Functional analysis of a mammalian odorant receptor subfamily

Abstract

Phylogenetic analysis groups mammalian odorant receptors into two broad classes and numerous subfamilies. These subfamilies are proposed to reflect functional organization. Testing this idea requires an assay allowing detailed functional characterization of odorant receptors. Here we show that a variety of Class I and Class II mouse odorant receptors can be functionally expressed in Xenopus laevis oocytes. Receptor constructs included the N-terminal 20 residues of human rhodopsin and were co-expressed with Galphaolf and the cystic fibrosis transmembrane regulator to allow electrophysiological measurement of receptor responses. For most mouse odorant receptors tested, these conditions were sufficient for functional expression. Co-expression of accessory proteins was required to allow functional surface expression of some mouse odorant receptors. We used this assay to examine the receptive ranges of all members of the mouse odorant receptor 42 (MOR42) subfamily. MOR42-1 responded to dicarboxylic acids, preferring a 10-12 carbon chain length. MOR42-2 responded to monocarboxylic acids (7-10 carbons). MOR42-3 responded to dicarboxylic acids (8-10 carbons) and monocarboxylic acids (10-12 carbons). Thus, the receptive range of each receptor was unique. However, overlap between the individual receptive ranges suggests that the members of this subfamily form one contiguous subfamily receptive range, suggesting that odorant receptor subfamilies do constitute functional units.

Figure 1. Functional surface expression of MOR42-3 in Xenopus oocytes

A) A TEVC recording of an oocyte injected with cRNA encoding β2-AR, Gα_olf and CFTR responding to 100 nM isoproterenol (Iso) and to 1 mM IBMX. Scale: 0.5 µA, 100 sec. B) A TEVC recording of an oocyte injected with cRNA encoding MOR42-3, Gα_olf and CFTR responding to 100µM nonanedioic acid (NDA) and to 1 mM IBMX. Scale: 0.5 µA, 100 sec. C) A TEVC recording of an oocyte injected with cRNA encoding Gα_olf and CFTR (but no OR cRNA) failing to respond to 100µM nonanedioic acid, but responding to 1 mM IBMX. Scale: 0.5 µA, 100 sec. D) A cryosection of an oocyte injected with cRNA encoding MOR42-3, Gα_olf and CFTR, and labeled with the 4D2 anti-rhodopsin antibody and a Cy3-conjugated secondary antibody, viewed in transmitted light. E) The same section as in panel D, viewed at 570 nm. F) A cryosection of an oocyte injected with cRNA encoding Gα_olf and CFTR (but no OR cRNA), and labeled with the 4D2 anti-rhodopsin antibody and a Cy3-conjugated secondary antibody, viewed in transmitted light. G) The same section as in panel E, viewed at 570 nm. In panels A – C, all applications were 15 sec in duration. In panels D – G, the cytoplasm is indicated by “c” and the scale bar is 5 µm.

Figure 2. The rhodopsin tag is essential for functional expression of MOR42-3

A) TEVC recordings from an oocyte injected with cRNA encoding untagged MOR42-3, Gα_olf and CFTR failing to respond to 100µM nonanedioic acid, but responding to 1 mM IBMX. Scale: 0.5 µA, 100 sec. B) TEVC recordings from an oocyte injected with cRNA encoding tagged MOR42-3, Gα_olf and CFTR responding to 100µM nonanedioic acid and to 1 mM IBMX. Scale: 0.5 µA, 100 sec. C) Functional responses from multiple oocytes are shown as the ratio of NDA responses to IBMX responses (mean ± SEM, n = 6–11). The NDA responses of oocytes injected with cRNA encoding tagged MOR42-3 are significantly greater than those of oocytes injected with cRNA encoding untagged MOR42-3 or oocytes not injected with receptor cRNA (p<0.05). In panels A and B, all applications were 15 sec in duration.

Figure 3. Functional responses of MOR42-3 to nonanedioic acid

A) An oocyte expressing MOR42-3, Gα_olf and CFTR is challenged with repeated applications of 30µM nonanedioic acid. Scale: 0.2 µA, 5 min. B) An oocyte expressing MOR42-3, Gα_olf and CFTR is challenged with a range of nonanedioic acid concentrations. Scale: 0.2 µA, 10 min. C) Dose-response curve for MOR42-3 responding to nonanedioic acid (mean ± SEM, n = 8). Data is fit to a Hill equation (see Experimental Procedures). In panels A and B, all applications were 15 sec in duration.

Figure 4. Functional expression of a variety of MORs in Xenopus oocytes

A) Current recordings from oocytes injected with cRNA encoding Gα_olf, CFTR and one of a variety of different MORs (or no receptor). Nonanedioic acid (NDA), nonanoic acid (NA) and 2-coumaranone (2-C) are applied at 100 µM. Scale: 0.5 µA, 200 sec. B) Current recordings from oocytes injected with cRNA encoding Gα_olf and CFTR, and either MOR174-9 (mOR-EG) or no receptor. Eugenol (EG) and methyl isoeugenol (MIEG) are applied at 100 µM. Scale: 0.25 µA, 200 sec. All applications were 15 sec in duration.

Figure 5. Accessory proteins are required for functional expression of some ORs

A) A cryosection of an oocyte injected with cRNA encoding MOR23-1, Gα_olf, CFTR, RTP1, RTP2 and REEP1, and labeled with the 4D2 anti-rhodopsin antibody and a Cy3-conjugated secondary antibody, viewed in transmitted light. B) The same section as in panel A, viewed at 570 nm. C) A cryosection of an oocyte injected with cRNA encoding MOR23-1, Gα_olf and CFTR, and labeled with the 4D2 anti-rhodopsin antibody and a Cy3-conjugated secondary antibody, viewed in transmitted light. D) The same section as in panel C, viewed at 570 nm. E) A cryosection of an oocyte injected with cRNA encoding Gα_olf, CFTR, RTP1, RTP2 and REEP1 (but no OR cRNA), and labeled with the 4D2 anti-rhodopsin antibody and a Cy3-conjugated secondary antibody, viewed in transmitted light. F) The same section as in panel E, viewed at 570 nm. In panels A–F, the cytoplasm is indicated by “c” and the scale bar is 5 µm. G) Upper traces, an oocyte injected with cRNA encoding MOR23-1, Gα_olf and CFTR fails to respond to 100µM Octanoic acid (OA). Lower traces, functional expression is achieved by coexpressing the accessory proteins RTP1, RTP2 and REEP1 (RTPs). Scale: 0.5 µA, 200 sec. H) Responses of MOR23-1 injected oocytes to 100µM OA, in the presence and absence of RTPs, are plotted as the ratio of OR response to IBMX response (mean ± SEM, n=7, p<0.001). I) Top traces, an oocyte injected with cRNA encoding MOR258-5, Gα_olf and CFTR fails to respond to 2-Coumaranone (2-C). Lower traces, functional expression is achieved by coexpressing the accessory proteins RTP1 and RTP2. Scale: 0.5 µA, 200 sec. J) Responses of MOR258-5 injected oocytes to 100µM 2-C, in the presence and absence of RTPs, are plotted as the ratio of OR response to IBMX response (mean ± SEM, n=5, p<0.05). In panels G and I, all applications were 15 sec in duration.

Figure 6. Ligand specificities of the MOR42 subfamily

A) Oocytes expressing Gα_olf and CFTR, and either MOR42-1 (left), MOR42-2 (middle), or MOR42-3 (right) are challenged with 30µM of dicarboxylic acids of varying carbon length. The MOR42-2 expressing oocyte responded to 30 µM octanoic acid (not shown). Scale: 0.25 µA, 300 sec. B) Oocytes expressing Gα_olf and CFTR, and either MOR42-1 (left), MOR42-2 (middle), or MOR42-3 (right) are challenged with 30µM of monocarboxylic acids of varying carbon length. The MOR42-1 expressing oocyte responded to 30 µM decanedioic acid and the MOR42-3 expressing oocyte responded to 30 µM nonanedioic acid (not shown). Scale: 0.25 µA, 300 sec. C) Responses of MOR42-1, MOR42-2 and MOR42-3 to dicarboxylic acids, monocarboxylic acids, aldehydes and alcohols of varying carbon chain length. Responses were normalized to the response of the same oocyte to 30µM decanedioic acid (MOR42-1), octanoic acid (MOR42-2) or nonanedioic acid (MOR42-3) and are presented as the mean of 6–8 separate oocytes. Standard errors are provided in Table 1. Unfilled squares on the floor of each graph represent compounds that were tested but yielded no response. These non-functional compounds were tested with oocytes that were expressing MOR42-1, MOR42-2 or MOR42-3, as confirmed by robust responses to 30 µM decanedioic acid, octanoic acid or nonanedioic acid, respectively. In panels A and B, all applications were 15 sec in duration.

Figure 7. Dose-response analysis of MOR42-1 and MOR42-3

A) Responses of MOR42-1 to a range of dicarboxylic acid concentrations. The dose-response curve for decanedioic acid (C10) was generated as described in Experimental Procedures. Responses to nonanedioic acid (C9), undecanedioic acid (C11) and dodecanedioic acid (C12) were normalized to the maximal response to decanedioic acid (see Experimental Procedures). Data are the mean ± SEM (n = 3–11). B) Responses of MOR42-3 to a range of dicarboxylic acid concentrations. The dose-response curve for nonanedioic acid (C9) is from Figure 3C. Responses to octanedioic acid (C8), decanedioic acid (C10), undecanedioic acid (C11) and dodecanedioic acid (C12) are normalized to the maximal response to nonanedioic acid (see Experimental Procedures). Data are the mean ± SEM (n = 3–12).