非同源性染色體轉位在人類生殖細胞發生機制之探討

Abstract

人類基因體計畫在2003年完成後，許多研究開始探討存在人類基因體的變異。國際人類基因體圖單型體計畫於2002年底至2007年底，已定出超過三百萬個單一核苷酸多型性(SNPs)位點。2004年開始，研究團隊以不同的策略，在表現型正常的個體間偵測>1kb之變異(結構變異)，到目前為止，人類基因體變異數據庫(Database of Genomic Variants, DGV)中的結構變異資料，總數已達將近四萬個，顯示出結構變異在構成人類基因體變異的重要性。 對結構變異做genome-wide的偵測，主要以array-based的策略進行，受限於方法的特性與解析度，故無法對平衡結構變異做偵測或將結構變異精確定位。實驗室先前建立了Restriction Enhanced Capturing of Rearranged DNA(RECORD)，此技術以Inverse PCR為基礎，用於目標基因結構變異的偵測，配合序列分析，進而可推測其可能的發生機制。 實驗室先前已利用RECORD技術對在新生兒及幼兒型血癌常見發生基因轉位的MLL，及在幼兒型骨癌常見發生基因轉位的EWSR1，在人類精蟲細胞中偵測到二種基因之轉位。本論文的目的是除選用已知基因轉位與兒童型淋巴性白血病相關的TCF3當作目標基因，探討在人類生殖細胞當中是否有類似結構變異牽涉到此基因；若有，則其可能的發生機制為何。另依據先前的實驗結果，顯示MLL-F2 (MLL-fragement 2)偵測到的染色體轉位斷裂點呈現明顯叢聚分布，故以此片段之環化模板利用RECORD技術配合基因轉位特異性引子進行確認試驗。 首先進行以HEK293與K562細胞株基因體混合DNA作為模版，檢測RECORD技術的偵測敏感度，約為5×10-4。接著在12個不同的精蟲細胞單套基因體DNA，共偵測到19個TCF3的結構變異，均為非同源性染色體轉位。經由序列對比判定4個轉位對象座落於基因內(geneic region)，另15個則位於基因間(intergenic region)。而就染色體轉位之機制而言，斷裂點周邊重複序列並未參與，我們可排除其為同源性重組(homologous recombination, HR)造成之可能。而在染色體轉位之斷裂點交界處均有1-13個核苷酸的微同源性序列(microhomology sequence)，與在體細胞由NHEJ及MMEJ(microhomology mediated end-joining)所造成的基因轉位類似。由於NHEJ核心因子Ku70/Ku80在減數分裂時並未表現，故並非NHEJ所造成。推測這些染色體轉位可能由類似MMEJ的機制所產生。 針對3個先前由同一捐贈者之精蟲細胞基因體DNA在MLL-F2偵測到的染色體轉位進行確認，但均未能被確認。故知其發生頻率可能低於或恰位於RECORD技術偵測之極限，應以敏感度更高的取代方法來驗證。此外，由十二個不同捐贈者的精蟲細胞基因體DNA，在MLL-F2偵測到20個染色體轉位。轉位對象8個位於基因內(geneic region)，12個位於基因間(intergenic region)。在轉位斷裂點交界處均具有2-10個核苷酸的微同源性序列(microhomology sequence)。 若將目標基因序列調整為正股時，依轉位對象位於染色體長臂或短臂、microhomology位於正股或負股及目標基因以5’端或3’端序列發生轉位，產物形式可分為八種，以5’或3’ PCR分別可偵測到四種。我們發現偵測到的染色體轉位均可依上述規則分類，由5’及3’ PCR均偵測到預期形式的產物，且屬於平衡互換(balanced exchange)；此外，亦顯示出這些轉位的發生需遵循微同源性序列(microhomology sequence)間的base-pairing rule，應為微同源性序列媒介的機制。為了與在體細胞發生的MMEJ有所區隔，我們將之命名為MMIT(microhomology mediated interchromosomal translocation)。 生殖細胞進行減數分裂重組的重要目的即是在修復減數分裂時產生的DNA雙股斷裂(double strand break)。由於我們在不同來源的精細胞基因體DNA中，普遍均能偵測到基因轉位的存在，且這些序列之特性，明顯異於由同源性重組(HR)修復者，故我們推測MMIT應屬於人類生殖細胞中修復DNA雙股斷裂的機制之ㄧ。

Since the human genome project finished in 2003, many studies focused on the variations in the human genome. From 2002 to 2007, the international HapMap project mapped over 3.1 million single nucleotide polymorphisms (SNPs). Using different approaches, until recently, about 40000 >1kb structural variations (SVs) had been detected between non-disease phenotype individuals. Theses observations reflected the important contribution of SVs to human genome variation. To interrogate which mechanisms involve in SVs formation in human genome, PCR-based approach is the most sensitive one. Previously, an inverse PCR-based approach, RECORD (Restriction Enhanced Capturing of Rearrange DNA), was used to identify MLL and EWSR1 translocations in human germline. MLL and EWSR1 are frequently involved in recurrent translocation associated with infant and childhood leukemia and childhood sarcoma, respectively. The aim of my project is to use human TCF3, which is a known target of recurrent translocation associated with childhood acute lymphoblastic leukemia (ALL), as an anchor gene to examine whether similar event occurs in human sperm and to explore the underlying mechanisms involved. Moreover, we will use RECORD and translocation specific PCR to validate germline translocations which identified previously. To assess the detection limit of RECORD, we used HEK293 spiked with K562 genomic DNA as template, and examined the recovery of BCR-ABL fusion from K562 cells. We found that the detection limit by RECORD is about 5×10-4. By using TCF3 as an anchor gene, we identified 19 non-homologous interchromosomal translocation events from 12 donors. These 19 translocation events include four translocation partners in geneic region and 15 in intergenic region. Sequence analysis of the 2 kb regions flanking both sides of breakpoints excluded the involvement of non-allelic homologous recombination (NAHR). A microhomology of 1-13 base pairs at the breakpoint junction indicates that either non-homologous end joining (NHEJ) or microhomology mediated endjoining (MMEJ) may be responsible for germline translocation. though the core factors, Ku70/Ku80 of NHEJ are not expressed during meiotic recombination. Using MLL as an anchor gene, we identified 20 additional MLL-F2 translocations from the haploid genomes of 12 sperm donors. Eight of which were mapped in the genic region and 12 in intergenic region of partners. At the breakpoint junctions, there are microhomologies of 2-10 nucleotides. When microhomology of the anchor gene is considered always on the top strand, the translocation partners can be classified into four groups, based on its location on the p or q arm, and its sequence from top or bottom strand. Thus 8 derivatives of translocation products can be defined by using 5’ and 3’ iPCR in our studies. This germline balanced exchange between non-homologous chromosomes is mediated by microhomology. To distinguish from MMEJ reported only in somatic cells, we proposed this putative mechanism as microhomology mediated interchromosomal translocation (MMIT). As the important purpose of meiotic recombination is to repair DNA double-strand breaks (DSBs), we suggest that MMIT may be another meiotic DSBs repair pathway in male germ cells.