An evolutionary conserved olfactory receptor for foodborne and semiochemical alkylpyrazines

Abstract

Molecular recognition is a fundamental principle in biological systems. The olfactory detection of both food and predators via ecological relevant odorant cues are abilities of eminent evolutionary significance for many species. Pyrazines are such volatile cues, some of which act as both human-centered key food odorants (KFOs) and semiochemicals. A pyrazine-selective odorant receptor has been elusive. Here we screened 2,3,5-trimethylpyrazine, a KFO and semiochemical, and 2,5-dihydro-2,4,5-trimethylthiazoline, an innate fear-associated non-KFO, against 616 human odorant receptor variants, in a cell-based luminescence assay. OR5K1 emerged as sole responding receptor. Tested against a comprehensive collection of 178 KFOs, we newly identified 18 pyrazines and (2R/2S)-4-methoxy-2,5-dimethylfuran-3(2H)-one as agonists. Notably, OR5K1 orthologs in mouse and domesticated species displayed a human-like, potency-ranked activation pattern of pyrazines, suggesting a domestication-led co-evolution of OR5K1 and its orthologs. In summary, OR5K1 is a specialized olfactory receptor across mammals for the detection of pyrazine-based key food odors and semiochemicals.

Keywords: chemical ecology; chemosensory evolution; odorant receptor; olfaction.

FIGURE 1

OR5K1 soley responds to 2,3,5‐trimethylpyrazine and TMT. OR5K1 is the sole responder in an automated screen of 616 OR variants against 100 µM 2,3,5‐TMP (A), and TMT at 300 µM (B). Data were normalized to the maximum response (OR5K1). OR families are color‐coded and sorted in ascending numerical order. Dashed line indicates 2σ‐threshold as signifier of activation. The negative control was a mock plasmid. Insert (A): Both the OR5K1 reference and its most common variant, OR5K1‐F₆₂L, presented measurable concentration/response‐relationships with. Inserts (B): Concentration/response‐relationship of various OR5K1 orthologs versus. Values are normalized to the highest signal of each respective receptor's maximum concentration/response‐relation (RLU/RLU_max). Data are presented as mean ± SD (n = 3). ● = apes, ◼ = prey species, ▲ = predators, bt = Bos taurus, cl = Canis lupus familiaris, oa = Ovis aries, pt = Pan troglodytes, pp = Pan paniscus, pc = Puma concolor, vv = Vulpes vulpes. C, Alkylpyrazines are the best KFO agonists of OR5K1. 177 key food odorants (listed in Table S2) according to Dunkel et al (2014) plus 2‐E‐3,5(6)‐DMP were screened at 300 µM against OR5K1. Data are means of triplicate determinations, and were normalized to the maximum response of 2,3‐DE‐5‐P. RLU = relative luminescence units. Bold numbers identify compounds according to Table 1

FIGURE 2

Mouse orthologs of the OR5K subfamily have diversified functions in response to 2‐Ethyl‐3,5(6)‐dimethylpyrazine and TMT. A, Maximum‐Likelihood‐Tree inferring the evolutionary relationships of the OR5K‐clade and its mouse orthologs plus Olfr1019. Numbers next to branches indicate percentage of replicate trees in which the associated taxa clustered together in a bootstrap test with 100 replicates. Grey boxes indicate human receptors. B‐F, Concentration/response curves of OR5K1 (B), mouse OR5K‐clade orthologs Olfr173 (C), Olfr175 (D), Olfr180 (E) and Olfr1019 (F), activated by both TMT and 2‐E‐3,5(6)‐DMP. ▲ indicates receptor interaction with TMT, while ● indicates interaction with 2‐E‐3,5(6)‐DMP. Data are normalized to the highest signal of each respective receptor's maximum response (RLU/RLU_max), and are presented as mean ± SD (n = 3). RLU, relative luminscence units

FIGURE 3

Compound and evolutionary clustering reveal evolutionary conserved agonist/receptor relationships for OR5K1‐orthologs. Clustering of compounds by their physico‐chemical properties was done using the clustering toolbox from ChemMine Tools. Key food odorants were identified according to Dunkel et al (2014), with addition of 2‐ethyl‐3,5(6)‐dimethylpyrazine (2‐E‐3,5(6)‐DMP). Studies that identified compounds as semiochemicals are referenced in Table S4. The evolutionary history of OR5K1, ppOR5K1 (Pan paniscus), ptOR5K1 (Pan troglodytes), Olfr173, btOR5K1 (Bos taurus), oaOR5K1 (Ovis aries), clOR5K1 (Canis lupus familiaris), vvOR5K1 (Vulpes vulpes), and pcOR5K1 (Puma concolor) was inferred from applying the Maximum‐Likelihood method and the Jones‐Taylor‐Thornton (JTT) matrix‐based model. The inferred evolutionary relationships and shared sequence identities of the receptors investigated are given in Figure S3. Color scale is based on EC₅₀ values ranging from 5.41 µM (lowest measured) to ≥1500 µM. The corresponding numerical values of the heat map are given in Table S6. ● indicates measurable concentration‐dependent responses that failed to fit to an EC₅₀. n.d., not determined. Data represent means of n = 3. Bold numbers in parentheses identify compounds according to Table 1

FIGURE 4

Pyrazines' in‐vitro EC₅₀ values on OR5K1 correlate with human odor thresholds. Odor thresholds and corresponding EC₅₀ values can be found in Table S5. Thresholds are detection thresholds, determined orthonasally in water. Bold numbers identify compounds according to Table 1. Pyrazines identified as both KFO and semiochemical are demarcated with red numbers. A non‐saturating concentration‐response relation of 2‐MOP prevented a logistic fit function‐derived EC₅₀ (Figure S4). Therefore, we assumed a least EC₅₀ value for 2‐MOP according to its concentration of maximum effect (1500 µM). R, Spearman's correlation coefficient. A 95%‐confidence interval is shaded in grey