Patricia F. Fitzpatrick Dimond Ph.D. Technical Editor of Clinical OMICs President of BioInsight Communications
As standalone technology, method has not delivered on promise.
In about the year 2000, many believed that combinatorial chemistry combined with high-throughput screening would create scores of new drug candidates and fill drug company pipelines. And combichem did solve a lot of problems presented by traditional synthetic organic chemical methods. The technology, combined with high-throughput screening, enabled less complex and time-consuming syntheses, generation of compound diversity to enable new lead discovery, higher compound output, more rapid development of structure-activity profiles, and lowered costs.
Pharma companies adopted the technologies, envisioning revitalized and rapidly filling pipelines, and combinatorial chemistry companies sprung up, funded by enthusiastic investors.
But when few drugs emerged from massive numbers of small molecule libraries, integrated technologies revealed the multiple subtleties involved in drug activity and, more notably, the need for greater molecular specificity.
ArQule, a pioneering combinatorial chemistry company, now approaches the design of novel kinase inhibitors with its AKIP™ platform. AKIP integrates in silico drug design, parallel robotic chemistry, and assay development. The company focuses its expertise on the creation of non-ATP competitive inhibitors that interact with these kinases in novel binding modes. Inhibitors identified through AKIP are optimized into drug candidates having the appropriate efficacy and selectivity for their target with minimal side effects.
“Since 2007, we have focused ArQule’s efforts in drug discovery to leverage our learnings in the parallel chemistry business together with our structural analysis of inactive kinases,” ArQule’s vp of chemistry Mark Ashwell, Ph.D., explained to GEN.
ArQule’s lead drug candidate is tivantinib, a c-MET kinase inhibitor currently in clinical trials for non-small-cell lung cancer (NSCLC). The company expects that it will enter Phase III trials to test its efficacy in liver cancer in late 2012/early 2013.
“The story of tivantinib’s discovery is unusual,” Dr. Ashwell said. “We found it while screening a small molecule kinase library in a cellular assay, looking for a cell signature that we felt would lead to an anticancer agent.
“We found the profile we were looking for, and discovered that the activity came from only one of a pair of enantiomers, leading us to conclude that the active small molecule had a highly specific target with a strict molecular recognition requirement.”
Dr. Ashwell noted that he and his colleagues screened the compound against a group of kinases and found it was a c-MET inhibitor.
Since then, he said, “We have been able to look at the activation states of kinases in very specific ways, and we found tivantinib prevents cMET from assuming an activated form. Now we are building an understanding of how other kinases become switched on and off from inactive to activated states and how inhibitors can be designed to interact with different conformations.
“To detect these types of inhibitors, we expanded our screening platform to include direct binding assays of the molecule with the kinase in its inactive form using thermal shift assays and size-exclusion mass spectroscopy. We have also included additional biochemical assays mimicking what is known about how cells activate kinases.”
According to Astex Pharmaceuticals, traditional high-throughput screening “has not delivered on its promise of increasing the numbers and quality of new drugs entering clinical trials.” One reason for this relates to the complexity and the relatively large size of the compounds routinely being screened.
Astex’ Pyramid platform, based on fragment-based drug discovery (FBDD) reportedly rapidly delivers customized, high-quality drug leads with enhanced therapeutic potential.
GEN asked David Rees, Ph.D., svp of medicinal chemistry, how combichem and screening technologies have evolved to produce viable drug candidates faster and more efficiently. Dr. Rees explained that, prior to his 10 year tenure at Astex, he worked in large pharma and was involved in “pushing combichem in those organizations.” While he believes that the technology successfully increased the number of compounds that chemists made, “Many people would question the quality that came out of it.”
“With combichem,” he said, “you make compounds of higher molecular weight and also with higher lipophilicity, a greater tendency to dissolve in fatty environments as opposed to aqueous environments. These compounds tend not to make good drugs.”
During the last decade, he explained, chemists have realized that there is a certain drug-like profile for molecules, described as “The rule of five.” “The guy who came up with the rule was Chris Lipinski, and the rules delineate drug-like properties including molecular weight, hydrogen bond donors, and lipophilicity.
“With combichem we were able to make loads of compounds but often going outside of Lipinski’s guidelines.” With the benefit of hindsight, Dr. Rees said, “It wasn’t such a smart idea.” Described in 1997, Lipinski’s rule has provided a proven “rule of thumb” to evaluate whether a chemical compound with a certain pharmacological or biological activity has properties that would make it a likely orally active drug in humans.
Astex’ FBDD approach involves “synthesizing as few compounds as possible—the opposite of combichem, but we put a lot of time and energy into designing each compound so it has the best chance of binding to the intended drug targets.”
Astex starts off screening very low molecular weight compounds that range from 150 to 200 daltons. “In high-throughput screening one would screen between a million to 10 million relatively high molecular weight compounds looking for a hit. In the fragment-based approach we screen only a few hundred compounds. Because these compounds are so much smaller they sample the possible binding interactions that you could get more comprehensively. HTS provides only a smaller slice of the total possible compounds. Statistically you have a higher chance of getting a hit screening a thousand fragments than a million HTS.”
Dr. Rees said that Astex also uses a lot of x-ray crystallography to “give us the three dimensional structure of our drug-protein targets—then we know the exact chemical structure of the binding site we want to block, in other words, we use structure-based drug design.
“The Lipinski-like properties of the candidates from Astex’ fragment approach are much more attractive than candidates arising from other methods, he says. If you look at most pharma companies these days you would find more interest in FBDD than in combichem.”
Astex currently has several drug candidates developed through its platform technology in clinical trials in the developmental hands of its big pharma partners and a number of internal programs.
In its 2011 “Year of Chemistry” report “The Changing Role of Chemistry in Drug Discovery,” Thomson Reuters said, “Already the job description for many chemists has changed from the pure synthetic chemist to include more computational chemistry.” And combinatorial chemistry as a technique for the rapid synthesis of drug-like compounds, integrated with novel technologies, will continue to impact the discovery of novel drugs.