Edward J. Louis and James E. Haber'. Rosenstiel Basic Medical Science Research Center and Department of Biology, Brandeis University, Waltham, ...
Copyright 0 1989 by the Genetics Societyof America
Nonrecombinant Meiosis I Nondisjunction in Saccharomyces cerevisiae Induced by tRNA Ochre Suppressors Edward J. Louis and JamesE. Haber’ Rosenstiel Basic Medical Science Research Center and Department of Biology, Brandeis University, Waltham, Massachusetts 0225491 10 Manuscript received March 2 1, 1989 Accepted for publication June 9, 1989 ABSTRACT The presence of the tRNA ochre suppressors SUPll and SUP5 is found to induce meiosis I nondisjunction in the yeast Saccharomyces cerevisiae. The induction increases with increasing dosage of the suppressor and decreases in the presence of an antisuppressor. The effect is independent of the chromosomal location of SUPl 1. Each of five different chromosomes monitored exhibited nondisjunction at frequencies of 0.1 %- 1.1% of random spores, which is a 16-160-fold increase over wildtype levels. Increased nondisjunction is reflected by a marked increase in tetrads with two and zero viable spores. In the case of chromosome ZZZ, for which a 50-cM map interval was monitored, the resulting disomes are all in the parental nonrecombinant configuration. Recombination along chromosome ZZZ appears normal both in meioses that have no nondisjunction and in meioses for which there was nondisjunction of another chromosome. We propose that a proportion of one or more proteins involved in chromosome pairing, recombination or segregation are aberrant due to translational read-through of the normal ochre stop codon. Hygromycin B, an antibiotic that can suppress nonsense mutations via translational read-through, also induces nonrecombinant meiosis I nondisjunction. Increases in mistranslation, therefore, increase the production of aneuploids during meiosis. There was no observable effect of SUPll on mitotic chromosome nondisjunction; however some disomes caused SUPl 1 ade2-ochre strains to appear white or red, instead of pink.
HE proper segregation of chromosomes during meiosis depends on several factors. These in-
T
clude the mechanical interactions of the meiotic spinof homologouschromosomes and may include processes involved in recombination. Many mutants that are defective in meiotic recombination in both Drosophila melanogaster (CARPENTER and SANDLER 1974) and the yeast Saccharomyces cerevisiae (ESPOSITOa n d KLAPHOLZ1981) are also defective in segregation. This coincident nondisjunction and lack of recombination has lead to the hypothesis that recombinationis necessary for proper segregation(BAKER et al. 1976).Nonrecombinant homologuesinDrosophilafemales are thoughtto segregate properly through a “distributivepairing” segregation pathway (GRELL 1962, 1964). This pathS. cerevisiae (MANNa n d way also appears to operate in DAVIS1986; DAWSON,MURRAY andSZOSTAK1986). A deeper understanding of the processes involved in meiotic segregation may come from a combination of mutational and structural analyses [see GIROUX (1 988) for review]. Disruption of segregation processes with external factors such as drugs may yield also information. Many
dle with thecentromeres
’
T o whom correspondence should be addressed, T h e publication costs of this article were partly defrayedby the payment of page charges. This article must therefore be hereby “advertisement” marked in accordance with 18 U.S.C. $1734 solely to indicate this fact. Genetics 123: 81-95 (September, 1989)
chemicals are known to inducenondisjunction[see BOND (1987) for review]. Some of these disrupt the mechanical interactions necessary for proper segregation. For example, benomyl inhibits tubulin polymerization. Other drugs may resultin the production of aberrant proteins through modification, which then disruptsegregation.Defectiveproteinsthatdisrupt segregation might alsobe produced by mistranslation. O n e way tocausemistranslation is through tRNA suppressors which aberrantly recognize stop codons as amino acid codons. ROTHSTEIN,ESPOSITOand ESPOSITO(1977) found (SUPIthat class I tyrosine inserting ochre suppressors SUP9) in theyeast S. cerevisiae reduced, and sometimes eliminated, sporulation when homozygous. Their effect on segregation was not tested. T h e class I1 tyrosine inserting ochre suppressorSUPl I did not affect sporulationwhenhomozygous.Ingeneral, class I1 ochresuppressors are less effectiveatsuppressing someochre alleles than class I suppressors(HAWTHORNE a n d MORTIMER1968). SUPl I has been used extensively as a dosage dependent marker to screen copy number of chromosomes and plasmids in mitotic segregation studies (HIETERet al. 1985). Zero, one and two copies of SUPl I are distinguishable as red, pink and whitein a n ade2-ochre background. Thelack of an apparent sporulation defect should make SUPl 1
E. J. Louis and J. E. Haber
82
useful in meiotic studies of segregation as well. In an attempt to follow recombination and segregation of Y' sequences, a subtelomeric repeated sequence family of S. cerevisiae (CHANand TYE1983), individual Y's were marked by insertion of a SUP1 I in which the dosage series is distinguishable in haploids (E. J. LOUISand J. E. HABER,unpublished data). From diploids that were homozygous for a Y'::SUPf1 at chromosome VZZ, XV or XZZZ, a screen for white spores was expected to yield haploids with a second copy of the Y ' : : S U P I l resulting from recombination or gene conversion. However,unexpectedly, the white spore colonies recovered still had only one copy of SUP1 1. These white colonies (15/1684 random spores) turned out to be MATaIMATa chromosome ZZZ disomes (MAT is 20 cM from the centromere on chromosome IZZ). A controlstrain with no S U P l l yielded 0/720 chromosome ZZZ disomic random spores ( P = 0.0047, exact test of independence). The reason why strains that hadbecome disomic for chromosome I l l were white is not known, however,as shown below, several different disomes exhibited altered pigment expression. In ordertounderstandthe effect of SUP11 on meiotic segregation, the following study was performed. MATERIALS AND METHODS Media and growth conditions: All strains were grown at 30" except for sporulation plates which were incubated at 25". Rich (YEPD), synthetic complete (SC), SC without specific amino acids added, sporulation, and canavanine containing and cycloheximide containing media were prepared as described in SHERMAN, FINKand HICKS(1986). aAmino adipatecontaining medium was prepared as described in CHATTOO et al. ( I 979). 5-Fluoro-orotic acid containing medium was prepared as described in BOEKE, LACROUTE and FINK(1984). YEPD, SC-adenine and sporulation media were prepared with 50 rg/ml, 100 rg/ml and 200 rg/ml of hygromycin B when appropriate. Plasmids and construction: The plasmids YRpl4 and YRpl5, containing URA3 and SUPl 1, were obtained from P. HIETER (HIETER et al. 1985). They differ in the orientation of cloned EcoRI fragmentcontainingSUPl l. The SUP11 in YRpl4 appears to have a higher level of suppression than that of YRpl5. The GAL::HO plasmid was ob(JENSEN and HERSKOWITZ, tained from I. HERSKOWITZ 1984) andconsists of URA3 and theHO endonuclease gene attached to a galactose inducible promoter. The plasmid pRHB 13 (obtained from R. H. BORTS) consists of URA3 and his4 sequence in which a four base pair insertion was created at the EcoRI site of h i d . pEL2 consists of URA? and the SUPl 1 from YRpl5 inserted into subtelomeric Y' sequence. PEL9 consists of URA3 and the LEU2 containing PstI fragment inwhich the LEU2 sequence from -646 to +762 is deleted. This deletes part of a resident Ty-6 element and a tRNA-leucine gene (SUP53) 5' to LEU2. The 3' end of the deletion is within the LEU2 coding sequence and does not affect the adjacent gene. There is no phenotype other than leucine auxotrophy associated with this deletion. PEL1 1 consists of TRPl sequence disrupted at the EcoRV site by the XhoI to Sal1 (-646 to +1584) LEU2 containing frag-
ment. All enzymes were obtained fromNew England Biolabs and used according to manufacturer's instructions. Strain construction:The strains used in this study (Table 1) were constructed as an isogenic, or in some cases, congenic series. Several steps, described below, were required in orderto constructastrainbackground with several chromosomes marked, inwhich the effects of different suppressor locations and dosages could be observed isogenically and congenically. The strain x23 10-12D was obtained from the Yeast Genetic Stock Center (Berkeley, CA). The haploid MATa strain YP1 and the diploid strain YP3 were obtained from P. HIETER(HIETERet al., 1985). In order to obtain a congenic MATa haploid, YP3 was sporulated and dissected and an appropriatesegregant was used. The canl allele, a recessive canavanine resistant mutation, was isolated spontaneously in YPI. Thecyh2 allele, a recessive cycloheximide resistant mutation, was obtained by mutagenesis of YP1with ultraviolet light. The canl version ofYP1 was crossed to the MATa segregant from YP3. A MATa, canl segregant from this diploid was obtained and crossed to the cyh2 derivative of YP1. From this diploid, aMATa, canl and cyh2 segregant was isolated. SUPll was inserted into Y' sequences in YPI by a one step gene transplacement(ROTHSTEIN 1983) using an EcoRI-PvuI fragmentfrom pEL2 which contains SUPll, URA3 and some pBR322 sequence flanked by Y' sequences. The chromosomal locations of the inserts were determined by Southern analysis (SOUTHERN 1975) of CHEF chromosome separating gels (CHU,VOLLRATH and DAVIS,1986) probed with pBR322. Two transformants, one with Y'::SUPll and URA3 at one end of chromosome VZZ or XV (not separable in this strain) and one with Y'::SUPll and URA3 at one endof chromosome XIII, were crossed to the MATa, canl, cyh2 strain. From each of these diploids, a MATa, canl, cyh2, Y'::SUP11 and URA3 segregant was isolated. The diploids EJL77 and EJL97 were created by backcrossing these segregants to the original transformants in YP1. The strain EJL146 was constructed by crossing YPl to the MATa, canl, cyh2 segregant and representsa nonsuppressor bearingstrain congenic to EJL77 and EJL97. An isogenic series of strains was constructed as follows. A MATa version of YP1was obtained by homothallic switching using the GAL::HO plasmid. Subclones that lost the plasmid were then backcrossed to YP1. The LEU2 sequence from -646 to +762 was deleted from the genomic location in the MATa version of YPI, by a two stepgenereplacement (ORR-WEAVER, SZOSTAK and ROTHSTEIN1983) using pEL9. The plasmid was integrated at the LEU2 locus and then 5fluoro-orotic acid resistant, Ura-, subclones were selected (BOEKE,LACROUTE and FINK 1984). These were screened for leucine auxotrophy and then checked by Southern analysis (SOUTHERN 1975)forthe desireddeletion.A his4-R mutation was introduced into YP1 by a two step gene replacement using pRHB13. A trpl::LEU2 disruption was created in YPl (with the LEU2 deletion) by gene transpiacement using a PstI to XbaI fragment from PEL 11. A spontaneous reversion oflys2-801was obtained in YPI. Subsequently, several new lys2 alleles were isolated using a-amino adipate selection. lys2-r2 was chosen as it recombines with lys2-801 leading to lysine prototrophy ata frequency of 5 X 10-4 meiotically. This frequency is over 50 fold greater than mitotic levels. A backcross to YPI (or the isogenic MATa version) was performed after every transformation and mutant isolation. The parents of EJL374, EJL360-ID and EJL363-12B (Table I), were constructed by serially crossing each of the YP1 derived mutants and transformants (YPI-MAT% y p l canl, YP1-cyh2, YPl-his4-R, YP1-MATa-leuZA,YPl-LYS2,
a3
Suppressor-Induced Nondisjunction TABLE 1 Strains used in this study Genotype"
Strain
YPl
MATa ura3-52 ade2-101" lys2-801
YP3b
MATa ura3-52 ade2-101"lys2-801 MATa ura3-52 ade2-101"lys2-801
EJL77'
MATa ura3-52 CANl CYHZ ade2-101lys2-801 Y ' : : S U P l l and U R A 3 (at VII or XV) MATa ura3-52 canlR cyhP ade2-101" lys2-801 Y ' : : S U P l l and URA3
EJL97b
MATa ura3-52 CANl CYHZ ade2-101" lys2-801 Y ' : : S U P l l and URA3 (at X l l l ) MATa ura3-52 canlR cyhP ade.2-101" lys2-801 Y ' : : S U P l l and URA3
EJL 1 46b
MATa ura3-52 CANl cyH2 ade2-101" lys2-801 MATa ura3-52 canlR cyhZR ade2-101" lys2-801
X2310-12D
MATa SUP5 ura3-1 ade2-lo lysl-1" canl-100" trp5-48" his5-2"
EJL360-1 D'
hisl-RI leu2-A MATa ura3-52 canlR lys2-801 CYH2 TRPl ade2-101"
EJL363-12B'
HIS4 leu2-A MATa ura3-52 CANllys2-r2 cyhZR trp1::LEUZ ade2-101"
EJL374'
his4-Rl l e d - A MATaura3-52canlR HIS4 leu2-A MATa ura3-52 CANl
EJL384'
his4-RIleu2-A MATaura3-52::YRp15(SUPll HIS4 leu2-A MATa ura3-52::YRp15 (SUPll
EJL389'
TRPl ade2-101' his4-RI led-AMATaura3-52::YRp14 (SUP11 and URA3) canlR lys2-801CYH2 HIS4 leu2-A MATa ura3-52::YRp14 (SUPll and LIRA?) CANl lys2-rZ cyhP trp1::LEUZade2-101"
EJL392'
his4-RI led-AMATaura3-52canlR HIS4 leu.2-A MATa ura3-52 CANl
EJL395'
his4-RIleu2-A MATaura3-52::YRp15(SUPll HIS4 leu.!?-Aura3-52 MATa CANl
EJL398'
his4-RI led-AMATaura3-52canlR HIS4 leu2-A MATa ura3-52 CANl
EJL386b
his4-RI leu2-A MATa ura3-52 canlR HIS4leu2-AMATaura3-52CANl
lys2-801 CYH2 TRPl ade2-101' Y ' : : S U P l l and URA3 ASUx (at VI1 or X V ) lys2-r2 CYHP trpl::LEUZade2-101Y'::SUPll and LIRA3 ASUx
EJL396'
his4-RI leu2-A MATa ura3-52 canl HIS4leu2-AMATaura3-52CANl
lys2-801 CYHZ T R P l ade2-101' SUP5 lys2-r2 cyhPtr@l::LEU2 ade.2-101 O -k
"
-
-
lys2-801 CYHZ TRPl ade2-101" lys2-7-2 cyhPtrp1::LEUZade2-101"
"
and URA3) canlR+!2-801CYH2TRPlade2-101" and URA3) CANl lys2-rZ cyhP trp1::LEUZade2-101'
-
-
lys2-801CYH2 TRPl ade2-101" Y ' : : S U P l l and URA3 (at V l l or X V ) lys2-r2 CYHP trp1::LEUZade2-101" Y ' : : S U P l l and LIRA3
-
and URA3) canlRlyse-801CYH2TRPl ade.2-101" lys2-rZ c y h p trp1::LEUZ ade2-101"
-
lys2-801CYH2 TRPl ade2-101" Y'::SUPll and URA3 (at VII or X V ) lys2-r2 cyhP trp1::LEUZ ade2-101" Y'
-
-
~~
~
-
-
-
Designates anochremutation. Congenic with Y P l . ' Isogenic to Y P l . a
YP1-lys2-r2 and YPl-leu2A-trpl::LEU2) together as described below. EJL374 is the control,nonsuppressor bearing strain, that is 100% YP1. The MATa leu2A version of YPl was crossed to the LYS2 version of YP1. A MATa leu2A LYSZ segregant was then isolated and crossed to each of the canl and cyh2 versions of YP1. A MATa leu2A LYS2 canl segregant was then crossed to the hisl-R version of YPl . The segregant EJL360-1 D was isolated from this diploid. A MATa leu2A LYS2 cyh2 segregant was crossed to the lys2-r2 version of YP1. The leu2A trpl::LEU2 version of YP1 was crossed to a MATa leu2A LYSZ cyh2 segregant. A MATa leu2A lys2-r2 cyh2 segregant was crossed to a MATa leu2A trpl::LEU2 cyh2 segregant. The haploid EJL363-12B was isolated from this diploid. The suppressor bearing strains EJL384, EJL389 and EJL395 were constructed by transforming YP1 derived strains with YRpl4 or Y R p l J such that the plasmid with SUP11 was integrated at theURA3 location on chromosome V. These transformants were then crossed to EJL360-1D and EJL363-12B and segregants with the appropriatemarkers (indicated in Table 1) were isolated to make the diploids. The Y'::SUPll and URA3 YP1 transformant with theSUPll
at one end of chromosome VZZ or XV was crossed to YP1 derived strains in order to obtain the markers shown (Table 1). For the parents of strain EJL392, a cyh2 segregant was not obtained making it homozygous for CYH2. SUP5 was crossed into the strain background by serially backcrossing SUP5 segregants five times to YPI and then crossing a final SUP5 segregant to EJL360-1D. An antisuppressor, ASUx, was isolated spontaneously in a segregant from EJL77 which had a singlecopyof SUPll present. It is a Mendelian mutation that behaves as previously described antisuppressors ofSUP11 (MCCREADY and COX 1973,1976). The strain containing the antisuppressor is not completely isogenic with YP1 but is congenic (five times backcrossed to YPl). A cyh2 segregant was not obtained for the parentsof EJL386 making it CYH2 homozygous. Each ofthe strains in Table 1 was constructed by mating the appropriately transformed or backcrossed parents of EJL374 such that different dosages of suppressors, as well as, different chromosomal locations are represented. Selection of random spores: After sporulation for three days at 25", asci were subjected to digestion with glusulase (SHERMAN, FINKand HICKS1986)and sonication to separate
84
E. J. Louis and J. E. Haber
spores (LICHTEN, BORTS and HABER 1985) or to digestion without sonication. Separated spores were then plated on selective media.Suppressor bearing strains were sensitive to digestion with glusulase and to sonication. Spore viability was virtually zero after normal levels of digestion and sonication and therefore these spores were treated gently. Selection of spores was performed either by selection for lysine prototrophy (LIGHTEN,BORTSandHABER1987) or on medium containing both canavanine and cycloheximide (SHERMAN, FINKHICKS and 1986). Lysine prototrophs result from meiotic recombination 50-fold more frequently than mitotic recombination and therefore are enriched for haploid spores. Colonies resistant for both canavanine and cycloheximide could result from two rare mitotic recombination events or more likelyby meiotic segregation which again enriches for haploid spores. For diploids EJL77 and EJL97 spores were selected on canavanine and cycloheximide containing medium such that about 100 colonies per plate grew. After growth at 30" for 3 days, these plates were replica plated onto synthetic complete medium in which there is less adenine than normal (5 pg/ml rather than 40 pg/ml) (HIETERet al. 1985). White spore colonies were identified in order to find spores with two copiesof Y' ::SUP11 and URA3. Spore colonies from the rest of the diploids were grown for two to three days at 30" and each colony was patched onto the same selective medium used for spore selection. After growth overnight these patches were replica plated onto media for scoring amino acid auxotrophies and canavanine and cycloheximide resistance. Testing for mating type phenotype was also performed. Each nonmater was subsequently plated onto sporulation medium. Vegetative diploid cells that survived the selection for haploid spores were identified as sporulation-proficient nonmaters and were not included in the analysis. Sporulation-deficientnonmaters were analyzed as potential chromosome III disomes as described below. Vegetative diploid lysine prototroph colonieswere found ata frequency of 1% in the lysine selected spores. None were found for the canavanine, cycloheximide double selected spores. Analysis of potential disomes: Potential disomes were identified genetically. Chromosome III MATaIMATa disomes are sporulation-deficient nonmaters. Heterozygous (TRPlltrpl::LEU2) chromosome ZV disomes are prototrophic for both tryptophan and leucine. Heterozygous chromosome V (CANl/can 1 ") and VI1 ( C Y H 2 / ~ y h 2disomes ~) are initially canavanine and cycloheximide sensitive,respectively.Uponselection each yields resistant papillae that result from mitotic recombination between the centromere and CAN1 or CYH2 followed by segregation. Heterozygous (lys2-801/lysZ-r2) chromosome IZ disomes are lysine requiring but yieldlysine prototroph papillaeupon selection. These result from intragenic mitotic recombination between the lysZ heteroalleles. Disomes that are homozygous for the marker to be scored can not be detected here. Most potential disomes were crossed to appropriate testers and theresulting diploids were sporulated, dissected and scored for trisomic segregation (+/+/-) at the marker in question. The sporulation-deficient non-maters were force mated for analysis as described below. Analysis of potential chromosome ZZZ disomes: MATa/ MATa chromosome ZII disomes were force mated to testers as described by CAMPBELL, FOCELand LUSNAK (1975). His+ disomeswere force mated to his4-R testers in order to determine whether the disome was homozygous HIS4/HIS4 or heterozygous his4-R/ffIS4. Forced mating occurs either by homozygosis of MAT via mitotic recombination and segregation or by loss of one of the two chromosomes III. Loss
TABLE 2 Induction of chromosomeZZZ disomy by SUPZl No. chromosome Ill disomes
Strain
-
SUP1 1
EJL77
SUPII
EJLg7
supll
SUP1 I
EJL146
+ +
No. spores -
9
908
6
776
0
720
of one of the two chromosomes IZI is accompanied by gene FOCEL conversion at HIS4 in 10% of the cases (CAMPBELL, and LUSNAK 1975). The linkage relationship between MAT and the HIS4 alleles can be determined by the proportion of his4-R alleles revealed when the MATa versus M A T a bearing chromosome is lost. For example, a MATa-his4-Rl MATa-HIS4 disome will yieldHis-diploidsupon forced mating to aM A T a tester roughly nine times more frequently than when force mated to a MATa tester. Tetrad and spore viability analysis:Diploids with different dosages and/or positions o f S U P l l were sporulated. Asci were dissected as described by SHERMAN, FINKand HICKS (1986). The viability of each diploid was determined and the viable spores were scored for all segregating markers. Tetrads with apparently no viable spores were checked by microscopic examination to confirm that there were spores which failed to germinate or grow. These were then allowed to incubate for several more days in order to allow for possibleslow germination and growth. Sister spores are defined as those that result from the second meiotic division and can be scored using a centromere linked marker. The trp1::LEUZ disruption which is tightly linked to the centromere of chromosome 1V was used to determine the sister versus nonsister spore status of two viable spored tetrads. Sister spores are then those that are both Trp+ and Leu- or both Trp- and Leu+.Map distances between the centromere and MAT or HIS4 and between MAT and HIS4 were calculated using a maximumlikelihood estimation procedure (SNOW1979). Significance of deviations between the observed spore viability and those expected from random spore inviability was tested using the G-test (SOKALand ROHLF 1969). Significance of deviations between observed spore viabilities in suppressor bearing strains to spore viabilities of the nonsuppressor bearing strain was also calculated using the G-test. Significance ofdeviations from 50:50 sisters to non-sister spores in the two viable spored tetrads was tested using the exact binomial distribution. Homogeneity of the recombination data of the differentdiploids was tested using the G-test of homogeneity (SOKAL and ROHLF 1969). The G-test (a x ' statistic) is used as it allows for small (
