SLC26A3 (DRA) is stimulated in a synergistic, intracellular Ca²⁺-dependent manner by cAMP and ATP in intestinal epithelial cells

Abstract

In polarized intestinal epithelial cells, downregulated in adenoma (DRA) is an apical Cl⁻/${\text{HCO}}_3^{-}$ exchanger that is part of neutral NaCl absorption under baseline conditions, but in cyclic adenosine monophosphate (cAMP)-driven diarrheas, it is stimulated and contributes to increased anion secretion. To further understand the regulation of DRA in conditions mimicking some diarrheal diseases, Caco-2/BBE cells were exposed to forskolin (FSK) and adenosine 5′-triphosphate (ATP). FSK and ATP stimulated DRA in a concentration-dependent manner, with ATP acting via P2Y1 receptors. FSK at 1 µM and ATP at 0.25 µM had minimal to no effect on DRA given individually; however, together, they stimulated DRA to levels seen with maximum concentrations of FSK and ATP alone. In Caco-2/BBE cells expressing the Ca²⁺ indicator GCaMP6s, ATP increased intracellular Ca²⁺ (Ca²⁺_i) in a concentration-dependent manner, whereas FSK (1 µM), which by itself did not significantly alter Ca²⁺_i, followed by 0.25 µM ATP produced a large increase in Ca²⁺ that was approximately equal to the elevation caused by 1 µM ATP. 1,2-Bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetrakis(acetoxymethyl ester) (BAPTA-AM) pretreatment prevented the ATP and FSK/ATP synergistically increased the DRA activity and the increase in Ca²⁺_i caused by FSK/ATP. FSK/ATP synergistic stimulation of DRA was similarly observed in human colonoids. In Caco-2/BBE cells, subthreshold concentrations of FSK (cAMP) and ATP (Ca²⁺) synergistically increased Ca²⁺_i and stimulated DRA activity with both being blocked by BAPTA-AM pretreatment. Diarrheal diseases, such as bile acid diarrhea, in which both cAMP and Ca²⁺ are elevated, are likely to be associated with stimulated DRA activity contributing to increased anion secretion, whereas separation of DRA from Na⁺/H⁺ exchanger isoform-3 (NHE3) contributes to reduced NaCl absorption.

NEW & NOTEWORTHY The BB Cl⁻/${\text{HCO}}_3^{-}$ exchanger DRA takes part in both neutral NaCl absorption and stimulated anion secretion. Using intestinal cell line, Caco-2/BBE high concentrations of cAMP and Ca²⁺ individually stimulated DRA activity, whereas low concentrations, which had no/minimal effect, synergistically stimulated DRA activity that required a synergistic increase in intracellular Ca²⁺. This study increases understanding of diarrheal diseases, such as bile salt diarrhea, in which both cAMP and elevated Ca²⁺ are involved.

INTRODUCTION

SLC26A3 [downregulated in adenoma (DRA)] is an intestinal brush border (BB) Cl⁻/${\text{HCO}}_3^{-}$ exchange protein that is linked to the BB SLC9A3 [Na⁺/H⁺ exchanger-3 (NHE3)] to produce the neutral NaCl absorptive process, the major way intestinal Na⁺ absorption occurs in the period between meals (1–9). DRA is expressed in the largest amount in the human proximal colon but is also present in significant amounts in the human duodenum and ileum (3, 6, 7, 10, 55). It is localized to the upper crypt and villus in the small intestine and upper crypt and surface cells of the colon (3, 7, 10); in these cells, it is colocalized with NHE3, and in many of these cells also with cystic fibrosis transmembrane conductance regulator (CFTR; 3, 11–14). DRA is functionally and physically linked to NHE3 and takes part in neutral NaCl absorption under baseline conditions. However, in addition, in some conditions associated with elevated cyclic adenosine monophosphate (cAMP), DRA also carries out increased rates of ${\text{HCO}}_3^{-}$ secretion and in the same cases, physically interacts with the BB Cl⁻ channel, CFTR (3, 12), and dissociates from NHE3. However, the role of DRA in diarrheal diseases is unclear; DRA is inhibited in inflammatory bowel disease (IBD) and Salmonella-related diarrheas, whereas its failure to function or failure to be produced causes congenital chloride diarrhea (2, 5, 15, 16).

There is increasing evidence that increased second messengers, as occurs with most diarrheal diseases, alter DRA activity (2, 3, 5, 8, 9, 12, 13, 16–23). Elevation in intracellular cyclic nucleotides and/or Ca²⁺ is part of the pathophysiology of many diarrheal diseases; for instance, cAMP in cholera toxin-related diarrhea, guanosine 3′,5′-cyclic monophosphate (cGMP) in E. coli heat-stable enterotoxin-related diarrhea, and Ca²⁺ in rotavirus-related diarrhea, whereas in bile salt-induced diarrhea both elevated cAMP and Ca²⁺ contribute to the pathogenesis (24–30). In addition to the role of cAMP and elevated intracellular Ca²⁺ signaling pathways in pathological processes, they also take part in the regulation of multiple physiological processes, in which their actions are independent, additive, synergistic, or inhibitory. In the current studies, we tested the hypotheses that both elevated cAMP and Ca²⁺ stimulated DRA activity and examined the interactions of elevated cAMP and elevated Ca²⁺ in DRA stimulation to determine whether they affected DRA entirely independently or if there were interactions that were additive or synergistic in polarized intestinal epithelia model cell lines.

MATERIALS AND METHODS

Chemicals, Reagents, and Antibodies

Reagents were obtained from Sigma Aldrich (St. Louis, MO) unless otherwise stated. 2,7-Bis(2-carboxyethyl)-5-carboxyfluorescein acetoxymethyl ester (BCECF-AM) was from Thermo Fisher. Alexa Fluor 488 and 568-conjugated goat anti-mouse and anti-rabbit secondary antibodies were from LI-COR. IRDye 800 (No. 926–32210) and 680 (No. 926–68071) were used with mouse monoclonal and rabbit polyclonal antibodies, respectively (1:10,000 dilutions). Rabbit polyclonal antibodies to human influenza hemagglutinin (HA; C29F4, 1:1,000) and rabbit monoclonal antibodies to HA (No. 3724, 1:1,000) were from Cell Signaling. β-Actin antibodies (No. A2228. 1:3,000) were from Sigma Aldrich. HA-cholinergic M3 receptor cDNA was from Addgene (No. 40753). BPTU N-[2-[2-(1,1-Dimethylethyl)phenoxy]-3-pyridinyl]-N′-[4-(trifluoromethoxy)phenyl]urea was from Tocris Bioscience (No. 6078). DRAinh-A250 was a gift from Dr. Alan Verkman (University of California, San Francisco).

Cell Culture

The Caco-2/BBE cell line originally derived from a human colon adenocarcinoma was obtained from M. Mooseker (Yale University) and J. Turner (Harvard University). Cells were grown on 0.4-µm membrane Transwell inserts in DMEM containing 25 mM NaHCO₃ supplemented with 0.1 mM nonessential amino acids, 10% fetal bovine serum (heat inactivated, 55°C for 30 min), 4 mM glutamine, 50 U/mL penicillin, and 50 μg/mL streptomycin, pH 7.4, in 5% CO₂/air at 37°C.

Endoscopic specimens of the human proximal colon from healthy human subjects were used to establish primary cultures of human colonoids as previously described (13). Colonoids were expanded and plated on Transwell inserts (0.4 µm, Νo. 3470, Corning Inc.) to form monolayers. For differentiation, colonoids were maintained in a medium that lacked WNT3A, R-spondin1, and SB202190 for 5 days. The procurement and study of human colonoids were approved by the Institutional Review Board of Johns Hopkins University School of Medicine (NA_00038329; 13).

Caco-2/BBE Cells and Human Colonoids Stably Transduced with Lentivirus GCaMP6s and Transiently Transduced with Adenovirus-HA-NHE3

Caco-2/BBE wild-type cells and human proximal colonoids were transduced with GCaMP6s lentiviral particles and stable Caco-2/GCaMP6s and colonic GCaMP6s cell lines were generated with puromycin selection (10 μg/mL) as described previously (31). In some cases, adenovirus-HA-NHE3 was prepared as described and studied ∼48 h after transduction in Caco-2 cells (31).

Measurement of Cl⁻/${\mathbf{\text{HCO}}}_{\mathbf{3}}^{\mathbf{-}}$ Exchange Activity

Cl⁻/${\text{HCO}}_3^{-}$ exchange activity was measured in Caco-2/BBE cells and colonoid monolayers fluorometrically using the intracellular pH (pH_i)-sensitive dye BCECF-AM and a custom chamber allowing separate apical and basolateral perfusion, as previously described (13, 32). Cells were incubated with 10 mmol/L BCECF-AM in NaCl solution (138 mmol/L NaCl, 5 mmol/L KCl, 2 mmol/L CaCl₂, 1 mmol/L MgSO₄, 1 mmol/L NaH₂PO₄, 10 mmol/L glucose, 20 mmol/L HEPES, pH 7.4) for 30 min at 37°C and mounted in a fluorometer (Photon Technology International, Birmingham, NJ). Cells were perfused on the apical surfaces with Cl⁻ solution (110 mmol/L NaCl, 5 mmol/L KCl, 1 mmol/L CaCl₂, 1 mmol/L MgSO₄, 10 mmol/L glucose, 25 mmol/L NaHCO₃, 1 mmol/L amiloride, 5 mmol/L HEPES, 95% O₂/5% CO₂) or Cl⁻-free solution (110 mmol/L Na-gluconate, 5 mmol/L K-gluconate, 5 mmol/L Ca-gluconate, 1 mmol/L Mg-gluconate, 10 mmol/L glucose, 25 mmol/L NaHCO₃, 1 mmol/L amiloride, 5 mmol/L HEPES, 95% O₂/5% CO₂) at a flow rate of 1 mL/min; the basolateral perfusion contained the Cl⁻ solution throughout the experiment. The switch from Cl⁻-containing to Cl⁻-free solution generates a Cl⁻-out concentration gradient that stimulates ${\text{HCO}}_3^{-}$ entry across the cell membrane performed by Cl⁻/${\text{HCO}}_3^{-}$ exchanger(s), and the resulting rate of change in pH_i was recorded. Multiple rounds of removing (indicated by closed arrowheads)/replenishing (indicated by open arrowheads) extracellular Cl⁻ were performed to determine the Cl⁻/${\text{HCO}}_3^{-}$ exchange activity under basal conditions as a time control as well as in the presence of several compounds, including FSK (1–10 µM, apical and basolateral, 7 min), adenosine 5′-triphosphate (ATP; 0.25–10 µM, apical), uridine-5′-triphosphate (UTP; 1 µM, apical), and carbachol (1 $\mathrm{\mu }$M, basolateral). At the end of each experiment, pH_i was calibrated using K-clamp solutions with 10 mmol/L nigericin (Cayman Chemical, Ann Arbor, MI) that were set at pH 6.8 and 7.8. The initial rate of alkalinization after the switch from Cl⁻ solutions to Cl⁻-free solutions was calculated using Origin 8.0 software (OriginLab, Northampton, MA) and in some cases, the stable pH_i reached after alkalinization was determined. To determine the effect of the DRA inhibitor, DRAinh-A250 on DRA activity, the control DRA activity was determined by one cycle of Cl removal and readdition to colonoids grown on monolayers. The colonoids were then exposed to 3 µM A250 in both apical and basolateral Cl⁻ perfusate for 20 min. Apical Cl⁻ perfusate was then switched to Cl⁻-free perfusate containing the same concentration of A250 for determining Cl⁻/${\text{HCO}}_3^{-}$ exchange activity. For determining FSK/ATP-stimulated Cl⁻/${\text{HCO}}_3^{-}$ exchange activity, FSK was included in Cl⁻/A250 perfusate for the last 7 min with A250. ATP was included in the apical Cl⁻-free/A250 perfusate.

Apical Cl⁻/${\text{HCO}}_3^{-}$ exchange in these Caco-2/BBE cells was shown to be contributed by DRA based on prevention of the change in apical Cl⁻ removal-induced alkalinization by pretreatment with DRA inhibitor-A250 (5 µM; 13). We previously reported that the IC₅₀ values of DRAinh-A250 in HEK293/DRA cells, Caco-2/BBE cells, and human colonoids were similar and were 0.12 ± 0.04 µmol/L, 0.53 ± 0.10 µmol/L, and 0.22 ± 0.08 µmol/L, respectively (13). DRA knockout in Caco-2 cells completely eliminated Cl⁻-dependent pH_i changes (13). We further confirmed that DRAinh-A250 inhibited basal DRA activity and FSK (0.5 µM)/ATP (10 µM) stimulated DRA activity in colonoids in a similar manner (81 ± 3% vs. 76 ± 8% inhibition, respectively) (Supplemental Fig. S1, A–C).

Intracellular Ca²⁺ Measurements

Plasmid pGP-CMV-GCaMP6s (http://addgene.org/40753) was cloned into lentiviral vector pCDH-EF1-MCS-IRES (puro) and lentiviral particles were produced. Caco-2/BBE wild-type cells were transduced with GCaMP6s lentiviral particles and stable Caco-2/GCaMP6s cell lines were generated with puromycin selection (10 μg/mL). Caco-2 cells stably expressing GCaMP were seeded into 12-well Transwell (polyester membrane, Corning No. 3460) plates (1 × 10⁵ cells/well). Cells grown for 12–15 days postconfluency were serum-deprived for 2–3 h before the study. Live images of GCaMP fluorescence were obtained every second with an Olympus FV30000RS confocal microscope with ×10/0.75 numerical aperture (NA) objective lens using the resonance mode at 37°C in the OkoLab stage top plus transparent shroud incubator. ATP (0.25 µM or 1 μM) was added apically at specific time points without interrupting the live cell imaging. Time series were acquired with 512 × 512 pixels, 300 × 300-μm field of view for 600 time points at ∼1 s per time point. In some studies, EGTA (2 μM) was added ∼1 min before FSK and/or ATP addition. In other studies, cells were treated with 25 µM 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetrakis(acetoxymethyl ester) (BAPTA-AM) for 20–30 min before FSK and/or ATP addition. Intensity of GCaMP6s fluorescence was quantitated by MetaMorph. To determine changes in Ca²⁺, ATP-induced changes in GCaMP6s fluorescence (F) were divided by baseline fluorescence (F₀) and compared with maximum F/F₀ that occurred with 1 µM ionomycin.

Statistical Analyses

Results are expressed as means ± SE. Statistical evaluation was done by Student’s t tests or ANOVA with Bonferroni correction when three or more comparisons were made. P values ≤ 0.05 were considered significant.

RESULTS

DRA Is Acutely Stimulated by Elevated Intracellular Ca²⁺

We previously demonstrated that cAMP (FSK) acutely stimulated DRA activity in Caco-2/BBE cells and human colonoids (13). Here, we determined the effect of elevating intracellular Ca²⁺ on DRA activity in these cells. Receptor-mediated Ca²⁺ release from intracellular calcium stores, including endoplasmic reticulum (ER), occurs through both intestinal P2Y purinergic receptors and cholinergic M3 receptors that includes activation of phospholipase C (PLC; 33, 34). As shown in Fig. 1, A and B, in which DRA was measured in the same Caco-2/BBE monolayer under basal conditions and after ATP (1 µM) exposure on the apical surface, ATP rapidly stimulated DRA activity. ATP (1 µM) caused ∼40% stimulation of basal BB DRA activity. Similar experiments determined the effects of 10 and 0.25 µM ATP. The ATP-induced stimulation at 10 µM ATP was slightly but not significantly greater than that at 1 µM ATP, whereas minimal (<10%) ATP stimulation occurred with apical 0.25 µM ATP exposure (Fig. 1C). The effect of apical addition of 1 µM UTP was also determined. Similar to the effect of ATP, apical UTP rapidly stimulated DRA activity (Fig. 1D). Because rotavirus, which causes diarrhea and inhibits Na⁺ absorption as well as induces intestinal fluid secretion, was shown to cause calcium waves by activation of P2Y1 receptors (33), it was determined if pretreatment with the P2Y1 receptor antagonist N-[2-[2-(1,1-dimethylethyl)phenoxy]-3-pyridinyl]-N′-[4-(trifluoromethoxy)phenyl]urea (BPTU) prevented the ATP (1 µM) stimulation of DRA activity. As shown in Fig. 1E, BPTU (10 µM) pretreatment prevented ATP stimulation of DRA activity.

Figure 1.

ATP and UTP acutely stimulate DRA activity in Caco-2/BBE cells. A: Cl⁻/${\text{HCO}}_3^{-}$ activity was determined by measuring intracellular pH with BCECF/fluorometry as removing apical bathing Cl⁻ solution stimulated intracellular alkalinization that was reversed by readdition of Cl⁻ after a constant pH_i was reached. Polarized Caco-2/BBE cells with apical and basolateral surface perfused with Cl⁻ reached a steady-state pH_i and then Cl⁻ was removed only from the apical media and initial rates of intracellular alkalinization were determined and in some cases, the steady state of pH_i reached was also determined. Multiple cycles of removal (closed arrowhead)/readdition (open arrowhead) of Cl⁻ were performed and the effect of apical ATP (1 µM) was determined. Results from a single monolayer are shown with increased initial rate and steady-state pH_i after ATP exposure. The Cl⁻/${\text{HCO}}_3^{-}$ exchange activity is considered to be due to DRA activity since Cl⁻ removal-induced alkalinization in these cells was entirely prevented by treatment with DRA inhibitor A250 (3–5 µM) (13, 56 and Supplemental Fig. S1). B: effect of ATP (1 µM) on Cl⁻/${\text{HCO}}_3^{-}$ exchange activity, determined as in A. C: experiments as in A and B studying the effects of multiple concentrations of ATP, which were compared with DRA activity in the same monolayer before ATP perfusion; results were calculated with untreated DRA activity set as 100% for each experiment and shown as the horizontal dashed line. ATP (0.25 µM) caused a minimal stimulation of DRA, whereas 1 and 10 µM ATP caused similar significant stimulation compared with basal activity. D: apical UTP (1 µM) acutely stimulated apical Cl⁻/${\text{HCO}}_3^{-}$ exchange activity. The magnitude of increased DRA activity (compare B and D) caused by UTP was similar to that caused by ATP. E: acute ATP stimulation of DRA activity was prevented by pretreatment with 10 µM BPTU, a specific P2Y1 inhibitor, added apically and basolaterally. Results in B–E are means ± SE. In all experiments, n represents the number of separate experiments. BCECF, 2,7-Bis(2-carboxyethyl)-5-carboxyfluorescein; BPTU, N-[2-[2-(1,1-dimethylethyl)phenoxy]-3-pyridinyl]-N′-[4-(trifluoromethoxy)phenyl]urea; DRA, downregulated in adenoma; pH_i, intracellular pH.

Carbachol is known to elevate intracellular Ca²⁺ in intestinal epithelial cells via the M3 cholinergic receptor, acting on the endoplasmic reticulum (ER) through phospholipase C (PLC)-inositol 1,4,5-trisphosphate (IP₃) signaling (34). The effect of carbachol was determined on DRA activity in Caco-2/BBE cells stably transduced to express the cholinergic M3 receptor, which is not present endogenously (Fig. 2A). Addition of 1 µM carbachol to the basolateral surface rapidly stimulated DRA activity (Fig. 2, B and C) by ∼45% above basal activity in the same cells. These results support that receptor-mediated intracellular elevation of Ca²⁺ stimulates DRA activity in polarized human intestinal epithelial cells.

Figure 2.

Carbachol (CCH) acutely stimulates DRA activity in Caco-2/BBE-M3 receptor cells. A: Caco-2 cells were stably transduced with the cholinergic M3 receptor epitope tagged with HA. The M3 receptors were lacking endogenously in these cells. Lysates from these cells were immunoblotted for the epitope-labeled M3 receptor using anti-HA antibody, along with the loading control, β-actin. B and C: experiments similar to those described in Fig. 1A were performed, but examining the effect of basolateral carbachol on Cl⁻/${\text{HCO}}_3^{-}$ exchange activity in Caco-2/BBE-M3 cholinergic receptor cells. B: single experiment. C: means ± SE of multiple experiments, comparing basal and carbachol effects on Cl⁻/${\text{HCO}}_3^{-}$ activity from the same monolayers. n = number of separate experiments. DRA, downregulated in adenoma; HA, human influenza hemagglutinin.

Having shown that acutely elevated Ca²⁺ stimulates DRA activity, it was determined whether there was an interaction in the DRA stimulation by elevated cAMP and elevated Ca²⁺; specifically, whether Ca²⁺ elevation and increased cAMP affected DRA entirely independently or whether there were interactions such that their effects were additive, synergistic, or competitive.

cAMP and ATP Acutely Stimulate DRA in a Synergistic Manner

We previously reported that 10 µM FSK acutely stimulated DRA in Caco-2 cells and in human colonoids (13) and confirmed that finding (Fig. 3, A and B). To evaluate the interactions of cAMP and elevated Ca²⁺ in the acute stimulation of DRA, lower concentrations of FSK were evaluated for effects on DRA activity in comparison with the effects of the higher concentration. Figure 3C demonstrates that in comparison with DRA stimulation by 10 µM FSK, exposure to 1 µM did not have a significant effect on DRA activity.

Figure 3.

Acute DRA stimulation by FSK in Caco-2/BBE cells. Apical and basolateral FSK (10 µM) was exposed to Caco-2 cells and ∼7 min later, effects on DRA activity were determined as in Fig. 1A by apical Cl⁻ removal. A: Cl⁻/${\text{HCO}}_3^{-}$ activity in a single experiment. Arrowheads as in Fig. 1. B: DRA activity, means ± SE from multiple experiments comparing the effects on DRA activity of 1 and 10 µM FSK, demonstrating significant stimulation of DRA by 10 µM but not by 1 µM FSK. n = number of separate experiments. Ap, apical; BL, basolateral; FSK, forskolin.

SLC26A6 [putative anion transporter 1 (PAT-1)] and CFTR were previously shown to be stimulated in a synergistic manner by elevation of cAMP and Ca²⁺ (12, 35). As 0.25 µM ATP caused a very minimal stimulation of DRA while 1 µM FSK did not alter DRA activity, the additive effects at these concentrations were determined. In these studies, FSK (1 µM) was exposed to the Caco-2/BBE cells for ∼7 min in Cl⁻-containing media and then ATP (0.25 µM) was added in a Cl⁻-free solution to determine DRA activity. Although 1 µM FSK did not alter DRA activity, this combination of FSK and ATP significantly increased DRA activity by ∼70% above baseline (Fig. 4, A–C). These results demonstrate that concentrations of FSK and ATP, which themselves cause minimal or no effects on DRA activity, together cause a robust stimulation, thus demonstrating a synergistic interaction of cAMP and Ca²⁺. Further studies examined the effects of maximal concentrations of FSK (10 µM) and ATP (10 µM) alone and together on DRA activity. Although both 10 µM ATP (140 ± 2%) and 10 µM FSK (164 ± 6%) alone acutely stimulated DRA, added sequentially, as done in Fig. 4A, their DRA stimulation was not additive (184 ± 11%; Fig. 4D).

Figure 4.

Concentrations of FSK and ATP that have minimal effects on DRA activity cause synergistic stimulation of DRA in Caco-2/BBE cells. A: single experiment in which Cl⁻/${\text{HCO}}_3^{-}$ exchange activity was determined under basal, untreated conditions and after FSK (1 µM) addition to the apical and basolateral perfusions followed in ∼7 min by apical addition of ATP (0.25 µM). B: multiple experiments as in A, showing DRA activity as means ± SE. n = number of separate experiments. While neither FSK (1 µM) nor ATP (0.25 µM) alone significantly altered DRA activity, together they can stimulate DRA activity. C: multiple experiments as in A and B showing DRA activity as means ± SE but with concentrations of FSK (10 µM) and ATP (10 µM) that each maximally stimulate DRA activity and studied together with a time course as in A. Maximum concentrations of DRA plus ATP did not demonstrate additive or synergistic DRA stimulation. In all individual experiments, results were normalized to untreated DRA activity, which was set as 100% and is shown as the horizontal line. Results are means ± SE. n = number of separate experiments. P values calculated by comparison to untreated conditions for each experiment set as 100%. DRA, downregulated in adenoma; FSK, forskolin.

As DRA has been shown to take part in both neutral NaCl absorption, interacting with NHE3, and anion secretion, while the Caco-2/BBE cells used in these studies express a minimal amount of endogenous NHE3, we determined whether increased NHE3 expression altered the FSK/ATP synergistic DRA stimulation. Adenoviral-NHE3 transduction was used to transiently increase NHE3 expression. The trace of a representative experiment in Fig. 5A demonstrates that DRA stimulation in NHE3 expressing Caco-2/BBE cells with ATP (1 µM) is characterized by both an increase in initial rate of DRA activity and an increase in the steady-state pH_i reached and this is similar to the response in wild-type cells (Fig. 1A). Stimulation of DRA by FSK and ATP using concentrations that produced synergistic stimulation, as described in Fig. 4, also caused synergistic stimulation of DRA activity in Caco-2/NHE3 cells, which was slightly but not significantly smaller than that which occurred without the increase in NHE3 expression [compare Fig. 5B, 49 ± 5% and Fig. 4B, 64 ± 8%, not significant (ns)].

Figure 5.

FSK/ATP-mediated synergistic stimulation of DRA is similar in Caco-2/BBE cells acutely transduced with HA-NHE3. Experiments were performed to determine whether the presence of HA-NHE3 in Caco-2/BBE cells affected the FSK/ATP-induced synergistic stimulation of DRA. Caco-2 cells were transduced with adenovirus-HA-NHE3 and studied for Cl⁻/${\text{HCO}}_3^{-}$ exchange activity ∼48 h later. A: single experiment from the same Caco-2 monolayer with DRA activity determined under basal conditions and after apical addition of 1 µM ATP. The traces are superimposed starting at time 0, although they were performed sequentially. Note that ATP stimulated the initial rate of DRA as well as increased the steady-state pH_i achieved. B: similar studies as those performed in Fig. 4 with basal DRA activity and effects of adding FSK (1 µM) followed by ATP (0.25 µM) demonstrated. Results are means ± SE. n = number of separate experiments. DRA, downregulated in adenoma; FSK, forskolin; HA, human influenza hemagglutinin; NHE3, Na⁺/H⁺ exchanger isoform-3.

Cytosolic-Free Ca²⁺ Is Increased by ATP Alone and Synergistically Increased by the Low Concentrations of FSK and ATP, Which Synergistically Stimulate DRA

Changes in free intracellular Ca²⁺ caused by ATP alone and following FSK were determined in Caco-2/BBE cells stably expressing the genetically encoded Ca²⁺ indicator, GCaMP6s, that responds to elevation in free Ca²⁺ with an increase in fluorescent intensity. GCaMP6s was stably expressed in Caco-2/BBE cells by lentiviral transduction with puromycin selection, which created a cell line with GCaMP6s expression in nearly all cells, as demonstrated by confocal microscopy (Fig. 6A). ATP, when added apically to these cells, caused a very rapid increase in intracellular Ca²⁺ that occurred in a concentration-dependent manner between 0.25 and 10 µM ATP. ATP (0.25 µM) caused a single small peak in Ca²⁺ elevation that occurred in some cells and even this partial response only occurred in some but not all experiments. One and 10 µM ATP caused higher initial peaks in Ca²⁺ but also induced several small Ca²⁺ peaks that occurred at later times (Fig. 6B). In contrast to the effect of ATP, FSK at 1 (Fig. 7) and 10 µM (data not shown) did not alter Ca²⁺.

Figure 6.

Apical ATP causes a concentration-dependent increase in intracellular Ca²⁺ in Caco-2/BBE cells stably expressing the Ca²⁺ indicator GCaMP6s. A: the Ca²⁺ sensor protein GCaMP6s was stably expressed in Caco-2/BBE cells grown on glass bottom dishes and the effects of several concentrations of apical ATP were determined on intracellular Ca²⁺ in cells in an OkoLab stage top plus transparent shroud incubator, studied with Olympus FV3000RS confocal microscope [×20/0.75 numerical aperture (NA) objective] using the resonance mode at 37°C; exposure to ionomycin (1 µM) produced an intracellular fluorescence signal demonstrating expression of GCaMP6s in nearly all cells. There were similar findings when this experiment was repeated three times. Scale bar is 100 µm. B: exposure to several concentrations of apical ATP was quantitated as the ratio of the maximum fluorescence signal achieved to that originally present (F/Fo); ionomycin (1 µM) was added as a positive control at the end of study of each monolayer. ATP caused a concentration-dependent increase in intracellular Ca²⁺. Single traces of each concentration of ATP are shown. These studies were repeated 7–14 times, with quantitation for 0.25 µM ATP shown in Fig. 7.

Figure 7.

FSK and ATP synergistically increase intracellular Ca²⁺ in Caco-2/BBE-GCaMP6s cells. A: intracellular Ca²⁺was determined as in Fig. 6 in Caco-2/BBE-GCaMP6s cells exposed to 0.25 µM ATP followed by 1 µM ATP and then after 1 µM ionomycin as a positive control. In parallel, the FSK (1 µM)/ATP (0.25 µM) conditions that led to synergistic stimulation of DRA were studied. Synergistic FSK/ATP produced elevated Ca²⁺ similar to that produced by 1 µM ATP and that produced by 1 µM ionomycin. B: intracellular Ca²⁺ in Caco-2/BBE-GCaMP6s cells showing maximum increased Ca²⁺ caused by ATP (0.25 and 1 µM), FSK (1 µM), and synergistic concentrations of FSK (1 µM)/ATP (0.25 µM); studies also performed after pretreatment with BAPTA-AM (25 µM, 20–30 min) or EGTA (2 µM, 1 min) added before ATP and/or FSK. Ionomycin (1 µM) was added as a positive control. Results are means ± SE of maximum Ca²⁺ level. n = number of monolayers studied. BAPTA-AM prevented any changes in intracellular Ca²⁺, whereas EGTA pretreatment did not alter the increase in Ca²⁺ caused by 1 µM ATP and by FSK (1 µM)/ATP (0.25 µM). BAPTA-AM, 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetrakis(acetoxymethyl ester); FSK, forskolin.

Whether the synergistic stimulation of DRA by FSK and ATP was reflected by a synergistic increase in intracellular Ca²⁺ was determined. Addition of FSK (1 µM) and ATP (0.25 µM) with the same time sequence as in Fig. 4 caused a rapid increase in intracellular Ca²⁺ following ATP addition (Fig. 7A). This increase was similar in magnitude to the effect of 1 µM ATP and to the increase caused by 1 µM ionomycin, which was used as a positive control (Fig. 7, A and B). The elevation of intracellular Ca²⁺ by ATP (0.25 and 1 µM) and the FSK/ATP concentrations causing the synergistic increase in Ca²⁺ were totally prevented by pretreatment of the Caco-2/BBE cells with BAPTA-AM (25 µM, 20–30 min pretreatment; Fig. 7B), whereas exposure to 2 µM ethylene glycol-bis(β-aminoethyl ether)-N,N,N′,N′-tetraacetic acid (EGTA) for 1 min before ATP exposure did not alter the 1-µM ATP-induced increase in Ca²⁺ or the synergistic increase caused by FSK (1 µM)/ATP(0.25 µM; Fig. 7B).

Elevated Ca²⁺ Is Necessary for Synergistic Stimulation of DRA by FSK/ATP

Whether the synergistic stimulation of DRA activity was Ca²⁺ dependent was determined, using the same conditions that were demonstrated to prevent the FSK/ATP increase in intracellular Ca²⁺. Pretreatment of Caco-2/BBE cells with 25 µM BAPTA-AM for 20–30 min abolished the synergistic stimulation of FSK (1 µM)/ATP (0.25 µM). In addition, BAPTA-AM pretreatment decreased basal DRA activity, although there was a wide range of basal DRA activities in different experiments (Fig. 8).

Figure 8.

BAPTA-AM prevents the FSK/ATP synergistic increase in DRA activity. A: Cl⁻/${\text{HCO}}_3^{-}$ exchange activity was determined in Caco-2/BBE cells under basal conditions and when stimulated by FSK (1 µM)/ATP(0.25 µM) as in Fig. 4. Similar studies were performed in monolayers exposed to BAPTA-AM (25 µM, 20–30 min pretreatment). BAPTA-AM pretreatment prevented the synergistic stimulation of DRA. Results are means ± SE. n = number of separate experiments. B: comparison of basal DRA activity from experiments including those in A, demonstrating that BAPTA-AM significantly reduced basal DRA activity. Results are means ± SE. n = number of separate experiments. BAPTA-AM, 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetrakis(acetoxymethyl ester); DRA, downregulated in adenoma; FSK, forskolin.

Human Colonoids Exhibit cAMP/ATP Synergy

Since the low concentrations of FSK and ATP synergistically stimulated DRA in Caco-2/BBE cells and we previously reported that 10 µM FSK acutely stimulated DRA in human colonoids, we tested whether FSK and ATP also synergistically stimulated DRA activity in human colonoids. Low concentrations of FSK and ATP were evaluated for the stimulation of DRA. As shown in Fig. 9A, 0.5 µM FSK had no significant effect on colonoid DRA activity. Similarly, concentrations of ATP up to 10 µM had no significant effect (Fig. 9B). The additive/synergistic effects of FSK and ATP at these concentrations on DRA activity were similarly determined as in Caco-2/BBE cells. In these studies, FSK (0.5 µM) was exposed to the colonoid cells for ∼7 min in Cl⁻-containing media and then ATP (10 µM) was included in Cl⁻-free solution to determine DRA activity (Fig. 9C). This combination of FSK and ATP significantly increased DRA activity by 76 ± 22% (Fig. 9, C and D), confirming that FSK and ATP synergistically stimulate DRA in colonoids.

Figure 9.

Concentrations of FSK and ATP that have minimal effects on DRA activity cause synergistic stimulation of DRA in human colonoid monolayers. A: single experiment in which Cl⁻/${\text{HCO}}_3^{-}$ exchange activity was determined under basal, untreated conditions and after FSK (0.5 µM) addition to the apical and basolateral perfusions. Cycles of Cl⁻ removal and readdition to apical surfaces were indicated by closed and open arrowheads. The basolateral surfaces were always perfused with Cl⁻ medium. At this concentration of FSK, there was no effect on colonoid DRA. B: representative experiment showed that apical ATP (10 µM) had no effect on DRA activity. After the basal DRA activity was determined, ATP was included in apical Cl⁻-free medium for studying its effect on DRA. C: single experiment showed that FSK (0.5 µM)/ATP (10 µM) synergistically stimulated DRA activity. After the basal DRA activity was determined, FSK was included in both apical and basolateral Cl⁻-containing perfusates for ∼7min and the apical perfusate was switched to Cl⁻ free with ATP. D: multiple experiments as in A, B, and C showing DRA activity as means ± SE. The control untreated DRA activity was set at 100% and is indicated by a horizontal line. DRA activities measured with 0.5 µM FSK or 10 µM ATP are not statistically different from control. n = number of separate experiments. DRA, downregulated in adenoma; FSK, forskolin.

To identify differences in ATP effects on DRA activity and the requirement for higher ATP concentrations to cause synergistic stimulation of DRA activity in human colonoids compared with Caco-2 cells, colonoids stably expressing GCaMPs were studied. As shown in Supplemental Fig. S2, 0.5 µM FSK and 0.25 µM ATP did not significantly alter intracellular Ca²⁺, whereas 1 and 10 µM ATP similarly increased Ca²⁺.

DISCUSSION

This study demonstrates that the FSK and apical ATP induced elevated Ca²⁺ to act synergistically to acutely stimulate DRA in polarized intestinal epithelial cells. Synergistic interaction of cyclic nucleotide and Ca²⁺ signaling is not unique for the regulation of DRA in intestinal epithelial cells. In fact, cAMP/Ca²⁺ synergy occurs in many aspects of signal transduction across species, organs, and physiologic responses, as well as is involved in transcriptional and posttranscriptional signaling. Several examples include effects on odor detection in drosophila (36), synergistic GH release from chicken pituitary (37), carbaprostacyclin (cPG12) induction of differentiation of preadipocytes to mature adipocytes (38), stimulation of c-FOS in Jurkat cells (39), and prostaglandin E2 (PGE2)-induced stimulation of prolactin transcription in Jurkat cells (40). Although cAMP/Ca²⁺ synergy occurs in many signaling pathways, the cross talk between the two-second messenger systems does not always involve synergy; for example, cAMP and Ca²⁺ are additive but not synergistic in extracellular signal-regulated kinase (ERK) activation in a mouse model of long-lasting long-term potentiation, related to learning (41), and there are examples in which cAMP and elevated Ca²⁺ have opposing effects on the same signaling pathway, such as in metabotropic glutamate receptor signaling (42).

Intracellular Ca²⁺ was necessary for the cAMP/Ca²⁺ synergistic acute stimulation of DRA in Caco-2 cells since both the effect of ATP alone on DRA and the synergistic stimulation were inhibited by pretreatment with BAPTA-AM but not by extracellular EGTA at a concentration that greatly lowered the extracellular free Ca²⁺ but was only present for a minute before ATP addition so as not to lower intracellular Ca²⁺. Concerning the upstream signaling, a P2Y1 receptor inhibitor prevented ATP stimulation of DRA activity. Of interest, this is the same signaling pathway involved in spreading the Ca²⁺ effect across many cells following rotavirus infection of small intestinal villus enterocytes (33).

Multiple mechanisms for cross talk/synergy between cAMP/Ca²⁺ signaling have been previously defined in detail. Given the involvement of apical P2Y1 receptors with PLC-IP₃ generation, the studies by Muallem and colleagues (43) are most relevant to our demonstration in Caco-2 cells that synergistic stimulation of DRA requires the magnitude of elevation of intracellular Ca²⁺ also to increase synergistically. In the studies by Muallem and colleagues, cAMP increased the binding and sensitivity of IP₃ to its ER Ca²⁺ channel receptor, IP₃R, increasing the amount of Ca²⁺ released per molecule of bound IP₃, to synergistically elevate cytosolic Ca²⁺. However, multiple other mechanisms of cAMP/Ca²⁺ synergy have been identified. These include, among others, Ca²⁺-dependent adenylyl cyclase activity (44) and direct effects of calmodulin (CaM) binding to the CFTR R domain, with Ca²⁺-CaM binding increasing the open probability similar to the effects of cAMP-induced stimulation via phosphorylation, with the overall effects of Ca²⁺/CaM and cAMP being additive but together achieving the maximum activity caused by either alone (45). The latter results indicate effects via a common pathway. We have previously reported that FSK/cAMP elevates intracellular Ca²⁺ via Epac1 in T84 cells (46). Of note, in the present study, we showed that FSK failed to increase intracellular Ca²⁺ in Caco-2 cells and in human colonoids (Fig. 7A and Supplemental Fig. S2).

The current studies appear to help explain previous studies in the model cell, Caco-2/BBE, in which DRA seemed to be both stimulated and inhibited by cAMP (8, 13, 14). We previously reported, based on use of proximity ligation assays in Caco-2/BBE cells expressing transduced NHE3 and endogenous DRA and CFTR, that under baseline conditions, DRA bound NHE3 and CFTR and NHE3 bound DRA and CFTR. These associations were dynamic. With elevation of cAMP, the interaction of NHE3 with DRA and CFTR was reduced and that between DRA and CFTR was increased, with overall activity of DRA increasing (3). The seeming contradiction of both DRA stimulation and inhibition can be explained by occurrence of reduced neutral NaCl absorption related to the previously demonstrated cAMP-induced decrease in NHE3 on the plasma membrane (8) that reduced the pool that interacts with BB DRA. In addition, the elevation of cAMP leads to increased DRA trafficking to the apical membrane where DRA interacts with CFTR and leads to increased anion secretion, as previously shown by us (13). Assuming that this occurs in cAMP/Ca²⁺ synergy in normal intestinal epithelial cells as well as in Caco-2 cells, it will be important to define the magnitude of the separation of NHE3 from DRA (which is part of the mechanism of inhibition of intestinal Na absorption that occurs in diarrhea as NHE3 trafficking from the plasma membrane to the early endosomes) and that of increased binding of DRA with CFTR (which is part of the stimulation of anion secretion). In the present studies, we showed that synergistic stimulation of DRA was observed whether NHE3 was present or not, and interestingly the magnitude of synergistic stimulation in the presence of NHE3 appeared to be slightly less than when it is minimally present or absent.

The results presented here differ from the detailed studies of Lamprecht and Seidler, also using Caco-2/BBE cells (23, 47). Those studies acutely elevated intracellular Ca²⁺ with ionomycin or UTP and found that both did not alter initial rates of DRA activity but lowered the steady-state pH_i achieved after ionomycin/UTP. This was interpreted to indicate that elevated intracellular Ca²⁺ probably changed the pH_i sensitivity of DRA as the major way Ca²⁺ altered DRA activity. What explains the differences in our findings is that both FSK and apical ATP, as well as synergistic concentrations of FSK and ATP, stimulate DRA activity that can only be speculated about. The current studies, however, show DRA stimulation that involves both initial rates and steady-state pH_i, effects that were reversed by BAPTA-AM, demonstrating that potential injury from elevated Ca²⁺ was not occurring. It must be noted that another major difference in these two Caco-2/BBE studies relates to the mechanism of DRA stimulation. In the study by Lamprecht et al. (23), neither ionomycin nor UTP altered the plasma membrane expression of DRA, while we previously reported that, using super-resolution combined with confocal microscopy, changes in surface expression of DRA and NHE3 supported that synergist stimulation of DRA was associated with increased apical membrane DRA expression and reduced NHE3 surface expression (3). Of note, there was a major difference in the ways our studies were performed, consisting of the support the Caco-2 cells were grown on. In our studies, Caco-2 cells were grown on semipermeable supports (Transwell filters) allowing access to both apical and basolateral surfaces to bathing solutions with standardized polarization based on days postconfluency studied. Previous demonstration of Caco-2 polarization and differentiation has generally been done with cells grown on such semipermeable supports rather than the solid supports apparently used in the Lamprecht studies (23, 47).

We expanded our studies to normal human colonoids. While similar to Caco-2/BBE cells in that in colonoids, 0.5 µM FSK failed to stimulate DRA activity (Fig. 9A) and 10 µM FSK significantly stimulated it (13), 10 µM ATP failed to significantly affect colonoid DRA activity (Fig. 9B). Nonetheless, the combination 0.5 µM FSK/10 µM ATP synergistically stimulated DRA activity. In parallel studies, the 1-µM ATP-induced increase in Ca²⁺ (peak F/Fo reached as percent of ionomycin induced maximum increase in F/Fo) was similar in both cell models (Caco-2 cells 69 ± 10% and human colonoids 66 ± 7%) (compare Fig. 7B and Supplemental Fig. S2). While we can conclude that ATP and cAMP synergistically stimulate DRA activity, the human colonoid studies demonstrate that the elevated Ca²⁺ is not sufficient to stimulate DRA activity and indicate that there are additional downstream factors that are necessary. Further studies are needed to identify the additional factors and to understand the differences in mechanisms of synergy in Caco-2 cells and human colonoids.

The demonstration that cAMP and Ca²⁺ synergistically stimulate DRA activity has relevance to pathophysiological mechanisms of multiple diarrheal diseases in which both cAMP and Ca²⁺ are elevated. In each of these, there is a potential role for cAMP/Ca²⁺ cross talk/synergy but in none have detailed evaluation of these interactions been achieved. PGE2 acutely elevates both intracellular Ca²⁺ and cAMP in intestinal epithelial cells and is part of the pathophysiology of IBD (48, 49), with effects on multiple intestinal cell types that likely contribute to the IBD pathophysiology (enteric nervous system, macrophages, dendritic cells, others). Bile salts signal by cAMP and protein kinase C (PKC) as part of cholerrheic enteropathy in which increased bile salt concentrations in the lumen of the colon stimulate CFTR activity while inhibiting NaCl absorption (50, 51). In mouse B cells, cross linking the B cell antigen receptor using anti-IgM antibody initially elevates cAMP and subsequently increases intracellular Ca²⁺, which leads to synergistic increase in CD80 transcription. The latter is part of B cell activation/interferon γ production that participates in multiple viral diarrheas, including rotavirus (52). In addition, low levels of apical ATP are known to be present in some polarized cells (53, 54) with the cause being less certain, although potential explanations include cell turnover/death, due to translocation of immunological and inflammatory cells through the tight junctions, and effects of interactions with the microbiome, among others. These low levels of ATP potentially are able to act synergistically with changes in cAMP that occur as part of digestion, as well as in diarrheal diseases, to synergistically participate in what is considered normal physiological regulation of intestinal salt and water transport. Future studies evaluating the relevance of increases in both cAMP and Ca²⁺ in the same intestinal cells or in the multiple intestinal cells beyond the enterocytes that contribute to diarrheal disease pathophysiology are likely to provide a new understanding of the pathophysiology of diarrhea and to provide drug targets to develop for treating diarrheal diseases.

DATA AVAILABILITY

Data will be made available upon reasonable request.

SUPPLEMENTAL DATA

Supplemental figure legends: https://doi.org/10.6084/m9.figshare.22306264.

Supplemental Figs. S1 and S2: https://doi.org/10.6084/m9.figshare.22306429.

GRANTS

Studies were supported in part by the NIH/National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) RO1 DK026523 (to M.D.), RO1 DK116352 (to C.-M.T./M.D.), R24 DK099803 (to M.D.), PO1AI125181(to M.D.), and Digestive Disease Research Core Center Grant P30 DK089502 (to M.D.).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

R.S., V.S., M.D., and C.-M.T. conceived and designed research; R.S., R.L., and C.-M.T. performed experiments; R.S., R.L., M.D., and C.-M.T. analyzed data; R.S., V.S., M.D., and C.-M.T. interpreted results of experiments; R.S., R.L., M.D., and C.-M.T. prepared figures; R.S., M.D., and C.-M.T. drafted manuscript; R.S., V.S., M.D., and C.-M.T. edited and revised manuscript; R.S., R.L., V.S., M.D., and C.-M.T. approved final version of manuscript.