Four out of five Americans between the ages of 12 and 24 develop acne on their skin, but scientists still struggle to explain its cause. Although clinicians have long assumed that bacteria play an important role, the latest evidence suggests that key genetic differences among the bacterial strains that live in your pores make the difference between picture-perfect skin and unsightly pimples.

A team of scientists from the University of California, Los Angeles, The Genome Institute at Washington University in Saint Louis and the Los Angeles Biomedical Research Institute has conducted the most exhaustive study of acne-associated bacteria yet. As they report in the Journal of Investigative Dermatology (JID) on February 28, the researchers found that some strains may contribute to skin disease whereas others could help fend it off.

“People have been treating acne as though all bacteria were the ‘bad guys’ in acne,” says U.C.L.A. molecular biologist Huiying Li, the study’s principal investigator. “Our study shows that some strains may be bad, but some strains may be good.”

Chances are that playing host to the wrong type of bacteria is not the only cause of the condition formally known as acne vulgaris. Dermatologists have also invoked the role played by hormones, oil glands and clogged pores. Treatments from soaps to prescription antibiotics often target the bacteria in pores. The JID study offers the strongest evidence yet, however, that bacteria play a key role in both causing as well as preventing at least some kinds of acne.

The research team collected samples of bacteria from within the nasal pores by using cleansing strips on 49 acne patients and 52 individuals with healthy skin. They found that a single type of bacteria—Propionibacterium acnes—predominated in the nasal pores of both clear-skinned and acne-prone individuals. To decipher what made the difference, the scientists had to analyze the strains of P. acnes present in pores.

Using select genetic markers, the researchers identified more than 11,000 strains of P. acnes. Further analysis pinpointed the 10 most common strains, dubbed ribotypes 1 through 10. Six of these strains were strongly associated with unhealthy skin. One strain, ribotype 6, was found only on healthy skin. Because these strains exist as part of a community of bacteria, or microbiome, the researchers next profiled the populations that exist within pores. They identified five common microbial combinations and found that two, dominated by ribotypes 4 and 5, were found primarily in acne patients.

These patterns provided more evidence that different strains of P. acnes might relate to healthy or sickly skin. But to investigate whether certain strains could be harming or benefiting human hosts, the researchers needed to study the genetic structure of each.

The scientists sequenced the DNA of 66 isolated samples of P. acnes from 10 different strains and then combined this information with five other published P. acnes genomes. Comparing these 71 genomes, they constructed a family tree that revealed how this species of bacteria had evolved and the genetic variations that had shaped each strain.

At last, the researchers could see why ribotypes 4 and 5 might cause so much trouble. Certain specimens had acquired DNA from other bacteria that could provoke an immune response in the skin or allow bacteria to stick more aggressively to the host. By contrast, the scientists identified a possible explanation for ‘helpful’ bacteria. Select specimens from other strains, such as ribotypes 2 and 6 appear to possess a defense system that could combat invading DNA from harmful viruses. University of California, San Diego, dermatologist Richard Gallo, who was not involved in the study, observes that this finding hints how treatments that wipe out entire bacterial populations may not be appropriate. “Many of us in the scientific community are showing how most bacteria on the skin benefit us,” Gallo says.

If further study confirms the link between these bacterial communities and acne, the study authors believe treatments for the condition could be further refined. “We could diagnose patients better [based on their microbiome] and predict better treatments,” says study author Noah Craft, a dermatologist of the Los Angeles Biomedical Research Institute. In addition, natural probiotic options—treatments that support the growth of certain microorganisms—could be used to encourage beneficial bacteria whereas more sophisticated antibiotic approaches could selectively target harmful bacteria.

Much remains to be understood, however. Craft adds that it is probable their findings might help explain acne in just one in five cases. “We didn’t specify the subtype of acne in this study,” he says. Craft suspects that more study will reveal how acne vulgaris is actually several different skin diseases.

Microbiologist David Relman of Stanford University, not associated with the study, adds that the complexity of the microbiome could further contribute to the disease’s occurrence. “We have to remember that nearly all microbes spend all of their time as part of a community,” Relman says. He suggests that their “chatter” within that community can influence bacterial behavior—whether beneficial or harmful.