Focal cerebral ischaemia increases the levels of several classes of heat shock proteins and their corresponding mRNAs

Abstract

The induction of focal cerebral ischaemia in rats by middle cerebral artery occlusion has previously been shown to increase, over time, the mRNA levels of the heat shock proteins (HSPs) 27 and 70. However, the levels of HSP90 mRNA remain constant. In contrast, during global ischaemia, HSP70 and HSP90 mRNA levels are both raised, particularly in the CA1 neurons in the hippocampus, an area that is resistant to the insult in comparison to the surrounding regions. HSP27 mRNA is raised in the neuroglia in the subregions of the hippocampus. However, the protein levels of HSP27, 70 and 90 have not been characterised in focal ischaemia. With this data in mind, we have carried out a comparative study of HSP27, 56, 60, 70 and 90 mRNA and protein levels during focal cerebral ischaemia in rats, up to 24 h post-occlusion. We have shown that HSP70 and HSP27 mRNA levels are increased and also that HSP60 mRNA levels (which had also not previously been characterised in this model of focal ischaemia) are significantly raised. HSP90 and HSP56 mRNAs were not significantly elevated. On Western blot analysis, the inducible HSP72 protein was first detected at 8 h post-occlusion, HSP27 protein was detected only at 24 h post-occlusion and HSP60 protein, although constitutive, appeared to increase at 24 h post-occlusion. HSP56 protein levels appeared to rise on the occluded side, but HSP90 protein levels remained constant.

Introduction

Cerebral ischaemia has been shown to alter gene expression in the mammalian brain in the infarcted tissue and in the surrounding cortical regions [23,24]. Sub-lethal stimuli such as heat shock and ischaemia prior to a lethal insult have also been demonstrated to confer protection against subsequent neuronal damage [25,26,40]. Many of the immediate early genes that are induced following an is-chaemic insult have been identified as members of the heat shock family of proteins (HSPs), in particular the HSPs 27, 47, 70 and 90, whose mRNA levels have been characterised temporally [16,22,24]. Both HSP27 and 70 mRNA levels were shown to rise over time both during global and focal ischaemia, whereas HSP90 mRNA levels were only raised during global ischaemia. HSP mRNA transcription and protein synthesis in neurons has also occurred in response to a number of other inducing agents, for example surgical incision or chemical-induced excitotoxicity leads to raised HSP70 levels [7,34].

It has been speculated that the role of these proteins during ischaemia is to protect against neuronal damage and cell death. Transfection of HSP70 cDNA into rat dorsal root ganglion neurons (DRGs) has been shown to protect these cells from severe heat stress [49], whereas overex-pressing HSP90 is not protective [3]. The most extensively characterized response to cerebral ischaemia is the induction of HSP72 mRNA, the inducible member of the HSP70 family, which is induced by kainic acid injury as well as ischaemia [14,38]. The regional localisation and temporal synthesis of HSP70 mRNA in the CNS during ischaemia differs according to the severity of the insult [45,50]. Furthermore, the CA1 pyramidal hippocampal neurons, which express particularly high levels of HSP70 mRNA after as little as two min of temporary global cerebral ischaemia, maintain a higher cell density on a subsequent ‘lethal’ ischaemic insult than the surrounding regions which sustain greater damage [25,26]. Indeed, heat shock treatment which increases expression of HSP70 mRNA in vitro in cultured neurons, has been demonstrated to induce a protective phenomenon which renders those cells which respond with high HSP70 mRNA expression more resistant to glutamate toxicity [40]. Greater levels of HSP70 protein are seen in neurons in vitro by pre-treatment with heat than with ischaemia, and the former gives greater protection than the latter irrespective of the type of subsequent lethal insult. Therefore it has been concluded that the degree of protection conferred to neuronal cells in vitro by pre-conditioning stimuli is correlated to the amount of HSP70 protein that is produced as a result, rather than the nature of the subsequent stress [2]. The raised levels of HSP70 and other immediate–early gene products through heat shock or ischaemic pre-treatment may therefore be responsible for the subsequent tolerance to ischaemia.

Although most studies in this area have concentrated on HSP70, other HSPs have been investigated to a lesser extent. HSP27, an αB-crystallin homologue, also induces thermotolerance [28] and has subsequently been shown to be raised in focal cerebral ischaemia at the mRNA level at least [16]. The β-isoform of HSP90 is also raised in global cerebral ischaemia [22], although its elevated expression has not yet been shown to produce the ‘ischaemic tolerance’ associated with HSP70. The protein levels of both of these have not been characterised in focal cerebral is-chaemia.

HSP56 is an FK506 binding protein and FK506, a synthetic immunosuppressant, has been shown to give neuroprotection during focal cerebral ischaemia [44]. HSP56 is also linked functionally with HSP90 and 27 as a part of the untransformed steroid receptor complex, which is formed under the influence of HSP70. However, HSP56 levels have not as yet been characterised during cerebral ischaemia. HSP60 is an auto-immunogen and chaperonin and has not been measured during focal ischaemia although its mRNA levels have been shown to be raised in the hippocampus after 3 h of reperfusion subsequent to a global ischaemic insult of 3.5 min [1].

The encouraging results from studies on HSP70 might suggest that other HSPs may also be protective. We believe it is necessary to understand the dynamics of the brain's physiological response to ischaemia in detail before the therapeutic possibilities of manipulating this response can be effectively assessed. With this in mind, the present study was carried out in order to examine the temporal expression of the mRNA and protein levels of HSP27, 56, 60, 70 and 90 in a rat model of focal cerebral ischaemia, produced by occlusion of the left middle cerebral artery with an intra-luminal suture. Not only have we examined the previously undocumented levels of HSP56 and HSP60, but in measuring the full range of the heat shock proteins together under similar laboratory conditions, comparisons of the mRNA and protein levels during the timecourse can be made. It is important to measure protein levels in parallel with mRNA as mRNA levels may not parallel protein levels after an ischaemic insult. Levels of mRNA were examined in the core ischaemic area by Northern blot analysis and further quantitated by slot blot analysis, with protein levels being examined by immunodetection on Western blots.