The process of drug discovery has been dominated in recent years by the screening of large libraries of structurally diverse small molecules for effects on target proteins associated with disease. Technologies such as high-throughput screening (HTS) have been the domain of the private sector for two reasons: (1) high costs associated with the equipment and running of screens and (2) the non-hypothesis driven nature of the work. The role of academic scientists in this process has largely been to explain the genetic/cellular basis of diseases for identifying useful targets and provide new chemical and computational methods to improve the pharmacological profile of potential drugs.

However, the recent National Institutes of Health Roadmap, which identifies priority areas for future NIH funding and support, suggests that many aspects of the drug discovery process will enter the academic arena. The Molecular Libraries initiative of the roadmap is specifically intended to put HTS and other drug discovery tools into the hands of academic researchers to help them develop new research tools and drugs. The upcoming shift in focus toward large-scale screening of small-molecule collections will provide both opportunities and challenges for graduate students and postdocs wishing to explore this new domain of academic research.

Drug discovery in an academic environment, the UCSF experience

So how does an academic scientist fit into the process of drug discovery/HTS and how does it work in the lab? Over the last few years our lab at the University of California, San Francisco (UCSF) has used HTS to discover activators and inhibitors of the cystic fibrosis transmembrane conductance regulator (CFTR) chloride channel. These studies have permitted a firsthand look at how academic drug discovery relates to the training and experience gained as a junior scientist.

An example of the academic drug discovery process and one that my own research has focused on is the discovery of CFTR chloride channel inhibitors and the role of CFTR in intestinal fluid secretion. To identify new high-affinity CFTR inhibitors, a large, diverse small-molecule compound library was screened using a cell-based fluorescence assay developed previously in the lab. After the initial screening, potential "hits" were optimized by medicinal chemistry and tested for potency, specificity, and toxicity.

The end result of this process was the discovery of a potent new class of CFTR inhibitors, ideal for investigating the role of CFTR in epithelial transport. Previous studies have shown that CFTR function is important in enterotoxin-induced intestinal fluid secretion occurring in cholera and traveler's diarrhea. Studies in rodents found that our new CFTR inhibitor potently blocked enterotoxin-induced fluid secretion in vivo and exhibited a favorable pharmacological profile, suggesting that it may be a potential drug candidate for the symptomatic treatment of secretory diarrhea. Diseases of the developing world, such as secretory diarrhea, have little commercial interest for the pharmaceutical industry and represent a gap that is productively filled by academic drug discovery efforts.

There are many stages in the drug discovery/development process from target identification, assay design, and screening through to compound optimization, pharmacology, and testing in animal models. These stages cover disciplines ranging from molecular biology and organic/computational chemistry to cell and whole-animal physiology. Each stage requires scientists with expertise in a specific area and consequently provides an opportunity for people from many different scientific backgrounds to work together.

As in industry, the actual screening process represents a relatively small amount of time in the overall project. The bulk of the time is spent trying to solve particular problems such as design of a robust screening assay or finding the best way of testing the effectiveness of a compound in vivo. These studies can lead to the development of new methods or the improvement of existing ones but are driven mainly by the primary goal of finding useful effectors of the target proteins. Essentially, the model for the drug discovery process in academia is the same as in industry. The difference lies in the objectives of screening, such as the development of new research tools or drugs for diseases of little commercial interest.

Multidisciplinary training

What are the benefits for junior scientists working in academic drug discovery? The development of a project from target protein to potential drug candidate requires the integration of many disciplines and allows scientists to be exposed to techniques far from their own expertise or knowledge. This exposure to an immense variety of techniques and approaches, and the resultant cross-fertilization of ideas, provides a valuable learning environment. A molecular biologist is not going to become an organic chemist, but the necessity of working in a data-rich environment does force scientists to work outside of their usual realm and fosters a more global perspective.

Many of the skills and techniques employed in drug discovery are quite marketable to pharmaceutical and biotech companies and, with the new NIH directive, perhaps even for academic positions. Drug discovery in academia, because of the financial constraints, is on a much smaller scale than industry and therefore allows "hands-on" experience of several stages in the process as well as more extensive training in a single area. It also allows scientists significant input into the overall direction and control of projects even as junior members of the team. It is worth noting that "academic drugs" need not be blockbusters and that most life sciences rely heavily on small-molecule effectors (e.g., A23187, phalloidin, or phorbol ester). The attraction of large-scale compound screening for life science researchers lies mainly in the technical challenge of assay design and the use of small-molecule effectors for hypothesis-driven research.

Designing a good assay

Assay design is perhaps the most important and also the most overlooked element of drug discovery. For a screen to successfully find a drug/small-molecule effector, the assay must be specific, sensitive, reproducible, robust, and relatively cheap. Whereas in the past assays were predominantly in vitro using purified components, there is now a realization that cell-based, high-content assays are a more productive path. Academic labs have always turned to ingenious methods to assay a particular cellular process and these methods are often amenable to screening.

The recent explosion of green fluorescent protein (GFP) based sensors of pH, halides, proteases, kinases, G proteins, and many second messengers (including Ca2+, cAMP, cGMP, nitric oxide, and IP3) have provided important tools for the design of screening assays. An example of these tools in our lab is the characterization and use of halide-sensitive variants of GFP to screen for activators and inhibitors of CFTR in cell-based assays.

New research tools

Screening large libraries of chemicals for activation or inhibition of particular targets can lead to the identification of highly specific and nontoxic compounds that are ideal for explaining the role of a protein in a particular process. For example, the identification of highly specific and potent CFTR inhibitors has led to a number of studies examining the role of CFTR ion channel function in basic physiological processes. These studies have further delineated the role of CFTR in airway gland secretion and identified it as a potential drugable target for secretory diarrheas such as cholera.

The development of high-affinity small-molecule effectors can also potentially be used for the creation of "chemical knockout" animal models. The pharmacological creation of such models ("chemical genetics") has a number of advantages over classical gene knockouts. Firstly, they explain the exact biological role of a protein without problems associated with compensatory regulation of other proteins. Secondly, they enable the use of animals with physiology more similar to humans rather than mice, which are the animal of choice for most mammalian gene knockout models. Finally, they allow the in vivo investigation of proteins where gene knockout leads to embryonic/neonatal fatality.

Overall, the upcoming involvement of academia in the drug discovery process will provide graduate students and postdocs the opportunity to gain experience and skills which are important for pharmaceutical and biotech companies whilst still engaging in hypothesis-driven academic research. Our studies and others have shown that screening of compound libraries for small-molecule effectors offer a powerful method for generating research tools. If graduate student and postdoc researchers can balance their time between basic science research and screening, both industrial and academic roads should remain open at the end of their training.

Jay Thiagarajah, Ph.D., is a CF Foundation Postdoctoral Fellow with the Laboratory for Cell and Membrane Biophysics, University of California, San Francisco.