(CANCER RESEARCH 54, 1566-1573, March 15. 1<XM] Increased Androgen Receptor Activity and Altered c-myc Expression in Prostate Cancer Cells after Long-Term Androgen Deprivation1 John Kokontis, Kenji Takakura, Nissim Hay, and Shutsung Liao2 The Ben May Institute ¡J.K., K. T., N. H., S. L.¡,Departments of Biochemistry and Molecular Biology [S. L.j and Pharmacology and Physiology [N. H/, The University of Chicago, Chicago. Illinois 60637 ABSTRACT cellular events probably occur which allow cell proliferation to bypass the androgen requirement. Loss of AR expression may, therefore, Proliferation of LNCaP 104-S cells, a donai subline of the human either drive the selection for such events or occur secondarily (4). prostate cancer cell line, was very slow in androgen-depleted medium but The androgen-responsive human prostatic carcinoma cell line LN increased 10-13-fold in the presence of 0.1 UMof a synthetic androgen, KISSI. This induction of proliferation was diminished at higher concen CaP (5, 6) has been used extensively as an in vitro model for exam trations of KISSI, indicating the biphasic nature of the androgen effect. ining the role of the AR in the control of prostatic carcinoma cell After 20-30 passages in androgen-depleted medium, these cells progressed proliferation. LNCaP cells express large amounts of AR mRNA and to 104-1 cells, which exhibited much lower proliferative sensitivity to 0.1 protein. The levels of AR mRNA but not protein appear to be under IIMRISSI. After another 20-30 passages, LNCaP 104-1 cells gave rise to moderate negative control by androgen (7-9). ARs in LNCaP cells 104-R cells, which proliferated rapidly without additional androgen. Pro possess a mutation in the androgen binding domain which alters the liferation of 104-R cells was induced 2-fold by 0.01 UMR1881 but was specificity of ligand binding and steroid-induced transactivation repressed by 0.1 MMR1881 and above. Thus, androgen induction and (10, 11) and is probably responsible for the aberrant proliferative repression of proliferation could be seen at lower concentrations of an response of LNCaP cells to antiandrogens (12-14). drogen as the cells progressed. During the transition of 104-S cells to 104-R cells, the androgen receptor mRNA level increased 2.5-fold whereas We were interested in determining whether changes might occur in the androgen receptor protein level increased 15-fold in the absence of the proliferative response of LNCaP cells to androgen when cells are androgen. Androgen receptor transcriptional activity, measured by andro cultured over a long period of time in androgen-depleted medium and gen induction of prostate-specific antigen mRNA and chloramphenicol if so, to identify the accompanying molecular changes. If androgen acetyltransferase activity in transfected cells, increased up to 20-fold dur sensitivity could be altered in a clonal isolate of androgen-sensitive ing the progression. LNCaP cells, therefore, appear to be able to adapt to LNCaP cells by long-term androgen deprivation, it should be helpful reduced androgen availability by increasing their sensitivity to androgen, in the study of molecular and/or cytogenetic changes which occur raising questions concerning the therapeutic strategies used against pros concomitantly with the observed changes in proliferative response. A tate cancer. Androgen induction of c-myc expression in 104-R cells occurred at a 10-fold lower concentration (0.01 nvi) than in 104-S cells demonstration that changes in specific gene expression take place over (0.1 IIMI.In all stages, cell proliferation and c-myc expression were re time in a clonal isolate would support the idea that prostate tumor cells pressed by androgen at a high concentration (20 IIMi, but the repression of are able to adapt to lowered androgen concentration in their environ cell proliferation was blocked by retroviral overexpression of c-myc. ment and that androgen-independent cells do not necessarily arise during androgen ablation therapy through selection of a preexisting, androgen-independent subpopulation (15). The use of a clonal subline also simplifies the interpretation of androgen response in a heterog- INTRODUCTION enous parental cell population. We were also particularly interested in Prostate cancer is currently the most common malignancy and is the changes in the expression of c-myc, because expression of this critical second leading cause of cancer death among males in the United growth-related protooncogene has been reported to be negatively States (1). The mechanism of prostatic carcinogenesis and tumori- regulated by high concentrations of androgen and positively correlated genesis likely involves a multistep progression from precancerous with proliferation in LNCaP cells (16, 17). Deregulated expression of cells to cells which proliferate and metastasize. The growth and de c-myc has been observed in a variety of neoplasias and has been velopment of prostate cancer appears to be androgen dependent ini implicated in neoplastic development (for reviews, see Refs. 18-21). tially, making it vulnerable to androgen ablation and antiandrogen Elevated c-myc expression has been observed in benign prostatic therapies (2). Most often, however, prostate cancer cells lose androgen hyperplasia and prostatic carcinoma (22, 23). In addition to having a dependency and responsiveness during the progression to malignant role in the induction of cell proliferation, c-myc also appears to func stages, and tumor cells which are resistant to endocrine therapy ulti tion under certain circumstances (overexpression during growth factor mately proliferate. While loss of AR3 expression may accompany loss withdrawal) in apoptosis (24-27). of androgen dependency and responsiveness (3), other prerequisite In this paper, we report on the results of experiments with an androgen-responsive LNCaP 104 clonal subline subcultured in andro Received 10/23/93; accepted 1/28/94. gen-depleted medium for about 60 passages. As the 104 cells pro The cosls of publication of this article were defrayed in part by Ihe payment of page gressed, they became more sensitive to low concentrations of andro charges. This article must therefore be hereby marked advertisement in accordance with gen, and proliferative response of this subline to R1881 at 0.1 UM 18 U.S.C. Section 1734 solely to indicate this fact. 1This work was supported by NIH Grams DK41670, DK 37694. and CA 58073. concentration changed from positive to negative. LNCaP 104 cells 2 To whom requests for reprints should be addressed, at The Ben May Institute, Box appeared to adapt to lowered androgen levels by increasing AR ex MC6027, The University of Chicago. 5841 South Maryland Avenue, Chicago, IL 60637. 1The abbreviations used are: AR. androgen receptor: RI881, 17ß-hydroxy-17-methyl- pression and transcriptional activity. These results show that androgen estra-4.°..ll-trien-3-one; ß;-MG.ß2-microglobulin: DMEM, Dulbecco's modified Eagle's regulation of c-myc expression may play an important role in the medium: DHT. Sor-dihydrotestoslerone; FBS. fetal bovine serum; CS-FBS, dextran-coated charcoal-stripped FBS; CAT, chloramphenicol acetyltransferase: PSA, prostate-specific apparent androgen sensitivity of prostate cancer cells. These findings antigen: cDNA, complementary DNA; PBS, phosphate-buffered saline; SDS-PAGE, so additionally suggest that in designing a therapeutic approach to the dium dodccyl sulfate-polyacrylamide gel electrophoresis; MMTV-LTR. mouse mammary later stages of prostate cancer that show apparent loss in androgen- tumor virus long terminal repeat; RSV LTR, Rous sarcoma virus long terminal repeat; DMNT, mibolerone or 7a,17a-dimethyl-19-nortestosterone; EGF. epidermal growth dependent proliferation, it is necessary to consider the underlying factor. mechanism and the usefulness of hormonal intervention. 1566 ANDROfiEN RECEPTOR ACTIVITY IN PROSTATE CANCER CELLS MATERIALS AND METHODS untreated FBS was equivalent to the activity of 0.05 HMR1881. LNCaP 104-R cells derived from 104-1 cells were also maintained in DMEM supplemented Materials. The LNCaP cell line was obtained from Dr. T. M. Chu at the with 10% CS-FBS. One passage consisted of 7 days of growth to confluency, Roswell Park Memorial Institute (Buffalo, NY), and the androgen-insensitive starting from a 1:9 dilution of a trypsinized cell suspension. Responsiveness of prostale cancer PC-3 cell line was obtained from American Type Tissue Col cell proliferation to various concentrations of R1881 was measured by deter lection. [a-32P]dCTP (3000 Ci/mmole), [a-12P]UTP (800 Ci/mmole), and the mining the cell density of trypsinized cell suspensions with a hemocytometer Multiprime Random Primer Kit were purchased from Amersham. Restriction after 6 days of growth. Cells were plated in triplicate in 12-well dishes (3 X endonucleases and other enzymes were purchased from Bochringcr Mann IO4 cells/well) in DMEM supplemented with 10% CS-FBS in the presence or heim, Stratagcne, BRL, and Promcga. AN-21, a polyclonal rabbit antibody absence of R1881. In instances of cell clumping, pelleted cells were treated raised against a 21-amino acid peptidc corresponding to the NH;-terminus of with a solution of 0.4% Nonidet P-40, 50 min Tris, and 5 mM MgCK, pH 7.5, human and rat AR (28, 29) was prepared in this laboratory by Drs. Ching Song and the density of free nuclei was determined with a hemocytometer. LNCaP and Richard Hiipakka using methods described by Tsang and Wilkins (30). An 104 sublines were infected with retrovirus generated from the pMV7 retroviral enhanced chemiluminescence detection kit was purchased from Amersham. A vector (32) using the procedures of Brown and Scott (33). The pMV7 vector lucifcrase assay system kit was purchased from Promega. R1881 was from contains the gene encoding ncomycin phosphotransferase, which confers se New England Nuclear, and other steroids were from Steraloids. The MMTV- lectable resistance to geneticin (G418; Gibco). A 1.5-kilobase ¿V»RI-////idlII CAT vector was derived from pMSG-CAT (Pharmacia) by the excision of the human c-myc cDNA fragment was inserted into £roRI-////idlll-digested 2-kilobase BamW fragment containing the gpt gene (11).