Read Time: 9 minutes, 791 words

Choosing a weapon against antibiotic resistance just got easier

A new screening method makes it easier for researchers and clinicians to identify pairs of antibiotics that work better together than alone in curing antibiotic resistant infections.


shadow

They say what doesn’t kill you makes you stronger, and this applies just as well to infectious diseases. Antibiotics have revolutionized medicine and dramatically increased human life expectancy, but we are in the midst of a post-antibiotic era. Many drugs that used to be effective against infections are becoming less so, and researchers and clinicians have to work hard to find new solutions that outpace bacterial adaptation.

There are three ways that researchers combat antibiotic resistance. The first is the most obvious – new drug discovery. With a steady stream of new antibiotics to try, doctors are able to provide options for their patients. When bacteria get used to one, a new one will catch them by surprise. Researchers can also test existing drugs that were designed for other uses to see if they have an impact where traditional drugs have failed. A third way is to find two drugs that work together more effectively than each would alone.

Antibiotics work by targeting a specific aspect of bacterial physiology and weakening it – such as the cell’s ability to reproduce or the molecules that make up the cell’s protective membrane. These aspects of physiology are referred to as “pathways.” Drugs that work together, called “synergistic pairs,” target different pathways in the same bacteria, attacking the organism from multiple sides.

However, finding pairs of drugs that do this is challenging. There are thousands of drugs to test, and to test one pair at a time would easily require a million separate tests. Thankfully, researchers at University of Utah developed a faster method for identifying pairs of drugs that work synergistically and, as a result, found a new drug pairing that may be effective against antibiotic resistant E. coli.

How did they do this? First, in a separate study, a group of researchers developed a large database of growth information for 4,000 E. coli “mutants” grown in over 300 different conditions – different nutrients, different drugs present, etc. A mutant means that a naturally occurring or “wild type” strain of E. coli was engineered to be missing one gene. There is one mutant in the dataset for each of 4,000 different genes. Therefore, each mutant represents a gene — this becomes important later!

So what kind of growth information did they collect? After growing the cultures, each mutant/drug combination was assigned a growth score. A positive growth score meant that growth was higher for the mutant than the wild type E. coli and a negative score meant that growth was lower for the mutant than wild type E. coli. For each drug tested, the collection of positive and negative growth scores from the set of mutants/genes is considered that drug’s “chemical-genetic signature.”

Then, using computers, the researchers at University of Utah searched the dataset for chemicals where the growth scores were significantly different for the mutant than the wild type E. coli. They predict that the missing gene in those mutants must be important for antibiotic resistance.

Next, researchers hypothesized that if an effective pair of antibiotics produces a certain chemical-genetic signature in a certain group of bacterial mutants, then testing different drug pairs in the same group of bacteria will produce the same signature. They were right.

They chose to search for the signature of trimethoprim and sulfamethizole, two antibiotics known to work together. For their test, the researchers selected only those mutants from the original dataset that showed significantly less growth in the presence of trimethoprim or sulfamethizole, and paired each of those drugs with 100 other drugs, creating 200 new pairs all together. Those 200 drug pairs were tested in the mutants. Drug pairs were considered synergistic if they were four times more effective together than separately and produced a chemical-genetic signature similar to the test pair.

After that, they were able to use that chemical-genetic signature for trimethoprim and sulfamethizole to screen even more FDA-approved drugs – 2,000 to be exact – and were able to find more pairs that worked together.

Among the many findings from these experiments, a significant and relevant one is the discovery that the AIDS/HIV drug AZT, when paired with either trimethoprim or sulfamethizole, is effective at treating multidrug-resistant clinical E. coli and Klebsiella pneumoniae. AZT is an antiviral drug, and while it is not known to be an antibiotic on its own, it seems to weaken bacteria and increase their susceptibility to antibiotics.

This finding as well as the methods used are exciting new advances in determining combinations of drugs to use for specific infections. It eliminates the need for testing millions of combinations of drugs one at a time. Also, using mutant bacteria for screening shows the pathways that the antibiotic pair is targeting — giving us clues about how the drug pair might be working. Perhaps there is hope in the post-antibiotic era after all.

Study Information

Original study: High-throughput identification and rational design of synergistic small-molecule pairs for combating and bypassing antibiotic resistance

Study was published on: June 20, 2017

Study author(s):

The study was done at:

The study was funded by:

Raw data availability:

Featured image credit: Flickr: https://c1.staticflickr.com/4/3175/3060243118_f22ac2d56a_b.jpg

This summary was edited by: Nadia Szeinbaum