The Prostate 31:168–174 (1997) Orthotopic Implantation of Human Prostate Cancer Cell Lines: A Clinically Relevant Animal Model for Metastatic Prostate Cancer Klaus Rembrink,1* Johannes C. Romijn,1 Theo H. van der Kwast,2 Herbert Rübben,3 and Fritz H. Schröder1 1 Institute of Urology, Erasmus University Rotterdam, Rotterdam, The Netherlands Institute of Pathology, Erasmus University Rotterdam, Rotterdam, The Netherlands 3 University of Essen Medical School, Essen, Germany 2 BACKGROUND. To study the metastatic behavior of human prostate cancer cell lines, orthotopic injection in nude mice was performed. METHODS. Local tumor growth, metastasis formation, prostate-specific antigen, and androgen receptor expression were examined. RESULTS. Hormone-independent cell lines (PC-3, PC-3-125-1L, and TSU-Pr1) were highly tumorigenic and had a higher rate of lymph node metastasis after orthotopic than after subcutaneous implantation. PC-3 cell lines also consistently metastasized to the lungs. The androgen-sensitive LNCaP cell line showed local growth in 7 of 10, and lymph node metastasis in 4 animals. Significant serum PSA levels and strong receptor expression in primary and metastatic tumor tissues were observed. CONCLUSIONS. These results demonstrates that orthotopic implantation of human prostate cancer cell lines, including LNCaP, reproducibly leads to metastasis formation in nude mice. Prostate 31:168−174, 1997. © 1997 Wiley-Liss, Inc. KEY WORDS: prostate cancer; orthotopic; metastasis; androgen receptor; experimental animal INTRODUCTION Prostatic carcinoma has become the most frequent cancer in men in Western society during the last decade : 25–40% of all patients already show metastatic disease at the time of diagnosis. At this stage, a cure cannot be achieved by any treatment modality and new approaches are needed. A first step towards more promising therapeutic strategies would be a better understanding of the metastatic process in general. This can only be studied in detail by the availability of valid experimental metastatic model systems for prostate cancer. Metastases of human prostate cancer after subcutaneous (s.c.) implantation in nude mice have only occasionally been reported . Injection into the spleen of nude mice resulted in a slightly enhanced rate of metastasis [3,4]. PC-3 cells injected into the tail vein © 1997 Wiley-Liss, Inc. while the vena cava inferior was clamped led to tumor growth in lumbar vertebras, pelvis, and femur . More realistic models of human prostate cancer can be achieved by orthotopic (o.t.) implantation . Within the last several years this approach has been used with a variety of other tumors . Recently, two papers concerning o.t. implantation of human prostate cancer cell lines were published. Fu et al.  implanted PC-3 and DU145 tumor fragments, obtained after s.c. tumor growth. Stephenson et al.  injected suspensions of PC-3M and LNCaP cells into the nude mouse prostate. Metastases were found in almost all animals *Correspondence to: Klaus Rembrink, M.D., University of Essen Medical School, Institute of Urology, Division of Experimental Oncology, Hufelandstr. 55, 45122 Essen, Germany. Received 29 November 1995; Accepted 19 March 1996 Orthotopic TX of Prostate Cancer Cells 169 Fig. 1. A: Nude mouse in supine position. Lower abdomen is opened, and the seminal vesicles (SV) are carefully retracted with a sterile cotton swab. A 30-gauge needle is inserted into the right dorsal lobe of the prostate. B: Whole-mount section of a very large prostatic tumor (T) grown after orthotopic implantation of 1 × 106 PC-3-125-1L cells. Adjacent to the tumor mass, the empty bladder (B) as well as the right and left seminal vesicles (SV), partly filled with semen, can be seen. implanted with PC-3M cells, but in none of the mice with LNCaP cells. In spite of this negative finding, we included LNCaP in our studies, because it is the only hormone-sensitive in vitro cell line presenting the human androgen receptor and secreting prostate-specific antigen. The aims of this study were: 1) to determine tumor incidence and development of metastases in nude mice injected s.c. and o.t. with hormoneindependent and hormone-responsive prostate cancer cell lines, and 2) to evaluate the expression of marker proteins like prostate-specific antigen (PSA) and human androgen receptor (hAR) in tumors growing at different anatomical sites. MATERIALS AND METHODS Prostate Cancer Cell Lines Four human prostate cancer cell lines (PC3, PC3125-1L, TSU-Pr1, and LNCaP-FGC) were used in this study. The PC-3 cell line, established by Kaighn et al.  from a human bone metastasis, was obtained from Flow Laboratories (Irvine, Scotland). The LNCaP-FGC, further referred to as LNCaP, was originally provided by Dr. J. Horoszewicz . The TSUPr1 cell line, established by Iiuzmi et al.  from a lymph node metastasis, was obtained from Dr. Jack Schalken (University of Nijmegen, Nijmegen, The Netherlands). The PC3-125-1L cell line was a kind gift from Dr. J.L. Ware, (Medical College of Virginia/VCU, Richmond, VA) isolated from a spontaneous lung metastasis of the PC-3 cell line . Cell Culture TSU-Pr1 and LNCaP cells were maintained in RPMI-1640 (GIBCO, Burlington, Ontario) + 10% or 7.5% fetal calf serum (FCS), respectively. PC-3 and PC3-125-1L cells were cultured in Minimal Essential Medium (GIBCO) + 10% FCS. All media were supplemented with glutamine and penicillin/streptomycin. Cell cultures were maintained at 37°C in a humidified incubator in an atmosphere of 5% CO2 in air. At near 170 Rembrink et al. confluence, cells were washed with PBS, harvested with trypsin/EDTA, solution and finally suspended in Ca++- and Mg++-free Hank’s balanced salt solution. Only single-cell suspensions of >90% viability (Trypan blue exclusion) were used. Animals Athymic male nude mice (NMRI nu/nu, 8–12 weeks old, from the breeding colony at the Erasmus Experimental Animal Center, Erasmus, The Netherlands) were kept in laminar-flow cage racks under standardized environmental conditions throughout the experiments. Sterilized food (RMS-GH, obtained from Hope Farms, Woerden, The Netherlands) and acidified drinking water were supplied ad libitum. Tumor Cell Injection The mice were anesthesized with tribromoethanol  and placed in supine position. A midline incision was made in the lower abdomen. Tumor cell suspensions (20 ml) were injected into the dorsal mouse prostate using a 30-gauge needle and a 50-ml glass syringe (Hamilton Bonaduz AG, Switzerland). A slight elevation of either the right or the left dorsal prostate capsule was considered indicative of correct deposition of tumor cells (Fig. 1A). The abdomen was closed by single-stitch sutures using 5-0 silk. Subcutaneous injections (1 × 106 cells) were made unilaterally in the area of the right shoulder. Autopsy and Histology The mice were sacrificed by CO2 at various times (27–91 days) after tumor cell injection. After careful inspection of gross tumor growth the prostate, bladder, and seminal vesicles were taken out as one unit. Iliac and preaortic lymph nodes (in cases of o.t. injection) as well as (ipsi- and contralateral) axillary nodes (after s.c. injection), lungs, liver, and spine were collected and fixed in 3.7% formalin. After paraffinembedding, specimens were serially sectioned and stained with hematoxylin and eosin. The human origin of the tumors and metastases was confirmed by bisbenzimid staining . Prostate-Specific Antigen Serum Analysis Immediately after sacrifice, 1 ml blood was obtained by heart puncture from each mouse and clotted in Eppendorf tubes in the refrigerator overnight. The serum was separated and stored at −20°C. Measurement of human PSA was carried out with an automated microparticle-based enzyme immunoassay TABLE I. Tumorigenicity and Metastases of Three Human Hormone-Independent Prostate Cancer Cell Lines After Subcutaneous Implantation of 1 × 106 Tumor Cells Cell lines Days postinjection Mean tumor weight Range Tumorigenicity Metastases Lymph nodes Lungs Liver or bone Side effects PC-3 PC-3-125-1L TSU-Pr1 45 1,148 mg 470–2,190 5/5 Yes 3/5 1/5 None None 43 1,386 mg 380–2,130 5/5 Yes 4/5 4/5 None None 43 1,134 mg 370–1,780 5/5 None 0/5 0/5 None None (IMx-MEIA; Abbott Diagnostics, Wiesbaden-Delkenheim, Germany). Sensitivity of this assay was 0.07 ng/ ml. PSA concentration was expressed as nanograms of PSA per ml of serum. Immunohistochemistry Immunostaining with MAb F39.4 (obtained from the Institute of Pathology, Erasmus University, Rotterdam) for hAR was performed on formalin-fixed 5-mmthick slides, as described previously . RESULTS Subcutaneous Implantation With hormone-independent cell lines, subcutaneous implantation always resulted in tumor take in murine hosts. Mean tumor weights were very similar for the three cell lines (Table I). The tumors were grossly necrotic, partly surrounded by a solid capsule, and they demonstrated invasiveness (muscle, lymph nodes). In about 50% of these tumors, tumorinfiltrated lymph nodes were seen adjacent to the primary tumor, suggesting direct growth infiltration. The PC-3 cell line showed a metastatic involvement of the axillary lymph nodes in 3 of 5 mice. The more aggressive subline PC-3-125-1L produced positive nodes in 4 of 5 animals. The lungs were affected in one animal injected with PC-3 and in 4 animals with PC-3-125-1L cells. In the mouse with PC-3 lung metastases, only a few tumor-cell colonies were detected, whereas the animals inoculated with PC-3-125-1L presented disseminated tumor growth all over their lungs. In contrast, TSU-Pr1 showed no metastatic spread at all after s.c. injection. In none of the animals injected with either cell line were contralateral lymph node metastases detected. Growth and metastasis of s.c.-injected LNCaP cells were not tested in parallel, but previous experiments Orthotopic TX of Prostate Cancer Cells TABLE II. Dose Dependence of Tumor Take and Metastasis After Orthotopic Implantation of TSU-Pr1 Cells Cell no. 1.25 × 104 2.5 × 104 5 × 104 1.25 × 105 2.5 × 105 5 × 105 1 × 106 2 × 106 Days Take rate LN-Metb 40 40 40 27 27 27 27 27 1/3 2/3 1/1a 3/3 3/3 3/3 2/2a 3/3 1/3 2/3 1/1 2/3 0/3 1/3 1/2 2/3 TABLE III. Tumorigenicity and Metastatic Capacity of Three Human Hormone-Independent Prostate Cancer Cell Lines After Orthotopic Implantation of 1 × 106 Tumor Cells Cell lines a One and two animals were lost in the first week after injection, respectively. b Lymph node metastasis. consistently revealed a take rate of about 10%, with no sign of metastatic spread to the lymph nodes or other sites. Orthotopic Implantation In an attempt to determine the appropriate amount of tumor cells for o.t. inoculation, we varied the TSUPr1 cell number from 1.25 × 104 to 2 × 106 cells. Tumorigenicity was 100% in all groups except for the two lowest cell doses. The extent of metastasis was variable (Table II). No lymph node metastases were detected after injection of 2.5 × 105 cells, but at a dose of 2 × 106 cells, metastases were found in 2 of 3 animals. In none of the animals within this experiment was any sign of metastatic spread to visceral organs apparent. Based on these results, we decided to inject a dose of 1 × 106 tumor cells in further experiments. With all cell lines used, local tumor growth was observed after o.t. implantation. Growth rate and growth patterns, however, appeared to be quite different for the various cell lines. As shown in Table III, intraprostatic tumor growth was reflected by a considerable weight increase of the urogenital organs. Local tumors with a size up to 15 mm in diameter were detected (Fig. 1B). Invasion into the blood vessels, urethra, bladder, and seminal vesicles occurred in almost all animals injected with the hormone-independent cell lines (Fig. 2A), causing symptoms like urinary retention, hydronephrosis, and kidney insufficiency, often leading to early death of the animal. After o.t. injection, positive lymph nodes were always located in the retroperitoneal space. Metastases were easily detectable macroscopically due to their large size (up to 5 mm) and induration. On the other hand, small tumor-cell nests in the subcapsular space of the lymph node were only 171 Days postinjection Mean weight P + SV + Ba Range Tumorigenicity Metastases Lymph nodes Lungs Liver or bone Side effects PC-3 PC-3-125-1L TSU-Pr1 43 48 32 782 mg 590–990 5/5 Yes 5/5 5/5 None Yesb 1,306 mg 630–1,820 5/5 Yes 5/5 5/5 None Yesc 1,122 mg 830–1,360 5/5 Yes 2/5 0/5 None Yesd a P + SV + B, prostate + seminal vesicles + bladder. Normal average weight is 350–450 mg. b Two mice showed urinary retention; one mouse died on day 41. c Four mice had urinary retention in combination with bilateral kidney insufficiency. d One mouse had urinary retention; two mice died on days 27 and 32. found after microscopic examination (Fig. 2B). Lung metastases were rarely diagnosed by macroscopic inspection. Microscopic evaluation mostly showed small tumor-cell aggregations distributed all through the lung (Fig. 3A). Lung metastases were always associated with the presence of preaortic lymph node metastases. Tumor growth after inoculation with PC-3 cells was always accompanied by local invasion, preaortic lymph node metastases, and microscopic spread to the lungs (Fig. 3A). With PC-3-125-1L, thought to be a more aggressive subline of PC-3, we achieved similar results (Table III). With TSU-Pr1 cells, the take rate was 100%, but lymph node metastases were noticed in only 40% of the animals. Spread of tumor cells to the lungs never occurred. With the LNCaP cell line, a take rate of 70% was obtained 89–91 days after o.t. inoculation. LNCaP tumors expanded by dislodging surrounding structures rather than invading them. Angioinvasive growth was occasionally seen. Typical for LNCaP was the presence of grossly hemorrhagic areas within tumors. Tumor growth was observed in 7 mice o.t.-implanted with LNCaP cells. Four of these showed metastatic spread to the preaortic nodes. PSA Measurements and Immunohistochemical Staining of hAR PSA was undetectable in the sera of nude mice not implanted with tumor cells. As expected, we could not 172 Rembrink et al. Fig. 2. A: Mouse prostate 43 days after intraprostatic injection of 1 × 106 PC-3 cells. Note invasion into seminal vesicles (arrows) (original magnification × 10). B: Subcapsular metastasis of PC-3 cells in a preaortic lymph node. Below, normal mouse lymphocytes are seen. Human tumor cells are dislodging the regular lymphatic population (original magnification × 20). Fig. 3. A: Lung metastasis (arrow) 48 days after orthotopic implantation of PC-3-125-1L cells (original magnification × 20). B: Immunohistochemical staining of human androgen receptor in a metastatic lymph node of the hormone-sensitive LNCaP-FGC line, 89 days after intraprostatic inoculation. Note large, darkly stained spots representing human tumor-cell nuclei; in contrast, smaller mouse lymphocytes (top, left) are not stained (original magnification × 40). detect PSA in the serum of mice inoculated with any of the hormone-independent tumor cell lines. In contrast, 7 of 10 mice implanted o.t. with LNCaP cells were found to be PSA-positive. The PSA level ranged from 3.7–219.0 ng/ml. The highest PSA level was found in the animal with the largest tumor mass. In the 3 animals without macroscopic or microscopic LNCaP tumor growth, the PSA concentration was below detection level (Table IV). Tumors and metastases of PC-3, PC-3-125-1L, and TSU-Pr1 did not show detectable hAR expression. In contrast, the tumors formed by the LNCaP cells were hAR-positive in all cases, including lymph node metastases (Fig. 3B). DISCUSSION The limited availability of metastatic in vivo models is still a barrier to increasing our knowledge of the Orthotopic TX of Prostate Cancer Cells TABLE IV. Tumor Growth, Metastases, and PSA Levels After Orthotopic Implantation of 1 × 106 Tumor Cells of the Hormone-Sensitive LNCaP-FGC Line No. Days P + SV + Ba Growth LN-Metb PSA (ng/ml) 217 218 219 220 221 397 398 399 401 402 89 89 89 89 89 91 91 91 91 91 390 mg 460 mg 910 mg 1,130 mg 570 mg 1,180 mg 980 mg 660 mg 370 mg 470 mg Yes No Yes Yes No Yes Yes Yes No Yes Yes No No Yes No Yes Yes No No No 3.7 <0.07 148.0 190.0 <0.07 219.0 164.0 21.0 <0.07 14.7 a P + SV + B, prostate + seminal vesicles + bladder. Normal average weight is 350–450 mg. b Lymph node metastasis. process leading to progression and metastasis, especially in the field of prostate cancer. Among other techniques, subcutaneous, intravenous, or intraperitoneal injections of human cancer cells into nude mice were the most commonly used in prostate cancer research, but metastases rarely occurred . Implantation into other sites, such as seminal vesicles or spleen, has resulted in local tumor growth only and in formation of liver metastases, respectively. But these approaches must be considered less physiologically relevant than o.t. implantation [3,16]. The technique of o.t. implantation has already been shown to be effective in other fields, always resulting in rapid local tumor growth and often also in metastatic disease . In line with the hypothesis of Paget , stating that tumor growth and metastasis require the presence of a suitable microenvironment, o.t. implantation might be expected to result in higher tumorigenicity and metastases rates. Recently, two papers concerning orthotopic implantation of human prostate cancer cell lines were published [8,9]. Our data confirmed the findings of both groups. Except for LNCaP, all cell lines produced a tumorigenicity of 100% irrespective of injection site, using 1 × 106 cells. As shown with TSU-Pr1, as few as 1.25 × 104 cells are able to develop tumors rapidly after o.t. injection. In general, metastatic behavior was increased after o.t. implantation as compared to s.c. injection. As an exception, PC-3-125-1L (thought to be more aggressive than the parental PC-3) revealed nearly no difference in this respect. Apparently PC-3-125-1L had already lost the requirement for specific prostatic factors that allowed the parental PC-3 and other cell lines to produce metastatic disease. An unexpected finding was the relatively low metastatic ability of TSU-Pr1 cells, in 173 spite of aggressive local tumor growth. Only 32 days after o.t. implantation the animals had to be sacrificed, because almost all of them demonstrated serious illness due to local invasiveness and large tumor size. Still, relatively few metastases were formed compared to PC-3 and PC-3-125-1L. Clearly, rapid local tumor growth and metastatic behavior are not necessarily related. A similar observation was made after o.t. implantation of the in vivo prostate cancer cell line PC133 (data not shown), which produced very large local tumors (>13 mm) in all animals but never caused metastatic disease. The relevance of the prostate environment for growth and metastasis of prostate cancer cells is illustrated best by the LNCaP cell line. After s.c. injection no growth occurred at all, but following o.t. inoculation 7 of 10 animals developed tumors. This result is similar to the findings of Stephenson et al. . However, no lymphatic spread was noted by these authors after 40 days, whereas we demonstrated positive lymph nodes 89–91 days after implantation. To our knowledge, this is the first report of lymph node metastases formed after injection of LNCaP cells into nude mice. The metastases found were only occupying a small part in the subcapsular space and had not yet dislodged all the lymphatic tissue. We believe that, in addition to lymph node metastasis formation, metastatic spread to the lungs and the bone can also occur with LNCaP cells, given more time to develop. LNCaP is the only in vitro cell line expressing important markers such as PSA and hAR. PSA measurements were positive in all mice exhibiting local LNCaP tumor growth. The amount of PSA seems to be a function of the local prostate tumor volume and is not related to the development of lymph node metastases. PSA is probably a very sensitive tumor marker in the mouse, because in earlier experiments (not shown here), significant levels of PSA (up to 1.12 ng/ ml) were measured in mice in which tumors were not (yet) detectable by histological means. The immunohistochemical examination for hAR revealed a strong expression pattern not only in the primary tumor but also in the metastatic lymph nodes. There was no apparent loss of hAR during the metastatic process. In contrast to former assumptions, recent examinations with human material have proven that hAR is not necessarily lost in metastatic disease, even in patients presenting hormone-resistant disease . The unequivocal demonstration of hAR in lymph node metastases confirms the relevance of the o.t. model system and the importance of the LNCaP cell line for the imitation of human metastatic prostate cancer. 174 Rembrink et al. CONCLUSIONS In conclusion, o.t. implantation of human prostate cancer cells into nude mice leads to the reproducible formation of lymph node and lung metastases, and is superior to the s.c. technique. Whereas the implantation site seems to be less critical for the development of metastatic disease from aggressive cell lines such as PC-3-125-1L, growth and subsequent metastasis of other prostate cancer cell lines, such as LNCaP, appear to be strongly stimulated in a prostatic environment. During the metastatic process there was no loss of hAR. Our data show that o.t. implantation with human prostate cancer cell lines, especially LNCaP, is a very promising approach for studying the mechanisms of human prostate cancer progression and metastasis. Further experiments will be aimed at the isolation of a more metastatic LNCaP subline, still expressing important markers such as PSA and hAR. 6. 7. 8. 9. 10. 11. 12. ACKNOWLEDGMENTS 13. We gratefully acknowledge the work of Mrs. Coby Peekstok in excellently processing all histologies, and of Dr. M.A. Blankenstein in PSA measurements. K.R. is the recipient of a research grant of the Dr. Mildred Scheel Stiftung für Krebsforschung (German Cancer Foundation). 14. 15. REFERENCES 1. Boring CC, Squires TS, Tong T: Cancer statistics, 1994. CA 44:7, 1994. 2. Ware JL, Paulson DF, Mickey GH, Webbs KS: Spontaneous metastasis of cells of the human prostate carcinoma cell line PC-3 in athymic nude mice. J Urol 128:1064–1067, 1982. 3. Shevrin D, Grony K, Kukreja S: Patterns of metastasis by the human prostate cancer cell line PC-3 in athymic nude mice. Prostate 15:187–194, 1989. 4. Sherwood E, Ford J, Lee C, Kozlowski J: Therapeutic efficacy of recombinant tumor necrosis factor alpha in an experimental model of human prostatic carcinoma. J Biol Response Mod 9: 44–52, 1990. 5. Shevrin D, Kukreja S, Ghosh L, Lad T: Development of skeletal 16. 17. 18. 19. metastases by human prostate cancer in athymic nude mice. Clin Exp Metastasis 6:401–409, 1988. Hoffman RM: Orthotopic is orthodox: Why are orthotopictransplant metastatic models different from all other models? J Cell Biochem 56:1–3, 1994. Kubota T: Metastatic models of human cancer xenografted in the nude mouse: The importance of orthotopic transplantation. J Cell Biochem 56:4–8, 1994. Fu X, Herrera H, Hoffman RM: Orthotopic growth and metastasis of human prostate carcinoma in nude mice after transplantation of histological intact tissue. Int J Cancer 52:987–990, 1992. Stephenson RA, Dinney CN, Gohji K, Ordonez NG, Killion JJ, Fidler IJ: Metastatic model for human prostate cancer using orthotopic implantation in nude mice. JNCI 84:951–957, 1992. Kaighn ME, Narayan KS, Ohnuki Y, Lechner J, Jones LW: Establishment of a human prostate cancer cell line (PC-3). Invest Urol 17:16–23, 1979. Horoszewicz JS, Leong SS, Kawinski E, Karr JP, Rosenthal H, Ming-Chu T, Mirand EA, Murphy GP: LNCaP model of human prostatic carcinoma. Cancer Res 43:1809–1818, 1983. Iizumi T, Yazuki T, Kanoh S, Kondo I, Koiso K: Establishment of a new prostatic carcinoma cell line (TSU-Pr1). J Urol 137:1304– 1306, 1987. Papaioannou VE, Fox JG: Efficacy of tribrimoethanol anesthesia in mice. Lab Anim Sci 43:189–192, 1993. Rygaard K: A rapid method for identification of murine cells in human malignant tumours grown in nude mice. In Rygaard J, Brunner N, Graen N, Spang-Thompson M (eds): ‘‘ImmuneDeficient Animals in Biomedical Research.’’ Basel: Karger, 1987, pp 268–272. Janssen PJ, Brinkmann AO, Boersma WJ, van der Kwast TH: Immunohistochemical detection of the androgen receptor with monoclonal antibody F39.4 in routinely processed, paraffinembedded human tissues after microwave pre-treatment. J Histochem Cytochem 42:1169–1175, 1994. Kozlowski JM, Fidler IJ, Campbell D: Metastatic behavior of human tumor cell lines grown in nude mouse. Cancer Res 44: 3522–3529, 1984. Fidler IJ: Rationale and methods for the use of nude mice to study the biology and therapy of human cancer metastasis. Cancer Metastasis Rev 5:29–49, 1986. Paget S: The distribution of secondary growths in cancer of the breast. Cancer Metastasis Rev 1:571–573, 1989. Ruizeveld de Winter JA, Janssen PJA, Sleddens HMEB, VerleunMooijman MCT, Trapmann J, Brinkmann AO, Santerse AB, Schröder FH, van der Kwast TH: Androgen receptor status in localized and locally progressive hormone refractory human prostate cancer. Am J Pathol 144:735–746, 1994.