【作者】 王璞；

【导师】 高天明；

【作者基本信息】 南方医科大学 ， 神经生物学， 2010， 博士

【摘要】 精神分裂症是一种严重危害人类健康的精神疾病,它影响着全世界约1%的人口。Neuregulin1 (NRG1)与ErbB4作为精神分裂症众多易感基因中的一员,在近年来正在被越来越多的人重视并已成为神经科学研究的热点之一。在冰岛、苏格兰、爱尔兰、中国和韩国等人群的分子遗传学研究结果表明NRG1和ErbB4是精神分裂症的易感基因。在精神分裂症患者的脑中也发现有NRG1或ErbB4表达量的异常增加,且患者大脑皮质中NRG1激活的ErbB4活性也有升高。再者,NRG1或ErbB4的敲除小鼠也表现出类精神分裂症样的症状。Neuregulin是一类隶属于营养因子家族的多肽因子,其由NRG基因编码产物的选择性剪切体所组成。在NRG的多种异构体当中(NRG 1-4),对NRG1的研究最为广泛。NRG1通过ErbB受体酪氨酸激酶起作用,ErbB受体同样具有多种异构体(ErbB1-4)。其中ErbB4在脑内高表达,并与NRG1结合后可激活其酪氨酸激酶的活性。NRG1-ErbB4信号通路在神经系统的发育中发挥着重要的作用,并且以往的研究大都集中于此。现已发现它参与了神经细胞的分化、神经元的迁移、神经突起的外向性生长和突触的形成等多个神经发育过程。尽管NRG1和ErbB4受体在成年的啮齿类动物和人的大脑中也高度表达,但目前对于NRG1-ErbB4信号通路在神经系统功能的认识绝大部分仅限于早期发育阶段,对于此信号通路在成年脑区中的作用以及对于其在精神分裂症中的病理机制仍然不甚明了。精神分裂症病人表现有情感异常,MRI的研究证明这些情感异常与杏仁核十分相关。近年的研究显示NRG1突变小鼠表现出情景恐惧记忆以及社交障碍。众所周知,恐惧记忆是反映大脑形成不良情感记忆的一种常用的模型。已有研究证明杏仁核在恐惧记忆模型中扮演着十分关键的角色,而杏仁核的长时程增强(LTP)被认为是形成恐惧记忆的一个重要机制。与其它脑区不同,杏仁核神经元动作电位的静息发放水平很低,这主要是由于该脑区具有很强的GABA能抑制活性,并且,GABA能抑制对杏仁核LTP也具有很强的控制作用。而目前对杏仁核这种强GABA能抑制的机制尚不清楚。在海马,NRG1可通过激活ErbB4抑制LTP的诱导,有研究表明ErbB4受体存在于GABA (γ-aminobutyric acid)能神经元的突触前末梢,NRG1可通过激活突触前的ErbB4受体增强活动依赖的GABA释放。因此有必要研究：杏仁核中NRG1-ErbB4信号通路是否通过调节GABA能系统来控制LTP的诱导水平。首先,我们记录了EPSC与IPSC的输入—输出关系(I-O),并使用IPSC/EPSC比证明了杏仁核相较于其它脑区具有更高GABA能水平。我们同时比较了外侧杏仁核,海马(CA1)与皮层(顶叶皮层,第五层细胞)所记录的IPSC的最大值。与海马(506.41±65.45 pA,n=10)及皮层(682.44±79.98 pA,n=9)相比,在外侧杏仁核的兴奋性神经元可以记录到更大幅度的IPSC (871.40±57.07pA,n=10),其中仅外侧杏仁核与海马两组间有显著性差异(P＜0.05, one way ANOVA)。随后观察了上述三个脑区的输入—输出关系,我们发现只有外侧杏仁核的IPSC I-O曲线保持在EPSC I-O曲线的上方,而海马和皮层IPSC I-O曲线均落在了EPSC I-O曲线的下方。而且,从IPSC/EPSC比可以看出外侧杏仁核与海马(其在20,25,30,35,40V刺激强度下,P<0.05,独立样本T检验)和皮层(在10,15V的刺激强度下,P<0.05,独立样本T检验)均有显著的差异。与以上结果相符,在外侧杏仁核中无法诱导出LTP。当给予丘脑传入纤维一个强直刺激(HFS)后,所记录到的fEPSP强度相对于刺激前的基线只有102.3±2.5%(n=8)。与此相反,HFS刺激则可在海马CA1区诱导出LTP(158.5±11.2%,n=14)。使用独立样本的T检验也显示出两个区域所诱导的LTP水平有显著性的差异(P=0.000)。这一结果提示外侧杏仁核中较高的GABA能水平阻止了LTP的产生。为了进一步肯定GABA能系统对LTP的抑制作用,在灌流液中加入20μM荷包牡丹碱(Bicuculline methiodidie, BMI, GABAA受体阻断剂)以观察诱导LTP的情况。我们发现在20μM荷包牡丹碱存在情况下,在外侧杏仁核可诱导出显著的LTP,其幅度为119.2±1.6%(n=6),而对照脑片中其幅度只有102.3±2.5%(n=8),且两组间有显著性差异(P<0.05)。这一结果表明GABA能系统抑制了外侧杏仁核中LTP的产生。为了探讨NRG1-ErbB4是否可以调节GABA能抑制系统,我们在灌流1mM犬尿喹啉酸(广谱的谷氨酸受体抑制剂)的条件下,使用双极电极刺激丘脑传入,记录了外侧杏仁核兴奋性神经元中诱发的抑制性突触后电流(eIPSC)。灌流NRG1(2nM,n=9)对IPSC的幅度无显著影响。NRG1灌流10分钟后标准化的IPSC幅度为0.90±0.05(P>0.05)。与此相反,灌流ecto-ErbB4(克隆蛋白,包含ErbB4的胞外段,可中和NRG1蛋白而起到抑制NRG1-ErbB4信号活性的作用)或AG1478(一种脂溶性的ErbB受体抑制剂,其可以阻止受体的自磷酸化并阻断其信号传递)均可显著抑制IPSC的幅度,ecto-ErbB4或AG1478处理后标准化IPSC幅度分别为0.65±0.05(2μg/ml,n=8,P<0.01)和0.70±0.04(5μM,n=10,P<0.01)。以上结果表明NRG1对eIPSC无明显影响,而NRG1-ErbB4通路的阻断剂则可以抑制eIPSC,提示生理条件下外侧杏仁核NRG1-ErbB4通路对GABA能系统的调控作用处于饱和状态。为了探讨NRG1-ErbB4是否可以调节外侧杏仁核中的LTP,我们进行了以下三组实验。空白对照组：强直刺激后fEPSP的幅度为刺激前基线的99.87±3.15%(n=8)；NRG1处理组：105.49±2.47%(n=6); ecto-ErbB4 (2μg/ml, n=7)处理组：117.97±3.72%。统计处理显示ecto-ErbB4处理组与其他两组之间都有显著性的差异(F=8.092,P=0.003,P<0.05与对照相比)。这一结果表明NRG1-ErbB4信号通路在外侧杏仁核区控制着LTP的诱导,并且处于一种相对饱和的状态而只能被负调控。同样,2μg/ml ecto-ErbB4预处理后,在皮层传入施加强直刺激可诱导出LTP(123.05±4.82%,n=8),与对照脑片(107.59±2.90%,n=6)相比有显著性的差异(P<0.05)。这一结果显示与丘脑传入相同,抑制NRG1-ErbB4信号通路可在外侧杏仁核的皮层传入也诱导出LTP,这说明NRG1-ErbB4信号对杏仁核LTP的调控无通路选择性。由于阻断GABAA受体或NRG1/ErbB4通路均可增强外侧杏仁核区中LTP的产生,所以我们认为,NRG1/ErbB4通路的阻断剂可能是通过降低GABA能抑制从而增强了LTP。为了验证这一假说,我们比较了单独灌流20μM BMI与同时灌流20μM BMI和2μg/ml ecto-ErbB4两种不同情况下所诱导LTP的幅度,以此观察两者之间是一种协同作用还是一种相互叠加的作用。我们发现这两种方式处理后都可以诱导出LTP,并且同时灌流BMI和ecto-ErbB4组与单独灌流BMI组所诱导的LTP几乎相同。单独灌流BMI组所诱导的LTP幅度为119.23±1.61%(n=6),同时灌流BMI和ecto-ErbB4组为120.76±5.68%(n=8),两组之间无统计学差异。以上结果提示NRG1/ErbB4对LTP的调控是通过GABA能机制来起作用的。我们进一步观察了ecto-ErbB4对基础兴奋性突触传递的影响。我们分别在加药前与加药20分钟后记录了fEPSP的输入输出曲线(I-O),发现ecto-ErbB4对I-O曲线没有影响。并且,加药前后的PPF(双脉冲易化,反映突触前的机制)同样也无显著性差异。随后,我们使用了一种ErbB4受体敲除鼠(ErbB4-/-HER4heart),以此进一步证明NRG1-ErbB4通路对外侧杏仁核的GABA能系统及LTP诱导的调控作用。首先我们比较了敲除鼠与野生型鼠中所记录的最大IPSC的差异。与野生型鼠相比(809.10±83.69pA,n=12),虽然ErbB4-/-HER4heart鼠显示了更小的IPSC幅度(619.09±72.65pA,n=18),但两者之间无统计学的差异(P>0.05,独立样本T检验)。与野生型相反,敲除鼠IPSC的I-O曲线落到了EPSC I-O曲线的下方。更重要的是,敲除鼠的IPSC/EPSC比值明显小于野生型(在15,20,25,30,35,40,50V刺激强度下,P<0.05,独立样本T检验)。以上结果进一步支持NRG1-ErbB4通路对外侧杏仁核区GABA能系统具有重要的调控作用。接下来,我们进一步比较了ErbB4敲除鼠与野生型之间LTP的差别。我们同样选择了丘脑传入给以强直刺激,刺激后在野生型鼠中记录到的fEPSP的幅度为基线的105.93±5.09%(n=7)。于此相反,在ErbB4-/-HER4heart敲除小鼠中可以诱导出幅度为121.94±3.42%的LTP (n=6),明显大于野生型(P<0.05)。此外,我们又在敲除鼠的脑片中加入BMI,以此观察其是否可进一步增加LTP,结果显示LTP的幅度为119.22±4.76%(n=5),与未处理敲除鼠相比无显著性差异。以上结果进一步支持NRG1-ErbB4通路通过GABA能系统调控LTP的诱导。综上所述,我们的研究结果提示,外侧杏仁核相对饱和的NRG1/ErbB4信号通路参与维持高水平的GABA能系统,进而阻止了LTP的产生。本研究为解释NRG1突变小鼠所表现出的情景恐惧记忆损害提供了新的实验证据,也可以增进人们对精神分裂症中情感异常机理的理解。更多还原

【Abstract】 Schizophrenia is a devastating psychiatric illness that affects about 1% of the world’s population. In recent years, by more and more people neuregulin1 (NRG1) and ErbB4, as members among multiple candidate susceptibility genes for schizophrenia, have become a hotspot of current neuroscience research.Genome-wide linkage analysis and large-scale profiling of gene expression have implicated that NRG1 and ErbB4 are the potential susceptibility genes for schizophrenia in diverse populations from Iceland, Scotland, China, Japan, and Korea. Increase in the expression level of NRG1 isoforms or ErbB4 have been reported in schizophrenia patients. And NRG1-induced activation of ErbB4 in cortex of schizophrenic patients is also increased. Furthermore, NRG1 and ErbB4 mutant mice also exhibit Schizophrenia-like phenotypes.NRG1 belongs to a family of trophic factors, and that are encoded by four individual genes (NRG1-4), of which NRG1 is the best characterized. NRG1 acts by stimulating a family of single-transmembrane receptor tyrosine kinases called ErbB proteins, which also have several isforms (ErbB1-4). Among them ErbB4 is the only autonomous NRG1-specific ErbB that can both interact with the ligand and become activated by it as a tyrosine kinase. Our present knowledge of NRG1 is mostly restricted to early development, like neuronal differentiation, migration, axon guidance and establishment etc.. Although NRG1 and its ErbB4 receptor are expressed highly in the adult rodent and human brain, little is known about their functions. The role of NRG1-ErbB4 signaling and its mechanisms underlying the pathology of schizophrenia are primarily unknown in the adult CNS.The emotional processing abnormalities were observed in patients with schizophrenia, and MRI evidences is suggestive of amygdala involvement. Recent studies showed NRG1 mutant mice exhibits deficits in physiology and behaviors like contextual fear conditioning and social interactions. As we know Fear conditioning is widely used as a model system for understanding how the brain forms and stores information about aversive emotional experiences. Research on the neural basis for the fear conditioning points to the amygdala as a key structure. And changes of long-term potentiation in the amygdala plays and important role in fear conditioning. Among the many mechanism underling LTP in amygdala, GABAergic inhibiton shows a powerful control of the LTP. For amygdala differs from other brain regions by the low firing due to a strong inhibitory tone. And given that NRG1-ErbB4 signaling regulating GABAergic inhibition is a candidate mechanism for the pathology of schizophrenia. NRG1/ErbB4 signaling may also regulate LTP in the amygdala by an GABAergic mechanism and that can also explain the changes in fear conditioning. So we decided to investigate the role of NRG1/ErbB4 signaling for GABAergic inhibition, and to see whether it regulates LTP in the lateral amygdala.Firstly, We recorded input-output relationship for EPSCs and IPSCs and confirmed the result that amygdala exhibits comparatively high GABAergic activity by the factor of IPSC/EPSC ratio. We investigate how the magnitude of the maximal IPSC differs in Lateral amygdala(LA), Hippocampus(CA1) and Cortex(parietal cortex, layer v). LA principle neurons exhibits higher IPSC amplitude (871.40±57.07pA, n=10) than that in Hip. (506.41±65.45 pA, n=10) and cortex (682.44±79.98 pA, n=9). And only between the two groups of LA and Hip. shows significant difference (P<0.05, one way ANOVA). And the IPSC I-O curve in LA lined above the EPSC I-O curve while that fell below in both hippocampus and cortex. Furthermore, the balance of inhibition/excitation in LA was significantly different from that in hippocampus (the 20,25,30,35,40V stimulus shown significant difference, P<0.5, Student T test) and cortex (the 10,15V stimulus shown significant difference, P<0.5, Student T test). Accordingly, LTP does not exist in LA in common situations, and the amplitude of fEPSP was 102.3±2.5%(n=8 slices,5 animals) after tetanus stimulation for thalamus inputs. In contrast, under the same conditions the LTP in CA1 is 158.5±11.2%(n=14 slices,10 animals). And independent-samples T test showed significant difference between the two areas for the averaged percentage after tetanus (P=0.000).This suggested the high level of GABAergic activity in LA hold the principle neurons to form LTP. To further confirm the theory, we gave thalamic inputs tetanus stimulation in LA under the application of 20μM bicuculline methiodidie(BMI, GABAA receptor blocker) to see whether it affects LTP. We found robust LTP can be induced in the LA in BMI treated slices with an fEPSP amplitude 1h after tetanus amounting to 119.2±1.6%(n=6 slices,4 animals)of its amplitude during baseline, while that in control slices is 102.3±2.5%(n=8 slices,5 animals). And they showed significant difference. This result indicates GABA-mediated inhibition suppressed LTP induction in LA.To further investigate whether NRG1-ErbB4 signaling regulates GABAergic inhibition, we recorded synaptically evoked inhibitory postsynaptic currents (IPSCs) form LA principal cells in slice incubated in the presence of kynurenic acid (1 mM, a broad spectrum ionotropic glutamate receptor antagonist), IPSCs were evoked with a bipolar stimulate electrode placed to thalamus inputs. The amplitude of IPSC was not affected after NRG1 perfusion (2nM, n=9). For each cell, IPSC amplitude was normalized to the value of mean IPSC before bath application of NRG1, after 10 min NRG1 treatment the normalized amplitude was 0.90±0.05(P>0.05). In contrast, ecto-ErbB4 (cloned protein, contain the ecto-domain of ErbB4, binding to NRG1 and inhibit its activity) and AG1478 (a potent and membrane permeable ErbB kinase inhibitor, which is known to prevent receptor autophosphorylation and signaling) remarkably attenuated the amplitude of IPSC, after ecto-ErbB4 and AG1478 treatment the normalized amplitude was 0.65±0.05 (2μg/ml, n=8, P<0.01) and 0.70±0.04(5μM, n=10, P<0.01), respectively. These results showed NRG1 have no effect on eIPSC, but the inhibitors for NRG1/ErbB4 signaling attenuate it, these suggested NRG1/ErbB4 signaling’s regulation on GABAergic inhibition is saturated in physiological conditions.In order to investigate whether NRG1/ErbB4 signaling can regulate LTP in the lateral amygdala, we exerted the next experiments of tree groups. Control group(vehicle treated, without Ecto-ErbB4 treatment):It exhibited an FP amplitude 1h after tetanus amounting to 99.87±3.15% (n=8 slices,5 animals) of its amplitude during baseline. NRG1 treated group:105.49±2.47% (n=6 slices,4 animals). ecto-ErbB4 (2μg/ml,7 slices,5 animals) treated group:117.97±3.72%. One way ANOVA for the averaged percentage after tetanus revealed a significant difference of the potentiation under ecto-ErbB4 application(F= 8.092, P= 0.003, P<0.05 compared with control and NRG1 treated group). This result indicates NRG1-ErbB4 signaling controls LTP induction in LA, and the signaling stands in a saturated state only can be negatively regulated.At the same time, ecto-ErbB4(2μg/ml) application show the same effects on LTP when tetanic stimulation was given to the cortical afferents. It exhibited an FP amplitude 1h after tetanus amounting to 123.05±4.82% (n=8 slices,5 animals) of its amplitude during baseline, while that in control slices amounting to 107.59±2.90%(vehicle treated,6 slices,5 animals). Independent-samples T test for the averaged percentage after tetanus revealed a significant difference of the potentiation under ecto-ErbB4 application (P=0.025). This result indicates NRG1-ErbB4 signaling take the same effect for the cortical afferents as that for thalamic afferents, shown NRG1-ErbB4 signaling have no selective effects on the two input pathways.Because blocking GABAA receptors or NRG1/ErbB4 signaling both enhanced the tetanus-induced LTP in LA. Thus, we hypothesized that ErbB4 inhibitor enhance the LTP induction by decreasing GABAA mediated inhibition in the LA. To test this hypothesis, we examined the effects of application of only ecto-ErbB4(2μg/ml) or both ecto-ErbB4(2μg/ml) and 20μM BMI on tetanus-induced LTP in amygdal slices, and to see whether they can produce synergistic or additive effects. We found that there are greatly enhanced tetanus-induced LTP in both two groups, and they showed almost the same potentiation extent on LTP. For the ecto-ErbB4 and BMI treated group, it exhibited an FP amplitude 1h after tetanus amounting to 120.76±5.68% (n=8 slices,4 animals) of its amplitude during baseline, while that in only BMI treated group amounting to 119.23±1.61%(20μM,6 slices,4 animals)). Independent-samples T test for the averaged percentage after tetanus revealed no significant difference between the two groups(P=0.825). These results suggested that NRG1/ErbB4 signaling regulates LTP induction through the GABAergic inhibition.Beside the effects of ecto-ErbB4 on LTP, it had no effect on basal synaptic transmission. We performed complete input-output (I-O) curves at a series of increasing stimulation intensities, after 20 min of ecto-ErbB4 superfusion additional I-O curves were performed at the same stimulation, and between them no obvious changes were detected. Also, ecto-ErbB4 superfusion has no influence on PPF that reflect an impact on presynaptic mechanisms.Next, we used a kind of ErbB4 knockout mice (ErbB4-/- HER4heart). We want to see how genetic alterations of NRG1-ErbB4 signaling influents GABAergic inhibition and LTP induction in the LA. Again, we recorded input-output relationship for EPSCs and IPSCs to investigate whether ErbB4 knockout mice have impaired GABAergic inhibition by the factor of IPSC/EPSC ratio. We first compared the maximal IPSCs between knockout mice and WT mice. The LA principle neurons of ErbB4-/-HER4heart mice exhibits smaller IPSC amplitude (619.09±72.65pA, n=18) than that in WT control (809.10±83.69pA, n=12). But the difference between them was not significant (P>0.05, student T test). Also the IPSC I-O curve in LA from ErbB4-/-HER4heart mice fell below the EPSC I-O curve while that lined above in the WT control mice. Even more important, the balance of inhibition/excitation in LA from ErbB4-/-HER4heart mice was significantly different from that in WT control mice (the15,20,25,30,35,40,50V stimulus shown significant difference, P<0.5, Student T test). This result further supported that NRG1-ErbB4 signaling plays and important role in regulating GABAergic inhibition.In the next step, we evaluated the difference of LTP induction between the ErbB4 knockout mice and WT mice. Tetanic stimulation of thalamic afferents can not induce LTP in the LA in control slices of WT mice. It exhibited an FP amplitude 1h after tetanus amounting to 105.93±5.09% (n=7 slices,4 animals) of its amplitude during baseline. In contrast, in the ErbB4-/-HER4heart mice (6 slices,4 animals), we found robust LTP can be induced in the LA with an FP amplitude 1h after tetanus amounting to 121.94±3.42% of its amplitude during baseline. It was significantly bigger than that in WT mice (P＜0.05). We further applied BMI for ErbB4-/-HER4heart mice slices to see whether they have synergic effects, the BMI treated group (5 slices, 3 animals) exhibited and LTP of 119.22±4.76%. And significant difference cannot be seen between ErbB4-/-HER4heart mice and BMI treated slices. This evidence confirmed the result that NRG1/ErbB4 signaling regulate LTP induction through GABAergic inhibition.In conclusion, our results here showed the saturated NRG1/ErbB4 signaling activity holds the robust GABAergic inhibition and suppressed the LTP induction in the lateral amygdala. This mechanism might explain the impairments of contextual fear conditioning in NRG1 mutant mice, and contribute to emotional dysfunctions in subjects with schizophrenia.更多还原