Join MySLAS Social

Uniting Phenotypic and Target-Based Drug Discovery

Many people see target-based drug discovery and the phenotypic approach to drug discovery as antagonistic but Fabien Vincent, associate research fellow in the department of Hit Discovery and Lead Profiling at Pfizer, sees them as complementary. This is the message he hopes to convey during the SLAS2017 session “Uniting Phenotypic and Target-based Drug Discovery,” part of the Drug Target Strategies Track.

When Vincent began his doctorate studies at the University of Houston in the mid-1990s, he already knew one thing: he was not interested in the epilepsy drug discovery program being conducted in the lab at the time, a project that involved synthesizing compounds on a 5 to 10-gram scale, sending them to the National Institutes of Health (NIH) and waiting several months for activity results in an animal model of epilepsy. Instead, he chose to work on the bacterial enzyme inhibited by the antibiotic bicyclomycin. Here he could use exciting tools and techniques that weren’t available in the epilepsy project, where the target was unknown. He learned enzymology while deciphering the mechanism of action of both the enzyme and its inhibitor and went on to enter the field of drug discovery where he spent several years working on a number of target-based drug discovery programs in biotech. “It was very exciting and interesting but it never led to a successful drug.”

Meanwhile, the in vivo safety and efficacy profile of one of the compounds from the epilepsy program had garnered the attention of a pharmaceutical company. In 2008, that compound was approved by the United States Food and Drug Administration (FDA) as lacosamide, a novel anticonvulsant of great value to patients with partial-onset seizures.  

It wasn’t until a few years later, after reading the 2011 landmark paper in Nature Reviews in Drug Discovery by David Swinney, that Vincent thought back to that experience and began developing an appreciation for the value of phenotypic screening conducted in a highly disease-relevant system. Today, his work is centered on improving the translation of pre-clinical research to patients by including phenotypic screening and atypical molecular mechanisms of action to increase the odds of successful translation from bench to clinic.

Vincent sees target-based discovery as having the advantage of allowing investigators to design biochemical assays against a known target to gain an understanding of the interaction between a compound and a specific protein. “Using tools such as X-ray crystallography, chemists and structural biologists can then derive a hypothesis as to which compound to pursue,” he says. This approach is appealing not only to scientists but also to clinicians. “It can be faster, more efficient, and knowing the target means researchers can design experiments to verify that the compound is occupying the desired target in patients and predict undesirable side effects. When there is a strong link between a disease and a target, this approach makes sense.” But, he adds, there is a very limited number of such targets.

Swinney, SLAS2017 Drug Target Strategies Track chair and CEO of the Institute for Rare and Neglected Diseases Drug Discovery, says “We’ve been taught to think the target is the therapeutic hypothesis.” He believes this target-based, hypothesis-driven approach has become too reductionist. He remembers the beginning of the molecular biology revolution when “the hope was that once the human genome was mapped it would provide an understanding that would enable rational design of new medicines; the hope was that the process of drug discovery could be engineered.” He believes the biology is too complex for such a reductionist approach to predict exact molecular and mechanistic details of complex human diseases. “We’ve been doing target discovery, not drug discovery. We would spend all this time in the preclinical phase learning how to modulate the target and developing a target hypothesis and it wasn’t until the drug was tested in the clinic that we discovered it just wouldn’t work in the larger system.”

And, as Vincent will attest, the pharmaceutical industry often picks the wrong target. “Simply binding a compound to the target doesn’t work if the target has no connection to the disease.”

Vincent says the phenotypic approach aims to address this drawback by starting with a model that resembles the disease as closely as possible rather than a target that may have only a weak link to the disease. “It is an unbiased approach. We look for what works instead of what we think should work.” Eliminating the bias of presuming too much about which target should work leaves the investigator open to discovering the complexities of the biology. “We don’t always have the ability to imagine a mechanism of action that is both safe and effective.” He gives the example of docosanol, a topical treatment for recurrent oral herpes or cold sores that was developed using a phenotypic approach. The mechanism of action doesn’t appear to involve direct activity against the herpes simplex virus proteins but rather seems to interfere with its lipid membrane. “That would not have been pursued in a target-based fashion,” Vincent says.

According to Vincent, there is a synergy to using both approaches together. He says the information about specific targets can help inform the design of assays, especially when there is genetic evidence of a link between a specific pathway and a disease. A recent example of this is the use of phenotypic screening in the discovery of a novel small molecule that inhibits the translation of PCSK9.

Pfizer isn’t alone in revisiting the value of phenotypic drug discovery methods. In fact, Swinney notes that since the publication of the Nature Reviews paper, and a follow-up review published in Clinical Pharmacology and Therapeutics in 2013 more and more people across the industry are beginning to look for ways to incorporate phenotypic screening into their discovery programs to deliver targets that accurately represent the biology. The U.S. Army Medical Research Institute of Infectious Diseases (USAMRIID) recently combined phenotypic and target-based approaches to help identify and characterize antiviral activity of GS-5734 against the ebola virus. But Swinney says this approach is not without its challenges.

Back to the Future?

Many people see phenotypic drug discovery as old-fashioned. But Vincent says that while phenotypic methods were used before the 1990s, that doesn’t mean we are going back to these same approaches. “We are taking the same philosophy of an unbiased system and using new 21st century tools to help with these programs.” Still, he acknowledges that it can be difficult to go back. “The current generation of scientists has been brought up in the target-based drug discovery era and many of the tools and processes have been optimized for target-based drug discovery. Some scientists are uncomfortable with having to conduct compound optimization using assays where we don’t know the target.”  

An important step to overcoming this challenge is finding a phenotypic assay that is highly relevant to the disease. Vincent explains that if the assays are designed appropriately, there is a greater likelihood that the compounds found to be active in the assays will also be effective in patients. But setting up these disease-relevant assays isn’t always straightforward. “It can be time consuming, costly and it requires a lot of expertise.” In his recently published paper, Developing Predictive Assays: The Phenotypic Screening 'Rule of 3,'” he and Pfizer colleagues describe the characteristics of the ideal phenotypic assay. Swinney agrees this is key and adds that another challenge is the different skill sets involved. “It takes bringing a team of people together who have the knowledge of biology to run the assays with people with the cellular and molecular expertise.” That was the focus of the Keystone Symposium on Phenotypic Drug Discovery that took place earlier this year, in which both Vincent and Swinney participated.

Shifting Paradigms

Swinney believes it may take a slight paradigm shift to overcome some of the challenges of combining the two approaches. “One way to look at it is these are just different tools in the same tool box and you’re probably going to use both depending on the stage of the program and information you have. It may take longer pre-clinically to find the right models to identify the right molecules using the phenotypic approach, but once these molecules are identified they will have a better chance for success in the clinic.” He also thinks scientists are going to need to become more comfortable with incomplete knowledge. “If you have an unmet medical need and you want to discover something totally new, an empirical approach is useful. That’s how you find out what you don’t know. Once you‘ve identified a new therapeutic mechanism empirically using phenotypic screening you can use the molecular, reductionist approach to identify the targets and molecular mechanistic details that will be of use for the development of the first in class medicine. Subsequently, followers and advances in class drugs will use this knowledge to inform targeted-based approaches.”  

Vincent sums it up: “It’s not us vs. them. We are likely going to be more successful when we intelligently combine the two.”

Learn More

Vincent is chairing a session in the SLAS2017 Drug Target Strategies Track highlighting the complementarity and synergies of utilizing both approaches to drug discovery. The session includes presentations by Paula Loria, associate research fellow at Pfizer, on how phenotypic screening identified a small molecule that inhibits PCSK9 protein translation and by Veronica Soloveva, senior research scientist at U.S. Army Medical Research Institute of Infectious Diseases, on how combining phenotypic and target-based approaches helped identify and characterize antiviral activity of gs-5734 against the ebola virus. Additionally, Swinney explores the differences in the therapeutic hypotheses associated with the two approaches, and Sharlene Velichko, director, research biology at DiscoverX, talks about combining the two strategies for drug repurposing.

December 4, 2016