Affymetrix last week ended months of speculation about whether and how it would compete in the gene-sequencing market by disclosing plans for a new technology that combines microarrays to capture regions of the genome with on-chip enzymatic sequencing.
The new technology, based on Affy’s arrays and sequencing chemistry from USB, a company it recently acquired, will likely compete with existing next-generation sequencing platforms for DNA resequencing applications. But Affy said its technology will be low in cost and high in throughput.
It will be able to analyze tens to hundreds of megabases of DNA on a single chip in a low-cost, high-throughput manner. “We are not talking tens of thousands of dollars per experiment here; we are talking about putting this level of information on single arrays in high-volume, high-throughput formats,” Affymetrix CEO Stephen Fodor told investors at the JPMorgan Healthcare Conference in San Francisco last week.
The array part of the technology is based on a new, higher-yield synthesis method Affymetrix has developed. The method allows the firm to manufacture microarrays with probes up to 75 bases in length that can be synthesized in either 3’-5’ or 5’-3’ direction, according to Fodor.
Affy is using this new method to “fabricate high-density whole-genome targeted arrays … that will capture portions of the genome,” Fodor said during his presentation, which was webcast.
The second piece of the technology consists of “enzymatic sequencing reactions” directly on the chip. Fodor did not reveal details about the sequencing chemistry, the read length, or the labels used but said that the chemistry “allows us to use both a polymerase and ligase labeling reactions directly on the surface” of the array. Affymetrix did not provide further information about the technology before deadline.
“Targeted genotyping all the way to a whole genome will be available with this technology,” he said. “[It] gives us genome-wide assays with high sensitivity and high specificity.”
Fodor did not mention when the company would launch the new sequencing-on-chip technology as a product.
The first product to be launched, “probably in the first half of this year,” will include the longer probes but not the on-chip sequencing reactions. It will be a set of arrays to measure copy number variation with “ultra-high resolution.” Each array in the three-chip set will have 10 million 50-mers “across the non-repetitive, working parts of the genome,” Fodor said.
Later this year, the company “will be talking about some next-generation systems we will be commercializing,” he added later.
Affymetrix said it believes the key advantages of its new technology are its cost and throughput. Fodor did not provide specific pricing information but said that “we will be able to apply all of our advances in high-volume manufacturing to get these at low cost, affordable cost, to the marketplace, and highly scalable to thousands of samples.”
“Together with USB, we are in accelerated development and commercialization of next-generation automated enzymology and sequencing reagents,” Affy President Kevin King said during the conference last week.
So far, Affymetrix has been using the sequencing-on-array technology internally to generate an in-house database of human sequence variation. The company is currently screening 1,100 human samples for “the entire known human variation database” — approximately 12.5 million known polymorphisms — a project the researchers hope to complete during the first half of the year. By comparison, Affy’s current products are based on data from 270 HapMap individuals, genotyped for approximately 4 million SNPs.
According to Fodor, the study “will generate an unprecedented database of human variation” that Affymetrix will use “to design our next-generation products.” It will also make this “extended variation map” available to customers of the new products. These products will deliver “up to 10 million assays per chip in the near future,” he said.
Though Fodor did not mention other next-generation sequencing technologies by name, he acknowledged that some of them also read out DNA sequence on a solid surface. Ilumina’s Genome Analyzer, for instance, grows DNA clusters in the channels of a flow cell, and ABI’s SOLiD spreads DNA-covered beads on a glass slide.
“When we look at many of the technologies that are being developed right now out in the marketplace, the real innovations occur in the enzymatic readout of sequence on solid phase,” Fodor said. “So this is a perfect opportunity for us to blend this, to incorporate these techniques directly into these products.”
“Targeted genotyping all the way to a whole genome will be available with this technology.”
Though details about the new technology are still unknown, in principle, sequencing on ordered arrays could have advantages over current next-generation sequencers, especially for resequencing applications.
“It is an established platform that users are familiar with and it is capable of generating millions of resequencing reactions in parallel,” Kalim Mir told In Sequence by e-mail last week.
Mir, a group leader at the Wellcome Trust Center for Human Genetics in Oxford, UK, has also been working on an enzymatic sequencing chemistry for use on microarrays (see In Sequence 3/27/2007).
“Microarrays will allow next-generation resequencing to be systematic, compared to the shotgun-style sequencing offered by existing [platforms],” he said.
However, “to me, the most compelling advantage of using micorarrays to do resequencing is that it will allow you to directly select whatever regions you want to sequence,” he added. “Solexa, Agencourt, Helicos, 454 technology can’t do this directly; you need plug-ins.”
Recently, several research studies have shown how such plug-ins, for example NimbleGen microarrays or PCR-like reactions, can be used to select portions of the genome for sequencing (see In Sequence 11/6/2007 and 10/16/2007). But these approaches add additional steps to the sample-prep process, Mir noted.
For sequencing entire small genomes, which next-gen sequencers can do without having to select portions of the genome first, “it remains to be seen whether microarray sequencing can offer an advantage,” he said.
A potential disadvantage of Affy’s approach, he noted, is that sequence information is required to generate the microarrays in the first place. As a result, “micorarray sequencing will be better for resequencing than it will be for de novo sequencing,” he said.
For Affy, USB’s sequencing chemistry is “one of the reasons why we made the acquisition,” Fodor said.
Affy announced its $75 million cash acquisition of privately held USB last month. The company makes molecular biology enzymes, biochemical reagents, and products used in membrane protein research. Affy plans to close the deal by the end of March.
USB has a history in DNA sequencing reagents. In the 1980s the company, then known as United States Biochemical, collaborated with Harvard researchers, leading to the commercialization of Sequenase DNA polymerase as well as DNA sequencing kits. According to Affymetrix, “this technology pioneered development of thermostable enzymes and automation for high-throughput sequencing.”
Amersham Life Science bought United States Biochemical in 1993. After Amersham merged with Pharmacia Biotech in 1997, USB managers acquired three product lines, including manual DNA sequencing reagents, and founded the current USB in 1998.