Stimulation by 1α,25(OH)2-Vitamin D3 of Whole Cell Chloride Currents in Osteoblastic ROS 17/2.8 Cells

1α,25-Dihydroxyvitamin D3(1α,25(OH)2D3) can generate biological responses via genomic and nongenomic mechanisms. This article reports for the first time the effects of 1α,25(OH)2D3 and structurally related analogs on whole cell chloride currents in osteoblastic cells. 1α,25(OH)2D3 promoted the rapid enhancement of outwardly rectifying Cl− currents in 93% of the osteoblasts in a concentration-dependent manner, with a maximal increase of about 4-fold between 0.5 and 5 nm. This effect of 1α,25(OH)2D3 was blocked by 1 nm stereoisomer 1β,25(OH)2D3when added to the bath before 1α,25(OH)2D3. On the other hand, 1 nm of the 6-s-cis locked analog 1α,25(OH)2-lumisterol3 significantly increased by about 2.2-fold outward Cl− currents in the ROS 17/2.8 cells, whereas the increase promoted by same concentration of the 6-s-trans locked analog 1α,25(OH)2-tachysterol (0.8-fold) was significantly lower, suggesting that the 6-s-cis locked or steroid-like form was preferred over the extended 6-s-trans conformer to promote these rapid effects of the hormone. We conclude that the agonist effects of 1α,25(OH)2D3 in osteoblasts at the cellular membrane level seem to be determined by some structural features of the molecule which may be crucial for its interaction with a putative membrane receptor in the cell surface.

-tachysterol (0.8-fold) was significantly lower, suggesting that the 6-s-cis locked or steroid-like form was preferred over the extended 6-s-trans conformer to promote these rapid effects of the hormone. We conclude that the agonist effects of 1␣,25(OH) 2 D 3 in osteoblasts at the cellular membrane level seem to be determined by some structural features of the molecule which may be crucial for its interaction with a putative membrane receptor in the cell surface.
Osteoblasts, which are bone-forming cells, are a main target cell for calciotropic hormones including 1␣,25-dihydroxyvitamin D 3 (1␣,25(OH) 2 D 3 ). 1 1␣,25(OH) 2 D 3 , the most biologically active metabolite of vitamin D 3 , generates biological responses via genomic as well as rapid, nongenomic mechanisms (reviewed in Refs. [1][2][3]. Genomic effects comprise the regulation of the transcription of different genes via interaction of the hormone with a nuclear receptor (nVDR) which in turn interacts with hormone response elements in the promoter region of those specific genes (4,5). In contrast, nongenomic effects (6,7) are likely initiated at the cell membrane level and seem to involve a putative membrane receptor for 1␣,25(OH) 2 D 3 (8), a variety of second messengers (2), and the modulation of ion channel activity (9 -11).
A wide variety of ion channels has been described in primary cultured osteoblasts and osteoblast-like cell lines (9 -21). In particular, chloride channel activity in osteoblasts has been postulated to be related to rapid changes in the electrical state of the cell and a regulatory cellular volume response observed under the influence of different hormones acting on the bone (22)(23)(24)(25).
To date, only a few published papers have shown the modulatory effects of 1␣,25(OH) 2 D 3 on ion channel activity in target cells, which are believed to represent one of the first examples of a nongenomic response. It has been demonstrated in osteoblasts that the hormone facilitates the opening of L-type Ca 2ϩ channels (9,10); these responses occurred within seconds after the addition of 1␣,25(OH) 2 D 3 . Recently, we reported in a preliminary communication that 1␣,25(OH) 2 D 3 also increases an outwardly rectifying anion conductance in ROS 17/2.8 cells (11). In this report we study via whole cell patch-clamp techniques the relative ability of the conformationally flexible 1␣,25(OH) 2 D 3 and related analogs to modulate chloride channels in ROS 17/2.8 cells. Our results demonstrate that the rapid effects of 1␣,25(OH) 2 D 3 on chloride channels are determined by specific structural components of the agonist ligand. Thus the 1-hydroxy epimer 1␤,25(OH) 2 D 3 was found to be an antagonist of 1␣,25(OH) 2 D 3 , and the 6-s-cis locked 1␣,25(OH) 2lumisterol was significantly more potent than the 6-s-trans locked 1␣,25(OH) 2 -tachysterol in stimulating Cl Ϫ currents in these bone cells. This is also the first extended report on the participation of Cl Ϫ channels in nongenomic effects of 1␣,25(OH) 2 D 3 .

EXPERIMENTAL PROCEDURES
Cell Culture-ROS 17/2.8 cells (obtained from M. C. Farach-Carson) were cultured in Ham's F-12 medium (Sigma) containing 5% fetal bovine serum (Sigma) and 5% Serum Plus (JRH Biosciences, Woodland, CA), as described previously (11). For patch-clamp experiments, cells were plated at very low density in 35-mm tissue culture dishes. Prior to recordings, the cells were washed at least three times with the electrophysiological external solution to remove the medium completely.
Patch-clamp recordings (26) were performed with an Axopatch 1C patch-clamp amplifier (Axon Instruments, Foster City, CA). Patch pipettes were fabricated from Drummond capillaries (Drummond Scien-tific Co., Broomall, PA), coated with Sylgard elastomer (Dow Corning Corp., Midland, MI) to reduce capacitative transients, and fire-polished. Experiments were carried out at room temperature. Currents were low-pass-filtered at 1 kHz and digitized every 100 s. Cell membrane capacitance and series resistance were electronically compensated prior to the recording of currents. Leak current compensation was done by subtracting the leak current elicited by an hyperpolarizing pulse from Ϫ70 mV to Ϫ80 mV. In all experiments, depolarizing pulses were applied at intervals of 2-4 s.
Chemicals-1␣,25(OH) 2 D 3 , 25-OH-D 3 , and 1␤,25(OH) 2 D 3 (analog HL) were obtained from M. Uskokovic (Hoffmann-La Roche); 1␣,25-(OH) 2 -lumisterol 3 (analog JN) and 1␣,25(OH) 2 -tachysterol (analog JB) from W. H. Okamura (University of California, Riverside, CA) (27). Vitamin D analogs were stored in the dark as stock solutions in absolute ethanol at Ϫ20°C and added to the bath solution containing 1 mg/ml cytochrome c, which was used as a nonspecific protein carrier. Appropriate controls for cytochrome c and a final ethanol concentration of 0.01% or less were carried out before the addition of the analogs to the bath. CdCl 2 (Sigma) and nifedipine (Sigma) were used as specific Ca 2ϩ channel blockers and were added to the bath from aqueous and ethanolic stock solutions, respectively. 4,4Ј-Diisothiocyanatostilbene-2,2Ј-disulfonic acid (DIDS, Sigma) was used as a specific Cl Ϫ channel blocker and was added to the bath from a stock solution made in the recording medium. Cholesterol (Calbiochem) and ␤-estradiol (Sigma) were added to the bath from a stock solution in ethanol. ROS 17/2.8 Cells-The structures of the conformationally flexible 1␣,25-(OH) 2 D 3 and other analogs employed in this study are shown in Fig. 1.

Effects of 1␣,25(OH) 2 D 3 on Ion Currents in
We first recorded the activity of previously described voltagedependent L-type Ca 2ϩ channels in the ROS 17/2.8 cells (9, 10, 15). Fig. 2 shows the effect mediated by 1␣,25(OH) 2 D 3 on Ba 2ϩ currents when it was added to the bath at a final concentration of 0.5 nM. This effect, which have been described before (9,10,15), was characterized by a drastic shift of I/V relations of about 25 mV to more negative potentials in the case of this particular cell and developed over the course of the first few minutes after the addition of the hormone. In this experiment, an increase in the concentration of 1␣,25(OH) 2 D 3 to 5 nM did not cause any additional effect, although this result varied among different cells. We recorded a mean shift of Ϫ10.4 Ϯ 2.6 mV in 8 out of 22 cells (36%) studied under the same conditions. These 1␣,25-(OH) 2 D 3 -sensitive inward Ba 2ϩ currents were almost completely blocked by 100 M Cd 2ϩ , a Ca 2ϩ channel blocker, subsequently added to the bath. As postulated previously (9), this modification of the voltage sensitivity of Ca 2ϩ channels in osteoblasts by 1␣,25(OH) 2 D 3 may result in the facilitation of Ca 2ϩ channel opening by the hormone for Ca 2ϩ uptake at membrane potentials close to the resting value, which has been  ) measurements were made after 5 min following the addition of each concentration of the hormone, which is the time required for achieving maximal effects. I Ba was measured at the peak of maximal activation of inward Ba 2ϩ currents elicited by a series of 200-ms duration depolarizing voltage steps to between Ϫ30 and 70 mV, applied every 2 s. I/V relations were obtained from the same single cell. A shift of about Ϫ25 mV of the peak value of the I/V curves was recorded after addition of 0.5 nM 1␣,25(OH) 2 D 3 . In the case of this cell, further addition of a higher concentration of the hormone (5 nM) did not cause any additional shift of the I/V curve. The shift of peak I/V values was estimated directly from the plots. Further addition of 100 M CdCl 2 to the bath almost completely blocked the inward Ba 2ϩ currents through Ca 2ϩ channels (closed triangles). The numbers in parentheses indicate the sequential order of the recordings. B, inward current traces corresponding to data shown in A, activated by a depolarizing step to 0 mV in the absence (control) and 5 min after the addition of 0.5 nM 1␣,25(OH) 2 D 3 . The holding potential was Ϫ70 mV.
reported to be in the range of Ϫ10 to Ϫ40 mV in osteoblasts (25,28,29).
In a preliminary communication we reported that 1␣,25(OH) 2 D 3 also affects Cl Ϫ channel activity in the ROS 17/2.8 cells (11). In the presence of 100 M Cd 2ϩ in the bath which effectively blocks Ca 2ϩ channel activity, an outward current was recorded in the range of Ϫ30 to 80 mV in the glutamate-containing external solution (see "Experimental Procedures") in approximately 80% of the cells, as shown in Fig. 3. The addition of a final concentration of 0.05 nM 1␣,25-(OH) 2 D 3 to the bath caused, in the case of this particular cell, a 1.2-fold increase of the outward current measured at 80 mV over the course of 2.5 min. The time course of this effect is shown in Fig. 3B. After a lag period of about 45 s, which could be attributed to the time required for diffusion of the steroid to the cell and/or the stimulation of cellular signal transduction pathways, the rapid increase of the current took place over the course of the next 30 s. 1␣,25(OH) 2 D 3 promoted the increase of outward currents in 14 out of 15 cells (93%) studied under the same recording conditions. This enhancing effect was dependent on the concentration of the hormone in the range of 0.05-50 nM (see also Fig. 5).
Permeability studies carried out by replacing the main anion in the recording solution are summarized in Table I. We found that the 1␣,25(OH) 2 D 3 -sensitive outward current was permeable to glutamate and Cl Ϫ , suggesting a poor anion discrimination of this channel. On the other hand, gluconate in the external solution significantly decreased outward currents, as expected for a less permeable anion. The addition of 200 M DIDS, a specific Cl Ϫ channel blocker, to the bath blocked the 1␣,25(OH) 2 D 3 -enhanced outward currents, as shown in Fig. 4. This blockade by DIDS was time-and voltage-dependent. It developed over the course of 1-2 min after the addition of the agent to the bath. As described before for the blockade by DIDS of the cAMP-activated Cl Ϫ currents in primary cultured rat osteoblasts (22) and of the mechanosensitive Cl Ϫ channels in ROS 17/2.8 cells (23), these 1␣,25(OH) 2 D 3 -sensitive Cl Ϫ currents were strongly reduced by the agent but were not completely blocked. DIDS was shown to be effective at a concentration of 200 M when any of the anions shown in Table I were used. Fig. 4 also shows the outward rectification of these Cl Ϫ currents. No inward currents were recorded at membrane potentials below the reversal potential (E rev ) for Cl Ϫ in the Cl Ϫcontaining solutions (see "Experimental Procedures"). The E rev for Cl Ϫ calculated from the Nernst equation gave a value of ϷϪ6 mV, which is very close to the value found from the recordings.
Antagonist Effects of the 1-Hydroxy Epimer 1␤,25(OH) 2 D 3 -The synthetic analog 1␤,25(OH) 2 D 3 (analog HL), which only differs from the natural metabolite in the orientation of the hydroxy group on carbon 1 (see Fig. 1), has been shown by this laboratory to inhibit the rapid activation of transepithelial calcium transport by 1␣,25(OH) 2 D 3 in chick intestine (30) and 45 Ca 2ϩ uptake in ROS 17/2.8 cells (31). Analog HL has also been shown to reduce the agonist potentiation by 1␣,25(OH) 2 D 3 of Ca 2ϩ channels in ROS 17/2.8 cells (32). In this work, we investigated the effects of 1␤,25(OH) 2 D 3 on outward Cl Ϫ currents in the osteoblastic cells. No enhancement of the Cl Ϫ currents by 1 nM 1␤,25(OH) 2 D 3 was found in 44% of the studied cells, while we recorded a modest 1.1 Ϯ 0.3-fold increase of the outward currents at 80 mV in the remaining 56% of the cases  compared with the increments recorded for the hormone alone (see Fig. 5). On the contrary, when the concentration of 1␣,25(OH) 2 D 3 was raised to 50 nM, the increase of the outward currents in the presence of the ␤-stereoisomer was similar to that measured for the hormone alone, suggesting that both ligands may be competing for the same receptor. These results with HL suggest that the 1-hydroxy epimer 1␤,25(OH) 2 D 3 may act as an antagonist of the 1␣,25(OH) 2 D 3 effects on Cl Ϫ channels in osteoblasts as it blocked the effects of the hormone on these cells. On the other hand, 25-OH-D 3 did not cause any significant modification of I/V relations for inward Ba 2ϩ currents in the range of 0.05-50 nM (see Fig. 6B), which agrees with previously published results (9,10). The specificity of this effect of 25-OH-D 3 on Cl Ϫ currents, which was not found on Ba 2ϩ currents in the same cellular system, may be indicative of different modulatory mechanisms underlying the mode of action of the vitamin D metabolites.

Effects of the Natural Metabolite 25-OH-D 3 -The natural metabolite 25-OH-D 3 promoted an increase in Cl
Enhancement of Cl Ϫ Currents by a 6-s-cis Locked and a 6-s-trans Locked Analog-It has been recently shown that synthetic vitamin D analogs locked in the 6-s-cis position (steroidlike molecules, see Fig. 1) are potent agonists for the rapid effects of the hormone, while 6-s-trans locked conformers (extended forms) are inactive for the same responses in target cells (27,33). Noticeable transcaltachia-promoting effects in chick intestinal epithelium and 45 Ca 2ϩ uptake by ROS 17/2.8 cells have been described specifically for the synthetic con- former 1␣,25(OH) 2 -lumisterol 3 (analog JN), which is locked in the 6-s-cis position (27). We found that 1-10 nM analog JN caused a 2.2 Ϯ 0.7-fold increase of the outward anion conductance in 7 out of 10 ROS 17/2.8 cells (70%). This response did not differ significantly from the effects promoted by 0.5 nM 1␣,25(OH) 2 D 3 (p Ͻ 0.05, see Fig. 7). On the other hand, 1␣,25(OH) 2 -tachysterol (analog JB), a synthetic 6-s-trans conformer which has been shown to be inactive in both transchaltachia and 45 Ca 2ϩ influx in osteoblasts (27) promoted only a modest 0.8 Ϯ 0.3-fold increase of Cl Ϫ currents at 80 mV when applied at 1-10 nM in 80% of the cells. In this case, the response differed significantly from the enhancing effect exerted by 0.5 nM 1␣,25(OH) 2 D 3 (p Ͻ 0.05, Fig. 7).
The magnitude of the effects on Cl Ϫ currents promoted by the different analogs of 1␣,25(OH) 2 D 3 used in this work is shown comparatively in Fig. 7. Note that the effects promoted by 1-10 nM JN and 1-10 nM JB on these Cl Ϫ currents differed significantly from each other (p Ͻ 0.01).
Specificity of the Response by 1␣,25(OH) 2 D 3 Analogs-The specificity of ion channel responses to 1␣,25(OH) 2 D 3 analogs was investigated by means of the evaluation of possible effects exerted by other steroids on the Cl Ϫ outward currents in ROS 17/2.8 cells. Steroid hormones have been demonstrated to modify ion channel activity in different cells (34 -36). As also shown in Fig. 7, 50 nM cholesterol and 0.01-10 nM ␤-estradiol did not have any significant effect on the Cl Ϫ currents studied in this work.

DISCUSSION
It has been demonstrated in osteoblasts that different hormones taking part in the process of bone remodeling affect ion channel activity in the plasma membrane. More specifically, electrophysiological measurements carried out on the boneforming cells have shown that the secosteroid 1␣,25(OH) 2 D 3 increases Ca 2ϩ influx through voltage-activated Ca 2ϩ channels by facilitating the opening of the channels at membrane potentials close to the resting value (9,10) and also enhances an outward K ϩ current activated by depolarization (37). However, the precise mechanisms for ion channel modulation by the hormone and their physiological role remain unknown.
We recently described in a preliminary report the enhancing effect of 1␣,25(OH) 2 D 3 on Cl Ϫ currents in the osteoblastic cell line ROS 17/2.8 cells (11). In the present work, we describe in more detail the effects of 1␣,25(OH) 2 D 3 and related structural analogs on these currents, this being the first extended report on a Cl Ϫ conductive membrane pathway involved in the rapid responses of this secosteroid.
We found an outwardly rectifying voltage-dependent Cl Ϫ current activated upon depolarization in ROS 17/2.8 cells that can be increased by external application of physiological concentrations of the hormonally active 1␣,25(OH) 2 D 3 . This effect was found in 93% of the studied cells. Since patch-clamp currents are defined on the basis of the direction of the movement of positive charges, an outward current in the case of anions represents an influx of the negative charges into the cell. Therefore, according to our results, the depolarization of the osteoblast membrane from the resting value (Ϫ40 to Ϫ10 mV, according to Refs. 28 and 29) activates an influx of anions, Cl Ϫ being the most abundant one under physiological conditions.
The potentiation of the anion currents by 1␣,25(OH) 2 D 3 was dependent upon the concentration of the hormone and had a maximal increase at 0.5-5 nM followed by an attenuation at 50 nM. A biphasic effect of 1␣,25(OH) 2 D 3 was also previously described for other nongenomic effects of the hormone including the promotion of Ca 2ϩ uptake in ROS 17/2.8 cells and the rapid Ca 2ϩ transport process in the chick intestinal epithelium (transcaltachia) (9,33).
The enhancing effect of 1␣,25(OH) 2 D 3 on the anion currents in the osteoblastic cell line developed rapidly, in the course of only a few seconds to minutes. The rapidity of these effects suggests that they are different from the classical steroid receptor-mediated nuclear effects and are part of an increasing list of 1␣,25(OH) 2 D 3 -mediated nongenomic effects described in different target cells (2, 3, 30, 33, 38 -41).
Structure-function studies carried out with different structurally related 1␣,25(OH) 2 D 3 analogs on different target cellular systems have proved to be very useful in defining the molecular steps involved in the mechanisms of action of the hormone (reviewed in Ref. 1). In the case of rapid nongenomic effects postulated to be initiated at the plasma membrane level, the use of synthetic analogs has led to the discovery that some structural forms are "preferred" over others which have been shown to be more effective in genomic processes. The natural secosteroid 1␣,25(OH) 2 D 3 exists as a continuum of potential shapes extended from the 6-s-cis (steroid-like) to the 6-s-trans (extended, see Fig. 1) which may interact with the receptors. It has been shown recently that synthetic 6-s-trans locked analogs are inactive in both rapid and some genomic responses, while 6-s-cis locked analogs are potent agonists of membraneinitiated nongenomic effects (27). In the present work, we demonstrate for the first time that the 6-s-cis locked analog 1␣,25(OH) 2 -lumisterol 3 significantly increased outwardly rectifying voltage-activated Cl Ϫ currents in ROS 17/2.8 cells, whereas the enhancing effects by 1␣,25(OH) 2 -tachysterol, the corresponding 6-s-trans conformer, were significantly lower (see Fig. 7).
In a related study the 1␤ epimer of 1␣,25(OH) 2 D 3 , 1␤,25-(OH) 2 D 3 , which has been shown to block the rapid effects of the hormone (30 -32) and to bind to the cellular membrane in osteoblasts (42), remarkably decreased the potentiation of Cl Ϫ currents by 1␣,25(OH) 2 D 3 in ROS 17/2.8 cells (see Fig. 5), suggesting that it may also act as an antagonist of ion channel responses by the hormone. We conclude that inversion of the orientation of the hydroxyl on carbon 1 of the hormone (1␤ for 1␣) may be enough to block the response of cell ion channels by competitively binding to the same surface receptor as a first step in the process.
Finally, the specific effects found for the natural metabolite 25-OH-D 3 on Cl Ϫ currents but not on Ba 2ϩ currents in ROS 17/2.8 cells (see Fig. 6) open the possibility to the existence of different cellular modulatory mechanisms underlying the control of membrane ionic pathways and the electrical state of the cell by active 1␣,25(OH) 2 D 3 .
Although there is growing evidence that 1␣,25(OH) 2 D 3 may exert its rapid, nongenomic effects by interacting with a putative mVDR and by initiating a series of molecular pathways involving second messengers, future experiments need to be carried out to elucidate the molecular steps linking the hormonal signal and the specific enhancement of voltage-dependent outward Cl Ϫ currents in osteoblasts.