This web page was produced as an assignment for an undergraduate course at Davidson College



A Review by Jon Palma of:

A Novel Family of Mammalian Taste Receptors

Adler E et al.  2000.  Cell. 100:693-702.

Back to my Molecular Biology Homepage


            In ‘A Novel Family of Mammalian Taste Receptors’, Adler et al. contend that they have identified a group of 40-80 human and rodent taste receptor candidates, which they designate T2Rs.  According to the authors, these candidates are expressed in particular sets of taste receptor cells of the tongue and palate epithelia, and they are only expressed in cells that contain the G protein α subunit gustducin.  Therefore, they postulate that T2Rs are gustducin-linked.  From the data gathered, the researchers hypothesize that “each [taste] cell may be capable of recognizing multiple tastants” (693).


            The five main categories of taste (sweet, bitter, sour, salty, and umami [the taste of monosodium glutamate]) are sensed by taste receptor cells in taste buds, which in turn are located on the surface of the tongue and palate.  The differences in perception of these basic tastes lie in different signal transduction pathways for each, which are thought to be separated into independent subsets of receptor cells.  That is, when a particle that elicits a taste perception, the physiological process that occurs is the activation of a receptor to trigger a particular signal cascade that eventually leads to neurotransmitter release.  The information from this release is processed by the cortical taste centers in the brain, thus yielding the perception of a taste.  Isolation of the genes that encode taste receptors could help elucidate this process further by enabling marking of the cells, definition of pathways, generation topographic maps, and tracing of neuronal circuits (693-4).


            The researchers had previously isolated the taste receptors T1R1 and T1R2, and using in situ hybridization showed they are expressed in about the same percentage of taste bud cells as gustducin, but not in the same cells as gustducin.  This led them to believe that other, gustducin-linked taste receptors might exist (in part because prior research has suggested that bitter taste perception is mediated by gustducin).  Therefore, Adler et al. began by searching “for G protein-coupled receptors (GPCRs) in genomic intervals linked to bitter taste perception” (695).  From previous experiments in humans, it was known that a locus existed that was integral in recognition of the bitter taste of 6-n-propyl-2-thiouracil (PROP).  The researchers postulated that this locus might encode a bitter taste receptor, and therefore searched the locus for open reading frames that might encode transmembrane proteins (receptors).  They identified a candidate for a bitter taste receptor, which they named T2R-1.  Sequence analysis of this candidate showed seven possible membrane spanning domains, as well as many conserved residues common to GPCRs.  Computer searches were then used to identify related sequences (i.e. related genes, related candidate receptors).  Numerous human receptors were identified that presumably belong to the novel group of T2Rs.  The estimate that the family is comprised of 40-80 T2Rs is an extrapolation from the number of potential T2R sequences found in various databases which represent different amounts of the human genome.  For example, one database representing approximately 50% of the human genome contained 36 functional T2Rs, suggesting that approximately 72 T2Rs may exist.


            Figure 1 is a comparison of the predicted amino acid sequences of several human and rodent T2R genes.  Seven predicted transmembrane domains are indicated.  Furthermore, identical and conservative amino acids are shaded to illustrate sequence similarities between various human and rodent T2Rs.  According to the researchers, these sequences range in similarity from 30% to 70%.  The most highly conserved regions appear to be the first three predicted transmembrane regions (TM1, TM2, TM3), the last predicted membrane spanning region (TM7), and the second cytoplasmic portion (between TM3 and TM4), all of which the authors note.  Because most of the variability in the genes exists in extracellular segments (between TM2 and TM3, TM4 and TM5, TM6 and TM7), it is thought that variability is related to the ability of T2Rs to be sensitive to different tastants, which are structurally different.


            In Figure 2, a cladogram is presented to show the relatedness of the T2R family, and the general relatedness of T2Rs to receptors in the vomeronasal organ (a chemosensory organ), and opsin (a protein involved in vision).  The roots of the cladogram are colored to represent the chromosomal location of the genes, and it is apparent that many of the genes are related in chromosomal location.  In general, the most closely related in terms of sequence are also related in locus.


            Further evidence for the researchers’ claims lies in the arrangement of T2Rs on chromosomes.  Figure 3 consists of representations of human chromosomes 5, 12, and 7, and mouse chromosomes 15 and 6.  Colors are used to indicate homologous regions, and it is evident that similarity exists between human chromosome 5 and mouse chromosome 15.  The portion of human chromosome 5 that encodes a T2R (hT2R-1) is associated with the location of the PROP gene (discussed earlier).  The related mouse T2R is mT2R-19.   Not only is this receptor on a homologous interval on mouse chromosome 15, it is also closely related according to the cladogram (figure 2), which is expected.  In fact, the researchers found that “each genomic interval containing mouse T2Rs is homologous to [a human locus] containing its closest human counterpart” (697).

            Mouse chromosome 6 contains two regions that encode T2Rs; these two regions exist on two separate chromosomes (12 and 7) in humans.  Furthermore, the distal end of mouse chromosome 6 is well characterized in that it contains a number of bitter taste sensing genes.  A closer inspection of three bacterial artificial chromosome (BAC) contigs indicates that several T2Rs are contained at this locus.  The researchers obtained the contigs by probing mouse genomic libraries with human T2Rs. 


At this point in the paper, Adler et al. assert, “T2Rs are intimately linked to loci implicated in bitter perception” which “substantiates the postulate that T2Rs may function as taste receptors” (697).  More specifically, they contend that T2Rs might function as bitter taste receptors.  In essence, the authors have demonstrated that a striking correlation exists between the loci of the genes they have isolated and the proposed functions of those loci.  They are wise to not overstate their findings by suggesting that they have without doubt isolated a family of taste receptor genes.  In an effort to take a step closer to making this claim, the researchers investigate the expression of T2R genes.


Logically, Adler et al. expect T2Rs to be expressed in taste receptor cells.  Figure 4 simply defines the regions of a rodent mouth that contain taste buds:  the fungiform papillae, the tip of the tongue; the foliate papillae, the posterior lateral edge of the tongue; circumvallate papillae, the very back of the tongue; and the taste sensitive areas of the palate epithelium, the geschmackstreifen and epiglottis.


Using in situ hybridization with T2R digoxigenin-labeled antisense RNA probes (figure 5), it was determined that various T2Rs were transcribed in approximately 15% of the cells in taste buds, except in the fungiform papillae, less than 10% of which contained T2Rs.  The researchers note that this was an unexpected finding, but it is my understanding that the tip of the tongue is a region that is more sensitive to perception of sweetness, in which case the result would correlate well with the notion that T2Rs encode receptors that perceive bitter stimuli.


Because T2Rs are so numerous, Adler et al. hypothesize that more than one receptor is present is each taste cell.  To test this proposal, they compared the number of taste cells labeled by in situ hybridization when different mixtures of 2, 5, or 10 T2R probes were used (figure 6, a-c).  If only one receptor were present in each taste cell, increasing the number of probes should correspondingly increase the total number of labeled cells.  However, the mixed probes only led to a slight increase in labeling, from approximately 15% of taste cells (for an individual receptor) to 20% of taste cells.  One possible explanation for this result is that despite the fact that mixed probes were being added, only a fraction of them bound (for example, if only 3 of the 10 had labeled in panel c), causing only a few more cells to be labeled.  To help convince the skeptical reader, the authors directly demonstrated that coexpression occurs by using a two-color double-label fluorescent in situ hybridization (FISH) technique (figure 6, d).  T2R-3 was labeled green and T2R-7 was labeled red in the same taste receptor cells, indicating that both receptors are expressed at the same time in the same cells.  Adler et al. also used other combinations of T2Rs in the double-label experiment, but the data are not shown.  This is good evidence that multiple T2Rs are expressed per taste cell.


The researchers summarize their findings up this point with three statements.  1) There are significant differences in expression of T2Rs in various taste buds (i.e. the lower number of cells labeled in the fungiform papillae compared to other taste cells).  2) The collection of T2Rs is numerous and complex.  3) Taste cells express multiple receptors.  The data the paper has presented support these claims, but the more salient claim of the paper, that a novel family of taste receptors has been discovered, has not been sufficiently demonstrated.


The next step taken by Adler and his colleagues was to determine whether T2Rs are coexpressed with gustducin.  They expect coexpression to occur; one of their original hypotheses was that because T1Rs are not coexpressed with gustducin, some class of taste receptors might exist that does coexpress with it.  Figure 7 utilizes the double-label FISH method, probing for T2Rs and gustducin in the same cells.  The results show that every taste cell containing a T2R also contains gustducin, though some gustducin-containing cells are devoid of T2Rs (even when as many as 10 T2R probes were used).  These data are evidence for the coexpression of T2Rs and gustducin.  Panel d is a FISH double labeling of T1Rs and T2Rs, demonstrating that the two families of receptors are distinct because there is no overlap in their expression.

The authors put forth the idea that gustducin-positive cells without T2Rs may contain yet another family of taste receptors.  Evidence for this is given by the fungiform papillae, where the vast majority of gustducin-containing cells do not express T2Rs.  The researchers employed two techniques to ensure that there was in fact no expression in most cells of the fungiform papillae, not merely a low level of expression in all the cells of that region.  First, they used mixed probes and extended developing times, but did not observe any additional labeling.  They also used PCR amplification with primers for T2R, but obtained no evidence for the presence of even low concentrations of T2Rs.  Because the cells of the fungiform papillae are capable of sensing taste, it is logical that they would contain some other class of taste receptors.


I believe the authors present a great deal of evidence that T2Rs are a novel family of taste receptors, but I have read enough papers this semester to know that although many correlations might signify that a research group is headed the right direction, it takes functional tests to be certain as to the role of proteins and the genes that encode them.  This paper alludes to a previous experiment involving gustducin knockout mice that have decreased sensitivity to sweet and bitter tastes, which is in agreement with the notion that taste receptors are gustducin-linked.  However, it is possible that T2Rs are not involved in perception of taste, but rather some other set of gustducin-linked receptors plays a role.  Adler et al. present a convincing argument; they show that T2Rs are present at chromosomal loci suspected to contain taste receptors, that they are expressed in the correct place and are distinct from T1Rs, and even that they appear to be gustducin-linked.  But the definitive test, a functional one, is not conducted in this paper.  Though T2Rs are certainly excellent candidates for taste receptors, their actual function is uncertain.  Since the authors did not claim to have found the taste receptors, but rather candidates for taste receptors, they do in fact fulfill their claims.

Future Research:


In order to determine whether T2Rs are actually taste receptors, I would use homologous recombination to alter several T2Rs to different extents (completely knock them out in some experiments) and examine the effects on taste perception.  If taste perception is not altered in T2R knockout mice, then I would conclude that T2Rs are in the right place at the right time, but do not play a critical role in sensing tastes.  On the other hand, if perception ability fluctuated according to the extent of T2R disruption by homologous recombination, I would be satisfied that T2Rs are taste receptors.  This method might also help to characterize which T2Rs are responsible for which bitter tastes.  For example, if T2R-14 is knocked out and only taste perception of cyclohexamide is affected, I would know that T2R-14 is related to the perceiving the taste of cyclohexamide.

Elucidation of the T2R promoter and its control mechanism would be a valuable tool in investigating the function of the family of genes.  Perhaps upregulating or inhibiting T2R expression would have an effect on taste perception.  It might be possible to carry out such experiments using a transposable element.

The yeast two-hybrid system could be used to examine the extent of the interaction between gustducin and T2Rs.  Do they actually interact?  If the yeast cells that the two chimeric proteins are placed in turn blue (in the case of a LacZ reporter), it can be deduced that the proteins have a close interaction.

A good place to start with this line of experiments would be the fungiform papillae, since they are known to contain gustducin and lack T2Rs.  I would use an approach similar to that of Adler et al.  I would search for GPCRs distinct from T2Rs, and then search for related sequences in humans and rodents.  Next I would do some FISH to see if the genes I found are expressed in taste cells and are distinct from T1Rs and T2Rs.  If all goes well, functional tests (knockout mice) will follow, and I will have found a novel family of taste receptors.

If wild type T2Rs are knocked out of mice, what level of taste perception can be restored by inserting human or rat T2R dna?

Davidson College Molecular Biology Homepage

Davidson College Homepage

Questions, comments - contact: