Maneuvers on PCNA Rings during DNA Replication and Repair

Abstract

DNA replication and repair are essential cellular processes that ensure genome duplication and safeguard the genome from deleterious mutations. Both processes utilize an abundance of enzymatic functions that need to be tightly regulated to ensure dynamic exchange of DNA replication and repair factors. Proliferating cell nuclear antigen (PCNA) is the major coordinator of faithful and processive replication and DNA repair at replication forks. Post-translational modifications of PCNA, ubiquitination and acetylation in particular, regulate the dynamics of PCNA-protein interactions. Proliferating cell nuclear antigen (PCNA) monoubiquitination elicits 'polymerase switching', whereby stalled replicative polymerase is replaced with a specialized polymerase, while PCNA acetylation may reduce the processivity of replicative polymerases to promote homologous recombination-dependent repair. While regulatory functions of PCNA ubiquitination and acetylation have been well established, the regulation of PCNA-binding proteins remains underexplored. Considering the vast number of PCNA-binding proteins, many of which have similar PCNA binding affinities, the question arises as to the regulation of the strength and sequence of their binding to PCNA. Here I provide an overview of post-translational modifications on both PCNA and PCNA-interacting proteins and discuss their relevance for the regulation of the dynamic processes of DNA replication and repair.

Keywords: DNA repair; DNA replication; post-translational protein modifications; proliferating cell nuclear antigen.

Figure 1

The structure of the proliferating cell nuclear antigen (PCNA) ring bound to DNA and the PIP-box of the CDK inhibitor p21. (A) Cartoon presentation of PCNA homotrimer bound to 10 bp dsDNA and p21 PIP-box peptide bound to the interdomain connector loop (IDCL) of each PCNA monomer. The image was obtained by overlaying PCNA-DNA co-structure (6GIS) [5] with PCNA-p21 PIP co-structure (1AXC) [31]. Three PCNA monomers are represented with different colors. (B) Interaction interface between PCNA and PIP-box shown for one PCNA monomer bound by one p21 PIP-box peptide. IDCL (pink), the central loop region (blue) and the C-terminal region (yellow) of PCNA anchor the PIP-box peptide through hydrophobic and electrostatic interactions. The sequence of the p21 PIP-box peptide is shown with the four critical residues indicated in bold. (C,D) Electron density distribution of PCNA from (A,B). The color-coded electrostatic surface potential of PCNA was drawn using the Adaptive Poisson-Boltzmann Solver package. The electrostatic potential ranges from −5 (red) to +5 (blue) kT/e. The images were generated using PyMOL [32].

Figure 2

Functions of PCNA in replication fork stalling. In fork reversal polyubiquitinated PCNA recruits ZRANB3 translocase/structure-specific endonuclease and FANCM. In translesion synthesis monoubiquitinated PCNA recruits translesion synthesis (TLS) polymerases η, κ, ι and REV1 to enable bypass of DNA lesions. Black dashed arrows indicate PCNA targets in different pathways. Ub (ubiquitin); SMARCAL1 (SWI/SNF-related matrix-associated actin-dependent regulator of chromatin subfamily A-like protein 1); HLTF (helicase-like transcription factor); RAD54 (DNA repair and recombination protein RAD54); BRCA1 (breast cancer type 1 susceptibility protein); BRCA2 (breast cancer type 2 susceptibility protein); WRN (Werner syndrome ATP-dependent helicase); RECQ1 (ATP-dependent DNA helicase Q1); DNA2 (DNA replication ATP-dependent helicase/nuclease DNA2); PARP1 (poly(ADP-ribose) polymerase 1); MRN (Mre11-Rad50-Nbs1); RAD51 (DNA repair protein RAD51); BLM (Bloom syndrome protein); TOPO3 (DNA topoisomerase 3).

Figure 3

Functions of PCNA in different DNA repair pathways. In homologous recombination PCNA enhances the processivity of the exonuclease EXO1 during end resection and of Pol δ during DNA repair synthesis. In mismatch repair, PCNA interacts with MutSα to recognize the mismatch, activates the endonuclease activity of MutLα to excise the mismatch, and recruits polymerase δ for DNA repair synthesis. In base excision repair, PCNA recruits polymerases β, δ or ε to displace the damaged base into a flap intermediate. In nucleotide excision repair PCNA interacts with the scaffold protein XPA, activates the endonuclease XPF, targets XPG for degradation and recruits polymerase δ to fill in the gap. Black dashed arrows indicate PCNA targets in different pathways. RPA (replication protein A); CtIP (CtBP-interacting protein); MSH2 (MutS homologue 2); MSH6 (MutS homologue 6); MLH1 (MutL homologue 1); PMS2 (PMS1 homologue 2); APE1 (apurinic/apyrimidinic endonuclease 1); TDP1 (tyrosyl-DNA phosphodiesterase 1); PNKP (polynucleotide kinase 3′-phosphatase); APTX (aprataxin); XRCC1 (X-ray repair cross-complementing protein 1); DDB1 (DNA damage-binding protein 1); DDB2 (DNA damage-binding protein 2); RAD23B (RAD23 homologue B); XPC (Xeroderma pigmentosum complementation group C); CSA (Cockayne syndrome group A); CSB (Cockayne syndrome group B); TFIIH (transcription factor II H); XPB (Xeroderma pigmentosum complementation group B); XPD (Xeroderma pigmentosum complementation group D).

Figure 4

Different binding modes of PIP-box motifs to PCNA. Electrostatic surface potential of the interaction interface on PCNA and stick representation of PIP-box peptides in green are shown for (A) p21 (1AXC) [31], (B) FEN1 [38], (C) PARG [27], and (D) DNA polymerase τ (2ZVM) [21]. The sequences of PIP-box peptides are shown with the four critical residues indicated in bold. The co-structure of PCNA and full-length FEN1 reveals additional binding sites within the FEN1 core, indicated in blue. PIP-boxes from p21 and FEN1 have a canonical sequence and a canonical mode of binding; PARG has a non-canonical sequence but a canonical mode of binding; Pol τ has a non-canonical sequence and a non-canonical mode of binding. The color-coded electrostatic surface potential of PCNA was drawn using the Adaptive Poisson-Boltzmann Solver package. The electrostatic potential ranges from −5 (red) to +5 (blue) kT/e. The images were generated using PyMOL [32].

Figure 5

Multivalent binding modules of PCNA-interacting proteins. (A) TLS polymerases bind PCNA IDCL through a PIP-box motif and monoubiquitinated PCNA through ubiquitin-binding domains (UBD). Overlay of monoubiquitinated PCNA (3TBL) [46], Pol τ PIP-PCNA co-structure (2ZVM) [21], and Pol τ UBM2-ubiquitin co-structure (2KHW) [62]. Pol τ PIP-box binds the IDCL at the front face of PCNA, while Pol τ UBM2 binds ubiquitin positioned radially towards the back face of PCNA. A dashed line represents the region between Pol τ PIP-box (445–457aa) and UBM2 (701–732aa). The 59Å distance between the last residue of the PIP-box and the first residue of UBM2 allows a flexible conformation of the intervening region. (B) Yeast Srs2 binds PCNA IDCL through a PIP-box motif (1148–1161aa) and SUMOylated PCNA on the back face of the PCNA ring through a SUMO-interacting motif (SIM) (1168–1174aa) (3V62) [20]. However, the 36Å distance between the last residue of the PIP-box and the first residue of SIM is too large to be filled by the six intervening residues, rendering a multivalent binding mode implausible according to this structure. The images were generated using PyMOL [32].

Figure 6

Two models illustrating protein maneuvers on PCNA rings. (A) In the ‘piggyback model’ a replicative polymerase binds to the front face of PCNA in the direction of DNA replication, whereas a TLS polymerase rides piggyback. Upon DNA damage, PCNA monoubiquitination instigates polymerase switching, whereby the replicative polymerase dissociates and is replaced by a TLS polymerase, which bypasses the lesion. (B) In the ‘toolbelt model’, PCNA trimers can accommodate simultaneously up to three proteins. This can facilitate their coordinate exchange, as in the case of Pol β, FEN1, and LIG1 during Okazaki fragment maturation. This can also facilitate the selection of the appropriate TLS polymerase (Pol η, Pol κ or Pol τ) during translesion synthesis.

Figure 7

Regulatory regions in PCNA-interacting proteins have PCNA-interacting modules and are modified by different PTMs. PCNA-interacting modules: PIP (PCNA-interacting protein motif); UBZ (ubiquitin-binding zinc finger motif); UBM (ubiquitin-binding motif); RIR (Rev1-interacting region motif). Other modules: IDCL (interdomain connector loop); NLS (nuclear localization signal); NES (nuclear export signal); MLS (mitochondrial localization signal). Colored dots represent modified residues (precise location is given in Supplementary Table S1). Black-circled dots denote overlapping modifications. Dots labelled with numbers indicate residues that were functionally characterized. PARG has an additional PIP-box at the N-terminus, which does not bind PCNA in vitro but is important for PARG localization within replication foci [27,141]. Pol η has two more internal PIP-boxes, which do now show binding to PCNA in vitro, but are important for PCNA monoubiquitination and stimulation of Pol η polymerase activity by PCNA [58]. PTMs were extracted from high-throughput mass spectrometry studies or low-throughput functional studies [69,71,76,78,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,131,132,133,134,135,136,137,138,139,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164].

Figure 8

Regulation of PCNA functions in DNA replication and repair by post-translational modifications. PCNA monoubiquitinatation on K164 after DNA damage promotes ‘polymerase switching’, whereby stalled replicative polymerase is replaced with a specialized TLS polymerase. PCNA polyubiquitination on K164 after DNA damage triggers replication fork reversal through recruitment of ZRANB3 translocase/structure-specific endonuclease. PCNA SUMOylation on K164 recruits anti-recombinases PCNA-interacting protein (PARI) or Srs2 in human cells and yeast respectively to prevent recombination during replication. PCNA SUMOylation on K127 in yeast weakens the interaction of the acetyl transferase Eco1 with PCNA, which in turn impairs cohesion establishment. PCNA acetylation on K14 in human cells facilitates PCNA dissociation upon completion of DNA repair. PCNA acetylation on K20 in yeast reduces the processivity of Pol δ and promotes homologous recombination. PCNA methylation on K248 enhances its interaction with FEN1. PCNA phosphorylation on Y211 weakens its interaction with mismatch repair proteins.

Figure 9

Regulation of PCNA-interacting proteins by post-translational modifications. Examples are given for two PCNA-interacting proteins: Pol η and FEN1. Monoubiquitination of Pol η suppresses monoubiquitinated PCNA and Pol η-mediated translesion synthesis. Pol η phosphorylation weakens its binding to PCNA and may promote its dissociation from replication forks upon completion of translesion synthesis. Pol η SUMOylation promotes Pol η recruitment to difficult-to-replicate common fragile sites. FEN1 methylation promotes its interaction with PCNA during Okazaki fragment maturation. FEN1 methylation antagonizes its phosphorylation, which abrogates its binding to PCNA. FEN1 phosphorylation promotes its SUMOylation and ubiquitination leading to FEN1 proteasome-mediated degradation in G2/M phase of the cell cycle. FEN1 acetylation negatively regulates its binding to DNA and its nuclease activity.