The functional organization of taste cells in the taste buds and of the gustatory neural circuits are still subject to debate. WGA labeling of the neurons that synapse upon discrete subsets of taste receptor cells will help determine how the gustatory signal is transmitted within and beyond the taste buds.
We set out to determine if transgenic expression of WGA in a subset of mouse taste receptor cells could be used to this effect. Our data show that WGA originally expressed in T1r3-expressing type II taste receptor cells is transported to the geniculate and petrosal ganglia, and not to neighboring taste cells. WGA is believed to be taken up by and transported across synapses in vesicles. Together, these data suggest that the T1r3-expressing cells do make contact with chorda tympani and glossopharyngeal neurons. T1r3-expressing taste cells are type II cells which do not express the presynaptic molecules SNAP-25 and voltage-gated calcium channels required for synaptic transmission [16
]. These cells activate the taste nerves by releasing ATP through pannexin hemichannels [19
]. The recently reported observation of subsurface cisternae of smooth endoplasmic reticulum at regions where nerves and type II taste cells are in close apposition [14
] suggests that this is the site of activation of the nerve by ATP released from the type II taste cell and may be the site where WGA crosses from taste cell to nerve.
As expected WGA immunostaining was observed in the geniculate and petrosal ganglia which contain the cell bodies of the nerve fibers of the chorda tympani and the glossopharyngeal nerves, respectively. Most neurons in the geniculate ganglion and many neurons in the petrosal ganglion were immunoreactive to the WGA antibody, although the intensity of the label varied from cell to cell. This pattern of staining is consistent with electrophysiological recordings from chorda tympani and glossopharyngeal single nerve fibers and geniculate ganglion cell bodies, which showed that there are broadly tuned and narrowly tuned fibers for each taste modality, and the fibers and cell bodies narrowly tuned to one modality also respond weakly to the other modalities [5
]. We speculate that the strongly stained cell bodies may correspond to sucrose or MSG best neurons, the weakly stained ones may correspond to neurons narrowly tuned to other modalities and the cell bodies with intermediate staining may correspond to broadly tuned neurons. We cannot however exclude the possibility that WGA is released in the pericellular space in the taste buds and is nonspecifically taken up by nerves other that the ones in close contact with T1r3-expressing taste cells.
Surprisingly, WGA was also found in the trigeminal ganglion, which contains the cell bodies of neurons that relay somatosensory information from the head and neck to the brain. The most likely explanation is that WGA has originated from the recently discovered solitary chemoreceptor cells of the nasal epithelium [29
], which also express T1r3 [26
]. However it cannot be excluded that WGA has migrated from taste receptor cells to the trigeminal ganglion. It was initially believed, based on studies with fluorescent nerve tracers that trigeminal fibers surround the taste buds in fungiform papillae but do not enter them [30
]. However, intragemmal nerve fibers immunoreactive for the vanilloid receptor TrpV1 were described, suggesting that trigeminal nerve fibers enter the taste buds [31
]. Furthermore, after ontophoretic injection of a fluorescent dye in fungiform papillae, fluorescence was detected in some cells of the trigeminal ganglion [32
]. This was attributed to direct injection of the perigemmal trigeminal fibers, but in the light of Ishida et al
and our results it could well be transsynaptic transport from taste receptor cells. It is also possible that WGA was transported to the trigeminal ganglion from the brain, but we found no evidence of GFP expression in the medulla. Finally it is possible that WGA originated from other as yet undiscovered T1r3-expressing cells innervated by the trigeminal nerve.
We found a complex pattern of WGA localization in the medulla. In addition to the nucleus of the solitary tract, WGA immunoreactivity was found in the nucleus ambiguus, the vestibular nucleus, the trigeminal nucleus and in the gigantocellular reticular nucleus. This complex pattern is not surprising given the contribution of the taste sensation to several functions, e.g. swallowing, oromotor reflexes and conditioned taste aversion. The synapses crossed by WGA in the brain may be stimulatory or inhibitory and may not be sufficient alone to activate the neurons in which they occur. This may explain the difference between the pattern we found here and the more restricted pattern of fos-like immunoreactivity in response to sucrose [33
] which results only from activated neurons. Others have found a more extensive labeling of the medulla by using a transsynaptic tracer. Bradley et al. [34
] found that WGA conjugated to horse radish peroxidase injected in the circumvallate papilla of rats was transported to the solitary nucleus, the trigeminal system, the nucleus ambiguus, and the inferior salivatory nucleus. A similar pattern was reported by Hamilton and Norgren [35
] after application of horse radish peroxidase to the cut lingual-tonsillar branch of the glosso-pharyngeal nerve. WGA transported through the trigeminal system also very likely contributes to the staining pattern we found in the medulla.
One limitation of this tracing approach is that the transport of WGA across synapses is both anterograde and retrograde, making it difficult to determine the order of the neurons in the neural circuits. The intensity of the WGA immunostain was strongest in the taste receptor cells and weakest in the medulla, and no GFP expression was detected in the ganglia or in the medulla. Therefore it is very likely that WGA migrated from taste receptor cells to the ganglia then to the medulla.
In this study, WGA is expressed from the promoter of a trangene, not knocked-in. Although unlikely, we cannot completely rule out ectopic expression of WGA in a small number of innervated cells outside the taste bud and migration of WGA to some nuclei in the medulla from those cells.
Two other publications have described WGA labeling of neural circuits originating from type II taste cells. The earlier publication by Sugita and Shiba [24
] described neural circuits originating from T1r3 or T2r5 expressing cells and reported WGA staining in the third order taste neurons and in the gustatory cortex of transgenic mice expressing WGA-DsRed under the control of the T1r3 or T2r5 promoter. The main conclusion of Sugita and Shiba was that neural circuits originating from T1r3 or T2r5 expressing cells are mostly separated. In a very recent publication Ohmoto et al [26
] described the neural circuits labelled by WGA expressed from the T1r3 promoter in transgenic mice. There are three major differences between the results of these two studies. First Sugita and Shiba detected WGA staining in the third order taste neurons and beyond, whereas Ohmoto et al did not. Second, Ohmoto et al detected WGA in the trigeminal ganglion whereas Sugita and Shiba did not. Third Sugita and Shiba reported that WGA was restricted to the taste areas in the central nervous system, whereas Ohmoto reported a wider pattern of WGA staining in the medulla. Our results are in full agreement with those of Ohmoto et al. and contrary to the main conclusions of Sugita and Shiba. Based on our results and those of Ohmoto et al., and contrary to Sugita and Shiba, the anatomical relationship between neural circuits originating from bitter responsive cells or sweet/umami responsive cells remains undetermined.