Further molecular support to indicate that may of the luminal cells in prostate atrophy are intermediate in phenotype between normal basal and luminal cells, and are perhaps in a transient state, is provided by the finding that they generally show low, albeit variable levels of staining for androgen receptors and prostate-specific antigen (33), which are much more highly expressed in the luminal cells of normal glands. Similarly to keratin 5, many of the atrophic luminal cells stain strongly for Bcl-2 (33,35,55). Interestingly, there is an inverse relation between staining intensity for Bcl-2 and androgen receptors, even within the same acini. Both C-met (25,74) and hepatocyte activator inhibi tor-1 (74) also show strong staining in normal basal cells, weak/negative staining in normal secretory cells, and moderate-to-strong staining in many of the atrophic luminal cells in focal atrophy.
4.3. Proliferation and Differentiation Related Proteins
Several proliferation markers have been examined in human prostate, including immunohis-tochemical analysis with antibodies against Ki-67, proliferating cell nuclear antigen, and topoisomerase Il-a (33). Most studies have used Ki-67. In the normal human prostate, the majority of proliferation occurs in the basal cell layer (75,76), which is considered by most to be the reserve layer of the prostate epithelium. Bonkhoff et al. (75) indicated that the relative distribution of Ki-67 positive cells was approx 70% in the basal layer and 30% in the luminal layer. This does imply, however, that some luminal cells can regenerate themselves. In focal atrophy, it is clear that the increase in the proliferative fraction, as measured by KI-67 staining in the epithelium (11,12,32-36), which ranges from approx 3- to 80-fold higher than the adjacent normal epithelium, occurs mostly in the luminal compartment (25). Proliferation in diffuse/hormonal atrophy has been studied in a few experiments and it seems that it decreases dramatically after a few days, but comes back after approx 1 week to levels observed without androgen blockade (77). This is similar to what has been reported for castrated rodents (see ref. 77). Interestingly, it seems that the proliferation in diffuse/hormonal atrophy also occurs more commonly in the luminal compartment (78).
NKX3.1 encodes a homeodomain protein that is selectively expressed at high levels in normal prostate luminal epithelial cells, is a potential target for deletion on chromosome 8p, and is downregulated in some prostate cancers (79-83). Targeted disruption of this gene in mouse models results in prostatic hyperplasia and prostatic intraepithelial neoplasia (PIN), and Nkx3.1 disruption in combination with Pten deletion or Cdkn1b (encoding p27) deletion results in invasive carcinoma. Recently, we have observed that many focal human atrophy lesions contain a marked downregulation of NKX3.1 protein in the luminal compartment, as compared with normal luminal cells (84).
P27 is highly expressed and localized to the nuclei of normal luminal cells (85) in the prostate, such that upwards of 80 to 85% of these cells are strongly positive (86). The protein is downregulated in many prostate cancers and decreased expression may imply a worse prognosis (85-92). The basal cells, by contrast, are very variable and are much more frequently negative for p27 (86). At times, in p27-stained sections, one can see a third layer of cells that is located between the basal and luminal cells and, when this occurs, the basal-most cells and the luminal-most cells are positive but the cells occupying the intermediate location show an absence of staining (86). This is found most commonly in the periurethral region and in prostates treated with androgen ablation therapy, but can also be observed in benign normal-appearing glands and in focal atrophy glands. A similar, although not identical, pattern of staining has been shown for involucrin, a protein associated with squamous differentiation (93).
Glutathione-S-transferase (GST) P1 is a stress response protein that is constitutively expressed in the human prostate basal cells, but infrequently expressed in normal-appearing luminal cells (48,94-96). By contrast, this protein is upregulated in focal prostate atrophy, in which many, albeit not all, of the luminal cells stain positively (33). GST-a is another stress-induced member of the GST family. It is only very infrequently expressed in the normal prostate, but is upregulated in most focal atrophy lesions (97). Cyclooxygenase 2 is another stress response gene that is upregulated in many of the luminal cells in focal atrophy lesions (98), and it has been shown that the frequency of staining in the luminal cells is higher in areas of inflammation (35).
Was this article helpful?