WASHINGTONOne 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
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
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
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
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.