Heterogeneous subpopulations of human prostatic adenocarcinoma cells: potential usefulness of P21 protein as a predictor for bone metastasis

Molecular markers of prostate cancer outcome

Molecular markers have the potential to serve not only as prognostic factors but may be targets for new therapeutic strategies and predictors of response in a range of cancers. Prostate cancer development and progression is predicated on a series of genetic and epigenetic events within the prostate cell and its milieu. Within this review, we identify candidate molecules involved in diverse processes such as cell proliferation, death and apoptosis, signal transduction, androgen receptor (AR) signalling, cellular adhesion and angiogenesis that are linked to outcome in prostate cancer. Current markers with potential prognostic value include p53, Bcl-2, p16INK4A, p27Kip1, c-Myc, AR, E-cadherin and vascular endothelial growth factor. Evolving technology permits the identification of an increasing number of molecular markers with prognosis and predictive potential. We also review the use of gene microarray analysis in gene discovery as a means of identifying and cosegregating novel markers of prostate cancer outcome. By integrating selected markers into prospective clinical trials, there is potential for us to provide specific targeted therapy tailored for an increasing number of patients.

Molecular characterization of human prostate carcinoma cell lines

BACKGROUND

This study presents a comprehensive survey and characterization of available prostate carcinoma cell lines, most of which have been widely used but are incompletely characterized.

METHODS

A total of 21 cell lines were investigated, including three "classical" (DU 145, LNCaP, and PC-3) and 18 "non-classical" lines (1013L, 22Rv1, ALVA-55, ALVA-101, ARCaP, CWR-R1, DuCaP, DuPro-1, LAPC-4, MDA PCa 1, MDA PCa 2a, MDA PCa 2b, NCI-H660, PC-346C, PC-93, PSK-1, UM-SCP-1, and VCaP). Cytogenetics, DNA profiling, expression of basal, luminal, and neuroendocrine differentiation markers, and mutation analyses of the TP53 and androgen receptor (AR) genes were performed.

RESULTS

Based on cytogenetics and DNA profiling analyses, out of the 18 "non-classical" lines, six were confirmed to be unique, eight (in four pairs) were confirmed to be related in origin, and four lines were identified as cross-contaminants. Of this latter group, PC-93 was found to be a derivative of HeLa, whereas DuPro-1, ALVA-55, and ALVA-101 were derivatives of PC-3. The 17 genuine prostate cell lines expressed keratin 8 (K8) and K18. Nine showed AR expression, of which five harbored mutations in the AR gene. Prostate-specific antigen and DD3 were exclusively detected in AR expressing cell lines. Seven lines expressed the basal cell marker K5, three of these lines showed co-expression of AR.

CONCLUSIONS

This study defines a collection of 17 genuine prostate carcinoma cell lines. This collection, although small, constitutes a variety of different types and stages of prostate cancer, while it also partly reflects the heterogeneous nature of this malignancy.

Expression of fas ligand in metastatic prostatic carcinoma: suggestive of possible clonal expansion of subpopulation with metastatic potential

Fas ligand (FasL) is a type II transmembrane tumor necrosis factor family protein, known to trigger apoptosis in cells that bear the FasL receptor, Fas. The authors found that normal prostate, benign hyperplasia, and most prostatic carcinoma cells at the primary site did not express FasL, whereas metastatic prostatic carcinoma cells in lymph nodes and bone marrow displayed almost uniform, immunohistochemically detectable, FasL expression. However, small foci of FasL-positive prostatic carcinoma cells amid a vast majority of FasL-negative tumor cells were noted at the primary sites in patients with distant metastases. Analysis of the FasL gene and its mRNA by polymerase chain reaction and reverse transcriptase-polymerase chain reaction, respectively, suggested that the expression of immunohistochemically detectable FasL in metastatic tumor cells was not due to mutation in the FasL gene with resulting overexpression. Further, FasL expression was detectable in the acinar epithelial cells of prostates with morphologic atrophic changes, suggesting that FasL also plays a role in the physiologic apoptosis process of noncancerous prostate. The current data suggest that a subpopulation of prostate carcinoma cells clonally expresses FasL, and this subpopulation may have metastatic potential. Evaluation of FasL expression in the primary tumor thus may provide a useful parameter for predicting metastatic potential of the tumor.

Gene amplifications in advanced-stage human prostate cancer

Gene amplification is a model of proto-oncogene alterations occasionally observed in human tumors. This amplification can, in some cases, have prognostic value (N-myc in neuroblastoma, c-erbB2 and int-2 in breast cancer, etc.). Amplifications of the proto-oncogenes c-myc, c-erbB2 and int-2 have not yet been report in prostate adenocarcinoma, which, like breast cancer, is hormone dependent. We sought amplifications of these three proto-oncogenes by means of Southern blotting in 15 human prostate adenocarcinoma specimens, most of which were advanced (7 stage C and 6 stage D1 or D2). We confirmed the lack of c-myc and c-erbB2 amplification, regardless of the stage, in contrast to the case of breast cancer. Int-2 amplification was observed in one advanced tumor with bone metastases, out of a total of six stage D tumors. The precise frequency of int-2 amplification and its role in prostate carcinogenesis remain to be determined.

Expression of nm23-H1 and nm23-H2 proteins in prostate carcinoma

The nm23 gene products/nucleoside diphosphate (NDP) kinase expression in prostate carcinomas and benign hyperplasias was evaluated immunohistochemically. Monoclonal antibodies against nm23-H1 and nm23-H2 proteins were prepared using the corresponding proteins fused with glutathione S-transferase as immunogens. Of the 80 cases of nonmetastatic prostate carcinoma examined, 74% (59/80) and 60% (48/80) were immunoreactive for nm23-H1 or nm23-H2 protein, respectively. Negative staining for nm23-H1 occurred in 83% of metastatic lesions, while 34% were negative for nm23-H2. All primary tumors corresponding to the metastases examined showed positive immunostaining for nm23-H1, indicating an inverse relationship between expression of this protein and metastatic status. nm23-H2 protein was detected in 83% of primary tumors and its expression appeared to be significantly correlated to the degree of histological differentiation. In contrast, all cases of benign prostatic hyperplasia showed elevated levels of both nm23-H1 and nm23-H2 expression. These data suggest that the nm23/NDP kinase may play a role in suppressing the expression of malignant potential in prostate carcinomas.

A pathologist's view of prostatic carcinoma

In this article, the pathologic findings of carcinoma of the prostate were reviewed. Criteria were discussed for the pathologic diagnosis of prostatic carcinoma (PC), premalignant lesions, lesions that simulate PC, immunopathologic findings, special types of PC, effects of therapy on the prostate, and recent efforts to improve diagnostic and prognostic capabilities. The possible role of the study of nucleolar organizing regions was reported. A new method for demonstrating chromosomes in formaldehyde-fixed paraffin-embedded tissue was mentioned. The need for research in all aspects of the pathology of prostatic cancer was emphasized.

K-ras activation and ras p21 expression in latent prostatic carcinoma in Japanese men

Twenty-three clinically silent prostatic carcinomas discovered in Japanese men at autopsy were surveyed for ras proto-oncogene mutations by mutation-specific oligonucleotide probe hybridization after polymerase chain reaction (PCR) amplification from a section of formalin-fixed, paraffin-embedded tissue. Six of the 22 that were satisfactory amplified contained activating point mutations in codon 12 of K-ras, a significantly higher frequency than has been reported in patients with clinically advanced disease in the United States. Of the six cases with activating point mutations in codon 12 of K-ras, one had a GGT----GAT transition, four had GGT----GTT transversions, and one had both GGT----GAT and GGT----GTT mutations. Sections from the same tissues were immunohistochemically stained with an anti-ras p21 antibody. Carcinoma cells stained for ras p21 to some degree in 13 cases. Immunohistochemically detectable expression of p21 was always focal and was not necessarily associated with K-ras mutation. K-ras oncogene activation in prostatic carcinoma appears to merit additional study as a significant event in the pathogenesis of this neoplasm.

Markers of the metastatic phenotype in prostate cancer

Metastatic malignant disease is the single most common cause of treatment failure and subsequent mortality of most human malignancies, including prostate cancer. Presently, cells expressing the metastatic phenotype cannot be identified within a primary tumor population. Hence, accurate assessment of the likely behavior of an individual primary malignancy cannot be made at the time of diagnosis. The studies now reported have been aimed at identifying some of the features that may be associated with the metastatic phenotype of prostatic cancer. Insight into those factors that may be involved in prostate cancer metastasis has been gained from a variety of experimental approaches as well as study of intact human prostate cancers.

Osteocalcin: is it a useful marker of bone metastasis and response to treatment in advanced prostate cancer?

Serum osteocalcin (OC) is derived largely from new cellular synthesis. It is a marker for bone formation and a noninvasive specific marker of osteoblastic activity. The clinical significance of OC in monitoring prostatic cancer bone metastases was evaluated. Pretreatment serum OC levels were determined with a radioimmunoassay kit in a total of 63 patients with prostate cancer (8 with stage B, 12 with stage C, 12 with stage D1, and 31 with metastatic bone disease). The OC levels in patients with skeletal metastasis were significantly higher than those in patients without bony lesions (P less than 0.01). The pattern of the initial changes in OC levels were analyzed in patients with skeletal metastasis who received endocrine treatment. The pretreatment OC value is of little use in predicting the response to treatment. The patients whose OC level initially increased and remained high tended to have a shorter interval to disease progression. On the other hand, the pattern of initial changes in OC varied according to the regimen of endocrine treatment. Our study suggests that OC seem to reflect the response to treatment and might lead to the improvement in follow-up procedures. However, the clinical significance of OC as a marker of the response of bone metastasis should be carefully discussed with regard to the direct hormonal effect on bone metabolism.

Correlation of blood group antigen expression and oncogene-related proteins in malignant prostatic tissues

The use of multiple tumor markers could better predict the behavior of malignant tumors. In prostate cancer, there are no reliably predictive markers for metastatic behavior but the histologic grade and clinical stage of tumor do influence prognosis. We have determined the expression of blood group antigens A, B, H, Lewis-a and Lewis-b and the proto-oncogene proteins v-erbB, c-fos and v-H-ras in both benign and malignant prostate tissues by immunohistochemistry. We determined the relationship between these markers and the grade of malignancy and by inference, clinical behavior. There was reduced expression of blood group antigens A, Lewis-a and Lewis-b in all grades of prostatic adenocarcinoma. We also found that the expression of v-erbB was greater in tumors of high grade. We suggest that loss of blood group antigens may be correlated with elevated v-erbB oncoprotein expression (related to epidermal growth factor receptor) and increasing grade of prostatic malignancy.

Screening for prostatic carcinoma

Routine testing for prostatic carcinoma by digital rectal examination, transrectal ultrasonography and prostate specific antigen determination has been proposed to reduce deaths by earlier diagnosis. The questionable reliability of results, cost of screening, and inability to establish a balance between the benefits of treatment and the adverse effects on the quality of life of the men screened make screening experimental until controlled studies prove its value.

Increased survival of patients with massive lymphadenopathy and prostate cancer: evidence of heterogeneous tumour behaviour

The survival of patients with prostate cancer and radiologically detectable lymph node enlargement has been studied prospectively over an 8-year period. Computed tomography in 108 patients presenting with symptoms, signs or biochemical results suggesting lymphatic spread revealed pelvic or abdominal node masses in 60 patients; in 29 (48%), the masses measured more than 4 cm and the maximum node diameter was 15 cm. Two-thirds of patients had advanced (T3/T4) tumour stage. Following treatment, actuarial survival in all 60 patients with nodal enlargement was 40% at 5 years. Within this group, survival in 22 patients with lymphadenopathy but negative bone scans at diagnosis was significantly better than that of 38 patients with both node and bone disease (70% vs 20% at 5 years). This improvement was related both to an apparent inability of certain tumours initially to progress and seed within bone and to a marked sensitivity of the node masses to subsequent hormonal manipulation. Primary tumour grade was proportionally similar in both groups. Unexpectedly, 6 of the 38 patients with combined disease obtained a complete remission after treatment. The reason for this heterogeneous biological behaviour remains unclear; but these observations underscore the importance of vigorous treatment in all patients with advanced lymph node disease.

Histochemical examination of expression of ras p21 protein and R 1881-binding protein in human prostatic cancers

Expression of ras p21 was examined with monoclonal antibody RASK-3 in normal, benign hyperplasic, and cancerous prostates. In patients with stage D2 disease who received endocrine therapy, the relation between ras p21 expression, response to therapy, and prognosis was studied. In these patients, R 1881-binding protein (androgen receptor and progestin-binding protein) was also examined. Non-cancerous cells and most cancer cells from stage A patients did not express ras p21, while expression increased with both higher staging and grading. Staging pelvic lymphadenectomy was done in some stage A2-C cases, and presence of nodal metastasis was correlated with ras p21 expressions in the primary tumours. In stage D2, there was no correlation between ras p21 expression and R 1881-binding protein. Response to therapy and survival did not correlate with expression of ras p21, but was influenced by presence of R 1881-binding protein.

Correlation between bone scans and serum levels of osteocalcin, prostate-specific antigen, and prostatic acid phosphatase in monitoring patients with disseminated cancer of the prostate

Response of prostatic cancer bone metastases to therapy (androgen withdrawal and Estracyt) was studied in 43 patients by applying scintiscanning and radioimmunodetective measurement of serum osteocalcin (OC) values. The prostate-specific antigen (PSA) and prostatic acid phosphatase (PAP) concentrations, as sensitive probes for the overall tumor spread, were used in parallel in a monitoring procedure. A significant rise in OC levels to values elevated from a pretreatment normal level has been found in patients with a partial osseous tumor remission, and this may be easily distinguished from normal and/or subnormal OC level in bony tumor progression (P less than 0.01) and during stabilization in metastatic spread (P less than 0.01). On these bases, differences between disease progression and the "no change" response category could not be statistically recognized (P greater than 0.05). A sharp increase in circulating OC level has been recorded 1 months after the beginning of the treatment leading to bone remodeling processes and precedes improvements in scintiscan appearance. Blood OC concentration seems also to be of utility 1) in distinguishing scintigraphic flare phenomenon from a slight bone scan progression and 2) when related to scans with regions of both disease improvement and worsening. Furthermore, serum OC concentration can frequently be measured through a noninvasive procedure, thus serving as a significant addition to bone scintigraphy.