Perceptual variation in umami taste and polymorphisms in TAS1R taste receptor genes

Abstract

Background: The TAS1R1 and TAS1R3 G protein-coupled receptors are believed to function in combination as a heteromeric glutamate taste receptor in humans.

Objective: We hypothesized that variations in the umami perception of glutamate would correlate with variations in the sequence of these 2 genes, if they contribute directly to umami taste.

Design: In this study, we first characterized the general sensitivity to glutamate in a sample population of 242 subjects. We performed these experiments by sequencing the coding regions of the genomic TAS1R1 and TAS1R3 genes in a separate set of 87 individuals who were tested repeatedly with monopotassium glutamate (MPG) solutions. Last, we tested the role of the candidate umami taste receptor hTAS1R1-hTAS1R3 in a functional expression assay.

Results: A subset of subjects displays extremes of sensitivity, and a battery of different psychophysical tests validated this observation. Statistical analysis showed that the rare T allele of single nucleotide polymorphism (SNP) R757C in TAS1R3 led to a doubling of umami ratings of 25 mmol MPG/L. Other suggestive SNPs of TAS1R3 include the A allele of A5T and the A allele of R247H, which both resulted in an approximate doubling of umami ratings of 200 mmol MPG/L. We confirmed the potential role of the human TAS1R1-TAS1R3 heteromer receptor in umami taste by recording responses, specifically to l-glutamate and inosine 5'-monophosphate (IMP) mixtures in a heterologous expression assay in HEK (human embryonic kidney) T cells.

Conclusions: There is a reliable and valid variation in human umami taste of l-glutamate. Variations in perception of umami taste correlated with variations in the human TAS1R3 gene. The putative human taste receptor TAS1R1-TAS1R3 responds specifically to l-glutamate mixed with the ribonucleotide IMP. Thus, this receptor likely contributes to human umami taste perception.

FIGURE 1

Variation in l-glutamate taste sensitivity. The left panel depicts a histogram of 242 subjects’ performances in a discrimination task: 29 mmol monosodium l-glutamate/L was tested against 29 mmol NaCl/L in 24 trials of a 3-alternative, forced-choice, triangle test. White bars depict chance performance and indicate monosodium l-glutamate–insensitive subjects. Dark gray bars indicate monosodium l-glutamate–hyposensitive subjects’ performances. Black bars indicate the performance of subjects who can significantly distinguish l-glutamate from sodium chloride. Two subjects depicted as light gray bars performed below chance. The right panel shows the test-retest correlation of 5 insensitive subjects and 5 sensitive subjects. The performances of 2 insensitive subjects were the same and thus are superimposed in the figure.

FIGURE 2

Mean (±SEM) modified Harris-Kalmus (mHK) recognition thresholds for umami taste. The same 10 subjects as were tested in Figure 1 (right panel) had their umami taste quality recognition thresholds measured by testing with monosodium l-glutamate (MSG). The dark bar represents the average response of umami-sensitive subjects, as categorized in Figure 1, and the light bar represents insensitive subjects. The y axis represents the concentration of MSG in mmol/L that was correctly identified as umami tasting; P < 0.05 (t test).

FIGURE 3

Mean (±SEM) concentration-intensity functions for sucrose and l-glutamic acid potassium salt (MPG). The left panel depicts the taste intensity rating functions on a general labeled magnitude scale (LMS) for sucrose and for MPG (right panel) for the same 10 subjects as were tested in Figures 1 and 2. Five concentrations of each stimulus and water were rated. The filled symbols represent the umami-sensitive subjects, and the open symbols represent the insensitive subjects as categorized in Figure 1. MPG was significantly different between the groups by using ANOVA at 50, 75, 100, and 200 mmol/L (P < 0.05).

FIGURE 4

A 2-alternative, forced-choice, intensity test of l-glutamic acid potassium salt (MPG) compared with sucrose. The same 10 subjects as were tested in Figures 1–3 each received 10 trials of 2 solutions in which they had to determine whether the 200 mmol MPG/L or 250 mmol sucrose/L tasted more intense. The solid bars represent the overall percentage of trials that MPG was selected as more intense, and the gray bars represent the overall percentage of trials that sucrose was selected as more intense. The left panel depicts data for 5 umami-sensitive subjects and the right panel for the 5 umami-insensitive subjects as categorized in Figure 1. P < 0.05 for the reversal (chi-square test).

FIGURE 5

Mean (±SEM) umami synergy tests with 200 mmol l-glutamic acid potassium salt (MPG)/L mixed with 3 mmol 5′-inosine monophosphate (IMP) or guanosine 5′-monophosphate (GMP)/L. The same 10 subjects as were tested in Figures 1–4 were tested. The dark bars represent the taste intensity ratings of solutions on a general labeled magnitude scale (LMS) for umami-sensitive subjects, and the white bars represent the insensitive subjects, as categorized in Figure 1. The x axis shows the stimuli: water, MPG, MPG mixed with IMP, and MPG mixed with GMP. Synergy was significant for both groups but was greater in intensity for the sensitive group; P < 0.05 (ANOVA).

FIGURE 6

Mean (±SEM) concentration-intensity curves for l-glutamic acid potassium salt (MPG) from a total of 87 subjects with extreme phenotypes whose TAS1R genes were sequenced. The figure depicts the overall average MPG concentration-taste intensity ratings on a general labeled magnitude scale (LMS). The x axis represents MPG concentration, and the y axis the average taste intensity ratings. The 87 subjects were divided into 2 approximately even groups of either umami-sensitive or -insensitive subjects. Differences in umami taste were evident between the 2 groups at 25, 50, 75, 100, and 200 mmol/L; P < 0.05 (ANOVA).

FIGURE 7

The open reading frame structures of the human TAS1R1 and TAS1R3 genes and their observed single nucleotide polymorphisms. The black boxes represent exons and the connecting spanners introns. The letter-number codes depict the locations and amino acid positions of identified polymorphisms. The first letter depicts the common amino acid and the second letter the substituted amino acid. The same letter indicates that the nucleotide change did not affect the amino acid code (synonymous change). The boxes indicate previously unreported polymorphisms.

FIGURE 8

Amino acid “ribbon” plots of the protein sequences for hTAS1R1 and hTAS1R3 with polymorphisms. TAS1R1 is represented on the left and TAS1R3 on the right. Note that TAS1R1 is a slightly larger protein than TAS1R3. The left side of each protein is the amino terminal, and the right side is the carboxy terminal. The dark horizontal lines depict the cellular membrane. Above these lines is extracellular space, and below these lines is intracellular space. Circles represent amino acids, and the single letter codes within each circle represent the amino acid identity. Black amino acids indicate the protein locations of nonsynonymous single nucleotide polymorphisms (SNPs), and amino acids with an asterisk above indicate the position of synonymous SNPs.

FIGURE 9

Expression and calcium mobilization by TAS1R1/TAS1R3. A, B: Human embryonic kidney 293 (HEK293) T cells were transiently transfected with plasmids that express TAS1R1 or TAS1R3. Surface expression of TAS1R1 (A) and TAS1R3 (B) was confirmed by staining with an antibody against an N-terminal epitope tag (FLAG). C: HEK-293 T cells were transiently transfected with plasmids that express TAS1R1, TAS1R3, and a chimeric G protein composed of G^a₁₆ containing the last 44 amino acids of G^a_-i3. Cell cultures were injected with a solution of either 5 mmol l-glutamate (L-Glu)/L and 1 mmol inosine 5′-monophosphate (IMP)/L or 5 mmol d-glutamate (D-Glu)/L and 1 mmol IMP/L. Data are represented as an average maximal fluorescence increase (n = 4; bars represent mean ± SEM). Only solutions containing l-glutamate induced calcium mobilization.