Vitamin D3 Regulates the Formation and Degradation of Gap Junctions in Androgen-Responsive Human Prostate Cancer Cells

1α-25(OH)2 vitamin D3 (1-25D), an active hormonal form of Vitamin D3, is a well-known chemopreventive and pro-differentiating agent. It has been shown to inhibit the growth of several prostate cancer cell lines. Gap junctions, formed of proteins called connexins (Cx), are ensembles of cell-cell channels, which permit the exchange of small growth regulatory molecules between adjoining cells. Cell-cell communication mediated by gap junctional channels is an important homeostatic control mechanism for regulating cell growth and differentiation. We have investigated the effect of 1-25D on the formation and degradation of gap junctions in an androgen-responsive prostate cancer cell line, LNCaP, which expresses retrovirally-introduced Cx32. Connexin32 is expressed by the luminal and well-differentiated cells of normal prostate and prostate tumors. Our results document that 1-25D enhances the expression of Cx32 and its subsequent assembly into gap junctions. Our results further show that 1-25D prevents androgen-regulated degradation of Cx32, post-translationally, independent of androgen receptor (AR)-mediated signaling. Finally, our findings document that formation of gap junctions sensitizes Cx32-expressing LNCaP cells to the growth inhibitory effects of 1-25D and alters their morphology. These findings suggest that the growth-inhibitory effects of 1-25D in LNCaP cells may be related to its ability to modulate the assembly of Cx32 into gap junctions.


Introduction
The role of Vitamin D 3 , and its active hormonal form 1a-25(OH) 2 vitamin D 3 (1-25D), as an anti-neoplastic, pro-differentiating, and pro-apoptotic agent has been established in a wide variety of normal and malignant epithelial cells, including prostate cancer (PCA) [1][2][3][4]. The actions of 1-25D are mediated by binding to vitamin D receptor, one of the members of nuclear receptor superfamily, which is expressed in a wide variety of cells, including prostate. The vitamin D receptor heterodimerizes with the RXR receptor and binds to vitamin D receptor response element to alter gene expression [1]. Based upon the observation that PCA mortality rates in the U.S are inversely proportional to the geographically incident ultraviolet radiation exposure from the sun, and that ultraviolet light is essential for vitamin D 3 synthesis in the skin, a role for this vitamin in decreasing the risk of developing PCA has been suggested [5,6]. Numerous in vitro studies show consistent growth inhibitory and differentiation-inducing effects of vitamin D 3 on prostate carcinoma cells, and animal studies show that it not only reduces the incidence of PCA by acting as a chemopreventive agent but also suppresses metastasis [7][8][9][10].
Gap junction (GJ)s are ensembles of cell-cell channels that signal non-canonically, by permitting the direct exchange of small molecules (#1500Da) between the cytoplasmic interiors of contiguous cells [11]. The constituent proteins of GJs, called connexins (Cxs), are coded by 21 genes, which have been designated according to their molecular mass [12]. Cell-cell channels are bicellular structures formed by the collaborative effort of two cells. To form a GJ cell-cell channel, Cxs first oligomerize in the endoplasmic reticulum or the trans-Golgi network as a hexamer, called connexon, which docks with the connexon displayed on a contiguous cell [13,14]. Multiple lines of evidence now lend credence to the notion that cell-cell communication mediated by gap junctional channels is an important homeostatic control mechanism for regulating cell growth and differentiation and for curbing tumor promotion. For example, impaired Cx expression, or loss of GJ function, has been implicated in the pathogenesis of several types of cancers and diseases [15][16][17][18][19]. Also, mutations in several Cx genes have been detected in genetic disorders characterized by aberrant cellular proliferation and differentiation [13,20].
Our previous studies showed that the expression of Cx32, which is expressed by the luminal cells of the prostate, coincided with the acquisition of the differentiated state of the luminal cells [21,22]. Moreover, we documented that the progression of PCA from an androgen-dependent state to an invasive, androgen-independent state was characterized by the aberrant trafficking of Cx32 and/or impaired assembly into GJs [22][23][24]. Furthermore, our studies showed that forced expression of Cx32 into androgen-responsive human PCA cell line, LNCaP, retarded cell growth in vivo and in vitro [22]. We have also shown that in LNCaP cells expressing Cx32, formation and degradation of GJs were regulated by the androgens, which controlled the expression level of Cx32 posttranslationally by preventing its degradation by endoplasmic reticulum associated degradation (ERAD) [25]. Androgens are required to maintain the secretory (differentiation-related) function of the luminal epithelial cells of normal prostate as depletion of androgens by surgical or chemical means triggers apoptosis and/ or dedifferentiation of these cells [26][27][28][29]. Our recent studies have shown that retinoids, which also regulate the proliferation and differentiation of prostate epithelial cells [28,30], also enhance the assembly of Cx32 into GJs [31]. These studies lend credence to the notion that formation and degradation of GJs may be linked to the proliferation and differentiation of luminal prostate epithelial cells.
Like androgens and retinoids, vitamin D 3 is essential for the normal development of the prostate and has also been documented to modulate PCA progression [7,9]. Recent studies have shown that vitamin D suppressed prostatic epithelial neoplasia in Nkx3.1/PTEN transgenic mice [32]. Epidemiologic, cell culture, and clinical studies have implicated antitumor effects of 1-25D for PCA and it has been suggested to be a potent chemopreventive agent [3,4]. However, in contrast to colon cancer [1,33], the potential of effectiveness of 1-25D in the chemoprevention of PCA has remained controversial despite numerous studies in transgenic mouse models of PCA and its use in clinical trials [1,2]. Earlier studies, including ours, have shown that the growth-inhibitory and differentiation-inducing effects of chemopreventive agents might be related to their ability to enhance gap junctional communication [34][35][36][37][38]. The luminal cells of normal prostate express Cx32 and form large GJs and progression of PCA is accompanied by loss of ability to form GJs [22,23]. Formation of GJs has been implicated in maintaining the polarized and differentiated state of epithelial cells [39]. These studies prompted us to examine the effect of 1-25D on the assembly of Cx32 into GJs in androgenresponsive human PCA cell line LNCaP. Because 1-25D has been shown to increase the expression of AR in LNCaP cells [40], we rationalized that it might modulate androgen-regulated formation and degradation of GJs and affect growth of androgen-responsive PCA cells that express Cx32. By using androgen-responsive LNCaP cells, which express retrovirally-introduced Cx32 [25], we show that 1-25D enhances the assembly of Cx32 into GJs. Moreover, we further show that 1-25D prevents androgenregulated degradation of GJs post-translationally, independent of AR-mediated signaling. Finally, our findings show that formation of GJs sensitizes LNCaP cells to growth-inhibitory effects of 1-25D and alters their morphology.

Cell Culture
Androgen-responsive human PCA cell line, LNCaP, was grown as described [41,42]. LNCaP-32 cells, one of the several clones of LNCaP cells expressing retrovirally-transduced rat Cx32, and LNCaP-N cells, one of the several control clones selected in G418 after infection with the control retrovirus, have been described [25,31]. Parental LNCaP cells, hereafter referred to as LNCaP-P cells, were grown in RPMI containing 5% fetal bovine serum in an atmosphere of 5% CO 2 /95% air whereas LNCaP-N and LNCaP-32 cells were maintained in RPMI containing 5% fetal bovine serum containing G418 at 200 mg/ml as described [25,31]. Steroid-depleted (charcoal-stripped) serum and phenol-red-free RPMI were obtained from HyClone Laboratories (Salt Lake City, UT).

Stock Solutions
Synthetic androgen mibolerone (MB) and a natural androgen dihydro-testosterone (DHT), 1-25D, and Casodex (Bicalutamide) were purchased from BIOMOL (ENZO Life Sciences, Inc., Farmingdale, NY). Stock solutions of MB and DHT were prepared at 1 mM in ethanol and stored at 220uC in small aliquots protected from light. Stock solution of 1-25D (10 mM) was prepared in ethanol and stored in aliquots at 280uC protected from light. Stock solution of Casodex (10 mM) was prepared in DMSO and stored in aliquots at 220uC. They were appropriately diluted in the medium at the time of treatment. All experiments were performed in yellow light as described [36,37].

Androgen Depletion and Other Treatments
Cells were seeded in six well clusters with glass cover slips (1.5610 5 cells per well) and in 6-cm (2610 5 cells per dish) and 10cm dishes (3.5610 5 cells per dish) in 2, 4 and 10 ml complete medium, respectively. Cells were treated by replenishing with fresh medium containing various reagents at the desired concentration when they attained 50% confluence. Controls were treated with ethanol such that the final concentration of the solvent did not exceed 0.1%. When cells were to be grown under androgendepleted conditions, normal cell culture medium was replaced with androgen-depleted cell culture medium (phenol-red-free RPMI containing 5% charcoal-stripped serum). The controls received fresh phenol-red-free medium containing normal serum. We used phenol-red-free medium because phenol-red has been documented to have steroidogenic effects on the growth of hormone-responsive cell lines, including LNCaP [46,47].

Colony Formation and Cell Growth Assays
Cell growth was assessed either by colony forming assay or by counting the number of cells as described [22,48]. For colony forming assay, 1610 3 cells were seeded in 6 cm dishes in triplicate in 3 ml culture medium. After 24 h, one ml medium containing 1-25D, MB or DHT was added to the dishes to give the desired final concentration. Cells were grown for 21 days, with a medium change every 4 days containing the appropriate concentration of the above reagents, when they formed visible colonies. Colonies in dishes were fixed with 3.7% buffered formaldehyde, stained with 0.025% solution of crystal violet in PBS, and photographed. For measuring cell growth, 5610 4 cells were seeded in 6 cm dishes in replicate and treated with 1-25D described above. Cells were Note that significant enhancement is observed even at 1 nM. The asterisks (**) indicate P value of #0.0001. A two tailed Student's t test was used to calculate P value assuming unequal variance. C. Kinetics of enhancement of Cx32 expression level upon treatment with 1-25D (10 nM) for the indicated times. Note that enhancement is observed as early as 12 h and plateaus at 72 h. D. Quantitative analysis of the expression level of the data shown in C. Each bar represents the Mean and the Standard Error of the Mean from 3-11 experiments. The asterisk (*) indicates P value of #0.0016 and asterisks (**) indicate P value of #0.0001. A two tailed Student's t test was used to calculate P value assuming unequal variance. doi:10.1371/journal.pone.0106437.g001 allowed to grow for 10 days with a medium change at day 5. Cells were trypsinized and counted in a hemocytometer.

Vitamin D 3 Enhances Cx32 Expression Level
We used LNCaP-32 cells that express retrovirally-transduced rat Cx32 described previously [25,31]. We previously showed that in LNCaP-32 cells androgens regulated the formation of GJs, posttranslationally, by controlling the expression level of Cx32 by inhibiting its ERAD-mediated degradation [25]. Our subsequent studies showed that androgen-regulated degradation of Cx32 was abrogated by all-trans retinoic acid (ATRA) and 9-Cis retinoic acid (9-CRA) [31]. We thus rationalized that 1-25D might act similar to ATRA and 9-CRA. Based on the earlier studies showing the effect of vitamin D on LNCaP cell growth [10,[49][50][51][52], we treated LNCaP-32 cells with various concentrations of 1-25D to examine its effect on the expression level of Cx32. We found that 1-25D increased Cx32 expression level in a dose-dependent manner ( Figure 1A). Significant enhancement was observed even at concentration as low as 1 nM ( Figure 1B, left graph). Concentrations higher than 10 nM were toxic to these cells as assessed by the colony formation assay (unpublished data). For subsequent studies we chose 10 nM 1-25D. Time course studies showed that a significant increase in Cx32 expression level occurred as early as 12 h post-treatment with 1-25D and reached a plateau at 72 h ( Figure 1CD, right graph). The effect of 1-25D on Cx32 expression level was as potent as of synthetic androgen, MB, and 9-CRA. Moreover, combined treatment with 1-25D and MB was more effective in increasing Cx32 expression level ( Figure  2AB). Vitamin D 3 had previously been shown to affect the expression of level of E-cad, a constituent protein of adherens junctions, in colon cancer cells [33]. To determine if 1-25D also affected the expression level of adherens junction associated proteins, we measured the expression level of E-cad and its associated proteins aand b-catenins 72 h after treatment with 1-25D. The results showed that 1-25D had no effect on the expression level of E-cad and aand b-catenins ( Figure 2C). As measured by semi-quantitative RT-PCR analysis, 1-25D neither induced the expression of endogenous Cx32 in LNCaP-P or LNCaP-N cells (data not shown) nor altered the expression level of retrovirally-transcribed Cx32 mRNA in LNCaP-32 as documented previously [25].

Vitamin D 3 Enhances Gap Junction Assembly and Junctional Communication
We next examined the effect of 1-25D on the assembly of Cx32 into GJs. We found that, concomitant with an increase in the expression level of Cx32, 1-25D also increased GJ assembly as assessed by immunocytochemical analysis ( Figure 3A) and bio- Combined treatment with 1-25D with MB or 9-CRA is more effective in increasing Cx32 expression level than treatment with the single agent alone. B. Quantitative analysis of the expression level of the data shown in A. Each bar represents the Mean and the Standard Error of the Mean from 4-13 experiments. The asterisks (**) indicate P value of #0.0001. A two tailed Student's t test was used to calculate P value assuming unequal variance. C. Effect of 1-25D on adherens junction associated proteins. Expression of adherens junction proteins E-cadherin (E-cad), a-catenin (a-cat), and b-catenin (b-cat) was analyzed by Western blot analysis of total cell lysate (10 mg). Note that there is no effect. doi:10.1371/journal.pone.0106437.g002 Vitamin D 3 and Gap Junction Assembly PLOS ONE | www.plosone.org chemically by Western blot analysis of total, TX100-soluble andinsoluble extracts at 48 h after treatment ( Figure 3BC). This biochemical method is based on the principle that Cxs, which are incorporated into GJs, become insoluble in TX100 whereas Cxs that are not incorporated into GJs remain soluble [53]. This assay has been reproducibly shown to measure the assembly of Cxs into GJs as documented by earlier studies [24,25,53]. Moreover, we found that enhancement of GJ assembly was accompanied by a 2-3 fold parallel increase in junctional communication as determined by the junctional transfer of three GJ permeable fluorescent tracers, Lucifer Yellow (MW 443), Alexa 488 (MW 570), and Alexa 594 (MW 760). For example, 1-25D increased junctional transfer of Alexa 594 by 2-3 folds compared to controls ( Table 1). The effect of 1-25D on junctional communication was as potent as of synthetic androgen, MB, and the natural androgen DHT (Table 1). To determine if 1-25D affected the assembly of other junctional complexes, we also examined the detergent-solubility of adherens and tight junction associated proteins. The rationale  behind these studies was that E-cad has been shown to facilitate the assembly of Cxs into GJs [43,44,54], and Cx expression has been shown to facilitate the assembly of tight junctions and their constituent proteins [39]. We found that 1-25D had no effect on the solubility of E-cad and its associated proteins, aand b-catenin, and tight junction associated proteins, ZO-1 and occludin, in TX-100 suggesting that their assembly was not further enhanced into respective cell junctions ( Figure 3B). Taken together, these data suggest that 1-25D, like androgens and retinoids, enhances the expression level of Cx32, and its subsequent assembly into GJs, without discernibly altering the expression level of other cell junction associated proteins.

Vitamin D 3 Modulates Androgen-regulated Formation and Degradation of Gap Junctions
Earlier studies with LNCaP-32 cells had shown that androgen depletion caused degradation of Cx32 by ERAD, and that androgens enhanced GJ formation by re-routing the ERADtargeted pool of Cx32 to the cell surface, making it amenable for GJ assembly [25]. In subsequent studies, we showed that androgen-regulated formation and degradation of GJs was prevented by 9-CRA and ATRAs [31]. We rationalized that 1-25D might enhance GJ assembly by rescuing the ERAD-targeted pool of Cx32 like 9-CRA and ATRA. Therefore, we examined the expression level of Cx32 and its assembly into GJs upon androgen depletion in the presence and absence of 1-25D in LNCaP-32 cells. For these studies, we used androgen-depleted (charcoalstripped) and phenol-red-free cell culture medium to grow cells because phenol-red has a weak steroidogenic effect [46]. As was observed in our earlier studies [25], we found that androgendepletion decreased the expression level of Cx32 within 12 h, which not only was prevented upon addition of MB but also by 1-25D ( Figure 4AB). Moreover, combined treatment with MB and 1-25D appeared to be more effective in enhancing the expression level of Cx32 ( Figure 4A, upper blot). We also found that androgen depletion decreased the expression level of AR, which was also prevented upon treatment with not only androgens but also with 1-25D ( Figure 4A, bottom blot). To substantiate the above data, we further assessed the formation of GJs immunocytochemically ( Figure 5A) and functionally by measuring the junctional transfer of Lucifer Yellow, Alexa 488, and Alexa 594 ( Table 2). The results showed that GJs were barely observed in cells grown in androgen-depleted medium as assessed by the lack of Cx32-specific immunostaining at cell-cell contact areas, while they were readily observed when androgen-depleted medium was supplemented with 1-25D and MB ( Figure 5A). Functional assays showed that the junctional transfer of Lucifer Yellow, Alexa 488 and Alexa 594 decreased significantly upon androgen depletion, which was prevented upon replenishing androgen-depleted medium with MB and 1-25D (Table 2), thus substantiating the immunocytochemical data.
The immunocytochemical and junctional transfer data were further corroborated by the TX100-solubility assay ( Figure 5BC). We also examined the effect of androgen-depletion on the detergent-solubility of E-cad and b-catenin and tight-junctionassociated proteins, ZO-1 and occludin. Consistent with our earlier studies [25], the results showed that androgen depletion increased the detergent-solubility of occludin but not of ZO-1 and E-cad and b-catenin ( Figure 5BD). We also examined the effect of MB and 1-25D either alone or in combination on the formation of GJs in parental LNCaP-P and G418-resistant LNCaP-N cells and found that they had no effect (data not shown). Collectively, these data suggest that 1-25D prevents androgen-regulated degradation of Cx32 and enhances GJ formation in LNCaP-32 cells. Because combined treatment with androgens and 1-25D did not enhance GJ assembly further, it is likely that the assembly was enhanced by rescuing the same pool of Cx32 that was targeted for ERAD upon androgen depletion. Moreover, as was observed in our earlier studies, the assembly and detergent-solubility of Cx32 and LNCaP-32 cells, seeded in six well clusters or 10 cm dishes, were switched to charcoal-stripped, androgen-depleted medium (ST). GJ assembly and the expression level of Cx32 were determined by immunocytochemical (A) and Western blot (B) analyses by TX100-solubility assay in the presence and absence of 1-25D (10 nM) and MB (2.5 nM). In (A), Cx32 is in green and E-cad is red and the nuclei (blue) are stained with DAPI. Scale bar in A = 20 mM. In (B), T = total fraction; S = soluble fraction and I = Insoluble fraction. TX100-soluble and insoluble fractions as well as immunocytochemical assay were performed 24 h post-stripping as described in Materials and Methods. Note that GJs are degraded upon androgen depletion and degradation is blocked upon 1-25D treatment (A). Note also that the TX100-soluble fraction of E-cad (E-cad), b-catenin (b-cat) and ZO-1 is not affected. Note also that androgen-depletion increases the soluble fraction of occludin (B) without affecting the total occludin levels as quantitated in D. The asterisk (*) indicates P value of #0.0016 and asterisks (**) indicate P value of #0.0001. A two tailed Student's t test was used to calculate P value assuming unequal variance. doi:10.1371/journal.pone.0106437.g005 Vitamin D 3 and Gap Junction Assembly PLOS ONE | www.plosone.org occludin into cell junctions, or vice versa, appears to be regulated coordinately [25].

1-25D Enhances Gap Junction Formation Independent of Androgen Receptor Function
Our data showed that 1-25D prevented the degradation of AR upon androgen depletion ( Figure 4A, bottom blot). Also treatment of LNCaP cells with 1-25D had been shown to enhance AR expression level [40]. Thus, we considered the possibility that the effect of 1-25D on the enhancement of GJ assembly depended on the function of AR alone -and not on the independent effect of 1-25D. To test this notion, we treated LNCaP-32 cells with the anti-androgen, Casodex (Bicalutamide), to block androgen action [55,56]. Both androgen depletion and treatment with Casodex caused degradation of AR ( Figure 6A, upper blot, Figure 6B) and abolished the effect of MB and DHT on Cx32 expression level ( Figure 6A, bottom blot; Figure 6B) as was observed in our earlier studies [25,31]. However, we found that Casodex had no effect on the enhancement of the expression level of Cx32 resulting from the treatment with 1-25D in androgen-depleted medium ( Figure 6AB). To substantiate these data, we next examined the formation of GJs immunocytochemically in cells treated with Casodex in the presence and absence of MB or 1-25D. We found that GJs were not formed when cells were treated with Casodex in normal serum or in androgen-depleted medium containing MB as was observed in our earlier studies (Figure 7) [25,31]. On the other hand, we found that GJs were abundantly formed when cells were treated with Casodex and 1-25D (Figure 7). Altogether, these data suggest that the mechanism by which 1-25D prevents the degradation of Cx32 and enhances GJ formation upon androgen depletion is independent of AR.

Connexin32 Expression Alters the Growth Response of LNCaP Cells to 1-25D
Our earlier studies showed that Cx32 expression potentiated the growth-inhibitory effect of 9-CRA and ATRA in LNCaP cells [31]. To test if 1-25D has similar effect on growth, we measured cell growth of LNCaP-32 cells at concentrations that were barely growth-inhibitory to LNCaP-P cells. We determined cell growth by the colony forming assay as well as by counting the number of cells (Figure 8, Table 3). As assessed visually by the size of the colonies, we found that the growth of LNCaP-32 cells was inhibited by 1-25D whereas the growth of LNCaP-P and LNCaP-N cells was minimally affected ( Figure 8AB). These data were substantiated by measuring the growth of LNCaP-32 cells at two different concentrations (Table 3). For example, the growth of LNCaP-P and LNCaP-N cells was inhibited by only 20-25% upon treatment with 1-25D (1nM and 2.5 nM) whereas the growth of LNCaP-32 cells was inhibited by 55-70%. Moreover, we found that higher concentrations of 1-25D (5 and 10 nM) altered the morphology of LNCaP-32 cells profoundly such that LNCaP-32 cells treated with 1-25D appeared flatter and more epithelial-like whereas these changes were minimally observed in LNCaP-P and LNCaP-N cells (Figure 9). Change occurred only in response to 1-25D; and only when cells had been growing in 1-25D-containing medium for at least 4 days and have begun to be contact-inhibited and growth-arrested.

Discussion
The main findings of this study are as follows: 1. 1-25D enhances the expression level of Cx32 and its assembly into functional GJs in androgen-responsive LNCaP cells through inhibition of Cx32's degradation. 2. Formation of GJs sensitizes LNCaP cells to the growth inhibitory effect of 1-25D. We previously showed that 9-CRA and ATRA, the two well-known chemopreventive agents, also enhanced GJ assembly and sensitized these cells to their growth inhibitory effects [31]. Thus, it appears that GJ assembly is also the down-stream target of 1-25D in androgen-responsive LNCaP-32 cells, leading to suppression of growth. Several independent lines of inquiry prompted us to undertake these studies. First, growth inhibitory and chemopreventive effects of retinoids and vitamin D 3 had been previously documented to correlate with their ability to enhance the assembly of Cxs into GJs in other cancer cell types [35][36][37]57]. Second, the differentiated and polarized state of epithelial cells of the prostate, as well as of several other exocrine glands and tissues, had generally been found to coincide with the expression of Cx32 and its assembly into GJs [22,23,58]. Third, numerous studies had shown that like androgens [28,29], retinoids [57,[59][60][61][62][63], and 1-25D were essential for the growth and differentiation of the prostate [4,7,64]. Hence, we rationalized that its expression and LNCaP-32 cells were seeded as described in Table 1   Vitamin D 3 and Gap Junction Assembly PLOS ONE | www.plosone.org assembly into GJs might as well as be regulated by 1-25D either alone or in conjunction with the androgens. How might 1-25D enhance GJ assembly in LNCaP-32 cells? Nearly 50% of newly synthesized Cx32 may be degraded in the endoplasmic reticulum by ERAD [65]. We had previously shown that in LNCaP-32 cells, androgen depletion caused the degradation of nearly 70-80% of Cx32 by ERAD, and degradation was prevented upon replenishment with the androgens, which allowed Cx32 to traffic to the cell surface and assemble into GJs [25]. These studies further showed that androgens neither induced the expression of Cx32 in Cx-null LNCaP-P cells nor increased the expression level of retrovirally-driven Cx32 in LNCaP-32 cells [25,31]. Thus, in LNCaP-32 cells, androgens enhanced the expression level of Cx32 posttranslationally [25]. Although not tested directly, our data seem to suggest that 1-25D also enhances GJ assembly by preventing the androgen-regulated degradation of Cx32 by ERAD posttranslationally both under normal and androgen-depleted conditions. Like androgens, 1-25D neither induced the expression of the endogenous Cx32 in LNCaP-32 cells nor affected retroviral driven Cx32 mRNA transcripts. Our previous studies with LNCaP-32 cells showed that AR-mediated signaling was the sole determining factor in enhancing Cx32 expression level and preventing GJ degradation both under androgen-depleted or androgen-containing medium as Casodex, which inhibits AR-function [66], annulled the effect of androgens on Cx32 expression level and its subsequent assembly into GJs [25]. With regard to 1-25D effect, our data showed that it enhanced the expression level of AR under androgen-depleted conditions ( Figure 4). Therefore, it is possible that the effect of 1-25D in the absence of androgens may be indirectly caused by persistent and increased level of AR and its activation by the trace amounts of androgens present in the charcoal-stripped medium. However, 1-25D also enhanced GJ assembly robustly in the presence Casodex in androgen-depleted medium, which robustly decreased AR level and inhibited AR function ( Figure 6). One plausible explanation for these findings is that 1-25D activates an AR-dependent mechanism under androgen-depleted conditions to rescue the ERAD-targeted pool of Cx32 yet triggers another signaling pathway to enhance GJ assembly when AR function is inhibited by Casodex both under normal and androgen-depleted conditions. Further studies are required to explore this possibility.  Of note here are the findings that similar effects were also observed with 9-CRA and ATRA [31].
Cadherins have been shown to facilitate the assembly of Cxs into GJs. However, we failed to observe any effect of 1-25D on the expression and degradation of adherens junction associated proteins, E-cad and its associated proteins a and b catenin under androgen-containing and androgen-depleted conditions (Figures 3  and 5). Therefore, E-cad and its assembly into adherens junctions are not the likely targets of 1-25D in enhancing GJ assembly in contrast to its effect on E-cad expression in human colon carcinoma cells [33]. The assembly of Cx32 into GJs has also been shown to affect tight junction assembly [39,67]. Our previous studies had shown that androgen depletion increased the detergent-solubility of occludin, without significantly altering its expression level, and that the trafficking of occludin to the cell surface and its detergent-solubility was controlled by the assembly of Cx32 into GJs under androgen-depleted conditions [25]. Similar observations were also made in other cell lines [39,68,69]. The present study also showed that androgendepletion increased the detergent-soluble fraction of occludin without altering its expression level, which was negated when LNCaP-32 cells were treated with 1-25D ( Figure 5BD). In this regard, our data suggest that the assembly of Cx32 and occludin might as well be coordinately regulated by 1-25D, and that this may be one of the additional mechanisms by which 1-25D maintains the polarized state of prostate epithelial cells and acts as a differentiating and chemopreventive agent. Further studies are required to substantiate this notion.
1-25D has been shown to induce G0/G1 arrest, differentiation and apoptosis of tumor cells by modulating different signaling pathways to delay tumor progression in different cancer cell types; moreover, it has also been known to potentiate the cytotoxic effects of many chemotherapeutic agents [1][2][3]. Several studies have shown that vitamin D inhibits the growth of PCA cell lines, including LNCaP, in both AR-dependent and -independent manner [10,49,50,70,71]. We found that the expression of Cx32 potentiated the growth inhibitory effect of 1-25D such that suppression of growth was observed at doses which had no significant effect on the growth of Cx-null LNCaP cells ( Figure 8, Table 3). For example, 1-25D at 1nM barely inhibited the growth of LNCaP-P and LNCaP-N cells but inhibited the growth of LNCaP-32 by more than 50% (see Table 3). Earlier studies had shown that LNCaP cells were sensitized to undergo apoptosis by tumor necrosis factor a, TRAIL, and anti-Fas antibodies when Cx43 was expressed via adenoviruses to which Cx-null LNCaP cells were resistant [72]. Moreover, expression of Cx32 not only inhibits the growth of cells but also induces differentiation in breast cancer cell lines as well as in LNCaP cells [22,73,74].
What might be the possible explanation for the growthsuppressive effects of 1-25D with regard to the assembly of Cx32 into GJs? The signaling pathways that are activated or suppressed upon formation and degradation of GJs to impact cell growth and differentiation are not well-understood [16,17,[75][76][77][78][79]. Elegant studies in Cx32 knockout mice revealed increased activation of mitogen-activated protein kinases and decreased level of tumor suppressor p27Kip1 [80][81][82]. While it is well-known that 1-25D suppresses the growth of several human PCA cell lines, the effect appear not to depend on the expression level of vitamin D receptor. For example, only LNCaP cells were found to be exquisitely sensitive to the growth inhibitory effect of 1-25D whereas other PCA cell lines, such as PC-3, DU-145 and ALVA-31, were barely sensitive despite the fact that all cell types expressed nearly similar levels of vitamin D receptor [50,71,83,84]. Also, in LNCaP cells the growth suppression by 1-25D was mediated via increased expression of cyclin-dependent kinase inhibitors p21waf1/cip1 and p27kip1 as well as through hyper-phosphorylation of retinoblastoma protein, resulting in G0/ G1 arrest [50]. Given the fact that Cx expression also has an impact on cell cycle [79,85], it is possible that transmission of growth-regulatory signals through channels composed of Cx32 activates signaling pathways that increase the expression of generegulatory proteins involved in the control of cell cycle progression as proposed [18,79]. Given the multiple effects of 1-25D on different tumor cell types, it is at present difficult to envisage how its chemopreventive, pro-differentiating and growth-inhibitory effects are related to its ability to regulate formation and degradation of GJs [16][17][18]77,79,86]. More elaborate studies are underway to explore the molecular basis of the augmentation of  An intriguing observation made during this study was that 1-25D caused a radical change in the morphology of LNCaP-32 cells compared to Cx-null LNCaP-P and LNCaP-N cells such that the treated cells became flatter and acquired an epithelial morphology ( Figure 9). Expression of Cx32 by itself had no conspicuous effect on the morphology and the change occurred only in response to 1-25D; and only when cells had been growing in 1-25D-containing medium for at least 4 days and have begun to be contact-inhibited and growth-arrested. One possible explanation for these data is that the formation of large GJs in response to 1-25D permits a more elaborate remodeling of the cortical actin network, which has emerged as a key regulator of cell morphology [87,88]. This notion is supported by studies that utilized embryonic fibroblasts from Cx43 knockout mice in wound-healing studies. Fibroblasts from Cx43 knockout mice showed cell polarity defects as characterized by the failure of the microtubule organizing center to reorient with the direction of wound closure as well as failure of actin stress fibers to appropriately align at the wound edge [89]. Whether expression of Cx32 in LNCaP cells also governs cell shape in response to 1-25D through modulation of actin-cortex or microtubule network remains to be explored in future studies.
Several pre-clinical and clinical trials have suggested that a decrease in vitamin D 3 levels contributes to the development and possibly to the progression of human PCA [1,2,90,91]. Con-nexin32 is expressed by the luminal epithelial cells of normal prostate and is aberrantly assembled in the epithelial cells of prostate tumors [22,23]. Because GJs have been implicated in maintaining the polarized and differentiated state of epithelial cells [39], we propose that the chemopreventive effects 1-25D in PCA may result from its ability to enhance the formation of GJs. Our results show that GJ assembly is the down-stream target of signaling initiated by 1-25D and that the formation of GJs sensitizes PCA cells to its growth modulatory influence. Because loss of cell junctions is a hallmark of PCA progression [29] and might occur as early as during prostatic intraepithelial neoplasia [92,93], understanding basic cell and molecular biological mechanisms by which 1-25D might govern the formation of GJs will provide new insights with regard to signaling pathways utilized to maintain the polarized and the differentiated state of epithelial cells in prostate tumors. This should open innovative avenues for designing new therapeutic approaches to delay the onset of malignancy as loss of polarization is the earliest changes that may initiate PCA progression [92].