Current Genetics 2, 207-210 (1980)
Repair of 2 um Plasmid DNA in Saccharomyces cerevisiae S. J. McCready and B. S. Cox Botany School, South Parks Road, Oxford OX1 3RA, England
Summary. We have developed a system for assaying pyrimidine dimers in the 2 /am DNA plasrnid of Saccharomyces cerevisiae, using Micrococcus luteus UV endonuclease to nick dimer-eontaining plasmid molecules and measuring percentages of nicked and eovalently closed circles on agarose gels. UV-irradiafion induced dimers in plasmid DNA, in vivo, at the same rate as in chromosomal DNA. After a dose of 20 Joules-m -2, approximately 86% of plasmid molecules had at least one dimer. After 3 h incubation under normal growth conditions only 4% still retained dimers in a wild-type strain. In a tad1 (excision-defective) mutant 81% of plasmid molecules still had dimers after 3 h, suggesting that excision repair operates to remove dimers from plasmid DNA in wild.type yeast. Dimers can be removed from 2/am DNA in a tad1 mutant by photoreactivation. Key words: Yeast - Plasmid - Repair
Saccharomyc.es cerevisiae contains about 50-100 copies
per cell of a circular, double-stranded DNA species of about 2/am contour length (Sinclair et al. 1967; Stevens and Moustacehi 1971; Guerineau et al. 1971; ClarkWalker and Miklos 1974). The location of the plasmid is unknown. It is not mitochondrial (Clark-Waiker 1972) and appears not to be within the nucleus (Clark-Walker and Miklos 1974; Livingston 1977). However, it is passed on to only about 50% of the haploid buds arising from a k a r - heterokaryon (in which there is no nuclear fusion) derived from one parent containing the plasmid and one lacking it (Livingston 1977), suggesting, if not a nuclear Offprint requests to: S. J. MeCready
location, restriction of the plasmid to, possibly, a cytoplasmic membrane or outer nuclear membrane attachment site. The 2 #m DNA is similar in several ways to chromosomal DNA. It shares the same buoyant density (Stevens and Moustacchi 1971), is replicated at a specific time (early S) in the cell division cycle (Zakian et al. 1979), is dependent on at least some of the same gene products for its replication (Petes and Williamson 1975;Livingston and Kupfer 1977) and exists as a condensed 'minichromosome' associated with proteins similar in eleetrophoretie mobility to calf thymus core histones (Livingston and Hahne 1979). It is of interest to know whether the same systems are available for repair of damaged plasmid DNA as for chromosomal DNA for two reasons: firstly, from the point of view of stability of both the naturally-occurring plasmid and genetically-engineered hybrid plasmids now widely cloned in yeast (e.g. Beggs 1978), and, secondly, because the 2/am DNA has a potential use in elucidating the significance of the genes known to be unvolved in DNA repair in yeast. We show, in this paper, that UV-induced pyrimidine dimers can be removed from the 2 #m plasmid by photoreaetivation and, unlike those in mitoehondrial DNA (Waters and Moustaechi 1974; Prakash 1975), by the excision repair pathway.
Materials and Methods Strains. The radl mutant used was a radl.1 lysl.1 and the wildtype, BSC 713/3b ade2.1 rad+. UV Irradiation. Cells were grown to early logarithmic phase (2-3 x 107 cells/ml)at 28 °C in YEPDmedium (10%yeastextract, 10% peptone, 2% dextrose) on an orbital shaker, harvested and suspended in distilled water at 2 x 107 cells/ml. 60 ml of cell sus-
S.J. McCready and B. S. Cox: Repair of 2 ~m Plasmid DNA in Yeast
pension were irradiated in a 12 cm x 20 cm plastic dish, using a Hanovia model 11 UV lamp deliveringmore than 95% of radiation at 254 nm. DNA Extraction. Cells were collected by centrifugation, held for 5 min in 0.45 M EDTA, 20 mM fl-mereaptoethanol, pH 8.0 at 23 °C and incubated in 1.2 M sorbitol, 0.45 M EDTA, pH 7.5 containing 100/~g/ml zymolase (Kirin Brewery, Japan) at 36 °C for 15 rain, by which time more than 95% of cells had been converted to spheroplasts. Spheroplasts were washed once in distilled water and DNA extracted by a method based on that of Cryer et al. (1975). After resuspending in 800 ~1 0.045 M EDTA pH 8.0, spheroplasts were lysed by the addition of 200 ~1 5% SDS in 0.05 M tris-HC1, 0.01 M EDTA, pH 8.0 and left for 5 minat 0 °C. Nucleic acid was precipitated, after one ehloroform/isoamyl alcohol extraction, with ethanol at -20 °C for 1 h, resuspended in 0.01 M tris-HC1, 0.01 M EDTA, pH 8.0, cleared by centrifugation, and reprecipitated with ethanol The precipitate was resuspended in 0.01 M tris-HC1, 5 mM EDTA, pH 7.5, incubated with 25/~g/ml heat-treated RNase A and the DNA reprecipitated. The proportion of covalenfly closed circles in this preparation was determined as described below and was normallygreater than 95%. UV-endonuclease Treatment. A crude extract of Micrococcus luteus containing UV-endonuclease activity was prepared by the method of Carrier and Setlow (1970) and stored in small aliquots at -20 °C. The final DNA precipitate, obtained as described above, was resuspended in 50 tzl 0.05 M potassium phosphate buffer, pH 6.5, 1 mM EDTA and 20 IB of this incubated at 22 °C for 15 min with 2 tzl of crude UV-endonuclease. Electrophoresis. Five microliters 20% glycerol, 0.1% bromophenol blue was added to the UV-endonuclease incubation mixture or to 20 ttl DNA extract without UV-endonuclease and loaded onto a 1% agarose gel The gel and running buffer was 0.01 M tris-HCl, 0.05 M phosphate, 0.01 M EDTA, pH 7.0 and 0.8/~g/ml ethidium bromide was added to gels only. Gels were run overnight at 60 mV, destained in running water for 15 rain, and illuminated on a long wave UV transilluminatorand photographed. The relative amount of DNA in the bands on the gels were estimated by tracing negatives with a Joyee-Loebl densitomcter and weighing the peaks corresponding to bands of interest. At the ethidium concentration used there is no significant difference in the fluorescence of equal amounts of supercoiled and open circles. This was confirmed using purified plasmid DNA, irradiating and measuring total fluorescence before and after treatment with UV-endonuctease. There was no detectable increase in total fluorescence with the increase in proportion of open circles. The number of dimers in circles was estimated by comparing the proportions of open and ciosed circles before and after UVendonuclease treatment and calculating the percentage of closed circles converted to open circles. Since one or more dimers in a circle results in a UV-endonuclease nick the mean number of dimers per circle could be calculated, assuming a Poisson distribution of dimers amongst plasmids, from the proportion of elfdes with no dimers.
Results and Discussion Figure 1 shows a plot of the mean number of pyrimidine dimers induced per 2 p m plasmid molecule at doses of UV light up to 40"J m - 2. The mean number of dimers induced was 0.08 per 2 # m per J.m - 2 , or 200 dimers per 5,000
3'0 20' 0f U V ( J r n -2)
Fig 1. Mean number of dimers per 2 ~m circle induced by doses of UV up to 40 J.m -2, calculated from percentages of direct-free circles, assuming a Poisson distributionof dimers amongst plasmid molecules
# m per J" m - 2, which is close to the dimers per genome (5,000 /~m) observed by Unrau et al. ( 1 9 7 3 ) f o r yeast chromosomal DNA. These two sets of data are both from DNA in vivo so any protection of the DNA from the effect of UV irradiation (Unrau et al. 1973) appears to operate on plasmid DNA to the same extent as on chromosomal DNA. The data also suggest that the UVendonuclease cuts close to 100% of the dimers in supercoiled plasmid molecules in the conditions used. Figure 2 shows gels on which DNA samples from wildtype yeast and a radl m u t a n t defective in excision repair (Unrau et al. 1971) were run before and after UV treatment and after various periods of post-UV incubation under normal growth conditions (inYEPD, at 28 °C aereated b y shaking). Figure 3 shows typical densitometer traces used to measure proportions of DNA in bands corresponding to open and closed circles, Figure 4 shows a plot of the proportions of dimer-containing plasmids after post-UV incubation in a wild.type strain and in a tad1 mutant. The results showed that: (1) very few (of the order of 5%) open circles were present in unirradiated or irradiated 2 pm DNA isolated from wild.type o f tad1 cells, showing that the isolation procedure does not result in a significant amount of nicking of DNA. Those circles n o t covalenfly closed were probably those engaged in DNA synthesis; (2) the crude UV-endonuclease preparation did n o t have any detectable non-specific nicking activity on unirradiated DNA; (3) a dose of 20 J.m - 2 resulted in dimers in about 86% of 2 pm plasmid mole-
S. J. McCready and B. S. Cox: Repair of 2 pm Plasmid DNA in Yeast
;/ Fig. 3. Densitometer scans of (a) track 4a and (b) track 4b of gel 2 in Fig. 2 100%
Fig. 2. DNA extracts, from a wild-type and a radl mutant, run on 1% agarose gels containing 0.8 pg/ml ethidium bromide. Samples had been treated as follows: (1) no irradiation (2) 20 J-m -2, no post-UV incubation (3) 20 I'm -2, 1/2 h post-UV incubation (4) 20 J ' m - 2, 3 h post-UV incubation (5) 20 J . m - 2, 3 h post-UV incubation in photoreactivating light (cells in saline). In each case (b) is treated with UV-endonuclease, (a) is untreated. (The 4 pm circles are dimeric forms of the 2 #m plasmid)
cules. After 3 h in YEPD medium only 4% of 2/~m circles contained dimers in wild-type cells. Two possible explanations for this are that either dimers were repaired or removed, or that molecules containing dimers were degraded and replaced by newly-synthesised dimer-free plasmids; (4) in a tad1, excision.defective, mutant 81% of 2/~m circles still had dimers after 3 h incubation. This suggests that the former o f the two possibilities above is correct and that the reduction in dimer-containing mole. cules in the wild-type strain was due to excision repair. This is in contrast to mitochondrial DNA which is not able to use the excision pathway (Waters and Moustacchi 1974; Prakash 1975). Presumably the excision repair enzymes or enzyme complex cannot penetrate the mito-
Fig. 4. Proportions of covalently closed circles which contained dimers after various periods of incubation following irradiation of a wild-type strain (A) and a tad1 mutant (=)
chondfial membrane. Exposure of the radl mutant to photo-reactivating light for 3 h resulted in complete loss of dimers from plasmid DNA. Photoreactivating enzyme is also, therefore, available to plasmid DNA. We do not know how DNA synthesis affects these resuits. It is not known whether dimer-containing plasmid molecules are replicated or how replication of undamaged
S.I. McCready and B. S. Cox: Repair of 2 ~m Plasmid DNA in Yeast
molecules is affected b y a dose o f 20 J.m - 2 o f UVirradiation. I f replication continued after UV irradiation one would expect an increase in the p r o p o r t i o n o f dimer-free plasmids b o t h in the presence or absence o f repair. This would, clearly, be true if only dimer-free plasmids could replicate but would also be true i f those containing dimers could also be replicated (since plasmids containing one dimer would give rise to 50% dimer-free daughter molecules). Some o f the increase in dimer-free molecules in the wild-type strain m a y be due to replication, therefore, as m a y the small increase observed in the tad1 strain. However, if replication proceeded uninterrupted and at a normal rate after irradiation, we would expect to see a much greater increase in dimer-free circles - to approx. 3 0 - 4 0 % in the rad'mutant 1. It seems, therefore, that replication o f plasmid DNA is suppressed for a time after irradiation at the dose used.
Acknowledgement. This work was funded by Science Resarch Council grant no. GR/A 82253.
Clark-Walker GD (1972) Proc Nat] Acad Sci USA 69:388-392 Clark-Walker GD, Miklos GLG (1974) Eur J Biochem 41:359365 Cryer DR, Eccleshall R, Marmur l (1975)Methods in Cell Biology XII: 39 -44 Guerineau M, Grandchamp C, Paoletti J, Slonimski P (1971) Biothem Biophys Res Commun 42:550-557 Livingston DM (1977) Genetics 86:73-84 Livingston DM, Hahne S (1979) Proc Nat] Acad Sei USA 76: 3727-3731 Livingston DM, Kupfer DM (1977) J Mol Biol 116:249-260 Petes TD, Williamson DH (1975) Cell 4:249-253 Prakash L (1975) J Mol Biol 98:781-795 Sinclair JH, Stevens BJ, Sanghavi P, Rabinowitz M (1967)Science 156:1234-1237 Stevens BJ, Mousacchi E (1971) Exp Cell Res 64:259-266 Unrau P, Wheatcroft R, Cox BS (1971) Mol Gen Genet 113: 359-362 Unrau P, Wheateroft R, Cox BS, Olive T (1973)Biochim Biophys Acta 312:636-632 Waters R, Moustaechi E (1974) Biochim Biophys Acta 366: 241-250 Zakian VA, Brewer BJ, Fangman WL (1979) Cell 17:923-934
References Communicated b y U. Leupold Beggs JD (1978) Nature 275:104-109 Carrier WL, Setlow RB (1970) J Bacteriol 102:178-186
In the absence of repair if both dimer-containing and direct-free plasmid molecules replicate normally: after a dose of 20 J'm -2 plasmid molecules contain an avergage of 1.6 dimers each. Assuming a Poisson distribution of dimers amongst 2 ~m circles, approximately 20% should be dimer-free (and give rise to 100% direct-free daughter molecules), 35% should have one dimer (and give rise to 50% dirner-free daughters), 25% should have two dimers (and give rise to 25% direct-free daughters since half of the dimer-contalming plasmids will have both dimers in the same strand) and so on. During one generation time (21/2) one would9 on this basis, expect approx. 43% of plasmids to be direct-free. If only dimer-free plasmid molecules can replicate, the proportion of dimer-free molecules after 3 h would be 33%. A temporary lag in DNA synthesis would reduce these percentages.
Received August 18, 1980