Mouse Prostate Cancer Cell Lines Established from Primary and Postcastration Recurrent Tumors

The clinical course of prostate cancer is grouped into two broad phases. The first phase, which is the growth of the androgen-dependent cancer (AD-Ca), responds well to androgen depletion treatment while the second phase that could be termed as androgen-depletion-independent cancer (ADI-Ca) does not. We used two separate prostate tumors, one AD-Ca and one ADI-Ca from the conditional Pten deletion mouse model, to generate from each a pair of cell lines. The AD-Ca cell lines (E2 and E4) and the ADI-Ca cell lines (cE1 and cE2) display biallelic deletion at the Pten gene locus, an event which is specific for the prostate epithelium for this mouse model and a fairly similar level of expression of the androgen receptor (AR). The ADI-Ca cell lines (cE series) grow well in the absence of androgen, display increased AR transcription under androgen-deprived environment, and retain the sensitivity to increased proliferation when androgen is supplemented. The AD-Ca cell lines (E series) grow slowly in the absence of androgen and, unlike cE cells, do not show increased AR expression when maintained in the absence of androgen. The detection of epithelial cell markers, such as CK8, CK14, CK18, and E-cadherin in the cE series is conforming with the polygonal epithelial morphology of these cells in culture. The E cells also present mostly polygonal-shaped morphology with a small percent of cells with fibroblastoid morphology and produce little or very low levels of cytokeratins but increased levels of vimentin, Twist, and Slug, the markers known to be associated with epithelial–mesenchymal transition. Each of the cell lines, when inoculated subcutaneously into male or female NOD.SCID mice induced tumors within 8 weeks with 100% incidence. Histopathological examinations of the tumor sections, however, led to noticeable biological differences. The cE series engenders adenocarcinomas, particularly in male hosts, and the E series induces sarcomatoid carcinomas (positively stained for CK8 and AR as well as vimentin expression) in either male or female hosts. These new cell lines are promising models for the elucidation of the androgen metabolism and their role in prostate cancer.