Brain Research 1663 (2017) 9–19

Contents lists available at ScienceDirect

Brain Research journal homepage: www.elsevier.com/locate/bres

Research report

Role of caveolin-1/vascular endothelial growth factor pathway in basic fibroblast growth factor-induced angiogenesis and neurogenesis after treadmill training following focal cerebral ischemia in rats Qiongyi Pang 1, Huimei Zhang 1, Zhenzhen Chen 1, Yudan Wu 1, Min Bai 1, Yidian Liu 1, Yun Zhao 1, Fengxia Tu 1, Chan Liu 1, Xiang Chen 1,⇑ Physical Medicine and Rehabilitation Center, The Second Affiliated Hospital and Yuying Children’s Hospital of Wenzhou Medical University, No. 109, Xueyuanxi Road, Wenzhou 325027, Zhejiang, China

a r t i c l e

i n f o

Article history: Received 14 November 2016 Received in revised form 7 March 2017 Accepted 8 March 2017 Available online 12 March 2017 Keywords: MCAO Treadmill bFGF shRNA Angiogenesis Neurogenesis

a b s t r a c t Exercise is known to aid functional recovery following ischemia, though the mechanisms responsible for the beneficial effects of exercise on recovery from ischemic stroke are not fully understood. Basic fibroblast growth factor (bFGF) contributes to angiogenesis and promotes neurologic functional recovery after stroke. The present study aimed to investigate the possible mechanisms whereby treadmill exercise ameliorated impaired angiogenesis and neurogenesis following transient cerebral ischemia in middle cerebral artery occlusion (MCAO) rats. Treadmill exercise was started 2 days after ischemia-reperfusion in MCAO rats and continued until 7 or 28 days after MCAO, after which the animals were sacrificed. Changes in neurological deficit, infarction volume, neuronal morphology, expression levels of bFGF, caveolin-1, and vascular endothelial growth factor (VEGF), and angiogenesis and neurogenesis in the ischemic penumbra were examined by reverse transcription-polymerase chain reaction, western blots, and/or double immunofluorescence. The results suggested that treadmill exercise promoted the expression of bFGF, improved neurological recovery, and reduced infarct volume compared with non-exercised rats, and also enhanced the expression of caveolin-1, VEGF, VEGF receptor 2(FIK-1)/CD34, and Brdu/nestin staining. Small interfering RNA targeting bFGF blocked the protective effects of bFGF. In addition, 4 weeks of post-stroke recovery still ameliorated ischemia-induced damage without bFGF shRNA. These findings suggest a novel mechanism underlying the beneficial effects of bFGF following stroke, and indicate that treadmill exercise may aid stroke recovery by regulating the caveolin-1/VEGF pathway in the ischemic zone. Ó 2017 Published by Elsevier B.V.

1. Introduction Strokes account for about 55% of neurological diseases and are considered to be the leading cause of permanent physical and Abbreviations: FGF, fibroblast growth factor; bFGF, basic FGF; MCAO, middle cerebral artery occlusion; VEGF, vascular endothelial growth factor; NVU, neurovascular unit; shRNA, small hairpin RNA; RNAi, RNA interference; S, shamoperation rats; M7 and M28, focal cerebral ischemia rats; EM7 and EM28, treadmill training after focal cerebral ischemia; siM7 and siM28, lateral ventricle injection of lentivirus-mediated bFGF shRNA before focal cerebral ischemia; siEM7 and siEM28, lateral ventricle injection of lentivirus-mediated bFGF shRNA and treadmill training after focal cerebral ischemia; NCEM7 and NCEM28, lateral ventricle injection of lentivirus-mediated negative control shRNA and treadmill training after focal cerebral ischemia; ECA, external carotid artery; Brdu, 5-bromo-20 -deoxyuridine. ⇑ Corresponding author. E-mail address: [email protected] (X. Chen). 1 All authors contributed equally to this work http://dx.doi.org/10.1016/j.brainres.2017.03.012 0006-8993/Ó 2017 Published by Elsevier B.V.

mental disability (Posada-Duque et al., 2014). However, therapeutic measures for ischemic stroke remain limited, and the only drug currently approved for stroke, recombinant tissue plasminogen activator, has side effects including increased bleeding risk and limited ‘golden time’ (Hosseini et al., 2015). In addition, blood reperfusion after thrombolysis may trigger oxidative and inflammatory events, resulting in ischemia-reperfusion injury to the brain (Rhim et al., 2013). Although our understanding of the molecular basis of the cerebral ischemia cascade has improved, the findings have not been successfully translated into clinical neuroprotective strategies, at least partly because of inappropriate target selection. Shen et al. suggested that growth factors may be an appropriate therapeutic target for ischemic brain disease (Shen et al., 2013). Fibroblast growth factors (FGFs) are involved in many biological processes, including angiogenesis, embryogenesis, differentiation,

10

Q. Pang et al. / Brain Research 1663 (2017) 9–19

and proliferation (Galzie et al., 1997), and have demonstrated survival-promoting and protective effects on brain neurons, and promotion of neural outgrowth, synapse formation, and neurogenesis in the brain (Rai et al., 2007). Basic FGF (bFGF) has been shown to contribute to functional neurologic recovery not only by increasing dendritic length and spine density (Nemati and Kolb, 2011), but also by driving the proliferation of neural stem cells to differentiate into neurons and astrocytes after ischemic brain injury in rats (Jinqiao et al., 2009). Recent research into cerebrovascular disease has focused on interactions among neurons, pericytes, astrocytes, and the extracellular matrix, leading to adoption of the term ‘neurovascular unit’ (NVU). However, the role of bFGF in the NVU remains unknown. Vascular endothelial growth factor (VEGF) is considered to have protective effects against stroke. VEGF was shown to participate in neuronal survival, increase microvascular density, and promote glial proliferation and migration after cerebral ischemia (Sanchez et al., 2010). Intimate cross-talk exists between bFGF and VEGF family members during angiogenesis, and vasculogenesis (Ito et al., 2013) (Yu et al., 2016). Previous studies demonstrated elevated expression levels of bFGF and VEGF during cerebral ischemia and reperfusion in rats (Wang et al., 2008), and several experiments have shown that both endothelial and exogenous bFGF modulate VEGF expression in endothelial cells (Seghezzi et al., 1998). Interestingly, recent reports also demonstrated that expression of FGF receptor (FGFR) 1 or FGFR2 on glioma cells decreased tumor vascularization in parallel with VEGF down-regulation (Auguste et al., 2001). Overall, this evidence suggests that bFGF may induce neovascularization indirectly via activation of VEGF. Caveolins and caveolae are novel pathologically activated factors that assist thrombolysis and neurorestoration in ischemic stroke (Xu et al., 2015). Caveolin-1, a major structural protein of caveolae, is known to be involved in vesicular trafficking, endocytosis, and signal transduction (Lisanti et al., 1994), and is present in different cell types in the central nervous system, especially in the NVU. Treatment of spinal cord injury model rats with bFGF maintained the blood–spinal cord barrier integrity associated with expression of caveolin-1, while introduction of caveolin-1 small hairpin RNA (shRNA) into brain microvascular endothelial cells eliminated the protective effect of bFGF under anaerobic conditions (Ye et al., 2016). We previously demonstrated that treadmill exercise activated the caveolin-1/VEGF signaling pathway to enhance angiogenesis in MCAO rats (Gao et al., 2014). These results suggest that bFGF may be a critical mediator of the caveolin-1/ VEGF pathway after MCAO. Therapeutic exercise is the most common mode of rehabilitation and can significantly reduce recurrence risk and minimize the severity of functional damage after stroke (Middleton et al., 2013). Physical exercise is associated with enhanced neurotropism and growth factor expression, synaptogenesis, neurogenesis and angiogenesis (Ward, 2004; van Praag et al., 2005). Niwa et al. (2016) found that voluntary wheel running enhanced the expression of bFGF, correlated with the proliferation and differentiation

Table 1 Results of the neurological scores in each group after MCAO (score). Groups

1 day

7 days

28 days

S M EM siM siEM NCEM

0 2.5 ± 0.58 2.25 ± 0.50 2.75 ± 0.50 2.75 ± 0.50 2.5 ± 0.58

0 2.17 ± 0.41 0.83 ± 0.75a 2.27 ± 0.47a 1.31 ± 0.48b 1.25 ± 0.50

0 1.67 ± 0.55 0.50 ± 0.55a 1.67 ± 0.75a 0.67 ± 0.52b 0.75 ± 0.50

Data were presented as mean ± SD, T-test, a: p < 0.05 as compared to the same period of M, b: p < 0.05 as compared to the same period of siM.

of hypothalamic neurons in stroke-prone spontaneously hypertensive rats. RNA interference (RNAi) has recently become widely used as a gene-silencing tool with high specificity and efficiency ，and has also emerged as an efficient approach for treating focal ischemic brain injury in rats (Hu et al., 2011). In the present study, we therefore transfected lentivirus-mediated shRNA against bFGF to elucidate the role of the caveolin-1/VEGF pathway in bFGF-enhanced angiogenesis and neurogenesis, and to determine the correlation between the beneficial effects of treadmill exercise and expression of bFGF. 2. Results 2.1. Treadmill training attenuated the neurological deficit score and decreased infarct size, while bFGF shRNA increased infarct injury There was a significant difference among the groups in terms of neurological function assessed by Zea Longa scores after ischemic injury (p < 0.05; Table 1 and Fig. 1e). Rats exposed to ischemia followed by exercise for 7 or 28 days after ischemic injury (EM group) had significantly better (lower) scores than rats without exercise (M group) (p < 0.05). Scores in rats injected in the lateral ventricle with lentivirus carrying shRNA against bFGF prior to ischemia (siM group) were higher at 7 or 28 days after exercise compared with the equivalent M groups. Staining with 2,3,5triphenyltetrazolium chloride showed that the EM groups had smaller infarct volumes while the siM groups had larger infarct volumes compared with the M groups (p < 0.05; Table 2 and Fig. 1a, b). Neuronal morphology of the rat brain was observed after MCAO by hematoxylin and eosin (HE) staining (Fig. 1c). No intact neurons were present in brain slices from the control groups, while the infarct site was observed in the ischemic area in the other groups. Brain tissues from rats in the EM groups showed a more regular arrangement than tissues from rats in the M groups. Changes in the siM groups were more severe compared with the M groups, with multiple vacuolated interspaces and dead neurons, correlated with significantly poorer motor function. 2.2. Treadmill training increased protein expression levels of bFGF, caveolin-1, and VEGF in the ischemic penumbra We examined the effect of treadmill training on protein levels by subjecting MCAO rats to treadmill exercise for 20 m/min, 30 min/day, 5 days/week. Western blot analysis 7 or 28 days after ischemic injury showed that expression levels of bFGF, caveolin, and VEGF were significantly improved in the EM groups compared with the M groups. These results indicated that treadmill training increased the expression of ischemia-induced bFGF, caveolin, and VEGF in the ischemic penumbra (p < 0.05; Fig. 5). 2.3. bFGF shRNA blocked the caveolin-1/VEGF pathway in the ischemic penumbra Although we previously demonstrated that the caveolin-1/VEGF pathway played a critical role in the ischemic zone, the relationship between bFGF and this pathway remains unclear. To investigate the hypothesis that the effect of bFGF is mediated via the caveolin-1/VEGF pathway, we transfected bFGF shRNA into rat brains using a lentivirus vector. After 3 days of virus injection rats were sacrificed. The GFP fluorescence showed the position of lateral ventricle was correct and the percentage of GFP fluorescence following virus injection of bFGF shRNA was about 37%. (Fig. 6) The level of bFGF knockdown was assessed by comparing bFGF mRNA expression levels following transfection with bFGF and neg-

11

Q. Pang et al. / Brain Research 1663 (2017) 9–19

ative control shRNA. Negative control shRNA had no effect on bFGF mRNA expression, while bFGF shRNA resulted in a knockdown rate of about 50% (Fig. 4c, d). bFGF shRNA also decreased the total expression levels of caveolin-1 and VEGF (Fig. 5). These results demonstrated that the caveolin-1/VEGF pathway was partly mediated by bFGF. 2.4. Treadmill training improved neural regeneration and enhanced angiogenesis, which effects were inhibited by bFGF shRNA, demonstrated by immunofluorescence The NVU consists of the endothelium, glia, neurons, pericytes, and basal lamina (Iadecola, 2004), which perform a coordinated response after cerebral ischemia, thus helping to establish and maintain blood flow and thereby exerting a neuroprotective effect. We investigated the interactions between high levels of bFGF after

Table 2 Results of infarct volume in each group after MCAO (%). Groups

7 days

28 days

S M EM siM siEM NCEM

0 49.48 ± 0.78 37.41 ± 1.91a 63.51 ± 1.84a 53.62 ± 8.70b 43.15 ± 6.86c

0 47.30 ± 4.46 39.90 ± 1.53 58.53 ± 7.73a 48.13 ± 5.23 46.01 ± 4.45

Data were presented as mean ± SD, ANOVA, a: p < 0.05 as compared to the same period of M, b: p < 0.05 as compared to the same period of siM, c: p < 0.05 as compared to the same period of siEM.

MCAO and components of the NVU in the pathophysiology of cerebral ischemia and recovery following therapy. Brain microvessels were labeled by co-staining for VEGFR2(FILK-1)/CD34 (Fig. 2).

Fig. 1. a, b Photomicrographs of ischemic lesion. c Effects of exercise and bFGF shRNA on HE staining (100 magnification) after ischemia. d Percentage of infarct volume in each group (mean ± SD, n = 5 for each group). e Neurological scores in different groups.

12

Q. Pang et al. / Brain Research 1663 (2017) 9–19

Quantitative analysis of immune-labeled vessels per unit area (Fig. 4a) showed intensive labeling in EM rats, compared with sparse labeling in the same regions in rats in the M groups. siM rats showed fewer vessels compared with the same regions in the M groups. However, there was no significant difference among M28, EM28, siM28, siEM28, NCEM28. Neuronal precursors were labeled by injection of 5-bromo-2-deoxyuridine (Brdu), and Brdu/nestinexpressing cells (Fig. 3) in the ischemic penumbra were quantified by epifluorescence. Brdu/nestin-labeled neuronal precursor cells were more abundant in the EM groups compared with the same regions in the M groups, while fewer labeled cells were detected

in siM rats compared with the same regions in M rats (Fig. 4b). In summary, treadmill training improved neural regeneration and enhanced angiogenesis, while bFGF shRNA inhibited the increase in microvessels and neuronal progenitor cells. 2.5. Effects of caveolin-1/VEGF pathway on angiogenesis and neurogenesis after MCAO may be induced by bFGF after treadmill training Silencing of bFGF significantly inhibited total caveolin-1 and VEGF in siM compared with M rats (Fig. 5). Furthermore,

Fig. 2. FIK-1/CD34 staining in the ischemic penumbra zone. Red: FIK-1; green: CD34; blue: DAPI; merge: FIK-1/CD34. Arrows indicate colored endothelial progenitor cells. Scale bar, 20 lm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Q. Pang et al. / Brain Research 1663 (2017) 9–19

caveolin-1 and VEGF expression in the siM groups were significantly reduced compared with the siEM groups. The relationship between bFGF expression and the effects of the caveolin-1/VEGF pathway on angiogenesis and neurogenesis were confirmed by immunofluorescence. These results further suggest that the effects of the caveolin-1/VEGF pathway on angiogenesis and neurogenesis may be induced by bFGF in the penumbra after treadmill training. We isolated total protein from the ischemic penumbra of the same brains at 7 or 28 days after MCAO. bFGF, caveolin-1, and VEGF levels peaked at 7 days and then declined at 28 days in the EM

13

compared with the M groups (Fig. 5). However, total protein levels at 28 days were still higher compared with the sham-operated (S) groups, demonstrating that 4 weeks of post-stroke recovery still ameliorated ischemia-induced damage.

3. Discussion The endogenous neurovascular pathways of cerebral ischemic cascades remain poorly understood, despite their role in improving

Fig. 3. Brdu/nestin staining in the ischemic penumbra zone. Red: Brdu; green: nestin; blue: DAPI; merge: Brdu/nestin. Arrows indicate colored neuronal precursor cells. Scale bar, 20 lm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

14

Q. Pang et al. / Brain Research 1663 (2017) 9–19

long-term recovery. Treadmill running has been shown to reduce brain infarction, improve angiogenesis, and induce neurogenesis (Yang et al., 2012; Matsuda et al., 2011). bFGF has also demonstrated protective effects against ischemic brain injury(Jin-qiao et al., 2009), though the mechanisms responsible for this are poorly understood. Previous studies demonstrated that caveolin-1 interacted with FGFR1 function as a platform for regulating bFGFinduced angiogenesis in ovine fetoplacental artery endothelial cells (Feng et al., 2012) and in the recovery of spinal cord injury (Ye et al., 2016). The current study also indicated a role for the caveolin-1/VEGF pathway in treadmill-mediated angiogenesis and neurogenesis, possibly induced by bFGF, thus providing a potential explanation for the mechanisms following focal cerebral stroke. Intravenous administration of bFGF within hours after stroke reduced infarct volume in some animal experiments, presumably as a consequence of a direct protective effect on cells in the ischemic penumbra during prolonged recovery (Ay et al., 1999). Others have reported that injection of bFGF in adult rats with cerebral ischemic injury resulted in increased cell proliferation and improved functional test results (Wang et al., 2008). bFGF has been shown to bind to its receptors (FGFRs 1–3) and stimulate downstream signaling pathways (e.g. MAPK or PI3-K) to generate cell proliferation, differentiation, and/or migration (Kalluri et al., 2007; Otaegi et al., 2006). Several FGFRs have been localized to vascular sites in a number of tissues (Hughes, 1997), suggesting that

bFGF may regulate vascular tone via its interaction with FGFRs. bFGF was also shown to be a central and peripheral vasodilator, decreasing infarct volume in endothelial cells in nitric oxide synthase-deficient mice without changing regional cerebral blood flow (Cuevas et al., 1991; Huang et al., 1997). In the current study, bFGF activity measured by quantitative reverse transcriptionpolymerase chain reaction (RT-qPCR) and western blot assays peaked at 7 days in rats subjected to treadmill exercise, in line with decreased infarct volumes and improved neurological outcome. bFGF shRNA accordingly abolished the effects of bFGF on functional recovery. We therefore hypothesized that bFGF had positive effects on stroke, which could be promoted by exercise, consistent with previous findings. Cerebral ischemia disrupts the NVU, involving death of neuronal and endothelial cells in the core and penumbra regions. bFGF promoted restoration of the NVU, potentially contributing to its therapeutic benefits after ischemic stroke. The current results indicated that bFGF induced neovascularization directly by activating the caveolin-1/VEGF pathway in the ischemic penumbra. Caveolins may exert different functions in different cell types in the central nervous system, especially in the NVU. bFGF with a caveolae-like microdomain has been reported in human neuroblastoma cells, and induced a compartmentalized signaling response involving the Scr family of tyrosine kinases and other factors, leading to cell proliferation (Davy et al., 2000). Moreover, bFGF was shown to activate sphingomyelinase to synthesize ceramide, which in turn

Fig. 4. a Quantification of FIK-1/CD34 cells in brain sections. The numbers of FIK-1/CD34 cells in the ischemic penumbra zone were counted to indicate blood-vessel density. b Quantification of Brdu/nestin cells in brain sections. The numbers of Brdu/nestin cells in the ischemic penumbra zone were counted to assess neurogenesis. Four slides per rat were counted. c, d RT-qPCR analysis of bFGF gene expression in ischemic penumbra at 7 or 28 days. Data are expressed as mean ± SD. Data analyzed by ANOVA. *p < 0.05, ** p < 0.01, ***p < 0.001, ****p < 0.0001.

Q. Pang et al. / Brain Research 1663 (2017) 9–19

allowed the dissociation of endothelial nitric oxide synthase from caveolin-1 to catalyze the synthesis of nitric oxide, which participated in bFGF proliferative activity (Florio et al., 2003). In terms of angiogenesis, caveolin-1 slowed down cell proliferation and speeded up endothelial differentiation and tube formation in stroke (Xu et al., 2015). VEGFR2 is localized in endothelial caveolae through physical association with caveolin-1 (Labrecque et al., 2003). Downregulation of caveolin-1 reduced VEGF-stimulated phosphorylation of VEGFR2 and subsequently affected downstream signaling factors such as PLCc1, Akt, and ERK1/2 (Tahir et al., 2009; Liao et al., 2009). In terms of neurogenesis, degradation and dissociation of caveolin proteins could induce a series of alterations in prosurvival signaling (e.g. MAPK/ERK pathway, PI3 K/Akt pathway, apoptotic and autophagic pathways, cellular prion protein pathway) (Xu et al., 2015), and could not maintain synapse stabilization and formation, and neuronal regeneration. We previously demonstrated that treadmill exercise promoted functional recovery after cerebral ischemic injury in the penumbra via caveolin-1/VEGF in association with angiogenesis (Gao et al., 2014) and enhanced pro-

15

liferation, migration, and differentiation of subventricular zonederived neural stem cells (Zhao et al., 2016). In the current study, bFGF shRNA inhibited the expression of caveolin-1 and VEGF. We therefore confirmed that treadmill training promoted angiogenesis and neurogenesis through the caveolin-1/VEGF pathway, which was induced by bFGF following focal cerebral ischemia in rats. The current findings suggest that bFGF apatite coating was developed as a slow-releasing drug delivery system can promote intrinsic angiogenic factors and decrease penumbra infarction in cerebral ischemic rats (Ito et al., 2013). Furthermore, intrinsic bFGF controlled angiogenesis after brain infarction by influencing VEGF and chemokines (Seghezzi et al., 1998) (Murakami and Simons, 2008) (Yuan et al., 2005). These findings imply that bFGF plays a key role in angiogenesis. We demonstrated that bFGF influenced microvessel density, as indicated by VEGFR2(FIK-1)/CD34 double immunofluorescent staining, while bFGF shRNA inhibited microvessel density, leading to decreased angiogenesis in the infarct region and increased infarct size. Although VEGFR2(FIK1)/CD34 expression was enhanced until to 28 days, there was no significant difference among M28, EM28, siM28, siEM28, NCEM28.

Fig. 5. a Protein levels of bFGF, caveolin-1, and VEGF in different groups. b Densitometric analyses of bFGF and tubulin. Data analyzed by ANOVA. *p < 0.05, **p < 0.01, *** p < 0.001， ****p < 0.0001. c Densitometric analyses of caveolin-1 and tubulin. Data analyzed by ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. d Densitometric analyses of VEGF and tubulin. Data analyzed by ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

16

Q. Pang et al. / Brain Research 1663 (2017) 9–19

Fig. 6. a A representative fluorescence image of GFP (200 magnification and 400 magnification) following lateral ventricle injection of bFGF shRNA. b Site of core and penumbra area. c Quantification of GFP fluorescence following lateral ventricle injection of bFGF shRNA. Data are expressed as mean ± SD.

It is possible that bFGF contributed to an early angiogenic response (Presta et al., 2005) and that pericytes, as a component of the NVU, upregulated platelet-derived growth factor receptor b via bFGF and FGFR to stabilize the newly formed vasculature and thus exert a neuroprotective effect after ischemic stroke (Nakamura et al., 2016). Additionally, the effects on microvessel density were not in line with the increase in Brdu/nestin cells and decreased infarct volume. This may indicate that the caveolin-1/VEGF pathway may be a necessary, but not the only pathway involved in bFGFmediated neuron proliferation. Further studies involving FGFR inhibition are needed to clarify the downstream signaling interactions of bFGF and the NVU. The chronic neurorestorative effects of long-term exercise after ischemic neuronal death have not been thoroughly examined. In the current study, total protein levels were elevated 28 days among M28, EM28, siM28, siEM28, NCEM28 after MCAO compared with sham-operated rats, associated with smaller infarct volume and better functional recovery, indicating that 4 weeks was a suitable period for demonstrating post-stroke recovery. A recent study suggested that a link between mental and physical training skills over 28 days increased brain plasticity via increased neurogenesis in the adolescent rat hippocampus (DiFeo and Shors, 2016). Treadmill exercise started 5 days after ischemia-reperfusion and lasting for 4 weeks facilitated memory recovery and enhanced cell proliferation and neuroblast differentiation in male Mongolian gerbils after stroke (Ahn et al., 2016). The current study had some limitations. Although injection with lentivirus vectors carrying bFGF shRNA specifically blocked bFGF expression, the relative efficacies of bFGF shRNA and bFGF inhibitors have not been studied. Our results showed that bFGF expression in the siM group peaked at 7 days and then decreased at 28 days, and we were unable to

determine if this was due to a limited-duration effect of shRNA. Notably, there are also some anatomic and physiological differences between rats and humans, and the safety of gene transfer would need to be evaluated in studies in larger animals before translation into clinical practice. 4. Conclusions We demonstrated that the caveolin-1/VEGF pathway is involved in bFGF-induced angiogenesis and neurogenesis after treadmill training following focal cerebral ischemia in rats. Evidence for this pathway furthers our understanding of the mechanisms activated following ischemic stroke. However, additional approaches such as a better-adapted treadmill-training protocol or stem cell therapies may be used to optimize this repair strategy. Further studies are needed to investigate the roles of other key factors, such as modulation of caveolin or other proteins by microRNAs, in treadmill-induced recovery after ischemic stroke. 5. Materials and methods 5.1. Animals Adult (8-week-old) Sprague-Dawley male rats weighing 250– 270 g were purchased from Shanghai Laboratory Animal Center (Shanghai, China). All the experimental protocols were approved by the Animal Research Committee of Wenzhou Medical University. The rats were housed in pairs in a temperature-controlled (22 ± 1 °C) laminar flow rack with a 12-h light:dark cycle, with free access to food and water. Rats were divided randomly into six

Q. Pang et al. / Brain Research 1663 (2017) 9–19

groups: sham-operation (n = 15; group S); focal cerebral ischemia (n = 30; 7 and 28 days, groups M7 and M28); focal cerebral ischemia followed by treadmill training (n = 30; 7 and 28 days, groups EM7 and EM28); lentivirus-mediated delivery of bFGF shRNA injected into the lateral ventricle before focal cerebral ischemia (n = 30; 7 and 28 days, groups siM7 and siM28); lentivirusmediated delivery of bFGF shRNA injected into the lateral ventricle before focal cerebral ischemia followed by treadmill training (n = 30; 7 and 28 days, groups siEM7 and siEM28); and lentivirus-mediated delivery of negative control shRNA into the lateral ventricle before focal cerebral ischemia followed by treadmill training (n = 30; 7 and 28 days, groups NCEM7 and NCEM28). All rats were familiarized with treadmill training for 3 days prior to MCAO using a motor-driven treadmill at 8 m/min for 10 min/day. In the exercise groups, treadmill training (0° slope, 20 m/min, 30 min/day, 5 days/week) was started at 2 days post-MCAO and continued until 7 or 28 days. Rats were tested for neurobehavioral deficits and then sacrificed on day 7 or 28 after MCAO.

5.2. Lateral ventricle injection of lentivirus Rats of siM, siEM, MCEM groups were anesthetized with 1.5– 2.0% isoflurane in air. A small borehole was then drilled into the bone 1.8 mm lateral to and 0.8 mm posterior to the bregma over the left hemisphere and a microinfusion pump was stereotactically introduced through the burr hole into the left lateral ventricle (3.8 mm beneath the dura). bFGF shRNA lentivirus and negative control lentivirus (5 ml, 5E + 8TU/mL, Gi Kai Gene Chemical Technology Co., Ltd., Shanghai, China) was infused into the left lateral ventricle at a rate of 1 ll/min and the rats were then allowed to recover for 3 days before MCAO. The thickness of brain was about 10 lm by using a freezing microtome (ThermoFisher Scientific Company, Hampton, NH, USA). Before that, we chose some rats of S group that injection of bFGF shRNA lentivirus to observe wether the lateral ventricle position was correct under fluorescence microscope (BX51, Olympus) and calculate the percentage of GFP fluorescence/4,6-Diamidino-2-phenylindole (DAPI) fluorescence.

5.3. MCAO model of focal brain ischemia MCAO was induced using standard microsurgical techniques. In brief, we isolated the common carotid artery and external carotid artery (ECA) from the surrounding connective tissues. We opened a small nick in the ECA stump and inserted a fine silicon-coated surgical nylon monofilament (0.36 ± 0.02 mm; L3600, Jia Ling Biotechnology Co., Ltd., Guangzhou, China) into the left internal carotid artery lumen through the ECA, and moved it forward to the origin of the middle cerebral artery. The rats were then kept in an incubator for 2 h at 37 °C. Sham-operated animals were treated identically, except for omission of the surgical nylon monofilament.

5.4. Neurological evaluation Rats were assessed for neurological deficit at 24 h, 7 days, and 28 days after MCAO by an investigator blinded to the experimental conditions. Neurological evaluations were scored on a five-point scale, based on the Zea Longa scale (Longa et al., 1989): 0, no observable neurologic deficit; 1, failing to extend right forepaw fully; 2, circling to the right; 3, falling to the right; 4, no movement or spontaneous walking and a depressed level of consciousness; and 5, death.

17

5.5. Measurement of infarctvolumes Following perfusion with phosphate-buffered saline to remove blood, rats were euthanized at 7 or 28 days after MCAO. Coronal sections of whole brains were taken at 2 mm intervals and stained with 2% 2,3,5-triphenyltetrazolium chloride. Infarct volumes were quantified using ImagePro-Plus6.0. Because of the cavities in the brain cortex, we calculated the percentage of infarct sizes as (contralateral hemispheric volume  non-infarcted tissue in the lesioned hemisphere)/contralateral hemispheric volume (Swanson et al., 1990). 5.6. Double immunofluorescence staining Rats were anesthetized and their brains were perfused and dehydrated as described previously (Yan et al., 2011). Coronal serial sections were cut using a freezing microtome at a thickness of 10 lm. For VEGFR2(FIK-1)/CD34 staining, the sections were incubated with a mixture of mouse anti-rat VEGFR2 (1:400, ab9530, Abcam, UK) and rabbit anti-rat CD34 (1:200, ab81289, AbcamAbcam) overnight at 4 °C. A goat anti-rabbit antibody conjugated with greenfluorescent protein (GFP) Alexa Fluor 488 (1:100, Yesen, Shanghai, China) and a goat anti-mouse antibody conjugated with red-fluorescent Alexa Fluor 594 (1:400, Yesen) were used to show immunoreactivity. Brdu staining was performed using 2 N HCl for 15 min to denature DNA followed by 0.1 M borate solution (pH 8.5) for 20 min. The sections were then incubated with a mixture of mouse anti-rat Brdu (1:400, ab8152, AbcamAbcam) and rabbit anti-rat nestin (1:50, ab92391,Abcam AbcamAbcom) overnight at 4 °C. A goat anti-rabbit antibody conjugated with green-fluorescent Alexa Fluor 488 (1:50, Yesen) and goat anti-mouse antibody conjugated with red-fluorescent Alexa Fluor 594 (1:500, Yesen) were used to show immunoreactivity. Fluorescent signals were observed under a fluorescence microscope (BX51, Olympus). 5.7. HE staining for neuronal morphology Brain tissue was prepared as described for doubleimmunofluorescence staining (Yan et al., 2011). Brain-tissue blocks containing ischemic penumbra were sectioned at 10 lm, stained with HE. This method adopts usage of hemalum. Hemalum stains nuclei of cells blue. Then, an aqueous eosin counterstains eosinophilic structures in various shades of red, pink and orange. We used an Olympus BH-2 microscope (Olympus Optical, London, UK) to capture image. 5.8. RT-PCR Three different bFGF shRNA sequences were designed: bFGF shRNA1, 50 -GGGCAGTATAAACTCGGAT-30 ; bFGF shRNA2, 50 -GCCT GGAGTCCAATAACTA-30 ; and bFGF shRNA3, 50 -TCAAGG 0 GAGTGTGTGCGAA-3 . The nonsense sequence used as a control was 50 -TTCTCCGAACGTGTCACGT-30 . bFGF shRNA3 resulted in the most effective knockdown of bFGF. Total RNA was isolated using TRIzol reagents (Cat#15596018, ThermoFisher Scientific, Kansas city, MO, USA) according to the manufacturer’s instructions (Invitrogen, USA). cDNA was generated from 0.5 lg RNA as a template using a reverse transcription kit (MJ Mini, Bio-Rad, USA). Reverse transcription products were quantified by real-time PCR (LightCycle 480, Rocheoche, Switzerland) in a final reaction volume of 10 ll using SYBR Green. Primers for bFGF were designed against known rat sequences as follows: forward, 50 -TCCATCAAGG GAGTGTGTGC-30 ; reverse, 50 -TCCGTGACCGGTAAGTGTTG-30 , with tubulin: forward, 50 -CCGCCTGCCTCTTCGTCT-30 ; reverse, 50 -GGTCT

18

Q. Pang et al. / Brain Research 1663 (2017) 9–19

ATGCCATGCTCGTCAC-30 as a housekeeping control. Comparative mRNA expression levels of bFGF were expressed as 2DDCt. 5.9. Western blot Approximately 100 mg of brain tissue was homogenized in icecold RIPA buffer (50 mM Tris buffer, pH 8.0; 150 mM NaCl; 1% NP40; 0.5% deoxycholate; and 0.1% sodium dodecyl sulfate) containing protease inhibitor cocktail (10 ll/ml), using a tissue homogenizer (MasterPrep, Bioer, Hangzhou, China). Insoluble material was precipitated in a mini-centrifuge at 12,000g for 10 min. The supernatant above were analyzed using bicinchoninic acid reagent. Thirty micrograms of protein was applied to a 12% of sodium dodecyl sulfate-polyacrylamide gel and transferred to a polyvinylidene fluoride membrane. Nonspecific sites were blocked with 5% nonfat milk in TBS with 0.05% Tween 20 (TBST) for 2 h. Blots were probed with anti-bFGF (1:500, Sc-79, Santa Cruz, CA, USA), anti-caveolin-1 (1:1000, ab2910, Abcam, anti-VEGF (1:800, ab46154, Abcam, and anti-tubulin (1:1000, AT819, Beyotime, China) antibodies overnight at 4 °C. The membranes were washed three times with TBST and treated with horseradish peroxidase-conjugated secondary antibodies (1:5000, EarthOx, LLC) for 2 h at room temperature. Blots were soaked in SuperSignal chemiluminescent substrate and visualized using a UVP gel-imaging system (Upland, CA, USA). Densitometric analysis was performed using AlphaEaseFC (version 4.0). 5.10. Statistical analysis Data were reported as mean ± SD. Statistical significance was determined using one-way analysis of variance (ANOVA) and Student’s t-tests. Values were considered significant at p < 0.05. Data were analyzed using GraphPad Prism 6.0 statistical software. References Ahn, J.H., Choi, J.H., Park, J.H., et al., 2016. Long-term exercise improves memory deficits via restoration of myelin and microvessel damage, and enhancement of neurogenesis in the aged gerbil hippocampus after ischemic stroke. Neurorehab. Neural Repair. 30 (9), 894–905. http://dx.doi.org/10.1177/ 1545968316638444. Auguste, P., Gursel, D.B., Lemiere, S., et al., 2001. Inhibition of fibroblast growth factor/fibroblast growth factor receptor activity in glioma cells impedes tumor growth by both angiogenesis-dependent and -independent mechanisms. Cancer Res. 61 (4), 1717–1726. Ay, H., Ay, I., Koroshetz, W.J., et al., 1999. Potential usefulness of basic fibroblast growth factor as a treatment for stroke. Cerebrovasc. Dis. (Basel, Switzerland) 9 (3), 131–135. doi:15941. Cuevas, P., Carceller, F., Ortega, S., et al., 1991. Hypotensive activity of fibroblast growth factor. Science (New York, NY) 254 (5035), 1208–1210. Davy, A., Feuerstein, C., Robbins, S.M., 2000. Signaling within a caveolae-like membrane microdomain in human neuroblastoma cells in response to fibroblast growth factor. J. Neurochem. 74 (2), 676–683. DiFeo, G., Shors, T.J., 2016. Mental and physical skill training increases neurogenesis via cell survival in the adolescent hippocampus. Brain Res. http://dx.doi.org/ 10.1016/j.brainres.2016.08.015. Feng, L., Liao, W.X., Luo, Q., et al., 2012. Caveolin-1 orchestrates fibroblast growth factor 2 signaling control of angiogenesis in placental artery endothelial cell caveolae. J. Cell. Physiol. 227 (6), 2480–2491. http://dx.doi.org/10.1002/ jcp.22984. Florio, T., Arena, S., Pattarozzi, A., et al., 2003. Basic fibroblast growth factor activates endothelial nitric-oxide synthase in CHO-K1 cells via the activation of ceramide synthesis. Mol. Pharmacol. 63 (2), 297–310. Galzie, Z., Kinsella, A.R., Smith, J.A., 1997. Fibroblast growth factors and their receptors. Biochem. Cell Biol. Biochim. Biol. Cellu. 75 (6), 669–685. Gao, Y., Zhao, Y., Pan, J., et al., 2014. Treadmill exercise promotes angiogenesis in the ischemic penumbra of rat brains through caveolin-1/VEGF signaling pathways. Brain Res. 1585, 83–90. http://dx.doi.org/10.1016/j.brainres.2014.08.032. Hosseini, S.M., Farahmandnia, M., Kazemi, S., et al., 2015. A novel cell therapy method for recovering after brain stroke in rats. Int. J. Stem Cells 8 (2), 191–199. http://dx.doi.org/10.15283/ijsc.2015.8.2.191. Hu, Q., Chen, C., Khatibi, N.H., et al., 2011. Lentivirus-mediated transfer of MMP-9 shRNA provides neuroprotection following focal ischemic brain injury in rats. Brain Res. 1367, 347–359. http://dx.doi.org/10.1016/j.brainres.2010.10.002.

Huang, Z., Chen, K., Huang, P.L., et al., 1997. BFGF ameliorates focal ischemic injury by blood flow-independent mechanisms in eNOS mutant mice. Am. J. Physiol. 272 (3 Pt 2), H1401–H1405. Hughes, S.E., 1997. Differential expression of the fibroblast growth factor receptor (FGFR) multigene family in normal human adult tissues. J. Histochem. Cytochem. 45 (7), 1005–1019. Iadecola, C., 2004. Neurovascular regulation in the normal brain and in Alzheimer’s disease. Nat. Rev. Neurosci. 5 (5), 347–360. http://dx.doi.org/10.1038/nrn1387. Ito, Y., Tsurushima, H., Sato, M., et al., 2013. Angiogenesis therapy for brain infarction using a slow-releasing drug delivery system for fibroblast growth factor 2. Biochem. Biophys. Res. Commun. 432 (1), 182–187. http://dx.doi.org/ 10.1016/j.bbrc.2013.01.013. Jin-qiao, S., Bin, S., Wen-hao, Z., et al., 2009. Basic fibroblast growth factor stimulates the proliferation and differentiation of neural stem cells in neonatal rats after ischemic brain injury. Brain Dev. 31 (5), 331–340. http://dx.doi.org/ 10.1016/j.braindev.2008.06.005. Kalluri, H.S., Vemuganti, R., Dempsey, R.J., 2007. Mechanism of insulin-like growth factor I-mediated proliferation of adult neural progenitor cells: role of Akt. Eur. J. Neurosci. 25 (4), 1041–1048. http://dx.doi.org/10.1111/j.14609568.2007.05336.x. Labrecque, L., Royal, I., Surprenant, D.S., et al., 2003. Regulation of vascular endothelial growth factor receptor-2 activity by caveolin-1 and plasma membrane cholesterol. Mol. Biol. Cell 14 (1), 334–347. http://dx.doi.org/ 10.1091/mbc.E02-07-0379. Liao, W.X., Feng, L., Zhang, H., et al., 2009. Compartmentalizing VEGF-induced ERK2/ 1 signaling in placental artery endothelial cell caveolae: a paradoxical role of caveolin-1 in placental angiogenesis in vitro (Baltimore, Md). Mol. Endocrinol. 23 (9), 1428–1444. http://dx.doi.org/10.1210/me.2008-0475. Lisanti, M.P., Scherer, P.E., Tang, Z., et al., 1994. Caveolae, caveolin and caveolin-rich membrane domains: a signalling hypothesis. Trends Cell Biol. 4 (7), 231–235. Longa, E.Z., Weinstein, P.R., Carlson, S., et al., 1989. Reversible middle cerebral artery occlusion without craniectomy in rats. Stroke 20 (1), 84–91. Matsuda, F., Sakakima, H., Yoshida, Y., 2011. The effects of early exercise on brain damage and recovery after focal cerebral infarction in rats. Acta Physiol. 201 (2), 275–287. http://dx.doi.org/10.1111/j.1748-1716.2010.02174.x. Middleton, L.E., Corbett, D., Brooks, D., et al., 2013. Physical activity in the prevention of ischemic stroke and improvement of outcomes: a narrative review. Neurosci. Biobehav. Rev. 37 (2), 133–137. http://dx.doi.org/10.1016/j. neubiorev.2012.11.011. Murakami, M., Simons, M., 2008. Fibroblast growth factor regulation of neovascularization. Curr. Opin. Hematol. 15 (3), 215–220. http://dx.doi.org/ 10.1097/MOH.0b013e3282f97d98. Nakamura, K., Arimura, K., Nishimura, A., et al., 2016. Possible involvement of basic FGF in the upregulation of PDGFRbeta in pericytes after ischemic stroke. Brain Res. 1630, 98–108. http://dx.doi.org/10.1016/j.brainres.2015.11.003. Nemati, F., Kolb, B., 2011. FGF-2 induces behavioral recovery after early adolescent injury to the motor cortex of rats. Behav. Brain Res. 225 (1), 184–191. http://dx. doi.org/10.1016/j.bbr.2011.07.023. Niwa, A., Nishibori, M., Hamasaki, S., et al., 2016. Voluntary exercise induces neurogenesis in the hypothalamus and ependymal lining of the third ventricle. Brain Struct. Funct. 221 (3), 1653–1666. http://dx.doi.org/10.1007/s00429-0150995-x. Otaegi, G., Yusta-Boyo, M.J., Vergano-Vera, E., et al., 2006. Modulation of the PI 3kinase-Akt signalling pathway by IGF-I and PTEN regulates the differentiation of neural stem/precursor cells. J. Cell Sci. 119 (Pt 13), 2739–2748. http://dx.doi. org/10.1242/jcs.03012. Posada-Duque, R.A., Barreto, G.E., Cardona-Gomez, G.P., 2014. Protection after stroke: cellular effectors of neurovascular unit integrity. Front. Cell. Neurosci. 8, 231. http://dx.doi.org/10.3389/fncel.2014.00231. Presta, M., Dell’Era, P., Mitola, S., et al., 2005. Fibroblast growth factor/fibroblast growth factor receptor system in angiogenesis. Cytokine Growth Factor Rev. 16 (2), 159–178. http://dx.doi.org/10.1016/j.cytogfr.2005.01.004. Rai, K.S., Hattiangady, B., Shetty, A.K., 2007. Enhanced production and dendritic growth of new dentate granule cells in the middle-aged hippocampus following intracerebroventricular FGF-2 infusions. Eur J. Neurosci. 26 (7), 1765–1779. http://dx.doi.org/10.1111/j.1460-9568.2007.05820.x. Rhim, T., Lee, D.Y., Lee, M., 2013. Drug delivery systems for the treatment of ischemic stroke. Pharm. Res. 30 (10), 2429–2444. http://dx.doi.org/10.1007/ s11095-012-0959-2. Sanchez, A., Wadhwani, S., Grammas, P., 2010. Multiple neurotrophic effects of VEGF on cultured neurons. Neuropeptides 44 (4), 323–331. http://dx.doi.org/ 10.1016/j.npep.2010.04.002. Seghezzi, G., Patel, S., Ren, C.J., et al., 1998. Fibroblast growth factor-2 (FGF-2) induces vascular endothelial growth factor (VEGF) expression in the endothelial cells of forming capillaries: an autocrine mechanism contributing to angiogenesis. J. Cell Biol. 141 (7), 1659–1673. Shen, F., Kuo, R., Milon-Camus, M., et al., 2013. Intravenous delivery of adenoassociated viral vector serotype 9 mediates effective gene expression in ischemic stroke lesion and brain angiogenic foci. Stroke; a journal of cerebral circulation. 44 (1), 252–254. http://dx.doi.org/10.1161/ STROKEAHA.112.662965. Swanson, R.A., Morton, M.T., Tsao-Wu, G., et al., 1990. A semiautomated method for measuring brain infarct volume. J. Cereb. Blood Flow Metab. 10 (2), 290–293. http://dx.doi.org/10.1038/jcbfm.1990.47.

Q. Pang et al. / Brain Research 1663 (2017) 9–19 Tahir, S.A., Park, S., Thompson, T.C., 2009. Caveolin-1 regulates VEGF-stimulated angiogenic activities in prostate cancer and endothelial cells. Cancer Biol. Ther. 8 (23), 2286–2296. van Praag, H., Shubert, T., Zhao, C., et al., 2005. Exercise enhances learning and hippocampal neurogenesis in aged mice. J. Neurosci. 25 (38), 8680–8685. http://dx.doi.org/10.1523/JNEUROSCI.1731-05.2005. Wang, J.H., Liu, N., Du, H.W., et al., 2008. Effects of adipose-derived stem cell transplantation on the angiogenesis and the expression of bFGF and VEGF in the brain post focal cerebral ischemia in rats. Xi bao yu fen zi mian yi xue za zhi Chin. J. Cell. Mol. Immunol. 24 (10), 958–961. Wang, Z.L., Cheng, S.M., Ma, M.M., et al., 2008. Intranasally delivered bFGF enhances neurogenesis in adult rats following cerebral ischemia. Neurosci. Lett. 446 (1), 30–35. http://dx.doi.org/10.1016/j.neulet.2008.09.030. Ward, N.S., 2004. Functional reorganization of the cerebral motor system after stroke. Curr. Opin. Neurol. 17 (6), 725–730. Xu, L., Guo, R., Xie, Y., et al., 2015. Caveolae: molecular insights and therapeutic targets for stroke. Expert Opin. Ther. Targets 19 (5), 633–650. http://dx.doi.org/ 10.1517/14728222.2015.1009446. Yan, W., Zhang, H., Bai, X., et al., 2011. Autophagy activation is involved in neuroprotection induced by hyperbaric oxygen preconditioning against focal

19

cerebral ischemia in rats. Brain Res. 1402, 109–121. http://dx.doi.org/10.1016/j. brainres.2011.05.049. Yang, Y.R., Chang, H.C., Wang, P.S., et al., 2012. Motor performance improved by exercises in cerebral ischemic rats. J. Mot. Behav. 44 (2), 97–103. http://dx.doi. org/10.1080/00222895.2012.654524. Ye, L.-B., Yu, X.-C., Xia, Q.-H., et al., 2016. Regulation of caveolin-1 and junction proteins by bFGF contributes to the integrity of blood-spinal cord barrier and functional recovery. Neurotherapeutics. http://dx.doi.org/10.1007/s13311-0160437-3. Yu, S., Yao, S., Wen, Y., et al., 2016. Angiogenic microspheres promote neural regeneration and motor function recovery after spinal cord injury in rats. Sci. Rep. 6, 33428. http://dx.doi.org/10.1038/srep33428. Yuan, A., Yu, C.J., Shun, C.T., et al., 2005. Total cyclooxygenase-2 mRNA levels correlate with vascular endothelial growth factor mRNA levels, tumor angiogenesis and prognosis in non-small cell lung cancer patients. Int. J. Cancer 115 (4), 545–555. http://dx.doi.org/10.1002/ijc.20898. Zhao, Y., Pang, Q., Liu, M., et al., 2016. Treadmill exercise promotes neurogenesis in ischemic rat brains via Caveolin-1/VEGF signaling pathways. Neurochem. Res. http://dx.doi.org/10.1007/s11064-016-2081-z.