PD173212

Diverse Calcium Channel Types are Present in the Human Placental Syncytiotrophoblast Basal Membrane

Abstract

The functional expression of calcium channels has been scarcely studied in human placental syncytiotrophoblast. We have presently sought to characterize Ca2+ currents of the healthy syncytiotrophoblast basal membrane using purified basal membranes reconstituted in giant liposomes subjected to patch-clamp recordings.

We detected presence of channels with high permeability to Ca2+ (relative PCa/PK up to 99.5) using K+ solutions in symmetric conditions. Recordings performed in Ba2+ gradients showed Ba2+-conducting channels in 100% of experiments. Ba2+ total patch currents were consistently blocked by addition of NiCl2, Nifedipine (L-type voltage-gated calcium channel blocker) or Ruthenium Red (TRPV5eTRPV6 channel blocker); Nifedipine and Ruthenium Red exerted a synergic blocking effect on Ba2+ total patch currents. Immunohistochemistry of placental villi sections evidenced presence of a1 subunit of voltage-gated calcium channels and TRPV5eTRPV6 channels in basal and apical syncytiotrophoblast plasma membranes; these three calcium channels were also detected in purified basal and apical fractions using Western blot. These results show the presence of three types of calcium channels in the syncytiotrophoblast basal membrane by both functional and molecular means. These basal membrane calcium channels would not be directly involved in mother-to-fetus Ca2+ transport, but could participate in other relevant tro- phoblast processes, such as exocytosis and Ca2+ transport regulation.

Keywords: Placenta; Syncytiotrophoblast; Basal membrane; Calcium channels

1. Introduction

As in all other cell types, Ca2+ appears to play a role as a second messenger in the human syncytiotrophoblast. This has been suggested by multiple studies, particularly consider- ing the participation of Ca2+ in hormone secretion by cultured trophoblast or placental explants [1e11]. Participation of L-type voltage-gated calcium channels (VGCC) in hormone secretion was studied in eight of the 11 mentioned studies by adding specific blockers (dihydropyridines and phenylalkyl- amines) to the culture media; in seven of these eight studies blocker addition resulted in suppression of intracellular Ca2+ concentration rise and hormone secretion, suggesting partici- pation of these channels in the secretory process. Some studies using placental explants, however, do not suggest the partici- pation of VGCC in hormone secretion [58]. On the other hand, no functional evidence for the presence of VGCC was demonstrated by Ca2+ uptake in BeWo cells [12] and using patch-clamp recording [13] or cytofluorimetric analysis [14] in cultured trophoblast cells, leading to a persistent contro- versy regarding the nature of Ca2+ channels in the human tro- phoblast. More recently, L-type Ca2+ channel a1-subunit expression (a1C and a1D) has been detected using RT-PCR in cultured trophoblast cells [15]. However, as would be ex- pected, it seems that VGCC are not the only Ca2+-conducting channels expressed in trophoblast cells, as expression of sev- eral Ca2+-conducting channels from the TRP (Transient Receptor Potential) superfamily has also been recently demonstrated using both RT-PCR in cultured trophoblast [15,16], RT-PCR in villous placental tissue and immunohisto- chemistry of villous placental tissue [17]. TRPV5 and TRPV6 channel plasma membrane activity was suggested by block of basal Ca2+ uptake in cultured trophoblast cells when adding magnesium and Ruthenium Red, non-specific blockers for these channels [16]; TRPC channel plasma membrane activity was suggested by block of depletion-activated Ca2+ entry in term placental villous fragments when adding gadolinium and SKF96365, non-specific blockers for some TRPC channels [17].

According to the above mentioned studies, candidates for ligand-independent trophoblast Ca2+ entry belong to two main families of Ca2+ channels: VGCC e particularly L-type VGCC e and TRP channels.Voltage-gated calcium channels (VGCC) are highly selective for Ca2+ over monovalent cations, choosing Ca2+ over Na+ in spite of their identical diameter (relative PCa/PNa z 1000) [18]. Current nomenclature is based on the pore-forming sub- unit gene subfamily of the channel (a1 subunit, subfamilies 1e3: Cav1, Cav2, Cav3) [19], although the first recorded chan- nels were originally named according to the electrophysio- logical characteristics of their currents [20]: L-type (large conductance, lasting current; presently Cav1), T-type (tiny con- ductance, transient current; presently Cav3), N-type (neuronal; presently Cav2 together with P/Q-type and R-type VGCC). The pharmacology of the three a1 subunit subfamilies is a useful tool to distinguish among Ca2+ currents, as channel blockers are quite specific. For example, the dihydropyridine Nifedipine is a specific blocker for the Cav1 (L-type) subfamily [19].

TRP (Transient Receptor Potential) ion channel subunit genes were first defined in the Drosophila visual system and named after the transient response of the transmembrane volt- age potential of mutant retinal neurons to light in contrast to the sustained response in wild type [21]. Current nomenclature of the TRP superfamily includes three families based on se- quence homology and functional similarity: TRPC (Canoni- cal) family, TRPV (Vanilloid receptor) family, and TRPM (Melastatin) family. TRP channels characteristically behave as non-selective cation channels with an equivalent permeabil- ity to monovalent cations Na+ and K+ (relative PNa/PK = 1); they share, however, the property of being permeable to Ca2+ with relative PCa/PNa ≤ 10, with the exception of TRPM4 and TRPM5 which are monovalent cation selective, and of TRPV5 (also called ECaC1 or CaT2) and TRPV6 (also called ECaC2 or CaT1) which are highly Ca2+-selective (relative PCa/
PNa > 100) [22].

In spite of the compelling evidence for the presence of members of these two families of Ca2+-conducting channels in the syncytiotrophoblast plasma membrane, there are few studies that deal with the electrophysiological characterization of plasma membrane Ca2+ currents. Two different non-selective cation channels of the human apical syncytiotrophoblast mem- brane have been shown to permeate Ca2+ [23,24]; one of them (Polycystin-2) belongs to the TRP superfamily. There are presently no equivalent studies in the human syncytiotropho- blast basal membrane, although Ca2+ flux measurements of basal membrane vesicles have shown the presence of ATP- independent Ca2+ fluxes with channel-like kinetics [25].The aim of our present work has been to detect Ca2+-con- ducting channels in the basal syncytiotrophoblast plasma membrane and to study their electrophysiological and pharma- cological behavior, according to the current evidence above mentioned. These results will hopefully aid in the understand- ing of the role of Ca2+ as an important signalling factor in the placental epithelium.

2. Materials and methods

2.1. Purification of basal membranes

Purification of basal membranes was achieved using the method described in Jimenez et al. [26], which allows simultaneous isolation of apical and basal membranes from the same placenta, and is based on the method described by Illsley et al. [27] and modified by incorporating steps from our previous pro- tocol for obtaining apical membrane [28] and one step to isolate plasma mem- brane free of mitochondrial membranes [29].

Placentas obtained from normal term pregnancies were collected immedi- ately after delivery from the San Jose´ Hospital Maternity Unit and transported to the laboratory on ice. The purification method involved precipitation of basal membrane with magnesium ions, differential centrifugation and sucrose step gradients. All solutions were buffered with 20 mmol/L TriseHEPES, pH 7.4. A portion of the basal membrane-enriched preparation containing about 6e8 mg of protein was overlaid on the sucrose gradient. The band at the 47/52% (w/v) sucrose interface was collected and diluted 10-fold with 20 mmol/L TriseHEPES, pH 7.4, before centrifugation at 110 000 × g for 30 min. The final pellet was resuspended in 300 mmol/L sucrose, 20 mmol/ L Trisemaleate, pH 7.4, and stored in liquid nitrogen.

The purity and enrichment of the basal membrane fraction were deter- mined routinely by assaying for classical marker protein activities. Adenylate cyclase and b-adrenergic receptors were used as basal membrane markers, al- kaline phosphatase as an apical membrane contamination marker and cyto- chrome-c oxidase/succinate dehydrogenase as mitochondrial membrane contamination markers. Our purified membranes show purity and enrichment parameters comparable to those of other preparations reported for single or paired apical and basal membrane purification (see Jimenez et al. [26]). Be- cause the BM and MVM enrichments are particularly comparable with the
method described by Illsley et al. [27], we considered our purified membranes to contain low levels (<12%) of contamination by non-syncytial plasma mem- branes reported for the isolated membrane fractions. Purified basal membranes were enriched over 9-fold in adenylate cyclase activity relative to the micro- somal fraction and 9-fold in dihydroalprenolol binding to b adrenergic recep- tors relative to the tissue homogenate, and were essentially free of apical membranes and mitochondrial membranes. Lack of contamination of purified basal membranes with apical membranes and mitochondrial membranes was confirmed using immunoblotting for placental alkaline phosphatase and cyto- chrome-c oxidase, respectively. The densitometric analysis of Western blots (see Fig. 8B) usually gives a relative mark of alkaline phosphatase for pure BM fractions of about 5% of the mark detected in MVM. Accordingly, the en- zymatic activity for alkaline phosphatase is also low in BM: enrichment was only 1.5e2-fold relative to the whole tissue homogenates, while the enrich- ment of alkaline phosphatase activity for MVM fractions is usually in the range of 17e21-fold. The degree of contamination, quantitated by a ratio of alkaline phosphatase activity enrichment of BM compared to MVM (BM/ MVM), is 0.15 in our study, lower or equivalent to that from several other reports [26]. 2.2. Reconstitution of the purified basal membrane into giant liposomes Giant liposomes were prepared by submitting a mixture of the reconsti- tuted basal membrane vesicles and asolectin lipid vesicles to a partial dehydration/rehydration cycle, as reported by Riquelme et al. [30]. A mem- brane aliquot containing 100e150 mg of protein was mixed with 2 mL of a 13 mmol/L (in terms of lipid phosphorus) suspension of the asolectin vesi- cles. After the partial dehydration/rehydration cycle, the diameter of the result- ing giant multilamellar liposomes ranged from 5 to 100 mm. 2.3. Patch-clamp recordings Aliquots of 3e6 mL of giant liposomes were deposited into the excised Patch chamber (RC-28, Warner Instruments Corporation, USA) mixed with 0.5 mL of the buffer of choice for electrical recording (bath solution). Sin- gle-channel recordings were obtained by patch-clamp techniques as described by Hamill et al. [31]. Giga seals were formed on giant liposomes with glass microelectrodes of 5e10 MU resistance. After sealing, withdrawal of the pipette from the liposome surface resulted in an excised patch. Current was recorded with an EPC-9 patch-clamp amplifier (Heka Elektronic, Lam- brecht/Pfalzt, Germany) at a gain of 50e100 mV/pA and a filter setting of 10 kHz. The holding potential was applied to the interior of the patch pipette, and the bath was maintained at virtual ground (V = Vbath — Vpipette). The bath was grounded via an agar bridge and the junction potential was compensated for when necessary. The signal was analyzed off-line using TAC (Bruxton Cor- poration), Pulse Fit (Heka, Lambrecht/Pfalz, Germany) and Microcal Origin 6.0 (Microcal Software, Inc., USA) software. All measurements were made at room temperature. 2.4. Pulse protocols Three main types of pulse protocols were used in our electrophysiological experiments: 1.- A 40 mV/s voltage ramps from —100 to +100 mV, which allowed mea- surement of experimental reversal potential (Erev) and of absolute currents generated by brief transmembrane holding voltages. This pulse protocol thus allowed the measurement of isolated Ba2+ currents (at VeqCl— = —55 mV) and of isolated Cl— currents (at VeqBa2+ = +76 mV) in pipette-to-bath BaCl2 concentration gradient conditions (solution com- position 3 in Section 2.5). 2.- Equilibrium potential pulses, consisting of 100 ms holding voltages at the Ba2+ theoretical equilibrium potential (VeqBa2+ = +76 mV) and at the Cl— theoretical equilibrium potential (VeqCl— = —55 mV) in pipette- to-bath BaCl2 concentration gradient conditions. This pulse protocol also allowed measurement of isolated Ba2+ and Cl— currents. 3.- Increasing voltage pulses, consisting of 5 s holding voltages separated by a 1 s pulse at 0 mV holding voltage, starting at —80 mV and increasing up to +80 mV in steps of 40 mV. This pulse protocol was used only to de- tect Cl— channels in solutions of NMDGCl. 2.5. Solutions and blockers The pipette and bath solutions were prepared with de-ionized water (Ca2+ concentration 10 mmol/L, measured using a fluorometric method e Fluoromax ISA Instruments), and had the following composition, according to the exper- imental strategy that was used (in mmol/L): 1.- Pipette and bath: 135 Kgluconate, 5 KCl, 10 NaHEPES, pH 7.4. 2.- Pipette and bath: 140 NMDGCl (N-methyl-D-glucamine chloride), 10 Na- HEPES, 2.5 CaCl2, 1.25 MgCl2, pH 7.4. 3.- Pipette: 40 BaCl2, 10 TrisCl, pH 7.4. Bath: 0.1 BaCl2, 10 TrisCl, pH 7.4. 4.- Pipette: 40 BaCl2, 10 TrisCl, 0.5 DIDS, pH 7.4. Bath: 0.1 BaCl2, 10 TrisCl, 0.5 DIDS, pH 7.4. Ba2+ ions were used as charge carriers for the following reasons. (a) Ba2+ ions, unlike Ca2+ ions, do not block calcium channels. (b) Ba2+ ions usually move through single calcium channels faster than other divalent cations. (c) Ba2+ ions block many types of monovalent cation channels that may complicate the single calcium channel current measurements. (d) Ba2+ ions, unlike Ca2+ ions, do not activate other conductances. Working blocker solutions were prepared from concentrated stock solutions by dilution with the corresponding experimental bath solution, so as not to change bath conditions. DIDS (4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid) and Nifedipine were dissolved in DMSO (di-methyl-sulphoxide), and cal- culations were made to reach a maximum DMSO bath concentration of 0.1% when adding 1 mmol/L DIDS or 10 mmol/L Nifedipine to the bath solution. Concentrated solutions of 9.8 and 392 mmol/L of Ruthenium Red in the corre- sponding working bath solution were used to raise the Ruthenium Red bath concentration to values ranging from 111 nmol/L to 18 mmol/L. Concentrated solutions of 10 and 50 mmol/L NiCl2 were used to raise Ni2+ bath concentra- tion to values ranging from 50 nmol/L to 20 mmol/L. 2.6. Immunohistochemistry Tissue samples from human placenta from normal pregnancies were rinsed in NaCl 0.9% and frozen in cryogel. Ten-micrometer thick sections were fixed with p-formaldehyde 4% for 30 min and rinsed three times with TBS buffer (Tris buffered saline, pH 7.6) for 5 min. Sections were blocked overnight with TCT buffer (carrageenan 0.7%, Triton X-100 0.5% in TBS, pH 7.6) at 4 ◦C. Carrageenans are sulphated linear polysaccharides extracted from red seaweeds used commonly as blockers for immunohistochemical techniques [32,33]. Primary antibody incubation was performed for 2 h at room temper- ature with anti-pan a1 subunit of VGCC antibody (Alomone Labs, product ACC-004) diluted 1:300 in TCT buffer, anti-TRPV5 polyclonal antibody (Alpha Diagnostic International, product CAT21-A) 10 mg/ml diluted in bidis- tilled water or anti-TRPV6 polyclonal antibody (Alpha Diagnostic Interna- tional, product CAT11-A) 10 mg/ml diluted in bidistilled water. The anti-pan a1 subunit of VGCC antibody is directed against an 18-residue peptide corre- sponding to a conserved region of the intracellular C-terminal part of the a1 subunit, which is present in the Cav1 and Cav2 calcium channel subfamilies from all known species. Additionally, double immunostaining was performed with each of the antibodies previously described, with mouse anti-cytokeratin 7 antibody (clone OVTL 12/30, Zymed Laboratories nc.) diluted 1:100 in PBS, to confirm the trophoblastic localization of these calcium channels. After rins- ing the samples with TBS, tissue sections were incubated for 1 h at room tem- perature with Rhodamine Red goat anti-rabbit antibody (Jackson Immunoresearch, Code Number 111-295-003) diluted 1:200 in PBS (for the anti-pan a1 subunit of VGCC antibody) or 1:200 in bidistilled water (for the anti-TRPV5 and anti-TRPV6 antibodies). Cy2 conjugated goat anti-mouse antibody (Jackson Immunoresearch, Code Number 115-225-003) was used as secondary antibody for anti-cytokeratin 7 diluted 1:200 in PBS. Control sec- tions were incubated with secondary antibody after incubation in TCT buffer or bidistilled water without primary antibody. Sections were viewed using a Carl Zeiss Laser Scanning Systems 510 Confocal Microscope and the Zeiss LSM 5 Image Browser. 2.7. Western blotting and densitometric analysis of blots Isolated basal and apical purified fractions were tested for the presence of the a1 subunit of VGCC (n = 7 placentas) and for the presence of TRPV5 and TRPV6 (n = 2 placentas) by SDS-PAGE and immunoblotting. Forty micro- grams of vesicle protein together with molecular weight markers (Invitrogen BenchMark 10748-010) were loaded on a 10% SDS-polyacrylamide gel. Elec- trophoresis was performed at 100 V and the gel was transferred to a nitrocellu- lose membrane (BioRad 162-0115) for 2 h at 100 V. The nitrocellulose membrane was blocked overnight at 4 ◦C with 2% non-fat milk in saline buffereTween (in mmol/L: 1.38 NaCl, 2.7 KCl and 0.05% Tween 20), and washed in saline buffereTween. Membranes were incubated with primary antibody for 2 h at room temperature; anti-pan a1 VGCC subunit polyclonal antibody (Alomone Labs, product ACC-004) was diluted 1:200 in saline buffereTween, anti-TRPV5 polyclonal antibody (Alpha Diagnostic Interna- tional, product CAT21-A) and anti-TRPV6 polyclonal antibody (Alpha Diag- nostic International, product CAT11-A) were diluted 1:400 in bidistilled water. Control antigenic peptide for each antibody was included to assess antibody specificity; for TRPV5 and TRPV6 antigenic peptides at 16 mg/mL (Alpha Diagnostic International, product CAT21-P and CAT11-P), and for a1 VGCC subunit antigenic peptide at 4.5 mg/mL (Alomone Labs, product ACC-004). Additionally, controls in absence of primary antibody were added in each case (only with antibody diluent). After washing with saline buffereTween, membranes were incubated with donkey anti-rabbit horseradish peroxidase- linked antibody (Amersham RPN 2108 NIF824) diluted 1:5000 in saline buf- fereTween for 1 h at room temperature. The final detection was done using a chemiluminiscence ECL Western Blotting Analysis System (Amersham, RPN 2106). The anti-pan a1 VGCC subunit polyclonal antibody developed film and its respective transferred gel were scanned using a SnapScan Touch Agfa Scanner. Protein content in lanes of the scanned images was quantified using UN-SCAN-IT gel Automated Digitizing System, version 4.1 for Windows (Silk Scientific Corporation). Densitometric values of selected lanes in the developed film were corrected by the protein densitometric quantification of the respective lane in the transferred gel. Molecular weight values for the detected bands could be obtained by scanning the molecular weight markers and using a built-in function of UN-SCAN-IT software developed for this purpose. 2.8. Data analysis The values of relative PCa/PK were calculated using the shift in experi- mental reversal potential obtained when applying voltage ramps from —100 to +100 mV (40 mV/s) in the presence of CaCl2 added to the Kgluconate bath solution. We used the approach described by Wollmuth et al. [34], which uses a formula derived from the Lewis equation [35] and assumes that Ca2+ and K+ are the only permeant ions in the external solution and that ErevCa (that is, Erev after Ca2+ addition to bath) is the same as DErev (which requires ErevK = 0, which is the case when working in symmetrical K+ solutions with compensation of junction potential); in this case, contribution of Na+ (10 mmol/L in pipette and bath) to DErev was neglected, assuming that PK/ PNa = 1, which is the case for calcium channels. The form of this equation is as follows: 3. Results The following results were obtained from a total of 99 ex- cised patch experiments from 11 healthy term placenta puri- fied basal membranes. Purification of basal membranes was achieved through the protocol described in Section 2, which resulted in adequate basal membrane enrichment markers and apical membrane/intracellular membrane contamination markers of the purified basal membrane fraction, as described fully in a previous paper [26]. 3.1. Detection of Ca2+ and Ba2+ currents Working with monovalent cation currents, we characterized the effect of Ca2+ addition to the bath solution as a strategy to detect the presence of Ca2+-permeable channels. Addition of CaCl2 in low millimolar concentrations (0.5e1.5 mmol/L) to the bath solution caused displacement of the reversal potential (Erev) in eight out of 11 experiments in symmetrical conditions using Kgluconate solutions (solution composition 1 in Section 2; total K concentration = 140 mmol/L), indicating Ca2+ conduction. The mean relative permeability ratio PCa/PK was where A = [Ba]in — [Ba]outexp—x. The calculation of relative PCa/PBa using the GHK current equation requires previous calculation of the value of relative PCl/PBa, which was obtained considering the contribution of Ba2+ and Cl— currents (assumed to be the only current carriers in the membrane patches) to the experimental reversal potential in conditions of BaCl2 pipette-to-patch gradient (condition 3 in Section 2.5). Contributions of Ba2+ and Cl— currents to the average measured Erev of +32 mV (n = 31 independent experiments), considering the maximum range for the current reversal as 131 mV, that is the difference between the theoretical equilibrium potential for chloride (VeqCl— = —55 mV) and that for Ba2+ (VeqBa2+ = +76 mV), were calculated as 66% and 34%, respectively, and corrected by the z values of each ion, finally obtaining a relative PCl/PBa value of 1.2. Fig. 2. Effect of millimolar calcium concentrations on Erev and total patch cur- rents. A characteristic shift in Erev when adding millimolar concentrations of CaCl2 to the bath solution in gradient conditions of BaCl2 (in mmol/L: 40 BaCl2 pipette, 0.1 BaCl2 bath); this result reflects a high pore affinity of the recorded channels for calcium with respect to barium (detailed analysis in Section 4). 3.2. Isolation of Ba2+ currents using the Cl— channel blocker DIDS followed by Ni2+ block of Ba2+ currents Given the constant presence of Cl— currents in our patches, we sought to obtain isolated Ba2+ currents by using the gener- alized Cl— channel blocker DIDS. Membrane excised patches were studied under gradient conditions and voltage ramps between —100 and +100 mV, as previously described, and total patch Ba2+ or Cl— currents were expressed as a ratio to control currents. Block of Cl— currents by DIDS was most effective when working in a BaCl2 concentration gradient with symmetrical (0.5 mmol/L) DIDS concentrations in pipette and bath (solution 4 in Section 2), as evidenced by an experimental Erev that shifted towards VeqBa2+ (mean Erev = +53 4.8 mV, n = 10, Fig. 3A). DIDS applied only to the bath solution caused a smaller shift in Erev together with equivalent reductions in Ba2+ and Cl— currents from 2.2 to 38.3 (n = 5). Addition of CaCl2 in concentrations of 0.1e1.5 mmol/L both diminished IBa to 50% of control values and increased ICl up to 180% of control currents (results not shown). Fig. 1. Detection of barium-conducting channels and of chloride-conducting channels in excised patches. (A) Typical recording obtained using a voltage ramp in gradient conditions of BaCl2 (in mmol/L: 40 BaCl2 pipette, 0.1 BaCl2 bath); notice channel activity at both barium and chloride Veqs, which implies the presence of barium-conducting channels and of chloride-conduct- ing channels. This type of recording was present in 99% of our excised patches. (B and C) Recordings obtained using a voltage ramp (B) and a pulse protocol (C), as described in Section 2, in symmetrical concentrations of NMDGCl (140 mmol/L). In these conditions, it was possible to isolate a chlo- ride channel with the classical biophysical properties of the Maxi-chloride channels: multiple substates and voltage-dependent open-probability. Arrows indicate the direction of channels openings (n = 6). Fig. 3. Effect of chloride channel blocker DIDS on Erev and total patch currents. (A) The most efficient chloride channel blocking effect, as judged by the shift of Erev towards the barium Veq, was achieved when applying DIDS to pipette and bath (n = 10) as compared to control conditions (*p-value < 0.005 relative to value in control conditions, one-way ANOVA). (B) DIDS exerted blocking effects on both chloride and barium currents, reaching 50e60% block of control currents at DIDS bath concentrations of 0.5 mmol/L (*p-value < 0.05 relative to control value of 1, dotted line), similar results are obtained with DIDS at 1 mmol/L (n = 9, 5 and 2 for 0.1, 0.5 and 1 mmol/L DIDS, respectively). Barium total patch currents were measured at VeqCl—, while chloride total patch currents were measured at VeqBa2+. Ba2+ and Cl— control theoretical Veqs as holding potentials (detailed explanation in Fig. 4A and B) and analyzed the ex- perimental reversal potentials in the presence and absence of NiCl2. This allowed us to obtain the relative change in charge generated by a voltage pulse in the presence of NiCl2 as com- pared to control conditions, and to determine the relative ‘‘weight’’ of Ba2+ and Cl— currents in the generation of the experimental Erev; block of Ba2+ currents would cause a pre- dominance of Cl— currents and therefore a shift of Erev to- wards the Cl— theoretical equilibrium potential (—55 mV). NiCl2 in concentrations of 50 nmol/L to 5 mmol/L caused a 50% decrease of Ba2+ total patch currents and a 300% in- crease in Cl— total patch currents (n = 8, Fig. 4C). The maxi- mum Ba2+ current block was reached at 0.5 mmol/L NiCl2, and addition of NiCl2 up to 20 mmol/L did not increase the current block. The maximum Cl— current increase was achieved with 10 mmol/L NiCl2, with a drop to control Cl— current values with 20 mmol/L NiCl2, which could be aided by the drop in Cl— pipette-to-patch electrochemical gradient at high NiCl2 concentrations in bath solution. Experimental Erev shifted towards VeqCl— with NiCl2 concentrations over 0.1 mmol/L (n = 8). The fact that Cl— bath concentrations are raised when adding NiCl2 to the bath causes a shift in VeqCl—, as shown in Fig. 4D. This fact can diminish the abso- lute shift towards VeqCl— caused by a Ba2+-conducting chan- nel block, therefore underestimating the NiCl2 blocking effect. To avoid this underestimation, we calculated the difference be- tween the post-NiCl2 application Erev and the corrected theo- retical VeqCl— to show the blocking effect of NiCl2 on Ba2+ currents (Fig. 4E). Using this difference, we show that NiCl2 application shifts Erev to values progressively nearer to the theoretical VeqCl—, only reaching it at a NiCl2 concentration of 20 mmol/L with a curve that could be well fit by a second order exponential decay. The existence of two decay constants for the block of Ba2+ currents by NiCl2 suggests two indepen- dent binding sites for this blocker, which could correspond either to two sites in a same Ba2+-conducting channel or to one site in two different Ba2+-conducting channels. In accor- dance with the described shift of Erev towards the corrected VeqCl—, calculation of total patch relative permeability ratio PCl/PBa (taking into account the variations in VeqCl— at in- creasing NiCl2 bath concentrations) showed increasing values (Fig. 5), reaching a value of 65.2 at 20 mmol/L bath NiCl2 (data not shown). This result reflects the progressive loss of permeability of the membrane patch to Ba2+ ions, reaching very low values at high NiCl2 concentrations. 3.3. Block of Ba2+ currents by Nifedipine and Ruthenium Red Altogether, the previous results could be due to the pres- ence of either one unique population or two different popula- tions of Ba2+-conducting channels. For this reason, we planned pharmacological strategies to determine the presence of one or more populations of these channels. Blockers were chosen taking into consideration the high values of relative PCa/PK obtained in the initial experiments, which suggested the presence of calcium channels in contrast to cation channels with Ca2+ permeability, as the latter do not present relative PCa/PCation ratios beyond 10. Among the numerous Ca2+ channel blockers, we chose two that could independently block populations of Ca2+ channels whose gene products have been previously detected in human trophoblast tissue by other authors, as mentioned previously: Nifedipine (L-type VGCC blocker) and Ruthenium Red (TRPV5 and TRPV6 cal- cium channel blocker). Fig. 4. Effects of nickel on Erev and pulse-generated charges. NiCl2 was added to the bath solution in millimolar concentrations. Voltage pulses used to obtain pulse-generated charges are schematically shown in (A), and consist of two 100 ms pulses whose potentials were the barium and chloride Veqs separated by a 100 ms at 0 mV. A representative current recording in response to this stimulus is shown in (B). The charges were calculated using an average of 100e300 current recordings for each experiment. Charges (q = i × t; where q is the charge and i is the total current during the time t) are obtained by calculating the area under the current average curve when applying either the barium Veq holding potential (to measure chloride charges) or the chloride Veq holding potential (to measure barium charges). (C) Nickel in concentrations of 0.5e20 mmol/L in bath solution reduced barium currents to 50% of control values and enhanced chloride currents to 300% of control currents (n = 8). (D) Addition of high NiCl2 concentrations to the bath solution shifts the theoretical chloride Veq, a fact that may lead to misinterpretation of data. This phenomenon was corrected for by considering the difference between the experimental Erev and the calculated VeqCl— for every NiCl2 bath concentration that was used, as shown in (E). A close to zero difference was achieved with nickel values of 20 mmol/L, meaning that the only permeant ions in these experimental conditions corresponded to chloride ions. This curve could be fit by a second order exponential decay, suggesting two binding sites for nickel (n = 8). Membrane excised patches in BaCl2 pipette-to-bath gradi- ent were studied using voltage ramps to obtain total patch cur- rents at Cl— and Ba2+ theoretical equilibrium potentials, which were expressed as a ratio to control currents. Nifedipine added to the bath solution at a concentration of 1 and 5 mM reduced Ba2+ currents to 60 4% and 48 8% of control currents, re- spectively (n = 5, Fig. 6A). Nifedipine in concentrations of up to 10 mM in bath solution did not increase this block (data not shown). Surprisingly, Nifedipine also blocked Cl— currents at all of the studied concentrations, as shown for 1 and 5 mM in Fig. 6A. Nifedipine shifted the experimental Erev towards VeqCl— in only one out of five experiments (data not shown); in the remaining experiments, Nifedipine addition did not alter the experimental Erev (n = 3) or shifted Erev towards VeqBa2+ (n = 1). These results suggest that L-type VGCC are present in the syncytiotrophoblast basal membrane, but are probably not the only calcium channels in this membrane, as Nifedipine is unable to completely block the Ba2+ currents in our membrane patches. Membrane excised patches in BaCl2 pipette-to-bath gradi- ent were studied using voltage ramps as mentioned for Nifed- ipine. Ruthenium Red was added to the bath solution in however, addition of 1 mmol/L Nifedipine to the bath solution after block by 9.5 mmol/L Ruthenium Red diminished the Ba2+ current an additional 30% in one experiment (47% of control Ba2+ current before Nifedipine, 18% of control Ba2+ current after Nifedipine; Fig. 6B). Ruthenium Red also blocked Cl— currents at all of the applied concentrations. The blocking effect of Ba2+ currents by Ruthenium Red at concentrations lower than those needed to block L-type volt- age-dependent calcium channels (L-type calcium channel IC50 = 25.4 mmol/L) suggests the presence of other Ba2+-con- ducting channels with high sensitivity to this blocker, such as TRPV channels or non-L-type VGCC (P/Q-type or N-type). 3.4. Immunohistochemical detection of VGCC and TRPV5eTRPV6 channels in syncytiotrophoblast basal and apical membranes We pursued the detection of a1 subunit of VGCC using concentrations that ranged between the IC50 described for TRPV5 and TRPV6 (IC50 TRPV5 = 111 nmol/L; IC50 TRPV6 = 9.5 mmol/L). The 50% block of total Ba2+ current was achieved at concentration of 555 nmol/L (n = 3; Fig. 6B) and Ruthenium Red blocked Ba2+ currents up to a 37% of control currents at 18 mM (data not shown). There were no statistically significant differences in the percentual block by 555 nmol/L and 9.5 mmol/L Ruthenium Red; a polyclonal antibody against a conserved segment in the a1 subunit of the Cav1 and Cav2 subfamilies, and of TRPV5 and TRPV6 channels using corresponding polyclonal anti- bodies. As can be seen in Fig. 7A, D and G for each antibody, respectively, transverse sections of placental villi show a pre- dominant stain for all three antibodies in the periphery, prob- ably corresponding to the syncytiotrophoblast sheet. Morphologically, the fluorescent immunostaining seems not to be present in intracellular membranes, but is clearly distin- guished in the syncytiotrophoblast basal and apical mem- branes, with a more intense stain in the latter. Additionally, double immunostaining was performed to confirm the tropho- blastic localization of these channels in microvillous and basal syncytiotrophoblast membranes. As shown in Fig. 7B, E and H for anti-cytokeratin 7 plus anti-a1 subunit of VGCC, anti- TRPV5 and anti-TRPV6 antibodies, respectively, the cytoskeletal protein cytokeratin 7 is detected only in the periphery of villous tissue, clearly staining the syncytiotropho- blast. The three types of calcium channels were detected in some areas that correspond to apical and basal syncytiotropho- blast membranes. TRPV5 and TRPV6, but not the a1 subunit of VGCC, also show a cytoplasmatic pattern of staining. Con- trol sections show an almost imperceptible stain (Fig. 7C, F and I). Fig. 5. Effect of nickel on total patch relative permeability ratio PCl/PBa.Nickel effect on barium currents was also analyzed by obtaining the chloride to barium relative permeability ratio using the experimental Erev, as described in the Section 2. The control relative PCl/PBa was calculated as 1.2; 0.5 mmol/ L DIDS in pipette and bath reduced this ratio to 0.47. Nickel addition of 5 mmol/L to the bath solution in these conditions increased relative PCl/PBa to 5.1. *p-Value < 0.05 relative to control conditions, **p-value < 0.01 rela- tive to control conditions. Fig. 6. Block of barium currents by Nifedipine and Ruthenium Red. (A) Nifedipine caused a significant decrease in barium currents with respect to control values, which were measured at the chloride Veq; surprisingly, Nifedipine also reduced chloride currents, which were measured at the barium Veq in voltage ramps (n = 5). (B) Ruthenium Red caused a significant decrease in barium currents measured at the chloride Veq in voltage ramps, reaching 50% of control current at 555 nmol/L (which corresponds to TRPV6 IC50, n = 3). Notice that addition of 1 mmol/L Nifedipine to the bath solution after addition of 9.5 mmol/L Ruthenium Red caused an additional barium current block of 29%, suggesting simultaneous presence of calcium channels with different pharmacological sensitivity. Chloride currents measured at the barium Veq in voltage ramps were unexpectedly and significantly blocked by both Nifedipine (n = 5) and Ruthenium Red (n = 3). *p-Value < 0.05 relative to control value of 1, **p-value < 0.01 relative to control value of 1. Fig. 7. Detection of voltage-gated calcium channels (VGCC) and TRPV5eTRPV6 channels in apical and basal membranes of normal term human placentas. Con- focal fluorescence micrographs with their respective transmitted light micrographs (insets) of immunohistochemical sections of placental villous tissue using pri- mary antibodies against a1 subunit of VGCC (A), TRPV5 (D) and TRPV6 (G). C, F, I, respective controls using only secondary antibody (without primary antibody). B, E, and H show double immunostaining of anti-cytokeratin 7 antibody with a1 subunit of VGCC, TRPV5 and TRPV6, respectively. MVM = micro- villous (apical) membrane, BM = basal membrane, FC = fetal capillary. 3.5. Western blot detection of VGCC, TRPV5 and TRPV6 in basal and apical syncytiotrophoblast purified membranes Western blots using a polyclonal antibody against the a1 subunit of VGCC of the Cav1 and Cav2 subfamilies (a1 sub- unit of VGCC) were pursued using samples of paired basal and apical purified fractions for a total of seven placentas,whose purified fractions were used in electrophysiological ex- periments. Fig. 8A shows a representative Western blot with bands corresponding to the molecular weight of the a1 subunit of VGCC in both basal (BM) and apical (MVM) membrane purified fractions, together with their respective control with antigen peptide and control without primary antibody, none of the controls showing a specific band. Fig. 8B shows a repre- sentative Western blot (inset) with bands corresponding to the molecular weight of the a1 subunit of VGCC in both BM and MVM purified fractions, together with bands corresponding to the molecular weight of placental alkaline phosphatase (PLAP). The PLAP band was much more marked in the apical purified fraction than in the basal purified fraction, as can be seen in the corresponding densitometric analysis; the basal PLAP band corresponds to less than 10% of the sum of apical and basal PLAP densitometrically quantified bands, while the basal a1 subunit of VGCC band corresponds to more than 30% of the sum of apical and basal a1 subunit of VGCC densito- metrically quantified bands. Western blots using polyclonal antibodies against TRPV5 and TRPV6 were pursued using samples of paired basal and apical purified fractions for a total of three placentas. Fig. 9 shows a representative Western blot with bands corresponding to the molecular weights of TRPV5 and TRPV6 in both basal and apical membrane purified fractions, with their respective antigenic peptide control. TRPV5 and TRPV6 show a strong band in both of the purified syncytiotrophoblast membranes, and each of them is almost completely eliminated with the preadsorption treatment with their respective antigenic pep- tide, similar to those obtained in the control without primary antibodies (data not shown). Usual values for Nifedipine block of L-type Ca2+ channels are in the micromolar range [43,44]. Addition of Nifedipine in concentrations between 1 and 10 mmol/L to the bath solution reduced Ba2+ currents to 50e60% of control values, which suggests the presence of L-type VGCC; there is presently no evidence that supports block of other Ba2+-permeable chan- nels by Nifedipine in these concentrations. The remaining Ba2+ current could be due either to the presence of dihydro- pyridine-insensitive channels or to unfavorable conditions for Nifedipine block of L-type channels (i.e. low probability of open or inactivated channel states). The lack of shift of Erev was not an expected result if the only effect of Nifedipine addition was block of Ba2+ currents. Since in our experimental conditions Erev depends on the balance between Ba2+ and Cl— currents, the only explanation for a lack of Erev shift is an equivalent fall in both currents when adding Nifedipine. In- deed, Cl— currents are reduced when adding Nifedipine to the bath. As there are no reports of dihydropyridine block of anion channels in the literature, a remaining possibility is the existence of a functional interaction between the recorded Ca2+ and Cl— channels that is kept in the conditions of puri- fied membrane reconstitution, such that Nifedipine block of Ca2+ channels affects conductance and/or gating kinetics of the accompanying Cl— channel. Ruthenium Red is an organometallic dye that acts as an in- hibitor of a wide range of Ca2+-binding proteins, including ion channels such as ryanodine receptors, Ca2+-activated K+ channels, VGCC and TRP channels [45]. Despite its wide- spread cation channel blocking action, differences in blocking concentrations allow its use to aid in the functional identifica- tion of Ca2+ channels; a particular case is that of the TRPV subfamily, as TRPC and TRPM channels are not sensitive to Ruthenium Red [46]. We have obtained block of Ba2+ currents by Ruthenium Red using concentrations much lower than those of the IC50 blocking concentrations for TRPV6 (9.5 mmol/L), compatible with the presence of both TRPV5 and TRPV6 channels. Addition of 1 mmol/L Nifedipine further raised the block of 9.5 mmol/L Ruthenium Red, supporting the presence of functional L-type VGCC and TRPV channels in our purified basal membranes. Immunohistochemical studies that show the presence of Ca2+-permeable channels in the syncytiotrophoblast plasma membrane using human trophoblast villous tissue have presently only been done by Clarson et al., who detected TRPC3 and TRPC4 in both basal and apical membranes [17]. Our immunohistochemical studies clearly show the pres- ence of VGCC, TRPV5 and TRPV6 channels in the syncytio- trophoblast villous layer, as can be seen in the double immunostaining experiments with anti-cytokeratin 7. This is an antibody that gives a cytoplasmatic pattern of staining and thus, it was used to confirm the trophoblastic localization of the Ca2+ channels’ antibodies used in some areas. The anti- cytokeratin 7 immunostaining results allowed us to demon- strate that the stain from Ca2+ channels antibodies used was for the most part in the syncytiotrophoblast. The presence of a1 subunits of VGCC, TRPV5 and TRPV6 both in the basal and apical purified syncytiotrophoblast membranes is in accor- dance with the immunohistochemical results and represents a molecular correlation of our electrophysiological findings in reconstituted purified basal membranes. Ba2+-conducting activity was a constant feature in our re- cordings, which is by itself an important result that reflects the high density of Ca2+ channels in our purified basal mem- brane. Another interesting and unexpected result was the con- stant association of Cl— currents to our Ba2+ currents; only one out of 76 excised patch experiments showed the presence of a solitary Ba2+ current (evidenced by an Erev that closely resembled VeqBa2+, data not shown), and no experiments showed Cl— currents on their own. One possible explanation is the presence of a high density of Ca2+-dependent Cl— chan- nels, as we never worked with Ca2+-free solutions and Ca2+ concentrations in de-ionized water can be in the low range of micromolar values, which can be enough to activate these types of currents [47]. Another possibility is the existence of a functional and/or topographical relationship between the de- tected calcium and Cl— channels, which is an interesting alter- native that must be explored. Although there are a few reports of block of Cl— channels by Ni2+ and Ruthenium Red [48,49], the constant block of Cl— currents by the Ca2+ channel blockers (including Nifedipine) is a result that suggests a func- tional relationship between Ca2+ and Cl— channels. Likewise, DIDS block of Ba2+ currents could either be due to a direct action of this classical Cl— channel blocker on Ca2+ channels, as is the case for Ryanodine Receptors and TRPC [50], or due to the mentioned functional relationship. The results of our study provide functional and immuno- histological evidence for the presence of VGCC and TRPV5eTRPV6 channels in the basal membrane of human syncytiotrophoblast. Results of other studies published in the literature to date are both in favor of and against the presence of functional VGCC in this plasma membrane, as mentioned in the introduction. An important contribution to the evidence against the presence of these channels is the study performed by Cronier et al., who did not find VGCC-type currents when using patch-clamp recordings in whole cell configuration in pri- mary villous trophoblastic cell culture [13]. On the other hand, our results are in agreement with those of other authors, who have detected the presence of mRNA for VGCC, TRPV5 and TRPV6 in trophoblasts from term placentas (see Belkacemi et al. [51] for a review of recent literature). The role of Ca2+ channels in the basolateral membrane of transport epithelia has generally been associated to functions other than transcellular transport of this ion [52], due to the en- ergy requirements of Ca2+ absorption: Ca2+ enters the apical epithelial membrane following a favorable electrochemical gradient, travels through the cytoplasm bound to Ca2+-binding proteins, and exits the basolateral membrane against the elec- trochemical gradient. This transport includes a passive apical step, which can be achieved by Ca2+-permeable channels or Ca2+ passive transporters, and an active basolateral step, which can be achieved by active transporters (ATP-coupled or gradient-coupled) [53]; Ca2+ channels cannot participate in this last step, so their presence in epithelial basal/basolateral membranes must respond to other necessities (i.e. syncytiotro- phoblast hormone secretion). Nevertheless, Ca2+ channels in the basal syncytiotrophoblast membrane could indirectly par- ticipate in transcellular Ca2+ transport through regulatory mechanisms that involve PTH and PTHrP [54]. Studies have suggested the participation of PTHrP in placental Ca2+ trans- fer and trophoblast growth and differentiation [55]. In a study by Farrugia et al. [56], both PTH and PTHrP enhanced Ca2+ efflux from basal, not apical, purified human syncytiotropho- blast membrane vesicles. PTH and PTHrP have been shown to exert their actions through modulation of L-type VGCC in multiple cell types [57]. The presence of L-type Ca2+ chan- nels in syncytiotrophoblast membranes is then an expected finding given the relevance of these hormones in placental function. Such analysis is at the moment not possible for TRP channels. Their role in the basal syncytiotrophoblast membrane must continue to be studied from a molecular and functional point of view. Overall, current evidence for the role of Ca2+ as a second messenger in the human placental syncytiotrophoblast de- mands for the functional study of plasma membrane Ca2+ channels in this epithelium. Our work contributes to the func- tional detection and molecular characterization of these channels, and we believe it is PD173212 a good starting point to explore the regulation of Ca2+ as a second messenger by fetal factors.