|Abstract: ||阿茲海默症是由於一種蛋白質在腦堆積形成斑塊和糾結而導致一種漸進性的神經退行性疾病。目前研究已知APP、PSEN1、PSEN2以及APOE這些常見危險因子之基因突變與阿茲海默症有高度致病風險。致病研究的探究雖然已行之多年，但是對於阿茲海默症的發病機制仍然尚未明瞭。因此，我們透過微陣列數據尋找正常腦老化以及因阿茲海默症病患特定五個腦區之差異共表達(DEC)基因對，五個腦區分別為:嗅皮層(entorhinal cortex, EC)、海馬 (hippocampus, HC)、內側顳中回 (medial temporal gyrus, MTG)、後扣帶 (posterior cingulate,PC)、額上回 (superior frontal gyrus, SFG)。利用差異共表達基因對分析，辨別基因與基因之間的關聯-共表達增加 (GOC) 或共表達失去(LOC)。接著利用蛋白質資料庫與差異共表達基因對交集，建立出具有蛋白質交互作用的核心基因網路。|
結果顯示在老化以及五個阿茲海默症不同腦區核心基因網路的基因有顯著的不同。老化的基因GOC數量大於LOC的數量，但是阿茲海默症因腦區而有所不同。在核心基因網路中阿茲海默症之不同腦區的樞紐基因也為已經知道與阿茲海默症有相關的基因，包含: TUBB、GAPDH、NEFL和AMPH。同樣的許多老化組的樞紐基因也為已經知道與老化有相關的基因，包含: YWHAZ, HSPAB, EEF2, 和HIF1A。在老化組研究當中也發現YWHAZ和XRCC6為老化組和阿茲海默症組都有的樞紐基因。
在功能路徑的結果顯示正常老化和阿茲海默症五個腦區的不同，例如:阿茲海默症病人的海馬回 (HC) 中並未發現任何異常癌症相關的路徑，但是在老化和後扣帶區(PC)顯著表現。大多數的功能路徑不是屬於GOC就是屬於LOC，但是也有少數例外:蛋白酶體以及大腸桿菌相關路經，它們同時在GOC和LOC中都有出現。;Alzheimer′s disease (AD) is one of most prevalent progressive neurodegenerative disease. It is believed that main pathology of AD in brain is due to accumulated neuronal plaque and tangle formations. Previous studies indicated that abnormally in the genes APP, PSEN1, PSEN2, and APOE are high risk factors to AD. Recent reports have also suggested that infec-tions from foreign pathogens may cause AD. In spite of much research efforts the underlying pathophysiology of AD remains unclear. Here, our goal is to gain insight into possible new genetic causes of normal brain aging (AG) and brain region-specific AD, Also, to identify changes in functional pathways associated with AG and region-specific AD, and to elaborate on the heterogeneity of AD in five brain regions.
The data we use for our study six sets of publically available microarray dataset: one set from brains tissues of people of a wide range of ages not known to have AD or similar brain disorders, and five sets from tissues of five brain regions – entorhinal cortex, hippocampus, medial temporal gyrus, posterior cingulate, and superior frontal gyrus – of AD patients and controls. We use the method of differential co-expression (DCE) analysis for our study. We identified pairs of genes whose correlation of expression levels in control versus test data in-creased significantly (gain of co-expression, or GOC), and those whose correlation decreased significantly (loss of co-expression, or LOC), integrated all the genes involved in GOC and LOC pairs with protein-protein interaction data to construct six core protein-protein interac-tion networks (cPPINs), one for each set of data, and used the networks for hub-gene identifi-cation (high-degree genes in cPPIN) and for pathway studies.
There was significant difference between the AG cPPIN and the AD cPPINs, and among regional AD cPPINs. The number of GOC genes greater than LOC genes in the AG cPPIN, but the GOC to LOC ratio is highly region-dependent in the AD networks.
Many of hub-genes in the AD cPPINs were also known AD target genes, including TUBB, GAPDH, NEFL, and AMPH. Similarly, many of hub-genes in the AG cPPIN were known AG target genes, including YWHAZ, HSPAB, EEF2, and HIF1A. YWHAZ and XRCC6 were among gene that were AG as well as AD hub-genes and were both AD and AG known target genes.
Classes of enriched pathways differed between AG and AD, and intra-regionally in AD. For instance, cancer pathways were seen to be highly enriched in AG, and PC/AD, but absent in HC/AD. Most pathways were either GOC or LOC enriched, but some, including pro-teasome and E. coli related pathways, were both GOC and LOC enriched.