Expression Analysis of Taste Signal Transduction Molecules in the Fungiform and Circumvallate Papillae of the Rhesus Macaque, Macaca mulatta

The molecular mechanisms of the mammalian gustatory system have been examined in many studies using rodents as model organisms. In this study, we examined the mRNA expression of molecules involved in taste signal transduction in the fungiform papillae (FuP) and circumvallate papillae (CvP) of the rhesus macaque, Macaca mulatta, using in situ hybridization. TAS1R1, TAS1R2, TAS2Rs, and PKD1L3 were exclusively expressed in different subsets of taste receptor cells (TRCs) in the FuP and CvP. This finding suggests that TRCs sensing different basic taste modalities are mutually segregated in macaque taste buds. Individual TAS2Rs exhibited a variety of expression patterns in terms of the apparent level of expression and the number of TRCs expressing these genes, as in the case of human TAS2Rs. GNAT3, but not GNA14, was expressed in TRCs of FuP, whereas GNA14 was expressed in a small population of TRCs of CvP, which were distinct from GNAT3- or TAS1R2-positive TRCs. These results demonstrate similarities and differences between primates and rodents in the expression profiles of genes involved in taste signal transduction.

The expression profiles of genes involved in taste signal transduction have been partially uncovered in primates, including humans [23,24,25,26,27]. In situ hybridization (ISH) demonstrated that human TAS2Rs were expressed in heterogeneous populations of TRCs [23], whereas the expression of multiple Tas2rs occurred in the same subset of TRCs in mice [7]. On the other hand, Matsunami and colleagues demonstrated that each Tas2r was expressed in a much smaller number of TRCs than Gnat3 in mice [9]. The tissue distribution of expression of genes involved in taste signal transduction, including TAS1Rs, TAS2Rs, PKDs, and TRPM5, was examined in the CvP of the cynomolgus macaque, Macaca fascicularis [24], but the co-expression relationships among these genes largely remain to be elucidated. Moreover, the tissue distribution of expression of the majority of genes in the FuP has not been examined by ISH, except for PKD1L3 and TRPM5 [25].
In this study, we examined the mRNA expression of genes involved in taste signal transduction in more detail in the FuP and CvP of the rhesus macaque, Macaca mulatta, by ISH. We compared the gene expression profiles in the FuP and CvP and examined the co-expression relationships among various genes. We found both similarities and differences between macaques and rodents. This In situ hybridization revealed that three TAS1Rs, TAS2R13, PKD1L3, GNAT3, GNA14, and PLCB2 were robustly expressed in subsets of the TRCs in the CvP. These genes, except for GNA14, were also expressed in subsets of the TRCs in the FuP. n$2

Macaques
This study was carried out in strict accordance with recommendations in the Guide for Care and Use of Nonhuman Briefly, animals were kept in cages with sufficient space (780 mm wide, 650 mm depth, and 800 mm height) in the air conditioned room with sufficient environmental enrichment. The animals were housed in 12hour light-dark cycle conditions with a daytime light intensity of 150-300 lux and their intake of water, food, or selected nutrients was not restricted. In addition to normal pellet foods, the animals were occasionally fed sweet potatoes, fruits, and vegetables for nutrimental enrichment. To ameliorate suffering, anesthesia was induced by intramuscular injection of ketamine (2.5 mg/kg) with medetomidine (0.1 mg/kg) into the femoral or brachial muscle of the animals at the initial step of the experiments. After the animals were anesthetized and immobilized, pentobarbital sodium (25 mg/kg) was infused intravenously on the autopsy table. After the animals were deeply anesthetized, which was confirmed by the absence of pain response, they were sacrificed by bloodletting from the jugular vein. After a sufficient amount of time had elapsed, respiratory arrest, cardiac arrest, and pupillary dilatation were confirmed. Then, taste tissues from five rhesus macaques (Macaca mulatta; approximately 3-year-old males), which were scheduled for euthanasia not only for this study but also for other experimental purposes, were collected and embedded in O.C.T. compound (Sakura Finetek, Tokyo, Japan) by ourselves.
All the experiments described above were performed at Primate Research Institute, Kyoto University.

Database Search and Cloning
The TBLASTN program was used to search for genomic sequences showing significant identity with human TAS1Rs, TAS2Rs, GNAT3, GNA14, and PLCB2 in the public genome database of the rhesus macaque (http://www.ensembl.org/ Macaca_mulatta/Info/Index). The macaque TAS2Rs were named following the nomenclature proposed by Dong et al [28]. The cDNA sequence corresponding to G756-E852 of T1R3, which was unknown due to the lack of a genomic sequence, was obtained by 39-RACE using 39-Full RACE Core Set (Takara Bio Inc., Shiga, Japan). The entire coding regions of TAS1R1, TAS1R2, TAS1R3, TAS2Rs, GNAT3, and GNA14 and partial coding regions of PKD1L3 (C42-Y749) and PLCB2 (M1-K192), which were amplified from macaque cDNA synthesized from epithelial tissues containing circumvallate papillae or genomic DNA extracted from tongue tissue, were used as probes.

Expression of Taste Receptors and Signal Transduction Molecules in the Fungiform and Circumvallate Papillae
To examine the tissue distributions of expression of genes involved in taste signal transduction, we conducted in situ hybridization on sections of the FuP and CvP using the following genes as probes: TAS1R1, TAS1R2, TAS1R3, TAS2Rs, PKD1L3, GNAT3, GNA14, and PLCB2. In the CvP, three TAS1Rs, PKD1L3, GNAT3, GNA14, and PLCB2 were robustly expressed in subsets of TRCs ( Figure 1A). Certain TAS2Rs, such as TAS2R13, were robustly expressed in subsets of TRCs, whereas only weak signals were observed for other TAS2Rs, including those located on chromosomes 3 and 6 ( Figures 1A and B; Figure S1). In the FuP, three TAS1Rs, GNAT3, and PLCB2, but not GNA14, were robustly expressed in subsets of TRCs, whereas TAS2R13 and PKD1L3 were weakly expressed in subsets of TRCs ( Figure 1A). It should be noted that TAS1R1, TAS1R2, and TAS2R13, as well as PKD1L3 [25], were expressed in both the FuP and the CvP.

Co-expression Relationships among Taste Signal Transduction Molecules
To compare the TRCs expressing each taste signal transduction molecule, we next performed double-label fluorescence ISH. T1R1 and T1R2 form heteromers with T1R3 to function as umami (savory) and sweet taste receptors, respectively [4,5,30]. TAS1R1 and TAS1R2 were exclusively expressed in different subsets of TAS1R3-positive TRCs in the FuP and CvP (Figure 2; Tables S1 and S2). In the CvP, approximately 20% and 40% of TAS1R3-positive TRCs were also positive for TAS1R1 and TAS1R2, respectively. Experiments using a mixed probe for TAS1R1 and TAS1R2 combined with a probe for TAS1R3 confirmed that approximately 40% of TAS1R3-positive TRCs were negative for both TAS1R1 and TAS1R2 (Figure 2A and data not shown). In the FuP, approximately 40% and 30% of TAS1R3positive TRCs were also positive for TAS1R1 and TAS1R2, respectively ( Figure 2B; Table S2). In summary, TAS1R3-positive TRCs in the FuP and CvP can be classified into three types of cells: cells expressing TAS1R1+TAS1R3, those expressing TAS1R2+TAS1R3, and those expressing TAS1R3 alone.
A previous gene expression analysis using microarrays in the taste buds of the cynomolgus macaque collected by laser capture microdissection revealed that TAS1R1 and TAS1R2 were more highly expressed in the taste buds of the FuP than in those of the CvP [24]. Our ISH analysis quantified the expression of the genes involved in taste signal transduction at the cellular mRNA level. Consequently, the majority of genes, including TAS1R1 and TAS1R2, showed more uniform expression patterns in the FuP and   the CvP of macaques than in those of rodents. These results are consistent with previous findings from gustatory nerve recordings in the rhesus macaque showing that both the chorda tympani and glossopharyngeal nerves, which innervate the FuP and CvP, respectively, responded to a variety of basic taste compounds [31].
TAS2R13, TAS2R15, and TAS2R23, which were robustly expressed in subsets of the TRCs in the CvP (Figure 1B), reside in different TAS2R gene clusters on chromosome 11 [28] ( Figure  S1; http://www.ensembl.org/Macaca_mulatta/Info/Index). Almost all the TAS2R15and TAS2R23-positive TRCs were also positive for TAS2R13, whereas TAS2R15-positive TRCs partially overlapped with those expressing TAS2R23 ( Figure 3A; Table S3). We used a mixed probe for TAS2R2, TAS2R3, TAS2R4, TAS2R5, and TAS2R6 because only weak signals were detected for each TAS2R located on chromosome 3 ( Figure 1B). When we compared the TRCs expressing TAS2Rs located on different chromosomes, almost all the TRCs labeled with the mixed probe were also positive for TAS2R13, but they partially overlapped with TRCs expressing TAS2R15 or TAS2R23 ( Figure 3B; Table S3). These results demonstrate that each TRC sensing bitter compounds expresses various combinations of TAS2Rs ( Figure 3C), as in the case of human TAS2Rs [23].
We next compared the TRCs expressing taste receptors for different basic taste modalities. We chose TAS2R13 as a representative TAS2R because almost all the TRCs expressing other TAS2Rs that we tested were included in the TAS2R13-positive TRCs, as described above (Figure 3; Table S3). TAS1R1 and TAS1R2 were exclusively expressed in different subsets of TAS1R3positive TRCs (Figure 2). TAS1R3-positive TRCs were negative for TAS2R13 in the CvP ( Figure 4A) and in the FuP ( Figure 4B). TAS1R1-, TAS1R2-, TAS1R3-, and TAS2R13-positive TRCs were also positive for PLCB2 in the CvP ( Figure 4A; Table S1) and in the FuP ( Figure 4B), as in the case of other vertebrates such as rodents and fish [29,32]. PLCB2-positive TRCs were negative for PKD1L3 in the CvP ( Figure 4A) and in the FuP ( Figure 4B). In summary, TAS1R1, TAS1R2, TAS2Rs, and PKD1L3 were exclusively expressed in different subsets of the TRCs in the FuP and CvP ( Figure 5D). Thus, these results suggest that the TRCs detecting different basic taste modalities are mutually segregated in macaque taste buds.
Finally, we focused on two genes encoding G protein a subunits, GNAT3 and GNA14, which are specifically expressed in subsets of rodent TRCs [10,13,22,33]. In the CvP, GNA14 was expressed in a much smaller population of TRCs than GNAT3 and in a mutually exclusive manner (Figures 1A and 5A; Table S1). GNA14-positive TRCs were distinct from those expressing TAS1R2 and TAS2R13, but they formed subsets of TAS1R3positive TRCs and partially overlapped with TAS1R1-positive TRCs ( Figure 5A; Table S1). In contrast, TAS1R2 and TAS2R13 were expressed in different subsets of GNAT3-positive TRCs, which partially overlapped with TAS1R1and TAS1R3-positive TRCs ( Figure 5B; Table S1). It should be noted that TAS1R2 was co-expressed with GNAT3 but not with GNA14 in macaques, whereas Tas1r2 was primarily co-expressed with Gna14 but not with Gnat3 in rodents [10,13,22]. In the FuP, GNAT3, but not GNA14, was expressed in TRCs ( Figure 1A). TAS1R1, TAS1R2, TAS1R3, and TAS2R13 were expressed in subsets of GNAT3positive TRCs ( Figure 5C; Table S2). These results suggest that GNAT3 plays a pivotal role in mediating sweet, bitter, and umami tastes in macaques.

Conclusions
We investigated the expression of taste receptors and signal transduction molecules in the FuP and CvP of the rhesus macaque and further examined the co-expression relationships among these genes. The majority of genes exhibited more uniform expression patterns in the macaque FuP and CvP than in these papillae in rodents ( Figure 5D). Intriguingly, there were several differences between the expression profiles of macaques and rodents. First, TAS1R1 and TAS1R2 were more uniformly expressed in both the FuP and the CvP of macaques than in rodents. Second, TAS1R2 was co-expressed with GNAT3 in the CvP but not with GNA14. Third, macaque TAS2Rs were expressed in heterogeneous populations of TRCs in the CvP. These results suggest that the molecular mechanisms underlying taste transduction in primates, including humans, may be different from those in rodents and that the macaque is an important model organism for taste perception in humans. Figure S1 Schematic drawing illustrating the locations of macaque TAS2Rs on the chromosomes. TAS2R1-8, TAS2R11, and TAS2R38 are located on chromosome 3, whereas only TAS2R26 is located on chromosome 6. The other 15 TAS2Rs are located on chromosome 11, although precise location of TAS2R9 has not been determined. TAS2R13, TAS2R15, and TAS2R23 reside in different TAS2R gene clusters on the chromosome 11. (TIF) Table S1 The percentages of TAS1Rs, GNAT3, GNA14, and PLCB2 co-expression in the circumvallate taste buds. The percentage values were calculated by dividing the number of cells expressing both gene X and gene Y by the number of cells expressing gene X. (DOCX)

Table S2
The percentages of TAS1Rs, TAS2R13, and GNAT3 co-expression in the fungiform taste buds. The percentage values were calculated by dividing the number of cells expressing both gene X and gene Y by the number of cells expressing gene X. (DOCX)

Author Contributions
Conceived and designed the experiments: YI. Performed the experiments: YI MA. Analyzed the data: YI MA. Contributed reagents/materials/ analysis tools: HI. Wrote the paper: YI TA KA. Figure 5. The co-expression relationships among taste receptors and G protein a subunits. (A) In the CvP, GNA14 was expressed in a much smaller population of TRCs than GNAT3 and in a mutually exclusive manner. The GNA14-positive TRCs were distinct from those expressing TAS1R2 and TAS2R13, but they were subsets of the TAS1R3-positive TRCs and partially overlapped with the TAS1R1-positive TRCs. n $1 (numbers of sections $2). (B) In the CvP, TAS1R2 and TAS2R13 were expressed in subsets of the GNAT3-positive TRCs, which partially overlapped with the TAS1R1-and TAS1R3-positive TRCs. n $2 (numbers of sections $4). (C) In the FuP, TAS1R1, TAS1R2, TAS1R3, and TAS2R13 were expressed in subsets of GNAT3positive TRCs. n $1 (numbers of sections $10). (D) Venn diagram illustrating the co-expression relationships among taste receptors and signal transduction molecules in the macaque and the mouse.