Granulocyte colony-stimulating factor attenuates delayed tPA-induced hemorrhagic transformation in ischemic stroke rats by enhancing angiogenesis and vasculogenesis

Abstract

Treatment with tissue plasminogen activator (tPA) beyond the therapeutic time window (>4.5 hours post stroke) may produce hemorrhagic transformation (HT). Strategies that could extend the narrow time window of tPA will benefit a significant number of stroke patients. Male Sprague-Dawley rats underwent middle cerebral artery occlusion (MCAo) and given vehicle, tPA (10 mg/kg), or tPA and granulocyte colony-stimulating factor (G-CSF, 300 μg/kg), at 6 hours after MCAo. Twenty-four hours post treatment, G-CSF+tPA-treated stroke rats displayed 25% improvement in neurological functions and 38.9% reduction of hemorrhage, with Western blots showing 1.9- and 1.2-fold increments in Ang-2 expression in the ischemic cortex and striatum, respectively, and 3-fold increase in phosphorylated endothelial nitric oxide synthase expression in the ipsilateral cortex relative to tPA-treated rats. Immunohistochemistry also showed 2- and 2.8-fold increase in von-Willebrand expression, 3.2- and 2.2-fold increased CD34+ expression, and 4- and 13-fold upregulation of VEGFR-2 expression in the ischemic cortex and striatum, respectively, in G-CSF+tPA-treated stroke rats relative to tPA-treated subjects. Altogether, these findings indicate that G-CSF attenuated delayed tPA-induced HT likely via the enhancement of angiogenesis and vasculogenesis. The use of G-CSF to protect the vasculature may improve the clinical outcome of tPA even outside the currently indicated therapeutic window for ischemic stroke.

Figure 1

Effects of granulocyte colony-stimulating factor (G-CSF) on delayed tissue plasminogen activator (tPA)-induced hemorrhage, cerebral infarction, and neurological deficits in MCAo rats. (A) Quantitative analysis of cerebral hemorrhage volume with spectrophotometric assay revealed incidence of hemorrhage (shown as increased levels of hemoglobin levels in the brain) in rats subjected to delayed tPA treatment. G-CSF treatment caused a 38% reduction of delayed tPA-induced hemorrhage (n=5–6 animals per group). (B) Quantitative analysis of infarct volume in vehicle- (n=9), tPA- (n=7) and G-CSF+tPA- (n=8) treated groups. (C) Photographs are representative coronal brain sections stained with triphenyltetrazolium chloride 24 hours after MCAo, showing infarct area (white) and intact areas (red). G-CSF had no effect on infarct volume. (D) Effects of G-CSF on delayed tPA-induced neurological deficits in MCAo rats. Twenty-four hours after drug treatment (Tx), rats injected with G-CSF displayed improvement of delayed tPA-induced neurological deficits (25% and 24.8%, relative to control and tPA-treated stroke rats, respectively). (n=8–10 animals per group), *P<0.05, **P<0.01; NS, not significant. Data are expressed as mean±s.e.m.

Figure 2

Western blotting for angiopoetins (Ang-1 and Ang-2), endothelial nitric oxide synthase (eNOS) and phosphorylated-eNOS. As shown in representative bands, treatment with granulocyte colony-stimulating factor+tissue plasminogen activator (G-CSF+tPA) upregulated Ang-2 (marker of angiogenesis) expression in the ischemic hemisphere. Quantitative analyses showed 1.9- and 1.2-fold increment in Ang-2 expression in the ischemic cortex and striatum, respectively, in G-CSF+tPA-treated rats relative to subjects treated with tPA only. Ang-1 expression in both ipsilateral and contralateral hemispheres was similar in all groups. Relative to tPA-treated rats, G-CSF+tPA-treated stroke rats also showed threefold increase in peNOS expression in the ischemic cortex, but not striatum, in cerebral hemisphere ipsilateral to MCAo. Total eNOS expression in the ipsilateral and contralateral hemispheres was similar in all treatment groups although there was a trend of increased eNOS expression in the cortex of G-CSF+tPA-treated versus tPA-treated rats. *P<0.05, **P<0.01, ***P<0.001 n=5–6. Data are expressed as mean±s.e.m. relative to sham group and normalized to β-actin.

Figure 3

Immunohistochemical analyses of von-Willebrand (vWF), CD34+ and vascular endothelial growth factor (VEGRF)-2 expression levels in the ischemic cortex. Representative merged images showing co-localization of vWF, CD34+, or VEGFR-2 with 4′,6-diamidino-2-phenylindole (DAPI; blue filter, nuclear staining). Analyses of fluorescence intensities showed 2-fold increase in the expression of vascular marker vWF, as well as 3.2- and 4-fold increment in endothelial progenitor cell markers CD34+ and VEGFR-2, respectively, in the ischemic cortex of G-CSF+tPA-treated rats compared with subjects administered with tPA only (at 6 hours post MCAO). **P<0.01, n=5–6. Data are expressed as mean±s.e.m. Horizontal bar indicates 100 μM.

Figure 4

Immunohistochemical analyses of von-Willebrand (vWF), CD34+ and vascular endothelial growth factor (VEGRF)-2 expression levels in the ischemic striatum. Representative merged images showing co-localization of vWF, CD34+, or VEGFR-2 with 4′,6-diamidino-2-phenylindole (DAPI; blue filter, nuclear staining). Analysis of fluorescence intensities showed 2.8-fold increase in the expression of vascular marker vWF, as well as 2.2- and 13-fold increment in endothelial progenitor cell markers CD34+ and VEGFR-2, respectively, in the ischemic cortex of G-CSF+tPA-treated rats compared with subjects administered with tPA only (at 6 hours post MCAo). *P<0.05, **P<0.01, n=5–6. Data are expressed as mean±s.e.m. Horizontal bar indicates 100 μM.