For decades taste quality coding has been a fiercely debated field (Smith et al. 2000, Herness 2000). In principle, two competing models exist (Hellekant et al. 1998, Smith and St John 1999). The labelled line model favours a separate coding of the five basic taste qualities (Hellekant et al. 1998). Therefore, this model suggests the existence of specialised taste receptor cells for each taste quality, which are innervated by dedicated fibres (Hellekant et al. 1998). Thus, for example, a sweet stimulus such as sucrose will activate sweet taste receptor cells. Subsequently, solely sweet taste receptor cell innervating fibres will convey the signal to the brain. Consequently, in this model, the information of the taste quality is encoded at the level of the taste receptor cells.
In the competing across fibre pattern model taste receptor cells and the innervating neurons are not strictly specialised and respond to all taste stimuli (Smith et al. 2000). Only the strength of responses differ amongst taste modalities (Smith et al. 2000). In this model, for example, sweet best taste receptor cells exist, that respond stronger to sweet stimuli than to bitter, umami, sour or salt stimuli. Similarly, bitter best, salt best, sour best and amino acid best taste receptor cells would exist. As the neurons always innervate several taste receptor cells, they might, by chance, innervate more sweet best taste receptor cells. Consequently, such a neuron would be a sweet best neuron and respond stronger to sweet stimuli than to the other taste stimuli. Due to the same mechanism, neurons with a higher response for the other taste qualities would exist. Dependent on the different activation pattern, stimulation of the oral cavity by sweet, bitter, sour, salty and umami compounds would generate excitation patterns across many neurons that are decoded by the brain (Smith et al. 2000).
Consequently, in this model the information about the taste qualities is also encoded in higher brain centres. As both models are supported by experimental data it is not easy to decide which model is correct. Evidence that supports the across-fibre pattern model stems from various electrophysiological recordings. Nerve recordings mainly obtained from rodents and amphibians show that fibres frequently respond to stimuli of more than one taste quality, although the nerve responses to stimuli of one taste quality are especially strong. (Dahl et al. 1997, Yamamoto and Yuyama 1987, Pritchard and Scott 1982, Smith et al. 2000, Smith and St John 1999, Woolston and Erickson 1979). Moreover, recordings of taste receptor cells equally clearly demonstrate, that isolated taste receptor cells respond to stimuli of multiple taste qualities (Sato and Beidler 1997, Bealer and Smith 1975, Tonosaki and Funakoshi 1984, Herness and Gilbertson 1999, Herness 2000). In addition, non-invasive in situ calcium imaging of rat taste buds essentially confirmed a broad tuning of taste receptor cells (Caicedo et al. 2002, Caicedo and Roper 2001). In marked contrast, other data strongly argue for a labelled line model. Nerve recordings in primates revealed relatively narrowly tuned fibres for sweet, bitter and salt transduction (Danilova et al. 1998, 2002, Hellekant and Ninomiya 1994, Hellekant et al. 1998, Sato et al. 1994). The Taslr receptors are just expressed in a subset of -30% of the taste receptor cells (Nelson et al. 2001). Moreover, Taslrl and Taslr2 do not colocalise, although they are almost always coexpressed with Taslr3 (Nelson et al. 2001, Adler et al. 2000, Hoon et al. 1999). These findings strongly argue that different subsets of taste receptor cells mediate sweet and umami taste. The observation that the Tas2r receptors, although expressed in -20% of the taste receptor cells, do not colocalise with the Taslr receptors (Adler et al. 2000, Nelson et al. 2001) suggests, that bitter taste is mediated by a third subpopulation of cells. In addition the salt taste receptor ENaC and the putative sour receptors HCN1 + 4 are also expressed in subsets of taste receptor cells (Kretz et al. 1999, Stevens et al. 2001). These findings are further supported by studies of taste transduction. Various signal transduction components such as gustducin, PLC/32, and TRPM5 are expressed in subsets of taste receptor cells (McLaughlin et al. 1992, Zhang et al. 2003). Apart of a specific expression also behavioural and physiological studies of various transgenic animals also argue for specificity in taste coding. Nerve recordings and behavioural studies of PLC/32 and TRPM5 knock out animals showed that sweet, bitter, and amino acid taste but not salt and sour taste were lost (Zhang et al. 2003). This demonstrates that sweet, bitter, and umami taste share some parts of their signal transduction pathways, whereas salt and sour taste use different molecules. The specificity of the signal transduction pathway is most evidently shown by a specific rescue of bitter taste. PLC/32 knock out animals that express PLC/32 under control of a Tas2r promoter selectively regain their capability to detect bitter compounds, whereas the ability to perceive sweet and umami substances is still impaired (Zhang et al. 2003). This strongly argues that bitter taste is mediated by Tas2r receptor expressing subpopulation of taste receptor cells. Analysis of Taslrl and Taslr2 knock out mice showed that umami taste depends on the Taslrl gene whereas sweet taste depends on the Taslr2 gene (Zhao et al. 2003). The Taslr3 knock out mice show a reduced sweet and umami taste (Zhao et al. 2003, Max et al. 2001). These findings are consistent with the specific expression pattern of the various taste receptor molecules in taste receptor cells. The expression of a receptor for the synthetic opiate spiralidone under control of the Taslr2 promoter resulted in a high preference of the transgenic animals for this opiate while wild type mice are indifferent to it (Zhao et al. 2003). This result strongly argues that sweet taste sensation is encoded by a specific subpopulation of taste receptor cells and not by the sweet receptor protein itself. Activation of any one receptor in this cell population would be perceived as sweet. Thus, a considerable amount of independent observations strongly point towards a labelled line coding of taste information in the periphery by different subtypes of taste receptor cells. This leads, of course, to the question how these results can be reconciled with the apparent broad tuning of taste receptor cells (Sato and Beidler 1997, Bealer and Smith 1975, Tonosaki and Funakoshi 1984, Herness and Gilbertson 1999, Herness 2000, Caicedo et al. 2002). One explanation might be based on the observation that a compound rarely has a pure taste (Skramlik 1926, Chon 1914). In many cases compounds elicit multiple taste qualities (Skramlik, 1926, Chon, 1914). For example the frequently used artificial sweetener saccharin, and other artificial sweeteners have, beside their sweet taste, a bitter aftertaste (Schiffman et al. 1995). Similarly, the prototype salt taste stimulus NaCl tastes sweet at low concentrations and salty at higher concentrations (Skramlik 1926). Moreover, many bitter compounds, such as the frequently used quinine, are pharmacologically active (Skramlik 1926) and might therefore influence various targets including ion channels and G-proteins in taste receptor cells. It has yet not been determined to what extent such 'unspecific' activation events occur and if such events lead to a release of neurotransmitters. Thus a certain degree of overlapping activation of taste receptor cells may be expected and does not necessarily argue against a labelled line coding. Moreover, there are clear indications that the taste receptors of different species have altered functional properties (see sections 1.3.1-1.3.5). The most drastic example might be the strong divergence of many TAS2R receptors in humans and rodents (Shi et al. 2003). This suggests strong species-specific variations in the bitter perception of humans and rodents. Such differences are also evident for salt and sweet taste (see sections 1.3.1 and 1.4.2). Thus, for some compounds, the perception in humans and rodents will drastically differ. Consequently, to some extent species differences might account for the apparent broad tuning. In addition, recent reports suggest that not all taste receptor cells might be directly connected with nerve fibres (Royer and Kinnamon 1994, Kinnamon et al. 1993). If this is true they must be, in an unknown fashion, connected to the so called type III cells in taste buds, which do contain synapses (Royer and Kinnamon 1994, Kinnamon et al. 1993). How these cells talk to each other and the implication for taste quality coding remains unknown. In summary, although much evidence argues for the labelled line model, both coding models still compete for validity. The final proof of the models depends on knowledge about the precise innervations of the cell types within taste buds and how these cells communicate. Thus transsynaptic tracing studies are necessary to elucidate how excitation of TRCs is conveyed to the brain.
Was this article helpful?