Combining degenerate oligonucleotide-primed PCR (DOP-PCR) with comparative genomic hybridization (CGH) has made it possible to analyze genomic changes in single cells. Although DOP-PCR-CGH methodology has been reported, the reproducibility of the method has been uncertain. We have developed a reproducible DOP-PCR-CGH protocol by systematically evaluating different labeling methods (including nick translation, PCR incorporation, and random-primed labeling) and different hybridization mixtures (including amplified test DNA vs. amplified reference DNA, termed homo-hybridization; and amplified test DNA vs. unamplified reference DNA or vice versa, termed hetero-hybridization). We have analyzed DNA samples obtained from 16 tissue sources including fresh/frozen normal and tumor samples, formalin fixed and paraffin embedded tumor tissue, and tumor cell lines by using differently labeled probes and hybridization combinations, and we calculated the corresponding rate (CR) of DOP-PCR-CGH with standard CGH. We found that homo-hybridization produced reproducible results with high CRs as compared to standard CGH (91-100% CR, mean 97%); In contrast, hetero-hybridization failed to generate reproducible hybridization with low CRs (57-97% CR, mean 80%; chi(2) = 1245.8, P<0.0001), high background, uneven hybridization, and false deletions or amplifications. In addition, our improved DOP-PCR protocol raised the amplification efficiency at least five times as compared to previously reported protocols, allowing for the detection of genomic imbalances in as little as 12.5 pg of starting DNA. In conclusion, the DOP-PCR-CHG homo-hybridization method, especially when combined with labeling by nick translation, is reliable and reproducible. The method can be used in screening for genomic imbalances using minute amounts of tumor DNA, thereby facilitating CGH application. Genes Chromosomes Cancer 28:395-403, 2000.
Copyright 2000 Wiley-Liss, Inc.