Functional identification and reconstitution of an odorant receptor in single olfactory neurons

Abstract

The olfactory system is remarkable in its capacity to discriminate a wide range of odorants through a series of transduction events initiated in olfactory receptor neurons. Each olfactory neuron is expected to express only a single odorant receptor gene that belongs to the G protein coupled receptor family. The ligand-receptor interaction, however, has not been clearly characterized. This study demonstrates the functional identification of olfactory receptor(s) for specific odorant(s) from single olfactory neurons by a combination of Ca2+-imaging and reverse transcription-coupled PCR analysis. First, a candidate odorant receptor was cloned from a single tissue-printed olfactory neuron that displayed odorant-induced Ca2+ increase. Next, recombinant adenovirus-mediated expression of the isolated receptor gene was established in the olfactory epithelium by using green fluorescent protein as a marker. The infected neurons elicited external Ca2+ entry when exposed to the odorant that originally was used to identify the receptor gene. Experiments performed to determine ligand specificity revealed that the odorant receptor recognized specific structural motifs within odorant molecules. The odorant receptor-mediated signal transduction appears to be reconstituted by this two-step approach: the receptor screening for given odorant(s) from single neurons and the functional expression of the receptor via recombinant adenovirus. The present approach should enable us to examine not only ligand specificity of an odorant receptor but also receptor specificity and diversity for a particular odorant of interest.

Figure 1

Isolation of olfactory receptor gene from tissue printed olfactory neurons that respond to odorants using Ca²⁺ imaging and single cell RT-PCR. (a) Schematic drawing of Ca²⁺ imaging set up to record odorant-induced Ca²⁺ increase in tissue-printed olfactory neurons shown by circle dots. Odorant solutions were applied sequentially to the recording chamber with a peristaltic pump at the flow rate of 1.5 ml/min. (b) The response of a single olfactory neuron to lyral (shown by arrow 2 in lower panel) in reflected changes in fura-2 fluorescence intensity ratios (340/380 nm). The odorants were applied to tissue-printed cells for 4 sec during times indicated by the bars. The cells were washed continuously between odorant applications. Names of odorants are abbreviated as follows: h4, 1-butanol; lCA, l-carvone; LY, lyral; EG, eugenol; GE, geraniol; EV, ethyl vanillin; LI, lilial; CR, cresol; PY, pyridine; dCA, d-carvone; Bz, benzene (each 100 μM). HK stands for high KCl buffer (145.6 mM KCl/2 mM CaCl₂/2 mM MgCl₂/9.4 mM glucose/5 mM Hepes, pH 7.4). Lower panels show pseudocolored images of Ca²⁺ measurements taken from tissue-printed cells at three representative time-points. The Ca²⁺ level of the lyral-responding cell (arrow 2) was increased 17-fold on lyral application from 246 nM (a basal level) to 4.2 μM. Arrow 1 depicts a dead cell. The graded color bar is a calibration of the imaging system. The white color indicates the highest Ca²⁺ level while the blue color represents the basal level and outlines the shape of each cell. (c) Single cell RT-PCR products separated on agarose gel. Minus and plus signs indicate RT-PCR experiments with or without reverse transcriptase using the same RNA preparations. (d) The putative amino acid sequence of the MOR23 coding region. The primers used for two-round PCR are shown by arrows. TM, putative transmembrane domains.

Figure 2

Localization of the endogenous MOR23 gene and expression of the recombinant adenovirus-mediated transgene in olfactory epithelium. (a) In situ hybridization of the MOR23 transcript in coronal sections of the olfactory epithelium. The antisense RNA for the MOR23 coding region was used as a probe. The presence of the transcripts was indicated by dispersed focal staining shown by arrows. The expression of MOR23 is totally restricted to the expression zone 1 at every section. [Bars = 0.5 mm (left), 250 μm (middle), and 100 μm (right).] (b) Schematic drawings of the expression units of the recombinant adenovirus vectors used in this study. Regions corresponding to the mature transcripts are indicated by boxes. The bicistronic adenovirus expression unit (Upper) contains MOR23 gene, IRES sequence, and GFP gene followed by the rabbit β-globin polyadenylation signal (GpA). The monocistronic expression unit (Lower) contains the GFP gene followed by GpA. The IRES allows internal translation initiation, resulting in expression of both MOR23 and GFP. These individual expression units are located at the E1A-deleted region of the adenovirus type-5 genome. (Bar = 0.4 kilobases). (c) Coronal sections of the nasal tissues infected with the monocistronic (Left and Center) and the bicistronic (Right) recombinant adenovirus vectors. (Left) Under light microscopy. (Center and Right) Under fluorescent illumination. There are no significant differences in the infected regions throughout the olfactory epithelium between two constructs. (Bars = 0.5 mm.) (d) Higher magnification of MOR23 virus-infected cells. The fluorescent cells are predicted to be olfactory neurons. (Bar = 100 μm.) (e) Tissue-printing of the GFP fluorescent olfactory epithelium result in identification of MOR23-infected individual single olfactory neurons under fluorescence illumination. (Bar = 100 μm.)

Figure 3

Characterization of odorant responses in MOR23-adenovirus infected cells. (a) Lyral-mediated Ca²⁺ increase in MOR23 adenovirus-infected cells shown by changes in fura-2 fluorescence intensity ratios (340/380 nm) (1 mM lyral application for 10 sec at the indicated bars). (Upper) Bicistronic MOR23-IRES-GFP. (Lower) Monocistronic GFP. High KCl buffer (HK) and forskolin (FK) (10⁻⁵ M) were applied at indicated bars (5 sec) to assess cell viability and olfactory neuronal properties. (b) Ca²⁺ increase in MOR23-adenovirus infected cells was induced by lyral-containing odorant mixture (mix + LY) but not by the mixture without the lyral (mix − LY). The mixture contains the same odorants used in Fig. 1b (10 or 100 μM). (c) Dose-dependent Ca²⁺ increase as a percentage of the response at 10 mM lyral. Data ±SE (n = 16 for 1 mM, n = 5 for others).

Figure 4

Ligand specificity of MOR23. Ca²⁺ increase in MOR23-expressing cells was induced by some odorants structurally similar to lyral at higher concentrations. MA, myrac aldehyde; DM, dihydromyrcerol; HC, hydroxycitronellol; HCA, hydroxycitronellol dimethyl acetal; TM, tetrahydromyrcenol. MOR23 responds to HC and HC dimethyl acetal (HCA) at 10 mM but not at 100 μM. Common structural motifs are shown by boxes. LY is a mixture of two isomers: 4-(4-hydroxy-4-methylpentyl)-3-cyclohexenecarbaldehyde and 3-(4-hydroxy-4-methylpentyl)-3-cyclohexenecarbaldehyde. MA is also a mixture of two isomers, which are the dehydrated forms of LY (hydroxy MA). Only 4-substituted isomers of LY and MA are depicted in the figure.

Figure 5

Lyral-MOR23 interaction leads to external Ca²⁺ influx. Lyral-induced Ca²⁺ increase is greatly reduced in the absence of external Ca²⁺. The buffer was exchanged to Ca²⁺-free Ringer’s solution for the indicated period, during which lyral in Ca²⁺-free buffer was applied. The response was rescued by changing the buffer to Ca²⁺-containing Ringer’s solution.