Neuroparhology and Applied Neurobiology 1991,17,139-147

Heat shock in cultured neurons and astrocytes: correlation of ultrastructure and heat shock protein synthesis

R. N. NISHIMURA*t, B. E. DWYER*tll, H. V. VINTERS$IJ, J. DE VELLISgII A N D R. COLE5 * Veterans Afairs Hospital, Sepulveda, California, USA, Departments of tNeurology and $Pathology (Neuropathology), §Mental Retardation Centre and I1 Brain Research Institute, UCLA, Los Angeles, California. USA

NISHIMURA R. N., DWYER B. E., VINTERS H. V., DE VELLIS J. & COLE R. (1991) Neuropathology and Applied Neurobiology 17,139-147

Heat shock in cultured neurons and astrocytes: correlation of ultrastructure and heat shock protein synthesis

Cultured cerebra1 cortical neurons and astrocytes were compared after a brief heat shock. Morphological findings were correlated with the synthesis of the 68 kD heat shock protein (HSP68). Heat shocked neurons demonstrated many severe morphological changes after exposure to temperatures of 43°C for 15 min and 45°C for 10 min. Nuclear membrane ‘blebbing’ with lysis of the membrane, chromatin clumping, and disappearance of the nucleolus were prominent after both conditions. Lysis of the cell membrane was noted in severely injured neurons; this was more prominent at the higher temperature. In addition, alterations to poly- ribosomes, Golgi apparatus, rough endoplasmic reticulum and mitochondria were noted in the cytoplasm of neurons after heat shock. In contrast, no significant changes were noted in either the nucleus or cytoplasm of heat shocked astrocytes. The severity of morphological changes in neurons directly correlated with the low level of induction of HSP68 in neurons. Neurons synthesized much less 68 kD heat shock protein than similarly heat shocked astrocytes. We conclude that cultured cerebral cortical neurons are more susceptible to injury after heat shock than heat resistant astrocytes and that one possible mechanism of injury is failure to synthesize adequate amounts of HSP68 after injury.

Keywords: tissue culture, astrocytes, neurons, heat shock proteins, protein synthesis

INTRODUCTION

The heat shock response refers to the activation of a set of cellular genes and the synthesis of heat shock proteins, and has been found in most cultured cells and animals studied (Craig, 1985; Lindquist, 1986; Subjeck & Shyy, 1986). In the mammalian central nervous system, heat shock proteins (HSPs) were synthesized after traumatic injury (Currie & White, 1981; Brown, Rush & Ivy, 1989; Gower et al., 1989), hypoxia-ischaemia (Jacewicz, Kiessling & Pulsinelli, 1986; Vass,

Correspondence to: Dr R. N. Nishimura, VA Medical Centre, 11 IN-I, 161 11 Plummer Street, Sepulveda, California 91343, USA. The research was supported by the Research Service of the Veterans Administration and the NIH (NS 263 12-02-HVV).

140 R. N. Nishimura et al.

Welch & Nowak, 1988; Dwyer, Nishimura & Brown, 1989, Gonzalez et af., 1989), and heat shock (Cosgrove & Brown, 1983; Sprang& Brown, 1987). Heat shock proteins were synthesized in vivo after heat stress in retinal neurons (Clark & Brown, 1985; Clark & Brown, 1986; Barbe et af., 1988) and cerebellar granule cells (Sprang et af., 1987). However, cortical neurons may not synthesize large amounts of HSP as compared with glial cells at comparable temperatures (Nishimura et af. , 1987; Sprang et al., 1987). These observations were significant in that the synthesis of these proteins, and in particular the major inducible 68 kD HSP, in various cells was associated with survival and tolerance to heat stress (Craig, 1985; Lindquist, 1986; Subjeck & Shyy, 1986). This study compares morphological changes of cultured cortical neurons with astrocytes after comparable heat stress and correlates the morphological changes with synthesis of HSP68.

In this study, members of the 70 kD family of heat shock proteins are denoted as HSP68 (major inducible HSP) and HSP70 (constitutively expressed 70 kD HSP) (Subjeck & Shyy, 1986).

METHODS

Cultures

Neurons were produced by the method of Syapin et af . (1985). Briefly, cerebral hemispheres from fetal rats at 16 days gestation were collected, mechanicaily dissociated and cultured on polylysine coated plastic six-well plates for 5 days in vitro (DIV). Cytosine arabinoside (10 p . ~ , final concentration) was added at 4 DIV to rid the cultures of dividing glial cells. This was continued until 10-14 DIV when the neurons were used for experiments. Purity of the cultures was approximately 90% by immunocytochemical staining with a primary antibody to neuron- specific enolase. Contamination by giial cells was monitored by staining for glial fibrillary acidic protein for astrocytes and galactocerebroside for oligodendrocytes.

Primary astrocyte cultures were prepared using mechanical dissociation (McCarthy & de Vellis, 1980). Secondary cultures were prepared by trypsin dissociation as previously described (Nishimura et al., 1988). Secondary cultures were grown for 14 DIV before use in experiments. Purity of the cultures was estimated at approximately 98% by immunocytochemical staining with glial fibrillary acidic protein (GFAP).

Heat stress

Neurons and astrocytes were switched to Dulbecco's modified Eagle's medium (DMEM)/ Ham's F-12 (1:l) prewarmed to 37°C for 15 min. This was to allow the cultured cells to equilibrate to physiological temperatures and medium change. Heat shock was performed by immersing cultures in a 43 or 45°C & 0.5"C water bath for 15 and 10 min, respectively. Cultures were returned to the 37°C incubator for recovery. At various times after heat stress, cultures were fixed in 2% formaldehyde, 2% glutaraldehyde, 0.25% CaC1, and 0.1 M cacodylic acid. Cultures were post-fixed in osmium tetroxide, then embedded in Medcast using routine pro- cedures (Hunter, 1984). One micron thick sections were stained with Toluidine Blue. Thin sections were stained with lead citrate and uranyl acetate and examined in a Siemens electron microscope. Two to three cultures were examined for each time point after heat shock. Control cultures were changed to DMEM/F-12 medium similar to experimental cultures but were not heat stressed. These cultures were fixed after 4 h in a 37°C incubator.

Heat shock in cultured neurons and astrocytes 14 1

Cell viability after heat shock was assessed by Trypan Blue staining, an indicator of acute cell death, as per Phillips (1973). Randomly sampled heat shocked astrocytes from this study (43°C for 15 min or 45°C for 10 min) showed no Trypan Blue uptake. In contrast, representative experiments of heat shocked neurons showed Trypan Blue staining of up to 5 1 % of surveyed cortical neurons 4 h after heat shock of 43°C for 15 min compared to 1 1 % of control cells. Four hours after heat shock of 45°C for 10 min revealed up to 65% stained neurons compared to 11 % of control neurons.

Labelling with [35S]-methionine

Cultures of neurons and astrocytes were labelled after heat shock as previously described (Nishimura et al., 1988). Briefly, medium was changed to prewarmed (37°C) methionine deficient DMEM. Cultures were heat shocked, then labelled with [35S]-methionine (ICN trans- label) for 3 h. Labelling was terminated by removing the medium, then washing the cells twice with ice cold 0.1 M phosphate buffer, pH 7.4, (PBS), and precipitated with 10% trichloroacetic acid. Cultures were scraped from the plates with PBS and the precipitated proteins were pelleted at 10 000 r.p.m. for 10 min. The supernatant was removed and the pellet was homogenized in 2% SDS sample buffer, then heated to 100°C for 3 min. Aliquots were taken to estimate the radioactivity in counts per minute (c.p.m.) and for protein determination (Lowry et al., 1951).

One dimensional polyacrylamide gel electrophoresis was performed as previously described (Nishimura et al., 1988).

RESULTS

Heat shocked cultured cortical astrocytes and neurons showed a distinct synthesis of a 68 kDa protein (Figure 1). However, induction of the synthesis of this 68 kDa protein was markedly higher in astrocytes compared to similarly treated neurons (Figure 1). This protein has been identified as the major inducible HSP68 on the basis of one and two dimensional polyacryla- mide gels and immunoblotting with specific antibodies to the protein (Nishirnura et al., 1988; Nishimura et al., 1991). In addition to the HSP68, two proteins of 70 and 97 kDa in size were also induced in astrocytes more than in neurons. The difference between neurons and astrocytes in their ability to induce HSPs and, in particular, HSP68, correlated with the astrocyte resistance and the neuron susceptibility to heat induced injury as visualized by morphological studies.

Astrocytes and neurons were compared by light and electron microscopy after heat stress of 43°C for 15 min and 45°C for 10 min. Astrocytes showed no morphological changes 2 and 4 h after heat shock when compared to unheated astrocytes in both conditions. Ultrastructural studies of control (not shown) and heat shocked astrocytes (Figure 2a) demonstrated flattened epithelioid cells with dispersed chromatin within the nuclei and normal appearing cytoplasmic structures including rough endoplasmic reticulum (RER), ribosomes, Golgi apparatus and mitochondria. Abundant intermediate filaments were present and there was the occasional membranous dense body (Figure 2a), a common occurrence in cultured astrocytes.

In contrast, heat shocked cortical neurons showed a variety of morphological alterations after 43 and 45°C heat stress. Light microscopic examination of heat shocked neurons with pale or absent cytoplasm and prominent peripheral chromatin clumping around the nuclear margin with clear nucleoli (not shown). After heat shock many degenerating cells were noted among

142 R. N . Nishimura et al.

- 9 7 kD

- 7 0 -6 8

Figure 1. Autoradiograph of protein synthesis of heat-shocked neurons and astrocytes. Neurons are represented in lanes a, b, c and astrocytes in lanes d, e, f; a and d are controls, b and e were heat shocked for 15 min at 43”C, and c and f were heat shocked for 10 min at 45°C. AIl experimental cultures were labelled for 3 h, 30 pCi/ml of medium. Lanes a-c and Mrepresent separate experiments with equal acid-precipitable counts run in each lane per experiment. Molecular size (kD) is noted in the right hand margin. Incorporation of labelled methionine into acid precipitable protein in neurons after 43 and 45°C heat shock were 9434 (39%) and 11 152 (46%) c.p.m./pg protein, respectively, of control neuron incorporation of 24 038 c.p.m./pg protein. Astrocyte incorporation after heat shock of 43 and 45°C was 18 191 (280%) and 4OOO (62%) c.p.m./pgprotein, respectively, compared with control astrocyte incorporation of 6505 c.p.m./pg protein. Values in percentages represent comparison with control incorporation.

neurons with normal cytoplasm and nuclei. No control cells demonstrated the severe pathologi- cal changes of the nucleus and cytoplasm found after heat shock. Ultrastructural examination of control neurons revealed generally rounded or oval cells with prominent nucleoli within a normal nucleus, normal cytoplasmic organelles; neuritic processes were often visible (Figure 2b). Many heat shocked neurons revealed dissolution or total loss of the cell membrane and early ‘blebbing’ of the outer lamellae of the nuclear membrane with clumping of the chromatin and disappearance of the nucleolus (Figure 2c). These were prominent findings in neurons after

Heat shock in cultured neurons and astrocytes 143

2 h or 4 h of recovery from a heat shock of 43°C for 15 min or 45°C for 10 min. Other findings were markedly swollen RER (not shown), loss of polyribosomes and Golgi apparatus, and swollen mitochondria. However, despite these findings many neurons survived both heat shock conditions and 4 h of recovery. The surviving cells appeared relatively normal when compared with controls (Figure 2d).

DISCUSSION

This study was undertaken in an effort to provide a possible biochemical explanation for neuronal vulnerability to hyperthermic stress. Synthesis of the major inducible HSP68 in some cells was associated with increased survival after heat stress and the development of heat tolerance (Craig, 1985; Lindquist, 1986; Subjeck & Shyy, 1986). In cultured cortical neurons and astrocytes, synthesis of HSP68 correlated well with resistance to morphological changes secondary to heat stress. Astrocytes which synthesized relatively large amounts of HSP68 were resistant to heat shock. In contrast neurons which synthesized low levels of HSP68 showed severe ultrastructural damage after being subjected to heat shock. Figure 1 showed decreased synthesis of HSP68 in heat shocked neurons compared to astrocytes. One explanation is that the transcribed HSP68 mRNA does not translate into protein due to damage to ribosomes after heat shock. This is probably in part the case but our previous studies (Nishimura, et al., 1991) showed decreased amounts of mRNA transcribed in neurons after heat stress compared with astrocytes. These findings suggest that perhaps neurons are unable to generate a similar induc- tion of HSP68 mRNA synthesis as noted in astrocytes. These findings raised the question of how much HSP is needed for resistance to heat stress. Our findings support those of Khan and Sotello (1989) who found that monoclonal antibodies raised against a 70 kDa cytoskeletal associated protein (but presumably recognizing the 70 and 73 kDa HSPs) injected into mouse cortical neurons led to decreased survival and neurite extension of injected neurons after a heat shock of 50°C for 15 min. Control neurons were unaffected. However, it is not clear from this study whether affected neurons synthesized the major inducible HSP.

The majority of reports support the association of resistance to heat shock with HSP68 or 70. However, in mouse plasmacytoma cells and human I316 melanoma cells heat resistance is not associated with induction of HSP68 (Anderson et a1.,1986; Aujame & Firko, 1988). These findings are puzzling but lend support to the findings of Riabowol, Mizzen and Welch (1988) that the constitutive HSP70 serves a similar protective function as HSP68 in these cell types. Two studies strongly support the association of HSP68 with heat resistance. Monoclonal anti- bodies specific for HSP68 and 70 injected into rat embyro fibroblasts failed to survive heat stress (Riabowol et al., 1988). In another study, the inducible HSP68 was selectively decreased by up to 90% after transfection of promoter elements into Chinese hamster ovary cells (Johnston & Kucey, 1988). These transfected cells were unable to survive heat stress.

The most severe ultrastructural changes found in surveyed cortical neurons after heat shock resulted in the ultimate death of the neuron. The lysis of the neuronal cell and nuclear membranes, chromatin clumping, nucleolar degeneration, and dissolution of cytoplasmic structure and organelles were in sharp contrast to the effects of comparable heat stress in astrocytes. Astrocytes demonstrated no nuclear and only minor cytoplasmic changes to organelles after heat shock. Heat shocked rat embryo fibroblasts, which synthesized HSP68, showed unravelling of the condensed nucleolus, and changes in the associated granular ribo- nucleoprotein and fibrillar reticulum (Welch & Suhan, 1985). These changes were not noted in astrocytes or neurons. Heat shocked neurons showed severe pathological changes within the

144 R. N . Nishimura et al.

Heat shock in cultured neurons and astrocytes 145

cytoplasm including swelling of RER and mitochondria, and loss of the Golgi apparatus and polyribosomes. Rat embryo fibroblasts showed mild abnormalities in the mitochondria, RER, and Golgi apparatus after heat stress. A striking change in some heat shocked cells was the collapse of the intermediate filaments around the nucleus (Welch & Suhan, 1985; Collier & Schlesinger, 1986). This was noted in this study. The morphological damage in neurons as opposed to astrocytes was not surprising in view of the known sensitivity of neurons to heat stroke, hypoxia-ischaemia and hypoglycaemia.

The inducible HSP68 and probably 70 kDa protein have been related to the intracellular transport of membrane proteins (Tomasovic, 1989). Since HSP68 is distributed in the nucleolus and the nucleus after heat shock, it is possible that they have a function in stabilizing nuclear membranes or constituents. After 30 to 60 min following heat shock, these proteins are redistri- buted diffusely in the cytoplasm. It can be hypothesized that the synthesis of HSP68 provides the cell with added protection after stress by being rapidly redistributed to sensitive or damaged sites within the cell. The inability to synthesize HSP68 rapidly may lead ultimately to membrane instability and cell death.

In spite of the severe morphological changes seen in many neurons, a subset of intact neurons remained apparently unaffected after heat shock. These cells appeared to be relatively intact which suggested that they might be responsible for synthesis of a large proportion of the HSP68 after heat shock noted in Figure 1. Other cells showed disruption of the rough endo- plasmic reticulum and Golgi apparatus and were unlikely to have synthesized proteins. However, it is possible that the unaffected population of neurons underwent minor reversible morphological changes during heat shock and within 2 h of recovery from heat shock. The basis for survival of this neuronal population is currently being investigated.

With reference to human disease, heat stroke is associated with temperatures of 106°F (41°C) or higher (Malamud, Haymaker & Custer, 1946). The duration of heat stress in reported cases of heat stroke was less than 1-2 h but in some cases persisted for several hours. Our in vitro data showed that irreversible damage occurs at 10 or 15 min exposures to high temperatures of between 109-1 13°F. Our in v i m observations of lethal changes in neurons after a brief heat shock supports Malamud et al. (1946), who showed severe loss of cortical neurons after heat stroke.

The low rate of HSP68 synthesis in cultured cortical neurons compared with astrocytes agrees well with the results of an in situ hybridization study by Sprang and Brown (1987). In that in vivo study, synthesis of the inducible HSP68 mRNA was prominent in rabbit glial cells after heat shock, but little synthesis was noted in cerebral cortical neurons. These results are supported by the findings of Marini et al. (1990), who found that after heat stress of adult rat brain, very little immunoreactivity for HSP68 was found in the cerebral cortical neurons. These results suggested that heat shocked cultured cortical neurons could be a useful model to study the biochemical mechanisms responsible for neuronal injury during and after heat stroke.

Figure 2. Ultrathin sections of heat shocked cultured cortical astrocytes and neurons. a, demonstrates astrocytes after heat shock (15 min at 43°C and 2 h recovery) with essentially normal nuclear and cytoplasmic components. A single membrane-bound structure containing electron-dense debris (arrow), possibly representing a lysosome, is identified ( x 28 200). b, (control after 4 h fresh medium) shows a normal neuron and normal neurites (arrow) ( x 1 1 400). c, shows a neuron exposed to heat shock (15 min, 4 3 T , 2 h recovery) with pronounced blebbing of the outer nuclear membrane (arrows), clumping of nuclear chromatin and dissolution of the cell membrane ( x 9400). d, (10 min, 4 5 ° C 4 h of recovery) reveals a remarkably well preserved neuron and associated neurites ( x 9400).

146 R. N . Nishimura et al.

However, major questions remain to be answered. Can the synthesis of HSP68 protect neurons from heat or other types of injury, and if so, how many copies of the protein per cell are required to afford protection? Also, can other proteins synthesized in response to heat and other stresses confer protection? In addition, HSP68 in neurons and astrocytes has been further characterized in this laboratory (Nishimura et af. , 1991). Not only are the amounts of the HSP68 synthesized in neurons less than in astrocytes, but also the characteristics of immunostaining were different. By immunoblotting with various antibodies for the inducible HSP68, it is apparent that a marked decrease in the synthesis of the HSP68 is noted in cultured neurons after heat stress and that the synthesis of the protein is not constitutive as noted in cultured astrocytes. This latter observation may account for the resistance of cultured astrocytes to heat stress. In addition, the HSP68 did not stain with the same antibodies as the HSP68 of astrocytes, suggesting that HSP68 in neurons and astrocytes was different. This finding was interesting because only one form of HSP68 may help protect a cell from injury.

The major inducible HSP68 has been noted in cerebral cortical neurons of animal models after hypoxia-ischaemia (Vass, Welch 8z Nowak, 1988; Ferriero et af., 1990) and status epilepti- cus (vass et af., 1989). These studies seem to contradict our findings and those of others. However, it may indicate that there are various inducing factors of the heat shock proteins in neurons and that direct heat stress is not a major inducer of HSP68 in cerebral cortical neurons. Exposure of astrocytes in v i m to a pH of 5.5-6.0 (Nishimura et af., 1989) and to hydrogen peroxide (Nishimura et af . , 1988b) induces HSP68. Both of these conditions may contribute to HSP68 induction in neurons after hypoxia-ischaemia and seizures.

ACKNOWLEDGEMENTS We gratefully acknowledge the technical contributions of Man-Anne Akers and Karen Picard to this work.

REFERENCES Anderson R.L., Tao T.W., Betten D.A. & Hahn G.M. (1986) Heat shock protein levels are not elevated in heat-resistant

Aujame L. & Firko H. (1988) The major inducible heat shock protein hsp68 is not required for acquisition of thermal

Barbe M.F., Tytell M., Gower D.J. & Welch W.J. (1988) Hyperthermia protects against light damage in the rat retina.

Brown I.R., Rush S. &Ivy G.O. (1989) Induction of a heat shock gene at the site of tissue injury in the rat brain. Neuron

Clark B.D. & Brown I.R. (1985) Axonal transport of a heat shock protein in the rabbit visual system. Proceedings ofthe National Academy of Sciences 82,1281-1285

Clark B.D. &Brown I.R. (1986) A retinal heat shock protein isassociatedwith elements ofthe cytoskeleton and binds to calmodulin. Biochemical and Biophysical Research Communications 139,974-98 1

Collier N.C. & Schlesinger M.J. (1986) The dynamic state ofheat shock proteins in chicken embryo fibroblasts. Journal of Cell Biology 103,1495-1507

Cosgrove J.W. & Brown I.R. (1983) Heat shock protein in munmalian brain and other organs after a physiologically relevant increase in body temperature induced by D-lysergic acid diethylamide. Proceedings of the National Academy of Sciences 80,569-573

B16 melanoma cells. Radiation Research 105,240-246

resistance in mouse plasmacytoma cell lines. Molecular and Cellular Biology 8,5486-5494

Science 241,1817-1820

2,1559-1564

Craig E.A. (1985) The heat shock response. Annual Reviewsof Biuchemistry55,I151-1191 Currie R.W. & White F.P. (1981) Trauma-induced protein in rat tissues: a physiological role for a 'heat shock' protein?

Dwyer B.E., Nishimura R.N. & Brown I.R. (1989) Synthesis of the major inducible heat shock protein in rat Science 214,72-73

hippocampus after neonatal hypoxia-ischemia. Experimental Neurology 104,28-3 1

Heat shock in cultured neurons and astrocytes 147

Femero D.M., Soberano H.Q., Simon R.P. & Sharp F.R. (1990) Hypoxia-ischemia induces heat shock protein-like

Gonzalez M.R., Shiraishi K., Hisanaga K., Sagar S.M., Mandabach M. & Sharp F.R. (1989) Heat shock proteins as

Gower D.J., Hollman C., Lee S. & Tytell M. (1989) Spinal cord injury and the stress protein response. Journal of

Hunter E.E. (1984) Practical Electron Microscopy, pp. 1-121. Praeger, New York Jacewicz M., KiessIing M. & Pulsinelli W.A. (1986) Selective gene expression in focal cerebral ischemia. Journal of

Johnston R.N. & Kucey B.L. (1988) Competitive inhibition of hsp 70 gene expression causes thermosensitivity. Science

Khan N. A. & Sotelo J. (1989) Heat shock stress is deleterious to CNS cultured neurons microinjected with anti-HSP7O

Lindquist S. (1986) The heat shock response. AnnualReviews of Biochemistry 55,1151-1 191 Lowry O.H., Rosebrough N.J., Rar, A.L. & Randall R.J. (1951) Protein measurements with the Folin phenol reagent.

Journal of Biological Chemistry 193,265-275 Malamud N., Haymaker W. & Custer R.P. (1946) Heat stroke-a clinicopathologic study of 125 fatal cases. Military

Surgeon 99,397-449. Marini A.M., Kozuka M., Lipsky R.H. & Nowak T.S. (1990) 70-Kilodalton heat shock protein induction in cerebellar

astrocytes and cerebellar granule cells in vitro: comparison with immunocytochemical localizations after hyperthermia in vivo. Journal of Neurochemistry 54,1509-1516

McCarthy K.D. & de Vellis J. (1980) Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue. Journal of Cell Biology 85,890-902

Nishimura R.N., Dwyer B.E., Liotta K., Cole R. & de Vellis J. (1987) Heat stress response in primary rat neuronal and ghal cultures. Neurology 37S, 370

Nishimura R.N., Dwyer B.E., Welch W.J., Cole R., de Vellis J. & Liotta K. (1988) The induction of the major heat-stress protein in purified rat glial cells. Journal of Neuroscience Research 20,12-I8

Nishimura R.N., Dwyer B.E., Cole R. & de Vellis J. (1988b) Induction of the 68/72 kDa heat-shock protein during hydrogen peroxide toxicity. In Neural Development and Regeneration-Cellular and Molecular Aspects, pp. 227-23 1. Springer-Verlag, Heidelberg

NishimuraR.N.,Dwyer B.E.,ColeR.,deVellisJ.&PicardK. (1989)Inductionofthemajorinducible68 kDaheat-shock protein after rapid changes ofextracellularpH in cultured rat astrocytes. Experimental CellResearch 180,276-280

Nishimura R.N., Dwyer B.E., Clegg K., Cole R. & de Vellis J. (1991) Comparison of the heat shock response in cultured neurons and astrocytes. Molecular Brain Research, 9,3945

Phillips H.J. (1973) Dye exclusion tests for cell viability. In Tissue Culture Merhoak and Applicarions, pp. 406-408. Academic Press, New York

Riabowol K.T., Mizzen L.A. & Welch W.J. (1988) Heat shock is lethal to fibroblasts microinjected with antibodies against hsp 70. Science 242,433-436

Sprang G.K. & Brown I.R. (1987) Selective induction of a heat shock gene in fibre tracts and cerebellar neurons of the rabbit brain detected by in situ hybridization. Molecular Brain Research 3,89-93

Subjeck J.R. & Shyy T.-T. (1986) Stress protein systems of mammalian cells. American Journal of Physiology 250 (Cell Physiology 19), CI-Cl7

Syapin P.J., Cole, R., de Vellis J. & Noble F.P. (1985) Benzodiazepine binding characteristics of embryonic rat brain neurons grown in culture. Journalof Neurochemistry 45,1797-1801

Tomasovic S.P. (1989) Functional aspects of the mammalian heat-stress protein response. Lqe Chemistry Reports 7, 33-63

Vass K., Welch W.J. & Nowak T.S. (1988) Localization of 70-kDa stress protein induction of gerbil brain after ischemia. Acta Neuropathologica 77,128-135

Vass K., Berger M.L., Nowak T.S., Welch W.J. & Lassmann H. (1989) Induction of stress protein HSP70 in nerve cells after status epilepticus in the rat. Neuroscience Letters 100,259-264

Welch W.J. & Suhan J.P. (1985) Morphological study of the mammalian stress response: characterization of changes in cytoplasmic organelles, cytoskeleton, and nucleoli, and appearance of intranuclear actin filaments in rat fibroblasts after heat-shock treatment. Journal of Cell Biology 101,1198-121 1

(HSP 72) immunoreactivity in neonatal rat brain. Beevelopmental Brain Research 53,145-1 50

markers of neural injury. Molecular Brain Research 6,93-I00

Neurosurgery 70,605-6 I 1

Cerebral Blood Flow and Metabolism 6,263-272

242,1551-1554

antibodies. Biology of the Cell 65,199-202.

Received 5 June 1990 Accepted 3 November 1990