Cl−-binding sites in the medaka fish taste receptor T1r2a/T1r3LBD. (A) Schematic drawing of the overall architecture of T1r2a/T1r3. The crystal structure (PDB ID: 5×2M) is shown at the LBD region, and helices B and C in T1r3 are labeled. (B) Anomalous difference Fourier map (4.5 σ, red) of the Br−-substituted T1r2a/T1r3LBD crystal. (C) Anomalous difference Fourier map (4.5 σ, red) of the Cl−-bound T1r2a/T1r3LBD crystal derived from the diffraction data collected at the wavelength of 2.7 Å. In panels B and C, the site originally identified the Cl−-binding was framed. (D) A close-up view of the Cl−-binding site in T1r3LBD in the Cl−-bound T1r2a/T1r3LBD (PDB ID: 5×2M). (E) Amino-acid sequence alignment of T1r proteins and the related receptors at the Cl−-binding site. The “h,” “m,” and “mf” prefixes to T1rs indicate human, mouse, and medaka fish, respectively. (F) The structures of ANPR (PDB ID: 1T34, left) (44) and mGluR2 (PDB ID: 5CNI, right) (45) bound with Cl−. (G) Superposition of the Cl−-binding site in T1r3, ANPR, and mGluR2.

Properties of the Cl−-binding to T1r2a/T1r3LBD.

The Cl−-binding properties of T1r2a/T1r3LBD. (A) Thermal melt curves of T1r2a/T1r3LBD in the presence of 0.003 - 10 mM concentrations of Cl−, measured by DSF. (B) Dose-dependent Tm changes of T1r2a/T1r3LBD by addition of Cl−. Data points represent mean and s.e.m. (n = 4). (C) Thermal melting curves of wild-type and the T1r3-T105A mutant of T1r2a/T1r3LBD in the presence and absence of Cl−, analyzed by FSEC-TS. Data points for the wild-type protein represent mean and s.e.m. (n = 3). (D) Dose-dependent FRET signal changes of the T1r2aLBD-Cerulean and T1r3LBD-Venus heterodimer by addition of Cl−. For comparison, the FRET index for the addition of 1 mM L-glutamine to the same protein sample (1.21 ± 0.003) was shown as an open triangle. (E) Dose-dependent FRET signal changes of the T1r2aLBD-Cerulean and T1r3LBD-Venus heterodimer induced by the addition of L-glutamine in the presence and absence of Cl−. For panels E and F, data points represent mean and s.e.m. (n = 3).

Results of single fiber recordings from the mouse chorda tympani nerve. (A) Representative recordings of sweet-responsive single fibers. The stimuli were 20 mM L-glutamine, 10 mM NMDG-Cl, or 20 mM L-glutamine + 10 mM NMDG-Cl. Lines indicate the application of stimuli to the tongue. After lingual treatment of T1rs blocker, gurmarin (Gur), responses were significantly inhibited. (B) Impulse frequencies in response to the concentration series of NMDG-Cl, NaCl, or KCl before and after Gur treatment. The mean number of net impulses per 10 seconds (mean response) ± s.e.m. in Gur-sensitive fibers (n = 5–6). (C) Impulse frequencies to 20 mM L-glutamine or 100 mM sucrose in the absence or presence of 10 mM NMDG-Cl before and after Gur treatment. Values are mean ± s.e.m. (n = 3). *, **: paired t-test; P < 0.05 (*), < 0.01 (**).

The structure of the regions relating to the Cl−-binding site in medaka fish T1r2a/T1r3LBD. (A) The Br−-binding site in T1r3. The anomalous difference Fourier map (4.5 σ, red) of the Br−-substituted crystal is overlayed to the Cl−-bound T1r2a/T1r3LBD structure (PDB ID: 5×2M). (B) The Cl−-binding site in T1r3. The anomalous difference Fourier map (4.5 σ, red) calculated from the data collected at the wavelength of 2.7 Å is overlayed. (C) The region in T1r2a, which corresponds to the Cl−-binding site in T1r3, in the Cl−-bound T1r2a/T1r3LBD. (D) The Br−-binding site in T1r2. The anomalous difference Fourier map (4.5 σ, red) is overlayed to the Cl−-bound T1r2a/T1r3LBD structure. (E) The Br−-binding site in T1r2. The 2.7 Å-anomalous difference Fourier map (4.5 σ, red) of the Cl−-bound T1r2a/T1r3LBD, which is the same map shown in panel B, is overlayed to the Br−-binding site in T1r2. No significant peak was observed.

The properties of T1r2a/T1r3LBD in the presence and absence of Cl−. (A) Binding analysis of gluconate by DSF. 0.1, 1, and 10 mM of L-glutamine, as a representative ligand, and sodium gluconate was added to T1r2a/T1r3LBD in 20 mM HEPES-NaOH, 300 mM NaCl, pH 7.5. The mean increases of Tm (ΔTm) from that in the absence of a ligand (54.1 °C, measured on the same sample, n = 2) are plotted. Error bars are in s.e.m. (n = 4). (B) Thermal melt curves of T1r2a/T1r3LBD in the presence or absence of Cl−, measured by DSF in the same condition in Figure 2C, and the panels C, D in this figure. (C, D) The L-glutamine-binding to T1r2a/T1r3LBD was measured by isothermal titration calorimetry. The upper and lower panels show the raw data and integrated heat signals upon L-glutamine injection to T1r2a/T1r3LBD in the presence (C) and absence (D) of Cl−. The binding isotherms were fitted assuming 1 ligand: 1 heterodimer binding.

X-ray data collection statistics of T1r2a/T1r3LBD-Fab16A complex.