Heat Shock Proteins and Neuroprotection

Yenari MA.

Publication Details

In response to many metabolic disturbances and injuries including stroke, neurodegenerative disease, epilepsy and trauma, the cell mounts a stress response with induction of a variety of proteins, most notably the 70 kD heat shock protein (Hsp70). The possibility that stress proteins might be neuroprotective was suspected because Hsp70, in particular, was induced to high levels in brain regions that were relatively resistant to injury. Hsp70 expression was also correlated with the phenomenon of induced tolerance. With the availability of transgenic animals and gene transfer, it has become increasingly clear that such heat shock proteins do indeed protect cells from injury. Several reports have now shown that selective overexpression of Hsp70 leads to protection in several different models of nervous system injury. This review will cover these studies, along with potential mechanisms by which Hsp70 might mediate cellular protection.


It is well known that cells respond to external stress in a highly conserved, stereotypical fashion. The stress response results in gene expression following such environmental challenges as high temperatures, ischemia, excitotoxin exposure and other stresses which result in protein denaturation (see recent reviews refs. 1,2). Heat shock proteins (Hsps) are induced by stressful stimuli and are thought to assist in the maintenance of cellular integrity and viability. Hsps consist of both stress-inducible and constitutive family members, as well as members that are associated with specific organelles. Constitutively synthesized Hsps perform housekeeping functions. They function as molecular chaperones by helping nascent polypeptides assume their proper conformation by binding to nascent proteins via their C-terminal domain. Hsps are also involved in antigen presentation, steroid receptor function, intracellular trafficking, nuclear receptor binding, and apoptosis.1,3 However, many are also upregulated by stress. Inducible Hsps prevent protein denaturation and incorrect polypeptide aggregation during exposure to physiochemical insults. Hsps may prevent protein unfolding or aggregation and enhance cell survival.

The 70 kD family of stress proteins is one of the most extensively studied. Included in this family are Hsc70 (heat shock cognate, the constitutive form), Hsp70 (the inducible form is also referred to as Hsp72), Grp75 (a constitutively expressed mitochondrial glucose regulated protein), Grp78 (a constitutively expressed glucose regulated protein found in the endoplasmic reticulum) and heme oxygenase-1 (HO-1, a stress protein involved in bilirubin metabolism). Other members of this family include Hsp40, Hsp90, Hsp27 and Hsp chaperone proteins such as the DNAJ homologs (HDJ-1 and HDJ-2 in humans), the latter being proteins which assist other Hsps in protein folding. Following a variety of central nervous system insults, Hsp70 is expressed at especially high levels and is present in the cytosol, nucleus and endoplasmic reticulum. Because inducible Hsp70 is not usually detectable under normal conditions and its expression is the most robust of the Hsps, it is often regarded as a diagnostic marker for stress. The stress proteins of the Hsp70 family function as chaperones, interacting transiently with many proteins in an ATP-dependent manner. Denatured proteins are thought to serve as the stimulus for stress protein induction. These proteins activate heat shock factors (HSFs) within the cytosol by dissociating other Hsps that are normally bound to HSFs. Once liberated, HSFs are phosphorylated and form trimers. The trimers then enter the nucleus and bind to heat shock elements (HSEs) within the promoters of different heat shock genes leading to transcription and Hsp translation.3,4 Once expressed, Hsp70 binds to then quickly releases denatured proteins in an ATP-dependent fashion (Fig. 1). There appear to be two main functional domains within Hsp70. The N-terminal end contains an ATP-binding domain whereas the C-terminal part contains a substrate binding domain. Substrate binding cannot occur in the absence of ATP binding which is regulated by an EEVD motif within the C terminus.5 Therefore, various stresses resulting in protein denaturation stimulate Hsp70 expression, which presumably acts to restore protein structure and function.

Figure 1. Mechanism of heat shock protein activation.

Figure 1

Mechanism of heat shock protein activation. Various stresses result in protein denaturation (1). Heat shock proteins (Hsp) are normally bound to heat shock factors (HSF) (2), but dissociate in the presence of denatured proteins (DP). Once dissociated, (more...)

In the nervous system, the heat shock proteins are induced in a variety of pathologic states including cerebral ischemia, neurodegenerative diseases, epilepsy and trauma. Expression has been detected in a variety of cell populations within the nervous system including neurons, glia, and endothelial cells.6 Although Hsp70 has been long thought to protect cells by preserving tertiary protein structure and preventing protein aggregation, direct evidence has been lacking. Gene transfer techniques and transgenic animal strain have now made it possible to selectively overexpress Hsps to better understand the precise role they play in cellular injury. This review will focus on recent results that help to elucidate the role of Hsp70 in neuroprotection.

Where and When Is Hsp70 Expressed?

Several groups using various models of experimental nervous system stress and injury have studied the anatomic and temporal expression of Hsp70. In a model of global cerebral ischemia (a model which results in cerebral injury similar to that following cardiac arrest), Hsp70 mRNA was detected in the hippocampus within hours of ischemia onset and decreased when neurons were lost.4 Protein expression followed a few hours later7. Brief periods of ischemia (3–8 minutes) resulted in Hsp70 protein expression within neurons and some glia after 24 hours, but longer durations of ischemia (10–20 minutes)4,7 showed decreased Hsp70 protein expression. Following 60 minutes of transient focal cerebral ischemia (a model of stroke), Hsp70 protein expression with similar patterns were observed with persistent expression as far out as 7 days.4,8 Hsp70 was observed within neurons and astrocytes at the infarct periphery, but only endothelial cells expressed it within brain regions where ischemia was the most severe (striatum).8 With increasing duration of middle cerebral artery (MCA) occlusion, graded levels of Hsp70 expression are observed within neurons, microglia and endothelial cells that decreased after the most severe ischemic insults.4,9,10 It has previously been suggested that glia can transfer Hsp70 to neurons; therefore, neuronal Hsp70 expression may be linked to a potential protective mechanism from glia.1

In models of excitotoxicity and seizures, kainic acid (KA) administration resulted in widespread Hsp70 induction, particularly within cortical and hippocampal neurons.11,12 Hsp70 mRNA was present in dentate granule cells within 6 hours of KA application and protein was observed by 12 hours, then decreased after 24 hours.11Protein expression has been observed to be especially prominent within brain regions known to be resistant to injury in this model, whereas less expression was seen in degenerating neurons.13On the other hand, Hsp70 was observed within hippocampal neuron populations such as CA1, CA3 and dentate granule cells,12 and appeared to precede and correlate with the extent of neuronal damage assessed a few days later.14 Furthermore, MK801, an antagonist of the N-methyl-D-aspartate receptor, reduced both Hsp70 expression and KA-induced neuronal injury.12

There appear to be similar patterns of Hsp70 expression in different in vivo models of cerebral stress, with graded expression depending on the severity of the insult and the cell population. In general, Hsp70 is observed in the brain within several hours of the insult, and persists for a few days.

Correlative Evidence for a Neuroprotective Role

Stress proteins are increased in resistant cell populations following injury: Whether Hsp70 serves a protective role against various cerebral insults has been debated. It has long been known that certain cell populations are less vulnerable than others to various cerebral insults such as global cerebral ischemia and KA toxicity. Following ischemia, several groups have reported that Hsp70 protein is expressed only in cell populations that will ultimately recover from cerebral ischemia10 such as infarct borders (focal cerebral ischemia) and dentate granules (global ischemia). With infarction, expression is absent within brain tissue and is restricted to vascular endothelial cells.10,13 Similar results were observed in the KA model.4,13These studies imply that Hsp70 is expressed in cell populations known to be resistant to injury; therefore, Hsp70 may be responsible for the observed resistance.

In contrast, other data suggest that Hsp70 is expressed in vulnerable cell populations or is expressed regardless of the fate of the cell.4,15 These contrasting observations could be limited by the severity of the insults, and whether or not protein translation has occurred. Dysjunction of Hsp70 mRNA and protein expression following focal cerebral ischemia has been observed. While Hsp70 mRNA was expressed extensively, protein was observed only within potentially salvageable peri-infarct regions.1,4 but not in damaged regions. Therefore, Hsp70 may be capable of rescuing injured cells provided the protein has been translated.

Hsp expression correlates with protection in tolerance paradigms: Whether Hsp70 is playing a protective role against various insults could not be addressed in the above studies. However, Hsp70's increased expression led investigators to study its potential role in the protective effect of “induced tolerance”, a phenomenon whereby a prior sublethal insult leads to protection against a subsequent severe insult. For instance, thermal or chemical stress protects against excitotoxic insults such as glutamate exposure in cultured neuronal cells and in whole animal models.16 Prior heat shock, sublethal ischemia or chemical toxins protect against subsequent ischemia.1,4 Induction of Hsp70 and protection from subsequent injury has been demonstrated in neuron cultures and in whole animal models of thermal stress, global and focal cerebral ischemia (Reviewed by Massa et al4 and Parsell et al17). The expression of Hsp70 also correlated with the period within which tolerance was observed, leading some to believe that Hsp70 may explain the observed phenomenon.18 Interestingly, the interval of tolerance was associated increases in neuronal Hsp70 and activated astrocytic Hsp27.19 However, these sublethal insults cause a host of changes in protein expression and metabolism rather than selectively increasing Hsp70 synthesis, and the role of Hsp70 in mediating the protective effects has been far from decisive. In fact, protection in tolerance experiments has been noted even when HSP70 and other new protein synthesis is blocked,20 suggesting that “induced tolerance” does not always require protein production.17

Neuroprotection with Hsp70 Overexpression

To directly test a protective role of Hsp70, cells can be made to selectively overexpress the protein, or expression can even be blocked. This can now be accomplished using transgenic animal models, gene transfer or antisense oligonucleotides. A number of studies have been published where Hsp70 has been overexpressed in various non-neuronal cell lines, and found protection against numerous stresses including heat shock,21 oxidative stress,22 apoptotic stimuli23,24and ischemia-like conditions.25,26 In the nervous system, Hsp70 overexpression in cultured hippocampal27,28 and peripheral29–32 neurons and glia29,30 similarly protected against insults such as heat shock and metabolic stresses. Hsp70 expression can also be suppressed with antisense oligonucleotides that inhibit transcription.33 Using this approach, Sato et al33 found that the protection from induced tolerance was reversed with Hsp70 blockade using antisense oligonucleotides. At the in vivo level, a few laboratories have generated transgenic mice capable of overexpressing hsp70 from a constitutive promoter.34–36 In models of myocardial ischemia, these mice were found to have reduced infarct volume,36 improved recovery of ATP stores35 and better contractile recovery.34

However, Hsp70 overexpression is not protective in all instances. Using a defective herpes simplex virus (HSV) vector, Fink et al28 showed that Hsp70 overexpression protected cultured hippocampal neurons from severe heat shock, but failed to protect against direct application of glutamate or 3-nitropropionic acid (3-NP), a mitochondrial toxin. Wagstaff et al37 showed that Hsp70 overexpression protected cultured peripheral neurons from thermal and simulated ischemia, but not apoptotic stimuli. Observations from our laboratory have shown that glial cell cultures from brains of transgenic mice that overexpress Hsp70 are resistant to hydrogen peroxide injury, but are less resistant to other injury paradigms.38 Interestingly, hippocampal neurons were more resistant to certain stresses compared to cortical neuron cultures. These results suggest that Hsp70 protects against some but not all kinds of central nervous system injury, and that the protective effects may be related to the nature and severity of the insults. It has previously been shown that the response to heat shock and Hsp70 expression can vary depending on the type and age of the cell;39 therefore, it is likely that the protective effects may similarly be dependent on such factors.

Hsp protection following ischemia and ischemia-like insults: Nervous system ischemia can be modeled in vitro by exposing cultures to ischemia like conditions such as substrate and/or oxygen deprivation,38,40 excitotoxin exposure28 or incubation in simulated ischemic buffers.32 Recent work from our group showed that cultured hippocampal, but not cortical neurons from transgenic mice overexpressing Hsp70 were protected from excitotoxin exposure and oxygen and glucose deprivation.38 Glial cultures isolated from this same mouse strain were also resistant to substrate deprivation. Higher levels of transgene expression could be attained by using retroviral vectors which increased Hsp70 protein greater than 10 fold (R. G. Giffard, personal communication). Xu et al41 and Papadopoulos et al40 used such retroviral vectors to overexpress Hsp70 in astrocyte cultures. Exposure to isolated glucose or combined oxygen and glucose deprivation led to robust glial survival following either insult. Conversely, suppression of Hsp70 in hippocampal neuron cultures with an antisense oligonucleotide worsened injury following heat shock.33

At the whole animal model level, cerebral ischemia models using Hsp70 overexpressing transgenic mice have been reported. Whereas various strains of Hsp70 transgenic mice are protected against myocardial ischemia,34–36 this has not always been the case for brain ischemia. The reasons for these discrepancies are not clear, but could be due to strain differences in transgene expression within different organ systems, or because brain regions have varying susceptibilities depending on the model studied. A few groups have studied permanent MCA occlusion in various Hsp70 transgenic mice strains with conflicting results. Constitutive overexpression of Hsp70 in one strain using a cytomegalovirus (CMV) enhancer combined with a β-actin promoter resulted in near complete protection determined by overall reduction in infarct size.42 Another group using a similar model but a different mouse strain found only improved hippocampal neuron survival by 24 h; however, overall infarct size was not affected.43 In this latter group, Hsp70 expression was under the control of a β-actin (constitutive) promoter. In yet a third study, our group38 examined a third strain of Hsp70 transgenic mice and failed to find any differences between infarct size or hippocampal neuron survival. These differing results could be due to background strain differences and the limitations of transgenic animals, such as developmental alterations in other biochemical systems caused by transgene overexpression. The extent of Hsp70 expression might also be different between the transgenic strains, since different promoters were used. The transgenic model containing the CMV enhancer increased transgene expression by 5–10 fold.42 In contrast, the model using the β-actin promoter alone resulted in only 2 fold increases.38 Therefore, higher transgene levels could account for the marked protection in the first study,42 and the less robust or lack of protection in the latter two studies.38,43

Whether lifelong Hsp70 expressing animals have alterations in other systems is unclear, though the gross phenotype appears normal. Therefore, viral vector mediated transfer of Hsp70 can be used to study neuroprotection in wildtype animals. Our laboratories have recently taken advantage of the neurotropic properties of herpes simplex virus (HSV) to preferentially overexpress Hsp70 in neurons.28 HSV vectors are capable of transfecting approximately 70% or more cortical neurons following direct injection into the brain.44 One of the major constraints in overexpressing genes in vivo has been the relatively low efficacy and efficiency of viral vector mediated gene transfer. To circumvent this problem, bipromoter vectors were developed which contained both the genes for hsp70 and a reporter gene, lacZ.45–47 With this approach, it becomes possible to identify the subset of transfected cells that also express the transgene of interest. Following direct intracerebral injection, vector mediated Hsp70 protein and reporter gene (β-galactosidase or β-gal) expression occurred about 9–12 hours later. The number of targeted Hsp70 overexpressing neurons were differentiated from cells that expressed endogenous Hsp70 by identifying the β-gal positive cells (Fig. 2A). By counting the number of β-gal-positive cells in the ischemic side of the brain and comparing it to β-gal counts from the non-ischemic contralateral side, we found greater numbers of striatal neurons 48 hours after one hour of MCA occlusion when vector was injected 12 hours prior to ischemia onset48 (Figs. 2A-B and 3). However, overall infarct size was not affected given the limited number of cells the vectors can target. Similar results were observed against KA-induced excitotoxicity where viral vectors were injected into the hippocampi of animals 15 hours before systemic KA administration (Fig. 3). Hippocampal dentate neuron survival at 24 hours was improved with Hsp70 overexpression. Hsp70 overexpression also protected against global cerebral ischemia. Using the HSV vectors, our group overexpressed Hsp70 within the vulnerable CA1 neurons of the hippocampus (Fig. 2C-D).49 Animals were then subjected to 8 minutes of forebrain ischemia, and the proportion of surviving, vector-targeted cells compared to sham-injured transfected controls were determined 3 days later. Like the results from the focal cerebral ischemia and KA models, more Hsp70-transfected CA1 hippocampal neurons remained compared to control vector transfected (Fig. 349). Another study by Kelly et al50 used adenoviral vectors to transfer Hsp70 to striata of mice. Following 20 minutes forebrain ischemia in their model, they, too, found neuroprotection following intrastriatal injection of Hsp70 expressing vectors compared to β-galactosidase expressing control vectors. These data indicate that Hsp70 is protective at the in vivo level in adult wild-type animals

Figure 2. Hsp70 protects neurons from a variety of central nervous system insults.

Figure 2

Hsp70 protects neurons from a variety of central nervous system insults. Several surviving, beta-galactosidase (β-gal) positive cells (arrows) which received a Hsp70 overexpressing vector are evident within an ischemic striatum 48 h after 1 hour (more...)

Figure 3. Hsp70 overexpression protects against a variety of insults in vivo.

Figure 3

Hsp70 overexpression protects against a variety of insults in vivo. Following focal cerebral ischemia, normalizing counts of β-galactosidase-positive neurons within the ischemic striatum to the contralateral non-ischemic side showed significantly (more...)

Hsp protection in neurodegenerative diseases: Hsps also appear to protect against neurodegeneration in models of polyglutamine expansion disease. Polyglutamine diseases constitute a heterogeneous set of diseases characterized by the cellular accumulation of glutamine aggregates. These diseases contain a genetic defect characterized by repeating trinucleotide CAG motifs. CAG repeats result in expanses of glutamine which result in misfolded, aggregated protein, leading to toxicity and cell death. Human “trinucleotide repeat” diseases include Huntington's disease, Kennedy's disease (spinobulbar muscular atrophy), certain spinocerebellar ataxic disorders and others. Overexpressing various Hsps in cells transformed with polyglutamine sequences has led to reduction in protein aggregates and less toxicity. By transfecting cells with genes that express proteins associated with spinocerebellar degeneration, such as mutant ataxin-151 or ataxin-3,52 protein aggregates could be observed. Many of these aggregates were associated with Hsps, particularly Hsp70, indicating that these expanded proteins can themselves elicit a stress response. After transfecting such cells containing these mutant ataxins, aggregates could be suppressed by overexpression of the Hsp40 chaperones HDJ-1 and HDJ-2.51,52 In similar models of Kennedy's disease, cell lines were transfected with mutant androgen receptor. Overexpression of Hsp40 and Hsp7053 or HDJ-154 could also suppress protein aggregates. In a model of Huntington's disease, these same Hsps could inhibit the assembly of abnormal polyglutamine containing huntingtin protein into amyloid fibrils.55 Recently, a murine model of spinocerebellar ataxia-1 (SCA-1) was crossed with a transgenic mouse model which overexpressed HSP70.56 SCA-1 mice overexpressing Hsp70 performed better on behavioral tests, and Pukinje cell neuropathology was improved by Hsp70 overexpression. However, characteristic nuclear inclusions were not altered by HSP70 expression. Nevertheless, such data are consistent with the notion that Hsps protect cells from injury through their chaperone functions, and possibly by preventing protein aggregation.

Hsps and clinical relevance: Whether Hsp70 can protect cells when delivered after an insult would have obvious clinical implications. Our group recently found that overexpressing Hsp70 up to 2 hours following middle cerebral artery occlusion resulted in improved striatal neuron survival.57 When Hsp70 expressing vector was delivered 5 hours post insult, neuroprotection was lost (Fig. 4). Since the HSV vectors do not begin to express transgene until 4-6 hours post transfection,45-47 this would suggest a temporal therapeutic window of 6-8 hours post stroke.

Figure 4. Hsp70 protects striatal neurons when delivered after stroke onset.

Figure 4

Hsp70 protects striatal neurons when delivered after stroke onset. The proportion of surviving vector-targeted neurons is improved even when Hsp70 vector is delivered 0.5 and 2 h post ischemia, but is no longer protective when administered at 5 h (HSP=Hsp70 (more...)

Another potentially clinically applicable approach for utilizing the neuroprotective properties of the Hsps might be to pharmacologically induce them in the brain. Geldanamycin, a benzoquinoid ansamycin, is one such compound currently being tested for cancer therapy.58 It binds to Hsp90, and possibly induces other Hsps by releasing HSFs to bind HSEs and increase Hsp70 expression.59 In a recent presentation by Lu et al,60 geldanamycin induced expression of several Hsps including Hsp70 and HO-1. Intrathecal and intraperitoneal administration lead to neuroprotection within hippocampal CA1 when given 6 hours prior to 5 minutes of forebrain ischemia in gerbils. Whether geldanamycin might protect the brain when given after the onset of cerebral ischemia is not yet known. However, Xiao et al61 showed that it can prevent glutamate-induced oxidative toxicity in hippocampal cell cultures, even if given 4 hours after glutamate exposure.

Potential Mechanisms of Protection

The mechanism of protection with Hsp70 is believed to be related to its chaperone functions leading to prevention of protein malfolding and aggregation. Polyglutamine diseases are known to have accumulations of abnormal protein chains. Following both focal62 and global63 cerebral ischemia, protein aggregates have been described within vulnerable cell populations. Hsps appear to interfere with the formation of such aggregates for some neurodegenerative disease, but has yet to be demonstrated in ischemia models.

Nevertheless, certain observations suggest that Hsp's chaperone functions may play a role in protection from injury through improving function of several different proteins. For instance, there may be a relationship between stress protein protection and oxidative injury. Work by Polla and colleagues64 demonstrated that Hsp70 induction correlated best with protection of a cell line from hydrogen peroxide induced oxidative injury. In cultured glial cells, Xu and Giffard41 noted that protection with Hsp70 overexpression against glucose deprivation or hydrogen peroxide exposure was associated with increased glutathione levels, suggesting that Hsp70 may protect cells through an antioxidant mechanism, perhaps by improving protein stability of endogenous antioxidants. Prolonged expression of hsp70 mRNA among Cu-Zn superoxide dismutase (SOD1) transgenic mice was observed following focal65 and global66 cerebral ischemia and kainic acid-induced injury.67 The reasons for these observations are not clear, but might be due to an overall reduction in oxidative stress, which permits extended expression of Hsp70. Prolonged Hsp70 expression could then enhance SOD1's protective effect through its chaperone functions. Another possibility is that increased levels of SOD1 without compensatory increases in glutathione peroxidase could lead to an accumulation of hydrogen peroxide, another reactive species. Consequently, hydrogen peroxide could lead to Hsp70 induction due to the increased oxidative stress.

Hsps may protect by mechanisms unrelated to their chaperone function as well. Insults such as ischemia, hypoxia and heat shock are notable for toxic increases in intracellular calcium that lead to activation of various proteases and endonucleases and ultimately, cell death. In non-neuronal cell lines, overexpression of Hsp70 attenuates this influx of calcium by desensitizing other ion regulating system such as the Na+/Ca++ exchanger (reviewed by Kiang & Tsokos3). Hsp70 may also alter other proteins or genes known to be involved in ischemic and excitotoxic injury and inflammation by inhibiting nitric oxide synthase generation and nuclear translocation of the transcription factor, NF-κB.68,69 Others have shown that prior thermal stress leads to inhibition of the inflammatory response, and this inhibition was associated with increased levels of Hsp70 induction and decreased nuclear NF-κB translocation.69,70 It has been speculated that Hsp70 could interact with NF-kB's inhibitor protein, IκB, and prevent IκB phosphorylation and NF-κB dissociation.68 Anther possibility is that Hsp70, which can also translocate to the nucleus, could also compete for the same nuclear pore as NF-κB.68,71

Recent studies in non-neuronal cell lines have shown that Hsp70's protective effect may also be due to anti-apoptotic mechanisms. Apoptosis, or programmed cell death is known to occur in pathological states either by activation of specific death receptors (Reviewed by Ashkenazi & Dixit72), or internally, via mitochondrial release of cytochrome c (Reviewed by Green & Reed73). Central to mitochondrial-based apoptosis is the assembly of the so-called apoptosome. This occurs when procaspase-9 binds to Apaf-1 in the cytosol, and becomes activated when cytochrome c is release from the mitochondria to the cytosol. This release of cytochrome c is blocked by the anti-apoptotic protein, Bcl-2. Activated caspase-9 then leads to activation of various effector caspases including caspase-3.

Several papers have now established that Hsps can interfere with apoptosis in various systems (Fig. 6). Overexpression of Hsp70 in lymphoid tumor cell lines appears to inhibit apoptosis by blocking caspase activation and activity.21,74,75 It has recently been shown that Hsp70 could inhibit caspase activation by interfering with Apaf-1, and prevent the recruitment of procaspase-9 to the apoptosome.76,77 How this occurs is not yet clear, but has been hypothesized to be due a direct competition between Hsp70 and procaspase-9 for Apaf-1 binding.77,78 However, Mosser et al21 demonstrated that Hsp70 did not appear to interfere with caspase-3 processing, and other data suggest that Hsp70 interferes with apoptosis downstream of caspase activation.23 Yet other studies in tumor cells lines have shown that Hsp70 may block stress kinase (SAPK/JNK) activation when Hsp70 overexpression is induced,21 but not when constitutively expressed.21,23,24 However, Hsps do not appear to block Fas mediated, receptor activated apoptosis.78,79 Together, these studies suggest that Hsp70 probably acts at multiple sites to confer protection in models of apoptosis.

Figure 6. Protection from apoptosis by Hsp70 may have multiple sites of action.

Figure 6

Protection from apoptosis by Hsp70 may have multiple sites of action. In non-neuronal cell cultures, Hsp70 appears to interfere with several different steps in apoptosis. Following an appropriate apoptotic insult (e.g., thermal stress, tumor necrosis (more...)

Whether Hsp70 prevents apoptosis from occurring in the brain has yet to be definitively shown; however, cells with DNA fragmentation (detected by DNA nick end labeling) following focal cerebral ischemia rarely expressed Hsp70 protein.80 Other related stress proteins have been shown to inhibit apoptosis in the brain. For instance, Bcl-2, an anti-apoptotic protein, is increased in neurons of mice which overexpress heme oxygenase (HO-1),81 and GRP78 appears to inhibit caspase activation in cultured hippocampal rat neurons.82 Our group also found that Hsp70 overexpression also increased levels of Bcl-2 protein. Following viral vector transfection, Bcl-2 was present in uninjured Hsp70 expressing neurons compared to control vector transfected cultures (Fig. 5A). Following oxygen glucose deprivation, Hsp70 overexpression still resulted in increased Bcl-2 levels (Fig. 5B). We observed similar patterns in vivo where higher proportions of CA1 hippocampal neurons transfected with Hsp70 also expressed Bcl-2 following 8 minutes of forebrain ischemia.49 These findings would be consistent with those of others who reported that Hsps decrease cytochrome c release following hydrogen peroxide exposure.83 In this scenario, it is conceivable that Hsps, by an as yet unknown mechanism, could increase Bcl-2 expression, which in turn block cytochrome c release and subsequent effector caspase activation. However, a few reports using in vitro models showed that Hsp70 and Hsp90 do not protect against apoptotic stress in peripheral neurons,84 but that Hsp27 does.32 These latter observations might suggest that stress proteins, though not necessarily Hsp70, do play a role in blocking neuronal apoptosis.

Figure 5. Hsp70 increases Bcl-2 expression: Cultured hippocampal and cortical neurons transfected with defective herpes simplex viral vectors express higher levels of Bcl-2 protein.

Figure 5

Hsp70 increases Bcl-2 expression: Cultured hippocampal and cortical neurons transfected with defective herpes simplex viral vectors express higher levels of Bcl-2 protein. Western blots show increased Bcl-2 protein in cell extracts of neurons 24 hours (more...)

Other mechanisms of protection might also be involved. Hsp70's protein binding properties are dependent on ATP;5 however, insults such as ischemia are especially notable for reduction in metabolic stores despite the observed neuroprotection from a few different laboratories. Interestingly, Hsp70 variants containing mutations in the ATP binding domain still protect tumor cells from heat shock.85 Therefore, ATP-independent mechanisms underlying the observed neuroprotection is plausible, but have yet to be tested in appropriate models.


Recent studies from different laboratories have now established that the stress response provides the organism with a cellular process for self preservation, and that stress proteins themselves can directly protect cells from death. The specific mechanisms underlying this protection are not well understood, but are likely multifactorial, encompassing a wide range of cellular chaperone functions from the prevention of protein aggregation to interfering with various death cascades. Capitalizing on the cell's natural response to stress is an attractive therapeutic target for a variety of nervous system diseases.


This work was funded in part by National Institutes of Health grants #P01 NS37520 and R01 NS40516 and American Heart Association Grant #0060091Y. The author would like to thank the many lab members in the Departments of Neurosurgery, Biology and Anesthesiology at Stanford University who contributed to some of the work presented here, and Ms. Beth Houle for preparation of the figures.


Sharp FR, Massa SM, Swanson RA. Heat-shock protein protection. Trends Neurosci. 1999;22(3):97–99. [PubMed: 10199631]
Yenari MA, Giffard RG, Sapolsky RM. et al. The neuroprotective potential of heat shock protein 70 (HSP70) Mol Med Today. 1999;5(12):525–531. [PubMed: 10562718]
Kiang JG, Tsokos GC. Heat shock protein 70 kDa: Molecular biology, biochemistry, and physiology. Pharmacol Ther. 1998;80(2):183–201. [PubMed: 9839771]
Massa SM, Swanson RA, Sharp FR. The stress gene response in brain. Cerebrovasc Brain Metab Rev. 1996;8(2):95–158. [PubMed: 8727184]
Freeman BC, Myers MP, Schumacher R. et al. Identification of a regulatory motif in Hsp70 that affects ATPase activity, substrate binding and interaction with HDJ-1. EMBO J. 1995;14(10):2281–2292. [PMC free article: PMC398335] [PubMed: 7774586]
Foster JA, Brown IR. Differential induction of heat shock mRNA in oligodendrocytes, microglia, and astrocytes following hyperthermia. Brain Res Mol Brain Res. 1997;45(2):207–128. [PubMed: 9149095]
Vass K, Welch WJ, Nowak T S Jr. Localization of 70-kDa stress protein induction in gerbil brain after ischemia. Acta Neuropathol (Berl) 1988;77(2):128–135. [PubMed: 3227811]
Kinouchi H, Sharp FR, Hill MP. et al. Induction of 70-kDa heat shock protein and hsp70 mRNA following transient focal cerebral ischemia in the rat. J Cereb Blood Flow Metab. 1993;13(1):105–115. [PubMed: 8416999]
Soriano MA, Planas AM, Rodriguez-Farre E. et al. Early 72-kDa heat shock protein induction in microglial cells following focal ischemia in the rat brain. Neurosci Lett. 1994;182(2):205–207. [PubMed: 7715811]
Nowak T S Jr., Jacewicz M. The heat shock/stress response in focal cerebral ischemia. Brain Pathol. 1994;4(1):67–76. [PubMed: 8025704]
Armstrong JN, Plumier J C L, Robertson HA. et al. The inducible 70,000 molecular/weight heat shock protein is expressed in the degenerating dentate hilus and piriform cortex after systemic administration of kainic acid in the rat. Neuroscience. 1996;74(3):685–93. [PubMed: 8884765]
Zhang X, Boulton AA, Yu PH. Expression of heat shock protein-70 and limbic seizure-induced neuronal death in the rat brain. Eur J Neurosci. 1996;8:1432–1440. [PubMed: 8758950]
Sharp FR. Stress proteins are sensitive indicators of injury in the brain produced by ischemia and toxins. J Toxicol Sci. 1995;20(4):450–453. [PubMed: 8531239]
Gass P, Prior P, Kiessling M. Correlation between seizure intensity and stress protein expression after limbic epilepsy in the rat brain. Neuroscience. 1995;65(1):27–36. [PubMed: 7753400]
Chopp M, Li Y, Dereski MO. et al. Neuronal injury and expression of 72-kDa heat-shock protein after forebrain ischemia in the rat. Acta Neuropathol (Berl) 1991;83(1):66–71. [PubMed: 1792866]
Yang YL, Lin MT. Heat shock protein expression protects against cerebral ischemia and monoamine overload in rat heatstroke. Am J Physiol. 1999;276(6 Pt 2):H1961–H1967. [PubMed: 10362676]
Parsell DA, Lindquist S. Heat shock proteins and stress tolerance In: Morimoto RI, Tissieres A, Georgopoulos C, eds. The Biology of Heat Shock Proteins and Molecular Chaperones New York: Cold Spring Harbor Laboratory Press,1994457–494.
Yang YL, Lu KT, Tsay HJ. et al. Heat shock protein expression protects against death following exposure to heatstroke in rats. Neurosci Lett. 1998;252(1):9–12. [PubMed: 9756346]
Currie RW, Ellison JA, White RF. et al. Benign focal ischemic preconditioning induces neuronal Hsp70 and prolonged astrogliosis with expression of Hsp27. Brain Res. 2000;863(12):169–181. [PubMed: 10773205]
Bader SB, Price BD, Mannheim-Rodman LA. et al. Inhibition of heat shock gene expression does not block the development of thermotolerance. J Cell Phys. 1992;151:56–62. [PubMed: 1560048]
Mosser DD, Caron AW, Bourget L. et al. Role of the human heat shock protein hsp70 in protection against stress-induced apoptosis. Mol Cell Biol. 1997;17(9):5317–5327. [PMC free article: PMC232382] [PubMed: 9271409]
Bellmann K, Jaattela M, Wissing D. et al. Heat shock protein hsp70 overexpression confers resistance against nitric oxide. FEBS Lett. 1996;391(12):185–188. [PubMed: 8706913]
Jaattela M, Wissing D, Kokholm K. et al. Hsp70 exerts its anti-apoptotic function downstream of caspase-3-like proteases. EMBO J. 1998;17(21):6124–6134. [PMC free article: PMC1170939] [PubMed: 9799222]
Buzzard KA, Giaccia AJ, Killender M. et al. Heat shock protein 72 modulates pathways of stress-induced apoptosis. J Biol Chem. 1998;273(27):17147–17153. [PubMed: 9642282]
Mestril R, Chi SH, Sayen MR. et al. Expression of inducible stress protein 70 in rat heart myogenic cells confers protection against simulated ischemia-induced injury. J Clin Invest. 1994;93(2):759–767. [PMC free article: PMC293923] [PubMed: 8113409]
Williams RS, Thomas JA, Fina M. et al. Human heat shock protein 70 (hsp70) protects murine cells from injury during metabolic stress. J Clin Invest. 1993;92(1):503–508. [PMC free article: PMC293638] [PubMed: 8326014]
Beaucamp N, Harding TC, Geddes BJ. et al. Overexpression of hsp70i facilitates reactivation of intracellular proteins in neurones and protects them from denaturing stress. FEBS Lett. 1998;441(2):215–219. [PubMed: 9883887]
Fink SL, Chang LK, Ho DY. et al. Defective herpes simplex virus vectors expressing the rat brain stress-inducible heat shock protein 72 protect cultured neurons from severe heat shock. J Neurochem. (1997);68(3):961–969. [PubMed: 9048741]
Uney JB, Kew JN, Staley K. et al. Transfection-mediated expression of human Hsp70i protects rat dorsal root ganglian neurones and glia from severe heat stress. FEBS Lett. 1993;334(3):313–316. [PubMed: 8243639]
Uney JB, Staley K, Tyers P. et al. Transfection with hsp70i protects rat dorsal root ganglia neurones and glia from heat stress. Gene Ther. 1994;1 Suppl 1:S65. [PubMed: 8542411]
Mailhos C, Howard MK, Latchman DS. Heat shock proteins hsp90 and hsp70 protect neuronal cells from thermal stress but not from programmed cell death. J Neurochem. 1994;63(5):1787–1795. [PubMed: 7931334]
Wagstaff MJ, Smith J, Collaco-Moraes Y. et al. Delivery of a constitutively active form of the heat shock factor using a virus vector protects neuronal cells from thermal or ischaemic stress but not from apoptosis. Eur J Neurosci. 1998;10(11):3343–3350. [PubMed: 9824447]
Sato K, Saito H, Matsuki N. HSP70 is essential to the neuroprotective effect of heat-shock. Brain Res. 1996;740(12):117–123. [PubMed: 8973805]
Plumier JC, Ross BM, Currie RW. et al. Transgenic mice expressing the human heat shock protein 70 have improved post-ischemic myocardial recovery. J Clin Invest. 1995;95(4):1854–1860. [PMC free article: PMC295725] [PubMed: 7706492]
Radford NB, Fina M, Benjamin IJ. et al. Cardioprotective effects of 70-kDa heat shock protein in transgenic mice. Proc Natl Acad Sci USA. 1996;93(6):2339–2342. [PMC free article: PMC39797] [PubMed: 8637874]
Marber MS, Mestril R, Chi SH. et al. Overexpression of the rat inducible 70-kD heat stress protein in a transgenic mouse increases the resistance of the heart to ischemic injury. J Clin Invest. 1995;95(4):1446–1456. [PMC free article: PMC295626] [PubMed: 7706448]
Wagstaff MJ, Collaco-Moraes Y, Smith J. et al. Protection of neuronal cells from apoptosis by Hsp27 delivered with a herpes simplex virus-based vector. J Biol Chem. 1999;274(8):5061–5069. [PubMed: 9988753]
Lee J, Yenari M, Sun G. et al. Differential neuroprotection from human heat shock protein 70 overexpression in in vitro and in vivo models of ischemia and ischemia-like conditions. Exp Neurol. 2001;170(1):129–139. [PubMed: 11421590]
Nishimura RN, Dwyer BE. Evidence for different mechanisms of induction of HSP70i: A comparison of cultured rat cortical neurons with astrocytes. Brain Res Mol Brain Res. 1996;36(2):227–239. [PubMed: 8965643]
Papadopoulos MC, Sun XY, Cao J. et al. Over-expression of HSP-70 protects against combined oxygen-glucose deprivation. Neuroreport. 1996;7(2):429–432. [PubMed: 8730798]
Xu L, Giffard RG. HSP70 protects murine astrocytes from glucose deprivation injury. Neuroscience Letters. 1997;224(1):9–12. [PubMed: 9132695]
Rajdev S, Hara K, Kokubo Y. et al. Mice overexpressing rat heat shock protein 70 are protected against cerebral infarction. Ann Neurol. 2000;47(6):782–791. [PubMed: 10852544]
Plumier JC, Krueger AM, Currie RW. et al. Transgenic mice expressing the human inducible Hsp70 have hippocampal neurons resistant to ischemic injury. Cell Stress Chaperones. 1997;2(3):162–167. [PMC free article: PMC312993] [PubMed: 9314603]
Yenari MA, Minami M, Sun GH. et al. Calbindin d28k overexpression protects striatal neurons from transient focal cerebral ischemia. Stroke. 2001;32(4):1028–1035. [PubMed: 11283407]
Yenari M, Dumas T, Sapolsky R. et al. Gene therapy for treatment of cerebral ischemia using defective herpes simplex viral vectors. Neurol Res. 2001;23:543–552. [PubMed: 11474812]
Yenari M, Dumas T, Sapolsky R. et al. Gene therapy for treatment of cerebral ischemia using defective herpes simplex viral vectors In: Slikker W, Trembly B, eds. Neuroprotective Agents: 5th International Conference, Ann NY Acad Sci Vol. 939 New York: New York Academy of Sciences, 2001340–357. [PubMed: 11462790]
Yenari MA, Dumas T, Sapolsky RM. et al. The role of gene therapy in acute cerebral ischemia In: Krieglstein K, S. Klumpp, eds. Pharmacology of Cerebral Ischemia Stuttgart: Medpharm Scientific Publishers, 2000457–472.
Yenari MA, Fink SL, Sun GH. et al. Gene therapy with HSP72 is neuroprotective in rat models of stroke and epilepsy. Ann Neurol. 1998;44(4):584–591. [PubMed: 9778256]
Zhang Z, Kelly S, Sun GH. et al. HSP72 gene transfer protects CA1 neurons from global cerebral ischemia by improving Bcl-2 expression. Soc Neurosci Abs. 2001;27:20.11.
Kelly S, Uney JB, McCulloch J. Adenovirus HSP70 gene transfer ameliorate damage following global ischaemia. J Cereb Blood Flow Metab. 2001;21(1):S23.
Cummings CJ, Mancini MA, Antalffy B. Chaperone suppression of aggregation and altered subcellular proteasome localization imply protein misfolding in SCA1. Nat Genet. 1998;19(2):148–154. [PubMed: 9620770]
Chai Y, Koppenhafer SL, Bonini NM. et al. Analysis of the role of heat shock protein (Hsp) molecular chaperones in polyglutamine disease. J Neurosci. 1999;19(23):10338–10347. [PubMed: 10575031]
Kobayashi Y, Kume A, Li M. et al. Chaperones Hsp70 and Hsp40 suppress aggregate formation and apoptosis in cultured neuronal cells expressing truncated androgen receptor protein with expanded polyglutamine tract. J Biol Chem. 2000;275(12):8772–8778. [PubMed: 10722721]
Stenoien DL, Cummings CJ, Adams HP. et al. Polyglutamine-expanded androgen receptors form aggregates that sequester heat shock proteins, proteasome components and SRC-1, and are suppressed by the HDJ-2 chaperone. Hum Mol Genet. 1999;8(5):731–741. [PubMed: 10196362]
Muchowski PJ, Schaffar G, Sittler A. et al. Hsp70 and hsp40 chaperones can inhibit self-assembly of polyglutamine proteins into amyloid-like fibrils. Proc Natl Acad Sci USA. 2000;97(14):7841–7846. [PMC free article: PMC16632] [PubMed: 10859365]
Cummings CJ, Sun Y, Opal P. et al. Over-expression of inducible HSP70 chaperone suppresses neuropathology and improves motor function in SCA1 mice. Hum Mol Genet. 2001;10(14):1511–1518. [PubMed: 11448943]
Hoehn B, Ringer T, Fink S. et al. Postischemic overexpression of HSP72 by gene therapy protects striatal neurons after experimental stroke. J Cereb Blood Flow Metab. 2001;21(suppl. 1):S410.
Neckers L, Schulte TW, Mimnaugh E. Geldanamycin as a potential anti-cancer agent: Its molecular target and biochemical activity. Invest New Drugs. 1999;17(4):361–373. [PubMed: 10759403]
Zou J, Guo Y, Guettouche T. et al. Repression of heat shock transcription factor HSF1 activation by HSP90 (HSP90 complex) that forms a stress-sensitive complex with HSF1. Cell. 1998;94(4):471–480. [PubMed: 9727490]
Lu A, Nee A, Parmentier S. et al. Geldanamycin, a HSP90 binding drug, induces heat shock proteins in vivo and in vitro, and protects against in vivo cerebral ischemia. J Cereb Blood Flow Metab. 2001;21(1):S231.
Xiao N, Callaway CW, Lipinski CA. et al. Geldanamycin provides posttreatment protection against glutamate- induced oxidative toxicity in a mouse hippocampal cell line. J Neurochem. 1999;72(1):95–101. [PubMed: 9886059]
Hu BR, Janelidze S, Ginsberg MD. et al. Protein aggregation after focal brain ischemia and reperfusion. J Cereb Blood Flow Metab. 2001;21(7):865–875. [PubMed: 11435799]
Hu BR, Martone ME, Jones YZ. et al. Protein aggregation after transient cerebral ischemia. J Neurosci. 2000;20(9):3191–3199. [PubMed: 10777783]
Polla BS, Kantengwa S, Francois D. et al. Mitochondria are selective targets for the protective effects of heat shock against oxidative injury. Proc Natl Acad Sci USA. 1996;93(13):6458–6463. [PMC free article: PMC39045] [PubMed: 8692837]
Kamii H, Kinouchi H, Sharp FR. et al. Prolonged expression of hsp70 mRNA following transient focal cerebral ischemia in transgenic mice overexpressing CuZn-superoxide dismutase. J Cereb Blood Flow Metab. 1994;14(3):478–486. [PubMed: 8163590]
Kondo T, Murakami K, Honkaniemi J. et al. Expression of hsp70 mRNA is induced in the brain of transgenic mice overexpressing human CuZn-superoxide dismutase following transient global cerebral ischemia. Brain Res. 1996;737(12):321–326. [PubMed: 8930384]
Kondo T, Sharp FR, Honkaniemi J. et al. DNA fragmentation and Prolonged expression of c-fos, c-jun, and hsp70 in kainic acid-induced neuronal cell death in transgenic mice overexpressing human CuZn-superoxide dismutase. J Cereb Blood Flow Metab. 1997;17(3):241–256. [PubMed: 9119897]
Feinstein DL, Galea E, Reis DJ. Suppression of glial nitric oxide synthase induction by heat shock: effects on proteolytic degradation of IkappaB-alpha. Nitric Oxide. 1997;1(2):167–176. [PubMed: 9701055]
Guzhova IV, Darieva ZA, Melo AR. et al. Major stress protein Hsp70 interacts with NF-kB regulatory complex in human T-lymphoma cells. Cell Stress Chaperones. 1997;2(2):132–139. [PMC free article: PMC312990] [PubMed: 9250404]
Heneka MT, Sharp A, Klockgether T. et al. The heat shock response inhibits NF-kappaB activation, nitric oxide synthase type 2 expression, and macrophage/microglial activation in brain. J Cereb Blood Flow Metab. 2000;20(5):800–811. [PubMed: 10826530]
Silver PA. How proteins enter the nucleus. Cell. 1991;64(3):489–497. [PubMed: 1991319]
Ashkenazi A, Dixit VM. Death receptors: Signaling and modulation. Science. 1998;281(5381):1305–1308. [PubMed: 9721089]
Green DR, Reed JC. Mitochondria and apoptosis. Science. 1998;281(5381):1309–1312. [PubMed: 9721092]
Gabai VL, Meriin AB, Mosser DD. et al. Hsp70 prevents activation of stress kinases.A novel pathway of cellular thermotolerance. J Biol Chem. 1997;272(29):18033–18037. [PubMed: 9218432]
Gabai VL, Meriin AB, Yaglom JA. et al. Role of Hsp70 in regulation of stress-kinase JNK: implications in apoptosis and aging. FEBS Lett. 1998;438(12):1–4. [PubMed: 9821948]
Saleh A, Srinivasula SM, Balkir L. et al. Negative regulation of the Apaf-1 apoptosome by Hsp70. Nat Cell Biol. 2000;2(8):476–483. [PubMed: 10934467]
Beere HM, Wolf BB, Cain K. et al. Heat-shock protein 70 inhibits apoptosis by preventing recruitment of procaspase-9 to the Apaf-1 apoptosome. Nat Cell Biol. 2000;2(8):469–475. [PubMed: 10934466]
Beere HM, Green DR. Stress management: Heat shock protein-70 and the regulation of apoptosis. Trends Cell Biol. 2001;11(1):6–10. [PubMed: 11146277]
Creagh EM, Cotter TG. Selective protection by hsp 70 against cytotoxic drug-, but not Fas- induced T-cell apoptosis. Immunology. 1999;97(1):36–44. [PMC free article: PMC2326807] [PubMed: 10447712]
States BA, Honkaniemi J, Weinstein PR. et al. DNA fragmentation and HSP70 protein induction in hippocampus and cortex occurs in separate neurons following permanent middle cerebral artery occlusions. J Cereb Blood Flow Metab. 1996;16(6):1165–1175. [PubMed: 8898689]
Panahian N, Yoshiura M, Maines MD. Overexpression of heme oxygenase-1 is neuroprotective in a model of permanent middle cerebral artery occlusion in transgenic mice. J Neurochem. 1999;72(3):1187–1203. [PubMed: 10037492]
Yu Z, Luo H, Fu W. et al. The endoplasmic reticulum stress-responsive protein GRP78 protects neurons against excitotoxicity and apoptosis: suppression of oxidative stress and stabilization of calcium homeostasis. Exp Neurol. 1999;155(2):302–314. [PubMed: 10072306]
Creagh EM, Carmody RJ, Cotter TG. Heat shock protein 70 inhibits caspase-dependent and -independent apoptosis in Jurkat T cells. Exp Cell Res. 2000;257(1):58–66. [PubMed: 10854054]
Wyatt S, Mailhos C, Latchman DS. Trigeminal ganglion neurons are protected by the heat shock proteins hsp70 and hsp90 from thermal stress but not from programmed cell death following nerve growth factor withdrawal. Brain Res Mol Brain Res. 1996;39(12):52–56. [PubMed: 8804713]
Li GC, Li L, Liu RY. et al. Heat shock protein hsp70 protects cells from thermal stress even after deletion of its ATP-binding domain. Proc Natl Acad Sci USA. 1992;89(6):2036–2040. [PMC free article: PMC48591] [PubMed: 1549562]