Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2012 Jan 1.
Published in final edited form as: Biochemistry (Mosc). 2011 Jan;76(1):49–60. doi: 10.1134/s000629791101007x

PARTICIPATION OF TRANSLESION SYNTHESIS DNA POLYMERASES IN THE MAINTENANCE OF CHROMOSOME INTEGRITY IN YEAST SACCHAROMYCES CEREVISIAE

O V Kochenova 1,2, J V Soshkina 2, E I Stepchenkova 2,3, S G Inge-Vechtomov 2,3, P V Shcherbakova 1,*
PMCID: PMC3174683  NIHMSID: NIHMS280374  PMID: 21568839

Abstract

We employed a genetic assay based on illegitimate hybridization of heterothallic Saccharomyces cerevisiae strains (the α-test) to analyze the consequences for genome stability of inactivating translesion synthesis (TLS) DNA polymerases. The α-test is the only assay that measures the frequency of different types of mutational changes (point mutations, recombination, chromosome or chromosome arm loss) and temporary changes in genetic material simultaneously. All these events are manifested as illegitimate hybridization and can be distinguished by genetic analysis of the hybrids and cytoductants. We studied the effect of Polζ, Polη and Rev1 deficiency on the genome stability in the absence of genotoxic treatment and in UV-irradiated cells. We show that, in spite of the increased percent of accurately repaired primary lesions, chromosome fragility, rearrangements and loss occur in the absence of Polζ and Polη. Our findings contribute to further refinement of the current models of translesion synthesis and the organization of eukaryotic replication fork.

Keywords: Saccharomyces cerevisiae, translesion synthesis, recombination, chromosome stability

INTRODUCTION

Translesion synthesis (TLS) DNA polymerases belong to the specialized group of DNA polymerases whose main function is to promote replication of damaged DNA by inserting nucleotides across from and past DNA lesions. In yeast Saccharomyces cerevisiae, several genes control translesion synthesis: the REV3 and REV7 genes encode for the two subunits of DNA polymerase ζ [1], the RAD30 gene encodes DNA polymerase η [2, 3], and the REV1 gene encodes a deoxycytidyl transferase, the enzyme that in addition to the polymerase activity has the ability to interact with Polζ, Polη and a subunit of the replicative DNA polymerase δ [4, 5]. Polζ can synthesize with low efficiency across several types of lesions, such as thymine glycols, T-T (6–4) photoproducts and abasic sites [6–8]. The main function of Polζ is extension of primers with a mismatched terminal nucleotide pair [9–11]. Polη is the primary TLS polymerase responsible for accurate bypass of cis-syn cyclobutane pyrimidine dimers, a major DNA lesion resulting from UV irradiation, and 7,8-dihydro-8-oxoguanine occurring during oxidative stress [3; 12]. However, Polη exhibits a low fidelity when bypassing other lesions and during replication of undamaged DNA in vitro [13–16]. Studies of Pol η mutants have important clinical implications, because mutations in the gene encoding for Polη in humans are associated with a complex disorder Xeroderma pigmentosum variant characterized by sunlight sensitivity [17, 18]. Rev1, which interacts with multiple DNA polymerases, is believed to play an important regulatory role in DNA polymerase switching and during DNA synthesis past lesions [8; 19]. Although the deoxycytidyl transferase activity of Rev1 is used during the bypass of some lesions [20–22], the essential function of Rev1 in TLS is structural rather than enzymatic.

Inactivation of certain TLS polymerases leads to a significant drop of point mutation frequency. For instance, deletions of the REV3, REV7 and REV1 genes reduce the spontaneous mutagenesis up to 70% and virtually abolish mutagenesis induced by many DNA damaging agents [23]. This assumes TLS to be a highly mutagenic process. Nevertheless, a lack of the TLS may result in destabilization of genetic material on a higher level and may provoke chromosome rearrangements and even chromosome loss. In higher eukaryotes, TLS DNA polymerases were shown to be involved in some processes other than translesion synthesis. REV1 was reported to induce Ig gene conversion and generate sister chromatid exchanges during DSB repair in higher eukaryotes independently from Polζ [24]. The hREV7 is thought to play an important role in the G2/M checkpoint. In higher eukaryotes, REV7 interacts with specific factors of anaphase promoting complex (APC/C) and with the spindle checkpoint protein MAD2 [25, 26]. These interactions may regulate the cell cycle progression and have provided the first link between a mutagenic DNA polymerase and chromosome segregation control [26, 27]. Studies of the chicken DT40 line have shown the contribution of the REV genes to the prevention of the chromosomal fragility/instability, increased chromosomal rearrangements and homologous recombination [24, 28, 29]. Rev3 was also shown to be involved in homologous recombination (HR)-dependent double-strand-break (DSB) repair after ionized radiation (IR). Inactivation of the REV3 gene in chicken caused a significant increase in the level of chromosomal breaks after IR. These data suggest the involvement of Rev3 in DNA synthesis during HR-mediated DSB repair [28]. Yeast Rev3 was also shown to participate in DNA synthesis during DSB repair [30, 31].

Thus, the contribution of TLS DNA polymerases to the genome stability maintenance is very complex. Here we undertake a comprehensive study of the role of TLS DNA polymerases in promoting and preventing genome instability in the same experimental system. To test the role of each DNA polymerase in the genome stability maintenance we used an approach described in a previous work and named the α-test for S. cerevisiae [32–34]. The α-test allows one to score a wide spectrum of changes in the genetic material. It simultaneously measures the frequency of gene mutation, recombination, loss of chromosome III and chromosome arm loss. The distinctive feature of the α-test is its ability to measure the frequency of primary (non-inherited) lesions in the genetic material and to estimate the proportion of primary lesions that were precisely eliminated by repair after mating, and the proportion of lesions that led to inherited genetic changes [32, 33]. In this study, we provide evidence that the activity of Polζ and Polη prevents global chromosome rearrangements and chromosome loss, potentially by suppressing double-strand breaks formation.

MATERIAL AND METHODS

Strains and growth conditions

The yeast strains used in the study are listed in Table 1. Strains were grown in rich media (YPD) or minimal media (MD) containing the corresponding selective amino acids required for selection. All liquid and solid media were prepared according to standard protocols [35]. Media with 24 ml/L glycerol as a sole source of carbon was used to determine respiratory competence. The media for cytoductants selection was MD lacking glucose and containing ethyl alcohol (20 ml/L) and cycloheximide (5 mg/L). Yeast cultures were grown at 30°C.

Table 1.

S. cerevisiae strains used

Strain Genotype Application
D926 MATα//MATα leu2Δ//leu2Δ lys2Δ//lys2Δ ura3Δ//ura3Δ his4Δ//his4Δ thr4Δ//thr4Δ CYHs Partner for copulation
K5-99-35B-D924 MATα ura3Δ leu2Δ met15Δ lys5 Strain-tester for illegitimate hybridization
K5-99-35B-D924-rad30Δ MATα ura3Δ leu2Δ met15Δ lys5 rad30Δ::KanMX4
K5-99-35B-D924-rev3Δ MATα ura3Δ leu2Δ met15Δ lys5 rev3Δ::LEU2
K5-99-35B-D924-rev3Δ-rad30Δ MATα ura3Δ leu2Δ met15Δ lys5 rev3Δ::LEU2 rad30::KanMX4
Κ5-99-35B-D924-rev7Δ MATα ura3 leu2Δ met15Δ lys5 cyhr rev7Δ::KanMX4
Κ5-99-35B-D924-rev1Δ MATα ura3 leu2Δ met15Δ lys5 cyhr rev1Δ::KanMX4
26B-D924 MATα uraΔ3 leu2Δ met15Δ kar1-1 Strain-tester for cytoduction
26B-D924-rad30Δ MATα ura3Δ leu2Δ met15Δ kar1-1 cyhr rad30Δ::kanMX4
26B-D924-rev3Δ MATα ura3Δ leu2Δ met15Δ kar1-1 cyhr rev3Δ::LEU2
26B-D924-rev3Δ-rad30Δ MATα ura3Δ leu2Δ met15Δ kar1-1 cyhr rev3Δ::LEU2 rad30::KanMX4
26B-D924-rev7Δ MATα ura3 leu2Δ met15Δ kar1-1 cyhr rev7Δ::KanMX4
26B-D924-rev1Δ MATα ura3 leu2Δ met15Δ kar1-1cyhr rev1Δ::KanMX4
78A-P2345 MATα his5 Testers for mating type determination
2G-P2345 MATa his5
2A-P143 MATα ura3Δ ade8Δ
Open in a new tab

Strain construction

Strains with a deletion of the REV3 gene were constructed as previously described [36]. The RAD30, REV7 and REV1 genes were disrupted by a selectable KanMX cassette as described [37]. The REV7 and REV1 gene deletions were confirmed by decreased UV-induced mutagenesis [38]. The RAD30 deletion was confirmed by increased sensitivity to UV irradiation [39].

Measurement of illegitimate hybridization and cytoduction frequency

For illegitimate hybridization, 50 μl of fresh overnight culture of the tester strain (K5-99-35B-D924 or its derivates) were transferred to selective media containing histidine, threonine, leucine and uracil for hybrid growth. In the case of UV light treatment, plates with the tester strain were irradiated with UV (25 J/m2). An aliquot (100 μl) of the partner strain D926 were then added. At the same time, the cultures of the tested strain were diluted appropriately and plated on YPD to determine the number of viable cells. Surviving colonies were scored after two days of incubation, and hybrids were scored after three days.

For cytoduction, fresh overnight independent cultures of strains to be tested were concentrated ten times (≈109 cells per ml). The aliquots (100 μl) of each strain (recipient and donor) were transferred onto rich media. Two strains mixed together were incubated for two days and then replica-plated onto the selective media containing uracile, leucine, methionine and ethyl alcohol (20 ml/L) and cycloheximide (5 mg/L) for cytoductants growth. To determine the survival of the recipient strain, cell cultures were diluted to appropriate cell densities and plated on YPD. Surviving colonies were scored after two days of incubation, and cytoductants were scored after ten days. For UV-induced cytoduction, 100 μl of the recipient strain (26B-D924 and its derivates) was irradiated with UV (25 J/m2) on plates. An equivalent amount of the donor strain was then added to the treated recipient strain.

The frequency of hybridization and cytoduction was calculated as F=(Ma)/(Nb), where M is the number of hybrids/cytoductants, N is the number of survivors, and a and b are the corresponding dilution factors. The significance of differences between samples was evaluated by Mann-Whitney test with P=0.05 [40]. For each experiment, at least nine independent cultures were used. Phenotypes of illegitimate hybrids and cytoductants were determined by replica-plating on media lacking selected amino acids. Based on their phenotypes, cytoductants and hybrids were placed into appropriate classes. The frequency of each class was calculated by multiplying the proportion of each class by the overall frequency hybrids or cytoductants.

RESULTS

A genetic system for the study of the genetoxic factors influence on the stability of genetic material

To study the effects of TLS DNA polymerases inactivation on genome stability we utilized a test-system, which allows for the detection of a wide spectrum of genetic events. The test-system used in the study was named the α-test [32, 33]. Criterion of genome instability in the α-test is the frequency of a mating type switch from mating type α to mating type a. Each event of a mating type switch may be scored in selective conditions. In the α-test two different strains of the same mating type are mixed on selective media which supports the growth of “illegitimate” hybrids only but not parental strains. Such “illegitimate” hybridization is possible when one of mating cells had changed its mating type from α to a. Different changes of genetic material may lead to a mating type switch in yeast strains of the mating type α. The mating type of S. cerevisiae cells is controlled by locus MAT on the right arm of chromosome III near the centromere [41–43]. The MAT locus determines the “a” or “α” cell type of haploid cells. Additional elements of the S. cerevisiae mating system are two cassettes (HMRa and HMRα), which contain silent genetic information for the “a” and “α” mating type, respectively (Fig. 1(A)). Normally, switching of the mating type occurs via the “cassette mechanism”: information from the HMRa or HMRα replaces the one in the MAT locus [43,46]. In addition, illegitimate hybridization between α-type cells of heterothallic strains may be caused by the loss of chromosome III, loss of the chromosome arm, recombination between the MAT locus and the cassette HMRa, point mutations and temporary changes (DNA lesions) in the MATα locus [47, 33]. The frequency of illegitimate hybridization increases after DNA damaging treatment due to the disturbance of the MAT locus expression [46]. All the events could be scored in the α-test by analyzing the phenotypes of illegitimate hybrids if both arms of the chromosome III are marked (Fig. 1(A, B), Table 2) [33].

Figure 1.

Figure 1

Open in a new tab

Genetic aspects of the α-test. (A) The structure of the S. cerevisiae chromosome III showing details of the MAT locus and markers on the right and left arms of the chromosome used in illegitimate hybridization experiments. P - bidirectional promoter of the MAT locus. Arrows designate the directions of transcription of the MATα1 and MATα2 genes in the MAT locus [44]. (B) Scheme of the selective system of illegitimate hybridization. Strain-tester bears wild-type alleles of the HIS4 and THR4 genes. The partner for hybridization bears mutant alleles or deletions of these genes. Illegitimate hybrids are selected on the media containing the appropriate amino acids for hybrid growth (see Materials and Methods), including histidine and threonine to allow for the growth of hybrids with lost chromosome III or rearrangements. The five possible phenotypes of hybrids are listed. The interpretation of these phenotypes is shown in Table 1. (C) Scheme of the selective system of illegitimate cytoduction (C) [45]. The tester strain, which is exposed to the genotoxic agent, bears allele of resistance to the antibiotic cycloheximide (cyhr) for nucleus selection and mutation kar1, which prevents karyogamy. The tester strain also lacks mitochondria, which results in [ρ0] phenotype (inability to utilize certain sources of carbon, e.g. ethanol) and serves as a cytoplasm marker. The partner for cytoduction bears markers of sensitivity to cycloheximide (CYHs), a wild-type allele of the KAR1 gene and has functional mitochondria [ρ+]. Cells with mixed cytoplasm and bearing the nucleus of the tester strain are selected on the medium containing ethyl alcohol and cycloheximide (see Materials and Methods).

Table 2.

Genetic events that lead to illegitimate mating of the MATα his4 thr4 strain to MATα HIS4 THR4 and phenotypes of the resulting hybrids and cytoductants [31].

Genetic event Phenotype of cytoductants Phenotype of illegitimate hybrids
Conversion between HMRa and MAT locus a n/m His+Thr+
Reciprocal recombination between MAT locus and HMRa Lethal n/m His+Thr−
Loss of the right arm of chromosome III α His +Thr−
Loss of chromosome III α His-Thr−
Mutations in MATα (matα1 or matα2) n/m α His+Thr+
Mutations in MATα (both in MATα1 and MATα2, or in the bidirectional promoter, MATα deletions) Recessive a or Alf- phenotype (from a- like fakers)
Transient lesions in the MATα locus (both in MATα1 and MATα2, or in the bidirectional promoter) α
Open in a new tab

n/m- non-mating phenotype

Moreover, a modification of the α-test (selective system of illegitimate cytoduction) allows us to detect primary lesions in the genetic material that are repaired after mating and do not result in heritable changes [45, 48]. Cytoduction is uncompleted hybridization, when the cytogamy is not followed by the nuclear fusion and diploid cell formation (Fig. 1(C), Table 2). Analysis of the phenotype of illegitimate cytoductants allows us to distinguish between the primary lesions and inherited changes of genetic material occurring both spontaneously and induced by genotoxic agents. Unfortunately, the system of cytoduction does not allow us to register the genetic events which are lethal in haploids, such as chromosome or chromosome arm loss. Therefore, these two variants of the α-test are complementary and together allow us to register a wide spectrum of genetic events [45].

Effect of TLS DNA polymerases inactivation on the frequency of illegitimate hybridization and cytoduction

The frequency of hybridization and cytoduction reflects the number of primary lesions that led to a disturbance of the MAT locus expression and, hence, mating type switching. Therefore, analysis of the frequency of induced and spontaneous hybridization or cytoduction allows us to estimate the activity of different genotoxic agents. In the present work, we investigated the effect of TLS polymerases inactivation on illegitimate hybridization and cytoduction in the absence of genotoxic treatment and after UV irradiation. The UV irradiation is a common environmental genotoxicant and a widely used model DNA-damaging agent that induces the formation of cyclobutane pyrimidine dimers and 6–4-photoproducts [49]. TLS polymerases are known to provide temporal tolerance to UV irradiation, as well as to various chemical agents [23]. We used the α-test to determine the consequences for the genome stability of inactivating this DNA damage tolerance pathway.

Normally, the frequency of spontaneous illegitimate hybridization in the wild type strain varies from 10−6 to 10−7 and the frequency of spontaneous illegitimate cytoduction is 10−7 to 10−8. In accordance with earlier studies [46, 45], we observed that UV light increases illegitimate hybridization and cytoduction in the wild-type strain up to five- and 3.5-fold, respectively, in comparison to the spontaneous frequency (legend in Fig 2, Table 3, respectively). The effects of rev3Δ, rev7Δ, rad30Δ and rev1Δ mutations on the frequency of spontaneous and UV-induced hybridization and cytoduction is shown in Fig. 2 and Table 3, respectively. The absolute frequencies of illegitimate hybridization varied widely between different experiments, but the observed effects of TLS mutations were similar in each experiment. Figures 2, 3 and 4, therefore, show the increase/decrease in the frequency of genetic events relative to the wild-type strain rather than the absolute frequency of these events.

Figure 2.

Figure 2

Open in a new tab

Relative frequency of illegitimate hybridization in rev3Δ, rev7Δ, rad30Δ, rev3Δ rad30Δ and rev1Δ derivatives of K5-35B-D924 strain. The hybridization frequency in the wild-type strain is designated as 1 and indicated by the bold horizontal line. All frequencies are means and standard errors for five experiments. The spontaneous frequencies in the wild-type strain were 130 (60–250); 45 (30–50); 170 (113–400); 168 (51–200); 30 (20 –46) (x10−6). The correspondent frequencies of illegitimate hybridization in the UV-treated wild type strain were 280 (190–610); 270 (170–300); 310 (230–1110); 230 (210–450); 96 (70–130) (x10−6). Asterisks indicate a statistically significant difference from the wild-type strain as determined by sign-test.

Table 3.

The frequency of spontaneous and UV-induced illegitimate cytoduction in the wild-type strain K5-99-35B-D924 and its derivates, × 10−7.

Strain Spontaneous frequency of cytoduction UV-induced frequency of cytoduction (25 J/m2)
WT 4 (3 –5) 18 (14–38)
rev3Δ 3 (1–5) 18 (9–26)
rev7Δ 5 (5–11) 106 (83–115)
rad30Δ 10 (8–17) 75 (54–110)
rev3Δ rad30Δ 9 (1–10) 100 (37–250)
Open in a new tab

All frequencies are medians and 95% confidence intervals. The frequencies that significantly differ from the wild-type strain are marked in bold letters.

Figure 3.

Figure 3

Open in a new tab

Relative frequency of chromosome loss (A) and chromosome arm loss (B) in TLS DNA polymerase mutants. All frequencies are means and standard errors for five experiments. All symbols are as shown in Figure 2.

Figure 4.

Figure 4

Open in a new tab

Relative frequency of gene conversion (A) and reciprocal recombination (B) in TLS DNA polymerase mutants in the absence of genotoxicant treatment (left) and UV-induced frequencies (right). All frequencies are means and standard errors for five experiments. All symbols are as shown in Figure 2.

We observed no significant effect of Polζ or Polη deficiency on the frequency of spontaneous illegitimate hybridization (Fig. 2(A)). However, UV-induced frequency of illegitimate hybridization was approximately eight-fold higher in the rev3Δ, rev7Δ and rad30Δ strains and approximately 20-fold higher in the double rev3Δ rad30Δ mutant in comparison to a UV-treated wild-type strain (Fig. 2(B)). In contrast, inactivation of the REV1 gene decreased the frequency of both spontaneous and UV-induced illegitimate hybridization 12- and 66-fold in comparison to the wild-type strain, respectively (Fig. 2(B)). Moreover, unlike in other studied strains, the UV treatment does not increase the frequency of illegitimate hybridization of the rev1Δ strain in comparison to its frequency of spontaneous hybridization. The genetic analysis of the rare hybrids obtained in the rev1Δ background further showed that this mutation decreased the frequency of all (both spontaneous and UV-induced) genetic events that we can distinguish in the illegitimate hybridization assay (chromosome loss, chromosome arm loss, reciprocal recombination and gene conversion).

We also examined the effect of the REV3, REV7 and RAD30 genes inactivation in the selective system of illegitimate cytoduction. Only RAD30 deletion showed an increased frequency of spontaneous cytoduction (approximately 2.5-fold), while deletion of the other genes had no significant effect (Table 3). In contrast, after exposure to UV light, the frequency of cytoduction was increased seven-fold in the rev7Δ strain, 4.6-fold in the rad30Δ strain and 6.5-fold in the double rev3Δ rad30Δ mutants (Table 3). Because Rev1 inactivation had a strong negative effect on the illegitimate mating, we were unable to measure the frequency of such a rare event as cytoduction in the rev1Δ strain.

Effect of TLS polymerase inactivation on chromosome or chromosome arm loss

Previous findings that the inactivation of the REV genes (REV1, REV3 and REV7) was shown to cause chromosome loss and breaks in vertebrates [24, 27] inspired us to study the effect of TLS polymerases inactivation on chromosome stability in yeast. Illegitimate mating of S. cerevisiae strains of the same mating type α could occur upon the loss of the right arm of chromosome III or the loss of the whole chromosome III [43]. The frequency of these events is strain-dependent and could be increased after DNA-damaging treatment [43, 33]. In our case, exposure of the wild-type strain to UV light increased the frequency of chromosome arm loss (approximately two-fold) and had no influence on the frequency of chromosome loss (Table 4).

Table 4.

Frequency of genetic events that led to illegitimate hybridization of the wild-type strain K5-99-35B-D924, × 10−6.

UV, J/m2 Chromosome III loss Chromosome III arm loss Conversion Recombination Mutations and primary lesions Hybrids analyzed
0 160 (124 – 230) 9 (7 – 13) 2 (1 – 3) 1.0 (0.5 – 1.5) 4 (3 – 6) 1362
25 160 (98 – 190) 22 (13 – 25) 3 (2 – 4) 3 (2 – 4) 160 (100 – 190) 1325
Open in a new tab

All frequencies are medians for 21 cultures with 95% confidence intervals in parentheses.

We observed that inactivation of Polζ and Polη does not induce significant spontaneous chromosome loss (Fig. 3(A)). After UV irradiation, the frequency of chromosome loss was 4.5-fold higher in strains with single REV3, REV7 and RAD30 deletions than in the wild type strain. Simultaneous deletion of REV3 and RAD30 leads to a more dramatic (approximately 15-fold) increase in the frequency of UV-induced chromosome loss (Fig. 3(A)) in comparison to a UV-treated wild-type strain. We also investigated the effect of TLS polymerases inactivation on the frequency of spontaneous and UV-induced chromosome arm loss. The REV3 and REV7 genes deletions led to an increase in chromosome arm loss, both spontaneously and after UV treatment (Fig. 3(B)). While inactivation of the catalytic and regulatory subunits of Polζ had a similar effect on the frequency of spontaneous chromosome arm loss (two-fold higher than in the wild-type strain), the REV3 deletion had a stronger effect on the chromosome arm loss after UV light treatment compared to the REV7 deletion (10- and three-fold, respectively) (Fig. 3(B)). The frequency of chromosome arm loss was elevated 6.5-fold in the UV-treated rev3Δ rad30Δ strain as well (Fig. 3(B)).

Inactivation of the REV1 gene led to a 20- and 40-fold decrease in the spontaneous frequency of chromosome and chromosome arm loss, respectively. The UV-induced frequency of both events was decreased 30-fold in comparison to the UV-treated wild-type strain.

Effect of TLS polymerase inactivation on gene conversion and reciprocal recombination

Polζ and Polη were suggested to participate in DNA synthesis during recombinational events in yeast and in higher eukaryotes [30, 31, 28, 50]. In the selective system of illegitimate hybridization, it is possible to determine the frequency of gene conversion and reciprocal recombination events that triggered the hybridization. Here we examined the effect of Polζ, Polη and Rev1 inactivation on these events in the α-test.

Inactivation of the REV7 gene slightly increased the frequency of spontaneous gene conversion and reciprocal recombination (Fig. 4). In contrast, the REV3 deletion had a negligible effect on the gene conversion and, in fact, decreased the reciprocal recombination almost two-fold (Fig. 4). We did not observe any significant effect of Polη inactivation on the frequency of spontaneous gene conversion or reciprocal recombination. Simultaneous inactivation of Polη and Polζ, however, decreased the frequency of spontaneous reciprocal recombination 3.5-fold (Fig. 4(B). Both conversion and reciprocal recombination are known to be induced by certain types of DNA damage. We observed that UV irradiation did not significantly affect the frequency of gene conversion, but increased the frequency of reciprocal recombination three-fold (Table 3). The frequency of gene conversion and reciprocal recombination events, however, is changed in some TLS DNA polymerase mutants. Inactivation of either subunit of Polζ increased the UV-induced gene conversion frequency approximately seven-fold (Fig. 4(A)), but only REV7 deletion increased the reciprocal recombination frequency after UV irradiation (Fig. 4(B)). Inactivation of Polη showed no effect on the frequency of UV-induced gene conversion, but inactivation of the RAD30 gene led to the eight-fold increase in UV-induced reciprocal recombination frequency (Fig. 4(A, B)). The frequency of UV-induced gene conversion and reciprocal recombination in the double rev3Δ rad30Δ mutants was elevated 20- and three-fold, respectively (Fig. 4(A, B)).

Inactivation of Rev1 dramatically decreased of both spontaneous and UV-induced frequencies of gene conversion and reciprocal recombination (Fig. 4).

Ratio of inherited and non-inherited changes of genetic material in strains with rev3Δ, rev7Δ, rad30Δ, and rev3Δrad30Δ deletions

Primary lesions in DNA that triggered illegitimate mating can either be accurately repaired or transformed into mutations. While mutations and non-inherited lesions in the MAT locus of tested strains could not be distinguished in illegitimate diploids, this could be done in the selective system of cytoduction by analysis of cytoductant classes. The cytoductants preserving their original mating type α are thought to result from the phenotypic expression of primary lesions that disappeared after mating due to correct repair.

We investigated the effect of TLS inactivation on the ratio of inherited (gene mutations and conversion) and non-inherited (primary lesions) changes in the genetic material using the selective system of cytoduction. While inactivation of Polη (rad30Δ) had only a slight effect on this ratio, the inactivation of Polζ (rev3Δ or rev7Δ) greatly increased the proportion of UV-induced primary (non-inherited) lesions in the MAT locus (Fig. 5). The effect of the double rev3Δrad30Δ mutation was not different from the single rev3Δ and rev7Δ mutations. Thus, the percent of non-inherited changes in genetic material is much higher in the absence of Polζ, consistent with the previously established role of this polymerase in mutagenic TLS [23].

Figure 5.

Figure 5

Open in a new tab

Percentage of primary lesions that were transformed to mutations (inherited changes) and precisely repaired (non-inherited changes) in TLS DNA polymerase mutants.

DISSCUSION

Translesion synthesis is thought to be a highly mutagenic process that helps cells to tolerate replication-blocking damage even at the expense of decreased accuracy of replication. Because of its ability to register different types of genetic events, the test-system used in this study allowed us to form a comprehensive view of the role of TLS in the genome stability maintenance. We were able to estimate the ratio of accurately repaired primary lesions to lesions that caused gene mutations, recombination or chromosome rearrangements both in the wild-type strain and mutants carrying different TLS defects. We provide evidence that TLS protects yeast cells from global chromosome rearrangements and chromosome loss in the presence of genotoxic factors.

According to the current concept of illegitimate αxα mating in yeast, it may occur when the transcription in the MATα locus of one of the copulating cells is repressed because of the presence of a lesion in it. The cell switches the mating type to the opposite (reversibly or irreversibly) and becomes able to copulate with another α cell. The lesion then can be repaired precisely after the completion of mating or become an inherited change if it is transformed into a mutation. In other cases, unrepaired primary lesions may provoke a double-strand break formation. Recombinational repair could rescue cells from such toxic double-strand breaks, but it would lead to genetic changes (gene conversion or large chromosomal deletions due to reciprocal recombination in our case). The double-strand breaks could also lead to chromosome or chromosome arm loss.

Investigation of TLS polymerases mutants revealed that the REV3, REV7 and RAD30 deficiency leads to an increased frequency of illegitimate hybridization of UV-irradiated cells. According to the current mating type switching model, this suggests that the inactivation of TLS polymerases increases the frequency of events that repress transcription of the MATα. An interesting possibility is that, in the absence of TLS, lesions may exist longer in DNA impairing the MAT locus transcription. Although TLS proteins are not expected to participate in lesion removal, it is conceivable that TLS deficiency increases the time, during which lesions remain in the single-stranded DNA region where they can not be repaired. In addition to blocking transcription, lesions in DNA could result in transcriptional mutagenesis due to inaccurate lesion bypass by the RNA polymerase [51, 52]. Transcriptional mutagenesis could potentially contribute to the mating type switching in yeast. In addition to the lesions themselves, stalled replication complexes could potentially impede the MATα locus transcription. Replicative DNA polymerases are blocked by DNA lesions [53]. It is conceivable that the number or the lifetime of the stalled replication complexes could increase if they are not acted upon by TLS in a timely manner. The stalled replication forks could significantly disturb the progression of transcription complexes. In the α-test, the double rev3Δ rad30Δ mutants show an additive increase in the frequency of illegitimate mating in comparison to the single mutant effects. This is consistent with the earlier data, suggesting the participation of Polζ and Polη in the bypass of different types of lesions [7].

According to one of the current TLS models (the gap-filling model), the replicative polymerase stalling at a lesion leads to a quick reinitiation of replication downstream of the lesion [54]. TLS polymerases are then recruited to bypass the lesion and to fill the gap remaining between the lesion and the site of replication reinitiation. Thus, the absence of TLS polymerases could potentially result in the persistence of single-stranded DNA regions. The increase in UV-induced chromosome loss and chromosome arm loss observed upon TLS polymerase inactivation (Fig. 3) may be a consequence of the accumulation of single-stranded gaps in DNA and their conversion to double-strand breaks. In addition, regions of single-stranded DNA are highly susceptible to spontaneous base damage [55, 56]. If such regions, indeed, accumulate in the TLS mutants, the additional DNA damage that they suffer could contribute to the increased frequency of the MATα inactivation and illegitimate mating. Cooperation of two TLS DNA polymerases helps cells to handle different DNA lesions that may lead to the accumulation of single-stranded DNA and double-strand breaks and, thus, chromosome fragility and loss.

In addition to increasing the UV-induced chromosome and chromosome arm loss, the inactivation of either subunit of Polζ increased the frequency of spontaneous chromosome arm loss (Fig. 3(B)). We propose that spontaneous chromosome arm loss occurs in the Polζ deficient cells due to the events similar to those occurring after UV irradiation. Polζ is well known to participate in the bypass of endogenously generated DNA lesions [57–64]. It also can contribute to the replication of undamaged DNA when the fork progression is impeded for reasons other than DNA lesions [65, 66] It is, therefore, likely that occasional re-priming of DNA synthesis and the accumulation of single-stranded gaps are also features of the normal DNA replication. As in the case of UV irradiation, the absence of Polζ could result in an increased accumulation of single-stranded DNA, double-strand breaks and, subsequently, increased chromosome arm loss. Replication fork impediment does not only provoke chromosome aberrations and chromosome loss due to the formation of double-strand breaks. Chromosome rearrangements can also result from homologous recombination repair of DSBs [67]. The increased frequency of spontaneous and UV-induced gene conversion in the Polζ deficient strains is consistent with the idea that Polζ participates in the avoidance of double-strand breaks in the DNA. Interestingly, the inactivation of Polζ and Polη had quite different effects on the frequency of UV-induced gene conversion and reciprocal recombination. This might reflect the participation of these TLS polymerases in the bypass of different types of lesions, and the processing of these lesions by distinct recombination pathways in the absence of these polymerases.

In several cases, we observed that the inactivation of catalytic (Rev3) and accessory (Rev7) subunits of Polζ had different effects on the genome stability. The frequency of spontaneous gene conversion and both spontaneous and UV-induced reciprocal recombination were significantly higher in the rev7Δ mutants in comparison to rev3Δ mutants (Fig. 4). At the same time, inactivation of Rev3 conferred a much stronger increase in the frequency of chromosome arm loss than Rev7 inactivation (Fig. 3(B)). This observation may suggest that Rev3 and Rev7 could function independent of each other. The Rev7 homolog in higher eukaryotes is involved in chromosome segregation control [26, 27]. The unique function of Rev7 observed in our study is unlikely to be related to chromosome segregation, because the frequency of chromosome loss is equally increased in the rev3Δ and rev7Δ mutants (Fig. 3A).

Interestingly, we observed that the inactivation of Rev1 impedes illegitimate copulation of yeast cells. We believe that this impediment could be related to the cell cycle-specific regulation of REV1 expression. It was shown recently that, unlike other TLS polymerases, Rev1 is preferentially produced in the G2 phase of the cell cycle [68]. This suggests that Rev1 may be required for damage tolerance pathways specific to the G2 phase. The Rev1-deficient cells could possibly have an altered cell cycle, which could affect their ability to copulate. In support of this idea, Jansen et al. reported that Rev1− chicken cells are characterized by a prolonged G2 phase of the cell cycle [69].

Numerous genetic syndromes, cancer and autoimmune diseases are associated with decreased fidelity of DNA replication. One of the approaches that are being developed for the prevention and treatment of such diseases is the inhibition of uncontrolled DNA replication and selective inhibition of mutagenic DNA polymerases, such as TLS polymerases [70]. TLS is a major source of spontaneous and, particularly, genotoxicant-induced point mutations that could contribute to cancer development. This view is supported by a recent study demonstrating that the inactivation of Rev1 results in a significant reduction in the incidence of tumors in mice exposed to benzo[a]pyrene [71]. The potential of using Polζ as a target for gastric and colorectal cancer prevention and therapy has also been discussed [72]. On the other hand, the REV3 deficiency has been previously found to induce chromosome instability in chicken, mice and mammalian cells [28, 29, 73]. The present study further illustrates that the function of TLS polymerases is important for limiting chromosome instability in yeast, suggesting that this function is widespread among eukaryotes. Consistent with the important role of Pol ζ in the prevention of large genome rearrangements, REV3L−/− mice have recently been found to have increased susceptibility to spontaneous tumorigenesis [73]. The selective inhibition of Polζ, thus, may not be an innocuous approach for cancer prevention and therapy because of the potential to provoke chromosome rearrangements. While it might be efficient for suppressing certain types of cancer in which the accumulation of point mutation plays a major role, the undesired side effect of such a treatment is likely to be the increase in other types of genomic instability and, subsequently, other types of cancer.

Acknowledgments

We thank Dr. Youri I. Pavlov for helpful discussion. This work was supported, in part, by NATO grant CBP.NR.NRCLG 982734, NIH grants ES011644 and ES015869 to P. V. S. and by the RFBR grant 09-04-13778 - ofi-c and RAS Presidium Program “Biodiversity and Dynamic of Genetic Pools” grant to S. G. I.-V.

References

RESOURCES