Chemical reaction

Medicinal chemistry is the very heart of the drug discovery process. A medicinal chemist requires the best possible set of leads derived from the hit generation process on which to start work. Thus, the starting point has to be the best possible screening collection that can be sourced in the first place. The definition and makeup of the ‘best’ screening collection has been and will continue to be the subject of controversy.

In some circumstances, it is highly likely that a well chosen designed library of novel compounds may well be an ideal starting point. This is probably true for novel targets with in a well-established family of targets, for example, Kinases. However, for truly novel targets, for which it is impossible to pre-select on the basis of knowledge, the focus has to be a highly diverse collection of ‘hit-like’ compounds, such as the Maybridge HitDiscover or HitFinder collections. The former is a highly diverse set of 58,000 compounds, on which an independent study showed that out of a number of commercial libraries that were produced in-house, the HitDiscover had the most diverse structure. i.e. the most singletons and the highest number of clusters. They have also been reviewed for suitability as potential hits by highly experienced Medicinal chemists, generally obey Lipinski’s ‘rule of five’ and have been shown to express ~87 percent of the ~400,000 pharmacophores in the world drug index.

Given that, going forward, much lead generation will occur in multidisciplinary groups within universities and other parts of the public sector, it is important that access to screening collections that are relatively inexpensive, but retain the required levels of high quality, are made available to all. The NIH has recognized this need and will attempt to address it though the Molecular Libraries initiative within the NIH Roadmap. The Maybridge contribution is through the HitFinder collection consisting of 16,000 compounds, selected to represent the diversity of the full 58,000 set, using a well known clustering algorithm based on standard Daylight Fingerprints and Tanimoto similarity.

Lead optimization will then become the critical step. The art is to make a number of appropriate compounds for test in as short a time period as possible. Too many chemists make one compound, see the results, and then design, make and test the next one – taking a ‘Brownian motion’ approach to the problem. A much better approach is to design and make a discrete set of compounds that will test a hypothesis, test them back to back, and review the complete data set before deciding what to do next. With the pressure from managers to produce a steady flow of results, this can only be done in one way. Automated or semi-automated synthesis based around building blocks that are commercially available. Speed will never be achieved if Medicinal Chemists are spending a significant amount of their time making the intermediates instead of using them.

Maybridge has recognized this need its building block offering and in particular with the innovative Reactive Intermediate range of products. This is a unique and rapidly expanding range of carefully designed sets of minimally substituted building blocks sharing a common ring structure, each functionalized with a selection of the most synthetically useful reactive groups. Minimal substitution means easier interpretation of SAR in lead optimization, while the diversity of functional groups attached to each ring structure allows the chemist maximum flexibility in library design and production.

As well as saving valuable chemist time, the availability of these compounds leads to significant other cost saving, as the chemist can buy and use just the amount they need, saving on the cost of basic starting materials and the disposal of byproducts and excess material.

An example of the merit of this approach is seen in the collaborative work that was undertaken jointly by Maybridge and the recently emerged biotech company ‘Kudos’.

Kudos are experts in targeting specific pathways for use in the treatment of cancer and other human diseases. They screened the Maybridge collection for hits in their Poly(ADP-Ribose) Polymerase-1 (PARP-1) assay. PARP-1 inhibition has wide-ranging therapeutic potential, from enhancing the efficacy of cytotoxic agents during cancer chemotherapy to the reduction of neuronal damage following cerebral ischemia. One of the key hits from that assay was S 15065, which through a relatively short optimization program involving the use of Reactive Intermediates and related building blocks, led directly to the invention of Series 1, members of which showed significant promise in a variety of assays.

Following the optimization of the series using semi-automated synthesis techniques as well as the use of bioisosteres and other traditional medicinal chemistry skills, candidates emerged which are now in Phase 1 clinical trials.

Medicinal chemistry has proven pivotal in driving the understanding of the optimization of the intrinsic activity of drug candidates. However, it has been singularly unsuccessful in leading to a thorough understanding of the other factors involved in pharmaceutical activity, namely the ADMET aspects – absorption, distribution, metabolism and toxicology. We have all worked on projects in which a weak hit had its intrinsic activity optimized to an exquisite degree – but still no biological effect is observed – or worse than that a totally unanticipated toxicological effect is seen. CAD and basic medicinal chemistry enable us to repeat this feat time and time again. However, the understanding of ADMET has simply not kept up.

The balance of understanding has to swing toward understanding effects other than just potency. An enormous amount of work is clearly underway, both practical – the determination of metabolism patterns – and computational, involving neural networks as well as more conventional predictive methodology. These techniques have the potential to significantly enhance the success rate of NCEs - if sufficient understanding is generated is generated and built into these tools before they are widely used ‘in anger’ by non-specialists.

The future of medicinal chemistry

The road going forward is going to become ever more complex and challenging before clarity emerges - if it ever does. The perceived failure of the pharmaceutical companies to produce sufficient numbers of drugs to justify the increases in speeding within the industry is clearly leading the major players down the road towards research that is concentrated on fewer and fewer targets that each have higher potential value – the ‘blockbusters’. However, this trend will be dramatically challenged by a desire within the pharmaceutical consuming public for ‘personalized medicine’, driven by wishful thinking that drugs can be made so specific for a particular phenotype that efficacy can be fully optimized and adverse reactions can be totally eliminated.

One possible positive outcome is that excluding the people who would not benefit from a new drug candidate, or even worse could be at risk from it, will increase the chance that a drug will show itself useful to a particular population group and will thus increase the chance that the same drug will make it into the marketplace. Pre-screening clinical trial subjects should also allow the clinical trials to be smaller, faster, and therefore less expensive; therefore, the consumer could benefit in reduced drug costs.

One conceivable outcome would have rather more negative consequences. It is perfectly possible that the regulators will be forced to continue to request ever larger and larger data sets before giving regulatory approval for transnational sales. And if the drug companies focus their attentions on just the rich Western nations and ignore the rest of the world, two effects are possible. One - healthcare becomes even more rationed that it is now and two – the accessible market size will shrink even further, leading to even fewer NCEs.

These two conflicting effects could eventually force either a step change in the efficiency of the drug discovery process, or a public debate on what is needed from the pharmaceutical industry (a very unlikely scenario!).