Effect Of Modulation Of Glial Glutamate Transporter Subtype GLT-1 On Neuro-protection Of Cerebral Ischemic Preconditioning In Rats

Ischemic cerebrovascular disease is one kind of diseases which severely impair human health. Since neuronal cells are very sensitive to ischemia, long time brain ischemic insult results in neuronal death, which usually can not revive and frequently causes severe sequelae even if blood supply recovers. With the progress of the studies on the ischemic cerebrovascular disease, more and more attention has been paid to the phenomenon of brain ischemic tolerance (BIT). It is found in animal experiment that suddenly severe cerebral ischemic insult causes severe delayed neuronal death (DND) in the CA1 hippocampus. But when the ischemic insult was preceded by a transient sublethal cerebral ischemic preconditioning, the DND normally induced by the ischemic insult could be significantly prevented. This phenomenon was called brain ischemic tolerance (BIT), and the sublethal cerebral ischemia given in advance was called cerebral ischemic preconditioning (CIP). It is important to clarify the mechanism of the neuro-protection induced by CIP for developing new therapeutic methods to enhance the tolerance of neurons to ischemia and hypoxia.Ischemia and hypoxia in brain tissue induced by various kinds of diseases can result in dysmetabolism in energy and inhibit the action of Na+-K+-ATPase, which would lead to an obvious increase of extracellular K+ concentration and intracellular Na+ concentration. The changes facilitate depolarization of neurons and release of glutamate in the nerve terminal. In addition, the high concentration of extracellular K+ and intracellular Na+ can reverse the activity of high-affinity glutamate transporter, which would lead to reversal transporting of glutamate from intracellular to extracellular space. These reasons result in an increase in extracellular glutamate concentration. The increased extracellular glutamate binds with and activates NMDA or non-NMDA receptors, and then induces influx of Na+ and Ca2+ and release of intracellular Ca2+ to cytoplasm. The neurons die of the overload of Na+ and Ca2+. Thus, the glutamate has been referred to as excitotoxin. Therefore, it is an impotant strategy to decrease the concentration of extracellular glutamate or to reduce the combination of glutamate with its specific receptor for preventing the excitotoxicity of the increased glutamate in brain ischemia. Excitatory amino acid transports (EAATs) are important mechanism for regulating the extracellular concentration of glutamate. It has been well-known that the glial glutamate transporter subtype GLT-1 plays a dominant role in terminating glutamate neurotransmission, and maintaining the extracellular glutamate below neurotoxic levels. Some studies paid attention to roles of GLT-1 in induction of tolerance of brain to ischemic or hypoxic insults. For example, Douen et al reported that preconditioning with cortical spreading depression could down-regulate the expression of glutamate transporters EAAT1 and EAAT2 of astrocytes to prevent the reversal transporting of glutamate during cerebral ischemia. Romera et al found that preconditioning with ischemia or hypoxia in vitro could increase the expression of glutamate transports. These results suggested that GLT-1 might associate with the induction of BIT. But there is no report yet about whether GLT-1 of astrocytes participates in the neuro-protection of CIP in vivo and whether the modulation of GLT-1 may enhance the tolerance of neuronal cells to ischemic insult. Therefore, the present study was undertaken to explore in vivo the effect of modulation of GLT-1 by Dihydrokainate (DHK), a selective inhibitor of GLT-1, GLT-1 antisense oligodeoxynucleotides (AS-ODNs) or sense oligodeoxynucleotides (S-ODNs) on the induction of BIT using rat global cerebral ischemic model. The results to be obtained would provide experimental evidence for clarifying the role of GLT-1 in the induction of BIT.1 Inhibiting function of GLT-1 with DHK blocked BIT induced by CIPThe role of GLT-1 in BIT was investigated in rat 4-vessel occlusion (4VO) global brain ischemia model by observing the effect of Dihydrokainate (DHK), a selective inhibitor of GLT-1, on the neuro-protection of CIP against delayed pyramidal neuronal death in the CA1 hippocampus normally induced by severe brain ischemic insult.Ninety-six Wistar rats with permanently occluded bilateral vertebral arteries for 2d were randomly assigned to eight groups:①Sham group (n=6): the bilateral common carotid arteries (BCCA) were separated, but without occluding the blood flow;②CIP group (n=6): the BCCA were clamped for 3 min;③Brain ischemic insult group (n=6): the BCCA were clamped for 8 min;④CIP+Brain ischemic insult group (n=6): a CIP for 3 min was preformed first, and then a brain ischemic insult for 8 min was given 2d after the CIP;⑤Distilled water group (n=6): 20μl distilled water was injected into the right lateral ventricle;⑥DHK group(n=30): 20μl DHK solution was injected into the right lateral ventricle. This group was further divided into 10 nmol, 100 nmol, 200 nmol, 500 nmol and 1000 nmol subgroups according to the doses of DHK;⑦DHK+CIP group(n=18): 20μl DHK solution was injected into the right lateral ventricle 20 min before CIP. This group was further divided into 10 nmol, 100 nmol and 200 nmol subgroups according to the doses of DHK;⑧DHK+CIP+brain ischemic insult group (n=18): 20μl DHK solution was injected into the right lateral ventricle 20 min before CIP. Other procedures were the same as those in CIP+brain ischemic insult group. According to the doses of DHK used, this group was also further divided into 10 nmol, 100 nmol and 200 nmol subgroups. The animals were sacrificed by decapitation at 7d after the sham operations or the last operations. The brain was sectioned in a thickness of 5μm and stained with thionin. Histological changes of the CA1 hippocampus were examined under light microscope to determine delayed neuronal death (DND) by neuronal density (ND) and histological grade (HG) of the CA1 hippocampus. The ND was determined by counting the number of surviving pyramidal neurons with intact cell membrane, full nucleus and clear nucleolus within 1 mm linear length of the CA1 hippocampus. The average of the number in 3 areas of the CA1 hippocampus was calculated as value of ND. HG was divided into the following 4 grades: grade 0, no neuron death; grade Ⅰ, scattered single neuron death; gradeⅡ, death of many neurons; gradeⅢ, death of almost complete neurons. The average HG of the bilateral hippocampus was counted as statistical data.The results showed that during the 4VO, the pupils of rats enlarged. EEG showed decreases in frequency and amplitude, even approached isoelectric level, and the righting reflex disappeared. These manifestations indicated that the rats were really subjected to global cerebral ischemia.It was found by thionine staining that there was no significant neuronal damage in the CA1 subfield of the hippocampus in the sham and CIP groups. HG was 0～Ⅰ, and values of ND were 208.25±5.97 and 202.93±4.32 in the two groups, respectively. Obvious destruction of the CA1 subfield was found in brain ischemic insult for 8 min group. The value of ND was 45.86±21.93, which was much low compared with that in the sham and CIP groups (P<0.01). HG was gradeⅡ～Ⅲ, which was much high compared with that in the sham and CIP. In CIP+ischemic insult group, no obvious neuronal damage in the hippocampal CA1 region was found. HG (grade 0～Ⅰ) was significantly lower than that in brain ischemic insult group (P<0.01), and the value of ND (208.07±5.87) was much higher than that in brain ischemic insult group (P<0.01). The results indicated that CIP protected pyramidal neurons of the CA1 hippocampus against DND induced by brain ischemic insult. There was no apparent DND in distilled water group, in which histological characteristics such as HG and ND were similar with those in the sham group. DHK in doses of 10, 100 and 200 nmol just caused slight DND of pyramidal neurons in the CA1 hippocampus. HG was gradeⅠin each dose, and the values of ND were 172.76±17.31, 162.64±6.12 and 155.43±9.82, respectively. While obvious DND was found in 500 and 1000 nmol groups; the value of ND was significantly decreased (105.35±3.84 and 6±1.39, respectively); HG was significantly increased (ⅡandⅢ, respectively) compared with 10, 100 and 200 nmol DHK groups (P<0.05). These results indicated that large dose DHK could lead to significant pyramidal neuron damage in the hippocampal CA1 region. Thus, we only observed effects of 10, 100 and 200 nmol DHK in the subsequent experiment. There was no apparent DND was found in DHK+CIP group in each dose of 10, 100 and 200 nmol DHK, in which histological characteristics such as HG and ND were similar with those in the sham group. Obvious DND was found in DHK+CIP+brain ischemic insult group. HG wereⅠ,Ⅰ～ⅡandⅢ, and the values of ND were 160.87±13.55, 117.07±10.11 and 4.87±2.02, respectively, in the doses of 10 nmol, 100 nmol and 200 nmol. The changes above in DHK+CIP+brain ischemic insult group were significant compared with CIP+brain ischemic insult group (P<0.01). Additionally, the correlative curve between the doses and effectiveness in DHK+CIP+brain ischemic insult groups shifted to left siginifantly, and the slope rate became sharper than that in DHK group. These results indicated that administration of DHK blocked in a dose dependent manner the neuroprotection of CIP against DND normally induced by severe brain ischemic insult in the CA1 hippoacmpus.The above results suggested the involvement of GLT-1 in the neuroprotection of CIP against DND normally induced by severe brain ischemic insult in the CA1 hippoacmpus.2 Inhibition of the expression of GLT-1 with GLT-1 AS-ODNs attenuated the neuro-protection of CIP against brain ischemic insultThe role of GLT-1 in BIT was investigated by observing the effect of antisense oligodeoxynucleotides (AS-ODNs) of GLT-1, which can inhibit the protein expression of GLT-1, on the neuro-protection of CIP against brain ischemic insult.2.1 The effect of AS-ODNs on the protein expression of GLT-1Forty-two Wistar rats with permanently occluded bilateral vertebral arteries for 36 h were randomly assigned to three groups.①control group (n=6): 5μl distilled water was injected into the right lateral ventricle and the animals were sacrificed by decapitation 12 h later.②AS-ODNs 9 nmol group(n=18):5μl AS-ODNs (9 nmol) solution was injected into the right lateral ventricle. This group was further divided into 12h, 24h and 36h subgroups according to the time of reperfusion after injection.③AS-ODNs 18 nmol group(n=18):5μl AS-ODNs (18 nmol) solution was injected into the right lateral ventricle. This group was also further divided into 12h, 24h and 36h subgroups according to the time of reperfusion after injection. The animals were sacrificed by decapitation at the determined endpoint of the experiment. The hippocampal CA1 was dissected out quickly on ice and determine the expression of GLT-1 using Western blotting. Integrated optical density (IOD) of each band was measured using image analysis system. The changes of GLT-1 expression were represented with the ratio of IOD of the aim protein band toβ-actin.The results of Western blotting showed that IOD in the control group was 0.65±0.22. In 9 nmol AS-ODNs groups, IOD were 0.37±0.07, 0.20±0.05 and 0.25±0.07, respectively. In 18 nmol AS-ODNs groups, IOD were 0.11±0.05, 0.05±0.02 and 0.17±0.16, respectively. IOD in AS-ODNs groups were decreased significantly compared with control group (P<0.05), and these changes were in a dose dependent manner.The results indicatedthatAS-ODNs injected into the lateral ventricle inhibited expression of GLT-1 in a dose dependent manner, and the inhibiting effect was much significant at 24h.2.2 AS-ODNs of GLT-1 inhibit the neuro-protection on pyramidal neurons of the CA1 hippocampus induced by CIPFifty-four Wistar rats with permanently occluded bilateral vertebral arteries were randomly assigned to seven groups.①Sham group (n=6): the bilateral common carotid arteries (BCCA) were separated, but without occluding the blood flow.②CIP group (n=6): the BCCA were clamped for 3 min.③Brain ischemic insult group (n=6): the BCCA were clamped for 8 min.④CIP+Brain ischemic insult group (n=6): 3 min CIP was preformed 2d prior to 8 min ischemic insult.⑤D istilled water group (n=6): 5μl distilled water was injected into the right lateral ventricle 12 h before, 12 h and 36 h after the BCCA were separated (but without occluding the blood flow), respectively.⑥AS-ODNs group(n=12): 5μl AS-ODNs solution was injected into the right lateral ventricle 12 h before, 12 h and 36 h after the BCCA were separated (but without occluding the blood flow), respectively. This group was further divided into 9 nmol and 18 nmol subgroups according to the doses of AS-ODNs (n=6 in each group).⑦A S-ODNs+CIP+Ischemic insult group (n=12): 5μl AS-ODNs solution was injected into the right lateral ventricle 12h before, 12h and 36h after CIP, respectively. This group was also further divided into 9 nmol and 18 nmol subgroups according to the doses of AS-ODNs. The other treatment was the same as CIP+Brain ischemic insult group, with six rats in each group. The animals were sacrificed by decapitation 7 d after sham operations or the last operations. Histological changes of the CA1 region of the hippocampus were examined using thionin staining (the methods were the same as that in part 1).Thionin staining showed that there was no significant neuronal damage in the CA1 subfield of the hippocampus in the sham and CIP groups. HG was 0～Ⅰ, and values of ND were 208.25±5.97 and 202.86±4.28 in the two groups, respectively. Obvious DND in the hippocampal CA1 region was found in ischemic insult for 8 min group. The value of ND was 45.86±21.93, which was much low compared with that in the sham and CIP groups (P<0.01). HG was gradeⅡ～Ⅲ, which was much high compared with that in the sham and CIP groups. In CIP+ischemic insult group, no obvious neuronal damage in the hippocampal CA1 region was found. HG (grade 0～Ⅰ) was significantly lower than that in 8 min brain ischemia group (P<0.01), and the value of ND was 208.98±5.9, which was much higher than that in 8 min brain ischemic insult group (P<0.01). The results indicated that CIP induced BIT of pyramidal neurons of the CA1 hippocampus against DND caused by brain ischemic insult. There was no apparent DND was found when distilled water and AS-ODNs (9 nmol and 18 nmol) were administrated into right lateral cerebral ventricle, and the histological changes were the same as those in sham and CIP groups. HG was grade 0～Ⅰand the values of ND were 176.22±6.29, 176.89±2.88 and 175.32±2.77, respectively. These results indicated that the injection of distilled water and AS-ODNs into the right lateral cerebral ventricle could not lead to apparent pyramidal neuron damage in the hippocampal CA1 region. In AS-ODNs +CIP + brain ischemic insult groups, DND were significantly obvious compared with that in CIP+ ischemic insult group. HG were gradeⅡ～Ⅲ, and ND (9 nmol and 18 nmol groups) were 133.56±3.42 and 70.94±7.38, respectively (P<0.01), Which indicated that AS-ODNs attenuated the neuro-protection on pyramidal neurons of the CA1 hippocampus induced by CIP via inhibiting the expression of GLT-1.The results indicated that the injection of AS-ODNs into the lateral cerebral ventricle induced the down-regulation of GLT-1 but no obvious DND in hippocampal CA1 region. However, the protective effect of CIP on pyramidal neurons of the CA1 hippocampus was attenuated by the injection of AS-ODNs. These results further suggested the association of GLT-1 with the induction of BIT by CIP in rats.3 Up-regulating the expression of GLT-1 with GLT-1 S-ODNs enhanced the tolerance to brain ischemic insult in the hippocampal CA1 regionThe role of GLT-1 in the brain ischemic tolerance was investigated by observing the effect of sense oligodeoxynucleotides (S-ODNs) of GLT-1 on tolerance to brain ischemic insult in the hippocampal CA1 region via up-regulating the protein expression of GLT-1.3.1 The effect of S-ODNs on the protein expression of GLT-1Forty-two Wistar rats with permanently occluded bilateral vertebral arteries for 36 h were randomly assigned to three groups.①control group (n=6): 5μl distilled water was injected into the right lateral cerebral ventricle and the animals were sacrificed by decapitation 12 h later.②S-ODNs 9 nmol group(n=18):5μl S-ODNs (9 nmol) solution was injected into the right lateral cerebral ventricle. This group was further divided into 12 h, 24 h and 36 h subgroups according to the time of reperfusion after injection.③S-ODNs 18 nmol group (n=18):5μl S-ODNs (18 nmol) solution was injected into the right lateral cerebral ventricle. This group was further divided into 12 h, 24 h and 36 h subgroups according to the time of reperfusion after injection. The animals were sacrificed by decapitation at the determined endpoint of the experiment. The hippocampal CA1 was dissected out quickly on ice and detected the expression of GLT-1 using Western blotting. IOD of each band was measured using image analysis system. The changes of GLT-1 expression were represented with the ratio of IOD of the aim protein toβ-actin.The results of Western blotting showed that IOD in the control group was 0.64±0.19. In 9 nmol S-ODNs groups, IOD were 0.78±0.02, 1.01±0.04 and 0.68±0.06, respectively. In 18 nmol S-ODNs groups, IOD were 0.84±0.02, 1.23±0.03 and 0.62±0.02, respectively. IOD in S-ODNs groups were up-regulated significantly compared with control group (P<0.05), and these changes were in a dose dependent manner.The results indicated that S-ODNs injected into the lateral ventricle up-regulated the protein expression of glutamate transporter subtype GLT-1 in a dose dependent manner, and this effect was much significant at 24 h. 3.2 S-ODNs of GLT-1 attenuated the injury of pyramidal neurons in the CA1 hippocampus induced by brain cerebral ischemiaFifty-four Wistar rats with permanently occluded bilateral vertebral arteries were randomly assigned to seven groups.①Sham group (n=6): the bilateral common carotid arteries (BCCA) were separated, but without occluding the blood flow.②CIP group (n=6): the BCCA were clamped for 3 min.③Brain ischemic insult group (n=6): the BCCA were clamped for 8 min.④CIP+Brain ischemic insult group (n=6): 3 min CIP was preformed 2d prior to 8 min ischemic insult.⑤D istilled water group (n=6): 5μl distilled water was injected into the right lateral cerebral ventricle 12h before, 12h and 36h after the BCCA were separated (but without occluding the blood flow) , respectively.⑥S-ODNs group(n=12): 5μl S-ODNs solution was injected into the right lateral cerebral ventricle 12h before, 12h and 36h after the BCCA were separated (but without occluding the blood flow), respectively. This group was further divided into 9 nmol and 18 nmol subgroups according to the doses of S-ODNs (n=6 in each group).⑦S -ODNs + Ischemic insult group (n=12): 5μl S-ODNs solution was injected into the right lateral cerebral ventricle 12 h before, 12 h and 36 h after the BCCA were separated (but without occluding the blood flow), respectively. The BCCA were clamped for 8 min after the BCCA were separated for 2 d. This group was also further divided into 9 nmol and 18 nmol subgroups according to the doses of S-ODNs. The animals were sacrificed by decapitation 7 d after sham operations or the last operations. Histological changes of the CA1 hippocampus were examined using thionin staining (the method was the same as that in part 1).Thionin staining showed that there was no significant neuronal damage in the CA1 subfield of the hippocampus in the sham and CIP groups. HG was 0～Ⅰ, and values of ND were 208.25±5.97 and 202.86±4.28 in the two groups, respectively. Obvious DND in the hippocampal CA1 region was found in ischemic insult for 8 min group and the value of ND was 45.86±21.93, which was much low compared with that in the sham and CIP groups (P<0.01). HG was gradeⅡ～Ⅲ, which was much high compared with that in the sham and CIP groups. In CIP+ischemic insult group, no obvious neuronal damage in the hippocampal CA1 region was found. HG (grade 0～Ⅰ) was significantly lower than that in 8 min brain ischemia group (P<0.01) and the value of ND was 208.98±5.9, which was much higher than that in 8 min brain ischemic insult group (P<0.01). The results indicated that CIP induced BIT of pyramidal neurons of the CA1 hippocampus against DND caused by brain ischemic insult. There was no apparent DND when distilled water and S-ODNs (9 nmol and 18 nmol) were administrated into right lateral cerebral ventricle, and the histological changes were the same as those in sham and CIP groups. HG was grade 0～Ⅰand the values of ND were 176.22±6.29, 186.53±5.64 and 204.4±12.99, respectively (P>0.05). In S-ODNs + Ischemic insult group, ND (they were 148.53±9.1 in 9 nmol group and 166.67±8.31 in 18 nmol group) was significantly increased compared with that in brain ischemic insult group (P<0.01). HG was gradeⅠwhich was significantly decreased compared with that in brain ischemic insult group. These results indicated that S-ODNs attenuate the damage of pyramidal neurons in the CA1 hippocampus induced by brain ischemic insult through up-regulating the expression of GLT-1.The results indicated that the injection of S-ODNs into the lateral cerebral ventricle led to the up-regulation of GLT-1 in hippocampal CA1 region. The damage induced by brain ischemic insult on pyramidal neurons of the CA1 hippocampus was attenuated by the injection of S-ODNs. These results further suggested the role of GLT-1 in the induction of BIT by CIP in rats.4 Conclusions(1) DHK, an inhibitor selective for glial glutamate transporter GLT-1, inhibited the protective effect induced by CIP on pyramidal neurons in the CA1 subfield of the hippocampus via inhibiting the function of GLT-1 in rats.(2) The injection of AS-ODNs of GLT-1 into the lateral cerebral ventricle inhibited the expression of glutamate transporter subtype GLT-1. At the same time, AS-ODNs of GLT-1 attenuated the neuro-protection on pyramidal neurons of the CA1 hippocampus induced by CIP.(3) The injection of S-ODNs of GLT-1 into the lateral cerebral ventricle enhanced the tolerance to brain ischemic insult in the hippocampal CA1 region as wll as up-regulating the protein expression of GLT-1.(4) These results indicated that GLT-1 participates in BIT induced by CIP.