Distinct Stress-Inducible and Developmentally Regulated Heat Shock Transcription Factors...

DEVELOPMENTAL BIOLOGY 181, 47–63 (1997)ARTICLE NO. DB968441

Distinct Stress-Inducible and DevelopmentallyRegulated Heat Shock Transcription Factors inXenopus Oocytes

Sandra Gordon,1 Steve Bharadwaj, Alex Hnatov,Adnan Ali, and Nick Ovsenek2

Department of Anatomy and Cell Biology, College of Medicine, University of Saskatchewan,107 Wiggins Road, Saskatoon, Saskatchewan, Canada S7N 5E5

The presence of a maternal pool of heat shock factor (HSF) in Xenopus oocytes has been suggested by two lines of evidencefrom previous studies. First, heat shock response element (HSE)-binding activity is induced in heat-shocked eggs andembryos prior to expression of zygotic HSF. Second, expression from microinjected heat shock protein promoters in oocytesis induced upon heat shock. To date, however, endogenous oocyte HSF molecules have not been detected, nor has inductionof HSE-binding activity been directly demonstrated. Here we report the detection of distinct stress-inducible and develop-mentally regulated HSE-binding activities of endogenous oocyte factors. Exposure of defolliculated oocytes to heat, cad-mium, and arsenite resulted in the formation of an HSE-specific complex detectable by gel mobility shift assay. Inductionof HSE-binding activity by each of these stressors corresponded to increased expression from a microinjected hsp70 pro-moter. The stress-inducible HSE-binding complex was recognized by antiserum against mammalian HSF1, but not by HSF2antiserum, suggesting that a Xenopus homologue of HSF1 is the major component of this activity. The HSE-binding activityof HSF1 was induced by stress treatments of stage I through VI oocytes, an indication that it is responsive to stressthroughout oogenesis. During recovery from heat shock, the HSF1–HSE complex rapidly declined to control levels, butwas induced for prolonged periods in oocytes exposed to continuous stress, a pattern unlike the transient activationpreviously observed in fertilized eggs or embryos. The kinetics of HSF1 activation in oocytes suggests that a key protein(s)regulating attenuation of the stress response is present at exceedingly low levels or is somehow modified during preembry-onic development. We also detected an unusual constitutive HSE-binding complex in unstressed stage I and II oocytes,but not in later stage oocytes, eggs, developing embryos, or A6 cells. This constitutive complex was unaffected by heat orchemical treatments and was not recognized by either HSF1 or HSF2 antiserum. Appearance of the constitutive HSE-binding activity during oogenesis corresponded closely with peak levels of hsp70 mRNA detected by Northern blot analysisof RNA from staged oocytes. We suggest that the constitutive HSE-binding activity in early oocytes is formed by a uniquedevelopmentally regulated heat shock factor that may play a role in the expression of heat shock proteins during earlystages of oogenesis. q 1997 Academic Press

INTRODUCTION wide variety of stress stimuli, as well as expression in theabsence of stress during development and differentiation,changes in the cell cycle, and exposure to growth factorsEukaryotic cells respond to stress by rapidly increasing(reviewed by Morimoto et al., 1990). In most eukaryotes,the synthesis of a set of heat shock proteins (HSPs) (re-the key biochemical step facilitating stress-induced tran-viewed by Morimoto et al., 1990). Genes encoding the HSPscription of HSPs involves posttranslational activation of afamily of proteins are subject to complex regulatory mecha-constitutively expressed transcription activator, the heatnisms, as illustrated by HSP expression in response to ashock factor (HSF), which binds to highly conserved heatshock response elements (HSEs) present in the upstreamregulatory region of all heat shock genes (Lis and Wu, 1993;1 The first two authors contributed equally to this manuscript.

2 To whom correspondence should be addressed. Morimoto, 1993). Upon exposure to stress, HSF is converted

47

0012-1606/97 $25.00Copyright q 1997 by Academic PressAll rights of reproduction in any form reserved.

AID DB 8441 / 6x15$$$$81 12-09-96 18:37:34 dbas

48 Gordon et al.

from an inactive monomer to a homotrimer with high-af- not induced upon heat shock and that HSP synthesis wascontrolled primarily at the level of translation (Bienz andfinity HSE binding activity (Wu et al., 1987; Mosser et al.,

1988; Westwood et al., 1991; Westwood and Wu, 1993). Gurdon, 1982; Bienz, 1984a,b). Although subsequent ex-periments confirmed that transcription of HSP genes inAlthough HSF is regulated primarily at the level of oligo-

merization and DNA binding, further posttranslational heat-shocked oocytes is exceedingly low, it was shown thatfollicle cells surrounding the oocyte contributed to the ex-modifications may be required for full activation of the

stress response (Larson et al., 1988). pression of HSPs first thought to represent translationalupregulation of preexisting mRNA (Horrell et al., 1987;In vertebrates, the HSF family of proteins includes HSF1,

HSF2, and HSF3 (Wiederrecht et al., 1988; Sorger and Pel- King and Davis, 1987). Inconsistent results have also beenobtained upon injection of reporter constructs under theham, 1988; Clos et al., 1990; Rabindran et al., 1991; Sarge

et al., 1991; Schuetz et al., 1991; Nakai and Morimoto, control of HSP promoters. While in some experimentsHSP70 promoters directed constitutive expression in oo-1993). HSF1 mediates the rapid induction of HSP gene ex-

pression characteristic of the classic stress response (Rabin- cytes (Bienz, 1984a; Bienz, 1986; Horrell et al., 1987), othershowed inducible transcription in response to heat stressdran et al., 1993; Sarge et al., 1993; Baler et al., 1993). In

contrast, HSF2 is not activated by stress, but appears to or coinjection of denatured protein inducers (Voellmy andRungger, 1982; Ananthan et al., 1986; Mifflin and Cohen,acquire HSE-binding ability in certain cells during differen-

tiation and development (Sistonen et al., 1992; Murphy et 1994). Wolffe and colleagues have recently accounted forvariability of expression from HSP70 promoters in the oo-al., 1994; Sarge et al., 1994; Sistonen et al., 1994). HSF3

shares some of the properties of HSF1, but its activity has cyte by demonstrating that appropriate regulation is depen-dent upon efficient assembly of injected template DNA intoonly been described in a restricted subset of avian cells

(Nakai et al., 1995). Sequence comparisons of HSF proteins chromatin (Landsberger and Wolffe, 1995; Landsberger etal., 1995). Thus, the Xenopus oocyte appears to be compe-from different species reveal several common features in-

cluding a conserved helix–turn–helix class DNA-binding tent for the stress response at the transcriptional level, eventhough increased HSP gene transcription is not apparentdomain (Harrison et al., 1994, Vuister et al., 1994), an adja-

cent amino-terminal trimerization domain containing leu- from the genomic DNA in a single nucleus (Horrell et al.,1987).cine zipper motifs (Sorger and Nelson, 1989; Peterandl and

Nelson, 1992), a carboxy-terminal leucine zipper which ap- Despite the controversy arising from earlier studies, theXenopus oocyte has emerged as a convenient model systempears to regulate oligomerization (Lis and Wu, 1993; Nakai

and Morimoto, 1993; Rabindran et al., 1993; Zou et al., in which to study the activities of injected HSF molecules.This has been illustrated by expression of cloned Drosophila1994), and transcription activation domains near the C-ter-

minus (Sorger, 1990; Green et al., 1995; Zou et al., 1995; HSF in oocytes showing partial suppression of HSE-bindingat the normal growth temperature of Xenopus (Clos et al.,Shi et al., 1995).

The mechanism by which cells detect various stress stim- 1990) and the use of oocytes by Voellmy and colleagues formutagenic analyses of human HSF1 (Baler et al., 1993; Zouuli and generate the signal(s) modulating HSF activity has

become one of the central questions of the stress response. et al., 1994, 1995). A key assumption of these experiments,however, is that endogenous oocyte HSF does not contrib-HSP gene expression appears to be strictly regulated to en-

sure that it is proportional to the severity of stress and ute substantially to the activity of injected factors. Whilethis is consistent with the failure to detect HSF with anti-switched off upon resumption of normal environmental and

physiological conditions. Attenuation of the transcriptional human HSF1 antibodies in oocyte extracts, or HSE-bindingactivity in gel shift assays (Zou et al., 1994; Stump et al.,response during prolonged exposure of cells to mild stress

is associated with conversion of the HSF trimer to inactive 1995; Landsberger and Wolffe, 1995), it is not compatiblewith strong evidence in support of a maternal pool of HSFmonomers (Clos et al., 1990; Rabindran et al., 1991; Nakai

and Morimoto, 1993). It has been proposed that the oligo- including HSE-binding activity in unfertilized eggs and em-bryos before expression from the embryonic genome at latemerization and DNA-binding state of HSF is regulated by

the intracellular level of free HSP70 family proteins (Craig blastula (Ovsenek and Heikkila, 1990; Karn et al., 1992)and transcriptional upregulation from microinjected HSP70and Gross, 1991; Morimoto, 1993). Although experiments

aimed at testing the autoregulatory effects of HSP70 in vivo promoters in the absence of exogenous HSF (Landsberger etal., 1995). Until now, however, the DNA-binding activity ofsupport an active role for modulating the HSE-binding activ-

ity of HSF (Mosser et al., 1993; Rabindran et al., 1994), a endogenous oocyte HSF has not yet been examined directly.Furthermore, very little is known about the early biochemi-mechanism involving free intracellular HSPs as a cellular

stress sensor has not been clearly established. cal steps of the stress response during vertebrate oogenesis.In this study, we have investigated the stress response ofA considerable degree of uncertainty has surrounded ear-

lier studies of the stress response in Xenopus oocytes. In the Xenopus oocyte, and report the detection of two distinctDNA-binding activities of endogenous HSF molecules. Anfact, the oocyte was at one time thought to represent an

exception to the universality of the heat shock response endogenous oocyte HSE-binding activity was induced byheat, cadmium, and arsenite at all stages of oogenesis. Thisbecause initial experiments suggested that HSP mRNA was

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID DB 8441 / 6x15$$$$81 12-09-96 18:37:34 dbas

49HSF Activities in Xenopus Oocytes

from Xenopus testes and rat ovary material were made essentiallystress-inducible activity was supershifted by antibodiesas described above.against HSF1, but not by HSF2 antibodies. The DNA-bind-

ing activity of oocyte HSF1 persisted throughout the dura-tion of prolonged stress treatments, suggesting that factors Gel Mobility Shift Assaysregulating attenuation of HSE-binding ability are limiting

DNA mobility shift assays were performed as previously de-or modified in the oocyte. In addition, we report a distinctscribed (Ovsenek and Heikkila, 1990). DNA-binding reactions withdevelopmentally regulated HSE-binding activity present ineither embryonic or stage V and VI oocyte samples contained 10stage I and II oocytes, but absent in later stage oocytes,ml extract (one embryo equivalent is approximately 20 mg solubleembryos, and somatic cells. This activity was not recog-protein). For experiments using early stage oocytes, protein levelsnized in vitro by either HSF1 or HSF2 antiserum and waswere quantified with a Bio-Rad kit (Bio-Rad, Hercules, CA), equiva-

unaffected by stress treatments of early oocytes. Northern lency was checked by Coomassie staining of SDS–polyacrylamideblot experiments showed that hsp70 mRNA levels are also gels, and extract volumes were adjusted so that equal protein (20maximal in early stage oocytes. Taken together, these data mg) was added to each binding reaction. HSE oligonucleotide probesprovide evidence of a unique, developmentally regulated used in our assays were as described in Ovsenek et al. (1990). Bind-HSF which may play a role in expression of heat shock ing reactions were performed in the presence of 1 mg poly(dI-dC),

10 mM Tris (pH7.8), 50 mM NaCl, 1 mM EDTA, 0.5 mM dithiothre-protein genes in previtellogenic oocytes.itol, and 5% glycerol, in a final volume of 20 ml. Reactions wereincubated on ice for 20 min and immediately loaded onto 5% non-denaturing polyacrylamide gels containing 6.7 mM Tris–ClMATERIALS AND METHODS(pH7.5), 1 mM EDTA, 3.3 mM sodium acetate. Gels were electro-phoresed for 2.5 hr at 150 V, dried, and exposed overnight to X-ray

Oocytes film (Kodak X-omat 5). For supershift analysis, polyclonal antise-rum against either HSF1 or HSF2 (kindly provided by Dr. KevinXenopus laevis frogs were purchased from Xenopus I (Ann Ar-Sarge, Department Biochemistry, Chandler Medical Center, Uni-bour, MI). Ovary portions were surgically obtained from adult fe-versity of Kentucky; Sarge et al., 1993) was incubated with extractsmale frogs and follicle cells were removed from oocytes by treat-for 15 min on ice prior to addition of extracts into binding reactions.ment in calcium-free OR2 buffer (82.5 mM NaCl, 2.5 mM KCl, 1

mM MgCl2, 1 mM NaH2PO4, 5 mM Hepes, pH 7.8, 10 mg/literstreptomycin sulfate, 10 mg/liter benzyl penicillin) containing 2 Northern Blot Analysismg/ml collagenase (type II, Sigma) for 2–3 hr at 187C. Oocytes werewashed extensively and allowed to recover overnight at 187C in Oocytes at different stages of oogenesis were collected andOR2 (as above /1 mM CaCl2; Wallace et al., 1973). Efficiency of total RNA was isolated after homogenization in TRIzol reagentfollicle cell removal was monitored by fluorescence microscopy (Gibco-BRL, Burlington, Ontario, Canada). Samples containingafter staining oocytes in 1 mg/ml Hoecsht dye (Horrell et al., 1987). total RNA (10 mg) were electrophoresed on 1.2% formaldehydeIn some experiments, oocytes were separated into pools represent- agarose gels and transferred to Hybond-N/ membrane (Amer-ing each of the individual stages according to the criteria described sham, Oakville, Ontario, Canada; Sambrook et al., 1989). Hybrid-by Dumont (1972). Control oocytes were maintained at 187C in ization reactions were performed using 32P-labeled XenopusOR2, chemically stressed oocytes were incubated in OR2 supple- hsp70B gene probe (Bienz, 1984a). After posthybridizationmented with sodium arsenite or cadmium chloride, and heat- washes, blots were exposed to X-ray film at 0807C. Equal loadingshocked oocytes were incubated in OR2 preequilibrated to indi- of the RNA gels was determined by spectrophotometry and bycated temperatures. In all experiments, at least 20 oocytes were visualization of the RNA after staining with ethidium bromide.used for each sample. Following stress treatments, oocytes were Autoradiographs were scanned on a UMAX Vista-S6E scannerquickly washed in OR2 and collected for protein extracts or for and relative message levels were determined using the NIH Im-CAT expression analysis. age Version 1.59.1 analysis program. Message sizes were deter-

mined using a standard RNA ladder (Gibco BRL).

Protein ExtractsOocyte Injections and CAT Assays

For protein extracts, oocytes were homogenized in Buffer C (50mM Tris–Cl, pH 7.9, 20% glycerol, 50 mM KCl, 0.1 mM EDTA, Plasmid constructs used for microinjection experiments were the

human CMV-CAT and the Xenopus Hsp70-CAT clones (kindly2 mM dithiothreitol, 10 mg/ml aprotinin, and 10 mg/ml leupeptin(Dignam et al., 1983)) in a Dounce homogenizer with a tight fitting provided by Dr. Alan Wolffe, NICHHD, National Institutes of

Health, Bethesda, MD) previously described in Landsberger et al.pestle. Homogenates were transferred to eppendorf tubes and spunfor 5 min at 15,000g (47C). The resultant supernatants were re- (1995). Following defolliculation, oocytes were incubated for sev-

eral hours at 187C, after which healthy oocytes were selected andmoved to a fresh tube, immediately frozen in liquid nitrogen, andstored at 0807C. Oocytes (stage V and VI) were homogenized in injected into nuclei with 20 nl of a solution containing 100 ng/ml

(2 ng) of either CMV-CAT or Hsp70-CAT plasmid using a Naras-10 ml buffer C per oocyte, which typically yielded a final proteinconcentration of 2.0 mg/ml. Early stage oocytes were homogenized hige IM 300 microinjector. After incubation overnight at 187C,

healthy oocytes were selected and stressed for 2 hr by heat shockin the following volumes in order to obtain approximately equiva-lent final protein concentration (2.0 mg/ml) for each sample: stages at 337C or by incubation for 2 hr in 50 mM cadmium or 5 mM

arsenite at 187C. Following these treatments, oocytes were incu-III and IV, 4 ml/oocyte; stages I and II, 1 ml/oocyte. Protein extracts


AID DB 8441 / 6x15$$$$82 12-09-96 18:37:34 dbas

50 Gordon et al.

bated for an additional 2 hr at 187C, washed in OR2, and assayed HSE-binding activity in the oocyte (Fig. 1B and 1C). HSE-for CAT activity. As a control for oocyte injections, DNA was binding activity was induced in oocytes after a 1-hr expo-recovered from at least 5 individuals out of the pool of injected sure to either cadmium chloride (Fig. 1B, lane 9) or sodiumoocytes, and the equivalency of injected plasmid DNA was con- arsenite (Fig. 1C, lanes 6–8). These results indicate thatfirmed by Southern blotting. CAT assays were performed using 1 putative HSF1 molecules are activated at the level of DNAoocyte equivalent of whole cell extract from uninjected or injected

binding by at least two different forms of chemical stress.oocytes as previously described (Ovsenek et al., 1990). A pool of atMoreover, the oocyte appears to be competent for the recep-least 20 oocytes were used for each analysis. The acetylated prod-tion and transduction of disparate stress signals.ucts were separated by thin-layer chromatography and visualized

We confirmed that the HSE-binding activity observed inby autoradiography. CAT activity was quantified by scintillationcounting and expressed as percentage conversion of chlorampheni- our experiments was not attributable to contaminating fol-col to acetylated forms. licle cells. Batches of oocytes used in these experiments

were determined to be free of somatic cells after stainingwith Hoescht dye and examination by fluorescence micros-copy (not shown). In addition, we directly compared theRESULTSlevels of inducible HSE-binding activity in enzymaticallydefolliculated oocytes and manually separated oocytes con-Stress Induction of an Endogenous HSE-Bindingtaining several thousand follicle cells (Fig. 1D). IncreasedActivity in Stage VI Oocyteslevels of inducible HSE-binding activity were not observedupon deliberate inclusion of follicle cells in stress experi-We investigated the DNA-binding activities of endoge-

nous HSF molecules in Xenopus oocytes. Results from ear- ments (Fig. 1D, compare lanes 2 and 3 to lanes 5 and 6).Thus, it appears as though any potential contribution fromlier work suggested that endogenous HSF molecules do not

contribute to the HSE-binding activity observed in gel shift follicle cell HSF activity is below the level of detection inthese assays.assays performed with extracts from oocytes injected with

cloned HSF molecules (Clos et al., 1990; Baler et al., 1993; In order to determine the composition of the stress-induc-ible HSE-binding activity (s1) in oocytes, gel super-shift ex-Zou et al., 1994, 1995). It has been established, however,

that the oocyte is capable of upregulating transcription from periments were performed using anti-HSF1 and anti-HSF2antibodies. For this experiment, extract from heat-shockedmicroinjected DNA constructs containing HSP70 promot-

ers (Landsberger et al., 1995). Since this expression was pre- stage VI oocytes was preincubated with polyclonal antise-rum raised against either mouse HSF1 or HSF2 proteinssumably attributable to induction of endogenous HSF mole-

cules, we reasoned it would be possible to detect oocyte (Sarge et al., 1983). These antisera have previously beenshown to exclusively recognize their respective antigens inHSE-binding activity in heat-shocked oocytes. In order to

test for HSF activation, a gel mobility shift assay was per- gel mobility shift assays (Sarge et al., 1993). Specific recogni-tion by either antiserum should result in neutralization orformed with oocyte extracts and an HSE oligonucleotide

probe (Fig. 1A). While HSE-binding activity was not de- decreased mobility of the HSE–HSF complex (s1). Additionof HSF1 antiserum to pre-stressed oocyte extract resultedtected in oocytes incubated at control temperatures (Fig.

1A, lane 1), a prominent HSE–protein complex was induced in the loss of heat shock-induced HSE-binding activity (Fig.2A, compare lanes 1–4). Conversely, addition of the poly-after treatment of oocytes between 30 and 367C (Fig. 1A,

lanes 4–9). Induction was rapid, as detectable levels of com- clonal antiserum against HSF2 did not affect the formationof the stress-inducible complex (lanes 5–7). The heat-induc-plex were observed immediately after 30 min of heat shock

(Fig. 1A, lanes 4, 6, and 8). The results of this experiment ible HSE-binding activity was unaffected even at very highHSF2 antibody concentration (1:10 dilution; data notsuggest that the Xenopus oocyte is competent to activate

trimerization of an endogenous HSF, the principal step in shown), whereas complete neutralization of binding wasobserved at a relatively low HSF1 antibody concentrationthe stress response pathway. Heat-inducible binding of en-

dogenous oocyte HSF is entirely consistent with previously (1:200 dilution; see lane 3). Neither antiserum affected thenonspecific complex. We also determined that the factorobserved transcriptional upregulation from injected HSP70

promoters (Landsberger et al., 1995; also see Fig. 2B). Ab- contributing to the stress-induced HSE-binding activity ob-served in these experiments is related to the previouslysence of significant HSE-binding activity in controls (Fig.

1A, lane 1) suggests that putative oocyte HSF1 molecules cloned XHSF1 (Stump et al., 1995). Incubation of heat-stressed stage VI extracts with a polyclonal rabbit antiserumare present as latent non-DNA-binding monomers, and may

be negatively regulated at the level of oligomerization under made against in vitro-synthesized XHSF1 (Mercier, Ov-senek, and Westwood, unpublished results) also resulted innonstress conditions.

Since earlier studies have shown that heat shock, heavy knockout of the stress-inducible complex (data not shown).We conclude from this experiment that the stress-inducedmetals, and arsenite treatments bring about activation of

the stress response and accumulation of HSP mRNA in HSE-binding activity in stage VI oocytes is formed by aXenopus homologue of mammalian HSF1. The presentdeveloping embryos (Heikkila et al., 1987), we were

prompted to examine the effects of metal and arsenite on analysis, however, does not allow us to conclude that the


AID DB 8441 / 6x15$$$$82 12-09-96 18:37:34 dbas


1 2 3 4 5 6 7 8 9 10 11

18 27 30 33 36 39- 0.5 1.0 0.5 1.0 0.5 1.0 0.5 1.0 0.5 1.0

Temp. (°C)Time (h)

ns

s1

A

1 2 3 4 5 6 7 8

C H 1 10 100 1 10 100

(µM)

[AsO2]

(mM)

ns

s1

C

1 2 3 4 5 6

_ _ _ + + +

_ HS Ars – HS ArsStressFollicle

ns

s1

D

FIG. 1. Stress-induction of HSE-binding activity in stage VI oocytes. (A) Gel mobility shift assays were performed with protein extractsprepared from stage VI oocytes incubated at control temperature (187C, lane 1) or different heat shock temperatures for either 30 min(lanes 2, 4, 6, 8, and 10) or 60 min (lanes 3, 5, 7, 9, and 11). Extract equivalent to one oocyte (20 mg) was used in each binding reaction.Time and temperature of heat treatments are indicated above each lane. The heat-inducible HSE-binding complex is indicated by an arrowlabeled s1 (shift-1). The relative intensity of the non-specific complex (ns), typically found in embryonic and oocyte extracts, is an indicationof cell viability. Treatments at 397C were lethal to oocytes, as indicated by disappearance of the ns complex associated with oocyte death.(B) HSE-binding activity was measured by gel mobility shift assay in stage VI oocytes exposed to different concentrations of cadmiumchloride (CdCl2) for 1 hr (lanes 4–9). Cadmium concentrations are indicated above each lane. Exposure to 100 mM CdCl2 induced formationof the HSE-binding complex (lane 9). Extracts prepared from untreated (187C, lane 2) and heat-shocked oocytes (337C for 1 hr, lane 3) wereincluded as controls. Lane labeled P (lane 1) contained probe alone. (C) HSE-binding activity was determined by gel mobility shift assayin stage VI oocytes exposed to increasing concentrations of sodium arsenite (NaAsO2) for 1 hr (lanes 3–8). Extracts from untreated controls(C, lane 1) and heat-shocked oocytes (H, lane 2) were included as controls. (D) Heat-inducible HSE-binding activities were compared inenzymatically defolliculated oocytes (lanes 1–3) and manually separated oocytes containing thousands of follicle cells as determined byHoescht staining (lanes 4–6). Binding reactions contained control oocytes at stage VI (lanes 1 and 4), oocytes heat shocked at 337C for 1hr (lanes 2 and 5), or oocytes treated with 10 mM arsenite for 2 hr (lanes 3 and 6).


AID DB 8441 / 6x15$$8441 12-09-96 18:37:34 dbas

52 Gordon et al.

1 2 3 4 5 6 7

No

Ab

1:50

0

1:20

0

1:10

0

1:50

0

1:20

0

1:10

0

Anti-HSF1 Anti-HSF2

ns

s1

A

0.1 0.1 0.1 0.1 0.1 89.2 84.3 83.7

C H Ars Cd C H Ars Cd

CMV-CAT HSP70-CATB

FIG. 2. Antibody recognition of the factor forming the stress-inducible HSF–HSE complex, and stress regulation of its transactivationability. (A) Aliquots of extract containing heat-shocked (337C, 1 hr) stage VI oocyte (20 mg) were incubated either alone (no antibody, lane1) or with a 1:500, 1:250, or 1:100 dilution of the HSF1 antiserum (lanes 2–4) or HSF2 antiserum (lanes 5–7) prior to gel mobility shiftanalysis. The specific HSE–HSF1 complex is indicated by an arrow (labeled s1), and the nonspecific complex is indicated as ns. (B) CATassays were performed on stage VI oocytes injected with CMV-CAT or hsp70-CAT plasmid DNAs and subjected to stress. Microinjectionswere performed as described under Materials and Methods. Oocyte treatments were as follows: control, 187C; heat shock, 337C for 1 hr;Cd, 50 mM Cd for 1 hr; Ars, 5 mM arsenite for 1 hr, as indicated above the panel. Percentage conversion of chloramphenicol to acetylatedforms is indicated below.

HSF1 protein present in oocytes is identical to XHSF1, and the CMV promoter were microinjected into oocyte nuclei,and oocytes were incubated overnight to allow for chroma-so the possibility remains that oocyte HSF1 is encoded by

a different gene. tin assembly (Landsberger and Wolffe, 1995). The next day,oocytes were subjected to the same stress treatments shownWe next examined the relationship between induction of

the HSE-binding ability of HSF1 and transcriptional regula- to result in activation of HSF1 and formation of the s1 com-plex (337C, 100 mM Cd and 5 mM arsenite; see Fig. 1). CATtion of the Xenopus hsp70 promoter microinjected into oo-

cytes. Solutions of plasmid DNAs containing the bacterial activity was undetectable in oocytes injected with CMVpromoter constructs (Fig. 2B, lanes 3–6) and unstressed oo-CAT reporter gene linked to either the hsp70 promoter or


AID DB 8441 / 6x15$$$$82 12-09-96 18:37:34 dbas


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

C H Cd C H Cd C H Cd C H Cd C H Cd C H CdI II III IV V VI

oocyte stage:

ns

s1

s2

FIG. 3. Developmental pattern of stress-inducible HSE-binding activity in developing oocytes. Oocytes were separated into pools repre-senting each individual stage (I to VI (Dumont, 1972), stages indicated above each lane). Staged oocytes were incubated under controlconditions (indicated as C; lanes 1, 4, 7, 10, 13, and 16), heat shocked at 337C for 1 hr (indicated as H; lanes 2, 5, 8, 11, 14, and 17), orexposed to 50 mM CdCl2 for 1 hr (indicated as Cd; lanes 3, 6, 9, 12, 15, and 18). Gel mobility shift analysis was performed essentially asin Fig. 1. Each binding reaction contained protein equivalent to one stage IV oocyte (20 mg; see Materials and Methods). An HSE-bindingcomplex (arrow labeled s1) was induced after heat shock or cadmium treatments at each stage of oogenesis. A second faster migratingcomplex (see arrow labeled s2) is detectable in extracts of control and stressed stage 1 and II oocytes, but not in later stage oocytes.

cytes injected with Hsp70-CAT (lane 7). In contrast, treat- heat (Fig. 3, lanes 2, 5, 8, 11, 14, and 17) and cadmium stress(lanes 3, 6, 9, 12, 15, and 18) in oocytes at stage I throughment of Hsp70-CAT-injected oocytes with heat, cadmium,

and arsenite resulted in very high levels of CAT activity to stage VI of development (see arrow labeled S1). The re-sults of this assay show that a stress-inducible HSE-binding(Fig. 2B, lanes 8–10). These experiments clearly establish a

functional relationship between formation of the s1 com- factor, presumably HSF1, is present in early oocytes andis responsive to heat and chemical stress at all stages ofplex in oocytes and transcription from an exogenous hsp70

promoter. It is clear, therefore, that a variety of stress treat- oogenesis. In order to confirm these results, we repeated thisexperiment several times (data not shown). The apparentments induce both the HSE-binding ability and the trans-

activation potential of oocyte HSF1. decline in the intensity of stress-inducible HSF1–HSE com-plex (s1) at stage V and VI (lanes 14, 15, 17, and 18) was notobserved in most experiments and therefore is not likely to

The Stress Response in Early Stage of Oocytes reflect real differences in the levels of induction of HSF1between early and late stage oocytes.

Very little is known about the stress response during oo-genesis, particularly at early stages. The majority of studiesexamining the fundamental molecular events of the heat Identification of a Constitutive, Developmentallyshock response in preembryonic development have focused Regulated HSE-Binding Activity in Early Stageon later stage (stage VI) Xenopus oocytes as a test system. OocytesWe therefore wished to examine whether oocytes at earlierstages were capable of responding to stress at the level of In our initial experiments using early stage oocytes, an

unusual HSE-binding activity was observed in unstressedHSF activation. In order to test this, individual oocytes wereseparated into pools representing each of the stages of oo- controls (Fig. 3, lanes 1–6; see arrow labeled s2). This com-

plex was electrophoretically distinct from heat-induciblegenesis as established by Dumont (1972). Oocytes at eachstage were exposed to heat shock or cadmium treatments, HSF activity, migrating between heat-inducible and nonspe-

cific bands. In order to clearly establish the stage-depen-and the results of the standard HSE-binding assay are shownin Fig. 3. Increased HSE-binding activity was detected after dency of this apparently constitutive complex, and to rule


AID DB 8441 / 6x15$$$$82 12-09-96 18:37:34 dbas

54 Gordon et al.

out the possibility that its formation in control oocytes (Fig. induction upon stress treatments. In order to extend thiscomparison, we examined the possibility that a similar ac-3, lanes 1 and 4) was a consequence of inadvertent stress,

staged oocytes were isolated from the ovary of a different tivity was present in testes isolated from adult Xenopusmales. This tissue was chosen because the DNA-bindingfemale and a gel mobility shift assay was performed as above

(Fig. 4A). An HSE-binding complex (s2) was detected in un- activity of HSF2 is known to be particularly strong in mam-malian testes (Murphy et al., 1994; Fiorenza et al., 1995).stressed stage I and II oocytes (Fig. 4A, lanes 1 and 2), but

not in later stage oocytes from stages III to VI (Fig. 4A, lanes Results of comparative gel shift analyses show that consti-tutive HSE-binding activity formed in stage II oocytes3–6). We also tested for the presence of this activity in

Xenopus embryos at various developmental stages and in closely co-migrates with an activity present in protein ex-tracts made from testes exposed to control and heat shocksomatic cells (Fig. 4B). The constitutive HSE-binding activ-

ity was not detected in extracts from unfertilized eggs, conditions (Fig. 5A, compare lane 1 to lanes 3 and 4). Wealso tested for the presence of a constitutive HSE-bindingcleavage or gastrula stage embryos, or in A6 cells (Fig. 4B,

lanes 2–5). Therefore, activity of the s2 complex appears to activity in mammalian ovary tissue. Interestingly, a similaractivity was also detected in rat ovary tissue (lane 2), im-be restricted to early oogenesis.

We next performed competition experiments to deter- plying the possible involvement of a related factor in mam-malian oogenesis. All of these activities were found to comi-mine the sequence specificity of both HSE–protein interac-

tions s1 and s2 (Fig. 4C). The faster-migrating constitutive grate with an HSF2 band in controls using muring testes(data not shown). Although this comigration strengthenedHSE-binding complex (s2) present in extracts from un-

stressed stage II oocytes was diminished by addition of HSE the correlation between the s2 complex and a putative Xen-opus HSF2 homologue, it was not sufficient to positivelyoligonucleotide competitor to DNA-binding reactions (Fig.

4C, lane 2), but not by excess amounts of oligonucleotides identify the HSE-binding factor as HSF2. In order to moreclearly determine the composition of the constitutive HSE-containing AP-2, CEBP, or GAL4 binding sites (lanes 3–5).

These results clearly establish the constitutive complex (S2) binding activity, we performed gel super-shift experimentswith the same anti-HSF1 and anti-HSF2 antibodies used inas a sequence-specific HSE-binding activity.

It was also important to determine whether stress treat- Fig. 2. For this experiment, extracts from unstressed andheat-shocked stage I oocytes were preincubated with thements would bring about an elevation in the levels of the

constitutive complex (Fig. 4C, lanes 6–10). Treatments of polyclonal antisera against mouse HSF1 or HSF2 proteins(Sarge et al., 1983). First, neither antiserum specifically rec-stage II oocytes with 50 mM cadmium resulted in formation

of the more slowly migrating HSE–HSF1 complex (s1), but ognized the constitutive (s2) complex in early oocytes (Fig.5B, compare lanes 2, 3, and 4). On the other hand, HSF1had no significant effect on formation of the s2 complex

(compare lanes 1–5 to lanes 6–10). In numerous experi- antiserum neutralized the stress-inducible HSE-binding ac-tivity (s1) in heat-shocked stage 1 oocytes (a supershiftedments using a variety of stress regimes, the constitutive

HSE-binding activity of stage I and II oocytes was not sig- complex is apparent in lane 6), but had no effect on the s2complex in the same extract (Fig. 5B, compare lanes 5 andnificantly affected by conditions known to elicit activation

of HSF1 (data not shown; also see Fig. 3). In addition, both 6). These results confirm the identity of the stress-induciblefactor as HSF1 and suggest that the constitutive complexheat-inducible and constitutive complexes were found to be

sequence specific, as shown by competition with unlabeled is formed by a distinct factor. HSF2 antiserum did not recog-nize either complex (compare lanes 4 and 7). HSF2 antise-HSE oligonucleotide (lane 7). Oligonucleotides containing

known binding sites for established transcription factors rum did, however, recognize the HSE-binding complex inrat ovary tissue, and control experiments confirm that the(AP-2, CEBP, and GAL4) had no competitive effect on for-

mation of either complex (lanes 8–10). antiserum used in these experiments supershifted the HSE–HSF2 complex present in extracts of murine testes (dataOn the basis of its behavior in stage I and II oocytes, we

speculated that the constitutive HSE-binding complex (s2) not shown). The simplest interpretation of this data is thatthe protein comprising the constitutive complex (s2) inwas formed by a Xenopus homologue of mammalian HSF2.

Previous experiments with mouse and human cells have early oocytes is a unique HSE-binding protein unrelatedto either HSF1 or HSF2, although it is possible that theshown that HSF2 is active in cells undergoing differentia-

tion and development, exhibits constitutive HSE-binding polyclonal antibodies present in the antiserum directedagainst mouse HSF2 do not sufficiently recognize a putativeactivity in the male germ line and embryonal carcinoma

cells, and is not induced by stress at the level of DNA- Xenopus HSF2 homologue. Incubation of stage I and II oo-cyte extracts with very high anti-HSF2 antiserum concen-binding activity (Mezger et al., 1989; Sistonen et al., 1992;

Murphy et al., 1994; Sarge et al., 1994; Sistonen et al., 1994; trations (1:10 dilution) did not affect the s2 complex (datanot shown).Goodson et al., 1995). The constitutive complex observed

in our experiments appeared to share several features in We were interested in examining a possible relationshipbetween the developmental pattern of the constitutive HSE-common with HSF2, including its presence in cells undergo-

ing growth and differentiation (stage I and II oocytes), affin- binding activity (s2) and the synthesis of HSP mRNA duringoogenesis. Northern blot analysis was performed using theity for the HSE in the absence of stress, and its lack of


AID DB 8441 / 6x15$$$$83 12-09-96 18:37:34 dbas


1 2 3 4 5 6

I II III IV V VI

ns

s2

oocyte stageA

1 2 3 4 5 6

stI E CL G A6

ns

s2

B

1 2 3 4 5 6 7 8 9 10

- HS

E

CE

BP

AP

-2

GA

L4

- HS

E

CE

BP

AP

-2

GA

L4

Control

Competitor:

50 mM CdStage II oocytes

ns

s1

s2

C

FIG. 4. A constitutive, noninducible, and developmentally regulated HSE-binding activity in stage I and II oocytes. (A) HSE-bindingactivity was examined by gel mobility shift analysis using extracts made from staged oocytes under control conditions (187C). A constitutiveHSE-binding complex (arrow labeled s2) was detected in binding reactions with stage I and II oocytes, but not with extracts from oocytesat stages III to VI. The nonspecific complex is indicated by an arrow labeled ns. (B) HSE-binding activity in extract from unstressed stage1 oocytes (lane 2) was compared to activity in unstressed extracts from unfertilized egg (lane 3), cleavage stage embryos at stage 6 (lane4), gastrula stage embryos at stage 10 (lane 5), and A6 cells (lane 6). Each binding reaction contained 20 mg of protein extract, except forlane 1 (P) which contained HSE oligonucleotide probe alone. (C) Protein extracts were prepared from oocytes at stage II that were eitherunstressed (Control; lanes 1–5) or incubated for 2 hr in cadmium (50 mM Cd, lanes 6–10) and used in a competition experiment todetermine the sequence specificity of stress-inducible (complex labeled s1) and constitutive (complex labeled s2) HSE-binding activities.Extract equivalent to approximately five stage II oocytes (20 mg) was used in each binding reaction. Competitor DNAs were added intobinding reactions at a 50-fold molar excess relative to labeled HSE probe. Formation of both constitutive (s2) and stress-inducible (s1)complexes was diminished by HSE competitor (lanes 2 and 7), but not by oligonucleotides containing CEBP (lanes 3 and 8), AP-2 (lanes4 and 9), or GAL4 (lanes 5 and 10) binding sites.


AID DB 8441 / 6x15$$8441 12-09-96 18:37:34 dbas

56 Gordon et al.

1 2 3 4

I-II C HSRatov.

X. testes

ns

s2

A

1 2 3 4 5 6 7

– + – – + +– – + – – +

oocyte stage I

control heat shock

Anti-HSF1Anti-HSF2

ns

s1

s2

B

FIG. 5. The developmentally regulated HSE-binding activity of early stage oocytes is unrelated to HSF2. (A) Comparison of constitutiveHSE-binding activities in early oocytes, testes, and rat ovary tissue. Migration of the constitutive complex (s2) was compared in extractsfrom stage I and II oocytes (lane 1), rat ovary (lane 2), unstressed Xenopus testes (lane 3), or heat-shocked testes (lane 4). (B) Antibodyrecognition of the constitutive HSE-binding activity (s2) in stage I oocytes. Extract containing unstressed (lanes 2–4) or heat-shocked(337C, 1 hr) stage I oocytes (lanes 5–7) was incubated either alone (no antibody; lanes 2 and 5), or with a 1:250 dilution of the HSF1antiserum (lanes 3 and 6) or HSF2 antiserum (lanes 6–8) prior to gel mobility shift analysis. 20 mg of extract was used in each bindingreaction. Lane 1 (P) contained probe alone.

Xenopus hsp70B gene probe (Bienz, 1984a) and total RNA shock (Horrell et al., 1987). Thus, the oocyte represents aunique system in which stress-induced activation of HSFisolated from oocytes at each stage of oogenesis (Fig. 6).

High levels of HSP70 mRNA was detected at stage I and can be assessed without concurrent elevation of intracellu-lar HSP concentrations. Therefore, we examined the kinet-II (Fig. 6, lanes 1 and 2). Levels declined throughout later

oogenesis (lanes 3–6). We note that the total RNA content ics of HSF1 activation in oocytes during prolonged stress.Stage VI oocytes were continuously exposed to heat shockof oocytes is known to increase approximately fivefold dur-

ing development from stage II to VI (Dolecki and Smith, and extracts prepared at various time points were used inthe HSE-binding assay presented in Fig. 7A. This experi-1979), and so this experiment should not be seen as a mea-

surement of absolute changes in HSP70 mRNA accumula- ment shows that the stress-inducible HSE-binding complexremained active for as long as 9 hr of heat shock at 337Ction in a single oocyte. Regardless, a clear correlation is

apparent between the early appearance of hsp70 mRNA and (Fig. 7A, lanes 3–12). HSE-binding activity of HSF1 per-sisted until oocyte death (Fig. 7A, lane 13). A similar patternthe presence of the s2 complex during oogenesis. We suggest

that the constitutive HSE-binding activity plays a role in was observed in several different experiments performed atthe expression of HSPs during early stages of oocyte devel- different heat shock temperatures, with HSF1 remainingopment. active throughout the duration of oocyte viability (data not

shown). This extended pattern contrasts with the transientupregulation and deactivation of HSE-binding activity

Activation of HSF1 during Continuous Stress (within 2 hr) observed under similar heat shock conditionsin Xenopus embryos (Ovsenek and Heikkila, 1990; Karn etIt has been postulated that the DNA-binding and oligo-al., 1992).meric state of HSF1 is regulated by the intracellular level

To extend this analysis, we also tested whether the HSE-of HSP70 family proteins (Morimoto, 1993). In oocytes,however, increased HSP70 synthesis is not induced by heat binding ability of HSF1 remained active during prolonged


AID DB 8441 / 6x15$$$$83 12-09-96 18:37:34 dbas


10) and also in oocytes exposed to 5 mM cadmium (Fig.7C, lanes 7–11). Stress-inducible HSE-binding activity waspresent in oocytes up until oocyte death at 36 hr, as detectedby diminishment of the nonspecific band (not shown).Lower concentrations of arsenite and cadmium were usedfor these time course experiments than in previous experi-ments examining acute activation of HSE-binding ability(Figs. 1B and 1C) because high concentrations resulted inoocyte death after only a few hours of treatment (data notshown). The same pattern of HSF activation was detectedwith chemical stress as with heat shock; HSF remainedactive for the duration of stress treatments up until oocytedeath, the time of which was dependent upon the severityof treatment (data not shown). These results suggest thatalthough trimerization of endogenous oocyte HSF1 mole-cules is apparently induced by stress, the cellular mecha-nisms regulating conversion of HSF1 to the monomericstate are absent or inactivated in the oocyte. Since HSP70family proteins appear to function in the disassembly ofHSF trimers and attenuation of HSE-binding activity (Mos-ser et al., 1993; Rabindran et al., 1994; Baler et al., 1996),we speculate that the extended pattern of HSF1 activationin oocytes is a consequence of low de novo HSP proteinsynthesis (Horrell et al., 1987).

We next tested the attenuation of HSF1 activity in oo-cytes during recovery from an acute heat shock (Fig. 8). Asexpected, a high level of HSE-binding complex was inducedafter 1 hr of heat shock (Fig. 8, lane 2). HSE-binding activityreturned to control levels within 5 min after resumption ofnormal temperatures (Fig. 8, lanes 3–6). In similar recoveryexperiments, HSF1 activity remained high in extracts madefrom oocytes following brief exposures to arsenite or cad-mium (data not shown). This was probably due to the persis-tent effects of arsenite or cadmium absorbed into oocytesand not removed even after extensive washes. Taken to-gether, the results of experiments examining the kineticsof HSF1 activation (Figs. 7 and 8) suggest that the cellularmechanism(s) required for disassembly of HSF1 trimers ispresent in the oocyte, but at levels insufficient for attenua-

1 2 3 4 5 6

l ll lll lV V Vloocyte stage:

28S

18S

18S

Hsp70

I II III IV V VIOocyte Stage

Hsp

70 m

RN

A L

evel

s

0

10

20

30

40

50

tion of the response during continuous stress.FIG. 6. Northern blot examination of hsp70 mRNA expressedduring Xenopus oogenesis. Total RNA isolated from oocytes atdifferent stages was electrophoresed on a 1.2% agarose formalde- DISCUSSIONhyde gel and then transferred to Hybond-N/ membrane. The blotswere hybridized against the 32P-labeled Xenopus hsp70B DNA

These experiments represent the first direct demonstra-probe. All lanes contain 10 mg of total RNA from each oocyte stagetion of a maternal pool of distinct stress-inducible and con-as indicated above. An arrow indicates the position of the 2.7-kb

hsp70 transcripts. Ribosomal RNA bands are shown on the left. As stitutive HSE-binding activities in Xenopus oocytes. Wea control, levels of 18S rRNA were detected with a radiolabeled have established that an endogenous HSF is inducible byrRNA oligonucleotide probe and are shown below the panel. Rela- heat shock and chemical stress and is readily detectable bytive levels of HSP70 mRNA were determined by densitometry and gel mobility shift assay. Gel supershift experiments usingare shown at the bottom of the panel. antisera against mammalian HSF1 and HSF2 suggest that

the stress-inducible HSE-binding activity detected through-out oogenesis is formed by a Xenopus homologue of HSF1.HSF1 appears to be present in unstressed oocytes in a latentexposure of oocytes to arsenite and cadmium (Figs. 7B and

7C). HSE-binding activity remained active for as long as 24 form that is converted to the HSE-binding conformationupon heat and chemical stress at all stages of oogenesis (Fig.hr in oocytes exposed to 500 mM arsenite (Fig. 7B, lanes 3–


AID DB 8441 / 6x15$$$$83 12-09-96 18:37:34 dbas

58 Gordon et al.

1 2 3 4 5 6 7 8 9 10 11 12 13

0 0.1 0.5 1 2 3 4 5 6 7 8 9 10

Time (h) at 33°C

ns

s1

A

FIG. 7. Time course of HSE-binding activity in oocytes continuously exposed to stress. (A) Stage VI oocytes were maintained at a heat-shock temperature of 337C, and protein extracts were obtained at the times indicated above each lane. A heat-inducible HSE-bindingcomplex (arrow labeled s1) was detected after 30 min of exposure (lane 3), and persisted throughout the time course of the experiment(lanes 3–12) up until oocyte death (lane 13). Diminishment of the ns complex at 10 hr is indicative of oocyte death. (B) Time course ofHSE-binding activity in oocytes incubated in 500 mM sodium arsenite. Lanes 1 and 2 contained a control extract (187C) taken at Time 0and at 24 hr, respectively. Time of arsenite treatment is indicated above each lane (lanes 3–10). The inducible complex is indicated byan arrow labeled s1. (C) Time course of HSE-binding activity in oocytes incubated in 5 mM cadmium chloride. Controls were taken atthe onset of treatment (Time 0; lane 1) and after 12 (lane 2) and 24 hr (lane 3). The inducible complex (arrow labeled s1) was present upuntil oocyte death after 24 hr. The nonspecific band is indicated by an arrow labeled ns.

2). Experiments with staged oocytes provide evidence for a binding activity in earlier studies (Clos et al., 1990; Baleret al., 1993; Zou et al., 1994, 1995), this was likely due toconstitutive HSE-binding activity that is electrophoreti-

cally distinct from the HSF1–HSE complex and does not the low level of maternal factors relative to the robust activ-ity of exogenously expressed HSF1 molecules. Here we haveappear to be affected by stress. This activity appears to be

developmentally regulated, as judged by its presence in established the in vivo cellular stress conditions for repro-ducible activation of HSE binding upon acute exposures tostage I and II oocytes, but not in later stage oocytes, eggs,

embryos, or somatic cells (Figs. 3 and 4). Although the be- heat shock (337C), cadmium (50 mM ), and arsenite (1 mM ).These data provide direct evidence that the frog oocyte hashaviour of the factor forming this complex is similar to that

of mammalian HSF2 (Fig. 5A), gel supershift experiments a vigorous response to heat shock and diverse forms ofchemical stress as measured by the activation of HSF1 DNAsuggest it may be a unique HSE-binding factor unrelated to

HSF1 or HSF2 (Fig. 5B). A close correlation was observed binding ability.Antibody recognition experiments (Figs. 2A and 5B) showbetween the developmental pattern of HSP70 mRNA accu-

mulation (Fig. 6) and formation of the constitutive complex that the factor binding to HSE forming the s1 complex instressed oocytes is related to mammalian HSF1. The stress-(Figs. 3 and 4), implying a role for this factor in the normal

expression of HSP genes during oogenesis. inducible properties of oocyte HSF1 detected in these exper-iments, both in DNA-binding ability and transactivationpotential (Figs. 1 and 2), closely resemble that of vertebrate

Stress-Inducible HSE Binding in Xenopus Oocytes HSF1 (Baler et al., 1993; Sarge et al., 1993). Furthermore,antibody recognition experiments using antiserum directedExperiments with stage VI oocytes show that stress-in-

ducible HSE-binding ability of endogenous oocyte HSF mol- against XHSF1 suggest the factor forming the heat-induciblecomplex is related to XHSF1 (data not shown), for which aecules is detectable using a standard gel mobility shift assay

(Figs. 1 and 2). To our knowledge the only previous demon- cDNA clone has recently been isolated (Stump et al., 1995).However, our analysis falls short of demonstrating a directstration of induced HSE-binding during preembryonic verte-

brate development was in mature ovulated mouse oocytes relationship between the XHSF1 cDNA and the gene encod-ing oocyte HSF1. Since XHSF1 cDNA was isolated from(Mezger et al., 1994). Induction of heat shock factors endog-

enous to the oocyte is entirely consistent with earlier re- and A6 cell library (Stump et al., 1995) it remains possiblethat oocyte HSF1 detected in this study is encoded by aports of transcriptional upregulation from injected HSP70

promoters in Xenopus oocytes (Landsberger et al., 1995). different gene.Our experiments suggest that HSF1 is normally presentAlthough we are unable to explain the failure to detect HSE-


AID DB 8441 / 6x15$$$$83 12-09-96 18:37:34 dbas


1 2 3 4 5 6 7 8 9 10

C1 C2 1 2 4 6 8 10 12 24

Time (h) in 500µM AsO2

ns

s1

B

1 2 3 4 5 6 7 8 9 10 11 12

0 12 24 1 2 3 4 6 8 12 18 24

Time (h) in 5 mM CdControlTime (h)

ns

s1

C

FIG. 7—Continued

in an inactive non-DNA-binding state in the oocyte and is oocytes. Landsberger and Wolffe (1995) reported that over-expression of XHSF1 in oocytes leads to inappropriate oligo-negatively regulated at the level of oligomerization under

nonstress conditions (Figs. 1 and 3). As with most cell types, merization and activation of HSP70 transcription undernonstress conditions. This improper regulation could be dueconversion of HSF1 to an active DNA-binding trimer ap-

pears to be an early step in the stress induction pathway in to titration of regulatory molecules that normally function


AID DB 8441 / 6x15$$$$83 12-09-96 18:37:34 dbas

60 Gordon et al.

scription of HSP mRNA in early oocytes and in the absenceof significant HSP mRNA synthesis in stage VI oocyte, de-serves further analysis.

A Developmentally Regulated HSF Activity inEarly Stage Oocytes

In the course of examining stress-inducible HSF activitiesin early stage oocytes, we also observed an unusual consti-tutive HSE-binding activity in unstressed controls. The con-stitutive HSE-binding complex (s2) was detected in un-stressed stage I and II oocytes, but not in later stage oocytes,eggs, embryos, or A6 cells (Figs. 3 and 4). Competition exper-iments established the sequence specificity of both consti-tutive and stress-inducible HSE–protein interactions (Fig.4C). The constitutive activity also was found to comigratewith an HSE–protein complex present in frog testes as wellas an abundant activity in rat ovary tissue (Fig. 4A). Initially

1 2 3 4 5 6

C HS 5 15 30 60

Recovery (min)

ns

s1

we compared this data with that the DNA-binding activityof HSF2 in mammalian tissues where HSF2 is restricted toFIG. 8. Time course of HSE-binding activity in oocytes recovering

from a 1-hr 337C heat shock. Lane 1 contains an untreated control cells undergoing differentiation and development, exhibitsextract (C; lane 1), and lane 2 contains extract obtained from oo- constitutive HSE-binding activity in testes, and is not sig-cytes immediately after heat treatment (HS; lane 2). Oocytes were nificantly induced by stress (Sistonen et al., 1992; Murphyallowed to recover at 187C for the time indicated above each lane et al., 1994; Sarge et al., 1994; Sistonen et al., 1994; Fiorenza(lanes 3–6). Stress-inducible and nonspecific complexes are indi- et al., 1995). The constitutive complex present in oocytescated with arrows (s1 and ns, respectively). behaved in a remarkably similar fashion: it comigrated with

an activity present in testes, exhibited constitutive HSEbinding, and did not appear to be stress-inducible. On thestrength of this comparison, we originally speculated thatto maintain HSF in an inactive monomeric state. In light

of this, the ability to detect DNA-binding activities of en- the constitutive complex observed in stage I and II oocytesis formed by a Xenopus homologue of HSF2. However, anti-dogenous oocyte HSF molecules greatly enhances the po-

tential of the oocyte as a model system to determine the body recognition experiments using antisera directedagainst mouse HSF1 and HSF2 in conjunction with gel mo-fundamental biochemical events and signal transduction

pathways of the stress response, especially since problems bility shift assays (Fig. 5B) provide strong evidence that thisfactor is unrelated to either HSF1 or HSF2. We propose thatassociated with the regulation of exogenous HSF can be

avoided. It is also noteworthy that HSE-binding activity is the s2 complex is formed by a novel developmentally regu-lated HSE-binding factor, but caution that the present analy-readily detectable in extract from a single oocyte, whereas

approximately 100 mature ovulated mouse oocytes are re- sis cannot rule out the possibility that this factor is relatedto HSF2 or negate the presence of a gene encoding an HSF2quired for detection (Mezger et al., 1994).

Experiments showing induction of HSF1 activity by heat homologue in Xenopus. In order to positively identify oo-cyte HSFs, we are presently attempting to isolate genes en-and chemical stress in early stage oocytes (Fig. 3) represent

the first demonstration that oocytes at all stages of develop- coding HSF family members from a Xenopus ovary cDNAlibrary. Clearly though, the implication of the present studyment are capable of a rigorous stress response at the level

of HSF trimerization. It is difficult to comment on the bio- is that additional HSF family members are expressed inoocytes and adult tissues of Xenopus.logical significance of HSF1 induction during oogenesis,

since HSPs are not expressed in response to stress in stage The detection of a constitutive HSE-binding activity inrat ovary and stage I and II oocytes suggests that HSP genesVI oocytes (Horrell et al., 1987) and since it is not known

whether HSP synthesis is upregulated by stress in early may be regulated during vertebrate oogenesis by a develop-mentally regulated HSF. A comparison between our datastage oocytes at either the transcriptional or translational

levels. In addition, very little is known about the early bio- showing high levels of constitutive HSE-binding in earlydeveloping oocytes, with the activity of HSF2 in cells ofchemical events of the stress response in the female germ

line as previous studies in Xenopus have focused primarily the testes (Sarge et al., 1994), gives rise to the intriguingpossibility that HSF family members may play related roleson the activities of exogenous proteins in stage VI oocytes

(Baler et al., 1993; Zou et al., 1994, 1995; Landsberger and in both male and female germ lines. As to the potentialfunction of the constitutively active HSF, we find it inter-Wolffe, 1995; Stump et al., 1995). It is clear, therefore, that

the potential role of stress-inducible HSF, both in the tran- esting that its HSE-binding activity coincides with the pres-


AID DB 8441 / 6x15$$$$83 12-09-96 18:37:34 dbas


reporter constructs, and Jeff Guttmann for assistance with figures.ence of HSP70 mRNA in stage I and II oocytes (Fig. 6).This work was supported by grants from the Medical ResearchFurthermore, this activity coincides with the most activeFoundation of Saskatchewan and the Medical Research Council ofperiod of mRNA synthesis in oocytes; most of the poly(A)/Canada to N.O. A.A. was supported by a postdoctoral fellowshipmRNA content of oocytes is synthesized by stage II andfrom the Health Sciences Utilization Research Council of Sas-turns over very slowly throughout the remainder of oogen-katchewan, and A.H. was supported by an NSERC (Canada) sum-

esis (Perlman and Rosbash, 1978; Dolecki and Smith, 1979). mer research scholarship.Therefore, it is likely that the constitutively active HSF isa positive regulator of HSP expression during very earlystages of oogenesis when the demand for cellular chaper- REFERENCESones is very high, and it will be interesting to determineof the mechanisms controlling the stage specificity of this

Ananthan, J., Golberg, A. L., and Voellmy, R. (1986). Abnormalfactor. proteins serve as eukaryotic stress signals and trigger the activa-

tion of heat shock genes. Science 232, 522–525.Baler, R., Welch, W. J., and Voellmy, R. (1992). Heat shock geneAttenuation of Inducible HSF Activity in Oocytes

regulation by nascent polypeptides and denatured proteins: hsp70Analysis of the kinetics of HSF1 activation in oocytes as a potential autoregulatory factor. J. Cell. Biol. 117, 1151–1159.

showed a rapid resumption to control HSE-binding levels Baler, R., Dahl, G., and Voellmy, R. (1993). Activation of humanupon return to normal temperatures (Fig. 8), but a tempo- heat shock genes is accompanied by oligomerization, modifica-rally extended activation profile upon continuous exposure tion, and rapid translocation of heat shock transcription factor

HSF1. Mol. Cell. Biol. 13, 2486–2496.of oocytes to heat, cadmium, and arsenite (Fig. 7). This pro-Baler, R., Zou, J., and Voellmy, R. (1996). Evidence for a role oflonged pattern of activation is fundamentally different than

hsp70 in the regulation of the heat shock response of mammalianthe transient induction of HSE binding previously reportedcells. Cell Stress Chaperones 1, 33–39.in embryos subjected to identical conditions (Ovsenek and

Bienz, M. (1984a). Xenopus hsp70 genes are constitutively ex-Heikkila, 1990; Karn et al., 1992). The rapid decline of HSE-pressed in injected oocytes. EMBO J. 3, 2477–2483.binding levels during recovery indicates that a mechanism

Bienz, M. (1984b). Developmental control of the heat shock re-for attenuation of HSF1 activity by trimer disassembly is sponse in Xenopus. Proc. Natl. Acad. Sci. USA 81, 3138–3142.present in the oocyte. It appears, however, that the cellular Bienz, M. (1986). A CAAT box confers cell-type specific regulationmachinery performing this task in oocytes is unable to over- of the Xenopus hsp70 gene in oocytes. Cell 46, 1037–1042.come the effects of continuous stress. Interestingly, HSF1 is Bienz, M., and Gurdon, J. B. (1982). The heat shock response in

Xenopus oocytes is controlled at the translational level. Cell 29,transiently activated in embryos under similar continuous811–819.stress regimes, even prior to the onset of HSP mRNA syn-

Clos, J., Westwood, J. T., Becker, P. B., Wilson, S., Lambert, K., andthesis and acquisition of thermotolerance at the midblas-Wu, C. (1990). Molecular cloning and expression of a hexamerictula stage (Heikkila et al., 1985; Ovsenek and Heikkila,Drosophila heat shock factor subject to negative regulation. Cell1990). The simplest interpretation of the difference between63, 1085–1097.oocytes and embryos is that a cellular factor(s) regulating

Dignam, J. D., Lebovitz, R. M., and Roeder, R. G. (1983). Accurateattenuation of the stress response and disassembly of HSF1 transcription initiation by RNA polymerase II in a soluble extracttrimers is either limited or postranslationally modified in from isolated mammalian nuclei. Nucleic Acids Res. 11, 1475–the oocyte. We propose that the delayed attenuation of 1489.HSF1 activity in stressed oocytes is linked to the low level Dolecki, G. J., and Smith, L. D. (1979). Poly (A/) RNA metabolism

during oogenesis in Xenopus laevis. Dev. Biol. 69, 217–236.of induced HSP synthesis (Horrell et al., 1987). In this re-Dumont, J. B. (1972). Oogenesis in Xenopus laevis (Daudin). I.gard, the idea that HSF1 oligomerization and acquisition of

Stages of oocyte development in laboratory maintained animals.DNA binding activity is regulated by intracellular levels ofJ. Morphol. 136, 153–180.HSP70 family proteins has been the subject of considerable

Fiorenza, M. T., Farkas, T., Dissing, M., Kolding, D., and Zimarino,speculation (Morimoto, 1993, and references therein). InV. (1995). Complex expression of murine heat shock transcrip-light of this, inefficient attenuation of HSF1 activity in thetion factors. Nucleic Acids Res. 23, 467–474.

absence of HSP synthesis in the oocyte would be consistent Goodson, M. L., Park-Sarge, O.-K., and Sarge, K. D. (1995). Tissue-with a potential role for HSP70 in the disassembly of HSF dependent expression of heat shock factor 2 isoforms with dis-trimers and attenuation of the stress response (Mosser et tinct transcriptional activities. Mol. Cell. Biol. 15, 5288–5293.al., 1993; Rabindran et al., 1995). Future experiments will Green, M., Schuetz, T. J., Sullivan, E. K., and Kingston, R. E. (1995).

A heat shock-responsive domain of human HSF1 that regulatesbe aimed at determining the potential role of HSPs on thetranscription activation domain function. Mol. Cell. Biol. 15,activities of endogenous HSF1 in Xenopus oocytes.3354–3362.

Harrison, C. J., Bohm, A. A., and Nelson, H. C. M. (1994). Crystalstructure of the DNA binding domain of the heat shock transcrip-ACKNOWLEDGMENTStion factor. Science 263, 224–227.

Heikkila, J. J., Kloc, M., Bury, J., Schultz, G. A., and Browder, L. W.The authors thank Brenda Bartnik for supplying rat ovary tissue,Kevin Sarge for HSF1 and HSF2 antisera, Alan Wolffe for CAT (1985). Acquisition of the heat shock response and thermotoler-


AID DB 8441 / 6x15$$$$84 12-09-96 18:37:34 dbas

62 Gordon et al.

ance during early development of Xenopus laevis. Devl. Biol. heat inducible before the midblastula transition. Development110, 427–433.107, 483–489.

Heikkila, J. J., Ovsenek, N., and Krone, P. (1987). Examination of Ovsenek, N., Williams, G. T., Morimoto, R. I., and Heikkila, J. J.(1990). Cis-acting sequences and trans-acting factors required forheat shock protein mRNA accumulation in early Xenopus em-

bryos. Biochem. Cell Biol. 65, 87–94. constitutive expression of a microinjected hsp70 gene after themidblastula transition of Xenopus laevis embryogenesis. Dev.Horrell, A., Shuttleworth, J., and Coleman, A. (1987). TranscriptGenet. 11, 97–109.levels and translational control of hsp70 synthesis in Xenopus

oocytes. Genes Dev. 1, 433–444. Peteranderl, R., and Nelson, H. C. M. (1992). Trimerization of theheat shock transcription factor by a triple stranded a-helicalKarn, H., Ovsenek, N., and Heikkila, J. J. (1992). Examination ofcoiled coil. Biochemistry 31, 12272–12276.the DNA sequence specific binding properties of heat shock tran-

scription factor in Xenopus laevis embryos. Biochem. Cell Biol. Rabindran, S. K., Giorgi, G., Clos, J., and Wu, C. (1991). Molecularcloning and expression of a human heat shock factor, HSF1. Proc.70, 1006–1013.Natl. Acad. Sci. USA 88, 6906–6910.King, M. L., and Davis, R. (1987). Do Xenopus oocytes have a heat

shock response? Dev. Biol. 119, 532–539. Rabindran, S. K., Haroun, R. I., Clos, J., Wisniewski, J., and Wu, C.(1993). Regulation of heat shock factor trimer formation: Role ofLandsberger, N., and Wolffe, A. P. (1995). Role of chromatin anda conserved leucine zipper. Science 259, 230–234.Xenopus laevis heat shock transcription factor in regulation of

transcription from the X. laevis hsp70 promoter in vivo. Mol. Rabindran, S. K., Wisniewski, J., Li, L., Li, G. C., and Wu, C. (1994).Interaction between heat shock factor and hsp70 is insufficientCell. Biol. 15, 6013–6024.to supress induction of DNA-binding activity in vivo. Mol. Cell.Landsberger, N., Ranjan, M., Almouzni, G., Stump, D., and Wolffe,Biol. 14, 6552–6520.A. P. (1995). The heat shock response in Xenopus oocytes, em-

bryos, and somatic cells: A regulatory role for chromatin. Dev. Sarge, K. D., Zimarino, V., Holm, K., Wu, C., and Morimoto, R. I.(1991). Cloning and characterization of two mouse heat shockBiol. 170, 62–74.factors with distinct inducible and constitutive DNA-bindingLarson, J. S., Schuetz, T. J., and Kingston, R. E. (1988). Activationability. Genes Dev. 5, 1902–1911.in vitro of sequence-specific DNA binding by a human regulatory

factor. Nature 335, 372–375. [Erratum 336, 184] Sarge, K. D., Murphy, S. P., and Morimoto, R. I. (1993). Activationof heat shock gene transcription by heat shock factor 1 involvesLis, J., and Wu, C. (1993). Protein traffic on the heat shock promoter:oligomerization, acquisition of DNA-binding activity, and nu-Parking, stalling, and trucking along. Cell 74, 1–4.clear localization and can occur in the absence of stress. Mol.Mezger, V., Renard, J-P., Christians, E., and Morange, M. (1994).Cell. Biol. 13, 1392–1407.Detection of heat shock element-binding activities by gel shift

Sarge, K. D., Park-Sarge, O.-K., Kirby, J. D., Mayo, K. E., and Mori-assay during mouse pre-implantation development. Dev. Biol.moto, R. I. (1994). Expression of heat shock factor 2 in mouse165, 627–638.testis: Potential role as a regulator of heat-shock protein expres-Mifflin, L. C., and Cohen, R. E. (1994). hsc70 moderates the heatsion during spermatogenesis. Biol. Reprod. 50, 1334–1343.shock (stress) response in Xenopus laevis oocytes and binds to

Schuetz, T. J., Gallo, G. J., Sheldon, L., Tempst, P., and Kingston,denatured protein inducers. J. Biol. Chem. 269, 15718–15723.R. E. (1991). Isolation of a cDNA for HSF2: Evidence for two heatMosser, D. D., Theodorakis, N. G., and Morimoto, R. I. (1988). Co-shock factor genes in humans. Proc. Natl. Acad. Sci. USA 88,ordinate changes in heat shock element-binding activity and

HSP70 gene transcription rates in human cells. Mol. Cell. Biol. 6911–6915.8, 4736–4744. Shi, Y., Kroeger, P. E., and Morimoto, R. I. (1995). The carboxyl-

Mosser, D. D., Duchaine, J., and Massie, B. (1993). The DNA-bind- terminal domain of heat shock factor 1 is negatively regulateding activity of the human heat shock transcription factor is regu- and stress responsive. Mol. Cell. Biol. 15, 4309–4318.lated in vivo by hsp70. Mol. Cell. Biol. 13, 5427–5438. Sistonen, L., Sarge, K. D., Abravaya, K., Phillips, B., and Morimoto,

Morimoto, R. I., Tissiers, A., and Georgopoulos, C. (Eds.), 1990. R. I. (1992). Activation of heat shock factor 2 (HSF2) during he-‘‘Stress proteins in Biology and Medicine.’’ Cold Spring Harbor min-induced differentiation of human erythroleukemia cells.Laboratory Press, Cold Spring Harbor, NY. Mol. Cell. Biol. 12, 4104–4111.

Morimoto, R. I. (1993). Cells in stress: Transcriptional activation Sistonen, L., Sarge, K. D., and Morimoto, R. I. (1994). Human heatof heat shock genes. Science 259, 1409–1410. shock factor 1 and 2 are differentially activated and can synergis-

tically induce hsp70 gene transcription. Mol. Cell. Biol. 14, 2087–Murphy, S. P., Gorzowski, J. J., Sarge, K. D., and Morimoto, R. I.(1994). Characterization of constitutive HSF2 DNA-binding ac- 2099.tivity in mouse embryonal carcinoma cells. Mol. Cell. Biol. 14, Sorger, P. K., and Pelham, H. R. B. (1988). Yeast heat shock factor5309–5317. is an essential DNA-binding protein that exhibits temperature-

dependent phosphorylation. Cell 54, 855–864.Nakai, A., and Morimoto, R. I. (1993). Characterization of a novelchicken heat shock transcription factor, heat shock factor 3, sug- Sorger, P. K., and Nelson, H. C. M. (1989). Trimerization of a yeastgests a new regulatory pathway. Mol. Cell. Biol. 13, 1983–1997. transcriptional activator via a coiled-coil motif. Cell 59, 807–

813.Nakai, A., Kawazoe, Y., Tanabe, M., Nagata, K., and Morimoto,R. I. (1995). The DNA-binding properties of two heat shock fac- Sorger, P. K. (1990). Yeast heat shock factor contains separable tran-tors, HSF1 and HSF3, are induced in the avian erythroblast cell sient and sustained response transcriptional activators. Cell 62,line HD6. Mol. Cell. Biol. 15, 5268–5278. 793–805.

Nieuwkoop, P. D., and Faber, J. (1956). Normal Table of Xenopus Stump, D. G., Landsberger, N., and Wolffe, A. P. (1995). The cDNAlaevis (Daudin). North Holland, Amsterdam. encoding Xenopus laevis heat-shock factor 1 (XHSF1): Nucleo-

tide sequence, deduced amino acid sequence and properties ofOvsenek, N., and Heikkila, J. J. (1990). DNA sequence specific bind-ing activity of the Xenopus heat shock transcription factor is the encoded protein. Gene 160, 207–211.


AID DB 8441 / 6x15$$$$84 12-09-96 18:37:34 dbas


Vuister, G. W., Kim, S.-J., Wu, C., and Bax, A. (1994). NMR evidence Wiederrecht, G., Shuey, D. J., Kibbe, W. A., and Parker, C. S. (1987).The Saccharomyces and Drosophila heat shock transcription fac-for similarities between the DNA-binding regions of Drosophilators are identical in size and DNA binding properties. Cell 48,melanogaster heat shock factor and the helix-turn-helix and507–515.HNF-3/forkhead families of transcription factors. Biochemistry

Wu, C., Wilson, S., Walker, B., Dawid, I., Paisley, T., Zimarino, V.,33, 10–16.and Ueda, H. Purification and properties of Drosophila heatVoellmy, R., and Rungger, D. (1982). Transcription of a Drosophilashock activator protein. (1987). Science 238, 1247–1253.heat shock gene is heat-induced in Xenopus oocytes. Proc. Natl.

Zuo, J., Baler, R., Dahl, G., and Voellmy, R. (1994). Activation ofAcad. Sci. USA 79, 1776–1780.the DNA-binding ability of human heat shock factor 1 may in-Wallace, R. A., Jared, D. W., Dumont, J. N., and Sega, M. W. (1973).volve the transition from an intramolecular to an intermolecularProtein incorporation by isolated amphibian oocytes. III. Opti-triple-stranded coiled-coil structure. Mol. Cell. Biol. 14, 7557–mum incubation conditions. J. Exp. Zool. 18, 321–334.7568.Westwood, J. T., Clos, J., and Wu, C. (1991). Stress-induced oligo-

Zuo, J., Rungger, D., and Voellmy, R. (1995). Multiple levels ofmerization and chromosomal relocalization of heat-shock factor.regulation of human heat shock transcription factor 1. Mol. Cell.Nature 353, 822–827.Biol. 15, 4319–4330.Westwood, J. T., and Wu, C. (1993). Activation of Drosophila heat

shock factor: Conformational change associated with a monomer Received for publication April 17, 1996Accepted October 11, 1996to-trimer transition. Mol. Cell. Biol. 13, 3481–3486.


AID DB 8441 / 6x15$$$$84 12-09-96 18:37:34 dbas

Distinct Stress-Inducible and Developmentally Regulated Heat Shock Transcription Factors...

Documents

Transcript of Distinct Stress-Inducible and Developmentally Regulated Heat Shock Transcription Factors...