phenicol resistant (CAP-R) were introduced in this strain by two independent and successive ..... frequencies expected on the basis of independent assortment.
GENETIC ANALYSIS OF MULTIPLE DRUG CROSS RESISTANCE IN SACCHAROMYCES CEREVZSZAE: A NUCLEAR-MITOCHONDRIAL GENE INTERACTION JOE D. COHEN*
AND
NORMAN R. EATON
Department of Biology, Brooklyn College, Brooklyn, New York Manuscript received May 1, 1977 Revised copy received June 20, 1978 ABSTRACT
A mutant of the yeast Saccharomyces cereuisiae, cross resistant to several antibiotics, was isolated in our laboratory and subjected to genetic analysis. Tetrad analysis of diploids obtained from crosses between the resistant mutant and a sensitive wild-type strain suggest that the multiple resistance to the five agents, oligomycin (OLI), rhodamine 6G (RHG), tetracycline (TCN), chloramphenicol (CAP) and cycloheximide (CHX) is determined by a single nuclear gene, anti, and requires several cytoplasmic genes for expression of resistance to oligomycin, rhodamine 6G and tetracycline.-Vegetatively growing diploid clones derived from the cross ant1 [RHO+] x f [RHO+] show mitotic segregation of two phenotypic classes for the drugs OLI, RHG TCN. Diploids derived from the two reciprocal crosses, ant1 [RHO+] x [RHO-] and ant1 [RHO-] x f [RHO+], fail to exhibit mitotic segregation. These results are consistent with our hypothesis concerning the involvement of cytoplasmic loci. They suggest, in addition, that these loci are associated with mitochondrial DNA (mtDNA) .-Evidence for this association is provided by the demonstration of genetic linkage between the cytoplasmic loci involved i n the interaction, RHG-1, TCN-1 and OLI-5, and two well-characterized mitochondrial loci, ERY and CAP.-We have mapped the nuclear ant1 locus 3.3 cM from the centromere-linked gene, l e d , on the same side of the centromere of chromosome VZZ as Zeu1.-In the light of these findings, we discuss the claims made by several authors of the episomal nature of mutations similar ta the one described here, as well as of the possible involvement of yeast 28 DNA in such mutations.
+
EVERAL recently published reports have described mutants of Saccharomyces cereuisiae that are cross resistant to a series of drugs. These drugs are diverse in their structure, mode of action and cellular target (RANKand BECH-HANSEN 1975; AVNERand GRIFFITHS1973; 1973; RANK,ROBERTSON and PHILLIPS et al. 1973; GUERINEAU, SLONIMSKI and AVNER1974; HOWELL et al. MITCHELL 1974; LANCASHIRE and GRIFFITHS1975; CARIGNANI,LANCASHIRE and GRIFFITHS 1977). Particular interest in these mutants was generated by the fact that most of them exhibited complex genetic behavior, reminiscent of both nuclear * Present address. Department of Biochemistry, Albert Einstein College af Medicine, 1300 Morris Park Ave., Bronx, New York 10461. Genetics 9 1 : 19-33 January, 1979
20
J. D. COHEN A N D N.
R. EATON
and cytoplasmic inheritance patterns. MITCHELL et al. (1973) have reported that their cross-resistant mutant showed mitotic segregation for oligomycin resistance and sensitivity among the diploid progeny of a cross between the resistant mutant and a sensitive strain. Furthermore, they observed that this cytoplasmic effect was eliminated upon treatment of either parent with ethidium bromide. These results are characteristic of mitochondrial mutations. However, upon tetrad analysis of either sensitive or resistant diploids, the resistance to oligomycin always segregated in a 2:2 Mendelian fashion. Practically identical results were reported by AVNERand GRIFFITHS (1973) €or their “class I” oligomycin resistant mutant. This anomalous behavior prompted different authors to propose various models as the possible genetic basis of these mutations. The models range from single nuclear gene inheritance (RANKand BECH-HANSEN 1975) or nuclear-mitochondrial genetic 1973; RANK,ROBERTSONand PHILLIPS interaction (MITCHELL et al. 1973; HOWELL et al. 1974), to the involvement of SLONIMSKI and AVNER1974; an episomal or plasmid-like system ( GUERINEAU, GRIFFITHS, LANCASHIRE and ZANDERS 1975; CARIGNANI, LANCASHIRE and GRIFFITHS 1977) similar to the bacterial systems (CLOWES 1972). However, a fully catisfactory demonstration of the genetic basis of these cross resistant mutations has not yet been offered. The findings presented in this paper are the result of the genetic analysis of one such cross-resistant mutant, isolated in our laboratory. This mutant bears striking phenotypic and genetic similarities to several of the mutants previously described in the literature. Our data demonstrate that, in our mutant, the pleiotropic resistant phenotype is controlled by a single nuclear gene, antl, and that in the case of at least three of the drugs, three mitochondrial genes, RHG-1, TCN-1 and QLI-5, interact with this nuclear gene. These interactions determine the response of the cells to the drugs rhodamine 6G, tetracyline and oligomycin, respectively. MATERIALS A N D METHODS
Yeast strains: Strain RD35, of genotype a ade2 leu1 pet-ts [ERY-R CAP-R RHO+] *, is the wild-type strain from which the cross resistant mutant, RD35-CR, was derived. Strain ZlEK27-D, of genotype (Y lysl [ERY-S C A P S RHO+], has been used in crosses as the wild-type sensitive parent. Strain S3B-CR is a meiotic segregant from the cross RD35-CR [RHO+] x ZlEK27-D [RHO-]. It exhibits the cross-resistant phenotype of RD35-CR and has the following genotype: a ad-2 leu1 pet-ts [ERY-R CAP-R RHO+]. The pet-ts mutation carried by some of the strains described above rapidly induces the “petite” (RHO-) condition in cells grown at elevated temperature ( 3 7 ” ) . RHO- “petites” from the strains listed above were obtained by treating the cells twice with ethidium bromide (50 pg/ml) in YEPD medium, for 48 hr. Although no biochemical determination of the mtDNA content of these “petites” was performed, it is likely in view of the treatment they were subjected to, that they are of the RHO” type, i.e., lacking detectable mtDNA (GOLDRING et al. 1970). Supporting this assumption is our observation that in crosses to RHO+ cells, they behave as neutral “petites.” Complete media: YEP ( 1 % Difco yeast extract; 2% Difco peptone) supplemented with either 2% Difco dextrose (YEPD) or 3% glycerol 2% ethanol (YEPGE).
+
* Genes between brackets refer t o the mitochondrial or cytoplasmic genotype of the strains.
MULTIPLE DRUG CROSS RESISTANCE I N YEAST
21
Minimum and drop-out media: YNBD (0.66% Difco nitrogen base without amino acids, 2% Difco dextrose). Drop-out media for tetrad analysis were made by adding the appropriate amino acids to YNBD medium. Drug media: YEPD-CHX is YEPD f 0.5 pg/ml of cycloheximide; YEPGE-OLI, YEPGECAP, YEPGE-ERY, YEPGE-RHG and YEPGE-TCN are YEPGE f, respectively, 3 &g/ml oligomycin, 4 mg/ml chloramphenicol, 2 mg/ml erythromycin, 25 pg/ml rhodamine 6G and 4 mg/ml tetracycline. The drugs were added to the media, after cooling t o 55”, as ethanol s o h tions, except cycloheximide and tetracycline hydrochloride, which were added as aqueous solutions. When required, the media were solidified by the addition of 2% Difca agar. Ethidium bromide, oligomycin, chloramphenicol, erythromycin, tetracycline hydrochloride and cycloheximide were purchased from Sigma Chemical Co., St. Louis, MO.Rhodamine 6G was a gift from R. B. NEEDLEMAN. Genetic analysis: Crosses, selection and sporulation of diploids, microdissections and tetrad analysis were performed as described i n Methods in Yeast Genetics, Cold Spring Harbor Laboratory. Cold Spring Harbor, New York. Determination of the drug resistance phenotype of individual diploid clones: Random diploids were allowed to grow for 30 to 40 generations, by successive subculturing in YNBD liquid medium. They were then plated on YNBD plates. Single colonies were picked and used to make master YNBD plates, each containing 50 individual diploid patches. After four to five days of incubation, the masters were replicated onto YEPGE and onto drug containing media. Drug plates were scored after six to 12 days of incubation, depending on the particular drug scored (see RESULTS for further details). Growth conditions: All plates and liquid cultures were routinely incubated at 25”, since strains RD35-CR and S3B-CR carry the temperature sensitive pet-ts mutation described earlier in this section. RESULTS
Origin and phenotype of the cross-resistant mutant: Strain RD35 has been used in our laboratory €or the study of the nuclear temperature sensitive “petite” mutaticn, pot-ts. Mutations to erythromycin resistance (ERY-R) and chloramphenicol resistant (CAP-R) were introduced in this strain by two independent and successive treatments with manganese chloride (MnC1,) , reportedly a speBARANOWSKI and PRAZNO 1973). cific mitochondrial mutagen ( PUTRAMENT, Upon genetic analysis, these two mutations conformed to the standard criteria for bona fide mitochondrial gene mutations, as proposed by BOLLOTINet al. (1971). The doubly resistant strain was then subjected to a third MnCL treatment in the hope of introducing an additional mitochondrial mutation to oligomycin resistance (OLI-R) . One of the several OLI-R mutants obtained following this treatment, strain RD35-CR, showed ambiguous genetic characteristics and, upon further analysis, proved to be cross resistant to cycloheximide and tetracycline, but sensitive to rhodamine EG. The work reported in the following sections was performed using the meiotic segregant, S3B-CR7obtained from a cross of RD35-CR [RH O+] x ZlEK27-D [RH 0-1. Strain S3B-CR has the same crossresistance phenotype as its RD35-CR parent, and carries the two mitochondrial mutations, ERY-R and CAP-R, inherited from that parent. Meiotic segregation of the cross-resistant phenotype: Random diploids from crosses of the drug-resistant mutan: to the sensitive wild-type strain, ZlEK27-D, were sporulated. dissected and subjected to tetrad analysis. Since several mutants of this type, previously described in the literature, gave ambiguous results upon
22
J. D. COHEN A N D N. R. EATON
genetic analysis, we analyzed tetrads from the three following crosses, in order to detect the possible involvement of cytoplasmic or mitochondrial genetic factors in the determination of the phenotype: Cross 1: Cross 2: Cross 3:
S3B-CR S3B-CR S3B-CR
X
[RHO+] [RHO+] [RHO-]
X X
ZlEK27-D ZlEK27-D ZlEK27-D
[RHO+] [RHO- ] [RHO+]
In addition to scoring the tetrads for all the drugs to which S3B-CR is resistant, we also scored them for resistance or sensitivity to the lipophilic dye, rhodamine 6G. The inclusion of this drug in our analysis was prompted by a report that it is a potent inhibitor of oxidative-phosphorylation in rat liver mitochondria that (GREAR1974) and by a personal communication from R. B. NEEDLEMAN, cross-resistant mutants under study in other laboratories showed resistance to rhodamine 6G. The results of these tetrad analyses are presented in Table 1. We should first point out that spores were simultaneously cross resistant or cross sensitive to all the drugs showing a 2:2 segregation for resistance and sensitivity. The data in Table 1 can be most readily explained by the following model: the genetic basis of the cross-resistance phenotype is the result of a single nuclear gene mutation, designated a n d . The phenotypic expression of this mutation depends, at least in the cases of oligomycin and rhodamine resistance, on several cytoplasmic, possibly mitochondrial, loci. The allelic form in which the latter are present determines whether a clone, derived from an ascospore carrying the ant2 mutation, will express resistance or sensitivity to oligomycin and/or rhodamine 6G. The putative cytoplasmic loci, designated OLI-5R or S and RHG-1R or S, can be eliminated concomitantly with the RH O+ factor (mtDNA) by treatment with ethidium bromide. According to this model, the parental strains in crosses 1, 2 and 3 can be assigned nuclear and cytoplasmic genotypes, with respect to the cross-resistance mutation, as shown in Table 2. Given these parental genotypes, TABLE 1 Meiotic segregation of drug resistance Number of tetrads analyzed in each cross
Segregation of resistance: sensitivity for each of the drugs listed
Tetrad classes
-
ERY Cross 1
Cross 2 Cross 3
48
16 13
I I1 I11 IV V VI I1 I
0:4
4:O 4:O 0:4
4:O 4:O 4:O 0:4
CAP 2:2 4:O
2:2 22 2:2 4:O
4:O 2:2
OLI 0:4 2:2 2:2 2:2 0:4 0:4 22 0:4
RHG TCN CHX 2:2 2:2 2:2 0:4 2:2 2:2 2:2 2:2 2:2 2:2 2:2 2:2 2:2 212 2:2 0:4 2:2 2:2 0:4 2:2 2:2 2:2 2:2 2:2
Number of tetrads in cach class
26 10 3 7 1 1 16 13
* Only complete (four-spored) tetrads in which all three nuclear markers, ade2/+, Zed/+ and l y s l / f segregated in a 2:2 fashion are included in this table.
MULTIPLE DRUG CROSS RESISTANCE I N YEAST
23
TABLE 2 Genolypes of parental strains used in crosses 1, 2 and 3, as inferred from the model proposed in the text Nuclear Comments
Parental strain
genotype
Cytoplasmic genotype
S3B-CR RHO+
antl
[OLI-5R RHG-IS RHO+]
+
[OLI-5s RHG-1R RHO+]
antl
[OLI-5"* RHG-I" RHO"]
+
[U,LI-5" RHG-1" RHO"]
ZIEK27-D RHO+
S3B-CR RHO-
ZIEK27-D RHO-
Resistant to TCN and CHX because of anti; Resistant to OLI because of the combined presence of anti and COLI-SRI. Sensitive to RHG because of the presence of the sensitive allele of the interacting cytoplasmic gene, i.e., [RHG-lS] . Sensitive to all drugs controlled by anti because of the presence of the wild-type allele of this gene, although it carries and can transmit to spore progeny the [RHG-lR] cytoplasmic allele. Petite. Cytoplasmic genes and mtDNA physically absent following ethidium bromide treatment. But can transmit antl to meiotic progeny in Mendelian fashion. Petite. Same as above as concerns the cytoplasmic genes. Transmits allele of nuclear gene to meiotic progeny.
+
* Superscript indicates the postulated physical absence of the genetic locus concerned. O
the diploids derived from cross 2 will have the genotype: a d / + [OLI-5R RHGIS], and as a result exhibit 2:2 segregation for oligomycin resistance:sensitivity, but 0:4 segregation for rhodamine 6G resistance:sensitivity. The diploids derived from cross 3, on the other hand, will have the genotype: antZ/+ [OLI-5s RHGlR], and segregate 0:4 for oligomycin, but 2:2 for rhodamine 6G. The diploids derived from cross 1 originate from zygotes having a genetically heterogenous cytoplasm and, therefore, the genotype: ant1J-I- [OLI-5R RHG-1S/OLI-SS RHG1RI. Assuming random assortment and/or recombination of the cytoplasmic alleles, and subsequent purification of the cytoplasmic types among the vegetative diploid progeny of these zygotes (as is the case for mitochondrial genes), four different genotypes are theoretically possible among the random diploids derived from cross 1: antl/+ anti/+ anti/+ anti/+
[OLI-5R [OLI-SS COLI-5R [OLI-SS
RHG-IS] RHG-IR] RHG-lR] RHG-IS]
These diploids will produce respectively the tetrad classes labeled in Table 1 as: Class 11,classes I and V, classes I11 and IV, and class VI.
24
J. D. COHEN A N D N . R. EATON
The tetrod data in Table 1, on which this model is based, do not permit any assumption concerning the possible involvement of cytoplasmic genes in the determination of resistance and sensitivity to the drugs cycloheximide and tetracycline. They do, however, indica:e that this part of the phenotype is, at least, controlled by the nuclear ant1 mutation, since 2:2 segregations were always obtained for these two drugs in all three crosses. The tetrad data further reveal that resistance to chloramphenicol is also imparted to the cells by the ant1 mutation. This is made clear by the results of cross 3 where chloramphenicol resistancesensitivity segregates in a 2:2 fashion, simultaneously with tetracycline and cycloheximide resistance:sensitiviq. This aspect of the phenotype could not be detected in the original mutant. because of the presence of the mitochondrial CAP-R mutation. Finally lhe data from cross 1 suggests that random assortment and/or recombination occurs among and between the mitochondrial loci, ERY and CAP and the putative cytoplasmic genes, OLI-5 and RHG-1. Segregation of phenotypes among vegetative diploid clones: A characteristic of mitochondrial genes, and of the phenotypes they control, is their segregation among vegetatively growing diploid clones. Mitotic segregation can reazonably be expected also from cytoplasmic genes not necessarily associated with mitochondrial DNA. We have analyzed the phenotypes of diploid clones in order to delermine whether or not such a phenomenon occurs in our system. It is clear, however, that in our mutant, mitotic segregation, if it occurs, may be complicated, or even masked, by the presence of the nuclear gene and by the dominance relationship between the alleles of this gene. When individual diploid clones, derived from crosses 1, 2 and 3, are tested for their resistance or sensiGvity to each o€ the drugs involved in this study, and the drug plates scored after four to five days of incubation, mitotic segregation is not observed, except in the case of erythromycin. The diploids from all three crosses are all sensitive to oligomycin, rhodamine 6G, tetracycline and cycloheximide, all semi-resistan: to chloramphenicol, and either sensitive or resistant to erythromycin. Upon further incubation (six to 12 days, depending on the particular drug tested). however, it becomes possible to distinguish two different phenotypes f o r all the drugs, except cycloheximide. Table 3 lists the phenotypes TABLE 3 Segregation of phenotypes among vegetative diploid clones Phenotypic classes observed for each drug tested and number of clones In each phenotypic class ~
Cross1 Cross 2 Cross3
ERY S* : R' 286 105 0 150 150 0
___
CAP SR' : R 289 102 0 150 150 0
OLI SP' : S 134 257 150 0 0 150
-~ -
RHG SP : S 307 84 0 150 150 0
TCNSR : S P 271 120 0 150 150 0
Number clones scored ~
CHX SP 391 150 150
391 150 150
* Phenotypes symbols: (S) = sensistive; (R) = resitant, confluent growth; (SR) = semiresistant, confluent growth; (SP) = sensitive with numerous resistant papillae.
M U L T I P L E DRUG CROSS RESISTANCE I N YEAST
25
observed. Although the phenotypes segregating among the diploid clones are not, except in the case of erythromycin, the usual resistant sensitive observed for mitochondrial mutations, they do illustrate mitotic segregation and can be explained in the following manner, consistent with our model: Erythromycin: The segregation of resistance and sensitivity occurs among the diploids from crosses 1, 2 and 3 as expected from a bona fide mitochondrial mutation (BOLLOTIN et al., 1971). Chloramphenicol: Diploids that have inherited the CAP-R mitochondrial mutation are fully resistant, while those having inherited the CAP-S mitochondrial allele are semi-resistant to the drug, because of the presence of the nuclear gene in the heterozygous, a n t l / f condition. This interpretation is fully consistent with the tetrad data obtained upon dissection of resistant and semi-resistant diploids: the former segregate 4:O while the latter give 2:2 segregation for resistance:sensitivity (see crosses 2 and 3 in Table 1). Oligomycin ‘and rhodamine 6G: The individual diploid patches either acquire an extensive number of resistant papillae or they do not, even after a prolonged period of incubation (12 to 15 days). We have shown that these papillae are homozygous for the ant1 allele, i.e., antl/antl (see later section on the mapping of the nuclear gene), most likely resulting from either mitotic recombination or mitotic gene conversion. Whether or not these antl/antl diploids will express resistance to oligomycin and/or to rhodamine 6G depends upon which cytoplasmic loci were inherited by the individual diploid from which they arose. This interpretation is consistent with the cytoplasmic genotypes that we have ascribed to the parental strains (Table 2). The results obtained for the diploid progeny of crosses 2 and 3, where all the patches are either heavily or not all papillated, argue against the possibility that the segregation of papillated and nonpapillated patches seen among the diploid progeny of cross 1 is merely the result of chance fluctuations in the number of mitotic recombinants or mitotic gene convertants occurring in any given patch. Figure 1 illustrates the lack of ambiguity in scoring the phenotypes. Although some resistant papillae occur on the oligomycin plate among the diploid progeny of cross 3, their number is usually small (one to three on any given patch) and far less than the number of resistant papillae occurring on all the patches of diploids derived from cross 2. They, therefore, do not constitute a major problem in scoring the phenotype of diploids derived from cross 1, and probably represent spontaneous mutations to oligomycin resistance. Tetracycline: Sensitive patches with numerous resistant papillae or semiresistant patches occur among the diploid progeny of cross 1, while crosses 2 and 3 produce diploids that are homogenous for one or the other of these phenotypes. This result indicates that the resistance to tetracycline also depends on a mitotically segregating cytoplasmic factor, designated TCN-1. However, in this case, both parental strains must carry resistant alleles of this factor. This conclusion is called for by the following observations: (a) the sensitive diploids (cross 2) can give rise to resistant papillae, and (b) diploids from both crosses 2 and 3 give only 2:2 segregation for resistancesensitivity to tetracycline (Table 1).
26
J . D. C O H E N A N D N. R. EATON
CROSS I
CROSS 2
CROSS 3
RHG
FIGURE l.--Sc~grcgntion of phrnotypes among diploids derived from crosses 1 , 2 and 3 on YEPGE-RHG platcs (upper three frames) and on YEPGE-OLI plates (lower three frames.)
The factors in the parental strains may be viewed as heteroalleles of the TCN-1R locus. The heteroallele in strain S3B-CR, TCN-1R-1, when present in a heterozygous diploid, ant2/+, produces a sensitive phenotype (with resistant papillae), whereas the heteroallele in strain ZlEK27-D, TCN-1R-2, allows an intermediate level of resistance to be expressed in such diploids. Cycloheximide: Diploids derived from all three crosses express the same phenotype: all are sensitive with nunierous resistant papillae occurring on every patch. Therefore, we do not detect the involvement of a cytoplasmic factor in this case. This, of course, does not constitute definitive evidence a p i n s t nuclear-cytoplasmic interaction, since it is possible to argue that the uniformity of the phenotypes of the diploids derived from all three crosses may be due to ?he identity of the allelcs of a possible cytoplasmic factor, CHX-1, in the two parental strains used in these crosses. Thus, the observation of mitotic segregation among the diploid progeny of cross 1 confirmed our hypothesis of the existence of cytoplasmic genetic factors involved in the determination of resistance or sensitivity to both oligomycin and rhodamine 6G, through their interaction with the nuclear mutation ant2. I n addition, it allowed us to detect a third cytoplasmic factor involved in the response of ant2 cells to tetracycline. The postulated sensitivity of the cytoplasmic factors to ethidium bromide is also confirmed by the failure of the diploids derived from crosses 2 and 3 to exhibit mitotic segregation. This again suggests their association with, or localization on, the mitochondrial genome.
27
MULTIPLE DRUG CROSS RESISTANCE IN YEAST
Recombination of phenotypes among vegetative diploid clones: Strong evidence f o r the localization of the cytoplasmic loci OLI-5, RHG-1 and TCN-1 on the mitochondrial genome would be provided if genetic linkage could be demonstrated between these three loci and the two mitochondrial loci, ERY and CAP. We have attempted to demonstrate such linkage by analyzing the recombination frequencies of the various drug resistance phenotypes among random diploid clones derived from cross 1 (RHO+ x RHO+). The results of this analysis are shown in Tables 4 and 5. For the purpose of clarity, the data are presented as a four-point cross involving ERY, CAP, RHG and TCN (Table 4a) and two three-point crosses involving ERY, CAP and OLI, and RHG, TCN and OLI, respectively (Table 4b and 4c). The observed recombination frequencies for each gene pair, shown in Table 5, were derived from the data in Table 4a, b and c. Table 5 also lists for comparison the recombination frequencies expected on the basis of independent assortment. These latter frequencies were derived TABLE 4 Recombination of phenotypes among vegetatiue diploid clones derived from cross 1 (RHO+ x RHO+) (4a.) Expected phenotypic claqses, and their obserued frequencies, for the jour drugs: ERY, CAP, RHG and T C N , listed in this order Phenotypes
Number of clones scored
RRSS RRSR RRRS RRRR RSSS RSSR RSRS RSRR 78 1 8 4 1 0 68 13 SRSS SRSR SRRS SRRR SSSS SSSR SSRS SSRR 6 4 3 4 0 0 24 245
39 1
(4b.) Expected phenotypic classes and their obserued frequencies, for the three drugs: ERY, CAP and OLI, listed in this order Number of clones scored
Phenotypes
RRR 71
RRS 14
RSR 7
RSS 13
SRR 9
SRS
8
SSR 47
SSS 222
391
(4c.) Expected phenotypic classes and their observed frequencies, for the three drugs: RHG, T C N and OLI, listed in this order Phenotypes
RRS 218
RRR 48
RSR 18
RSS 23
SRR 2
Number of clones scored
SRS 3
SSS 13
SSR 66
391
The cross analyzed in this table is essentially a five-point genetic cross, with the two parents having the following cytoplasmic genotypes: S3B-CR [ERY-R CAP-R RHG-IS TCN-1R1 OLI-5R RHO+] x ZIEK27-D [ERY-S CAP-S RHG-1R TCN-1R-2 OLI-5s RHO+]. For clarity, we have used the following symbols in reference to the phenotypes described i n Table 3: (R) = Erythromycin resistant; (S) = Erythromycin sensitive; (R) = Chloramphenicol resistant; (S) =Chloramphenicol semi-resistant; (R) = Rhodamine 6G sensitive with resistant papillae; (S) = Fthodamine sensitive; (R) = Tetracycline semi-resistant; (S) = Tetracycline sensitive with resistant papillae; (R) = Uligomycin sensitive with resistant papillae; (S) = Oligomycin sensitive. Therefore, the parental phenotypic classes in Table 4a, b and c are, respectively: RRSS and SSRR, RRR and SSS, and RRS and SSR. All other classes represent all the possible recombinant phenotypes.
28
J . D. C O H E N A N D N. R . EATON
TABLE 5 Percent recombination between the cytoplasmic markers ERY, CAP, RHG-I, TCN-1 and OLI-5, taken pairwise
Gene pair
ERY-CAP ERY-RHG-1 CAP-RHG-1 ERY-TCN-I CAP-TCN-1 RHG-1-TCN-I ERY-OLIB CAP-OLI-5 RHG-1-OLI-5 TCN-1-OLI-5
Percent recombination observed
Percent recombination expected on the basis of random assortment
9.5 (37) 10.5 (41) 5.1 (20) 13.0 (51) 11.2 (4'4) 11.8 (46) 21.2 (83) 19.4 (76) 21.0 (82) 22.0 (86)
38.9 (152) 36.7 (143) 36.1 (141) 41.3 (161) 40.9 (160) 39.0 (152) 42.6 (166) 42.3 (165) 40.7 (159) 43.9 (172)
The data appearing in the left column represent the product of the frequencies with which each allele occurs among the diploid progeny. Numbers in parentheses represent the observed or expected number of recombinants for a particular gene pair.
from the frequencies with which each allele occurs among the diploid progeny of cross 1. Although the observed recombination frequencies cannot be extrapolated to actual map distances, they can, however, be used as a qualitative indication of genetic linkage by statistically comparing them to the recombination frequencies expected on the basis of independent assortment. The data in Table 5, therefore, clearly demonstrate that the cytoplasmic loci RHG-1 and TCN-1 are genetically linked to the mitochondrial loci ERY and CAP, and thus are themselves mitochondrial loci. The recombination frequencies between OLI-5 and the four other markers are, in all cases, approximately 20%. This represents the upper limit of recombination observed for mitochondrial genes, and is generally interpreted as indicative of unlinked genetic markers carried by the same mtDNA molecule (DUJON,SLONIMSKY and WEILL1974). Therefore, although a doubt may persist as to the localization of the OLI-5 locus, it is likely that it also is located on the mitochondrial genome. Mapping of the antl locus: The segregation of the antl and alleles during meiosis can be followed by scoring the ascospore clones for resistance or sensi-
+
TABLE 6 Mapping of the antl locus Segregation of leuf / f and anti in complete and true tetrads
+
2 leul anti : 2+ 2 leu1 :2 antl 1 leu1 anti : 1 leul
+ +
/+
+ : 1 + antl : 1 ++
Type of tetrads
Number of tetrads of each type
PD NPD
223 0 16
TT
M U L T I P L E DRUG CROSS RESISTANCE IN YEAST
29
tivity to tetracycline and cycloheximide, since the phenotypes for these two drugs always segregate in a 2:2 fashion in the tetrads. It is therefore possible to map the ant1 locus, relative to the other nuclear markers present in the crosses, by determining %herelative frequencies of the various types of tetrads obtained for the two markers considered. The results of such an analysis, for the two markers antl and leul, are reported in Table 6. The obvious inequality PD >> NPD indicates linkage between the two genes. The distance between them can be derived from the equation:
X=
100 X ( T T f GNPD) 2 x (TTfPD-tNPD)
where X is the distance, in centimorgans, between the two markers (PERKINS 1949). In this instance X = 3.3 cM. In addition, we have attempted to determine the position of the antl locus relative to the Zeul locus and to their common centromere. Thirty resistant papillae were collected from diploid patches from the mitotic segregation experiment described previously and purified on the appropriate drug media. Among these, 18 proved to be leucine prototrophs, while 12were auxotrophic for leucine. Five leucine prototrophic papillae and seven auxotrophic ones were sporulated and subjected to tetrad analysis. All 12 diploids segregated 4:O for resistance: sensitivity to cycloheximide and tetracycline. The five leucine prototrophs segregated 2: 2 for leucine, while the seven leucine auxotrophs gave 0: 4 segregation for leucine (+: Zeul).These data confirm our earlier assumption that the resistant papillae were homozygous antljantl . In addition, the proportion of leucine auxotrophs, homozygous Zeul/Zeul, among them, suggest that the antl locus is located on the same side of the centromere as Zeul. The uncertainty as to the nature of the genetic event that gave rise to the antl/antl isolates (mitotic recombination or mitotic gene conversion), does not allow a more specific ordering of the or antl-leulgenes. Therefore two orders remain possible: leul-antl-cVII CVII. DISCUSSION
The data reported in this paper clearly demonstrate that the genetic basis of the cross-resistance mutation in strain S3B-CR depends on an interaction between a single nuclear gene and three cytoplasmic loci. The nuclear gene, antl, maps near the centromere-linked marker, Zeul, on chromosome VIZ. Two of the cytoplasmic loci, RHG-1 and TCN-1, are closely linked to the ERY and CAP markers on the mtDNA. The third cytoplasmic gene, OLI-5, is genetically unlinked in ERY, CAP, RHG-1 and TCN-1, but it is most probably also located on the mitochondrial genome. This type of nuclear-mitochondria1 genetic interaction has been suggested by MITCHELLet al. (1973) in the case of a mutant very similar to ows in its phenotypic and genetic characteristics. More recently, ROTMAN (1975) demonstrated a dual nuclear-cytoplasmic control in the case of a mutant resistant to primaquin, a nucleic acid intercalating dye. The principal contribution of our work is the additional demonstration that, in our mutant,
30
J . D. COHEN A N D N. R . EATON
the cytoplasmic loci involved are located on the mitochondrial genome. This finding must be viewed in the context of the striking similarities between mutants of this type reported in the literature, including ours, and the various models proposed in order to explain their genetic properties. It is likely that several, if not all, of these mutants share some common genetic features. Supporting this assumption is our determination that two of these mutants, the ones reported by RANK and BECH-HANSEN (1973) and GUERINEAU, SLONIMSKI and AVNER (1974), which were kindly made available to us by J. MARMUR, are allelic (or heteroallelic) to ours, so far as the nuclear gene is concerned (COHEN 1977). Thus, our results raise some questions concerning the report by GUERINEAU, and AVNER(1974) that the phenotype of their mutant is controlled, SLONIMSKI in part, by episomal genetic elements. In a more recent paper (GUERINEAU, and SLONIMSKI (1976) claim that three nuclear genes and two GRANDCHAMPS episomes (U and T) are involved in the control of the phenotype of their mutant. Moreover, they claim that episome U, controlling the resistance to oligomycin, is identical to a nonmitochondrial, presumably cytoplasmic, DNA species found in yeast, known as the 2 p circular DNA (CLARK-WALKER 1973). Most of the reported data supporting their claim is of a circumstantial and biochemical nature, which can be summarized as follows: (a) Recent studies on the structure of the 2 p DNA have shown it to consist of covalently closed circles containing et al. 1976) similar to the insertion two inverted repeat sequences (HOLLEMBERG sequences (IS) found in bacterial plasmids. (b) A certain correlation (but not an absolute one) exists between the loss of resistance to oligomycin and the loss of the 2 p DNA (GUERINEAU, SLONIMSKI and AVNER1974; GUERINEAU, GRANDCHAMPS and SLONIMSKI 1976). It is clear, however, that in the final analysis, genetic data not yet published by these authors will have to be provided in order to demonstrate the episomal nature of their mutation. Similarly, in a recent LANCASHIRE and GRIFFITHS(1977) have presented data publication, CARIGNANI, suggesting the involvement of both nuclear and cytoplasmic genetic factors in the determination of the phenotype of a mutant cross resistant to rhodamine 6G, venturicidin, trietliyltin, bongkrekic acid and cycloheximide. These authors have concluded that the cytoplasmic factors involved in their mutant are not located on the mitochondrial genome. This conclusion is based on their observation that the genetic determinant for rhodamine resistance is apparently retained in at least some RHO- and RHO” petites derived from the rhodamine-resistant mutant. They propose the 2 p DNA as a possible candidate for the location of this genetic determinant. This conclusion, however, is not necessarily justified. Indeed, the demonstrated involvement of a nuclear gene, as well as the possibility that the RHO+ wild-type strain (to which the petites derived from the resistant mutant must be crossed in order to determine the retention o r loss of cytoplasmic factors) may be a “carrier” of a cytoplasmic allele for rhodamine resistance. does not allow the elimination of the mitochoiidrial genome as a possible location for this cytoplasmic factor. Our own findings do not support these episomal theories and/or the involvement of 2 p DNA in the determination of the phenotype of multiple cross-resistant
MULTIPLE DRUG CROSS RESISTANCE IN YEAST
31
mutants. They do, however, provide a plausible alternative genetic model and suggest that the claims concerning possible genetic markers located on the 2~ DNA should be re-evaluated. Another interesting aspect of our mutant, as well as of mutants of this type described by others, is their bearing on the question of the interaction between the nuclear and mitochondrial genetic systems in the biogenesis of mitochondria, as well as in the control of other cellular characteristics. It is, by now, a welldocumented fact that the synthesis of several mitochondrial enzyme complexes requires the cooperative interaction of the nuclear and mitochondrial genomes, as well as the contribution of the cytoplasmic and organelle protein synthesis systems. The cytochrome oxidase complex (MASON and SCHATZ 1972) and the 1971) have been shown to consist of protein subATPase complex (TZAGLOFF units synthesized on both cytoplasmic and mitochondrial ribosomes. Numerous nuclear mutations (BECKet al 1971; EBNER, MASONand SCHWATZ 1973; TZAGOLOFF, AKAIand NEEDLEMAN 1975a) and several mitochondrial mutations (TZAGOLOFF, AKAI and NEEDLEMAN 1975b), affecting one or more of these enzymes, have been described. Recently, there has also been several circumstantial lines of evidence suggesting that the mitochondrial genome may be implicated in the determination of cellular characteristics other than those of the organelle itself (PUGLISIand ALGERI1974; EVANS and WILKIE1975,1976). We hope that further genetic and physiological study of mutations such as the one reported on in this paper will provide new insight into the problems of the integration and cooperation of the nuclear and mitochondrial genomes in the specification of mitochondrial and other cellular functions and structures. LITERATURE CITED
AVNER,P. R. and D. E. GRIFFITHS,1973 Studies of energy-linked reactions. Genetic analysis of oligomycin-resistant mutants of Saccharomyces cereuisiae. European J. Biochem. 32 : 312-321. BOMTIN,M., D. COEN,J. DEUTSCH, B. DUJON,P. NETTER,E. PETROCHILQ and P. P. SLONIMSKI, 1971 La recombinaison des mitochondries chez Saccharomyces cereuisiae. Bull. Inst. Pasteur (Paris) 69: 215-239. BECK,J. C., J. H. PARKER, W. X. BALCAVAGE and J. R. MATTOON, 1971 Mendelian genes affecting development and function of yeast mitochondria. pp. 194-204. In: Autonomy and biogenesis of mitochondria and chloroplasts. Edited by N. K. BOARDMAN, A. W. LINNANEand R. M. SMILLIE.North Holland, Amsterdam. CARIc"Nr, G., w. E. LANCASHIRE and D. E. GRIFFITHS,1977 Extrachromosomal inheritance of rhodamine 6G resistance in Saccharomyces cereuisiae. Molec. Gen. Genet. 151 : 49-56. CLARK-WALKER, C. D., 1973 Size distribution of circular DNA from petite mutant yeast lacking p DNA. European J. Biochem. 32: 263-267. CLOWES, R. C., 1972 Molecular structure of bacterial plasmids. Bacteriol. Rev. 36: 361-405. COHEN,J. D., 1977 Genetic analysis of multiple drug cross resistance in Saccharomyces cereuisiae: a nuclear-mitochondrial gene interaction. Ph.D. Thesis, City University of New York. DUJON, B., P. P. SLONIMSKIand L. WEILL,1974 Mitochondrial genetics IX: A model for recombination and segregation of mitochondrial genomes in Saccharomyces cereuisiae. XI1 International Congress of Genetics. Genetics 78: 415-437.
32
J . D. C O H E N A N D N . R. EATON
EBNER,E., T. MASONand G. SCHATZ,1973 Mitochondrial assembly in respiration-deficient mutants of Saccharomyces cereuisiae. 11. Effect of nuclear and extrachromosomal mutations on the formation of cytochrome c oxydase. J. Biol. Chem. 248: 5369-5378. EVANS,I. H. and D. WILKIE,1975 Cellular effects of mitochondrial inhibition by acriflavine in Saccharomyces cereuisiae. pp. 179-1 82. In: Molecular biology of nucleocytoplasmic relationElsevier, Amsterdam. __ , 1976 Mitochondrial ships. Edited by A. PUISSEUX-DAO. factors in the utilization of sugars in Saccharomyces cereuisize. Genet. Res. Camb. 27: 89-93. GOLDRING, E. S., I. L. GROSSMAN, D. KRUPNICK, D. R. CRYERand J. MARMUR, 1970 The petite mutation in yeast. Loss of mitochondrial deoxyribonucleic acid during induction of petites with ethidium bromide. 5. Mol. Biol. 52: 323-335. GREAR,A. R. L., 1974 Rhodamine 6G. A potent inhibitor of mitochondrial oxidative phosphorylation. J. Biol. Chem. 249: 3628-3637. GRIFFITHS,D. E., W. E. LANCASHIRE and E. D. ZANDERS, 1975 Evidence of and extrachromosomal element involved in mitochondrial function: A mitochondrial episome? FEBS Letters 53: 126-130. GUERINEAU, M., P. P. SLONIMSKI and P. R. AVNER,1974 Yeast episome: Oligomycin resistance associated with a small covalently closed nonmitochondrial circular DNA. Biochem. Biophys. Res. Commun. 61 : 412-419. GUERINEAU, M., C. GRANDCHAMPS and P. P. SLONIMSKI, 1976 Structure and genetics of the 2fi circular DNA in yeast. pp. 557-564. In: Genetics and biogenesis of chloroplasts and mitochondria Edited by T. H. BUCHER,W. NEUPERT,W. SEBALD and S. WERNER. North Holland, Amsterdam. HOLLEMBERG, C. P., A. A. DEGELMANN, D. KUSTERMANN-KUHN and H. D. ROYER,1976 Characterization of 28 DNA of Saccharomyces cereuisiae by restriction fragments analysis and integration in an E. coli plasmid. Proc. Natl. Acad. Sci. U.S. 73: 2072-2076. HOWELL, N., P. L. MOLLOY, A. W. LINNANEand H. B. LUKINS,1974 Biogenesis of mitochondria 34. The synergistic interaction of nuclear and mitochondrial mutations to produce resistance to high levels of mikamycin in Saccharomyces cereuisiae. Mol. Gen. Genet. 128: 43-54.
w.
LANcAsHmE, E. and D. E. GRIFFITHS,1975 Studies of energy-linked reactions: Genetic analysis of venturicidin-resistant mutants. European J. Biochem. 51 : 403-413. MASON,T. and G. SCHATZ, 1973 Cytochrome-c oxidase from Baker’s yeast. 11. Site of translation of the protein components. J. Biol. Chem. 248: 1355-1361. MITCHELL,C. H., C. L. BUNN,H. G. LUKINSand A. LINNANE,1973 Biogenesis of mitochondria 23. The biochemical and genetic characteristics of two different oligomycin resistant mutants of Saccharomyces cereuisiae under the influence of cytoplasmic genetic modification. Bioenergetics 4: 161-177.
PERKINS, D. D., 1949 Biochemical mutants in the smut fungus Ustilago maydis. Genetics 34: 607-626. PUGLISI, P. and A. A. ALGmr, 1974 Interaction of mitochondrial protein synthesis on the regulation of gene activity in Saccharomyces cereuisiae. pp. 169-177. In: The biogenesis of mitochondria. Edited by A. M. KROONand C. SACCONE. Academic Press, New York.
PUTRAMENT, A., H. BARANOWSKA and W. PRAZNO, 1973 Induction by manganese of mitochondrial antibiotics resistance mutations in yeast. Mol. Gen. Genet. 126: 357-366. RANK,G. H. and N. T. BECH-HANSEN, 1973 Single nuclear gene inherited cross resistance and collateral sensitivity to 17 inhibitors of mitochondrial function in Saccharomyces cereuisiae. Mol. Gen. Genet. 126: 93-102.
M U L T I P L E DRUG CROSS RESISTANCE IN YEAST
33
RANK, G. H., A. J. ROBERTSON and K. L. PHILLIPS, 1975 Modification and inheritance of pleiotropic cross resistance and collateral activity in Saccharomyces cereuisiae. Genetics 80 : 483-493. ROTMAN, A., 1975 Genetics of a primaquin resistant yeast. J. Gen. Microbiol. 89: 1-10.
TZAGOLOFF, A., 1971 Assembly of the mitochondrial membrane system. IV. The role of mitochondrial and cytoplasmic protein synthesis in the biogenesis of the rutamycin-sensitive adenosine-triphosphatase. J. Biol. Chem. 246 : 3050-3056. TZAGOLOFF, A., A. AKAI and R. B. NEEDLEMAN, 1975a Characterization of nuclear mutants of Saccharomyces cereuisiae with defects in mitochondrial ATPase and respiratory enzymes. J. Biol. Chem. 250: 8228-8235.
TZAGOLOFF, A., A. AHAI and R. B. NEEDLEMAN, 1975b Cytoplasmic mutants of Saccharomyces cerevisiae with lesions in enzymes of the respiratory chain and in the mitochondrial ATPase. J. Biol. Chem. 250: 8236-8242. Corresponding editor: F. SHERMAN
