October 1, 2010 (Vol. 30, No. 17)
Information Generated from Struture-Based Programs Finds Niche in Product Development
Structure dictates function. That is the mantra uttered by many a life scientist. In fact, there is a branch of life science called structural biology that is completely based on this mantra. To the pharmaceutical scientist, this mantra takes on a different meaning: the structure of an investigational new drug can impact its function. And the more planning and design that goes into a drug structure, the better its predicted function.
This is a major reason why so many pharmaceutical scientists use structural biology tools, and why they both attended and presented at the recent CHI conference on structural biology.
Receptos is using structural biology to get an accurate picture of the binding pocket of drug targets. The structural biology of a binding pocket can be used to determine if binding to particular residues can lead to optimal drug-target interactions and where (if any) modifications can be made to a drug to increase its interaction with its target.
Receptos uses standard structural biology tools to work on G-protein-coupled receptors (GPCRs). Because membrane proteins are difficult to crystallize, the company uses lipidic cubic phase to crystallize GPCRs, explained Michael Hanson, Ph.D., associate director of structural biology.
“One of the major problems of generating crystals of any membrane protein is that you need to satisfy the hydrophobic requirement of the protein. Prior to lipidic cubic-phase development, the only way you could do that was to solubilize the protein in a detergent micelle and then you would try to crystallize the protein and the detergent at the same time,” explained Dr. Hanson. The lipidic cubic-phase method involves inserting the GPCR back into a lipid membrane that is capable of supporting crystal growth.
Dr. Hanson went on to demonstrate how data collected from lipidic cubic phase can be applied to small molecule drug discovery. Receptos is currently using this structural biology tool to screen molecules against specific drug targets in a specific GPCR family.
“Structural biology is the approach of choice for these studies because it gives you a wealth of information that you can use to interpret all the other information that you are gaining through the more traditional routes such as cell-based assays and a structure-activity relationships that you are doing as part of the process,” noted Dr. Hanson.
“You can now interpret all that information in the context of the structure, and that tends to make a lot more sense and eliminates some of the dead ends that you would normally go down.” Structure-based drug discovery also reduces the amount of time and money that a company would traditionally encounter in drug development.
According to Dr. Hanson, just two years ago it was not possible to do structure-based drug design on GPCRs because the technology was not available. “We now have a rapid turnaround of structural information that will actually inform medicinal chemistry of a compound’s potential in less than one year.”
Dr. Hanson believes that this work could have a significant impact on the industry.
Incorporating Bioinformatics
Emanuele Perola, Ph.D., research fellow I at Vertex Pharmaceuticals, presented a recently published study based on the binding efficiencies of 60 pairs of commercially available drugs and their originating leads. The goal of the study was to develop some guidelines on how to use binding efficiencies to help improve drug discovery programs.
A number of papers on binding efficiency have been published in recent years, but clear guidelines on the use of this metric had yet to be established. “I wanted to derive some useful lessons that could be applied prospectively in drug discovery programs,” said Dr. Perola, who added that “a critical part of the study consisted of mining the literature for data.”
Dr. Perola had to search a number of databases, examine numerous papers, and explore various other sources to retrieve the origin of each drug. The main data mined from these sources—the chemical structure of the drug and its parent compound as well as their binding affinities—was used in the meta-analysis.
Dr. Perola’s paper was published in the Journal of Medicinal Chemistry earlier this year. Although it is a bit premature to judge the paper’s impact on the pharmaceutical community, Dr. Perola said that he thinks the study will help people at different stages of drug discovery programs—from the selection of viable lead series to the development of effective lead optimization strategies.
Vertex Pharmaceuticals recently evaluated the binding efficiencies of 60 pairs of commercially available drugs and their originating leads. One of the lead/drug pairs analyzed in the study with the corresponding binding efficiencies is shown. The substructure in common between lead and drug is drawn in black. The binding efficiency index used in the study is defined at the bottom.
High-Throughput Screening
Rajarshi Guha, Ph.D., research informatics scientist at NIH Chemical Genomics Center (NCGC), was scheduled to speak at the meeting about his research. NCGC is a high-throughput screening center with expertise in HTS assay development and assay optimization.
Upon receipt of an assay and subsequent optimization, the center screens the Molecular Libraries Small Molecule Repository (MLSMR) of about 400,000 molecules, along with other libraries. “From the primary screening, we get our set of hits and, in some projects, have access to structural information for follow-up structure-activity relationship (SAR) studies,” he explained.
“Our use of structural biology at that point is to identify useful molecular series for secondary screening, and, in some cases, it’s also used to directly elucidate more specific SARs in the lead-optimization process.”
Dr. Guha has focused on linking small molecule ligand SARs with RNAi screening assays to probe novel drug targets. His group is also trying to develop a systems-based approach in which results from RNAi screens are integrated with results from small molecule screens using MLSMR. “We are trying to get more high-resolution information and, for that reason, we are moving to high-content screening on both small molecules and RNAi.”
There is a major challenge in performing such phenotypic matching, though—the mechanism for small molecule hits and RNAi molecule hits differ substantially from one other. “So when a molecule and an siRNA are both identified as active in their respective assays, that doesn’t mean that the small molecule is hitting the same target as that of the RNAi knockdown,” said Dr. Guha.
As a result, Dr. Guha and his team are spending a lot of time performing pilot testing and tweaking their methods of phenotypic matching. “I think there’s a wide scope in this approach, in that, even though high-content imaging screens are time-consuming, the end result is a rich set of data that allows us to go toward a holistic picture of what’s going on at the small molecule level.”
Structure for Stronger Bonding
Avila Therapeutics uses a structure-based platform to discover and then develop covalent inhibitors, a class of drugs that it says has not been well explored by the pharmaceutical industry. Avila’s platform creates molecules that are designed to bind specifically and tightly to their cognate targets and then form a bond with those targets.
A lot of companies create reversible inhibitors that come on and off the target, but Avila’s covalently bonded drugs are different from the pack. “We believe there is a pharmacological advantage in having drugs that actually form a bond and remain on the target—they can achieve better selectivity, they can achieve a longer duration of action, and can lead to fewer side effects,” said Juswinder Singh, Ph.D., co-founder and CSO.
Although covalent inhibitors are not well explored by the pharmaceutical industry, in general, there are a lot of highly successful covalent inhibitors, such as Plavix. “There is a disconnect between the prevalence of covalent drugs and the lack of attention and focus on that area,” says Dr. Singh, who added that there are a few reasons for this lack of attention: prejudice against the actual approach itself, concern about more side effects, and a lack of valid methods to enable discovery of these types of drugs.
According to Dr. Singh, Plavix “was not designed, it was discovered through serendipity. And that is why structural biology becomes an important component because we believe that we can use design algorithms to discover drugs against targets that form specific bonds with them.” Avila discovers these drugs by examining the drug binding sites for residues that are unique to that target, then creating chemistry that will form a bond to that drug target.
At the meeting, Dr. Singh described Avila’s efforts to use its platform to discover hepatitis C virus (HCV) protease inhibitors. Current HCV drugs are susceptible to resistance. “We believe that covalent inhibitors as a mechanism of action can be more active against drug-resistant mutants than reversible inhibitors because, if the covalent inhibitor is on the target long enough to form a bond, that target is inactivated until a new protein is re-synthesized,” he explained.
“We discovered that there was a residue in the HCV protease that did not appear in host proteases that can be acted upon by covalent inhibitors.” This inhibitor is currently in preclinical development.
3-D structure of Avila Therapeutics’ covalent drug complexed with the antiviral drug target HCV protease
