Girlscientist Consulting
http://about.me/girlscientist
Variety is the spice of life, and in the case of DNA, measurement of single nucleotide polymorphisms (SNPs) between genomes has become the predominant way researchers assess genomic variety between individuals of all species.
Many labs conduct genome-scale analyses, for example to determine SNPs associated with increased risk for common diseases in humans. These are the headline-grabbing “genome-wide association studies” (GWAS). But genotyping small numbers of SNPs in large numbers of individuals is also common.
For example, after a group finds a few hundred or thousand SNPs that appear interesting from a genome-wide study, it needs to test those polymorphisms in many more samples to see whether the associations hold. Similarly, agricultural studies often use genotyping when breeders want to follow a few, specific traits in large numbers of individuals, but whole-genome sequencing is neither feasible nor cost-effective. Infectious-disease studies tracking specific pathogen genomic variants as they spread between hosts also make use of small-SNP but large-sample genotype tests.
TaqMan
One of the earliest technologies developed for SNP detection is the TaqMan® assay, now offered by Life Technologies. Researchers can choose from more than 4.5 million predesigned TaqMan SNP assays or use the Custom TaqMan Assay Design Tool for their own SNPs of interest. Either way, the reaction uses standard forward and reverse PCR primers flanking the polymorphism, plus assay-specific oligonucleotide probes binding the SNP of interest.
Each SNP probe contains one of two fluorophores on one end and a quencher on the other. Most probes are designed to adopt a stem-loop structure in solution, ensuring the probes are dark. When added to a PCR mix containing DNA with the allele in question, the appropriate probe binds to the DNA and is cleaved into nucleotides by the exonuclease activity of Taq DNA polymerase during primer extension. That frees the fluorophore from the quencher, producing a detectable signal.
TaqMan technologies have the advantages of being relatively fast, with less than three-hour turnaround, and greater specificity than other methods. But they’re also relatively expensive because of all of the primers and probes needed. Several groups have developed multiplexing ability for TaqMan probes and primers, but most users still test one SNP per tube or reaction well.
GoldenGate
At the HudsonAlpha Institute for Biotechnology (where the author was the director of research affairs for several years), Faculty Investigator Devin Absher and his lab use TaqMan assays primarily for relatively small numbers of SNPs at a time and Illumina’s GoldenGate® assays for up to 3,072 SNPs per sample. Beyond that, the lab moves into either next-generation sequencing or Illumina microarrays.
Daniel Peiffer, market manager for DNA applications at Illumina, recommends the company’s Infinium® array product, which “has become the most widely adopted large-scale genotyping technology and was designed to be complementary to GoldenGate, such that it targets 3,072 to 1 million custom SNPs.”
“The GoldenGate assays have worked well for us, but they become most cost-effective when you have 384 SNPs or more to test,” Absher says. For example, when Absher’s collaborators wanted to map regions of the stickleback fish’s genome that influenced lateral line morphology, they created a custom GoldenGate assay with 1,536 SNPs. More of the assays on the chip failed than might be expected with the human genome, as there was not a high-quality reference-genome sequence at the time on which to base the SNP tests. But the team was still able to identify linkage groups influencing both morphology and epistasis between genomic regions [1].
GoldenGate is a PCR-based assay based on allele-specific primer extension. The reaction uses two forward primers, each with an allele-specific nucleotide at its 3’ end and one of two universal sequences on its 5’ end. The reverse primer contains a third universal sequence, plus an “IllumiCode address” for detection.
To perform the assay, the genomic DNA is mixed with the three primers, DNA polymerase and ligase. Only the forward primer corresponding to the polymorphism in the template is incorporated into the final product, which is then amplified using one of two dye-labeled primers and detected on an Illumina microarray. (Using this strategy, a heterozygous allele should produce a pool of products labeled half with one dye and half with the other.)
Sequenom
For typing less than 40 SNPs in a large number of samples, Associate Professor Michael Zwick of Emory University’s Integrated Genetics Core recommends the Sequenom MassARRAY® Genotyping System.
At its heart, MassARRAY is a MALDI-TOF mass spectrometer. A Sequenom iPLEX® genotyping assay involves multiplexed PCR around the SNPs of interest, followed by a primer-extension reaction for SNP detection. The reaction products for each sample are then spotted onto the company’s SpectroCHIP® arrays and separated by MALDI-TOF mass spectrometry. Because different nucleotides have different masses, two alleles can be differentiated in a mass spectrometer.
“Samples are processed easily,” says Zwick, and the data quality is well suited for this application. But as a caveat, the web site for Partners Center for Personalized Genetic Medicine tells potential customers of its Sequenom services that, as with many SNP-based detection assays, “There is a possible 5-20% dropout rate of SNPs” because of genomic features around each one.
Given the falling cost of DNA sequencing, many labs that have established the technology are already moving some of their genotyping to sequence-based platforms. Peiffer says many of Illumina’s customers “are quite interested in supplementing their current menu of array-based tools with an amplicon-based method for targeted resequencing.” These methods also allow for pooling of samples and have spurred significant software development, primarily in the quest to make them even more cost-effective. But traditional SNP assays lend themselves particularly to in-the-field applications where sequencing operations are just not practical, and are likely to be with us for some time to come.
Reference
[1] Wark, AR, et al., “Genetic architecture of variation in the lateral line sensory system of threespine sticklebacks,” G3, 2:1047-56, 2012. [PubMed]