Heterotrimeric G protein subunit Gγ13 is critical to olfaction

Abstract

The activation of G-protein-coupled olfactory receptors on the olfactory sensory neurons (OSNs) triggers a signaling cascade, which is mediated by a heterotrimeric G-protein consisting of α, β, and γ subunits. Although its α subunit, Gαolf, has been identified and well characterized, the identities of its β and γ subunits and their function in olfactory signal transduction, however, have not been well established yet. We, and others, have found the expression of Gγ13 in the olfactory epithelium, particularly in the cilia of the OSNs. In this study, we generated a conditional gene knock-out mouse line to specifically nullify Gγ13 expression in the olfactory marker protein-expressing OSNs. Immunohistochemical and Western blot results showed that Gγ13 subunit was indeed eliminated in the mutant mice's olfactory epithelium. Intriguingly, Gαolf, β1 subunits, Ric-8B and CEP290 proteins, were also absent in the epithelium whereas the presence of the effector enzyme adenylyl cyclase III remained largely unaltered. Electro-olfactogram studies showed that the mutant animals had greatly reduced responses to a battery of odorants including three presumable pheromones. Behavioral tests indicated that the mutant mice had a remarkably reduced ability to perform an odor-guided search task although their motivation and agility seemed normal. Our results indicate that Gαolf exclusively forms a functional heterotrimeric G-protein with Gβ1 and Gγ13 in OSNs, mediating olfactory signal transduction. The identification of the olfactory G-protein's βγ moiety has provided a novel approach to understanding the feedback regulation of olfactory signal transduction pathways as well as the control of subcellular structures of OSNs.

Figure 1.

Targeted disruption of the Gng13 gene. The wild-type (WT) allele of the gene consists of three exons: E1, E2, and E3, with the start codon ATG in E2 and the stop codon TGA in E3. The entire Gng13 gene was translocated into the targeting vector by homologous recombination in the Escherichia coli cells through two flanking homologous sequences: H-A and H-B. The same recombination approach was used to insert a loxP site and a Frt-Neo-Frt-loxP cassette into intron 1 and 3′-end sequence downstream of exon 3, respectively, which is followed by the thymidine kinase gene (TK) on the targeting vector for embryonic stem (ES) cell counter-selection. The recombination at the 5′- and 3′- homologous arms (5′-, 3′-Hom) in the mouse ES cells resulted in the generation of the knock-in allele, which was identified by PCR screening with the two primers P1 and P2 located upstream of the first loxP site and downstream of 3′-Hom, respectively. DNA templates prepared from the ES cells containing both the recombinant and WT alleles (ES-DNA PCR inset: R/W lane) produced two PCR products of 6.1 and 4.2 kb, whereas that containing only WT allele generated a single amplified product of 4.2 kb (W/W lane). In the olfactory sensory neurons that express Cre recombinases, the sequence between two loxP sites was excised, deleting all coding exons of the Gng13 gene, and PCR with P1 and P2 primers and genomic DNA from this olfactory epithelium generated two PCR products of 0.8 kb from the excised allele and 2.1 kb from WT allele (OE-DNA PCR inset: R/W lane), whereas genomic DNA from WT olfactory epithelium generated a single amplified product of 2.1 kb (W/W lane).

Figure 2.

Immunostaining of WT and KO olfactory epithelia with antibodies against olfactory signaling components. Eight-month-old male mice of WT and KO were used for this experiment. Gγ13, Gαolf, Gβ1, ACIII, acetylated tubulin (Ac-Tubulin), and OMP immunostaining on WT (A–G) and KO (H–N) olfactory epithelium. Scale bar, A, H, 100 μm; B–G, I–N, 25 μm.

Figure 3.

Western blot and real-time PCR analyses of the effect of Gng13 KO on expression of olfactory proteins. A, Western blots were prepared from lysates of the olfactory epithelia isolated from WT and KO mice, which were incubated with the primary antibodies against Gγ13, Gβ1, Ric-8B(RIC8B), CEP290, ACIII, OMP, and β-Actin, followed by the horseradish peroxidase-conjugated secondary antibodies. B, Plot of signal intensities from Western blots in A. The relative levels of other proteins were normalized against those of β-actin in each sample and then the intensities of other proteins in WT were set to 100 to determine the relative intensities of the corresponding proteins in KO sample. C, Relative transcript levels. As with B, β-actin transcript levels were used as an internal control to normalize other genes' transcript levels in different samples, and WT levels for other genes were set to 100. Error bars represent SEM. ***p > 0.001; *p > 0.05.

Figure 4.

Immunostaining of olfactory bulb sections. While OMP was detected in axon bundles and glomeruli of both WT and KO tissues (A, C), Gγ13 immunoreactivity was found only in the WT section (B) and not in the KO section (D). Scale bar, 80 μm.

Figure 5.

Representative EOG responses from WT and KO mice. Gng13-KO epithelia (gray traces) had very small EOG responses to odorants compared with WT epithelia (black traces). All odorants/pheromones were applied for 100 ms at 0.1 m in solution. TEA elicited a biphasic response in WT epithelia and a positive response in KO epithelia.

Figure 6.

EOG amplitudes to odorants were diminished in Gng13-KO epithelia. A, Responses to a set of 15 odorants from WT (n = 10) and KO (n = 14) mice (p < 0.0001 for all odorants tested). B, Response to 0.1 m heptanal from endoturbinates 2, 2b, 3, and 4 from WT (n = 5) and KO (n = 7) mice. The recordings were made from 12 points on each turbinate (right, graphical representation on the epithelium). WT and KO data are significantly different (p < 0.05), except for positions 4 and 10 on T4. Error bars represent SE.

Figure 7.

Gng13-KO mice showed significantly impaired food-seeking capabilities. KO mice took significantly longer to locate a 1 g chow pellet, scented with 5 μl of 10% peanut butter, buried under the bedding compared with heterozygous (HT) or WT littermates. However, all three types of mice located food placed on top of the bedding in similarly short times. Values are means ±SE. ***p < 0.001.