Comparative Biochemistry and Physiology Part B 136(2003) 275–282

1096-4959/03/$ - see front matter� 2003 Elsevier Inc. All rights reserved.doi:10.1016/S1096-4959(03)00205-7

Heat- and cold-inducible regulation of HSP70 expression inzebrafish ZF4 cells

Susanna Airaksinen*, Terhi Jokilehto , Christina M.I. Rabergh , Mikko Nikinmaa1 2˚

Department of Biology, Laboratory of Animal Physiology, University of Turku, Turku FIN-20014, Finland

Received 19 December 2002; received in revised form 30 April 2003; accepted 30 June 2003

Abstract

Elevated temperature induces a rapid heat shock transcription factor(HSFs)-mediated expression of heat shock(hsp)genes. The effect of cold exposure onhsp gene expression has hardly been investigated, although ectothermic animalsexperience both cold and heat stress. We have previously shown in zebrafish that the expression ofhsf1a and a uniqueisoform hsf1b vary in a tissue-specific manner upon heat stress. In the current study, using a zebrafish(Danio rerio)embryonic cell line(ZF4), we have compared the effects of heat shock(28™37 8C) vs. cold shock(28™20 8C) onthe expression ofahsf1a, zhsf1b and hsp70. Concomitantly, the suitability of the ZF4 cells as a model system wasverified. The expression pattern of HSP70 proteins following heat or cold exposure is distinct, and the total HSP70 levelis upregulated or stable, respectively. Moreover, heat exposure specifically increases the ratio ofzhsf1ayb expression(10-fold), whereas cold exposure decreases it to one half. These data suggest that thezhsf1ayzhsf1b ratio is regulated ina temperature-dependent manner, and the ratio may be indicative of the stressor-specific HSP70 expression. Furthermore,the response in ZF4 cells upon heat shock resembles the response observed in zebrafish liver and thus, supports the useof this cell line in stress response studies.� 2003 Elsevier Inc. All rights reserved.

Keywords: Heat shock; Cold shock; Temperature; Heat shock protein(HSP); Heat shock factor(HSF); Zebrafish;Danio rerio; ZF4

1. Introduction

Thermal stress among other stressors triggers acellular program called the heat shock response,i.e. elevated expression of specific classes of pro-teins known as heat shock proteins(HSPs). HSPs

*Corresponding author.Present address: Finnish Game andFisheries Research Institute, Turku Game and FisheriesResearch, Itainen Pitkakatu 3, Turku FIN-20520, Finland. Tel.:¨ ¨q358-205-751-688; fax:q358-205-751-689.

E-mail address: susanna.airaksinen@rktl.fi(S. Airaksinen).Current address: Turku Centre for Biotechnology, Univer-1

sity of Turku, Abo Akademi University, P.O. Box 123, FIN-˚20521 Turku, Finland.

Current address: Schering Oy, Pansiontie 47, FIN-20101,2

Finland.

are associated with many cellular processes includ-ing protein synthesis, folding and translocation aswell as assembly of larger protein complexes, allof which can be impaired upon stress(Lindquistand Craig, 1988). HSFs(heat shock factors) aretranscription factors that regulate the expression ofHSPs through interaction with a specific heatshock element(HSE) in the promoter region ofthe hsp genes(for review, see Morimoto, 1998).Four members of the HSF family have beendescribed in vertebrates, whereas inSaccharomy-ces cerevisiae and in Drosophila melanogaster asingle HSF, HSF1 homolog, appears to be suffi-cient for the regulation of the heat shock responseboth in stressful and non-stressful conditions

276 S. Airaksinen et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 275–282

(Wiederrecht et al., 1988; Sorger and Pelham,1988; Jedlicka et al., 1997; Pirkkala et al., 2001).Several HSF members have multiple isoforms.

Two HSF1 isoforms exist in mouse(Fiorenza etal., 1995; Goodson and Sarge, 1995) and in aparasitic helminthSchistosoma mansoni (LevyHoltzman and Schechter, 1996). According to theexpression patterns, the isoforms of HSF1 havebeen suggested to participate in the tissue-specificregulation in mouse and in the developmentalstage-specific regulation inS. mansoni. However,the exact function of these isoforms has not beenelucidated. Two distinct HSF2 isoforms(Fiorenzaet al., 1995; Goodson et al., 1995) participate inhemin-mediated erythroid differentiation of K562cells with opposing effects(Leppa et al., 1997).¨The HSF2b isoform functions as a negative regu-lator, whereas overexpression of HSF2a can coun-teract the inhibitory effect of HSF2b and potentiatethe transcriptional activity of HSF2 thereby allow-ing differentiation(Leppa et al., 1997). Also, the¨isoforms of mammalian HSF4, HSF4a and HSF4bact either as transcriptional repressors or activators,respectively(Tanabe et al., 1999; Zhang et al.,2001).Elevated temperature is a well-known inducer

of the heat shock response. However, the role ofcold shock is still controversial. Although coldshock has been intensively studied in bacteria andplants, the investigation in higher organisms hasreceived minor attention, except for its applicationin cryosurgery and cold preservation(Thieringeret al., 1998; Fujita, 1999; Phadtare et al., 1999;Sonna et al., 2002). Eurythermic and especiallyectothermic animals, such as fish, may encounterlarge fluctuations in their body temperature. Ingeneral, a decrease in temperature is as frequentlyexperienced as an increase in temperature. To ourknowledge, cold exposure in respect to heat shockresponse has not been addressed in fish. On thewhole, the research on the regulation of HSPexpression in fish has only recently been initiated(for review, see Basu et al., 2002).In our previous studies, we have focused on the

regulation of the heat shock response in zebrafishand shown that there are two isoforms ofhsf1,zhsf1a and zhsf1b (Rabergh et al., 2000). zhsf1b˚is a unique isoform found only in zebrafish. Theratio of thezhsf1 isoforms altered upon heat stressin a tissue-specific manner(Rabergh et al., 2000).˚However, no change was observed in the ratio ofthe zhsf1 isoforms in response to heavy metal

exposure even though the target gene,hsp70 wasupregulated, indicating a stressor-specific patternof zhsf1 expression(Airaksinen et al., 2003). Inthe present study, we have focused on the zebrafishcell line (ZF4) in order to evaluate its suitabilityas a model system for studying the expression ofHSP70 and its regulatory mechanisms. We studiedthe effects of heat vs. cold shock and investigatedwhether the change inzhsf1 ratio is regulated in atemperature-dependent manner. This was accom-plished by determining the ratio of thezhsf1isoforms as well as the level ofhsp70 mRNA andprotein upon exposures.

2. Materials and methods

2.1. Cell culture and temperature treatments

The studies were performed with the zebrafish(Danio rerio) embryonic fibroblast cell line, ZF4(ATCC CRL 2050; Driever and Rangini, 1993).The cells were grown to confluency at 288C, 5%CO , in Dulbecco’s modified Eagle’s mediumy2

F12 nutrient mix (DMEMyF12) supplementedwith 10% FBS, 1%L-glutamine, 100 uyml peni-cillin, 100 mgyml streptomycin and 0.25mgymlFungizone. Heat(37 8C) and cold(20 8C) shocktreatments were performed by incubating the cellsin a water bath(Lauda K4R Electronic, Germany)for 1, 2, 4 or 6 h. Control cells were incubated ina water bath at 288C. Each sample is representa-tive of cells grown to semi-confluency in a cellculture bottle. All treatments were replicated sep-arately for 5 independent times. All cell culturemedium components were purchased from GibcoBRL Life Technologies Ltd(USA).

2.2. Immunoblotting for HSP70

Whole cell extracts were prepared from cells asdescribed by Mosser et al.(1988). Protein samples(20 mg) were then separated on 8% SDS-poly-acrylamide gels and transferred to nitrocellulosemembranes(Schleicher and Schuell, Keene, NH,USA) using a semi-dry transfer apparatus(Bio-Rad, CA, USA). Membranes were blocked for 1h in 3% non-fat dry milk in PBS with 0.3% Tween20. Both the constitutive and inducible forms ofHSP70 were detected using a monoclonal mouseanti-HSP70 antibody(dilution 1:10 000; clone 3a3,Affinity Bioreagents, Golden, CO, USA). Horse-radish-peroxidase-conjugated rabbit anti-mouse

277S. Airaksinen et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 275–282

Fig. 1. Representative western blot analysis of HSP70 expression in ZF4 cells(28 8C). HSP70 family members were detected after 1,2, 4 or 6-h heat shocks at 378C (a) and cold shocks at 208C (b) as indicated above the panels. C denotes an untreated control sample.

immunoglobulin(Amersham Life Sciences, Buck-inghamshire, UK) was used as a secondary anti-body, and the signals were visualized using anenhanced chemiluminescence method according tothe manufacturer’s instructions(Amersham). Thewestern images were captured using a Chemi-imager(Alpha Innotech Corporation, San Leandro,CA, USA). Samples from five heat shock experi-ments and four cold shock experiments were ana-lyzed and the results evaluated visually.

2.3. Reverse transcription-polymerase chain reac-tion (RT-PCR) of zhsf1a and zhsf1b mRNAs

Total RNA was isolated from the cells using theRNAzolB� kit (TEL-TEST Inc., Friendswood,TX, USA). The cDNA synthesis was performedwith 5 mg total RNA isolated from the ZF4 cellsat 428C with oligo(dT) primers(Invitrogen, Carls-bad, CA, USA) and AMV (Avian myeloblastosisvirus) reverse transcriptase(Finnzymes, Espoo,Finland). Expression levels were analyzed usingPCR with specific primers(Rabergh et al., 2000)˚amplifying a 605 bp fragment of thezhsf1a and a683 bp fragment of thezhsf1b, a 457 bp fragmentof the partial zebrafishhsp70 sequence availablein the GeneBank(Lele et al., 1997) and a 411 bpfragment of the zebrafishhsc70 sequence availablein the GeneBank(Graser et al., 1996). PCRreactions were optimized to amplify bothzhsf1isoforms within a sample. Therefore, an annealingtemperature of 658C and 35 cycles of PCRamplification were used(PCRExpress, Hybaid,London, UK). To avoid variability between ampli-fication reactions bothhsp70 and hsc70 weresimultaneously amplified.

2.4. Statistical analysis

RT-PCR data from 4(cold shock) to 5 (heatshock) independent samples were analyzed using

a Chemi-imager with the AlphaEase� software(Alpha Innotech Corp.). The statistical signifi-cance of the differences relative to controls wastested with a pairedt-test(PF0.05).

3. Results

3.1. HSP70 isoforms are differently induced byheat and cold exposure

HSP70 protein levels were markedly elevatedalready after a 1-h heat shock at 378C (Fig. 1a).The HSP70 continued to accumulate at 2, 4 and 6h as was observed in five separate experiments.There appeared to be several heat-inducible HSP70isoforms. One isoform showed a slightly highermolecular weight than the other constitutively andinducibly expressed counterparts and was observedafter 6-h heat shock. The lower molecular weightfraction of HSP70 was especially induced inresponse to heat shock already at the earlier timepoints. However, the exact number of the HSP70isoforms cannot be determined using one-dimen-sional electrophoresis. It is known that HSC70 ispresent also during heat shock(Santacruz et al.,1997). However, we were not able to distinguishit from other isoforms due to its close proximityto the high-intensity HSP70 bands, and due to thefact that the used antibody crossreacts with severalHSP70 family members(Affinity Bioreagents).Thus, it remains unknown which proportion of theobserved expression is due to upregulation of theconstitutively expressed members of HSP70. How-ever, we cannot exclude the possibility that someadditional HSP70 isoform(s) remained unidenti-fied by the used antibody.Upon cold shock(20 8C), the induction of

HSP70 was mitigated at the total HSP70 level.However, a new isoform was observed after 4 and6 h (Fig. 1b). Furthermore, the upregulated HSP70isoform had a higher molecular weight than the

278 S. Airaksinen et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 275–282

Fig. 2. Analysis of expression patterns ofzhsf1a, zhsf1b, hsp70 andhsc70 mRNA by RT-PCR.(a) ZF4 cells(28 8C) were heat-shockedat 378C for 1, 2, 4 or 6 h as indicated above the panels. C denotes an untreated control sample. The positions of the twozhsf1 isoforms,a andb, are indicated next to the panel.(b) ZF4 cells(28 8C) were cold-shocked at 208C as in panel a.(c) The levels ofhsp70 aftertemperature treatment are presented in relation tohsc70 levels obtained by densitometric analysis(d). The ratio ofhsf1 isoforms aftertemperature treatments obtained by densitometric analysis. Values represent mean"S.E. of five independent experiments in heat shockand four independent experiments in cold shock and asterisks denote a significant effect of temperature compared to controls(*P-0.05, **P-0.01, ***P-0.001).

constitutively expressed isoform(s) (Fig. 1b). Dur-ing the cold exposure the expression of the consti-tutive form of HSP70 was constant, until at 6 hwhen the expression of the induced HSP70 becamedominating (Fig. 1b). In some of the cases, theshift between the expressed isoforms was not asevident. However, in contrast to heat shock, wherethe total level of HSP70 was prominentlyincreased, the total HSP70 expression during coldshock exposure appeared more stable. Althoughthe HSP70 isoforms detected upon heat and coldstress overlapped, the observed expression patternwas distinct. In contrast to cold shock, also thelower molecular weight isoform(s) of HSP70 washighly inducible upon heat stress. Northern blotanalysis using humanhsp70 as a probe showed atransient increase in accumulation ofhsp70 mRNAupon heat shock, whereas it remained unchangedupon cold shock(data not shown).

3.2. The expression of zhsf1 isoforms shows tem-perature sensitivity

RT-PCR data demonstrate a significant inductionof hsp70 and a concurrent change in the ratio ofzhsf1 isoforms upon heat shock(Fig. 2a), whichis consistent with our previous studies in vivo(Rabergh et al., 2000; Airaksinen et al., 2003). An˚upregulation ofhsf1a upon a 2 to 6-hheat exposure(Fig. 2a, upper panel), though more prominenthere, corresponds to the change observed in zebraf-ish liver (Rabergh et al., 2000). hsf1b starts to˚disappear already at 1 h and it becomes virtuallyundetectable by 6 h. No significant induction ofhsp70 occurred upon cold exposure(Fig. 2b),which is consistent with the observations in north-ern analysis(data not shown). However, there wasa change in the ratio ofzhsf1ayhsf1b followingcold shock but it was opposite to that observed

279S. Airaksinen et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 275–282

upon heat stress, i.e.hsf1b, and not hsf1a, wasupregulated at the 6-h time point(Fig. 2b, upperpanel).The densitometric analyses of the RT-PCR

results are shown in Fig. 2c and d. Following heatshock, thehsp70 expression is significantly ele-vated at 1 h and tends to be upregulated duringthe entire exposure, whereashsp70 expressionfollowing cold shock remains unchanged(Fig. 2c).The PCR primers that were used forhsp70 amplifya region within the ATPase domain, which is wellconserved between different isoforms ofhsp70.All the currently knownhsp70 andhsc70 sequenc-es of zebrafish, should thus be amplified with theused primers. The changes in the ratio ofhsf1isoforms are visualized in Fig. 2d. The heat shockresults in a 10-fold increase in the ratio ofhsf1ayhsf1b during the exposure, due to downregulationof the hsf1b and upregulation of thehsf1a isoform(black bars). The ratio is even greater at the 4-htime point. During the cold shock, the ratiodecreases gradually from a value of 1–0.5 mostlydue to downregulation ofhsf1a isoform (greybars). The decrease becomes significant at 4 and6 h. Although, the quantity ofhsp70 inductioncannot be exactly determined by RT-PCR, theobserved shift in thehsf1 isoforms is clearlydemonstrated and is unambiguously caused by thedifferent temperature exposures.

4. Discussion

4.1. Heat vs. cold shock

As the nomenclature indicates the heat shockresponse is readily induced by elevated tempera-tures. Interestingly, many of the physiologicaleffects of cold exposure that are observed oncellular level are similar to those seen in heat-stressed cells, including decelarated protein synthe-sis and cell cycle progression, reduced membranepermeability and cytoskeletal structure changes, aswell as protein denaturation(Sonna et al., 2002).However, there are few studies, either in vitro orin vivo that have focused on low temperatures.The number of available studies becomes evensmaller when studies resembling our experiment,in which cold exposure without recovery periodsat higher temperatures are considered. An earlystudy withDrosophila suggested that the inductionof HSPs following cold exposure(0 8C) was dueto the shift from 08C back to the control temper-

ature of 258C, i.e. heat shock, rather than to thecold treatment itself(Burton et al., 1988). Also,in a mammalian study the induction after coldtreatment(4 8C) occurred upon recovery at controltemperature(37 8C). The magnitude and the kinet-ics of the response were, however, related to theduration of cold stress(Liu et al., 1994). Further-more, the same study suggested that phosphoryla-tion and trimerization of HSF1 is involved in thecold-inducible stress response in human cells. Incontrast to heat shock, no hyperphosphorylation ofHSF1 was observed in mice exposed to cold invivo (Sarge et al., 1993; Cullen and Sarge, 1997).Brown adipose tissue is a specific tissue, which

becomes metabolically active upon cold exposure.In this tissue in mouse, the cold ambient temper-ature inducedhsp70 and increased HSP70 proteinlevel within an hour of cold exposure(Matz et al.,1995). This is in contrast to our study were totalhsp70 mRNA and protein levels remained stablefollowing cold shock. In midge,Culicoides vari-ipennis sonorensis, a subset of proteins that wereinduced by heat and cold were the same, but alsoadditional proteins were induced in response tocold (Nunamaker et al., 1996). It is possible thatectothermic animals may have an enhanced regu-latory machinery to cope with temperature changescompared to euthermic animals, which are morelikely to be subject to thermal stress by elevatedtemperature(fever) rather than both heat and cold.For example, a specific family of small cold-inducible proteins is present in prokaryotes(for areview, see Thieringer et al., 1998).

4.2. Zebrafish HSF1

The definite role ofhsf1a andhsf1b isoforms inzebrafish is still unknown. In this study, coldexposure led to a lowerhsf1ayb ratio with con-comitant stability of HSP70 expression, whereasheat shock led to a higher ratio ofhsf1ayb togetherwith HSP70 induction. The heat-inducible shift isconsistent with our previous findings in zebrafishin vivo (Rabergh et al., 2000). When zebrafish˚were exposed to heavy metals at constant temper-ature, we did not observe any change in the ratioof hsf1 isoforms even though HSP70 expressionwas slightly increased(Airaksinen et al., 2003).These results indicate that zHSF1b might act byinhibiting the expression of HSP70 upon coldstress. Hypothetically, the inhibition could be

280 S. Airaksinen et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 275–282

mediated either directly or via an interaction withzHSF1a.Alternative splicing is a plausible explanation

for the multiple zhsf1 mRNAs. RNA splicing assuch is a flexible process. It has been regarded asa tool for evolution of genes and organisms. Morethan 50 protein or RNA molecules are involved inthe splicing event and, therefore, it would not besurprising if one or several of them could beregulated by a temperature shift leading to amodified ratio of mRNA products. A pure temper-ature effect on splicing would suggest identicalregulation in different tissues and cells. Cold-inducible alternative splicing has been recentlyreported in addition to potato also in human(Bournay et al., 1996; Ars et al., 2000). In neu-rofibromatosis type I(NFI) an alternatively splicedform in response to cold exposure was observed.Other stressors did not induce the alternative splic-ing and the splicing did not vary in different cells(Ars et al., 2000). This is in contrast to our study,in which high temperature exposure caused tissue-specific change in the ratio ofzhsf1 isoforms(Rabergh et al., 2000). Thus, a regulated change˚in splicing specificity is suggested in zebrafish.So far fish heat shock transcription factors have

been cloned only from zebrafish(Rabergh et al.,˚2000). Wang et al.(2001) have reported a thirdisoform of hsf1 in zebrafish. This isoform calledhsf1c is expressed throughout the early embryonicdevelopment. Although no direct evidence this farexists to show that all splicedzhsf1 isoforms aretranslated into functional proteins in vivo, zHSF1aand zHSF1c translated in vitro bind to HSE andfurthermore, the binding is heat-inducible(Wanget al., 2001). In the same study, a morpholinooligonucleotide specific for the DNA-bindingdomain, which is common to allzhsf1 isoforms,was used to inhibit the translation ofhsf1. Thereby,it was shown that the heat-induced apoptosis isenhanced by the inhibition of HSF1-mediatedHSP70 expression. This suggests that zHSF1 pro-tects zebrafish against heat-induced apoptosis in asimilar fashion as HSF1 in murine cells(McMillanet al., 1998).According to the current model, in the control

conditions HSF1 exists as an inactive monomerassociated with a multichaperone complex, whichis localized to the cytoplasm(Morimoto, 1998;Zou et al., 1998). It has been postulated that stresscauses a competitive release of HSP70 and HSP90from the monomeric HSF1 due to their high

affinity to other, denatured, proteins. This enablesnuclear translocation of HSF1, followed by trimer-ization and DNA-binding(Morimoto, 1998). Theactivation can further be regulated by phosphory-lation (Sarge et al., 1993; Cotto et al., 1996;Morimoto, 1998; Holmberg et al., 2002). HumanHSF1 has been found to be a target for a stress-inducible sumoylation and the consensus siteinvolved in this modification is conserved also inthe zebrafishhsf1 (Hong et al., 2001; Hietakangaset al., 2003). The role of splice variants has beenimplicated in differential transactivation potentialupon stress and physiological signals, as well asin signal transduction in a cell type-specific man-ner (Levy Holtzman and Schechter, 1996; Leppa¨et al., 1997; Fiorenza et al., 1995; Goodson andSarge, 1995; Tanabe et al., 1999). Additionally,human HSF4a functions as a direct repressor ofHSF1-mediated transcription(Zhang et al., 2001).All above-mentioned functions are also feasible inzebrafish(Rabergh et al., 2000; Wang et al., 2001;˚Airaksinen et al., 2003).In summary, our data indicate that in addition

to the temperature increase, also temperaturedecrease causes a shift in the proportions of thezhsf1 isoforms. The cold-induced change is oppo-site to that observed with elevated temperature andis associated with the unique expression pattern ofHSP70 family members. It will be important tocharacterize the mechanisms by which the ratio ofzhsf1a and zhsf1b isoforms is altered followingtemperature shift, and whetheryhow zHSF1b actsas a suppressor of HSP70 expression. Furthermore,it remains to be elucidated if this would serve tosave energy during the cold exposure or to regulatesome unidentified HSP70 isoforms or other cold-inducible proteins. Finally, the present study showsthat the ZF4 cell line offers a valuable method tostudy the detailed regulation of the heat shockresponse in fish.

Acknowledgments

We thank Lea Sistonen(Turku Centre for Bio-technology, University of Turku, Abo Akademi˚University) for critically reading the manuscript.This work was supported by the Academy ofFinland, projects 40830, 42186 and 50748.

References

Airaksinen, S., Rabergh, C.M.I., Lahti, A., Kaatrasalo, A.,˚Sistonen, L., Nikinmaa, M., 2003. Stressor-dependent regu-

281S. Airaksinen et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 275–282

lation of the heat shock response in zebrafish,Danio rerio.Comp. Biochem. Physiol. A 134, 839–846.

Ars, E., Serra, E., de la, L.S., Estivill, X., Lazaro, C., 2000.Cold shock induces the insertion of a cryptic exon in theneurofibromatosis type 1(NF1) mRNA. Nucl. Acids Res.28, 1307–1312.

Basu, N., Todgham, A.E., Ackerman, P.A., Bibeau, M.R.,Nakano, K., Schulte, P.M., et al., 2002. Heat shock proteingenes and their functional significance in fish. Gene 295,173–183.

Bournay, A.S., Hedley, P.E., Maddison, A., Waugh, R., Mach-ray, G.C., 1996. Exon skipping induced by cold stress in apotato invertase gene transcript. Nucl. Acids Res. 24,2347–2351.

Burton, V., Mitchell, H.K., Young, P., Petersen, N.S., 1988.Heat shock protection against cold stress ofDrosophilamelanogaster. Mol. Cell Biol. 8, 3550–3552.

Cotto, J.J., Kline, M., Morimoto, R.I., 1996. Activation of heatshock factor 1 DNA binding precedes stress-induced serinephosphorylation. J. Biol. Chem. 271, 3355–3358.

Cullen, K.E., Sarge, K.D., 1997. Characterization of hypother-mia induced cellular stress response in mouse tissues. J.Biol. Chem. 272, 1742–1746.

Driever, W., Rangini, Z., 1993. Characterization of a cell linederived from zebrafish(Brachydanio rerio) embryos. InVitro Cell Dev. Biol. Anim. 29A, 749–754.

Fiorenza, M.T., Farkas, T., Dissing, M., Kolding, D., Zimarino,V., 1995. Complex expression of murine heat shock tran-scription factors. Nucl. Acids Res. 23, 467–474.

Fujita, J., 1999. Cold shock response in mammalian cells. J.Mol. Microbiol. Biotechnol. 1, 243–255.

Goodson, M.L., Park-Sarge, O.-K., Sarge, K.D., 1995. Tissuedependent expression of heat shock factor 2 isoforms withdistinct transcriptional activities. Mol. Cell Biol. 15,5288–5293.

Goodson, M.L., Sarge, K.D., 1995. Regulated expression ofheat shock factor 1 isoforms with distinct leucine zipperarrays via tissue-dependent alternative splicing. Biochem.Biophys. Res. Commun. 211, 943–949.

Graser, R.T., Malnar-Dragojevic, D., Vincek, V., 1996. Cloningand characterization of a 70 kd heat shock cognate(hsc70)gene from the zebrafish(Danio rerio). Genetica 98,273–276.

Hietakangas, V., Ahlskog, J.K., Jakobsson, A.M., Hellesuo,M., Sahlberg, N.M., Holmberg, C.I., et al., 2003. Phospho-rylation of serine 303 is a prerequisite for the stress-inducible SUMO modification of heat shock factor 1. Mol.Cell Biol. 23, 2953–2968.

Holmberg, C.I., Tran, S.E., Eriksson, J.E., Sistonen, L., 2002.Multisite phosphorylation provides sophisticated regulationof transcription factors. Trends Biochem. Sci. 27, 619–627.

Hong, Y., Rogers, R., Matunis, M.J., Mayhew, C.N., Goodson,M.L., Park-Sarge, O.K., et al., 2001. Regulation of HSF1by stress-induced SUMO-1 modification. J. Biol. Chem.276, 40263–40267.

Jedlicka, P., Mortin, M.A., Wu, C., 1997. Multiple functionsof Drosophila heat shock transcription factor in vivo. EMBOJ. 16, 2452–2462.

Lele, Z., Engel, S., Krone, P.H., 1997.hsp47 andhsp70 geneexpression is differentially regulated in a stress- and tissue-

specific manner in zebrafish embryos. Dev. Genet. 21,123–133.

Leppa, S., Pirkkala, L., Saarento, H., Sarge, K.D., Sistonen,¨L., 1997. Overexpression of HSF2-beta inhibits hemin-induced heat shock gene expression and erythroid differen-tiation in K562 cells. J. Biol. Chem. 272, 15293–15298.

Levy Holtzman, R., Schechter, I., 1996. Expression of differentforms of the heat-shock factor during the life cycle of theparasitic helminthSchistosoma mansoni. Biochim. Biophys.Acta 1317, 1–4.

Lindquist, S., Craig, E.A., 1988. The heat-shock proteins.Annu. Rev. Genet. 22, 631–677.

Liu, A.Y.C., Bian, H., Huang, L.E., Lee, Y.K., 1994. Transientcold shock induces the heat shock response upon recoveryat 37 8C in human cells. J. Biol. Chem. 269, 14768–14775.

Matz, J.M., Blake, M.J., Tatelman, H.M., Lavoi, K.P., Hol-brook, N.J., 1995. Characterization and regulation of cold-induced heat shock protein expression in mouse brownadipose tissue. Am. J. Physiol. 269, R38–R47.

McMillan, D.R., Xiao, X., Shao, L., Graves, K., Benjamin,I.J., 1998. Targeted disruption of heat shock transcriptionfactor 1 abolishes thermotolerance and protection againstheat-inducible apoptosis. J. Biol. Chem. 273, 7523–7528.

Morimoto, R.I., 1998. Regulation of the heat shock transcrip-tional response: cross talk between a family of heat shockfactors, molecular chaperones and negative regulators. GenesDev. 12, 3788–3796.

Mosser, D.D., Theodorakis, N.G., Morimoto, R.I., 1988. Coor-dinate changes in heat shock element-binding activity andHSP70 gene transcription rates in human cells. Mol. CellBiol. 8, 4736–4744.

Nunamaker, R.A., Dean, V.C., Murphy, K.E., Lockwood, J.A.,1996. Stress proteins elicited by cold shock in the bitingmidge,Culicoides variipennis sonorensis Wirth and Jones.Comp. Biochem. Physiol. B 113, 73–77.

Phadtare, S., Alsina, J., Inouye, M., 1999. Cold-shock responseand cold-shock proteins. Curr. Opin. Microbiol. 2, 175–180.

Pirkkala, L., Nykanen, P., Sistonen, L., 2001. Roles of the heatshock transcription factors in regulation of the heat shockresponse and beyond. FASEB J. 15, 1118–1131.

Rabergh, C.M.I., Airaksinen, S., Soitamo, A., Bjorklund, H.V.,¨˚Johansson, T., Nikinmaa, M., et al., 2000. Tissue-specificexpression of zebrafish(Danio rerio) heat shock factor 1mRNAs in response to heat stress. J. Exp. Biol. 203,1817–1824.

Santacruz, H., Vriz, S., Angelier, N., 1997. Molecular charac-terization of a heat shock cognate cDNA of zebrafish,hsc70, and developmental expression of the correspondingtranscripts. Dev. Genet. 21, 223–233.

Sarge, K.D., Murphy, S.P., Morimoto, R.I., 1993. Activationof heat shock gene transcription by heat shock factor 1involves oligomerization acquisition of DNA-binding activ-ity, and nuclear localization and can occur in the absenceof stress. Mol. Cell. Biol. 13, 1392–1407.

Sonna, L.A., Fujita, J., Gaffin, S.L., Lilly, C.M., 2002. Effectsof heat and cold stress on mammalian gene expression. J.Appl. Physiol 92, 1725–1742.

Sorger, P.K., Pelham, H.R.B., 1988. Yeast heat shock factor isan essential DNA-binding protein that exhibits temperature-dependent phosphorylation. Cell 54, 855–864.

282 S. Airaksinen et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 275–282

Tanabe, M., Sasai, N., Nagata, K., Liu, X.D., Liu, P.C., Thiele,D.J., et al., 1999. The mammalian HSF4 gene generatesboth an activator and a repressor of heat shock genes byalternative splicing. J. Biol. Chem. 274, 27845–27856.

Thieringer, H.A., Jones, P.G., Inouye, M., 1998. Cold shockand adaptation. Bioessays 20, 49–57.

Wang, G., Huang, H., Dai, R., Lee, K.Y., Lin, S., Mivechi,N.F., 2001. Suppression of heat shock transcription factorHSF1 in zebrafish causes heat-induced apoptosis. Genesis30, 195–197.

Wiederrecht, G., Seto, D., Parker, C.S., 1988. Isolation of thegene encoding theS. cerevisiae heat shock transcriptionfactor. Cell 54, 841–853.

Zhang, Y., Frejtag, W., Dai, R., Mivechi, N.F., 2001. Heatshock factor-4(HSF-4a) is a repressor of HSF-1 mediatedtranscription. J. Cell Biochem. 82, 692–703.

Zou, J., Guo, Y., Guettouche, T., Smith, D.F., Voellmy, R.,1998. Repression of heat shock transcription factor HSF1activation by HSP90(HSP90 Complex) that forms a stress-sensitive complex with HSF1. Cell 94, 471–480.