Maris John M, Hii George, Gelfand Craig A, Varde Shobha, White Peter S, Rappaport Eric, Surrey Saul, Fortina Paolo
Division of Oncology, The Children's Hospital of Philadelphia, and Department of Pediatrics, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104, USA.
Genome Res. 2005 Aug;15(8):1168-76. doi: 10.1101/gr.3865305.
Many cancers are characterized by chromosomal aberrations that may be predictive of disease outcome. Human neuroblastomas are characterized by somatically acquired copy number changes, including loss of heterozygosity (LOH) at multiple chromosomal loci, and these aberrations are strongly associated with clinical phenotype including patient outcome. We developed a method to assess region-specific LOH by genotyping multiple SNPs simultaneously in DNA from tumor tissues. We identified informative SNPs at an average 293-kb density across nine regions of recurrent LOH in human neuroblastomas. We also identified SNPs in two copy number neutral regions, as well as two regions of copy number gain. SNPs were PCR-amplified in 12-plex reactions and used in solution-phase single-nucleotide extension incorporating tagged dideoxynucleotides. Each extension primer had 5' complementarity to one of 2000 oligonucleotides on a commercially available tag-array platform allowing for solid-phase sorting and identification of individual SNPs. This approach allowed for simultaneous detection of multiple regions of LOH in six human neuroblastoma-derived cell lines, and, more importantly, 14 human neuroblastoma primary tumors. Concordance with conventional genotyping was nearly absolute. Detection of LOH in this assay may not require comparison to matched normal DNAs because of the redundancy of informative SNPs in each region. The customized tag-array system for LOH detection described here is rapid, results in parallel assessment of multiple genomic alterations, and may speed identification of and/or assaying prognostically relevant DNA copy number alterations in many human cancers.
许多癌症的特征是染色体畸变,这些畸变可能预示疾病的预后。人类神经母细胞瘤的特征是体细胞获得性拷贝数变化,包括多个染色体位点的杂合性缺失(LOH),并且这些畸变与包括患者预后在内的临床表型密切相关。我们开发了一种方法,通过对肿瘤组织DNA中的多个单核苷酸多态性(SNP)进行基因分型来评估区域特异性LOH。我们在人类神经母细胞瘤中反复出现LOH的九个区域,以平均293 kb的密度鉴定了信息丰富的SNP。我们还在两个拷贝数中性区域以及两个拷贝数增加区域鉴定了SNP。SNP在12重反应中进行聚合酶链反应(PCR)扩增,并用于掺入标记双脱氧核苷酸的溶液相单核苷酸延伸反应。每个延伸引物在5'端与市售标签阵列平台上2000个寡核苷酸之一互补,从而实现固相分选和单个SNP的鉴定。这种方法能够同时检测六种人类神经母细胞瘤衍生细胞系以及更重要的14例人类神经母细胞瘤原发肿瘤中的多个LOH区域。与传统基因分型的一致性几乎是绝对的。由于每个区域中信息丰富的SNP具有冗余性,因此该检测方法中LOH的检测可能不需要与匹配的正常DNA进行比较。本文所述的用于LOH检测的定制标签阵列系统速度快,可并行评估多个基因组改变,并且可能会加速许多人类癌症中预后相关DNA拷贝数改变的鉴定和/或检测。