A Review of:
A Novel Family of Mammalian Taste Receptors

Adler, E., M.A. Hoon, K.L. Mueller, J. Chandrashekar, N.J.P. Ryba, and C.S. Zulker. 2000. Cell. Col. 100: 693-702.

It is believed that sweet, bitter, sour, salty, and umami are the only five types of taste that are distinguishable by mammals.  These five taste types, or "modalities" are thought to be "mediated by distinct transduction pathways expressed in subsets of receptor cells" (p. 693) which are found in the taste buds.  The taste buds are located on palate and tongue epithelia.  Bitter, the modality being focused on here,  as well as sweet and umami seem to be "mediated by G protien-coupled receptor (GPCR) signaling pathways" (p.693).  Once the above receptors interact with "tastants, they trigger "signaling cascades" which ultimately cause the release of neurotransmitter which in turn leads to the transmitting of signals to the cortical taste centers through the thalamus by the "afferent nerve fibers from cranial nerve ganglia."  The taste information is then processed in the the cortical taste centers.  The authors of this paper previously isolated "two novel GPCRs," T1R1 and T1R2.  They also found that these two receptors "are putative receptors expressed in subsets of taste receptor cells of the tongue ad palate epithelia.  Further, it was found through In situ hybridization that T1Rs are transcribed in about 30% of taste bud cells which is close to the percentage of cells in which the G protien alpha subunit gustducin is transcribed.  Despite this commonality,  the researchers believe that there is another group of GPCRs that are expressed in cells containing gustducin because it appears that T1Rs are mainly "not co-expressed with gustducin." There is evidence to support the idea that gustducin plays a role in the detection of bitter taste.

The first experiments discussed in this paper were those conducted to identify "novel candidate taste receptors" (p. 696).  Due to the background information above, "GPCRs in genomic intervals linked to bitter taste perception" (p. 695) were searched for.   Resulting from previous evidence linking "a locus at 5p15" to response to a bitter compound,  this 5p15 was searched for "genes encoding candidate transmembrane proteins" using "DNA sequencing databases".  Open reading frames of DNA of "six sequenced human genomic BAC clones were analyzed revealing T2R-1, a novel GPCR.  T2r-1 was found to have multiple conserved sequences found in may GPRCs as well as seven possible transmembrane domains as seen in Figure 1.  T2R-1 and "related sequences" were used in further computerized searches.  These searches resulted in the discovery of 12 more complete receptors and 7 pseudogenes which are also shown on Figure 1.  Figure 1 illustrates the predicted amino acid sequence of various T2R genes of human, mouse, and rat.  Clusta/W was used.  As can be seen by the black and gray shaded boxes, 30%-70% of the amino acid sequence is shared by each of the T2Rs found***.  Furthermore, "most share highly conserved sequence motifs in the first three and last transmembrane segments, and also in the second cytoplasmic loop" (p. 695).  The investigators also claim that "The most divergent regions between T2Rs are the extracelluar segments, extending part way into the transmembrane helices.  We presume that the high degree of variability between T2Rs reflects the need to recognize many structurally diverse ligands" (p. 695).  The main weakness of this data and thus the theories based upon it is that Figure 1 is computer generated data rather than data gathered in the lab.  However, this figure does help to support the authors' theories and illustrates similarites that may constitute a novel family of taste receptors.  The conclusion was made from Figure 1 along with Figure 2 that the T2Rs found did in fact constitute*** "a novel family of GPCRs distantly related to V1R vomeronasal receptors and opsins" (p. 695).  Figure 2 shows a phylogenic tree demonstrating the "sequence relationships" between T2Rs of mouse, rat, and human, as well as opsin and V1R.  This figure does clearly illustrate these relationships and shows that the various T2Rs do appear to be more related to each other than to the V1Rs or opsin.  Again, the weakness, though small,  in this figure is that it is computer generated.  However, this data does support the conclusions.

The next the authors dealt with the organization of the Human T2R genes.  While the clarity and detail of the methods leave some to be desired, it appears that PCR of Genomic DNA was used to separate various  complete hT2Rs.  These hT2Ra were then used to probe mouse BAC filter arrays and a rat circumvallate cDNA library.  Then, "Southern hybridization experiments were used to identify a nonredundant set of positive BACs and to order overlapping BACs."  This blott was not shown in the paper**.  In addition to a "mouse/hamster radiation hybrid panel," inspection of "the strain distribution pattern of single nucleotide polymorphisms in a panel of C57BL/6J X DBA/2J recombinant inbred lines was used to map the T2Rs of mice.  Figure 3 illustrates the placements of various human and mouse T2R genes.  The figure shows that the currently known T2R genes are located on chromosomes 5, 12, and 7 and tend to be found in clustered in "head-to-tail arrays." Interestingly, the degree of similarity between the receptors varies with an array.  In addition, Figure 3 demonstrates that the organization of mouse T2Rs seems to reflect that of the human T2Rs.  The authors also conclude that the organization of these genes is similar to that seen in mouse, fly, human, and worm olfactory receptor genes.  The authors also estimated the the size of the T2R gene family to be about 40 to 80.  This was done by  first utilizing the Genome Sequence Survey database.  This database found only 12 partial sequences of interest.  However, the authors estimated that there might actually be around 90 T2R genes in the human genome because this database only demonstrates a sampling of about 14% of the human genome.  36 complete and 15 partial sequences were then obtained from other databases which only illustrate 50% of the human genome thus suggesting again that there are around 100 T2R genes in the entire genome.  Rightly so, the authors acknowledge the limitations of these databases and therefore estimated 80 to 120 T2R genes.  This number was lowered to the final estimate of 40-80 functional genes because about 1/3 of the complete T2Rs were found to be pseudogenes.  The estimate of 40 - 80 is close to that seen for the olfactory receptors of humans.

The third task of the authors was to demonstrate the "T2R gene Are Linked to Loci Involved in Bitter Taste."  Going back to Figure 3, Figure 3 also shows regions of gene clusters that have been determined to be linked to responses to bitter substances in previous experiments by other labs.  As seen in Figure 3, The authors found that a cluster of nine T2R genes on chromosome 12 "maps to an interval that is homologous with the mouse chromosome 6 bitter cluster" in addition to including three PRP (salivary proline rich protein).  The authors concluded that this information was further evidence in support of the theory that T2R receptors are receptors for bitter taste.  The methods mentioned in the last section on the organization of human T2R genes, were used to map the genes and show the relationship between the human and mouse T2Rs and the bitter clusters.  Screening of mouse mouse libraries with hT2Rs resulted in 61 BAC clones with 28  mouse T2Rs.  The radiation hybrid along with the "recombinant inbred strain mapping studies" served to illustrate the the locations of the mouse gene clusters***.  With this information it was observed (see Figure 3) that every known  mouse and corresponding human T2R gene were found on homologous sections of each species' genome.  The C57BL/6J X DBA/2J panel analysis mentioned above was used discovering that Prp and the bitter cluster on mouse chromosome 6 "cosegregate" with the T2R receptors found in the three BAC contigs.  This data in not shown.  Further, The authors also found that the mouse Prp gene is located in the T2R cluster on mouse chromosome 6.  Figure 3 and the above experiments are more convincing than the last two figure because it is now actual experimental data that is being dealt with.  While it is disappointing that none of the raw data from the expirements is shown, the information illustrated in Figure 3 does seem to adequately support the authors' conclusion that "these results demonstrate that T2Rs are intimately linked to loci implicated in bitter perception, and substantiate the postulate that T2Rs may function as receptors."

The rest of the experiments conducted were in situ hybridizations.  The first of these experiments was to determine if "T2Rs Are Expressed in Taste Receptor Cells."  Figure 5 illustrates the results from in situ hybridization conducted on the rat taste buds found in the circumvallate (5a-e), foliate (5f), and fungiform papillae (5i), epiglottis (5h), and geschmackstreifen (5g).  Five T2R dioxigenin-labled antisense RNA probes were used.  These probes were determined by screening a rat circumvallate cDNA library and distinguishing 14 T2Rs cDNAs.  As seen in Figure 5, all taste buds, except those found in the fungiform papillae, have cells which appear to express T2Rs.  It seems that less than 10% of the fungiform buds have cells with seem to express T2Rs.  Further, it can be seen that the T2Rs seem to only be expressed in certain "subsets" of cells.  The authors also stated that similar results to those above were obtained with 11 rat T2Rs as well as when mouse hybridizations were done with 17 separate mT2R probes.  This data, however, is not shown.  The above results do seem to support the theory that T2Rs are expressed in taste receptor cells and therefore may function as taste receptors.  However, it is a concern that in the Experimental Procedures both negative and positive controls for the experiment shown in Figure 5 is described, yet neither is shown.  Sense probes were used in the  negative control and cDNA of the gene for the G alpha subunit was used in the positive control.  Both controls illustrated the expected result.  In addition to the controls, showing the data, or at least representative data from experiments in -volving the mouse hybridizations and the 11 rat T2Rs would have been helpful and more convincing than simply stating the results.

Next the authors looked at whether "Individual Receptor Cells Express Multiple T2R Receptors."  Because it was shown in the previous experiment that about 15% of cells in all  but the fungiform taste buds express a T2R (based on the fact that the sections from the last part were 1/5-1/3 the thickness of the taste bud combined with the number of positive cells) and the genome of a rodent possesses over 30 T2Rs, the hypothesis was made that "a taste cell must express more than one receptor."  Figure 6 illustrates the result of in situ hybridizations using circumvallate taste buds and mixtures of two different T2R probes (a), five probes (b), and 10 probes (c).  6d shows in situ hybridization using two fluorescent labels on different T2Rs (T2R-3, green, and T2R-7, red).  According to the figure legend and the Experimental Procedures, this experiment was also conducted with various other combinations of probes yielding similar results.  The authors claim that the number of positive cells seen using multiple probes is only slightly higher than the number of positive cells found when only one probe was used, but they say that intensity of the signal was "significantly enhanced" in the hybridizations using more than one probe.  In addition, such results were also found when this experiment was conducted on taste buds of other sections of the mouth.  There are two main problems here.  First of all, no controls are shown to allow the reader to be totally confident comparing the signal intensities seen in Figures 5 and 6.  Furthermore, the data from hybridizations conducted on the different regional taste buds is not shown.  Also, according to the Experimental Procedures, a Northern blott and another hybridizations were done to make sure that cells other than those found in taste tissue do not "widely" express T2Rs.  Although, panel 6d, the double fluorescent labeling, was conducted on samples from minimum three animals, 50 taste buds each with similar, and the results of this representative panel are very convincing and lend further credibility to the other data.  Thus the authors' claims that their results (a) show that "there are marked topographical differences in the expression patterns of candidate signaling molecules in the various taste buds and papillae," (b) "the complexity of the receptor repertoire is significantly larger that previously though, and (c) "each cell expresses multiple receptors," are supported but not as well as they could be.

Finally, the authors explored whether "T2R Genes Are Selectively Expressed in Gustducin-Expressing Cells."  In Figure 7, the results from double-label fluorescent in situ hybridization of taste receptor cells with (a) T2R probes, (b) gustducin probes, (c) gustducin and T2R probes, and (d) T1R and T2R probes.   From this figure it is concluded that T1Rs and T2Rs are expressed by different cells, T2Rs are expressed in the solely in cells expressing gustducin, and gustducin appears to be express in cells not expressing T1Rs.  The authors addressed two problems with this data.  First, about 1/3 of the taste bud cells expected to express gustducin were not labeled by the mixture of 10 T2R probes used.  The possible explanations given were that these cells were in a different stage of development, or these cells express receptors that less related.  The second problem was with fungiform taste buds.  A large amount of the gustducin cells found "in the front of the tongue do not co express members of the T2R family of receptor" due to the previous conclusion that 10% of the fungiform buds express have any cells that express T2R.  The authors said that T2R expression levels may simply be to low to detect in these cells, yet the authors feel this is not correct.  The authors site experiments for which no data is shown including one in which "mixed probes and extended developing times" were used and the number of positive cells was the same.  Another experiment included T2R primers and fungiform taste buds in PCR illustrating a negative result for the expression of T2Rs.  Finally the authors claim that all the T2Rs probed for were expressed in those fungiform buds that did express any T2R :suggesting that all receptors are also expressed in the front of the tongue, but in a much smaller subset of taste buds."  The results of these last experiments would in fact support this conclusion, however, it would not be prudent for the reader to take the word of the author without seeing any actual data.  Further,  another positive experiment would be beneficial considering the negative result PCR.  Further data supporting the linkage between gustducin and T2Rs is not shown including results which illustrate the presence of gustducin in other tissue and the expression of certain T2Rs in these gustducin cells.  Again, this data definitely does support the idea that T2Rs are "gustducin-linked taste receptors," however the reader would more convinced if the data were shown.

All the above data and results do in support the authors' original claim that "We have identified a novel family of 40-80 human and rodent G protien-coupled receptors expressed in subsets of taste receptors cells of the tongue and palate epithelia."  While the data does provided support for this conclusion, there is not enough support for the reader to be fully convinced.  As mentioned above, much of the results are estimates taken from computer programs, not experimental data (although this can generally be accepted according to Dr. Campbell and the use of multiple databases does increase the credibility).  Further, the controls are not shown and while it is understandable that not all data can be shown in such a paper, there were experiments for which no representative data was even shown.  This paper could have actually been broken down into several papers may have allowed the authors to go into more depth and show more data, thus hopefully being more convincing.  Perhaps time and money were an issue as usual.

Despite some problems, the data is enough to suggest multiple future experiments.  For example, the authors of this paper wrote an accompanying paper entitled "T2Rs Function as Bitter Taste Receptors" (Chandrashekar, Mueller, Hoon, Adler, Feng, Guo, Zuker, and Ryba, 2000) in which the authors demonstrate that certain T2Rs do in fact function as bitter taste receptors. A "heterologous expression system" is used.  Another possibility would be that mentioned in the conclusion of this paper, "How might co-expression of T2Rs be controlled?"  It is also mentioned in the conclusion that a common "motif" was found upstream of the start codon of most human T2Rs.  Thus homologous recombination could be used to knock out this common motif probably as it occurs upstream from the mT2R-4, mT2R-5, and mT2R-14 cluster as suggested in the paper because of the "head-to-tail array" organization of the cluster.  Once the parent generation mice had been breed, in situ hybridization using T2R-4, T2R-5, and T2R-14 cDNA probes (antisense to the mRNA for these three T2Rs) separately could be conducted on each of the regions of the mouth containing taste buds to determine if these T2Rs are still expressed or if they are expressed more which  can be seen by intensity.  Controls consisting of the same probes hybridized with the same sections on wild type mice would be needed both to confirm that the hybridization worked and to an intensity comparison.  The same could be done using double-label fluorescent in situ hybridization to see if co-expression is changed.  The same type of control would be needed. RT-PCR analysis with primers specific to each of the three T2Rs could also be used to determine changes in expression.  Again, a wild type control would be needed.

References:
Campbell, A. Malcolm. conversation. 20 April, 2000.

Chandrashekar, J., K.L. Mueller, M.A. Hoon, E. Adler, L. Feng, W. Guo, C.S. Zuker, and N.J.P. Ryba. 2000. T2Rs Function as Bitter Taste Receptors. Cell. Vol. 100: 703-711.