New Biosensor Could Help Identify Antibiotic-Producing Bacteria

Drug development is a notoriously long and difficult process, often taking decades for new antibiotics to move from discovery and development to the shelves at the local pharmacy. But now, a group of researchers believe their new biosensor will speed up this process.

The sensor operates by detecting macrolides, molecules produced by bacteria that have antibiotic or anti-fungal properties. They’re useful molecules, but as soon as doctors start to use them to treat infections, bacteria start to develop resistance. Finding new sources quickly is key to staying ahead of the curve.

“We’re essentially in a constant struggle to produce new macrolide derivatives to help combat bacterial infections,” said Gavin Williams, an associate professor at North Carolina State University, who worked on developing the sensor.

Illustration of the MphR biosensor binding to its target DNA sequence

One of the first steps in producing new antibiotics is to identify the strains of bacteria that produce them.

Researchers used to have to screen each strain of bacteria to reveal if it produced the molecule they wanted, but with this new biosensor, that process could be much faster. “In high-throughput types of experiments, this biosensor would increase the throughput of the screen by hundreds if not thousand-fold,” said Ashton Cropp, an associate professor at Virginia Commonwealth University who worked on developing the biosensor.

The ability to conduct high-throughput screens is essential when working with macrolides. Finding the right strain might mean looking at thousands of cultures, so how successful you are depends on how many bacteria you can look at. “It’s a numbers game,” Christian Kasey, who worked on this sensor as a graduate research assistant, said.

The biosensor is built on bacteria’s natural ability to detect macrolides. Some bacteria produce these antibacterial compounds to eliminate competition, and as a defense some other bacteria have developed resistance to these compounds. The bacteria’s resistance is dependent on their ability to detect when the macrolides are present in their environment.

One of the methods bacteria have developed for detection is the MphR protein. As Cropp puts it, this protein is a natural on/off switch for resistance. Normally the protein is bound to DNA to prevent the expression of resistance, meaning the switch is off.

But when a macrolide comes in contact with a bacteria the MphR protein binds to it, turning the switch on. The protein releases the DNA, allowing the bacteria to express its resistance genes. These genes encode an enzyme that can deactivate the macrolide, neutralizing the threat.

Cropp and Williams co-opted this natural switch to tell them when the macrolide they’re interested in is present. They did this by replacing the resistance genes that get switched on in nature with a gene that encodes for fluorescence.

“Essentially you’re adapting that natural system for a fluorescent detection system by changing the genes that get expressed,” Cropp said. Where before, researchers could only detect macrolides through time consuming expensive analysis, now they can simply look to see if a culture is glowing or not.

The green glow makes detection easy, but the researchers had to make sure that the sensor was specific as well. It wouldn’t do them any good if they weren’t sure which macrolide was being detected.

The MphR protein is naturally paired with erythromycin, but it also binds to many similar macrolides. Meaning the researchers couldn’t be sure exactly what macrolide was activating the switch. They had to modify the protein to identify a specific macrolide.

So the researchers used directed evolution. In this process, researchers control the evolution of bacteria by introducing new genes, and selecting strains with desired traits to reproduce. “It’s really just a process that adapts the natural evolution cycle to work in a laboratory,” Cropp said.

The researchers ended up with a MphR protein that could reliably turn on fluorescent genes in response to a specific macrolide, erythromycin.

The team is planning to expand the scope of the sensor to pair it with different antibiotics. “We believe we can expand that approach to detect essentially any macrolide,” Williams said.

In the future the researchers hope to apply this biosensor to not only detect natural macrolides, but to build entirely new macrolides. “I think that we have a couple of important steps to demonstrate still, but I think that this advance is going to be very important,” Williams said.