Expression of basal cell keratins in human prostate cancer metastases and cell lines

GJLH van Leenders, TW Aalders… - The Journal of …, 2001 - Wiley Online Library
GJLH van Leenders, TW Aalders, CA Hulsbergen‐van de Kaa, DJ Ruiter, JA Schalken
The Journal of Pathology: A Journal of the Pathological Society of …, 2001Wiley Online Library
Within normal human prostate epithelium, basal and luminal cells can be discriminated by
their expression of keratins (K). While basal cells express K5/14, luminal cells show
expression of K8/18 and an intermediate cell population can be identified by co‐expression
of K5/18. Prostate cancer is predominantly composed of luminal and neuroendocrine cells,
while a minority of cells have a basal phenotype. In order to distinguish between basal and
intermediate cells, and to assess the effects of androgen deprivation on prostate cancer, 56 …
Abstract
Within normal human prostate epithelium, basal and luminal cells can be discriminated by their expression of keratins (K). While basal cells express K5/14, luminal cells show expression of K8/18 and an intermediate cell population can be identified by co‐expression of K5/18. Prostate cancer is predominantly composed of luminal and neuroendocrine cells, while a minority of cells have a basal phenotype. In order to distinguish between basal and intermediate cells, and to assess the effects of androgen deprivation on prostate cancer, 56 human prostate cancer metastases and three cancer cell lines were characterized using antibodies to K5, K14, K18, and the neuroendocrine marker chromogranin A (ChA). The staining was performed on paraffin tissue and visualized by the avidin–biotin–peroxidase complex method. Protein expression was quantified as the number of positive cells in 20 high power fields (HPF; 400×). Keratin expression in the prostate cancer cell lines LNCaP, DU145, and PC3 was analysed by immunofluorescence with triple staining and confocal laser scanning microscopy. Prostate cancer metastases were consistently positive for K18 and negative for K14, irrespective of hormonal therapy. K5 expression was displayed in 28.9% of the tumours without treatment, in 75% after androgen deprivation, and in 57.1% of hormone‐escaped prostate carcinomas. After androgen deprivation, the number of K5‐expressing cells increased significantly. While androgen‐dependent prostate cancer showed a median of 0 cells/20 HPF (range 0–50), regressed tumours displayed 22.5 (range 0–65) and hormone‐escaped tumours 7.5 (range 0–361) positive cells/20 HPF. Expression of ChA was observed in 47.4% of the androgen‐dependent tumours. The number of neuroendocrine cells was not significantly affected in regressed or hormone‐escaped disease. The androgen‐dependent cell line LNCaP stained for K18, while the androgen‐independent lines DU145 and PC3 both expressed K5 and 18. Expression of K5 in the absence of K14 identifies the existence of an intermediate cell population in prostate carcinoma. Accumulation of intermediate cells in regressed and hormone‐escaped prostate cancer indicates that for their survival, these cells are androgen‐independent. Copyright © 2001 John Wiley & Sons, Ltd.
Wiley Online Library