藉由Akt1基因缺損小鼠及P19細胞株去探討AKT1對於γ-氨基丁酸訊號傳遞及相關認知功能

Abstract

精神分裂症是一嚴重精神疾病，其盛行率約1%。主要的症狀有正性症狀、負性症狀及認知缺損。在精神分裂症的研究中有兩個很重要的主題：一是致病機制，另一個是尋找有效的治療藥物。越來越多證據顯示AKT1跟γ-胺基丁酸(GABA)的A型受體皆為精神分裂症致病的候選基因。在病人死後的組織上也發現到GABA相關的神經元有減少的現象。為了釐清特定生理因子跟認知功能的因果關係，基因剔除小鼠是一個適當的生物模型。因此在第一個研究中我們使用Akt1基因剃除小鼠去研究AKT1、GABA突觸跟相關行為的因果關係。在Akt1基因剃除的母鼠海馬迴發現特定的GABAergic 神經元(含有parvalbumin)會減少50%，細胞膜上的GABA的A型受體表現量也下降。此外也發現有在海馬迴所記錄的神經震盪及海馬迴相關行為作業(Y行迷津及莫氏水迷津)有顯著地改變。然而為了模擬病人體內AKT表現的程度，Akt1異型合子的小鼠 (Akt1 HET)是相對合適的選擇。針對過去在Akt1基因剃除的母鼠發現到的行為缺損，在第二部分的研究，使用Akt1 HET的母鼠，並分別使用懸尾作業及Y型迷津去測試類憂鬱行為及空間記憶表現。同時使用GABA的促進劑valproate來測試是否可以回復這些行為缺損。結果顯示Akt1 HET的母鼠只有在懸尾作業發現缺損而非Y型迷津，此一缺損可以被valproate的長期注射回復。然而使用動物模型仍然有一些美中不足之處。例如實驗動物的飼養與繁殖相對耗時費力，生物體內也有可能會產生補償作用等。而這些問題，可以透過使用細胞株先進行初步實驗來克服。也適合做為大量篩選藥物的初期生物模型。因此在第三部分的研究，藉由ASCL1導致P19細胞株進行神經元分化的現象，嘗試發展成一個初步研究的實驗平台，以鋰鹽對於AKT所參與的神經元分化及神經元軸突成長有何作用作為範例。首先，使用AKT 1/2的抑制劑先測試AKT在GABA神經分化(DIV 5)及神經元軸突成長(DIV 3)所扮演的角色。結果顯示含有GAD67的神經元有40%的下降，含有parvalbumin的神經元有60%的下降，但細胞膜上的GABA A型受體則無顯著下降。同時也觀察到AKT 1/2抑制劑也會使得神經元軸突生長長度下降60%的現象。因此在其後的鋰鹽實驗中，進一步證實鋰鹽可以隨著劑量的增加有效地回復AKT 1/2抑制劑所導致的神經元軸突長度下降的情況。綜合以上三個部分的研究結果，支持Akt1基因剃除小鼠中觀察到的GABA突觸改變的直接原因是AKT1失去作用而非補償作用所造成的現象，此外鋰鹽跟valproate都是在此現象具有治療潛力的藥物。

Schizophrenia is a serve mental disorder with 1% prevalence comprising positive, negative and cognitive symptoms. There are two important fields in the study of schizophrenia: the etiology of disorder and the discovery/development of potential antipsychotic drugs. Accumulating evidence suggests that AKT1 and GABA (A) receptor were both candidate genes of schizophrenia. The reduction of GABA system was reported in the postmortem tissue of some schizophrenic patients. For clarifying the cause effect between specific biological components and cognitive functions, gene knockout mice provide a feasible approach for elucidating causal relationships between susceptibility gene(s) and related functions. Taking advantage of knockout (homozygous) mice, we study causality between Akt1gene and GABA-related function in study 1. We found that ~50% reduction of parvalbumin-positive interneurons, a subpopulation GABAergic interneurons, was found in hippocampus of female Akt1 knockout mice but no difference was found in calretinin-positive interneurons. A reduction of GABA (A) receptors subunit, β2 subunit, was also oobserved on plasma membrane of hippocampal neurons. Interestingly, Akt1 knockout female mice further displayed impaired hippocampal oscillation power and behavioral deficits in hippocampus related cognitive functions (especially in Y-maze task and Morris water maze). Our data in study 1 suggest that Akt1 deficiency resulted in the alteration of specific GABA synapse which might lead up to the impairment of hippocampus-dependent cognitive functions in female Akt1 knockout mice. In contrast to Akt1 homologous mutant mice, the use of Akt1 heterozygous (HET) mutant mice offers a more feasible tool to mimic human subjects with a genetic deficiency of AKT1 and to evaluate drug effects in vivo. In study 2, female Akt1 heterozygous mice were used to evaluate the rescue effect of valproate, the GABA function facilitator, on the behavioral deficits in Tail suspension test (TST) and Y-maze task. Our result revealed that female Akt1 HET mice only displayed impairment in TST but not in Y-maze. Such impairment can be rescued by chronic injections of valproate. In study 3, to minimize time spent on mouse breeding and to reduce potential compensatory effect in mutant mouse models, a cell culture model was developed as a high-throughput platform for in vitro drug screening. The in vitro model using Ascl1 to differentiate P19 embryonal carcinoma cells into neurons was established and AKT1/2 inhibitor resulted in a reduction of neurite outgrowth and neuron differentiation. Taking advantage of this model, the rescue effects of lithium on the neurite outgrowth in DIV 3 and the expression of GABAergic neuron in DIV 5 were examined. Our data revealed a 40% reduction of GAD67-positive neurons and a 60% reduction of parvalbumin-positive neurons in neurons treated with AKT1/2 inhibitor. The AKT 1/2 inhibitor also resulted in a 60% reduction of the neurite length in DIV 3. Importantly, the reduction of neurite outgrowth observed in AKT inhibitor-treated neurons can be rescued by the treatment of lithium. Taken together, these findings suggest that the deficiency of Akt1 has a causal-effect on the alteration of GABAergic systems which can resulted in the impairment of GABA-related hippocampal functions in Akt1 mutant female mice. Lithium and valproate might have some therapeutic potentials in the development of treatments for Akt1-related GABAergic impairments.