Peng Zhen, Zhou Weichen, Fu Wenqing, Du Renqian, Jin Li, Zhang Feng
State Key Laboratory of Genetic Engineering and Ministry of Education Key Laboratory of Contemporary Anthropology and Department of Biology, Washington University in St. Louis, St. Louis, MO 63130, USA.
State Key Laboratory of Genetic Engineering and Ministry of Education Key Laboratory of Contemporary Anthropology and.
Hum Mol Genet. 2015 Mar 1;24(5):1225-33. doi: 10.1093/hmg/ddu533. Epub 2014 Oct 16.
Non-allelic homologous recombination (NAHR) is one of the key mechanisms of DNA rearrangement. NAHR occurring between direct homologous repeats can generate genomic copy number variation (CNV) and make significant contributions to both genome evolution and human diseases such as cancer. Intriguingly, previous observations on the rare CNVs at certain genomic disorder loci suggested that NAHR frequency could be dependent on homology properties. However, such a correlation remains unclear at the other NAHR-mediated CNV loci, especially the common CNVs in human populations. Different from the rare CNVs associated with genomic disorders, it is challenging to identify de novo NAHR events at common CNV loci. Therefore, our previously proposed statistic M was employed in estimating relative mutation rate for the NAHR-mediated CNVs in human populations. By utilizing generalized regression neural network and principal component analysis in studying 4330 CNVs ascertained in 3 HapMap populations, we identified the CNVs mediated by NAHR between paired segmental duplications (SDs) and further revealed the correlations between SD properties and NAHR probability. SD length and inter-SD distance were shown to make major contributions to the occurrence of NAHR, whereas chromosomal position and sequence similarity of paired SDs are also involved in NAHR. An integrated effect of SD properties on NAHR frequency was revealed for the common CNVs in human populations. These observations can be well explained by ectopic synapsis in NAHR together with our proposed model of chromosomal compression/extension/looping (CCEL) for homology mis-pairing. Our findings showed the important roles of SDs in NAHR and human genomic evolution.
非等位基因同源重组（NAHR）是DNA重排的关键机制之一。发生在直接同源重复序列之间的NAHR可产生基因组拷贝数变异（CNV），并对基因组进化和癌症等人类疾病产生重大影响。有趣的是，先前对某些基因组疾病位点罕见CNV的观察表明，NAHR频率可能取决于同源性特性。然而，在其他NAHR介导的CNV位点，尤其是人群中的常见CNV，这种相关性仍不清楚。与与基因组疾病相关的罕见CNV不同，在常见CNV位点识别新生NAHR事件具有挑战性。因此，我们先前提出的统计量M被用于估计人群中NAHR介导的CNV的相对突变率。通过在研究3个HapMap人群中确定的4330个CNV时利用广义回归神经网络和主成分分析，我们鉴定了由配对节段重复序列（SD）之间的NAHR介导的CNV，并进一步揭示了SD特性与NAHR概率之间的相关性。结果表明，SD长度和SD间距离对NAHR的发生起主要作用，而配对SD的染色体位置和序列相似性也与NAHR有关。我们揭示了SD特性对人群中常见CNV的NAHR频率的综合影响。这些观察结果可以通过NAHR中的异位联会以及我们提出的用于同源性错配的染色体压缩/延伸/环化（CCEL）模型得到很好的解释。我们的研究结果表明了SD在NAHR和人类基因组进化中的重要作用。
