SNPs Are All the Rage—Part I
Contents
Introduction
What's in an SNP?
SNP's the Old Fashioned Way—One Base at a Time
SNPs on Chips
Next Up
Introduction (Back to Top)
Three press releases in just two days last week announcing various agreements between companies working with single nucleotide polymorphisms (SNPs) attest to just how hot this branch of genomics is right now.
- DeCODE genetics (Reykjavik, Iceland) and Affymetrix (Santa Clara, CA) announced their plans to collaborate on optimizing chip-based SNP genotyping applications.
- PE Biosystems (Foster City, CA) obtained an exclusive license to use the ligase detection reaction for mutation detection, which it plans to automate to speed up SNP detection.
- Lynx Therapeutics (Haywood, CA) has reached an agreement with Oxagen Ltd. (Oxford, England) to collaborate on a program to discover and validate disease-associated SNPs.
And that's just what is being made public. Doubtless, companies with technologies are striking deals with researchers who have access to patient populations to move this emerging technology from the lab bench to the clinic as soon as possible.
What's in an SNP? (Back to Top)
How is it that something as seemingly trivial as a single base change could hold out so much promise? And how close are we to realizing any of this potential?
SNPs, which occur naturally and frequently in DNA, reflect the highly polymorphic nature of genomes—that is, the genome's ability to accept single nucleotide changes without disrupting the functioning of the genes or genome. This is not to say all SNPs are benign. In fact, some common genetic diseases—sickle cell anemia being the premier example—can be traced back to a single base change. However, since much of the genome is non-coding, and several million years of evolution have gone into weeding out potentially deleterious mutations, most SNPs extant today have no discernible phenotype.
One reason why SNPs have captured so much attention is their expected utility in pinpointing regions of the genome involved with disease. The fact that they are scattered throughout the genome—estimated to occur once every several hundred base pairs in humans—makes them convenient markers for genetic mapping studies. Also of interest, especially to big pharma, is the possibility that SNPs could be used to screen patients for adverse drug reactions prior to prescribing them.
However, to make a SNP useful, a couple of things need to happen. It first needs to be confirmed—the going rate appears to be that a given SNP must appear in 10% of the population, otherwise it's considered too infrequent to be informative. Second, association studies that typically require thousands of individuals need to be done to link up SNPs with particular pathologies. Do the math—thousands of SNPs screened against thousands of individuals adds up to a prodigious effort. And therein lies the reason for the furious technology development for SNP discovery and screening.
Recognizing the potential of SNPs in disease diagnosis and treatment, the NIH launched a three-year, $36 million program to catalog some 50,000 to 100,000 human SNPs. But to some in the industry, this effort won't be enough, and a host of pharmaceutical companies have thrown their considerable weight into this. In fact, just last month, ten giants in the industry formed a consortium to provide support for five research centers. The goal: to identify some 300,000 SNPs in two years (click here for related story).
So where do we stand? Of the possible six million SNP's in a human genome (assuming that they occur once every 500 base pairs in the three billion base pair genome) only a few thousand (4,588 to be exact, as of May 20, 1999) have been placed in the public domain SNP database maintained by the Nation Center for Biotechnology Information—a mere drop in the proverbial bucket. But hold onto your hats! With all the effort in this area and the new technologies being applied to this, this number is likely to jump in short order.
SNP's the Old Fashioned Way—One Base at a Time (Back to Top)
Early work on SNP detection relied strictly on sequence information—in silico screening, as it were, of gene sequences derived from different individuals. But a paper published last year paved the way of the future. A group from the Whitehead Institute for Biomedical Research (Cambridge, MA) conducted parallel searches for SNPs using conventional sequencing, with which some 200 SNPs were found (Ref. 1). Using a human genome chip array, they found roughly ten times the number of SNPs, some 2,748 in all. But 2,000, or even 20,000, is still a far cry from the millions of SNPs believed to be buried in the genome, and to get even this much (or this little) data required some 150 different chip designs.
However, sequencing approaches to SNP detection are not dead. Celera Genomics (Rockville, MD), the company that wedded PE Biosystem's high throughput sequencing technology with Craig Ventor's sequencing strategies, is engaged in a massive effort to sequence completely three to five individuals. This project, which a company spokesperson estimates will be completed by the end of the 2001, is expected to yield 10 to 30 million polymorphisms (six million per individual genome times 5 individuals, in case this number throws you). A database containing all known candidate and validated polymorphisms will be accessible to Celera's customers and collaborators.
Other companies in the process of amassing proprietary SNP databases are CuraGen (New Haven, CT), which is compiling cSNPs (coding SNPs) as well as SNPs; GenSet (Paris, France); and Incyte (Palo Alto, CA).
SNPs on Chips (Back to Top)
In addition to the brute force approach, new technologies for rapid detection and screening for SNPs are coming on line.
Affymetrix will be introducing a product called the HuSNP Mapping Assay. The product will interrogate approximately 1500 SNPs derived from sites randomly located across the 22 autosomes and the X chromosome. The entire scan will be performed on a single GeneChip probe array. A multiplex PCR strategy further simplifies the assay—sample preparation for an entire genome scan can be done in 24 wells of a micro plate. The SNPs represented in the HuSNP Chip were discovered in a collaboration between Affymetrix and the Whitehead Institute using Affymetrix GeneChip technology. All of the SNPs on the product will be on the SNP map viewable at Whitehead's web site.
Orchid Biocomputer (Princeton, NJ) has shown its commitment to SNP analysis in several ways. The recent acquisition of Molecular Tools will enable Orchid to combine its expertise in microfluidics with the genetic technologies developed by Molecular Tools, in particular its Genetic Bit Analysis (GBA). This technique uses a combination of hybridization and primer extension in a multi-well format to screen for SNPs without the use of gels or sequencing. Single-stranded fragments, protected by a proprietary technology, are captured through hybridization to oligos attached to the surface of wells. Primer extension is then used to attach one nucleotide to the ends of the fragments. Detector molecules for each of the four nucleotides are applied one at a time, and the nucleotide sequence at each position can be read off. If the fragments under study are polymorphic, more than one detector will register for a given position.
A high throughput robotic system using GBA is in the works at Orchid, which allows one person to perform some 20,000 genotypes per day. This machine is designed to process GBA formatted plates that the company will prepare and bar-code for its customers. It's expected to be useful for confirming suspected SNPs, but might find its greatest utility in association studies, which require large-scale screening.
Also in the works at Orchid are SNPs on chips, array-based multiplexed GDA, miniaturized to reduce the sample volumes required, which will provide significant cost savings to large users. An even smaller chip, with volumes in the sub-microliter range, is being developed for its in-house genotyping service.
Finally, just last month, Orchid launched a web site (SNPs.com) "dedicated to the creation, collection, and dissemination of information pertaining to SNPs," according to a company press release. This site will provide a comprehensive overview of this research area, portals to other genomic Internet sites, and a dynamic forum for exchange between members of the its Clinical Genetics Network.
Next Up (Back to Top)
SNP features to follow will look at other SNP detection and screening technologies, including Lynx Therapeutics' SNP-extracting and genotyping technologies, and PE Biosystem's Ligase Detection Method, as well as some instruments, old and new, that can be used for SNP detection in the run-of-the-mill lab environment.
References
- D.G. Wang et al., "Large-Scale Identification, Mapping, and Genotyping of Single-Nucleotide Polymorphisms in the Human Genome," Science 280: 1088-82 (1998).
For more information on Genetic Bit Analysis: Russel Granzow, Senior Director of Business Development and Marketing, Orchid Biocomputer, 303 College Road East, Princeton, NJ 08540. Tel: 609-750-2200.
By Laura DeFrancesco