JOIJRNAL OF CELLULAR PHYSIOL,OGY 152:568-577 (19921

Transforming Growth Factor-pl Rapidly Induces Hsp70 and Hsp9O Molecular Chaperones in Cultured Chicken Embryo Cells IVONE M. TAKENAKA AND LAWRENCE E. HIGHTOWER* Department of Molecular and Cell Brology, / hc UnrvProty ot ConncctmJt, Storrc, Connecticut 06269- 3044 In this report we show that: ( 1 ) molecular chaperones in the heat shock protein (hsp) family are a new class of cellular proteins induced by Transforming Growth Factor-PI (TGFP), a cytokine present in serum, (2)rapid induction of Hsc70 precedes a general increase in protein synthesis and may be a preparatory event, (3) TGFP i s a potent regulator of overall protein synthesis in chicken embryo cells (CEC), and (4) isoforms of tisp90 with different biochemical properties exist, raising the possibility that they may have different functions. TGFP can substit~ite for serum in stimulating synthesis of members ol the Hsp90 and Hsp70 families of stress proteins, whereas other cytokines, including PDGF, FGF, and EGF, were not effective nor did they enhance the stimulatory effect of TGFP on the hsp‘s. Analysis of the induction of hsp’s using one- and two-dimensional polyacrylamide gel electrophore5is indicated that members of the Hsp70 family of molecular chaperones were induced rapidly by TGFP, reaching maximum rates of accumulation by 5 hours of treatment. Total protein synthesk increased more slowly, undergoing an -twofold increase in 24 hours. Using a modified protocol for two-dimensional gel electrophoresis, the Hsp90 protein family was separated into four isoelectric forms, two of which were phosphorylated (Hsp90-2 and -4). These phosphorylated isoforrns turned over faster than the unphosphorylated forms of HspY0. All four isoforms were heat inducible, but only Hsp90-2 and -3 were induced rapidly by TGFP, again within 5 hours of treatment. The effects oiserum on these protein families were similar to those of TGFP, suggesting that this cytokine may be the serum component primarily responsible for up-regulating members of the Hsp90 and Hsp70 families. We hypothesize that cells rapidly increase thcir chaperoning capacity for newly synlhesired polypeptides in preparation for an increase in the rate of synthesis of proteins up-regulated by TGFP. 1992 Wiley-Lis5, Inc.

Molecular chaperones are proteins that facilitate the folding and assembly of other polypeptides in vivo but are not part of the finished complex. The reversible binding of Hsp7O family members to a variety of unfolded proteins suggests t h a t they may function a s molecular chaperones (Ellis and Hemmingsen, 1989; Rothman, 1989). Genetic evidence points to a n essential role for nucleocytoplasmic Hsc70, endoplasmic reticular Grp78, and mitochondrial Hsp70 in facilitating the translocation and subsequent folding of many secretory and mitochondrial proteins (Deshaies et al., 1988; Kang et al., 1990; Vogel et al., 1990). In bacteria, the Hsp7O homologue, DnaK, plays a seminal role in DNA replication by facilitating assembly of the preprimosomal complex (Liberek et al., 1988; Hwang et al., 1990).These experiments and those showing the association of Hsc7O with nascent and newly synthesized polypeptides (Beckmann et al., 1990) suggest that Hsp70 family proteins function both early and late in the folding pathways of many if not most polypeptides in normal cells. The Hsp9O family of chaperones is more restricted in 0 1992 WILEY-LJSS. INC

its interactions with other proteins. Hsp9O associates with a variety of steroid hormone receptors, forming a complex known as the “untransformed” form of these receptors. Upon binding of the hormone, t,his complex dissociates releasing Hsp9O (Sanchez et al., 1985j. Recently, Picard et al. (1990) showed that reduced levels of Hsp9O gene expression compromise steroid receptor activity mainly because Hsp9O seems to facilitate the subsequent response of the aporeceptor to the hormonal signal. Members of the Hsp9O family also transiently associate with a number of retroviral tyrosine kinases (Courtneidge and Bishop, 1982; Brugge, 1986) and other kinds of cellular kinases (Rose et al., 1989).

Received February 18, 1992; accepted April 9,1992.

.KTowhom reprint requestslcorrespondence should be addressed. Ivone M. Takenaka is now at the Department of Molecular Biology, Bristol-Myers Squibb, 5 Research Parkway, Wallingford, CT 06492.

STRESS PROTEINS

The roles described above for hsp’s would fulfill essential requirements for normal, unstressed cells and are carried out by the constitutive or cognate members of these families such a s Hsc70 and inducible members that are present at significant basal levels such as Hsp9O and Grp78. An important consideration in evaluating data on the basal or constitutive levels of expression of hsp’s is that most studies of the vertebrate heat shock response have used cultured cells. Under these conditions, “unstressed” cells are somewhat arbitrarily defined as those not subjected to stresses beyond the unavoidable disturbances associated with the establishment and maintenance of cells in culture. In particular, the serum supplements of cell culture media contain components such as platelet-derived cytokines a t substantially higher concentrations than found in blood plasma and healthy tissues, some of which might cause increases in expression of certain members of heat shock gene families. During a previous study of metal ions and other stressors in cultured CEC, it was noticed that the rates of accumulation of Hsp9O and Hsp70 family proteins decreased in serum-free cultures used a s controls (unpublished observations; Whelan and Hightower, 1985). This observation suggested that one or more components of serum were responsible for maintaining the basal levels of expression of certain heat shock genes in cultured CEC. Herein, we show that of the major known cytokines found in serum, TGF-6 is the only one that regulates the levels of members of the Hsp7O and Hsp9O families in cultured CEC. Although this cytokine was originally identified a s a factor that allows certain types of anchorage-dependent cells to proliferate without a substratum, TGF-@more frequently has growth inhibitory effects on cultured cells. This cytokine is present in tissues containing rapidly dividing cells of mesodermal origin during embryonic development where TGF-fi may function a s a master switch to set in motion a cascade of events that first prepare and ultimately stimulate certain embryonic cell types into rapid proliferation and others into differentiation pathways (reviewed by Nilsen-Hamilton, 1990). TGF-@also plays a major role in wound responses of somatic tissues where i t may promote chemotaxis and extracellular matrix production (reviewed by Sporn e t al., 1987). This spectrum of responses is determined in part by the presence of other growth factors, by the intracellular metabolic state, by the kinds of receptors present on the cell surface, and by the influence of structural elements such as extracellular matrix (ECM) (Feng et al., 1988). TGF-P is known to regulate the expression of two broad categories of genes: (1) growth and differentiation factors (Thompson et al., 1988), and (2) ECM components including structural proteins such as collagen and fibronectin, cellular membrane receptors called integrins (Ignotz et al., 1989), and matrix proteases and protease inhibitors (reviewed by Nilsen-Hamilton, 1990). In the present study, evidence of a third category of TGFp regulated genes is presented, a category composed of molecular chaperones that may be necessary for production of proteins encoded by genes in the other two categories. A preliminary report has appeared in a monograph (Takenaka et al., 1991). In a paper to follow, we will show that TGFp regulates the levels of

569

Hsp70 and Hsp9O family members posttranscriptionally.

MATERIALS AND METHODS Cell culture Primary cultures were prepared from 10-day-old chicken embryos by the methods of Sekellick and Marcus (1986). Confluent primary cultures incubated a t 37°C in a humidified, 5% CO, atmosphere were disrupted with trypsin (0.05% w/v)-EDTA (0.53 mM) solution and resuspended in Nutrient Colorado Inositol (NCI) medium supplemented with 6% calf serum. Secondary CEC were grown to confluency on either 35mm plates or 150mm plates and used for experiments 2 days postseeding. Unless otherwise indicated, secondary CEC, prepared a s indicated above, were washed with Eagle’s Minimal Essential Medium containing Earle’s salts (MEM) and preincubated for 12 hours in MEM without serum. Sample-specific conditions were designated as follows: after preincubation, cells were incubated in either fresh MEM (designated minus serum, -S), or in MEM containing bovine TGFp a t a concentration of 10 ngiml or in MEM supplemented with 6% calf serum (designated plus serum, +S). All cultures were incubated for 5 hours. Platelet-derived growth factor (PDGF), epidermal growth factor (EGF), and basic fibroblast growth factor IFGF) were kindly provided by L. Wakefield and M. Sporn. TGFpl was a gift from Collagen Corporation. Protein analyses using polyacrylamide gels Secondary CEC were grown to confluency on 35mm plates and treated as indicated. Cells were labeled with Selectamine medium (Gibco) containing 2% (viv) dialysed calf serum, methionine (0.15 pgiml, i.e., 11100 of the normal concentration) and “S-methionine (ICN, 1,100 Ci/mM) at 10 pCi/ml for 30 minutes at 37°C. For sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE), cells were lysed with a buffer containing 2.3% ( w h ) SDS, 10% (viv) glycerol, 5% (viv) P-mercaptoethanol and 62.5 mM Tris-HC1 pH 6.8. Either equal volumes of lysates or equal counts were loaded onto 7.5% (wiv) acrylamide: 0.075% ( w h ) bisacrylamide slab gels using a modification (Hightower, 1980) of a published procedure (Laemmli, 1970). Gels were treated with 2,5-diphenyloxazole (PPO)-dimethyl sulfoxide (DMSO) in preparation for fluorography (Laskey and Mills, 1975). Dried gels were exposed to preflashed X-ray films (Kodak) at -70°C and fluorograms were quantified by laser densitometry (LKB Ultroscan XL). For isoelectric focusing-SDS-PAGE (IEF-SDSPAGE), cells were lysed in a modified O’Farrell buffer (O’Farrell, 1975; Duncan and Hershey, 1984) consisting of 9.5 M urea, 2% (wiv) nonidet P-40 (NP-40), 1.6% (viv) pH 3.5-10.0 and 0.4% (v1v) pH P 6 ampholytes (LKB), 5% (viv) P-mercaptoethanol and 62.5 mM TrisHC1 pH 6.8. Aliquots of lysates (6-7 Fg of protein as determined by the Bradford (1976) method were treated with RNAse A1 (50 pg/ml) for 15 minutes a t room temperature, and equal amounts of lysates were loadedonto a cylindrical gel containing 3.5% (wiv)acryl-

570

TAKENAKA ANT) HIGHTOWER

amide and 0.2% (wiv) bis-acrylamide, 0.4% (viv) pH 4-6 and 1.6% (viv) pH 3.5-10.0 ampholytes in O’Farrell buffer. The gels were run for -13,000 Volt-hrs using 50 mM NaOH a s the upper buffer and 25 mMH,PO, as the lower buffer. The cylindrical gels were then mounted on polyacrylamide slab gels containing 7.5%:0.075%acrylamide:bis-acrylamide for SDS-PAGE a s a second dimension. Fluorograms were obtained and analysed by laser densitometry (Molecular Dynamics Model 300A Computing Densitometer). For ulse-chase experiments, samples were labeled with FS-methionine at a concentration of 200 KCiiml for 5 minutes, washed, and chased with MEM supplemented with l o x methionine (150 pgiml) for 5 hours. Samples were analyzed by IEF-SDS-PAGE as described above.

Protein phosphorylation CEC were labeled for 6 hours in phosphate-free MEM containing 32P-orthophosphate (ICN, 8,500-9,000 Cii mM) at 200 pCi/ml. Cells were lysed in O’Farrell buffer, aliquots were precipitated with trichloroacetic acid, spun in a microcentrifuge for 1 minute, and the pellet was resuspended in O’Farrell buffer. Proteins were analyzed by IEF-SDS-PAGE and autoradiograms were obtained. Western blots Cells treated a s indicated were lysed and proteins were separated by IEF-SDS-PAGE. Gels were blotted onto a nitrocellulose membrane using 25 mM Tris-base, 192 mM glycine, 20% (viv) methanol for 3 hours at 100 Volts a t 4°C. Antibody-binding and visualization was performed using recommended manufacturer’s procedures for the biotinylated alkaline phosphatase reaction kit (Vector Labs). Monoclonal antibody 7.10 raised against Drosophila Hsp70, which recognizes a conserved epitope present in all but the mitochondria1 members of the Hsp7O family (kindly provided by Susan Lindquist) (Kurtz et al., 1986) or monoclonal antibody 7 D l l raised against chicken Hsp9O (kindly provided by David Toft) were used.

RESULTS Screen of serum cytokines In order to determine whether one or more of the cytokines normally found in serum is responsible for the previously observed serum effects on hsp synthesis, several cell types, including HeLa cells (obtained from the American Type Culture Collection), a spontaneously transformed chicken cell line designated LSCCH32 (Kaaden et al., 19821, r a t embryo cells (REC), secondary CEC, the human fibroblast strain FS-4, and the mouse fibroblast cell line L929, were screened with TGFP, PDGF, EGF (which shares a receptor with transforming growth factor a [TGFa]), and FGF. TGFa, TGFP, and PDGF are factors released from platelets during the process of blood clotting. Under the conditions tested, only TGFp significantly increased the accumulation of newly synthesized Hsp70 and Hsp9O family members and only in CEC (Fig. 1).Since the chicken cognate (Hsc7O) and inducible (Hsp7O) forms were not separated from one another on these gels, the composite band was marked simply Hsp7O. Each gel

lane was loaded with approximately equal amounts of cell protein. A qualitative comparison of the samples from cultures treated with serum (lane 2) and TGFp (lanes 3 and 4) relative to control cplls without serum (lane 1) showed a n overall increase in 3’S-methionine incorporation into proteins and substantial increases in newly synthesized Hsp7O and HspSO. Cultures treated with PDGF (lanes 5 and 6) or EGF (lanes 7 and 8) or FGF (lanes 13 and 14) did not show this stimulation. Addition of other platelet growth factors such as PDGF (lane 9), EGF (lane lo), or both (lane 11)along with TGFp did not enhance the effect of the latter factor alone. Combinations of factors other than TGFP, such as PDGF plus EGF (lane 12), did not stimulate protein synthesis or induce hsp’s. In fact, the combinations of either PDGF or EGF and TGFp did not show the stimulated incorporation of 3’SS-methionineexhibited by the latter cytokine alone.

Effects of serum deprivation and TGFp on protein synthesis Our initial experiments on the effects of cytokines on protein synthesis by CEC were done using cells grown to confluency in the presence of serum, washed with serum-free medium, and incubated with cytokines in the absence of serum, e.g., Figure 1.Under these conditions, the basal rates of synthesis of Hsp9O and Hspi Hsc7O were substantial in the control cultures incubated in serum-free medium, which interfered with efforts to quantify the induction of hsp’s by TGF-P. A likely cause was residual serum still present in the control cultures. Therefore, the effect of longer periods of serum deprivation on protein synthesis was estimated by ”S-methionine incorporation into newly synthesized proteins (Fig. 2). Confluent CEC incubated for various periods of time in -S medium showed the lowest rate of protein synthesis, estimated by radioisotopic incorporation as a 2.2-fold decrease, by 12 hours of serum deprivation. Incubation under - S conditions for a n additional 24 hours did not change the rate of protein synthesis (not shown). Furthermore, other cultures incubated in parallel without serum showed substantial decreases by 12 hours in accumulation of newly synthesized Hsp9O and Hsp70 when their proteins were analyzed by SDS-PAGE (data not shown). Therefore, serum depletion was routinely carried out for 12 hours. When TGFp was then added to CEC deprived of serum for 12 hours (marked by arrow in Fig. 2). total radioisotopic incorporation increased steadily, achieving a 2.5-fold increase after 24 hours of treatment relative to the 12 hours -S cultures. A more detailed analysis of the time course of the stimulation of protein synthesis in confluent, serum depleted CEC treated with TGFp (10 ng/ml) was carried out using SDS-PAGE to separate newly synthesized proteins pulse-labeled with 35S-methionine (Fig. 3). The effect of TGFp on protein synthesis in CEC shown in Figure 3 was quantified by scanning densitometry of fluorograms. The absorbance of each gel lane was normalized to the same absorbance units-full scale (AUFS). Peak areas of several size classes of proteins were determined and percentage of maximum absorbance was plotted against time. The term “size class” is used here to indicate t h a t each peak may contain more than one kind of protein.

571

STRESS PROTEINS

1 2

3

4

5

6

7 8

91011121314

HSP9OHSP70,

Fig. 1. Effects of growth factors on CEC. CEC were grown to confluency in NCI medium containing 6% calf serum. Cultures were rinsed three times with MEM, incubated for 30 minutes in the same medium, and then incubated for 6 hours with MEM containing several different cytokines. Cells were labeled with 3%methionine (10 p.Ciiml) for 30 minutes after the 6 hours of cytokine treatment. Lysates were made and equal amounts were loaded onto a 9% acrylamide: 0.2% bis-acrylamide gel. After SDS-PAGE, the gels were prepared for fluorography.

CI

f x E

8

O!

0

10

20

30

40

TIME (hrs)

Pig. 2. Analysis of total protein synthesis under serum deprivation and TGFp treatment on CEC. Secondary CEC grown in NCI plus 6% calf serum were washed with MEM and incubated for various periods of time in -S mcdiurn. At 12 hours of serum deprivation, TGFlj (10 ngiml) was added (time marked by arrow) and further incubated for various periods of time. Samples, collected a t 1, 3, 5, 9, 12, 13, 15, 17. 24, and 36 hours, were labeled with 3SS-methionine(20 pCi/ml) for 30 minutes, lysed with 0.1 N NaOH and TCA precipitated. Total radioisotopic incorporation into acid precipitable material was determined by scintillation counting. The cpm of duplicate samples wcre averaged and plotted against time. Since confluent CEC cultures contain the same number of cells and amount of protein per plate, it was not necessary to do protein determinations (Hightower and Bratt, 1974).

As shown in Figure 4A, the 70 kDa size class that includes HspiHsc70 (closed circles) was synthesized very rapidly, reaching a maximum rate by 3 hours after the addition of TCFp and then decreasing slowly. In contrast, the 90 kDa size class that includes Hsp9O (closed squares) was synthesized more slowly than Hsp/ Hsc7O in the first 5 hours, but its synthesis continued to increase for a t least 24 hours of TGFP treatment. The

Lane 1, MEM (-S); Lane 2, MEM containing 6% calf serum; Lane 3, TGFp (10 ngiml); Lane 4, TGFp (50 ngiml); Lane 5, PDGF (10 ngiml); Lane 6, PDGF (50 nglml); Lane 7, EGF (5 ngiml!; Lane 8, EGF 125 nglml); Lane 9, TGFB plus PDGF ihoth a t 10 ngiml); Lane 10, TGFP pius EGF (10 and 5 ngiml, respectively); Lane 11, TGFp plus PDGF plus EGF (10, 10 and 5 ngiml respectively); Lane 12, PDOP plus EGP (1 0 and 5 ng:ml, respectively); Lane 13, FGF (5 nglml): Lane 14, FGF (25 ngimll.

synthesis of the 78 kDa size class that includes Grp78 (open squares) roughly followed that of the 90 kDa proteins. For a more thorough survey of the effects of TGFp on protein synthesis, overall rates of protein synthesis, as estimated by the total absorbance in each gel lane, as well as a selection of additional size classes of proteins were analyzed. For the purpose of these analyses, unknown proteins were named according to their apparent molecular masses in kilodaltons as determined by SDS-PAGE. The overall rate of protein synthesis (Fig. 4B, closed squares) increased two- to threefold over 24 hours. However, not all the patterns of induction of different size classes of proteins were the same. For instance, actin (Fig. 4B, open squares), p35 (Fig. 4B, open circles), and p85 (Fig. 4C, closed squares) accumulated slowly with a pattern similar to Grp78, Hsp90, and total protein synthesis, reaching the maximum only after 24 hours of TGFp treatment. Synthesis of p210 (Fig. 4C, closed circles) was not affected during the first 5 hours but underwent a delayed increase between 5 and 10 hours, whereas synthesis of p200 (Fig. 4C, open circles) declined during the first 5 hours and then slowly recovered. The rapid synthesis of Hsp/ Hsc7O represents yet another response pattern.

Separation of the isoforms of Hsp7O and Hsp9O Although several laboratories have reported isoforms of Hsp9O with different molecular masses in mice and humans, these isoforms are difficult to separate, and the characteristics described for Hsp9O usually refer to the mixture of isoforms. Chicken Hsp9O poses even greater difficulties because the isoforms have approximately the same molecular masses and often focus poorly during IEF. In order to overcome these problems, a modified IEF-SDS-PAGE method was developed that separates Hsp9O as well as Hsp70 isoforms. Using this method, chicken Hsp9O was separated into four different isoforms as shown by the protein pattern from 35S-

TAKENAKA AND HIGHTOWER

572 HS 0 1 3

5 12 24

120

I

A

1

-P210 -200

HSP90, 0

HSdHSC70t 120

I

I

10

20

I 30

I

1

B

-actin

4P35

Fig. 3. Analysis by SDS-PAGE of the time course of induction of Hsp9O and Hsp7O by TGFP. CEX were incubated for 12 hours in - S medium (0 hr), and then further incubated with TGFp (10 ngiml) for 1, 3, 5, 12, and 24 hours. Cells were labeled with ”S-methionine (20 WCiiml) for 30 minutes at the end of each cytokine treatment period at 37’C and equal amounts of cell lysate were loaded per well. An extract from CEC incubated for 1 hour at 44°C prior to labeling was included in the lane marked €IS as a source of authentic hsp’s for comparison. Proteins were separated by SUS-PAGE and fluorographs were obtained as described in Materials and Methods. Proteins were labeled using their molecular masses in kilodaltons (HSP heat shock protein, HSC = heat shock cognate, and GRP = glucose-regulated protein). Actin and the protein size classes marked p210, p200, p85, and p35 were included in the quantification in Figure 4.

0

10

20

30

methionine labeled control cultures (- S) in Figure 5A. Small closed arrowheads mark Hsp9O isoforms that have been designated 90-1,90-2,90-3,and 90-4 starting from the basic side of the gel. The Hsp9O isoform pattern was affected strongly by the amount of protein & O loaded on the gels and larger amounts of proteins re0 10 20 30 sulted in streaking and incomplete focusing. At lower TIME (hrs) concentrations, the pattern shown was highly reproducible. All four isoforms were induced by heat shock (Fig. 5B). Fig. 4. Quantification of the effect of TGFp on the accumulation of Hsp9O is known to be a phosphorylated protein various proteins. The fluorograph of Figure 3 was scanned using a (Brugge et al., 1981; Welch et al., 1983; Walker et al., densitometer. The absorhances of each lane were normalized to the same absorbance units full-scale (AUFS) value. Peak areas (AU)were 1985; Legagneux et al., 1988; Lees-Miller and Ander- plotted as percentage of maximum absorbance against time of TGFp son, 1989a, b). In order to determine which of the treatment. Unknown proteins designated p35, p85, p200, and p210 chicken Hsp9O isoforms are phosphorylated, CEC were were named according to their apparent molecular masses in SDSlabeled continuously for 6 hours with 32P-orthophos- PAGE (positions marked in Fig. 3 ) . A. Hsp70 is represented by closed Grp78 by open squares; Hsp9O by closed squares. B. Total phate in phosphate-free MEM and proteins were sepa- circles; absorbance of each lane is represented by closed squares; Actin by rated by IEF-SDS-PAGE (Fig. 5C). Two spots corre- open squares; p35 by open circles. C. p85 is closed squares; p200 is sponding to 90-2 and 90-4 were phosphorylated. Spot open circles; p210 is closed circles. 90-3 contained little or no phosphate and the phosphorylation state of 90-1 could not be determined unambiguously due to the high 32P-background in the same In order t o assure that all of these spots are members region of the gel. Grp78 (Fig. 5C, spot d) was also phos- of either the Hsp9O or Hsp7O families, these proteins phorylated as reported previously (Welch et al., 1983; were analyzed by immunoblotting. The same four Lee et al., 1984). Hsp9O isoforms were recognized by the antichicken

“ 1

STRESS PROTEINS

SDS-PAGE

V

IEF,

H+

A.

573

H+

490 -78 470

C.

D.

Fig. 5. Identification of Hsp9O and Hsp7O family of proteins on twodimensional gels. Hsp9O and Hsp70 protein families were separated by IEF-SDS-PAGE as described in Materials and Methods. Portions of gels including the 70-90 kDa region are shown. Small arrowheads numbered 1 to 4 are Hsp9O isoforms. Large arrowheads marked a to c are Hsp7O isoforms. Letter d indicates Grp78. Small open arrowheads indicate two 90 kD proteins that are probably not part of Hsp9O family. H i is the acidic side of the gel. A. Fluorograph of a -S culture sample labeled with 35S-methionine(100 pCi/ml) for 30 minutes. B. Fluorobp-aph of a - S culture heat-shocked for 1 hour at 44"C, and labeled as described above. C. Fluorograph of a -S culture labeled with 32P-orthophosphate 1200 kCi!ml) for 6 hours. Cells were lysed with O'Farrell buffer and TCA precipitated. The resulting pellets were resuspended in O'Farrell buffer and separated by IEP-SDS-PAGE. D. Western blot of a -Sculture sample separated by IEF-SDS-PAGE and reacted with an antichicken Hsp9O monoclonal antibody (7Dll). Protein complexes were visualized using an alkaline phosphatase kit (Vectastain).

Hsp9O monoclonal antibody 7 D l l (Fig. 5D). The two 90 kDa proteins, indicated by the open arrowheads in Figure 5A, B are probably not Hsp9O family members since they were not recognized by monoclonal antibody 7D11 and their levels did not change with TGFP, serum or heat shock treatments. In the fluorogram shown in Figure 5A, the Hsp70 family was separated into the heat-inducible Hsp7O (spot a),which is the most basic isoform, and the constitutive or cognate form Hsc7O (spot c), the most acidic form. Spot b, which was located between Hsp70 and Hsc70, may be the mitochondria1 form, which is not recognized by the 7.10 monoclonal antibody. The inducible Hsp70, Hsc7O and Grp78 (spot d) were all recognized by the monoclonal antibody 7.10 in immunoblots (data not shown). There are several features of the chicken Hsp70 family shown in Figure 5A, B that distinguish it from the more thoroughly studied mammalian Hsp7O family. First, Hsp70, Hsc70, and Grp78 are all synthesized a t relatively high basal or constitutive rates in CEC under normal culture conditions. Second. these three proteins are all heat inducible in CEC, but the magnitude of induction is much greater for Hsp7O than for Hsc7O and Grp78. Thus for chicken cells, the distinctions between cognate and inducible members of this hsp family are

-90 -78 -70

-ac

Fig. 6. EfYect of TGFp on Hsp90 and Hsp70 isofurms separated on two-dimensional gels. - S and TGFB treated cultures (5hrs exposure) were labeled with '"S-methionine (100 +Ci/ml) for 30 minutes a t 37"C, lysates were prepared and proteins from equal volumes of lysate were separated by IEF-SDS-PAGE as described. A. Fluorograph of -S control. B. TGFB treated sample. Actin is indicated by "ac." Hsp70 and Hsp9O isoforms are marked as in Figure 5. Open arrowheads mark same proteins as in Figure 5.

not a s clearcut as for mammalian cells. In rodent cells in culture, basal expression of Hsp70 is undetectable (Hightower and White, 19811, whereas the situation in human and monkey cells is intermediate with Hsp70 synthesized a t a detectable basal rate (Welch et al., 1983; Milarski and Morimoto, 1986; Ferris et al., 1988).

Effect of TGFp on Hsp9O and Hsp7O isoforms Having separated and identified Hsp9O isoforms using IEF-SDS-PAGE, we pursued the question of whether TGFp has different effects on Hsp9O and IIsp70 isoforms and whether these isoforms are metabolically stable. "S-methionine labeled proteins from CEC treated as the -S control (Fig. 6A) or with TGFp (Fig. 6B) were analyzed by IEF-SDS-PAGE. All of the Hsp9O isoforms were induced to some extent by TGFP. A qualitative evaluation of the fluorograms indicated that a number of other proteins including Grp78, inducible Hsp70, Hsc70, and aclin (Ac) were also induced by TGFP. However, there were many other proteins whose synthesis was either reduced or not altered by TGFp after 5 hours of treatment, indicating some selectivity in the cohort of proteins induced by this cytokine in CEC. Pulse-chase experiments were performed to determine whether Hsp9O and Hsp7O isoforms were metabolically stable and to detect any precursor-product relationships. CEC were pulse-labeled for 5 minutes as

574

TAKENAKA AND HIGHTOWER

described in Materials and Methods and chased with excess methionine. Proteins were subsequently analysed by IEF-SDS-PAGE and fluorograms were scanned using a densitometer. Twenty-eight other spots consistently present in every gel were quantified and the sum of their absorbances was taken as a reference absorbance for each fluorogram (Figs. 7, 8). The absorbance of each IIsp90 and Hsp70 isoform was expressed as a percentage of the reference absorbance from the same fluorogram. This procedure allows one to follow changes in the fraction of protein synthesis for which each isoform accounts under each of the three treatment conditions, but does not reflect absolute rates of synthesis of hsps between two treatment conditions. All of the data shown in Figures 7 and 8 are from the same representative experiment. The distribution of Hsp9O isoforms is shown in Figure 7. In CEC pulse-labeled for 5 minutes with no chase (Fig. 7A), Hsp9O-1 and Hsp90-4 showed similar relative rates of synthesis in comparison to the reference absorbances under -S (closed columns), TGFp (hatched columns), and +S (open columns) treatments. This result does not mean that synthesis of these two isoforms was not stimulated but rather that the stimulation is no greater than that experienced by the reference proteins. As shown in Figure 2, there was a modest increase in overall protein synthesis in TGFp treated cultures by 5 hours of treatment (and also for +S cultures, Fig. 1, lane 21, and qualitative evaluation of the twodimensional gels shown in Figure 6 indicated that synthesis of all four isoforms was stimulated to some extent. In contrast, the percentage of reference absorbances calculated for Hsp90-2 and Hsp90-3 increased by 3.5fold and twofold under TGFp and + S treatments, respectively, compared to the percentage of reference absorbance calculated for these isoforms for the -S controls. The results of the quantitative two dimensional gel analysis (Fig. 7A) indicate that the rapid induction of Hsp90-2 and 90-3 was masked by the majority of more slowly induced proteins in the 90 kDa size class in the quantitative one dimensional gel analysis (Fig. 4A, closed squares). Our interpretation is that Hsp90-2 and 90-3 synthesis is stimulated much faster than that of the other two isoforms and they belong to the same rapid response class a s HspiHsc70. The levels of both phosphorylated forms Hsp90-2 and Hsp90-4 decreased during the chase period (Fig. 7B). The data in Figure 6 indicated that newly synthesized Hsp90-4 accumulated to higher levels after a 30minute labeling period than after a 5-minute labeling period (Fig. 7A) relative to the other Hsp9O isoforms. This may indicate the existence of both rapidly turning over and more slowly accumulating stable populations of Hsp90-4. Hsp90-3 levels did not change significantly upon chase, indicating that the phosphorylated forms were probably degraded and not simply dephosphorylated, which would have resulted in elevated levels of Hsp9O-1 and 90-3. Because of the high background in the region of Hsp9O-1, we are less confident in its quantification. However, our analysis suggests that this isoform was intermediate in stability. Interestingly, heat shock increases the turnover of phosphate groups on mammalian HspSO, which may be yet another level of

regulation manifested during stress (Legagneux et al., 1991). Among the Hsp70 family (Fig. 8,4), serum (open columns) stimulated synthesis of Hsc7O (marked 70C) and Grp78 but had no effect on Hsp70 (marked 70A). In contrast, TGFp treatment (hatched columns) increased the percentages of Hsp70 and Hsc7O -threefold and Grp78 about twofold compared to the percentages of these proteins in -S controls (closed columns). This result suggests that serum may contain a factor that blocks the stimulatory effect of TGFp on Hsp70 but not on Hsc70 and Grp78. All four of these Hsp7O family members (assuming that 70B is mitochondria1 Hsp70) appeared to be relatively stable during chase (Fig. 8B).

A

90-1

90-2

90-3

90-4

HSP90 ISOFORMS

3

w 2

u z

1

1

W

ez

w

ez Ei?

0

90-1

90-2 90-3 HSP9O ISOFORMS

90-4

Fig. 7. Pulse-chase experiments-Hsp9O: -S contrnl, TGFP or +S CEC treated cultures (5 hrs exposure) were pulse-labeled with "Smethinnine (300 pCi:mll for 5 minutes (0 h chase) (A);or chased with MEM supplemented with 1Ox methionine for 5 hours (5 h chase) (B). Lysates were separated by IEF-SDS-PAGE and fluorographs were scanned using a densitometer. Twenty-eight other spots, present in all gels, were quantified by densitometry and the sums were taken as reference absorbance. The percentage of the reference absorbance was calculated for each hsp isoform and was plotted on the y-axis. Hsp9O isoforms were identified as indicated on previous figures. -S sample is rcpresented by closed columns, TGFP by hatched columns, and +S by open columns.

STRESS PROTEINS

A

70 A

70 B

70 C

Grp78

HSP70 ISOFORMS

70 A

70 B

70 C

Grp78

HSP7O ISOFORMS

Fig. 8. Pulse-chase experiments-Hsp70. Experiments were done and columns designated a s described in the legend of Figure 7. Hsp70 was designated as 70A, the presumptive mitochondria1 Hsp70 as 70B, Hsc70 as 70C, and the 78 kDa glucose-regulated protein as Grp78. Panel A shows the 0 h chase and panel B shows the 5 h chase.

As was the case for two of the Hsp9O isoforms, the early stimulation of Grp78 synthesis was revealed by the higher resolution two-dimensional gel analysis, but not by quantitative one-dimensional gel analysis of 78 kDa size proteins.

DISCUSSION Our data indicate t h a t TGFp is a regulator of protein synthesis in CEC and that the rate of protein synthesis, estimated by incorporation of radioactive methionine into proteins, increases two- to threefold by 24 hours of exposure. Gel analyses revealed that the synthesis of individual proteins may increase, decrease, or remain substantially unchanged. The complexity of the protein synthetic response to TGFp may help to explain the cell type- and differentiation state-specific biological effects of this cytokine, since which proteins are up- or down-

575

regulated will likely depend upon the gene regulatory network of the particular target cell. In cultured CEC, synthesis of the molecular chaperones Hsp70, Hsc70, Grp78, Hsp90-2, and Hsp90-3 are stimulated early during the overall increase in protein synthesis. The rapid up-regulation of molecular chaperones could be part of a n overall increase in cellular protein synthesis. Such a n increase may require as a prelude a rapid increase in the levels of proteins that assist the folding and transport of newly synthesized proteins, since recent work (Beckmann et al., 1990) suggests that many if not all nascent polypeptide chains may interact with Hsc7O. Alternatively, TGFp may up-regulate synthesis of a group of proteins that place a high demand on the cell’s molecular chaperoning capacity. Many of the known proteins stimulated by TGFp are components of ECM, which are folded in the lumen of the endoplasmic reticulum and processed through the secretory pathway. Translocation of a t least some proteins across intracellular membranes requires both Hsc7O in the cytoplasm and Grp78 in the lumen (Vogel et al., 1990). All of the previous studies of serum effects on hsp levels used cultured mammalian cells. Serum-stimulation of serum-deprived HeLa and 293 cells results in increased synthesis and accumulation of Hsp7O (Wu and Morimoto, 1985), which is also regulated temporally during the cell cycle (Milarski and Morimoto, 1986). Both Hsp9O and Hsp70 levels are elevated in human cells by serum, and a promoter element designated the serum regulated element has been implicated (Wu et al., 1987). In serum-deprived human lung cells, a protein designated P72, which is probably Hsc70, increases in the cytoplasmic compartment and has been implicated in translocation of proteins into lysosomes (Chiang et al., 1989). Mitogenic factors such a s phytohemagglutinin A and interleukin-2 increase hsp7O and hsp90 gene expression in human T cells (Ferris et al., 1988). In contrast, growth stimulatory conditions cause a rapid decrease in Hsp70 mRNA levels in freshly isolated (nonproliferative) human peripheral blood mononuclear cells (Kaczmarek et al., 1987). And treatment of human erythroleukemic cells (K562 line) with antiproliferative prostaglandins induces Hsp70 (Santoro et al., 1989). In in vitro cell differentiation models, Hsp7O accumulates in hemin-treated K562 cells (Singh and Yu, 1984; Theodorakis et al., 19891,whereas Hsp7O synthesis decreases in N-methylformamide-treated HL-60 promyelocytic cells (Richards et al., 1988).These seemingly contradictory effects on Hsp70 and Hsp9O levels may be linked to different requirements for chaperoning functions in different, cell types. We suggest that the same hypothesis proposed above to explain the different effects of TGFp on cells may apply to these cellular responses, which may be determined in part by the effects of a particular factor on overall rates of protein synthesis and the chaperoning needs of the cohort of proteins that are up- or down-regulated. According to our hypothesis, both heat shock and TGFp treatment increase the cellular need for molecular chaperones. Interestingly both stimuli affect at least one common structure, the ECM and its associated proteins. For example, TGFP increases collagen and fibronectin synthesis in cultured cells (Ignotz et al., 1987). Heat shock induces a collagen-binding protein

576

TAKENAKA AND HIGHTOWER

called Hsp47 in CEC (Hirayoshi et al., 1991) and increases collagenase and stromelysin mRNA levels in rabbit synovial fibroblasts (Vance et al., 1989). TGFp stimulates chemotactic activity in some cell types, and such activity characteristically involves assembly and disassembly of cytoskeletal elements, which in turn could increase the need for molecular chaperones. In support of this possibility, Hsp9O and Hsp7O family proteins are associated with isolated cytoskeletal elements (Lim et al., 1984; Napolitano et al., 1985; Koyasu et al., 1986; Sanchez et al., 1988). Both Hsp7O family proteins and TGFp have been implicated in normal development and in wound responses, but a s yet there is no evidence linking them mechanistically. Our observations that TGFP induces Hsp70 and Hsp9O family members in CEC supports the existence of such a link. For example, the expression of the hsc4 gene of Drosophila is enhanced during development in cells actively undergoing endocytosis, rapid growth, and shape changes (Perkins et al., 1990) and both Hsp70 and Hsp9O family members are induced during mammalian spermatogenesis (Lee, 1990; Zakeri et al., 1990). TGFP is detectable after day 9.5 of murine development and is concentrated in mesenchymal tissues (Heine et al., 1987; for additional references, see Takenaka et al., 1991). Both Hsp7O and Hsp9O are induced in rat tissue slices incubated in vitro (Hightower and White, 1981) and hsp70 gene expression increases dramatically in rat brain tissue at the site of a surgical incision (Brown et al., 1989), suggesting a role for these molecular chaperones in wound responses. Evidence of a role for TGFp in wound responses has been obtained as well (Sporn e t al., 1987). The Hsp9O family in mammals is composed of two closely related phosphoproteins with slightly different molecular masses on SDS-PAGE, which are encoded by two genes (Hickey et al., 1989). HspSOa, the larger form, is induced by serum, the adenoviral E1A protein (Simon et al., 1987), and is developmentally regulated in the mouse. Based on the fact that TGFp and serum rapidly induce chicken Hsp90-2 and Hsp90-3, we speculate that these two isoforms may be the chicken homologs of mammalian Hsp9Ooi. Both mammalian Hsp9O's are heat inducible as are all of the chicken Hsp9O isoforms. New features of Hsp9O reported herein include the existence of a n unphosphorylated isoform (Hsp90-3) and the metabolic instability of the phosphorylated isoforms Hsp90-2 and Hsp90-4. These biochemical, metabolic, and regulatory differences indicate that the isoforms have distinctive properties and raise the possibility that the isoforms may have different functional roles.

ACKNOWLEDGMENTS We thank Michael Sporn, Anita Roberts, and Lalage Wakefield for helpful discussions and gifts of reagents a t the start of this study. We thank Susan Lindquist and David Toft for antibodies and David Toft for sharing preprints. Carol White and Mary Brown provided helpful comments on the manuscript. This work was supported by grants from the National Science Foundation (DCB-8916418) and the National Institutes of Health (GM35334). We benefited from the use of the

Cell Culture Facility of the University of Connecticut Biotechnology Center.

LITERATURE CITED Beckman, R.P., Mizzen, L A . , and Welch, W.J. (1990) Interaction of hsp70 with newly synthesized proteins: implications for protein folding and assembly. Science, 248t850-854. Bradford, M. (19761 A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle ofprotcindye binding. Anal. Biochem., 72t248-254. Brown?LR., Rush, S.J., and Ivy, G.O.(1989) Induction of a heat shock gene at the site of tissue injury in the rat brain. Neuron, 2t15591564. Brugge, J.S. (1986) Interaction of the Rous sarcoma virus protein, pp60"", with the cellular proteins pp50 and pp90. Curr. Topics Microbiol. Immunol., 123t1-23. Brugge, J.,Erikson, E., and Erikson, R.L. (3981) The specific interaction of the Rous sarcoma virus transforming protein pp6ObrC,with two cellular proteins. Cell, 25t363-372. Chiang, H.L., Terlecky, S.R., Plant, C.P., and Dice. J.F. (1989) A role for a 70-kilodalton heat shock protein in lysosomal degradation of intracellular proteins. Hature, 246r382-385. Courtneidge, S.A., and Bishop, J . M . (1982) Transit of ~ p 6 0 " ~to ~ "the plasma membrane. Proc. Natl. Acad. Sci. USA, 79.7117-7121. Deshaies, R., Koch, B., Werner-Washburne, M., Craig, E.A., and Scheckman, R. (1988) 70 kD stress protein homologues facilitate translocation of secretory and mitochondria1 precursor polypeptides. Nature, 3323300-805. Duncan, R., and Hershey, W.B. (19841 Evaluation of isoelectric focusing running conditions during two-dimensional isoelectric focusing/ sodium dodecyl sulfate-polyacrylamide gel electrophoresis: variation of gel patterns with changing conditions and optimized isoelectric focusing conditions. Anal. Biochem., 138.144-155. Ellis, R.J., and Hemmingsen, S.M. (1989) Molecular chaperones: proteins essential for the biogenesis of some macromolecular structures. Trends Biochem. Sci., 24t339-343. Feng, P., Catt, K.J., and Knecht, M. (1988) Transforming growth factor-p stimulates meiotic maturation of the rat oocyte. Endocrinology, 122t181-186. Ferris, D.K., Harel-Bellan, A., Morimoto, R.I., Welch, W.J., and F'arrar, W.L. (1988)Mitogen and lymphokine stimulation of heat shock proteins in T lymphocytes. Proc. Natl. Acad. Sci. USA, 85t38503854. Heine, U.I., Munoz, E.F., Flanders, K.C., Ellingsworth, L.R., Lam, H.-Y.P.,Thompson, N.L., Roberts,A.B., and Sporn, M.B. (1987)Role of transforming growth factor-p in the development of the mouse embryo. J. Cell. Biol., 105:2861-2876. Hickey, E., Brandon, S., Smale, G., Lloyd, D., and Weber, L.A. (1989) Sequence and regulation of a gene encoding a human 89 kilodalton heat shock protein. Mol. Cell. Biol., 92615-2626, Hightower, L.E. (1980) Cultured animal cells exposed t o amino acid analogues or puromycin rapidly synthesize several polypeptides, J. Cell. Physiol., 102.4077427. Hightower, L.E., and Bratt, M.A. (19741 Protein synthesis in Newcastle disease virus-infected chicken embryo cells. J. Virol., 13,788800.

Hightower, L.E., and White, F.P. (1981) Cellular responses to stress: comparison of a family of 71-73 kilodalton proteins rapidly synthesized in rat tissue slices and canavanine-treated cells in culture. J . Cell. Physiol., I08t261-275. Hirayoshi, K., Kudo, H., Takechi, H., Nakai, A,, Iwamatsu, A,, Yamada, K., and Nagata, K. (1991) HSP47: a tissue-specific, transformation-sensitive, collagen-binding protein of chicken embryo fibroblasts. Mol. Cell. Biol., 11:4036-4044. Hwang, D.S., Crooke, E., and Kornberg, A. (1990) Aggregated DnaA protein is dissociated and activated for DN.4 replication by phospholipase or DnaK protein. J . Bid. Chcm., 265:19244-19248. Ignotz, R.A., Endo, T., and Massague, J . 11987)Regulation of fibronectin and type I collagen mRNA levels by transforming growth factor+. J. Biol. Chem., 262.6443-6446, Ignotz, R.A., Heino, J., and Massague, J. (1989) Regulation of cell adhesion receptors by transforming growth factor-p: regulation of vitronectin receptor and LFA-1. J. Biol. Chem., 264.389392. Kaaden, O.R., Lange, S., and Stiburek, B. (1982) Establishment and characterization of chicken embryo fibroblast clone LSCC-H32. In Vitro, 18t827-834.

STRESS PROTEINS

Kaczmarek, L., Calabretta, B., Kao, H.-T., Heintz, N., Nevins, J., and Baserga, R. (1987) Control of hsp7O RNA levels in human lymphocytes. J. Cell. Biol., 104t183-187. Kang, P . J . , Ostermann, J.,Shilling, J., Neupert, W., Craig, E.A., and Pfanner, N. (1990) Requirement for hsp70 in the mitochondria1 matrix for translocation and folding of precursor proteins. Nature, 348r137-143. Koyasu, S., Nishida, E., Kadowaki, T., Matsuzaki, I?., Iida, K., Harada, F., Kasuga, M.S., H., and Yahara, I. (1986) Two mammalian heat shock proteins, HSP9O and HSP100, are actin binding proteins. Proc. Natl. Acad. Sci. USA4:83t8054-8058. Kurtz, S., Rossi, J . , Petko, L., and Lindquist, S. (1986) An ancient developmental induction: Heat-shock proteins induced in sporulation and oogenesis. Science, 231:1154-1157. Laemmlj, U.K. (1970) Cleavage of the structural proteins during the assembly of the head of bacteriophage T4. Nature (Lond.),277t680685. Laskey, R.A., and Mills, A.D. (1975) Quantitative film detection of 'H and 14C in polyacrylaniide gels by fluorography. Eur. J. Biochem., 56t335-341. Lee, S.J. 11990) Expression of HSP86 in male germ cells. Mol. Cell. Biol., 10t3239-3242. Lee, A S , Bell, J., and Ting, J . (1984)Biochemical characterization of the 94- and 78-kilodalton glucose-regulated proteins in hamster fibroblasts. J. Biol. Chem., 259t46164621. Lees-Miller, S., and Anderson, C.W. (1989a) Two human 90-kDa heat shock proteins are phosphorylated in vivo at conserved serines that are DhosDhorvlated in vitro bv casein kinase 11. J. Riol. Chem.. 264t243f-2437. Lees-Miller, S.P., and Anderson, C.W. (1989b) The human doublestranded DNA activated protein kinase phosphorylates the 90-kDa heat shock protein, hsp90a a t two NH,-terminal threonine residues. J. Biol. Chem., 264:17275-17280. Legagneux, V., Dubois, M.F., Morange, M., and Bensaude, 0. (1988) Phosphorylation of the 90 kDa heat shock protein in heat shocked HeLa cell lysates. FEBS Letters, 231t417-420. Legagneux, V., Morange, M., and Bensaude, 0. (1991) Heat shock increases turnover of 90 kDa heat shock protein phosphate groups in HeLa cells. FEBS Let.,291t359-362. Liberek, K.C.G., and Zylicz, M. (1988) Role of the Escherichia coli DnaK and DnaJ heat shock proteins in the initiation ofbacteriophage A DNA replication. Proc. Natl. Acad. Sci. USA, 85:66324636. Lim, L., Hall, C., Leung, T., and Whatley, S. (1984)The relationship of the rat brain 68 kDa microtubule-associated protein with synaptosoma1 plasma membranes and with the Drosophila 70 kDa heatshock protein. Biochem. J.,224;677-680. Milarski, K.L., and Morimoto, R.I. (1986)Expression of human HSP70 during the synthetic phase of the cell cycle. Proc. Natl. Acad. Sci. USA, 83r9517-9521. Napolitano, E.W., Pachter, J.S., Chin, S.S.M., and Leim, R.K.H. (1985) p-internexin, a ubiquitous intermediate filament-associated protein. J. Cell. Biol., 101;1323--1331. Nilsen-Hamilton, M. (1990) Transforming growth factor+ and its action on cellular growth and differentiation. In: Growth Factors and Development. M. Nilsen-Hamilton, eds. Academic Press, San Diego, pp. 96-136. O'Farrell, P.H. (1975) High resolution two-dimensional electrophoresis ofproteins. J. Biol. Chem., 250t4007-4021. Perkins, L A , Doctor, J.S., Zhang, K., Stinson, L., Perrimon, N., and Craig, E.A. (1990) Molecular and developmental characterization of the heat shock cognate 4 gene ofDrosophila melunogaster. Mol. Cell. Biol., 10,3232-3238, Picard, D.,Khursheed, B., Garabedian, M.J., Fortin, M.G., Lindquist, S.,and Yamamoto, K.R. 11990)Reduced levels of hsp90 compromise steroid receptor action in vivo. Nature, 348t166-168. Richards, F.M., Watson, A., and Hickman, J.A. (1988) Investigations on the effects of heat shock and agents which induce a heat shock response on the induction of differentiation of HL-60 cells. Cancer Res., 48t67156720. Rose, D.W., Welch, W.J., Kramer, G., and Hardesty, B. 11989)Possible

577

involvement of the YO-kDa heat shock protein in the regulation of protein synthesis. J. Biol. Chem., 264t6239-6244. Rothman, J.E. (1989) Polypeptide chain binding proteins: catalysts of protein folding and related processes in cells. Cell, 59591-601. Sanchez, E.R., Toft, D.O., Schlesinger, M.J., and Pratt, W.B. (1985) Evidence that the 90 kDa phosphoprotein associated with the untransformed L-cell glucocorticoid receptor is a murine heat shock protein. J . Biol. Chem., 26Ot1235X-12403. Sanchez, E.R., Redmond, T., Schemer, L.C., Rresnick, E.H., Welsh, M.J., and Pratt, W.B. (1988) Evidence that the 90 kD heat shock protein is associated with tubulin containing complexes in I, cell cytosol and in intact PtK cells. Mol. Endocrinol., 2,756-760. Santoro, M.G., Garaci, E., and Amici, C. (1989) Prostaglandins with antiproliferative activity induce the synthesis of a heat shock protein in human cells. Proc. Natl. Acad. Sci. USA, 86t8407-8411. Sekellick, M.J., and Marcus, P.I. (1986) Induction of high titer chicken interferon. Methods Enzymol., 119,115-125. Simon. M.C., Kitchener, K., Kao, H.T., Hickey, E., Weber, L., Voellmy, R., Heintz, N., and Nevins, J.R. (1987) Selective induction of human heat shock gene transcription by the adenovirus E l A gene products, including the 125 E l a product. Mol. Cell. Biol., 7,28842890. Singh, M.K., and Yu, J. (1984) Accumulation of a heat shock-like protein during differentiation of human erythroid cell line K562. Nature, 309t631-633. Sporn, M.B., Roberts, A.B., Wakefield, L.M., and de Crombrugghe, B. 11987)Some recent advances in the chemistry and biology of transforming growth factor-beta. J . Cell Biol., 105,1039-1049. Takenaka, I.M., Sadis, S., and Hightower, L.E. (1991) Transforming Growth Factor-p regulates basal expression of the hsp70 gene family in cultured chicken embryo cells. In: Heat Shock and Development. L.E. Hightower and L. Nover, eds. Springer-Verlag, Heidelberg, pp. 188-209. Theodorakis, N.G., Zand, D.J., Kotzbauer. P.T.. Williams. G.T.. and Morimoto, R.I. (1989) Hemin induced transcriptional activation of the HSP70 gene during erythroid maturation in K562 cells is due to a heat shock factor mediated stress response. Mol. Cell. Biol., 9t3166-3173. Thompson, K.L., Assoian, R., and Rosner, M.R. (1988) Transforming growth factor-p increases transcription of the genes encoding the epidermal growth factor receptor and fibronectin in normal rat kidney fibroblasts. J. Biol. Chem., 263:19519-19524. Vance, B.A.,Kowalski, C.G., and Brinckerhoff, C.E. (1989)Heat shock of rabbit synovial fibroblasts increases expression of mRNAs for two metalloproteinases, collagenase and stromelysin. J. Cell. Biol., Z08t2037-2043. Vogel, J.P., Misra, L.M., and Rose, M.D. (1990) Loss of BiP/GRP78 function blocks translocation of secretory proteins in yeast. .J. Cell Biol., IZUt7885-1895. Walker, A.I., Hunt, T., Jackson, R.J., and Anderson, C.W. (1985) Double-stranded DNA induces the phosphorylation of several proteins including the 90,000 mol wt heat-shock protein in animal cell extracts. EMBO J.,4,139-145. Welch, W .J., Garrels, J.I., Thomas, G.P., Lin, J.J.C., and Feramisco, J.R. (1983) Biochemical characterization of the mammalian stress proteins and identification of two stress proteins as glucose- and Ca" ionophore-regulated proteins. J. Biol. Chem., 258:7102-7111. Whelan, S.A., and Bghtower, L.E. (1985) Induction of stress proteins in chicken embryo cells by low-level zinc contamination in amino acid-free media. J. Cell. Physiol., Z22t20.5-209. Wu, B.J., and Morimoto, R.I. (198.5)Transcription of the human hsp70 gene is induced by serum stimulation. Proc. Natl. Acad. Sci. USA, 82t60706074. Wu, B., Williams, G.T., and Morimoto, R.I. (1987) Detection of three protein binding sites in the serum regulated promoter of human gene encoding 70 kD heat shock protein. Proc. Natl. Acad. Sci. USA, 8412203-2207. Zakeri, Z.F., Welch, W.J., and Wolgemuth, D.J. (1990) Characterization and inducibility of hsp70 proteins in the male mouse germ line. J. Cell Biol., 111t1785-1792.