Reinduction of Androgen-Responsiveness in Prostate Cancer: The Role of Caveolin

basic science with clinical application

Synopsis: Hormone-independent human prostate cancer cells usually express increased levels of a protein called caveolin, a major component of caveolae, cave-like structures involved in membrane trafficking and import of molecules into the cell. Decreasing the level of expression of caveolin experimentally with an antisense vector resulted in a return of hormone-responsiveness. It is possible that a therapeutic intervention that was capable of interfering with caveolin expression could restore hormone sensitivity in advanced prostate cancer.

Source: Nasu Y, et al. Nature Med 1998;4:1062-1064.

The progression of prostate cancer coincides with a change in the influence of androgens on tumor growth. Early in the course of disease, prostate cancer is dependent on androgen for survival. Without androgen, prostate cancer cells undergo apoptosis. Clinical treatment approaches take advantage of this hormone dependence. Following primary treatment (surgery or radiation therapy), men with disease spread beyond the prostate gland typically receive either medical or surgical castration and the majority of patients experience some objective benefit from androgen deprivation. Over time, however, the prostate cancer acquires additional genetic defects and becomes more abnormal in its cytologic appearance. And, among the most important clinical markers of disease progression is the development of hormone-independent prostate cancer.

The nature of the changes in the cells that accompany hormone independence is slowly being elucidated. For example, it appears that the antiapoptotic protein bcl-2 is overexpressed primarily in hormone-independent prostate cancer and is virtually undetectable in hormone-dependent prostate cancer,1 and consequently, the hormone-independent cells are relatively refractory to apoptosis. Bcl-2 overexpression also contributes to drug resistance.

An approach to understanding tumor progression (and other differences between cell types) that is beginning to bear unexpected fruit is an experimental procedure called differential display. In this method, messenger RNA from two (or more) cell lines with different phenotypic characteristics is first expanded and then compared by running the nucleic acids on a polyacrylamide gel. If one examines cells that are similar, except for a particularly easily identifiable feature, the two cell lines will have some messages in common and some that are expressed distinctively in one or the other. Bands can be cut out of the gel and sequenced using micro-methods, and, in this fashion, genes can be identified that are "differentially expressed," that is, expressed in one cell but not the other. When such an approach was taken to examine hormone-dependent and hormone-independent prostate cancer cell lines, Yang and colleagues identified caveolin as a gene overexpressed in hormone-independent cells.2

Caveolin is the major protein component of cell structures located near the cell membrane called caveolae or little caves. Caveolae are one of at least three forms of coated vesicles, regions of the cell membrane with characteristic proteins associated with them that face the cytosol. They form a unique endocytic and exocytic compartment at the surface of most cells and are capable of importing molecules and delivering them to specific locations within the cell, exporting molecules to extracellular space, and compartmentalizing a variety of signaling activities. Caveolins form a scaffold onto which many signaling molecules can assemble to generate multi-component signaling complexes. In addition to concentrating these signal transducers within a distinct region of the plasma membrane, caveolin binding may functionally regulate the caveolae-associated signaling molecules. The best evidence is that caveolin most often inhibits signal transduction.

Androgen-independent prostate cancer cells are usually not androgen-independent because they fail to express androgen receptors. Indeed, amounts of androgen receptor are often increased in hormone-unresponsive prostate cancer cells. Given that caveolin expression increases in hormone-independent prostate cancer cells and, paradoxically, the androgen receptor gene undergoes amplification and increased expression in most prostate cancers, it is reasonable to ask whether these two findings are related. Is the caveolin expression interfering with the function of the androgen receptor?

Nasu and colleagues examined the relationship by interfering with caveolin expression in hormone-independent mouse and human prostate cancer cells using an antisense construct stably expressed by infecting tumor cells with adenovirus containing the construct. When caveolin levels decreased in the cells, sensitivity to androgen withdrawal returned. In vivo, cells with lower levels of caveolin had much slower rates of tumor progression than did cells expressing large amounts of caveolin. And, when antisense-containing clones were selected for androgen resistance in vivo, they uniformly were overexpressing caveolin. Conversely, when androgen-sensitive cells were made to express caveolin through adenovirus-mediated infection, the cells became androgen insensitive.

Thus, the correlation appears direct. Prostate cancer cells may undergo a variety of changes in the course of progressing from androgen-sensitive to androgen-insensitive. However, this process appears to depend in a crucial way on the increased expression of caveolin. Many important and interesting experiments are needed to follow up on these interesting findings. Does increased caveolin mean increased numbers of caveolae? Is the androgen receptor associated with caveolae or are the effects of caveolin indirect, or perhaps even caveolae-independent? Does caveolin form a complex with the androgen receptor? Does caveolin bind to other molecules in an androgen signaling cascade?

However, the most clinically relevant application of this important work is the need for therapeutic agents capable of inhibiting the expression of caveolin. It is hoped that blocking the expression of caveolin does not result in some horrific toxicity, a possibility that is of great concern in light of the ubiquitous nature of caveolae and the as yet uncertain dependence of the caveolae on caveolin for their function. Thus, in addition to the developmental therapeutic approaches that aim to overcome bcl-2 expression and to interfere with other genetic lesions that characterize prostate cancer, it is also relevant to consider caveolin antagonists as being a high priority for drug discovery.


1. McDonnell TJ, et al. Cancer Res 1992;52:6940-6944.

2. Yang G, et al. Clin Cancer Res 1998;8:1873-1880.