A Linarin Derivative Protects against Ischemia-Induced Neuronal Injury in Mice by Promoting Cerebral Blood Flow Recovery via KDELR-Dependent CSPG4 Activation

Abstract

The cerebral ischemic microvascular response and collateral circulation compensatory capacity are important for the outcome of ischemic stroke. Here, we sought to evaluate the effect of a linarin derivate 4'-benzylapigenin-7-β-rutinoside (BLR) on neurological function and cerebral blood flow restoration in ischemic stroke. A mouse model of middle cerebral artery occlusion (30 min) with reperfusion (24 h) was used to mimic ischemic stroke in vivo, and 2,3,5-triphenyltetrazolium chloride (TTC) staining, terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assays, and immunofluorescence microscopy were used to assess the protective effects of BLR on infarct volume, neurological function, neuronal apoptosis, and inflammatory damage. Cerebral blood flow was assayed by laser speckle contrast imaging. Double immunostaining of GFAP-collagen IV and brain lucidification were performed to determine the protective effects of BLR on the disruption of brain vasculature. Differential gene expression was assessed by RNA sequencing. Coimmunoprecipitation and western blotting were used to explore the mechanism of BLR-induced neuroprotection. The results of in vivo experiments showed that BLR administration after reperfusion onset reduced infarct volume, improved neurological function, and decreased the neural cell apoptosis and inflammatory response in the ischemic brain, which was accompanied by increased cerebral blood flow and reduced detachment of astrocyte endfeet from the capillary basement membrane. The RNA sequencing data showed that BLR promoted the upregulation of extracellular matrix and angiogenesis pathway-related genes; in particular, BLR significantly increased the expression of the chondroitin sulfate proteoglycan 4 (CSPG4) gene, enhanced the membrane location of CSPG4, and promoted its downstream signaling protein expression, which is associated with KDEL receptor (KDELR) activation. In addition, activated KDELR further increased the phosphorylation of mitogen-activated protein kinases after BLR treatment. Taken together, our data showed that BLR could protect against ischemic brain injury and may serve as a new promising therapeutic candidate drug for ischemic stroke, and that KDELR might act as both a sensor and effector to activate CSPG4 to increase cerebral blood flow.

Figure 1

Preliminary and synthesis of a linarin derivative 4′-benzylapigenin-7 -β-rutinoside (BLR). (a) The names and structures of linarin and its derivates and (b) the synthetic route and chemical structure of BLR from linarin. The neuroprotective effects of linarin and its derivatives were evaluated by the mouse MCAO model, and the result of preliminary study was not shown.

Figure 2

Delivery of BLR improved short-term stroke outcomes. The mice were subjected to 30min MCAO and 24h reperfusion and treated with BLR at the dose of 4, 20, and 40mg/kg or vehicle via intraperitoneal at 1h after reperfusion onset; the series of battery behavior tests were performed 24h post stroke. (a) Schematic diagram of the experimental design. (b) Body weight was monitored over time. ^##P < 0.01 compared with the sham group (n = 10 per group). (c) LDF recorded the CBF changes in ischemic core of the MCA territory (n = 10 per group). ^##P < 0.01 compared with the sham group. (d) Representative TTC-stained images of brain slices. (e) Infarct volume of TTC-stained brain slices was quantitated (n = 10 per group). ^###P < 0.001 compared with the sham group. ^∗∗∗P < 0.01 versus MCAO group. (f) Neurological defects were evaluated by neurologic scores (n = 10 per group). ^###P < 0.001 compared with the sham group. ^∗∗∗P < 0.001 versus MCAO group. (g) Open field and (h) rotarod tests were evaluated presurgery (day 0) at baselines and at 24 hours after MCAO (n = 10 per group). Data were analyzed by two-way ANOVA, followed by two-tailed unpaired t-tests. ^#P < 0.05 compared with the sham group. ^∗P < 0.05 versus MCAO group.

Figure 3

BLR attenuated MCAO-induced apoptosis and inflammation in mice. (a) The images of apoptotic cells in brain slices assessed by TUNEL assay (n = 6 for each group). White arrows represent TUNEL-positive cells, scale bar = 25 μm. (b) Immunofluorescence images of Iba-1 (n = 6 per group). Scale bar = 25 μm. (c) The levels of inflammatory cytokines including TNF-α, IL-1β, IL-6, and ICAM-1 were detected by ELISA (n = 6 per group). ^##P < 0.01 compared with the sham group. ^∗∗P < 0.01 compared with MCAO group.

Figure 4

BLR inhibited cerebral blood flow decline in ischemic zone induced by MCAO. (a) Cerebral blood flow was detected by LSCI in entire MCAO process (including pre-MCAO, during MCAO, and post-MCAO) (n = 6 per group). Representative laser speckle contrast images of surface cortical CBF. Red oval represents ROI of ipsilateral hemisphere; green oval represents ROI area of contralateral. (b) Blood flow index changes (lpsi ROI/control ROI) of CBF in MCAO or BLR treatment (n = 6 per group). ^∗P < 0.05 compared with the MCAO group. (c) 3D volume rendering of the vascular architecture in ischemic stroke. Brain transparent was realized by cubic clarity in the mouse MCAO model. The brain of the mice was clarified by CUBIC (n = 6 per group) after MCAO and BLR treatment. The ischemic regions labeling with red circle were detected by ultramicroscope (n = 6 per group). White arrows point to the areas which represent the blood vessels. Scale bar = 2000 μm. (d) Representative immunofluorescence images of CD31 (n = 6 per group) to evaluate the capillary density in ischemic brain tissue. Scale bar = 25 μm.

Figure 5

BLR treatment reduced disconnection of the astrocyte endfeet from the capillary basement membrane after ischemic stroke. Double immunofluorescence images of GFAP and collagen IV (n = 6 per group). Arrowheads indicate to the detachment degree of the GFAP-positive astrocyte endfeet from the collagen IV-positive basement membrane. Scale bar = 25 μm.

Figure 6

Effects of BLR on the vasculature and angiogenesis pathway-related gene expression by RNA sequencing analysis. (a) Scatter plot showed the differences of gene expression between MCAO group and BLR group. Red dots represented as upregulated genes, blue dots represented as downregulated genes, and gray dots indicated no change. (b) Pathway enrichment analyses were performed on the clusterProfiler software using ClueGO module at 0.05 significant level. (c) Interaction network and cluster relationship of differentially expressed genes involved in MCAO mice treated with BLR or saline were shown. (d) Heatmap revealing RNA sequencing result of upregulation and downregulation of vasculature and angiogenesis pathway-related gene expression between MCAO group and BLR group (n = 3 per group). Red blocks indicated high expression, and blue blocks indicated low expression.

Figure 7

BLR induced transport and membrane localization of CSPG4 mediated by KDELR after MCAO. (a) Effects of BLR on the expression of CSPG4 in different subcellular fraction (cytoplasm, ERGIC, and membrane). (b) There is no contamination among subcellular fraction (cytoplasm, ERGIC, and membrane). (c) Effects of BLR on the expression of KDELR and CSPG4, together with the combined of KDELR to CSPG4 measured with Co-IP assays. The results of three independent experiments were termed as the mean ± SD. ^##P < 0.01 compared with the sham group. ^∗∗P < 0.01 compared with the MCAO group. n = 6 per group.

Figure 8

BLR both aroused CSPG4 pathway and KDELR-associated MAPK cascades in ischemic stroke. (a) Effects of BLR on the phosphorylation levels of CSPG4 downstream targets (phosphorylated Akt and phosphorylated Src). (b) Effects of BLR on the phosphorylation levels of KDELR-associated MAPK cascades (p38 and CREB). (c) BLR affected the expression levels of ER-resident molecular chaperones including GRP94 and GRP78. (d) A schematic diagram in which a derivate of linarin (BLR) protects against neurovascular unit injury via KDELR-dependent CSPG4 activation in ischemic stroke (n = 6 per group). ^##P < 0.01 compared with the sham group. ^∗∗P < 0.01 compared with the MCAO group.