Exercise attenuates neurological deficits by stimulating a critical HSP70/NF-κB/IL-6/synapsin I axis in traumatic brain injury rats

Eight groups of rats were randomly assigned as the following: (i) no exercise-preconditioned sham controls received siRNA-vector (EP⁻ + sham + siRNA-vector); (ii) no exercise-preconditioned sham controls received siRNA-HSP70 (EP⁻ + sham + siRNA-HSP70); (iii) exercise-preconditioned sham controls received siRNA-vector (EP⁺ + sham + siRNA-vector); (iv) exercise-preconditioned sham controls received siRNA-HSP70 (EP⁺+ sham + siRNA-HSP70); (v) no exercise-preconditioned TBI rats received siRNA-vector (EP⁻ + TBI + siRNA-vector); (vi) no exercise-preconditioned TBI rats received siRNA-HSP70 (EP⁻ + TBI + siRNA-HSP70); (vii) exercise-preconditioned TBI rats received siRNA-vector (EP⁺ + TBI + siRNA-vector); and (viii) exercise-preconditioned TBI rats received siRNA-HSP70 (EP⁺ + TBI + siRNA-HSP70) (Table 1).

In experiment 1, the optical density (O.D.) values of HSP70 of ipsilateral hemisphere were determined 3 days after a TBI or a sham operation in all groups.

In experiment 2, 1 day before, 1–3 days after a TBI or a sham operation, we determined the neurological motor functions for all eight groups of rats.

In experiment 3, in all eight groups of rats 3 days after the operation, we measured both cerebral contusion and Evans Blue extravasations.

In experiment 4, both neuronal loss and apoptosis were determined by using immunofluorescence stain in all eight groups 3 days after a TBI.

In experiment 5, qPCR arrays were used to evaluate blood levels of inflammatory cytokines and chemokines in six groups as follows: (i) EP⁻ + sham + siRNA-vector group, (ii) EP⁺ + sham + siRNA-vector group, (iii) EP⁺ + sham + siRNA-HSP70 group, (iv) EP⁻ + TBI + siRNA-vector group, (v) EP⁺ + TBI + siRNA-vector group, and (vi) EP⁺ + TBI + siRNA-HSP70 group.

In experiment 6, the correlation between IL-6 and HSP70 in the peripheral blood and brain tissues (evaluated by qPCR or ELISA and Western blot), NF-κB transcription activation (evaluated by ChIP), and synapsin I (evaluated by Western blot) in six groups as follows: (i) EP⁻ + sham + siRNA-vector group, (ii) EP⁺ + sham + siRNA-vector group, (iii) EP⁺ + sham + siRNA-HSP70 group, (iv) EP⁻ + TBI + siRNA-vector group, (v) EP⁺ + TBI + siRNA-vector group, and (vi) EP⁺ + TBI + siRNA-HSP70 group.

We purchased male Wistar rats (300–320 g) from BioLASCO Taiwan Co., Ltd. (Taipei, Taiwan). The BioLASCO’s policies on the care and use of laboratory animals were followed. The Institutional Animal Care and Use Committee of Chi Mei Medical Center approved the experiments (IACUC Approval No. 101122405). All efforts were made to minimize animal suffering and reduce the number of rats used. The rats were housed under controlled laboratory conditions with a 12-h light/dark cycle, a temperature of 22 ± 1 °C, and a humidity of 60–70% for at least 1 week before surgery or drug treatment. We provide enough chow and unlimited fresh drinking water for rats.

The exercise training protocol was implemented according to the procedure described previously [22]. Animals were trained on a treadmill 5 days a week for 3 weeks.

TBI was induced using a fluid percussion injury device on rats placed in a Kopf stereotaxic frame (Kopf Instruments, Tujunga, CA, USA) detailed previously [25]. Each injured or sham-injured rat was housed individually and closely evaluated for behavioral recovery immediately after a TBI.

pSUPER vector (Oligoengine, Seattle, WA, USA) contains H1 RNA polymerase III promoter, which can direct the synthesis of the siRNA-like transcript. The target sequence for HSP70 (Gen Bank Accession No. NM_031971) was chemically synthesized (Tri-I Biotech, Taipei, Taiwan) as complementary oligonucleotides [26]. A BLAST search of the Rattus genome database was done to ensure that the sequence did not target another gene transcript (http://​blast.​ncbi.​nlm.​nih.​gov/​Blast.​cgi). The clone HSP70 target sequence was sequence confirmed using a DNA sequencer as shown in Chang et al. [22]. Three days before a TBI or a sham operation, rats were anesthetized using sodium pentobarbital (25 mg/kg, i.p.) (Sigma-Aldrich, St. Louis, MO, USA) and placed in a Kopf stereotaxic frame. An acute dose of pSUPER plasmid expression HSP70 siRNA (siRNA-HSP70) (12.5 μg/rat) in 25 μl of pSUPER RNAi delivery media (siRNA-vector) was microinjected into the frontal cortex at 0.5 μl/min flow rate using a microinfusion pump (CMA 100, Carnegie Medicine AB, Stockholm, Sweden) according to the coordinates of the atlas of Paxinos and Watson [27]: from bregma, anteroposterior (AP) −3 mm, lateral (L) 4 mm, and dorsoventral (DV) 2 mm. After the injection was completed, the cannula was left in place for 5 min and then removed at 1 mm/min.

The acute neurological injury was assessed in all rats the day before and 1–3 days after surgery using a modified neurological severity score, a composite of the motor, sensory, and reflex test scores [28]. The higher of the score was given, the more severe of the injury. The degrees of the inclined plane in which the rats can hold were defined as the extents of limb muscle strength [29]. The angle of the inclined plane was increased or decreased in 5^o increments to determine the maximal angle in which a rat could hold to the plane. The data for each were the mean of left- and right-side maximal angle. Neurological and motor functions were examined and scored blindly.

The triphenyl tetrazolium chloride (TTC) (Sigma-Aldrich, St. Louis, MO, USA) staining procedures are used for determining cerebral ischemia extent caused by a TBI [30]. An image analysis system (Image Pro Plus 4.50.29, Media Cybernetics, Inc., Silver Spring, MD, USA) measured the volume of the contusion as revealed by negative TTC stains. The volume of infarction was calculated as 1 mm (thickness of the slice) × [sum of the infarction area in all brain slices (mm²)].

Evans Blue dye (Sigma-Aldrich, St. Louis, MO, USA) was administered intravenously at 3 days after the onset of a TBI. After the dye had circulated 2 h, under deep anesthesia, the rats were transcardially perfused with physiological saline containing 10 U/ml of heparin. Brains were removed for determination of Evans Blue extravasation or brain water contents according to the methods described previously [31].

At predetermined time points after a TBI, the rats were anesthetized with sodium pentobarbital and perfused via cardiac puncture with 0.1 M of phosphate buffered saline (PBS; pH 7.4). Western blotting of the brain tissues samples was done 3 days after a TBI. The methods used for determination of brain expression of HSP70 (1:1000; Enzo Life Sciences, Farmingdale, NY, USA), synapsin I (1:4000; GeneTex, Inc., San Antonio, TX, USA), and β-actin (1:4000; Santa Cruz Biotechnology, Santa Cruz, CA, USA) were modified from our previous study [22].

Rats were perfused with paraformaldehyde (Sigma-Aldrich, St. Louis, MO, USA) solution in PBS (pH 7.4) at 3 days following a TBI, and brains were removed and fixed with 10% buffered formalin (Sigma-Aldrich, St. Louis, MO, USA). Ten-micrometer sections were cut from paraffin-embedded tissue blocks. The sections were xylene- and ethanol-treated for deparaffinization and dehydration. Triple immunofluorescence was performed using the neuronal marker anti-NeuN (1:100, Abcam, Cambridge, UK) and 4′,6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich, St. Louis, MO, USA) in combination with terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) (Roche Inc., Mannheim, Germany), as previously described [30, 32]. The sections were coverslipped with the mounting medium (DAKO’s fluorescent mounting medium; Agilent Technologies, Denmark). For negative control sections, all the procedures were without the primary antibody. Images were made with a Carl Zeiss upright fluorescence microscope (Carl Zeiss, Jena, Germany) using Plan-Apochromat 63×/1.4 oil and Plan-Neofluar 40×/1.30 oil objectives and analyzed with an Axiovision image analysis software (Carl Zeiss, Jena, Germany). The NeuN/DAPI/TUNEL triple-labeled cells were calculated in five coronal sections from each rat and counted in at least six rats per group and expressed as the mean number of cells per section.

For messenger RNA (mRNA) analysis, we killed the animals with an overdose of sodium pentobarbital and collected the blood samples via cardiac puncture. Blood was collected with heparinized RNA protect animal blood tubes (Qiagen, Valencia, CA, USA) to prevent the blood from coagulating. Whole blood was separated into supernatant and pellet fractions within 30 min after blood was collected. Then pellet fractions were discarded, and supernatant were stored at −80 °C. RNA was isolated using Trizol reagent (Invitrogen, Grand Island, NY, USA) and manufacturer’s recommended protocol. Total blood RNA was subsequently treated with DNase I (Qiagen) to remove any traces of contaminating DNA and further purified using an RNeasy Protect Animal Blood Kit (Qiagen). Total RNA concentration was determined by a spectrophotometric optical density measurement (260 nm [OD260] and 280 nm [OD280]). For each sample tested, the ratio between the spectrophotometric readings at 260 and 280 nm (OD260/OD280) in all samples ranged between 1.8 and 2.2. One microgram (1 ug) of high-quality total RNA (RNA integrity number (RIN) >7) was then reverse transcribed using the First Strand Synthesis kit (Qiagen) and reverse transcriptase with reverse transcription cocktail (Molecular Genetic Resources, Tampa, FL, USA) to obtain cDNA and subsequently loaded onto a RT² Profiler^TM PCR Array Rat Cytokines & Chemokines profiler array according to manufacturer’s instructions (Qiagen). Quantitative real-time polymerase chain reaction (qPCR; with gene-specific primers and SYBR green assay) was performed to quantify the genes encoding 84 cytokines and chemokines. QPCR was performed in triplicate in six separate experiments on an Applied Biosystems 7500 system (Applied Biosystems, Foster City, CA, USA). Qiagen’s online web analysis tool (http://​www.​qiagen.​com/​us/​shop/​genes-and-pathways/​data-analysis-center-overview-page/​) was utilized to produce comparative cluster gram, and fold change was calculated by determining the ratio of mRNA levels to control values using the ΔCt method (2^−ΔΔCt). The RT² Profiler™ PCR Array Rat Cytokines & Chemokines profiles the expression of five sets of commonly used housekeeping genes. The five sets of housekeeping genes used in our assessing include B2m (beta-2 microglobulin, a cytoskeletal protein), Hprt1 (hypoxanthine phosphoribosyltransferase 1, involved in the metabolic salvage of purines), Rplp1 (ribosomal protein large P1, involved in structural constituent of ribosome), Ldha (lactate dehydrogenase A, involved in catalytic activity), and Actb (actin beta, a cytoskeletal structure protein). This commercialized array kit can be easily used to identify genes with a constant level of expression among our different experimental conditions for use in normalizing our 84 relative gene expression profiling experiment. All data from EP⁻ + sham + siRNA-vector group was normalized to 100%. The PCR conditions used hold for 10 min at 95 °C, followed by 40 cycles of 15 s at 95 °C, 40 s at 55 °C, and 30 s at 72 °C.

Lists of differentially expressed genes (DEG) from qPCR analysis were generated using a cutoff >2 fold between EP⁻ + TBI + siRNA-vector vs. EP⁺ + TBI + siRNA-vector groups and used as the input for QIAGEN’s Ingenuity® Pathway Analysis (IPA®, Redwood City, CA, USA; https://​www.​qiagenbioinforma​tics.​com/​products/​ingenuity-pathway-analysis/​), and the top five canonical pathways were observed. Canonical pathways that were enriched in the DEG datasets were determined. From the IPA library, we identified the canonical molecular pathways that were most significant to the dataset. We further analyzed the genes from the dataset that were related to canonical pathways.

IL-6 levels of animals were measured by enzyme-linked immunosorbent assay (ELISA) (BD Biosciences, San Jose, CA, USA) at 3 days after a TBI, an exercise preconditioning or an HSP70 siRNA injection. The frozen lesioned side of rat’s cortex was mechanically homogenized and centrifuged at 12,000 rpm for 10 min at 4 °C. The cerebral levels of inflammatory cytokines were quantified using specific ELISA kits for rats according to the manufacturers’ instructions. The inflammatory mediators were expressed as picogram per milligram protein (pg/mg protein).

Chromatin immunoprecipitation (ChIP) assay (EpiTect ChIP qPCR Assay kit; Qiagen) was performed to evaluate the extent of NF-κB binding to the DNA elements in the IL-6 promoter regions respectively using “EpiTect ChIP qPCR Assays kit” from Qiagen. ChIP assays were done as previously described [33]. Ipsilateral lesion side cortical tissue was exposed to 1% formaldehyde to cross-link with DNA proteins. We added glycine to each sample tube to quench unreacted formaldehyde. Cortical tissue was broken open with a sodium dodecyl sulfate (SDS) lysis buffer containing Protease Inhibitor Cocktail II (Calbiochem: Merck Millipore, Darmstadt, Germany). We sheared the chromatin to a manageable size by using a sonicator. Generally, between 200 and 1000 bp of DNA is used to achieve a high degree of resolution during the detection step. We added Protein G Agarose (Millipore, Billerica, MA) to each IP tube to remove proteins or DNA. We used an anti-NF-κB Ab (Millipore) or rabbit IgG (Millipore) to immunoprecipitate the precleared chromatin. It was then incubated overnight at 4 °C while being rotated. Protein G Agarose was added to each IP tube and incubated for 1 h at 4 °C with rotation to collect the antibody/antigen/DNA complex. After four consecutive washes using four different wash buffers, the DNA from the DNA-protein complexes from all the samples, including the input and negative control, was reverse cross-linked by incubation with 2 μL of Proteinase K for 2 h at 65 °C. DNA was purified to remove chromatin proteins and to prepare it for the detection step. Primers specific for IL-6 promoters were used to determine the extents of both immunoprecipitated DNA and quantitative PCR. The following IL-6 primers were used: forward, 5′-GCG ATG GAG TCA GAG GAA AC-3′, and reverse 5′-TGA GGC TAG CGC TAA GAA GC-3′. Results are presented as a percentage of input.

Data are mean ± standard deviation (SD) and were analyzed using one-way analysis of variance (ANOVA) with Fisher’s post hoc test where appropriate. Analyses for behavioral variables used Student’s unpaired t test to compare variables between groups. Bonferroni’s multiple comparison test was then used when appropriate to determine post hoc significance at individual time points. Sigma Plot version 12.0 software (Systat Software Inc., San Jose, CA, USA) was used to analyze all data. A p value of less than 0.05 was considered to be significant.

As shown in previous studies [20‐22], HSP70 expression was significantly higher in ipsilateral brain regions including the frontal cortex (Fig. 1a), hippocampus (Fig. 1b), striatum (Fig. 1c), and hypothalamus (Fig. 1d) in both the (EP⁺ + TBI + siRNA-vector) and the (EP⁺ + sham + siRNA-vector) rats evaluated 3 days post-injury. In addition, the increased cerebral HSP70 expression caused by EP⁺ were significantly attenuated by an acute dose of pSUPER plasmid expression of HSP70 small interfering RNA (siRNA-HSP70) (5 μg/rat in 5 μl of pSUPER RNAi delivery media [siRNA-vector]) into the injured cortex to induce gene silencing [21, 26, 34]. Western blotting analyses revealed that such an intracerebral injection of siRNA-HSP70, but not siRNA-vector, significantly reduced HSP70 expression in the ipsilateral brain tissue of EP⁺ + TBI + siRNA-HSP70 and EP⁺ + sham + siRNA-HSP70 rats (Fig. 1).

All the neurological motor deficits (Fig. 2a, b), brain contusions (Fig. 2c, d), brain edema (evidenced by increased Evans Blue extravasation and increased brain water contents) (Fig. 3a–c), and neuronal loss and apoptosis (evidenced by decreased numbers of neurons and increased numbers of apoptotic neurons respectively) (Fig. 4a–f) 3 days post-injury were all significantly attenuated in EP⁺ rats. In addition, all of the TBI-induced histopathological outcomes and neurological motor deficits in rats treated with HSP70 siRNA, but not siRNA-vector, were significantly attenuated (Figs. 2, 3, and 4). Our data provide first evidence showing the HSP-70-mediated EP protect against traumatic brain injury. Several previous investigations support the present results. For example, a single oral dose of geranylgeranylacetone upregulated brain levels of HSP70 and attenuated kainic acid-induced cognitive, affective, and sensorimotor function deficits in rats [35, 36]. In fact, a promising treatment strategy is that of preconditioning [37].

We used a qPCR-based array to analyze 84 mRNAs expression levels in the peripheral blood at 3 days post-TBI (Fig. 5): 23 mRNAs were upregulated (Table 2). A comparison of mRNA expression profiles showed that 9 mRNAs were significantly upregulated after a TBI and were driving pro-inflammatory and neurodegenerative processes (Table 3): Cxcl10 (a promoter of detrimental effects of TBI), IL-18 (a proinflammatory cytokine), IL-16 (a proinflammatory cytokine), Cd70 (a proinflammatory cytokine), Mif (a suppressor of anti-inflammatory effects), Ppbp (a promoter of leukocyte migration), Ltb (a product of endotoxin), Tnfrsf 11b (also known as osteoprotegerin, an activator of tumor necrosis factor receptor), and Faslg (a promoter of apoptosis) (Tables 2 and 3 and Fig. 6). Moreover, 14 mRNAs driving anti-inflammatory and/or neuroregenerative events (Table 3) are also upregulated: Ccl 19, (a neuroprotective agent), Ccl 3 (an agent against microglial neurotoxicity), Cxcl 19 (an agent that induces the M1 type of macrophages or microglia into the M2 type), IL-10 (a product of M2 macrophages or microglia), IL-22 (a member of the IL-10-related cytokine), IL-6 (a product of the M2 macrophages or microglia), Bmp 6 (an inhibitor of apoptosis), Ccl 22 (a product of M2 macrophages or microglia), IL-1γn (an inhibitor of IL-1α and IL-1β), Ccl 2 (an inducer of angiogenesis), IL-7 (an anti-apoptotic agent), Bmp 7 (a neuroregenerative agent), GPi (a neurotrophic factor), and Ccl17 (also known as thymus- and activation-regulated chemokine (TARC), an anti-microbial agent). Six out of the 9 genes driving pro-inflammatory and/or neurodegenerative processes are significantly inhibited in EP⁺ + TBI rats but not in EP⁻ + TBI rats: Cxcl 10, IL-18, IL-16, Cd70, Mif, and Faslg (Table 2). In contrast, 4 out of 14 genes driving anti-inflammatory and/or neuroregenerative events after a TBI were significantly augmented in EP⁺ + TBI rats but not in EP⁻ + TBI rats: IL-10, IL-22, IL-6, and Bmp 6 (Table 2). The effects of the 6 genes driving pro-inflammatory and/or neurodegenerative processes and the 4 genes driving anti-inflammatory and/or neuroregenerative events are significantly reversed in HSP-70 gene silence rats (Tables 2 and 3 and Fig. 6). The residual 13 mRNAs upregulated after TBI were not significantly affected by EP (Table 2).

Figure 7a depicts a positive Pearson correlation between the blood levels of IL-6 (measured by qPCR assay) and blood levels of HSP70 (measured by Western blotting assay). In the ipsilateral brain regions, a positive Pearson correlation also exists between cortical IL-6 (measured by ELISA) and cortical HSP70 (measured by Western blotting assay) (Fig. 7b), hippocampal IL-6 and hippocampal HSP70 (Fig. 7c), striatal IL-6 and striatal HSP70 (Fig. 7d), and hypothalamic IL-6 and hypothalamic HSP70 (Fig. 7e). The overexpression of NF-κB binding to the DNA elements in the IL-6 promoter regions in the ipsilateral frontal cortex after an EP in both TBI rats and sham-operated rats are significantly attenuated by prior depletion of cortical HSP70 with gene silence (Fig. 8a, b). Again, the overexpression of synapsin I in the ipsilateral brain regions including the frontal cortex (Fig. 9a), hippocampus (Fig. 9b), striatum (Fig. 9c), and hypothalamus (Fig. 9d) caused by EP in both TBI rats and sham-operated rats are significantly reduced by depleting cortical levels of HSP70 with gene silence (Fig. 9). According to the IPA core analysis-based network of mRNA interaction [19], a critical HSP70/NF-κB/IL-6/synapsin I signaling system in the damaged brain in rats can be confirmed by our present results.