7. Embryol. exp. Morph. 83, Supplement, 147-161 (1984)

Printed in Great Britain © The Company of Biologists Limited 1984

Heat shock — a comparison of Drosophila and yeast

By SUSAN LINDQUISTThe Department of Molecular Genetics and Cell Biology, The University of

Chicago, 1103 E. 57th Street, Chicago, Illinois 60637, U.S.A.

TABLE OF CONTENTS

IntroductionThe response to high temperatureDevelopmental inductionConclusionsReferences

INTRODUCTION

When cells or whole organisms are exposed to temperatures slightly abovetheir optimum for growth, they respond by synthesizing a small group ofproteins, called the heat shock proteins (hsps), which help protect them from thetoxic effects of heat. The same set of proteins can also be induced by a widevariety of other stresses including exposure to ethanol, heavy metal ions, andinhibitors of respiratory metabolism. Their induction is apparently a very generalreaction to adverse conditions. (See Schlessinger, Ashburner & Tissieres, 1982,for review.)

This response is the most highly conserved genetic regulatory system known.The proteins produced by fruit flies are homologous to those produced byenterobacteria, corn plants, slime moulds, yeasts, sea urchins, and humans(Bardwell & Craig, 1984; Key, Lin & Chen, 1981; Loomis & Wheeler, 1980;McAlister, Strausberg, Kulaga & Finkelstein, 1979; Giudice, Roccheri &Debernardo, 1980; Thomas, Welch, Matthews & Feramisco, 1982). Relatedspecies have been detected even in archaebacteria (Bardwell & Craig, 1984;Daniels, McKee & Doolittle, 1984). The list of species in which the heat shockresponse has now been studied is enormous.

Determining how this small group of proteins is able to protect such a diversespectrum of organisms from so many different types of stress is a fascinatingbiological problem. Unfortunately, it is one which still awaits a solution at themolecular level. But the response has attracted study for another reason as well:the rapid and extremely reproducible nature of the induction makes it an attrac-tive model system for learning about the mechanisms cells use to alter their

148 S. LINDQUIST

patterns of gene expression. In this respect investigations have met with great suc-cess: heat shock promoter elements have been dissected and used to place othergenes under heat shock regulation (Pelham, 1982; Pelham & Bienz, 1982);changes in chromatin structure associated with the activation of the genes hasbeen described in remarkably fine detail (Wu, 1980,1984; Keene, Corces, Lo wen-haupt & Elgin, 1983); factors required for preferential transcription have beenfractionated and characterized in vitro (Bonner, 1981; Topol & Parker, 1984).

For the past few years my laboratory has been studying the heat shock res-ponses of two very different organisms - the fruit fly, Drosophila, and yeast,Saccharomyces cerevisiae. Both organisms are able to achieve a rapid synthesisof heat shock proteins in response to elevated temperatures. In keeping with theenormous differences in their biology, however, different regulatory mechan-isms are employed to accomplish this induction. Many other basic features of theresponse differ in the two organisms. It is somewhat surprising, therefore, thatin both organisms a particular subset of the heat shock proteins is induced duringthe normal course of development, in the absence of heat. The similarity be-tween the two developmental inductions indicates it is a very ancient pattern andsuggests that the heat shock proteins not only provide protection from stress butmay also serve an important role in the natural life cycles of many organisms.

THE RESPONSE TO HIGH TEMPERATURE

Both yeast and Drosophila grow well at 25 °C, the temperature normally usedfor their culture in the laboratory. In Drosophila, the optimal temperature forinduction of heat shock proteins is between 36 and 37°C (Lindquist, 1980a,b)\in Saccharomyces cerevisiae, it is between 39 and 40 °C (Lindquist et al. 1982).Fig. 1 compares the patterns of protein synthesis in these organisms at theirstandard growing temperatures and during heat shock.

In both cases, transfer to high temperature is accompanied by a rapid induc-tion of heat shock proteins. In Drosophila the major proteins have relativemolecular masses (Mr) of 83000, 70000, 68000, 28000, 26000, 23000, and22 000 (designated hsp83, hsp70, etc. for obvious reasons). Yeast has proteins of84000, 70000, 69000, and 26000, which appear to be analogous to the corres-pondingly sized proteins of Drosophila. Nucleotide sequence analysis of the70 000 Mx protein coding genes of the two species predicts 72 % homology at theamino acid level (Ingolia, Slater & Craig, 1982). Sequence analysis for the otheryeast proteins is not yet complete, but the genes have all been cloned and com-parative data should be available soon. A difference between the two organismsis that yeast cells produce only one small heat shock protein while Drosophilacells produce four closely related species. An additional difference is that yeastcells synthesize a protein of 96 000 Mr which does not appear to have a counter-part in Drosophila. In this regard Drosophila is unusual, as most organismssynthesize a high relative molecular mass species.

Heat shock - a comparison of Drosophila and yeast 149Drosophila Yeast

C HS C HS

^ - 9 684

_ -70

-26

- 2 8- 2 6- 2 3- 2 2

Fig. 1. Heat shock proteins in Drosophila and yeast. Cells were grown at 25 °C tomid-log phase and then either maintained at this temperature (C) or heat shockedfor 1 h (HS). 3H-labelled isoleucine was added during the final 30 min of incubation.Drosophila cells (Schneider's line No. 2) were heat shocked at 36-5 °C and theproteins were analysed in 10 % SDS polyacrylamide gels. Yeast cells (strain A364a)were heat shocked at 39°C and the proteins analysed on 12 % gels. Mt x 10~3.

A major distinction between the Drosophila and yeast responses is in the effectheat treatment has on normal protein synthesis. As shown in Fig. 1, theDrosophila response is much cleaner than the yeast response. Normal proteinsynthesis is virtually undetectable in Drosophila cells, while in yeast cells it is

150 S. LINDQUIST

only partially reduced. A time course of the induction in the two organisms ismore revealing. As shown in Fig. 2, at the maximal induction temperaturejbrhsp70, 37 °C, the repression of normal protein synthesis in Drosophila cells issudden and absolute. Even at more moderate temperatures, where normalprotein synthesis is only partially repressed, the repression is rapidly achievedand maintained at a constant level. In yeast cells, by contrast, even at hightemperatures the repression is progressive and relatively slow.

The underlying cause of this difference is now known to be a difference inregulation. In both organisms, the exposure to high temperature results in im-mediate transcriptional activation of the heat shock genes. Only in Drosophila,however, is this accompanied by the induction of a specific translational controlmechanism which rigorously discriminates against pre-existing messages.

The evidence for a specific translational mechanism operating in Drosophilacells is very strong. First, the disappearance of normal cellular protein synthesisduring heat shock is not due to the degradation of pre-existing messages. If heatshocked cells are treated with actinomycin D (to prevent the synthesis of newmessenger RNAs) and then returned to normal temperatures, the full spectrumof normal protein synthesis is still restored (McKenzie, 1977; Storti, Scott, Rich& Pardue, 1980; Lindquist, 1981). That the pre-existing messages have not

•37°C

nISit*

H i mm *•* mm mm mm mm *— « • •"• mmmt

|PH ^m ^ ^ ^ ^ ^ ^ ^ ^ u M g B i im mmmmi^b ^ a a a > » • • ^mmm

1Drosophila Yeast

Fig. 2. Heat shock time course in Drosophila and yeast. Small aliquots of cells grownat 25 °C were heat shocked by immersing tubes in water baths at the indicatedtemperatures. Every 15 min an aliquot was labelled with [3H]isoleucine. Drosophilacells were grown and labelled in Shields and Sang medium; yeast cells were grownin a synthetic acetate medium.

Heat shock - a comparison of Drosophila and yeast 151changed appreciably in concentration or undergone any major physical modifi-cations has been demonstrated by in vitro translation of total cellular RNAs incell-free lysates and by hybridization of electrophoretically separated RNAs tocloned probes (Mirault etal. 1978; DiDomenico, Bugaisky & Lindquist, 1982a).

Second, the messages for normal cellular proteins are not simply swamped outof translation by the massive influx of new heat shock mRNAs. Within 10 min oftemperature elevation, pre-existing polysomes have all but disappeared. Newpolysomes translating the heat shock mRNAs do not appear in substantial num-bers for another 10 or 15 min (see Fig. 3). Furthermore, if heat shock messagesynthesis is blocked by the addition of actinomycin D before heat shock, pre-existing polysomes still disappear.

Third, the change in translational specificity is not simply due to a direct effectof temperature on the secondary structure of the messenger RNA. Cell-freelysates made from heat shocked Drosophila tissue culture cells retain the capac-ity to discriminate against normal cellular messenger RNAs (Storti et al. 1980;Scott & Pardue, 1981). Lysates made from control cells incubated under thesame conditions show no discrimination. When heat shock messages are mixedwith messages for normal cellular proteins and translated in vitro at varioustemperatures, the relative translational efficiency of the two classes is not effec-ted. Finally, when tissue culture cells are heat shocked and returned to 25 °C,normal patterns of protein synthesis are not restored until a specific quantity ofheat shock protein has been produced, indicating the discrimination againstnormal cellular messengers occurs in response to a physiological need for heatshock proteins and not as a consequence of the temperature shift per se(DiDomenico et al. 19826).

A change in the pattern of protein synthesis can be produced by a non-specificdecrease in the efficiency of ribosome initiation since different messages natur-ally have different intrinsic affinities for ribosomes (Lodish, 1976). If heat shockcaused a drastic decline in the overall rate of initiation, heat shock messengerRNAs might be preferentially translated by virtue of out-competing normalcellular messages for the limiting initiations. This possibility was eliminated bymeasuring the actual rates of ribosome initiation in heat shocked cells. The rateswere found to compare favourably with those of rabbit reticulocytes, one of thehighest known initiation rates in a eucaryotic cell (Lindquist, 19806). The changein protein synthesis which occurs in Drosophila cells immediately after heatshock involves a highly specific system of translational control which stronglyinhibits the translation of pre-existing messenger RNAs and promotes the rapidand efficient translation of heat shock messages.

Yeast cells have no comparable translational mechanism. The gradual declinein normal cellular protein synthesis during heat shock parallels a gradualdepletion of normal cellular messages from the cell. RNAs isolated at varioustimes after a shift to 39 °C were analysed by in vitro translation and Northern blothybridization. In contrast with results from Drosophila cells, the patterns of

152 S. LINDQUIST

in vitro protein synthesis closely matched the in vivo labellings. Hybridization ofindividual message species with cloned genomic probes indicated that the declinein translation was due to the loss of message sequences from the cell. As isapparent in Fig. 2, each individual messenger RNA appears to have its own rateof disappearance. The most likely explanation for this effect is a combination oftranscriptional inhibition and varying intrinsic rates of message turnover.(Messenger RNA half-lives in yeast have previously been shown to vary between4 and 90min with an average of 20 to 25 min.)

It would certainly be a mistake to assume that temperature has no effect on thetranslational efficiency of yeast messenger RNAs. Indeed, McLaughlin and hiscolleagues (Plesset, Foy, Chia & McLaughlin, 1983) have shown that pre-existing messenger RNAs differ in relative translational efficiencies at low andhigh temperatures. It is clear, however, that yeast cells have not evolved the sortof translational mechanism observed in Drosophila cells in which pre-existingmessages are blocked from translation and preserved for later use. This dif-ference in regulation makes biological sense. In Drosophila cells messengerRNAs typically have half-lives on the order of 6 to 9 h. Furthermore, in healthyand rapidly growing cells the majority of ribosomes are already engaged inprotein synthesis. Even with a massive change in transcription, then, it wouldtake several hours to achieve a high level of heat shock protein synthesis withouta special mechanism to clear out the competition. Yeast cells do not have thisproblem. With message half-lives on the order of 20 to 30 min high levels of heatshock protein synthesis can be achieved by simply allowing pre-existing messagesto decay at their own rates.

Another difference in regulation between Drosophila and yeast cells is in thelevel of transcriptional induction. In Drosophila, messenger RNA for hsp70 ispresent in extremely low concentrations in normal healthy cells. After one hourof heat shock, the concentration has increased by approximately three orders ofmagnitude, reaching a final concentration of several thousand molecules per cell.(See Fig. 3.) In yeast cells, on the other hand, messenger RNA for hsp70 ispresent in substantial quantities at normal temperatures. The increase in con-centration with heat shock is less than ten-fold.

The significance of this difference is not yet clear. Both Drosophila and yeastcells contain a family of proteins which are closely related to hsp70 but con-stitutively expressed (Ingolia & Craig, 1982). It is likely, although by no meanscertain, that these proteins have some overlap in function. If so, the balance offunction between the heat induced proteins and their constitutively synthesizedcognates may be different in yeast and Drosophila, accounting for the differentlevels of hsp70 inducibility. An answer to this question must await quantitativeanalysis of cognate protein expression and biochemical or genetic characteriza-tion of function.

In Fig. 4, yet another important difference between yeast and Drosophila isillustrated. Yeast cells growing by fermentative metabolism have the capacity to

Heat shock - a comparison of Drosophila and yeast 153

25 °C

30min37°C

10min37°C+ act. T = 0

5min37°C

60min37°C

10min37°C

60min37°C+ act 0 T = 40

Fig. 3. Effect of heat shock on polysome profiles. Small flasks of tissue culture cellswere submerged in a 37 °C water bath for the indicated times. Actinomycin wasadded (1 /ig/ml) before heat shock (F and G) or after 60min. Cytoplasmic lysateswere analysed on 0-5 to l-5M-sucrose gradients.

return to normal patterns of protein synthesis when maintained at high tem-perature. Yeast cells are able to grow by both fermentation and respiration,depending upon the carbon source available. In acetate medium they growperforce by respiration, but when supplied with dextrose, they prefer to ferment.Under standard laboratory culture conditions, and presumably during commer-cial fermentation, the yeast heat shock response is transient, and cells willresume growth at temperatures as high as 39-40 °C. Drosophila cells and yeastcells growing by respiratory metabolism do not have this capacity.

154 S. LINDQUIST

Fermentative growthFig. 4. Transient heat shock in yeast cells. Small aliquots of yeast cells grown to mid-log phase at 25 °C in dextrose medium were immersed in a 39 °C water bath andlabelled with [3H]leucine every 20min.

The basis of the ability to recover is not known. Fermentative metabolism isnot the whole story, since yeast cells deficient in mitochondrial metabolism arealso unable to resume growth at normal temperatures, even when growing bydextrose fermentation. At any rate, transient heat shock, followed by adaptationto high temperatures, is a property of many other organisms besides yeast. Itseems likely that in some organisms, the heat shock proteins may be involved in

Heat shock - a comparison of Drosophila and yeast 155

adjusting cellular physiology to life at high temperature as well as in protectingorganisms from short-term exposure to extremes.

DEVELOPMENTAL INDUCTION

It has recently been determined that certain of the Drosophila heat shockgenes are induced during the normal course of development, in the absence ofheat. The expression is both tissue and stage specific. Two stages in the lifecycleare involved, puparium formation and oogenesis.

During puparium formation, in both sexes, transcripts for all four of the smallheat shock proteins, hsp22, 23, 26, and 28 are produced (Sirotkin & Davidson,1982). At least one of these, hsp23, is translated into protein (Cheney & Shearn,1983). The RNAs are observed in imaginal wing discs but not in salivary glandsor in fat bodies. The messages first appear in late third instar larvae, shortly afterrelease of the moulting hormone ecdysone and disappear from late pupae.

Heat shock genes are not expressed in earlier larval stages nor in adult males.In adult females, however, messenger RNAs for hsp28, hsp26 and hsp83 areabundantly transcribed in ovarian nurse cells (Zimmerman, Petri & Meselson,1983). Transcripts are undetectable in stage-1 to -7 egg chambers, increasethereafter to reach a maximum at stage 10 and are then passed into the develop-ing oocyte. They remain abundant during the first 3h of embryogenesis anddisappear with formation of the cellular blastoderm.

A distinctive feature of these developmental inductions is that hsp70 is notinduced. In fact, in developing oocytes, hsp70 cannot be induced even with a heattreatment (Zimmerman et al. 1983). In all other cells and tissues this protein isthe most abundantly synthesized species during heat shock. Its presence seemsto be a very sensitive indicator of stress, since it is produced in tissue culture cellsgrowing in anything less than ideal conditions (Velazquez, Sonoda, Bugaisky &Lindquist, 1983). Thus, the developmental induction is very different from atypical stress response. While hsp70 itself is not developmentally induced, oneof its relatives, the 72000 Mr cognate, is strongly induced in developing oocytes(Craig & Palter, personal communication).

The signal for the developmental induction appears to be release of the moult-ing hormone ecdysone. Imaginal wing discs and tissue culture cells treated withecdysone in vitro display up to a 15-fold increase in messenger RN A for the smallheat shock proteins (Ireland & Berger, 1982). The induction follows a typicalphysiological dosage-response curve, with half-maximal induction at 10~8M.

Ecdysone causes profound changes in cell growth and morphology as well. No-tably, cell lines isolated on the basis of resistance to these effects of the hormonealso do not display the hormonal induction of the small hsps (Berger, Vitek &Morgenelli, 1984).

In investigating the induction of heat shock proteins during development inyeast, vegetatively growing cells were transferred into sporulation medium and

156 S. LINDQUIST

RNAs were extracted every few hours. Each sample was analysed for heat shockmessenger RNAs by in vitro translation and Northern blot hybridization.

We were surprised to find that the yeast heat shock proteins showed the sameuncoupled pattern of expression in development as observed in Drosophila.Messenger RNAs for hsp26 and hsp83 are strongly induced; messenger RNA forhsp70 is not. Fig. 5 displays the patterns of hybridization obtained when RNAsisolated at various times during sporulation are electrophoretically separated,transferred to nitrocellulose, and hybridized with radioactively labelled cloneprobes for the heat shock genes. In vitro translation of the same RNAs resultsin strong synthesis of hsp26 and hsp83 but not of hsp70. Although we can't becertain that hsp83 messenger RNA is translated in vivo, experiments withantibodies against hsp26 demonstrate that the protein is abundant within 6h.

25 27 29 31 33 35 37 25 39

Drosophila YeastFig. 5. Induction of hsp70 messenger RNA with heat shock. Drosophila and yeastcells grown at 25 °C to mid-log phase were heat shocked for 1 h at the indicatedtemperatures. Total cellular RNAs were extracted, separated on CH3-Hg agarosegels, transferred to nitrocellulose, and hybridized with a nick-translated plasmidcontaining the Drosophila or yeast hsp70 gene.

Heat shock - a comparison of Drosophila and yeast 157Experiments with another antibody, which recognizes the entire hsp70 and cog-nate family, extends the analogy with the Drosophila oocyte induction stillfurther. Although hsp70 itself is not induced in developing spores at normal

V 2 8 12 24h

26

84

70

Fig. 6. Developmental induction of hsp mRNAs during yeast sporulation. Totalcellular RNAs were extracted from yeast cells during vegetative growth and 2, 4, 8,and 24 h after transfer to sporulation medium. The RNAs were analysed as for Fig.5 except that in addition to the gene for hsp70, they were also hybridized with thegenes for yeast hsp26 and hsp83.

EMB 83S

158 S. LINDQUIST

temperatures, the 72000 cognate gene is induced (Kurtz, Rossi & Lindquist,manuscript in preparation).

We do not yet know what the role of the heat shock proteins is in development.Yeast spores, of course, are characterized by high tolerance to heat. The earlyDrosophila embryo, however, is the most heat-sensitive stage in the life cycle(Graziosi etal. 1980), indicating that in this case, at least, the proteins are servingsome function specific to development. The recent finding that hsp26 is an RNA-binding protein, taken together with the fact that both Drosophila embryos andyeast spores store large quantities of maternal message, leads to the conjecturethat this protein may be involved in message storage. We have recently createddisruption mutants of hsp26 in yeast which will allow us to test this hypothesis.

An equally interesting question is why, of all the heat shock proteins, hsp70is specifically excluded from developmental induction. Here, too, the answermay lie in its distinctive nucleic-acid-binding properties. We have found that theprotein is a single-stranded nucleic-acid-binding protein which concentrates innuclei and binds to chromosomes in a specific way (Velazquez, DiDomenico &Lindquist, 1980; Velazquez & Lindquist, 1984). It may be that the presence ofthe protein in meiotic cells would poison some normal nuclear process, such asDNA replication or recombination. This might explain why hsp70 is quan-titatively removed from nuclei during recovery from a standard heat shock.Placing the hsp70 coding sequence under control of the hsp26 promoter, toactivate it during sporulation, should provide an answer to this question.

CONCLUSIONS

Over the past several years the heat shock response has been the subject ofintense investigation. The speed of the induction, the intensity of the response,and the reproducibility of the effect have combined to make it an excellent modelsystem for studying gene expression in an amazing variety of different organisms.

One lesson that has been learned is that different cells and organisms usedifferent means to achieve the same ends. The point has been illustrated herewith respect to heat shock protein production in Drosophila and yeast. Anotherinteresting example was provided by Bienz & Gurdon (1982) in studies ofXenopus. In somatic cells heat shock induction is regulated by a combination oftranscriptional and translational control, much as it is in Drosophila cells. Inoocytes, however, the responses is regulated entirely at the level of translation.Heat shock messenger RNAs are already present in Xenopus oocytes but theyare not translated efficiently unless the cells are exposed to heat. This againmakes perfect biological sense. The oocyte is so large that, if it had to dependupon new transcription from a single haploid genome, it would take an inor-dinate amount of time to mount an effective response.

Another outcome of these comparative studies is the discovery of amazinglysimilar developmental patterns of heat shock gene expression in Drosophila

Heat shock - a comparison 0/Drosophila and yeast 159oogenesis and yeast sporulation. The finding has several interesting implications.First, it underscores the fundamental biological similarity of the two processesin these vastly different organisms. Second, it suggests that the developmentalinduction of these proteins may be as ancient a phenomenon as the heat induc-tion; their roles in development may be as important, on an evolutionary scale,as their roles in thermotolerance. Third, the fact that in both organisms thesynthesis of hsp70 has been uncoupled from the other heat shock proteins in-dicates that there has been a strong selection to evolve independent regulatoryelements. It may be that hsp70 is toxic to some fundamental meiotic process.

Our two major goals in studying the heat shock response are to discover howthese proteins confer protection from the toxic effects of stress and to exploittheir induction as a model system for investigating mechanisms of gene regula-tion. Both yeast and Drosophila have unique advantages for these studies. Yeastoffers haploid and diploid genetics as well as the powerful advantage ofhomologous transformation. Drosophila offers a richer and more complex pat-tern of differentiation and development together with the extraordinary cytologyof polytene tissues. The benefits of studying these organisms individually areenhanced by their comparison. The ways in which the responses differ illustratethe principles by which regulatory mechanisms are honed to particular biologicalconstraints. The ways in which they are similar provide perspective on the func-tion and evolution of the response.

REFERENCESBARDWELL, J. C. A. & CRAIG, E. A. (1984). Major heat shock gene of Drosophila and the

Escherichia coli heat inducible dnaK gene are homologous. Proc. natn. Acad. Sci., U.S.A.81, 848-852.

BERGER, E., VITEK, M. & MORGENELLI, C. M. (1984). Genetic reprogramming of Drosophilacells in culture by ecdysterone. In Invertebrate Systems in Vitro 11, (ed. E. Kurstak, H.Oberlander & L. A. Dubendorfer). Amsterdam: Elsevier Press.

BIENZ, M. & GURDON, J. (1982). The heat-shock response in Xenopus oocytes is controlledat the translational level. Cell 29, 811-819.

BONNER, J. J. (1981). Induction of Drosophila heat-shock puffs in isolated polytene nuclei.Devi Biol. 86, 401-414.

CHENEY, C. M. & SHEARN, A. (1983). Developmental regulation of Drosophila imaginal discproteinsd: synthesis of a heat shock protein under non-heat shock conditions. Devi Biol. 95,325-330.

DANIELS, C. J., MCKEE, A. H. Z. & DOOLITTLE, W. F. (1984). Archaebacterial heat shockproteins. EMBO Journal 3, 745-749.

DIDOMENICO, B. J., BUGAISKY, G. E. & LINDQUIST, S. (1982a). Heat shock and recovery aremediated by different translational mechanisms. Proc. natn. Acad. Sci., U.S.A. 79,6181-6185.

DIDOMENICO, B. J., BUGAISKY, G. E. & LINDQUIST, S. (19826). The heat shock response is selfregulated at both the transcriptional and post-transcriptional levels. Cell 31, 593-603.

GUIDICE, G., ROCCHERI, M. C. &DEBERNARDO, M. G. (1980). Synthesis of heat shock proteinsin sea urchin embryos. Cell Biol. Int. Reports 4, 69-73.

GRAZIOSI, G., MICALI, F., MAZANI, R., DECRISTINI, F. & SAVIONI, A. (1980). Variability ofresponse of early Drosophila embryos to heat shock. /. exp. Zool. 214,141-145.

160 S. LINDQUIST

INGOLIA, T. D. & CRAIG, E. (1982). Drosophila gene related to the major heat shock-inducedgene is transcribed at normal temperatures and not induced by heat. Proc. natn. Acad. ScL,U.S.A. 79, 525-529.

INGOLIA, T. D., SLATER, M. & CRAIG, E. A. (1982). Yeast contains a multigene family relatedto the major heat shock inducible gene of Drosophila. Mol. Cell. Biol. 2, 261-21 A.

IRELAND, R. C. & BERGER, E. M. (1982). Synthesis of low molecular weight heat shockproteins stimulated by ecdysterone in Drosophila tissue culture cells. Proc. natn. Acad. ScL,U.S.A. 79, 855-859.

KEENE, M., CORCES, V., LOWENHAUPT, K. & ELGIN, S. C. R. (1981). DNAse 1 hypersensitivesites in Drosophila chromatin occur at the 5' ends of regions of transcription. Proc. natn.Acad. ScL, U.S.A. 78, 143-146.

KEY, J. L., LIN, C. Y. & CHEN, Y. M. Heat shock proteins of higher plants. Proc. natn. Acad.ScL, U.S.A. 78, 3526-3530.

LINDQUIST, S. (1980a). Varying patterns of protein synthesis in Drosophila: Implications forregulation. Devi Biol. 77, 463-469.

LINDQUIST, S. (19806). Translational efficiency of heat induced messages in Drosophilamelanogaster cells. /. mol. Biol. 137, 151-158.

LINDQUIST, S. (1981). Regulation of protein synthesis during heat shock. Nature 294,311-314.LINDQUIST, S., DIDOMENICO, B., BUGAISKY, G., KURTZ, S., PETKO, L. & SONODA, S. (1982).

Heat-shock regulation in Drosophila and yeast. In Heat Shock from Bacteria to Man. ColdSpring Harbor, New York: Cold Spring Harbor Press.

LODISH, H. F. (1976). Translational Control. Ann. Rev. Biochem. 45, 39-72.LOOMIS, W. F. & WHEELER, S. (1980). Heat shock response of Dictyostelium. Devi Biol. 79,

399-408.MCALISTER, L., STRAUSBERG, S., KULAGA, A. & FINKELSTEIN, D. B. (1979). Altered patterns

of protein synthesis inducted by heat shock in yeast. Curr. Genet. 1, 64-73.MCALISTER, L. & FINKELSTEIN, D. B. (1980). Alterations in translatable ribonucleic acid after

heat shock of Saccharomyces cerevisiae. J. Bacteriol. 143, 603-619.MCKENZIE, S. L. (1977). Translational control of heat shock protein synthesis. /. Cell Biol.

75, (2/2), 336A.MIRAULT, M. E., GOLDSCHMIDT-CLEMONT, M., MORAN, L., ARRIGO, A. P. & TISSIERES, A.

(1978). The effect of heat shock on gene expression in Drosophila melanogaster. ColdSpring Harbor Symp. quant. Biol. 42, 819-827.

PELHAM, H. (1982). A regulatory upstream promoter element in the Drosophila hsp 70 heat-shock gene. Cell 30, 517-528.

PELHAM, H. & BIENZ, M. (1982). A synthetic heat-shock promoter element confers heat-inducibility on the herpes simplex virus thymidine kinase gene. EMBO Journal 11,1473-1477.

PLESSET, J., FOY, J. J., CHIA, L. L. & MCLAUGHLIN, C. S. (1983). In Transcription/Transla-tional regulation of gene expression, (ed. M. Grunberg-Manago & B. Safer). Amsterdam:Elsevier North Holland.

SCOTT, M. & PARDUE, M. L. (1981). Translational control in lysates of Drosophila melano-gaster cells. Proc. natn. Acad. ScL, U.S.A. 78, 3353-3358.

SIROTKIN, K. & DAVIDSON, N. (1982). Developmentally regulated transcription fromDrosophila melanogaster chromosomal site 67B. Devi Biol. 89, 196-210.

STORTI, R., SCOTT, M., RICH, A. & PARDUE, M. L. (1980). Translational control of proteinsynthesis in response to heat shock in D. melanogaster cells. Cell 22, 825-834.

SCHLESINGER, M. J., ASHBURNER, M. & TISSIERES, A. (1982). Heat Shock from Bacteria toMan. Cold Spring Harbor, New York: Cold Spring Harbor Press.

THOMAS, G. P., WELCH, W. J., MATHEWS, M. B. & FERAMISCO, J. R. (1982). Molecular andcellular effects of heat shock and related treatments on mammalian tissue culture cells. ColdSpring Harbor Symp. quant. Biol. 46, 985-996.

TOPOL, J. & PARKER, C. (1984). A Drosophila RNA polymerase 11 transcription factor bindsto the regulatory site of an hsp 70 gene. Cell 37, 273-283.

VELAZQUEZ, J. M., DIDOMENICO, B. J. & LINDQUIST, S. (1980). Intracellular localization ofheat shock proteins in Drosophila. Cell 20, 679-690.

Heat shock - a comparison of Drosophila and yeast 161VELAZQUEZ, J. M., SONODA, S., BUGAISKY, G. & LINDQUIST, S. (1983). Is the major

Drosophila heat shock protein present in cells that have not been shocked? /. Cell Biol. 96,286-290.

VELAZQUEZ, J. M. & LINDQUIST, S. (1984). hsp70: nuclear concentration during environment-al stress and cytoplasmic storage during recovery. Cell 36, 655-662.

Wu, C. (1980). The 5' ends of Drosophila heat shock genes in chromatin. Nature 286,854-859.

Wu, C. (1984). Two protein-binding sites in chromatin implicated in the activation of heatshock genes. Nature (in press).

ZIMMERMAN, J. L., PETRI, W. & MESELSON, M. (1983). Accumulation of a specific subset ofD. melanogaster heat shock mRNAs in normal development without heat shock. Cell 32,1161-1170.