Skip to main content
Premium Trial:

Request an Annual Quote

The Missing Lynx


Lynx Therapeutics is misnamed, low on funds, and short on marketing talent. But if it can overcome all that, it has one hell of a product to sell you


By Aaron J. Sender


“One of the most fulfilling things in this job,” says Lynx Therapeutics CEO Norrie Russell, whose role is to enlighten outsiders about his company’s product, “is when they stop being glassy-eyed and say, ‘Holy shit! I’ve got it.’”

No doubt, Russell’s is no simple task. Many in the industry consider Lynx’s technology the most complicated biology tool ever created.

If you can get past the complexity, though, Lynx’s Megaclone micro-bead-based tool sounds like every molecular biologist’s wet dream: It approaches, in an area smaller than a square centimeter, the sequencing capacity of Celera’s entire warehouse of ABI 3700s. And in addition to strings of As, Cs, Gs, and Ts, it generates high-resolution gene expression profiles.

The bead-based sequencing process, called Massively Parallel Signature Sequencing, reveals the identity and quantity of nearly all transcripts in a tissue sample, detecting even the weakest signals. Says Leroy Hood, a board member who considers Lynx’s platform to be in tune with his systems biology view, “What Lynx technology allows you to do is find whole new transcriptomes. It’s the next big step in DNA sequencing — parallelizing the process and miniaturizing it.”

All things considered, selling this thing should be a breeze.

But it hasn’t been. As Tycho Peterson of JP Morgan H&Q says, Lynx has “a technology that’s been validated, but they haven’t signed a lot of deals.”

In fact, Lynx has a mere $18.8 million in the bank and its shares are down 90 percent from its high of $99.50.

It has lured a few agricultural and biotech customers. And to be sure, a recent deal with a high-visibility customer will boost the Lynx image: As this story went to press last month, the ink was still wet on a contract Russell had just penned with Celera, which will add expression content generated by Lynx to its database package. Still, Lynx has yet to land a big pharma deal.

Stumped Stars

The Silicon Valley company’s circumstances are surely not for want of talent. Its founder and original CEO Sam Eletr is a cofounder and former chairman of Applied Biosystems. Several other top-level ABI engineers and software developers have joined him. Hood’s Institute for Systems Biology uses Lynx technology. And the company is staffed with folks who know what big pharma customers would want: Pfizer’s former senior director of molecular genetics, Richard Woychik, is CSO. CEO Russell formerly headed AstraZeneca’s biological science and technology group. Furthermore, the brains behind Lynx’s patented inventions belong to none other than Sydney Brenner, renowned genetics pioneer, known best for discovering mRNA and its triplet code of translation, as well as for introducing C. elegans as a model organism.

Nor is the product lacking in the proof-of-principle department. “It works,” says Scott Tingey, head of genome sciences at DuPont, an early adopter of Megaclone. “They can literally deliver to us, in a single sequencing run, 17 megs of sequence from a single hybridization platform.”

In 1998, DuPont shelled out $10 million and committed another $12 million over three years for exclusive rights to apply Megaclone to corn, soybean, wheat, and rice. “It’s the equivalent to using Craig Venter’s capacity to sample extremely deeply into a cDNA library and observe gene expression, without the complexity of the Venter warehouse,” Tingey says of MPSS. “You can sample a million lanes.”

The Fast Lane

Those lanes are actually one million five-micron diameter beads, each bearing 100,000 copies of a unique cDNA fragment. The process of cloning an entire cDNA library onto the beads is what Lynx calls Megaclone, its core technology. Once the DNA is on the beads, it can be sorted, counted, compared, and sequenced all at once, without electrophoretic separation. Lynx doesn’t sequence the entire clone, however ¯ just enough for a unique tag, or signature, to identify the sequence. The reads are 17 bases long, but one million deep.

In an experiment for DuPont, Lynx isolated more than 520,000 corn cDNAs. The fragments’ signature sequences revealed some 56,000 unique transcripts — many that were not in the public or DuPont’s own databases. By simply counting the number of beads carrying each signature, the experiment revealed the relative abundance of each expressed gene. Lynx earned a milestone payment for this and for similar work on two microbial samples from BASF of Ludwigschafen, Germany.

Russell says the company also hopes to excavate its own gold with these tools. “When you’re a technology-based company, you can sell the technology,” he says, “but you would be absolutely derelict if you didn’t apply it for yourself as well.”

To that end, the joint venture BASF-Lynx Bioscience in Heidelberg has generated about 20 patent applications for potential neurological-disorder gene targets, which the two companies will share, and is developing toxicogenomic assays and databases. Another venture, a wholly owned Lynx subsidiary, also in Heidelberg, is seeking out immunopathology-related targets.


Nevertheless, Lynx has yet to fulfill its inventor’s vision. “My aim is to provide everybody with the equivalent of the whole of Celera on a bench,” says Brenner. “And I think it’s doable.”

For now, customers and collaborators who want sequence data must turn their samples over to Lynx. “We’ve invented the microscope,” explains Russell. “There are 31 microscopes in the world and they’re all in Hayward, California.”

The first floor of Lynx headquarters is a large open room with rows of desks for the 150-plus employees. There are no corner offices here. The environment encourages mingling among biologists, chemists, engineers, programmers, and executives, says Russell.

But the second floor is where the real horsepower resides. In several wet labs, young scientists manipulate cell cultures and isolate DNA. A flow cytometry lab across the hall processes beads. Automated machines use fluidics, lasers, and light optics to examine the beads and discard those that don’t make the cut. Next door, MPSS instruments lined up on lab benches do indeed resemble microscopes.

Sequencing reactions take place on glass slides, and fluorescent images of the microbeads bristling with DNA are captured by lenses attached to each instrument. To the naked eye, the one million beads — each one-fortieth the width of a human hair — appear inconsequential, a barely noticeable clump at the bottom of a test tube.

But the significance of that clump becomes apparent as a Lynx technician loads the beads, teeming with DNA, into a microchannel-etched slide for sequencing. Viewed through a microscope, a flurry of beads rushes by like a blizzard of perfectly spherical snowflakes. Indeed, no two are alike — each carries a distinct genetic transcript.

In yet another room, electrical and mechanical engineers hover over instruments in various stages of deconstruction, tinkering with next-generation fixes. “As the microscopes get refined, more and more people will have [them], until ultimately every lab has one,” Russell says.

Brenner prefers a different analogy: “I view this as a huge transition equivalent to the change from mainframe computers to the integrated circuit.”

Like the early advocates of the PC, 74-year-old Brenner has met with resistance for his ideas, but he’s resilient enough to persist. Brenner divides his time between Cambridge, UK, and La Jolla, Calif., and is known to pop up in places like Singapore or Princeton on a moment’s notice.

Still, his speech is slow and deliberate, punctuated with pauses as if to allow listeners to digest his message fully. In the 1960s nobody understood why Brenner insisted on focusing his energy on a microscopic roundworm. Some 30 years later C. elegans was the first multicellular organism sequenced. In the 1980s, he proposed puffer fish as a model for humans. The Human Genome Project flat out rejected it and Brenner has only recently been vindicated.

So it came as no surprise when he began pitching his next brainchild to British venture capitalists in the late ’80s that no one would listen. Brenner’s mind was on a post-genomics universe. He thought: Decades from now we’ll have the complete genome sequence of a handful of people. Then what?

“We’re going to need to find the degree of human variation, and not just by the odd SNP,” says Brenner. The only way to do this, short of six billion human genome projects, is to analyze all the genetic material of cells in parallel.

Brenner says the VCs considered his plan too big a risk. “It needed a really complex group of people of many disciplines. I realized you could only do that in a high-tech region like California.”

Bead Brothers

In Silicon Valley Brenner found a kindred soul. Sam Eletr had left Applied Biosystems, and in 1992 spun out Lynx Therapeutics to exploit some antisense patents ABI had acquired from academic collaborators. Having spent years designing sequencing machines, Eletr appreciated the power of parallel sequencing. Fed up with VCs who didn’t, in 1993 he appointed Brenner to his board and soon after set up a subsidiary called Spectragen to develop Brenner’s bead technology.

Meanwhile, antisense was turning out not to be the magic bullet Lynx had hoped for. “Antisense didn’t have much future,” says Brenner. Besides, he adds, “It was more clinically oriented and Sam felt more comfortable with developing technology.”

In 1996 Spectragen overcame its status as the awkward stepchild and was welcomed as a full-fledged member of the Lynx family. Two years later, when the company gave up its antisense business for adoption (to Inex Pharmaceuticals for $3 million and 1.2 million shares), Brenner’s baby beads were attended to like an only child. “It was the most promising thing that Lynx could do,” says Brenner.

Implementing the technology was difficult and took a bit longer than expected. “We knew that Sydney was starting to camp out with Lynx and so we watched that evolve,” says DuPont’s Tingey. “Sydney’s brilliant. And from an engineering perspective, Sam’s brilliant. That’s a powerful combination. So when you get close to a technology coming out of their heads, you know it’s going to be pretty good.”

In Tingey’s opinion, what they produced is “the best way to do gene expression analysis.”

Tony Kerlavage, senior director of product strategy at Celera says, “This technology has generated a wealth of mRNA expression data that gives us a catalog of transcripts in a broad variety of tissues.”

So why aren’t more customers beating a path Lynx’s microbeads? And why is the company teetering on the brink of solvency?

One reason, says H&Q’s Peterson, who expects the company to break even in early 2003, is that partners must send their samples to Lynx. “A lot of companies are reluctant to do that,” he says. “Especially when there are competing technologies they can bring in-house like DNA arrays.”

Certainly, chip vendors can outdo Lynx on cost. “Microarray technology has the advantage that it is less expensive and much easier to do,” says Hood. To be sure, to build a microarray you need to know what you’re looking for. With Lynx’s beads you don’t — you can even measure the extent to which the genes in a cell are turned on in oysters, pine trees, or rattlesnakes, because no prior sequence knowledge is required.

But for a “quick and dirty” look at a handful of known genes, microarrays are still a better option. At an estimated price of a penny-per-base, a single MPSS run puts a $170,000 dent in the customer’s budget.

Competitive expression analysis methods aside, the bigger reason for Lynx’s failure to wow might be poor planning. Until recently Lynx’s marketing strategy was to have no strategy at all. “We have been reluctant to hype our technologies and their uses,” Eletr told the Wall Street Transcript in December 1998. In the “balance between aggressively promoting one’s technologies and overselling them … we prefer to err on the conservative side,” he said.

In fact, Lynx still has no marketing or PR department and only recently posted a website to get its message out. Management has been so focused on fine-tuning all the enzymatic and engineering particulars that it seems to have forgotten that the point of it all was to build a business.

New Attitude

Prior to his appointment to Lynx two years ago, Russell, 48, headed Zeneca’s target discovery efforts in the UK. He’d spent a lot of time in the Bay Area, scouting out microarray, informatics, and SNP technologies to build the pharmaceutical company’s genomic platform. “California is cool!” he exclaims with a Scottish brogue. In his business-casual gray shirt, no tie, and unbuttoned collar, Russell appears fit for the California culture.

His soft-spoken, low-key manner turns to contagious enthusiasm when he describes his product. Recalling his reaction when Eletr first explained the Lynx platform, he says, “If this really does what he says it can do, then this is the answer.”

It quickly became apparent to Russell that he had inherited an intriguing technology, but no clear business plan. “What I noticed when I came was that Lynx was very internally focused and mostly arranged on getting the technology to work and not well positioned to get out there and market,” he says. “The way I think about it is that Lynx started in January 2000. And that’s sometimes difficult for people to understand because Lynx has been in existence since 1992.”

Making the rounds at investor conferences and speaking to potential customers, he realized that even more difficult to understand than Lynx’s transformations was how the bead technology works. Even the eyes of senior pharma scientists would glaze over during his Power Point presentations.

“The technology is very, very complicated,” says Tingey. “Within my experience in molecular biology for the last 26 years, this is by far the most technically complex system that has ever been brought to fruition.”

Russell found himself wasting much of his presentation time explaining the technical details of the Megaclone process and its various applications — replete with numerous rounds of combinatorial chemistry and a tangled web of enzymology. “It’s what it does that sells it, not how it works,” he says.

He threw more than $100,000 into a state-of-the-art website, in which the knotty details unfold in full animation “I said to the guys here, ‘I want a Nintendo version of our website. I want the Sony PlayStation website — and as quickly as possible.’”

o Russell’s delight, by the first quarter of 2000, Lynx’s outlook had started to improve. The company had validated its technology and had prepared two publications in major journals, including the cover of Nature Biotechnology, leading to mainstream media coverage.

The company had received key patents on its technology and met milestones with several collaborators. Its share price, which had been hovering around $10 since being listed on NASDAQ in 1997, was now approaching the triple digits — a landmark at which Russell had promised to shave his characteristic broad black mustache.

It was the perfect time to raise some funds. Lynx filed for a follow-on public offering of 1.5 million shares on March 14, 2000. Considering its share price at the time, the number could have been enough to put more than $100 million in the bank.

But the date was ill-fated. Russell recalls with a smirk, “That was the day that Bill Clinton and Tony Blair got their PhDs.” Reacting to the politicians’ proclamations that genomic sequence belongs in the public domain, investors bailed. Along with the rest of the sector, Lynx’s stock sank. It has never recovered.

Today, the company has enough money to carry it through mid-2002. “The intention was and still is to go out and raise more money when it’s appropriate,” says Russell.

Hard Sell

For a business that has long followed the “if you build it, they will come” school of marketing, emerging into an intensely competitive climate is not easy. Lynx is preparing to unleash Megatype, an adaptation of Megaclone to identify and isolate SNPs and their disease associations en masse, into a market saturated with at least 45 genotyping technologies. It is also developing a protein analysis alternative to 2D gels — a space already swarming with companies boldly promoting their vaporware.

Add to that the daunting task of selling a technology that, in all its detailed complexity, is as easy to grasp as a fistful of water.

“Even I fully admit that when I heard the Lynx story the first time, while I quickly got what it can do, it took me a while to realize and understand how it worked,” says Russell. “It requires you to think differently about DNA analysis. You’ve got to do some mental gymnastics to get it.”

To his credit, Russell brokered five new deals for Lynx in 2000: an exclusive Asian distribution deal with Takara Shuzo and collaborations with the Institute of Molecular and Cell Biology of Singapore on cancer and infectious diseases, Molecular Engines on tumor reversion, Genomics Collaborative on Type 2 diabetes-associated SNPs, and Hybrigenics on obesity.

nd so far this year, aside from the high-profile Celera deal, the company inked a deal with Phytera to identify genes in several plants where little or no prior gene sequence information is available, and has entered five academic collaborations: with Nobel laureates Michael Brown and Joseph Goldstein of the University of Texas Southwestern on cholesterol metabolism; Sanders Williams, also of UT Southwestern, on cardiac tissue repair; John Todd of University of Cambridge on Type 1 diabetes; Robert Margolskee of Mount Sinai School of Medicine on taste receptors; and David Dempster of Columbia University on osteoporosis.

And, finally, a marketing team is incubating. William Wong has come aboard from Nexell as VP of business development.


Russell is not worried that Lynx is scraping the bottom of the cash barrel. “We’re not desperate,” he says. “When you look at something as powerful as Lynx and you look at its market cap — people are rushing to offer us money.”

nstead Russell thinks it more appropriate to expend his energy on getting the Lynx message out. “I really worked hard at refining how you tell the Lynx story, because it’s so damn complicated,” says Russell. “I’ve seen so many companies advertising their enormous impact on the world of genomics. I think it’s about time we did that.”

Perhaps signing the most visible company in genomics as a customer is just the catalyst the company needs. Buoyed by the recent Celera deal, Russell is confident of things to come. “I’ll tell you what, if share price gets to $100, I will still shave my mustache,” says Russell. But with plenty of work still to be done, “I think my mustache is safe for a while.”

The Scan

Tens of Millions Saved

The Associated Press writes that vaccines against COVID-19 saved an estimated 20 million lives in their first year.

Supersized Bacterium

NPR reports that researchers have found and characterized a bacterium that is visible to the naked eye.

Also Subvariants

Moderna says its bivalent SARS-CoV-2 vaccine leads to a strong immune response against Omicron subvariants, the Wall Street Journal reports.

Science Papers Present Gene-Edited Mouse Models of Liver Cancer, Hürthle Cell Carcinoma Analysis

In Science this week: a collection of mouse models of primary liver cancer, and more.