Inside Applera: Celera Genomics Becomes Applied Bio's Diagnostics Angel

Applera Corp. 's July 24 announcement of a $75 million high-throughout genotyping initiative to be equally funded by its three businesses, Celera Genomics Group , Applied Biosystems Group (ABI), and Celera Diagnostics (a joint venture of Celera Genomics and ABI), coupled with its conference call two days later to discuss earnings, has provided a context for the company to begin to reveal its diagnostics strategy. New information about the structure of the joint venture, in which Celera Genomics is absorbing the short-term financial risk by covering the JV's initial operating losses, also sheds more light on the acceleration of Celera Genomics' business strategy as it moves away from an information-based genomics platform model toward that of a partnering-oriented pharmaceutical company. (See "The Proteomic Evolution of Celera, Incyte and Myriad," START-UP, June 2001 [A#2001900098.)

The new genotyping program is a way for Applera to further leverage the return on its human genome sequencing investment. It calls for Celera Genomics to resequence genes and regulatory regions from 40-50 individuals over the next 12-15 months in order to identify patterns of genetic variation—groups of SNPs or haplotypes. Assuming process improvements along the way, Celera Genomics says it will then be able to map all of the genes and regulatory regions for any person with any disease in about a week, for $1 million or less. ABI will use the haplotype data to develop validated reagent sets—oligonucleotide probes and primers based on its sequence detection system(SDS) platform—which it expects to begin selling to the research community within a year, the first revenue source to flow from the initiative.

Celera Diagnostics, meanwhile, is establishing an industrial-scale facility for high-volume genotyping and analysis of gene expression, including genes expressed at low levels that cannot be detected using conventional DNA chips. It is also gathering clinical samples in anticipation of starting association studies in which it will use ABI's oligos to identify diagnostic markers by correlating haplotypes with disease. In the near term, Celera Diagnostics will develop molecular diagnostic tests on ABI's SDS platform, including commercialization of the ViroSeqHIV genotyping system it submitted to FDA for 510(k) approval in July. In the intermediate term, Celera Diagnostics will convert the data from the genotyping initiative into content for diagnostic tests, which it will outlicense to CROs or reference labs or develop internally. Farther out, it will discover, develop, and commercialize proprietary tests and, with ABI, will develop a proprietary proteomics-based systems platform.

Applera has been eyeing the molecular diagnostics opportunity –currently pegged at $750 million to $1 billion per year and growing at 25-30%—even before it began its direct genome sequencing effort eight years ago. Since then, it participated in the business in a small way, mostly through the development of an HIV genotyping system. Encouraged by the success of its SDS platform and TaqManprobes—and having obtained a license from Roche in July 2000 [See Deal] to use PCR in clinical diagnostics—last fall it announced the diagnostics JV and the hiring of Kathy Ordonez, formerly president and CEO of Roche Molecular Systems, to head it and actively formulate the diagnostics strategy. (See "PE's Diagnostics Start-Up," START-UP, November 2000 [A#2000900204.)

The diagnostics program started out as an ABI venture, notes Tony White, Applera president and CEO, then evolved to include Celera Genomics. In a sense, Celera Genomics, with a billion dollars of assets in cash, is playing the role of a deep-pocketed drug company with a long time horizon for development. Celera Genomics will cover Celera Diagnostics' initial operating losses up to $300 million—which protects Applied Bio's P&L in the early days of the JV. (Including the contribution to the genotyping initiative, Applera estimates Celera Diagnostics' losses at $55-65 million for the next fiscal year, mostly from R&D expenses.) Applied Bio will also reimburse Celera Genomics for all tax benefits generated by Celera Diagnostics that Applied Bio uses. That amounts to 35% of Celera Diagnostics' losses, in effect making the split of cash losses 65/35 in favor of Celera Genomics. Profits from Celera Diagnostics will be similarly split 65-35 until losses are equalized, and 50/50 thereafter.

"We worked it out so Celera [Genomics] shareholders could participate in a reasonable way," White explains, with Celera and ABI benefiting equally even though ABI donated a sizable asset to the JV—its molecular diagnostics business, including its genotyping program, principally in HIV, and the rights to use PCR for diagnostics. The value of the PCR rights over time "can conservatively be estimated at tens of millions, if not hundreds of millions, of dollars," White says. Now, the information from the genotyping initiative will accelerate PCR-based assay development, enabling ABI to monetize the value of that PCR license more broadly and rapidly. On the downside, however, ABI will lose revenues from the sale of reagents and instruments to Celera Genomics, which it will now be providing at cost as part of the genotyping initiative.

As part of the initiative, each of Applera's businesses will have access to technology and materials—instruments and reagents—supplied by sister businesses at cost. Intellectual property claims will flow according to business focus—therapeutics to Celera Genomics, diagnostics to Celera Diagnostics, and instrumentation and consumables to ABI. Celera Diagnostics will also have access to data from Celera Genomics' proteomics efforts, which Celera Diagnostics can use to develop biomarkers for development or out-licensing. Over time, IP from discoveries made by Celera Diagnostics may also flow back to Celera Genomics' therapeutic target discovery program.

Arguably, the genotyping initiative doesn't divert resources from either ABI or Celera Genomics and is core to the success of Celera Diagnostics—the best of three worlds. For its part, ABI would have been developing validated reagent sets anyway, including an announced plan for a 200,000 SNP set for linkage disequilibrium studies. Celera Genomics similarly would have embarked on a resequencing program even without the diagnostics opportunity as a driver, based on its analysis of data from the five individuals it initially sequenced and the resulting observations that the number of genes in the human genome was lower (around 30,000) than expected, the number of expression variants was higher, and that less than 1% of gene variances coded for proteins.

While the Celera revelations didn't drive Applera's decision to initiate the high-throughput genotyping program, says Tony White, they may have changed how the company approached diagnostic marker identification. Adds Kathy Ordonez: "The low number of genes identified and the large number of expression variants indicated that gene expression could be extremely important. [The observations] helped catalyze excitement about the power of genetic approaches and encouraged Celera Diagnostics to think about new approaches for identifying disease associations."

Indeed, those observations provided the rationale for the size and scope of the genotyping initiative: The data set will be small enough so that the company won't have to deal with millions of SNPs, claims Celera Genomics' president Craig Venter, PhD, because in any one individual only a few thousand SNPs will be medically relevant. "The number of SNPs is not overwhelming," he says, "and we should see clear-cut differences in population studies." Celera Genomics will use the haplotype data both to support internal R&D (including for predictive efficacy and toxicity of new compounds) and to support the Celera Discovery Systemon-line subscription business. Not that there's zero opportunity cost: Celera Genomics may have opted for a slightly different strategy, perhaps by doing a deeper resequencing of fewer genes (disease-associated genes it has already identified) in a larger group of people—a capability it may still offer drug development partners.

Celera's unexpectedly strong financial position accelerated its drug development plans and the corresponding shift to a discovery outlicensing and commercialization-via-partnering model. "We thought we'd have to spend five years getting Celera Genomics' on-line business to profitability to be ready for drug development," White says. With that business now expected to be profitable by the end of the fiscal year, "the five-year time frame became a two-year time frame." Also unexpected was Celera Genomics' ability to raise $944 million in a March 2000 follow-on offering [See Deal], which it is using to build its biological capabilities and scale up its proteomics efforts. The financing also enabled it to acquire Axys Pharmaceuticals Inc.in June 2001, giving Celera small-molecule screening expertise to apply to the genomics targets it has identified [See Deal], and a month later to enlist the aid of Isis Pharmaceuticals Inc. 's GeneTrove functional genomics division to analyze gene function [See Deal].