Scientists seek a way to predict antibiotic resistance

Tim van Opijnen has an unusual library.

Instead of books, it holds over 10,000 mutant strains of Streptococcus pneumoniae, each with a gene disabled — though a different one than its neighbor. By knocking out a single gene, van Opijnen’s lab at Boston College is trying to understand individual genes’ importance and function in the presence of an antibiotic.

His choice of bacterium is intentional: there are 1.2 million drug-resistant pneumococcal infections per year in the US, joining several other species of bacteria that are growing in immunity to antibiotic treatment.

Countering that threat requires a number of different approaches. Scientists are trying to develop new antibiotics, new ways to make old antibiotics effective, and new procedures that cut down on antibiotic overuse.

More From Stat News

But another initiative that’s gotten less attention is predicting how microbes will become resistant in the first place. Scientists say they know what the end results of resistance look like. But the path to resistance is blurry, akin to seeing a starting and ending destination on a map, but none of the roads in between.

“There are major mechanisms that we know will confer a very high level of resistance against a specific antibiotic,” van Opijnen said. “But what is not well understood is the process. How do you get there?”

Van Opijnen along with colleagues from Tufts University School of Medicine, St. Jude Children’s Research Hospital, and the Pittsburgh School of Medicine recently received a $10 million grant from the NIH to help answer that question. They’re looking for the specific genetic changes, or mutations, that enable any given type of bacteria to become a superbug.

Finding those “warning signs” could, theoretically, pave the way for a surveillance system for bugs on their way to resistance. Hospitals already sample their environment for resistant pathogens; expanding that practice to track the rise and fall of resistance genes globally could help guide which antibiotics doctors prescribe when.

Those are the motivations behind a growing movement of scientists using genomics to understand resistance and, hopefully, predict it before it arrives in our hospitals.

Multiple roads to the same place

At the Pittsburgh School of Medicine, van Opijnen’s collaborator, evolutionary biologist Vaughn Cooper, is watching bacteria evolve in action.

Using a cloned strain of Streptococcus pneumoniae, Cooper doses the bacteria with increasing amounts of an antibiotic and takes a genetic snapshot at each stage. Putting the sequences side by side shows the evolutionary steps, in detail, that the bacteria are taking.

What Cooper is seeing, so far, is promising — a set of fewer than 10 genes that are mutating to create resistance. In other words, “there aren’t that many ways for the bacterium to immediately solve the problem” of antibiotics, Cooper said. The key word, though, being ‘immediately.’ Cooper’s team is only looking at the initial mutations in response to the antibiotic. “The challenge is whether the next layer of genes that are affected is more diverse,” he said.

The team is still in the early phases of its experiments. Next up will be studies in mice, which are much more realistic since they have a working immune system.

But a cautionary note exists in van Opijnen’s own findings. Recently, using his mutant libraries, he exposed two strains of Streptococcus pneumoniae to the antibiotic daptomycin. What he found surprised him: although the two strains were of the same species, over half of the genes used by one strain to grow in the presence of the antibiotic could be thrown away in the other strain. They were using different genes to achieve the same result — survival and growth.

“Ideally that is not what you want,” van Opijnen said.

The results, published in PLOS Pathogens in September, provided clues as to which genes might mutate to enable resistance, but they also complicated the roadmap. van Opijnen’s results imply that adaptation is partly strain-dependent, and the number of pathways could be more numerous than initially thought.

“This could be one of the reasons why it remains so difficult to predict the emergence of antibiotic resistance,” he said.

Even so, their five-year grant will continue to try to do just that, initially focusing on Streptococcus pneumoniae and Acinetobacter baumanii, and gradually expanding to other superbugs.

A strategy against resistance

Van Opijnen’s collaboration isn’t the only team working toward a genetic warning system for resistant bacteria. Julian Parkhill, the group leader at the Parkhill Group at the Sanger Institute, and his team are conducting genome-wide association studies (GWAS) on large populations of resistant bacteria to try and associate specific mutations with resistance. GWAS allows for the examination of all genetic changes, no matter how many there are.

“There are numerous different pathways to resistance,” Parkhill said. “And hopefully by taking this kind of top-down approach, then you don’t pre-judge that. You just find all the genetic changes that are associated and some of them will be known and some of them will be novel, which leads you in different directions as to what’s causing resistance.”

This kind of knowledge could have impacts at the level of the individual patient and societally as well. At the patient level, van Opijnen envisions a future where he can sequence a bacterium causing someone’s infection and see if it’s already partway down the path of resistance. If it is, doctors could use an alternative antibiotic. That would mean less time wasted on treatments that don’t work and less chance that the infection becomes life-threatening.

“The rate of change of genomics is stunning,” Parkhill said. “And I don’t think it’s unreasonable to think that there will be point of care sequencing in a short time frame. What that enables us to do is not just ask what antibiotics is this organism resistant to and therefore we shouldn’t use, but also conversely what antibiotics is this organism sensitive to and therefore, which ones we can use.”

At a societal level, once we know the genetic steps to resistance, we could conceivably begin a surveillance system for pre-superbugs. Hospitals routinely sample their facilities for resistant bacteria. Collating those genetic data from across the country or around the world could help scientists catch trends in emerging resistance; with that knowledge, medical officials could temporarily restrict certain antibiotics until the trend reverses.

In many ways, these genomic experiments are bringing us back to basics of bacteria physiology — which we had taken for granted that we had a handle on until antibiotics stopped working.

And with the fast pace of evolution, it will be a strategy of combatting resistance that will have to be ongoing.

“There will always be new mechanisms arising,” Parkill said. “We’re not going to do all of our experiments now and say, ‘That is a finished data set, and we’ll use that to predict for the rest of time.’ What we’ll say is, ‘This is our best prediction at the moment.’ Obviously we have to maintain vigilance and constantly test new examples of strains to look for new resistance mechanisms, which we know will arise.”