The level of the protein encoded by the PTEN tumor suppressor gene determines whether prostate tumors are benign, cancerous but slow growing, or aggressive with early metastasis, according to an article in the initial issue of the Public Library of Science, published online on October 27th. One or both copies of the PTEN gene have been lost in 70 percent of prostate cancer patients at time of diagnosis, and scientists had previously believed that one copy would protect against progression to advanced, metastatic cancer.

In the current work, American researchers studied mouse models in which varying levels, or doses, of protein were produced by tumor cells with one copy of the PTEN gene. They found that expression of the PTEN gene is not the determinant of tumor behavior, but that behavior becomes more aggressive in a dose-dependent manner, that is, aggression increases as level of protein decreases. This new understanding of the natural history of the disease could allow researchers to develop novel clinical strategies to diagnose, treat, and possibly prevent prostate cancer.

The article is available online at http://www.plosbiology.org and will appear in the initial print issue on December 23rd.

“We have shown that prostate cancer development is not just affected by mutation and loss of the PTEN gene but that its progression is dose-dependent on the PTEN protein, which we have measured for the first time,” said Pier Paolo Pandolfi, M.D., Ph.D., the study's senior author. “Two men, each with one PTEN gene left, could have totally different disease outcomes depending on the actual dose of PTEN protein coming from that gene.”

Earlier studies by Pier Paolo Pandolfi and Antonio Di Cristofano had demonstrated that loss of the PTEN tumor suppressor gene in mice is responsible for a variety of malignant tumors. In humans, these were shown to include melanoma and cancers of the breast, prostate, and brain. Although the loss of just one PTEN gene is enough to affect cell signaling, the loss has only been associated with slow-growing, mild lesions in the mouse prostate that are comparable with the early stages of human disease. Therefore, many scientists in the field assumed that one copy of the PTEN gene was sufficient to prevent progression to aggressive cancer, in agreement with the classic definition of tumor suppressor genes.

To test this assumption, two sets of mouse models were generated. In one, the PTEN gene was engineered to be removed completely from the prostate only (whole body deletion cannot be studied because it causes a lethal defect in the embryo). In the second model, mice were engineered to have only one half-active copy of the PTEN gene left (roughly 30 percent protein level).

In stark contrast to mice with one gene copy, the mice with no PTEN gene showed aggressive, invasive prostate cancer that developed in just a short period, perhaps suggesting that the major danger in having only one copy of the PTEN gene would be to lose it. However, mice with one half-active gene also developed prostate tumors while those with the fully active copy did not. This refuted the notion that only complete loss of the gene can cause prostate cancer and instead suggests that prostate tumor development correlates closely with the actual PTEN protein level.

“We analyzed the mice at a time when they should have been healthy but instead found massive prostate enlargement and cancer,” explained Lloyd Trotman, Ph.D., first author of the study. “Most importantly, this showed that dropping the PTEN protein dose slightly below the 50 percent level has dramatic consequences for disease progression in just a short period.”

“This study shows the consequences of serial reductions in a critical gene on prostate cancer development and progression,” said Dr. Howard Scher, Chief of the Genitourinary Oncology Service at Memorial Sloan-Kettering, where the work was done. “It shifts the focus from targets of the PTEN gene to the PTEN protein itself. Restoring the function of the gene to stabilize the PTEN level may be clinically beneficial. These findings also show that to understand an individual's prognosis and to optimize the therapeutic approach to an individual patient's tumor, it will be necessary to determine the absolute level of key signaling proteins, and not simply whether the protein is present or absent. Developing these methods is an area of active investigation.”