PKCαβγ- and PKCδ-dependent endocytosis of NBCe1-A and NBCe1-B in salivary parotid acinar cells

Clint Perry, Olga J. Baker, Mary E. Reyland, Irina I. Grichtchenko

Abstract

We examined membrane trafficking of NBCe1-A and NBCe1-B variants of the electrogenic Na+-HCO3 cotransporter (NBCe1) encoded by the SLC4A4 gene, using confocal fluorescent microscopy in rat parotid acinar cells (ParC5 and ParC10). We showed that yellow fluorescent protein (YFP)-tagged NBCe1-A and green fluorescent protein (GFP)-tagged NBCe1-B are colocalized with E-cadherin in the basolateral membrane (BLM) but not with the apical membrane marker zona occludens 1 (ZO-1). We inhibited constitutive recycling with monensin and W13 and detected that NBCe1-A and NBCe1-B accumulated in vesicles marked with the early endosomal marker early endosome antigen-1 (EEA1), with a parallel loss from the BLM. We observed that NBCe1-A and NBCe1-B undergo massive carbachol (CCh)-stimulated redistribution from the BLM into early endosomes. We showed that internalization of NBCe1-A and NBCe1-B was prevented by the general PKC inhibitor GF-109203X, the PKCαβγ-specific inhibitor Gö-6976, and the PKCδ-specific inhibitor rottlerin. We verified the involvement of PKCδ by blocking CCh-induced internalization of NBCe1-A-cyan fluorescent protein (CFP) in cells transfected with dominant-negative kinase-dead (Lys376Arg) PKCδ-GFP. Our data suggest that NBCe1-A and NBCe1-B undergo constitutive and CCh-stimulated endocytosis regulated by conventional PKCs (PKCαβγ) and by novel PKCδ in rat epithelial cells. To help develop a more complete model of the role of NBCe1 in parotid acinar cells we also investigated the initial phase of the secretory response to cholinergic agonist. In an Ussing chamber study we showed that inhibition of basolateral NBCe1 with 5-chloro-2,3-dihydro-3-(hydroxy-2-thienylmethylene)-2-oxo-1H-indole-1-carboxamide (tenidap) significantly decreases an initial phase of luminal anion secretion measured as a transient short-circuit current (Isc) across ParC10 cell monolayers. Using trafficking and functional data we propose a model that describes a physiological role of NBC in salivary acinar cell secretion.

  • internalization
  • model of saliva secretion
  • Ussing chamber
  • SLC4A4

we aimed to investigate endocytosis of two members of the SLC4A4 gene family, NH2-terminal variants of the electrogenic Na+-HCO3 cotransporter (NBCe1), NBCe1-A and NBCe1-B, which have been detected in the same tissues, e.g., pancreas, kidney, intestine, eye, and submandibular gland (9, 11, 19, 69, 72, 73). Moreover, a few recent studies have detected both variants in the same cells, e.g., renal proximal tubule (PT) and pancreatic islet cells (11, 82). NBCe1-A and NBCe1-B move Na+ and HCO3 across the membrane to mediate transepithelial absorption or secretion of HCO3 and maintain cell homeostasis through pH regulation (68). The electrogenic Na+-HCO3 cotransporters, found in parotid acinar cells (59, 67, 70, 84, 90), are likely to play similar roles helping regulate secretion in response to cholinergic signaling via muscarinic receptors, as well as maintain optimal intracellular pH (pHi) and homeostasis.

We previously reported (56) that the functional activity of NBCe1 is regulated (inhibited) in an endocytosis-dependent manner in Xenopus oocytes. Since endocytosis is an important mechanism that controls surface expression and regulates functional activity of many membrane proteins (8, 10, 15, 25, 36, 74, 91), we investigated whether cholinergic agonist treatment induces endocytosis of NBCe1-A and NBCe1-B in cultured parotid acinar cells (ParC5 and ParC10).

Endocytosis of many plasma membrane transporters is PKC dependent and in some cases mediated by specific PKC isoforms (20, 21, 33, 7981, 92). We previously showed (58) PKC dependence of endocytosis of endogenous NBCe1 in cultured parotid acinar cells. We also reported (57) that PKCε mediates endocytosis of the NBCe1-A variant in Xenopus oocytes. Here we aimed to investigate the putative PKC isoform dependence of regulated endocytosis of NBCe1-A and NBCe1-B in ParC5 and ParC10 cells. Our previous study was unable to differentiate between both variants of NBCe1; therefore, we utilized confocal fluorescent microscopy to study cells transfected with NBCe1-A-yellow fluorescent protein (YFP)or NBCe1-B-green fluorescent protein (GFP) or cotransfected with NBCe1-A-cyan fluorescent protein (CFP) together with a dominant-negative kinase-dead (Lys376Arg) PKCδ-GFP mutant. Our studies show that NBCe1-A and NBCe1-B localize at the basolateral membrane (BLM) of ParC5 cells and undergo constitutive and PKC-dependent internalization stimulated by the PKC activator phorbol 12-myristate 13-acetate (PMA) and carbachol (CCh). We also show that conventional PKCαβγ and novel PKCδ isoforms control endocytosis of the two NH2-terminal variants of NBCe1. To investigate whether functional inhibition of NBCe1 decreases parotid acinar cell secretion, we used 5-chloro-2,3-dihydro-3-(hydroxy-2-thienylmethylene)-2-oxo-1H-indole-1-carboxamide (tenidap), reported to be an NBCe1 inhibitor in renal PT and pancreatic islet cells (18, 82). Tenidap caused a concentration-related inhibition of cholinergic agonist-stimulated luminal anion secretion, measured as the transient short-circuit current (Isc) across ParC10 cell monolayers in Ussing chambers. We propose a physiological model of the role of Na+-HCO3 cotransporters in salivary acinar cell secretion based on our previous reports and our present trafficking and functional data.

EXPERIMENTAL PROCEDURES

Materials.

GF-109203X (GF), Gö-6976 (Gö), rottlerin, W13, CCh, monensin, and atropine were purchased from Sigma Chemical (St. Louis, MO). Zona occludens 1 monoclonal antibody (ZO-1) and E-cadherin monoclonal antibody (E-Cad) were from Invitrogen (Carlsbad, CA). Early endosome antigen-1 (EEA1) monoclonal antibody and Cy3 were from BD Transduction Laboratories (San Jose, CA). PMA was purchased from Calbiochem (La Jolla, CA). Secondary antibody Alexa Fluor-TR was purchased from Invitrogen. Fugene HD transfection reagent was purchased from Roche (Indianapolis, IN). ECL Cell Attachment Matrix was purchased from Upstate-Millipore (Billerica, MA). Filters used in confocal microscopy were purchased from Corning (Corning, NY). Filters used in the Ussing chamber studies were purchased from Becton Dickinson (Franklin Lakes, NJ). Vectashield mounting medium was purchased from Vector Laboratories (Burlingame, CA). Tenidap was a gift of Pfizer (Groton, CT). Fura-2 acetoxymethyl ester (AM) was purchased from Molecular Probes (Eugene, OR). The EasyMount electrode set was purchased from Physiologic Instruments (San Diego, CA). The Dual-Wavelength Fluorescence Imaging System was purchased from Intracellular Imaging (Cincinnati, OH). All other chemicals were purchased from Sigma Chemical.

Drug treatments.

Monensin was made as a 10,000× stock solution in MeOH. The final working concentration of MeOH was 0.01%. CCh, atropine, and W13 were made as 1,000× stock solutions in distilled H2O. PMA and tenidap were made as 1,000× stock solutions in DMSO. The final working concentration of DMSO was 0.1%. All drugs were diluted in culture medium to final concentrations before use. Where indicated, drugs were diluted in serum-free medium. As controls, we used 0.01% MeOH and 0.1% DMSO. Drugs or vehicles were added to serum-free culture medium, and cells were kept at 37°C in a humidified atmosphere of 5% CO2 and 35% oxygen balanced with air. For the Ussing chamber studies, tenidap and CCh were added to the basolateral side of cell monolayers.

ParC5 and ParC10 cell culture and transfection.

ParC5 and ParC10 cell lines were derived from the same freshly isolated rat parotid gland acinar cells by transformation with simian virus 40 and exhibit morphological, biochemical, and functional characteristics similar to freshly isolated acinar cells (85). Cells were cultured in modified DMEM-Ham's F-12 medium at 37°C in a humidified atmosphere of 5% CO2 and 35% oxygen balanced with air as described previously (5, 58). YFP-tagged NBCe1-A (NBCe1-A-YFP), CFP-tagged NBCe1-A (NBCe1-A-CFP), and GFP-tagged NBCe1-B (NBCe1-B-GFP) constructs were made by subcloning of the open reading frame of NBCe1 into mammalian pEYFP-C1, pECFP-C1, and pEGFP-C1 vectors, respectively (Clontech). The GFP-tagged wild-type PKCδ (PKCδ-WT-GFP) and GFP-tagged kinase-dead (Lys376Arg) PKCδ mutant (PKCδ-KD-GFP) were made as previously described (13). Cells were grown up to 40–60% confluence for 24 h before transfection. Fugene HD transfection reagent in serum-free medium was used to transfect cells with 0.5 μg/ml cDNA.

Immunofluorescence and confocal microscopy.

Polarized cells were grown in 24-well plates on 12-mm glass coverslips treated for 30 min at 37°C with ECL Cell Attachment Matrix and rinsed with Ca2+-free PBS, or, where indicated, cells were plated on permeable filter supports. After 4 days in culture cells were serum starved for 15 min, and drugs or vehicles were added to the serum-free culture medium. Cells were kept at 37°C in a humidified atmosphere of 5% CO2 and 35% oxygen balanced with air. For immunocytochemistry cells were washed with PBS, fixed, permeabilized, and incubated in blocking solution. After fixation cells were labeled with anti-E-Cad (1:1,000), anti-ZO-1 (1:250), and anti-EEA1 (1:500) primary antibodies, followed by the respective Alexa Fluor-labeled secondary antibodies, and mounted with Vectashield on slides for qualitative and quantitative analysis through confocal microscopy as described previously (58). Fluorescent images were taken with an Olympus Spinning Disk confocal microscope controlled by Slidebook 4.0.10.2. Confocal images were captured in z-stack intervals of 0.5 μm with a 60× oil-immersion objective (1.45 numerical aperture). Under mercury illumination, the filter sets were FITC: excitation 460–480 nm, emission 495–540 nm; tetramethylrhodamine isothiocyanate (TRITC): excitation 535–555 nm, emission 570–625 nm; CFP: excitation 420–440 nm, emission 460–480 nm. Collapsed z stacks were done by creating a projection image in cells with Slidebook 4.2.0.10 (Olympus Spindisk fluorescent microscope).

Measurements of short-circuit current in ParC10 cell monolayers in Ussing chambers.

ParC10 cells in monolayers (up to 50 passages) were grown for ∼4 days on Falcon filters. Cell monolayers on permeable supports were mounted in modified Ussing chambers equipped with a recirculating water jacket. Five milliliters of Krebs-Ringer-HCO3 medium (in mM: 118 NaCl, 3 KCl, 1.2 MgSO4, 25 NaHCO3, 1.0 CaCl2, and 10 glucose, pH 7.5, 37°C) was added to the apical and basolateral reservoirs and oxygenated by bubbling with 95% O2 and 5% CO2. The transepithelial resistance (TER) of monolayers was measured by an EVOM epithelial voltohmmeter with miniature dual chopstick electrodes (World Precision Instruments, New Haven, CT). TER (Ω·cm2) was obtained after subtraction of ∼120-Ω bare filter resistance as tissue resistance values in ohms multiplied by effective membrane area. As described previously (5), agonist-stimulated changes in IscIsc) were measured continuously and transepithelial potential differences were monitored intermittently with a VCC-600 automatic voltage-clamp apparatus and an EasyMount electrode set (Physiologic Instruments) connected to the chamber system with 3% (wt/vol) agar-KCl bridges. Isc and automatic fluid resistance compensation current were applied through Ag-AgCl electrodes connected with 4% (wt/vol) agar-KCl bridges. ΔIsc (μA/cm2 membrane) was expressed as the peak change obtained in response to CCh minus the basal value in a resting cell.

Measurements of intracellular free Ca2+ concentration in ParC10 cells.

We quantified intracellular free Ca2+ concentration ([Ca2+]i) in single cells within polarized monolayers preloaded with the Ca2+-sensitive fluorescent dye fura-2 AM, using the InCyt Dual-Wavelength Fluorescence Imaging System (Intracellular Imaging). For preloading, cells were incubated in assay buffer [in mM: 120 NaCl, 4 KCl, 1.2 KH2PO4, 1.2 MgSO4, 1 CaCl2, 10 glucose, 15 mM HEPES, and 0.1% (wt/vol) BSA, pH 7.4] containing 2.0 μM fura-2 AM for 45 min at 37°C and then washed and further incubated for 20 min at 37°C. The cell monolayers on permeable supports were positioned on the stage of a fluorescence microscope, stimulated with CCh at 37°C, and exposed sequentially to 340/380-nm excitation. Fluorescence emission was detected at 505 nm and converted to [Ca2+]i with a standard curve created with solutions containing known concentrations of free Ca2+. Increases in [Ca2+]i were expressed as the peak response obtained under the indicated conditions minus the basal value.

Data analysis.

Qualitative Pearson's r correlation coefficient analysis was done to ascertain whether colocalization occurred between YFP-tagged NBCe1-A or GFP-tagged NBCe1-B and Alexa Fluor-Texas red or Cy3. At least three independent experiments with consistent results were done for each experimental condition. Images chosen for analysis were selected for whole cells that fluoresced (using bright field to determine outline and region of interest in analysis). Each experiment contained three separate coverslips of cells from which we took at least five images from different areas of each coverslip. Representative images of these experiments are shown in Figs. 25. Pearson's correlation coefficients (r values) were calculated to quantify endocytosis of NBCe1-A and NBCe1-B measured as increased colocalization of internalized NBCe1-A-YFP, NBCe1-A-CFP, or NBCe1-B-GFP with early endosomes marked with EEA1. Although images presented in results are collapsed z stacks, Pearson r values were obtained by using all planes of the z stack separately, essentially calculating values for a three-dimensional (3D) image and therefore eliminating the false correlation between different planes. Pearson's correlation is the correlation of intensity in one channel with another; a 0 value signifies no correlation, while a value of 1.0 signifies perfect correlation. A negative value signifies anticorrelation. The equation used is shown below. Ri is the intensity in channel 1 for pixel i, and Gi is the intensity in channel 2 for the same pixel. Rav and Gav are the average (mean) intensity values over all pixels. r=i(RiRav)·(GiGav)[i(RiRav)2·ii(GiGav)2]12 Obtained Pearson's r values were compared with standard critical values at P = 0.05 to determine significant correlation (i.e., colocalization of NBCe1 variants with EEA1). Using Slidebook 4.2.0.10 Cross Channel software, we acquired the Pearson's correlation coefficients that matched the intensity of the green fluorescence through the FITC filter channel (NBCe1-A-YFP or NBCe1-B-GFP) or the blue fluorescence through the CFP filter channel (NBCe1-A-CFP) with red fluorescence through the TRITC filter channel (EEA1). All averages are reported as means ± SD. The statistical significance of data was determined with an unpaired Student's t-test. Differences were considered significant at a level of P < 0.05.

RESULTS

Basolateral membrane localization of NBCe1-A and NBCe1-B in ParC5 cells.

Here we utilized YFP-tagged NBCe1-A (NBCe1-A-YFP) or GFP-tagged NBCe1-B (NBCe1-B-GFP) to transfect ParC5 cells grown on Transwell filters to form a tight monolayer (see low-magnification view in Fig. 1, A and C). After fixation, we stained some of the cells with the lateral membrane marker E-Cad, a type I cadherin transmembrane glycoprotein that is specifically expressed on the surface of epithelial cells, where it mediates the formation of cell-to-cell adherens junctions (55, 75, 88). We stained other cells with an antibody against the tight junction protein ZO-1, which is an apical membrane marker for truly polarized cells (17). In the typical apical and lateral view images shown in Fig. 1B, top, the “green” fluorescence of NBCe1-A-YFP is colocalized with the “red” E-Cad staining (yellow in merged images signifies colocalization of cotransporters with E-Cad). In contrast, staining for ZO-1 is visibly separate from the fluorescence of NBCe1-A-YFP (Fig. 1B, bottom). Similarly, the “green” fluorescence of NBCe1-B-GFP colocalized with the “red” E-Cad staining (Fig. 1D, top; yellow in a merged image signifies such colocalization), while the “green” fluorescence of NBCe1-B-GFP did not colocalize with ZO-1 staining (Fig. 1D, bottom). Our findings demonstrate BLM localization of NBCe1-A and NBCe1-B variants in ParC5 cells.

Fig. 1.

Yellow fluorescent protein (YFP)-labeled electrogenic Na+-HCO3 cotransporter (NBCe)1-A (NBCe1-A-YFP) and green fluorescent protein (GFP)-labeled NBCe1-B (NBCe1-B-GFP) are localized at the basolateral membrane (BLM) of ParC5 cells. A: low-magnification view of cells transfected with NBCe1-A-YFP. Left: monolayer of cells stained with the “red” BLM marker E-cadherin (E-Cad) with a few cells expressing NBCe1-A-YFP shown in merged “yellow” (high-magnification image of cell inside white rectangle is shown in B, top). Right: monolayer of cells stained with the “red” apical membrane marker zona occludens 1 (ZO-1), with 2 cells expressing NBCe1-A-YFP shown in “green” (lack of merged yellow) (high-magnification image of cell inside white rectangle is shown in B, bottom). B: high-magnification view of NBCe1-A-YFP-expressing cell. Top: apical view (AV) and lateral view (LV) images show fluorescence of NBCe1-A-YFP (“green,” left) in a single transfected cell stained with E-Cad (“red,” center). Merged image (“yellow,” right) represents overlap of NBCe1-A-YFP and E-Cad confirming BLM localization of NBCe1-A. Bottom: AV and LV images show fluorescence of NBCe1-A-YFP (“green”) in a single transfected cell stained with ZO-1 (“red”). The strong separation of “green” and “red” staining (lack of merged yellow) indicates that NBCe1-A does not colocalize with ZO-1 and is not present at the apical membrane. C: low-magnification view of cells transfected with NBCe1-B-GFP. Left: monolayer of cells stained with E-Cad (“red”), with 1 cell expressing NBCe1-B-GFP shown in merged “yellow” (high-magnification image of this cell is shown in D, top). Right: cells stained with ZO-1 (“red”), with 2 cells expressing NBCe1-B-GFP shown in “green” (lack of merged yellow) (high-magnification image of cell inside white rectangle is shown in D, bottom). D: high-magnification view of NBCe1-B-GFP-expressing cell. Top: AV and LV images show fluorescence of NBCe1-B-GFP (“green,” left) in a single transfected cell stained with E-Cad (“red”, center). Merged (“yellow,” right) represents overlap of NBCe1-B-GFP and E-Cad, confirming BLM localization of NBCe1-B. Bottom: AV and LV images show fluorescence of NBCe1-B-GFP (“green”) in a single transfected cell stained with ZO-1 (“red”). The strong separation of “green” and “red” staining indicates that NBCe1-B does not colocalize with ZO-1 and is not present at the apical membrane. ParC5 cells were grown on Transwell filters for 3 days. A z stack of optical sections was acquired through the FITC (“green”) and tetramethylrhodamine isothiocyanate (TRITC, “red”) filter channels. Collapsed z stacks of the entire cell through the z-axis are shown. Representative of at least 20 images from each of 3 independent experiments.

Recycling inhibitors “lock” NBCe1-A and NBCe1-B in early endosomes: evidence for constitutive endocytosis.

We used the recycling inhibitor monensin (3, 6, 56, 57, 60, 83) to determine whether A and B variants undergo constitutive endocytosis in ParC5 cells transfected with NBCe1-A-YFP and NBCe1-B-GFP. Fig. 2, A and B, show a strong BLM localization of the “green” fluorescence of NBCe1-A-YFP and NBCe1-B-GFP and a low level of colocalization with the “red” staining of the early endosomal marker EEA1 in untreated cells (Fig. 2, A and B, top) and in cells treated with the vehicle (0.01% MeOH; Fig. 2, A and B, 3rd rows). In contrast, in cells treated for 60 min with 50 μM monensin, we found a significant accumulation of the “green” fluorescence of NBCe1-A-YFP and NBCe1-B-GFP in intracellular endosomal-like formations and clear colocalization of these cotransporters with EEA1-stained early endosomes (Fig. 2, A and B, 2nd rows, insets; yellow in the merged images signifies colocalization of “green” cotransporters with “red” EEA1). We also found that a 60-min treatment with 100 μM W13, a calmodulin inhibitor that also acts as a recycling inhibitor (58), resulted in increased colocalization of NBCe1-A and NBCe1-B with EEA1 (Fig. 2, A and B, bottom).

Fig. 2.

Recycling inhibitors monensin and W13 increase localization of NBCe1-A-YFP and NBCe1-B-GFP in early endosomes: evidence for constitutive endocytosis in ParC5 cells. A and B: Images of NBCe1-A-YFP (A) and NBCe1-B-GFP (B). Top rows: control untreated cells (Untr). Left: strong BLM expression of “green” NBCe1-A-YFP (A) and NBCe1-B-GFP (B). Center: early endosome antigen-1 (EEA1) staining (“red”). Right: merged images. Second rows: monensin-treated cells (Mon). Left: loss from BLM and redistribution into endosome-like compartments of NBCe1-A-YFP (A) and NBCe1-B-GFP (B). Right: merged images (“yellow”) represent costaining of NBCe1-A-YFP with EEA1 (A) and NBCe1-B-GFP with EEA1 (B); this indicates that monensin “locks” cotransporters in early endosomes. Third rows: control vehicle-treated cells (MeOH). Left: strong BLM presence of NBCe1-A-YFP (A) and NBCe1-B-GFP (B). Bottom rows: W13-treated cells. Left: loss from BLM and redistribution into endosome-like compartments of NBCe1-A-YFP (A) and NBCe1-B-GFP (B). Right: merged images (“yellow”) represent costaining of NBCe1-A-YFP with EEA1 (A) and NBCe1-B-GFP with EEA1 (B); this indicates that W13 also “locks” cotransporters in early endosomes. In A and B, live cells were untreated or treated for 60 min with 50 μM monensin or vehicle [0.01% methanol (MeOH)]. A z stack of optical sections was acquired through the FITC (green) and TRITC (red) filter channels. Collapsed z stacks near BLM are shown. Images are representative of 3 independent experiments. Scale bars, 5 μm. Insets: high-magnification images of cell regions indicated by white rectangles. C and D: quantitative analysis using the Pearson's r correlation coefficient shows that monensin and W13 increase the localization of NBCe1-A and NBCe1-B in early endosomes. The Pearson's r values in untreated cells (Untr) were insignificant for NBCe1-A-YFP and EEA1 (C) and for NBCe1-B-GFP and EEA1 (D). Similarly insignificant values were found for vehicle-treated cells (MeOH). Pearson's r values for monensin-treated (Mon) and W13-treated cells were significant. C and D summarize data generated by analyzing cell images from 3 different preparations in experiments like those shown in A and B. Values are means ± SD. F ratio = 70.15 (NBCe1-A) and 70.98 (NBCe1-B) by 1-way ANOVA. *P < 0.0005 vs. untreated by 2-tailed Student's t-test. nd, No difference.

We performed qualitative 3D Pearson's r correlation coefficient analysis (see experimental procedures) to determine the degree of colocalization of NBCe1-A-YFP with EEA1 and NBCe1-B-with EEA1 for each treatment. As Fig. 2, C and D, show, the Pearson's r values in untreated cells were insignificant for both NBCe1-A-YFP (r = 0.13 ± 0.06, n = 94) and NBCe1-B-GFP (r = 0.13 ± 0.07, n = 85). In vehicle-treated cells the Pearson's r values were also insignificant for both NBCe1-A-YFP (r = 0.14 ± 0.06, n = 41) and NBCe1-B-GFP (r = 0.12 ± 0.05, n = 34). In striking contrast, in monensin-treated cells we found that Pearson's r values were significant for both NBCe1-A-YFP (r = 0.39 ± 0.08, n = 31) and NBCe1-B-GFP (r = 0.39 ± 0.08, n = 28). In W13-treated cells the Pearson's r values for colocalization of NBCe1-A-YFP with EEA1 (r = 0.42 ± 0.11, n = 28) and NBCe1-B-GFP with EEA1 (r = 0.43 ± 0.10, n = 23) were also significant.

In summary, we found that monensin and W13 recycling inhibitors cause a statistically significant, approximately threefold increase in the Pearson's r value for colocalization of NBCe1-A-YFP or NBCe1-B-GFP with EEA1 (P < 0.0005 in 2-tailed Student's t-test), indicating significant accumulation of cotransporters into early endosomes compared with untreated cells. Thus, with application of recycling inhibitors, internalized NBCe1 cotransporters are probably unable to return back to the surface and are essentially trapped within endosomes, including early endosomes. Therefore, our data suggest that under conditions of our experiments NBCe1-A and NBCe1-B undergo constitutive endocytosis in transfected mammalian secretory cells.

PMA stimulates endocytosis of NBCe1-A and NBCe1-B in ParC5 cells.

The PKC activator PMA is commonly used to stimulate endocytosis of membrane proteins in mammalian cells (43, 46, 80). Here we examined whether PMA causes endocytosis of A and B variants of NBCe1 in cells transfected with NBCe1-A-YFP or NBCe1-B-GFP. A 15-min treatment with 1 μM PMA significantly reduced the presence of NBCe1-A and NBCe1-B at the BLM (Fig. 3, A and B). At the same time, PMA significantly increased redistribution of NBCe1-A and NBCe1-B to early endosomes marked with EEA1 (Fig. 3, A and B, 2nd rows, insets; yellow in the merged images signifies colocalization of “green” cotransporters with “red” EEA1). In untreated cells and cells treated with vehicle (0.1% DMSO), the “green” fluorescence of NBCe1-A-YFP and NBCe1-B-GFP is strongly present at the BLM, and the level of their colocalization with “red” staining EEA1 is low (Fig. 3, A and B, top and bottom).

Fig. 3.

Phorbol 12-myristate 13-acetate (PMA) induces endocytosis of NBCe1-A-YFP and NBCe1-B-GFP in ParC5 cells. A and B: images of NBCe1-A-YFP (A) and NBCe1-B-GFP (B). Top rows: control untreated cells (Untr). Left: strong BLM expression of “green” NBCe1-A-YFP (A) and NBCe1-B-GFP (B). Center: EEA1 staining (“red”). Right: merged images. Second rows: PMA-treated cells. Left: loss from BLM and redistribution into early endosomes of NBCe1-A-YFP (A) and NBCe1-B-GFP (B). Right: merged images (“yellow”) represent costaining of NBCe1-A-YFP and EEA1 (A) and NBCe1-B-GFP and EEA1 (B). This indicates that PMA induces endocytosis of NBCe1-A (A) and NBCe1-B (B). Third rows: cells were treated with a mixture of PMA and GF-109203X (GF). Left: strong BLM presence as in untreated cells of NBCe1-A-YFP (A) and NBCe1-B-GFP (B). Bottom rows: control vehicle (DMSO)-treated cells. Left: strong BLM presence of NBCe1-A-YFP (A) and NBCe1-B-GFP (B). In A and B, cells were untreated or treated for 15 min with 1 μM PMA, a mixture of 1 μM PMA and 500 nM GF, or vehicle (0.1% DMSO). Scale bars, 5 μm. Insets: high-magnification images of cell regions indicated by white rectangles. C and D: quantitative analysis using Pearson's r correlation coefficient shows that PMA causes accumulation of NBCe1-A and NBCe1-B in early endosomes. Pearson's r values for NBCe1-A-YFP with EEA1 (C) and NBCe1-B-GFP with EEA1 (D) were insignificant in untreated (Untr) and vehicle DMSO-treated cells (DMSO), significant in cells treated with PMA, and insignificant in cells treated with mixture of PMA and GF. C and D summarize data generated by analyzing cell images from 3 different preparations in experiments like those shown in A and B. Values are means ± SD. F ratio = 44.10 (NBCe1-A) and 49.42 (NBCe1-B) by 1-way ANOVA. *P < 0.0005 vs. untreated by 2-tailed Student's t-test.

To prove that internalization of NBCe1 variants is regulated via PMA-induced activation of PKC, we applied a mixture of PMA with 0.5 μM GF, an inhibitor specific for PKC that inhibits all classes of PKC isoforms (Fig. 3, A and B, 3rd rows). In cells treated with a mixture of PMA and GF we observed a strong BLM presence of the “green” fluorescence of NBCe1-A-YFP and NBCe1-B-GFP combined with a decreased level of their colocalization with “red” EEA1 staining (Fig. 3, A and B, 3rd rows). As summarized in Fig. 3, C and D, in PMA-treated cells the Pearson's r values of colocalization of NBCe1-A-YFP with EEA1 (r = 0.40 ± 0.11, n = 34) and NBCe1-B-GFP with EEA1 (r = 0.42 ± 0.10, n = 24) were significant. In contrast, the Pearson's r values were insignificant in cells treated with vehicle (0.1% DMSO) or treated with PKC inhibitor GF added together with PMA (see Fig. 3, C and D). Thus in vehicle-treated cells we obtained r = 0.10 ± 0.08 (n = 24) for NBCe1-A and r = 0.12 ± 0.06 (n = 27) for NBCe1-B. In cells treated with a mixture of PMA and GF we found r = 0.14 ± 0.07 (n = 25) for NBCe1-A and r = 0.12 ± 0.06 (n = 17) for NBCe1-B. Our data demonstrate that in 15 min after application of PMA, NBCe1-A and NBCe1-B undergo significant redistribution from the BLM to early endosomal compartments, and this process was inhibited by the PKC antagonist GF. These data demonstrate that PMA stimulates PKC-regulated endocytosis of NBCe1 cotransporters in secretory cells.

Carbachol stimulates endocytosis of NBCe1-A and NBCe1-B in ParC5 cells.

We next examined cholinergic regulation of the membrane trafficking of NBCe1-A-YFP and NBCe1-B-GFP in transfected cells. Figure 4, A and B (2nd rows, insets), show that in CCh-treated cells (50 μM, 15 min) the presence of NBCe1-A and NBCe1-B at the BLM was significantly reduced, with parallel redistribution of both A and B variants to early endosomes marked with EEA1 (yellow in the merged images signifies colocalization of NBCe1 with EEA1). To examine whether CCh-induced endocytosis of NBCe1-A and NBCe1-B is mediated via muscarinic receptor signaling, we utilized the cholinergic receptor inhibitor atropine (58). One micromolar atropine added together with 50 μM CCh completely prevented CCh-induced loss of NBCe1-A and NBCe1-B from the BLM and noticeably decreased their colocalization with EEA1 (Fig. 4, A and B, 3rd rows). As summarized in Fig. 4, C and D, the Pearson's r values for colocalization of NBCe1-A-YFP and NBCe1-B-GFP with EEA1 were increased approximately threefold by CCh compared with untreated cells and this was significantly decreased by atropine inhibition of muscarinic receptor signaling. In CCh-treated cells we obtained r = 0.41 ± 0.09 (n = 42) for NBCe1-A and r = 0.44 ± 0.09 (n = 35) for NBCe1-B. In cells treated with a mixture of CCh with atropine we obtained r = 0.12 ± 0.06 (n = 28) for NBCe1-A and r = 0.13 ± 0.06 (n = 14) for NBCe1-B. These data suggest that cholinergic agonist via muscarinic receptor signaling stimulates internalization of NBCe1-A and NBCe1-B in rat parotid acinar cells.

Fig. 4.

Carbachol (CCh) induces muscarinic receptor-mediated, PKC-dependent endocytosis of NBCe1-A-YFP and NBCe1-B-GFP. A and B: images of NBCe1-A-YFP (A) and NBCe1-B-GFP (B). Top rows: control untreated cell (Untr). Left: strong BLM expression of “green” NBCe1-A-YFP (A) and NBCe1-B-GFP (B). Center: EEA1 staining (“red”). Right: merged images. Second rows: cells were treated with CCh. Left: loss of NBCe1-A-YFP (A) and NBCe1-B-GFP (B) from BLM and redistribution into early endosomes. Right: merged images (“yellow”) represent costaining of NBCe1-A-YFP and EEA1 (A) and NBCe1-B-GFP and EEA1 (B). Third rows: cells were treated with CCh and atropine (Atr). Fourth rows: cells were treated with CCh and GF. Fifth rows: cells were treated with CCh and Gö-6976 (Gö). Sixth rows: cells were treated with CCh and rottlerin (Rott). In A and B cells were untreated or treated for 15 min with 50 μM CCh or CCh mixture with 1 μM Atr, 500 nM GF, 1 μM Gö, or 5 μM Rott. Scale bars, 5 μm. Insets: high-magnification images of cell regions indicated by white rectangles. C and D: quantitative analysis using the Pearson's r correlation coefficient shows that CCh increases localization of NBCe1-A (C) and NBCe1-B (D) in early endosomes. Pearson's r values for NBCe1-A-YFP and NBCe1-B-GFP with EEA1 were insignificant in untreated cells (Untr), significant in CCh-treated cells, and insignificant in cells treated with CCh+Atr, CCh+GF, CCh+Gö, or CCh+Rott. C and D summarize data generated by analyzing cell images from 3 different preparations in experiments like those shown in A and B. Values are means ± SD. F ratio = 45.43 (NBCe1-A) and 49.56 (NBCe1-B) by 1-way ANOVA. *P < 0.0005 vs. untreated by 2-tailed Student's t-test.

Specific PKC inhibitors prevent CCh-induced endocytosis of NBCe1-A and NBCe1-B in ParC5 cells.

First we utilized GF, an inhibitor specific for conventional and novel PKC isoforms, to investigate PKC dependence of the CCh-induced endocytosis of NBCe1-A-YFP and NBCe1-B-GFP. In cells treated with a mixture of 0.5 μM GF and 50 μM CCh we observed that GF prevents CCh-induced internalization of NBCe1-A-YFP and NBCe1-B-GFP from the BLM, noticeably decreasing colocalization of both cotransporters with EEA1 staining (Fig. 4, A and B, 4th rows). Second, we used Gö, a specific inhibitor of conventional PKCαβγ isoforms, and observed that 1 μM Gö inhibits CCh-induced endocytosis of NBCe1-A-YFP and NBCe1-B-GFP (Fig. 4, A and B, 5th rows). Next, we used rottlerin, which is commonly used as an inhibitor of the novel PKCδ isoform (45). Figure 4, A and B (bottom rows), show that 1 μM rottlerin inhibits CCh-induced endocytosis of NBCe1-A-YFP and NBCe1-B-GFP.

As summarized in Fig. 4, C and D, the Pearson's r values for colocalization of NBCe1-A-YFP and NBCe1-B-GFP with EEA1 were insignificant when CCh was applied with PKC inhibitor GF, PKCαβγ inhibitor Gö, or PKCδ inhibitor rottlerin. In cells treated with CCh and GF we obtained r = 0.15 ± 0.07 (n = 18) for NBCe1-A and r = 0.14 ± 0.06 (n = 17) for NBCe1-B. In cells treated with CCh and Gö we found r = 0.14 ± 0.06 (n = 52) for NBCe1-A and r = 0.16 ± 0.07 (n = 37) for NBCe1-B. In cells treated with CCh and rottlerin we obtained r = 0.13 ± 0.08 (n = 25) for NBCe1-A and r = 0.11 ± 0.07 (n = 29) for NBCe1-B. This was in striking contrast to the significantly increased Pearson's r values in CCh-treated cells (Fig. 4, C and D; see values above). These data suggest PKC dependence and, in particular, specific PKCαβγ isoform dependence of CCh-stimulated internalization of NBCe1 variants. Our data also suggest that a specific PKCδ isoform could be involved in the regulation of CCh-stimulated endocytosis of NBCe1 in mammalian cells.

Dominant-negative kinase-dead (Lys376Arg) PKCδ mutant inhibits endocytosis of NBCe1-A in ParC5 cells.

Rottlerin indirectly inhibits activation of PKCδ in parotid acinar cells (28, 78). Therefore we aimed to abolish endogenous PKCδ activity to investigate further whether PKCδ is involved in regulation of CCh-stimulated endocytosis of NBCe1-A. We used a downregulation approach by transfecting cells with the GFP-tagged kinase-dead (Lys376Arg) PKCδ mutant (PKCδ-KD-GFP). As a control we used GFP-tagged wild-type PKCδ (PKCδ-WT-GFP). Next, we tagged NBCe1-A variant with CFP (NBCe1-A-CFP). This allowed us to simultaneously record “blue” fluorescence of NBCe1-A-CFP and “green” fluorescence of PKCδ-WT-GFP or PKCδ-KD-GFP in the same cells. First, we showed that NBCe1-A-CFP is localized to the BLM in untreated cells (data not shown) similar to NBCe1-A-YFP (Figs. 2A, 3A, and 4A, top). Next, we showed that CCh causes endocytosis of NBCe1-A-CFP (data not shown) similar to that of NBCe1-A-YFP (Fig. 4A, 2nd row), confirming that there is no difference due to tagging with cyan versus yellow fluorescent proteins.

To test our hypothesis that PKCδ regulates CCh-induced endocytosis of NBCe1-A, we used cells cotransfected with NBCe1-A-CFP and kinase-dead mutant PKCδ-KD-GFP (Fig. 5A). For control experiments we utilized cells cotransfected with NBCe1-A-CFP and wild-type PKCδ-WT-GFP (Fig. 5B). After 2 days in culture, cells were untreated or treated with 50 μM CCh and fixed. In a control study, Fig. 5B (bottom, insets) shows a typical image of the CCh-induced endocytosis of NBCe1-A-CFP in cells cotransfected with wild-type PKCδ-WT-GFP. Purple signifies colocalization of NBCe1-A-CFP (“blue”) with EEA1 staining (“red”) in merged images. The extent of colocalization was similar to that observed in CCh-treated cells transfected with NBCe1-A-CFP alone (data not shown) or with endocytosis of NBCe1-A-YFP (described above and shown in Fig. 4A, 2nd row). In striking contrast, we observed that CCh did not induce endocytosis of NBCe1-A-CFP in cells cotransfected with the kinase-dead mutant PKCδ-KD-GFP (Fig. 5A, 2nd row).

Fig. 5.

CCh induces PKCδ-dependent endocytosis of cyan fluorescent protein (CFP)-labeled NBCe1-A (NBCe1-A-CFP) in ParC5 cells. A: kinase-dead (KD) (Lys376Arg) PKCδ mutant inhibits CCh-induced endocytosis of NBCe1-A. Top: untreated (Untr) cell cotransfected with NBCe1-A-CFP and kinase-dead PKCδKD-GFP. First panel: strong BLM presence of NBCe1-A-CFP (“blue”). Second panel: EEA1 staining (“red”). Third panel: merged image of “blue” and “red.” Fourth panel: cytosolic distribution of PKCδKD-GFP (“green”). Bottom: CCh-treated cell cotransfected with NBCe1-A-CFP and PKCδKD-GFP; images show that NBCe1-A-CFP (“blue”) remains at the BLM. B: wild-type PKCδ does not inhibit CCh-induced endocytosis of NBCe1-A. Top: control untreated cell cotransfected with NBCe1-A-CFP and wild-type PKCδWT-GFP; images how a strong presence of NBCe1-A-CFP (“blue”) at the BLM, cytosolic distribution of PKCδWT-GFP (“green”), and EEA1 staining (“red”). Bottom: CCh-treated cell cotransfected with NBCe1-A-CFP and PKCδWT-GFP. First panel: loss of NBCe1-A-CFP (“blue”) from BLM and its redistribution into endosome-like compartments. Second panel: EEA1 staining (“red”). Third panel: “purple” in a merged image represents colocalization of “blue” NBCe1-A-CFP with “red” EEA1. Fourth panel: cytosolic PKCδWT-GFP (“green”). Scale bars, 5 μm. Insets: high-magnification images of the cell regions indicated by white rectangles. C: quantitative analysis using Pearson's r correlation coefficient shows PKCδ dependence of CCh-induced increase of NBCe1-A-CFP localization in early endosomes. Pearson's r values for NBCe1-A-CFP with EEA1 were insignificant in untreated and CCh-treated cells cotransfected with kinase-dead PKCδ mutant (bars 1 and 2); insignificant in untreated and significant in CCh-treated cells cotransfected with wild-type PKCδ (bars 4 and 5); and insignificant in untreated and significant in CCh-treated cells transfected with NBCe1-A-CFP alone (bars 5 and 6). C summarizes data generated by analyzing cell images from 3 different preparations in experiments like those shown in A and B. Values are means ± SD. *P < 0.0005 vs. untreated by 2-tailed Student's t-test.

As summarized in Fig. 5C, CCh treatment resulted in an insignificant Pearson's r value for colocalization of NBCe1-A-CFP with EEA1 in cells cotransfected with kinase-dead (Lys376Arg) mutant PKCδ-KD-GFP (r = 0.13 ± 0.07, n = 28) similar to that for untreated cells (r = 0.12 ± 0.07, n = 19) (bars 1 and 2). In contrast, the Pearson's r values for CCh-treated cells cotransfected with NBCe1-A-CFP and wild-type PKCδWT-GFP (r = 0.41 ± 0.09, n = 36) were significant and insignificant for untreated cells (r = 0.12 ± 0.06, n = 22) (Fig. 5C, bars 3 and 4). Pearson's r values obtained in CCh-treated cells transfected with NBCe1-A-CFP alone (r = 0.41 ± 0.09, n = 29) were also significant and insignificant for untreated cells (r = 0.11 ± 0.03, n = 16) (Fig. 5C, bars 5 and 6). These data strongly suggest that the novel PKCδ isoform regulates CCh-induced internalization of NBCe1-A in secretory cells.

Inhibition of NBCe1 by tenidap significantly decreased luminal anion secretion in ParC10 cell monolayers.

Previous studies have shown that CCh-stimulated secretion in ParC10 cell monolayers is due to an increase in transepithelial anion secretion carried by luminal HCO3 and Cl effluxes (5, 85). It was suggested that transport via basolateral NBCe1 supports luminal HCO3 efflux in several secretory epithelia (11, 69, 72, 73). Here we studied the effect of the selective NBCe1 inhibitor tenidap on CCh-stimulated transepithelial anion secretion measured as Isc in monolayers formed by ParC10 cells (5, 85). To perform these studies, the basolateral side of ParC10 cell monolayers was treated with 100 μM tenidap for 30 min. Next, monolayers were mounted in Ussing chambers, CCh was applied on the basolateral side, and immediate changes in Isc were monitored. Figure 6A-1 shows a typical trace of CCh-induced Isc in untreated or tenidap-treated ParC10 monolayers. The inhibitory effect of 100 μM tenidap is summarized in Fig. 6A-2. These results indicate that 100 μM tenidap almost completely blocks CCh-induced Isc. To determine the concentration-dependent effects of tenidap, a dose-response study was performed. Figure 6A-3 shows that tenidap blocked CCh-induced Isc in a concentration-dependent manner with an IC50 value of ∼2.3 × 10−8 M, without affecting TER (data not shown).

Fig. 6.

NBCe1 inhibitor tenidap blocks luminal secretion in CCh-stimulated ParC10 cell monolayers. A: tenidap inhibits CCh-induced short-circuit current (Isc). A-1: typical Isc traces were recorded as a function of time in monolayers stimulated with CCh (solid line) or pretreated for 30 min with 100 μM tenidap and subsequently stimulated with CCh (dashed line). Changes in IscIsc) were measured in response to 100 μM CCh added to the basolateral compartment of the Ussing chamber. ΔIsc (μA/cm2 membrane) is expressed as the peak change obtained in response to CCh minus the basal value in a resting cell. A-2: summarized data of CCh-induced ΔIsc (μA/cm2) obtained from monolayers that were not treated with tenidap (CCh) or treated with 100 μM tenidap (CCh+100 μM Ten). Values are means ± SD obtained in 12 individual measurements. *P < 0.05 by 2-tailed Student's t-test. A-3: dose-response curve of tenidap inhibition of Isc was obtained in monolayers pretreated for 30 min with tenidap concentrations from 0.001 to 100 μM. Normalized ΔIsc (%) was acquired as the ratio of ΔIsc in tenidap-treated to ΔIsc in tenidap-untreated cells. Values are means ± SD of results from 12 individual measurements at each concentration of tenidap. No changes in the basal current were detected when tenidap was added alone (data not shown). B: 100 μM tenidap does not alter intracellular calcium signaling in ParC10 cells. B-1: typical traces of intracellular free Ca2+ concentration ([Ca2+]i) in response to 100 μM CCh were similar in tenidap-untreated cells (left) and cells pretreated for 30 min with 100 μM tenidap applied to the basolateral side (center). No changes in basal [Ca2+]i were recorded in response to acutely applied 100 μM tenidap (right). Cells were loaded with fura-2 AM fluorescent dye and sequentially excited at 340/380 nm, and emission was detected at 505 nm and converted to [Ca2+]i with the InCyt Dual-Wavelength Fluorescence Imaging System. B-2: summarized data of changes in [Ca2+]i (Δ[Ca2+]i) were obtained in response to 100 μM CCh in tenidap-untreated cells [148.8 ± 11.4 nM (n = 8), CCh], in response to 100 μM CCh in cells pretreated for 30 min with 100 μM tenidap [122.4 ± 22.5 nM (n = 8), CCh+Ten], in cells acutely treated with 100 μM tenidap [5.1 ± 1.6 nM (n = 8), Ten], and untreated cells [9.5 ± 6.4 nM (n = 8), Control]. Δ[Ca2+]i (nM) was expressed as the peak change obtained in response to CCh minus the basal value in a resting cell. Values are means ± SD. *P < 0.005.

Tenidap has a negligible effect on the inhibition of Ca2+-activated anion conductance in oocytes (39, 48). However, it was reported that in human gingival fibroblasts tenidap may prevent Ca2+ influx through the plasma membrane and decrease [Ca2+]i (14, 24). Thus the above observed effect of tenidap on Isc in ParC10 cells could be due to alterations in intracellular Ca2+ signaling in these cells. Therefore, we investigated the effect of tenidap (100 μM) on cholinergic Ca2+ signaling by measuring [Ca2+]i in ParC10 cells loaded with the Ca2+-sensitive dye fura-2 AM. We found that CCh similarly stimulates [Ca2+]i in tenidap-treated or untreated cells, and tenidap alone does not change background [Ca2+]i in untreated cells, as indicated in the [Ca2+]i traces from single cells within the monolayers (see Fig. 6B-1) and summarized in Fig. 6B-2. These results indicate that tenidap does not affect CCh-mediated intracellular Ca2+ signaling in ParC10 cells.

In summary, our data suggest that tenidap may block Isc by inhibiting luminal HCO3 efflux, possibly by inhibiting basolateral NBCe1 in ParC10 cells. We found here that basolateral application of 100 μM tenidap produced ∼85% inhibition of Isc, which is possibly carried by 50% HCO3 and 50% Cl efflux (66, 85). Tenidap may inhibit HCO3 efflux by inhibiting NBCe1, suggesting the importance of NBCe1 for acinar cell HCO3 secretion. The major mechanism for Cl efflux is basolateral AE2, which is partially tenidap sensitive (40, 42, 48, 51, 70, 87). Thus tenidap may inhibit Isc by blocking NBCe1 and partially inhibiting AE2. However, future studies are needed to investigate the role of NBCe1 by using small interfering RNAs (siRNAs) targeting NBCe1.

DISCUSSION

Basolateral localization of NBCe1-A and NBCe1-B.

We previously described (7, 58) basolateral localization of the endogenous electrogenic NBCe1 in rat parotid acinar ParC5 cells, using an antibody raised against a COOH-terminal sequence that therefore does not differentiate between the NH2-terminal A and B variants of NBCe1. In the present study we examined localization of fluorescent-tagged NBCe1-A-YFP and NBCe1-B-GFP transfected in the same cell type. We show that in ParC5 cells NBCe1-A and NBCe1-B are expressed predominantly basolaterally, with insignificant cytosolic presence (Fig. 1). Our new data are in agreement with reports by others who have shown basolateral location of NBCe1-A and NBCe1-B in several secretory epithelia (1, 4, 9, 16, 26, 29, 32, 37, 41, 53, 61, 63, 69, 70, 72). Distribution of NBCe1 at the BLM domain may help to maintain transepithelial HCO3 efflux in secretory epithelia, including parotid acinar cells. This idea is supported by our functional data showing that inhibition of basolateral NBCe1 prevents an apical anion secretion measured as Isc across ParC10 cell monolayers in an Ussing chamber (see Fig. 6A and discussion below).

Constitutive endocytosis of NBCe1-A and NBCe1-B.

The activity of a membrane cotransporter depends on the number of molecules at the cell surface that can be regulated by constitutive or stimulated endocytosis in resting or cholinergic agonist-treated cells, respectively. Therefore, we first analyzed constitutive internalization of NBCe1-A and NBCe1-B by utilizing inhibitors of constitutive recycling (monensin and W13) that we (57, 58) and others (80) described previously. We found that the 60-min application of monensin or W13 (Fig. 2) “locks” NBCe1-A-YFP and NBCe1-B-GFP in large endosomal formations, thus preventing constitutively internalized cargo from recycling back to the cell surface. Figure 2 shows that some of the endosomal formations containing NBCe1-A (Fig. 2A) and NBCe1-B (Fig. 3B) are marked as early endosomes by EEA1. Our data show that both NBCe1-A and NBCe1-B undergo constitutive endocytosis, which may play an important role in the functional regulation of cotransporters in resting cells. The reader may note that although there is colocalization of EEA1 and fluorescent-tagged protein, not all dots are colocalized after treatment with monensin or other treatments discussed below. This is likely due to the intracellular trafficking process of these fluorescent-tagged proteins, which do not normally stop in early endosomes but continue on to be recycled or degraded. Even in cells treated with monensin where recycling is inhibited, some proteins will be “locked” in recycling endosomes and late endosomes rather than exclusively in early endosomes. Indeed, we (I. I. Grichtchenko, unpublished data) found that some endosomal formations that contain NBCe1-A and NBCe1-B are marked with anti-Rab11 (recycling endosomes) or anti-LAMP1 (late endosomes). This does not affect the conclusions drawn from our observations and analysis. These new observations, together with our previous reports (57, 58), suggest that constitutive near-membrane cycling may be an intrinsic feature of NBCe1 independent of NH2-terminal amino acid differences but dependent on structural determinants that are identical in these transporter proteins. However, future studies are needed to determine whether constitutive endocytosis of A and B variants of NBCe1 is dependent or independent of cell type.

CCh-induced PKC-dependent endocytosis of NBCe1-A and NBCe1-B.

We also examined trafficking of NBCe1-A and NBCe1-B in ParC5 cells stimulated by cholinergic signaling via endogenous muscarinic receptors that activate PKC and intracellular Ca signaling and stimulate salivary secretion. We found that 5–15 min after its application CCh stimulates internalization of NBCe1-A-YFP and NBCe1-B-GFP in ParC5 cells. Figure 4 shows that some of the endosomes that contain NBCe1-A and NBCe1-B are early endosomes marked by EEA1. We also observed that some of the transporter molecules could be present in recycling and late endosomes marked with anti-Rab11 and anti-LAMP1 antibodies in CCh-treated cells (I. I. Grichtchenko, unpublished data). These data, together with our previous observations that ANG II via ANG II-type 1 receptor (AT1R) stimulates endocytosis of NBCe1-A in Xenopus oocytes (56) and CCh via M3R stimulates endocytosis of endogenous NBCe1 in ParC5 cells (58), suggest that cholinergic signaling induces internalization of NH2-terminal variants of NBCe1 and is possibly regulated by that part of the molecule that is common to these variants. Several groups have reported that the cell type could play a role in the regulation of ion transporters, e.g., the sodium/proton exchanger (NHE) and the cystic fibrosis transmembrane conductance regulator (CFTR) (35, 47, 49). Further studies are needed to determine whether agonist-stimulated endocytosis of NBCe1-A and NBCe1-B is cell specific.

To investigate PKC dependence of NBCe1-A and NBCe1-B internalization we first used PMA, which freely diffuses across the cell membrane and directly activates conventional and novel PKC isoforms (46, 5658, 80). We found that PMA stimulates NBCe1-A-YFP and NBCe1-B-GFP internalization and that this can be blocked by GF, which inhibits all classes of PKC isoforms (Fig. 3). Next we found that GF also blocked CCh-induced internalization of two NBCe1 variants, strongly suggesting PKC dependence (Fig. 4). We followed up on this discovery and determined that conventional PKCαβγ (blocked by Gö) and novel PKCδ (blocked by rottlerin) mediate internalization of NBCe1-A and NBCe1-B. To support our rottlerin data, we performed a study using a kinase-dead (Lys376Arg) PKCδ mutant. Figure 5 shows that downregulation of the endogenous PKCδ with the dominant-negative kinase-dead (Lys376Arg) PKCδ mutant prevents CCh-stimulated internalization of NBCe1-A. This result suggests that PKCδ regulates NBCe1 endocytosis in ParC5 cells. It was reported that PKCδ can regulate trafficking of various receptors for degradation (12, 31, 34). Therefore our observation suggests that PKCδ may control trafficking of NBCe1 to lysosomes. Interestingly, it was reported that cholinergic stimulation activates PKCδ and accelerates the BLM trafficking of endocytosed cargo to lysosomes in lacrimal acinar cells (64, 89). Further experiments with lysosomal markers and inhibitors are needed to investigate the role of PKCδ in NBCe1 endocytosis.

Luminal anion secretion depends on basolateral NBCe1 in parotid acinar cells.

To help develop a more complete model of the role of NBCe1 in salivary secretion, we investigated the initial phase of the secretory response in parotid acinar cells. We conducted an Ussing chamber study measuring Isc across ParC10 cell monolayers with or without basolaterally applied tenidap to study the role of basolateral NBCe1 in apical anion secretion. Tenidap, an NBCe1 inhibitor, has also been shown to modulate cytoplasmic pH and thus pH-dependent cellular activities in vivo (42). We hypothesized that NBCe1 supports luminal anion secretion since previous studies showed that the initial transient increase in Isc is carried by HCO3 and Cl fluxes in CCh-treated ParC10 cells (65, 85). Figure 6A shows that tenidap blocks Isc in CCh-stimulated ParC10 cell monolayers. It was reported that the basolateral application of 100 μM DIDS does not alter CCh-induced transient Isc in ParC10 cells (85). Another study indicated that a higher concentration of DIDS (500 μM) is able to inhibit basolateral NBCe1 in acutely dissociated bovine parotid acinar cells (90). Therefore, it is likely that lower DIDS concentration fails to block NBCe1 in ParC10 cells (85). Although future studies using siRNAs targeting NBCe1 are needed to fully clarify the role of NBCe1 in CCh-stimulated Isc in ParC10 cells, our present data may suggest that NBCe1 supports an initial phase of agonist-stimulated acinar cell secretion. Cholinergic stimulation causes an immediate transient increase in the rates of NBCe1-A and NBCe1-B in several secretory tissues (4, 71). A similar mechanism could support a transient increase in the anion efflux underlying the active secretory phase of salivary acinar cell secretion, before NBCe1 undergoes internalization during transition to the sustained secretory phase.

Possible role of NBCs in salivary secretion.

As the model in Fig. 7 shows, cholinergic stimulation that can induce an instant transient increase in the HCO3 influx via NBCe1-B can support a large transient initial phase of acinar cell secretion. Because of its electrogenicity, NBCe1-B may support HCO3 secretion not only by mediating basolateral HCO3 influx, but also by transporting a net negative charge inside the cell to provide the driving force for eventual HCO3 efflux across the luminal membrane (via an anion channel) (38, 59, 86, 90). This is similar to reports that the electrogenicity of NBCe1-B sustains secretion in the pancreas (27, 29, 30, 76). Subsequent endocytosis of NBCe1-B causes a gradual inhibition of cotransporter activity and slows down initial luminal secretion.

Fig. 7.

Physiological role of NBC in salivary parotid acinar cell secretion. The basolateral step of HCO3 transcellular secretion occurs through HCO3 entry via NBCe1-B and electroneutral NBCn1. 1: The electrogenic NBCe1-B transports a net negative charge across the BLM that may help provide the driving force for the initial transient phase of cholinergic-stimulated secretion in salivary acinar cells. 2: Subsequent PKC-dependent internalization of NBCe1-B coincides with the acinar cell's transition between an initial phase of increased secretion and a subsequent steady-state phase of decreased secretion. NBCe1-B internalization may acutely regulate secretion and homeostasis of acinar cells. 3: The electroneutral NBCn1 does not undergo endocytosis and thus likely regulates pHi in cholinergic-stimulated acinar cells. Loss of surface NBCe1-B in the presence of constant activity of NBCn1 may lead to an adjustment to a new steady state characteristic of continuous secretory response in parotid acinar cells.

Figure 7 also illustrates the role of electroneutral NBCn1 (encoded by SLC4A7 gene) in luminal secretion by parotid acinar cells. We showed previously (58) that NBCn1 is stably present at the BLM in ParC5 cells. Others have shown that transport via NBCn1 is important for HCO3 flux and pHi regulation in various secretory epithelia (1, 4, 9, 16, 26, 29, 32, 37, 41, 53, 61, 63, 69, 70, 72). Basolateral NBCn1 may constantly support pHi regulation during all phases of salivary secretion. Indeed, it has been suggested that NBCn1 regulates pHi in the secretory epithelia of the rat distal nephron (52). A loss of electrogenic NBCe1 from the BLM and a stable presence of electroneutral NBCn1 may support partial recovery of acinar cell intracellular ion composition and pHi during the recovery phase of luminal secretion (22, 23, 44, 50, 54).

Possible role of NBCe1-A endocytosis.

We found that NBCe1-A undergoes internalization in response to cholinergic stimulation in ParC5 cells. The ability of NBCe1-A to change its expression at the cell surface is supported by the reported redistribution of NBCe1-A toward the cytosol during acute metabolic alkalosis in the renal PT (11). This is also in agreement with functional studies that have demonstrated decreased Na+-HCO3 cotransporter activity during metabolic alkalosis and is consistent with a role of the renal PT in the regulation of systemic acid-base homeostasis (2, 62, 77). Thus internalization of NBCe1-A could be an important part of the adaptive mechanism to support a decrease in HCO3 reabsorption by the kidneys under pathological conditions during acid-base disturbances. NBCe1-A is also present in several secretory tissues and the brain; therefore its endocytosis may be a means of adaptation during other tissue pathologies.

GRANTS

This work is supported by The Pilot Project Grant in Women's Health Research, University of Colorado Denver (I. I. Grichtchenko) and National Institute of Dental and Craniofacial Research Grants DE-015648 (M. E. Reyland) and K08-DE-017633 (O. J. Baker).

DISCLOSURES

No conflicts of interest are declared by the author(s).

REFERENCES

ACKNOWLEDGMENTS

The authors thank Dr. Alexander Sorkin, Dr. Manuel Miranda, Tatiana Sorkina, and Angela Ohm for constant intellectual inspiration and technical help. The authors are thankful to Dr. William J. Betz for his invaluable intellectual support and Clinton Cave and the late Steven Fadul for their assistance with fluorescent microscopy.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
  65. 65.
  66. 66.
  67. 67.
  68. 68.
  69. 69.
  70. 70.
  71. 71.
  72. 72.
  73. 73.
  74. 74.
  75. 75.
  76. 76.
  77. 77.
  78. 78.
  79. 79.
  80. 80.
  81. 81.
  82. 82.
  83. 83.
  84. 84.
  85. 85.
  86. 86.
  87. 87.
  88. 88.
  89. 89.
  90. 90.
  91. 91.
  92. 92.
View Abstract