- Email: [email protected]
The Ultrastructural Localization of Prostatic Specific Antigen and Prostatic Acid Phosphatase in Hyperplastic and Neoplastic Human Prostates
0022-534 7/85/1343-0607$02.00/0 Vol. 134, August
THE JOURNAL OF UROLOGY
Printed in U.S.A.
Copyright © 1985 by The Williams & Wilkins Co.
THE ULTRASTRUCTURAL LOCALIZATION OF PROSTATIC SPECIFIC ANTIGEN AND PROSTATIC ACID PHOSPHATASE IN HYPERPLASTIC AND NEOPLASTIC HUMAN PROSTATES MICHAEL J. WARHOL*
AND
JANINA A. LONGTINE
From the Department of Pathology, the Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts
ABSTRACT
A low temperature embedding, protein A-gold technique was used to localize prostatic specific antigen and prostatic acid phosphatase at the ultrastructural level in hyperplastic and neoplastic human prostates. Prostatic specific antigen immunoreactivity was localized over the endoplasmic reticulum, cytoplasmic vesicles and vacuoles, and within the lumina of prostatic glands. In contrast, prostatic acid phosphatase immunoreactivity was localized to lysosomal granules. The pattern of labelling was similar in both hyperplastic glands and adenocarcinomas. This is the first localization of prostatic specific antigen at the ultrastructural level. The localization of prostatic acid phosphatase by an immunochemical technique confirms and expands previous histochemical observations. . Human prostatic tissue produces at least two specific substances: prostatic acid phosphatase and prostatic specific antigen. The first of these substances identified was prostatic acid phosphatase. 1•2 Acid phosphatases are found in a variety of tissues: in erythrocytes, and in glandular epithelium including prostate, breast, stomach and colon. Low levels of activity are detected in thyroid, kidney and ovary. Enzyme histochemistry at the ultrastructural level has suggested that this enzyme is lysosomal in origin. 3 The tissue levels of acid phosphatase in the prostate are under hormonal influence and can be induced by androgens but not estrogens. Lysosomal granules are similarly influenced by these same hormones. 4 This is indirect morphologic evidence that this enzyme is lysosomal in origin. Although three acid phosphatases have been identified in the prostate, only one appears to be prostate specific. 1• 2 This isoenzyme has a molecular weight of approximately 100 kd and a pH optimum of 5 to 7. This acid phosphatase is immunologically distinct from acid phosphatases present in other tissues. This unique antigenicity allows serum determinations and tissue immunohistochemistry to be sensitive and highly specific tests for the diagnosis of carcinoma of the prostate. 5 A more recently described marker of prostatic tissue is prostatic specific antigen.6-8 This antigen has a molecular weight of 33 kd and is chemically and immunologically distinct from prostatic acid phosphatase. No ultrastructural site of prostatic specific antigen has been suggested. Light microscopic observations have suggested that this marker is concentrated in perinuclear regions on the luminal side of the gland. We employed a low temperature embedding procedure with Lowicryl K4M using protein A-gold as a marker to localize both prostatic acid phosphatase and prostatic specific antigen immunochemically at the ultrastructural level.
4C until embedding. Prior to embedding, the tissue was rinsed in 0.5 M NH4 Cl to quench unreacted aldehyde groups. The tissues were embedded in Lowicryl K4M according to a protocol previously described.9 Briefly, the tissues were treated with a series of one hour incubations at -20C in increasing concentrations of ethanol, 50, 75, 95 and 100 per cent. Tissue infiltration with Lowicryl K4M was accomplished by one hour incubations with increasing concentrations of Lowicryl in ethanol, 1:1, 2:1, and finally pure Lowicryl, all at -35C. Final infiltration was accomplished by an overnight incubation with pure Lowicryl at -35C. The tissues were then transferred to gelatin capsules and initial polymerization begun at -35C with UV light as a catalyst. After twenty-four hours the tissues were brought to room temperature where polymerization was allowed to continue for an additional 48 hours. After polymerization, thick 1 µ sections were cut and stained with toluidine blue to select appropriate areas for thin sectioning. Thin sections were cut and picked up on formvar coated nickel grids. At least two blocks per specimen were selected for immunochemical staining. The immunochemical procedures were performed as previously described. 9 The grids were first incubated with 0.5 per cent albumin for five minutes. They were then incubated for one hour at room temperature with the respective antisera. The anti-prostatic acid phosphatase (Immulok) was incubated at dilutions of 1:2 and 1:3 in PBS. The anti-prostatic specific antigen (Dakopatts) was incubated at dilutions of 1:200 and 1:400 in PBS. The grids were then washed with PBS and incubated with protein A-gold for one hour at room temperature. The protein A-gold was prepared as previously described. 9 The grids were then washed in PBS followed by distilled water and stained with uranyl acetate followed by lead citrate. Controls included grids incubated with non-immune rabbit serum and grids incubated with Tris buffer. A JEOL CoJEMS 100 electron microscope was used for observation and photography.
MATERIALS AND METHODS
Human prostatic tissue was obtained promptly after surgical removal. This material included three specimens from hyperplastic glands obtained by transurethral resection, and two carcinomas of the prostate obtained by radical prostatectomy. The tissue was minced into 1 mm. cubes and fixed for one hour in 3 per cent paraformaldehyde-0.1 per cent glutaraldehyde. The fixative was removed by three washes in phosphate buffered saline, pH 7.4 (PBS). The tissue was stored in buffer at Accepted for publication April 24, 1985. *Requests for reprints: Dept. of Pathology, Brigham & Women's Hospital, 75 Francis St., Boston, MA 02115.
RESULTS
The prostate morphology obtained by the fixation and embedding protocol described closely approximates that of conventionally fixed and embedded tissues. 3• 10 The lack of osmication renders intracellular membrane structures indistinct. However, endoplasmic reticulum and Golgi areas could be clearly identified because of their configuration, their location within the cell, and in the case of the endoplasmic reticulum, the presence of ribosomes (figs. 2, 4). Numerous lysosomes were present in both benign and malignant epithelium.
607
608
WARHOL AND LONGTINE
FIG. l. Anti-prostatic specific antigen. A, hyperplastic prostatic epithelium with vacuoles within cytoplasm labelled with gold particles. Lumen (L) contains amorphous granular material labelled with gold particles (arrow) (X 8,000) (bar= 350 nm.). B, both large and small vacuoles are labelled. Large vacuole contains amorphous granular material (open arrow) similar to material seen within lumen (small arrow) (X 15,000) (bar = 200 nm.).
609
ULTRASTRUCTURE OF PROSTATIC ANTIGENS
plasmic reticulum (Fig. 2A). Structures whose configurations were consistent with Golgi lamellae were unlabelled; however, vesicle-like structures adjacent to the lamellae were frequently labelled (fig. 2B). The supporting basal epithelial cells failed to label with the antisera (fig. 3A, 3B). A similar pattern of labelling was noted with malignant epithelium (figs. 4, 5). However, many of these glands lacked label within their lumina. The pattern of labelling was similar in all fields examined, with little variation between cases. Prostatic acid phosphatase antiserum localized primarily to lysosomal granules (fig. 6A, 6B). The majority of lysosomal granules were labelled. The perinuclear localization of some of the granules suggested that they might be originating within the proximal face of the Golgi apparatus. The indistinct rendering of the Golgi apparatus in these preparations precludes a definite identification. Label was irregularly noted within the lumina in both hyperplastic and adenocarcinomatous glands (figs. 7, 8). No label was noted within basal epithelial cells. Control grids revealed only weak, non-specific background labelling. DISCUSSION
Our results suggest that prostatic acid phosphatase and prostatic specific antigen are segregated into different storage compartments within the glandular epithelial cell. Our immunochemical technique confirms earlier observations employing enzyme histochemistry3 that prostatic acid phosphatase is primarily a lysosomal enzyme. The failure of this antiserum to label other cytoplasmic structures, specifically rough endoplasmic reticulum, may indicate that this antigen does not assume its exclusive form until it is packaged into a lysosome. An alternate hypothesis is that the sensitivity of the method only permits the detection of high concentrations of antigen. Longer incubations and higher concentrations of antisera failed to produce different patterns of labelling. Therefore, we favor the hypothesis that prostatic acid phosphatase obtains its unique idiotype as it is packaged into granules. In contrast to acid phosphatase, prostatic specific antigen failed to label lysosomal structures in hyperplastic glands. Rare lysosomes adjacent to the lumen were labelled in the adenocarcinomas. Antigenicity was detected over the rough endoplasmic reticulum, suggesting its synthesis by this structure. Structures which presumably represent Golgi lamellae were unlabelled, but vesicles were labelled. Subnuclear vesicles were also labelled. In contrast to the hyperplastic glands which had heavy labelling within their lumina, the adenocarcinomas exhibited irregular labelling of luminal secretions. Our observations suggest that prostatic specific antigen is synthesized by the rough endoplasmic reticulum, stored in vesicles and vacuoles, and released into the glandular lumina by exocytosis. The functional nature of prostatic specific antigen remains unknown. Although the supporting basal epithelial cells are presumably the precursors of the glandular epithelial cells, no immunoreactivity for either PrSA or PrAP could be detected within these cells. These antigens appear to be unique for glandular epithelium. FIG. 2. Anti-prostatic specific antigen. A, endoplasmic reticulum (R) is labelled but lysosomal granules are unlabelled (X 10,000) (bar= 300 nm.). B, cellular organelle with stacked cisternae consistent with Golgi lamellae (G) is unlabelled. Vesicles at lateral aspect of this structure are labelled (arrow) (X 8,000) (bar= 350 nm.).
Different patterns of labelling were observed with the prostatic specific antigen (PrSA) and the prostatic acid phosphatase (PrAP). Using the PrSA antiserum, label was present in heaviest concentration over cytoplasmic vacuoles (fig. IA). Many of these vacuoles contained amorphous faintly electron dense material (fig. lB). Similar material was present in the glandular lumina. Gold label was also detected over the rough endo-
REFERENCES 1. Manley, P. N., Mahan, D. E., Bruce, A. W. and Franchi, L.: Prostate-specific acid phosphatase. In: Diagnostic Immunohistochemistry. Edited by R. A. DeLeLellis. Masson, New York, pp. 313-324, 1981. 2. Romas, N. A., Rose, N. R. and Tannenbaum, M.: Acid phosphatase: new developments. Human Pathol., 10: 501, 1979. 3. Fisher, E. R. and Sieracki, J. C.: Ultrastructure of human normal and neoplastic prostate. In: Pathology Annual. Edited by S. C. Sommers. Appleton-Century-Crofts, New York, pp. 1-26, 1970. 4. Merk, F. B., Leav, I., Kwan, P. W. and Afner, P.: Effects of estrogen and androgen on the ultrastructure of secretory granules and intercellular junctions in regressed canine prostate. Anat. Rec., 197: 111, 1980.
610
WARHOL AND LONG TINE
FIG. 3. Anti-prostatic specific antigen. A, glandular cell (G) demonstrates labelling of presumably secretory vacuoles (arrowheads). Supporting basal cell (S) is unlabelled. Basal lamina is clearly seen (arrow). There is weak background staining of stroma (X 5,000) (bar = 400 nm.). B, vacuoles (arrowheads) within glandular cell (G) are labelled, but supporting basal cell (S) is unlabelled. Both cells are surrounded by basal lamina (arrow) (X 8,000) (bar= 350 nm.).
ULTRASTRUCTURE OF PROSTATIC ANTl:GENS
611
FIG. 4. Anti-prostatic specific antigen. Adenocarcinoma cell demonstrates labelling of vesicles (arrows) adjacent to presumed Golgi apparatus (G) and labelling of vacuoles in apical cytoplasm (arrowheads). Lumen is unlabelled (X 6,000) (bar= 375 nm.).
FIG. 6. Anti-prostatic acid phosphatase. A, perinuclear (N) granule is labelled (x 10,000) (bar= 300 nm.). B, some, but not all, lysosomes are labelled (X 8,000) (bar= 350 nm.).
612
WARHOL AND LONG TINE
FIG. 5. Anti-prostatic specific antigen. Adenocarcinoma with labelling of secretory vacuoles. Tumor cell also contains prominent bundles of tonofilaments (arrows) (X 10,000) (bar= 300 nm.).
ULTRASTRUCTURE OF PROSTATIC ANTIGENS
613
FIG. 7. Anti-prostatic acid phosphatase. Adenocarcinoma cell with large nucleolus (N) exhibits labelling of lysosomal structures, some of which appear to fuse with vesicles (X 10,000) (bar= 300 nm.).
5. Nadji, M., Taber, S. Z., Castro, A., Ming Chu, T. and Morales, A.: Prostatic origin of tumors, an immunohistochemical study. Am. J. Clin. Pathol., 73: 735, 1979. 6. Nadji, M., Taber, S. Z., Castro, A., Ming Chu, T., Murphy, G. P., Wang, M. C. and Morales, A.: Prostatic-specific antigen: an immunohistochemical marker for prostatic neoplasms. Cancer, 48: 1229, 1981. 7. Papsidero, L. D., Wang, P. C., Valenzuela, L.A., Murphy, G. P. and Ming Chu, T.: A prostate antigen in sera of prostate cancer patients. Cancer Res., 40: 2428, 1980. 8. Stein, B. J., Petersen, R. 0., Vangore, S. and Kendall, A. R.: Immunoperoxidase localization of prostate specific antigen. Am. J. Surg. Pathol., 6: 553, 1982.
FIG. 8. Anti-prostatic acid phosphatase. There is labelling of lysosomal granules (arrows) and vacuoles within apical cytoplasm. Wef 1c labelling is noted in lumen (L) (X 8,000) (bar= 350 nm.). 9. Warhol, M. J.: The ultrastructural localization of keratin proteins and carcinoembryonic antigen in malignant mesotheliomas. Am. J. Pathol., 116: 385, 1984. 10. Brandes, D., Kirdheim, D. and Scott, W.: Ultrastructure of the human prostate: normal and neoplastic. Lab. Invest., 13: 1541, 1964.
THE JOURNAL OF UROLOGY
Printed in U.S.A.
Copyright © 1985 by The Williams & Wilkins Co.
THE ULTRASTRUCTURAL LOCALIZATION OF PROSTATIC SPECIFIC ANTIGEN AND PROSTATIC ACID PHOSPHATASE IN HYPERPLASTIC AND NEOPLASTIC HUMAN PROSTATES MICHAEL J. WARHOL*
AND
JANINA A. LONGTINE
From the Department of Pathology, the Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts
ABSTRACT
A low temperature embedding, protein A-gold technique was used to localize prostatic specific antigen and prostatic acid phosphatase at the ultrastructural level in hyperplastic and neoplastic human prostates. Prostatic specific antigen immunoreactivity was localized over the endoplasmic reticulum, cytoplasmic vesicles and vacuoles, and within the lumina of prostatic glands. In contrast, prostatic acid phosphatase immunoreactivity was localized to lysosomal granules. The pattern of labelling was similar in both hyperplastic glands and adenocarcinomas. This is the first localization of prostatic specific antigen at the ultrastructural level. The localization of prostatic acid phosphatase by an immunochemical technique confirms and expands previous histochemical observations. . Human prostatic tissue produces at least two specific substances: prostatic acid phosphatase and prostatic specific antigen. The first of these substances identified was prostatic acid phosphatase. 1•2 Acid phosphatases are found in a variety of tissues: in erythrocytes, and in glandular epithelium including prostate, breast, stomach and colon. Low levels of activity are detected in thyroid, kidney and ovary. Enzyme histochemistry at the ultrastructural level has suggested that this enzyme is lysosomal in origin. 3 The tissue levels of acid phosphatase in the prostate are under hormonal influence and can be induced by androgens but not estrogens. Lysosomal granules are similarly influenced by these same hormones. 4 This is indirect morphologic evidence that this enzyme is lysosomal in origin. Although three acid phosphatases have been identified in the prostate, only one appears to be prostate specific. 1• 2 This isoenzyme has a molecular weight of approximately 100 kd and a pH optimum of 5 to 7. This acid phosphatase is immunologically distinct from acid phosphatases present in other tissues. This unique antigenicity allows serum determinations and tissue immunohistochemistry to be sensitive and highly specific tests for the diagnosis of carcinoma of the prostate. 5 A more recently described marker of prostatic tissue is prostatic specific antigen.6-8 This antigen has a molecular weight of 33 kd and is chemically and immunologically distinct from prostatic acid phosphatase. No ultrastructural site of prostatic specific antigen has been suggested. Light microscopic observations have suggested that this marker is concentrated in perinuclear regions on the luminal side of the gland. We employed a low temperature embedding procedure with Lowicryl K4M using protein A-gold as a marker to localize both prostatic acid phosphatase and prostatic specific antigen immunochemically at the ultrastructural level.
4C until embedding. Prior to embedding, the tissue was rinsed in 0.5 M NH4 Cl to quench unreacted aldehyde groups. The tissues were embedded in Lowicryl K4M according to a protocol previously described.9 Briefly, the tissues were treated with a series of one hour incubations at -20C in increasing concentrations of ethanol, 50, 75, 95 and 100 per cent. Tissue infiltration with Lowicryl K4M was accomplished by one hour incubations with increasing concentrations of Lowicryl in ethanol, 1:1, 2:1, and finally pure Lowicryl, all at -35C. Final infiltration was accomplished by an overnight incubation with pure Lowicryl at -35C. The tissues were then transferred to gelatin capsules and initial polymerization begun at -35C with UV light as a catalyst. After twenty-four hours the tissues were brought to room temperature where polymerization was allowed to continue for an additional 48 hours. After polymerization, thick 1 µ sections were cut and stained with toluidine blue to select appropriate areas for thin sectioning. Thin sections were cut and picked up on formvar coated nickel grids. At least two blocks per specimen were selected for immunochemical staining. The immunochemical procedures were performed as previously described. 9 The grids were first incubated with 0.5 per cent albumin for five minutes. They were then incubated for one hour at room temperature with the respective antisera. The anti-prostatic acid phosphatase (Immulok) was incubated at dilutions of 1:2 and 1:3 in PBS. The anti-prostatic specific antigen (Dakopatts) was incubated at dilutions of 1:200 and 1:400 in PBS. The grids were then washed with PBS and incubated with protein A-gold for one hour at room temperature. The protein A-gold was prepared as previously described. 9 The grids were then washed in PBS followed by distilled water and stained with uranyl acetate followed by lead citrate. Controls included grids incubated with non-immune rabbit serum and grids incubated with Tris buffer. A JEOL CoJEMS 100 electron microscope was used for observation and photography.
MATERIALS AND METHODS
Human prostatic tissue was obtained promptly after surgical removal. This material included three specimens from hyperplastic glands obtained by transurethral resection, and two carcinomas of the prostate obtained by radical prostatectomy. The tissue was minced into 1 mm. cubes and fixed for one hour in 3 per cent paraformaldehyde-0.1 per cent glutaraldehyde. The fixative was removed by three washes in phosphate buffered saline, pH 7.4 (PBS). The tissue was stored in buffer at Accepted for publication April 24, 1985. *Requests for reprints: Dept. of Pathology, Brigham & Women's Hospital, 75 Francis St., Boston, MA 02115.
RESULTS
The prostate morphology obtained by the fixation and embedding protocol described closely approximates that of conventionally fixed and embedded tissues. 3• 10 The lack of osmication renders intracellular membrane structures indistinct. However, endoplasmic reticulum and Golgi areas could be clearly identified because of their configuration, their location within the cell, and in the case of the endoplasmic reticulum, the presence of ribosomes (figs. 2, 4). Numerous lysosomes were present in both benign and malignant epithelium.
607
608
WARHOL AND LONGTINE
FIG. l. Anti-prostatic specific antigen. A, hyperplastic prostatic epithelium with vacuoles within cytoplasm labelled with gold particles. Lumen (L) contains amorphous granular material labelled with gold particles (arrow) (X 8,000) (bar= 350 nm.). B, both large and small vacuoles are labelled. Large vacuole contains amorphous granular material (open arrow) similar to material seen within lumen (small arrow) (X 15,000) (bar = 200 nm.).
609
ULTRASTRUCTURE OF PROSTATIC ANTIGENS
plasmic reticulum (Fig. 2A). Structures whose configurations were consistent with Golgi lamellae were unlabelled; however, vesicle-like structures adjacent to the lamellae were frequently labelled (fig. 2B). The supporting basal epithelial cells failed to label with the antisera (fig. 3A, 3B). A similar pattern of labelling was noted with malignant epithelium (figs. 4, 5). However, many of these glands lacked label within their lumina. The pattern of labelling was similar in all fields examined, with little variation between cases. Prostatic acid phosphatase antiserum localized primarily to lysosomal granules (fig. 6A, 6B). The majority of lysosomal granules were labelled. The perinuclear localization of some of the granules suggested that they might be originating within the proximal face of the Golgi apparatus. The indistinct rendering of the Golgi apparatus in these preparations precludes a definite identification. Label was irregularly noted within the lumina in both hyperplastic and adenocarcinomatous glands (figs. 7, 8). No label was noted within basal epithelial cells. Control grids revealed only weak, non-specific background labelling. DISCUSSION
Our results suggest that prostatic acid phosphatase and prostatic specific antigen are segregated into different storage compartments within the glandular epithelial cell. Our immunochemical technique confirms earlier observations employing enzyme histochemistry3 that prostatic acid phosphatase is primarily a lysosomal enzyme. The failure of this antiserum to label other cytoplasmic structures, specifically rough endoplasmic reticulum, may indicate that this antigen does not assume its exclusive form until it is packaged into a lysosome. An alternate hypothesis is that the sensitivity of the method only permits the detection of high concentrations of antigen. Longer incubations and higher concentrations of antisera failed to produce different patterns of labelling. Therefore, we favor the hypothesis that prostatic acid phosphatase obtains its unique idiotype as it is packaged into granules. In contrast to acid phosphatase, prostatic specific antigen failed to label lysosomal structures in hyperplastic glands. Rare lysosomes adjacent to the lumen were labelled in the adenocarcinomas. Antigenicity was detected over the rough endoplasmic reticulum, suggesting its synthesis by this structure. Structures which presumably represent Golgi lamellae were unlabelled, but vesicles were labelled. Subnuclear vesicles were also labelled. In contrast to the hyperplastic glands which had heavy labelling within their lumina, the adenocarcinomas exhibited irregular labelling of luminal secretions. Our observations suggest that prostatic specific antigen is synthesized by the rough endoplasmic reticulum, stored in vesicles and vacuoles, and released into the glandular lumina by exocytosis. The functional nature of prostatic specific antigen remains unknown. Although the supporting basal epithelial cells are presumably the precursors of the glandular epithelial cells, no immunoreactivity for either PrSA or PrAP could be detected within these cells. These antigens appear to be unique for glandular epithelium. FIG. 2. Anti-prostatic specific antigen. A, endoplasmic reticulum (R) is labelled but lysosomal granules are unlabelled (X 10,000) (bar= 300 nm.). B, cellular organelle with stacked cisternae consistent with Golgi lamellae (G) is unlabelled. Vesicles at lateral aspect of this structure are labelled (arrow) (X 8,000) (bar= 350 nm.).
Different patterns of labelling were observed with the prostatic specific antigen (PrSA) and the prostatic acid phosphatase (PrAP). Using the PrSA antiserum, label was present in heaviest concentration over cytoplasmic vacuoles (fig. IA). Many of these vacuoles contained amorphous faintly electron dense material (fig. lB). Similar material was present in the glandular lumina. Gold label was also detected over the rough endo-
REFERENCES 1. Manley, P. N., Mahan, D. E., Bruce, A. W. and Franchi, L.: Prostate-specific acid phosphatase. In: Diagnostic Immunohistochemistry. Edited by R. A. DeLeLellis. Masson, New York, pp. 313-324, 1981. 2. Romas, N. A., Rose, N. R. and Tannenbaum, M.: Acid phosphatase: new developments. Human Pathol., 10: 501, 1979. 3. Fisher, E. R. and Sieracki, J. C.: Ultrastructure of human normal and neoplastic prostate. In: Pathology Annual. Edited by S. C. Sommers. Appleton-Century-Crofts, New York, pp. 1-26, 1970. 4. Merk, F. B., Leav, I., Kwan, P. W. and Afner, P.: Effects of estrogen and androgen on the ultrastructure of secretory granules and intercellular junctions in regressed canine prostate. Anat. Rec., 197: 111, 1980.
610
WARHOL AND LONG TINE
FIG. 3. Anti-prostatic specific antigen. A, glandular cell (G) demonstrates labelling of presumably secretory vacuoles (arrowheads). Supporting basal cell (S) is unlabelled. Basal lamina is clearly seen (arrow). There is weak background staining of stroma (X 5,000) (bar = 400 nm.). B, vacuoles (arrowheads) within glandular cell (G) are labelled, but supporting basal cell (S) is unlabelled. Both cells are surrounded by basal lamina (arrow) (X 8,000) (bar= 350 nm.).
ULTRASTRUCTURE OF PROSTATIC ANTl:GENS
611
FIG. 4. Anti-prostatic specific antigen. Adenocarcinoma cell demonstrates labelling of vesicles (arrows) adjacent to presumed Golgi apparatus (G) and labelling of vacuoles in apical cytoplasm (arrowheads). Lumen is unlabelled (X 6,000) (bar= 375 nm.).
FIG. 6. Anti-prostatic acid phosphatase. A, perinuclear (N) granule is labelled (x 10,000) (bar= 300 nm.). B, some, but not all, lysosomes are labelled (X 8,000) (bar= 350 nm.).
612
WARHOL AND LONG TINE
FIG. 5. Anti-prostatic specific antigen. Adenocarcinoma with labelling of secretory vacuoles. Tumor cell also contains prominent bundles of tonofilaments (arrows) (X 10,000) (bar= 300 nm.).
ULTRASTRUCTURE OF PROSTATIC ANTIGENS
613
FIG. 7. Anti-prostatic acid phosphatase. Adenocarcinoma cell with large nucleolus (N) exhibits labelling of lysosomal structures, some of which appear to fuse with vesicles (X 10,000) (bar= 300 nm.).
5. Nadji, M., Taber, S. Z., Castro, A., Ming Chu, T. and Morales, A.: Prostatic origin of tumors, an immunohistochemical study. Am. J. Clin. Pathol., 73: 735, 1979. 6. Nadji, M., Taber, S. Z., Castro, A., Ming Chu, T., Murphy, G. P., Wang, M. C. and Morales, A.: Prostatic-specific antigen: an immunohistochemical marker for prostatic neoplasms. Cancer, 48: 1229, 1981. 7. Papsidero, L. D., Wang, P. C., Valenzuela, L.A., Murphy, G. P. and Ming Chu, T.: A prostate antigen in sera of prostate cancer patients. Cancer Res., 40: 2428, 1980. 8. Stein, B. J., Petersen, R. 0., Vangore, S. and Kendall, A. R.: Immunoperoxidase localization of prostate specific antigen. Am. J. Surg. Pathol., 6: 553, 1982.
FIG. 8. Anti-prostatic acid phosphatase. There is labelling of lysosomal granules (arrows) and vacuoles within apical cytoplasm. Wef 1c labelling is noted in lumen (L) (X 8,000) (bar= 350 nm.). 9. Warhol, M. J.: The ultrastructural localization of keratin proteins and carcinoembryonic antigen in malignant mesotheliomas. Am. J. Pathol., 116: 385, 1984. 10. Brandes, D., Kirdheim, D. and Scott, W.: Ultrastructure of the human prostate: normal and neoplastic. Lab. Invest., 13: 1541, 1964.