Implications of Translesion DNA Synthesis Polymerases on Genomic Stability and Human Health

Abstract

Replication fork arrest-induced DNA double strand breaks (DSBs) caused by lesions are effectively suppressed in cells due to the presence of a specialized mechanism, commonly referred to as DNA damage tolerance (DDT). In eukaryotic cells, DDT is facilitated through translesion DNA synthesis (TLS) carried out by a set of DNA polymerases known as TLS polymerases. Another parallel mechanism, referred to as homology-directed DDT, is error-free and involves either template switching or fork reversal. The significance of the DDT pathway is well established. Several diseases have been attributed to defects in the TLS pathway, caused either by mutations in the TLS polymerase genes or dysregulation. In the event of a replication fork encountering a DNA lesion, cells switch from high-fidelity replicative polymerases to low-fidelity TLS polymerases, which are associated with genomic instability linked with several human diseases including, cancer. The role of TLS polymerases in chemoresistance has been recognized in recent years. In addition to their roles in the DDT pathway, understanding noncanonical functions of TLS polymerases is also a key to unraveling their importance in maintaining genomic stability. Here we summarize the current understanding of TLS pathway in DDT and its implication for human health.

INTRODUCTION

DNA damage in cells is a continuous process resulting in the creation of DNA lesions via exogenous and endogenous factors (Fig. 1). Exogenous sources include chemical compounds, ionizing radiation (IR), ultraviolet (UV), chemotherapeutic-induced DNA damage, tobacco smoke carcinogens, and alcohol, among many others, whereas endogenous sources include reactive oxygen species (ROS), spontaneous deamination, DNA methylation, apurinic or apyrimidinic (AP) sites, and active enzymatic DNA cleavage.¹ Error-free replication and transmission of DNA are needed to maintain genomic stability, which translates the necessary information for life.^2,3 Cell death may occur due to an imbalance in DNA integrity, leading to premature aging and cancer. However, all organisms have a DNA damage repair mechanism that correctly senses and repairs DNA lesions and move their genome to the next generation with high fidelity.^2,4 The mammalian genome on average gets approximately 10⁵ spontaneous lesions per day,⁵ most of which are removed by DNA repair mechanisms; however, when the DNA template is missing and the DNA is unwound at replication forks, DNA lesions persist through S-phase. To avoid replication fork collapse and DNA breaks, cells undergo a lesion bypass process via DNA damage tolerance (DDT) pathways and restart the DNA replication process (Fig. 1). DNA damage tolerance can occur mainly through error-free template switching (TS) and error-prone translesion synthesis (TLS).⁶ Template switching utilizes the newly synthesized nascent sister chromatid as a template for lesion bypass, while TLS employs specialized DNA polymerases to replicate across the otherwise replication-blocking lesion. These pathways are tightly regulated to avoid more severe consequences arising from stalled replication forks.⁷

FIG 1.

The endogenous, exogenous, and spontaneous decay-induced DNA damage.

Replicative DNA polymerases with stringent requirements can replicate the entire genome with high fidelity but cannot use a template with replication-blocking lesions. When encountered with DNA damage, the replication fork stalls, resulting in long stretches of single-stranded DNA (ssDNA) attached to the replication protein A (RPA) [8]. This leads to the initiation of Ataxia telangiectasia mutated (ATM) and Rad3-related (ATR) checkpoint and initiation of lesion bypass by TLS through monoubiquitinated proliferating cell nuclear antigen (PCNA).^8–11 Translesion DNA synthesis uses specialized DNA polymerases to perform lesion bypass using damaged DNA as a template for nucleotide insertion opposite to lesions. Generally, TLS polymerases do not have 3′–5′ exonucleolytic proofreading activity, which helps TLS to avoid additional kinetic barriers. Moreover, they have a more significant active site that helps accommodate bulky DNA adducts, facilitating replication by directly bypassing the lesion.¹²

To date, it is reported that multiple processes such as transcription, post-translational modifications (PTMs), protein-protein interactions (PPIs), and noncoding RNAs tightly control the critical events of TLS in mammalian cells. Furthermore, this involves polymerase switching and PCNA monoubiquitination (PCNA-mUb). Such tight regulation not only ensures the recruitment of lesion-appropriate TLS polymerase to the damaged site to salvage replication, but also blocks access of these mutagenic DNA polymerases to the undamaged DNA. In addition, TLS is involved in the initiation, progression as well as chemoresistance of many cancers, signifying its importance as a potential target for cancer treatment.

This review summarizes the type of TLS polymerases and their structural properties, TLS polymerase-induced DNA damage, and DNA damage tolerance pathways, which will facilitate the development of potential TLS inhibitors for sensitizing tumor cells to chemotherapy. We will also discuss the noncanonical functions of TLS polymerases and TLS polymerase-associated human diseases.

TYPE OF TLS POLYMERASES AND THEIR STRUCTURAL PROPERTIES

TLS polymerases are a particular class of DNA polymerases capable of utilizing damaged DNA as a template for DNA synthesis.^13,14 TLS polymerases are found across all forms of life and consist of the following: Dpo4 (dpo4) and Dbh (dbh) in archaea, Pol II (polB), Pol IV (dinB), and Pol V (umuDC) in prokaryotes, and Rev1 (REV1), Pol κ (POLK/DINB1), Pol η (RAD30/POLH), Pol ζ [Rev3 (REV3L) and Rev7 (REV7/MAD2BL2) subunits], Pol ι (POLI/RAD30B), Pol λ (POLL), Pol β (POLB; not a classical TLS polymerase), Pol θ (POLQ), and the newly identified primase and polymerase enzyme PrimPol (PRIMPOL/CCDC111) in eukaryotes [15–17]. Depending on the protein sequence, eukaryotic domain organization of human TLS polymerases is shown in Fig. 2. TLS polymerases are classified into A, B, X, Y, and PrimPol families.^15–17 The majority of the TLS polymerases, including Dpo4, Dbh, Pol IV, Pol V, Rev1, Pol κ, Pol η, and Polι, belong to the Y family DNA polymerases.¹⁸ The eukaryotic Pol ζ and prokaryotic Pol II belong to the B family, which also includes high-fidelity replicative DNA polymerases. Unlike other B family polymerases, Pol ζ is a low-fidelity polymerase capable of carrying out TLS and has a unique property to extend from primer-template pairs that are mismatched and/or deformed, including those that are opposite to DNA lesions.¹⁹ Pol λ and Pol β belong to the X family whereas Pol θ belongs to the A family.

FIG 2.

Human TLS polymerases and their protein domain.

In general, all Y family polymerases feature a predefined active site where deoxyribonucleoside triphosphate (dNTPs) and pyrophosphate may smoothly migrate in and out due to a tiny finger and thumb domain.²⁰ In Pol κ the fifth domain, known as N-clasp is located at the N-terminus.²¹ Through interaction with the catalytic core, the little finger, and the DNA duplex region, the Pol κ N-clasp engages with DNA and stabilizes the ternary complex polymerase coupled with DNA and dNTP.²² Despite having 870 amino acid residues, human Pol κ utilizes a primer extension strategy by truncating proteins that keep residues 19–526 active enough to maintain wild-type-like polymerization.²¹ Until recently, it was difficult to determine the nature of interactions between N-clasp residues and other components of the complex from Pol κ crystal structures; however, a 2.0-Å resolution crystal structure demonstrated structural traits and interactions that were essential for our understanding of this function.²³ In a recent study, the reconstruction of structural details of Pol κ in interaction with PCNA and DNA using cryo-electron microscopy (Cryo-EM) mapping revealed the flexible nature of Pol κ C-terminal region (aa residues 535–870), as this region remained invisible in the Cryo-EM map.^24,25 Likewise, the N-clasp residues 21–45 also appeared invisible on the map, suggesting a partly flexible nature of the N-clasp as well.²⁵ Besides this, the Cryo-EM mapping also revealed that the clash between the inner rim of the PCNA and the DNA duplex emerging from the catalytic core of Pol κ is avoided by bending the DNA by ∼30°.²⁵ Similar to Pol κ, the three dimensional structure of Rev1 contains a huge gap that separates the little finger (LF) domain from the ring-shaped catalytic core that surrounds the DNA.²⁶ Rev1 structurally has a detachable LF and an N-terminal extension known as the N-digit that connects the LF to the catalytic core.^24,25 It has been proposed that two substantial inserts in the human Rev1's finger and palm domains, which are specific to its catalytic core, have a significant role in conferring Rev1 the unique property of carrying out protein-protein interaction during TLS.²⁷

Human Pol ι is an 80-kDa protein with 715 amino acids (GenBank Accession number AF140501.1); however, some isoforms have 740 amino acids, and the commonly found isoform has only 415 amino acids.²⁸ The POLI gene, which is found on human chromosome 18q21.2, has several organismal homologs and is evolutionarily conserved.²⁹ Pol ι structurally resembles the highly conserved Y family polymerase right-hand configuration, with fingers (aa 38–98), thumb (aa 225–288), and palm (aa 25–37 and 99–224) forming the N-terminally located catalytic active site.³⁰ The crystal structure of Pol ι in a ternary complex with DNA substrates and an incoming nucleotide provides important new information about the unique base selectivity, which is characterized by a 10⁵-fold change in fidelity depending on the template base.^31,32 For the catalytic activity, all DNA polymerases require a divalent cation. Mg²⁺ is often thought to be the physiological cofactor for replicative DNA polymerases in vivo.³³ The catalytic affinity of pol ι for the metal ions of Mn²⁺ was higher than that of Mg²⁺ during TLS.³⁴ Also, Mn²+ achieves more optimal octahedral coordination geometry than Mg²⁺, with lower values in average coordination distance geometry within the catalytic metal A-site, and crystal structures of Pol ι ternary complex include the primer terminus 3-OH and a nonhydrolyzable dCTP analog opposite G.³⁴ Pol ι appears to contain an intrinsic 5’-deoxyribose phosphate (dRP) lyase activity;³⁵ however, dRP lyase of Pol ι does not shield cells Pol β lacking dRP lyase activity from methylation-induced cytotoxicity, consistent with the exclusion of pol ι from BER of certain lesions.³⁶

Human Pol η is encoded by POLH located on chromosome 6p21.1.^37–39 It contains 713 amino acids and plays a vital role in bypassing thymine-thymine (TT) cyclobutane pyrimidine dimers (CPDs) caused by UV irradiation in an error-free manner. It does so by inserting correct nucleotides across the lesion.^38,40 While the carboxy-terminus of Pol η is made up of two PCNA interaction protein (PIP) domains and a ubiquitin-binding zinc finger domain (UBZ), the N-terminus of Pol η contains protein-arginine deiminase (PAD) and catalytic domains (Fig. 2).^38,40 Because the Pol η active site is bigger than that of other Y-family polymerases, it can accommodate two nucleotides of the CPDs and allow Pol η to synthesize DNA over large adducts.²⁰ This is obvious from the structure of yeast Pol η, which was simulated with a cis-syn TT dimer in the template DNA and incoming deoxyadenosine triphosphate (dATP) (Table 1).^41,42 Pol η often causes both clustered and unclustered mutations at A:T nucleotide pairings in numerous cancer types.⁴³ Some carcinogens promote the activity of error-prone noncanonical mismatch repair (ncMMR) pathway involving Pol η, leading to an increased relative mutation rate in the H3K36me3-marked region.^44,45 ncMMR is mainly independent of DNA replication, lacks strand directionality, induces PCNA monoubiquitylation, and increases the recruitment of the error-prone polymerase to chromatin,^44,45 implying that some environmental factors cause cancer by redistributing mutations to targeted genome regions rather than raising the overall mutation rate.

TABLE 1.

Types of human TLS polymerases, their clinical role, and bypass lesions

TLS polymerase	Organism; gene name	Mechanisms	Clinical role/preclinical roles	Bypassed lesion
Pol η	Human DNA polymerase; POLH gene	Nucleotide insertion and extension (extends up to 2 nucleotides beyond the lesion) is carried out by the enlarged active site of the Pol η polymerase catalytic domain (residues 1–432) [20]. The Pol η’s wide active site stably accommodates the two covalently bonded thymines (T-T dimmer). Unlike other Y family polymerases, Pol η operates by stabilizing the DNA's B-form conformation through a strong electrostatic interaction with four consecutive DNA template phosphates [20, 51]. In the TT bypass, Pol η functions with a biased fidelity, bypassing 5’ T with a low error rate than the 3’ T [52].	Humans who lack active Pol η due to mutation(s) suffer a type of UV-sensitive condition known as Xeroderma pigmentosum variant (XPV), which increases the risk for skin cancer [53, 54]. Pol η expression levels are negatively associated with the survival rate of people with non-small-cell lung cancer, metastatic gastric adenocarcinoma, and head and neck squamous cell cancer [55].	Cisplatin-induced guanine-guanine intra-strand adducts, 8-ooGuanine Abasic sites, UV-induced lesions, particularly T-T cyclobutane pyrimidine dimers, N-2-acetylaminofluorene (AAF)-modified guanine [35, 120, 121].
Pol ι	Human DNA Polymerase; POLI gene	Pol ι can efficiently incorporate deoxythymidine triphosphate (dTTP) opposite dA but due to its unusual active site, it promotes Hoogsteen pairing instead (3–10 fold increased dG-dT mispairing) [12, 56]. It may be that Pol ι can avoid N1-methylated deoxyadenosine by using the dA(syn): dTTP (anti) combination [57].	The human cell cycle constrained by UV-induced DNA damage is improved by elevated levels of pol ι triggered by p53 [58].	5-hydroxyuracil (5-OHU), N2-guanine adduct, 5-hydroxycytosine (5-OHC), thymine-thymine 6–4 Photoproducts, 5,6-dihydrouracil (5,6-DHU), 8-oxoGuanine [127, 128].
Pol κ	Human DNA polymerase; POLK gene	Polycyclic aromatic hydrocarbon adducts, such as benzo[a]pyrene diol epoxide (BPDE) covalently bound to N2 of guanine in the minor groove, are specifically avoided by Pol κ [59].	Pol κ deficient mice have a spontaneous mutator phenotype, and the mutation rate rises as dietary cholesterol rises, suggesting that Pol κ can avoid naturally occurring steroid adducts [60, 61].	Benzo[a]pyrene-guanine adducts (BP-G), Thymine glycol,8-oxo-d-guanine [14, 116–119].
Rev1	Human DNA polymerase; REV1 gene	When the template is abasic, deoxycytidylic (dCMP) transferase activity of Rev1 is highest, followed by dG and dA [50], and later, it was confirmed that Rev1 bypasses abasic lesions in vivo [62].	In somatic hypermutation, Rev1 adds deoxycytidine residues opposite the abasic sites [52]. In human colon carcinoma cells, loss of p53 increases Rev1 expression [58].	8-oxoguanine (8-oxoG), 1, N 6-ethenoadenine adducts, UV-induced lesions, Trans-anti-benzo[a]pyrene-N 2-dG [123, 124].
Pol ζ	Human DNA Polymerase; REV3L and REV7 genes	Pol ζ was thought to integrate mismatched dNTPs and circumvent various DNA lesions, but later on, its main role was identified in the primer extension step of TLS [10, 63, 64]. Rev7 can cause biallelic mutation in the TLS and the direct binding of CHAMP1 to REV7 lowers the concentration of the Shieldin complex, increasing double-strand break end excision (PMID: 36044844).	Increased DNA damage tolerance and decreased spontaneous tumorigenesis are two ways that Pol ζ improves genomic stability [64]. Impairment of Rev1 precursors has been found to cause bone marrow failure in Fanconi anemia (PMID: 27500492).	cyclobutane pyrimidine dimers, Extender polymerase for numerous lesions, thymine-thymine 6–4Photoproducts [125, 126].
Pol θ	Human DNA Polymerase; POLQ gene	Pol θ has a role in DSB repair by error-prone end-joining, referred to as alt-EJ or MMEJ [65]. Additionally, Pol θ can expand from mismatched bases opposite large lesions such as 6–4 photoproducts and thymine glycol [66, 67].	Pol θ is overexpressed in various types of cancers [65] which is related to the lymphoid tissue and mainly found associated with lung, stomach, and colon cancer [58–60]. Additionally, in women with an unknown breast cancer gene (BRCA) 1/2 profile, a mutation in the promoter region of POLQ (c.-1060A > G) was found strongly linked to both hereditary breast and ovarian cancer [68].	Thymine glycols, UV-induced lesions, 1, N 6-ethenoadenine adducts [116, 122].
Pol λ	Human DNA polymerase; POLL gene	PCNA and Pol λ engage in physical and functional interaction without changing the rate of nucleotide incorporation. This interaction stabilizes the binding of Pol λ to the primer template and improves the processivity of DNA synthesis [69–71].	The R438W hPol λ mutation causes chromosomal aberrations, impairs NHEJ-mediated DSB repair, and increases mutation frequency when expressed ectopically in mammalian cells, a potential cause for cancer due to chromosome level genomic instability [72].	1,2-dihydro-2-oxoadenine (2-OH-A) (PMID: 17666409).
PrimPol	Human DNA Polymerase; PRIMPOL gene	PrimPol can avoid templated oxidative lesions such as 8-oxo-guanine (8-oxoG), abasic sites, cyclobutane pyrimidine dimer (CPD), and pyrimidine pyrimidone (6–4) photoproducts (6–4 PPs) [73, 74].	Human cells are vulnerable to UVC-induced cell death only in the absence of functional Pol η, as in XPV patients [75].	Apyrimidinic/apurinic site [67].

Pol ζ has a pentameric ring-like design with a catalytic subunit Rev3, two noncatalytic Rev7 subunits, and auxiliary Pol31 and Pol32 subunits, forming a continuous daisy chain of protein-protein interactions [38,69]. REV7 was first recognized as a gene responsible for UV-induced mutagenesis in Saccharomyces cerevisiae⁴⁶ and later findings referred to the Rev3-Rev7 complex as DNA polymerase ζ since it is the sixth eukaryotic DNA polymerase (Table 1).¹⁹ The human mitotic checkpoint protein hMAD2 shares 23% identity and 53% similarity with hRev7, which also shares 23% identity and 54% similarity with ScRev7. Although no deleterious nucleotide alteration(s) have been detected in hREV3 or hREV7 genes in primary human tumors or human tumor cell lines, hREV7 is mapped to chromosome 1p36 in a location that often suffers from loss of heterozygosity in human cancers and hREV3 is located in the common fragile site, a hotspot of chromosomal rearrangements.^19,47–49 The major role of Rev7 is to connect between Rev3 and Rev1, while a new structural study has shown possible interactions between Rev7 and Pol32 (POLD3 in mammals), a regulatory component shared by Pol ζ and the replicative polymerase δ.^49,50

PrimPol is emerging as an important factor in the tolerance to DNA damage, notably in vertebrate and human cells. PrimPol is a 560 amino acid long protein in humans and is found in vertebrates, plants, and lower eukaryotes with a varied number of amino acid residues.^51–53 The PrimPol nucleotidyltransferase activity, necessary for its primase and polymerase activities, is regulated by two highly conserved modules, ModN (residues 35–105) and ModC (residues 108–200 and 261–348) in the catalytic archaeo-eukaryotic primase (AEP) core.⁵⁴ Changes to the catalytic residues Asp114, Glu116, or Asp280 of ModC abolish both primase and polymerase activities of PrimPol.⁵⁴ The DNA binding affinity of ModN is diminished when critical residues are modified, resulting in a decreased enzymatic activity.^55,56 A zinc finger (ZnF) domain, an RPA binding domain, and an N-terminal helix that stabilizes PrimPol’s interaction with ssDNA are all present in PrimPol in addition to the AEP core.^54,55 PrimPol has a low processivity while synthesizing DNA, infrequently adding more than four nucleotides to an intact template. The ZnF domain of PrimPol, which is linked to the catalytic core by a 140-amino-acid flexible linker and regulates the amount of nucleotide incorporation, is one of the mechanisms responsible for this poor processivity.^56,57

TLS POLYMERASE-INDUCED MUTAGENESIS

Correct replication by DNA polymerases is a critical step for normal biological processes and the maintenance of genetic integrity. Various human diseases have been associated with mutations in genes involved in the DDT pathway, and importantly the TLS polymerase genes. For instance, mutations in POLB result in lymphomas in mouse and human breast, lung, colorectal, and prostate carcinomas.⁵⁸ Likewise, mutations in POLG cause progressive external ophthalmoplegia.⁵⁹ POLD1 and POLE encode human DNA polymerases δ and ε, respectively, and their mutations cause hereditary colorectal cancer (doi: 10.1186/s13046-022-02422-1). Xeroderma pigmentosum variant (XPV), a well-characterized syndrome associated with a defective TLS pathway, is caused by a mutation in POLH.^60,61 As evident from several studies,^4,5,12,16 the importance of DNA polymerases in maintaining genomic stability and, therefore, human health is beyond doubt.

Numerous studies have suggested that TLS polymerases exhibit decreased replication fidelity due to their lack of 3'>5' proofreading activity and the increased use of altered DNA as templates. Compared to TLS polymerases, replicative DNA polymerases exhibit error rates in the range of one nucleotide for every 10⁶ to 10⁸ replicated bases.⁶² In contrast, TLS polymerases when replicating undamaged DNA exhibit error rates in the range of one base for every 10–10,000 bases.^41,62–65 Moreover, unlike replicative polymerases, TLS polymerases do not proceed with an induced fit model upon nucleotide binding,^65,66 allowing bypass of replication-blocking lesions. Furthermore, some TLS polymerases, such as Rev1 and DNA Pol ι, do not use the Watson–Crick canonical base pairing.^42,67 Hence, TLS polymerase activities are largely mutagenic.

TLS POLYMERASES AND DNA DAMAGE TOLERANCE

DNA regularly comes into contact with endogenous and exogenous substances that cause DNA damage. The presence of unpaired DNA lesions in the template can arrest the replication fork movement that leads to genomic instability via fork collapse and DSB formation. To complete the replication and avoid strand breaks, DDT plays an essential role in allowing replication in the presence of template lesions.⁶⁸ There are two main DDT pathways: (i) TLS, which involves the recruitment of TLS DNA polymerase to bypass the lesion, which can occur either directly at the site of replication arrest or behind the arrest by repriming DNA synthesis at the daughter strand gaps (DSGs) (Fig. 3); and (ii) homology-directed damage tolerance, which is accomplished by fork reversal (FR) and/or TS (Fig. 3).⁶⁹ Monoubiquitination of the clamp protein PCNA in response to DNA damage results in the recruitment of TLS polymerases, which are capable of bypassing the lesion. At the site of an arrested fork, lesion bypass takes place directly or during gap filling following replication restart away from the lesion site.^26,70 However, despite certain exceptions mentioned below, TLS polymerase-mediated lesion bypass tends to misincorporate bases resulting in increased mutations.^15,70 The error-prone characteristics of TLS polymerases have been implicated in both carcinogenesis and resistance to chemotherapy.^16,71,72 K63-linked polyubiquitination of monoubiquitinated PCNA initiates FR and/or TS to facilitate a temporary switch from the lesion-containing strand to a newly synthesized complementary sister chromatid strand of the homologous chromosome. Because the undamaged template is being copied, FR and TS are considered error-free lesion bypass pathways [98]. A structure resembling a chicken foot is created by fork reversal, allowing the replisome on the halted nascent strand to reach the homologous sister template.^73,74 Different from FR, TS arises after repriming at DSGs generated at lesion locations behind the replication fork.^75,76 Template switch is characterized by strand invasion, in which the newly synthesized strand serves as a template for nascent strand synthesis, which avoids the lesion from replication machinery.^77,78

FIG 3.

Schematic diagram representing main DNA damage tolerance pathways. Yellow color indicates the TLS pathways; Teal color indicates the homology-based damage tolerance in eukaryotic cells.

Mechanism of DNA damage tolerance initiation

The TS and TLS pathways share the same “substrate” at the DNA lesion site; stalling of replicative DNA polymerase along with the ongoing helicase activity at the replication fork results in the formation of ssDNA on the template strand to which the RPA binds. The ssDNA-RPA complex brings in the ATR interacting protein (ATRIP), which activates the ATR-dependent replication checkpoint.^77,79 Simultaneously, the INO80 ATPase, a chromatin-remodeling protein, attaches to the stalled replication fork.^80–82 Moreover, Rad6 is recruited by Rad18, resulting in the E2-E3 ubiquitinase complex formation at the lesion site that monoubiquitinates PCNA at K164.^67,83–86 Apart from this, monoubiquitination may take place through different E3 ligases, such as ring finger protein domain 8 (RNF8) and ubiquitin-conjugating enzyme H5c (UbcH5c).⁸⁷ At this point, the two DDT pathways separate, one with monoubiquitinated PCNA leading to TLS induction, and the second to further polyubiquitinate PCNA, resulting in homology-directed DDT.

Translesion synthesis

Subsequent to PCNA-K164 monoubiquitination and dissociation of the replicative polymerases, one or more TLS polymerases get recruited depending on the nature of the replication-blocking lesion (Fig. 4). For Y-family TLS polymerases, the conserved N-terminal domain consists of an active site that catalyzes lesion bypass, whereases the variable C-terminal region helps in the enrolment of protein to stalled forks.^88,89 They can attach directly to K164-ubiquitinated PCNA by ubiquitin-binding motifs (UBM) found in Pol ι and Rev1 or UBZ present in Pol η and κ.⁸⁹ TLS polymerases binding to PCNA is also facilitated through the BRCA1C-terminus (BRCT) domain located in the N-terminus of Rev1 and PIP boxes on Pol ι, η, and κ.^50,89,90 The ubiquitin-binding domain (UBD) and PIP box mutations impair the damage-induced interaction of TLS polymerase with monoubiquitinated PCNA and their recruitment at the replication forks site [110, 111]. Generally, translesion synthesis is accomplished by the sequential action of two TLS polymerases. A distributive Y-family TLS polymerase inserts nucleotide(s) across the lesion site and then another TLS polymerase continues to extend the replicated DNA strand before it is replaced by replicative polymerase.^88,90,91 In S. cerevisiae, during the REV1 mediated bypass, the lesion on the template strand is flipped into an extra-helical position and gets stabilized inside a hydrophobic pocket of Rev1, where it stays during the integration of incoming cytosine.⁹² Rev1 incorporates a single dCTP opposite to the lesion site.^51,54,55 The R324 residue in the Rev1 side chain shifts the DNA lesion and acts as a substitutive template in Watson–Crick base pairing for incoming cytosine.⁹² Hydrogen bonding between R324 and cytosine is broken after the coupling of hydrolysis of pyrophosphate and phosphodiester bond. Rev1 is then separated from the DNA, and the lesion is reinserted in a double helix.⁹² Y family DNA polymerases lack 3′≥5′ proofreading exonuclease activity and have open active sites that can accommodate the altered bases.^15,20,93 For example, in UV-induced cis-syn thymine-thymine CPD lesion bypass, the two covalently connected thymine bases can fit in Pol η active site.^20,93 The protein domain in the β-strand in the LF provides a molecular splint that leads to the stabilization of newly synthesized dsDNA into a structure of B-form. It prevents CPD-induced framework shift and duplex distortion, which helps efficient and correct Pol η associated bypass of thymine-thymine CPDs.^20,88,93

FIG 4.

Schematic diagram representing translesion DNA synthesis.

The accuracy of TLS polymerase in lesion bypass depends on the lesion type, and bypassing it seems to be at the cost of replication fidelity.⁹⁴ For instance, Pol θ often inserts the correct base during replication across a 1,N⁶-ethenodeoxyadenosine lesion in human cells.⁹⁵ In contrast, Pol θ also plays an essential role in the error-prone bypass of PP lesions and UV-induced cis-syn CPDs.^26,95 The overall accuracy with which a TLS polymerase bypasses the lesion depends on many factors, such as the polymerases-lesion affinity, the biochemical properties of the individual TLS polymerases, and the sequence context of the lesion^14,15,96 (Table 1).

Homology-directed DDT

As mentioned above, TS and FR are the two homology-directed DDT pathways that bypass the lesion in an error-free manner during S-phase (Fig. 3). The choice of error-free DDT pathways during lesion bypass is initiated by the polyubiquitination of the previously Rad6-Rad18-mediated monoubiquitinated PCNA, a process required in the signaling of TLS-mediated lesion bypass.⁹⁷ In yeast, the polyubiquitination step requires the formation of a Mms2-Ubc13-Rad5 complex, whereas, in mammals, the polyubiquitination step is carried out by the complex of Ubc13-MMS2 and one of the two mammalian Rad5 homologs – SNF2 histone linker PHD RING helicase (SHPRH) and helicase-like transcription factor (HLTF).^97,98 The direct binding of helicase protein SMARCAL1 to ssDNA removes the bound RPA, resulting in the remodeling of stalled replication fork to a “chicken-foot” structure in the FR pathway.^99,100 Post RPA removal, translocase zinc-finger RANBP2-type containing 3 (ZRANB3) initiates further FR.^69,101,102 The RAD51, BRCA1, and BRCA2 proteins help in the stabilization of reversed fork through which binding to the lagging and leading strand prevents the exonucleolytic degradation mediated by MRE11.^103–105 Protein-DNA complex and the Fanconi anemia complementation group M (FANCM) helicase, when combined, lead to the formation of a four-way junction.^106,107 After the successful lesion bypass, the original three-way junction formed by the regression of the reversed fork forms a four-way junction. This is catalyzed by RECQ-like helicase (WRN), DNA replication helicase/nuclease 2 (DNA2), and RecQ-like helicase (RECQ1).^69,108,109

In the TS pathway, recruitment of exonuclease-1 (Exo1) is facilitated through the attachment of the 9-1-1 complex to the 5’ end of the gap on the nascent DNA strand.^7,68 The Rad51-ssDNA presynaptic filament formed above the ssDNA portion of the template strand gets stabilized by Rad55/Rad57.^110,111 The nucleofilament is disrupted by ATP-dependent DNA helicase Srs2 and restricts the action of Rad55/Rad57; the stability of Rad51-ssDNA presynaptic filament is determined by the balance between these processes.^75,112,113 In sister chromatids, the Rad52 and Rad54 nucleofilament induce homology search and strand invasion.^75,114 Following complementary base-pairing between the homologous template and the invading strand, DNA Pol δ gets recruited, which continues DNA replication and synthesizes the D-loop and later a sister-chromatid junction (SCJ).^68,114,115 The Srs2 negatively correlates with the D-loop formation.^78,116 Lastly, DNA topoisomerase 3 or RecQ-mediated genome instability protein 1 (Rmi1) or the slow growth suppressor 1 complex separates SCJ and regenerates the regular double-helical DNA structure.^68,117,118

Regulation of DDT pathway choice

The selection of TLS or homology-directed DDT pathways is highly dependent on the type of PCNA ubiquitination.^15,69,119 TLS pathway takes place when PCNA monoubiquitination occurs; however, polyubiquitination of PCNA leads to the initiation of homology-directed DDT. Regulation of ubiquitinated PCNA level is achieved via the enhanced level of genomic instability 1 (Elg1), ubiquitin-specific protease 7 (USP7), and USP1/upstream activation factor (UAF1) complex.^119–122 After UV irradiation, USP1 proceeds toward inactivation through autocleavage, thereby upregulating the level of modified PCNA.¹²³ It has been proposed that the type of PCNA ubiquitination depends on how long the replication arrest stays. Therefore, in promoting a switch to homology-directed DDT, extended replication occurs, resulting in the polyubiquitination of PCNA molecules that remain bound at the arrest site.¹¹⁹ The homology-directed DDT in mammals is first activated by the recruitment of HLTF along with the RAD6/RAD18 complex, leading to the instant polyubiquitination of PCNA.¹¹⁹ Apart from ubiquitination, PCNA experiences other related modifications. A previous study suggests that SUMOylation of PCNA at K164 mediated by protein inhibitors of STAT 1 and 4 (PIAS1 and PIAS4) encourages TS instead of TLS.¹²⁴ When the TLS process completes, modification happens in monoubiquitinated PCNA via the accumulation of interferon-stimulated gene 15 (ISG15) molecules, resulting in USP10 recruitment and PCNA deubiquitination.^125,126 Hence, a deep understanding of the connection between the choice of DNA damage tolerance and the PCNA modification pathway is a vital area for future study. Yeast RAD5 also interacts with REV1 to be involved in TLS and controls pathway choice.¹²⁷

Regulation of TLS

Interaction between a RAD18 E3 ligase and PCNA accessory proteins with TLS polymerases maintains the regulation of TLS.⁹ However, there is some uncertainty about the role of terminal nucleotidyltransferase 4 A (TENT4A; formerly referred to as PAPD7), a noncanonical poly(A) polymerase in the TLS regulation. It was recently identified as another potential TLS regulator via its action on mRNA stability by controlling the poly(A) tail and/or Pol η and RAD18 translation.¹²⁸ As discussed above, PCNA monoubiquitination by the RAD6/RAD18 complex is the key signal that allows the recruitment of TLS polymerases to the site of DNA damage. In addition, indirect regulation of RAD18 by TENT4A was also reported in the same study involving tumor suppressor gene CYLD and the long non-coding antisense RNA PAXIP1-AS2, suggesting the involvement of CYLD and PAXIP1-AS2 in the TLS regulation as well.¹²⁸ In prokaryotes, the access of TLS polymerase Pol IV to the DNA was recently shown to be regulated by MutS, a mismatch repair protein that blocks the interaction of Pol IV with the clamp processivity factor, thus limiting the mutagenic impact of Pol IV during replication under normal conditions.¹²⁹ Other factors involved in TLS regulation include HLTF, SHPRH, TIMELESS, SIVA1, CLASPIN, CHK1, and SprT-like domain at the N-terminus (SPARTAN), all having a crucial role in the monoubiquitination of PCNA.^130–133 The differential role of HLTF and SHPRH in the recruitment of lesion-specific TLS polymerase Pol η (UV induced) and Pol κ (MMS induced), respectively¹³² highlights the complex yet essential facets of TLS regulation.

TLS polymerases regulate TLS pathways individually by undergoing post-translational modifications like SUMOylation, ubiquitination, and phosphorylation. Apart from ubiquitination, Pol η is also regulated through SUMOylation.¹³⁴ Protein inhibitors of STAT1-mediated SUMOylation at K163 targets Pol η to the difficult-to-replicate genomic regions, including fragile sites in the absence of any exogenous DNA damage.¹³⁵ After the lesion bypass is over, multiple lysine residues of Pol η are SUMOylated, thereby preventing it from interacting with ubiquitinated PCNA, resulting in SUMO-targeted ubiquitin ligase (STUbL) induced Pol η ubiquitination and its elimination from damage sites.¹³⁵

Pol η is phosphorylated in the C-terminus involving multiple sites by protein kinases CDK2, PKC, and ATR. After DNA damage, ATR phosphorylates Pol η on S601, leaving it from sequestration by Pol δ-interacting protein of 38 kDa (PDIP38) and allowing Pol η to bind to the monoubiquitinated PCNA.^131,136 This correlates with ATR activation via replication arrest-induced ssDNA, leading to TLS polymerase enrolment at the arrested replication fork.¹³⁶ Furthermore, Pol η gets phosphorylated by PKC on S587 and T617, and by CDK2 on S687, leading to its stabilization in late S- and G2/M-phases.^137,138

Regulation of TLS can also occur at the transcriptional level. After DNA damage, the expression of Pol η relies on p53,¹³⁹ whereas the expression of Pol κ is governed by an aryl hydrocarbon receptor (AhR).¹⁴⁰ A recent study reported that the function of Pumilio 1 (PUM1), an RNA-binding protein, is negatively correlated with TLS, as activation of TLS in response to DNA damage depends on the inhibition of PUM1-mediated mRNA decay.¹⁴¹

Regulation of homology-directed DDT

During fork reversal, the interplay between fork-protective and fork-degradative factors significantly contributes to the regulation of homology-directed repair.¹⁰⁶ The MRE11 exonuclease degrades the nascent DNA strand at the reversed fork, which is protected by RAD51, BRCA1, and BRCA2.^69,103,104 Subsequently, the DNA2-mediated fork degradation and structure-specific endonuclease subunit (SLX4)-facilitated fork cleavage is prevented by WRN helicase interacting protein 1 (WRNIP1).^142,143 Progression of the fork slows down due to the interaction of ZRANB3 with polyubiquitinated PCNA, resulting in fork reversal by the ZRANB3 translocase activity.^101,144 Furthermore, properly regulated DNA translocase SMARCAL1 is necessary to facilitate stalled fork repair and restart, and to ensure no aberrant fork progression takes place.¹⁴⁴ This is promoted by the ATR-mediated phosphorylation of SMARCAL1 on S652, thereby restricting its fork regression activity and controlling its recruitment to the stalled fork site.^142,145 Fork reversal and fork restart are modified by poly(ADP-ribose) polymerase 1 (PARP1) through the inhibition of RECQ1 helicase; further, it extends RECQ1-facilitated reversed forks to the three-way junction regression.^74,108,142 Template switching is regulated at multiple points, including PCNA polyubiquitination, the formation of RAD51-ssDNA presynaptic filament, and SCJ formation.⁶⁷ In human cells, SHPRH, HLTF, and INO80 chromatin remodeling proteins are essential for PCNA polyubiquitination.¹⁴⁶ In the remodeling of chromatin, INO80 assists the assembly of K63-linked polyubiquitin chains to PCNA through SHRPH and HLTF.^{69,82,147,148} The Rad51-ssDNA presynaptic filament is disrupted by Srs2 and later participates in homology search;^68,149–151 however, high mobility group protein 1 (HMO1), (EXO1, and INO80 promote the SCJ formation.^{68,82,151,152}

NONCANONICAL FUNCTIONS OF TLS POLYMERASE

As discussed above, each TLS polymerase has certain known functions as part of recognized mechanisms. However, new functions and pathways mediated by TLS polymerases have emerged recently, linking these highly specialized polymerases with numerous cellular functions, including somatic hypermutation (SHM), DNA repair, and maintenance of genomic integrity in the absence of exogenous lesions, spindle checkpoint, cell cycle regulation, and fragile site maintenance.

SHM is a mutagenesis process that takes place in secondary lymphoid organs and produces B cells with the greatest affinity receptors for a particular antigen. This process has been linked to several TLS polymerases with Pol η and REV1 playing a major role.¹⁵³ The frequency of mutations during SHM is a million times higher than the rest of the genome in the hypermutable region of the immunoglobulins around 2 kB in size.¹⁵⁴ These mutations are mostly base substitutions that contribute to the diversity of antibodies. A well-defined SHM spectrum of C/G transversions, C/G transitions, and A/T mutation around the initial lesion has been reported by five mutagenic DNA damage response mechanisms, starting with the U as the original lesion: (1) replication opposite template U generates transitions at C/G; (2) UNG2-dependent TLS; (3) a hybrid pathway consisting of UNG2-dependent TLS and ncMMR; (4) ncMMR; and (5) PCNA ubiquitination and UNG2-dependent mutations at A/T [194]. Other TLS polymerases, such as Pol ζ [195,196], Pol θ,^155–158 Pol ι,^159,160 and Pol κ,¹⁵⁹ have also been linked to SHM. However, it appears that these polymerases play a more supporting role. For instance, it was reported that Pol κ only contributed to the production of mutations when Pol η was completely absent.¹⁵⁹

As Pol ζ can replicate beyond non-B DNA types that prevent the replicative polymerases from working, it also plays a crucial role in the genome maintenance. In mice, the absence of Pol ζ catalytic subunit Rev3L is embryonically lethal, and primary cells exhibit large-scale genomic instability as a result of an increase in spontaneous chromosomal translocations.¹³ Intriguingly, the codisruption of REV3L and POLH in chicken DT40 cells can correct this REV3L mutant-associated genomic instability and chromosomal aberrations, suggesting that Pol ζ and Pol η work together and that Pol η generates a toxic intermediate that requires Pol ζ to get resolved.¹⁶¹ In general, Rev1 and Pol ζ are both missing the ubiquitous hydrophobic motif that many proteins utilize to link with PCNA at its interdomain connecting loop. Pol ζ and Rev1 are distinctive to PCNA-interacting proteins that use the new binding surface close to the intermolecular interface of PCNA. The reported novel form of Rev1-PCNA binding raises the possibility of a mechanism by which Rev1 acquires a catalytically inactive conformation at the replication fork.¹⁶² Breakage at both inverted repeats and GAA/TTC repeats is exacerbated by DNA replication flaws. Increased fragility is linked to higher mutation levels in reporter genes located up to 8 kB away from both sides of the repeats. The existence of inverted or GAA/TTC repeats, as well as the activity of Pol ζ, influences increased mutagenesis.¹⁶³

Fragile motifs cause it to create extended single-stranded areas in the disrupted chromosome, invasion of the undamaged sister chromatid for repair, and incorrect DNA synthesis using Pol ζ.¹⁶⁴ The TLS-independent role of Rev3 in maintaining common fragile sites (CFSs) and the Rev3-independent, therefore TLS-independent roles of Rev7 in spindle formation, mitotic checkpoint, NER pathway, and many more roles of Pol ζ subunits again highlight the significance of these polymerases in the TLS-independent cellular functions.^165–168

A previous study revealed that Pol η recruits DHX9 helicase to promote replication across guanine quadruplex structures.¹⁶⁹ The DNA replication protein DHX9 was discovered to interact with the WRN helicase and PCNA, promoting the WRN activity in unraveling Okazaki fragment-like nucleic acid structures.^170,171 Through its interaction with monoubiquitinated PCNA, Pol η is brought into DNA replication forks that are halted.⁹ Similarly, when replication stress occurs at CFSs, Pol η makes it easier for CFS loci to replicate, indicating its broader role. When Pol η is absent, CFS DNA sequences that match certain pause sites may result in non-B DNA structures.¹⁷²

The involvement of Pol λ in TLS opposite photoproducts strongly argues that Pol λ would similarly insert a proper nucleotide from where polymerase would continue synthesis in the absence of numerous such additional DNA distorting lesions. Complex formation with Pol λ, such as Pol λ/Pol ζ during the error-free TLS opposite Tg and εdA damage, may allow Pol ζ to act in an error-free manner opposite additional DNA lesions when its structuring role is required.¹⁷³ One study reported that the incorporation of A, G, or C opposite dA in mouse embryonic fibroblasts or human follicular lymphoma (HF) cells is coordinated by Tyr-2387 and Tyr-2391 via a Hoogsteen base pairing mechanism, as opposed to the actual role of Pol θ, which uses an AP-like mode to misincorporate A against εdA. Most significantly, only εdA adopting syn conformation at the Pol θ active site could account for 92% incorporation of T opposite εdA.¹⁷⁴ Besides this, Pol θ-mediated end-joining (TMEJ) has been linked to chromosomal repair. TMEJ is an end-joining pathway uniquely capable of repairing substrates when conventionally established NHEJ is inefficient. The consequences of Pol θ deficiency indicate that Pol θ/TMEJ is necessary for the majority of what was previously characterized as MMEJ or Alt-NHEJ. A deficiency of Pol θ has only a moderate effect on the activity of other repair pathways (NHEJ and HR) in Cas9-induced DSBs.^169–171

UBE2V2, the human homolog of MMS2, was reported to play a modest or redundant role in the Rad18-dependent DDT. However, UBE2V2 has been implicated in the abrogation of UV-induced gene conversion following antisense suppression of the transcript in HFs. To ascertain the underlying mechanism, it was demonstrated that DT40 cells do not exhibit considerable hypersensitivity to DNA damage or the enhanced sister chromatid exchange as seen in vertebrate Rad18 mutants.¹⁷⁵ In addition, despite the crucial role Rad18 plays in DNA damage-induced TLS, neither UBE2V2 nor Rad18 in DT40 cells exhibited abnormally high levels of immunoglobulin gene conversion, and unlike Rev1, Rad18 in DT40 cells does not exhibit a defect in the nontemplated immunoglobulin gene mutation. Together, the data revealed that immunoglobulin diversification in DT40 cells is not dependent on the PCNA ubiquitination signaling.¹⁷⁵ A study has shown that localization of Pol η to the chromatin is reduced in FANCD2- and Rad51-deficient cells and that Pol η knockdown alone leads to increased hydroxyurea (HU) sensitivity in the cells.¹⁷⁶ The DNA damage response processes the “U” lesion in a mutagenic manner, producing a full range of base substitutions that define SHM at and near the initial “U” lesion.

Pol µ lengthens primers with 3'-terminal nucleotides opposite the abasic site. Most notably, this extension occurs through a mechanism of nucleotidyl transferase activity that is independent of the template sequence. This is not attributable to simple terminal nucleotidyl transferase activity, because, under typical conditions, Pol µ cannot add dNTPs to a blunt end duplex oligonucleotide or an oligo(dT)29 primer. Pol µ serves a dual mechanism of DNA-synthesizing enzyme that can function as a traditional DNA polymerase or a noncanonical, template-dependent but sequence-independent nucleotidyl transferase. This is the first time such a DNA-synthesizing enzyme has been described.¹⁷⁷

Although TLS polymerases possess high accuracy for certain lesions, they have poor fidelity on undamaged DNA, suggesting their tight regulation in vivo to avoid mutagenesis. When dysregulated, some TLS polymerases confer a hypermutability state^178–181 and alter replication fork progression.^182–184 The loss of activity of some TLS polymerase may have a stronger effect on multicellular organisms than their upregulation.^38,185,186 The early embryonic lethality in mice due to loss of Pol ζ,^187–189 but not due to deletion of Pol κ or Pol ι highlights the distinctive roles some TLS have acquired which are not attributed to their TLS function. There might be a substantial genetic redundancy between TLS polymerases. For instance, mutant phenotype has been observed only upon simultaneous loss of more than one TLS polymerase. The majority of mutations are harmful, and organisms have a protective mechanism that keeps the mutational rate negligible.¹⁹⁰ Even an increase in genetic discrepancy in the population may act beneficial under harmful or adverse conditions, as it sometimes results in a better variant that can withstand stress conditions.^191,192 Furthermore, in response to stress, mutations that occur by TLS polymerase can serve as an important factor in the evolution of genetic variability. Consistent with this possibility, TLS polymerase induces adaptive mutagenesis in bacteria upon cellular stress.^191,192 Moreover, the mutagenic ability of TLS polymerase has been hampered for SHM in higher eukaryotes, resulting in mutation in variable regions of antibodies synthesized by B cell lymphocytes.¹⁹³ Thus, even though the strong deleterious mutagenic effect of TLS polymerases is rather disadvantageous, it may also provide many advantages to the cells.

TLS POLYMERASES AND HUMAN HEALTH

TLS activity mediated mutagenesis and human diseases

Selection of the wrong error-prone polymerase during lesion bypass or mutation(s) in the TLS polymerase genes that compromises their function often results in mutagenesis, a cause of many human diseases, including cancer. Several studies have shown that the levels of Pol ι are considerably higher in esophageal squamous cell carcinoma (ESCC) patients with lymph node metastasis than those without lymph node metastasis. The Kaplan–Meier method demonstrated a negative relationship between Pol ι expression and patient prognosis. In ESCC tissues, the expression levels of matrix metalloproteinase-2 (MMP-2) and matrix metalloproteinase-9 (MMP-9) were found to be favorably linked with Pol ι expression.¹⁹⁴ Besides ESCC, Pol ι overexpression was correlated with breast cancer cells,¹⁹⁵ and bladder cancer.¹⁹⁶

In a recent study, Fanconi anemia cell lines were shown to upregulate Pol ι and rely on Y-family DNA polymerase for survival.¹⁹⁷ The inactivation/impairment of alternative end joining (alt-EJ) by Pol θ depletion slows the development of pancreatic cancer and increases the longevity of experimental mice, but it does not prevent the formation of pancreatic ductal adenocarcinoma (PDAC), which results in full-blown PDAC with widespread metastases.¹⁹⁸ In the initial S phase, poly(ADP-ribose) polymerase inhibitor (PARPi) causes ssDNA gaps behind replication forks in a PrimPol-dependent manner. Even though cells engage postreplication repair mechanisms to patch the gaps, gap repair cannot be completed in the presence of PARPi until the following S phase, resulting in DSBs in a trans-cell cycle manner. Poly(ADP-ribose) polymerase inhibitor has the unusual capacity to produce DSBs progressively in a trans-cell cycle way.¹⁹⁹ BRCA1/2 loss exacerbates DSB accumulation spanning numerous cell cycles while also disrupting repair, making BRCA1/2-deficient cells particularly vulnerable to PARPi. This can be the potential basis for targeting cancer through PARPi.¹⁹⁹ In primary hematopoietic cells, silencing of REV7 reduced progenitor activity, indicating that DNA repair deficiency is a major concern of bone marrow failure in Fanconi anemia.²⁰⁰ The overexpression of error-prone Pol θ which is involved in the DSB repair causes homologous recombination deficiency tumors.²⁰¹ Pol θ-inserted errors are susceptible to error-prone microhomology-mediated end-joining of the DSB, which explains the genetic profile of BRCA-altered cancers.²⁰² Most cancer types accumulate a unique mutation signature generated predominantly by the error-prone polymerases and other error-prone repair activities²⁰³ and to date more than 30 single-base substitution (SBS) mutation signatures have been found in various cancer types with four SBS signatures (SBS2, 5, 13 and 9) attribute to TLS polymerases.²⁰⁴ Apolipoprotein B MRNA editing catalytic polypeptides (APOBEC) family of proteins deaminate cytosine to uracil and its subsequent removal creates AP sites. The insertion of cytosines across these AP sites by Rev1 results in C > T or C > G mutations which are commonly observed in breast and bladder cancers.^205–207 Pol η-dependent somatic hypermutation (SHM) activity is believed to generate the SBS9 mutation signature as seen in chronic lymphocytic leukemia (CLL) and malignant B-cell lymphomas and supported by an animal study;²⁰⁸ however, direct experimental evidence in support of this hypothesis is still missing.

The identification of cancer/tests antigen (CTA) melanoma antigen A4 (MAGE-A4) as a RAD18 stabilizer which in turn promotes PCNA monoubiquitination that can induce mutagenesis in cancer cells by favoring recruitment of TLS polymerases to the replisome.^209,210 RAD18 is also seen overexpressed in the majority of the cancer types,²⁰⁴ thus hinting towards a link between the TLS pathway and cancer. An increased level of E3 ubiquitin ligase MDM2 seen in many cancer types causes depletion of Pol η, thereby increasing cancer predisposition similar to XPV syndrome.²¹¹ Besides these, another TLS influencer RNF168, an E3 ubiquitin ligase is frequently overexpressed in cancer cells.²¹² Although not experimentally validated, this evidence strongly supports the potential role of the TLS pathway in the initiation and progression of cancer, thus making the TLS polymerase a potential therapeutic target. In addition, by looking at the catastrophic connection between TLS and other DNA repair mechanisms could also help look forward to a therapeutic approach.

Apart from their role in cancer, TLS polymerases also have an extended role(s) in other comorbidities, such as the preservation of stem cells, and aging.²¹³ It was found that Pol κ deficiency increases spontaneous mutations during aging in an organ-specific manner, whereas Pol η deficiency causes UV-induced mutagenesis in the skin.^214,215 The premature aging of hematopoietic stem cells (HSCs), skewing of differentiation toward the myeloid/erythroid-associated multipotent progenitor 2 (MPP2) lineage at the expense of MPP4 lineage, and significantly higher DNA damage in lineage, Sca-1⁺, cKit⁺ (LSK) cells are all observed in DDT deficient PCNA^K164R/K164R mice.²¹⁶ Increased DNA damage in HSCs and skewing toward myeloid/erythroid progenitor lineages is an indicator of accelerated aging.^216,217 Furthermore, ATR is a crucial enzyme of the DNA damage response involved in maintaining fork stability and controlling replication stress.²¹⁸ In addition to greater dwarfism, higher replication stress, and hastened aging, the homozygous ATR-deficient mice do not show increased malignancy.²¹⁹ In conclusion, abnormalities in DDT tolerance pathways can increase DNA damage and replication stress, which can speed up aging. It will be interesting to find out whether DDT has any impact on how quickly other tissues and their stem cells age. Studies reported that depletion of Pol κ affects the gap-filling DNA synthesis phase of the NER process in dorsal root ganglion (DRG) neurons, delaying the removal of cisplatin adducts and, consequently, the restart of RNA synthesis. While the expression of other DNA polymerases remained unaltered, Pol κ was shown to have a transcriptional upregulation.²²⁰ DNA Pol κ is thought to be a key component in cellular survival under harmful circumstances because it participates in the bypass synthesis of various DNA damage.¹⁴ Pol κ levels were decreased using siRNA to evaluate the effects of Pol κ deficiency on DRG neurons that had been treated with cisplatin. As evidenced by a lower incorporation of thymidine analog into nuclear DNA, Pol κ -targeting siRNA reduced the level of cisplatin-induced DNA repair synthesis and nuclear Pol κ immunoreactivity in DRG neurons.²²⁰ Moreover, cisplatin-induced global transcriptional inhibition in DRG neurons was significantly worsened by Pol κ deletion.²²⁰ Thus, Pol κ may be essential for repairing neuronal DNA damage and maintaining neuronal function and its loss might make DRG neurons more susceptible to genotoxic assaults. Since DRG neurons do not have strong blood–brain barrier protection, it is conceivable that Pol κ gives these neurons a stronger capability to deal with DNA damage, as the basal levels of Pol κ are higher in DRG neurons than in the cortical neurons.^220,221 Therefore, Pol κ may be essential for the repair of neuronal DNA damage.

TLS polymerase gene mutations and human diseases

The link between TLS polymerase gene mutation(s) and human diseases dates back to the time when XPV syndrome was linked with a point mutation in the POLH. As mentioned earlier the elevated rates of UV-induced mutagenesis in XPV patients are ascribed to their high incidence of skin cancer. The mutated Pol η hampers the lesion bypass accuracy of UV-induced TT-dimers, causing mutations at TT sites and the development of skin cancer²²² by activating compensatory, albeit error-prone TLS of CPD lesions by other “inserter” Y-family DNA polymerases. Since then several other mutations and SNPs have been identified in the TLS polymerase genes, with the majority showing significant association with cancer.¹⁶ The biallelic inactivating mutation in REV7 that codes for a mutant REV7-V85E protein was recently reported in an infant with acute marrow failure.²⁰⁰ When exposed to DNA crosslinking agents, cells derived from the patient showed an extended Fanconi anemia phenotype comprised of increased chromosomal breaks and G2/M accumulation, the buildup of γH2AX and 53BP1 foci, and higher p53/p21 activation.²⁰⁰ In the Han Chinese population, rs10077427 SNP of the POLK gene and rs462779 SNP of the REV3 gene were found significantly associated with breast cancer and colorectal cancer, respectively.^223,224 Our group recently reported a significant association of three REV3 SNPs; rs1002481, rs462779, and rs465646 with nonsquamous cell lung cancer in the North Indian population.²²⁵ A study comparing individuals with a POLK C-C (rs5744533-rs5744724) haplotype to those with a POLK C-G haplotype revealed that the latter had a lower chance of developing lung cancer. The heterozygous condition of REV1 rs3087386 and rs3792136 were also an independent predictive factor for lung cancer survival, with hazard ratio (HR) values of 1.54 (95% CI: 1.12–2.12) and 1.44 (95% CI: 1.06–1.97), respectively.²²⁶ Interestingly, while REV7 mutations were connected to Fanconi anemia, mutations in the catalytic subunit REV3 in humans are autosomal dominantly responsible for the Mobius syndrome.²²⁷ Baring few studies, such as the association of Y89D mutation with myopia in a Chinese population,⁵⁵ the primase and polymerase abnormalities caused by the PRIMPOL Y89D mutation in the DT40 B-cell lymphoma,²²⁸ and the Y100H mutation detected in various cancers.²²⁹ there is limited evidence to support the role of PRIMPOL in human diseases. The Y100H mutation promotes the incorporation of ribonucleotides (NTPs) at the expense of dNTPs and is believed to facilitate the survival of cells at an early stage of tumorigenesis.²²⁹

TLS activity and therapeutic drug resistance

The primary objective of TLS is to replicate preceding DNA damage brought on by first-line genotoxic drugs, which leads to decreasing effectiveness and the development of chemoresistance. Different tumor types exhibit different levels of relative expression of the Y family polymerase Pol κ. In colorectal and stomach cancers, for instance, the expression of Pol κ was found to decrease^230,231 whereas, in lung and brain cancer samples the levels were found significantly elevated.^232–234 Surprisingly, two-thirds of head and neck mucosal-derived squamous cell carcinoma (HNSCC) specimens had high expression of Pol η, in contrast to a decreased level in colorectal, lung, and stomach malignancies.^235,236 While Pol η is unquestionably crucial for the emergence of platinum drug resistance, Rev1, and Pol ζ are also crucial for the pathways behind chemoresistance and mutagenesis in malignancies treated with genotoxic chemicals. In the case of cisplatin resistance, Rad18- and PCNA-mediated ubiquitination aids in coordinating REV1, Pol η, and Pol ζ bypass of intrastrand cross-links, while REV1 and Pol ζ work independently of PCNA-Ub to allow repair of interstrand cross-links brought on by cisplatin.²³⁷ Furthermore, the pace at which tumor cells develop cisplatin resistance is slowed down by the reduction of either REV1 or Pol ζ.²³⁸ In a murine model of Burkitt’s lymphoma, REV1 expression was downregulated, leading to a decrease in mutations and the development of cyclophosphamide resistance, tumor regression, and an improvement in overall survival.²³⁹ Additional research on TLS and chemoresistance has revealed a potential application for TLS inhibition in combination therapy with various other DNA-damaging agents. In the TGCT models mentioned above, REV7 depletion increased susceptibility to the DNA intercalator doxorubicin and the alkylating agent mitomycin C.²⁴⁰ The acquired resistance and tumor relapse that are frequently observed after first therapy are ultimately caused by cancer cells that survive first-line genotoxic medicines by showing enhanced mutation rates. TLS is thought to be crucial for this process, enabling malignancies to resist DNA-damaging chemicals.^93,241

Strategies to enhance cancer therapy

TLS DNA polymerases are attractive targets for improving the efficacy of chemotherapy, and some DNA polymerase mutations, polymorphisms, and activity levels can serve as useful prognostic indicators for determining the best course of treatment for a variety of oncological conditions. Undoubtedly, a viable strategy for improving the efficiency and lowering cytotoxicity of chemotherapeutic medications is to find novel, precise, and efficient inhibitors of TLS polymerases and crucial protein-protein interactions of the translesome.²⁴² Studies are now focusing on the development of TLS polymerase inhibitors which can hamper the exacerbation of the diseases including chemoresistance by targeting the TLS polymerases and their pathways. One study reported the identification of JH-RE-06, a small molecule that inhibits the recruitment of mutagenic Pol ζ.²⁴³ Interestingly, JH-RE-06 concentrates on a region of essentially featureless surface of REV1 where it interacts with the REV7 subunit of Pol ζ. When JH-RE-06 binds to REV1, REV1 dimerization is induced, which prevents the interaction between REV1 and REV7, thus failing to recruit Pol ζ to the DNA. In cultured human and mouse cell lines, JH-RE-06 suppresses mutagenic TLS and increases cisplatin-induced toxicity.²⁴⁴ The development of TLS inhibitors as a new class of chemotherapy adjuvants was made possible by the co-administration of JH-RE-06 and cisplatin, which inhibited the growth of xenograft human melanoma in mice.²⁴³ In another study, a small molecule inhibitor of REV1 UBM2 prevented the DDT with the treated cells showing delayed removal of UV-induced CPDs from nuclei, reduced UV-induced mutations of the HPRT gene, and reduced ability of cells exposed to cyclophosphamide or cisplatin to proliferate through clonogenic processes.²⁴⁵ Novobiocin, an aminocoumarin antibiotic also known as cathomycin or albamycin, has been demonstrated to bind directly to the Pol θ ATPase domain, reduce its ATPase function, and mimic Pol θ depletion.²⁴⁶ In genetically engineered mouse models, xenograft, and patient-derived xenograft models, novobiocin kills HR-deficient breast and ovarian cancers.²⁴⁶

The predetermined active sites of TLS polymerases that can readily accommodate a wide range of DNA lesions may overcome the damage caused by chemotherapy drugs. As mentioned earlier, TLS involves two-step processes with Y family polymerase proceeded by B family DNA polymerases in cancerous cells.^247,248 Hence, by using short hairpin RNA or other target protein depletion strategies, it is possible to enhance the therapeutic effects of the chemotherapeutic drug treatment through the depletion of TLS polymerases.^{239,249–251} This further helps by not allowing drug resistance and reoccurrence of the cancerous cells. Based on the capacity to skip the damaged site it can lead to the prevention of collapsing of the replication fork leading to apoptosis. The wrong nucleotide incorporation can either lead to mutation postcell division or help cells to survive; hence, TLS is a key process by which cancer cells develop resistance to genotoxic treatment. Recent research reveals that DNA replication-associated single-stranded DNA (ssDNA) gaps cause PARPi toxicity in HR-deficient (HRD) cells. This susceptibility raises the possibility that the TLS pathway may serve as a novel therapeutic target for HRD tumors. In addition, the development of small molecule inhibitors of the TLS polymerase machinery has been identified as a therapeutic approach with promise for improving the efficacy of first-line chemotherapy and reducing cellular acquired resistance to DNA damaging drugs.

FUTURE PERSPECTIVES

The cumulative evidence suggests that inhibition of DDT, particularly the error-prone TLS pathway, can be used as a chemotherapeutic strategy for conventional treatment-resistant cancer cells. Current research on understanding the molecular mechanism of DNA-damage tolerance will create ideas and opportunities for further advancement in this area. Under the circumstances, an in-depth investigation is needed to understand the pathway choice and factors that influence it. In the TLS pathway, it is important to understand the individual role of TLS polymerases in bypassing the lesion in normal and disease cells. This will provide additional information for designing specific inhibitors and targeting TLS polymerases and accessory proteins.

Furthermore, agents that will directly block the protein function or interactions that are crucial for DDT can be effective therapeutic drugs. Tremendous attention has been given to the inhibition of TLS. However, with the recent understanding of genetics, fork reversal biochemistry, and template switching, there is an opportunity to identify inhibitors targeting these pathways as well. Additional challenges include the clinical utility of synthesized molecules and the development of target-specific inhibitory molecules that produce a cellular phenotypic response. Ongoing and future research gives a potential hope to target DDT pathways as a therapeutic approach for various diseases and associated complications.

CONCLUSION

In response to DNA damage, cells depend on the highly conserved DNA damage tolerance pathways, including translesion synthesis to prevent replication fork collapse and carry on DNA replication unhindered across the lesion-containing sites or at stalled replication forks. However, inequities in these pathways in neoplastic cells cause substantial mutagenesis, which may lead to chemoresistance and enhanced cancer cell survival. Inhibiting DDT pathways, especially the error-prone TLS pathway could therefore possibly sensitize cancer cells to existing chemotherapeutic agents and may overcome chemoresistance. Therefore, the design and characterization of DDT inhibitors might eventually lead to the development of a new class of cancer therapeutics that have the potential to improve the treatment efficacy of existing chemotherapeutic medicines, especially in chemoresistant cancers. In addition, understanding the noncanonical functions of TLS polymerases in human cells is also going to be an area of active research since many essential functions have already been attributed to TLS polymerases.

Funding Statement

A.B acknowledges the financial support from Indian Council of Medical Research (grant nos. 5/10/15/CAR-SMVDU/2018-RBMCH and 6719/2020-DDl/BMS) and A.D acknowledges the financial support from Indian Council of Medical Research (grant no. 6719/2020-DDl/BMS).

AUTHORS’ CONTRIBUTIONS

Conceptualization, A.B., T.K.P.; methodology, A.B., A.D., J.V., and G.L.; writing—original draft preparation, J.V., and G.L.; writing – review and editing, A.B., A.D., T.K.P., K.M.B and W.X.; visualization, A.B., and A.D.; supervision, A.B.; project administration, A.B. and A.D. All authors have read and agreed to the published version of the manuscript.

DATA AVAILABILITY STATEMENT

Data sharing not applicable – no new data generated.