A. Photo Credit: Jack Kearse / B. BacMan image: (c)2014 Life Technologies Corporation (www.lifetechnologies.com). Used under permission.
"There are very few targets that actually come all ready – born by nature – with small molecules bound to them for which you can analyze or form druggable sites. So, you need to develop other approaches to go after these challenging targets."
− James A. Wells, University of California, San Francisco
James A. Wells, along with fellow innovators Haian Fu of Emory University and Mary Ellen Digan of Novartis, are doing just that. Their work is featured in the Spring 2014 SLAS Webinar Series, Protein-Protein Interactions as Small Molecule Drug Targets. In three one-hour presentations, they each will share progress made to date and encourage others to investigate this growing area of research.
Wells has a 25-year history of tackling the troubles of protein-protein interactions (PPI) – and some outstanding results. His lab's work in the early to mid-1990s was responsible for discovering hot spots in proteins using alanine scanning and, since then, many others have verified this in other systems.
"Biophysical data we collected led to the discovery of hot spots and the further understanding that hot spots on protein-protein interfaces might be druggable," Wells explains. "The epitopes they bind on proteins are actually centered on a small set of residues."
To advance work in this area, Wells founded Sunesis in 1998 and developed a fragment-based site-directed approach, Tethering, to deliberately target around the interfaces alanine scanning deemed important. An important discovery was made while working with interleukin-2 (IL-2) signaling molecule – they found that compounds bound in "nooks and crannies" that were not evident in the apo structure. Editor's Note: Additional references are here and here.
"The structure of IL-2 without the ligand shows the site to be kind of closed off; the compound finds nooks and crannies that open up in the presence of the compound," Well states. "We suggest that these interfaces are dynamic; that they flex and bend. Since then, a number of other protein-protein interfaces have been attacked by drug discovery methods and they have similarly found that small molecules go to hot spots."
In the first of the three SLAS Webinars, Wells will address another area of his group's work with Tethering technology, allosteric sites and its interesting findings with PDK1.
"It was known that PIFtide binds to PDK1, so we used our Tethering technology to go into that site and introduce cysteines in and around the presumed binding site asking the question, if fragments bound to that site would they actually affect the function of the kinase; would they work allosterically?"
Wells says they expected to find compounds that would activate PDK1 – and they did. Plus, they found even more. Some compounds introduced were small – only 250 molecular weight, as compared to the 1,400 molecular weight of the natural PIFtide.
"We found some smaller compounds that, with some chemical modification, would not only activate PDK1 even more than the natural PIFtide but more importantly they would tether at these allosteric sites and actually act as inhibitors of the enzyme," Wells recalls.
Intrigued by these findings, the group continued to explore. One direction was to compare assay structures of the inhibited enzyme with the activated enzyme. They saw a systematic shift of the C-helix above where the PIFtide sits that shifts the helix down into the active position and up into an inactive position.
"That was cool," Wells notes. "We then developed a fluorescent polarization assay using the PIFtide and using high-throughput screening to identify freestanding compounds that bound to the PIFtide binding site as our next step."
More work in this area uncovered additional information on compound function and structure, including the realization that the PIFtide bind was "kind of like a three-pronged wall plug" with two primary sites that the aromatics fit into and another ancillary site. Wells feels many of their findings also will apply to other kinases.
The final area Wells will address in the first SLAS Webinar deals with activation of capsases, or cell-death enzymes, by small molecules. As "there is evidence to suggest that cancer cells are much more sensitive to capsase activation than normal cells," Wells' group sought – and found – compounds that would activate capsases. They found that these capsase-activating compounds actually form "long fibrous aggregates."
"This was something that had never been seen before," Wells notes. "The capsases actually line up along the fibril much like how Brussels sprouts grow on a thick vine."
Wells continues to study these interesting fibrils, and soon will publish the findings. For now, he points to the lessons of his capsase work.
"Be prepared for the unexpected opportunities when attacking these challenging protein-protein targets," he says. "We didn't expect to find fibrils; we were looking for things that activated capsases and that's what we found. It turns out that biology uses large protein scaffolds to activate capsases naturally."
Fu also believes there is much to be learned from interrogating protein-protein interactions. His most recent work at Emory University Chemical Biology Discovery Center (ECBDC) is focused on emerging opportunities enabled by HTS technology advancements and cancer genomics for protein-protein interaction-targeted discovery. That work will be shared in his SLAS Webinar presentation.
"Protein-protein interactions act as communication centers that integrate, propagate and transmit biological signals in cellular networks; thus, they encode the functional dimension of the genome," Fu states. "The ECBDC has established specialized expertise in targeting PPIs to develop chemical probes for understanding cellular signaling pathways and to identify therapeutic targeting opportunities.
"Proteins network just like human beings," he continues. "What happens in cancer is that communication goes wrong. Dysregulated protein-protein interactions can lead to disease, but at the same time can offer a potential vulnerability for targeting in drug discovery."
Fu's interest in protein-protein interactions started when he was a junior research faculty at Harvard Medical School. "I started working on a family of proteins called 14-3-3 twenty years ago," Fu states. "Through this work, I recognized that many critical regulatory proteins function without enzymatic activity and alter the function of other proteins. 14-3-3 led me to focus my career on this fascinating field of protein-protein interactions."
Fu adds that the interrogation of 14-3-3-mediated protein-protein interactions have "emerged as promising therapeutic targets for diseases such as cancer and neurodegenerative disorders." His Emory team has studied 14-3-3 in depth and one study led to the discovery of a novel class of 14-3-3 small molecule inhibitors, including FOBISIN101, which disrupt 14-3-3/client protein interactions.
Fu indicates more than half of the Center's screening projects have involved protein-protein interactions. Other studies include the translational regulator complex eIF4E/eIF4G for the development of anticancer and antiviral therapies and the redox modulator p47/phox interaction for inflammatory disorder therapeutics.
This once undruggable area has become a driving force, Fu says. "A large number of compounds have already been discovered as potential inhibitors of various PPIs and clinical results with some of these are promising. Inhibitors of MDM2–p53, BCL2, XIAP, and Hsp90 PPIs are in Phase I-III trials. This provides a proof-of-concept from the clinical point of view. Examples of other promising PPI targets with recently identified novel inhibitors include MLL1–WDR5, beta-catenin–TCF, BCLAF1–L3MBTL3, and BRD4–histone H4 interactions."
Fu also will describe new work focused on high-throughput discovery of novel cancer-related PPI targets as part of the National Cancer Institute (NCI) Cancer Target Discovery and Development (CTD2) network. The Emory Chemical Biology Discovery Center, of which Fu is director, is part of the CTD2 Network. Emory leverages the vast amounts of cancer genomics data to delineate the functional effects of genetic alterations by interrogating the protein-protein interaction networks of tumors for therapeutic target discovery.
"By defining PPI networks in tumors, researchers may identify novel therapeutic strategies," Fu states in a recent guest editorial, Decoding the Functional Dimension of the Cancer Genome: Protein-Protein Interaction Networks, for the NCI Office of Cancer Genomics. "Many of the cancer driver mutations revealed by cancer genome projects are found in gene products that are traditionally challenging to target therapeutically, such as tumor suppressors and genes that encode proteins with no enzymatic activity. Targeting the nodes and hubs downstream of inactivated tumor suppressors, for example, could restore tumor suppressive function. Thus, understanding how cancer driver mutations are integrated within growth control signaling networks may present new opportunities for pathway perturbation and novel therapeutic discovery in tumors, even with 'undruggable' targets."
The ability to "leverage genomic and clinical data for high-throughput PPI network mapping of disease-related proteins to reveal promising nodes for therapeutic intervention" has resulted in an expanding list of non-enzymatic targets.
In the webinar, Fu also will describe the rationale for using high-throughput screening to target protein-protein interactions for small molecule discovery as well as some commonly employed technology platforms. He indicates that SLAS is the perfect audience to present his work due to its two major membership components – screening and technology innovators – and he looks forward to sharing through the SLAS Webinar how high-throughput technology-enabled novel protein-protein interaction research offers new opportunities for discovery.
"I want to know how cells work."
That seemingly simple overall sentiment has stemmed a robust career for Digan, who has come at the question from different angles, vantage points and methods. Trained as a Drosophila geneticist in molecular, cellular and developmental biology, she went on to work in the yeast, Pichia pastoris, on its genetics and protein expression before landing at the Novartis Institutes for Biomedical Research (NIBR).
"It is an exciting time not only for the understanding of basic biology but also for high-throughput screening where people are starting to move toward more complex cellular models," she offers. "Our understanding of all the previous years of work has built to where we're now able to work in a more complex world."
At the Center for Proteomic Chemistry (CPC) at NIBR, Digan designs screens and cell-based assays to discover low molecular weight inhibitors of different protein-protein interactions of therapeutic importance. CPC collaborates closely with all of NIBR's departments, applying the suitable combination of lead finding approaches to each drug discovery project. These include high- and medium-throughput screening, structural biology, fragment-based drug discovery, biophysics and various in silico approaches. CPC is the steward of the Novartis Compound Bank, a collection of over one million chemical compounds.
In the SLAS Webinar, Digan will share basic information on various methods to interrogate protein-protein interactions in cells and then hone in on specific examples from Novartis.
"We've been using BRET2 [bioluminescence resonance energy transfer] since the end of 2008," Digan states.
They selected BRET2 following some direct comparisons using Bcl-2 and Bad against a little used protein complementation system for beta-lactamase.
"Through this comparison, we got our hands around the biggest issue with protein complementation assays which is reversibility," she says. "With BRET2, you measure a proximity-based energy transfer. The first member, the donor, is an enzyme named Renilla luciferase, which emits light in a certain wavelength. If the GFP, the acceptor, is close to the Renilla, that closeness should be mediated by the protein-protein interaction and not by the affinity of the enzyme for its other half."
They measured light emissions from both Renilla luciferase and from GFP to get the BRET2 ratio. What they found was that BRET2 had an attractive separation of the two emissions and also the advantage of having no inherent affinity of Renilla for GFP.
"They are completely different organisms," Digan shares. "It was the confluence of the right detectors with an attractive substrate that made it superior to protein complementation for our work. When directly comparing only these two, we chose BRET2. It's been useful, versatile and easily adaptable for many protein-protein interactions in our hands."
Digan will continue discussion of these methods with an in-depth example of assays employing BacMam, which uses viral insect particles to deliver and express genes in mammalian cells with minimum effort and toxicity.
"BacMam vectors don't replicate or integrate in mammalian cells so, in general, they are biosafety level 1," Digan states. "That's really handy because screening labs are full of moving parts and people so if you can lower biosafety, that's good. BacMam vectors also offer the ability to titrate expression and maximize sensitivity of BRET2 assays.
"It is my hope that cell biologists, or anyone interested in pathway biology or protein-protein interactions, will start to think about all of the things you can do with BRET2 and consider BacMam a versatile tool for many cell-based assays in addition to protein-protein interactions," Digan concludes.
The three webinars in the series, Protein-Protein Interactions as Small Molecule Drug Targets, will be held live on the dates and times listed below and will then remain available on demand at SLAS.org. Webinars are free to dues-paid SLAS members. If you currently are not a member, or have yet to renew your membership for 2014, join or renew today!
Targeting Protein-Protein Interactions as an Anticancer Strategy
April 15, 2014 at 11:30 AM EDT (convert time)
Haian Fu, Ph.D.
Professor, Pharmacology, Hematology & Medical Oncology
Director, Emory Chemical Biology Discovery Center
Director, Discovery & Developmental Therapeutics Program of Winship Cancer Institute Emory University
Free to SLAS members
Cellular Protein:Protein Interaction Assays Implemented using BacMam
May 13, 2014 at 11:30 AM EDT (convert time)
Mary Ellen Digan, Ph.D.
Senior Research Investigator I
Center for Proteomic Chemistry
Novartis Institutes for BioMedical Research, Inc.
Free to SLAS members
March 17, 2014