| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Department of Internal Medicine and 2Department of Physiology, Virginia Commonwealth University, Richmond 23298; 3McGuire Veterans Affairs Medical Center, Richmond, Virginia 23249; and 4National Institutes of Health, Bethesda, Maryland 20892-1603
Submitted 20 August 2003; accepted in final form 29 October 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGMENTSREFERENCES |
|---|
). TRCs are polarized neuro-epithelial cells that can be studied in vitro while maintaining their natural polarity in the lingual epithelium. Stimulating the apical membranes of TRCs with acids induced sustained decreases in intracellular pH (pHi) (Lyall et al. 2001, 2002a,b). During acid stimulation, a decrease in TRC pHi, rather than a decrease in extracellular pH (pHo), is the stimulus intensity variable that correlates specifically with increased chorda tympani (CT) taste nerve activity. Inhibiting acid-induced TRC acidification also inhibits the acid-evoked CT response (Lyall et al. 2001, 2002b). These results indicate that a decrease in TRC pHi is the proximate stimulus for sour taste. Further studies indicate that, in the case of strong mineral acids (HCl), both apical H+ entry and H+ exit across the basolateral membrane of TRCs are regulated by second messengers, which also modulate CT responses to HCl (Lyall et al. 2002a). An increase in intracellular cAMP enhanced the sour taste of strong acids by activating a Ca2+- and amiloride-insensitive apical H+ conductance and inhibited pHi recovery in TRC membranes. In contrast, an increase in intracellular Ca2+ concentration ([Ca2+]i) stimulated pHi recovery in TRCs, which increased sensory adaptation to acids (Lyall et al. 2002a). Overall, the data indicate that pH recovery mechanisms in TRC membranes play an important role in sour taste transduction and adaptation, but these have not been characterized so far.
Isolated rat (Lyall et al. 1997) and hamster (Stewart et al. 1998) TRCs, maintained at pHo 7.4, demonstrated spontaneous pHi recovery from intracellular acid loading. However, during acid stimulation, i.e., when changes in TRC pHi were induced by a decrease in pHo, no spontaneous recovery of TRC pHi was observed. Together, these results suggest that TRCs regulate pHi when perturbations in pHi occur at constant pHo of 7.4, but pH regulatory mechanisms are attenuated during acid stimulation (i.e., when pHo is decreased). Because TRCs are epithelial cells, with distinct apical and basolateral membrane domains, pH regulatory mechanisms can differ in kind and function between membrane domains (Noel and Pouysségur 1995; Ritter et al. 2001). We have therefore used a method that allows us to make measurements on a single fungiform taste bud while maintaining a normal polarized epithelial environment (Lyall et al. 2001, 2002a,b; Simon 2002). Using pH imaging, Na+ imaging, RT-PCR, and immunocytochemical methods, we have identified both Na+-H+ exchanger isoform 1 (NHE-1) and isoform 3 (NHE-3) in fungiform and circumvallate TRCs. We present evidence that NHE-3 is present in the apical membranes of fungiform TRCs, but it is quiescent under the experimental conditions examined so far. A functional NHE-1 activity is present in the basolateral membrane of TRCs in the fungiform papillae. The functional characteristics of basolateral NHE-1, as they relate to TRC pHi regulation and its suggested role in sour taste transduction, are presented below.
Preliminary reports of this work have been published in abstract form (Lyall et al. 2000a; Vinnikova et al. 2001).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGMENTSREFERENCES |
|---|
Rat fungiform and circumvallate taste buds were isolated by collagenase treatment (Béhé et al. 1990). Taste buds were aspirated with a micropipette (
100 µm diam) from either fungiform papillae in the anterior tongue or from the circumvallate papilla in the posterior tongue. About 100 taste buds (pooled from 3 rats) from each region of the tongue were individually transferred onto cover slips, avoiding contaminating cells and debris. Total RNA was extracted using the RNeasy Mini kit (Qiagen) according to the manufacturer's instructions. RT-PCR was performed with a One-Step RT-PCR Kit (Qiagen) using NHE-1 and NHE-3 primer sequences and PCR conditions described elsewhere (Borensztein et al. 1995). The PCR products were analyzed by direct sequencing (Davis Sequencing).
Immunocytochemistry
Rats were anesthetized with isoflurane and perfused via the left ventricle with PBS followed by 2% paraformaldehyde solution in PBS for 10 min. Tongues and kidneys were harvested, embedded in paraffin, and sectioned. The immunocytochemical procedure was performed as described previously (Hager et al. 2001) with some modifications. Briefly, 5-µm sections were deparaffinized, rehydrated, and subjected to antigen unmasking using High pH Target Retrieval Solution (Dako) (for NHE-3) or 1% SDS (for NHE-1). Following the blocking procedure, sections were incubated overnight at 4°C with primary antibodies. The monoclonal anti-NHE-1 antibodies were obtained from Chemicon. The polyclonal NHE-3 antibodies were generated in Dr. Mark Knepper's laboratory and have been characterized previously (Fernandez-Llama et al. 1998; Kim et al. 1999). The monoclonal anti-Na+-K+-ATPase antibodies were obtained from Upstate Biotechnology and used as a marker of the basolateral membrane. NHE-1 and Na+-K+-ATPase immunoreactivity was detected using Alexa-488-labeled goat-anti-mouse IgG (Molecular Probes). NHE-3 immunoreactivity was detected using the multistep immunoperoxidase technique with biotinylated anti-rabbit IgG followed by Cy3-labeled antiperoxidase antibodies (Jackson ImmunoResearch Laboratories). Alexa-488 and Cy3 fluoroprobes have nonoverlapping spectra and were used in dual-labeling experiments. Sections were mounted in ProLong Antifade medium (Molecular Probes) and were imaged using a Nikon laser scanning confocal microscope. The images were analyzed using Photoshop software.
pH and Na+ imaging
Rats were anesthetized with isoflurane and killed by cervical dislocation. The tongues were rapidly removed and stored in ice-cold Ringer solution pH 7.4 (R; Table 1). The lingual epithelium was isolated by collagenase treatment (Lyall et al. 1997; Stewart et al. 1998). A small piece of the anterior lingual epithelium containing a single fungiform papilla was mounted in a special microscopy chamber (Chu et al. 1995) as described before (DeSimone et al. 2001b; Lyall et al. 2001, 2002a,b). For the measurement of pHi, TRCs within the taste bud were loaded with BCECF, and for Na+ measurement ([Na+]i), the taste cells were loaded with either 1,3-benzenedi-carboxylic acid, 4,4'-[1,4,10-trioxa-7,13-diazacyclopentadecane-7, 13-diylbis(5-methoxy-6,12-benzofurandiyl]bis-, tetrakis[(acetyloxy) methyl] ester (SBFI) or Na-green. The detailed methods for the measurement of pHi using BCECF (DeSimone et al. 2001b; Lyall et al. 2001, 2002a,b) and the measurement of [Na+]I using SBFI or Na-green have been described earlier (Lyall et al. 2002b). In brief, TRCs in the taste bud were visualized from the basolateral side through a 40x objective (Zeiss; 0.9 NA) with a Zeiss Axioskop 2 plus upright fluorescence microscope and imaged with a set up consisting of: a cooled CCD camera (Imago, TILL Photonics, Applied Scientific Instrumentation, Eugene, OR) attached to an image intensifier (VS4-1845, Videoscope, Washington, DC), an epifluorescent light source (TILL Photonics Polychrome IV), dichroic beam splitters, and emission filters for BCECF and fura-2 (Omega Optical). Small regions of interest (ROIs) within the taste bud (2-3 µm2 diam) were chosen in which the changes in fluorescence of the single wavelength dye (sodium green) or the fluorescence intensity ratio (FIR) for a dual excitation dye (BCECF or SBFI) was analyzed using TILLvisION v3.1 imaging software. Each ROI contained two to three receptor cells. Thus the fluorescence intensity recorded for a ROI represents the mean value from two to three receptor cells within the ROI. In a typical experiment, the FIR measurements were made in an optical plane in the taste bud containing four to six ROIs (12-18 cells). The background and autofluorescence were corrected from images of a taste bud without the dye. Amiloride, 5-(N-methyl-N-isobutyl)-amiloride (MIA), 5-(N-ethyl-N-isopropyl)-amiloride (EIPA), and cariporide (HOE642) exhibit strong UV fluorescence at 380 nm. For this reason, SBFI is not suitable for measuring changes in TRC [Na+]i in the presence of these drugs. Accordingly, these experiments were performed with the single wavelength excitation dye, Na-green (Lyall et al. 2002b).
|
The composition of the solutions used in the in vitro experiments is given in Tables 1 and 2. The control solution was Ringer solution without Na-pyruvate (RC; Table 1) (Lyall et al. 2002a,b). In Na+-free solutions, NaCl was replaced with an equivalent amount of N-methyl-D-glucamine chloride (NMDGCl; R0Na; Table 1). Some solutions contained amiloride, MIA, EIPA (Sigma), the nonspecific blockers, or HOE642 (Aventis Pharma, Germany), a specific blocker of NHE-1 (Scholz et al. 1995). In some experiments, the apical membrane was perfused with a solution without HEPES containing 58.3 mM acetic acid (RAA; pH 3.0; Table 1) or 1 mM HCl (pH 3.0). For the measurement of intrinsic buffering capacity (
1), we used Na+- and Cl--free solutions (R0Na0Cl; Table 2). To block NH4+ flux via the K+ channels, 10 mM tetraethylammonium acetate (TEAA) was added to the solutions. In these experiments, the apical solutions also contained, in addition, 2 mM cetylpyridinium chloride (CPC) to block the NH4+ flux via the amiloride-insensitive CPC-sensitive cation channels in the apical membranes of fungiform TRCs (DeSimone et al. 2001b). Different concentrations of NH4+ were obtained by mixing the R0Na0Cl and R(NH4)2SO4 (Table 2) solutions. At the end of each experiment, TRC pHi was calibrated using the calibrating solutions (CS; pH 6.5-8.0; Table 2) containing 10 µM nigericin.
|
The changes in TRC pHi were expressed as the mean ± SE of n, where n represents the number of ROIs within the taste bud. In TRCs loaded with sodium-green, the changes in [Na+]i were expressed relative to the fluorescence intensity (F490) under control conditions. The F490 under control conditions for each ROI was taken as 100%. For TRCs loaded with SBFI, the relative changes in FIR (F340/F380) were compared between individual ROIs under different experimental conditions. The data were also presented as the mean ± SE from different tissue preparations (N). In this case, N represented the number of individual polarized lingual preparations studied. Student's t-test was employed to analyze the differences between sets of data.
All in vivo and in vitro protocols were approved by the Institutional Animal Care and Use Committee of Virginia Commonwealth University.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGMENTSREFERENCES |
|---|
; Stewart et al. 1998) and polarized (Lyall et al. 2002a,b) TRCs spontaneously recovered their pHi to resting levels. This suggests that the TRC membranes must contain pH compensatory mechanisms that spontaneously restore pHi to resting levels. We therefore hypothesized that the most common epithelial isoforms of NHE, namely apical NHE-3 and basolateral NHE-1, would be present and functional in TRCs. In this study, we first used RT-PCR to detect the presence of NHE mRNA transcripts in fungiform and circumvallate TRCs. Second, we used specific antibodies to localize the presence of NHEs in TRCs membranes; last, we used pH and Na+ imaging to characterize the functional role of NHEs in TRC pHi regulation. Presence of NHE-1 and NHE-3 mRNA in TRCs
Figure 1 shows RT-PCR in fungiform taste buds for NHE-1 in lanes 2-5. Using specific primers for NHE-1 (Borensztein et al. 1995), a 422-bp product of expected size was observed in taste buds (lane 2) and in positive kidney control (lane 4). Figure 1 also shows the RT-PCR for NHE-3 mRNA in lanes 7-10. Using specific primers for NHE-3 (Borensztein et al. 1995), a 321-bp product of expected size and a larger size genomic band was seen (lane 7). In taste buds, in which the reverse transcription step was omitted, only the genomic band was seen (lane 8). In kidney positive control (lane 9), a single 321-bp product was observed (genomic DNA was not amplified). The negative control without template is shown in lane 10. Identity of the PCR products as rat NHE-1 and NHE-3 was confirmed by sequencing. In parallel experiments, we were also able to confirm the presence of NHE-1 and NHE-3 mRNAs in circumvallate taste buds (data not shown).
|
Using specific NHE-3 antibodies and confocal imaging, NHE-3 was found to be present in the apical region (arrow) of the fungiform (FF) TRCs (Fig. 2B; red label). The kidney (K) control (Fig. 2A; red label) demonstrated the specific binding of the NHE-3 antibodies to the brush border of the proximal tubules. The green label shows the basolateral Na+-K+-ATPase. In some sections of fungiform papillae, the NHE-3 was also localized in the intracellular compartment (Fig. 2C; red label). The NHE-3 label was not present in the basolateral membranes of TRCs since NHE-3 did not co-localize with the basolateral Na+-K+-ATPase (Fig. 2C; green label).
|
|
In polarized TRCs initially perfused on both sides with control Ringer solution (RC; Table 1), switching to a Na+-free solution (R0Na; Table 1) in the apical compartment (Fig. 4A, a to b) did not alter resting TRC pHi. In contrast, switching to R0Na in the basolateral compartment (Fig. 4A, b to c) decreased the mean TRC pHi from 7.26 ± 0.03 to 6.70 ± 0.03 (n = 6). Re-perfusing the control Ringer solution (RC) in the basolateral compartment (c to d) promptly increased pHi to its original level. In three preparations, no changes in TRC resting pHi were observed on perfusing Na+-free solution in the apical compartment. In contrast, perfusing the Na+-free solution in the basolateral compartment decreased resting TRC pHi by 0.62 ± 0.03 pH units (P < 0.001; n = 3). These results indicate that TRC pHi is dependent on the Na+ concentration in the basolateral compartment but is independent of apical Na+ concentration.
|
The TRC pHi was monitored during step changes in basolateral NaCl concentration from 0 to 1, 2.5, 5, 10, 20, 50, 100, and 150 mM. Varying concentrations of Na+ were achieved by mixing RC and R0Na in appropriate proportions (Table 1). The data in Fig. 5A show that in a polarized TRC preparation perfused on both sides with Na+-free solution (R0Na; Table 1), successively increasing the basolateral Na+ concentration from 0 to 150 mM, increased TRC pHi in a stepwise manner. The steady-state pHi values at each successive Na+ concentration are plotted in Fig. 5B. The curve was drawn according to a kinetic model of NHE activity (cf. APPENDIX). In the complete absence of Na+, the mean pHi was 6.62 ± 0.003 (n = 4). At a basolateral Na+ concentration of 50 mM, the pHi increased to 7.47 ± 0.005. Increasing the Na+ concentration from 50 to 100 mM and then to 150 mM increased pHi to 7.61 ± 0.005 and 7.65 ± 0.005, respectively. These results indicate that increasing the basolateral Na+ concentration from 0 to 50 mM increased pHi to 82.7% of its resting value under control conditions.
|
pHi/min) achieved its maximum value (-0.17 pH units/min) during a step change in the basolateral Na+ concentration from 150 to 0 mM and decreased to -0.03 pH units/min during a step change in Na+ concentration from 150 to 50 mM. Small but significant changes in pHi occured when the basolateral Na+ concentration was varied from 50 to 150 mM (Fig. 5, A and B), suggesting that the overall resting TRC pHi is determined by several Na+-dependent pH regulatory mechanisms in TRC membranes.
These results suggest that a Na+-dependent acid extrusion mechanism is present in the basolateral membranes of fungiform TRCs. In the apical membranes of TRCs, a Na+-dependent acid extrusion mechanism seems to be either absent or is quiescent under our experimental conditions. To test the possibility that this mechanism is a Na+-H+ exchanger, further experiments were performed in the presence of basolateral NHE blockers, amiloride, MIA, EIPA, and HOE642 (Scholz et al. 1995).
Effect of NHE blockers on TRC pHi
The data presented in Fig. 4A also show that perfusing the basolateral membrane of polarized TRCs with control Ringer solution (RC; Table 1) containing 1 µM HOE642, a specific blocker of the NHE-1, decreased resting TRC pHi (d to e) from 7.26 ± 0.01 to 7.15 ± 0.03 (
pHi = -0.11 ± 0.002; P < 0.01; n = 6). In the continuous presence of HOE642, lowering the basolateral NaCl concentration from 150 to 0 mM (switching to R0Na + 1 µM HOE642; Table 1) decreased TRC pHi from 7.15 ± 0.03 to 6.91 ± 0.02. Thus in the presence of the 1 µM HOE642, the magnitude of TRC pHi (pHi = -0.24; e to f) was decreased by approximately 59% relative to control (pHi = -0.58; b to c). The initial rates of changes in TRC pHi were measured for the first 100 s following a change in basolateral Na+ concentration. In the absence of HOE642, decreasing the basolateral Na+ concentration from 150 to 0 mM decreased TRC pHi at the mean rate of -0.065 ± 0.009 pH units/min (Fig. 4A; b to c), and on re-perfusing with control Ringer solution containing 150 mM NaCl, increased pHi at the mean rate of 0.227 ± 0.017 pH units/min (Fig. 4A; c to d; n = 6). In contrast, in the presence of HOE642, the corresponding mean rates of changes in pHi on Na+ removal (e to f) and re-addition (f to g) were -0.021 ± 0.003 (67.7% inhibition) and 0.014 ± 0.003 pH units/min (93.8% inhibition), respectively. Perfusing control Ringer solution (RC) without HOE642 promptly increased TRC pHi to its resting level (g to h). In the final step, re-perfusing the control Ringer solution (RC) in the apical compartment did not produce any further changes in TRC pHi (h and i). In additional experiments, in polarized TRCs perfused on both sides with the Na+-free solution, perfusing the control Ringer solution in the apical compartment also had no effect on TRC pHi (data not shown). In an additional experiment, perfusing the basolateral membrane of polarized TRCs with Na+-free Ringer solution decreased TRC pHi from 7.30 ± 0.01 to 6.79 ± 0.01 (Fig. 4B; j to k to l; mean pHi = 0.51). In the presence of 10 µM HOE642, perfusing 0 Na+ Ringer solution in the basolateral compartment decreased TRC pHi from 7.21 ± 0.01 to 7.10 ± 0.01 (Fig. 4B; m to n; mean pHi = 0.11). Thus in the presence of 10 µM HOE642, the magnitude of TRC pHi was decreased by approximately 78.4% relative to control.
Similar effects were also observed with amiloride, a nonspecific blocker of the NHEs. In the absence of amiloride, decreasing basolateral Na+ concentration from 150 (RC) to 0 mM (R0Na) decreased the mean resting TRC pHi from 7.23 ± 0.02 to 6.55 ± 0.01 at a rate of 0.108 ± 0.002 pH units/min (n = 5). On reintroducing 150 mM NaCl solution in the basolateral compartment, pHi increased at a rate of 0.337 ± 0.004 pH units/min. In the presence of basolateral amiloride (1 mM), the resting TRC pHi decreased from 7.30 ± 0.01 to 7.26 ± 0.01 (pHi = -0.04 ± 0.002; P < 0.001, n = 6). In the continuous presence of amiloride, lowering basolateral Na+ concentration from 150 to 0 mM decreased TRC pHi at a mean rate of 0.063 ± 0.004 pH units/min (40% inhibition). Reintroducing 150 mM NaCl solution in the basolateral compartment increased pHi at a rate of 0.029 ± 0.002 pH units/min (90% inhibition). These rates are significantly less than the corresponding rates in the absence of amiloride (P < 0.001; paired; n = 5). Taken together, the data suggest the presence of a HOE642-sensitive Na+-H+ exchange mechanism in the basolateral membranes of TRCs. In the next series of experiments, we investigated the involvement of this exchanger in the regulation of TRC pHi.
Effect of intracellular acid loading on TRC pHi
To investigate if the basolateral Na+-H+ exchange activity is involved in TRC pHi regulation, we monitored the rate of spontaneous pHi recovery following intracellular acid loading with NH4Cl or Na-acetate at constant pHo (Lyall et al. 2002a,b).
Studies with NH4Cl
Figure 6 shows the effect of a short basolateral NH4Cl pulse on TRC pHi. Immediately following NH4Cl perfusion (RNH4Cl; Table 1), TRC pHi rapidly alkalinized (a to b), presumably due to the entry of NH3 and the conversion of free intracellular H+ ions to NH4+ ions (Roos and Boron 1981). This was followed by a slow decline of pHi toward baseline (b to c), presumably reflecting NH4+ entry or pH compensation mechanism(s) in TRC membranes (Roos and Boron 1981). On replacing NH4Cl solution with a Na+-free solution (R0Na; Table 1) in the basolateral compartment, TRC pHi acidified (c to d) and became lower than its resting value due to the combined effect of rapid NH3 exit from the cells, the conversion of NH4+ to NH3 + free H+ ions, and the removal of basolateral Na+ (cf. Fig. 4). No spontaneous recovery of pHi was observed in the absence of basolateral Na+ (Fig. 6; d to e). Perfusing the basolateral compartment with control Ringer solution (RC; Table 1) promptly increased TRC pHi to its resting value (Fig. 6; e to f). The spontaneous pHi recovery rate following an NH4Cl pulse was dependent on the basolateral Na+ concentration (Fig. 7). Increasing the Na+ concentration in a stepwise manner from 0 to 150 mM caused the spontaneous pHi recovery rate to increase as a saturating function of [Na+]Bl (Km = 8.3 mM). This is consistent with the predicted kinetics of NHE (Grinstein and Rothstein 1986; APPENDIX).
|
|
HOE642 inhibited the spontaneous rate of pHi recovery after the NH4Cl pulse in a dose-dependent manner (Fig. 8; open circles and dotted line). The data were fit to a Michaelis-Menten-type equation. In three independent polarized TRC preparations, the mean Ki value (the concentration that inhibits the spontaneous rate of pHi recovery by 50%) for HOE642 was 0.23 µM (range, 0.14-0.40 µM). In separate experiments, we also tested the inhibition of NHE-1 activity by MIA (filled triangles and dash-dot line), EIPA (filled squares and short dash line), and amiloride (filled circles and solid line). In three taste buds, the mean Ki for HOE642, MIA, EIPA, and amiloride was 0.23, 0.46, 0.84, and 29 µM, respectively (Fig. 8).
|
Figure 9 shows the effect of a short basolateral side Naacetate (NaA) pulse on TRC pHi. Consistent with our previous observations (Lyall et al. 2002b), immediately following the perfusion of Ringer solution containing 30 mM NaA (RNaA; Table 1) in the basolateral compartment, pHi rapidly acidified (a to b), presumably due to the entry of membrane-permeable undissociated acetic acid and its conversion to free intracellular H+ ions plus acetate anion (Roos and Boron 1981). The intracellular acidification was transient and was followed by a spontaneous recovery of pHi to baseline (b to c). On Na-acetate washout (RC; Table 1), TRC pHi alkalinized and became higher than its resting value (c to d). This is due to the rapid exit of the undissociated acetic acid from cells and a decrease in intracellular H+ ions. The spontaneous recovery of alkaline pHi toward baseline (d to e) reflects the presence of as yet an unknown pH recovery mechanism in TRCs that allows base (OH-) exit or entry of acid equivalents at alkaline pHi. In the presence of 10 µM HOE642, the spontaneous mean initial pHi recovery rate (f to g; 0.009 ± 0.003 pH units/min) was significantly reduced (81.2% inhibition) compared with control (b to c; 0.048 ± 0.004 pH units/min; n = 5). However, HOE642 did not affect the exit of the undissociated acetic acid from cells (g to h) and the subsequent spontaneous recovery of alkaline pHi toward baseline (h to i). Similar results were obtained in two additional experiments (data not shown). The results indicate that HOE642 specifically inhibits pHi recovery from an intracellular acid load and does not affect mechanisms involved in base (OH-) exit or entry of acid equivalents at alkaline pHi. The results further suggest that the basolateral NHE-1 is involved in the regulation of TRC pHi.
|
Relationship between TRC pHi and [Na+]i
Figure 10A shows the effect of external Na+ concentration on the relative changes in TRC [Na+]i in a polarized fungiform taste bud preparation loaded with Na-green. On lowering the apical Na+ concentration from 150 (RC) to 0 mM (R0Na), there was a decrease in resting F490 (a to b), indicating a decrease in TRC [Na+]i. In the second step, lowering the basolateral Na+ concentration from 150 to 0 mM further decreased F490 (b to c). These changes in TRC [Na+]i were reversed on addition of external Na+ to the basolateral (c to d) and apical (d to e) compartments, respectively. The fluorescence of single wavelength dyes can be affected by changes in cell volume, dye leakage, dye bleaching, and changes in focus due to tissue movement. Measurements using a ratiometric dye are not affected by the above factors. Therefore we also performed some experiments with the dual excitation, single emission dye, SBFI. In polarized TRCs loaded with SBFI, the temporal changes in [Na+]i were measured as changes in FIR (F340/F380). Consistent with the data shown in Fig. 10A, a decrease in Na+ concentration from 150 (RC) to 0 mM (R0Na; Table 1) in the basolateral compartment decreased FIR (Fig. 10B; a to b), indicating a decrease in TRC [Na+]i. On re-perfusing the control Ringer solution (RC; Table 1) in the basolateral compartment, the FIR promptly increased to its original value (b to c).
|
) shown that apical Na+ flux occurs, in part, via the amiloride (or benzamil)-sensitive epithelial Na+ channels (ENaCs) and by an amiloride (or benzamil)-insensitive pathway, which is modulated by cetylpyridinium chloride (CPC) (DeSimone et al. 2001b). In contrast, the basolateral Na+ flux is accompanied by changes in TRC pHi (Fig. 4A, b to c vs. Fig. 10A, b to c). To investigate if the basolateral Na+ flux is coupled to changes in TRC pHi via the basolateral Na+-H+ exchanger activity, further experiments were done in the presence of amiloride and HOE642. Data summarized in Fig. 11A show that lowering basolateral Na+ concentration from 150 (RC; Table 1) to 0 mM (R0Na; Table 1) reversibly decreases F490 (a to b to c), and 250 µM amiloride in the basolateral compartment attenuated the changes in F490 induced by Na+ removal (d to e to f) by 50%. The initial mean rate of change in F490 was measured for the first min following the basolateral Na+ removal. In three individual taste buds, under control conditions, the F490 declined at the rate of 20.4 ± 1.7%/min, and this rate decreased to 6.3 ± 1.9%/min in the presence of 250 µM amiloride (P < 0.01; paired; n = 3; n = 21). Amiloride also inhibited the initial mean rate of increase in F490 induced by changing the basolateral Na+ concentration from 0 to 150 mM by 39.3 ± 3.4% (P < 0.01; paired). Similar results were obtained with 10 µM HOE642 (Fig. 11B). As shown in Fig. 4, HOE642 (and amiloride) inhibited changes in TRC pHi under identical conditions. Thus both Na+ and H+ fluxes across the basolateral membrane are inhibited by HOE642 and amiloride. Taken together, these results support the presence of a NHE-1 in the basolateral membranes of TRCs and indicate that a part of the Na+ flux across the basolateral membrane occurs via this mechanism.
|
1) of TRCs
We used the pHi decrease caused by the washout of NH4+ to compute TRC 1 (Boyarsky et al. 1988; Vaughan-Jones and Wu 1990). Since, at the present time, many of the pH regulatory mechanisms in TRCs have not been characterized, these experiments were done under the conditions in which all acidbase transport mechanisms were inhibited by ion substitution. The experiments were done in nominally CO2/HCO3--free, Na+-free, and Cl--free solutions to block Na+-, Cl--, and HCO3--dependent pH regulatory mechanisms in TRC cell membranes. The NH4+ flux via the apical and basolateral K+ channels was blocked by the addition of 10 mM TEAA. The NH4+ flux via the apical amiloride-insensitive, CPC-sensitive cation pathway was inhibited by the addition of 2 mM CPC in the apical solution (R0Na0Cl; Table 2). The 1 (mM/pH unit) was calculated as ([H+]i/pHi), where [H+]i is the amount (in mM) of acid introduced into the cell and pHi is the resultant change in pHi. [H+]i is assumed to equal the intracellular concentration of NH4+ ([NH4+]i) at the moment of its removal from the external solution (Vaughan-Jones and Wu 1990) and is given by [NH4+]i = [NH4+]o x 10(pHo-pHi), where [NH4+]o = C/[10(pHo-pK) + 1]. Here C = the total concentration of external NH4+ and pK = 9.02.
In the example shown in Fig. 12A, lowering total NH4+-NH3 from 30 to 20 mM caused the mean TRC pHi to decrease in 16 ROIs within a fungiform taste bud from 7.51 to 7.38 (pHi = 0.13). We calculate that, at 30 mM basolateral NH4+ concentration, the mean [NH4+]i was 22.4, and at 20 mM basolateral NH4+ concentration, the mean [NH4+]i was 20.6 ([NH4+]i = 1.8). Thus the average 1 in the small pH range during a change in the basolateral NH4+ concentration from 30 to 20 mM (average pHi 7.44) was (1.8/0.13) =13.8 mM/pH unit. In a similar manner, subsequently decreasing the NH4+-NH3 concentration to 10 and 5 mM gave mean 1 value of 24.7 and 26.9 at the average pHi values of 7.28 and 7.10, respectively. Figure 12A, inset, shows the relationship between mean TRC pHi and 1 from three individual taste buds containing 30 ROIs. The line of best fit (r2 = 0.92; n = 30) gave a mean slope of -36.9 (mM/pH)/pH.
|
1 value of 18.7 ± 1.2 mM/pH unit at the mean TRC pHi of 7.38 ± 0.02. In the 22 ROIs studied, the mean resting pHi value varied from 7.27 to 7.61, and the 1 value ranged between 30.7 and 7.6 mM/pH unit. This indicates that there are significant variations in pHi and 1 value among individual TRCs within the taste bud. It is likely that TRCs with lower 1 values respond with a greater decrease in pHi to an acid stimulus and hence participate most in sour taste transduction. The results further indicate that 1 increases with a decrease in TRC pHi. The 1 value was used to calculate the net efflux of acid or the acid-extrusion rate (JH+; mM/min) as the product of 1 and the rate of pHi recovery in TRCs. The 1 values reported here and their dependence on TRC pHi are comparable to those reported in type 1 carotid body cells, which are involved in CO2/H+ sensing for arterial blood (Buckler et al. 1991b).
In Fig. 12B, the resting pHi values under control conditions from 78 ROIs in seven individual polarized fungiform TRCs preparations were plotted against the number of ROIs that fall within a given pHi value. The data demonstrate that two distinct subpopulations of TRCs can be separated based on the initial value of resting pHi. The first group demonstrates a normal distribution with a mean pHi of around 7.2. The second group demonstrates a skewed distribution with a mean pHi around 7.45. Similarly, two subpopulations of TRCs were distinguished by their responses to acid-induced changes in intracellular Ca2+ activity (Liu and Simon 2001; Richter et al. 2003). These results further indicate that there are significant variations in resting pHi values among individual TRCs within the taste buds. The TRCs having relatively more alkaline resting pHi values, and consequently lower 1 values, will produce larger changes in pHi when stimulated by acids. On this basis, this cell population might be expected to show higher sensitivity to acids and therefore may be the more important cell population for sour taste transduction.
Effect of pHo on the basolateral NHE-1 activity
Our results (Figs. 6 and 9) indicate that a decrease in TRC pHi activates the basolateral NHE-1. To investigate the effect of pHo on the basolateral NHE-1 activity, we monitored the spontaneous rate of pHi recovery from short NH4Cl pulses at basolateral pH (pHBl) of 7.8, 7.4, and 6.8. The pHi recovery rate was measured for the first 2 min following the washout of the NH4Cl pulse. In a representative experiment (Fig. 13), unilaterally increasing pHBl from 7.4 to 7.8, increased pHi from 7.04 ± 0.02 to 7.31 ± 0.02 (e to f) and increased the initial spontaneous pHi recovery rate from 0.060 ± 0.002 (d to e) to 0.076 ± 0.004 (i to j; n = 7). Lowering the basolateral pH from 7.8 to 6.8, decreased pHi from 7.29 ± 0.03 to 6.75 ± 0.01 (j to k) and attenuated the spontaneous rate of pHi recovery from 0.076 ± 0.004 (i to j) to 0.019 ± 0.002 (n to o). In the final step, raising the basolateral pH to 7.4 increased TRC pHi from 6.75 ± 0.01 to 6.96 ± 0.02 (o to p). Using the mean 1 values of 28.8 and 39.5 (cf. Fig. 12A, inset) at the mean TRC pHi values of 7.04 and 6.75, respectively, one can calculate the change in the net mean flux of H+ ions (JH+) at the two pHi values. Thus a decrease in mean TRC pHi from 7.04 to 6.75 (j to k) decreased the mean JH+ from 1.73 ± 0.06 to 0.75 ± 0.08 mM/min (P < 0.01; n = 7; paired). In three tissue preparations containing 21 ROIs (n = 3; n = 21), a decrease in basolateral pH from 7.4 to 6.8 decreased the mean resting pHi by 0.38 ± 0.04 (P < 0.025) and decreased the mean JH+ by 47.3 ± 1.4% (P < 0.001; paired). This decrease in JH+ represents the decrease in H+ flux due to the inhibition of the basolateral NHE-1. The mean data from four such experiments are plotted in Fig. 14.
|
|
), there was a linear relation between mean steady-state TRC pHi (at points a, f, j, k, and p) and pHBl (Fig. 14; 
). The line with slope 0.85 is the predicted relation between pHi and pHo based on NHE kinetics (see APPENDIX). NHE kinetics also predicted that the mean spontaneous rate of pHi recovery is an increasing nonlinear function of pHBl. The mean rates of recovery represented by d to e, i to j, and n to o from four experiments are plotted in Fig. 14 as a function of pHBl (Fig. 14, 
). The curve is a least-squares fit of the data according to the model developed in the APPENDIX. The results indicate that, unlike changes in TRC pHi, changes in pHo have an opposite effect on NHE-1 activity. External acidification inhibits while external alkalinization increases the activity of basolateral NHE-1 activity.
Data presented in Fig. 15 show that lowering the apical pH (pHAp) from 7.4 to 3.0 (with acetic acid) decreased resting TRC pHi (a to b) from 7.20 ± 0.03 to 6.96 ± 0.03 (pHi = -0.24 ± 0.007 pH units; P < 0.001; n = 5). In the presence of apical acetic acid, the spontaneous initial pHi recovery rate (measured during the 1st 100 s) from an NH4Cl pulse was 0.025 ± 0.003 (pHi/min; c to d). Returning the pHAp back to 7.4 increased TRC pHi (d to e) from 6.87 ± 0.03 to 7.04 ± 0.03 (pHi = 0.17 ± 0.003 pH units; P < 0.001; n = 5) and increased pHi recovery rate (pHi/min; f to g) to 0.041 ± 0.002. Thus decreasing apical pH from 7.4 to 3.0 inhibited the initial pHi recovery rate by 39.0 ± 5.4% (P < 0.01; paired; n = 5). Using the mean 1 values of 35.1 and 28.8 (Fig. 12A; inset) at the mean TRC pHi of 6.91 and 7.04, respectively, one can calculate the change in the net mean flux of H+ ions (JH+) at the two pHi values. Thus in the presence of apical acetic acid (pHi = 6.87 ± 0.03), the mean JH+ was 0.88 ± 0.10 mM/min. On acetic acid washout, the mean pHi increased to 7.04 ± 0.03 and the mean JH+ increased to 1.20 ± 0.05 mM/min (P < 0.05; paired; n = 5). In three polarized fungiform taste bud preparations containing 22 ROIs, stimulating the apical membrane with acetic acid (pH 3) produced variable responses in individual ROIs (Lyall et al. 2001). The acetic acid-induced decrease in pHi ranged from 0.05 to 0.30 pH units, with a mean of 0.17 ± 0.015 (SE; n = 22). In the same ROIs, the decrease in JH+ ranged from 0.03 to 0.69 mM/pH unit (0.31 ± 0.04; mM/pH unit). Similar results were obtained when the apical membrane was stimulated with HCl solution at pH 3.0 (data not shown). The decrease in JH+ in the presence of apical acetic acid represents the decrease in H+ flux due to the inhibition of the basolateral NHE-1. Taken together, the results suggest that the NHE-1 activity is inhibited by pHo. The results further indicate that taste buds are comprised of a heterogeneous population of TRCs that demonstrate variable responses to acid stimuli. Variable responses to acid stimuli may arise due to the differences in resting pHi, 1, and NHE-1 activity in individual TRCs. Overall, the data support the hypothesis that only a subset of TRCs act as acid-sensing cells that participate in sour taste transduction.
|
pHi/pHAp) was 0.055, a value 10-fold less than that observed across the basolateral membrane (Figs. 13 and 14). The results suggest that the apical pore region and the paracellular pathway between TRCs present a significant diffusion barrier to acids (Lyall et al. 2001). Lowering pHAp decreases resting TRC pHi and inhibits pHi recovery rate (Grinstein and Rothstein 1986; Vaughan-Jones and Wu 1990), suggesting that, during apical acid stimulation, the pHi recovery by the basolateral NHE-1 is blunted. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGMENTSREFERENCES |
|---|
). NHEs mediate electro-neutral exchange of Na+ for H+ and thereby play a central role in pH regulation and Na+ homeostasis. Our RT-PCR studies provide the first molecular evidence that both NHE-1 and NHE-3 messages are expressed in TRCs (Fig. 1). Using immunocytochemical methods and confocal microscopy, the NHE-3 and NHE-1 were localized to the apical and basolateral membranes of TRCs, respectively (Figs. 2 and 3). NHE-1 has been shown to be present in the basolateral membranes of most epithelial cells (Noel and Pouysségur 1995). The basal NHE-1 activity serves a housekeeping function in maintaining resting
), and MIA (
). In each case, rate of spontaneous pHi recovery in the absence of drug was taken as 100%. Spontaneous pHi recovery rates are presented as means ± SE of N, where N (number of lingual epithelial preparations) = 3.
