Keywords

6.1 Introduction

Animal adaptation requires constant monitoring of the chemical environment. In mammals, chemosensation is based on three main mechanisms, olfaction , taste and chemesthesis. The first two allow the detection and identification of chemical compounds that are airborne or dissolved in the saliva, respectively, whereas the third is based on detection at the level of the skin, mucosa and viscera. The initial step in these processes is the interaction of compounds with specialized receptors expressed in sensory cells. This event is then transduced into electric signals, which are carried by nerve fibers to high centers of the nervous system. In the case of taste, transduction takes place in taste buds located in the tongue and palate. These structures contain four types of taste cells (I–IV) and the endings of gustatory nerves (chorda tympani and glossopharyngeal). Type I cells seem to provide mechanical support to the rest of the structure (Pumplin et al. 1997) , play buffering functions that maintain low levels of extracellular ATP (Bartel et al. 2006; Vandenbeuch et al. 2013) and K+ (Dvoryanchikov et al. 2009) and are thought to contribute to salt taste transduction (Vandenbeuch et al. 2008). Type II cells express G protein-coupled receptors (GPCRs) for molecules that we perceive as sweet , bitter or umami (glutamate) . Expression of these receptors is more or less exclusive, in the sense that GPCRs for bitter compounds (T2Rs) are not co-expressed with receptors for sweet and umami (heterodimers of T1R1, T1R2 and T1R3) (Chaudhari and Roper 2010) . However, it is important to note that the taste quality of a particular taste receptor cell is not determined by the receptors they express, but rather the nerve fiber which they are connected to (Mueller et al. 2005) . Type III cells are stimulated by sour and by salts. Type II and type III cells differ in the way they connect to the taste sensory nerves. Upon stimulation, type II cells release ATP and there is evidence indicating that this may occur through the passage of ATP through pannexin 1 (Huang et al. 2007; Romanov et al. 2007; Huang and Roper 2010) and through CALHM1 channels (Taruno et al. 2013a; Taruno et al. 2013b) , but it remains unclear whether these mechanisms are mutually exclusive. Extracellular ATP activates P2X receptors in taste nerves and P2Y receptors in Type III cells (Kinnamon and Finger 2013) . The latter are endowed with classical synaptic proteins, as well as with voltage-gated Ca2+ channels. Type III cells, also known as ‘presynaptic’, are thought to communicate with the gustatory nerves via vesicular release of GABA and serotonin (Chaudhari and Roper 2010) . Finally, type IV cells, also called basal cells, are considered to be precursors of taste receptor cells. These structural and functional aspects of taste were covered by recent reviews (Feng et al. 2012; Liman et al. 2014; Roper 2014) .

Several TRP channels have been implicated in distinct aspects of taste transduction. For instance, TRPV1 has been proposed in multiple studies as mediator of the amiloride-insensitive component of salt taste transduction (Lyall et al. 2004, 2005, 2009; Treesukosol et al. 2007; Katsumata et al. 2008) . In addition, TRPP3/PKD2L3 were suggested as sour receptors (Ishimaru et al. 2006; LopezJimenez et al. 2006; Ishimaru and Matsunami 2009) . However, negative or inconclusive evidence from knockout mice, and lack of definitive proof for the functional expression of these channels in taste receptor cells, have been raised as doubts for their classification as bona fide elements of taste transduction pathways (Roper 2014) . In contrast, although many issues remain unclear, TRPM5 has drawn consensus about its role in the transduction of several taste qualities. For these reasons, and given the availability of very recent excellent reviews on this subject (Kinnamon 2012; Nilius and Appendino 2013; Liman 2014; Liman et al. 2014; Roper 2014) , this chapter focuses on the most salient basic properties of TRPM5 and how the mechanisms of modulation of this channel influence gustatory transduction at the level of taste receptor cells.

6.2 TRPM5 and its Role in Gustatory Transduction and Modulation

6.2.1 Basic Properties in Heterologous and Native Expression Systems

Of all TRP channels thought to be involved in taste transduction TRPM5 is the best characterized. TRPM5 was first identified in a differential screening of cDNAs from taste receptor cells and was shown to co-express with molecules known to participate in taste signaling, including α-gusducin and phospholipase C-β2 (Perez et al. 2002, 2003) . TRPM5 was originally described as a Ca2+-permeable store-operated channel (Perez et al. 2002, 2003) , but soon afterwards it was established that it is in fact a non-selective Ca2+-impermeable cation channel activated by increase in intracellular Ca2+ concentration (Hofmann et al. 2003; Liu and Liman 2003; Prawitt et al. 2003) .

The effective intracellular Ca2+ concentration (EC50) for TRPM5 activation has been reported for recombinant channels between 0.7 and 32 µM (Hofmann et al. 2003; Liu and Liman 2003; Ullrich et al. 2005) for mouse and 0.84 µM for human (Prawitt et al. 2003). The reasons for these divergences seem to be related to the degree of desensitization of the channel in different recording conditions. Indeed, Ullrich et al. showed that in the whole-cell configuration the EC50 was 0.7 µM, whereas in inside-out patches it was 32 µM (Ullrich et al. 2005). This desensitization was found to be Ca2+-dependent and partially reverted by phosphatidylinositol-4,5-bisphosphate (PIP2) (Liu and Liman 2003). The mechanisms underlying the sensitivity of TRPM5 to intracellular Ca2+ remain poorly understood, although it has been discarded that it is mediated by calmodulin (Hofmann et al. 2003). Intracellular Ca2+ uncaging experiments revealed that TRPM5 is rapidly activated (τ < 2 ms) upon a strong increase of intracellular Ca2+ concentration (Ullrich et al. 2005) .

The characterization of the biophysical properties of TRPM5 has been limited. The single-channel conductance has been estimated between 16 and 25 pS (Hofmann et al. 2003; Liu and Liman 2003; Prawitt et al. 2003; Zhang et al. 2007) , but detailed properties and modulation of single-channel kinetics have not been reported. It is known though that TRPM5 is a voltage-gated channel, activated by membrane depolarization. Like for other voltage-gated TRP channels, the voltage-dependence of TRPM5 activation is quite shallow, with a slope factor between 53 and 58 mV (Talavera et al. 2008c; Gees et al. 2014) . This is suggestive of a low apparent gating valence of ~ 0.6 per channel (Talavera et al. 2005) , similar to what has been directly estimated for TRPM8 (Voets et al. 2007) . In regard to the kinetics of voltage-dependent gating, it has been shown that current relaxation can be described with a single exponential time course. The resulting time constant has a bell-shaped voltage dependence, which is a classical indicative of faster channel closing and opening at very negative and very positive potentials, respectively (Talavera et al. 2008c; Gees et al. 2014). The relationship between the voltage and Ca2+ dependencies of gating is not yet known, and no gating models have been put forward. As detailed below, TRPM5 was shown to be also strongly stimulated by heat (Talavera et al. 2005).

Most of the properties of recombinant TRPM5 channels have been confirmed in the native environment of taste receptor cells. These include lack of selectivity between Na+ and Cs+ as charge carriers, impermeability to NMDG + and Ca2+, stimulation by increase in intracellular Ca2+ concentration and membrane depolarization and high sensitivity to heat (Romanov et al. 2007; Zhang et al. 2007) .

6.2.2 TRPM5 in the Perception of Sweet, Bitter and Umami: Two Knockout Mice Tell Slightly Different Stories

The role of TRPM5 in taste transduction has been directly tested by evaluating the responses of Trpm5 knockout mice. The first report, by Zhang et al., utilized mice with a partial deletion of the Trpm5 gene and retaining the promoter region, which suggest that a large part of the amino terminal of TRPM5 could be still transcribed (Zhang et al. 2003) . Such mice were reported not to be different from wild type mice regarding the responses to salt and acid stimuli, but completely insensitive to sweet , bitter and umami stimuli at the level of behavior and gustatory nerve responses (Zhang et al. 2003). Thus, it was concluded that TRPM5 is absolutely required for the transduction of these taste modalities. However, other studies that used knockout mice fully lacking the TRPM5 protein consistently yielded different results. In several separate experimental series it was found that gustatory nerve responses to sweet , bitter and umami stimuli were significantly reduced, but not completely abolished in knockout mice (Talavera et al. 2005, 2008c; Damak et al. 2006) . Damak et al. (2006) found that Trpm5 knockout mice were sensitive to sweet, bitter and umami tastes in a long access two-bottle preference test but not in a brief-access licking test. It was argued that this difference could result from a lower sensitivity of the brief-access test and/or from the influence of post-ingestive taste-independent effects (Damak et al. 2006). Interestingly, it was later found that the sweet-blind Trpm5 knockout mice from Zhang et al. displayed preference for sucrose in a long-term test due to post-ingestive reward (de Araujo et al. 2008) . This preference was associated with calorie intake because the non-energetic sweetener sucralose failed to trigger preference over water in the same test.

The reasons underlying the differences in the results from studies with the two Trpm5 knockout strains remain unknown. It has been suggested that the lack of response of knockout mice in the study of Zhang et al. could be due to a putative dominant-negative action of the TRPM5 N-terminal protein fragment that is presumably produced in this mice on alternative transduction pathways (Damak et al. 2006) . Additionally, differences in genetic background of these mice were also put forward (Damak et al. 2006). However, none of these hypotheses has been tested. Nevertheless, the observations of Damak et al. (2006) and several other studies (Devantier et al. 2008; Eddy et al. 2012 and see below) tilt the balance in favor of a model of transduction of sweet, bitter and umami featuring TRPM5-dependent as well as TRPM5-independent mechanisms. Not surprisingly, the relative importance of these mechanisms depends on the tastant, experimental paradigm and gustatory field being evaluated (Damak et al. 2006) .

These and other studies served to establish a qualitative model for the role of TRPM5 in transduction of sweet , bitter and umami stimuli (Fig. 6.1). Activation of G-protein-coupled taste receptors by these tastants stimulates PLCβ2 with the consequent production of IP3 and Ca2+ release from the endoplasmic reticulum through type 3 IP3 receptors (IP3R3) (Hisatsune et al. 2007) . The increase in intracellular Ca2+ concentration activates TRPM5, leading to cation influx and depolarization of the taste receptor cell (Hofmann et al. 2003; Liu and Liman 2003; Prawitt et al. 2003; Zhang et al. 2003) . This depolarization activates voltage-gated Na+ channels leading to action potential firing (Gao et al. 2009; Vandenbeuch and Kinnamon 2009) . It seems that the combination of increase in intracellular Ca2+ concentration and depolarization triggers the release of ATP through pannexin 1 (Huang et al. 2007; Huang and Roper 2010) and/or CALHM1 channels (Taruno et al. 2013a, b) .

Fig. 6.1
figure 1

Cartoon of the TRPM5-dependent signaling pathway for the detection of sweet, bitter and umami stimuli. Type II taste receptor cells express TRPM5 in the basolateral membrane. TRPM5 channels are activated by increase in intracellular Ca2+ that results from the activation of G-protein-coupled taste receptors sensitive to bitter, sweet and umami compounds. TRPM5 is inhibited by extracellular acidosis, quinine, quinidine and nicotine and stimulated by membrane depolarization, heat and arachidonic acid (AA)

Full size image

6.2.3 Role of TRPM5 in the Perception of Salts and Fats

Intriguingly, the Trpm5 knockout mice from Damak et al. (2006) displayed significantly lower responses to NaCl than wild type mice. This again contrasts with the results of Zhang et al. who reported that their Trpm5 knockout and wild type mice have the same sensitivity to NaCl, but are in agreement with reduced responses to NaCl in α-gustducin (Glendinning et al. 2005) and of Plcβ2 (Dotson et al. 2005) knockout mice. Notably, if we consider TRPM5 to be exclusively expressed in cells transducing sweet, bitter or umami stimuli we should conclude that the TRMP5-dependent responses to NaCl do not elicit salty taste. Instead, they should elicit sweet, bitter and/or umami qualities. Alternatively, TRPM5 may participate in the transduction of saltiness of NaCl in a small proportion of taste receptor cells reported to express this channel but not α-gustducin (Damak et al. 2006) or T1R or T2R (Zhang et al. 2003) . This latter possibility is supported by the findings that Trpm5 knockout mice, but not Plcβ2 knockout mice, display smaller chorda tympani responses to low concentrations (100 mM) of NaCl in the presence of benzamil, a potent blocker of the epithelial Na+ channel ENaC (Ren et al. 2013) . On the other hand, the responses of the chorda tympani to salts at high concentrations (500 mM KCl or 500 mM NaCl plus amiloride) were shown to be about 50 % weaker in Trpm5 knockout mice and Plcβ2 knockout mice than in wild type animals (Oka et al. 2013) . Interestingly, wild type mice clearly preferred solutions containing low concentrations of divalent salts such as FeSO4 and ZnSO4 over water, but this was not observed in Trpm5 knockout mice nor in T1r3 knockout mice (Riera et al. 2009) . Furthermore, recent evidence indicates that T1R3 is involved in the detection of calcium and magnesium (Tordoff et al. 2008, 2012) .

Finally, there is one study showing that linoleic acid elicits non-selective monovalent cation currents in taste receptor cells that are reduced by application of the TRPM5 inhibitor triphenylphosphine oxide (TPPO) and by genetic ablation of Trpm5 (Liu et al. 2011) . However, it remains unclear why these maneuvers also reduced the increase in intracellular Ca2+ concentration triggered by linoleic acid. This is not expected because TRPM5 activation is supposed to be downstream the Ca2+ signal. Of note, Trpm5 knockout mice were found to be deficient in the detection of linoleic acid in tests for long-term preference and brief-access licking recordings after conditioned taste aversion. This body of evidence indicates that TRPM5 is involved in the detection of salts and fats, but the precise underlying pathways remain to be fully characterized.

6.2.4 Modulators of TRPM5 and their Influence on Taste Transduction

The key role of TRPM5 in transduction mechanisms in the taste receptor cells implies that factors that modulate this channel are expected to affect taste. It is notable, however, that the influences of the Ca2+-dependent desensitization and the modulation by PIP2 and voltage on taste transduction remain completely unknown .

As other voltage-gated TRP channels (Nilius et al. 2005; Talavera et al. 2008a; b) , TRPM5 displays a strong thermal sensitivity, being stimulated by heating. The underlying mechanism entails a shift of the voltage-dependence of channel activation to negative potentials (Talavera et al. 2005) , very much like that described for activation of TRPV1 (Voets et al. 2004) . Of note, channel stimulation by either depolarization or heating can not be observed at low intracellular Ca2+ concentration (Talavera et al. 2005). Notably, it was shown that heating strongly enhances chorda tympani responses to sweet compounds in wild type but not in Trpm5 knockout mice. The strong sensitivity of TRPM5 to heat was therefore proposed to underlie the enhanced sweetness perception at high temperatures. It is also tempting to speculate that heat-induced stimulation of TRPM5 mediates “thermal taste”, which consists in the trigger of taste sensations in the absence of tastants by warming the tongue from 20 to 30 °C (Cruz and Green 2000) . This effect would require the stimulation of TRPM5 currents by heating at basal (low) intracellular Ca2+ concentrations. Alternatively, heating could induce on its own slight increase in the intracellular Ca2+ concentration and act as initial trigger for Ca2+-induced activation of TRPM5. However, none of these possibilities have been addressed experimentally. It has been argued that, besides TRPM5, other elements of the taste transduction pathways could be sensitive to heat and contribute to enhanced taste response (Liman 2006, 2014) . However, for any element to be significantly important, it would have to feature a thermal sensitivity similar to that of TRPM5 (Q10 ~ 10 between 25 and 35 °C). For any TRP channel lover who recognizes that the “TRP-centric” view of sensory biology has done more harm than good (Meseguer et al. 2011) , it would be extremely interesting to identify such heat-sensitive element outside the TRP protein family. At the moment, TRPM5 remains the most plausible candidate (Talavera et al. 2007) .

It has been also pointed out that, considering the role of TRPM5 in the transduction of bitter and umami, heat-induced activation of TRPM5 should result in enhanced responses to these stimuli (Liman et al. 2014; Roper 2014) . However, this was not observed experimentally (Talavera et al. 2005). To explain this apparent contradiction, it is important to notice that only the responses that strongly rely on TRPM5 activation are expected to be significantly enhanced by heating. In the original report, which used the Trpm5 knockout strain generated by Damak et al. (2006) , only the chorda tympani responses to sweeteners were strongly reduced in the mice lacking TRPM5 (Talavera et al. 2005). As explained below, a similar discussion rose around the inhibitory effect of quinine on TRPM5.

Another potent modulator of TRPM5 is extracellular acidosis. Liu et al. showed that TRPM5 currents were sharply inhibited by drops in extracellular pH between 7 and 5.5 (Liu et al. 2005) . This effect was voltage-independent, which is in accordance with the location of residues important for the inhibitory effect at the extracellular face of the channel. Notably, acidosis enhanced the desensitization of TRPM5 currents. Given the restricted localization of TRPM5 at the basolateral membrane of taste receptor cells, the relevance of its inhibition by extracellular protons has been taken with caution (Roper 2014) . Also with reserves one should take the reports on stimulation of TRPM5 by arachidonic acid (Oike et al. 2006) and on the inhibition of this channel by extracellular Zn2+ (Uchida and Tominaga 2013) . It is currently not clear whether and how these effects are relevant for taste transduction.

The rather extensive literature on interactions between bitter and sweet tastants motivated the study of the possible effects of membrane-permeable bitter compounds on TRPM5 (Talavera et al. 2008c; Gees et al. 2014) . It was found that quinine, the bitter additive in tonic waters, and its stereoisomer quinidine inhibit TRPM5 at micromolar concentrations. Furthermore, both compounds inhibited the responses of the chorda tympani to sweet tastants in wild type mice, but did not alter the residual responses of sweet-sensitive gustatory fibers of Trpm5 knockout mice (Talavera et al. 2008c) (Fig. 6.2). These findings can readily explain the long-known inhibitory effect of quinine on sweet taste perception (Öhrwall 1891) and its partial localization at the gustatory periphery (Formaker and Frank 1996) . This phenomenon was considered as a bitter-sweet taste interaction, but it is now clear that the quinine-induced inhibition of the sweet pathway is at least in part dissociated from the detection and processing of the bitter signal because the bitter tastant inhibits directly the sweet transduction pathway.

Fig. 6.2
figure 2

Quinine inhibits the responses of the chorda tympani to sucrose in wild type but not in Trpm5 knockout mice. a Effects of quinine (10 mM) on the responses of single sweet-sensitive fibers from wild type and Trpm5 knockout mice to sucrose (0.5 M). b, c Concentration-dependent effects of quinine on the number of nerve impulses elicited by sucrose in single-sweet sensitive fibers from wild type and Trpm5 knockout mice. d Correlation between the inhibitory effects of quinine of the responses of mouse chorda tympani to different tastants with the degree to which these responses depend on TRPM5. (Figure adapted from (Talavera et al. 2008c), with permission)

Full size image

TRPM5 currents were later found to be inhibited also by nicotine, a compound that has been usually linked to gustatory transduction (Grunberg 1982, 1985; Simons et al. 2006; Lyall et al. 2007; Tomassini et al. 2007) . The exact way in which nicotine acts on the perception of these tastes is still not well understood, but very likely multiple different mechanisms are involved. These include nicotine-induced reduction of the size of fungiform papillae and the number of taste cells (Tomassini et al. 2007), inhibition of chorda tympani responses to KCl and NaCl (Lyall et al. 2007) , suppression of gustatory responses of neurons in the nucleus of the solitary tract (Simons et al. 2006), inhibition of voltage-dependent sodium channels, sensitization of TRPV1 (Liu et al. 2004) and activation of TRPA1 (Talavera et al. 2009) . The inhibition of TRPM5 by nicotine represents yet another peripheral mechanism of inhibition of gustatory perception. Given the relatively high nicotine concentrations required for TRPM5 inhibition (IC50 of 1.3 mM at− 50 mV), it is virtually impossible that this effect could be relevant during normal smoking. On the other hand, it could be relevant during tobacco chewing and the use of snuss, in nicotine-replacement therapies, and certainly for the interpretation of experimental studies that used nicotine concentrations of up to 600 mM (Carstens et al. 1998; Boucher et al. 2003; Simons et al. 2006) .

Besides shedding light on the mechanisms underlying bitter-sweet taste interactions, the effects of quinine and nicotine on TRPM5 have important implications for our understanding of the mechanisms of bitter taste transduction . The inhibitory effects of these compounds on TRPM5 currents indicate that the perception of their bitter taste cannot be mediated solely by a process requiring TRPM5 activation. Alternative pathways should be recruited to elicit the aversive behavior towards quinine and nicotine at high concentrations. Accordingly, it has been recently shown that nicotine also activates a nAChR-dependent and TRPM5-independent gustatory pathway (Oliveira-Maia et al. 2009) . Notably, it is expected that quinine, quinidine and nicotine may also interfere with the perception of umami . However, as for the case of the stimulatory effect of heat on TRPM5 (see above), the effect of TRMP5 inhibitors on the transduction of a certain compound must be related to the extent to which this process depends on TRPM5 activation. This contention was clearly demonstrated for quinine (Talavera et al. 2008c) (Fig. 6.2).

In contrast to quinine, quinidine and nicotine, other bitter compounds known to have an influence in sweet taste perception, including saccharine, caffeine and the related xanthines theophylline and theobromine had no effect on TRPM5 currents, indicating multiple mechanisms of bitter-sweet taste interactions (Gees et al. 2014) .

6.3 Concluding Remarks

In recent years we have witnessed evident advances in the understanding of the signaling processes underlying taste transduction , but many issues remain unsolved. For instance, it is particularly challenging that despite the large amount of evidence implicating TRPV1 and TRPP channels, no definitive conclusions have been reached. Regarding TRPM5 we are in a more luxurious position, as some tools, including knockout mice, channel modulators and the inhibitor TPPO, are available to study the roles of this channel. We are particularly expectant about the publication of a full report on the discovery and characterization of a group of selective agonists that may help developing new taste modulators and treatments for type II diabetes (Philippaert 2013) . However, it is clear that we still know very little about the details of how this channel is gated in basal conditions and how it is modulated. Furthermore, most findings have been obtained in heterologous expression systems and need to be confirmed in taste cells. It is also clear that studying the pharmacological properties of individual compounds in detail is of outmost importance for the understanding of their specific chemosensory properties.