Overexpression of heat shock protein 27 reduces ... - Semantic Scholar

Journal of Cerebral Blood Flow & Metabolism (2010) 30, 849–856 & 2010 ISCBFM All rights reserved 0271-678X/10 $32.00 www.jcbfm.com

Overexpression of heat shock protein 27 reduces cortical damage after cerebral ischemia Louise van der Weerd1,2, Mohammed Tariq Akbar3, Romina Aron Badin1,2, Lauren M Valentim2, David L Thomas4, Dominic J Wells3, David S Latchman2, David G Gadian1, Mark F Lythgoe5,6 and Jackie S de Belleroche3,6 1

RCS Unit of Biophysics, Institute of Child Health, University College London, London, UK; Medical Molecular Biology Unit, Institute of Child Health, University College London, London, UK; 3 Division of Neuroscience and Mental Health, Department of Cellular and Molecular Neuroscience, Faculty of Medicine, Hammersmith Hospital, Imperial College London, London, UK; 4The Advanced Magnetic Resonance Imaging Group, Department of Medical Physics and Bioengineering, University College London, London, UK; 5Centre for Advanced Biomedical Imaging, Department of Medicine, University College London, London, UK 2

Heat shock protein 27 (HSP27) has a major role in mediating survival responses to a range of central nervous system insults, functioning as a protein chaperone, an antioxidant, and through inhibition of cell death pathways. We have used transgenic mice overexpressing HSP27 (HSP27tg) to examine the role of HSP27 in cerebral ischemia, using model of permanent middle cerebral artery occlusion (MCAO). Infarct size was evaluated using multislice T2-weighted anatomical magnetic resonance imaging (MRI) after 24 h. A significant reduction of 30% in infarct size was detected in HSP27tg animals compared with wild-type (WT) littermates. To gain some insight into the mechanisms contributing to cell death and its attenuation by HSP27, we monitored the effect of induction of c-jun and ATF3 on tissue survival in MCAO and their effects on the expression of endogenous mouse HSP25 and HSP70. It is important that, the c-jun induction seen at 4 h tended to be localized to regions that were salvageable in HSP27tg mice but became infarcted in WT animals. Our results provide support for the powerful neuroprotective effects of HSP27 in cerebral ischemia. Journal of Cerebral Blood Flow & Metabolism (2010) 30, 849–856; doi:10.1038/jcbfm.2009.249; published online 9 December 2009 Keywords: focal ischemia; heat shock protein 27; MRI; permanent middle cerebral artery occlusion

Introduction Heat shock proteins (HSPs) form a heterogeneous family of proteins that exhibit a diverse range of protein chaperoning properties facilitating protein folding, trafficking, and removal of aberrant proteins. In addition, certain HSPs are induced during cellular

Correspondence: Dr JS de Belleroche, Faculty of Medicine, Neurogenetics Group, Division of Neuroscience and Mental Health, Department of Cellular and Molecular Neuroscience, Imperial College London, Hammersmith Hospital, London W12 0NN, UK. E-mail: [email protected] and Dr MF Lythgoe, Centre for Advanced Biomedical Imaging, University College London, London WCIE 6DD, UK. E-mail: [email protected] 6 These corresponding authors contributed equally to this work. We are grateful for support from the British Heart Foundation, the Wellcome Trust, and Biotechnology Biological Sciences Research Council. Received 10 August 2009; revised 24 September 2009; accepted 10 November 2009; published online 9 December 2009

stress and contribute to cell survival through antioxidant and antiapoptotic properties. These properties have been particularly well established in cell culture and using in vivo models for two inducible HSPs, HSP70 and, a small 27-kDa HSP, HSP27. In the mouse, this HSP is smaller and known as HSP25. These HSPs show distinct patterns of cellular activation in vivo, induction of HSP70 occurring mainly in neuronal populations, whereas HSP27 induction is most prevalent in astrocytes (Akbar et al, 2001). A number of cytoprotective effects have been shown using HSP27 transgenic lines in a range of in vivo disease models, including cardiac ischemia (Efthymiou et al, 2004; Hollander et al, 2004), kainate-induced hippocampal cell death (Akbar et al, 2003), nerve injury (Sharp et al, 2006), and the SOD1G93A model of amyotrophic lateral sclerosis (Sharp et al, 2008). Furthermore, in vivo viral delivery of HSP27 before and after middle cerebral artery occlusion (MCAO) has been shown to be protective (Badin et al, 2006; Badin et al, 2009).

HSP27 reduces infarct size in cerebral ischemia L van der Weerd et al 850

Similarly, using a model of transient cerebral ischemia, HSP27 transgenic mice have been shown to develop a reduced infarct size (Stetler et al 2008). The well-documented antiapoptotic properties of HSP27 exhibited in cell culture (Garrido et al, 1999; Bruey et al, 2000; Paul et al, 2002) have also been shown in vivo in the kainate seizure model, in which HSP27 attenuates caspase 3 induction and reduces apoptotic cell death (Akbar et al, 2003). The formation of HSP27 oligomers necessary for reducing oxidative damage in cellular models has also been shown in cardiac ischemia (Hollander et al, 2004). More recently, the upstream mechanisms of HSP27 action have been elucidated in cerebral ischemia models, in which the importance of the MKK4/c-Jun N-terminal kinase (JNK) kinase cascade in the progression to cell death has been shown by Stetler et al (2008), using HSP27 transgenic mice and cell culture models. They showed that HSP27 overexpression specifically leads to the suppression of the MKK4/JNK cascade and in this way HSP27 is able to reduce cell death. In this study, we have used transgenic mice overexpressing HSP27 (HSP27tg) to examine the role of HSP27 in a permanent MCAO (pMCAO) model of cerebral ischemia, compared with the transient ischemia model used by Stetler et al, (2008). We assessed the severity of injury using magnetic resonance imaging (MRI) to determine the overall lesion volume, estimated after 24 h of ischemia. Our results provide support for the powerful neuroprotective effects of HSP27 in cerebral ischemia in this model and further provide evidence that the presence of HSP27 enhances cell survival in regions that undergo cell death in wild-type (WT) littermates. This difference occurs despite marked induction of c-jun and ATF3 early after MCAO and shows a profound inhibitory effect of HSP27 on the c-junmediated cell death cascade.

Materials and methods Animals All animal care and procedures were in accordance with the UK Animals (Scientific Procedures) Act 1986. Transgenic mice overexpressing human HSP27 (HSP27o/e) (under the control of a chicken b-actin promoter and a cytomegalovirus enhancer element) on a C57BL10/CBA background were used. The line was maintained by continuously backcrossing with C57BL10/CBA mice (Akbar et al, 2003).

Middle Cerebral Artery Occlusion Model Adult male (weighing 24 to 28 g; 10 to 13 weeks old) HSP27tg mice (n = 6) and WT mice (n = 5) were subjected to pMCAO using an intraluminal suture method based on the model of Koizumi et al (1986). Briefly, mice were anesthetized with 2.5% isoflurane and maintained on 1.75% isoflurane with pure oxygen. The rectal temperature Journal of Cerebral Blood Flow & Metabolism (2010) 30, 849–856

was maintained at 36.81C±0.31C throughout the procedure. A ventral neck incision was made to expose the right carotid artery. A 6-0 monofilament suture (Monocryl; Ethicon, Edinburgh, UK) with a 180 to 200 mm rounded tip was advanced from the common carotid artery into the internal carotid artery past the MCA junction (B9 mm). The incision was sutured, after which the animals were immediately transferred to the MRI scanner to confirm the occlusion of the middle cerebral artery using the cerebral blood flow (CBF) measurements. After all scans were completed, the animals were allowed to recover and were transferred to the animal house, with free access to food and water. At 24 h, the mice were re-anesthetized and scanned again.

Magnetic Resonance Imaging Coronal images were obtained with the central slice B0.5 mm from bregma at 2 and 24 h after the onset of stroke. Magnetic resonance imaging measurements were performed on a 2.35-T horizontal bore magnet (Oxford Instruments, Oxford, UK) interfaced to a SMIS console (Surrey Medical Imaging Systems, Guildford, UK). The radio frequency pulses were transmitted with a volume coil of 6.5 cm length. The signal was received with a separate 1-cm-diameter surface coil. After MCAO, the animals were transferred to the magnet and imaged using the following protocol: 2 h: Single slice CBF (continuous arterial spin labeling): inversion plane positioned in the neck (1.5 mm posterior to the cerebellum); spin labeling time of 3 secs; post-labeling delay 500 ms; recovery delay 1,000 ms; TE (echo time) 36 ms; and 44 averages (Alsop and Detre, 1996). Field of view of 32 16 mm, 2 mm slice thickness, and 128 64 pixels. 24 h: CBF, T1 as before. Multislice T2-weighted spin-echo; field of view of 20 mm; slice thickness 1 mm; nine slices; TR (repetition time) 1,500 ms; TE 120 ms; and eight averages.

Image Processing Perfusion maps were obtained by subtracting the labeled image from the control image. The T1 data were used to calculate the CBF based on the equation in Alsop and Detre (1996). For the calculation of CBF, the following values were used: T1a = 1.5 secs, l = 0.9, efficiency of the spin labeling pulse = 0.71, and tissue transit time = 1 s. The lesion area was determined manually by an observer blinded to the experiment, using the multislice anatomical T2-weighted scan, in the SMIS image analysis program. The relative infarcted area per slice was defined as the ratio of the lesion area to the size of the contralateral hemisphere. An example of T2-weighted images at 24 h after MCAO is shown in Supplementary Figure 1. Cerebral blood flow images were analyzed on a region-ofinterest basis. Ten regions of interest were placed in the medial cortex, the dorsolateral cortex, the ventrolateral cortex, the lateral caudate putamen, and the medial caudate putamen of each hemisphere. The CBF of the affected hemisphere was expressed relative to the contralateral side.

HSP27 reduces infarct size in cerebral ischemia L van der Weerd et al

In Situ Hybridization For in situ hybridization studies, male HSP27tg and WT mice were killed at 4 h (WT, n = 3; HSP27tg, n = 4) or 24 h (WT, n = 4; Tg, n = 4) after pMCAO. Sham-operated WT animals (n = 3) were subjected to the same pMCAO procedure, but without advancing the suture past the MCA junction. Brains were rapidly removed and frozen on dry-ice. Sections were cut (10 mm) on a cryostat (Bright Instruments, Huntingdon, Cambridge, UK), thaw mounted onto Superfrost slides (VWR, Lutterworth, UK) and stored at 801C until use. Specific antisense oligonucleotides were synthesized (Sigma-Genosys, Cambridge, UK) for use in the in situ hybridization studies. A 33-mer oligonucleotide complimentary to nucleotides 593 to 625 of the human HSP27 gene (Carper et al, 1990) was used for the detection of human HSP27 mRNA. A 34-mer oligonucleotide complimentary to nucleotides 358 to 391 of the mouse HSP25 sequence (Frohli et al, 1993) was used for mouse HSP25 mRNA. A 45-mer oligonucleotide complementary to nucleotides 1,119 to 1,163 of the mouse GFAP gene (Lewis et al, 1984) was used for mouse glial fibrillary acidic protein (GFAP) mRNA. A 39mer oligonucleotide complimentary to nucleotides 1,872 to 1,910 of c-jun gene (Ryder and Nathans 1988) was used for cjun mRNA. A 45-mer oligonucleotide complimentary to nucleotides 987 to 1,031 of the mouse AFT3 gene (Drysdale et al, 1996) was used for mouse ATF3 mRNA. A 30-mer oligonucleotide complementary to nucleotides 539 to 578 was used for inducible HSP70 mRNA (Longo et al, 1993). All oligonucleotide probes have been previously characterized (Akbar et al, 2001, 2003, Tsujino et al, 2000, Kolko et al, 2002). Oligonucleotide probes were 30 -end-labeled using terminal deoxynucleotidyl transferase (Promega, Southampton, UK) and [35S]dATP (GE Healthcare, Little Chalfont, Buckinghamshire, UK). In situ hybridization was carried out as previously described (Akbar et al, 2001). In brief, slides were fixed in 4% paraformaldehyde in phosphate-buffered saline (41C) for 10 min, rinsed in phosphate-buffered saline for 2 5 min and then treated with 0.25% acetic anhydride in 0.1 mol/L triethanolamine/0.9% NaCl for 10 min. Following dehydration in increasing concentrations of ethanol, the sections were delipidated in chloroform for 5 min, rinsed in ethanol, and allowed to air-dry. Sections were then hybridized overnight at 421C with 1 to 2 106 c.p.m. of labeled probe in 100 ml of hybridization buffer (5 Denhardt’s solution (0.02% bovine serum albumin, 0.02% Ficoll, and 0.02% polyvinylpyrrolidone), 4 standard saline citrate (SSC), 50% formamide, 10% dextran sulfate, 200 mg/mL polyadenylic acid, 200 mg/mL sheared singlestranded salmon sperm DNA, 50 mmol/L phosphate buffer, and 20 mmol/L dithiothreitol). The next day, the sections were stringently washed (1 SSC at room temperature for 30 min, 1 SSC at 551C for 30 min thrice, and 0.1 SSC at room temperature for 5 min), dehydrated in an ascending series of ethanol and air-dried. When dry, the sections were exposed to X-ray film (BioMax MR, Eastman Kodak, Rochester, NY, USA) for 15 to 23 days and developed. Sections were then dipped in photographic emulsion (Ilford K5) and exposed for 16 to 20 days at 41C,

851

developed, and counter-stained with toluidine blue for light microscopic analyses. All hybridization experiments were carried out on near adjacent sections from the same animals to allow comparisons to be made.

Statistics The lesion volume per slice was compared using a degree of freedom adjusted repeated-measures analysis of variance (ANOVA). For each analysis, the appropriate within-subject variables and between-subject factors were adopted. Differences were considered significant when the P-value was < 0.05. The HSP27 expression levels, as determined by western blots, were analyzed using standard t-tests. Wherever a multicomparison correction was appropriate, a Bonferroni correction of the significance level was performed, as stated in the text. All data are presented as mean±s.d.

Results Lesion volume

The HSP27 transgenic mouse line used in this study has been characterized extensively in earlier studies (Akbar et al, 2003) and shows no effect of the genetic modification on development or survival. The transgene is widely expressed across multiple tissues and is particularly abundant in the cerebral cortex, the hippocampus, and the striatum (Figure 1). To determine whether HSP27 overexpression confers

Figure 1 Representative in situ hybridization film autoradiographs of heat shock protein 27 (HSP27) transgene expression in horizontal (A and C) and coronal (B and D) sections from wildtype (WT) and transgenic mice overexpressing HSP27 (HSP27tg) mice. Scale bar = 2 mm. Journal of Cerebral Blood Flow & Metabolism (2010) 30, 849–856


Figure 2 Effect of heat shock protein 27 (HSP27) overexpression on the lesion area after permanent middle cerebral artery occlusion. Relative lesion area for transgenic mice overexpressing HSP27 (HSP27tg) and control mice. The lesion area was calculated relative to the contralateral hemispherical area (horizontal bar indicates the group mean). Mixed model analysis of variance (ANOVA) showed a significant difference in lesion area between the HSP27tg mice and the WT mice.

protection against ischemic injury, HSP27tg (n = 6) and WT (n = 5) mice were subjected to permanent MCA occlusion. The lesion size was evaluated using a multislice T2-weighted anatomical MRI scan after 24 h (Figure 2; Supplementary Figure 1). This type of occlusion induces an ischemic region in the ipsilateral hemisphere, including the basal ganglia and a large part of the cerebral cortex. A blinded observer quantified the lesion size in all slices. At 24 h after permanent MCA occlusion, the mean lesion area per slice was reduced for the HSP27 mice. Statistical analysis using a mixed model ANOVA indicated that this was a significant difference between HSP27tg mice and WT mice (P < 0.05). No interaction was found between the animal groups and the different slices. Overall, the lesion volume comprised 30±3.2% of the whole brain for WT mice compared with 20±7.0% for HSP27tg mice, a 33% decrease in lesion volume. Successful occlusion of the MCA results in a severe reduction of CBF in a large part of the ipsilateral hemisphere. However, there were no significant differences in CBF reduction within the lesion area between WT and HSP27tg at 2 or 24 h after stroke (Supplementary Figure 2) as previously found with HSP70tg mice (van der Weerd et al, 2005).

Glial Fibrillary Acidic Protein expression

The spatial and temporal expression profile of three sensitive markers of tissue injury, GFAP, c-jun, ATF3, and two endogenous HSPs, mHSP25 and HSP70, in WT and HSP27tg mice after pMCAO was assessed using in situ hybridization. The degree of cellular protection provided by HSP27 overexpression seen in the MRI analysis was also evident in the expression studies where the extent of viable tissue was greatly enhanced ipsilateral to MCAO in HSP27tg mice. This is clearly seen in the expression of GFAP, Journal of Cerebral Blood Flow & Metabolism (2010) 30, 849–856

a glial marker induced in response to tissue injury, present in the region surrounding the infarcted tissue. Figure 3 shows representative animals of each genotype at each time point. The large infarct present in WT animals at 24 h extended throughout the MCA territory affecting the ipsilateral cortical and striatal regions (Figures 3C, 3G, and 3K). In contrast, in HSP27tg mice, although the striatum was severely affected, the majority of the cerebral cortex in the MCA was preserved (Figures 3D, 3H, and 3L). Increased GFAP expression was seen in tissue adjacent to the infarct at 24 h in WT mice, being localized to regions largely outside the MCA territory. In HSP27tg mice, GFAP was present in rescued tissue ipsilateral to MCAO and little upregulation was detected outside the MCA territory.

c-jun Expression

In sham-operated control mice, there were low-tomoderate levels of constitutive expression of c-jun mRNA throughout the cortex. In contrast, at 4 h after MCAO, a profound induction of c-jun was observed in most animals in both WT and HSP27tg mice (Figures 4A, 4B, and 4F). c-jun mRNA expression in these animals was evident throughout the ischemic hemisphere, the outer and deep cortical layers showing the highest levels compared with the intermediate layers. Variable expression was seen in the caudate and putamen in the ischemic hemisphere. However, the most intense level of c-jun labeling was seen in the cortical and striatal regions of HSP27tg mice. These regions coincided with the areas of tissue preservation in HSP27tg mice at 24 h (Figures 4B and 4F). This contrasts with the extensive infarcts that develop in the same regions in control animals. Intense labeling was also seen in the pyriform cortex of all mice in the ipsilateral cortex at this time point with moderate expression in the pyriform cortex of the contralateral side (Figures 4A, 4B, 4E, 4F, and 4J). There was a low level expression of c-jun mRNA in the contralateral cortex in both WT and HSP27tg mice. At 24 h, the expression levels of c-jun mRNA were generally reduced. In WT mice, there was moderate expression throughout the ventral, lateral, and dorsal cortex in the contralateral hemisphere, whereas strong labeling in the pyriform cortex was maintained. In the ipsilateral hemisphere, c-jun mRNA was confined to the penumbral tissue with no signal detected in the infarct area. In HSP27tg mice at 24 h after MCAO, expression of c-jun mRNA in the contralateral hemisphere was similar to that described for WT mice. However, expression in the ipsilateral hemisphere of HSP27tg mice showed a rather different pattern with intense to moderate labeling seen in the larger area of surviving cortical and striatal tissue (penumbral tissue/peri-infarct area) and more intense labeling in the pyriform cortex compared with WT mice (Figures 4D, 4H, and 4L).


Figure 3 The effect of middle cerebral artery occlusion (MCAO) on glial fibrillary acidic protein (GFAP) expression. c-jun mRNA detected by in situ hybridization in wild-type (WT) (A, C, E, G, and K) and transgenic mice overexpressing HSP27 (HSP27tg) (B, D, F, H, J, and L) mice at 4 h (A, B, E, F, and J) and 24 h (C, D, G, H, K, and L) at the level of the caudate putamen after MCAO. Representative animals of each genotype are shown at each time point. Scale bar = 3 mm.

Figure 4 The effect of middle cerebral artery occlusion (MCAO) on c-jun expression. c-jun mRNA detected by in situ hybridization in wild-type (WT) (A, C, E, G, and K) and transgenic mice overexpressing HSP27 (HSP27tg) (B, D, F, H, J, and L) mice at 4 h (A, B, E, F, and J) and 24 h (C, D, G, H, K, and L) at the level of the caudate putamen after MCAO. Representative animals of each genotype are shown at each time point. Scale bar = 3 mm. Journal of Cerebral Blood Flow & Metabolism (2010) 30, 849–856


Figure 5 The effect of middle cerebral artery occlusion (MCAO) on expression ATF3 compared with c-jun and HSP70 mRNA. In situ hybridization of c-jun, ATF3, and HSP70 mRNA in coronal sections from wild-type (WT) mice (A, B and C, respectively) and transgenic mice overexpressing HSP27 (HSP27tg) mice (D, E, and F, respectively) at 4 h after permanent MCAO. (Note the marked induction of ATF3 mRNA in the caudate putamen of WT mice compared with the moderate expression in Tg mice.) Scale bar = 3 mm.

ATF3 expression

The expression of ATF3 mRNA in forebrain sections from control nonoperated/naive or sham-operated animals showed no significant hybridization signal (data not shown). A similar pattern of ATF3 mRNA expression was seen throughout the nonischemic (contralateral) hemisphere of both WT and HSP27tg mice killed 4 h after pMCAO (Figures 5B and 5E). In contrast, a marked induction of ATF3 mRNA was seen in the caudate/putamen in the ischemic (ipsilateral) hemisphere in sections from WT and HSP27tg mice (Figures 5B and 5E). Moderate labeling was seen in the overlying cortex, predominantly in the lateral and ventral cortex. These regions define the core of the infarct area. A strong signal was also apparent in the pyriform cortex. It is noteworthy that the animals with the highest levels of c-jun expression were also accompanied by the highest levels of ATF3 and HSP70 expression (Figures 5A, 5B, and 5C and , respectively). At 24 h after pMCAO, ATF3 mRNA expression levels had declined to background levels in both WT and HSP27tg mice (data not shown).

Expression of HSP70 and HSP25

There was no signal detected for HSP70 mRNA in sections from sham-operated mice (data not shown). HSP70 mRNA expression generally showed a similar labeling pattern to that observed for c-jun. At 4 h after MCAO onset, there was robust expression of HSP70 mRNA in the ipsilateral hemisphere in both WT and HSP27tg mice (Supplementary Figures 3A, 3B, 3E, 3F, and 3J). Intense labeling was evident in the pyriform cortex, in the outer and inner layers of the cortex, with moderate signal detected in the intermediate cortical layers. Robust expression was also evident in the striatum. The distribution of HSP70 Journal of Cerebral Blood Flow & Metabolism (2010) 30, 849–856

induction at 4 h showed considerable overlap with the expression of c-jun mRNA in near adjacent sections from the same animals (Figure 5). At 24 h after pMCAO, the expression of HSP70 mRNA was much reduced. In WT animals, HSP70 mRNA was present in midline brain structures adjacent of the infarct region, although expression was granular/patchy in appearance (Supplementary Figures 3C, 3G, and 3K). In HSP27tg mice at this time point, strong expression of HSP70 mRNA was present in the outer and deep layers of the cortex in the ipsilateral hemisphere (Supplementary Figures 3D, 3H, and 3L). There was no hybridization signal detected in the contralateral hemisphere in both WT and HSP27tg mice at both time points after induction of focal cerebral ischemia. There was moderate diffuse expression of endogenous inducible mHSP25 mRNA in the striatum and cortex of WT mice at 4 h after ischemia, which was reduced in HSP27tg mice (Supplementary Figure 4). At 24 h, there was a greater intensity of expression in midline peri-infarct areas and in the dorsal cortex in WT mice. No signal was detected in the infarct region. In HSP27tg mice at 24 h, there was robust expression of mHSP25 mRNA in the penumbral tissue with clear labeling in the deep and superficial layers of the cortex and low to moderate labeling in the intermediate cortical layers. There was no mHSP25 mRNA detected in the contralateral hemisphere. Overall, the expression of mHSP25 mRNA at 24 h in cells adjacent to infarcted tissue highlighted the extent of salvaged tissue in the HSP27tg animals.

Discussion The MRI results described in this study clearly show that HSP27 over-expression reduces infarct size.


This is also substantially supported by expression studies. The large infarct seen in WT animals affects the whole MCA territory in ipsilateral cortical and striatal regions. In contrast, in HSP27tg mice, although the striatum is severely affected, the majority of the cerebral cortex in the MCA is preserved at 24 h after pMCAO. Increased GFAP expression is seen in tissue adjacent to the infarct at 24 h in WT mice, being localized to regions largely outside the MCA territory. In HSP27tg mice, GFAP is present in rescued tissue ipsilateral to MCAO and little upregulation was detected outside the MCA territory. To gain some insight into the mechanisms contributing to cell death and its attenuation by HSP27, we analyzed the expression of c-jun, ATF3, and HSP70. c-jun is upregulated within minutes of a cellular stress (Honkaniemi et al, 1997) and in the form of homodimers or heterdimers regulates transcription through the AP-1 binding site. Induction of c-jun contributes both to cell death through the JNK cascade and also survival responses to central nervous system insults mediated through the Akt pathway. In neutrophils, HSP27 also provides an essential role in regulating the antiapoptotic effects of Akt (Rane et al, 2003). Widespread expression of cjun occurs after focal cerebral ischemia (Ohba et al, 2003, Honkaniemi et al, 1997). ATF3 is also induced after cerebral ischemia and nerve injury, and has been shown to colocalize in surviving neurons (Ohba et al, 2003, Takeda et al, 2000, Tsujino et al, 2000). ATF3 is known to protect vulnerable CA3 neurons from kainate toxicity (Francis et al, 2004), and furthermore, ATF3 has been shown to prevent JNKinduced neuronal death in PC12 cells by promoting HSP27 expression and Akt activation (Nakagomi et al, 2003). In the pMCAO model, the ischemic insult caused a rapid induction of c-jun and ATF3, which was mainly localized to the cortical and striatal regions served by the MCA at 4 h. At this stage, severe loss of cellularity was evident across both regions in WT animals, but was mainly restricted to striatal regions in HSP27tg mice. Importantly, the c-jun induction seen at 4 h tended to be localized to regions that were salvageable. The 4-h time point coincides with high/maximal levels of JNK and c-jun phosphorylation occurring in MCA (Stetler et al, 2008). This is particularly marked in the HSP27 animals where cortical regions served by the MCA showed profound induction; at the later time point of 24 h, these regions showed greater cellularity. Interestingly, the distribution of HSP70 induction at 4 h showed considerable overlap overall and within animals with the expression of c-jun mRNA indicating a highly coordinated activation of the two pathways of transcriptional activation, mediated through heat shock factor-1 (HSF-1) in the case of HSPs and MAP kinases and their targets, such as cyclic AMP-responsive element binding protein (CREB), in the case of c-jun. The induction of ATF3 was of much smaller magnitude than that of c-jun and was not detectable

at 24 h. The distribution of c-jun mRNA was similar in WT and HSP27 tg mice but was absent in areas that had already undergone infarction, which was more extensive in WT animals. By 24 h, c-jun induction ipsilateral to MCAO was restricted to a midline region, which is not served by the MCA and coincides with the zone of GFAP induction. Interestingly, c-jun was induced throughout the contralateral hemisphere. In contrast, in HSP27 tg mice, c-jun was mainly present at low levels in the ipsilateral cortex. c-jun and other immediate early genes are known to show a widespread upregulation, which is attributed to spreading depression caused by glutamate release at the site of injury (Kiessling and Gass 1994). The induction of HSP25 showed a marked difference between control and transgenic animals, being particularly evident at 24 h and mainly present in peri-infarct areas, which in control animals are seen outside the MCA territory in midline regions and in transgenic animals lie within the MCA territory associated with the much attenuated infarct. HSP27 is thought to be regulated by c-jun/ATF3 heterodimers, which have been shown to increase HSP25/ 27 transcription by binding to an atypical CRE motif in the promoter region of HSP27 (Nakagomi et al, 2003). This mechanism may be responsible for the induction of mHSP25 in striatal regions of WT animals at 4 h, at which time both c-jun and ATF3 are elevated. However, as ATF3 induction is transient, this regulatory mechanism is unlikely to explain the delayed more widespread increase in mHSP25 mRNA. This paper shows that HSP27 mediates substantial neuroprotection in a model of pMCAO. In HSP27tg animals, infarct development is attenuated even in tissue producing abundant levels of c-jun, the substrate for JNK, which is critical for mediating neuronal cell death in cerebral ischemia (Gao et al, 2005). Furthermore, this effect is consistent with the ability of HSP27 to suppress JNK/MKK4 activity and subsequent cell death (Stetler et al, 2008). In summary, the neuroprotective properties of HSP27 shown in this and other studies have important therapeutic implications for the treatment of stroke.

References Akbar MT, Wells DJ, Latchman DS, de Belleroche J (2001) Heat shock protein 27 shows a distinctive widespread spatial and temporal pattern of induction in CNS glial and neuronal cells compared to heat shock protein 70 and caspase 3 following kainate administration. Brain Res Mol Brain Res 93:148–63 Akbar MT, Lundberg AM, Liu K, Vidyadaran S, Wells KE, Dolatshad H, Wynn S, Wells DJ, Latchman DS, de Belleroche J (2003) The neuroprotective effects of heat shock protein 27 over-expression in transgenic animals against kainate-induced seizures and hippocampal cell death. J Biol Chem 278:19956–65 Alsop DC, Detre JA (1996) Reduce transit-time sensitivity in non-invasive magnetic resonance imaging of Journal of Cerebral Blood Flow & Metabolism (2010) 30, 849–856


human cerebral blood flow. J Cereb Blood Flow Metab 16:1236–49 Badin RA, Lythgoe MF, van der Weerd L, Thomas DL, Gadian DG, Latchman DS (2006) Neuroprotective effects of virally delivered HSPs in experimental stroke. J Cereb Blood Flow Metab 26:371–81 Badin RA, Modo M, Cheetham M, Thomas DL, Gadian DG, Latchman DS, Lythgoe MF (2009) Protective effect of post-ischaemic viral delivery of heat shock proteins in vivo. J Cereb Blood Flow and Metab 29:254–63 Bruey JM, Ducasse C, Bonniaud P, Ravagnan L, Susin SA, Diaz-Latoud C, Gurbuxani S, Arrigo AP, Kroemer G, Solary E, Garrido C (2000) Hsp27 negatively regulates cell death by interacting with cytochrome c. Nat Cell Biol 2:645–52 Carper SW, Rocheleau TA, Storm FK (1990) cDNA sequence of a human heat shock protein HSP27. Nucleic Acids Res 18:6457 Drysdale BE, Howard DL, Johnson RJ (1996) Identification of a lipopolysaccharide inducible transcription factor in murine macrophages. Mol Immunol 33:989–98 Efthymiou CA, Mocanu MM, de Belleroche J, Wells DJ, Latchmann DS, Yellon DM (2004) Heat shock protein 27 protects the heart against myocardial infarction. Basic Res Cardiol 99:392–4 Francis JS, Dragunow M, During MJ (2004) Over expression of ATF-3 protects rat hippocampal neurons from in vivo injection of kainic acid. Brain Res Mol Brain Res 124:199–203 Frohli E, Aoyama A, Klemenz R (1993) Cloning of the mouse hsp25 gene and an extremely conserved hsp25 pseudogene. Gene 128:273–7 Gao Y, Signore AP, Yin W, Cao G, Yin XM, Sun F, Luo Y, Graham SH, Chen J (2005) Neuroprotection against focal ischemic brain injury by inhibition of c-Jun N-terminal kinase and attenuation of the mitochondrial apoptosissignaling pathway. J Cereb Blood Flow Metab 25: 694–712 Garrido C, Bruey JM, Fromentin A, Hammann A, Arrigo AP, Solary E (1999) HSP27 inhibits cytochrome c-dependent activation of procaspase-9. FASEB J 13:2061–70 Hollander JM, Martin JL, Belke DD, Scott BT, Swanson E, Krishnamoorthy V, Dillmann WH (2004) Over-expression of wild-type heat shock protein 27 and a nonphosphorylatable heat shock protein 27 mutant protects against ischemia/reperfusion injury in a transgenic mouse model. Circulation 110:3544–52 Honkaniemi J, States BA, Weinstein PR, Espinoza J, Sharp FR (1997) Expression of zinc finger immediate early genes in rat brain after permanent middle cerebral artery occlusion. J Cereb Blood Flow Metab 17:636–46 Kiessling M, Gass P (1994) Stimulus-transcription coupling in focal cerebral ischemia. Brain Pathol 4:77–83 Koizumi J, Yoshida Y, Nakazawa T, Ooneda G (1986) Experimental studies of ischaemic brain edema: 1. An experimental model of cerebral embolism in rats in which recirculation can be introduced in the ischemic area Jpn. J Stroke 8:1–8 Kolko M, Nielsen M, Bazan NG, Diemer NH (2002) Secretory phospholipase A2 Induces delayed neuronal

COX-2 expression compared with glutamate. J Neurosci Res 69:169–77 Lewis SA, Balcarek JM, Krek V, Shelanski M, Cowan NJ (1984) Sequence of a cDNA clone encoding mouse glial fibrillary acidic protein: structural conservation of intermediate filaments. Proc Natl Acad Sci USA 81:2743–6 Longo FM, Wang S, Narasimhan P, Zhang JS, Chen J, Massa SM, Sharp FR (1993) cDNA cloning and expression of stress-inducible rat hsp70 in normal and injured rat brain. J Neurosci Res 36:323–5 Nakagomi S, Susuki Y, Namikawa K, Kiryu-Seo S, Kiyama H (2003) Expression of the activating transcription factor 3 prevents c-Jun N-terminal kinaseinduced neuronal death by promoting heat shock protein 27 expression and Akt activation. Neurosci 23: 5187–96 Ohba N, Maeda M, Nakagomi S, Muraoka M, Kiyama H (2003) Biphasic expression of activating transcription factor-3 in neurons after cerebral infarction. Brain Res Mol Brain Res 115:147–56 Paul C, Manero F, Gonin S, Kretz-Remy C, Virot S, Arrigo AP (2002) Hsp27 as a negative regulator of cytochrome C release. Mol Cell Biol 22:816–34 Rane MJ, Pan Y, Singh S, Powell DW, Wu R, Cummins T, Chen Q, McLeish KR, Klein JB (2003) Heat shock protein 27 controls apoptosis by regulating Akt activation. J Biol Chem 278:27828–35 Ryder K, Nathans D (1988) Induction of protooncogene c-jun by serum growth factors. Proc Natl Acad Sci USA 85:8464–7 Sharp P, Krishnan M, Pullar O, Navarrete R, Wells D, de Belleroche J (2006) Heat shock protein 27 rescues motor neurons following nerve injury and preserves muscle function. Exp Neurol 198:511–8 Sharp PS, Akbar MT, Bouri S, Senda A, Joshi K, Chen HJ, Latchman DS, Wells DJ, de Belleroche J (2008) Protective effects of heat shock protein 27 in a model of ALS occur in the early stages of disease progression. Neurobiol Dis 30:42–55 Stetler RA, Cao G, Gao Y, Zhang F, Wang S, Weng Z, Vosler P, Zhang L, Signore A, Graham SH, Chen J (2008) Hsp27 protects against ischemic brain injury via attenuation of a novel stress-response cascade upstream of mitochondrial cell death signaling. J Neurosci 28: 13038–55 Takeda M, Kato H, Takamiya A, Yoshida A, Kiyama H (2000) Injury-specific expression of activating transcription factor-3 in retinal ganglion cells and its colocalized expression with phosphorylated c-Jun. Invest Ophthalmol Vis Sci 41:2412–21 Tsujino H, Kondo E, Fukuoka T, Dai Y, Tokunaga A, Miki K, Yonenobu K, Ochi T, Noguchi K (2000) Activating transcription factor 3 (ATF3). Induction by axotomy in sensory and motoneurons: a novel neuronal marker of nerve injury. Mol Cell Neurosci 15:170–82 van der Weerd L, Lythgoe MF, Badin RA, Valentim LM, Akbar MT, de Belleroche JS, Latchman DS, Gadian DG (2005) Neuroprotective effects of HSP70 over-expression after cerebral ischaemia—an MRI study. Exp Neurol 195:257–66

Supplementary Information accompanies the paper on the Journal of Cerebral Blood Flow & Metabolism website (http:// www.nature.com/jcbfm)

Journal of Cerebral Blood Flow & Metabolism (2010) 30, 849–856