Inhibition of Reactive Astrocytes with Fluorocitrate Ameliorates Learning and Memory Impairment Through Upregulating CRTC1 and Synaptophysin in Ischemic Stroke Rats

Abstract

Ischemic stroke often causes motor and cognitive deficits. Deregulated glia gap junction communication, which is reflected by increased protein levels of glial fibrillary acidic protein (GFAP) and connexin 43 (Cx43), has been observed in ischemic hippocampus and has been associated with cognitive impairment in animal stroke models. Here, we tested the hypothesis that reactive astrocytes-mediated loss of synaptophysin (SYP) and CREB-regulated transcription coactivator 1 (CRTC1) contribute to dysfunction in glia gap junction communication and memory impairment after ischemic stroke. Male Sprague-Dawley rats were subjected to a 90-min middle cerebral artery occlusion (MCAO) with 7-day reperfusion. Fluorocitrate (1 nmol), the reversible inhibitor of the astrocytic tricarboxylic acid cycle, was injected into the right lateral ventricle of MCAO rats once every 2 days starting immediately before reperfusion. The Morris water maze was used to assess memory in conjunction with western blotting and immunostaining to detect protein expression and distribution in the hippocampus. Our results showed that ischemic stroke caused significant memory impairment accompanied by increased protein levels of GFAP and Cx43 in hippocampal tissue. In addition, the levels of several key memory-related important proteins including SYP, CRTC1, myelin basic protein and high-mobility group-box-1 were significantly reduced in the hippocampal tissue. Of note, inhibition of reactive astrocytes with fluorocitrate was shown to significantly reverse the above noted changes induced by ischemic stroke. Taken together, our findings demonstrate that inhibiting reactive astrocytes with fluorocitrate immediately before reperfusion may protect against ischemic stroke-induced memory impairment through the upregulation of CRTC1 and SYP.

Keywords: CRTC1; Cx43; Fluorocitrate; Ischemic stroke; Memory; Reactive astrocytes.

Fig. 1

Diagram shows the experimental procedure

Fig. 2

Effect of FC treatment on stroke-induced GFAP expression in hippocampus after 90-min MCAO and 7-day reperfusion. Western blot and immunofluorescence were used to detect the GFAP expression and spatial distribution in the nonischemic (NI) and ischemic (I) hemispheric tissue. Representative western blot revealed that MCAO induced a significant increase in GFAP protein level in hippocampus (A, left panel) and the relative band intensities of GFAP were quantitated (a, right panel). FC treatment significantly prevented this increase (^#P < 0.05 versus vehicle group, n = 4/group). Representative photomicrographs of fluorescent staining of GFAP in CA1 (b) and CA3 (c) in hippocampus. MCAO induced a significant increase GFAP in the CA1 (b) and CA3 (c) and FC treatment significantly prevented this increase. N = 3/group, scale bar = 100 μm

Fig. 3

Effect of FC treatment on stroke-induced learning, memory and motor impairment after 90-minute MCAO and 7-day reperfusion. Morris water maze was used to detect learning (a–c) and memory (d, e). Diagram shows the track of motion (a). Treatment with FC significantly decreased the escape latency (P < 0.01 versus vehicle group, b), but had no effect on mean speed (P > 0.05 versus vehicle group, c). Treatment with FC significantly decreased the latency to first entry (P < 0.05, d), but not the duration in the target quadrant (P > 0.05, e). Treatment with FC significantly decreased neurological scores (P < 0.01, f) and foot-faults (P < 0.01, g). N = 10 for vehicle group and n = 11 for FC group

Fig. 4

Effect of FC treatment on stroke-induced Cx43 expression change in hippocampus after 90-minute MCAO and 7-day reperfusion. Representative western blot revealed that MCAO induced a significant increase in Cx43 protein level in hippocampus (a, left panel). The relative band intensities of Cx43 were quantitated (a, right panel). *P < 0.05 versus NI hemisphere in vehicle group. Treatment with FC significantly prevented this increase.#P < 0.05 versus vehicle group, n = 5/group. Representative photomicrographs of fluorescent staining of Cx43 in CA1 (b) and CA3 (c) in hippocampus. MCAO induced a significant decrease Cx43 in the CA1 (B) and CA3 (c). Treatment with FC significantly prevented this decrease. N = 3/group, scale bar = 100 μm

Fig. 5

Effect of FC treatment on stroke-induced SYP, CRTC1, p-CRTC1 and MBP expression in hippocampus after ischemic stroke. Representative western blot revealed protein level of SYP (a, upper panel) and p-CRTC1, CRTC1 (b, upper panel) in hippocampus. The relative band intensities of SYP (a, bottom panel) and CRTC1 (b, bottom panel) were quantitated. Ischemic stroke induced a significant decrease of SYP and CRTC1, but no significant change of p-CRTC1. Treatment with FC significantly prevented this decrease of SYP and CRTC1. *P < 0.05 versus NI hemisphere in vehicle group.#P < 0.05 versus vehicle group, n = 4/group for SYP, n = 4/group for CRTC1. Representative photomicrographs of fluorescent staining of MBP from CA1 and CA3 in hippocampus. Ischemic stroke induced a significant decrease MBP in CA1 (c) and CA3 (d). FC treatment significantly prevented this decrease. N = 3/group, scale bar = 100 μm

Fig. 6

Effect of FC treatment on stroke-induced HMGB1 expression in hippocampus after ischemic stroke. Representative western blot revealed protein level of HMGB1 (a, left panel) in hippocampus. The relative band intensities of HMGB1 (a, right panel) were quantitated. Ischemic stroke induced a significant decrease of HMGB1 in hippocampus and treatment with FC significantly prevented this decrease of HMGB1. *P < 0.05 versus NI hemisphere in vehicle group. ^#P < 0.05 versus vehicle group, n = 4/group for HMGB1. Representative photomicrographs of fluorescent staining of HMGB1 and GFAP from CA1 and CA3 in hippocampus. Ischemic stroke induced a significant increase of GFAP and a significant decrease of HMGB1 in CA1 (b) and CA3 (c). FC treatment significantly prevented this change. HMGB1 did not have a good colocalization with GFAP. N = 3/group, scale bar = 100 μm