The computational drug discovery company IDs ‘Lipinski-enhanced’ compounds.
May 18, 2010 | The British computational drug discovery company Domainex has announced it is participating in a new lead optimization collaboration with the Institute of Cancer Research (ICR) in the search for new treatments for breast cancer.
The ICR team, lead by Alan Ashworth of the Breakthrough Breast Cancer Research Centre, is searching for drug-like compounds that target tankyrase and other enzymes in the PARP (Poyl(ADP ribose) polymerase) superfamily, a promising approach to halting the growth of breast cancer cells. Last year, the program received a $6-million research grant under the Wellcome Trust’s Seeding Drug Discovery initiative.
Domainex was established in 2002, spun out of University College London, the ICR, and Birkbeck College in London. It provides computational and medicinal chemistry services, notably a virtual screening technology called LeadBuilder to select compounds suitable for further investigation and medicinal chemistry approaches to develop drug candidates for progression into clinical trials. The company also offers a variety of protein expression services to biotech and pharma—molecular biology resources (Combinatorial Domain Hunting) to clone and express challenging proteins, thereby helping big pharma restart stalled pipelines.
CEO Eddy Littler is willing to admit that the services his company provides pays the bills, while external investment funds the firm’s own pipeline, which focuses on oncology. “We thought, if we’re so damn smart, let’s raise some cash and use it ourselves. The services cover operational costs and investment covers our own pipeline.” Domainex’ first internal target is a kinase called IKK–epsilon, implicated in breast and other cancers (not to mention inflammation and obesity). Another target is the lysine methyltransferases involved in epigenetic regulation.
Littler told Bio•IT World that the ICR collaboration highlights the value of Domainex’ drug discovery offerings to academia. “Because we tend to work with smaller start-ups and small biotechs, we need a cost-effective technology to identify robust hits,” he says.
Ashworth approached Domainex in 2009 having identified tankyrase as a promising potential breast cancer target. Ashworth is perhaps best known for helping to discover and isolate one of the hereditary breast cancer genes, BRCA2, in 1996. More recently, his team reported that PARP inhibitors showed promise in treating hereditary breast and ovarian cancers.
All cell types—even cancer cells—sustain DNA damage in their lifespan and need to fix it in order to keep dividing. PARP is one of the cell’s crucial repair pathways. In a recent interview, Ashworth said that the effect of adding PARP inhibitors to breast cancer cells harboring mutations in BRCA1 or BRCA2 was “astonishing.” The cancer cells were 1,000 times more susceptible to PARP inhibitors than normal cells. Ashworth theorizes that attacking the PARP DNA repair pathway, in addition to the genetically compromised BRCA repair pathway, cripples BRCA cells. But the promise of PARP inhibitors extends far beyond breast cancer. Many other cancers, including ovary, colon, skin and uterine cancer, have other kinds of DNA repair defects that make them candidates for PARP inhibitor treatment.
The Domainex virtual screening tool begins with the universe of small-molecule compounds available for purchase—Littler estimates approximately 15 million. Domainex curates those 15 million compounds by using assessing their chemoinformatics properties, and eliminating compounds that are deemed unlikely to make good drug candidates on the basis of molecular weight, unacceptable log Ps, toxicicity profiles, and so on.
Littler ends up with about 2 million “Lipinski-enhanced” compounds, likely to be acceptable in terms of molecular weight and other physico-chemical properties. In addition, medicinal chemists identify “privilege structures,” asking what types of structures will most likely interact with the target? (The target can be a crystal structure or, as in the ICR case, based on a homology model.)
Domainex then selects the top-ranked compound and recommends the partner purchase those. “These are pragmatic tools that actually work,” says Littler. “It is aimed at small end users, because we can tailor the number of compounds according to the client’s budget.” Academic groups, in particular, may not have the resources to screen tens of thousands of compounds.
Typically, Domainex will recommend 2-5,000 candidates for a client to purchase. They will then test those compounds in their own system and assays, while Domainex assists in analyzing the results and synthesizing analogues.
Based on experience over the past two years, Littler says the approach is producing very high hit rates and is proving very robust. It should appeal especially to clients working with a new target and assay but with only a limited number of compounds on hand, and looking for some credible hits.
After some promising initial results, ICR and Domainex presented their data to the Wellcome Trust, which evidently found the model attractive for further funding. “The Wellcome Trust wants to make a difference,” says Littler. “They’ve done an excellent job.” •
This article also appeared in the May-June 2010 issue of Bio-IT World Magazine. Subscriptions are free for qualifying individuals. Apply today.