Join MySLAS Social

The New Future of Phenotypic Drug Discovery

While there may be different strategies, varied techniques and multiple diseases studied, universally accepted across the board by the talented men and women working in drug discovery is each person wants to put safe and effective medicines into the hands of physicians treating sick patients as quickly as scientifically feasible.


The best ways to accomplish this, of course, remain in flux.

A strong voice for phenotypic drug discovery (PDD) has been heard the last couple years within the SLAS community. PDD advocates have been sharing information widely at the SLAS Phenotypic Drug Discovery Special Interest Group (SIG) meeting at the annual conference and through year-round networking on the SLAS Phenotypic Drug Discovery SIG on LinkedIn.

Likewise, SLAS is offering additional outlets for exploration, including a three-part SLAS Webinar Series (September to November 2013 live; then On Demand), a two-part special issue of the Journal of Biomolecular Screening (JBS December 2013 and January 2014) and scientific sessions at SLAS2014, Jan. 18-22, 2014 in San Diego.

Driving the Interest

In 2012, Nature Publishing's Science-Business eXchange Senior Editor Joanne Kotz ran an article, "Phenotypic Screening, Take Two." A brief excerpt:

Two broad types of screens have sequentially dominated early stage drug development over the last hundred years or so—phenotypic screens and target-based screens. The former looks at the effects, or phenotypes, that compounds induce in cells, tissues or whole organisms, whereas the latter measures the effect of compounds on a purified target protein via in vitro assays.

Phenotypic screens used to be the mainstay of drug development. Such screens can potentially lead to the identification of a molecule that modifies a disease phenotype by acting on a previously undescribed target or by acting simultaneously on more than one target. However, subsequently determining the relevant target or targets of molecules identified by phenotypic screening has often proven slow or impossible.

Beginning in the 1980s, advances in molecular biology and genomics led to phenotypic screens largely being replaced by screens against defined targets implicated in disease.

Over the last decade, however, some drug developers have questioned whether an over-reliance on genetic approaches to validating targets for subsequent target-based drug discovery has resulted in reduced success in discovering first-in-class medicines.

The last paragraph from Kotz referenced a journal article from Nature Reviews whose co-author David C. Swinney, Ph.D., is one of three leaders for the Fall SLAS Webinar Series, Phenotypic and Signaling Network Approaches as an Alternative to Target Based Drug Discovery, sponsored by Cellular Dynamics International; and a contributor to the special issue of JBS.

In that paper, Swinney and Jason Anthony reported on their studies analyzing the discovery strategies and the molecular mechanism of action for new molecular entities and new biologics approved by the U.S. Food and Drug Administration between 1999 and 2008. They reported:

Out of the 259 agents that were approved, 75 were first-in-class drugs with new MMOAs (molecular mechanism of action), and out of these, 50 (67%) were small molecules and 25 (33%) were biologics. The results also show that the contribution of phenotypic screening to the discovery of first-in-class small-molecule drugs exceeded that of target-based approaches — with 28 and 17 of these drugs coming from the two approaches, respectively — in an era in which the major focus was on target-based approaches. We postulate that a target-centric approach for first-in-class drugs, without consideration of an optimal MMOA, may contribute to the current high attrition rates and low productivity in pharmaceutical research and development.

Swinney explains that during this study period, 90-95 percent of drug discovery efforts were by a molecular approach making the strong phenotypic showing stand out even more. The JBS special issue guest editors and SIG chair agree this manuscript has likely triggered more companies to take a deeper look at phenotypic methods and jumpstarted a major turning point in activity.

Swinney's Exploration

Swinney and Anthony delved into their study, as the two scientists needed to better understand why there weren't higher success rates for drug development.

"I had been with Roche Palo Alto for many years doing drug discovery," Swinney says. "Success rates were not as great as we would have liked, so I analyzed this from a biochemical pharmacology perspective and concluded that how drugs work, their mechanism of action, was important to success because it helped define the therapeutic index. This work was published in Nature Reviews in 2004 and subsequently, I tried to apply this knowledge to the programs at Roche. But it was not clear how to a priori select a target and mechanism of action that would be successful – have an acceptable therapeutic index. So I did a further analysis to learn how the mechanisms of action of approved medicines were discovered." This analysis led to the 2011 Nature Reviews paper.

"Learning of the phenotypic screening success was a surprising result," Swinney notes. "I'm a molecular person from my training so it was a good experience to bring the two together."

His webinar presentation, The Value of Phenotypic-Based Drug Discovery, will be held live on September 26, at 11:30 a.m. ET. Swinney will begin by giving an overall perspective of phenotypic-based assays in drug discovery – past, present and future – beginning before the 1900s with herbal medicines and then moving into more classic pharmacology and how to incorporate phenotypic assays into drug discovery to increase overall success.

"Some people say we are 'going back to the future' with this movement to increased phenotypic assays," Swinney says. "I say we're going to a new future, which really integrates hypothesis and empirical driven research. I tend to think that biology is just too complex and too dynamic at a molecular level to ever predict all the interactions. With phenotypic assays, you are able to identify starting points. Then, once you've identified them, the molecular approach evolves and becomes more relevant."

Swinney is a regular speaker on phenotypic assays and binding kinetics in addition to running the company he co-founded in 2010 – iRND3, the Institute for Rare and Neglected Diseases Drug Discovery. iRND3 is a non-profit corporation seeking to "discover new medicines to address health inequities for rare and neglected diseases afflicting patients throughout the world."

For the JBS special issue, Swinney updated the 1999-2008 data that appeared in his original, groundbreaking paper. His "The Contribution of Mechanistic Understanding to Phenotypic Screening for First-in-Class Medicines" appears ahead-of-print at JBS Online, as do many other manuscripts from the special issue. Additionally, Swinney will take part in the SLAS2014 special session, Phenotypic and Target Based Drug Discovery: Point-Counterpoint, chaired by Lorenz Mayr of Astra Zeneca.

When and Why to use Phenotypic Screens

Ellen L. Berg, scientific director of BioSeek (a division of DiscoveRx), feels her contribution to the field and to the SLAS Webinar Series is to talk about some details – what makes a good phenotypic drug discovery screen and how to manage the data. Her November 12 webinar at 11:30 a.m. ET is titled Phenotypic Drug Discovery Using Primary Human Cells and Co-Cultures: Lessons Learned.

"Typically the data is multivariate," she offers. "It comes from all kinds of different platforms and requires either different kinds of normalization or ways to pull different data types together to make the best use of a phenotypic screening assay. You also need to understand the cascade of assays that one wants to use in a phenotypic drug discovery program; it's very different than target-based drug discovery (TDD). I will share lessons learned on what to do when in my webinar presentation."

Berg has been perfecting phenotypic screens for some 14 years now. "In the early days, we would test compounds from other people's screens and many programs were pulled during this time because scientists had not realized their compounds had unexpectedly hit relatively important cellular pathways. They were not cytotoxic, not deadly to the cells, but they were uninteresting to develop as a drug.

In her webinar presentation, she will talk about the most frequently found mechanisms in phenotypic screens and why some of these common pathways that show up must be quickly triaged out. Additionally, Berg will review common findings, like epigenetic targets that show up frequently and are hard to deconvolute because nobody has panels of assays for many such mechanisms.

"I have a pharmacology background and recognized many years ago that target-based assays just don't predict in vivo effect," she says. "We'd identify a target, validate it in a particular way so we knew some of the biology of the target, but it was always a surprise when you'd go into animals or people with drugs that inhibit the target and all kinds of other biology shows up. And sometimes it is not very good biology. That's why I was always very interested in supportive phenotypic assays that would help us better understand the target. You would have a compound that was very target specific. You would follow that up with a phenotypic screen and, well, you always moved forward with the compounds that were active in the secondary, phenotypic screens. So it is kind of a no-brainer to make those phenotypic screens high-throughput and screen on them to begin with."

Berg adds, however, that depending upon your goal, there are certain times when target-based discovery might be the better choice. "While phenotypic assays have given us most of the first-in-class drugs, the best follow-up drugs are often target-based," she says. "But certainly for innovation, you would want to try something new. You can reduce the risk by having diverse approaches."

Berg believes webinar participants will gain more confidence and therefore be more willing to consider phenotypic assays after hearing her presentation.

"I believe if they incorporate some of my 'lessons learned' into what they do, it will improve their success and reduce their timelines," Berg notes. "I want to make participants more willing to try phenotypic drug discovery."

As co-guest editor of the JBS special issue on Phenotypic Drug Discovery, Berg was amazed at the number of success stories, as well as variety of biologies and therapeutic areas captured in the submitted manuscripts.

"We received 95 excellent manuscript proposals and through this special issue of JBS, the diversity of the biology, the different analysis techniques and the novel ways to analyze data will be captured in one place," she says. "Typically, these would be dribbled out over many issues or contained in various journals. It will be terrific to have these successful phenotypic reports together in one place."

"I'll be a session chair at SLAS2014," Berg notes. "I have been very impressed with SLAS and its electronic channels, training and information. This is something scientists really like and appreciate. And, the phenotypic drug discovery topic was an easy fit for me to become involved. At SLAS2013, the SIG room was overflowing. There is definitely a resurgence of interest in phenotypic drug discovery. I want to be a part of it!"

Making the Business Case for PDD

Jonathan Lee, senior research advisor at Eli Lilly Research Laboratories, is quite pleased that SLAS has provided a platform for phenotypic drug discovery discussion. Propelled by the SLAS2012 discussion, Lee worked with SLAS to form the SLAS Phenotypic Drug Discovery SIG and set up a LinkedIn subgroup. Members flocked to these opportunities, robust discussion began and it continues to this day. The SLAS PDD SIG is the most active year-round SLAS SIG. Lee is chair of the SLAS PDD SIG and co-guest editor of the JBS special issue.

"SLAS has provided the infrastructure for those interested in phenotypic to have thoughtful discussions, but the work is not finished of course," Lee says.

In many ways, Lee indicates PDD has a renaissance sort of a feel, as it is a rebirth of an earlier drug discovery method. Lee explains that before scientists were able to readily clone and express molecular targets and before genomes were sequenced, the information that drove most drug discovery was biological function.

Pharma workflows started to change in the mid-1980s. Lee offers. "Molecular biology and genomics was integrated into the drug discovery process and offered a catalog of new targets and the ability to express and purify these proteins relatively quickly. At the same time there were advancements in automation – high-throughput screening became something to be reckoned with. In addition, there were advancements in parallel chemistry with the promise of new chemical diversity and rapid synthesis."

Lee believes it was the convergence of all these capabilities – automation, biology and chemistry – that fueled the concept of developing high-throughput, high-capacity assays that are principally biochemical in nature.

Through this period of time, the cost of doing pharmaceutical research was increasing exponentially; new approaches were needed and the confluence of these technologies promised a revolution in drug discovery. "Focusing on molecular targets made a lot of sense; understanding and predicting biology in molecular terms was exciting and elegant; molecular approaches also promised to make drug discovery more process oriented, faster and cheaper," Lee says.

Lee takes seriously, however, that there has not been a dramatic increase in the rate of production of new molecular entities approved by the FDA in the post molecular/genomic period.

"The Swinney paper gave the industry a surprise; more first-in-class drugs were being identified by old-fashioned physiologic, functional assays and not by the biochemical, target-directed assays," Lee says. "And I think an even more important finding in this research was that the overall rate of first-in-class drug launches from the phenotypic approach was two to three times higher than targeted strategies even though it was likely less than 10 percent of lead generation in pharma was using PDD during that period," Lee continues. "TDD and PDD are complementary approaches to drug discovery, each with their inherent strengths and weaknesses. The major feature that differentiates TDD from PDD is the requirement for identification of a physiologically relevant and validated molecular target for the former while the latter utilizes a target agnostic approach."

The PDD LinkedIn subgroup has discussed how Swinney's paper has been validated by other analyses.

"Two papers (Prinz et al Nat. Rev. Drug Discov 2011, 10, 712; Begley and Ellis, Nature 2012, 483, 531-533) indicated how difficult it is to reproduce published target-validation data," Lee explains. "In fact, it was said that about 65 percent of the studies based on target validation had to be terminated because certain pivotal parts of the study linking the target with the disease were loose enough that scientists felt they could not go forward. Blog conversation revealed that many actually were surprised that 35 percent were correct. Part of the problem is that you can get a lot of data points with high-throughput, target-directed million compound screens, but the underlying hypothesis linking it with disease may not be that strong. In my limited experience, there have been projects where we had developed very specific cell-permeable, cell-active compounds but it didn't actually elicit the overall functional biology that we wanted it to do."

Lee credits the management at Lilly for encouraging investigation of potential benefits of PDD and supporting Lilly pilot studies to determine whether PDD systems are compatible with pharma workflow. Seeing positive results and interesting molecules, they began work within Lilly therapeutic areas to incorporate PDD into the portfolio.

"Space was cleared out by our management to investigate a different way to do discovery research," Lee concludes.

Lee will share information from those pilot studies during his webinar presentation, Modern Phenotypic Drug Discovery is a Viable Pharma Strategy, at 11:30 a.m. ET on December 5. He hopes sharing examples will help clarify many common perceptions of PDD limitations held by many in pharma.

"The Swinney paper stirred the pot," Lee notes. "At that point in time, the vast majority in pharma – perhaps 85-90 percent thought PDD was not a good idea. The paper made people really start to think about it."

Lee says there are a lot of intellectual, emotional and business hurdles to overcome, as pharma holds a molecular mindset. "Almost all practicing scientists today – both chemists and biologists – were trained when molecular cloning, purified proteins and x-ray analysis were pretty commonplace. We have much re-education to do and not just with the scientists. But we are not throwing out all our molecular technologies or appreciation for the capabilities we have developed over the last 30 years. There are so many tools now; it's an exciting time."

Where Do We Go From Here?

Swinney, Berg and Lee feel the webinars, JBS special issue and SLAS2014 sessions will offer loads of information and bring people together to discuss PDD and TDD. They believe people need to consider all knowledge gained over the years and then use that knowledge to instigate change within their work as required.

"We can lay the foundation about what we learned in our studies, and people then need to pull that into their own work and become comfortable," Swinney concludes.

"It's not an 'I'm right and you're wrong' scenario," Lee says.

SLAS Webinar Series

Phenotypic and Signaling Network Approaches as an Alternative to Target Based Drug Discovery
Sponsored by Cellular Dynamics International

September 26, 2013, 11:30 a.m. ET
The Value of Phenotypic-Based Drug Discovery
David C. Swinney, Ph.D.

November 12, 2013, 11:30 a.m. ET
Phenotypic Drug Discovery Using Primary Human Cells and Co-Cultures: Lessons Learned
Ellen L. Berg, Ph.D.

December 5, 2013, 11:30 a.m. ET
Modern Phenotypic Drug Discovery is a Viable Pharma Strategy
Jonathan Lee, Ph.D.

Journal of Biomolecular Screening Special Issue on Phenotypic Drug Discovery

JBS 18.10 (December 2013), part one
JBS 19.1 (January 2014), part two
Editor's Note: many manuscripts from the special issue are published ahead-of-print at JBS Online.


SLAS2014 PDD Scientific Sessions

Phenotypic and Target Based Drug Discovery: Point-Counterpoint
Lorenz Mayr, Astra Zeneca, chair
David Swinney, iRND3
Frank Sams-Dobb, Willingsford Ltd.
Mel Reichman, Lankenau Institute for Medical Research
Bernard Munos, InnoThink Center for Research in Biomedical Innovation

Case Studies in Modern Phenotypic Drug Discovery
Vance Lemmon, University of Miami, chair

SLAS2014 PDD Special Interest Group Meeting
Wednesday, January 22
8:00-9:15 a.m.

September 23, 2013