Random mutagenesis-PCR to introduce alterations into defined DNA sequences for validation of SNP and mutation detection methods

Sensitive and high throughput techniques are required for the detection of DNA sequence variants such as single nucleotide polymorphisms (SNPs) and mutations. One problem, common to all methods of SNP and mutation detection, is that experimental conditions required for detection of DNA sequence variants depend on the specific DNA sequence to be analyzed. Although algorithms and other calculations have been developed to predict the experimental conditions required to detect DNA sequence variation in a specific DNA sequence, these algorithms do not always provide reliable information and experimental conditions for SNP and mutation detection must be devised empirically. Determination of experimental conditions for detection of DNA sequence variation is difficult when samples containing only wild type sequence are available. When patient derived positive controls are used, increasingly there are valid concerns about commercial ownership and patient privacy. This report presents a rapid and efficient method, employing random mutagenesis-PCR (RM-PCR) using low fidelity DNA polymerase, to randomly introduce single and multiple base substitutions and deletions into DNA sequences of interest. Clones with sequence changes were used to validate denaturing HPLC (DHPLC) algorithm predictions, optimize conditions for mutation detection in exon 15 of the tyrosine kinase domain of the MET proto-oncogene, and to confirm the association between specific DNA sequence changes and unique DHPLC chromatographic profiles (signatures). Finally, DNA from 33 papillary renal carcinoma (PRC) patients was screened for mutations in exon 15 of MET using "validated" DHPLC conditions as a proof of principle application of RM-PCR. Use of RM-PCR for DHPLC and other SNP/mutation detection methods is discussed along with challenges associated with detecting sequence alterations in mixed tumor/normal tissue, pooled samples, and from regions of the genome that have been amplified during tumorigenesis or duplicated during evolution. Hum Mutat 17:210-219, 2001. Published 2001 Wiley-Liss, Inc.     

Profiling of copy number variations (CNVs) in healthy individuals from three ethnic groups using a human genome 32 K BAC-clone-based array

To further explore the extent of structural large-scale variation in the human genome, we assessed copy number variations (CNVs) in a series of 71 healthy subjects from three ethnic groups. CNVs were analyzed using comparative genomic hybridization (CGH) to a BAC array covering the human genome, using DNA extracted from peripheral blood, thus avoiding any culture-induced rearrangements. By applying a newly developed computational algorithm based on Hidden Markov modeling, we identified 1,078 autosomal CNVs, including at least two neighboring/overlapping BACs, which represent 315 distinct regions. The average size of the sequence polymorphisms was approximately 350 kb and involved in total approximately 117 Mb or approximately 3.5% of the genome. Gains were about four times more common than deletions, and segmental duplications (SDs) were overrepresented, especially in larger deletion variants. This strengthens the notion that SDs often define hotspots of chromosomal rearrangements. Over 60% of the identified autosomal rearrangements match previously reported CNVs, recognized with various platforms. However, results from chromosome X do not agree well with the previously annotated CNVs. Furthermore, data from single BACs deviating in copy number suggest that our above estimate of total variation is conservative. This report contributes to the establishment of the common baseline for CNV, which is an important resource in human genetics.     

Concordance between semen-derived HIV-1 proviral DNA and viral RNA hypervariable region 3 (V3) envelope sequences in cases where semen populations are distinct from those present in blood

Sequence analysis of the third hypervariable region (V3) of the envelope gene of the HIV-1 was carried out on HIV proviral and viral populations present in blood and semen. Phylogenetically distinct populations of virus were observed in three of the 10 patients analysed. Although the majority of the viruses were predicted to have an R5 phenotype, amino acid differences between blood and semen-derived virus and provirus sequences were observed at sites previously shown to affect cell tropism. Importantly, the semen proviral population was representative of that observed for cell-free virus. This indicates that seminal fluid mononuclear cells are possible sources for the cell-free virus in found in semen.     

Identification and evaluation of 55 genetic variations in the <Emphasis Type="Italic">BRCA1</Emphasis> and the <Emphasis Type="Italic">BRCA2</Emphasis> genes of patients from 50 Japanese breast cancer families

We sequenced approximately 23 kb genomic regions containing all the coding exons and their franking introns of two breast cancer susceptibility genes, BRCA1 and BRCA2, of 55 individuals from 50 unrelated Japanese breast cancer families. We identified 55 single-nucleotide polymorphisms (SNPs) (21 in BRCA1 and 34 in BRCA2) containing nine pathogenic protein-truncating mutations (four in BRCA1 and five in BRCA2 from ten patients). Among the remaining 46 SNPs, allele frequencies of 40 were examined in both the breast cancer patients and 28 healthy volunteers with no breast cancer family history by PCR-RFLP or by direct DNA sequencing. Twenty-eight SNPs were common and were also found in the healthy volunteers and/or a SNP database. The remaining 18 were rare (allele frequency <0.05) and were not found in the healthy volunteers and/or the database. The pathogenic significance of these coding SNPs (cSNPs) remains to be clarified. The SNP information from this study will be useful in the future genetic testing of both BRCA1 and BRCA2 genes in the Japanese population.The first two authors contributed equally to this study.