Essential Role of the Main Olfactory System in Social Recognition of Major Histocompatibility Complex Peptide Ligands

The mouse MOE detects MHC class I peptide ligands. A, Peptide ligands for prototypical representatives of MHC class I molecules used in this study. Sequences (in one-letter code) and sources were taken from Rammensee et al. (1997), which provides detailed primary references. Anchor residues defining haplotype-specific binding characteristics are highlighted in bold, and mutated residues in control peptides are underlined. In the ANPRAFDTE peptide, anchor positions 5 and 9 of the WT peptide were replaced with residues unfavorable for binding to the D^b MHC molecule (Rammensee et al., 1997), whereas the overall amino acid composition remains identical. minor H Ag, Minor histocompatibility antigen; TK, tyrosine kinase. B, Examples of peptide-evoked MOE field potentials and their dose dependence (each series from a single location) to varying concentrations of either AAPDNRETF (representative of 43 recordings in 5 animals) or SYFPEITHI (representative of 38 recordings in 12 animals). Responses are produced by 500 ms pulses of peptides as indicated. C, Average dose–response plots of peak responses to AAPDNRETF (filled circles, black curve) and SYFPEITHI (open circles, red curve). The results for two control peptides, AAPDARETA (solid triangles) and SAFPEITHA (open triangles), respectively, are also shown. Smooth curves are fitted by the Hill equation, with K_1/2 values and Hill coefficients of 190 pm, 0.6 (SYFPEITHI); 240 pm, 0.6 (AAPDNRETF); 5.4 nm, 0.8 (SAFPEITHA); and 8.3 nm, 1.3 (AAPDARETA). D, Medial view of the mouse endoturbinate system after dissection. Roman numerals designate individual turbinates. Sixty spatial locations of 96 spots stimulated that exhibited a field potential response to AAPDNRETF and/or SYFPEITHI at 10⁻¹⁰ m are shown. The other 36 locations that did not respond are included in the map in Figure 3A, which documents the entire data set. Scale bar, 1000 μm. E, F, Specificity of peptide-evoked MOE field potentials assessed by two control peptides in which the characteristic anchor residues of MHC class I ligands were replaced by alanines [i.e., AAPDARETA (8 recordings, 3 animals) and SAFPEITHA (14 recordings, 4 animals), respectively]. These peptides failed to elicit a field potential at 10⁻¹⁰ m (E). A scrambled version of the H2-D^b ligand AAPDNRETF, ANPRAFDTE (10⁻¹⁰ m), and a mixture containing all amino acids (aa mix; in free form, each at 10⁻¹⁰ m) that constitute the SYFPEITHI peptide elicited only very small or no responses (E, F). Each pair of responses in E corresponded to the same spatial location. Error bars indicate SEM.

Spatial distribution of peptide-evoked field potentials on endoturbinates IIa, IIb, III, and IV. A, Map of 96 spatial locations that were stimulated sequentially with AAPDNRETF, SYFPEITHI, and 2-heptanone (each at 10⁻¹⁰ m). Color-coded symbols indicate whether a given location responded to none, one, two, or all three of these stimuli. Scale bar, 1000 μm. B, Examples of field potential responses registered in different locations. The numerals (1–7) refer to a given location as shown in A. Ejecting bath solution (control) onto the MOE gave no responses. C, Analysis of the data shown in A with respect to the frequency of observing a response to each of the three ligands (at 10⁻¹⁰ m) on a given endoturbinate (n, number of locations tested on each turbinate).

Recognition of MHC peptides by the MOE requires cAMP signaling and the CNG channel subunits A2 and A4. A, Inhibition of a peptide-induced field potential by SQ22536 (300 μm). B, Summary of the effects of pharmacological blockers on peptide-evoked potentials. Each response was normalized to its own control response obtained before application of each drug as shown in A. C, Representative peptide-induced responses from the MOE of WT and CNGA2^−/− mice. Responses to 2-heptanone (10⁻¹⁰ m) are shown for comparison. In accord with Lin et al. (2004), we observed residual EOG responses to 2-heptanone in CNGA2^−/− mice at a much higher concentration (10⁻⁴ m; n = 9) (data not shown). D, Average size of peptide-evoked MOE potentials in mice deficient for CNGA2 or CNGA4 compared with WT littermate controls. E, Representative responses to a prolonged 6 s peptide pulse in WT and CNGA4^−/− MOE. Fit is a single-exponential decay with a time constant (τ) of 1.1 s. F, G, Diminished adaptation rate of peptide-evoked potentials in CNGA4^−/− MOE as observed by measuring plateau/peak ratios (F) or desensitization time constants (G), respectively (means ± SEM; Student’s t test; **p < 0.001). The peptide-evoked potentials were recorded on endoturbinate III in response to SYFPEITHI (10⁻¹⁰ m). The number of experiments is indicated above each bar. See Results for rationale.

In situ Ca²⁺ mapping reveals distinct populations of peptide-sensitive OSNs. A, Transmitted light image of a mouse coronal MOE slice (P7). DR, Dorsal recess; S, nasal septum. A portion of the sensory epithelium delimited by the white box is shown at a higher magnification in B and C. B, Confocal fluorescence image of the fluo-4-loaded MOE acquired at rest (grayscale). C, The same region is depicted as a merged pseudocolor image of the relative increase in peptide-induced Ca²⁺-dependent dye fluorescence (ratio between the peak fluorescence before and after stimulation, ΔF/F). In this example, AAPDNRETF (10⁻⁹ m; green) activated three OSNs (cells 1, 2, and 4), and SYFPEITHI (10⁻⁹ m; red) activated one OSN (cell 3). In some cases, individual somata, dendrites, and dendritic knobs can be clearly distinguished. D, Time course of peptide-induced Ca²⁺ responses from the same cells that are shown in C. The tissue was successively stimulated with five different peptides (each at 10⁻⁹ m), followed by forskolin (20 μm). Note that cells 1 and 2 each recognized two peptides with different anchor residues. The arrows indicate the time points at which stimulus application was turned on. Scale bars: A, 200 μm; B, C, 10 μm.