Neurovascular Research Laboratory, Institut de Recerca Vall d'Hebron, Universitat Autònoma de Barcelona, Barcelona, Spain
Neurovascular Unit, Department of Neurology, Vall d'Hebron University Hospital, Universitat Autònoma de Barcelona, Hospital Vall d'Hebron, Barcelona, Spain
Address correspondence and reprint requests to Dr Joan Montaner, Neurovascular Research Laboratory, Neurovascular Unit, Institut de Recerca, Hospital Vall d'Hebron, Pg. Vall d'Hebron 119-129, 08035 Barcelona, Spain. E-mail: email@example.com
Finding an efficient neuroprotectant is of urgent need in the field of stroke research. The goal of this study was to test the effect of acute simvastatin administration after stroke in a rat embolic model and to explore its mechanism of action through brain proteomics. To that end, male Wistar rats were subjected to a Middle Cerebral Arteria Occlusion and simvastatin (20 mg/kg s.c) (n = 11) or vehicle (n = 9) were administered 15 min after. To evaluate the neuroprotective mechanisms of simvastatin, brain homogenates after 48 h were analyzed by two-dimensional fluorescence Difference in Gel Electrophoresis (DIGE) technology. We confirmed that simvastatin reduced the infarct volume and improved neurological impairment at 48 h after the stroke in this model. Considering our proteomics analysis, 66 spots, which revealed significant differences between groups, were analyzed by matrix-assisted laser desorption/ionization-time of flight mass spectrometry allowing the identification of 27 proteins. From these results, we suggest that simvastatin protective effect can be partly explained by the attenuation of the oxidative and stress response at blood–brain barrier level after cerebral ischemia. Interestingly, analyzing one of the proteins (HSP75) in plasma from stroke patients who had received simvastatin during the acute phase, we confirmed the results found in the pre-clinical model.
Our aim was to study statins benefits when administered during the acute phase of stroke and to explore its mechanisms of action through brain proteomics assay. Using an embolic model, simvastatin-treated rats showed significant infarct volume reduction and neurological improvement compared to vehicle-treated group. Analyzing their homogenated brains by two-dimensional fluorescence Difference in Gel Electrophoresis (DIGE) technology, we concluded that the protective effect of simvastatin can be attributable to oxidative stress response attenuation and blood–brain barrier protection after cerebral ischemia.
matrix-assisted laser desorption/ionization-time of flight
middle cerebral arteria occlusion
Stroke is the third leading cause of death and the most common cause of permanent disability among adults worldwide (Roger et al. 2012). Nowadays, the only existing treatment in acute phase of stroke approved by the Food and Drug Administration is the thrombolytic rt-PA (NINDS rt-PA Stroke Study Group 1995). Owing to its known side-effects (NINDS rt-PA Stroke Study Group 1997), new treatments with wider therapeutic window, which can be easily and rapidly administered and are available before the patients’ arrival in emergency units, are needed for neuroprotection. Many studies have demonstrated beneficial effects of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins) in decreasing the incidence of stroke in patients at high risk of cardiovascular disease (Jonsson and Asplund 2001; Martí-Fàbregas et al. 2004; Greisenegger et al. 2004; Yoon et al. 2004), although such effects have not been attributed to their cholesterol-lowering therapeutic actions. Alternatively, pleiotropic effects such as increasing endothelial nitric oxide (Laufs et al. 1998; Endres et al. 1998), reduction of oxidative stress (Balduini et al. 2001), inhibition of platelet activation (Rosenson 1999) and anti-inflammatory actions (Jain and Ridker 2005) seem to be implicated on statins’ protective role after stroke.
In this study, we have focused on the benefits that statins can offer when administered during the acute phase of stroke. Although different animal models have previously been used to study the utility of statins on brain ischemia (García-Bonilla et al. 2012), the mechanisms of neuroprotection are less clear and are mainly attributed to endothelial nitric oxide synthase up-regulation (Laufs et al. 1998; Amin-Hanjani et al. 2001 and Laufs et al. 2000). We aimed to explore the neuroprotective effects of simvastatin in an embolic model of stroke, as it closely resembles the nature of human stroke and has yet to be adequately explored. Moreover, the effects of simvastatin on different brain areas were evaluated by means of proteomics. Therefore, this will be the first systematic study using hypothesis-free system biology tools to test the global effect of simvastatin administered in the hyperacute phase of the ischemic event. Our data are likely to corroborate the idea of stroke patients being able to benefit from simvastatin treatment at very early time points, in an ambulance, for instance, before arriving at the hospital.
Materials and methods
All procedures were approved by the Ethics Committee of the Vall d'Hebron Research Institute (protocol number 02/09) and were conducted in compliance with the Spanish legislation and in accordance with the Directives of the European Union. Experiments were performed in male Wistar rats (300–325 g; Charles River Laboratories Inc., Wilmington, MA, USA). Rats were kept in a climate-controlled environment on a 12-h light/12-h dark cycle. Food and water were available ad libitum and analgesia (Magnesic methamizol, Boehringer Ingelheim, España) was given to all rats every 24 h from the first until the last day of the experiment protocol to minimize their pain and discomfort.
Infarction in the territory of the middle cerebral artery (MCA) was induced trough a clot injection as described in detail elsewhere (Zhang et al. 1997). Animals were anesthetized under spontaneous respiration with 2% isoflurane (Abbot Laboratories, Kent, UK) in oxygen during surgery and body temperature was maintained at 37°C. Arterial blood from a donor rat was withdrawn to form two identical clots (length: 1.5 cm; diameter: 0.3 mm, each). Continuous laser–Doppler flowmetry (Moor Instruments, Devon, UK) was used to monitor regional cerebral blood flow and only animals that exhibited a reduction > 75% in regional cerebral blood flow during middle cerebral arteria occlusion (MCAO) were included in the study.
A total of 77 animals were needed to obtain all data. For the study of simvastatin efficacy 55 animals were used. From these, 19 animals were excluded after applying the following criteria: inappropriate occlusion of the middle cerebral artery after embolization (n = 10); spontaneous reperfusion within 10 min of occlusion (n = 4), death during the following minutes after the occlusion (n = 2), or poor cerebral blood flow register (n = 3). Sixteen animals died before the completion of the experimental protocol (12 during the first 24 h and 4 between 24 and 48 h after the occlusion).
For the proteomic study, four animals per group were evaluated. To that end, 22 animals were needed. Seven animals were excluded following the same criteria: inappropriate occlusion (n = 3), spontaneous reperfusion (n = 3), and death just after the occlusion (n = 1). Seven animals died during the experimental protocol (six during the first 24 h and one at 48 h post-ischemia). All animals were subjected to MCAO and randomly allocated to experimental groups (simvastatin or vehicle) using a computer-generated randomization list. One mL of Simvastatin solution (20 mg/kg (diluted in vehicle); Uriach Laboratories, Barcelona, Spain) or 1 mL of vehicle was subcutaneously injected 15 min after occlusion in a blinded manner. Simvastatin was prepared by opening the lacton ring and activating it. In short, simvastatin was dissolved in vehicle [distilled H2O (75%), absolute ethanol (10%), and 0.1 M NaOH (15%)], incubated at 50°C for 2 h and pH adjusted at 7.2.
Rats were assessed by a blind experimenter using a 9-point neurological deficit scale (modified Bederson test), as previously described (Perez-Asensio et al. 2005). Four consecutive tests were conducted: (i) spontaneous activity (moving and exploring = 0, moving without exploring = 1, no moving or moving only when pulled by the tail = 2); (ii) left drifting during displacement (none = 0, drifting only when elevated by the tail and pushed or pulled = 1, spontaneous drifting = 2, circling without displacement, or spinning = 3), (iii) parachute reflex (symmetrical = 0, asymmetrical = 1, contralateral forelimb retracted = 2), and (iv) resistance to left forepaw stretching (stretching not allowed = 0, stretching allowed after some attempts = 1, no resistance = 2). Neurological score was assessed at 90 min, 24 and 48 h after occlusion.
Infarct volume was evaluated by an investigator blinded to the treatment using thionin stain. Animals were killed and transcardially perfused with Paraformaldehyde 4%. Brains were removed and fixed during 24 h in Paraformaldehyde 4% at 4°C and immersed in sucrose for another 24 h also at 4°C. Afterward, brains were embedded in OCT and placed at −80°C. Coronal brain sections (30 μm) were cryostat-cut, mounted on slides and stained with thionin. Non-injured areas were stained in deep blue, whereas ischemic areas remained pale. Images of the stained slices were captured using a CanoScan 4200F scanner (Canon USA Inc., New York, NY, USA) and infarct, contralateral (CL) and ipsilateral (IP) areas were outlined and quantified using an image analysis system (Scion Image v4.02, Scion Corporation, Frederick, MD, USA). Infarct volume was measured by integration of infarcted areas and expressed as a percentage of the IP hemisphere. Edema correction of infarct volume was evaluated taking into account the following equation: volume correction = (infarct volume × CL volume)/IP volume.
Brain tissue samples
To maintain brain tissue structure for the proteomic study, whole brains from a second group of rats were used to obtain brain homogenates as described below. Owing to infarct volume, calculation was not possible in this set of animals, only those that showed a reduction > 75% in regional cerebral blood flow and a neurological score > 3 at 90 min and ≥ 2 at 24 and 48 h were included in the study. The inclusion criteria applied guaranteed that all the rats included in the study suffered an ischemic episode. Rats were transcardially perfused with cold saline and brains were removed and divided into four areas after discarding the hippocampus: cortex IP, cortex CL, striatum IP, and striatum CL. Each separated area separately was snap frozen in liquid nitrogen.
Frozen samples were homogenized using cold difference in gel electrophoresis (DIGE) lysis buffer (7 M urea, 2 M thiourea, 4% CHAPS, 30 mM Tris, pH 8.5), for protein extraction. After sonication and centrifugation at 12 000 g for 5 min at 4°C, the cleared protein extract was further purified by a modified trichloroacetic acid–acetone precipitation (2-D-CleanUp kit, Amersham Biosciences, Munich, Germany) and re-dissolved in DIGE lysis buffer. Protein concentration was determined using the Bio-Rad RCDC Protein Assay (Bio-Rad, Hercules, CA, USA).
Two-dimensional fluorescence difference in gel electrophoresis
Samples of each experimental condition (cortex or striatum and IP or CL) and a pool consisting of equal protein amounts of each of these samples were analyzed by two-dimensional fluorescence DIGE technology. The pool was prepared to be used as an internal standard for quantitative comparisons as described (Alban et al. 2003). Samples were labeled with Cy3 or Cy5 cyanine dyes; internal standard pooled sample was labeled with Cy2 dye, by the addition of 400 pmol of Cy dye in 1 μL of anhydrous N,N-dimethylformarnide per 50 μg of protein. To avoid possible bias introduced by labeling efficiency, the samples from each group were alternatively labeled with both Cy3 and Cy5 dyes. After 30-min incubation on ice in the dark, the reaction was quenched with 10 mM lysine and further incubated for 10 min. Samples were finally combined according to the experimental design, at 50 μg of protein per Cy dye per gel, and diluted 2-fold with Isoelectric Focosing IEF sample buffer (7 M urea, 2 M thiourea, 4% wt/vol CHAPS, 2% dithiothreitol, 2% pharmalytes pH 3–10 and 0.002% bromophenol blue). The two-dimension electrophoresis was performed using GE Healthcare reagents and equipment (Buckinghamshire, UK). First dimension isoelectric focusing was performed on immobilized pH gradient strips (24 cm; linear gradient pH 3–10) using an Ettan-immobilized pH gradientphor system (GE Healthcare Europe, Freiburg, Germany). First, strips were incubated overnight in 450 μL of re-hydration buffer (7 M urea, 2 M thiourea, 4% wt/vol CHAPS, 1% pharmalytes pH 3–10, 100 mM DeStreak and 0.002% bromophenol blue). Then, samples were applied via cup loading near the acidic end of the strips. After focusing at a global voltage of 67 kV, strips were equilibrated first for 15 min in 6 mL of reducing solution [6 M urea, 100 mM Tris–HCl pH 8, 30% vol/vol glycerol, 2% wt/vol sodium dodecyl sulfate (SDS), 5 mg/mL dithiothreitol, and 0.002% bromophenol blue] and then in 6 mL of alkylation solution (6 M urea, 100 mM Tris–HCl pH 8, 30% vol/vol glycerol, 2% wt/vol SDS, 22.5 mg/mL iodoacetamide, and 0.002% bromophenol blue) for 15 min, on a rocking platform. Second-dimensional SDS–polyacrylamide gel electrophoresis were run by overlaying the strips on 12.5% isocratic Laemmli gels (24 × 20 cm), casted in low fluorescence glass plates on an Ettan DALT six system (GE Healthcare Europe). Gels were run at 20°C at constant power 2.5 W per gel for 30 min followed by 17 W per gel until the bromophenol blue tracking front reached the end of the gel. Fluorescence images of the gels were acquired on a Typhoon 9400 scanner (GE Healthcare). Cy2, Cy3, and Cy5 images were scanned at 488/520, 532/580, and 633/670 nm excitation/emission wavelengths, respectively, at a 100 μm resolution. Image analysis and statistical quantification of relative protein abundances were performed using Progenesis SameSpots software (Non-linear Dynamics Limited, Newcastle upon Tyne, UK).
Protein identification by mass spectrometry
To investigate treatment effects on protein expression in ischemic animals, brain samples obtained from ischemic IP and CL hemispheres of both cerebral cortex and striatum from simvastatin-treated ischemic animals (Simv) and vehicle-treated ischemic animals (Veh) as control, were resolved by two-dimensional DIGE gels. The samples were analyzed by pairs of simvastatin and control samples mixed with the pooled internal standard. These pairs of samples were used for comparisons in the differential protein expression analysis. The selected comparisons for protein expression analysis were as follows: ipsilateral hemisphere from simvastatin-treated animals (Simv-IP) versus vehicle-treated animals (Veh-IP); contralateral hemisphere from simvastatin-treated animals (Simv-CL) versus vehicle-treated animals (Veh-CL); ipsilateral versus contralateral hemisphere from vehicle-treated animals (Veh-IP and Veh-CL, respectively); and ipsilateral versus contralateral hemisphere from simvastatin-treated animals (Simv-IP and Simv-CL, respectively). All comparisons were made in both cerebral cortex (C) and striatum (S). The differential protein expression analysis of two-dimensional DIGE gels was performed in samples from four different animals. Spots present in four of four animal samples per group, with significant anova test (p < 0.05) and an averaged ratio > 1.3 between comparisons, were considered and selected for protein identification by matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry (MS).
Protein spots were excised from the gel using an automated Spot Picker (GE Healthcare). In-gel trypsin digestion was performed as described (Shevchenko et al. 1996) using autolysis stabilized trypsin (Promega, Madison, WI, USA). Tryptic digests were purified using ZipTip microtiter plates (Millipore, Cork, Ireland). MALDI-TOF MS analysis of tryptic peptides was performed on an Ultraflex TOF-TOF Instrument (Bruker, Bremen, Germany). Samples were prepared using Prespotted AnchorChip (PAC96) target with alpha-cyano-4-hydroxycinnamic acid matrix for 96 sample spots and 24 calibration spots (Bruker Daltonics). The spectra were processed using Flex Analysis 3.2 software (Bruker Daltonics). Identification of the proteins was carried out by peptide-mass fingerprinting data from MALDI-TOF MS. Database searches were performed using the MASCOT program (Matrix Science, London, UK). Protein identification was confirmed by TOF-TOF post-source decay fragmentation spectra or by ion trap mass spectrometry as described (Esselens et al. 2008).
Validation by western blot
With the same samples previously used in DIGE analysis, validation of DIGE and MS results of selected proteins in brain homogenates was performed by western blot. Briefly, equal protein amounts (35 μg) were resolved by 5% or 8% sodium dodecyl sulfate–polyacrylamide gel electrophoresis. Separated proteins were transferred to a polyvinylidene difluoride membrane using a mini Trans-Blot® Electrophoretic Transfer Cell (Bio-Rad Laboratories) for 1 h at 100 V. Non-specific bindings were blocked with 10% non-fat milk and 0.05% tween-phosphate buffer saline before membranes were incubated overnight at 4°C with the following antibodies: mouse anti-Guanin nucleotide-binding-protein (Go protein alpha, 1 : 3000, catalog no.ab78218; Abcam, Cambridge, UK), mouse anti-heat-shock protein 60 (HSP60, 1 : 100, catalog no. sc-376240; Santa Cruz Biotechnology, Santa Cruz, CA, USA), mouse anti-HSP75 (1 : 200, catalog no sc-135944; Santa Cruz Biotechnology), mouse anti-HSP70 (1 : 1000; Calbiochem, Dramstadt, Germany), rabbit anti-Hemoglobin α (H-80, 1 : 1000, catalog no. sc-21005; Santa Cruz Biotechnology), rabbit anti-Dynamin (H-300, 1 : 200, catalog no. sc-11362, Santa Cruz Biotechnology), rabbit anti-DRP-2 (1 : 200, catalog no. sc-30228; Santa Cruz Biotechnology), rabbit anti-Alpha Fodrin (Spectrin α, 1 : 1000, catalog no. ab75755; Abcam) in 10% non-fat milk. Secondary antibodies (Chemicon International, Inc, Rockford, IL, USA) anti-rabbit-horseradish peroxidase or anti-mouse-horseradish peroxidase, were diluted 1 : 2000 in 10% non-fat milk and incubated at 25°C for 1 h. Before and after incubations, membranes were washed three times (10 min each) with 0.05% tween-phosphate buffer saline. The substrate reaction was developed with chemiluminescent reagent Luminol (Amersham Biosciences, GE Healthcare) and visualized with a luminescent image analyzer (Las-3000; Fujifilm Medical Systems, Stamford, CT, USA). Scanned western blots were quantified with the image Quantity One software package (Bio-Rad Laboratories). All protein levels were corrected by a loading control detected using the antibody mouse anti-Actin (1 : 20 000, catalog no. A1978, Sigma-Aldrich, St Louis, MO, USA) except for Guanin Go. Levels, which was corrected using anti-mouse-αTubulin (1 : 1000, catalog no,T6199; Sigma-Aldrich).
HSP75 protein level in stroke patients’ plasma
A pilot study was performed to explore in human plasma from stroke patients the changes found in HSP75 in the proteomic study. Blood samples from stroke patients included in a clinical trial (Montaner et al. 2008) who received simvastatin or placebo at 3–12 h from symptom onset at an initial dose of 40 mg/day for the first week were used. Samples were drawn in citrate tubes just before treatment administration and plasma fractions were obtained by centrifugation (1500 g during 15 min at 4°C) and stored at −80°C until analysis was performed. A sandwich enzyme immunoassay (Human HSP75 kDa, mitochondrial TRAP1/HSP75 ELISA kit; Cusabio Biotech CO., LTD, Wuhan, China) was performed according to manufacturer's instructions. Optical density was detected with a microplate reader (Synergy Mx, Biotek, China). Each sample was assayed twice. If intra-assay coefficient of variation (CV) was < 20%, the mean value of both measurements was used in the statistical analysis, whereas samples with a CV > 20% were excluded from the analysis. Only patients with a complete temporal profile: basal time (before treatment administration), third day and seventh day after stroke, were included in the analysis (placebo = 5 patients, simvastatin = 6 patients). All results were expressed as a percentage of basal values. Clinical and demographic characteristics of the included patients were taken into consideration and are shown in Table S1 (Supporting Information).
Statistical significance for intergroup differences was assessed by anova followed by Bonferroni post hoc test using Progenesis SameSpots software (Non-linear Dynamics Limited). Statistical significance for quantified data from western blots was analyzed by Student's t-test using GraphPad Prism_v5 software. For non-parametric data, Mann–Whitney test, Kruskal–Wallis test, and Friedman test followed by Dunn's Multiple Comparison test were assayed. Bars represent mean ± SD for parametric data and box plots represent median (IQR) for non-parametric data. For all data, a p < 0.05 was considered statistically significant at a 95% confidence level.
Infarct lesion, neurological deficit, and mortality
Animals treated with simvastatin 15 min after the occlusion showed a 43% reduction on the infarct lesion measured 48 h post-ischemia compared to vehicle-treated animals [21.95 (15.44, 42.95)% for vehicle vs.12.92 (10.24, 19.63)% for simvastatin, p < 0.05] (Fig. 1a, b).
Neurological deficit assessed by the Bederson test, at 90 min post-occlusion, showed no differences between groups. At 24 h, simvastatin-treated rats scored lower [6 (5, 7) for vehicle vs. 5.75 (4, 7) for simvastatin], although differences were not significant. However, at 48 h post-occlusion, rats treated with simvastatin showed a 35% significant improvement in neurological behavior [5.50 (4.25, 6) vs. 3 (1.25, 5.75), p < 0.05] (Fig. 1c).
Mortality rates were high, as already reported for this MCAO model (Lapergue et al. 2010; Hernández-Guillamon et al. 2010), but no significant differences were found between vehicle- and simvastatin-treatment groups (43.75% vs. 42.86%, respectively) Fig. 1d.
Protein expression profile in simvastatin-treated animals
Brain samples of each experimental condition (cortex or striatum of IP or CL hemispheres) were analyzed following the experimental design showed in the flow chart (Fig. 2). In the two-dimensional gels we detected 186 protein spots present in all samples (Fig. 3a). We found a total of 74 spots with significant expression differences which were further analyzed by MALDI-TOF MS. Among them, 38 spots and 27 different proteins were identified, as detailed in Table S2 Supporting Information. For expression analysis, only spots present in all samples (4 per treatment group), showing a significant p value (p < 0.05) and fold change> 1.3 between groups were selected for protein identification by MALDI-TOF MS (Fig. 3b). The effect of simvastatin on the ischemic brain was analyzed by the comparison of the Simv-IP group with respect to the Veh-IP group, in cerebral cortex (Fig. 4A; Simv-IP vs. Veh-IP) and striatum (Fig. 4B; Simv-IP vs. Veh-IP). The differential protein expression analysis in cerebral cortex detected nine spots with significant differences and five proteins were successfully identified. The results showed higher induction of the alpha subunit of guanine nucleotide-binding protein Go (2 isoforms), and decreased levels of mitochondrial 60 kDa HSP60 and fumarate hydratase (2 isoforms d). In striatum, eight spots were detected with significant differences and two proteins were identified; microtubule-associated protein RP/EB showed higher levels in the Simv-IP group, whereas the 75 kDa HSP75 presented lower levels in this group of animals.
The effect of simvastatin on non-ischemic tissue was analyzed when compared the Simv-CL group with the Veh-CL control group, in cerebral cortex (Fig. 4A; Simv-CL vs. Veh-CL) and striatum (Fig. 4B; Simv-CL vs. Veh-CL). The differential protein expression analysis in cerebral cortex detected four spots corresponding to two proteins. The result showed higher levels of the beta subunit of mitochondrial ATP synthase, and decreased expression of fumarate hydratase in Simv-CL group. In striatum, four spots corresponding to two proteins were detected; the 70 kDa HSP70 presented higher levels in Simv-CL, whereas phosphoglycerate mutase 1 presented lower levels in the same group.
The effect of ischemia in vehicle-treated animals was studied by comparing the changes in protein expression of the Veh-IP group with respect to the Veh-CL control. In cerebral cortex, 16 spots were detected and eight proteins were successfully identified (Fig. 4A; Veh-IP vs. Veh-CL). The results showed higher levels of hemoglobin alpha and beta, and serine protease inhibitor A3N (Serpin A3N), and decreased levels of dynamin 1 (detected in three isoforms), pyruvate carboxylase, and hexokinase 1. In striatum, 10 spots were detected and eight proteins were identified (Fig. 4B; Veh-IP vs. Veh-CL). Among them, albumin, hemoglobin alpha and beta (2 isoforms), dihydropyrimidinase-related protein 2 (DRP2), and HSP75 were detected presenting higher levels in Veh-IP, whereas neuromodulin/GAP43 (2 isoforms) was identified with decreased levels.
Finally, the effect of simvastatin was studied in the ischemic groups by comparing the protein changes in the Simv-IP group with respect to the Simv-CL control. In cerebral cortex, 15 spots were detected and eight proteins successfully identified (Fig. 4A; Simv-IP vs. Simv-CL). The results showed higher induction of alpha spectrin (fragment), 6-phosphogluconate dehydrogenase decarboxylating and actin-related protein 2/3 complex subunit 4, and decreased levels of DRP2, synapsin 2 (2 isoforms), beta tubulin 2A and dynamin-like 120 kDa protein. In striatum, 20 spots were detected and 13 proteins were identified (Fig. 4B; Simv-IP vs. Simv-CL). Hemoglobin alpha and beta, DRP2, in addition to myelin basic protein S (MBP-S) and fructose-bisphosphate aldolase A, were detected showing higher levels in Simv-IP; whereas dynamin 1 (3 isoforms), neurofilament L (NF-L, 2 isoforms), neurofilament M (NF-M), pyruvate carboxylase, and hexokinase 1 presented lower levels.
Once the proteins were identified, they were classified and grouped into functional role according to data base of Pathway Commons, Pather, and NCBI (Fig. 5). Proteins showed specific functions in the nervous system (DRP2, dynamin 1, dynamin-like 120 kDa protein, MBP-S, neuromodulin, and synapsin 2); another six were cytoskeleton proteins (actin-related protein 2/3 complex subunit 4, microtubule-associated protein RP/EB, NF-L, NF-M, alpha spectrin, and beta tubulin 2A); five were stress–response-related proteins (HSP60, HSP70, HSP75, fumarate hydratase and Serpin A3N); five belonged to the carbohydrate metabolic process (6-phosphogluconate dehydrogenase decarboxylating, fructose-bisphosphate aldolase A, hexokinase 1, phosphoglycerate mutase 1, and pyruvate carboxylase); three proteins were from blood circulation (albumin and hemoglobin alpha and beta); one was a signal-transduction protein(the guanine nucleotide-binding protein Go);and one was an electron transport protein (the ATP synthase).
Then, the changes observed in some of the proteins identified by MALDI-TOF MS were also evaluated by western blot. Among the 27 proteins differentially expressed between groups, eight (corresponding to 13 comparisons) were chosen regarding antibody availability, fold change and functional role. From these 13 comparisons, seven were found in the same expression-direction although differences did not reach statistical significance, whereas the other six could not be validated (Figures S1 and S2).
Levels of HSP75 in stroke plasma samples
Next, to study the clinical impact of our study performed using a rat stroke model, we tested the levels of one of the proteins, HSP75, which showed relevant lower levels in simvastatin-treated ischemic brains. In our cohort of stroke patients we observed that, while patients who received placebo did not show significant differences at several time points in terms of HSP75 plasma concentration, patients receiving simvastatin after stroke, presented lower levels at third and seventh day after the stroke compared to basal HSP75 level (p < 0.05) (Fig. 6).
Our results clearly show neuroprotective effects of simvastatin on brain ischemia when administered in a single dose (20 mg/kg) during the acute phase. The dose was selected according to previously published meta-analysis (García-Bonilla et al. 2012) which determined that the typical range in in vivo rodent studies was 10–100 mg/kg/day. Although the dose chosen differs with the one used in human studies (40 mg/day), we are of the opinion that more studies are needed to clarify the extrapolation of statin doses between human and rodent studies.
Despite our efforts in reducing the pain and discomfort of the rats subjected to the embolic stroke, the mortality associated to this experimental model was around 40%, as shown in other studies (Hernández-Guillamon et al. 2010). This high mortality rate, which is considered a limitation of this study, is on account of the severe brain injury and some secondary complications such as weight loss.
Previous studies, compiled in a recent meta-analysis (García-Bonilla et al. 2012), reflected statins’ neuroprotection effect in both infarct size reduction and neurological deficit improvement in the short term. However, to our knowledge, only a recent study (Pirzad Jahromi et al. 2012), has tested simvastatin efficacy after ischemia in an embolic model, the one that best resembles human stroke (Carmichael 2005). Previous investigations conducted with the intraluminal model showed a lesion reduction of 15% administrating a very low dose of simvastatin (1 mg/kg) (Chen et al. 2003) and up to 45.62% or 30.41% infarct reductions when higher doses of simvastatin were administered (20 and 40 mg/kg), respectively (Laufs et al. 1998; Nagaraja et al. 2006). This beneficial effect was explained, at least in part, by one of the main strengths of statins: their ability to up-regulate endothelial nitric oxide synthase expression and activity, promoting a cerebral blood flow increase that leads to a cerebral infarct size reduction (Kureishi et al. 2000). It has also been reported that statins interfere with the coagulation system by up-regulating tissue-type plasminogen activator (tPA) and down-regulating plasminogen activator inhibitor (PAI-1) (Essig et al. 1998). Moreover, statins alter the inflammatory response to ischemia, by minimizing the inflammatory components of the ischemic cascade and reperfusion injury. They also modulate the immune responses by the recruitment, differentiation and secretory activity of monocytes, macrophages, and T cells (Palinski and Tsimikas 2002).
Because of statins pleiotropic effects, other proteomic approaches have been carried out so as to explore their effect in different cells or tissues. Despite our study being the first, to our knowledge, to explore statins’ neuroprotective effect after cerebral ischemia, the conclusions of some of these other papers may also contribute to the understanding of this mechanism. Pienaar et al. (2009), for instance, evaluated rats’ substantia nigra and reported that statins may prevent the increase in oxygen-free radical productions. Gu et al. (2012), after studying the translocation of lipid-raft-related proteins in endothelial cells, revealed an up-regulation on anti-oxidative proteins, as well as a down-regulation on proteins associated with inflammation and cell adhesion. Brioschi et al. (2013) analyzed endothelial cells and demonstrated that statins influenced the thrombogenic response of the vessel wall as well as thrombotic factors in blood.
In our study, however, we sought to study the brain proteomic changes after a stroke embolic model and explore the effect that simvastatin may exert on different ischemic brain regions (striatum and cortex) and in healthy tissue (contralateral brain hemisphere).
Apart from the neuroprotection study, brains of four animals per treatment group were analyzed through proteomics technique. Although this low sample size can be assumed as a limitation of the study, we could observe appealing results. The time point of the analysis (48 h after the ischemia) was chosen in accordance with neuroprotection study results, which showed that infarct lesion was well established and that simvastatin had exerted its protection at this time point. Interestingly, we have found different protein expression changes considering the analyzed area, possibly because of the distinct effects that neuroprotectants are described to induce on cortex and striatum after submitting the rats to an eMCAO (Wang et al. 2001). In addition, regarding simvastatin effect on infarct size, we cannot rule out that the differences on protein expression were in partly explained by the smaller infarct lesion.
First, our data show simvastatin influencing cell signaling by up-regulating the levels of the guanine nucleotide-binding protein Go, and reducing cellular stress by decreasing the HSP60 and the fumarate hydratase. As described by others (Worley et al. 1986 and Neves et al. 2002), guanine nucleotide-binding protein Go seems to play an important role in signal transduction by modulating adenilate cyclase (Huff et al. 1985) and coupling to ion channels (Pfaffinger et al. 1985) whereas HSP60 is reported to be at high levels after many stressors, and to be indicative of mitochondrial biogenesis (Yin et al. 2008). Thus, both actions would be helpful for recovery from ischemic damage in our system. Our results are in accordance with other studies which conclude that simvastatin therapy is associated with significant reduction of serum HSPs antibody titers in general (Moohebati et al. 2011) and plasma HSP60 antibody titers in particular (Ghayour-Mobarhan et al. 2005). These studies claim that HSPs levels decrease may be modulated by simvastatin through its immunomodulatory properties.
Second, the effect of simvastatin into the striatum, showed an increase of the microtubule-associated protein RP/EB (MAP1) and a decrease in the HSP75 concentration. Both processes are indicative of cell reparation and lower stress induction, respectively, and are useful for tissue preservation. MAP1, a structural microtubule-associated protein, was found to be highly expressed in developing neurons playing an important role in neurite and axon extension (Tortosa et al. 2013). Regarding HSP75, it has been previously published that levels of mitochondrial HSP75 are high during ischemia conditions (Kiang and Tsokos 1998). Other authors, however, showed that Hsp75 levels decreases reactive oxygen species production and preserves mitochondrial membrane potential during glucose deprivation in astrocyte cell cultures (Voloboueva et al. 2008). Overall, as simvastatin-treated rats showed a less severe brain injury, we can infer that lower HSPs levels are needed to exert neuroprotection. Moreover, an ischemic insult involving immunological mechanisms such as autoimmune reaction against HSPs could be mitigated through immunomodulatory statin effect on lymphocyte function (Guisasola et al. 2009).
Third, the effect of simvastatin in contralateral hemispheres revealed that only four proteins show significant differences regarding its concentration levels, in both cerebral cortex and striatum regions. These results may indicate a limited effect of simvastatin administered acutely in the non-ischemic tissue, although we noted that simvastatin induces HSP70 expression in healthy striatum.
On the other hand, when we studied the effect of ischemia in simvastatin-treated animals we detected 15 and 20 spots in cerebral cortex and striatum, respectively. In cerebral cortex, the increase of spectrin fragment may reflect possible neuroregeneration and repair in response to the degeneration caused by the ischemia (Aikman et al. 2006), whereas 6-phosphogluconate dehydrogenase decarboxylating up-regulation is indicative of glycolisis (Belfiore et al. 1975). Moreover, the increased levels of actin-related protein 2/3 complex subunit 4 would show cell reparation (Hetrick et al. 2013). In addition, the decrease in functional and structural neuronal proteins, such as synapsin 2, dynamin-like 120 kDa protein, and beta tubulin 2A, would indicate neuronal damage. Furthermore, the decreased concentration of DRP2 could show less significant ischemic injury, although some controversy exists about this protein as it has been reported to be up or down-regulated depending on its phosphorilation status (Zhou et al. 2008). In striatum, we identified blood circulation proteins, in addition to DRP2, fructose-bisphosphate aldolase A, and MBP-S. The latter may indicate developing of glial cells (Lashgari et al. 1990). In addition, the decrease in functional or structural neuronal proteins, such as dynamin 1, NF-L and –M, and proteins from carbohydrate metabolic process, such as pyruvate carboxylase and hexokinase 1, would point out again to neuronal damage (Julien and Biebuyck 1990).
Finally, we studied the effect of ischemia in un-treated animals. Sixteen and 10 spots in cerebral cortex and striatum, respectively, were detected and identified as blood circulation proteins, hemoglobin alpha and beta, or albumin, in addition to the decrease of proteins essential in the CNS function or in the carbohydrate metabolic process as dymanin 1 and neuromodulin, or pyruvate carboxylase and hexokinase 1. Accordingly with previous proteomics approaches (Chen et al. 2007; Koh 2010) after cerebral ischemia in rat, in striatum it was detected increased levels of DRP2, a protein which was reported to be involved in neuronal differentiation and axonal guidance (Quinn et al. 1999) by promoting microtubule formation (Fukata et al. 2002) and HSP75, a stress response protein involved in mitochondria biogenesis (Bellizzi et al. 2009).
In summary, when comparing the effect of ischemia in simvastatin-treated animals respect to the un-treated animals, we found that in the infarct core region (striatum) the blood circulation proteins were significantly decreased or absent, such as hemoglobin beta or albumin (additional information is presented in Supporting information) as well as HSP75. Thus, the most important finding of this study was simvastatin capacity to attenuate stroke-induced changes in blood–brain barrier (BBB) permeability and to protect against cell stress and signal transduction stimulation. This statement is consistent with other in vitro (Kahles et al. 2007) and in vivo studies which concluded that statin BBB protection was associated with the abolition of isoprenylation processes (Law et al. 2006). In addition, it has experimentally been proved (Nagaraja et al. 2006) that rats receiving simvastatin 30 min after MCA occlusion had reduced volumes of AIB leakage at 6 h post-ischemia.
Moreover, when comparing this effect in the cortex region, the blood circulation proteins were absent, as well as Serpin A3N, an inflammatory response protein (Law et al. 2006). Simvastatin also decreased DRP2, a characteristic protein of ischemic damage (Chen et al. 2007). In addition to the protection against cell stress, simvastatin demonstrated again a protection against BBB damage in both regions, striatum and cortex (additional information is presented in Supporting information). All in all, BBB protection could lead to a milder inflammatory process as well as in a neurological improvement, as we could see in our experimental study.
Regarding global protein changes previously observed in brain rodents after an ischemic event, they seem to depend on the animal stroke model used and on evaluation time point. It is likely that the earlier the brains are evaluated, the more differences can be found. Chen et al. (2007) and Koh (2011) for instance, reported proteins up-regulation more than 3-fold 24 h after submitting rats to a permanent MCAO. Other studies performed in mice (Hori et al. 2012; Föcking et al. 2006) reported expression changes of 1.5-2 fold when evaluating at 1, 6, or 24 h after the ischemia. Nevertheless, and in accordance with our results, some proteins show up in most of the studies such as HSP's or DRP-2.
Furthermore, the validation study by western blot was performed with the same samples used in the proteomics study. This second assay tested only eight of the 27 proteins found by proteomics analysis and reflected that 53.85% of the analyzed comparisons coincided with previous results, meaning that they presented similar expression patterns than those found in the proteomic study, although no statistical significance was reached. This poor validation percentage could be attributed to the small sample size (n = 4) and also to the incapability of western blot technique to distinguish between different protein species (isoelectric point) of the same protein.
Considering the consistent fold change found in the proteomic assay, we moved one step forward with the analysis of HSP75 in plasma samples of stroke patients treated with simvastatin in the acute phase of stroke. Accordingly with our proteomics data, simvastatin-treated patients appear to have a more homogenous distribution and lower HSP75 plasma levels after stroke than placebo-treated patients. Related to the well-documented role of HSPs in maintaining cellular homeostasis in response to stress (Bellizzi et al. 2009) it has been published that both mitochondrial and cytosolic inducible HSPs are up-regulated during ischemic conditions (Kregel 2002). Moreover, and in agreement with our results, it has also been demonstrated that simvastatin therapy is associated with significant reductions in serum HSPs (Moohebati et al. 2011) owing to statins immunosupressive properties.
We assume this substudy entails some limitations such as the small sample size or the fact of analyzing only the HSP75 protein in human samples. Moreover, a better gender-matched between simvastatin and placebo-treated patients would have been useful to rule out the possibility of HSP75 levels being influenced by gender or other demographic factors.
All in all, however, it should be emphasized the importance of the concordant results found in both experimental and clinical assays, support embolic cerebral ischemia model as capable of reproducing human stroke conditions.
In conclusion, our study identifies, for the first time through proteomic techniques, the key proteins that are likely involved in the neuroprotective mechanisms of simvastatin. Although simvastatin seems to be a promising neuroprotective drug, clinical trials are needed to demonstrate its beneficial effects also in stroke patients in the hyperacute phase. In addition, the use of surrogate markers such as HSP75 may be useful in future studies.
Acknowledgments and conflict of interest disclosure
M.C is supported by a pre-doctoral fellowship grant (FI10/00508); and M.H-G. and A.R. are senior researchers of the Miguel Servet Programme (CP12/03259 and CP09/00265) all those from the Instituto de Salud Carlos III from the Spanish Ministry of Economy. The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreements no. 201024 and no. 202213 (European Stroke Network). Neurovascular Research Laboratory takes part in the Spanish stroke research network INVICTUS (RD12/0014/0005) and the Proteomics Laboratory at VHIO belongs to the Spanish networked platform ProteoRed-ISCIII. This study was partially funded by projects FIS 11/0176 on stroke biomarkers research and EC07/90195 on increasing safety and efficacy of simvastatin neuroprotection.
All experiments were conducted in compliance with the ARRIVE guidelines. The authors have no conflict of interest to declare.