The spectrum of spontaneous mutations caused by deficiency in proteasome maturase Ump1 in Saccharomyces cerevisiae
- First Online:
- Cite this article as:
- McIntyre, J., Baranowska, H., Skoneczna, A. et al. Curr Genet (2007) 52: 221. doi:10.1007/s00294-007-0156-8
- 86 Views
Ump1 is responsible for maturation of the catalytic core of the 26S proteasome. Dysfunction of Ump1 causes an increase in the frequency of spontaneous mutations in Saccharomyces cerevisiae. In this study we analyze the spectrum of mutations occurring spontaneously in yeast deficient in Ump1 by use of the SUP4-o system. Single base substitutions predominate among the mutations analyzed (73 of the 91 alterations examined). Two major classes are GC to TA transversions and GC to AT transitions (∼50 and ∼30% of base substitutions, respectively). Besides base substitutions, almost all the major types of sequence alterations are represented. The specificity and distribution of mutations occurring in the ump1 strain are unique compared to the spectra previously established for other yeast mutators. However, the profile of mutations arising in this strain is similar to that observed in wild type. The same similarity has previously been reported for yeast deficient in Mms2, a protein involved in Rad6-dependent postreplication DNA repair (PRR). The specificity of the mutator effect caused by ump1 is discussed in light of the proposed role of the proteasome activity in the regulation of the PRR mechanisms.
KeywordsSpontaneous mutagenesisProteasomePostreplication DNA repairMms2Polymerase eta
Ump1 is a small chaperone-like protein functioning in the assembly of the 26S proteasome (Ramos et al. 1998), the major eukaryotic multi-proteinase consisting of a catalytic core, the 20S proteasome, and one or two regulatory caps, the 19S complexes (Coux et al. 1996; Hershko and Ciechanover 1998). Proteins destined for proteasomal degradation are covalently tagged with chains of ubiquitin in a tightly controlled multi-step reaction. The 19S regulatory cap is responsible for the recognition of tagged proteins and their conversion into a form competent for degradation. Proteolysis proceeds inside the 20S proteasome, a cylindrical structure built of different but related subunits of α- and β-type, which form four heptameric rings (Groll et al. 1997). In the 20S proteasome assembly process, the proteasome maturase Ump1 facilitates conversion of double-ring 16S precursor forms into the correct four-ring structure. Dysfunction of Ump1 leads to defective proteasome maturation, resulting in decreased activity of proteasomal peptidases and reduction in the overall level of ubiquitin-mediated proteolysis (Ramos et al. 1998).
It has previously been shown that defects in proteasome maturase cause an increase in the level of spontaneous mutagenesis, suggesting that among its many regulatory functions, the proteasome also plays a role in the maintenance of genetic stability (Podlaska et al. 2003). This suggestion was further supported by the description of mutator phenotypes of yeast strains deficient in proteasomal peptidases Pup1 or Pre2 (McIntyre et al. 2006). Mutations affecting proteasome activities also resulted in sensitivity to UV radiation (Podlaska et al. 2003). A number of studies have linked proteasomal activity to nucleotide excision repair (NER) and repair of DNA double-strand breaks (DSBs) (Schauber et al. 1998; Marston et al. 1999; Moynahan et al. 2001; Gillette et al. 2001; Ng et al. 2003; Ramsey et al. 2004; Sone et al. 2004). However, an analysis of genetic interactions between mutations affecting proteasome activity and homologous recombination or NER has not confirmed the requirement for proteasome activity in recombination- or excision repair-mediated protection against spontaneous mutagenesis (McIntyre et al. 2006). Instead, the results suggested a role for proteasome activity in Rad6/Rad18-governed postreplication DNA repair (PRR) mechanisms. The Rad6/Rad18 complex monoubiquitinates lysine164 in the DNA replication processivity factor PCNA in response to agents that block progression of the DNA replication fork (Hoege et al. 2002). Monoubiquitination of PCNA facilitates the recruitment of specialized non-replicative DNA polymerases to the stalled replication complex (Stelter and Ulrich 2003; Kannouche et al. 2004; Garg and Burgers 2005). This results in translesion DNA synthesis (TLS) which allows replication to continue despite the DNA damage. In yeast, the TLS polymerases Pol eta and Pol zeta are responsible for error-free or error-prone DNA damage bypass (McDonald et al. 1997; Broomfield et al. 2001).
The monoubiquitinated PCNA can be further modified by a chain of ubiquitins bound via non-canonical lysine63, in a reaction dependent on the Ubc13/Mms2/Rad5 complex (Hofmann and Pickart 1999; Hoege et al. 2002). The polyubiquitination of PCNA promotes the error-free DNA-damage avoidance pathway. This recombination-like pathway of PRR probably employs a transitory template switch by the replicative DNA polymerase delta, to continue replication without DNA-damage removal (Broomfield et al. 1998, 2001).
The mutator effect caused by Ump1 dysfunction is dependent on Rad6/Rad18 as well as on the activities of Pol zeta and Pol eta (McIntyre et al. 2006). It has also recently been shown that the ump1 mutation causes stabilization of Pol eta (Skoneczna et al. 2007), suggesting that increased cellular activity of this DNA polymerase may be responsible for the mutator phenotype connected with proteasome deficiency.
In order to further characterize the mutator phenotype caused by the Ump1 proteasome maturase deficiency we have established the spectrum of mutations occurring spontaneously in the ump1 strain. The detailed spectrum turned out to be unique compared to the spectra found in other mutator yeast strains. However, the mutation profile resembles that established previously for the yeast strain defective in Mms2. These results suggest that in ump1, as in the mms2 strain, the processing of preexisting DNA lesions switches from an error-free to a mutagenic processing. We also show, however, that the mechanisms of error-prone lesion processing caused by Ump1 or Mms2 deficiency differ in their requirements for the TLS polymerases.
Materials and methods
Haploid Saccharomyces cerevisiae strains used in this work were constructed by standard gene disruption, performed by direct transformation of yeast cells with a PCR product representing the desired disruption cassette. Deletions of the appropriate genes were verified by PCR. The derivatives of strain WCG4-a (Heinemayer et al. 1997): YAS13 (ump1::kanMX), YJM4 (rad30::HIS3) and YJM5 (ump1::kanMX rad30::HIS3) have been described previously (Podlaska et al. 2003). Isogenic strains carrying the mms2::kanMX4 (YJM56) and mms2::kanMX4 rad30::HIS3 (YJM57) mutations were constructed using a disruption cassette, produced by PCR amplification of DNA isolated from BY4741 mms2 (mms2::kanMX4, Euroscarf) with primers: 5′ ATCTCTGCTCATTACATT and 5′ ATTATTGGCTTGGACTGG, introduced into WCG4-a and YJM4, respectively. ump1::kanMXmms2::kanMX4 (YHB20-2A) is a haploid segregant from a genetic cross of YAS13 (ump1::kanMX) with YHB57α (mms2::kanMX4). The desired progeny spores from tetrads showing 2:2 segregation with respect to G-418 resistance and temperature sensitivity were selected. The presence of ump1::kanMX and mms2::kanMX4 was confirmed by PCR. The construction of YAS6 (ump1::TRP1), a derivative of MKP-o (Pierce et al. 1987), was described by Podlaska et al. (2003).
Spontaneous mutagenesis assay
Yeast cultures were grown in minimal YNBG medium, supplemented as described previously (Mieczkowski et al. 2000), at 24°C, to mid-logarithmic growth phase (1–3 × 107 cfu/ml) and plated (in duplicate) on minimal YNBG omission media (lacking amino acids corresponding to the auxotrophic mutation being tested) or on CAN-plates supplemented with canavanine sulfate to 30 μg/ml. Plates were incubated at 24°C for 5–6 days before the mutant colonies were counted. Total colony-forming units were counted on non-selective minimal plates after 4–5 days of incubation at 24°C. In each experiment six to ten independent cultures of the tested yeast strain were analyzed. Mutation rates were calculated using the equation μ = f/ln(N × μ) (where μ is the mutation rate, f the mutant frequency and N the final cell population size) which was solved by iteration (Drake 1991). Fluctuation tests were repeated three to four times for each strain and experimental condition.
Spectrum of mutations
The spectrum of mutations occurring spontaneously was analyzed in YAS6, a derivative of MKP-o, defective in proteasome maturase Ump1, transformed with plasmid YCpMP2 carrying a functional SUP4-o tRNA gene (Pierce et al. 1987; Kunz et al. 1991; Roche et al. 1994). SUP4-o suppresses three ochre mutations (can1-100, ade2-1 and lys2-1) present in MKP-o derivatives. The nature of mutations in this gene was determined by sequencing the SUP4-o locus on YCpMP2 isolated from mutant colonies, which displayed a pink color (Ade−), were resistant to canavanine (CanR), and were unable to grow on plates lacking lysine (Lys−). To permit a direct comparison of the obtained results with those previously reported for other DNA repair mutants, the strategy for mutant isolation, growth conditions and procedures for determining plasmid retention and mutation rate were exactly as described in the protocol kindly provided by Dr. B. A. Kunz (Pierce et al. 1987; Kunz et al. 1991; Roche et al. 1994). YCpPM2 plasmids carrying mutant SUP4-o alleles were isolated using the glass bead method and introduced into Escherichia coli by transformation. Plasmid DNA was prepared from bacterial cultures by alkaline lysis and the SUP4-o alleles were sequenced. To statistically evaluate the differences in mutational classes and mutation distribution within the SUP4-o gene, between ump1 and the wt or other DNA repair mutants, a computer program (Cariello et al. 1994) available in the public domain was used and differences were considered significant at P < 0.05.
Spectrum of mutations arising spontaneously in yeast deficient in proteasome maturase Ump1
Profiles of spontaneous sup4-o mutations occurring in wild type and ump1 mutant
Percent of total eventsa
Percent of total events
Sequencing data for 121 mutations in SUP4-o selected in the control MKP-o strain are shown here for comparison (Table 1). In general, the profile of the main classes of mutations occurring in the ump1 mutant and the control wild-type strain did not differ significantly (P value > 0.5), suggesting that the rates of different types of mutational events increased proportionally as a result of proteasome dysfunction.
Specificity of single base substitutions occurring in the wild type and ump1mutant
Percent of total substitutionsa
Percent of total substitutions
Location of multiple mutations, deletions and insertions in the SUP4-o gene
6 → 7
11 → 12
17 ↔ 18
35 → 37
37 ↔ 38
43 ↔ 44
58 → 64
69 ↔ 70
79 → 83
84 ↔ 86
88 → 96
89 → 96
Interestingly, insertions of the yeast Ty transposon into SUP4-o were found at a rate over fourfold higher in ump1 than in the control strain. Almost 80% of the Ty elements in the control strain occurred at position 38 which has been previously reported as a hotspot for Ty transposition (Giroux et al. 1988; Kunz et al. 1990, 1991). In the Ump1-deficient strain the Ty insertions were more scattered throughout the sequence. However, a third of them were also found at the same hotspot, at a rate almost twice that of the wild type. In the ump1 strain two mutations in SUP4-o consisted of single base-pair insertions and one mutation classified as a complex event was identified. The locations of these mutations are listed in Table 3.
Different requirement for Pol eta in ump1- and mms2-mediated mutagenesis
Although the detailed specificity of spontaneous mutations occurring in the ump1 strain was unique, the profile of the mutation types occurring in this strain was similar to that established for the wild type. This mutational profile was also statistically similar (P > 0.7) to that established previously for the mms2 mutant (Xiao et al. 1999). Mutations inactivating MMS2 cause a defect in the PCNA polyubiquitination-dependent branch of Rad6-mediated PRR (Broomfield et al. 2001; Hoege et al. 2002).
Another approach to explore the relationship between the ump1- and mms2-mediated mutagenic effects is to compare the genetic requirements of these processes. It has previously been shown that mutagenesis in yeast deficient in the Ump1 proteasomal maturase (McIntyre et al. 2006), and in yeast deprived of Mms2 (Xiao et al. 1999), requires the REV3 gene, encoding the catalytic subunit of Pol zeta. Recently, it has also been shown that the mutagenic effect caused by Ump1 deficiency is largely dependent on another TLS polymerase, Pol eta, encoded by RAD30 (McIntyre et al. 2006). We verified this result and established the rates of mutations occurring per generation in ump1 vs. ump1 rad30. In agreement with our previous findings, the rate of mutations caused by the defect in Ump1 was reduced in cells that were also deprived of Pol eta (Fig. 2). However, in contrast to the ump1-mediated mutator effect, mutagenesis in the Mms2-deficient strain did not require the presence of Pol eta and the mutation rate in mms2 rad30 was in fact slightly higher than that established for the mms2 single mutant (Fig. 2). This result indicated that despite similarities in the mutational profiles, different mechanisms are responsible for the occurrence of mutations in yeast deficient in the proteasomal maturase Ump1 and Mms2.
Spontaneous mutations occur as the result of error-prone processing of preexisting DNA lesions and/or infidelity of the replicating polymerase. It is possible that an increase in the spontaneous mutation frequency not accompanied by a change in the profile of mutational classes may reflect a shift in the balance between error-free and error-prone processing of preexisting DNA damage. In accordance with this speculation, the only defect, reported to date, that results in an increased level of spontaneous mutagenesis without meaningful change in the mutational profile, is the mms2 mutation (Xiao et al. 1999). We have previously shown that UMP1, like MMS2, belongs to the RAD6/RAD18 epistasis group of genes engaged in PRR (Podlaska et al. 2003; McIntyre et al. 2006). The mutational spectrum identified in this study indicates that the relative contribution of various classes of mutations to the mutator effect caused by ump1 is statistically similar to the profile of mutations occurring in wt and to that previously established for the mms2 strain (Xiao et al. 1999). This similarity suggests that, by analogy to the situation in the mms2 mutant, the error-free avoidance pathway in ump1 may be partially compromised, and the mutator phenotype caused by the defect in proteasome maturation reflects increased processing of endogenous DNA damage by the error-prone TLS pathway, as has been postulated for mms2.
Interestingly, however, the similarities in mutational profiles between strains deficient in Mms2 and the proteasomal maturase are accompanied by differences in the specificity and distribution of base substitutions in SUP4-o, occurring spontaneously in these mutants. These differences might reflect the different impact of individual DNA polymerases on the error-prone processing of DNA lesions in mms2 and ump1 mutants. The mutator effect conferred by the mms2 mutation is almost entirely dependent on Pol zeta (Broomfield et al. 1998; Xiao et al. 1999). The ump1-mediated spontaneous mutagenesis requires Pol eta in addition to Pol zeta (McIntyre et al. 2006). Here we have confirmed this requirement and also showed that, in contrast to ump1-mediated mutagenesis, the mutator phenotype caused by Mms2 deficiency does not require Pol eta. The finding that endogenous DNA lesions are processed by different mechanisms in the ump1 and mms2 mutants implies that the ump1-mediated mutator effect is not initiated by a defect in the error-free pathway, as it is in the mms2 mutant. The apparent bias against the Mms2-dependent error-free pathway in the ump1 mutant may be explained by promotion of error-prone processing of DNA damage, which can compete with the error-free repair pathway and in consequence cause its limitation. Consistent with this interpretation, it has recently been shown that the low fidelity DNA polymerase eta is a short-lived, proteasomally degraded protein and that the ump1 mutation raises the cellular level of this TLS polymerase in yeast (Skoneczna et al. 2007). In the process of recruitment of Pol eta to the stalled replication fork, the polymerase interacts with PCNA monoubiquitinated at lysine164 (Kannouche et al. 2004) and the same ubiquitinated lysine is the target for polyubiquitination by the Rad5/Ubc13/Mms2 complex (Hoege et al. 2002). Since this polyubiquitination activates the Mms2-dependent error-free DNA damage avoidance pathway, an increased level of Pol eta is likely to interfere with the function of this pathway by competition between Pol eta and Rad5/Ubc13/Mms2 complex for monoubiquitinated PCNA.
It remains to be established whether the increased level of Pol eta in the ump1 mutant results in the direct involvement of this low fidelity polymerase in replication of intact DNA or in TLS, or if the mutator phenotype is only an indirect effect of perturbations in the replication fork caused by Pol eta. The specificity of base substitutions in ump1, among which mutations at GC base pairs predominate, argues against the concept that mutagenesis in ump1 is the consequence of direct replication errors caused by Pol eta, since a strong bias towards AT base pair mutations was observed when DNA was replicated by mammalian Pol eta (Matsuda et al. 2000). However, the increased involvement of Pol eta in error-prone TLS cannot be excluded, since the specificity of mutations would then be determined by the nature of the DNA lesions processed. Error-prone TLS as the source of ump1-mediated mutagenesis would be consistent with the requirement in this process for PCNA ubiquitination and Pol zeta (McIntyre et al. 2006), which has been proposed to assist polymerase eta in the bypass of some DNA lesions (Prakash et al. 2005).
On the other hand, a mutator effect with a predominance of mutations at GC base pairs was observed in yeast overexpressing the RAD30 gene, encoding Pol eta, in the msh6 (mismatch repair deficient) background (Pavlov et al. 2002). However, the same mutator effect was observed irrespective of whether wt RAD30 or an allele encoding Pol eta mutated at the active center of the polymerase was overexpressed. This finding supports the hypothesis that an excess of Pol eta perturbs the replication fork and causes mutations by an as yet unknown mechanism. It is important to note, however, that in the experiments of Pavlov et al. (2002), Pol eta was massively overproduced, while in the ump1 mutants the cellular content of Pol eta was increased by only two–fourfold (Skoneczna et al. 2007). As has recently been shown for other nonessential DNA polymerases (Pillaire et al. 2007), the consequences of up-regulation of error-prone polymerases differ, depending on the extent of the increase in cellular levels.
Regardless of whether or not the polymerizing activity of Pol eta is directly involved in mutagenesis caused by Ump1-deficiency, the spectrum of mutations reported in this study confirms the role of the proteasome in controlling the equilibrium between error-free and error-prone processing of endogenous DNA damage by PRR.
We thank Dr. Bernard Kunz for providing the SUP4-o system and the experimental protocol for determination of the spectrum of mutagenesis. This work was supported by grant N303 040 31/1306 from the Polish Ministry of Education and Science.