Stable interactions between DNA polymerase δ catalytic and structural subunits are essential for efficient DNA repair

Abstract

Eukaryotic DNA polymerase δ (Pol δ) activity is crucial for chromosome replication and DNA repair and thus, plays an essential role in genome stability. In Saccharomyces cerevisiae, Pol δ is a heterotrimeric complex composed of the catalytic subunit Pol3, the structural B subunit Pol31, and Pol32, an additional auxiliary subunit. Pol3 interacts with Pol31 thanks to its C-terminal domain (CTD) and this interaction is of functional importance both in DNA replication and DNA repair. Interestingly, deletion of the last four C-terminal Pol3 residues, LSKW, in the pol3-ct mutant does not affect DNA replication but leads to defects in homologous recombination and in break-induced replication (BIR) repair pathways. The defect associated with pol3-ct could result from a defective interaction between Pol δ and a protein involved in recombination. However, we show that the LSKW motif is required for the interaction between Pol3 C-terminal end and Pol31. This loss of interaction is relevant in vivo since we found that pol3-ct confers HU sensitivity on its own and synthetic lethality with a POL32 deletion. Moreover, pol3-ct shows genetic interactions, both suppression and synthetic lethality, with POL31 mutant alleles. Structural analyses indicate that the B subunit of Pol δ displays a major conserved region at its surface and that pol31 alleles interacting with pol3-ct, correspond to substitutions of Pol31 amino acids that are situated in this particular region. Superimposition of our Pol31 model on the 3D architecture of the phylogenetically related DNA polymerase α (Pol α) suggests that Pol3 CTD interacts with the conserved region of Pol31, thus providing a molecular basis to understand the defects associated with pol3-ct. Taken together, our data highlight a stringent dependence on Pol δ complex stability in DNA repair.

Introduction

Central to the chromosomal DNA replication apparatus in eukaryotic cells are three essential DNA polymerases, namely DNA polymerase α, δ and ɛ [1]. The primase subunit of the Pol α/primase complex initiates synthesis by oligomerizing short RNA primers on both leading and lagging strands. These primers are initially extended by Pol α and subsequently transferred to processive polymerases Pol ɛ and Pol δ. In contrast to the quite obvious role of Pol α, the specific roles for Pol ɛ and Pol δ in DNA replication were much debated for years. However, recent evidence strongly supports a model of the replication fork wherein the leading and the lagging strand templates are primarily copied by Pol ɛ and Pol δ, respectively [2], [3].

Pol α, Pol δ and Pol ɛ belong to the B class of DNA polymerases [4] and their catalytic subunits are phylogenetically related and conserved from yeast to humans [5], [6]. Similarly, among the different accessory subunits of the replicative Pols, only the so-called B subunit [7], is clearly conserved in eukaryotic organisms [8]. Interestingly, Pol α, Pol δ and Pol ɛ share unifying features of their subunit organization that again underlines their evolutionary relationship. Indeed, the sequence conservation of the catalytic subunit extends past the polymerase fold to a cysteine-rich C-terminal domain (CTD) that contains two zinc fingers. A large body of experimental evidence has highlighted that this CTD interacts with the B subunit and that this interaction is of functional importance in Pol α [9], Pol δ [10], [11] and Pol ɛ [12]. Recently, a 3D architecture of Pol α has been published that allows a clearer overall picture of the catalytic and B subunit interactions for all three replicative polymerases [13]. An interesting conclusion drawn from this structural analysis is that it defines the C-terminal region of the replicative Pols α, δ and ɛ as an independently folded protein domain, separate from the polymerase fold.

DNA polymerase δ from Saccharomyces cerevisiae, has three subunits: Pol3/Cdc2, the catalytic subunit of 125 kDa that contains the polymerase and 3′ to 5′ exonuclease active site domains; Pol31/Hys2, the essential structural B subunit of 55 kDa; Pol32, the additional C subunit of 40 kDa [14]. The enzymes from Schizosaccharomyces pombe and humans have an additional small fourth subunit that functions to stabilize the complex [15], [16], [17]. The enzymes from the three different sources show roughly similar structure characteristics [7] and contain one of each of the subunits, i.e., they are a monomeric catalytic complex [18], [19]. The catalytic and the B subunits form a stable complex (as discussed above), to which the C subunit is tethered solely via interactions with the B subunit [14], [20]. The interaction between the B subunit and the C subunit of Pol δ participates to the functional integrity of the whole complex. Indeed, while POL32 is dispensable for growth in S. cerevisiae, the POL32 deletion mutant nevertheless shows poor growth and synthetic lethality with numerous genes that function in DNA metabolism [14], [21]. In addition, the orthologous S. pombe Cdc27 gene is essential for growth [18], [22]. In both S. cerevisiae and S. pombe, the C subunit interacts with the B subunit via its N-terminal part, whereas its C-terminal part contains a conserved PCNA-binding motif and a motif that mediates interaction with Pol α [18], [23], [24]. It is noteworthy that deletion of the POL32 N-terminal end exhibits the same phenotype than the complete deletion [24]. Using two-hybrid screening, the human C subunit, p66, has been shown to contain p50 (the human Pol δ B subunit) and PCNA binding domains within the 144 N-terminal and 20 C-terminal amino acids, respectively [25]. Recently, the crystal structure of the complex between p50 and the 144 amino acids N-terminal domain of p66 of human Pol δ has been reported [26]. It appears that p66 makes an intricate complex with p50 and has a significantly large surface area buried at the heterodimer interface. Interestingly, in this study, it is noted that in spite of the formation of a very tight complex between p50 and p66 subunits, the residues involved in dimerization correspond to poorly conserved regions among eukaryotic sequences. The analysis of the secondary structure elements of p50 and p66 reveals, however, that this complex is likely to share the same overall fold with its orthologs [26].

The interaction between the Pol δ subunits plays a crucial role in the stability of the Pol δ complex and allows full proficiency of Pol δ in DNA replication. Interestingly, essential functions of Pol δ in DNA repair are even more sensitive to defective interactions within the complex. This observation comes essentially from the study of Pol δ mutants in S. cerevisiae. The pol3-13 mutation leads to the substitution of a Pol3 cysteine at position 1074 by a serine (C1074S). Pol3 contains a CTD with two zinc fingers, ZnF1 and ZnF2. Residue C1074 is in the last ZnF2 cysteine. Two-hybrid studies have shown that the mutated C1074S ZnF2 module can no longer interact with Pol31 [11]. In vivo, the pol3-13 allele confers thermosensitivity, but at the permissive temperature of 30 °C, it confers UV and γ-rays sensitivity and almost a complete defect in both γ- and UV-induced recombination and UV-induced mutagenesis [10]. Similarly, Pol32 interacts with Pol31 but also with Pol1, the catalytic subunit of Pol α [24], [27] and the translesion DNA polymerase Rev1 [28]. Deletion of POL32 is viable but confers UV sensitivity and abolishes UV-induced mutagenesis [29]. In addition, break-induced replication (BIR) repair is seriously impaired in the pol32Δ mutant and gives rise to aberrant double-strand break (DSB) repair resulting in half crossovers [30], [31], [32]. Interestingly, Rev1 is required for UV-induced mutagenesis [33] while Pol α is involved in BIR [31]. Thus, Pol32 interaction with Rev1 and Pol α could have provided a clear molecular basis to understand the role of Pol32 in UV-induced mutagenesis and BIR. However, the deletion of neither the Pol1 or Rev1 interacting domains of Pol32 leads to defects in the corresponding DNA repair pathway [24], [31]. By contrast, the essential functions of Pol32 in DNA repair are comprised in a small amino-terminal domain of Pol32, not much larger than that necessary to specify the interaction with Pol31, suggesting that the Pol32 subunit mediates its role in UV-induced mutagenesis and BIR through the Pol δ complex [24], [31].

In our previous studies, we described the pol3-ct allele that changes a Leu codon to a stop codon at position +1094 in S. cerevisiae resulting in the loss of the four last C-terminal amino acids (LSKW) of Pol3 [34], [35]. Yeast cells carrying this allele show no obvious defects in genome replication since they are neither thermosensitive (ts) nor mutators and they do not display high levels of spontaneous recombination. In addition, they are not sensitive to UV and γ-rays irradiation. We found, however, that the pol3-ct mutant, although repairing DSB as efficiently as WT, yields shorter recombination intermediates in both meiosis and mitosis. This interesting phenotype of the pol3-ct mutant implies that the role of Pol δ during HR is not restricted to gap filling but participates as well in recombination intermediates extension.

The phenotype associated with the pol3-ct allele raises the attractive possibility that in the pol3-ct mutant, Pol δ fails to interact with a protein involved in recombination intermediates extension. However, in our present study, we provide genetic and structural evidence showing that pol3-ct leads to a destabilized interaction between Pol3 CTD and Pol31. Therefore, we reach the alternative conclusion that DNA repair synthesis is clearly dependent on Pol δ complex stability.