Electrogenic NBCe1 (SLC4A4), but not electroneutral NBCn1 (SLC4A7),cotransporter undergoes cholinergic-stimulated endocytosis in salivaryParC5 cells
Clint Perry,1 David O. Quissell,2 Mary E. Reyland,2 and Irina I. Grichtchenko1
1Department of Physiology and Biophysics, and 2Department of Craniofacial Biology, University of Colorado Denver,Anschutz Medical Campus, Aurora, Colorado
Submitted 14 March 2008; accepted in final form 22 September 2008
Perry C, Quissell DO, Reyland ME, Grichtchenko II. ElectrogenicNBCe1 (SLC4A4), but not electroneutral NBCn1 (SLC4A7), cotrans-porter undergoes cholinergic-stimulated endocytosis in salivary ParC5cells. Am J Physiol Cell Physiol 295: C1385–C1398, 2008. First pub-lished September 24, 2008; doi:10.1152/ajpcell.00153.2008.—Cholin-ergic agonists are major stimuli for fluid secretion in parotid acinarcells. Saliva bicarbonate is essential for maintaining oral health.Electrogenic and electroneutral Na�-HCO3
� cotransporters (NBCe1and NBCn1) are abundant in parotid glands. We previously reportedthat angiotensin regulates NBCe1 by endocytosis in Xenopus oocytes.Here, we studied cholinergic regulation of NBCe1 and NBCn1 mem-brane trafficking by confocal fluorescent microscopy and surfacebiotinylation in parotid epithelial cells. NBCe1 and NBCn1 colocal-ized with E-cadherin monoclonal antibody at the basolateral mem-brane (BLM) in polarized ParC5 cells. Inhibition of constitutiverecycling with the carboxylic ionophore monensin or the calmodulinantagonist W-13 caused NBCe1 to accumulate in early endosomeswith a parallel loss from the BLM, suggesting that NBCe1 is consti-tutively endocytosed. Carbachol and PMA likewise caused redistri-bution of NBCe1 from BLM to early endosomes. The PKC inhibitor,GF-109203X, blocked this redistribution, indicating a role for PKC. Incontrast, BLM NBCn1 was not downregulated in parotid acinar cellstreated with constitutive recycling inhibitors, cholinergic stimulators,or PMA. We likewise demonstrate striking differences in regulation ofmembrane trafficking of NBCe1 vs. NBCn1 in resting and stimulatedcells. We speculate that endocytosis of NBCe1, which coincides withthe transition to a steady-state phase of stimulated fluid secretion,could be a part of acinar cell adjustment to a continuous secretoryresponse. Stable association of NBCn1 at the membrane may facilitateconstitutive uptake of HCO3
� across the BLM, thus supporting HCO3�
luminal secretion and/or maintaining acid-base homeostasis in stim-ulated cells.
EEA1; protein kinase C; muscarinic type-3 receptor; PDZ; clathrin
SALIVA IS ESSENTIAL FOR MAINTAINING oral health and function(57). Impairment of isotonic fluid secretion in parotid acinarcells causes dry mouth syndrome in millions of patients (41).The parasympathetic nervous system is responsible for theproduction of salivary gland fluid and electrolyte secretionfollowing stimulation by the neurotransmitter ACh (6). TheACh signaling via the muscarinic type-3 receptors (M3R) isprimarily responsible for the initiation of fluid and ion secre-tion in parotid acinar cells (11, 62).
Normal bicarbonate secretion makes it possible to neutralizeharmful acid produced by oral bacteria, thus preventing oral
infections. The Na�-HCO3� cotransporters (NBC) encoded by
SLC4 gene family are recognized now as important mecha-nisms for HCO3
� flux in many secretory epithelia (1, 4, 9, 12,21, 24–26, 29, 33, 42, 46, 47, 52, 53, 55).
Moreover, mutations and deletions of NBC genes are majorculprits of poor dentition in humans and mice (16, 19, 23).NBC was functionally detected in bovine and sheep parotidacinar cells before cloning of the SLC4 gene family (45, 50,59). These reports were confirmed by a more recent report thatelectrogenic NBCe1 (SLC4A4) is present and functional inbovine parotid acinar cells (66). Additionally, several groupsreported abundant expression of NBCe1 and NBCn1 in parotidacinar cells of different species (53). However, nothing isknown about regulation of NBC by cholinergic secretorystimuli in parotid acinar cells. In fact, so little attention hasbeen given to potential roles that NBC could play in the fluidsecretion process that current models of fluid secretion pro-posed for parotid acinar cell do not include NBC (20, 39, 41).It is critical to clarify the roles of NBC in parotid acinar cellsfor a thorough understanding of the molecular physiology ofthe fluid secretion process.
We (43, 44) reported earlier that signaling via the angioten-sin receptor (AT1R) stimulates endocytosis and inhibits func-tional activity of NBCe1 in Xenopus oocytes. Here, we ex-panded our early observations to investigate regulation of thesurface expression of NBCe1 and NBCn1 in untreated andcholinergically stimulated parotid acinar cells. For these stud-ies, we have used the ParC5 cell line, developed by Dr. DavidO. Quissell, which is functionally similar to native parotidacinar cells (48, 49, 63). Using confocal fluorescent micros-copy in fixed ParC5 cells, NBCe1 and NBCn1 antibodies andmarkers for polarized epithelial cells and endocytosis, andsurface biotinylation, we found the following. 1) The basolat-eral membrane (BLM) of rat ParC5 cells expresses NBCe1 andNBCn1, which would support HCO3
� ions influx into the cellvia BLM. In addition, a net negative charge enters the cell viaelectrogenic BLM NBCe1. 2) NBCe1, but not NBCn1, under-goes constitutive endocytosis. 3) Cholinergic stimulation in-duces PKC-dependent endocytosis of NBCe1. Removal ofelectrogenic NBCe1 from the BLM decreases HCO3
�, andnegative charge entry may support recovery of the acinar cellfrom massive loss of salt and water (63). 4) In contrast,cholinergic stimulation did not affect NBCn1, which maycontinuously transport HCO3
� inside to possibly neutralize
Address for reprint requests and other correspondence: I. I. Grichtchenko,Univ. of Colorado and Denver Health Sciences Center, Dept. of Physiologyand Biophysics, Mail Stop 8307, P.O. Box 6511, Aurora, CO 80045 (e-mail:[email protected]).
The costs of publication of this article were defrayed in part by the paymentof page charges. The article must therefore be hereby marked “advertisement”in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Am J Physiol Cell Physiol 295: C1385–C1398, 2008.First published September 24, 2008; doi:10.1152/ajpcell.00153.2008.
intracellular acidification in secreting acini. Our data suggestthat NBCe1 and NBCn1 are important molecular mechanismsfor the fluid secretion process in parotid acinar cells.
EXPERIMENTAL PROCEDURES
Materials. GF-109203X (GF), W-13, carbachol (CCh), and mo-nensin were purchased from Sigma (St. Louis, MO). Texas red-conjugated transferrin (Tfn-TR), zona occludens 1 monoclonal anti-body (ZO-1), and E-cadherin monoclonal antibody (E-Cad) were fromInvitrogen (Carlsbad, CA). The EEA1 monoclonal antibody was fromBD Transduction Laboratories (San Jose, CA). PMA was purchasedfrom Calbiochem (La Jolla, CA). KIA, a rabbit polyclonal antibodyagainst electrogenic NBCe1 (SLC4A4), was a kind gift from Dr.Walter F. Boron, Yale University School of Medicine, New Haven,CT. The COOH terminus of NBCe1 variants are identical, thus theantibody does not distinguish among renal NBCe1-A and pancreasNBCe1-B (7). NBCn1-106 and ntNBCn12977, two rabbit polyclonalantibodies raised against the COOH terminus of the rat sequence ofelectroneutral NBCn1, were kind gifts from Dr. Søren Nielson,University of Aarhus, Denmark (64). Another NBCn1 affinity purifiedrabbit polyclonal antibody raised against the COOH terminus of therat sequence of electroneutral NBCn1 was a kind gift from Dr.Inyeong Choi, Emory University (49a). NBCn1 antibodies do notdistinguish among NBCn1B, NBCn1C, NBCn1D, and NBCn1E be-cause the COOH terminus of all four rat NBCn1 variants are identical(21, 30, 64). Secondary antibodies Alexa 488 and Alexa Fluor-Texasred were from Invitrogen. Vectashield mounting media was purchasedfrom Vector Laboratories (Burlingame, CA).
ParC5 cell culture. We used a well-established rat parotid acinarcell line, ParC5 (passages 10–40; Refs. 48, 49, 63). Cells were culturedat 37°C in humidified atmosphere of 5% CO2 and 35% O2. Cellswere split by dry trypsinization with 0.05% trypsin (Gibco, Carls-bad, CA) and grown to confluency before experiments were per-formed. ParC5 cells were grown in DMEM/F-12 media with 15mM HEPES and L-glutamine (Invitrogen) with the followingadded: 5 �g/ml insulin (Invitrogen), 10�7 M retinoic acid (Sigma),5 mM glutamine (Invitrogen), 0.4 �g/ml hydrocortisone (Sigma),5 �g/ml transferrin (Invitrogen), 10% standard FBS (HyClone),1% nonessential amino acids (Invitrogen), and 50 �g/ml gentami-cin (Invitrogen). Experiments were performed by first exchangingmedia with serum-free media before drug treatments. Serum-freemedia consisted of DMEM/F-12 media with 15 mM HEPES andL-glutamine (Invitrogen).
Drug treatments. Monensin was made as a 10,000� stock solutionin MeOH. The final working concentration of MeOH was 0.01%.CCh, atropine, and W-13 were made as 1,000� stock solution indistilled water (dH2O). PMA was made as a 1,000� stock solutions inDMSO. The final working concentration of DMSO was 0.1%. Alldrugs were diluted in ParC5 culture media to final concentrationsbefore use. Where indicated, drugs were diluted in serum-free media.As controls, we used 0.1% DMSO and 0.01% MeOH.
Immunofluorescence and confocal microscopy. Cells were culturedon 12-mm glass coverslips in 24-well plates or, where indicated,plated on Transwell filters (Corning, Corning, NY). After 3 days(filters) or 4 days (coverslips) in culture, cells were incubated withmonensin, MeOH, W-13, PMA, CCh, ACh, and/or, in some experi-ments, with GF or atropine, as indicated. Cells were serum-starved for15 min before treatment. Following treatment, cells were washed withPBS and fixed with 2% paraformaldehyde for 15 min at roomtemperature. For immunofluorescence staining, cells were permeabil-ized with 0.1% Triton X-100 in PBS for 10 min and incubated inblocking solution (0.05% Triton X-100, 1% bovine serum albumin,and 10% FBS in PBS). After blocking was completed, cells wereincubated with anti-NBCe1 (1:500), anti-NBCn1 (1:500), anti-E-Cad(1:1,000), anti-ZO-1 (1:250), or anti-EEA1 (1:500) primary antibod-ies, followed by a series of washes (1% bovine serum albumin and
10% FBS in PBS), and then the respective Alexa Fluor-labeledsecondary antibodies for 1 h. Subsequently, cells and coverslips werewashed in PBS and then dH2O and mounted with Vectashield onslides for qualitative analysis through confocal microscopy.
Fluorescent images were taken using a Spinning Disk confocalmicroscope (Olympus) controlled by SlideBook 4.2.0.10 (Light Mi-croscopy Facility at University of Colorado Denver; http://www.uchsc.edu/lightmicroscopy/). Confocal images were captured in z-stackintervals of 0.5 �m using a �60 oil immersion objective (1.45numerical aperture). Under mercury illumination, the filter sets were:FITC, excitation 460–480 nm, emission 495–540 nm; tetramethyl-rhodamine isothiocyanate (TRITC), excitation 535–555 nm, emission570–625 nm. Qualitative and quantitative analysis was done toascertain whether colocalization occurred between Alexa 488 andAlexa Fluor-Texas red or Cy3.
Surface biotinylation. ParC5 cells were grown in 35-mm dishes andused at 100% confluency. Cells were washed twice with cold PBScontaining 0.1 mM CaCl2 and 1 mM MgCl2 and incubated for 20 minon ice with 1 mg/ml sulfosuccinimidyl 2-(biotinamido)-ethyl-1,3-dithiopropionate (EZ-Link Sulfo-NHS-SS-Biotin; Pierce, Rockford,IL) in PBS, followed by a second incubation with fresh NHS-SS-Biotin. After biotinylation, the cells were washed twice with coldPBS, incubated for 30 min on ice with 0.1 M glycine in PBS, andwashed with PBS. The cells were then solubilized by scraping with arubber policeman in lysis buffer [50 mM NaCl, 2 mM MgCl2, 10 mMTris �HCl (pH 6.8), 10% glycerol, 1 mM EGTA, 1 mM EDTA, 10 mMsodium fluoride, 1 mM phenylmethylsulfonyl fluoride, 10 �g/mlleupeptin, 544 �M iodoacetamide, 10 �g/ml aprotinin, 1% TritonX-100, 1% sodium deoxycholate] and by further incubating for 30min at 4°C. Cell lysates were precleared by centrifugation at 16,000 g;the biotinylated proteins were precipitated with NeutrAvidin beads(Pierce), washed five times with lysis buffer, and denatured by heatingthe beads in sample buffer at 95°C for 5 min. NeutrAvidin beads wereresolved on 7.5% SDS-PAGE, and the proteins were transferred to themembrane. Western blotting was performed with polyclonal antibod-ies to NBCe1 or NBCn1 followed by species-specific secondaryantibodies conjugated with horseradish peroxidase. The enhancedchemiluminescence kit was purchased from Pierce. NBCe1 or NBCn1proteins were detected using Kodak Image Station 440CF, and theirintensities were measured using ImageJ software (http://rsb.info.nih.gov/ij/). For negative controls, we incubated cells in the absence ofEZ-Link Sulfo-NHS-SS-Biotin and precipitated cell lysates in theabsence of NeutrAvidin beads.
Two-electrode oocyte voltage clamp. All studies were performed inaccordance with guidelines from and submitted to and approved bythe University of Colorado Denver Animal Care and Use Committee.Female Xenopus laevis frogs (NASCO) were anesthetized with 1.5mg/ml tricaine. The ovarian lobes were surgically removed, dissected,and treated with 2 mg/ml collagenase type IA, and oocytes wereincubated as described previously (43, 44). The cDNAs encodinghuman NBCe1 and M3R receptors were each subcloned into thepGH19 expression vector. DNAs were transcribed in vitro using themMessage Machine kit (Ambion, Austin, TX) to generate synthesizedcapped mRNAs. Oocytes were injected with 50 nl of 0.5 ng/nl NBCe1mRNA; 25 nl of 1 ng/nl NBCe1 mRNA plus 25 nl of 1 ng/nl M3RmRNA; or 50 nl of dH2O. Three days later, oocytes were subjected toelectrophysiological experiments as described before (43, 44). Briefly,oocytes were voltage-clamped at room temperature using a two-electrode oocyte clamp (Warner Instruments, New Haven, CT) andmicroelectrodes made by pulling borosilicate glass capillary tubing(Warner Instruments) on a microelectrode puller. The oocytes wereimpaled with microelectrodes filled with 3 M KCl (resistance �0.3–1.0 M�). We clamped oocytes to �50 mV holding potential (Vh)and measured NBC current (INBC) in response to a 60-s depolarizationfrom �50 to 0 mV. The currents were filtered at 20 Hz (4-pole Besselfilter) and digitized. An oocyte was placed in a chamber with a 4ml/min constant superfusion using bath solutions described previously
C1386 ENDOCYTOSIS OF NBCe1 BUT NOT NBCn1 IN PAROTID CELLS
AJP-Cell Physiol • VOL 295 • NOVEMBER 2008 • www.ajpcell.org
(43, 44). Bath solutions were delivered with syringe pumps (HarvardApparatus, South Natick, MA), and solutions were switched withpneumatically operated valves (Clippard Instrument Laboratory, Cin-cinnati, OH).
Data analysis. Collapsed z-stacks were done by creating a projec-tion image in ParC5 cells using SlideBook 4.2.0.10. At least threeindependent experiments with consistent results were done for eachexperimental condition. Each experiment contained three separatecoverslips of cells from which we took at least five images fromdifferent areas of each coverslip. Representative images of theseexperiments are shown in the figures. The Pearson R correlationcoefficients were calculated to quantify endocytosis of NBCe1 mea-sured as increased colocalization of internalized NBCe1 with earlyendosomes marked with EEA1. Using SlideBook 4.2.0.10 CrossChannel software (Olympus Spindisk fluorescent microscope), weacquired the Pearson R correlation coefficients that match the intensityof the green fluorescence FITC filter channel (NBCe1 or NBCn1) withthe red TRITC filter channel (EEA1). Pearson R coefficient 0.0signifies no correlation trend, 1.0 means complete co-correlation, and�1.0 implies anticorrelation. We normalized intensity of biotinylatedprotein bands in treated cells to that in untreated control cells. InXenopus oocyte experiments, we normalized amplitude of the “test” to“control” voltage-clamped NBCe1 current. All averages are reportedas means � SD. For ratios, averages are presented as log-normalmeans. The statistical significance data were determined using anunpaired Student’s t-test (28). Differences were considered significantat a level of P � 0.05.
RESULTS
Localization of endogenous electrogenic NBCe1 and elec-troneutral NBCn1 to the BLM of polarized ParC5 cells. Here,we investigated specifically where NBCe1 and NBCn1 co-transporters are localized in polarized ParC5 cells. Cells weregrown on Transwell filters for 3–4 days to achieve polarization(see EXPERIMENTAL PROCEDURES) and costained with antibodiesagainst NBCe1 and the BLM marker E-Cad or the apicalmarker ZO-1. Note that only truly polarized cells form tightjunctions between the individual cells and display clear ZO-1staining (15). In Fig. 1A-1, the fluorescent staining for endog-enous electrogenic NBCe1 (green) colocalizes almost entirelywith E-Cad (red) staining both in apical view (AV) images andlateral view (LV) images. On the other hand, staining for ZO-1(red) does not colocalize with, and is visibly separate from, thestaining for NBCe1 (Fig. 1A-2). Images are collapsed z-stacksof the entire cell through the z-axis (AV) and collapsed z-stacksof the region of interest (white line) through the y-axis (LV).Therefore, within both of these images, one can see that thereis clearly no apical presence of NBCe1 staining in ParC5 cells.This is most notable in the LV images showing strong colo-calization of NBCe1 and E-Cad and total separation of NBCe1and ZO-1.
In parallel experiments, we also stained cells with antibodiesagainst endogenous electroneutral NBCn1 along with antibod-ies against the BLM marker E-Cad and the apical membranemarker ZO-1 (Fig. 1B; here, we used NBCn1-106 antibody).Similar to the results found with NBCe1, we found that NBCn1is also localized at the BLM of ParC5 cells. This is evidencedby colocalization of NBCn1 staining with E-Cad staining anda clear separation of NBCn1 and ZO-1 staining. We also foundthe same staining pattern by using two other NBCn1 antibodies(see EXPERIMENTAL PROCEDURES; data not shown). In our immu-nolabeling control experiments, in which we preabsorbed eachantibody with its immunizing peptide, no NBC staining was
detected (data not shown). Again, a lack of apical staining forNBCn1 suggests there is no NBCn1 present at the apicalsurface of ParC5 cells. These data illustrate that rat parotidacinar ParC5 cells express electrogenic NBCe1 and electro-neutral NBCn1, both of which are localized at the BLM ofthese cells.
Recycling inhibitors significantly increased accumulation ofNBCe1 but not NBCn1 in early endosomes. We (44) previouslyreported that recombinant NBCe1 undergoes constitutive en-docytosis in Xenopus oocytes. Here, we noticed that althoughmost endogenous NBCe1 and NBCn1 are present at the BLM,there can also be found NBCe1 and, to a lesser extent, NBCn1staining within the cytoplasm of the untreated ParC5 cells(Figs. 1 and 2). This suggested that NBCe1 and possiblyNBCn1 may go through some degree of constitutive endocy-tosis in ParC5 cells. To test this idea, we used two recyclinginhibitors: the carboxylic ionophore monensin and the calcium-binding protein calmodulin (CaM) inhibitor W-13, which we(44) previously used to study recombinant NBCe1. It is knownthat monovalent carboxylic ionophore monensin raises pHwithin endosomes and inhibits recycling of internalized pro-teins to the plasma membrane (5, 58, 65). It is also known thatCaM controls the recycling step of the endocytic pathway ofseveral membrane proteins, and its antagonist, W-13, blocksthe exit of internalized cargo from early endosomes resulting inthe formation of large endosomes (3, 13, 22, 32, 61). Othershave shown that CaM is involved in the regulation of func-tional activity of endogenous NBCe1, whereas we reported thatCaM is involved in trafficking of recombinant NBCe1 inXenopus oocytes (4, 17, 44, 54).
To study the effect of monensin on NBCe1 distribution, wetreated ParC5 cells with 50 �M monensin added to serum-freeculture media for 60 min and kept in a humidified atmosphereof 5% CO2 and 35% oxygen at 37°C (Fig. 2A-1, 2nd row fromthe top). In a parallel control experiment, we applied thevehicle MeOH at 0.01% for 60 min at 37°C (Fig. 2A-1, 3rd rowfrom the top). Subsequently, we fixed cells and stained withantibodies against NBCe1 and the early endosomal markerEEA1. We compared confocal microscopy images of cellsuntreated and monensin-treated for differences both in NBCe1staining and in colocalization of NBCe1 with EEA1. As seen inFig. 1A, in untreated ParC5 cells, the staining of NBCe1 isalmost entirely colocalized with E-Cad and therefore at theBLM. In Fig. 2A-1 (top row), in addition to strong BLMstaining of NBCe1, we also detected NBCe1 staining thatcolocalized with EEA1-stained early endosome in an untreatedcell (Fig. 2A-1, arrow). In the monensin-treated cells, we founda marked decrease of NBCe1 at the BLM, significant accumu-lation of NBCe1 in intracellular endosomal-like formations,and increased colocalization with EEA1-stained early endo-somes (Fig. 2A-1, 2nd row from the top). This was in remark-able contrast to untreated (Fig. 2A-1, top row) or vehicle-treated (Fig. 2A-1, 3rd row from the top) cells. Figure 2A-1 (seebottom row) shows a typical experiment in which we applied100 �M W-13 for 60 min to serum-starved ParC5 cells at37°C. We found that W-13 caused a significant loss of NBCe1from the BLM, a parallel accumulation in endosomal-likeformations, and increased colocalization of cotransporters withEEA1-stained early endosomes (Fig. 2A-1, insets). To quantifyour results, we used SlideBook 4.2.0.10 Cross Channel soft-ware (Olympus fluorescent Spindisk microscope) to acquire
C1387ENDOCYTOSIS OF NBCe1 BUT NOT NBCn1 IN PAROTID CELLS
AJP-Cell Physiol • VOL 295 • NOVEMBER 2008 • www.ajpcell.org
C1388 ENDOCYTOSIS OF NBCe1 BUT NOT NBCn1 IN PAROTID CELLS
AJP-Cell Physiol • VOL 295 • NOVEMBER 2008 • www.ajpcell.org
the Pearson R correlation analysis (see EXPERIMENTAL PROCE-DURES). As shown in a bar graph in Fig. 2A-2, the Pearson Rcorrelation coefficients that match colocalization of NBCe1with EEA1 were low in untreated (0.16 � 0.06, n � 124) andvehicle-treated cells (0.15 � 0.05, n � 80). The colocalizationof NBCe1 with EEA1 were 3-fold increased in cells treatedwith monensin (0.32 � 0.06, n � 60) and W-13 (0.47 � 0.06,n � 84). These results suggest that endogenous electrogenicNBCe1 undergoes constitutive endocytosis in ParC5 cells.With application of these recycling inhibitors, internalizedNBCe1 is, possibly, unable to return to the surface and isessentially trapped within endosomes. As the recycling inhib-itor W-13 is also a CaM antagonist, these findings may proveuseful in future experiments to study the role of Ca2� and CaMin NBCe1 trafficking and regulation.
The effect of the recycling inhibitor monensin on the elec-troneutral NBCn1 is shown in Fig. 2B. This figure shows thatin contrast to NBCe1, NBCn1 does not undergo constitutiveendocytosis. We found that after application of 50 �M monen-sin for 1 h, there was no detectable loss of NBCn1 from theBLM (Fig. 2B-1, bottom row). The bar graph in Fig. 2B-2shows that the Pearson R correlation coefficient of the colo-calization of NBCn1 with EEA1 is low both in untreated cells(0.13 � 0.02, n � 61) and monensin-treated cells (0.104 �0.044, n � 68). This suggests that NBCn1 does not undergoconstitutive endocytosis or does so at a very slow rate. Toinvestigate this latter possibility, we applied monensin over-night and found little to no change in staining from untreatedcells (data not shown). Therefore, we conclude that either theconstitutive endocytosis of NBCn1 is extremely slow or simplydoes not occur.
Surface biotinylation quantitative analysis showed that re-cycling inhibitors cause removal of NBCe1 but not NBCn1from the BLM. To extend support of our confocal microscopydata, we asked whether recycling inhibitors cause removal ofthe electrogenic NBCe1 but not the electroneutral NBCn1 fromthe BLM. For this, we performed surface biotinylation studiesusing a membrane-impermeable derivative of biotin (see EX-PERIMENTAL PROCEDURES). First, we studied surface expressionof the electrogenic NBCe1. Western blot analysis with anti-NBCe1 polyclonal antibodies (KIA; 1:500) detected biotinyl-ated NBCe1 protein as an 130-kDa band on SDS-PAGE(Fig. 5A-1). We found that recycling inhibitors caused a sig-nificant decrease of the intensity of the surface-biotinylatedNBCe1 protein. Thus the intensity of the NBCe1 band fromcells treated with 50 �M monensin as well as with 100 �MW-13 was 23.8% � 19.6% (n � 4) and 33.3% � 19.5% (n �4), respectively, of that of untreated cells (Fig. 5A-2). The
intensity of the NBCe1 band in vehicle-treated (0.01% MeOH)cells was 88.9% � 11.1% (n � 4) of that of untreated cells.This suggests that the possible inhibition of the constitutiverecycling of NBCe1 induces removal of the cotransporter fromthe BLM of ParC5 cells. Next, we examined the electroneutralNBCn1. Western blot analysis with anti-NBCn1 affinity puri-fied polyclonal antibodies (NBCn1-Ct; 1:500) detected sur-face-biotinylated NBCn1 protein as an 127-kDa band onSDS-PAGE (Fig. 5B-1). In striking contrast to NBCe1, wefound that recycling inhibitor monensin did not alter the levelof the surface expression of NBCn1 in ParC5 cells. Thesurface-biotinylated NBCn1 band from cells treated for 60 minwith 50 �M monensin was 99.1% � 26.9% (n � 6) of that ofuntreated cells, indicating that this recycling inhibitor does notinduce removal of NBCn1 from the BLM of ParC5 cells.Negative controls showed a complete absence of biotinylatedNBCe1 or NBCn1 (data not shown). Taken together, ourconfocal fluorescent microscopy findings and surface biotiny-lation data suggest that the electrogenic NBCe1 undergoesconstitutive endocytosis, whereas the electroneutral NBCn1 isstably present at the BLM of ParC5 cells.
PMA stimulates endocytosis of NBCe1 but not NBCn1.Given the differences in constitutive endocytosis of NBCe1and NBCn1, we hypothesized that there may also be differ-ences in stimulated endocytosis of these cotransporters. Forthis, we utilized PMA, which is commonly used to stimulateendocytosis of membrane proteins in mammalian cells (34, 40,56), in an attempt to induce endocytosis of NBCe1 and NBCn1in ParC5 cells. Results for endogenous electrogenic NBCe1 aresummarized in Fig. 3A-1 (middle row) and A-2 (2nd bar). Inthis experiment, ParC5 cells were treated with 1 �M PMAadded to serum-free culture media for 15 min and kept in ahumidified atmosphere of 5% CO2 and 35% oxygen at 37°C.Subsequently, we fixed cells and stained with antibodiesagainst NBCe1 and EEA1. We compared confocal microscopyimages of untreated cells and PMA-treated cells (Fig. 3A-1, topand middle rows) for differences in staining of NBCe1 and inits colocalization with EEA1. In PMA-treated cells, we founda marked loss of the BLM staining of NBCe1 and its redistri-bution into endosome-like formations, including EEA1-stainedearly endosomes within the cytoplasm of the cell (Fig. 3A-1,middle row). To quantify our results, we performed a PearsonR correlation analysis (see EXPERIMENTAL PROCEDURES). Asshown in the bar graph in Fig. 3A-2, the Pearson R correlationcoefficient of colocalization of electrogenic NBCe1 with EEA1was low (0.15 � 0.05, n � 499 cells) in control untreated cells(see “UNTR” bar). PMA increased by 2.5-fold the colocal-ization of NBCe1 with EEA1 with a Pearson R correlation
Fig. 1. Endogenous electrogenic NBCe1 and electroneutral NBCn1 cotransporters are expressed at the basolateral membrane (BLM) of ParC5 cells.A: endogenous electrogenic NBCe1. A-1: apical view (AV) and lateral view (LV) images show cells that were costained with antibodies against NBCe1 (green)and the BLM marker E-cadherin (E-Cad; red). Following primary antibody staining, cells in A and B were subsequently treated with corresponding secondaryantibodies Alexa 488 and Alexa Fluor-Texas red. Merged (yellow) represents overlap of NBCe1 and E-Cad, confirming BLM localization of NBCe1. A-2: AVand LV images show cells that were costained with antibodies against NBCe1 (green) and the apical membrane marker zona occludens 1 (tight junction proteinZO-1; red). The strong separation of green and red staining (lack of merged yellow) indicates that NBCe1 does not colocalize with ZO-1 and is not present atthe apical membrane. B: endogenous electroneutral NBCn1. B-1: AV and LV images show cells that were costained with antibodies against NBCn1 (green) andE-Cad (red). Merged (yellow) represents overlap of NBCn1 and E-Cad, confirming BLM localization of NBCn1. B-2: AV and LV images show cells that werecostained with antibodies against NBCn1 (green) and ZO-1 (red). The strong separation of green and red staining (lack of merged yellow) indicates that NBCn1(green) does not colocalize with ZO-1 (red) and is not present at the apical membrane. ParC5 cells were grown on Transwell filters for 3 days. Cells were fixed,filters were removed from inserts and mounted on slides, and a z-stack of optical sections was acquired through the FITC (green) and tetramethylrhodamineisothiocyanate (TRITC; red) filter channels. Collapsed z-stacks of the entire cell through the z-axis (AV) and collapsed z-stacks of the region of interest (whiteline) through the y-axis (LV) are shown. Images are representative of at least 15 images from each of 3 independent experiments. Scale bars, 5 �m.
C1389ENDOCYTOSIS OF NBCe1 BUT NOT NBCn1 IN PAROTID CELLS
AJP-Cell Physiol • VOL 295 • NOVEMBER 2008 • www.ajpcell.org
coefficient of 0.38 � 0.07 (n � 170 cells) (see “PMA” bar inFig. 3A-2). These results indicate that PKC activator PMAinduces significant endocytosis of NBCe1.
Next, we investigated the effect of PMA on endogenouselectroneutral NBCn1 (see Fig. 3B-1). ParC5 cells were treatedwith PMA and stained with antibodies against NBCn1 andEEA1. In PMA-treated cells, we found that NBCn1 (Fig. 3B-1,middle row) in striking contrast to NBCe1 (Fig. 3A-1, middle
row) remains at the BLM, does not redistribute within thecytoplasm, and does not colocalize with EEA1-stained earlyendosomes. As summarized in the bar graph in Fig. 3B-2, thePearson R correlation coefficient of colocalization of NBCn1with EEA1 in the PMA-treated cells (0.15 � 0.04, n � 100)was similar to that in the control untreated cells (0.17 � 0.03,n � 270). These results suggest that PMA does not stimulateendocytosis of NBCn1.
C1390 ENDOCYTOSIS OF NBCe1 BUT NOT NBCn1 IN PAROTID CELLS
AJP-Cell Physiol • VOL 295 • NOVEMBER 2008 • www.ajpcell.org
CCh stimulates endocytosis of NBCe1 but not NBCn1. Cho-linergic stimulation is a major stimulus for isotonic salt andfluid secretion in rat parotid salivary cells (18). Here, we usedmuscarinic receptor (MR) agonists ACh and CCh to examinewhether cholinergic stimulation regulates trafficking of endog-enous NBCe1 and NBCn1. Results for endogenous electro-genic NBCe1 are shown in Fig. 3A-1 (bottom row). This figureshows a typical experiment in which we applied 50 �M CChfor 15 min to live cells and then fixed and stained cells withantibodies against NBCe1 and EEA1. Similar to PMA, wefound that in CCh-treated cells, the staining at the BLM wasreduced, and colocalization of NBCe1 and EEA1 increased.Again, in response to CCh, NBCe1 redistributed to endosomal-like formations within the cell. Similar robust endocytosis ofNBCe1 was detected in ParC5 cells treated with neurotrans-mitter ACh (data not shown). As shown in the bar graph in Fig.3A-2, cholinergic stimulation increases by 2.6-fold the Pear-son R correlation coefficient of colocalization of electrogenicNBCe1 with EEA1 from 0.15 � 0.05 (n � 499 cells) in controluntreated cells to 0.40 � 0.08 (n � 167) in CCh-treated cells.These data suggest that in 15 min following application ofcholinergic agonists, NBCe1 undergoes a significant stimu-lated endocytosis in ParC5 cells.
In striking contrast to NBCe1, we found that CCh does notstimulate endocytosis of endogenous electroneutral NBCn1 inParC5 cells (see Fig. 3B-1). Briefly, we applied 50 �M CCh for15 min to live cells and then fixed and stained cells withantibodies against NBCn1 and EEA1. We found that NBCn1staining remains strong at the BLM, its accumulation in endo-somal formations was absent, and there is no increase in thecolocalization of NBCn1 and EEA1 in CCh-treated cells (Fig.3B-1, bottom row) compared with untreated cells (Fig. 3B-1,top row). The Pearson R correlation coefficient of colocaliza-tion of NBCn1 with EEA1 in the CCh-treated cells (0.18 �0.04, n � 170) was similar to that in the control untreated cells(0.17 � 0.03, n � 270) (see bar graph in Fig. 3B-2).
Surface biotinylation quantitative analysis showed that PMAand CCh remove NBCe1 but not NBCn1 from the BLM. Bysurface biotinylation, we demonstrated that PMA and CChremoved the electrogenic NBCe1 but not the electroneutral
NBCn1 from the BLM of ParC5 cells. We found a significantdecrease in the intensity of the surface-biotinylated NBCe1band from cells treated with PMA or CCh (Fig. 5A-3). Assummarized in the bar graph in Fig. 5A-4, intensity of theNBCe1 band from cells treated with 1 �M PMA or 50 �MCCh was 36.2% � 21.1% (n � 4) and 50.2% � 13.1% (n �4), respectively, of that of untreated cells. In striking contrast toNBCe1, we found that neither PMA nor CCh altered the levelof the surface expression of NBCn1 in ParC5 cells. Intensitiesof the surface-biotinylated NBCn1 band from cells treated with1 �M PMA or 50 �M CCh were 110.2% � 8.9% (n � 6) and101.2% � 17.4% (n � 6) of that of untreated cells (see bargraph in Fig. 5B-2). Negative controls showed a completeabsence of biotinylated NBCe1 or NBCn1 (data not shown).Taken together, our confocal fluorescent microscopy findingsand surface biotinylation data suggest that PMA and CChinduce endocytosis of electrogenic NBCe1 but not electroneu-tral NBCn1 in ParC5 cells.
CCh-stimulated endocytosis of NBCe1 occurs through theMR. It was reported that cholinergic muscarinic receptors areexpressed in the BLM of ParC5 cells (8, 63). To ensure that theobserved above effects of CCh are mediated via muscarinicreceptors, we used the cholinergic receptor inhibitor atropine.Figure 4B (bottom row) shows staining of NBCe1 and EEA1 inParC5 cells treated with a mixture of 1 �M atropine and 50 �MCCh. Atropine completely prevents loss of NBCe1 from theBLM (Fig. 4B, bottom row) in contrast to that found in cellstreated with CCh alone (bottom rows in Figs. 3A-1 and 4A).We also observed a noticeable decrease of colocalization ofNBCe1 and EEA1 staining. Thus the Pearson R correlationcoefficient dropped from its value of 0.40 � 0.08 (n � 167)in the CCh-treated cells to its value of 0.14 � 0.06 (n � 143cells) in the cells treated with a mixture of atropine withCCh. Note that its value was 0.15 � 0.05 (n � 499 cells) inthe control untreated cells (see bar graph in Fig. 3A-2).Additionally, our surface biotinylation quantitative analysisshowed that atropine prevents CCh-induced removal ofNBCe1 from the BLM. As summarized in the bar graph inFig. 5A-4, a normalized intensity of the biotinylated NBCe1band in the cells treated with a mixture of atropine with CCh
Fig. 2. Recycling inhibitors monensin and W-13 increase localization of electrogenic NBCe1 but not electroneutral NBCn1 in early endosomes. A-1: endogen-ous electrogenic NBCe1. Top row: control cells were untreated. Left panel shows strong BLM expression of electrogenic NBCe1 (green). Middle panel shows earlyendosomes stained with EEA1 (red). Right panel is a merged image of both staining with relatively infrequent presence (arrow) of intracellular NBCe1 in theearly endosomes. Second row: cells were treated with monensin. Left panel shows loss of NBCe1 (green) from BLM and its redistribution into intracellularendosomal-like compartments. Merged (yellow) in right panel represents costaining of NBCe1 and EEA1, indicating that monensin “locks” some of NBCe1cotransporters in early endosomes. Third row: control cells were treated with vehicle. Left panel shows strong BLM presence of NBCe1 (green) as in untreatedcells. Bottom row: cells were treated with W-13. Left panel shows loss of NBCe1 (green) from BLM and its redistribution into endosomes. Merged (yellow) inright panel represents costaining of NBCe1 and EEA1, indicating that W-13 locks some of NBCe1 cotransporters in early endosomes. Insets showhigh-magnification images of the cell regions indicated by white rectangles. B-1: endogenous electroneutral NBCn1. Top row: control cells were untreated. Leftpanel shows strong BLM expression of electroneutral NBCn1 (green). Bottom row: cells were treated with monensin. Strong membrane staining and lack ofcolocalization with intracellular EEA1 indicate that NBCn1 is not affected by monensin. In A-1 and B-1, ParC5 cells were grown on coverslips for 4 days in24-well dishes and were untreated or incubated with 50 �M monensin, vehicle 0.01% MeOH, or 100 �M W-13 for 60 min in culture media at 37°C. Cells werefixed, permeabilized, costained with primary antibodies against NBCe1 or NBCn1 and the early endosomal marker EEA1, subsequently treated withcorresponding secondary antibodies Alexa 488 and Cy3, and mounted on slides. 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 at least 60 cells from 15 coverslips from 3 independent experiments.Scale bars, 5 �m. A-2 and B-2: quantitative Pearson R correlation analysis shows that monensin and W-13 increase localization of NBCe1 but not NBCn1 inearly endosomes. A-2: the Pearson R correlation coefficients of colocalization of electrogenic NBCe1 with EEA1 were low in untreated (0.16 � 0.06, n � 124)and vehicle-treated cells (0.15 � 0.05, n � 80). The colocalization of NBCe1 with EEA1 were significantly increased in cells treated with monensin (MON;0.32 � 0.06, n � 60) and W-13 (0.47 � 0.06, n � 84). B-2: the colocalization of NBCn1 with EEA1 were low both in untreated cells (0.13 � 0.02, n � 61)and monensin-treated cells (0.104 � 0.044, n � 68). The Pearson R correlation coefficient matches intensity of the green fluorescence of FITC filter channel(NBCe1 or NBCn1) with the red TRITC filter channel (EEA1) acquired by SlideBook 4.2.0.10 Cross Channel software (Olympus fluorescent Spindiskmicroscope). Bar graphs summarize data generated by analyzing cell images from 15 coverslips in 3 different preparations in experiments like those shown inA-1 and B-1. Bars represent means � SD. *P � 4.01E-10 vs. untreated by 2-tailed Student’s t-test.
C1391ENDOCYTOSIS OF NBCe1 BUT NOT NBCn1 IN PAROTID CELLS
AJP-Cell Physiol • VOL 295 • NOVEMBER 2008 • www.ajpcell.org
was 90.4% � 13.9% (n � 4), which is a 2-fold increasefrom its value of 50.2% � 13.1% (n � 4) in the CCh-treatedcells. Thus our confocal microscopy findings and surfacebiotinylation data indicate that the CCh-induced endocytosisof NBCe1 is indeed due to stimulation of muscarinic recep-tors in ParC5 cells.
PMA- and CCh-stimulated endocytosis of NBCe1 is PKCdependent. Previously, we (43) reported that PMA-inducedinhibition of activity of NBCe1 in Xenopus oocytes is PKCdependent. Here, we investigate the PKC dependency ofNBCe1 endocytosis caused by PMA, which activates theconventional and novel isoforms of PKC. In the experiment in
Fig. 3. PMA and carbachol (CCh) increase endocytosis of electrogenic NBCe1 but not electroneutral NBCn1 in ParC5 cells. A-1: endogenous electrogenicNBCe1. Top row: control cells were untreated. Left panel shows strong BLM expression of electrogenic NBCe1 (green). Middle panel shows EEA1 (red). Rightpanel is a merged image of both staining. Second row: cells were treated with PMA. Merged (yellow) in right panel represents costaining of NBCe1 (green) andEEA1 (red), indicating that PMA induces loss of NBCe1 from BLM and redistribution into early endosomal compartments. Bottom row: cells were treated withCCh. Merged (yellow) image in right panel shows that CCh induces loss of NBCe1 (green) from BLM and its redistribution into intracellular early endosomes(red). Insets show high-magnification images of the cell regions indicated by white rectangles and show points of colocalization of NBCe1 with EEA1 suggestingthat PMA and CCh cause internalization of NBCe1 via early endosomes. B-1: endogenous electroneutral NBCn1. Top row: control cells were untreated. Leftpanel shows strong BLM expression of electroneutral NBCn1. Second row: cells were treated with PMA. Left and right panels show strong membrane stainingof NBCn1 and insignificant colocalization with early endosomes. Bottom row: cells were treated with CCh. Left and right panels show strong membrane stainingof NBCn1 and lack of colocalization with early endosomes. Experiments like these suggest that NBCn1 is not affected by PMA or CCh. ParC5 cells were grownon coverslips for 3 days, and 24-well dishes were untreated or incubated with PMA (1 �M) or CCh (50 �M) in media at 37°C for 15 min. Cells were fixed,permeabilized, costained with primary antibodies against NBCe1 or NBCn1 with EEA1, and subsequently treated with corresponding secondary antibodies Alexa488 and Cy3. 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 at least 60 cells from 15 coverslips from 3 independent experiments. Scale bars, 5 �m. A-2 and B-2: quantitative Pearson Rcorrelation analysis shows that PMA and CCh increase localization of NBCe1 but not NBCn1 in early endosomes. A-2: the Pearson R correlation coefficient ofcolocalization of electrogenic NBCe1 with EEA1 was low (0.15 � 0.05, n � 499) in control untreated cells (UNTR). Both PMA and CCh significantly increasedcolocalization of NBCe1 with EEA1 with the Pearson R correlation coefficients of 0.38 � 0.07 (n � 170) and 0.40 � 0.08 (n � 167), respectively. SpecificPKC inhibitor GF-109203X (GF) significantly decreases PMA- and CCH-induced colocalization of electrogenic NBCe1 with EEA1 (see Fig. 4). The PearsonR correlation coefficients of colocalization of electrogenic NBCe1 with EEA1 were 0.10 � 0.05 (n � 80) and 0.11 � 0.04 (n � 65) in cells treated with a mixtureof PMA plus GF and CCh plus GF, respectively. Also, atropine (Atr) added together with CCh significantly lowered colocalization of electrogenic NBCe1 withEEA1 (0.14 � 0.06, n � 143). B-2: the Pearson R correlation coefficients of colocalization of NBCn1 with EEA1 were similarly low in the control untreatedcells (0.17 � 0.03, n � 270) and in the PMA-treated (0.15 � 0.04, n � 100) or CCh-treated cells (0.18 � 0.04, n � 170). Bar graphs summarize data generatedby analyzing n cells from 15 coverslips in 3 different preparations. Bars represent means � SD (filled bars, *P � 4.01E-10 vs. untreated cells by 2-tailed Student’st-test).
C1392 ENDOCYTOSIS OF NBCe1 BUT NOT NBCn1 IN PAROTID CELLS
AJP-Cell Physiol • VOL 295 • NOVEMBER 2008 • www.ajpcell.org
Fig. 4B (top row), we applied a mixture of 1 �M PMA and 500nM GF (a specific PKC inhibitor) to ParC5 cells for 15 min andthen fixed cells and stained with antibodies against NBCe1 andEEA1. We found that BLM staining remained strong, and therewas no noticeable increase in endosomal formation stainingwith NBCe1. We also found no increase in colocalization ofNBCe1 and EEA1 staining (see Fig. 4, A and B). Thus, in cellstreated with a mixture of PMA and GF, the Pearson R corre-lation coefficient was 0.10 � 0.05 (n � 80), which is similar toits value of 0.15 � 0.05 (n � 499) in untreated cells (see bargraph in Fig. 3A-2). Additionally, our surface biotinylationdata show that the PKC inhibitor GF prevents PMA-induced
removal of NBCe1 from the BLM (see Fig. 5A-3). As sum-marized in the bar graph in Fig. 5A-4, the intensity of thebiotinylated NBCe1 band in the cells treated with a mixture ofPMA and GF was 85.3% � 21.3% (n � 4) of that in untreatedcells. This was in striking contrast to the intensity of thebiotinylated NBCe1 band in cells treated with PMA alone,which was 36.2% � 21.1% (n � 4) of that in untreated cells(see Fig. 5A-4). Thus our confocal microscopy findings andsurface biotinylation data support the view that PMA-inducedendocytosis of NBCe1 in ParC5 cells is PKC dependent.
We (44) also reported that PKC mediates angiotensin II(ANG II)/AT1-stimulated endocytosis of NBCe1 in Xenopus
Fig. 4. Effect of the PKC inhibitor GF and atropine on NBCe1 endocytosis. A: endocytosed NBCe1s are colocalized with transferrin receptors. Top row: controlcells were untreated. Left panel shows strong BLM expression of electrogenic NBCe1. Second row: cells were treated with PMA. Left panel shows substantialloss of NBCe1 from BLM and its redistribution into endosomes. Middle panel shows endosomes stained with Texas red-conjugated transferrin (Tfn-TR). Insetsshow high-magnification images of the cell regions indicated by white rectangles. Insets show points of colocalization of NBCe1 with Tfn-TR. Bottom row: cellswere treated with CCh. Left panel shows substantial loss of NBCe1 from BLM and its redistribution into endosomes. Middle panel shows endosomes stainedwith Tfn-TR. Insets show points of colocalization between NBCe1 and Tfn-TR. ParC5 cells were untreated or incubated with PMA (1 �M) or CCh (50 �M)in media containing Tfn-TR (5 �g/ml) at 37°C for 15 min. Fixed cells were stained with a primary antibody against NBCe1 and treated with secondary antibodyAlexa 488. Merged images in conjunction with insets help show colocalization. B: PKC inhibitor GF prevents PMA- and CCh-induced endocytosis of NBCe1.Atropine blocks CCh-induced endocytosis of NBCe1. Top row: cells were treated with PMA and 0.5 �M GF, a specific inhibitor of all PKC isoforms. Strongmembrane staining of NBCe1 and lack of colocalization with EEA1 suggest that PKC mediates PMA-induced endocytosis of NBCe1. Second row: cells weretreated with CCh and GF. Strong membrane staining of NBCe1 and lack of its colocalization with EEA1 suggest that CCh-induced endocytosis of NBCe1 isalso PKC dependent. Bottom row: Cells were treated with CCh and atropine (10 �M), an antagonist of the muscarinic receptor. Strong membrane staining ofNBCe1 suggests that binding of CCh agonist to muscarinic receptors induces endocytosis of NBCe1. A z-stack of optical sections was acquired through the FITC(green) and TRITC (red) filter channels. Images are representative of at least 6 images from at least 3 independent experiments. Collapsed z-stacks near BLMare shown. Cells were costained with primary antibodies against NBCe1 and EEA1 and subsequently treated with the corresponding secondary antibodies Alexa488 and Cy3. Bars, 5 �m. C: PKC-dependent ACh inhibition of NBCe1 current in Xenopus oocytes. ACh inhibits functional activity of NBCe1 in atime-dependent manner. This inhibition is partially prevented by PKC blocker GF. Voltage-clamp Na�-HCO3
� cotransporter (NBC) currents were obtained fromoocytes coexpressing NBCe1 and muscarinic type-3 receptor (M3R). Oocytes were superfused with 5% CO2-33 mM HCO3
� solution, voltage clamped to �50mV for 10 min, and depolarized from �50 to 0 mV to record either control currents before treatment or test currents following treatment (see EXPERIMENTAL
PROCEDURES). Values are means � SD of relative NBCe1 peak currents (INBC) acquired as the ratio of test to control peak currents in 6 independent experiments[as described earlier (28), and see EXPERIMENTAL PROCEDURES]. Oocytes were untreated or treated with 0.5 �M ACh (open bars; *P � 0.005 vs. untreated cellsby Student’s t-test) or 0.5 �M ACh and 0.5 �M GF (filled bars; *P � 0.005 vs. untreated cells by Student’s t-test) for 5, 10, 15, and 20 min. We corrected ourdata for a small background current of 25 nA recorded in HCO3
�-free solution in the presence of the NBCe1 inhibitor DIDS (200 nM) or in water-injectedoocytes.
C1393ENDOCYTOSIS OF NBCe1 BUT NOT NBCn1 IN PAROTID CELLS
AJP-Cell Physiol • VOL 295 • NOVEMBER 2008 • www.ajpcell.org
oocytes. It is known that, similar to AT1R activation, bindingof agonists to MR activates PKC in ParC5 cells (51, 67). Here,we aimed to investigate the PKC dependence of the CCh-stimulated endocytosis of NBCe1. First, we examined the PKCdependence of the cholinergic regulation of the activity ofNBCe1 coexpressed with the M3R in Xenopus oocytes. Forthis, we performed study using the MR agonist ACh. Results ofthese experiments are summarized in the bar graph in Fig. 4C.Briefly, oocytes were superfused with 5% CO2-33 mM HCO3
�
solution, voltage clamped to �50 mV for 10 min, and depo-larized from �50 to 0 mV to record either control currentsbefore treatment or test currents following treatment. At 0 min,control currents were measured, and oocytes were treated with500 nM ACh or with a mixture of 500 nM ACh and 500 nMGF. NBCe1 currents were recorded every 5 min. Figure 4Cshows that the NBCe1 currents decrease over time with AChapplication. With addition of the PKC inhibitor GF to the AChtreatment, the NBCe1 current stays statistically the same over
at least a 15-min period. Thus this study shows that CCh-induced inhibition of activity of NBCe1 in Xenopus oocyteswas also PKC dependent.
Second, to investigate the PKC dependence of the CCh-stimulated endocytosis of NBCe1, we applied a mixture of 50�M CCh and 500 nM GF to ParC5 cells followed by stainingwith antibodies against NBCe1 and EEA1 (Fig. 4B, middlerow). We found that BLM staining remained strong, and therewas no noticeable increase in endosomal formation stainingwith NBCe1. We also found that PKC inhibitor significantlydecreases CCh-induced colocalization of NBCe1 and EEA1. Inthe cells treated with a mixture of CCh and GF, the Pearson Rcorrelation coefficient dropped to a value of 0.11 � 0.04 (n �65) from its value of 0.40 � 0.08 (n � 167) in the CCh-treatedcells. Note that its value was 0.15 � 0.05 (n � 499 cells) in thecontrol untreated cells (see bar graph in Fig. 3A-2). Addition-ally, our surface biotinylation studies showed that the PKCinhibitor GF prevents CCh-induced removal of NBCe1 from
Fig. 5. Surface biotinylation of NBCe1 and NBCn1 in ParC5 cells. A: surface biotinylation shows removal of the electrogenic NBCe1 from the cell surface.A-1 and A-3: typical Western blots of the cell surface-biotinylated NBCe1 proteins. In A-1, cells were untreated or treated for 60 min at 37°C in the presenceof vehicle 0.01% MeOH, 50 �M monensin, or 100 �M W-13. In A-3, cells were untreated or treated for 30 min at 37°C in the presence of 1 �M PMA, mixtureof 1 �M PMA and 1 �M GF, 50 �M CCh, mixture of 50 �M CCh with 1 �M GF, or mixture of 50 �M CCh with 1 �M atropine. Arrow indicates NBCe1band at 130 kDa. A-2 and A-4: graphical representation of quantitative analysis of surface-biotinylated NBCe1 protein in treated cells normalized to that inuntreated cells from experiments like those shown in A-1 and A-3. In A-2, normalized level of the surface-biotinylated NBCe1 proteins were 88.9% � 11.1%(n � 4), 23.8% � 19.6% (n � 4), and 33.3% � 19.5% (n � 4) in vehicle-treated (MeOH), monensin-treated, and W-13-treated cells, respectively. In A-4,normalized level of the surface-biotinylated NBCe1 proteins were 36.2% � 21.1% (n � 4), 85.3% � 21.3% (n � 4), 50.2% � 13.1% (n � 4), 86.1% � 10.1%(n � 4), and 90.4% � 13.9% (n � 4) in cells treated with PMA, mixture of PMA with GF, CCh, mixture of CCh with GF, and mixture of CCh with atropine,respectively. B: surface biotinylation shows constant presence of the electroneutral NBCn1 at the cell surface. B-1: typical Western blots of the surface-biotinylated NBCn1 proteins in cells, which were untreated or treated for 30 min at 37°C in the presence of 1 �M PMA, 50 �M CCh, or treated for 60 min at37°C in the presence of 50 �M monensin. The arrow indicates band of NBCn1 at 127 kDa. B-2: graphical representation of the quantitative analysis of thesurface-biotinylated NBCn1 protein in the treated cells normalized to that in the untreated cells from experiments like those shown in B-1. The normalized levelsof the surface-biotinylated NBCn1 protein were 110.2% � 8.9% (n � 6), 101.2% � 17.4% (n � 6), and 99.1% � 26.9% (n � 6) in cells treated with PMA,CCh, and monensin, respectively. In all experiments, untreated and treated cells were subsequently incubated in the presence of EZ-Link Sulfo-NHS-SS-Biotinfor 1 h at 4°C. Biotinylated proteins were recovered from cell lysates by immunoprecipitation overnight with NeutrAvidin-conjugated beads and resolved bySDS-PAGE. Biotinylated NBCe1 band was detected using anti-NBCe1 polyclonal antibodies (KIA; 1:500). Biotinylated NBCn1 band was detected usinganti-NBCn1 affinity purified polyclonal antibodies (NBCn1; 1:500). Values are means � SD (filled bars; *P � 0.005 vs. untreated cells by Student’s t-test). Datawere generated using 6 different preparations of ParC5 cells.
C1394 ENDOCYTOSIS OF NBCe1 BUT NOT NBCn1 IN PAROTID CELLS
AJP-Cell Physiol • VOL 295 • NOVEMBER 2008 • www.ajpcell.org
the BLM. As summarized in the bar graph in Fig. 5A-4, anormalized intensity of the biotinylated NBCe1 band in thecells treated with a mixture of GF with CCh was 86.1% �10.1% (n � 4) compared with its value of 50.2% � 13.1%(n � 4) in the CCh-treated cells. Therefore, our confocalmicroscopy findings and surface biotinylation data demon-strated that PKC inhibitor GF completely prevents PMA- andCCh-induced endocytosis of NBCe1, suggesting PKC depen-dence of NBCe1 membrane trafficking in ParC5 cells.
PMA and CCh increase colocalization of NBCe1 with trans-ferrin-stained endosomes. Since transferrin is a well-studiedligand of the transferrin receptor that endocytoses through theclathrin-dependent mechanism (35), we used Tfn-TR to inves-tigate possible mechanisms of internalization of NBCe1 inParC5 cells (Fig. 4A). We found that internalized NBCe1colocalizes with Tfn-TR within the same endosomes in PMA-and CCh-treated cells (see insets in Fig. 4A, middle and bottomrows). These data suggest possible involvement of clathrin instimulated endocytosis of NBCe1 in ParC5 cells. Furtherstudies manipulating clathrin or its mechanisms must be done,however, to determine whether it is, in fact, involved.
DISCUSSION
BLM staining of both electrogenic NBCe1 and electroneu-tral NBCn1 in polarized ParC5 cells. Cholinergic MR agonistsare major secretory stimuli with HCO3
� luminal secretion beingan important component of isotonic fluid secretion in ratparotid acinar cells (6, 20, 39). A dual function of acinar cellsis to carry out HCO3
� flux and to maintain neutral intracellularpH (pHi) during stimulated secretion. HCO3
� transport viaNBC could support HCO3
� secretion and/or normalize pHi ofHCO3
�-secreting cells. Indeed, electrogenic NBCe1 and elec-troneutral NBCn1 are found in the HCO3
�-secreting epitheliaof pancreas, guts, distal renal tubule, and salivary glands (1, 4,9, 12, 21, 24–26, 29, 33, 42, 46, 47, 52, 53, 55). Here, weinvestigated the subcellular localization and regulation ofmembrane trafficking of NBCe1 and NBCn1 in the ParC5 cellsthat are functionally similar to native parotid acinar cells (48,49, 63). The ParC5 cell is also an established model forMR-mediated secretion (8). We observed strong expression ofNBCe1 and NBCn1 in the BLM of polarized ParC5 cells (Fig.1). Our results showing BLM staining of electrogenic cotrans-porter is in agreement with reports by others that NBCe1 ispresent at the BLM in rat, human, mouse, and guinea pigparotid acinar cells (27, 31, 42, 53).
In contrast to our detection of electroneutral cotransporter atthe BLM in ParC5 cells, Gresz et al. (21) did not find anyNBCn1 in rat parotid acini. Gresz et al. (21) used an indirectimmunoperoxidase labeling of dewaxed and rehydrated paraf-fin tissue sections, whereas here we used confocal immunoflu-orescence microscopy in fixed cultured cells. We also supportour finding with Western blot analysis of the biotinylatedNBCn1 protein extracted from ParC5 cells (Fig. 5B). There-fore, this discrepancy may be explained by differences in thesample preparation and/or in the sensitivity of these differentapproaches. In agreement of our finding of the BLM staining ofNBCn1 in HCO3
�-secreting ParC5 cells is the reported detec-tion of NBCn1 at the BLM in HCO3
�-secreting renal distalmedullary B intercalated cells (46, 47). Therefore, HCO3
� entry
via BLM NBC could be one of the important features of thesalivary fluid secretory response.
Constitutive endocytosis of electrogenic NBCe1 but notelectroneutral NBCn1 in ParC5 cells. Functional activity ofbicarbonate transporters depends on the amount of their pro-teins at the cell surface, which could be regulated by endocy-tosis, as we (43) previously showed in Xenopus oocytes. Here,we detected a striking difference in membrane trafficking ofelectrogenic NBCe1 vs. electroneutral NBCn1 in ParC5 cells.In particular, our evidence strongly suggests that NBCe1, butnot NBCn1, undergoes constitutive endocytosis. In support ofthis, a 60-min blockage of constitutive recycling by monensinor W-13 leads to loss of surface NBCe1 and redistribution ofthe cotransporter from the BLM into EEA1-marked cytosoliccompartments (Figs. 2A and 5A). In contrast, electroneutralNBCn1 was constantly present at the cell surface of ParC5cells after 60-min (Figs. 2B and 5B) or overnight treatmentwith monensin (data not shown).
PKC-dependent endocytosis of electrogenic NBCe1 but notelectroneutral NBCn1 in stimulated ParC5 cells. Next, weinvestigated the effect of cholinergic agonists on the membranetrafficking of NBC in ParC5 cells. We found here that the MRagonist CCh stimulates massive endocytosis of electrogenicNBCe1. A 15-min CCh application leads to a loss of NBCe1from the cell surface and its redistribution into endosomes,some of which are marked with EEA1 early endosomal marker(Fig. 3A-1, bottom row). ACh, another MR agonist, alsoinduced similar endocytosis of NBCe1 in ParC5 cells (data notshown). Working with Xenopus oocytes, we found that AChinhibits NBCe1 activity, which could reflect similar activityinhibition due to endocytosis of NBCe1 in ParC5 cells (Fig.4C). This speculation is supported by our reports (43, 44) ofAT1R-induced endocytosis of NBCe1 and parallel inhibition ofits activity in Xenopus oocytes. Thus endocytosis of NBCe1could significantly reduce translocation of HCO3
� ions and anet negative charge across the BLM in ParC5 cells. We foundthat the MR blocker, atropine, completely inhibits CCh-in-duced (Figs. 3A-2, 4B, and 5A) and ACh-induced (data notshown) endocytosis of NBCe1. Therefore, signaling via Gprotein-coupled MR receptor mediates endocytosis of NBCe1in stimulated ParC5 cells. This new finding is in agreementwith our previous report (43) that signaling via G protein-coupled AT1R stimulates endocytosis of NBCe1 in Xenopusoocytes.
Stimulation of MR and AT1R G protein-coupled receptorsactivates PKC, which can be delivered to the plasma mem-brane to phosphorylate its multiple target proteins (14). Here,we showed that the PKC-specific inhibitor GF prevents CCh-stimulated endocytosis of NBCe1 in ParC5 cells (Figs. 3A, 4B,and 5A). To additionally support the view that PKC regulatesendocytosis of NBCe1, we used the PKC activator PMAknown to stimulate endocytosis of many membrane proteins inmammalian cells (56). We found that PMA stimulates PKC-dependent endocytosis of NBCe1, which can be blocked by GF(see Fig. 3A, 2nd row, Fig. 4A, 2nd row, and Fig. 4B, 1st row).These new data are in agreement with our previous reports (43,44) that AT1R-activated PKC binds to recombinant NBCe1,which undergoes PKC-dependent endocytosis in Xenopus oo-cytes. Here, we used Tfn-TR, an agonist of the transferrinreceptor known to endocytose via the clathrin-dependent path-way (56). We found that internalized NBCe1 was colocalized
C1395ENDOCYTOSIS OF NBCe1 BUT NOT NBCn1 IN PAROTID CELLS
AJP-Cell Physiol • VOL 295 • NOVEMBER 2008 • www.ajpcell.org
with Tfn-TR-positive endosomes in ParC5 cells treated withCCh or PMA (Fig. 4A). Further study is needed to clarifywhether PKC-dependent CCh- and/or PMA-stimulated endo-cytosis of NBCe1 is also clathrin dependent.
In striking contrast, the electroneutral NBCn1 was con-stantly present at the BLM after 15-min of CCh or PMAapplication (Figs. 3B and 5B). Longer 30-min treatment withCCh or PMA produced no changes in BLM expression ofNBCn1 (data not shown). Interestingly, Kwon et al. (29) alsoreported that the abundance of the BLM NBCn1 was un-changed in response to ANG II treatment in the medullarythick ascending limb (mTAL) cells of rat kidney.
Why would NBCn1 not go through constitutive endocytosis,and why would NBCn1 not endocytose in response to cholin-ergic stimulation of ParC5 cells? We can speculate thatNBCn1, unlike NBCe1, is not regulated by PKC due to lowprobability of putative PKC phosphorylation sites in its aminoacid sequence (analysis by NetPhos 2.0 at http://www.cbs.dtu.dk; data not shown). Therefore, lack of PKC regulation ofNBCn1 could possibly prevent its endocytosis in response tocholinergic stimulation of ParC5 cells. Further support for thisspeculation is found in our earlier report (44) that in Xenopusoocytes, AT1R-activated PKC failed to stimulate endocytosisof excitatory amino acid transporter (EAAT3), which is notregulated by PKC. These speculations may be able to explainthe absence of activated endocytosis but would not suffice toexplain constitutive endocytosis. However, the amino acidsequence of NBCn1 has a putative PDZ-binding site, the last 4amino acids (ETSL) of its COOH terminus. We can thereforespeculate that NBCn1 could be anchored to the BLM cytoskel-eton by PDZ adapter proteins. This could stabilize NBCn1 atthe BLM in ParC5 cells, similar to a model of potassiumchannel stabilization proposed by Tamkun’s group (60).
Possible physiological roles of NBCe1 and NBCn1 in ParC5cells. Differences in the regulation of membrane trafficking ofNBCe1 vs. NBCn1 indicates that electrogenic and electroneu-tral processes of Na� and HCO3
� translocation across the BLMwould play different roles in HCO3
� and/or total salt and fluidluminal secretion and/or in pHi regulation in ParC5 cells. BothNBCe1 and NBCn1 translocate HCO3
� across the BLM topossibly support a net HCO3
� secretion in acinar cells. Inaddition to Na� and HCO3
� influx, the electrogenic NBCe1translocates a net negative charge across the BLM as reportedin bovine parotid acinar cells (66). Therefore, transport viaNBCe1 would contribute to electronegative membrane poten-tial, which drives luminal secretion in parotid acinar cells (20).As we observed, NBCe1 undergoes removal from the BLM ofParC5 cells in 5–15 min following cholinergic stimulation.Interestingly, we (63) reported earlier that within the first 30 sof agonist stimulation, ParC5 cells undergo transient increasein anion, i.e., Cl� and HCO3
�, luminal secretion, which, afterapproximately 2–5 min, was gradually decreased to a lowsteady-state level of secretion lasting at least for 10 min. Thistransient decrease of anion secretion coincides with reports byothers of a transition between the initial and the sustainedphases of stimulated fluid secretion in parotid acinar cells(36–38). Therefore, we speculate that the timing of endocyto-sis and inhibition of NBCe1 could be a part of acinar celladjustment to a new steady state characteristic of a continuoussecretory response.
In striking contrast, the electroneutral NBCn1 is stablypresent at the BLM throughout cholinergic stimulation, possi-bly continuously supporting its portion of the HCO3
� flux viathe BLM. For NBCn1, a low cellular Na� concentration isusually sufficient to drive inwardly directed ionic transport (2,10). We speculate that the role of the BLM NBCn1 would beto allow for an accompanying HCO3
� entry to maintain neutralpHi in secreted parotid acinar cells. This is similar to the BLMK� channels, which allow for an accompanying K� efflux tomaintain intracellular electroneutrality in acinar cells (20).
Parotid acinar cells secrete salt and fluid in response toparasympathetic muscarinic stimulation. It is the resultinganion loss, including massive HCO3
� loss, across the apicalmembrane that drives transepithelial fluid secretion in acinarcells. Therefore, the BLM Na�, HCO3
�, and a net negativecharge entry via two types of NBC could be important featuresof the salivary fluid secretory response. Undoubtedly, futurestudies are needed to measure activity of NBCe1 in stimulatedparotid acinar cells, especially during the first 30 s of increasedanion, i.e., Cl� and HCO3
�, secretion. Future studies are alsoneeded to measure the activity of NBCn1 in resting andstimulated acinar cells. Since current models of fluid secretionproposed for parotid acinar cells do not include NBC (20, 39,41), determining the roles that NBC plays within these modelswould ultimately lead to a more thorough understanding of themolecular physiology of the fluid secretion process in parotidacinar cells.
GRANTS
This work is supported by The Pilot Project Grant in Women’s HealthResearch, University of Colorado Denver (I. I. Grichtchenko) and NationalInstitute of Dental and Craniofacial Research Grant DE-015648 (M. E. Rey-land).
REFERENCES
1. Abuladze N, Lee I, Newman D, Hwang J, Boorer K, Pushkin A, KurtzI. Molecular cloning, chromosomal localization, tissue distribution, andfunctional expression of the human pancreatic sodium bicarbonate co-transporter. J Biol Chem 273: 17689–17695, 1998.
2. Amlal H, Burnham CE, Soleimani M. Characterization of the Na�:HCO3
3. Apodaca G, Enrich C, Mostov KE. The calmodulin antagonist, W-13,alters transcytosis, recycling, and the morphology of the endocytic path-way in Madin-Darby canine kidney cells. J Biol Chem 269: 19005–19013,1994.
4. Bachmann O, Reichelt D, Tuo B, Manns MP, Seidler U. Carbacholincreases Na�/HCO3
� cotransport activity in murine colonic crypts in aM3
�, Ca2�/calmodulin-, and PKC-dependent manner. Am J Physiol Gas-trointest Liver Physiol 291: G650–G657, 2006.
5. Basu SK, Goldstein JL, Anderson RG, Brown MS. Monensin interruptsthe recycling of low density lipoprotein receptors in human fibroblasts.Cell 24: 493–502, 1981.
6. Baum BJ. Principles of saliva secretion. Ann N Y Acad Sci 694: 17–23,1993.
� cotransporter (NBC) with a novel COOH-termi-nus, cloned from rat brain. Am J Physiol Cell Physiol 278: C1200–C1211,2000.
8. Bockman CS, Bradley ME, Dang HK, Zeng W, Scofield MA, DowdFJ. Molecular and pharmacological characterization of muscarinic recep-tor subtypes in a rat parotid gland cell line: comparison with native parotidgland. J Pharmacol Exp Ther 297: 718–726, 2001.
9. Bok D, Schibler MJ, Pushkin A, Sassani P, Abuladze N, Naser Z,Kurtz I. Immunolocalization of electrogenic sodium-bicarbonate cotrans-porters pNBC1 and kNBC1 in the rat eye. Am J Physiol Renal Physiol281: F920–F935, 2001.
C1396 ENDOCYTOSIS OF NBCe1 BUT NOT NBCn1 IN PAROTID CELLS
AJP-Cell Physiol • VOL 295 • NOVEMBER 2008 • www.ajpcell.org
12. Damkier HH, Nielsen S, Praetorius J. Molecular expression of SLC4-derived Na�-dependent anion transporters in selected human tissues. Am JPhysiol Regul Integr Comp Physiol 293: R2136–R2146, 2007.
13. de Figueiredo P, Brown WJ. A role for calmodulin in organelle mem-brane tubulation. Mol Biol Cell 6: 871–887, 1995.
14. Dempsey EC, Newton AC, Mochly-Rosen D, Fields AP, Reyland ME,Insel PA, Messing RO. Protein kinase C isozymes and the regulation ofdiverse cell responses. Am J Physiol Lung Cell Mol Physiol 279: L429–L438, 2000.
15. Denker BM, Nigam SK. Molecular structure and assembly of the tightjunction. Am J Physiol Renal Physiol 274: F1–F9, 1998.
16. Dinour D, Chang MH, Satoh J, Smith BL, Angle N, Knecht A, SerbanI, Holtzman EJ, Romero MF. A novel missense mutation in the sodiumbicarbonate cotransporter (NBCe1/SLC4A4) causes proximal tubular ac-idosis and glaucoma through ion transport defects. J Biol Chem 279:52238–52246, 2004.
17. Espiritu DJ, Bernardo AA, Robey RB, Arruda JA. A central role forPyk2-Src interaction in coupling diverse stimuli to increased epithelialNBC activity. Am J Physiol Renal Physiol 283: F663–F670, 2002.
18. Evans RL, Turner RJ. Upregulation of Na�-K�-2Cl� cotransporteractivity in rat parotid acinar cells by muscarinic stimulation. J Physiol 499:351–359, 1997.
20. Gin E, Crampin EJ, Brown DA, Shuttleworth TJ, Yule DI, Sneyd J.A mathematical model of fluid secretion from a parotid acinar cell. J TheorBiol 248: 64–80, 2007.
21. Gresz V, Kwon TH, Vorum H, Zelles T, Kurtz I, Steward MC,Aalkjaer C, Nielsen S. Immunolocalization of electroneutral Na�-HCOcotransporters in human and rat salivary glands. Am J Physiol GastrointestLiver Physiol 283: G473–G480, 2002.
22. Hidaka H, Takara T. Modulation of Ca2�-dependent regulatory systemsby calmodulin antagonists and other agents. In: Calmodulin Antagonistsand Cellular Physiology, edited by Hidaka H and Hartshorne DJ. NewYork: Academic Press, 1985, p. 13–23.
23. Inatomi J, Horita S, Braverman N, Sekine T, Yamada H, Suzuki Y,Kawahara K, Moriyama N, Kudo A, Kawakami H, Shimadzu M,Endou H, Fujita T, Seki G, Igarashi T. Mutational and functionalanalysis of SLC4A4 in a patient with proximal renal tubular acidosis.Pflugers Arch 448: 438–444, 2004.
29. Kwon TH, Nielsen J, Kim YH, Knepper MA, Frokiaer J, Nielsen S.Regulation of sodium transporters in the thick ascending limb of ratkidney: response to angiotensin II. Am J Physiol Renal Physiol 285:F152–F165, 2003.
30. Kwon TH, Pushkin A, Abuladze N, Nielsen S, Kurtz I. Immunoelectronmicroscopic localization of NBC3 sodium-bicarbonate cotransporter in ratkidney. Am J Physiol Renal Physiol 278: F327–F336, 2000.
31. Li J, Koo NY, Cho IH, Kwon TH, Choi SY, Lee SJ, Oh SB, Kim JS,Park K. Expression of the Na�-HCO3
� cotransporter and its role in pHi
regulation in guinea pig salivary glands. Am J Physiol Gastrointest LiverPhysiol 291: G1031–G1040, 2006.
32. Llorente A, Garred O, Holm PK, Eker P, Jacobsen J, van Deurs B,Sandvig K. Effect of calmodulin antagonists on endocytosis and intracel-lular transport of ricin in polarized MDCK cells. Exp Cell Res 227:298–308, 1996.
33. Luo X, Choi JY, Ko SB, Pushkin A, Kurtz I, Ahn W, Lee MG,Muallem S. HCO3
� salvage mechanisms in the submandibular glandacinar and duct cells. J Biol Chem 276: 9808–9816, 2001.
34. Melikian HE, Buckley KM. Membrane trafficking regulates the activityof the human dopamine transporter. J Neurosci 19: 7699–7710, 1999.
35. Mellman I. Endocytosis and molecular sorting. Annu Rev Cell Dev Biol12: 575–625, 1996.
36. Melvin JE, Kawaguchi M, Baum BJ, Turner RJ. A muscarinic agonist-stimulated chloride efflux pathway is associated with fluid secretion in ratparotid acinar cells. Biochem Biophys Res Commun 145: 754–759, 1987.
37. Melvin JE, Moran A, Turner RJ. The role of HCO3� and Na�/H�
exchange in the response of rat parotid acinar cells to muscarinic stimu-lation. J Biol Chem 263: 19564–19569, 1988.
38. Melvin JE, Turner RJ. Cl� fluxes related to fluid secretion by the ratparotid: involvement of Cl�-HCO3
� exchange. Am J Physiol GastrointestLiver Physiol 262: G393–G398, 1992.
39. Melvin JE, Yule D, Shuttleworth T, Begenisich T. Regulation of fluidand electrolyte secretion in salivary gland acinar cells. Annu Rev Physiol67: 445–469, 2005.
40. Miranda M, Wu CC, Sorkina T, Korstjens DR, Sorkin A. Enhancedubiquitylation and accelerated degradation of the dopamine transportermediated by protein kinase C. J Biol Chem 280: 35617–35624, 2005.
41. Nakamoto T, Srivastava A, Romanenko VG, Ovitt CE, Perez-CornejoP, Arreola J, Begenisich T, Melvin JE. Functional and molecularcharacterization of the fluid secretion mechanism in human parotid acinarcells. Am J Physiol Regul Integr Comp Physiol 292: R2380–R2390, 2007.
42. Park K, Hurley PT, Roussa E, Cooper GJ, Smith CP, Thevenod F,Steward MC, Case RM. Expression of a sodium bicarbonate cotrans-porter in human parotid salivary glands. Arch Oral Biol 47: 1–9, 2002.
43. Perry C, Blaine J, Le H, Grichtchenko II. PMA- and ANG II-inducedPKC regulation of the renal Na�-HCO3
� cotransporter (hkNBCe1). Am JPhysiol Renal Physiol 290: F417–F427, 2006.
44. Perry C, Le H, Grichtchenko II. ANG II and calmodulin/CaMKIIregulate surface expression and functional activity of NBCe1 via separatemeans. Am J Physiol Renal Physiol 293: F68–F77, 2007.
46. Praetorius J, Kim YH, Bouzinova EV, Frische S, Rojek A, Aalkjaer C,Nielsen S. NBCn1 is a basolateral Na�-HCO3
� cotransporter in rat kidneyinner medullary collecting ducts. Am J Physiol Renal Physiol 286: F903–F912, 2004.
47. Pushkin A, Yip KP, Clark I, Abuladze N, Kwon TH, Tsuruoka S,Schwartz GJ, Nielsen S, Kurtz I. NBC3 expression in rabbit collectingduct: colocalization with vacuolar H�-ATPase. Am J Physiol RenalPhysiol 277: F974–F981, 1999.
48. Quissell DO, Barzen KA, Redman RS, Camden JM, Turner JT.Development and characterization of SV40 immortalized rat parotid acinarcell lines. In Vitro Cell Dev Biol Anim 34: 58–67, 1998.
49. Quissell DO, Turner JT, Redman RS. Development and characteriza-tion of immortalized rat parotid and submandibular acinar cell lines. EurJ Morphol 36, Suppl: 50–54, 1998.
49a.Rajbhandari I, Kim E, Choi I. Interaction of the Na/HCO3 cotransporterSLC4A7 (NBCn1) and the N-methyl-D-aspartate (NMDA) receptor sub-unit NR2A. FASEB J 22: 759.8, 2008.
50. Reshkin SJ, Lee SI, George JN, Turner RJ. Identification, character-ization and purification of a 160 kD bumetanide-binding glycoproteinfrom the rabbit parotid. J Membr Biol 136: 243–251, 1993.
51. Rosado JA, Salido GM, Garcia LJ. Activation of m3 muscarinic recep-tors induces rapid tyrosine phosphorylation of p125(FAK), p130(cas), andpaxillin in rat pancreatic acini. Arch Biochem Biophys 377: 85–94, 2000.
52. Roussa E, Nastainczyk W, Thevenod F. Differential expression ofelectrogenic NBC1 (SLC4A4) variants in rat kidney and pancreas. Bio-chem Biophys Res Commun 314: 382–389, 2004.
53. Roussa E, Romero MF, Schmitt BM, Boron WF, Alper SL, ThevenodF. Immunolocalization of anion exchanger AE2 and Na�-HCO3
� cotrans-porter in rat parotid and submandibular glands. Am J Physiol GastrointestLiver Physiol 277: G1288–G1296, 1999.
C1397ENDOCYTOSIS OF NBCe1 BUT NOT NBCn1 IN PAROTID CELLS
AJP-Cell Physiol • VOL 295 • NOVEMBER 2008 • www.ajpcell.org
54. Ruiz OS, Qiu YY, Cardoso LR, Arruda JA. Regulation of the renalNa-HCO3 cotransporter: VII. Mechanism of the cholinergic stimulation.Kidney Int 51: 1069–1077, 1997.
55. Satoh H, Moriyama N, Hara C, Yamada H, Horita S, Kunimi M,Tsukamoto K, Iso-O N, Inatomi J, Kawakami H, Kudo A, Endou H,Igarashi T, Goto A, Fujita T, Seki G. Localization of Na�-HCO�
3
cotransporter (NBC-1) variants in rat and human pancreas. Am J PhysiolCell Physiol 284: C729–C737, 2003.
56. Sorkina T, Hoover BR, Zahniser NR, Sorkin A. Constitutive andprotein kinase C-induced internalization of the dopamine transporter ismediated by a clathrin-dependent mechanism. Traffic 6: 157–170, 2005.
57. Sreebny LM, Valdini A. Xerostomia. Part I: relationship to other oralsymptoms and salivary gland hypofunction. Oral Surg Oral Med OralPathol 66: 451–458, 1988.
58. Stein BS, Bensch KG, Sussman HH. Complete inhibition of transferrinrecycling by monensin in K562 cells. J Biol Chem 259: 14762–14772, 1984.
59. Steward MC, Poronnik P, Cook DI. Bicarbonate transport in sheepparotid secretory cells. J Physiol 494: 819–830, 1996.
60. Tamkun MM, O’connell KM, Rolig AS. A cytoskeletal-based perimeterfence selectively corrals a sub-population of cell surface Kv2.1 channels.J Cell Sci 120: 2413–2423, 2007.
61. Tebar F, Villalonga P, Sorkina T, Agell N, Sorkin A, Enrich C.Calmodulin regulates intracellular trafficking of epidermal growth factor
receptor and the MAPK signaling pathway. Mol Biol Cell 13: 2057–2068,2002.
62. Tobin G, Giglio D, Gotrick B. Studies of muscarinic receptor subtypes insalivary gland function in anaesthetized rats. Auton Neurosci 100: 1–9, 2002.
64. Vorum H, Kwon TH, Fulton C, Simonsen B, Choi I, Boron W,Maunsbach AB, Nielsen S, Aalkjaer C. Immunolocalization of electro-neutral Na-HCO3
� cotransporter in rat kidney. Am J Physiol Renal Physiol279: F901–F909, 2000.
65. Whittaker J, Hammond VA, Taylor R, Alberti KG. Effects of monen-sin on insulin interactions with isolated hepatocytes. Evidence for inhibi-tion of receptor recycling and insulin degradation. Biochem J 234: 463–468, 1986.
66. Yamaguchi S, Ishikawa T. Electrophysiological characterization of na-tive Na�-HCO3
� cotransporter current in bovine parotid acinar cells.J Physiol 568: 181–197, 2005.
67. Yule DI, Williams JA. CCK antagonists reveal that CCK-8 and JMV-180interact with different sites on the rat pancreatic acinar cell CCKAreceptor. Peptides 15: 1045–1051, 1994.
C1398 ENDOCYTOSIS OF NBCe1 BUT NOT NBCn1 IN PAROTID CELLS
AJP-Cell Physiol • VOL 295 • NOVEMBER 2008 • www.ajpcell.org
