A Degenerate PCNA-Interacting Peptide (DPIP) box targets RNF168 to replicating DNA to limit 53BP1 signaling

TL;DR: In this article, the E3 ligase RNF168 is recruited to DNA replication factories independent of the canonical DSB response pathway regulators and identify a degenerate PCNA-Interacting Peptide (DPIP) motif in the C-terminus of RNF 168 which mediates its binding to PCNA.

Abstract: The E3 ligase RNF168 has been suggested to have roles at DNA replication forks in addition to its canonical functions in DNA double-strand break (DSB) signaling. However, the precise role of RNF168 in DNA replication remains unclear. Here we demonstrate that RNF168 is recruited to DNA replication factories independent of the canonical DSB response pathway regulators and identify a degenerate PCNA-Interacting Peptide (DPIP) motif in the C-terminus of RNF168 which mediates its binding to PCNA. An RNF168 mutant harboring substitutions in the DPIP box fails to interact with PCNA and is not recruited to sites of DNA synthesis, yet fully retains its ability to promote DSB-induced 53BP1 foci. Surprisingly, the RNF168 DPIP mutant also retains the ability to support ongoing DNA replication fork movement, demonstrating that PCNA-binding is dispensable for normal S-phase functions. However, replisome-associated RNF168 functions to suppress the DSB-induced 53BP1 DNA damage response during S-phase. Moreover, we show that WT RNF168 can perform PCNA ubiquitylation independently of RAD18 and also synergizes with RAD18 to amplify PCNA ubiquitylation. Taken together, our results identify non-canonical functions of RNF168 at the replication fork and demonstrate new mechanisms of cross talk between the DNA damage and replication stress response pathways.

Summary

Introduction

• Ubiquitin signaling cascades play central roles in the sensing and repair of DNA damage.
• Once recruited, RNF168/UBC13 mediates additional K63-linked poly-ubiquitylation of the ubiquitylated H2A histones.
• Although TLS confers DNA damage tolerance, Y-family polymerases are error-prone and can increase the risk of mutagenesis.
• Given the impact of E3 ubiquitin ligases on DNA damage tolerance and genome stability in both normal development and pathological states it is important to understand how these enzymes are targeted to different genome maintenance pathways.
• Using targeted mutagenesis, the authors demonstrate that the DSB and DNA replication functions of RNF168 are completely separable and that the DPIP motif is required to both limit the 53BP1-dependent DDR pathway at replication forks and promote ubiquitylation of PCNA at replication forks.

Materials and Methods

• Cell culture and transfection - Cancer cell lines H1299, U2OS, 293T were purchased from the American Type Culture Collection (ATCC) and used for the described experiments without further authentication.
• U2OS RNF168-/- and U2OS RNF8-/- cell lines were gifted by Dr. Daniel Durocher, University of Toronto, Canada.
• Recombinant adenovirus clones were isolated by plaque purification and verified by restriction analysis and Southern blotting.
• For experiments in which cells were stained with PCNA antibody, cells were fixed and permeabilized with ice-cold methanol for 20 minutes prior to antibody staining.
• Fiber track lengths were measured and recorded using ImageJ software (US National Institutes of Health; NIH).

Results

• RNF168 is recruited to DNA replication factories and DSB via independent mechanisms.
• The authors sought to define the new non-canonical mechanism of RNF168 recruitment to sites of DNA replication.
• In similar experiments, a peptide corresponding to the Polh PIP box (which binds PCNA with low affinity, Kd ~400 nM) (40) when compared with p21 (Kd ~76 nM) did not disrupt PCNA-RNF168 interactions as effectively (Fig. 2D).
• As shown in Fig. 2F, RNF168 WT showed reduced binding to PCNA K164R when compared with PCNA WT (indicating that PCNA modification at K164 promotes RNF168 binding).
• The authors conclude that the RNF168-PCNA interaction is dispensable for recruiting 53BP1 to PCNA in NCS-treated cells.

RNF168-PCNA-interactions promote Trans-Lesion Synthesis (TLS)

• Therefore, the PCNA-binding ability of RNF168 is essential for RNF168-induced PCNA ubiquitylation.
• The results of Fig. 5C suggested that RNF168 activates a PCNA-directed E3 ubiquitin ligase other than RAD18, or that PCNA is a direct target of RNF168-mediated E3 ubiquitin ligase activity.
• Given that RNF168 has been reported to form a complex with RAD6 (47), the authors initially used RAD6 as the E2 conjugating enzyme in the in vitro ubiquitylation reactions.

Discussion

• Here the authors identify a new PCNA-interacting motif at the C-terminus of RNF168.
• Recent work (29,30,54) including this study demonstrates similar extensive crosstalk between DSB- and DNA replication stress-induced ubiquitylation cascades.
• Clearly further experiments are necessary to determine the impact of excessive RNF168 on mutability of cancer genomes.
• In addition to ways in which pathological RNF168-PCNA interactions impact cancer biology, it is likely that RNF168-dependent PCNA-Ubiquitylation also participates in normal physiological genome maintenance.
• Further experiments are underway to define the ways in which PCNA interactions influence RNF168 functions at the replication fork.

Figures

Fig. 2 RNF168 contains a Degenerate PCNA-Interacting Peptide (DPIP) domain

Fig. 4 Wild-type and PCNA-interaction-deficient RNF168 support DSB-induced formation of 53BP1 foci

Fig 1 RNF168 is recruited to DNA replication and DSB via separable mechanisms

Fig. 6 Summary of mechanisms by which RAD18 and RNF168 participate in genome maintenance at DSB (left) and DNA replication forks (right)

Fig. 3 Wild-type and PCNA-interaction-deficient RNF168 sustain DNA replication

Fig. 5 RNF168 synergizes with RAD18 to mono-ubiquitinate PCNA and stimulate TLS.

Full text

1
A Degenerate PCNA-Interacting Peptide (DPIP) box targets RNF168 to replicating
DNA to limit 53BP1 signaling
Yang Yang
1#
, Deepika Jayaprakash
1,2#
, Robert Hollingworth
3
, Steve Chen
4,5
, Amy E. Jablonski
4,6
,
Yanzhe Gao
1
, Jay Ramanlal Anand
1
, Elizabeth Mutter-Rottmayer
1
, Jing An
1,7
, Xing Cheng
1,8
, Kenneth H.
Pearce
9
, Sophie-Anne Blanchet
10
, Amélie Fradet-Turcotte
11
*, Grant S. Stewart
3
*, and Cyrus Vaziri
1
*
1
Department of Pathology and Laboratory Medicine, University of North Carolina at Chapel Hill, 614 Brinkhous-
Bullitt Building, Chapel Hill, NC 27599, USA.
2
Oral and Craniofacial Biomedicine Program, Adam’s School of Dentistry, University of North Carolina at
Chapel Hill, NC 27599, USA.
3
Institute of Cancer and Genomic Sciences, University of Birmingham, Birmingham, B15 2TT UK.
4
Cytiva Life Sciences, Global Life Sciences Solutions USA LLC, Marlborough, MA 01752, USA
5
Molecular Devices, LLC, 3860 N First Street, San Jose, CA 95134
6
Advanced imaging group at Miltenyi Biotec, Inc. Gaithersburg, MD.
7
Institute of Cancer Prevention and Treatment, Harbin Medical University, Harbin, China
8
Department of Neuro-Oncology, Chongqing University Cancer Hospital & Chongqing Cancer Institute &
Chongqing Cancer Hospital, Chongqing, China
9
Center For Integrated Chemical Biology and Drug Discovery, Marsico Hall, 125 Mason Farm Road, CB#
7363, Chapel Hill, NC, 27599
10
CHU de Québec-Université Laval Research Center (Oncology division), Université Laval Cancer Research
11
Center and Department of Molecular Biology, Medical Biochemistry and Pathology, Faculty of Medecine,
Université Laval, Québec, Canada
#
Co-first authors
*Correspondence: cyrus_vaziri@med.unc.edu (C.V.), g.s.stewart@bham.ac.uk (G.S.S.), Amelie.fradet-
turcotte@crchudequebec.ulaval.ca (A.F.T.)
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted March 18, 2021. ; https://doi.org/10.1101/2021.03.17.435897doi: bioRxiv preprint

2
Abstract
The E3 ligase RNF168 has been suggested to have roles at DNA replication forks in addition to its canonical
functions in DNA double-strand break (DSB) signaling. However, the precise role of RNF168 in DNA
replication remains unclear. Here we demonstrate that RNF168 is recruited to DNA replication factories
independent of the canonical DSB response pathway regulators and identify a degenerate PCNA-Interacting
Peptide (DPIP) motif in the C-terminus of RNF168 which mediates its binding to PCNA. An RNF168 mutant
harboring substitutions in the DPIP box fails to interact with PCNA and is not recruited to sites of DNA
synthesis, yet fully retains its ability to promote DSB-induced 53BP1 foci. Surprisingly, the RNF168 DPIP
mutant also retains the ability to support ongoing DNA replication fork movement, demonstrating that PCNA-
binding is dispensable for normal S-phase functions. However, replisome-associated RNF168 functions to
suppress the DSB-induced 53BP1 DNA damage response during S-phase. Moreover, we show that WT
RNF168 can perform PCNA ubiquitylation independently of RAD18 and also synergizes with RAD18 to
amplify PCNA ubiquitylation. Taken together, our results identify non-canonical functions of RNF168 at the
replication fork and demonstrate new mechanisms of cross talk between the DNA damage and replication
stress response pathways.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted March 18, 2021. ; https://doi.org/10.1101/2021.03.17.435897doi: bioRxiv preprint

3
Introduction
Ubiquitin signaling cascades play central roles in the sensing and repair of DNA damage. Failure to sustain
DNA damage-inducible ubiquitin signaling compromises DNA repair resulting in increased sensitivity to
genotoxic agents and genomic instability. DNA double-strand breaks (DSBs) such as those induced by
ionizing radiation (IR) or radiomimetic agents are a potent inducer of the ubiquitin-dependent cellular DNA
damage response (Ub-DDR), which is mediated primarily by the E3 ligases, RNF8 and RNF168. The
mechanism of RNF8-RNF168 activation in response to DSBs is well-established (1-3) and is initiated by the
DSB-responsive checkpoint kinase ATM which phosphorylates the histone H2A variant H2AX located in
chromatin proximal to a break. Phosphorylated H2AX (termed γH2AX) recruits MDC1 which is also
phosphorylated by ATM. The phosphorylation of MDC1 facilitates the recruitment of RNF8 to sites of damage,
whereby it catalyzes the ubiquitylation of H1 and H2A-type histones surrounding the break. RNF8-dependent
ubiquitylation of histones stimulates the recruitment of another E3 ligase complex, RNF168/UBC13 via direct
interaction. Once recruited, RNF168/UBC13 mediates additional K63-linked poly-ubiquitylation of the
ubiquitylated H2A histones. The sequential RNF8 and RNF168-dependent ubiquitylation of histones, together
with other post-translational modifications already present at the break creates a platform for repair factors
such as 53BP1, RNF169, and BRCA1 (4-6). 53BP1 is a critical DNA repair factor that dictates the pathway
choice between NHEJ (non-homologous DNA end-joining) and homologous recombination (HR)-dependent
DSB repair through its ability to suppress DNA end-resection, a process that is essential for initiating HR (7,8).
The importance of the RNF168-dependent Ub-DDR in genome maintenance is evident from the existence of
patients with 'RIDDLE' syndrome which is caused by bi-allelic mutations in RNF168. Patients with RIDDLE
syndrome are immunodeficient and unable to produce IgG owing to defects in class switch recombination
(CSR). Cells from RIDDLE syndrome patients are hypersensitive to ionizing radiation and fail to recruit 53BP1
to sites of DNA DSBs, recapitulating phenotypes of cells lacking H2AX, RNF8 and MDC1 (4,9).
In addition to the vital roles of ubiquitin signaling in the DSB response, ubiquitylation also plays key roles in
sensing and processing of DNA damage or 'stress' that arises during DNA replication. DNA replication stalling
can result from exogenously-induced bulky DNA lesions or DNA synthesis through highly repetitive regions of
chromatin. The uncoupling of replicative DNA helicase and polymerase activities at stalled DNA replication
forks generates ssDNA tracts (10) which are prone to breakage, a process often described as 'fork collapse'.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted March 18, 2021. ; https://doi.org/10.1101/2021.03.17.435897doi: bioRxiv preprint

4
To sustain S-phase and avert the lethal consequences of fork collapse, eukaryotic cells employ genome
maintenance processes termed 'DNA damage tolerance' and 'DNA damage avoidance' which are also
regulated by ubiquitin signaling cascades (11). The DNA damage tolerance pathway is triggered by activation
of RAD18, an E3 ligase which interacts with ssDNA at stalled replication forks (12,13). RAD18 cooperates with
the E2 ubiquitin conjugating enzyme RAD6 to mono-ubiquitylate PCNA at stalled replication forks (14). PCNA
mono-ubiquitylation helps recruit specialized Y-family Trans-Lesion Synthesis (TLS) DNA polymerases to
stalled replication forks (15). The Y-family DNA TLS polymerases (Polκ, Polη, Polι and REV1) have ubiquitin-
binding domains and PCNA-interacting peptide motifs (termed 'PIP boxes') that favor PCNA binding when
PCNA is in its ubiquitylated state (16). TLS polymerases are damage-tolerant when compared with the
replicative DNA polymerases (Polδ, Polε) and are able to sustain DNA synthesis in cells harboring damaged
genomes. TLS-deficient cells are intolerant of DNA damage and typically accumulate excessive ssDNA in S-
phase which leads to checkpoint activation and lethal abortive mitoses (17,18). However, although TLS
confers DNA damage tolerance, Y-family polymerases are error-prone and can increase the risk of
mutagenesis. Each Y-family polymerase has a preferred cognate lesion that is bypassed with relative accuracy
when compared with other TLS polymerases. For example, Polη bypasses UV-induced cyclobutane
pyrimidine dimers (CPD) in an error-free manner and represses sunlight-induced mutagenesis (19). In Polη-
deficient individuals (a congenital disease termed xeroderma pigmentosum-Variant or XPV), error-prone
compensatory bypass of CPD adducts by alternative TLS polymerases leads to mutations and skin cancer-
propensity (20).
Interestingly, RAD18-induced PCNA mono-ubiquitylation can also promote a relatively error-free DNA damage-
avoidance pathway termed 'Template Switching' (TS) to sustain S-phase in cells with damaged genomes
(21,22). TS uses fork-reversal as a mechanism for the leading strand to avoid damaged DNA and instead use
the intact replicated lagging DNA strand as a template. In mammalian cells, the template-switching process
requires two E3 ubiquitin ligases, HLTF and SHPRH, which further extend the K164-mono-ubiquitin moiety on
PCNA, generating K63-linked poly-ubiquitin chains. K63-poly-ubiquitylated PCNA recruits helicases and
chromatin-remodeling factors that mediate fork reversal and the template- switching process (23,24). Thus,
ubiquitin signaling is crucial for normal responses to both DNA DSB and DNA replication stress.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted March 18, 2021. ; https://doi.org/10.1101/2021.03.17.435897doi: bioRxiv preprint

5
Increasingly it is clear that E3 ubiquitin ligases are versatile enzymes that can ubiquitylate multiple substrates,
have non-catalytic functions and participate in diverse biological processes. For example, RAD18 responds to
DSB and potentiates HR via a non-catalytic mechanism involving chaperoning of the recombinase RAD51D to
DSB sites (25). RAD18 was recently shown to mono-ubiquitylate histones in response to DSB (26). Although
RAD18 is the main PCNA-directed E3 ubiquitin ligase in most cells, other E3 ligases including HLTF (27) and
RNF8 may also mono-ubiquitylate PCNA (28). RNF168 was recently detected at DNA replication forks (29-31)
and shown to promote normal S-phase progression (29). Therefore, E3 ligase-directed Ub-dependent signaling
cascades are rarely specific to an individual type of DNA damage but rather form a complex network of
integrated pathways that respond to a variety of different cellular stresses.
The mechanisms by which E3 ubiquitin ligases are targeted to different pathways potentially have enormous
impact on DNA damage tolerance in cancer cells. Inappropriate expression levels of E3 ubiquitin ligases, or
their inappropriate deployment to potential downstream substrates could lead to imbalance in DNA repair
pathway choice that results in genetic instability and/or therapy-resistance. Interestingly, RNF168 and RAD18
are both aberrantly overexpressed in many cancers (32-35), although the impact of altered RNF168/RAD18 on
tumorigenic phenotypes is poorly understood.
Given the impact of E3 ubiquitin ligases on DNA damage tolerance and genome stability in both normal
development and pathological states it is important to understand how these enzymes are targeted to different
genome maintenance pathways. Therefore, we sought to elucidate the mechanisms that target RNF168 to
sites of DNA synthesis. Here we identify a degenerate PIP box (DPIP) located within the C-terminus of
RNF168 that mediates its recruitment to replication forks. Using targeted mutagenesis, we demonstrate that
the DSB and DNA replication functions of RNF168 are completely separable and that the DPIP motif is
required to both limit the 53BP1-dependent DDR pathway at replication forks and promote ubiquitylation of
PCNA at replication forks.
Materials and Methods
Cell culture and transfection - Cancer cell lines H1299, U2OS, 293T were purchased from the American
Type Culture Collection (ATCC) and used for the described experiments without further authentication. U2OS
RNF168
-/-
and U2OS RNF8
-/-
cell lines were gifted by Dr. Daniel Durocher, University of Toronto, Canada.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprintthis version posted March 18, 2021. ; https://doi.org/10.1101/2021.03.17.435897doi: bioRxiv preprint

Frequently asked questions

Q1. What contributions have the authors mentioned in the paper "A degenerate pcna-interacting peptide (dpip) box targets rnf168 to replicating dna to limit 53bp1 signaling" ?

The mechanism of RNF8-RNF168 activation in response to DSBs is well-established and is initiated by the DSB-responsive checkpoint kinase ATM which phosphorylates the histone H2A variant H2AX located in chromosome proximal to a break this paper. 

Q2. What is the role of RNF168 in DNA replication forks?

Because RNF168 is a rate-limiting mediator of histone ubiquitylation following acquisition of DSB, there has been tremendous interest in determining the downstream targets and consequences of excessive RNF168 activity. 

Q3. What mechanism would disrupt the interaction between RNF168 and PCNA?

The authors reasoned that if RNF168 binds to PCNA via a PIP-box-mediated mechanism, the interaction would be competitively disrupted by a PIP box-containing peptide. 

Q4. What mechanism is involved in the interaction between RNF168 and PCNA?

The authors conclude that RNF168 associates with PCNA via a degenerate PIP-box and that PCNA K164 mono-ubiquitylation further stimulates this binding. 

Q5. What mechanism is required for RNF168 to sustain normal replication fork rates?

The authors surmise that global RNF168 activity may be sufficient for sustaining normal DNA replication fork rates, or that additional RNF8-dependent mechanisms exist that recruit RNF168 to replication forks. 

Q6. What is the role of RNF168 in regulating 53BP1?

To test the role of RNF168-PCNA interactions in regulating 53BP1, the authors developed an accurate and quantitative live cell imaging platform for measuring RNF168-dependent and cell cycle phase-specific subcellular localization of 53BP1 in response to DNA DSB in asynchronous cultures. 

Q7. What mechanism could be used to limit the recruitment of 53BP1 to DSBs?

The authors reasoned that the PCNA-association of RNF168 during S-phase might provide a new mechanism for restraining RNF168 and limiting the recruitment of 53BP1 to DSBs. 

Q8. What is the effect of RNF168 on DSB repair?

ectopically-expressed RNF168 promotes mutagenic NHEJ but does not affect physiological NHEJ-mediated repair of DSBs that arise during immunoglobulin class switch recombination or CSR (35). 

Q9. What is the effect of mutating the DPIP to a high-affinity?

In contrast, mutating the DPIP to a high-affinity PIP box (super-PIP) significantly enhanced the RNF168-dependent ubiquitylation of PCNA.