Vitamin D metabolism and effects on pluripotency genes and cell differentiation in testicular germ cell tumors in vitro and in vivo

Abstract

Testicular germ cell tumors (TGCTs) are classified as either seminomas or nonseminomas. Both tumors originate from carcinoma in situ (CIS) cells, which are derived from transformed fetal gonocytes. CIS, seminoma, and the undifferentiated embryonal carcinoma (EC) retain an embryonic phenotype and express pluripotency factors (NANOG/OCT4). Vitamin D (VD) is metabolized in the testes, and here, we examined VD metabolism in TGCT differentiation and pluripotency regulation. We established that the VD receptor (VDR) and VD-metabolizing enzymes are expressed in human fetal germ cells, CIS, and invasive TGCTs. VD metabolism diminished markedly during the malignant transformation from CIS to EC but was reestablished in differentiated components of nonseminomas, distinguished by coexpression of mesodermal markers and loss of OCT4. Subsequent in vitro studies confirmed that 1,25(OH)(2)D(3) (active VD) downregulated NANOG and OCT4 through genomic VDR activation in EC-derived NTera2 cells and, to a lesser extent, in seminoma-derived TCam-2 cells, and up-regulated brachyury, SNAI1, osteocalcin, osteopontin, and fibroblast growth factor 23. To test for a possible therapeutic effect in vivo, NTera2 cells were xenografted into nude mice and treated with 1,25(OH)(2)D(3), which induced down-regulation of pluripotency factors but caused no significant reduction of tumor growth. During NTera2 tumor formation, down-regulation of VDR was observed, resulting in limited responsiveness to cholecalciferol and 1,25(OH)(2)D(3) treatment in vivo. These novel findings show that VD metabolism is involved in the mesodermal transition during differentiation of cancer cells with embryonic stem cell characteristics, which points to a function for VD during early embryonic development and possibly in the pathogenesis of TGCTs.

Figures

Figure 1

IHC expression of VDR in fetal germ cells. (A) IHC expression in serial sections from fetal testis 16 GW of OCT4, MAGE-A4, VDR, and negative control. (B) IHC from 24 GW. Arrowheads mark germ cells. Bar corresponds to 20 µm.

Figure 2

Expression of VDR and VD-metabolizing enzymes in normal testis and TGCTs. (A) IHC detection of proteins in normal and CIS tubules (marked with asterisk). (B) IHC expression in CIS with no counterstaining. (C) IHC expression in seminoma. (D) IHC expression in EC, arrowhead indicates OCT4-negative elongated EC cells. (E) IHC expression in mixed nonseminoma, arrowhead indicates OCT4-positive EC cells. (F) IHC expression in yolk sac tumor, arrowhead indicates α-fetoprotein-positive cells (positive control). (G) IHC expression in choriocarcinoma, and positive control shows hCG-producing cells. All control samples are negative control with no primary antibody and counter-stained with Mayer except for F (positive control α-fetoprotein) and G (positive control hCG), in which negative control is placed in the upper left corner. C–G: serial sections. Bar corresponds to 20 µm.

Figure 3

Gene expression following treatment with 1,25(OH)2D3, RA, and DMSO in NTera2 and TCam-2 cells. (A) OCT4 and NAONOGexpression in NTera2 cells. (B) OCT4 and NANOG expression in TCam-2 cells. (C) Expression of CYP24A1, SNAI1, T(brachyury), RUNX2, BGLAP (osteocalcin),OPN, and FGF23 in NTera2 cells. (D) Effect of VDR antagonist ZK159222 on expression of pluripotency genes. All cells are treated with either 100 nM 1,25(OH)2D3, 10 µM RA, or DMSO for 18 days. Except for (D) where cells were treated 6 days. Expression is normalized to β2M and expression level at day 0. Values represent mean ± SD. All experiments are conducted in triplicates and have been repeated twice. Note different scales. *P < .01.

Figure 4

ICC expression of OCT4, VDR, CYP24A1, FGF23, gla (osteocalcin), OPN, and negative control in NTera2 and TCam-2 cells treated with DMSO, 100 nM 1,25(OH)2D3, or 10 µM RA for 18 days.

Figure 5

Influence of VD on NTera2 xenograft tumors. Graphs show total tumor burden for each mice tumor growth on the right [pretreated with 100 nM 1,25(OH)2D3 for 20 days] and left flanks (vehicle) until day 19. Seven mice received cholecalciferol-supplemented diet (1100 IU/kg) from the day of inoculation, whereas 28 mice received standard chow (600 IU/kg). At day 19, the 28 animals were randomized to four different treatment groups: 1) control received vehicle (0.1% ethanol in sterile saline) three times weekly i.p., 2) cisplatin (6 mg/kg) i.v. once weekly, 3) calcitriol (0.05 µg) i.p. three times weekly, and 4) fortified diet with cholecalciferol (1100 IU D3/kg diet). Values represent mean ± SEM. *P < .05.

Figure 6

Changes in histology, gene, and protein expression following treatment with 1,25(OH)2D3 in NTera2 xenograft tumors. (A) Changes in gene expression of OCT4, NANOG, CYP24A1, SNAI1, and Runx2. Expression is normalized to β2M and expression level at day 0. All experiments are conducted in triplicates and have been repeated twice. Note different scales. *P < .05. Changes in cellular protein expression (OCT4, SOX2, KI-67, and VDR) evaluated by IHC in the right- and left-sided tumors from animals treated in vivo with vehicle or 1,25(OH)2D3. Data presented as mean ± SEM. (B) Hematoxylin and eosin stainings of xenograft tumors from right and left flanks. NTera2 cells invaded the surrounding muscle and adipose tissue. Notice the aberrant morphology in a subpopulation of the left-sided NTera2 xenograft cells. R: right-sided tumor. L: left-sided tumor. (C) IHC on serial sections of OCT4, SOX2, CYP27B1, VDR, CYP24A1, osteocalcin (gla), and Ki-67 in representative NTera2 xenograft tumors.