The Molecular Mechanisms of Taste
Summary: Robert Margolskee's laboratory is using molecular, transgenic, and structural techniques to study the peripheral and central mechanisms of taste transduction and coding and to determine how these pathways and circuits regulate gustatory behaviors in vivo.
Making Sense of Taste
The sensation of taste is initiated by the interaction of sapid molecules (tastants) with receptors and ion channels in the apical microvilli of taste receptor cells (TRCs). This phylogenetically primitive sense enables higher organisms to avoid toxins and find nutrients. Many taste transduction pathways convert chemical information into cellular second messenger codes utilizing cyclic nucleotides (cNMPs) or inositol trisphosphate (IP3). These messengers are typically part of a signaling cascade that leads to TRC depolarization and Ca2+ release. Responses to bitter, sweet, and umami (glutamate) compounds are transduced by specific receptors linked to guanine nucleotidebinding regulatory proteins (G proteins). Current psychophysical models suggest that taste is composed of five distinct qualities: sweet, sour, bitter, salty, and umami. How the vertebrate taste cell responds to a given tastant (signal transduction) and how this information is processed in the periphery and encoded in the gustatory areas of the brain (sensory coding) are both areas of keen interest to my research program.
We use bioinformatics, biochemistry, molecular cloning, structural biology, electrophysiology, and behavioral analysis of transgenic mice to identify and characterize functionally TRC components involved in taste transduction and coding. Using these methods, we have identified taste transduction elements of the following types: G proteincoupled receptors, G protein subunits, effector enzymes, and ion channels. Our goal is to understand the peripheral and central mechanisms of taste coding at the cellular and molecular levels and relate the functioning of these pathways and circuits to the regulation of complex gustatory behaviors in vivo.
Taste Transduction and Coding
G protein subunits. We have determined that the G protein subunits α-gustducin, α-transducin, γ13, β1, and β3 are expressed at high levels in TRCs. α-Gustducin is expressed selectively in TRCs and other chemosensory cells. Its biochemical properties closely resemble those of the α-transducins, including the ability to activate retinal phosphodiesterase (PDE6). α-Gustducin and α-transducin activate TRC-expressed PDE1A; gustducin's βγ component, β3γ13, activates the β2 form of phospholipase C (PLCβ2). Knockout mice lacking α-gustducin have diminished nerve and behavioral responses to bitter, sweet, and umami compounds, demonstrating α-gustducin's involvement in mediating the taste responses. Studies with α-gustducin/α-transducin double-knockout mice have shown that α-transducin does not mediate responses to bitter or sweet compounds, but plays a role in umami detection.
Effector enzymes and downstream signaling components. In response to bitter compounds, TRC cNMP levels drop (mediated by α-gustducin activation of PDE1A), while IP3 and diacylglycerol levels rise (mediated by β3γ13 activation of PLCβ2). PDE1A is coexpressed with α-gustducin in many TRCs. We have shown that this Ca2+/calmodulin-activated PDE1A can be activated by α-gustducin, α-transducin, and αi. G protein α-subunit regulation of PDE1A constitutes a previously unrecognized transduction mechanism of potential broad importance in taste and in other organ systems where Gi-type G proteins and PDE1A are coexpressed. β3, α13, PLCβ2, type 3 IP3 receptors, and a transient receptor potential (TRP) channel, Trpm5 (see below), are coexpressed in a subset of TRCs and define the pathway from receptor to second messenger to Ca2+ elevation and neurotransmitter release.
Ion channels. We had previously identified a cNMP-suppressed cation channel in TRCs that responds to gustducin/PDE regulation of cNMP levels and may serve as the end target in this pathway. We subsequently used a differential screen of α-gustducin-positive versus α-gustducin-negative TRCs to identify Trpm5, a TRP channel that is selectively expressed in a subset of TRCs along with other components of the phosphoinositide pathway. Heterologous expression of Trpm5 shows that it functions as a cationic channel that is gated when calcium is released from internal stores. Trpm5-knockout mice display diminished responses to bitter, sweet, and umami compounds.
Bitter taste receptors. Behavioral and electrophysiological analyses of α-gustducin-knockout mice, as well as numerous in vitro experiments, clearly implicate the gustducin heterotrimer in transduction of responses to bitter compounds. We previously characterized the native taste receptors that activate gustducin and determined that they respond to quinine, strychnine, atropine, and several other bitter compounds. A family of ~25 G protein-coupled receptors (GPCRs), the T2R/TRB receptors, were identified independently by the research groups of Linda Buck (HHMI, Fred Hutchinson Cancer Research Center) and Charles Zuker (HHMI, University of California, San Diego). T2R/TRB receptors have been shown to be expressed selectively in a subset of α-gustducin-positive TRCs and to couple preferentially to gustducin. In collaboration with Steve Roper (University of Miami), we have used confocal microscopy and Ca2+ imaging of TRCs in taste budcontaining slices to monitor TRC responses in situ to various bitter compounds. About half of the bitter-responsive TRCs in a slice were gustducin-positive. When we compared Ca2+ responses of TRCs in slices from wild-type and gustducin-knockout mice, we found a much higher percentage of bitter-responsive TRCs in the wild-type mice. Furthermore, the TRCs in the wild-type mice responded at significantly lower concentrations of bitter compounds than did those from the knockout mice. These results argue for the existence of bitter-response mechanisms independent of T2R/TRB receptors and gustducin, although the most sensitive responses appear to involve these signaling molecules.
Sweet taste receptors.The mouse sac gene is the primary determinant of behavioral and electrophysiological responsiveness of mice to sweeteners. The sac gene encodes T1R3, the third member of the type 1 taste receptor family of GPCRs (mT1R in mouse and hT1R in humans). T1R3 is expressed selectively in TRCs and is most closely related to two other taste receptors, T1R1 and T1R2. The T1R receptors are family 3 GPCRs with ligand-binding sites within large extracellular amino-terminal domains. Heterologous expression of T1R2 plus T1R3 yields a sweet-responsive receptor, while expression of T1R1 plus T1R3 yields an umami-responsive receptor.
T1R3-knockout mice show diminished behavioral preference for and taste nerve responses to sucrose and other sugars and no preference for or nerve responses to artificial sweeteners. We conclude from this that T1R3 is essential for artificial sweetener detection, but that there are T1R3-independent means to detect sugars. Responses of T1R3-knockout mice to glutamate and other umami compounds indicate that there are T1R3-independent means to detect amino acids.
A wide variety of chemically diverse compounds taste sweet, including natural sugars such as glucose and sucrose, sugar alcohols, small-molecule artificial sweeteners such as saccharin, and proteins such as brazzein. Brazzein is a naturally occurring plant protein that only humans, apes, and Old World monkeys perceive as tasting sweet. Differential sensitivity of mice and humans to brazzein's sweetness provided us with a means to identify the portions of the T1R2 plus T1R3 sweet receptor that interact with brazzein.
Using interspecies pairs of human and mouse T1Rs, we determined that hT1R2 plus hT1R3 responds to brazzein, but hT1R2 plus mT1R3 does not, indicating that residues in hT1R3 are required for receptor activity toward brazzein. Using mouse/human chimeric receptors, we determined that human-specific residues within a limited segment of the C-rich region of T1R3 (amino acids 536545) are required for sensitivity to brazzein. Replacement of the C-rich region of mT1R3 with the corresponding human segment enabled hT1R2 plus humanized mT1R3 to respond to brazzein. Pairing humanized mT1R3 with hT1R2 better supports brazzein activity than pairing with mT1R2. Thus, there are likely to be contacts within hT1R2 that interact with brazzein; i.e., brazzein physically interacts with both hT1R2 and hT1R3.
Grants from the National Institutes of Health provided partial support for some of these projects.
Last updated November 05, 2004