以方法學的觀點探索小鼠腦中之基因組印記

Abstract

基因組印記是一個會受制於父方或母方來源的染色體之影響的現象，經由此模式而造成基因只從其中一方來源的染色體上表達。印記基因的異常牽涉到許多神經或精神上的疾病，而其在腦中所扮演的角色至今仍尚未釐清。基因組印記的表達是相當動態的，不同時間、空間、物種、外在環境、組織以及細胞型態都會調控其表現情形。神經幹細胞位於中樞神經系統具有自我更新及分化成新神經的能力。神經幹細胞對於大腦是相當重要的，且與許多發育、神經退行性及精神疾病皆有關聯。事實上，已有兩篇研究指出印記基因與神經幹細胞之間的關聯性。因此，我們想鑑定出所有小鼠神經幹細胞中的印記基因。我們利用螢光激活細胞分選技術來收集神經幹細胞，抽取核醣核酸後送去做次世代定序。然而免疫螢光染色及次世代定序的結果，皆不如我們的預期。最後我們試著尋找其他的小鼠模型來達到我們的研究目標。 由於X染色體在男女性別間具有不對稱遺傳的特性，進而造成與X染色體有關的基因組印記會伴隨著性別雙型性。根據本實驗室先前的核醣核酸定序資料，我們發現Ndufb11穩定的表現母方來源的等位基因。其原因為Ndufb11位於X染色體上，而我們先前又都使用公小鼠的核醣核酸來做定序。其實不只是Ndufb11，還有X染色體上的所有基因都會假性地表現母方來源的等位基因。因此，這次我們使用母小鼠的核醣核酸，且利用核酸質譜分析技術來驗證Ndufb11到底是否為一個與X染色體有關之母方表達印記基因。有趣的是，我們的數據指出Ndufb11在小鼠視覺皮質是一個特定種族 (CAST/EiJ) 之單方等位表達印記基因。 一般來說，我們都知道印記是動態的且會受到外在環境所調控，因此我們想瞭解印記基因是否會受到光線調控。我們透過將小鼠飼養在正常環境 (12小時光照、12小時黑暗) 以及無光線飼養 (24小時全黑暗) 的方式操弄小鼠模型的生活環境。然後，我們分別測量有光線及無光線飼養小鼠運動皮層的印記基因之核醣核酸表現量。理論上，我們推測他們之間應該會沒有顯著的差異。令人驚訝的是，我們發現在21個已知的印記基因中，Airn-160932和Copg2-48774的表現量在有光線及無光線飼養小鼠運動皮層間顯著地改變，但卻在小鼠視覺系統 (視覺皮層、視叉上核、視網膜) 維持穩定的表現量。我們認為在有光線及無光線飼養小鼠運動皮層間，此兩個印記基因表現量的改變可能是受到光線調控的次級效應所影響，而非初級效應。然而，這樣的假說還需要未來更多的證據支持才得以證實。 基於本實驗室先前的桑格定序以及核酸質譜分析資料，我們發現Ago2可能在小鼠視覺皮層之其他細胞型態中為一個局部母方表達印記基因，而不是在興奮性神經元、抑制性神經元及星形膠質細胞中。我們成功地在本實驗室建立了核醣核酸之螢光原位雜合技術，但是我們驚訝地發現，組織切片在製備中的煮沸步驟處理後，內源性的螢光會消失。這樣一來，我們就無法利用標記特定細胞的基因轉殖小鼠之腦切片，來觀察Ago2核醣核酸在該特定細胞的表現情形。即使如此，我們還是成功地利用核醣核酸之螢光原位雜合技術展現出Ago2之單方等位或雙方等位表達型態。總而言之，若我們未來想在特定細胞偵測基因的單方等位或雙方等位表達，結合核醣核酸之螢光原位雜合技術及免疫螢光染色可能是必須的。

Genomic imprinting is a parent-of-origin effect which causes monoallelic gene expression. Dysregulation of imprinted genes is involved in various neurological and psychiatric disorders, but their roles in the brain are still unclear. Genomic imprinting is spatiotemporally dynamic, and varies between different species, experiences, tissues and cell types. NSCs have the capacity to self-renew and differentiate into new neurons of the central nervous system. NSCs are crucial to brain and are associated with several developmental disorders, neurodegenerative and psychiatric diseases. In fact, there had been already two studies reported the relationship between imprinted genes and NSCs. Consequently, we wanted to identify all the imprinted genes in the mouse NSCs. We used FACS technique to collect NSCs, extracted the RNA and sent for RNA sequencing (RNA-Seq). However, the immunofluorescence staining and RNA-Seq results were not as good as we hoped. We then tried to find another mouse model to achieve our goal. X-linked imprinting can accompany with sexual dimorphism because of the asymmetrical inheritance of the X chromosome. Based on the previous RNA-Seq data in our lab, we found that Ndufb11 revealed stable maternal expression. That was because Ndufb11 is located on X chromosome, and we used male mouse sample to do the RNA-Seq. Actually, not only Ndufb11 but all the genes located on X chromosome would show falsely maternal expression. Therefore, this time we used female mice sample and MassArray system to validate whether Ndufb11 was an X-linked maternally expressed imprinted gene or not. Intriguingly, our data indicated that Ndufb11 was a strain-specific (CAST/EiJ-specific) monoallelic expressed gene in mouse visual cortex. Generally speaking, we all known that imprinting is dynamic and can be regulated by the environment; hence we wanted to understand whether the imprinted genes can be regulated by the light. We manipulated the environment of mouse model by maintaining them in normal circadian cycle (normal-rearing) or complete darkness (dark-rearing). Then, we measured the imprinted genes mRNA expression level from normal- and dark-reared mouse motor cortex respectively. Theoretically, we supposed that it would be no significant change in normal- and dark-reared mouse motor cortex. Surprisingly, we found that among 21 known imprinted genes, Airn-160932 and Copg2-48774 expression changed significantly in mouse motor cortex under light manipulation, but their expression remain stable in mouse visual system (visual cortex, SCN and retina). We considered that the change of these two imprinted genes in normal- and dark-reared mouse motor cortex was possibly due to the secondary effect but not the primary effect of the light manipulation. However, this hypothesis still needs more evidence for further verification in the future. On the basis of the previous Sanger sequencing and MassArray data in our lab, we discovered that Ago2 might be a partial maternally expressed gene not in the excitatory neurons, interneurons and astrocytes but in the other cell types in mouse visual cortex. We successfully established the RNA-FISH technique in our lab whereas we surprisingly found that the endogenous fluorescence disappeared after the boiling step of tissue sections preparation. Accordingly, we could not use the cell type-specific Cre transgenic mice to observe the Ago2 mRNA expression in specific cell types. Even so, we still succeeded to show both the monoallelic and biallelic expression pattern of Ago2 mRNA by using RNA-FISH technique. In summary, if we want to detect monoallelic or biallelic gene expression in specific cell types in the future, combine the RNA-FISH technique and immunofluorescent staining may be necessary.