Formamide Sensitivity: a Novel Conditional Phenotype in Yeast

Formamide Sensitivity: A Novel Conditional Phenotype in Yeast

Andres Aguilera Departamento de Genitica, Facultad de Biologia, Universidad de Sevilla, E-41012 Sevilla, Spain Manuscript received June 16, 1993 Accepted for publication September 28, 1993

ABSTRACT Yeast mutants unable to growin the presence of 3% formamide have been isolatedin parallel with mutants sensitive to either 37" or 6% ethanol. The number of formamide-sensitivemutations that affect different genes that can be identified from yeast cells is at least as large as the number of thermosensitive or ethanol-sensitive mutations. These mutations are of two types: those that are sensitive to formamide, temperature and/or ethanol simultaneously; and those that are specific for formamide sensitivity and show no temperature or ethanolsensitivity phenotype. Those genes susceptible to giving rise to formamide-sensitive alleles include the structural gene for DNA ligase, CDC9, and the structural gene for arginine permease, CANl. The results indicate that formamide sensitivitycan be used as a novelconditional phenotype for mutations on bothessential and nonessential genes. This work also confirms that ethanol-sensitivity can beused as a conditional phenotype._ to identify mutations in at least as many genes as those susceptible to temperature or formamide sensitive mutations.

HE understanding of any biological process re- the differentbiological processes studied. Second, not T quiresthe identificationof theproteins and all amino acids of a protein are equally susceptible to genes involved in it. The contributionof genetic lead to athermosensitive phenotype (ALBERet al. methods to this understanding through the isolation 1987). Third, the identification and genetic analysis ofmutants is very important.However, there are of extragenic suppressors of thermosensitive muta- obvious limitations for the identification of mutations tions in essential genes require a second conditional in essential genes. Conditional mutations, that is mu- phenotype (MOIR and BOTSTEIN1982). And fourth, tations that only express the mutant phenotype under the use of different temperatures is not possible for some restrictive conditions (HOROWITZ1948), serve homeothermic eukaryotes. to overcome this problem. Recently BARTELand VARSHAVSKY(1988) used Temperature has been used extensively to isolate heavy water (D20) as a conditional mutant phenotype conditional mutantssince first proposedby HOROWITZ in yeast. Their work provides an important tool to and LEUPOLD(1 95 1). The mutant phenotype can be identify novel conditionalmutations and open new expressed at eitherhigh (thermosensitive) or low (cry- ways to isolate conditional mutants in homeothermic osensitive) temperatures. This technique has allowed organisms. the identification of many genes involved in essential A characteristic that can also be used to isolate biochemical and cellular processes, such as DNA rep- conditional mutants is sensitivity to ethanol,which has lication, cell division, cytoskeleton structure, etc., been characterized at the genetic and physiological since the first thermosensitive mutants were reported level (AGUILERAand BENITEZ1986). A great many in phage (EDGARand LIELAUSIS1964; EPSTEINet al. genes can be mutated to produce anethanol-sensitive 1963), bacteria (EIDLICand NEIDHART1965, KOHI- phenotype. A number of these mutations also lead to YAMA et al. 1966) and yeast (HARTWELL, 1967). Dif- thermosensitivity. That study suggests that many ferentmutant proteins contain changes in charge, genes,not necessarily involved in glucose or lipid hydrophobicity, solvent structure and hydrogen bond- metabolism, are susceptible to mutations that lead to ing that influence its secondary or tertiary structure an ethanol-sensitivity phenotypefor growth. How- and that can confer a thermolability phenotype (AL- ever, the use of ethanol-sensitivity as aconditional BER et d. 1987). phenotype may berestricted to organisms such as There areseveral reasons to believe that alternative yeast that tolerate ethanol concentrationsof 6%. conditionalphenotypes are necessary. First, not all In this study, formamide sensitivity as a novel con- proteins are equally susceptible to aminoacid changes ditional phenotype in yeast was investigated for the that make them thermolabile; the extensive work of following reasons: (1) formamide is an ionizing solvent HARTWELL(1 967) on yeast suggested that the isola- that destabilizes noncovalent bonds such as hydrogen tion of thermosensitive mutants is not random among bonds; (2) formamide is not metabolized by cells and Genetics 136: 87-91 (January, 1994) 88 A. Aguilera there are no reasons to believe a priori that yeasts TABLE 1 have evolved as highly formamide-tolerant organisms Classification of 41 conditional mutants isolated as is the case for ethanol; and(3) formamideis a small molecule (HCONH2) that should enter cellseasily. Strains We have isolated some mutants that are simultane- Conditional Total phenotype" AYW3-3A AYW3-4b mutants ously formamide, ethanol and temperature sensitive for growth, andsome others that areexclusively form- fs ts es 3 6 9 amide sensitive. We have also identified formamide- fs ts 2 3 5 0 6 6 sensitive alleles of the CDC9 and CAN1 genes. The fs es fs 7 5 12 genetic analysis of the formamide sensitive mutants ts es 0 0 0 isolated in this study suggests that any type of gene, ts 4 1 5 whether essential or not, is susceptible to formamide es 1 3 4 sensitive mutations. Total 17 24 41 "3:sensitivity to 3% formamide; ts: sensitivity to 37°C; es: sensitivity to 6% ethanol. Sensitivity is defined in this table as the MATERIALS AND METHODS inability to form colonies. Strains: The yeast strains usedin this study arethe congenic strains W303-1B (MATa ade2-1 leu2-3, I12 ura3- AYW3-4B were mutagenized with nitrosoguanidine 1 trpl-1 his3-11, 15 canl-loo), AYW3-3A (MATa ade2 leu2- as described in MATERIALS AND METHODS. A total of 3, 112 ura3 trpl his3), and AYW3-4B (MATa ade2 leu2-3, 2635 surviving colonies (1887 from strain AYW3-4B 112 ura3 trpl his3), 5a-P3631 (MATa his7-1 lys2-18), and and748 from AYW3-3A) were picked onto new C9Y-1B (MAT& leu2 his3 cdc9-I). Twelve other strains car- YEPD plates andafter 2 days at 30°, they were rying 1 1different thermosensitive mutations in 10 cell cycle genes were used.These mutations were hpr6 (a CDC2 allele), replica-plated onto YEPD-3% formamide, YEPD-6% cdc4, cdc5, cdc6,cdc8-I, cdc9-1, cdcl3-I, cdcl4, cdcl7-I, hpr3 ethanol and YEPD. The first two groups of plates (a CDCl7 allele) and cdc28. The cdc alleles used are original were cultured at 30O , and the YEPD plates at 37 " C, from L. Hartwell, C.Newlon and H. Klein laboratories. in orderto scorefor a formamide-sensitive (The allele number is not given when it is not known.) (fs), ethanol-sensitive (es) or thermosensitive (ts) pheno- Media and growth conditions: Standard media such as rich medium YEPD, synthetic complete medium with bases type for growth. After three rounds of scrutiny, a and amino acids omitted as specified, and sporulation me- total of 41 mutants (17 from strain AYW3-3A and 24 dium were prepared according to published procedures from AYW3-4B) with an fs, es or ts phenotype were (SHERMAN,FINK and HICKS1986). L-Canavanine sulfate was selected. From Table 1 two results can be highlighted: added to synthetic medium at concentrations of 60 pg/ml. Formamide or ethanol were added to the medium at the of the 41 conditional mutants isolated 9 were simul- indicated concentrations after autoclaving. All yeast strains taneouslyfs, es and ts; and of 32fs mutants isolated, were grown at 30" with horizontal shaking for liquid cul- 12 were exclusively fs, and the other 20 were also ts tures. or es or both. These results indicate that: first, the Mutagenesis: Yeast strains were grown overnight in 5-ml number of mutations leadingto a formamide-sensitive YEPD. Cells in early stationary phase were resuspended in Tris-maleate pH 7.8 to a density of 10' cells/ml and muta- phenotypearose at afrequency similar tothat of genized with a final concentration of 20 rg/ml N-methyl- mutations leading to a ts phenotype or an es pheno- N'-nitro-N-nitrosoguanidine (nitrosoguanidine) for 15 min type;second, mutations conferring a fs phenotype according to CALDERONand CERDA-OLMEDO(1983). Muta- could lead simultaneously to a ts and/or es phenotype; tions were allowed to segregate by culturing the mutagenized cells in liquid YEPD for 6 hr at 30" prior to plating. and third, many mutations were specific for an), es The viability of the cells after mutagenesis was approxi- or ts phenotype. mately 30%.Mutagenized cells were plated on YEPD plates Genetic analysis of formamide-sensitive mutants: to isolatesingle colonies and on SC-can and SC-can-2% Twenty formamide-sensitive mutants (1 2 from strain formamide plates to select for CanR colonies in either the AYW3-4B and 8 from AYW3-3A) were selected for absence or the presence of 2% formamide, respectively. After 2 days at 30°,colonies from YEPD plates were replica- further genetic studies. All of them were crossed with plated onto YEPD-3% formamide, YEPD-6% ethanol and either strain AYW3-3A or AYW3-4B to determine YEPD. Those plates containing formamide or ethanol were the recessivity or dominance of the fs phenotype. In cultured at 30" and the simple YEPD plates at 37". all cases, diploid strains were able to grow on YEPD- Genetic analysis:Genetic analysis was performed accord- like ing to SHERMAN,FINK and HICKS(1 986). 3% formamide the wild-type haploid parental strains, indicatingthat all 20 mutations were recessive. That most of the mutations were monogenic was RESULTS confirmed by a genetic analysis of 12 of the selected Isolation of formamide-sensitive mutants: Wild- fs mutants(10 from strain AYW3-4B and 2 from type cells grow very poorly at formamide concentra- strain AYW3-3A). Between 9 to 14 tetrads were dis- tions above 3% v/v and not at all at a concentration sected and subjected to genetic analysis. Nine of the of 10% (datanot shown). Strains AYW3-3A and 12 crosses analyzed showed the 2:2 segregation ex- Forrnarnide Sensitivity 89 TABLE 2 Genetic analysis of the cross of nine selected formamide- sensitive mutants with a wild-type strain"

f~ sepegationb Complete Mutants tetradsMutants 2Js: 2 wt Others

AYW3-4BlC 11 11 0 AYW3-4B3C 9 9 0 AYW3-4B4 9 9 0 AYW3-4B5C 7 6 1 AYW3-4BW 8 8 0 AYW3-4B10C 9 9 0 AYW3-4B125 5 0 AYW3-4B20 7 7 0 AYW3-4B21 7 6 1

a The wild-type parental strain used was either W303-1B or 5a- P363 1. bfs: forrnamide-sensitive; wt: wild-type. FIGURE1 .-Segregation analysis of seven complete tetradsfrom Crosses where the therrnosensitivity phenotype was shown to the cross W303-1B (canl-100) X A3A-CANSF (canIfi)on SC-can cosegregate with thefs phenotype; the rest were onlyfs. (upper) and SC-can-2%forrnamide (lower). The four spores of each tetrad are ordered vertically. To facilitate the identification of the pected for a monogenic character (Table 2). Also, in tetrads, the first spore of each tetrad is displayed as a number. those cases in which the thermosensitive phenotype could be scored, there was a cosegregation of the ts when grown in the presence of 6% ethanol or at 37". phenotype with thefs phenotype,which indicates that To determine if this mutation maps at the CAN1 the same mutation is responsible forthe fs and ts locus, the mutant strainA3A-CAN2F was crossed with phenotypes. W303-3B carrying the nonconditional allele canl-100. To establish complementation groups, each of the Tetrad analysis of dissected spores showed that the eightmutants derived from strain AYW3-3A were formamide-conditional CanR mutationof strain A3A- crossed with each of the 12 mutants derivedfrom CANSF was linked to the CANI locus (Figure 1). A 4 AYW3-4B. All combinations studied complemented CanR;0 CanS segregation was obtained on SC-can-2% as shown by the fact that all 96 diploids obtained were formamide for the seven complete tetrads analyzed, able to grow on YEPD-3% formamide. Thus, as far whereas a 2 CanR:2 CanS was obtained on SC-can as as could be tested, most of thef mutations identified expected for the canl-100 allele that is the only allele affected different genes. expressing the CanR phenotype onthis latter medium. Isolation of a cad formamide sensitive allele: A thermosensitive cdc9 allele that is formamide The results shown suggest that formamide could be sensitive: The large number offs mutants obtained used to isolate conditionalmutations in essential in this study that were simultaneously ts suggests genes. To confirmthat formamide-sensitive alleles that mutations initially isolated on the basis of a tem- could beobtained from a wide range of genes, perature-sensitive conditional phenotype may also be whether essential for growth or not, a procedure that formamide-sensitive. To test this possibility, the sen- identifies CanR mutants that express the canavanine sitivity to formamide of 13 different mutant strains resistance phenotype only in the presence of 2% form- carrying 1 1 different thermosensitive mutations in 10 amide was carriedout. Canavanine is anarginine cell cycle genes was determined. The mutations were analogue that is toxic to thecell. Mutations conferring hpr6 (a CDC2 allele), cdc4, cdc5, cdc6, cdc8, cdc9, cdcl3, a CanR phenotype map at the CANI locus, which is cdcl4, cdcl7, hpr3(a CDCI 7 allele) and cdc28; and all thestructural gene for arginine permease, can be of them confer cellcycle arrest phenotype at 37". easily selected on SC-can media. Only the two strainscarrying the same cdc9 allele From the same mutagenesis mentioned above, 1.5 (original from L. HARTWELLstrain H9ClAl) were X lo7 cells from strain AYW3-3A and 3.8 X lo7 cells unable to grow on YEPD-3% formamide at 30". The from AYW3-4B were plated on SC-can containing 2% fact that the cdc9 mutation itself led to thefs pheno- formamide. A total of 641 CANR colonies was ob- type was confirmed by demonstrating that cdc9 strains tained (143 from AYW3-3A and 498 from AYW3- showed the same cell-cycle arrest phenotype in liquid 4B). Of these mutants, 640 were also CanRin SC-can YEPD-2% formamide medium at 30" and YEPD at without formamide, indicating that they were not 37" (Figure 2). conditional forformamide. One mutant, however, was CanRwhen the medium contained 2% formamide, DISCUSSION but was unable to grow on SC-can without formamide. In this work it has been shown that sensitivity to 3% This mutant also showed a wild-type CanS phenotype v/v formamide can be used as a novel and comple- 90 A. Aguilera fs mutants obtained show sensitivity to temperature and ethanol, but 18 were not ts and 17 were not es (Table 1). This suggests that formamide, temperature and ethanol sensitivity share common molecular bases but that there aredifferences that give rise to proteins sensitive to only one agent. In this sense, BARTELand VARSHAVSKY(1988) did not find a high degree of overlap between heavy water sensitivity and thermosensitivity. Also, it was previously shown (ACUILERA and BENITEZ1986) that the lower the ethanol concentration es mutants could resist, the less likely to be ts. Of 10 cdc ts alleles tested in this study, only one was sensitive to 2% formamide. It could be that either the stringency of conditions used lead to a bias in the type of allele or the type of genes that can become sensitive to a particular solvent or that lethal mutations classi- fied with only one conditional phenotype such as>, share otherphenotypes such as ts or es, but that at the conditions used (37" or 6% ethanol) they are partly functional so that cells are viable. The formamide sensitivity phenotype can be obtained for any type of gene independent of its func- C tion. A DNA-ligase allele, cdc9, initially isolated as ts has been shown to be also fs, and a new can1 allele has been isolated which expresses the CanR phenotype only in the presence of 2% formamide. The fact that a protein involved in DNA metabolism and a protein involved in transport across the membrane canbe- come formamide sensitive suggests that the sensitivity to formamide should be related to the structure of the protein itself more than to the type of function FIGURE 2.-Microphotograph of yeast cells from strain CSY-IB (cdc9) after 6 hr in YEPD at 30" (A), YEPD-2% formamide at 30" which is involved. (B) and YEPD at 37" (C). Many different types of noncovalent interactions (ion pairs, hydrogen bonds, van der Waals contacts, mentary conditional phenotype for mutationson both hydrophobic interactions) contribute to the stability essential and nonessential genes as suggested by the of a protein (MATTHEW, WEAVER andKESTER 1974). following results: (1) the number of formamide-sensi- Changes in charge, hydrophobicity, hydrogen bond- tive lethal mutations affecting different genes thatcan ing and folding of a protein will depend either on the be isolated from yeast cells is at least as large as that nature of the amino acid changed or the structural of thermosensitive or ethanol-sensitive lethal muta- contextwhere that amino acid is located (i.e., an tions; (2) these mutations either confer sensitivity to amino acid may be located in a substructure unacces- formamide, temperature and/or or ethanol simulta- sible to solvents). Formamideand ethanol contain neously or are exclusively formamide sensitive; and chemical groups (H, 0, OH, NH) that compete for (3) those genes susceptible to formamide-sensitive mu- noncovalent bonds with the aminoacids of the protein tations include the structural gene for DNA ligase, so that the functional conformation of the protein can CDC9, and the structural gene for arginine permease, bedisrupted. This, together with the fact thatthe CANI. This work alsoserves to confirm, together with denaturationproperties of anumber ofchemical a previous report (ACUILERAand BENITEZ1986), that agents are different (GORDONand JENCKS 1963), can ethanol sensitivity can be used as a conditional phe- explain that some amino acid changes lead to proteins notype to identify mutations in at least as many genes thatare unstable at high-temperature,ethanol or asthose susceptible totemperature or formamide formamideconcentrations, whereas others lead to sensitive mutations. proteins that are unstable only under one or two of The frequencies at which>, es and ts mutants can the conditions mentioned. be obtainedby mutagenesis are very similar. The same Finally, one conditional phenotype may not be overlap shown for ts and fs phenotypes exists for fs enough to isolate mutations in all essential genes of and es and for ts and es phenotypes. Nine of a total 32 an organism (KABACKet al. 1984). Mutations leading Form am ide Sensitivity Formamide 91 to one conditional phenotype may not be distributed ALBER,T., S. DAO-PIN,J. A. NYE, D. C. MUCHMOREand B. W. randomly among allgenes. HARTWELL(1967) ob- MATHEWS,1987 Temperature-sensitive mutations of bacte- served that not allbiological processes are equally riophage T4 lysozyme occur at sites with low mobility and low solvent accessibilityin the folded protein. Biochemistry 26: affected by ts mutation. Some macromolecular struc- 3754-3758. tures or biochemical reactions may be more sensitive BARTEL,B., and A. VARSHAVSKY,1988 Hypersensitivity to heavy to a particular agent than others. High order macro- water: a new conditional phenotype. Cell 52: 935-94 1. molecular structures or their associatedbiological CALDERON,1. L., and E. CERDA-OLMEDO,1983 Induction byN- processes (cytoskeleton, ribosomes,replication, trans- methyl-N’-Nitro-N-Nitrosoguanidineof nuclear and cyto- port, etc.), and nonproteic structures (membranes), plasmic mutations in Saccharomyces cereuisiae. Mutation Res. could be specially sensitive to a particular agent. If 108 133-146. EDGAR,R. S., and I. LIELAUSIS,1964 Temperature-sensitive mu- this were the case the conditional phenotype would tants of bacteriophage T4D: their isolation and genetic char- be caused by the lability of the entire structure (it?., acterization. Genetics 49 649-662. cryosensitivity of yeast cytoskeleton, MOIR and BOT- EIDLIC,L., and F. NEIDHARDT,1965 Protein and nucleic acid STEIN (1982) or biological process (i. e., cryosensitivity synthesis in two mutants of Escherichia coli with temperature- of prokaryotic protein export, P~CLIANOand BECK- sensitive aminoacyl ribonucleic acid synthetases. J. Bacteriol. WITH, 1993) and not by the lability of the mutated 89: 706-7 1 1. EPSTEIN,R. H., A. BOLLE,C. M. STEINBERG,E. KELLENBERGER,E. protein itself. The fact that the$ and es phenotypes BOY DE LA TOUR,R. CHEVALLEY,R. S. EDGAR,M. SUSMAN,G. correlate with the ts phenotype in a number of mu- H. DENHARDTand A.LIELAUSIS, 1963 Physiological studies tants suggests that the conditions used in this work of conditional lethal mutants of bacteriophage T4D. Cold affect the labilityof the protein itself. However, it Spring Harbor Symp. Quant. Biol. 28: 375-394. cannot be discarded that some macromolecular struc- GORDON,J. A., and W. P. JENCKS, 1963 The relationship of tures could be specially sensitive to a particular sol- structure tothe effectiveness of denaturing agents for proteins. vent. Biochemistry 2: 47-57. In summary, experimental conditions can be estab- HARTWELL,L. H., 1967 Macromolecule synthesis in temperature- sensitive mutants of yeast. J. Bacteriol. 93: 1662-1670. lished under which a particular amino acid change HOROWITZ,N. H., 1948 The one gene-one enzyme hypothesis. will make a protein susceptible to selective inactivation Genetics 33: 6 12-61 3. by a specific chemical agent, such as formamide or HOROWITZ,N. H., and U. LEUPOLD,1951 Some recent studies ethanol, to a degree that leads to expression of the bearing on the one gene-one enzyme hypothesis. Cold Spring mutant phenotype. The use of different conditional Harbor Symp. Quant. Biol. 16: 65-72. phenotypes should contribute greatly to the identifi- KOHIYAMA,M., D. COUSIN, A.RYTER and F. JACOB, 1966 Mutants cation of conditional lethal mutations in most of the thermosensibles d’Escherichia coli K12. I. Isolement et charac- essential genes of an organism and to isolate novel tkrisation rapide. Ann. Inst. Pasteur (Paris) 110 464-486. KABACK,D. B., P. W. OELLER,H. Y. STEENSMA,J. HIRSCHMAN, D. conditional alleles of known genes. Formamide and RUEZINSKY,K. G. COLEMANand J. R. PRINGLE,1984 temperature could be used to create amore stringent Temperature-sensitive lethal mutations on yeast chromosome condition to identify conditional mutations with a I appear to define only a small number of genes. Genetics 108 profound phenotype. Finally it should open new per- 67-90. spectivesto the isolationof conditional mutants in MATTHEW,B. W., L. H. WEAVERand W. R. KESTER, 1974 The other organisms. conformation of thermolysin. J. Biol. Chem. 249 8030-8044. MOIR, D., and D. BOTSTEIN,1982 Determination of the order of I would like to thank I. L. CADER~N andT. BEN~TEZ for their gene function in the yeast nuclear division pathway using cs critical reading of the manuscript, R. RHETTfor style correction, and ts mutants. Genetics 100 565-577. and L. HARTWELL,C. NEWLON,D. A. GORDENIN,R. ROTHSTEIN POGLIANO,K. J., and J. BECKWITH,1993 The Cs sec mutants of and H. L. KLEIN for providing yeast strains. This work was sup- Escherichia coli reflect the cold sensitivity of protein export ported by DGICYT grant PB90-0800C02-02. itself. Genetics 133: 763-773. SHERMAN,F., G. R. FINKand J. B. HICKS,1986 Methods in Yeast LITERATURECITED Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, AGUILERA,A., and T. BENITEZ,1986 Ethanol-sensitive mutants N.Y. of Saccharomyces cereuisiae. Arch. Microbiol. 143: 337-344. Communicating editor: M. CARLSON