¨ller and monitored the pool of ‘new H3.1’ and ‘new H3.3’ incorporated Almouzni, 2014). In this study, we propose a new role for into chromatin after one complete cell cycle of HeLa cells CUL4A/B–DDB1, together with RBBP7, in CENP-A centromeric expressing H3.1–SNAP or H3.3–SNAP transgenes (Ray-Gallet deposition. Biochemical analysis showed that the levels of et al., 2011) synchronized at the G1/S transition (Fig. 5A). The chromatin-bound CENP-A were reduced after downregulation of efficiency of quenching, chasing and pulse labelling was CUL4A/B, DDB1 and RBBP7, and microscopy-based assays optimized for both H3.1–SNAP and H3.3–SNAP cell lines identified defects in loading of CENP-A at centromeric locations. (supplementary material Fig. S3C–F), and the mean nuclear We showed that the CRL4RBBP7 complex has no impact on other TMR-Star fluorescence of both new H3.1 and new H3.3 histones aspects of CENP-A dynamics such as pre-nucleosomal stabilization was quantified with the automated image analysis pipeline for in and centromeric localization maintenance. Importantly, siRNA silico synchronized cells as described in Fig. 2A. Interestingly, against CUL4A/B, DDB1 and RBBP7 led to prolonged mitotic the mean nuclear levels of new H3.1 cells were unaffected in cells progression, reminiscent of the CENP-A homozygous knockout with downregulated CUL4A/B and DDB1 (Fig. 5D,F), whereas phenotype (Fachinetti et al., 2013). Our work thus suggests a new the loading of new H3.3 was unaffected by siCUL4A/B, but role for CRL4-based ubiquitin ligases in regulating centromeric possibly slightly reduced after siDDB1 treatment (Fig. 5E,G). As loading of CENP-A during G1. expected, downregulation of HJURP had no effect on loading of the H3.1 and H3.3 variants, demonstrating the specificity of the Automated image analysis quantifies levels and dynamics of assay. Based on these results, we conclude that CRL4RBBP7 CENP-A at individual centromeres activity is required for efficient loading of CENP-A at Automated processing of imaging data allows the analysis of large centromeres, and that this function does not involve a general datasets, thereby facilitating extraction of multiple features and histone deposition function but rather a CENP-A-specific enabling consistent unbiased processing of various experimental mechanism. conditions. In order to analyse the centromeric fluorescent signal, we developed a pipeline for in-depth automated image analysis by combining computer-assisted analysis with CellProfiler and CRL4RBBP7 is not required for maintaining loaded CENP-A MATLAB in-house scripts. We were able to analyse a large at centromeres number of centromeres per nuclei on a single-cell basis, which was CENP-A remains constitutively associated with centromeres not possible with previous methods (Bodor et al., 2012). Validation throughout the cell cycle, and is only diluted by DNA revealed that the analysis pipeline detected over 92% of the replication and subsequently replenished after mitosis (Jansen centromeres, and allowed a quantification of the levels and et al., 2007). To test whether CRL4RBBP7 is not only required to distribution of proteins on single centromeres in an unbiased load but also to maintain CENP-A at centromeres, we followed manner. Combined with appropriate cell cycle markers, this the fate of centromeric CENP-A in a maintenance assay. To this method not only permitted us to determine average numbers but end, synchronized HeLa cells expressing SNAP–CENP-A also cell-to-cell and centromere-to-centromere variations. The (Jansen et al., 2007) were fluorescently pulse labelled with image-based method will thus be of broad interest, and could also TMR-Star (Fig. 6A), thereby specifically labelling SNAP–CENP- be adapted to analyse centromere dynamics in more complex A that is expressed before the pulse labelling (‘old CENP-A’). samples such as 3D cultures or tissues. Finally, the developed The dilution of ‘old CENP-A’ at centromeres was then quantified image-processing pipeline could be applied to study other after different chasing periods ranging from 0 to 3 days using our biological processes, including quantifying aneuploidy or the automated imaging pipeline (Fig. 6A). With control siRNA, a detection and quantification of other types of foci such as DNA progressive decrease of the old CENP-A centromeric signal was damage markers (data not shown). observed over the course of the chase period (Fig. 6B,C), with kinetics comparable to other published reports (Bodor et al., RBBP7 and RBBP4 are required to stabilize CENP-A and the 2013; Fachinetti et al., 2013). Interestingly siRNA-mediated histone chaperone HJURP in vivo downregulation of CENP-A, CUL4A/B, DDB1 and RBBP7 RBBP7 and RBBP4 are components of multiple chromatin resulted in decay kinetics similar to those with control siRNA modification complexes, and interact with histone target (Fig. 6B,C), suggesting that CRL4RBBP7 is not required for substrates (Loyola and Almouzni, 2004). In contrast to their CENP-A maintenance at centromeres. To monitor proper cell counterpart in fission yeast, human RBBP7 and RBBP4 proteins cycle progression under the experimental conditions, we pulse are not enriched at centromeric regions, but are required for the labelled S-phase cells with BrdU 16 h before harvesting stability and centromeric localization of CENP-A (Hayashi et al., (supplementary material Fig. S4C). The appearance of BrdU- 2004). It was shown previously that RBBP4 is associated with labelled cells in G1-phase demonstrates that all cells underwent pre-nucleosomal CENP-A, and that RBBP7 and RBBP4 are mitosis, and the expected cell cycle delay was only observed for required to stabilize the soluble fraction of the CENP-A-specific DDB1-depleted cells after a 72-h siRNA treatment. Taken histone chaperone HJURP (Dunleavy et al., 2009). Our data together, these results suggest that the CRL4RBBP7 E3 ubiquitin confirm that RBBP7 and RBBP4 stabilize HJURP and pre- ligase complex is required to specifically load CENP-A on nucleosomal CENP-A, supporting a model in which RBBP7 and

1740 RESEARCH ARTICLE Journal of Cell Science (2015) 128, 1732–1745 doi:10.1242/jcs.162305

Fig. 6. CRL4RBBP7 is not required for centromeric CENP-A maintenance. (A) Timeline of cell handling prior to imaging of the SNAP–CENP-A maintenance assay. (B) Representative images of cells treated with siControl oligonucleotides show centromeric fluorescently labelled SNAP–CENP-A protein (upper row) after the indicated chase period (days). Staining with CREST antibodies was included as a control (lower row). (C) Quantification of centromeric SNAP- CENP-A maximum pixel levels normalized to siControl conditions in in silico synchronized HeLa cells treated with the indicated siRNA. ‘n’ represents the number of centromeres measured, and results are median6s.d. siCUL4A/B indicates treatment with siRNA against both CUL4A and CUL4B. (D) Model of the timing of HJURP, RBBP4 and RBBP7, and CRL4RBBP7–RBBP4 activity in CENP-A dynamics. Our data suggest that CRL4RBBP7–RBBP4-dependent ubiquitylation promotes CENP-A loading at the G1/S transition.

1741 RESEARCH ARTICLE Journal of Cell Science (2015) 128, 1732–1745 doi:10.1242/jcs.162305

RBBP4 together with HJURP protect soluble CENP-A from promoting CENP-A deposition. However, biochemical in vitro proteolysis. reconstitution assays will be necessary to test this exciting HJURP dimers recognize and stabilize the heterodimeric CENP- hypothesis. A–histone-H4 pre-nucleosomal complex by interacting with the CATD domain of CENP-A (Bassett et al., 2012; Foltz et al., 2009; CRL4 is not required to regulate deposition of other histone Zasadzin´ska et al., 2013). Our data suggest that pre-nucleosomal H3 variants in vivo CENP-A stabilization requires both RBBP7 and RBBP4, but that CENP-A is a variant of histone H3 that marks centromeric their functions are not redundant. RBBP4 directly interacts with chromatin (Ekwall, 2007). Additional histone H3 variants have histone H4 (Murzina et al., 2008; Saade et al., 2009) and been described that as well as performing functions at colocalizes with pre-nucleosomal CENP-A (Dunleavy et al., centromeres also do so at other heterochromatic regions with 2009), suggesting that the pre-nucleosomal CENP-A complex is unrelated functions (Biterge and Schneider, 2014). In contrast to composed of a HJURP homodimer and RBBP7–RBBP4 and CENP-A, by using quench-chase-pulse SNAP-tagged labelling CENP-A–histone-H4 heterodimers (Fig. 6D). Recently, it has been we could not observe a defect in loading of the histone H3 shown that in fruit flies the CUL3-based complex CRL3RDX mono- variants H3.1 and H3.3 in cells downregulated for CUL4A/B, ubiquitylates CENP-A/CID, which stabilizes CENP-A/CID and DDB1 and RBBP7, implying that CRL4RBBP7 is specifically the HJURP ortholog CAL1 (Bade et al., 2014). Whether required for CENP-A deposition. Moreover, CENP-A, H3.1 and mammalian CRL3 similarly regulates the stability of CENP-A H3.3 maintenance at centromeres was independent of CUL4, and/or HJURP remains to be determined. DDB1 and RBBP7. Interestingly, Rtt101, the functional budding yeast CUL4 CRL4 is required for efficient loading of new CENP-A to ortholog (Zaidi et al., 2008), has recently been shown to be centromeres in vivo involved in histone dynamics by a conserved mechanism that In contrast to RBBP7 and RBBP4 downregulation, siCUL4A/B involves CUL4A- and DDB1-dependent turnover of H3.1 and and siDDB1 do not lead to decreased levels of soluble CENP-A or H3.3 (Han et al., 2013). Several aspects could explain the HJURP. Instead, CUL4A/B and DDB1 appear to be specifically differences observed. First, we have experimentally synchronized required for centromeric recruitment of CENP-A, without altering the cells and monitored H3.1 and H3.3 dynamics at the same cell global CENP-A levels (Fig. 6D). Indeed, CUL4A and DDB1 are cycle stage. Second, we implemented a longer chase period to both present at centromeres during interphase (Obuse et al., 2004), compare cells with a similar history regarding their cell cycle and fractionation experiments revealed that the neddylated (active) progression. Finally, we downregulated both CUL4A and forms of CUL4A and CUL4B are both enriched in the chromatin CUL4B, not only CUL4A, to avoid compensatory mechanisms fraction at the G1/S transition (Fig. 4B), a point when pre- from these two very close paralogs. Taken together, our results nucleosomal CENP-A is loaded on centromeres. Available suggest that CRL4RBBP7 is primarily required for CENP-A evidence suggests that RBBP7 and RBBP4 function as substrate- deposition, and is not generally involved in histone dynamics. specific adaptors in CRL4 complexes to load CENP-A. RBBP7 co- immunoprecipitates in vivo with CUL4A in a DDB1-dependent What are the physiological CRL4RBBP7 substrates for CENP- manner (Fig. 3A,B; He et al., 2006), and both RBBP7 and RBBP4 A deposition? directly bind DDB1 in vitro in heterologous expression systems Our results indicate that CUL4 and DDB1 are required for CENP- (Fig. 3C–E). Although a sequence motif search identified RBBP7 A deposition at centromeres, most likely by forming an E3 ligase and RBBP4 as DDB1–CUL4A interactors possessing a DDB1- complex together with RBBP7. This raises many questions for binding WD40 protein box, RBBP7 and RBBP4 lack the helix- future study. Clearly, it would be of prime interest to identify the loop-helix motif characteristic for established DCAFs, which crucial CRL4RBBP7 substrates. Recent work in Drosophila has mediates the interaction with DDB1 (Fischer et al., 2011). RBBP7 identified CENP-A as a potential candidate (Bade et al., 2014), and Nevertheless, we were able to reconstitute stable CRL4 it is possible that ubiquitylation of newly synthesized CENP-A is complexes in vitro, and it will be interesting to analyse them required for its centromere deposition. Another line of questioning structurally and measure their ubiquitylation activity on potential derives from the observation that centromeric regions are enriched substrates. in histone methylation and display a low degree of acetylation, Overall, our results suggest that RBBP7 and probably also which appears to influence the structural organization of chromatin RBBP4 act at two distinct stages in CENP-A dynamics: (1) during (Bergmann et al., 2011; Kim et al., 2012; Ohzeki et al., 2012). stabilization of pre-nucleosomal CENP-A and HJURP, and (2) RBBP7 Interestingly, the ortholog of RBBP7 and RBBP4 in fission yeast, during formation of a CRL4 complex at centromeres to Mis16, is necessary for maintaining the deacetylated status of deposit CENP-A at chromatin (Fig. 6D). Although total CENP-A centromeres (Hayashi et al., 2004). Therefore, the CRL4RBBP7 levels were not altered in cells lacking any CRL4 core subunits, we RBBP7 complex might regulate the acetylation status of centromeric cannot exclude that CRL4 affects CENP-A stability proteins, thereby promoting recruitment of the pre-nucleosomal specifically during the deposition process. We could not fully CENP-A complex or other factors important for chromatin investigate the involvement of RBBP4 for CENP-A loading due to structure of centromeres. Because ubiquitylated substrates at its inefficient downregulation at chromatin. Nonetheless, given that least transiently interact with specific DCAFs, the search for RBBP4 also directly interacts with DDB1, it is likely that RBBP4 RBBP7–RBBP4 proteins interacting with RBBP7 and RBBP4 may be a fruitful CRL4 and CRL4 complexes similarly promote focus for future research. CENP-A deposition on centromeres (Fig. 6D). Given that CRLs might dimerize to regulate their activity (Merlet et al., 2009), it is MATERIALS AND METHODS tempting to speculate that CRL4 recognizes the pre-nucleosomal Cell culture, cell transfection, siRNAs and cell synchronization CENP-A complex containing RBBP7 and RBBP4 and that HeLa cell lines were grown in a humidified incubator at 37˚C and 5% CO2 in CRL4RBBP7–CRL4RBBP4 heterodimers might be responsible for Dulbecco’s modified Eagle medium (DMEM, Gibco, Invitrogen, Carlsbad,

1742 RESEARCH ARTICLE Journal of Cell Science (2015) 128, 1732–1745 doi:10.1242/jcs.162305

CA) supplemented with 10% fetal calf serum (FCS, PAA Laboratories) and from the median pixel value of the shrunken segmented nuclear area. 1% penicillin-streptomycin (Gibco). The stable SNAP–CENP-A-expressing Centromere detection was performed using the A-Trous Wavelet cell line was produced by transfecting HeLa E1 cells with a pSEMS1-SV40- transform algorithm applied to the CREST channel. Each centromere CENP-A plasmid using FUGENE6 reagents (Roche, Basel, Switzerland) detected was affiliated to the segmented nucleus according to their according to the manufacturer’s instructions. 5 mM puromycin (Gibco) and coordinates. We developed a library of in-house MATLAB scripts 2 mM blasticidin S (PAA Laboratories, GE Healthcare, Little Chalfont, UK) (version R2011b, MathWorks) that allow us to extract in an automated were used as selection antibiotics for HeLa SNAP–CENP-A and SNAP– fashion values from the CellProfiler MATLAB files. The centromeric H3.1 or SNAP–H3.3 cell lines (Ray-Gallet et al., 2011), respectively. HeLa fluorescent signal was determined by subtracting the median pixel value ‘Kyoto’ cells stably expressing H2B–mCherry and IBB–mEGFP (IBB is from the edge of a 7-pixel wide circle centred on the detected centromere, the importin-b-binding domain of importin-a) (Held et al., 2010) were corresponding to the local background signal, from the maximum pixel grown in 0.5 mg/ml geneticin (Calbiochem, Merck, Darmstadt, Germany) value of a 3-pixel wide circle centred on the same detected centromere. and 0.5 mg/ml puromycin (Gibco). DNA plasmid was transiently transfected For the CENP-A maintenance assay, the brightest 30% of centromeres using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s were used for quantification. The CENP-A loading assay results were instructions. For RNAi experiments, cells were transfected with 50 nM assessed by manually counting by visual inspection from three siRNAs (Qiagen, Hilden, Germany or Microsynth, Balgach, Switzerland) independent experiments. for 72 h using Oligofectamine (Invitrogen) or Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer’s instructions. The target Cell fractionation, western blot and antibodies sequences of double-stranded RNA used in this study were as follows: For cell fractionation, cells were harvested by trypsinization and washed siControl, 59-GGACCTGGAGGTCTGCTGT-39 (McHedlishvili et al., with cold PBS. Cytosol and nuclei were extracted using the NE-PER 2012); siCENP-A, 59-AACACAGTCGGCGGAGACAAG-39 (Carroll extraction kit (Pierce, Thermo Fisher, Rockford, USA). Nucleoplasm and et al., 2009); siHJURP, 59-CTACTGGGCTCAACTGCAA-39 (Dunleavy chromatin were subsequently extracted by sucrose cushion centrifugation et al., 2009); siCUL4A, 59-AAGAATCCTACTGCTGATCGA-39 (Piwko after lysis of the nuclei in 100 mM KCl, 10 mM Hepes (pH 7.7), 50 mM et al., 2010); siCUL4B, 59-CACCGTCTCTAGCTTTGCTAA-39 sucrose, 0.25% Triton X-100 and with addition of various inhibitors (Piwko et al., 2010); siDDB1, 59-CCACTAGATCGCGATAATAAA-39 (inhibitor cocktail, Roche; NaF; b-glycerophosphate; Leupeptin; (Piwko et al., 2010); siRBBP7, 59-GCGGATAAGACCGTAGCTTTA-39 Pepstatin; PMSF). Cell lysate preparations were applied to SDS-PAGE (Piwko et al., 2010); siRBBP4, 59-GCCACTCAGTTGATGCTCA-39 gels and proteins were transferred onto PVDF membranes. (Hayashi et al., 2004); siCDC20, 59-AACCTTGTGGATTGGAGTTCT- Immunoreactive bands were visualized with enhanced 39 (Piwko et al., 2010); and siMAD2L1, 59-CAGAAAGCTATCCAGG- chemiluminescence. The following antibodies were used: mouse anti- ATGAA-39 (Piwko et al., 2010). Cells were synchronized at the G1/S CENP-A (1:1000, Abcam), rabbit anti-RBBP7 (1:1000, Abcam), mouse transition by 16 h incubation in 2 mM thymidine (Sigma, St. Louis, MO), anti-RBBP4 (1:1000, Abcam), mouse anti-GAPDH (1:2000, Sigma), washed twice with pre-warmed phosphate-buffered saline (PBS) solution, rabbit anti-HJURP (1:1000) (Foltz et al., 2009), mouse anti-DDB1 incubated for 9 h in culture medium and then incubated for another 16 h in (1:1000; BD Biosciences, San Jose, CA), rabbit anti-CUL4A (1:1000) 2 mM thymidine. (Olma et al., 2009), mouse anti-RPA (1:2000, Imgenex, Novus Biologicals, Littleton, CO), mouse anti-a-tubulin (1:10,000, Sigma), Immunofluorescence, SNAP tag labelling and microscopy mouse anti-histone H3 (1:1000, Upstate, Merck) and peroxidase- Cells were fixed as described previously (McHedlishvili et al., 2012). The conjugated goat anti-mouse or anti-rabbit IgG (1:5000, Pierce) antibodies. following antibodies were used: anti-CREST antisera (1:250, Antibodies Incorporated), rabbit anti-cyclin-A (1:500, Santa Cruz Biotechnology) and Flow cytometry and BrdU labelling mouse anti-CENP-A (1:2000, Abcam, Cambridge, UK). Cross-absorbed Cells were collected by trypsinization, ixed in 70% ethanol overnight at fluorescently labelled antibodies were used (Invitrogen). SNAP-tag 220˚C, washed in PBS 0.25% Triton X-100 and blocked for 30 min in PBS labelling was performed as described previously (Bodor et al., 2012; 5% FCS 0.25% Triton X-100. Immunostaining was performed after 1 h Jansen et al., 2007). For quenching, cells were incubated for 30 min in incubation with rabbit anti-phosphorylated Ser10 histone H3 (1:50, 2 mM bromothenylpteridine (New England Biolabs, Ipswich, MA) Upstate) at room temperature, followed by 30 min incubation with dissolved in culture medium followed by two washes with pre-warmed conjugated anti-rabbit IgG Alexa 488 (1:500, Invitrogen). Cells were PBS, a 30-min incubation in culture medium and another wash with pre- resuspended for 30 min at 37˚Cin50mg/ml propidium iodide, 20 mg/ml warmed PBS. For pulse labelling, cells were incubated for 15 min in 2 mM ribonuclease and 38 mM Na3Citrate (pH 7.5) solution. Flow-cytometry TMR-Star (New England Biolabs) dissolved in culture medium followed was performed with a FACScalibur flow cytometer (BD Biosciences) using by the same sequence of washes described above for the SNAP-tag CellQuest software. Data analysis was performed using FlowJo software. quenching. Different chase timing was performed for the deposition and For assessing cell proliferation by BrdU labelling, cells were incubated in maintenance assays as depicted in Fig. 5A and Fig. 6A, respectively. z- culture media supplemented with 30 mM 5-Bromo-29-Deoxyuridine stacks of cells were acquired in 0.2-mm z-steps over a 50 mm depth using a (BrdU, Sigma) for 30 min and washed twice with culture media. After 1006oil (NA 1.4) UPlanSApo objective (Olympus, Shinjuku, Japan) with 16 h, cells were harvested, fixed and incubated in 2 M HCl containing 0.5% a Cell-R epifluorescence microscope (Olympus) mounted with a Orca ER Triton X-100 for 30 min at room temperature. 0.1 M sodium tetraborate camera (Hamamatsu, Hamamatsu City, Japan) or a 1006 oil (NA 1.45) (Sigma) was then added and cells were washed with PBS containing 1% Plan Apochromat objective (Olympus) with an Eclipse Ti epifluorescence BSA (BSA, Fraction V, Roche) before staining with conjugated anti-BrdU microscope (Nikon Instruments, Melville, NY) mounted with a Orca- FITC antibody (1:100, eBioscience, San Diego, CA). DNA labelling, data Flash4.0 camera (Hamamatsu). Images were deconvolved using the acquisition and analysis were performed as described. Huygens deconvolution software (SVI), prepared with Fiji and assembled with Illustrator (Adobe). Pull downs and immunoprecipitation Transiently transfected HeLa ‘Kyoto’ cell lines were collected from Image analysis tissue culture plates by scraping in lysis buffer [10 mM Tris-HCl pH 8, Non-deconvolved z-stack images were projected using a maximum 100 mM KCl, 2 mM MgCl2, 0.5% NP-40, 300 mM sucrose, 10 mM b- intensity projection method (Fiji). A modified version of CellProfiler glycerophosphate, 0.2 mM NaF, 0.2 mM PMSF, 0.5 mM DTT, (version 1.0.5122) (Carpenter et al., 2006) was used to analyse the dataset leupeptine, protease inhibitors mix (Roche) and nuclease (Pierce)] and images. Global image thresholding using Otsu’s method was applied to cells were lysed by several passage through a 27G needle on ice. Cell the DAPI channel to segment nuclei. Cyclin A nuclear intensity was lysates were centrifuged at 11,000 g for 15 min at 4˚C, and cleared determined by subtracting the median pixel value of the expanded supernatant was applied to non-reactive agarose resins for 10 min at 4˚C segmented nuclear area edge, corresponding the local background signal, to reduce non-specific binding to resins. The protein concentration of the

1743 RESEARCH ARTICLE Journal of Cell Science (2015) 128, 1732–1745 doi:10.1242/jcs.162305

pre-cleared supernatants was measured by a Bradford assay, and equal Competing interests amounts of protein were applied to anti-HA-coupled agarose resin (clone The authors declare no competing or financial interests. HA-7, Sigma) for 2 h at 4˚C. Elution was performed using 100 mM glycine (pH 2) and eluates were neutralized by addition of ammonium bicarbonate. Author contributions To immunoprecipitate endogenous CUL4A, HeLa ‘Kyoto’ cells were lysed J.M., S.G., M.G.M., F.L. and M.B. performed the experiments; J.M., S.G., M.G.M., as described above and incubated with Affi-Prep Protein A Support resin F.L., P.M. and M.P. participated in the experimental design; and J.M., S.G. and M.P. wrote the paper. (Biorad). Rabbit IgG and rabbit anti-CUL4A antibodies (Olma et al., 2009) were incubated for 1 h at room temperature and were subsequently washed Funding extensively with 0.2 M sodium tetraborate pH 9 (Sigma) before chemical Work in the Peter laboratory was supported by the Swiss National Science cross-linking using 20 mM dimethyl pimelimidate (Sigma) for 30 min Foundation (SNF) [grant number 31003A_141148], an ERC senior award [ERC followed by extensive washes with 250 mM Tris-HCl (pH 8). Pre-cleared Rubicon 268930] and ETH Zurich (ETHZ) [grant number FEL01092]. P.M. was cell lysates were incubated 20 min at 4˚C with cross-linked resins, and supported by the SNF [grant number 31003A_151256], the University of Geneva elution was performed as described above using 100 mM glycine (pH 2). and the Louis-Jeantet Foundation. J.M. was supported by a Marie-Curie Actions Intra-European Fellowship [grant agreement number 225141] and an ETHZ Handling of insect cells, protein expression and pull-down assays Postdoctoral Fellowship. F.L. was supported by a postdoctoral fellowship from the Insect cell lines were grown in suspension in SF-900 II SFM medium Human Frontiers Science Foundation (HFSP) [grant number HFSP_LT000376/ (Invitrogen) in a humidified incubator at 27˚C under constant shaking. 2014]. M.G.M. was supported by a Predoctoral grant of the SNF [grant number RBBP7 and RBBP4 complete open reading frames were cloned into PDFMP3-124904]. pFastBac Dual vectors (Invitrogen) containing either an N- or C-terminal Strep tag. Bacmids were produced in DH10Bac/Multibac E. coli bacterial Supplementary material strains and extracted according to manufacturer’s instructions. Log-phase Supplementary material available online at Sf-9 cell lines (Invitrogen) were transformed with Bacmid DNA using http://jcs.biologists.org/lookup/suppl/doi:10.1242/jcs.162305/-/DC1 Cellfectin II (Invitrogen) according to the manufacturer’s instructions. Strep-tagged RBBP7 and RBBP4 viruses were amplified after successive References rounds of infections. Viruses for His–DDB1 infection were a kind gift of Allshire, R. C. and Karpen, G. H. (2008). Epigenetic regulation of centromeric Nicolas H. Thoma¨ (Scrima et al., 2008). Protein expression for pull-downs chromatin: old dogs, new tricks? Nat. Rev. Genet. 9, 923-937. was performed after co-infection of High Five cells (Invitrogen). Infected Bade, D., Pauleau, A.-L., Wendler, A. and Erhardt, S. (2014). The E3 ligase CUL3/RDX controls centromere maintenance by ubiquitylating and stabilizing cells were lysed by sonication in cold extraction buffer (50 mM Tris-HCl CENP-A in a CAL1-dependent manner. Dev. Cell 28, 508-519. pH 8, 250 mM NaCl, 4 mM MgCl2, 0.5% NP-40, 5% glycerol, 2 mM ATP, Barnhart, M. C., Kuich, P. H. J. L., Stellfox, M. E., Ward, J. A., Bassett, E. A., 2 mM DTT, 1 mM PMSF, benzonase, benzamidine, proteases inhibitor Black, B. E. and Foltz, D. R. (2011). HJURP is a CENP-A chromatin assembly mix, Roche). Lysates were centrifuged at 12,000 g for 30 min at 4˚C, and factor sufficient to form a functional de novo kinetochore. J. Cell Biol. 194, 229- 243. pre-cleared supernatant was applied to magnetic beads for 1 h at 4˚C. M- Bassett, E. A., DeNizio, J., Barnhart-Dailey, M. C., Panchenko, T., Sekulic, N., 280 Streptavidin dynabeads (Invitrogen) or TALON dynabeads (Invitrogen) Rogers, D. J., Foltz, D. R. and Black, B. E. (2012). HJURP uses distinct were used for Strep- or His-tag pulldown assays, respectively. CRL4ARBBP7 CENP-A surfaces to recognize and to stabilize CENP-A/histone H4 for and CRL4BRBBP7 complexes harbouring a FLAG-tag fused to the N- centromere assembly. Dev. Cell 22, 749-762. Bergmann, J. H., Rodrı´guez, M. G., Martins, N. M. C., Kimura, H., Kelly, D. A., terminus of the respective CUL4 subunit were expressed from a single Masumoto, H., Larionov, V., Jansen, L. E. T. and Earnshaw, W. C. (2011). vector (pFL MultiBac) in Sf-9 cells. The cell pellets were resuspended in Epigenetic engineering shows H3K4me2 is required for HJURP targeting and lysis buffer [50 mM Tris-HCl pH 7.5, 350 mM NaCl, 1 mM EDTA, 0.5% CENP-A assembly on a synthetic human kinetochore. EMBO J. 30, 328-340. Triton-X, 1 mM DTT, protease inhibitor cocktail (Roche)] and purified Biterge, B. and Schneider, R. (2014). Histone variants: key players of chromatin. Cell Tissue Res. 356, 457-466. with M2 FLAG-affinity agarose (Sigma-Aldrich) for 1 h, gently rotating at Black, B. E., Jansen, L. E. T., Maddox, P. S., Foltz, D. R., Desai, A. B., Shah, J. V. 4˚C. Subsequently, the resin was washed three times with lysis buffer and and Cleveland, D. W. (2007). Centromere identity maintained by nucleosomes three times with wash buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl). assembled with histone H3 containing the CENP-A targeting domain. Mol. Cell Elution of CRL complexes was achieved by applying one resin volume 25, 309-322. Blower, M. D., Sullivan, B. A. and Karpen, G. H. (2002). Conserved organization of 36FLAG peptide at final concentration of 2 mg/ml in wash buffer with of centromeric chromatin in flies and humans. Dev. Cell 2, 319-330. 10% glycerol. Bodor, D. L., Rodrı´guez, M. G., Moreno, N. and Jansen, L. E. T. (2012). Analysis of protein turnover by quantitative SNAP-based pulse-chase imaging. Statistical analysis Curr. Protoc. Cell Biol. 55, 8.8.1-8.8.34. Lilliefors’ composite goodness-of-fit test was applied to test normality Bodor, D. L., Valente, L. P., Mata, J. F., Black, B. E. and Jansen, L. E. T. (2013). Assembly in G1 phase and long-term stability are unique intrinsic features of of the datasets. A Wilcoxon rank sum test or paired two-sided Student’s CENP-A nucleosomes. Mol. Biol. Cell 24, 923-932. t-test was performed to compare datasets when appropriate. Statistical Carpenter, A. E., Jones, T. R., Lamprecht, M. R., Clarke, C., Kang, I. H., analyses were performed using MATLAB software (MathWorks). Friman, O., Guertin, D. A., Chang, J. H., Lindquist, R. A., Moffat, J. et al. Standard deviation and standard error of the mean were calculated (2006). CellProfiler: image analysis software for identifying and quantifying cell phenotypes. Genome Biol. 7, R100. using Excel (Microsoft). Dataset sizes are indicated as ‘n’. Carroll, C. W., Silva, M. C. C., Godek, K. M., Jansen, L. E. T. and Straight, A. F. (2009). Centromere assembly requires the direct recognition of CENP-A Note added in proof nucleosomes by CENP-N. Nat. Cell Biol. 11, 896-902. Carroll, C. W., Milks, K. J. and Straight, A. F. (2010). Dual recognition of CENP- After acceptance of this work for publication, it has come to the A nucleosomes is required for centromere assembly. J. Cell Biol. 189, 1143- attention of the authors the recent work by Niikura et al. (2015), 1155. who report that CRL4 in complex with COPS8/CSN8 is also Dambacher, S., Deng, W., Hahn, M., Sadic, D., Fro¨ hlich, J., Nuber, A., Hoischen, C., Diekmann, S., Leonhardt, H. and Schotta, G. (2012). CENP-C required for CENP-A centromeric localization. facilitates the recruitment of M18BP1 to centromeric chromatin. Nucleus 3, 101- 110. Acknowledgements Dunleavy, E. M., Roche, D., Tagami, H., Lacoste, N., Ray-Gallet, D., We thank D. Foltz (University of Virginia, USA), L. Jansen (IGC, Portugal) and N. Nakamura, Y., Daigo, Y., Nakatani, Y. and Almouzni-Pettinotti, G. (2009). Thoma¨ (FMI, Switzerland) for sharing reagents; R. Aebersold and P. Picotti for HJURP is a cell-cycle-dependent maintenance and deposition factor of CENP-A at centromeres. Cell 137, 485-497. providing access to MS equipment; J. Gerez and E. Milani for MS support; the Dunleavy, E. M., Almouzni, G. and Karpen, G. H. (2011). H3.3 is deposited at ETH ScopeM facility for microscopy support; M. Unger and P. Horvath for help centromeres in S phase as a placeholder for newly assembled CENP-A in G1 with MATLAB; S. Erhardt for sharing unpublished results; members of the Peter phase. Nucleus 2, 146-157. and Meraldi laboratories for helpful discussions; and W. Piwko and A. Smith for Ekwall, K. (2007). Epigenetic control of centromere behavior. Annu. Rev. Genet. critical reading of the manuscript. 41, 63-81.

1744 RESEARCH ARTICLE Journal of Cell Science (2015) 128, 1732–1745 doi:10.1242/jcs.162305

Fachinetti, D., Folco, H. D., Nechemia-Arbely, Y., Valente, L. P., Nguyen, K., Niikura, Y., Kitagawa, R., Ogi, H., Abdulle, R., Pagala, V. and Kitagawa, K. Wong, A. J., Zhu, Q., Holland, A. J., Desai, A., Jansen, L. E. T. et al. (2013). A (2015). CENP-A K124 ubiquitylation is required for CENP-A deposition at the two-step mechanism for epigenetic specification of centromere identity and centromere. 32, 598-603. function. Nat. Cell Biol. 15, 1056-1066. O’Connell, B. C. and Harper, J. W. (2007). Ubiquitin proteasome system (UPS): Fischer, E. S., Scrima, A., Bo¨ hm, K., Matsumoto, S., Lingaraju, G. M., Faty, what can chromatin do for you? Curr. Opin. Cell Biol. 19, 206-214. M., Yasuda, T., Cavadini, S., Wakasugi, M., Hanaoka, F. et al. (2011). The Obuse, C., Yang, H., Nozaki, N., Goto, S., Okazaki, T. and Yoda, K. (2004). molecular basis of CRL4DDB2/CSA ubiquitin ligase architecture, targeting, and Proteomics analysis of the centromere complex from HeLa interphase cells: UV- activation. Cell 147, 1024-1039. damaged DNA binding protein 1 (DDB-1) is a component of the CEN-complex, Foltz, D. R., Jansen, L. E. T., Bailey, A. O., Yates, J. R., III, Bassett, E. A., while BMI-1 is transiently co-localized with the centromeric region in interphase. Wood, S., Black, B. E. and Cleveland, D. W. (2009). Centromere-specific Genes Cells 9, 105-120. assembly of CENP-a nucleosomes is mediated by HJURP. Cell 137, 472-484. Ohzeki, J., Bergmann, J. H., Kouprina, N., Noskov, V. N., Nakano, M., Kimura, Han, J., Zhang, H., Zhang, H., Wang, Z., Zhou, H. and Zhang, Z. (2013). A Cul4 H., Earnshaw, W. C., Larionov, V. and Masumoto, H. (2012). Breaking the E3 ubiquitin ligase regulates histone hand-off during nucleosome assembly. Cell HAC Barrier: histone H3K9 acetyl/methyl balance regulates CENP-A assembly. 155, 817-829. EMBO J. 31, 2391-2402. Hayashi, T., Fujita, Y., Iwasaki, O., Adachi, Y., Takahashi, K. and Yanagida, M. Olma, M. H., Roy, M., Le Bihan, T., Sumara, I., Maerki, S., Larsen, B., Quadroni, (2004). Mis16 and Mis18 are required for CENP-A loading and histone M., Peter, M., Tyers, M. and Pintard, L. (2009). An interaction network of the deacetylation at centromeres. Cell 118, 715-729. mammalian COP9 signalosome identifies Dda1 as a core subunit of multiple He, Y. J., McCall, C. M., Hu, J., Zeng, Y. and Xiong, Y. (2006). DDB1 functions Cul4-based E3 ligases. J. Cell Sci. 122, 1035-1044. as a linker to recruit receptor WD40 proteins to CUL4-ROC1 ubiquitin ligases. Perpelescu, M., Nozaki, N., Obuse, C., Yang, H. and Yoda, K. (2009). Active Genes Dev. 20, 2949-2954. establishment of centromeric CENP-A chromatin by RSF complex. J. Cell Biol. Held, M., Schmitz, M. H. A., Fischer, B., Walter, T., Neumann, B., Olma, M. H., 185, 397-407. Peter, M., Ellenberg, J. and Gerlich, D. W. (2010). CellCognition: time- Pickart, C. M. (2004). Back to the future with ubiquitin. Cell 116, 181-190. resolved phenotype annotation in high-throughput live cell imaging. Nat. Piwko, W., Olma, M. H., Held, M., Bianco, J. N., Pedrioli, P. G. A., Hofmann, K., Methods 7, 747-754. Pasero, P., Gerlich, D. W. and Peter, M. (2010). RNAi-based screening Hemmerich, P., Weidtkamp-Peters, S., Hoischen, C., Schmiedeberg, L., identifies the Mms22L-Nfkbil2 complex as a novel regulator of DNA replication in Erliandri, I. and Diekmann, S. (2008). Dynamics of inner kinetochore assembly human cells. EMBO J. 29, 4210-4222. and maintenance in living cells. J. Cell Biol. 180, 1101-1114. Ranjitkar, P., Press, M. O., Yi, X., Baker, R., MacCoss, M. J. and Biggins, S. Hewawasam, G., Shivaraju, M., Mattingly, M., Venkatesh, S., Martin-Brown, S., (2010). An E3 ubiquitin ligase prevents ectopic localization of the Florens, L., Workman, J. L. and Gerton, J. L. (2010). Psh1 is an E3 ubiquitin centromeric histone H3 variant via the centromere targeting domain. Mol. ligase that targets the centromeric histone variant Cse4. Mol. Cell 40, 444-454. Cell 40, 455-464. Hotton, S. K. and Callis, J. (2008). Regulation of cullin RING ligases. Annu. Rev. Ray-Gallet, D., Woolfe, A., Vassias, I., Pellentz, C., Lacoste, N., Puri, A., Plant Biol. 59, 467-489. Schultz, D. C., Pchelintsev, N. A., Adams, P. D., Jansen, L. E. T. et al. (2011). Jansen, L. E. T., Black, B. E., Foltz, D. R. and Cleveland, D. W. (2007). Dynamics of histone H3 deposition in vivo reveal a nucleosome gap-filling Propagation of centromeric chromatin requires exit from mitosis. J. Cell Biol. mechanism for H3.3 to maintain chromatin integrity. Mol. Cell 44, 928-941. 176, 795-805. Re´gnier, V., Vagnarelli, P., Fukagawa, T., Zerjal, T., Burns, E., Trouche, D., Kato, T., Sato, N., Hayama, S., Yamabuki, T., Ito, T., Miyamoto, M., Kondo, S., Earnshaw, W. and Brown, W. (2005). CENP-A is required for accurate Nakamura, Y. and Daigo, Y. (2007). Activation of Holliday junction recognizing chromosome segregation and sustained kinetochore association of BubR1. Mol. protein involved in the chromosomal stability and immortality of cancer cells. Cell. Biol. 25, 3967-3981. Cancer Res. 67, 8544-8553. Ribeiro, S. A., Vagnarelli, P., Dong, Y., Hori, T., McEwen, B. F., Fukagawa, T., Kim, I. S., Lee, M., Park, K. C., Jeon, Y., Park, J. H., Hwang, E. J., Jeon, T. I., Flors, C. and Earnshaw, W. C. (2010). A super-resolution map of the Ko, S., Lee, H., Baek, S. H. et al. (2012). Roles of Mis18a in epigenetic vertebrate kinetochore. Proc. Natl. Acad. Sci. USA 107, 10484-10489. regulation of centromeric chromatin and CENP-A loading. Mol. Cell 46, 260-273. Saade, E., Mechold, U., Kulyyassov, A., Vertut, D., Lipinski, M. and Ogryzko, Lacoste, N., Woolfe, A., Tachiwana, H., Garea, A. V., Barth, T., Cantaloube, S., V. (2009). Analysis of interaction partners of H4 histone by a new proteomics Kurumizaka, H., Imhof, A. and Almouzni, G. (2014). Mislocalization of the approach. Proteomics 9, 4934-4943. centromeric histone variant CenH3/CENP-A in human cells depends on the Schuh, M., Lehner, C. F. and Heidmann, S. (2007). Incorporation of Drosophila chaperone DAXX. Mol. Cell 53, 631-644. CID/CENP-A and CENP-C into centromeres during early embryonic anaphase. Lagana, A., Dorn, J. F., De Rop, V., Ladouceur, A.-M., Maddox, A. S. and Curr. Biol. 17, 237-243. Maddox, P. S. (2010). A small GTPase molecular switch regulates epigenetic Scrima, A., Konı´ckova´, R., Czyzewski, B. K., Kawasaki, Y., Jeffrey, P. D., centromere maintenance by stabilizing newly incorporated CENP-A. Nat. Cell Groisman, R., Nakatani, Y., Iwai, S., Pavletich, N. P. and Thoma¨,N.H. Biol. 12, 1186-1193. (2008). Structural basis of UV DNA-damage recognition by the DDB1-DDB2 Lee, J. and Zhou, P. (2007). DCAFs, the missing link of the CUL4-DDB1 ubiquitin complex. Cell 135, 1213-1223. ligase. Mol. Cell 26, 775-780. Shuaib, M., Ouararhni, K., Dimitrov, S. and Hamiche, A. (2010). HJURP binds Loyola, A. and Almouzni, G. (2004). Histone chaperones, a supporting role in the CENP-A via a highly conserved N-terminal domain and mediates its deposition limelight. Biochim. Biophys. Acta 1677,3-11. at centromeres. Proc. Natl. Acad. Sci. USA 107, 1349-1354. Mchedlishvili, N., Wieser, S., Holtackers, R., Mouysset, J., Belwal, M., Amaro, Silva, M. C. C., Bodor, D. L., Stellfox, M. E., Martins, N. M. C., Hochegger, H., A. C. and Meraldi, P. (2012). Kinetochores accelerate centrosome separation Foltz, D. R. and Jansen, L. E. T. (2012). Cdk activity couples epigenetic to ensure faithful chromosome segregation. J. Cell Sci. 125, 906-918. centromere inheritance to cell cycle progression. Dev. Cell 22, 52-63. Mellone, B. G., Grive, K. J., Shteyn, V., Bowers, S. R., Oderberg, I. and Stellfox, M. E., Bailey, A. O. and Foltz, D. R. (2013). Putting CENP-A in its place. Karpen, G. H. (2011). Assembly of Drosophila centromeric chromatin proteins Cell. Mol. Life Sci. 70, 387-406. during mitosis. PLoS Genet. 7, e1002068. Sullivan, B. A. and Karpen, G. H. (2004). Centromeric chromatin exhibits a Merlet, J., Burger, J., Gomes, J.-E. and Pintard, L. (2009). Regulation of cullin- histone modification pattern that is distinct from both euchromatin and RING E3 ubiquitin-ligases by neddylation and dimerization. Cell. Mol. Life Sci. heterochromatin. Nat. Struct. Mol. Biol. 11, 1076-1083. 66, 1924-1938. Tagami, H., Ray-Gallet, D., Almouzni, G. and Nakatani, Y. (2004). Histone H3.1 Moree, B., Meyer, C. B., Fuller, C. J. and Straight, A. F. (2011). CENP-C recruits and H3.3 complexes mediate nucleosome assembly pathways dependent or M18BP1 to centromeres to promote CENP-A chromatin assembly. J. Cell Biol. independent of DNA synthesis. Cell 116, 51-61. 194, 855-871. Walczak, C. E., Cai, S. and Khodjakov, A. (2010). Mechanisms of chromosome Moreno-Moreno, O., Medina-Giro´ , S., Torras-Llort, M. and Azorı´n, F. (2011). behaviour during mitosis. Nat. Rev. Mol. Cell Biol. 11, 91-102. The F box protein partner of paired regulates stability of Drosophila centromeric Wang, J., Liu, X., Dou, Z., Chen, L., Jiang, H., Fu, C., Fu, G., Liu, D., Zhang, J., histone H3, CenH3(CID). Curr. Biol. 21, 1488-1493. Zhu, T. et al. (2014). Mitotic regulator Mis18b interacts with and specifies the Mu¨ ller, S. and Almouzni, G. (2014). A network of players in H3 histone variant centromeric assembly of molecular chaperone HJURP. J. Biol. Chem. 289, deposition and maintenance at centromeres. Biochim. Biophys. Acta 1839, 241- 8326-8336. 250. Westhorpe, F. G. and Straight, A. F. (2013). Functions of the centromere and Mu¨ ller, S., Montes de Oca, R., Lacoste, N., Dingli, F., Loew, D. and Almouzni, kinetochore in chromosome segregation. Curr. Opin. Cell Biol. 25, 334-340. G. (2014). Phosphorylation and DNA binding of HJURP determine its Zaidi, I. W., Rabut, G., Poveda, A., Scheel, H., Malmstro¨ m, J., Ulrich, H., centromeric recruitment and function in CenH3(CENP-A) loading. Cell Hofmann, K., Pasero, P., Peter, M. and Luke, B. (2008). Rtt101 and Mms1 in Reports 8, 190-203. budding yeast form a CUL4(DDB1)-like ubiquitin ligase that promotes replication Murzina, N. V., Pei, X.-Y., Zhang, W., Sparkes, M., Vicente-Garcia, J., Pratap, through damaged DNA. EMBO Rep. 9, 1034-1040. J. V., McLaughlin, S. H., Ben-Shahar, T. R., Verreault, A., Luisi, B. F. et al. Zasadzin´ ska, E., Barnhart-Dailey, M. C., Kuich, P. H. J. L. and Foltz, D. R. (2008). Structural basis for the recognition of histone H4 by the histone- (2013). Dimerization of the CENP-A assembly factor HJURP is required for chaperone RbAp46. Structure 16, 1077-1085. centromeric nucleosome deposition. EMBO J. 32, 2113-2124.

1745 2. CRL4RBBP7 is required for efficient CENP-A deposition at centromeres

2.1. Supplementary figures

Figure S1. CUL4A/B, DDB1 and RBBP7 are required for correct mitotic progression and cell proliferation. (A) Average metaphase duration quantified from live-cell time-lapse imaging using a HeLa cell line expressing histone H2B- mCherry as shown in Fig. 1A. CellCognition software was used to classify chromatin morphology and to quantify metaphase timing. Error bars represent standard deviation, and p-values from a Wilcoxon rank sum test are indicated accordingly for each RNAi conditions. For each experiment there were more than 4 replicates. ‘n’ represents the number of time-lapse movie acquired. Each movie contains at least 30 cells tracked. (B) Flow- cytometry estimating cell populations positive for the mitotic marker phosphorylated Ser10 histone H3 (pHisH3). The pHisH3 positive, mitotic (M) cells are indicated by the black box. The percentage (%) indicates the proportion of mitotic cells in the sample. The DNA content was measured by propidium iodide (PI) staining. (C, D) The proliferation rate over a 24h-period and the rate of apoptosis was quantified from live-cell time-lapse movies as shown in Suppl. Fig. S1A. Error bars represent standard deviation. CellCognition was used for the analysis. For each experiment there were more than 4 replicates. ‘n’ represents the number of time-lapse movie acquired. Each movie contains at least 30 cells tracked.

2. CRL4RBBP7 is required for efficient CENP-A deposition at centromeres

Figure S2. Verification of the automated image analysis pipeline developed to quantify CENP-A levels at centromeres. (A-B) Asynchronous and DTBR-treated HeLa cell populations were compared for their nuclear cyclin A levels by immunofluorescence. Example images (B) and the distribution of cyclin A levels per nuclei after automated analysis are shown (A). The cyclin A levels for the presumed G1/S population based on the distribution after DTBR is highlighted, and the percentage of G1/S cells was estimated for both conditions. (C) Comparison of centromeres detection and quantification of pixel levels by two different methods: a semi-automated method using the Imaris software and the automated method described in Fig. 2 combining CellProfiler software and MATLAB in-house developed scripts. ‘n’ represents number of centromeres. (D) Comparison between manual counting and the automated image analyses pipeline described in Fig. 2. ‘n’ represents number of centromeres. (E, G) Flow cytometry analyses of cell cycle profiles in different RNAi conditions following DTBR treatment shown in Figs 2, 4. The DNA content was measured by PI-staining. (F, H) Western blot of whole cell extracts (WCE) prepared from G1/S synchronized HeLa cells controlling the RNAi efficiency of the different RNAi conditions shown in Figs 2, 4.

2. CRL4RBBP7 is required for efficient CENP-A deposition at centromeres

Figure S3. Controls to estimate efficiency of the SNAP-tag labeling steps. (A, C, E) Representative images of HeLa cell lines expressing SNAP-tagged histone H3 variants illustrating the quantification from (B), (D) and (F). ‘Chase’ corresponds to cell treatment as depicted in Fig. 5A. ‘No Quench’ corresponds to cell treatment as depicted in Fig. 6A with a chase incubation of 24 hours following fluorescent labeling. ‘No Chase’ indicates that quenching and fluorescent labeling were applied sequentially without any incubation time in-between. Background fluorescence controls are indicated either as ‘No Pulse’ when only quenching solution was applied or as ‘DMSO’ when neither quenching nor pulse labeling solutions were applied to the cells. (B, D, F) Quantification of chromatin SNAP-tagged histone H3 variants maximum pixel levels in (D) and (F) or fluorescent cells proportion in (B) for in silico synchronized HeLa cells. ‘n’ represents number of nuclei measured and error bars are the standard error of the mean.

2. CRL4RBBP7 is required for efficient CENP-A deposition at centromeres

Figure S4. Control of release efficiency and cell cycle progression in cells depleted as indicated for CENP-A, HJURP, CUL4A/B, DDB1 or RBBP7. (A-B) Cells expressing SNAP-CENP-A constructs were treated by DTBR as described in Fig. 5 and cell cycle profile using PI-staining and flow cytometry analysis (A) or cyclin A nuclear levels by immunofluorescence (B) were used to monitor cell cycle restart following DTBR treatment. (figure legend continues in the next page) 64

2. CRL4RBBP7 is required for efficient CENP-A deposition at centromeres

Figure S4 (continued). Thymidine was added back 9h after release as described in Fig. 5. ‘n’ represents number of nuclei. (C) Cells expressing SNAP-CENP-A constructs were treated by DTBR as described in Fig. 6. A BrdU pulse was performed 16h before harvesting either at day 2 (upper row) or day 3 (lower row) after release and BrdU incorporation was analyzed by flow cytometry. Each histogram shows cells that stain positive for BrdU, either at the time of labeling (red) or at the time of harvest, 16h after labeling (grey). Note that all cells progressed into S-phase after the release, and a cell cycle delay was observed in DDB1-depleted cells only after 3 days of RNAi treatment.

3. Manuscript in preparation: Hyperphosphorylation repurposes the CRL4B E3 ligase to coordinate mitotic entry and exit

Manuscript, not yet submitted

Contribution (first author) o Writing the manuscript and assembling the figures o Planning and execution of the presented experiments

Additional information The supplementary figures S1 to S17 are included after the references section within this chapter. The following additional information can be found in Appendix B: B.1 – Conservation of CUL4B protein B.2 – Analysis of CUL4 interactions by mass spectrometry B.3 – Obtainment of CUL4B knockout clones B.4 – CRISPR knock-in of CUL4B mutants B.5 – Optimization of immunofluorescence with pre-extraction

3. CRL4B repurposing by mitotic phosphorylation

Hyperphosphorylation repurposes the CRL4B E3 ligase to coordinate mitotic entry and exit

Samuel Gilberto1, Fabienne Lampert1, Wojciech Piwko1, and Matthias Peter1,*

1 Department of Biology, Institute of Biochemistry, Swiss Federal Institute of Technology, 8093 Zurich, Switzerland

Running title: CRL4B repurposing by mitotic phosphorylation

*Corresponding author: Prof. Dr. Matthias Peter Tel: +41 44 633 65 86; Fax: +41 44 633 12 98 E-mail: [email protected]

Keywords: Ubiquitin / Mitosis / Cullin-4 / CDK1 / Phosphorylation

3. CRL4B repurposing by mitotic phosphorylation

Abstract

Ubiquitylation depends on an enzyme cascade that allows transfer of ubiquitin to a target protein by an E3 ligase. The E3 ligase family Cullin-RING ligases (CRLs) include Cullin-4A (CUL4A) and CUL4B which together regulate multiple cellular processes, in particular chromatin-associated processes such as chromatin remodeling and DNA repair. However, CUL4B mutations result in syndromic X-linked intellectual disability (XLID), never reported for CUL4A, which denotes a non-overlapping function. Through cell cycle analysis in HeLa cells, we discovered that CUL4B is important for mitotic entry. CUL4B is subject to mitotic phosphorylation in its unique N-terminal sequence, mediated by CDK1, PLK1 and possibly other kinases and consequently no longer localizes to the chromatin. Interestingly, proliferation assays and live-cell imaging experiments revealed that the overexpression of the phosphomimetic mutant is lethal in HeLa cells due to aberrant metaphase progression and impaired mitotic exit, with a subset of cells failing to separate their chromosomes. Together, our data suggests that CUL4B phosphorylation repurposes this E3 ligase for a non-chromatin function that is both important for the early mitotic stages and must be abolished by mitotic exit. Our study establishes a clear regulatory distinction between CUL4B and CUL4A, with functional implications, which may aid in clarifying why CUL4A cannot compensate for CUL4B loss in human XLID patients.

3. CRL4B repurposing by mitotic phosphorylation

Introduction

The cell cycle is essentially successive alternating phases of DNA synthesis and separation, culminating in the process of cell division. Biochemically, the conjugation of ubiquitin (termed ubiquitylation) takes an essential role ensuring this alternate behavior, by regulating the activity of many critical cell cycle effectors, including that of cyclin-dependent kinase (CDK) (Morgan and Roberts, 2002; Bassermann et al., 2014; Teixeira and Reed, 2013; Gilberto and Peter, 2017). Ubiquitylation is a versatile post-translational modification known to regulate protein turnover, activity or subcellular localization. The most notable outcome of ubiquitylation is protein targeting to the proteasome, a proteolytic machine (Livneh et al., 2016). A cascade of E1, E2 and E3 enzymes are responsible for ubiquitin conjugation, being that the E3 ligase enzyme is the one responsible for providing specificity to this reaction. Several E3 ligases, predominantly the anaphase-promoting complex/cyclosome (APC/C), SKP1-CUL1-F-box (SCF) and cullin- RING E3 ligases (CRLs) CRL3 and CRL4, have been involved to different extents in cell cycle regulation (Bassermann et al., 2014; Teixeira and Reed, 2013; Gilberto and Peter, 2017). CRL4-type E3 ligases contain either a cullin-4A (CUL4A) or CUL4B scaffold that brings together the substrate and the ubiquitin-loaded E2 conjugating enzyme so that the reaction can occur. Complexes containing either CUL4A or CUL4B (collectively called CUL4) are comprised by a substrate-binding module, where the substrate adaptor, termed DDB1- and CUL4-associated factor (DCAF), is recruited by the linker protein DNA-damage binding protein 1 (DDB1) (Lee and Zhou, 2007). The separate catalytic domain of the complex is comprised by the RING subunit and is responsible for recruitment of the ubiquitin-loaded E2 enzyme. Interestingly, both CUL4A and CUL4B share the same interacting partners and their structure is virtually indistinguishable (Fischer et al., 2011), with the single difference being the extended N-terminus of CUL4B (Ohtake et al., 2007). CRL4A and CRL4B are both involved in the regulation of multiple cellular events, in particular chromatin-associated processes such as DNA replication, chromatin remodeling and DNA repair. To provide the means to ubiquitylate many substrates in a regulated fashion, CRL4 assemble more than 50 distinct complexes by recruiting a variety of DCAF proteins (Lee and Zhou, 2007). Although functional details for most of these are lacking, evidence suggests that several complexes function in cell cycle-related processes (Piwko et al., 2010), thus arguing that important functions of CRL4 in the regulation of the cell cycle are yet to discover. One additional layer of complexity that remains to clarify is whether both CUL4A and CUL4B are fully functionally redundant. Indeed, functional studies report overlapping functions of the two paralogue cullins, with little distinction between their substrates (Hannah and Zhou, 2015). Despite such reports,

3. CRL4B repurposing by mitotic phosphorylation phenotypic observations come to challenge this view. Unexpectedly, mutations in CUL4B were reported to be associated with syndromic X-linked intellectual disability (XLID) in human males, never the case for CUL4A (Cabezas et al., 2000; Tarpey et al., 2007; Zou et al., 2007; Badura-Stronka et al., 2010; Isidor et al., 2010; Londin et al., 2014; Vulto-van Silfhout et al., 2015; Grozeva et al., 2015; Okamoto et al., 2017; Kerzendorfer et al., 2011). Moreover, phenotypic studies carried in mice demonstrated that CUL4B is essential during mouse embryogenesis, with cells arresting at G2/M (Liu et al., 2012b). Surprisingly, embryonic lethality nor overt growth defects are observed upon deletion of CUL4A (Kopanja et al., 2011; Yin et al., 2011). These observations, in conjunction with potential undiscovered roles in the cell cycle of somatic cells, lead us to examine whether CUL4A and CUL4B can be functionally distinguishable in their cell cycle functions. Our observations show that indeed the deletion of the CUL4B gene impairs proper cell proliferation and G2/M progression in cells where CUL4A is present. CUL4B is specifically hyperphosphorylated in its N-terminus during mitosis, and chromatin fractionation experiments demonstrate that it no longer localizes to the DNA during this phase, whereas CUL4A remains chromatin- bound. Notably, proliferation experiments where a phosphomimetic mutant of CUL4B was overexpressed revealed a lethal phenotype that was related to a failure in mitotic exit and adequate nuclear and cell separation. In light with our data, mitosis defines a point where CUL4A and CUL4B do not act redundantly, with CUL4B acquiring a distinct function which cannot be preserved outside of the cell division context.

3. CRL4B repurposing by mitotic phosphorylation

Results

CUL4B is distinctively regulated by phosphorylation in mitosis CUL4A and CUL4B share extensive sequence homology (82% sequence identity and 89% sequence similarity), with multiple fully conserved structural features (Figure 1A and Supplementary Figure S1). That they perform redundant functions as E3 ligase scaffolds is expected, but we found very intriguing that embryos of CUL4B knockout mice arrest during mitosis (Liu et al., 2012b). We thus set to investigate whether the CUL4 paralogues, in particular CUL4B, function during this stage in somatic human cells. An interesting early observation was that HeLa cells arrested in mitosis display an upshifted CUL4B band (Figure 1B), whereas the migration of the CUL4A band remains unaltered. This upshift was only observed when cells were treated with nocodazole or taxol, which causes a prometaphase or metaphase-anaphase transition arrest, respectively, and not if cells were halted in S or G2 phases. Treatment with λ- Phosphatase, either after CUL4B immunoprecipitation (IP) or directly in whole cell lysates, confirmed that CUL4B is modified by phosphorylation (Figure 1C and Supplementary Figure S2A). Because the main difference between the CUL4 paralogues is the extended serine-rich N-terminus in CUL4B (Figure 1A), we expressed a truncated version of CUL4B lacking its N-terminus (termed ΔN), which was not modified by phosphorylation (Figure 1C and Supplementary Figure S2C). Restoring a nuclear localization signal (NLS) in the ΔN mutant did not recover phosphorylation, arguing that indeed it is the unique N-terminus of CUL4B that is directly modified, and the lack of phosphorylation is not a result of impaired interphase localization. Accordingly, inactivation of the NLS of CUL4B produced a construct with migration pattern identical to the wild-type (WT) (Supplementary Figure S2C,D). Importantly, a chimera CUL4A construct containing CUL4B’s N-terminus is equally phosphorylated as CUL4B WT (Supplementary Figure S2E). We then set to identify the N-terminal regions and the kinase(s) that can potentially phosphorylate CUL4B in mitosis. Several of the serines or threonines have been found phosphorylated in unbiased high-throughput mass spectrometry studies, and based on sequence proximity and frequency of detection as phosphorylated, we subdivided the N-terminus of CUL4B in several phospho- serine/threonine clusters (Figure 1D). Bioinformatic analysis suggests that consensus sites for several mitotic kinases can be found within the sequence of the N-terminus. Amongst the sites most often found phosphorylated are several TP or SP sites (CDK has a mandatory constraint for a proline residue at position +1), which can be potentially phosphorylated by CDK1 (clusters 1, 2, 4, 5 and 6) (Supplementary Figure S3). We set to test several of these kinases in in vitro kinase assays and indeed demonstrate that CDK1- cyclin B can efficiently phosphorylate CUL4B and not CUL4A. CUL4B is nevertheless a promiscuous kinase

3. CRL4B repurposing by mitotic phosphorylation substrate, and can be phosphorylated by at least two mitotic kinases, CDK1-cyclin B and PLK1, but also CK1 – a non-mitotic kinase with several predicted consensus sites in the CUL4B N-terminus (Figure 1E). The absence of CDK1-cyclin B phosphorylation in the ΔN, Pnull-NT (non-phosphorylatable N-terminus) and Pnull 0-4 mutants (similar to Pnull-NT, but the non-phosphorylatable region spans only until cluster 4) demonstrates that, at least for this kinase, phosphorylation takes place exclusively in the CUL4B N- terminal region (Figure 1F). We found especially intriguing that the residue most often found phosphorylated, T49 (Figure 1D), is adjacent to P50, which if mutated to leucine has repercussions in XLID (Supplementary Figure S4A,B). The CUL4B mutations described in XLID are likely loss-of-function mutations (in particular the mutations leading to extensive truncations of the protein sequence), and the reasoning as to why P50L is deleterious for CUL4B function is not understood. We found that the stability and localization of P50L remains unchanged (Supplementary Figure S4C,D). Despite the observation that hyperphosphorylation in arrested cells resembles the WT (Supplementary Figure S4E,F), indicative that the bulk of phosphorylation remains the same, it remains a possibility that this mutation impairs mitotic phosphorylation in specific sites. In particular, T49 is predicted to be phosphorylated by CDK1 (Supplementary Figure S3), and the P50L mutation could abolish its phosphorylation by this kinase. We tested this notion in vitro, but the P50L mutation alone did not abolish general CUL4B phosphorylation by CDK1-cyclin B (Figure 1F). As CDK1 can putatively phosphorylate multiple sites, we hypothesized that CDK1-cyclin B still phosphorylates additional residues and thus we fail to observe a relevant difference when only P50 is mutated. We thus performed the reverse approach, and compared the activity of CDK1- cyclin B towards the Pnull 0-4 mutant with a mutant where cluster 2 is engineered to contain the WT residues (Pnull 0-4 2:WT). Remarkably, this construct was now efficiently phosphorylated by CDK1-cyclin B (Figure 1F), suggesting that T49 is indeed phosphorylated by this kinase, as this is the only putative CDK1 phosphorylation site in cluster 2 (Supplementary Figure S3 and S4B). In summary, we have found that the cullin paralogues CUL4A and CUL4B are distinctively regulated by phosphorylation in mitosis, indicating that both cullins in fact do not function fully redundantly in the cell cycle.

3. CRL4B repurposing by mitotic phosphorylation

Figure 1. The N-terminus of CUL4B is phosphorylated by mitotic kinases. (A) Conserved and non-conserved structural and functional domains in the sequences of CUL4A and CUL4B. Note the Ser-rich N-terminal extension of CUL4B. Amino acid positions refer to CUL4B, and the annotations are according to (Fischer et al., 2011) and UniProtKB (The UniProt Consortium, 2015). The RBX1 binding region and neddylated site were inferred from homology with CUL1 (Zheng et al., 2002). (B) Migration of the CUL4B band analyzed by western blot upon HeLa cell treatment with diverse drugs that cause cell cycle arrest in S phase (S), G2 or mitosis (M). H3 pS10 (Phosphorylated histone H3 in serine 10) is used as a mitotic marker. (Figure legend continues on the next page) 73

3. CRL4B repurposing by mitotic phosphorylation

Figure 1 (continued). (C) Treatment of immunoprecipitated FLAG-CUL4B with lambda phosphatase. HeLa cells were transfected with WT or ΔN (N-terminal truncation)-expressing vectors 48h before IP, and left untreated or arrested in mitosis for 24h. Async – Asynchronous; Noc – Nocodazole. (D) Annotation of the serine and threonine residues on the N-terminus of CUL4B. The chart displays the frequency each residue was found phosphorylated in high- throughput mass spectrometry studies (HTP-MS), according to the PhosphoSitePlus database (Hornbeck et al., 2015). Phosphoclusters were categorized and the furthermost N-terminal region was labeled ‘region 0’ due to the absence of found phosphorylations in these residues. (E) In vitro kinase assay using CUL4B/DDB1/RBX1 (CRL4B; upper panel) or CRL4A (lower panel) as substrates, and a panel of kinases with putative phosphorylation sites in CUL4B. The Coomassie gel denotes equal amount of CRL4A/B as used in the assays. (F) In vitro phosphorylation of CUL4B mutants by CDK1-cyclin B. Pnull-NT: serines and threonine residues on the N-terminus of CUL4B were mutated to alanine. Pnull 0-4: similar to Pnull-NT, but the mutations were performed until cluster 4, leaving the remain of the N-terminus WT. Pnull 0-4 2:WT: Pnull 0-4 mutant where cluster 2 was reverted to the WT sequence. The Coomassie gel is a picture of the gel exposed to the phosphorimager.

Mitotic CUL4B does not retain binding to chromatin CRL4A and CRL4B are nuclear proteins (Supplementary Figure S6A) commonly associated with chromatin processes, in particular DNA repair and the regulation of DNA replication and transcription (Hannah and Zhou, 2015). Due to the mitotic-specific phosphorylation of CUL4B, we set out to uncover whether chromatin binding of CUL4A and CUL4B is affected during mitosis. We first set out to visualize how much CUL4A and CUL4B co-localize with mitotic condensed DNA, in immunofluorescence (IF) experiments. While the available CUL4A-specific antibody allows for a clear detection of CUL4A, there was no available suitable CUL4B-specific antibody for IF usage. To overcome this limitation, we employed a CRISPR-based method to tag endogenous CUL4B with an HA tag (Supplementary Figure S7). IF analysis indicated that a significant pool of these cullins is present in the mitotic cytoplasm and does not co-localize with mitotic condensed DNA (Supplementary Figure S6B). However, while it appears that CUL4A does still co-localize with DNA, CUL4B appears more absent. To explore this possibility, we performed similar IF studies, but first washed off the soluble pool of proteins by detergent treatment before fixing the cells. This approach allowed us to look exclusively at the chromatin-bound pool of each protein (Figure 2A). CUL4A remains chromatin-bound in interphase and all mitotic stages, from prophase to telophase (Figure 2A, see arrowheads, and Figure 2B for quantification). Remarkably, the same is not the case for CUL4B: while it is retained at the chromatin in interphase, it shows a characteristic behavior during mitosis, where it appears to be excluded from the condensed chromosomes in prometaphase, metaphase and anaphase (Figure 2A,B). CUL4B is observed at the chromatin in prophase and telophase. For the latter, it appears that CUL4B readily adopts nuclear localization upon nuclear envelope reformation (Supplementary Figure S6B). Cell fractionation experiments corroborate these findings: whereas a portion of CUL4A and CUL4B is found in the chromatin fraction of a cell cycle-asynchronous population, CUL4B is absent from this fraction in mitotic cells, but not CUL4A (Figure 2C). Due to their functions as E3 ligase scaffolds, it is expected that

3. CRL4B repurposing by mitotic phosphorylation

CUL4A and CUL4B are recruited to DNA via their substrate-binding module, i.e. a chromatin-bound DCAF recruits CRL4 to the site of substrate ubiquitylation (Fischer et al., 2011). We verified that this is indeed the case, as both CUL4A and CUL4B are absent from the chromatin fraction upon DDB1 depletion, in all cell cycle stages (Figure 2D; see Supplementary Figure S8A for a side-by-side comparison of DDB1 depletion). Because DDB1 remains present in the mitotic chromatin fraction (Figure 2C), this raised the possibility that CUL4B is excluded from the chromatin in mitosis as a result of impaired interaction with DDB1. Nevertheless, co-precipitation (co-IP) experiments demonstrate that CUL4B does actually maintain

Figure 2. CUL4B is excluded from the chromatin in mitosis, but not CUL4A or DDB1. (A) Exemplificative immunofluorescence images for evaluation of chromatin binding of CUL4B and CUL4A. Cells were treated with Triton X-100 before formalin fixation, for extraction of the chromatin-unbound soluble fraction. Endogenous CUL4B is tagged with an N-terminal HA tag. Shown are maximal Z-projections of the acquired Z-stacks. White arrows denote cells in the mitotic stage stated on the left. Cells in interphase are included for signal comparison. In all cases, the contrast settings are identical between images of a single channel. Scale bar: 10 μm. (B) The fluorescent integrated intensity of the maximal Z-projection images exemplified in (A) was measured using ImageJ, with subtraction of the background signal. “Mitosis” denotes cells in prometaphase, metaphase or anaphase. The median is shown and error bars depict standard deviation. (C) Cell fractionation experiments in asynchronous and nocodazole-arrested mitotic cells. T, Total fraction, CP, cytoplasm; NP, Nucleoplasm; Chr, Chromatin; C/NP, combined cytoplasm and nucleoplasm in mitotic cells.

3. CRL4B repurposing by mitotic phosphorylation its interaction with DDB1 throughout the cell cycle (Supplementary Figure S9A,B), suggesting another factor for chromatin exclusion. Together, these findings indicate that although CUL4A and CUL4B are found associated with chromatin during interphase, CUL4B is selectively excluded during mitosis in a manner independent on the modulation of CUL4B-DDB1 interaction.

Phosphorylation in a subset of Ser/Thr clusters excludes CUL4B from mitotic chromatin Taking into account the two independent observations that 1) the N-terminus of CUL4B is phosphorylated in mitosis and 2) CUL4B is excluded from mitotic chromatin, we set to investigate what is the relevance of phosphorylation for chromatin exclusion. We discovered that mutagenesis of all serine and threonine residues to alanine within the N-terminus (Pnull-NT) prevents CUL4B from being excluded from the chromatin during mitosis, even though the chromatin recruitment in asynchronous cells resembles the WT (Figure 3A,B). To clarify whether Pnull-NT is indeed chromatin-bound or may integrate other bodies present in the chromatin fraction, we confirmed that Cajal bodies and nuclear speckles do not integrate this fraction in mitosis, even though nuclear speckles are present in interphase (Supplementary Figure S8C). While this data argues that N-terminal phosphorylation is necessary for chromatin exclusion, we inquired if it is also sufficient. Indeed, the concomitant full N-terminus phosphomimetic mutant (Pmimic- NT, serine/threonine to glutamate) appears to be always absent from the chromatin fraction (Figure 3A,B). The utilized mutants span across the entire N-terminus, and we thus set to uncover whether there are sub-regions whose phosphorylation is more relevant for chromatin exclusion. To this end, we assessed the mitotic chromatin binding of a set of mutants with varied non-phosphorylatable (Pnull) and WT regions in cell fractionation experiments (Figure 3C, Supplementary Figure S10 and S11). To exclude indirect effects from abnormal subcellular localization, we confirmed that the tested mutants retain their interphase nuclear localization (Supplementary Figure S12). We discovered that within the N-terminus, it is not simply the absolute number of putative phosphorylation sites that promotes chromatin exclusion. Instead, there are defined regions whose putative phosphorylation appear more relevant to this effect, in particular that cluster 2 and to some extent of region 0 (Figure 3D). In fact, preventing the phosphorylation of the regions/clusters 0 to 4 grants CUL4B the capacity to remain chromatin-bound in mitosis, but it is sufficient to revert cluster 2 to the WT sequence for CUL4B to be again excluded from DNA (Figure 3C). Of lesser importance are clusters 1, 3 and 4, and clusters 5, 6 and 7 are mostly irrelevant for the exclusion of CUL4B from the chromatin (Figure 3C and Supplementary Figure S10C,D). Importantly, the truncation of CUL4B’s N-terminus partly allows for its relocalization to mitotic chromatin, though not to full extent. Moreover, a chimeric CUL4A where its own N-terminal sequence is replaced with CUL4B’s N-terminal

3. CRL4B repurposing by mitotic phosphorylation sequence also shows chromatin unbinding, though not full exclusion like in the case of the full-length CUL4B (Supplementary Figure S10B). Taken together, here we show that CUL4B N-terminal phosphorylation is necessary to for its chromatin exclusion, with particular importance for cluster 2. Concomitantly, phosphomimetic mutations are sufficient to permanently exclude CUL4B from the chromatin.

CUL4B regulates mitotic entry While there was no appreciable delay for the cell population to reach the G2 state in both conditions, 8h after the release, the great majority of control cells went through cell division between 8 and 10h release, whereas only about 20% of CUL4B KO cells had done so. At this stage, most KO cells hadn’t entered mitosis (56.2%). This data suggests that CUL4B knockout delays mitotic entry, but we find no evidence that mitosis itself is delayed. siRNA depletion similarly produces in a delay at the point of cell division, although it is

Figure 3. N-terminal phosphorylation of CUL4B is necessary and sufficient for its chromatin exclusion. Fractionation experiments in cells transfected with FLAG-CUL4B-expressing vectors, encoding for (A,B) WT CUL4B, the nonphosphorylatable mutant Pnull-NT and phosphomimetic Pmimic-NT or (C) additional nonphosphorylatable mutants, as shown. The illustration depicts which clusters are mutated (green) or kept with the WT sequence. P – potential phosphorylation site. (B,C) cells were nocodazole-arrested prior to fractionation. (D) representation of the clusters that appear more likely to drive mitotic chromatin exclusion if phosphorylated. “–“, unlikely to have a role; “+”, somewhat involved; “++”, important role; “+++”, highest relevance. T, Total fraction, CP, cytoplasm; NP, Nucleoplasm; Chr, Chromatin; C/NP, combined cytoplasm and nucleoplasm in mitotic cells. 77

3. CRL4B repurposing by mitotic phosphorylation

Figure 4. CUL4B is important for G2/M progression. Unedited and CUL4B knockout cells were synchronized at the G1/S border via a double-thymidine block (dTB), and released for 8, 10 or 12h before harvest. Top: histogram showing cell count vs DNA content (X axis). Bottom: intensity of phosphorylated histone H3 in serine 10. A.u., arbitrary units. The percentage of cells in G2, mitosis (M) and G1 is shown. Numbers in red and blue represent an increase and decrease in cell percentage, respectively, when compared to the unedited cells. not clear whether it is due to an effect on mitotic entry or progression (Supplementary Figure S14). Importantly, knockout of CUL4B yields a visible consequence on normal cellular growth that appears independent of impaired DNA repair, as we find the levels of a marker for DNA damage, phosphorylated histone H2A.X (γH2A.X), unchanged (Supplementary Figure S13D,E). These data reveal a previously unforeseen function of CUL4B for G2 and/or mitotic progression.

Both the phosphorylation and dephosphorylation of CUL4B are important for mitotic fidelity Provided that the chromatin localization of CUL4B is regulated by phosphorylation specifically in mitosis, we postulated two hypotheses for the functional consequence of such regulation: either CUL4B has a specific chromatin interphase function which is unwanted at the point of cell division, or CUL4B acquires a novel function at this stage that requires its exclusion from DNA (even though the G2/M delay upon CUL4B knockout rather supports the latter). In line with these hypotheses, we tested whether the overexpression of the Pnull-NT or Pnull 0-4 mutants shows a dominant deleterious effect, which in clonogenic assays does not appear to be the case (Figure 5A), as both cell viability and growth appear equal as compared to expression of the WT. Strikingly, overexpression of both the Pmimic-NT or Pmimic 0-4 mutants dramatically impacted cell viability, in the presence of endogenous CUL4B (Figure 5A). Despite the high infection efficiency for the lentiviral vectors used, selection for infected cells with puromycin demonstrated that most of the surviving colonies were negative for the infection, further arguing that overexpression of Pmimic-NT CUL4B mutants alone results in cell death. Identical analysis

3. CRL4B repurposing by mitotic phosphorylation

Figure 5. CUL4B dephosphorylation during mitotic exit ensures nuclear division. (A) Unedited HeLa cells infected with lentivirus expressing untagged CUL4B mutants were seeded 4h post-infection. Bottom: 2 days post- transfection, puromycin was added to select for infected cells only. (B) Selected CUL4B knockout cells, infected with WT or Pmimic-NT-expressing vectors were analyzed 2 weeks post-infection. Ploidy was evaluated in cells arrested at the G1/S border via a double-thymidine block, where all diploid cells are expected to have ~2N DNA. (C) CUL4B knockout cells infected with lentivirus expressing untagged CUL4B mutants for 3h were imaged for 28h. Red: Fluorescent DNA staining with SiR-DNA. Greyscale: Brightfield imaging. White arrows indicate normal mitotic events; Green arrows indicate cells with metaphase defects; Cyan arrows indicate failed karyokinesis and cytokinesis, a possible consequence of defective chromosome segregation or telophase. Scale bar, 10 μm. 79

3. CRL4B repurposing by mitotic phosphorylation with knockout of the endogenous CUL4B yielded comparable results (Supplementary Figure S15A). In this case, parallel analysis of surviving colonies revealed an enrichment for cells containing 4N DNA. To distinguish between a G2/M delay and tetraploid G1 cells, we arrested this population at G1/S with a double-thymidine block. Because one observes two histogram peaks in the cells previously infected with Pmimic-NT reveals the presence of both diploid and tetraploid G1/S cell populations (Figure 5B). Indeed, live-cell imaging analysis starting 3 hours after infection, with endogenous CUL4B knockout, shows that cells infected with Pmimic-NT and Pmimic 0-4 are severely impaired in cell proliferation (Supplementary Figure S15B). Indeed, live-cell imaging experiments exposed various cases of failed karyokinesis and cytokinesis (Figure 5C and Supplementary Figs. S15B and S16, cyan arrows). Moreover, careful analysis indicates that chromosome congression during metaphase is somewhat impaired (Supplementary Figure S16). The expression of nonphosphorylatable constructs resulted in sporadic impairments at the metaphase-to-anaphase transition, as evaluated by prolonged and catastrophic metaphase. Importantly, the pronounced Pmimic-NT effect on cell division appears dependent on the expression levels of these CUL4B mutants, because we did not observe distinct proliferation issues on a stable cell line expressing low levels of the same construct (Supplementary Figure S17A,B). In this setup, with depletion of the endogenous CUL4B, we did observe markedly reduced cell proliferation upon Pnull-NT expression. Together, our results indicate that the overexpression of phosphomimetic CUL4B mutants prevents either adequate chromosome segregation or karyokinesis, whereas preventing CUL4B phosphorylation negatively impacts chromosome congression at the metaphase plate.

3. CRL4B repurposing by mitotic phosphorylation

Discussion CUL4A and CUL4B are known to share a number of functions, and have been taken as being mostly functionally redundant, in particular on what comes to cell cycle roles (Hannah and Zhou, 2015). Nevertheless, studies in mice and the observation that CUL4B loss-of-function mutations are associated with syndromic X-linked intellectual disability argue against full functional redundancy (Jiang et al., 2012; Liu et al., 2012b; Chen et al., 2012; Kerzendorfer et al., 2011). It has been suggested that a differential subcellular localization of both cullins is responsible for their separate function, with CUL4A found more abundant in the cytoplasm (Chen et al., 2012; Hannah and Zhou, 2015; Nakagawa and Xiong, 2011), we and others have demonstrated that both endogenous CUL4A and CUL4B are abundant in the nucleus of diverse human cell lines, in particular in HeLa cells used in this study (Supplementary Figure S6; Olma et al. 2009; see the Human Protein Atlas project entries for CUL4A and CUL4B). Regardless, we show that the distinction between CUL4A and CUL4B goes beyond subcellular localization, and that the unique CUL4B N-terminal sequence is in fact a regulatory region responsible for mediating CUL4B’s chromatin binding state and pivotal in the establishment of a distinct CUL4B function in the regulation of mitotic fidelity. In fact, we are not the first to suggest unique CUL4B functions arising from its N-terminal region, as it was demonstrated that in the presence of environmental toxins, CUL4B assembles an atypical E3 ligase independent on DDB1 by binding the Dioxin receptor via CUL4B’s N-terminus (Ohtake et al., 2007). Nevertheless, here we show that this region also takes up physiological regulatory purposes, with additional functional consequences. Our study provides evidence that these cullins are independently regulated in mitosis and that CUL4B alone appears to take up a novel function to ensure mitotic fidelity. We observed that CUL4B, but not CUL4A, is hyperphosphorylated in mitosis and that this modification excludes the former from the chromatin without disrupting the assembly of the E3 ligase itself.

Mechanisms of CUL4B hyperphosphorylation at mitotic entry The N-terminus of CUL4B contains an extraordinary content of serine and threonine residues, together accounting for 36% of the 194 amino acids. We observe a high shift of the CUL4B band, indicating that this sequence is hyperphosphorylated. We anticipated that the conjunction of several serine/threonine kinases would perform this task, which we show to be the case for CDK1, PLK1 and a non-mitotic kinase, CK1 (Figure 1). We cannot exclude that other kinases than the ones tested phosphorylate the N-terminus of CUL4B. In fact, this is somewhat expected to take place for such a S/T-rich region. It is likely that CUL4B will not appear phosphorylated during interphase due to the action of phosphatases and due to its nuclear localization, which could shield it from cytosolic kinases (although we do not see a band shift in the ΔNLS

3. CRL4B repurposing by mitotic phosphorylation mutant, Supplementary Figure S2D). An additional effect might be that phosphorylation by mitotic kinases primes phosphorylation by other kinases: for instance, the CK1 consensus sequence includes a phosphorylated serine at position -3 (Ubersax and Ferrell Jr, 2007). It has also been reported for numerous proteins that CDK1 phosphorylation is followed by PLK1 binding and phosphorylation, which might occur in CUL4B (Elia et al., 2003; Watanabe et al., 2005). Similar mechanisms could account for the mitotic- specific hyperphosphorylation, even that performed by non-mitotic kinases. One standing question is which kinases have a bigger impact on the mitotic function of CUL4B. Our data supports a model where it is not simply the number of phosphorylated serine/threonine residues that exclude CUL4B from the chromatin (Figure 3), but the location where these phosphorylations take place is relevant (in particular region 0 and cluster 2). This indicates that perhaps specific phosphosites are more relevant for chromatin exclusion, and therefore phosphorylation by the respective kinases is of greater importance. Alternatively, certain residues might be phosphorylated more promptly, and others not phosphorylated at all. Within cluster 2, we explored the possibility of phosphorylation by CDK1-cyclin B – the key kinase in mitotic entry (Gavet and Pines, 2010). CDK1-Cyclin B is imported into the nucleus at the onset of mitosis (Porter and Donoghue, 2003), which represents the stage where we begin to observe reduced CUL4B binding to the chromatin (Figure 2). From these observations, we presume that CDK1- Cyclin B is pivotal in modulating the function of CUL4B. Within cluster 2, T49 is the only site that matches the CDK1 consensus, suggesting a special role for phosphorylation for this site. T49 is the site most often found in unbiased high-throughput mass spectrometry studies, which could indicate that it is also tone of the most abundant. T49 phosphorylation might therefore be the first line of mitotic CUL4B phosphorylation, directly or indirectly allowing subsequent phosphorylation of several other residues. It is important to mention that there is a time gap between full CDK1-cyclin B activation, in prophase, and complete CUL4B chromatin exclusion, in prometaphase (Figure 2). This likely reflects a requirement for several phosphorylated residues for a complete chromatin exclusion. Indeed, it is thought that a substrate that requires many phosphorylations to have its function changed will be regulated later than a substrate that requires fewer phosphorylated residues (Kamenz and Ferrell, 2017). It is thus possible that a requirement for multiple CUL4B phosphorylation is a way of delaying full chromatin exclusion towards the end of prophase/prometaphase.

Mitotic exit: governing CUL4B dephosphorylation Following phosphorylation upon mitotic entry, our observations identify telophase as the point that CUL4B comes back to the chromatin (Figure 2), hence this is most likely the stage when it’s

3. CRL4B repurposing by mitotic phosphorylation dephosphorylated. It is well established that the activity of CDK1-cyclin B drops due to cyclin B degradation at anaphase onset (Gavet and Pines, 2010). Other kinases, such as PLK1, remain active until later stages (Zitouni et al., 2014). Nevertheless, we believe that the rapid recruitment of CUL4B to the chromatin at telophase (but not in anaphase, including late anaphase) represents active dephosphorylation instead of simply unmaintained phosphorylation events. The bulk of dephosphorylation during mitotic exit, which restores the phosphorylation status of proteins back to their low interphase level, is made by PP2A and PP1 phosphatases. PP2A and PP1 complexes are kept inactive from mitotic entry to anaphase onset due to high CDK1 activity (Wieser and Pines, 2015; de Castro et al., 2016). We presume that these phosphatases (both or individually) likewise act on CUL4B. In the event of dephosphorylation by these phosphatases, it would remain to clarify why doesn’t CUL4B re-bind the chromatin during anaphase, but only does so in telophase. This timing difference is not incompatible with a model where CUL4B is dephosphorylated by PP2A or PP1. In fact, a recent report revealed that mitotic PP2A-B55 substrates (B55 is a type B regulatory subunit) are dephosphorylated sequentially, based on a preference of the phosphatase for phosphothreonine residues and flanking basic regions (Cundell et al., 2016; Godfrey et al., 2017). PP2A-B55 first dephosphorylates spindle proteins, in early anaphase, to which follows the dephosphorylation of nuclear envelope and nuclear pore proteins, towards telophase. Perhaps PP1 acts similarly, and better recognition sites are dephosphorylated first. A second effect might be the hyperphosphorylation of CUL4B, where the sequential removal of multiple phospho-groups is necessary to allow the cullin to re-bind the chromatin.

The mechanism of phosphorylation-driven chromatin unbinding Our data provides a direct link between CUL4B N-terminal phosphorylation and its residence at the chromatin. CUL4B chromatin recruitment shows dependence on DDB1, likely because DDB1 links the cullin with chromatin-binding substrate adaptors such as CDT2, DDB2, WDR5 or RBBP7/4 (Hannah and Zhou, 2015; Mouysset et al., 2015; Hu et al., 2012). Thus, while we initially thought that CUL4B N-terminal phosphorylation disrupts its N-terminal interaction with DDB1, we observed that this is not the case (Supplementary Figure S9). Mitotic-specific phosphorylation of other proteins has been associated with their chromatin exclusion. As an example, the replication licensing factor ORC1 does not bind DNA during mitosis, after being hyperphosphorylated (mediated by CDK1), and concomitantly returns to the chromatin during mitotic exit in order to function in the establishment of origins of replication (Li et al., 2004). Another

3. CRL4B repurposing by mitotic phosphorylation example is the cAMP response element-binding protein (CREB) transcription factor, whose mitotic CDK1- dependent phosphorylation reduces its intrinsic DNA binding potential (Trinh et al., 2013). How CUL4B hyperphosphorylation impacts its binding to the chromatin is unknown. Although there is a regional bias for region 0 and cluster 2, rendering these stretches unphosphorylatable does not guarantee that CUL4B remains at the chromatin (Supplementary Figure S10C). This observation suggests that there is no strict sequence specificity. Perhaps the proximity of a highly phosphorylated stretch impairs DNA binding by CRL4B, simply due to the local abundance of negative charges. Because clusters more towards the N-terminus appear more relevant for chromatin exclusion (Figure 3D), this implies that these regions come closer to DNA.

Phosphorylated CUL4B is repurposed for a distinct mitotic function We show that the knockout of CUL4B affects cell growth, and particularly mitotic entry. Our observations are in line with previously observed G2/M delays in rat neural progenitor and human NT-2 cells treated with CUL4B shRNA (Liu et al., 2012a). Together with mitotic-specific phosphorylation, our discoveries hint that phosphorylated CUL4B takes up a distinct mitotic function that cannot be compensated by CUL4A. This is further supported by the observation that CUL4B may indeed assemble CRL4B complexes in mitosis, as judged by DDB1 binding (Supplementary Figure S9). Such regulatory mechanism resembles the mitotic- specific of the APC/CCDC20, which also becomes active upon phosphorylation by CDK1-cyclin B and PLK1 (Kramer et al., 2000; Golan et al., 2002; Qiao et al., 2016; Zhang et al., 2016; Fujimitsu et al., 2016). We expected that nonphosphorylatable CUL4B-expressing cells cannot progress normally through the initial stages of mitosis. Indeed, this seems to be the case and for reasons yet to identify, these cells have a somewhat impaired metaphase-to-anaphase progression (Figure 5, Supplementary Figure S15). This provides evidence that the phosphorylated, chromatin-unbound CUL4B is indeed important for G2/M and/or the initial stages of mitosis. So far we cannot specify the relative individual contributions of phosphorylation and chromatin exclusion on the new function of CRL4B. Because CUL4B is nevertheless found outside of the chromatin context (Figure 2), we speculate that phosphorylation alone is responsible for the new mitotic role, and the incapacitation to integrate the chromatin pool aids in this function.

3. CRL4B repurposing by mitotic phosphorylation

In line with these conclusions, we postulated that a phosphomimetic mutant would have no issue in G2/M progression. While this does seem to be the case, to our surprise we observed that overexpression of this mutant has a lethal effect provoked by issues in the later stages of mitosis, thereby hindering karyokinesis (Figure 5). The observation that this phenotype is dominant (in the presence of the endogenous) provides further evidence that CUL4B phosphorylation is indeed involved on a new function that begins at G2/M and is terminated in telophase (See Figure 6 for a model that integrates our findings). It also indicates that this function must end by telophase, otherwise karyokinesis cannot be adequately executed. We also observe impaired chromosome congression at metaphase upon overexpression of such mutants. Together with the phenotype of prolonged metaphase upon expression of nonphosphrylatable mutants, these results might indicate that CUL4B is important for chromosome alignment and/or the regulation of the spindle assembly checkpoint. Nevertheless, anaphase progression does not seem

Figure 6. Proposed mode of regulation and action of CUL4B throughout mitosis. Our data suggests that CUL4B phosphorylation during prophase-prometaphase creates a CRL4B population that cannot bind the chromatin and takes up a new regulatory role by ubiquitylating a new substrate. CUL4B phosphorylation is performed by early mitotic kinases (CDK1-cyclin B1, PLK1 and perhaps others) and supposedly counteracted by PP1 and/or PP2A, which are activated during anaphase. Black lines/grey cloud depict condensed and uncondensed chromosomes, respectively. Perpendicular green cilinders depict the centrosome. Green lines, microtubules. Dashed line, nucleus undergoing nuclear envelope breakdown (NEBD) or reformation. Red circles, ubiquitin.

3. CRL4B repurposing by mitotic phosphorylation impaired, and thus one would expect the formation of two nuclei, even if aneuploidy, which is not the case – thereby unveiling a function specific at mitotic exit. How these two observations are correlated is at present unknown. We find strong parallels with the timing of CUL4B phosphorylation/chromatin binding and the observed phenotypes: CUL4B knockout and depletion or lack of phosphorylation impair adequate early mitotic progression at the point where CUL4B is phosphorylated/removed from chromatin. Concomitantly, the phosphomimetic mutant prevents proper progression at later stages, i.e. telophase, where we find CUL4B to re-bind the chromatin and hence dephosphorylated. In other words, phosphorylated CUL4B takes up a new function in mitotic entry that is maintained until telophase, and no further (Figure 6).

A novel function. But which? Additional clarification is required to precisely identify which novel function CUL4B performs. In accordance with the timings at which phenotypes are observed, we hypothesize that CUL4B is important for the execution of an early mitotic event that must be obliterated in telophase, possibly chromatin condensation, nuclear envelope breakdown (NEBD) or the removal of cohesin from chromosomal arms. NEBD is a multi-stage process that occurs from prophase to early prometaphase and involves the disassembly of nuclear pore complexes and nuclear lamina and finally retraction of the nuclear envelope membranes into the ER (Fernández-Álvarez and Cooper, 2017; Ungricht and Kutay, 2017). Because these processes are tightly coordinated by post-translational modifications of the effector proteins, in particular phosphorylation, perhaps ubiquitylation by CUL4B represents an additional regulatory mechanism. Chromatin condensation also requires post-translational modifications, in particular histone phosphorylation (Antonin and Neumann, 2016). The removal of cohesin from chromosomal arms takes place in a phosphorylation-dependent manner (Leman and Noguchi, 2014). These processes share the peculiarity that they must be reversed in telophase (Ungricht and Kutay, 2017; Antonin and Neumann, 2016; Gerlich et al., 2006), and we could envisage that the constitutively active phosphomimetic CUL4B mutant would prevent their execution. One possibility is that CUL4B acts upstream of these processes, for example by itself regulating the kinases or phosphatases that orchestrate these mitotic processes (establishing double feedback loops which are in fact a hallmark of cell cycle transitions (Barr et al., 2016; Kim and Ferrell, 2007)). Indeed, PP1 and PP2A are central in the occurrence of these late mitotic events (Wieser and Pines, 2015; de Castro et al., 2016; Cundell et al., 2016). Concomitantly, the inactivation of these phosphatases is important for the early mitotic stages

3. CRL4B repurposing by mitotic phosphorylation

(Krasinska et al., 2011). PP2A absence is important for G2/M, as it promotes CDK1-cyclin B inhibition by dephosphorylating activatory CDC25 sites and promoting WEE1 activation (Forester et al., 2007). CUL4B- dependent PP2A inactivation would perhaps be more fitting to the observation that CUL4B regulates the G2/M transition itself (Figure 4). In regards to mitotic kinase regulation, one hypothesis would be if CUL4B phosphorylation positively regulates PLK1. PLK1 is a master orchestrator of the G2/M progression, chromatin condensation, chromosome congression and mitotic exit (Zitouni et al., 2014). Hence, CUL4B- mediated regulation of PLK1 could explain the pleiotropic phenotypes that we observe upon CUL4B deletion and the overexpression of the phosphomimetic mutant.

CUL4B mitotic role: the missing link between CUL4B loss-of-function and XLID? Despite much research, the demonstration of a CUL4B-specific role that justifies why loss-of- function mutations result in XLID is still lacking. While there have been reports of CUL4B involvement in gene expression important neuronal cell differentiation (Nakagawa and Xiong, 2011; Zhao et al., 2015), we believe that the link between CUL4B loss-of-function and syncromic XLID still remains unclarified, and in particular why CUL4A cannot compensate for lack of CUL4B. In the case of our study, we describe a function of CUL4B that indeed cannot be compensated by CUL4A, regardless of its expression of subcellular localization, because it resides within the non-conserved region between the two cullins. Indeed, we found most puzzling that the N-terminal missense mutation P50L, located in this unique CUL4B region, results in XLID with severe phenotypes (Vulto-van Silfhout et al., 2015). The reasoning for this effect is still unknown, because the P50L mutation does not affect the stability of CUL4B like other disease-related mutations do (Supplementary Figure S4; Vulto-van Silfhout et al., 2015; Nakagawa and Xiong, 2011), and is not located at any of the known vital regions of CUL4B (Figure 1A). It had been anticipated that the nuclear localization of CUL4B is disrupted (Vulto-van Silfhout et al., 2015), but we show that this is not the case. Related to our report of a specific role related to the N-terminus of CUL4B, future research should clarify whether P50L results in impaired mitotic function of CUL4B. While we did not observe obvious growth defects of HeLa cells in colony assays, which was also the case for the nonphosphorylatable mutants, it is possible that a mild deregulation of this function has amplified effects in early developmental stages. This would also mean that this newly-identified role is the key role whose loss impairs adequate development of the brain and other organs. One exciting hypothesis is that the P50L mutation disrupts the T49 CDK1 phosphorylation site in cluster 2, thereby affecting the mitotic function of the cullin. While we did not see an overall change in the hyperphosphorylation pattern of the

3. CRL4B repurposing by mitotic phosphorylation

P50L mutant in nocodazole-arrested cells, it is possible that the kinetics of phosphorylation in normally cycling cells is impaired, but not overall phosphorylation if cells are arrested. The possibility that CUL4B loss-of-function-mediated XLID is due to the loss of a mitotic function of CRL4B also implies that loss-of-function of the yet-to-identify substrate receptor has implications in intellectual disability. Indeed, mutations in the genes of several putative CRL4 substrate receptors has been associated with intellectual disability (CRBN, DCAF17, KATNB1, LIS1, WDR62, PHIP) (Jin et al., 2006; de Ligt et al., 2012; Vissers et al., 2015). Besides, mutations in the genes of mitotic proteins have been associated with intellectual disability (ID), in particular genes involved in chromosome alignment or the spindle assembly checkpoint (BUB1B, CDK5RAP2, USP9X, CHAMP1, POGZ, KIF11, NSUN2). Remarkably, mutations in cohesin subunits (SMC1, SMC3, SCC1/RAD21, STAG2) and regulators (NIPBL, HDAC8, ESCO2, PP2A scaffold A subunit PR65), whose normal gene functions are also important for proper chromosome segregation and mitotic progression (Sonoda et al., 2001; Tang et al., 2006), are similarly connected to ID (Vissers et al., 2015; Mullegama et al., 2017). This renders plausible that CUL4B mutations result in syndromic XLID due to the disruption of a mitotic function, and unveils the possibility that CUL4B acts on the same pathways as these factors. In the future, we will attempt to identify whether the indicated ID- related DCAF proteins are indeed mitotic CRL4B-specific substrate adaptors that act in a concerted manner with CUL4B to regulate mitotic progression. Besides, perhaps the function of CUL4B resides in the same pathway of other proteins described in ID, which will be explored in light with the proposed functions discussed above. In summary, our study establishes a novel elegant regulatory mechanism of a CRL complex, via cell-cycle regulated phosphorylation of the cullin scaffold. We provide evidence that CRL4B is a novel mitotic E3 ligase, acting from mitotic entry to mitotic exit, and show that phosphorylation by CDK1 and other kinases is directly correlated to CUL4B chromatin occupancy. Future research will determine the molecular mechanisms how CUL4B, but not CUL4A, regulates mitosis, i.e. which substrates are ubiquitylated by this complex and what is the identity of the mitotic CRL4B complex. We believe that such analysis will shed light on what are the molecular mechanisms behind lethality in mice knockout embryos and syndromic X-linked intellectual disability in human patients.

Materials and methods Antibodies used in this study. For western blot, flow cytometry and immunofluorescence analysis, we used the following commercial and non-commercial primary antibodies: Anti-Coilin (GeneTex); Anti-CREST antiserum (Antibodies Incorporated); Anti-CUL4 antibodies (rabbit) (Olma et al., 2009); Anti-DDB1

3. CRL4B repurposing by mitotic phosphorylation

(mouse, BD Biosciences); Anti-FLAG (mouse or rabbit, both Sigma); Anti-GAPDH (mouse, Sigma); Anti-HA (mouse or rabbit, both Covance); Anti-Phospho Histone H2A.X S139 (mouse, Millipore); Anti-Phospho Histone H3 S10 (rabbit, Upstate); Anti-Histone H4 (rabbit, Abcam); Anti-Nedd8 (rabbit, Abcam); Anti- Tubulin (mouse, Sigma); Anti-SF2 (GeneTex); Secondary antibodies: Anti-mouse or rabbit IgG-HRP (both Biorad) and anti-mouse or rabbit IgG-Alexa 488 and 568 (both Invitrogen)

Cell culture, RNAi, double thymidine block-release. Cells were cultured in DMEM (Gibco) supplemented with 10% FCS and penicillin-streptomycin-glutamine (Gibco) in an incubator at 37ºC with 5% CO2. Cells were arrested in mitosis with either 100 ng/µL nocodazole (Sigma) or taxol (Paclitaxel, Sigma), as indicated. G2 cell cycle arrest was performed with 10 µM RO3306 (Calbiochem), and S phase arrest with 2 mM thymidine (Sigma). Double-thymidine block-release experiments were performed by arresting cells in S phase with 2 mM thymidine for 16h, after which cells were washed 2x with PBS (137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4 pH 7.4) and 2x with DMEM and allowed to resume the cell cycle for 9h. 2 mM thymidine was subsequently re-added for 16h to arrest cells at the G1/S border or early S phase, followed by new washes and release. Cells were harvested at the indicated time points post- release by trypsinization (for flow cytometry) or by addition of Laemmli buffer (for SDS-PAGE analysis). For RNAi experiments, cells were transfected with 25 nM double stranded siRNA (Microsynth) using Lipofectamine RNAiMax (Invitrogen) according to the manufacturer’s instructions. siRNA used: siCUL4B #1 (5’-CCGTCTCTAGCTTTGCTAA), siCUL4B #2 (5’- AGCGCCTGTTAGTCGGAAA) and AllStars Negative Control siRNA (Qiagen).

Plasmid transient transfection, cycloheximide chase and cell fractionation. Human CUL4A and CUL4B cDNA were PCR-amplified from total cell cDNA and cloned into a pcDNA5/FRT/TO vector (Invitrogen). CUL4B N- terminal nonphosphorylatable and phosphomimetic DNA was synthesized (Genscript) and subsequently subcloned into pcDNA5. Cells were transfected using lipofectamine 3000 (Invitrogen; according to the manufacturer’s instructions). For cycloheximide chase experiments, cells were transfected with pcDNA5 vectors expressing the indicated construct and split into multiwells 24h after. 48h post-transfection, 100 µg/mL cycloheximide (Sigma) was added and cells harvested at the indicated time points by addition of Laemmli buffer and analyzed by western blot. For cell fractionation, 4 million cells were harvested by trypsinization 48h after transfection and nuclei isolated using the NE-PER Nuclear and Cytoplasmic Extraction Reagents (Thermo Fischer), according to the manufacturer’s instructions. Subsequently, nuclei were resuspended in a buffer containing 10 mM Hepes-KOH pH 7.7, 100 mM KCl, 50 µM sucrose, 0.25% Triton X-100 and protease inhibitor cocktail, and

3. CRL4B repurposing by mitotic phosphorylation lysed with a 27G syringe. Nuclear lysates were then loaded on top of a 1M sucrose cushion and centrifuged at 8000g for 30 min. The pellet and soluble fractions were collected and boiled in Laemmli buffer, followed by SDS-PAGE and Western blotting.

Lentiviral infection and clonogenic assays. For lentiviral vector production we subcloned untagged CUL4B constructs into a pTRIPZ vector (GE Dharmacon), replacing the turboRFP gene with the gene of interest. 10 million HEK293T cells were co-transfected with the pTRIPZ vector and the envelope pMD2.G and packaging psPAX2 plasmids. The latter vectors were a kind gift from Didier Trono (Addgene plasmid #12259 and #12260, respectively). 4h post-transfection, the medium was replaced and the lentivirus- containing medium was collected 48h later, filtered through a 0.45 µm pore filter and kept at 4ºC. For clonogenic assays, 700 cells of the indicated cell lines or cells infected for 4h with lentivirus expressing the indicated construct were seeded in 6-well cell culture plates and colonies were allowed to grow. 6 days post-infection, cells were left untreated or 1 µg/mL puromycin (Gibco) was added until 11 days post- transfection. At this point, cells were washed with PBS and stained with 0.25 % (w/v) crystal violet (Sigma) in 80% (v/v) methanol for 30 min, followed by 2 washes with water. Plates were allowed to air-dry.

Live-cell imaging. For imaging in freshly lentivirus-infected cells, 10 000 cells were seeded in µClear bottom 96-well microplates (Greiner Bio-One) and infected the day after for 5h with the indicated lentivirus in the presence of polybrene (Sigma). For live-cell imaging using stably-expressing cell lines, cells were infected and selected with puromycin for at least 1 week. 10 000 cells of the indicated cell line were treated with CUL4B siRNA and expression was induced with 1 µg/mL doxycycline 48h before imaging. 25 000 cells were seeded in a 8-well µ-Slide chamber coated with ibiTreat (Ibidi). In all cases, 1h before imaging the cell medium was replaced with FluoroBrite DMEM media (Gibco) supplemented with 10% FCS and penicillin- streptomycin-glutamine, as well as 1 µM SiR-DNA, 1 µM verapamil (both Spirochrome) and 1 µg/mL doxycycline. Cells were imaged in Eclipse Ti epifluorescence microscope (Nikon Instruments, Melville, NY),

with controlled temperature, CO2 and humidity, using a 10x objective (NA 0.45 CFI Plan Apochromat). Images were acquired every 30 min (for freshly infected cells) or every 20 min (for cell lines).

Flow cytometry. Cells were harvested with 0.5% Trypsin solution (Thermo Fischer), washed with PBS after centrifugation at 500g for 5 min and then resuspended in 70% ethanol solution pre-cooled to -20ºC. After an incubation at -20ºC for min. 2h, cells were centrifuged, washed 2x with PBS supplemented with 0.05% Triton X-100 (Sigma) and resuspended in blocking buffer (PBS, 0.25% Triton X-100, 5% FCS) for 30 min. Cells were incubated for 1h in blocking buffer supplemented with the indicated antibody, washed 2x with washing buffer (PBS, 0.1% Triton X-100, 5% FCS) and incubated for 1h in blocking buffer supplemented

3. CRL4B repurposing by mitotic phosphorylation with the appropriate secondary anti-IgG antibody, conjugated to Alexa 488. Following 2x washes with washing buffer, cells were resuspended in PBS supplemented with 50 µg/mL propidium iodide (PI; Sigma) and 20 µg/mL RNAse A (Thermo Fischer). Flow cytometry was carried out in a FACScalibur instrument (BD Biosciences). If no antibody staining was performed, blocking and antibody incubation steps were omitted.

Immunofluorescence and western blotting. For immunofluorescence experiments, cells previously seeded in coverslips were fixed with cold methanol or room-temperature 4% formalin for 10 min with or without prior incubation with extraction buffer (20 mM PIPES pH 6.8, 10 mM EGTA, 1 mM MgCl2 and 0.25% Triton X-100) for 4 min. If not subject to pre-extraction, cells were subsequently permeabilized with 0.25% Triton X-100 in PBS. Cells were washed with PBS supplemented with 0.05% Triton X-100 and incubated in blocking buffer (PBS, 0.05% Triton X-100, 5% FCS) for 30 min. Cells were subsequently incubated for 1h with antibodies directed against the indicated antigens prepared in blocking buffer. Cells were washed 3x and incubated for 1h with secondary anti-mouse IgG and anti-rabbit IgG antibodies, conjugated to Alexa 488 and Alexa 568, respectively. Cells were again washed 3x, and 0.1 µg/mL DAPI (Sigma) was added during the second wash. Coverslips were mounted in microscope slides using Shandon Immu-mount (Thermo Scientific). Z-stacks of cells were acquired in 0.3 µm steps in an Eclipse Ti epifluorescence microscope using a 60x oil objective (NA 1.4 Plan Apochromat). For western blotting, samples were boiled in 1x Laemmli buffer, followed by SDS-PAGE and transfer to PVDF membranes using the Tris-glycine buffer system. PVDF membranes were incubated in blocking buffer (5% Skimmed milk in PBS, 0.1% Tween-20), sequentially incubated with the primary antibody diluted in blocking buffer HRP-coupled Anti-Mouse or anti-rabbit IgG diluted 1:5000 for 1h at RT, intercalated by 4 washes with PBS containing 0.1% Tween-20 (PBS-T). After 4 additional washes with PBS- T, membranes were incubated with ECL substrate (Pierce Thermo Fischer) and developed with a Fusion FX system (Vilber Lourmat).

Recombinant protein expression, purification and in vitro kinase assays. FLAG-tagged CUL4 constructs, DDB1 or RBX1 cDNA were subcloned into pFL multiBac vectors, bacmids were produced in DH10Bac/Multibac E. coli bacterial strains and extracted as previously described (Fitzgerald et al., 2006). 2 million log‐phase SF9 cell lines (Invitrogen) were transformed with bacmid DNA using Cellfectin II (Invitrogen) according to the manufacturer's instructions. Infected cells were resuspended lysed in buffer containing 50 mM Tris-HCl pH 8.0, 350 mM NaCl, 0.4% Triton X-100, 1 mM EDTA, 1 mM DTT and protease inhibitor cocktail (Roche), using a Dounce homogeneizer. Lysates was cleared by centrifugation at 12 000

3. CRL4B repurposing by mitotic phosphorylation g for 30 min at 4ºC and loaded into anti-FLAG M2 affinity gel (Sigma). Washes were performed in buffer containing 50 mM Tris-HCl pH 8.0 and 150 mM NaCl and CRL4 complexes were subsequently eluted with 3xFLAG peptide diluted in 50 mM Tris-HCl pH 8.0, 150 mM NaCl and 10% glycerol, frozen in liquid nitrogen and stored at -80ºC until analysis. For in vitro kinase assays, 0.5 µg of recombinant CUL4A or CUL4B (in complex with DDB1 and RBX1) or 0.15 µg native swine myelin basic protein (MBP; SignalChem) were added to 1x protein kinase buffer (NEB) supplemented with 200 µM “cold” ATP (Sigma) and 500 µCi/µmol ATPγ32P (Hartmann Analytic). The reaction was initiated by addition of 10 U (pmol/min) of recombinant CK1 (NEB), CDK1-cyclin B1 (NEB), PLK1 (SignalChem), Aurora A (SignalChem) or Aurora B (SignalChem) kinases, incubated for 1h at 30ºC. Samples were boiled in 1x Laemmli buffer and subjected to SDS-PAGE using the NuPAGE Bis-Tris system with MOPS SDS buffer (Invitrogen). Gels were stained with InstantBlue coomassie stain (Expedeon Protein Solutions), vacuum-dried for 30 min and exposed overnight to a storage phosphor screen (GE Healthcare). Screens were scanned in a Typhoon FLA 9000 system (GE Healthcare)

CRISPR-mediated knock-out and knock-in. For SpCas9-mediated genome editing, DNA oligos (Microsynth) corresponding to the gRNA target sequence were cloned into a pX458 plasmid, kindly provided by F. Zhang (Addgene plasmid #48138), following the indicated protocol (Ran et al., 2013). For all editing experiments, HeLa cells were co-transfected with two pX458 plasmids encoding for SpCas9-2A-GFP and one of two gRNAs, using lipofectamine 3000 transfection reagent (Invitrogen). CUL4B knockout was performed by simultaneous targeting the intron between exons 2 and 3 (target DNA: 5’TCCCGCGGGACCGTTAAGGA) and between exons 4 and 5 (target DNA: 5’TAACGGCTACCTATATGGTA). For U6-mediated transcription of the gRNA sequence, an additional guanine was included upstream (Ran et al., 2013). Four days post- transfection, cells were sorted for GFP in a BD FACSAria IIIu cell sorter (BD Biosciences). After two weeks, individual clones were tested by western blot for CUL4B deletion. Knock-in of the WT CUL4B gene was performed by simultaneous targeting two regions within exon 3 (target DNA 1: GGTGCTGGTATTACCATCAG; target DNA 2: ACTAACTTCTTAGCAGAGCC). The two respective pX458 plasmids were co-transfected with a third plasmid containing the DNA template sequence for homologous recombination. The Puro-P2A-HA-containing template vector (prepared for siRNA-resistance) was cloned into a pUC18 plasmid. Homology regions were PCR-amplified from HeLa cell genomic DNA, and the puromycin N-acetyl-transferase (pac) gene PCR-amplified from a pIRESpuro2 vector (Takara Bio Inc). The regions on the template that can be targeted by the used gRNAs were mutated using QuikChange site- directed mutagenesis with Pfu turbo DNA polymerase (Agilent). Cells were selected with 1 µg/mL puromycin starting 4 days post-transfection and clones were tested by PCR on genomic DNA.

3. CRL4B repurposing by mitotic phosphorylation

Immunoprecipitation and lambda phosphatase treatment. For protein precipitation experiments, approximately 20 million cells expressing HA-CUL4B, or transfected with FLAG-CUL4B-encoding pcDNA5 were harvested with a rubber policeman, washed once with PBS and lysed for 20 min using a 27G needle syringe in IP buffer supplemented with 125 U pierce universal nuclease (Thermo Fischer). For FLAG IPs, this buffer is composed of 20 mM Tris-HCl pH 7.0, 150 mM NaCl, 2 mM MgCl2, 10% glycerol, 0.5% NP-40,

10 mM NaF, 10 mM β-glycerophosphate, 0.2 mM NaVO4, 1 mM PMSF, whereas to precipitate HA-tagged

CUL4B, we used buffer containing 20 mM HEPES pH 7.9, 100 mM KCl, 2 mM MgCl2, 0.5% NP40, 300 mM sucrose, 10 mM NaF, 10 mM β-glycerophosphate, 0.2 mM NaVO4 and 1 mM PMSF. Following centrifugation for 30 min at 16000g, supernatants were incubated with anti-FLAG M2 affinity gel (Sigma) or anti-HA-agarose beads (Sigma; clone HA-7) for 1h at 4ºC. The supernatant was then discarded and the beads were washed 5x with IP buffer. For analysis of interactions, proteins were eluted with 1 mg/mL 3x FLAG peptide if bait is FLAG-tagged, or eluted with 100 mM glycine pH 1.8 for HA-CUL4B. For lambda phosphatase treatment, we did not perform elution at this stage and instead the beads were resuspended in 1x buffer for protein metallophosphatases (PMP buffer, NEB) supplemented with 1 mM MnCl2 (NEB) and split in two tubes. Either 200 U λ-Phosphatase or PMP buffer (control) were added, following incubation for 1h at 30ºC. Proteins were eluted with 1x Laemmli buffer. For λ-Phosphatase treatment of whole-cell lysates, 4 million cells were lysed described, in 1x PMP buffer supplemented with 1 mM MnCl2, 0.5 % NP40, 0.5 mM PMSF and protease inhibitor cocktail. Lysates were split in two tubes and either 200 U λ-Phosphatase or lysis buffer were added, following incubation for 1h at 30ºC.

Acknowledgements We thank Feng Zhang for sharing vectors, the Flow Cytometry Core Facility at ETH for experimental support and members of the Peter lab for helpful discussions. This work was funded by the Eidgenössische Technische Hochschule Zürich (ETH-20 14-1 and ETH-46 16-1), the Swiss National Science Foundation (SNF 310030B_160321/1), and the European Research Council (ERC 268930 Rubinet).

3. CRL4B repurposing by mitotic phosphorylation

References Antonin, W., and H. Neumann. 2016. Chromosome condensation and decondensation during mitosis. Curr. Opin. Cell Biol. 40:15–22. doi:10.1016/j.ceb.2016.01.013. Badura-Stronka, M., A. Jamsheer, A. Materna-Kiryluk, A. Sowińska, K. Kiryluk, B. Budny, and A. Latos- Bieleńska. 2010. A novel nonsense mutation in CUL4B gene in three brothers with X-linked mental retardation syndrome. Clin. Genet. 77:141–144. doi:10.1111/j.1399-0004.2009.01331.x. Barr, A.R., F.S. Heldt, T. Zhang, C. Bakal, and B. Novák. 2016. A Dynamical Framework for the All-or-None G1/S Transition. Cell Syst. 2:27–37. doi:10.1016/j.cels.2016.01.001. Bassermann, F., R. Eichner, and M. Pagano. 2014. The ubiquitin proteasome system - implications for cell cycle control and the targeted treatment of cancer. Biochim. Biophys. Acta. 1843:150–162. doi:10.1016/j.bbamcr.2013.02.028. Burma, S., B.P. Chen, M. Murphy, A. Kurimasa, and D.J. Chen. 2001. ATM phosphorylates histone H2AX in response to DNA double-strand breaks. J. Biol. Chem. 276:42462–7. doi:10.1074/jbc.C100466200. Cabezas, D.A., R. Slaugh, F. Abidi, J.F. Arena, R.E. Stevenson, C.E. Schwartz, and H.A. Lubs. 2000. A new X linked mental retardation (XLMR) syndrome with short stature, small testes, muscle wasting, and tremor localises to Xq24-q25. J. Med. Genet. 37:663–8. de Castro, I.J., E. Gokhan, and P. Vagnarelli. 2016. Resetting a functional G1 nucleus after mitosis. Chromosoma. 125:607–619. doi:10.1007/s00412-015-0561-6. Chen, C.-Y., M.-S. Tsai, C.-Y. Lin, I.-S. Yu, Y.-T. Chen, S.-R. Lin, L.-W. Juan, Y.-T. Chen, H.-M. Hsu, L.-J. Lee, and S.-W. Lin. 2012. Rescue of the genetically engineered Cul4b mutant mouse as a potential model for human X-linked mental retardation. Hum. Mol. Genet. 21:4270–85. doi:10.1093/hmg/dds261. Cundell, M.J., L.H. Hutter, R. Nunes Bastos, E. Poser, J. Holder, S. Mohammed, B. Novak, and F.A. Barr. 2016. A PP2A-B55 recognition signal controls substrate dephosphorylation kinetics during mitotic exit. J. Cell Biol. 214:539–554. doi:10.1083/jcb.201606033. Elia, A.E.H., P. Rellos, L.F. Haire, J.W. Chao, F.J. Ivins, K. Hoepker, D. Mohammad, L.C. Cantley, S.J. Smerdon, and M.B. Yaffe. 2003. The molecular basis for phosphodependent substrate targeting and regulation of Plks by the Polo-box domain. Cell. 115:83–95. doi:10.1016/S0092-8674(03)00725-6. Fernández-Álvarez, A., and J.P. Cooper. 2017. Chromosomes Orchestrate Their Own Liberation: Nuclear Envelope Disassembly. Trends Cell Biol. 27:255–265. doi:10.1016/j.tcb.2016.11.005. Fischer, E.S., A. Scrima, K. Böhm, S. Matsumoto, G.M. Lingaraju, M. Faty, T. Yasuda, S. Cavadini, M. Wakasugi, F. Hanaoka, S. Iwai, H. Gut, K. Sugasawa, and N.H. Thomä. 2011. The molecular basis of CRL4 DDB2/CSA ubiquitin ligase architecture, targeting, and activation. Cell. 147:1024–1039. doi:10.1016/j.cell.2011.10.035. Fitzgerald, D.J., P. Berger, C. Schaffitzel, K. Yamada, T.J. Richmond, and I. Berger. 2006. Protein complex expression by using multigene baculoviral vectors. Nat. Methods. 3:1021–1032. doi:10.1038/nmeth983. Forester, C.M., J. Maddox, J. V. Louis, J. Goris, and D.M. Virshup. 2007. Control of mitotic exit by PP2A regulation of Cdc25C and Cdk1. Proc. Natl. Acad. Sci. 104:19867–19872. doi:10.1073/pnas.0709879104. Fujimitsu, K., M. Grimaldi, and H. Yamano. 2016. Cyclin-dependent kinase 1-dependent activation of

3. CRL4B repurposing by mitotic phosphorylation

APC/C ubiquitin ligase. Science. 352:1121–4. doi:10.1126/science.aad3925. Gavet, O., and J. Pines. 2010. Progressive Activation of CyclinB1-Cdk1 Coordinates Entry to Mitosis. Dev. Cell. 18:533–543. doi:10.1016/j.devcel.2010.02.013. Gerlich, D., B. Koch, F. Dupeux, J.-M. Peters, and J. Ellenberg. 2006. Live-Cell Imaging Reveals a Stable Cohesin-Chromatin Interaction after but Not before DNA Replication. Curr. Biol. 16:1571–1578. doi:10.1016/j.cub.2006.06.068. Gilberto, S., and M. Peter. 2017. Dynamic ubiquitin signaling in cell cycle regulation. J. Cell Biol. doi:10.1083/jcb.201703170. Godfrey, M., S.A. Touati, M. Kataria, A. Jones, A.P. Snijders, and F. Uhlmann. 2017. PP2ACdc55 Phosphatase Imposes Ordered Cell-Cycle Phosphorylation by Opposing Threonine Phosphorylation. Mol. Cell. 65:393–402.e3. doi:10.1016/j.molcel.2016.12.018. Golan, A., Y. Yudkovsky, and A. Hershko. 2002. The Cyclin-Ubiquitin Ligase Activity of Cyclosome/APC Is Jointly Activated by Protein Kinases Cdk1-Cyclin B and Plk. J. Biol. Chem. 277:15552–15557. doi:10.1074/jbc.M111476200. Grozeva, D., K. Carss, O. Spasic-Boskovic, M.-I. Tejada, J. Gecz, M. Shaw, M. Corbett, E. Haan, E. Thompson, K. Friend, Z. Hussain, A. Hackett, M. Field, A. Renieri, R. Stevenson, C. Schwartz, J.A.B. Floyd, J. Bentham, C. Cosgrove, B. Keavney, S. Bhattacharya, M. Hurles, and F.L. Raymond. 2015. Targeted Next-Generation Sequencing Analysis of 1,000 Individuals with Intellectual Disability. Hum. Mutat. 36:1197–204. doi:10.1002/humu.22901. Hannah, J., and P. Zhou. 2015. Distinct and overlapping functions of the cullin E3 ligase scaffolding proteins CUL4A and CUL4B. Gene. 573:33–45. doi:10.1016/j.gene.2015.08.064. Hornbeck, P. V., B. Zhang, B. Murray, J.M. Kornhauser, V. Latham, and E. Skrzypek. 2015. PhosphoSitePlus, 2014: mutations, PTMs and recalibrations. Nucleic Acids Res. 43:D512–D520. doi:10.1093/nar/gku1267. Hu, H., Y. Yang, Q. Ji, W. Zhao, B. Jiang, R. Liu, J. Yuan, Q. Liu, X. Li, Y. Zou, C. Shao, Y. Shang, Y. Wang, and Y. Gong. 2012. CRL4B Catalyzes H2AK119 Monoubiquitination and Coordinates with PRC2 to Promote Tumorigenesis. Cancer Cell. 22:781–795. doi:10.1016/j.ccr.2012.10.024. Isidor, B., O. Pichon, S. Baron, A. David, and C. Le Caignec. 2010. Deletion of the CUL4B gene in a boy with mental retardation, minor facial anomalies, short stature, hypogonadism, and ataxia. Am. J. Med. Genet. Part A. 152:175–180. doi:10.1002/ajmg.a.33152. Jiang, B., W. Zhao, J. Yuan, Y. Qian, W. Sun, Y. Zou, C. Guo, B. Chen, C. Shao, and Y. Gong. 2012. Lack of Cul4b, an E3 ubiquitin ligase component, leads to embryonic lethality and abnormal placental development. PLoS One. 7. doi:10.1371/journal.pone.0037070. Jin, J., E.E. Arias, J. Chen, J.W. Harper, and J.C. Walter. 2006. A Family of Diverse Cul4-Ddb1-Interacting Proteins Includes Cdt2, which Is Required for S Phase Destruction of the Replication Factor Cdt1. Mol. Cell. 23:709–721. doi:10.1016/j.molcel.2006.08.010. Kamenz, J., and J.E. Ferrell. 2017. The Temporal Ordering of Cell-Cycle Phosphorylation. Mol. Cell. 65:371– 373. doi:10.1016/j.molcel.2017.01.025. Kerzendorfer, C., L. Hart, R. Colnaghi, G. Carpenter, D. Alcantara, E. Outwin, A.M. Carr, and M. O’Driscoll. 2011. CUL4B-deficiency in humans: Understanding the clinical consequences of impaired Cullin 4-

3. CRL4B repurposing by mitotic phosphorylation

RING E3 ubiquitin ligase function. Mech. Ageing Dev. 132:366–373. doi:10.1016/j.mad.2011.02.003. Kim, J.H., S.R. Lee, L.H. Li, H.J. Park, J.H. Park, K.Y. Lee, M.K. Kim, B.A. Shin, and S.Y. Choi. 2011. High cleavage efficiency of a 2A peptide derived from porcine teschovirus-1 in human cell lines, zebrafish and mice. PLoS One. 6:1–8. doi:10.1371/journal.pone.0018556. Kim, S.Y., and J.E. Ferrell. 2007. Substrate Competition as a Source of Ultrasensitivity in the Inactivation of Wee1. Cell. 128:1133–1145. doi:10.1016/j.cell.2007.01.039. Kopanja, D., N. Roy, T. Stoyanova, R.A. Hess, S. Bagchi, and P. Raychaudhuri. 2011. Cul4A is essential for spermatogenesis and male fertility. Dev. Biol. 352:278–287. doi:10.1016/j.ydbio.2011.01.028. Kramer, E.R., N. Scheuringer, A. V Podtelejnikov, M. Mann, and J.M. Peters. 2000. Mitotic regulation of the APC activator proteins CDC20 and CDH1. Mol. Biol. Cell. 11:1555–69. doi:10.1091/mbc.11.5.1555. Krasinska, L., M.R. Domingo-Sananes, O. Kapuy, N. Parisis, B. Harker, G. Moorhead, M. Rossignol, B. Novák, and D. Fisher. 2011. Protein Phosphatase 2A Controls the Order and Dynamics of Cell-Cycle Transitions. Mol. Cell. 44:437–450. doi:10.1016/j.molcel.2011.10.007. Lee, J., and P. Zhou. 2007. DCAFs, the Missing Link of the CUL4-DDB1 Ubiquitin Ligase. Mol. Cell. 26:775– 780. doi:10.1016/j.molcel.2007.06.001. Leman, A.R., and E. Noguchi. 2014. Linking chromosome duplication and segregation via sister chromatid cohesion. Methods Mol. Biol. 1170:75–98. doi:10.1007/978-1-4939-0888-2_5. Li, C. -j., A. Vassilev, and M.L. DePamphilis. 2004. Role for Cdk1 (Cdc2)/Cyclin A in Preventing the Mammalian Origin Recognition Complex’s Largest Subunit (Orc1) from Binding to Chromatin during Mitosis. Mol. Cell. Biol. 24:5875–5886. doi:10.1128/MCB.24.13.5875-5886.2004. de Ligt, J., M.H. Willemsen, B.W.M. van Bon, T. Kleefstra, H.G. Yntema, T. Kroes, A.T. Vulto-van Silfhout, D.A. Koolen, P. de Vries, C. Gilissen, M. del Rosario, A. Hoischen, H. Scheffer, B.B.A. de Vries, H.G. Brunner, J.A. Veltman, and L.E.L.M. Vissers. 2012. Diagnostic Exome Sequencing in Persons with Severe Intellectual Disability. N. Engl. J. Med. 367:1921–1929. doi:10.1056/NEJMoa1206524. Liu, H.C., G. Enikolopov, and Y. Chen. 2012a. Cul4B regulates neural progenitor cell growth. BMC Neurosci. 13:112. doi:10.1186/1471-2202-13-112. Liu, L., Y. Yin, Y. Li, L. Prevedel, E.H. Lacy, L. Ma, and P. Zhou. 2012b. Essential role of the CUL4B ubiquitin ligase in extra-embryonic tissue development during mouse embryogenesis. Cell Res. 22:1258–1269. doi:10.1038/cr.2012.48. Livneh, I., V. Cohen-Kaplan, C. Cohen-Rosenzweig, N. Avni, and A. Ciechanover. 2016. The life cycle of the 26S proteasome: from birth, through regulation and function, and onto its death. Cell Res. 26:869– 885. doi:10.1038/cr.2016.86. Londin, E.R., J. Adijanto, N. Philp, A. Novelli, E. Vitale, C. Perria, G. Serra, V. Alesi, S. Surrey, and P. Fortina. 2014. Donor splice-site mutation in CUL4B is likely cause of X-linked intellectual disability. Am. J. Med. Genet. Part A. 164:2294–2299. doi:10.1002/ajmg.a.36629. Morgan, D.O., and J.M. Roberts. 2002. Cell cycle: Oscillation sensation. Nature. 418:495–496. doi:10.1038/418495a. Mouysset, J., S. Gilberto, M.G. Meier, F. Lampert, M. Belwal, P. Meraldi, and M. Peter. 2015. CRL4RBBP7 is required for efficient CENP-A deposition at centromeres. J. Cell Sci. 7:1732–1745.

3. CRL4B repurposing by mitotic phosphorylation

doi:10.1242/jcs.162305. Mullegama, S. V., S.D. Klein, M. V. Mulatinho, T.N. Senaratne, K. Singh, D.C. Nguyen, N.M. Gallant, S.P. Strom, S. Ghahremani, N.P. Rao, J.A. Martinez-Agosto, and J.A. Martinez-Agosto. 2017. De novo loss- of-function variants in STAG2 are associated with developmental delay, microcephaly, and congenital anomalies. Am. J. Med. Genet. Part A. 173:1319–1327. doi:10.1002/ajmg.a.38207. Nakagawa, T., and Y. Xiong. 2011. X-Linked Mental Retardation Gene CUL4B Targets Ubiquitylation of H3K4 Methyltransferase Component WDR5 and Regulates Neuronal Gene Expression. Mol. Cell. 43:381–391. doi:10.1016/j.molcel.2011.05.033. Nguyen Ba, A.N., A. Pogoutse, N. Provart, and A.M. Moses. 2009. NLStradamus: a simple Hidden Markov Model for nuclear localization signal prediction. BMC Bioinformatics. 10:202. doi:10.1186/1471- 2105-10-202. Ohtake, F., A. Baba, I. Takada, M. Okada, K. Iwasaki, H. Miki, S. Takahashi, A. Kouzmenko, K. Nohara, T. Chiba, Y. Fujii-Kuriyama, and S. Kato. 2007. Dioxin receptor is a ligand-dependent E3 ubiquitin ligase. Nature. 446:562–566. doi:10.1038/nature05683. Okamoto, N., M. Watanabe, T. Naruto, K. Matsuda, T. Kohmoto, M. Saito, K. Masuda, and I. Imoto. 2017. Genome-first approach diagnosed Cabezas syndrome via novel CUL4B mutation detection. Hum. genome Var. 4:16045. doi:10.1038/hgv.2016.45. Olma, M.H., M. Roy, T. Le Bihan, I. Sumara, S. Maerki, B. Larsen, M. Quadroni, M. Peter, M. Tyers, and L. Pintard. 2009. An interaction network of the mammalian COP9 signalosome identifies Dda1 as a core subunit of multiple Cul4-based E3 ligases. J. Cell Sci. 122:1035–44. doi:10.1242/jcs.043539. Piwko, W., M.H. Olma, M. Held, J.N. Bianco, P.G. a Pedrioli, K. Hofmann, P. Pasero, D.W. Gerlich, and M. Peter. 2010. RNAi-based screening identifies the Mms22L-Nfkbil2 complex as a novel regulator of DNA replication in human cells. EMBO J. 29:4210–4222. doi:10.1038/emboj.2010.304. Porter, L.A., and D.J. Donoghue. 2003. Cyclin B1 and CDK1: nuclear localization and upstream regulators. Prog. Cell Cycle Res. 5:335–47. Qiao, R., F. Weissmann, M. Yamaguchi, N.G. Brown, R. VanderLinden, R. Imre, M.A. Jarvis, M.R. Brunner, I.F. Davidson, G. Litos, D. Haselbach, K. Mechtler, H. Stark, B.A. Schulman, and J.-M. Peters. 2016. Mechanism of APC/CCDC20 activation by mitotic phosphorylation. Proc. Natl. Acad. Sci. U. S. A. 113:E2570–E2578. doi:10.1073/pnas.1604929113. Ran, F.A., P.D. Hsu, J. Wright, V. Agarwala, D.A. Scott, and F. Zhang. 2013. Genome engineering using the CRISPR-Cas9 system. Nat. Protoc. 8:2281–2308. doi:10.1038/nprot.2013.143. Sievers, F., A. Wilm, D. Dineen, T.J. Gibson, K. Karplus, W. Li, R. Lopez, H. McWilliam, M. Remmert, J. Söding, J.D. Thompson, and D.G. Higgins. 2011. Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol. Syst. Biol. 7:539. doi:10.1038/msb.2011.75. Sonoda, E., T. Matsusaka, C. Morrison, P. Vagnarelli, O. Hoshi, T. Ushiki, K. Nojima, T. Fukagawa, I.C. Waizenegger, J.-M. Peters, W.C. Earnshaw, and S. Takeda. 2001. Scc1/Rad21/Mcd1 Is Required for Sister Chromatid Cohesion and Kinetochore Function in Vertebrate Cells. Dev. Cell. 1:759–770. doi:10.1016/S1534-5807(01)00088-0. Tang, Z., H. Shu, W. Qi, N.A. Mahmood, M.C. Mumby, and H. Yu. 2006. PP2A Is Required for Centromeric Localization of Sgo1 and Proper Chromosome Segregation. Dev. Cell. 10:575–585. doi:10.1016/j.devcel.2006.03.010.

3. CRL4B repurposing by mitotic phosphorylation

Tarpey, P.S., F.L. Raymond, S. O’Meara, S. Edkins, J. Teague, A. Butler, E. Dicks, C. Stevens, C. Tofts, T. Avis, S. Barthorpe, G. Buck, J. Cole, K. Gray, K. Halliday, R. Harrison, K. Hills, A. Jenkinson, D. Jones, A. Menzies, T. Mironenko, J. Perry, K. Raine, D. Richardson, R. Shepherd, A. Small, J. Varian, S. West, S. Widaa, U. Mallya, J. Moon, Y. Luo, S. Holder, S.F. Smithson, J. a Hurst, J. Clayton-Smith, B. Kerr, J. Boyle, M. Shaw, L. Vandeleur, J. Rodriguez, R. Slaugh, D.F. Easton, R. Wooster, M. Bobrow, A.K. Srivastava, R.E. Stevenson, C.E. Schwartz, G. Turner, J. Gecz, P.A. Futreal, M.R. Stratton, and M. Partington. 2007. Mutations in CUL4B, which encodes a ubiquitin E3 ligase subunit, cause an X-linked mental retardation syndrome associated with aggressive outbursts, seizures, relative macrocephaly, central obesity, hypogonadism, pes cavus, and tremor. Am. J. Hum. Genet. 80:345–352. doi:10.1086/511134. Teixeira, L.K., and S.I. Reed. 2013. Ubiquitin ligases and cell cycle control. Annu. Rev. Biochem. 82:387– 414. doi:10.1146/annurev-biochem-060410-105307. The UniProt Consortium. 2015. UniProt: A hub for protein information. Nucleic Acids Res. 43:D204–D212. doi:10.1093/nar/gku989. Trinh, A.T., S.H. Kim, H. Chang, A.S. Mastrocola, and R.S. Tibbetts. 2013. Cyclin-dependent Kinase 1- dependent Phosphorylation of cAMP Response Element-binding Protein Decreases Chromatin Occupancy. J. Biol. Chem. 288:23765–23775. doi:10.1074/jbc.M113.464057. Ubersax, J.A., and J.E. Ferrell Jr. 2007. Mechanisms of specificity in protein phosphorylation. Nat. Rev. Mol. Cell Biol. 8:530–541. doi:10.1038/nrm2203. Ungricht, R., and U. Kutay. 2017. Mechanisms and functions of nuclear envelope remodelling. Nat. Rev. Mol. Cell Biol. 18:229–245. doi:10.1038/nrm.2016.153. Vissers, L.E.L.M., C. Gilissen, and J.A. Veltman. 2015. Genetic studies in intellectual disability and related disorders. Nat. Rev. Genet. 17:9–18. doi:10.1038/nrg3999. Vulto-van Silfhout, A.T., T. Nakagawa, N. Bahi-Buisson, S.A. Haas, H. Hu, M. Bienek, L.E.L.M. Vissers, C. Gilissen, A. Tzschach, A. Busche, J. Müsebeck, P. Rump, I.B. Mathijssen, K. Avela, M. Somer, F. Doagu, A.K. Philips, A. Rauch, A. Baumer, K. Voesenek, K. Poirier, J. Vigneron, D. Amram, S. Odent, M. Nawara, E. Obersztyn, J. Lenart, A. Charzewska, N. Lebrun, U. Fischer, W.M. Nillesen, H.G. Yntema, I. Järvelä, H.H. Ropers, B.B.A. de Vries, H.G. Brunner, H. van Bokhoven, F.L. Raymond, M.A.A.P. Willemsen, J. Chelly, Y. Xiong, A.J. Barkovich, V.M. Kalscheuer, T. Kleefstra, and A.P.M. de Brouwer. 2015. Variants in CUL4B are associated with cerebral malformations. Hum. Mutat. 36:106–117. doi:10.1002/humu.22718. Watanabe, N., H. Arai, J.-I. Iwasaki, M. Shiina, K. Ogata, T. Hunter, and H. Osada. 2005. Cyclin-dependent kinase (CDK) phosphorylation destabilizes somatic Wee1 via multiple pathways. Proc. Natl. Acad. Sci. U. S. A. 102:11663–11668. doi:10.1073/pnas.0500410102. Waterhouse, A.M., J.B. Procter, D.M.A. Martin, M. Clamp, and G.J. Barton. 2009. Jalview Version 2-A multiple sequence alignment editor and analysis workbench. Bioinformatics. 25:1189–1191. doi:10.1093/bioinformatics/btp033. Wieser, S., and J. Pines. 2015. The Biochemistry of Mitosis. Cold Spring Harb. Perspect. Biol. 7:a015776. doi:10.1101/cshperspect.a015776. Xue, Y., J. Ren, X. Gao, C. Jin, L. Wen, and X. Yao. 2008. GPS 2.0, a tool to predict kinase-specific phosphorylation sites in hierarchy. Mol. Cell. Proteomics. 7:1598–608. doi:10.1074/mcp.M700574- MCP200.

3. CRL4B repurposing by mitotic phosphorylation

Yin, Y., C. Lin, S.T. Kim, I. Roig, H. Chen, L. Liu, G.M. Veith, R.U. Jin, S. Keeney, M. Jasin, K. Moley, P. Zhou, and L. Ma. 2011. The E3 ubiquitin ligase Cullin 4A regulates meiotic progression in mouse spermatogenesis. Dev. Biol. 356:51–62. doi:10.1016/j.ydbio.2011.05.661. Zhang, S., L. Chang, C. Alfieri, Z. Zhang, J. Yang, S. Maslen, M. Skehel, and D. Barford. 2016. Molecular mechanism of APC/C activation by mitotic phosphorylation. Nature. 533:260–264. doi:10.1038/nature17973. Zhao, W., B. Jiang, H. Hu, S. Zhang, S. Lv, J. Yuan, Y. Qian, Y. Zou, X. Li, H. Jiang, F. Liu, C. Shao, and Y. Gong. 2015. Lack of CUL4B leads to increased abundance of GFAP-positive cells that is mediated by PTGDS in mouse brain. Hum. Mol. Genet. 24:4686–4697. doi:10.1093/hmg/ddv200. Zheng, N., B.A. Schulman, L. Song, J.J. Miller, P.D. Jeffrey, P. Wang, C. Chu, D.M. Koepp, S.J. Elledge, M. Pagano, R.C. Conaway, J.W. Conaway, J.W. Harper, and N.P. Pavletich. 2002. Structure of the Cul1– Rbx1–Skp1–F boxSkp2 SCF ubiquitin ligase complex. Nature. 416:703–709. doi:10.1038/416703a. Zitouni, S., C. Nabais, S.C. Jana, A. Guerrero, and M. Bettencourt-Dias. 2014. Polo-like kinases: structural variations lead to multiple functions. Nat. Rev. Mol. Cell Biol. 15:433–452. doi:10.1038/nrm3819. Zou, Y., Q. Liu, B. Chen, X. Zhang, C. Guo, H. Zhou, J. Li, G. Gao, Y. Guo, C. Yan, J. Wei, C. Shao, and Y. Gong. 2007. Mutation in CUL4B, which encodes a member of cullin-RING ubiquitin ligase complex, causes X-linked mental retardation. Am. J. Hum. Genet. 80:561–6. doi:10.1086/512489. Zou, Y., J. Mi, J. Cui, D. Lu, X. Zhang, C. Guo, G. Gao, Q. Liu, B. Chen, C. Shao, and Y. Gong. 2009. Characterization of nuclear localization signal in the N terminus of CUL4B and its essential role in cyclin E degradation and cell cycle progression. J. Biol. Chem. 284:33320–33332. doi:10.1074/jbc.M109.050427.

3. CRL4B repurposing by mitotic phosphorylation

3.1. Supplementary Figures

Supplementary Figure S1. Alignment of the human CUL4A and CUL4B protein sequences. Alignment was performed using Clustal Omega (Sievers et al., 2011) and analyzed with Jalview 2.10.1 (Waterhouse et al., 2009). Coloring: dark blue and white lettering, identical residue; lighter shades of blue correlate with decreased similarity. Identity is 82% and similarity 89% for 759 amino acids. The possible location for a CUL4A NLS is marked, predicted using the NLStradamus online tool using a prediction cutoff of 0.4 (Nguyen Ba et al., 2009). There is no experimental confirmation that this is a true NLS for CUL4A. CRD1-3, cullin repeat domains.

100

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S2. The serine and threonine residues in the N-terminus of CUL4B are necessary and sufficient for the observed hyperphosphorylation (related to Figure 1). (A) Treatment of whole-cell lysates with λ-Phosphatase and analysis of the migration of endogenous CUL4A and CUL4B. (B-E) Cells were transfected with the indicated FLAG- CUL4A or FLAG-CUL4B constructs and the migration pattern was analyzed. ΔN refers to the truncation of the N- terminus, without or with an added SV40 NLS sequence (denoted NLS-ΔN) to replace the loss of CUL4B’s own NLS. ΔNLS denotes the mutation of CUL4B’s N-terminus to render it unfunctional, according to (Zou et al., 2009). Pnull, Nonphosphorylatable mutants where all serine and threonine residues in the N-terminus of CUL4B were substituted by alanine (Pnull-NT), or these residues were replaced in clusters 0 to 4 (Pnull 0-4). bNT-CUL4A denotes a chimera of CUL4A with the N-terminus of CUL4B (bNT). Async, cell-cycle asynchronized; M, mitotic arrest; Noc, treatment with nocodazole for 24h; Thy, treatment with thymidine for 24h; S, S-phase arrest (S).

101

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S3. N-terminal CUL4B phosphorylation site prediction for selected mitotic kinases. (A) Top, residues found phosphorylated in high-throughput mass spectrometry studies (HTP-MS), according to PhosphoSitePlus (Hornbeck et al., 2015), as in Figure 1A. Below, prediction of phosphorylation by the indicated kinases. The score is an estimate of the likelihood that a residue is phosphorylated by the indicated kinase, obtained using the GPS 3.0 web tool (gps.biocuckoo.org/) (Xue et al., 2008). Grey and yellow horizontal lines denote the low and high confidence thresholds, respectively. Scores for a given site are marked red if above the high threshold and violet if between the high and low thresholds. Note that score cutoffs are independent for each kinase, defined according to the calculated maximal false positive rate, which is of 2% for the high threshold and 10% for the low threshold. (B) Sequence of the phosphorylation clusters identified. A red background illustrates that a specific site was found more often phosphorylated in HTP-MS studies, as specified above.

102

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S4. The XLID CUL4B mutant P50L localizes, is stable and hyperphosphorylated indistinguishably from the WT in mitotic cells. (A) Annotation of CUL4B mutations that have been implicated in syndromic XLID, Cabezas type (Cabezas et al., 2000; Tarpey et al., 2007; Zou et al., 2007; Badura-Stronka et al., 2010; Isidor et al., 2010; Londin et al., 2014; Vulto-van Silfhout et al., 2015; Grozeva et al., 2015; Okamoto et al., 2017). (B) Location of the proline 50 residue within cluster 2. (C) Cycloheximide chase experiments in cells previously arrested in S-phase or mitosis. (D) Immunofluorescence of cells transfected with plasmids expressing FLAG-CUL4B WT or P50L (methanol fixation) and stained for DNA with DAPI. The maximal Z-projection is displayed. (E,F) Migration of the P50L mutant in mitosis-arrested cells. Async, asynchronized; Noc, nocodazole arrest; M, mitosis; Thy, thymidine; S, S-phase. Scale bar, 10 μm.

103

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S5. In vitro kinase assays using recombinant CUL4B or myelin basic protein (MBP) as substrate and the indicated kinases (related to Figure 1).

104

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S6. Subcellular localization of endogenous CUL4A and endogenous HA-CUL4B (CRISPR-edited) in interphase and mitosis (related to Figure 2). (A,B) Immunofluorescence. Asynchronized cells were fixed with 4% formaldehyde and not subjected to detergent pre-extraction, for analysis of all the pools of CUL4. (A) interphase cells are shown. (B) Representative mitotic cells on the indicated mitotic stages. Scale bar, 10 μm.

105

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S7. Knock-in strategy for HA-tagging of CUL4B and insertion of siRNA-resistance mutations. (A) Top, CUL4B gene, with annotation of sequences corresponding to transcript exons (vertical lines or boxes); middle, region corresponding to the transcript exon 3 and the sites that the used gRNAs target; bottom, gene product of the unedited CUL4B gene. Light blue, coding sequence of the unique CUL4B N-terminus. Dark blue, remaining CUL4B coding sequence. (B) Editing strategy for modification of the exon 3 region in the CUL4B loci in genomic DNA. The color assignments to the different inserts are as indicated for the gene product. P2A is a self-cleaving peptide (Kim et al., 2011). (C) Sequencing of the region of insertion upon homologous recombination. Left, unedited and right, edited cells. (D) Successful gene editing was evaluated by PCR, where the lower band corresponds to the unedited gene and the upper band corresponds to the edited exon 3. (E) HA-CUL4B expression was evaluated by flow cytometry upon staining with an HA antibody. Single cell information and relation of HA expression in different stages of the cell cycle is shown.

106

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S8. Controls related to the chromatin fractionation experiments (related to Figures 2 and 3). (A) Evaluation of the DDB1 depletion efficiency by siRNA. ASN, AllStars negative control siRNA. (B) Confirmation by cell fractionation that the edited HA-CUL4B localizes indistinguishably from the unedited endogenous gene in interphase and mitosis. (C) Evaluation of the fractions containing nuclear bodies in cell fractionation experiments (samples from Figure 2C). Note that these bodies are absent from the chromatin fraction in mitosis, substantiating that CUL4B Pnull mutants indeed relocalize (Figure 3) to the chromatin at this stage and not to nuclear bodies. T, total fraction; C, Cytoplasm; N, nucleoplasm; Chr, chromatin fraction.

107

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S9. The CUL4B-DDB1 interaction is maintained in mitosis. (A) Immunoprecipitation experiments of FLAG-CUL4B WT in cells arrested in S-phase or Mitosis with thymidine (Thy) or nocodazole (Noc), respectively. (B) Immunoprecipitation of HA-CUL4B (CRISPR-edited) in cells synchronized at different cell cycle stages, as indicated. “–“ indicates unedited cells. dTB+10h, double thymidine block and release for 10h.

108

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S10. Chromatin localization of other FLAG-CUL4B mutants and FLAG-CUL4B (related to Figure 3). (A-E) Chromatin fractionation in cells transfected with the indicated flag constructs and arrested in mitosis with nocodazole or untreated (asynchronous). (C-D) Diverse CUL4B nonphosphorylatable mutant variants were expressed for cell fractionation experiments. The regions whose serine and threonine residues are mutated to alanine are labeled in green. bNT, N-terminus of CUL4B; T, total fraction; CP, Cytoplasm; NP, nucleoplasm; Chr, chromatin fraction.

109

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S11. Identification of the serine and threonine residues in the N-terminus of CUL4B that are mutated to alanine (green background) in the diverse FLAG-tagged full-length nonphosphorylatable mutants used. The sequence numbers refer to the WT untagged CUL4B, starting at Met-2 (the starting methionine Met-1 was not included).

110

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S12. Effects on chromatin localization/exclusion are not a consequence of impaired phosphomutant localization. Immunofluorescence of cells transfected with plasmids expressing FLAG-CUL4B mutants (methanol fixation) and stained for DNA with DAPI. The maximal Z-projection is displayed. Refer to Supplementary Figure S4 for the subcellular localization of the WT construct. Scale bar, 10 μm.

111

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S13. The knockout of CUL4B has consequences for cell proliferation independent of DNA damage. (A) Strategy for the knockout of CUL4B. The gRNAs used result in the removal of exons 3 and 4. (B) Upon non-homologous repair of the induced double-strand DNA break, it is expected that exon 2 is joined with exon 5, resulting in a frameshift and a premature stop codon. Only 22 amino acids of CUL4B are expected to be correctly translated. (C) Confirmation of the deletion of CUL4B in two independent clones. (D) Effect of CUL4B knockout on cell proliferation in clonogenic assays. Note that the colony size is affected, not colony number. Thus, growth is itself slower and viability is not affected. (E) Flow cytometry analysis of cells immunostained for phosphorylated (γ) histone variant H2A.X, a marker for double-strand DNA breaks (Burma et al., 2001). X axis: DNA content (propidium iodide staining).

112

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S14. CUL4B depletion delays mitotic progression and/or the onset of mitosis. Cells treated with the indicated siRNAs were synchronized using a double-thymidine block and released for the indicated times. Flow cytometry analysis was performed with immunostaining for phosphorylated histone H3 (serine 10). X axis: DNA content (propidium iodide staining).

113

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S15. Overexpression of CUL4B phosphomutants reveals chromosome alignment issues and problems in mitotic exit. (A) Cell proliferation evaluated by clonogenic assays in CUL4B knockout cells infected with lentivirus that induce the expression of the stated CUL4B constructs. (B) Live-cell imaging experiments (here portrayed in 4h time intervals) of CUL4B knockout cells infected with the stated lentiviral constructs. The squares indicate cells that were highlighted in Figure 5. Brightfield is in greyscale; DNA marker (SiR-DNA) in red. Scale bar, 50 μm.

114

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S16. DNA marker for the cells presented in Figure 5C. Upon overexpression of Pnull mutants, cells do undergo chromosome congression, but fail to transit to anaphase. Concomitantly, metaphase cells appear not to properly undergo chromosome congression upon overexpression of Pmimic mutants and fail to generate two nuclei and two daughter cells.

115

3. CRL4B repurposing by mitotic phosphorylation

Supplementary Figure S17. Cell proliferation in HeLa cell lines stably expressing CUL4B mutants. (A) Cells were treated with CUL4B siRNA and the expression of the indicated siRNA-resistant constructs was induced with doxycycline. The values displayed refer to the number of nuclei counted in live-cell imaging experiments, as a measure of cell proliferation. (B) evaluation of the expression level of the CUL4B constructs in the used cell lines.

116

4. Investigation of the role of CRL4WDTC1 in cell cycle progression

4.1. Introduction

WD and tetratricopeptide repeats protein 1 (WDTC1) is a WD40 and TPR-containing putative substrate adaptor of CRL4. It is found to be widely expressed, and is often found to co-precipitate with CRL4 core subunits in interaction studies using HeLa and HEK293T cell lines (Angers et al., 2006; Jin et al., 2006; Li et al., 2010; Olma et al., 2009; Bennett et al., 2010). Despite the early identification as a substrate adaptor of CRL4 complexes, its function remains still obscure, and a role for CRL4WDTC1 was only first described in 2016 (Groh et al., 2016). WDTC1 (DCAF9), also known as adipose (ADP) due to its role as an antiobesity gene, was first described in naturally occurring fly mutants with hypertrophied fat bodies (Doane, 1960a; Häder et al., 2003). These flies were also completely sterile: eggs were characterized by abnormal mitotic figures, not growing beyond the 6th cleavage division (Doane, 1960a; b, 1961). Transplanted wild-type ovaries into homozygous mutant flies produced normal eggs, indicating that this phenotype was autonomous. Surprisingly, ovarian wild-type transplants into mutant flies was sufficient for the regression of the fat body hypertrophy. These data suggest that WDTC1 deletion affects the normal ovarian function, including the putative impairment of an endocrine function and a role in egg development. Although these data suggest that fly WDTC1 does not influence lipid storage directly, several more recent studies argue that it indeed does so in mammals. A WDTC1 heterozygous mice knockout model shows an obesity phenotype, and the authors demonstrate a transcriptional regulatory role of WDTC1 by interacting with a HDAC3 co- repressor complex (Suh et al., 2007). In humans, abdominal subcutaneous WDTC1 expression in adipose tissue resulted in lower fat mass (Galgani et al., 2013). Finally, a recent study established the first connection between the antiobesity properties and a function of WDTC1 as part of a CRL4 E3 ligase (Groh et al., 2016). CRL4WDTC1 was shown to promote histone H2AK119 monoubiquitylation and subsequent transcriptional repression during adipogenesis. Despite the clear involvement of WDTC1 in adipose tissue homeostasis, some observations suggest that the functions of WDTC1 go beyond adipogenesis. First of all, it is widely expressed across murine (Suh et al., 2007) and human tissues, according to the Human Protein Atlas project database (proteinatlas.org). Secondly, it was observed that homozygous knockout in mice is lethal (though not always), which is inconsistent with a function exclusively in adipose tissue (Suh et al., 2007). WDTC1 mutations were identified as drivers in microsatellite-unstable colorectal cancer (Alhopuro et al., 2012),

117

4. CRL4WDTC1 in cell cycle progression indicative of a role as a tumor suppressor regulating cell proliferation. Moreover, analysis of homozygous mutant Drosophila fertilized eggs, which do not develop past 5 to 6 cleavage stages, demonstrated that mitotic spindles are usually blocked at metaphase and there is evidence for chromosomal fragmentation (Doane, 1960a). In line with the hypothesis of a cell proliferation-related function, a screening performed by our group (Piwko et al., 2010; Mouysset et al., 2015) revealed that depletion of WDTC1 leads to multiple delays across interphase and mitosis, especially during metaphase and telophase. We thus set out to uncover how WDTC1 works to drive proper cell cycle progression, as a substrate adaptor of CRL4. Our approach can be split in two main sections: 1) the confirmation and description of the cell cycle delays observed by siRNA screening and 2) the identification of cell cycle-specific interactions of CRL4WDTC1 that could reveal which function it takes up during particular cell cycle stages. Despite that we could not confirm a cell cycle-specific phenotype for WDTC1, we did detect an interaction with a mitotic regulator, taking place both in G2 and mitosis itself.

4.2. Results

4.2.1. WDTC1: a function in the cell cycle? The first evidence that WDTC1 takes up a cell cycle function came from a high-content live-cell imaging siRNA screening focusing on putative CRL4 interactors (see chapter 2). The aim of this screening was to identify cell cycle functions of CRL4 substrate adaptors. By usage of a marker for S-phase – PCNA-GFP – and for DNA – Histone H2B-mCherry – the duration of all cell cycle stages could be measured in HeLa cells. Our group identified functions of novel CRL4 complexes and CRL4-related proteins in DNA repair (Piwko et al., 2010, 2016), DNA replication (Brodersen et al., 2016) and centromere determination (chapter 2) (Mouysset et al., 2015). Another interesting candidate for cell cycle functions was WDTC1, showing various delays throughout all cell cycle stages. We started by confirming the cell cycle stage length determinations made in HeLa cells by the automated platform (Piwko et al., 2010). Indeed, manual evaluation of the live-cell imaging of WDTC1 siRNA-treated cells and cells treated with a control siRNA (siASN, all-star negative) determined that the former population is characterized by significantly longer durations of S-phase, G2 and mitosis (Figure IV- 1). The increased duration of multiple stages indicates a pleiotropic effect, either because CRL4WDTC1 has more than one cell cycle function or the effect at a specific stage results in impaired further progression at other stages (due to DNA damage, for example).

118

4. CRL4WDTC1 in cell cycle progression

Figure IV-1. Measurements of the duration of each cell cycle stage in HeLa cells treated with WDTC1 siRNA #1, in live-cell imaging experiments (Piwko et al., 2010; Mouysset et al., 2015). (A) Data points are manual counts from the live-cell imaging videos and the median ± standard deviation is shown. (B) Depletion efficiency of this particular siRNA, tested by qRT-PCR. NS, not significant (P-value (P) > 0.05). ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001.

The automated platform detected similar delays in 3 out of 4 WDTC1 siRNAs used. However, attempts to rescue cell cycle phenotypes with the overexpression of siRNA-resistant WDTC1 constructs were unsuccessful. Hence, we opted for developing a new approach as an alternative to siRNA to efficiently deplete an endogenous protein. Previous studies have shown that it is possible to induce the degradation of a target protein by creating a protein fusion with the auxin-inducible degron (AID). Auxin is a family of plant hormones which regulates gene transcription by inducing the degradation of specific plant transcriptional repressors. Degradation is performed by the proteasome, following polyubiquitylation by the SCFTIR1 E3 ligase (Figure IV-2A). Binding of the degron to the E3 ligase does not take place in the absence of auxin, but is rapidly induced in its presence, leading to swift destruction of the degron and any covalently-bound protein (Nishimura et al., 2009). This system has been shown to work in yeast and mammalian cells in the degradation of a protein of interest, provided that this protein is tagged with AID and also that TIR1, which only exists in plants, is heterologously expressed in the organism of interest (Nishimura et al., 2009; Holland et al., 2012).

119

4. CRL4WDTC1 in cell cycle progression

Figure IV-2. Overview of the auxin-degron system and generation of an HeLa TIR1-myc cell line. (A) Degradation of an auxible inducible degron (AID)-tagged protein is executed by the ubiquitin-proteasome system. The E3 ligase responsible for this function is SCF, which utilizes a plant-specific substrate adaptor, TIR1. Heterologously expressed TIR1 functions with SCF of other organisms, including yeast and mammalian cells. The AID degron protein, and any covalently-attached moiety are degraded upon the addition of auxin, because this small molecule induces binding to SCFTIR1 (Nishimura et al., 2009). (B) We generated an HeLa cell line constitutively expressing TIR1. Note the varying loading control levels (tubulin). Further testing indicated that degradation with the HeLa clone #10 was most efficient.

We considered this system as a valid alternative to siRNA, bearing additional advantages such as the swift drop in protein levels, demonstrated to occur in human cells regardless of the subcellular localization of the protein of interest (Holland et al., 2012). Such fast downregulation could additionally be employed in dissecting the precise cell cycle stage where WDTC1 acts. However, to achieve downregulation of the endogenous protein it is necessary to perform gene editing so that the AID is introduced and can mediate binding of the endogenous protein of interest to TIR1 and thereby induce its degradation. Development of this approach began by the generation of a HeLa TIR1 cell line constitutively expressing OsTIR1 (from rice, Oryza sativa, Figure IV-2B). From the set of obtained clones, we picked three candidates with higher TIR1/tubulin ratio which were tested for their capacity to degrade a transiently transfected GFP-AID construct (Figure IV-3A). Both clones 10 and 16 appeared to mediate the degradation of GFP-AID and GFP-3miniAID, and we chose clone 10 for all subsequent studies. To further test this system, we derived this cell line for stable inducible expression of a panel of GFP constructs bearing an N- terminal or C-terminal AID or the 3miniAID – 3 copies of a region of the full-length AID that preserves binding to TIR1 upon auxin addition (Kubota et al., 2013).

120

4. CRL4WDTC1 in cell cycle progression

Figure IV-3. Testing of GFP-AID degradation in HeLa cell lines expressing TIR1. (A) Cells from three TIR1-expressing clones were transiently transfected with pcDNA5 plasmids expressing GFP constructs with a C-terminal AID tag. Besides the full AID degron (IAA17 protein), we also employed 3 copies of “miniAID” – the degron region that is most important for TIR1 binding (Kubota et al., 2013). Cells were treated with auxin for 24h, starting 4h after transfection. (B) Induced degradation of either N-terminally or C-terminally AID-tagged GFP (inducible stable expression in TIR1 HeLa cell line, clone 10) by addition of auxin for the indicated times. Doxycycline was added 24h before the start of the experiment to induce GFP expression, and kept throughout the experiment.

Treatment of these cell lines with auxin (24h after inducing the expression of the construct with doxycycline) revealed rapid degradation of GFP, especially if tagged in its C-terminus (Figure IV-3B). GFP was virtually absent in the first time-point, 6h after auxin treatment. Note that doxycycline was kept during the course of the experiment, to analyze exclusively the effect of induced degradation by auxin. While upon C-terminal tagging it appeared that AID and 3miniAID were equally efficient as degrons, upon N-terminal tagging it became apparent that the 3miniAID was more efficient, albeit still not as much as upon C-terminal tagging. After we tested auxin-induced GFP construct degradation, we aimed to infer whether WDTC1 can equally be efficiently and quickly targeted for degradation by auxin addition. Due to the better results for

121

4. CRL4WDTC1 in cell cycle progression

GFP (Figure IV-3B), we only tagged WDTC1 C-terminally with AID or 3miniAID. We discovered that indeed inducible degradation of WDTC1 also takes place swiftly and efficiently (Figure IV-4). Similar to GFP, efficient degradation was observed already on the first tested time-point, 6h after the addition of auxin. Overexposure of the western blot revealed that the remaining WDTC1 continues nevertheless to be degraded in the course of the 24h of the experiment, but the bulk of degradation takes place in the first hours. We determined that in this case, 3miniAID is a better degron than the full-length AID, judged by the remaining levels on WDTC1 at 6h (compared to the auxin-untreated control).

Figure IV-4. Induced degradation of C-terminally AID or 3miniAID-tagged GFP (inducible stable expression in TIR1 HeLa cell line, clone 10) by addition of auxin for the indicated times. The expression of the indicated constructs was induced 24h before the addition of auxin.

The initial testing stage (Figures IV-3 and IV-4) made clear that this system works with several GFP and WDTC1 constructs in inducing protein degradation. From these experiments, it appeared that very little protein of interest remains in cells after only 6h of auxin treatment, demonstrating that this is a much more efficient and fast downregulation method than RNAi or other inducible degron systems (Park et al., 2014). Nevertheless, we aimed to more precisely determine the kinetics of protein depletion. To do so, cells were treated with auxin for 30 min up to 8h and the remaining WDTC1 was quantified by densitometry, normalized to the GAPDH level (Figure IV-5). These quantifications revealed that WDTC1 depletion fits a first-order kinetics model, which was used for the regression curve (Figure IV-5, bottom). By applying the same model, we calculated the pseudo-half-life of WDTC1 after the addition of auxin, which is of ~50 min. Please note that this is not a true value for protein half-life, as the induction of protein expression is maintained.

122

4. CRL4WDTC1 in cell cycle progression

Figure IV-5. Calculation of the pseudo-half-life of C-terminally 3miniAID-tagged WDTC1. The levels of WDTC1 following auxin addition (top) were measured by densitometry (bottom). A first order kinetics model was employed for the calculation of the half-life.

These data also revealed that only ~20% of WDTC1 is left after 2h of auxin treatment and it is mostly absent after 4h, consistent with the data shown above (Figure IV-4). Due to the apparent success of this approach in inducing the degradation of WDTC1 in inducible cell lines, we set to tag the endogenous WDTC1 alleles with the 3minAID degron. To this end, we designed a CRISPR-based approach that relies on the nuclease activity of Streptococcus pyogenes Cas9 (Hsu et al., 2014) to induce double-strand breaks (DSBs) in the WDTC1 locus. If a DNA template with sufficient homology exists, the cell may use homologous recombination (HR) to repair this DSB, thereby referred to as homology-directed repair. It has been shown by others that a transfected DNA fragment with two flanking 800-1000 bp homology regions can be used by the HR machinery of the cell and allows for gene editing, i.e. the insertion of an intented DNA region (Hsu et al., 2014; Jiang and Marraffini, 2015). In principle, any DNA insert can be included in-between the homology region. In our approach, we generated a HR template so that WDTC1’s own stop codon is eliminated and the 3miniAID inserted, before the 3’ UTR region (Figure IV-6). The upstream 800 bp homology region includes part of the intron between exons 15 and 16 (last exon),

123

4. CRL4WDTC1 in cell cycle progression

Figure IV-6. Strategy for the CRISPR-mediated knock-in of the 3miniAID degron on the C-terminus of WDTC1. 3 different gRNAs that span across the stop codon were independently tested. The template for homologous recombination (HR) is supplied on a pUC18 vector and is composed of the insert and two flanking 800 bp homology regions. whereas the downstream 800 bp homology region is part of the 3’ UTR. Regarding the gRNA design, we used the online tool crispr.mit.edu to generate 3 high-scoring gRNAs that would span across the stop codon, so that the induced DSB is at the site of editing, known to increase HR efficiency (Ran et al., 2013). We first set to optimize the conditions that would enhance the efficiency of editing. To this end, we co-transfected HeLa TIR1 cells with a pX458 vector (expressing Cas9 and the gRNA) and a pUC18 vector (containing the HR template sequence for 3miniAID insertion) on a 1:5 ratio, respectively. Editing was evaluated by PCR, where the amplified region spans the edited site on genomic DNA. Please note that we did not pick individual clones at this point, hence most alleles in the population will not be edited (band at around ~900 bp). Nevertheless, this optimization approach should allow the visualization of a distinct band of ~1700 bp, corresponding to the edited allele, indicating that some cells within the population had their WDTC1 gene appropriately modified. Two main condition settings were tested at this point: 1) the editing efficiency when using either of the three different gRNAs and 2) the contribution of DNA repair modulatory drugs for successful editing

124

4. CRL4WDTC1 in cell cycle progression

Figure IV-7. PCR analysis of the edited region in genomic DNA of a cell population (on top, expected band sizes). Cells were co-transfected with the indicated pX458-Cas9-GFP-gRNA vector and the HR template-containing vector, using lipofectamine 3000. RS-1 and/or SCR7 were added 4h post-transfection at the indicated concentrations. Control refers to non-transfected cells.

with the provided HR template (Figure IV-7). In the latter case, we tested a stimulator of HR, RS-1 (enhances RAD51 activity), previously shown to increase the CRISPR-mediated HR efficiency (Song et al., 2016). We also utilized an inhibitor of nonhomologous end joining (NHEJ), SCR7 (inhibits DNA ligase IV), which has been used for the same purpose (Maruyama et al., 2015; Chu et al., 2015). Both inhibitors were added for 24h post-transfection. For most conditions, there was a clear appearance of a band of higher size, indicating that this genomic region is indeed being edited (Figure IV-7). Moreover, it appeared that the usage of gRNA #24 led to better results than the remaining, as evaluated by a stronger upshifted band as compared to the control-sized band. Higher concentrations of RS-1 appeared to increase the efficiency of editing. However, we noticed that the usage of this inhibitor greatly impacted cell survival. Regarding SCR7, we did not observe a benefit for the usage of this drug. We additionally tested whether changing the transfection method would improve the editing efficiency. We tested both lipofectamine 3000 and electroporation, but the usage of the latter method did not support the integration of the insert, despite the high efficiency of pX458 transfection (Appendix C). With the aim of sorting the cells and subsequent individual clone picking, we nevertheless employed the above-mentioned drugs and extended the time of treatment up to 48h post-transfection (Figure IV-8).

125

4. CRL4WDTC1 in cell cycle progression

Figure IV-8. PCR analysis of the edited region in genomic DNA of a cell population. Cells were co-transfected with pX458-Cas9-GFP-gRNA #24 and the HR template vector, and RS-1 and/or SCR7 were added 4h post-transfection at the indicated concentrations and time. The migration of all bands appears faster, possibly due to the high concentration of DNA sample. *A faint defined band is visible instead of the smear observed for other samples.

In this particular co-transfection experiment, the knock-in efficiency appeared much lower. Nevertheless, for most conditions we could observe a smear rather than discrete bands (Figure IV-8). The exception appeared to be the cells treated with SCR7 0.1 μM for the duration of 48h, which showed a faint discrete band. Cells transfected in these conditions were sorted for GFP, which is expressed together with Cas9 in the pX458 vector (Figure IV-9A), after which we proceeded by individual clone picking.

Figure IV-9. Editing of the WDTC1 gene with the 3miniAID sequence in HeLa cells (expressing TIR1). (A) Cells co- transfected with pX458-Cas9-GFP-gRNA #24 and the HR template-containing vector and treated with SCR7 0.1 μM for 48h were sorted for GFP and seeded in 15-ch dishes). After approximately 12 days, clones were picked. (B) Shown are the results of genomic PCR reactions performed on 33 different clones, around the editing site. If unedited, the PCR yields a product of 961 bp, whereas if successfully edited 1729 bp. The high concentration of PCR product likely distorts band migration, which does not appear to be at the exact expected size (including the control).

126

4. CRL4WDTC1 in cell cycle progression

From the 33 clones tested, clone 5 in particularly had a clear band upshift and appeared not to have any unedited-sized band, indicating that the modification was homozygous, i.e. in all alleles (Figure IV-9B). We utilized clone 5 for further testing, as we needed to determine whether the features of our insert were conserved in this construct, especially because the band migration on the above agarose gels does not match exactly the expected band size, including the unedited PCR product. We tested for the appearance of a new band upon blotting with an anti-HA antibody (the insert contains two HA epitopes, plus the 3miniAID degron). Indeed, as compared to control cells, we did verify that a band appears on the respective western blot in edited cells (clone 5 and a heterozygous clone), which matches the expected size for the WDTC1-2HA-3miniAID edited protein (predicted: 102 kDa). Remarkably, treatment of these cells with auxin led to the disappearance of this band (Figure IV-10). This data thus demonstrates that the knock-in of the 3miniAID degron on the WDTC1 was correctly performed. More importantly, we demonstrate that the auxin-inducible degron system indeed works to quickly induce the degradation of an endogenous target protein. Following the successful knock-in of the 3miniAID degron in WDTC1, we set to clarify whether a CRL4WDTC1 E3 ligase indeed does work in cell cycle regulation. We performed a simple analysis by flow

Figure IV-10. CRISPR-mediated editing results (C-terminal knock-in of HA-3miniAID). Shown is the HA signal of unedited cells, as well as of a homozygous edited clone and an heterozygous clone, as evaluated by PCR. Auxin was added to demonstrate the targeted degradation of the endogenous tagged WDTC1. NUDC is a WDTC1 interactor (see below), which prompted us to also test whether its protein levels are altered upon WDTC1 downregulation

127

4. CRL4WDTC1 in cell cycle progression cytometry to determine if the phenotypes obtained upon siRNA treatment are reproducible, i.e. S-phase, G2 and mitotic delays (Figure IV-1). We used the obtained edited clone 5 cells for this analysis (and the unedited cell line as a control). Treatment with auxin for 48h or 72h, although clearly leading to the disappearance of WDTC1 (Figure IV-10), did not result in altered relative cell populations at the different cell cycle stages, indicating that this downregulation does not impact the cell cycle (Figure IV-11).

Figure IV-11. Analysis of cell cycle effects of edited WDTC1-HA-3miniAID downregulation by auxin treatment for the course of 48h (top) or 72h (bottom). Quantifications were performed from flow cytometry measurements.

Collectively, these experiments do not support a cell cycle role for WDTC1, and suggest that the observed phenotypes upon siRNA depletion are due to off-target effects. Hence, we did not perform further phenotypic characterization of WDTC1 depletion, even though a role of CRL4WDTC1 in the cell cycle cannot be completely ruled out (see the next results section and the discussion). In particular, interactome analysis suggests that WDTC1 does interact with a cell cycle regulator, which we demonstrate in the following section.

128

4. CRL4WDTC1 in cell cycle progression

4.2.2. Identification of a CRL4WDTC1 function: analysis of WDTC1 interactions Despite that the demonstration of a function of WDTC1 in the cell cycle is lacking, we examined the interactome of WDTC1 in HeLa cells. We aimed to perform this analysis by immunoprecipitation-mass spectrometry (IP-MS) in order to employ an unbiased method with the capacity of revealing novel protein- protein interactions. Thus, we initiated a collaboration with the laboratory of Paola Picotti, benefiting from direct support by the former PhD student Andre Melnik. For the IP-MS experiment, we were aiming at detecting inclusively E3 ligase-substrate interactions. However, binding of a substrate to its corresponding E3 ligase is expected to be transient, and is often not visible in immunoprecipitation experiments. Therefore, we attempted to retrieve transient interactions by usage of the bifunctional chemical crosslinker DTSSP (Figure IV-12). DTSSP has two functional groups that can react with two lysine residues which are on two distinct proteins. In this manner, transiently interacting proteins can be covalently crosslinked, thus converting unstable binding into a covalent tether. Applied to our experimental setup, we crosslink proteins immediately after cell lysis, to which follows the quenching of unbound DTSSP functional groups. At this point, we perform the immunoprecipitation step, followed by trypsin digestion of all co-precipitating

Figure IV-12. Identification of transient interactions: usage of a chemical cross-linker, DTSSP, in immunoprecipitation-mass spectrometry. Collaboration with Andre Melnik and Paola Picotti.

129

4. CRL4WDTC1 in cell cycle progression proteins (Figure IV-12). The mass-spectrometry based analysis of crosslinked peptides is complex very complex, in particular when one considers for the analysis that the two peptides may originate from any protein in the proteome. To greatly facilitate this process, we destroy the crosslink at this point: DTSSP contains a disulfide bond which can be reduced with DTT treatment. Although we lose the information of specific interacting regions, mass spectrometry analysis can nevertheless identify all the proteins that co- precipitated. To apply this strategy, we generated an inducible stable cell line expressing WDTC1-TAP (tandem affinity purification; 2xStrep-HA), where WDTC1 expression is under the CMV promoter. We first confirmed that we are able to detect peptides that have been modified by DTSSP, because a reminiscent

Figure IV-13. Detection of crosslinked peptides by tandem mass spectrometry. Example for a DDB1 peptide, after WDTC1 immunoprecipitation. KCAM is a lysine residue modified by DTSSP and iodoacetamide.

130

4. CRL4WDTC1 in cell cycle progression stretch of this crosslinker stays covalently bound even after reduction of its disulfide bond. We performed a WDTC1 IP as described above and searched for modified peptides on WDTC1 interactors. We used a peptide from DDB1, the CRL4 linker, to exemplify this effect (Figure IV-13). This peptide, found unmodified in the absence of DTSSP, is detected to contain a modified lysine in the DTSSP-treated sample. Peptide sequencing by tandem mass spectrometry revealed that indeed the mass/charge ratio of the lysine residue shifts from ~128 Da to ~273 Da, matching the expected increase due to DTSSP crosslinking. This modified lysine residue is labeled KCAM (Lysine-CAMthiopropanoyl). These data show that we are able to 1) successfully perform chemical crosslinking in cell lysates (although we do not demonstrate that we crosslinked two proteins to each other) and 2) detect peptides that have been modified by DTSSP. Subsequently, we tested whether applying this chemical crosslinker indeed contributed to an easier detection of interactions. In mass spectrometry studies, the spectral counts (number of spectra identified for a protein) are used as a label-free semi-quantitative measure of protein abundance in a sample. In our case, we used spectral counts as a measure of the amount of co- precipitated protein, and compared this value for immunoprecipitations performed with and without DTSSP treatment of the cell lysate (Figure IV-14).

Figure IV-14. Comparison between a WDTC1 immunoprecipitation without and with the DTSSP crosslinker, analyzed by the detected co-precipitation of known CRL4 interactors. Left: numbers are the difference in percentage of the spectral counts between conditions (with DTSSP - without DTSSP), being that spectral counts are a semi-quantitative measure of protein abundance in a sample. Higher numbers mean that more of a given protein was co-precipitated with WDTC1 if DTSSP was used, negative numbers signify that less interaction was observed if DTSSP was used. Right: detected co-precipitating proteins that are known subunits of CRL4 (blue) or regulators (red), analyzed with STRING (string-db.org/). The bait, WDTC1, is labeled green. *COPS7A is only present with DTSSP

131

4. CRL4WDTC1 in cell cycle progression

We first analyzed the difference in abundance of known components or interactors of CRL4. In fact, it appeared that the direct interactor DDB1, or the cullin, were not increasingly co-precipitated in the WDTC1 IP when DTSSP was employed (Figure IV-14), indicating that strong interactions do not benefit from this approach. However, we also analyzed this effect for a known regulator of CRLs termed COP9 signalosome (CSN). It appears that CSN subunits are more abundant if DTSSP is used, as judged by the increase in its detected subunits COPS2, COPS5, COPS8 or COPS7A/B (Figure IV-14, left). In conclusion, it appears that interactions that are more transient benefit from DTSSP crosslinking, in the sense that they are easier to detect, whereas interactions that are anyway strong do not. These data meet our objective of enhancing the method so that we could increase the chances of finding CRL4WDTC1 substrates. Next, we employed this chemical crosslinking approach to identify WDTC1 interactions. We analyzed all WDTC1 co-precipitating proteins and used the online tool CRAPome to calculate the likelihood that a precipitated protein is indeed a true WDTC1 interactor (Figure IV-15). All top hits are known interactors of CRL4 (highlighted in blue), validating the approach and the scoring algorithm applied. Of the putative novel interactors, one particular protein that sparked our interest was Nuclear distribution protein C homolog (NUDC; highlighted in yellow), a known regulator of the mitotic kinase PLK1 (Nishino et al., 2006; Aumais et al., 2001; Zhou et al., 2003).

Figure IV-15. Top co-precipitating proteins detected by mass spectrometry in a WDTC1 immunoprecipitation with the DTSSP crosslinker. The list of detected proteins, with spectral count information, was submitted to the CRAPome online tool (Mellacheruvu et al., 2013), which scores the likelihood of an interaction based on information on known contaminants in other mass spectrometry studies and based on this experiment’s internal control (no expression of the bait). Dark blue, known direct interactor; Light blue, known indirect interactors; Green, bait; Yellow, NUDC (mitotic regulator).

132

4. CRL4WDTC1 in cell cycle progression

Based on the mass spectrometry analysis, we aimed at validating the WDTC1-NUDC interaction. To do so, we performed similar immunoprecipitation assays, but instead of analyzing by mass spectrometry, we detected this interaction by western blot. For this approach, we also varied the concentration of DTSSP applied up to 2.5 mM (for the experiments above, we used 0.25 mM). Indeed, we confirmed the interaction of WDTC1 with both the E3 ligase (via DDB1) and with NUDC (Figure IV-16). It is clear that 0.25 mM DTSSP enhances the amount of co-precipitated NUDC, but its usage is not necessary to allow the detection of this interaction. Nevertheless, the amount of detected NUDC is likely a substantial underestimation of co-precipitated protein, because we noticed that DTSSP hinders the detection of this protein, probably by blocking the epitope (Figure IV-16, input). In connection with the possibility of WDTC1 as a cell cycle regulator, and following the confirmation of NUDC as a true interactor of WDTC1, we set to uncover if this interaction is specific to some cell cycle stages, or is maintained throughout the cell cycle. Indeed, IP experiments performed in cells synchronized at different cell cycle stages revealed that the WDTC1-NUDC interaction is enhanced in

Figure IV-16. Immunoprecipitation (IP) of WDTC1-TAP (strep-strep-HA) from HeLa cells and western blot analysis. Proteins in the cell lysate were crosslinked with several concentrations of DTSSP after lysis (see scheme on top), after which WDTC1 was immunoprecipitated and the eluate analyzed for the presence of DDB1 and NudC (bottom).

133

4. CRL4WDTC1 in cell cycle progression

Figure IV-17. Immunoprecipitation of WDTC1-TAP in several cell cycle stages, after double-thymidine block release or nocodazole treatment (mitosis). (A) Evaluation of the quality and cleanliness of the immunoprecipitation by SDS- PAGE and silver nitrate protein staining. (B) Analysis of the interaction of WDTC1 with DDB1 and NudC. (C) Flow cytometry analysis to evaluate the quality of cell synchronization. X axis: DNA content; Y axis: cell count. Async, asynchronous; S, S-phase. mitosis, present in G2 but not in other cell cycle stages (Figure IV-17B,C). Please note that the levels of WDTC1 vary because expression from a CMV promoter is cell-cycle dependent (Brightwell et al., 1997), and indeed treatment with the proteasome inhibitor MG132 does not affect this oscillation, arguing that it is due to fluctuating expression (appendix C). The finding of a mitotic-specific interaction between WDTC1 and NUDC might represent that this is a substrate of the CRL4WDTC1 E3 ligase, or perhaps that this protein is rather a regulator of WDTC1. We tested the former hypothesis by measuring the levels of NUDC, in auxin-mediated WDTC1 depletion (Figure IV-10), but the levels of NUDC remained unchanged. This suggests that NUDC is not a substrate of CRL4WDTC1, or at least that WDTC1 does not mediate its proteasomal degradation. Put together, our data argues that WDTC1 interacts with the mitotic regulator NUDC, and does so in mitosis. We also demonstrated that the usage of chemical crosslinking enhances the amount of co- precipitated proteins, especially if these are not constitutive interactors, but rather transient interactors. In line with this observation, we conclude that WDTC1 and NUDC are rather transient interactors because there is a clear benefit to the usage of DTSSP in NUDC co-precipitation. Nevertheless, we can detect this interaction without chemical crosslinking, which might indicate that NUDC is not a substrate of CRL4WDTC1, but perhaps takes up other functions.

134

4. CRL4WDTC1 in cell cycle progression

4.3. Discussion

4.3.1. CRISPR-mediated gene editing: a technical advancement In the course of this study, we decided to explore the usage of the auxin-inducible degron as an alternative to siRNA-mediated protein depletion of WDTC1. After careful testing of the functionality of this system in overexpressed constructs, we performed CRISPR-based genome editing to introduce the 3miniAID degron in the C-terminus of WDTC1. This system was very efficient in quickly degrading WDTC1. This represents a clear technical advancement as compared to other similar methods, and a clear advantage over siRNA- mediated depletion. In particular, this method 1) greatly decreases the likelihood of off-target effects, due to the necessity of DNA integration based on homology (besides Cas9 specificity), 2) allows for fast depletion for cell cycle studies or other studies requiring quick downregulation and 3) results in very efficient depletion, best for functional studies that require close-to-completion downregulation and a knockout is not possible (e.g. essential proteins). This approach has nevertheless a number of caveats, namely 1) the protein of interest has to be tagged, which may be deleterious to its normal function, 2) the generation of a parental cell line expressing TIR1 is required and 3) it is required to perform the technically challenging step of CRISPR-mediated knock-in of the degron. We nevertheless argue that the advantages outweigh the disadvantages of using this method and this system offers a wide array of applications. We should also mention that during the step of endogenous tagging, one may include additional small tags for functional studies of the endogenous protein, such as IPs or immunofluorescence in case good antibodies are not available. Recently, the application of this method in a manner identical to what is displayed here has been published for HCT116 cells and mouse embryonic stem cells (Natsume et al., 2016).

4.3.2. WDTC1: is a cell cycle function ruled out? Even though our current data does not support a clear cell cycle impact for WDTC1 depletion, it is possible that a CRL4WDTC1 E3 ligase does operate in the cell cycle. In regards to siRNA rescue experiments, it is possible that the construct used does not correspond to a functional WDTC1. For example, other putative isoforms exist and perhaps the one tested is not the one who takes part in cell cycle regulation. In regards to the CRISPR editing, although by PCR analysis it appears that we did edit all alleles, we did not fully confirm that there is no detectable unedited endogenous WDTC1 by western blot. To overcome this caveat, we did attempt to develop antibodies anti-WDTC1 that were successful in detecting the overexpressed protein, but unfortunately not the endogenous (Appendix C). One possibility for the expression of WT WDTC1 even though we do not observe the WT-sized band by PCR is if for some alleles

135

4. CRL4WDTC1 in cell cycle progression the primer binding site has been destroyed during the repair of the DSB by NHEJ. This would make it impossible to detect this NHEJ-edited allele with the primers used, but cells could nevertheless express WDTC1 copies that are very close to the full-length WT. Another possibility is if the loss of function of CRL4WDTC1 does in fact not perceptibly result in a strong cell cycle phenotype, even it operates in its regulation. The result of this E3 ligase loss-of-function is likely the upregulation of its substrate(s), which in fact might not be deleterious, in particular in the cell type used (HeLa cells).

4.3.3. Significance of NUDC-WDTC1 interaction In support of a cell cycle function for CRL4WDTC1, we detected cell cycle dependent interaction with NUDC. This protein is a regulator of the mitotic kinase PLK1, which is vital for the proper execution of early mitotic events, but also for chromosome congression and cytokinesis (Zitouni et al., 2014). In the context of mitosis, NUDC impacts chromosome segregation and cytokinesis (Zhou et al., 2003; Aumais et al., 2003) and indeed it was shown that it is required for proper kinetochore localization of PLK1 in prometaphase. Interestingly, the localization of NUDC to the kinetochore requires phosphorylation by PLK1 itself, establishing a positive feedback loop necessary for proper PLK1 localization (Nishino et al., 2006). The present data does not support CRL4WDTC1-mediated degradation of NUDC. Despite that we did observe an interaction between WDTC1 and NUDC especially in mitosis, the significance of this interaction is yet to discover. One possibility is that CRL4WDTC1 does ubiquitylate NUDC, but without the consequence of proteasomal degradation. Another hypothesis is that NUDC is itself a regulator of WDTC1, for example to target CRL4WDTC1 to the kinetochore to fulfill a novel function. It is interesting that we detect the interaction of WDTC1 and NUDC at the point where NUDC is phosphorylated by PLK1 and targeted to the kinetochore (prometaphase), which could indicate that the NUDC-WDTC1 interaction requires phosphorylation of NUDC.

4.3.4. Analysis of interactions: enhancing the method The detection of E3 ligase-substrate interactions is a common issue in the field, which is why many putative E3 ligase complexes have no described function, in particular CRL4 complexes (Piwko et al., 2010). Here, we employed the usage of chemical crosslinking as the means of enhancing the detection of transiently interacting proteins. Our data supports that this objective was partly accomplished, as we demonstrate that indeed in IP-MS studies one can increase the amount of co-precipitating proteins. However, this method still requires optimization, and we believe that we are yet to discover a substrate for CRL4WDTC1. Nevertheless, perhaps we did not detect such interaction because we did not perform these studies in cells synchronized at the point where binding between proteins takes place. Besides, the usage

136

4. CRL4WDTC1 in cell cycle progression of a proteasome inhibitor would probably be greatly beneficial to find this E3 ligase-substrate interaction, and we set such optimizations as the goal for the future.

4.4. Materials and Methods

For the methods not included here, please see chapter 3. qRT-PCR. Total cell RNA was isolated an RNeasy mini kit (Qiagen) according to the manufacturer’s instructions. 5 µg total cell RNA were used for reverse transcription with 200 U SuperScript II reverse transcriptase (Invitrogen) in the provided First-strand buffer with 10 mM DTT, in the presence of 40 U RNaseOUT ribonuclease inhibitor (Invitrogen), 500 µM dNTP mix (Fermentas) and 2.5 µM hexamer random primers (Microsynth). Real-Time quantitative reverse transcription PCR was performed on a Light Cycler 480 (Roche) using the SYBR Green reagent (Life Technologies), according to the manufacturer’s instructions.

Vectors, cell line generation and auxin treatment. Cloning and transfections were essentially performed as described (chapter 3). AID, 3miniAID and OsTIR1 cDNA were PCR-amplified from pMK43, pMK154 and pMK243 plasmids, respectively, kindly offered by M. Kanemaki, and subcloned in pcDNA5 vectors together with EGFP or WDTC1 cDNA. OsTIR1 was cloned into a pIRESpuro2 vector (Takara Bio Inc). For the homology recombination template, homology regions were PCR-amplified from HeLa genomic DNA and AID and 3miniAID sequences were subcloned from the previously generated pcDNA5 vectors. Inducible cell lines were generated using the HeLa Flp-In T-Rex system (Invitrogen), according to the vendor’s instructions. Induction of degradation was initiated by addition of 500 µM Indole-3-acetic acid (Heteroauxin, Sigma) for the indicated times.

CRISPR knock-in. As described in chapter 3, though only one pX458 was transfected at a time. The DNA target sequence of the used gRNA sequences are: gRNA #12 (CAGGGCTGGAGGGTCTAGCT), gRNA #13 (CCAGGGCTGGAGGGTCTAGC) and gRNA #24 (GCTGGAGGGTCTAGCTGGGC). The DNA template was generated as described (chapter 3) by subcloning the vectors mentioned above. Cells were sorted for GFP, due to the absence of a selection marker.

Immunoprecipitation and mass spectrometry. IP experiments were performed identically to chapter 3 (HA tag) if western blot analysis was performed and no DTSSP was used. For experiments where we performed chemical crosslinking, DTSSP (Thermo Fischer) was added at the indicated concentrations during the lysis step and the lysate was incubated for 2h. To stop the reaction, 50 mM Tris-HCl was added. For mass

137

4. CRL4WDTC1 in cell cycle progression spectrometry, precipitating proteins were not eluted from the beads, but instead the beads were resuspended in 5x bead volume of 100 mM NH4HCO3. The beads were incubated with 12 mM DTT for 30 min at 60ºC and iodoacetamide was added to a final concentration of 40 mM and incubated for 45 min at

RT. The solution was then diluted with 5x bead volume of 100 mM NH4HCO3 and trypsin (Promega) added at a 1:100 protein mass ratio. Following an overnight incubation at 37ºC, the supernatant was recovered, acidified with 0.2% formic acid and run in an C18 column pre-activated with 0.2% formic acid and 10% acetonitrile. After 3 washes with 0.2% formic acid, 10% acetonitrile solution, peptides were eluted with 50% acetonitrile. 2 µg of peptides were run in a Q Exactive mass spectrometer (Thermo Fischer). Peptide and protein assignments were performed and analyzed with Progenesis QI (Nonlinear Dynamics). Protein identifications were analyzed with the CRAPome web tool (crapome.org), with built-in significance analysis of interactome (SAINT) analysis which assigns a confidence/probability score for the interaction. The FC-B score estimates the fold change in comparison to negative controls. The Venn diagram was created with Venny 2.1 (bioinfogp.cnb.csic.es/tools/venny).

4.5. References

Alhopuro, P., H. Sammalkorpi, I. Niittymäki, M. Biström, A. Raitila, J. Saharinen, K. Nousiainen, H.J. Lehtonen, E. Heliövaara, J. Puhakka, S. Tuupanen, S. Sousa, R. Seruca, A.M. Ferreira, R.M.W. Hofstra, J.P. Mecklin, H. Järvinen, A. Ristimäki, T.F. Ørntoft, S. Hautaniemi, D. Arango, A. Karhu, and L.A. Aaltonen. 2012. Candidate driver genes in microsatellite-unstable colorectal cancer. Int. J. Cancer. 130:1558–1566. doi:10.1002/ijc.26167. Angers, S., T. Li, X. Yi, M.J. MacCoss, R.T. Moon, and N. Zheng. 2006. Molecular architecture and assembly of the DDB1-CUL4A ubiquitin ligase machinerary (Supplemental figures). Nature. 443. Aumais, J.P., J.R. Tunstead, R.S. McNeil, B.T. Schaar, S.K. McConnell, S.H. Lin, G.D. Clark, and L.Y. Yu-Lee. 2001. NudC associates with Lis1 and the dynein motor at the leading pole of neurons. J Neurosci. 21:RC187. doi:20015895 [pii]. Aumais, J.P., S.N. Williams, W. Luo, M. Nishino, K.A. Caldwell, G.A. Caldwell, S.H. Lin, and L.Y. Yu-Lee. 2003. Role for NudC, a dynein-associated nuclear movement protein, in mitosis and cytokinesis. J Cell Sci. 116:1991–2003. doi:10.1242/jcs.00412. Bennett, E.J., J. Rush, S.P. Gygi, and J.W. Harper. 2010. Dynamics of cullin-RING ubiquitin ligase network revealed by systematic quantitative proteomics. Cell. 143:951–965. doi:10.1016/j.cell.2010.11.017. Brightwell, G., V. Poirier, E. Cole, S. Ivins, and K.W. Brown. 1997. Serum-dependent and cell cycle- dependent expression from a cytomegalovirus-based mammalian expression vector. Gene. 194:115–123. doi:10.1016/S0378-1119(97)00178-9. Brodersen, M.M.L., F. Lampert, C.A. Barnes, M. Soste, W. Piwko, and M. Peter. 2016. CRL4WDR23- Mediated SLBP Ubiquitylation Ensures Histone Supply during DNA Replication. Mol. Cell. 62:627– 635. doi:10.1016/j.molcel.2016.04.017. Chu, V.T., T. Weber, B. Wefers, W. Wurst, S. Sander, K. Rajewsky, and R. Kühn. 2015. Increasing the

138

4. CRL4WDTC1 in cell cycle progression

efficiency of homology-directed repair for CRISPR-Cas9-induced precise gene editing in mammalian cells. Nat. Biotechnol. 33:543–548. doi:10.1038/nbt.3198. Doane, W.W. 1960a. Developmental Physiology of the Mutant Female Sterile(2)Adipose of Drosophila melanogaster. I. Adult morphology, longevity, egg production, and egg lethality. J. Exp. Zool. 145:1– 21. doi:10.1002/jez.1401450102. Doane, W.W. 1960b. Developmental physiology of the mutant female sterile(2)adipose of Drosophila melanogaster. II. Effects of altered environment and residual genome on its expression. J. Exp. Zool. 145:23–41. doi:10.1002/jez.1401450103. Doane, W.W. 1961. Developmental physiology of the mutant female sterile(2)adipose of Drosophila melanogaster. III. Corpus allatum-complex and ovarian transplantations. J. Exp. Zool. 146:275–298. doi:10.1002/jez.1401460307. Galgani, J.E., D.E. Kelley, J.B. Albu, J. Krakoff, S.R. Smith, G.A. Bray, E. Ravussin, and Look AHEAD Adipose Research Group. 2013. Adipose tissue expression of adipose (WDTC1) gene is associated with lower fat mass and enhanced insulin sensitivity in humans. Obesity (Silver Spring). 21:2244–8. doi:10.1002/oby.20371. Groh, B.S., F. Yan, M.D. Smith, Y. Yu, X. Chen, and Y. Xiong. 2016. The antiobesity factor WDTC1 suppresses adipogenesis via the CRL4WDTC1 E3 ligase. EMBO Rep. 17:e201540500. doi:10.15252/embr.201540500. Häder, T., S. Müller, M. Aguilera, K.G. Eulenberg, A. Steuernagel, T. Ciossek, R.P. Kühnlein, L. Lemaire, R. Fritsch, C. Dohrmann, I.R. Vetter, H. Jäckle, W.W. Doane, and G. Brönner. 2003. Control of triglyceride storage by a WD40/TPR-domain protein. EMBO Rep. 4:511–6. doi:10.1038/sj.embor.embor837. Holland, A.J., D. Fachinetti, J.S. Han, and D.W. Cleveland. 2012. Inducible, reversible system for the rapid and complete degradation of proteins in mammalian cells. Proc. Natl. Acad. Sci. U. S. A. 109:E3350- 7. doi:10.1073/pnas.1216880109. Hsu, P.D., E.S. Lander, and F. Zhang. 2014. Development and Applications of CRISPR-Cas9 for Genome Engineering. Cell. 157:1262–1278. doi:10.1016/j.cell.2014.05.010. Jiang, W., and L.A. Marraffini. 2015. CRISPR-Cas: New Tools for Genetic Manipulations from Bacterial Immunity Systems. Annu. Rev. Microbiol. 69:209–228. doi:10.1146/annurev-micro-091014-104441. Jin, J., E.E. Arias, J. Chen, J.W. Harper, and J.C. Walter. 2006. A Family of Diverse Cul4-Ddb1-Interacting Proteins Includes Cdt2, which Is Required for S Phase Destruction of the Replication Factor Cdt1. Mol. Cell. 23:709–721. doi:10.1016/j.molcel.2006.08.010. Kubota, T., K. Nishimura, M.T. Kanemaki, and A.D. Donaldson. 2013. The Elg1 Replication Factor C-like Complex Functions in PCNA Unloading during DNA Replication. Mol. Cell. 50:273–280. doi:10.1016/j.molcel.2013.02.012. Li, T., E.I. Robert, P.C. van Breugel, M. Strubin, and N. Zheng. 2010. A promiscuous alpha-helical motif anchors viral hijackers and substrate receptors to the CUL4-DDB1 ubiquitin ligase machinery. Nat. Struct. Mol. Biol. 17:105–111. doi:10.1038/nsmb.1719. Maruyama, T., S.K. Dougan, M.C. Truttmann, A.M. Bilate, J.R. Ingram, and H.L. Ploegh. 2015. Increasing the efficiency of precise genome editing with CRISPR-Cas9 by inhibition of nonhomologous end joining. Nat. Biotechnol. 33:538–42. doi:10.1038/nbt.3190. Mellacheruvu, D., Z. Wright, A.L. Couzens, J.-P. Lambert, N.A. St-Denis, T. Li, Y. V Miteva, S. Hauri, M.E. Sardiu, T.Y. Low, V.A. Halim, R.D. Bagshaw, N.C. Hubner, A. al-Hakim, A. Bouchard, D. Faubert, D.

139

4. CRL4WDTC1 in cell cycle progression

Fermin, W.H. Dunham, M. Goudreault, Z.-Y. Lin, B.G. Badillo, T. Pawson, D. Durocher, B. Coulombe, R. Aebersold, G. Superti-Furga, J. Colinge, A.J.R. Heck, H. Choi, M. Gstaiger, S. Mohammed, I.M. Cristea, K.L. Bennett, M.P. Washburn, B. Raught, R.M. Ewing, A.-C. Gingras, and A.I. Nesvizhskii. 2013. The CRAPome: a contaminant repository for affinity purification–mass spectrometry data. Nat. Methods. 10:730–736. doi:10.1038/nmeth.2557. Mouysset, J., S. Gilberto, M.G. Meier, F. Lampert, M. Belwal, P. Meraldi, and M. Peter. 2015. CRL4RBBP7 is required for efficient CENP-A deposition at centromeres. J. Cell Sci. 7:1732–1745. doi:10.1242/jcs.162305. Natsume, T., T. Kiyomitsu, Y. Saga, and M.T. Kanemaki. 2016. Rapid Protein Depletion in Human Cells by Auxin-Inducible Degron Tagging with Short Homology Donors. Cell Rep. 15:210–218. doi:10.1016/j.celrep.2016.03.001. Nishimura, K., T. Fukagawa, H. Takisawa, T. Kakimoto, and M. Kanemaki. 2009. An auxin-based degron system for the rapid depletion of proteins in nonplant cells. Nat. Methods. 6:917–922. doi:10.1038/nmeth.1401. Nishino, M., Y. Kurasawa, R. Evans, S.H. Lin, B.R. Brinkley, and L. yuan Yu-Lee. 2006. NudC Is Required for Plk1 Targeting to the Kinetochore and Chromosome Congression. Curr. Biol. 16:1414–1421. doi:10.1016/j.cub.2006.05.052. Olma, M.H., M. Roy, T. Le Bihan, I. Sumara, S. Maerki, B. Larsen, M. Quadroni, M. Peter, M. Tyers, and L. Pintard. 2009. An interaction network of the mammalian COP9 signalosome identifies Dda1 as a core subunit of multiple Cul4-based E3 ligases. J. Cell Sci. 122:1035–44. doi:10.1242/jcs.043539. Park, A., S.T. Won, M. Pentecost, W. Bartkowski, and B. Lee. 2014. CRISPR/Cas9 allows efficient and complete knock-in of a destabilization domain-tagged essential protein in a human cell line, allowing rapid knockdown of protein function. PLoS One. 9:1–8. doi:10.1371/journal.pone.0095101. Piwko, W., L.J. Mlejnkova, K. Mutreja, L. Ranjha, D. Stafa, A. Smirnov, M.M. Brodersen, R. Zellweger, A. Sturzenegger, P. Janscak, M. Lopes, M. Peter, and P. Cejka. 2016. The MMS22L-TONSL heterodimer directly promotes RAD51-dependent recombination upon replication stress. EMBO J. 35:2584–2601. doi:10.15252/embj.201593132. Piwko, W., M.H. Olma, M. Held, J.N. Bianco, P.G. a Pedrioli, K. Hofmann, P. Pasero, D.W. Gerlich, and M. Peter. 2010. RNAi-based screening identifies the Mms22L-Nfkbil2 complex as a novel regulator of DNA replication in human cells. EMBO J. 29:4210–4222. doi:10.1038/emboj.2010.304. Ran, F.A., P.D. Hsu, J. Wright, V. Agarwala, D.A. Scott, and F. Zhang. 2013. Genome engineering using the CRISPR-Cas9 system. Nat. Protoc. 8:2281–2308. doi:10.1038/nprot.2013.143. Song, J., D. Yang, J. Xu, T. Zhu, Y.E. Chen, and J. Zhang. 2016. RS-1 enhances CRISPR/Cas9- and TALEN- mediated knock-in efficiency. Nat. Commun. 7:10548. doi:10.1038/ncomms10548. Suh, J.M., D. Zeve, R. McKay, J. Seo, Z. Salo, R. Li, M. Wang, and J.M. Graff. 2007. Adipose Is a Conserved Dosage-Sensitive Antiobesity Gene. Cell Metab. 6:195–207. doi:10.1016/j.cmet.2007.08.001. Zhou, T., J.P. Aumais, X. Liu, L.Y. Yu-Lee, and R.L. Erikson. 2003. A role for Plk1 phosphorylation of NudC in cytokinesis. Dev. Cell. 5:127–138. doi:10.1016/S1534-5807(03)00186-2. Zitouni, S., C. Nabais, S.C. Jana, A. Guerrero, and M. Bettencourt-Dias. 2014. Polo-like kinases: structural variations lead to multiple functions. Nat. Rev. Mol. Cell Biol. 15:433–452. doi:10.1038/nrm3819.

140

5. Extended discussion and future perspectives

In this thesis, I presented the product of a collective effort to identify novel cell cycle functions of CRL4 complexes. I believe that we are moving one step forward in the understanding of how ubiquitylation regulates cell division, as we unveiled a novel mechanism involved in centromere and thus kinetochore assembly (chapter 2), a non-redundant role for the two paralog cullins (chapter 3) and propose yet another potential mode of mitotic regulation by CRL4 (chapter 4). Nevertheless, these associations still demand further mechanistic exploration, so that we can expand the understanding of the mechanisms governing the cell cycle. In this chapter I focus on some of the main conclusions and implications of our work and investigate how they connect with the observations made by others, while suggesting topics for future exploration. In chapter 2, we described a mechanism how CRL4 can have indirect, time-separated repercussions on cell division by regulating centromeric maintenance. Because RBBP7 is also part of several methyltransferase complexes (Loyola and Almouzni, 2004; Vaute et al., 2002; Kuzmichev et al., 2002), our observations also come to support an additional intriguing feature of CRL4 complexes: it appears that substrate adaptors are in several cases not dedicated to exclusively fulfil a function associated with CRL4, but take part in CRL4-independent cellular roles. This is indeed the case for another methyltransferase subunit, WDR5 (Higa et al., 2006). VPRBP (DCAF1) is also a protein kinase that can phosphorylate histone H2A and negatively influence gene transcription (Kim et al., 2013). Moreover, the substrate adaptor COP1 can in fact also induce ubiquitylation of substrates on its own, because it is itself a RING E3 ligase (Suriben et al., 2015). Just like for CUL4RBBP7, what determines this functional shift remains to be explored. One direction that could be followed for the case of RBBP7 is the search for functional regions that support one function but not the other. Otherwise, perhaps modulation of its WD40 domain by post-translational modifications redefines its interactions and thus function. Another possibility is that both functions are inter-dependent, indeed supported by the joint action of CRL4B with PRC2 or SIN3A- HDAC in transcription repression (both complexes contain RBBP7 as a core subunit (Loyola and Almouzni, 2004; Vaute et al., 2002; Kuzmichev et al., 2002)). Like in chapter 2, chapter 4 aims to investigate the action of yet another obscure CRL4 complex in the cell cycle. Nevertheless, whether CRL4WDTC1 has a function at this stage remains to be clarified, because functional studies rather propose that this is not the case. Rather, CRL4WDTC1 has been associated with fat metabolism due to a role in gene repression (Groh et al., 2016). However, while the function of WDTC1

141

5. Extended discussion and future perspectives remains to be clarified, our interactome data argues that it could act in the regulation of mitosis, albeit not with an essential role. However, the knockout in mice is most often lethal (Suh et al., 2007). Consistent with this, a more general role of WDTC1 than the regulation of fat metabolism would also justify the observations that its expression (evaluated by mRNA and protein levels) is widespread in human tissues, with prevalence for the testis, parathyroid gland and retina, according to the Human protein atlas project. Indeed, the original studies in flies suggested that the observed influence of WDTC1 (ADP) on the fat body appears rather to be due to its influence on the endocrine system (Doane, 1960a; b, 1961). If this role is conserved, one should investigate the function of WDTC1 outside of adipose tissue, and examine if WDTC1 acts in the modulation of endocrine functions in the testis or the parathyroid gland, where its expression is upregulated. In chapter 3, I provide evidence for a novel mechanism of mitotic E3 ligase regulation, where it is the phosphorylation of the cullin that changes its targets. It is still unclear whether phosphorylation alone is responsible for this change, or whether chromatin exclusion also contributes. Mitotic phosphorylation is a known regulatory mode for another major cell cycle E3 ligase – the APC/C (Kramer et al., 2000; Golan et al., 2002; Qiao et al., 2016; Zhang et al., 2016; Fujimitsu et al., 2016) – and our study reveals that regulating the E3 ligase scaffold (and not only the adaptor) is indeed a method of choice used by the cell to timely direct E3 ligase activity. Several questions remain open, including 1) what is precisely the time of phosphorylation and dephosphorylation of CUL4B (and thus, of its functional repurposing), 2) which precise function does CRL4B do, and 3) whether this function leads to syndromic XLID or mice embryonic lethality when disrupted. Answers for these questions might be provided by interactome studies. In fact, we performed CUL4B immunoprecipitation-mass spectrometry in collaboration with the Aebersold group, though so far we couldn’t validate these interactions. Nevertheless, we found that CUL4B might interact with PPP1R12A, a regulatory subunit of protein phosphatase 1 (PP1). Interestingly, PP1-PPP1R12A has been involved in the dephosphorylation of PLK1 at the end of mitosis (Ma and Poon, 2011), which could indicate that this is the phosphatase that acts on CUL4B during mitotic exit. Nevertheless, this hypothesis demands further exploration, with better characterization of CUL4B’s interactome and functional analysis of potential interactions.

5.1. Mitotic function of CUL4B: a molecular explanation for XLID? To date, mutations in more than 700 human genes have been involved in intellectual disability (ID) and ID-associated disorders. One subset are gene alleles located in the X chromosome, and which includes CUL4B (Vissers et al., 2015). Despite some suggestions on how specific CRL4/CRL4B functions

142

5. Extended discussion and future perspectives could be implicated in XLID (Kerzendorfer et al., 2011; Nakagawa and Xiong, 2011), described mechanisms have only loosely been connected to disease development. For most of the functions, an involvement of CUL4A was not thoroughly addressed, and there’s little explanation as to why CUL4A cannot compensate for the loss of CUL4B, in particular because we find that CUL4A and CUL4B to localize to the nucleus indistinctly. In fact, we believe that the discovery of a XLID mutation in the unique N-terminal region of CUL4B (P50L) that is correlated to the disease is evidence for a functional distinction that resides within this N-terminus. Admittedly, so far we did not find evidence for an overt phenotype resulting from the expression of the P50L mutant. It is important to clarify whether the discovered mitotic function of CUL4B is indeed connected with the mechanisms how syndromic XLID comes about. To this end, exploring the P50L mutant will be key, because it might define the separation-of-function between a N-terminal specific role involved in XLID and the N-terminal-independent roles of CUL4B.

5.2. CUL4B loss-of-function in mouse embryonic lethality, syndromic XLID and mitosis: how do they come together? The establishment of mouse knockout models revealed that while CUL4B can compensate for the absence of CUL4A (Kopanja et al., 2011; Yin et al., 2011), the opposite is not true. Cul4b knockout is lethal in male mice (Cul4b–/Y), with a severe growth impairment discernable in E7.5 embryos, which die before E9.5 (Liu et al., 2012; Jiang et al., 2012). The reason of cell death hasn’t been fully determined, but Liu et al. report a marked increase in cells positive for phosphorylated Histone H3 S10 (pH3), indicative of a G2/M arrest (Liu et al., 2012). Growth impairment, increased pH3 and apoptosis are observed in embryonic and extraembryonic tissues. In light with our observations of an independent mitotic function of human CUL4B, we suggest that the observed growth impairments are, at least in part, due to this newly- identified function. Importantly, due to the presence of a serine-rich region in mouse CUL4B, we indeed expect this function to be conserved (Appendix B). To test this hypothesis, one should clarify whether lethality in mice is due to a specific function of CUL4B, instead of simply the lack of CUL4A. To this end, we suggest the generation of Cul4b knockout mice that also express Cul4a under control of the endogenous Cul4b promoter. If Cul4a expression does not rescue the Cul4b knockout phenotype, we believe that this would be clear evidence arguing for a functional distinction between the cullins in mouse embryogenesis.

143

5. Extended discussion and future perspectives

5.2.1. CUL4B: essential for mitotic fidelity during gastrulation? Considering that CUL4B is indeed important for cell proliferation (our own observations), this raises the question why can Cul4b be deleted around E6.5 (Sox2-Cre) without overt phenotypes in the remaining developmental stages. As a side note, we would like to highlight that the deletion of the gene does not reflect immediate gene product depletion: for example, DDB1 deletion in mouse embryonic fibroblasts is reflected in dramatic DDB1 level drop by 4 days after Cre transfection (with an observed concomitant growth delay from day 3 onwards) (Cang et al., 2006). The observation that the Cul4b gene can be deleted at E6.5 stage implies that the critical function of CUL4B, in mice and perhaps in humans, takes place latest at this stage of embryogenesis. While it was suggested that this is due to critical roles in extraembryonic tissue development (Liu et al., 2012), evidence suggests that CUL4B is equally important for the development of the embryo. It is important to mention that at E6.5, the embryo dramatically increases its cell number to form the gastrula, which contains the three germ layers (termed gastrulation) (Mac Auley et al., 1993; O’Farrell et al., 2004; Farrell and O’Farrell, 2014). In fact, gastrulation fails to occur in knockout mutants (Liu et al., 2012; Jiang et al., 2012). Indeed, loss of the extraembryonic ectoderm has been determined to prevent gastrulation (Donnison et al., 2005), supporting the hypothesis suggested by Liu et al. However, in the event of loss of the extraembryonic ectoderm, embryos are not detected only by E12.5 (Donnison et al., 2005), whereas they couldn’t be detected in CUL4B knockout mice anymore by E9.5 (Liu et al., 2012), suggesting a more direct role in maintaining the viability of the embryo. Moreover, heterozygous knockout mice where the WT Cul4b allele is in the paternal X chromosome, inactivated in mice extraembryonic tissues, die by E12.5-E13.5 and not before E9.5 as their hemizygous Cul4B–/Y counterparts. This is indicative that the complete absence of CUL4B from the embryo is additive to the phenotype arising from impaired extraembryonic tissue proliferation. Finally, Cul4b-null embryonic cells are selected against (in favour of cells that do not inactivate the X chromosome containing the Cul4b WT allele), meaning that its deletion is not innocuous in these tissues. Taken together, we suggest that CUL4B is essential for the proliferation of extraembryonic tissues but also for the proliferation of embryonic cells. This proliferation defect would be compatible with our observed mitotic-specific function. Importantly, in humans it does not appear to be critical for extraembryonic tissue proliferation, though the embryo proliferative function may be maintained. Perhaps the disruption of such an early developmental stage justifies the observation that loss of CUL4B results in syndromic XLID, where multiple organs are affected. It is important to mention the curious observation that the deletion of other human XLID genes in mice also prevents embryonic development (Cox et al., 2010).

144

5. Extended discussion and future perspectives

The proposed hypothesis does not clarify why would CUL4B be so important at this stage of embryonic development and not for other stages. The reasoning might lie within the length of the cell cycle upon gastrulation. In mammals, the cell cycle accelerates and exceedingly rapid cycles are observed (each lasting only 2-3h in mice), reminiscent of the very fast early cleavage cycles in non-amniotic metazoans (Duncan and Tin Su, 2004; O’Farrell et al., 2004; Farrell and O’Farrell, 2014). We speculate that effects of Cul4b disruption are amplified in these fast-cycling cells, resulting in cell division defects and induction of apoptosis, as observed (Liu et al., 2012; Jiang et al., 2012). In order to test the hypothesis that CUL4B is also directly involved in embryonic cell proliferation, we suggest performing the Cul4b knockout before gastrulation (E5.5, for instance by using an inducible CRE promoter (Hayashi and McMahon, 2002)). Moreover, performing the knockout only in extraembryonic tissues would be equally informative, for example by using Cdx1-CRE (Hierholzer and Kemler, 2009).

5.2.2. Spermatogenesis: is CUL4B function common to mitosis and meiosis? Remarkably, a hypogonadism phenotype reminiscent of observations in human males with CUL4B mutations was noticed in mice, provided that the knockout had been induced at the embryonic stage (Sox2-Cre). The reasoning for the observed depletion of the stem cell pool upon Cul4b knockout was not determined, but careful assessment indicated that Sertoli, Leydig or macrophage cell pools (all involved in stem cell self-renewal) are not affected (Yin et al., 2016). It was also observed that spermatids (the haploid product cells of meiosis II) often underwent apoptosis and displayed aberrant acrosome and nuclear morphology (Lin et al., 2016). In the detected sperm cells, heads were misshapen and nuclei were often disintegrated or even absent. Although it was suggested that these phenotypes are due to impaired acrosome formation, nuclear morphogenesis and post-mitotic histone turnover, we also consider (in light with our data) that both the failure in maintaining the stem cell niche and the spermatid/sperm cell aberrations arise from accumulated cell division issues and subsequent compromised genome integrity. In line with these conclusions, it is important to mention that general protein degradation is prominent during the transition from the pachytene (meiosis prophase I) to the round spermatid stage (i.e. from the end of meiotic prophase I to completion of meiosis II), but not before (Gan et al., 2013). Hence, it is possible that CUL4B would take part as a contributing E3 ligase during this transition. Interestingly, CUL4B is predicted to be absent at the pachytene-diplotene stages of prophase I due to meiotic sex chromosome inactivation (Kopanja et al., 2011; Yin et al., 2011). Although some X chromosome genes do remain repressed for a longer time, de novo transcription can occur (Turner, 2007), and the CUL4B gene can potentially be derepressed by the end of prophase I. This effect would implicate

145

5. Extended discussion and future perspectives absence of CUL4B function at the pachytene stage of prophase I, but CUL4B could still function afterwards, towards the end of prophase I and during meiosis II. Indeed, CUL4B expression is observed in round spermatids, the earliest post-meiotic cells (stage P20), consistent with its early derepression, though protein levels during meiosis II itself weren’t determined (Lin et al., 2016). Accordingly, it was observed that several early post-meiotic round spermatids undergo apoptosis in the absence of Cul4b, consistent with an impairment just before this stage, i.e. during meiosis. Taken together, we speculate that CUL4B regulates not only mitosis, but perhaps also progression throughout meiosis, and might regulate the same processes in both: at the end of prophase I, CUL4B could similarly regulate a late prophase I event such nuclear envelope breakdown. In support of our remark, kinases that phosphorylate CUL4B are equally active at this stage (Bennabi et al., 2016). It is additionally remarkable that a possible meiotic role for CUL4B also cannot be compensated by CUL4A, because CUL4A is present and functions during mouse meiosis (Kopanja et al., 2011; Yin et al., 2011). Our observations of CUL4B phosphorylation in the regulation of cell division could equally provide an explanation for this effect. To explore this possibility, it would be of prime interest to examine whether CUL4B is also phosphorylated during meiosis. Because the repression of its expression during some prophase I stages could also indicate that its function is deleterious (yet sex chromosome inactivation is of course not exclusively directed to modulate CUL4B function), an interesting experiment would be to constitutively express CUL4B throughout prophase I, for example by making use of the CUL4A promoter.

5.3. References

Mac Auley, A., Z. Werb, and P.E. Mirkes. 1993. Characterization of the unusually rapid cell cycles during rat gastrulation. Development. 117. Bennabi, I., M.-E. Terret, and M.-H. Verlhac. 2016. Meiotic spindle assembly and chromosome segregation in oocytes. J. Cell Biol. Cang, Y., J. Zhang, S.A. Nicholas, J. Bastien, B. Li, P. Zhou, and S.P. Goff. 2006. Deletion of DDB1 in Mouse Brain and Lens Leads to p53-Dependent Elimination of Proliferating Cells. Cell. 127:929–940. doi:10.1016/j.cell.2006.09.045. Cox, B.J., M. Vollmer, O. Tamplin, M. Lu, S. Biechele, M. Gertsenstein, C. van Campenhout, T. Floss, R. Kuhn, W. Wurst, H. Lickert, and J. Rossant. 2010. Phenotypic annotation of the mouse X chromosome. Genome Res. 20:1154–1164. doi:10.1101/gr.105106.110. Doane, W.W. 1960a. Developmental Physiology of the Mutant Female Sterile(2)Adipose of Drosophila melanogaster. I. Adult morphology, longevity, egg production, and egg lethality. J. Exp. Zool. 145:1– 21. doi:10.1002/jez.1401450102. Doane, W.W. 1960b. Developmental physiology of the mutant female sterile(2)adipose of Drosophila melanogaster. II. Effects of altered environment and residual genome on its expression. J. Exp. Zool.

146

5. Extended discussion and future perspectives

145:23–41. doi:10.1002/jez.1401450103. Doane, W.W. 1961. Developmental physiology of the mutant female sterile(2)adipose of Drosophila melanogaster. III. Corpus allatum-complex and ovarian transplantations. J. Exp. Zool. 146:275–298. doi:10.1002/jez.1401460307. Donnison, M., A. Beaton, H.W. Davey, R. Broadhurst, P. L’Huillier, and P.L. Pfeffer. 2005. Loss of the extraembryonic ectoderm in Elf5 mutants leads to defects in embryonic patterning. Development. 132:2299–2308. doi:10.1242/dev.01819. Duncan, T., and T. Tin Su. 2004. Embryogenesis: Coordinating Cell Division with Gastrulation. Curr. Biol. 14:R305–R307. doi:10.1016/j.cub.2004.03.050. Farrell, J.A., and P.H. O’Farrell. 2014. From Egg to Gastrula: How the Cell Cycle Is Remodeled During the Drosophila Mid-Blastula Transition. Annu. Rev. Genet. 48:269–294. doi:10.1146/annurev-genet- 111212-133531. Fujimitsu, K., M. Grimaldi, and H. Yamano. 2016. Cyclin-dependent kinase 1-dependent activation of APC/C ubiquitin ligase. Science. 352:1121–4. doi:10.1126/science.aad3925. Gan, H., T. Cai, X. Lin, Y. Wu, X. Wang, F. Yang, and C. Han. 2013. Integrative proteomic and transcriptomic analyses reveal multiple post-transcriptional regulatory mechanisms of mouse spermatogenesis. Mol. Cell. Proteomics. 12:1144–57. doi:10.1074/mcp.M112.020123. Golan, A., Y. Yudkovsky, and A. Hershko. 2002. The Cyclin-Ubiquitin Ligase Activity of Cyclosome/APC Is Jointly Activated by Protein Kinases Cdk1-Cyclin B and Plk. J. Biol. Chem. 277:15552–15557. doi:10.1074/jbc.M111476200. Groh, B.S., F. Yan, M.D. Smith, Y. Yu, X. Chen, and Y. Xiong. 2016. The antiobesity factor WDTC1 suppresses adipogenesis via the CRL4WDTC1 E3 ligase. EMBO Rep. 17:e201540500. doi:10.15252/embr.201540500. Hayashi, S., and A.P. McMahon. 2002. Efficient Recombination in Diverse Tissues by a Tamoxifen-Inducible Form of Cre: A Tool for Temporally Regulated Gene Activation/Inactivation in the Mouse. Dev. Biol. 244:305–318. doi:10.1006/dbio.2002.0597. Hierholzer, A., and R. Kemler. 2009. Cdx1::Cre allele for gene analysis in the extraembryonic ectoderm and the three germ layers of mice at mid-gastrulation. genesis. 47:204–209. doi:10.1002/dvg.20484. Higa, L.A., M. Wu, T. Ye, R. Kobayashi, H. Sun, and H. Zhang. 2006. CUL4-DDB1 ubiquitin ligase interacts with multiple WD40-repeat proteins and regulates histone methylation. Nat. Cell Biol. 8:1277–83. doi:10.1038/ncb1490. Jiang, B., W. Zhao, J. Yuan, Y. Qian, W. Sun, Y. Zou, C. Guo, B. Chen, C. Shao, and Y. Gong. 2012. Lack of Cul4b, an E3 ubiquitin ligase component, leads to embryonic lethality and abnormal placental development. PLoS One. 7. doi:10.1371/journal.pone.0037070. Kerzendorfer, C., L. Hart, R. Colnaghi, G. Carpenter, D. Alcantara, E. Outwin, A.M. Carr, and M. O’Driscoll. 2011. CUL4B-deficiency in humans: Understanding the clinical consequences of impaired Cullin 4- RING E3 ubiquitin ligase function. Mech. Ageing Dev. 132:366–373. doi:10.1016/j.mad.2011.02.003. Kim, K., J.-M. Kim, J.-S. Kim, J. Choi, Y.S. Lee, N. Neamati, J.S. Song, K. Heo, and W. An. 2013. VprBP Has Intrinsic Kinase Activity Targeting Histone H2A and Represses Gene Transcription. Mol. Cell. 52:459– 467. doi:10.1016/j.molcel.2013.09.017. Kopanja, D., N. Roy, T. Stoyanova, R.A. Hess, S. Bagchi, and P. Raychaudhuri. 2011. Cul4A is essential for spermatogenesis and male fertility. Dev. Biol. 352:278–287. doi:10.1016/j.ydbio.2011.01.028. Kramer, E.R., N. Scheuringer, A. V Podtelejnikov, M. Mann, and J.M. Peters. 2000. Mitotic regulation of the APC activator proteins CDC20 and CDH1. Mol. Biol. Cell. 11:1555–69.

147

5. Extended discussion and future perspectives

doi:10.1091/mbc.11.5.1555. Kuzmichev, A., K. Nishioka, H. Erdjument-Bromage, P. Tempst, and D. Reinberg. 2002. Histone methyltransferase activity associated with a human multiprotein complex containing the Enhancer of Zeste protein. Genes Dev. 16:2893–2905. doi:10.1101/gad.1035902. Lin, C.-Y., C.-Y. Chen, C.-H. Yu, I.-S. Yu, S.-R. Lin, J.-T. Wu, Y.-H. Lin, P.-L. Kuo, J.-C. Wu, and S.-W. Lin. 2016. Human X-linked Intellectual Disability Factor CUL4B Is Required for Post-meiotic Sperm Development and Male Fertility. Sci. Rep. 6:20227. doi:10.1038/srep20227. Liu, L., Y. Yin, Y. Li, L. Prevedel, E.H. Lacy, L. Ma, and P. Zhou. 2012. Essential role of the CUL4B ubiquitin ligase in extra-embryonic tissue development during mouse embryogenesis. Cell Res. 22:1258–1269. doi:10.1038/cr.2012.48. Loyola, A., and G. Almouzni. 2004. Histone chaperones, a supporting role in the limelight. Biochim. Biophys. Acta. 1677:3–11. doi:10.1016/j.bbaexp.2003.09.012. Ma, H.T., and R.Y.C. Poon. 2011. How protein kinases co-ordinate mitosis in animal cells. Biochem. J. 435. Nakagawa, T., and Y. Xiong. 2011. X-Linked Mental Retardation Gene CUL4B Targets Ubiquitylation of H3K4 Methyltransferase Component WDR5 and Regulates Neuronal Gene Expression. Mol. Cell. 43:381–391. doi:10.1016/j.molcel.2011.05.033. O’Farrell, P.H., J. Stumpff, T.T. Su, P.P. Tam, M.C. Ellis, T.B. Kornberg, C.S. Goodman, K. Nakayama, M. Nakanishi, A. Bradley, and et al. 2004. Embryonic cleavage cycles: how is a mouse like a fly? Curr. Biol. 14:R35-45. doi:10.1016/J.CUB.2003.12.022. Qiao, R., F. Weissmann, M. Yamaguchi, N.G. Brown, R. VanderLinden, R. Imre, M.A. Jarvis, M.R. Brunner, I.F. Davidson, G. Litos, D. Haselbach, K. Mechtler, H. Stark, B.A. Schulman, and J.-M. Peters. 2016. Mechanism of APC/CCDC20 activation by mitotic phosphorylation. Proc. Natl. Acad. Sci. U. S. A. 113:E2570–E2578. doi:10.1073/pnas.1604929113. Suh, J.M., D. Zeve, R. McKay, J. Seo, Z. Salo, R. Li, M. Wang, and J.M. Graff. 2007. Adipose Is a Conserved Dosage-Sensitive Antiobesity Gene. Cell Metab. 6:195–207. doi:10.1016/j.cmet.2007.08.001. Suriben, R., K.A. Kaihara, M. Paolino, M. Reichelt, S.K. Kummerfeld, Z. Modrusan, D.L. Dugger, K. Newton, M. Sagolla, J.D. Webster, J. Liu, M. Hebrok, and V.M. Dixit. 2015. β-Cell Insulin Secretion Requires the Ubiquitin Ligase COP1. Cell. 163:1457–1467. doi:10.1016/j.cell.2015.10.076. Turner, J.M.A. 2007. Meiotic sex chromosome inactivation. Development. 134:1823–1831. doi:10.1242/dev.000018. Vaute, O., E. Nicolas, L. Vandel, and D. Trouche. 2002. Functional and physical interaction between the histone methyl transferase Suv39H1 and histone deacetylases. Nucleic Acids Res. 30:475–81. Vissers, L.E.L.M., C. Gilissen, and J.A. Veltman. 2015. Genetic studies in intellectual disability and related disorders. Nat. Rev. Genet. 17:9–18. doi:10.1038/nrg3999. Yin, Y., C. Lin, S.T. Kim, I. Roig, H. Chen, L. Liu, G.M. Veith, R.U. Jin, S. Keeney, M. Jasin, K. Moley, P. Zhou, and L. Ma. 2011. The E3 ubiquitin ligase Cullin 4A regulates meiotic progression in mouse spermatogenesis. Dev. Biol. 356:51–62. doi:10.1016/j.ydbio.2011.05.661. Yin, Y., L. Liu, C. Yang, C. Lin, G.M. Veith, C. Wang, P. Sutovsky, P. Zhou, and L. Ma. 2016. Cell autonomous and non-autonomous function of CUL4B in Mouse Spermatogenesis. 291. jbc. M115.699660 pp. Zhang, S., L. Chang, C. Alfieri, Z. Zhang, J. Yang, S. Maslen, M. Skehel, and D. Barford. 2016. Molecular mechanism of APC/C activation by mitotic phosphorylation. Nature. 533:260–264. doi:10.1038/nature17973.

148

6. Appendices

A. Complementary data to chapter 2

A.1. Optimization of immunoprecipitation assays

Figure A.1-1. Optimization of immunoprecipitation assays. To visualize co-precipitating endogenous RBBP7 in endogenous CUL4A immunoprecipitation assays, buffer or lysis conditions were altered as indicated (while the remaining conditions, described in chapter 3 materials and methods were maintained. Alternatively, the buffer described in (He et al. 2006)* was used. “–“, immunoprecipitation using unspecific IgG coupled to affiprep beads; “+” Immunoprecipitation using anti-RBBP7 or anti-CUL4A antibodies, as indicated.

*He, Y.J., C.M. McCall, J. Hu, Y. Zeng, and Y. Xiong. 2006. DDB1 functions as a linker to recruit receptor WD40 proteins to CUL4-ROC1 ubiquitin ligases. Genes Dev. 20:2949–2954. doi:10.1101/gad.1483206

149

6. Appendices

B. Complementary data to chapter 3

B.1. Conservation of CUL4B protein

Figure B.1-1. Multiple sequence alignment (Clustal) of the gene products of CUL4B orthologs from several vertebrates. The region of high homology with CUL4A begins at amino acid 195 of the human CUL4B sequence (isoform 1), position 256 in the alignment. The shades of grey indicate conservation (darker color for less conserved).

150

6. Appendices

B.2. Analysis of CUL4 interactions by mass spectrometry

Figure B.2-1. Analysis of FLAG-CUL4A vs FLAG-CUL4B interactions in S-phase. HeLa cell lines expressing these proteins were arrested in S-phase with thymidine and used in immunoprecipitation assays. Co-precipitating proteins were analyzed by mass spectrometry, in collaboration with Tatjana Sajic and Ruedi Aebersold. The method used is as described in chapter 4, but no DTSSP was added. Protein identifications were analyzed with the CRAPome web tool (crapome.org), with built-in significance analysis of interactome (SAINT) analysis which assigns a confidence/probability score for the interaction. The FC-B score estimates the fold change in comparison to negative controls. The Venn diagram was created with Venny 2.1 (bioinfogp.cnb.csic.es/tools/venny).

151

6. Appendices

Figure B.2-2. Proteins identified in the datasets presented in figure B.2.1. SP, SAINT probability.

152

6. Appendices

Figure B.2-3. Analysis of FLAG-CUL4A vs FLAG-CUL4B interactions in mitosis. HeLa cell lines expressing these proteins were arrested in mitosis with nocodazole and used in immunoprecipitation assays, as described in Figure B.2-1.

153

6. Appendices

Figure B.2-4. Proteins identified in the indicated datasets presented in Figure B.2-3. SP, SAINT probability. Please refer for figure B.2.5 for proteins identified exclusively upon CUL4B IP.

154

6. Appendices

Figure B.2-5. Proteins identified in the CUL4B dataset presented in Figure B.2-3. SP, SAINT probability.

155

6. Appendices

Figure B.2-6. Analysis of FLAG-CUL4B interactions: S-phase vs mitosis. For the control samples, no construct was expressed. Here, the FC-B score is compared, which estimates the fold change in comparison to negative controls. The Venn diagram was created with Venny 2.1. Green, known CUL4B interacting partners.

156

6. Appendices

B.3. Obtainment of CUL4B knockout clones

Figure B.3-1. FACS sorting of cells transfected with pX458 (Cas9-2A-GFP) used for CUL4B gene deletion. GFP-positive cells were seeded and hand-picked two weeks afterwards, followed by testing of the absence of CUL4B protein by western blot. The method is described in chapter 3, materials and methods section.

157

6. Appendices

B.4. CRISPR knock-in of CUL4B mutants

Figure B.4-1. PCR testing of CUL4B gene editing with CUL4B mutants and imaging of obtained clones (see chapter 3 – the strategy applied for the mutants is the same as for the WT). Cytokinesis defects were observed for several obtained clones (with separation of nuclei). Because this was found in a heterozygous WT knock-in clone (clone 7) whose function is disrupted due to flawed DNA repair (see figure B.4.2), we conclude that these phenotypes are due to loss-of-function. In the Pmimic-NT clone a strong phenotype was not observed, possibly due to the low expression (as compared to the overexpression assays presented in chapter 3).

158

6. Appendices

Figure B.4-2. Sequencing of the two observed bands in the WT clone 7. Top, the upper band yielded a sequence that indicates several products from the PCR, possibly none is a correct allele editing (though this clone survived puromycin selection). Bottom, the endogenous band of this clone is not identical to the WT, as it contains a deletion of four nucleotides. The subsequent frameshift creates a premature truncated product.

159

6. Appendices

B.5. Optimization of immunofluorescence with pre-extraction

Figure B.5-1. Optimization of the pre-extraction conditions for the immunofluorescence presented in chapter 3. Cells were fixed for 10 min with 4% formaldehyde/paraformaldehyde before or after incubation with extraction buffer. Extraction was performed for 10 min or for the indicated times in case of pre-extraction before fixation. Metaphase cells were imaged for the optimization, for which one representative cell is shown. CUL4B is mostly absent from metaphase chromatin when using 2 or 4 min pre-extraction. Because less background was obtained for 4 min treatment, this condition was used in chapter 3.

160

6. Appendices

C. Complementary data to chapter 4

C.1. Optimization of co-transfection conditions for CRISPR knock-in

Figure C.1-1. Transfection optimization for CRISPR knock-in of the 3miniAID degron in the endogenous WDTC1 loci. Co-transfection of the pX458-gRNA and DNA template vector was performed with the indicated conditions (see chapter 4). Cells were either transfected with lipofectamine 3000 (Invitrogen) or with an Amaxa Nucleofector II (Lonza), according to the manufacturer’s instructions. Knock-in efficiency was evaluated on a GFP-positive population, 1 week after sorting. Please note the absence of upshifted bands for cells transfected with the nucleofector, indicating that following this protocol does not support gene editing in this case.

161

6. Appendices

C.2. Development of an anti-WDTC1 antibody

Figure C.2-1. An anti-WDTC1 antibody was produced in rabbit (Eurogenetec) using their standard peptide antibody generation approaches. The antibody was purified from rabbit serum using the antigen peptide coupled to AminoRLink Plus Coupling Resin (Thermo Scientific). Top, several elution methods employed for the elution of the antibody were used, as indicated, and compared for their capacity to detect purified HA-tagged WDTC1. A commercial antibody (Sigma) was also tested, but did not detect purified WDTC1. Bottom panels, the MgCl2-eluted antibody was tested for endogenous WDTC1 binding but no clear band was perceived.

162

6. Appendices

C.3. CMV expression pattern in the cell cycle

Figure C.3-1. The levels of WDTC1, expressed under a CMV promoter, oscillate throughout the cell cycle in a proteasome-independent manner. Cells from an inducible stable HeLa cell line expressing C-terminally TAP-tagged WDTC1 (strep-strep-HA) were synchronized via double-thymidine block/release at the indicated cell cycle stages and analyzed by (A) flow cytometry to verify cell synchrony and (B) western blot for WDTC1 levels. (C) Western blot analysis on an identical experiment performed in the presence of the proteasome inhibitor MG132. Async, asynchronous. S, S-phase.

163

6. Appendices

D. ImageJ macros written for this study

D.1. Z-projection and color merge for multiple-position immunofluorescence acquisitions

// Open "superfolder" e.g. folderA": folderA > folderB1 > Pos0 // Saves channel files individually, plus RGB // ------// ------* Input the number of channels used * ------// ------nChannels = 3; // Number of channels

// ------setBatchMode(true); dir = getDirectory("Input directory "); print("\nMother folder selected: " + dir+"\n"); listA = getFileList(dir); for (b=0; b

164

6. Appendices

D.2. Automated Z-projection for live-cell imaging experiments

// Open "superfolder" e.g. folderA": folderA > folderB1 > Pos0 // Saves SiR-DNA only // Uses Grouped Z projector plugin

// ------// ------* Input the number of channels and Z-slices used * ------// ------nChannels = 2; // Number of channels nSlices_time = 6; //Number of slices per time point

// ------setBatchMode(true); dir = getDirectory("Input directory "); print("\nMother folder selected: " + dir+"\n"); listA = getFileList(dir); for (b=0; b

165

6. Appendices

D.3. Automated cropping of multi-channel images and color merge

//Welcome to ROItify //This macro loads selections stored in ROI Manager and saves them as separate images, in all 2 channels //INPUT must be a hyperstack with 2 channels //ALL selections must be made in the DAPI channel, which must be the first channel of the 2 nChannels = 2 //Number of channels - order: DAPI, RFP dir=getDirectory("image"); File.makeDirectory(dir+"/Crop"); sub=getInfo("image.subtitle"); za=indexOf(sub,"z:"); zz=indexOf(sub,"/",za); z=substring(sub,za+2,zz); //retrieves the Z (current slice number) pa=indexOf(sub,"("); pz=indexOf(sub,"_"); coreP=substring(sub,pa+1,pz); //retrieves the current Pos number title = getTitle(); n = roiManager("count"); print("\\Clear"); print("Welcome to ROItify"); print("\nFile: "+title); print("Slice: " + z + " ("+coreP+")"); print(n +" selections"); if (n==0){ exit("Cancelled: The ROI Manager is empty");} for (i=0; i

selectWindow(coreP+"_DAPI"); run("Grays"); resetMinAndMax(); run("8-bit"); saveAs("png", dir+"/Crop/" + coreP + "_" + i+1 + "_4DAPI"); rename("Proj. 1");

selectWindow(coreP+"_RFP"); run("Grays"); resetMinAndMax(); run("8-bit"); saveAs("png", dir+"/Crop/" + coreP + "_" + i+1 + "_3RFP"); rename("Proj. 3");

run("Merge Channels...", "red=[Proj. 3] green=[Proj. 1] blue=*None* grey=*None*"); saveAs("png", dir+"/Crop/" + coreP + "_" + i+1 + "_1RGB"); close(coreP + "_" + i+1 + "_1RGB.png");

close("Proj. 1"); close("Proj. 3"); } for (i=0; i

166

6. Appendices

D.4. Live imaging time-point saver Takes a live-cell imaging hyperstack and saves individual images with the selected time interval; Also adds the scale bar and stamps the time on the image

//Live-cell imaging slicer //Takes time-points of live-cell imaging experiments and saves as separate files //Input is a hyperstack: RGB composite, with Z information //Optional: adds the scale bar (important: first set the scale as global)

Time=30; //Time (in min) between each time-point interval=6; //Spacing in-between time-points to be saved. In number of time-points, not time

title = getTitle(); dir=getDirectory("image"); sub=getInfo("image.subtitle"); za=indexOf(sub,"z:"); zend=indexOf(sub,";"); zz=indexOf(sub,"/",za); maxzz=substring(sub,zz+1,zend); pz=indexOf(title,".tif"); coreP=substring(title,0,pz); //retrieves the current File

print("Image: " + coreP + "; Max position: " +maxzz); run("Stack to RGB", "slices keep"); //Reduces the dimensionality rename("StacktoRGB"); run("Time Stamper", "starting=0 interval=30 x=5 y=45 font=36 '00 decimal=0 anti-aliased or=h"); //Time-stamper

File.makeDirectory(dir+"/"+coreP); n=0;

while (interval*n

//------Scale bar------

run("Scale Bar...", "width=10 height=3 font=14 color=White background=None location=[Lower Right] hide");

//------

saveAs("png", dir+"/"+coreP+"/"+coreP+"_"+interval*n+1); print("Saved "+ coreP+"_"+interval*n+1); selectWindow(coreP+"_"+interval*n+1+".png"); close(); n=n+1; } selectWindow("StacktoRGB"); close(); print("Job done");

167

7. Publications

7.1. CRL4RBBP7 is required for efficient CENP-A deposition at centromeres Research article: Journal of Cell Science, 2015

Authors Julien Mouysset*, Samuel Gilberto*, Michelle G Meier, Fabienne Lampert, Mukta Belwal, Patrick Meraldi, Matthias Peter *Authors contributed equally

See chapter 2 for the manuscript and contributions

7.2. Dynamic ubiquitin signaling in cell cycle regulation Review article: The Journal of Cell Biology, 2017

Authors Samuel Gilberto and Matthias Peter

Contribution Writing the manuscript and assembling the figures

Full citation Gilberto, S., and M. Peter. 2017. Dynamic ubiquitin signaling in cell cycle regulation. J. Cell Biol. doi:10.1083/jcb.201703170

168

JCB: Review

Dynamic ubiquitin signaling in cell cycle regulation

Samuel Gilberto1,2 and Matthias Peter1

1Department of Biology, Institute of Biochemistry, Swiss Federal Institute of Technology, Zurich, Switzerland 2Molecular Life Science PhD Program, Life Science Zurich Graduate School, Zurich, Switzerland

The cell division cycle is driven by a collection of enzymes (Teixeira and Reed, 2013; Bassermann et al., 2014). However, that coordinate DNA duplication and separation, ensur- ubiquitylation is not necessarily linked to protein degradation, and in recent years, an increasing number of nonproteo- ing that genomic information is faithfully and perpetually lytic outcomes of protein ubiquitylation have been reported maintained. The activity of the effector proteins that per- to play important cellular roles (Komander and Rape, 2012). form and coordinate these biological processes oscillates Proteasome-independent regulation of an ubiquitylation target by regulated expression and/or posttranslational modifi- is achieved by changes in protein–protein interactions, subcellular localization, or enzyme activity (Fig. 1 B). As opposed to cations. Ubiquitylation is a cardinal cellular modification the irreversible fate of degradation, nonproteolytic outcomes of and is long known for driving cell cycle transitions. In this ubiquitylation allow for functional fine-tuning, dynamically and

review, we emphasize emerging concepts of how ubiqui- reversibly responding to intracellular cues instead of requiring Downloaded from de novo protein synthesis. tylation brings the necessary dynamicity and plasticity Ubiquitin conjugation to its targets requires the concerted that underlie the processes of DNA replication and mito- action of an E1 ubiquitin-activating enzyme, E2 ubiquitin- sis. New studies, often focusing on the regulation of chro- conjugating enzyme, and E3 ubiquitin ligase. The latter binds mosomal proteins like DNA polymerases or kinetochore specifically to the substrate and promotes the transfer of ubiquitin to one of its lysine residues (see text box for an overview of

kinases, are demonstrating that ubiquitylation is a versa- E3 ligases involved in cell cycle regulation). Because of multi- jcb.rupress.org tile modification that can be used to fine-tune these cell ple reactive sites on ubiquitin, more moieties may be added, es- cycle events, frequently through processes that do not in- tablishing complex oligomers or chains (Fig. 1 A). This enables volve proteasomal degradation. Understanding how the that multiple ubiquitin topologies generate individual signals, which are collectively referred to as the ubiquitin code (Koman- increasing variety of identified ubiquitin signals are trans-

der and Rape, 2012). This code is read by downstream factors on August 6, 2017 duced will allow us to develop a deeper mechanistic per- containing ubiquitin-binding domains, referred to as readers or ception of how the multiple factors come together to decoders, which specifically recognize the chain topology and induce the appropriate signal (Husnjak and Dikic, 2012). For ex- faithfully propagate genomic information. Here, we dis- ample, a polyubiquitin chain in which ubiquitin conjugates via cuss these and additional conceptual challenges that are its lysine-48 (K48) and/or K11 residues is read and as a result currently under study toward understanding how ubiqui- rapidly degraded by the 26S proteasome, an irreversible process tin governs cell cycle regulation. that is often observed in cell cycle transitions (Grice and Na- than, 2016). Conversely, a monoubiquitin moiety or K63-linked chain can recruit factors that allow for a specific localized re- THE JOURNAL OF CELL BIOLOGY CELL OF JOURNAL THE Introduction sponse, such as the recruitment of a DNA damage–tolerant Cell proliferation is a continuous cycle of DNA synthesis and polymerase to a site of replication stress (García-Rodríguez et subsequent chromosome separation. Posttranslational mod- al., 2016). In many cases, ubiquitylated proteins first need to ifications of effector proteins ensure that these major events be extracted from interacting partners or chromatin, a function and their transitions are orchestrated so that genomic informa- typically attributed to the ATPase valosin-containing protein tion is preserved. The covalent conjugation of the small pro- (VCP)/p97 (Cdc48 in yeast; Meyer et al., 2012; Franz et al., tein ubiquitin through a process called ubiquitylation plays a 2016). Importantly, specific proteases termed deubiquitylating critical role in the overall regulation of cell division. It is well enzymes (DUBs) can cleave off ubiquitin moieties and reverse established that ubiquitylation is a signal for protein degrada- the signal (Lim et al., 2016). tion by the proteasome (Fig. 1, A and B), with special impor- In this review, we summarize the main ubiquitin- tance in assuring ordered and well-timed cell cycle transitions mediated regulatory mechanisms that are believed to fine-tune DNA replication and segregation. We emphasize how E3 ubiq- Correspondence to Matthias Peter: [email protected] uitin ligases orchestrate these processes in space and time, with Abbreviations used: APC/C, anaphase-promoting complex/cyclosome; CRL, cullin-RING E3 ligase; DUB, deubiquitylating enzyme; ESCRT, endosomal © 2017 Gilberto and Peter This article is distributed under the terms of an Attribution– sorting complexes required for transport; FACT, facilitates chromatin tran- Noncommercial–Share Alike–No Mirror Sites license for the first six months after the scription; MCC, mitotic checkpoint complex; PCNA, proliferating cell nuclear publication date (see http://www.rupress.org/terms/). After six months it is available under antigen; SAC, spindle assembly checkpoint; SLBP, stem-loop binding protein; a Creative Commons License (Attribution–Noncommercial–Share Alike 4.0 International UBD, ubiquitin-binding domain; VCP, valosin-containing protein. license, as described at https://creativecommons.org/licenses/by-nc-sa/4.0/).

The Rockefeller University Press J. Cell Biol. https://doi.org/10.1083/jcb.201703170 JCB 1 Downloaded from jcb.rupress.org on August 6, 2017

Figure 1. Writing, reading, and editing ubiquitin. (A) Ubiquitin is added to a substrate protein (writing ubiquitin) by E3 ligases, and ubiquitin moieties can be removed (editing ubiquitin) by deubiquitylating enzymes (DUBs). Protein degradation is mostly associated with polyubiquitin chains, in which ubiquitin moieties attach to each other via homotypic lysine-48 (K48) linkages or heterotypic K11/K48 linkages (mixed chain or branched; see Ubiquitin signals produced by CRLs and the APC/C text box). The result of K63-linked polyubiquitylation is distinct and, together with monoubiquitylation (monoUb), it is associated with nonproteolytic outcomes. (B) Ubiquitylation produces a signal that is often dependent on effector proteins or complexes (ubiquitin readers). These include the proteasome, which is a proteolytic machine, or the segregase VCP/p97 (Cdc48 in yeast), which extracts proteins from chromatin, cellular compartments, or protein complexes for recycling or degradation. Other ubiquitin-binding proteins can fulfill a specific function with nonproteolytic outcomes when they are recruited to ubiquitylated substrates (e.g., damage tolerance by error-prone polymerases), potentially altering the localization or activity of the ubiquitylated substrate. By affecting protein interactions or conformations, ubiquitylation may directly alter protein localization or activity. A challenge in present research is to distinguish between a passive effect of ubiquitylation and the action of an unidentified ubiquitin reader.

a special focus highlighting nonproteolytic consequences of For example, to prevent rereplication by prematurely assem- ubiquitylation. We aim to pinpoint current research challenges bling origins on newly replicated DNA, replication licensing and suggest novel research approaches to decipher the complex factors are degraded in S phase, and cells degrade DNA rep- ubiquitin-dependent network orchestrating cell cycle regulation. lication factors such as the nuclease FEN1 after replication is complete (Guo et al., 2012; Moreno and Gambus, 2015). Al- Dynamic control of DNA replication together, the prevailing paradigm suggests that degradation of by ubiquitin replication effectors is required to restrict their function to a A cell duplicates its genomic information during S phase. Syn- narrow temporal window. thesis of the complementary DNA strands begins at localized Regulation of lagging-strand synthesis. In re- replication origins, which are established during mitosis and G1 cent years, localized proteolytic and several nonproteolytic during replication licensing (Fragkos et al., 2015). After DNA ubiquitin-mediated regulatory processes have been discovered duplex unwinding by the replicative helicase, the polymerases to regulate replication (Table 1 summarizes nonproteolytic cell (Pol) Polε and Polδ elongate the “leading” and “lagging” DNA cycle ubiquitylation events). An example of replication fine- strands (Fig. 2 A). Once the duplication of a DNA stretch is tuning through selective and localized degradation arises during complete, replication is terminated and components are removed the process of lagging-strand synthesis (Fig. 2 B). The discon- from chromatin. Ubiquitylation impacts all stages of DNA rep- tinuous synthesis of DNA requires a constant exchange of fac- lication (Moreno and Gambus, 2015; García-Rodríguez et al., tors to prime, elongate, process, and ligate the so-called Okazaki 2016). Past research has focused on the global degradation of fragments. Priming is performed by Polα, which synthesizes a replication effectors when their function is no longer needed. RNA primer that is removed during the maturation step. Polδ

2 JCB • 2017 Cullin-RING and APC/C E3 ligases functions during lagging-strand synthesis for consecutive extension of the primer and also for gap-filling during nick translation, a far less processive event (Zheng and Shen, 2011). It appears that in humans the composition of the four-subunit Polδ enzyme (Polδ4) is altered in order to promote this activity shift. Recent evidence argues that the cullin RING ligase (CRL) CRL4CDT2 mediates the destruction of the regulatory p12 subunit of Polδ4 during S phase (Cullin-RING and APC/C E3 ligases text box; Zhang et al., 2013), resulting in the formation of Polδ3, which has specialized properties such as increased proofreading activity. Polδ3 was also associated with gap-filling during DNA repair (Lee et al., 2012). Hence, one model is that the conversion from Polδ4 to Polδ3 generates a polymerase that is more suitable for gap-filling during Okazaki fragment processing (Fig. 2 B), explaining how the processivity of Polδ is locally adjusted (Lin et al., 2013; Lee et al., 2014), with local Polδ4 clearance important for the proper execution of DNA replication. Moreover, there is also a role for nonproteolytic ubiquitylation in lagging-strand synthesis through modulation of protein–protein interactions. MCM10 is a replication fork

scaffolding protein involved in the recruitment of the replicative Downloaded from polymerases. Early evidence in yeast suggested that dimonoubiquitylation of MCM10 changes its interactions. Al- though the affinity of MCM10 for the primase Polα decreases, dimonoubiquitylation likely facilitates the recruitment of the elongating Polε/δ because of the concomitant increased Of the three described E3 ligase families, HECT, RING, and MCM10 affinity to proliferating cell nuclear antigen (PCNA), RING-between-RING E3 ligases (Spratt et al., 2014), the bulk of jcb.rupress.org cell cycle regulation is performed by RING E3 ligases. In particular, the sliding clamp that brings these polymerases to DNA the major family Cullin-RING E3 ligases (CRLs) and the anaphase- (Das-Bradoo et al., 2006; Thu and Bielinsky, 2014). Whether promoting complex/cyclosome (APC/C) take up most known cell cycle analogous mechanisms also regulate this switch in higher ubiquitylation events. CRLs use one of the six cullin proteins encoded eukaryotes remains to be established. by the human genome as a scaffolding subunit that brings together Control of chromatin assembly during DNA the ubiquitin-loaded E2 enzyme and the substrate. The E2 enzyme is recruited by the C-terminally bound RING subunit (RBX1 or RBX2). Sub- replication. Recent work also uncovered a crucial nonproteo- on August 6, 2017 strates associate to CRLs via an N-terminal receptor module composed lytic role for ubiquitin signaling in regulating the dynamic nu- of a variable substrate-specific adaptor and a cullin-bound linker sub- cleosomal chromatin structure at advancing replication forks unit, except in the case of CRL3. CRLs are activated by modification (Fig. 2, C and D). Nucleosome histones must be evicted from with NEDD8, termed neddylation, and they associate dynamically with regulators that modulate the neddylation state, block substrate access, DNA and deposited in a semiconservative manner onto new or promote substrate receptor release and exchange (Lydeard et al., DNA strands and the remaining gaps filled with newly synthe- 2013). CRLs are thus regarded as modular, dynamic assemblies with sized histones. Thus, nucleosome assembly during S phase ne- substrate-specific adaptors that associate and dissociate in a regulated cessitates an adequate histone supply (Alabert and Groth, manner to ensure timely and specific substrate ubiquitylation (Craney 2012), regulated through transcriptional induction and histone and Rape, 2013). Specific adaptors are linked to individual cullins. CRL1 or SCF (SKP1–CUL1–F-box) E3 ligases contain an F-box protein; mRNA maturation by the processing factor stem-loop binding CRL3 contains a Broad complex, Tramtrack, and Bric-a-brac (BTB) do- protein (SLBP; Fig. 2 D). Interestingly, histone mRNA process- main–containing protein; and CRL4 has a DDB1- and CUL4-associated ing is activated by human CRL4WDR23 through multimonoubiq- factor (DCAF) protein (Lydeard et al., 2013). Subdivided into CRL2 uitylation of SLBP (Brodersen et al., 2016). Indeed, cells and CRL5, the Elongin B-C–CUL2/CUL5–SOCS box protein (ECS) E3 ligases recruit BC-box–containing adaptors, in particular VHL-box and lacking WDR23 or SLBP exhibit severe DNA replication de- SOCS-box proteins (Cai and Yang, 2016). Multiple cell cycle transitions fects caused by slow replication forks, suggesting that incorpo- critically depend on SCF E3 ligases, in particular for targeting degra- ration of newly synthesized histones is tightly coupled to fork dation of cyclin-dependent kinase (CDK) inhibitors such as p27 and progression. How ubiquitylation mechanistically impacts SLBP WEE1 at G1/S and G2/M, respectively. CRL4 complexes have been function remains to be investigated, but it is conceivable that described for their functions in preventing DNA re-replication, whereas CRL3 is probably the most emergent CRL in cell cycle control, in partic- ubiquitylation regulates its binding to interacting partners or di- ular by regulating mitosis. Several nonproteolytic functions of CRL3 and rectly affects enzymatic activity (Lampert et al., 2017). After S CRL4 complexes are now attributed (Table 1; Teixeira and Reed, 2013; phase, SLBP is rapidly degraded by SCFcyclin F complexes Bassermann et al., 2014). DNA replication is one of the few cell cycle (Dankert et al., 2016), and this proteolytic destruction is critical functions currently attributed to ECS E3 ligases (Table 1 and main text). Although the APC/C is closely related to CRLs and contains the for genome maintenance upon genotoxic stress. Thus, nonpro- cullin-homology subunit APC2 (Yu et al., 1998), it is structurally diver- teolytic and proteolytic regulation of SLBP by ubiquitin coop- gent. The APC/C is composed of at least 14 different subunits, includ- erate in space and time to restrict histone synthesis to S phase ing the RING subunit APC11, plus one of two coactivators (CDC20 and and thereby maintain genome stability. CDH1) that also participate in substrate binding (Sivakumar and Gorbsky, Both histone eviction and deposition require so-called 2015). The APC/C operates in mitosis and G1 and is mostly known for its ability to degrade mitotic cyclins and other mitotic factors so that chromo- histone chaperones. Available data suggest that nonproteolytic somes are separated and mitotic exit ensues (Zhou et al., 2016). ubiquitin signaling mediated by cullin-4 and its putative yeast homologue, Rtt101 (Zaidi et al., 2008), coordinate histone-

Dynamic ubiquitylation in the cell cycle • Gilberto and Peter 3 Downloaded from jcb.rupress.org on August 6, 2017

Figure 2. The dynamic regulation of unperturbed DNA replication by ubiquitin. Proteolytic and nonproteolytic mechanisms are depicted with light orange and light blue background, respectively, and gray if a determination is incomplete. (A) Overview of the primary events occurring during DNA replication. Activation of the active CMG helicase (CDC45, MCM hexamer, GINS complex) induces the recruitment of the sliding clamp PCNA (depicted in red), which serves as an interaction platform to tether DNA polymerases to chromatin (Moldovan et al., 2007). (B) Polymerase switches occurring in lagging strand synthesis are mediated by ubiquitylation. (C and D) Concomitant with DNA replication, nucleosomes are disassembled and reassembled in a semiconservative manner, incorporating newly synthesized histones, which requires nondegradative ubiquitylation. (E) Termination of chromosomal replication in yeast and Xenopus requires Cdc48/p97 for CMG eviction from the chromatin. Red crosses depict targets of proteasomal degradation, and red circles depict ubiquitin. Ac, acetylation; Sc, Saccharomyces cerevisiae; Xl, Xenopus laevis.

related processes by acting either on histone chaperones or on histone H3–H4 dimers by ubiquitylating new, acetylated his- histones themselves. Rtt101 is required to target the histone tone H3. The consequence is a switch in interactions between chaperone facilitates chromatin transcription (FACT) complex H3–H4 and the respective histone chaperones that allows their to the replication fork through nonproteolytic polyubiquityla- loading onto nucleosomes (Fig. 2 D; Han et al., 2013). A recent tion of the FACT Spt16 subunit (Fig. 2 C; Han et al., 2010). The study clarified that Rtt101 is indeed tethered to replisomes to same E3 ligase promotes the deposition of newly synthesized locally restrict its function to the vicinity of the replication fork

4 JCB • 2017 (Buser et al., 2016). In humans, CRL4CDT2 is also recruited to ubiquitylation of factors in the normal progression of replica- active forks (Havens and Walter, 2009; Havens et al., 2012) and tion forks. In the case of SCFDia2, the best described function is may thus perform an equivalent function. to promote the termination of DNA replication (Fig. 2 E). Unloading of the replicative helicase. Rtt101 is Hence, although Rtt101 is necessary during fork progression, not the only resident E3 ligase functioning at yeast replication SCFDia2 rather operates when chromosomal replication is fin- forks. The replisome also binds the SCFDia2 E3 ligase (Moro- ished. Because binding of SCFDia2 to the fork is important, it hashi et al., 2009), further underscoring the importance of local appears that SCFDia2 in some way senses when replisome

Table 1. Nonproteolytic ubiquitylation: Selected substrates of E3 ubiquitin ligases that operate in an unperturbed cell cycle

Phase and E3 ligase Chain topology Evidence Role of ubiquitylation Counteracting DUB Reference substrate (or deubiquitylation, if indicated)

G1 PALB2 CRL3KEAP1 MultimonoUb? vv (uPD), int, Prevents BRCA1-PALB2-BRCA2 complex USP11 Orthwein et al., 2015 vt, m assembly, inhibiting homology-directed DNA repair S Histone H2A RING1A,B MonoUb vv (ChIP, IF) Pericentromeric DNA replication Multiple; not tested Bravo et al., 2015; for this function Lim et al., 2016 Histone H2B BRE1 MonoUb vv (ChIP), m Promotes nucleosome reassembly — Trujillo and Osley, 2012 and/or stability Downloaded from Histone H3 Rtt101Mms22 (Sc); MultimonoUb vv (sIP), vv Promote H3 deposition in newly — Han et al., 2013 CRL4? (Wb, ChIP), synthesized DNA int, vt, m MCM3 CRL3KEAP1 MultimonoUb? vv (sIP), int Undetermined — Mulvaney et al., 2016 MCM7 SCFDia2 (Sc), K48-linked vv (sIP, uPD), vt Replication termination: Disassembly of — Maric et al., 2014; CRL2LRR1 (Xl) (degradation the replicative CMG helicase Moreno et al., 2014; uncertain) Dewar et al., 2017

MCM10 ? (Sc) DimonoUb vv (sIP) Promote PCNA recruitment for elongation — Das-Bradoo et al., 2006 jcb.rupress.org during DNA replication SLBP CRL4WDR23 MultimonoUb vv (K-GG), vt, Promote histone mRNA expression — Brodersen et al., 2016 m, int Spt16 Rtt101 (Sc) K63-linked chain vv (sIP, uPD), Stabilizes FACT complex at replication — Han et al., 2010 int, vt origins to promote MCM binding

S and G2 on August 6, 2017 Aurora A CRL3KLHL18 ? vv (sIP), int, vt Activation of centrosomal Aurora A to — Moghe et al., 2012 promote mitotic entry TOP2A BRCA1 K63-linked chain? vv (sIP), int Increase decatenation activity of — Lou et al., 2005 topoisomerase IIα TOP2A RNF168 K63-linked chain vv (sIP), int, Promote DNA decatenation by increasing USP10 Guturi et al., 2016 vt, m topoisomerase IIα chromatin association Mitosis Aurora B CRL3KLHL21 MonoUb? int, vt, m Promote UBASH3B-dependent Aurora B — Maerki et al., 2009; translocation to the spindle midzone Krupina et al., 2016 in anaphase Cyclin B1 ? K63-linked chain vv (sIP), int Stabilize cyclin B1 — Zhang et al., 2015 Dishevelled ? K63-linked chain DUBa: vv (sIP, DUB: Promotes spindle orientation, by CYLD Yang et al., 2014 DVL3 uIP), m promoting correct localization of NuMA/dynein at the cell cortex NuMA BRCA1? K63-linked chain DUBa: vv (sIP) DUB: Promotes spindle assembly by BRISC complex Yan et al., 2015 stimulating the incorporation of NuMA into spindle poles PLK1 CRL3KLHL22 MonoUb? int, vt, m Remove PLK1 from the kinetochore upon USP16 Beck et al., 2013; chromosome bi-orientation Zhuo et al., 2015 Survivin ? K63-linked chain DUBa: vv (sIP) DUB: Dissociates Survivin and the CPC USP9X Vong et al., 2005 from centromeres Late M/early G1 CENP-A (Dm) CRL3RDX ? vv (uIP), vt Stabilize CENP-A to promote its — Bade et al., 2014 incorporation into centromeres CENP-A CRL4COPS8, MonoUb vv (sIP), vt, m Promote interaction with the HJURP — Mouysset et al., 2015; CRL4RBBP7? histone chaperone and CENP-A Niikura et al., 2015 loading at centromeres

Shown substrates are not thought to be targeted for proteasomal degradation. Depicted E3 ligase/substrate pairs refer to human proteins, unless indicated. If known, the type of ubiquitylation topology is indicated. A question mark denotes unknown information or a speculative hypothesis. ChIP, chromatin immunoprecipitation; Dm, Drosophila melanogaster; IF, immunofluorescence; int, E3 ligase interaction with substrate; K-GG, ubiquitin profiling; m, mutagenesis of ubiquitylated sites (lysine to arginine); monoUb, monoubiquitylation; Sc, Saccharomyces cerevisiae; sIP, substrate immunoprecipitation and ubiquitin detection; uPD or uIP, ubiquitin pull-down or immunoprecipitation and substrate detection; vt, in vitro ubiquitylation assays; vv, in vivo (method indicated between parentheses); Wb, Western blot; Xl, Xenopus laevis; —, not described. aAvailable evidence designates the function of the DUB, not an E3 ligase.

Dynamic ubiquitylation in the cell cycle • Gilberto and Peter 5 Downloaded from jcb.rupress.org on August 6, 2017

Figure 3. Ubiquitin in the regulation of protein dynamics and localization in mitosis. Proteolytic and nonproteolytic mechanisms are depicted with light orange and light blue background, respectively. (A) Upon mitotic entry, chromosomes condense and the cell assembles a bipolar mitotic spindle. The kinases Aurora B (AurB) and PLK1 both contribute to the establishment of correct, bioriented, kinetochore–microtubule attachments by destabilizing incorrect attachments and stabilizing correct ones, respectively. (B) Transit to anaphase occurs when the APC/C is no longer inhibited by the MCC and is activated by CDC20. MCC disassembly is promoted by autoubiquitylation of CDC20 within the APC/C-bound MCC and by the ubiquitin reader CUEDC2. The irreversibility of this transition necessitates cyclin B destruction, as otherwise the SAC is reactivated (Clijsters et al., 2014; Rattani et al., 2014; Vázquez- Novelle et al., 2014). (C) Kinetochore recruitment and exclusion of the chromosomal passenger complex (CPC), which includes Aurora B and Survivin, depend on nonproteolytic ubiquitylation. Exclusion of PLK1 from the kinetochore in case of bioriented microtubule attachments also depends on its ubiquitylation. (D and E) Microtubule transport can be promoted by cargo ubiquitylation, as is the case for the spindle assembly factor NuMA and Aurora B. Whether the ubiquitylation of PLK1 promotes its transport to the spindle midzone has not yet been determined. Red crosses depict targets of proteasomal degradation, red circles depict ubiquitin, and purple circles depict Aurora B kinase.

function is complete, after which it ubiquitylates the Mcm7 sub- from DNA by Cdc48/p97 and hence the disassembly of the en- unit of the replicative helicase (Maculins et al., 2015). Mcm7 tire replisome, thereby terminating replication (Maric et al., ubiquitylation promotes the extraction of the replicative helicase 2014; Moreno et al., 2014). A similar mechanism exists in

6 JCB • 2017 Ubiquitin signals produced by CRLs and the APC/C to identify, and the process of local cell cycle effector regulation may be more common than current evidence suggests. Like- Unlike the other known classes of E3 ligases, RING E3 ligases work by fa- wise, nonproteolytic ubiquitylation is expected to rely on DUBs cilitating the direct transfer of ubiquitin from the E2 to the substrate lysine residue. A different E2 enzyme may be used to initiate and elongate a that remove the modification. However, only a few DUBs have polyubiquitin chain (Deshaies and Joazeiro, 2009; Ye and Rape, 2009). been identified to date that regulate DNA replication, all but Alternatively, as for a subset of CRLs, an independent E3 ligase may be ensuring that a new chapter of discovery awaits. For example, is recruited to catalyze the initiation step (Scott et al., 2016). The E2 enzyme a DUB also tethered to the replication fork to reverse MCM10 used for chain elongation is the major determinant of ubiquitin chain topology (Deshaies and Joazeiro, 2009; Ye and Rape, 2009). In the case dimonoubiquitylation? At which point is newly synthesized his- of CRLs, UBCH5 E2 enzymes allow for mono or multimonoubiquitylation, tone H3 deubiquitylated? Furthermore, we do not understand whereas CDC34 drives chain extension, forming canonical K48-linked how regulatory ubiquitylation signals are translated into their polyubiquitin chains (Lydeard et al., 2013; Grice and Nathan, 2016). response and which ubiquitin readers are involved. Finally, Surprisingly, the metazoan APC/C appears to be the major cel- there are certainly more ubiquitylation substrates and perhaps lular source of atypical K11-linked polyubiquitin chains, which is part of a signal for proteasomal degradation. The APC/C makes use of the more E3 ligases with functions in DNA replication awaiting E2 enzymes UBE2C and UBE2S to initiate and elongate these atypical discovery. For example, in Caenorhabditis elegans, the chains, respectively (Jin et al., 2008; Garnett et al., 2009; Williamson et CRL2LRR-1 complex regulates DNA replication, and in Xenopus, al., 2009; Matsumoto et al., 2010; Min et al., 2015; Brown et al., 2016). the homologous E3 ligase is involved in replication termination Despite considerable effort, a consensual structure for K11-linked chains is lacking (Bremm et al., 2010; Matsumoto et al., 2010; Castañeda et al., (Ossareh-Nazari et al., 2016; Dewar et al., 2017). The direct 2013). Nevertheless, recent studies clarified that homotypic K11 chains substrates mediating DNA replication regulation are not known, are not sufficient to signal proteasome-mediated degradation. Rather, and whether either function of CRL2LRR1 is conserved in hu- heterotypic K11/K48–polyubiquitinated proteins are efficient proteolytic mans remains to be tested. Finally, although the replicative he- signals (Grice et al., 2015). Moreover, it was also observed that several

licase subunit MCM3 is ubiquitylated, the biological Downloaded from ubiquitin chains can be extended from preformed ubiquitin oligomers, constituting branched K48/K11–polyubiquitin chains that appear to significance of this regulation is elusive despite considerable be better signals for proteasomal recognition (Meyer and Rape, 2014). study (Mulvaney et al., 2016). Collectively, many questions re- These branched chains were suggested to facilitate the degradation main to be answered, especially in identifying the players that of prometaphase APC/C substrates, a mitotic stage characterized by erase and read critical ubiquitin signals during S phase. low APC/C activity (Meyer and Rape, 2014). The ability of the human APC/C to synthesize heterotypic ubiquitin chains does not appear to be conserved across all eukaryotes, as at least yeast APC/C substrates are Ubiquitin regulation of DNA segregation modified with canonical K48-linked polyubiquitin (Rodrigo-Brenni and Sister chromatids are segregated during mitosis in a process jcb.rupress.org Morgan, 2007). Yeast might instead make use of complementary mech- that involves chromosome condensation, nuclear envelope anisms that reassure the similarly ordered degradation pattern (Lu et al., breakdown in animal cells, and centrosome separation to op- 2014, 2015a). Despite the importance of K11 chains as a degradation signal, the posite poles. The activity of cyclin-dependent kinase 1, with its respective E2 UBE2S is not essential for cyclin B1 degradation (a canoni- positive regulator cyclin B (CDK1/cyclin B), is the main trig- cal APC/C substrate; Garnett et al., 2009; Dimova et al., 2012), leading

ger of these events (Gavet and Pines, 2010). In addition, the on August 6, 2017 to the conclusion that multimonoubiquitylation can also constitute a signal centromere of condensed chromosomes plays an important role for proteasomal degradation (Dimova et al., 2012). Indeed, single-molecule kinetic studies support the view that multimonoubiquitylation can in the assembly of kinetochores that mediate chromosome– efficiently induce substrate binding to the proteasome (Lu et al., 2015b). spindle attachments and allow chromosome congression at the Hence, it appears that higher local concentration of ubiquitin moieties metaphase plate (Fig. 3 A). Finally, the spindle assembly check- enhances binding to proteasomal ubiquitin readers, even though binding point (SAC) monitors microtubule–kinetochore attachments to to the proteasome does not necessarily correlate with an increased rate ensure faithful separation of sister chromatids. of degradation (Lu et al., 2015b; Yau and Rape, 2016). Future research will likely reveal the determinants of the commitment of a substrate to Regulating APC/C E3 ligase activity. Well-timed degradation once it is bound to the proteasome. protein degradation is a common event in the cell cycle, known to drive mitotic entry (G2/M) as well as the metaphase-to- anaphase transition (Teixeira and Reed, 2013; Bassermann et Xenopus laevis, and the E3 ligase was recently identified to be al., 2014). A frequent general question in these and other cell CRL2LRR1 (Moreno et al., 2014; Dewar et al., 2017). Of note, cycle processes is what defines the functional time window of CRL2LRR1 seems to be specifically recruited to the chromatin at an E3 ligase. In principle, either the activity of the E3 ligase the time of termination instead of being tethered to the repli- may itself be regulated, or the substrate binding to the E3 ligase some like SCFDia2 (Dewar et al., 2017). It is currently unclear may depend on third-party factors such as kinases or scaffold- whether the K48-polyubiquitylated Mcm7 subunit is targeted to ing proteins. Mitosis provides a remarkable example of how an the proteasome or recycled. E3 ligase can be dynamically regulated, in this case to tightly Ubiquitin in DNA replication: Open questions. coordinate the status of kinetochore–microtubule attachments Altogether, ubiquitin can be used to signal specific and conse- with the onset of chromosome separation. It is long known that quential modulation of the DNA replication machinery, espe- the metaphase-to-anaphase transition is driven by the E3 ligase cially for lagging-strand synthesis factor switching and anaphase-promoting complex/cyclosome (APC/C; see Cullin- nucleosome reassembly. Both proteasomal and nonproteolytic RING and APC/C E3 ligases text box), activated by its subunit pathways can contribute to this behavior. Importantly, the fine- CDC20 (Teixeira and Reed, 2013; Bassermann et al., 2014). tuned response requires reversible effects, because a modified High APC/CCDC20 activity triggers anaphase and mitotic exit by protein must be rapidly unmodified or replaced to initiate a new mediating the degradation of cyclin B and securin, an inhibitor synthesis cycle. After local factor degradation, a sufficiently of the protease separase that cleaves the cohesin complex hold- large protein pool must be available to allow dynamic regula- ing sister chromatids together (Hirano, 2015). Before anaphase, tion, as in the case of Polδ4 (Lee et al., 2014). Local degrada- APC/CCDC20 is kept inhibited by the SAC until appropriate tion and replenishment of factors is experimentally challenging kinetochore–microtubule attachments are established for all

Dynamic ubiquitylation in the cell cycle • Gilberto and Peter 7 chromosomes. A critical product of the SAC is the mitotic from the APC/C. As a result, CDC20 would be available to the checkpoint complex (MCC), which inhibits APC/CCDC20 activ- proteasome, with subsequent MCC disassembly. ity to prevent premature separation of sister chromatids Ordered degradation of the targets of a single (Lischetti and Nilsson, 2015). E3 ligase. Another concept currently in focus is the pattern of Further studies provided deeper mechanistic insight into ordered degradation of substrates of a single E3 ligase. Such the dynamic regulation of the APC/CCDC20 E3 ligase (Fig. 3 B). pattern was observed for S phase targets of CRL4CDT2 and is Surprisingly, the APC/CCDC20 can itself promote the release established by distinct substrate binding affinities to the E3 li- of its inhibitor MCC through autoubiquitylation of CDC20, a gase (Coleman et al., 2015). APC/CCDC20 likewise represents a process antagonized by the DUB USP44 (Reddy et al., 2007; prime example of coordinated sequential degradation of E3 li- Stegmeier et al., 2007). More recently, it was clarified that gase substrates, though it does not make use of identical mech- CDC20 ubiquitylation is brought about by a peculiar structural anisms. Early observations debated that despite the fact that the rearrangement, triggering CDC20 destruction and MCC disas- MCC precludes the degradation of its late metaphase substrates, sembly (Mansfeld et al., 2011; Varetti et al., 2011; Foster and MCC-bound APC/CCDC20 can ubiquitylate other targets in pro- Morgan, 2012; Yamaguchi et al., 2016). Rather than occurring metaphase, namely cyclin A and the kinase NEK2A (Fig. 3 B; only at the point of anaphase onset, a model has been pro- den Elzen and Pines, 2001; Geley et al., 2001; Hames et al., posed in which constant MCC disassembly during metaphase 2001). Thus, the very same E3 ligase mediates the destruction generates a pool of uninhibited APC/C that can either rebind of several substrates at different time points. The mechanistic the MCC when unattached kinetochores are present or bind basis for selective substrate targeting includes increased affinity free CDC20 and thus be activated, triggering anaphase onset of the early substrates for APC/C binding, and APC/CCDC20 can (Fig. 3 B). This dynamic view of APC/C release from inhibi- generate branched ubiquitin chains that are better signals for

tion is complemented by other specific mechanisms of MCC proteasomal degradation ( Meyer and Rape, 2014; Boekhout Downloaded from extraction (Westhorpe et al., 2011; Miniowitz-Shemtov et al., and Wolthuis, 2015; Di Fiore et al., 2015; Lu et al., 2015a). A 2015; Kaisari et al., 2017). Interestingly, MCC disassembly is summary of the current information on proteolytic ubiquitin enhanced by the ubiquitin reader CUEDC2 (Fig. 3 B; Gao et al., signals generated by the APC/C and CRLs can be found in 2011). Although experimental evidence demonstrated that the the respective text box. ubiquitin-binding domain (UBD) of CUEDC2 is important for Fine-tuning kinetochore protein localization. its function, the ubiquitylated factor to which CUEDC2 binds Other E3 ligases operate in mitosis, providing critical regulation

remains to be determined. The UBD is not required for consti- often through nonproteolytic ubiquitylation. These signals jcb.rupress.org tutive binding to CDC20, but we speculate that it might be the during mitosis contribute to the remarkable resilience of the key in detecting CDC20 ubiquitylation to trigger MCC release system so that cells readily adapt to changing conditions such as

Table 2. Proposed human ubiquitin readers with cell cycle functions on August 6, 2017 UBD Ubiquitin binding mode Number of UBD- Examples with cell Ubiquitin-binding role in cell cycle Reference containing proteins (cell cycle functions functiona cycle associated/total)

UBA MonoUb, polyUb 14/55 UBQLN2,b KPC2,b,c KPC2 (E3 ligase subunit): Promotes Hara et al., 2005; (predominant for K48) FAF1,d UBASH3B,c the transfer of ubiquitylated p27 Krupina et al., 2016 BRSK1/2, LATS1/2, to the proteasome; UBASH3B: MARK4 Targets ubiquitylated Aurora B to microtubules in mitosis CUE MonoUb 1/13 CUEDC2c CUEDC2: Promotes MCC release Gao et al., 2011 from APC/CCDC20 UIM MonoUb, polyUb (K48, 6/28 DDI1,b RPN10,b alpha4: Prevents polyubiquitylation Kong et al., 2009; K63) Epsin-1, MAT1, of the PP2A catalytic subunit McConnell et al., 2010 alpha4c UBP PolyUb 1/3 USP39 None described — NZF MonoUb, polyUb (K63) 2/7 NPL4,d HOIL-1 None described — UEV MonoUb 1/2 TSG101 None described — UBAN DimonoUb (M1) 2/7 ALIX, Optineurin None described — WD40 (subset) MonoUb 1/4 BUB3 None described — Unique or Various 10/21 RPN13,b VCP,d UFD1,d SMURF2 (E3 ligase): Stabilize Ogunjimi et al., 2010 uncharacterized NUP62, ERα, ubiquitylated substrate binding domains SMURF2c to promote polyUbe Domains not found in Various 0/22 — — — cell cycle regulators Total — 38/162 — — —

We considered UBD-containing proteins annotated in UniProtKB/Swiss-Prot, a curated protein database (UniProt Consortium, 2015), together with a manual literature search. Our criteria include (1) proteins containing UBDs described previously (Husnjak and Dikic, 2012), (2) other proteins annotated as binding ubiquitin, and (3) manual exclusion in case of covalent ubiquitin binding (e.g., E2s or ubiquitin modification) and of active DUBs. Assignment of cell cycle–related functions was determined by gene ontology and complemented with manual literature search. Ubiquitin-binding mode is according to Husnjak and Dikic (2012). MonoUb, monoubiquitylation; polyUb, polyubiquitylation; —, not applicable. aExcluding proteins involved in general proteasome function or that of VCP/p97. bProteasome subunit, or proteasome associated (mostly according to Grice and Nathan, 2016). cProteins have a reported cell cycle function for their UBDs that is distinct from general proteasome or VCP/p97 functions. dVCP/p97 component or cofactor (Meyer et al., 2012). eProbably a general mechanism, which includes its cell cycle functions.

8 JCB • 2017 the status and quality of kinetochore–microtubule attachments. indicating that NuMA ubiquitylation likely promotes its transport The importance of these ubiquitylation signals is twofold. by dynein (Fig. 3 D; Yan et al., 2015). Cytoplasmic dynein was Ubiquitylation triggers removal of factors from local chromo- previously implicated in the transport of ubiquitylated protein ag- somal pools when their function is no longer required, and it gregates, tethered by the ubiquitin-binding protein HDAC6, and can promote microtubule transport of effectors to their new sites perhaps another reader transports NuMA in a similar fashion of action. During metaphase, correct kinetochore–microtubule (Kawaguchi et al., 2003; Ouyang et al., 2012). attachments must be stabilized, whereas erroneous attachments Ubiquitin in mitosis: Open questions. Overall, in are destabilized in order to prevent chromosome instability. mitosis, ubiquitin operates to ensure genome integrity and well- These processes are coordinated by two kinases, PLK1 and Au- timed DNA segregation by essentially two pathways. First, the rora B (Zitouni et al., 2014; Krenn and Musacchio, 2015). Inter- peculiar regulation of APC/C by autoubiquitylation provides estingly, the mitotic localization of Aurora B is regulated by the necessary flexibility for the cell to quickly recognize chang- nonproteolytic ubiquitylation at multiple points, including for ing conditions in the kinetochore–microtubule attachment state. its microtubule-mediated translocation (Fig. 3, C and E). Au- Second, the plasticity of PLK1, Aurora B, and NuMA ubiqui- rora B works at the kinetochore to destabilize incorrect micro- tylation ensures that the spindle is correctly assembled and that tubule attachments. The VCP–p97 complex ensures exclusive proper kinetochore–microtubule attachments are established. kinetochore localization by removing Aurora B from chromo- Today, cell cycle research faces the challenge of understanding somal arms, possibly after CRL3KLHL9-KLHL13-mediated poly- how the observed dynamicity in ubiquitylation is achieved. The ubiquitylation (Ramadan et al., 2007; Sumara et al., 2007; increased knowledge of APC/CCDC20 regulation might facilitate Dobrynin et al., 2011). In anaphase, Aurora B translocates to understanding of how other E3 ligases are regulated in space the spindle midzone, a process initiated by CRL3KLHL21-dependent and time. For example, it seems that CRL3KLHL22 dynamically

monoubiquitylation of Aurora B at attached kinetochores. Re- responds to microtubule–kinetochore tension to ubiquitylate Downloaded from markably, this ubiquitin signal is decoded by the UBA-containing PLK1, but the underlying mechanism remains elusive (Beck et protein UBASH3B, which recruits ubiquitylated Aurora B to al., 2013). To which extent other cell cycle E3 ligases are regu- microtubules in the vicinity of the attached kinetochore (Maerki lated in a comparable dynamic fashion will likely demand con- et al., 2009; Krupina et al., 2016). The microtubule-dependent siderable research efforts. For instance, the APC/C E3 ligase translocation of Aurora B to the spindle midzone in anaphase is was an early discovery in cell cycle research (Irniger et al., mediated by the kinesin MKLP2 (Gruneberg et al., 2004). In- 1995; King et al., 1995; Sudakin et al., 1995), yet APC/C regu- deed, UBASH3B tethers MKLP2 and ubiquitylated Aurora B, lation is still an area of active investigation. Dynamic ubiquityl- jcb.rupress.org thereby promoting microtubule-dependent Aurora B transloca- ation can also be modulated at the level of the substrate by tion (Fig. 3, C and E). Whether ubiquitylated Aurora B first DUBs, but information regarding how their activity is modu- needs to be extracted by VCP/p97 remains to be investigated. lated is mostly lacking. As another pressing and relatively ob- Although PLK1 stabilizes correct kinetochore–micro- scure topic, further functional analysis will be required to

tubule attachments, its removal from kinetochores is required identify specific readers involved in mitotic processes regulated on August 6, 2017 for faithful metaphase progression (Liu et al., 2012). Bipolar by nonproteolytic ubiquitin signals. Finally, it will be of interest attachment creates tension across the kinetochore, and recent to determine whether ubiquitin-dependent microtubule motor data suggest that this may activate CRL3KLHL22 to trigger rapid binding is an ordinary feature in microtubule cargo transport. removal of PLK1 (Fig. 3 C; Beck et al., 2013). Ubiquitylation is counteracted by the DUB USP16, and thus a balance between Perspective: Reading ubiquitin signals CRL3KLHL22 and USP16 ensures the correct localization and In this review, we summarized examples in which both proteo- function of PLK1 (Zhuo et al., 2015). This balance provides lytic and nonproteolytic ubiquitin signals regulate cell cycle plasticity to this system, as ubiquitylation can be added or re- events. Ubiquitylation of key factors can be reversible, either by moved to fine-tune the localization of a subpopulation of PLK1. a DUB or through the rapid replenishment of a locally degraded Because CRL3KLHL22 regulates PLK1 by nonproteolytic ubiqui- factor, such as p12 or CDC20. Despite a growing catalog of tylation, probably by monoubiquitylation, its displacement from nonproteolytic ubiquitin signals, surprisingly little is known kinetochores likely depends on a dedicated ubiquitin-binding about the mechanisms underlying cell cycle regulation that go protein such as VCP/p97. Because PLK1 is also translocated by beyond proteasome targeting. Although monoubiquitylation the MKLP2 kinesin to the spindle midzone in anaphase (Neef is widespread (Nakagawa and Nakayama, 2015), assessing et al., 2003), it also needs to be clarified whether ubiquitylated nonproteolytic ubiquitin signals and elucidating how ubiqui- PLK1 is similarly recognized and translocated by UBASH3B or tin mechanistically alters the activity of a given target requires whether this process requires a different reader. detailed understanding of the underlying process. Therefore, DUBs reveal additional roles of ubiquitin in reading the information encoded in ubiquitin chains is now a microtubule transport. Kinase translocation in anaphase is major challenge in cell cycle research for nondegradative out- not the only example of how protein ubiquitylation determines comes. The action of CUEDC2 and UBASH3B, in addition to cargo for mitotic microtubule-based transport. Two studies re- VCP/p97, provides the first clues toward a more comprehensive ported that the DUBs CYLD and the BRISC complex are in- understanding. We have summarized information regarding cell volved in the assembly and positioning of the mitotic spindle by cycle proteins with UBDs and discovered that ∼25% of the pu- regulating the function of the spindle assembly factor NuMA tative human ubiquitin readers are also proteins associated with (Yang et al., 2014; Yan et al., 2015). NuMA promotes the tether- cell cycle regulation (Table 2). Nevertheless, in the majority of ing of microtubules to the spindle poles and also to the cell cortex these cases, we do not yet understand the role of the UBD or that and is transported to these sites along microtubules by cytoplas- of the ubiquitylated binding proteins and subsequent response mic dynein (Radulescu and Cleveland, 2010). The BRISC com- in the context of the cell cycle. For example, the yeast MCC plex appears to deliver ubiquitylated NuMA to spindle poles, component BUB3 can bind ubiquitin, but how it contributes

Dynamic ubiquitylation in the cell cycle • Gilberto and Peter 9 to APC/C regulation remains elusive (Pashkova et al., 2010). Submitted: 23 March 2017 Other examples are the endosomal sorting complexes required Revised: 11 May 2017 for transport (ESCR T)–related proteins TSG101 and ALIX, Accepted: 25 May 2017 which regulate cytokinesis (Morita et al., 2007). Although their interaction with ubiquitin needs to be investigated (Bishop et al., 2002; Dowlatshahi et al., 2012), ALIX and other ESCR T proteins recruit ESCR T-III to promote cytokinetic abscission References (Christ et al., 2016). Interestingly, ESCR T-III is directed to the Alabert, C., and A. Groth. 2012. Chromatin replication and epigenome reforming nuclear envelope by a VCP/p97-dependent mecha- maintenance. Nat. Rev. Mol. Cell Biol. 13:153–167. http://dx.doi.org/10 .1038/nrm3288 nism to aid in nuclear envelope reformation after chromosome Bade, D., A.-L.L. Pauleau, A. Wendler, and S. Erhardt. 2014. The E3 ligase segregation (Olmos et al., 2015). Although speculative, it is CUL3/RDX controls centromere maintenance by ubiquitylating and thus possible that binding of ALIX to an ubiquitylated factor stabilizing CENP-A in a CAL1-dependent manner. Dev. Cell. 28:508– may similarly help to recruit ESCR T-III during late mitosis. 519. http://dx.doi.org/10.1016/j.devcel.2014.01.031 Our efforts to compile cell cycle–associated readers (Table 2) Bassermann, F., R. Eichner, and M. Pagano. 2014. The ubiquitin proteasome system - implications for cell cycle control and the targeted treatment of are likely incomplete, and it is therefore clear that much re- cancer. Biochim. Biophys. Acta. 1843:150–162. http://dx.doi.org/10.1016 mains to be discovered before the underlying processes of non- /j.bbamcr.2013.02.028 proteolytic ubiquitylation are well understood. Beck, J., S. Maerki, M. Posch, T. Metzger, A. Persaud, H. Scheel, K. Hofmann, D. Rotin, P. Pedrioli, J.R. Swedlow, et al. 2013. Ubiquitylation-dependent Technically, addressing nondegradative ubiquitylation localization of PLK1 in mitosis. Nat. Cell Biol. 15:430–439. http://dx.doi can be a challenging task. In particular, when the bulk levels .org/10.1038/ncb2695 of a given target protein remain unchanged, it can be diffi- Bishop, N., A. Horman, and P. Woodman. 2002. Mammalian class E vps proteins recognize ubiquitin and act in the removal of endosomal protein-ubiquitin

cult to experimentally distinguish local degradation of a small Downloaded from conjugates. J. Cell Biol. 157:91–101. http://dx.doi.org/10.1083/jcb but specific pool from ubiquitin-dependent changes promot- .200112080 ing protein translocations and/or activity changes. Tagging Boekhout, M., and R. Wolthuis. 2015. Nek2A destruction marks APC/C specific proteins with a photoswitchable fluorescent protein activation at the prophase-to-prometaphase transition by spindle- checkpoint-restricted Cdc20. J. Cell Sci. 128:1639–1653. http://dx.doi (Zhou and Lin, 2013) and/or pulse-chase–type labeling with .org/10.1242/jcs.163279 stable protein markers provide powerful tools to visualize Bravo, M., F. Nicolini, K. Starowicz, S. Barroso, C. Calés, A. Aguilera, and ubiquitin-dependent translocations. The identification of spe- M. Vidal. 2015. Polycomb RING1A- and RING1B-dependent histone

cific ubiquitin readers may require siRNA or CRISPR-based H2A monoubiquitylation at pericentromeric regions promotes S-phase jcb.rupress.org progression. J. Cell Sci. 128:3660–3671. http://dx.doi.org/10.1242/jcs screenings and/or mutagenesis of their UBDs. Because of .173021 the lack of tools for their detection, another technically chal- Bremm, A., S.M.V. Freund, and D. Komander. 2010. Lys11-linked ubiquitin lenging task is addressing the synthesis and functions of het- chains adopt compact conformations and are preferentially hydrolyzed erotypic (including branched) polyubiquitin chains in vivo. by the deubiquitinase Cezanne. Nat. Struct. Mol. Biol. 17:939–947. http://dx.doi.org/10.1038/nsmb.1873

Ubiquitin linkage in polyubiquitin chains is often distin- on August 6, 2017 Brodersen, M.M.L., F. Lampert, C.A. Barnes, M. Soste, W. Piwko, and M. Peter. guished by linkage-specific polyubiquitin antibodies, but they 2016. CRL4(WDR23)-mediated SLBP ubiquitylation ensures histone cannot discern between homotypic and heterotypic chains. To supply during DNA replication. Mol. Cell. 62:627–635. http://dx.doi.org overcome this limitation, bispecific bivalent antibodies that /10.1016/j.molcel.2016.04.017 Brown, N.G., R. VanderLinden, E.R. Watson, F. Weissmann, A. Ordureau, simultaneously and exclusively bind two distinct types of K.-P. Wu, W. Zhang, S. Yu, P.Y. Mercredi, J.S. Harrison, et al. 2016. Dual ubiquitin linkages within the same polyubiquitin chain have RING E3 architectures regulate multiubiquitination and ubiquitin chain been developed (Rape, M., personal communication). Per- elongation by APC/C. Cell. 165:1440–1453. http://dx.doi.org/10.1016/j .cell.2016.05.037 haps research will also lead to the identification of specific Buser, R., V. Kellner, A. Melnik, C. Wilson-Zbinden, R. Schellhaas, L. Kastner, ubiquitin readers for these noncanonical linkages that in ad- W. Piwko, M. Dees, P. Picotti, M. Maric, et al. 2016. The replisome- dition to their functional characterization could be exploited coupled E3 ubiquitin ligase Rtt101Mms22 counteracts Mrc1 function to and employed to discriminate linkage types. We believe that tolerate genotoxic stress. PLoS Genet. 12:e1005843. http://dx.doi.org/10 .1371/journal.pgen.1005843 new tools will be required to decipher the ubiquitin code. Cai, W., and H. Yang. 2016. The structure and regulation of Cullin 2 based E3 Despite the numerous challenges, it is clear that studying the ubiquitin ligases and their biological functions. Cell Div. 11:7. http://dx roles of proteins that noncovalently bind ubiquitin will con- .doi.org/10.1186/s13008-016-0020-7 tinue to shed light into how the complex network of ubiquitin- Castañeda, C.A., T.R. Kashyap, M.A. Nakasone, S. Krueger, and D. Fushman. 2013. Unique structural, dynamical, and functional properties of k11- dependent signals cooperate to perpetually drive cells through linked polyubiquitin chains. Structure. 21:1168–1181. http://dx.doi.org ordered cycles of DNA synthesis and separation. /10.1016/j.str.2013.04.029 Christ, L., E.M. Wenzel, K. Liestøl, C. Raiborg, C. Campsteijn, and H. Stenmark. Acknowledgments 2016. ALIX and ESCR T-I/II function as parallel ESCR T-III recruiters in cytokinetic abscission. J. Cell Biol. 212:499–513. http://dx.doi.org/10 .1083/jcb.201507009 We apologize to colleagues whose work could not be included be- Clijsters, L., W. van Zon, B.T. Riet, E. Voets, M. Boekhout, J. Ogink, C. Rumpf- cause of space limitations. We are grateful to Michael Rape for shar- Kienzl, and R.M.F. Wolthuis. 2014. Inefficient degradation of cyclin B1 re-activates the spindle checkpoint right after sister chromatid disjunction. ing results prior to publication, and we thank F. Lampert, R. Dechant, Cell Cycle. 13:2370–2378. http://dx.doi.org/10.4161/cc.29336 P. Kimmig, and A. Smith for helpful discussions and critical reading Coleman, K.E., G.D. Grant, R.A. Haggerty, K. Brantley, E. Shibata, of the manuscript. B.D. Workman, A. Dutta, D. Varma, J.E. Purvis, and J.G. Cook. 2015. Work in the Peter laboratory is funded by the Eidgenössische Sequential replication-coupled destruction at G1/S ensures genome stability. Genes Dev. 29:1734–1746. http://dx.doi.org/10.1101/gad Technische Hochschule Zürich (ETH-20 14-1 and ETH-46 16-1), the .263731.115 Swiss National Science Foundation (SNF 310030B_160321/1), and Craney, A., and M. Rape. 2013. Dynamic regulation of ubiquitin-dependent cell the European Research Council (ERC 268930 Rubinet). cycle control. Curr. Opin. Cell Biol. 25:704–710. http://dx.doi.org/10 The authors declare no competing financial interests. .1016/j.ceb.2013.07.004

10 JCB • 2017 Dankert, J.F., G. Rona, L. Clijsters, P. Geter, J.R. Skaar, K. Bermudez-Hernandez, modifications program FEN1 degradation during cell-cycle progression. E. Sassani, D. Fenyö, B. Ueberheide, R. Schneider, and M. Pagano. 2016. Mol. Cell. 47:444–456. http://dx.doi.org/10.1016/j.molcel.2012.05.042 Cyclin F-mediated degradation of SLBP limits H2A.X accumulation and Guturi, K.K.N., M. Bohgaki, T. Bohgaki, T. Srikumar, D. Ng, R. Kumareswaran, apoptosis upon genotoxic stress in G2. Mol. Cell. 64:507–519. http://dx S. El Ghamrasni, J. Jeon, P. Patel, M.S. Eldin, et al. 2016. RNF168 and .doi.org/10.1016/j.molcel.2016.09.010 USP10 regulate topoisomerase IIα function via opposing effects on Das-Bradoo, S., R.M. Ricke, and A.-K. Bielinsky. 2006. Interaction between its ubiquitylation. Nat. Commun. 7:12638. http://dx.doi.org/10.1038/ PCNA and diubiquitinated Mcm10 is essential for cell growth in budding ncomms12638 yeast. Mol. Cell. Biol. 26:4806–4817. http://dx.doi.org/10.1128/MCB Hames, R.S., S.L. Wattam, H. Yamano, R. Bacchieri, and A.M. Fry. 2001. APC/ .02062-05 C-mediated destruction of the centrosomal kinase Nek2A occurs in early den Elzen, N., and J. Pines. 2001. Cyclin A is destroyed in prometaphase and can mitosis and depends upon a cyclin A-type D-box. EMBO J. 20:7117– delay chromosome alignment and anaphase. J. Cell Biol. 153:121–136. 7127. http://dx.doi.org/10.1093/emboj/20.24.7117 http://dx.doi.org/10.1083/jcb.153.1.121 Han, J., Q. Li, L. McCullough, C. Kettelkamp, T. Formosa, and Z. Zhang. 2010. Deshaies, R.J., and C.A.P. Joazeiro. 2009. RING domain E3 ubiquitin ligases. Ubiquitylation of FACT by the cullin-E3 ligase Rtt101 connects FACT to Annu. Rev. Biochem. 78:399–434. http://dx.doi.org/10.1146/annurev DNA replication. Genes Dev. 24:1485–1490. http://dx.doi.org/10.1101/ .biochem.78.101807.093809 gad.1887310 Dewar, J.M., E. Low, M. Mann, M. Räschle, and J.C. Walter. 2017. CRL2(Lrr1) Han, J., H. Zhang, H. Zhang, Z. Wang, H. Zhou, and Z. Zhang. 2013. A Cul4 E3 promotes unloading of the vertebrate replisome from chromatin during ubiquitin ligase regulates histone hand-off during nucleosome assembly. replication termination. Genes Dev. 31:275–290. http://dx.doi.org/10 Cell. 155:817–829. http://dx.doi.org/10.1016/j.cell.2013.10.014 .1101/gad.291799.116 Hara, T., T. Kamura, S. Kotoshiba, H. Takahashi, K. Fujiwara, I. Onoyama, Di Fiore, B., N.E. Davey, A. Hagting, D. Izawa, J. Mansfeld, T.J. Gibson, and M. Shirakawa, N. Mizushima, and K.I. Nakayama. 2005. Role of the J. Pines. 2015. The ABBA motif binds APC/C activators and is shared UBL-UBA protein KPC2 in degradation of p27 at G1 phase of the cell by APC/C substrates and regulators. Dev. Cell. 32:358–372. http://dx.doi cycle. Mol. Cell. Biol. 25:9292–9303. http://dx.doi.org/10.1128/MCB.25 .org/10.1016/j.devcel.2015.01.003 .21.9292-9303.2005 Dimova, N.V., N.A. Hathaway, B.-H. Lee, D.S. Kirkpatrick, M.L. Berkowitz, Havens, C.G., and J.C. Walter. 2009. Docking of a specialized PIP box onto S.P. Gygi, D. Finley, and R.W. King. 2012. APC/C-mediated multiple chromatin-bound PCNA creates a degron for the ubiquitin ligase monoubiquitylation provides an alternative degradation signal for cyclin

CRL4Cdt2. Mol. Cell. 35:93–104. http://dx.doi.org/10.1016/j.molcel Downloaded from B1. Nat. Cell Biol. 14:168–176. http://dx.doi.org/10.1038/ncb2425 .2009.05.012 Dobrynin, G., O. Popp, T. Romer, S. Bremer, M.H.A. Schmitz, D.W. Gerlich, Havens, C.G., N. Shobnam, E. Guarino, R.C. Centore, L. Zou, S.E. Kearsey, and H. Meyer. 2011. Cdc48/p97-Ufd1-Npl4 antagonizes Aurora B during and J.C. Walter. 2012. Direct role for proliferating cell nuclear antigen chromosome segregation in HeLa cells. J. Cell Sci. 124:1571–1580. in substrate recognition by the E3 ubiquitin ligase CRL4Cdt2. J. Biol. http://dx.doi.org/10.1242/jcs.069500 Chem. 287:11410–11421. http://dx.doi.org/10.1074/jbc.M111.337683 Dowlatshahi, D.P., V. Sandrin, S. Vivona, T.A. Shaler, S.E. Kaiser, F. Melandri, Hirano, T. 2015. Chromosome dynamics during mitosis. Cold Spring Harb. W.I. Sundquist, and R.R. Kopito. 2012. ALIX is a Lys63-specific Perspect. Biol. 7:1–14. http://dx.doi.org/10.1101/cshperspect.a015792 polyubiquitin binding protein that functions in retrovirus budding. Dev.

Cell. 23:1247–1254. http://dx.doi.org/10.1016/j.devcel.2012.10.023 Husnjak, K., and I. Dikic. 2012. Ubiquitin-binding proteins: Decoders of jcb.rupress.org ubiquitin-mediated cellular functions. Annu. Rev. Biochem. 81:291–322. Foster, S.A., and D.O. Morgan. 2012. The APC/C subunit Mnd2/Apc15 promotes http://dx.doi.org/10.1146/annurev-biochem-051810-094654 Cdc20 autoubiquitination and spindle assembly checkpoint inactivation. Mol. Cell. 47:921–932. http://dx.doi.org/10.1016/j.molcel.2012.07.031 Irniger, S., S. Piatti, C. Michaelis, and K. Nasmyth. 1995. Genes involved in sister chromatid separation are needed for B-type cyclin proteolysis Fragkos, M., O. Ganier, P. Coulombe, and M. Méchali. 2015. DNA replication in budding yeast. Cell. 81:269–278. http://dx.doi.org/10.1016/0092 origin activation in space and time. Nat. Rev. Mol. Cell Biol. 16:360–374. -8674(95)90337-2 http://dx.doi.org/10.1038/nrm4002 on August 6, 2017 Franz, A., L. Ackermann, and T. Hoppe. 2016. Ring of change: CDC48/p97 Jin, L., A. Williamson, S. Banerjee, I. Philipp, and M. Rape. 2008. Mechanism drives protein dynamics at chromatin. Front. Genet. 7:73. http://dx.doi of ubiquitin-chain formation by the human anaphase-promoting complex. .org/10.3389/fgene.2016.00073 Cell. 133:653–665. http://dx.doi.org/10.1016/j.cell.2008.04.012 Gao, Y.-F., T. Li, Y. Chang, Y.-B. Wang, W.-N. Zhang, W.-H. Li, K. He, R. Mu, Kaisari, S., D. Sitry-Shevah, S. Miniowitz-Shemtov, A. Teichner, and A. Hershko. C. Zhen, J.-H. Man, et al. 2011. Cdk1-phosphorylated CUEDC2 promotes 2017. Role of CCT chaperonin in the disassembly of mitotic checkpoint spindle checkpoint inactivation and chromosomal instability. Nat. Cell complexes. Proc. Natl. Acad. Sci. USA. 114:956–961. http://dx.doi.org Biol. 13:924–933. http://dx.doi.org/10.1038/ncb2287 /10.1073/pnas.1620451114 García-Rodríguez, N., R.P. Wong, and H.D. Ulrich. 2016. Functions of ubiquitin Kawaguchi, Y., J.J. Kovacs, A. McLaurin, J.M. Vance, A. Ito, and T.P. Yao. 2003. and SUMO in DNA replication and replication stress. Front. Genet. 7:87. The deacetylase HDAC6 regulates aggresome formation and cell viability http://dx.doi.org/10.3389/fgene.2016.00087 in response to misfolded protein stress. Cell. 115:727–738. http://dx.doi .org/10.1016/S0092-8674(03)00939-5 Garnett, M.J., J. Mansfeld, C. Godwin, T. Matsusaka, J. Wu, P. Russell, J. Pines, and A.R. Venkitaraman. 2009. UBE2S elongates ubiquitin chains on King, R.W., J.M. Peters, S. Tugendreich, M. Rolfe, P. Hieter, and M.W. Kirschner. APC/C substrates to promote mitotic exit. Nat. Cell Biol. 11:1363–1369. 1995. A 20S complex containing CDC27 and CDC16 catalyzes the http://dx.doi.org/10.1038/ncb1983 mitosis-specific conjugation of ubiquitin to cyclin B. Cell. 81:279–288. http://dx.doi.org/10.1016/0092-8674(95)90338-0 Gavet, O., and J. Pines. 2010. Progressive activation of CyclinB1-Cdk1 coordinates entry to mitosis. Dev. Cell. 18:533–543. http://dx.doi.org/10 Komander, D., and M. Rape. 2012. The ubiquitin code. Annu. Rev. Biochem. .1016/j.devcel.2010.02.013 81:203–229. http://dx.doi.org/10.1146/annurev-biochem-060310-170328 Geley, S., E. Kramer, C. Gieffers, J. Gannon, J.M. Peters, and T. Hunt. 2001. Kong, M., D. Ditsworth, T. Lindsten, and C.B. Thompson. 2009. α4 is an Anaphase-promoting complex/cyclosome-dependent proteolysis of essential regulator of PP2A phosphatase activity. Mol. Cell. 36:51–60. human cyclin A starts at the beginning of mitosis and is not subject to the http://dx.doi.org/10.1016/j.molcel.2009.09.025 spindle assembly checkpoint. J. Cell Biol. 153:137–148. http://dx.doi.org Krenn, V., and A. Musacchio. 2015. The Aurora B kinase in chromosome bi- /10.1083/jcb.153.1.137 orientation and spindle checkpoint signaling. Front. Oncol. 5:225. http:// Grice, G.L., and J.A. Nathan. 2016. The recognition of ubiquitinated proteins by dx.doi.org/10.3389/fonc.2015.00225 the proteasome. Cell. Mol. Life Sci. 73:3497–3506. http://dx.doi.org/10 Krupina, K., C. Kleiss, T. Metzger, S. Fournane, S. Schmucker, K. Hofmann, .1007/s00018-016-2255-5 B. Fischer, N. Paul, I.M. Porter, W. Raffelsberger, et al. 2016. Ubiquitin Grice, G.L., I.T. Lobb, M.P. Weekes, S.P. Gygi, R. Antrobus, and J.A. Nathan. receptor protein UBASH3B drives Aurora B recruitment to mitotic 2015. The proteasome distinguishes between heterotypic and homotypic microtubules. Dev. Cell. 36:63–78. http://dx.doi.org/10.1016/j.devcel lysine-11-linked polyubiquitin chains. Cell Reports. 12:545–553. .2015.12.017 http://dx.doi.org/10.1016/j.celrep.2015.06.061 Lampert, F., M.M.L. Brodersen, and M. Peter. 2017. Guard the guardian: A Gruneberg, U., R. Neef, R. Honda, E.A. Nigg, and F.A. Barr. 2004. Relocation CRL4 ligase stands watch over histone production. Nucleus. 8:134–143. of Aurora B from centromeres to the central spindle at the metaphase to http://dx.doi.org/10.1080/19491034.2016.1276143 anaphase transition requires MKlp2. J. Cell Biol. 166:167–172. http://dx Lee, M.Y.W.T., S. Zhang, S.H.S. Lin, J. Chea, X. Wang, C. LeRoy, A. Wong, .doi.org/10.1083/jcb.200403084 Z. Zhang, and E.Y.C. Lee. 2012. Regulation of human DNA polymerase Guo, Z., J. Kanjanapangka, N. Liu, S. Liu, C. Liu, Z. Wu, Y. Wang, T. Loh, delta in the cellular responses to DNA damage. Environ. Mol. Mutagen. C. Kowolik, J. Jamsen, et al. 2012. Sequential posttranslational 53:683–698. http://dx.doi.org/10.1002/em.21743

Dynamic ubiquitylation in the cell cycle • Gilberto and Peter 11 Lee, M.Y.W.T., S. Zhang, S.H.S. Lin, X. Wang, Z. Darzynkiewicz, Z. Zhang, and Moghe, S., F. Jiang, Y. Miura, R.L. Cerny, M.-Y. Tsai, and M. Furukawa. 2012. E.Y.C. Lee. 2014. The tail that wags the dog: p12, the smallest subunit of The CUL3-KLHL18 ligase regulates mitotic entry and ubiquitylates DNA polymerase δ, is degraded by ubiquitin ligases in response to DNA Aurora-A. Biol. Open. 1:82–91. http://dx.doi.org/10.1242/bio.2011018 damage and during cell cycle progression. Cell Cycle. 13:23–31. http:// Moldovan, G.L., B. Pfander, and S. Jentsch. 2007. PCNA, the maestro of the dx.doi.org/10.4161/cc.27407 replication fork. Cell. 129:665–679. http://dx.doi.org/10.1016/j.cell.2007 Lim, K.H., M.H. Song, and K.H. Baek. 2016. Decision for cell fate: .05.003 Deubiquitinating enzymes in cell cycle checkpoint. Cell. Mol. Life Sci. Moreno, S.P., and A. Gambus. 2015. Regulation of unperturbed DNA replication 73:1439–1455. http://dx.doi.org/10.1007/s00018-015-2129-2 by ubiquitylation. Genes (Basel). 6:451–468. Lin, S.H.S., X. Wang, S. Zhang, Z. Zhang, E.Y.C. Lee, and M.Y.W.T. Lee. 2013. Moreno, S.P., R. Bailey, N. Campion, S. Herron, and A. Gambus. 2014. Dynamics of enzymatic interactions during short flap human Okazaki Polyubiquitylation drives replisome disassembly at the termination of fragment processing by two forms of human DNA polymerase δ. DNA DNA replication. Science. 346:477–481. http://dx.doi.org/10.1126/ Repair (Amst.). 12:922–935. http://dx.doi.org/10.1016/j.dnarep.2013.08.008 science.1253585 Lischetti, T., and J. Nilsson. 2015. Regulation of mitotic progression by the Morita, E., V. Sandrin, H.-Y. Chung, S.G. Morham, S.P. Gygi, C.K. Rodesch, and spindle assembly checkpoint. Mol. Cell. Oncol. 2:e970484. http://dx.doi W.I. Sundquist. 2007. Human ESCRT and ALIX proteins interact with .org/10.4161/23723548.2014.970484 proteins of the midbody and function in cytokinesis. EMBO J. 26:4215– Liu, D., O. Davydenko, and M.A. Lampson. 2012. Polo-like kinase-1 regulates 4227. http://dx.doi.org/10.1038/sj.emboj.7601850 kinetochore-microtubule dynamics and spindle checkpoint silencing. Morohashi, H., T. Maculins, and K. Labib. 2009. The amino-terminal TPR J. Cell Biol. 198:491–499. http://dx.doi.org/10.1083/jcb.201205090 domain of Dia2 tethers SCF(Dia2) to the replisome progression complex. Lou, Z., K. Minter-Dykhouse, and J. Chen. 2005. BRCA1 participates in DNA Curr. Biol. 19:1943–1949. http://dx.doi.org/10.1016/j.cub.2009.09.062 decatenation. Nat. Struct. Mol. Biol. 12:589–593. http://dx.doi.org/10 Mouysset, J., S. Gilberto, M.G. Meier, F. Lampert, M. Belwal, P. Meraldi, .1038/nsmb953 and M. Peter. 2015. CRL4(RBBP7) is required for efficient CENP-A Lu, D., J.Y. Hsiao, N.E. Davey, V.A. Van Voorhis, S.A. Foster, C. Tang, and deposition at centromeres. J. Cell Sci. 128:1732–1745. http://dx.doi.org D.O. Morgan. 2014. Multiple mechanisms determine the order of APC/C /10.1242/jcs.162305 substrate degradation in mitosis. J. Cell Biol. 207:23–39. http://dx.doi.org Mulvaney, K.M., J.P. Matson, P.F. Siesser, T.Y. Tamir, D. Goldfarb, T.M. Jacobs, /10.1083/jcb.201402041 E.W. Cloer, J.S. Harrison, C. Vaziri, J.G. Cook, and M.B. Major. 2016.

Lu, D., J.R. Girard, W. Li, A. Mizrak, and D.O. Morgan. 2015a. Quantitative Identification and characterization of MCM3 as a kelch-like ECH- Downloaded from framework for ordered degradation of APC/C substrates. BMC Biol. associated protein 1 (KEAP1) substrate. J. Biol. Chem. 291:23719– 13:96. http://dx.doi.org/10.1186/s12915-015-0205-6 23733. http://dx.doi.org/10.1074/jbc.M116.729418 Lu, Y., B.H. Lee, R.W. King, D. Finley, and M.W. Kirschner. 2015b. Substrate Nakagawa, T., and K. Nakayama. 2015. Protein monoubiquitylation: Targets and degradation by the proteasome: A single-molecule kinetic analysis. diverse functions. Genes Cells. 20:543–562. http://dx.doi.org/10.1111/ Science. 348:1250834. http://dx.doi.org/10.1126/science.1250834 gtc.12250 Lydeard, J.R., B.A. Schulman, and J.W. Harper. 2013. Building and remodelling Neef, R., C. Preisinger, J. Sutcliffe, R. Kopajtich, E.A. Nigg, T.U. Mayer, and Cullin-RING E3 ubiquitin ligases. EMBO Rep. 14:1050–1061. http://dx F.A. Barr. 2003. Phosphorylation of mitotic kinesin-like protein 2 by polo-like kinase 1 is required for cytokinesis. J. Cell Biol. 162:863–875. .doi.org/10.1038/embor.2013.173 jcb.rupress.org http://dx.doi.org/10.1083/jcb.200306009 Maculins, T., P.J. Nkosi, H. Nishikawa, and K. Labib. 2015. Tethering of SCF(Dia2) to the replisome promotes efficient ubiquitylation and Niikura, Y., R. Kitagawa, H. Ogi, R. Abdulle, V. Pagala, and K. Kitagawa. 2015. disassembly of the CMG helicase. Curr. Biol. 25:2254–2259. http://dx CENP-A K124 ubiquitylation is required for CENP-a deposition at the .doi.org/10.1016/j.cub.2015.07.012 centromere. Dev. Cell. 32:589–603. http://dx.doi.org/10.1016/j.devcel .2015.01.024 Maerki, S., M.H. Olma, T. Staubli, P. Steigemann, D.W. Gerlich, M. Quadroni, I. Sumara, and M. Peter. 2009. The Cul3-KLHL21 E3 ubiquitin ligase Ogunjimi, A.A., S. Wiesner, D.J. Briant, X. Varelas, F. Sicheri, J. Forman- targets aurora B to midzone microtubules in anaphase and is required Kay, and J.L. Wrana. 2010. The ubiquitin binding region of the Smurf on August 6, 2017 for cytokinesis. J. Cell Biol. 187:791–800. http://dx.doi.org/10.1083/jcb HECT domain facilitates polyubiquitylation and binding of ubiquitylated .200906117 substrates. J. Biol. Chem. 285:6308–6315. http://dx.doi.org/10.1074/jbc .M109.044537 Mansfeld, J., P. Collin, M.O. Collins, J.S. Choudhary, and J. Pines. 2011. APC15 drives the turnover of MCC-CDC20 to make the spindle assembly Olmos, Y., L. Hodgson, J. Mantell, P. Verkade, and J.G. Carlton. 2015. ESCRT- checkpoint responsive to kinetochore attachment. Nat. Cell Biol. III controls nuclear envelope reformation. Nature. 522:236–239. http://dx 13:1234–1243. http://dx.doi.org/10.1038/ncb2347 .doi.org/10.1038/nature14503 Maric, M., T. Maculins, G. De Piccoli, and K. Labib. 2014. Cdc48 and a Orthwein, A., S.M. Noordermeer, M.D. Wilson, S. Landry, R.I. Enchev, ubiquitin ligase drive disassembly of the CMG helicase at the end of A. Sherker, M. Munro, J. Pinder, J. Salsman, G. Dellaire, et al. 2015. A DNA replication. Science. 346:1253596. http://dx.doi.org/10.1126/ mechanism for the suppression of homologous recombination in G1 cells. science.1253596 Nature. 528:422–426. http://dx.doi.org/10.1038/nature16142 Matsumoto, M.L., K.E. Wickliffe, K.C. Dong, C. Yu, I. Bosanac, D. Bustos, Ossareh-Nazari, B., A. Katsiarimpa, J. Merlet, and L. Pintard. 2016. RNAi-based L. Phu, D.S. Kirkpatrick, S.G. Hymowitz, M. Rape, et al. 2010. K11- suppressor screens reveal genetic interactions between the CRL2LRR-1 linked polyubiquitination in cell cycle control revealed by a K11 linkage- E3-ligase and the DNA replication machinery in Caenorhabditis elegans. specific antibody. Mol. Cell. 39:477–484. http://dx.doi.org/10.1016/j G3 (Bethesda). 6:3431–3442. .molcel.2010.07.001 Ouyang, H., Y.O. Ali, M. Ravichandran, A. Dong, W. Qiu, F. MacKenzie, S. Dhe- McConnell, J.L., G.R. Watkins, S.E. Soss, H.S. Franz, L.R. McCorvey, Paganon, C.H. Arrowsmith, and R.G. Zhai. 2012. Protein aggregates are B.W. Spiller, W.J. Chazin, and B.E. Wadzinski. 2010. Alpha4 is a recruited to aggresome by histone deacetylase 6 via unanchored ubiquitin ubiquitin-binding protein that regulates protein serine/threonine C termini. J. Biol. Chem. 287:2317–2327. http://dx.doi.org/10.1074/jbc phosphatase 2A ubiquitination. Biochemistry. 49:1713–1718. http://dx .M111.273730 .doi.org/10.1021/bi901837h Pashkova, N., L. Gakhar, S.C. Winistorfer, L. Yu, S. Ramaswamy, and R.C. Piper. Meyer, H.J., and M. Rape. 2014. Enhanced protein degradation by branched 2010. WD40 repeat propellers define a ubiquitin-binding domain that ubiquitin chains. Cell. 157:910–921. http://dx.doi.org/10.1016/j.cell regulates turnover of F box proteins. Mol. Cell. 40:433–443. http://dx.doi .2014.03.037 .org/10.1016/j.molcel.2010.10.018 Meyer, H., M. Bug, and S. Bremer. 2012. Emerging functions of the VCP/ Radulescu, A.E., and D.W. Cleveland. 2010. NuMA after 30 years: The matrix p97 AAA-ATPase in the ubiquitin system. Nat. Cell Biol. 14:117–123. revisited. Trends Cell Biol. 20:214–222. http://dx.doi.org/10.1016/j.tcb http://dx.doi.org/10.1038/ncb2407 .2010.01.003 Min, M., T.E. Mevissen, M. De Luca, D. Komander, and C. Lindon. 2015. Ramadan, K., R. Bruderer, F.M. Spiga, O. Popp, T. Baur, M. Gotta, and Efficient APC/C substrate degradation in cells undergoing mitotic exit H.H. Meyer. 2007. Cdc48/p97 promotes reformation of the nucleus by depends on K11 ubiquitin linkages. Mol. Biol. Cell. 26:4325–4332. extracting the kinase Aurora B from chromatin. Nature. 450:1258–1262. http://dx.doi.org/10.1091/mbc.E15-02-0102 http://dx.doi.org/10.1038/nature06388 Miniowitz-Shemtov, S., E. Eytan, S. Kaisari, D. Sitry-Shevah, and A. Hershko. Rattani, A., P.K. Vinod, J. Godwin, K. Tachibana-Konwalski, M. Wolna, 2015. Mode of interaction of TRIP13 AAA-ATPase with the Mad2- M. Malumbres, B. Novák, and K. Nasmyth. 2014. Dependency of the binding protein p31comet and with mitotic checkpoint complexes. Proc. spindle assembly checkpoint on Cdk1 renders the anaphase transition Natl. Acad. Sci. USA. 112:11536–11540. http://dx.doi.org/10.1073/pnas irreversible. Curr. Biol. 24:630–637. http://dx.doi.org/10.1016/j.cub .1515358112 .2014.01.033

12 JCB • 2017 Reddy, S.K., M. Rape, W.A. Margansky, and M.W. Kirschner. 2007. Williamson, A., K.E. Wickliffe, B.G. Mellone, L. Song, G.H. Karpen, and Ubiquitination by the anaphase-promoting complex drives spindle M. Rape. 2009. Identification of a physiological E2 module for the human checkpoint inactivation. Nature. 446:921–925. http://dx.doi.org/10.1038 anaphase-promoting complex. Proc. Natl. Acad. Sci. USA. 106:18213– /nature05734 18218. http://dx.doi.org/10.1073/pnas.0907887106 Rodrigo-Brenni, M.C., and D.O. Morgan. 2007. Sequential E2s drive Yamaguchi, M., R. VanderLinden, F. Weissmann, R. Qiao, P. Dube, N.G. Brown, polyubiquitin chain assembly on APC targets. Cell. 130:127–139. D. Haselbach, W. Zhang, S.S. Sidhu, J.M. Peters, et al. 2016. Cryo-EM http://dx.doi.org/10.1016/j.cell.2007.05.027 of mitotic checkpoint complex-bound APC/C reveals reciprocal and Scott, D.C., D.Y. Rhee, D.M. Duda, I.R. Kelsall, J.L. Olszewski, J.A. Paulo, conformational regulation of ubiquitin ligation. Mol. Cell. 63:593–607. A. de Jong, H. Ovaa, A.F. Alpi, J.W. Harper, and B.A. Schulman. 2016. http://dx.doi.org/10.1016/j.molcel.2016.07.003 Two distinct types of E3 ligases work in unison to regulate substrate Yan, K., L. Li, X. Wang, R. Hong, Y. Zhang, H. Yang, M. Lin, S. Zhang, Q. He, ubiquitylation. Cell. 166:1198–1214.e24. http://dx.doi.org/10.1016/j.cell D. Zheng, et al. 2015. The deubiquitinating enzyme complex BRISC is .2016.07.027 required for proper mitotic spindle assembly in mammalian cells. J. Cell Sivakumar, S., and G.J. Gorbsky. 2015. Spatiotemporal regulation of the Biol. 210:209–224. http://dx.doi.org/10.1083/jcb.201503039 anaphase-promoting complex in mitosis. Nat. Rev. Mol. Cell Biol. 16:82– Yang, Y., M. Liu, D. Li, J. Ran, J. Gao, S. Suo, S.-C. Sun, and J. Zhou. 2014. 94. http://dx.doi.org/10.1038/nrm3934 CYLD regulates spindle orientation by stabilizing astral microtubules Spratt, D.E., H. Walden, and G.S. Shaw. 2014. RBR E3 ubiquitin ligases: New and promoting dishevelled-NuMA-dynein/dynactin complex formation. structures, new insights, new questions. Biochem. J. 458:421–437. Proc. Natl. Acad. Sci. USA. 111:2158–2163. http://dx.doi.org/10.1073/ http://dx.doi.org/10.1042/BJ20140006 pnas.1319341111 Stegmeier, F., M. Rape, V.M. Draviam, G. Nalepa, M.E. Sowa, X.L. Ang, Yau, R., and M. Rape. 2016. The increasing complexity of the ubiquitin code. E.R. McDonald III, M.Z. Li, G.J. Hannon, P.K. Sorger, et al. 2007. Nat. Cell Biol. 18:579–586. http://dx.doi.org/10.1038/ncb3358 Anaphase initiation is regulated by antagonistic ubiquitination and Ye, Y., and M. Rape. 2009. Building ubiquitin chains: E2 enzymes at work. Nat. deubiquitination activities. Nature. 446:876–881. http://dx.doi.org/10 Rev. Mol. Cell Biol. 10:755–764. http://dx.doi.org/10.1038/nrm2780 .1038/nature05694 Yu, H., J.M. Peters, R.W. King, A.M. Page, P. Hieter, and M.W. Kirschner. 1998. Sudakin, V., D. Ganoth, A. Dahan, H. Heller, J. Hershko, F.C. Luca, Identification of a cullin homology region in a subunit of the anaphase- J.V. Ruderman, and A. Hershko. 1995. The cyclosome, a large complex promoting complex. Science. 279:1219–1222. http://dx.doi.org/10.1126/ containing cyclin-selective ubiquitin ligase activity, targets cyclins for science.279.5354.1219 Downloaded from destruction at the end of mitosis. Mol. Biol. Cell. 6:185–197. http://dx.doi Zaidi, I.W., G. Rabut, A. Poveda, H. Scheel, J. Malmström, H. Ulrich, .org/10.1091/mbc.6.2.185 K. Hofmann, P. Pasero, M. Peter, and B. Luke. 2008. Rtt101 and Mms1 Sumara, I., M. Quadroni, C. Frei, M.H. Olma, G. Sumara, R. Ricci, and in budding yeast form a CUL4(DDB1)-like ubiquitin ligase that promotes M. Peter. 2007. A Cul3-based E3 ligase removes Aurora B from mitotic replication through damaged DNA. EMBO Rep. 9:1034–1040. http://dx chromosomes, regulating mitotic progression and completion of .doi.org/10.1038/embor.2008.155 cytokinesis in human cells. Dev. Cell. 12:887–900. http://dx.doi.org/10 Zhang, S., H. Zhao, Z. Darzynkiewicz, P. Zhou, Z. Zhang, E.Y.C. Lee, and .1016/j.devcel.2007.03.019 M.Y.W.T. Lee. 2013. A novel function of CRL4Cdt2: Regulation of the Teixeira, L.K., and S.I. Reed. 2013. Ubiquitin ligases and cell cycle control. subunit structure of DNA polymerase in response to DNA damage and

Annu. Rev. Biochem. 82:387–414. http://dx.doi.org/10.1146/annurev during the S phase. J. Biol. Chem. 288:29550–29561. http://dx.doi.org/10 jcb.rupress.org -biochem-060410-105307 .1074/jbc.M113.490466 Thu, Y.M., and A.K. Bielinsky. 2014. MCM10: one tool for all—Integrity, Zhang, X., J. Cai, Z. Zheng, L. Polin, Z. Lin, A. Dandekar, L. Li, F. Sun, maintenance and damage control. Semin. Cell Dev. Biol. 30:121–130. R.L. Finley Jr., D. Fang, et al. 2015. A novel ER-microtubule-binding http://dx.doi.org/10.1016/j.semcdb.2014.03.017 protein, ERLIN2, stabilizes Cyclin B1 and regulates cell cycle Trujillo, K.M., and M.A. Osley. 2012. A role for H2B ubiquitylation in DNA progression. Cell Discov. 1:15024. http://dx.doi.org/10.1038/celldisc replication. Mol. Cell. 48:734–746. http://dx.doi.org/10.1016/j.molcel .2015.24

.2012.09.019 Zheng, L., and B. Shen. 2011. Okazaki fragment maturation: Nucleases take on August 6, 2017 UniProt Consortium. 2015. UniProt: A hub for protein information. Nucleic centre stage. J. Mol. Cell Biol. 3:23–30. http://dx.doi.org/10.1093/jmcb Acids Res. 43(D1):D204–D212. http://dx.doi.org/10.1093/nar/gku989 /mjq048 Varetti, G., C. Guida, S. Santaguida, E. Chiroli, and A. Musacchio. 2011. Zhou, X.X., and M.Z. Lin. 2013. Photoswitchable fluorescent proteins: Ten years Homeostatic control of mitotic arrest. Mol. Cell. 44:710–720. http://dx of colorful chemistry and exciting applications. Curr. Opin. Chem. Biol. .doi.org/10.1016/j.molcel.2011.11.014 17:682–690. http://dx.doi.org/10.1016/j.cbpa.2013.05.031 Vázquez-Novelle, M.D., L. Sansregret, A.E. Dick, C.A. Smith, A.D. McAinsh, Zhou, Z., M. He, A.A. Shah, and Y. Wan. 2016. Insights into APC/C: From D.W. Gerlich, and M. Petronczki. 2014. Cdk1 inactivation terminates mitotic cellular function to diseases and therapeutics. Cell Div. 11:9. http://dx.doi checkpoint surveillance and stabilizes kinetochore attachments in anaphase. .org/10.1186/s13008-016-0021-6 Curr. Biol. 24:638–645. http://dx.doi.org/10.1016/j.cub.2014.01.034 Zhuo, X., X. Guo, X. Zhang, G. Jing, Y. Wang, Q. Chen, Q. Jiang, J. Liu, and Vong, Q.P., K. Cao, H.Y. Li, P.A. Iglesias, and Y. Zheng. 2005. Chromosome C. Zhang. 2015. Usp16 regulates kinetochore localization of Plk1 to alignment and segregation regulated by ubiquitination of survivin. promote proper chromosome alignment in mitosis. J. Cell Biol. 210:727– Science. 310:1499–1504. http://dx.doi.org/10.1126/science.1120160 735. http://dx.doi.org/10.1083/jcb.201502044 Westhorpe, F.G., A. Tighe, P. Lara-Gonzalez, and S.S. Taylor. 2011. p31comet- Zitouni, S., C. Nabais, S.C. Jana, A. Guerrero, and M. Bettencourt-Dias. 2014. mediated extraction of Mad2 from the MCC promotes efficient mitotic Polo-like kinases: Structural variations lead to multiple functions. Nat. exit. J. Cell Sci. 124:3905–3916. http://dx.doi.org/10.1242/jcs.093286 Rev. Mol. Cell Biol. 15:433–452. http://dx.doi.org/10.1038/nrm3819

Dynamic ubiquitylation in the cell cycle • Gilberto and Peter 13 8. Acknowledgements

First of all, I would like to thank Matthias for giving me the opportunity to perform my graduate studies in his lab. I very much appreciated all the support throughout the years, all the discussions in the most exciting (and not so exciting) moments and for the nice lab atmosphere. Of course, without Matthias this work wouldn’t have been possible, and for this reason I will always be grateful to him. Secondly, I would like to thank my thesis committee members, for valuable guidance provided throughout the years, especially through the changes my project(s) suffered along the course of my studies. Specifically, I would like to thank Prof. Karsten Weis and Prof. Anton Wutz for accepting this challenge, and also Pierre Gönczy for the early guidance. A big thank you for the people that always encouraged me throughout my studies before arriving to Zurich, especially Gonçalo Conde da Costa and Mónica Bettencourt-Dias! Moreover, I very much appreciate the opportunity to integrate the MLS PhD program here in Zurich, thank you Susanna Bachmann! Thank you to all the Peter lab members whom I have the pleasure of coming across throughout these years. We had great discussions and an awesome atmosphere! Many of you are not only my colleagues but also truly my friends! In particular, thank you Shady, Wojtek, Phil, Reinhard, Remy, Julien, Gea, Eduardo, Fabi, Diana, Vanessa, Amandine, Anna and all the other fantastic people that were or are in the Peter lab. I’m also very grateful to all the great friendships I made within the IBC (past and present), you make the life of a PhD student far more enjoyable! Big thanks to Carina, Martin Soste, Phil “the Wild”, Arun, Martin Peterka, Sebastian, Qian and Yuehan. And in particular, thank you dude Melnik for the awesome home atmosphere (you will always be welcome to 13.02)! Still within the IBC, thank you very much to all the staff, especially Tony and Barbara Schuhbeck-Wagner! More than just within the IBC, much of my enjoyment in life is due to all the friends I have outside of the working environment. Thank you to the Vinzenzheimers, especially Spitzi, Dominik, Lyo and Marta(s)! And of course, my friends from Portugal that have been incessantly pushing me forward: Custódio, Nuno (Jet), Márcia, André, Samuel and Inês. A special thank you to Maria João, a true friend!

182

8. Acknowledgements

This thesis wouldn’t have been possible without my family. A very special thank you goes to them, for which I would like to leave some words in Portuguese:

Muito obrigado à minha família, sem vocês esta tese não teria sido escrita. É impossível descrever o quanto agradecido me sinto por vos ter como família. Sei que não é nada fácil não me terem por perto (para mim também é difícil), mas agradeço imenso todo o apoio incessante ao longo destes anos. Nem uma vez me disseram “não vás”, ainda que provavelmente o vosso coração queria dizê-lo. Sempre me deram toda a força e apoio, e por isso muito vos agradeço. Muito obrigado Mãe, Pai, Nelson e Jorge. Mas claro, também aos pequenos, Isac e Sofia, e à Daniela. O vosso apoio incondicional significa muito para mim. Vocês têm um lugar muito especial no meu coração!

Last but definitely not least, I wouldn’t be able to express in words how thankful I am to you, Hanna. My heater.

183

9. Curriculum vitae

PERSONAL DATA

NAME: Samuel Filipe da Graça Gilberto

DATE OF BIRTH: February 21st 1988

NATIONALITY: Portuguese

ADDRESS: Tramstrasse 30, 8050 Zurich, Switzerland

TELEPHONE: +41 764497408

E-MAIL: [email protected]

EDUCATION

2012-now Candidate for Dr. sc. ETH Zurich; Work conducted at the Institute of Biochemistry, Eidgenössische Technische Hochschule (ETH) Zürich – Zurich, Switzerland. • Planning and experimental execution of scientific research projects, as part of a multidisciplinary research team • Teaching and mentoring undergraduate students in analytical methods • Presentation of own research in multiple national and international scientific meetings • Organization of scientific meetings and social events

2009-2011 M.Sc. in Biochemistry, graded 19/20 – University of Lisbon, Portugal

2010 ERASMUS Exchange (part of M.Sc.) – University of Wrocław, Poland

2006-2009 B.Sc. in Biochemistry, graded 17/20 – University of Lisbon, Portugal

2003-2006 High School, graded 185/200 – Escola Sec. Santa Maria do Olival – Tomar, Portugal

184

9. Curriculum vitae

PROFESSIONAL EXPERIENCE

2012-now PhD research project: “Non-redundant functions of CUL4A and CUL4B” in the Peter group, ETH Zurich.

2011-2012 Research assistant, investigating the role of the kinase PLK4 in centriole biogenesis, at Instituto Gulbenkian de Ciência (IGC) – Oeiras, Portugal.

2010-2011 Master’s research project: “Effect of fibrinogen in the formation of TTR fibers in patients with familial amyloidosis” at the Faculty of Sciences of the University of Lisbon.

2010 Undergraduate research assistant, under the Erasmus Programme, at the Faculty of Biotechnology of the University of Wrocław.

2008-2009 Undergraduate research assistant under the FAD/UL scholarship at the Faculty of Sciences of the University of Lisbon.

2006-2009 Laboratory technician at the water treatment company EPAL, Asseiceira under the youth-oriented ‘OTL Empresa’ programme (1 month/year).

AWARDS

Jul 2016 Travel grant of the Zurich Molecular Life Sciences PhD programme

Jul 2014 Travel grant of the Zurich Molecular Life Sciences PhD programme

Sep 2011 Social merit scholarship granted by the Faculty of Sciences of the University of Lisbon (for class monitoring on the “complements of analytical biochemistry” master course)

Dec 2010 Best Poster presentation, at the 4th Portuguese Mass Spectrometry Meeting

2010 ERASMUS scholarship to study as an exchange student in Wrocław, Poland

Sep 2008 Research initiation scholarship granted by the Amadeu Dias Foundation and the University of Lisbon (FAD/UL)

Apr 2008 Honorable mention granted by the Faculty of Sciences of the University of Lisbon Foundation

Sep 2006 Diploma of merit attributed by Escola Secundária Santa Maria do Olival (high school)

Sep 2004 Diploma of merit attributed by Escola Secundária Santa Maria do Olival (high school)

185

9. Curriculum vitae

PUBLICATIONS

Gilberto S, Peter M. Dynamic ubiquitin signaling in cell cycle regulation. J Cell Biol 2017. pii: jcb.201703170 [review]

Fonseca D*, Gilberto S*, Ribeiro-Silva C, Ribeiro R, Guinote IB, Saraiva S, Gomes RA, Mateus É, Viana A, Barroso E, Freire AP, Freire P, Cordeiro C, da Costa G. The role of fibrinogen glycation in ATTR: evidence for chaperone activity loss in disease. Biochem J 2016, 473 (14): 2225-2237 *Equal contribution

Zitouni S, Francia ME, Leal F, Montenegro Gouveia S, Nabais C, Duarte P, Gilberto S, Brito D, Moyer T, Kandels- Lewis S, Ohta M, Kitagawa D, Holland AJ, Karsenti E, Lorca T, Lince-Faria M, Bettencourt-Dias M. CDK1 Prevents Unscheduled PLK4-STIL Complex Assembly in Centriole Biogenesis. Curr Biol 2016, 26 (9): 1127-1137

Lopes CAM, Jana SC, Cunha-Ferreira I, Zitouni S, Bento I, Duarte P, Gilberto S, Freixo F, Guerreiro A, Francia M, Lince-Faria M, Carneiro J, Bettencourt-Dias M. PLK4 trans-Autoactivation Controls Centriole Biogenesis in Space. Dev Cell 2015, 35: 1-14 da Costa G, Ribeiro-Silva C, Ribeiro R, Gilberto S, Gomes RA, Ferreira A, Mateus É, Barroso E, Coelho AV, Freire AP, Cordeiro C. Transthyretin Amyloidosis: Chaperone Concentration Changes and Increased Proteolysis in the Pathway to Disease. PLoS One 2015, 10 (7)

Mouysset J*, Gilberto S*, Meier MG, Lampert F, Belwal M, Meraldi P, Peter M. CRL4RBBP7 is required for efficient CENP-A deposition at centromeres. J Cell Sci 2015, 128 (9): 1732-45 *Equal contribution

Cunha-Ferreira I, Bento I, Pimenta-Marques A, Jana SC, Lince-Faria M, Duarte P, Borrego-Pinto J, Gilberto S, Amado T, Brito D, Rodrigues-Martins A, Debski J, Dzhindzhev N, Bettencourt-Dias M. Regulation of autophosphorylation controls PLK4 self-destruction and centriole number. Curr Biol 2013, 23 (22): 2245-2254

Guerreiro A, da Costa G, Gomes RA, Ribeiro-Silva C, Gilberto S, Mateus E, Monteiro E, Barroso E, Coelho AV, Ponces Freire A, Cordeiro C. α-Synuclein aggregation in the saliva of familial transthyretin amyloidosis: a potential biomarker. Amyloid 2012, 19 (2): 74-80

Gilberto S, Borrego-Pinto J, Bettencourt-Dias M. Centrosomes and cilia in health and disease: mechanisms of biogenesis and function. CanalBQ 2012, 9: 28-38 [review]

Ribeiro-Silva C, Gilberto S, Mateus E, Monteiro E, Barroso E, da Costa G, Ponces Freire A, Cordeiro C. The relative amounts of plasma transthyretin forms in familial transthyretin amyloidosis: a quantitative analysis by Fourier transform ion-cyclotron resonance mass spectrometry. Amyloid 2011, 18 (4): 191-199

EXTRA ACADEMIC ACTIVITIES

2017 Member of the organizing committee of the 2nd retreat of the Institute of Biochemistry, ETH Zurich, Adelboden, Switzerland

2015-2017 Board member of the departmental academic staff association (AMB) at ETH Zurich • Organization of social events • Representative of the academic staff of the Institute of Biochemistry

2015 Member of the organizing committee of the 1st retreat of the Institute of Biochemistry, ETH Zurich, Adelboden, Switzerland

186

9. Curriculum vitae

2014-2016 PhD student representative in the Institute of Biochemistry, ETH Zurich, Switzerland • Member of the board of the Institute of Biochemistry

2014 Member of the steering committee of the Molecular Life Sciences PhD program, Zurich, Switzerland • Decision-making privilege • Organization of social events for respective students

2014 Member of the Retreat Organizing Committee of the Molecular Life Sciences PhD program, Zurich, Switzerland

2006-2009 Student representative and member of the Biochemistry bachelor student committee, FCUL, Lisbon

OTHER RELEVANT INFORMATION

IT skills Office software knowledge (Word, Excel, Powerpoint) Adobe Photoshop CC and Adobe Illustrator CC Scientific software (e.g. ImageJ, CLC Genomics Workbench) Python v2 programing language ImageJ macro programming language

Languages Portuguese – Mother tongue English – Fluent German – Basic French – Basic

Personal interests Reading (Historical fiction, Sci-fi), Computer and console gaming, Traveling (urban and nature sight-seeing)

187