The Journal of Neurophysiology Vol. 79 No. 2 February 1998, pp. 911-921
Copyright ©1998 by the American Physiological Society

Self-Inhibition in Ca²⁺-Evoked Taste Responses: A Novel Tool for Functional Dissection of Salt Taste Transduction Mechanisms

Mamoun A. Kloub, Gerard L. Heck, and John A. Desimone

Department of Physiology, Medical College of Virginia, Virginia Commonwealth University, Richmond, VA 23298-0551

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Kloub, Mamoun A., Gerard L. Heck, and John A. DeSimone. Self-inhibition in Ca²⁺-evoked taste receptors: a novel tool for functional dissection of salt taste transduction mechanisms. J. Neurophysiol. 79: 911-921, 1998. Rat chorda tympani (CT) responses to CaCl₂ were obtained during simultaneous current and voltage clamping of the lingual receptive field. Unlike most other salts, CaCl₂ induced negatively directed transepithelial potentials and gave CT responses that were auto-inhibitory beyond a critical concentration. CT responses increased in a dose-dependent manner to ~0.3 M, whereafter they decreased with increasing concentration. At concentrations where Ca²⁺ was self-inhibitory, it also inhibited responses to NaCl, KCl, and NH₄Cl present in mixtures with CaCl₂. Ca²⁺ completely blocked the amiloride-insensitive component of the NaCl CT response, the entire KCl-evoked CT response, and the high-concentration-domain CT responses of NH₄Cl (0.3 M). The overlapping Ca²⁺-sensitivity between the responses of the three Cl salts (Na⁺, K⁺, and NH⁺₄) suggests a common, Ca²⁺-sensitive, transduction pathway. Extracellular Ca²⁺ has been shown to modulate the paracellular pathways in different epithelial cell lines by decreasing the water permeability and cation conductance of tight junctions. Ca²⁺-induced modulation of tight junctions is associated with Ca²⁺ binding to fixed negative sites. This results in a conversion of ion selectivity from cationic to anionic, which we also observed in our system through simultaneous monitoring of the transepithelial potential during CT recording. The data indicate the paracellular pathway as the stimulatory and modulatory site of CaCl₂ taste responses. In addition, they indicate that important transduction sites for NaCl, KCl, and NH₄Cl taste reception are accessible only through the paracellular pathways. More generally, they show that modulation of paracellular transport by Ca²⁺ in an intact epithelium has functional consequences at a systemic level.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Taste buds and ion-transporting epithelia share common topology, including polarized receptor cells with apical and basolateral domains separated by tight junctions (TJs) (Holland et al. 1989). The early events in the detection of Na⁺ salts in the rat peripheral taste system also share similarities with other ion-transporting epithelia. The Na⁺-detecting elements include an amiloride and voltage-sensitive transduction site (Avenet and Lindemann 1988; DeSimone et al. 1981; Heck et al. 1984; Schiffman et al. 1983; Ye et al. 1993). This suggests that one transducer is an apical membrane ion channel, similar to that involved in epithelial sodium transport (Smith and Benos 1991). However, amiloride, does not suppress the entire chorda tympani (CT) response to NaCl in rats, even at high concentration (Brand et al. 1985; DeSimone and Ferrell 1985; Elliot and Simon 1990; Formaker and Hill 1988; Lundy and Contreras 1997). Evidence suggests that the amiloride-insensitive component (AIC) of the NaCl neural response arises from transduction sites along the basolateral membranes of receptor cells, access to which is assumed to be through paracellular pathways (Elliot and Simon 1990; Mierson et al. 1996; Simon et al. 1993; Stewart et al. 1995; Ye et al. 1993). The principal barriers in these pathways are the TJs that connect the apical poles of the taste receptor cells, which act as weakly cation-selective barriers (DeSimone et al. 1984; Simon and Garvin 1985; Ye et al. 1993). In addition, these pathways likely participate in K⁺ and NH⁺₄ salt taste responses (Kloub et al. 1997; Ye at al. 1994). If so, transduction sites for NaCl, KCl, and NH₄Cl may be accessible only by means of a common pathway across the TJ complex. However, until now, a reliable means of probing paracellular pathways, involved in taste reception, has been unavailable.

Calcium modulation of TJs has been documented in various epithelia. Removing it increases the permeability of the rat intestine (Tidball 1964), opens the junctional complex between the oxyntic cells (Sedar and Forte 1964) and pancreatic acinar cells (Meldolesi et al. 1978), and produces fragmentation of the TJs in mammary glands (Pitelka et al. 1983). Removal of Ca²⁺ from the medium of Madin-Darby canine monolayer cells opens their TJs, and its subsequent restoration causes them to reseal (Martinez-Palmo et al. 1980). Ca²⁺ triggers the sealing of TJs at a critical concentration by acting on an extracellular site (Contreras et al. 1991). TJ permeability has been correlated with changes in transepithelial conductance (TEC) and extracellular Ca²⁺ concentration is inversely proportional to TEC (Contreras et al. 1991). These observations suggest that Ca²⁺ can be used to probe the common paracellular pathways believed to be involved in part of the CT response to NaCl and NH₄Cl and virtually all of that to KCl. There are suggestions that Ca²⁺ effects on TJs can modulate other intact epithelia (Mooseker 1985), but this has not yet been demonstrated. Our results from an intact taste sensory system verify this speculation.

At concentrations where Ca²⁺ is self-inhibitory, it also inhibits responses to NaCl, KCl, and NH₄Cl in mixtures with CaCl₂. Moreover, transepithelial potential (TEP) recordings, made simultaneously with the CT responses, confirm that Ca²⁺ converts the paracellular region from cation- to anion-selective, consistent with observations on TEP in vitro in other epithelia (Moreno and Diamond 1974; Prather and Wright 1969; Smyth and Wright 1966). The data and their interpretation using electrodiffusion theory show that Ca²⁺ blocks transduction sites for Na⁺, K⁺, and NH⁺₄ taste reception by reducing the permeability of the paracellular pathway to cations in a highly cooperative fashion.

	METHODS

Abstract Introduction Methods Results Discussion References

Solutions and chemicals

Stimulus salts included NaCl, NH₄Cl (Mallinckrodt Chemical, Paris, KY) KCl, and CaCl₂ and amiloride hydrochloride (Sigma Chemical, St. Louis, MO). All chemicals were reagent grade and were prepared in distilled water. A rinse consisting of 15 mM KHCO₃ + 15 mM KCl (pH 8.3) was applied for 1 min before and after each test stimulus. NaCl-depleted Krebs-Henseleit buffer was applied periodically to the tongue as an artificial saliva (cf. Ye et al. 1994).

Nerve preparation and voltage-clamp recording

Neural responses were obtained from the CT nerves of female Sprague-Dawley rats (180-240 g) during chemical stimulation of the tongue, according to the protocols approved by the Virginia Commonwealth University Animal Research Committee. The surgical procedure has been described in detail (Ye et al. 1993, 1994). Rats were preanesthetized with ether and then given an intraperitoneal injection of pentobarbital sodium (65 mg/kg) with additional injections as needed. The trachea was cannulated, the head immobilized, and the left chorda tympani nerve was exposed, cut, and placed on a platinum electrode. Petroleum jelly was placed around the CT, and a platinum reference electrode was positioned nearby. A stimulation chamber was held on the anterior tongue by vacuum. Solutions were injected in 3-ml aliquots at 1 ml/s and remained in the chamber for 1 or 2 min. The whole CT neural activity was analyzed and displayed as described previously (Ye et al. 1993). Transepithelial voltage or current clamp was maintained with a four-electrode voltage clamp amplifier as described elsewhere (Ye et al. 1993). A periodic (15 s) biphasic pulse of 1 µA (current clamp) or 20 mV (voltage clamp) was generated to measure the transepithelial resistance.

Data analysis

Integrated CT responses were analyzed off-line as previously described (Ye et al. 1993). The area under an integrated CT response curve for 1 min from the onset of neural activity was used as its numerical value. All stimulus series were bracketed by application of 0.1 M NaCl. and their CT responses were included only when bracketing NaCl responses varied by <20%. All responses for a given animal were normalized to those of 0.1 M NaCl. The normalized TEC for CaCl₂ was expressed as the conductance per unit ionic strength. Phasic CT responses were not analyzed separately because of stimulus flow rate sensitivity (Heck and Erickson 1973; Smith and Bealer 1975). Numerical results are expressed as the means ± SE. Statistical significance was determined by paired Student's t-test.

	RESULTS

Abstract Introduction Methods Results Discussion References

CaCl₂ transepithelial potential and CT responses

Figure 1 (top) shows TEP changes while recording the CT responses to CaCl₂ at zero current clamp. The TEP evoked by NaCl showed an electropositive increase on the submucosal side because paracellular regions normally behave as leaky cation-exchangers (Ye et al. 1993, 1994). In contrast, the TEPs evoked by CaCl₂ increased in the electronegative direction, evidence of a reversal in ion exchange selectivity in favor of anions (cf. Fig. 2C). Figure 1 also shows CT recordings to NaCl before and after a series of responses to three concentrations of CaCl₂ under zero current clamp. CaCl₂ CT responses are unusual in displaying strong self-inhibition: <0.3 M, CT responses to CaCl₂ increase in a dose-dependent manner, however, >0.3 M, responses decline significantly. This is established further by the complete concentration-response relations, under zero current clamp, shown in Fig. 2A. Recording the CT concentration-response relation for CaCl₂ with the lingual receptive field under voltage clamp (Fig. 2A) preserves the fundamental shape of the curve but shifts it to the right on the concentration axis at V = +50 mV and to the left at V = 50 mV. This suggests that the effect of voltage is mainly to increase (negative clamp voltage) or decrease (positive clamp voltage) the Ca²⁺ concentration at sites at the main permeability barriers. Monitoring changes in TEC along with the CT response provides some insight into the origin of the unusual CT response changes. Figure 2B shows a plot of the normalized conductance as a function of CaCl₂ concentration. As concentration increased, the normalized CaCl₂ TECs decreased. Between 0.1 and 0.3 M CaCl₂ there was a rapid drop in the TECs, whereafter (>0.3 M) a smaller drop was observed. These data demonstrate that Ca²⁺ binding causes a reduction in paracellular CaCl₂ conductance. (see Analysis of self-inhibition in Ca²⁺-evoked taste responses and DISCUSSION).

View larger version (27K): [in this window] [in a new window]   FIG. 1. Simultaneously recorded transepithelial potentials (TEPs; top) and the integrated chorda tympani (CT) responses (bottom) to 0.1 M NaCl and 0.1, 0.3, and 0.5 M CaCl2 under 0 current clamp. A 1-µA bipolar current pulse was passed periodically through the tissue patch (period, 15 s) for measurement of transepithelial conductance (TEC). Current pulse also produces a transient neural response superimposed on the chemically evoked response. Note that TEPs evoked by CaCl2 increased in the electronegative direction, whereas the NaCl-evoked potential showed an increase in the electropositive direction (typical salt pattern). Note also that the 0.5 M CaCl2 CT response is smaller than that of either 0.1 or 0.3 M CaCl2.

View larger version (17K): [in this window] [in a new window]   View larger version (10K): [in this window] [in a new window]   FIG. 2. A: normalized CT responses over CaCl2 concentration ranges from 0.1 to 0.6 M under 0 current clamp (), 50 mV (), +50 mV () voltage clamp. Each point represents the mean ± SE (n = 6, except for 0.6 M CaCl2 where n = 3). Each experiment consisted of a concentration series for CaCl2. Order of presentation was varied. Note that the concentration-response relations exhibit an auto-inhibition profile at critical CaCl2 concentrations. Note also that the main effect of the applied voltage on the CT response-concentration curve is a displacement to the right along the concentration axis at V = +50 mV and to the left at V = 50 mV. Theoretical lines were obtained by using Eq. 16 (see Fitting the data). Fit was obtained with the following values of parameters: a = 26.8, b = 8.64, k = 0.2 M, n = 4. Additional parameter, , was added under voltage-clamp condition:  = 0.12, at 50 mV and  = 0.11, at +50 mV. B: normalized conductances at 0 current clamp as a function of CaCl2 concentration. Each point represents the mean ± SE (n = 6). Note that between 0.1 and 0.3 M, there is a rapid drop in the TEC, whereas a slower drop in the TEC is observed at CaCl2 concentrations >0.3 M. Theoretical line was obtained using Eq. 21 (see Interpretation of conductance data). Best fit of the data was obtained with A = 0.14, B = 0.067. C: normalized TEP for CaCl2 (relative to that of 0.1 M NaCl) as a function of log[CaCl2] (M). Each point represents the mean ± SE (n = 6). Potentials increase in the negative direction with increasing CaCl2 concentration as a result of Ca2+-induced reduction in Ca2+ permeability. Theoretical line was obtained using Eq. 22 (see Transepithelial potential). Best fit of the data was obtained with A1 = 0.76, B1 = 2.54.

Effect of Ca²⁺ on NaCl CT responses and TEPs

Figure 3 shows the effect of CaCl₂ on the CT response to 0.3 M NaCl. The NaCl CT response displayed the usual electropositive-going TEP. In mixture with 0.1 M CaCl₂, the TEP for NaCl, was electropositive but clearly reduced, indicating diminishing relative cation conductance, but the CT response was not significantly reduced. However, given that 0.1 M CaCl₂ alone presents a CT response (cf. Fig. 1), a significant suppression in overall CT response, therefore, occurred. With 0.1 M CaCl₂, the NaCl TEC showed no further increase even though the ionic strength had doubled, so the conductance per unit ionic strength declined by 50%. Similar to CaCl₂ CT responses, the NaCl + CaCl₂ mixture suppression then was correlated with a large drop in TEC. In mixture with 0.2 M CaCl₂ and higher concentrations, the TEP for NaCl became electronegative, the TEC continued to drop but more slowly, and the corresponding CT responses showed obvious suppression. Figure 4 illustrates CaCl₂-induced CT response suppression for both 0.1 and 0.3 M NaCl. In each case, suppression became pronounced beyond 0.2 M CaCl₂. At 0.3 M CaCl₂, the response to 0.1 M NaCl was suppressed by 8% and that to 0.3 M NaCl by 47%, figures approximating the AICs of the responses to 0.1 and 0.3 M NaCl, respectively (Ye et al. 1993). Nearly a constant difference between the mixture responses and that due to CaCl₂ remained at all CaCl₂ concentrations >0.3 M. If CaCl₂ blocked access to paracellular transduction sites (presumed locus of the AIC for NaCl), the difference remaining ought to be suppressed entirely by amiloride at all CaCl₂ concentrations >0.3 M because it must arise exclusively from apical Na-channel transduction sites.

View larger version (29K): [in this window] [in a new window]   FIG. 3. Simultaneously recorded TEPs (top) and the integrated CT responses (bottom) to 0.3 M NaCl and 0.3 M NaCl mixed with 0.1, 0.2, 0.3, and 0.4 M CaCl2, under 0 current clamp. First arrow from the left indicates the application of 0.3 M NaCl, which gives the typical electropositive change in TEP. 0.3 M NaCl + 0.1 M CaCl2 (2nd arrow) shows a reduced electropositive TEP (reduced cation conductance). Corresponding CT response is unchanged but the TEC (per unit ionic strength) declined by 50%. The 0.3 M NaCl + higher concentrations of CaCl2 (arrows 3-5) show TEP becoming progressively more electronegative (declining cation conductance), a continued decline in TEC, and CT responses substantially smaller than control. Ability of Ca2+ to transform the paracellular pathway from cation to anion selective with reduced conductance is unaffected by the presence of NaCl, consequently, CaCl2 inhibits the CT response to NaCl as well as its own.

View larger version (28K): [in this window] [in a new window]   FIG. 4. Normalized CT responses as a function of CaCl2 concentration in mixture with 0.3 M NaCl () or 0.1 M NaCl () or without NaCl (), under 0 current clamp. Each point represents the mean ± SE (n = 5, except for [] where n = 6). Note that 0.3 M CaCl2, in mixtures with NaCl, exerts a differential suppression on the responses to each concentration of NaCl (see inset). Note also that nearly a constant difference between the mixture responses and that due to CaCl2 remained at all concentrations of CaCl2 >0.3 M.

Figure 5 displays records testing the preceding hypothesis. Figure 5, top, shows a response to 0.3 M NaCl and its partial inhibition by amiloride. This is followed by a response to 0.3 M CaCl₂ and then a mixture of CaCl₂ with NaCl. The next record demonstrates that all of the NaCl response that occurs in a mixture with CaCl₂ is amiloride-sensitive, i.e., CaCl₂ blocked the AIC of the NaCl response, as suggested in Figs. 3 and 4. This effect is even more striking in mixtures with 0.5 M CaCl₂ (Fig. 5, bottom) because the CT response to CaCl₂ at this, and higher concentrations, is often at or near the baseline. As trace N + C₂ + A shows again, all of the AIC was eliminated and the response to NaCl was driven to baseline by the presence of both Ca²⁺ and amiloride. These results demonstrate that the AIC of NaCl CT responses is entirely Ca²⁺ sensitive.

View larger version (29K): [in this window] [in a new window]   FIG. 5. Integrated CT responses obtained from the same animal under 0 current clamp. Notation as is follows: N = 0.3 M NaCl; R = rinse; A = 30 µM amiloride; C1 = 0.3 M CaCl2; C2 = 0.5 M CaCl2. Top: response to 0.3 M NaCl followed by the response to 0.3 M NaCl + amiloride. Latter represents the amiloride-insensitive component (AIC) of NaCl CT response. Next, response to 0.3 M CaCl2 followed by a repeat stimulation with 0.3 M CaCl2. On achieving steady state response, 0.3 M CaCl2 was replaced by the mixture of NaCl and CaCl2, each at 0.3 M. Additional response caused by NaCl in mixture with CaCl2 represents the amiloride-sensitive component (ASC) of NaCl as is demonstrated in the next trace where the NaCl/CaCl2 mixture contains amiloride. Note the elimination of the NaCl-evoked response. Bottom: response to 0.3 M NaCl followed by the response to 0.3 M NaCl + 0.5 M CaCl2, the later represents the ASC of NaCl CT response. This is evident from the next trace where the application of amiloride completely eliminated the entire response.

Effect of Ca²⁺ on KCl CT responses

Figure 6 shows the effect of using a series of CaCl₂ concentrations in mixture with 0.3 M KCl. Complete suppression of KCl CT responses by CaCl₂ did not occur until the CaCl₂ concentration in the mixture was ~0.3 M. When the CaCl₂ concentration in the mixture exceeded 0.3 M, KCl responses were, in fact, not significantly different from responses observed with CaCl₂ alone (Fig. 6, inset). Ca²⁺ affects K⁺ responses in a manner similar to the AIC of NaCl response (cf. Figs. 4 and 5). This would be expected if transduction sites for both Na⁺ and K⁺ are accessible only by way of a paracellular pathway, blockable by Ca²⁺.

View larger version (21K): [in this window] [in a new window]   FIG. 6. Normalized CT responses as a function of CaCl2 concentration in a mixture with 0.3 M KCl () or without KCl () under 0 current clamp. Each point represents the mean ± SE (n = 5). Note that complete suppression of KCl responses that occur in the mixture with CaCl2 was not apparent until the CaCl2 concentration in the mixture was 0.3 M. This can be seen when the mean difference between the mixture responses and those to CaCl2 alone was plotted as a function of CaCl2 concentration (see inset).

Effect of Ca²⁺ on NH₄Cl CT responses

Figure 7A shows recordings from experiments demonstrating the effect of two concentrations of CaCl₂ in mixture with either 0.1 or 0.3 M NH₄Cl on the CT responses to NH₄Cl. The chosen NH₄Cl concentrations are representative of the low (0.1 M) and high (0.3 M) concentration regimes observed in CT responses to NH₄Cl (Kloub et al. 1997). Figure 7A (top) shows a response to 0.3 M NH₄Cl (1st record) followed by responses to mixtures of 0.3 M NH₄Cl and 0.3 M CaCl₂ (2nd record) and 0.3 M NH₄Cl and 0.5 M CaCl₂ (3rd record). The suppression of 0.3 M NH₄Cl response by both concentrations of CaCl₂ is obvious. In contrast, Fig. 7A (bottom) shows that the responses to mixtures of 0.1 M NH₄Cl and either concentration of CaCl₂ significantly exceed the response to 0.1 M NH₄Cl alone. An examination of the concentration-response relation for the two concentrations of NH₄Cl over a range of CaCl₂ (Fig. 7B) shows that the responses to 0.1 M NH₄Cl and CaCl₂ are nearly additive over the range of CaCl₂ concentration. On the other hand, responses to 0.3 M NH₄Cl and CaCl₂ follow the KCl pattern, i.e., a much less-than-additive response <0.3 M CaCl₂ and a response not significantly different from that of CaCl₂ alone at CaCl₂ concentrations >0.3 M. Similar Ca²⁺ effects on NH₄Cl CT responses were observed when presented in mixtures with 0.5 M NH₄Cl (data not shown). The major qualitative differences between 0.1 and 0.3 M NH₄Cl in mixtures with CaCl₂ once again, emphasize that NH₄Cl CT responses are the result of two transduction mechanisms. One of these predominates at concentrations approximately >0.3 M and is believed to involve NH₄Cl transport across paracellular pathway (Kloub et al. 1997). The fact that the high NH₄Cl concentration CT response regime is more sensitive to Ca²⁺ is further confirmation of this view.

View larger version (22K): [in this window] [in a new window]   FIG. 7. A: integrated CT responses under 0 current clamp for 2 concentrations of NH4Cl (0.1 and 0.3 M) alone or in a mixture with 2 concentrations of CaCl2 (0.3 and 0.5 M). Top: response to 0.3 M NH4Cl followed by the responses ofthe mixtures, 0.3 M NH4Cl + 0.3 M CaCl2 (C1) and 0.3 M NH4Cl + 0.5 M CaCl2 (C2). Note that NH4Cl CT responses were suppressed in the presence of CaCl2. Bottom: response to 0.1 M NH4Cl followed by the responses of the mixtures, 0.1 M NH4Cl + 0.3 M CaCl2 (C1) and 0.1 M NH4Cl + 0.5 M CaCl2 (C2). Note that the responses to mixtures of 0.1 M NH4Cl and either concentration of CaCl2 significantly exceed the response to 0.1 M NH4Cl alone. B: mean normalized CT responses as a function of CaCl2 concentration in mixture with 0.3 M NH4Cl () or 0.1 M NH4Cl () or without NH4Cl () under 0 current clamp. Each point represents the mean ± SE (n = 5). Note that CaCl2 exhibits differential effects on the responses to each concentration of NH4Cl. Responses of the mixture of 0.1 M NH4Cl and CaCl2 and those to CaCl2 alone are nearly additive, but nearly a complete suppression of 0.3 M NH4Cl responses occurred in mixtures with CaCl2 when the CaCl2 concentration in the mixture was 0.3 M.

Analysis of self-inhibition in Ca²⁺-evoked taste responses

Figures 1 and 2 suggest that the self-inhibition in the CaCl₂ response results from CaCl₂ having to diffuse across a paracellular barrier to reach the Ca²⁺ sensors on the basolateral membranes of taste-bud cells. The reversal in the TEP (Fig. 2C) suggests that in passing through the diffusion barrier Ca²⁺ causes a reduction in Ca²⁺ ion permeability relative to that of Cl. From the rapid decline in conductance observed between 0.1 and 0.3 M CaCl₂ (Fig. 2B), this appears to occur through a highly cooperative reaction between Ca²⁺ and fixed anionic sites. With a minimum of further assumptions, therefore, the nonmonotonic CT response-concentration curves at zero current and under voltage clamp, the CaCl₂ conductance-concentration curve and the electronegative shift in TEP with increasing CaCl₂ concentration should emerge from a straightforward application of the electrodiffusion equations. The analysis will help illustrate that the permeability barrier attenuates large disturbances in ion concentration, osmotic pressure, and pH that may arise in the oral cavity. The extracellular fluid on the submucosal side of the barrier is only slightly perturbed, consequently, cell volume changes will be negligible.

Zero-current clamp

The CaCl₂ flux across the diffusion barrier between the mucosal side (oral cavity) and the submucosal side is

(1)

where c is the CaCl₂ concentration difference between the submucosal (s) side, c_s and the mucosal (m) side, c_m, and

(2)

Here P₁ and P₂ are the permeability coefficients of Ca²⁺ and Cl, respectively. The corresponding potential difference under zero current is

(3)

We restrict the analysis to the pseudo-steady state (tonic phase) of the CT response. This state is assumed to be sustained by a first order process that removes Ca²⁺ ions from the submucosal compartment at the same rate that they enter by diffusion from the mucosal side. This would be the case if, for example, the inflowing CaCl₂ can diffuse into the extracellular fluid space. The steady state condition is

(4)

where J is the Ca²⁺ influx into the s compartment (cf. Eq. 1),  is rate constant, V is the volume, and A is the area of the s compartment. Substituting for J in Eq. 4 using Eq. 1 gives the s concentration, c_s, as a function of the m concentration, c_m, viz

(5)

where  =  V/A. Ca²⁺ self-inhibition of Ca²⁺ permeability can be modeled as

(6)

(7)

Voltage clamp

Under voltage-clamp the Ca²⁺ influx, J₁, is given by

(8)

(9)

(10)

where  is the dimensionless perturbation voltage. If  is restricted to values less than one, i.e., to ~26 mV, a solution has the form

(11)

where x₀ is c_s/c_m under zero current conditions, and  is the voltage modulator function, which depends on a quantity q, given by

(12)

which is twice the permeability factor that determines open-circuit potential (cf. Eq. 3). The voltage modulator function,  is 1/h(q) where h(q) is

(13)

From Eq. 11 the submucosal Ca²⁺ concentration, c_s under voltage clamp is given by c_s0 exp[2], and with Eq. 7 this is

(14)

where the voltage dependent factor in k (i.e., exp[2]) also has been included. The predicted values of c_s as a function of c_m are displayed in Fig. 8 (top) for the three voltage conditions:  = 1,  = 0, and  = +1, and for = 150 (Fig. 8, curves A-C, respectively; see legend for the values of the other parameters). The extent of diffusion control in the system, as expressed in the value of , is unknown a priori. However, because  determines the maximum concentration of Ca²⁺ achieved on the submucosal side of the diffusion barrier, it is possible to put bounds on it. Table 1 shows the maximum value of c_s at zero current as a function of the corresponding index of diffusion control,  for the choice of parameters used in Fig. 8. To minimize the chances of cytotoxicity, it seems likely that the degree of diffusion control will be such to prevent the maximum value of c_s from rising no more than a few millimolar above its basal level. This would tend to restrict  to values greater than ~25 but with an upper limit of ~150. The profiles of c_s as a function of c_m in Fig. 8, accurately reflect the CT response-concentration curves for CaCl₂ in Fig. 2A. This suggests that taste cell responses are simply proportional to the changes in Ca²⁺ that occur in the submucosal microenvironment.

View larger version (15K): [in this window] [in a new window]   FIG. 8. Top: theoretical value of the Ca2+ concentration on the submucosal side of the paracellular diffusion barrier in the taste-bud (cs) as a function of the mucosal concentration (cm) according to Eq. 14. Curve A: voltage clamp at v = 1 (26 mV); curve B: 0 current clamp; curve C: voltage clamp at v = +1 (26 mV). Parameter values are P = 2, k = 0.2, n = 4, and  = 150. Note that although cm spans a range from 0 to 1 M, cs never increases more than 1.3 mM above baseline. Bottom: voltage modulator function,  as a function of cm. ( = 1/h) is plotted according to Eq. 13.  represents the fraction of the applied voltage affecting the Ca2+ flux through the paracellular region. This decrease parallels the declining normalized CaCl2 transepithelial conductance (cf. Fig. 2B and text). Values of p, k, n, and  are as in top.

Voltage modulator function

Figure 8 (bottom) shows  as a function of c_m for the same parameter values used in Fig. 8 (top).  has a maximum value of 0.3 at c_m = 0, so diffusing Ca²⁺ ions "feel" at maximum only 30% of the clamp voltage at low Ca²⁺ concentration where the diffusion barriers are still cation selective. As they become increasingly anion selective, the influence of potential on Ca²⁺ influx diminishes, reflecting the decreasing Ca²⁺ permeability as c_m increases. The qualitative form of  as a function of c_m should, therefore, parallel that of the TEC (per unit ionic strength). A comparison of Fig. 8 with Fig. 2B confirms that expectation (see also following text). On that basis, we might expect the form of  to hold also for voltages >26 mV (i.e., for  > 1) but with different limits. The average value of  for c_m between 0 and 0.6 M is 0.15. Accordingly, in fitting the CT responses in Fig. 2A to Eq. 14,  is treated as an adjustable constant, expected to lie between 0.05 and 0.3 and most probably near the average of 0.15.

Fitting the data

To fit the CT response data for CaCl₂, we assume that the response, R, is simply proportional to c_s. Equation 14 can be recast as

(15)

(16)

(17)

It follows that

(18)

indicating that b must be >1 if negative permeabilities are to be avoided. In fitting the CT responses to Eq. 16, there are as many as five free parameters. The zero current constraint reduces the number to four, still a high degree of freedom. When a four parameter fit was implemented values of 0.37 M and 3.8 were obtained for k and n, respectively. However, b was 0.62, i.e., <1. The value n = 3.8 agrees well with values obtained by fitting the conductance data in Fig. 2B empirically to a Hill equation. Also in Fig. 2B, the most rapid drop in conductance occurs between 0.1 and 0.3 M, suggesting an estimate of k of ~0.2 M. If we constrain k and n at 0.2 M and 4, respectively, and seek a two-parameter fit, then a and b are: 26.8 and 8.64, respectively. This two-parameter fit is displayed in Fig. 2A for the CT response data under zero current clamp and represents the data well. With b = 8.64, the ratio of the Ca²⁺ permeability coefficient to that of Cl in the limit of zero CaCl₂, p, is seen to be narrowly constrained. For  = 25, P = 3.6, in the limit as  approaches infinity, P = 3.8. This is consistent with the value of P = 2 chosen a priori in plotting the theoretical curves shown in Fig. 8. In fitting the data in Fig. 2A under voltage clamp (±50 mV), we required that the current clamp values of a, b, k, and n be maintained, and that the fit be achieved by varying the single parameter . The fit shown in Fig. 2A for the clamp voltage, 50 mV, was obtained with  = 0.12, and at +50 mV with  = 0.11, values that are consistent with the range of expected values.

Interpretation of the conductance data

The curve in Fig. 2B of the conductance per unit ionic strength of CaCl₂ was obtained as follows. A perturbation in potential imposed on the open circuit potential (Eq. 3) will produce a small current. Provided the voltage-perturbation is fast compared with diffusional relaxation, the steady state zero-current concentration profiles will prevail. The conductance per unit ionic strength, L_t (in ¹ mol¹ cm¹), is then

(19)

where x = (c_s/c_m), and L_n is a nonspecific constant leak conductance per unit ionic strength. The dimensionless conductance (normalized to 0.1 M NaCl), L_d, is

(20)

where A and B may be regarded as scaling constants independent of c_m. However, B also represents the proportion of the conductance outside the taste-receptive regions. For self-consistency, we require that the conductance data be satisfied by the same parameter values obtained in fitting the CT response curves (Fig. 2A). Accordingly we use Eq. 15 with  = 0 to calculate x, keeping k = 0.2, n = 4 and using the least squares fit parameter, b = 8.64, to calculate, p, the ratio of the Ca²⁺ permeability coefficient to that of Cl at zero CaCl₂ concentration. Using, as before,  = 150, constrains, p to a value of 3.78 (cf. Eq. 18). In Eq. 20, therefore, x and P are determined fully and the only parameters remaining are A and B. Using least-squares criteria on Eq. 20 provides the fit of the conductance-concentration data shown in Fig. 2B with A = 0.14 and B = 0.067.

The conductance per unit ionic strength of CaCl₂ is only 30% that of NaCl at maximum (at c_m = 0). In this limit, the percent of the total conductance attributable to taste-receptive regions (i.e., the conductance in excess of that of nonspecific leaks) is ~79%. At 0.3 M, it has dropped to ~47%, and by 0.5 M, it is but 27%. As seen in Figs. 1 and 2A, the maximum CT response occurs near 0.3 M. CT responses begin their decline, therefore, when the conductance through the taste-receptive regions falls to <50% of the total conductance. This finding further supports the view that transport of the stimulus across the paracellular regions in the taste buds (as reflected in the conductance) is critical in sustaining the taste response.

Transepithelial potential

The zero current potential, ₀ with c_m has the expected properties of the system. In Fig. 1, the potential increases in the electronegative direction as the CaCl₂ concentration increases, and this occurs (between 0.1 and 0.3 M) even as the CT response is increasing. The model shows this same characteristic. Normalizing the CaCl₂ potential to that of 0.1 M NaCl gives the dimensionless potential _d plotted in Fig. 2C as a function log c_m. The curve was plotted according to Eq. 3, put in the form

(21)

We have retained all of the parameters used in fitting Figs. 2, A and B, except for the scaling constants A₁ and B₁, determined as 0.76 and 2.54, respectively. In accord with the data, the changing ionic selectivity of the diffusion barrier therefore is seen in the potential before the submucosal Ca concentration (and therefore the CT response) reaches its maximum. Thus the changes in potential seen with CaCl₂ are not, themselves, the cause of the self-inhibition in the CT response induced by CaCl₂ but rather the consequence of the cooperative changes in the Ca²⁺ permeability of the paracellular barrier, resulting in increased relative Cl permeability but overall decreased salt permeability (i.e., conductance, see preceding text).

	DISCUSSION

Abstract Introduction Methods Results Discussion References

Paracellular Ca²⁺-induced changes in TEP and self-inhibition of CT responses

The paracellular pathway mediates passive Ca²⁺ transport in various epithelia including intestine (Bronner and Spence 1988), renal (Bronner 1989), and lingual (Sostman and Simon 1991) epithelia. Our data are consistent with that view, i.e., Ca²⁺ permeation of the taste-bud epithelium occurs principally through paracellular shunts. The TEP, established across these shunts, is normally moderately cation selective (Berry et al. 1978; DeSimone et al. 1984; Simon et al. 1988; Reuss 1991; Ye et al. 1993), but CaCl₂ produces an electronegative change in the TEP (Figs. 1 and 2C). This is consistent with similar observations in small intestine (Smyth and Wright 1966), choroid plexus (Prather and Wright 1969), and gallbladder (Moreno and Diamond 1974).

Ca²⁺ binding to fixed anionic sites will buffer Ca²⁺, initially reducing the effective Ca²⁺ concentration during Ca²⁺ diffusion. However, fixed charge density is limited so buffering will be transient. The major effect of Ca²⁺ binding is reduced Ca²⁺ permeability and consequently reduced CaCl₂ conductance (cf. Fig. 2B). The CT response initially increases (e.g., between 0.1 and 0.3 M, cf. Fig. 2A). This suggests that, although the Ca²⁺ permeability is decreasing (cf. Fig. 2B), it is still sufficiently high to cause an increase in the submucosal Ca²⁺ concentration. The model predicts an increase in submucosal Ca²⁺ concentration as stimulus Ca²⁺ concentration increases between 0 and 0.3 M (cf. Fig. 8). However, >0.3 M, the response decreases as the CaCl₂ concentration on the mucosal side increases. As seen in Fig. 2B, this coincides with the low CaCl₂ conductance regime, where the model indicates (cf. Fig. 8) that increasing submucosal concentrations no longer can be sustained. The CT response (Fig. 2A) and the TEC (Fig. 2B) begin their decline over a narrow range of CaCl₂ concentrations, suggesting that both derive from a highly cooperative process. As we have shown, the single assumption of a Ca²⁺ permeability regulated cooperatively by the Ca²⁺ concentration itself is sufficient to account for: the electronegative TEP, the falling TEC, and ultimately the self-inhibiting CT response. Extracellular Ca²⁺ has been shown to regulate the degree of tightness of the TJs (Contreras et al. 1991), suggesting that the TJs mediate Ca²⁺-induced reduction in CaCl₂ permeability. Modulation of TJ properties has been demonstrated in various species including amphibia (Moreno and Diamond 1974), suggesting the nonmonotonic concentration-response relationships for CaCl₂, obtained from single-unit recording from amphibian peripheral taste nerves (Kitada 1995), also may arise through paracellular processes.

Although the fibrous material of TJs is not fully characterized (Madara 1991), their ion-exchange (polyelectrolyte) properties are well established. Moreover, polyelectrolyte fibers in model systems, subjected to an ion exchange reaction with Ca²⁺, undergo length changes resembling phase transitions in that the changes occur at critical inducing concentrations as in a highly cooperative process (Katchalsky and Oplatka 1971). In addition, Ca²⁺ and other divalent cations have been shown to induce contraction of the perijunctional actomyosin ring, located just below the TJs (Burgess 1982; Rodewald et al. 1976). If Ca²⁺, through such mechanochemical transformations, can reduce its own TJ permeability and therefore its own CT response, it is reasonable to expect that it will inhibit the CT responses of other stimuli depending on paracellular transport.

Ca²⁺-sensitivity of NaCl, KCl, and NH₄Cl CT responses

Part of the NaCl-evoked CT response seems to arise from transduction sites located below the TJs (Bradley 1973; Elliot and Simon 1990; Mierson et al. 1996; Simon et al. 1993; Stewart et al. 1995; Ye et al. 1993). If Ca²⁺ can block access to its own paracellular transduction sites, it should block completely the AIC of the NaCl CT response attributed to transduction sites below the TJs (Elliot and Simon 1990; Mierson et al. 1996; Simon et al. 1993; Stewart et al. 1995; Ye et al. 1993). As seen in Figs. 3 and 4, Ca²⁺ inhibited the CT responses of NaCl. The percentage inhibition was higher at 0.3 M NaCl than at 0.1 M NaCl, corresponding approximately to the percentage of the CT response due to the AICs at these NaCl concentrations (Elliot and Simon 1990; Ye et al. 1993). Ca²⁺ also can inhibit Na⁺ channels, but inhibition constants exceed 0.3 M (Palmer 1986). In mixtures with CaCl₂ concentration >0.3 M, some further suppression occurred, probably attributable to some direct inhibition of the apical Na⁺ channels by Ca²⁺. However, direct suppression of the Na⁺ channels by Ca²⁺ would not explain the different levels of suppression of 0.1 and 0.3 M NaCl.

Although the magnitude of the NaCl CT response blocked by Ca²⁺ is quantitatively similar to that of the AIC, this is insufficient to prove that they are identical. However, the results of Fig. 5, in which Ca²⁺ block and amiloride block of NaCl CT responses were implemented together, provide strong evidence in favor of this view. They demonstrate that the AIC of NaCl CT responses is entirely Ca²⁺ sensitive. The block occurs rapidly between 0.2 and 0.3 M CaCl₂ (cf. Figs. 3 and 4), which coincides with the range over which the conductance declines rapidly as a function of CaCl₂ concentration (Fig. 2B), suggesting that Ca²⁺ blocks Na⁺ permeability (and therefore CT response) by the same cooperative process. This is further evidence that taste transduction sites for Na⁺ are accessible only by way of a paracellular shunt. Although the AIC comprises <50% of the NaCl-evoked taste intensity in the rat, in human NaCl perception it appears to be more dominant (Ossebaard and Smith 1995). It is unknown, however, if the human AIC is Ca²⁺ sensitive in a manner comparable with rat.

The paracellular pathway also has been implicated in taste transduction for K⁺ salts (Ye at al. 1994). If so, Ca²⁺ should block completely KCl-evoked taste responses. Figure 6 demonstrates that 0.3 M KCl CT responses, which occurred in mixtures with CaCl₂ 0.3 M, were eliminated completely. These results support the view that KCl transduction is mediated largely through a paracellular pathway susceptible to block by Ca²⁺.

Earlier work shows that <0.3 M, NH₄Cl mainly uses a transcellular transduction pathway, whereas, 0.3 M, a paracellular pathway is favored (Kloub et al. 1997). Therefore if the Ca²⁺ effects observed here are exerted mainly on the ion permeability of the TJs, Ca²⁺ block of the NH₄Cl CT response should be most effective at higher NH₄Cl concentrations. This is demonstrated in Fig. 7. The differential effect of CaCl₂ on the two concentration domains of the NH₄Cl CT response reinforces the earlier conclusions regarding dual transduction mechanisms for NH₄Cl, with the one predominating at higher NH⁺₄ concentrations being principally shunt-mediated.

Concentration modulator properties of the paracelluar diffusion barrier

The concentration range of salt taste sensitivity is typically from millimolar to hundreds of millimolar. Changes of this order in the intercellular milieu almost certainly would be injurious to the sensory cells. In the case of CaCl₂, exceeding the intracellular buffering capacity for Ca²⁺ would lead to cell death (McConkey and Orrenius 1996). Ca²⁺-induced inhibition of its own CT response and that of other salts is reversible and without obvious cytotoxicity, suggesting that Ca²⁺ concentrations at the effector sites are far below those in the applied stimulus solution. We conclude that CaCl₂ concentration attenuation is achieved by placing the common transduction pathway for Ca²⁺, K⁺, Na⁺, and NH⁺₄ salts in series with a significant diffusion barrier in a manner demonstrated in the analysis. Ca²⁺, of course, is also an important intracellular regulatory agent. Modulation of TJ permeability in various epithelial and endothelial tissues may be subject therefore to physiological regulation by Ca²⁺ supplied to the submucosal microenvironment from internal stores.

	ACKNOWLEDGEMENTS

  We thank Dr. Steven Price for helpful suggestions and Dr. Oxana Kloub for technical assistance.

  This research was supported by National Institute of Deafness and Other Communication Disorders Grants DC-00122 and DC-02422.

	FOOTNOTES

  Address for reprint requests: J. A. DeSimone, Dept. of Physiology, Virginia Commonwealth University, PO Box 980551, Richmond, VA 23298-0551.

  Received 24 April 1997; accepted in final form 30 October 1997.

	REFERENCES

Abstract Introduction Methods Results Discussion References

• AVENET, P., LINDEMANN, B. Amiloride-blockable sodium currents in isolated taste receptor cells. J. Membr. Biol. 105: 245-255, 1988.[Medline]
• BERRY, C. A., WARNOCK, D. G., RECTOR, F. C. Ion selectivity and proximal salt reabsorption. Am. J. Physiol. 235: F234-F245, 1978.
• BRADLEY, L. M. Electrophysiological investigations of intravascular taste using perfused rat tongue. Am. J. Physiol. 224: 300-304, 1973.[Free Full Text]
• BRAND, J. G., TEETER, J. H., SILVER, W. L. Inhibition by amiloride of chorda tympani responses evoked by monovalent salts. Brain Res. 334: 207-214, 1985.[Medline]
• BRONNER, F. Renal calcium transport: mechanisms and regulation- an overview. Am. J. Physiol. 257: F707-F711, 1989.[Abstract/Free Full Text]
• BRONNER, F., SPENCE, K. Non-saturable Ca transport in the rat intestine is via the paracellular pathway. Prog. Clin. Biol. Res. 252: 277-283, 1988.[Medline]
• BURGESS, D. R. Reactivation of intestinal epithelial brush border motility: ATP-dependent contraction via a terminal web contractile ring. J. Cell. Biol. 95: 853-863, 1982.[Abstract/Free Full Text]
• CONTRERAS, R. G., PONCE, A., BOLIVAR, J. J. Calcium and tight junctions. In: Tight Junctions, edited by and M. Cereijido . Boca Raton, FL: CRC, 1991, p. 139-149
• DESIMONE, J. A., FERRELL, F. An analysis of amiloride inhibition of the chorda tympani taste response of the rat to NaCl. Am. J. Physiol. 249: R52-R61, 1985.[Abstract/Free Full Text]
• DESIMONE, J. A., HECK, G. L., DESIMONE, S. K. Active ion transport in dog tongue: a possible role in taste. Science 214: 1039-1041, 1981.[Abstract/Free Full Text]
• DESIMONE, J. A., HECK, G. L., MIERSON, S., DESIMONE, S. K. The active ion transport properties of canine lingual epithelia in vitro: implications for gustatory transduction. J. Gen. Physiol. 83: 633-656, 1984.[Abstract/Free Full Text]
• ELLIOTT, E. J., SIMON, S. A. The anion in salt taste: a possible role for paracellular pathways. Brain Res. 535: 9-17, 1990.[Medline]
• FORMAKER, B. K., HILL, D. L. An analysis of residual NaCl taste response after amiloride. Am. J. Physiol. 255: R1002-R1007, 1988.[Abstract/Free Full Text]
• HECK, G. L., ERICKSON, R. P. A rate theory of gustatory stimulation. Behav. Biol. 8: 687-712, 1973.[Medline]
• HECK, G. L., MIERSON, S., DESIMONE, J. A. Salt taste transduction occurs through an amiloride-sensitive sodium transport pathway. Science 223: 403-405, 1984.[Abstract/Free Full Text]
• HOLLAND, V. F., ZAMPIGHI, G. A., SIMON, S. A. Morphology of fungiform papillae in canine lingual epithelium: location of intercellular junctions in the epithelium. J. Comp. Neurol. 279: 13-27, 1989.[Medline]
• KATCHALSKY, A., OPLATKA, A. Mechano-chemical conversion. In: Handbook of Sensory Physiology, Principles of Receptor Physiology, edited by and W. R. Loewenstein . Berlin: Springer-Verlag, 1971, vol. 1, p. 1-17
• KITADA, Y. Anions modulate cation-induced responses of single units of the frog glossopharyngeal nerve. Brain Res. 694: 253-263, 1995.[Medline]
• KLOUB, M. A., HECK, G. L., DESIMONE, J. A. Chorda tympani responses under lingual voltage clamp: implications for NH₄ salt taste transduction. J. Neurophysiol. 77: 1393-1406, 1997.[Abstract/Free Full Text]
• LUNDY, R. F. JR, CONTRERAS, R. J. Temperature and amiloride alter taste nerve responses to Na⁺, K⁺, and NH⁺₄ salts in rats. Brain Res. 744: 309-317, 1997.[Medline]
• MADARA, J. L. Anatomy of the tight junction: vertebrates. In: Tight Junctions, edited by and M. Cereijido . Boca Raton, FL: CRC, 1991, p. 15-22
• MARTINEZ-PALOMO, A., MEZA, I., BEATY, G., CEREIJIDO, M. Experimental modulation of occluding junctions in a cultured transporting epithelium. J. Cell Biol. 87: 736-745, 1980.[Abstract/Free Full Text]
• MCCONKEY, D. J., ORRENIUS, S. The role of calcium in the regulation of apoptosis. J. Leukocyte Biol. 59: 775-783, 1996.[Abstract]
• MELDOLESI, J., CASTIGLIONI, G., PARMA, R., NASSIVERA, N., DECAMILLI, P. Ca⁺⁺-dependent disassembly and reassembly of occluding junctions in guinea pig pancreatic acinar cells. J. Cell Biol. 79: 165-172, 1978.
• MIERSON, S., OLSON, M., TIETZ, A. E. Basolateral amiloride-sensitive Na⁺ transport pathway in rat tongue epithelium. J. Neurophysiol. 76: 1297-1309, 1996.[Abstract/Free Full Text]
• MOOSEKER, M. S. Organization, chemistry, and assembly of the cytoskeletal apparatus of the intestinal brush border. Annu. Rev. Cell Biol. 1: 209-242, 1985.
• MORENO, J. H., DIAMOND, J. M. Discrimination of monovalent inorganic cations by "tight" junctions of gallbadder epithelium. J. Membr. Biol. 15: 277-318, 1974.[Medline]
• OSSEBAARD, C. A., SMITH, D. V. Effect of amiloride on the taste of NaCl, Na-gluconate and KCl in humans: implications for Na⁺ receptor mechanisms. Chem. Senses 20: 37-46, 1995.[Abstract/Free Full Text]
• PALMER, L. G. The epithelial sodium channel. In: New Insights Into Cell and Membrane Transport Processes, edited by G. Poste, and S. T. Crooke . New York: Plenum, 1986, p. 327-344
• PITELKA, D. R., TAGGART, B. N., HAMAMOTO, S. T. Effects of extracellular calcium depletion on membrane topography and occluding junctions of mammary epithelial cells in culture. J. Cell Biol. 96: 613-624, 1983.[Abstract/Free Full Text]
• PRATHER, J. W., WRIGHT, E. M. Choroid plexus permeability to non-electrolytes. J. Physiol. (Lond.) 204: 29P-30P, 1969.
• REUSS, L. Tight junction permeability to ions and water. In: Tight Junctions, edited by and M. Cereijido . Boca Raton, FL: CRC, 1991, p. 49-66
• RODEWALD, R., NEWMAN, S. B., CARNOVSKY, M. J. Contraction of isolated brush borders from the intestinal epithelium. J. Cell Biol. 70: 541-545, 1976.[Abstract/Free Full Text]
• SCHIFFMAN, S. S., LOCKHEAD, E., MAES, F. W. Amiloride reduces the taste intensity of Na⁺ and Li⁺ salts and sweeteners. Proc. Natl. Acad. Sci. USA 80: 6136-6140, 1983.[Abstract/Free Full Text]
• SEDAR, A. W., FORTE, J. G. Effects of calcium depletion on the junctional complex between oxyntic cells of gastric glands. J. Cell Biol. 22: 173-188, 1964.[Abstract/Free Full Text]
• SIMON, S. A., GARVIN, J. L. Salt and acid studies on canine lingual epithelium. Am. J. Physiol. 249: C398-C408, 1985.[Abstract/Free Full Text]
• SIMON, S. A., HOLLAND, V. F., BENOS, D. J., ZAMPHIGHI, G. A. Transcellular and paracellular pathways in lingual epithelia and their influence in taste transduction. Microsc. Res. Tech. 26: 196-208, 1993.[Medline]
• SIMON, S. A., ROBB, R., SCHIFFMAN, S. S. Transport pathways in rat lingual epithelium. Pharmacol. Biochem. Behav. 29: 257-267, 1988.[Medline]
• SMITH, D. V., BEALER, S. L. Sensitivity of the rat gustatory system to the rate of stimulus onset. Physiol. Behav. 15: 303-314, 1975.[Medline]
• SMITH, P. R., BENOS, D. J. Epithelial Na⁺ channels. Annu. Rev. Physiol. 53: 509-530, 1991.[Medline]
• SMYTH, D. H., WRIGHT, E. M. Streaming potentials in the rat small intestine. J. Physiol. (Lond.) 182: 591-602, 1966.[Abstract/Free Full Text]
• SOSTMAN, A. L., SIMON, S. A. Trigeminal nerve responses in rat elicited by chemical stimulation of the tongue. Arch. Oral Biol. 36: 95-102, 1991.[Medline]
• STEWART, R. E., LASITER, P. S., BENOS, D. J., HILL, D. L. Immunohistochemical correlates of peripheral gustatory sensitivity to sodium and amiloride. Acta Anat. 153: 310-319, 1995.[Medline]
• TIDBALL, C. S. Magnesium and calcium as regulators of intestinal permeability. Am. J. Physiol. 206: 243-246, 1964.[Abstract/Free Full Text]
• YE, Q., HECK, G. L., DESIMONE, J. A. Voltage dependence of the rat chorda tympani response to Na⁺ salts: implications for the functional organization of taste receptor cells. J. Neurophysiol. 70: 167-178, 1993.[Abstract/Free Full Text]
• YE, Q., HECK, G. L., DESIMONE, J. A. Effects of voltage perturbation of the lingual receptive field on chorda tympani responses to Na⁺ and K⁺ salts in the rat: implications for gustatory transduction. J. Gen. Physiol. 104: 885-907, 1994.[Abstract/Free Full Text]

0022-3077/98 $5.00 Copyright ©1998 The American Physiological Society