巨核細胞分化時染色體間之互動結構

Abstract

於1990年，美國國家衛生院和能源部提出了人類基因體計畫並在2003年將人類基因體定序完成。此計畫的目的是希望能夠針對演化、發育、分化和疾病的產生做更深入地探討。早期在研究人類基因體時都認為DNA是單純的由線性序列組成，但事實上會經過纏繞和折疊形成具有三度空間結構的染色體。染色體的空間結構組織對基因體的功能有很大的影響力，例如DNA複製、轉錄和DNA修復；除此之外，染色體三度空間結構會更進一步地影響一些生理功能，例如細胞發育和分化，免疫反應和疾病產生。舉例來說，在紅血球分化過程中α-globin的轉錄會受到染色體空間結構的調控。 在早期研究染色體結構時最困難的是技術上的問題，並沒有一個很理想的方法來觀察染色體的空間結構。在2002年，Dekker等人建立了專門用來研究DNA結構的技術，叫做chromosome conformation capture (3C)；此方法的確改善了過去技術上的缺點，因此近年來對DNA結構的了解也愈來愈清楚。但是3C的缺點就是必須對目標基因的順式作用因子等等很了解才比較適合拿來做探討，因此在2006年將3C更進一步修改成4C，叫做circular chromosome conformation capture。此方法的優點就是能利用某一段DNA來找出許多未知的互動區域，染色體內和染色體間的互動結構都可以偵測到。 我的論文主要想研究在TPA刺激後的K562細胞中，染色體結構的互動情形。選擇在巨核細胞分化過程中會表現增加的ITGA2B當研究對象，利用4C技術偵測和此基因的啟動子有互動的區域。我們總共找到了11個染色體間的互動區域，其中8個位在基因內，剩下的3個位於兩基因間。ITGA2B和KSR2間的互動僅在TPA刺激後的K562才可以被驗證，可見在巨核細胞分化過程中染色體有結構改變情形發生。利用半定量RT-PCR偵測互動基因的表現，發現KSR2和NAV2為誘發性表現，而PLEKHA5、PCSK7、DACH1、LPHN3、SFRS15和C1orf125都是持續性表現。這些結果顯示會表現的基因在細胞核內容易聚集在一起，因此推測此為一個轉錄工廠，這些基因共享著相同的轉錄因子複合物。 同時利用生物資訊方法預測位於這些互動基因上游1kb的啟動子區域的轉錄因子，發現除了LPHN3之外，其他基因都類似ITGA2B缺少了TATA box。所以此工廠如果真的存在，應是供給缺少TATA box的啟動子使用，不論是誘發性還是持續性表現的基因。雖然沒有TBP存在，但帶有其他轉錄因子 (例如SP1和ETS)可能可調控這些互動基因的表現。更進一步用生物資訊方法分析互動區域上的轉錄因子發現互動區域和ITGA2B啟動子上的轉錄因子都不一致，且位於SFRS15的互動區域還沒有任何轉錄因子存在，這些結果顯示出轉錄的動態機制，即轉錄進行時DNA會沿著不動的蛋白質複合物移動，此不動的複合物即為轉錄工廠中的轉錄因子組成的。

The Human Genome Project was initiated by the National Institute of Health and the Department of Energy of the United States in 1990 and completed its ultimate goal of decoding the genomic sequence of humans in 2003. A complete book of life enabled us to study evolution, growth and development, and diseases in a genome-wide scale. The human genome was conventionally considered as a contiguous, linear sequence entity. In fact, DNA is folded into multilayered, high-order structures that eventually packaged into chromosomes. The spatial organization of chromatin dramatically affects the function of the genome, including DNA replication, transcription, and repair. Furthermore, physiological processes at the organismal and cellular levels, such as differentiation, development, immune response and pathogenesis, are greatly influenced by the three-dimension (3D) chromosomal structures in the nucleus. For example, the transcription of mammalian α-globin gene is determined by the unique spatial arrangement of chromatin structure during erythroid differentiation. In the past, the analysis of 3D chromatin structures was limited by technical difficulties. Recently, a new methodology, termed chromosome conformation capture ( 3C ) has been developed to study chromosomes that are physically associated with nuclear protein complex. Such interactions require a prior knowledge of defined chromosomal regions. In order to examine physical associations in the unknown regions, a modified, high-throughput method of circular chromosome conformation capture ( 4C ) was developed to detect the intra- and interchromosomal interactions. The objective of my thesis research was to describe the structural organization of chromosomal interactions in human K562 cells during TPA-induced megakaryocytic differentiation. We adopted the 4C approach to investigate chromosomal interactions with the promoter sequence of the ITGA2B gene that is up-regulated in TPA-treated K562 cells. Altogether we identified 11 interchromosomal interactions. Eight of which were involved in the intragenic regions, whereas the remaining three were mapped to the intergenic regions. In this study, we also confirmed the interaction between ITGA2B and KSR2 only in the TPA-induced differentiation of K562 cells. By using the semi-quantitative RT-PCR assay to examine the transcripts of the ITGA2B-interacting genes, we found that the expression of both KSR2 and NAV2 was inducible after TPA treatment and that PLEKHA5, PCSK7, DACH1, LPHN3, SFRS15 and C1orf125 were constitutively expressed in K562 cells. These results suggest that actively transcribed genes are potentially clustered together in a nuclear space, as proposed to be in a transcription factory, which can be detected by 4C through the physical association with shared transcription complex. By using the bioinformatics approach, we also explored the transcription factor (TF)-binding motifs on the promoter sequences in the 1kb regions upstream of the ITGA2B-interacting genes. All but LPHN3, like ITGA2B, contain TATA-less promoters. It is likely that such a transcription factory, if indeed exists, may transcribe specifically the housekeeping and inducible genes without the TATA promoter with a transcription complex incorporating other TFs, such as SP1 and ETS. Further examination of the TF-binding motifs on the interactor sequences revealed that the potential binding sites, except SFRS15, were different from the ITGA2B promoter. These results likely reflect the dynamic nature of transcription machinery, in which during transcription the DNA sequence moves through immobilized protein complex.