Researchers have developed a new computer program that accurately predicts genetic changes in bacteria that can make them drug-resistant, offering a potentially powerful tool in the fight against global antibiotic resistance.
“This gives us a window into the future to see what bacteria will do to evade drugs that we design before a drug is deployed,” says study co-author Dr. Bruce Donald, a professor of computer science and biochemistry at Duke.
Drug resistance happens when disease-causing bacteria adapts to antibiotics and becomes less responsive, or completely unresponsive, to treatment. Over the past 40 years, treatments for common illnesses — everything from malaria to pneumonia to gonorrhea — have been increasingly losing their effectiveness. Currently, antibiotic-resistant bacteria cause at least 2 million infections and 23,000 deaths in the United States each year, but some estimates suggest the global death toll could rise to 10 million annually by the year 2050.
There is clearly a desperate need for new drugs to fight these so-called ‘superbugs’. However, drug development is a lengthy process and evolution will eventually drive bacteria to evolve defenses (genetic mutations) to new lines of treatment. Physicians, therefore, need to find methods of lengthening the effective lifespan of new medicines.
In the past, researchers looked at “libraries” of resistance mutations that had occurred to predict resistance and combat it, but this technique was less than efficient. To get around this, researchers at Duke University developed a protein design algorithm, dubbed OSPREY, to identify DNA sequence changes in bacteria that would lead to drug resistance.
In a new study in the journal Proceedings of the National of Sciences, the Duke University describe how they tested OSPREY with the common drug-resistant bacteria called methicillin-resistant Staphylococcus aureus, or MRSA, which kills an estimated 11,000 people in the US every year — more than HIV.
Algorithm correctly predicted genetic changes in MRSA
The researchers programmed the algorithm to identify the genetic changes that MRSA would have to undergo in order to become resistant to a promising new class of experimental drug called propargyl-linked antifolates, which attack a bacterial enzyme called dihydrofolate reductase (DHFR), used for building DNA and other tasks. The drugs – still to be tested in humans – are showing promise as a new treatment for MRSA infections. Identifying the most likely mutations while drugs are still under development, the team believes, means the medicine is better positioned for success when it hits the market.
“If we can somehow predict how bacteria might respond to a particular drug ahead of time, we can change the drug, or plan for the next one, or rule out therapies that are unlikely to remain effective for long,” said study co-author Pablo Gainza-Cirauqui in a statement.
Using OSPREY, the team came up with a ranked list of possible mutations. They picked out four – none of which had been seen before. When the researchers exposed MRSA to the new drugs, they found that the software had successfully predicted the genetic changes that would occur. Specifically, the analysis revealed that more than half of the bacteria that survived carried the mutation they predicted would give the organism the greatest amount of resistance: a tiny change in the bacterial DNA that reduced the effectiveness of the new drugs by 58 percent.
“The fact that we actually found the new predicted mutations in bacteria is very exciting,” says Dr. Donald. The hope now, he says, is that with time and practice the software algorithm will be able to predict genetic changes more than one mutation ahead. “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,” he adds
The team is now enhancing OSPREY to predict resistance mutations to drugs designed to treat E. coli and Enterococcus infections. They believe OSPREY will be useful for predicting drug resistance in cancer, HIV, flu and other diseases where culturing resistant strains is harder than it is with bacteria.
Dr. Donald and colleagues are developing OSPREY in open source format so it is freely available for any researcher to use. The team hopes the approach they are developing will give drug designers a head start in the race against superbugs, which — if predictions are correct — could surpass cancer as a leading cause of death in the coming decades.