HCO3(-) secretion by murine nasal submucosal gland serous acinar cells during Ca2+-stimulated fluid secretion

Abstract

Airway submucosal glands contribute to airway surface liquid (ASL) composition and volume, both important for lung mucociliary clearance. Serous acini generate most of the fluid secreted by glands, but the molecular mechanisms remain poorly characterized. We previously described cholinergic-regulated fluid secretion driven by Ca(2+)-activated Cl(-) secretion in primary murine serous acinar cells revealed by simultaneous differential interference contrast (DIC) and fluorescence microscopy. Here, we evaluated whether Ca(2+)-activated Cl(-) secretion was accompanied by secretion of HCO(3)(-), possibly a critical ASL component, by simultaneous measurements of intracellular pH (pH(i)) and cell volume. Resting pH(i) was 7.17 +/- 0.01 in physiological medium (5% CO(2)-25 mM HCO(3)(-)). During carbachol (CCh) stimulation, pH(i) fell transiently by 0.08 +/- 0.01 U concomitantly with a fall in Cl(-) content revealed by cell shrinkage, reflecting Cl(-) secretion. A subsequent alkalinization elevated pH(i) to above resting levels until agonist removal, whereupon it returned to prestimulation values. In nominally CO(2)-HCO(3)(-)-free media, the CCh-induced acidification was reduced, whereas the alkalinization remained intact. Elimination of driving forces for conductive HCO(3)(-) efflux by ion substitution or exposure to the Cl(-) channel inhibitor niflumic acid (100 microM) strongly inhibited agonist-induced acidification by >80% and >70%, respectively. The Na(+)/H(+) exchanger (NHE) inhibitor dimethylamiloride (DMA) increased the magnitude (greater than twofold) and duration of the CCh-induced acidification. Gene expression profiling suggested that serous cells express NHE isoforms 1-4 and 6-9, but pharmacological sensitivities demonstrated that alkalinization observed during both CCh stimulation and pH(i) recovery from agonist-induced acidification was primarily due to NHE1, localized to the basolateral membrane. These results suggest that serous acinar cells secrete HCO(3)(-) during Ca(2+)-evoked fluid secretion by a mechanism that involves the apical membrane secretory Cl(-) channel, with HCO(3)(-) secretion sustained by activation of NHE1 in the basolateral membrane. In addition, other Na(+)-dependent pH(i) regulatory mechanisms exist, as evidenced by stronger inhibition of alkalinization in Na(+)-free media.

Figures

Figure 1.

Measurement of extracellular CO2–HCO3− buffer pHo and in vivo calibration SNARF pHi indicator fluorescence. (A) Measurement of pHo under experimental conditions. The experimental perfusion system was loaded with either Solution A (5% CO2–25 mM HCO3−) or Solution B (HEPES with pHo = 6.8, 7.2, or 7.6 and ungassed) containing 0.5 μg/ml BCECF free acid. BCECF 495/440 nm fluorescence excitation ratios (emission collected at 535 nm) were measured upon switching between the solutions with different pHo. (B) Mean BCECF fluorescence ratios (from experiment shown in A) were plotted vs. pHo (diamonds). Mean ratios were 2.74, 3.71, and 4.71 for solutions of pHo 6.8, 7.2, and 7.6, respectively. Solid line represents linear fit to the data with slope 2.46 ± 0.02 (s.d.) and intercept −14.0 ± 0.2 (s.d.). The BCECF fluorescence ratio in CO2–HCO3−-buffered Solution A (large black circle) was 4.32, corresponding to pHo of ∼7.44. (C) Example of an in vivo SNARF fluorescence calibration experiment. A SNARF-loaded serous acinar cell was treated with 10 μg/ml nigericin and exposed to 140 mM [K+]o solutions with pHo = 6.8, 7.2, and 7.6. (D) Composite calibration of SNARF 640/580 ratio measured in vivo as shown in C. Plotted are 78 raw data points (gray circles) from 26 experiments. Mean SNARF ratio values (±SEM; black diamonds) for pH 6.8, 7.2, and 7.6 were 1.29 ± 0.02, 1.84 ± 0.05, and 2.52 ± 0.08, respectively. Solid line represents linear regression fit to the raw data, with slope = 1.6 ± 0.1 (s.d.) and intercept −9.89 ± 0.7 (s.d.), used to convert SNARF fluorescence ratio values to pHi values in all subsequent experiments.

Figure 2.

Measurement of serous acinar cell pHi buffering capacity. (A) Representative NH4Cl pulse experiment showing pHi changes in response to varying [NH4Cl]o (from 0 to 20 mM) in Na+- and HCO3−-free Solution E. Intrinsic cellular buffering capacity (βi) was determined from NH4Cl−-induced pHi changes as described in Materials and methods. (B) Measurements of serous acinar cell βi plotted as a function of pHi. Data points (circles) represent 184 measurements made in 34 separate cells/acini from experiments as shown in A. Igor Pro software was used to fit βi data with an exponential function (equation of fitted line: βi = 0.88259 + 249.8·e((6.69-pHi)/0.243)). The dashed line represents CO2–HCO3−-dependent buffering (βHCO3-), calculated as 2.3 × [HCO3−]i (as described in Materials and methods). The total buffering curve (βt; u-shaped curve) is the sum of the βi and βHCO3- curves. Conversions of changes in pHi to OH− eq flux values were performed using either the βt or βi functions in the presence or absence of CO2–HCO3−, respectively.

Figure 3.

Serous acinar cell pHi responses to CCh stimulation. (A) Representative experiments showing pHi (black circles; left axes) and cell volume (red triangles; right axes) responses of SNARF-loaded serous acinar cells to stimulation with 100 μM CCh in CO2–HCO3−-buffered solution, showing transient acidification occurring concomitantly with CCh-induced cell shrinkage/Cl− efflux followed by a prolonged alkalinization. Upon removal of agonist, pHi relaxed back to near resting levels. A second identical response was observed upon reapplication of agonist (second panel). (B) Representative traces from cells stimulated with CCh in nominally CO2–HCO3−-free (HEPES-buffered) solution, resulting in ∼50% smaller acidification (compared with CO2–HCO3− conditions) but a similar subsequent pHi increase. (C) Summary of resting pHi (left), CCh-induced pHi decrease (middle), and CCh-induced OH− eq flux (right) in CO2–HCO3−- and HEPES-buffered conditions. (D) Summary of cell shrinkage (left) and Cl− fluxes (right).

Figure 4.

DMA enhances and prolongs CCh-induced acidification. (A) Representative experiment showing enhanced and prolonged CCh-induced acidification in serous acinar cells exposed to 30 μM DMA in presence of CO2–HCO3−. (B) After removal of CCh, 30 μM DMA significantly slowed pHi recovery but did not affect cell swelling (Cl− uptake). (C) Serous cells in CO2–HCO3−-free conditions stimulated with CCh in the presence of DMA exhibited small acidification nearly identical to that observed in DMA-free conditions. However, DMA strongly blocked the subsequent pHi increase observed in presence of CCh alone. (D) Summary comparing CCh-induced acidification (left) and OH− eq flux data (right) ± 30 μM DMA in CO2–HCO3−- and HEPES-buffered conditions.

Figure 5.

Blocking the driving force for conductive HCO3− efflux significantly inhibits CCh/DMA-induced acidification. (A) Representative experiments showing serous acinar cells stimulated with 100 μM CCh after pretreatment with 30 μM DMA in high (89 mM) [K+]o/low (103 mM) [Cl−]o conditions (Solution G), demonstrating that both cell shrinkage and acidification are blocked under these conditions. (B) Mean CCh-induced pHi decrease and cell shrinkage (top graph) along with peak CCh-induced OH− eq and Cl− fluxes (bottom graph) in Solution A (normal [K+]o/[Cl−]o) vs. Solution G (high [K+]o/low [Cl−]o). Asterisks represent significance when compared with 100 μM CCh stimulation in Solution A in the presence of 30 μM DMA.

Figure 6.

NFA inhibits CCh/DMA-induced acidification and cell shrinkage/Cl− efflux, suggesting that both Cl− and HCO3− effluxes occur through a similar Cl− channel pathway. (A) Representative experiment showing that application of 100 μM NFA completely blocked CCh-induced cell shrinkage/Cl− efflux without significantly inhibiting CCh-induced [Ca2+]i signals. (B) Representative experiment showing that NFA blocks CCh-induced acidification (∼80%) in the presence of 30 μM DMA. NFA inhibited cell shrinkage to a similar extent. (C) Summary of acidification and cell shrinkage (top) along with peak CCh-induced Cl− and OH− eq fluxes (bottom). Asterisks represent significance compared with 100 μM CCh stimulation in the presence of 30 μM DMA but in the absence of NFA.

Figure 7.

Inhibition of NKCC does not affect CCh-induced pHi dynamics, suggesting that CCh/DMA-induced acidification is not dependent on changes in [Cl−]i and that NFA block of HCO3− efflux is not dependent on block of Cl− efflux. (A) Recovery of pHi after CCh + DMA-induced acidification was not dependent on influx of Cl− (cell swelling; blocked with 100 μM bumetanide). (B) Representative experiment showing NFA inhibition of CCh + DMA-induced acidification under conditions of low [Cl−]i (shrunken cell with Cl− uptake inhibited by bumetanide).

Figure 8.

NHE isoform mRNA transcript expression in murine airway gland serous acinar cells. Messenger RNAs harvested from small airway gland serous acini were amplified and subject to rtPCR using NHE isoform-specific primers (shown in Table II). (A–I) Representative EtBr-stained agarose gels showing PCR products demonstrating expression of NHE isoforms 1–9. Serous acinar cells (SMG acinar aRNA) expressed NHEs 1, 2, 3, 4, 7, 8, and 9. These isoforms were also detected in RNA isolated from nasal tissue.

Figure 9.

The NHE1 isoform is the major alkalinizing mechanism during CCh stimulation. (A–E) Representative experiments showing pHi recovery (alkalinization) following acidification by 100 μM CCh stimulation in the presence of 30 μM DMA (∼90 s; solid black bar) and subsequent removal of CCh and exposure to buffer only (A), 1 μM DMA (B), 30 μM DMA (not depicted), 100 μM DMA (C), 0.5 μM EIPA (not depicted), 1.5 μM cariporide (D), or 0.8 μM S3226 (E; open bars). Scales of x axes are identical, with the time of CCh removal set as t = 0 to facilitate comparisons among experiments. All experiments performed in presence of CO2–HCO3−. (F) Summary from replicate experiments as shown above. Low concentrations of DMA and EIPA (1 μM and 0.5 μM, respectively) and the NHE1 inhibitor cariporide inhibited alkalinization similarly to a high concentration of DMA (100 μM), whereas the NHE3 inhibitor S3226 had no effect. Asterisks represent significance compared with buffer only. (G) Representative experiments of serous cells stimulated with 100 μM CCh in the presence of either 1 μM DMA (left trace), 1.5 μM cariporide (middle), or 1 μM S3226 (right). Cells stimulated in presence of DMA or cariporide exhibited enhanced prolonged acidification, while cells exposed to S3226 showed only transient acidification.

Figure 10.

Confocal immunofluorescence microscopy indicates that NHE1 localizes to the basolateral membrane of serous acinar cells and acini.(A) Isolated, fixed serous acini exhibited NKCC1 immunofluorescence localized to the basolateral membranes. Top and bottom panels represent two separate focal planes imaged through same acinus. (B) Basolateral NKCC1 immunofluorescence did not overlap with CFTR, shown previously to be an apical serous cell marker. (C) NHE1 immunostaining revealed basolateral immunofluorescence pattern similar to that for NKCC1. (D) NHE1 immunofluorescence did not overlap with CFTR immunofluorescence. (E) Preincubation of NHE1 antibody with antigenic peptide reduced NHE1 immunofluorescence. Micrographs in E taken with identical system settings (filters, camera gain, exposure time, laser power) as used for images in D. Scale bar in each micrograph represents 10 μm.

Figure 11.

Removal of Na+o enhances CCh-induced acidification and completely blocks pHi recovery. Representative experiments illustrating dependence of serous cell pHi on Na+o. (A) Removal of Na+o (0 Na+) in CO2–HCO3−(Solution C) caused slow acidification. Cells recovered to near resting pHi upon reintroduction of Na+o. (B) Stimulation of acinar cells with 100 μM CCh in 0 Na+ solution elicited enhanced acidification compared with 0 Na+ alone or CCh alone. (C) Cells stimulated with CCh in 0 Na+ exhibited slow, progressive acidification following immediate rapid CCh-induced acidification, and lacked pHi or volume recovery upon removal of CCh. (D) Removal of Na+o in HEPES buffer (Solution D) also caused prolonged acidification, with realkalinization upon reintroduction of Na+o. (E) CCh stimulation in HEPES-buffered 0 Na+ solution enhanced CCh-induced acidification, but to a smaller extent than observed in CO2–HCO3− buffer. (F) Summary of CCh-induced acidification (left) and peak OH− eq fluxes (right) in Na+-containing and 0 Na+ CO2–HCO3− and HEPES buffers.