Testicular Maturation Arrestto Testis Cancer: Spectrum of Expression of theVitamin D Receptor and Vitamin D Treatment In VitroAjay K. Nangia,* Oya Hill, Maudine D. Waterman, Catherine E. B. Schwender and Vince MemoliFrom the Section of Urology and Department of Pathology (MDW, VM), Dartmouth-Hitchcock Medical Center, Lebanon,New Hampshire
Purpose: We investigated the qualitative distribution of vitamin D receptor in human testis pathologies and performed anin vitro study of vitamin D receptor expression in a human testis cancer cell line model.Materials and Methods: Qualitative immunohistochemical analysis of vitamin D receptor in testis tumors, normal testisand specimens from infertile patients was performed. The human embryonal carcinoma cell line NT2/D1 (American TypeCulture Collection, Manassas, Virginia) was cultured. Vitamin D receptor expression was examined by Western immunoblotanalysis after incubating the cells with 250 to 800 nM vitamin D, 10 to 70 nM testosterone, 2 nM calcium or a combinationof the 3 products.Results: Negative controls, synctiotrophoblasts and interstitial stroma did not stain positive for vitamin D receptor.Spermatogenic, Sertoli’s, Leydig and tumor cells stained positive in all specimens. Embryonal carcinoma demonstrated morenuclear and cytoplasmic staining than other tumors. Vitamin D receptor expression was seen at 50 kDa in the cell line.Sequential concentrations of vitamin D increased vitamin D receptor expression intensity. Simultaneous addition of vitamin D andtestosterone decreased the vitamin D receptor signal, as did testosterone alone. Delayed administration of vitamin D 5 hours aftertestosterone showed the return of vitamin D receptor expression. A combination of calcium, testosterone and vitamin D showeddecreased or no vitamin D receptor expression. Calcium alone increased vitamin D receptor expression at later passages.Conclusions: To our knowledge this is the first description of vitamin D receptor in different primary testis pathologies andin an embryonal carcinoma cell line. The in vitro model showed that vitamin D receptor is an active receptor and it is induciblewith the addition of vitamin D. Testosterone may be important for vitamin D receptor down-regulation. Calcium may be animportant co-factor in vitamin D receptor expression.
Vitamin D in its active form (1,25-dihydroxycholecal-ciferol) is involved with the calcium mobilization/metabolism pathway through interaction with VDR.
Vitamin D binds to VDR and then heterodimerizes with reti-noid X receptor and binds to the vitamin D response elementlocated in the promoter regions of target genes.1 VDR interactswith transcription factors, such as co-activator proteins, andtranscription integrators, eg calcium binding proteins.2
VDR has been identified in homogenates and nuclearmatrix preparations from human testis tissue.3,4 VDR isseen in smooth muscle of the epididymis, spermatogonia andSertoli’s cells in rodents.5,6 These findings suggest that vi-tamin D and VDR may have an important genomic andnongenomic role in the production and function of sperm.
A possible common pathophysiological pathway betweentestis tumors and spermatogenic arrest is thought to exist.7
Submitted for publication November 20, 2006.Study received institutional review board approval.Supported by the Harme’s Scholarship from the Department of
Surgery, Dartmouth-Hitchcock Medical Center and an endowmentfrom the Smith Family in honor of their son, who had testis cancer.
* Correspondence: Section of Urology, Dartmouth-HitchcockMedical Center, 1 Medical Center Dr., Lebanon, New Hampshire
Vitamin A, which like vitamin D is a member of the retinoidX receptor superfamily, has been shown to interact withchromosome 12 gene expression in the embryonal cancer cellline NT2/D1, which has also been shown to be important inmale gamete pluripotent embryonal stem cell differentiationand may explain the role in spermatogenesis.8 A study of thespectrum of testis pathologies may help explain vitaminD/VDR physiology in the testis and the role of vitamin D inmale infertility and testis cancer. To our knowledge this hasnot been studied before.
We defined the presence of VDR on cell types in a spec-trum of human testicular pathologies and studied the func-tionality of the receptor using the NT2/D1 embryonal testiscancer cell line in vitro. We hypothesized that VDR is foundin different cell types with normal and abnormal testicularfunction, and receptor distribution qualitatively differs inthe spectrum of testis pathologies. We also hypothesizedthat VDR is present and functional in the cell line and it isaffected by vitamin D, androgens and calcium.
MATERIALS AND METHODS
Institutional review board approval was obtained to identify
archived paraffin blocks of removed testis tissue.
Vol. 178, 1092-1096, September 2007Printed in U.S.A.
DOI:10.1016/j.juro.2007.05.009
TESTICULAR MATURATION ARREST TO TESTIS CANCER 1093
ImmunohistochemistryAutomated immunohistochemistry was performed using anI6000 Immunostainer (BioGenex, San Ramon, California).Reagents come in a pre-diluted kit for set times, includingperoxide blocking (10 minutes), normal blocking (10 minutes),avidin block (15 minutes), biotin block (15 minutes), primaryantibody (not in kit and added separately) (30 minutes), mul-tilink secondary antibody (20 minutes), streptavidin labeling(20 minutes), 3.3 diaminobenzidine (10 minutes) and hematox-ylin counterstaining (1 minute). Mouse anti-human VDRmonoclonal primary antibody (Diagnostic Systems Laborato-ries, Webster, Texas) was validated for positive stainingusing different titers of antibody against normal kidney.Negative control with secondary antibody only was per-formed. The best dilution of the antibody was 1:30, whichwas used for all sections in the study.
Cell CultureThe testicular embryonal cell carcinoma NT2/D1 line wascultured at a density of 5 � 106 viable cells per 75 cm2 T75flask in Dulbecco’s modified Eagle’s medium with 4 mML-glutamine, adjusted with 1.5 gm/l sodium bicarbonate, 4.5gm/l glucose (American Type Culture Collection) and 10%fetal bovine serum at 37C with 5% CO2.
VDR Ligand Binding AssayCultured cells were passaged 3 to 14 times. They weretrypsinized, seeded at a density of 5 � 105 cells per ml in T75tissue culture flasks and incubated with the different li-gands (vitamin D, testosterone and/or calcium) at 37C with5% CO2 for 24 hours.
Vitamin D (2 mM 1,25-dihydroxycholecalciferol) stock(Sigma®) was prepared in 100% ethanol. Testosterone stocksolution (10 mM) (Sigma) was dissolved in 2-hydroxypropyl-�-cyclodextrin 45% solution in water (Sigma). The ligandstimulus effect of vitamin D was tested on NT2/D1 cells atconcentrations of 250, 400, 500 and 800 nM. Cells were alsoincubated with 2 mM Ca�� (Sigma) and 10 to 70 nM tes-tosterone. Different combinations of vitamin D, and/or tes-tosterone and/or calcium were added to the cells simulta-neously for 24 hours before processing for immunoprecipitation(figs. 1 to 3).
Cells were harvested after the incubation period intomicrocentrifuge tubes, washed twice with cold PBS andspun down at 1,100 rpm for 5 minutes at 4C. Pellets were
FIG. 1. Immunoblot for VDR in human embryonal cell cancerNT2/D1 cell line at passage 4 shows vitamin D effect.
resuspended in cold 0.8 ml mild cell lysis buffer (CytoSignal,Irvine, California) with 16 �l protease inhibitors (10 � stocksolution tablets dissolved in 1 ml H2O) and 26.5 �l 1 �g/�lPefabloc® SC (50 � stock solution). Cell lysates were homog-enized by pipetting the solution through a 22 gauge hypo-dermic needle on ice. Lysates were allowed to solubilize for20 minutes on ice. Lysates were clarified by centrifugation at4C for 10 minutes at 12,000 � gravity.
Cell extract protein concentrations were determined bythe Bradford method using an MBA200 spectrophotometer(PerkinElmer®). After determining the cell protein concen-trations of the samples the cell lysates were prepared forimmunoprecipitation.
ImmunoprecipitationTo remove nonspecific protein binding protein A/G resinbeads were incubated with an equal volume of pre-immunerabbit serum (CytoSignal) (200 �l beads per 200 �l pre-immune serum) in microcentrifuge tubes with gentle agita-tion at room temperature for 30 minutes. Protein A/G resinbeads were centrifuged at 14,000 � gravity for 5 minutes atroom temperature. IgG bound protein beads were resus-pended to the original volume with mild lysis buffer. TheIgG bound A/G resin beads were added to ice-cold cell pro-
FIG. 2. Immunoblot for VDR in NT2/D1 cell line at passage 4 revealseffect of testosterone and vitamin D.
FIG. 3. Immunoblot for VDR in NT2/D1 cell line at late passage 14demonstrates effect of testosterone, calcium and vitamin D.
witedu
TESTICULAR MATURATION ARREST TO TESTIS CANCER1094
tein lysates and incubated with gentle shaking for 2 hours at4C. The mixture was centrifuged for 10 minutes at 14,000 �gravity at 4C. Supernatants (80 to 100 �l) from the lysateswere incubated overnight with 10 �l polyclonal rabbit anti-VDR IgG (Active Motif, Carlsbad, California). The immunecomplexes were then washed 3 times with mild lysis buffer(CytoSignal) and pelleted after centrifugation at 14,000 �gravity for 5 minutes. The samples were incubated with 25�l fresh protein A/G resin beads for 30 minutes at roomtemperature. The pellets were resuspended in 30 �l 2 � SDSsample buffer, boiled for 5 minutes at 95C and analyzed bySDS-polyacrylamide gel electrophoresis.
Negative control immunoprecipitation was performed us-ing water with A/G resin bead pellets. As positive controls,0.2 �l of a 10 � dilution of recombinant human VDR protein(Panvera/Invitrogen™) and 5 �g MG-63 osteoblastic cellnuclear extract (Active Motif) were used, which did notrequire immunoprecipitation with A/G beads. They werealso analyzed by SDS-polyacrylamide gel electrophoresis.
Western ImmunoblottingTreated and nontreated lysate immobilized protein A/Gresin beads (500 to 1,300 �g), controls, Bio-Rad™ markerand Magic Marker (Invitrogen) were electrophoresedthrough 10% SDS polyacrylamide gels and electroblottedonto nitrocellulose. Equal amount of test cell protein lysateswere loaded in each well per gel. The blots were incubatedovernight at 4C in PBS, pH 7, containing 10% NFDM and0.05% Tween-20 (PBST). VDR immunolabeling was detectedby incubating the blots for 2 hours at 37C with primary VDRantibody, VDR (Ab-1) rat monoclonal IgG clone 97A (EMDBioscience, San Diego, California), in PBST, 0.25% bovineserum albumin and 5% NFDM. The blots were then washedtwice in PBST and 1% NFDM solution, and once in PBSTalone. The blots were incubated for 90 minutes with AlexaFluor™ 488 goat anti-rat IgG as a secondary antibody at adilution of 1:1,000 in PBST and 0.25% bovine serum albu-
FIG. 4. Qualitative VDR immunochemistry of human testis tissuespermatogenesis stages and Leydig cells (arrow). B, Sertoli-cell-onlycell tumor cells stained lighter for VDR than normal Leydig cells. R
FIG. 5. A, seminoma showed VDR in nucleus and cytoplasm. Redu
min at room temperature. They were then washed twice asdescribed in PBST and 1% NFDM, solution and PBST alonefor the last wash, allowing a 5-minute incubation period atroom temperature for each wash. Blotted proteins on themembranes were determined by Typhoon scanning at anexcitation of 488 nM and an emission of 526 nM.
RESULTS
The archived pathology specimens examined were 2 embry-onal carcinomas, 2 seminomas, 1 choriocarcinoma, 1 Leydigcell tumor and 1 normal testis removed due to prostatecancer. Biopsies for infertility showed 3 with normal sper-matogenesis, 3 with hypospermatogenesis and 2 with Sertoli-cell-only. Spermatogenic, Sertoli’s, Leydig and tumor cellsstained for VDR with significantly more staining in Leydigcells in tissue with normal spermatogenesis (figs. 4 and 5).Interstitial stroma was negative for VDR. Leydig cells in theLeydig cell tumor showed less intensity than Leydig cells inspecimens with normal spermatogenesis (fig. 4, A and C).Cytoplasmic and nuclear staining was seen at all stages ofspermatogenesis and Leydig cells. Leydig cells stainingtended to clump next to the nucleus in a manner similar toGolgi apparatus.
Seminoma demonstrated light staining for VDR through-out the nucleus and cytoplasm with no significant mem-brane staining. Choriocarcinoma showed strong cytoplasmicstaining with no staining in the synctiotrophoblasts. Embry-onal cell tumor had strong nuclear and cytoplasmic mem-brane staining. Cytoplasm stained mildly (fig. 5).
Western blot analysis with monoclonal anti-VDR anti-body against recombinant human VDR protein confirmedantibody functioning with expression at a molecular weightof 50 kDa (data not shown), which is the known size of VDR.Immunoprecipitation in the MG63 osteoblastic cell line alsoconfirmed VDR at 50 kDa (positive control). No expression at50 kDa was seen with water (negative control) (fig. 1). Anti-
ormal spermatogenesis with cytoplasmic and nuclear VDR at allh VDR predominantly in Sertoli’s cell cytoplasm (arrow). C, Leydigced from �20.
rom �20. B, in choriocarcinoma there was no syncytiotrophoblast
. A, n
ced f
ated strong nuclear and cytoplasmic membrane staining. Reduced
TESTICULAR MATURATION ARREST TO TESTIS CANCER 1095
body binding was also noted at lower molecular weights inpositive and negative controls. Immunoprecipitation withmonoclonal anti-VDR antibody and early passage (passage4) embryonal cell carcinoma NT2/D1 demonstrated expres-sion at 50 kDa. Adding vitamin D at sequential concentra-tions of 250 to 800 nM to passage 4 cells demonstrated aqualitative increase in the expression of 50 kDa VDR protein(figs. 1 and 2). Adding 20 to 70 nM testosterone only to themedium showed a loss of expression of 50 kDa VDR protein.Simultaneous addition of 500 nM vitamin D and 20 to 70 nMtestosterone showed no 50 kDa protein expression. Adding500 nM vitamin D to the medium 5 hours after 70 nMtestosterone showed re-expression of VDR protein in earlypassage cells (fig. 2). These experiments were repeated 3 to4 times and findings were consistent.
Immunoprecipitation of the cells at passage 14 after theaddition of 250 nM vitamin D with or without the addition2 nM calcium demonstrated weak VDR protein expression.Adding 250 nM vitamin D, 20 nM testosterone and 2 nMcalcium simultaneously resulted in the loss of protein ex-pression. This was also seen after the addition of 20 nMtestosterone with or without 2 nM calcium. Adding 2 nMcalcium alone resulted in higher protein expression thanwith the addition of vitamin D (fig. 3).
DISCUSSION
The presence of the VDR in different cell types in the testishas been described previously.3–6 The role of the receptor isunknown. An animal model has shown that vitamin D defi-ciency decreased fecundity, which was reversible with cal-cium replacement only.9 This questioned the role of vitaminD beyond calcium homeostasis. To our knowledge the mech-anism of this interaction in spermatogenesis and testis tu-mor formation is unknown and it may occur in a nongenomicmanner. Before we could analyze this on the molecular level,we wanted to define the qualitative distribution of VDR inthe normal human testis to different human testicular pa-thologies.
The presence of VDR on Sertoli’s, Leydig and spermato-genic cells was confirmed. To our knowledge the presence ofVDR in all testis tumors is the first description of the VDRin this cancer type. This finding and VDR in tissues withspermatogenic problems were anticipated, considering thecommon cells of origin for these pathological conditions, butthe uniqueness of receptor distribution, especially in Leydigcells, embryonal carcinoma and synctiotrophoblasts, was notanticipated. The limitations of a qualitative study are rec-ognized but staining was highly standardized and deliberatewith an automated immunostainer and all slides were pre-pared and processed at the same time with the same re-agents and antibody batch. Semiquantitative analysis ofVDR staining was considered unreliable with current soft-ware. The significance of increased VDR staining in differ-ent pathological conditions is unclear but it may suggestthat the increased density is important for vitamin D func-tion in the different cell types. The dense staining for VDR inLeydig cells led us to investigate the role of testosterone andvitamin D in an in vitro model.
There are no readily available cell lines for studyingindividual cell types in the human testis. The in vitro testistumor model used in this study served as a surrogate to
study the possible functionality and interaction of VDR in
the testis. It was an ideal model since it was a humanembryonal tumor cell line (NT2/D1), the same tumor typeseen with the densest qualitative staining for VDR. VDRwas noted in the cell line at the known size of the receptor(50 kDa) and this was confirmed with 2 positive controls,including commercially available recombinant VDR proteinand MG63 osteoblastic cell nuclear extract. To our knowl-edge this is the first description of VDR in the embryonalNT2/D1 cell line. The antibody was noted to bind at lowermolecular weights with the positive and negative controls,and it confirmed a degree of nonspecific binding secondary tothe use of Immunobeads™ with the immunoprecipitationtechnique. This was not seen with recombinant VDR be-cause a pure Western blot without immunoprecipitation waspossible due to the high concentration of protein in theavailable product.
Manipulating medium conditions demonstrated in-creased VDR expression with increasing doses of added vi-tamin D. This suggested that the receptor was functionaland inducible or recruited by incubation with vitamin D.Testosterone at concentrations of 20 to 70 nM incubatedsimultaneously with vitamin D or alone resulted in a loss ofreceptor expression. This suggests that testosterone maydown-regulate VDR through co-factors, although it is un-clear how and why at this stage. To our knowledge suchinteraction in the human testis has not been studied previ-ously. Akerstrom et al noted that vitamin D induced cellularcalcium uptake was decreased when testosterone and vita-min D were added simultaneously to a mouse Sertoli’s cellline.10 Leman et al reported that vitamin D increased VDRand androgen receptor protein expression in prostate neo-natal cells.11 They also observed that dihydrotestosteronewith or without vitamin D increased androgen receptor andVDR DNA binding activity. This suggests that VDR bindingactivities may be regulated by androgens. In our study de-creased VDR expression was noted but we did not test forVDR DNA binding activity. Up-regulation of androgen re-ceptor with vitamin D has been demonstrated in an in vitrostudy using the LNCaP prostate cancer cell line.12 Interac-tion between androgens and VDR has also been shown inovarian and breast cells.13 Study of androgen receptor wasoutside of the objective of our study. The role of calciumremains unclear in view of effect on cell age/passage. Cal-cium may have served as a second messenger to recoverVDR expression from older cells.
Understanding vitamin D physiology in the testis mayhelp us understand the possible role in the treatment ofspermatogenic arrest to testis cancer, as demonstrated byvitamin A in the NT2/D1 cell line with an effect on chromo-some 12 expression and the role of vitamin D in the suppres-sion of prostate and other cancer growth.8,14–18
CONCLUSIONS
To our knowledge this is the first description of VDR in differ-ent primary testis pathologies and in an embryonal carcinomacell line. The in vitro model showed that VDR is an activereceptor and inducible with the addition of vitamin D. Testos-terone may be important in VDR down-regulation. Calciummay be an important co-factor in VDR expression. Furtherwork is needed to study the role of vitamin D in the testistissue/cancer gene and protein expression associated with
1. Jones G, Strugnell S and DeLuca H: Current understanding ofthe molecular actions of vitamin D. Physiol Rev 1998; 78:1193.
2. Jenster G, Spencer TE, Burcin MM, Tsai SY, Tsai MJ andO’Malley BW: Steroid receptor induction of gene transcrip-tion: a two-step model. Proc Natl Acad Sci U S A 1997; 94:7879.
3. Habib FK, Maddy SQ and Gelly KJ: Characterisation of recep-tors for 1,25-dihyrdoxyvitamin D3 in the human testis. JSteroid Biochem 1990; 35: 195.
4. Nangia AK, Butcher JL, Konety BR, Vietmeier BN andGetzenberg RH: Association of vitamin D receptors with thenuclear matrix of human and rat genitourinary tissues. JSteroid Biochem Molec Biol 1998; 66: 241.
5. Merke J, Hugel U and Ritz E: Nuclear testicular 1,25-dihyro-droxyvitamin D3 receptors in Sertoli cells and seminiferoustubules of adult rodents. Biochem Biophys Res Commun1985; 127: 303.
6. Osmundsen BC, Huang HFS, Anderson MB, Christakos S andWalters MR: Multiple sites of action of the vitamin D en-docrine system: FSH stimulation of testis 1,25-dihydroxy-vitamin D3 receptors. J Steroid Biochem 1989; 34: 339.
7. Crepieux P, Lecureuil C, Marion S, Kara E, Piketty V, MartinatN et al: A mechanistic overview on male infertility and germcell cancers. Curr Pharm Des 2004; 10: 449.
8. Clark AT, Rodriguez RT, Bodnar MS, Abeyta MJ, Cedars MI,Turek PJ et al: Human STELLAR, NANOG and GF3 genes
are expressed in pluripotent cells and map to chromosome
12p13, a hotspot for teratocarcinoma. Stem Cell 2004; 22:169.
10. Akerstrom VL and Walters MR: Physiological effects of 1,25-dihydroyxvitamin D3 in TM4 Sertoli cell line. Am J Physiol1992; 262: E884.
11. Leman ES, DeMiguel F, Gao AC and Getzenberg RH: Regula-tion of androgen and vitamin D receptors by 1,25 dihy-droxyvitamin D3 in human prostate epithelial and stromalcells. J Urol 2003; 170: 235.
12. Zhao XY, Ly LH, Peehl DM and Feldman D: Induction ofandrogen receptor by 1 alpha, 25 dihydroxyvitamin D3 and9-cis retinoic acid in LNCaP human prostate cancer cells.Endocrinology 1999; 140: 1205.
13. Escaleria MT, Sonohara S and Brentani MM: Sex steroidsinduced upregulation of 1,25–(OH)2 vitamin D3 receptors inT47D breast cancer cells. J Steroid Biochem Mol Biol 1993;45: 257.
14. Konety BR, Johnson CS, Trump DL and Getzenberg RH:Vitamin D in the prevention and treatment of prostatecancer. Semin Urol Oncol 1999; 17: 77.
15. Sandgren M, Danforth L, Plasse TF and DeLuca HF: 1,25-Dihydroxyvitamin D3 receptors in human carcinomas: apilot study. Cancer Res 1991; 51: 2021.
16. Frampton RJ, Omond SA and Eisman JA: Inhibition of humancancer cell growth by 1,25 hydroxyvitamin D3 metabolites.Cancer Res 1983; 43: 4443.
17. Sinella MJ, Kitareewan S, Mellado B, Sekula D, Khoo KS andDmitrovsky E: Specific retinoid receptors cooperate to sig-nal growth suppression and maturation of human embryo-nal carcinoma cells. Oncogene 1998; 16: 3471.
18. Giuliano CJ, Kerley-Hamilton JS, Bee T, Freemantle SJ,Manickaratnam R, Dmitrovsky E et al: Retinoic acid re-presses a cassette of candidate pluripotency chromosome 12pgenes during induced loss of human embryonal carcinoma
