Regulation of Transforming Growth Factor- b Type II Receptor Expression in Human Breast Cancer MCF-7 Cells by Vitamin D 3 and Its Analogues*

In view of the tumor suppressor role of the transforming growth factor- b (TGF b ) type II receptor (RII), the identification and characterization of agents that can induce the expression of this receptor are of potential importance to the development of chemoprevention approaches as well as treatment of cancer. To date, the identification of exogenous agents that control RII expression has been rare. We demonstrated that proliferation of MCF-7 early passage cells (MCF-7 E), which express RII and are sensitive to TGF b growth inhibition activity, was significantly inhibited by vitamin D 3 and its analogue EB1089. In contrast, proliferation of MCF-7 late passage cells (MCF-7 L), which have lost cell surface RII and are resistant to TGF b , was not affected by these two compounds. TGF b -neutralizing antibody was able to block the inhibitory effect on MCF-7 E cells by these compounds, indicating that treatment induced auto-crine-negative TGF b activity. An RNase protection assay showed approximately a 3-fold induction of the RII mRNA, while a receptor cross-linking assay revealed a 3–4-fold induction of the RII protein. In contrast, there was no change in either RII mRNA or protein in the MCF-7 L cells. Transforming growth

Transforming growth factor-␤ (TGF␤) 1 comprises a family of hormone-like polypeptides that affects cell growth, adhesion, and differentiation (1). They act as growth inhibitors for most epithelial cells and some cancer cells. Two pathways are primarily involved in mediating effects of TGF␤ on cell growth and differentiation. One pathway involves blockade of cell cycle transit, while the other involves alteration of the extracellular matrix environment.
TGF␤s elicit their effects by binding to cell surface receptors. Three major types of receptors have been shown to be present in most TGF␤-responsive cell lines. They are designated as type I (RI), type II (RII), and type III (RIII), respectively. RIII is a 280 -330-kDa glycoprotein that has no functional signaling domain but rather serves as a ligand storage protein and presents TGF␤ to the signaling receptors (2). RI and RII, which are glycoproteins of ϳ55 and 85 kDa, respectively, form a hetero-meric receptor complex. Both are serine/threonine kinases, and each appears to be indispensable for TGF␤ signaling (3)(4)(5). The direct involvement of both RI and RII in conferring TGF␤ effects indicates that loss of either of the functional receptors would contribute to loss of autocrine TGF␤ activity. Loss of negative autocrine TGF␤ activity results in a growth advantage caused by an imbalance in positive and negative regulators, possibly leading to tumor formation and progression (6,7). Recent evidence has shown a loss of RII is often associated with the failure to respond to autocrine and exogenous TGF␤. We have previously demonstrated that re-expression of this receptor in an RII-deficient breast cancer cell line (late passage MCF-7) leads to restoration of TGF␤ sensitivity and reduced malignancy in athymic nude mice (6). In addition, it has been shown that mutational inactivation of RII occurs frequently in a subset of colon tumors with microsatellite instability (7), and reconstitution of RII expression by stable transfection also leads to reversal of malignancy in these cells (8). Others have noted that loss of RII expression is important in other types of malignancies (9 -15). These lines of evidence suggest that RII is a critical determinant for conferring TGF␤ tumor suppression as well as negative autocrine TGF␤ growth function. Consequently, agents that can induce RII expression would be valuable in the development of approaches for cancer treatment and prevention where receptor expression appears to be repressed, such as estrogen receptor-positive (ER ϩ ) breast cancer (16,17). To date, no such agents have been carefully characterized for their ability to induce RII. Although Cohen et al. (55) showed increased RII mRNA in a human neuroblastoma cell line after retinoic acid (RA) treatment, they were not able to detect cell surface RII. In addition, they failed to test for increased autocrine activity and increased responsiveness/ growth inhibition to TGF␤ after RA treatment. A study by Turley et al. (56) in RL human B lymphoma cells demonstrated increased RII protein levels following treatment with RA and vitamin E succinate. However, they did not investigate whether this correlated with increased levels of cell surface RII.
Development of effective therapeutic and preventive approaches for breast cancer remains an issue, since conventional treatment by antiestrogens such as tamoxifen often leads to resistance in estrogen receptor-positive tumors (18), and chemotherapy of estrogen receptor-negative tumors is even less effective (19). Since there is a high incidence of vitamin D receptors (VDRs) in human breast cancer tumors (21,22), vitamin D 3 is an appealing candidate as a new therapeutic agent. Like other steroid hormones, it mediates its effect through interaction of its nuclear receptor (VDR) with DNA-responsive elements in the target genes (20). Moreover, many breast cancer cell lines are responsive to vitamin D 3 antiproliferative effects both in vitro and in athymic mice (23). However, a major drawback for its clinical application is that the doses effective for suppressing tumor growth often cause hypercalcemia. Consequently, analogues have been developed to reduce the calcemic effects while increasing the potency of inhibition of proliferation (23,24). Two analogues, EB1089 and MC903, both of which are derived by modification of the C17 side chain of vitamin D 3 , have been shown to be effective against rat breast tumors in vivo (24) or as an antiproliferative agent when given topically for psoriasis as well as for cutaneous metastatic breast cancer (25). However, the mechanisms of vitamin D 3mediated growth inhibition and in particular its anti-tumor action remain largely unresolved.
In this report, we show a correlation between RII expression and vitamin D 3 inhibition in MCF-7 sublines that differ dramatically in their RII expression and hence their TGF␤ sensitivity as well. We hypothesized that vitamin D 3 's mechanism of inhibition might involve induction of TGF␤ autocrine activity through increased expression of RII. This hypothesis was confirmed by RNase protection assays showing approximately 3-fold induction of the RII mRNA and a 3-4-fold induction of cell surface RII protein. The increased inhibition by vitamin D 3 /analogues was blocked by TGF␤-neutralizing antibodies, indicating an induction of negative autocrine TGF␤ activity.
The use of an essential dietary nutrient with antiproliferative and anti-tumor properties represents an attractive approach for chemoprevention and/or therapy. This is particularly true of vitamin D compounds, since the high stress western style diet associated with colon and breast cancer is also associated with low levels of vitamin D and calcium (26). Thus, increased autocrine negative TGF␤ activity mediated by vitamin D 3 compounds in MCF-7 E cells may provide a novel mechanism for blocking malignant progression by chemopreventive approaches.

MATERIALS AND METHODS
Cell Cultures-MCF-7 E cells (passage number 150) were kindly provided by Drs. Robert J. Pauley and Herbert D. Soule from the Michigan Cancer Foundation. MCF-7 L cells were obtained from the ATCC and used at a passage number greater than 500. These cell lines were cultured in McCoy's 5A medium supplemented with 10% fetal bovine serum, pyruvate, vitamins, amino acids, and antibiotics. Working cultures were maintained at 37°C in a humidified atmosphere of 5% CO 2 .
Vitamin D 3 Compounds-1,25-(OH) 2 vitamin D 3 as well as its analogues EB1089 and MC903 were generous gifts from Dr. Lise Binderup of LEO Pharmaceutical Products (Ballerup, Denmark). Stock solutions were prepared in isopropyl alcohol at 4 mM. Serial dilutions were made in absolute ethanol and stored at Ϫ20°C protected from light. These diluted solutions were added to the experimental culture media at a final ethanol concentration of 0.1%. Control cells received 0.1% ethanol vehicle, which had no effect on cell proliferation.
DNA Synthesis Assay-[ 3 H]thymidine incorporation into DNA was measured as described previously to determine TGF␤ and vitamin D 3 sensitivity (6). Briefly, MCF-7 cells were seeded in 24-well tissue culture plates at a density of 1.5 ϫ 10 4 cells/well in 1 ml of medium. Various concentrations of compounds (1,25-(OH) 2 D 3 , EB1089, MC903, or TGF␤) were added after cell attachment (approximately 2 h). Following 4 days of incubation, cells received a 2-h pulse with [ 3 H]thymidine (7 Ci, 46 Ci/mmol, Amersham Pharmacia Biotech). DNA was then precipitated with 10% ice-cold trichloroacetic acid, and the amount of [ 3 H]thymidine incorporated was analyzed by liquid scintillation counting in a Beckman LS 7500 scintillation counter as described previously (6). To determine whether there is an increase in the inhibitory effects by TGF␤ 1 following EB1089 treatment, MCF-7 E cells, which are TGF␤-responsive, were plated as described above. Various concentrations of EB1089 plus 0.1 ng/ml of TGF␤ 1 were added after attachment. Cells were incubated and [ 3 H]thymidine incorporation was determined as described above.
TGF␤-neutralizing Antibody Assay-Cells were resuspended at a concentration of 1.5 ϫ 10 4 cells/ml and plated into 24-well tissue culture plates (1 ml/well) either untreated or in the presence of 10 g/ml TGF␤ 1 neutralizing antibody (R & D Systems) or control normal IgG. After 3 h of incubation, different concentrations of vitamin D 3 compounds were added as indicated. Cells were allowed to grow for 72 h without changing the media, followed by determination of [ 3 H]thymidine incorporation as described above.
RNA Analysis-RNase protection assays were performed to determine RII RNA expression levels after vitamin D 3 treatment. A 476-base pair fragment of the RII cDNA within the cytoplasmic region was obtained by polymerase chain reaction with the following primers: 5Ј-TGGACCCTACTCTGTCTGTG-3Ј and 5Ј-TGTTTAGGGAGCCGTCT-TCA-3Ј. The fragment was subcloned into a pBSK (Ϫ) plasmid (Stratagene, La Jolla) for making the RII riboprobes. In vitro transcription using T3 RNA polymerase yields antisense riboprobes that protect a 476-base pair RII fragment. RNase protection assays were performed as described previously (27). Briefly, exponentially growing cells were treated with EB1089 at 1 ϫ 10 Ϫ8 M for the indicated time periods. Cells were solubilized in guanidine thiocyanate, and total RNA was obtained by cesium chloride gradient ultracentrifugation (28). 40 g of total RNA was used for overnight hybridization with 32 P-labeled antisense riboprobes. Following RNase A and T 1 treatment, the protected doublestranded RNA fragments were heat-denatured at 95°C and analyzed by urea-polyacrylamide gel electrophoresis, and the radioactive probes were visualized by autoradiography. Actin was used as an internal control for normalizing the amount of sample loading.
Receptor Cross-linking-Simian recombinant TGF␤ 1 was purified as described (29) and iodinated by the chloramine T method (30). MCF-7 cells were seeded into 35-mm 2 tissue culture wells at a density of 10 5 cells/well. In the kinetic studies, exponentially growing cells were treated with various concentrations of the compounds for 24 h or with a single concentration for the indicated time periods. Cell monolayers were then incubated with 200 pM 125 I-labeled TGF␤ 1 at 4°C for 4 h followed by chemical cross-linking with disuccinimidyl suberate for 15 min (31). Labeled cell monolayers were solubilized in 200 l of 1% Triton X-100 with 1 mM phenylmethylsulfonyl fluoride. Equal amounts of cell lysate protein were separated by 4 -10% gradient SDS-polyacrylamide gel electrophoresis under reducing conditions and exposed for autoradiography.
Mink Lung Epithelial Cell Growth Inhibition Assay-MCF-7 E cells were plated into 100-mm 2 tissue culture dishes and allowed to reach 70 -80% confluency. The medium was then removed and replaced with 5 ml of fresh McCoy's 5A medium supplemented with pyruvate, vitamins, amino acids, and antibiotics (SM). The cells were then treated with EB1089 (1 ϫ 10 Ϫ8 M) or with vehicle only for 24 h. Following treatment, the conditioned medium was collected, and the indicated volumes were used to treat mink lung epithelial cells plated in 96-well tissue culture plates at a density of 1500 cells/well. In addition, a standard TGF␤ growth inhibition curve was generated by treating the cells with various concentrations of TGF␤ 1 . The mink lung epithelial cells were allowed to incubate for 3 days at which time the medium was removed and replaced by 100 l of fresh SM. Colonies were immediately visualized by staining with 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (Sigma) for 2 h. The stained cells were solubilized with Me 2 SO (dimethyl sulfoxide) (Mallinckrodt) and the relative cell numbers were then determined by the resultant absorbance at 595 nm.
Luciferase Assay-The TGF␤-responsive cyclin A promoter in tandem with a luciferase reporter construct (Ϫ133/Ϫ2) was used as described previously (32). The reporter construct (Ϫ133/Ϫ2) contains only the activating transcription factor site, which has been shown to be the site required to mediate down-regulation of cyclin A promoter activity by TGF␤ 1 in mink lung epithelial cells (32). MCF-7 E cells were transiently transfected with 30 g of luciferase reporter construct and 7 g of ␤-galactosidase plasmid by electroporation with a Bio-Rad gene pulser at 250 mV and 960 microfarads. Cells were plated into a six-well tissue culture plate and treated with TGF␤-neutralizing antibody (10 g/ml) or control normal IgG and allowed to attach for 3 h. Following attachment, cells were treated with EB1089 (1 ϫ 10 Ϫ8 M), while control cells were treated with vehicle only. At 51 h post-transfection, cells were harvested with 100 l of lysis buffer (Luciferase assay system, Promega). Luciferase activity was determined according to the manufacturer's instruction using a luminometer (Berthold Lumat LB 0501) and expressed as relative units after normalization to ␤-galactosidase.

RESULTS
TGF␤ Sensitivity of MCF-7 Cells-Inconsistent response of MCF-7 cells to TGF␤ has been observed in several laboratories (34,35), probably due to growth selection during long term passage of cultures. Having obtained both early (150) and late (Ͼ500) passage MCF-7 cells, we decided to first test whether they responded differently to TGF␤ (Fig. 1). MCF-7 E cells showed a significant dose-dependent inhibition by TGF␤ with an IC 50 of 0.2 ng/ml. In contrast, MCF-7 L cells demonstrated complete resistance to TGF␤ up to 25 ng/ml (Fig. 1). As described below, MCF-7 E cells expressed RII mRNA and protein in contrast to MCF-7 L cells, which had 5-fold less mRNA and no detectable cell surface protein.
Vitamin D 3 Sensitivity-Effects of 1,25-(OH) 2 D 3 and its analogues on cell proliferation of MCF-7 cells were investigated by assessing [ 3 H]thymidine incorporation following treatment by these compounds as described under "Materials and Methods." MCF-7 E cells showed a dose-dependent inhibition by vitamin D 3 with an IC 50 of 5 ϫ 10 Ϫ8 M. In contrast, MCF-7 L cells were not affected by vitamin D 3 (Fig. 2A). Vitamin D 3 analogues EB1089 and MC903 demonstrate similar growthinhibitory patterns (Fig. 2, B and C). The overall potency of growth inhibition by EB1089 was approximately 2 orders of magnitude higher than vitamin D 3 . MCF-7 E cells showed an IC 50 of 2.5 ϫ 10 Ϫ10 M, and MCF-7 L cells did not respond to EB1089 treatment up to 1 ϫ 10 Ϫ7 M.
TGF␤ Autocrine Activity-The correlation between TGF␤ and vitamin D 3 sensitivity suggested that vitamin D 3 may function through increasing TGF␤ autocrine-negative activity in MCF-7 E cells. To test this hypothesis, TGF␤-neutralizing antibodies were used to determine whether they were capable of blocking the growth inhibition induced by these compounds (Fig. 3). At 10 g/ml, TGF␤ 1 -neutralizing antibody reversed the inhibitory effect of vitamin D 3 and its analogues, generating an approximately 60% increase in DNA synthesis as compared with the normal chicken IgG treatment. In contrast, MCF-7 L cells did not respond to TGF␤ 1 -neutralizing antibody, indicating a lack of induction of autocrine TGF␤ activity. These results indicate that the growth-inhibitory mechanism of vitamin D 3 involves induction of TGF␤ autocrine-negative activity in MCF-7 E cells.
Alteration of RII Expression-Increased autocrine TGF␤ activity could result from enhanced expression of TGF␤ isoforms and/or their receptors. To test these possibilities, RNase protection assays were initially carried out on MCF-7 E cells to determine whether there were alterations of TGF␤ isoform expression upon treatment with vitamin D 3 compounds. MCF-7 E cells expressed high levels of TGF␤ 1 mRNA and low levels of TGF␤ 2 and TGF␤ 3 mRNA. Treatment with EB1089 did not generate altered mRNA expression for any of the three TGF␤ isoforms (data not shown). In addition, enzyme-linked

FIG. 3. Effect of TGF␤-neutralizing antibody on DNA synthesis of vitamin D 3 -or EB1089-treated MCF-7 cells. MCF-7 E (f) and MCF-7 L cells (3)
were plated at a density of 1.5 ϫ 10 4 cells/ well. The addition of TGF␤-neutralizing antibodies and vitamin D 3 compounds was performed as described under "Materials and Methods." The results were expressed as the percentage increase in DNA synthesis relative to their respective control antibody-treated cells. Each point represents the mean Ϯ S.E. of triplicate determinations. immunosorbent assay analysis of the conditioned medium showed no significant increase in the levels of activated TGF␤ 1 protein (data not shown). Since the levels of activated TGF␤ cannot be determined by enzyme-linked immunosorbent assay analysis, a growth inhibition bioassay on mink lung epithelial cells was performed. The condition medium from EB1089treated and -untreated MCF-7 E cells was added to mink lung epithelial cells as described under "Materials and Methods." After exposure to either treated or untreated conditioned medium, no significant difference in growth inhibition was observed in the mink lung epithelial cells (Fig. 4). These results indicate that EB1089 treatment did not alter the activation of secreted growth and/or inhibitory peptides from MCF-7 E cells, one of which is likely to be TGF␤ 1 , as demonstrated by enzymelinked immunosorbent assay analysis. Taken together, these results suggest that the enhanced TGF␤ autocrine activity by vitamin D 3 did not result from modulation of ligand expression or activation.
The other possibility for increased autocrine TGF␤ activity upon treatment with vitamin D 3 compounds was induction of receptor expression; therefore, we determined whether vitamin D 3 analogue treatment modulated expression of RII mRNA. EB1089 (10 Ϫ8 M) was utilized to determine the kinetic effects on RII expression. MCF-7 E cells expressed 5-fold higher RII mRNA than MCF-7 L cells. After exposure to EB1089, a 3-fold increase in the RII mRNA levels of MCF-7 E cells was observed. In contrast, no significant modulation was noted for the MCF-7 L cells after exposure to EB1089 (Fig. 5). EB1089 treatment did not effect the levels of RI or RIII mRNA (data not shown).
The increase in MCF-7 E RII mRNA expression led us to examine whether this corresponded to an increase in cell surface RII protein. This was tested by receptor cross-linking with 125 I-labeled TGF␤ (Fig. 6A) EB1089 resulted in a 3-4-fold induction of RII, while expression levels of RI remained relatively unchanged (Fig. 6A). Compared with the MCF-7 E cells, no cell surface RII protein was detected in MCF-7 L cells. Treatment with the vitamin D 3 compounds did not result in any change in receptor expression of these cells (Fig. 6B). To determine if the induction of RII protein in MCF-7 E cells was also time-dependent, a kinetic study was performed. Receptor cross-linking assays revealed a time-dependent increase of cell surface RII protein (Fig. 7A).

Induction was detected as early as 8 h with a 3-4-fold increase after 24 h of treatment.
Responsiveness to Autocrine TGF␤-To evaluate whether RII induction by treatment with EB1089 enhanced autocrine TGF␤ sensitivity, TGF␤-dependent promoter activity was analyzed using the TGF␤-responsive cyclin A luciferase reporter construct (32). TGF␤ induces down-regulation of cyclin A promoter activity but requires a functional TGF␤ type I and II receptor complex (32,33). Thus, an increase in functional receptor levels would result in enhanced down-regulation of cyclin A promoter activity. The cyclin A reporter construct (Ϫ133/Ϫ2) contains only the activating transcription factor site, which has been shown to mediate down-regulation of cyclin A promoter activity by TGF␤ 1 in mink lung epithelial cells (32). This reporter construct was transiently transfected into MCF-7 E cells, which are sensitive to TGF␤, followed by treatment with TGF␤neutralizing antibody and EB1089 as described under "Materials and Methods." As expected, a decrease in luciferase activity was induced in MCF-7 E cells following treatment with EB1089. TGF␤-neutralizing antibody reversed the decrease in cyclin A luciferase activity by EB1089, increasing it by approximately 70%. TGF␤-neutralizing antibody alone had no significant affect on cyclin A luciferase activity (Fig. 8) The fact that EB1089 did not increase expression or activation of any of the three TGF␤ isoforms indicates that enhanced responsiveness to autocrine TGF␤ after EB1089 treatment is due to the increased expression of RII. DISCUSSION TGF␤ signaling requires a heteromeric assembly of its two Ser-Thr kinase receptors, designated RI and RII, respectively (36). A recent model illustrating the physical and functional interactions between the two receptors proposes that upon ligand binding, the constitutively active RII recruits RI and transphosphorylates the RI, which subsequently initiates downstream cytoplasmic events (37). Defects in expression of either receptor would contribute to loss of response to exogenous as well as endogenous TGF␤. In particular, loss and/or lack of autocrine TGF␤ response plays a major role in enhancing tumor progression. As was demonstrated previously, anti-sense TGF␤ transfection in two early malignant colon carcinoma cell lines eliminated autocrine negative TGF␤ activity but did not block response to exogenous TGF␤ (27,38). The TGF␤ antisense-transfected cells showed increased tumor growth and incidence in athymic nude mice, indicating that autocrine TGF␤ plays a key role in blocking tumor progression. Reestablishment of autocrine TGF␤ responsiveness leading to decreased tumorigenicity in cells with deficient TGF␤ receptor function was achieved by stable transfection of RII in breast cancer MCF-7 L cells (6) and in the human colon carcinoma cell line HCT116 (8). Based on these studies, agents that can control the expression of TGF␤ receptors may have therapeutic implications. In particular, agents that can enhance the expression level of RII in cells where it appears to be repressed may be an effective chemopreventive approach. To date, no such agents have been carefully and fully characterized for their ability to induce RII and subsequently enhance autocrine- negative TGF␤ activity. Characterization of these agents should lead to a better understanding of TGF␤-mediated growth inhibition and its anti-tumor effects.
In the present study, we demonstrated for the first time that an increase in autocrine TGF␤ function by the active metabolite of vitamin D 3 is solely due to an increase in RII expression. However, the effective dose of active vitamin D 3 , which induces negative autocrine TGF␤ activity would also cause hypercalcemia, leading to unwanted side effects. To overcome this problem, analogues such as EB1089 have been developed that have increased potency and reduced hypercalcemic effects (39). We have shown here that the analogue EB1089 has similar effects to the parental compound at lower concentrations, which make it an attractive and potential chemopreventive agent. Thus, vitamin D 3 and its analogues can inhibit malignant cell growth through a novel mechanism of induction of negative autocrine TGF␤ activity.
A number of studies have reported that expression of RI and RII protein can be regulated by factors such as cell density (40,41); exposure to parathyroid, adrenal, or androgenic hormones (42)(43)(44); and TGF␤ (45). However, these agents do not readily lend themselves to chemopreventive approaches. Moreover, these studies were restricted to cell surface analysis utilizing 125 I-TGF␤ in receptor cross-linking and did not determine biological effects with respect to potential autocrine activity changes. In this study, we demonstrated a 3-fold increase in steady state RII mRNA levels (Fig. 4) by vitamin D 3 treatment, which correlated with a 3-4-fold increase in cell surface RII protein (Figs. 5 and 6). This suggests that translational modifications were unlikely to be responsible for up-regulation of the RII protein. Induction of RII mRNA by vitamin D 3 may involve transcriptional and post-transcriptional mechanisms. Vitamin D 3 association with VDR can either increase the affinity of VDR binding to its target DNA sequence or cause conformational changes in the receptor leading to alterations in gene activation (46). VDR proteins were detected by Western analysis in both MCF-7 E and MCF-7 L cells (data not shown). The VD-VDR complex in combination with other steroid receptors could be directly involved in stimulation of the RII promoter activity or act indirectly by increasing the quantity or activity of related transcriptional activators. Examination of the recently characterized RII promoter region (47) did not reveal sequences analogous to the well established vitamin D 3 -responsive elements, indicating that induction of RII mRNA in MCF-7 E cells by vitamin D 3 was unlikely to be a direct effect. The lack of RII induction in MCF-7 L cells suggests that the RII gene might be suppressed by factors or mechanisms that were not present in MCF-7 E cells and that certain transcriptional factors that were essential to activation of the VDR signaling pathway might be deficient in MCF-7 L cells. Unraveling these mechanisms may lead to novel approaches for reactivation of the RII tumor suppressor gene.
Modulation of TGF␤ expression or secretion by vitamin D 3 has been shown in keratinocytes, chondrocytes, rat prostatic epithelial cells, and one human breast cancer cell line BT-20 (48 -51). Danielpour (51) was able to demonstrate in a nontumorigenic rat prostate epithelial cell line that induction of TGF␤ autocrine activity by vitamin D 3 was mediated by increases in all three isoforms of TGF␤. In addition, other steroid hormones have been shown to increase activation of latent TGF␤ while not affecting total levels (52,53). Interestingly, modulation of TGF␤ levels or its activation by vitamin D 3 compounds was not observed in this study of this strain of human breast cancer cell line (MCF-7 E). A potential difficulty with chemopreventive approaches involving TGF␤ ligand induction rather than receptor induction resides in the tumor- enhancing effects associated with TGF␤ overexpression, such as angiogenesis and immunosuppression (54). However, EB1089 may offer an advantage in that induction of autocrinenegative TGF␤ activity occurs through RII and not its TGF␤ ligand. Enhancement of autocrine-negative TGF␤ activity, without the increase in TGF␤ ligand and the potential tumorenhancing effects associated with it, makes the use of these compounds an attractive approach by offering a potential novel mechanism for cancer prevention and/or therapy.
In addition to vitamin D 3 , other related members of the steroid hormone family have been shown to modulate TGF␤ receptor expression. In human neuroblastoma cells, RA increased RII mRNA levels and RI protein as well as increasing expression and secretion of TGF␤ 1 (55). However, cell surface RII was undetectable by receptor cross-linking in this study. The up-regulation of TGF␤ 1 and the TGF␤ receptors occurred only in the neuroblastoma cell line that was responsive to RA-induced growth arrest. RA treatment of RL human B lymphoma cells induced a 2-fold increase in total RII protein.
However, the study failed to examine whether that correlated to increased cell surface RII (54). In addition to RA, vitamin E succinate, which demonstrated potent growth inhibition, induced RI, RII, and TGF␤ proteins but did not affect their mRNA levels in RL human B lymphoma cells (56). Treatment of the lymphoma cells with TGF␤ 1 -neutralizing antibodies could partially block the growth inhibitory functions of vitamin E succinate and RA, indicating that treatment induced a TGF␤ autocrine-negative loop. However, this study did not examine cell surface receptor levels; thus, the increase in negative TGF␤ autocrine activity could be due to increased ligand levels. Both of these studies demonstrated the ability of other agents to induce RII mRNA or protein levels but failed to either detect or investigate RII cell surface levels. In addition, these compounds also enhance ligand expression, which may have adverse effects on surrounding tissue. The observation that vitamin D 3 induced autocrine negative TGF␤ activity through increased cell surface RII and not TGF␤ ligand may prove to be of significance in breast cancer therapy and/or prevention.