YEAST
VOL.
7: 539-546 (1991)
The Determinants of Heat-Shock Element-Directed lac2 Expression in Saccharomyces cerevisiae NIALL KIRK*? AND PETER W. PIPER*$ Departments of * Biochemistryand Molecular Biology and ?Chemical and BiochemicalEngineering, University College London, London WClE 6BT. U.K. Received 17 September 1990; revised 5 February 1991
Heat-shock induction of heat-shock protein genes is due to a specific promoter element (the heat-shock element, HSE). This study used lac2 under HSE control (HSE-lacZ) to characterize HSE activity in Saccharomyces cerevisiae cells of different physiological states and differing genetic backgrounds. In batch fermentations HSE-IacZinduction by heat shock was maximal in exponential growth, and showed marked decline with the approach to stationary phase. Expression in the absence of heat shock was unaffected by growth phase, indicating that the growth-dependent expression of many yeast heat-shock genes uses promoter elements in addition to the HSE. Heat-induced expression was strongly influenced by the temperature at which cultures were grown. While basal, uninduced expression was constant during growth at different temperatures to 3 0 T , induction by transfer to 39°C was reduced by increases in growth temperature as low as 18-24°C. Maximal HSE-lacZ induction (30- to 50-fold) was in cultures grown at low temperatures (18-24"C), then heat shocked at 39°C. Ethanol was a poor inducer. Mutations having little effect on HSE-ZacZ expression included a respiratory petite; ubi4 (which inactivates the polyubiquitin gene); also ubc4 and ubc5 (which each inactivate one of the ubiquitin ligases involved in degradation of aberrant protein). pep4-3 increased both basal and induced P-galactosidaseabout two-fold, probably because of slower turnover of this enzyme inpep4-3 strains. KEY WORDS - Saccharomyces cerevisiae; heat-shock
element; IacZ induction; growth state; ubi4, ubcl, ubc5, pep4-3
mutants.
INTRODUCTION Temperature-regulated yeast expression systems achieve gene induction through either a temperature downshift (Brake et al., 1988; Kramer et al., 1984; Sledewski et al., 1988) or a temperature upshift (DaSilva and Bailey, 1989; also the heat-shock element (HSE)-directed expression investigated in this study). Important considerations need to be met before temperature shifts can be used to control the expression of a heterologous gene. Firstly, since Saccharomyces cerevisiae grows more slowly with reduced biomass yield at temperatures above 34-36°C it is usually desirable to grow cultures at temperatures slightly below this value. Secondly, while systems using temperature downshift for induction may be straightforward to operate on a small scale, they are not as attractive for larger-scale processes as those systems that use temperature upshift. This is because heat addition to large fermenters is easier than heat extraction. $Author to whom correspondence should be addressed. 0749-503X/9 1/060539-08 $05.00 0 1991 by John Wiley & Sons Ltd
Induction by temperature upshift is readily achieved using the promoter element (HSE) which directs the activation of heat-shock genes at elevated temperatures. Sequences placed under HSE control then become activated together with other heat-shock genes as the heat-shock response is induced. The HSE is the promoter binding site for heat-shock factor (HSF), a protein which partly acquires its activity as a transcriptional transactivator by heat-shock-induced phosphorylation (Sorger and Pelham, 1987,1988; Sorger, 1990). HSE-directed expression systems allow yeast to be grown under conditions enabling efficient biomass accumulation. Also unlike other yeast temperatureregulated expression systems described to date, they can operate in a variety of genetic backgrounds. They are not, however, without their potential problems, notably: (i) The heat-shock response is transient at sublethal temperatures, being switched off within
540
N. KIRK AND P. W. PIPER
Table 1 . Yeast strains used in this study Strain CT3b* CT3c* SUB60t SUB62t ubc4t ubc5t
Genotype leu2 trpl leu2 trpl leu2 trpl leu2 trpl leu2 trpl leu2 trpl
ura3 pep4-3 ura3 ura3 his3 lys2 ubi4::LEUZ ura3 his3 lys2 ura3 his3 lys2 ubc4::HIS3 ura3 his3 lys2 ubcS::LEU2
Source J. Wingfield J. Wingfield
D. Finley D. Finley S. Jentsch S. Jentsch
*CT3b and CT3c are products of a single meiotic tetrad (Wingfield and Dickinson. 1990). tCongenic' with' diploid DF5 (Finley et al., 1987; Seufert and Jentsch, 1990).
30-120 min of induction (Lindquist and Craig, 1988). (ii) Induction, as measured with an Escherichia coli lacZ reporter gene, is usually not greater than 30- to 50-fold. This is due both to the limited expression period and measurable HSF activity in the absence of heat shock (Wei et al., 1986; Sorger and Pelham, 1987; Sorger, 1990; also results described below). (iii) Secretion may not be very efficient at those temperatures optimal for HSF activation. (iv) The temperatures used to heat-stress yeast may detrimentally affect certain heterologous products. Since heat shock induces certain genes of the ubiquitination cycle for protein turnover (Finley et al., 1987; Seufert and Jentsch, 1990) it may stimulate the turnover of aberrant (but possibly not native) protein. Despite such possible disadvantages HSEdirected expression might find applications where a limited expression period will suffice, and where authentic rather than aberrant product can accumulate intracellulary at 37-39°C. We therefore extended previous studies of expression of lacZ under HSE control (HSE-lacZ), so as to determine how this expression is influenced by: (i) growth state, and growth temperature prior to heat shock; (ii) magnitude of temperature upshift, and final temperature of heat shock; (iii) other inducers of heat-shock genes; (iv) respiratory-deficiency (rhocells); (v) inactivation of genes of the ubiquitination system for turnover of aberrant protein; and (vi) a mutation (pep4-3)which substantially reduces vacuolar protease activities.
MATERIALS AND METHODS Yeast strains and yeast transformation The S . cerevisiae strains employed are listed in Table 1. Each was transformed to uracil prototrophy using plasmid pHSE2 (Sorger and Pelham, 1987) and the transformation procedure of Hinnen et al. (1978). These transformants are denoted by a -pHSE2 suffix in conjunction with strain designations (e.g. CT3b-pHSE2). pHSE2 contains a tandem copy of the HSE sequence known to selectively bind yeast heat-shock transcription factor (Sorger and Pelham, 1987). This HSE is inserted within the cycI promoter of the cytochrome cl-lacZ fusion gene of pHSE2, where it replaces the normal upstream activating sequences of this promoter. Spontaneous petites were selected from SUB62pHSE2 as small colonies on standard mineral salts glucose-defined (SD) medium, rho- colonies which both lacked mitochondria1 cytochromes and displayed no capacity for gluconeogenic growth. Culture media Transformants were grown aerobically in vigorously-shaken flask cultures, either on SD medium (Sherman et al., 1983) which has 2 mg I - ' meso-inositol; or on SD with an increased (5 mg 1 ') meso-inositol supplement (data in Figure 1 only). Appropriate amino acid supplements were also present, omitting uracil to ensure pHSE2 maintenance. To heat-shock cells, flask cultures were placed in water baths at the appropriate temperature, temperature equilibration occurring over less than 3 min. In preliminary experiments HSE-lacZ expression 1 h after heat shock to 39°C was found to -
DETERMINANTS OF HEAT-SHOCK ELEMENT -DlRECTED L A C 2 EXPRESSION IN SACCHAROMYCES CEREVISIAE
54 1
RESULTS Basal HSE-lacZ expression is not growth phase-dependent
In the absence of heat shock, SD medium batch cultures of pHSE2 transformants displayed a low, constant HSE-lacZ expression. This uninduced f3galactosidase did not rise with transitions to slower growth, either at the start of transition phase or during entry to stationary phase (Figure I ) . Similar low basal activities, unaffected by growth phase, were found during batch fermentations on complex (YEPD) medium (not shown). Figure I . Batch fermentation of SUB62-pHSE2 in 23°C YNBG medium (see Methods) with either 2 mg 1 I ( A ) or a 5 mg I I (2.)inositol supplement. 1, I1 and I11 approximate to the exponential, transition and stationary phases of growth respectively. P-galactosidase activity both before (0) and I h after heat shock 10 39°C ( 0 )was identical for both medium inositol concentrations. This figure shows the progressive decline of the heat inducibility of the HSE-lucZgene during transition phase and the entry to stationary phase.
be the same irrespective of whether cultures had taken 30 s or 5 min to attain this temperature.
The optimal temperature for HSE-lacZ induction
With SUB62-pHSE2 initially in exponential growth at 23°C only a narrow range of heat-shock temperatures gave high f3-galactosidase induction (Figure 2). Induction was maximal at 39"C, 3 7 T giving about half the induction of 39°C. Table 2 also shows that the final temperature of heat stress is a major determinant of induction level.
- 507 L
E
C
5 LOa
/IGalac - t osidase ussays
f3-Galactosidase was assayed as in Sorger and Pelham (1987), either in the absence of heat shock or at the stated intervals after heat shock. Each data point is the mean of two experimental determinations (S.D. 20%).
U
m
-g 33-3 0
22910-
/'
Plasmid copy level determinations
Total yeast DNA was isolated from untransformed and pHSE2-transformed cultures as in Piper er al. (1988). These DNA samples were slotblotted in duplicate, then hybridized (as in Piper et al., 1988) to either a URA3 gene probe (the 0.8 kb EcoI-Smul fragment of pHSE2) or a PGK gene probe (the 1.95 kb HindIII fragment of pMA1; Piper et al., 1988). The increased URA3 hybridization signal in DNA samples from pHSE2 transformants indicated fairly constant copy number (7 & 3) for this plasmid in all cultures used for lac2 expression measurements. Observed differences in lacZ expression could not therefore be attributed to changes in plasmid copy level.
Figure 2. The importance of final temperature of heat shock on induction. Using SUB62-pHSE2 in exponential growth at 23°C. p-galactosidase levels were measured I h after shift to the temperatures indicated.
HSE-lacZ induction is highly growth phasedependent and maximal in exponential growth Heat-induced P-galactosidase was maximal with cells that were in exponential growth at the time of heat shock (Figure 1). Marked decreases in heat-shock induction accompanied the approach to stationary phase, although a slight induction (about
542
N. KIRK AND P. W. PIPER
Table 2. Effects of the magnitude of temperature upshift and final temperature on heat-shock induction levels
Heat shock By 13°C upshift: 2&33"C 22-35°C 24-37°C 2639°C By 15°C upshift: 18-33°C 2&35"C 22-37°C 2639°C By 17°C upshift: 18-35°C 2&37"C 22-39°C By 19°C upshift: 18-37°C 2&39"C
Units P-galactosidase in heat-shocked cells*
Induction ratio?
12 20 28.2 68.8
4.4 7.4 10.4 25.5
13.3 23.5 38.0 80.3
4.9 8.7 14.1 29.7
25.8 37.5 83.8
10.5 13.9 31.0
28.4 92.6
27.8 34.3
*Measured 60 min after temperature upshift. tIn the SUB60-HSE2 cultures used for these determinations basal p-galactosidase activity at 18,20,22,24 and 26°C prior to heat shock was 2.7 units.
two-fold) could still be obtained in stationary phase (Figure 1). These stationary cells could regain maximal capacity for heat induction within 2 h if diluted 20-fold in fresh 23°C SD medium (not shown). Both basal and heat-induced HSE-lac2 expression were unaffected by increasing the inositol supplement of SD medium (Figure 1) even though this allows the attainment of slightly higher cell densities (Henry et al., 1977). Growth temperature has relatively minor efects on basal HSE-lacZ expression levels, but strongly influences induction by heat shock Growth temperature was found to be extremely important in determining levels of induction by heat shock. Uninduced P-galactosidase was constant during growth at temperatures from 18°C to 30"C, only increasing in cultures growing above 30°C (Figure 3). In contrast, HSE-IacZinduction by heat shock showed a dramatic decline with increasing growth temperatures from 18°C to 30°C (Table 2 and Figure 3), being greatest for those cells grown at 18-20°C prior to heat shock at 39°C. S. cerevisiae is often grown at 28-33"C, the temperatures that enable optimal growth rates and
!OOr
Growth temperature
Figure 3. Variation in heat inducibility of the HSE-lucZ gene in SUB62-pHSE2 growing at different temperatures. Cells were grown at the indicated temperatures, their P-galactosidase levels being measured both before (0)and 1 h after ( 0 )heat shock to 39°C.
substrate conversions. Although SUB62-pHSE2 growing at 30°C gave only ten-fold HSE-IacZ induction after shift to 39°C (Figure 3), these same cells could display > 30-fold induction if placed at 23°C for 2 h prior to heat shock at 39°C (not shown). This indicates that it may be necessary to cool cells for a period prior to heat shock in order to obtain maximal HSE induction with yeast cultured
DETERMINANTS OF HEAT-SHOCK ELEMENT -DIRECTED L A C Z EXPRESSION IN SACCHAROMYCES CEREVISIAE
Table 3.
HSE-kuc% induction by various inducers of
stress responses
Agent Hcal shock to 39°C Ethanol 8 % (v'v)
Dimcthylsulphoxide 10% ( w / \ ) Canavaninc I mg:ml NaCl 1.7 M Paromomycin I m g m l 4-nitroquinoline- I -oxide I pg/ml
0-Galactosidasc induction > 30XA 2X" Nilb Nilb
Nilb 2X' 2Xh
Induction wasmeasuredeither I h(a), 1 4 h(b)orS h(c)afterthe addition of each agent to SUB62-pHSE2 cells in exponential growth at 23 C. the fraction of the cells surviving at these times k i n g > 9X'!AI in all cdscs.
at those temperatures that allow optimal growth rates. HSE-lacZ is poorly induced b y many inducers of heat-shock genes A striking feature of the heat-shock response is its multiplicity of environmental and chemical stress agents (Ananthan et al.. 1985; Lindquist and Craig, 1988). We tested a number of known inducers of stress-responsive genes in S . cerevisiae to see if they gave appreciable HSE-lucZ induction (Table 3). Without exception all gaveonlyvery slight induction (ethanol. paromomycin and 4-nitroquinoline- 1oxide) or no detectable change in expression (dimethylsulphoxide, canavanine and high osmolarity).
Eflects qfyeast genetic background on HSE-lac2 expression
To determine how respiratory deficiency affects HSE-IucZ expression, spontaneous rho- mutants were isolated from SUB62-pHSE2. Both the basal and the heat-shock induced P-galactosidase of these r/io cells were approximately two-fold higher than in the rho ' parents (not shown). rho did not therefore greatly enhance intracellular accumulation of P-gdlactosidase. The effects of rho- may be more dramatic on the synthesis of secreted products since it increases human lysozyme secretion by yeast up to ten-fold, especially at stationary phase (Kaisho et ul., 1989). Heat shock activates certain genes of the ubiquitination system for intracellular protein turnover.
543
These include UBZ4 (encoding polyubiquitin; Finley et ul., 1987); also UBC4 and UBCS (encoding the ubiquitin ligases which join ubiquitin to substrates destined for proteolysis; Seufert and Jentsch, 1990). Although P-galactosidase normally has a long halflife ( > 20 h) in yeast (Varshavsky et al., 1988), when made at heat-shock temperatures it could conceivably turn over more rapidly due to simultaneous activation of the ubiquitination pathway. ubi4 expression hosts might therefore give higher andl or more prolonged synthesis. We compared pgalactosidase induction in SUB60-pHSE2 and SUB62-pHSE2, two transformants that are isogenic but for a LEU2 gene disruption of the UBM locus in SUB60-pHSE2 (Finley et al., 1987). The uhi4 strain did not give appreciably higher or more prolonged HSE-lacZ expression, its P-galactosidase synthesis after heat shock being only slightly greater than in UBI4' cells (Figure 4). Should P-galactosidasc synthesized at heat-shock temperatures be subject to appreciable degradation by the ubiquitination pathway, reductions in ubiquitin ligase activity might also increase its stability. Two such ligases, UBC4 and UBC5, are involved in degradation of abnormal proteins (Seufert and Jentsch, 1990). Preventing the expression of either UBC4 or UBCS might therefore conceivably increase heat-induced HSE-IacZexpression. This was not found, the use of ubc5 (Figure 4) or uhc4 (not shown) expression hosts having no major effect on heat-shock induction of p-galactosidase. This is consistent with earlier findings that the turnover of long-lived and short-lived normal proteins (unlike proteolysis of abnormal canavanyl-peptides) is barely affected by ubc4 or ubc5 as single mutations (Seufert and Jentsch. 1990). The highly-selective protein turnover catalysed by the ubiquitination system (Ciechanover and Schwartz, 1989) contrasts with the non-specific proteolysis catalysed by the vacuolar proteases. pep4.3, a defect in the P R A l gene, reduces the activity of several vacuolar hydrolases by over 90% (Jones et al., 1982). pep4.3 expression strains are therefore in widespread use to minimize the breakdown of products made in yeast during their recovery. Basal and heat-induced HSE-fucZ expression were both approximately two-fold higher in a pep4.3 strain as compared to a wild-type strain of the same parentage (Figure 4). These effects appeared not to be due to any higher instability of P-galactosidase in crude cell extracts from wild-type cells (the rates of inactivation of this enzyme in crude extracts from 6T3b-pHSE2 and 6T3c-
544
N. KIRK AND P. W. PIPER 150
I ~
IOC
0 v)
0 0 c
-00 a v)
c
5
5c
i
I
I
60
120
I
I
180
60
I
120
1
180
Minutes after 23-39OC shift
Figure 4. Heat-shock induction of P-galactosidase in SUB62-pHSE2 (m), SUB60pHSE2 (A),ubc5-pHSE2( A ) , 6T3b-pHSE2( 0 )and 6T3c-pHSE2(0). All cultures were in exponential growth at 23°C prior to heat shock to 39°C for the indicated times.
pHSE2 were similar irrespective of the presence or absence of protease inhibitors; data not shown). P-galactosidase increases due to pep4-3 have also been found by Wingfield and Dickinson (1990) for lucZ expressed from the complete CYCl promoter. They have also concluded this is due to pep4-3 increasing P-galactosidase during growth, rather than any artefact of the procedure for making cell extracts prior to enzyme assay. DISCUSSION The synthetic heat-shock promoter used here has previously been employed to measure the activity of HSF in S. cerevisiue (Sorger and Pelham, 1987). Variants of this promoter bearing point mutations in the HSE that interfere with HSF binding cause p-galactosidase activity to be low under all conditions (Sorger and Pelham, 1987).Therefore the pgalactosidase levels in pHSE2-transformants reflect the activity of HSF in enhancing transcription. Here we have extended these studies in order to identify the extent to which successful operation of HSEbased induction systems may be dependent on the physiological state, the culture conditions and the genetic background of S. cerevisiue cells. Our results demonstrated that HSE-lucZ induction is highly dependent on growth state and temperature of growth
(Figures 1 and 2; Table 2); also the final temperature of heat stress (Figure 2 and Table 2). Table 2 clearly shows that the magnitude of temperature upshift is not the sole determinant of induction level. Several S . cerevisiue heat-shock genes are under growth phase control, being activated in the absence of heat shock as batch fermentations approach stationary phase. Such genes include SSA3 (Boorstein and Craig, 1990), H S P l 2 (Praekelt and Meacock, 1990) and UBZ4 (Finley et ul., 1987; Tanaka et ul., 1988). However, there is no increase in basal HSE-ZucZ expression at the approach to stationary phase (Figure 1). The activation of many natural heat-shock promoters towards the end of batch fermentation would therefore appear to require promotor-controlling elements in addition to HSE and HSF. At least one of these promoters (SSA3) has a cyclic AMP-responsive element (Boorstein and Craig, 1990). Certain of the agents in Table 3 induce a subset of the heatshock proteins of S. cerevisiue (e.g. 8% ethanol, which strongly induces Hsp70, Hsp90 and Hsp104, unpublished observations; paromomycin, which causes translational errors, Grant et ul., 1989; and 4-nitroquinoline- 1-oxide, a DNA-damaging agent, McClanahan and McEntee, 1986). The fact that none of these agents gave strong HSE-
DETERMINANTS OF HEAT-SHOCK ELEMENT -DIRECTED LACZ EXPRESSION IN SACCHAROMYCES CEREVISIAE
lac2 induction suggests they are n o t s t r o n g HSF activators. T h e activation o f UBi4 by heat shock m a y help overcome a n y shortage o f free ubiquitin t h a t arises from t h e need f o r increased turnover o f thermallydenatured protein by t h e ubiquitination system ( M u n r o a n d Pelham, 1985). If a heterologous protein induced by heat shock is degraded more rapidly because of this activation, prevention o f UBI4 expression m a y lead t o higher expression levels. Alternatively, if shortage of free, unconjugated ubiquitin is t h e induction signal f o r the heatshock response ( M u n r o a n d Pelham, 1985), a ubi4 genetic background might give m o r e sustained heatshock-induced transcription. Our results (Figure 4) s h o w t h a t b o t h the stability o f heat-induced 0galactosidase a n d the switch-off o f HSF-directed HSE-lmZ expression a t 39°C are largely independ e n t o f UBI4 expression and its influence on ubiquitin levels. ACKNOWLEDGEMENTS W e thank Peter Sorger f o r pHSE2, and J o n a t h a n Wingfield, Daniel Finley a n d Stefan Jentsch f o r strains. T h i s w o r k w a s supported by a SERC g r a n t (GRiF;72932) a n d CASE studentship (N.K.) held in conjunction with ICI Pharmaceuticals.
REFERENCES Ananthan, J., Goldberg, A. L. and Voellmy, R. (1985). Abnormal proteins serve as eukaryotic stress signals and trigger the activation of heat shock genes. Science 232,522-524. Boorstein, W. R.and Craig, E. A. (1990). Transcriptional regulation or SSA3, an HSP70 gene from Saccharomyces cerevisiae. Mol. Cell. Biol. 10,3262-.3269. Brake, A. J.. Merryweather, J. P., Coit, D. G., Heberlain, V. A,. Maziarz, F. R., Mullenbach, G. T., Urdea, M. S., Valenzuela. P. and Barr, P. J. (1984) a-factor-directed synthesis and secretion of mature foreign proteins in Sacchnromyces cerevisiae. Proc. Natl. Acud. Sci. USA 81,46424646. Ciechanover, A. and Schwartz, A. L. (1989). How arc substrates recognised by the ubiquitin-mediated proteolytic system? Trends Biochem. Sci. 14,483487. DaSilva, N . A. and Bailey, J. E. (1989). Construction and characterisation of a temperature-sensitive expression system in east. Biotech. Prog. 5, 18-26. Finley, D., dzkaynak, E. and Varshavsky, A. (1987). The ycast polyubiquitin gene is essential for resistance to high temperatures, starvation, and other stresses. Cell 48, 1035--1046. Grant, C. M., Firoozan, M. and Tuite, M. F. (1989). Mistranslation induces the heat shock response in the yeast Succharomyces cerevisiae. Mol. Microbiol. 3 , 2 1 5-220.
545
Henry, S. A., Atkinson, K. D., Kolat, A. I. and Culbertson, M. R. (1977). Growth and metabolism of inositol-starved Saccharomyces cerevisiae. J. Bacreriol. 130,472474. Hinnen, A., Hicks, J. B. and Fink, G . (1978). Transformation of yeast. Proc. Natl. Acad. Sci. USA 75, 1929-1933. Jones, E. W., Zubenko, G. S. and Parker, R. R. (1982). PEP4 gene function is required for expression of several vacuolar hydrolases in Saccharomyces cerevisiae. Genetics 102,66%77. Kaisho, Y., Yoshimura, K. and Nakahama, K. (1989). Increase in gene expression by respiratory-deficient mutation. Yeast 5,91-98. Kramer, R. A., DeChiara, T. M., Schaber, M. D. and Hilliker, S. (1984). Regulated expression of a human interferon gene in yeast: Control by phosphate concentration or temperature. Proc. Natl. Acad. Sci. USA 81, 367-370. Lindquist, S. and Craig, E. A. (1988). The heat shock proteins. Ann. Rev. Genet. 55, 1151-1 191. McClanahan, T. and McEntee, K. (1986). DNAdamage and heat shock dually regulate genes in Saccharomyces cerevisiae. Mol. Cell. Biol. 6,9&96. Munro, S. and Pelham, H. R. B. (1985). What turns on heat shock genes? Nature 317,477478. Piper, P. W., Curran, B., Davies, M. W., Hirst, K., Lockheart, A., Ogden, J. E., Stanway, C., Kingsman, A. J. and Kingsman, S. M. (1988). A heat shockelement within the phosphoglycerate kinase gene promoter of Saccharomyces cerevisiae. Nucl. Acids Res. 16, 1333-1348. Praekelt, U. and Meacock, P. A. (1990). HSP12, a new small heat shock gene of Saccharomyces cerevuiae: Analysis of structure, regulation and function. Mol. Gen. Genet. 223,97-106. Seufert, W. and Jentsch, S. (1990). Ubiquitin-conjugating enzymes, UBC4 and UBC5 mediate selective degradation of short-lived and abnormal proteins. EMBO J 9,543-550. Sherman, F., Fink, G. R. and Hicks, J. B. (1983). Synthetic complete medium. In Methodv in Yeast Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, pp. 62-84. Sledewski, A. Z., Bell, A., Kelsay, K. and MacKay. V. L. (1988). Construction of temperature-regulated yeast promoters using the MAT2 repression system. Biofechnology 6 , 4 1 1 416. Sorger, P. K. and Pelham, H. R. B. (1987). Purification and characterisation of a heat-shock element binding protein from yeast. EMBO 56,3035-3041. Sorger, P. K. and Pelham, H. R. B. (1988). Yeast heat shock factor is an essential DNA-binding protein that exhibits temperature-dependent phosphorylation. Cell 54,855-864. Sorger, P. K. (1990). Yeast heat shock factor contains separable transient and sustained response transcriptional activators. Cell 62,793-805.
546 Tanaka, K., Matsumoto, K. and Toh-e, A. (1988). Dual regulation of the expression of the polyubiquitin gene by CAMPand heat shock in yeast. EMBO 57,495499. Varshavsky, A,, Bachmair, A., Finley, D., Gonda, D. and Wunning, 1. (1988). The N-end rule of selective protein turnover-mechanistic aspects and functional implications. In Rechsteiner, M. (Ed.), Ubiquitin. Plenum Press, New York, pp. 287-324. Wei, R., Wilkinson, H., Pfeifer, K., Schneider, C., Young, R. and Guarente, L. (1986). Two or more
N. KIRK AND P.W. PIPER
copies of Drosophila heat shock consensus sequence serve to activate transcription in yeast. Nucl. Acids Res. 14,8183-8188. Wingfield, J. M. and Dickinson, J. R. (1990). The activity of a model heterologous protein in S. cerevisiae strains with reduced vacuolar proteinase activity. Yeast 6, S421 .
VOL.
7: 539-546 (1991)
The Determinants of Heat-Shock Element-Directed lac2 Expression in Saccharomyces cerevisiae NIALL KIRK*? AND PETER W. PIPER*$ Departments of * Biochemistryand Molecular Biology and ?Chemical and BiochemicalEngineering, University College London, London WClE 6BT. U.K. Received 17 September 1990; revised 5 February 1991
Heat-shock induction of heat-shock protein genes is due to a specific promoter element (the heat-shock element, HSE). This study used lac2 under HSE control (HSE-lacZ) to characterize HSE activity in Saccharomyces cerevisiae cells of different physiological states and differing genetic backgrounds. In batch fermentations HSE-IacZinduction by heat shock was maximal in exponential growth, and showed marked decline with the approach to stationary phase. Expression in the absence of heat shock was unaffected by growth phase, indicating that the growth-dependent expression of many yeast heat-shock genes uses promoter elements in addition to the HSE. Heat-induced expression was strongly influenced by the temperature at which cultures were grown. While basal, uninduced expression was constant during growth at different temperatures to 3 0 T , induction by transfer to 39°C was reduced by increases in growth temperature as low as 18-24°C. Maximal HSE-lacZ induction (30- to 50-fold) was in cultures grown at low temperatures (18-24"C), then heat shocked at 39°C. Ethanol was a poor inducer. Mutations having little effect on HSE-ZacZ expression included a respiratory petite; ubi4 (which inactivates the polyubiquitin gene); also ubc4 and ubc5 (which each inactivate one of the ubiquitin ligases involved in degradation of aberrant protein). pep4-3 increased both basal and induced P-galactosidaseabout two-fold, probably because of slower turnover of this enzyme inpep4-3 strains. KEY WORDS - Saccharomyces cerevisiae; heat-shock
element; IacZ induction; growth state; ubi4, ubcl, ubc5, pep4-3
mutants.
INTRODUCTION Temperature-regulated yeast expression systems achieve gene induction through either a temperature downshift (Brake et al., 1988; Kramer et al., 1984; Sledewski et al., 1988) or a temperature upshift (DaSilva and Bailey, 1989; also the heat-shock element (HSE)-directed expression investigated in this study). Important considerations need to be met before temperature shifts can be used to control the expression of a heterologous gene. Firstly, since Saccharomyces cerevisiae grows more slowly with reduced biomass yield at temperatures above 34-36°C it is usually desirable to grow cultures at temperatures slightly below this value. Secondly, while systems using temperature downshift for induction may be straightforward to operate on a small scale, they are not as attractive for larger-scale processes as those systems that use temperature upshift. This is because heat addition to large fermenters is easier than heat extraction. $Author to whom correspondence should be addressed. 0749-503X/9 1/060539-08 $05.00 0 1991 by John Wiley & Sons Ltd
Induction by temperature upshift is readily achieved using the promoter element (HSE) which directs the activation of heat-shock genes at elevated temperatures. Sequences placed under HSE control then become activated together with other heat-shock genes as the heat-shock response is induced. The HSE is the promoter binding site for heat-shock factor (HSF), a protein which partly acquires its activity as a transcriptional transactivator by heat-shock-induced phosphorylation (Sorger and Pelham, 1987,1988; Sorger, 1990). HSE-directed expression systems allow yeast to be grown under conditions enabling efficient biomass accumulation. Also unlike other yeast temperatureregulated expression systems described to date, they can operate in a variety of genetic backgrounds. They are not, however, without their potential problems, notably: (i) The heat-shock response is transient at sublethal temperatures, being switched off within
540
N. KIRK AND P. W. PIPER
Table 1 . Yeast strains used in this study Strain CT3b* CT3c* SUB60t SUB62t ubc4t ubc5t
Genotype leu2 trpl leu2 trpl leu2 trpl leu2 trpl leu2 trpl leu2 trpl
ura3 pep4-3 ura3 ura3 his3 lys2 ubi4::LEUZ ura3 his3 lys2 ura3 his3 lys2 ubc4::HIS3 ura3 his3 lys2 ubcS::LEU2
Source J. Wingfield J. Wingfield
D. Finley D. Finley S. Jentsch S. Jentsch
*CT3b and CT3c are products of a single meiotic tetrad (Wingfield and Dickinson. 1990). tCongenic' with' diploid DF5 (Finley et al., 1987; Seufert and Jentsch, 1990).
30-120 min of induction (Lindquist and Craig, 1988). (ii) Induction, as measured with an Escherichia coli lacZ reporter gene, is usually not greater than 30- to 50-fold. This is due both to the limited expression period and measurable HSF activity in the absence of heat shock (Wei et al., 1986; Sorger and Pelham, 1987; Sorger, 1990; also results described below). (iii) Secretion may not be very efficient at those temperatures optimal for HSF activation. (iv) The temperatures used to heat-stress yeast may detrimentally affect certain heterologous products. Since heat shock induces certain genes of the ubiquitination cycle for protein turnover (Finley et al., 1987; Seufert and Jentsch, 1990) it may stimulate the turnover of aberrant (but possibly not native) protein. Despite such possible disadvantages HSEdirected expression might find applications where a limited expression period will suffice, and where authentic rather than aberrant product can accumulate intracellulary at 37-39°C. We therefore extended previous studies of expression of lacZ under HSE control (HSE-lacZ), so as to determine how this expression is influenced by: (i) growth state, and growth temperature prior to heat shock; (ii) magnitude of temperature upshift, and final temperature of heat shock; (iii) other inducers of heat-shock genes; (iv) respiratory-deficiency (rhocells); (v) inactivation of genes of the ubiquitination system for turnover of aberrant protein; and (vi) a mutation (pep4-3)which substantially reduces vacuolar protease activities.
MATERIALS AND METHODS Yeast strains and yeast transformation The S . cerevisiae strains employed are listed in Table 1. Each was transformed to uracil prototrophy using plasmid pHSE2 (Sorger and Pelham, 1987) and the transformation procedure of Hinnen et al. (1978). These transformants are denoted by a -pHSE2 suffix in conjunction with strain designations (e.g. CT3b-pHSE2). pHSE2 contains a tandem copy of the HSE sequence known to selectively bind yeast heat-shock transcription factor (Sorger and Pelham, 1987). This HSE is inserted within the cycI promoter of the cytochrome cl-lacZ fusion gene of pHSE2, where it replaces the normal upstream activating sequences of this promoter. Spontaneous petites were selected from SUB62pHSE2 as small colonies on standard mineral salts glucose-defined (SD) medium, rho- colonies which both lacked mitochondria1 cytochromes and displayed no capacity for gluconeogenic growth. Culture media Transformants were grown aerobically in vigorously-shaken flask cultures, either on SD medium (Sherman et al., 1983) which has 2 mg I - ' meso-inositol; or on SD with an increased (5 mg 1 ') meso-inositol supplement (data in Figure 1 only). Appropriate amino acid supplements were also present, omitting uracil to ensure pHSE2 maintenance. To heat-shock cells, flask cultures were placed in water baths at the appropriate temperature, temperature equilibration occurring over less than 3 min. In preliminary experiments HSE-lacZ expression 1 h after heat shock to 39°C was found to -
DETERMINANTS OF HEAT-SHOCK ELEMENT -DlRECTED L A C 2 EXPRESSION IN SACCHAROMYCES CEREVISIAE
54 1
RESULTS Basal HSE-lacZ expression is not growth phase-dependent
In the absence of heat shock, SD medium batch cultures of pHSE2 transformants displayed a low, constant HSE-lacZ expression. This uninduced f3galactosidase did not rise with transitions to slower growth, either at the start of transition phase or during entry to stationary phase (Figure I ) . Similar low basal activities, unaffected by growth phase, were found during batch fermentations on complex (YEPD) medium (not shown). Figure I . Batch fermentation of SUB62-pHSE2 in 23°C YNBG medium (see Methods) with either 2 mg 1 I ( A ) or a 5 mg I I (2.)inositol supplement. 1, I1 and I11 approximate to the exponential, transition and stationary phases of growth respectively. P-galactosidase activity both before (0) and I h after heat shock 10 39°C ( 0 )was identical for both medium inositol concentrations. This figure shows the progressive decline of the heat inducibility of the HSE-lucZgene during transition phase and the entry to stationary phase.
be the same irrespective of whether cultures had taken 30 s or 5 min to attain this temperature.
The optimal temperature for HSE-lacZ induction
With SUB62-pHSE2 initially in exponential growth at 23°C only a narrow range of heat-shock temperatures gave high f3-galactosidase induction (Figure 2). Induction was maximal at 39"C, 3 7 T giving about half the induction of 39°C. Table 2 also shows that the final temperature of heat stress is a major determinant of induction level.
- 507 L
E
C
5 LOa
/IGalac - t osidase ussays
f3-Galactosidase was assayed as in Sorger and Pelham (1987), either in the absence of heat shock or at the stated intervals after heat shock. Each data point is the mean of two experimental determinations (S.D. 20%).
U
m
-g 33-3 0
22910-
/'
Plasmid copy level determinations
Total yeast DNA was isolated from untransformed and pHSE2-transformed cultures as in Piper er al. (1988). These DNA samples were slotblotted in duplicate, then hybridized (as in Piper et al., 1988) to either a URA3 gene probe (the 0.8 kb EcoI-Smul fragment of pHSE2) or a PGK gene probe (the 1.95 kb HindIII fragment of pMA1; Piper et al., 1988). The increased URA3 hybridization signal in DNA samples from pHSE2 transformants indicated fairly constant copy number (7 & 3) for this plasmid in all cultures used for lac2 expression measurements. Observed differences in lacZ expression could not therefore be attributed to changes in plasmid copy level.
Figure 2. The importance of final temperature of heat shock on induction. Using SUB62-pHSE2 in exponential growth at 23°C. p-galactosidase levels were measured I h after shift to the temperatures indicated.
HSE-lacZ induction is highly growth phasedependent and maximal in exponential growth Heat-induced P-galactosidase was maximal with cells that were in exponential growth at the time of heat shock (Figure 1). Marked decreases in heat-shock induction accompanied the approach to stationary phase, although a slight induction (about
542
N. KIRK AND P. W. PIPER
Table 2. Effects of the magnitude of temperature upshift and final temperature on heat-shock induction levels
Heat shock By 13°C upshift: 2&33"C 22-35°C 24-37°C 2639°C By 15°C upshift: 18-33°C 2&35"C 22-37°C 2639°C By 17°C upshift: 18-35°C 2&37"C 22-39°C By 19°C upshift: 18-37°C 2&39"C
Units P-galactosidase in heat-shocked cells*
Induction ratio?
12 20 28.2 68.8
4.4 7.4 10.4 25.5
13.3 23.5 38.0 80.3
4.9 8.7 14.1 29.7
25.8 37.5 83.8
10.5 13.9 31.0
28.4 92.6
27.8 34.3
*Measured 60 min after temperature upshift. tIn the SUB60-HSE2 cultures used for these determinations basal p-galactosidase activity at 18,20,22,24 and 26°C prior to heat shock was 2.7 units.
two-fold) could still be obtained in stationary phase (Figure 1). These stationary cells could regain maximal capacity for heat induction within 2 h if diluted 20-fold in fresh 23°C SD medium (not shown). Both basal and heat-induced HSE-lac2 expression were unaffected by increasing the inositol supplement of SD medium (Figure 1) even though this allows the attainment of slightly higher cell densities (Henry et al., 1977). Growth temperature has relatively minor efects on basal HSE-lacZ expression levels, but strongly influences induction by heat shock Growth temperature was found to be extremely important in determining levels of induction by heat shock. Uninduced P-galactosidase was constant during growth at temperatures from 18°C to 30"C, only increasing in cultures growing above 30°C (Figure 3). In contrast, HSE-IacZinduction by heat shock showed a dramatic decline with increasing growth temperatures from 18°C to 30°C (Table 2 and Figure 3), being greatest for those cells grown at 18-20°C prior to heat shock at 39°C. S. cerevisiae is often grown at 28-33"C, the temperatures that enable optimal growth rates and
!OOr
Growth temperature
Figure 3. Variation in heat inducibility of the HSE-lucZ gene in SUB62-pHSE2 growing at different temperatures. Cells were grown at the indicated temperatures, their P-galactosidase levels being measured both before (0)and 1 h after ( 0 )heat shock to 39°C.
substrate conversions. Although SUB62-pHSE2 growing at 30°C gave only ten-fold HSE-IacZ induction after shift to 39°C (Figure 3), these same cells could display > 30-fold induction if placed at 23°C for 2 h prior to heat shock at 39°C (not shown). This indicates that it may be necessary to cool cells for a period prior to heat shock in order to obtain maximal HSE induction with yeast cultured
DETERMINANTS OF HEAT-SHOCK ELEMENT -DIRECTED L A C Z EXPRESSION IN SACCHAROMYCES CEREVISIAE
Table 3.
HSE-kuc% induction by various inducers of
stress responses
Agent Hcal shock to 39°C Ethanol 8 % (v'v)
Dimcthylsulphoxide 10% ( w / \ ) Canavaninc I mg:ml NaCl 1.7 M Paromomycin I m g m l 4-nitroquinoline- I -oxide I pg/ml
0-Galactosidasc induction > 30XA 2X" Nilb Nilb
Nilb 2X' 2Xh
Induction wasmeasuredeither I h(a), 1 4 h(b)orS h(c)afterthe addition of each agent to SUB62-pHSE2 cells in exponential growth at 23 C. the fraction of the cells surviving at these times k i n g > 9X'!AI in all cdscs.
at those temperatures that allow optimal growth rates. HSE-lacZ is poorly induced b y many inducers of heat-shock genes A striking feature of the heat-shock response is its multiplicity of environmental and chemical stress agents (Ananthan et al.. 1985; Lindquist and Craig, 1988). We tested a number of known inducers of stress-responsive genes in S . cerevisiae to see if they gave appreciable HSE-lucZ induction (Table 3). Without exception all gaveonlyvery slight induction (ethanol. paromomycin and 4-nitroquinoline- 1oxide) or no detectable change in expression (dimethylsulphoxide, canavanine and high osmolarity).
Eflects qfyeast genetic background on HSE-lac2 expression
To determine how respiratory deficiency affects HSE-IucZ expression, spontaneous rho- mutants were isolated from SUB62-pHSE2. Both the basal and the heat-shock induced P-galactosidase of these r/io cells were approximately two-fold higher than in the rho ' parents (not shown). rho did not therefore greatly enhance intracellular accumulation of P-gdlactosidase. The effects of rho- may be more dramatic on the synthesis of secreted products since it increases human lysozyme secretion by yeast up to ten-fold, especially at stationary phase (Kaisho et ul., 1989). Heat shock activates certain genes of the ubiquitination system for intracellular protein turnover.
543
These include UBZ4 (encoding polyubiquitin; Finley et ul., 1987); also UBC4 and UBCS (encoding the ubiquitin ligases which join ubiquitin to substrates destined for proteolysis; Seufert and Jentsch, 1990). Although P-galactosidase normally has a long halflife ( > 20 h) in yeast (Varshavsky et al., 1988), when made at heat-shock temperatures it could conceivably turn over more rapidly due to simultaneous activation of the ubiquitination pathway. ubi4 expression hosts might therefore give higher andl or more prolonged synthesis. We compared pgalactosidase induction in SUB60-pHSE2 and SUB62-pHSE2, two transformants that are isogenic but for a LEU2 gene disruption of the UBM locus in SUB60-pHSE2 (Finley et al., 1987). The uhi4 strain did not give appreciably higher or more prolonged HSE-lacZ expression, its P-galactosidase synthesis after heat shock being only slightly greater than in UBI4' cells (Figure 4). Should P-galactosidasc synthesized at heat-shock temperatures be subject to appreciable degradation by the ubiquitination pathway, reductions in ubiquitin ligase activity might also increase its stability. Two such ligases, UBC4 and UBC5, are involved in degradation of abnormal proteins (Seufert and Jentsch, 1990). Preventing the expression of either UBC4 or UBCS might therefore conceivably increase heat-induced HSE-IacZexpression. This was not found, the use of ubc5 (Figure 4) or uhc4 (not shown) expression hosts having no major effect on heat-shock induction of p-galactosidase. This is consistent with earlier findings that the turnover of long-lived and short-lived normal proteins (unlike proteolysis of abnormal canavanyl-peptides) is barely affected by ubc4 or ubc5 as single mutations (Seufert and Jentsch. 1990). The highly-selective protein turnover catalysed by the ubiquitination system (Ciechanover and Schwartz, 1989) contrasts with the non-specific proteolysis catalysed by the vacuolar proteases. pep4.3, a defect in the P R A l gene, reduces the activity of several vacuolar hydrolases by over 90% (Jones et al., 1982). pep4.3 expression strains are therefore in widespread use to minimize the breakdown of products made in yeast during their recovery. Basal and heat-induced HSE-fucZ expression were both approximately two-fold higher in a pep4.3 strain as compared to a wild-type strain of the same parentage (Figure 4). These effects appeared not to be due to any higher instability of P-galactosidase in crude cell extracts from wild-type cells (the rates of inactivation of this enzyme in crude extracts from 6T3b-pHSE2 and 6T3c-
544
N. KIRK AND P. W. PIPER 150
I ~
IOC
0 v)
0 0 c
-00 a v)
c
5
5c
i
I
I
60
120
I
I
180
60
I
120
1
180
Minutes after 23-39OC shift
Figure 4. Heat-shock induction of P-galactosidase in SUB62-pHSE2 (m), SUB60pHSE2 (A),ubc5-pHSE2( A ) , 6T3b-pHSE2( 0 )and 6T3c-pHSE2(0). All cultures were in exponential growth at 23°C prior to heat shock to 39°C for the indicated times.
pHSE2 were similar irrespective of the presence or absence of protease inhibitors; data not shown). P-galactosidase increases due to pep4-3 have also been found by Wingfield and Dickinson (1990) for lucZ expressed from the complete CYCl promoter. They have also concluded this is due to pep4-3 increasing P-galactosidase during growth, rather than any artefact of the procedure for making cell extracts prior to enzyme assay. DISCUSSION The synthetic heat-shock promoter used here has previously been employed to measure the activity of HSF in S. cerevisiue (Sorger and Pelham, 1987). Variants of this promoter bearing point mutations in the HSE that interfere with HSF binding cause p-galactosidase activity to be low under all conditions (Sorger and Pelham, 1987).Therefore the pgalactosidase levels in pHSE2-transformants reflect the activity of HSF in enhancing transcription. Here we have extended these studies in order to identify the extent to which successful operation of HSEbased induction systems may be dependent on the physiological state, the culture conditions and the genetic background of S. cerevisiue cells. Our results demonstrated that HSE-lucZ induction is highly dependent on growth state and temperature of growth
(Figures 1 and 2; Table 2); also the final temperature of heat stress (Figure 2 and Table 2). Table 2 clearly shows that the magnitude of temperature upshift is not the sole determinant of induction level. Several S . cerevisiue heat-shock genes are under growth phase control, being activated in the absence of heat shock as batch fermentations approach stationary phase. Such genes include SSA3 (Boorstein and Craig, 1990), H S P l 2 (Praekelt and Meacock, 1990) and UBZ4 (Finley et ul., 1987; Tanaka et ul., 1988). However, there is no increase in basal HSE-ZucZ expression at the approach to stationary phase (Figure 1). The activation of many natural heat-shock promoters towards the end of batch fermentation would therefore appear to require promotor-controlling elements in addition to HSE and HSF. At least one of these promoters (SSA3) has a cyclic AMP-responsive element (Boorstein and Craig, 1990). Certain of the agents in Table 3 induce a subset of the heatshock proteins of S. cerevisiue (e.g. 8% ethanol, which strongly induces Hsp70, Hsp90 and Hsp104, unpublished observations; paromomycin, which causes translational errors, Grant et ul., 1989; and 4-nitroquinoline- 1-oxide, a DNA-damaging agent, McClanahan and McEntee, 1986). The fact that none of these agents gave strong HSE-
DETERMINANTS OF HEAT-SHOCK ELEMENT -DIRECTED LACZ EXPRESSION IN SACCHAROMYCES CEREVISIAE
lac2 induction suggests they are n o t s t r o n g HSF activators. T h e activation o f UBi4 by heat shock m a y help overcome a n y shortage o f free ubiquitin t h a t arises from t h e need f o r increased turnover o f thermallydenatured protein by t h e ubiquitination system ( M u n r o a n d Pelham, 1985). If a heterologous protein induced by heat shock is degraded more rapidly because of this activation, prevention o f UBI4 expression m a y lead t o higher expression levels. Alternatively, if shortage of free, unconjugated ubiquitin is t h e induction signal f o r the heatshock response ( M u n r o a n d Pelham, 1985), a ubi4 genetic background might give m o r e sustained heatshock-induced transcription. Our results (Figure 4) s h o w t h a t b o t h the stability o f heat-induced 0galactosidase a n d the switch-off o f HSF-directed HSE-lmZ expression a t 39°C are largely independ e n t o f UBI4 expression and its influence on ubiquitin levels. ACKNOWLEDGEMENTS W e thank Peter Sorger f o r pHSE2, and J o n a t h a n Wingfield, Daniel Finley a n d Stefan Jentsch f o r strains. T h i s w o r k w a s supported by a SERC g r a n t (GRiF;72932) a n d CASE studentship (N.K.) held in conjunction with ICI Pharmaceuticals.
REFERENCES Ananthan, J., Goldberg, A. L. and Voellmy, R. (1985). Abnormal proteins serve as eukaryotic stress signals and trigger the activation of heat shock genes. Science 232,522-524. Boorstein, W. R.and Craig, E. A. (1990). Transcriptional regulation or SSA3, an HSP70 gene from Saccharomyces cerevisiae. Mol. Cell. Biol. 10,3262-.3269. Brake, A. J.. Merryweather, J. P., Coit, D. G., Heberlain, V. A,. Maziarz, F. R., Mullenbach, G. T., Urdea, M. S., Valenzuela. P. and Barr, P. J. (1984) a-factor-directed synthesis and secretion of mature foreign proteins in Sacchnromyces cerevisiae. Proc. Natl. Acud. Sci. USA 81,46424646. Ciechanover, A. and Schwartz, A. L. (1989). How arc substrates recognised by the ubiquitin-mediated proteolytic system? Trends Biochem. Sci. 14,483487. DaSilva, N . A. and Bailey, J. E. (1989). Construction and characterisation of a temperature-sensitive expression system in east. Biotech. Prog. 5, 18-26. Finley, D., dzkaynak, E. and Varshavsky, A. (1987). The ycast polyubiquitin gene is essential for resistance to high temperatures, starvation, and other stresses. Cell 48, 1035--1046. Grant, C. M., Firoozan, M. and Tuite, M. F. (1989). Mistranslation induces the heat shock response in the yeast Succharomyces cerevisiae. Mol. Microbiol. 3 , 2 1 5-220.
545
Henry, S. A., Atkinson, K. D., Kolat, A. I. and Culbertson, M. R. (1977). Growth and metabolism of inositol-starved Saccharomyces cerevisiae. J. Bacreriol. 130,472474. Hinnen, A., Hicks, J. B. and Fink, G . (1978). Transformation of yeast. Proc. Natl. Acad. Sci. USA 75, 1929-1933. Jones, E. W., Zubenko, G. S. and Parker, R. R. (1982). PEP4 gene function is required for expression of several vacuolar hydrolases in Saccharomyces cerevisiae. Genetics 102,66%77. Kaisho, Y., Yoshimura, K. and Nakahama, K. (1989). Increase in gene expression by respiratory-deficient mutation. Yeast 5,91-98. Kramer, R. A., DeChiara, T. M., Schaber, M. D. and Hilliker, S. (1984). Regulated expression of a human interferon gene in yeast: Control by phosphate concentration or temperature. Proc. Natl. Acad. Sci. USA 81, 367-370. Lindquist, S. and Craig, E. A. (1988). The heat shock proteins. Ann. Rev. Genet. 55, 1151-1 191. McClanahan, T. and McEntee, K. (1986). DNAdamage and heat shock dually regulate genes in Saccharomyces cerevisiae. Mol. Cell. Biol. 6,9&96. Munro, S. and Pelham, H. R. B. (1985). What turns on heat shock genes? Nature 317,477478. Piper, P. W., Curran, B., Davies, M. W., Hirst, K., Lockheart, A., Ogden, J. E., Stanway, C., Kingsman, A. J. and Kingsman, S. M. (1988). A heat shockelement within the phosphoglycerate kinase gene promoter of Saccharomyces cerevisiae. Nucl. Acids Res. 16, 1333-1348. Praekelt, U. and Meacock, P. A. (1990). HSP12, a new small heat shock gene of Saccharomyces cerevuiae: Analysis of structure, regulation and function. Mol. Gen. Genet. 223,97-106. Seufert, W. and Jentsch, S. (1990). Ubiquitin-conjugating enzymes, UBC4 and UBC5 mediate selective degradation of short-lived and abnormal proteins. EMBO J 9,543-550. Sherman, F., Fink, G. R. and Hicks, J. B. (1983). Synthetic complete medium. In Methodv in Yeast Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, pp. 62-84. Sledewski, A. Z., Bell, A., Kelsay, K. and MacKay. V. L. (1988). Construction of temperature-regulated yeast promoters using the MAT2 repression system. Biofechnology 6 , 4 1 1 416. Sorger, P. K. and Pelham, H. R. B. (1987). Purification and characterisation of a heat-shock element binding protein from yeast. EMBO 56,3035-3041. Sorger, P. K. and Pelham, H. R. B. (1988). Yeast heat shock factor is an essential DNA-binding protein that exhibits temperature-dependent phosphorylation. Cell 54,855-864. Sorger, P. K. (1990). Yeast heat shock factor contains separable transient and sustained response transcriptional activators. Cell 62,793-805.
546 Tanaka, K., Matsumoto, K. and Toh-e, A. (1988). Dual regulation of the expression of the polyubiquitin gene by CAMPand heat shock in yeast. EMBO 57,495499. Varshavsky, A,, Bachmair, A., Finley, D., Gonda, D. and Wunning, 1. (1988). The N-end rule of selective protein turnover-mechanistic aspects and functional implications. In Rechsteiner, M. (Ed.), Ubiquitin. Plenum Press, New York, pp. 287-324. Wei, R., Wilkinson, H., Pfeifer, K., Schneider, C., Young, R. and Guarente, L. (1986). Two or more
N. KIRK AND P.W. PIPER
copies of Drosophila heat shock consensus sequence serve to activate transcription in yeast. Nucl. Acids Res. 14,8183-8188. Wingfield, J. M. and Dickinson, J. R. (1990). The activity of a model heterologous protein in S. cerevisiae strains with reduced vacuolar proteinase activity. Yeast 6, S421 .
