Issue Date: March 7, 2011
If drug discovery were a fishing expedition, then Eli Lilly & Co. would have the fish jumping into its boat. About 18 months after launching its Phenotypic Drug Discovery (PD2) initiative, the company has announced its first agreement with outside researchers. Lilly identified potential drug candidates in a collection of compounds that University of Notre Dame chemistry professor Marvin J. Miller and research associate Garrett Moraski submitted for screening through the open innovation program.
Lilly designed PD2 to tap into new sources of compounds and the science behind them, explains Alan D. Palkowitz, vice president of discovery chemistry research and technologies. “It gives us more options to consider as we try to uncover innovative structures and modes of action.” The 1 to 5% hit rate for submitted compounds is comparable with, if not better than, what Lilly gets by screening its own collections.
Collaborations between academia and industry are not new, but Lilly may be the first major drug firm to ask researchers to join in an open scientific exchange. The agreement with Notre Dame will test whether open innovation benefits both the partners and drug discovery itself. Lilly is negotiating other agreements with other scientists, Palkowitz says.
Although some of the compounds Lilly is catching will be worth keeping, many will get thrown back. Since September 2009, about 300 scientists in 26 countries have uploaded 30,000 compounds to the PD2 website. Roughly two-thirds of 200 participating institutions are universities and research organizations. The rest are small drug firms.
Submitted structures are initially kept confidential. Lilly gets an encrypted electronic fingerprint that it can compare in silico with known compounds. Filters rule out compounds with reactive centers, metal atoms, or other features that would limit their use as drugs. For structures that turn out to have novel, druglike properties, Lilly requests samples for testing against disease targets.
About 20,000 compounds have been physically accepted, and Lilly has completed cell-based assays on nearly half. These assays allow Lilly to rule out compounds with known mechanisms or that interact with undesirable targets and pathways. Through collaborative agreements, the pharma company may explore further hits found against cancer, diabetes, osteoporosis, or Alzheimer’s disease targets.
In return for first rights to license the compounds, Lilly provides the analysis and a full data report for free. Researchers must sign a materials transfer agreement (MTA), but all the intellectual property remains with them or their organization. If no deal results, the researchers can keep and use the data for publication and other purposes.
The data Lilly provides are comprehensive, valuable, and exactly what university scientists need, Miller says. “We are good at innovative and early-stage research, but what we lack is the link to fuller evaluation and development,” he says. “These data are something we never would have had before.”
Over the years, Miller’s group had synthesized many compounds but only tested a few. “We had several thousand compounds around our lab, most of which had never been screened,” he explains. “We gathered a representative set of the different structural types to try to hit as many things as we could.”
For Lilly, an early concern was that PD2 would become “open season for scientists to take everything off their shelves and send it to us,” Palkowitz says. To avoid this, Lilly helps scientists select a manageable number of samples to yield worthwhile data. “We want to be fair and equitable and not have just one group exhaust all of our screening capacity,” he adds.
He believes that many researchers are using the process to test their hypotheses about biological activity. “We are not dealing with a lot of compounds just flowing in without any rationale, but they actually are backed up by some very good ideas and initial scientific investment,” Palkowitz says.
When Miller got involved, he had already consulted with Lilly for more than 30 years. In fact, he helped test PD2’s website, submission process, and shipping logistics. “From the early discussions, it was clear to me that Lilly wouldn’t just take our compounds and run with them but wanted to learn together about what could be the next stage in development.”
From Notre Dame’s initial 200 compounds, some “serendipitously floated to the top for reasons we hopefully are going to discover within the coming year,” says Moraski, Miller’s lab manager. Lilly and the Notre Dame researchers have signed a one-year agreement under which they will further characterize the mode of action for selected compounds and try to understand these compounds’ underlying biology. The deal includes undisclosed funding.
Designed but not successful as antituberculosis agents, the Notre Dame compounds are small, druglike heterocycles that showed activity in Lilly’s angiogenesis assay. The next step is to invest in more chemistry to optimize their activity and see whether they can be advanced into a more mature discovery program, Palkowitz says.
Calling the deal “very collaborative and inclusive,” he says Miller and Moraski will be extended members of the Lilly team who see all new compounds and data. For his part, Moraski is excited about learning about what his molecules can do. “Understanding the biology and its optimization is what med chemists live for,” he says. “I’m really lit up by the opportunity to work with Lilly and not just be on the outside looking in.”
For Lilly, the initiative clearly facilitates finding compounds and lining up partners. A small PD2 team, which continues to see several new users each month, adheres to strict processes for evaluating data and responding quickly to contributors, Palkowitz says. And because of the MTA and shared data, “we are all on the same page in terms of what to do next,” he adds. Moving forward, the Notre Dame researchers now must keep confidential the information they’ve been given.
Lilly is fishing in a pool fed largely by government and industrial funding, so researchers must be sure they own the compounds. Essentially all Miller’s work is funded by the National Institutes of Health, he says, with a few grants from Lilly and other companies and institutes. The university has been kept informed of all interactions and was involved in the MTA and subsequent agreement. Patent applications have been filed on the compounds.
Lilly developed the PD2 program with help from the Association of University Technology Managers, which wanted to ensure that publication and other rights were preserved. Listing the initiative among its technology-transfer resources, AUTM states that participation offers access to relevant assays, publication-quality data, and potential collaborations.
Open innovation still holds cultural challenges for industrial and academic partners, says Carol Mimura, assistant vice chancellor for intellectual property and industry research alliances at the University of California, Berkeley. Such challenges include different attitudes toward publication, timing, and intellectual property.
But the drug development paradigm needs reworking, she points out. Open innovation in many areas, including drug discovery, Mimura says, “has produced results faster, better, and without compromising the participants’ individual missions and goals.”
- Chemical & Engineering News
- ISSN 0009-2347
- Copyright © American Chemical Society