Potential role of transient receptor potential channel M5 in sensing putative pheromones in mouse olfactory sensory neurons

Abstract

Based on pharmacological studies of chemosensory transduction in transient receptor potential channel M5 (TRPM5) knockout mice it was hypothesized that this channel is involved in transduction for a subset of putative pheromones in mouse olfactory sensory neurons (OSNs). Yet, in the same study an electroolfactogram (EOG) in the mouse olfactory epithelium showed no significant difference in the responses to pheromones (and odors) between wild type and TRPM5 knockout mice. Here we show that the number of OSNs expressing TRPM5 is increased by unilateral naris occlusion. Importantly, EOG experiments show that mice lacking TRPM5 show a decreased response in the occluded epithelia to putative pheromones as opposed to wild type mice that show no change upon unilateral naris occlusion. This evidence indicates that under decreased olfactory sensory input TRPM5 plays a role in mediating putative pheromone transduction. Furthermore, we demonstrate that cyclic nucleotide gated channel A2 knockout (CNGA2-KO) mice that show substantially decreased or absent responses to odors and pheromones also have elevated levels of TRPM5 compared to wild type mice. Taken together, our evidence suggests that TRPM5 plays a role in mediating transduction for putative pheromones under conditions of reduced chemosensory input.

Figure 1. Naris occlusion upregulates GFP immunofluorescence in the occluded side of the OE of TRPM5-GFP mice.

A. This panel shows a representative image of a 14 µm coronal section of the OE taken from a naris-occluded animal. The right side in the image is ipsilateral to the occluded naris (Closed) (scale bar  = 500 µm). GFP immunofluorescence is green. B. Averaged GFP immunofluorescence (intensity ranges from 0 to 1, with a gain set so that OSN GFP immunofluorescence ranged from 0 to 0.2). The left side in the image is ipsilateral to the open side of naris (Open). Averaged GFP immunofluorescence intensity in the OE in the septum (i–iii) and the lateral regions (iv) were compared between open and closed sides. Averaged fluorescence intensity was significantly higher in the closed side of epithelium in all locations. (i; p = 0.02, ii; p = 0.006, iii; p = 0.002, iv; p = 0.007, p value FDR corrected 0.05, paired t-test, n = 4).

Figure 2. Naris occlusion upregulates the intensity of ciliary layer immunolabeling with a TRPM5 antibody.

Immunolabeling for TRPM5 (red) and GFP (green) is shown in the naris open and occluded sides under two different magnifications. A. Immunolabeling in the open and closed nostrils in endoturbinate II (bar is 50 µm, D is dorsal and M is medial). Notice that there is more intense labeling for both GFP and TRPM5 in the olfactory epithelium soma layer (OS) and cilia (cil) respectively in the closed nostril and that within each turbinate the labeling was inhomogeneous (some areas in the section show higher intensity than others). In addition, as expected from our earlier study there is intense labeling of microvillar cells that are not being studied here (mic). B. GFP immunolabeling intensity in the olfactory soma layer vs. TRPM5 immunolabeling intensity in the ciliary layer (red) measured in 20 degree sections around the endoturbinates in A (see the methods). Intensity in the image ranged from 0 to 1. Blue: closed, red: open. The straight line is a best fit for all the points. The correlation coefficient is 0.72 (different from zero, p<0.0001). The intensity for both GFP and TRPM5 immunolabeling in the open nostril are statistically significantly different compared to intensity in the closed nostril (t-test, p<0.000001). C. Higher magnification figures for the TRPM5 (red) and GFP (green) immunohistochemistry (bar is 20 µm, no microvillar cells are found in these images). The two images for the epithelium in the closed naris were from areas that displayed different intensity for labeling for the two antibodies.

Figure 3. Naris occlusion upregulates GFP presence in the OB.

A shows the representative coronal 18 µm section of the naris-occluded OB of a TRPM5 GFP animal (scale bar = 500 µm). The right OB in the image is ipsilateral to the occluded (Closed) naris, and the left OB is ipsilateral to the open naris. The section was immunostained with an antibody against GFP (green). As expected, the OB is smaller in the occluded side . (B) Histogram of the number of pixels as a function of the fluorescence intensity (0–4095) after subtracting intensity taken from the external plexiform layer (EPL) just underneath the glomerular layer. GFP immunofluorescence is higher in the occluded side (ii) compared to open side (i). (C) Cumulative histogram for fluorescence intensity after subtraction of EPL fluorescence levels for all four animals examined. Occluded OBs (red) express GFP at significantly higher level than open OB (blue) (t-test for mean intensity, p = 0.0286, n = 4). D shows a 2D color map of the glomeruli displaying GFP immunofluorescence as a function of percentage distance from the dorsal most point (*) around the glomerular layer in the olfactory bulb in A. Three representative OB sections were taken from the rostral, medial and caudal one-third and analyzed for GFP immunofluorescence intensity around the glomerular layer. (E) Mean fluorescence intensity around the glomerular layer of the occluded (red) and the open OB (blue). Thin lines represent the standard error of the mean (SEM). The intensity was averaged for caudal, middle and rostral images. Occluded side of OB (red) significantly differs from the open side of the OB (blue) (p<0.0001, N-Way ANOVA, n = 4). % peripheral distance was measured starting from the dorsal most point. d = dorsal, l = lateral, v = ventral, m = medial.

Figure 4. CNGA2 knockout OB shows wider distribution of glomeruli displaying GFP immunofluorescence that differs from the WT OB.

A shows a representative coronal section of the OB of CNGA2 knockout/TRPM5-GFP mouse (CNGA2 KO in the figure). Eighteen µm OB section was immunoreacted with antibody against GFP (green, scale bar = 500 µm). B shows the mean fluorescence intensity around the glomerular layer as a function of percent peripheral distance from the dorsal most point. Thin lines represent the SEM. The CNGA2 knockout OB (black) significantly differs from the open side of naris occluded OB (blue) of TRPM5-GFP animals (Open in the figure). (two way ANOVA for CNGA2 knockout/TRPM5-GFP vs. TRPM5-GFP open naris F _(1,49) = 189, p<0.0001, n = 4–6). % peripheral distance was measured starting from the dorsal most point. d = dorsal, l = lateral, v  =  ventral, m = medial.

Figure 5. Decreased odor-evoked OSNs activity leads to an elevated level of TRPM5 mRNA.

The mRNA levels of TRPM5, OMP, and Gα_olf were compared between: A. left (open) and right (closed, occluded) OE of the naris-occluded TRPM5-GFP animals. B. CNGA2 knockout/TRPM5-GFP and TRPM5-GFP control OE. mRNA levels are normalized to levels of 18 S rRNA. In the occluded OE, there was significantly higher expression of mRNA of TRPM5 (p = 0.024, n = 10), OMP (p = 0.008, n = 10), and Gα_olf (p = 0.033, n = 10, paired t-test, p value FDR corrected p = 0.05).Untreated control animals showed no significant difference between left and right OE (p>0.05, paired t-test, n = 5, data not shown). The mRNA expression level of TRPM5 was significantly higher in CNGA2 knockout/TRPM5-GFP compared to TRPM5-GFP (p = 0.016, unpaired t-test, p value FDR corrected p = 0.0167, n = 5). There was no significant difference between CNGA2 knockout/TRPM5-GFP and TRPM5-GFP in mRNA expression levels for OMP and Gα_olf (p>0.05, unpaired t-test FDR corrected, n = 5). Error bars are SEM.

Figure 6. Underwater EOG response to DMP and 2-heptanone was affected by unilateral naris occlusion in TRPM5(-) epithelia.

A; Example traces of EOG. The responses to DMP and isoamylacetate (ISO) are depicted in red and the response to IBMX is depicted in blue. The black bars above the traces indicate when the stimuli reach the epithelia and the duration of the stimulus application ( = 1 second). B; Putative pheromones (2,5-dimethylpyrazine (DMP) and 2-heptanone (2-HEP)), diluted urine (1∶200 or 1∶100) and general odorants (lilial and isoamylacetate) were tested for underwater EOG. What is reported is the ratio of the response to the pheromone/odorant divided by the response to IBMX [odor/IBMX]. In TRPM5(+) olfactory epithelia, naris occlusion had no effect on the ratio [odor/IBMX] (p value>0.05, paired t-test, FDR corrected). However, in TRPM5(-) olfactory epithelia, naris occlusion significantly reduced the ratio [odor/IBMX] to 2-heptanone or DMP (p<0.01, paired t-test, FDR corrected). The numbers in parentheses show the number of epithelia tested. The error bars are SEM.