Feature: Making drug resistance futile

Read Part I, Shortening the odds in drug design

At this year’s Lorne Conference on Protein Structure and Function, Associate Professor Martin Scanlon discussed his group’s use of fragment- based drug design to search for novel antimicrobial compounds and a way to combat the increasing problem of bacterial resistance.

Once the fragments that bind efficiently at a biophysical level are identified – by mass spectrometry in Scanlon’s case – the job to optimise potency and reach the desired biological activity for a drug-like molecule begins. According to Scanlon, the route taken to do this is what structure-based drug design is all about.

“We first rank the ‘positive’ fragments based on their binding efficiencies, and then enter the best ones into X-ray crystallography screens. The resultant structural data then drives a process of rational drug design to elaborate the different fragments and improve their potencies.

The classical way to do this is by identifying two fragments that bind efficiently, and at two adjacent sites on the protein you’re interested in. Those two fragments are then ‘evolved’ chemically to become a single, more potent molecule, such that each original ‘fragment’ can still bind optimally to its specific pocket.

In this way, you effectively multiply the component affinities due to the relationship between energy of binding and the binding potency.

Thus, two fragments with potencies of around 100µM and 1mM, respectively – which would not even be detected by HTS – become a new molecule with around 100nM binding affinity, and that is potentially the making of a lead series of therapeutic compounds.”

A current theme within both Scanlon’s group and the MIPS is developing novel antimicrobial agents or novel strategies for the treatment of microbial infection to get around some of the problems of antimicrobial resistance.

“As one strategy to do this, we are screening potential targets in bacterial cells that regulate virulence. We know, for instance, that humans have all sorts of bacteria living on and in them without causing any problem, but also that, in response to some trigger, that bacterium can become virulent.

---PB---

“Our idea is to turn these pathogenic bacteria back into commensals, so rather than trying to kill it or stop it growing, we just want to turn off the bacterial drive and/or ability to infect. We know that, to some extent, the development of resistance is inevitable, but by regulating virulence, we might slow down the rate at which resistance will develop. No such agents used in this way exist as yet.”

The target used in Scanlon’s fragment- based screening approach was carefully chosen to cause the biggest impact on virulence. Virulence mechanisms often act simultaneously, and shutting down only one or two at a time might not be sufficient to block infectivity.

Many virulence factors are proteins that must get outside of a bacterial cell to work; they are either presented on the outer membrane or are secreted out of the bacteria via the cell’s trafficking machinery.

Either route of exit requires correct protein folding which often relies on a specific oxidoreductase enzyme to make sure the secreted protein has all its disulphide bonds present and correct. In other words, this enzyme is a key regulator of multiple bacterial mechanisms, and a perfect target.

Gene deletion studies by various groups confirmed that bacteria lacking this enzyme still grow happily, but generally have no capacity for virulence.

Simply put, inhibiting this disulphide bond-forming oxidoreductase in bacteria inhibits oxidative protein, with the aim of simultaneously disrupting multiple different virulence pathways and hopefully enabling the bacterium to regenerate a commensal phenotype.

“So far our screens have identified quite a few compounds that bind, and we have some structural data for small molecules bound to our target enzyme,” Scanlon says. “We have already improved the potency of some of these promising compounds and, with these, are reproducing some of the commensal phenotypes present in the knock-out strain.”

The group is currently using cell- based assays to understand more about the potency and optimal dosages of the small molecules in vitro before taking the project into the pointy end of animal models, as a first step on the road to new drug molecules.