|Home | About | Journals | Submit | Contact Us | Français|
A genome-wide screen in Saccharomyces cerevisiae identified LSM1 as a new gene affecting sensitivity to ultraviolet (UV) radiation. Lsm1p is a member of a cytoplasmic complex composed of Lsm1p–7p that interacts with the yeast mRNA degradation machinery. To investigate the potential role of Lsm1p in response to UV radiation, we constructed double mutant strains in which LSM1 was deleted in combination with a representative gene from each of three known yeast DNA repair pathways. Our results show that lsm1Δ increases the UV-radiation sensitivity of the rad1Δ and rad51Δ mutants, but not the rad18Δ mutant, placing LSM1 within the post-replication repair/damage tolerance pathway (PRR). When combined with other deletions affecting PRR, lsm1Δ increases the UV-radiation sensitivity of the rev3Δ, rad30Δ and pol30-K164R mutants but not rad5Δ. Furthermore, the UV-radiation sensitivity phenotype of lsm1Δ is partially rescued by mutations in genes involved in 3′ to 5′ mRNA degradation, and mutations predicted to function in RNA interactions confer the most UV-radiation sensitivity. Together, these results suggest that Lsm1p may confer protection against UV-radiation damage by protecting the 3′ ends of mRNAs from exosome-dependent 3′ to 5′ degradation as part of a novel RAD5-mediated, PCNA-K164 ubiquitylation-independent subpathway of PRR.
Using a genome-wide screen in Saccharomyces cerevisiae, we have previously identified LSM1 as a new gene affecting sensitivity to UV radiation (1). Lsm1p is part of a protein complex, Lsm1p–7p/Pat1p, that is involved in the regulation of mRNA degradation (2). In yeast, polyadenylated mRNAs are degraded by two general pathways, both of which require shortening of the 3′ end poly(A) tail (deadenylation). In the major pathway, deadenylation is followed by removal of the 5′ cap structure and subsequent 5′ to 3′ exonucleolytic degradation (3–6). In the minor pathway, deadenylation is directly followed by 3′ to 5′ exonucleolytic degradation mediated by the exosome complex (6–10).
The Lsm1p–7p/Pat1p complex localizes to discrete cytoplasmic structures called P-bodies where the 5′ to 3′ mRNA degradation process occurs (11, 12). The complex has been implicated in various mRNA degradation functions, including facilitating the decapping step of mRNA degradation (13, 14) as well as protecting the 3′ ends of mRNAs from partial degradation (12, 15). Consistent with our finding that deletion of LSM1 causes sensitivity to UV radiation is that other proteins involved in mRNA turnover may play a role in response to UV radiation. For example, deletion of DHH1, a decapping activator that interacts with Lsm1p (14), causes decreased survival after UV irradiation (16). Deletion of PAT1, a gene encoding a protein that associates with the Lsm1p–7p complex (13), also results in decreased survival after UV irradiation (17). The role of DHH1 in the UV-radiation damage response has been linked to recovery dependent on the G1/S DNA damage checkpoint (16); however, the mechanism of action by which LSM1 confers protection against UV radiation is currently unknown.
Because the human ortholog of LSM1 was reported to play a possible role in carcinogenesis (18, 19), LSM1 is an attractive gene for investigation in view of the known relationship between sensitivity to DNA-damaging agents and cancer (20). Genes that affect cell sensitivity to killing by UV radiation have classically been assigned to three major repair groups, each controlling a different type of DNA repair (21). The RAD3 group mediates nucleotide excision repair (NER), a mechanism by which UV-radiation-induced thymine dimers, photoproducts and other bulky lesions are repaired (22). Mutants in this pathway are highly sensitive to UV radiation. The RAD51 group mediates recombination repair, a mechanism by which DNA double-strand breaks and other forms of lesions are repaired using a homologous template (23). Mutants in this pathway are highly sensitive to ionizing radiation, and some are mildly sensitive to UV radiation. The RAD6 group, the most complex and least characterized pathway, allows replication through UV-radiation lesions by mutagenic translesion synthesis, error-free translesion synthesis, and postreplication repair of discontinuities (24). Mutants in this pathway show variable sensitivity to many different DNA-damaging agents, including UV radiation. Strains carrying mutations in two genes within the same repair group show UV-radiation sensitivity no greater than that of either single mutant (and are therefore in the same epistasis group), whereas strains carrying mutations in two genes in different groups show UV-radiation sensitivity greater than that of either single mutant.
In this study, we have used epistasis analysis to address the role of LSM1 in response to UV radiation. Genetic analysis shows that LSM1 is in the same epistasis group as RAD18 and is specifically placed in a novel RAD5-dependent subpathway of PRR that does not require PCNA-K164 ubiquitylation. We also demonstrate that protection against UV-radiation damage is conferred by the whole Lsm1p–7p/Pat1p complex and is mediated via predicted RNA contact residues of Lsm1p, and the UV-radiation sensitivity phenotype of lsm1Δ is rescued by mutations in genes required for 3′ to 5′ mRNA degradation. Based on these results, we propose a model in which the Lsm1p–7p/Pat1p complex binds to the 3′ ends of transcripts involved in a novel RAD5-mediated, PCNA-K164 ubiquitylation-independent subpathway and protects them from exosome-mediated 3′ to 5′ degradation.
The S. cerevisiae strains used in this study are listed in Table 1. Yeast media were prepared according to standard protocols (25). For nonselective growth, cells were grown in YEP-rich medium consisting of 2% glucose, 1% bactopeptone, and 0.5% yeast extract. For selective growth, cells were grown in synthetic medium lacking uracil. pol30-K164R lsm1ΔKanMX4 was constructed by transforming an lsm1ΔKanMX4 fragment into haploid pol30-K164R and selecting for G418-resistant clones. lsm1ΔKanMX4 disruption fragment was made by PCR amplification of the gene locus using template genomic DNA from YJL124C and LSM1A, LSM1D primers taken from the Saccharomyces Genome Deletion Project (http://sequence-www.stanford.edu/group/yeast_deletion_project/deletions3.html).
Plasmids pST11 (LSM1), pST26 (lsm1-6), pST33 (lsm1-13), pST28 (lsm1-8), pST29 (lsm1-9), pST34 (lsm1-14), pST25 (lsm1-5), pST21 (lsm1-1), pST36 (lsm1-16), and pST45 (lsm1-25) have been described (26). Plasmids pGal and pGal-RAD5 and pRP1000 (pSKI4) have been described (10, 27).
Genetic crosses, sporulation and tetrad dissection were performed according to standard protocols (28). The genotype of each constructed strain was confirmed by PCR with gene-specific primers. The sequences of gene-specific primers A and D for each locus were designed as instructed from the Saccharomyces Genome Deletion Project.
Cells in logarithmic growth (2.2 × 107 cells/ml) were serially diluted in sterile water, and different dilutions were plated on solid YEPD medium for different irradiations. Each plate was irradiated at the indicated dose using a Sankyo Denki UVC germicidal lamp giving most of its radiation at 254 nm at a rate of 1 J/m−2 or 10 J/m−2, depending on the distance of the plate from the source. Plates were incubated in the dark for 4 days, and survival was calculated by counting visible colonies. UV-radiation sensitivity of lsm1 point mutants was measured by spotting 5 µl of fivefold serial dilutions of cells onto plates containing synthetic medium lacking uracil. The plates were irradiated at 80 J/m2, and viability was assessed after 3 days of incubation at 30°C.
Total RNA was isolated from yeast cells with a ToTALLY RNA Isolation Kit (Ambion, Inc.) according to the manufacturer’s instructions. Total RNA was reverse transcribed into cDNA using a SuperScriptIII First-Strand Synthesis Kit with Oligo d(T)20 primers (Invitrogen, Carlsbad, CA) according to the manufacturer’s protocol. The final cDNA product was stored at −20°C until further analysis.
Quantitative real-time PCR was carried out in duplicate in two independent experiments using SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA) on a TaqMan Real-Time PCR Instrument (Applied Biosystems). A PCR reaction mixture of 20 µl contained 10 µl of 2× SYBR Green PCR Master Mix, 1 µl of 10 µM gene-specific forward and reverse primer, 6 µl cDNA (diluted 1:8), and 2 µl of water. The amplification program consisted of one cycle at 95°C for 2 min followed by 40 cycles of 95°C for 30 s and 55°C for 1 min. A negative control (water) was included in each run. The oligonucleotide sequences used for gene-specific detection are summarized in Table 2.
We previously identified lsm1Δ as sensitive to UV radiation (1) but with wild-type sensitivities to other damaging agents, including ionizing radiation (data not shown). To determine whether the UV-radiation sensitivity of lsm1Δ was due to a functional defect in any one of the three major yeast repair pathways, we crossed lsm1Δ with isogenic strains carrying a representative deletion in each repair group and compared the UV-radiation sensitivity of double mutants to that of the corresponding single mutants.
To analyze the relationship between LSM1 and nucleotide excision repair, we measured survival after UV irradiation in lsm1Δ, rad1Δ and lsm1Δ rad1Δ strains. The UV-radiation sensitivity of the double mutant was dramatically increased compared to the rad1Δ single mutant (Fig. 1A), suggesting that LSM1 affects a pathway other than nucleotide excision repair, and both pathways compete for a common substrate.
To analyze the relationship between LSM1 and recombination repair, we measured survival after UV irradiation in lsm1Δ, rad51Δ and lsm1Δ rad51Δ mutants. The UV-radiation sensitivity of the double mutant was significantly increased compared to the rad51Δ single mutant (Fig. 1B), suggesting that LSM1 affects a pathway other than recombination repair. Cells defective in recombination repair primarily exhibit a failure to repair double-strand breaks caused by agents such as ionizing radiation; thus this result is consistent with the finding that lsm1Δ is not hypersensitive to ionizing radiation (data not shown).
To analyze the relationship between LSM1 and postreplication damage tolerance, we measured survival after UV irradiation in lsm1Δ, rad18Δ and lsm1Δ rad18Δ mutants. Our results show that the UV-radiation sensitivity of the lsm1Δ rad18Δ mutant was not increased compared to the rad18Δ single mutant (Fig. 1C), suggesting that LSM1 affects the function of some aspect of the postreplication damage tolerance pathway. It is possible that the effect of lsm1Δ in the rad18Δ background was not detectable because of the extreme UV-radiation sensitivity of rad18Δ; however, there is clearly a difference between the rad1Δ lsm1Δ interaction and the rad18Δ lsm1Δ interaction at similar doses. The assignment of LSM1 to the RAD6 epistasis group is further supported by the finding that lsm1Δ rad6Δ is not any more sensitive than rad6Δ (data not shown).
Rad18p forms a heterodimer with Rad6p, and the Rad6p/Rad18p complex promotes replication through DNA lesions via three different subpathways: error-free translesion synthesis, mutagenic translesion synthesis, and Rad5-dependent postreplication repair (PRR) of discontinuities (24). Since deletion of RAD18 blocks the activity of all three subpathways (24), we sought to determine whether LSM1 affects any particular sub-branch of the RAD6 epistasis group. We examined the UV-radiation sensitivity of the lsm1Δ strain in combination with deletion of RAD30 (a gene affecting the error-free translesion synthesis subpathway), deletion of REV3 (a gene affecting the error-prone translesion synthesis subpathway), and deletion of RAD5 (a gene affecting the error-free postreplicative repair of discontinuities). Deletion of LSM1 enhanced the UV-radiation sensitivity of rev3Δ (Fig. 2C) and rad30Δ (Fig. 2B) but not rad5Δ (Fig. 2A). These results suggest that RAD5 is epistatic to LSM1, thus placing LSM1 in the RAD5-dependent subpathway of PRR.
Our yeast deletion pool study also identified pat1Δ as being in the top 100 strains sensitive to UV radiation (1). This finding suggests that Lsm1p does not act alone in conferring protection against UV radiation but functions as a member of the cytoplasmic Lsm1p–7p complex, which interacts with Pat1p. To test this hypothesis, we examined the UV-radiation sensitivity of pat1Δ, lsm6Δ and lsm7Δ strains, which are deleted in the nonessential genes of the complex. Figure 3 shows that all three mutants display similar UV-radiation sensitivities to lsm1Δ. To test our hypothesis further, we constructed the lsm1Δ pat1Δ double mutant and compared its UV-radiation sensitivity to corresponding single mutants. The double deletion strain is no more sensitive to UV radiation than each of the single deletion strains (Fig. 3). Together, these results suggest that Lsm1p confers protection against UV radiation as a member of the Lsm1p–7p/Pat1p complex and is consistent with the finding that lsm1 alleles defective in predicted inter-subunit contacts are also sensitive to UV radiation (Fig. 4A).
Recently, several lsm1 point mutants were generated whose mutated residues are predicted to be defective in RNA binding and inter-subunit contacts. Each of these mutants demonstrated a different degree of deficiency in mRNA decay and 3′ end mRNA protection (26). We tested the UV-radiation sensitivity of these mutants to gain better insight into which residues may be implicated in mediating protection against UV radiation and whether UV-radiation sensitivity correlates with the defect in 3′ end mRNA protection and mRNA decay. Our results show that lsm1 alleles with mutated RNA contact residues, lsm1-8 (pST28), lsm1-9 (pST29), and lsm1-14 (pST34), and alleles with mutated inter-subunit contacts residues were indeed sensitive to UV radiation (Fig. 4). All other lsm1 alleles that did not exhibit UV-radiation sensitivity (Fig. 4) also did not show any significant defects in mRNA metabolism (26). All together, the results suggest that the RNA binding property of Lsm1p as well as its ability to form a functional Lsm1p–7p complex are important for protection against UV-radiation-induced damage.
The temperature sensitivity phenotype of lsm1Δ was found to be suppressed by mutations in the exosome or the functionally related Ski proteins, which are required for efficient 3′ to 5′ mRNA degradation (15). To examine whether mutations in the exosome would also suppress the UV-radiation sensitivity phenotype of lsm1Δ, we assessed the UV-radiation survival of double mutants carrying lsm1Δ and a second lesion in SKI2, SKI3, SKI4 and SKI8 genes, all of which are required for 3′ to 5′ mRNA degradation (7, 10). Our results show that the ski4-1 mutation rescued the UV-radiation sensitivity phenotype of lsm1Δ to almost wild-type levels (Fig. 5A), ski2Δ rescued the phenotype slightly less effectively (Fig. 5B), and deletions in SKI3 and SKI8 were able to only partially suppress the UV-radiation sensitivity of lsm1Δ (data not shown). The ski4-1 allele is a point mutation in a core component of the exosome and has the strongest effect on mRNA turnover of all the ski mutations (10). This might explain why the ski4-1 mutant has the strongest effect on suppression of UV-radiation sensitivity whereas the other ski mutants show only partial suppression. To confirm that the suppression of the phenotype was indeed due to the absence of Ski4p function in lsm1Δ, we transformed a plasmid encoding the wild-type SKI4 gene into lsm1Δ ski4-1 and tested its survival after UV irradiation. Replacing the wild-type Ski4p in lsm1Δ ski4-1 resensitized the cells to UV radiation (Fig. 5C), thus confirming that Ski4p function is required to confer UV-radiation sensitivity in lsm1Δ cells. These results suggest that the UV-radiation sensitivity phenotype of lsm1Δ is at least partially due to 3′ to 5′ exosome-dependent degradation of unknown mRNAs and that the phenotype is alleviated by the absence of the exosome.
A simple interpretation of the above results is that the Lsm1p–7p complex binds to the 3′ ends of mRNAs and sterically inhibits their exosome-dependent 3′ to 5′ degradation, resulting in stabilization of the transcripts. It is currently not clear whether Lsm1p–7p complex can regulate a specific subset of mRNAs by binding to their 3′ ends, but it has been proposed that there may be some specificity for its substrates during growth at high temperatures (15). In view of the fact that LSM1 is in the same epistasis group as RAD18, we sought to determine whether Lsm1p–7p might affect mRNA levels of genes downstream of RAD18. To assess the transcript levels, we isolated total RNA from wild-type and lsm1Δ growing cultures, reverse transcribed RNA to cDNA, and performed real-time quantitative PCR. The Ct value was calculated for each transcript in both strains, normalized to the actin housekeeping gene control and expressed as a ratio (lsm1Δ/wild type). Figure 6 shows that the Rad18, Rad5 and Rev3 transcripts are not differentially expressed in lsm1Δ mutants, whereas Rad6 and Rad30 transcript levels are slightly elevated. These results indicate that transcripts from the RAD18 group do not appear to be preferentially degraded in the absence of Lsm1.
Given the assignment of LSM1 to the RAD5-mediated subpathway, we reasoned that overexpression of Rad5p and/or its downstream target PCNA might rescue the UV-radiation sensitivity phenotype of lsm1Δ. To test this hypothesis, we transformed wild-type, lsm1Δ and rad5Δ cells with a plasmid encoding Rad5p under a galactose-inducible promoter and a plasmid encoding PCNA under its native promoter. Cultures transformed with Gal-Rad5p plasmid were grown in nonrepressive raffinose medium or inducible galactose medium, plated on raffinose or galactose plates, and assessed for survival after UV irradiation. Our results show that overexpression of Rad5p rescued the UV-radiation sensitivity of rad5Δ whereas the empty plasmid did not, indicating that the plasmid encoding RAD5 was functional (Fig. 7B). Overexpression of Rad5p, however, did not rescue the UV-radiation sensitivity of lsm1Δ, suggesting that protein levels of Rad5p are likely sufficient in lsm1Δ. Overexpression of PCNA partially rescued the UV-radiation sensitivity phenotype of rad5Δ (Fig. 7A), indicating that high levels of mono-ubiquitylated PCNA at lysine 164 can compensate for the lack of PCNA multi-ubiquitylation in rad5Δ. It did not, however, rescue the UV-radiation sensitivity of lsm1Δ (Fig. 7A), indicating that a high level of PCNA is not able to compensate for the defect in lsm1Δ.
Since rad5Δ does not enhance the UV-radiation sensitivity of pol30-K164R (29) and overexpression of PCNA partially rescued the UV-radiation sensitivity of rad5Δ but not lsm1Δ (Fig. 7A), we wished to determine the effect of lsm1Δ on UV-radiation sensitivity in pol30-K164R background. Our results show that deletion of LSM1 enhanced the UV-radiation sensitivity of pol30-K164R (Fig. 8), suggesting that Lsm1p has a function independent of PCNA-K164 ubiquitylation and that Rad5p may mediate additional pathways independently of PCNA-K164 ubiquitylation in which the Lsm1p–7p/Pat1p complex participates.
In this study, we show that deletion of the LSM1 gene causes sensitivity to UV radiation, and several lines of evidence suggest that the UV-radiation sensitivity phenotype is linked to Lsm1p’s role in mRNA metabolism: (1) lsm6Δ, lsm7Δ and pat1Δ mutants, the non-essential members of the Lsm1p–7p/Pat1p complex, display similar UV-radiation sensitivity to lsm1Δ, and the double-deletion lsm1Δ pat1Δ strain has the same UV-radiation sensitivity as each single deletion strain (Fig. 3); (2) mutations in predicted RNA contact residues of Lsm1p correlate with increased UV-radiation sensitivity of these mutants. (Fig. 4); and (3) the UV-radiation sensitivity phenotype of lsm1Δ is rescued by inactivating the core component of the exosome required for 3′ to 5′ mRNA degradation. Taken together, these observations suggest that the role of Lsm1p in conferring protection against UV radiation requires the integrity of the Lsm1p–7p/Pat1p complex and is linked to some aspect of mRNA metabolism.
One model for the role of Lsm1p in mRNA decay proposes that after deadenylation, the Lsm1p–7p complex binds to the 3′ end of the deadenylated mRNAs and triggers decapping activation and subsequent degradation by the major 5′ to 3′ mRNA pathway. The binding to 3′ ends of mRNA could, however, also result in protecting these ends from the minor 3′ to 5′ mRNA degradation pathway (26). The Lsm1p–7p complex could therefore have a dual function in the mRNA turnover processes: for mRNAs undergoing 5′ to 3′ degradation, binding of the Lsm1p–7p complex to their 3′ ends would promote decapping and subsequent decay, whereas for mRNAs undergoing 3′ to 5′ degradation, binding of the Lsm1p–7p complex to their 3′ ends would result in transcript protection and stabilization. Consistent with this model of protection of the 3′ ends of mRNAs is the accumulation of several deadenylated mRNAs truncated at their 3′ ends by ~10 to 20 nucleotides in cells lacking any of the Lsm1p–7p/Pat1p complex components. The truncation of these mRNAs is more severe at high temperatures, and the temperature sensitivity of lsm1Δ is thought to be due to the increased susceptibility of a subset of truncated mRNAs essential during high temperatures to 3′ to 5′ exosome-mediated degradation (15). Analogously, the lack of Lsm1p–7p/Pat1p complex might increase the susceptibility of a subset of trimmed mRNAs essential during the UV-radiation response to 3′ to 5′ exosome-dependent degradation. This model is consistent with the finding that the UV-radiation sensitivity phenotype is suppressed by mutation in SKI4 (Fig. 5), the core component of the exosome complex required for efficient 3′ to 5′ mRNA degradation.
A recent study reported that deletion of DHH1 also causes sensitivity to UV-radiation damage (16). Dhh1p physically interacts with the Lsm1p–7p complex (14), but decreased survival of dhh1Δ after UV irradiation was linked to a G1/S DNA damage checkpoint recovery defect (16). We confirmed that dhh1Δ has a G1/S checkpoint recovery defect in our genetic background, but we found no such defect in our lsm1Δ strain (data not shown). This suggests that the mechanism of action by which Lsm1p and Dhh1p provide protection against UV radiation is likely to be different. An intriguing possibility is that decapping activators might not only function at a global level during mRNA turnover, but each might also regulate the decay, stability and/or translation of a specific subset of mRNAs. For example, efficient recovery from G1/S checkpoint arrest would require Dhh1p to regulate the decay, stability and/or translation of a specific subset of mRNAs important for the release from G1 arrest, whereas Lsm1p might be required to regulate the decay, stability and/or translation of a specific subset of mRNAs in the RAD5-mediated subpathway of PRR during the UV-radiation damage response. Identifying the specific substrates of Lsm1p–7p/Pat1p complex will be crucial for understanding the complex interactions that occur during mRNA metabolism and UV-radiation response.
In yeast cells treated with UV radiation, yeast proliferating cell nuclear antigen (PCNA) becomes mono-ubiquitylated at the K164 residue by the Rad6p/Rad18p complex and subsequently becomes poly-ubiquitylated via a K63-linked chain by the Rad5p/Mms2/Ubc13p complex (29–32). Several analyses of PCNA have shown that a variant with a mutated K164 residue is sensitive to UV radiation, and PCNA mono-ubiquitylation at K164 is a prerequisite for Rad30- and Rev3-dependent translesion synthesis as well as subsequent Rad5-dependent poly-ubiquitylation (29, 33, 34). These findings suggest that all subpathways downstream of the Rad18p/Rad6p complex are regulated by mono-ubiquitylation of PCNA at K164, supported by the observation that deletions in RAD5, RAD30 and REV3 do not further enhance the UV-radiation sensitivity of the pol30-K164 mutant (29, 33, 34). In this study, we present evidence that the Rad18p/Rad6p complex may mediate an additional subpathway that functions independently of PCNA-K164 ubiquitylation. We show that LSM1 is in the same epistasis group as RAD18 and RAD5 but that deletion of LSM1 in a pol30-K164 background further enhances the UV-radiation sensitivity phenotype of the mutant. There is some evidence that PCNA may indeed function in a K164-independent manner. For example, the PCNA mutant pol30-46 has four separate point mutations in charged residues with an intact K164 residue, yet it exhibits a UV-radiation sensitive phenotype (35). Genetic analysis of the pol30-46 mutant placed it in the RAD6 epistasis group with respect to UV-radiation sensitivity (36), and the pol30-46 rad5Δ double mutant showed a synergistic increase in UV-radiation sensitivity (37), as did the pol30-46 rev3Δ double mutant (36). In contrast, the pol30-K164R mutant is in the same epistasis group as rad30Δ, rev3Δ (34) and rad5Δ (33), indicating that separate pathways are blocked in the pol30-K164R and pol30-46 mutants. The finding that LSM1 functions independently of K164 ubiquitylation indicates that it might play a role in the subpathway that is blocked in the pol30-46 mutant and that is distinct from the one blocked in the pol30-K164R mutant. Testing the UV-radiation sensitivity of pol30-46 lsm1Δ and pol30-46 pol30-K164R double mutants will be essential to confirm this hypothesis.
In addition to showing that LSM1 functions in a PCNA-K164 ubiquitylation-independent pathway, we also show that LSM1 is in the same epistasis subpathway as RAD5. These results suggest that RAD5 may mediate more than one subpathway: one participating in poly-ubiquitylation of PCNA at K164 residue in conjunction with the Ubc13p/Mms2 complex and a second one participating in a yet-to-be-determined pathway to which LSM1 belongs. Given the complexity of the cellular response to damaging agents, it would not be surprising if Rad5p had additional substrates and/or functions, and there are in fact several lines of evidence indicating that Rad5 affects at least two separate repair subpathways in response to UV-radiation-induced damage (24, 32, 38–41).
In summary, we propose that RAD18 is upstream of at least two major subpathways: one functioning in mono-ubiquitylation of PCNA-K164 leading to the activation of the known error-free and error-prone PRR subpathways and a second pathway involving a novel function mediated by RAD5. The Lsm1p–7p complex possibly regulates the stability of transcripts in this novel RAD5 subpathway and protects them from exosome-mediated 3′ to 5′ degradation by binding to their 3′ ends. In the absence of Lsm1p–7p complex, these transcripts become susceptible to 3′ to 5′ exosome-dependent degradation and lead to the observed UV-radiation sensitivity. By inactivating the exosome complex, these transcripts become stabilized and the UV-radiation sensitivity phenotype is consequently suppressed. These transcripts most likely play a minor role in coping with the UV-radiation-induced stress since cells lacking any of the components of the Lsm1p–7p/Pat1p complex show only moderate UV-radiation sensitivity. However, the effect of this pathway becomes apparent in the absence of nucleotide excision repair, the major pathway dealing with UV-radiation damage, as shown by the synergistic increase in UV-radiation sensitivity in the lsm1Δ rad1Δ double mutant, indicating that both pathways compete for a common substrate resulting from UV-radiation damage.
The authors thank Dr. Roy Parker for advice and generously providing strains and plasmids and Dr. John Game for many helpful comments and suggestions. This investigation was supported by PHS Training Grant CA09302 and Grants CA67166 and GM073879 awarded by the National Cancer Institute and the National Institute of General Medical Science, DHHS, respectively.