Molecular Brain Research, 12 (1992) 203-208 © 1992 Elsevier Science Publishers B.V. All rights reserved. 0169-328X/92/$03.50

203

BRESM 70370

Characterization of the major 68 kDa heat shock protein in a rat transformed astroglial cell line Robert

N . N i s h i m u r a 1'4, B a r n e y

E . D w y e r 2'4'5, J e a n

de Vellis 6 and Kerry

B. C!egg 3

1Regeneration Research Laboratory, 2Molecular Neurobiology Laboratory and 3Hypertension Research Laboratory, Veterans Affairs Hospital, Sepulveda, CA 91343 (U.S.A.), 4Department of Neurology and 5The Brain Research Institute, UCLA, Los Angeles, CA 90024 (U.S.A.) and 6Laboratory of Neurobiochemistry, Mental Retardation Research Center, Department of Anatomy, UCLA, Los Angeles, CA 90024 (U.S.A.) (Accepted 30 July 1991) Key words: Heat shock protein; Protein synthesis; Glial cell; Cancer; Brain tumor

The heat shock response in a transformed astrocyte line was compared with nontransformed astrocytes. The synthesis of HSP 68, the major inducible heat shock protein (HSP 68) was induced by a non-lethal 45 °C, 10 min heat shock. Although the incorporation of [35S]methionine into HSP 68 suggested that similar amounts of protein were being synthesized after heat shock, Western immunoblotting demonstrated striking differences in the HSP immunostaining between the two cell types. By one- and two-dimensional gel electrophoresis the major 68 kDa heat shock protein (HSP 68) was similar in both cell types. However, HSP 68 from heat shocked, transformed astrocytes did not immunostain with the monoclonal antibody, C-92, which is specific for the major inducible heat shock proteiil of HeLa cells. In contrast HSP 68 from heat shocked, nontransformed astrocytes immunostained quite well. A polyclonal antibody raised against the inducible 72 kDa heat shock protein of HeLa cells immunostained the HSP 68 from both astrocytes and transformed astrocytes. Analysis of the mRNA from the two cell types after heat shock revealed two bands of approximately 2.5 and 2.8 kb in astrocytes but only a single 2.5 kb band in the heat shocked transformed astroglia. These results suggest that structural differences in the HSP 68 may be present in the transformed astrocytes compared to the normal astrocytes. INTRODUCTION The heat shock response refers to the induction of a set of cellular heat inducible genes and the synthesis of their protein products in cells after exposure to a variety of noxious stimuli t°'19.2x. The major heat shock proteins (HSPs) most commonly noted are 68, 70, 89 and 110 kDa in size 19. In addition, a variety of low molecular weight HSPs and a family of glucose-regulated proteins has been described 19. The 70 k D a family of HSPs con~ists of HSP 70 which is constitutively synthesized by most cells as well as the highly stress-inducible HSP 68. HSP 68 is important because it appears to be directly associated with cellular survival after heat stress and likely is involved in the mechanism of acquired thermotolerance L9'~8. In the central nervous system of mammals it is noted that the members of the 70 k D a family of HSPs are found in limited areas of the brain after hypoxic-ischemic injury 7'16'24, trauma 3, and exposure to neurotoxins s'23. This laboratory is investigating the role of HSPs after neural injury. We have found that astrocytes in culture synthesize more HSP 68 after heat shock than cultured cortical neurons ~5 and have hypothesized

that the inability of neurons to promptl3: synthesize adequate amounts of HSPs in a timely fashion may provide a basis for the selective vulnerability of neurons after acute injury. In that communication we also showed that the 70 k D a heat shock proteins synthesized by neurons and astrocytes differed in their immunostaining characteristics. In our efforts to characterize HSP 68 derived from central nervous system cells, we have compared the heat shock response in rat forebrain astrocytes with that of a spontaneously transformed astroglial cell line (AT) which was derived from a subpopulation of newborn forebrain astroglial cells which transformed in culture conditions 2. The asr~:c3~es and ATs have a common glial cell lineage. Morphologically, these cells demonstrated loss of glial filaments and the only visible cytoskeletal elements were microtubules. They expressed glycerol-phosphate dehydrogenase and lactate dehydrogenase characteristic of oligodendroglial cell~ but lost the glutamine synthetase induction characteristic of normal astrocytes in the central nervous system. Furthermore, these transformed cells were tumorigenic in rats. Here we describe major differences between astrocytes a~d ATs in the m R N A and the immunostaining properties of

Correspondence: R.N. Nishimura, VA Medical Center 111 N-l, 16111 Plummer Street, Sepulveda, CA 91343, U.S.A. Fax: (1) (818) 895-9554.

2~ HSP 68. These results and those from our previous study t5 have important implications for i m m u n o c y tochemical staining of H S P 68 in the nervous system. This work was presented in preliminary form in the abstracts of the International Workshop on H e a t Shock, II GB Press, Naples, Italy, 1990.

MATERIALS AND METHODS

Materials Newborn Wistar rats were obtained from a breeding colony maintained in the laboratory of J. de Vellis. The transformed astroglial cell line was obtained from the laboratory of J. de Vellis. [35S]Methionine Translabel (1000 Ci/mmol) was purchased from ICN, Irvine Scientific, Costa Mesa, CA. Calf serum was obtained from Hyclone L~bs, Logan, UT. All chemicals were analytical grade or tissue culture grade. Antibodes used in this study were the gift ,~' "" ~v"~ch University of California, San Francisco, CA. The clone,. - .gment of the major inducible human heat shock proteir . ~ spanning the entire coding region was kindly provided by .~,..~iorimoto, Northwestern University, Evanston, IL. Preparation of astrocyte cultures The primary purified glial culture were prepared as previously described'*. Secondary cultures of astrocytes were prep~ired by washing with 0.02% EDTA in calcium- and magnesium-free PBS, then trypsinized (2.5% in 0.9% NaCl) until completely dissociated. Cells were then resuspended in Dulbecco's modified essential mediumtHam's F-12 (1.~1)supplemented with calf serum (5% final concentration) and replated at 1 x 106 cells/25 cm2 flask. Astroglial cells were grown for 14-21 days (16-29 DIV) until ready for use. Astroglia were greater than 98% pure by staining for glial fibrillary acidic protein. Transformed astroglial cells were thawed then passed in medium identical to the medium for the nontransformed astrocytes. Induction of heat stress Heat stress was induced as described previouslyt3. Prior to heat stress medium was changed from serum supplemented to serum free DMEM/F.12 (1:1) medium warmed to 37 °C and left to equilibrate at 37 °C for 15 min. After a heat stress of 45 °C for 10 min, the flasks were removed and three mi of pre-warmed (37 °C) fresh methionine deficient DMEM was added to each flask with 100/~Ci [3SS]methionine Translabel. Cultures were incubated at 37 °C for three hours then harvested. Medium was removed and cultures were washed with two 5 ml washes of cold 0.1 M phosphate-buffered saline (PBS), pH 7.4. Total cellular protein was precipitated in 10% trichloracetic acid supplemented with 0.1% methionine. After two brief washes with PBS the cells were homogenized in 2% SDS. Aliquots for measurement of CPMs and protein t~ were taken. Analysis of proteins by gel electrophoresis Gels of sample of total cellular protein were processed and electrophoresed on 12,0% SDS gels as previously described ~3. Two-dimensional gel electrophoresis was performed as previously described t3. After electrophoresis, gels were stained with Commassie blue, destained in acetic acid, methanol, and water (!:5:5 by volume), then dried and exposed to Kodak X-Omat film for one week. Western blot analysis was performed by the method of Towbin et al. 22. Niirocellulose membranes with transferred proteins were immunostained as previously described using a Vectastain Kit, Vecta Labs. t3 Total cellular RNA was prepared as described in detail TM. Preparation of the HSP probe was previously described TM. Briefly, a HindIII-BamHI fragment from the original plasmid was subcloned and grown in a dual pro.mcter plasmid pT7/T3-19 (BRL) 32p-Labelin~ was performed using the Riboprobe Gemini System

34

1

2

a

'4~

kS

6

Fig. 1. Autoradiograph of [aSS]methionine incorporation into total cellular proteins of astrocytes and ATs. 12% polyacrylamide gels of total cellular proteins shows induction of the major 68 kDa heat shock protein in astrocytes and ATs at 45 °C or 10 min and 45 °C for 20 min. Control astrocytes are in lane 1, and control ATs in lane 4. Heat shocked (45 °C, 10 min) astrocytes and ATs lanes 2 and 4, respectively. Heat shocked (45 °C, 20 min) astrocytes and ATs lane 3 and 6, respectively. Also induced are 97, 89 and 70 kDa proteins in both astrocytes and ATs. Heat shocked ATs also synthesized a 30/34 kDa protein. Equal cpms were run in each lane.

(Promega). Northern blot and slot blot analysis of total cellular RNA was performed as previously described sas. RESULTS The pattern of [35S]methionine incorporation into transformed astrocytes is similar to that seen in astrocytes with a notable exception (Fig. 1). Most prominent is the reduced labeling in transformed astrocytes (ATs) of a 33 k D a protein which we have tentatively identified as heme oxygenase 6. Transformed astrocytes exposed to heat stress appear to synthesize the major inducible 68 k D a heat shock protein (HSP 68) as well as primary astrocytes (Fig. 1). As in astrocytes after heat stress, the ATs also show induction of 70, 89 and 97 k D a proteins. The identity of HSP 68 from ATs and/~strocytes is suggested by their similar migration during one- and twodimensional gel electrophoresis (Fig. 2). However, Western immunoblot analysis revealed that the immunostaining characteristics of HSP 68 from these

205

I.

a

Fig. 2. Two-dimensional gels of co,trol and heat shocked astrocytes and ATs (45 °C, 20 min). Control astrocytes (c) and ATs (a) are compared with heat shocked astrocytes (d) and ATs (b). Arrows indicate in each gel the location of the major inducible 68 kDa HSP. The location of the 68 kDa HSP is approximately the same in heat-shocked astrocytes and ATs.

two cell types differed dramatically (Fig. 3). Monoclonal antibody C-92, which was shown by Welch and Suhan (1986) to specifically stain the 72 k D a heat shock protein of H e L a cells, immunostained, as expected, HSP 68 from heat-shocked astrocytes. Se~rpfisingly, C-92 failed to immunostain HSP 68 from transformed astrocytes. No staining in controls of either cell type was detected. This pattern has been reproduced more than 6 times. When

the polyclonal antibody, specific for the human inducible 72 kDa heat shock protein (which is equivalent to our HSP 68) was used to immunostain the Western blot of the astrocytes and transformed astroglia~ the immunostaining pattern of astrocytes and transformed astrocytes differed significantly. First, the polyclonal antibody was able to detect HSP 68 in unstressed astrocytes, but there was no detectable HSP 68 in unstressed ATs (Fig. 3).

"--"--"

F

1

2

3

5

6

- 6 8 kD

7

8

Fig. 3. lmmunoblot of the major inducible 68 kDa lISP in astrocytes and ATs using a polyclonal and monoclonal antibody specific for the 68 kDa HSP. Lanes 1 and 2 are control and heat-shocked astrocytes, respectively, Lanes 3 and 4 are control and heat-shocked ATs, respectively. Lanes 1 through 4 were immunostained with a'polyclonal antibody specific for HSP 68. Note that the HSP 68 stained in both heat shocked astrocytes and ATs as well as the control astrocytes. Lanes 5 and 6 are control and heat-shocked astrocytes, respective'.y. Lanes 7 and 8 are control and heat-shocked ATs, respectively. Lanes 5 through 8 were stained with a monoclonal antibody specific for HSP 68. Note that the HSP 68 stained only the HSP 68 of astrocytes but not ATs. Controls were unstained. All lanes represent 100 ~ug of total cellular precipitable protein transferred to nitrocellulose blots aRer electrophoresis. Lanes ~ and 5, 2 and 6, 3 and 7, and 4 and 8 are from the same experimental samples, respectively.

206

i a

b

c

I II d

e

f

g

h

i

Fig. 4. Slot-bot of total cellular RNA from control and heat shocked ATs (45 °C, 10 min) probed for HSP 68 mRNA. Lanes (a) control, (b) heat-shocked ATs 1 h after heat shock, (c) heat-shocked ATs 3 h after heat shock, (d) heat-shocked ATs 7 h after heat shock. Lanes (e) through (i) represent a standard HSP 68 mRNA ladder blotted at the same time: (e) 0.05 ng, (f) 0.1 ng, (g) 0.5 ng, (h) 1.0 ng, and (i) 2.5 ng. HSP 68 mRNA content: (a) undetectable, (b) 0.045 ng, (c) 0.30 ng, and (d) 0.025 rig. Note that at the 45 °C, 10 rain heat shock, peak levels of mRNA are noted at 3 h. Two/~g of total cellular RNA was applied to each slot.

This most likely reflects increased sensitivity of the polyclonal antibody immunostaining. Previously we found that unstressed rat forebrain astrocytes acc,amulate immunostainable HSP 68 (detected with the C-92 antibody) as they aged in culture (data not shown) suggesting that a minimal level of protein is needed for detection when C-92 is used. Not surprisingly, this antibody recognized

4o

the HSP 68 of both the heat shocked astrocytes and transformed astroglia (Fig. 3). These data suggest that the HSP 68 synthesized in ATs after heat shock is structurally different than astrocyte HSP 68. We next analyzed HSP 68 mRNA in heat shocked transformed astrocytes. Slot-blot hybridization (Fig. 4) showed that HSP 68 mRNA levels changed in a time-dependent fashion very similar to that reported previously in heat shocked astrocytes 5. HSP 68 mRNA was detected at 1 h after heat shock with highest levels being present at 3 h; by 7 h after heat shock, levels had declined to 8% of those at 3 h. A rough estimate of the half-life of HSP 68 mRNA based on the decline in levels between 3 and 7 h was 60 min, This assumes a maximum of HSP 68 mRNA at 3 h and a constant rate of degradation of HSP mRNA through 7 h without appreciable synthesis of HSP 68 mRNA. However, Northern blot analysis of the total cellular RNA revealed a striking difference between the astrocytes and ATs (Fig. 5), The 32p-labeled HSP 68 riboprobe did not detect HSP 68 mRNA in either unstressed, control astrocytes or ATs. As expected, this probe hybridized to two RNA bands approximately 2.8 and 2.5 kb in size in heat shocked astrocytes (Fig. 5). Surprisingly, this riboprobe hybridized to only a single 2.5 kb RNA band in heat shocked transformed astro. cytes (Fig. 5). DISCUSSION

a

b

c

d

~

Fig. 5. Northern blot of heat-shocked astrocytes and ATs. Lanes

(a) and (c) represent control astrocytesand ATs respectively. Lanes (b) and (d) represent heat-shocked astrocytes and ATs after a 20 rain heat shock at 45 °C. Note that two bands of 2.5 and 2.8 kb hybridize in heat-shocked astrocytes, while only the 2.5 kb band hybridizes in the heat-shocked ATs. Ten ~tg of total cellular RNA was applied to each lane. O represents the origin and E represent the end of the run.

Major qualitative differences in the heat shock response between transformed rat forebrain astrocytes (ATs) and their nontransformed counterparts have been found. The major heat shock proteins synthesized by rat forebrain astrocytes in culture are 68, 70, 89 and 97 kDa in size t3. Transformed astrocytes appear capable of synthesizing all four HSPs after a brief, non-lethal heat shock, although some differences are noted. ATs appear less able to induce HSP 97 synthesis by visual comparison of gels. In contrast, the labeling of HSP 70, the constitutive member of the 70 kDa heat shock protein family appears to be far lower in unstressed ATs but appears far more stress inducible compared to astrocytes. Synthesis of HSP 68, the major inducible member of the 70 kDa family of heat shock proteins, and HSP 89 were both increased after heat shock in each cell type. A 33 kDa band representing heine oxygenase 6 was induced in ATs but not significantly in astrocytes. Further investigation revealed significant qualitative differences in HSP 68 between the two cell types. HSP 68 of astrocytes and ATs share the same molecular mass and migrate similarly on two dimensional gels. However, the two proteins immunostain differently with the monoclonal antibody C-92 which is specific for the human inducible HSP 72. Mi-

207 larski and Morimoto ~2 reported that the C-92 binds to the carboxy-terminal end of the HSP 68. This would suggest that the carboxy-terminus of the AT 68 kDa HSP is a different amino acid sequence from the HSP 68 of astrocytes. Although HSP 68 is one of the most highly conserved proteins known, the 3" terminus of the gene, which codes for the carboxy end of the protein, is also known to be the region of the gene in which the nucleotide sequence is most divergent. The presence of a small structural difference is supported by the fact that a polyclonal antibody, which presumably recognizes many epitopes on the HSP 68 does stain the HSP 68 of the transformed astrocytes. An alternative explanation for the differences in immunostaining of the AT HSP 68 is that the epitope for the HSP 68 is masked by a cellular protein or membrane which does not dissociate in SDS. If this were the case, then we would expect the molecular mass of the resultant complex to be significantly different than the HSP 68 of the astrocytes. The one and two dimensional gels do not support this possibility. In addition, Northern blotting revealed that a single 2.5 kb mRNA from transformed astrocytes hybridized to our HSP 68 riboprobe, in sharp contrast to the doublet found at 2.8 and 2.5 kb in astrocytes. The two bands may represent alternate length of poly(A) + tailing on the mRNA 12. The specific difference in HSP 68 from ATs which results in the different pattern of immunostaining is not known but may be revealed by peptide mapping of purified HSP 68 from both cell types and sequencing the gene. The differences in HSP 68 induction between ATs and astrocytes may have far-reaching implications. Though the basis of transformation in the ATs is unknown it raises speculation that the transformation produces changes in the expression of HSP 68. It has been reported that transfected p53 mutant mouse protein in SV40 transformed monkey COS cells coprecipitates with HSP 72 and 73, corresponding to HSP 68 and 70 in rat astrocytes 2°. p53 is a cellular phosphoprotein which is elevated in transformed cell lines. It has also been asso-

REFERENCES 1 Barbe, M.E, Tytell, M., Gower, D.J. and Welch, W.J., Hyperthermia protects against light damage in the rat retina, Science, 241 (1988) 1817-1820. 2 Bressler, J.P. and de Vellis, J., Neoplastic transformation of newborn rat astrocytes in culture, Brain Res., 348 (1985) 2127. 3 Brown, I.R., Rush, S.J. and Ivy, G.O., Induction of a heat shock gene at the site of tissue injury in the rat brain, Neuron, 2 (1989) 1559-1564. 4 Cole, R. and de Vellis, J., Preparation of astrocyte and oligodendrocyte cultures from primary rat glial cultures. In A. Shahar, J. de Vellis, A. Vernadakis and B. Haber (Eds.), A DISsection and Tissue Culture Manual of the Nervous System, Alan

ciated with immortalization of primary cells in vitro. The interaction of the p53 and HSP 68 and 70 may result in stabilization of p5317. If the presence of p53 is required for immortalization of cells then, the complex of HSP 68 and 70 with p53 may decrease degradation of p53. The present study also implies that regulation of HSP 68 is different in ATs compared with astrocytes. This possibility is suggested by the qualitative differences we have found with the absence of HSP 68 immunostaining in unheated control ATs versus HSP 68 accumulatkm in unheated control astrocytes. Previously we showed that HSP 68 synthesis is strictly regulated in heat shocked astrocytes 5. Altered expression of HSP 68 may lead to altered sensitivity of ATs to stressful conditions. Since the transformed astrocytes are dedifferentiated cells which are tumorigenic in rats, understanding how these cells respond to stress will provide insight into strategies which will allow selectl~ e killing of tumor cells in the presence of normal tissue. Finally, the results of differences of immunostaining of the HSP 68 between the related cell types raises a major caution for those using an antibody to characterize a heat shock response. A negative antibody staining does not necessarily translate into the absence of synthesis of the HSP 68. Previous work frcm our laboratory underscores this point. We have shown major differences in the expression of m R N A and immunostaining of HSP 68 between neurons and astrocytes from rat after a brief heat shock and recovery 15. In that study HSP 68 of neurons does not stain with the C-92 moaoclonal antibody used in the present study, but the polyclonal antibody for the same protein will immunostain not only the HSP 68 but also the constitutively synthesized HSP 70 after heat shock.

Acknowledgements. This work was supported in part by the Research Service of the (Jffice of Veterans Affairs. We wou!d like to acknowledge the helpful discussions with W. Welch and R. Morimoto in the preparation of this paper. The technical help of K. Picard and R. Cole is gratefully acknowledged.

R. Liss; New York, 1989, pp. 121-133. 5 Dwyer, B.E., Nishimura, R.N., de Vellis, J. and Clegg, K.B., Regulation of heat shock protein synthesis in rat astrocytes, Z Neurosci. Res., ~ (1991) 352-358. 6 Dwyer, B.E., Nishimura, R.N., de Vellis, J. and Tadashi, Y., Heme oxygenase is a heat shock protein and PEST protein in rat astroglial cells, Gila, in press. 7 Dwyer, B.E., Nishimura, R.N. and Brown, I.R., Synthesis of the major inducible heat shock protein in rat hippocampus after neonatal hypoxia-ischemia, Exp. Neurol., 104 (1989) 28-31. 8 Gonzalez, M.F., Shiraishi, K., I-Iisanaga, K., Sagar, S.M., Mandabach, M. and Sharp, ER., Heat shock proteins as markers of neural injury; Mol. Brain Res., 6 (1989) 93-100. 9 Johnston, R.N. and Kucey, B.L., Competitive inhibition of hsp 70 gene expression causes thermosensitivity, Science, 242 (1988)

208 1551-1554. 10 Lindquist, S. and Craig, E.A., The heat-shock proteins, Annu. Rev. Genet., 22 (1988) 631-677. 11 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J., Protein measurements with the Folin phenol reagent, Y. Biol. Chem., 193 (1951) 265-275. 12 Milarski, K.L. and Morimoto, R.I., Mutational analysis of the human HSP 70 protein: distinct domains for nucleolar localization and adenosine triphosphate binding, l. Cell. Biol., 169 (1989) 1947-1962. 13 Nishimura, R.N., Dwyer, B.E., Welch, W., Cole, R., de Vellis, J. and Liotta, K., The induction of the mahor heat-stress protein in purified rat glial cells, Y. Neurosci. Res., 20 (1988) 12-18. 14 Nishimura, R.N., Dwyer, B.E., de Vellis, J. and Clegg, K., Transformed and primary rat astrocytes have different shock responses: analysis of the major inducible 68 kDa protein, Trans. Int. Heat Shock Conf., 2 (1990) 161. 15 Nishimura, R.N., Dwyer, B.E., Clegg, K., Cole, R. and de Vellis, J., Comparison of the heat shock response in cultured neurons and astrocytes, Mol. Brain Res., 9 (1991) 39-45. 16 Nowak, T.S., Synthesis of a stress protein following transient ischemia in the gerbil, l. Neurochem., 45 (1985) 1635-1641. 17 Pinhvsi-Kimhi, O., Michalovitz, D., Ben-Zeev, A. and Oren, M., Specific interaction between the p53 cellular tumour antigen and major heat shock proteins, Nature, 320 (1986) 182-185.

18 Riabowol, K.T., Mizzen, L.A. and Welch, W.J., Heat shock is lethal to fibroblasts microinjected with antibodies against hsp 70, Science, 242 (1988) 433-436. 19 Subjeck, J.R. and Shyy, T.-T., Stress protein systems of mammalian cells, Am. l. Physiol., 250 (Cell Physiol. 19) (1986) C1C17. 20 Sturzbecher, H.-W., Chumakov, Welch, W.J. and Jenkins, J.R., Mutant p53 proteins bind hsp 72/73 cellular heat shock-related proteins in SV40-transformed monkey cells, Oncogene, 1 (1987) 201-211. 21 Tomasovic, S.E, Functional aspects of the mammalian heatstress protein response, Life Chem. Reports, 7 (1989) 33-63. 22 Towbin, H., Staehelin, T. and Gordon, J., Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications, Proc. Natl. Acad. Sci. U.S.A., 76 (1979) 4350-4354. 23 Uney, J.B., Leigh, EN., Marsden, C.D., Lees, A. and Anderton, B.H., Stereotaxic injection of kainic acid into the striatum of rats induces synthesis of mRNA for heat shock protein 70, FEBS Lett., 235 (!9~8) 215-218. 24 Vass, K., Weld,, W.J. and Nowak, T.S., Localization of 70-kDa stress protein induction in gerbil brain after ischemia, Acta Neuropathol., 77 (1988) 128-135. 25 Welch, W.J. and Suhan, J.E, Cellular and biochemical events in mammalian cells during and after recovery from psychiological stress, J. Cell. Biol., 103 (1986) 2035-2052.