Effects of castration and replacement of androgen on the separation of rat ventral prostate cells by Ficoll gradient centrifugation.

When regressing or growing (hypertrophic) cells from collagenase-digested ventral prostates were centrifuged on isokinetic Ficoll gradients for 6-8 min, they distributed into four fractions. Because of changes in epithelial cell morphology and density following castration to induce regression and replacement of androgens to cause cell growth, and contrary to results with normal rat ventral prostate, stromal cell fraction 2 was contaminated to a greater extent with regressing epithelial cells, as judged by their morphology and binding of radioactive androgens. However, centrifugation for 3 min increased the purity of epithelial cell fraction 4, although the yield of desired cells was reduced. Most cells from endocrine-manipulated rats were viable, as judged by exclusion of trypan blue and the initial incorporation of 3H-uridine. Cells centrifuged on a similar gradient of Percoll separated by a 'sieving' effect, which inverted the order of cellular fractions and removed red blood cells from fraction 2. Metrizamide offered no advantages, compared with Ficoll or Percoll. Neither physiologic nor pharmacologic amounts of testosterone returned the morphology of isolated epithelial cells to normal. To obtain consistent results with prostates from normal or hormone-manipulated rats, one should take care to select an active preparation of collagenase, avoid the use of very old animals, cool the tissue after it is dissociated, and do not apply undigested clumps of cells or overload the gradient. If attention is paid to these details, populations enriched in viable regressing or growing prostate epithelial or stromal cells can be obtained from hormonally manipulated rats.