West Coast Editor

It almost sounded like a hedge bet.

When business leaders and their hired scientists latched onto the word “proteomics,” trumpeting that the more refined field of research would push genomics past the point where (at least in the view of many investors) it had become stalled, anyone might have been forgiven for suspecting desperation.

Times had grown hard for genomics. By 2002, at the year’s bellwether JPMorgan H&Q Healthcare Conference in San Francisco in January, no less seasoned an authority than Franklin Berger, managing director of biotechnology for the organization, said genomics “was the favorite word last year [but] has fallen into ill repute, if not disrepute, in the last six months.”

Enter proteomics, which loosely means the study of all the proteins encoded by the genome, and specifically the structure and function of those proteins.

Sounds simple enough. But proteins, vital as they are to the study of disease, are nowhere near as straightforward as genes. They fold, even as they interact and change. How would one measure, say, the way a bunched-up hand towel lies against another especially if both towels are figuratively churning inside the body’s biological washing machine?

The job can be done. Moreover, it’s seriously worth doing, hardly a hedge bet at all. And plenty of firms are entering the space. In fact, so many have joined the fray that keeping track of them, much less dividing them into sectors and finding the leaders in each, is nearly impossible.

Big pharmaceutical companies have figured out the much-ballyhooed promise of proteomics is less likely to fizzle even if it will demand more of an investment than the promise of genomics, which may require proteomics to make it real. Proteins, after all, do the “heavy lifting” in the body. They’re directly implicated in disease, and therefore treatments.

Deemed the veteran in proteomics is Large Scale Biology Corp., which disclosed in June 2001 the first results of its effort to create the Human Protein Index, a database with proteins measured through the firm’s ProGEx platform. A combined set of 157 medically relevant tissue databases yielded 115,000 protein forms, the company said, with many of the proteins showing up in multiple tissues and versions, though encoded by a single gene.

LSBC also has a deal with Biosite Diagnostics Inc. for protein biochips, with the first focus on cardiovascular diseases and liver toxicology.

Another solid player in protein biochips is Ciphergen Biosystems Inc., with its ProteinChip System, launched in May 1999. By the third quarter of 2001, the firm was boasting revenues up more than 130 percent over the same quarter from the year before, a jump from $2.3 million to $5.4 million.

In 2001, Ciphergen bought Invitrogen Corp.’s chromatography business, BioSepra, for $12 million in cash; chalked up orders for its new ProteinChip Biomarker System, its flagship product; and entered a deal with Pfizer Inc. to discover proteins in serum from rat disease models, as well as to evaluate the importance of such proteins as drug-safety biomarkers.

An up-and-comer in the chip space is LumiCyte Inc., which deploys what it calls SELDI, or surface-based molecular capture and imaging, a process that uses pulses of laser light to read molecular profiles on small chips and other molecular probes. LumiCyte raised $20 million in 2001, saying it planned global expansion.

In the proteomics subsector of analytical equipment, a big name is Applied BioSystems Inc., a company of Applera Corp., whose other operating arm is the Celera Genomics Group, considered a leader in the discipline of isolating differentially expressed proteins, noted Charles Duncan, analyst with Dresdner Kleinwort Wasserstein Securities Inc.

Protein-protein interactions also are critical, and firms exploring that zone include CuraGen Corp., Myriad Genetics Inc. and Rigel Pharmaceuticals Inc., with Myriad having been paid the most for its work along these lines.

Other strong names in proteomics: Syrrx Inc., Structural GenomiX Inc. and the German firm Cellzome GmbH, each with a different approach.

Heidelberg-based Cellzome made major news early in 2002 when it published in the journal Nature a paper titled, “Functional organization of the yeast proteome by systematic analysis of protein complexes.” It reported on the first draft of a functional map of the yeast proteome, looking at a network of proteins and their interactions in Saccharomyces cerevisiae.

MDS Proteomics Inc., of Toronto, offered a paper in the same issue on identifying the complexes by mass spectrometry. MDS started with 10 percent of the predicted yeast protein, and was able to find 3,617 associated proteins, covering 25 percent of the yeast proteome.

In May 2001, MDS made news when it teamed with IBM Life Sciences Inc. to form a nonprofit organization that is compiling what it calls the Biomolecular Interaction Network Database of protein interactions, available at www.blueprint.org.

Structural GenomiX uses high-throughput X-ray crystallography to determine the 3-dimensional structures of proteins. The tactic netted SGX a pair of deals in January 2002: one with Millennium Pharmaceuticals Inc. and one with Aventis Pharmaceuticals Inc.

Syrrx works with automated structural proteomics. The company also started 2002 with a bang, hauling down $18.5 million in a preferred stock financing, which brought the amount raised by Syrrx in two years to $98.5 million.

In 2001 came Syrrx’s first deal, an agreement with Cubist Pharmaceuticals Inc., under which Syrrx is helping to characterize anti-infective drug targets, with a plan for designing antibiotics.

There’s no shortage of news to prove the point that proteomics is booming. For investors, competitors in the proteomics field and even curious bystanders, there’s almost too much news to make sense of. Of all the subsectors in proteomics, the structural field is potentially the most valuable, in Duncan’s view.

“People look at protein-protein interactions, and they understand that, if you were able to come up with small-molecule modulators and model the interactions, it would be very important,” he allowed. “But structural proteomics does it faster, and the knowledge of protein structure can be immediately translated into a drug candidate.”

Although the likes of Vertex Pharmaceuticals Inc. and “a slew of private companies” are busy with structural proteomics, Syrrx is the clear leader, he added.

Biology Mediated By Protein Machines’

Nathaniel David, co-founder and director of business development for Syrrx, said the company has made its way forward by “constantly asking, How have we thought about this problem incorrectly?’”

Most of biology, David explained, is “mediated by protein machines, these multiprotein complexes. The dream of what would happen if you knew all the networks and how they interact is that you could disrupt them through a small-molecule drug, and make the disease go away.”

Alas, he said, that seldom works.

“It’s done in a handful of examples,” he said. “It’s kind of like wedging a little piece of something between two boulders, and trying to break them apart. What works is to gum up the actual mechanics of how the protein is catalyzing a chemical reaction. That’s how aspirin works, how penicillin works, it’s how people have been developing drugs for 40 years.”

And it’s what structural proteomics companies do.

“That’s their value proposition,” David said. “You can see how an enzyme-active site is put together, using 3-dimensional measurements. Then you make a drug you can stick in there and switch off the function of the enzyme, and cause the disease to go away.”

Syrrx expects to “solve” visualize the 3-dimensional structure of between 150 and 200 targets in 2002, and use virtual screening as the next step.

“We basically take the protein structure and put it in our computer and fit millions of small molecules into the protein structure,” David said. “The computer figures out which ones look like they will fit. From this set, the computer nominates, out of a million, let’s say 1,000. Vertex also does this solves the structures of proteins and uses computers to figure out what small molecules will stick.”

In a test tube, Syrrx finds that only about 10 percent are found to bind to the protein.

“Then, we start optimizing those, and they go into the clinic,” David said.

So far, though, none has gone that far.

“It’s a very highly attritioned process,” he noted. “And only one in five things that go into the clinic get commercialized, so the way you have to succeed in this is to have a large number of simultaneous programs. Because we can do so many things in parallel, we’ll have 70 full-blown drug discovery programs active over the next 24 to 36 months.”

That’s partly thanks to Syrrx’s November 2001 buyout of Onyx Pharmaceuticals Inc., David said.

Syrrx got certain drug targets, including related reagents and assays; compound libraries; and licenses to intellectual property assets and technologies. In return, Onyx got equity in Syrrx, future milestone payments and royalties on resulting products and Onyx ended its research efforts for the discovery of novel small molecules to focus on development of products that resulted from prior collaborations with Warner-Lambert Co. and Bayer AG.

“We jump-started our program by acquiring the assets of Onyx, which already had nine programs in development, some of which [in cancer and inflammation] we were already optimizing,” David said.

“I think we’ll have our first investigational new drug candidate at the end of 2003 or the first part of 2004,” he added. “It will either be cancer or anti-infective [from the Cubist deal], I’ll bet.”

As time goes on, the sorting of genomics and proteomics companies will continue, and only the strong will survive, David said.

“Just look, for analogies, at biology,” he said. “It’s a natural process of any dynamic growth system. There’s always an ongoing birth, selection and death.”

Smart companies “were lucky enough to raise cash in 2000, vis-a-vis CuraGen Corp., Celera and Human Genome Sciences Inc., and they used it to evolve,” David said. “Celera went out and boldly acquired [Axys Pharmaceuticals Inc.], a small-molecule company. We’re one of their partners.” Celera agreed to pay $173.4 million for Axys in June 2001.

Keeping cash until the industry pans out is the key, David said.

“Once you have a drug, pharmaceuticals is the highest-margin business in the U.S., and that’s why people are so hot to trot on this stuff,” he said. “The only question is how to handle your expenses.”

Another question, of course, is whether the whole effort broad scale and even scattershot as it might be characterized will bear any fruit that the FDA will approve.

“Some people think there will be a higher attrition rate,” David said. “There may be a huge value sink, with all these things dying in Phase II. That would be logical, wouldn’t it? So what’s next is, we need some clever way of being able to do more sophisticated target validation. I don’t know [if there can be one] yet. It’s an unaddressed issue, but it represents an opportunity for people who are very clever.”

Uncertainty, as much as opportunity, abounds.

“We’re not sure this is right; we just believe this is right,” David said. “It’s a good story, but you’ve got to watch it play out, and make sure it works.”