By Kevin Davies
January 28, 2010 | Imagine administering a small-molecule drug that could effectively, specifically, and permanently silence its protein target. That should be the ideal of any drug-discovery program, but most small-molecule drugs interact with their target in a reversible, transient and depressingly non-specific manner. One solution to the problem would be to focus on a different class of small molecules—compounds that form a strong, covalent bond with their desired target.
“Because the industry’s focused exclusively on reversible drugs, they’ve ignored the opportunities of covalent drugs and novel ways of solving pharmacological problems,” says Avila Therapeutics co-founder Juswinder Singh. “Three of the most popular drugs on the market are covalent. They were discovered through serendipity. Our platform has been geared toward making this a predictable approach. The opportunity of irreversible inhibitors opens up new ways of looking at chemistry and biology and pharmacology that aren’t really as accessible as small molecule reversible inhibitors.”
Singh’s childhood roots in northeast England are still plainly evident in his thick Geordie accent. He did his Ph.D. in computational drug design with Tom Blundell and Janet Thornton, two of the foremost structural biologists in the U.K. and Simon Campbell, the former head of Pfizer research. That got him thinking about rational drug design. He worked at Parke Davis for three years in the early 1990s on the epidermal growth factor receptor (EGFR) before joining Biogen, which combined the creativity of small pharma with the resources of big pharma. There he worked on protein design, but Biogen’s disappointing results with the drug Tysabri in 2005 prompted a reshuffle and Singh began hatching plans for his own company.
Singh’s ideas took shape in his garage before he sought funding from venture capitalists armed with some novel ideas and “complete naivety.” His key ally was Abingworth’s Roy Lobb, whom Singh knew from his days at Biogen. Lobb quickly saw the opportunity in covalent drugs and became a staunch advocate and co-founder. Lobb’s wife, a South American, contributed the company name, based on the name of a mountain range in Venezuela.
Katrine Bosley, another former Biogen executive, came on board as CEO in early 2009. “It was an opportunity to realize a new drug class, not just an interesting piece of biology or chemistry, but to open up a new area of human drugs that can do things you couldn’t do before,” she says.
Early Insight
It was during his early days at Parke Davis that Singh saw—in the shape of a cysteine residue in the EGFR binding site—the potential for covalently binding drugs. In 1997, Singh and his Parke-Davis colleagues published a proof-of-principle paper in the Journal of Medicinal Chemistry on the structure-based drug design of covalent inhibitors against two tyrosine kinases, members of the EGFR family. As Singh predicted in that paper, “This approach may have application in the design of selective irreversible inhibitors against other members of the kinase family.”
As is well known, early drugs targeting the EGFR such as AstraZeneca’s Iressa (see, “Iressa’s Trials and Tribulations, Bio•IT World, Oct 2003) were only partially successful, benefiting patients with mutated forms of the target protein, but also susceptible to drug resistance. Molecules that bind covalently should, in principle, have greater success. Compounds such as Wyeth’s HKI-272 (neratinib—currently in Phase III) and EKB-569 and Boehringer-Ingelheim’s BIBW2992 target EGFR and HER2, and show considerable promise in treating breast, lung and colon cancers. They target a unique cysteine residue in the kinase domains, and retain their activity against drug-resistant tumors.
Avila identifies targets through structural insights from X-ray crystallography and molecular modeling approaches, coupled with virtual screening to identify small molecules. The ideal drugs have several key attributes. Call it “Avilomics”—high specificity, irreversible binding, and optimal reactivity.
Some protein families contain hundreds of members, with very similar binding sites. Says Singh: “We can exploit residues unique to the target to form a covalent bond, and therefore solve the problem of selectivity.” Covalent bonding introduces a second tier of selectivity. In theory, one can minimize off-target effects because you’re not having to maintain the usual drug levels.
A second advantage is drug resistance. For many drugs, weakened affinity toward mutant forms of the target can result in a loss of efficacy and resistance to therapy. However, because covalent drugs must only engage with the protein once to form a resilient bond, they will not readily disengage from a mutated protein and can retain efficacy.
Another virtue of covalent drugs is the property of uncoupling pharmacokinetics (PK) from pharmacodynamics (PD). Once the drug is irreversibly bound, the protein target is essentially out of action until the cell synthesizes new protein. “A lot of effort in pharma is to get the pharmacokinetics so you’ve got 24-hour coverage of the target. For some targets, we’ve removed that necessity,” says Singh. “We [still] need to get to the target and bind, but our sweet spot of pharmacokinetics is fundamentally different from pharma’s. And that leads to profound opportunities.”
Singh argues that big pharma primarily uses the same toolkit and ends up with a lot of the same answers. He showed data in which the Avila lead compounds provide complete occupancy of their target almost instantaneously. Impressively, it takes 18-24 hours for 50% of the target protein to recover.
Bosley is encouraged by what this means for going into humans. “We’ll be able to administer drug to patients, withdraw blood, and see how long that effect is persisting,” she says. “What are you going to learn in your first dose in your first person ever? If I give you this much, how much of your [target] will be occupied in 2 hours, at 24 hours, etc. So it’s a very different way of thinking about your early drug development.”
Small Molecules
Singh won’t divulge details of Avila’s key chemistry steps, other than to say: “We now realize that each drug target has its own personality. The way to influence that chemistry is using some of these compounds to exploit the microenvironment, which can enable us to come up with highly specific modifications… That’s an area that hasn’t been explored.”
Bosley adds: “If you’re harnessing the local environment, you can use bond-forming elements that are much lower reactivity. We want to use a low reactivity entity, something positioned and designed so that it can only respond when it’s in the right place.” She even lets slip the phrase “intelligently designing” small molecules, before thinking better of the expression.
Singh displays assay results showing the effects of an Avila compound compared to an approved drug against three classes of drug target—a protease, kinase, and a lipid kinase. The contrast in inhibition is much more persistent than the current standards. For some targets, it’s the duration of effect that’s exciting, for others it’s more the selectivity.
“Some of the approaches we’re using are cutting edge—nobody’s thought about how do you exploit reversible interactions then try to engineer specificity using algorithms for computational chemistry.” Singh doesn’t do high-throughput screening, but a lot of rational engineering by looking at structure-based drug design, protein binding, and ways to exploit reversible interactions, specificity. He has a group of seven crack chemists to synthesize compounds in house.
Avila’s first two programs focus on the Hepatitis C protease and Bruton’s tyrosine kinase (BTK)—block activation of B cells. (Natural human mutation, XLA). Both drugs have undergone preclinical testing and Bosley expects to start Phase I clinical trials in 2010. The importance of BTK is that it is an intractable target—it has been very difficult to develop drugs with the requisite specificity. The selective covalent approach should provide that. In the case of HCV protease, while companies such as Vertex have some promising drugs, Bosley says there is an unmet need to combine current inhibitors with drugs that have different mechanisms of action, just like HIV cocktails. “We need to hit from multiple mechanistic angles,” she says. The covalent approach shows promise in combining efficacy, tolerability, and overcoming resistant mutants. After all, says Bosley, “Once you’re on, you’re on.”
Singh says that Avila is now able to target multiple amino acids in its target sites. “We’ve gone beyond cysteines and we’ve gone beyond kinases.” The range of potential targets from the “druggable genome” using this approach is sizeable, and could include membrane proteins such as G-protein coupled receptors and ion channels. One question is how far can they push it? Bosley says several criteria will be considered: What is the biological or clinical problem? Why is this worth spending time and energy on? How does it help us validate the technology? For example, is the drug attacking a non-cysteine residue or another kind of target?
Will Avila’s approach expedite the drug development process itself? Possibly. “Early in clinical development, we should have a much clearer understanding of, are we achieving the pharmacological result we want?” says Bosley. For any target, then there’s also the question of efficacy. In the case of BTK, there is little experience in humans to match the preclinical profile. If the lead drug is hitting BTK, then the question becomes, does the drug have the intended clinical impact? For HCV, the biology is better understood, so the question is more about hitting the expected mutants and tracking the pharmacokinetic and safety profiles predicted by the preclinical data.
Avila’s successful series B round last July takes it through the initial clinical trials. Phase II would need extra funds. Bosley says she is open to partnering with big pharma, whether on specific programs or discovery alliances. “It’s an important way to realize the breadth of covalent drugs. We couldn’t do it all, even if we were ten times our current size,” she says.
This article also appeared in the January-February 2010 issue of Bio-IT World Magazine.
Subscriptions are free for qualifying individuals. Apply today.