肺癌之基因體缺失及其與雙股斷裂修補基因多型性之相關性研究

Abstract

研究背景：癌症的形成常常是由於腫瘤抑制基因 (tumor suppressor genes, TSGs) 不表現，或是致癌基因 (oncogenes) 過度活化所造成，因此利用分佈在基因體的微衛星序列 (microsatellite markers)，針對基因體進行異質性喪失 (loss of heterozygosity, LOH) 分析，可以協助我們尋找未知的腫瘤抑制基因。大部分的肺癌組織都有很高的對偶基因缺失頻率 (fractional allelic loss, FAL)，顯示肺癌組織的基因體經常發生DNA缺失，這可能意味著癌細胞的DNA雙股斷裂 (DNA double-strand break, DSB) 修補機制效率不佳，無法將DNA雙股斷裂做適當的修復。細胞中的DNA雙股斷裂修補機制有兩個，分別是同源染色體重組路徑 (homologous recombination pathway, HR) 和非同源染色體末端接合路徑 (non-homologous end-joining pathway, NHEJ)。過去許多研究發現，基因多型性的差異會在某種程度上影響基因的功能或效率，導致不同的疾病發生敏感性，因此我們推測DNA雙股斷裂修補基因的基因多型性可能會影響肺癌的發生風險。 目的與方法：在LOH分析中，本研究使用112個微衛星序列針對84個非小細胞肺癌 (non-small cell lung cancer, NSCLC) 病人進行基因體缺失分析，並且定義最小缺失區域 (minimal deletion region, MDR)，以尋找和肺癌發生相關的腫瘤抑制基因。另外，我們針對NHEJ路徑中的四個基因：Ku70、Ku80、Ligase IV和XRCC4，進行單一核苷酸多型性 (single nucleotide polymorphism, SNP) 分析，以瞭解基因型對於肺癌發生敏感性的影響，並分析這四個基因之間是否有交互作用 (joint effect)。 結果：肺腺癌 (adenocarcinoma, AD) 和鱗狀上皮肺癌 (squamous carcinoma, SQ) 的LOH頻率分別是48%和42%。我們還發現25個MDRs，平均長度是5.18 cM，在AD和SQ中，分別有12和13個MDRs。其中MDRA7p1 (D7S1818-D7S506, 7p12.1-7p12.3) 的缺失和AD有關，而MDRS9p1 (D9S2169-D9S286, 9p24.1-9p24.3) 的缺失則是和SQ及抽菸有關。在SNP分析部分，Ku70和Ku80的基因型分佈在肺癌病人及無癌症族群間沒有顯著的差異 (P>0.05, multivariate logistic regression model)；但是我們發現當個體在Ligase IV (G3539A, Thr9Ile) 帶有高風險基因型 (A/A and A/G) 時，其得肺癌的危險比會明顯的提高 (adjusted odds ratio, aOR=1.64, 95% confidence interval, 95% CI=1.03-2.62, adjusted P=0.038)。此外，當個體在XRCC4 (G275626A, splice-site) 帶有高風險基因型 (G/G) 時，也會有較高的肺癌發生風險 aOR=2.38, 95% CI=1.05-5.76, adjusted P=0.043)。我們還發現在Ku70和Ku80之間，以及Ligase IV和XRCC4都有交互作用的存在，這樣的交互作用會使該個體的肺癌發生風險提高為一般人的4.65倍 (95% CI=1.19-24.08, adjusted P=0.040) 和8.75倍 (95% CI=2.27-57.77, adjusted P=0.006)。 結論：LOH分析以及MDRs的定義，將有利於尋找和肺癌發生有關的基因。SNP的分析，則讓我們瞭解NHEJ基因的基因多型性以及基因之間的交互作用，確實會影響肺癌發生敏感性。

Background: Most tumor suppressor genes (TSGs) are recessive. Both copies of TSGs need to be inactivated for losing their biological function. One allele may be inactivated by point mutation and methylation change. The other is frequently inactivated by a large deletion involving the gene of interest as well as adjacent stretches of DNA. Thus, searches for genomic regions frequently deleted by loss of heterozygosity (LOH) in cancer have helped to identify or confirm the location of several TSGs. In addition, lung cancer cells frequently exhibit marked chromosome instability such as high fractional allelic loss (FAL, an indication of genomic instability). We then postulated that genetic polymorphism of the double strand-break repair (DSBR) genes, Ku70, Ku80, Ligase IV, and XRCC4, may involve in lung cancer susceptibility. Purposes and Methods: To define the minimal deletion regions (MDRs) that may contain TSGs, we used 112 microsatellite markers to perform LOH analysis in tumors and corresponding normal tissues from 48 adenocarcinomas (ADs) and 36 squamous carcinomas (SQs) lung cancer patients. In addition, we investigated the frequency of four NHEJ (non-homologous end-joining) pathway genes Ku70 (C217G), Ku80 (G92504A), Ligase IV (G3539A, Thr9Ile), and XRCC4 (G275626A, splice-site) polymorphisms in 151 lung cancer patients and 162 cancer-free individuals. The joint effects of these four genes contributed to an increased cancer risk were also analyzed. Results: The average LOH frequency was 48% in AD and 42% in SQ. In addition, 12 MDRs and 13 MDRs were revealed in AD and SQ, respectively. The data suggested that different mechanisms may be involved in the development of AD and SQ. The average interval of 25 MDRs was 5.18 cM. We found that the loss of MDRA7p1 (D7S1818-D7S506, 7p12.1-7p12.3) was associated with AD, and the loss of MDRS9p1 (D9S2169-D9S286, 9p24.1-9p24.3) was associated with SQ and smoking. Furthermore, stage-specific and common deletion markers were identified, suggesting their role in tumor progression. Several candidate genes were annotated by the web-search tools and they can be target genes for further analyses. There was no significant difference in the genotype distribution of Ku70 (C217G), Ku80 (G92504A) polymorphism between lung cancer patients and cancer-free controls (P>0.05, multivariate logistic regression model). However, we found that patients with homozygous or heterozygous variant genotype (A/A or A/G) of Ligase IV (G3539A, Thr9Ile) polymorphism had a tendency to be associated with lung cancer risk (adjusted odds ratio, aOR=1.64, 95% confidence interval, CI=1.03-2.62, adjusted P=0.038). The patients with homozygous variant genotype (G/G) of XRCC4 (G275626A, splice-site) polymorphism also showed a significantly increase lung cancer risk (aOR=2.38, 95% CI=1.05-5.76, adjusted P=0.043). In addition, we found that the G/G frequency of XRCC4 was significant higher in high FAL patients than low FAL patients. (adjusted P=0.016). Furthermore, the joint effect was found between Ku70 and Ku80 as well as Ligase IV and XRCC4. The increased risk of developing lung cancer were 4.65 (95% CI=1.19-24.08, adjusted P=0.040) and 8.75 (95% CI=2.27-57.77, adjusted P=0.006), respectively, when patients harboring two putative high-risk genotypes in these genes. Conclusions: The LOH analysis and definition of MDRs can potentially be used for the clinical association study and the cloning of new TSGs whose aberration contributes to lung tumorigenesis. In addition, the polymorphic results suggest that the polymorphism in four NHEJ genes may be associated with the susceptibility to lung cancer. Our data provide new evidence that the NHEJ putative risk alleles may have a joint effect on the genomic instability of lung cancer.