New Hypothesis on Tamoxifen Resistance Is Under Study

WASHINGTON—One major problem in the adjuvant treatment of breast cancer is that tumors may become resistant to endocrine therapy, even if estrogen/progesterone (ER/PR) receptors remain. Furthermore, some tumor cells appear to be stimulated by the treatment itself. “An antagonist can act as an agonist,” Kathryn B. Horwitz, PhD, said at a basic science symposium sponsored by the National Foundation for Cancer Research.

Recent research indicates that the ratio of co-activators to co-repressors in a tumor can determine whether an ER/PR antagonist will act more like an agonist or more like an inhibitor, said Dr. Horwitz, professor of medicine and endocrinology, University of Colorado School of Medicine.

She hypothesizes that in ER-positive breast cancer, that ratio will determine whether a tumor is inhibited by tamoxifen (Nolvadex) (tamoxifen responsive) or is stimulated by it (tamoxifen resistant).

To screen for such co-activators and co-repressors, Dr. Horwitz and her colleagues used a yeast two-hybrid assay, looking for co-regulatory proteins that interacted with antagonist-occupied

ER/PR receptors. Her team isolated several relevant proteins, including a DNA-binding protein co-activator (L7/SPA) and a co-repressor.

“These were interesting because steroid receptors are not supposed to bind co-repressors,” she said.

PR Receptors and RU-486

Dr. Horwitz looked at transcription from a PR receptor in the presence of a progestin agonist (R50-20), the mixed antiprogestin RU-486, and the pure antiprogestin ZK98. In the absence of hormone, she observed no transcription. But adding an agonist induced a very strong transcription.

This transcription of the progestin agonist was not affected by the co-activator L7/SPA. The mixed antagonist RU-486 had a very weak partial agonist activity in the absence of L7/SPA, but was strongly active in the presence of L7/SPA. The pure antagonist, ZK98, was not affected by L7/SPA.

“We suspect that in tumors that become resistant to hormone treatment, the antagonist activity switches from an inhibitory phenotype to a stimulatory phenotype because the co-regulators in the tumor are changing,” Dr. Horwitz said. “L7/SPA had no effect on agonists, only on antagonists.”

The same appeared to be true with estrogen receptors and tamoxifen. The co-activator L7/SPA had no effect on estradiol, enhanced the transcriptional effect of tamoxifen, and had no effect on a pure antiestrogen. “A co-repressor can bind a steroid receptor,” Dr. Horwitz said “but only when it is occupied by an antagonist, not by an agonist or an unliganded steroid receptor. It binds in the hormone-binding domain.”

Currently, Dr. Horwitz is using a quantitative RT-PCR assay to measure the expression of co-activator to co-repressor mRNA levels in breast cancers. Preliminary evidence hints that the co-activator is more prevalent in ER/PR-positive cells. Whether this is a true relationship remains to be seen. “However, it does suggest the possibility that the co-activator itself is regulated by a steroid hormone.” She is now testing her hypothesis on paired sets of primary tamoxifen sensitive and resistant cells supplied by the San Antonio Tumor Bank.