Attempting to emulate the ability of chess grandmasters to think multiple moves ahead, scientists are learning to anticipate the most likely move–countermove scenarios that give rise to drug resistance. For example, scientists at Duke University and the University of Connecticut have developed a computational approach that can predict resistance mechanisms, including previously unidentified mutations—that pathogens might use to defeat new drugs. The scientists point out that their approach can be used long before new drugs are tested on patients.
The scientists used their approach to evaluate a potential treatment for methicillin-resistant Staphylococcus aureus (MRSA) infections. Encouraged by the accuracy of their resistance forecasts in this work with Staphylococcus aureus, the scientists plan to broaden their efforts by evaluating drug resistance mutations that may occur in Escherichia coli and Enterococcus. Looking even farther ahead, the scientists say that their approach could help forecast drug resistance mutations in diseases such as cancer, HIV, and influenza. In these diseases, the scientist note, raising resistant cells or strains in the lab is more difficult to do than with bacteria.
The scientists presented their results December 31 in the Proceedings of the National Academy of Sciences, in an article entitled, “Protein design algorithms predict viable resistance to an experimental antifolate.”
“In this study, a structure-based protein design algorithm (K* in the OSPREY suite) was used to prospectively identify single-nucleotide polymorphisms that confer resistance to an experimental inhibitor effective against dihydrofolate reductase (DHFR) from Staphylococcus aureus,” wrote the authors. “Four of the top-ranked mutations in DHFR were found to be catalytically competent and resistant to the inhibitor. Selection of resistant bacteria in vitro revealed that two of the predicted mutations arise in the background of a compensatory mutation.”
Essentially, the researchers used their algorithm to identify DNA changes in bacteria that would alter a drug’s protein target, an enzyme the bacteria need to build DNA. The researchers were particularly interested in DNA changes that would disguise the target from an experimental drug while allowing it to retain its biological function. Such changes would be expected to enhance antibiotic resistance.
“We wanted to find out what countermoves the bacteria are likely to employ against these novel compounds. Will they be the same old mutations we've seen before, or might the bacteria do new things instead?” said study co-author Bruce Donald, a professor of computer science and biochemistry at Duke.
From a ranked list of possible mutations, the researchers zeroed in on four single nucleotide polymorphisms, or SNPs. Though none of the mutations they identified had been reported previously, experiments with live bacteria in the lab showed that the researchers’ predictions were right.
When the scientists treated MRSA with the new drugs and sequenced the bacteria that survived, more than half of the surviving colonies carried the predicted mutation that conferred the greatest resistance—a tiny change that reduced the drug’s effectiveness by 58-fold.
The accurate forecast encouraged the researchers that their approach could be used to enable preemptive strategies in drug design, giving drug developers a head start on the next line of compounds that could remain effective despite resistance mutations. The model could also be expanded to anticipate a microbe's response more than one move ahead, Donald said: “We might even be able to coax a pathogen into developing mutations that enable it to evade one drug, but that then make it particularly susceptible to a second drug, like a one-two punch.”