New tool detects drug-resistant bacteria before treatment

When penicillin was discovered, it was hailed as a “miracle drug” because of its unprecedented ability to kill disease-causing bacteria without harming the human body. Since then, numerous other antibiotics have been developed that specifically target a wide range of bacteria. However, the more widely they are used, the greater the risk of antibiotic-resistant strains forming.

In a recently published study in Limits of MicrobiologyResearchers at Osaka University have discovered that bacteria exhibit characteristic shape differences when they are resistant to drug treatment.

Antibiotic resistance is a major public health problem worldwide because it means we have fewer and fewer options for treating bacterial infections. Rapidly identifying antibiotic-resistant bacteria is important to ensure patients receive effective treatment, but the most readily available method for doing this is to grow the bacteria in the laboratory for several days and treat them with drugs to see how they respond.

“There is some evidence that antibiotic resistance also manifests itself in other ways. For example, the morphology of gram-negative rod bacteria changes when they are exposed to antibiotics,” says study leader Miki Ikebe.

We were interested in whether this function could be used to detect antibiotic resistance without actually treating the bacteria with antibiotics.”

Miki Ikebe, Osaka University

The researchers revealed Escherichia coli to fixed concentrations of different antibiotics, causing them to develop antibiotic resistance. They then removed the antibiotic treatment and used machine learning to determine the shapes, sizes, and other physical characteristics of the bacteria from microscope images.

“The results were very clear,” explains Kunihiko Nishino, the lead author. “The antibiotic-resistant strains were fatter or smaller than their parent strains, especially those that were resistant to quinolone and β-lactams.”

Next, the researchers examined the genetic makeup of the antibiotic-resistant bacteria to see if there was a link between bacterial shape and antibiotic resistance. The results showed that genes related to energy metabolism and antibiotic resistance were indeed associated with the shape changes observed in the antibiotic-resistant bacteria.

“Our results show that machine learning can be used to identify drug-resistant bacteria in microscope images, even when no antibiotics are present,” says Ikebe.

Because the bacteria resistant to quinolone, β-lactams, and chloramphenicol all had similar shapes and sizes, it is likely that the same genetic mechanism is responsible for antibiotic resistance in all of these strains. In the future, a machine learning tool could be used to quickly evaluate patient samples and prescribe the right drug to treat their infection.