Few would argue that today’s leading manufacturers of DNA microarrays, or “chips,” have much in common with mail order clothing companies like Land’s End or Eddie Bauer. But if chip manufacturer ProtoGene has its way, that will change next year when the Menlo Park company begins to offer a service it calls “chips at your doorstep.” Almost as easily as one can log onto a retailer’s website, enter neck size, hem and sleeve lengths, and have clothing delivered by Fedex, researchers will be able to log onto ProtoGene’s website any time of day and order microarray chips.
Bob Barrett, ProtoGene’s vice president of marketing, explains: “You’ll be able to go into your secure space and design a chip with whatever sequences you’re looking for — your internal proprietary sequences, or sequences from GenBank, or Celera, or another source. These will go through a bioinformatics algorithm, and the design will be sent back to you for approval.” After pushing the “order microarray” button, your customized microarrays with the specified oligonucleotides will be manufactured at ProtoGene’s automated facility and shipped to you overnight.
AN UNTAPPED MARKET
ProtoGene hopes to succeed in the microarray marketplace not by going head-to-head with established giants like Affymetrix that produce high-density microarrays with tens of thousands of “spots” or features, but by producing customizable, lower density chips (2,500 features and under) on a fast turnaround. These would help ProtoGene tap into a market that its management believes is currently underserved — academic and industry laboratories that cannot afford to incorporate conventional DNA microarray technology into their research programs.
Barrett explains that labs trying to adopt microarray technology have several options — none of them ideal. One is to order high-density, pre-manufactured chips from a company like Affymetrix that also provides the peripheral tools — scanners and software — necessary to read and make sense of the data acquired by the chip. Researchers can also purchase “spotting” equipment that allows them to create customized, lower density microarrays. Alternatively, labs can contract out to “garage shops” that use spotting to make low-density microarrays.
But both high-density and lower density spotted microarrays involve initial costs and “hassle factors” that are barriers for many researchers. Barrett says that these costs mean that decisions about microarray purchases are often made at higher administrative levels. The result: Well-funded, “A-tiered” laboratories can adopt the technology while more modestly funded laboratories often go without.
ProtoGene’s plan is to help researchers overcome these barriers by introducing more affordable and customizable microarrays to compete with the lower density chips currently made by spotting technology. If costs are reduced, says Barrett, researchers themselves can make decisions about microarray technology. “We want to democratize the process. We want to drive the decision down to the people doing the research, who know what they need on a chip, and make it as easy as possible for them to get it,” Barrett says.
Richard Myers, professor of genetics at Stanford University and co-director of the Stanford Human Genome Center, is also vice president for scientific research at ProtoGene. He says, “Right now, many people want to do this and they’re frustrated because it is hard to get into, because there is no easy customization and costs are high.”
Myers explains that researchers often use high-density microarray chips to measure the expression of a large number of genes in a tissue sample, but once they have identified a subset of relevant genes, they switch to lower density, customized chips that contain only the genes they’re interested in. Currently, these custom microarrays are created by spotting technology. The costs and logistics of ordering and storing the long chain DNA spotted onto the arrays are drawbacks.
One way ProtoGene says its customizable, “chips on your doorstep” approach will compete with the lower density, spotted chips will be by eliminating the need for researchers to deal with these logistics. The company will prepare glass slides by a method that it patented in 1991 for in situ synthesis of oligonucleotides on a differential surface tension array.
Myers emphasizes that the “global pass” using high-density microarrays will continue to be a very valuable research tool. “I’m a real advocate of it,” he says. “It’s just that it’s not the only experiment most people want to do. They want a more modest number [of features] designed around their hypothesis.”
Industry insiders say ProtoGene’s proposal sounds promising. Michael McKenna, vice president of operations for CuraGen, a genomics company that uses proprietary, non-microarray technology to measure gene expression, says, “Many companies as well as labs used to have oligo synthesizers, but no one bothers any more. Now, you e-mail in your order and you get your oligos in two days. I think that’s eventually going to happen with microarray chips, as well. It’s not a commodity yet, but it’s definitely on the road to becoming one.”
Larry Hunter, director of computational pharmacology at the University of Colorado School of Medicine, agrees that rapid-turn around, customizable chips would be an important service, but he says there are some caveats. “If an array technology isn’t extremely consistent, then it’s much less useful,” Hunter says.
Noting that consistency and reproducibility are problems with conventional custom spotted arrays, he notes that oligonucleotide microarrays face problems with cross-hybridization. “Picking the right sequence is tricky. It’s a problem with oligos in general.”
Hunter adds that cost is another factor that makes it difficult for academic labs to do the kind of experiments they’d like. “A typical oligo array costs hundreds of dollars and we need dozens of them to perform a well-designed study, and this results in sticker shock.” The price, he says, especially of small arrays, “has to be an order of magnitude less to dramatically change” how arrays are used in academic labs.
John D’Errico, global product manager for Genomic Solutions of Ann Arbor, predicts that the introduction of low-priced, customizable microarrays will “open up the market to researchers who don’t have the resources” to use current technology. But, he adds that there is much more to performing genomic analysis than just obtaining a good microarray. Because of the time and cost involved for a laboratory to set up an in-house system, D’Errico anticipates continued business for his company, which sells integrated genomics and proteomics systems and performs sample analysis on a contract basis.
He warns, “The biggest problem is information flow. When people buy separate components that don’t talk to one another, they can do the research but it’s very difficult to get data out quickly.”
Adds Hunter, “We’re getting to the point where it’s no longer good enough to get the data off the chip in some reliable but proprietary way, it’s also important that things interoperate. It’s really incumbent on the chip provider to make sure that their informatics works and plays well with others.”
EXPANDING BUT “BRUTAL” MARKETPLACE
ProtoGene believes that the “chips on your doorstep” model will make the company an “oligo-opoly” of the microarray industry, an area that by most accounts is undergoing explosive growth. A report released this August by Frontline Strategic Management Consulting of Foster City, Calif., predicts the microarray chip industry will expand at supernova rates, growing to $3.5 billion by 2003.
But while the market is expanding, it is also unpredictable. Noting the announcement of new hollow-fiber gene expression technology by Mitsubishi Rayon in late September (a technology that circumvents Affymetrix’s patents for DNA hybridization on flat glass surfaces), CuraGen’s McKenna sounds a cautionary note. “I think technology is continuing to move, competition is encroaching on all sides, and it’s going to be a brutal marketplace over the next five years.”
One player brushes off the threat of new products in the low-density chip market. Anne Bowdidge, head of investor relations for Affymetrix, says her company still regards high-density chips as its leading products. Acknowledging that high-density and low-density chips are complementary, she adds, “We believe the market is going toward high density.” Bowdidge says Affymetrix continues to increase the number of features on its chips while keeping prices stable. Its most popular product, she says, is GeneChip R, Human Genome U-95, which has over 60,000 human genes and ESTs on five chips.
Bowdidge notes that Affymetrix also supplies spotters and other equipment for customers who want to make their own low-density chips. She says a new “pin and ring” spotting technology that Affymetrix acquired when it bought GMS last year provides more precise spotting and better performing microarrays than conventional spotters. Bart Selby, product manager for instrumentation for Affymetrix, echoes Larry Hunter when he adds that some researchers prefer to use the high-density arrays for most phases of their research rather than lower density, customized chips.
SHORT TIME LINE
ProtoGene’s novel approach might seem overambitious, but CEO Chris Wolf argues that rapid growth is possible because ProtoGene is a “picks and shovels business” for the genomics industry. “What I like is that this is a technology that we can put rapidly into commercialization,” says Wolf. “That is a different time horizon than a biopharmaceutical model, which goes out 7-12 years. The gene chip field is exploding, and with the particular approach of ProtoGene, the time line is fairly short, and we can see results fairly fast. We’re through our proof of principle phase and heading into commercialization.”
Wolf was hired as ProtoGene’s sixteenth employee less than two years ago from Hyseq, a genomics company focused on drug discovery. As Hyseq’s vice president of business development Wolf steered that company through the largest biotech IPO of 1997. Today, ProtoGene has almost 100 employees at its 50,000-square-foot facility in Menlo Park, which houses both corporate offices and manufacturing facilities.
When he came on board, Wolf says, ProtoGene was still in its research phase and the company’s business model was largely undefined. “We stepped back and said, Where’s this business evolving independent of whether or not there’s a ProtoGene?”
Rejecting the vertical integration model, where users have to buy chips and scanners and software from the same suppliers, Wolf and his team decided that a horizontal model made more sense. “We believe the model is going to be evolving as other businesses have, into more open systems,” Wolf says.
Citing the classic example of how the personal computer industry converted from a vertically to a horizontally integrated market, Wolf foresees a microarray market that will be supportive of ProtoGene’s pure chip strategy. “Because it’s so large and growing so fast, it’s going to be very attractive for companies to provide peripherals — the scanners, the software, and so forth,” he says.
FAST-TRACK MANAGEMENT TEAM
Another of Wolf’s tasks was assembling a management team that could bring ProtoGene’s product to the market rapidly, drawing on management talent from the medical device and diagnostic industries. These industries have experience “doing rapid scale up with a lot of discipline,” he notes.
Wolf says his vice president of marketing was a key hire. Bob Barrett was recruited from Chiron, where he oversaw the rollout of Chiron’s branch DNA viral load test for HIV and hepatitis. Barrett comments, “In medical devices and diagnostics, unlike in some of the biotech fields, you have to be very focused because you have the regulatory and other issues that have to be dealt with.”
Also recruited from Chiron was Peter Chris, who was vice president of business development at his previous employer, and has the same title at ProtoGene. Also, Bill Picht, vice president of operations, previously managed 700 people as VP of operations at Behring Diagnostics; CFO Steve Van Dick was CFO at both Perclose and Cardiothoracic Systems; and Rich Meyer, vice president of research and development, was previously research VP at Genelabs Technologies and Epoch Pharmaceuticals.
In addition to bringing scientific expertise to the company, Myers of Stanford is also representative of
ProtoGene’s user base. He says, “I was one of those frustrated people who has worked around lots of microarrays.”
When he first heard ProtoGene’s business plan, he says his reaction was, “This is the way this ought to be done.” A dedicated academician, he admits he has never participated in a private company “on this scale.” He is currently working part-time for ProtoGene while on sabbatical.
Myers predicts that ProtoGene’s chips would initially be used mostly for gene expression analysis, but that the company’s technique of synthesizing oligonucleotides directly on glass could help extend the usefulness of microarrays. “Oligonucleotides are small and you can probe more finely in a gene; it allows you to do something you can’t do with larger pieces of DNA — genetic variation, or genotyping.”
Wolf adds that, for reasons of quality control and added capabilities, “the industry as a whole is going in the direction of shorter length oligonucleotides rather than the longer chain DNA used in conventional spotting.”
At first glance, ProtoGene’s current production space is reminiscent of a machine shop — the room filled by two rows of automated chip synthesizers roughly the size and shape of table saws.
In contrast to the surgical orderliness of the current production room, across the hall is a cavernous, cement-floored factory space with at least ten times the square footage. Bookcases lying on their side and an odd-looking piece of equipment resembling a high-tech weaver’s loom are pushed into the corners, waiting to be cleared to make room for the chip-synthesizers that sometime next year will start working 24-hours-a-day to fill custom chip orders placed over the Web.
In a darkened office, a monitor displays a close-up image of the fluorescing features on a ProtoGene chip. Red and green points are arranged on a black background like dots on a domino. Barrett proudly notes that ProtoGene’s patented surface tension technology produces crisp, round features. “The chip surface is prepared so there are hydrophilic features or spots surrounded by hydrophobic regions. These unique surface tension properties enable us to use ink jet-like printing technology to rapidly and accurately populate the chip.”
It is this combination of chip surface preparation and in situ synthesis of oligonucleotides that Barrett says will distinguish ProtoGene’s from other DNA chip methodologies. Like the marketing maestro he is, Barrett adds, “This technology will greatly impact the genomics industry by providing researchers with rapid access to affordable, customized chip methodologies.”
Envisioning the road ahead, CEO Wolf says, “This has been the year to build, and next year is the year of the battle.”