Structural and Functional Insight into Proliferating Cell Nuclear Antigen

Abstract

Proliferating cell nuclear antigen (PCNA) is a critical eukaryotic replication accessory factor that supports DNA binding in DNA processing, such as DNA replication, repair, and recombination. PCNA consists of three toroidal-shaped monomers that encircle double-stranded DNA. The diverse functions of PCNA may be regulated by its interactions with partner proteins. Many of the PCNA partner proteins generally have a conserved PCNA-interacting peptide (PIP) motif, located at the N- or C- terminal region. The PIP motif forms a 3₁₀ helix that enters into the hydrophobic groove produced by an interdomain-connecting loop, a central loop, and a C-terminal tail in the PCNA. Post-translational modification of PCNA also plays a critical role in regulation of its function and binding partner proteins. Structural and biochemical studies of PCNA-protein will be useful in designing therapeutic agents, as well as estimating the outcome of anticancer drug development. This review summarizes the characterization of eukaryotic PCNA in relation to the protein structures, functions, and modifications, and interaction with proteins.

Introduction

Proliferating cell nuclear antigen (PCNA) is a highly conserved protein in archaebacteria and several eukaryotic species [67]. Studies of all known PCNAs show that they are conserved in amino acid sequences, functions, and structures [5,54,68]. PCNA was initially reported as a fundamental auxiliary protein for the processes of DNA replication and repair, and it works as a DNA sliding clamp; that is, it stabilizes the interaction with other proteins through a sliding platform [14,30]. The DNA sliding clamps perform an essential function in DNA metabolic processes, and are found in eukaryotes, prokaryotes, and archaea [24,31,37,71,74,83]. They are ring-shaped six-domain proteins that encircle duplex DNA and enable highly processive DNA replication by serving as binding sites for DNA polymerases. As a DNA clamp, PCNA acts as a scaffold protein that organizes various components for DNA replication, chromatin remodeling, DNA damage repair, and cell cycle progression [47,55]. PCNA is a homotrimer ring and it forms in a head-to-tail arrangement of three monomers [36,47]. This ring-shaped trimeric PCNA is loaded onto DNA. The inner surface of the PCNA is formed by 12 positively charged α-helices that interact with DNA, and the outer layer contains 54 β-sheets and interdomain-connecting loops (IDCLs) for protein-protein interactions [47,82]. In addition, PCNA interacts with several translesion synthesis (TLS) DNA polymerases [55,84] and participates in repair of DNA that has been damaged by chemotherapy agents [78], thereby conferring chemotherapy resistance. PCNA was initially discovered to be an antigen to the autoimmune disease systemic lupus erythematosus [46,71]. The level of expression of PCNA changes during cell cycles and is associated with cell proliferation or transformation [4]. A great deal of work was conducted to determine the roles of PCNA in DNA replication and one of the functions identified was a sliding clamp for DNA polymerase δ [54,68]. Deregulation of PCNA expression is a hallmark of many proliferative diseases and PCNA acts as a general proliferative marker, particularly in determining cancer prognosis [34,65].

The homotrimeric complex of PCNA consists of three domains; the amino-terminal domain, the IDCL, and the carboxyl-terminal domain. The IDCL links the N- and C-terminal domains and makes significant contributions to the diverse cellular activity of PCNA [21]. Investigations of PCNA have shown that the IDCL is a major interaction site for various binding proteins [19], including polymerases Polδ, p21, DNA-(cytosine-5) methyltransferase (MeCTr), DNA ligase 1 (LIG1), flap endonuclease 1 (FEN1), cyclin-dependent kinase 2 (CDK2), and cyclin D, etc. Whereas it is known that many proteins and peptides bind to PCNA through conserved motifs, the binding mechanisms between PCNA and partner proteins are still unknown. A large part of the PCNA-binding proteins comprises a conserved motif referred to as the PCNA-interacting peptide (PIP) box [24,28,49]. The IDCL of PCNA is the binding site that contains the consensus PIP-box motif, Q-X-X-(I/L/M)-X-X-(F/Y)-(F/Y) [32,84]. The PIP-box has I/L/M small hydrophobic residues, F/Y aromatic side chains, and X residues. Recently, another PCNA-binding motif termed the KA-box [39], K-A(A/L/I)-(A/L/Q)-x-x-(L/V), has been identified. It is also introduced in many PCNA-interacting proteins and is associated with the canonical PIP-box [62,87]. Polη, Polι, and Polκ have a noncanonical PIP-box sequence [25]. Other binding regions are placed in the N-terminal region, including the interaction between α-helices of PCNA and cyclin D [52]. In addition, the region is in the C-terminal region, which is important for the interaction with Polε, CDK2, replication factor C (RFC), and growth-arrest and DNA-damage-inducible protein 45 (Gadd45) [39,52]. This review focuses on the structural characterization and function of PCNA. The information will be useful in designing therapeutic agents, as well as estimating the outcome of anticancer drug development.

 

Major Role of PCNA in the DNA Replication and Repair System

PCNA is a crucial replication factor that binds with many partner proteins included in DNA replication and damage repair. In mammals, DNA replication is started by phosphorylation of the origin recognition complex at the site of origin [8,67]. PCNA is a homolog of the β subunit of the DNA polymerases in prokaryotes, known as the sliding clamp for eukaryotic DNA polymerases δ/ε [54,68]. It interacts with the Polδ switch to facilitate extension of Okazaki fragments [79]. PCNA functions as a clamp platform for polymerases δ and ε as well as a variety of proteins at the replication. It is loaded onto DNA junctions by the action of a multiple clamp loader, RFC, which couples with ATP hydrolysis to open and close the PCNA ring during replication and repair [23,41]. PCNA exists as stable trimers that form a closed ring with a central hole, which encircles the DNA (Fig. 1). It interacts with and recruits FEN1 and LIG1, which are required for Okazaki fragment p rocessing (Fig. 2) [39]. The PCNA complex for the Okazaki fragment maturation ensures the directivity of replication, and PCNA is a crucial moderator of DNA replication. During the DNA replication, synthesis over damaged templates is accomplished by polymerases via translesion synthesis (TLS) [56,90]. Conserved TLS polymerases identified in several species includes Y-family polymerases η, ι, and κ, and Rev1 and Rev7 as well as B-family polymerase ζ [7,53]. Mono-ubiquitination of PCNA stimulates the recruitment of TLS polymerase to the replication fork [23,63]. PCNA provides the central stool on which TLS polymerase can interact to acquire access to the replicative stalled at the lesion site and to perform their roles in modulation of different lesion bypass processes. TLS polymerase works individually or in pairs according to the damage type. PCNA enhances polymerase activity, resulting in up-regulated ability of nucleotide on the other sites to the damaged template [53,61].

Fig. 1.Structural features of PCNA. (A) Front view of the three-dimensional structure of human PCNA in complex with the p21 peptide (PDB ID, 1AXC). PCNA is a trimeric molecule, with each monomer containing a PIP-box binding site located between the C-terminus of the protein and the interdomain connector loop (IDCL). The side view shows the location of the IDCL on the front side of PCNA and the extended loops on the back side. The N-terminal domain is colored in red, the C-terminal domain is colored in blue, and the IDCL is colored in green. (B) The molecular surface of PCNA showing its charge distribution. Negatively charged residues are shown in red; positively charged residues are shown in blue.

Fig. 2.DNA replication and repair. (A) During DNA replication, replicative polymerases, including DNA polymeraseδ (Polδ), are associated with PCNA and ensure progression of the replication fork mediated by the MCM complex. (B) Upon encountering a replication block such as DNA damage, PCNA modified by ubiquitylation (Ub) plays a key role in recruiting translesion synthesis DNA polymerases, including Polη, which initiates damage bypass. PCNA exists as stable trimers that form a closed ring with a hole in the center that encircles duplex DNA.

PCNA plays an important role in the DNA damage repair and DNA replication system [72]. The major metabolic pathways for DNA repair systems implicate nucleotide excision repair (NER), base excision repair (BER), double-strand break repair, and mismatch repair (MMR) [10,39,85]. DNA damage by certain chemicals and UV-irradiation results in bulky lesions, which are then repaired via the NER pathway. During this process, PCNA binds to the endonuclease, XPG (xeroderma pigmentosum complementation group G) and facilitates new DNA fragment resynthesis, which occurs after reactions catalyzed by XPG [20]. PCNA is specifically loaded at the major cellular activity site of the XPG 3’-incision to the lesion for repairing [39]. BER is responsible for replacing chemically altered nucleotide bases in DNA and can operate in either short- or long-patch modes. PCNA is associated with DNA repair in long-patch mode, which involves a DNA polymerase δ/ε-dependent mechanism. It has been observed to bind with various BER proteins, such as AP-endonuclease 1 (APE1), AP-endonuclease 2 (APE2), uracil-DNA glycosylase 2 (UNG2), nth endonuclease III-like 1, methylpurine-DNA glycosylase (MPG), human MutY homolog, and X-ray repair cross-complementing protein 1 (XRCC1) [49]. It is possible that PCNA functions as a bridge for BER proteins and stimulates their activities and acts as a coordinator for the repair process [67]. MMR amends misincorporated bases, which can produce from small insertion/deletion and polymerase error loops achieved during recombination and replication. It also operates for the beginning stages of damage recognition [74]. The repair machinery takes away the error-comprising part of a freshly synthesized strand, and repairs targets to the newly generated single-stranded gap [47]. In MMR, PCNA is needed for repair synthesis and the beginning stages of damage recognition. MMR needs to discriminate between the original and newly synthesized strand to function properly. Because PCNA is loaded onto the DNA in the only possible orientation, facing the 3’-end of the daughter strand, discrimination is possible. Indeed, exonuclease excisions of incorrectly incorporated nucleotides in the growing strand are carried out in the 5’-3’ direction. PCNA interacts directly with Msh6 (MutS homolog 6), Msh3 (MutS homolog 3), Mlh1 (MutL homolog 1), and EXO1 (exonuclease 1). MLH1 possesses PIP-boxes, MSH3, and MSH6 [65].

 

Post-Translational Modifications of PCNA

PCNA is modified by several post-translational modifications, including ubiquitylation, sumoylation, acetylation, and phosphorylation. PCNA was recently shown to be subject to nitrosylation at specific cysteine residues for which the biological significance remains to be determined [1,26]. Post-translational modification of PCNA seems to be important for the polymerase switch, with post-translational modification by ubiquitin being the best known. Monoubiquitiylation of PCNA at Lys164 is induced by Rad6 and Rad18 in a DNA damage-dependent manner [29,60,64], which serves as a signal for activation of the translesion synthesis pathway [75]. It is achieved by consequent movement of the ubiquitin-activating enzyme E1, specific ubiquitin-conjugation enzyme E2 (which in humans might be either Rad6A or Rad6B), and RING (really interesting new gene) finger-containing E3 ubiquitin ligase (Rad18) [26,91]. Mono-ubiquitylation of PCNA directs a switch between processive DNA polymerase and TLS DNA polymerase and results in error-prone bypass replication [47]. In addition, polyubiquitylation of PCNA at Lys164 and Lys63 needs the heterodimeric ubiquitin-conjugation enzyme Ubc13-Mms2 and a specific RING-finger-containing E3 ubiquitin ligase, Rad5 (in yeast) [75]. Rad5 promotes PCNA polyubiquitylation via interactions with both Ubc13-Mms2 and PCNA. In humans, Rad5 orthologs, SNF2 histone linker PHD RING helicase, helicase-like transcription factor, and RING finger protein 8, have been found to be catalyzed by Mm2-Ubc13-dependent polyubiquitylation of PCNA [11,76]. Another PCNA modification is sumoylation on Lys164, and to a lesser portion, Lys127, by the E2 SUMO conjugating enzyme Ubc9 combination with E3 SUMO ligase Siz1. The crystal structures of mono-ubiquitylated, polyubiquitylated, and sumolylated PCNA were solved [17,18] and they showed that the modified positions were located on the backside of PCNA (Figs.3A and 3B). PCNA sumoylation shows inhibition effects on the interaction of PIP-box protein and PCNA. However, it has not yet been characterized in mammals. Modification of PCNA by ubiquitin and SUMO modulates the function of its target protein by modifying, creating, or blocking the binding motif. Both SUMO and ubiquitin are targeted to Lys164 residues of PCNA, and these modifications control the various functions of PCNA [64]. In order to understand the difference and similarity between post-translationally modified forms of PCNA, the structures of UbiPCNA and SUMOPCNA were superimposed (Fig.3 C). Although these modification positions coincide, the modifier positions are quite different in both structures. UbiPCNA and SUMOPCNA structures revealed that the ubiquitin and SUMO moieties are located on the back face of the PCNA ring and interact with the loop of PCNA via its hydrophobic surface [73]. The SUMO moiety is a more radial position than the ubiquitin. This difference in modifier position is probably caused by the longer flexible linker at the C-terminus of ubiquitin in comparison with that of SUMO [11,17]. Recently, new modifications of PCNA have been reported. Aspartic acid and glutamic acid residues of PCNA undergo esterification: methyl esterification on several aspartic acids and glutamic acids residues. Interestingly, PCNA methylation is associated with breast cancer and believed to be cancer-specific [27]. Use of 2D-PAGE and a specifically generated antibody revealed that an acidic isoform of PCNA, cancer-specific PCNA (csPCNA), is exclusively expressed in malignant tissues, including breast cancer, prostate cancer, and esophageal adenocarcinoma, but not in normal cells [82]. The functional results of these modifications are not clearly known. The methyl esterification of PCNA is likely to induce conformational changes in the structure of PCNA, and it may facilitate or disrupt the interaction with partner proteins. Therefore, PCNA has the potential for use in the development of new cancer markers and targeting of csPCNA in cancer cells.

Fig. 3.PCNA modifications. (A and B) Side view of SUMO- and ubiquitin-modified PCNA. The SUMO (orange) and ubiquitin (magenta) groups are located on the back of PCNA (cyan). SUMO and ubiquitin are oriented differently relative to PCNA. Left, ribbon representation; middle, analogous space-filling representation; right, molecular surface representation of the SUMO- and ubiquitin-modified PCNA. (C) Structure of a single SUMOPCNA subunit superimposed with the structure of ubiquitin from UbiPCNA is shown. The PDB ID is 3PGE for the SUMO-modified PCNA and 3L10 for the ubiquitin-modified PCNA.

 

PCNA in Cell Cycle and Apoptosis

PCNA-interacting proteins play major roles in the regulation of the cell cycle. PCNA itself is a cyclin [49] that is highly up-regulated during the S-phase and it binds to cyclin-CDK complexes [86] as well as the CDK inhibitor, p21 [21]. These interactions produce a PCNA-p21/CDK-cyclin quaternary complex that could be independent from the DNA replication machinery [89]. In the regulation of cell cycle, p21 modulates critically the function of PCNA [16]. p21Cip1/Waf1 is the main mediator of growth arrest induced by p53 in response to DNA damage [13,38,69]. Tumor suppressor protein p21 having the PIP-box modulates cell cycle progression by directly binding to PCNA through its C-terminal region [48]

PCNA is involved in the regulation of damage-induced apoptosis and programed cell death. The physical interaction of PCNA with Gadd45 and MyD118 (myeloid cell differentiation protein) has also been shown. They have similar domains that mediate interaction of PCNA, leading to negative cell growth [65,77]. ING1 (inhibitor of growth 1) is a tumor suppressor protein that binds PCNA through the site used by growth regulatory proteins [70]. Specifically, the p33ING1 isoform of ING1 includes a PIP domain that binds with PCNA [59]. Therefore, PCNA can function as a bridging molecule that targets proteins with distinct roles in DNA-based processes [39].

The PCNA structure has been conserved during evolution. Human, rat, mouse, and Drosophila melanogaster PCNAs are highly conserved in primary sequences [80]. The sequences of full-length rat and human PCNAs are conserved, with the exception of four amino acids (Fig. 4A). Human PCNA, which consists of 261 amino acid residues, includes a central hole used for interaction with DNA (Fig. 5) [21]. It is composed of an N-terminal domain (amino acids 1-117), a flexible IDCL (amino acids 118–135) and a C-terminal domain (amino acids 136-261) (Fig.4 B). Crystallographic study has shown that PCNA is composed of a toroid shape structure in a head-to-tail manner [58]. The crystal structure of yeast PCNA was determined and followed by the human PCNA-p21 complex structure [88]. The structures obtained from analysis of yeast, human, archaeal, and plant PCNAs revealed similarities for the DNA polymerase III β subunit [21,36,43,66]. The DNA polymerase III β subunit forms a homodimer with a six-fold symmetrical ring, wherein each monomer consists of three repeating domains [33]. Similarly, PCNA also exhibits a six-fold symmetrical ring that encircles DNA. In contrast to the two-subunit structure of DNA polymerase III β rings, most of PCNA proteins have homotrimeric rings composed of three PCNA homologs (PCNA1, PCNA2, and PCNA3) (Fig. 6A). Head-to-tail arrangement of the three monomers (29 kDa in human) gives rise to two distinct faces, the back and front (C-terminus) [47,67]. Each PCNA monomer consists of two domains connected with an extended β-sheet across the IDCL (Fig. 4A). PCNA monomers bind to antiparallel interactions, resulting in six-fold symmetry [26]. The PCNA ring has an overall negative charge with a positively charged inner surface owing to the existence of Lys and Arg [21]. The positively charged inner surface is formed by α-helices interacted with the negatively charged DNA backbone and the outer surface is composed of β-sheets [45]. The P CNA m onomer b elongs to the α/β protein family, which contains four α-helices and a twisted β-sheet composed of 18 antiparallel β-strands (Fig. 4B) [51]. The interaction with partner proteins occurs on the front side of PCNA where the IDCL is located. The back side of PCNA is the site for post-translational modifications and contains several loops [65,71]. PCNA structures are known to complex with DNA, proteins, and peptides [2,6,21,57]. The complexes formed through binding of PCNA to DNA and proteins/peptides have provided valuable insight into the mechanism by which PCNA functions during DNA processing.

Fig. 4.Domain and secondary structures of PCNAs. (A) Secondary structures of hPCNA, mPCNA, DmPCNA, and rPCNA. The secondary structure is shown according to the hPCNA structure. Alpha helices are shown as rectangles and β-sheets as arrows. Loops are shown as black lines. Residues that are identical between the four species are indicated blue-shadded Box. Sequence alignment was performed by ClustalW and Jalview. (B) Schematic diagram showing domains of the full-length PCNA.

Fig. 5.Structure of PCNA bound to DNA. PCNA-DNA model derived from PDB ID 3K4X. DNA forms a ~40° angle with the central axis of PCNA.

Fig. 6.Binding sites on PCNA interaction partners. (A and B) PCNA interacts with protein partners through the frontal hydrophobic groove organized by the central loop (CL, amino acids 41-44, colored purple), the C-terminal tail (CT, amino acids 254-257, colored magenta), and the interdomain-connecting loop (IDCL, amino acids 118-135, colored orange). (C) PIP-box and p21 PIP-box motifs are shown.

 

PCNA Interaction with Partner Proteins

Interaction with PCNA-partner proteins is a key regulatory role in various PCNA cellular functions. PCNA-interacts directly with many of the proteins involved in various cellular processes [47]. Table 1 represents the major PCNA-dependent activities and the respective PCNA-binding proteins. The interaction sites of PCNA are shown with its partner proteins in Fig.6A. PCNA interacts with partner proteins via the hydrophobic patch on its front-facing side created by the IDCL (amino acids 118-135), the central loop (amino acids 41-44), and the C-terminal tail (amino acids 254-257) [15]. The ICDL is a major interaction site for several proteins, including p21, Pol δ, MeCTr, DNA ligase 1, and FEN1 [39]. The globular N-terminus including α-helices is the interaction site with cyclin D, and the extreme C-terminus is crucial for interactions with Polε, RFC, CDK2, and Gadd45 [9]. Many PCNA-interacting partners include a conserved PIP-box PCNA-binding motif, Q-X-X-(L/M/I)-X-X-(F/Y)-(F/Y), and they are consisted of hydrophobic and aromatic residues for interaction (Fig. 6C) [83]. A general motif controlling PCNA-protein interactions is marked in quite a few of the human proteins [47]. The PIP motif forms a 310 helix that enters into the hydrophobic groove in the PCNA (Fig.6B) [2,6]. Recent studies have focused on PIP-box interaction and identification of an additional modulation protein-protein interface. The crystal structure of human PCNA-p21CIP1/WAF1 showed the typical interaction of a PCNA-PIP box (PDB ID, 1AXC) [21]. Interestingly, the mutation of PCNA with extensive affinity for PIP-box revealed the DNA replication defects and the elevated plasticity of PCNA for partner protein affinities [40]. Some PCNA-interacting proteins do not possess the PIP-box sequence. Instead, a PCNA interaction motif, K-A-(A/L/I)-(A/L/Q)-x-x-(L/V), mediates PCNA interactions [39,87]. However, some PCNA-binding proteins can interact with PCNA independently of the classical PIP-box or the KA-box through another binding site on PCNA [49]. For example, FEN1 utilizes the PIP-box motif and its flanking sequences located at the extreme C-terminal tail, as well as several additional contacts from its globular N-terminal region, to interact with the C-terminus and IDCL on PCNA [57]. Because a large number of PCNA partner candidates by database search studies contain pathways that are irrelevant to DNA replication, DNA repair, or chromatin assembly, the identification of these additional bindings may announce new functions of PCNA [15].

Table 1.PIP-box seqence containing proteins indicated by a bold letter. The proteins are originated from mammals.

 

Summary

PCNA plays crucial roles in DNA replication, DNA repair, the cell cycle, and apoptosis, and it interacts with many partner proteins to accomplish these roles. Conversion of PCNA-binding proteins can be started by phosphorylation, proteolysis, affinity competition, and modification of PCNA by sumoylation and ubiquitylation. PCNA function is controlled by interaction partners as well as post-translational modifications of itself. Many PCNA partner proteins include PIP-box motifs and their binding modes are conserved. A hydrophobic groove at the front of PCNA serves as a docking site for the consensus PIP-box motifs. An interdomain-connecting loop on the front-facing PCNA ring serves as a major interaction site. PCNA expression relates to cell proliferation and it has diagnostic value in many types of cancers. In addition, it is also a target for cancer therapy and its inhibitors are currently being developed as potential anticancer drugs. These studies for the interaction with partner proteins, signaling regulation, or trimer formation of PCNA can approach to develop new therapeutic agents. Two types of PCNA-targeting peptide agents have been reported and some peptides disrupt protein interactions and others prevent the phosphorylation of Tyr211. Functional activation of PCNA may contribute to a favorable patient response and provide a therapeutic tool to overcome the development of resistance to other therapies. Structural and biochemical studies of the PCNA protein may provide a model target for designing therapeutic agents, as well as evaluating the efficacy of anticancer drugs. Further studies are being conducted to identify the complex structures in PCNA and its partner proteins/compounds, as well as structure-based regulation of PCNA.