Urine Bacteria Turns a Steroid Precursor Into Testosterone

For seven days, bacterial colonies grown from men’s urine sat under anaerobic conditions in 96 well plates with steroid molecules to work on. Then the researchers removed the cells, extracted the steroids from the spent culture medium with solvent, evaporated the solvent and added an enzyme system that turned certain steroid changes into a signal a plate reader could see at 340 nm.

The point was not simply to ask which bacteria were present in urine. It was to catch them doing chemistry.

The screen began with urine collected from 27 men before prostate biopsy. It found several steroid-metabolizing bacteria, including three isolates of Actinobaculum massiliense. That species carried an activity the team had not gone looking for by name: when one isolate was supplied with DHEA (dehydroepiandrosterone), it converted the human steroid precursor into testosterone.

One Urinary Isolate Turned DHEA Into Testosterone

The conversion did not happen in a single leap. DHEA fell as androstenedione appeared, then declined, while androstenediol and testosterone accumulated. The bacterium could use either intermediate route. In the culture experiment, testosterone measured 4.14 micromolar after 24 hours and 17.15 micromolar after 72 hours. The result showed that a microbe isolated from the human urinary tract possessed enough enzymatic machinery to perform a piece of steroid chemistry that human tissues also use.

That matters because DHEA is a precursor that human tissues can turn into more potent sex steroids, including testosterone. And testosterone matters especially in prostate cancer: androgens activate the androgen receptor, which supports the growth of many prostate cancers, and treatments for advanced disease often work by lowering androgen production or blocking that signaling. The bacterial result therefore raises an obvious question. Could microbes living near the prostate contribute, even locally, to the androgen environment around it?

The paper can’t answer that question, but it did give the researchers something more concrete to chase. Sequencing that isolate’s genome pointed to two genes, which they named dirA and dirB. When the genes were expressed and tested separately, DirA proved to be the versatile one. It could handle both sides of the route from DHEA toward testosterone. DirB could carry out only part of the chemistry.

A Wider Enzyme Pocket Made More Chemistry Possible

The difference became clearer when the team modeled the enzymes and watched steroid molecules move inside them. DirA’s catalytic cleft was broad and open. A steroid could bind, rotate and present different reactive parts of itself to the catalytic machinery. DirB’s pocket was narrower. It could hold the steroid, but the molecule often ended up flipped or misaligned, with the group that needed to react pointed away from the right residues. Molecular dynamics simulations followed those arrangements for 100 ns and supported the same split seen in the enzyme experiments.

“At this scale, chemistry depends on choreography,” Auburn University physicist Rafael Bernardi, a study co-author, said in the school’s press release. The useful distinction is physical: binding is not enough. A steroid has to face the right direction, at the right distance, with the right part of the molecule exposed. DirA gave it room to do that. DirB often did not.

That mechanism is the strongest part of the study. The researchers did not infer testosterone production from a gene list alone. They first screened living isolates for steroid activity, confirmed products with mass spectrometry, identified candidate genes, tested the encoded enzymes and then used structural modeling and simulations to explain why the two proteins behaved differently.

The Cancer Link Stops at the Laboratory Door

Steroid-metabolizing bacteria were found in 9 of the 27 men, and 7 of those 9 were later diagnosed with prostate cancer. All three A. massiliense isolates came from men who received a cancer diagnosis. But this was a small, clinically selected group already headed for biopsy, not a study designed to compare cancer risk or estimate how common these bacteria are. Those numbers cannot show that the microbes caused, promoted or even accompanied prostate cancer in the wider population.

“We are not saying that these bacteria cause cancer,” Bernardi said. The study never measured whether A. massiliense produces meaningful amounts of testosterone inside a person’s urinary tract, whether any microbial androgen crosses into prostate tissue, or whether it changes androgen receptor signaling there. Those are separate biological steps, and each remains untested.

What the work does establish is narrower and more unusual: a urinary bacterium can take DHEA and, through a defined pair of enzymes, reach testosterone in culture. The next experiment is not another proof that the pathway exists. It is the missing bridge between that pathway and a human organ, where the decisive question is whether enough bacterial hormone ever reaches the prostate to matter.

FAQ