Opportunities for new studies of nuclear DNA replication enzymology in budding yeast

Abstract

Three major eukaryotic DNA polymerases, Polymerases α, δ and ε (Pols α, δ and ε), perform the fundamental process of DNA synthesis at the replication fork both accurately and efficiently. In trying to understand the necessity and flexibility of polymerase usage, we recently reported that budding yeast cells lacking Pol ε exonuclease and polymerase domains (pol2-16) survive but have severe growth defects, checkpoint activation, increased level of dNTP pools as well as significant increase in the mutation rates. Herein we suggest new opportunities to distinguish the roles of Pol ε from those of two other eukaryotic B family DNA polymerases, Pols δ and ζ.

Eukaryotic cells perform highly efficient and accurate chromosomal DNA replication using replisomes, molecular machines composed of proteins bearing multiple enzymatic DNA manipulation and synthesis activities. Key among these replisome components are the DNA polymerases, enzymes tasked with faithful synthesis of new DNA. Among numerous DNA polymerases encoded by eukaryotic genomes are four B family enzymes involved in nuclear DNA replication: DNA polymerases (Pols) α, δ, ε and ζ. A great deal of evidence supports the idea that Pol ε is the major leading strand replicase, and that Pols δ and α replicate the lagging strand ((Burgers and Kunkel 2017) and references therein)). Moreover, recent evidence further suggests that Pol δ plays a key role in initiating leading strand replication (Yeeles et al. 2017; Aria and Yeeles 2018; Garbacz et al. 2018). This replication model stems from the DNA strand-specific error specificities of these polymerases and from their patterns of ribonucleotide incorporation during replication (Clausen et al. 2015; Lujan et al. 2016). Also included among the evidence in support of the model are several properties of a budding yeast pol2-16 mutant bearing in-frame deletion of residues 176–1134 of POL2 that inactivates both the polymerase and 3′-exonuclease activities of Pol ε. Although pol2-16 mutants survive, they have a severe growth defect and tiny spore colonies (Garbacz et al. 2019). They also have larger than normal cells, an increased doubling time, aberrant nuclei, and higher than normal dNTP concentrations.

Moreover, the pol2-16 mutant also has a strong mutator phenotype can be suppressed and that partly depends on both the TLS polymerase Pol ζ and the 3′-exonuclease proofreading activity of Pol δ (Garbacz et al. 2019). Collectively then, current results support a model wherein slow-moving replication forks caused by the lack of Pol ε’s catalytic activities result in greater involvement of Pol δ and Pol ζ in DNA replication. In conjunction with numerous other observations on the roles of the B family replicases in nuclear DNA replication, these data suggest new experiments that could further advance our knowledge of the roles of the B family polymerases in cellular transactions.

One example involves the role of Pol ζ in replicative DNA synthesis. This polymerase shares with Pol α the lack of a 3′-exonuclease, such that neither Pol α nor Pol ζ can proofread its own replication errors. While Pol α is essential for starting replication by initiating new DNA chains, Pol ζ is thought to operate later in replication, as a translesion DNA polymerase (Makarova and Burgers 2015; Szwajczak et al. 2018). Pol ζ’s role in lesion bypass can now be examined during leading and lagging strand replication, using specific reporter genes and/or whole genome sequencing. When performed in conjunction with mutations in genes involved in specialized replication (or DNA repair) reactions, such experiments should advance our understanding of Pol ζ’s roles, be they in copying undamaged yet intrinsically difficult-to replicate DNA sequences, lesions generated endogenously during cellular metabolism (e.g., from oxidative stress or alkylating agents), and/or during bypass of lesions generated by external environmental stressors. Pol ζ is also the least accurate of the four eukaryotic B family polymerases and the most prone of these four polymerases to incorporate ribonucleotides that are the most-frequently incorporated non-canonical nucleotides in DNA (Makarova et al. 2014). Although Pol ζ is thought to synthesize far less of the nuclear genome than Pols α, δ and ε, its ability to incorporate ribonucleotides in vivo, its ability to bypass newly incorporated ribonucleotides, and the consequences of such incorporation events largely remain to be investigated.

Another opportunity to study replication enzymology is suggested by recent studies of Pol δ proofreading in the pol2-16 strain and earlier work on exonucleolytic proofreading. Along with the high selectivity of the three major replicases in selecting the correct nucleotide for initial incorporation, and the final correction of replication errors by DNA mismatch repair, proofreading is the intermediate, third major process responsible for the very high fidelity of replication of the nuclear genome (reviewed in (Bębenek and Ziuzia-Graczyk 2018)). In addition to their polymerase activities, Pol δ and Pol ε each has a 3′-exonuclease that can proofread replication errors by fraying the mismatched primer terminus and moving the mismatched base from the polymerase active site directly to the exonuclease active site for removal. Motivated by biochemical studies indicating that the exonuclease active site can also bind directly to a mismatched primer terminus (Perrino and Loeb 1990), a study in vivo suggests that yeast Pol δ can also extrinsically proofread mismatches made by Pol α during synthesis of primers used for replication at origins and during Okazaki fragment synthesis of the lagging strand (Pavlov et al. 2006). That study, and subsequent work published by us (St Charles et al. 2015) and by others(Flood et al. 2015), suggest that extrinsic proofreading may be more prevalent that is currently appreciated. If so, defects in extrinsic proofreading could also contribute to the origin of diseases, including cancer.

The severe phenotypes of the yeast pol2-16 mutant mentioned above may also partly be reflected in the consequences of Pol ε deficiency in higher eukaryotes. For example, during the last decade, human FILS (Pachlopnik Schmid et al. 2012) and IMAGe syndromes (Logan et al. 2018), have been identified that have decreased levels of Pol ε. Defects in human Pol δ are thought to underlie MDP syndrome, lipodystrophy, and atypical Werner’s syndrome (Weedon et al. 2013; Pelosini et al. 2014; Lessel et al. 2015; Reinier et al. 2015; Nicolas et al. 2016). Mutations in human Pol α are thought to underlie several X-linked disorders with symptoms including intellectual disability, short stature, microcephaly and hypogonadism (Starokadomskyy et al. 2016; Van Esch et al. 2019). Moreover, recent studies of Pole4 −/− p53+/− mouse mutant with decreased cellular level of Pol ε revealed that such mice exhibit growth abnormalities and also have increased rates of lymphomas (Bellelli et al. 2018). That study and numerous studies in recent years reveal that mutations in the genes encoding human Pol ε and Pol δ are associated with various types of cancer (Cerami et al. 2012; Kandoth et al. 2013; Palles et al. 2013; Forbes et al. 2015; Barbari and Shcherbakova 2017; Guenther et al. 2018; Vande Perre et al. 2019). This includes a recent study of one mutant of Pol ε (POLE^P286R) that when mimicked in mouse, results in an ‘ultramutator’ phenotype (Li et al. 2018; Parkash et al. 2019; Xing et al. 2019). These results strongly support the hypothesis that errors in DNA replication produce a mutator phenotype that is involved in generating the multiple mutations required for tumor development (Loeb et al. 1974). Thus, deficiencies in the B family polymerases that generate genome instability and have consequences for human health remains a very important topic for future study.

DNA polymerase specialization is found throughout all kingdoms of life (Lujan et al. 2016). Pol α is highly specialized for initiating DNA replication, Pol ζ appears to be highly specialized for synthesis of difficult-to replicate sequences, and Pol ε is highly specialized for leading strand DNA replication. In contrast, Pol δ has roles in initiating replication, in bulk lagging strand replication, in all stages of break-induced replication (Lydeard et al. 2007), as well as in replication restart at replication fork barriers (Miyabe et al. 2015). The strategic use of defined mutants of Pol δ should assist in further defining its role(s) in each of these critical DNA transactions.