| IDENTIFICATION
OF NOVEL ANDROGEN RECEPTOR INTERACTING PROTEINS INVOLVED
IN RESPONSE TO PROSTATE CANCER THERAPY.
Dr Charlotte Bevan, Professor R.C. Coombes Imperial College, London |
 |
Background and aims:
Prostate cancer is the most commonly diagnosed cancer in Western males and its incidence is increasing. Prostate tumours, like the prostate itself, grow in response to the male hormones androgens, the most abundant of which is testosterone. Androgens exert their effects on growth via the androgen receptor, which is hence a key molecule in prostate cancer research. Androgens bind to and activate this protein resulting in a cascade of events culminating in growth of the prostate and/or tumour cells.
Several of the most successful prostate cancer therapies in current use (“antiandrogens”, e.g. bicalutamide, flutamide) work by blocking the androgen receptor and are very effective for a limited period of time. Unfortunately, eventually they stop working and tumours recur, often in an aggressive form. Currently there is no effective therapy for such advanced stages of the disease. It is therefore vital that we understand the mechanisms by which antiandrogens work in order to avoid or delay this failure.
Although the exact details of how antiandrogen therapies work are not known, it appears that it involves proteins that interact with the androgen receptor and affect its activity, called cofactors. However, it is not clear which cofactors are required in human tumours for antiandrogens to be effective. The aim of this study was to identify these essential interacting proteins and study their expression in human prostate tumours. As well as being potential targets for future therapies, we may be able to use them to predict which therapies would be most effective. This would improve quality of life for prostate cancer patients by reducing unnecessary side-effects and prolonging the disease-free period.
Results
Using human prostate cancer cells grown in the laboratory, we have identified a number of proteins that interact with the androgen receptor in the cell nucleus. It is the nucleus that is most relevant to our work since it is here that the androgen receptor exerts its effects on cell growth, by acting on the DNA. We are especially interested in proteins that show different interaction with the androgen receptor in the presence of activating androgen as compared to inhibitory antiandrogens, since this patern of interaction would indicate a possible role in the action of antiandrogens. An initial “nonbiased” screen, i.e. one that simply highlights changes in protein interactions regardless of the identity of the proteins involved, showed a number of differences between androgen versus antiandrogen treatment (Figure 1). However, it was not technically possible to subsequently identify the proteins involved. Hence we used a different “candidate protein” approach, studying proteins identified in our laboratory previously as having the potential to be involved in these mechanisms. This approach yielded several proteins that showed different interactions with androgen receptor in the presence of androgens and antiandrogens, and also had an effect on androgen signalling. Further, they were found in human prostate cancer cells. Hence these proteins are potential candidates for involvement in antiandrogen action and/or prostate cancer progression, and may also represent potential therapeutic targets for advanced prostate cancer. The most promising candidate protein was p23.
For the androgen receptor to bind androgens, it requires processing by a number of “chaperone” proteins, including one termed p23. However, p23 also appears to be involved in androgen signalling at later stages – within the nucleus, where androgen receptor acts on DNA. In a previous study we had found p23 can increase androgen receptor activity. Now, we confirmed that p23 and androgen receptor interact in the nucleus in protate cancer cells. To confirm whether p23 is a candidate protein for involvement in human prostate cancer progression, we investigated its expression in human prostate cancers compared with normal tissue from the same patients. As seen in Figure 2, p23 is strongly expressed in the nuclei of prostate epithelial cells. These are the cells that also express androgen receptor, and in which over 90% of prostate tumours arise.
Further, when we analysed a large number of human prostate cancers and compared them to normal prostate tissue from the same patients, we saw that p23 expression was increased in the cancer cells in every case. Hence it may be acting to increase androgen receptor activity in cancer and contribute to tumour growth even after antiandrogen therapy. These results have possible clinical relevance in terms of improving prostate cancer therapies, since compounds are currently being developed that can target p23 and our study shows these may be useful for treatment of advanced prostate cancer.
Summary
We have identified a number of nuclear proteins that interact with the androgen receptor and are found in human prostate cancer cells. These have the potential to be involved in antiandrogen action and/or prostate cancer progression. The most promising candidate protein is the small molecular chaperone p23, which was found to be increased in all prostate tumours studied so is a promising potential target for future therapy. Ongoing work is currently being undertaken to establish the importance of this and other proteins to androgen signalling and the pathogenesis of prostate cancer.
Research summary final report dated 20 December 2007
Project 2005/08