Analysing polyubiquitin chain signalling by 
means of tailor-made ubiquitin ligases 
 
 
Dissertation 
Zur Erlangung des Grades 
“Doktor der Naturwissenschaften” 
(Dr. rer. nat.) 
 
am Fachbereich Biologie 
der Johannes Gutenberg-Universität Mainz 
 
 
Kirill Petriukov 
geboren am 30.05.1995 in Sankt Petersburg 
 
 
Juli 2023  
  
2 
 
Dekan:  
1st examiner:  
2nd examiner:  
Date: 11.01.2024  
3 
 
  
4 
 
Declaration 
 
 
 
I, Kirill Petriukov, confirm that the work presented in this thesis is my own. Where 
information has been derived from other sources, I confirm that it has been indicated in 
the thesis.  
  
5 
 
 
 
  
6 
 
Abstract 
Ubiquitylation is a widespread posttranslational modification involved in all major 
cellular pathways. Ubiquitin can be attached to target proteins as a single moiety or form 
conjugates of higher complexity, creating unique interaction surfaces on the substrates, 
called the ‘ubiquitin code’. DNA sliding clamp PCNA is one of the substrates for which 
mono- and polyubiquitylation have been shown to perform different functions during 
replication over damaged DNA. Monoubiquitylation of PCNA recruits translesion synthesis 
polymerases, whereas modification of PCNA with a K63-linked polyubiquitin chain 
activates an error-free damage bypass pathway in a mechanistically poorly characterised 
manner.  
To explore the roles of PCNA polyubiquitylation in human cells, tailor-made 
ubiquitin ligases were designed and applied to create ubiquitin chains on PCNA 
independently of the responsible cellular enzymes. These tailor-made enzymes harbour 
a PCNA-interaction motif combined with linkage-specific ubiquitin ligase domains, 
designed to create M1-, K63- and K48-linked ubiquitin chains. By applying tailor-made 
ligases in human cells, I demonstrate that unscheduled modification of PCNA with K63-
linked polyubiquitin chains leads to global replication collapse. Mass spectrometry-based 
analysis identifies the ubiquitin-dependent segregase VCP as the responsible effector. By 
performing a series of independent in vitro and in vivo assays, I demonstrate that K63-
linked chains on PCNA are further modified by the E2 enzyme UBE2K with K48 linkages, 
leading to the recruitment of VCP, PCNA degradation and checkpoint activation. I further 
demonstrate that the build-up of branched K63-K48 chains is not an artefact of tailor-made 
enzymes but is a general cellular pathway, which takes place upon replication over 
damaged DNA and underlies the synthetic lethality between the tumour suppressor 
BRCA1 and the deubiquitylating enzyme USP1. 
Finally, I describe the ‘Ubiquiton’ technology recently designed in our laboratory. It 
allows inducible polyubiquitylation of any protein of interest with M1-, K63- and K48-linked 
chains. By applying this technology to human histone H2B, I demonstrate the feasibility of 
this approach in human cells. Altogether, this work focuses on two protein engineering 
technologies, which allowed to explore the functions of PCNA polyubiquitylation and 
provides the basis for a study of ubiquitin signalling on other substrates. 
7 
 
  
8 
 
Zusammenfassung 
Die Ubiquitinierung ist eine weit verbreitete posttranslationale Modifikation, die an 
allen wichtigen zellulären Signalwegen beteiligt ist. Ubiquitin kann als einzelnes Molekül 
an das Zielprotein angehängt werden oder Konjugate in höherer Komplexität bilden, die 
als 'Ubiquitin-Code' bezeichnet werden. Einige Substrate wie beispielsweise das DNA-
Ringklemmenprotein PCNA üben je nach Ubiquitinierungsstatus unterschiedliche 
Funktionen aus. Während der Replikation beschädigter DNA dient die 
Monoubiquitinierung von PCNA als Rekrutierungssignal für Transläsionspolymerasen, 
während die Modifikation von PCNA mit einer K63-verknüpften Polyubiquitinkette, auf 
noch mechanistisch unzureichend charakterisierte Weise, den fehlerfreien ‘Template 
Switching’ Signalweg aktiviert.  
Um die Rollen der PCNA-Polyubiquitinierung in menschlichen Zellen zu 
erforschen, wurden maßgeschneiderte Ubiquitin-Ligasen entworfen und angewendet, 
welche unabhängig von den verantwortlichen zellulären Enzymen Ubiquitinketten auf 
PCNA synthetisieren. Diese maßgeschneiderten Enzyme enthalten ein PCNA-
Interaktionsmotiv kombiniert mit verknüpfungsspezifischen Ubiquitin-Ligase-Domänen, 
die so konzipiert sind, dass sie M1-, K63- und K48-verknüpfte Ubiquitinketten erzeugen. 
In der vorliegenden Arbeit werden die maßgeschneiderten Ubiquitin-Ligasen verwendet 
um die Rollen der PCNA-Polyubiquitinierung in menschlichen Zellen zu untersuchen. Ich 
weise nach, dass eine außerplanmäßige Modifikation von PCNA mit K63-verknüpften 
Polyubiquitinketten zum Zusammenbruch der Replikation führt. Eine 
Massenspektrometrie-basierte Analyse identifizierte die Ubiquitin-Segregase VCP als 
Interaktor von polyubiquitiniertem PCNA in vivo. Durch eine Reihe von unabhängigen in 
vitro und in vivo Assays zeige ich, dass K63-verknüpfte Ketten auf PCNA durch das E2-
Enzym UBE2K mit K48-Verknüpfungen weiter modifiziert werden. Dies resultiert in der 
Rekrutierung von VCP, dem Abbau von PCNA und der Aktivierung des Checkpoints. 
Darüber hinaus zeige ich weiter, dass der Aufbau verzweigter K63-K48-Ketten kein 
Artefakt der maßgeschneiderten Enzyme ist, sondern vielmehr ein allgemeiner zellulärer 
Prozess, der bei der Replikation geschädigter DNA abläuft und der der synthetischen 
Letalität zwischen dem Tumorsuppressor BRCA1 und der Deubiquitinase USP1 zugrunde 
liegt. 
9 
 
Abschließend beschreibe ich die kürzlich in unserem Labor entwickelte 
Technologie namens 'Ubiquiton', die eine induzierbare Polyubiquitinierung jedes 
beliebigen Proteins mit M1-, K63- oder K48-verknüpften Ketten ermöglicht. Indem ich 
diese Technologie bei dem menschlichen Histon H2B einsetze, demonstriere ich die 
Anwendbarkeit dieses Systems in menschlichen Zellen. Insgesamt konzentriert sich diese 
Arbeit auf zwei Technologien des Protein-Engineerings, die es ermöglichen, die 
Funktionen der PCNA-Polyubiquitinierung zu erforschen und die zusätzlich die Grundlage 
für die Untersuchung der Ubiquitin-Signale anderer Substrate schaffen.  
10 
 
Acknowledgements 
 
  
  
11 
 
Contents 
Declaration .................................................................................................................................... 5 
Abstract ......................................................................................................................................... 7 
Zusammenfassung ........................................................................................................................ 9 
List of figures ............................................................................................................................... 17 
List of tables ................................................................................................................................ 19 
Chapter 1: Introduction ................................................................................................................ 20 
1.1 The Ubiquitin system ......................................................................................................... 20 
1.1.1 Ubiquitin as a posttranslational modifier ..................................................................... 20 
1.1.2 Enzymatic core of the ubiquitin system ....................................................................... 22 
1.1.3 Ubiquitin chains ........................................................................................................... 24 
1.1.3.1 K48-linked ubiquitin chains ................................................................................... 24 
1.1.3.2 K63-linked ubiquitin chains ................................................................................... 24 
1.1.3.3 M1-linked (linear) ubiquitin chains ........................................................................ 25 
1.1.3.4 K11-linked ubiquitin chains ................................................................................... 26 
1.1.3.5 K27-linked ubiquitin chains ................................................................................... 27 
1.1.3.6 K29-linked ubiquitin chains ................................................................................... 27 
1.1.3.7 K6-linked ubiquitin chains ..................................................................................... 28 
1.1.3.8 K33-linked ubiquitin chains ................................................................................... 29 
1.1.4 The ubiquitin-proteasome system ............................................................................... 29 
1.1.5 Deubiquitylating enzymes ........................................................................................... 31 
1.1.6 Tools to study the ubiquitin system ............................................................................. 32 
1.1.6.1 Mass spectrometry ............................................................................................... 32 
1.1.6.2 High-affinity binders to polyubiquitin chains ......................................................... 33 
1.1.6.3 Protein engineering approaches ........................................................................... 34 
1.2 Genome stability ................................................................................................................ 36 
1.2.1 Types of DNA damage ................................................................................................ 36 
1.2.1.1 Base oxidation ...................................................................................................... 36 
1.2.1.2 Base alkylation ..................................................................................................... 36 
1.2.1.3 Bulky adducts ....................................................................................................... 37 
1.2.1.4 UV photoproducts ................................................................................................. 37 
1.2.1.5 Interstrand crosslinks ............................................................................................ 38 
1.2.1.6 Abasic sites .......................................................................................................... 39 
1.2.1.7 DNA-protein crosslinks ......................................................................................... 40 
12 
 
1.2.1.8 Single-strand breaks ............................................................................................. 40 
1.2.1.9 Double-strand breaks ........................................................................................... 41 
1.2.2 Replication stress ........................................................................................................ 41 
1.2.3 DNA damage bypass .................................................................................................. 44 
1.2.3.1 Translesion synthesis ........................................................................................... 44 
1.2.3.2 Template switching ............................................................................................... 45 
1.2.3.3 Fork reversal ......................................................................................................... 47 
1.2.3.4 Repriming ............................................................................................................. 49 
1.2.3.5 Roles of ubiquitylated PCNA in DNA damage bypass ......................................... 51 
1.3 Aims of this work ................................................................................................................ 55 
Chapter 2: Materials and methods .............................................................................................. 56 
2.1 Reagents ........................................................................................................................... 56 
2.1.1 Chemicals and recombinant proteins .......................................................................... 56 
2.1.2 Antibodies ................................................................................................................... 57 
2.2 Media and solutions ........................................................................................................... 58 
2.2.1 Media for bacteria ....................................................................................................... 58 
2.2.2 Media for mammalian cells ......................................................................................... 58 
2.2.1 Solutions ..................................................................................................................... 59 
2.3 DNA oligonucleotides ........................................................................................................ 60 
2.4 RNA oligonucleotides ........................................................................................................ 62 
2.5 Plasmids ............................................................................................................................ 63 
2.6 Strains and cell lines .......................................................................................................... 68 
2.6.1 Bacterial strains ........................................................................................................... 68 
2.6.2 Mammalian cell lines ................................................................................................... 68 
2.7. General methods for DNA manipulation ........................................................................... 69 
2.7.1 Measurement of DNA concentration ........................................................................... 69 
2.7.2 Agarose gel electrophoresis ........................................................................................ 69 
2.8 Methods for molecular cloning ........................................................................................... 69 
2.8.1 Polymerase chain reaction (PCR) ............................................................................... 69 
2.8.2 Site-directed mutagenesis ........................................................................................... 70 
2.8.3 Gibson assembly ......................................................................................................... 70 
2.8.4 Gateway cloning .......................................................................................................... 70 
2.8.5 Restriction cloning ....................................................................................................... 71 
2.8.6 DNA sequencing ......................................................................................................... 71 
2.9 Methods for protein manipulation ...................................................................................... 71 
13 
 
2.9.1 SDS polyacrylamide gel electrophoresis ..................................................................... 71 
2.9.2 Western blotting .......................................................................................................... 71 
2.10 Methods for E. coli ........................................................................................................... 72 
2.10.1 Cultivation of E. coli ................................................................................................... 72 
2.10.2 Transformation of chemically competent E. coli cells ............................................... 72 
2.10.3 Isolation of plasmid DNA ........................................................................................... 73 
2.11 Methods for mammalian cells .......................................................................................... 73 
2.11.1 Cell thawing and freezing .......................................................................................... 73 
2.11.2 Cell passaging ........................................................................................................... 74 
2.11.3 Cell harvesting .......................................................................................................... 74 
2.11.4 Cell counting ............................................................................................................. 74 
2.11.5 Transient transfection of DNA ................................................................................... 74 
2.11.5.1 Cell transfection with polyethyleneimine ............................................................. 74 
2.11.5.2 Cell transfection with Fugene HD ....................................................................... 74 
2.11.6 Transient transfection of siRNA ................................................................................ 75 
2.11.7 Creation of stable cell lines via Flp-In integration ...................................................... 75 
2.11.8 Lentiviral transduction ............................................................................................... 75 
2.11.9 DNA fibre assay ........................................................................................................ 76 
2.11.10 Preparation of cell lysates ....................................................................................... 77 
2.11.11 Cell cycle analysis by flow cytometry ...................................................................... 77 
2.11.12 Detection of single-stranded DNA by immunofluorescence .................................... 78 
2.11.13 SILAC labelling of cells ........................................................................................... 78 
2.11.14 Cell viability assay ................................................................................................... 78 
2.12 Detection of protein ubiquitylation .................................................................................... 79 
2.12.1 Detection of PCNA ubiquitylation by PCNA immunoprecipitation ............................. 79 
2.12.2 Detection of protein ubiquitylation by denaturing Ni-NTA pulldown .......................... 80 
2.12.3 Ubiquitin chain restriction assay (UBICREST) .......................................................... 80 
2.13 In vitro ubiquitylation assay .............................................................................................. 81 
Chapter 3: Results ....................................................................................................................... 82 
3.1 Basis of the study. Design of PCNA-selective linkage-specific ubiquitin ligases ............... 82 
3.2 Optimisation of PIP-E3s in mammalian cells ..................................................................... 87 
3.2.1 PIP-E363 ...................................................................................................................... 87 
3.2.2 PIP-E348 ...................................................................................................................... 88 
3.2.3 PIP-E31........................................................................................................................ 89 
3.3 Excessive K63-linked PCNA polyubiquitylation results in replication catastrophe............. 91 
14 
 
3.3.1 Creation of a cell line with inducible expression of PIP-E363 ....................................... 91 
3.3.2 PIP-E363-expressing cells activate checkpoint signalling and exhibit S phase arrest . 94 
3.3.3 ATR inhibition exacerbates effects of PIP-E363 ........................................................... 97 
3.3.4 Effects of PIP-E363 expression are dependent on its interaction with PCNA and 
catalytic activity .................................................................................................................... 99 
3.3.5 PCNA monoubiquitylation is a prerequisite for PIP-E363 activity in vivo .................... 101 
3.3.6 Negative effects of PIP-E363 are independent of known interactors of   
polyubiquitylated PCNA ..................................................................................................... 105 
3.4 Analysing the interactome of polyubiquitylated PCNA ..................................................... 107 
3.5 K63-linked chains on PCNA are converted into conjugates of higher complexity in vivo 111 
3.5.1 VCP activity underlies PIP-E363-induced replication collapse ................................... 111 
3.5.2 Analysing the interactome of polyubiquitylated PCNA upon VCP inhibition ............. 114 
3.5.3 K63-linked chains on PCNA are converted into branched K63-K48 conjugates ....... 116 
3.5.3.1 UBICREST assay ............................................................................................... 116 
3.5.3.2 Targeted deubiquitylation of PCNA in vivo recapitulates the UBICREST assay 121 
3.5.3.3 K63-linked chains on PCNA are branched via K48 and K11 .............................. 124 
3.5.4 UBE2K is involved in the build-up of branched chains on PCNA .............................. 127 
3.6 Analysing the significance of polyubiquitin linkage type on PCNA in triggering      
replication collapse ................................................................................................................ 131 
3.6.1 Establishment of a cell line with inducible K48-linked polyubiquitylation of PCNA ... 132 
3.6.2 Direct comparison of K63- versus K48-linked PCNA polyubiquitylation ................... 134 
3.7 Analysing the physiological relevance of ubiquitin chain branching on PCNA ................ 138 
3.7.1 Introduction to the synthetic lethal relationship between BRCA1 and USP1 ............ 138 
3.7.2 VCP mediates PCNA degradation upon inhibition of USP1 in BRCA1-deficient       
cells .................................................................................................................................... 139 
3.7.3 Ubiquitin chain branching underlies the synthetic lethality between BRCA1 and    
USP1 .................................................................................................................................. 142 
3.7.4 Removal of PCNA impacts the damage bypass pathway in BRCA1-deficient cells . 144 
3.8 Expanding the technology beyond PCNA: the “Ubiquiton” system .................................. 147 
3.8.1 Design of the “Ubiquiton” system .............................................................................. 147 
3.8.2 Application of the Ubiquiton system to human histone H2B ..................................... 148 
Chapter 4: Discussion ............................................................................................................... 153 
4.1 Tailor made E3s as a new engineering approach to study the ubiquitin code. ................ 153 
4.2 Exploring the roles of PCNA polyubiquitylation by means of PIP-E3s ............................. 155 
4.2.1 Which biological questions are addressed by the PIP-E3 system in yeast and     
human cells? ...................................................................................................................... 155 
4.2.3 Effects of PIP-E363 expression are on-target ............................................................ 157 
15 
 
4.2.5 PCNA is modified with branched ubiquitin chains ..................................................... 159 
4.2.6 Ubiquitin chain branching underlies synthetic lethality between BRCA1 and USP1 . 162 
4.2.7 USP1 slows down replication under stressed conditions in BRCA1-deficient cells .. 163 
4.2.8 Potential clinical significance of the findings ............................................................. 164 
4.3 Future perspectives ......................................................................................................... 165 
4.3.1 PCNA polyubiquitylation ............................................................................................ 165 
4.3.2 The Ubiquiton technology ......................................................................................... 166 
Chapter 5: Appendix .................................................................................................................. 168 
5.1 Abbreviations ................................................................................................................... 168 
5.2 Publications ..................................................................................................................... 171 
5.3 Curriculum Vitae .............................................................................................................. 172 
Bibliography ............................................................................................................................... 173 
 
 
  
16 
 
List of figures 
 
Figure 1: Ubiquitin as a posttranslational modifier. ..................................................................... 22 
Figure 2: Enzymatic machinery of the ubiquitin system. ............................................................. 23 
Figure 3: Structure of 26S proteasome and VCP. ....................................................................... 31 
Figure 4: Different types of DNA lesions. .................................................................................... 38 
Figure 5. Structure of an abasic site and the mechanism of its conversion into a DNA-protein 
crosslink ...................................................................................................................................... 39 
Figure 6: Schematic representation of the template switching pathway ..................................... 46 
Figure 7: Fork reversal. ............................................................................................................... 48 
Figure 8: Nascent DNA degradation ........................................................................................... 49 
Figure 9: Daughter-strand gaps .................................................................................................. 51 
Figure 10: Roles of PCNA ubiquitylation in DNA damage tolerance ........................................... 54 
Figure 11: Design and implementation of tailor-made ubiquitin ligases ...................................... 85 
Figure 12: PIP-E363 is active in vitro and in human cells. ............................................................ 87 
Figure 13: Optimisation of PIP-E348 system in mammalian cells. ............................................... 89 
Figure 14: Optimisation of PIP-E31 system in mammalian cells. ................................................. 90 
Figure 15: Creation and characterization of an RPE1 hTERT PIP-E363 cell line......................... 93 
Figure 16: PCNA polyubiquitylation induces checkpoint activation and cell cycle arrest. ........... 95 
Figure 17: Lower EdU incorporation and replication speed in PIP-E363-expressing cells ........... 97 
Figure 18: ATR prevents DSB formation upon excessive PCNA polyubiquitylation. .................. 98 
Figure 19: Negative effects of PIP-E363 require interaction with PCNA and catalytic activity.. . 100 
Figure 20: The negative effects of PIP-E363 are dependent on lysine 164 of PCNA. ................ 103 
Figure 21: The negative effects of PIP-E363 are dependent of RAD18 activity ......................... 104 
Figure 22: The negative effects of PIP-E363 are independent of fork reversal enzymes,        
RAP80 and WRNIP1. ................................................................................................................ 106 
Figure 23: Immunoprecipitation of chromatin-bound PCNA preserves its interactions with a  
panel of known PCNA binders. ................................................................................................. 107 
Figure 24: Analysing the interactome of polyubiquitylated PCNA ............................................. 109 
Figure 25: VCP activity mediates the toxic effects of PIP-E363 expression ............................... 112 
Figure 26: VCP extracts polyubiquitylated PCNA from chromatin under replication stress 
conditions. ................................................................................................................................. 114 
Figure 27: Interactome of polyubiquitylated PCNA upon VCP inhibition. .................................. 115 
17 
 
Figure 28: Scenarios explaining accumulation of K48 and K11 linkages in PCNA 
immunoprecipitates. .................................................................................................................. 116 
Figure 29: The UBICREST assay reveals ubiquitin chain branching on PCNA. ....................... 119 
Figure 30: K63-linked chains on PCNA are branched with K48 linkages .................................. 120 
Figure 31: Replication stress induces formation of K63-K48 branched chains on PCNA……  .121 
Figure 32: Targeting of OTUB1 to PCNA interferes with chain branching in vivo ..................... 123 
Figure 33: K63-linked chains on PCNA are branched via K48 and K11.. ................................. 126 
Figure 34: UBE2K forms K63-K48 branched ubiquitin chains in the PIP-E363 system .............. 129 
Figure 35: Effects of UBE2K on PCNA polyubiquitylation in vivo .............................................. 130 
Figure 36: UBE2K affects PCNA polyubiquitylation both in the PIP-E363 system and during 
endogenous damage response. ................................................................................................ 131 
Figure 37: Design and creation of a cell line with inducible K48-linked polyubiquitylation of 
PCNA.. ...................................................................................................................................... 133 
Figure 38: Self-cleavable and uncleavable fusions of PIP-E348 to Ubc7 display similar activity 
towards PCNA in vivo. ............................................................................................................... 135 
Figure 39: Comparison of K63- and K48-linked PCNA polyubiquitylation. ................................ 136 
Figure 40: UBE2K mediates checkpoint activation after K63-, but not K48-linked PCNA 
polyubiquitylation. ...................................................................................................................... 137 
Figure 41: HU- and USP1i-dependent PCNA degradation in BRCA1-deficient cells.. .............. 141 
Figure 42: Toxic effects of USP1 inhibition in BRCA1-deficient cells is dependent on RAD18, 
UBE2K and RFWD3. ................................................................................................................. 143 
Figure 43: USP1 inhibition in MDA-MD-436 cells affects replication upon replication stress. ... 145 
Figure 44: Design of the Ubiquiton system.. ............................................................................. 150 
Figure 45: K63- and K48-Ubiquiton applied to human H2B ...................................................... 151 
Figure 46: M1-Ubiquiton applied to human H2B ....................................................................... 152 
Figure 47: Schematic representation of the pathway, initiated by PCNA polyubiquitylation in     
the PIP-E363 and PIP-E348 systems .......................................................................................... 161 
Figure 48: A model explaining rapid PCNA degradation upon USP1 inhibition in BRCA1- 
deficient compared to wild-type cells ......................................................................................... 163 
 
  
18 
 
List of tables 
Table 1: List of primary antibodies used in this study. ........................................................................ 57 
Table 2: A list of secondary antibodies used in this thesis ................................................................. 58 
Table 3: A list of DNA oligonucleotides used in this thesis. ............................................................... 60 
Table 4: A list of RNA oligonucleotides used in this thesis ................................................................ 62 
Table 5: A list of plasmids created in this study. .................................................................................. 63 
Table 6: A list of plasmids generated by others and used in this study. .......................................... 67 
Table 7: A list of the E. coli strains used in this study. ........................................................................ 68 
Table 8: A list of the mammalian cell lines used in this study............................................................ 68 
 
  
19 
 
Chapter 1 
 
Introduction 
 
1.1 The Ubiquitin system 
1.1.1 Ubiquitin as a posttranslational modifier 
 Protein-protein interactions form the basis of life. Whereas some proteins tend to 
interact with each other constitutively, many protein-protein interactions emerge or 
disappear as a response to certain stimuli. Regulation of protein interactions often occurs 
at a posttranslational level through protein modification, as it does not require time-
consuming de novo protein synthesis. Modification of a protein can change its 
interactome, stability or subcellular localisation and may initiate or terminate cellular 
pathways. More than 400 posttranslational modifications (PTMs) are known, affecting side 
chains of amino acids as well as N- and C-termini of proteins (Khoury et al., 2011). Many 
PTMs are small modifications of amino acids: typical examples are phosphorylation, 
methylation or acetylation. The addition of a modification changes the charge distribution 
and spatial characteristics of a given amino acid, thereby affecting corresponding 
interactions. Importantly, due to their size, small modifications are recognised in the 
context of surrounding amino acids: the precise position of a small modification in a protein 
may drastically change its function. A canonical example is the modification of histone 
tails: trimethylation of lysines K4 and K9 of histone H3, which are separated by just four 
amino acids, leads to opposite effects on gene expression (Lu et al., 2018; Bannister and 
Kouzarides, 2011).  
 An 8.6 kDa protein Ubiquitin, as well as a family of structurally related proteins, 
termed ubiquitin-like modifiers (UBLs), can also serve as a posttranslational modification. 
Attachment of ubiquitin to target proteins occurs via an isopeptide bond between the C-
terminal glycine of ubiquitin and the ε-amino group of the substrate lysine. Although lysine 
is the most common ubiquitylation target, other amino acids such as serine, cysteine, 
threonine or N-terminal methionine have also been shown to be ubiquitylated (McClellan 
et al., 2019; Ciechanover, 2005; Ben-Saadon et al., 2004). In striking contrast to small 
20 
 
modifications, ubiquitin is several magnitudes larger and, for small substrates, can be 
comparable in size to its target proteins. Not surprisingly, the attachment of ubiquitin may 
result in a conformational change of a region surrounding the attachment site, as 
suggested by in silico simulations for Ubc7 and p19 (Hagai and Levy, 2010). Due to its 
size, recognition of ubiquitin usually occurs by means of dedicated ubiquitin-binding 
domains (Husnjak and Dikic, 2012), allowing for certain flexibility in terms of an attachment 
site (discussed later in section 1.1.6.3). The importance of ubiquitin is highlighted by its 
extremely high conservation: just three amino acids out of 76 differ between yeast and 
human ubiquitin (Zuin et al., 2014). Being present at approximately 108 copies per cell, 
ubiquitin is one of the most abundant proteins, amounting to up to 5% of all proteins in a 
cell (Yewdell 2001; Haas and Bright, 1985). The structure of ubiquitin involves a β-sheet, 
an α-helix and a short 310 helix. It is compact except for the highly flexible C-terminal 
region, which protrudes from the globular fold (Figure 1A) (Vijay-Kumar et al., 1987). A 
characteristic feature of ubiquitin is the presence of two water-exposed hydrophobic 
patches: the I44 patch (L8-I44-V70) and the I36 patch (I36-L71-L73). Most known 
ubiquitin-binding domains interact with hydrophobic patches (Grabbe and Dikic, 2009), 
although for several types of ubiquitin-binding domains the non-hydrophobic D58 patch 
(R54-T55-S57-D58) and F4 patch (Q2-F4-T14) have the highest impact on the interaction 
(Figure 1B) (Lee et al., 2006; Sato et al., 2011; Wright et al., 2016).  
 There are multiple ways in which ubiquitin can modify a target protein. The first one 
is monoubiquitylation – attachment of a single ubiquitin moiety. This type of modification 
has been observed for PCNA, histones, RAD18, CENP-A and other substrates and 
typically plays signalling non-degradative roles (Hicke 2001; Hoege et al., 2002; Stelter 
and Ulrich, 2003; Zeman et al., 2014; Niikura et al., 2019). A substrate can also carry 
several monoubiquitin moieties (multi-monoubiquitylation) (Lai et al., 2001; Haglund et al., 
2003). Importantly, ubiquitin itself can be a target for ubiquitylation, with seven internal 
lysines (K6, K11, K27, K29, K33, K48, K63) or the N-terminal methionine of ubiquitin 
acting as attachment sites (Figure 1C). Therefore, 8 linkage types between two ubiquitin 
moieties are possible, all detected in mammalian cells (discussed in detail in section 
1.1.3). Furthermore, multiple lysine residues of a single ubiquitin can be ubiquitylated, 
giving rise to branched ubiquitin chains. Given that ubiquitin itself can be 
posttranslationally modified and considering the existence of non-lysine ubiquitylation, 
21 
 
ubiquitin can form a great variety of structures with different topologies. This phenomenon 
has been called the ‘Ubiquitin code’ (Figure 1D) (Komander and Rape, 2012).  
 
Figure 1: Ubiquitin as a posttranslational modifier. (A) Crystal structure of ubiquitin in cartoon 
representation (PDB: 1UBQ). Highlighted are the following secondary structure elements: β-sheet 
(pink), α-helix (light blue) and 310 helix (dark blue). (B) Crystal structure of ubiquitin as in (A) with 
highlighted interaction patches: I44 patch (yellow), I36 patch (orange), D58 patch (light blue) and 
an F4 patch (pink). (C) Crystal structure of ubiquitin as in (A) with highlighted N-terminal 
methionine and lysine residues. (D) Different modes of protein ubiquitylation. 
1.1.2 Enzymatic core of the ubiquitin system 
Attachment of ubiquitin to the target proteins takes place via a cascade of 
enzymatic reactions involving ubiquitin-activating enzymes (E1), ubiquitin-conjugating 
enzymes (E2) and ubiquitin ligases (E3), which mediate the final step and are responsible 
for the choice of a substrate (Stewart et al., 2016; Zheng and Shabek, 2017). Human cells 
22 
 
possess 2 E1, at least 38 E2 and at least 600 E3 enzymes. E1 enzymes couple the C-
terminus of ubiquitin to their active site via a thioester bond, followed by a transfer of 
ubiquitin to the catalytic site of E2 enzymes. The next step in the ubiquitylation cascade 
highly depends on the type of E3 enzymes. There are three different classes of ubiquitin 
ligases: RING/U-box, HECT and RING-between-RING (RBR). Ligases of the RING-type 
mediate interactions between E2 enzymes and substrates: ubiquitin is directly transferred 
from the E2 enzyme onto the substrate. As the E3 enzyme does not directly participate in 
the ubiquitin transfer, the linkage type in the case of RING-type E3s is determined by the 
E2 enzyme. On the contrary, HECT-type E3s first load ubiquitin on their catalytic cysteine 
and then transfer it onto the substrate. In this case, the linkage type is dependent on the 
E3 enzyme. RBR-type E3 enzymes can be considered RING-HECT hybrids, as they 
possess two RING domains, one interacting with the ubiquitin-charged E2 and the other 
accepting ubiquitin from the E2. As for HECT-type E3s, the linkage type in the case of 
RBR enzymes is determined by the E3 enzyme itself (Figure 2).  
 
 
 
Figure 2: Enzymatic machinery of the ubiquitin system. Schematic representation of the 
ubiquitylation cascade starting from the free ubiquitin and resulting in a ubiquitylated substrate. 
23 
 
1.1.3 Ubiquitin chains 
As introduced in section 1.1.1, two ubiquitins in a chain can be connected with at 
least 8 linkage types. Whereas certain linkage types have been shown to fulfil specific 
functions in a cell, for many chain types the significance of the linkage remains 
unexplored. Here I summarise the known functions of ubiquitin chains, starting with the 
best and finishing with the least well-characterised ones, as well as describe functions of 
the known ubiquitin structures of higher complexity. 
1.1.3.1 K48-linked ubiquitin chains 
 K48 linkages are the most abundant linkage type in a cell, representing a canonical 
signal for protein degradation by the proteasome (section 1.1.4) (Pickart, 1997). This 
pathway is indispensable for cell viability, as expression of the K48R mutant of ubiquitin 
cannot compensate for the loss of ubiquitin genes (Finley et al., 1994). Whereas a 
minimum chain length of 4 ubiquitin units has been initially suggested to be a signal for 
proteasomal degradation, later findings suggest that high local ubiquitin concentration is 
more important than chain length: two di-ubiquitin K48-linked chains on a cyclin B1 
represent a stronger degradation signal than a single tetra-ubiquitin chain of the same 
linkage type (Thrower et al., 2000; Lu et al., 2015). Despite their predominant role in 
protein degradation, several reports indicate that K48-linked chains may fulfil regulatory 
non-degradative functions. An example is the transcriptional factor Met4, which binds 
K48-linked chains with high specificity. This interaction competes with the binding of the 
transcription machinery to the same region, thereby repressing the transcriptional activity 
of Met4 (Villamil et al., 2022; Li et al., 2019). Another example is the translesion 
polymerase kappa, which has been shown to associate with K48-linked ubiquitin chains 
on PCNA upon conditions of nucleotide starvation. However, the relevance of chain 
linkage on PCNA as well as its functions need to be clarified (Tonzi et al., 2018). 
1.1.3.2 K63-linked ubiquitin chains 
 K63-linked polyubiquitylation performs a variety of predominantly non-degradative 
functions in a cell. Assembly of these chains by RING-type ubiquitin ligases often depends 
on the E2 enzyme UBC13. It possesses unique linkage specificity due to its obligatory 
interaction with the catalytically inactive E2 variant MMS2. In yeast, Ubc13 functions with 
24 
 
ubiquitin ligases Rad5 and Pib1, whereas its human homologue has a broader spectrum 
of cognate E3s, including HLTF, SHPRH, TRAF6, RNF8, CHFR and CHIP (Hodge et al., 
2016). 
Modification of the DNA sliding clamp PCNA with K63-linked chains is 
indispensable for error-free damage bypass and fork reversal in yeast and human cells, 
respectively (section 1.2.3.5), and modification of histones with this chain type is required 
for the proper DNA damage response during repair of double-strand breaks (Schwertman 
et al., 2016; Stelter and Ulrich, 2003; Vujanovic et al., 2017). In the latter case, the 
responsible reader is RAP80, which brings the BRCA1A complex to the site of damage. 
The binding of K63-linked chains by RAP80 occurs via its highly linkage-selective tandem 
ubiquitin-binding motifs, which served as a basis for the design of K63-linked polyubiquitin 
sensor probes (Sims and Cohen, 2009; Sims et al., 2012). K63-linked chains, assembled 
by ubiquitin ligases TRAF6, cIAP1 and cIAP2, are also crucial for inflammation and 
immune responses via the recruitment of TAB2 and promoting NF-κB activation (Deng et 
al., 2000; Wajant and Scheurich, 2011). Functions of K63-linked ubiquitin chains in 
membrane trafficking are demonstrated by the strict requirement of this chain type for the 
endocytosis of yeast uracil permease Fur4; a similar pathway has also been observed for 
mammalian proteins, such as epidermal growth factor receptor EGFR or dopamine 
transporter DAT (Galan and Haguenauer-Tsapis, 1997; Vina-Vilaseca and Sorkin, 2010). 
 K63-linked chains have also been shown to be modified with K48 linkages, 
resulting in the formation of branched ubiquitin chains, which can be assembled by the 
ubiquitin ligases UBE2K, HUWE1 and UBR5 and fulfil both degradative and non-
degradative functions in a cell (Ohtake et al., 2018; Ohtake et al., 2016; Pluska et al., 
2021). A recent proteomics-based study identifies a number of proteins, which selectively 
bind to branched K63-K48 chains, but not to homotypic K63- or K48-linked ones (Lange 
et al., 2023). 
1.1.3.3 M1-linked (linear) ubiquitin chains 
In S. cerevisiae, M1 linkages between ubiquitins are not produced enzymatically 
and are only present in the product of the UBI4 gene, which is a head-to-tail fusion of 5 
ubiquitins. In higher eukaryotes, M1-linked chains are produced by the LUBAC complex, 
which is, up until now, the only E3 ligase complex known to synthesise linear ubiquitin 
25 
 
chains. The LUBAC complex consists of three components: ubiquitin ligases HOIL1 and 
HOIP, as well as the structural component SHARPIN. HOIL1 mainly participates in the 
transfer of the first ubiquitin to the substrate (lysine, serine or threonine residues), whereas 
HOIP extends ubiquitins on the substrate into an M1-linked chain (Kelsall et al., 2019; 
Smit et al., 2013). Several substrates for linear ubiquitylation (RIPK1/2, TRADD, TNFR1, 
etc.), as well as readers of linear ubiquitin chains (ABIN1-3, Optineurin, A20, etc.), have 
been reported (Fennell et al., 2018; Dittmar and Winklhofer, 2019). Interestingly, NF-κB 
essential modulator NEMO is both a selective reader of linear ubiquitin chains and a 
substrate of linear polyubiquitylation. Functions of linear ubiquitin chains mainly involve 
inflammation and immune responses. However, the role of linear polyubiquitylation in the 
ubiquitin fusion degradation pathway in human cells has also been reported (Tokunaga 
and Iwai, 2012; Kirisako et al., 2006). In addition to homotypic linear chains, mixed M1/K63 
chains have been implicated in the IKK activation pathway (Emmerich et al., 2013). M1-
linked chains were also found on the TRAF6 proteins in combination with K48 and K63 
linkages, suggesting the presence of a complex ubiquitin code on this substrate (Ohtake 
et al., 2016). 
1.1.3.4 K11-linked ubiquitin chains 
K11-linked chains have been most extensively characterised in the context of 
branched chains generated by the APC/C complex. The E2 enzyme UBE2S assembles 
K11 linkages on top of the substrates, pre-ubiquitylated by the UBE2C enzyme with K63, 
K48 and K11 linkages. Branched K11/K48 conjugates have been shown to possess a 
unique interface, increasing their affinity to p97/VCP and the proteasome receptor S5a 
and facilitating faster degradation of the APC/C substrates (Boughton et al., 2020; Meyer 
and Rape, 2014). Branched K11/K48 chains have also been found on misfolded nascent 
proteins. Their formation depends on the ubiquitin ligases UBR4 and UBR5 (Yau et al., 
2017). Another example is the degradation of the secretory factor EVI/WLS, which is 
dependent on VCP activity and ubiquitin-conjugating enzymes UBC13, UBE2K and 
UBE2J2. Ubiquitin code on EVI/WLS includes K48, K63 and K11 linkages. High linkage 
specificity of UBC13 (K63) and UBE2K (K48) suggests that UBE2J2 might be the enzyme 
that builds K11 chains on the EVI/WLS substrate. Similar ubiquitin architecture, namely a 
combination of K63, K48 and K11 linkages, is triggered by K63-linked polyubiquitylation 
26 
 
in mammalian cells. This pathway has been employed to create a new class of PROTACs, 
which rely on K63 ubiquitylation (Akizuki et al., 2023). Apart from a role in protein turnover, 
several non-degradative roles of K11-linked chains have also been proposed. 
cIAP1/UBCH5-dependent polyubiquitylation of RIP1 with K11 linkages promotes NF-κB 
activation upon TNFα signalling (Dynek et al., 2010). Indirect evidence, based on 
depletion of the K11-selective DUB Cezanne, suggests that K11 chains are assembled 
on top of K63-linked chains during DNA damage response, limiting recruitment of 
RAP80/BRCA1-A complex to the sites of DNA damage (Wu et al., 2019).  
1.1.3.5 K27-linked ubiquitin chains 
Ubiquitin replacement strategy revealed that K27 linkages, along with K63 ones, 
are essential for the proper activation of the DNA damage response upon double-strand 
break induction. Ubiquitin ligase RNF168 was shown to be responsible for the 
polyubiquitylation of chromatin proteins with K27-linked chains, and a mass spectrometry-
based approach revealed histones as the main targets (Gatti et al., 2015). Another area 
where K27-linked chains were shown to be highly important is immune signalling. In 
response to cytoplasmic DNA accumulation, ubiquitin ligase AMFR modifies STING with 
K27-linked ubiquitin chains. This modification is crucial for recruiting the TBK1 kinase and 
downstream activation of the immune response (Wang et al., 2014). K27-linkages have 
also been found as a component of complex ubiquitin conjugates involving K6, K29 and 
K48 linkages (Zucchelli et al., 2010; Ben-Saadon et al., 2006). Apart from that, K27-linked 
ubiquitin dimers possess one additional interesting characteristic – compared to other 
linkages, they are resistant to cleavage by several linkage non-selective DUBs (Ubp6, 
USP5, USP2) – physiological relevance of this phenomenon remains to be understood. 
1.1.3.6 K29-linked ubiquitin chains 
K29-linked chains have been implicated in the ubiquitin fusion degradation (UFD) 
pathway, which is responsible for the degradation of proteins C-terminally fused to an 
uncleavable ubiquitin moiety. Ufd4 builds a K29-linked on a ubiquitin moiety of a UFD 
substrate. This chain is further modified with K48 linkages by Ufd2, leading to the 
degradation of the substrate (Johnson et al., 1995; Koegl et al., 1999). UFD pathway is 
generally conserved in human cells. Ubiquitin ligase TRIP12, which is a human 
27 
 
homologue of Ufd4, assembles K29-linked chains in vitro and in vivo and is involved in 
the degradation of UFD substrates (Park et al., 2009). Together with CRL2, TRIP12 is 
also involved in the formation of branched K29/K48 chains during PROTAC-induced 
degradation of BRD4, suggesting that roles of K29-linked polyubiquitylation by TRIP12 
extend beyond the UFD pathway (Kaiho-Soma et al., 2021). 
The only well-characterised example of non-proteolytic K29-linked 
polyubiquitylation is ubiquitylation of AXIN1 by Smurf1. This modification inhibits the 
interaction of AXIN1 with LRP5/6, leading to the repression of the Wnt/β-catenin pathway 
(Fei et al., 2013). 
 
1.1.3.7 K6-linked ubiquitin chains 
Mass-spectrometry analysis revealed enrichment of K6 linkages after UV treatment 
but not after ionising radiation (Elia et al., 2015). BRCA1/BARD1 complex was shown to 
build K6-linked chains as a response to replication stress (Morris and Solomon, 2004). 
However, the targets and readers of K6-linked ubiquitylation in these conditions are not 
yet identified. Another condition that leads to a build-up of K6 linkages is the depolarisation 
of mitochondria – in this case, the responsible ligase is Parkin. Ubiquitin replacement 
strategy reveals that K6-linked ubiquitylation in this condition is crucial for the timely 
initiation of mitophagy (Ordureau et al., 2015). HUWE1 is another ligase that creates K6-
linked chains in vitro and in vivo (Michel et al., 2017). Mitochondrial membrane protein 
MFN2 is a target of HUWE1-dependent K6-linked polyubiquitylation, however, the 
existence of other HUWE1 substrates is not excluded. Recently, K6-linked ubiquitylation 
has been linked to the quality control pathway at the stalled ribosomes, with RFN14 being 
the responsible ubiquitin ligase (Oltion et al., 2023). Finally, proteomic analysis reveals a 
unique behaviour of K6-linked chains with respect to VCP and proteasome inhibition: 
whereas K48 linkages accumulate in a cell upon inhibition of either VCP or proteasome, 
the abundance of K6 linkages significantly increases upon VCP inhibition but remains 
rather unchanged upon inhibition of proteasome (Heidelberger et al., 2018). This pinpoints 
K6 linkages as VCP substrates and, at the same time, highlights the non-degradative role 
of these linkages. 
28 
 
1.1.3.8 K33-linked ubiquitin chains 
Similar to K6-linked chains, K33 linkages accumulate in mammalian cells upon UV 
but not ionising irradiation (Elia et al., 2015). One of the described substrates of K33-
linked polyubiquitylation is XRCC1. Once allosterically activated by 5-
hydroxymethylcytosine, ubiquitin ligase UHRF2 polyubiquitylates XRCC1 with K33 
linkages, leading to the recruitment of RAD23B and activating base excision repair 
pathway (Liu et al., 2021). Apart from DNA damage, K33-linked polyubiquitylation occurs 
during T-cell activation, AMPK family kinase signalling and post-Golgi trafficking (Yuan et 
al., 2014; Al-Hakim et al., 2008; Huang et al., 2010). Recently developed split GFP-based 
polyubiquitin imaging approach PolyUB-FC revealed that the autophagosome cargo 
protein SQSTM1/p62 interacts with K33-linked chains via its UBA binding domain, 
promoting the transport of K33-polyubiquitylated cargo to the autophagosomes (Nibe et 
al., 2018). 
 
1.1.4 The ubiquitin-proteasome system 
Together with autophagy, the ubiquitin-proteasome system (UPS) is a key 
regulator of cellular metabolism, accounting for 80-90% of total protein degradation (Lee 
and Goldberg, 1998). The discovery of the UPS resulted in the award of the 2004 Nobel 
Prize to A. Ciechanover, A. Hershko and I. Rose (Ciechanover et al., 1980; Hershko et 
al., 1980). The main enzyme complex of the UPS is the 26S proteasome, which consists 
of the 20S core protease subunit and two regulatory 19S particles. 20S proteasome has 
a cylindrical shape, resulting from the stacking of two α-rings and two β-rings, each 
consisting of seven subunits. Proteolysis is performed by β1, β2 and β5 subunits, which 
cleave peptide bonds after acidic, basic and hydrophobic amino acids, respectively. 19S 
regulatory subunits cap the 20S proteasome on one or both sides (Figure 3A). Their 
primary role is to recognise polyubiquitylated substrates, unfold them and translocate 
them into the proteasome inner chamber, where the hydrolysis of peptide bonds takes 
place (Tanaka, 2009). Although the 26S proteasome possesses intrinsic ubiquitin 
receptors (Martinez-Fonts et al., 2020), several factors have been shown to bind 
polyubiquitylated substrates and shuttle them to the proteasome. An example is RAD23, 
which, on the one hand, selectively binds K48-linked chains and, on the other hand, 
29 
 
interacts with proteasome via its ubiquitin-like domain (Liang et al., 2014; Chen and 
Madura, 2002). As a part of the 19S regulatory particles, several deubiquitinating enzymes 
are associated with the 26S proteasome - USP14, RPN11, and UCH37. They remove 
ubiquitin moieties from polyubiquitylated substrates upon binding to the proteasome, 
protecting ubiquitin from degradation and replenishing the cellular ubiquitin pool (Shin et 
al., 2020). 
In some instances, direct shuttling of polyubiquitylated proteins to the proteasome 
is impossible. This happens, for example, when proteins reside in the membranes or are 
tightly bound to DNA. In this case, polyubiquitylated proteins undergo unfolding by the 
ATPase VCP/p97 concomitantly with their extraction from the corresponding cellular 
compartments (Meyer and Weihl, 2014; Meyer et al., 2012). VCP is a highly conserved 
protein, which belongs to type II AAA+ ATPase family due to the presence of two tandem 
ATPase domains (D1 and D2) and a large N-terminal domain, which mediates interactions 
of VCP with its substrates. VCP exists as a homohexamer, although two VCP hexamers 
can interact tail-to-tail and form a dodecamer: interaction between the subunits is 
promoted by ATP binding to the D1 domain (Wang et al., 2003) (Figure 3B). Unfolding of 
the substrates by VCP occurs via a hand-over-hand mechanism: sequential ATP 
hydrolysis by subunits leads to a conformational change and a step-by-step translocation 
of the substrate polypeptide chain through the central pore (Cooney et al., 2019). Although 
VCP can directly interact with and unfold a subset of its targets, typical VCP substrates 
are conjugated to ubiquitin (Ahlstedt et al., 2022). Recognition of polyubiquitylated 
proteins by VCP takes place with the help of its multiple adaptors. Interestingly, whereas 
some adaptors (UFD1, NPLOC4, UBXN1) prefer to bind K48-linked and K48-K11 
branched ubiquitin chains, marked preference for branched K63-K48 chains was 
observed for the VCP-associated proteins ZFAND2B, ATXN3 and RHBDD1 (Yau et al., 
2017; Lange et al., 2023).  
30 
 
 
Figure 3: Structure of 26S proteasome and VCP. (A) Structure of a human 26S proteasome, 
obtained by electron microscopy, in spheres representation (Zhang et al., 2022). (B) Structure of 
human VCP, obtained by electron microscopy, in cartoon representation. Different colours 
correspond to different subunits of a hexamer (Yu et al., 2021). 
1.1.5 Deubiquitylating enzymes 
Attachment of ubiquitin to target proteins can be reverted by the action of ubiquitin-
specific hydrolases, or deubiquitylating enzymes (DUBs). The activity of DUBs results in 
the termination of ubiquitin signalling (including prevention of protein degradation) and 
replenishment of the free ubiquitin pool. Human cells have ~100 DUBs, which can be 
roughly divided into two classes: cysteine proteases, which use cysteine as a catalytic 
residue, and metalloproteases, which employ a catalytic serine and zinc as a cofactor. 
Like E3 enzymes, many DUBs have their preferred substrates and linkage types. For 
example, USP1 removes monoubiquitin moiety from PCNA and FANCD2 (Huang et al., 
2006; Nijman et al., 2005); AMSH and OTUB1 are selective towards K63- and K48-linked 
polyubiquitin chains, respectively (McCullough et al., 2006; Wang et al., 2009). 
Deubiquitylating enzyme A20/TNFAIP3 is the only known protein that has ubiquitin ligase 
activity, deubiquitylating activity and binds to polyubiquitin chains. Interestingly, these 
three functions are directed towards different chain types: A20 selectively binds to linear 
ubiquitin chains, cleaves K63-linked and synthesises K48-linked ubiquitin chains 
(Tokunaga et al., 2012; Wertz et al., 2004). All these functions are essential for the NF-κB 
31 
 
activation, highlighting the complexity of the ubiquitin code in this pathway (Bai et al., 
2022). 
1.1.6 Tools to study the ubiquitin system 
1.1.6.1 Mass spectrometry 
Detection of protein ubiquitylation by mass spectrometry relies on the specific 
signature that ubiquitin leaves on the substrate after trypsin digestion: ubiquitin is removed 
from the substrate via cleavage of the peptide bond between R74 and G75, leaving two 
C-terminal glycine residues of ubiquitin attached to the substrate via an isopeptide bond. 
Detection of such peptides by mass spectrometry allows to identify those residues of the 
substrate, which had been ubiquitylated. Early proteomics approaches to study 
ubiquitylation relied on the expression of tagged ubiquitin, which allows isolating and 
identifying ubiquitylated proteins by means of affinity chromatography or 
immunoprecipitation (Peng et al., 2003). Development of the monoclonal antibody against 
GG dipeptide coupled via an isopeptide bond to a lysine allowed to identify global 
ubiquitylation profile of a cell in unperturbed conditions, i.e. without overexpression of 
tagged ubiquitin (Kim et al., 2011; Wagner et al., 2011). Absolute quantification of ubiquitin 
linkages became possible with the development of AQUA (absolute quantification) mass 
spectrometry, where synthesised isotope-labelled ubiquitin peptides are used as an 
internal standard. This method allowed direct quantification of ubiquitin linkages on the 
cyclin B1, which is heavily ubiquitylated by the anaphase-promoting complex: rather 
surprisingly, K48, K63 and K11 linkages amount up to 40% of total ubiquitin forms, the 
rest being monoubiquitylation events (Kirkpatrick et al., 2006). Detection of ubiquitin chain 
branching by mass spectrometry is more challenging: during trypsin digestion, the majority 
of the ubiquitin peptides, originating from branched chains and therefore containing two 
diGly remnants at the same time, will be cleaved between the two ubiquitylation sites. This 
makes it later impossible to understand whether these remnants originate from the same 
ubiquitin molecule. An interesting approach to overcome this problem was used to prove 
the presence of branched K63-K48 chains in vivo: R54, the only residue between K48 and 
K63 that leads to trypsin cleavage, was mutated to alanine. Tryptic digestion of this 
ubiquitin mutant produces peptides, which contain both K48 and K63 residues, allowing 
32 
 
to simultaneously monitor their ubiquitylation. Combining this approach with AQUA 
technology revealed that nearly one-fifth of all K63 linkages are branched via K48. The 
amount of branched K63-K48 linkages increases upon proteasome inhibition, suggesting 
their role in protein degradation (Ohtake et al., 2016).   
1.1.6.2 High-affinity binders to polyubiquitin chains 
Different conformations of ubiquitin linkages underlie the existence of high-affinity 
binders, which prefer one linkage type over the others. The first class of binders is linkage-
selective antibodies, which are available for all linkage types (Matsumoto et al., 2010; 
Matsumoto et al., 2012; Newton et al., 2008). Linkage-selective antibodies allow the 
detection of their cognate linkages by western blotting or immunofluorescence. However, 
their use in vivo is limited due to the complex maturation process required for the correct 
antibody folding. The use of non-antibody binders can overcome this limitation. A major 
class of binders to ubiquitin linkages is based on the existing linkage-selective ubiquitin-
binding domains. For example, the NZF domain of TAB2 and the UBAN domain of NEMO 
have been used as sensors for K63-linked and linear ubiquitin chains, respectively (Qin 
et al., 2022). Linkage selectivity of these domains is achieved due to their binding to the 
interface between two ubiquitin moieties around the (iso)peptide bond. Another approach 
relies on combining several weak ubiquitin-interacting motifs separated by appropriate 
linkers. As demonstrated for tandem ubiquitin-interacting motifs (tUIMs) of RAP80 and 
Ataxin-3, linker length determines the linkage selectivity of the domain. Whereas high-
avidity binding to K63-linked chains by RAP80 tUIMs requires 7 amino acid linkers 
between ubiquitin-binding entities, 2 amino acid linkers are optimal for binding to more 
compact K48-linked chains by Ataxin-3 (Sims and Cohen, 2009). Varying the type of 
ubiquitin-binding modules and amino acid composition of the linkers allows to create 
binders with different affinities toward desired linkages (Sims et al., 2012). 
Apart from adapting naturally occurring ubiquitin-binding domains, a number of 
artificial protein binders to ubiquitin linkages have been developed. These binders are 
typically based on a specific protein frame with several variable regions, which dictate 
specificity to a target. Existing high-throughput screening platforms allow the selection of 
binders by screening a library of peptide fragments incorporated in the variable regions. 
One example is affimers, ~12 kDa proteins based on a cystatin fold with two variable 
33 
 
regions that dictate their specificity (Tiede et al., 2017). An affimer developed for the K6 
linkage has found widespread use and revealed the involvement of K6-linked 
polyubiquitylation in the DNA damage response and VCP metabolism, as well as led to 
the identification of proteins ubiquitylated with this linkage type (Michel et al., 2017; 
Heidelberger et al., 2018). Another example is nanobodies – single-chain camelid 
antibodies, which have found use in fundamental science and also have therapeutic 
applications (Salvador et al., 2019). Selection of nanobodies on high-throughput display 
platforms in vitro bypasses the need of animal immunisation. Nanobodies have been 
developed not only for pure K48 and K63 linkages, but also for K63-K48 branched 
ubiquitin chains (Gonzalez-Santamarta et al., 2023; Lange et al., 2023). The latter 
nanobody was used to confirm the degradative function of branched K63-K48 linked 
chains and identify the previously underestimated role of these linkages in DNA damage 
response.  
 
1.1.6.3 Protein engineering approaches 
As discussed in section 1.1.1, recognition of ubiquitylated substrates by effector 
proteins differs from the recognition of small modifications, such as phosphate or methyl 
groups. Affinity to ubiquitylated substrates is often achieved by combining separate 
substrate-binding and ubiquitin-binding domains. It allows for certain conformational 
flexibility both for the modified substrate and the effector protein. Moreover, ubiquitin is 
attached to the substrate via its flexible C-terminus, contributing to the conformational 
freedom of the product. Indeed, X-ray scattering analysis and molecular simulations 
confirm the conformational flexibility of ubiquitylated PCNA (Powers et al., 2018). This is 
further supported by the fact that the formation of a complex between ubiquitylated PCNA 
and its reader protein – polymerase eta – involves significant rotation of the ubiquitin 
moiety (Lau et al., 2015). Therefore, a natural approach to mimic the ubiquitylation at the 
native lysine is to fuse a ubiquitin moiety N- or C-terminally to the substrate. Notably, the 
design of the fusion proteins should include removal or mutation of the terminal GG 
peptide, as it may lead to cleavage by DUBs in case of N-terminal fusions or uncontrolled 
conjugation to cellular proteins via its C-terminus in case of C-terminal fusions (Qian et 
al., 2002; Qin et al., 2016).  If the flexibility of the effector protein and ubiquitin-substrate 
34 
 
fusion allows their interaction, linear fusions to ubiquitin can rescue the phenotypes that 
originate from the loss of a target lysine residue or the responsible ubiquitin ligases. The 
number of known cases when linear fusions can substitute for the native ubiquitylation 
events is surprisingly high: for more than 20 proteins, linear fusions to ubiquitin or 
ubiquitin-like modifiers can functionally replace modifications at the native lysine residues 
(Asimaki et al., 2022). These known substrates belong to entirely different cellular 
pathways: DNA damage response (PCNA, RAD18, FANCD2), regulation of chromatin 
structure (H2A, H2B), endocytosis (EGFR, EEA1), immune response and related 
signalling pathways (RIP1, NEMO, p53). An interesting alternative approach has been 
used to mimic PCNA ubiquitylated at K164: PCNA was split between residues 163 and 
164, followed by an attachment of ubiquitin to the N-terminus of the second fragment 
(Freudenthal et al., 2010). The resulting fusion protein resembles PCNA ubiquitylated at 
K164: it forms stable trimers and supports translesion synthesis in vitro and in vivo. 
Unlike monoubiquitylation, there are only a few described attempts to mimic 
polyubiquitin chains using linear fusions. As fusions of multiple ubiquitins in-frame 
represent a linear chain, this approach may accurately represent linear and, in some 
cases, conformationally similar K63-linked chains, whereas their ability to mimic differently 
linked chains remains questionable. Linear fusion of three ubiquitins was successfully 
used to mimic a K63-linked chain on the yeast membrane transporter Ypq1 in the 
approach termed RapiDeg (Rapamycin-induced degradation). Recruitment of the chain to 
Ypq1 by means of the FKBP-FRB dimerisation system resulted in its rapid internalisation 
and degradation. Thus, a linear fusion of three ubiquitins can successfully mimic 
endogenous K63-linked ubiquitin chains in the membrane trafficking pathways (Zhu et al., 
2017). Another example comes from the mitophagy field: linear tetraubiquitin fusions 
harbouring the S65D phospho-mimicking mutation are proficient in recruiting phospho-
mimetic Parkin (Okatsu et al., 2015). This approach directly targets the ubiquitin chain to 
the outer mitochondrial membrane via a mitochondrial targeting domain. Therefore, a 
phosphorylated ubiquitin chain can lead to Parkin recruitment independently of the 
substrate to which it is attached. Significantly, targeting linear di-, tetra-, or hexa-ubiquitin 
chains to the mitochondrial outer membrane induces mitophagy independently of Parkin 
(Yamano et al., 2020). A third example is PCNA polyubiquitylation: Unlike native K63-
linked chains, linear polyubiquitin fusions to PCNA do not support error-free damage 
35 
 
bypass in S. cerevisiae (Zhao and Ulrich, 2010). At the same time, an extension of a 
ubiquitin moiety of a ubiquitin-PCNA fusion protein in a K63-linked manner supports error-
free damage bypass, even if chain length is limited to a single K63 linkage (Takahashi et 
al., 2020). This may reflect that combining a non-native linkage type with a non-native 
attachment site results in the fusion protein not being recognised by the endogenous 
effector(s) of polyubiquitylated PCNA. 
1.2 Genome stability   
Maintenance of genetic information is essential to sustain an organism's life and 
ensure the lives of the next generations. Endogenous and exogenous factors constantly 
threaten the integrity of DNA – the carrier of the entire genetic information of a cell. In this 
section, I will describe the most common sources of DNA damage and the mechanisms a 
cell can employ to maintain genetic information in challenging conditions. 
1.2.1 Types of DNA damage 
1.2.1.1 Base oxidation 
Reactive oxygen species (ROS) constantly threaten DNA as they are produced 
during cellular respiration due to ionising radiation or other pathways (Markkanen, 2017). 
Among all four DNA bases, guanine oxidation at position C8 (8-oxoG) (Figure 4A) is one 
of the most well-known examples of oxidative DNA damage, the other common ones 
being thymine glycol and 8-hydroxy adenine. Typical agents used to induce oxidative 
damage of DNA are hydrogen peroxide (H2O2) and hydroxyurea (HU). The latter affects 
cells in at least two different ways, depleting the dNTP pool and leading to the formation 
of H2O2 and oxidative damage of DNA (Sakano et al., 2001). Repair of oxidised bases 
typically includes their removal from DNA by OGG1 and MYH DNA glycosylases and 
subsequent abasic site repair (section 1.2.1.6). 
1.2.1.2 Base alkylation 
Alkylation of DNA bases (predominantly methylation of purine bases) is another 
prominent type of DNA damage. The most common methylated DNA bases are N7-
methylguanine (Figure 4B), O6-methylguanine and N3-methyladenine. 
Methylmethanesulfonate (MMS) and N-methyl-N′-nitro-N-nitrosoguanidine (MNNG) are 
36 
 
two DNA methylating drugs widely used to study the cellular response to DNA alkylation. 
Cellular enzymes can directly demethylate several methylated bases. For example, O6-
methylguanine can be directly demethylated by E. coli Ogt or mammalian MGMT (Dolan 
et al., 1990; Sedgwick and Lindahl, 2002). Methylated bases can be also excised by DNA 
glycosylases (Bjelland et al., 1993; Singer and Hang, 1997), leading to the formation of 
abasic sites (section 1.2.1.6). 
1.2.1.3 Bulky adducts 
This term applies to large (polycyclic) aromatic compounds covalently attached to 
DNA bases. The main difference between small base modifications (oxidation, 
methylation) and bulky adducts is that the latter significantly distort the DNA helix (Gómez-
Pinto et al., 2004). A typical example of a bulky DNA adduct is benzo[a]pyrene and its 
oxidised derivative BPDE, which efficiently modifies the N2 position of guanine (Hess et 
al., 1997) (Figure 4C). Repair of bulky adducts typically includes the excision of a 
damaged DNA strand with the subsequent filling of the resulting ~30 nt gap (nucleotide 
excision repair, NER) (Schärer, 2013). Importantly, DNA duplex distortion is a critical step 
in the NER-dependent repair of bulky adducts, as the distorted helix is recognised by the 
XPC–RAD23B complex, favouring the nucleotide excision repair pathway (Sugasawa, 
2001). 
1.2.1.4 UV photoproducts 
UVB (280-315 nm) and UVC (<280 nm) light can lead to DNA damage via the 
formation of cyclobutane pyrimidine dimers and (6-4) photoproducts (Figure 4D and 3E). 
Both lesions induce distortion of the DNA helix, which is more prominent in the case of (6-
4) photoproducts. Although certain organisms possess enzymes that can directly repair 
both types of photoproducts (photolyases) (Zhang et al., 2017), their repair in human cells 
takes place via nucleotide excision repair (Tornaletti and Pfeifer 1996). The importance of 
this pathway for the repair of UV-induced lesions is highlighted by the fact that mutations 
in the members of the NER machinery are associated with severe photosensitive genetic 
diseases in humans: Xeroderma pigmentosum, Cockayne syndrome and 
trichothiodystrophy (Bergoglio and Magnaldo, 2006). 
37 
 
1.2.1.5 Interstrand crosslinks 
Interstrand crosslinks (ICLs) represent a lesion where two complementary DNA 
strands become covalently linked. Unlike all described above types of lesions, interstrand 
crosslinks affect both DNA strands simultaneously, which makes their repair more 
challenging. Furthermore, the covalent linkage of two DNA strands poses a roadblock for 
transcription and replication machinery. This ability of ICLs to cause severe problems in 
cellular metabolism is reflected by the widespread use of ICL-inducing agents, such as 
cisplatin or oxaliplatin, as potent chemotherapeutic drugs (Figure 4F). Apart from the 
mentioned platinum-based compounds, ICLs can be induced by nitrogen mustards, 
mitomycin C, psoralen in combination with UVA light exposure and many other chemicals 
(Deans and West, 2011). Repair of ICLs involves the coordinated action of multiple 
enzymes, including structure-specific nucleases, DNA helicases and polymerases. 
Mutation of the components of the ICL repair system results in a rare genetic disorder, 
Fanconi anaemia (Moreno et al., 2021), which gives rise to the name of the repair 
mechanism (Fanconi anaemia pathway). 
 
Figure 4: Different types of DNA lesions. (A) 8-oxoguanine. (B) N7-methylguanine. (C) Adduct 
of guanine to BPDE. (D) Thymidine dimer. (E) 6-4 photoproduct formed by adjacent thymidines. 
(F) DNA crosslink, formed by cis-platin between two guanine residues. 
38 
 
1.2.1.6 Abasic sites 
Abasic sites can be considered as an extreme form of base modification, as the 
whole nitrogen base is lost from the nucleotide due to the hydrolysis of the glycosidic bond 
– a process termed depurination (for A and G bases) or depyrimidination (for C and T 
bases). Cleavage of a glycosidic bond can happen spontaneously (An et al., 2014) or as 
a result of the enzymatic activity of DNA glycosylases. Abasic sites in DNA can contribute 
to genomic instability in multiple ways. First, losing a nitrogen base at a specific position 
in DNA means losing genetic information, and bypassing abasic sites by translesion 
synthesis polymerases (TLS) can be mutagenic (Chan et al., 2013). Second, the ribose 
ring in the absence of a base resides in equilibrium with its aldehyde form and can 
therefore react with amino groups of neighbouring proteins, such as histones, leading to 
the formation of DNA-protein crosslinks (Figure 5) (Sczepanski et al., 2010). Although 
crosslinking of abasic sites to HMCES protein was shown to serve a protective role 
(Nowotny, 2019), nonspecific crosslinking to chromatin proteins may pose a serious threat 
to transcription and replication. Hydrolysis of a glycosidic bond in vivo is an irreversible 
process, and repair of abasic sites requires base excision repair - excision of the 
deoxyribose, filling of the gap and ligation of DNA strands (Krokan and Bjørås, 2013).  
 
 
Figure 5. Structure of an abasic site and the mechanism of its conversion into a DNA-protein 
crosslink. 
39 
 
1.2.1.7 DNA-protein crosslinks 
DNA-protein crosslinks (DPCs) represent a broad class of modifications where a 
protein is covalently attached to a DNA base or backbone. Aldehydes represent one class 
of potent DPC-inducing substances: a formyl group can sequentially react with two 
nucleophiles, ultimately linking them together via a methylene bridge. A DPC is formed if 
these nucleophiles are represented by DNA bases (typically amino- and imino-groups) 
and proteins (typically N-terminus and several amino acid side chains). The high 
crosslinking activity of formaldehyde led to the development of the chromatin 
immunoprecipitation (ChIP) technique, which enables the identification of DNA-binding 
profiles of any proteins of interest (Orlando, 2000). Another source of DNA-protein 
crosslinks is the enzymatic activity of certain enzymes, such as topoisomerases I and II. 
These enzymes create single- and double-strand breaks in the DNA duplex, respectively, 
while being attached to the DNA backbone via a transient covalent bond between the 
phosphate group in DNA and a tyrosine residue in the enzymes (Champoux, 2001). While 
the normal enzymatic cycle of topoisomerases I and II involves rejoining of the DNA ends 
with the simultaneous release of the enzyme, inhibition of topoisomerases (for example, 
by camptothecin and etoposide, respectively) results in the enzyme remaining covalently 
attached to DNA. The repair of DPCs highly depends on the nature of the attached protein. 
It may involve direct reversal of a crosslink (for topoisomerase adducts), cleavage of the 
trapped protein by SPRTN/Wss1 protease, conversion of DPCs into clean double-strand 
breaks, nucleotide excision repair and other pathways (Stingele et al., 2017). 
1.2.1.8 Single-strand breaks 
Together with abasic sites, single-strand breaks (SSBs) or nicks represent one of 
the most common types of DNA insults. They can arise due to ROS activity towards the 
DNA backbone, as an intermediate of the BER pathway or TOP1 activity, and they are 
naturally present during DNA replication due to discontinuous synthesis of the lagging 
strand. When not being timely repaired, SSBs result in highly poisonous one-ended 
double-strand breaks during DNA replication. Typical ways to induce SSB formation in 
cells include H2O2 and MMS as damaging agents, which trigger SSBs due to BER 
activation. However, the spectrum of lesions for these drugs is quite diverse. Inhibition of 
TOP1 also results in SSBs concomitantly with the formation of DPCs. A much cleaner way 
40 
 
to create SSBs relies on the activity of the D10A mutant of Cas9 (Ren et al., 2014), which 
inactivates one of two nuclease domains of Cas9 and therefore turns its DSB-inducing 
activity into a nickase one. Repair of SSBs is initiated by the poly (ADP-ribose) polymerase 
PARP1, which binds to and is activated by DNA breaks (Ali et al., 2012). PARP1 activity 
results in the accumulation of single-strand break repair proteins, which include DNA end 
processing factors and DNA ligases, leading to the sealing of the DNA nick (Caldecott, 
2008).  
1.2.1.9 Double-strand breaks 
Double-strand breaks, which occur when two strands of a DNA duplex are 
simultaneously broken, represent highly deleterious structures: a single DSB can be 
sufficient to kill a cell (Cui and Bikard, 2016). DSBs can arise from two adjacent SSBs in 
the opposite DNA strands or exposure to ionising radiation or certain chemicals 
(neocarzinostatin, zeocin). Inhibition of human DNA topoisomerase TOP2 by etoposide 
also leads to the formation of DSBs. However, DNA ends in this case remain covalently 
attached to the enzyme. The primary pathway of DSB repair in human cells is non-
homologous end joining (NHEJ), which involves the processing of DNA ends and their 
ligation by DNA ligase IV and operates throughout the cell cycle. In the case of defective 
NHEJ machinery, DNA ends can be joined via an alternative end-joining pathway, which 
relies on microhomology and often involves base insertions by DNA polymerase theta 
(Chang et al., 2017). During S and G2 phases, when a sister chromatid is available, DSBs 
can be repaired by homologous recombination: resection of DNA ends leads to an 
invasion of RAD51-coated ssDNA into a sister chromatid, DNA synthesis on a sister 
chromatid template and resolution of Holliday-junction intermediates (Wright et al., 2018).  
Pathway choice is highly dependent on the cell cycle stage and includes multiple layers 
of regulation involving phosphorylation and ubiquitin signalling  (Yun and Hiom, 2009; 
Schwertman et al., 2016; Ceccaldi et al., 2016).  
1.2.2 Replication stress 
The time a cell spends between two divisions is called a cell cycle and can be 
subdivided into distinct stages. During the G1 phase, the primary goal of a cell is to grow 
and accumulate sufficient nutrients for DNA replication. The S phase is defined by 
41 
 
replication and finishes once the genome information has been duplicated. During the 
following G2 phase, a cell prepares for division in the subsequent M phase. Two daughter 
cells, which appear after mitosis, enter the G1 phase, and the cell cycle repeats. The S 
phase is the most critical for preserving genetic information: mistakes during DNA 
replication in the form of mutations or genomic rearrangements are propagated to 
following generations. Replication starts from discrete genomic regions – replication 
origins – and involves a series of events called ‘origin licencing’, which is necessary to 
initiate replication. It includes loading the origin recognition complex (ORC) onto double-
stranded DNA, followed by recruitment of CDT1, CDC6 and MCM2-7 double hexamers. 
This complex is further converted into bidirectional replication forks upon activation by 
DDK and CDK kinases (Parker et al., 2017). Processive DNA synthesis is carried out by 
the replisome – a multiprotein complex involving the helicase, replicative polymerases 
Polε and Polδ, primase Polα, replication sliding clamp PCNA and other factors. In vitro 
reconstitution of a human replisome requires 43 proteins, and more proteins may be 
involved in vivo (Baris et al., 2022). As DNA synthesis by polymerases occurs exclusively 
in the 5’-3’ direction, two DNA strands are replicated differently: one strand is replicated 
continuously by Polε in the 5’-3’ direction (leading strand) and the other discontinuously 
in small patches (lagging strand). These patches, termed Okazaki fragments, originate 
from ~35 nucleotide-long RNA primers, which Polδ extends into ~200 nucleotide-long 
fragments. Maturation of the lagging strand, which is one of the most common DNA 
metabolic processes, requires removing the RNA primer and ligating the neighbouring 
Okazaki fragments (Maga et al., 2001; Raducanu et al., 2022). When DNA replication is 
complete, converging replication forks meet and are disassembled in a VCP-dependent 
manner (Dewar and Walter, 2017). 
Replication stress is defined as a condition which leads to the slowdown or stalling 
of individual replication forks or global DNA synthesis (Zeman and Cimprich, 2014). The 
primary source of replication stress is damaged DNA bases, which often stall replicative 
DNA polymerases (section 1.2.1). Ribonucleotides, often incorporated in DNA by Polε 
and Polδ, can also block processive DNA synthesis (Dalgaard, 2012). Another source of 
replication stress arises from DNA sequences prone to form non-B form secondary 
structures. These typically involve hairpins, DNA triplexes and G-quadruplexes. 
Secondary structures can lead to replication stress by inhibiting helicase-dependent DNA 
42 
 
unwinding or stalling DNA polymerases if they are formed by ssDNA after the helicase 
unwinding (Sharma, 2011). Transcription-replication conflicts (TRCs) can also lead to 
replication stress as collisions with transcription machinery prevent DNA synthesis by 
DNA polymerases. TRCs are a common source of genomic instability in highly transcribed 
DNA regions, such as ribosomal DNA (Lindström et al., 2018; Murakumo et al., 2001). 
Depletion of the cellular nucleotide pool is another common trigger of replication stress. 
Hydroxyurea is a commonly used replication stress-inducing drug, which leads to dNTP 
pool depletion by inhibiting ribonucleotide reductase (Turner et al., 1966). Nucleotide 
depletion can also occur due to the excessive number of origins being fired simultaneously 
and is often a consequence of oncogene overexpression (Beck et al., 2012; Srinivasan et 
al., 2013; Jones et al., 2013). 
Replication stress often leads to persistent exposure of single-stranded DNA, 
which appears due to the unwinding of the parental DNA duplex by the replicative helicase 
after the polymerase has stalled. Single-stranded DNA stretches are bound and protected 
by ssDNA-binding protein complex RPA. The protective role of RPA is demonstrated by 
the fact that replication collapse upon substantial replication stress is often dependent on 
RPA levels and can be prevented by RPA overexpression (Toledo et al., 2013). The 
regions where dsDNA of the newly synthesised duplex turns into the exposed ssDNA – 
also called primer-template junctions – have been shown to activate checkpoint signalling, 
involving the Ataxia telangiectasia and Rad3 related (ATR) kinase (MacDougall et al., 
2007).  
One pathway to cope with replication stress relies on the regulation of origin firing. 
In order to control the distribution of resources, such as dNTPs, not all the origins are fired 
at the same time, and a significant proportion of the origins is not fired at all during the 
unperturbed S phase (dormant origins). However, in the presence of replication stress, 
firing of dormant origins increases the number of active replication forks, allowing to 
complete replication in time despite the stress conditions. Importantly, firing of new origins 
is the primary pathway that allows the completion of replication upon long-term replication 
block (Petermann et al., 2010). Nevertheless, stressed forks do not always have to rely 
on the neighbouring forks and are often able to sustain replication in the presence of DNA 
damage without removing DNA lesions. This involves a series of pathways, which will be 
discussed in detail in the following sections.  
43 
 
1.2.3 DNA damage bypass 
As two DNA strands are complementary, either of them contains all the genetic 
information of the cell. A lesion in one DNA strand can be eliminated and replaced with a 
correct nucleotide using the second DNA strand as a template. However, in a number of 
cases, a second DNA strand is not available. This can happen when an interstrand 
crosslink affects complementary DNA bases or during DNA replication when DNA strands 
are unwound by the MCM2-7 helicase. The latter situation is hazardous, as the excision 
of damaged nucleotides from single-stranded DNA regions during DNA replication results 
in single-ended double-strand breaks, which may lead to cell death if not timely repaired. 
Therefore, cells have developed a series of pathways that allow the completion of DNA 
replication in the presence of DNA damage without repairing the damaged bases. This 
series of pathways is collectively termed ‘DNA damage bypass’. In the following sections, 
I will describe the critical components of the damage bypass system in yeast and 
mammalian cells and how they are coordinated to enable the faithful completion of DNA 
replication. 
1.2.3.1 Translesion synthesis 
High fidelity of the replicative DNA polymerases ε and δ is achieved by the structure 
of their catalytic pockets, which can only accommodate a perfect Watson-Crick base pair, 
as well as by their ability to proofread and replace wrongly incorporated bases utilising the 
3’-5’ exonuclease activity (Bębenek and Ziuzia-Graczyk, 2018). Translesion synthesis 
(TLS) polymerases have more relaxed catalytic pockets and therefore reduced 
requirements for an incoming base pair. On the one hand, this allows them to catalyse 
DNA synthesis over damaged DNA. On the other hand, less strict control of base pairing 
and the lack of exonuclease activity can result in a higher mutagenesis rate than that of 
replicative polymerases. Whereas there are three TLS polymerases in S. cerevisiae (Polζ, 
Polη and Rev1), their number increases to 10 in human cells, at least 5 of which can 
function in DNA damage bypass. These are Polζ and the members of a so-called Y family 
of polymerases - Polη, Polκ, Polι and REV1. Unique structural features allow these 
polymerases to bypass specific lesions in an error-free manner. For example, Polη can 
perform error-free DNA synthesis over CPDs, making this polymerase crucial for the 
cellular response to UV damage (Hendel et al., 2008). In line with this, levels of Polη 
44 
 
drastically increase after UV irradiation in a p53-dependent manner, increasing cellular 
resistance to UV light (Lerner et al., 2017). Polκ can insert correct bases opposite of 
benzo[a]pyrene adducts and thymine glycol (Jha et al., 2016; Yoon et al., 2010). Although 
there are examples of an error-free mode of action of TLS polymerases, certain lesions, 
such as (6-4) photoproducts or abasic sites, tend to be bypassed by TLS polymerases in 
a mutagenic way. However, the danger of increased mutagenesis is often lower than that 
of the persistent replication blockade, as the latter may result in fork collapse, double-
strand breaks and, ultimately, cell death. TLS polymerase REV1 is quite different from 
other Y-family members: in addition to performing TLS synthesis, it acts as a scaffold for 
the recruitment of other TLS polymerases, including Polη, Polκ, Polι and Polζ (Ohashi et 
al., 2004; Murakumo et al., 2001). At least for Polκ and Polζ, it has been shown that their 
recruitment to the UV-damaged chromatin depends on REV1, confirming its scaffolding 
role (Gallina et al., 2021). For many DNA lesions, their bypass by TLS polymerases results 
in imperfectly paired bases, unable to be extended by replicative DNA polymerases. B-
family TLS polymerase Polζ, consisting of a catalytic (REV3) and structural (REV7) 
subunits, has been shown to efficiently elongate unpaired primer termini, suggesting a 
two-step lesion bypass mechanism: dependent on the type of DNA damage, the first TLS 
polymerase inserts a base opposite the lesion, followed by an extension step by Polζ 
(Shachar et al., 2009; Martin and Wood, 2019). Similar to Polη, expression of REV3 
increases after DNA damage in a p53-dependent manner, representing an adaptive 
response to replication stress (Krieg et al., 2006). 
1.2.3.2 Template switching 
In some instances, stalled primer terminus can use a sister chromatid as a template 
for extension in a DNA damage bypass pathway called template switching. In budding 
yeast, template switching is activated by K63-linked PCNA polyubiquitylation by 
Rad5/Ubc13/Mms2, although the effectors of this modification are unknown. Many steps 
of template switching resemble those from the homologous recombination pathway of 
DSB repair: following DNA unwinding and resection by Pif1 and Exo1, respectively, the 
formation of Rad51 filaments leads to the annealing of the ssDNA to the complementary 
strand of the homologous sister chromatid. Following branch migration and DNA synthesis 
based on the sister chromatid template, the resolution of Holliday structure-like 
45 
 
intermediates by Sgs1-Top3-Rmi1 complex completes the pathway, resulting in the error-
free bypass of the lesion (Figure 6) (García-Rodríguez et al., 2018b; García-Rodríguez 
et al., 2018a; Branzei and Szakal, 2016). The proposed mechanism was confirmed by 
means of two-dimensional gel electrophoresis and electron microscopy of the pathway 
intermediates (Giannattasio et al., 2014; Branzei et al., 2008). Despite being best 
characterised in budding yeast, template switching also operates in human cells, as was 
shown for the bypass of UV photoproducts and bulky adducts (Izhar et al., 2013; Piberger 
et al., 2020). 
In budding yeast, template switching is initiated by PCNA polyubiquitylation with 
K63-linked chains. The responsible E3 enzyme, Rad5, in addition to ubiquitin ligase 
activity, can catalyse fork regression and recruit the TLS polymerase Rev1 (Blastyák et 
al., 2007; Gallo et al., 2019). However, all three activities of Rad5 are genetically 
separable and, therefore, may be necessary in different physiological circumstances (Choi 
et al., 2015). 
 
 
 
 
Figure 6: Schematic representation of the template switching pathway. Main steps are 
depicted, including gap invasion into the sister chromatid, DNA synthesis on a sister chromatid 
template and resolution of the Holliday junction-like intermediates.  
46 
 
1.2.3.3 Fork reversal 
One of the mechanisms activated when a replisome encounters a DNA lesion or 
stalls due to other reasons is a transformation of a replication fork into a Holliday junction 
structure – a process termed replication fork reversal (Figure 7). Detection of reversed 
forks became possible due to advances in electron microscopy techniques, which allow 
direct visualisation of these replication intermediates. In budding yeast, fork reversal is a 
rare event: in cells treated for 30 min with HU, only 1.5% of replication forks appear to be 
reversed. Their number, however, increases if checkpoint kinase Rad53 is absent, rising 
to 11,2% (Sogo et al., 2002). In striking contrast to yeast, replication fork reversal in 
mammalian cells is considered one of the main pathways initiated as a response to 
replication stress. Electron microscopy experiments, which allow direct visualisation of 
reversed forks, revealed that 15-30% of forks are reversed after treating human cells with 
10 common DNA-damaging agents (Zellweger et al., 2015). 
The roles of the replication fork reversal and whether it can be considered a mode 
of damage bypass is a matter of debate. On the one hand, fork reversal transfers a 
damaged nucleotide into the parental DNA duplex, where it can be faithfully repaired. On 
the other hand, if the 5’-DNA overhang in the regressed arm is longer than the 3’ one, it 
can serve as a template for DNA synthesis and thus lead to damage bypass. This 
pathway, however, lacks experimental evidence. Alternatively, cleavage of a reversed fork 
by structure-specific nucleases results in a double-strand break, which can be repaired by 
a break-induced replication (BIR) pathway (Hanada et al., 2007; Donnianni and 
Symington, 2013). Another role for fork reversal has been recently proposed by Mutreja 
and colleagues (Mutreja et al., 2018). Using immunolabelling of individual ICLs, the 
authors demonstrate that not only those forks that encountered the lesion undergo 
reversal, but global ATR-dependent fork reversal takes place as a response to replication 
stress. The fact that even unaffected forks undergo reversal suggests that the formation 
of reversed forks may be a protective mechanism which slows down global replication, 
thereby providing stressed forks more time to deal with the damage. 
Several enzymes can regress replication forks in vivo and in vitro, among which 
SMARCAL1, HLTF and ZRANB3 are the best-characterised ones. SMARCAL1 is targeted 
to replication forks via RPA and reanneals DNA strands, which were unwound by the 
47 
 
helicase. HLTF performs a dual function: it promotes fork remodelling via its HIRAN 
domain and polyubiquitylates PCNA, leading to ZRANB3 recruitment. Importantly, these 
three translocases are not redundant in vivo, suggesting they participate in different 
stages of fork regression (Halder et al. 2022; Achar et al. 2015; Motegi et al. 2008; Ciccia 
et al. 2012). Recent studies also suggest that fork reversal is a two-step pathway: 
coordinated activity of SMARCAL1, HLTF and ZRANB3 leads only to initial steps of fork 
remodelling, leading to DNA supercoiling, sumoylation of TOP2B and recruitment of the 
DNA translocase PICH, which catalyses extensive fork reversal (Tian et al. 2021). 
 
Figure 7: Fork reversal. Schematic representation of a fork reversal pathway. Electron 
microscopy images of a stalled and reversed fork are adapted from (Zellweger et al., 2015). 
Replication fork reversal is responsible for the nascent DNA degradation 
phenotype: depletion of certain factors, termed replication fork protectors, leads to 
degradation of newly synthesised DNA. Some of the best-characterised fork protectors 
are BRCA1 and BRCA2 (Lemaçon et al. 2017). Prolonged replication stress in the 
absence of these factors leads to MRE11 and EXO1-dependent DNA degradation, 
dependent on fork remodellers SMARCAL1, HLTF and ZRANB3, and thus likely 
originating from reversed replication forks (Taglialatela et al. 2017). Depletion of another 
fork protector, WRNIP1, also leads to nascent DNA degradation. However, in this case, it 
is dependent on the nuclease DNA2. WRNIP1 has been proposed to directly bind Holliday 
junctions of reversed replication forks and protect them from SLX4-dependent cleavage 
and formation of double-strand breaks (Porebski et al. 2019). Thus, nascent DNA 
48 
 
degradation may result from DNA2-dependent resection of double-strand breaks. 
Interestingly, nuclease CtIP and the cytoskeleton motor protein MYO6 behave similarly to 
WRNIP1 regarding fork protection (Przetocka et al., 2018; Shi et al., 2023). In all the 
abovementioned cases, nascent DNA degradation depends on all three major fork 
reversal enzymes (Figure 8). However, there is one exception to this pattern, which will 
be discussed in section 1.2.3.5. 
 
Figure 8: Nascent DNA degradation. Schematic representation of the two major pathways of 
fork protection. BRCA1/2 prevent resection of a reversed fork by the nucleases MRE11 and EXO1, 
whereas WRNIP1 protects the Holliday junction from the activity of structure-specific nucleases 
with subsequent DSB resection. 
 
1.2.3.4 Repriming 
Another pathway cells often utilise when a replication fork encounters a lesion is 
reinitiation of DNA synthesis downstream of the lesion. As replicative polymerases cannot 
start DNA synthesis on the blank ssDNA template and require the presence of a short 
primer, this pathway is often referred to as repriming. Repriming is an essential part of 
unperturbed replication of the lagging strand due to its discontinuous mode of synthesis. 
However, it is not typical for the leading strand in the absence of replication stress. 
Repriming downstream of a lesion results in a daughter-strand gap, essentially a region 
of ssDNA flanked by dsDNA. These postreplicative gaps differ from single-strand gaps 
49 
 
that appear during nucleotide excision repair, as they contain a lesion in the ssDNA stretch 
(Figure 9A). The presence of postreplicative gaps after replication over damaged DNA 
has been proven by single-molecule techniques, such as electron microscopy and DNA 
fibre assays (Figure 9B).  
In budding yeast, Polα is responsible for repriming downstream of a lesion 
(Fumasoni et al., 2015). In contrast, human cells have an additional enzyme, PRIMPOL, 
which is targeted to stressed replication forks by RPA and can initiate DNA synthesis on 
the leading strand ~14 nucleotides past the lesion (Guilliam et al., 2017). Repriming 
neither repairs DNA damage nor leads to damage bypass but merely allows the replication 
fork to continue DNA replication past the lesion, which needs to be repaired later in a 
postreplicative manner via translesion synthesis or template switching. Significantly, 
postreplicative damage bypass can be separated from bulk DNA replication: yeast cells 
that lack Rad18, which is essential for damage bypass, are able to complete the S phase 
even in the presence of DNA damage, accumulating gaps behind the replication forks. 
Subsequent re-expression of Rad18 in the G2 phase leads to gap filling, which can be 
visualised by microscopy as short DNA stretches scattered along the bulk DNA. 
Interestingly, whereas both TLS and TS pathways can fill postreplicative gaps, for the UV-
induced lesions, TLS by Rev1 and Polζ is a dominant pathway of gap filling, with template 
switching acting as a backup (Daigaku et al., 2010). It was further demonstrated that 
damage bypass can occur in a postreplicative manner in Rad18-proficient cells. The loci 
where it takes place are visualised by RPA foci, resulting from repriming and gap 
expansion (Wong et al., 2020). Evidence for the postreplicative gap filling has also been 
found for mammalian cells: bypass of BPDE-induced bulky adducts requires repriming by 
PRIMPOL and homologous recombination at gaps in a postreplicative manner (Piberger 
et al., 2020).  
Although repriming allows the cells to continue replication after encountering DNA 
lesions, extensive use of this pathway can be detrimental to cell viability. The formation of 
daughter-strand gaps has been identified as an underlying pathway for the sensitivity of 
BRCA-deficient cells to chemotherapeutic drugs (Panzarino et al., 2021). Furthermore, 
hyperaccumulation of daughter-strand gaps can lead to elevated mutagenesis rates due 
to the increased use of TLS in gap filling (Somyajit et al., 2021). 
50 
 
 
 
Figure 9: Daughter-strand gaps. (A) Schematic representation of daughter-strand gaps and their 
comparison to NER-dependent gaps. Note that lesion is not eliminated from DNA upon daughter-
strand gap formation. (B) Electron microscopy image of a ssDNA gap behind the fork, adapted 
from (Zellweger et al., 2015). P: parental DNA duplex; D: daughter DNA strands. The arrow 
indicates a daughter-strand gap. 
Finally, it should be noted that once a replication fork encounters the lesion, there 
is competition between repriming, fork reversal and translesion synthesis. HLTF knockout 
cells, which lack a functional fork reversal pathway, do not slow down replication in the 
presence of low-dose HU due to PRIMPOL-dependent repriming. Conversely, increased 
expression of PRIMPOL interferes with fork reversal and prevents nascent DNA 
degradation in BRCA1-deficient cells. At the same time, mutation of the HLTF HIRAN 
domain, which is responsible for fork reversal, leads to unrestrained fork progression 
dependent on the TLS polymerase REV1 (Bai et al., 2020). There are multiple other 
examples of the interplay between the discussed pathways, extensively reviewed in 
(Quinet et al., 2021). 
1.2.3.5 Roles of ubiquitylated PCNA in DNA damage bypass 
DNA sliding clamp PCNA is subject to multiple posttranslational modifications, 
including phosphorylation, acetylation, sumoylation and ubiquitylation. Among these, 
mono- and polyubiquitylation of PCNA play an essential role in the cellular response to 
DNA damage and replication stress (Figure 10). PCNA monoubiquitylation, carried out 
51 
 
by the E3 enzyme RAD18, is a signal for the recruitment of TLS polymerases in yeast and 
human cells.  A common feature of TLS polymerases is their interaction with PCNA and 
ubiquitin: REV1 has two ubiquitin-binding motifs and a PCNA-interacting BRCT domain; 
Polη – two PCNA-interacting peptides and a ubiquitin-binding zinc finger; Polκ - two 
PCNA-interacting peptides and two ubiquitin-binding UBZ domains (Guo et al., 2006; 
Yoon et al., 2014; Acharya et al., 2008). In yeast, monoubiquitylation of PCNA is essential 
for TLS, as damage-induced mutagenesis is completely abolished in strains harbouring 
the PCNA K164R mutation (Stelter and Ulrich, 2003). In human cells, PCNA 
monoubiquitylation favours TLS, but appears to be not essential for it (Temviriyanukul et 
al., 2012). Interestingly, in avian DT40 cells, recruitment of the TLS polymerase REV1 
requires PCNA monoubiquitylation at gaps but not at replication forks (Edmunds et al., 
2008). In addition to TLS polymerases, metalloprotease Spartan has been shown to bind 
ubiquitylated PCNA via a combination of a PIP box and a UBZ domain (Centore et al., 
2012). By recruiting VCP, Spartan promotes the removal of Polη from the damage site, 
preventing excessive mutagenesis (Davis et al., 2012). Finally, a combination of a PIP 
box and a UBZ domain was found in a nuclease SNM1A (Yang et al., 2010). In the context 
of alternative telomere lengthening, ubiquitylated PCNA recruits SNM1A, initiating 
resection and template-switching pathway at damaged telomeres (Zhang et al., 2023). 
A ubiquitin moiety on PCNA can be further extended into a K63-linked polyubiquitin 
chain. In budding yeast, the responsible E3 is Rad5, which operates with highly K63-
specific E2 Ubc13/Mms2. Although human cells possess two Rad5 homologs, HLTF and 
SHPRH, they are not essential for PCNA polyubiquitylation. Recent findings implicate E3 
ligase RFWD3 in promoting this PCNA modification. However, it needs to be clarified 
whether RFWD3 directly polyubiquitylates PCNA or it recruits other E3 ligases to the 
damage site. In S. cerevisiae, PCNA polyubiquitylation activates an error-free branch of 
the damage bypass by means of template switching (section 1.2.3.2). This pathway 
seems to be conserved in higher eukaryotes, as K63-linked polyubiquitylation was shown 
to protect human cells against TLS-induced mutations. Interestingly, whereas K63-linked 
polyubiquitylation in yeast is vital for cell survival after MMS and UV damage, human cell 
lines with K63R mutant as the only source of ubiquitin are sensitive to cisplatin, but not to 
UV irradiation unless Polη is co-depleted (Chiu et al., 2006). This implies a high level of 
redundancy for the TS and TLS in the bypass of UV-induced photoproducts in human 
52 
 
cells. In addition to the activation of template switching, K63-linked PCNA 
polyubiquitylation is essential in human cells for the replication fork reversal, where it acts 
as a recruitment signal for the DNA translocase ZRANB3 (Ciccia et al., 2012). Another 
possible effector of polyubiquitylated PCNA is the TLS polymerase Polζ. K63-linked PCNA 
polyubiquitylation is essential for Polζ-dependent lesion bypass during replication in X. 
laevis egg extracts (Gallina et al., 2021).  Consistent with these data, the S phase gap-
filling pathway in human cells relies on UBC13 and REV1-Polζ (Tirman et al., 2021).  
PCNA polyubiquitylation in human cells has recently been linked to the replication 
fork protection pathway: prolonged replication stalling in PCNA polyubiquitylation-deficient 
cells leads to DNA2-dependent nascent DNA degradation (Thakar et al., 2020). 
Interestingly, this pathway is substantially different from the “BRCA1” and “WRNIP1” 
pathways, as it depends entirely on SMARCAL1, only partially on ZRANB3 and is 
independent of HLTF. This suggests that PCNA polyubiquitylation may be important to 
protect replication intermediates that originate from SMARCAL1 and ZRANB3 but not 
HLTF activity, underlining differential roles of these enzymes in fork regression.  
Finally, PCNA is also decorated with K48-linked polyubiquitin chains upon 
hydroxyurea-induced fork stalling. K48-linked polyubiquitylated PCNA was shown to 
interact with Polκ, implicating this modification in the pathway of fork restart (Tonzi et al., 
2018). However, the writers of this modification, as well as its functions, are currently 
unknown. 
53 
 
 
 
Figure 10: Roles of PCNA ubiquitylation in DNA damage tolerance. Depicted are only the best 
characterised pathways: translesion synthesis, template switching and fork reversal. For the 
details see main text. 
  
54 
 
1.3 Aims of this work 
Among all posttranslational modifications, ubiquitylation is one of the most complex 
due to the variety of topologies that ubiquitin units may adopt on a given substrate. At 
least 8 different ubiquitin linkages, branched chains and multi-ubiquitylation events create 
distinctive interaction surfaces on the modified proteins, determining their interaction 
profiles. PCNA polyubiquitylation is unique among other instances of protein 
polyubiquitylation: although the fact that PCNA is subject to modification with K63-linked 
chains is known for more than 20 years, responsible readers of this modification are yet 
not characterised. Even though human cells possess the enzyme ZRANB3, which binds 
to polyubiquitylated PCNA, several cellular pathways in human cells cannot be 
mechanistically explained if ZRANB3 is assumed to be the only effector of 
polyubiquitylated PCNA.  
This work aims to provide a better understanding of the pathways acting 
downstream of PCNA polyubiquitylation in human cells. For this purpose, artificial tailor-
made enzymes are used to build a K63-linked polyubiquitin chain on PCNA in the absence 
of exogenous DNA damage. This technology, which has previously been implemented in 
S. cerevisiae, allows to mechanistically separate ubiquitin signalling on PCNA from other 
damage-induced pathways. Furthermore, exchanging the K63-specific enzyme with a 
K48-specific one allows direct comparison of differently linked polyubiquitin chains on 
PCNA. Finally, the expression of artificial enzymes will be compared to physiological 
scenarios where excessive PCNA polyubiquitylation takes place.  
  
55 
 
Chapter 2 
 
Materials and methods 
 
 
2.1 Reagents 
2.1.1 Chemicals and recombinant proteins 
Unless indicated differently, chemicals and recombinant proteins used in this study 
were purchased from Sigma Aldrich or Thermo Fisher Scientific. Chemicals purchased 
from other companies are listed below. 
Carl Roth:  Ethanol 96% and 99.5%, isopropanol, acetic acid glacial, 
guanidine hydrochloride, imidazole, urea 
Macherey-Nagel:   NucleoBond Xtra Midi Plus 
MedChemExpress:   ML-323 
New England Biolabs:  Restriction endonucleases, 10x CutSmart buffer 
Takara Bio:   2x DNA ligation kit 
Polyscience:    Polyethyleneimine linear (Molecular weight ~ 25000 g/mol) 
Promega:    FUGENE HD transfection reagent 
QIAGEN:  QIAquick gel extraction kit, FlexiTube siRNA solutions, Ni-
NTA agarose 
Recombinant proteins for the in vitro ubiquitylation assay (section 2.13) were 
generously provided by Sabrina Wegmann (former member of Helle Ulrich laboratory). 
Recombinant proteins for the ubiquitin chain restriction assay (section 2.12.3) 
were generously provided by Christian Renz (a member of Helle Ulrich laboratory). 
 
 
56 
 
 
2.1.2 Antibodies 
Table 1: List of primary antibodies used in this study. Unless otherwise indicated, the 
dilutions refer to western blotting. 
ID Name Species Source Dilution 
11 BrdU (BU1/75) Rat mAb Abcam ab6326 DNA fibre 
assay: 1:50 
38 EXO1 Rabbit pAb GeneTex 1:1000 
GTX109891 
42 Flag (M2) Mouse mAb Sigma Aldrich  1:2500 
F1804 
72 HLTF Rabbit pAb Alexandra Belayew, 1:1000 
UMONS (Belgium) 
122 PCNA (PC10) Mouse mAb IMB PPCF 1:5000 
125 PCNA  Rabbit pAb Abcam ab18197 1:2500 
133 γH2AX (JBW301) Mouse mAb Millipore 05-636-I 1:1000 
153 POLH Rabbit pAb Abcam ab17725 1:1000 
167 RAD18 (79B1048) Mouse mAb Abcam ab12007 1:1000 
199 RPA32 (12F3.3) Mouse mAb Abcam ab12320 1:2500 
261 Ubiquitin (VU1) Mouse mAb Tebu-bio VU101 1:1000 
262 Ubiquitin (K48- Rabbit mAb  Millipore 05-1307 1:1000 
selective, Apu2) 
263 Ubiquitin (K63- Mouse mAb Enzo BML-PW- 1:1000 
selective, HWA4C4) 0600 
265 Ubiquitin (P4D1) Mouse mAb CST 3936 1:5000 
272 VSV (P5D4) Mouse mAb Roche 1:2500 
11667351001 
276 ZRANB3 Rabbit pAb Bethyl A303-033A 1:1000 
324 Phospho-RPA S4-S8 Rabbit pAb Biomol A300-245A 1:2500 
334 Phospho-RPA S33 Rabbit pAb Biomol A300-246A 1:2500 
335 Phospho-RPA T21 Rabbit pAb R&D Systems 1:2500 
AF6654 
339 CHK1 (2G1D5) Mouse mAb CST 2360 1:1000 
340 Phosho-CHK1 S345 Rabbit pAb CST 2341 1:1000 
372 H2B (53H3) Mouse mAb CST 2934 1:2500 
392 Linear Ubiquitin Mouse mAb Merck MABS451 1:1000 
(LUB9) 
395 WRNIP1 Rabbit pAb Bethyl A301-389A-T 1:1000 
402 SMARCAL1 Mouse mAb Santa Cruz sc- 1:1000 
376377 
404 GAPDH Goat pAb Novus Biologicals 1:10000 
NB300-320 
408 RAP80 Rabbit pAb Novus Biologicals 1:1000 
NBP1-87156 
57 
 
417 HUWE1 Rabbit pAb ThermoFisher PA5- 1:1000 
20374 
423 BrdU (B44) Mouse mAb BD Bio 347580 DNA fibre 
assay: 1:50 
IF: 1:500 
425 PCNA-Ub (D5C7P) Rabbit mAb CST 13439 1:2000 
436 UBE2K/HIP2 (H-6) Mouse mAb Santa Cruz sc- 1:2000 
390339 
442 Ubiquitin 1002A Rabbit mAb Bio-Techne 1:2500 
MAB8595 
 
Table 2: A list of secondary antibodies used in this thesis. Unless otherwise indicated, 
the dilutions refer to western blotting. 
ID Name Species Source Dilution 
86 Mouse Alexa Fluor 488 Goat Life A-11001 IF: 1:500 
88 Mouse Alexa Fluor 647 Goat Life A-21236 DNA fibre assay 
1:100 
93 Mouse HRP Goat Pierce 1:10000 
156 Rabbit HRP Goat Pierce 1:10000 
186 Rat Alexa Fluor 488 Goat Invitrogen A-11006 DNA fibre assay 
1:100 
 
2.2 Media and solutions 
2.2.1 Media for bacteria 
All E. coli strains were grown in liquid Luria Broth (LB) medium or on LB-agar 
plates. If needed, liquid medium was supplemented with 100 μg/ml ampicillin or 30 μg/ml 
kanamycin, or antibiotic-containing LB-agar plates were used. Liquid LB medium, LB-agar 
plates and stock solutions of antibiotics were prepared by the IMB Media Laboratory.  
2.2.2 Media for mammalian cells 
 HEK 293T and MDA-MB-436 cells were maintained in Dulbecco’s Modified Eagle 
Medium (DMEM) supplemented with 2 mM L-glutamine, 100 U/ml penicillin, 100 μg/ml 
streptomycin and 10% (v/v) Fetal bovine serum (FBS).  RPE1 hTERT cells were cultured 
in RPMI 1640 medium with the same supplements as stated for the DMEM medium. If 
indicated, 2 μg/ml puromycin, 100 μg/ml blasticidin, 100 μg/ml zeocin or 400 μg/ml G418 
were added to the medium.  
58 
 
2.2.1 Solutions 
Following solutions were prepared by IMB Media Laboratory: 5x PBS, 1M Tris-HCl 
pH 6.3, 1M Tris-HCl pH 7.5, 1M Tris-HCl pH 8.0, 1M KCl, 2M MgCl2, 0.5 M EDTA pH 8.0, 
5M NaCl, 10x TBE, 10x wet transfer buffer. 
Blocking buffer for DNA fibre assay: 2% (w/v) BSA in PBST 
Blocking buffer for immunofluorescence: 3% (w/v) BSA in PBST 
Blocking buffer for western blotting: 5% (w/v) skim milk powder in PBST 
Buffer A for denaturing pulldown: 6M Gua-HCl, 50 mM NaH2PO4, 50 mM Na2HPO4, 10 
mM Tris-HCl pH 8.0, 0.1% Tween 20 
Buffer B for denaturing pulldown: 8M Gua-HCl, 80 mM NaH2PO4, 20 mM Na2HPO4, 10 
mM Tris-HCl pH 6.3, 0.1% Tween 20 
Cell fractionation buffer: 10 mM HEPES pH 7.5, 10 mM KCl, 1.5 mM MgCl2, 350 mM 
sucrose, 10% glycerol 
Cell lysis buffer for DNA fibre assay: 200 mM Tris-HCl pH 7.5, 50 mM EDTA, 0.5% SDS 
Dilution buffer for immunoprecipitation: 20 mM HEPES pH 7.5, 125 mM KOAc, 1.5 mM 
MgCl2, 10% glycerol 
Elution buffer for denaturing pulldown: 2x NuPAGE LDS sample buffer (Thermo Fisher 
Scientific), 200 mM imidazole, 10 mM DTT 
MES running buffer: 50 mM MES hydrate, 50 mM Tris, 0.1% SDS, 1mM EDTA 
MOPS running buffer: 50 mM MOPS, 50 mM Tris, 0,1 % (v/v) SDS, 1 mM EDTA 
PBST: 1x PBS, 0.1% (v/v) Tween 20 
PI staining buffer: 1x PBS, 20 μg/ml RNAse A (Sigma Aldrich), 100 μg/ml propidium iodide 
(PI) 
Ponceau S solution:  0.1% (w/v) Ponceau S in 5% (v/v) acetic acid 
Pre-extraction buffer for immunofluorescence: 100 mM PIPES pH 6.8, 1 mM EGTA, 100 
mM NaCl, 300 mM sucrose, 0,5% Triton X-100 
59 
 
RIPA buffer: 50 mM Tris-HCl pH 7.5, 150 mM NaCl, 1% IGEPAL CA-630, 0.5% sodium 
deoxycholate, 0.1% SDS, 2.5 mM MgCl2 
Trans-Blot Turbo buffer: 1x Trans-Blot Turbo buffer (Bio-Rad), 20% ethanol 
Ubiquitylation buffer: 40 mM HEPES pH 7.4, 8 mM magnesium acetate, 50 mM NaCl 
Wet transfer buffer: 192 mM glycine, 25 mM Tris-HCl pH 8.3, 15% (v/v) methanol 
 
2.3 DNA oligonucleotides 
Table 3: A list of DNA oligonucleotides used in this thesis. 
ID Name Sequence 
3343 Ub-K63R-fw CCCTGTCTGACTACAACATCCAGAGAGAGTCCAC
CCTGCACCTG 
3344 Ub-K63R-rev CAGGTGCAGGGTGGACTCTCTCTGGATGTTGTAG
TCAGACAGGG 
4455 Ub-K48R-fw GATTGATCTTTGCCGGTCGACAGCTAGAAGACGG
TAGAAC 
4456 Ub-K48R-rev GTTCTACCGTCTTCTAGCTGTCGACCGGCAAAGA
TCAATC 
4457 Ub-K63R-fwII GTCTGATTACAACATTCAGCGAGAGTCCACCTTA
CATCTTG 
4458 Ub-K63R-revII CAAGATGTAAGGTGGACTCTCGCTGAATGTTGTA
ATCAGAC 
4950 Ubc7-His6-NotI-rev GCAAGCGGCCGCTTAGTGATGGTGATGGTGATG
AG 
5085 VSV-fw TACACTGACATCGAAATGAATAGATTG 
5086 E363-NotI-rev GCAAGCGGCCGCTTAAAAAACCGCATCGTGAAAT
G 
5194 Ube2G2-NotI-rev GCAAGCGGCCGCTCACAGTCCCAGAGACTTCTG 
5196 Ubc7-NcoI-fw GACCACCATGGGATCGAAAACCGCTCAGAAACG 
5197 Ube2G2-NcoI-fw GACCACCATGGCGGGGACCGCGCTCAAGAG 
5285 Flag-NotI-rev GCAAGCGGCCGCTTACTTGTCGTCGTCATCCTTG
TAG 
5288 hcoUbc7-ATG-fw AGCAAGACCGCCCAGAAGCG 
5289 hcoUbc7-NotI-rv GCAAGCGGCCGCTCAGAAGCCCAGGCTCTTCAG 
5305 hcoCue1-SmaI-fw CGATCCCGGGGCAGAGCAACCAGCACCCAAG 
5436 E363 I227A-fw GAGTGTCCTGCGTGTTTTGAAAACATGG 
5437 E363 I227A-rev TTCAAAACACGCAGGACACTCTTTAATAGC 
5438 E363-fw AATCTCCTTGTCTCAAGCACTAATAG 
5439 VSV-rev CTTCCCCAATCTATTCATTTCGATGTC 
60 
 
5447 E363RINGdel-rev CTCTTTAATAGCGTTATCAGTCGTG 
5448 E363RINGdel-fw GATGCGGTTTTTTAAGCGGC 
5454 hcoE31 RBR-N-BglII-fw CAGTAGATCTGCTAGCGGTGCTGGAGG 
5455 hcoE31 RBR-N-SbfI-rev TTCTCCTGCAGGTACATGGCCAGGCC 
5456 hcoE31 RBR-C-SbfI-fw TGTACCTGCAGGAGAACGGCATCGACTG 
5457 hcoE31 RBR-C-NotI-rev GAGTGCGGCCGCCTACTTGCGGCGGCGAGGG 
5799 hcoUbc7-P2A-PacI-fw GATCTTAATTAACGGAAGCGGAGCCACGAACTTC
TCTCTGTTAAAGCAAGCAGGAGACGTGGAAGAAA
ACCCCGGTCCTGACTACAAAGACCATGACGGT 
5880 Ub-K48R-fwII CAGCAGAGGTTGATCTTTGCTGGGAGACAGCTGG
AAGATGGACGC 
5881 Ub-K48R-revII GCGTCCATCTTCCAGCTGTCTCCCAGCAAAGATC
AACCTCTGCTG 
5896 His-Ub-HindIII-fw GCATAAGCTTCGGGCTGCAGATGCATCACC 
5897 Ub-NotI-rev GCATGCGGCCGCTCACCCACCTCTGAGACGGAG
G 
5981 HOIP-C885A-fw CGCCCTGGCCCGAGGAGGCGCTATGCACTTTCA
CTGTACCC 
5982 HOIP-C885A-rev GGGTACAGTGAAAGTGCATAGCGCCTCCTCGGG
CCAGGGCG 
6000 His8-H2B-HindIII-fw GCATAAGCTTACCATGCATCACCATCACCATCAC
CATCACATGCCAGAGCCAGCGAAGTCTGCTC 
6002 H2B-BamHI-rev CGATGGATCCCTTAGCGCTGGTGTACTTGGTG 
6006 P2A-mutagenesis-fw CGTGGAAGAAAAACTCGCTTCTGACTACAAAGAC 
6007 P2A-mutagenesis-rev GTCTTTGTAGTCAGAAGCGAGTTTTTCTTCCACG 
6008 Ub-K11R-fw GAAGACCCTGACTGGTAGGACCATCACTCTCGAA
G 
6009 Ub-K11R-rev CTTCGAGAGTGATGGTCCTACCAGTCAGGGTCTT
C 
6060 Ub-K6R-fw GGCATGCAGATCTTCGTGCGGACCCTGACTGGT 
6061 Ub-K6R-rev ACCAGTCAGGGTCCGCACGAAGATCTGCATGCC 
6062 Ub-K27R-K29R- GACACCATTGAGAATGTCAGGGCAAGAATCCAAG
K33R-fw ACCGGGAAGGCATCCCTCCTGAC 
6063 Ub-K27R-K29R- GTCAGGAGGGATGCCTTCCCGGTCTTGGATTCTT
K33R-rev GCCCTGACATTCTCAATGGTGTC 
6100 H2B-GA-fw ATGCCAGAGCCAGCGAAG 
6101 H2B-GA-rev CTTAGCGCTGGTGTACTTGGTG 
6102 H2B-FRB*-GA-rev GACTTCGCTGGCTCTGGCATACTAGTCTTGCTGA
TGCGGCG 
6103 H2B-His8-GA-fw CACCAAGTACACCAGCGCTAAGCATCACCATCAC
CATCACCATCACTAAGCGGCCGCTCGAGTCTAGA
G 
6165 PIP box-fw CATGGGCATGAAGCAAAGCTCATTGCTGTCATTC
TTTTG 
6166 PIP box-rev GATCCAAAAGAATGACAGCAATGAGCTTTGCTTC
ATGCC 
61 
 
6167 OTUB1 C91S-fw CAGGCCTGACGGCAACTCTTTCTATCGGGCTTTC 
6168 OTUB1 C91S-rev GAAAGCCCGATAGAAAGAGTTGCCGTCAGGCCT
G 
6187 His6-HindIII-fw AGCTTATGCATCACCATCACCATCACA 
6188 His6-HindIII-rev AGCTTGTGATGGTGATGGTGATGCATA 
6219 yUbc13hco-HindIII-fw GATCAAGCTTCCACCATGGCCAGCCTGCCCAAGC 
6220 yUbc13hco-NotI-rev CGATGCGGCCGCTCACTCGGGCTTCTTCTTGGC 
6221 yMMS2 hco-HindIII-fw GATCAAGCTTCCACCATGAGCAAGGTGCCCCGCA
AC 
6222 yMMS2 hco-NotI-fw CGATGCGGCCGCTCAGAACGTCTCGCCCTCTTTG 
6264 His6-NUbo-ATG CACCATCACCATCACAAGCTTGGGCAGATCTTCG
mutagenesis-fw TGAAGACCC 
6265 His6-NUbo-ATG GGGTCTTCACGAAGATCTGCCCAAGCTTGTGATG
mutagenesis-rev GTGATGGTG 
 
 
2.4 RNA oligonucleotides 
Table 4: A list of RNA oligonucleotides used in this thesis. Information is provided for 
the sense (5’-3’) strand of the duplex. 
Name Sequence 5’-3’  Source 
siHLTF GCGAAAUGACUUAUACUAUTT Life Technologies 
siHUWE1 GGUCUAAUCAUGCCGCAGATT Life Technologies 
siRAD18#1 GUUCAGACAUCAUAAGAGATT Life Technologies 
siRAD18#2 GGAUUAUCUAUUCAAGGAATT Life Technologies 
siRAP80#1 GAUACAGUAUUGACUCGGATT Life Technologies 
siRAP80#2 GGGUUGCAGAAAACGAAGATT Life Technologies 
siSMARCAL1 CAGCUUUGACCUUCUUAGCAA Qiagen 
siUBE2K ACAGGAUGCUGUAGUAGCAAA Qiagen 
siWRNIP1#1 CCAAGGCUGUCAUUUUAUATT Life Technologies 
siWRNIP1#2 GAAACAUAGCAUAAGGUUUTT Life Technologies 
siZRANB3#1 CCGGAUUCACAUCUAUACUAA Qiagen 
siZRANB3#2 UGCCAGUGUAUGACACCUUAA Qiagen 
 
 
 
 
 
62 
 
2.5 Plasmids 
 
Table 5: A list of plasmids created in this study 
 
ID Name Construction  Use 
4528 pENTR4-Ubc7-His6 Ubc7-His6 was amplified from Gateway cloning 
p3365 with o5196 + o4950. PCR 
product was cloned into pHU1788 
via NcoI/NotI 
4596 pENTR4-VSV-PIP- VSV-PIP-E363 was amplified with Gateway cloning 
E363 o5085 + o5086 from a G-block 
(IDT). PCR product was cloned into 
p1788 via NcoI/NotI 
4599 pDEST/FRT/TO- Gateway LR recombination Overexpression of 
Flag-VSV-PIP-E363 between p4596 and p3543.  Flag-VSV-PIP-
E363 in 
mammalian cells; 
creation of FlpIn 
cell lines 
4635 pENTR4-VSV-PIP- E348 sequence was amplified from Gateway cloning 
E348-Flag a G-block (IDT) with o5305 + 
o5285. PCR product was cloned 
into pHU4597 via XmaI/NotI 
4645 pENTR4-hcoUbc7 hcoUbc7 sequence was amplified Gateway cloning 
from a G-block (IDT) with o5288 + 
o5289. PCR product was cloned 
into pHU1788 via NcoI/NotI 
4650 pENTR4-Ube2G2 Ube2G2 sequence was amplified Gateway cloning 
from p4393 with o5194 + o5197. 
PCR product was cloned into 
pHU1788 via NcoI/NotI 
4651 pDEST-3xFlag- Gateway LR recombination Overexpression of 
Ube2G2 between p4650 and p1804 3xFlag-Ube2G2 in 
mammalian cells 
4660 pDEST-3xFlag- Gateway LR recombination Overexpression of 
hcoUbc7 between p4645 and p1804 3xFlag-hcoUbc7 in 
mammalian cells 
4715 pDEST/FRT/TO- PCR with 5’-phosphorylated o5436 Overexpression of 
Flag-VSV-PIP-E363 + o5437 on pHU4599. Self-ligation Flag-VSV-PIP-
(I227A) of the PCR product E363 (I227A) in 
mammalian cells; 
creation of FlpIn 
cell lines 
63 
 
4720 pDEST/FRT/TO- PCR on p4599 with 5’- Overexpression of 
Flag-VSV-E363 phosphorylated oligos o5438 + Flag-VSV-E363 
o5439. Self-ligation of the PCR (I227A) in 
product mammalian cells; 
creation of FlpIn 
cell lines 
4756 pDEST/FRT/TO- PCR on p4599 with 5’- Overexpression of 
Flag-VSV-PIP- phosphorylated oligos o5447 + Flag-VSV-PIP-
E363ΔRING o5448. Self-ligation of the PCR E363ΔRING in 
product mammalian cells; 
creation of FlpIn 
cell lines 
4894 pLentiCMV/Tre/3G Gateway LR recombination Creation of stable 
Neo -VSV-PIP-E363 between p4893 and p4596 cell lines with 
inducible 
overexpression of 
VSV-PIP-E363 
5109 pENTR4-VSV-PIP- Ligation of three DNA fragments: Gateway cloning 
hcoE31 1. PCR on the G-block (IDT) with 
o5454 + o5455, digested with 
BglII/SbfI 
2. PCR on the G-block (IDT) with 
o5456 + o5457, digested with 
NotI/SbfI 
3. p4510 digested with NotI/BglII 
 
5110 pDEST/FRT/TO- Gateway LR recombination Overexpression of 
Flag-VSV-PIP-E31 between p3543 and p4510 Flag-VSV-PIP-E31 
in mammalian 
cells 
5111 pDEST/FRT/TO- Gateway LR recombination Overexpression of 
Flag-VSV-PIP-hcoE31 between p3543 and p5109 Flag-VSV-PIP-
hcoE31 in 
mammalian cells 
5112 pDEST/FRT/TO- Gateway LR recombination Overexpression of 
Flag-VSV-PIP-E348- between p3543 and p4635 Flag-VSV-PIP-
Flag E348-Flag in 
mammalian cells; 
creation of FlpIn 
cell lines 
64 
 
5136 pDEST/FRT/TO- 3xFlag-Ubc7 was amplified from Overexpression of 
Flag-VSV-PIP-E348- p4660 with o5799 + o5289. PCR Flag-VSV-PIP-
Flag-P2A-3xFlag- product was cloned into p5112 E348-Flag-P2A-
hcoUbc7 backbone via PacI/NotI 3xFlag-hcoUbc7 in 
mammalian cells; 
creation of FlpIn 
cell lines 
5242 pDEST-His10- p2685 was amplified with o5896 + Overexpression of 
Ubiquitin o5897 and cloned into p4530 via His10-Ubiquitin in 
NotI/HindIII  mammalian cells 
5340 pDEST-His10- Site-directed mutagenesis on Overexpression of 
Ubiquitin (K48R) p5242 with o5880 + o5881 His10-Ubiquitin 
(K48R) in 
mammalian cells 
5341 pDEST-His10- Site-directed mutagenesis on Overexpression of 
Ubiquitin (K63R) p5242 with o3343 + o3344 His10-Ubiquitin 
(K63R) in 
mammalian cells 
5342 pDEST-His10- Site-directed mutagenesis on Overexpression of 
Ubiquitin (K48R p5340 with o3343 + o3344 His10-Ubiquitin 
K63R) (K48R K63R) in 
mammalian cells 
5346 pDEST-His10- Site-directed mutagenesis on Overexpression of 
Ubiquitin (K11R p5342 with o6008 + o6009 His10-Ubiquitin 
K48R K63R) (K11R K48R 
K63R) in 
mammalian cells 
5347 pDEST-His10- Site-directed mutagenesis on Overexpression of 
Ubiquitin (K11R p5340 with o6008 + o6009 His10-Ubiquitin 
K48R) (K11R K48R) in 
mammalian cells 
5351 pDEST/FRT/TO- Site-directed mutagenesis on Overexpression of 
Flag-VSV-PIP- p5136 with o6006 + o6007 Flag-VSV-PIP-
hcoE348-P2A*-3xFlag- hcoE348-P2A*-
hcoUbc7 3xFlag-hcoUbc7 in 
mammalian cells 
5370 pcDNA5/FRT/TO- H2B sequence was amplified from Overexpression of 
His8-H2B-FKBP-CUb p3234 with o6000 + o6002 and His8-H2B-FKBP-
cloned into p5049 via CUb in 
HindIII/BamHI mammalian cells 
5372 pcDNA5/FRT/TO- Site-directed mutagenesis on Overexpression of 
His8-H2B-FKBP- p5370 with o4455 + o4456 His8-H2B-FKBP-
CUb(K48R) CUb(K48R) in 
mammalian cells 
65 
 
5374 pcDNA5/FRT/TO- Site-directed mutagenesis on Overexpression of 
His8-H2B-FKBP- p5370 with o4457 + o4458 His8-H2B-FKBP-
CUb(K63R) CUb(K63R) in 
mammalian cells 
5379 pDEST-His10- Site-directed mutagenesis on Overexpression of 
Ubiquitin (K6R K11R p5347 with o6060 + o6061 His10-Ubiquitin 
K48R) (K6R K11R K48R) 
in mammalian 
cells 
5380 pDEST-His10- Site-directed mutagenesis on Overexpression of 
Ubiquitin (K63-only) p5379 with o6062 + o6063 His10-Ubiquitin 
(K63-only) in 
mammalian cells 
5381 pDEST-His10- Site-directed mutagenesis on Overexpression of 
Ubiquitin (K0) p5380 with o3343 + o3344 His10-Ubiquitin 
(K0) in mammalian 
cells 
5382 pcDNA5/FRT/TO- Gibson assembly for: Overexpression of 
NUa-HA-FRB*-H2B- 1. PCR on p5046 with o6102 + NUa-HA-FRB*-
His8 o6103 H2B-His8 in 
2. PCR on p5370 with o6100 + mammalian cells 
o6101 
5425 pENTR4 PIP-OTUB1 p3457 was digested with Gateway cloning 
NotI/BamHI and ligated with the 
annealed oligos o6165 + o6166 
5431 pENTR4 PIP-OTUB1 Site-directed mutagenesis on Gateway cloning 
(C91S) p5425 with o6167 + o6168 
5457 pDEST/FRT/TO- Gateway LR recombination Overexpression of 
Flag-PIP-OTUB1 between p3543 and p5425 Flag-PIP-OTUB1 
in mammalian 
cells 
5458 pDEST/FRT/TO- Gateway LR recombination Overexpression of 
Flag-PIP-OTUB1 between p3543 and p5431 Flag-PIP-OTUB1 
(C91S) (C91S) in 
mammalian cells 
5477 pcDNA5/FRT/TO- yhcoUBC13 sequence was amplified Overexpression of 
yhcoUBC13  from a G-block (IDT) with o6219 + yhcoUBC13 in 
o6220. PCR product was cloned mammalian cells 
into p4636 backbone via 
HindIII/NotI 
5478 pcDNA5/FRT/TO- yhcoUBC13 sequence was amplified Overexpression of 
yhcoMMS2  from a G-block (IDT) with o6221 + yhcoMMS2 in 
o6222. PCR product was cloned mammalian cells 
into p4636 backbone via 
HindIII/NotI 
66 
 
5703 pcDNA5/FRT/TO- Site-directed mutagenesis on Overexpression of 
Myc-hcoE31(C885A)- p5236 with o5981 + o5982 Myc-
L20-FKBP-CUb hcoE31(C885A)-
L20-FKBP-CUb in 
mammalian cells 
5706 pcDNA5/FRT/TO- p5382 was digested with HindIII Overexpression of 
His6-NUa-HA-FRB*- and ligated with the annealed oligos His6-NUa-HA-
H2B-His8 o6187 + o6188, followed by site- FRB*-H2B-His8 in 
directed mutagenesis with o6264 + mammalian cells 
o6265 
 
Table 6: A list of plasmids generated by others and used in this study. 
ID  Name Use Source/Received from 
1788 pENTR4 Gateway cloning Simon Boulton 
1804 pDEST/N- Gateway cloning Stephen West 
3xFLAG/FRT 
1809 pOG44 Creation of stable cell Simon Boulton 
lines 
3543 pDEST/TO/FLAG Gateway cloning Martin Möckel 
FRT Puro 
3457 pENTR4-OTUB1 Gateway cloning Stephanie Nick 
4636 pcDNA5/FRT/TO Overexpression in Petra Beli laboratory 
mammalian cells; 
creation of stable cell 
lines 
4693 pcDNA5/FRT/TO- Overexpression in Evrydiki Asimaki 
Nua-HA-FRB*- mammalian cells 
E348-Flag  
4892 pLenti CMV Lentiviral vector for Vassilis Roukos laboratory 
rtTAE3 Blast the expression of the 
reverse tetracycline 
transactivator rtTAE3 
4893 pLentiCMV/Tre/3G Gateway cloning Vassilis Roukos laboratory 
Neo Dest 
5046 pcDNA5/FRT/TO- Overexpression of Evrydiki Asimaki 
Nua-HA-FRB*- Nua-HA-FRB*-E363-
E363-Flag Flag in mammalian 
cells 
5236 pcDNA5/FRT/TO- Overexpression of Evrydiki Asimaki 
Myc-hcoE31-L20- Myc-hcoE31-L20-
FKBP-CUb FKBP-Cub in 
mammalian cells 
 
67 
 
2.6 Strains and cell lines 
2.6.1 Bacterial strains 
Table 7: A list of the E. coli strains used in this study 
ID Strain Genotype Use Source 
14 Top Ten F-mcrA Δ(mrr-hsdRMS- Cloning Invitrogen 
mcrBC) Φ80LacZΔM15 Δ 
LacX74 recA1 araD139 
Δ(araleu) 7697 galU galK 
rpsL (StrR) endA1 nupG 
18 ccdB Survival 2 F-mcrA Δ(mrr-hsdRMS- Cloning of ccdB- Invitrogen 
mcrBC) Φ80lacZΔM15 containing vectors 
ΔlacX74 recA1 araΔ139 
Δ(ara-leu)7697 galU galK 
rpsL (StrR) endA1 nupG 
fhuA::IS2 
 
2.6.2 Mammalian cell lines 
Table 8: A list of the mammalian cell lines used in this study 
ID  Cell line Antibiotics Source 
120 RPE1 hTERT FlpIn TREX Blasticidin, Jonathon Pines 
Zeocin, G418 laboratory 
157 HEK 293T  Merck 
249 RPE1 hTERT FlpIn TREX Flag- Blasticidin, This study 
VSV-PIP-E363 Puromycin, G418 
257 RPE1 hTERT FlpIn TREX Flag- Blasticidin, This study 
VSV-E363 Puromycin, G418 
258 RPE1 hTERT FlpIn TREX Flag- Blasticidin, This study 
VSV-PIP-E363 (I227A) Puromycin, G418 
259 RPE1 hTERT FlpIn TREX Flag- Blasticidin, This study 
VSV-PIP-E363 (ΔRING) Puromycin, G418 
268 HEK 293T WT/PCNA K164R  George-Lucian 
Moldovan laboratory 
269 HEK 293T PCNA K164R  George-Lucian 
Moldovan laboratory 
270 RPE1 hTERT WT  Anja-Katrin Bielinsky 
laboratory 
271 RPE1 hTERT   Anja-Katrin Bielinsky 
laboratory 
273 RPE1 hTERT WT Tet-On VSV-PIP- Blasticidin, G418 This study 
E363 
68 
 
274 Rpe1 hTERT PCNA K164R clone Blasticidin, G418 This study 
2B10 Tet-On VSV-PIP-E363 
275 HEK 293T Tet-On VSV-PIP-E363 Blasticidin, G418 This study 
277 HEK 293T WT/PCNA K164R Tet-On Blasticidin, G418 This study 
VSV-PIP-E363 
278 HEK 293T PCNA K164R Tet-On Blasticidin, G418 This study 
VSV-PIP-E363 
308 RPE1 hTERT FlpIn TREX Flag- Blasticidin, This study 
VSV-PIP-E348 Puromycin, G418 
313 RPE1 hTERT FlpIn TREX Flag- Blasticidin, This study 
VSV-PIP-E348-P2A-3xFlag-hcoUbc7 Puromycin, G418 
391 MDA-MB-436  Petra Beli laboratory 
 
 
2.7. General methods for DNA manipulation 
2.7.1 Measurement of DNA concentration 
Concentration of DNA in solution was measured with the Nanodrop 2000 
spectrophotometer (Thermo Scientific) or with the DeNovix DS-11 spectrophotometer 
(Biozym). 
2.7.2 Agarose gel electrophoresis 
Agarose gels were prepared by dissolving 1% (w/v) agarose in 1x TBE buffer, 
supplemented with SYBR Safe DNA stain (Invitrogen). DNA was mixed 5:1 with 6x DNA 
loading dye, loaded on a 1% agarose gel and run at 100 V until the desired separation of 
DNA bands took place. 1 kbp or 100 bp New England Biolab DNA ladder was used as a 
size standard. 
2.8 Methods for molecular cloning  
2.8.1 Polymerase chain reaction (PCR) 
DNA amplification by polymerase chain reaction (PCR) was performed using HF 
polymerase (IMB Protein production core facility). 1-10 ng of DNA were mixed with 5 μl of 
10x HF buffer, 2.5 μl of 10 μM forward and reverse oligonucleotides, 1 μl of 10 mM dNTPs 
and 0.5 μl of HF polymerase (2 cU/μl) in a total volume of 50 μl. The reaction was 
performed in a Professional TRIO cycler (Biometra). PCR protocol included a denaturation 
69 
 
step at 98°C for 30 s followed by 30 cycles of 98°C for 10 s (denaturation), 50-72°C 
(depending on the melting temperatures of the oligonucleotides) for 30 s (annealing) and 
72°C for 30 s per kbp of the amplified DNA fragment (extension). PCR was completed by 
a final extension step at 72 °C for 10 min.  
2.8.2 Site-directed mutagenesis 
In order to introduce point mutations in a plasmid, a PCR with primers containing 
the desired mutations was performed as described in section 2.8.1 with the following 
modifications. The annealing temperature was set to 42°C during the first 3 cycles and 
increased to 58°C during the next 18 cycles. The total number of cycles was 21, and the 
extension time was 9 min. After the completion of a PCR, 5.5 μl of 10x CutSmart buffer 
(New England Biolabs) and 0.5 μl DpnI restriction nuclease (20 U/μl) were added, followed 
by overnight incubation at 37°C. PCR product was purified with a GeneJET PCR 
purification kit (Thermo Fisher Scientific) and transformed into chemically competent E. 
coli cells (section 2.10.2). Mutations were analysed by DNA sequencing of plasmid DNA 
isolated from individual bacterial colonies.  
2.8.3 Gibson assembly 
DNA fragments containing ~20 nt long overlapping ends were mixed 1:1 (molar 
ratio) in a total volume of 10 μl, followed by the addition of 10 μl of the 2x Gibson assembly 
master mix (IMB Protein Production Core Facility). The reaction mix was incubated at 
50°C for 1 h and transformed into chemically competent E. coli cells (section 2.10.2). 
Individual bacterial clones were analysed for the presence of the desired plasmid by DNA 
sequencing.  
2.8.4 Gateway cloning 
100 ng of a Gateway entry vector was mixed with 100 ng of a Gateway destination 
vector in a total volume of 7.5 μl. 2.5 μl of the 4x LR clonase mix (IMB Protein Production 
Core Facility) were added to the mix, followed by incubation at room temperature for 1 h. 
2 μg of Proteinase K were added to the reaction mix and incubated at 37°C for 10 min. 
The reaction mix was transformed into chemically competent E. coli cells (section 2.10.2), 
and the presence of the desired plasmid DNA was analysed in individual bacterial clones 
by DNA sequencing.  
70 
 
2.8.5 Restriction cloning 
  For DNA digestion with endonucleases, 1-2 μg of plasmid DNA or PCR product 
were digested in 1x CutSmart buffer with 10-20 U of the desired endonuclease overnight 
at 37°C. DNA fragments were separated by agarose gel electrophoresis, and the desired 
fragment was purified using the QIAquick Del Extraction kit (Qiagen).  
 For DNA ligation, vector (backbone) and insert DNA fragments were combined in 
a 1:3 molar ratio in a total volume of 10 μl. 10 μl of the 2x DNA Ligation mix (Takara) were 
added, and the reaction mix was incubated at 16°C for 30 min. Ligation mix was 
transformed into chemically competent E. coli cells (section 2.10.2). Plasmid DNA was 
isolated from individual bacterial clones, and the fragments of interest were sequenced. 
2.8.6 DNA sequencing 
  400-700 ng of plasmid DNA were mixed with 1 μl of 10 μM DNA primer in a total 
volume of 7 μl and sent for sequencing to StarSEQ GmbH.  
 
2.9 Methods for protein manipulation 
2.9.1 SDS polyacrylamide gel electrophoresis 
SDS polyacrylamide gel electrophoresis was performed using 4-12% NUPAGE 
gels (Life Technologies). Protein samples were mixed 3:1 (v/v) with 4x NUPAGE LDS 
Sample buffer (Thermo Fisher), supplemented with 100 mM DTT, and incubated at 95°C 
for 5 min. After loading on a gel, the samples were run at 140V in 1x MOPS buffer for 60-
90 min. Prestained PageRuler (Thermo Fisher Scientific) was used as a size marker. For 
the detection of free ubiquitin and low-molecular-weight ubiquitin chains,1x MES buffer 
was used. 
 
2.9.2 Western blotting  
 In most cases, semi-dry transfer with Trans-Blot Turbo Transfer System (Bio-Rad) 
was used. Filter paper stacks and nitrocellulose membranes were equilibrated in 1x Trans-
Blot Turbo buffer for 1 min. A filter paper stack was placed at the anode (bottom) part of 
71 
 
a Trans-Blot cassette, followed by a membrane, a gel, and another stack of filter paper. 
Air bubbles were rolled out, the cassette was closed, and a constant voltage of 15 V was 
applied for 12 min, with the maximum allowed current being 1.3A. Following the transfer, 
membranes were optionally stained with Ponceau S solution to visualise the efficiency of 
the transfer, and the membranes were further incubated in the blocking buffer for 1-2 h at 
room temperature with constant shaking. Membranes were incubated with primary 
antibodies overnight at 4°C with constant shaking. Following 4× 5 min washes with PBST, 
secondary antibodies were applied for 1-2 h at room temperature. Membranes were then 
washed again 4× 5 min with PBST and developed with Amersham’s ECL Prime or Select 
Western blotting detection reagent. Chemiluminescence signal was recorded on a Fusion 
FX system (Vilber).  
 For the detection of PCNA polyubiquitylation, wet transfer in a Mini Trans-Blot Cell 
(Bio-Rad) was used. The technique was similar to the one described above, with the 
exception of buffer composition (Wet blot buffer) and transfer conditions (100 V, 100 min). 
 For the detection of free ubiquitin in the UBICREST assay with VU1 anti-Ubiquitin 
antibody (section 3.5.3.1), after Trans-Blot Turbo transfer, membranes were washed 2× 
5 min with PBS and incubated with 0.5% glutaraldehyde solution in PBS, followed by a 
2 h incubation in the blocking buffer at room temperature. Subsequent steps were identical 
to the ones described above.  
 
2.10 Methods for E. coli 
2.10.1 Cultivation of E. coli 
All E. coli strains were cultured in liquid LB medium or on LB-agar plates 
supplemented with the appropriate antibiotic at 37°C. 
2.10.2 Transformation of chemically competent E. coli cells 
 An aliquot of chemically competent E. coli Top Ten cells was thawed on ice, mixed 
with ~100 ng plasmid DNA or up to 10% (v/v) of a ligation mix and incubated for 20 min 
on ice. Following 45 sec of heat shock at 42°C, cells were diluted with 1 ml of LB medium, 
incubated for 45 min at 37°C with constant shaking and plated on an agar plate with the 
72 
 
respective antibiotic. In case ampicillin was used, cells were plated on an agar plate 
directly after the heat shock.  
2.10.3 Isolation of plasmid DNA 
To isolate plasmid DNA for sequencing or genetic manipulations, a single E. coli 
colony was incubated in 5 ml of LB medium with the respective antibiotic and grown 
overnight at 37°C with constant shaking (200 rpm). Bacteria were pelleted by 
centrifugation for 10 min at 4500 ×g, followed by plasmid DNA isolation with the GeneJET 
Plasmid Miniprep kit (Thermo Fisher Scientific) according to the manufacturer’s 
instructions. For the preparation of transfection-grade DNA, a single E. coli colony was 
inoculated in 5 ml of LB medium with the respective antibiotic for 8 h, then diluted 1:1000 
into 220 ml of the antibiotic-containing LB medium and grown overnight at 37°C with 
constant shaking (200 rpm). DNA was subsequently isolated with the NucleoBond Xtra 
Midi Plus kit (Macherey-Nagel) according to the manufacturer’s instructions. 
 
2.11 Methods for mammalian cells 
2.11.1 Cell thawing and freezing 
 For cell thawing, a cryovial with frozen cells was transferred from the -150°C freezer 
into a water bath pre-warmed to 37°C. After thawing, cell suspension was added dropwise 
to a petri dish containing the respective cell medium. Cells were allowed to adhere to the 
dish overnight, and the following day the media was exchanged. 
 For cell freezing, exponentially grown cells were washed once with pre-warmed to 
37°C Dulbecco’s Phosphate-Buffered Saline (DPBS) (Gibco) and incubated with 0.05% 
Trypsin-EDTA-Phenol red (Gibco) until complete cell detachment. Cell suspension was 
mixed with complete DMEM or RPMI 1640 medium, followed by centrifugation at 300 ×g 
for 3 min. Cell pellet was resuspended in freezing medium (90% complete growth medium, 
10% DMSO), distributed in cryovials and transferred to the -80°C freezer in a CoolCell 
Freezing Container. The next day cryovials were transferred to the -150°C freezer for 
long-term storage.  
73 
 
2.11.2 Cell passaging 
 Cells were passaged once they reached ~90-100% confluency. After washing with 
DPBS, pre-warmed to 37°C, cells were incubated with 0.05% Trypsin-EDTA-Phenol red 
until complete cell detachment. Trypsin was inactivated by the addition of the complete 
growth medium, and cells were transferred to a new dish in a 1:8 to 1:10 dilution. Due to 
their slow growth rate, MDA-MB-436 cells were typically split in a 1:4 dilution.  
2.11.3 Cell harvesting 
In case cell lysates were needed, cells were washed on the plate once with DPBS, 
collected with a cell scraper in 1-2 ml of ice-cold DPBS and centrifuged at 300 ×g for 3 
min. DBPS was removed, and cell pellets were stored at -80°C until further processing. 
Cell harvesting for flow cytometry is described in section 2.11.11. 
2.11.4 Cell counting 
 The number of cells in a cell suspension after trypsinisation was determined with a 
TC20 Automated Cell Counter (BioRad) according to the manufacturer’s instructions. 
2.11.5 Transient transfection of DNA 
HEK 293T cells were transfected with polyethyleneimine (PEI) and RPE1 hTERT 
cells with the Fugene HD (Promega) reagent.  
2.11.5.1 Cell transfection with polyethyleneimine 
For transfection in 6-well plates, 2*105-5*105 HEK 293T cells were plated 24 h prior 
to transfection. 1.5 μg of plasmid DNA was mixed with 6 μl of PEI solution (1 mg/ml) in 
250 μl of DMEM media without FBS but with all other aforementioned supplements, 
vortexed and incubated for 10 min at room temperature. 750 μl of full DMEM media was 
then added to the DNA-PEI mix and the resulting 1 ml of the transfection mix was added 
to cells. For the transfection of larger cell plates, the amounts of reagents were scaled up 
accordingly. 
2.11.5.2 Cell transfection with Fugene HD 
 For transfection in 10-cm dishes, 1-2*106 RPE1 hTERT cells were plated 24 h prior 
to transfection. 3 μg of plasmid DNA were mixed with 9 μl of the Fugene HD reagent in 
74 
 
510 μl of Opti-MEM medium (Life Technologies) and incubated for 10 min at room 
temperature. Cell medium was replaced with 10 ml of fresh full RPMI 1640, followed by 
the dropwise addition of the transfection mix. The medium was replaced 24 h after 
transfection. 
2.11.6 Transient transfection of siRNA 
 Transient cell transfection with siRNA duplexes was performed with the 
Lipofectamine RNAiMAX reagent (Thermo Fisher Scientific). For transfection in 10-cm 
dishes, 1-2*106 RPE1 hTERT cells were plated 24 h prior to the transfection. 10 μl of 
siRNA duplex solution was mixed with 25 μl of the Lipofectamine RNAiMAX reagent in 1 
ml of Opti-MEM medium and incubated for 20 min at room temperature. Cell medium was 
exchanged with 4 ml of Opti-MEM, pre-warmed to 37°C, followed by dropwise addition of 
the transfection mix. For RPE1 hTERT cells, medium was exchanged 8 h, for MDA-MB-
436 cells – 4 h after transfection. Functional experiments were performed 72 h after siRNA 
transfections. 
2.11.7 Creation of stable cell lines via Flp-In integration 
  2.7 μg of the Flp-recombinase expression vector (pOG44, ID 1809) was mixed with 
0.3 μg of the target plasmid and transfected in RPE1 hTERT FlpIn TREX cells (ID 120) as 
described in section 2.12.6. 24 h after transfection, cell medium was replaced, and 48 h 
after the transfection puromycin was added to the final concentration of 2 μg/ml. During 
the following 2 weeks, the medium was replaced every 2-3 days until bulk cell death took 
place and colonies of puromycin-resistant cells appeared. Surviving cells were 
trypsinised, pooled and further cultured with the standard cell culture techniques.  
2.11.8 Lentiviral transduction 
All manipulations that included lentiviral particles were conducted in the IMB S2 
laboratory. Recipient cells were plated in 10-cm dishes 24 h prior to transduction at 5*105 
(for RPE1 hTERT) or 2*106 (for HEK 293T) cells per plate. For the generation of lentiviral 
particles, lentiviral packaging plasmids (ID 3805, 3806 and 3807) were co-transfected in 
HEK 293T cells together with the rtTAE3-expressing vector (ID 4892) or the target transfer 
plasmid (based on the pLENTI CMV vector, ID 4894), as described in section 2.11.5.1. 
For each transfection, 4*106 of HEK 293T cells, 2,5 μg of each packaging plasmid and 4 
75 
 
μg of rtTAE3-expressing vector or the target plasmid were used. 7 h after transfection, the 
medium was changed with 10 ml of full DMEM, and cells were allowed to express the 
proteins for 24 h. After that, supernatants containing rtTAE3 and target plasming-
containing viral particles were combined, mixed with 16 μl polybrene (10 mg/ml stock 
solution), filter-sterilised and added to the cells-recipients. Media was replaced 24 h after 
transduction. 48 h after transduction, 100 μg/ml blasticidin and 400 μg/ml G418 were 
added to the medium. During the following 2 weeks of selection, cell medium was changed 
every 2-3 days, and cells were split 1:5 if they reached confluency.  
2.11.9 DNA fibre assay 
Cells were incubated with 50 μM CldU, washed 3× with pre-warmed to 37°C DPBS, 
followed by labelling with 50 μM IdU. The exact labelling times are described in the figure 
legends. If indicated, hydroxyurea was added during or after the second pulse, as 
described in the figure legends. After labelling, cells were washed 3× with ice-cold DPBS, 
trypsinised for 2-3 min at 37°C, diluted with complete cell media, centrifuged for 3 min at 
300 ×g, resuspended in DPBS, counted and diluted to the concentration of 1.75*105 
cells/ml. 7,5 μl of the lysis buffer were mixed with 4 μl of the cell suspension directly on 
the surface of a SuperFrost Ultra Plus Gold adhesion slide (Thermo Fisher Scientific). 
Cells were lysed on the slide at room temperature for 8.5 min. After that, the slide was 
tilted, allowing the drop to run down. The precise angle at which the slide was tilted 
depended on the temperature and the humidity of the air and was adjusted for every 
experiment in a way that it took at least 1 min for each drop to reach the end of the slide. 
The slides were air-dried for 30 min and fixed overnight in a fixation solution. Next day, 
the slides were washed 2× 3 min in PBS, dipped in water and denatured in 2,5M HCl for 
1,5 h at room temperature. Slides were washed with PBS 6× 2,5 min and blocked with a 
blocking solution for 40 min at room temperature. After that, the slides were incubated 
with primary antibodies (anti-BrdU ID 11 and 423) for 2,5h at room temperature, washed 
3× 5 min in PBST, followed by the incubation with the secondary antibodies (ID 88 and 
186) for 1 h at room temperature in the dark. The slides were washed 3× 5 min in PBST, 
2× 1 min in water, air-dried and mounted in ProLong Diamond Antifade Mountant (Thermo 
Fisher Scientific). Images of DNA fibres were acquired with a Widefield Fluorescence 
Microscope (Thunder, LASX software, Leica) (magnification: 63x, NA 1.44 HC PL APO oil 
76 
 
immersion objective; LED illumination and the corresponding emission filters: 635 nm, 
642/80 and 475 nm, 535/70). Lengths of DNA fibres were quantified manually using the 
Fiji/ImageJ software. 
2.11.10 Preparation of cell lysates 
  Cells were collected with a cell scraper in ice-cold PBS, centrifuged at 1000 ×g for 
3 min, the cell pellet was resuspended in RIPA buffer, supplemented with 1x cOmpleteTM 
protease inhibitor cocktail (Roche) and Sm nuclease (0.625 cU/ μl), and incubated for 1 h 
on ice. Lysates were centrifuged at 21500 ×g for 10 min, and the supernatant was 
transferred to a new tube. Protein concentration was determined with the Protein Assay 
Dye Reagent (Bio-Rad) according to the manufacturer’s instructions.   
2.11.11 Cell cycle analysis by flow cytometry 
For cell cycle analysis by flow cytometry, cells were washed once on a plate with 
DPBS, pre-warmed to 37°C, and incubated with 0.05% Trypsin-EDTA-Phenol red (Gibco) 
until complete cell detachment. Trypsin was inactivated by addition of complete cell media, 
cells were centrifuged at 300 ×g for 5 min, washed once with PBS and centrifuged again 
at 300 ×g for 5 min. The supernatant was discarded, and the cell pellet was resuspended 
in 50 μl of ice-cold PBS. 1 ml of 70% ethanol, pre-cooled to -20°C, was added dropwise 
to the cell suspension with constant vortexing. Cell suspension was stored overnight at 
4°C and then transferred to -20°C until further processing. For the propidium iodide (PI) 
staining, cells were centrifuged at 1000 ×g for 10 min, washed once with 1 ml of ice-cold 
PBS and centrifuged again at 1000 ×g for 10 min. The supernatant was discarded, cells 
were resuspended in 500-1000 μl of PI staining buffer and incubated for 30 min at room 
temperature. Cells were analysed on the BD LSRFortessa SORP system (BD 
Biosciences) equipped with the BD FACSDIVA software. Data were analysed with the 
Flow Jo software. 
For 2D cell cycle analysis, 10 μl of EdU were added to cell medium 30 min before 
harvest. Coupling of the AlexaFluor 647 fluorophore was performed with Click-ITTM 647 
Flow Cytometry Assay kit (Thermo Fisher Scientific) according to the manufacturer’s 
instructions. Analysis of cells by flow cytometry was performed as described above.  
77 
 
2.11.12 Detection of single-stranded DNA by immunofluorescence 
RPE1 hTERT cells were plated in 6-well plates with coverslips (Carl Roth) at 
650000 cells/well and cultured in the presence of 10 μM BrdU for 24 h. Next, cells were 
treated as described in the respective figure legend, coverslips were transferred to a 24-
well plate, washed once briefly with ice-cold PBS and incubated for 5 min with pre-
extraction buffer on ice. Following another wash with PBS, cells were incubated with 4% 
formaldehyde in PBS for 10 min at room temperature and permeabilised with 0.3% Triton 
X-100 in PBS for 15 min at room temperature. Coverslips were then washed again in PBS 
and incubated in the blocking buffer for 1 h. Primary anti-BrdU antibody (clone B44, ID) 
solution was applied overnight at 4°C. Coverslips were washed 3× 5 min in PBST, 
incubated with secondary antibody (ID 86) for 1 h at room temperature, washed again 3× 
5 min in PBST and incubated with 1 μg/ml Hoechst 33342 (Thermo Fisher Scientific) in 
PBS for 5 min at room temperature for total DNA staining. Image acquisition and analysis 
were performed as described in section 2.11.9. 
2.11.13 SILAC labelling of cells 
For the analysis of the interactome of polyubiquitylated PCNA, stable isotope 
labelling by amino acids in cell culture (SILAC) mass spectrometry-based approach was 
used (Ong and Mann 2006). HEK 293T were labelled with ‘light’ (Lys-0, Arg-0), ‘medium’ 
(Lys-4, Arg-6) or ‘heavy’ (Lys-8, Arg-10) amino acids for 2 weeks and transfected as 
described in section 2.11.5.1. PCNA was immunoprecipitated as described in section 
2.12.1. During the last wash, beads from different samples belonging to the one replicate 
were pooled. Downstream processing and data analysis were performed by Ivan Mikicic 
(Petra Beli laboratory). 
2.11.14 Cell viability assay 
Cells were transfected with siRNAs as described in section 2.11.6 and, after 24 h, 
plated on a 96-well plate at 2000 cells per well. Next day, cell medium was replaced, and 
cells were treated as indicated in the respective figure legend for 72 h. Viability of cells 
was assessed with the MTT Cell Viability Assay Kit (Sigma Aldrich) according to the 
manufacturer’s instructions. 
78 
 
2.12 Detection of protein ubiquitylation 
2.12.1 Detection of PCNA ubiquitylation by PCNA immunoprecipitation 
Cells (typically 1 sub-confluent 15-cm plate per condition) were treated as indicated 
in the respective figure legends and collected with a cell scraper in ice-cold PBS. Cells 
were centrifuged at 1000 ×g for 3 min, the cell pellet was resuspended in the cell 
fractionation buffer, supplemented with 1× cOmpleteTM protease inhibitor cocktail (Roche), 
and incubated for 15 min on ice. The suspension was then centrifuged at 1300 ×g for 5 
min, the supernatant was discarded, and cell pellets were frozen at -80°C for at least 30 
min. Pellets were then resuspended in 1 ml of ice-cold RIPA buffer, supplemented with 1x 
cOmpleteTM protease inhibitor cocktail (Roche) and Sm nuclease (0.625 cU/ μl), followed 
by incubation for 45 min on ice. 60 μl of 5M NaCl was added to the lysates, which were 
briefly vortexed and incubated on ice for 15 min. Lysates were centrifuged at 21500 ×g 
for 10 min, supernatants were transferred to new tubes, and protein concentration was 
determined with the Protein Assay Dye Reagent (Bio-Rad) according to the 
manufacturer’s instructions.  Equal amounts of proteins for different conditions were mixed 
with RIPA buffer to make the total volume up to 1 ml, and 300 μl of dilution buffer for 
immunoprecipitation were added. 13 μl of each sample were taken out, mixed with 5 μl of 
4x NUPAGE LDS Sample buffer (Thermo Fisher), supplemented with 100 mM DTT, 
incubated at 95°C for 5 min and later used as the ‘input’ samples. 1.5 μg of monoclonal 
PC10 were added to each sample except for the negative control, and the samples were 
incubated for 2 h at 4°C. 40 μl of 50% Protein A-agarose slurry per sample were washed 
once with 500 μl of RIPA buffer and added to the cell lysates, followed by an overnight 
incubation at 4°C. Beads were then centrifuged at 1000 ×g for 3 min, the supernatant was 
discarded, and the beads were washed 6× 5 min with 1 ml of RIPA buffer at 4°C. After the 
last wash, the supernatant was removed completely with a syringe, and proteins were 
eluted in 60 μl of 2× NUPAGE LDS Sample buffer (Thermo Fisher), supplemented with 
50 mM DTT, at 95°C for 15 min. Beads were then centrifuged at 1000 ×g for 3 min, and 
the eluate was used for SDS-PAGE and western blotting (sections 2.9.1 and 2.9.2). 
 
79 
 
2.12.2 Detection of protein ubiquitylation by denaturing Ni-NTA pulldown 
HEK 293T cells were plated at 5*106 cells per 10-cm plate (for the pulldown of His-
tagged histone H2B, 48 h prior to harvest) or at 2.5*106 cells per 10-cm plate (for the 
pulldown of his-tagged ubiquitin, 72 h prior to harvest) and transfected with the required 
constructs 24 h after plating. On the day of harvest, cells were optionally treated as 
indicated in the respective figure legends and collected in ice-cold PBS with a cell scraper. 
After centrifugation at 1000 ×g 3 min, the cell pellet was lysed in 1.5 ml of buffer A. Lysates 
were sonified with the Branson Ultrasonics sonifier (2-3 pulses at the lowest intensity) to 
reduce the viscosity. 22.5 μl of 1M imidazole and 40 μl of 50% Ni-NTA slurry were added 
to each sample and incubated overnight on a rotation wheel at 4°C. Beads were washed 
4× 5 min buffer A and 2× 5 min buffer B at room temperature, followed by incubation with 
50 μl elution buffer for 10 min at 95°C. Beads were centrifuged for 5 min at 1000 ×g, 
eluates were collected and analysed by SDS-PAGE and western blotting (sections 2.9.1 
and 2.9.2) 
2.12.3 Ubiquitin chain restriction assay (UBICREST) 
For the in vitro ubiquitin chain restriction assay (UBICREST), PCNA was 
immunoprecipitated from mammalian cells as described in section 2.12.1 before the 
elution step. Beads were additionally washed 2 times with 500 μl of 1x ubiquitylation 
buffer, and after the last wash, the supernatant was completely removed with a syringe. 
Beads were resuspended in 80 μl of the 1x ubiquitylation buffer. If indicated, AMSH, 
OTUB1 or USP2cc were added to the final concentrations of 3 μM, 10 μM and 1 μM, 
respectively. Beads were incubated at 37°C for 1 h with constant shaking and manually 
vortexed every 10 min. Beads were then centrifuged for 5 min at 1000 ×g, 60 μl of the 
supernatant were mixed with 20 μl of 4x NUPAGE LDS Sample buffer (Thermo Fisher), 
supplemented with 100 mM DTT, and incubated at 95°C for 5 min (‘Supernatant’ fraction). 
The remaining supernatant was completely removed from the beads with a syringe, and 
proteins were eluted from the beads with 60 μl of 2x NUPAGE LDS Sample buffer (Thermo 
Fisher), supplemented with 50 mM DTT, at 95°C for 15 min. Beads were centrifuged at 
1000 ×g for 3 min, and the eluate was collected (‘Elution’ fraction). The presence of 
ubiquitin chains on ubiquitylated proteins in the ‘Supernatant’ and ‘Elution’ fractions was 
analysed by SDS-PAGE and western blotting (sections 2.9.1 and 2.9.2). 
80 
 
2.13 In vitro ubiquitylation assay 
 In vitro ubiquitylation assay was performed in 1x ubiquitylation buffer supplemented 
with 100 μM ATP in a 10 μl scale. 2 μM PCNA or PCNAUb(K164), 5 μM ubiquitin, 0.1 μM 
Uba1, 0.2 μM Ubc13 or UBE2N, 0.2 μM Mms2 or UBE2V2 and 1 μM PIP-E363 were 
incubated at 30°C for 40 min. The reaction was stopped by the addition of 3 μl of 4x 
NUPAGE LDS Sample buffer (Thermo Fisher) and analysed by SDS-PAGE and western 
blotting (sections 2.9.1 and 2.9.2). 
  
81 
 
Chapter 3 
 
Results 
 
 
3.1 Basis of the study. Design of PCNA-selective linkage-specific ubiquitin 
ligases 
DNA repair mechanisms are essential to preserve the integrity of genomic 
information and to sustain life at both cellular and organismal levels. DNA damage can be 
hazardous for cells during DNA replication, as it may lead to incorrect duplication of the 
genetic information and, ultimately, cell death or the development of cancer. To minimise 
possible devastating effects of DNA lesions during S phase, cells have developed a series 
of pathways, collectively termed “DNA damage bypass”. They allow to complete DNA 
replication in the presence of DNA lesions, which can be removed in a postreplicative 
manner once replication has finished. One of the critical layers of DNA damage bypass 
regulation is posttranslational modifications of the DNA sliding clamp PCNA (Ulrich and 
Takahashi, 2013). Upon accumulation of replication intermediates due to replication over 
damaged DNA, PCNA undergoes Rad18-dependent monoubiquitylation – a signal that 
recruits translesion synthesis polymerases to the sites of DNA damage. Subsequent 
extension of this monoubiquitin into a K63-linked chain activates an error-free branch of 
damage bypass. Although K63-linked PCNA polyubiquitylation is a phenomenon known 
for over 20 years (Hoege et al., 2002; Stelter and Ulrich, 2003), a comprehensive 
understanding of the pathway is still missing. In mammalian cells, DNA translocase 
ZRANB3 has been shown to bind polyubiquitylated PCNA via a combination of its PCNA-
binding motifs (PIP and APIM) and NZF domain, selective towards K63-linked ubiquitin 
chains (Ciccia et al., 2012). Recruitment of ZRANB3 to stalled replication forks promotes 
their transformation into Holliday junctions via a process termed replication fork reversal. 
However, yeast cells lack a convincing homologue of ZRANB3, and replication fork 
reversal is a rare event unless checkpoint signalling is deactivated (Sogo et al., 2002), 
indicating the presence of another reader of PCNA polyubiquitylation. In yeast, PCNA 
polyubiquitylation is catalysed by the E3 ligase Rad5, which, together with its cognate E2 
82 
 
Ubc13/Mms2, extends monoubiquitin moiety on PCNA into a polyubiquitin chain. As Rad5 
is a multi-domain protein that also recruits TLS polymerases and has a helicase activity 
(Gallo et al., 2019; Blastyák et al., 2007), it was initially not clear whether the presence of 
a ubiquitin chain on PCNA is sufficient to activate error-free DNA damage bypass in yeast, 
or ubiquitin-independent functions of Rad5 are also essential for this pathway. 
Furthermore, the importance of linkage type in this process has not yet been 
characterised. These biological problems have been successfully solved by replacing 
Rad5 with artificial enzymes that mimic its ubiquitin ligase activity and are able to create 
not only native K63 but also other types of polyubiquitin linkages (Wegmann et al., 2022). 
The catalytic core of these artificial enzymes is based on previously identified and further 
developed in our laboratory highly linkage-selective ubiquitin ligases. These enzymes 
have been successfully designed for three chain types, namely K63, K48 and M1 (linear) 
linkages, and will be addressed further in the text as E363, E348 and E31, respectively. The 
design of these enzymes is described below and graphically presented in Figure 11A. 
 K63-linked chains are assembled by an S. cerevisiae RING-type ubiquitin ligase 
Pib1, which, together with its cognate highly K63-selective E2 Ubc13-Mms2, 
functions in the multivesicular body (MVB) pathway (Renz et al., 2020). In order to 
keep the ubiquitin ligase activity but abrogate MVB-related functions, the N-
terminal phosphatidylinositol 3-phosphate-binding FYVE domain was removed 
from the protein (Shin et al., 2001).  
 K48-linked chains are assembled by a S. cerevisiae protein Cue1 that acts as an 
activator for the highly K48-selective E2 Ubc7 (Bagola et al., 2013). The 
transmembrane domain of Cue1 was removed in order to prevent its targeting to 
ER membrane (Ravid and Hochstrasser, 2007). 
 M1-linked chains are assembled by the H. sapiens RING-Between-RING ubiquitin 
ligase HOIP, which is a component of the LUBAC complex and is known to function 
with several E2 enzymes. As HOIP is a large and multi-domain protein, only the 
catalytic RING-in-between-RING and linear chain determination (LDD) regions 
were used as an E31 (Smit et al., 2012). Additionally, as HOIP is allosterically 
activated by ubiquitin (Lechtenberg et al., 2016), two uncleavable (G76L) ubiquitin 
83 
 
moieties together with appropriate linkers were added N-terminally to the catalytic 
domain (Wegmann et al., 2022). 
In order to bring the enzymes to PCNA, a canonical PCNA-interacting motif (PIP-
box) from the mismatch repair protein Msh6 was fused N-terminally to the described 
catalytic domains of the E3s. The resulting enzymes are called PIP-E363, PIP-E348 and 
PIP-E31. To promote the nuclear localisation of PIP-E31, a nuclear localisation signal 
(NLS) was added upstream of a PIP box.  
High linkage selectivity of the tailor-made E3s implies that they only take a specific 
ubiquitin lysine as a substrate and therefore cannot initiate a chain on any protein other 
than ubiquitin (elongation specificity). Thus, Rad18-dependent PCNA monoubiquitylation 
is a prerequisite for the activity of the enzymes. By expressing the tailor-made E3s in a 
rad5 background, one can explore whether PCNA polyubiquitylation is sufficient to initiate 
error-free damage bypass and which linkages would support this pathway. Strikingly, PIP-
E363 is able to reverse the damage sensitivity of a rad5 strain (Wegmann et al., 2022). 
Although to a lesser extent, the same holds true for PIP-E31, which most likely reflects the 
similar conformation of K63- and linear polyubiquitin chains (Komander et al., 2009). On 
the contrary, expression of PIP-E348 causes a dominant negative effect: not only is error-
free bypass not supported, but also translesion synthesis is inhibited due to the 
proteasomal degradation of PCNA (Figure 11B). These results not only indicate that 
PCNA polyubiquitylation is sufficient to trigger error-free damage bypass in yeast but also 
delineate the geometric requirements for ubiquitin chains in this pathway. This experiment 
was the first example of an in vivo manipulation of ubiquitin signalling: depending on the 
linkage type, differently linked ubiquitin chains, assembled on the same acceptor site of 
the same substrate, provide different biological outputs.  
 
84 
 
 
 
Figure 11: Design and implementation of tailor-made ubiquitin ligases.  (A) Design of PIP-
E3s: domain arrangement of linkage-selective ubiquitin ligases and PIP-E3s that use their catalytic 
domains. Numbers below the protein schemes indicate amino acid positions in the protein 
sequence. (B) Application of PIP-E3s in S. cerevisiae reveals linkage requirements in DNA 
damage bypass. 
 
85 
 
  The roles of PCNA polyubiquitylation in human cells are even less well understood. 
Although this signal has been associated with ZRANB3-mediated fork reversal, an 
accumulating number of cellular pathways that depend on this modification cannot be 
explained simply assuming that ZRANB3 is the only reader of polyubiquitylated PCNA 
(Tirman et al., 2021; Thakar et al., 2020). My goal in the present work was to use PIP-E3s 
in human cells to address the following two questions: 
 What are the functional consequences of PCNA polyubiquitylation in human 
cells? 
 How important is K63 linkage in ubiquitin signalling on PCNA in human 
cells? 
In the following sections, I address these questions and discuss the optimisation 
steps that were needed to transfer the yeast PIP-E3 system into human cells, the effects 
that are caused by K63-linked PCNA polyubiquitylation, a comparison of K63 versus K48 
polyubiquitylation, the physiological relevance of PCNA polyubiquitylation and, finally, 
further development of tailor-made ubiquitin ligases, which enables their application to 
other substrates. 
  
86 
 
3.2 Optimisation of PIP-E3s in mammalian cells 
3.2.1 PIP-E363  
The original PIP-E3 constructs contain a canonical PIP-box from yeast Msh6, 
which interacts with yeast PCNA. Although PCNA is a highly conserved protein - the 
human and yeast orthologues display 35% identity and 62% similarity - I first tested 
whether the yeast PIP-E363 can polyubiquitylate human PCNA in vitro. As a substrate, I 
used PCNA pre-monoubiquitylated by UbcH5c at its native lysine K164 (Hibbert and 
Sixma, 2012). As demonstrated in Figure 12A, PIP-E363 readily polyubiquitylates 
monoubiquitylated PCNA with a similar efficiency for yeast Ubc13-Mms2 pair or their 
human homologs UBE2N-UBE2V2. In accordance with the elongation specificity of PIP-
E3s, no reaction takes place if unmodified PCNA is used as a substrate. Instead, free 
ubiquitin chains are accumulated in this condition. This justifies the use of the PIP-E3 
constructs with the original PIP box in human cells.  
 
 
Figure 12: PIP-E363 is active in vitro and in human cells. (A) In vitro ubiquitylation assay with 
PCNA and PCNA-Ub(K164) as substrates and PIP-E363 as a ligase. Yeast and human E2s were 
compared. (B) PIP-E363 was expressed in HEK 293T cells for 24 h, followed by PCNA 
immunoprecipitation from chromatin. PCNA polyubiquitylation was analysed in the 
immunoprecipitate fraction, and expression of the ligase in whole cell extracts.  
87 
 
In order to proceed with in vivo studies, a codon-optimised version of PIP-E363 was 
cloned under the control of the CMV promoter and, as a proof-of-concept, overexpressed 
in HEK 293T cells. As shown in Figure 12B, expression of PIP-E363 induces robust 
polyubiquitylation of PCNA, visualised by the anti-Ubiquitin and anti-PCNA-Ub (native 
K164 linkage) antibody after immunoprecipitation of PCNA from chromatin. Importantly, 
in striking contrast to yeast, the basal levels of PCNA monoubiquitylation, likely originating 
from endogenous DNA damage or due to replication over difficult-to-replicate regions 
(Tubbs and Nussenzweig, 2017), are sufficient to support PIP-E363 activity.  
 
3.2.2 PIP-E348 
 Unlike K63-linked, K48-linked ubiquitin chains represent a canonical signal for 
proteasomal degradation and are rapidly deconjugated upon degradation of a substrate 
(Shin et al., 2020). Despite strong genetic evidence for PIP-E348 activity in vivo, K48-linked 
chains have not been detected on yeast PCNA, likely because of their transient nature 
(Wegmann et al., 2022). To overcome this problem, His10-tagged ubiquitin was 
overexpressed in human HEK 293T cells together with codon-optimised PIP-E348. 
Subsequent Ni-NTA pulldown under fully denaturing conditions allows the identification of 
all proteins that have incorporated His-tagged ubiquitin, including PCNA. Furthermore, in 
order to boost E3 activity, I co-overexpressed PIP-E348 together with the original yeast 
Ubc7, a codon-optimised version of it (hcoUbc7) or its human homologue UBE2G2. The 
results of the experiment are shown in Figure 13A. If only His-tagged ubiquitin, but no 
PIP-E348 is transfected, the anti-PCNA antibody detects a single species, likely 
representing PCNA modified with a single His-tagged ubiquitin. Expression of PIP-E348 
results in the formation of higher molecular weight PCNA forms, likely representing K48-
linked chains. Co-expression of non-codon-optimised Ubc7 further augments chain 
formation, and co-expression of UBE2G2 results in longer ubiquitin chains. Strikingly, 
when the codon-optimised version of Ubc7 is used, I observed the disappearance of 
monoubiquitylated PCNA concomitantly with the formation of high molecular weight 
species. This may indicate that the combination of PIP-E348 with codon-optimised Ubc7 
leads to very efficient polyubiquitylation of PCNA with K48 linkages and its subsequent 
rapid degradation. As PIP-E348 is elongation-specific, only monoubiquitylated PCNA 
88 
 
should undergo polyubiquitylation and degradation, which could explain the 
disappearance of the PCNA-Ub form. These results correlate with the expression levels 
of the enzymes, shown in Figure 13B: while the original yeast Ubc7 is hardly detectable 
by western blotting, codon optimisation drastically increases its levels. The disappearance 
of the monoubiquitylated PCNA in total cell extracts upon co-expression of PIP-E348 with 
hcoUbc7 correlates with the results of the Ni-NTA pulldown. Taken together, these results 
show that, depending on the co-expressed E2, PIP-E348 may create chains of different 
lengths and intensities. The most active combination is PIP-E348 together with hcoUbc7. 
 
 
 
Figure 13: Optimisation of PIP-E348 system in mammalian cells. (A) HEK 293T cells were 
transfected with the indicated constructs and allowed to express the proteins for 48 h. His-tagged 
ubiquitin conjugates were isolated via Ni-NTA pulldown and analysed for presence of PCNA by 
western blotting. (B) Expression of the enzymes as well as PCNA ubiquitylation were analysed by 
western blotting in the lysates of cells that were used for Ni-NTA pulldown. 
 
3.2.3 PIP-E31  
Linear polyubiquitylation by PIP-E31 was investigated analogously. Due to the high 
repetitiveness and GC content of the full-length PIP-E31, I was able to optimise the codon 
usage of this construct only partially (Figure 14A). Nevertheless, I detect the activity of 
89 
 
the PIP-E31 on PCNA after performing denaturing pulldown from cells expressing His-
tagged ubiquitin (Figure 14B). Importantly, unlike K48-linked chains, the incorporation of 
a His-tagged ubiquitin into a growing M1-linked chain terminates it because its N-terminus 
is blocked by the His-tag itself. Therefore, it is not surprising to see that, when co-
expressed with His-tagged ubiquitin, PIP-E31 predominantly creates di- and tri-
ubiquitylated PCNA forms. Interestingly, even though the codon usage of this construct 
could not be fully optimised, partial optimisation still results in higher expression levels of 
the enzyme and more potent PCNA polyubiquitylation.  
 
 
 
 
Figure 14: Optimisation of PIP-E31 system in mammalian cells. (A) Schematic representation 
of the PIP-E31 enzyme. The region that was codon optimised for expression in human cells is 
encircled with a red dashed line. (B) HEK 293T cells were transfected with the indicated constructs 
and allowed to express the proteins for 48 h. His-tagged ubiquitin conjugates were isolated via Ni-
NTA pulldown and analysed for presence of PCNA by western blotting (left). Expression of the 
enzymes was analysed by western blotting in total lysates from the cells that were used for the 
Ni-NTA pulldowns. 
  
90 
 
3.3 Excessive K63-linked PCNA polyubiquitylation results in replication 
catastrophe  
3.3.1 Creation of a cell line with inducible expression of PIP-E363 
Up until now, the DNA damage field has accumulated a lot of data that originate 
from different model organisms and, particularly, from different human cell lines. As 
mutations in DNA damage response pathways are one of the key mechanisms employed 
by cancer cells to achieve immortality, the genetic background of a cell line should be 
taken into account when interpreting experimental results (Erenpreisa and Cragg, 2013). 
Importantly, some DNA damage and replication stress-related phenotypes are cell line-
specific, and the reasons for this specificity remain unclear (Liu et al., 2020). As a starting 
point, I chose a human retinal epithelial cell line RPE1 hTERT for two main reasons. First, 
this cell line is diploid, chromosomally stable and untransformed, immortality being 
achieved via constitutive telomerase expression.  Second, PCNA polyubiquitylation has 
been shown to be essential for fork protection in these cells (Thakar et al., 2020). Data 
from the Ulrich laboratory, as well as others (Igoucheva et al., 2006), suggest that the 
introduction of exogenous double-stranded DNA in mammalian cells via transient 
transfection leads to activation of DNA damage response, which may be the consequence 
of lesions present in the transfected DNA. Therefore, in order to carefully analyse the 
impact of K63-linked PCNA polyubiquitylation on damage signalling, PIP-E363 was stably 
integrated in RPE1 hTERT FlpIn cells via FRT recombination (Figure 15A). In the 
resulting cell line, hereafter called RPE1 hTERT PIP-E363, the open reading frame of the 
ligase was integrated into the genome and placed under the control of a doxycycline-
inducible promoter. In this system, two copies of a TetO sequence are placed immediately 
downstream of a CMV promoter and, in the absence of doxycycline, are tightly bound by 
the TetR protein expressed in the host line. The addition of doxycycline releases TetR 
from the DNA, allowing transcription from the CMV promoter. As the initial PIP-E363 
construct contained a VSV tag at the N-terminus and the vector for FlpIn recombination 
provided a Flag tag, the resulting protein expressed in the RPE1 hTERT PIP-E363 cell line 
carried both tags. As expected, the addition of doxycycline induced robust expression of 
PIP-E363, detectable by western blotting with the anti-Flag or anti-VSV antibodies (Figure 
15B). Furthermore, I observed that expression of the construct begins as early as 2 h after 
91 
 
the addition of doxycycline and gradually increases over time (Figure 15C). To analyse 
whether PIP-E363 expression level, arising from a single-copy integration, is sufficient to 
induce PCNA polyubiquitylation, I analysed the presence of ubiquitin chains on PCNA 
after 24 h expression of the enzyme by means of PCNA pulldown from chromatin. Albeit 
less efficient than after a transient overexpression of PIP-E363 in the HEK 293T cell line, 
PCNA polyubiquitylation was also induced by adding doxycycline to RPE1 hTERT PIP-
E363 cells (Figure 15D). Importantly, when PCNA monoubiquitylation was enhanced by 
UV light treatment, the amount of PIP-E363-generated chains increased correspondingly. 
I also observe a decrease in monoubiquitylated PCNA levels after expression of the 
ligase, consistent with the conversion of PCNA-Ub into polyubiquitylated forms. These 
data justify the use of an RPE1 hTERT PIP-E363 cell line as a model system to study 
PCNA polyubiquitylation. 
92 
 
 
 
Figure 15: Creation and characterization of an RPE1 hTERT PIP-E363 cell line. (A) Schematic 
representation of cell line generation based on FRT recombination. (B) Indicated cell lines were 
treated with 1 μg/ml doxycycline for 24 h and expression of PIP-E363 was analysed by western 
blotting. Asterisks indicate signal remaining from Flag or VSV detection. (C) RPE1 hTERT PIP-
E363 cells were treated with 10 nM or 100 nM doxycycline for the indicated times, followed by 
western blotting. (D) RPE1 hTERT PIP-E363 cells were treated with 1 μg/ml for 24 h and, if 
indicated, irradiated with 40 J/m2 4h prior to cell harvesting. After chromatin fractionation and lysis, 
PCNA was immunoprecipitated from the lysates and analysed by western blotting with the 
indicated antibodies. 
93 
 
3.3.2 PIP-E363-expressing cells activate checkpoint signalling and exhibit S phase 
arrest 
A common hallmark of cells experiencing replication stress or DNA damage is the 
activation of checkpoint signalling. This involves hyperactivation of certain kinases, 
typically ATR, ATM and DNA-PK, characterised by increased self-phosphorylation and 
phosphorylation of substrates such as CHK1, CHK2, RPA, H2AX and many others (Lanz 
et al., 2019). PCNA polyubiquitylation is also a cellular response to replication stress; 
however, it is not yet known how a cell would react to this modification if no DNA damage 
is present. To understand whether polyubiquitylation of PCNA without exposing cells to 
DNA-damaging agents also induces checkpoint activation in cells, we first analysed the 
phosphorylation status of the checkpoint kinase CHK1. Surprisingly, we observe strong 
activation of CHK1 phosphorylation, starting from 6 h post-PIP-E363 induction (Figure 
16A). Interestingly, while phosphorylated CHK1 peaks at early time points, it almost 
disappears after 24 h, together with a drop in total CHK1 levels. As phosphorylation of 
CHK1 has been shown to be a destabilizing modification (You-Wei Zhang et al., 2005), 
the results suggest that excessive and persistent CHK1 phosphorylation during extensive 
PIP-E363 expression leads to substantial degradation of this kinase. We also observe 
phosphorylation of the histone variant H2AX at serine 139 (γH2AX), which is a common 
marker for the presence of DNA double-strand breaks. Unlike phospho-CHK1, we do not 
see an increase in γH2AX at the early time points but rather late, suggesting that DNA 
double-strand breaks occur as a consequence of long-term replication stress induced by 
PIP-E363. As CHK1 phosphorylation is carried out by ATR, which is typically activated 
upon persistent exposure of single-stranded DNA, I analysed RPA phosphorylation at S33 
and T21 as a proxy for the presence of single-stranded DNA. Indeed, both phosphorylated 
RPA forms appear after the expression of PIP-E363, and the kinetics of this modification 
is more similar to that of phospho-CHK1 than γH2AX (Figure 16B). Interestingly, the 
extent of checkpoint activation is comparable or even higher, as for the 24 h time point, 
than that induced by 40J/m2 UV light, which is considered a high dose of DNA damage. 
Taken together, these data suggest that expression of PIP-E363 leads to exposure of 
single-stranded DNA and checkpoint activation, ultimately resulting in DNA double-strand 
breaks. 
94 
 
As activation of checkpoint signalling slows down cell cycle progression, I analysed 
the DNA content of RPE1 hTERT cells expressing PIP-E363 for 6–48 h. As shown in 
Figure 16C, expression of PIP-E363 leads to a strong early S-phase arrest: the effect is 
already noticeable 12 h after the addition of doxycycline and becomes stronger at the later 
time points.  
 
 
 
Figure 16: PCNA polyubiquitylation induces checkpoint activation and cell cycle arrest. (A) 
RPE1 hTERT PIP-E363 cells were treated with 1 μg/ml doxycycline for the indicated time periods 
and, if indicated, irradiated with 40 J/m2 4h prior to cell harvesting. Activation of checkpoint 
signalling was analysed by western blotting with indicated antibodies. Ponceau S staining serves 
as a loading control. (B) RPA phosphorylation was analysed by western blotting for cells treated 
as in (A). (C) RPE1 hTERT PIP-E363 cells were treated as in (A), followed by cell cycle analysis 
by flow cytometry. 
 
95 
 
The significant accumulation of PIP-E363-expressing cells in the S phase suggests 
they struggle to complete replication. This phenomenon could be caused by lower total 
replication speed, which can be monitored by nucleotide incorporation. 2D cell cycle 
analysis reveals a robust decrease in the incorporation of the thymidine analogue EdU 
24 h and, even stronger, 48 h after the expression of PIP-E363 (Figure 17A). This 
indicates that PIP-E363-expressing cells require more time to complete replication, thus 
explaining their accumulation in the S phase. At the same time, lower EdU incorporation 
for cells in the S phase can have two main reasons. First – the number of replication forks 
remains unchanged, but their velocity decreases. Second – the speed of individual forks 
per se is unaffected, but the total number of active replication forks decreases. The second 
scenario is less likely, as the general response to replication stress involves excessive 
firing of new origins, leading to more replication forks being active at a given time (Courtot 
et al., 2018). Nevertheless, one cannot exclude that PCNA polyubiquitylation may 
permanently stall replication forks, decreasing the number of moving replisomes. In order 
to discriminate between these two scenarios, I performed a DNA fibre assay, which allows 
monitoring of replication speed at the level of individual replication forks. This assay 
involves the treatment of the cells with two different thymidine analogues, CldU and IdU, 
followed by cell lysis and spreading of the DNA on a glass slide. Subsequent 
immunofluorescent detection of CldU and IdU allows direct visualisation of replication 
tracts and analysis of fork speed in given conditions. Importantly, in this assay, ongoing 
replication results in two-colour tracts, whereas one-colour tracts may represent initiation 
and termination of replication (depending on the colour) or fork stalling. As shown in 
Figure 17B, expression of PIP-E363 for 12 h results in a massive decrease in replication 
fork speed, indicating that the first scenario, namely reduced replication speed at the level 
of individual forks, is responsible for the cell cycle arrest. 
 
96 
 
 
Figure 17: Lower EdU incorporation and replication speed in PIP-E363-expressing cells. (A) 
RPE1 hTERT PIP-E363 cells were mock-treated or treated with 1 μg/ml doxycycline for 24 h or 48 
h, followed by a treatment with 10 μM EdU for 30 min. Levels of incorporated EdU as well as total 
DNA content were analysed by flow cytometry. (B) DNA fibre assay was performed in RPE1 
hTERT PIP-E363 cells that were mock-treated or treated with 1 μg/ml doxycycline for 12 h. 
Representative images and quantification of total tract lengths are shown. Scale bar corresponds 
to 30 μm. ****p<0.0001 (Mann-Whitney U test). 
 
3.3.3 ATR inhibition exacerbates effects of PIP-E363 
The above-described data demonstrate that K63-linked PCNA polyubiquitylation 
by PIP-E363 leads to dramatic consequences for cells: intra-S checkpoint signalling is 
activated, fork speed decreases, and cells arrest in the S phase. This persistent replication 
arrest ultimately results in the formation of double-strand breaks. It is known that activation 
of checkpoint signalling per se even in the absence of real DNA damage may inhibit cell 
cycle progression (Soutoglou and Misteli, 2008). Therefore, I aimed to understand 
whether inhibition of checkpoint signalling would prevent any long-term adverse effects of 
PIP-E363 expression. Phosphorylation of CHK1 and RPA at S33 upon expression of PIP-
E363 suggests that ATR is one of the kinases responsible for checkpoint generation. 
Therefore, I reasoned that if ATR activity is the reason for replication collapse after PIP-
E363 expression, inhibition of this kinase should prevent the negative effects of PCNA 
polyubiquitylation, including DSB formation. As shown in Figure 18, the addition of the 
97 
 
ATR inhibitor VE-821 essentially prevents CHK1 phosphorylation after PIP-E363 
expression. Phospho-RPA T21 levels are partially reduced, and phospho-RPA S4-S8 
levels are unaffected, consistent with the known contribution of other checkpoint kinases 
to the phosphorylation of these residues (Liu et al., 2012). Strikingly, I observe a heavy 
increase in a γH2AX signal in the conditions where ATR is inhibited on top of PIP-E363 
expression. As H2AX phosphorylation is a canonical marker of DNA double-strand 
breaks, this indicates that ATR activity prevents DNA breakage after prolonged PCNA 
polyubiquitylation. In this respect, the expression of PIP-E363 is similar to the treatment of 
cells with the replication stress-inducing agent hydroxyurea (HU). Inhibition of ATR leads 
to excessive firing of new origins and RPA exhaustion, leading to the formation of single-
stranded DNA that is not coated by RPA and therefore prone to breakage (Toledo et al., 
2013). ATR has a second known activity that prevents the formation of breaks during HU 
treatment: it phosphorylates SMARCAL1, preventing excessive fork reversal and SLX4-
dependent cleavage of reversed forks (Couch et al., 2013). It is likely that one of these 
mechanisms is responsible for the observed PIP-E363-induced γH2AX signalling upon 
ATR inhibition. 
 
Figure 18: ATR prevents DSB formation upon excessive PCNA polyubiquitylation. RPE1 
hTERT PIP-E363 cells were mock-treated or treated with 1 μg/ml doxycycline or 10 μM VE-821 for 
the indicated time periods. Checkpoint signalling was analysed by western blotting. Ponceau S 
staining serves as a loading control.  
98 
 
3.3.4 Effects of PIP-E363 expression are dependent on its interaction with PCNA 
and catalytic activity 
To prove that the observed phenotypes are not the result of potential off-target 
activity of PIP-E363, I created a panel of mutants of this ligase: 1) lacking the PIP-box 
(ΔPIP or, alternatively, E363), 2) containing the point mutation I227A, which has been 
shown to abolish interaction with Ubc13 in yeast (Renz et al., 2020) or 3) completely 
lacking the RING domain (ΔRING) (Figure 19A). I created RPE1 hTERT-based cell lines 
that express these mutants of PIP-E363 and compared them to the cell line expressing the 
original PIP-E363. As shown in Figure 19B, interfering with either PCNA binding or with 
the catalytic activity of E3 results in no checkpoint activation, although all constructs, 
except for the probably unstable PIP-E363ΔRING, were expressed at similar levels. 
Interestingly, I227A reduces but does not completely abolish the activity of PIP-E363 – 
nevertheless, this is sufficient to completely abrogate checkpoint signalling. Moreover, the 
expression of none of these constructs interferes with cell cycle progression either (Figure 
19C). Therefore, this means that both the interaction with PCNA and the catalytic activity 
of PIP-E363 are important to induce replication catastrophe in cells. 
99 
 
 
Figure 19: Negative effects of PIP-E363 require interaction with PCNA and catalytic activity. 
(A) Scheme of the PIP-E363 enzyme and its mutant forms: lacking a PIP box, containing an I227A 
mutation or lacking a RING domain. (B) RPE1 hTERT cells carrying doxycycline-inducible PIP-
E363 variants described in panel (A) were treated with 1 μg/ml doxycycline for 6 h. Expression of 
enzymes and checkpoint activation were analysed by western blotting. Ponceau S staining serves 
as a loading control. (C) Cell cycle profiles of cells treated as in (B) but for 24 h.  
100 
 
3.3.5 PCNA monoubiquitylation is a prerequisite for PIP-E363 activity in vivo 
The findings described earlier show that replication stress induced by PIP-E363 
results from its ubiquitin ligase activity in the vicinity of PCNA. However, there is a 
possibility that our tailor-made E3 extends ubiquitin(s) present on substrates other than 
PCNA. To test this hypothesis, I integrated PIP-E363 by means of lentiviral transduction in 
an RPE1 hTERT K164R (KR) mutant cell line, in which K164 of PCNA is mutated to 
arginine and therefore cannot be ubiquitylated by RAD18 (Thakar et al., 2020). As a 
control, the corresponding parental RPE1 hTERT (WT) cell line was used. Together with 
PIP-E363, TetR was integrated in the same cells in order to make expression of the ligase 
doxycycline-inducible. Additionally, to confirm that the effects are not cell line-specific, I 
also used the following panel of HEK 293T-derived cells (Thakar et al., 2020):  
1. Wild-type HEK293 (WT) 
2. Heterozygous PCNA K164R, in which other PCNA alleles were inactivated.  
The loss of PCNA levels was compensated by re-expressing wild-type 
PCNA (WT/KR). 
3. Heterozygous PCNA K164R mutant, in which other PCNA alleles were 
inactivated.  Loss of PCNA levels was compensated by re-expressing 
mutant PCNA (KR/KR). 
Figure 20A shows the expression of PIP-E363 and activation of the checkpoint after 
its integration in the described cell lines. The RPE1 hTERT FlpIn PIP-E363 cell line serves 
as a positive control (here, the ligase is ~2 kDa heavier due to the presence of an 
additional Flag tag). Despite having a higher expression of PIP-E363, the cell lines that 
harbour the PCNA K164R mutation do not activate checkpoint signalling after expression 
of the ligase, as demonstrated by the absence of phospho-RPA T21 or γH2AX. In order 
to show that the K164R mutation abolishes PIP-E363-induced polyubiquitylation, PCNA 
was immunoprecipitated from these cell lines after the expression of PIP-E363. In 
accordance with PIP-E3s being elongation-specific, polyubiquitylation of PCNA by PIP-
E363 is completely abolished when the acceptor lysine is not available (Figure 20B). 
Interestingly, compared to homozygous wild-type cells, we do not observe a decrease in 
PCNA polyubiquitylation in HEK 293T WT/KR cells. This may be due to the fact that 
exogenous PCNA is overexpressed in the WT/KR cell line, leading to slightly higher total 
101 
 
PCNA levels and may ultimately interfere with the pathway. Alternatively, it is possible that 
checkpoint activation by PIP-E363 requires all three subunits of PCNA to be 
simultaneously monoubiquitylated, which is less likely to be the case in the WT/KR cell 
line. It has been reported that such simultaneous ubiquitylation of PCNA subunits acts as 
a switch for the HLTF mode of action (Masuda et al., 2018). It remains to be understood 
what the underlying molecular mechanism of HLTF activity towards the fully and partially 
monoubiquitylated PCNA trimers is and whether it can also be applied to the PIP-E363 
system. 
Consistent with no checkpoint signalling being active in K164R cells, only wild-type, 
but not mutant RPE1 hTERT cells exhibit cell cycle arrest after PIP-E363 expression 
(Figure 20C). Interestingly, HEK 293T maintain the normal cell cycle progression even 
when the checkpoint is active, highlighting that the genetic background of a cell line may 
influence its reaction to replication stress and checkpoint signalling (Figure 20D). 
To further prove that PIP-E363-induced effects are specific for PCNA, the E3 was 
expressed in cells transiently depleted of Rad18 by means of two independent siRNAs. 
S-phase arrest (Figure 21A, 11B) and robust checkpoint activation (Figure 21C) were 
observed only in cells, transfected with control siRNA and were absent upon knock-down 
of Rad18. Of note, siRNA #1 led to a moderate decrease in expression levels of PIP-E363; 
however, this is not the case for siRNA #2 and, altogether, these results place Rad18 
upstream of PIP-E363 in the PCNA ubiquitylation cascade. 
 
102 
 
 
 
Figure 20: The negative effects of PIP-E363 are dependent on lysine 164 of PCNA. (A) 
Indicated cell lines with integrated PIP-E363 were treated 1 μg/ml doxycycline for 24 h. Expression 
of the E3, PCNA levels and checkpoint activation were analysed by western blotting of total cell 
lysates with the indicated antibodies. Ponceau S staining serves as a loading control.  (B) 
Indicated cell lines with integrated PIP-E363 were treated as in panel (A), followed by chromatin 
isolation, PCNA immunoprecipitation and western blotting with indicated antibodies. (C) Cell cycle 
profiles of wild-type and PCNA K164R RPE1 hTERT TetOn PIP-E363 cells, treated with 1 μg/ml 
doxycycline for 24 h. (D) Cell cycle profiles of wild-type, heterozygous and homozygous PCNA 
K164R mutant HEK 293T TetOn PIP-E363 cells, treated with 1 μg/ml doxycycline for 24 h. 
103 
 
 
 
 
Figure 21: The negative effects of PIP-E363 are dependent of RAD18 activity. (A) Cell cycle 
profiles of RPE1 hTERT PIP-E363 cells, transfected with control siRNA or 2 independent siRNAs 
against RAD18 for 72 h and treated with doxycycline during the last 12 h or 24 h. (B) Quantification 
of cell cycle distribution of cells treated as described in (A). (C) RPE1 hTERT PIP-E363 cells were 
treated as described in (A), followed by lysis and western blotting with the indicated antibodies. 
Ponceau S staining serves as a loading control. 
 
104 
 
3.3.6 Negative effects of PIP-E363 are independent of known interactors of 
polyubiquitylated PCNA 
To date, two proteins have been reported to interact with polyubiquitylated PCNA. 
One is the DNA translocase ZRANB3, which catalyses the reversal of replication forks. 
Therefore, we hypothesised that excessive fork reversal upon PIP-E363 overexpression 
may cause activation of checkpoint signalling. However, I observed that depletion of 
ZRANB3 with 2 independent siRNAs does not abolish PIP-E363-induced checkpoint 
activation (Figure 22A), indicating that the negative effects of the ligase expression are 
likely ZRANB3-independent. ZRANB3 catalyses fork reversal with at least two other DNA 
translocases – HLTF and SMARCAL1. Although a link between HLTF and SMARCAL1 
recruitment to stalled replication forks and PCNA ubiquitylation, as well as a possible 
functional redundancy between all three translocases, have not been reported, I tested 
whether depletion of HLTF and SMARCAL1 would counteract the negative effects of PIP-
E363. Similar to ZRANB3, depletion of HLTF or SMARCAL1 does not significantly change 
checkpoint activation upon PCNA polyubiquitylation (Figure 22B).  
As the factors that must act downstream of PCNA polyubiquitylation in the PIP-
E363 system do not have to bind to PCNA itself and can merely be general readers of K63-
linked chains on chromatin, I hypothesised that the responsible factor could be RAP80. 
This protein is involved in double-strand break repair by binding to K63-linked chains on 
histones and recruiting the BRCA1-A complex to damage sites to counteract resection 
(Hu et al., 2011). However, no change in checkpoint activation is observed when RAP80 
is depleted (Figure 22C). The same holds for the second known interactor of 
polyubiquitylated PCNA, WRNIP1 (Figure 22D). Thus, the phenotypes observed in the 
PIP-E363 system suggest the presence of other readers of PCNA polyubiquitylation. 
105 
 
 
 
 
 
Figure 22: The negative effects of PIP-E363 are independent of fork reversal enzymes, 
RAP80 and WRNIP1. RPE1 hTERT PIP-E363 cells were transfected with siRNAs against ZRANB3 
(A), HLTF (B), SMARCAL1 (B), RAP80 (C) and WRNIP1 (D) for 72 h and treated with doxycycline 
for 12 h or 24 h. Expression of PIP-E363 and levels of RPA phosphorylation were analysed in total 
cell lysates by western blotting. Ponceau S staining serves as a loading control. 
  
106 
 
3.4 Analysing the interactome of polyubiquitylated PCNA 
 
Given that the depletion of putative interactors of polyubiquitylated PCNA did not 
lead to the rescue of PIP-E363–induced checkpoint activation, I set out to identify 
interactors of polyubiquitylated PCNA in an unbiased manner in vivo by mass-
spectrometry. In order to prove that pulling down PCNA in my experimental setup 
preserves binding of PCNA to its known interactors, I analysed whether ZRANB3, HLTF, 
EXO1 and POLH, all of which have been reported to directly interact with PCNA (Ciccia 
et al., 2012; Seelinger and Otterlei, 2020; Chen et al., 2013; Bienko et al., 2005), co-
immunoprecipitate with PCNA. I also compared whether stable or transient 
overexpression of PIP-E363 influences these interactions.  
 
 
Figure 23: Immunoprecipitation of chromatin-bound PCNA preserves its interactions with 
a panel of known PCNA binders. PIP-E363 was expressed either in HEK 293T PIP-E363 cells by 
addition of 1 μg/ml doxycycline for 24 h or via a transient overexpression of PIP-E363 in wild-type 
HEK 293T cells, followed by chromatin isolation and immunoprecipitation of PCNA. The levels of 
PCNA ubiquitylation as well as co-immunoprecipitated ZRANB3, HLTF, EXO1 and POLH were 
analysed by western blotting.  
107 
 
As shown in Figure 23, all four factors bind to PCNA. However, these interactions 
seem to be independent of the ubiquitylation status of PCNA. I also observe that transient 
overexpression of the enzyme leads to more intense PCNA polyubiquitylation, justifying 
the use of this method for further MS-based experiments. 
To analyse the interactome of polyubiquitylated PCNA in mammalian cells, I 
performed SILAC labelling in HEK 293T cells with heavy or light amino acids, followed by 
transient overexpression of PIP-E363 or, as a control, PIP-E363ΔRING (Figure 24A). The 
experiment was performed in triplicates: twice PIP-E363 was overexpressed in heavy-, and 
PIP-E363ΔRING in light-labelled cells, and for the third replicate, a label switch was 
performed. Preparation of samples for mass-spectrometry, as well as subsequent data 
analysis, was carried out by Ivan Mikicic (Petra Beli laboratory). Figure 24B shows the 
results of this experiment as a volcano plot; factors that preferentially interact with 
polyubiquitylated PCNA are shown on the right side. Importantly, we observe WRNIP1 as 
one of the top interactors. ZRANB3 peptides, on the contrary, were not identified in this 
experiment, probably due to the transient nature of its interaction with polyubiquitylated 
PCNA or low expression levels. Somewhat unexpectedly, I observed the preferential 
binding of VCP (valosin-containing protein, also known as p97 or Cdc48 in yeast) and the 
proteasomal subunit PSMD4 to polyubiquitylated PCNA. VCP is an ATP-dependent 
unfoldase that extracts polyubiquitylated proteins from chromatin and other cellular 
compartments (Meyer and Weihl, 2014). VCP is targeted to its substrates via a set of 
adaptor proteins (Meyer and van den Boom, 2023); however, a high preference of VCP 
for K48-linked and branched, but not K63-linked chains has been reported (Yau et al., 
2017; Lange et al., 2023). PSMD4 is a ubiquitin receptor of the proteasome, which has 
been shown to bind both K63 and K48-linked polyubiquitin with a marked preference for 
longer ubiquitin chains (Wang et al., 2005). Factors that preferentially bind to 
polyubiquitylated PCNA in the aforementioned interactome analysis also include the 
ubiquitin ligases HUWE1 and UBR5, both of which have been shown to bind K63-linked 
ubiquitin chains and to be involved in DNA damage repair pathways in mammalian cells 
(Ohtake et al., 2016; Choe et al., 2016; Cipolla et al., 2019).  
Table 9 contains a list of proteins identified in the interactome analysis that have 
been implicated either in DNA damage response or in ubiquitin-mediated transactions.  
108 
 
 
 
Figure 24: Analysing the interactome of polyubiquitylated PCNA. (A) MS-based experimental 
approach to identify interactors of polyubiquitylated PCNA. (B) Volcano plot of the results of the 
experiment. Factors that preferentially interact with polyubiquitylated PCNA are highlighted in red. 
 
Table 9: Selected interactors of polyubiquitylated PCNA, identified in the MS experiment 
Protein name Fold change  Brief description 
PCNA-Ubn/PCNA 
HUWE1 1.5763 HECT-type ubiquitin ligase. Interacts with 
PCNA (Choe et al., 2016), as well as K63-
linked polyubiquitin, assembles K48 linkages 
on top of K63-linked chains (Ohtake et al., 
2016). 
PSMD4 1.3209 Ubiquitin receptor subunit of the 26S 
proteasome. Prefers long ubiquitin chains 
(Wang et al., 2005). 
FBXL12 1.3202 Substrate-recognition component of the SCF 
E3 ubiquitin ligase complex. Mediates 
ubiquitylation of Ku80 at damage sites (Postow 
and Funabiki 2013). 
WRNIP1 1.2845 Interacts with polyubiquitylated PCNA (Saugar 
et al., 2012; Kanu et al., 2016). Involved in 
replication fork protection (Leuzzi et al., 2016; 
Porebski et al., 2019) and intra-strand crosslink 
repair (Socha et al., 2020). 
109 
 
VCP 1.2697 VCP/p97 extracts polyubiquitylated proteins 
from membranes or cellular structures (Meyer 
and van den Boom 2023). Is involved in many 
pathways that safeguard genome stability (Vaz 
et al., 2013) 
ERBB2 1.2715 Receptor tyrosine kinase often overexpressed 
in breast cancers. Modulates repair of certain 
DNA lesions produced by chemotherapy 
(Boone et al., 2009) 
H1.3 1.2573 Linker histone. Involved in chromatin 
compaction into higher order structures 
(Hergeth and Schneider 2015) 
UBR5 1.2224 HECT-type ubiquitin ligase. Interacts with the 
components of replication fork and controls the 
levels of histone H2A monoubiquitylation 
(Cipolla et al., 2019). Limits spreading of 
chromatin ubiquitylation (Gudjonsson et al., 
2012). Forms branched K63-K48 ubiquitin 
chains (Ohtake et al., 2018). 
 
 Thus, in the interactome of polyubiquitylated PCNA, I identified several factors that 
have not been previously known to be associated with this modification. In the following 
sections, I will analyse several of these factors to unravel the pathway triggered by 
excessive K63-linked PCNA polyubiquitylation. 
 
 
 
 
  
110 
 
3.5 K63-linked chains on PCNA are converted into conjugates of higher 
complexity in vivo 
3.5.1 VCP activity underlies PIP-E363-induced replication collapse 
In the interactome of polyubiquitylated PCNA, we observe components of the VCP 
– proteasome pathway. This is a common mechanism of extraction and degradation of 
polyubiquitylated substrates; however, the relevance of ubiquitin chain linkage in this 
context remains underinvestigated. Therefore, I hypothesised that VCP might be one of 
the factors that mediated checkpoint activation following the expression of PIP-E363. Since 
VCP is an essential protein and its knockdown can lead to heavy alterations of cellular 
metabolism, I set out to inhibit its activity via allosteric inhibition with NMS-873 (hereafter 
referred to as VCPi) (Magnaghi et al., 2013). Regardless of how long either PIP-E363 is 
expressed, or VCP is inhibited, blocking VCP activity prevents checkpoint activation by 
PIP-E363 (Figure 25A). As inhibition of VCP for 12 h leads to a reduction of the ligase 
levels, probably due to a toxic effect of the compound, only shorter (6 h) times of VCP 
inhibition are used for future experiments. Consistent with the role of VCP in the turnover 
of ubiquitylated proteins, VCP inhibition leads to the accumulation of a high molecular 
weight ubiquitin smear. Yet, interestingly, the pool of free ubiquitin seems to be unaffected 
by this treatment. Concomitantly with an abrogation of checkpoint signalling, VCP 
inhibition leads to the accumulation of polyubiquitylated PCNA on the chromatin (Figure 
25B). I also observe that long polyubiquitin conjugates on PCNA are more efficiently 
extracted from chromatin than shorter forms, as demonstrated by stronger enrichment of 
high molecular weight smear upon VCP inhibition.  
111 
 
 
Figure 25: VCP activity mediates the toxic effects of PIP-E363 expression. (A) RPE1 hTERT 
PIP-E363 cells were treated with 1 μg/ml doxycycline or 5 μM NMS-873 (VCPi) for the indicated 
time periods, followed by western blotting with the indicated antibodies. (B)  RPE1 hTERT PIP-
E363 cells were treated as described in (A), followed by chromatin isolation, PCNA 
immunoprecipitation and western blotting with the indicated antibodies. (C) RPE1 hTERT PIP-
E363 cells were labelled with 10 μM BrdU for 24 h, treated as described in panel (A) followed by 
immunofluorescence staining against BrdU in non-denaturing conditions. Images represent an 
overlay of DAPI (blue) and BrdU (green) signals. The scale bar corresponds to 30 μm. (D) 
Quantification of the experiment described in (C). Graph represents the number of nuclear foci per 
nucleus in cells that have at least one nuclear focus. *p<0.05; ***p<0.001 (Mann-Whitney test). 
 
112 
 
 Expression of PIP-E363 induces replication catastrophe, which is characterised by 
activation of checkpoint signalling, exposure of single-stranded DNA and cell cycle arrest. 
To exclude the possibility that VCP inhibition acts similarly to ATR inhibition - reducing the 
checkpoint signalling without affecting the other adverse effects of PIP-E363 expression - 
I analysed the exposure of single-stranded DNA upon expression of PIP-E363 and 
inhibition of VCP. This was achieved by labelling the cells with BrdU and performing 
immunofluorescence staining in non-denaturing conditions: only BrdU carried within 
single-stranded DNA can be visualised by this method. As shown in Figure 25C, D, 
expression of PIP-E363 induces massive exposure of single-stranded DNA; however, this 
is largely prevented by VCP inhibition. These data indicate that multiple negative effects 
of PIP-E363 overexpression rely on the removal of polyubiquitylated PCNA from chromatin 
by VCP. 
 Next, I wanted to understand whether the observed activity of VCP towards 
polyubiquitylated PCNA is specific to the PIP-E363 system or is also present when PCNA 
polyubiquitylation is induced by replication stress. Therefore, I induced replication stress 
in wild-type RPE1 hTERT cells by means of three DNA-damaging agents – HU, MMS and 
UV, and analysed the impact of VCP inhibition on PCNA polyubiquitylation in these 
conditions. All three genotoxins create polymerase-stalling base lesions, and HU 
additionally leads to fork stalling due to the depletion of the free nucleotide pool. As shown 
in Figure 26A, PCNA polyubiquitylation is induced by all three treatments. Inhibition of 
VCP increases the intensity of chromatin-bound PCNA polyubiquitylation, the effect being 
the strongest for HU and MMS. To make sure that all ubiquitin-reactive species observed 
in the PCNA IP/western blot depend on PCNA K164 monoubiquitylation and do not 
correspond to de novo PCNA ubiquitylation events on other residues, I analysed PCNA 
ubiquitylation in the RPE1 hTERT PCNA K164R cell line in the presence HU and VCPi. 
As demonstrated in Figure 26B, in RPE1 hTERT PCNA K164R cells, PCNA gets neither 
mono- nor polyubiquitylated. These results highlight the similarity between the PIP-E363 
system and the replication stress response and indicate that VCP-mediated extraction of 
PCNA is a common pathway downstream of PCNA polyubiquitylation.  
113 
 
 
Figure 26: VCP extracts polyubiquitylated PCNA from chromatin under replication stress 
conditions. (A) RPE1 hTERT cells were left untreated, treated with 4 mM HU for 4,5h, 0.01% 
MMS for 1 h followed by 4,5h recovery or 40J/m2 UV followed by 4,5h recovery. During recovery 
or HU treatment, 5 μM VCPi (NMS-873) or the corresponding amount of DMSO were also added. 
PCNA polyubiquitylation was analysed by immunoprecipitation from the chromatin fraction and 
western blotting. (B) HU-induced PCNA polyubiquitylation in RPE1 hTERT (WT) or RPE1 hTERT 
PCNA K164R (KR) was analysed as in panel (A). 
 
3.5.2 Analysing the interactome of polyubiquitylated PCNA upon VCP inhibition 
Up until now, it remains unclear what happens to the interactors of polyubiquitylated 
proteins upon their removal by VCP from different cellular compartments. If 
polyubiquitylated PCNA is removed from chromatin, its binders will likely leave it as well. 
Therefore, some interactions could be lost in our MS-based experiment. Also, although 
K48 and K11 linkages were identified in the interactome, they showed no enrichment after 
PCNA polyubiquitylation. As these linkage types are well-characterised substrates for 
VCP-dependent extraction, VCP inhibition would be necessary in order to analyse their 
presence on chromatin upon polyubiquitylation of PCNA. For that reason, I performed an 
114 
 
analogous mass spectrometry-based experiment as described in section 3.4, except that 
VCP was inhibited for 6 h both during PIP-E363ΔRING or PIP-E363 expression. 
Preparation of samples for mass-spectrometry analysis as well as subsequent data 
analysis was carried out by Ivan Mikicic (Petra Beli laboratory). Figure 27A shows the 
results of the experiment as a volcano plot, with factors preferentially interacting with 
PCNA upon PIP-E363 expression positioned on the right half of the plot. I reproducibly 
observe WRNIP1 and HUWE1 as interactors of polyubiquitylated PCNA. The DNA 
translocase ERCC6L (PICH), which has been shown to catalyse processive fork reversal 
(Tian et al., 2021), is also enriched upon PCNA polyubiquitylation. No other proteins with 
known functions in DNA replication or repair were enriched in the condition of PIP-E363 
expression. Interestingly, analysis of the ubiquitin linkages in the described experiment 
reveals an increase not only of K63 but also of K48 and K11 linkages after K63-linked 
PCNA polyubiquitylation (Figure 27B). In the following sections, I will discuss the origin 
of these linkages and their impact on VCP-dependent PCNA extraction. 
 
 
Figure 27: Interactome of polyubiquitylated PCNA upon VCP inhibition. (A) Volcano plot 
showing interactors of PCNA upon expression of PIP-E363ΔRING (left) or PIP-E363 after 6 h 
treatment with VCPi (NMS-873). Proteins with known functions in replication or DNA repair are 
highlighted in yellow. (B) Enrichment of ubiquitin linkages in the conditions described in panel (A).   
 
115 
 
3.5.3 K63-linked chains on PCNA are converted into branched K63-K48 conjugates 
3.5.3.1 UBICREST assay 
It was intriguing to observe that polyubiquitylation of PCNA with K63-linked chains 
leads to an increase of K48- and K11-linked ubiquitin chains that co-immunoprecipitate 
with PCNA. As immunoprecipitation of PCNA is performed under non-denaturing 
conditions, there are three scenarios that can explain these results (Figure 28): 
1. K48 and K11 linkages are assembled on top of K63-linked chains (branched 
ubiquitin chains) 
2. K63-linked chain remains intact. However, PCNA lysines other than K164 undergo 
K48- or K11-linked polyubiquitylation 
3. K48 and K11-linked chains reside on PCNA interactors and therefore are not 
covalently attached to (polyubiquitylated) PCNA 
 
Figure 28: Scenarios explaining accumulation of K48 and K11 linkages in PCNA 
immunoprecipitates. From left to right: K48 and K11 linkages are assembled on top of K63-
linked chains, on PCNA (except K164) or on PCNA interactors. 
 
To discriminate between the first scenario and the latter two, I employed a 
deubiquitinase-based assay called UBICREST (Ubiquitin Chain Restriction) (Hospenthal 
et al., 2015) to determine the precise architecture of ubiquitin chains on PCNA. In the 
following experiments, I focus mainly on K48 linkages due to the availability of a high-
quality linkage-specific antibody and established UBICREST assays for this linkage type. 
116 
 
The experimental set-up required immunoprecipitation of PCNA from cellular extracts, 
followed by treatment with linkage-selective deubiquitinating enzymes (hereafter called 
DUBs) and analysis of the ubiquitylation status of PCNA (‘elution from beads’) or soluble 
ubiquitin forms that are released into solution upon DUBs cleavage (‘supernatant’). This 
discrimination between soluble and insoluble (coupled to beads via PCNA) ubiquitin forms 
explains whether K48 linkages were assembled on top of K63-linked chains or have 
another attachment site. If a K63-linked chain on PCNA remains intact and is not further 
modified with other linkages, its cleavage by the K63-selective DUB AMSH (McCullough 
et al., 2006) should produce ubiquitin in the soluble fraction. Concomitantly, 
monoubiquitylated PCNA should remain on the beads, as AMSH is not able to cleave the 
first ubiquitin attached to a substrate. At the same time, if ubiquitin units in a K63-linked 
chain are further modified with other linkages, AMSH will not be able to cleave them, and 
they will appear in the soluble fraction as di-ubiquitin, tri-ubiquitin or higher conjugates 
depending on the length of branched units in an initial chain. This reaction is possible if 
AMSH activity towards K63-linked chains is unaffected by chain branching, which is most 
likely the case (Ohtake et al., 2016). We can also expect a loss of K48 linkages from the 
beads, which should be released into the supernatant. However, if a non-K63 linkage 
resides on a ubiquitin closest to PCNA, this linkage will not be targeted by AMSH cleavage 
and, therefore, should remain coupled to PCNA (Figure 29A).  
Firstly, I applied UBICREST-AMSH to the chains generated by PIP-E363. Four 
conditions were compared: mock treatment (DMSO), expression of PIP-E363, VCP 
inhibition or a combination of both. As demonstrated in Figure 29B, VCP inhibition alone 
does not lead to PCNA polyubiquitylation, whereas expression of PIP-E363 results in the 
robust formation of chains on PCNA, which are further stabilised by VCP inhibition. 
Consistent with the linkage specificity of PIP-E363, almost all ubiquitin signal on the beads 
disappears after incubation with AMSH. In striking contrast to the total ubiquitin signal, an 
anti-K48 linkage-selective antibody reveals the presence of K48 linkages only if both PIP-
E363 is expressed and VCP is inhibited, but in neither of these conditions alone. This may 
indicate that once formed, K48 linkages are promptly removed from chromatin by VCP 
and therefore are undetectable in the chromatin fraction. Similarly to total ubiquitin, the 
anti-K48 signal is lost upon AMSH treatment. As AMSH is unable to cleave K48 linkages, 
this suggests that K48 linkages are assembled on top of K63-linked chains. Analysis of 
117 
 
the ubiquitin forms released in the soluble fraction after AMSH treatment reveals the 
presence of low-intensity chain branching even in unperturbed cells, which is mainly 
limited to di-ubiquitin. This chain branching increases with PIP-E363 expression and is 
strongly stabilised by concomitant VCP inhibition. Interestingly, in the latter condition, the 
chain branching pattern shows more di-ubiquitin and tri-ubiquitin forms than 
monoubiquitin. Although precise quantification of the degree of chain branching from this 
experiment is difficult, we can roughly estimate that approximately every second ubiquitin 
in a chain is modified with non-K63 linkages. It is also worth pointing out that inhibition of 
VCP mainly affects uncleaved linkages but not monoubiquitin, indicating that branched 
chains are a better substrate for VCP than homotypic K63 linkages. 
To understand the nature of the branched chains on PCNA, I applied DUBs with 
different linkage selectivities to PCNA polyubiquitylated in vivo by PIP-E363. Suppose 
ubiquitin linkages released into the supernatant upon treatment with AMSH are 
conjugated through K48. In that case, co-treatment with the K48-selective DUB OTUB1 
(Wang et al., 2009) should decrease their amount, whereas treatment with a non-selective 
DUB USP2 should result in only monoubiquitin in the soluble fraction (Figure 30A). As 
shown in Figure 30B, amounts of di- to tetraubiquitin in the soluble fraction after AMSH 
cleavage are reduced when OTUB1 is added to the reaction. Some chains may remain 
upon combined treatment with AMSH + OTUB1 due to incomplete digestion by OTUB1 
or because other linkages are present (for example, K11, as their mobility in a gel is similar 
to that of K48-linked chains (Michel et al., 2018)). Surprisingly, reaction with OTUB1 alone 
also releases a chain in the supernatant, which is absent when both AMSH and OTUB1 
are used together. This indicates that the architecture of the chains may be more complex 
than initially anticipated: K48 linkages, assembled on top of the K63-linked chain, are 
further modified by K63 linkages, likely due to the presence of PIP-E363. As predicted, 
USP2 releases monoubiquitin as the only soluble form. Taken together, these data 
confirm that K63 chains on PCNA are branched in vivo and that these branched chains 
consist, at least partially, of K48 ones.  
 
118 
 
 
 
Figure 29: The UBICREST assay reveals ubiquitin chain branching on PCNA. (A) Schematic 
representation of a UBICREST assay applied to a homotypic (top) or branched ubiquitin 
conjugates (bottom). (B) On-beads fraction of the UBICREST assay applied to PCNA ubiquitylated 
via transfection of HEK 293T cells with PIP-E363 and, if indicated, treatment with 5 μM VCPi (NMS-
873) for 6 h. (C) Supernatant fraction of the UBICREST assay described in (B). Ubiquitin was 
detected with the VU1 monoclonal antibody. 
119 
 
 
 
Figure 30: K63-linked chains on PCNA are branched with K48 linkages. (A) Schematic 
representation of the UBICREST assay with the K63-selective DUB AMSH, the K48-selective DUB 
OTUB1 and the linkage non-selective DUB USP2 (cc = catalytic centre). (B) Supernatant fraction 
of the UBICREST assay applied to PCNA ubiquitylated via transfection of HEK 293T cells with 
PIP-E363 and, if indicated, treatment with 5 μM VCPi (NMS-873) for 6 h. Ubiquitin was detected 
with the VU1 anti-ubiquitin antibody. 
 
Finally, I applied the UBICREST assay to analyse the architecture of ubiquitin 
chains formed on PCNA upon replication stress independently of PIP-E363. HU, MMS and 
UV light were used as replication stress-inducing agents, and VCP was inhibited to 
increase the amount of chromatin-bound polyubiquitylated PCNA. Similar to the PIP-E363 
system, AMSH cleavage releases ubiquitin chains into the soluble fraction, indicating that 
branching of K63-linked chains takes place (Figure 31). Interestingly, among all three 
treatments, adding HU together with VCP inhibition results in the highest amount of both 
total K63 linkages, as indicated by the strong accumulation of monoubiquitin in the 
supernatant, and branched conjugates. One possible explanation for these differences 
between the three genotoxins may lie in their mechanism of action: MMS and UV directly 
damage DNA bases, and PCNA ubiquitylation is triggered once a replisome encounters 
damaged DNA. On the contrary, hydroxyurea affects all the active replication forks due to 
120 
 
the depletion of the dNTP pool and, therefore, may affect more replication forks at a given 
time, thus inducing stronger PCNA polyubiquitylation. Altogether, both the PIP-E363 
system and replication stress produce similar ubiquitin architectures on PCNA, which 
suggests that they should trigger similar downstream mechanisms in vivo. 
 
3.5.3.2 Targeted deubiquitylation of PCNA in vivo recapitulates the UBICREST 
assay 
In a UBICREST assay, DUBs are applied to polyubiquitylated PCNA 
immunoprecipitated from cell extracts. I aimed to understand whether targeting the K48-
selective DUB OTUB1 to PCNA can interfere with chain branching in vivo.  A similar 
strategy has been employed in the DUBTAC method, where OTUB1 is targeted to 
substrates via heterobifunctional small molecules to promote substrate stabilisation  
(Henning et al., 2022). To promote the interaction of OTUB1 with PCNA, I fused a PIP-
box N-terminally to OTUB1, analogously to the architecture of tailor-made ubiquitin ligases 
(Figure 32A). An N-terminal Flag tag was used for the detection of the enzyme. 
Furthermore, since OTUB1 has been shown to directly bind and inhibit UBC13 and, 
therefore, may cause side effects upon its overexpression, the catalytically inactive 
OTUB1 mutant C91S was used as a control (hereafter referred to as “CS“). Figure 32B 
shows the expression levels of WT and CS PIP-OTUB1. This fusion protein is expressed 
at drastically higher levels than PIP-E363, which may be due to the codon usage of the 
constructs or the higher stability of OTUB1 in comparison to E363. Inhibition of VCP 
reduces the levels of PIP-OTUB1, possibly due to the general toxic effects of VCPi. Figure 
32C demonstrates the effects of PIP-OTUB1 expression on PIP-E363-induced PCNA 
polyubiquitylation. Surprisingly, expression of PIP-E363 alone or in combination with PIP-
OTUB1 CS results in similar levels of PCNA polyubiquitylation despite much higher levels 
of PIP-OTUB1 CS compared to PIP-E363, which I expected to have a dominant-negative 
effect. When catalytically active PIP-OTUB1 is used, polyubiquitin chains on PCNA, 
visualised by total ubiquitin antibody, become shorter. At the same time, the K48-linked 
polyubiquitin signal is reduced in PCNA immunoprecipitates. Thus, similarly to the in vitro 
assays, PIP-OTUB1 can de-branch K63-K48 chains on PCNA. Interestingly, the increase 
121 
 
in PCNA-Ub and PCNA-Ub2 levels in the condition of PIP-E363 and PIP-OTUB1 co-
expression likely indicates that these PCNA forms (and potentially others) are further 
modified with K48 linkages in vivo. These data confirm the results of the in vitro 
UBICREST assay, which suggested that not only distant but also the first ubiquitin in a 
K63-linked chain, can be a substrate for K48-linked ubiquitylation. Altogether, the 
consequences of targeting OTUB1 to PCNA in vivo confirm the results of the UBICREST 
assay and validate the approach as a tool to interfere with the branching of K63-linked 
chains in vivo.  
 
 
 
Figure 31: Replication stress induces formation of K63-K48 branched chains on PCNA. 
RPE1 hTERT WT cells were mock-treated, treated with 4 mM HU for 2 h, 0.01% MMS for 1 h 
followed by 2 h recovery or 40 J/m2 UV followed by 2 h recovery. In all cases, NMS-873 (VCPi) or 
the corresponding amount of DMSO were added to a final concentration of 5 μM during the last 2 
h of treatments. Soluble (left) and bead (right) fractions of the UBICREST assay are shown 
122 
 
 
 
Figure 32: Targeting of OTUB1 to PCNA interferes with chain branching in vivo. (A) Scheme 
of the PIP-OTUB1 construct and its activity towards PCNA modified with branched ubiquitin 
chains. (B) Western blot analysis of whole cell extracts from HEK 293T cells, transfected with the 
specified constructs and, if indicated, treated with 5 μM VCPi (NMS-873) for 6 h. Ponceau S 
staining serves as a loading control. (C) PCNA polyubiquitylation in HEK 293T cells, transfected 
and treated as described in (B). Asterisk indicates PCNA modified with a K48-linked di-ubiquitin. 
   
123 
 
3.5.3.3 K63-linked chains on PCNA are branched via K48 and K11 
The UBICREST assay and the in vivo targeting of the K48-selective DUB OTUB1 
to PCNA reveal that the K63-linked chains built by PIP-E363 or by cellular enzymes upon 
replication stress are further modified with K48 linkages. At the same time, MS-based 
interactome analysis suggests that the build-up of not only K48 but also K11-linked chains 
is induced by PIP-E363 expression upon inhibited VCP activity. Because linkages other 
than K63, K48 and K11 were not identified in this experiment, they cannot be quantified, 
yet they may still be involved in the branching of K63-linked chains on PCNA. The 
involvement of specific ubiquitin lysines in certain cellular pathways can be conveniently 
analysed by expressing the corresponding ubiquitin mutants. In yeast, ubiquitin 
replacement strains, in which exogenous ubiquitin mutants are expressed upon the knock-
out of endogenous ubiquitin genes, have found widespread use (Meza Gutierrez et al., 
2018). Although a similar strategy exists for mammalian cells, it involves transient knock-
down of ubiquitin genes (Xu et al., 2009), which is highly labour-consuming and may 
cause unwanted side effects due to changes in total ubiquitin levels. Another commonly 
used approach in human cells relies on overexpressing a tagged ubiquitin mutant without 
depleting its endogenous pool. Subsequent tag-based pulldown allows for the enrichment 
of proteins that incorporated the overexpressed ubiquitin. To analyse which of the six 
ubiquitin lysine residues are involved in the branching of the K63-linked chains on PCNA, 
I created the following panel of His10-tagged ubiquitin mutants: K48R, K11R/K48R and 
K6R/K11R/K27R/K29R/K33R/K48R (“K63-only”-Ubiquitin). In addition, for each ubiquitin 
variant (including the wild-type), I added the K63R mutation. This modification acts as a 
negative control as it does not support the assembly of K63-linked chains. The addition of 
this mutation to the “K63-only”-Ubiquitin results in a “K0-Ubiquitin” that lacks any lysines 
and can modify proteins only as a single unit. My hypothesis regarding the effects of the 
ubiquitin mutants on PIP-E363-induced PCNA polyubiquitylation is graphically represented 
in Figure 33A. In the case of wild-type ubiquitin overexpression, branching of K63-linked 
chains should occur, leading to the degradation of the polyubiquitylated PCNA forms. On 
the contrary, the K48R mutation should protect K63-linked chains from branching via K48, 
stabilising polyubiquitylated PCNA. If additional lysines are involved in chain branching, 
their mutation may lead to even further stabilisation of polyubiquitylated PCNA. Adding a 
124 
 
K63R mutation should inhibit PCNA polyubiquitylation by PIP-E363 in all cases. Figure 
33B shows the ubiquitylated PCNA forms that were enriched with the Ni-NTA pulldown 
and therefore were ubiquitylated with His-tagged ubiquitin. Without PIP-E363, only a single 
prominent band is present, corresponding to monoubiquitylated PCNA. Interestingly, upon 
expression of PIP-E363, I do not observe ubiquitin chains but rather the disappearance of 
monoubiquitylated PCNA. The same effect was observed for PIP-E348 (Figure 13). These 
observations indicate that, upon overexpression of ubiquitin, the pathway is accelerated, 
and equilibrium is shifted towards PCNA degradation. As expected, accumulation of 
chains takes place when the K48R ubiquitin mutant is used. Strikingly, even further 
stabilisation is observed upon overexpression of the double K11R/K48R mutant, whereas 
mutation of all lysines except K63 (K63-only form) does not further increase the amount 
of detected polyubiquitylated PCNA. As expected, the K0-Ubiquitin mutant abolishes 
chain formation for all chain types, therefore confirming that the observed 
polyubiquitylated PCNA forms contain K63 linkages. This clearly pinpoints K48 and K11 
as the two substrate lysines for branching of a K63-linked polyubiquitin chain on PCNA. 
125 
 
 
 
Figure 33: K63-linked chains on PCNA are branched via K48 and K11. (A) Schematic 
representation of the chain architecture analysis by overexpression of ubiquitin mutants. (B) 
Indicated ubiquitin mutants were expressed in HEK 293T cells together with PIP-E363, followed by 
Ni-NTA pulldown and analysis of PCNA ubiquitylation by western blotting.  
 
126 
 
3.5.4 UBE2K is involved in the build-up of branched chains on PCNA 
One of the important characteristics of PIP-E363 is its exclusive linkage selectivity, 
as proven in vitro and in vivo in S. cerevisiae (Wegmann et al., 2022). Therefore, it is 
reasonable to envision that certain cellular enzymes use K63-linked chains on PCNA as 
a substrate and extend them via lysines K48 and K11, into branched conjugates. This 
would also explain the appearance of branched chains on PCNA upon replication stress 
in the absence of PIP-E363, when K63-linked polyubiquitylation is carried out by cellular 
enzymes. Several enzymes have been reported to exhibit chain branching activity towards 
K63-linked ubiquitin chains. The first is the ubiquitin-conjugating (E2) enzyme UBE2K and 
its yeast homologue UBC1. In addition to a UBC domain, characteristic of all E2 enzymes 
(Stewart et al., 2016), it possesses a highly K63 linkage-selective UBA domain (Figure 
34A). This specificity is demonstrated by the ability of this UBA domain to discriminate 
between K63-linked and conformationally similar M1-linked ubiquitin chains (Pluska et al., 
2021). Upon binding to a K63-linked ubiquitin dimer, UBE2K ubiquitylates K48 of the 
proximal ubiquitin. Importantly, UBE2K is much more processive in creating a single K48 
linkage on a K63-linked dimer than in synthesising long K48-linked chains. The second 
ligase with chain branching activity is the HECT-type E3 HUWE1, which was also 
identified as a binder of polyubiquitylated PCNA in our mass-spectrometry-based 
interactome analysis. HUWE1 turns homotypic K63-linked chains on TRAF6 into 
branched structures and, in cooperation with UBR4 and UBR5, branches K63-linked 
chains on TXNIP (Ohtake et al., 2016; Ohtake et al., 2018). In order to understand whether 
UBE2K and HUWE1 are involved in chain branching in the PIP-E363 system, these 
enzymes were depleted in RPE1 hTERT PIP-E363 cells. As shown in Figure 34B, 
depletion of UBE2K almost completely abolishes checkpoint activation after the 
expression of PIP-E363. Depletion of HUWE1 leads to a moderate decrease in checkpoint 
activation, and co-depletion of both enzymes behaves like depletion of UBE2K alone. 
Importantly, the knockdown of UBE2K does not affect the expression levels of PIP-E363, 
which means that the deleterious effects of excessive PCNA polyubiquitylation depend on 
the enzymatic activity of UBE2K.  
Next, I analysed whether UBE2K can branch ubiquitin chains on PCNA in vitro 
(these reactions were performed by Nils Krapoth, Ulrich Laboratory). As demonstrated in 
127 
 
Figure 34C, when unmodified PCNA or PCNA monoubiquitylated on K164 were used as 
substrates, UBE2K does not display any significant activity. However, when K63-linked 
polyubiquitylated PCNA is used as a substrate, the addition of UBE2K drastically changes 
the chain pattern. Anti-ubiquitin and anti-K63-polyubiquitin signals shift to the high 
molecular weight range, and the anti-K48 signal appears only when PCNA is pre-
polyubiquitylated with K63 linkages and UBE2K is added to the reaction. At the same time, 
there is a reduction in shorter ubiquitylated PCNA forms, thus suggesting that they are 
converted into heavier conjugates due to UBE2K activity. This experiment confirms the 
specificity of UBE2K towards K63-linked chains in comparison to monoubiquitin and non-
ubiquitylated substrates.  
To prove that UBE2K also affects PCNA polyubiquitylation in vivo in the PIP-E363 
system, the tailor-made ligase was expressed in RPE1 hTERT cells transfected with an 
siRNA against UBE2K or a corresponding control siRNA, and VCP was inhibited to further 
accumulate chains on chromatin-bound PCNA. As expected, the knockdown of UBE2K 
leads to a decrease in both anti-ubiquitin and anti-K48-linked polyubiquitin signal after 
PCNA immunoprecipitation, whereas the levels of PCNA monoubiquitylation in whole cell 
extracts appear to slightly increase upon loss of UBE2K (Figure 35). This indicates that 
the above-described drop in PCNA polyubiquitylation is not the result of lower levels of 
monoubiquitylated PCNA but rather confirms the chain branching activity of UBE2K. The 
UBICREST assay also confirms this hypothesis: when UBE2K is depleted, less di- and 
tri-ubiquitin forms are released into the soluble fraction after AMSH-mediated cleavage of 
polyubiquitylated PCNA without significantly affecting the amount of cleaved 
monoubiquitin (Figure 36A). Overall, these observations show that UBE2K is involved in 
the formation of short K48 linkages on top of PCNA-conjugated K63-linked chains. Yet, it 
is also clear that a significant amount of chain branching in the PIP-E363 system takes 
place even in the absence of UBE2K, thus suggesting that ubiquitin ligases may be 
involved in this pathway.  
 
128 
 
 
Figure 34: UBE2K forms K63-K48 branched ubiquitin chains in the PIP-E363 system. (A) 
Schematic representation of the UBE2K domain structure and catalytic activity. (B) Checkpoint 
activation in RPE1 hTERT PIP-E363 cells after depletion of HUWE1 and UBE2K and doxycycline-
induced expression of the ligase for the indicated times. (C) In vitro ubiquitylation assay with 
UBE2K as a ligase and PCNA, PCNA-Ub and PCNA-Ubn(K63) as substrates. 
129 
 
 
 
Figure 35: Effects of UBE2K on PCNA polyubiquitylation in vivo. RPE1 hTERT PIP-E363 cells 
were depleted of UBE2K and treated with 1 μg/ml doxycycline and 5 μM NMS-873 (VCPi) for 6 h. 
PCNA ubiquitylation was analysed after PCNA immunoprecipitation from the chromatin fraction or 
in total cell extracts. Ponceau S staining serves as a loading control.  
 
Finally, I analysed the impact of UBE2K on PCNA polyubiquitylation upon 
replication stress. I compared different replication stress-inducing agents: HU (2 h and 
4 h), MMS and UV. As shown in Figure 36B, MMS and UV induce stronger PCNA mono- 
and polyubiquitylation than HU, and depletion of UBE2K leads to a reduction in some 
polyubiquitylated forms of PCNA. Although the UBICREST assay is needed to strictly 
prove that these forms represent branched chains on PCNA, the first evidence suggests 
that UBE2K is active on K63-linked chains on PCNA independently of the enzyme(s) that 
create them. Interestingly, checkpoint signalling is unaffected by UBE2K depletion. In the 
case of MMS and UV, it is likely that a significant portion of the checkpoint signalling 
originates from the nucleotide excision repair and gap resection outside of the S phase 
(Balogun et al., 2013), and for all tested damaging agents, PCNA polyubiquitylation seems 
to play a minor role in checkpoint activation in wild-type RPE1 cells. 
130 
 
 
 
Figure 36: UBE2K affects PCNA polyubiquitylation both in the PIP-E363 system and during 
endogenous damage response. (A) The UBICREST assay was performed on PCNA, 
immunoprecipitated from RPE1 hTERT PIP-E363 cells treated with 1 μg/ml doxycycline and 5 μM 
NMS-873 (VCPi). Soluble ubiquitin forms were visualised by western blotting with the VU1 
ubiquitin antibody. (B) RPE1 hTERT cells were mock-treated, treated with 4 mM HU for 2 h or 4h, 
with 0.01% MMS for 4h or with 40J/m2 UV followed by a 4h recovery in complete growth medium. 
PCNA ubiquitylation and checkpoint signalling were analysed in total cell extracts by western 
blotting with the indicated antibodies. Ponceau S staining serves as a loading control. 
 
3.6 Analysing the significance of polyubiquitin linkage type on PCNA in 
triggering replication collapse 
K63- and K48-linked polyubiquitin chains are the most studied polyubiquitin conjugates. 
The difference between these chain types is unambiguously demonstrated by their 
different effects on DNA damage bypass in S. cerevisiae (Wegmann et al., 2022). In the 
human PIP-E363 system, I observe the downstream modification of the K63-linked chain 
131 
 
on PCNA with K48 linkages, leading to VCP recruitment and extraction of the 
polyubiquitylated PCNA from chromatin. Therefore, unlike the situation in budding yeast, 
in human cells one would expect that modifying PCNA with a K48-linked chain should 
lead to a similar, rather than opposite, phenotype, as both VCP and the proteasome have 
a high affinity towards K48-linked chains. I decided to test whether polyubiquitylation of 
PCNA with K48-linked chains leads to checkpoint activation in a manner similar to K63-
linked chains and, if so, which chain type is more efficient in recruiting VCP and promoting 
replication catastrophe. 
 
3.6.1 Establishment of a cell line with inducible K48-linked polyubiquitylation of 
PCNA 
Our initial optimisation of PIP-E348 in HEK 293T cells showed that maximal activity 
towards PCNA requires co-overexpression of the yeast E2 Ubc7, optimised in terms of 
codon usage for H. sapiens. PIP-E348 construct contains a VSV and a Flag tag at its N-
terminus, and Ubc7 is additionally N-terminally tagged with a triple Flag sequence. For 
the sake of simplicity, in the following sections, I will omit the tags from the construct 
names. In order to create a cell line that expresses both PIP-E348 and Ubc7, I created a 
single fusion protein containing both enzymes separated by a self-cleavable P2A peptide. 
The latter consists of 22 amino acids and is co-translationally cleaved between the C-
terminal glycine and proline (Ryan et al., 1991). Upon expression and self-cleavage of a 
fusion protein, the first enzyme (PIP-E348) receives 21 additional amino acids on its C-
terminus, and the second protein (Ubc7) starts with a proline just upstream of its genuine 
initiator methionine (Figure 37A). The fusion construct, hereafter called PIP-E348-P2A-
Ubc7, was integrated into RPE1 hTERT FlpIn cells in a manner analogous to the RPE1 
hTERT PIP-E363 cell line. A cell line that carries PIP-E348 without the P2A-Ubc7 extension 
was used as a control. Figure 37B shows the expression of the enzymes in the described 
cell lines after the addition of doxycycline and VCP inhibition. As expected, only PIP-E348 
is expressed in the RPE1 hTERT PIP-E348 cell line. An additional species can be observed 
in the RPE1 hTERT PIP-E348-P2A-Ubc7 cell line, corresponding to P2A-cleaved Ubc7. 
Notably, the apparent molecular weight of PIP-E348, originating from the fusion construct, 
132 
 
is slightly higher than that of the enzyme expressed in RPE1 hTERT PIP-E348 cells, 
reflecting the presence of the post-cleavage peptide remnant of P2A at the C-terminus of 
the ligase. Consistently with results obtained in HEK 293T cells upon transient 
overexpression of the enzymes, efficient polyubiquitylation of PCNA requires co-
overexpression of Ubc7 together with PIP-E348 (Figure 37C). Interestingly, even upon co-
expression of Ubc7 and PIP-E348, the levels of polyubiquitylated PCNA are quite low if 
VCP is not inhibited. This indicates that even single K48 linkages on PCNA, along with 
longer chains, trigger efficient VCP-mediated removal from chromatin. 
 
Figure 37: Design and creation of a cell line with inducible K48-linked polyubiquitylation of 
PCNA. (A) Schematic representation of a self-cleavable fusion of PIP-E348 to Ubc7 via a P2A 
peptide. The full amino acid P2A sequence as well as the cleavage site are shown. (B) Cell lines 
with doxycycline-inducible expression of PIP-E348 or its self-cleavable fusion with Ubc7 were 
treated for 6 h with 1 μg/ml doxycycline or 5 μM VCPi (NMS-873). Enzyme levels were analysed 
by western blotting of cell lysates against the Flag tag. Ponceau S staining serves as a loading 
control. (C) Indicated cell lines were treated as in (B), followed by PCNA immunoprecipitation from 
the chromatin fraction and western blotting with the indicated antibodies. Note that the order of 
samples is opposite of those in (B). 
 
133 
 
A certain proportion of the PIP-E348-P2A-Ubc7 construct in RPE1 hTERT cells 
remains uncleaved and appears as a ~50 kDa product (Figure 37B). Since this 
permanent fusion could be more active than co-expressed but separate E3 and E2 
enzymes, I investigated whether rendering the P2A peptide uncleavable could further 
boost the activity of the system. Provided that the four last amino acids of the P2A peptide 
are critical for the self-cleavage (Hahn and Palmenberg 1996), I created a mutant version 
with the C-terminal NPGP sequence mutated to KLAS sequence (hereafter referred to as 
P2A*, Figure 38A). As expected, the mutant version appears in a western blot as a ~ 50 
kDa protein with no detectable free PIP-E348 or Ubc7 (Figure 38B). Interestingly, 
permanent fusion of Ubc7 to PIP-E348 via P2A* leads to a similar level of PCNA 
polyubiquitylation as making the E3 and E2 enzymes from the self-cleavable precursor. 
Consistently with the experiments described above, K48-linked ubiquitin chains produced 
by the uncleavable mutant on chromatin-bound PCNA can be detected only upon VCP 
inhibition (Figure 38C). 
On the one hand, these results could indicate that head-to-tail fusion of the E2 to 
E3 results in their optimal orientation for the ubiquitin transfer. On the other hand, the 
amount of polyubiquitin chains on PCNA depends on the level of PCNA 
monoubiquitylation, which may become a limiting factor for the reaction. Therefore, for the 
PCNA system, I did not pursue using the permanent fusion of PIP-E348 to Ubc7 because 
it does not display any additional activity beyond that provided by the self-cleavable fusion. 
However, for the future development of tailor-made enzymes, for example, for the 
‘Ubiquiton’ technology (section 3.8), these constructs may be of practical use. 
 
3.6.2 Direct comparison of K63- versus K48-linked PCNA polyubiquitylation 
The data described above suggest that K63- and K48-linked PCNA 
polyubiquitylation may similarly affect cell replication, as they both are substrates for VCP 
activity. In order to directly compare these chain types on PCNA, they were induced in 
RPE1 hTERT cells by expressing PIP-E363 or PIP-E348-P2A-Ubc7 (Figure 39). I observed 
that K63- and K48-linked ubiquitin chains on PCNA behave differently with respect to VCP 
inhibition. For K63-linked chains, VCP does not extract oligo-ubiquitylated (containing 2-
4 ubiquitin units) PCNA from chromatin but is rather active towards heavily 
134 
 
polyubiquitylated forms, likely consisting of branched ubiquitin conjugates. This is different 
for K48-linked chains, where all ubiquitin conjugates on PCNA, including a single K48 
linkage, are substrates of VCP. Yet, K63-linked polyubiquitylation of PCNA leads to a 
much stronger checkpoint activation than K48-linked one, as demonstrated by higher 
levels of RPA phosphorylation 6 h after the expression of the corresponding E3 enzymes 
for the former. Remarkably, for both chain types, checkpoint activation is dependent on 
VCP activity. Altogether, these data indicate that although both chain types lead to the 
activation of checkpoint signalling, probably due to the removal of PCNA from chromatin, 
K63-linked chains are significantly more active in triggering this pathway. 
 
 
Figure 38: Self-cleavable and uncleavable fusions of PIP-E348 to Ubc7 display similar 
activity towards PCNA in vivo. (A) Schematic representation of the self-cleavable construct PIP-
E348-P2A-Ubc7 in comparison to its uncleavable version PIP-E348-P2A*-Ubc7. (B) HEK 293T cells 
were transfected with the indicated constructs and harvested after 24 h. If stated, VCPi (NMS-
873) was added to growth medium to a final concentration of 5 μM 6 h prior to cell harvesting. Cell 
lysates were analysed by western blotting against the Flag tag. Ponceau S staining serves as a 
loading control. (C) HEK 293T cells were treated as described in (B), followed by PCNA 
immunoprecipitation from the chromatin fraction and western blotting with the indicated antibodies. 
 
135 
 
 
Figure 39: Comparison of K63- and K48-linked PCNA polyubiquitylation. RPE1 hTERT PIP-
E363 or PIP-E348-P2A-Ubc7 cells were treated with 1 μg/ml doxycycline or 5 μM VCPi (NMS-873) 
for 6 h, followed by chromatin fractionation and PCNA immunoprecipitation (left) or lysis and 
western blotting analysis with the indicated antibodies. Ponceau S staining serves as a loading 
control.  
One possible explanation for why K63-linked chains trigger stringer checkpoint 
activation could be that branched ubiquitin chains, originating from K63 ones, are a better 
substrate for VCP than homotypic K48-linked chains. Despite the fact that it has not 
directly been investigated for K63 chains, this effect has been demonstrated for K48 and 
K11 linkages: branched conjugates have higher affinity to VCP (and its adaptors) than 
homotypic chains (Yau et al., 2017). I previously demonstrated that UBE2K is at least 
partially responsible for the branching of K63-linked chains in the PIP-E363 system. If our 
mechanistic understanding of the pathway is correct, UBE2K depletion should not affect 
checkpoint activation in the PIP-E348 system because UBE2K cannot branch K48-linked 
chains. As demonstrated in Figure 40, this is indeed the case. While checkpoint activation 
by PIP-E363 largely depends on UBE2K, PIP-E348-induced checkpoint activation is 
significantly less intense and does not depend on the levels of UBE2K. Interestingly, if 
136 
 
DNA damage is directly induced by MMS, I do not observe a significant dependency of 
checkpoint activation on UBE2K. This indicates that in the presence of exogenous DNA 
damage, most checkpoint signalling originates from sources other than polyubiquitylated 
PCNA. 
 
 
 
 
Figure 40: UBE2K mediates checkpoint activation after K63-, but not K48-linked PCNA 
polyubiquitylation. RPE1 hTERT cells inducibly expressing PIP-E363 or PIP-E348-P2A-Ubc7 
were transfected with an siRNA against UBE2K or a non-targeting siRNA for 72 h and treated, if 
indicated, with 1 μg/ml doxycycline for 24 h. 90 min treatment with 0.01% MMS was used as a 
positive control for checkpoint activation. The levels of the E3s, UBE2K and phosphorylated RPA 
were analysed in total cell extracts by western blotting. Ponceau S staining serves as a loading 
control. 
  
137 
 
3.7 Analysing the physiological relevance of ubiquitin chain branching on 
PCNA 
3.7.1 Introduction to the synthetic lethal relationship between BRCA1 and USP1 
In human cells, the endogenous levels of PCNA monoubiquitylation are sufficient 
to support PIP-E363 activity. Due to the elongation specificity of the tailor-made E3s, the 
amount of polyubiquitylated PCNA must be governed by the availability of the 
monoubiquitylated form. The ubiquitin ligases HLTF and SHPRH display activities similar 
to PIP-E363: they directly interact with PCNA and can extend a monoubiquitin moiety into 
a K63-linked chain (Seelinger and Otterlei, 2020; Motegi et al., 2008; Unk et al., 2006). 
Nevertheless, contrary to the endogenously expressed HLTF and SHPRH, PIP-E363 
expression can lead to a replication catastrophe. Therefore, I hypothesised that certain 
mechanisms may counteract the activity of HLTF and SHPRH but not PIP-E363. One of 
these mechanisms could rely on recruiting the DUB USP1, which deubiquitylates 
monoubiquitylated PCNA and FANCD2 (Huang et al., 2006; Nijman et al., 2005). The 
tonicity-responsive enhancer-binding protein TonEBP has been shown to recruit both 
SHPRH and USP1 to the sites of DNA damage, preventing persistent PCNA 
polyubiquitylation (Kang et al., 2019). Therefore, I investigated whether inhibition of USP1 
activity can, in certain cellular backgrounds, phenocopy the expression of PIP-E363. 
Recently, USP1 and BRCA1 were shown to be in a synthetic lethal relationship: inhibition 
or depletion of USP1 in BRCA1-deficient cells results in cell death concomitantly with a 
drastic decrease in total PCNA levels. This synthetic lethality depends on RAD18 and 
UBE2K, suggesting that PCNA, rather than FANCD2 ubiquitylation underlies the observed 
phenotypes (Simoneau et al., 2023). The reported effects of USP1 inhibition in BRCA1-
deficient cells resemble the phenotypes of PIP-E363 expression, including impaired cell 
growth, accumulation of cells in the S phase and a lower replication speed. Dependency 
of these phenotypes on UBE2K makes it tempting to imply that chain branching could be 
the underlying pathway. Additionally, inhibition of USP1 in a BRCA1-deficient background 
leads to deprotection of replication forks, as observed by the degradation of nascent DNA 
after blocking replication with hydroxyurea. Although fork deprotection has been 
demonstrated to depend on RAD18 activity, other details of the pathway remain unknown 
138 
 
(Lim et al., 2018). In the following sections, I will investigate whether the replication 
problems in BRCA1-deficient cells where USP1 is also inhibited have the same origin as 
those in PIP-E363-expressing cells. 
 
3.7.2 VCP mediates PCNA degradation upon inhibition of USP1 in BRCA1-deficient 
cells 
As a model cell line lacking functional BRCA1, I used breast cancer MDA-MB-436 
cells (kind gift from Petra Beli Laboratory, IMB). These cells contain a mutation in exon 20 
of BRCA1, resulting in a truncated protein (Elstrodt et al., 2006). If inhibition of USP1 in 
the background of BRCA1 deficiency is indeed functionally similar to overexpression of 
PIP-E363 in wild-type cells, VCP inhibition should reverse these effects because 
polyubiquitylated PCNA cannot be removed from chromatin and subsequently degraded. 
Since inducing significant PCNA degradation requires 24 h of USP1 inhibition (Simoneau 
et al., 2023), and inhibition of VCP for too long is cytotoxic, in order to further boost PCNA 
monoubiquitylation and therefore accelerate the pathway, I treated the cells with a high 
dose of hydroxyurea (HU). Indeed, although 4 h treatment with hydroxyurea alone does 
not lead to a significant loss of PCNA in total cell extracts, the addition of the USP1 
inhibitor ML-323 (hereafter USP1i) induces PCNA degradation (Figure 41). The observed 
degradation is VCP-dependent, as it can be reversed upon co-treatment with the VCP 
inhibitor. A long exposure of the western blot for PCNA reveals the presence of short 
degradation products when hydroxyurea is used together with the USP1 inhibitor, likely 
originating from incomplete proteasomal degradation of PCNA. Given that inhibition of 
VCP not only stabilises PCNA but also prevents the formation of partially degraded PCNA 
forms, this means that, as for the artificially modified PCNA in the PIP-E363 system, 
extraction of natively polyubiquitylated PCNA from chromatin by VCP is a prerequisite for 
its proteasomal degradation. Notably, checkpoint activation, as represented by levels of 
RPA phosphorylation, correlates with PCNA levels, pinpointing the role of PCNA in 
checkpoint activation in this system. Time course analysis reveals consistent PCNA 
degradation 0-6 h after the addition of both HU and USP1 inhibitor in both total cell extracts 
139 
 
(Figure 41B) and the chromatin fraction (Figure 41C), which can be rescued by VCP 
inhibition. 
The behaviour of monoubiquitylated PCNA in this experiment is surprising: 
although co-treatment with VCPi, HU and USP1i massively stabilises total PCNA levels, 
the levels of the monoubiquitylated form remain relatively unchanged. This could be 
explained if constant recycling of PCNA takes place at the HU-induced replication 
intermediates. Once polyubiquitylated, PCNA is removed from chromatin and degraded 
by the proteasome. If new PCNA trimers replace the removed ones, they should undergo 
the same process as forks remain in a compromised state. In this scenario, the total 
amount of PCNA should drop due to continuous proteasomal degradation, but the levels 
its monoubiquitylated form should be determined by the number of stressed forks or other 
replication intermediates and therefore remain similar in a given timescale.  
.  
140 
 
 
 
Figure 41: HU- and USP1i-dependent PCNA degradation in BRCA1-deficient cells. (A) MDA-
MB-436 cells were treated with 4 mM HU, 30 μM ML323 (USP1i) or 5 μM NMS-873 (VCPi) for 
4 h, followed by western blotting with the indicated antibodies. Ponceau S serves as a loading 
control. (B) MDA-MB-436 cells were treated with 4 mM HU, 30 μM ML323 (USP1i) or 5 μM NMS-
873 (VCPi) for the indicated time periods followed by western blotting with the indicated antibodies. 
(C) MDA-MB-436 cells were treated as in (B) followed by isolation of the chromatin fraction and 
western blotting with the indicated antibodies. 
141 
 
3.7.3 Ubiquitin chain branching underlies the synthetic lethality between BRCA1 
and USP1 
The synthetic lethality between BRCA1 and USP1 was shown to originate from a 
massive loss of PCNA and to be dependent on RAD18 and UBE2K (Simoneau et al., 
2023). According to the in vitro data and the PIP-E363 system, UBE2K is unlikely to act on 
monoubiquitylated PCNA but rather on polyubiquitylated one. As the ubiquitin ligase 
RFWD3 promotes PCNA polyubiquitylation without significantly affecting 
monoubiquitylated PCNA levels (Moore et al., 2023), I tested whether depletion of RFWD3 
would prevent USP1i toxicity in BRCA1-deficient cells. As shown in Figure 42A, the USP1 
inhibitor ML-323 affects MDA-MB-436 cell viability in a dose-dependent manner. 
Importantly, depletion of RAD18, UBE2K or RFWD3 completely prevents ML-323 toxicity. 
Furthermore, treatment of MDA-MB-436 cells with increasing ML-323 concentrations for 
72 h results in bulk PCNA degradation, which is prevented if RAD18, UBE2K or (partially) 
RFWD3 are depleted (Figure 42B). Knocking down RAD18 prevents the ML-323-induced 
increase in PCNA monoubiquitylation, which is consistent with the role of RAD18 in 
monoubiquitylating PCNA. Depleting UBE2K does not prevent PCNA monoubiquitylation 
but still reverts PCNA degradation. This is consistent with our model, where I propose that 
UBE2K acts on polyubiquitylated PCNA without affecting its monoubiquitylation. 
Furthermore, I observe RAD18-, UBE2K- and RFWD3-dependent activation of checkpoint 
signalling, as demonstrated by an increase in phospho-RPA T21 levels, concomitantly 
with PCNA degradation. Thus, inhibiting USP1 in a BRCA1-deficient background affects 
cells in the same way as expressing PIP-E363 in wild-type cells. Thus, I propose that chain 
branching on PCNA is the pathway that mediates the degradation of this protein in MDA-
MB-436 cells and underlies BRCA1-USP1 synthetic lethality. 
142 
 
 
 
Figure 42: Toxic effects of USP1 inhibition in BRCA1-deficient cells is dependent on 
RAD18, UBE2K and RFWD3. (A) MTT cell viability assay performed on MDA-MB-436 cells, 
transfected with the indicated siRNAs for 48 h and cultured for 96 h in the presence of indicated 
concentrations of the USP1 inhibitor ML-323. (B) MDA-MB-436 cells were transfected and treated 
as described in (A). USP1 was inhibited for 72 h. The levels of PCNA and phosphorylated RPA 
were analysed by western blotting. Ponceau S staining serves as a loading control.  
143 
 
3.7.4 Removal of PCNA impacts the damage bypass pathway in BRCA1-deficient 
cells 
K63-linked PCNA polyubiquitylation in human cells has been functionally linked to 
the pathway of replication fork reversal (Vujanovic et al., 2017). Formation of reversed 
forks during replication over damaged DNA is thought to protect forks from collapse 
(Mutreja et al., 2018) and simultaneously transfer damaged bases into the parental DNA 
duplex, where they can be repaired by the base or the nucleotide excision repair 
machinery. In terms of replication speed, fork reversal slows down replication:  interfering 
with the pathway results in higher replication rates upon replication stress. Potentially, 
branching of K63-linked chains on PCNA can affect replication over damaged DNA in at 
least three different ways. First, removal of PCNA from replication forks is important for 
the completion of fork reversal (Park et al., 2019). Second, if this reaction takes place 
before ZRANB3 recruitment, it can rather prevent fork reversal as it should remove K63-
linked polyubiquitylated PCNA, which is necessary for ZRANB3 recruitment, from 
stressed forks. In this case, alternative pathways, such as TLS at the fork or repriming, 
should be favoured. Third, the branching could impact the interactions of polyubiquitylated 
PCNA with ZRANB3 or WRNIP1, or there may be specific readers of the branched chains.  
To gain insights into the roles of UBE2K-VCP pathway in DNA damage bypass, I 
used the MDA-MB-436 cell line and analysed how well it replicates over damaged DNA 
in the presence of the USP1 inhibitor, which should lead to excessive chain branching on 
PCNA. As a replication stress-inducing agent I used hydroxyurea, which induces oxidative 
damage of DNA bases and simultaneous depletion of the dNTP pool, and analysed the 
speed of replication by means of the DNA fibre assay. As shown in Figure 43A, inhibition 
of USP1 results in faster replication in the presence of hydroxyurea. The observed effect 
is likely due to PCNA removal from chromatin, as it does not take place if VCP is inhibited.  
Another assay that has been extensively used to study DNA transactions at 
stressed forks is the nascent DNA degradation (NDD) assay. It is based on the 
observation that in certain conditions, newly synthesised DNA is degraded by nucleases 
upon prolonged fork stalling, which can be visualised as a shortening of the replication 
tracts. These conditions typically involve depleting or inhibiting certain proteins, 
collectively termed “fork protectors”. BRCA1 is one of the best-characterised fork 
144 
 
protectors: knock-down of BRCA1 or loss-of-function BRCA1 mutants result in MRE11-
dependent DNA degradation upon prolonged HU treatment. This degradation is thought 
to originate from reversed forks, and the elimination of the fork reversal pathway rescues 
nascent DNA degradation in BRCA1 mutants. As shown in Figure 43B, I could reproduce 
the above-described phenotype in MDA-MB-436 cells: shortening of newly synthesised 
DNA takes place when HU is added for 5 h after cell labelling. At the same time, 
simultaneous inhibition of USP1 rescues this phenotype. 
 
 
 
 
Figure 43: USP1 inhibition in MDA-MD-436 cells affects replication upon replication stress. 
(A) DNA fibre assay performed in MDA-MD-436 cells treated as depicted in the scheme, following 
DNA spreading and staining. At least 100 fibres per condition were quantified. (B) Nascent DNA 
degradation assay in MDA-MD-436 cells, treated as indicated in the scheme. At least 100 fibres 
per condition were quantified. 
  
145 
 
In this assay, I cannot conclude whether the effects observed following USP1 
inhibition take place due to PCNA extraction because even VCP inhibition alone rescues 
nascent DNA degradation, and the simultaneous treatment with both inhibitors acts no 
differently. At this moment, I do not know precisely why VCP inhibition rescues nascent 
DNA degradation in MDA-MB-436 cells: possible scenarios and experiments to address 
this question are described in the Discussion. Nevertheless, the presented data indicate 
that excessive chain branching and VCP-dependent removal of PCNA interfere with fork 
reversal, favouring higher replication speeds upon replication stress and preventing the 
degradation of nascent DNA upon prolonged fork stalling. Future experiments will be 
needed to analyse which pathways compensate for the loss of fork reversal in these 
conditions and whether this could affect the long-term survival of cancer cells.  
  
146 
 
3.8 Expanding the technology beyond PCNA: the “Ubiquiton” system 
3.8.1 Design of the “Ubiquiton” system 
Tailor-made ubiquitin ligases represent a powerful tool to study ubiquitin signalling, 
as they allow direct manipulation of ubiquitin chain linkage. PCNA is one of the few 
proteins for which mono- and polyubiquitylation with different linkage types have been 
shown to fulfil different functions. Nevertheless, the DNA damage response alone involves 
ubiquitylation of numerous substrates (García-Rodríguez et al., 2016), for which the 
significance of chain linkage is not understood. Therefore, the development of a system 
that would allow applying linkage-selective E3s to any substrate would be of great 
practical use. The catalytic domains of the E363, E348 and E31 highly prefer ubiquitin as a 
substrate (Section 3.1). In the PIP-E3 system, the first ubiquitin is provided to PCNA in 
vivo by RAD18/RAD6 upon replication stress in yeast. In human cells, it is present in 
significant amounts even without exogenous sources of damage. Although for certain 
proteins, such as histones, FANCD2 and RAD18, monoubiquitylation is a well-
characterised modification, for many others this is not the case and generalising the PIP-
E3 system would require solving the problem of chain initiation. Furthermore, a general 
dimerisation strategy is needed to bring the substrate and ligases in close proximity. The 
“Ubiquiton” technology provides a solution for these challenges; it allows the application 
of linkage-specific E3s to any protein of interest (Figure 44A). Chain initiation in this 
system is achieved using the split-ubiquitin principle, which is based on the ability of two 
ubiquitin halves (containing amino acids 1-34 and 35-76, respectively) to refold when 
brought in close proximity (Johnsson and Varshavsky 1994). Upon refolding, the ubiquitin 
halves mimic the tertiary structure of a wild-type ubiquitin and, therefore, can be 
recognised by the ubiquitylation machinery. When one half of ubiquitin is fused to a 
substrate and the other to a ligase, ubiquitin refolds when the interaction between the 
substrate and the ligase is induced, allowing for polyubiquitylation by an E3. In addition to 
splitting ubiquitin into two halves, two point mutations were also introduced: G76V in the 
C-terminal moiety of the split-ubiquitin to make it resistant to DUB cleavage and I13A in 
the N-terminal half to reduce background interaction of ubiquitin halves in the absence of 
a dimerisation signal (Pratt et al., 2007). Finally, the FKBP-FRB dimerisation system is 
used to bring the E3s and the substrates together. These two domains come from 
147 
 
immunophilin FKBP12 and the protein kinase mTOR and interact in the presence of the 
small molecule drug rapamycin (Chen et al., 1995). In the resulting system, a fusion of an 
N-terminal ubiquitin half to the FRB domain (hereafter called NUbo) and a fusion of the 
FKBP domain to a C-terminal ubiquitin half (hereafter called CUbo) are appended to 
substrates and E3s as single tags. Addition of rapamycin induces interaction between 
FKBP and FRB, which is followed by the re-folding of ubiquitin from its two halves. The 
E3s then extend the reformed ubiquitin in a linkage-specific manner. Figure 44B shows 
how the Ubiquiton system works for K63, K48 and M1-linked polyubiquitylation. 
Importantly, for K63- and K48-Ubiquiton, NUbo tags are applied to E3s and the CUbo tag 
to a substrate. In this case, the resulting chains are covalently attached to a substrate. 
Since, unlike K63 and K48, ubiquitin’s M1 residue is located in an NUbo tag, for the M1-
Ubiquiton, the NUbo tag is applied to a substrate and the CUbo tag to the E3s.  
 
3.8.2 Application of the Ubiquiton system to human histone H2B 
 Histone mono- and polyubiquitylation is an essential cellular response to DNA 
damage. Nevertheless, the role of ubiquitin chain linkage in histone-coupled chains is not 
yet well investigated. Therefore, I used human histone H2B as a model substrate for the 
Ubiquiton system. All constructs were cloned under the control of a CMV promoter and, if 
not originating from H. sapiens, codon-optimised for human cells. The E31 sequence was 
partially codon-optimised, as described in section 3.2.3.  
 For the K63-Ubiquiton, N-terminally His8-tagged H2B-CUbo was used as the 
substrate and NUbo-E363 as the ligase. Additionally, we co-expressed yeast Ubc13 and 
Mms2 to boost the E363 activity. Isolation of the substrate under denaturing conditions 
reveals robust polyubiquitylation upon the addition of rapamycin (Figure 45A). As 
expected, given the underlying properties of the Ubiquiton system, mutation of the K63 
acceptor lysine in the CUbo tag prevents rapamycin-induced polyubiquitylation.  
 Because both K63 and K48 are located at the C-terminal half of a split-ubiquitin, 
the same substrate as for the K63-Ubiquiton was used for the K48-Ubiquiton. Similarly to 
the PIP-E348 system, Ubc7 was co-expressed with NUbo-E348. Figure 45B shows a 
rapamycin-inducible build-up of K48-linked chains on H2B-CUbo, which depends on the 
availability of K48 of ubiquitin. 
148 
 
 Finally, the activity of E31-CUbo was detected in total cell extracts with the LUB9 
antibody, which detects linear ubiquitin linkages and the NUbo tag and, therefore, the 
unmodified NUbo-H2B as well. Figure 46 shows the polyubiquitylation of NUbo-H2B by 
E31-CUbo in the presence of rapamycin. An N-terminally blocked substrate (His6-NUbo-
H2B-His8) is a negative control for this system, similar to the K63R and K48R mutations 
in the previous experiments. It is important to note that blocking the N terminus of ubiquitin 
prevents the LUB9 antibody from recognising this protein, which complicates the 
interpretation of this control. Therefore, in order to make sure that the observed 
polyubiquitylation results from the E31 catalytic activity, I utilised the catalytic point mutant 
(C885A) of the E31. As expected, no build-up of chains is observed when this mutant is 
used. Thus, all three chain types can be assembled on a model H2B substrate, which 
validates the use of the Ubiquiton system to study ubiquitin signalling in human cells.  
 
  
149 
 
 
Figure 44 – Design of the Ubiquiton system. (A) Three principal components of the Ubiquiton 
system. (B) Schematic representation of the Ubiquiton system for the three chain types.  
150 
 
 
Figure 45: K63- and K48-Ubiquiton applied to human H2B. (A) The indicated constructs for 
K63-Ubiquiton were expressed in HEK 293T cells for 24 h. Cells were treated, if indicated, with 1 
μM rapamycin for 30 min, followed by a Ni-NTA pulldown under denaturing conditions. Substrate 
polyubiquitylation and expression of the constructs was analysed by western blotting of the Ni-
NTA pulldown (left) and whole-cell extracts (right). Ponceau S serves as a loading control. (B) 
Cells were treated as in panel (A), but the K48-Ubiquiton constructs were used. 
  
151 
 
 
Figure 46: M1-Ubiquiton applied to human H2B. Indicated constructs were expressed in HEK 
293T cells for 24 h. Cells were treated, if indicated, with 1 μM rapamycin for 240 min, followed by 
western blotting of whole-cell extracts with the indicated antibodies. Ponceau S staining serves as 
a loading control. 
   
152 
 
Chapter 4 
Discussion 
 
4.1 Tailor made E3s as a new engineering approach to study the ubiquitin 
code 
There are several approaches to studying the ubiquitin code. The majority of them 
rely on the analysis of the existing ubiquitin conjugates. These include mass-
spectrometry, linkage-selective binders and a deubiquitylase-based approach 
UBICREST. All these methods allow to unravel the architecture of the ubiquitin coat on a 
given substrate without providing information on how important each particular element of 
the ubiquitin code is. Indeed, in order to understand the significance of the ubiquitin code, 
one has to be able to directly manipulate it – change the ubiquitin architecture on a 
substrate and analyse the physiological consequences of this manipulation. The task is 
challenging because the writers of the ubiquitin code are often unknown. In addition, 
ubiquitin ligases often have multiple functions and, therefore, cannot be depleted without 
severe side effects for a cell. Therefore, protein engineering approaches, which allow to 
re-create specific ubiquitin structures, are of great importance. There are a lot of known 
examples where permanent fusions of ubiquitin to a protein successfully mimic 
monoubiquitylation events (Asimaki et al., 2022). This approach, however, has relatively 
limited scope in mimicking polyubiquitylation. Whereas conformations of head-to-tail 
ubiquitin functions can be similar to that of K63-linked chains, other linkage types are 
unlikely to be functionally replaced by linear polyubiquitin fusions. The present work 
focuses on two protein engineering-based systems, which allow polyubiquitylation of 
desired substrates with K63-, K48- or M1-linked polyubiquitin chains. The first system – 
PIP-E3s – is specific for PCNA and was successfully implemented before in S. cerevisiae. 
In this study, PIP-E3s were optimised for human cells. Whereas PIP-E363 did not require 
further optimisation steps except for optimisation of codon usage, I observed that the 
activity of PIP-E348 drastically increased in the presence of its cognate E2 Ubc7, which 
can be co-overexpressed or produced as a self-cleavable or a permanent fusion with PIP-
E348. Unlike PIP-E363 and PIP-E348, the activity of PIP-E31 was detectable only in 
153 
 
denaturing conditions and upon overexpression of ubiquitin. This may reflect that linear 
chains are more prone to DUB cleavage in human cells or that this enzyme needs further 
optimisation to become as active as K63- and K48-specific E3s. The fundamental 
difference between PIP-E3s and other approaches, where E3s are targeted to the 
substrates, is their elongation specificity and, therefore, reliance on PCNA 
monoubiquitylation. One can envision the development of similar tailor-made E3s by 
means of fusing linkage-selective E3s to substrate binding domains for substrates other 
than PCNA. In order to promote the activity of E3s, the substrates need to be 
monoubiquitylated in vivo, which limits the scope of potential candidates. The second 
system I used in this study is the Ubiquiton technology, which can be applied to any 
desired protein and overcomes the abovementioned problem of chain initiation. While 
catalytic domains remain the same as in PIP-E3s, several advances have been 
implemented in the system: the FKBP-FRB dimerisation system as a means of substrate 
targeting and the split ubiquitin approach for chain initiation. In this thesis, I demonstrate 
the applicability of the Ubiquiton system to the human histone H2B, and my co-workers 
Evrydiki Asimaki and Christian Renz have successfully used it to polyubiquitylate human 
EGFR and GFP as a proof-of-concept substrate. Unlike PIP-E3s, overexpression of E2s 
was needed for both ‘K63’ and ‘K48’-Ubiquiton. One possible explanation for this 
phenomenon could lie in the different availability of the substrates in these systems. While 
PIP-E3s can only act on monoubiquitylated PCNA, which is a relatively rare PCNA 
modification in unchallenged conditions, the substrate for the Ubiquiton system was 
overexpressed. In this case, more components of the ubiquitylation machinery may be 
needed to support the ubiquitylation of the substrate, explaining the need for E2 
overexpression.  
There are several described approaches, where ubiquitin ligases are targeted to 
the substrates of interest. One well-known example is proteolysis-targeting chimaeras 
(PROTACs) – heterobifunctional molecules, which bring the desired substrate and the E3 
enzyme in close proximity. A wide range of ubiquitin ligases has been used as a ‘degrader’ 
component of the PROTAC system (Békés et al., 2022). In contrast PROTACs, the 
Ubiquiton system allows not only to degrade proteins but also to study numerous non-
degradative functions of K63- and M1- polyubiquitin chains. Even in the context of protein 
degradation, the Ubiquiton system provides an alternative way for targeted protein 
154 
 
degradation based on the addition of rapamycin, thereby allowing, in combination with 
other technologies, independent degradation of several cellular proteins.  
Another example of a technology which is based on the targeted protein 
ubiquitylation is the ProxE3 system: catalytic domain of the HECT-type K63-selective E3 
NEDD4 was in an inducible manner targeted to the mitochondria-localised substrate 
(Richard et al., 2020). The authors demonstrate that the activity of the E3 results in the 
sequestration of mitochondria but not mitophagy induction. However, although the 
catalytic domain of NEDD4 is predominantly K63-selective, it also multi-monoubiquitylates 
the substrate, making it difficult to differentiate between multi-mono and polyubiquitylation. 
In the Ubiquiton system, this is achieved by the split ubiquitin system: mutation of the 
acceptor lysine on a Ubiquiton tag prevents chain formation while retaining the re-folding 
of the first ubiquitin. Thus, tailor-made E3s and, in particular, the Ubiquiton system provide 
a series of conceptual advances over the existing ubiquitin technologies and have a great 
potential to explore the function of the ubiquitin code.  
 
4.2 Exploring the roles of PCNA polyubiquitylation by means of PIP-E3s 
4.2.1 Which biological questions are addressed by the PIP-E3 system in yeast and 
human cells? 
Despite the same mechanism of PIP-E3s’ activity in yeast in human cells, 
fundamental biological questions addressed by the expression of these artificial enzymes 
are different. The underlying reason is the extent of PCNA monoubiquitylation in cycling 
cells in undamaged conditions. In yeast, this modification is damage-inducible. Therefore, 
DNA damage is a prerequisite for PIP-E3 activity in this organism. Thus, a biological 
question for this system can be whether PIP-E363 can functionally act as Rad5 – an E3 
enzyme responsible for PCNA modification with K63-linked chains in the DNA damage 
bypass. Enzymes with alternative linkage specificity (PIP-E348 and PIP-E31) provide 
insight into the relevance of K63 linkage in the pathway. As demonstrated by Wegmann 
et al., PIP-E363 and PIP-E31 can functionally replace Rad5 in the DNA damage bypass 
pathway, whereas PIP-E348 counteracts damage bypass owing to PCNA degradation and 
exerts dominant-negative effects.  
155 
 
In human cells, PCNA is modified with monoubiquitin during normal S-phase 
progression, whereas polyubiquitylation becomes evident only upon substantial 
replication stress. The activity of PIP-E3s is therefore not damage-inducible: endogenous 
levels of PCNA monoubiquitylation are sufficient to support the activity of PIP-E3s, which, 
in the case of PIP-E363, leads to replication catastrophe as early as 6 h after the induction 
of the ligase. Therefore, it is impossible to utilize PIP-E3s to address the same biological 
question as in budding yeast. Instead, the activity of PIP-E363 simulates a damage 
response in a cell without ‘real’ DNA damage. This experimental setup is conceptually 
similar to tethering DNA damage markers to chromatin without any exogenous damage 
(Soutoglou and Misteli, 2008). As demonstrated by Soutoglou et al., forced recruitment of 
DSB repair proteins NBS1, MRE11, MDC1, and ATM to chromatin leads to G2/M arrest 
of cells, although ‘real’ DSBs are not present. In this sense, polyubiquitylation of PCNA by 
PIP-E363 in otherwise unchallenged cells creates a signal, which would typically appear 
in cells that experience replication stress. In theory, cells can either ignore this 
unscheduled PCNA modification or undergo a response to it. My results show that the 
second scenario is the case: human cells respond to PIP-E363-induced PCNA 
polyubiquitylation as if they would replicate over damaged DNA: they activate checkpoint 
signalling and arrest in the S phase.  Moreover, PIP-E348, albeit less efficient, leads to the 
same phenotype, highlighting the differences in the damage response pathway in yeast 
and human cells.  
 
4.2.2 Expression of PIP-E363 phenocopies HU-induced replication stress and USP1 
inhibition in BRCA1-deficient cells 
 One of the first observed phenotypes of PIP-E363 expression was S phase arrest 
and activation of checkpoint signalling. Whereas checkpoint activation begins as early as 
6 h post PIP-E363 induction, cells exhibit S-phase accumulation starting from 12 h after 
expression of PIP-E363. Alongside S-phase arrest, late (> 24 h) response to PIP-E363 
expression includes reduced EdU incorporation, reduced replication speed and formation 
of double-strand breaks, visualised by the appearance of γH2AX signal. Notably, at early 
time points, ATR activity prevents the formation of DNA breaks, as ATR inhibition by VE-
821 leads to increase of γH2AX 6 h after PIP-E363 expression. All the observed 
156 
 
phenotypes are common for cells experiencing replication stress. Indeed, treatment of 
cells with high-dose hydroxyurea leads to activation of checkpoint signalling due to 
exposure of single-stranded DNA, reduced replication speed, S-phase arrest and 
formation of DSBs upon prolonged treatment. Moreover, at early time points, ATR 
prevents the formation of DSBs by preventing excessive origin firing and RPA exhaustion. 
This similarity between PIP-E363 and HU treatment is surprising: treatment with HU affects 
all replication forks simultaneously, whereas PIP-E363 creates a chain only at the sites 
where PCNA is monoubiquitylated. As monoubiquitylated PCNA levels are relatively low, 
it suggests that K63-linked polyubiquitylation is a potent signal for checkpoint activation. 
One crucial difference between HU-induced replication stress and PIP-E363 expression is 
PCNA levels: activation of checkpoint signalling in PIP-E363 occurs concomitantly with a 
drastic decrease in total PCNA levels, whereas massive PCNA degradation upon HU 
treatment has not been reported. 
 Another situation which affects cells in a way similar to PIP-E363 expression is the 
inhibition of USP1 in BRCA1-deficient cells: this treatment results in RAD18-dependent S 
phase arrest and checkpoint activation, ultimately leading to cell death. Although the 
importance of BRCA1 deficiency in this scenario is poorly understood and will be 
discussed in section 4.2.6, USP1 inhibition leads to an increase in PCNA 
monoubiquitylation. As monoubiquitylated PCNA is a substrate for subsequent 
polyubiquitylation, inhibition of USP1 in a BRCA1-deficient background is likely to induce 
replication collapse via the same mechanism as PIP-E363 expression. Thus, expression 
of PIP-E363 leads to similar effects as replication stress triggered by at least two different 
sources.   
 
4.2.3 Effects of PIP-E363 expression are on-target 
In the course of the work, I performed a series of controls to prove that the observed 
effects are PCNA-specific. PIP-E363-induced checkpoint activation and cell cycle arrest 
do not take place if: 
1. PIP-box of the ligase is deleted, interaction with UBC13 is compromised via a point 
mutation of a RING domain, or the catalytic domain is completely deleted 
157 
 
2. PCNA K164R cell line is used, deficient in PCNA ubiquitylation 
3. RAD18 is depleted, preventing PCNA monoubiquitylation 
Taken together, these data demonstrate that unscheduled PCNA polyubiquitylation 
with K63-linked chains leads to replication catastrophe. 
Results from HEK293T-based cell lines (George-Lucian Moldovan laboratory) 
provide additional information about the nature of the adverse effects of PIP-E363 
expression. Whereas HEK293T WT cells do, and HEK293T PCNA K164R do not activate 
checkpoint upon expression of PIP-E363, cells expressing both wild-type and mutant 
PCNA forms behave like full mutant cells. One possible explanation could be that multiple 
subunits of the same PCNA trimer need to be polyubiquitylated to lead to the observed 
phenotype. Several studies indicate that simultaneous ubiquitylation of multiple PCNA 
subunits of the same trimer may be essential for damage tolerance in human cells. First, 
human cells expressing two different PCNA forms – WT and PCNA K164R – are sensitive 
to UV and cisplatin, suggesting that heterotrimeric PCNA forms interfere with damage 
bypass (Kanao and Masutani, 2017). Second, HLTF preferentially polyubiquitylates PCNA 
trimers, pre-monoubiquitylated on all three subunits (Masuda et al., 2018).  Finally, 
HUWE1, identified in this study as an interactor of polyubiquitylated PCNA, was shown to 
amplify the ubiquitin signal various of substrates in different cellular compartments (Zhou 
et al., 2023). One could also speculate that the activity of HUWE1 would include the 
propagation of ubiquitin signalling from one PCNA subunit onto others. This needs to be, 
however, further investigated.  
 
4.2.4 Ubiquitylated PCNA is a VCP substrate 
Depletion of the two factors that have been previously shown to interact 
preferentially with polyubiquitylated PCNA – ZRANB3 and WRNIP1 – do not prevent 
checkpoint activation by PIP-E363. Moreover, as there might be redundancy between 
ZRANB3 and other fork remodellers, I tested whether HLTF and SMARCAL1, as well as 
the interactor of K63-linked polyubiquitin chains RAP80, are responsible for the harmful 
effects of PIP-E363 expression. However, none of these factors seems to be involved in 
the pathway, suggesting the presence of other readers of PCNA polyubiquitylation. This 
158 
 
was a reason to conduct a mass-spectrometry-based identification of factors that 
preferentially interact with polyubiquitylated PCNA in vivo. Alongside WRNIP1, which 
serves as a positive control, I identified components of a VCP-proteasome system 
enriched in polyubiquitylated PCNA conditions. Importantly, VCP has been shown to 
regulate the disassembly of replisomes upon replication termination, and several studies 
suggest an involvement of VCP in the process of DSB repair (Meerang et al., 2011). 
However, the functions of VCP in the DNA damage bypass pathway have not yet been 
reported. VCP inhibition by NMS-873 indeed prevents checkpoint signalling after PIP-E363 
expression. Moreover, one can observe a massive accumulation of polyubiquitylated 
PCNA on chromatin upon VCP inhibition. These data pinpoint VCP activity towards 
polyubiquitylated PCNA as the reason for replication collapse after PCNA 
polyubiquitylation. Importantly, this pathway is not an artefact of the PIP-E363 system and 
also operates when PCNA polyubiquitylation is induced by HU, MMS or UV light, 
suggesting that removal of polyubiquitylated PCNA from chromatin is a general response 
to replication stress.  
 
4.2.5 PCNA is modified with branched ubiquitin chains 
Mass-spectrometry-based analysis of PCNA, immunoprecipitated from PIP-E363-
expressing cells upon VCP inhibition, reveals upregulation of K63, K48 and K11 linkages. 
Interestingly, this combination of ubiquitin linkages on a single substrate has already been 
reported for several proteins. For example, the ubiquitin coat, consisting of K63, K48 and 
K11 chains, is created by the APC/C complex on the cyclin B1, mediating its rapid 
degradation at the anaphase onset (Kirkpatrick et al., 2006). Similar ubiquitin architecture 
is present on the secretory factor EVI/WLS, the responsible ubiquitin-conjugating 
enzymes being UBE2N, UBE2K and UBE2J2 (Wolf et al., 2021). As UBE2N and UBE2K 
are highly K63- and K48-linkage specific, UBE2J2 is likely responsible for generating the 
K11 component of the ubiquitin code. CRISPR screen also identified that UBE2K, UBE2J2 
and a K63-specific ubiquitin ligase MARCH5 are responsible for the degradation of the 
pro-apoptotic factor NOXA. Importantly, UBE2K and UBE2J2 do not mediate NOXA 
degradation in MARCH5-deficient cells, suggesting that a K63-linked chain serves as a 
basis for the build-up of a complex branched K63-K48-K11 architecture (Nakao et al., 
159 
 
2023). The same pathway likely underlies the mechanism of action of cIAP-based 
degraders: a K63-linked chain is further modified with K48- and K11-linkages, leading to 
rapid degradation of a target protein (Akizuki et al., 2023). Therefore, the branching of 
K63-linked chains on PCNA may not be specific to PCNA but rather represent a particular 
case of a general pathway initiated by K63-linked chains.  
 The model that K48 ubiquitin linkages on PCNA after PIP-E363 expression are 
assembled on top of K63-linked chains was further confirmed by means of the UBICREST 
assay, where linkage-selective DUBs are used to delineate the architecture of the ubiquitin 
conjugates. Treatment of PCNA, immunoprecipitated from PIP-E363-expressing cells, with 
a K63-linkage selective DUB AMSH leads to the removal of both K63 and K48 linkages 
from PCNA. Analysis of soluble ubiquitin forms after AMSH cleavage confirmed that 
PCNA is decorated with K63-K48 branched ubiquitin chains. I further demonstrate that 
ubiquitin-conjugating enzyme UBE2K is at least partially responsible for the branching of 
K63-linked chains on PCNA in vitro (collaboration with Nils Krapoth) and in vivo. Transient 
depletion of UBE2K diminishes the amount of branched ubiquitin chains on PCNA and 
completely prevents checkpoint activation by PIP-E363. The latter observation is quite 
interesting because it suggests that a certain amount of ubiquitin chain branching on 
PCNA can be tolerated by cells; however, its excessive amount results in a replication 
collapse. By directly comparing the consequences of K63- versus K48-linked PCNA 
polyubiquitylation, I observed that both chain types lead to VCP-dependent replication 
collapse. However, due to UBE2K-dependent branching, K63-linked chains lead to much 
more substantial adverse effects than K48-linked chains (Figure 47). This suggests that 
branched K63-K48 chains are a better substrate for VCP-mediated extraction than 
homotypic chains, which agrees with similar findings for the branched K11-K48 chains 
(Yau et al., 2017). A recent study suggests that branched K63-K48 chains play a role in 
VCP metabolism: several VCP-interacting proteins (ATXN3, ZFAND2B, RHBDD1) 
preferentially bind to branched K63-K48 conjugates compared to homotypic K63 and K48-
linked chains. It is worth pointing out, that in the same study, VCP itself was found to 
preferentially bind to K48-linked chains. Therefore, it is likely that certain branched chain-
specific VCP adaptors are involved in the removal of PCNA from chromatin. As none of 
them was identified in the MS experiment, alternative approaches, such as crosslinking 
160 
 
MS or siRNA/CRISPR screens, are needed to identify VCP adaptors involved in removing 
polyubiquitylated PCNA from DNA. 
Another remaining question is the nature of identified K11 linkages on PCNA. 
Theoretically, these linkages can be assembled either on top of K63-linked or K48-linked 
chains. This can be understood in the future by performing a UBICREST assay with K63- 
or K48-selective DUBs and analysing the presence of K11 linkages in soluble and PCNA-
bound fractions by K11-selective antibodies or mass spectrometry. K11-selective DUB 
Cezanne and branched K48-K11 selective antibody may also be of great use in this 
direction.  
 
 
  
 
 
Figure 47: Schematic representation of the pathway, initiated by PCNA polyubiquitylation 
in the PIP-E363 and PIP-E348 systems  
 
161 
 
4.2.6 Ubiquitin chain branching underlies synthetic lethality between BRCA1 and 
USP1 
Why does inhibition of USP1 in BRCA1-deficient cells phenocopy overexpression 
of PIP-E363? Whereas USP1 inhibition increases the levels of mono- and, consequently, 
polyubiquitylated PCNA, the role of BRCA1 deficiency in this pathway is unclear. The 
difference between BRCA1-proficient and -deficient cells is evident even during otherwise 
unchallenged replication, as BRCA1-deficient cells tend to accumulate single-strand DNA 
gaps, which likely originate from PRIMPOL repriming at DNA lesions or difficult-to-
replicate structures (Cong et al., 2021; Quinet et al., 2020). A recent study suggests 
BRCA1 monoubiquitylates PCNA during unperturbed replication, promoting continuous 
DNA synthesis (Salas-Lloret et al., 2023). A similar phenotype, namely accumulation of 
daughter strand gaps, is exhibited by PCNA K164R cells, confirming the proposed role of 
BRCA1-mediated PCNA ubiquitylation in gap suppression. Studies in yeast suggest that 
PCNA can remain at the 3’ of a gap while the replisome further continues DNA synthesis 
(Daigaku et al., 2010). Increased formation of DNA gaps may result in more PCNA trimers 
being loaded on DNA at any time point: in addition to active replication forks, PCNA is 
likely to be present at the gaps in BRCA1-deficient cells, increasing DNA-bound PCNA 
pool, which can be a substrate for ubiquitylation. Once PCNA is sequentially 
monoubiquitylated, polyubiquitylated, undergone chain branching by UBE2K, extracted 
from DNA by VCP and degraded, a new PCNA trimer can be loaded on the same 
daughter-strand gap by RFC clamp loader, and the cycle can repeat (Schrecker et al., 
2022). In this scenario, total PCNA levels will be a trade-off between degradation of 
chromatin-bound PCNA and PCNA translation in the cytoplasm. Due to the presence of 
DNA gaps, the rate of chromatin-bound PCNA degradation should be higher in BRCA1-
deficient cells, explaining the observed rapid degradation of PCNA (Figure 48). Proof of 
this hypothesis would require performing the experiments described in section 3.7.3 upon 
depletion of PRIMPOL when the formation of ssDNA gaps, at least on the leading DNA 
strand, is prevented. Alternatively, if the hypothesis is correct, inhibition of VCP or 
depletion of UBE2K should prevent excessive gap formation, which can be visualised by 
DNA fibre assay with S1 nuclease treatment (Quinet et al., 2017). 
162 
 
 
Figure 48: A model explaining rapid PCNA degradation upon USP1 inhibition in BRCA1- 
deficient compared to wild-type cells. Excessive formation of daughter-strand gaps may lead 
to more PCNA being loaded onto DNA and undergo ubiquitylation and degradation. More detail 
in the main text. 
   
4.2.7 USP1 slows down replication under stressed conditions in BRCA1-deficient 
cells  
Inhibition of USP1 in BRCA1-deficient cells for 24 – 72 h results in bulk PCNA 
degradation and cell death. I hypothesised that this treatment may also affect replication 
on a shorter timescale. As described in section 1.2.3.4, once a replication fork stalls due 
to encountering a lesion or nucleotide starvation, there are multiple ways to solve this 
issue, the main ones being translesion synthesis, fork reversal and repriming. Of these 
three, only repriming is favoured upon the conditions of PCNA depletion, as both TLS and 
fork reversal require PCNA ubiquitylation. Indeed, my results suggest that fork reversal is 
impaired in BRCA1-deficient cells, as nascent DNA degradation in this cell line is 
prevented by USP1 inhibition. Previously, depletion of USP1 was shown to further 
aggravate, rather than rescue, nascent DNA degradation upon BRCA1 deficiency (Lim et 
163 
 
al., 2018). However, it should be noted that there is no discrepancy between these data. 
Conditions for DNA fibre assay in the Lim et al. study include either siRNA-mediated 
depletion of USP1 or long-term (24 h) treatment with a USP1 inhibitor. These conditions 
already induce significant PCNA degradation, which may alter the composition or 
structure of replication forks when HU is applied, enhancing the effects of BRCA1 
deficiency (Simoneau et al., 2023). On the contrary, I applied the USP1 inhibitor together 
with HU, ensuring that USP1 inhibition would not affect the replication forks before they 
encountered stress conditions. In addition, excessive repriming in BRCA1-deficient cells 
upon USP1 inhibition is supported by higher replication speed in these conditions. This 
acceleration of forks upon USP1 inhibition does not take place if VCP activity is inhibited, 
confirming the proposed role of PCNA extraction from DNA in counteracting TLS and fork 
reversal and facilitating repriming. Although strict proof of USP1 influence on fork reversal 
would require performing electron microscopy of the replication intermediates, the rescue 
of the nascent DNA degradation upon BRCA1 deficiency has been widely used as a proxy 
for impaired fork reversal. Rescue of the nascent DNA degradation in BRCA1-deficient 
cells can happen even in the presence of a functional fork reversal pathway, for example, 
if MRE11 nuclease activity is compromised. However, there is no known connection 
between USP1 and MRE11, so the hypothesis that USP1 inhibition impairs fork reversal 
in BRCA1-deficient cells seems more plausible. 
 
4.2.8 Potential clinical significance of the findings 
As PCNA is an essential factor in DNA replication and repair, compromising its 
functions may be a promising approach to inhibit the growth of tumour cells. Chang et al. 
describe a system for targeted PCNA degradation based on a CRL adaptor SPOP fused 
to a PCNA-targeting peptide Con1 (Chang et al., 2022). Once delivered into cells, Con1-
SPOP induces robust PCNA degradation, resulting in mitotic defects, mitochondria 
dysregulation and overall complete inhibition of tumour growth in vivo. These effects 
depend on PCNA degradation, as they are not observed upon stoichiometric inhibition of 
PCNA. In this study, I demonstrated that PCNA decoration with K63-linked chains leads 
to more efficient PCNA degradation than K48-linked polyubiquitylation. Therefore, there 
is a possibility that the PIP-E363 construct may be even more efficient in inhibiting tumour 
164 
 
growth than Con1-SPOP fusion. As cancer cells often experience replication stress and 
may have elevated levels of ubiquitylated PCNA at replication forks and daughter-strand 
gaps, PIP-E363 may preferentially inhibit the proliferation of cancer cells. More research is 
needed to assess the viability of this approach. 
Moreover, the results obtained in this study highlight the toxicity of PCNA 
ubiquitylation in BRCA1-deficient cells. Analysis of mutations in the factors involved in the 
pathway (RAD18, UBC13, RFWD3, UBE2K, VCP adaptors and others) may serve as a 
basis for prediction of how BRCA1-deficient tumours react to USP1, VCP or TLS inhibition 
or treatment with HU. The results may also be expanded beyond BRCA1-deficient 
tumours, as discussed in section 4.3.1.  
 
4.3 Future perspectives 
4.3.1 PCNA polyubiquitylation 
Excessive PCNA polyubiquitylation with K63-linked chains, induced either by the 
expression of PIP-E363 or by USP1 inhibition in a BRCA1-deficient background, results in 
replication collapse. As this effect is potentially helpful for drug discovery applications, it 
is crucial to explore whether it also occurs in other cellular backgrounds. My results 
suggest that ubiquitin chain branching on PCNA and its removal from chromatin take place 
even in wild-type cells as a part of a replication stress response. However, unlike BRCA1-
deficient, wild-type cells do not experience substantial PCNA degradation upon treatment 
with genotoxins. As mentioned in section 4.2.6, an increased presence of daughter-
strand gaps in BRCA1-deficient cells can be a reason for a higher PCNA degradation rate 
in this genetic background. However, there are other conditions that lead to increased 
formation of gaps during DNA replication. One example is the inhibition of PARP, which 
leads to an increase in replication speed due to the excessive accumulation of daughter-
strand gaps (Maya-Mendoza et al., 2018). As this situation is similar to BRCA1 deficiency, 
it is worth investigating whether co-treatment of wild-type cells with PARP and USP1 
inhibitors would lead to PCNA degradation and checkpoint activation. 
165 
 
Another open question is whether the PCNA-Ubn – UBE2K – VCP pathway protects 
replication forks from DNA2-mediated degradation. Indeed, the reason for the fork 
deprotection in PCNA K164R cell line or upon depletion of RAD18 or UBC13 is not yet 
identified. One possible explanation could be that VCP-dependent removal of PCNA from 
stressed replication forks is required for the proper fork reversal and protection. PCNA 
unloading by ATAD5 is essential for the fork reversal pathway in vitro. It is possible that 
PCNA unloading in vivo may involve other factors, such as VCP. 
Finally, since the translesion synthesis polymerase Polκ preferentially associates 
with PCNA modified with K48 ubiquitin linkages, it is tempting to speculate that ubiquitin 
chain branching by UBE2K may serve as a recruitment signal for this polymerase, 
facilitating fork restart after prolonged HU treatment. Future research is needed to address 
this question and investigate the timely coordination between the signalling functions of 
PCNA-conjugated ubiquitin chains and VCP-dependent termination of the signalling. 
 
4.3.2 The Ubiquiton technology 
 The Ubiquiton system generalises the PIP-E3 technology, allowing the 
polyubiquitylation of any protein of interest with one of three (M1-, K63-, K48-linked) chain 
types. Although the ubiquitin chains in the Ubiquiton system are assembled on the 
respective tags instead of native attachment sites, there are no known cases where an 
attachment site of a polyubiquitin chain would impact its functions on a substrate. I 
demonstrated that the Ubiquiton system can be applied to polyubiquitylate human histone 
H2B with three different linkage types. This experiment not only serves as a proof-of-
concept for the Ubiquiton technology in human cells but also allows to explore the role of 
ubiquitin signalling in the DSB response. Future experiments can include the modification 
of histones with polyubiquitin chains in the absence of ‘real’ DNA damage and analysing 
of a cellular response to this chromatin modification. Should this approach be 
unsuccessful, one can target the Ubiquiton components to chromatin by other means, 
such as the TetO-TetR tethering system or non-catalytic Cas9. Since ubiquitin signalling 
is involved in all major cellular pathways, the Ubiquiton technology will likely find 
applications beyond DNA damage signalling. It will be useful to study the roles of 
polyubiquitylation of those substrates that are known to be decorated with different 
166 
 
linkages. Examples are p53, RIPK1 and IRAK1 (Guo et al., 2021; Newton et al., 2008). 
Combining the Ubiquiton system with the proximity labelling mass spectrometry approach 
will allow to identify the interactors of the polyubiquitylated substrates. 
 Another possible application of the Ubiquiton system originates from the results of 
this work, namely the branching of K63-linked chains on PCNA. The apparent absence of 
a PCNA-targeting domain in UBE2K suggests that K63-linked chains on other substrates 
may also be decorated by UBE2K with K48 linkages. Indeed, nearly one-fifth of K63 
linkages are branched at K48, although the contribution of UBE2K to the ‘bulk’ branching 
of K63-linked chains is yet unknown. This question can be answered with the help of the 
Ubiquiton system: by assembling K63-linked chains on a panel of substrates (including 
non-physiological ones, e.g., GFP) and analysing their modification in vivo, one can 
explore the substrate specificity of UBE2K and other chain-branching enzymes. 
The expansion of the Ubiquiton technology in terms of other non-canonical linkages 
is one of the priorities of our laboratory. Several promising candidates will be characterised 
in vitro and in vivo, aiming to cover all 8 linkage types in the future. 
Finally, the Ubiquiton technology can be combined with the Sortase-based 
approach in order to site-specifically polyubiquitylate a protein of interest. Sortase-
mediated ubiquitylation, in combination with genetic code expansion, has been 
successfully applied to attach ubiquitin to the internal residues of target proteins (Fottner 
et al., 2019). Subsequent recruitment of tailor-made E3s would allow site-specific 
attachment of polyubiquitin chains, eliminating the need to attach tags to proteins. Thus, 
we anticipate further development of the Ubiquiton technology, leading to its broader 
applicability to different cellular pathways. 
  
167 
 
Chapter 5 
Appendix 
 
5.1 Abbreviations 
 
aa   Amino acid 
ADP   Adenosine-5’-diphosphate 
APC   Anaphase-promoting complex 
ATM   Ataxia-telangiectasia mutated 
ATP    Adenosine-5’-diphosphate 
ATR   Ataxia telangiectasia and Rad3 related 
AQUA   Absolute quantification 
BER   Base excision repair 
bp   Base pair 
BPDE   Benzo[a]pyrene dihydrodiol epoxide 
BRCA   Breast cancer 
BrdU   5-Bromo-2'-deoxyuridine 
BSA   Bovine serum albumine 
CHIP   Chromatin immunoprecipitation 
CldU   5-Chloro-2′-deoxyuridine 
CMV   Cytomegalovirus 
CPD   Cyclobutane pyrimidine dimer 
DMEM   Dulbecco's Modified Eagle Medium 
DNA   Deoxyribonucleic acid 
dNTP   Deoxynucleotide triphosphate 
Dox   Doxycycline 
DPC   DNA-protein crosslink 
DSB   Double-strand break 
dsDNA   Double-stranded deoxyribonucleic acid 
168 
 
DUB   Deubiquitylating enzyme 
EDTA   Ethylenediaminetetraacetic acid 
EdU   5-Ethynyl-2'-deoxyuridine 
hco   Human codon-optimised 
HECT   Homologous to E6-AP carboxyl terminus 
HIRAN   HIP116 Rad5p N-terminal 
HLTF    Helicase-like transcription factor 
HR   Homologous recombination 
hTERT   Human telomerase reverse transcriptase 
HU   Hydroxyurea 
ICL   Interstrand crosslink 
IdU   5-Iodo-2′-deoxyuridine 
IF   Immunofluorescence 
LDD   Linear chain determination domain 
LUBAC  Linear ubiquitin chain assembly complex 
MCM   Minichromosome maintenance protein complex 
MMS   Methyl methanesulfonate 
MNNG   N-Methyl-N'-nitro-N-nitrosoguanidin 
MVB   Multivesicular body 
NEMO   NF-κB essential modulator 
NER   Nucleotide excision repair 
NF-κB   Nuclear factor kappa-light-chain-enhancer of activated B cells 
NHEJ   Non-homologous end joining 
nt   Nucleotide  
NTA   Nitrilotriacetic acid 
NZF   Npl4 zinc finger 
ORC   Origin recognition complex 
PARP   Poly (ADP-Ribose) Polymerase 
PBS   Phosphate-buffered saline 
PCNA   Proliferating cell nuclear antigen 
169 
 
PEI   Polyethyleneimine 
PRIMPOL  Primase and DNA directed polymerase 
PROTAC  Proteolysis targeting chimera 
PTM   Post-translational modification 
RBR   RING-between-RING  
RING   Really interesting new gene 
RNA   Ribonucleic acid 
ROS   Reactive oxygen species 
RPA   Replication protein A 
RPE   Retinal pigment epithelium 
SDS   Sodium dodecyl sulfate 
SHPRH  SNF2 Histone Linker PHD RING Helicase 
siRNA   Small interfering ribonucleic acid 
SSB   Single-strand break 
ssDNA   Single-stranded deoxyribonucleic acid 
TLS   Translesion synthesis 
TRC   Transcription-replication conflict 
TS   Template switching 
UBAN   Ubiquitin-binding domain in ABIN proteins and NEMO 
UBD   Ubiquitin-binding domain 
UBICREST  Ubiquitin chain restriction 
UBLs   Ubiquitin-like proteins  
UBZ   Ubiquitin-binding zinc finger 
UIM   Ubiquitin-interacting motif 
UPS   Ubiquitin-Proteasome system 
UV   Ultraviolet light 
VCP   Valosin-containing protein  
VSV   Vesicular stomatitis virus    
170 
 
5.2 Publications 
Asimaki, Evrydiki; Petriukov, Kirill; Renz, Christian; Meister, Cindy; Ulrich, Helle D. 
(2022): Fast friends - Ubiquitin-like modifiers as engineered fusion partners. In Seminars 
in cell & developmental biology 132, pp. 132–145. DOI: 10.1016/j.semcdb.2021.11.013. 
Golovina, Anna Y.; Dzama, Margarita M.; Petriukov, Kirill S.; Zatsepin, Timofei S.; 
Sergiev, Petr V.; Bogdanov, Alexey A.; Dontsova, Olga A. (2014): Method for site-specific 
detection of m6A nucleoside presence in RNA based on high-resolution melting (HRM) 
analysis. In Nucleic acids research 42 (4), e27. DOI: 10.1093/nar/gkt1160. 
Oo, James A.; Pálfi, Katalin; Warwick, Timothy; Wittig, Ilka; Prieto-Garcia, Cristian; 
Matkovic, Vigor; Tomaskovic, Ines; Boos, Frederike; Ponce, Judit Izquierdo; Teichmann, 
Tom; Petriukov, Kirill et al., (2022): Long non-coding RNA PCAT19 safeguards DNA in 
quiescent endothelial cells by preventing uncontrolled phosphorylation of RPA2. In Cell 
reports 41 (7), p. 111670. DOI: 10.1016/j.celrep.2022.111670. 
Sergiev, Petr V.; Golovina, Anna Ya; Osterman, Ilya A.; Nesterchuk, Michail V.; Sergeeva, 
Olga V.; Chugunova, Anastasia A.; Evfratov, Sergey A.; Andreianova, Ekaterina S.; 
Pletnev, Philipp P.; Laptev, Ivan G.; Petriukov, Kirill S. et al., (2016): N6-Methylated 
Adenosine in RNA: From Bacteria to Humans. In Journal of molecular biology 428 (10 Pt 
B), pp. 2134–2145. DOI: 10.1016/j.jmb.2015.12.013. 
Wong, Ronald P.; Petriukov, Kirill; Ulrich, Helle D. (2021): Daughter-strand gaps in DNA 
replication - substrates of lesion processing and initiators of distress signalling. In DNA 
repair 105, p. 103163. DOI: 10.1016/j.dnarep.2021.103163. 
   
171 
 
5.3 Curriculum Vitae 
Personal information 
Name:   Kirill Petriukov 
Address:   Josefsstrasse 1, 55118 Mainz 
E-mail:   k.petriukov@imb-mainz.de 
Date of birth:  30.05.1995 
Place of birth:  Saint-Petersburg, Russian Federation 
 
Education 
Since 08/2018  PhD student, Institute of Molecular Biology (IMB), Mainz 
    Research group of Prof. Dr. Helle Ulrich 
Research topic: Analysing polyubiquitin chain signalling by 
means of tailor-made ubiquitin ligases 
09/2012 – 07/2018  Specialist in Chemistry (equivalent of M.Sc.) 
Final grade: 1.0 
Master thesis: Search for targets of the putative 
methyltransferase NSUN7 
09/2001 – 07/2012  High school diploma, Gymnasium 526 Saint-Petersburg 
Final grade: 1.0 
 
 
 
 
 
 
 
 
 
 
172 
 
Bibliography 
Achar, Yathish Jagadheesh; Balogh, David; Neculai, Dante; Juhasz, Szilvia; Morocz, Monika; 
Gali, Himabindu et al. (2015): Human HLTF mediates postreplication repair by its HIRAN domain-
dependent replication fork remodelling. In Nucleic acids research 43 (21), pp. 10277–10291. DOI: 
10.1093/nar/gkv896. 
Acharya, Narottam; Yoon, Jung-Hoon; Gali, Himabindu; Unk, Ildiko; Haracska, Lajos; Johnson, 
Robert E. et al. (2008): Roles of PCNA-binding and ubiquitin-binding domains in human DNA 
polymerase eta in translesion DNA synthesis. In Proceedings of the National Academy of 
Sciences of the United States of America 105 (46), pp. 17724–17729. DOI: 
10.1073/pnas.0809844105. 
Ahlstedt, Brittany A.; Ganji, Rakesh; Raman, Malavika (2022): The functional importance of VCP 
to maintaining cellular protein homeostasis. In Biochemical Society transactions 50 (5), pp. 1457–
1469. DOI: 10.1042/BST20220648. 
Akizuki, Yoshino; Morita, Mai; Mori, Yuki; Kaiho-Soma, Ai; Dixit, Shivani; Endo, Akinori et al. 
(2023): cIAP1-based degraders induce degradation via branched ubiquitin architectures. In Nature 
chemical biology 19 (3), pp. 311–322. DOI: 10.1038/s41589-022-01178-1. 
Al-Hakim, Abdallah K.; Zagorska, Anna; Chapman, Louise; Deak, Maria; Peggie, Mark; Alessi, 
Dario R. (2008): Control of AMPK-related kinases by USP9X and atypical Lys(29)/Lys(33)-linked 
polyubiquitin chains. In The Biochemical journal 411 (2), pp. 249–260. DOI: 10.1042/BJ20080067. 
Ali, Ammar A. E.; Timinszky, Gyula; Arribas-Bosacoma, Raquel; Kozlowski, Marek; Hassa, Paul 
O.; Hassler, Markus et al. (2012): The zinc-finger domains of PARP1 cooperate to recognize DNA 
strand breaks. In Nature structural & molecular biology 19 (7), pp. 685–692. DOI: 
10.1038/nsmb.2335. 
An, Ran; Jia, Yu; Wan, Baihui; Zhang, Yanfang; Dong, Ping; Li, Jing; Liang, Xingguo (2014): Non-
enzymatic depurination of nucleic acids: factors and mechanisms. In PloS one 9 (12), e115950. 
DOI: 10.1371/journal.pone.0115950. 
Asimaki, Evrydiki; Petriukov, Kirill; Renz, Christian; Meister, Cindy; Ulrich, Helle D. (2022): Fast 
friends - Ubiquitin-like modifiers as engineered fusion partners. In Seminars in cell & 
developmental biology 132, pp. 132–145. DOI: 10.1016/j.semcdb.2021.11.013. 
Bagola, Katrin; Delbrück, Maximilian von; Dittmar, Gunnar; Scheffner, Martin; Ziv, Inbal; Glickman, 
Michael H. et al. (2013): Ubiquitin binding by a CUE domain regulates ubiquitin chain formation 
by ERAD E3 ligases. In Molecular Cell 50 (4), pp. 528–539. DOI: 10.1016/j.molcel.2013.04.005. 
Bai, Gongshi; Kermi, Chames; Stoy, Henriette; Schiltz, Carl J.; Bacal, Julien; Zaino, Angela M. et 
al. (2020): HLTF Promotes Fork Reversal, Limiting Replication Stress Resistance and Preventing 
Multiple Mechanisms of Unrestrained DNA Synthesis. In Molecular Cell 78 (6), 1237-1251.e7. 
DOI: 10.1016/j.molcel.2020.04.031. 
Bai, Wenya; Huo, Siying; Li, Junjie; Shao, Jianlin (2022): Advances in the Study of the Ubiquitin-
Editing Enzyme A20. In Frontiers in pharmacology 13, p. 845262. DOI: 
10.3389/fphar.2022.845262. 
173 
 
Balogun, Fiyinfolu O.; Truman, Andrew W.; Kron, Stephen J. (2013): DNA resection proteins Sgs1 
and Exo1 are required for G1 checkpoint activation in budding yeast. In DNA repair 12 (9), 
pp. 751–760. DOI: 10.1016/j.dnarep.2013.06.003. 
Bannister, Andrew J.; Kouzarides, Tony (2011): Regulation of chromatin by histone modifications. 
In Cell research 21 (3), pp. 381–395. DOI: 10.1038/cr.2011.22. 
Baris, Yasemin; Taylor, Martin R. G.; Aria, Valentina; Yeeles, Joseph T. P. (2022): Fast and 
efficient DNA replication with purified human proteins. In Nature 606 (7912), pp. 204–210. DOI: 
10.1038/s41586-022-04759-1. 
Bębenek, Anna; Ziuzia-Graczyk, Izabela (2018): Fidelity of DNA replication-a matter of 
proofreading. In Current genetics 64 (5), pp. 985–996. DOI: 10.1007/s00294-018-0820-1. 
Beck, Halfdan; Nähse-Kumpf, Viola; Larsen, Marie Sofie Yoo; O'Hanlon, Karen A.; Patzke, 
Sebastian; Holmberg, Christian et al. (2012): Cyclin-dependent kinase suppression by WEE1 
kinase protects the genome through control of replication initiation and nucleotide consumption. 
In Molecular and cellular biology 32 (20), pp. 4226–4236. DOI: 10.1128/MCB.00412-12. 
Békés, Miklós; Langley, David R.; Crews, Craig M. (2022): PROTAC targeted protein degraders: 
the past is prologue. In Nature reviews. Drug discovery 21 (3), pp. 181–200. DOI: 
10.1038/s41573-021-00371-6. 
Ben-Saadon, Ronen; Fajerman, Ifat; Ziv, Tamar; Hellman, Ulf; Schwartz, Alan L.; Ciechanover, 
Aaron (2004): The tumor suppressor protein p16(INK4a) and the human papillomavirus 
oncoprotein-58 E7 are naturally occurring lysine-less proteins that are degraded by the ubiquitin 
system. Direct evidence for ubiquitination at the N-terminal residue. In The Journal of biological 
chemistry 279 (40), pp. 41414–41421. DOI: 10.1074/jbc.M407201200. 
Ben-Saadon, Ronen; Zaaroor, Daphna; Ziv, Tamar; Ciechanover, Aaron (2006): The polycomb 
protein Ring1B generates self atypical mixed ubiquitin chains required for its in vitro histone H2A 
ligase activity. In Molecular Cell 24 (5), pp. 701–711. DOI: 10.1016/j.molcel.2006.10.022. 
Bergoglio, V.; Magnaldo, T. (2006): Nucleotide excision repair and related human diseases. In 
Genome dynamics 1, pp. 35–52. DOI: 10.1159/000092499. 
Bienko, Marzena; Green, Catherine M.; Crosetto, Nicola; Rudolf, Fabian; Zapart, Grzegorz; Coull, 
Barry et al. (2005): Ubiquitin-binding domains in Y-family polymerases regulate translesion 
synthesis. In Science (New York, N.Y.) 310 (5755), pp. 1821–1824. DOI: 
10.1126/science.1120615. 
Bjelland, S.; Bjørås, M.; Seeberg, E. (1993): Excision of 3-methylguanine from alkylated DNA by 
3-methyladenine DNA glycosylase I of Escherichia coli. In Nucleic acids research 21 (9), 
pp. 2045–2049. DOI: 10.1093/nar/21.9.2045. 
Blastyák, András; Pintér, Lajos; Unk, Ildiko; Prakash, Louise; Prakash, Satya; Haracska, Lajos 
(2007): Yeast Rad5 protein required for postreplication repair has a DNA helicase activity specific 
for replication fork regression. In Molecular Cell 28 (1), pp. 167–175. DOI: 
10.1016/j.molcel.2007.07.030. 
Boone, Julien J. M.; Bhosle, Jaishree; Tilby, Mike J.; Hartley, John A.; Hochhauser, Daniel (2009): 
Involvement of the HER2 pathway in repair of DNA damage produced by chemotherapeutic 
agents. In Molecular cancer therapeutics 8 (11), pp. 3015–3023. DOI: 10.1158/1535-7163.MCT-
09-0219. 
174 
 
Boughton, Andrew J.; Krueger, Susan; Fushman, David (2020): Branching via K11 and K48 
Bestows Ubiquitin Chains with a Unique Interdomain Interface and Enhanced Affinity for 
Proteasomal Subunit Rpn1. In Structure (London, England : 1993) 28 (1), 29-43.e6. DOI: 
10.1016/j.str.2019.10.008. 
Branzei, Dana; Szakal, Barnabas (2016): DNA damage tolerance by recombination: Molecular 
pathways and DNA structures. In DNA repair 44, pp. 68–75. DOI: 10.1016/j.dnarep.2016.05.008. 
Branzei, Dana; Vanoli, Fabio; Foiani, Marco (2008): SUMOylation regulates Rad18-mediated 
template switch. In Nature 456 (7224), pp. 915–920. DOI: 10.1038/nature07587. 
Caldecott, Keith W. (2008): Single-strand break repair and genetic disease. In Nature reviews. 
Genetics 9 (8), pp. 619–631. DOI: 10.1038/nrg2380. 
Ceccaldi, Raphael; Rondinelli, Beatrice; D'Andrea, Alan D. (2016): Repair Pathway Choices and 
Consequences at the Double-Strand Break. In Trends in cell biology 26 (1), pp. 52–64. DOI: 
10.1016/j.tcb.2015.07.009. 
Centore, Richard C.; Yazinski, Stephanie A.; Tse, Alice; Zou, Lee (2012): Spartan/C1orf124, a 
reader of PCNA ubiquitylation and a regulator of UV-induced DNA damage response. In Molecular 
Cell 46 (5), pp. 625–635. DOI: 10.1016/j.molcel.2012.05.020. 
Champoux, J. J. (2001): DNA topoisomerases: structure, function, and mechanism. In Annual 
review of biochemistry 70, pp. 369–413. DOI: 10.1146/annurev.biochem.70.1.369. 
Chan, Kin; Resnick, Michael A.; Gordenin, Dmitry A. (2013): The choice of nucleotide inserted 
opposite abasic sites formed within chromosomal DNA reveals the polymerase activities 
participating in translesion DNA synthesis. In DNA repair 12 (11), pp. 878–889. DOI: 
10.1016/j.dnarep.2013.07.008. 
Chang, Howard H. Y.; Pannunzio, Nicholas R.; Adachi, Noritaka; Lieber, Michael R. (2017): Non-
homologous DNA end joining and alternative pathways to double-strand break repair. In Nature 
reviews. Molecular cell biology 18 (8), pp. 495–506. DOI: 10.1038/nrm.2017.48. 
Chang, Shih Chieh; Gopal, Pooja; Lim, Shuhui; Wei, Xiaona; Chandramohan, Arun; Mangadu, 
Ruban et al. (2022): Targeted degradation of PCNA outperforms stoichiometric inhibition to result 
in programed cell death. In Cell chemical biology 29 (11), 1601-1615.e7. DOI: 
10.1016/j.chembiol.2022.10.005. 
Chen, J.; Zheng, X. F.; Brown, E. J.; Schreiber, S. L. (1995): Identification of an 11-kDa FKBP12-
rapamycin-binding domain within the 289-kDa FKBP12-rapamycin-associated protein and 
characterization of a critical serine residue. In Proceedings of the National Academy of Sciences 
of the United States of America 92 (11), pp. 4947–4951. DOI: 10.1073/pnas.92.11.4947. 
Chen, Li; Madura, Kiran (2002): Rad23 promotes the targeting of proteolytic substrates to the 
proteasome. In Molecular and cellular biology 22 (13), pp. 4902–4913. DOI: 
10.1128/MCB.22.13.4902-4913.2002. 
Chen, Xiaoqing; Paudyal, Sharad C.; Chin, Re-I; You, Zhongsheng (2013): PCNA promotes 
processive DNA end resection by Exo1. In Nucleic acids research 41 (20), pp. 9325–9338. DOI: 
10.1093/nar/gkt672. 
Chiu, Roland K.; Brun, Jan; Ramaekers, Chantal; Theys, Jan; Weng, Lin; Lambin, Philippe et al. 
(2006): Lysine 63-polyubiquitination guards against translesion synthesis-induced mutations. In 
PLoS genetics 2 (7), e116. DOI: 10.1371/journal.pgen.0020116. 
175 
 
Choe, Katherine N.; Nicolae, Claudia M.; Constantin, Daniel; Imamura Kawasawa, Yuka; 
Delgado-Diaz, Maria Rocio; De, Subhajyoti et al. (2016): HUWE1 interacts with PCNA to alleviate 
replication stress. In EMBO reports 17 (6), pp. 874–886. DOI: 10.15252/embr.201541685. 
Choi, Koyi; Batke, Sabrina; Szakal, Barnabas; Lowther, Jonathan; Hao, Fanfan; Sarangi, Prabha 
et al. (2015): Concerted and differential actions of two enzymatic domains underlie Rad5 
contributions to DNA damage tolerance. In Nucleic acids research 43 (5), pp. 2666–2677. DOI: 
10.1093/nar/gkv004. 
Ciccia, Alberto; Nimonkar, Amitabh V.; Hu, Yiduo; Hajdu, Ildiko; Achar, Yathish Jagadheesh; Izhar, 
Lior et al. (2012): Polyubiquitinated PCNA recruits the ZRANB3 translocase to maintain genomic 
integrity after replication stress. In Molecular Cell 47 (3), pp. 396–409. DOI: 
10.1016/j.molcel.2012.05.024. 
Ciechanover, A.; Heller, H.; Elias, S.; Haas, A. L.; Hershko, A. (1980): ATP-dependent conjugation 
of reticulocyte proteins with the polypeptide required for protein degradation. In Proceedings of 
the National Academy of Sciences of the United States of America 77 (3), pp. 1365–1368. DOI: 
10.1073/pnas.77.3.1365. 
Ciechanover, Aaron (2005): N-terminal ubiquitination. In Methods in molecular biology (Clifton, 
N.J.) 301, pp. 255–270. DOI: 10.1385/1-59259-895-1:255. 
Cipolla, Lina; Bertoletti, Federica; Maffia, Antonio; Liang, Chih-Chao; Lehmann, Alan R.; Cohn, 
Martin A.; Sabbioneda, Simone (2019): UBR5 interacts with the replication fork and protects DNA 
replication from DNA polymerase η toxicity. In Nucleic acids research 47 (21), pp. 11268–11283. 
DOI: 10.1093/nar/gkz824. 
Cong, Ke; Peng, Min; Kousholt, Arne Nedergaard; Lee, Wei Ting C.; Lee, Silviana; Nayak, Sumeet 
et al. (2021): Replication gaps are a key determinant of PARP inhibitor synthetic lethality with 
BRCA deficiency. In Molecular Cell 81 (15), 3128-3144.e7. DOI: 10.1016/j.molcel.2021.06.011. 
Cooney, Ian; Han, Han; Stewart, Michael G.; Carson, Richard H.; Hansen, Daniel T.; Iwasa, Janet 
H. et al. (2019): Structure of the Cdc48 segregase in the act of unfolding an authentic substrate. 
In Science (New York, N.Y.) 365 (6452), pp. 502–505. DOI: 10.1126/science.aax0486. 
Couch, Frank B.; Bansbach, Carol E.; Driscoll, Robert; Luzwick, Jessica W.; Glick, Gloria G.; 
Bétous, Rémy et al. (2013): ATR phosphorylates SMARCAL1 to prevent replication fork collapse. 
In Genes & development 27 (14), pp. 1610–1623. DOI: 10.1101/gad.214080.113. 
Courtot, Lilas; Hoffmann, Jean-Sébastien; Bergoglio, Valérie (2018): The Protective Role of 
Dormant Origins in Response to Replicative Stress. In International journal of molecular sciences 
19 (11). DOI: 10.3390/ijms19113569. 
Cui, Lun; Bikard, David (2016): Consequences of Cas9 cleavage in the chromosome of 
Escherichia coli. In Nucleic acids research 44 (9), pp. 4243–4251. DOI: 10.1093/nar/gkw223. 
Daigaku, Yasukazu; Davies, Adelina A.; Ulrich, Helle D. (2010): Ubiquitin-dependent DNA 
damage bypass is separable from genome replication. In Nature 465 (7300), pp. 951–955. DOI: 
10.1038/nature09097. 
Dalgaard, Jacob Z. (2012): Causes and consequences of ribonucleotide incorporation into nuclear 
DNA. In Trends in genetics : TIG 28 (12), pp. 592–597. DOI: 10.1016/j.tig.2012.07.008. 
176 
 
Davis, Emily J.; Lachaud, Christophe; Appleton, Paul; Macartney, Thomas J.; Näthke, Inke; 
Rouse, John (2012): DVC1 (C1orf124) recruits the p97 protein segregase to sites of DNA damage. 
In Nature structural & molecular biology 19 (11), pp. 1093–1100. DOI: 10.1038/nsmb.2394. 
Deans, Andrew J.; West, Stephen C. (2011): DNA interstrand crosslink repair and cancer. In 
Nature reviews. Cancer 11 (7), pp. 467–480. DOI: 10.1038/nrc3088. 
Deng, L.; Wang, C.; Spencer, E.; Yang, L.; Braun, A.; You, J. et al. (2000): Activation of the 
IkappaB kinase complex by TRAF6 requires a dimeric ubiquitin-conjugating enzyme complex and 
a unique polyubiquitin chain. In Cell 103 (2), pp. 351–361. DOI: 10.1016/s0092-8674(00)00126-
4. 
Dewar, James M.; Walter, Johannes C. (2017): Mechanisms of DNA replication termination. In 
Nature reviews. Molecular cell biology 18 (8), pp. 507–516. DOI: 10.1038/nrm.2017.42. 
Dittmar, Gunnar; Winklhofer, Konstanze F. (2019): Linear Ubiquitin Chains: Cellular Functions and 
Strategies for Detection and Quantification. In Frontiers in chemistry 7, p. 915. DOI: 
10.3389/fchem.2019.00915. 
Dolan, M. E.; Moschel, R. C.; Pegg, A. E. (1990): Depletion of mammalian O6-alkylguanine-DNA 
alkyltransferase activity by O6-benzylguanine provides a means to evaluate the role of this protein 
in protection against carcinogenic and therapeutic alkylating agents. In Proceedings of the 
National Academy of Sciences of the United States of America 87 (14), pp. 5368–5372. DOI: 
10.1073/pnas.87.14.5368. 
Donnianni, Roberto A.; Symington, Lorraine S. (2013): Break-induced replication occurs by 
conservative DNA synthesis. In Proceedings of the National Academy of Sciences of the United 
States of America 110 (33), pp. 13475–13480. DOI: 10.1073/pnas.1309800110. 
Dynek, Jasmin N.; Goncharov, Tatiana; Dueber, Erin C.; Fedorova, Anna V.; Izrael-Tomasevic, 
Anita; Phu, Lilian et al. (2010): c-IAP1 and UbcH5 promote K11-linked polyubiquitination of RIP1 
in TNF signalling. In The EMBO journal 29 (24), pp. 4198–4209. DOI: 10.1038/emboj.2010.300. 
Edmunds, Charlotte E.; Simpson, Laura J.; Sale, Julian E. (2008): PCNA ubiquitination and REV1 
define temporally distinct mechanisms for controlling translesion synthesis in the avian cell line 
DT40. In Molecular Cell 30 (4), pp. 519–529. DOI: 10.1016/j.molcel.2008.03.024. 
Elia, Andrew E. H.; Boardman, Alexander P.; Wang, David C.; Huttlin, Edward L.; Everley, Robert 
A.; Dephoure, Noah et al. (2015): Quantitative Proteomic Atlas of Ubiquitination and Acetylation 
in the DNA Damage Response. In Molecular Cell 59 (5), pp. 867–881. DOI: 
10.1016/j.molcel.2015.05.006. 
Elstrodt, Fons; Hollestelle, Antoinette; Nagel, Jord H. A.; Gorin, Michael; Wasielewski, Marijke; 
van den Ouweland, Ans et al. (2006): BRCA1 mutation analysis of 41 human breast cancer cell 
lines reveals three new deleterious mutants. In Cancer research 66 (1), pp. 41–45. DOI: 
10.1158/0008-5472.CAN-05-2853. 
Emmerich, Christoph H.; Ordureau, Alban; Strickson, Sam; Arthur, J. Simon C.; Pedrioli, Patrick 
G. A.; Komander, David; Cohen, Philip (2013): Activation of the canonical IKK complex by K63/M1-
linked hybrid ubiquitin chains. In Proceedings of the National Academy of Sciences of the United 
States of America 110 (38), pp. 15247–15252. DOI: 10.1073/pnas.1314715110. 
Erenpreisa, Jekaterina; Cragg, Mark S. (2013): Three steps to the immortality of cancer cells: 
senescence, polyploidy and self-renewal. In Cancer cell international 13 (1), p. 92. DOI: 
10.1186/1475-2867-13-92. 
177 
 
Fei, Cong; Li, Zhenfei; Li, Chen; Chen, Yuelei; Chen, Zhangcheng; He, Xiaoli et al. (2013): Smurf1-
mediated Lys29-linked nonproteolytic polyubiquitination of axin negatively regulates Wnt/β-
catenin signaling. In Molecular and cellular biology 33 (20), pp. 4095–4105. DOI: 
10.1128/MCB.00418-13. 
Fennell, Lilian M.; Rahighi, Simin; Ikeda, Fumiyo (2018): Linear ubiquitin chain-binding domains. 
In The FEBS journal 285 (15), pp. 2746–2761. DOI: 10.1111/febs.14478. 
Finley, D.; Sadis, S.; Monia, B. P.; Boucher, P.; Ecker, D. J.; Crooke, S. T.; Chau, V. (1994): 
Inhibition of proteolysis and cell cycle progression in a multiubiquitination-deficient yeast mutant. 
In Molecular and cellular biology 14 (8), pp. 5501–5509. DOI: 10.1128/mcb.14.8.5501-5509.1994. 
Fottner, Maximilian; Brunner, Andreas-David; Bittl, Verena; Horn-Ghetko, Daniel; Jussupow, 
Alexander; Kaila, Ville R. I. et al. (2019): Site-specific ubiquitylation and SUMOylation using 
genetic-code expansion and sortase. In Nature chemical biology 15 (3), pp. 276–284. DOI: 
10.1038/s41589-019-0227-4. 
Freudenthal, Bret D.; Gakhar, Lokesh; Ramaswamy, S.; Washington, M. Todd (2010): Structure 
of monoubiquitinated PCNA and implications for translesion synthesis and DNA polymerase 
exchange. In Nature structural & molecular biology 17 (4), pp. 479–484. DOI: 10.1038/nsmb.1776. 
Fumasoni, Marco; Zwicky, Katharina; Vanoli, Fabio; Lopes, Massimo; Branzei, Dana (2015): 
Error-free DNA damage tolerance and sister chromatid proximity during DNA replication rely on 
the Polα/Primase/Ctf4 Complex. In Molecular Cell 57 (5), pp. 812–823. DOI: 
10.1016/j.molcel.2014.12.038. 
Galan, J. M.; Haguenauer-Tsapis, R. (1997): Ubiquitin lys63 is involved in ubiquitination of a yeast 
plasma membrane protein. In The EMBO journal 16 (19), pp. 5847–5854. DOI: 
10.1093/emboj/16.19.5847. 
Gallina, Irene; Hendriks, Ivo A.; Hoffmann, Saskia; Larsen, Nicolai B.; Johansen, Joachim; 
Colding-Christensen, Camilla S. et al. (2021): The ubiquitin ligase RFWD3 is required for 
translesion DNA synthesis. In Molecular Cell 81 (3), 442-458.e9. DOI: 
10.1016/j.molcel.2020.11.029. 
Gallo, David; Kim, TaeHyung; Szakal, Barnabas; Saayman, Xanita; Narula, Ashrut; Park, Yoona 
et al. (2019): Rad5 Recruits Error-Prone DNA Polymerases for Mutagenic Repair of ssDNA Gaps 
on Undamaged Templates. In Molecular Cell 73 (5), 900-914.e9. DOI: 
10.1016/j.molcel.2019.01.001. 
García-Rodríguez, Néstor; Morawska, Magdalena; Wong, Ronald P.; Daigaku, Yasukazu; Ulrich, 
Helle D. (2018a): Spatial separation between replisome- and template-induced replication stress 
signaling. In The EMBO journal 37 (9). DOI: 10.15252/embj.201798369. 
García-Rodríguez, Néstor; Wong, Ronald P.; Ulrich, Helle D. (2016): Functions of Ubiquitin and 
SUMO in DNA Replication and Replication Stress. In Frontiers in genetics 7, p. 87. DOI: 
10.3389/fgene.2016.00087. 
García-Rodríguez, Néstor; Wong, Ronald P.; Ulrich, Helle D. (2018b): The helicase Pif1 functions 
in the template switching pathway of DNA damage bypass. In Nucleic acids research 46 (16), 
pp. 8347–8356. DOI: 10.1093/nar/gky648. 
Gatti, Marco; Pinato, Sabrina; Maiolica, Alessio; Rocchio, Francesca; Prato, Maria Giulia; 
Aebersold, Ruedi; Penengo, Lorenza (2015): RNF168 promotes noncanonical K27 ubiquitination 
to signal DNA damage. In Cell reports 10 (2), pp. 226–238. DOI: 10.1016/j.celrep.2014.12.021. 
178 
 
Giannattasio, Michele; Zwicky, Katharina; Follonier, Cindy; Foiani, Marco; Lopes, Massimo; 
Branzei, Dana (2014): Visualization of recombination-mediated damage bypass by template 
switching. In Nature structural & molecular biology 21 (10), pp. 884–892. DOI: 
10.1038/nsmb.2888. 
Gómez-Pinto, Irene; Cubero, Elena; Kalko, Susana G.; Monaco, Vania; van der Marel, Gijs; van 
Boom, Jacques H. et al. (2004): Effect of bulky lesions on DNA: solution structure of a DNA duplex 
containing a cholesterol adduct. In The Journal of biological chemistry 279 (23), pp. 24552–24560. 
DOI: 10.1074/jbc.M311751200. 
Gonzalez-Santamarta, Maria; Ceccato, Laurie; Carvalho, Ana Sofia; Rain, Jean-Christophe; 
Matthiesen, Rune; Rodriguez, Manuel S. (2023): Isolation and Mass Spectrometry Identification 
of K48 and K63 Ubiquitin Proteome Using Chain-Specific Nanobodies. In Methods in molecular 
biology (Clifton, N.J.) 2602, pp. 125–136. DOI: 10.1007/978-1-0716-2859-1_9. 
Grabbe, Caroline; Dikic, Ivan (2009): Functional roles of ubiquitin-like domain (ULD) and ubiquitin-
binding domain (UBD) containing proteins. In Chemical reviews 109 (4), pp. 1481–1494. DOI: 
10.1021/cr800413p. 
Gudjonsson, Thorkell; Altmeyer, Matthias; Savic, Velibor; Toledo, Luis; Dinant, Christoffel; Grøfte, 
Merete et al. (2012): TRIP12 and UBR5 suppress spreading of chromatin ubiquitylation at 
damaged chromosomes. In Cell 150 (4), pp. 697–709. DOI: 10.1016/j.cell.2012.06.039. 
Guilliam, Thomas A.; Brissett, Nigel C.; Ehlinger, Aaron; Keen, Benjamin A.; Kolesar, Peter; 
Taylor, Elaine M. et al. (2017): Molecular basis for PrimPol recruitment to replication forks by RPA. 
In Nature Communications 8, p. 15222. DOI: 10.1038/ncomms15222. 
Guo, Caixia; Sonoda, Eiichiro; Tang, Tie-Shan; Parker, Joanne L.; Bielen, Aleksandra B.; Takeda, 
Shunichi et al. (2006): REV1 protein interacts with PCNA: significance of the REV1 BRCT domain 
in vitro and in vivo. In Molecular Cell 23 (2), pp. 265–271. DOI: 10.1016/j.molcel.2006.05.038. 
Guo, Yafei; Li, Qin; Zhao, Gang; Zhang, Jie; Yuan, Hang; Feng, Tianyu et al. (2021): Loss of 
TRIM31 promotes breast cancer progression through regulating K48- and K63-linked 
ubiquitination of p53. In Cell death & disease 12 (10), p. 945. DOI: 10.1038/s41419-021-04208-3. 
Haas, A. L.; Bright, P. M. (1985): The immunochemical detection and quantitation of intracellular 
ubiquitin-protein conjugates. In The Journal of biological chemistry 260 (23), pp. 12464–12473. 
Hagai, Tzachi; Levy, Yaakov (2010): Ubiquitin not only serves as a tag but also assists 
degradation by inducing protein unfolding. In Proceedings of the National Academy of Sciences 
of the United States of America 107 (5), pp. 2001–2006. DOI: 10.1073/pnas.0912335107. 
Haglund, Kaisa; Sigismund, Sara; Polo, Simona; Szymkiewicz, Iwona; Di Fiore, Pier Paolo; Dikic, 
Ivan (2003): Multiple monoubiquitination of RTKs is sufficient for their endocytosis and 
degradation. In Nature cell biology 5 (5), pp. 461–466. DOI: 10.1038/ncb983. 
Hahn, H.; Palmenberg, A. C. (1996): Mutational analysis of the encephalomyocarditis virus 
primary cleavage. In Journal of virology 70 (10), pp. 6870–6875. DOI: 10.1128/JVI.70.10.6870-
6875.1996. 
Halder, Swagata; Ranjha, Lepakshi; Taglialatela, Angelo; Ciccia, Alberto; Cejka, Petr (2022): 
Strand annealing and motor driven activities of SMARCAL1 and ZRANB3 are stimulated by 
RAD51 and the paralog complex. In Nucleic acids research 50 (14), pp. 8008–8022. DOI: 
10.1093/nar/gkac583. 
179 
 
Hanada, Katsuhiro; Budzowska, Magda; Davies, Sally L.; van Drunen, Ellen; Onizawa, Hideo; 
Beverloo, H. Berna et al. (2007): The structure-specific endonuclease Mus81 contributes to 
replication restart by generating double-strand DNA breaks. In Nature structural & molecular 
biology 14 (11), pp. 1096–1104. DOI: 10.1038/nsmb1313. 
Heidelberger, Jan B.; Voigt, Andrea; Borisova, Marina E.; Petrosino, Giuseppe; Ruf, Stefanie; 
Wagner, Sebastian A.; Beli, Petra (2018): Proteomic profiling of VCP substrates links VCP to K6-
linked ubiquitylation and c-Myc function. In EMBO reports 19 (4). DOI: 10.15252/embr.201744754. 
Hendel, Ayal; Ziv, Omer; Gueranger, Quentin; Geacintov, Nicholas; Livneh, Zvi (2008): Reduced 
efficiency and increased mutagenicity of translesion DNA synthesis across a TT cyclobutane 
pyrimidine dimer, but not a TT 6-4 photoproduct, in human cells lacking DNA polymerase eta. In 
DNA repair 7 (10), pp. 1636–1646. DOI: 10.1016/j.dnarep.2008.06.008. 
Henning, Nathaniel J.; Boike, Lydia; Spradlin, Jessica N.; Ward, Carl C.; Liu, Gang; Zhang, Erika 
et al. (2022): Deubiquitinase-targeting chimeras for targeted protein stabilization. In Nature 
chemical biology 18 (4), pp. 412–421. DOI: 10.1038/s41589-022-00971-2. 
Hergeth, Sonja P.; Schneider, Robert (2015): The H1 linker histones: multifunctional proteins 
beyond the nucleosomal core particle. In EMBO reports 16 (11), pp. 1439–1453. DOI: 
10.15252/embr.201540749. 
Hershko, A.; Ciechanover, A.; Heller, H.; Haas, A. L.; Rose, I. A. (1980): Proposed role of ATP in 
protein breakdown: conjugation of protein with multiple chains of the polypeptide of ATP-
dependent proteolysis. In Proceedings of the National Academy of Sciences of the United States 
of America 77 (4), pp. 1783–1786. DOI: 10.1073/pnas.77.4.1783. 
Hess, M. T.; Gunz, D.; Luneva, N.; Geacintov, N. E.; Naegeli, H. (1997): Base pair conformation-
dependent excision of benzoapyrene diol epoxide-guanine adducts by human nucleotide excision 
repair enzymes. In Molecular and cellular biology 17 (12), pp. 7069–7076. DOI: 
10.1128/MCB.17.12.7069. 
Hibbert, Richard G.; Sixma, Titia K. (2012): Intrinsic flexibility of ubiquitin on proliferating cell 
nuclear antigen (PCNA) in translesion synthesis. In The Journal of biological chemistry 287 (46), 
pp. 39216–39223. DOI: 10.1074/jbc.M112.389890. 
Hicke, L. (2001): Protein regulation by monoubiquitin. In Nature reviews. Molecular cell biology 2 
(3), pp. 195–201. DOI: 10.1038/35056583. 
Hodge, Curtis D.; Spyracopoulos, Leo; Glover, J. N. Mark (2016): Ubc13: the Lys63 ubiquitin chain 
building machine. In Oncotarget 7 (39), pp. 64471–64504. DOI: 10.18632/oncotarget.10948. 
Hoege, Carsten; Pfander, Boris; Moldovan, George-Lucian; Pyrowolakis, George; Jentsch, Stefan 
(2002): RAD6-dependent DNA repair is linked to modification of PCNA by ubiquitin and SUMO. In 
Nature 419 (6903), pp. 135–141. DOI: 10.1038/nature00991. 
Hospenthal, Manuela K.; Mevissen, Tycho E. T.; Komander, David (2015): Deubiquitinase-based 
analysis of ubiquitin chain architecture using Ubiquitin Chain Restriction (UbiCRest). In Nature 
protocols 10 (2), pp. 349–361. DOI: 10.1038/nprot.2015.018. 
Hu, Yiduo; Scully, Ralph; Sobhian, Bijan; Xie, Anyong; Shestakova, Elena; Livingston, David M. 
(2011): RAP80-directed tuning of BRCA1 homologous recombination function at ionizing 
radiation-induced nuclear foci. In Genes & development 25 (7), pp. 685–700. DOI: 
10.1101/gad.2011011. 
180 
 
Huang, Haining; Jeon, Myung-Shin; Liao, Lujian; Yang, Chun; Elly, Chris; Yates, John R. 3rd; Liu, 
Yun-Cai (2010): K33-linked polyubiquitination of T cell receptor-zeta regulates proteolysis-
independent T cell signaling. In Immunity 33 (1), pp. 60–70. DOI: 10.1016/j.immuni.2010.07.002. 
Huang, Tony T.; Nijman, Sebastian M. B.; Mirchandani, Kanchan D.; Galardy, Paul J.; Cohn, 
Martin A.; Haas, Wilhelm et al. (2006): Regulation of monoubiquitinated PCNA by DUB 
autocleavage. In Nature cell biology 8 (4), pp. 339–347. DOI: 10.1038/ncb1378. 
Husnjak, Koraljka; Dikic, Ivan (2012): Ubiquitin-binding proteins: decoders of ubiquitin-mediated 
cellular functions. In Annual review of biochemistry 81, pp. 291–322. DOI: 10.1146/annurev-
biochem-051810-094654. 
Igoucheva, O.; Alexeev, V.; Yoon, K. (2006): Differential cellular responses to exogenous DNA in 
mammalian cells and its effect on oligonucleotide-directed gene modification. In Gene therapy 13 
(3), pp. 266–275. DOI: 10.1038/sj.gt.3302643. 
Izhar, Lior; Ziv, Omer; Cohen, Isadora S.; Geacintov, Nicholas E.; Livneh, Zvi (2013): Genomic 
assay reveals tolerance of DNA damage by both translesion DNA synthesis and homology-
dependent repair in mammalian cells. In Proceedings of the National Academy of Sciences of the 
United States of America 110 (16), E1462-9. DOI: 10.1073/pnas.1216894110. 
Jha, Vikash; Bian, Chuanbing; Xing, Guangxin; Ling, Hong (2016): Structure and mechanism of 
error-free replication past the major benzoapyrene adduct by human DNA polymerase κ. In 
Nucleic acids research 44 (10), pp. 4957–4967. DOI: 10.1093/nar/gkw204. 
Johnson, E. S.; Ma, P. C.; Ota, I. M.; Varshavsky, A. (1995): A proteolytic pathway that recognizes 
ubiquitin as a degradation signal. In The Journal of biological chemistry 270 (29), pp. 17442–
17456. DOI: 10.1074/jbc.270.29.17442. 
Johnsson, N.; Varshavsky, A. (1994): Split ubiquitin as a sensor of protein interactions in vivo. In 
Proceedings of the National Academy of Sciences of the United States of America 91 (22), 
pp. 10340–10344. DOI: 10.1073/pnas.91.22.10340. 
Jones, R. M.; Mortusewicz, O.; Afzal, I.; Lorvellec, M.; García, P.; Helleday, T.; Petermann, E. 
(2013): Increased replication initiation and conflicts with transcription underlie Cyclin E-induced 
replication stress. In Oncogene 32 (32), pp. 3744–3753. DOI: 10.1038/onc.2012.387. 
Kaiho-Soma, Ai; Akizuki, Yoshino; Igarashi, Katsuhide; Endo, Akinori; Shoda, Takuji; Kawase, 
Yasuko et al. (2021): TRIP12 promotes small-molecule-induced degradation through K29/K48-
branched ubiquitin chains. In Molecular Cell 81 (7), 1411-1424.e7. DOI: 
10.1016/j.molcel.2021.01.023. 
Kanao, Rie; Masutani, Chikahide (2017): Regulation of DNA damage tolerance in mammalian 
cells by post-translational modifications of PCNA. In Mutation research 803-805, pp. 82–88. DOI: 
10.1016/j.mrfmmm.2017.06.004. 
Kang, Hyun Je; Park, Hyun; Yoo, Eun Jin; Lee, Jun Ho; Choi, Soo Youn; Lee-Kwon, Whaseon et 
al. (2019): TonEBP Regulates PCNA Polyubiquitination in Response to DNA Damage through 
Interaction with SHPRH and USP1. In iScience 19, pp. 177–190. DOI: 10.1016/j.isci.2019.07.021. 
Kanu, N.; Zhang, T.; Burrell, R. A.; Chakraborty, A.; Cronshaw, J.; DaCosta, C. et al. (2016): 
RAD18, WRNIP1 and ATMIN promote ATM signalling in response to replication stress. In 
Oncogene 35 (30), pp. 4009–4019. DOI: 10.1038/onc.2015.427. 
181 
 
Kelsall, Ian R.; Zhang, Jiazhen; Knebel, Axel; Arthur, J. Simon C.; Cohen, Philip (2019): The E3 
ligase HOIL-1 catalyses ester bond formation between ubiquitin and components of the 
Myddosome in mammalian cells. In Proceedings of the National Academy of Sciences of the 
United States of America 116 (27), pp. 13293–13298. DOI: 10.1073/pnas.1905873116. 
Khoury, George A.; Baliban, Richard C.; Floudas, Christodoulos A. (2011): Proteome-wide post-
translational modification statistics: frequency analysis and curation of the swiss-prot database. In 
Scientific reports 1. DOI: 10.1038/srep00090. 
Kim, Woong; Bennett, Eric J.; Huttlin, Edward L.; Guo, Ailan; Li, Jing; Possemato, Anthony et al. 
(2011): Systematic and quantitative assessment of the ubiquitin-modified proteome. In Molecular 
Cell 44 (2), pp. 325–340. DOI: 10.1016/j.molcel.2011.08.025. 
Kirisako, Takayoshi; Kamei, Kiyoko; Murata, Shigeo; Kato, Michiko; Fukumoto, Hiromi; Kanie, 
Masato et al. (2006): A ubiquitin ligase complex assembles linear polyubiquitin chains. In The 
EMBO journal 25 (20), pp. 4877–4887. DOI: 10.1038/sj.emboj.7601360. 
Kirkpatrick, Donald S.; Hathaway, Nathaniel A.; Hanna, John; Elsasser, Suzanne; Rush, John; 
Finley, Daniel et al. (2006): Quantitative analysis of in vitro ubiquitinated cyclin B1 reveals complex 
chain topology. In Nature cell biology 8 (7), pp. 700–710. DOI: 10.1038/ncb1436. 
Koegl, M.; Hoppe, T.; Schlenker, S.; Ulrich, H. D.; Mayer, T. U.; Jentsch, S. (1999): A novel 
ubiquitination factor, E4, is involved in multiubiquitin chain assembly. In Cell 96 (5), pp. 635–644. 
DOI: 10.1016/s0092-8674(00)80574-7. 
Komander, David; Rape, Michael (2012): The ubiquitin code. In Annual review of biochemistry 81, 
pp. 203–229. DOI: 10.1146/annurev-biochem-060310-170328. 
Komander, David; Reyes-Turcu, Francisca; Licchesi, Julien D. F.; Odenwaelder, Peter; Wilkinson, 
Keith D.; Barford, David (2009): Molecular discrimination of structurally equivalent Lys 63-linked 
and linear polyubiquitin chains. In EMBO reports 10 (5), pp. 466–473. DOI: 
10.1038/embor.2009.55. 
Krieg, Adam J.; Hammond, Ester M.; Giaccia, Amato J. (2006): Functional analysis of p53 binding 
under differential stresses. In Molecular and cellular biology 26 (19), pp. 7030–7045. DOI: 
10.1128/MCB.00322-06. 
Krokan, Hans E.; Bjørås, Magnar (2013): Base excision repair. In Cold Spring Harbor perspectives 
in biology 5 (4), a012583. DOI: 10.1101/cshperspect.a012583. 
Lai, Z.; Ferry, K. V.; Diamond, M. A.; Wee, K. E.; Kim, Y. B.; Ma, J. et al. (2001): Human mdm2 
mediates multiple mono-ubiquitination of p53 by a mechanism requiring enzyme isomerization. In 
The Journal of biological chemistry 276 (33), pp. 31357–31367. DOI: 10.1074/jbc.M011517200. 
Lange, Sven M.; McFarland, Matthew R.; Lamoliatte, Frederic; Kwaśna, Dominika; Shen, Linnan; 
Wallace, Iona et al. (2023): Comprehensive approach to study branched ubiquitin chains reveals 
roles for K48-K63 branches in VCP/p97-related processes. In Biorxiv, Article 
https://doi.org/10.1101/2023.01.10.523363. 
Lanz, Michael Charles; Dibitetto, Diego; Smolka, Marcus Bustamante (2019): DNA damage 
kinase signaling: checkpoint and repair at 30 years. In The EMBO journal 38 (18), e101801. DOI: 
10.15252/embj.2019101801. 
182 
 
Lau, Wilson C. Y.; Li, Yinyin; Zhang, Qinfen; Huen, Michael S. Y. (2015): Molecular architecture 
of the Ub-PCNA/Pol η complex bound to DNA. In Scientific reports 5, p. 15759. DOI: 
10.1038/srep15759. 
Lechtenberg, Bernhard C.; Rajput, Akhil; Sanishvili, Ruslan; Dobaczewska, Małgorzata K.; Ware, 
Carl F.; Mace, Peter D.; Riedl, Stefan J. (2016): Structure of a HOIP/E2~ubiquitin complex reveals 
RBR E3 ligase mechanism and regulation. In Nature 529 (7587), pp. 546–550. DOI: 
10.1038/nature16511. 
Lee, D. H.; Goldberg, A. L. (1998): Proteasome inhibitors: valuable new tools for cell biologists. In 
Trends in cell biology 8 (10), pp. 397–403. DOI: 10.1016/s0962-8924(98)01346-4. 
Lee, Sangho; Tsai, Yien Che; Mattera, Rafael; Smith, William J.; Kostelansky, Michael S.; 
Weissman, Allan M. et al. (2006): Structural basis for ubiquitin recognition and autoubiquitination 
by Rabex-5. In Nature structural & molecular biology 13 (3), pp. 264–271. DOI: 
10.1038/nsmb1064. 
Lemaçon, Delphine; Jackson, Jessica; Quinet, Annabel; Brickner, Joshua R.; Li, Shan; Yazinski, 
Stephanie et al. (2017): MRE11 and EXO1 nucleases degrade reversed forks and elicit MUS81-
dependent fork rescue in BRCA2-deficient cells. In Nature Communications 8 (1), p. 860. DOI: 
10.1038/s41467-017-01180-5. 
Lerner, Leticia K.; Francisco, Guilherme; Soltys, Daniela T.; Rocha, Clarissa R. R.; Quinet, 
Annabel; Vessoni, Alexandre T. et al. (2017): Predominant role of DNA polymerase eta and p53-
dependent translesion synthesis in the survival of ultraviolet-irradiated human cells. In Nucleic 
acids research 45 (3), pp. 1270–1280. DOI: 10.1093/nar/gkw1196. 
Leuzzi, Giuseppe; Marabitti, Veronica; Pichierri, Pietro; Franchitto, Annapaola (2016): WRNIP1 
protects stalled forks from degradation and promotes fork restart after replication stress. In The 
EMBO journal 35 (13), pp. 1437–1451. DOI: 10.15252/embj.201593265. 
Li, Yanchang; Dammer, Eric B.; Gao, Yuan; Lan, Qiuyan; Villamil, Mark A.; Duong, Duc M. et al. 
(2019): Proteomics Links Ubiquitin Chain Topology Change to Transcription Factor Activation. In 
Molecular Cell 76 (1), 126-137.e7. DOI: 10.1016/j.molcel.2019.07.001. 
Liang, Ruei-Yue; Chen, Li; Ko, Bo-Ting; Shen, Yu-Han; Li, Yen-Te; Chen, Bo-Rong et al. (2014): 
Rad23 interaction with the proteasome is regulated by phosphorylation of its ubiquitin-like (UbL) 
domain. In Journal of molecular biology 426 (24), pp. 4049–4060. DOI: 
10.1016/j.jmb.2014.10.004. 
Lim, Kah Suan; Li, Heng; Roberts, Emma A.; Gaudiano, Emily F.; Clairmont, Connor; Sambel, 
Larissa Alina et al. (2018): USP1 Is Required for Replication Fork Protection in BRCA1-Deficient 
Tumors. In Molecular Cell 72 (6), 925-941.e4. DOI: 10.1016/j.molcel.2018.10.045. 
Lindström, Mikael S.; Jurada, Deana; Bursac, Sladana; Orsolic, Ines; Bartek, Jiri; Volarevic, Sinisa 
(2018): Nucleolus as an emerging hub in maintenance of genome stability and cancer 
pathogenesis. In Oncogene 37 (18), pp. 2351–2366. DOI: 10.1038/s41388-017-0121-z. 
Liu, Shengqin; Opiyo, Stephen O.; Manthey, Karoline; Glanzer, Jason G.; Ashley, Amanda K.; 
Amerin, Courtney et al. (2012): Distinct roles for DNA-PK, ATM and ATR in RPA phosphorylation 
and checkpoint activation in response to replication stress. In Nucleic acids research 40 (21), 
pp. 10780–10794. DOI: 10.1093/nar/gks849. 
183 
 
Liu, W.; Krishnamoorthy, A.; Zhao, R.; Cortez, D. (2020): Two replication fork remodeling 
pathways generate nuclease substrates for distinct fork protection factors. In Science advances 6 
(46). DOI: 10.1126/sciadv.abc3598. 
Liu, Xiaoping; Xu, Bosen; Yang, Jianguo; He, Lin; Zhang, Zihan; Cheng, Xiao et al. (2021): UHRF2 
commissions the completion of DNA demethylation through allosteric activation by 5hmC and K33-
linked ubiquitination of XRCC1. In Molecular Cell 81 (14), 2960-2974.e7. DOI: 
10.1016/j.molcel.2021.05.022. 
Lu, Chunwan; Yang, Dafeng; Sabbatini, Maria E.; Colby, Aaron H.; Grinstaff, Mark W.; Oberlies, 
Nicholas H. et al. (2018): Contrasting roles of H3K4me3 and H3K9me3 in regulation of apoptosis 
and gemcitabine resistance in human pancreatic cancer cells. In BMC cancer 18 (1), p. 149. DOI: 
10.1186/s12885-018-4061-y. 
Lu, Ying; Lee, Byung-Hoon; King, Randall W.; Finley, Daniel; Kirschner, Marc W. (2015): 
Substrate degradation by the proteasome: a single-molecule kinetic analysis. In Science (New 
York, N.Y.) 348 (6231), p. 1250834. DOI: 10.1126/science.1250834. 
MacDougall, Christina A.; Byun, Tony S.; Van, Christopher; Yee, Muh-ching; Cimprich, Karlene A. 
(2007): The structural determinants of checkpoint activation. In Genes & development 21 (8), 
pp. 898–903. DOI: 10.1101/gad.1522607. 
Maga, G.; Villani, G.; Tillement, V.; Stucki, M.; Locatelli, G. A.; Frouin, I. et al. (2001): Okazaki 
fragment processing: modulation of the strand displacement activity of DNA polymerase delta by 
the concerted action of replication protein A, proliferating cell nuclear antigen, and flap 
endonuclease-1. In Proceedings of the National Academy of Sciences of the United States of 
America 98 (25), pp. 14298–14303. DOI: 10.1073/pnas.251193198. 
Magnaghi, Paola; D'Alessio, Roberto; Valsasina, Barbara; Avanzi, Nilla; Rizzi, Simona; Asa, 
Daniela et al. (2013): Covalent and allosteric inhibitors of the ATPase VCP/p97 induce cancer cell 
death. In Nature chemical biology 9 (9), pp. 548–556. DOI: 10.1038/nchembio.1313. 
Markkanen, Enni (2017): Not breathing is not an option: How to deal with oxidative DNA damage. 
In DNA repair 59, pp. 82–105. DOI: 10.1016/j.dnarep.2017.09.007. 
Martin, Sara K.; Wood, Richard D. (2019): DNA polymerase ζ in DNA replication and repair. In 
Nucleic acids research 47 (16), pp. 8348–8361. DOI: 10.1093/nar/gkz705. 
Martinez-Fonts, Kirby; Davis, Caroline; Tomita, Takuya; Elsasser, Suzanne; Nager, Andrew R.; 
Shi, Yuan et al. (2020): The proteasome 19S cap and its ubiquitin receptors provide a versatile 
recognition platform for substrates. In Nature Communications 11 (1), p. 477. DOI: 
10.1038/s41467-019-13906-8. 
Masuda, Yuji; Mitsuyuki, Satoshi; Kanao, Rie; Hishiki, Asami; Hashimoto, Hiroshi; Masutani, 
Chikahide (2018): Regulation of HLTF-mediated PCNA polyubiquitination by RFC and PCNA 
monoubiquitination levels determines choice of damage tolerance pathway. In Nucleic acids 
research 46 (21), pp. 11340–11356. DOI: 10.1093/nar/gky943. 
Matsumoto, Marissa L.; Dong, Ken C.; Yu, Christine; Phu, Lilian; Gao, Xinxin; Hannoush, Rami N. 
et al. (2012): Engineering and structural characterization of a linear polyubiquitin-specific antibody. 
In Journal of molecular biology 418 (3-4), pp. 134–144. DOI: 10.1016/j.jmb.2011.12.053. 
Matsumoto, Marissa L.; Wickliffe, Katherine E.; Dong, Ken C.; Yu, Christine; Bosanac, Ivan; 
Bustos, Daisy et al. (2010): K11-linked polyubiquitination in cell cycle control revealed by a K11 
184 
 
linkage-specific antibody. In Molecular Cell 39 (3), pp. 477–484. DOI: 
10.1016/j.molcel.2010.07.001. 
Maxwell, Brian A.; Gwon, Youngdae; Mishra, Ashutosh; Peng, Junmin; Nakamura, Haruko; 
Zhang, Ke et al. (2021): Ubiquitination is essential for recovery of cellular activities after heat 
shock. In Science (New York, N.Y.) 372 (6549), eabc3593. DOI: 10.1126/science.abc3593. 
Maya-Mendoza, Apolinar; Moudry, Pavel; Merchut-Maya, Joanna Maria; Lee, MyungHee; 
Strauss, Robert; Bartek, Jiri (2018): High speed of fork progression induces DNA replication stress 
and genomic instability. In Nature 559 (7713), pp. 279–284. DOI: 10.1038/s41586-018-0261-5. 
McClellan, Amie J.; Laugesen, Sophie Heiden; Ellgaard, Lars (2019): Cellular functions and 
molecular mechanisms of non-lysine ubiquitination. In Open biology 9 (9), p. 190147. DOI: 
10.1098/rsob.190147. 
McCullough, John; Row, Paula E.; Lorenzo, Oscar; Doherty, Mary; Beynon, Robert; Clague, 
Michael J.; Urbé, Sylvie (2006): Activation of the endosome-associated ubiquitin isopeptidase 
AMSH by STAM, a component of the multivesicular body-sorting machinery. In Current biology : 
CB 16 (2), pp. 160–165. DOI: 10.1016/j.cub.2005.11.073. 
Meerang, Mayura; Ritz, Danilo; Paliwal, Shreya; Garajova, Zuzana; Bosshard, Matthias; Mailand, 
Niels et al. (2011): The ubiquitin-selective segregase VCP/p97 orchestrates the response to DNA 
double-strand breaks. In Nature cell biology 13 (11), pp. 1376–1382. DOI: 10.1038/ncb2367. 
Meyer, Hemmo; Bug, Monika; Bremer, Sebastian (2012): Emerging functions of the VCP/p97 
AAA-ATPase in the ubiquitin system. In Nature cell biology 14 (2), pp. 117–123. DOI: 
10.1038/ncb2407. 
Meyer, Hemmo; van den Boom, Johannes (2023): Targeting of client proteins to the 
VCP/p97/Cdc48 unfolding machine. In Frontiers in molecular biosciences 10, p. 1142989. DOI: 
10.3389/fmolb.2023.1142989. 
Meyer, Hemmo; Weihl, Conrad C. (2014): The VCP/p97 system at a glance: connecting cellular 
function to disease pathogenesis. In Journal of cell science 127 (Pt 18), pp. 3877–3883. DOI: 
10.1242/jcs.093831. 
Meyer, Hermann-Josef; Rape, Michael (2014): Enhanced protein degradation by branched 
ubiquitin chains. In Cell 157 (4), pp. 910–921. DOI: 10.1016/j.cell.2014.03.037. 
Meza Gutierrez, Fernando; Simsek, Deniz; Mizrak, Arda; Deutschbauer, Adam; Braberg, Hannes; 
Johnson, Jeffrey et al. (2018): Genetic analysis reveals functions of atypical polyubiquitin chains. 
In eLife 7. DOI: 10.7554/eLife.42955. 
Michel, Martin A.; Komander, David; Elliott, Paul R. (2018): Enzymatic Assembly of Ubiquitin 
Chains. In Methods in molecular biology (Clifton, N.J.) 1844, pp. 73–84. DOI: 10.1007/978-1-
4939-8706-1_6. 
Michel, Martin A.; Swatek, Kirby N.; Hospenthal, Manuela K.; Komander, David (2017): Ubiquitin 
Linkage-Specific Affimers Reveal Insights into K6-Linked Ubiquitin Signaling. In Molecular Cell 68 
(1), 233-246.e5. DOI: 10.1016/j.molcel.2017.08.020. 
Moore, Chandler E.; Yalcindag, Selin E.; Czeladko, Hanna; Ravindranathan, Ramya; Wijesekara 
Hanthi, Yodhara; Levy, Juliana C. et al. (2023): RFWD3 promotes ZRANB3 recruitment to regulate 
the remodeling of stalled replication forks. In The Journal of cell biology 222 (5). DOI: 
10.1083/jcb.202106022. 
185 
 
Moreno, Olga María; Paredes, Angela Camila; Suarez-Obando, Fernando; Rojas, Adriana (2021): 
An update on Fanconi anemia: Clinical, cytogenetic and molecular approaches (Review). In 
Biomedical reports 15 (3), p. 74. DOI: 10.3892/br.2021.1450. 
Morris, Joanna R.; Solomon, Ellen (2004): BRCA1 : BARD1 induces the formation of conjugated 
ubiquitin structures, dependent on K6 of ubiquitin, in cells during DNA replication and repair. In 
Human molecular genetics 13 (8), pp. 807–817. DOI: 10.1093/hmg/ddh095. 
Motegi, Akira; Liaw, Hung-Jiun; Lee, Kyoo-Young; Roest, Henk P.; Maas, Alex; Wu, Xiaoli et al. 
(2008): Polyubiquitination of proliferating cell nuclear antigen by HLTF and SHPRH prevents 
genomic instability from stalled replication forks. In Proceedings of the National Academy of 
Sciences of the United States of America 105 (34), pp. 12411–12416. DOI: 
10.1073/pnas.0805685105. 
Murakumo, Y.; Ogura, Y.; Ishii, H.; Numata, S.; Ichihara, M.; Croce, C. M. et al. (2001): Interactions 
in the error-prone postreplication repair proteins hREV1, hREV3, and hREV7. In The Journal of 
biological chemistry 276 (38), pp. 35644–35651. DOI: 10.1074/jbc.M102051200. 
Mutreja, Karun; Krietsch, Jana; Hess, Jeannine; Ursich, Sebastian; Berti, Matteo; Roessler, 
Fabienne K. et al. (2018): ATR-Mediated Global Fork Slowing and Reversal Assist Fork Traverse 
and Prevent Chromosomal Breakage at DNA Interstrand Cross-Links. In Cell reports 24 (10), 
2629-2642.e5. DOI: 10.1016/j.celrep.2018.08.019. 
Nakao, Fumihiko; Setoguchi, Kiyoko; Semba, Yuichiro; Yamauchi, Takuji; Nogami, Jumpei; 
Sasaki, Kensuke et al. (2023): Targeting a mitochondrial E3 ubiquitin ligase complex to overcome 
AML cell-intrinsic Venetoclax resistance. In Leukemia 37 (5), pp. 1028–1038. DOI: 
10.1038/s41375-023-01879-z. 
Newton, Kim; Matsumoto, Marissa L.; Wertz, Ingrid E.; Kirkpatrick, Donald S.; Lill, Jennie R.; Tan, 
Jenille et al. (2008): Ubiquitin chain editing revealed by polyubiquitin linkage-specific antibodies. 
In Cell 134 (4), pp. 668–678. DOI: 10.1016/j.cell.2008.07.039. 
Nibe, Yoichi; Oshima, Shigeru; Kobayashi, Masanori; Maeyashiki, Chiaki; Matsuzawa, Yu; 
Otsubo, Kana et al. (2018): Novel polyubiquitin imaging system, PolyUb-FC, reveals that K33-
linked polyubiquitin is recruited by SQSTM1/p62. In Autophagy 14 (2), pp. 347–358. DOI: 
10.1080/15548627.2017.1407889. 
Niikura, Yohei; Kitagawa, Risa; Fang, Lei; Kitagawa, Katsumi (2019): CENP-A Ubiquitylation Is 
Indispensable to Cell Viability. In Developmental cell 50 (6), 683-689.e6. DOI: 
10.1016/j.devcel.2019.07.015. 
Nijman, Sebastian M. B.; Huang, Tony T.; Dirac, Annette M. G.; Brummelkamp, Thijn R.; 
Kerkhoven, Ron M.; D'Andrea, Alan D.; Bernards, René (2005): The deubiquitinating enzyme 
USP1 regulates the Fanconi anemia pathway. In Molecular Cell 17 (3), pp. 331–339. DOI: 
10.1016/j.molcel.2005.01.008. 
Nowotny, Marcin (2019): Crosslink and shield: protecting abasic sites from error-prone repair. In 
Nature structural & molecular biology 26 (7), pp. 530–532. DOI: 10.1038/s41594-019-0264-4. 
Ohashi, Eiji; Murakumo, Yoshiki; Kanjo, Naoko; Akagi, Jun-Ichi; Masutani, Chikahide; Hanaoka, 
Fumio; Ohmori, Haruo (2004): Interaction of hREV1 with three human Y-family DNA polymerases. 
In Genes to cells : devoted to molecular & cellular mechanisms 9 (6), pp. 523–531. DOI: 
10.1111/j.1356-9597.2004.00747.x. 
186 
 
Ohtake, Fumiaki; Saeki, Yasushi; Ishido, Satoshi; Kanno, Jun; Tanaka, Keiji (2016): The K48-K63 
Branched Ubiquitin Chain Regulates NF-κB Signaling. In Molecular Cell 64 (2), pp. 251–266. DOI: 
10.1016/j.molcel.2016.09.014. 
Ohtake, Fumiaki; Tsuchiya, Hikaru; Saeki, Yasushi; Tanaka, Keiji (2018): K63 ubiquitylation 
triggers proteasomal degradation by seeding branched ubiquitin chains. In Proceedings of the 
National Academy of Sciences of the United States of America 115 (7), E1401-E1408. DOI: 
10.1073/pnas.1716673115. 
Okatsu, Kei; Koyano, Fumika; Kimura, Mayumi; Kosako, Hidetaka; Saeki, Yasushi; Tanaka, Keiji; 
Matsuda, Noriyuki (2015): Phosphorylated ubiquitin chain is the genuine Parkin receptor. In The 
Journal of cell biology 209 (1), pp. 111–128. DOI: 10.1083/jcb.201410050. 
Oltion, Keely; Carelli, Jordan D.; Yang, Tangpo; See, Stephanie K.; Wang, Hao-Yuan; Kampmann, 
Martin; Taunton, Jack (2023): An E3 ligase network engages GCN1 to promote the degradation 
of translation factors on stalled ribosomes. In Cell 186 (2), 346-362.e17. DOI: 
10.1016/j.cell.2022.12.025. 
Ong, Shao-En; Mann, Matthias (2006): A practical recipe for stable isotope labeling by amino 
acids in cell culture (SILAC). In Nature protocols 1 (6), pp. 2650–2660. DOI: 
10.1038/nprot.2006.427. 
Ordureau, Alban; Heo, Jin-Mi; Duda, David M.; Paulo, Joao A.; Olszewski, Jennifer L.; 
Yanishevski, David et al. (2015): Defining roles of PARKIN and ubiquitin phosphorylation by PINK1 
in mitochondrial quality control using a ubiquitin replacement strategy. In Proceedings of the 
National Academy of Sciences of the United States of America 112 (21), pp. 6637–6642. DOI: 
10.1073/pnas.1506593112. 
Orlando, V. (2000): Mapping chromosomal proteins in vivo by formaldehyde-crosslinked-
chromatin immunoprecipitation. In Trends in biochemical sciences 25 (3), pp. 99–104. DOI: 
10.1016/s0968-0004(99)01535-2. 
Panzarino, Nicholas J.; Krais, John J.; Cong, Ke; Peng, Min; Mosqueda, Michelle; Nayak, Sumeet 
U. et al. (2021): Replication Gaps Underlie BRCA Deficiency and Therapy Response. In Cancer 
research 81 (5), pp. 1388–1397. DOI: 10.1158/0008-5472.CAN-20-1602. 
Park, Su Hyung; Kang, Nalae; Song, Eunho; Wie, Minwoo; Lee, Eun A.; Hwang, Sunyoung et al. 
(2019): ATAD5 promotes replication restart by regulating RAD51 and PCNA in response to 
replication stress. In Nature Communications 10 (1), p. 5718. DOI: 10.1038/s41467-019-13667-4. 
Park, Yoon; Yoon, Sungjoo Kim; Yoon, Jong-Bok (2009): The HECT domain of TRIP12 
ubiquitinates substrates of the ubiquitin fusion degradation pathway. In The Journal of biological 
chemistry 284 (3), pp. 1540–1549. DOI: 10.1074/jbc.M807554200. 
Parker, Matthew W.; Botchan, Michael R.; Berger, James M. (2017): Mechanisms and regulation 
of DNA replication initiation in eukaryotes. In Critical reviews in biochemistry and molecular biology 
52 (2), pp. 107–144. DOI: 10.1080/10409238.2016.1274717. 
Peng, Junmin; Schwartz, Daniel; Elias, Joshua E.; Thoreen, Carson C.; Cheng, Dongmei; 
Marsischky, Gerald et al. (2003): A proteomics approach to understanding protein ubiquitination. 
In Nature biotechnology 21 (8), pp. 921–926. DOI: 10.1038/nbt849. 
Petermann, Eva; Orta, Manuel Luís; Issaeva, Natalia; Schultz, Niklas; Helleday, Thomas (2010): 
Hydroxyurea-stalled replication forks become progressively inactivated and require two different 
187 
 
RAD51-mediated pathways for restart and repair. In Molecular Cell 37 (4), pp. 492–502. DOI: 
10.1016/j.molcel.2010.01.021. 
Piberger, Ann Liza; Bowry, Akhil; Kelly, Richard D. W.; Walker, Alexandra K.; González-Acosta, 
Daniel; Bailey, Laura J. et al. (2020): PrimPol-dependent single-stranded gap formation mediates 
homologous recombination at bulky DNA adducts. In Nature Communications 11 (1), p. 5863. 
DOI: 10.1038/s41467-020-19570-7. 
Pickart, C. M. (1997): Targeting of substrates to the 26S proteasome. In FASEB journal : official 
publication of the Federation of American Societies for Experimental Biology 11 (13), pp. 1055–
1066. DOI: 10.1096/fasebj.11.13.9367341. 
Pluska, Lukas; Jarosch, Ernst; Zauber, Henrik; Kniss, Andreas; Waltho, Anita; Bagola, Katrin et 
al. (2021): The UBA domain of conjugating enzyme Ubc1/Ube2K facilitates assembly of K48/K63-
branched ubiquitin chains. In The EMBO journal 40 (6), e106094. DOI: 
10.15252/embj.2020106094. 
Porebski, Bartlomiej; Wild, Sebastian; Kummer, Sandra; Scaglione, Sarah; Gaillard, Pierre-Henri 
L.; Gari, Kerstin (2019): WRNIP1 Protects Reversed DNA Replication Forks from SLX4-
Dependent Nucleolytic Cleavage. In iScience 21, pp. 31–41. DOI: 10.1016/j.isci.2019.10.010. 
Postow, Lisa; Funabiki, Hironori (2013): An SCF complex containing Fbxl12 mediates DNA 
damage-induced Ku80 ubiquitylation. In Cell cycle (Georgetown, Tex.) 12 (4), pp. 587–595. DOI: 
10.4161/cc.23408. 
Powers, Kyle T.; Lavering, Emily D.; Washington, M. Todd (2018): Conformational Flexibility of 
Ubiquitin-Modified and SUMO-Modified PCNA Shown by Full-Ensemble Hybrid Methods. In 
Journal of molecular biology 430 (24), pp. 5294–5303. DOI: 10.1016/j.jmb.2018.10.017. 
Pratt, Matthew R.; Schwartz, Edmund C.; Muir, Tom W. (2007): Small-molecule-mediated rescue 
of protein function by an inducible proteolytic shunt. In Proceedings of the National Academy of 
Sciences of the United States of America 104 (27), pp. 11209–11214. DOI: 
10.1073/pnas.0700816104. 
Przetocka, Sara; Porro, Antonio; Bolck, Hella A.; Walker, Christina; Lezaja, Aleksandra; Trenner, 
Anika et al. (2018): CtIP-Mediated Fork Protection Synergizes with BRCA1 to Suppress Genomic 
Instability upon DNA Replication Stress. In Molecular Cell 72 (3), 568-582.e6. DOI: 
10.1016/j.molcel.2018.09.014. 
Qian, Shu-Bing; Ott, David E.; Schubert, Ulrich; Bennink, Jack R.; Yewdell, Jonathan W. (2002): 
Fusion proteins with COOH-terminal ubiquitin are stable and maintain dual functionality in vivo. In 
The Journal of biological chemistry 277 (41), pp. 38818–38826. DOI: 10.1074/jbc.M205547200. 
Qin, Weihua; Steinek, Clemens; Kolobynina, Ksenia; Forné, Ignasi; Imhof, Axel; Cardoso, M. 
Cristina; Leonhardt, Heinrich (2022): Probing protein ubiquitination in live cells. In Nucleic acids 
research 50 (21), e125. DOI: 10.1093/nar/gkac805. 
Qin, Zhoushuai; Bai, Zhiqiang; Sun, Ying; Niu, Xiaohong; Xiao, Wei (2016): PCNA-Ub 
polyubiquitination inhibits cell proliferation and induces cell-cycle checkpoints. In Cell cycle 
(Georgetown, Tex.) 15 (24), pp. 3390–3401. DOI: 10.1080/15384101.2016.1245247. 
Quinet, Annabel; Carvajal-Maldonado, Denisse; Lemacon, Delphine; Vindigni, Alessandro (2017): 
DNA Fiber Analysis: Mind the Gap! In Methods in enzymology 591, pp. 55–82. DOI: 
10.1016/bs.mie.2017.03.019. 
188 
 
Quinet, Annabel; Tirman, Stephanie; Cybulla, Emily; Meroni, Alice; Vindigni, Alessandro (2021): 
To skip or not to skip: choosing repriming to tolerate DNA damage. In Molecular Cell 81 (4), 
pp. 649–658. DOI: 10.1016/j.molcel.2021.01.012. 
Quinet, Annabel; Tirman, Stephanie; Jackson, Jessica; Šviković, Saša; Lemaçon, Delphine; 
Carvajal-Maldonado, Denisse et al. (2020): PRIMPOL-Mediated Adaptive Response Suppresses 
Replication Fork Reversal in BRCA-Deficient Cells. In Molecular Cell 77 (3), 461-474.e9. DOI: 
10.1016/j.molcel.2019.10.008. 
Raducanu, Vlad-Stefan; Tehseen, Muhammad; Al-Amodi, Amani; Joudeh, Luay I.; Biasio, Alfredo 
de; Hamdan, Samir M. (2022): Mechanistic investigation of human maturation of Okazaki 
fragments reveals slow kinetics. In Nature Communications 13 (1), p. 6973. DOI: 10.1038/s41467-
022-34751-2. 
Ravid, Tommer; Hochstrasser, Mark (2007): Autoregulation of an E2 enzyme by ubiquitin-chain 
assembly on its catalytic residue. In Nature cell biology 9 (4), pp. 422–427. DOI: 10.1038/ncb1558. 
Ren, Xingjie; Yang, Zhihao; Mao, Decai; Chang, Zai; Qiao, Huan-Huan; Wang, Xia et al. (2014): 
Performance of the Cas9 nickase system in Drosophila melanogaster. In G3 (Bethesda, Md.) 4 
(10), pp. 1955–1962. DOI: 10.1534/g3.114.013821. 
Renz, Christian; Albanèse, Véronique; Tröster, Vera; Albert, Thomas K.; Santt, Olivier; Jacobs, 
Susan C. et al. (2020): Ubc13-Mms2 cooperates with a family of RING E3 proteins in budding 
yeast membrane protein sorting. In Journal of cell science 133 (10). DOI: 10.1242/jcs.244566. 
Richard, Thibaud J. C.; Herzog, Laura K.; Vornberger, Julia; Rahmanto, Aldwin Suryo; Sangfelt, 
Olle; Salomons, Florian A.; Dantuma, Nico P. (2020): K63-linked ubiquitylation induces global 
sequestration of mitochondria. In Scientific reports 10 (1), p. 22334. DOI: 10.1038/s41598-020-
78845-7. 
Ryan, M. D.; King, A. M.; Thomas, G. P. (1991): Cleavage of foot-and-mouth disease virus 
polyprotein is mediated by residues located within a 19 amino acid sequence. In The Journal of 
general virology 72 (Pt 11), pp. 2727–2732. DOI: 10.1099/0022-1317-72-11-2727. 
Sakano, K.; Oikawa, S.; Hasegawa, K.; Kawanishi, S. (2001): Hydroxyurea induces site-specific 
DNA damage via formation of hydrogen peroxide and nitric oxide. In Japanese journal of cancer 
research : Gann 92 (11), pp. 1166–1174. DOI: 10.1111/j.1349-7006.2001.tb02136.x. 
Salas-Lloret, Daniel; García-Rodríguez, Néstor; Giebel, Lisanne; Ru, Arnoud de; van Veelen, 
Peter; Huertas, Pablo et al. (2023): BRCA1/BARD1 ubiquitinates PCNA in unperturbed conditions 
to promote replication fork stability and continuous DNA synthesis. DOI: 
10.1101/2023.01.12.523782. 
Salvador, J-Pablo; Vilaplana, Lluïsa; Marco, M-Pilar (2019): Nanobody: outstanding features for 
diagnostic and therapeutic applications. In Analytical and bioanalytical chemistry 411 (9), 
pp. 1703–1713. DOI: 10.1007/s00216-019-01633-4. 
Sato, Yusuke; Fujita, Hiroaki; Yoshikawa, Azusa; Yamashita, Masami; Yamagata, Atsushi; Kaiser, 
Stephen E. et al. (2011): Specific recognition of linear ubiquitin chains by the Npl4 zinc finger 
(NZF) domain of the HOIL-1L subunit of the linear ubiquitin chain assembly complex. In 
Proceedings of the National Academy of Sciences of the United States of America 108 (51), 
pp. 20520–20525. DOI: 10.1073/pnas.1109088108. 
189 
 
Saugar, Irene; Parker, Joanne L.; Zhao, Shengkai; Ulrich, Helle D. (2012): The genome 
maintenance factor Mgs1 is targeted to sites of replication stress by ubiquitylated PCNA. In 
Nucleic acids research 40 (1), pp. 245–257. DOI: 10.1093/nar/gkr738. 
Schärer, Orlando D. (2013): Nucleotide excision repair in eukaryotes. In Cold Spring Harbor 
perspectives in biology 5 (10), a012609. DOI: 10.1101/cshperspect.a012609. 
Schrecker, Marina; Castaneda, Juan C.; Devbhandari, Sujan; Kumar, Charanya; Remus, Dirk; 
Hite, Richard K. (2022): Multistep loading of a DNA sliding clamp onto DNA by replication factor 
C. In eLife 11. DOI: 10.7554/eLife.78253. 
Schwertman, Petra; Bekker-Jensen, Simon; Mailand, Niels (2016): Regulation of DNA double-
strand break repair by ubiquitin and ubiquitin-like modifiers. In Nature reviews. Molecular cell 
biology 17 (6), pp. 379–394. DOI: 10.1038/nrm.2016.58. 
Sedgwick, Barbara; Lindahl, Tomas (2002): Recent progress on the Ada response for inducible 
repair of DNA alkylation damage. In Oncogene 21 (58), pp. 8886–8894. DOI: 
10.1038/sj.onc.1205998. 
Seelinger, Mareike; Otterlei, Marit (2020): Helicase-Like Transcription Factor HLTF and E3 
Ubiquitin Ligase SHPRH Confer DNA Damage Tolerance through Direct Interactions with 
Proliferating Cell Nuclear Antigen (PCNA). In International journal of molecular sciences 21 (3). 
DOI: 10.3390/ijms21030693. 
Shachar, Sigal; Ziv, Omer; Avkin, Sharon; Adar, Sheera; Wittschieben, John; Reissner, Thomas 
et al. (2009): Two-polymerase mechanisms dictate error-free and error-prone translesion DNA 
synthesis in mammals. In The EMBO journal 28 (4), pp. 383–393. DOI: 10.1038/emboj.2008.281. 
Sharma, Sudha (2011): Non-B DNA Secondary Structures and Their Resolution by RecQ 
Helicases. In Journal of nucleic acids 2011, p. 724215. DOI: 10.4061/2011/724215. 
Shi, Jie; Hauschulte, Kristine; Mikicic, Ivan; Maharjan, Srijana; Arz, Valerie; Strauch, Tina et al. 
(2023): Nuclear myosin VI maintains replication fork stability. In Nature Communications 14 (1), 
p. 3787. DOI: 10.1038/s41467-023-39517-y. 
Shin, Ji Yeong; Muniyappan, Srinivasan; Tran, Non-Nuoc; Park, Hyeonjeong; Lee, Sung Bae; Lee, 
Byung-Hoon (2020): Deubiquitination Reactions on the Proteasome for Proteasome Versatility. In 
International journal of molecular sciences 21 (15). DOI: 10.3390/ijms21155312. 
Shin, M. E.; Ogburn, K. D.; Varban, O. A.; Gilbert, P. M.; Burd, C. G. (2001): FYVE domain targets 
Pib1p ubiquitin ligase to endosome and vacuolar membranes. In The Journal of biological 
chemistry 276 (44), pp. 41388–41393. DOI: 10.1074/jbc.M105665200. 
Simoneau, Antoine; Engel, Justin L.; Bandi, Madhavi; Lazarides, Katherine; Liu, Shangtao; Meier, 
Samuel R. et al. (2023): Ubiquitinated PCNA Drives USP1 Synthetic Lethality in Cancer. In 
Molecular cancer therapeutics 22 (2), pp. 215–226. DOI: 10.1158/1535-7163.MCT-22-0409. 
Sims, Joshua J.; Cohen, Robert E. (2009): Linkage-specific avidity defines the lysine 63-linked 
polyubiquitin-binding preference of rap80. In Molecular Cell 33 (6), pp. 775–783. DOI: 
10.1016/j.molcel.2009.02.011. 
Sims, Joshua J.; Scavone, Francesco; Cooper, Eric M.; Kane, Lesley A.; Youle, Richard J.; Boeke, 
Jef D.; Cohen, Robert E. (2012): Polyubiquitin-sensor proteins reveal localization and linkage-type 
dependence of cellular ubiquitin signaling. In Nature methods 9 (3), pp. 303–309. DOI: 
10.1038/nmeth.1888. 
190 
 
Singer, B.; Hang, B. (1997): What structural features determine repair enzyme specificity and 
mechanism in chemically modified DNA? In Chemical research in toxicology 10 (7), pp. 713–732. 
DOI: 10.1021/tx970011e. 
Smit, Judith J.; Monteferrario, Davide; Noordermeer, Sylvie M.; van Dijk, Willem J.; van der 
Reijden, Bert A; Sixma, Titia K. (2012): The E3 ligase HOIP specifies linear ubiquitin chain 
assembly through its RING-IBR-RING domain and the unique LDD extension. In The EMBO 
journal 31 (19), pp. 3833–3844. DOI: 10.1038/emboj.2012.217. 
Smit, Judith J.; van Dijk, Willem J.; El Atmioui, Dris; Merkx, Remco; Ovaa, Huib; Sixma, Titia K. 
(2013): Target specificity of the E3 ligase LUBAC for ubiquitin and NEMO relies on different 
minimal requirements. In The Journal of biological chemistry 288 (44), pp. 31728–31737. DOI: 
10.1074/jbc.M113.495846. 
Socha, Anna; Di Yang; Bulsiewicz, Alicja; Yaprianto, Kelvin; Kupculak, Marian; Liang, Chih-Chao 
et al. (2020): WRNIP1 Is Recruited to DNA Interstrand Crosslinks and Promotes Repair. In Cell 
reports 32 (1), p. 107850. DOI: 10.1016/j.celrep.2020.107850. 
Sogo, José M.; Lopes, Massimo; Foiani, Marco (2002): Fork reversal and ssDNA accumulation at 
stalled replication forks owing to checkpoint defects. In Science (New York, N.Y.) 297 (5581), 
pp. 599–602. DOI: 10.1126/science.1074023. 
Somyajit, Kumar; Spies, Julian; Coscia, Fabian; Kirik, Ufuk; Rask, Maj-Britt; Lee, Ji-Hoon et al. 
(2021): Homology-directed repair protects the replicating genome from metabolic assaults. In 
Developmental cell 56 (4), 461-477.e7. DOI: 10.1016/j.devcel.2021.01.011. 
Soutoglou, Evi; Misteli, Tom (2008): Activation of the cellular DNA damage response in the 
absence of DNA lesions. In Science (New York, N.Y.) 320 (5882), pp. 1507–1510. DOI: 
10.1126/science.1159051. 
Srinivasan, Seetha V.; Dominguez-Sola, David; Wang, Lily C.; Hyrien, Olivier; Gautier, Jean 
(2013): Cdc45 is a critical effector of myc-dependent DNA replication stress. In Cell reports 3 (5), 
pp. 1629–1639. DOI: 10.1016/j.celrep.2013.04.002. 
Stelter, Philipp; Ulrich, Helle D. (2003): Control of spontaneous and damage-induced mutagenesis 
by SUMO and ubiquitin conjugation. In Nature 425 (6954), pp. 188–191. DOI: 
10.1038/nature01965. 
Stewart, Mikaela D.; Ritterhoff, Tobias; Klevit, Rachel E.; Brzovic, Peter S. (2016): E2 enzymes: 
more than just middle men. In Cell research 26 (4), pp. 423–440. DOI: 10.1038/cr.2016.35. 
Stingele, Julian; Bellelli, Roberto; Boulton, Simon J. (2017): Mechanisms of DNA-protein crosslink 
repair. In Nature reviews. Molecular cell biology 18 (9), pp. 563–573. DOI: 10.1038/nrm.2017.56. 
Taglialatela, Angelo; Alvarez, Silvia; Leuzzi, Giuseppe; Sannino, Vincenzo; Ranjha, Lepakshi; 
Huang, Jen-Wei et al. (2017): Restoration of Replication Fork Stability in BRCA1- and BRCA2-
Deficient Cells by Inactivation of SNF2-Family Fork Remodelers. In Molecular Cell 68 (2), 414-
430.e8. DOI: 10.1016/j.molcel.2017.09.036. 
Takahashi, Tomio S.; Wollscheid, Hans-Peter; Lowther, Jonathan; Ulrich, Helle D. (2020): Effects 
of chain length and geometry on the activation of DNA damage bypass by polyubiquitylated PCNA. 
In Nucleic acids research 48 (6), pp. 3042–3052. DOI: 10.1093/nar/gkaa053. 
191 
 
Tanaka, Keiji (2009): The proteasome: overview of structure and functions. In Proceedings of the 
Japan Academy. Series B, Physical and biological sciences 85 (1), pp. 12–36. DOI: 
10.2183/pjab.85.12. 
Temviriyanukul, Piya; van Hees-Stuivenberg, Sandrine; Delbos, Frédéric; Jacobs, Heinz; Wind, 
Niels de; Jansen, Jacob G. (2012): Temporally distinct translesion synthesis pathways for 
ultraviolet light-induced photoproducts in the mammalian genome. In DNA repair 11 (6), pp. 550–
558. DOI: 10.1016/j.dnarep.2012.03.007. 
Thakar, Tanay; Leung, Wendy; Nicolae, Claudia M.; Clements, Kristen E.; Shen, Binghui; 
Bielinsky, Anja-Katrin; Moldovan, George-Lucian (2020): Ubiquitinated-PCNA protects replication 
forks from DNA2-mediated degradation by regulating Okazaki fragment maturation and chromatin 
assembly. In Nature Communications 11 (1), p. 2147. DOI: 10.1038/s41467-020-16096-w. 
Thrower, J. S.; Hoffman, L.; Rechsteiner, M.; Pickart, C. M. (2000): Recognition of the polyubiquitin 
proteolytic signal. In The EMBO journal 19 (1), pp. 94–102. DOI: 10.1093/emboj/19.1.94. 
Tian, Tian; Bu, Min; Chen, Xu; Ding, Linli; Yang, Yulan; Han, Jinhua et al. (2021): The ZATT-
TOP2A-PICH Axis Drives Extensive Replication Fork Reversal to Promote Genome Stability. In 
Molecular Cell 81 (1), 198-211.e6. DOI: 10.1016/j.molcel.2020.11.007. 
Tiede, Christian; Bedford, Robert; Heseltine, Sophie J.; Smith, Gina; Wijetunga, Imeshi; Ross, 
Rebecca et al. (2017): Affimer proteins are versatile and renewable affinity reagents. In eLife 6. 
DOI: 10.7554/eLife.24903. 
Tirman, Stephanie; Quinet, Annabel; Wood, Matthew; Meroni, Alice; Cybulla, Emily; Jackson, 
Jessica et al. (2021): Temporally distinct post-replicative repair mechanisms fill PRIMPOL-
dependent ssDNA gaps in human cells. In Molecular Cell 81 (19), 4026-4040.e8. DOI: 
10.1016/j.molcel.2021.09.013. 
Tokunaga, Fuminori; Iwai, Kazuhiro (2012): LUBAC, a novel ubiquitin ligase for linear 
ubiquitination, is crucial for inflammation and immune responses. In Microbes and infection 14 (7-
8), pp. 563–572. DOI: 10.1016/j.micinf.2012.01.011. 
Tokunaga, Fuminori; Nishimasu, Hiroshi; Ishitani, Ryuichiro; Goto, Eiji; Noguchi, Takuya; Mio, 
Kazuhiro et al. (2012): Specific recognition of linear polyubiquitin by A20 zinc finger 7 is involved 
in NF-κB regulation. In The EMBO journal 31 (19), pp. 3856–3870. DOI: 10.1038/emboj.2012.241. 
Toledo, Luis Ignacio; Altmeyer, Matthias; Rask, Maj-Britt; Lukas, Claudia; Larsen, Dorthe Helena; 
Povlsen, Lou Klitgaard et al. (2013): ATR prohibits replication catastrophe by preventing global 
exhaustion of RPA. In Cell 155 (5), pp. 1088–1103. DOI: 10.1016/j.cell.2013.10.043. 
Tonzi, Peter; Yin, Yandong; Lee, Chelsea Wei Ting; Rothenberg, Eli; Huang, Tony T. (2018): 
Translesion polymerase kappa-dependent DNA synthesis underlies replication fork recovery. In 
eLife 7. DOI: 10.7554/eLife.41426. 
Tornaletti, S.; Pfeifer, G. P. (1996): UV damage and repair mechanisms in mammalian cells. In 
BioEssays : news and reviews in molecular, cellular and developmental biology 18 (3), pp. 221–
228. DOI: 10.1002/bies.950180309. 
Tubbs, Anthony; Nussenzweig, André (2017): Endogenous DNA Damage as a Source of Genomic 
Instability in Cancer. In Cell 168 (4), pp. 644–656. DOI: 10.1016/j.cell.2017.01.002. 
192 
 
Turner, M. K.; Abrams, R.; Lieberman, I. (1966): Meso-alpha, beta-diphenylsuccinate and 
hydroxyurea as inhibitors of deoxycytidylate synthesis in extracts of Ehrlich ascites and L cells. In 
The Journal of biological chemistry 241 (24), pp. 5777–5780. 
Ulrich, Helle D.; Takahashi, Diane T. (2013): Readers of PCNA modifications. In Chromosoma 
122 (4), pp. 259–274. DOI: 10.1007/s00412-013-0410-4. 
Unk, Ildiko; Hajdú, Ildikó; Fátyol, Károly; Szakál, Barnabás; Blastyák, András; Bermudez, Vladimir 
et al. (2006): Human SHPRH is a ubiquitin ligase for Mms2-Ubc13-dependent polyubiquitylation 
of proliferating cell nuclear antigen. In Proceedings of the National Academy of Sciences of the 
United States of America 103 (48), pp. 18107–18112. DOI: 10.1073/pnas.0608595103. 
Vaz, Bruno; Halder, Swagata; Ramadan, Kristijan (2013): Role of p97/VCP (Cdc48) in genome 
stability. In Frontiers in genetics 4, p. 60. DOI: 10.3389/fgene.2013.00060. 
Vijay-Kumar, S.; Bugg, C. E.; Cook, W. J. (1987): Structure of ubiquitin refined at 1.8 A resolution. 
In Journal of molecular biology 194 (3), pp. 531–544. DOI: 10.1016/0022-2836(87)90679-6. 
Villamil, Mark; Xiao, Weidi; Yu, Clinton; Huang, Lan; Xu, Ping; Kaiser, Peter (2022): The Ubiquitin 
Interacting Motif-Like Domain of Met4 Selectively Binds K48 Polyubiquitin Chains. In Molecular & 
cellular proteomics : MCP 21 (1), p. 100175. DOI: 10.1016/j.mcpro.2021.100175. 
Vina-Vilaseca, Arnau; Sorkin, Alexander (2010): Lysine 63-linked polyubiquitination of the 
dopamine transporter requires WW3 and WW4 domains of Nedd4-2 and UBE2D ubiquitin-
conjugating enzymes. In The Journal of biological chemistry 285 (10), pp. 7645–7656. DOI: 
10.1074/jbc.M109.058990. 
Vujanovic, Marko; Krietsch, Jana; Raso, Maria Chiara; Terraneo, Nastassja; Zellweger, Ralph; 
Schmid, Jonas A. et al. (2017): Replication Fork Slowing and Reversal upon DNA Damage 
Require PCNA Polyubiquitination and ZRANB3 DNA Translocase Activity. In Molecular Cell 67 
(5), 882-890.e5. DOI: 10.1016/j.molcel.2017.08.010. 
Wagner, Sebastian A.; Beli, Petra; Weinert, Brian T.; Nielsen, Michael L.; Cox, Jürgen; Mann, 
Matthias; Choudhary, Chunaram (2011): A proteome-wide, quantitative survey of in vivo 
ubiquitylation sites reveals widespread regulatory roles. In Molecular & cellular proteomics : MCP 
10 (10), M111.013284. DOI: 10.1074/mcp.M111.013284. 
Wajant, Harald; Scheurich, Peter (2011): TNFR1-induced activation of the classical NF-κB 
pathway. In The FEBS journal 278 (6), pp. 862–876. DOI: 10.1111/j.1742-4658.2011.08015.x. 
Wang, Qiang; Liu, Xing; Cui, Ye; Tang, Yijun; Chen, Wei; Li, Senlin et al. (2014): The E3 ubiquitin 
ligase AMFR and INSIG1 bridge the activation of TBK1 kinase by modifying the adaptor STING. 
In Immunity 41 (6), pp. 919–933. DOI: 10.1016/j.immuni.2014.11.011. 
Wang, Qing; Song, Changcheng; Li, Chou-Chi H. (2003): Hexamerization of p97-VCP is promoted 
by ATP binding to the D1 domain and required for ATPase and biological activities. In Biochemical 
and biophysical research communications 300 (2), pp. 253–260. DOI: 10.1016/s0006-
291x(02)02840-1. 
Wang, Qinghua; Young, Patrick; Walters, Kylie J. (2005): Structure of S5a bound to monoubiquitin 
provides a model for polyubiquitin recognition. In Journal of molecular biology 348 (3), pp. 727–
739. DOI: 10.1016/j.jmb.2005.03.007. 
Wang, Tao; Yin, Luming; Cooper, Eric M.; Lai, Ming-Yih; Dickey, Seth; Pickart, Cecile M. et al. 
(2009): Evidence for bidentate substrate binding as the basis for the K48 linkage specificity of 
193 
 
otubain 1. In Journal of molecular biology 386 (4), pp. 1011–1023. DOI: 
10.1016/j.jmb.2008.12.085. 
Wegmann, Sabrina; Meister, Cindy; Renz, Christian; Yakoub, George; Wollscheid, Hans-Peter; 
Takahashi, Diane T. et al. (2022): Linkage reprogramming by tailor-made E3s reveals 
polyubiquitin chain requirements in DNA-damage bypass. In Molecular Cell 82 (8), 1589-1602.e5. 
DOI: 10.1016/j.molcel.2022.02.016. 
Wertz, Ingrid E.; O'Rourke, Karen M.; Zhou, Honglin; Eby, Michael; Aravind, L.; Seshagiri, 
Somasekar et al. (2004): De-ubiquitination and ubiquitin ligase domains of A20 downregulate NF-
kappaB signalling. In Nature 430 (7000), pp. 694–699. DOI: 10.1038/nature02794. 
Wolf, Lucie M.; Lambert, Annika M.; Haenlin, Julie; Boutros, Michael (2021): EVI/WLS function is 
regulated by ubiquitylation and is linked to ER-associated degradation by ERLIN2. In Journal of 
cell science 134 (16). DOI: 10.1242/jcs.257790. 
Wong, Ronald P.; García-Rodríguez, Néstor; Zilio, Nicola; Hanulová, Mária; Ulrich, Helle D. 
(2020): Processing of DNA Polymerase-Blocking Lesions during Genome Replication Is Spatially 
and Temporally Segregated from Replication Forks. In Molecular Cell 77 (1), 3-16.e4. DOI: 
10.1016/j.molcel.2019.09.015. 
Wright, Joshua D.; Mace, Peter D.; Day, Catherine L. (2016): Noncovalent Ubiquitin Interactions 
Regulate the Catalytic Activity of Ubiquitin Writers. In Trends in biochemical sciences 41 (11), 
pp. 924–937. DOI: 10.1016/j.tibs.2016.08.003. 
Wright, William Douglass; Shah, Shanaya Shital; Heyer, Wolf-Dietrich (2018): Homologous 
recombination and the repair of DNA double-strand breaks. In The Journal of biological chemistry 
293 (27), pp. 10524–10535. DOI: 10.1074/jbc.TM118.000372. 
Wu, Xiao; Liu, Shichang; Sagum, Cari; Chen, Jianji; Singh, Rajesh; Chaturvedi, Apurva et al. 
(2019): Crosstalk between Lys63- and Lys11-polyubiquitin signaling at DNA damage sites is 
driven by Cezanne. In Genes & development 33 (23-24), pp. 1702–1717. DOI: 
10.1101/gad.332395.119. 
Xu, Ming; Skaug, Brian; Zeng, Wenwen; Chen, Zhijian J. (2009): A ubiquitin replacement strategy 
in human cells reveals distinct mechanisms of IKK activation by TNFalpha and IL-1beta. In 
Molecular Cell 36 (2), pp. 302–314. DOI: 10.1016/j.molcel.2009.10.002. 
Yamano, Koji; Kikuchi, Reika; Kojima, Waka; Hayashida, Ryota; Koyano, Fumika; Kawawaki, 
Junko et al. (2020): Critical role of mitochondrial ubiquitination and the OPTN-ATG9A axis in 
mitophagy. In The Journal of cell biology 219 (9). DOI: 10.1083/jcb.201912144. 
Yang, Kailin; Moldovan, George-Lucian; D'Andrea, Alan D. (2010): RAD18-dependent recruitment 
of SNM1A to DNA repair complexes by a ubiquitin-binding zinc finger. In The Journal of biological 
chemistry 285 (25), pp. 19085–19091. DOI: 10.1074/jbc.M109.100032. 
Yau, Richard G.; Doerner, Kerstin; Castellanos, Erick R.; Haakonsen, Diane L.; Werner, Achim; 
Wang, Nan et al. (2017): Assembly and Function of Heterotypic Ubiquitin Chains in Cell-Cycle and 
Protein Quality Control. In Cell 171 (4), 918-933.e20. DOI: 10.1016/j.cell.2017.09.040. 
Yewdell, J. W. (2001): Not such a dismal science: the economics of protein synthesis, folding, 
degradation and antigen processing. In Trends in cell biology 11 (7), pp. 294–297. DOI: 
10.1016/s0962-8924(01)02030-x. 
194 
 
Yoon, Jung-Hoon; Acharya, Narottam; Park, Jeseong; Basu, Debashree; Prakash, Satya; 
Prakash, Louise (2014): Identification of two functional PCNA-binding domains in human DNA 
polymerase κ. In Genes to cells : devoted to molecular & cellular mechanisms 19 (7), pp. 594–
601. DOI: 10.1111/gtc.12156. 
Yoon, Jung-Hoon; Bhatia, Gita; Prakash, Satya; Prakash, Louise (2010): Error-free replicative 
bypass of thymine glycol by the combined action of DNA polymerases kappa and zeta in human 
cells. In Proceedings of the National Academy of Sciences of the United States of America 107 
(32), pp. 14116–14121. DOI: 10.1073/pnas.1007795107. 
You-Wei Zhang; Diane M. Otterness; Gary G. Chiang; Weilin Xie; Yun-Cai Liu; Frank Mercurio; 
Robert T. Abraham (2005): Genotoxic Stress Targets Human Chk1 for Degradation by the 
Ubiquitin-Proteasome Pathway. In Molecular Cell 19 (5), pp. 607–618. DOI: 
10.1016/j.molcel.2005.07.019. 
Yu, Guimei; Bai, Yunpeng; Li, Kunpeng; Amarasinghe, Ovini; Jiang, Wen; Zhang, Zhong-Yin 
(2021): Cryo-electron microscopy structures of VCP/p97 reveal a new mechanism of 
oligomerization regulation. In iScience 24 (11), p. 103310. DOI: 10.1016/j.isci.2021.103310. 
Yuan, Wei-Chien; Lee, Yu-Ru; Lin, Shu-Yu; Chang, Li-Ying; Tan, Yen Pei; Hung, Chin-Chun et al. 
(2014): K33-Linked Polyubiquitination of Coronin 7 by Cul3-KLHL20 Ubiquitin E3 Ligase 
Regulates Protein Trafficking. In Molecular Cell 54 (4), pp. 586–600. DOI: 
10.1016/j.molcel.2014.03.035. 
Yun, Maximina H.; Hiom, Kevin (2009): CtIP-BRCA1 modulates the choice of DNA double-strand-
break repair pathway throughout the cell cycle. In Nature 459 (7245), pp. 460–463. DOI: 
10.1038/nature07955. 
Zellweger, Ralph; Dalcher, Damian; Mutreja, Karun; Berti, Matteo; Schmid, Jonas A.; Herrador, 
Raquel et al. (2015): Rad51-mediated replication fork reversal is a global response to genotoxic 
treatments in human cells. In The Journal of cell biology 208 (5), pp. 563–579. DOI: 
10.1083/jcb.201406099. 
Zeman, Michelle K.; Cimprich, Karlene A. (2014): Causes and consequences of replication stress. 
In Nature cell biology 16 (1), pp. 2–9. DOI: 10.1038/ncb2897. 
Zeman, Michelle K.; Lin, Jia-Ren; Freire, Raimundo; Cimprich, Karlene A. (2014): DNA damage-
specific deubiquitination regulates Rad18 functions to suppress mutagenesis. In The Journal of 
cell biology 206 (2), pp. 183–197. DOI: 10.1083/jcb.201311063. 
Zhang, Meng; Wang, Lijuan; Zhong, Dongping (2017): Photolyase: Dynamics and Mechanisms of 
Repair of Sun-Induced DNA Damage. In Photochemistry and photobiology 93 (1), pp. 78–92. DOI: 
10.1111/php.12695. 
Zhang, Shuwen; Zou, Shitao; Yin, Deyao; Zhao, Lihong; Finley, Daniel; Wu, Zhaolong; Mao, 
Youdong (2022): USP14-regulated allostery of the human proteasome by time-resolved cryo-EM. 
In Nature 605 (7910), pp. 567–574. DOI: 10.1038/s41586-022-04671-8. 
Zhang, Tianpeng; Rawal, Yashpal; Jiang, Haoyang; Kwon, Youngho; Sung, Patrick; Greenberg, 
Roger A. (2023): Break-induced replication orchestrates resection-dependent template switching. 
In Nature. DOI: 10.1038/s41586-023-06177-3. 
Zhao, Shengkai; Ulrich, Helle D. (2010): Distinct consequences of posttranslational modification 
by linear versus K63-linked polyubiquitin chains. In Proceedings of the National Academy of 
195 
 
Sciences of the United States of America 107 (17), pp. 7704–7709. DOI: 
10.1073/pnas.0908764107. 
Zheng, Ning; Shabek, Nitzan (2017): Ubiquitin Ligases: Structure, Function, and Regulation. In 
Annual review of biochemistry 86, pp. 129–157. DOI: 10.1146/annurev-biochem-060815-014922. 
Zhou, Mengying; Fang, Rui; Colson, Louis; Donovan, Katherine; Hunkeler, Moritz; Song, Yuyu et 
al. (2023): HUWE1 Amplifies Ubiquitin Modifications to Broadly Stimulate Clearance of Proteins 
and Aggregates. In Biorxiv. DOI: 10.1101/2023.05.30.542866. 
Zhu, Lu; Jorgensen, Jeff R.; Li, Ming; Chuang, Ya-Shan; Emr, Scott D. (2017): ESCRTs function 
directly on the lysosome membrane to downregulate ubiquitinated lysosomal membrane proteins. 
In eLife 6. DOI: 10.7554/eLife.26403. 
Zucchelli, Silvia; Codrich, Marta; Marcuzzi, Federica; Pinto, Milena; Vilotti, Sandra; Biagioli, Marta 
et al. (2010): TRAF6 promotes atypical ubiquitination of mutant DJ-1 and alpha-synuclein and is 
localized to Lewy bodies in sporadic Parkinson's disease brains. In Human molecular genetics 19 
(19), pp. 3759–3770. DOI: 10.1093/hmg/ddq290. 
Zuin, Alice; Isasa, Marta; Crosas, Bernat (2014): Ubiquitin signaling: extreme conservation as a 
source of diversity. In Cells 3 (3), pp. 690–701. DOI: 10.3390/cells3030690. 
 
 
 
196