Mol Gen Genet (t986) 204:496-504 © Springer-Verlag1986

Heterothallic mating type switching in Saccharomyces cerevisiae is RAD 52 dependent Robert Schiestl* Institute for Tumorbiology - Cancer Research, University of Vienna, Borschkegasse 8 a, A-1090 Wien, Austria and Department of Genetics, University of Alberta, Edmonton, Alberta T6G 2E9, Canada Summary. Mating type interconversion of the yeast Saceharomyces cerevisiae is mediated by intrachromosomal gene conversion. Whereas homothallic switching is initiated by an endonuclease that produces a D N A double-strand cut within M A T, heterothallic strains lack this activity. In order to identify functions essential for initiation and realisation of heterothallic switching, repair-deficient strains carrying the rad52 or the rad3 mutation were constructed and tested for spontaneous and induced heterothallic switching frequencies. The wild type R A D 5 2 function is essential for spontaneous and induced switching as well as for the intrachromosomal crossing over which produces a deleted ring chromosome III. The rad3 mutation had almost no influence on spontaneous or X-ray induced switching, but it does reduce induction by ultraviolet radiation. The data are interpreted to indicate that heterothallic switching is accomplished via recombinogenic repair, perhaps of a double-strand break. The conversion event as well as the crossing over event leading to a change in mating type are equally affected by the rad52 mutation and therefore perhaps associated. Key words: Heterothallic mating type switching - Saccharomyces cerevisiae - R A D 5 2 - R A D 3

Introduction Haploid cells of the yeast Saccharomyces cerevisiae can be of mating type a or ~. Cells of the opposite type can mate with each other to produce a/~ diploids which can sporulate, but are unable to mate. The mating type is determined by the M A T locus on chromosome III. To show an a or phenotype expression of a or ~ information is required at MAT, whereas diploids are heterozygous for M A T and expression of both alleles is needed for sporulation (review: Herskowitz and Oshima 1981). In addition, two silent mating type loci are present on chromosome III, HMLc~ and H M R a respectively, left and right from M A T . Haploid cells can switch from one mating type to the other by unidirectional intrachromosomal gene conversion, replacing the sequence at M A T by one from a silent locus * Present address: Department of Biology, University of Roches-

ter, Rochester, NY 14627, USA Abbreviations: DSB, Double Strand Break; nm, nonmater

(Oshima and Takano 1971 ; Hicks et al. 1979). Homothallic strains can interconvert their mating types as often as every cell division. They contain the H O gene and a site-specific endonuclease, YZendo, which cuts exclusively at the M A T locus and thus initiates switching (Kostriken et al. 1983). Heterothallic strains, on the other hand, interconvert only about once per 10 6 cells, contain the ho allele and lack any detectable YZendo activity. The R A D 5 2 function is involved in double-strand-break repair. The mutation rad52 was originally isolated on the basis of its X-ray sensitivity (cf. Game and Mortimer 1974). The sporulation frequency of rad52 is also much reduced. In addition, it was shown that the R A D 5 2 gene is essential for meiotic recombination (Game et al. 1980), a normal level of spontaneous, as well as UV- and 7-ray induced mitotic recombination (Prakash et al. 1980), and for homothallic mating type switching (Malone and Esposito 1980; Weiffenbach and Haber 1981). Homothallic strains which do not carry any information of the opposite mating type, and therefore cannot change their type, accumulate DSBs in M A T D N A (Strathern et al. 1982). It is interesting to note that some other recombination events are R A D 5 2 independent; integration (Orr-Weaver et al. 1981) and excision (Jackson and Fink 1981) of circular plasmids, spontaneous sister-chromatid exchange (Prakash and Taillon-Miller 1981 ; Zamb and Petes 1981), and A--A recombination of T Y elements (M. Ciriacy, personal communication). A high percentage of recombination events between homologs requires the R A D 5 2 function, whereas a low percentage is independent (Malone and Esposito 1980; Prakash et al. 1980). The R A D 3 gene product is involved in excision repair (Cox and Game 1974; Prakash 1977). The rad3-2 mutant has been reported to exhibit an elevated frequency of spontaneous and a reduced frequency of UV-induced mitotic gene conversion (Kern and Zimmermann 1978). Deletion of the R A D 3 gene results in inviability (Naumovski and Friedberg 1983; Higgins et al. 1983). Schiestl and Wintersberger (1982) developed a system for measuring the frequency of heterothallic switching. Cells in which the mating type was changed were trapped by mating to tester cells with complementing markers and the diploids were selected on a special medium. An enhancement of the switching frequency was demonstrated by treating cells with DNA-damaging agents (Schiestl and Wintersberger 1983; Wintersberger and Schiestl 1982). In order to learn more about the mechanism responsible for this

497 process, repair-deficient strains were used to test their influence on heterothallic switching. The rad52-1 and rad3-2 mutations were chosen because they are deficient in relatively well known repair systems. Materials and methods

Strains. The yeast strains used are indicated in Table 1. E. coli strain R R I was obtained from M. Ciriacy. Media. Growth, minimal and sporulation media were pre-

pared according to Sherman et al. (1983). Determination o f the frequency o f heterothallic switching.

The principle of the system used is as follows: two strains with complementary auxotrophic markers are needed. Both strains are of the same mating type so that switching of the mating type of one cell and subsequent mating with a cell of the other strain results in a diploid prototrophic colony which can be selected. To test influences on this process accurately, switching is limited to one strain. One strain, called the tester strain, contains information of only one mating type in the genome. The other strain called the switching strain, possesses ~ information at the silent locus H M L and a or ~ type information at M A T and a information at H M R . If a diploid prototrophic colony capable of sporulation occurs on selective medium from a mixture of two strains of the same type, it must have resulted from a change of the mating type of a cell from the switching strain. To demonstrate genetically that the information is duplicated in the switching process, the tester strain N1/6-8C contains the marl mutation. This mutation permits expression of all three mating type loci. A haploid cell with the marl mutation and a as well as c~ information exhibits a nonmater phenotype (Klar et al. 1979). Tetrad analyses with the selected diploid colonies reveal the segregation pattern of mating types and thus indicate the number of c~ genes in a diploid cell (Schiestl and Wintersberger 1982).

Segregants of the crosses $9, $23 and $25 (Table 1) were tested for markers. Appropriate R A D and rad segregants, as well as strain RM38-7D, were used as switching strains with the tester strains N1/6-8C and J39~. Cells of the switching strains were grown to stationary phase in liquid complete medium, freed from aggregates by sonication, and, after centrifugation and washing, suspended in 0.1 M potassium phosphate. The cells were irradiated with various doses of X-radiation and ultraviolet radiation, and then incubated in liquid complete medium either until buds became visible or for one or two generations thereafter. This treatment reduced any unequal mating chances due to different lag periods of the control and irradiated cells. The number of survivors was determined by the plating of cells onto complete medium. 4 x 108 stationary phase tester cells were mixed with 2-8 x 107 switching cells per millilitre. 0.25 ml of the suspension was plated per petri dish containing selective medium for the growth of prototrophic diploid colonies. It was necessary to supplement the selection medium with 5% complete medium so that mating of the switched cells could take place. Diploid colonies were counted after incubation for five days at 30 ° C. Irradiation o f cells. For the application of X-rays, stationary phase cells were suspended in 0.1 M potassium phosphate at a density of about 5 x 107 cells/ml. Cells were irradiated with a Phillips RT 100 X-ray source ( l . 7 m m Al-filter, 100 KV, 8 mA). UV-light irradiation was carried out with a Camag low-pressure mercury lamp (Switzerland) providing most of its radiant energy at 254 nanometers. Suspensions of 2 x 107 cells/ml in 0.1 M potassium phosphate (pH 7) were used. In order to determine the presence of the rad52 and rad3 mutations in spore clones, complete medium plates containing the cells were irradiated immediately after replica plating, and incubated afterwards to test for growth ability. In the case of the rad52 mutation, a dose of 20-30 Krads of X-rays and for rad3 a fluence of 50 Jm-2 UV light incident radiation were found to be appropriate.

Table l. Strains used

Strain

Genotype

Source

N1/6-8C

MA Ta HMLa HMRa metx marl ural ade3 aro2 his4 ho

This study

S23/2a-9B

MATa HMLc~ HMRa lys5 ade2-119 trp5 his3-11,15 leu2-3,112 ho

This study

RM38-7D

MATa leul trp5 ade2 lys2 his7 tyrl ura3 rad52-1 ho

Malone

$23

MA Ta met13 lys5 ural ade2 trp5 ilvl LEU2 HIS3 rad52-1ho MA T~ MET13 LYS5 URA1 ADE2 TRP5 ILV1 leu2-3,112 his3-11,15 RAD52 ho

This study

$25

MA Ta leu2-3,112 his3-11,15 ADE2 TRP5 ILV1 RAD3 ho MATo~ LEU2 HIS3 ade2 trp5 ilvl rad3-2 ho

This study

J39a

MA Ta HMLc~ HMRo~ cryl thr4 ura3 his4 leu2 ho

Klar

J39c~

MA T~ HMLc~ HMR~ cryl thr4 ura3 his4 leu2 ho

This study

$9

MATa leu2-3,112 his3-11,15 trpl LYS5 URA1 ILV1 RAD52 MATc~ leu2-3,112 his3-11,15 TRP1 lys5 ural ilvl rad52-1

This study

P502

MATa ade2 leul trp5 metl3 lys2 tyrl his7 HO MATc~ ade2 leul trp5 metl3 lys2 tyrl his7 HO

Marmiroli

HR161-1C

MATa ura3-52 leu2-3,112 trpl HO-lacZ

Herskowitz

If not indicated, strains contain HMLe and HMRa as determined by Southern blotting. To obtain the diptoids, several rounds of backcrosses were performed with our laboratory strains. The rad52-1 allele was obtained from R Malone, and rad3-2 was from F. Zimmermann

498 Genetic analyses. Isolation, sporulation, dissection of asci, and testing of spore colonies for markers and mating type were performed according to standard methods (cf. Sherman et al. 1983). The test for the presence of M A T a or MA T~ is as follows: This test was based upon the observation that MA Ta HO rad52 cells do not change their mating type, but divide mitotically, whereas MA Tg HO rad52, as well as M A T ~ H O rad52 cells die (Malone and H y m a n 1983). Several of the rad52 switching strains were crossed with strain P502 HO and tetrad analyses were done. Presence of the gt allele was indicated if the number of a rad52 and e rad52 segregants were equal; presence o f the a allele was indicated if the number of a rad52 was greater than the number of e rad52 segregants. Construction of a strain containing c~information at all three mating type loci. Strain J39 containing HML~, M A T a or H M R ~ was transformed with the plasmid YEpHO (Jensen et al. 1983) which shows HO activity. Colonies with c~mating type which subsequently have lost the plasmid were isolated, checked for presence of the genes HMLc~, M A T ~ and HMRo~ by Southern blotting and used as "tester" strain. Southern blotting. Southern blotting was carried out for the determination of the mating type loci on chromosome III. A plasmid containing pBR322 and mata was subcloned (Schiestl and Wintersberger 1983). Yeast D N A extraction was carried out as described by Struhl et al. (1979). Plasmid preparation and electrophoresis were performed according to Maniatis et al. (1982). Gene Screen Membrane (NEN) and 32p Nick Translation Kit (Amersham, Arlington Heights, I1, USA) were used as indicated by the suppliers. Restriction enzymes were purchased from Boehringer/ Mannheim (FRG) and Bethesda Research Laboratories (Gaithersburg, MD, USA).

Results The RAD52 function is required for heterothallic switching In Saccharomyces cerevisiae different pathways of recombination exist with respect to their dependence on the function of the RAD52 gene. In order to test to which pathway heterothallic mating type switching belongs, the cross $23 (Table 1) containing the rad52-i allele was performed. Haploid RAD52 + and rad52 segregants of the cross were used to test for heterothallic switching ability. A prescreen for the mating ability of a and 0~ rad52-strains gave mating frequencies equal to wildtype strains. If one cell of the switching strain changes the mating type from a to ~ it can mate with a tester cell to produce a diploid capable of sporulation. The tester cells cannot switch their mating type because they bear M A T a information at all three loci. N1/6-8C was constructed as a tester strain. Each experiment was carried out with a freshly isolated colony of the switching strain and spontaneous as well as UV-induced switching frequencies were measured. In these experiments non-sporulating diploids were very rare, less than one per 10 s switching cells. The results in Table 2 show that all tested RAD52 strains of the cross could switch their mating type spontaneously and that the frequency could be enhanced using UV irradiation. F r o m all the experiments with rad52 strains, only two sporulation proficient colonies were

HMLoc

m

I

I

m

HMLa MAT¢ HMRa MAT a

a

b

c d

e

f

g

h

i

j

k

Fig. l. Examples of Southern blots obtained with DNA from S. eerevisiae clones, digested with HindIII and BglII and probed with a plasmid containing a 3.8 Kb fragment with the complete mata information. As BglII cuts once within the a, but not within the e information, both are easily discernible, a, tester strain (HMLa MATa HMRa); b, switching strain (HMLe MATa HMRa); c, d, e, f, g: spore colonies obtained by doing tetrad analyses with the one diploid prototrophic colony originating from mating of a MATa rad52-1 cell with a tester cell. All cells containing HMLe are phenotypically nonmaters (Table 2b); h, i, j, k: spore colonies of a whole tetrad from the one diploid colony occurring after irradiation of rad52-1 cells (Table 2c). This diploid obviously stems from a bona fide switching event as no MATe information was present originally in either parent (tester: lane a; switching: lane b), but from the tetrad two spores contain MATe and two spores contain HMLe informations

obtained. Tetrad analyses demonstrated that these exhibited segregation of the rad52 gene. This indicated that the rad52 mutation was present at the time of the switching process. The one spontaneously occurring diploid exhibited a 2:0 pattern of spore viability and all viable spore clones were either of the a or nonmater type. This segregation type will be discussed in the section dealing with the crossing-over event. The other diploid colony occuring after UVirradiation produced four viable spores of the a, ~ and nm type and in Southern blots exhibited a as well as information of MA T (Fig. 1, lanes h-k). This is an indication of a bona fide gene conversion event. An explanation for this very rare event could be that it might be due to some reversion o f the rad52-1 mutation resulting in very low activity of the product. This possibility matches the observation that the rad52 phenotype segregating from the diploid was not as pronounced as in other crosses with respect to their X-ray sensitivity. Several data demonstrate that switching deficiency is really due to the rad52 mutation: first rare M A T a/a diploids could be taken as p r o o f that switching and tester strains complemented markers and secondly, the a/e diploids failing in segregation of rad52 show that reversion of the rad52 mutation resulted in switching ability (data not shown). The results of Table 2 show that 306 spontaneous heterothallic switching events were found in RAD52 strains, whereas in rad52 strains from the same cross no switching event was observed. In the case of UV irradiation all RAD52 strains showed enhanced switching frequency. Only one bona fide conversion event was found originating from a rad52 cell compared to 4573 events from RAD52 strains. Other workers have reported that spontaneous gene conversion in rad52 strains at different loci is lowered 5-fold

499 Table 2. Influence of rad52 on the switching frequency Switching strain $23

1 a-2B

la-SB lb-6A lb-8B 2a-9B 2a-11A 3a-9A

RAD allele

RAD52 RAD52 RAD52 RAD52 RAD52 RAD52 RAD52

Sum la-6D lb-5D lb-9C 2b-2C 3b-3D 3b-7C 3b-8C 4a-5C 4b-3B Sum

tad52 rad52 rad52 rad52 rad52 rad52 tad52 rad52 rad52

Spontaneous switching

UV induced switching

Number of tested cells x 10 v (survivors)

UV dose J/m z

Actual count

7.0 8.7 7.0 5.1 20.5 5.5 5.5

3 74 39 7 100 72 11

59.3

306

Switched cells a

2.3 3.5 7.0 12.0 38.0 33.0 21.0 24.0 24.0

0 0 Ib 0 0 0 0 0 0

174.8

Ib

Number per 106 survivors 0.04 0.85 0.56 0.14 0.49 1.31 0.20

<0.043 <0.029 0.014 <0.008 <0.003 <0.003 <0.005 <0.004 <0.004

20 20 20 20 20 20 20

10 10 5 5 5 5 5 5 5

Number of tested cells x 107 (survivors)

Switched cells" Actual count

3.2 4.4 2.9 5.4 3.2 13.8 7.0

7 576 157 396 1,319 1,900 218

39.9

4,573

5 15 10 7 14 9 5 12 9

0 0 0 0 0 1~ 0 0

86

1°

Number per 106 survivors 0.2 13.1 5.4 7.3 41.2 13.8 3.1

Relative increase: induced/ spontaneous frequency 5x 15 x 10 x 52 x 84 x 11 x 16 x

<0.020 <0.007 <0.010 <0.014 <0.007 <0.011 0.020 <0.008 <0.011

" Switched cells were measured by counting prototrophic colonies capable of sporulation, approximately 10 clones were tested from each batch. All data were obtained by growing the cells for two generations before mixing with tester cells to avoid rescue of irregular switched cells (see Sehiestl and Wintersberger 1983) b Upon tetrad analyses and Southern blotting, this clone obviously does not descend from a bona-fide switch ° See "Results" section (Prakash et al. 1980) to 10-fold (Malone and Esposito 1980; Saeki et al. 1980), but all o f these workers found a detectable n u m b e r o f conversion events. In contrast to this, the present study showed that, c o m p u t e d for equal cell numbers, the gene conversion controlling heterothallic mating-type switching was lowered at least 10a-fold in fact no spontaneous event was detected and induced conversion was decreased 104-fold. So it is concluded that the presence o f the RAD52 gene is an absolute requirement for spontaneous a n d induced heterothallic mating type switching.

The RAD52 function is required for the crossing over event leading to a ho switched phenotype It was reported that reciprocal i n t r a c h r o m o s o m a l recombination between duplicated genetic elements in rad52 mutants occurs at wild type level whereas the frequency o f the conversion event was much reduced (Jackson and F i n k 1981). A reciprocal recombination would delete essential sites on c h r o m o s o m e I I I and result in an inviable cell expressing the opposite mating type (Strathern et al. 1979). This is found spontaneously in a p p r o x i m a t e l y 20% o f heterothallic switching events and can be detected by rescue mating. In the previous experiments, bonafide gene conversion events were selected by growing cells for two generations before measuring the frequency o f mating type switching. In this manner, all types o f rescue mating (Schiestl and Wintersberger 1982) were excluded.

F o r the detection of lethal events leading to a change in mating type, stationary cells rather than logarithmic cells o f the switching strain were employed. RAD52 + and rad52strains as indicated in Table 3 were used and diploid colonies were selected. Ten or all colonies present per batch were isolated and analysed for their mating type. All but one o f the isolates from RAD52 events showed on a/0~ nonmater p h e n o t y p e and were capable o f sporulation. Tetrad analysis was done with 20 tetrads o f the diploids originating from two RAD52 strains. Two events showed a 2: 0 segregation pattern of survival with all surviving colonies being o f the a mating type, thus indicating a crossing-over event. Similar results have been obtained with other RAD strains (Schiestl and Wintersberger 1982; Strathern et al. 1979). All o f the isolates descending from rad52 strains, with the exception of two strains, showed an a/a diploid type (Table 3), as tested by crossing one respresentative o f each group to an c~ haploid as well as an ~/e diploid tester and by assaying spore viability. Surprisingly all the isolates descending from the remaining two o f the rad52 strains showed an a/c~ diploid phenotype. Tetrad analysis was done with 5 o f these diploid strains each. Only a and n m but no c~ haploids were obtained for 10 tetrads dissected from each strain. Southern blots from 6 o f these diploid strains tested showed heterozygosity at HML. Neither the diploids nor any o f the 15 tested haploid spore colonies showed any MATc~ information, all a maters contained H M L a , MATa, H M R a and all nonmaters contained HMLct, MA Ta, HMRa. This result is the same as the one obtained

500 Table 3. Test forthe crossing-over event (~ lethal ring chromosome) Switching strain $23

RAD allele

Spontaneous switching Number of tested cells x 107 (survivors)

la-2B la-8B lb-6A II~8B 2a-9B 2a-11A 3b-5C

RAD52 RAD52 RAD52 RAD52 RAD52 RAD52 RAD52

Sum I a-6D lb-5D lb-9C 2b-2C 3b-3D 3b-7C 3b-8C 4a-5C 4b-3B

18.4 18.2 14.8 15.8 13.4 13.0 16.4 110.0

rad52 rad52 rad52 rad52 rad52 rad52 rad52 rad52 rad52

Sum

Phenotype of the diploids

Tetrad analysis segregation of phenotypes

a/a

a/~

I

II

III

Diploid cells Actual count

Number per 106 survivors

43 240 339 94 83 255 366

0.23 1.32 2.29 0.60 0.62 1.96 2.23

0 0 0 0 0 1 0

10 10 l0 10 10 9 10

8 10 N/T N/T N/T N/T N/T

2 0 N/T N/T N/T N/T N/T

0 0 N/T N/T N/T N/T N/T

0.264 0.111 0.012 <0.006 0.015 0.056 0.038 0.008 0.063

0 0 2 0 2 8 5 1 6

10 10 0 0 0 0 0 0 0

0 0 N/A N/A N/A N/A N/A N/A N/A

0 0 N/A N/A N/A N/A N/A N/A N/A

5 5 N/A N/A N/A N/A N/A N/A N/A

•,420

10.6 14.4 16.4 16.4 13.2 14.2 13.2 13.2 9.6

28 16 2 0 2 8 5 1 6

121.2

68

I: 4 surviving spores of the a, ~ or nm type, indicative of a bonafide switching event II : 2 surviving spores of the a mating type, indicative of an e lethal ring chromosome III: 2 or 4 surviving spores of the a or nm type, indicative of an irregular mating between a cells For rescue mating of events leading to a lethal change of the mating type (~ lethal ringchromosomes) stationary phase cells of the switching and the tester strain were used. Different frequencies for the same strains as used in Table 2 are due to different treatment N/T: not tested; N/A : not applicable

from a third rad52 strain (b in Table 2 and Fig. 1, lanes d, e, g) with the other experimental design. These diploids are not genuine a / e diploids. Their genetics suggest that an event allelic with or linked to marl gave rise to these diploids. W i t h all the rad52 strains of Table 3 no crossing over event changing the mating type was detected with 1.2 x 109 cells. The RAD52 strains with a similar number o f cells tested gave rise to 1420 diploid colonies. If a p p r o x i m a t e l y 10% o f the events are accompanied by e lethality, indicative of a crossing over event, it could be concluded that the frequency o f the crossing over event is decreased at least 102 fold in rad52 mutants. This result shows that gene conv e r s i o n and reciprocal recombination is similarly affected by the rad52 m u t a t i o n in case o f heterothallic mating-type switching. To address the question of whether the p r o d u c t i o n o f the lethal H a w t h o r n e deletion (Hawthorne 1963; Weiffenbach and H a b e r 1981) is also affected, an e system was constructed. Strain J39e bearing HMLe, M A T e and H M R e was used as tester and RAD52 as well as rad52 segregants of cross $9 (Table 1) were used as switching strain. Three tad52 and three RAD52 strains all with mating type e were tested and diploid p r o t o t r o p h i c colonies were isolated. Twenty colonies each were screened for sporulation ability and not a single colony was found to be able to sporulate.

M u t a t i o n or deletion of the e l as well as the e2 function can lead to an a mating type without conferring sporulation ability to the diploid after mating (c.f. Herskowitz and Oshima 1981). In a similar a p p r o a c h by Paquin and A d a m s (1982) no sporulation proficient colony was detected from 128 diploid colonies isolated. A n o t h e r study (McCusker and H a b e r 1981) shows that most o f the events leading to diploid f o r m a t i o n in two heterothallic e populations is a c h r o m o s o m e break which deletes M A T e and all o f the k n o w n markers distal on the right arm on c h r o m o s o m e III. Only 1% o f their events was due to crossing over between M A T and HMR. Because of these difficulties and the low frequency o f the crossing-over events even in RAD52 strains (1% o f the total events) the a p p r o a c h with the e system was n o t pursued.

The RAD52 function is required for ho switehing of both M A T a alleles, M A T a and MATfi Two M A T a alleles exist in terms o f homothallic mating type switching in rad52 strains (Weiffenbach and H a b e r 1981; M a l o n e and H y m a n 1983). In the latter w o r k these alleles were designated MA Ta and MA Tfi. Cells o f the genotype M A T a rad52 HO remain as stable haploids instead of switching and cells o f the M A T e or MA Tgt rad52 HO type die; presumably because o f an unrepaired DSB in the M A T D N A . Both allelic forms o f M A T a switch normally

501 b

cells and thus in the same range as the switching frequency of ho strains. Tetrad analyses revealed a normal segregation of a and c~ cells. So no obvious difference between ho switching and H O : : l a c Z switching was found. p e r 10 6

,0L .b, c 181" a

/ OI-8C

The RAD3 gene product is needed for efficient UV induction of heterothallic mating-type switching

16 14

5

i

%

5C

12

r

lO

f/

1-7B

't l//'""

-

~8

~~

6,

2 -

14D

/2-7A

..a

.o o.

2

1

,

" 0 4 - 7A 4 - 50

100

50

%survival

10

100

50

10

%survival

Fig. 2. Spontaneous and X-ray (a) or UV (b) induced switching frequencies of control (o) and tad3 (o) strains

in RAD52 HO strains. As there was an unexpected difference found for homothallic switching, a test was made to see whether this difference in M A T a alleles has any bearing on the process of heterothallic switching. Four rad52 switching strains were crossed with the HO strain P502 and tetrad analyses were performed. As spore phenotype one would expect, in the case of M A Ta, approximately two times more M A T a rad52-1 HO than MATo~ rad52-1 HO, and in the case of MATfi equal numbers. As most of the crosses resulted in even more M A Tc~ rad52 than M A Ta rad52 spores, the M A T~ allele was obviously present (data not shown). To determine the switching ability of the M A T a allele, strain RM38-7D ho rad52-1 was used as "switching" strain with N1/6-8C as tester, diploid prototrophic colonies were obtained at a frequency of about one per 107 cells. The colonies obtained exhibited the MATa/a type and did not sporulate. Thus both types of M A Ta strains failed to show heterothallic switching in the absence of RAD52.

There is no obvious difference in mating type switching of ho strains and one HO null mutation strain As the ho gene produces a transcript identical in size and regulation to the HO gene (Jensen et al. 1983), the ho alleles might consist of several mutations with very low activity of the product. In this case ho switching would reflect just a low frequency of HO switching. If this were true it would not be surprising if switching were impaired in ho strains as was shown for HO strains (Malone and Esposito 1980; Weiffenbach and Haber 1981). To test this possibility, strain HR161-1C, which contains a laeZ insertion in the HO gene (Stern et al. 1984), was used as a null mutation. The switching frequency of this strain was about 0.6 events

To test the influence of the RAD3 gene on heterothallic switching, the cross $25 (Table 1) was performed and haploid rad3 and RAD3 segregants were used to measure spontaneous as well as UV and X-ray induced frequencies. Data in Fig. 2 and Table 4 show for the rad3 mutants that the spontaneous frequency was about 2-fold that of the wild type. X-rays caused no dramatic effect on the induction kinetics of the rad3 strains compared to wild type strains, but after UV irradiation a significant decrease in inducibility was observed for the rad3 mutant. Further, with rad3 strains an induction profile with different doses 0-0.6 J/m 2 resulted in a decrease below the level of unirradiated cells (Table 4). But, in an experiment using the same doses of UV light for RAD3 and rad3 strains, rather than similar survival levels, rad3 strains show an induction level exceeding that of RAD3 by approximately 2-fold. It is worth noting that in the RAD3 case all low doses of UV (0.2-0.6 J/m 2) showed an enhanced switching frequency even though no effect on viability compared to unirradiated cells was observed. This demonstrates the sensitivity of the system used. Rad3 strains exhibited an enhanced spontaneous and decreased UV induced level of heterothallic mating type switching, thus showing no obvious difference between the system used here and results for gene conversion reported in the literature. It was however reported that UV-induced reciprocal crossing over was unaffected by the rad3 mutation (Kern and Zimmermann 1978). They observed approximately a 100-fold increase for the mutant even at low survival levels (2%). If excision repair affects only gene conversion in heterothallic switching, an increase of switching events due to the production of a deleted chromosome III would be expected at lower survival levels. To check this, a system to test for the occurrence of rescue mating events was used. After irradiation of the haploid rad3 parent with 0.5 Jim z, five diploid clones were picked and tested genetically; all arose after conversion events and no deleted chromosome III indicative of crossing over was present. This suggested that crossing over does not occur at a higher frequency than conversion and that both events show some correlation. Alternatively it might be that only conversion but not crossing over is enhanced by irradiation as is likely for RAD cells (Schiestl and Wintersberger 1983).

Discussion

The results indicate that the RAD52 function is essential for heterothallic mating type switching. In fact, the spontaneous frequency shows at least a 10S-fold and the UV induced frequency a 10g-fold decrease. As one major function of the RAD52 gene is repairing of D N A double strand breaks, the DSB and gap repair model, originally developed for meiotic recombination (Resnick 1976; Szostak et al. 1983), is proposed as a possible mechanism for heterothallic mating type interconversion.

502 Table 4. Spontaneous and irradiation induced switching frequencies of control and rad3 strains Switching strain $25

1~4D 1-5C 2-6B 2-13B 4-5D 4-7A

RAD allele

RAD3 RAD3 rad3 rad3 rad3 rad3

Spontaneous rate of switching per 106 survivors

1.6 1.6 2.7 2.5 1.8 4.8

X-ray Dose Krad

Survival rel. to control

Plated surviving cells x 10v

Induced rate of mating

37 32 24 22 31 33

4.8 4.4 4.0 2.0 4.3 4.1

160 211 266 85 302 278

(%)

5 5 5 5 5 5 J/m 2

1-5C 1-8C

RAD3 RAD3

0.8 0.63

2-8C I-3D

RAD3 rad3

0.78 t.04

I-7B 2-7A 2-14D 4-5D

rad3 rad3 rad3 rad3

2.3 1.3 2.2 1.05

4-7A

rad3

1.6

20 10 20 10 0.05 0.2 0.1 0.2 0.2 0.2 0.4 0.6 0.2 0.4 0.6

Actual Count

Frequency per ]06 survivors

Relative induced/ spontaneous rate

3.3 4.8 6.6 4.2 7.0 6.8

2x 3x 2.5 x 1.7 x 3.8 x t.4 x

20 11.6 17.4 14.6 4.3 4.1 9.5 7.3 8.4 3 1.6 < 0.3 3.3 1.7 < 1.3

25 x 18 x 28 x 19 x 4x 4x 4x 6x 4x 3x 1.5 x < 0.3 x 2x 1x < 0.8 x

UV 40 70 36 77 77 54 51 40 34 41 30 10 34 18 2

1.4 5.0 2.0 1.6 2.3 3.0 7.0 t .4 0.9 1.5 0.9 0.3 1.3 0.6 0.08

285 563 364 234 99 105 159 101 79 44 14 0 41 9 0

In the case of X-ray induction the switching rates were measured after one generation of growth after the treatment, whereas in case of UV induction, determination was carried out after the first buds were visible; therefore spontaneous rates of the same strains are not wholly reproducible

Rad52 has been reported to show only a slight effect on crossing over. The rad52 m u t a t i o n causes a decrease in crossing over between homologous chromosomes of 6-fold (Saeki et al. 1980) to 13-fold (Malone and Esposito 1980). It has no effect on intrachromosomal crossing over, such as sister chromatid exchange (Prakash and TaillonMiller 1981 ; Z a m b and Petes 1981), crossing over between duplicated genes (Jackson and F i n k 1981), or A - A recombination (M. Ciriacy, personal communication). However, in the present study for heterothallic switching a decrease of intrachromosomal crossing over - at least 10 z fold was found. Since conversion is abolished in rad52 strains this may indicate that all crossover events in ho mating type switching are associated with gene conversion. Several pieces of information suggest that mitotic conversion and crossing over are not tightly associated. Rad52substantially decreases the frequency of conversion between duplicated elements whereas crossing over occurs at wild type levels (Jackson and F i n k 1981). Rad3 m u t a n t s which show decreased levels of U V induced conversion show normal levels of U V induced crossing over (Kern and Zimmerm a n n 1978). Studies of the time of occurrence of conversion and crossing over suggested that these were separable events ( R o m a n and Fabre 1983). I n contrast the present study on heterothallic mating type switching shows a concomitant decrease of crossing over with conversion for rad52- mutants (Table 3) and no increase of U V induced crossing-over relative to conversion for rad3- mutants. I n these respects

heterothallic mating type switching is more similar to meiotic than mitotic events. Meiotic conversion as well as crossing over is abolished in rad52 m u t a n t s (Game et al. 1980; Prakash etal. 1980) and rad3 shows no effect on meiotic recombination (Snow 1968; Dicaprio and Hastings 1976). Finally it was suggested for meiotic recombination that crossing over is causally associated with conversion (Hurst et al. 1972). With an expectation that the RAD3 product would be directly involved in the U V induction of ho switching, there remains the problem of residual induction in the rad3 background (Fig. 2). There are at least two different explanations for this phenomenon. First, it is k n o w n that rad3-2 is a leaky allele since deletion of the RAD3 gene leads to inviability of the cell (Naumovski and Friedberg ] 983; Higgins et al. 1983). If RAD3 is essential for induction of conversion after UV-treatment, the residual enzyme activity might be responsible for the residual induction. Alternatively, one could assume that conversion functions normally in a rad3 background as the spontaneous frequency is even higher than in a wild type background and X-ray induction of conversion is not influenced by the m u t a n t (Table 4 and Fig. 2; K e r n and Z i m m e r m a n n 1978). Thus it has been suggested that the presence of unrepaired dimers may hinder the conversion process (Kern and Z i m m e r m a n n 1978). If the second explanation is true, the RAD3 product does not have any direct action in induction of ho switching. Homothallic switching can occur as often as every gen-

503 eration and is initiated by a site-specific DSB introduced by YZendo (Kostriken et al. 1983). Homothallic switching is associated with reciprocal recombination at a frequency of about 1% (Haber et al. 1980). Spontaneous heterothallic switching is associated with reciprocal exchange at about 10-20%, but from mutagen induced heterothallic switching not one reciprocal exchange was found in 126 cases examined (combined data from Table 3; Schiestl and Wintersberger 1983). This possibly provides a further indication that heterothallic switching might be induced by DSBs, as is homothallic switching and that in the induced state (homothallic or m u t a g e n treated heterothallic) crossing over is depressed. If ho switching would just be a low frequency of HO switching due to low activity of the gene product, it would not be surprising that ho switching is RAD52 dependent since this was shown for HO switching. But, as no differences could be f o u n d between mating type switching of an HO insertional m u t a n t and ho switching, it is not likely that ho switching reflects just a low frequency of HO switching. Also the inducibility of ho switching by D N A damaging agents (Schiestl and Wintersberger 1983) matches with the hypotheses that repair of D N A damage is involved.

Acknowledgement. The author expresses his thanks to Dr. U. Wintersberger and Dr. P.J. Hastings for advice and helpful discussion, and to Dr. H.J. Becker for allowing the use of the X-ray source at the Institut fiir Allgemeine Biologic, Vienna, Austria. For providing very useful yeast strains and plasmids Drs. Andrew Murray, Ira Herskowitz, Amar Klar, Bob Malone, Nelson Marmiroli and Fritz Zimmermann, and for initial help with some molecular techniques Dr. H. Ruis and collaborators are acknowledged. Susan Hampson, Anne Galloway as well as Cayatana Hsu are thanked for dedicated technical assistance. This work was sponsored in part by a grant from the "Fonds Osterreichischer Krebsforschungs-Institute" to Dr. U. Wintersberger, by a grant of the Natural Sciences and Engineering Research Council of Canada to P.J. Hastings and by a postdoctoral fellowship of the Alberta Heritage Foundation for Medical Research, No. 3498, to the author. References Cox BS, Game JC (1974) Repair systems in yeast. Mutation Res 26 : 257-264 Dicaprio L, Hastings PJ (1976) Post-meiotic segregation in strains of Saccharomyces cerevisiae unable to excise pyrimidine dimers. Mutation Res 37:137-140 Game JC, Mortimer RK (1974) A genetic study of X-ray sensitive mutants in yeast. Mutation Res 24:281 292 Game JC, Zamb TJ, Braun RJ, Resnick M, Roth RM (1980) The role of radiation genes in meiotic recombination in yeast. Genetics 94:51-68 Haber JE, Rogers DT, McCusker JH (1980) Homothatlic conversion of yeast mating-type genes occur by intrachromosomal recombination. Cell 22:277 289 Hawthorne DC (1963) A deletion in yeast and its bearing on the structure of the mating-type locus. Genetics 48:1727-1729 Herskowitz I, Oshima Y (1981) Control of cell type in Saccharomyees cerevisiae: Mating and mating-type interconversion. In: Strathern JN, Jones EW, Broach JR (eds) The molecular biology of the yeast Saccharomyces, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York pp 181-209 Hicks JB, Strathern JN, Klar AJS (1979) Transposable mating type genes in Saccharomyces cerevisiae. Nature 82:478-483 Higgins DR, Prakash S, Reynolds P, Polakowska R, Weber S, Prakash L (1983) Islation and characterization of the RAD3 gene of Saccharomyees eerevisiae and inviability of rad3 deletion mutants. Proc Natl Acad Sci USA 80 : 5680-5684

Hurst D, Fogel D, Mortimer RK (1972) Conversion-associated recombination in yeast. Proc Natl Acad Sci USA 69:101-105 Jackson JA, Fink GR (1981) Gene conversion between duplicated genetic elements in yeast. Nature 292: 306-311 Jensen R, Sprague Jr GF, Herskowitz I (1983) Regulation of yeast mating-type interconversion: Feedback control of HO gene expression by the mating-type locus. Proc Natl Acad Sci USA 80:3035-3039 Kern R, Zimmermann FK (1978) The influence of defects in excision and error prone repair on spontaneous and induced mitotic recombination and mutation in Saceharomyces cerevisiae. Mol Gen Genet 161:81-88 Ktar AJS, Fogel S, Macleod K (1979) MAR1 - a regulation of the HMa and HMe loci in Saccharomyees cerevisiae. Genetics 93 : 37-50 Klar AJS, Hicks JB, Strathern JN (1982) Directionality of yeast mating-type interconversion. Cell 28:551-561 Kostriken R, Strathern JN, Klar AJS, Hicks JB, Heffron F (1983) A site-specific endonuclease essential for mating-type switching in Saccharomyces cerevisiae. Cell 35 : 167-174 Malone RE, Esposito RE (1980) The RAD52 gene is required for homothallic interconversion of mating types and spontaneous mitotic recombination in yeast. Proc Natl Acad Sci USA 77: 503-507 Malone RE, Hyman D (1983) Interactions between the MAT locus and the tad52-1 mutation in yeast. Current Genetics 7:439-447 Maniatis T, Fritsch EF, Sambrook J (1982) Molecular cloning, a laboratory manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York McCusker JH, Haber JE (1981) Evidence of chromosomal breaks near the mating-type locus of Saecharomyces eerevisiae that accompany MATe x MATe matings. Genetics 99: 383403 Mowat MRA, Jachymczyk WJ, Hasting PJ, yon Borstel RC (1983) Repair of 7-ray induced DNA strand breaks in the radiationsensitive mutant radl8-2 of Saceharomyees cerevisiae. Mol Gen Genet 189:256-262 Naumovski L, Friedberg EC (1983) A DNA repair gene required for the incision of damaged DNA is essential for viability in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 80:4818-4821 Orr-Weaver TL, Szostak JW, Rothstein RJ (1981) Yeast transformation: a model system for the study of recombination. Proc Natl Acad Sci USA 78:6354-6358 Oshima Y, Takano I (1971) Mating types in Saccharomyces: their convertibility and homothallism. Genetics 67:327-335 Paquin C, Adams T (1982) Isolation of sets of a, e, a/a, a/a and e/e isogenic strains in Saccharomyces cerevisiae. Current Genetics 6:21-24 Prakash L (1977) Repair of pyrimidine dimers in radiation-sensitive mutants rad3, tad4, rad6 and rad9 of Saccharomyees cerevisiae. Mutation Res 45:13 20 Prakash L, Taillon-Miller P (1981) Effects of the rad52 gene on sister chromatid recombination in Saecharomyces eerevisiae. Current Genetics 3 : 247-250 Prakash S, Prakash L, Burke W, Montelone B (1980) Effect of the RAD52 gene on recombination in Saccharomyces cerevisiae. Genetics 94:31 50 Resnick MA (1976) The repair of double-strand breaks in DNA: a model involving recombination. J Theoret Biol 59:97-106 Roman H, Fabre F (1983) Gene conversion and associated reciprocal recombination are separable events in vegetative ceils of Saccharomyces cerevisiae. Proc Natl Acad Sci USA 80:6912-6916 Saeki T, Machida I, Sayaka N (1980) Genetic control of diploid recovery after 7-irradiation in the yeast Saccharomyces cerevisiae. Mutation Res 73:251-265 Schiestl R, Wintersberger U (1982) X-ray enhances mating type switching in heterothallic strains of Saecharomyces cerevisiae. Mol Gen Genet 186:512-517 Schiestl R, Wintersberger U (1983) Induction of mating-type inter-

504 conversion in a heterothallic strain of Saccharomyces cerevisiae by DNA-damaging agents. Mol Gen Genet 191 : 59-65 Sherman F, Fink Gr, Hicks JB (1983) Methods in yeast genetics, a laboratory manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York Snow R (1968) Recombination in ultraviolet sensitive strains of Saccharomyces cerevisiae. Mutation Res 6:409418 Stern M, Jensen R, Herskowitz I (1984) Five SWI genes are required for expression of the HO gene in yeast. J Mol Biol 178:853-868 Strathern JN, Newlon CS, Herskowitz I, Hicks JB (1979) Isolation of a circular derivative of yeast chromosome III: Implications for the mechanisms of mating type interconversion. Cell 18:309 -319 Strathern JN, Klar AJS, Hicks JB, Abraham JA, Ivy JM, Nasmyth KA (1982) Homothallic switching of yeast mating type cassettes is initiated by a double-stranded cut in the MAT locus. Cell 31 : 183-192 Sruhl K, Stinchcomb DT, Scherer S, Davis RW (1979) High frequency transformation of yeast: Autonomous replication of hybrid DNA molecules. Proc Natl Acad Sci USA 76:1035-1039

Szostak JW, Orr-Weaver TL, Rothstein RJ, Stahl FW (1983) The double-strand-break repair model for recombination. Cell 33 : 25-35 Weiffenbach B, Haber JE (1981) Homothallic mating-type switching generates lethal chromosome breaks in tad52 strains of Saccharomyces eerevisiae. Mol Cell Biol 1 : 522--534 Wintersberger U, Schiestl R (1982) The yeast mating type system a model for the regulation of gene expression by the position of a certain gene within the genome. In: Jaenicke L (ed) 33. Colloquium Mosbach on Biochemistry of Differentiation and Morphogenesis, Springer, Berlin, Heidelberg, New York, pp 50-53 Zamb TJ, Petes TD (1981) Unequal sister-strand recombination within yeast ribosomal DNA does not require the RAD52 gene product. Current Genetics 3 : 125-132

C o m m u n i c a t e d by C.P. Hollenberg Received February 11 /April 1, 1986