Cyclic-AMP regulates postnatal development of neural and behavioral responses to NaCl in rats

During postnatal development rats demonstrate an age-dependent increase in NaCl chorda tympani (CT) responses and the number of functional apical amiloride-sensitive epithelial Na+ channels (ENaCs) in salt sensing fungiform (FF) taste receptor cells (TRCs). Currently, the intracellular signals that regulate the postnatal development of salt taste have not been identified. We investigated the effect of cAMP, a downstream signal for arginine vasopressin (AVP) action, on the postnatal development of NaCl responses in 19–23 day old rats. ENaC-dependent NaCl CT responses were monitored after lingual application of 8-chlorophenylthio-cAMP (8-CPT-cAMP) under open-circuit conditions and under ±60 mV lingual voltage clamp. Behavioral responses were tested using 2 bottle/24h NaCl preference tests. The effect of [deamino-Cys1, D-Arg8]-vasopressin (dDAVP, a specific V2R agonist) was investigated on ENaC subunit trafficking in rat FF TRCs and on cAMP generation in cultured adult human FF taste cells (HBO cells). Our results show that in 19–23 day old rats, the ENaC-dependent maximum NaCl CT response was a saturating sigmoidal function of 8-CPT-cAMP concentration. 8-CPT-cAMP increased the voltage-sensitivity of the NaCl CT response and the apical Na+ response conductance. Intravenous injections of dDAVP increased ENaC expression and γ-ENaC trafficking from cytosolic compartment to the apical compartment in rat FF TRCs. In HBO cells dDAVP increased intracellular cAMP and cAMP increased trafficking of γ- and δ-ENaC from cytosolic compartment to the apical compartment 10 min post-cAMP treatment. Control 19–23 day old rats were indifferent to NaCl, but showed clear preference for appetitive NaCl concentrations after 8-CPT-cAMP treatment. Relative to adult rats, 14 day old rats demonstrated significantly less V2R antibody binding in circumvallate TRCs. We conclude that an age-dependent increase in V2R expression produces an AVP-induced incremental increase in cAMP that modulates the postnatal increase in TRC ENaC and the neural and behavioral responses to NaCl.

The results presented here demonstrate that in 19-23 day old rats, an incremental increase in TRC cAMP mimics the postnatal age-dependent increase in the NaCl CT response and behavior. These changes are accompanied by a concomitant increase in the number of functional apical ENaC density in FF TRCs. Both dDAVP and cAMP increased ENaC expression and ENaC subunit trafficking from cytosolic compartment to the apical compartment in FF TRCs. Relative to adult rats, 14 day old rats demonstrated significantly less V2R antibody binding in circumvallate TRCs. We conclude that an age-dependent increase in V2R expression produces an AVP-induced incremental increase in cAMP that modulates the postnatal increase in TRC ENaC and the neural and behavioral responses to NaCl.

Antibodies and q-PCR primers
Western blots were used to detect the presence of α-and γ-ENaC subunits in rat TRCs. Antiserum against rat α-or γ-ENaC subunit was generated by immunizing rabbits with specific α-or γ-rENaC subunit peptides described earlier [53] using a standard immunization protocol (Pierce Biotechnology, Rockford, IL). V2R ((P-20)-R sc-1800-R) rabbit polyclonal antibody was obtained from Santa Cruz Biotechnology. The primary antibody for δ-hENaC (rabbit polyclonal anti-SCNN1D antibody (aa411-460) LS-C119717) was obtained from LifeSpan Biosciences and was used only on HBO cells. The secondary antibody, donkey anti-rabbit IgG-CFL 488 was obtained from Santa Cruz Biotechnology. Q-PCR primers for α-and γ-rENaC were obtained from Invitrogen. The primary antibodies for β-actin and HRP-conjugated secondary antibody were obtained from Santa Cruz Biotechnology. ECL Western Blotting Substrate was obtained from Life Technologies.
vacuum to a 28 mm 2 patch of anterior dorsal lingual surface. The chamber was fitted with separate Ag-AgCl electrodes for measurement of current and potential and served as inputs to a voltage-current clamp amplifier that permitted the recording of CT responses with the chemically stimulated receptive field under zero current-clamp or voltage-clamp. The clamp-voltages were referenced to the mucosal side of the tongue [54]. For stimulation or rinsing, 3-ml aliquots were injected at a rate of 1 ml/s into the perfusion chamber. Typically, stimulus solutions remained on the tongue for 1-2 min. Control stimuli consisting of 0.3 or 0.5 M NH 4 Cl applied at the beginning and at the end of experiment were used to assess preparation stability. The preparation was considered stable only if the difference between the magnitude of the control stimuli at the beginning and at the end of the experiment was less than 10% [21,55]. Neural responses were differentially amplified with a custom built, optically-coupled isolation amplifier. For display, responses were filtered using a band pass filter with cutoff frequencies 40 Hz-3 KHz and fed to an oscilloscope. Responses were then full-wave rectified and integrated with a time constant of 1s. Integrated neural responses were recorded on a chart recorder and also captured on disk using Labview software and analyzed off-line [21,55]. The CT responses were recorded while the rat tongue was stimulated with a rinse solution (10 mM KCl) and then with stimulating solutions containing NaCl (0.1, 0.3 and 0.5 M) under current-clamp or voltage-clamp conditions. Benzamil (Bz; 5 μM) and cetylpyridinium chloride (CPC; 2 mM) were added to the NaCl solutions to block the ENaC-dependent and the Bz-insensitive component of the NaCl CT response, respectively [50].
CT responses were measured before and after topical lingual application of 8-CPT-cAMP, a membrane permeable form of cAMP. 8-CPT-cAMP (5-20 mM) was dissolved in H 2 O and was topically applied to the tongue for 10-15 min to enhance ENaC activity in TRCs. A membrane permeable form of cGMP, 8-CPT-cGMP was used as a control. In some experiments, adult rats were injected intravenously with AVP (1 nano mole/Kg BW) or dDAVP (0.1 nano moles/Kg BW) dissolved in normal saline and CT responses were monitored before and 15 min post-injection. The data were fitted to an apical ENaC kinetic model [51]. The equations used for data analysis are described in the Appendix.
Behavioral studies. Behavioral studies were performed in 19-27 day old and 60+ day old (adult) rats using standard two bottle/24h NaCl preference tests. Rats (21 day old) were anesthetized with isoflurane (2.5%) using the Surgivet/Anesco ISOTEC4 apparatus [56]. A small piece of a filter paper containing 50 μl of either H 2 O (Control) or 50 μl of 20 mM 8-CPT-cAMP dissolved in H 2 O or 50 μl of 20 mM 8-CPT-cGMP dissolved in H 2 O was placed on the anterior tongue for 20 min. Following this, the rat tongues were thoroughly washed with H 2 O to remove any remaining applied cAMP or cGMP. Individually 19 to 27 day old rats consume a small volume of fluid/day that is difficult to measure accurately. Accordingly, 3-5 rat pups were placed in each plastic cage and were given a choice between 2 bottles, one containing H 2 O and the other containing NaCl (0.035, 0.075 or 0.150 M). Their combined 24h fluid intake was monitored daily for the next 6 days. The volumes consumed for each solution were converted to g of fluid consumed by taking into account the density of each solution. The preference ratio for a taste stimulus was calculated as the g of the test solution consumed/24h/g mean body weight (BW) of rats divided by the total fluid intake (g of H 2 O/24h/g mean BW + g of the test solution/24h/g mean BW). The bottles containing H 2 O and the NaCl solution were switched from left to right at each trial.
Behavioral studies were performed in 3 groups of 14-16 adult rats/group. Before the start of the experiment rats were maintained on two bottles containing H 2 O for 7 days. After 7 days, the mean preference ratio for two bottles containing water was 0.50 ± 0.03. Rats were anesthetized with isoflurane and their tongues were treated with H 2 O or 20 mM 8-CPT-cAMP or 20 mM 8-CPT-cGMP for 20 min as described above. Rats were then placed individually in plastic cages and were given a choice between 2 bottles, one containing H 2 O and the other varying concentrations (0-0.5 M) of NaCl. Their 24h fluid intake was monitored daily for the next 6 days as described before [57]. The bottles containing H 2 O and the NaCl solution were switched from left to right at each trial. The data are expressed as means ± SEM of the number of rats in each group, and statistical significance was determined using Student's t-test for unpaired values.
Western-blot. To prepare protein samples, fifteen 14-19 day old rats and ten 60+ day old rats were used. The circumvallate (CV) papillae were isolated as described earlier [58] and pooled together and were lysed in modified RIPA buffer (50 mM Tris-Cl; pH 7.4, 1% NP-40, 150 mM NaCl, 1 mM EDTA, 1 mM PMSF, 1 μg/ml each of aprotinin and leupeptin, and 1 mM Na 3 VO 4 ). The Western blot was performed with SDS-PAGE electro-blotting system (Bio-Rad). Briefly, 40 μg total protein samples were resolved by 10% SDS-PAGE and transferred to nitrocellulose membranes (Cat: 162-0094, Bio-Rad). Membranes were immune-blotted with rabbit antiserum containing α-rENaC antibody and a primary β-actin antibody followed by HRP-conjugated secondary antibodies. Reactions were visualized by ECL Western Blotting Substrate. Beta-actin was used as a protein loading control.
Immunofluorescence studies. Fifteen day old rats were injected intravenously with 1 nano moles of dDAVP/Kg BW, 3 times, 2h apart, and were sacrificed 24h later. The control rats were injected with saline. Animals were perfused with 4% paraformaldehyde/1×PBS for 5-10 min under anesthetic (isoflurane). The tongues were excised and fixed in 4% paraformaldehyde/1 × PBS for 2h at 4˚C, and dehydrated in 40% sucrose/1×PBS overnight at 4˚C before embedding in O.C.T. Compound (Andwin Scientific, Cat 14-373-65). Sections (8 μm thick) were prepared using a CM3050S cryostat (Leica Microsystems) and applied on pre-coated microscope slides (Fisher Scientific, Cat 12-550-15). The sections were dried at room temperature for 20 min and immediately used for immunofluorescence. After washing with 1X PBS for 5 min and blocking with 3% donkey serum for 1h at room temperature, sections were stained with rabbit antiserum containing γ-rENaC antibody (1:100 dilution in 3% donkey serum) at 4˚C for overnight. After washing, sections were incubated with fluorescent-conjugated secondary antibodies for 1h at room temperature. Nuclei were visualized with 1 μg/ml of 4',6-diamidino-2-phenylindole (DAPI). Images were acquired with a 40X oil immersion objective on a Zeiss LSM 700 confocal laser scanning microscope and processed using Photoshop software (Adobe System). In another set of naïve adult rats (90+ day old) and 14 day old rat pups, V2R antibody binding was investigated in CV taste buds. Rat kidney was used as a positive control. 85˚C X 5 sec and cooled to 4˚C. Subsequently, 200 ng total cDNA was used as template, 35 cycles of PCR amplification were performed (initial denaturation at 95˚C for 1 min, denaturation at 95˚C for 15 sec, annealing for 15 sec at 59-60˚C, and extension for 10 sec at 72˚C). RT-PCR products were subjected to electrophoresis on a 1% agarose gel to determine the expression of ENaC subunits and other taste receptors. The primers used to detect the presence of mRNAs in HBO cells for the α-, β-, γ-, and δ-ENaC subunits were synthesized by Thermo Fisher Scientific. The primer pairs used to detect the mRNA of the ENaC subunits are listed in Table 1.
Detection of δ-ENaC protein by western blot. HBO cells (2 x 10 6 ) were lysed in 200 μl modified RIPA buffer (50 mM Tris-Cl; pH 7.4, 1% NP-40, 150 mM NaCl, 1 mM EDTA, 1 mM PMSF, 1 μg/ml each of aprotinin and leupeptin, and 1 mM Na 3 VO 4 ). Samples containing 30 μg total protein were used for the experiment. The Western blot was performed with SDS-PAGE electro-blotting system (Bio-Rad) as described above for CV taste bud cells. HEK 293-cells were used as positive control. HEK cells were obtained from (American Type Culture Collection, Manassas, VA) and grown in a complete 293 SFMII growth medium (Thermo Fisher Scientific) supplemented with 4 mM L-glutamine. Cells were maintained at 37˚C in 5% CO 2 .

cAMP assay
Cyclic-AMP levels were measured in the presence of the phosphodiesterase blocker IBMX [59] as described earlier [60]. HBO cells (2 x 10 6 cells/ml) were incubated for 10 min in the presence of 100 μM IBMX in the absence and presence of 0.01 μM dDAVP, 0.1 μM dDAVP or 10 μM forskolin. The reaction was terminated with 1 ml of 6% ice-cold trichloroacetic acid, and cAMP was extracted by freeze thawing. The acid was removed by ether extraction, and cAMP was measured in triplicates by radioimmunoassay using 10 μl aliquots of reconstituted samples. The results were expressed in picomoles per mg protein.

Immunofluorescence
HBO cells (1 x 10 4 cells/well) were plated into 8-wells chamber slides and treated with 10 μM 8-CPT-cAMP or 10 μM 8-CPT-cGMP or 10 nM dDAVP for 10 min. Following that HBO cells were fixed with 4% PFA for 10 min at 4˚C. The procedure of immunofluorescence staining was same as described for taste bud cells above. After washing with 1X PBS for 5 min and blocking with 3% donkey serum for 1h at room temperature, cells were stained with rabbit antiserum containing γ-rENaC antibody (1:400 dilution in 3% donkey serum) or rabbit polyclonal δ-ENaC antibody (1:100 dilution in 3% donkey serum) at 4˚C for overnight. After washing, cells were incubated with fluorescent-conjugated secondary antibodies for 1h at room temperature. Nuclei were visualized with 1 μg/ml of 4',6-diamidino-2-phenylindole (DAPI). Images were acquired with a 40X oil immersion objective on a Zeiss LSM 700 confocal laser scanning microscope and processed using Photoshop software (Adobe System). To quantitate the number of δ-ENaC positive HBO cells, the confocal images were divided into 4 regions of interest (ROIs). In each ROI, the total number of cells were counted as DAPI labelled nuclei and the δ-ENaC positive cells as green fluorescence labelled cells.

Separation of ENaC-dependent and ENaC-independent NaCl CT response
In adult rats, the NaCl CT response is composed of two components (Fig 1). One component is blocked by CPC and the second component is blocked by Bz. Bz is a specific blocker of ENaC. Thus, the Bz-sensitive response represents the component of the NaCl CT response that arises due to Na + influx through apical ENaCs in salt sensing FF TRCs. The Bz-insensitive response is blocked by CPC. It represents the ENaC-independent component of NaCl CT response. This component most likely arises due to apical Na + influx through non-selective cation channel(s) in TRCs that are sensitive to CPC [50]. To investigate specifically changes in ENaC-dependent NaCl CT responses during postnatal development, NaCl CT responses in 19-23 day old rats were monitored in the presence of CPC. CPC completely eliminates the ENaC-independent component of the NaCl CT response.

Effect of 8-CPT-cAMP and 8-CPT-cGMP on NaCl CT responses in 19-23 day old rats
Under open-circuit conditions, in 19-23 day old rats, the CT response increased with increasing NaCl concentration from 0.1 to 0.5 M (Fig 2A and 2F). Adding CPC to each NaCl stimulus eliminated most of NaCl CT response (Fig 2B and 2F). This suggests that in 19-23 day old rats a small number of functional apical ENaCs are present in FF TRCs that contribute to a small NaCl CT response just above the rinse baseline level. Following topical lingual application of 20 mM 8-CPT-cAMP for 10 min, the response to NaCl was significantly enhanced relative to control (Fig 2C and 2F). The enhanced response was partially blocked by CPC (Fig 2D and  2F), but was inhibited to the rinse baseline level in the presence of CPC + Bz ( Fig 2E). This suggests that in 19-23 day old rats, the cAMP-induced enhanced NaCl CT response is the result of ENaC activation. The post-8-CPT-cAMP NaCl CT response (Fig 2C), consisted of two components, an ENaC-dependent (Bz-sensitive) component and a Bz-insensitive but CPC-sensitive component. No change in the Bz-insensitive NaCl CT response ((NaCl)-(NaCl + CPC)) was observed at any NaCl concentration ( Fig 2F; grey bars). In 19-23 day old rats treated with 20 mM 8-CPT-cAMP, the mean normalized ENaCdependent tonic CT response at 0.1 M NaCl was not different from the ENaC-dependent tonic CT response at 0.1 M NaCl in control adult rats ( Fig 2G). However at 0.3 and 0.5 M NaCl, post-8-CPT-cAMP-dependent NaCl tonic CT responses in 19-23 day old rats remained significantly lower than ENaC-dependent NaCl tonic CT responses in control adult rats (Fig 2G).
In contrast to 8-CPT-cAMP, topical lingual application of 20 mM 8-CPT-cGMP for 10-15 min had no effect on the NaCl CT responses in 19-23 day old rats (data not shown). These results indicate that NaCl CT responses are specifically enhanced by cAMP.
Under control conditions, in 19-23 day old rats, the Bz-sensitive component (NaCl + CPC; r aso (0)) was a minor component at all NaCl concentrations tested (Fig 3A; ). In the presence of both Bz and CPC the CT response decreased to the baseline rinse level (Fig 3A; ▲). This shows that the solid curve for the whole NaCl CT response (r o (0); Eq 3) in Fig 3A (■) is, therefore, the sum of curves derived for r aso (0) using Eq 1 representing the Bz-sensitive component, and r aio (0) using Eq 2 representing the amiloride-and Bz-insensitive component.
The changes in the mean open-circuit CT response to NaCl + CPC (r aso ) as a function of NaCl concentration before and after the topical lingual application of 20 mM 8-CPT-cAMP in 19-23 day old rats are shown in Fig 3B. The data were fitted to Eq 1. The values of the maximal ENaC-dependent response (r asmc ) and K asc for the control curve and for the post-8-CPT-cAMP curve are shown in Table 2. Following 8-CPT-cAMP treatment r asm increased (p = 0.0001) and K as decreased.
As shown in Fig 2F (grey bars), relative to control, 20 mM 8-CPT-cAMP treatment did not produce any significant changes in the Bz-insensitive NaCl CT response ((NaCl)-(NaCl + CPC)). Therefore, the normalized values of the Bz-insensitive NaCl CT response at each NaCl concentration before and after 8-CPT-cAMP treatment were combined to estimate the parameters in Eq 2 for r aio . Using the least squares fit of the combined data points gave r aim and K ai values also shown in Table 2. These results indicate that the cAMP effects on the NaCl CT response in 19-23 day old rats are independent of the Bz-insensitive component of the CT response.

Concentration-dependence of 8-CPT-cAMP on r asm and K as
In 19-23 day old rats, CT responses to NaCl + CPC were obtained before and after topical lingual application of 5, 10, 15, and 20 mM 8-CPT-cAMP ( Fig 4A). The data were fitted to Eq 1 and yielded values of r asm and K as corresponding to a particular 8-CPT-cAMP concentration, p. The Bz-sensitive maximum CT response, r asm , was a saturating sigmoidal function of 8-CPT-cAMP concentration as p ! 1 (Fig 4B). The maximum increase in r asm relative to r asmc , was found by least squares fit of the data to Eq 8, where it is expressed by parameter a, whose value was 0.417 ± 0.081 (Table 3). At zero 8-CPT-cAMP r asm was 0.191 ± 0.043 (from Fig 3B). The ratio a/r asmc is 2.18 (218%) and represents the 8-CPT-cAMP-induced increase in r asm relative to r asmc . This ratio is slightly less than the value 2.96 (0.565/0.191) calculated from values shown in Table 2. In the ENaC channel model [51], r asm is proportional to the density of functional apical ENaCs. This suggests that in 19-23 day old rats, 8-CPT-cAMP increases the functional ENaC density by 218%. The parameter k expresses the 8-CPT-cAMP concentration at which the increase in r asm relative to r asmc is one-half maximal ( Table 3). The large n value (Table 3) is an expression of the high threshold 8-CPT-cAMP concentration applied to the tongue (about 10 mM) before a cAMP effect is observed. As cAMP concentration (p) increases K as decreases ( Fig 4C). The curve is the least squares fit to Eq 9.

Fig 1. Effect of CPC and Bz on the NaCl CT response in adult rats. (A)
Shows a representative NaCl CT response in an adult (60 + day old) rat. CT responses were monitored under open-circuit conditions while the tongue was first rinsed with a rinse solution R (10 mM KCl) and then with 0.3 M NaCl. The CT responses were also monitored with NaCl stimulating solutions containing 2 mM CPC or 5 μM Bz or 2 mM CPC + 5 μM Bz. (B) Shows the mean normalized tonic CT responses to NaCl, NaCl + CPC, NaCl + Bz and the added response ((NaCl + CPC) + (NaCl + Bz)) from 6 rats. The data were normalized to 0.5 M NH 4 Cl tonic response. *p = 0.0031; **p = 0.0001. The mean tonic CT responses to NaCl + CPC and NaCl + Bz were also significantly different with p = 0.0002. In a 19 day old rat, the ENaC-dependent open-circuit CT response to 0.3 M NaCl was small relative to baseline and virtually insensitive to the applied voltages ( Fig 5A). Following treatment with 20 mM 8-CPT-cAMP for 10 min, the same rat demonstrated a significantly increased NaCl CT response and marked sensitivity to voltage (Fig 5B). The ENaC-dependent NaCl CT response varied linearly with applied voltage between -60 mV and 60 mV ( Fig 5C) [21,55]. Treatment of the tongue with 8-CPT-cAMP increased both r aso (the open-circuit response to NaCl + CPC), but also the rate of increase in r as per unit change in translingual epithelial voltage (response conductance, κ as ). For each condition r as is described by Eq 5, and the calculated parameters are given in Table 4. In 19-23 day old rats, 20 mM 8-CPT-cAMP treatment increased κ as by a factor of 11.7 (0.00367/0.000313; Table 3).
Since r asm , and K as are known from open-circuit analysis (Fig 4B and 4C), the response conductance (Eq 6) yields values of δ (0.14) under control conditions and (0.93) post-20 mM 8-CPT-cAMP (Table 4). Thus, only 14% of the clamp voltage is dropped across functional apical membrane ENaCs under control conditions and 93% of the voltage is dropped across the significantly increased number of functional ENaCs post-20 mM 8-CPT-cAMP treatment.
Effect of 8-CPT-cAMP and 8-CPT-cGMP on behavioral responses to NaCl in the developing rats Two bottle/24h NaCl preference tests were performed on 21 day old rats following topical lingual treatment with H 2 O or 20 mM 8-CPT-cAMP for 20 min (Fig 6A). The control and treated rats were housed in two separated cages and were given access to H 2 O and 0.15 M NaCl and their combined fluid intake was measured each day every 24h. In control rats, the mean NaCl preference for a period of 6 days (postnatal day 22-27) was 0.58 ± 0.02 (N = 3). In 8-CPT-cAMP treated rats (N = 4), there was a significant decrease in H 2 O intake and a significant increase in NaCl intake, and the mean NaCl preference for the 6 day period was 0.82 ± 0.018 (P = 0.0001; N = 4).
In a separate experiment, rat (21 day old) tongues were treated with H 2 O (N = 5), 20 mM 8-CPT-cAMP (N = 5) or 20 mM 8-CPT-cGMP (N = 5) for 20 min. The above 3 groups of rats were housed in three separated cages and were given a choice between H 2 O and 0.035 M NaCl or H 2 O and 0.075 M NaCl. Their combined fluid intake was measured each day. For the group of rats whose tongues were treated with H 2 O, 8-CPT-cAMP or 8-CPT-cGMP the mean combined NaCl preference at 0.035 and 0.075 M NaCl was 0.56 ± 0.055, 0.90 ± 0.039 (P = 0.0001), and 0.45 ± 0.094, respectively.   (Fig 7). The control and treated rats were housed individually in separated cages and were given access to H 2 O and varying concentrations of NaCl (0.01-0.5 M) and their fluid intake was measured each day. There was no difference between H 2 O intake and NaCl intake between 8-CPT-cGMP treated rats and control rats. Accordingly, the data from both groups was combined. In control rats and in rats treated with 20 mM 8-CPT-cGMP, the NaCl intake increased linearly between 0.01 and 0.15 M NaCl (Fig 7; grey bars). At 0.25 and 0.5 M NaCl the NaCl intake was significantly reduced relative to 0.15 M NaCl. In rats treated with 20 mM 8-CPT-cAMP, the NaCl intake increased linearly between 0.01 and 0.10 M NaCl (Fig 7; black bars). At 0.15, 0.25 and 0.5 M NaCl the NaCl intake was significantly reduced relative to 0.1 M NaCl. These results suggest that in both young and adult rats, the behavioral responses to NaCl are specifically modulated by 8-CPT-cAMP and not by 8-CPT-cGMP.

Age-dependent increase in ENaC expression
In the anterior taste receptive field ENaC is expressed in a subset of FF TRCs and in squamous epithelial cells. Fungiform taste buds are difficult to separate without contamination from the surrounding non-taste tissue. Accordingly, we investigated ENaC expression in pure taste buds isolated from CV taste papillae from young and adult rats. In our Q-PCR studies, α-and γ-ENaC mRNA levels were significantly higher in adult CV taste buds than in 15 day old rats ( Fig  8A). Relative to adult rats, in CV taste bud lysates a significantly lower expression of α-ENaC protein was observed in 14-18 day old rats (Fig 8B). These results indicate that during development there is an age-dependent increase in ENaC expression in CV TRCs (Fig 8C). We hypothesize that a similar age-dependent increase in ENaC expression also occurs in FF TRCs.

Spaciotemporal relationship between cAMP formation and ENaC trafficking
The spaciotemporal relationship between cAMP formation and ENaC trafficking was investigated in HBO cells. In HBO cells, mRNAs for α-, β-, γ-, and δ-ENaC subunits were detected by RT-PCR ( Fig 9A) and δ-ENaC protein by Western blot (Fig 9B). HBO cells treated with dDAVP for 10 min produced a dose-dependent increase in intracellular cAMP (Fig 10). In control HBO cells δ-ENaC antibody binding was observed in a subset of cells (Fig 11; Panel A). The antibody binding was observed mainly in the cytosolic compartment (Fig 11; Panel B). Out of 113 cells examined, 27 cells were δ-ENaC positive (23.9%). In contrast, incubating HBO cells with 10 μM 8-CPT-cAMP (Fig 12; Panel A) or 10 nM dDAVP (Fig 12; Panel B) for 10 min, resulted in trafficking of δ-ENaC from the cytosolic compartment to the apical compartment of a subset of HBO cells. However, similar to the control HBO cells (Fig 11), in HBO cells treated with 10 μM 8-CPT-cGMP for 10 min, the δ-ENaC antibody binding was observed mainly in the cytosolic compartment of a subset of HBO cells (Fig 12; Panel C).
We also observed γ-ENaC antibody binding to a subset of HBO cells, mainly in the cytosolic compartment (Fig 13; Panel A). Treating HBO cells with 10 μM 8-CPT-cAMP for 10 min induce trafficking of γ-ENaC subunit from the cytosolic compartment to the apical compartment (Fig 13; Panel B). These results show that dDAVP and cAMP induce trafficking of δ and γ ENaC subunits from the cytosolic compartment to the apical membranes of salt sensing HBO cells. We hypothesize that dDAVP and cAMP also induce trafficking of α and β ENaC subunits to form functional ENaCs in the apical membrane of HBO cells.
In addition to γ and δ ENaC subunit trafficking, dDAVP and 8-CPT-cAMP also induced changes in the cell shape, resulting in a slight narrowing of HBO cells.

Effect of AVP and dDAVP on NaCl CT responses in adult rats
In adult rats, a single tail vein injection of AVP (1 nano moles/Kg BW) enhanced the CT response to 0.1 and 0.3 M NaCl 20 min post-AVP injection (Fig 14A). In additional rats, a single tail vein injection of dDAVP (0.1 nano moles/Kg BW) enhanced the CT response to 0.1 and 0.3 M NaCl 20 min post-dDAVP injection. The increase in NaCl CT response was specifically due to an increase in the ENaC-dependent component of the NaCl CT response ( Fig  14B).

Effect of dDAVP on ENaC expression and trafficking in rat FF TRCs
In dDAVP-injected 15 day old rats the γ-ENaC antibody binding was observed both in the cytoplasmic compartment and in the apical membranes of a subset of FF TRCs 24h post   (Fig 15B and 15C; arrows). In rats injected with saline, the antibody binding was seen mainly in the cytosolic compartment in a subset of FF TRCs (Fig 15A). In addition, the γ-ENaC binding was more intense in dDAVP-injected rats relative to control rats. This indicates that dDAVP administration enhances γ-ENaC expression in FF TRCs and induces trafficking of γ-ENaC from cytosolic compartment to the apical membrane.

Age-dependent increase in V2R expression
In adult rats, V2R antibody demonstrated specific binding to kidney cortical collecting duct cells (Fig 16A) and to a subset of CV taste bud cells (Fig 16C). In CV TRCs V2R labeling was predominantly intracellular with some staining in the basolateral membrane domains ( Fig  16C) [61]. While in adult rats, a subset of CV TRCs demonstrated V2R antibody binding ( Fig  16C), in 14 day old rats, only a faint V2R antibody binding was observed in CV TRCs (Fig 16E  and 16G). These results indicate that there is an age-dependent increase in V2R expression in TRCs.
Taken together, the above results indicate that there is a spaciotemporal relationship between AVP/dDAVP binding to V2Rs in the basolateral membrane of FF TRCs, activation of adenylyl cyclase, increase in intracellular cAMP, ENaC trafficking and the increase in the NaCl CT response and behavior.

Discussion
In a previous study [51] tonic NaCl CT responses in rats under open circuit conditions increased in a progressive manner between 12-14 and 29-31 postnatal days. In 29-31 day old rats tonic NaCl CT responses were not different from 60+ day old rats. In addition, the sensitivity of the NaCl CT responses to applied ±60 mV across the lingual epithelium increased with age. The data were fitted to an apical Na + channel kinetic model that predicted that in FF TRCs the apical Na + channel density increased with age monotonically [51]. However, in this study, developmental changes in NaCl CT response and Na + channel density were calculated using NaCl CT responses in the absence of amiloride or Bz. In the present study, we specifically monitored the ENaC-dependent NaCl CT response in the presence of CPC, a blocker of the amiloride-insensitive component of the NaCl CT response (Fig 1).
Our results demonstrate that in 19-23 day old rats, an incremental increase in FF TRC cAMP induced a dose-dependent increase in the magnitude of the ENaC-dependent NaCl CT response (Fig 4A). The ENaC-dependent maximum CT response (r asm ) was a saturating   (Fig 4B). The large n value (Fig 4B; Table 3) indicates that a high threshold concentration of intracellular cAMP must be reached before a cAMP effect is observed on r asm (Fig 4B). With an increase in 8-CPT-cAMP concentration there was a near linear decrease in K as (Fig 4C). Consistent with these results, a saturating sigmoidal relationship was observed between the maximum NaCl CT response (CT max ) and the postnatal age of rat pups. And the apparent affinity of ENaC for Na + declined in rats between 12-14 and 19-23 postnatal days [51]. In 19-23 day old rats, 20 mM 8-CPT-cAMP increased the maximum ENaC-dependent NaCl CT response (r asm ) by a factor of 2.18 ( Fig 4B). In the Na + channel model [51], r asm is proportional to the density of functional ENaCs. This suggests that in 19-23 day old rats, 8-CPT-cAMP increases the functional ENaC density by 218%. An increase in r asm was accompanied by an increase in the voltage sensitivity of ENaC-dependent NaCl CT response (Fig 5).

Fig 8. Age-dependent increase in ENaC expression. (A)
CV papillae from naïve rats at different age were collected for Q-PCR to determine the age dependent mRNA level of α-ENaC and γ-ENaC. The results showed the mRNA levels of both subunits is higher in adult rat CV papillae compared with the pups. (B) For Western blot studies, CV papillae were pooled from fifteen 14-18 day old rats and eight adult (60+ day old) rats. Forty μg total protein was used for the assay. Beta-actin was used as a protein loading control. (C) Shows the intensity of the α-ENaC band in 14-18 day old rats relative to adult rats from 3 different batches of CV tissues normalized to β-actin (p = 0.0001).
doi:10.1371/journal.pone.0171335.g008  (Table 4). We have previously shown that in adult rats, 8-CPT-cAMP enhanced ENaC-dependent NaCl CT response and the voltage sensitivity of the response at ±60 mV [21]. In adult rats, the control (κ as0 ) and the post-cAMP (κ ascAMP ) mean values at 0.1 M NaCl normalized to 0.3 M NH 4 Cl were 0.0046 and 0.0071 response units/mV, respectively (κ ascAMP /κ as0 = 1.54) [21]. The corresponding values of κ as0 and κ ascAMP in 19-23 day old rats for 0.1 M NaCl also normalized to 0.3 M NH 4 Cl were 2.4 x 10 −4 and 5.1 x 10 −3 response units/mV, respectively (κ ascAMP /κ as0 = 21.3). This large difference in κ asAMP /κ as0 between 19-23 day old rats and adult rats is due to the fact that in developing rats the starting functional ENaC levels are just detectable and cAMP brings them up to adult levels, while in adult rats ENaC levels are closer to their maximum value to start with.
Both in adult rats [21] and in 19-23 day old rats, 8-CPT-cAMP did not have any effect on the Bz-insensitive NaCl CT response (Fig 2F). Therefore, the cAMP-induced increase in NaCl CT response is specifically due to an increase in the ENaC-dependent component of the NaCl CT response. Consistent with this, in rats the amiloride-insensitive NaCl CT response is functional at birth and appears not to vary between 12 and 100 days postnatal [11,62]. Taken together, the above results indicate that in 19-23 day old rats, an incremental increase in intracellular cAMP mimics the age-dependent increase in the magnitude of NaCl CT response, voltage-sensitivity, and increase in functional ENaCs in the apical membrane of salt sensing FF TRCs.
In our previous studies [63,64], topical lingual application of 8-CPT-cAMP also enhaned the CT response to strong acids (e.g. HCl) by activating a proton conductance in the apical membrane of sour sensing TRCs. However, no effect of 8-CPT-cAMP was observed on CT responses to sucrose or NaCl + Bz. These studies indicate that 8-CPT-cAMP enhances salty  and sour taste responses by activating two different ion channels, ENaC and proton channel (s), respectively, expressed in non-overlapping salty and sour sensing TRCs [20].
In 19-23 day old rats topical lingual application of 20 mM 8-CPT-cAMP enhanced the ENaC-dependent NaCl CT response at 0.1 M NaCl to the level observed in control adult rats ( Fig 2F). However, the post-cAMP responses in 19-23 day old rats at 0.3 and 0.5 M NaCl remained significantly lower than in adult rats. The ability of cAMP to enhance ENaC-dependent NaCl CT response in 19-23 day old rats is limited by the expression levels of ENaC subunits in FF TRCs. An increment in ENaC expression and activity is seen in TRCs (Fig 8), reabsorptive epithelia and non-epithelial tissues in a development-dependent manner [26]. Both AVP (Fig 15) and aldosterone [15] have been shown to increase ENaC expression in TRCs and in other epithelial tissues [37][38][39]. We hypothesize that as TRC ENaC expression increases with postnatal age, 20 mM 8-CPT-cAMP will enhance the ENaC-dependent NaCl CT profile to the level observed in adult rats over the whole range of NaCl concentrations. The presence of amiloride-sensitive whole cell currents in FF TRCs isolated from rats as young as 2 days old has been demonstrated [65]. These results suggest that ENaC is largely in place and functional in rat neonates that nevertheless do not have a detectable ENaC-dependent NaCl CT response until 7-10 days postnatal. Unlike the mice in which α-ENaC is knocked out globally [29], mice in which α-ENaC has been conditionally knocked out only in TRCs, survive. These mice demonstrate the lack of ENaC-dependent NaCl CT response and preference for appetitive NaCl concentrations [20]. These results indicate in salt sensing FF TRCs both ENaC expression and ENaC trafficking to the apical membrane is necessary for salt taste development.
Placing female pregnant rats on a Na + -restricted (0.01% NaCl) diet produces offspring that do not develop functionally active TRC ENaC, provided they are also maintained on the Na +restricted diet into maturity [66]. Normal responses were restored with a latency of about 20 days by providing the Na + -restricted rats a one-time access to saline [67]. These studies led to the hypothesis that the ingestion of physiological saline initiated systemic changes leading to hormonal action on the taste system and the activation of TRC ENaC. Rats reared on high (3%) NaCl diet from conception to postnatal day 30 demonstrated greater amiloride-sensitive NaCl responses than rats reared on 1% NaCl diet but gave normal responses when switched to 1% NaCl diet [68,69]. These results further support the hypothesis that hormonal changes during development regulate TRC ENaC expression and trafficking, and thus the neural and behavioral responses to NaCl.

Spaciotemporal relationship between cAMP generation and ENaC trafficking
In isolated TRCs, the amiloride-sensitive Na + current is increased by AVP and membrane-permeable cAMP analogues. This effect was predominantly due to an increase in the number of functional apical ENaCs [34]. The maximum increase in the Na + -current was observed 15-20 min post-AVP treatment. These results indicate that both AVP-induced increase in intracellular cAMP and ENaC trafficking is a rapid process that occurs in the time frame of minutes. We investigated the spatiotemporal relationship between cAMP generation and ENaC trafficking in HBO cells. HBO cells demonstrated the mRNAs for α-, β-, γ-, and δ-ENaC subunits (Fig 9) and δ-and γ-ENaC proteins (Figs 11-13) [70]. HBO cells treated with 0.1 μM dDAVP or 10 μM forskolin (an activator of adenylyl cyclase) for 10 min increased cAMP (Fig 10), and cAMP, in turn, induced trafficking of δ-and γ-ENaC from cytosolic compartment to the apical compartment in ENaC positive cells (Figs 11-13). Consistent with this, cAMP increased the density of ENaC subunits in the apical membrane of MDCK cells in direct proportion to amiloride-sensitive Na + transport [38]. Cyclic AMP rapidly mobilized wild type human renal ENaC expressed in Xenopus laevis oocytes and increased the membrane density of ENaCs in oocyte membranes but had no effect on the Liddle-mutated hENaC [46]. In our studies intravenous injections of dDAVP induced increased expression and trafficking of γ-ENaC from the cytosolic compartment to the apical membrane of rat FF TRCs (Fig 15). It is important to note that even though 15 day old rats express low levels of TRC V2Rs, repeated injections of high doses of dDAVP (1 nano moles/Kg BW) can induced ENaC expression and trafficking. These results suggest that TRC ENaC is regulated by similar intracellular signaling mechanisms in rodents and humans.
The effect of cAMP on ENaC trafficking is due, at least in part, to inhibition of neural precursor cell expressed developmentally down-regulated protein 4-2 (Nedd4-2)-induced ENaC ubiquitination via phosphorylation of Nedd4-2 on Ser 327 , Ser 221 , and Thr 246 . Notably, these are the same residues that are phosphorylated by serum and glucocorticoid-inducible kinase-1 (SGK-1) and protein kinase A (PKA), although the cAMP effect is SGK independent [71,72]. Nedd4-2 controls ENaC surface expression by catalyzing its ubiquitination, which targets ENaC for degradation [41,45]. Nedd4-2 binds to the PY motif located in the cytoplasmic C terminus of each ENaC subunit decreasing ENaC surface expression. SGK-1 binds to Nedd4-2 and phosphorylates it which reduces the binding of Nedd4-2 to ENaC [71][72][73][74][75]. Aldosterone acts through the mineralocorticoid receptor to alter the transcription of specific genes, including SGK-1 [42]. Aldosterone has been shown to increase ENaC expression and translocation of ENaC subunits from the cytosolic compartment to the apical membrane of TRCs [15].
In isolated TRCs from adult mice, insulin (5-20 nM) enhanced Na + influx in both patchclamp and imaging studies. Adult mice injected with insulin showed significant avoidance of NaCl at lower concentrations than the control group [30]. Stimulation of ENaC by insulin involves activation of SGK-1 via phosphoinositide (PI)3-kinase and phosphoinositide-dependent kinase (PDK1) [75]. SGK-1 has been shown to be expressed in vallate and FF taste bud cells [30]. Insulin had no effect on Na + transport in taste cells from SGK −/− mice. In contrast, taste cells from SGK +/+ mice evoked greater Na + responses to insulin. Functional ENaC activity was dramatically reduced in SGK −/− taste cells. These studies suggest that SGK is essential to maintain normal ENaC function and that the absence of SGK protein in the null mice severely inhibits insulin's effects in the gustatory system [30]. Thus, it is likely that in addition to cAMP and AVP, aldosterone and insulin also play a role in postnatal development of FF TRC ENaC.
Hormonally induced increase in intracellular cAMP or increasing cell cAMP by direct application of membrane permeable forms of cAMP has been shown to induce a change in cell shape in both epithelial and non-epithelial cells [76][77][78][79][80]. In some studies [77][78][79], changes in cell shape were observed following incubation of cells with 1 or 2 mM dibutyryl-cAMP or 8-CPT-cAMP. When mesangial cells were treated with 1 mM dibutyryl-cAMP, 85% of cells underwent a shape change by 40 min [76]. Treating cultured rat ovarian granulosa cells with dibutyryl-cAMP resulted in a dose-dependent increase in the percentage of cells that demonstrated morphological change [78]. The above studies suggest that the changes in cell shape induced by an increase in intracellular cAMP is both dose-and time-dependent. In the present study, HBO cells were treated with 10 μM 8-CPT-cAMP for 10 min, a concentration 100 times lower than that used in the above studies. Under these conditions, we expect to observe some changes in the morphology of a subset of HBO cells (Figs 11-13). While there is some change in the cell shape after cAMP or dDAVP treatment, the significant finding is the translocation of γ-and δ-ENaC from the cytosolic compartment to the apical compartment in a subset of HBO cells. It is likely that changes in cell morphology and ENaC subunit translocation from the cytosolic compartment to apical compartment occur in the same time frame. Elevation of intracellular cAMP resulted in a change in cell shape that resembled dome formation in cultured rat glomerular epithelial cells [77]. In Madin-Darby canine kidney epithelial cells (MDCK) an increase in cell cAMP caused an increase in the number and size of hemicysts [80]. This was attributed to a decrease in cell adhesion to the substrate [81]. Cyclic AMPinduced morphological changes are counteracted by the activated RhoA small GTPase and the Rho kinase ROKα [79]. The confocal images are overlay of transmitted image, nuclei labelled with DAPI (blue) and secondary antibody IgG-CFL 488 binding (green). V2R antibody binding is shown in adult (90+ day old) rat CV taste bud cells with overlay of DAPI and green fluorescence (C). The corresponding NC without primary antibody is an overlay of transmitted image, DAPI and green fluorescence (D). The V2R antibody binding is shown in CV taste bud cells from 14 day old rat pups (E and G). The confocal images are overlay of DAPI and green fluorescence. The corresponding transmitted images (T) are shown in (F and H). In two bottle preference tests, 19-23 day old rats did not discriminate between H 2 O and NaCl concentrations of 0.035, 0.075 and 0.15 M. However, after a single topical lingual application of 20 mM 8-CPT-cAMP for 20 min the 19-23 day old rats demonstrated a clear preference for the above NaCl concentrations (Fig 6). The effects of 8-CPT-cAMP persisted for 5-6 days. 8-CPT-cAMP also produced behavioral changes in adult rats. While in control or 8-CPT-cGMP treated adult rats the maximum intake was observed at 0.15 M NaCl, in 8-CPT-cAMP treated rats the maximum intake was observed at 0.1 M NaCl (Fig 7). This indicates that in adult rats, cAMP-induced increase in TRC ENaC results in shifting the NaCl preference to lower NaCl concentrations. Thus, even in adult rats 8-CPT-cAMP effects could be observed on NaCl behavior over a 7 day period. It is likely that 8-CPT-cAMP is resistant to phosphodiesterases, and therefore, persists intracellularly for extended periods, and thus, maintains the elevated levels of functional apical ENaCs in TRCs. Alternately, during development the cAMP-induced increase in functional apical ENaCs may have a lower rate of degradation due to hormone induced decrease in Nedd4-2 abundance [82].
In summary, the results presented here demonstrate that an incremental increase in FF TRC cAMP mimics the postnatal age-dependent increase in NaCl CT response, increase in functional apical ENaCs, and preference for appetitive NaCl concentrations. AVP binds to V2R coupled to Gα s . Gα s stimulates adenylyl cyclase and increases cAMP. During postnatal development an AVP-induced incremental increase in cAMP occurs due to an age-dependent increase in V2Rs in the basolateral membrane of TRCs (Fig 16). This leads to a gradual increase in Gα s -induced activation of adenylate cyclase and an incremental increase in TRC cAMP.
Voltage effects and response conductance. We have previously shown that in adult rats, 8-CPT-cAMP increased the maximal open-circuit NaCl CT response (r asm ) by increasing the response conductance (κ as ). The response conductance is defined as the slope of the CT response curve measured by varying the lingual clamp voltage at a fixed NaCl concentration [21,55]. The response (r) is approximately linear with applied voltage (V) (see Fig 5C). Accordingly, the voltage dependence of the response in Eqs 1 and 2 can be linearized and total response (r) appears as follows: where, κ is the total response conductance: Here, and k ai ¼ r aio Fl RT From Eqs 6 and 7 we note that the response conductance is proportional to the open-circuit response magnitude for each ion channel type. cAMP concentration dependence of r asm and K as . The 8-CPT-cAMP-induced increase in r asm was fitted to a sigmoidal saturating function of 8-CPT-cAMP concentration to: r asm ¼ r asmo þ a p n k n þ p n ð8Þ and a decrease in K as was fitted to: In Eq 8 r asmc is the value of r asm when the 8-CPT-cAMP concentration, p, is zero, and a, k, and n are positive constants. In Eq 9 K as is the NaCl concentration (M) at which r aso becomes halfmaximal and p is the 8-CPT-cAMP concentration (mM) applied to the rat tongue. The parameter b is the K as value in the untreated tongue for which p = 0, q is the 8-CPT-cAMP concentration for which K as reaches half its untreated value (i.e. value at p = 0), and n is a positive constant.