AIM: To elucidate the effect and underlying mechanisms of omeprazole action on Mg²⁺ transport across the intestinal epithelium.

METHODS: Caco-2 monolayers were cultured in various dose omeprazole-containing media for 14 or 21 d before being inserted into a modified Ussing chamber apparatus to investigate the bi-directional Mg²⁺ transport and electrical parameters. Paracellular permeability of the monolayer was also observed by the dilution potential technique and a cation permeability study. An Arrhenius plot was performed to elucidate the activation energy of passive Mg²⁺ transport across the Caco-2 monolayers.

RESULTS: Both apical to basolateral and basolateral to apical passive Mg²⁺ fluxes of omeprazole-treated epithelium were decreased in a dose- and time-dependent manner. Omeprazole also decreased the paracellular cation selectivity and changed the paracellular selective permeability profile of Caco-2 epithelium to Li⁺, Na⁺, K⁺, Rb⁺, and Cs⁺ from series VII to series VI of the Eisenman sequence. The Arrhenius plot revealed the higher activation energy for passive Mg²⁺ transport in omeprazole-treated epithelium than that of control epithelium, indicating that omeprazole affected the paracellular channel of Caco-2 epithelium in such a way that Mg²⁺ movement was impeded.

CONCLUSION: Omeprazole decreased paracellular cation permeability and increased the activation energy for passive Mg²⁺ transport of Caco-2 monolayers that led to the suppression of passive Mg²⁺ absorption.

Magnesium plays an important role in numerous biological functions. Mg²⁺ deficiency is associated with several diseases, e.g. Alzheimer’s disease[1], osteoporosis[2], and hypertension[3]. Therefore, its plasma level is tightly regulated within a narrow range (0.7-1.1 mmol/L) by intestinal absorption and renal excretion[4]. In human intestine, fractional Mg²⁺ absorption varies from 11% to 65% depending on the amount of Mg²⁺ intake[5]. Intestinal epithelium absorbs Mg²⁺via both saturable transcellular and non-saturable paracellular pathways. Transcellular Mg²⁺ transport is an active process that requires the activity of transient receptor potential melastatin 6 (TRPM6) and the basolateral Na⁺/Mg²⁺ exchanger[6,7]. On the other hand, paracellular Mg²⁺ transport is a passive mechanism and is implicated in about 90% of intestinal Mg²⁺ absorption[7]. The paracellular Mg²⁺ transport process is modulated by the tight junction proteins, i.e. Claudin-16 and Claudin-19[8].

Omeprazole is a common therapeutic tool for acid-peptic disorders. Its active sulphenamide form selectively and covalently interacts with the H⁺/K⁺-ATPase, particularly the extracellular cysteine 813, leading to potent inhibition of H⁺/K⁺-ATPase activity[9]. Previous reports demonstrated that prolonged omeprazole administration led to hypomagnesemia and hypomagnesuria in humans[10,11]. Withdrawal of omeprazole and intravenous Mg²⁺ replacement, but not high dose oral Mg²⁺ supplement, could normalize the plasma and urinary Mg²⁺ levels[10,12]. Renal Mg²⁺ handling was normal in patients with severe hypomagnesemia associated with long-term use of omeprazole[12-14]. This body of evidence suggested an inhibitory effect of omeprazole on intestinal Mg²⁺ absorption. However, the direct action of omeprazole on intestinal Mg²⁺ transport is still elusive. The present study, therefore, aimed to elucidate the effect of omeprazole as well as obtain information regarding possible mechanisms of omeprazole action on Mg²⁺ transport across the intestinal epithelium. This study employed a monolayer of Caco-2 cells which is a suitable in vitro model for studying intestinal transport of divalent cations, e.g. Ca^2+[15] and Mg^2+[16].

Omeprazole decreased Mg_A→B transport and P_Mgin both a dose- and time-dependent manner

As demonstrated in Figure 1, the Mg_A→B transport vs Mg²⁺ concentration plots of Caco-2 monolayers were curvilinear similar to that reported in humans[5]. After 14 d in the omeprazole-treated groups, Mg_A→B transport was inhibited when compared with its corresponding untreated group (Figures 1A-F). The level of inhibition progressively increased with higher concentrations of omeprazole. Omeprazole selectively decreased non-saturable Mg_A→B transport, but not the saturated component, as clearly demonstrated by the lower rate constant for non-saturable Mg_A→B transport (Table 1). For 21 d omeprazole-treated groups, the results were similar to those of the 14 d omeprazole-treatment (Figure 1G-L, Table 1). When the same omeprazole concentration was considered, 21 d-treated groups showed a significantly lower Mg_A→B transport than the 14 d-treated groups (Figure 1, Table 1). Therefore, omeprazole decreased Mg_A→B transport in a dose- and time-dependent manner. According to the Mg_A→B transport, omeprazole also decreased the apical to basolateral P_Mg in a dose- and time-dependent mechanism (Figure 2). Moreover, omeprazole significantly increased TER, but not PD or I_SC, of Caco-2 monolayers (Table 2), indicating the lower paracellular permeability to ion transport.

Omeprazole inhibited paracellular Mg²⁺ transport

Theoretically, ions can move transversely across Caco-2 monolayer via four transport mechanisms, i.e. solvent drag-induced active, voltage dependent active, transcellular active, and paracellular passive transport. Therefore, the present experiment aimed to identify the relative involvement of each mechanism in Mg²⁺ transport across Caco-2 monolayers. Inhibitors of solvent drag-induced ion transport (phlorizin and phloretin) had no effect on Mg_A→B transport (40 mmol/L Mg²⁺ concentration gradient) in both control and omeprazole-treated monolayers (Figure 5D). In another set of experiments, Caco-2 monolayers received continuous application of Isc, simultaneously with the Mg²⁺ flux study, to nullify trans-epithelial PD and to abolish voltage dependent Mg²⁺ transport. The Mg_A→B transport in both control and omeprazole-treated monolayers were unaffected by Isc, (Figure 5A). The results indicated that solvent drag-induced and voltage dependent Mg_A→B transport were negligible.

Open in New Tab   Full Size Figure   Download Figure

Figure 5 Mechanism of Mg_A→B transport. A: Mg_A→B transport 14 d control and 600 ng/mL omeprazole-treated monolayers with or without I_SC; B and C: 20 μmol/L ruthenium red (RR); D: 100 μmol/L phlorizin or 100 μmol/L phloretin. For each data point, n = 9 in A; n = 6 in B and C. E, Arrhenius plot of 14 d control and 600 ng/mL omeprazole-treated monolayers. ^bP < 0.01 vs the control group. For each data point, n = 6.

Since transcellular Mg²⁺ transport required apical Mg²⁺ influx, inhibition of Mg²⁺ influx should abolish Mg²⁺ transport. When 20 μmol/L ruthenium red (RR), a TRPM6 inhibitor[23], was added to the apical solution, a linear relationship between Mg_A→B transport and Mg²⁺ concentration was observed (Figure 5B). The rate constant for non-saturable Mg_A→B transport of control monolayers (2.18 ± 0.13) was not different from the slope of Mg_A→B transport (2.09 ± 0.06) of RR-treated control monolayers. In parallel experiments, 14 d-600 ng/mL omeprazole-treated monolayers were bathed in bathing solution with or without 20 μmol/L RR (Figure 5C). Similar to control conditions, RR inhibited the saturable component, but not the non-saturable component, of Mg_A→B transport in omeprazole-treated monolayers. In RR-treated monolayers, the 14 d-omeprazole-treated group showed a less steep slope when compared with that of the control group (1.29 ± 0.04 vs 2.09 ± 0.06, P < 0.001, Figure 5B and C). Therefore, omeprazole suppressed the non-saturable passive Mg²⁺ transport across Caco-2 monolayers.

The present study demonstrated the effect of omeprazole on Mg²⁺ transport across Caco-2 intestinal epithelium. Omeprazole-treated monolayers showed a dose- and time-dependent decrease in Mg²⁺ transport and P_Mg (Figures 1-3). Omeprazole selectively inhibited the non-saturable passive component, but not the saturable active component, of transepithelial Mg²⁺ transport (Table 1 and Figure 5). The paracellular cation selectivity of the monolayers was also reduced after prolonged exposure to omeprazole (Figure 4). Results of the Arrhenius plot (Figure 5) showed the higher E_a in the omeprazole-treated group, indicating impediment of the paracellular channel to Mg²⁺ movement.

In humans, intestinal Mg²⁺ absorption vs Mg²⁺ intake exhibited a curvilinear relationship[5] from the combination of saturable active and non-saturable passive absorption. Moreover, lower intestinal passive Mg²⁺ absorption as compared with passive Ca²⁺ absorption was also demonstrated5,24]. Similarly, in Caco-2 monolayers, a plot of Mg_A→B transport (representing Mg²⁺ absorption) against Mg²⁺ concentration (in apical solution) was also curvilinear (Figure 1) and Mg_A→B transport was lower than the apical to basolateral Ca²⁺ transport[21]. Therefore, the Caco-2 monolayer was a suitable in vitro model of intestinal Mg²⁺ absorption[16].

Several case reports demonstrated severe hypomagnesemia associated with prolonged omeprazole usage[10-14], suggesting that intestinal Mg²⁺ absorption, but not renal Mg²⁺ handling, was defective. On the other hand, short-term omeprazole administration had no effect on intestinal Mg²⁺ absorption[25] because its bioavailability was low and its half-life was short[9,26]. Therefore, the later development of hypomagnesemia was probably associated with the depletion of Mg²⁺ store in the human body[13]. The present study demonstrated an inhibitory effect of omeprazole on Mg²⁺ fluxes across 14 and 21 d-omeprazole-treated Caco-2 monolayers, suggesting that the intestinal Mg²⁺ flux defect could not be responsible for later development of hypomagnesemia in omeprazole use.

There are two transport mechanisms for Mg²⁺ absorption, i.e. transcellular active and paracellular passive transport, across the intestinal epithelium[7]. Previous reports suggested that omeprazole inhibited active intestinal Mg²⁺ absorption and TRPM6 activity because high dose oral Mg²⁺ supplement partially[13] and totally[14] resolved hypomagnesemia in prolonged omeprazole use. On the other hand, other reports showed different results i.e. high dose oral Mg²⁺ supplement, but not intravenous Mg²⁺ replacement and withdrawal of omeprazole, failed to normalize plasma and urinary Mg²⁺ levels[10,12]. The later evidence indicated that omeprazole inhibited passive Mg²⁺ absorption, which agreed with the present findings. In the present study, omeprazole inhibited the non-saturable passive, but not saturable active, Mg²⁺ transport across Caco-2 monolayers (Table 1). In addition, the role of transcellular active Mg²⁺ transport was examined using the TRPM6 inhibitor RR. Inhibition of TRPM6 in Caco-2 cells[27] abolished the saturable active Mg²⁺ transport and revealed the inhibition of non-saturable passive Mg²⁺ transport in omeprazole-treated monolayers (Figure 5B and C). Therefore, the paracellular passive Mg²⁺ absorption defect should be recognized in omeprazole usage.

Consistent with previous findings that the paracellular passive transport of cations, such as Na⁺, Cs⁺, H⁺, Ca²⁺, and Mg²⁺, was a temperature variance mechanism[18,20], the present Arrhenius plot (Figure 5E) showed the temperature-dependent Mg²⁺ transport. Since the temperature coefficient Q₁₀ of passive ion diffusion through the open ion channel ranged from 1.2 to 1.4[28] and the paracellular pore of the tight junction behaved as the channel[20], therefore, the Q₁₀ of control (1.22) and omeprazole treated (1.31) monolayers indicated that Mg²⁺ mainly moved through the paracellular channels of Caco-2 epithelium. The paracellular passive H⁺ transport occurred via the claudin-8 channel of MDCK II epithelium[18]. The paracellular claudin-8 channel was found to impede H⁺ transport by increasing the E_a[18]. Therefore, the higher E_a of omeprazole-treated Caco-2 epithelium suggested that the paracellular channel of Caco-2 epithelium impeded Mg²⁺ transport. In addition, the higher TER (Table 2) also indicated lower paracellular permeability. The present study supported a previous report by Hou et al[29], who demonstrated that the epithelium with higher TER showed lower passive Mg²⁺ transport.

The paracellular transport of Mg²⁺ was regulated by the paracellular charge selectivity, i.e. cation selectivity, of the tight junction[29,30]. Caco-2 epithelium was a cation selective epithelium (Figure 4A-C)[21,22] that favored the transport of cations through the paracellular pathway. Similar to a previous report[21], the paracellular selective permeability profile of Caco-2 monolayers to monovalent cations was Na⁺ > K⁺ > Rb⁺ > Cs⁺ > Li⁺ (Figure 4D) which was classified as series VII of the Eisenman sequence[31]. Series VII indicated the presence of moderate negative electrical field strength in the paracellular channel of Caco-2 epithelium. However, omeprazole changed the selective permeability profile to series VI of the Eisenman sequence (K⁺ > Na⁺ > Rb⁺ > Cs⁺ > Li⁺)[31]. Since series VI was characterized by lower negative electrical field strength than that of series VII[31], the paracellular cation selectivity was decreased when the monolayers were exposed to omeprazole (Figure 4A-C). Hou et al[29] also demonstrated lower paracellular Mg²⁺ transport due to lower paracellular cation selectivity of the epithelium. Thereby, omeprazole-induced suppression of paracellular cation selectivity led to the inhibition of paracellular Mg²⁺ transport across Caco-2 epithelium.

The present study demonstrated the inhibitory effect of omeprazole on passive Mg²⁺ transport which was consistent with previous reports[29,30,32-34]. The paracellular passive Mg²⁺ transport was mainly mediated by claudins at the tight junction[29,30], the distribution of which could be affected by the change in extracellular fluid pH. Inhibition of H⁺/K⁺-ATPase activity in Caco-2 cells[35] by omeprazole might decrease the extracellular H⁺ concentration, which in turn increased the sensitivity of the extracellular calcium sensing receptor (CaSR)[32,33], which was expressed in Caco-2 cells[36,37]. Ikari et al[34] clearly demonstrated that the activation of CaSR led to the translocation of claudin-16 from the tight junction into the cell, thus inhibiting paracellular Mg²⁺ transport. Therefore, omeprazole-inhibited passive Mg²⁺ transport appeared to involve the CaSR-tight junction-dependent mechanism.

In conclusion, omeprazole inhibited paracellular passive Mg²⁺ transport across Caco-2 epithelium in a dose- and time-dependent fashion. The inhibition of passive Mg²⁺ transport was due to the decrease in paracellular cation selectivity. The results from the present study provided evidence for the regulation of intestinal Mg²⁺ absorption. However, the underlying mechanism of omeprazole inhibiting passive Mg²⁺ transport requires further study.