The roles of acid-sensing ion channel 1a and ovarian cancer G protein-coupled receptor 1 on passive Mg²⁺ transport across intestinal epithelium-like Caco-2 monolayers

Abstract

Intestinal passive Mg²⁺ absorption, which is vital for normal Mg²⁺ homeostasis, has been shown to be regulated by luminal proton. We aimed to study the regulatory role of intestinal acid sensors in paracellular passive Mg²⁺ transport. Omeprazole enhanced the expressions of acid-sensing ion channel 1a (ASIC1a), ovarian cancer G protein-coupled receptor 1 (OGR1), and transient receptor potential vanilloid 4 in Caco-2 cells. It also inhibited passive Mg²⁺ transport across Caco-2 monolayers. The expression and activation of OGR1 resulted in the stimulation of passive Mg²⁺ transport via phospholipase C- and protein kinase C-dependent pathways. ASIC1a activation, on the other hand, enhanced apical HCO₃ ⁻ secretion that led, at least in part, by a Ca²⁺-dependent pathway to an inhibition of paracellular Mg²⁺ absorption. Our results provided supporting evidence for the roles of OGR1 and ASIC1a in the regulation of intestinal passive Mg²⁺ absorption.

Introduction

Magnesium (Mg²⁺) is essential for many physiological processes such as muscle contraction and relaxation, energy metabolism, neuronal function, and bone formation. Dietary intake is the sole source of Mg²⁺ in human, therefore adequate intestinal absorption of Mg²⁺ is vital for normal Mg²⁺ balance. Intestinal Mg²⁺ uptake comprises saturable transcellular active and non-saturable paracellular passive mechanisms [1–3], with the latter contributing about 90 % of the total intestinal Mg²⁺ absorption [3]. This transport mechanism is driven by the electrochemical gradient set up by a higher luminal Mg²⁺ concentration and lumen positive voltage with respect to the basolateral side [1, 3, 4]. The tight junction-associated claudins (Cldn) have been reported to act as a paracellular Mg²⁺ channel within the tight junction [5, 6]. However, the regulation of the paracellular passive intestinal absorption of Mg²⁺ is unknown.

Apical proton has been shown to modulate paracellular Mg²⁺ transport [6–8]. Suppression of apical proton accumulation by a proton pump inhibitor (PPI) omeprazole altered paracellular permselectivity, suppressed Cldn-7 and -12 expressions, and inhibited paracellular passive Mg²⁺ transport across the intestinal-like Caco-2 monolayers [6, 9]. On the other hand, apical acidity (pH 7.0–5.5) was reported to increase paracellular passive Mg²⁺ uptake, affinity of paracellular channel for Mg²⁺, and expression of Cldn-7 and -12 in both control and omeprazole-exposed epithelia [6]. However, the underlying mechanism of apical proton regulation of the passive Mg²⁺ absorption remains elusive.

Epithelial cells in the small intestine are regularly exposed to strong gastric acid. When luminal pH decreases, the intestinal epithelium cells can directly detect and modulate their cellular response through the proton sensors, e.g., ASIC1a, OGR1, and transient receptor potential vanilloid 4 (TRPV4) [10–14]. The pH of half (pH_0.5) and full activation of OGR1 are 7.4–7.2 and 6.8, respectively [15, 16]. OGR1 is associated with G_q proteins and acts through phospholipase C (PLC)–protein kinase C (PKC) signaling pathway to activate the epithelial Na⁺/H⁺ exchanger (NHE) and H⁺-ATPase activity [15, 17]. ASIC1a is activated by extracellular pH below 6.9 with pH_0.5 of 6.2–6.8 [13], whereas TRPV4 requires a more acidic pH for activation (pH > 6.0) and is fully activated at pH 4.0 [18]. Both ASIC1a and TRPV4 are Ca²⁺ channels that trigger Ca²⁺ signaling to regulate epithelial HCO₃ ⁻ secretion [11, 19]. Activation of TRPV4 also modulates paracellular permeability and Cldn expressions by a Ca²⁺-dependent mechanism [20]. It is not known whether intestinal acid sensing ASIC1a, OGR1, and TRPV4 have any role in apical acidity-induced stimulation of paracellular passive Mg²⁺ absorption. The present study investigated the role of intestinal acid sensors in the regulation of paracellular passive Mg²⁺ absorption. The results showed that OGR1 enhanced whereas ASIC1a decreased the passive intestinal Mg²⁺ absorption.

Methods

Cell culture

The human intestinal Caco-2 cells (ATCC No. HTB-37) were grown and maintained for 14 days as previously described [21]. For the experiments, the cells were plated on the permeable polyester Transwell-clear inserts (1.0 × 10⁶ cells cm⁻²; Corning, Corning, NY, USA), 6-well plates (5.0 × 10⁵ cells per well; Corning), or 96-well plates (5.0 × 10⁴ cells per well; Corning) and maintained for 7 days [6]. From days 8 to 14 of culture, cells were grown in media with or without 5 mM HCL-activated omeprazole (200 or 400 ng/mL; Calbiochem, San Diego, CA, USA). In some experiments, the apical side of Caco-2 monolayers was intermittently exposed over a 2-h period, 3 times a day, to acidic culture media (pH 6.5 or 5.5) with or without 5 mM HCL-activated omeprazole (Calbiochem) from days 8 to 14 (Fig. 1d). The cell monolayers grown on Transwell-clear inserts, 6-well plates, or 96-well plates were used for ion flux studies, western blot analysis, or MTT reduction assay, respectively. Apical pH of Caco-2 monolayers grown for 14 days in media with or without 5 mM HCL-activated omeprazole was also determined as previously described [6].

Fig. 1

Apical proton regulates intestinal passive Mg²⁺ transport. The relative cell viability of Caco-2 exposed to acidic apical culture medium, omeprazole, and omeprazole plus acidic was elucidated by MTT assay (a). Passive Mg²⁺ transport across Caco-2 monolayer was measured in the presence of acidic apical medium, omeprazole plus acidic apical medium (b), acidic apical bathing solution, and omeprazole plus acidic apical bathing solution (c). Representative time line of Caco-2 monolayer experiments (d). *P < 0.05, **P < 0.01, ***P < 0.001 compared with the corresponding pH 7.4 group. ^†† P < 0.01, ^††† P < 0.001 compared with the omeprazole-free pH 7.4 group. (n = 6)

Bathing solutions

For the apical to basolateral 40 mM concentration gradient-driven passive Mg²⁺ transport studies, the apical solution contained (in mM) 40 MgCl₂, 1.25 CaCl₂, 4.5 KCl, 12 d-glucose, 2.5 l-glutamine, 115 d-mannitol, and 10 HEPES pH 7.4, whereas the basolateral solution contained (in mM) 1.25 CaCl₂, 4.5 KCl, 12 d-glucose, 2.5 l-glutamine, 250 d-mannitol, and 10 HEPES pH 7.4. In the apical acid-activated Mg²⁺ transport studies, the apical solution of pH 7.4 was substituted with apical solution of pH 7.0, 6.5, 6.0, or 5.5 [6]. During Mg²⁺ transport studies, pH of all apical bathing solutions was simultaneously measured and not changed throughout the experiments.

For apical HCO₃ ⁻ secretion experiments, the composition of NaHCO₃-free apical solution was as follows (in mM) 1.25 CaCl₂, 4.5 KCl, 1 MgCl₂, 12 d-glucose, 2.5 l-glutamine, 230 d-mannitol, and 10 HEPES pH 7.4; and the NaHCO₃-containing basolateral solution contained (in mM) 25 NaHCO₃, 1.25 CaCl₂, 4.5 KCl, 1 MgCl₂, 12 d-glucose, 2.5 l-glutamine, 200 d-mannitol, and 10 HEPES pH 7.4.

All solutions were maintained at 37 °C, pre-gassed with 100 % O₂ for 30 min, and had an osmolality of 290–295 mosM as measured by a freezing-point depression-based Fiske^® micro-osmometer (model 210; Fiske^® Associates, Norwood, MA, USA). All chemicals were purchased from Sigma (St. Louis, MO, USA). Bioresearch grade deionized water used in the present work had a resistance of >18.3 MΩ cm, total organic compound of <10 part per billion, and pyrogen contamination of <0.005 endotoxin units/mL.

Measurements of paracellular Mg²⁺ flux

Apical to basolateral 40 mM MgCl₂ gradient-driven paracellular Mg²⁺ flux study and determination of Mg²⁺ concentration were performed as previously described [9]. Apical acidity-dependent passive Mg²⁺ transport was observed under apical bathing solution pH 7.0, 6.5, 6.0, or 5.5 with basolateral bathing solution remaining at pH 7.4. In some experiments, Caco-2 monolayers were pre-incubated with various compounds, namely; ASIC1a activator (1 or 10 μM amitriptyline; AMT; Sigma), ASIC1a inhibitor [30 nM psalmotoxin 1 (PcTx1); Phoenix Pharmaceuticals, Burlingame, CA, USA], TRPV4 inhibitor [10 μM ruthenium red (RR); Sigma], OGR1 inhibitor [10 μM CuCl₂ (Cu²⁺) or 10 μM ZnCl₂ (Zn²⁺); Sigma], polyclonal antibody raised against an extracellular domain of human OGR1 (1:10 or 1:100 OGR1 Ab; Santa Cruz Biotechnology, Santa Cruz, CA, USA), PLC inhibitor [10 μM 1-[6-((17β-3-Methoxyestra-1,3,5(10)-trien-17-yl)amino)hexyl]-1H-pyrrole-2,5-dione; U-73122; Calbiochem], PKC inhibitor (1 μM Bisindolylmaleimide I; Gö 6850; Calbiochem), phosphoinositide 3-kinase (PI3K) inhibitor (200 nM Wortmannin, Wort.; Calbiochem], and intracellular Ca²⁺ chelator [50 μM 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetra(acetoxymethyl ester); BAPTA-AM; Calbiochem] prior to the performed experiments.

Measurements of HCO₃ ⁻ secretion

Bicarbonate secretion was measured in humidified atmosphere with 5 % CO₂ at 37 °C. After removal of the culture media, the Caco-2 monolayer was gently rinsed 3 times and incubated for 15 min in the physiological bathing solution [21]. Then, apical and basolateral solutions were substituted with apical and basolateral solutions for the HCO₃ ⁻ secretion experiment. To inhibit apical membrane-bound carbonic anhydrase (CA) activity and prevent apical HCO₃ ⁻ degradation, the selective CA IX/XII inhibitor (45 nM 4-[[[(4-fluorophenyl)amino]carbonyl]amino]-benzenesulfonamide; U-104; Sigma) was added to the apical solution. After 20 min, HCL was added to the apical solution (the final concentration of 10 mM) and incubation proceeded for 5 min. After removal of the HCL-containing apical solution, the apical side of the monolayer was gently rinsed and further incubated for 50 min in the apical solution. Aliquots of apical solution at various time points (Fig. 7a) were individually sampled. The concentration of HCO₃ ⁻ was immediately determined by using a clinical chemistry analyzer (ILab Taurus; Instrumentation Laboratory, Bedford, MA, USA). In some experiments, Caco-2 monolayers were pre-incubated with 30 nM PcTx1 (Phoenix Pharmaceuticals), 50 μM BAPTA-AM (Calbiochem) or cystic fibrosis transmembrane conductance regulator (CFTR) inhibitor [50 μM N-(2-naphthalenyl)-[(3,5-dibromo-2,4-dihydroxyphenyl)methylene]glycine hydrazide; GlyH-101; Calbiochem].

MTT reduction assay

Percent viability of control Caco-2 cells and cells exposed to 200 or 400 ng/mL omeprazole (Calbiochem), and intermittently exposed to acidic apical medium with or without omeprazole was evaluated as previously described [6].

Western blot analysis

Western blot analysis was performed as previously described [6]. In brief, protein samples were prepared by using Piece^® Ripa Buffer (Thermo Fisher Scientific, Rockford, IL, USA). 35 μg of protein sample or 5 μL Cruz Marker™ Molecular Weight Standards (Santa Cruz Biotechnology) were separated on 12.5 % SDS-PAGE gel, and then transferred onto a nitrocellulose membrane (Amersham, Buckinghamshire, UK) by electroblotting. Membranes were blocked and probed overnight at 4 °C with 1:1,000 rabbit polyclonal antibodies (Santa Cruz Biotechnology) raised against human ASIC1a, OGR1, or TRPV4. Membranes were also reprobed with rabbit polyclonal antibodies (Santa Cruz Biotechnology) raised against actin (1:5,000), ASIC1a, OGR1, or TRPV4 antibodies. After 2 h incubation at 25 °C with 1:10,000 goat anti-rabbit IgG-HRP-conjugated secondary antibodies (Santa Cruz Biotechnology), blots were visualized by Thermo Scientific SuperSignal^® West Pico Substrate (Thermo Fisher Scientific) and captured on CL-XPosure Film (Thermo Fisher Scientific). Densitometric analysis was performed using ImageJ for Mac Os X [22].

Statistical analysis

Results were expressed as mean ± SE. Two sets of data were compared using unpaired Student’s t test. One-way analysis of variance with Dunnett’s post test was used for comparison of multiple sets of data. The level of significance was P < 0.05. All data were analyzed by GraphPad Prism (GraphPad Software, San Diego, CA, USA).

Results

Apical acidity nullified omeprazole effect on passive Mg²⁺ transport

These series of experiments were performed to demonstrate the effect of direct exposure to intermittent acidic apical culture medium, apical acidity, and omeprazole on paracellular passive Mg²⁺ transport. Since an acidic extracellular pH and omeprazole have previously been reported to affect cell viability [23, 24], MTT cell proliferation assay was performed. As shown in Fig. 1a, neither exposure to intermittent acidic apical culture medium (6.5 and 5.5) nor 200 or 400 ng/mL omeprazole, concentrations resembling those found in human plasma [25], had cytotoxic effect on cell viability when compared to control. Similar to our previous report [6], 200 and 400 ng/mL omeprazole significantly increased apical pH compared to control condition (data not shown), therefore omeprazole suppressed apical acidification.

Intermittent exposure to acidic apical culture medium at pH 6.5 and 5.5 had no effect on passive Mg²⁺ transport when compared to control pH 7.4 (Fig. 1b). Both 200 and 400 ng/mL omeprazole (pH 7.4) significantly decreased the passive Mg²⁺ transport from 134.82 ± 3.02 to 110.45 ± 4.32 and 96.42 ± 4.86 nmol h⁻¹ cm⁻², respectively (Fig. 1b). In the 200 ng/mL omeprazole-exposed groups, intermittent exposure to acidic apical culture medium of pH 6.5 and 5.5 abolished the inhibitory effect of omeprazole on passive Mg²⁺ transport (120.71 ± 6.36 and 130.70 ± 4.44 nmol h⁻¹ cm⁻², respectively; Fig. 1b). Acidic apical culture medium of pH 6.5 and 5.5 also normalized passive Mg²⁺ transport (115.16 ± 3.98 and 122.45 ± 2.93 nmol h⁻¹ cm⁻², respectively) in the 400 ng/mL omeprazole exposed groups (Fig. 1b).

To study the effect of apical acidity on passive Mg²⁺ transport, Caco-2 monolayers grown in culture medium pH 7.4 with or without omeprazole were used. As shown in Fig. 1c, apical acidity significantly increased passive Mg²⁺ transport from 133.60 ± 3.81 nmol h⁻¹ cm⁻² at pH 7.4 to 163.71 ± 6.94, 214.78 ± 7.43, 241.56 ± 9.07, and 270.42 ± 10.54 nmol h⁻¹ cm⁻² when apical bathing solution was changed to pH 7.0, 6.5, 6.0, and 5.5, respectively. Similar to Fig. 1b, 200 and 400 ng/mL omeprazole (apical pH 7.4) significantly decreased passive Mg²⁺ transport to 105.85 ± 5.19 and 98.95 ± 4.05 nmol h⁻¹ cm⁻², respectively (Fig. 1c). Apical acidity abolished the inhibitory effect of 200 and 400 ng/mL omeprazole on passive intestinal Mg²⁺ transport (Fig. 1c). These results demonstrated that apical proton directly stimulated passive intestinal Mg²⁺ absorption.

Intermittent acidic apical culture medium and omeprazole altered intestinal proton sensor expression

Since previous results demonstrated the stimulatory effect of apical acidity on the intestinal passive Mg²⁺ transport (Fig. 1) [6], these experiments aimed to examine the expression of proton sensitive ASIC1a, OGR1, and TRPV4 in Caco-2 cells. Omeprazole (200 and 400 ng/mL) significantly increased the expressions of ASIC1a, OGR1, and TRPV4 (Fig. 2a–d) in Caco-2 cells. On the other hand, intermittent exposure to acidic apical culture medium (pH 6.5 and 5.5) markedly suppressed ASIC1a, OGR1, and TRPV4 expressions (Fig. 2e–h). Intermittent exposure to acidic apical culture medium also abrogated the up-regulatory effect of 400 ng/mL omeprazole on ASIC1a, OGR1, and TRPV4 expressions (Fig. 3). These findings demonstrated the regulatory role of extracellular protons on the expression of intestinal acid sensors.

Fig. 2

Acidic apical culture medium and omeprazole regulate intestinal acid sensor expression. Representative immunoblotting and densitometric analysis of OGR1 (a, b, respectively), ASIC1a (a, c, respectively), and TRPV4 expressions (a, d, respectively) in Caco-2 cells exposed to 200 or 400 ng/mL omeprazole. Representative immunoblotting and densitometric analysis of OGR1 (e, f, respectively), ASIC1a (e, g, respectively), and TRPV4 expressions (e, h, respectively) in Caco-2 cells intermittently exposed to acidic apical culture medium. *P < 0.05, **P < 0.01 compared with the corresponding control group. (n = 5)

Fig. 3

Acidic apical culture medium abolishes omeprazole effect on acid sensor expression. Representative immunoblotting and densitometric analysis of OGR1 (a, b, respectively), ASIC1a (a, c, respectively), and TRPV4 expressions (a, d, respectively) in Caco-2 cells intermittently exposed to acidic apical culture medium and 400 ng/mL omeprazole. *P < 0.05, **P < 0.01, ***P < 0.001 compared with the corresponding control group. (n = 5)

Intestinal acid sensors modulated passive Mg²⁺ transport

This experiment aimed to investigate the effect of activation or inhibition of ASIC1a, OGR1, and TRPV4 on passive Mg²⁺ transport across Caco-2 monolayers. Unexpectedly, at apical pH 7.4, OGR1 inhibitors (10 μM Cu²⁺ and 10 μM Zn²⁺ [15, 17]) and 1:10 OGR1 Ab significantly decreased passive Mg²⁺ transport from 133.60 ± 3.82 to 96.93 ± 5.09, 99.20 ± 5.66, and 102.65 ± 6.18 nmol h⁻¹ cm⁻², respectively (Fig. 4a). Neither ASIC1a inhibitor 30 nM PcTx1 [26], TRPV4 inhibitor 10 μM RR [20], nor 100 °C-heated OGR1 Ab had modulatory effect on passive Mg²⁺ transport in apical pH at 7.4 (Fig. 4a). In the presence of mild acidic apical pH 7.0, the rate of passive Mg²⁺ transport was significantly decreased from 161.18 ± 10.34 to 92.80 ± 7.58 and 96.67 ± 8.29 nmol h⁻¹ cm⁻² in the presence of Cu²⁺ and Zn²⁺, respectively, and to 126.91 ± 7.44 and 107.71 ± 8.65 nmol h⁻¹ cm⁻² by 1:100 and 1:10 OGR1 Ab, respectively (Fig. 4b). In the 400 ng/mL omeprazole exposed groups, the rate of passive Mg²⁺ transport (nmol h⁻¹ cm⁻²) of 98.95 ± 4.05 was decreased by Cu²⁺ to 63.79 ± 4.36 at apical pH 7.4 (Fig. 5a) and from 122.36 ± 4.72 to 78.78 ± 3.52 at apical pH 7.0 (Fig. 5b), and by Zn²⁺ to 66.23 ± 4.63 at apical pH 7.4 and 81.39 ± 3.92 at apical pH 7.0, and by 1:10 OGR1 Ab to 67.33 ± 3.82 at apical pH 7.4 and 83.53 ± 3.58 at apical pH 7.0. These results indicated that OGR1 activation could have stimulatory effect on passive intestinal Mg²⁺ absorption.

Fig. 4

OGR1 and ASIC1a regulate intestinal passive Mg²⁺ transport. In this prior of passive Mg²⁺ transport studies, Caco-2 monolayers were pre-incubated with 30 nM PcTx1, 10 μM RR, 10 μM Cu²⁺, 10 μM Zn²⁺, 1:10 or 1:100 OGR1 Ab, 100 °C-heated OGR1 Ab, or 100 °C-heated 30 nM PcTx1 before being exposed to apical bathing solution at pH 7.4 (a), 7.0 (b), 6.5 (c), 6.0 (d) or 5.5 (e). The passive Mg²⁺ transport was also studied in Caco-2 monolayers pre-incubated with 1 or 10 μM AMT (f). *P < 0.05, **P < 0.01, ***P < 0.001 compared with the corresponding control group. (n = 6)

Fig. 5

OGR1 and ASIC1a regulate passive Mg²⁺ transport in omeprazole exposed epithelium. Passive Mg²⁺ transport by 400 ng/mL omeprazole-exposed Caco-2 monolayers that were pre-incubated with 30 nM PcTx1, 10 μM RR, 10 μM Cu²⁺, 10 μM Zn²⁺, 1:10 OGR1 Ab, 100 °C-heated OGR1 Ab, or 100 °C-heated 30 nM PcTx1 in apical bathing solution at pH 7.4 (a), 7.0 (b), 6.5 (c), 6.0 (d) or 5.5 (e). Passive Mg²⁺ transport was also studied in 400 ng/mL omeprazole-exposed Caco-2 monolayers that were pre-incubated with 1 or 10 μM AMT (f). *P < 0.05, **P < 0.01, ***P < 0.001 compared with the corresponding control group. (n = 6)

As expected, PcTx1 significantly increased passive Mg²⁺ transport by about 14 % at acidic apical pH of 6.5 (241.56 ± 5.15 vs. 211.32 ± 5.58 nmol h⁻¹ cm⁻² of control group; Fig. 4c), by 17 % at acidic apical pH of 6.0 (280.96 ± 8.14 vs. 239.32 ± 7.58 nmol h⁻¹ cm⁻² of control group; Fig. 4d), and by 14 % at acidic apical pH of 5.5 (308.51 ± 8.74 vs. 270.42 ± 10.54 nmol h⁻¹ cm⁻² of control group; Fig. 4e). On the other hand, ASIC1a activator AMT [27] at 1 and 10 μM significantly decreased passive Mg²⁺ transport at apical pH 7.4 from 133.60 ± 3.82 to 99.41 ± 6.00 and 105.39 ± 4.92 nmol h⁻¹ cm⁻², and at apical pH 7.0 from 165.18 ± 6.60 to 111.54 ± 9.04 and 117.26 ± 6.76 nmol h⁻¹ cm⁻², respectively (Fig. 4f). In the 400 ng/mL omeprazole exposed groups, PcTx1 also increased passive Mg²⁺ transport by about 42 % at acidic apical pH of 6.5 (Fig. 5c), by about 49 % at acidic apical pH of 6.0 (Fig. 5d), and by about 67 % at acidic apical pH of 5.5 (Fig. 5e). On the other hand, ASIC1a activator AMT at 1 and 10 μM significantly reduced passive Mg²⁺ transport at apical pH 7.4 from 98.95 ± 4.05 to 73.31 ± 4.14 and 75.87 ± 6.47 nmol h⁻¹ cm⁻², and at apical pH 7.0 from 122.36 ± 4.72 to 80.89 ± 6.24 and 83.62 ± 8.05 nmol h⁻¹ cm⁻², respectively (Fig. 5f). These results demonstrated the inhibitory effect of ASIC1A on the passive intestinal Mg²⁺ absorption.

Signaling pathway of passive Mg²⁺ transport modulation

These series of experiments were performed to elucidate the signaling pathways of OGR1 and ASIC1a activation in the regulation of passive Mg²⁺ transport. Previous reports showed that active OGR1 acted through PLC and PKC to stimulate epithelial ion transport [15, 17]. In the present study, OGR1 inhibitor 10 μM Cu²⁺, PLC inhibitor 10 μM U-73122, and PKC inhibitor 1 μM Gö 6850 significantly decreased the rate of passive Mg²⁺ transport (nmol h⁻¹ cm⁻²) in the absence (92.80 ± 7.58, 84.55 ± 6.51, and 94.95 ± 7.20, respectively, vs. 161.18 ± 10.34 of control monolayers) and presence of 400 ng/mL omeprazole (78.78 ± 3.52, 82.44 ± 5.52, and 78.11 ± 6.53, respectively, vs. 122.36 ± 4.72 of control monolayers) (Fig. 6a). Neither PI3K inhibitor 200 nM Wort. nor intracellular Ca²⁺ chelator 50 μM BAPTA-AM had a modulatory effect on passive Mg²⁺ transport in mild acidic apical pH at 7.0 (Fig. 6a). Therefore, active OGR1 probably stimulated intestinal passive Mg²⁺ absorption by PLC- and PKC-dependent mechanisms.

Fig. 6

Signaling pathways of OGR1 and ASIC1a regulate passive Mg²⁺ transport. Passive Mg²⁺ transport by using control or omeprazole-exposed Caco-2 monolayers that were pre-incubated with 30 nM PcTx1, 10 μM Cu²⁺, or 10 μM Zn²⁺, 10 μM U-73122, 1 μM Gö 6850, 200 nM Wort., or 50 μM BAPTA-AM in apical bathing solution pH at 7.0 (a) or 5.5 (b). DMSO 0.3 % (vol/vol) was used as vehicle for preparation of inhibitors. *P < 0.05, **P < 0.01, ***P < 0.001 compared with the corresponding control group. (n = 6)

At acidic apical pH 5.5, 30 nM PcTx1 and 50 μM BAPTA-AM were found to increase the rate of passive Mg²⁺ transport (nmol h⁻¹ cm⁻²) in the absence (328.46 ± 10.42 and 324.95 ± 6.31, respectively, vs. 266.68 ± 10.81 of control) and presence of 400 ng/mL omeprazole (299.97 ± 13.67 and 252.97 ± 16.35, respectively, vs. 178.61 ± 6.89 of control) (Fig. 6b). Inhibitors of Ca²⁺-sensitive signaling mediator (200 nM Wort. and 1 μM Gö 6850) had no effect on passive Mg²⁺ transport at acidic apical pH of 5.5 (Fig. 6b). Therefore, inhibition of ASIC1a and prevention of intracellular Ca²⁺ elevation could stimulate the intestinal passive Mg²⁺ absorption.

ASIC1a-stimulated HCO₃ ⁻ secretion

After having demonstrated the expression and function of ASIC1a in Caco-2 epithelium, we sought to discover the possible underlying mechanism of ASIC1a-induced decrease in the passive Mg²⁺ transport. It was previously reported that activation of ASICs stimulated duodenal HCO₃ ⁻ secretion in vivo [11]. As shown in Fig. 7a, apical acid was found to stimulate HCO₃ ⁻ secretion in both control and omeprazole-exposed monolayers at various time points. HCO₃ ⁻ secretion in response to acid peaked at 50 min in control, 200 ng/mL omeprazole-exposed, and 400 ng/mL omeprazole-exposed groups (Fig. 7a). Omeprazole at 200 and 400 ng/mL significantly increased both the basal (4.33 ± 0.31 and 5.09 ± 0.5, respectively, vs. 2.43 ± 0.48 μmol h⁻¹ cm⁻² of control group) and peak acid-stimulated HCO₃ ⁻ secretion (12.15 ± 0.55 and 13.85 ± 0.45, respectively, vs. 9.07 ± 0.54 μmol h⁻¹ cm⁻² of control group) (Fig. 7b, c). CFTR inhibitor 50 μM GlyH-101 on the other hand markedly decreased the basal HCO₃ ⁻ secretion in both control and 400 ng/mL omeprazole-exposed monolayers (Fig. 7d). As expected, PcTx1, GlyH-101, and BAPTA-AM significantly reduced peak acid stimulated HCO₃ ⁻ secretion in both control (3.01 ± 0.37, 2.53 ± 0.83, and 2.90 ± 0.67 μmol h⁻¹ cm⁻², respectively) and 400 ng/mL omeprazole exposed monolayers (5.92 ± 1.19, 5.03 ± 0.53, and 5.85 ± 1.01 μmol h⁻¹ cm⁻², respectively) (Fig. 7e). These results suggested that omeprazole induced HCO₃ ⁻ secretion by ASIC1a-, Ca²⁺-, and CFTR-dependent mechanisms.

Fig. 7

ASIC1a mediates apical HCL stimulating HCO₃ ⁻ secretion. Time course of HCO₃ ⁻ secretion by control or omeprazole-exposed Caco-2 monolayers that was apically induced by 10 mM HCL (a). Basal (at 20 min; b, d) and peak acid-stimulated HCO₃ ⁻ secretion (at 50 min; c, e) by control or omeprazole-exposed Caco-2 monolayers that were pre-treated with 30 nM PcTx1, 50 μM BAPTA-AM, or 50 μM GlyH-101. DMSO 0.3 % (vol/vol) was used as vehicle for preparation of inhibitors. *P < 0.05, **P < 0.01, ***P < 0.001 compared with the corresponding control group. (n = 5)

Discussion

Adequate intestinal absorption of Mg²⁺ is vital for maintaining normal Mg²⁺ balance. We have previously reported a direct effect of apical acidity on passive Mg²⁺ transport across human intestinal Caco-2 epithelium. Suppression of apical proton secretion and elevation of apical pH by omeprazole decreased the paracellular cation selectivity, negative electrical field strength of paracellular pore, Mg²⁺ affinity of the paracellular channel, Cldn-7 and -12 expression, and paracellular Mg²⁺ absorption [6, 9]. On the other hand, increase in apical acidity markedly enhanced Mg²⁺ affinity of the paracellular channel, Cldn-7 and -12 expression, and paracellular Mg²⁺ absorption [6]. In the present study, we proposed that intestinal passive Mg²⁺ absorption was regulated by the proton-sensitive OGR1 and ASIC1a.

In human tissues, OGR1 has previously been detected in the small intestine, spleen, testis, brain, lung, placenta, heart, and kidney, but not in the colon, liver, or skeletal muscle [14]. OGR1 is known as a proton-sensitive G-protein-coupled receptor [15, 17] that requires extracellular histidine residues 17, 20, 84, 169, and 269 for its proton detection [15]. It is coupled to G_q proteins and PLC that triggers an increase in the intracellular Ca²⁺ transient and PKC, which in turn activates the epithelial NHE and H⁺-ATPase in OGR1-transfected HEK293 cells [15, 17]. In the present study, activation of OGR1 by mild acidic apical pH of 7.0 was found to stimulate passive Mg²⁺ transport. This effect was inhibited by OGR1 inhibitors and OGR1 Ab. However, because OGR1 activity declined to an inactivation state when the extracellular pH declined to 6.5 [15], the stimulation of passive Mg²⁺ transport seen at more acidic apical pH of 6.5, 6.0, and 5.5 must be OGR1-independent. Our results also showed that the OGR1 induced stimulation of passive Mg²⁺ transport was mediated by PLC and PKC signaling pathways, but not by the intracellular Ca²⁺-dependent mechanism. The controversial reports regarding the role of intracellular Ca²⁺ as a downstream mediator of OGR1 activation in the present and previous studies [15, 17] was probably due to the different cell types used in the studies. Moreover, the observed up-regulation of OGR1 expression was probably a compensatory response of omeprazole-exposed monolayers to oppose the inhibitory effect of omeprazole on passive intestinal Mg²⁺ absorption.

Previously, the expression and function of ASIC1a has been extensively studied in the nervous system [12, 13, 28]. Although ASICs were cloned from the human small intestine [29], expression and function of ASICs were mainly observed in neurons in the gastrointestinal tract [12, 13]. Dong et al. [11] demonstrated the expression and function of ASIC1a in rat duodenal epithelium and human HT29 cells. In the present study, we proposed that ASIC1a activation had an inhibitory effect on intestinal passive Mg²⁺ absorption. Since ASIC1a is active at extracellular pH below 6.9 [13], it might directly regulate the passive Mg²⁺ transport under acidic apical pH of 6.5, 6.0, and 5.5. By using specific inhibitor or activator of ASIC1a in the monolayer incubation, we were able to show the inhibitory role of activated ASIC1a on passive intestinal Mg²⁺ absorption. Since ASIC1a can act as a Ca²⁺ channel [11], activation of ASIC1a by apical acidity probably involves an intracellular Ca²⁺ elevation leading to inhibition of intestinal passive Mg²⁺ absorption that in turn was relieved by PcTx1 and BAPTA-AM. However, it was possible that other acid sensors may also mediate the stimulatory effect of apical pH below 7.0 on the intestinal passive Mg²⁺ absorption.

Although TRPV4 has been reported to regulate paracellular permeability in mammary HC11 epithelium [20] and act as a proton sensor that is fully activated at pH 4.0 in Chinese hamster ovary cells [18], it was probably not involved in the apical acidity-induced paracellular passive Mg²⁺ transport in human intestinal Caco-2 epithelium. There are four possible reasons to explain this assumption: (1) being located in the basolateral membrane of human Caco-2 epithelium [10], TRPV4 cannot be directly activated by apical acidity, (2) the acidic apical pH used in the present study is not low enough to effectively activate TRPV4, (3) TRPV4 activation has not been known to affect passive Mg²⁺ transport, and (4) in the present study used different cell type from those used in the previous studies [18, 20].

It is widely accepted that duodenal mucosal HCO₃ ⁻ secretion is an important mechanism of enterocyte epithelium to defend itself against exposure to strong gastric acid [30, 31]. The duodenal epithelial cells can directly detect and regulate mucosal HCO₃ ⁻ secretion via ASIC1a and purinoceptors P2Y [11, 32]. Our results agreed with the previous study [11] that apical HCL stimulated HCO₃ ⁻ secretion in ASIC1a by an intracellular Ca²⁺-dependent pathway. Apical HCO₃ ⁻ secretion in the small intestine occurs via: (1) anion conductive CFTR-dependent, (2) anion Cl⁻/HCO₃ ⁻ exchanger-dependent, and (3) paracellular hydrostatic pressure-dependent mechanisms [31]. Because the specific CFTR inhibitor GlyH-101 could not totally suppress HCO₃ ⁻ secretion in the present study, therefore the Cl⁻/HCO₃ ⁻ exchanger-dependent and paracellular pathways may also play a role in the acid-stimulated HCO₃ ⁻ secretion. Furthermore, from the finding that ASIC1a activation suppressed paracellular Mg²⁺ absorption, as well as enhanced apical HCO₃ ⁻ secretion, it was possible that apical HCO₃ ⁻ secretion could contribute to the suppression of paracellular Mg²⁺ absorption. In the small intestine, luminal proton from gastric acid was believed to play an important role in providing appropriate environment for mineral absorption by stabilizing their ionized forms [33]. Not only reducing luminal proton [30, 31], secreted HCO₃ ⁻ may also trigger the precipitation of luminal Mg²⁺ as MgCO₃ [34], thus reducing the intestinal Mg²⁺ absorption. In addition, secreted HCO₃ ⁻ may reduce the luminal positive voltage that is required as a driving force for passive Mg²⁺ absorption [1, 3, 4]. Therefore, apical secreted HCO₃ ⁻ may, at least in part, suppress the intestinal passive Mg²⁺ absorption by decreasing Mg²⁺ bioavailability and luminal positive voltage.

Hypomagnesemia that occurred with a long-term use of PPI probably resulted from a decrease in the intestinal absorption [35–38]. Previously, we reported an inhibitory effect of the PPI omeprazole on the paracellular passive Mg²⁺ absorption in human enterocyte-like Caco-2 monolayers [6, 9]. In the present study, we showed an enhancing effect of omeprazole on the expression of ASIC1a, an activation of which increased the apical HCO₃ ⁻ secretion and suppressed paracellular passive Mg²⁺ absorption in CFTR- and intracellular Ca²⁺-dependent manners. Since omeprazole increased paracellular anion selectivity [9], it was likely that paracellular HCO₃ ⁻ secretion was also involved in the omeprazole-induced HCO₃ ⁻ secretion in the present study. Our results agreed with a previous report of omeprazole stimulating human duodenal mucosal HCO₃ ⁻ secretion in vivo [39]. Therefore, the higher rate of HCO₃ ⁻ secretion by omeprazole-exposed epithelium in the present study could, at least in part, decrease the intestinal passive Mg²⁺ absorption by increasing MgCO₃ precipitation [34] and reducing the luminal-positive voltage that acted as a driving force for passive Mg²⁺ absorption.

In conclusion, we showed that the expression and activation of OGR1 resulted in the stimulation of paracellular intestinal Mg²⁺ absorption via PLC- and PKC-dependent pathways. ASIC1a activation in the presence of acidic apical condition, on the other hand, enhanced apical HCO₃ ⁻ secretion that led, at least in part, by a Ca²⁺-dependent pathway to an inhibition of paracellular Mg²⁺ absorption.