長鏈非編碼RNA在發育之運動神經元所扮演之角色

Abstract

近年來的研究顯示，人類完整轉錄體（transcriptome）裡至少含有超過一半以上是屬於非編碼RNA（non-coding RNA，ncRNA），它們不像大家熟知的信使RNA（mRNA）會被轉錄成有功能性的蛋白質，而是直接以RNA的形式去執行其任務。至於ncRNA如何參與胚胎發育相關基因的調控一直是生命科學界非常感興趣的課題。先前的研究文獻指出，這些ncRNA在發育的過程中會大量的在神經系統中表現，但在生物體內的功能仍然隱晦不明。因此，我們藉由體外培養的方式在實驗室裡將小鼠胚胎幹細胞大量分化成運動神經元後，近一步執行次世代定序（strand-specific RNA-seq）的實驗來鑑定出運動神經元特異性表現的長鏈型非編碼核糖核酸(long non-coding RNAs, lncRNAs)。在這個分化的過程中，我們發現在哺乳動物的基因組中，一個稱為Dlk1-Dio3的銘印基因群（imprinted genes）會從母源染色體產生一群長鍊鏈型非編碼核醣核酸，其中一群最廣為探討的長鏈非編碼核糖核酸為Meg3（又名Gtl2），Rian以及Mirg會在成熟的運動神經元中大量表現。雖然這個銘印基因群組在幹細胞及腫瘤細胞的研究領域裡已經很常被探討，但是對於它們在神經發育的過程中卻還未被深入的探討研究。另外，我們的研究發現，Meg3，Rian以及Mirg三個長鏈非編碼核糖核酸可以協同組蛋白甲基轉移酶（Ezh2 and Jarid2）的活性作用，當這群長鏈非編碼核糖核酸被移除時，對於它們所催化的H3K27me3組蛋白修飾也會跟著活性下降，導致於神經先驅細胞基因及尾部端表現之Hox基因 (Hox8) 在運動神經元中異常表現。目前的實驗結果顯示Meg3, Rian以及Mirg三個長鏈非編碼核糖核酸，尤其是Meg3，在維持運動神經元細胞命運的過程中，無論是抑制先驅細胞基因的表現使成熟運動神經元細胞能夠正常表現，或是決定運動神經元亞型的細胞特性都扮演著極重要的角色。

Long noncoding RNAs (lncRNAs) are an emerging new class of mediator that regulate epigenetics and gene expression. However, only a small set of lncRNAs have been examined in vivo through genetic deletion in animal models (Sauvageau et al., 2013). Additionally, it remains elusive whether lncRNAs function to regulate gene expressions in the central nervous system (CNS). In my PhD study, I explored the lncRNAs that were essential for spinal cord development by profiling cell types at each stage during embryonic stem cell (ESC) differentiation into motor neurons (MNs). Among the lncRNAs candidates, I uncovered that three lncRNAs Meg3 (i.e., Gtl2), Rian and Mirg, which are expressed from the maternally inherited homolog in Dlk1-Dio3 imprinted locus (Lin et al., 2003a), are predominantly and gradually enriched in rostral motor neurons. Although Meg3 has well-characterized functions in stem cell and tumor contexts, its role during motor neuron development is unknown. A recent study revealed that Meg3, Rian and Mirg facilitate Ezh2 and Jarid2 interaction to direct H3K27 trimethylation (H3K27me3) in vitro (Kaneko et al., 2014b). In my study, I found that loss of these lncRNAs compromises the H3K27me3 landscape, leading to aberrant expression of progenitor and caudal Hox genes in postmitotic MNs. My results thus illustrate that these lncRNAs in the Dlk1-Dio3 locus, particularly Meg3, play a critical role in maintaining postmitotic MN cell fate. They shape MN subtype identity by repressing progenitor genes as well as regulating caudal Hox genes. All together, these findings not only provide critical information in regards to lncRNA function, but also in understanding the specific role of lncRNAs during motor neuron development.