|Home | About | Journals | Submit | Contact Us | Français|
The taste-selective G protein, α-gustducin (α-gus) is homologous to α-transducin and activates phosphodiesterase in vitro. α-Gustducin-knockout mice are compromised to bitter, sweet and umami taste stimuli, suggesting a central role in taste transduction. Here, we suggest a different role for Gα-gustducin. In taste buds of α-gustducin knockout mice, basal (unstimulated) cAMP levels are high compared to those of wild-type mice. Further, H-89, a cAMP-dependent Protein Kinase inhibitor, dramatically unmasks responses to the bitter tastant denatonium in gustducin-lineage cells of knockout mice. We propose that an important role of α-gustducin is to maintain cAMP levels tonically low to ensure adequate Ca2+ signaling.
Bitter, sweet, and umami (glutamate) taste stimuli are transduced by taste G protein-coupled receptors (taste GPCRs) and downstream signaling effectors. Two families of taste GPCRs exist, the T1Rs, for sweet and umami, and T2Rs, for bitter [1-3]. Taste GPCRs activate heterotrimeric G proteins that contain Gβ3 and Gγ13 . The heterodimer, β3γ13, is released from tastant-bound receptors, where it stimulates phospholipase Cβ2 (PLCβ2) [5,6] to produce inositol trisphosphate (IP3) and activate the type 3 IP3 receptor, IP3R3 [7,8] to release stored Ca2+. Mice in which these effector genes are knocked out show taste afferent nerve and behavioral responses to sweet, bitter and umami stimuli that are either eliminated in the case of PLCβ2, , or compromised in the case of IP3R3, . These findings underline the central roles of PLCβ2 and IP3R3 in taste transduction.
Gα-gustducin (α-gus)  is a frequent partner of Gβ3γ13 and αGus-/- mice are compromised to bitter, sweet, and umami stimuli . Yet, despite its discovery over 15 years ago, the precise role of α-gus in taste transduction is still unclear. As with the closely related Gα-transducins, effector-interacting peptides derived from α-gus can activate a retinal PDE in vitro to decrease cyclic nucleotide levels [12,13].
We considered the possibility that α-gus in taste cells may regulate cAMP levels in a continuous fashion, in the absence of taste ligands. If this were the case, genetic ablation of α-gus should markedly alter basal levels of cAMP in taste cells. We directly tested this hypothesis by measuring cAMP levels in taste buds of α-gus+/+ and α-gus-/- mice. We selected circumvallate (CV) taste buds for this analysis because (a) taste buds are numerous, (b) earlier studies linking gus to bitter transduction were performed in CV and (c) α-gus couples primarily to bitter receptors in this taste field . We show that taste buds of α-gus-/- mice have highly elevated basal levels of cAMP relative to those in α-gus+/+ mice. Further, we show in taste buds from α-gus-/- mice, that elevated cAMP in α-gus-lineage cells activates cAMP-dependent Protein Kinase A (PKA) causing a chronic inhibition of Ca2+ responses to bitter stimuli. Our data lead us to propose a novel explanation of the α-gus knockout phenotype. We suggest that taste buds in α-gus-/- mice exist in a chronically depressed state, unable to generate robust release of stored Ca2+ in response to any of the taste GPCR-mediated taste qualities.
Mouse housing and experimental procedures were approved by Colorado State University’s Animal Care and Use Committee. Animals were killed by exposure to CO2 followed by cervical dislocation before tongues were removed. Adult α-gus-/- mice  and α-gus+/+ littermates were used for cAMP measurements. For Ca2+ imaging experiments, transgenic mice in which the α-gus promoter drives expression of GFP, i.e. gus-GFP , were crossed with α-gus-/- mice. GFP-positive, α-gus-negative progeny were identified. In these mice, taste cells of the α-gus lineage express the GFP label, while lacking α-gus itself.
Tyrode’s solution contained (in mM): 140 NaCl, 5 KCl, 1 MgCl2, 1 CaCl2, 10 HEPES, 10 glucose, and 1 pyruvate (pH 7.4 with NaOH). Calcium-Magnesium-free (CMF) Tyrode’s solution was similar to the above except that MgCl2 and CaCl2 were omitted (i.e. nominally CMF) or were replaced with 1 mM BAPTA for isolating taste buds. The PKA inhibitor, H-89, and the Protein Kinase C (PKC) inhibitor, bisindolymaleimide I (Bis, Calbiochem; San Diego) were diluted from dimethyl sulfoxide stocks into Tyrode’s solution before use.
Taste bud-enriched CV epithelia from α-gus-/- and α-gus+/+ mice were enzymatically delaminated, dissected free of adjacent non-taste epithelium, and were processed in parallel as we described previously [14,15]. Tissues were lysed to extract cAMP into a soluble supernatant and total cAMP in each tissue extract was measured using enzyme immunoassay (Amersham Biosciences, Piscataway, NJ) [14,15]. Total protein in each tissue piece was quantified using a Nano-Orange Kit (Invitrogen, Carlsbad, CA). In all cases, cAMP titer is presented normalized to total protein in the tissue.
α-gus was immunodetected in paraformaldehyde fixed cryosections, using a rabbit polyclonal anti-α-gus (Santa Cruz Biotechnology, Santa Cruz, CA; SC-395; 1:500) and Cy5 goat anti-rabbit secondary (Jackson ImmunoResearch Laboratories) . Controls for antibody specificity included omitting primary antibody and the lack of staining in taste buds of the α-gus-/- mouse.
CV taste epithelia were isolated and placed in CMF-Tyrode’s solution for 5 min. Taste buds were removed by gentle suction with a fire-polished pipet and plated onto cover slips coated with poly-L-lysine (Sigma) .
Intracellular Ca2+ measurements were obtained from fura-2-loaded taste cells as described previously . Images were acquired with the CCD Sensicam QE camera (Cooke Corp, Romulus, MI) using a 40X oil immersion objective lens. Emission at ~510nm was captured following sequential excitation at 350nm and 380nm. Calcium levels are reported as a ratio of fluorescence emissions, F350/F380, captured every 1-5 seconds using Imaging Workbench 5.2 (Indec Biosystems, Inc.). Denatonium (Sigma, St. Louis, MO), H-89 (10 μM, Calbiochem, San Diego, CA), and bisindolymaleimide I (0.15 μM, Calbiochem) were bath applied using gravity flow perfusion. In most experiments, denatonium was applied at 10mM, the lowest concentration that elicits significant afferent nerve responses and behavioral rejection in α-gus-/- mice , and is a sub-maximal concentration for nerve recordings in wild-type mice .
Paired and unpaired two-tailed t-test and Bonferroni corrections were performed using Prism v5.00 (GraphPad, San Diego CA).
α-Gus activates a retinal PDE in vitro [12,13]. If a similar activity occurs in taste cells, chronic absence of α-gus might result in altered cAMP levels. Hence, we measured total cAMP in CV taste bud-enriched epithelium from α-gus-/- and α-gus+/+ mice. We also carried out parallel measurements of cAMP levels in adjacent non-taste lingual epithelium as a control. The resting level of cAMP in α-gus+/+ circumvallate epithelium was 2.65 ± 0.35 (mean ± s.e.m.) pmole cAMP/μg protein. The value is similar to that measured from rat CV epithelium [14,15]. The basal level of cAMP was approximately 3.8 fold higher in CV epithelium from α-gus-/- mice as compared to α-gus+/+, a highly significant difference (p=0.0005; n=6; t-test; Figure 1). We attributed this difference to taste buds because adjacent regions of non-taste epithelium had much lower resting levels of cAMP, and these did not differ between genotypes (p=0.504; n=6; t-test). The data suggest that cAMP levels in taste buds are controlled by a PDE that is primarily regulated by α-gus, and may be active in the absence of taste stimulation.
Because α-gus-/- mice have elevated basal levels of cAMP, phosphorylation by PKA could also be chronically elevated and this could underlie the decreased bitter sensitivity observed in α-gus-/- mice . In that earlier study, it could not be determined whether residual responses to bitter tastants were in cells of the α-gus lineage. To resolve this uncertainty, we used mice in which GFP was expressed in α-gus lineage taste cells of α-gus-/- mice (Figure 2). Taste cells, loaded with fura-2, were stimulated with denatonium (10 mM). Denatonium elicited little or no increase in intracellular Ca2+ in the GFP-labeled taste cells of the α-gus-/- mice. To determine if elevated cAMP and PKA-dependent phosphorylation were responsible for this loss of sensitivity, we also examined bitter responses in the presence of the membrane-permeant PKA inhibitor, H-89 (10 μM). After treatment with H-89, Ca2+ responses in GFP-labeled α-gus-/- cells were larger (Figure 3A). Responses to denatonium were unmasked by H-89 in 7 of 21 GFP-labeled α-gus-/- taste cells that were previously unresponsive to denatonium (e.g. Figure 3A). On average, responses to denatonium in α-gus-/- cells were significantly enhanced, 6.5-fold in the presence of H-89 (Figure 3A,C). We noted that H-89 caused a slight elevation of the baseline Ca2+ in many taste cells, even prior to taste stimulation. This suggests that both basal Ca2+ levels and taste-evoked Ca2+ signals are regulated by PKA activity.
Next, we asked if H-89 had a similar enhancing effect on Ca2+ responses to denatonium in gus-expressing taste cells of α-gus+/+ mice. Surprisingly, we found that H-89 also significantly increased Ca2+ responses to denatonium in α-gus+/+ mice (2.9-fold; Figure 3B, C). After H-89 treatment, responses to denatonium were unmasked in 5 of 26 previously unresponsive Gus-GFP-positive cells.
The effect of H-89 on individual taste cells was reversible and repeatable (Figure 3D) in both knockout and wildtype mice. The vehicle, DMSO, had no effect on either resting Ca2+ level or denatonium responses (n=4; data not shown). Although 10μM H-89 is selective for PKA, it may also block PKC slightly. Thus, we applied a robust, membrane permeant inhibitor of PKC, bisindolymaleimide I (Bis) to GFP-labeled cells of α-gus+/+ mice. Bis (0.15 μM) had no effect on either resting Ca2+ levels or denatonium responses (Figure 3D; p=0.78; paired t-test; n=4), suggesting that the effect of H-89 on bitter responses arises from a specific block of PKA.
Ca2+ responses to 10mM denatonium were on average, 2.5-fold larger in GFP-labeled taste cells of gus+/+ mice relative to gus-/- mice (Figure 3C; a significant difference, p=0.0072; t-test with Bonferroni correction for 3 way comparison across genotypes, and for H-89 within each genotype). We infer that the depressed responses to denatonium in the α-gus-/- cells arise from the elevated basal levels of cAMP, which would constitutively stimulate PKA.
Two important findings are presented here. First, α-gus-/- mice have elevated basal (unstimulated) levels of cAMP. The result suggests that α-gus may be active and stimulate PDE in taste cells, even under basal conditions. Either α-gus has some constitutive (non-receptor dependent) activity, or α-gustducin couples to GPCRs (for tastants or physiological ligands) that have ligand-independent activity or are chronically stimulated. The loss of α-gus in knockout mice would allow cAMP to accumulate. Second, we show that in α-gus-lineage taste cells, cAMP-regulated PKA dramatically inhibits intracellular Ca2+ responses to bitter stimuli. This inhibition is PKA-selective; it could be overcome by a membrane permeant inhibitor of PKA (H-89), but not of PKC (Bis). Two key proteins that generate tastant-evoked Ca2+ responses, PLCβ2 and IP3R3, are both known to be inhibited by PKA-mediated phosphorylation  . Thus, an important function of α-gus may be to tonically maintain low levels of cAMP in taste cells. When α-gus is ablated, Ca2+ responses to bitter stimuli would be dampened through chronic phosphorylation and inhibition of both PLCβ2 and IP3R3. Even in wild-type cells, resting cAMP levels may be sufficiently high to maintain these proteins in a partially phosphorylated state, permitting modulation in either direction - enhanced or dampened responsiveness.
Wild-type taste buds stimulated with bitter tastants display Ca2+ responses in both gus-expressing, and non-expressing cells . Whereas taste buds from knockout mice contained fewer bitter-responsive cells, it was not possible to determine whether such cells belonged to the gus-lineage. Caicedo et al  suggested that α-transducin or αi might substitute for α-gus in the knockout. Our present findings suggest that those related Gα subunits are insufficient to keep cAMP levels low in taste cells. Further, we show that gus-lineage cells in the knockout have very small Ca2+ responses to bitter compounds, but that these can be unmasked by a PKA inhibitor, even in the absence of α-gus.
Margolskee and colleagues originally proposed that ligand-bound taste receptors, by activating Gα-gus, would stimulate PDE, trigger a drop in cyclic nucleotide levels, modulate the activity of cyclic nucleotide-gated (CNG) channels, and eventually alter membrane conductance . In frog taste cells, a cyclic nucleotide-suppressed CNG channel was reported  while in mammalian taste cells, a cAMP-activated cone CNG channel is expressed . However, neither of these conductances has been shown to be sensitive to taste stimulation.
Our data support another model (Figure 4). We propose that α-gus regulates PKA-mediated inhibition of the Ca2+ signaling effectors, PLCβ2 and IP3R3. Our model may explain why the PKA inhibitor, H-89, increases action potentials elicited by sweet stimuli , and Ca2+ responses to ATP in taste cells . We suggest that gus-expressing cells in taste buds of α-gus-/- mice exist in a chronically depressed state, unable to generate Ca2+ responses to any of the taste GPCR-mediated taste qualities. Acutely blocking cAMP-dependent phosphorylation renders knockout taste cells fully competent to respond to the tastant, denatonium. Thus, a key role for α-gustducin in taste may be to keep taste buds in an active state, ready to respond to taste stimuli with robust Ca2+ signals.
Supported by NIH grants DC00766 to SK; DC006021 to NC; DC003055 to RFM.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.