Temperature-Sensitive Defects in Both Mitochondrial Protein Import ...

MOLECULAR AND CELLULAR BIOLOGY, May 1991, p. 2647-2655 0270-7306/91/052647-09$02.00/0 Copyright t 1991, American Society for Microbiology

Vol. 11, No. 5

A Mutation in the Yeast Heat-Shock Factor Gene Causes Temperature-Sensitive Defects in Both Mitochondrial Protein Import and the Cell Cycle BARBARA J. SMITH AND MICHAEL P. YAFFE*

Department of Biology, University of California, San Diego, La Jolla, California 92093 Received 11 October 1990/Accepted 15 February 1991

Yeast cells containing the recessive mas3 mutation display temperature-sensitive defects in both mitochondrial protein import and the cell division cycle. The import defect is characterized by two pools of mitochondrial precursors and a dramatically slower rate of posttranslational import. The effect of mas3 on cell cycle progression occurs within one cell cycle at the nonpermissive temperature and retards progression through the G2 stge. The mas3 mutation maps to the gene encoding yeast heat-shock transcription factor (HSF), and expression of wild-type HSF complements the temperature-sensitive defects. The mas3 lesion has no apparent effect on protein secretion. In mas3 cells, induction of a major heat-shock gene, SSA1, is defective at 37C. The properties of the mas3 mutant cells indicate that HSF mediates the response to stress of two basic cellular processes: mitochondrial protein import and cell cycle progression.

Heat shock is a ubiquitous response of cells to environmental stress (12). A central feature of this response is the rapid, increased synthesis of specific heat-shock proteins (hsps) which are highly conserved among diverse organisms (25). Some of these hsps or homologous proteins are expressed in cells under all growth conditions and play essential roles in cellular function and proliferation (13). While widely assumed to confer thermotolerance on heat-stressed cells, the specific functions of individual hsps in the heatshock response and the relationship between hsp function under stressed and nonstressed conditions are unknown. One function of hsps or their cognate proteins is to facilitate the import and assembly of mitochondrial proteins. Studies in yeast cells have implicated the 70-kDa hsps (HSP70s) of the cytoplasm (7, 14) and two mitochondrial matrix proteins, HSP60 (3, 18) and Ssclp, an HSP70 homolog (5, 11), as components important for the mitochondrial assembly process. The cytoplasmic hsps may unfold or maintain an import-competent conformation of mitochondrial precursor proteins, while the matrix proteins may catalyze refolding or assembly of imported subunits. These hsp functions are required by cells even under nonstressed conditions, but their additional roles in the heat-shock response are unknown. We have been studying mutants of Saccharomyces cerevisiae defective in mitochondrial protein import to identify components of the import apparatus. Our previous investigations of two mutants, masi and mas2, led to the identification of products of the MASI and MAS2 genes as subunits of the protease that processes precursor proteins upon their import into mitochondria (10, 37, 39). In this report we describe a third mutant, mas3, that was identified by its defects in growth and mitochondrial protein import at an elevated temperature. Cells with the mas3 mutation display two properties unique from cells with masi and mas2 lesions: the defect in protein import appears to affect primarily posttranslational import, and mas3 cells exhibit a temperature-dependent block in cell cycle progression. We demon-

*

strate below that the mas3 lesion lies in the gene for heat-shock factor (HSF) and suggest a genetic basis for the interrelationship of three essential cellular processes: heat shock, mitochondrial protein import, and progression through the cell division cycle.

MATERIALS AND METHODS Strains and genetic techniques. The parent strain AH216 (MATa leu2 his3 phoC phoE) and isolation of the mas mutants have been described previously (39). Strains MYY290 and MYY291 are ura3 derivatives of strain AH216 with mating types MATa and MATa, respectively. Strain MYY260 (MATa leu2 his3 ura3 trpl lys2 mas3) was derived as a haploid spore from a cross of the original mas3 mutant strain, MYY243 (MATa leu2 his3 phoC phoE mas3), to strain YPH54 (MATa ura3 lys2 ade2 trpl his3) (27). Strain DL1 (MATa leu2 his3 ura3) was described previously (33). Strain MYY385 (MATa leu2 his3 phoC phoE ura3 mas3) was derived as a haploid spore from a cross of strain MYY243 to strain MYY291. Strain H146-5-3 (MATot cdc21-1 adel ural leu2) was the gift of B. Garvick and L. Hartwell (University of Washington). Strain MS177 (MATa kar2-159 ura3 ade2) was the gift of M. Rose (Princeton University). Plasmid DNA was prepared in Escherichia coli DH5a and MH6. Standard yeast genetic techniques and media were as described previously (26). Cellular labeling and immunoprecipitation. Cells were grown to an optical density at 600 nm (OD6.) of between 1.0 and 3.0 in semisynthetic medium (6) containing 2% glucose or 2% lactate or to an OD6. of 1.0 to 2.0 in minimal (synthetic) medium with glucose (for cells with plasmids). Cellular labeling, protein extraction, and immunoprecipitation were performed essentially as described previously (39) except that cells were resuspended in either fresh semisynthetic-glucose medium or minimal glucose medium and labeled for 5 min with [35S]-Translabel (100 mCi/10 OD6. of cells; ICN). Treatment with carbonyl cyanide m-chlorophenylhydrazone (CCCP) and subsequent import studies were done as described previously (20) except that semisyntheticglucose medium was used, cells were incubated with CCCP

Corresponding author. 2647

2648

SMITH AND YAFFE

to a final concentration of 20 ,uM, [35S]-Translabel was added to 100 ,uCi/10 OD6. of cells for 10 min, and P-mercaptoethanol, unlabeled methionine, and cycloheximide were added after the labeling period to final concentrations of 0.05%, 2 mM, and 0.1 mg/ml, respectively. Antisera. Antiserum against the , subunit of Fl-ATPase against yeast (F1l) was described previously (10).giftAntibody of A. Lewin (Univercitrate synthase was the generous sity of Florida). Antisera against yeast a-factor and against yeast carboxypeptidase Y were gifts of R. Schekman (University of California, Berkeley). Analysis of cell cycle arrest. The effect of the mas3 mutation on progression through the cell cycle and the ability of various DNA constructs to complement the cell cycle defect was analyzed by microscopic examination. Cells were grown first on either YPD or selective (synthetic minimal) medium (if plasmid selection was necessary) at 23°C. Cultures were then shifted to 37°C and incubated for various times. Random fields of cells were counted in a hemacytometer, and cellular morphologies (e.g., dumbbell-shaped cells) were recorded. A large-budded cell was defined as one having a bud at least half the size of the mother cell. Duplicate samples were analyzed, and 150 to 300 cells were counted for each sample. In some experiments cells were treated with the yeast mating pheromone a-factor (Sigma), hydroxyurea, or nocodazole (Sigma) at final concentrations of 4 ,ug/ml, 0.1 M, and 12.5 ,ug/ml, respectively. To examine the fate of dumbbell-shaped cells after an incubation at the nonpermissive temperature, individual cells were separated on thin agar slabs (YPD medium; prepared with or without 0.2 M hydroxyurea) by micromanipulation. Slabs were incubated for various periods at 23°C, and cells were then observed microscopically. Analysis of secretion. Secretion of a-factor was analyzed by pulse-labeling of cells with [35S]methionine, protein extraction, and immunoprecipitation as described above except that cells were preincubated at 37°C for 30 min and labeled with [35S]methionine (300 ,jCi/2 A6. of cells) at 37°C for 30 min, and cycloheximide (to 0.1 mg/ml) and unlabeled wild-type (MATa) cells were added to the samples prior to protein extraction. Isolation of the MAS3 gene. The MAS3 gene was isolated by genetic complementation of the mas3 temperature-sensitive lethal mutation. The mas3 strain MYY260 was transformed with a yeast genomic plasmid library in the vector YCp5O (21) as described by Ito et al. (9). Ura+ transformants were selected at 23°C and screened for growth at the nonpermissive temperature by two successive replica platings to 37°C. Of 4,000 Ura+ transformants analyzed, one complementing plasmid, YCpM3, was isolated. This plasmid contained a 12.6-kbp insert of yeast DNA. To further localize the MAS3 gene, portions of the 12.6kbp fragment were subcloned into YCp5O, and the ability of these constructs to complement the temperature-sensitive defect in mas3 cells was analyzed. Complementing activity was localized to a 5.8-kbp XhoI-SalI fragment contained within plasmid YCpCE. Integrative transformation. A 3.5-kbp BamHI fragment isolated from YCpM3 was inserted in to the URA3-containing vector YIp5 (31), and the resulting plasmid, YIpB3, was linearized at the unique XhoI site prior to transformation of strain DL1. Stable Ura+ transformants were crossed to strain MYY260, and the diploids were sporulated at 23°C. The meiotic products were tested for growth at 37°C and for growth on media lacking uracdil. In crosses to two independently isolated integrants, no recombinants between mas3

MOL. CELL. BIOL.

and URA3 were observed in 35 tetrads analyzed. All tetrads were of the parental ditype (2 Ura+, 2 Mas+:2 Ura-, 2 Mas-). Thus, YIpB3 integrated within 1.4 centimorgans of MAS3. Sequencing of the MAS3 gene. A 5.8-kbp XhoI-Sall fragment which complemented all of the mas3 mutant phenotypes was ligated into the SalI site of plasmid pUC19, yielding plasmid pucMAS3. A DNA fragment containing MAS3 was then reisolated from pucMAS3 by cutting in the polylinker region with PstI and XbaI. This fragment was inserted into the PstI and XbaI sites in the vector Bluescript SK (Stratagene) to create plasmid BsMAS3. The nucleotide sequence of the 5.8-kbp yeast sequence was determined by using the Sequenase 2.0 DNA sequencing kit (U.S. Biochemical Corp.). Fragments to be sequenced were obtained by (i) exonuclease III digestion of BsMAS3 and pucMAS3, creating nested deletions from either end of the yeast insert; (ii) subcloning of small restriction fragments by using convenient sites within pucMAS3 or BsMAS3 and reinserting them into Bluescript SK; and (iii) constructing internal deletions by using sites within the fragment to be sequenced and sites in the polylinker of pucMAS3 or BsMAS3. Plasmid complementition. Various DNA fragments (see Fig. 5) were tested for their ability to complement the protein import and cell division cycle defects of mas3 cells. Portions of the 5.8-kbp SalI-XhoI DNA fragment containing the HSF (MAS3) gene were subcloned into yeast vectors, the plasmids were transformed into the mas3 strain MYY260, and cellular morphology and mitochondrial protein import were examined at 37°C as described above. Plasmid YCHSF5.3 was constructed by isolating the 5.8-kbp ClaI-EagI fragment from plasmid BsMAS3 and inserting it in the ClaI and EagI sites in plasmid YCp5O. Plasmid pADH-HSF was constructed by isolating the 2.8-kbp PvuII fragment containing the HSF gene (and some flanking sequence from the vector) from pucMAS3 and ligating this piece into the filled EcoRI site of plasmid pACi (8) behind the active ADHI promoter. Plasmid YCHSFA3' was made by cutting plasmid YCHSF5.3 with EagI, filling the ends, cutting with BamHI, and replacing most of the wild-type HSF gene by inserting a 900-bp BamHI-blunt DNA fragment whose sequence extends from the BamHI site to 16 nucleotides past the Ball site. Plasmid YCHSF5.3B was constructed by cutting plasmid YCHSF5.3 with BamHI, filling the DNA ends with Klenow polymerase, and religating. The DNA sequence changes created in plasmid YCHSF5.3B were confirmed by nucleotide sequencing, using a synthetic oligonucleotide primer (5'-GGTTTCCTTG AGTCATGG-3'; Operon, Alameda, Calif.) which hybridizes adjacent to the unique (filled) BamHI site in the plasmid. Fluorescence microscopy. Fluorescence microscopy of yeast cells was performed essentially as described by Thomas and Botstein (32). Cells used for this analysis were grown to early log phase in YPD, back-diluted to an ODwo of 0.1 to 0.2, and shifted to 37°C for various times prior to processing for 4,6-diamidino-2-phenylindole (DAPI) staining. Analysis of gene expression. Yeast strains were grown at 23°C in SD medium to an OD6. of 0.5 to 0.7. Cells were concentrated in fresh SD to 5 OD6.Jml, transferred to prewarmed flasks, and incubated at 37°C with vigorous shaking. Aliquots of cells were removed after 0, 15, 30, and 60 min, resuspended in RNA extraction buffer (0.5 M NaCl, 200 mM Tris [pH 7.5], 1 mM EDTA), and quick-frozen in a dry ice-ethanol bath. Total RNA was extracted as described previously (1). RNA (-5 ,ug) was separated by electrophoresis in a 1.1%

VOL . 1 l, 1991

YEAST TEMPERATURE-SENSITIVE HEAT-SHOCK FACTOR wt

chase-

0

mas

2

0

3

2

ma s

10

0

2

2

mas 3

wt

2649

mas 2

10 CHASE

0

15

60

0

15 40 60 0

15 40 60

A

%'.40w

-_.

B .-

_ _ _

_.

_ l! _ nz--

mmb.

FIG. 1. Evidence that the pattern of protein import into mitochondria in mas3 cells is consistent with two pools of precursor. The wild-type (wt) strain, mas3, and mas2 cells were grown at 23°C to an OD600 of 1 to 2, concentrated to an OD6w of 10, and preincubated at 37°C for 15 min. Cells were labeled by incubation with ["SI methionine (0.1 mCilml) for 5 min at 37°C. Labeling was stopped by addition of cycloheximide to 0.1 mg/ml, and incubations were continued at 37°C. Aliquots of 1 ml were removed periodically, and proteins were extracted from these samples. The dissociated samples were subjected to immunoprecipitation with antisera against mitochondrial F13 and citrate synthase. Immunoprecipitates were analyzed by polyacrylamide gel electrophoresis and fluorography. (A) F1l; (B) citrate synthase. p, Precursor; m, mature. Chase indicates minutes after addition of cycloheximide, with the 0-min point being the sample removed at the beginning of the chase, immediately following the addition of cycloheximide to stop labeling.

agarose-formaldehyde gel, blotted onto GeneScreen (Dupont, NEN Research Products, Boston, Mass.), and treated as described by Werner-Washburne et al. (35). Probes for hybridizations were made from portions of the SSAI and ACT] (actin) genes. A 1.0-kbp ClaI-PstI fragment was isolated from plasmid cen30-SSAlVB (obtained from E. Craig), and a 1.0-kbp EcoRV-BgII fragment containing a portion of the yeast actin gene was isolated from plasmid pRP37 (obtained from R. Parker). DNA fragments were labeled by random priming, using a kit from Bethesda Research Laboratories. Hybridizations were performed at 42°C. Radioactive bands were detected by autoradiography. Radioactivity was measured by scintillation counting of individual bands excised from the blot. The values determined for the SSAI bands were corrected for differences in recovery by dividing by the counts per minute corresponding to the actin message determined for the same samples. Induction was calculated as the normalized SSA1 mRNA level for each time point divided by the level prior to induction (0 min). Percent maximal induction was calculated by dividing the induction value for each time point by the value for wild-type (MAS3) cells after 15 min at 37°C. RESULTS

Isolation of the mas3 mutant. The mas3 mutant was isolated in a previous screen for yeast mutants defective in mitochondrial protein import (39). Briefly, a collection of temperature-sensitive strains was screened for the accumulation of precursor to F1l by Western immunoblotting of total cell protein after incubation of cells at 37°C. Unlike two previously described mas mutants, masi and mas2 (38), mas3 cells no longer accumulated large amounts of pre-Flp after several backcrosses to the wild-type parental strain (data not shown). However, mas3 cells still displayed tem-

FIG. 2. Evidence that the rate of posttranslational import of

pre-Fl, is greatly reduced in mas3 cells. Wild-type (wt), mas3, and mas2 cells to an

were grown

at 23°C to an

OD6. of 2 to 4, concentrated

OD600 of 10, and incubated at 37°C for 5 min. All subsequent

incubations were at 37°C. Cells were labeled with [35S]methionine in the presence of CCCP for 10 min. Next, cycloheximide, unlabeled methionine, and ,B-mercaptoethanol were added, and 1-ml aliquots were removed periodically. Protein extraction, immunoprecipitation, and analysis of precursor (p) and mature (m) forms of F1p were as described for Fig. 1.

perature-sensitive growth and a defect in mitochondrial protein import when analyzed by pulse-labeling. This latter phenotype was apparent when labeled proteins were analyzed after a 5-min pulse at the nonpermissive temperature: greater amounts of precursor of the imported mitochondrial proteins F1l and citrate synthase were present in mas3 cells than in wild-type cells (Fig. 1, lane 0). Precursors behave as two pools in mas3 cells. To examine how the rate of import was affected in mas3 cells, we examined the import of two mitochondrial proteins, F13 and citrate synthase, with isotope tracer (pulse-chase) experiments. The precursor labeling and chase patterns in mas3 cells differed from those observed in both the wild type and another import mutant, mas2. At the end of a 5-min labeling period at the nonpermissive temperature, a substantial amount of mature F1B was already present along with an elevated level of precursor in mas3 cells (Fig. 1A, lane 0); however, the labeled precursor of F1l was not converted to mature form completely even after a 10-min chase, reflecting a conversion rate slower than that seen in the mas2 mutant (Fig. 1A). Similar results were observed with another mitochondrial protein, citrate synthase (Fig. 1B). These results are consistent with the idea that the labeled precursor exists in two pools in mas3 cells, one that is rapidly imported and the other which is imported only very slowly. While localization of precursor pools in mas3 cells would be valuable for understanding the import defect, efforts to detect labeled precursors in any subcellular fractions have not been successful (data not shown). The precursors may be unstable and rapidly degraded upon cell disruption, as has been observed by others (2, 7). The rapidly imported pool may correspond to precursor imported cotranslationally, while the second pool could represent precursor imported posttranslationally. To examine the rate of posttranslational import in mas3 cells, the rate of protein import was analyzed after first accumulating a cytoplasmic pool of precursors by uncoupling mitochondria in vivo with CCCP. Following neutralization of CCCP with 13-mercaptoethanol, import of certain precursor proteins (including pre-F13) resumes (20). Cellular proteins were first labeled at 37°C in the presence of CCCP to build up labeled

2650

SMITH AND YAFFE

MOL. CELL. BIOL.

A

FIG. 3. Evidence that the mas3 mutation causes the accumulation of cells with large buds and single nuclei. Wild-type (A and B) and mas3 (C and D) cells were grown at 23°C to an OD6N of 0.1, shifted for 4 h to 37°C, and fixed with methanol-acetone as described by Thomas and Botstein (32). Samples were stained with DAPI and viewed with phase-contrast (A and C) and fluorescence (B and D) microscopy.

precursor. P-Mercaptoethanol was then added with excess, unlabeled methionine, and the fate of labeled precursor was examined periodically during a chase at 37°C. Pre-Fl behaved as a single pool in mas3 cells under these import conditions, and this precursor was imported only very slowly compared with its rate of import in wild-type and even mas2 cells (Fig. 2). A possible explanation for this slowed import rate is that mas3 cells recover from the CCCP block more slowly than do mas2 or wild-type cells. To test this possibility, mas3 cells were first treated with CCCP alone, the uncoupler was neutralized with ,B-mercaptoethanol, proteins were labeled, and the fate of newly labeled precursor was analyzed. The pattern of import under these conditions was identical to that seen in the original pulsechase experiment: two pools of precursor were again evident (data not shown), indicating that mas3 cells have no difficulty reestablishing the mitochondrial inner membrane potential upon which import depends. These experiments demonstrate that precursor behaves as two pools in mas3 cells and that the mas3 mutation affects the rate of posttranslational import. Progression through the cell cycle is defective in mas3 cells. The two previously described import mutants, masi and mas2, die at a variety of stages of the cell cycle following incubation for a number of hours at the nonpermissive temperature (37a). Surprisingly, microscopic examination of mas3 cells after a 4-h incubation at the nonpermissive

temperature revealed that a large fraction of the population was composed of large cells with single, large buds (dumbbells) (Fig. 3). Additionally, these dumbbells contained a single nucleus usually located in the mother portion of the cell (Fig. 3). After a similar 4-h incubation at 37°C, wild-type (Fig. 3) and mas2 (data not shown) cultures contained only a small fraction of dumbbells, and the nuclei in such cells had divided and were distributed in both mother and bud regions. Microscopic observations of mas3 cells after a shift for various times to 37°C revealed that the percentage of dumbbells increased to a plateau of approximately 75% of the population (Table 1). A mas3 cell population shifted after synchronization at the G1 (unbudded) stage of the cell cycle with the mating pheromone a-factor behaved similarly to unsynchronized cells but reached the plateau more rapidly (Table 1), indicating that the lesion affected cell cycle progression within the first division cycle after the temperature shift. The cell cycle defect in mas3 cells was largely reversible for up to 6 to 8 h of incubation at 37°C: upon return to the permissive temperature, the majority of cells formed colonies when plated on solid medium. The fraction of mas3 cells which did not display the dumbbell morphology after 4 to 5 h at the nonpermissive temperature might represent cells retarded at another point in the cell cycle or could reflect a continued, albeit extremely delayed, cycling at 37°C. To explore these possibilities, mas3 cells were incubated at 37°C for 5 h (to reach the

YEAST TEMPERATURE-SENSITIVE HEAT-SHOCK FACTOR

VOL . 11, 1991

TABLE 1. Effect of mas3 on cell cycle progression at 37°C

2651

A

% of cell population Expt

Ab

BC

Time (h)a

0 2.5 4 5 0 2.5 4

mas3

Wild type

U

S

L

U

S

L

30 22 16 27 83 33 38

49 52 71 52 14 49 48

21 26 13 21 3 18 14

23 21 16 15 76 28 10

48 52 20 9 23 36 14

29 27 64 76 1 35 76

Time after the cells were shifted to 37°C. b Cells were grown to early log phase at 23°C in YPD medium and shifted to 37°C. Aliquots were removed at the indicated times, and the numbers of unbudded (U), small-budded (S), and large-budded (L) cells were counted in a hemacytometer. The total number of cells counted for each time point ranged between 150 and 300. c Cells were incubated with a-factor at 23°C for 2 h, diluted 10-fold into fresh medium, shifted to 37°C for the indicated times, and counted as for experiment A. a

plateau of 75% dumbbells), and then nocodazole, a mitotic inhibitor, was added to the culture and incubation continued at 37°C. This treatment caused a further increase in the number of dumbbells (to 89%; data not shown), indicating that the non-dumbbell cells were not blocked at another stage of the cell cycle but could progress until they reached the nocodazole block. Many temperature-sensitive yeast mutants with defects both in DNA synthesis and in some later events of the cell cycle arrest as dumbbells following incubation at the nonpermissive temperature (17). To examine whether the mas3 lesion affected DNA replication or a later event, we examined the fate of mas3 dumbbells after a return to the permissive temperature in the presence of an inhibitor of DNA synthesis, hydroxyurea. Cells were incubated at 37°C for 4 h and transferred to agar slabs with or without hydroxyurea, and then individual dumbbell-shaped cells were separated by micromanipulation. The cells were observed periodically during a continued incubation at 23°C. The majority of these dumbbells (74%) completed the cell cycle in the presence of hydroxyurea and began the next division cycle (Fig. 4); 69o did so in the absence of hydroxyurea. Thus, the nuclear DNA had already duplicated in these cells during incubation at the nonpermissive temperature. A fraction of the cells did not progress beyond the dumbbell stage upon a return to the permissive temperature in the presence or absence of hydroxyurea; these cells might have difficulty recovering from the incubation at 37°C. In contrast to the results for the mas3 mutation, only 17% of cells with cdc2l, a mutation preventing progression through S phase (17), were able to continue the cell cycle at the permissive temperature in the presence of hydroxyurea; in the absence of hydroxyurea, 94% were able to do so. The cell cycle block in the majority of mas3 cells thus appears to occur following DNA replication but before the events of mitosis (e.g., nuclear elongation and division), indicating that these cells are retarded in progression through the G2 phase of the cell cycle. mas3 cells contain a single, recessive, nuclear mutation. To demonstrate that the defects in mitochondrial protein import, cell cycle progression, and temperature-sensitive growth were all caused by a single mutation in mas3 cells, the inheritance of the phenotypes was evaluated by back-

FIG. 4. Evidence that the mas3 mutation retards the cell cycle after DNA replication. mas3 cells were grown at 23°C to an OD6. of 0.15, and cells were then incubated for 4 h at 37°C. Hydroxyurea was added to 0.1 M, and cells were returned to 23°C. After a 5-h additional incubation, cells were fixed, stained with DAPI, and processed for microscopy as for Fig. 4. A typical mas3 dumbbell treated as described above is shown. (A) Phase-contrast microscopy; (B) fluorescence microscopy.

crossing the mas3 cells to a wild-type parental strain, sporulating the resulting diploid cells, and analyzing the meiotic segregation of traits by tetrad analysis. In eight tetrads examined, two of four spores exhibited the mutant phenotypes, and perfect cosegregation of the protein import defect, cell cycle block, and temperature-sensitive growth were observed (data not shown). The 2:2 segregation of the temperature-sensitive lethality was observed in 60 additional tetrads. Furthermore, a single fragment of wild-type DNA complemented all of the mutant phenotypes (described below). The heterozygous diploid cells grew normally at 37°C and displayed normal mitochondrial protein import, indicating that the mas3 lesion is recessive. Therefore, a single, recessive, nuclear mutation causes the import, cell cycle, and temperature-sensitive growth defects in mas3 cells.

2652

SMITH AND YAFFE

MOL. CELL. BIOL.

A S

R

P B

L

XR

C

1 kb

H-i

PLASMID

B

Growth 37°C

WT ---IMPORT

NORMAL CELL CYCLE

-----

YCp5O

s

x

_p

YCHSF5.3

+

+

+

pADH-HSF

+

+

+

YCHSFA3T

-

YCHSF5.3B

-

-

-

R

pADH

S

L

S

B FILL s X~19

""Ir

FIG. 5. Complementation of the mas3 mutation by the HSF gene. (A) Map of some restriction endonuclease sites in the 5.8-kbp SalI-XhoI fragment which complemented the mas3 lesion. The solid arrow above the map indicates the position of the ORF encoding HSF. The dashed arrow indicates a small ORF on the strand opposite that encoding HSF. (B) DNA fragments and constructs tested for complementation of mas3. (C) Results of complementation tests. Details of the DNA constructs and complementation tests are given in the text. Restriction sites: S, SalI; R, EcoRI; P, PvuII; B, BamHI; L, Ball; X, XhoI.

The complementing DNA encodes HSF. No previous studies have identified a relationship between mitochondrial protein import and the cell cycle. To better understand how the mas3 mutation could affect both of these essential processes, the MAS3 gene was isolated by complementation of the temperature-sensitive defect with plasmids containing yeast genomic DNA. Only one type of complementing plasmid was isolated, and it contained a 12.6-kbp insert. To prove that this fragment contained the MAS3 gene (rather than a suppressor of the mas3 mutation), a DNA fragment from the complementing clone was inserted into plasmid YIp5, the plasmid was integrated into the yeast chromosome by homologous recombination, and the site of integration was mapped. Meiotic mapping demonstrated that plasmid DNA integrated close to or at the site of the MAS3 locus, indicating that the plasmid contained the MAS3 gene (see Materials and Methods). Deletion analysis was used to further localize the complementary DNA within the 12.6-kbp DNA fragment. A 5.8-kbp XhoI-Sall fragment (Fig. 5) complemented both the protein import and cell cycle defects. Nucleotide sequence analysis of this region identified a large open reading frame (ORF) encoding a putative protein product of 93.3 kDa (data not shown). Comparison of the putative protein sequence with sequences in the GenBank data base revealed that it is identical to that of the previously identified heat-shock transcription factor HSF (29, 36). DNA sequence analysis also revealed a second, short ORF (324 bp) on the strand opposite that encoding HSF, and this ORF overlaps regions corresponding to the extreme amino-terminus and 5'-upstream sequences of HSF (Fig. 5). To examine the possibility that this second ORF encodes a component important for either cell cycle progression or mitochondrial protein import, we tested various plasmid constructs for their ability to complement the defects displayed by mas3 cells (Fig. 5). Gene constructs containing an intact small ORF but containing deletions or disruptions in the HSF-coding region failed to complement any of the mas3 phenotypes. Moreover, a construct in which the HSF-coding region was placed behind the active yeast promoter pADHI (disrupting the small ORF) complemented defects in both

mitochondrial protein import and cell cycle progression. Additionally, no evidence for expression of the small ORF was obtained. These results demonstrate that mas3 is an allele of HSF. The mas3 mutation affects the induction of the heat-shock gene SSAI at 3rc. If mas3 is a temperature-sensitive lesion in HSF, the induction of some hsps should be defective at 37°C. To test this possibility, we examined the mRNA levels of SSAI, a gene encoding a major 70-kDa hsp (35). After a shift of wild-type cells to 37°C, expression of SSAI increased to a maximum of 32-fold over its level of expression at 23°C (Fig. 6). This induction is consistent with that previously reported for SSAI (35). In contrast, induction of SSAI was only twofold (at most) in the mas3 cell at 37°C (Fig. 6). This result indicates that induction of a major hsp is defective at 37°C in cells containing the mas3 allele of HSF. mas3 cells display no defect in secretion. Previous studies (4, 7, 11, 14, 23, 34) have implicated hsps or their homologs in both mitochondrial protein import and protein secretion across the endoplasmic reticulum. To examine whether secretion was also affected by the mas3 mutation, mutant cells were synchronized at the unbudded (G1) stage with mating pheromone and then incubated at 37°C following removal of the pheromone block. During this incubation, most of the mutant cells developed large, single buds (Table 1). This generation of dumbbells at 37°C indicated that secretion was not grossly affected, since bud growth is entirely dependent on continued protein secretion (24, 34). To further analyze a possible effect of the mutation on the secretory process, the secretion of the mating pheromone a-factor was compared in wild-type cells and cells containing the mas3 or kar2-159 mutation at 37°C by pulse-labeling analysis. a-Factor secretion is particularly sensitive to a depletion of the cytoplasmic 70-kDa hsps (7). As previously reported (34), the secretory defect in kar2-159 cells caused an accumulation of prepro-a-factor (Fig. 7A, lane 4); however, no such accumulation was detected in the mas3 strain (Fig. 7A, lane 3). In these same cells, a defect in mitochondrial protein import was observed only in the mas3 mutant (Fig. 7B). The mitochondrial import defect in this experiment was not as dramatic as that displayed in Fig. 1, since


VOL. 11, 1991

1 2

A

3

2653

4

A z 0

-

pp

C-) cD 5

B x

-p imam m 25

B minutes:

0

mas3 15 30 60

wt

0 15 30 60

SSA I

ACT 1

FIG. 6. Defective heat-shock induction of SSAI in mas3 cells. Yeast strains MYY290 (wild type) and MYY385 (mas3) were grown at 23°C in SD medium. Following a shift of cells to 37°C, aliquots were removed periodically, and RNA was extracted from the cells. RNA samples were separated by electrophoresis and blotted as described in Materials and Methods. Blots were probed with 32Plabeled fragments of the SSA1 and ACT) (actin) genes. Radioactive bands were detected by autoradiography. Radioactivity was determined, and percent maximal induction was calculated as described in Materials and Methods. (A) Percent maximal induction of SSA1 in cells shifted for various times to 37°C; 10lo is the induction observed in wild-type (MAS3) cells after 15 min at 37°C. (B) Autoradiograms of blots of total cellular RNA probed with SSAI and ACT] (actin) DNA. Minutes indicates times for which cells were incubated at 37°C. wt, wild type.

the cells in the experiment shown in Fig. 7 were labeled for 30 min (necessary for detection of a-factor), while those in Fig. 1 were labeled for only 5 min. The longer labeling results in a decreased ratio of labeled precursor to labeled mature form. Additionally, the delivery of carboxypeptidase Y to the vacuole, a process dependent on the early steps of the secretory pathway (30), occurred at the same rate in the mutant and wild-type cells at the nonpermissive temperature (data not shown). These observations indicate that the mas3 mutation does not affect the secretory process. DISCUSSION We have characterized yeast cells containing the mas3 mutation. A number of observations indicate that mas3 is a lesion in the HSF gene. First, DNA sequence analysis revealed that a genomic DNA fragnent which complemented the mas3 mutant phenotypes contained the HSF gene. Second, the HSF gene mapped to the mas3 locus. Third, mutant forms of HSF failed to complement the mas3 mutation. Finally, controlled expression of HSF in the mutant cells corrected all of the aberrant traits. The mas3 mutation appears to affect primarily the post-

FIG. 7. Evidence that mas3 cells display no defect in secretion of a-factor. Cells were grown at 23°C, incubated at 37°C for 30 min, and then labeled with [35S]methionine for 30 min at 37°C. Tunicamycin was added to 10 pg/ml to one sample (lane 2) 15 min prior to addition of the radioactive label. Proteins were extracted following the addition of 3 A6. units of unlabeled, MATa, wild-type cells. a-Factor and F1lo were isolated from the samples by successive immunoprecipitations. Immunoprecipitates were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and fluorography. (A) Prepro-a-factor; (B) F1o. Lanes: 1, wild type (AH216); 2, wild type plus tunicamycin; 3, mas3 (MYY243); 4, kar2-159 (MS177). pp, Prepro-a-factor; **, unglycosylated pro-afactor; p, pre-Fl; m, mature Fl. The slower migration of unglycosylated pro-a-factor than prepro-a-factor was noted previously (22).

translational import of mitochondrial proteins. This is suggested by the two pools of precursor proteins observed in pulse-chase experiments (Fig. 1) and by the conversion of these two populations into a single pool (which is imported only very slowly) under conditions in which the import of labeled precursors is exclusively posttranslational (Fig. 2). In the wild-type cell, the import of mitochondrial proteins is likely to proceed by a mixture of cotranslational and posttranslational routes, with a balance between these two modes dictated by differences in the rates of protein synthesis and protein import (19). For example, the amino terminus of a large protein or one which is synthesized slowly might engage the import apparatus before the completion of translation, allowing the nascent protein to be imported cotranslationally. Other proteins might be released from the ribosome before they initiate the import process and then follow an entirely posttranslational mode. Components specific for this posttranslational route might include proteins that unfold precursors or prevent their acquisition of a tight tertiary structure and probably act early in the import pathway (15). Such proteins may be deficient in mas3 cells during incubation at the nonpermissive temperature. The link between a defect in HSF and aberrant protein import in the mas3 cells is likely to involve an hsp that is required for posttranslational import and depends on HSF for its induction at 37°C. Previously, specific hsps have been identified as essential components for mitochondrial protein import and for protein secretion in yeast. These hsps include the cytoplasmic HSP70s, required for both mitochondrial protein import and secretion (7, 14), the KAR2 protein required for efficient transfer of proteins into the endoplasmic reticulum (34), and two mitochondrial matrix proteins, Ssclp (5, 11) and HSP60 (3, 18), required for the import and assembly, respectively, of mitochondrial polypeptides. These hsps are essential proteins, required by cells to facilitate protein traffic even under nonstressed conditions. It is not known whether higher levels or additional hsps are

2654

SMITH AND YAFFE

required for protein traffic under heat-shock conditions; however, the phenotype of the mas3 allele suggests that increased levels of one or several hsps are required for mitochondrial protein import at 37°C. Supplying such a requirement may be one of the basic cellular functions of hsp induction in response to heat-shock conditions. An additional requirement of hsps at elevated temperatures apparently is not needed for the secretory process. Previous studies (7, 14) have identified the cytoplasmic HSP70 proteins as components important for an early step in mitochondrial protein import such as precursor unfolding or maintaining the import competence of precursor proteins, and the HSP70s might be required for posttranslational import. However, secretion, a process also dependent on cytoplasmic HSP70s (4, 7) is not affected by the mas3 lesion. Additionally, the import defect in mas3 cells is not corrected by increased expression of Ssalp, one of the cytoplasmic HSP70s (27a), suggesting that the aberrant import in the mutant is not due to inadequate amounts of these proteins. It is possible that mitochondrial protein import requires much higher levels of HSP70s (or higher levels of a specific HSP70 isozyme) at 37°C than does protein secretion and that adequate, high-level expression of one or several cytoplasmic HSP70s would correct the import defect. Cells containing the mas3 mutation are largely retarded in progression through the G2 phase of the cell cycle. This finding is the first evidence of a link between HSF and cell cycle progression. The blockage in mas3 cells presumably reflects a deficiency in one or several components which are required during this portion of the cell cycle and whose synthesis is mediated by HSF. A shortage of these components might greatly extend the time required for completion of the G2 stage (rather than causing an absolute blockage in the division cycle) and lead to the observed increase in dumbbells in the population. This continued but delayed cycling could explain the approximately 25% of a mas3 population found at other stages of the cell cycle even after many hours at 37°C. The identities and specific functions of hsps required in G2 have yet to be determined, but the increased synthesis of these proteins during heat shock might allow cells to complete the division cycle under stress conditions. Such a function would be consistent with a proposed role of hsps (16) in enabling stressed cells to reach a Go state, where they would have a greater resistance to additional stress. Yeast HSF is normally activated at elevated temperatures, and the activated factor greatly stimulates the transcription of genes encoding hsps (13, 28). In mas3 cells at 37°C such increased transcription is absent for SSAI, a gene encoding a major 70-kDa hsp (Fig. 6). Induction at elevated temperatures is also likely to be defective for other genes encoding hsps. The aberrant gene expression could result from deficient activation of the mas3 allele of HSF or a defect in the factor's activity or stability at 37°C. A defect in the induction of various hsps is likely to account for the mutant properties of the mas3 cell at the nonpermissive temperature. HSF plays an essential role in the constitutive expression of many genes under nonstressed conditions (29, 36), and some of this activity apparently is preserved in the mas3 mutant, which displays normal cellular functions at 23°C. The mas3 mutation prevents the hsp induction that is mediated by HSF at elevated temperatures. This leads to defects in mitochondrial protein import and cell cycle progression at 37°C. HSF thus mediates the stress response of two separate and essential cellular processes and may also function to coordinate these processes in nonstressed cells. Further

MOL. CELL. BIOL.

study should lead to the identification of specific hsps required for these different cellular functions and to a better understanding of the mechanisms underlying the cell's response to environmental stress. ACKNOWLEDGMENTS We are grateful to Roy Parker for early characterizations of the mas3 mutant cells and for critical reading of the manuscript. We thank Michelle Apperson for expert technical assistance. We are grateful to Al Lewin and Randy Schekman for antisera, Roy Parker and Betty Craig for plasmids, and Lee Hartwell, Barbara Garvick, and Mark Rose for yeast strains used as controls for certain experiments. We thank Leslie Stewart for critical reading of the manuscript. The mas3 mutant was originally isolated by M.P.Y. in the laboratory of Gottfried Schatz, and we appreciate his support in early aspects of the study. This work was supported by grant MV-419 from the American Cancer Society and by a Searle Scholarship (to M.P.Y.) from the Searle Scholars Program of the Chicago Community Trust. REFERENCES 1. Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. A. Smith, J. G. Seidman, and K. Struhl. 1987. Current protocols in molecular biology. John Wiley & Sons, Inc., New York. 2. Baker, K. P., A. Schaniel, D. Vestweber, and G. Schatz. 1990. A yeast mitochondrial outer membrane protein essential for protein import and cell viability. Nature (London) 348:605-609. 3. Cheng, M. Y., F.-U. Hartl, R. A. Martin, R. A. Poilock, F. Kalousek, W. Neupert, E. M. Hallberg, R. L. Hallberg, and A. L. Horwich. 1989. Mitochondrial heat-shock protein hsp60 is essential for assembly of proteins imported into yeast mitochondria. Nature (London) 337:620-625. 4. Chirico, W. J., M. G. Waters, and G. Blobel. 1988. 70K heat shock proteins stimulate protein translocation into microsomes. Nature (London) 332:805-810. 5. Craig, E. A., J. Kramer, J. Shilling, M. Werner-Washburne, S. Holmes, J. Kosic-Smithers, and C. M. Nicolet. 1989. SSCI, an essential member of the yeast HSP70 multigene family, encodes

a mitochondrial protein. Mol. Cell. Biol. 9:3000-3008. 6. Daum, G., P. C. Bohni, and G. Schatz. 1982. Import of proteins into mitochondria. Cytochrome b2 and cytochrome c peroxidase are located in the intermembrane space of yeast mitochondria. J. Biol. Chem. 257:13028-13033. 7. Deshaies, R. J., B. D. Koch, M. Werner-Washburne, E. A. Craig, and R. Schekman. 1988. 7OkD stress protein homologues facilitate translocation of secretory and mitochondrial precursor polypeptides. Nature (London) 332:800-805. 8. Hurt, E. C., U. Muller, and G. Schatz. 1985. The first twelve amino acids of a yeast mitochondrial outer membrane protein can direct a nuclear-encoded cytochrome oxidase subunit to the mitochondrial inner membrane. EMBO J. 4:3509-3518. 9. Ito, H., Y. Fukuda, K. Murata, and A. Kinura. 1983. Transformation of intact yeast cells treated with alkali cations. J. Bacteriol. 153:163-168. 10. Jensen, R. E., and M. P. Yaffe. 1988. Import of proteins into yeast mitochondria: the nuclear MAS2 gene encodes a component of the processing protease that is homologous to the MAS1-encoded subunit. EMBO J. 7:3863-3871. 11. Kang, P.-J., J. Ostermann, J. Shilling, W. Neupert, E. A. Craig, and N. Pfanner. 1990. Requirement for hsp70 in the mitochondrial matrix for translocation and folding of precursor proteins. Nature (London) 348:137-142. 12. Lindquist, S. 1986. The heat-shock response. Annu. Rev. Biochem. 55:1151-1191. 13. Lindquist, S., and E. A. Craig. 1988. The heat-shock proteins. Annu. Rev. Genet. 22:631-677. 14. Murakami, H., D. Pain, and G. Blobel. 1988. 70-kD heat shock-related protein is one of at least two distinct cytosolic factors stimulating protein import into mitochondria. J. Cell Biol. 107:2051-2057. 15. Ohta, S., and G. Schatz. 1984. A purified precursor polypeptide requires a cytosolic protein fraction for import into mitochon-

VOL.

11, 1991


dria. EMBO J. 3:651-657. 16. Plesset, J., J. R. Ludwig, B. S. Cox, and C. S. McLaughlin. 1987. Effect of cell cycle position on thermotolerance in Saccharomyces cerevisiae. J. Bacteriol. 169:779-784. 17. Pringle, J. R., and L. H. Hartwell. 1981. The Saccharomyces cerevisiae cell cycle, p. 97-142. In J. N. Strathern, E. W. Jones, and J. R. Broach (ed.), Molecular biology of the yeast Saccharomyces cerevisiae: life cycle and inheritance. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 18. Reading, D. S., R. L. Hallberg, and A. M. Myers. 1989. Characterization of the yeast HSP60 gene coding for a mitochondrial assembly factor. Nature (London) 337:655-659. 19. Reid, G. A., and G. Schatz. 1982. Import of proteins into mitochondria. Extramitochondrial pools and post-translational import of mitochondrial protein precursors in vivo. J. Biol. Chem. 257:13062-13067. 20. Reid, G. A., and G. Schatz. 1982. Import of proteins into mitochondria. Yeast cells grown in the presence of carbonyl cyanide m-chlorophenylhydrazone accumulate massive amounts of some mitochondrial precursor polypeptides. J. Biol. Chem. 257:13056-13061. 21. Rose, M. D., P. Novick, J. H. Thomas, D. Botstein, and G. R. Fink. 1987. A Saccharomyces cerevisiae genomic plasmid bank based on a centromere-containing shuttle vector. Gene 60:237243. 22. Rothblatt, J. A., R. J. Deshaies, S. L. Sanders, G. Daum, and R. Schekman. 1989. Multiple genes are required for proper insertion of secretory proteins into the endoplasmic reticulum in yeast. J. Cell Biol. 109:2641-2652. 23. Sadler, I., C. T. Kuribara, J. Rothblatt, J. Way, and P. Silver. 1898. A yeast gene important for protein assembly into the endoplasmic reticulum and the nucleus has homology to DnaJ, an Escherichia coli heat shock protein. J. Cell Biol. 109:26652675. 24. Schekman, R. W., and P. Novick. 1982. The secretory process and yeast cell surface assembly, p. 361-398. In J. N. Strathern, E. W. Jones, and J. R. Broach (ed.), Molecular biology of the yeast Saccharomyces cerevisiae: metabolism and gene expression. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 25. Schlesinger, M. J. 1990. Heat shock proteins. J. Biol. Chem. 265:12111-12114. 26. Sherman, F., G. R. Fink, and J. B. Hicks. 1979. Methods in

2655

yeast genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 27. Sikorski, R. S., and P. Hieter. 1989. A system of shuttle vectors and yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae. Genetics 122:19-27. 27a.Smith, B., and M. Yaffe. Unpublished data. 28. Sorger, P. K., M. J. Lewis, and H. R. B. Pelham. 1987. Heat shock factor is regulated differently in yeast and HeLa cells. Nature (London) 329:81-84. 29. Sorger, P. K., and H. R. B. Pelham. 1988. Yeast heat shock factor is an essential DNA-binding protein that exhibits temperature-dependent phosphorylation. Cell 54:855-864. 30. Stevens, T. H., B. Esmon, and R. Schekman. 1982. Early stages in the yeast secretory pathway are required for the transport of carboxypeptidase Y to the vacuole. Cell 30:439-448. 31. Struhl, K., D. T. Sfinchcomb, S. Scherer, and R. W. Davis. 1979. High frequency transformation of yeast: autonomous replication of hybrid DNA molecules. Proc. Natl. Acad. Sci. USA 76:1035-1039. 32. Thomas, J. H., anid D. Botstein. 1986. A gene required for the separation of chromosomes on the spindle apparatus in yeast. Cell 44:65-76. 33. Van Loon, A. P. G. M., E. Van Eik, and L. A. Grivell. 1983. Biosynthesis of the ubiquinol-cytochrome c reductase complex in yeast. Discoordinate synthesis of the 11-kd subunit in response to increased gene copy number. EMBO J. 2:1765-1770. 34. Vogel, J. P., L. M. Misra, and M. D. Rose. 1990. Loss of BiP/GRP78 function blocks translocation of secretory proteins in yeast. J. Cell Biol. 110:1885-1895. 35. Werner-Washburne, M., J. Becker, J. Kosic-Smithers, and E. A. Craig. 1989. Yeast Hsp7O RNA levels vary in response to the physiological status of the cell. J. Bacteriol. 171:2680-2688. 36. Wiederrecht, G., D. Seto, and C. S. Parker. 1988. Isolation of the gene encoding the S. cerevisiae heat shock transcription factor. Cell 54:841-853. 37. Witte, C., R. E. Jensen, M. P. Yaffe, and G. Schatz. 1988. MASI, a gene essential for yeast mitochondrial assembly, encodes a subunit of a mitochondrial processing protease. EMBO J. 7:1439-1447. 38. Yaffe, M. Unpublished data. 39. Yaffe, M. P., and G. Schatz. 1984. Two nuclear mutations that block mitochondrial protein import in yeast. Proc. Natl. Acad. Sci. USA 81:4819-4823.