© Springer-Verlag 1983
UraciI-DNA Glycosylase Defective Mutants of
Ustilagomaydis
Yoko Yamamoto 1 and Robin Holliday Genetics Division, National Institute for Medical Research, The Ridgeway,Mill Hill, London, NW7 1AA, England
Summary. Uracil-DNA glycosylase activity can be measured in cell-free extracts of Ustilago maydis and in its general properties the enzyme resembles the same glycosylase from other organisms. A rapid assay was used to screen nearly 1,000 clones from cells treated with N-methyl-N'-nitro-N-nitrosoguanidine, and five isolates had < 1 - 3 0 % of wild-type enzyme activity. All these ungstrains were unstable since they frequently recovered normal enzyme activity (ung+), and two were lost for this reason. The ung- strains were also variable in morphology, and two which were slow growing, produced faster growing ung + sectors. However, in preliminary experiments, ung- strains were shown to be only weak mutators. Two ung- mutants were crossed to wild-type and ung- progeny were recovered, which were also variable in morphology and reverted to ung +. It was found that a standard metabolic enzyme, glucose-6-phosphate dehydrogenase, was significantly more heat-labile in ung- strains than wild-type. The results indicate that the presence of uracil in DNA may have more severe physiological effects in eukaryotes than in prokaryotes. Over 1,000 isolates form Chinese hamster ovary (CHO) cells treated with ethyl methane sulphonate, were also tested for uracil-DNA glycosylase activity, but no stable ungstrain was recovered. Key words: Uracil-DNA glycosylase - Ustilago maydis Instability - G-6-PD - CHO cells
Offprint requests to: R. Holliday 1 Present address: Department of Radiation Biophysics, Kobe University School of Medicine, Kusunoki-cho7-5-1 Chuo-ku, Kobe 650, Japan
Introduction
Uracil can be introduced into DNA in two ways. One is the direct deamination of cytosine residues in DNA. This can occur spontaneously at a low rate under physiological conditions (Lindahl and Nyberg 1974) and can be induced by specific reagents, such as nitrous acid or bisulphite (Freese 1971; Hayatsu 1976). The other depends on the misincorporation of dUMP into DNA instead of dTMP by DNA potymerase during DNA replication in vivo (Tye et al. 1977; Tye et al. 1978;Tye and Lehamn 1977; Brynolf et al. 1978; Grafstrom et al. 1978). Uracil-DNA glycosylase recognizes uracil residues in DNA, whichever way it is introduced, and catalyzes the hydrolysis of uracil-deoxyribose bonds so that free uracil bases are released from DNA. Since the discovery of this enzyme activity in E. coli by Lindahl (1974),uracil-DNA gtycosylase has been found widely from prokaryotes to eukaryotes; for example, B. subtilis, yeast, wheat germ, human cells (reviewed by Lindahl 1979), as well as cultured Chinese hamster ovary cells (Y. Yamamoto and P. A. Jeggo, unpublished data). Uracil-DNA glycosylase has been highly purified from E. coli (Lindahl et al. 1977) and partially purified from B. subtilis (Cone et ai. 1977), calf thymus (Talpaert-Borl6 et al. 1979), human leukaemia cells (Charradonna and Cheng 1980), and HeLa cells (Krokan and Wittwer 1981). All of these enzymes showed similar properties; they are of low molecular weight, have a strict specificity for uracil in DNA and are fully active in the presence of EDTA. It would be expected that the conversion of cytosine to uracil would lead to transition mutations (C-G -+ T - A ) during DNA replication, if the uracil was not previously removed. Mutants lacking uracil-DNA glycosylase should therefore be mutable. The incorporation of dUMP in place of dTMP would not produce mutations, but might have other physiological effects, if the uracil remained in
Y. Yamamoto and R. Holliday: Uracil Glycosylase Mutants of UstiIago
290 Table 1. Ustilago maydis strains Strain
Genotype or phenotype
Source
483 617 YY47 YY57 YY34 YY7 YY7L YY43 YY43-164 R1 R2
narl-1, inosl-3, a2b 2 panl-1, adel-1, nicl-1 a, b, The same as 483, but ung The same as 483, but ung The same as 483, but ung The same as 483, but ung The same as YY7, but the colony size was bigger The same as 483, but ung pan1-1, nat1-1, ions1-3, nicl-1, ung, a,b, The same as YY43-164, but ung + The same as YY43-164, but ung +
MNNG mutagenesis MNNG mutagenesis MNNG mutagenesis MNNG mutagenesis Spontaneously isolated from YY7 MNNG mutagenesis Meiotic segregant from the cross 617 x YY43 Spontaneously isolated from YY43-164 Spontaneously isolated from YY43-164
DNA. So far, the biological significance of the enzyme has only been examined in prokaryotes. Mutants deficient in this enzyme activity were isolated by direct enzyme assay from both B. subtilis (Makino and Munakata 1977) a n d E . coli K-12 (Duncan et al. 1978). The B. subtilis m u t a n t ( u r g - ) , which had less than 1% of the wildtype enzyme, could grow at normal rate and exhibited normal sensitivity to nitrous acid, whereas one of the E. coli mutants (ung.1), which showed only 0.02% activity of its parental strain, was more sensitive to nitrous acid and bisulphite than wild-type (Da Roza et al. 1977; Simmons and Friedberg 1979). It could grow normally and was also a weak mutator. Mutants ofE. coli deficient in dUTPase have also been isolated, and these incorporate significant amounts of uracil into DNA (Hochhauser and Weiss 1976; Tye et al. 1977). From the study of both the ung m u t a n t and the dUPTase defective m u t a n t , it is now believed that uracil-DNA glycosylase has an important role in the repair of uracil in DNA, whether or n o t a uracil-guanine mismatch is present (Duncan and Weiss 1978; Duncan and Miller 1980). It is important to study the biological function of uracii-DNA glycosylase in eukaryotic systems. For this purpose, we selected the simple eukaryotic organism, Ustilago maydis, which can grow like yeast and has m a n y advantages for biochemical and also genetical studies (Holliday 1974). In this paper, we first show that crude extracts from Ustilago maydis exhibit uracil-DNA glycosylase activity, and then describe the isolation and the characterization of mutants deficient in this enzyme activity.
Materials and Methods Chemicals Poly[d(AT)] was purchased from Boehringer and Mannheim. Poly[d(AT)] was dissolved with 50 mM Tris-HC1 (pH 7.5) f.c.
(50OD26o) and stored at -20°C. [3H]dUTP (11Ci/mmot, 1 mCi/ml, Cat. No.TRK351) was purchased from the Radio Chemical Centre. The highly purified E. eoli DNA polymerase I was kindly given by Dr. G. T. Yarranton. Sodium bisulphite and sodium thioglycolate were purchased from Nakarai Chemicals Ltd. NADP and D-glueose-6-phosphate were obtained from Kohjin Co. Ltd. and Sigma Chemical Co., respectively. Chioroneb (1,4 dichloro-2,5-dimethoxybenzene) was a gift from E. I. Dupont de Nemours and Co., Delaware, U.S.A.
Media Complete medium, nitrate minimal medium and yeast extract peptone (YEP) medium were prepared as described previously (Holliday 1974).
Strains The strains of Ustilago maydis used for the experiments in this paper are listed up in Table 1.
Polyld(A-[ 3H lU) ] Preparation Poly[d(A-[3HIU)] was synthesized by E. coli DNA polymerase I in vitro. The reaction mixture (100 #1) contained 66 mMTrisacetate buffer (pH 7.4), 6 mM MgC12, 5 mM DTT, 3 nmole dATP, 0.3 nmole dUTP, [3HIdUTP (2 #Ci, 0.18 nmole), poly[d(AT)] (0.25 OD26o) and 51,1 of E. coli DNA polymerase (0.5 unit; one unit was defined as the activity which incorporated 10 nmole total nucleotides into DNA in 30 min at 37 ° C). Ten assay tubes (each tube contained 100 #1 of the reaction mixture) were incubated at 37 °C for 20 min, then sodium dodecyl sulphate (f.c. 1%) was added. The reaction mixtures were gathered, treated with phenol and then DNA was precipitated by ethanol. DNA was resuspended with lml of 10 mM phosphate buffer (pH 7.0) - 10 mM EDTA (ph 7.5) and was dialyzed against the same buffer. The net count of 5 t~l of this DNA solution was about 130,000 cpm. The DNA solution was used after appropriate dilution with poty[d(AT)] (2.5 OD260).
Y. Yamamoto and R. Holliday: Uracil Glycosylase Mutants of Ustilago
Uracil-DNA Glycosylase Assay 1) Crude Extract Preparation
Freezing-and-Thawing Method. Fresh colonies were inoculated into 5 ml or 10 ml of complete medium and were shaken at 26 °C for 24 h. After 24 h, ung+ strains (483 and 617) and YY34 were grown to saturation (approximately 2 x 10 s cells/m), while the cell concentration of cultures of YY7L and YY43 was around 1 x 107 cells/ml. Cells were harvested, washed with 10 ml of water twice, washed with 500 tal of buffer (50 mM Tris-HCl (pH 7.5) - 1 mM DTT - 1 mM EDTA - 10% glycerol) and finally resuspended with 50 to 200 ~1 of the same buffer. The cells were broken by repeating freezing-and-thawing four times and immediately centrifuged at 15,000 rpm for 20 min at 4 °C. The supernatants were re-centrifuged under the same conditions and were stored at - 7 0 °C.
French-Pressure-Cell Method. Freshly prepared cells (2 x 109 cells) were washed and resuspended with 10 ml of buffer (50 mM phosphate buffer (pH 7.0) - 1 mM DTT - 1 mM EDTA - 10% glycerol). Cells were broken by passing through a French-pressure-cell at 18,000 psi twice and were centrifuged twice as described above. The crude extracts were stored at - 7 0 °C. The protein concentrations of crude extracts were determined by the method of Lowry et al. (1951).
2) Assay of the Uracil-DNA Glycosylase The assay procedures were basically the same as the methods described by Duncan et al. (1978). The crude extract was incubated at 32 °C in an assay mixture (50 tsl), which contained 50 mM phosphate buffer (pH 7.0) (or 50 mM Tris-HC1 (pH 7.5)), 30 mM EDTA (pH 7.5, adjusted with NaOH), 1 mM DTT and poly[d(A-[3H]U)]. Uracil-DNA glycosylase activity was usually measured by the assay based on the acid-solubility of uracil. The acid-soluble products obtained by the crude extract of strain 483 (ung +) were analysed by paper chromatography in the solvent system; isobutyric acid : water : 0.1 M sodium EDTA :conc. ammonia:toluene; 160:22 : 3 : 2:20 (Reeves et al. 1969). 80% of the total counts of the acid-soluble fraction were located at free uracil position, not at deoxyuridine or deoxyuridine monophosphate. The reaction products were also examined by Dowex-1. After incubation, 10 #1 of 2% sodium dodecyl sulphate was added to the reaction mixture and applied to the Dowex column. Almost the same amount of radioactivity as that of the acid-soluble fraction was eluted with water.
Screening of u ng- Mu tan ts Wild-type strain 483 was mutagenized with N-methyl-N'-nitro-Nnitrosoguanidine (MNNG) by using two different methods: (1) CeBs from the logarithmic growth phase in complete medium at 32 °C were spun down and resuspended with 9 ml of water (Lc. 1.25 x 107 cells/ml). One ml of freshly prepared MNNG solution (500 ~g/ml in 25% ethanol) was added to the ceil suspension and it was shaken at 32 °C for 20 rain (S/S 0 = 1.6 x 10-4). The reaction was stopped by adding 1 ml of cold 0.5 M Tris-HC1 (pH 7.5) and the cells were spun down at room temperature for 5 rain. The cells were washed with 10 ml of 50 mM Tris-HC1 (pH 7.5) and then resuspended with 10 ml of complete medium. Each 1 ml of
291
this suspension was added to 9 ml of complete medium (total, two tubes) and incubated at 32 °C until saturation (stationaryphase). A portion of the culture from each tube was diluted and spread on complete medium agar and incubated at 32 °C. 95 colonies were picked from each plate and transferred to complete medium agar. They were incubated at 32 °C for several days and kept at 4 °C as master plates. (2) Cells were mutagenized in almost the same way as method (1). 4.5 ml of the cell suspension was added to 0.5 ml of MNNG solution (500 ~g/ml in 25% ethanol) and incubated at 32 °C for 5 rain (S/S 0 = 29%). The reaction was stopped by adding 5 ml of cold 50 mM phosphate buffer (pH 7.0). The cells were washed with the same buffer and, after the appropriate dilutions, were immediately plated on complete medium agar. The plates were incubated at 26 °C in order to get temperature-sensitive uracil-DNA glycosylase mutants. 800 colonies were picked up and transferred to complete medium agar. They were incubated at 26 °C for several days and kept at 4 °C as master plates. The screening assay for ung- mutants was based on the acidsolubility of uracil, using the assay procedure developed by Duncan et al. (1978) with minor modifications. Each colony from the master plates was inoculated into 1 ml of complete medium in trays (16 mm well diameter and 20 mm well depth, 24 cultures per tray) and was grown for two days at 26 °C. Wildtype cells and most of the mutagenized cells could grow to saturation ( - 108 cells/mt). Each culture was transferred to a small Eppendorf tube. The cells were spun down at 4 °C, resuspended with 40 ~1 of the buffer (50 mM phosphate buffer (pH 7.0), 1 mM DTT, 1 mM EDTA) and then broken by freezing-and-thawing four times. They were immediately centrifuged at 15,000 rpm for 5 rain at 4 °C. Usually forty tubes were put on the same rack, extracts of them were prepared at once and assayed immediately. For the screening assay, 5 ~tl of poly[d(A-[3H]U)] (3,000 ~ 9,000 cpm) was added to each tube and they were incubated at 32 °C for 30 rain. During this time, about 80% of the radioactivity was acid-solubilized by the extract prepared from wild-type cells. The reaction was stopped by putting the rack into liquid nitrogen for a few min and then adding 25 ttl of 20% trichloroacetic acid. After 10 rain at 0 °C, tubes were centrifuged. The supernatants were removed and counted in 4 ml of Bray's scintillation fluid. With this procedure we screened 50 colonies per day, and in all a total of 953 colonies were examined for uracil-DNA glycosylase activity.
Glucose-6-phosphate Dehydrogenase {G-6-PD) Assay The assay procedures for G-6-PD activity of crude extracts have been described before (Holliday and Tarrant 1972). The assay mixture consisted of 0.02 ml of enzyme and 0.78 ml of the reaction substrate (0.25 M Tris-HC1 (pH 8.6)containing 9.5 x 10 - 3 M glucose-6-phosphate (Na salt), 3.2 x 10 - 4 M NADP and 1.9 x 10 - 2 M MgC12).
Heat Inactivation Experiments Crude extracts stored at - 7 0 °C were used for this assay, and in all cases the specific activity after thawing was similar. The protein concentration was adjusted to 0.3 mg/ml and the extract was incubated at 45 °C. Aliquots (0.01 ml) were withdrawn at 1 or 2 min intervals and kept on ice until all the samples were taken (13 rain). Then they were added to 0.5 ml of the reaction substrate kept at 23 °C and were assayed for G-6-PD activity, as described above.
Y. Yamamoto and R. Holliday: Uracil Glycosylase Mutants of Ustilago
292
Random Products of Meiosis The procedure for the analysis of the random products of meiosis is discussed by Holliday (1974). Two crosses were analysed: 617 x YY43 (cross t) and 617 x YY7L (cross 2). Large numbers of teliospores were spread on plates of complete medium. After 2 days incubation at 26 °C, all the colonies from one plate were suspended with water and plated on complete medium agar (50 to 100 cells per plate). These plates were incubated at 26 °C. 84 colonies from cross 1 and 52 colonies from cross 2 were picked up, respectively, and were streaked on complete medium agar. After incubation for 3 days at 26 °C, they were examined for both auxotrophy by replica-plating and UracilDNA glycosylase activity by enzyme assay based on the acidsolubility of uracil.
Spontaneous Mutation Frequency At present there is no direct assay for GC ~ AT transitions in U. maydis, so forward mutation to chloroneb-resistance (Holliday et al. 1979) or cycloheximide-resistance was measured in several parallel populations. For chloroneb-resistanee, individual ung colonies were picked after 3 days growth, the cells were suspended in 0.2 ml water and were plated on YEP medium agar, containing 50/~g/ml chloroneb. In the case ofung- YY43164, the colonies were so srnall after 3 days that 10 were picked for each separate cell suspension. For cycloheximide-resistance, 10-12 tubes containing 5 ml complete medium, plus tetracycline (15 ~g/ml), were inoculated with 104 cells and groven 3 days to stationary-phase. Cells were suspended in 1 ml water and a 10-fold dilution were plated on nitrate minimum medium (supplemented with inositol and ammonium sulphate) containing 10 tzg/ml cycloheximide, and incubated for up to 2 weeks. Appropriate dilutions were made to obtain viable cell counts. q-
.
sure-cell method exhibited uracil-DNA glycosylase activity which catalyzed the release of free uracil bases from the radioactive polynucleotide substrate, poly[d(A-[3H]U)]. The product of this enzyme was free uracil rather than deoxyuridine phosphate (see Methods). The best assay conditions for this enzyme activity in the crude cell extracts were determined. The optimum temperature was 32 °C, although the activity was almost the same from 27 °C to 37 °C. The enzyme in the crude extract (2/~g p r o t e i n / 5 0 pl assay) was rather heat-labile, since only 60% of the original activity was retained after 30 min incubation at 32 °C. The glycosylase exhibited a broad pH optimum from 6.5 to 8.9 at 32 °C. Activity was slightly stimulated by dithiothreitol (1 or 5 mM) or NaC1 (15, 30 or 45 mM), and was unaffected by EDTA, even at 30 mM. The glycosylase was inhibited by uracil in the assay mixture: 6 mM uracil inhibited 92% of activity and 12 mM uracil inhibited 96%.
.
Sensitivity to Sodium Bisulphite Sodium bisulphite and sodium thioglycolate solutions were freshly prepared and sterilized by millipore filtration. Logarithmic growing ceils at 25 °C in YEP medium were harvested, washed with water twice, then were finally suspended with water. The reaction mixture was as follows: 0.2 M succinate buffer (pH 5.6), 0.05% sodium thioglycolate, 0.2 M NaHSO3, 0.8 M NaC1 and approximately 1 x 106 cells/ml, total volume 8 ml. For the control: 0.2 M succinate buffer (pH 5.5), 0.05% sodium thioglycolate, 1 M NaC1 and approximately 1 x 106 cells/ml, total volume 4 ml. Final pH of both mixtures was 5.3 and they were incubated at 26 °C. Aliquots (0.1 ml) were withdrawn at 20, 40 and 60 rain. They were diluted with 0.5 M Tris-HC1 (pH 8.6) and were examined for the survival. For the induced mutagenesis, aliquots (1 ml) were added to 1 ml of 1 M Tris-HCl (pH 8.6). The cells were spun down, washed and suspended with 0.5 ml of YEP medium. They were incubated at 25 °C till saturation, and were plated on YEP medium agar containing chloroneb (f.c. 50/~g/ml).
Results Uracil-DNA Glycosylase Activity in 6~ude Extracts Crude cell extracts prepared from Ustilago maydis, strain 483, by either freezing-and-thawing or the French-pres-
Isolation o f Uraeil-DNA Glycosylase Defective (ung- ) Mutants Ustilago maydis haploid strain 483 was mutagenized with MNNG and crude extracts prepared from 953 colonies (clones) were examined for their uracil-DNA glycosylase activity by a direct enzyme assay (see Methods). The crude extracts from 19 isolates released .. I.- 6 0 I-o ,,~
50
_o r7 4° i r~ IJA -.I Ill
3c
-.1
o o 20 o 10
dium and formed much smaller colonies than the parental strain on complete medium agar at 26 °C. YY43 was unable to grow at 32 °C, whereas all other ung- mutants could do so. YY7 and YY43 exhibited great variation in colony morphology (Fig. le and d) and both strains produced rapidly growing sectors which grew almost as fast as the parental strain 483. These faster growing isolates were much less variable than the small colonies from which they were derived, and they had recovered full glycosylase activity. It was difficult to handle YY7 for further analysis because there was so much variation in colony morphology. Therefore a colony of medium size which grew from part of an original small colony was picked up and was checked for glycosylase activity. Since this isolate showed low glycosylase activity and showed much less variation in colony morphology (Fig. 1f), it was designated YY7L and it was used, together with YY43, for meiotic analysis.
PROTEIN CONC. (I.Ig/ASSAY)
Fig. 2. Uracil-DNA glycosylase activity of strain YY7L (ung-), YY43 (ung-), 483 (ung+) and 617 (ung+). The assay mixture (50 ~1) contains 10 mM phosphate buffer (pH 7.0), 30 mM EDTA (pH 7.5) 15 mM NaC1, 1 mM DTT, poly[d(A-[3H]U)] (3,900 cpm), and various amounts of the crude extract (0, 5, 10 and 15 ~g). They were incubated at 32 °C for 20 rain and the radioactivity of the acid-soluble fractions was determined. H strain 483, o-o strain 617, 0 - 0 YY43, and A--A YY7L. G-6-PD activity of these crude extracts, strain 617, YY7L and YY43, was 82% 92% and 66% of that of strain 483, respectively
ung + revertants survive better than the original ungstrain. The problem is exacerbated by the genetic instabilitiy of the ung- strains (see below).
Variability o f ung- Mutants YY47 showed striking features of colony variability, namely, sectored white colonies and non-sectored brown and white colonies (Fig. lb). However, this mutant was lost (see above) and was not studied further. YY34 showed two types of colony morphology and some colonies had sectors (Fig. lc). Initially, cultures from each type of colony had less than 14% of the glycosylase activity of the parental strain. However, after several cycles of single colony purification, it was found that the isolates regained uracil-DNA glycosylase activity and the differences in colony morphology were no longer clear. We could not find a phenotype which allowed us to pick up ung- mutants selectively from the colonies of strain YY34. Whereas YY34 grew at the same rate as 483 at 26 °C, YY7 and YY43 grew very slowly in liquid complete me-
Genetic Analysis o f ung- Mutants YY43 and YY7L were crossed to 617 (see Table 1 for genotypes) and the random products of meiosis were examined. The uracil glycosylase activities of 617, YY43, YY7L and 483 are shown in Fig. 2. Both YY7L and YY43 exhibited 7 to 8% activity of ung + strains. However, it has not been clear yet whether the residual activity of both ung- strains was due to the leakiness of the mutations or to the revertants which appeared during the culture. 1) Cross YY43 x617. After meiosis, 84 independent colonies were isolated and scored for auxotrophic markers, colony morphology and the specific activity of uracilDNA glycosylase. 61 colonies were found out to be ung + and 23 to be ung-, among which 20 colonies had less than 30% activity and 3 colonies had about 57% activity of the other 61 colonies. The segregation of auxotrophic markers (nic, ade, inos, nar and pan) was close to 1 : 1 in both ung + and ung- colonies, and the expected linkage between nic and inos was observed in 84 colonies (the recombination frequency between them was 17%). It is probable that during the growth of teliospore colonies ung + isolates had a selective advantage. There was also the possibility that some ung + isolates might be just revertants from ung- colonies selected during the experimental procedure from picking up the meiotic segregants until crude extracts were made. For this reason, it is more reliable to score the segregation of auxotrophic markers amongst the ung- isolates, even though there were only 23. From this small sample there was no evidence of close linkage between ung and any of the five auxotrophic markers. All the ung- segregants had variable colony morphology. Some were similar to the
Y. Yamamoto and R. Holliday: Uracil Glycosylase Mutants of Ustilago 483 ung + 1(~3
I
YY43-164 ung t
R-2 ung +
295
(a)
I
100
_> >O
0 < O a < rc
o
o~o 0 0 ~
h
I
5
8 8
°
,,o,
UraciI-DNA Glycosylase Defective Mutants of
Ustilagomaydis
Yoko Yamamoto 1 and Robin Holliday Genetics Division, National Institute for Medical Research, The Ridgeway,Mill Hill, London, NW7 1AA, England
Summary. Uracil-DNA glycosylase activity can be measured in cell-free extracts of Ustilago maydis and in its general properties the enzyme resembles the same glycosylase from other organisms. A rapid assay was used to screen nearly 1,000 clones from cells treated with N-methyl-N'-nitro-N-nitrosoguanidine, and five isolates had < 1 - 3 0 % of wild-type enzyme activity. All these ungstrains were unstable since they frequently recovered normal enzyme activity (ung+), and two were lost for this reason. The ung- strains were also variable in morphology, and two which were slow growing, produced faster growing ung + sectors. However, in preliminary experiments, ung- strains were shown to be only weak mutators. Two ung- mutants were crossed to wild-type and ung- progeny were recovered, which were also variable in morphology and reverted to ung +. It was found that a standard metabolic enzyme, glucose-6-phosphate dehydrogenase, was significantly more heat-labile in ung- strains than wild-type. The results indicate that the presence of uracil in DNA may have more severe physiological effects in eukaryotes than in prokaryotes. Over 1,000 isolates form Chinese hamster ovary (CHO) cells treated with ethyl methane sulphonate, were also tested for uracil-DNA glycosylase activity, but no stable ungstrain was recovered. Key words: Uracil-DNA glycosylase - Ustilago maydis Instability - G-6-PD - CHO cells
Offprint requests to: R. Holliday 1 Present address: Department of Radiation Biophysics, Kobe University School of Medicine, Kusunoki-cho7-5-1 Chuo-ku, Kobe 650, Japan
Introduction
Uracil can be introduced into DNA in two ways. One is the direct deamination of cytosine residues in DNA. This can occur spontaneously at a low rate under physiological conditions (Lindahl and Nyberg 1974) and can be induced by specific reagents, such as nitrous acid or bisulphite (Freese 1971; Hayatsu 1976). The other depends on the misincorporation of dUMP into DNA instead of dTMP by DNA potymerase during DNA replication in vivo (Tye et al. 1977; Tye et al. 1978;Tye and Lehamn 1977; Brynolf et al. 1978; Grafstrom et al. 1978). Uracil-DNA glycosylase recognizes uracil residues in DNA, whichever way it is introduced, and catalyzes the hydrolysis of uracil-deoxyribose bonds so that free uracil bases are released from DNA. Since the discovery of this enzyme activity in E. coli by Lindahl (1974),uracil-DNA gtycosylase has been found widely from prokaryotes to eukaryotes; for example, B. subtilis, yeast, wheat germ, human cells (reviewed by Lindahl 1979), as well as cultured Chinese hamster ovary cells (Y. Yamamoto and P. A. Jeggo, unpublished data). Uracil-DNA glycosylase has been highly purified from E. coli (Lindahl et al. 1977) and partially purified from B. subtilis (Cone et ai. 1977), calf thymus (Talpaert-Borl6 et al. 1979), human leukaemia cells (Charradonna and Cheng 1980), and HeLa cells (Krokan and Wittwer 1981). All of these enzymes showed similar properties; they are of low molecular weight, have a strict specificity for uracil in DNA and are fully active in the presence of EDTA. It would be expected that the conversion of cytosine to uracil would lead to transition mutations (C-G -+ T - A ) during DNA replication, if the uracil was not previously removed. Mutants lacking uracil-DNA glycosylase should therefore be mutable. The incorporation of dUMP in place of dTMP would not produce mutations, but might have other physiological effects, if the uracil remained in
Y. Yamamoto and R. Holliday: Uracil Glycosylase Mutants of UstiIago
290 Table 1. Ustilago maydis strains Strain
Genotype or phenotype
Source
483 617 YY47 YY57 YY34 YY7 YY7L YY43 YY43-164 R1 R2
narl-1, inosl-3, a2b 2 panl-1, adel-1, nicl-1 a, b, The same as 483, but ung The same as 483, but ung The same as 483, but ung The same as 483, but ung The same as YY7, but the colony size was bigger The same as 483, but ung pan1-1, nat1-1, ions1-3, nicl-1, ung, a,b, The same as YY43-164, but ung + The same as YY43-164, but ung +
MNNG mutagenesis MNNG mutagenesis MNNG mutagenesis MNNG mutagenesis Spontaneously isolated from YY7 MNNG mutagenesis Meiotic segregant from the cross 617 x YY43 Spontaneously isolated from YY43-164 Spontaneously isolated from YY43-164
DNA. So far, the biological significance of the enzyme has only been examined in prokaryotes. Mutants deficient in this enzyme activity were isolated by direct enzyme assay from both B. subtilis (Makino and Munakata 1977) a n d E . coli K-12 (Duncan et al. 1978). The B. subtilis m u t a n t ( u r g - ) , which had less than 1% of the wildtype enzyme, could grow at normal rate and exhibited normal sensitivity to nitrous acid, whereas one of the E. coli mutants (ung.1), which showed only 0.02% activity of its parental strain, was more sensitive to nitrous acid and bisulphite than wild-type (Da Roza et al. 1977; Simmons and Friedberg 1979). It could grow normally and was also a weak mutator. Mutants ofE. coli deficient in dUTPase have also been isolated, and these incorporate significant amounts of uracil into DNA (Hochhauser and Weiss 1976; Tye et al. 1977). From the study of both the ung m u t a n t and the dUPTase defective m u t a n t , it is now believed that uracil-DNA glycosylase has an important role in the repair of uracil in DNA, whether or n o t a uracil-guanine mismatch is present (Duncan and Weiss 1978; Duncan and Miller 1980). It is important to study the biological function of uracii-DNA glycosylase in eukaryotic systems. For this purpose, we selected the simple eukaryotic organism, Ustilago maydis, which can grow like yeast and has m a n y advantages for biochemical and also genetical studies (Holliday 1974). In this paper, we first show that crude extracts from Ustilago maydis exhibit uracil-DNA glycosylase activity, and then describe the isolation and the characterization of mutants deficient in this enzyme activity.
Materials and Methods Chemicals Poly[d(AT)] was purchased from Boehringer and Mannheim. Poly[d(AT)] was dissolved with 50 mM Tris-HC1 (pH 7.5) f.c.
(50OD26o) and stored at -20°C. [3H]dUTP (11Ci/mmot, 1 mCi/ml, Cat. No.TRK351) was purchased from the Radio Chemical Centre. The highly purified E. eoli DNA polymerase I was kindly given by Dr. G. T. Yarranton. Sodium bisulphite and sodium thioglycolate were purchased from Nakarai Chemicals Ltd. NADP and D-glueose-6-phosphate were obtained from Kohjin Co. Ltd. and Sigma Chemical Co., respectively. Chioroneb (1,4 dichloro-2,5-dimethoxybenzene) was a gift from E. I. Dupont de Nemours and Co., Delaware, U.S.A.
Media Complete medium, nitrate minimal medium and yeast extract peptone (YEP) medium were prepared as described previously (Holliday 1974).
Strains The strains of Ustilago maydis used for the experiments in this paper are listed up in Table 1.
Polyld(A-[ 3H lU) ] Preparation Poly[d(A-[3HIU)] was synthesized by E. coli DNA polymerase I in vitro. The reaction mixture (100 #1) contained 66 mMTrisacetate buffer (pH 7.4), 6 mM MgC12, 5 mM DTT, 3 nmole dATP, 0.3 nmole dUTP, [3HIdUTP (2 #Ci, 0.18 nmole), poly[d(AT)] (0.25 OD26o) and 51,1 of E. coli DNA polymerase (0.5 unit; one unit was defined as the activity which incorporated 10 nmole total nucleotides into DNA in 30 min at 37 ° C). Ten assay tubes (each tube contained 100 #1 of the reaction mixture) were incubated at 37 °C for 20 min, then sodium dodecyl sulphate (f.c. 1%) was added. The reaction mixtures were gathered, treated with phenol and then DNA was precipitated by ethanol. DNA was resuspended with lml of 10 mM phosphate buffer (pH 7.0) - 10 mM EDTA (ph 7.5) and was dialyzed against the same buffer. The net count of 5 t~l of this DNA solution was about 130,000 cpm. The DNA solution was used after appropriate dilution with poty[d(AT)] (2.5 OD260).
Y. Yamamoto and R. Holliday: Uracil Glycosylase Mutants of Ustilago
Uracil-DNA Glycosylase Assay 1) Crude Extract Preparation
Freezing-and-Thawing Method. Fresh colonies were inoculated into 5 ml or 10 ml of complete medium and were shaken at 26 °C for 24 h. After 24 h, ung+ strains (483 and 617) and YY34 were grown to saturation (approximately 2 x 10 s cells/m), while the cell concentration of cultures of YY7L and YY43 was around 1 x 107 cells/ml. Cells were harvested, washed with 10 ml of water twice, washed with 500 tal of buffer (50 mM Tris-HCl (pH 7.5) - 1 mM DTT - 1 mM EDTA - 10% glycerol) and finally resuspended with 50 to 200 ~1 of the same buffer. The cells were broken by repeating freezing-and-thawing four times and immediately centrifuged at 15,000 rpm for 20 min at 4 °C. The supernatants were re-centrifuged under the same conditions and were stored at - 7 0 °C.
French-Pressure-Cell Method. Freshly prepared cells (2 x 109 cells) were washed and resuspended with 10 ml of buffer (50 mM phosphate buffer (pH 7.0) - 1 mM DTT - 1 mM EDTA - 10% glycerol). Cells were broken by passing through a French-pressure-cell at 18,000 psi twice and were centrifuged twice as described above. The crude extracts were stored at - 7 0 °C. The protein concentrations of crude extracts were determined by the method of Lowry et al. (1951).
2) Assay of the Uracil-DNA Glycosylase The assay procedures were basically the same as the methods described by Duncan et al. (1978). The crude extract was incubated at 32 °C in an assay mixture (50 tsl), which contained 50 mM phosphate buffer (pH 7.0) (or 50 mM Tris-HC1 (pH 7.5)), 30 mM EDTA (pH 7.5, adjusted with NaOH), 1 mM DTT and poly[d(A-[3H]U)]. Uracil-DNA glycosylase activity was usually measured by the assay based on the acid-solubility of uracil. The acid-soluble products obtained by the crude extract of strain 483 (ung +) were analysed by paper chromatography in the solvent system; isobutyric acid : water : 0.1 M sodium EDTA :conc. ammonia:toluene; 160:22 : 3 : 2:20 (Reeves et al. 1969). 80% of the total counts of the acid-soluble fraction were located at free uracil position, not at deoxyuridine or deoxyuridine monophosphate. The reaction products were also examined by Dowex-1. After incubation, 10 #1 of 2% sodium dodecyl sulphate was added to the reaction mixture and applied to the Dowex column. Almost the same amount of radioactivity as that of the acid-soluble fraction was eluted with water.
Screening of u ng- Mu tan ts Wild-type strain 483 was mutagenized with N-methyl-N'-nitro-Nnitrosoguanidine (MNNG) by using two different methods: (1) CeBs from the logarithmic growth phase in complete medium at 32 °C were spun down and resuspended with 9 ml of water (Lc. 1.25 x 107 cells/ml). One ml of freshly prepared MNNG solution (500 ~g/ml in 25% ethanol) was added to the ceil suspension and it was shaken at 32 °C for 20 rain (S/S 0 = 1.6 x 10-4). The reaction was stopped by adding 1 ml of cold 0.5 M Tris-HC1 (pH 7.5) and the cells were spun down at room temperature for 5 rain. The cells were washed with 10 ml of 50 mM Tris-HC1 (pH 7.5) and then resuspended with 10 ml of complete medium. Each 1 ml of
291
this suspension was added to 9 ml of complete medium (total, two tubes) and incubated at 32 °C until saturation (stationaryphase). A portion of the culture from each tube was diluted and spread on complete medium agar and incubated at 32 °C. 95 colonies were picked from each plate and transferred to complete medium agar. They were incubated at 32 °C for several days and kept at 4 °C as master plates. (2) Cells were mutagenized in almost the same way as method (1). 4.5 ml of the cell suspension was added to 0.5 ml of MNNG solution (500 ~g/ml in 25% ethanol) and incubated at 32 °C for 5 rain (S/S 0 = 29%). The reaction was stopped by adding 5 ml of cold 50 mM phosphate buffer (pH 7.0). The cells were washed with the same buffer and, after the appropriate dilutions, were immediately plated on complete medium agar. The plates were incubated at 26 °C in order to get temperature-sensitive uracil-DNA glycosylase mutants. 800 colonies were picked up and transferred to complete medium agar. They were incubated at 26 °C for several days and kept at 4 °C as master plates. The screening assay for ung- mutants was based on the acidsolubility of uracil, using the assay procedure developed by Duncan et al. (1978) with minor modifications. Each colony from the master plates was inoculated into 1 ml of complete medium in trays (16 mm well diameter and 20 mm well depth, 24 cultures per tray) and was grown for two days at 26 °C. Wildtype cells and most of the mutagenized cells could grow to saturation ( - 108 cells/mt). Each culture was transferred to a small Eppendorf tube. The cells were spun down at 4 °C, resuspended with 40 ~1 of the buffer (50 mM phosphate buffer (pH 7.0), 1 mM DTT, 1 mM EDTA) and then broken by freezing-and-thawing four times. They were immediately centrifuged at 15,000 rpm for 5 rain at 4 °C. Usually forty tubes were put on the same rack, extracts of them were prepared at once and assayed immediately. For the screening assay, 5 ~tl of poly[d(A-[3H]U)] (3,000 ~ 9,000 cpm) was added to each tube and they were incubated at 32 °C for 30 rain. During this time, about 80% of the radioactivity was acid-solubilized by the extract prepared from wild-type cells. The reaction was stopped by putting the rack into liquid nitrogen for a few min and then adding 25 ttl of 20% trichloroacetic acid. After 10 rain at 0 °C, tubes were centrifuged. The supernatants were removed and counted in 4 ml of Bray's scintillation fluid. With this procedure we screened 50 colonies per day, and in all a total of 953 colonies were examined for uracil-DNA glycosylase activity.
Glucose-6-phosphate Dehydrogenase {G-6-PD) Assay The assay procedures for G-6-PD activity of crude extracts have been described before (Holliday and Tarrant 1972). The assay mixture consisted of 0.02 ml of enzyme and 0.78 ml of the reaction substrate (0.25 M Tris-HC1 (pH 8.6)containing 9.5 x 10 - 3 M glucose-6-phosphate (Na salt), 3.2 x 10 - 4 M NADP and 1.9 x 10 - 2 M MgC12).
Heat Inactivation Experiments Crude extracts stored at - 7 0 °C were used for this assay, and in all cases the specific activity after thawing was similar. The protein concentration was adjusted to 0.3 mg/ml and the extract was incubated at 45 °C. Aliquots (0.01 ml) were withdrawn at 1 or 2 min intervals and kept on ice until all the samples were taken (13 rain). Then they were added to 0.5 ml of the reaction substrate kept at 23 °C and were assayed for G-6-PD activity, as described above.
Y. Yamamoto and R. Holliday: Uracil Glycosylase Mutants of Ustilago
292
Random Products of Meiosis The procedure for the analysis of the random products of meiosis is discussed by Holliday (1974). Two crosses were analysed: 617 x YY43 (cross t) and 617 x YY7L (cross 2). Large numbers of teliospores were spread on plates of complete medium. After 2 days incubation at 26 °C, all the colonies from one plate were suspended with water and plated on complete medium agar (50 to 100 cells per plate). These plates were incubated at 26 °C. 84 colonies from cross 1 and 52 colonies from cross 2 were picked up, respectively, and were streaked on complete medium agar. After incubation for 3 days at 26 °C, they were examined for both auxotrophy by replica-plating and UracilDNA glycosylase activity by enzyme assay based on the acidsolubility of uracil.
Spontaneous Mutation Frequency At present there is no direct assay for GC ~ AT transitions in U. maydis, so forward mutation to chloroneb-resistance (Holliday et al. 1979) or cycloheximide-resistance was measured in several parallel populations. For chloroneb-resistanee, individual ung colonies were picked after 3 days growth, the cells were suspended in 0.2 ml water and were plated on YEP medium agar, containing 50/~g/ml chloroneb. In the case ofung- YY43164, the colonies were so srnall after 3 days that 10 were picked for each separate cell suspension. For cycloheximide-resistance, 10-12 tubes containing 5 ml complete medium, plus tetracycline (15 ~g/ml), were inoculated with 104 cells and groven 3 days to stationary-phase. Cells were suspended in 1 ml water and a 10-fold dilution were plated on nitrate minimum medium (supplemented with inositol and ammonium sulphate) containing 10 tzg/ml cycloheximide, and incubated for up to 2 weeks. Appropriate dilutions were made to obtain viable cell counts. q-
.
sure-cell method exhibited uracil-DNA glycosylase activity which catalyzed the release of free uracil bases from the radioactive polynucleotide substrate, poly[d(A-[3H]U)]. The product of this enzyme was free uracil rather than deoxyuridine phosphate (see Methods). The best assay conditions for this enzyme activity in the crude cell extracts were determined. The optimum temperature was 32 °C, although the activity was almost the same from 27 °C to 37 °C. The enzyme in the crude extract (2/~g p r o t e i n / 5 0 pl assay) was rather heat-labile, since only 60% of the original activity was retained after 30 min incubation at 32 °C. The glycosylase exhibited a broad pH optimum from 6.5 to 8.9 at 32 °C. Activity was slightly stimulated by dithiothreitol (1 or 5 mM) or NaC1 (15, 30 or 45 mM), and was unaffected by EDTA, even at 30 mM. The glycosylase was inhibited by uracil in the assay mixture: 6 mM uracil inhibited 92% of activity and 12 mM uracil inhibited 96%.
.
Sensitivity to Sodium Bisulphite Sodium bisulphite and sodium thioglycolate solutions were freshly prepared and sterilized by millipore filtration. Logarithmic growing ceils at 25 °C in YEP medium were harvested, washed with water twice, then were finally suspended with water. The reaction mixture was as follows: 0.2 M succinate buffer (pH 5.6), 0.05% sodium thioglycolate, 0.2 M NaHSO3, 0.8 M NaC1 and approximately 1 x 106 cells/ml, total volume 8 ml. For the control: 0.2 M succinate buffer (pH 5.5), 0.05% sodium thioglycolate, 1 M NaC1 and approximately 1 x 106 cells/ml, total volume 4 ml. Final pH of both mixtures was 5.3 and they were incubated at 26 °C. Aliquots (0.1 ml) were withdrawn at 20, 40 and 60 rain. They were diluted with 0.5 M Tris-HC1 (pH 8.6) and were examined for the survival. For the induced mutagenesis, aliquots (1 ml) were added to 1 ml of 1 M Tris-HCl (pH 8.6). The cells were spun down, washed and suspended with 0.5 ml of YEP medium. They were incubated at 25 °C till saturation, and were plated on YEP medium agar containing chloroneb (f.c. 50/~g/ml).
Results Uracil-DNA Glycosylase Activity in 6~ude Extracts Crude cell extracts prepared from Ustilago maydis, strain 483, by either freezing-and-thawing or the French-pres-
Isolation o f Uraeil-DNA Glycosylase Defective (ung- ) Mutants Ustilago maydis haploid strain 483 was mutagenized with MNNG and crude extracts prepared from 953 colonies (clones) were examined for their uracil-DNA glycosylase activity by a direct enzyme assay (see Methods). The crude extracts from 19 isolates released .. I.- 6 0 I-o ,,~
50
_o r7 4° i r~ IJA -.I Ill
3c
-.1
o o 20 o 10
dium and formed much smaller colonies than the parental strain on complete medium agar at 26 °C. YY43 was unable to grow at 32 °C, whereas all other ung- mutants could do so. YY7 and YY43 exhibited great variation in colony morphology (Fig. le and d) and both strains produced rapidly growing sectors which grew almost as fast as the parental strain 483. These faster growing isolates were much less variable than the small colonies from which they were derived, and they had recovered full glycosylase activity. It was difficult to handle YY7 for further analysis because there was so much variation in colony morphology. Therefore a colony of medium size which grew from part of an original small colony was picked up and was checked for glycosylase activity. Since this isolate showed low glycosylase activity and showed much less variation in colony morphology (Fig. 1f), it was designated YY7L and it was used, together with YY43, for meiotic analysis.
PROTEIN CONC. (I.Ig/ASSAY)
Fig. 2. Uracil-DNA glycosylase activity of strain YY7L (ung-), YY43 (ung-), 483 (ung+) and 617 (ung+). The assay mixture (50 ~1) contains 10 mM phosphate buffer (pH 7.0), 30 mM EDTA (pH 7.5) 15 mM NaC1, 1 mM DTT, poly[d(A-[3H]U)] (3,900 cpm), and various amounts of the crude extract (0, 5, 10 and 15 ~g). They were incubated at 32 °C for 20 rain and the radioactivity of the acid-soluble fractions was determined. H strain 483, o-o strain 617, 0 - 0 YY43, and A--A YY7L. G-6-PD activity of these crude extracts, strain 617, YY7L and YY43, was 82% 92% and 66% of that of strain 483, respectively
ung + revertants survive better than the original ungstrain. The problem is exacerbated by the genetic instabilitiy of the ung- strains (see below).
Variability o f ung- Mutants YY47 showed striking features of colony variability, namely, sectored white colonies and non-sectored brown and white colonies (Fig. lb). However, this mutant was lost (see above) and was not studied further. YY34 showed two types of colony morphology and some colonies had sectors (Fig. lc). Initially, cultures from each type of colony had less than 14% of the glycosylase activity of the parental strain. However, after several cycles of single colony purification, it was found that the isolates regained uracil-DNA glycosylase activity and the differences in colony morphology were no longer clear. We could not find a phenotype which allowed us to pick up ung- mutants selectively from the colonies of strain YY34. Whereas YY34 grew at the same rate as 483 at 26 °C, YY7 and YY43 grew very slowly in liquid complete me-
Genetic Analysis o f ung- Mutants YY43 and YY7L were crossed to 617 (see Table 1 for genotypes) and the random products of meiosis were examined. The uracil glycosylase activities of 617, YY43, YY7L and 483 are shown in Fig. 2. Both YY7L and YY43 exhibited 7 to 8% activity of ung + strains. However, it has not been clear yet whether the residual activity of both ung- strains was due to the leakiness of the mutations or to the revertants which appeared during the culture. 1) Cross YY43 x617. After meiosis, 84 independent colonies were isolated and scored for auxotrophic markers, colony morphology and the specific activity of uracilDNA glycosylase. 61 colonies were found out to be ung + and 23 to be ung-, among which 20 colonies had less than 30% activity and 3 colonies had about 57% activity of the other 61 colonies. The segregation of auxotrophic markers (nic, ade, inos, nar and pan) was close to 1 : 1 in both ung + and ung- colonies, and the expected linkage between nic and inos was observed in 84 colonies (the recombination frequency between them was 17%). It is probable that during the growth of teliospore colonies ung + isolates had a selective advantage. There was also the possibility that some ung + isolates might be just revertants from ung- colonies selected during the experimental procedure from picking up the meiotic segregants until crude extracts were made. For this reason, it is more reliable to score the segregation of auxotrophic markers amongst the ung- isolates, even though there were only 23. From this small sample there was no evidence of close linkage between ung and any of the five auxotrophic markers. All the ung- segregants had variable colony morphology. Some were similar to the
Y. Yamamoto and R. Holliday: Uracil Glycosylase Mutants of Ustilago 483 ung + 1(~3
I
YY43-164 ung t
R-2 ung +
295
(a)
I
100
_> >O
0 < O a < rc
o
o~o 0 0 ~
h
I
5
8 8
°
,,o,
