Positive Allosteric Modulation of Insect Olfactory Receptor Function by ORco Agonists

Abstract

Insect olfactory receptors (ORs) are heteromeric ligand-gated cation channels composed of a common olfactory receptor subunit (ORco) and a variable subunit (ORx) of as yet unknown structures and undetermined stoichiometries. In this study, we examined the allosteric modulation exerted on Anopheles gambiae heteromeric ORx/ORco olfactory receptors in vitro by a specific class of ORco agonists (OAs) comprising ORcoRAM2 and VUAA1. High OA concentrations produced stronger functional responses in cells expressing heteromeric receptor channels relative to cells expressing ORco alone. These OA-induced responses of ORx/ORco channels were also notably much stronger than those obtained upon administration of ORx-specific ligands to the same receptors. Most importantly, small concentrations of OAs were found to act as strong potentiators of ORx/ORco function, increasing dramatically both the efficacy and potency of ORx-specific odorants. These results suggest that insect heteromeric ORs are highly dynamic complexes adopting different conformations that change in a concerted fashion as a result of the interplay between the subunits of the oligomeric assemblies, and that allosteric modulation may constitute an important element in the modulation and fining tuning of olfactory reception function.

Keywords: Anopheles gambiae; ORco agonists; cell-based screening; ligand discovery; malaria; mosquito olfaction; olfactory function enhancement; olfactory receptor pharmacology.

FIGURE 1

Direct agonism of heteromeric and homomeric OR receptor function by ORco agonists (OAs). Results of experiments employing OrcoRAM2 as an OA are shown. (A) Magnitude of responses obtained from cells expressing ORco, OR1, OR2, OR9, OR53 alone or the same ORx subunits as heteromers with ORco, in the presence of 100 μM of the OA. (n = 2 for OR9, OR53, and OR53/ORco; 3 for OR1, OR2, OR9/ORco; 6 for OR1/ORco, OR2/ORco, and ORco alone, the latter being tested in all experiments). Positive responses only were assessed by one-way ANOVA, followed by Bonferroni’s multiple comparison test, and significances of each heteromer’s response relative to ORco are depicted: ^∗∗P < 0.01; ^∗∗∗P < 0.001. (B) Left. Detection of ORco by western blot analysis in cells expressing ORco homomeric or ORx/ORco heteromeric complexes. Flag-tagged version of ORco was used in this experiment, and detection was performed by monoclonal antibody against the Flag epitope. Right. Detection of expression of ORs 1, 2, 9, and 53. Myc-tagged versions of ORx subunits were used in this experiment, and detection was performed by monoclonal antibody against the Myc epitope. The mock sample contains lysates from untransfected cells. (C) Comparison of the magnitudes of the heteromers’ responses to their specific ligands (SLs) (4MP for OR1, IN for OR2, 2EP for OR9, and LIN for OR53) relative to those obtained with the OA, both applied at a concentration of 100 μM. The inset in the bargraph for OR53 presents more clearly the low response of the cells that express the OR53/ORco heteromer to the partial agonist LIN (n = 3, 6, 4, and 2 for ORs 1, 2, 9, and 53, respectively).

FIGURE 2

Dose-dependent agonism of heteromeric receptor function by OAs. (A) Functional responses of cells expressing ORco alone or OR1/ORco, OR2/ORco, OR9/ORco, and OR53/ORco heteromers to increasing concentrations of OrcoRAM2. Results shown are from 3 to 4 independent experiments, each performed in triplicates; the responses for each receptor are normalized to the highest one (100%) obtained with 200 μM of the OA. (B) Comparison of the pEC₅₀ values for ORco homomer and the four studied heteromers (^∗∗P < 0.01; ^∗∗∗P < 0.001, for each heteromer relative to ORco).

FIGURE 3

OAs act as allosteric enhancers of heteromeric receptor function. (A) Cells expressing OR1/ORco, OR2/ORco, and OR9/ORco were challenged with 100 μM of their specific agonists (4MP, IN, and 2EP, respectively) in the absence (–) or presence (+) of a low concentration (10 μM) of OA (ORcoRAM2). OA-dependent potentiation was observed for all three tested heteromers (n = 5, 6, and 3, respectively). (B) Potentiation of responses to 100 μM of partial agonists (3MP for OR1/ORco, BA for OR2/ORco, 3MP and 4MP for OR9/ORco, and LIN for OR53/ORco) by 10 μM of the OA. Individual application of odorants alone is depicted by – signs, while the combined application of the OA and odorants is indicated by + signs (n = 2 for ORs 1, 2, and 9; n = 5 for OR53). (C) Lack of measurable potentiation by cells expressing OR1, OR2, OR9, and OR53 heteromers with ORco upon administration of odorants that do not normally activate these receptors and a low concentration (10 μM) of the OA. In all experiments, the SLs (4MP for OR1, IN for OR2, 2EP for OR9, and the partial agonist LIN for OR53) were used as positive controls providing maximal (100%) OA-potentiated responses (n = 3 for OR1 and OR2; 2 for OR9; and 3–5 for OR53). The cases of IN (for OR1) and OCT (for OR53), which appear as apparent exceptions to the behavior of odorants not recognized by the respective receptor heteromers are discussed in the main text (significance of OA+IN relative to OA and OA+OCT relative to OA for OR1 and OR53, respectively, is depicted: ^∗P < 0.05).

FIGURE 4

The OA is an allosteric modulator that enhances both the affinity and efficacy of odorant recognition by cognate olfactory receptors in vitro. (A) Upward and leftward shifts of the dose-response curves of OR1, OR2, OR9, and OR53 heteromers with ORco, to their SLs 4MP, IN, 2EP, and (partial agonist) LIN, respectively, in the presence of a low concentration of OA (10 μM). The upper and lower panels present the same results with different types of normalization: in the upper panels, the responses for each receptor were normalized relative to the highest responses (100%) obtained with each SL in the presence of 10 μM of the OA (SL+OA); while in the lower panels, the responses for each receptor were separately normalized to the maximum value (100%) obtained at the highest concentration of each SL alone or in the simultaneous presence of 10 μM of the OA (SL+OA) to make more evident the leftward shift. (B) Comparison of the pEC₅₀ values of the specific odorants alone or together with 10 μM of the OA (black and gray bars, respectively) for the four tested heteromers. The pEC₅₀ values (also listed in Table 1) are (mean ± SE): 4MP 5.927 ± 0.03852 and 4MP+OA 6.981 ± 0.2329 for OR1/ORco; IN 5.258 ± 0.03010 and IN+OA 7.280 ± 0.1809 for OR2/ORco; 2EP 4.110 ± 0.009062 and 2EP+OA 5.213 ± 0.08414 for OR9/ORco; and LIN 3.742 ± 0.04453 and LIN+OA 4.899 ± 0.1493 for OR53/ORco. (n = 2 for OR1/ORco and OR9/ORco, 3 for OR2/ORco and OR53/ORco).

FIGURE 5

Models of ORco homomeric and ORx/ORco heteromeric channels and schematic overview of allosteric modulation by OAs. Channel subunits are indicated by rectangular or oval shapes of different colors (brown for ORco and green for ORx), with conformational changes induced on subunits as a result of ORco or ORx-specific agonist binding being indicated by changes in shape (from rectangular to oval) and channel pore openings of various sizes induced by different ligands indicating magnitudes of ion permeability. ORco and ORx ligands are indicated by brown hexagons and green triangles or green ellipsoids, respectively, inside the corresponding receptor subunits. For all cases, conformational changes occurring in any given channel subunit have notable effects on the structure of its interacting subunit due to altered protein-protein interactions, and induce changes in magnitudes of ion permeability indicated by commensurate changes in channel pore sizes. (A) ORco channel: the different states of the homomeric ORco channel, which is presented here as a dimer, are shown with the unliganded, inactive state indicated by misaligned subunits and a closed pore, and the partially (+OA left) or fully active (+OA right) states indicated with the partly or fully changed shapes of one or both channel subunits, depending on agonist concentrations, alignment of channel subunits and commensurately increasing pore widths. Arrow 1 indicates the final effect of ligand gating on the channel pore. (B) ORx/ORco channel: the different states of the heteromeric channels (ORx/ORco channel) are shown, with dimers and tetramers illustrated for simplicity. Schematic in (I) shows heteromers containing an ORco-based channel pore, while (II) illustrates heteromers with the channel pore formed with contributions by both ORx and ORco subunits. For (I), a low concentration of the OA is indicated by a single binding site per receptor complex (Ib, left), while ORco and ORx-specific agonist concentrations at or higher than EC₅₀ are indicated by two bound agonists per receptor complex (Ib, right, and Ia). In the case of (II), the high ORco or ORx agonist states are indicated with the respective ligands being highlighted with black contour (IIb, right, and IIa), while the enhanced potency of the ORx agonist resulting from the simultaneous binding of the OA, observed under incomplete agonist occupancy conditions, is indicated by the change of the ligand’s shape from triangle to ellipsoid and the presence of the black contour (IIc). Arrows 2 and 3 indicate the effects of specific ORx ligands or the OAs, respectively, on channel pore permeabilities, while arrows 4 and 5 (with dotted lines) point to the elements on which the OA is hypothesized to act and thus affect the nature of the binding pocket (potency) and efficacy, respectively, of the SL. Note that for reasons of figure clarity, only one set of the predicted interactions is illustrated in the case of the high ligand concentration (Ic).