Calcium Transport and Local Pool Regulate Polycystin-2 (TRPP2) Function in Human Syncytiotrophoblast

Abstract

Polycystin-2 (PC2, TRPP2) is a Ca²⁺-permeable, nonselective cation channel implicated in Ca²⁺ transport and epithelial cell signaling. Although PC2 may contribute to Ca²⁺ transport in human term placenta, the regulatory mechanisms associated with Ca²⁺ handling in this tissue are largely unknown. In this work we assessed the regulation by Ca²⁺ of PC2 channel function from a preparation of apical membranes of human syncytiotrophoblast (PC2_hst) reconstituted in a lipid bilayer system. Addition of either EGTA or BAPTA to the cis hemi-chamber, representing the cytoplasmic domain of the channel, and lowering Ca²⁺ to ∼0.6–0.8 nM, inhibited spontaneous PC2_hst channel activity, with a time response dependent on the chelator tested. EGTA reduced PC2_hst channel currents by 86%, with a t_1/2 = 3.6 min, whereas BAPTA rapidly and completely (100%) eliminated channel activity with a t_1/2 = 0.8 min. Subsequent titration with Ca²⁺ reversed the inhibition, which followed a Hill-type function with apparent dissociation constants of 1–5 nM, and 4 Ca²⁺ binding sites. The degree of inhibition by the cis Ca²⁺ chelator largely depended on increasing trans Ca²⁺. This was consistent with measurable Ca²⁺ transport through the channel, feeding the regulatory sites in the cytoplasmic domain. Interestingly, the reconstituted in vitro translated PC2 (PC2_iv) was completely insensitive to Ca²⁺ regulation, suggesting that the regulatory sites are not intrinsic to the channel protein. Our findings demonstrate the presence of a Ca²⁺ microdomain largely accessible through the channel that controls PC2 function in human syncytiotrophoblast of term placenta.

Introduction

The human syncytiotrophoblast (hST) transfers Ca²⁺ from mother to the fetus, to meet developmental needs, particularly during the third trimester of pregnancy (1,2). Ca²⁺ entry by the hST is thought to be mediated by ion channels, whose molecular identity remains to be fully determined. A number of Ca²⁺-permeable ion channels present in human placenta may contribute to this function, including L-type and T-type voltage-gated channels, and the highly Ca²⁺ permeable TRP channel isotypes TRPV5 and TRPV6 (3–6). Limited information is presently available, however, regarding the possible contribution of these channels to Ca²⁺ transport in the placenta. Studies from our laboratory determined the abundant expression of another TRP isotype, polycystin-2 (PC2, TRPP2), in the hST of term placenta (7). PC2 is the product of the gene PKD2, whose mutations cause autosomal dominant polycystic kidney disease (ADPKD) (8). PC2 may be relevant to epithelial Ca²⁺ transport due to its large nonselective cation conductance and slight selectivity to divalent cations (7).

PC2 is expressed in various epithelial tissues and other organs, where it is largely associated with Ca²⁺ transport and cell activation (9,10). This channel protein is found in several cell locations, including the primary cilium, intracellular Ca²⁺ stores, and the plasma membrane (10,11). PC2 shares homology with voltage-gated Na⁺, and Ca²⁺ channels (8), both in the pore region, and in putative Ca²⁺-binding domains such as the EF-hand (8,12), that may contribute to Ca²⁺-dependent channel regulation (13,14). Cytoplasmic Ca²⁺ regulation of PC2 function in the plasma membrane may entail direct interactions with the channel protein. Vassilev et al. (15) reported the transient activation of wild-type PC2 by increasing the intracellular Ca²⁺ concentration from 0.1 to 1 μM in Xenopus laevis oocytes overexpressing the protein. This regulation was absent in the truncated PC2 carrying the mutation R742X, which lacks the cytoplasmic terminus, and causes ADPKD. Cai et al. (16) determined that wild-type PC2 Ca²⁺-dependent, bell-shaped open probability in response to voltage shifts in the mutated PC2-S812A carrying an alanine substitution in Ser⁸¹², which represents a casein kinase phosphorylation site. Previous studies from our laboratory demonstrated that intracellular Ca²⁺ regulation of PC2 function in the hST implicates interactions with the actin cytoskeleton and actin-associated proteins. The PC2 stimulatory effect of the actin filament disrupter cytochalasin D in reconstituted hST apical membranes was mimicked by addition of gelsolin from the cis (cytoplasmic) site of the reconstitution chamber in the presence, but not the absence, of micromolar Ca²⁺ (17). Thus, Ca²⁺ regulation of PC2 may implicate diverse mechanisms, entailing both the channel itself and/or regulatory proteins associated with it.

Herein, we explored the response to Ca²⁺ of PC2_hst from reconstituted hST apical membranes and the in vitro translated PC2 (PC2_iv) after addition of Ca²⁺ chelators EGTA and BAPTA. Whereas PC2_iv was insensitive to changes in Ca²⁺ concentrations, PC2_hst activity decreased to different degrees and kinetics responses depending on the chelator, and the trans Ca²⁺ concentration. The evidence suggested the presence of cooperative Ca²⁺ binding sites, which are not intrinsic to, but regulate, PC2_hst. The evidence supports a feedback mechanism of a local cytoplasmic Ca²⁺ pool, with intracellular high-affinity Ca²⁺ binding sites, which are in turn modulated by Ca²⁺ entry through the channel in hST.

Materials and Methods

Human placenta membrane preparation

Apical hST plasma membranes from term human placenta were obtained as previously described in González-Perrett et al. (7). Briefly, normal placenta from vaginal deliveries were obtained and immediately processed. The villous tissue was fragmented, washed with ice-cold unbuffered NaCl saline (150 mM), and minced into small pieces. The fragmented tissue was processed, filtered, and centrifuged, as previously reported in González-Perrett et al. (7). The final pellet was resuspended in a buffer solution containing HEPES 10 mM, sucrose 250 mM, and KCl 20 mM and adjusted to pH 7.4, and was then aliquoted and stored frozen until the time of the experiment. Apical hST enrichment usually was higher than 26-fold.

Preparation of in vitro translated PC2

In vitro translated PC2 (PC2_iv) was prepared as previously reported in González-Perrett et al. (7), with a reticulocyte lysate system TnT T7 (Promega, Fitchburg, WI). The plasmid pGEM-PKD2 encoding PC2, was in vitro transcribed and translated with a reticulocyte lysate system TnT T7 (Promega) by incubation of plasmid DNA (1 μg) and 50 μL of the reaction mixture for 90 min at 30°C. The PC2_iv was introduced by dialysis into liposomes as previously reported in González-Perrett et al. (7).

Ion channel reconstitution

PC2 containing vesicles were reconstituted into lipid bilayers as previously reported in González-Perrett et al. (7) in a KCl chemical gradient (cis/trans, 150:15 mM). PC2 containing vesicles were incorporated into lipid bilayers of a reconstitution system. The lipid mixture was a 7:3 ratio of POPC and POPE (20–25 mg/mL; Avanti Polar Lipids, Birmingham, AL) in n-decane. Unless otherwise stated, the cis chamber contained a solution of KCl 150 mM, CaCl₂ 10 μM, and HEPES 10 mM, at pH 7.40. The trans side contained a similar solution with lower KCl (15 mM), to create a KCl chemical gradient. PC2_hst was identified as previously reported in González-Perrett et al. (7), by a large conductance (∼170 pS), K⁺-conducting channel, which was inhibited by trans (external) amiloride, and cis (cytoplasmic side of PC2) anti-PC2 antibody, properties that also ensured its orientation in the reconstituted membrane (7).

Reagents and Ca²⁺ chelation

EGTA was dissolved in NaOH and titrated with HCl to reach pH 7.1. BAPTA was dissolved in dimethylsulfoxide. The concentrated reagents (16 μL and 8 μL, respectively) were diluted in either cis (1600 μL) or trans (1000 μL) chambers and buffered at pH 7.4 (10 mM HEPES), to reach final concentrations of 1 mM and 2 mM, respectively (see Results). Neither chelator nor vehicle alone affected the final pH. The Ca²⁺ concentration was calculated by

$$\left[\text{Ca}\right]=\frac{K_Q\left[\text{Ca}Q\right]}{\left[Q\right]},$$ (1)

where K_Q, is the dissociation constant of the Ca²⁺-chelator complex, [Q] is the concentration of the free chelating agent, and [CaQ] is the concentration of Ca²⁺ bound to Q. The final free Ca²⁺ concentration was estimated to be either 0.6 or 0.8 nM (pH ∼7.4) in the presence of EGTA or BAPTA, respectively (more details in the Supporting Material).

Data acquisition and analysis

Single channel currents obtained at 40–60 mV with a PC501A patch-clamp amplifier (Warner Instruments, Hamden, CT) with a 10 gigaOhm feedback resistor and signal were driven and processed with the software pCLAMP 6.2 (7). Output (voltage) signals were low-pass-filtered at 700 Hz (3 dB) with an eight-pole, Bessel-type filter (Frequency Devices, Haverhill, MA). Single channel current tracings were further filtered for display purposes only. Unless otherwise stated, the software pCLAMP, Ver. 10.0 (Axon Instruments, Foster City, CA), was used for data analysis and the software SIGMAPLOT, Ver. 11.0 (Jandel Scientific, Corte Madera, CA), was used for statistical analysis and graphics. Unless otherwise stated, all tracings shown in this study were obtained at holding potentials between 40 and 60 mV. PC2 channel identification was conducted as previously reported in González-Perrett et al. (7). Statistical significance was obtained by unpaired Student’s t-test comparison of sample groups of similar size, and accepted at p < 0.05. Average data values were expressed as the mean ± SE (N) under each condition, where n represents the total number of experiments analyzed.

Results

Effect of Ca²⁺ chelation on PC2_hst channel function

To assess the effect of cytoplasmic Ca²⁺ on PC2_hst channel function, hST apical vesicles were reconstituted in the presence of a KCl chemical gradient, (150 mM KCl in cis, and 15 mM trans), with symmetrical Ca²⁺ (10 μM), and pH 7.4 (Fig. 1 a, Ctrl). Once spontaneous channel activity was observed, EGTA (1 mM) was added to the cis chamber to reduce the free Ca²⁺ concentration to ∼0.6 nM. Ca²⁺ chelation decreased the PC2_hst-mediated K⁺ mean currents by 86% (6.25 ± 1.8 vs. 0.88 ± 0.08 pA, N = 24, p < 0.05) with a t_1/2 of 3.6 min (Fig. 1, a and c). The EGTA inhibition was never complete, leaving a remainder current (∼14%, Fig. 1 c). A lag in the response was usually observed, before a decrease in current was noticeable after EGTA addition (Fig. 1 c). The inhibitory effect of lowering cis Ca²⁺ on PC2_hst channel function was confirmed with another Ca²⁺-chelating agent, BAPTA (2 mM), which produced a faster (t_1/2 = 0.4 min) and complete (100%, N = 8) inhibition of channel function (Fig. 1 c). It is important to note that higher concentrations of EGTA never reproduced the effect of BAPTA, nor achieved complete inhibition but instead generated membrane instability (data not shown). Thus, only 1 mM EGTA data are reported herein.

Figure 1.

Effect of Ca²⁺ chelation on PC2 channel function. (a) Representative single channel tracings of reconstituted PC2_hst. Addition of cis EGTA (1 mM) inhibited channel function after 5 min (N = 24). (b) Representative single channel tracings of reconstituted PC2_iv as in panel a. Addition of EGTA (1 mM) to the cis chamber had no effect on channel function (N = 5). (c) Relative current (I_m/I_ctrl) after addition of either EGTA (1 mM, solid circles) or BAPTA (2 mM, open circles) to PC2_hst and EGTA (1 mM, triangles) to PC2_iv. Experimental data are the mean ± SE of 24, 8, and 5 experiments, respectively. (d) All-point histograms of single channel tracings from panel a.

Effect of cis Ca²⁺ chelation on PC2_iv channel activity

The above results indicated the presence of Ca²⁺ regulatory sites in PC2_hst that control its function. Thus, PC2_iv was also reconstituted in a lipid bilayer system. Spontaneous PC2_iv single channel currents were neither modified by addition of EGTA (1 mM) to the cis side (8.1 ± 0.4 pA vs. 8.2 ± 0.3 pA, N = 5, Fig. 1, b and c) nor by BAPTA (2 mM) (data not shown). Actually, the PC2_iv protein did not respond to high Ca²⁺ in the trans compartment either (see Table S1 in the Supporting Material). Thus, in contrast to PC2_hst, Ca²⁺ chelation did not affect single channel currents through the PC2_iv isolated protein.

PC2_hst function recovery by Ca²⁺addition to the cis chamber

To determine whether the inhibition of PC2_hst by Ca²⁺ chelation was reversible, the cis chamber was titrated with increasing concentrations of Ca²⁺ in the presence of chelator (Fig. 2 a). The cis free Ca²⁺ was correlated with the recovery of PC2_hst channel function that was obtained as fractional currents after either EGTA or BAPTA conditions (Fig. 2, a and b). Inhibition from either chelator was entirely reversible. Mean recovery data as a function of [Ca²⁺]_cis for both chelating agents were best fitted with a Hill-type equation (18) (Fig. 2 b) that followed

$$\frac{I_m}{I_{\max }}=\frac{{\left[\text{Ca}\right]}^{n_H}}{{\left[\text{Ca}\right]}^{n_H}+K_D^{n_H}}+I_b,$$ (2)

where K_D is the Ca²⁺ apparent dissociation constant for the binding sites associated with the n_H, namely the Hill coefficient, and I_b and I_max are the basal (Ca²⁺ insensitive) and maximal currents, respectively. The results showed a K_D = 5.01 ± 0.06 nM (I_b = 0.08, n_H = 4.0 ± 0.3) in the presence of EGTA, and a K_D = 1.75 ± 0.13 nM (I_b = 0.04, n_H = 4.0 ± 0.2) in the presence of BAPTA (Fig. 2 b). These results suggested the presence of at least four regulatory Ca²⁺ binding sites of PC2_hst channel function, although n_H does not represent the actual number of binding sites. In its simplified form, the Hill equation describes an allosteric interaction between originally identical, unoccupied binding sites (18,19). Changes in affinity occur as these sites become increasingly occupied. Thus, n_H is, strictly speaking, a measurement of cooperation, not of the actual number of binding sites (or their intrinsic binding affinity). The simplified Hill equation will have an n_H, which will only represent the actual number of binding sites, as positive cooperativity becomes very strong (n_H → n). Strong cooperation was suggested by the nonlinear Scatchard and Hanes plots (Fig. 2, c and d) and calculation of R_s which, as defined by Bisswanger (19), is the ratio of ligand concentration at 90 and 10% saturation, respectively, with a value of 81 for a normal hyperbolic saturation curve. The value R_s decreases with the strength of positive cooperativity. We defined R_s as the ratio between 90 and 10% saturation values from the titration curves, respectively (18,19). The R_s values, 4.31 and 3.44 obtained after recovery from EGTA and BAPTA inhibition, respectively, were in agreement with a strong positive cooperativity (R_s ≪ 81 (18,19)).

Figure 2.

Recovery by Ca²⁺ of PC2_hst channel function after Ca²⁺ chelation. (a) Time response of current change after addition of different Ca²⁺ concentrations to the cis chamber after inhibition by EGTA. (b) Recovery curves as a function of cis Ca²⁺ concentration, after addition of either EGTA (solid circles) or BAPTA (open circles). Experimental data are the mean ± SE of seven and five experiments, respectively. (Solid lines) Best fitted Hill-type equation. Scatchard (c), and Hanes (d) plots of fitted EGTA (black line) and BAPTA (shaded line) recovery curves. (Dashed lines) Mean ± SE of fitted results.

We further evaluated an expanded Hill equation considering that if ligand binding cooperativity is very marked, then the fraction of all channel-ligand states containing fewer than n molecules of ligand would be negligible at any concentration that is appreciable compared to the intrinsic constant, K_s (18). In such cases, the Hill equation may be rewritten with the assumption that n_H = 4, such that

$$\frac{I}{I_{\text{max}}}=\frac{{\left[\text{Ca}\right]}^4}{{\left[\text{Ca}\right]}^4+a^3{b}^{2}c{K}_s^4}+I_b,$$ (3)

where a, b, and c are factors by which the intrinsic binding constants have been increased in each step, and a³b²cK_s⁴ = K_D⁴, in Eq. 3. The parameters obtained from the recovery dose-response fitted data to Eq. 3 are shown in Table S2, where the highest dissociation constant for the first binding site of either EGTA or BAPTA conditions were 111.6 and 105 nM, respectively.

Diffusional limitation corrections

Although Ca²⁺ binding interactions to the putative binding sites after recovery from inhibition by either chelator were almost identical (see Table S2), current decay (Fig. 1 c) was faster and stronger in the presence of BAPTA as compared with EGTA (7:1). This difference could not be explained by either binding interaction or by the forward rate constants of the chelators, which relate as 100:1 (20). Thus, we explored the possibility that a diffusional limitation of the chelators could be limited to the Ca²⁺ binding sites. Thus, we corrected the recovery curves by a transport coefficient (h) ((19), also see the Supporting Material), considering the diffusive properties of EGTA and BAPTA (20,21). This coefficient represented the existence of a diffusional layer between the channel and bulk Ca²⁺, which would have distinct diffusive properties for either chelator. Under these conditions, diffusional Ca²⁺ flow (J_diff) from the bulk solution (C₂) to the diffusional limitation layer (C₁) would be described as

$$J_{\text{diff}}=h\left(\left[C_2\right]-\left[C_1\right]\right)=\frac{D_C\left(\left[C_2\right]-\left[C_1\right]\right)}{\delta },$$ (4)

where δ represents the layer thickness, and D_C the diffusion constant of the chelating agent (20). Once Ca²⁺ accesses the regulatory sites, PC2_hst activation will follow the function described by Hill (Eq. 2). Under steady-state conditions, J_diff and the activation process would occur at the same velocity (19), such that

$$h\left(\left[C_2\right]-\left[C_1\right]\right)=\frac{V{\left[C_1\right]}^n}{K_D^n+{\left[C_1\right]}^n}.$$ (5)

To estimate h, we plotted I_m/I_max versus time after Ca²⁺ addition to the cis chamber following either EGTA or BAPTA inhibition, respectively (see Fig. S1 a in the Supporting Material). The slopes of the curves were plotted versus the cis Ca²⁺ concentration (see Fig. S1 b). The transport coefficients obtained, 4.3 ± 1.3 × 10⁻³ s⁻¹ nM⁻¹ (r = 0.82) and 2.5 ± 1.7 × 10⁻³ s⁻¹ nM⁻¹ (r = 0.58) for EGTA and BAPTA, respectively, were not statistically different from each other (p > 0.05), indicating that the two chelators did not display any relevant diffusional differences. Even further correction taking into account the actual diffusional coefficient ratio of the chelators (D_EGTA/D_BAPTA = 1.1), based on their difference in molecular size (molecular mass 380.35 vs. 476.43 g/mol for EGTA and BAPTA, respectively), did not render further differences, but actually favored slightly EGTA over BAPTA (data not shown). These coefficients were corrected considering that Ca²⁺ was added at an average distance (d) of 500 μm from Ca²⁺ regulatory sites (approximately center of the cuvette), and h could then be calculated as

$$h=m_{i}d,$$ (6)

where i represents either chelator.

The h values obtained were 2.15 μm s⁻¹ and 1.25 μm s⁻¹, for EGTA and BAPTA, respectively. The corrected PC2_hst reactivation curves (see Fig. S1 c), following Bisswanger (19), showed a left shift in the corrected K_D (K_D′) that was greater in the case of EGTA than of BAPTA, with corrected values of 4.70 ± 0.02 nM (n_H = 4.1 ± 0.07), and a K_D of 1.26 ± 0.03 nM (n_H = 4.5 ± 0.04) for Ca²⁺, respectively. Thus, the values fell within the experimental error, such that the diffusional contribution would be negligible for either chelator.

Effect of external Ca²⁺ on PC2_hst channel function

To explore whether external Ca²⁺ also had an effect on PC2_hst channel function, we increased the Ca²⁺ concentration from the trans side after channel inhibition with EGTA. A rise in trans Ca²⁺ rapidly increased PC2_hst channel function, which reached a peak current higher (240%) than control (before EGTA addition) at ∼5 mM Ca²⁺ (Fig. 3 a). Further addition of external Ca²⁺ (10–50 mM CaCl₂) decreased the mean currents through PC2_hst. In the presence of cis BAPTA, however, increasing concentrations of trans Ca²⁺ did not restore PC2_hst channel function (Fig. 3 b). These findings raised the hypothesis that the differences in inhibition elicited by either chelator may be associated with a remainder channel function, as in the case of EGTA, which should drive Ca²⁺ through, from the trans compartment. This was explored in the next section.

Figure 3.

Effect of trans Ca²⁺ on PC2_hst function. (a) Kinetics of the I_m/I_ctrl ratio under control conditions (Ctrl), after cis EGTA addition, followed by increasing Ca²⁺ concentrations added to the trans side. Experimental data (solid circles) are the mean ± SE (N = 5). (b) Similar data as in panel a, after addition of BAPTA. Experimental data (solid circles) are the mean ± SE (N = 3).

Effect of Ca²⁺ chelation at different initial trans Ca²⁺ concentrations

To further explore the changes observed in PC2_hst channel function by titration of external Ca²⁺, experiments were also designed to test the effect of the cis chelator at a given initial trans Ca²⁺ concentration. The experiment was initiated with a given trans Ca²⁺ already present in the chamber before addition of the cis chelator. In the presence of low external Ca²⁺ (0.6 nM), addition of EGTA (Fig. 4 a) elicited a rapid and almost complete inhibition with t_1/2 = 0.4 min, which was rather similar to the BAPTA effect with 10 μM Ca²⁺ trans. Conversely, in the presence of high (1 mM) external Ca²⁺, cis EGTA only had a transient and incomplete effect, reaching 50% inhibition (Fig. 4 a). Because the BAPTA effect was fast and complete in symmetrical 10 μM Ca²⁺, we also tested the effect of BAPTA while in the presence of high (1 mM) trans Ca²⁺, which only elicited a transient inhibition of the currents by 20%, returning to control values in ∼1 min (Fig. 4 b). Thus, the inhibitory effect of cis Ca²⁺ chelation (with either chelator) on PC2_hst channel function strongly depended on the trans Ca²⁺ concentration. This would suggest that the Ca²⁺ chemical gradient, and thus permeation through the channel pore, modifies the availability of Ca²⁺ in the proximity to the regulatory sites in the cis compartment.

Figure 4.

Effect of trans Ca²⁺ on PC2_hst channel inhibition. (a) Kinetics of channel recovery (I_m/I_ctrl) after EGTA addition at different initial trans Ca²⁺ concentrations. Experimental data were obtained in 1 mM (triangles) and 0.6 nM (stars) Ca²⁺ conditions, respectively. Experimental data are expressed as mean ± SE (N = 3). (b) Kinetics of channel recovery (I_m/I_ctrl) after BAPTA addition in the presence of 1 mM trans Ca²⁺ concentration (triangles). Experimental data are expressed as mean ± SE (N = 3). (Dashed lines) Values obtained in the presence of 10 μM trans Ca²⁺ (Fig. 1c).

Calculation of Ca²⁺ flux coefficient through PC2_hst

The results obtained in the presence of different trans Ca²⁺ concentrations allowed us to determine the contribution PC2_hst made to Ca²⁺ transport and its delivery to the cis side. Ca²⁺ transport through the channel was calculated as the rate of PC2_hst channel recovery. To this end, we increased the trans Ca²⁺ concentration after cis EGTA channel inhibition, and then obtained the fractional recovery current I_m/I_max as a function of [Ca²⁺]_trans (Fig. 5 a, left). The relationship was well approximated by a Michaelis-Menten-type function with an apparent dissociation constant of 2.8 ± 0.92 mM and n = 1.0 ± 0.4. The relationship t_1/2 versus [Ca²⁺]_trans evidenced a linear function (Fig. 5 b). Because the results with cis Ca²⁺ titration indicated that the regulatory effect of Ca²⁺ on PC2_hst was only possible from the cis side, we compared the I_m/I_max for both [Ca²⁺]_trans and [Ca²⁺]_cis titrations, thus obtaining the ratio between trans and cis Ca²⁺ to reach similar recovery values (i.e., Fig. 5 a, dashed line). This recovery ratio was thus associated with a Ca²⁺ influx (J_Ca) for each trans Ca²⁺ concentration. Further, the trans Ca²⁺ concentration was inversely correlated with the average time it took for the channel to recover to a given value. We next obtained a curve of cis Ca²⁺ versus trans Ca²⁺ (to obtain the same fractional current) divided by the time necessary to reach a steady recovery response (Fig. 5 c). The best fitted theoretical curve had a slope of 2 × 10⁻⁵ s (Fig. 5 c), which represents the Ca²⁺ flux coefficient (t_Ca) through the PC2_hst channel. A summary of conclusions obtained from these calculations, including the Ca²⁺ permeability at physiological Ca²⁺ gradients, is given in the Discussion (see below).

Figure 5.

Recovery of PC2_hst channel activity by either cis or trans Ca²⁺. (a) (Left) PC2_hst function (I_m/I_max) recovery curve in the presence of increasing trans Ca²⁺ concentrations. (Right) Recovery curve in the presence of increasing cis Ca²⁺ concentrations (from Fig. 2b). (Dashed line) I_m/I_max = 0.75 reached in both graphs. (b) Experimental t_1/2 as a function of trans Ca²⁺ concentration. The curve was fitted with t_1/2 = m [Ca²⁺]_trans + t₀, where m = −100 min.M⁻¹ and t₀ = 0.40 min. (c) Correlation between cis and trans Ca²⁺ concentrations generating similar I_m/I_max ratios.

Phenomenological model

The encompassed results were explored in terms of a phenomenological model (Eq. 7) that approximated the temporal response of the PC2_hst currents (I_m/I_ctrl) after addition of either EGTA or BAPTA for the various cis and trans Ca²⁺ concentrations, as

$$\frac{I_m}{I_{\text{ctrl}}}=A+Bt-Ct{e}^{-Dt},$$ (7)

where the parameters A, B, C, and D are constants, and t is the time after addition of the chelator. The fitting parameters are shown in Table S3. This model considered that the overall kinetics of current decay after chelation would occur due to two simultaneous processes taking place after addition of the cis chelator. The first contribution would be a deficit in local Ca²⁺ which, in turn, modified the Ca²⁺ bound to the regulatory sites associated with, but not intrinsic to, the PC2_hst channel. A second contribution to this local pool is associated with an entry step feeding Ca²⁺ through the open channel. Thus, the Ca²⁺ associated with the regulatory binding sites would primarily be determined by:

• 1.The initial Ca²⁺ concentration (represented by A, relative value);
• 2.Ca²⁺ entry through the channel (Bt); and
• 3.Ca²⁺ retrieval from the regulatory sites as the result of the chelating reaction and diffusion (Cte^−Dt).

This equation best approximated all experimental data from the various conditions tested (Fig. 6, a and b).

Figure 6.

Phenomenological model. (a) Fitted curves obtained with the phenomenological model (Eq. 7) for EGTA additions. Experimental data were obtained after cis EGTA addition, for different trans Ca²⁺ concentrations, 1 mM, 10 μM, and 0.6 nM (triangles, circles, and stars, respectively). (b) Fitted curves obtained for BAPTA conditions and two different trans Ca²⁺ concentrations, 1 mM, and 10 μM (triangles and circles, respectively). The fitting parameters obtained are listed in Table S3 in the Supporting Material. (c) Diagram of Ca²⁺ interactions with PC2_hst. Ca²⁺ flows through the channel pore. The mobile chelating agents (Q) compete for a local pool of Ca²⁺ (fixed buffer) on cis side, contained in a microdomain accessible from the trans side while the channel is open. PC2_hst channel function will depend on the Ca²⁺ level reached in the cis microenvironment by interaction with the Ca²⁺ binding region (CBR).

Discussion

The human placenta transfers ∼30 grams of Ca²⁺ during late gestation. The mechanisms implicated in this transepithelial transport are the subject of active research (4,22,23). PC2 is implicated in Ca²⁺ transport in Ca²⁺ handling epithelia including the kidney and the placenta. The first functional characterization of PC2 as a Ca²⁺-permeable nonselective cation channel was made in hST from term placenta (7). Little is known, however, as to how Ca²⁺ controls PC2 function in this epithelium. Herein, we explored the regulation by Ca²⁺of PC2_hst from apical hST membranes we have previously characterized (7). We observed that a decrease in free cis Ca²⁺ to subnanomolar levels obtained with either EGTA or BAPTA extensively inhibited PC2_hst channel function. Current decay after chelation with EGTA (symmetrical 10 μM Ca²⁺) was slower and incomplete, compared with the inhibition observed with BAPTA.

Surprisingly, similar experiments conducted with PC2_iv showed no changes in channel activity. This discrepancy suggests that the PC2_hst channel complex is somewhat different from the isolated protein. A difference may relate to a channel complex in the hST preparation that contains other channel subunits that are absent in the in vitro material and/or themselves regulated by Ca²⁺. TRP channel monomers other than PC2 have been reported to interact with PC2 (24). In such a case, the heterocomplex may respond to Ca²⁺ in a distinct manner not observed for the isolated protein. We previously have compared PC2 channel function from various preparations in membranes, including the hST, with isolated protein from PC2-overexpressing Sf9 cells (7) and renal epithelial cells (25,26) as well as the in vitro translated material (7,24,27). We thereby identified a number of intrinsic features of the channel that were confirmed throughout. All these PC2 preparations essentially show the same functional channel properties. TRPC1-PC2 heterocomplexes, for example, show distinct structural-functional properties, including overall and subconductance states, as well as pH and amiloride sensitivity. Thus, it is likely that the PC2_hst channel may actually present distinct associated properties instead, which modified Ca²⁺ binding, but not its biophysical properties.

Changes in Ca²⁺ are already known to regulate PC2 (16,26), although these studies were conducted with channel preparations that may contain PC2-associated proteins as well. Ehrlich’s group recently found Ca²⁺-dependent conformational changes of the isolated carboxy terminus of PC2 (27), supporting the contention that the PC2 channel has Ca²⁺-interacting domains. It is presently unknown as to whether this phenomenon conveys any functional properties to the complete channel complex. Our contemporary observations suggest that the cis Ca²⁺ regulatory sites were not intrinsic to the channel, but instead related to associated proteins. The actual Ca²⁺-binding linkers implicated in this regulation remain to be identified. Nonetheless, we have preliminary evidence that the PC2_iv channel can be made responsive to Ca²⁺ in a manner rather similar to the PC2_hst material, simply by adding α-actinin (M. d. R. Cantero and H. F. Cantiello, unpublished). This is in agreement with previous studies (17,28), which demonstrated a direct structural-functional interaction between PC2 and the actin-associated protein α-actinin, which binds to, and modulates PC2_hst channel function (28). Further, F-actin severing proteins, such as gelsolin, also regulate PC2_hst in a Ca²⁺-dependent manner (17). Thus, cytoskeletal structures present in apical hST vesicles may provide a kinetic compartment, forming a Ca²⁺-dependent microdomain that controls PC2 channel function.

Number of regulatory Ca²⁺ binding sites associated with the PC2 channel

The reversal from inhibition by titration with Ca²⁺ after chelation followed a well-approximated Hill equation. The n_H obtained in the presence of either chelator, was consistent with at least four Ca²⁺ regulatory sites, with apparent K_D in the range of 1–5 nM. It is important to mention that although the K_D and n_H in Eq. 2 are mathematically possible, Hill assumptions are not physically possible except for the condition in which n = 1 (29). Thus, evaluation of the intrinsic dissociation constant for the first site, K_S, was necessary to better understand the potential effect of Ca²⁺ at physiological levels. A strong positive cooperativity between the Ca²⁺ binding sites was observed by the nonlinear curves in the Scatchard and Hanes plots, and confirmed by the large shift in R_s to values much lower than 81 (4.31 and 3.44 for EGTA and BAPTA, respectively; see Results) (18,19). Although n_H will be between the limits 1 < n_H < n, it is also important to note that the more pronounced the positive cooperativity, the closer n_H will approach n, which is the true number of binding sites, and when cooperativity is maximal, n_H equals n. We further explored the decrease in the affinity constant for the first binding by expansion of the Hill equation, considering the interaction factors (Eq. 3). The summarized data in Table S2 clearly indicated that the first occupied site has an affinity of 105–111.6 nM, which is consistent with basal cytosolic Ca²⁺, and an increase in affinity between 10 and 100 times for the other sites, as expected (18,19).

Differences between chelator responses

There were clear kinetic differences between the electrical responses of BAPTA and EGTA in symmetrical Ca²⁺ (10 μM). BAPTA was seven-times faster (3.6 vs. 0.4 min) than EGTA. This phenomenon could not be explained by differences in the forward rate constant (k_on) for either chelator, because the BAPTA k_on rate is a hundred times faster than that of EGTA (20). Moreover, EGTA produced a partial inhibition whereas BAPTA completely inhibited PC2_hst channel function. Despite the fact that the experiments were conducted in a bilayer reconstitution system which may be envisioned as an entirely exchangeable infinite reservoir, the constraints of a kinetic compartment were in agreement with previous findings in cell models (30,31). The fast exocytotic response in melanotropic cells, a phenomenon coupled to voltage-gated Ca²⁺ channels, is very sensitive to mobile Ca²⁺ buffers, such that EGTA elicits a complete block of the response (31). Kits et al. (31) found that BAPTA blocked exocytosis only twice as effectively as EGTA. Using computer simulations, Kits et al. (31) demonstrated that these results could be explained neither by free diffusion nor by the binding rates of the mobile buffers, but instead that saturation of local Ca²⁺ buffers was involved. In their studies, a diffusional barrier for both Ca²⁺ and buffer molecules reduced diffusion 1000–10,000 times to generate similar local Ca²⁺ concentrations for specific concentrations of EGTA and BAPTA. Kits et al. concluded that when the effects of a fixed buffer are taken into consideration, the experimental results are not explained by diffusion or mobile buffer properties. This is qualitatively in agreement with our findings that show that the transport coefficient (h) obtained to correct the diffusional contribution of Ca²⁺ to both chelators was unable to explain the experimental differences for either chelator. Local buffer saturation may occur due to the existence of diffusional barriers formed by a kinetic constraint to Ca²⁺ binding. Thus, the influence of mobile Ca²⁺ buffers on Ca²⁺-dependent processes far from the site of Ca²⁺ entry may actually depend largely on submembranous compartmentalization (31).

Other studies (32,33) similarly showed that it is not the difference in binding rate but rather the difference in affinity constants that is relevant in the Ca²⁺ chelation by EGTA and BAPTA. In general, fixed buffers will slow down and prolong the occurrence of Ca²⁺-dependent processes because they contribute to initial chelating of ions but release Ca²⁺ when free levels drop, thus allowing for extended duration of the process (34). Naraghi and Neher (20) showed that fixed buffers do not affect the steady-state concentration of Ca²⁺ in the microdomain, but somewhat prolong the time course leading to the steady-state condition. Only very high levels of fixed buffer cause a drop in Ca²⁺ concentration very close to the membrane. Fixed buffers then will become dominant over mobile buffers, which results in equal effects of different concentrations of EGTA and BAPTA close to the membrane. This is particularly relevant in the context of our findings, which require kinetic constraints to slow down the response in the range of minutes, several orders-of-magnitude slower than either diffusion or binding rates.

Ca²⁺ transport through the PC2 channel

The differences in response between cis EGTA and BAPTA in the presence of 10 μM trans Ca²⁺ were twofold. On the one hand, there was a faster inhibition kinetics with BAPTA than EGTA; on the other hand, the inhibition with EGTA was incomplete, leaving a remainder channel function. This remaining activity proved to be efficient in reversing the inhibitory effect of the chelator by titration with trans Ca²⁺. Failure to recover channel function by addition of trans Ca²⁺ after complete inhibition by cis BAPTA suggested that the Ca²⁺ regulatory sites were only accessible from the cis side, and thus inaccessible once the channel closed. Titration after (cis) chelator inhibition therefore depended on both the chelator and the concentration of trans Ca²⁺ at the beginning of the experiment. This was confirmed by the second set of experiments in which cis Ca²⁺ chelation was conducted at different initial trans Ca²⁺ concentrations. In the presence of higher trans Ca²⁺ (1 mM), for example, cis EGTA only elicited a transient and lower inhibition. Similarly, cis BAPTA inhibition in higher initial trans Ca²⁺ (1 mM) was also transient, resembling the EGTA effect at lower trans Ca²⁺. This is consistent with a scenario whereby PC2-mediated Ca²⁺ fed through the channel into the cytoplasmic domain would help maintain channel function. This contribution could be quantified from the inverse relationship observed between t_1/2 of the activating response, and the external Ca²⁺ concentration, such that the higher the Ca²⁺ gradient, the faster the saturation of the cis Ca²⁺ regulatory sites.

From the cis-to-trans Ca²⁺ response ratio, we obtained a flux coefficient (t_Ca) of 2 × 10⁻⁵ s, reflecting the average time it took for Ca²⁺ to pass through the pore (trans→cis), then access and activate the cis-located Ca²⁺ regulatory binding sites. The t_Ca also provided a tool for estimating the Ca²⁺ flux (J_Ca) through the PC2_hst channel pore at physiologically relevant Ca²⁺ concentrations. To turn this value into either a meaningful J_Ca or current (I_Ca), we created a geometrical model of the channel pore, envisioned as a cylinder of area π × r², and volume π × r² × L. We assumed an inner radius of the channel to be at least 5 nm, based on its cationic permeability properties (33). Other dimensions included a maximum diameter of 20 nm and total length of 10 nm, based on a 1:5 total ratio, and 10 times the protruding section of the channel (L), respectively, as we recently determined by atomic force microscopy (35). Thus, the internal volume (V_pore) that a given ion should travel would be between 0.079 and 1.25 × 10⁻¹⁷ cm³. The theoretical Ca²⁺ permeability of the pore could then be estimated by

$$P_{\text{Ca}}=\frac{V_{\text{pore}}}{t_{\text{Ca}}},$$ (8)

where P_Ca is the Ca²⁺ permeability coefficient in this concentration range. The calculated P_Ca was between 0.39 and 6.28 × 10⁻¹³ cm³/s. Thus, J_Ca through PC2_hst would then be obtained as postulated by Läuger (36) such that

$$J_{\text{Ca}}=P_{\text{Ca}}\left[\frac{za}{2\sinh \left(a/2\right)}\right]\left({\left[\text{Ca}\right]}_{trans}{e}^{\frac{a}{2}}-{\left[\text{Ca}\right]}_{cis}{e}^{-\frac{a}{2}}\right),$$ (9)

where a = RTV/zF and V is the applied holding potential. The value obtained, J_Ca, in the range of 9.32 × 10⁻²⁰ mol/s and 1.48 × 10⁻¹⁸ mol/s, would be consistent with a PC2_hst Ca²⁺ conductance of ∼0.15–2.38 pS at 1 mM external Ca²⁺. These results were in agreement with a boundary condition obtained by calculation of the maximal Ca²⁺ permeability $P_{\text{Ca}}^{\text{max}}$, as estimated by Läuger (36),

$$P_{\text{Ca}}^{\text{max}}=2\pi r_o{D}_{\text{Ca}},$$ (10)

where r_o is a hemispherical surface parameter that represents an effective capture radius. Considering an r_o = 0.1 nm (36) and the Ca²⁺ diffusion coefficient of 660 μm²/s (20), the $P_{\text{Ca}}^{\text{max}}$ and $J_{\text{Ca}}^{\text{max}}$ would be 4.15 × 10⁻¹³ cm³/s and 3.92 × 10⁻¹⁵ mol s⁻¹, respectively, giving a maximal conductance of 6.30 pS. This value represents a very high Ca²⁺ transport rate compared to most Ca²⁺-permeable channels under the imposed electrochemical conditions (37). In addition, we validated the Ca²⁺ transport data by direct electrodiffusional experiments (see the Supporting Material). We obtained current-to-voltage (I/V) relationships of PC2_hst single channel currents under biionic conditions with a cis→trans K⁺ gradient and a high trans Ca²⁺ concentration (90 mM), such that Ca²⁺ currents from trans to cis could be determined (see Fig. S2). The I/V data obtained were well fitted with both the Goldman-Hodgkin-Katz equation and an absolute rate theory model containing two internal Ca²⁺ binding sites within the pore of the channel (38). The K_D values calculated were used to obtain a Michaelis-Menten-type g_PC2 versus Ca²⁺_trans curves. A Ca²⁺ single channel conductance in the range of 0.12–1.33 pS was obtained for 1 mM trans Ca²⁺, in close agreement with 0.15–2.38 pS, obtained from the Ca²⁺ transfer data.

Size of the Ca²⁺ microdomain

From the $J_{\text{Ca}}^{\max }$ range obtained under our experimental conditions, and the average time it took for Ca²⁺ to attain a given activation response, say an increase in channel activity from 20 to 80%, it could be estimated that the delivery of ∼5 × 10⁻¹⁷ moles of Ca²⁺ to the cis side would be required. This estimation suggests a local volume associated with the kinetic differences between EGTA and BAPTA of ∼8–10 nL, which reflects a large microdomain associated with the PC2_hst channel. Considering the geometry and expected diffusional properties of the cis chamber, it is expected that this microdomain is most likely associated with the cytoskeletal network linked to the channel in the hST apical vesicles (as seen in Montalbetti et al. (17)).

Phenomenological model

The results in this report were best followed by a phenomenological model (Eq. 7) describing the kinetic changes in PC2_hst channel activity as a function of Ca²⁺, and entails, as indicated above, the presence of a local pool of Ca²⁺ (microdomain) buffering Ca²⁺ regulatory sites associated with the cytoplasmic side of the PC2_hst channel, whose saturation would depend on two contributing processes: the high affinity binding sites of an associated protein and the rate of Ca²⁺ delivered through the channel. The first contribution is evidenced as Ca²⁺ is depleted by addition of the chelating agent. The other (compensating) contribution lies in the Ca²⁺ entry through the channel pore. This is described as a linear parameter related to the electrochemical Ca²⁺ gradient. The parameters obtained using this model (see Table S3) could be interpreted as the contribution of both Ca²⁺ input and output to and from the microdomain near the channel, respectively.

Competition between Ca²⁺ influx through the channel, and Ca²⁺ chelation by either mobile buffer, determines the degree of depletion (or not) of the microdomain, and therefore the effected system, PC2_hst function. Thus, in the presence of BAPTA and 10 μM cis Ca²⁺, the contribution of the exponential term is larger, compared to that obtained in the presence of EGTA. This indicates a greater depletion of the microdomain in the presence of BAPTA. In the presence of BAPTA and 1 mM trans Ca²⁺, however, the absence of channel inhibition would reflect the larger contribution of Ca²⁺ entry through the channel, from the trans side. Moreover, in the presence of either chelator and 1 mM trans Ca²⁺, the linear term of Eq. 3 representing Ca²⁺ influx through the channel is greater than in the presence of lower concentrations (0.6 nM and 10 μM). Thus, Ca²⁺ transport (trans → cis) through the channel largely contributes to maintain the channel open (Fig. 6 c).

Conclusions

The data in this report indicate that a local cytoplasmic Ca²⁺ binding region associated with PC2 (CBR, Fig. 6 c) regulates its channel activity in the hST preparation. This Ca²⁺ regulatory region is absent in the isolated protein, and represents a Ca²⁺ binding protein(s) that associate with, and confer Ca²⁺-dependence to, the channel. Ca²⁺ feed through the channel would replete and control this Ca²⁺ microdomain. The data suggest a high intrinsic Ca²⁺ permeability by PC2.