ABSTRACT
The recently cloned Saccharomyces cerevisiae MMS19 gene appears to be involved in both nucleotide excision repair (NER) and transcription, which is also the case for components of the NER/transcription complex TFIIH. Unlike TFIIH however, the Mms19 protein does not affect NER in a highly purified in vitro system. In order to investigate the role of Mms19 in NER, we have analysed the repair capacity of the mms19 disruption mutant. We find that a cell-free extract of this mutant is deficient for NER in vitro. Since mms19 mutants are only moderately sensitive to irradiation with ultraviolet (UV) light, it is possible that such mutants are specifically deficient in one of the two modes of NER, i.e. transcription-coupled or global genome repair. To investigate this possibility, we have analysed the removal of cyclobutane-pyrimidine dimers (CPDs) at the nucleotide level in an mms19 mutant. Repair of CPDs was not detectable for both transcribed and non-transcribed sequences in this mutant, demonstrating a requirement for Mms19 in both transcription-coupled and global genome repair. Our data, combined with those obtained by others, suggest that Mms19 is required for NER in yeast, although it seems likely that the protein plays an indirect role in this process.
Nucleotide excision repair (NER) is a multi-step process capable of removing a variety of lesions from the DNA (reviewed in 1 ). The characteristic step of the NER reaction comprises single stranded incisions on both sides of the lesion in the damaged strand. NER has been reconstituted in defined systems by using purified yeast or mammalian proteins (2 -4 ). Besides the minimal set of proteins necessary for the damage-dependent dual incisions that is revealed by those experiments, other proteins may influence the proficiency of NER inside the cell. The kinetics of NER is not equal for different regions of the genome, and NER can be divided into two subpathways. Transcription-coupled repair is a mode of NER that accomplishes fast repair of transcribed strands, while non-transcribed DNA relies on the generally slower global genome repair subpathway. Specific proteins are implicated in each of these pathways. Such proteins are accessory to the NER reaction, not being required for the reconstituted in vitro NER reaction with highly purified components, but greatly influencing the NER reaction in vivo. Besides the phenomenon of transcription-coupled repair, in which NER enzymes are specifically attracted towards lesions on the transcribed strand that have stalled the transcription machinery (5 ), another intimate link between NER and transcription exists. The multisubunit complex TFIIH is required for transcription and for NER (reviewed in 6 ). TFIIH is required for repair of both transcribed and non-transcribed DNA (7 ,8 ), demonstrating that its role is not confined to the transcription-coupled repair subpathway of NER.
Saccharomyces cerevisiae mms19 mutants are moderately UV-sensitive (9 ,10 ), as compared to several NER-deficient rad mutants that are highly UV-sensitive. This suggests an accessory role for the Mms19 protein in the NER process. Recently the MMS19 gene has been cloned, and the encoded protein appeared to be involved in both NER and in transcription, like TFIIH components (10 ). However, Mms19 is not a subunit of TFIIH, nor appears to be stably associated with this NER/transcription factor (10 ). Instead, Mms19 might influence TFIIH as an upstream factor, since a thermolabile transcription defect of mms19 cell-free extracts could be complemented by addition of purified TFIIH and not by purified Mms19 protein (10 ). The pleiotropic effects of mms19 mutations might be partly caused as a secondary consequence of a transcription defect in such mutants. Although it could be concluded that mms19 mutants have an NER-defect based on a deficiency in removal of UV-induced lesions and UV-sensitivity (9 ,10 ), the purified Mms19 protein does not affect the damage-dependent incisions in a highly defined reconstituted NER system (10 ). It was, therefore, of considerable interest to determine the repair capacity of mms19 mutants, compared to other mutants in which the proficiency of NER is affected, in order to gain insight into the role of Mms19 in NER. The moderate UV-sensitivity of mms19 mutants could result from a deficiency in NER of a subset of lesions, as has been found for rad7 and rad16 mutants, which also display a moderate UV-sensitivity. These latter mutants are specifically deficient in repair of lesions in non-transcribed DNA, whereas transcription-coupled repair is unaffected (11 ). On the other hand, Mms19 is known to affect transcription (10 ), and therefore, it is also possible that the mms19 mutant is specifically deficient for transcription-coupled repair. Therefore, this mutant might be disturbed exclusively in either one of the subpathways of NER. Alternatively, NER of all lesions might be affected. Here we describe the characterization of the NER defect in the mms19 disruption mutant. We demonstrate that such a mutation augments the UV-sensitivity of several partially NER deficient mutants disturbed in transcription-coupled repair, global genome repair or both. We reveal a NER deficiency in cell-extracts from the mms19 mutant in vitro. Finally, we show that this mutant is defective in both transcription-coupled and global genome NER in vivo. These repair data extend the observations of Lauder et al. (10 ), and are in accordance with a role of the Mms19 protein upstream of TFIIH, as suggested by these authors.
The S.cerevisiae wild-type strain used for this study is W303-1B (MAT[alpha] ade2-1 trp1-1 can1-100 his3-11,15 leu2-3,112 ura3-1). An isogenic mms19[Delta]mutant was constructed by a PCR based strategy (12 ). The following primers were used for PCR: 5'-CACAGTAACGGATTCTATTGTACACCGATCATCTTTGACAGCTTATCATC-3' and 5'-GTGTCCTTCAGAGTTTCTAAAGCGGATACCCACTCGTGCACCC-3'. PCR was carried out for five cycles at 45oC, followed by 30 cycles at 72oC, using pYES2 as a target. The PCR fragment was transformed to W303-1B. The resulting mms19[Delta] strain MGSC217, in which almost the entire open reading frame of the MMS19 gene was replaced by URA3, had characteristics consistent with mms19 mutants described in literature (10 ). The PCR-fragment was also used to construct the mms19[Delta] mutation in the following isogenic yeast strains that were described before (11 ,13 ,14 ): MGSC126 (rad16[Delta]), MGSC102 (rad26[Delta]), MGSC107 (rad16[Delta] rad26[Delta]) and MGSC139 (rad14[Delta]) to obtain strains MGSC219 (rad16[Delta] mms19[Delta]), MGSC218 (rad26[Delta] mms19[Delta]), MGSC220 (rad16[Delta] rad26[Delta] mms19[Delta]) and MGSC221 (rad14[Delta] mms19[Delta]). All strains were kept on selective YNB (0.67% yeast nitrogen base, 2% glucose, 2% bacto agar) supplemented with the appropriate markers (15 ). Cells were grown in complete medium (YEPD: 1% yeast extract, 2% bacto peptone, 2% glucose) at 28oC (15 ).
Yeast cell-free extracts were prepared as described (16 ). Yeast cells were converted to spheroplasts with yeast lytic enzyme (ICN). Spheroplasts were washed thoroughly and stored as pellets at -80oC until further use. The pellet was resuspended at 4 ml/g spheroplast in a hypotonic buffer (10 mM Tris-HCl pH 8, 1 mM EDTA, 5 mM DTT) containing protease inhibitors (benzamidin 300 µg/ml, leupeptin 1 µg/ml, antipain 1 µg/ml, pepstatin 1 µg/ml, chymostatin 1 µg/ml, PMSF 1 mM) and stirred for 20 min on ice. A sucrose solution (50 mM Tris-HCl pH 8, 10 mM MgCl2, 25% sucrose, 50% glycerol, 2mM DTT) (4 ml/g spheroplasts) was added dropwise to this suspension while stirring gently. After another 20 min of stirring, 290 µl/ml of a neutralized solution of 4M (NH4)2SO4 was added dropwise, and stirring on ice was continued for another 30 min. The suspension was centrifuged at 170 000 g for 1 h at 4oC (Beckman L7 ultracentrifuge, 75Ti rotor). The supernatant was transferred to a beaker on ice, leaving some residual supernatant. Per ml of supernatant, 330 mg of powdered (NH4)2SO4 was added over the course of 30-40 min, while gently stirring on ice. After neutralization of the solution with 1M NaOH [10 µl/g (NH4)2SO4 added], stirring on ice was continued for 30 min. The suspension was centrifuged at 25 000 g for 15 min at 4oC and the pellet was resuspended in 1/40-1/30 vol. of the ultracentrifugation supernatant in dialysis buffer [20 mM HEPES-KOH pH 7.5, 20% (v/v) glycerol, 10 mM MgSO4, 10 mM EGTA, 5 mM DTT] containing protease inhibitors. After dialysis overnight on ice against dialysis buffer containing 1 mM PMSF, precipitates were removed by centrifugation at 15 000 g for 15 min at 4oC. The resulting yeast cell-free extract was snap-frozen using liquid nitrogen and stored at -80oC. Under these conditions the extract was stable for at least several months and could be used for in vitro NER experiments, even upon repeated thawing and freezing of the extract.
To measure in vitro NER activity (method as in ref. 16 ,17 ), supercoiled plasmid pNP81 (4.4 kb) was purified as an undamaged control plasmid, while purified pUC18 (2.7 kb) was treated with N-acetoxy-2-acetylaminofluorene to obtain pUC18-AAF. Standard in vitro NER reactions (50 µl) contained 300 ng each of pNP81 and pUC18-AAF DNA, NER buffer (45 mM HEPES-KOH pH 7.8, 40 mM KCl, 7 mM MgCl2, 1 mM DTT, 0.4 mM EDTA, 20 µM dGTP, 20 µM dATP, 20 µM dTTP, 4 µM dCTP, 1 µCi [[alpha]-32P]dCTP, 2 mM ATP, 40 mM creatine phosphate, 2.5 µg creatine phosphokinase, 4% glycerol, 18 µg BSA, 5% polyethylene glycol (PEG 6000) and 250 µg yeast cell-free extract. Repair reactions were incubated for 2 h at 28oC. Reactions were stopped by addition of 2 µl of 0.5 M EDTA and incubation for 2 min on ice. Samples were treated with RNaseA (2 µg) for 10 min at 37oC, and DNA was subsequently purified by treatment of the reaction mixtures with 0.5% SDS and 1 µg/µl proteinase K at 50oC for 30 min and precipitation of protein with ammonium acetate (final concentration 2.4 M). DNA was precipitated with ethanol, resuspended in TE (10 mM Tris-HCl pH 8, 1 mM EDTA) and digested with BamHI to linearize the plasmids. The samples were electrophoresed in 1% agarose gels containing ethidium bromide (EtBr) and photographed. The gels were dried and autoradiographs were prepared.
Yeast cells were grown until stationary phase in YEPD at 28oC. Serial dilutions of the cells were prepared in water and spread on YEPD plates. The plates were irradiated with increasing doses of UV-light and incubated in the dark at 28oC. After 4-6 days, colonies were counted and survival was calculated.
Yeast cells diluted in chilled phosphate-buffered saline were irradiated with 254 nm UV light (Philips T UV 30W) with 70 J/m2. Cells were collected by centrifugation, resuspended in complete medium and incubated for various times in the dark at 28oC prior to DNA isolation. DNA samples were purified on CsCl gradients (18 ). DNA samples (25 µg) were digested with appropriate endonucleases, precipitated and URA3 or RPB2 fragments were isolated and end-labeled as described previously (19 ,20 ) using fragment-specific oligonucleotides (sequences available upon request). CPDs were identified using T4endoV. DNA samples were divided in two equal parts. One was incubated with T4endoV, the other was mock treated. Samples were subjected to spun column chromatography and lyophilized to small volumes. Approximately equal amounts of c.p.m. were loaded on 6% denaturing acrylamide gels. After drying, autoradiographs were prepared from the gels.
From each experiment, multiple autoradiographs were obtained with different exposure times to allow signal determination within the linear range of X-OMATtm (AR) Scientific imaging films (Kodak) for each individual CPD. Autoradiographs were scanned using an LKB Ultrascan XL densitometer (Pharmacia) and analysed using ImageMastertm software (Pharmacia). Background levels were subtracted and gel-band intensities were corrected for loading variations. OD-values were plotted against repair time for each CPD that gave sufficient signal to background ratio. Repair half-times (t1/2), defined as the time at which 50% of the initial damage (signal at t = 0) was removed, were derived from these plots.
The purified Mms19 protein does not influence the NER reaction when purified proteins are used to reconstitute this reaction (10 ). Nevertheless, an mms19 mutant is sensitive to UV and epistatic with NER deficient mutants, suggesting that Mms19 is involved in NER. We have determined the in vitro NER capacity of a cell-free extract of a yeast strain lacking Mms19. To this purpose an mms19[Delta] mutant was constructed by a PCR-based strategy (12 ). This mutant is a methionine auxotroph, grows slowly at 37oC and is sensitive towards MMS and UV-light, in agreement with the mms19 mutants described before (10 ). The in vitro NER capacity of this mutant was investigated by determining the repair synthesis activity of a cell-free extract on a plasmid containing AAF damage, compared to a non-damaged plasmid (17 ).
In vivo, the mms19 mutant may be deficient in NER of all lesions, or in either transcription-coupled or global genome repair (see Introduction). By constructing the mms19 mutation in genetic backgrounds that are already affected in one or both of the subpathways of NER, the influence of the mms19 mutation on the remaining NER system can be studied by UV-survival experiments. When the mms19 mutation leads to higher UV-sensitivity than found in the single mutant, Mms19 affects the remaining subpathway(s). As an example, rad16 rad26 double mutants are more UV-sensitive than either single mutant, since rad16 mutations lead to deficiencies in global genome repair while the rad26 mutation affects the other, transcription-coupled, subpathway of NER (14 ). Therefore, we constructed the mms19 mutation in strains that carry deletions of the RAD16 gene (deficient for global genome repair, 11 ), the RAD26 gene (partially deficient for transcription-coupled repair, 13 ,14 ), or both. To confirm epistasis with the rad3 group of NER mutants (9 ,10 ), an mms19 mutant was constructed in a completely NER deficient rad14 mutant. Figure 2 shows the survival of these mutants after irradiation with UV light. The mms19 mutation appears to increase the UV-sensitivity of the other partially NER deficient mutants in either subpathway, suggesting that Mms19 affects both transcription-coupled and global genome repair.
To test this directly, we performed repair analysis of UV-induced cyclobutane-pyrimidine dimers (CPDs) at the nucleotide level in the mms19 mutant in vivo, using a method we previously described (19 ,20 ). Yeast cells were irradiated with UV and incubated at various times to allow repair, after which DNA was isolated. Lesions were detected by sizing end-labeled DNA, cleaved at the site of CPDs with the T4endoV enzyme, thus giving rise to a distinct band on a gel, reflecting the occurrence in the sequence of UV-sensitive pairs of adjacent pyrimidines (19 ). Disappearance of these bands from the lanes treated with T4endoV in the later time points is indicative for removal of CPDs. Figure 3 A shows repair data for the transcribed strand of the URA3 gene and Figure 3 shows repair of the non-transcribed strand of this locus. In the repair proficient strain, removal of CPDs after UV irradiation is evident from both the transcribed and the non-transcribed strand. Because of transcription-coupled repair, CPDs from the transcribed strand are removed faster than from the non-transcribed strand (compare Fig. 3 A and B) in the RAD+ strain (20 ). In contrast, CPDs are not removed after repair periods up to 2 h from either DNA strand of the mms19 mutant (Fig. 3 A and B). In addition, we analysed the removal of CPDs from the template and non-template strand around the transcription initiation site of the RPB2 locus, allowing repair analysis of the transcribed strand and of non-transcribed DNA (promoter sequences and the non-transcribed strand of an active gene). In Figure 4 A and B, the repair half-times of CPDs at specific dinucleotides in this gene are represented graphically, as calculated by quantification of the intensity of the corresponding bands on autoradiographs (19 ). Like the URA3 gene, removal of CPDs from either strand of the RPB2 gene is not detected in the mms19 mutant, while lesions in these regions are repaired in the RAD+ strain (Fig. 4 A and B). From these data we conclude that the mms19 mutant is deficient in transcription-coupled and global genome repair.
We have investigated the NER capacity of the yeast mms19 mutant and find that it is deficient for NER in vitro and in vivo. Despite the moderate UV-sensitivity of this mutant, which might suggest a specific defect of either global genome or transcription-coupled repair, we demonstrate that this mutant is deficient in both subpathways of NER.
The Mms19 protein influences transcription (10 ), and probably acts upstream of TFIIH in the cell, giving a possible explanation for the pleiotropic phenotype of the mms19 mutant. However, in contrast to the defect in transcription that is observed most clearly at 37oC, we observe a complete NER defect at 28oC, at which transcription still takes place in this mutant. Hence, it is not likely that the NER defect of an mms19 mutant is caused by an indirect transcription defect. This idea is corroborated by the following observations: (i) the NER defect in vivo is not confined to the transcription-coupled repair subpathway and (ii) cell-free extracts of repair deficient strains that have functional TFIIH can be complemented by the mms19 mutant, demonstrating that other NER proteins are still functional in the mms19 mutant.
The reason for the only moderate UV-sensitivity of the mms19 mutant is not known, as we failed to detect NER activity in this mutant. Probably, a residual NER capacity which leads to very slow removal of lesions exists in this mutant, but escapes detection in our in vivo repair analysis. A similar finding was reported for the rad23 mutant (21 , and discussion therein). Since complete absence of TFIIH is lethal, the mms19 mutation probably affects the stability of TFIIH, without completely abrogating its function. This might lead to a slow turnover or a very low capacity of the NER reaction, rendering it difficult to measure repair activity in this mutant. The time for a survival experiment can suffice for residual repair. We cannot formally exclude the possibility that another class of UV-induced lesions, pyrimidine (6-4) pyrimidone photoproducts [(6-4)PPs], is repaired in the mms19 mutant. However, we consider this very unlikely since the genetic requirements for NER of CPDs and (6-4)PPs are the same (22 ). The development of a method to analyse repair of (6-4)PPs at the nucleotide level in yeast is currently in progress in our laboratory, allowing to obtain experimental data on this presumption. The absence of detectable NER activity of cell-free mms19 extracts in vitro is shared with the other moderate UV-sensitive rad7, rad16 and rad23 mutants (23 ,24 , our unpublished results). From these, only rad23 strains have a mutation in a protein that is required for damage-dependent incisions in a reconstituted NER reaction with purified components. This demonstrates that besides the minimal set of proteins required for NER in a purified system, other proteins are required for this complex process in less well defined systems in vitro, as is the case inside the living cell.
The experiments of Lauder et al. strongly suggested that Mms19 acts upstream of TFIIH in the cell (10 ). Our experiments are fully consistent with this hypothesis. Indeed, the NER defect of a cell-free extract from an mms19 mutant cannot be complemented by an extract derived from a strain that has no functional TFIIH because one of the subunits of TFIIH (Rad3) is mutated, whereas complementation by NER-deficient extracts that contain intact TFIIH is possible. Furthermore, the mms19 mutant is deficient for both transcription-coupled and global genome repair, as are yeast strains with defects in TFIIH components (7 ,8 ). Despite the absence of a role for the purified Mms19 protein in NER with purified components as described by Lauder et al. (10 ), this report directly demonstrates a NER defect in the mms19[Delta] mutant. Therefore, Mms19 is involved in NER, albeit probably not in a direct manner but rather via upstream regulation of the integrity of TFIIH (10 ).
It has been suggested that mutations in a human homolog of MMS19 might be responsible for certain diseases in which both transcription and NER are affected (10 ). Our data suggest that patients with such mutations, if they exist, will have a general NER deficiency rather than being specifically deficient for transcription-coupled or global genome repair.
We thank Michiel van Oosterwijk for providing the initial pUC18-AAF preparation. This research was partially financed by the J.A. Cohen Institute for Radiation Protection and Radiopathology, project 4.2.13.
*To whom correspondence should be addressed. Tel: +31 71 5274755; Fax +31 71 5274537; Email: brouwer@chem.leidenuniv.nl
REFERENCES