The challenge with natural products is that many have not been synthesized, and it turns out they are often difficult to isolate in sufficient quantities from the organisms that produce them. One emerging solution is to use natural products extracts, mixtures of compounds that are isolated from microorganisms (5). But producing extracts that are clean enough and homogenous enough for screening is complicated. Such extracts might include thousands of compounds of similar molecular weight, yet only a handful might be active in a screen. According to Brian Bachmann of Vanderbilt University isolation is the real problem. “That's where we're putting our energy right now. If you had a method that could beam compounds out of an extract and put them in tubes, that's what we need.” –SW

The follow-up gap

Outside the specialized groups that NIH has funded to carry out medicinal chemistry work, finding the resources to follow up on initial screening hits can present a bottleneck in chemical genomics research, says Hakim Djaballah, Director of the High Throughput Drug Screening Facility at Memorial Sloan-Kettering Cancer Center. Even with these NIH funded centers, the resources devoted to this piece of the puzzle are relatively small compared with the number of screening centers that are producing hits from high throughput screening. Following up on screening hits is risky and is not necessarily hypothesis driven, which can make it difficult to secure funding. “I'm involved in at least 3 or 4 projects like that that are tangibly parked because we don't have any money to move them forward,” Djaballah says.

Even once medicinal chemistry is done there's still the question of doing secondary and tertiary screening on the optimized compounds. “The professor who develops the assays then has to come up with the funding for the secondary and tertiary assays and the functional assays to discern whether these [optimized] compounds are biologically relevant,” notes Rathnam Chaguturu, director of the High Throughput Screening Laboratory at the University of Kansas.

Non-profit foundations with an interest in a particular rare disease can form one potential source of funding for projects aimed at questions in particular disease areas. Inglese, who studying Charcot-Marie-Tooth disease, a rare peripheral neuropathy, has funding from a foundation that supports this research. Another approach that could shorten the process of lead optimization is to start with a smaller pharmaceutical library of known drug compounds, such as the one NCGC assembled and published in Science Translational Medicine (3). That process of repurposing known drugs, whose potency and activity have already been thoroughly tested, for new indications provides a fast track to the clinic. “A disease foundation wants that,” says Inglese. “They could use that information to begin a clinical trial with patients.”

For those molecules that might have clinical applications, researchers and institutions would like to hold on to the intellectual property rights, but use of the NIH Molecular Libraries Program resources requires researchers to deposit the structures of the molecules that they discover into a public database. “By definition and by mandate, once done you have to post the structures into Pubchem. And once you've posted in Pubchem database, you've lost your intellectual property,” notes Chaguturu. “That's one of the biggest drawbacks.”

Quantitative high throughput screening (qHTS) allows researchers to obtain dose-response relationships during their initial screen of compounds. (Click to enlarge)

Although currently funded through 2014, questions surrounding future funding for the Molecular Libraries Program loom. High throughput screening is expensive, and it's a kind of business transaction in the academic world, says Djaballah. “The initial investment is written off by a university, but maintaining [HTS centers] is expensive.”

“This network has put into place some really fantastic infrastructure and has enabled some tremendously talented scientists from all over the country to engage in this work,” says Aubé. “It would be wonderful to see those capabilities leveraged in the future, and I'm not exactly sure how that will happen.”