FRAMINGHAM (09/24/2003) - The naysayers who warned it would take 100 years to complete the Human Genome Project (HGP) will have plenty to fret over about the latest proposal percolating at the National Human Genome Research Institute (NHGRI). Its chemical libraries plan has all the breadth and potential of the HGP and raises similar questions about focus and feasibility. But the plan is part of a larger scheme that moves the National Institutes of Health (NIH) into territory that has traditionally been restricted to industry -- high-throughput screening and small-molecule development. Indeed, some biotechs are voicing concern that it might seriously limit their own prospects for success.
NHGRI Director Francis Collins has been testing the idea on a variety of audiences over the past year. The project, which Collins calls "the centerpiece" of the agency's new five-year plan, would establish several NIH-based centers for high-throughput screening of small molecules.
The proposal already has many NIH supporters, including director Elias Zerhouni, whom Collins says is "strongly in favor of it." Zerhouni has made the development of chemical libraries an NIH-wide priority. According to Collins, 20 NIH institutes and centers are working on the project.
Chris Austin, senior advisor to the director for translational research at NHGRI and an architect of the chemical library plan, says that NHGRI will primarily use the libraries to "determine function and therapeutic potential of genes and to define molecular networks." While NHGRI hopes to pull new drug targets from the genome, other institutes will seek targets for specific medical conditions, including orphan diseases -- typically neglected by industry. "Some of those (chemical) probes just might turn out to be good drug candidates," Collins says.
Costs are not yet available, but the first concrete steps could happen within a year. NHGRI would start by acquiring small-molecule libraries and assembling information on functional assays and chemical synthesis protocols. Scientists would submit screening assays, and collaborate in their adaptation to high-throughput platforms. "Search algorithms analogous to BLAST for gene sequences" would allow researchers to trawl for useful compounds, Austin says. Samples of small molecules would also be made available.
The ultimate goal is to have at least one small-molecule modulator of every human gene product, which could take decades with current technologies. However, a simultaneous goal is to develop novel tools for chemical diversity generation, functional assay development, high-throughput screening, and cheminformatics.
Biotechs have no problems with the technology development aspect of the project. "Having more smart people working on this is bound to help -- I'm excited about that part of it," says Exelixis Inc. Chief Science Officer Geoffrey Duyk. But he and others are concerned about some aspects of the proposal.
A major worry is that researchers new to screening will overreact to early results, clogging the scientific literature with manuscripts "talking about hits, and with a tagline implying that 'cures are around the corner,'" Duyk says. Most active molecules found through high-throughput screening do not pan out as drug candidates. "Hits have long been sold as a milestone in screening," says Anton Fliri, research advisor in Pfizer Inc.'s CNS technology group. "But if you take 100 hits, you may come up with only one drug."
"The unintended consequence (of the libraries plan) could be a lot of noise in the literature and not many practical results -- and another wave of genome hype," echoes Joshua Boger, CEO of Vertex Pharmaceuticals Inc., wary of the suffering biotechs have experienced from investors' overblown expectations.
Intellectual property (IP) is another problem. Boger has a "free-floating anxiety" that this project could shrink the pool of commercially exploitable targets. "Once things come into the public domain, even if they are not already patented, it could be harder to claim a novel use for them," he says.
Specific high-throughput "friendly" functional assays are also extremely valuable. But getting even academics to part with them could be tough. "Certain academic institutions take a hard line on intellectual property," says Mark Namchuk, head of high-throughput screening at Vertex.
Plan architects agree that the IP issues are complicated. "We don't have a final decision on how to proceed, but we believe they can be worked out," Austin says. He adds that NHGRI is committed to giving all interested parties, including industry, freedom to operate with the probes in their collection. The agency hopes that researchers will recognize the community benefits that sharing resources brings.
Tom Insel, director of the National Institute of Mental Health (NIMH), supports the initiative, saying, "We want to bring the power of academic researchers into this effort." That would be researchers like Huda Zoghbi, at Baylor College of Medicine. Her group identified a kinase involved in the development of spinocerebellar ataxia type 1, and is looking for modulators of this gene product. The NHGRI screening plan "could help us," Zoghbi says, "but it depends on what type of molecules they have in the library."
A further concern is whether this is the best use of NHGRI resources. Drug makers contend that screening isn't the slow point: The real pain lies when they move from hits to drugs, trying to find improved ways of characterizing "druggable" and elucidating pathways in animal models. "It would be nice if NHGRI was addressing some of the really hard problems in drug discovery," says Nat Goodman, a senior research scientist at the Institute for Systems Biology. "This is an interesting project, but it's not an obvious choice as a centerpiece."
The NHGRI, meanwhile, is struggling against the perception that its charter goal has been met with the completion of the HGP. Recently, a joint National Research Council/Institute of Medicine report recommended merging NHGRI with the National Institute of General Medical Sciences (NIGMS) -- a suggestion Collins has tactfully deflected for the moment. Zerhouni's office, meanwhile, at press time said it was still considering the document.
But finding new drugs was always a goal of the HGP, and now there is more pressure than ever for that to happen. "Everyone involved with the genome project, including Congress, is anxious to see that translation take place," Austin says.
As part of the plan, the agency will recruit medicinal chemists to refine leads into drug candidates, but it will rely mainly on industry collaborators for subsequent steps. Collins says industry should see the plan as "complementary, not competitive," particularly since the data will be freely available. "Would we use the data? Of course we would," says Pfizer's Fliri. "It all depends on how good it is."
Clearly many researchers are being seduced into small-molecule research. Guiding that process, says Collins, would be a "transforming opportunity" for NHGRI. "Isn't it our job to look into the future to identify areas of intense promise that need further development?" he asks rhetorically.
The Case for Chemical Libraries
The NIH's plan is inspired in part by Harvard University chemist Stuart Schreiber, a champion of the integration of chemistry with biology.
Schreiber's approach is winning fans in high places in academia. He recently became a founding faculty member of Eric Lander's new Broad Institute. Harvard's Institute of Chemistry and Cell Biology (ICCB), which Schreiber founded and co-directs, recently received funding from the National Cancer Institute (NCI)'s Initiative in Chemical Genetics (ICG). That effort centers around building a chemical library (ChemBank) to help accelerate drug discovery.
"Genomics can only take you so far," says NCI's Daniela Gerhard. "Being able to modulate the cell more easily is the next step."
More than 15 groups have contacted ICCB to set up screening facilities, according to Schreiber protégé Caroline Shamu, head of screening at ICCB. Programs are in place at Case Western Reserve University, the University of Texas Southwestern, the University of Kansas, and The Rockefeller University, among others. Most of these groups are focused on studying gene function, but with an eye for drug candidates as well.
However, the first nonprofit screening efforts in drug discovery are also popping up. For example, Harvard recently set up a Laboratory for Drug Discovery in Neurodegeneration, which does high-throughput screening, and has plans for a similar effort in oncology.
The National Institute of Neurological Disorders and Stroke recently established a screening program, and the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) is launching a chemical biology core facility. "We will utilize novel structural scaffolds for identifying new leads and optimize existing leads," says Kenneth A. Jacobson, chief of the NIDDK's molecular recognition section in the laboratory of bioorganic chemistry. "Small molecules have been underutilized at NIH," he says. Jacobson's group has two compounds in clinical trials.
Collins maintains the chemical library initiative is the logical next step for NHGRI: "The importance of small molecules in drug discovery and development is undeniable," he says. But even Schreiber concedes that his particular approach, matching traditional screening with nontraditional molecules, is relatively unproven: Only a few clinical candidates have emerged so far. But, "it's theoretically very attractive," he says. "Let's not fool ourselves that the pharmaceutical industry has already found all the therapeutic targets."
Most pharmaceutical libraries traditionally use the same kinds of guidelines for what makes a putative drug. According to Schreiber, the complexity of biology has hampered drug discovery, and more complex chemicals are needed to fill the missing space.
Jacobson says he's not concerned that so few clinical candidates have yet emerged from Schreiber's lab. "Many of the molecules Schreiber has come up with have unique biological properties," he says. "That's the initial proof to me.