A bacterium isolated from the human urinary tract can convert a common steroid precursor into testosterone, according to research that opens a line of inquiry nobody had previously pursued.
The researchers who made the finding were unusually careful about what it does not mean, and that caution should carry into how readers interpret it.
"We are not saying that these bacteria cause cancer," said Rafael Bernardi, associate professor in Auburn University's Department of Physics and a co-author of the study. He added that what is now known is that the bacteria have the molecular machinery to produce testosterone, and that this is worth understanding because prostate cancer is closely connected to androgen signaling.
A Bacterium with Enzymes for a Job Nobody Expected
Urine was long assumed to be sterile, and the urinary tract was left out of the original Human Microbiome Project. Scientists now recognize that it hosts its own community of organisms, sometimes called the urobiome.
Most work in that young field has cataloged which microbes are present and whether those communities differ between healthy people and people with disease. This team asked a different question, which was what the organisms actually do.
Researchers examined bacteria isolated from urine collected from men before prostate biopsy and developed a rapid screening method to find organisms capable of transforming steroids. Among the organisms identified was Actinobaculum massiliense. Supplied with DHEA, a steroid the human body produces naturally and uses as a starting material for more potent hormones, the bacterium generated intermediate steroid molecules and ultimately testosterone.
Searching the bacterium's genome, the team identified two genes responsible, which they named dirA and dirB, short for DHEA isomerase reductase. Laboratory experiments showed DirA was unusually versatile, performing several steroid transformations and allowing more than one route to testosterone. DirB could carry out only part of that chemistry.
The Line the Researchers Drew Themselves
The study demonstrates that the bacterium has the biochemical capacity to produce testosterone under laboratory conditions. It does not show that the pathway is active inside the human body.
It does not show that bacterial testosterone reaches prostate tissue, how much might be produced there, or whether any of it affects prostate biology or the course of disease. Those are the questions that would have to be answered before any of this became clinically meaningful, and none of them have been.
That distinction is worth holding onto, because the combination of testosterone, bacteria, and prostate cancer is exactly the sort of material that gets compressed into misleading claims. Many prostate cancers depend on androgens to grow, which is why treatments for advanced disease often aim to suppress androgen production or block androgen signaling. A microbial contribution to the local hormone environment is a plausible idea, not a demonstrated one.
Jason M. Ridlon, who conceptualized and supervised the study, framed the value in terms of what researchers can now investigate. The discovery of the androgen-producing pathway gives the field specific genes and enzymes to examine in the context of urinary health and disease, he said.
Two Enzymes, One Difference, and a Set of Molecular Simulations
The puzzle the team faced was why two broadly similar enzymes behaved so differently. Auburn's contribution was computational.
Bernardi's group built three-dimensional models of both enzymes and ran simulations tracking what happened when steroid molecules entered each one, treating the proteins as moving structures rather than fixed shapes.
DirA turned out to have a broad, open pocket giving the steroid room to reposition itself, so different parts of the molecule could approach the enzyme's catalytic machinery at different stages. DirB has a much narrower internal tunnel, and the restricted space frequently left the steroid misaligned.
"At this scale, chemistry depends on choreography," Bernardi said. The steroid, he explained, has to be in the right place and facing the right way at the right moment, and one enzyme gives it room to do that while the other does not.
Raissa Rosa, a doctoral candidate in Bernardi's group, performed the computational studies. A static structure can show that a molecule fits inside a protein, she said, while the simulations reveal whether it can reach the precise orientation the reaction needs and whether that arrangement stays stable.
The work was published in Nature Communications and brought together researchers from the University of Illinois Urbana-Champaign, Auburn University, Virginia Commonwealth University and Carle Foundation Hospital.
A Second Field That May Need to Take Notice
Beyond prostate research, the finding raises a question for anyone who measures steroids in urine.
Urinary steroid levels are used in medical research and diagnostics and in testing for performance-enhancing drugs. If microorganisms in the urinary tract can transform those compounds, microbial metabolism could eventually become another factor scientists need to account for when interpreting a urinary steroid profile. That is a hypothesis raised by the work rather than an established problem, but it is a practical one.
For readers, there is no action to take here. No test, supplement or treatment follows from this study, and none should be sought on the basis of it. Men concerned about prostate cancer risk should discuss screening with a clinician based on age, family history and personal risk factors, which is guidance this research does not change in any way. The Centers for Disease Control and Prevention develops communication tools and decision aids intended to support that conversation.
Bernardi described the next step as moving from molecular capability to physiological relevance. Researchers can now search existing urinary microbiome datasets for dirA and dirB, determine how common the genes are, and examine whether their presence tracks with differences in urinary hormones or prostate conditions. That work has not been done yet, and it is what would determine whether this remains a laboratory curiosity or becomes something more.
Key Questions Answered
What did researchers discover? A bacterium from the human urinary tract, Actinobaculum massiliense, can convert the steroid precursor DHEA into testosterone under laboratory conditions using two enzymes the team named DirA and DirB.
Does this mean bacteria cause prostate cancer? No. The researchers stated explicitly that it does not. The study shows the bacterium has the biochemical capacity to make testosterone, not that this happens in the body or affects disease.
Why is testosterone relevant to prostate cancer? Many prostate cancers depend on androgens, including testosterone, to grow, which is why treatments for advanced disease often aim to reduce androgen production or block its effects.
Where did the bacteria come from? They were isolated from urine samples collected from men before prostate biopsy.
Should men change anything based on this? No. There is no test or treatment that follows from this study. Prostate cancer screening decisions should be discussed with a clinician based on age, family history, and personal risk.
Could this affect drug testing? Possibly. Urinary steroids are measured in diagnostics and in testing for performance-enhancing drugs. If microbes transform those compounds, that may need to be considered when interpreting results. This is a hypothesis raised by the work.
What is the next step? Researchers can search existing urinary microbiome datasets for the two genes, determine how common they are, and examine whether their presence relates to urinary hormone levels or prostate conditions.
