What could be the explanation for a drug that works at low doses to help children with one type of early puberty - but requires doses ten times higher to treat a similar disease? For years, this puzzle has intrigued pediatric endocrinologists, the doctors who care for children with abnormal hormone systems.

Now scientists at UC San Francisco, working with a new experimental model that turns yeast cells into little versions of human adrenal glands, have learned the answer: they have shown how the drug, medroxyprogesterone acetate (MPA) inhibits one of the key enzymes necessary to produce steroid hormones.

MPA more commonly is known by its brand name, Provera. Often prescribed as a hormone replacement therapy during menopause, it sometimes is used to treat small children who develop breasts, enlarged testicles and other sex characteristics. In high doses it is used to treat women with breast cancer.

The research, published in the June issue of The Journal of Clinical Endocrinology and Metabolism, was conducted by second year Tufts University medical student Tim Lee and UCSF research scientist Richard Auchus, MD, PhD, in the laboratory of Walter L. Miller, MD, UCSF professor of pediatric endocrinology. The study, conducted on a summer research grant from the Society for Pediatric Research (SPR), earned Lee a prestigious Medical Student Research Award at the annual meeting of the Society for Pediatric Research in May.

Lee and Auchus were looking for a way to explain MPA’s dual mechanism of action. Normally, puberty starts in the pre-teen years, when certain cells in the brain begin pulsing out small timed doses of a chemical signal called GnRH, for gonadotropin releasing hormone. GnRH instructs the ovaries and testes to start making steroids, including sex hormones. For children who start to develop sex characteristics very early - as young as toddler age - Provera is one of a number of drugs that can be used to delay the onset of puberty. Scientists already have shown that small amounts of the drug bind to receptors in the brain and inhibit GnRH production.

However some children have a condition called gonadotropin-independent sexual precocity. Though their brains are not yet pulsing out GnRH, their ovaries and testes begin making steroid hormones anyway. They have disorders such as McCune-Albright Syndrome, which causes ovarian cysts in young girls, and testotoxicosis, which causes early masculinization in young boys. Because these children’s bodies are making steroid hormones without a GnRH signal from the brain, low doses of MPA are ineffective. Instead, UCSF pediatric endocrinologists have successfully treated these children with higher doses of the drug.

The success of the treatment proved that larger doses of MPA must interfere with some step of hormone production other than GnRH.

Miller and his lab study the process that the body uses to produce steroid hormones. Sex steroids are produced when the ovaries and testes convert cholesterol into androgens, estrogens and related substances. The process takes several steps, and each step is mediated by a different enzyme, a protein that speeds and guides a biological process.
With Auchus’ guidance, Lee was assigned to investigate whether MPA works at high doses by blocking the action of one of these enzymes.
Studies of rats had shown that large doses of MPA might block the action of cytochrome P450c17, a key enzyme at work in the conversion of cholesterol into hormones. But when Lee and Auchus tested MPA on the human form of P450c17, the enzyme’s action was unaffected. “This emphasizes a point: rats are not always a good model for human biology,” Miller said.

Lee and Auchus next tested MPA on 3HSDII (3-beta-HSD-2), the next step after P450c17 in the cascade of enzymes essential to making human hormones. They found that MPA inhibits 3HSDII by binding to the site that the enzyme uses to help synthesize hormones. They also showed that this action is dose-related - their biochemical data corroborated physicians’ clinical experience that relatively high doses are needed for a therapeutic effect.

Thus at low doses, MPA disrupts the production of GnRH, the chemical signal sent out by the brain to sex glands. At higher doses it is effective in the sex glands themselves.

“This knowledge gives us some new tools in working with these diseases,” Auchus said. “For example, there are other drugs that people use in gonadotropin independent precocity, and we now know that they work by inhibiting different parts of the hormone-producing pathway than MPA does. That may mean that additive therapy would be effective - several drugs working on different pathways, each at lower doses than if it were used alone.”

In addition, Auchus said, now that they have shown that inhibiting this one enzyme can influence sex hormone production, 3HSDII may turn out to be a good target enzyme to enhance treatments of breast cancer.


Lee’s and Auchus’ discovery actually was a small part of a much larger project, undertaken in the Miller lab under Auchus’ initiative. The group has developed a strain of “humanized” yeast that express the DNA of human enzymes. They plan to use yeast cells, converted into little versions of human adrenal glands, as a scientific model to study the action of the human enzymes that are involved in sex steroid hormone production.

Lee worked in the Miller lab on this project during a year-long stint at UCSF, between his graduation from UC Berkeley and matriculation as a Tufts University medical student. The MPA project is one of several to use the new yeast strains. 

With their new model, for the first time the scientists can study the biochemical action of a single enzyme. Also, they can study the actions of drugs on each enzyme needed for human steroid production.

“All human cells handle cholesterol, whereas yeast contains no cholesterol,” Auchus explained. “The value of using yeast cells, instead of human cells, for these studies is that we can put in known amounts of the enzyme, the cholesterol and all the proteins the enzyme has to interact with, to work out the individual biochemical steps in the enzyme’s action.

“With this system we can ask questions about how human adrenal glands and gonads make hormones,” Auchus said. “We don’t know, for example, how a cell in the testes or the ovaries decides to make one steroid as opposed to another. The yeast system will permit us for the first time to work this out with rigor.”

Auchus, Miller and Lee first used “humanized” yeast to study the biochemistry of the key cholesterol-converting enzyme cytochrome P450c17. The group’s paper on the biochemical activity of cytochrome P450c17 was published in February, 1998 in the Journal of Biological Chemistry.