Search

Despite many recent advances in high-throughput single nucleotide polymorphism (SNP) genotyping technologies, there is still a great need for inexpensive and flexible methods with a reasonable throughput. Here we report substantial modifications and improvements to an existing homogenous allele-specific PCR-based SNP genotyping method, making it an attractive new option for researchers engaging in candidate gene studies or following up on genome-wide scans. In this advanced version of the melting temperature (Tm)-shift SNP genotyping method, we attach two GC-rich tails of different lengths to allele-specific PCR primers, such that SNP alleles in genomic DNA samples can be discriminated by the Tms of the PCR products. We have validated 306 SNP assays using this method and achieved a success rate in assay development of greater than 83% under uniform PCR conditions. We have developed a standalone software application to automatically assign genotypes directly from melting curve data. To demonstrate the accuracy of this method, we typed 592 individuals for 6 SNPs and showed a high call rate (>98%) and high accuracy (>99.9%). With this method, 6-10,000 samples can be genotyped per day using a single 384-well real-time thermal cycler with 2-4 standard 384-well PCR instruments.

The ideal technology for screening single-nucleotide polymorphisms requires high throughput with minimal cost per sample, minimal usage of valuable DNA resources, and maximal flexibility for assessment of new polymorphisms. We demonstrate here the feasibility of kinetic allele-specific PCR with DNA pooling (S. Germer et al., Genome Res., 10: 258-266, 2000) in a population study that satisfies all of the mentioned criteria and offers a powerful new tool for detecting meaningful polymorphic differences in candidate gene association studies and genome-wide linkage dysequilibrium scans. Three individuals prepared pooled DNA samples from 269 individuals separated into three racial/ethnic groups: Caucasians (n = 56), African-Americans (n = 86), and Hispanics (n = 127). We used kinetic allele-specific PCR to determine the allele frequencies of the common paraoxonase 1 polymorphism, PON1 Q191R, in these pools. Paraoxonase 1 is a critical enzyme for inactivating neurotoxic intermediates in the metabolism of organophosphates. In a blinded test of the technology, these nine pooled DNA samples were sent to Roche for genotyping by kinetic allele-specific PCR. The allele frequencies found were 0.266 +/- 0.011, 0.386 +/- 0.011, and 0.617 +/- 0.010, respectively, which were comparable to the frequencies of 0.269, 0.403, and 0.622 determined by PCR-restriction fragment length polymorphism analysis. These same samples were genotyped on two kinetic PCR platforms from different manufacturers, using three different DNA polymerases. The results were comparable between both platforms and among all three polymerases. The results demonstrate a powerful new technology for determining frequencies of single-nucleotide polymorphisms in an epidemiological study.

We have developed an accurate, yet inexpensive and high-throughput, method for determining the allele frequency of biallelic polymorphisms in pools of DNA samples. The assay combines kinetic (real-time quantitative) PCR with allele-specific amplification and requires no post-PCR processing. The relative amounts of each allele in a sample are quantified. This is performed by dividing equal aliquots of the pooled DNA between two separate PCR reactions, each of which contains a primer pair specific to one or the other allelic SNP variant. For pools with equal amounts of the two alleles, the two amplifications should reach a detectable level of fluorescence at the same cycle number. For pools that contain unequal ratios of the two alleles, the difference in cycle number between the two amplification reactions can be used to calculate the relative allele amounts. We demonstrate the accuracy and reliability of the assay on samples with known predetermined SNP allele frequencies from 5% to 95%, including pools of both human and mouse DNAs using eight different SNPs altogether. The accuracy of measuring known allele frequencies is very high, with the strength of correlation between measured and known frequencies having an r(2) = 0.997. The loss of sensitivity as a result of measurement error is typically minimal, compared with that due to sampling error alone, for population samples up to 1000. We believe that by providing a means for SNP genotyping up to thousands of samples simultaneously, inexpensively, and reproducibly, this method is a powerful strategy for detecting meaningful polymorphic differences in candidate gene association studies and genome-wide linkage disequilibrium scans.