Overlapping but distinct topology for zebrafish V2R-like olfactory receptors reminiscent of odorant receptor spatial expression zones

Abstract

Background: The sense of smell is unrivaled in terms of molecular complexity of its input channels. Even zebrafish, a model vertebrate system in many research fields including olfaction, possesses several hundred different olfactory receptor genes, organized in four different gene families. For one of these families, the initially discovered odorant receptors proper, segregation of expression into distinct spatial subdomains within a common sensory surface has been observed both in teleost fish and in mammals. However, for the remaining three families, little to nothing was known about their spatial coding logic. Here we wished to investigate, whether the principle of spatial segregation observed for odorant receptors extends to another olfactory receptor family, the V2R-related OlfC genes. Furthermore we thought to examine, how expression of OlfC genes is integrated into expression zones of odorant receptor genes, which in fish share a single sensory surface with OlfC genes.

Results: To select representative genes, we performed a comprehensive phylogenetic study of the zebrafish OlfC family, which identified a novel OlfC gene, reduced the number of pseudogenes to 1, and brought the total family size to 60 intact OlfC receptors. We analyzed the spatial pattern of OlfC-expressing cells for seven representative receptors in three dimensions (height within the epithelial layer, horizontal distance from the center of the olfactory organ, and height within the olfactory organ). We report non-random distributions of labeled neurons for all OlfC genes analysed. Distributions for sparsely expressed OlfC genes are significantly different from each other in nearly all cases, broad overlap notwithstanding. For two of the three coordinates analyzed, OlfC expression zones are intercalated with those of odorant receptor zones, whereas in the third dimension some segregation is observed.

Conclusion: Our results show that V2R-related OlfC genes follow the same spatial logic of expression as odorant receptors and their expression zones intermingle with those of odorant receptor genes. Thus, distinctly different expression zones for individual receptor genes constitute a general feature shared by teleost and tetrapod V2R/OlfC and odorant receptor families alike.

Keywords: In situ hybridization; Microvillous neurons; Olfactory receptors; Spatial distribution; V2Rs; Zebrafish.

Fig. 1

Selection of seven representative genes from a phylogenetic tree of the zebrafish OlfC family. A phylogenetic tree of 60 full length OlfC gene sequences was constructed using a maximum likelihood (ML) algorithm (see Methods for details). Bootstrap support in percent is indicated at relevant nodes. As in previous analyses, the OlfC family appears polyphyletic, with the calcium sensor gene CaSR intercalating between OlfCa1, OlfCb1, and the remainder of the OlfC family. Mouse and zebrafish T1R taste receptors were used as outgroup. Single asterisk, gene was predicted in [23] as pseudogene (OlfCb1, e1, q10) or fragment (OlfCf1, m2), but is intact and full length in the current prediction; double asterisk, gene is lost in GRCz10; triple asterisk, novel gene. OlfC genes highlighted with yellow were selected for expression analysis. To the right of the tree the core cell distribution parameters for the genes analysed are shown as bar graphs. Light blue, median radial position; dark blue, median height position, all values normalized to the respective maximal values. No correlation is apparent between position in the phylogenetic tree and median radius or height

Fig. 2

v2r-related OlfC genes generally are expressed in small subsets of scattered olfactory sensory neurons. Horizontal sections of adult zebrafish olfactory epithelium were hybridized with probes for OlfCg1, OlfCn1, OlfCq1, OlfCj1, OlfCu1, and OlfCt1. Column a shows representative complete sections labeled with the respective probes. The scale bars correspond to 40 μm. Column b, higher magnifications from different sections. The hybridization signal was observed in sparse cells within the sensory region of the olfactory epithelium, as expected; arrowheads point to some labeled neurons. The scale bars correspond to 20 μm. c Bar graphs representing number of labeled cells per section for each OlfC gene (mean +/− SEM, n = 78–265 sections/gene)

Fig. 3

Quantitative assessment of laminar height distributions of v2r-related OlfC-expressing neurons. Laminar height of OlfC-expressing cells was quantified for seven OlfC genes, including OlfCc1. Complete series of sections from three to five olfactory epithelia were evaluated for each OlfC gene. Height within the lamina was normalized to maximal laminar thickness. a The resulting distributions of relative laminar height (from 0, most basal to 1, most apical, i.e. bordering to the lumen) are shown binned (histogram, top row) and unbinned (empirical cumulative distribution function, ECDF, bottom row). The color code for the OlfC genes is the same as for later figures to facilitate comparisons between different positional parameters. Light grey curves in the ECDF plot for each of the tested OlfC genes represent the distribution for individual olfactory organs. Due to the smaller number of cells, scatter is increased. b Overlay of the seven distributions shown individually in panel a), both as histogram (left panel) and ECDF (right panel)

Fig. 4

Quantitative assessment of the radial distribution of v2r-related OlfC-expressing neurons. The distribution of radial positions of OlfC-expressing cells was quantified for seven OlfC genes, including OlfCc1, using the same set of sections, for which laminar height was determined, except the very first sections, where the sensory surface does not yet extend toward the median raphe. Radial position within the section was normalized to maximal radius, i.e. length of the lamella containing the respective labeled cell. In other words, radial distance was measured from the apex of the lamellar ‘curve’, i.e. closest to the median raphe, to the cell soma center, and normalized to the distance between this apex position (most central) and the border of the epithelial section (most peripheral). a The resulting distributions of relative radius (from 0, innermost to 1, outermost) are shown binned (histogram, top row) and unbinned (ECDF, bottom row). The color code for the OlfC genes is the same as in Fig. 3 to facilitate comparisons between different positional parameters. Light grey curves in the ECDF plot for each of the tested OlfC genes represent the distribution for individual olfactory organs. Due to the smaller number of cells, scatter is increased. b Overlay of the seven distributions shown individually in panel a), both as histogram (left panel) and ECDF (right panel)

Fig. 5

Quantitative assessment of distribution of v2r-related OlfC expressing neurons along the vertical z-axis (height within the organ). Height within the olfactory organ was quantified as section number in a series of horizontal sections, and normalized to the total number of sections containing sensory epithelium, using the same set of cells, for which laminar height was determined. Relative height within the organ ranges from 0 (top section, near to the opening of the bowl-shaped olfactory organ) to 1 (bottommost section). a The resulting distributions are shown binned (histogram, top row) and unbinned (ECDF, bottom row). The color code for the OlfC genes is the same as in Figs. 3 and 4 to facilitate comparisons between different positional parameters. Light grey curves in the ECDF plot for each of the tested OlfC genes represent the distribution for individual olfactory organs. Due to the smaller number of cells, scatter is increased. b Overlay of the distributions shown individually in panel a), both as histogram (left panel) and ECDF (right panel)

Fig. 6

Simultaneous labeling of two OlfC genes confirms distinctly different distributions. a Representative micrograph of two-color in situ hybridization, depicting sparse expression of OlfCg1 (in magenta) and OlfCu1 (in green), within a single horizontal section of the olfactory epithelium. b-d Quantitative assessment of the distributions of the labeled cells for laminar height (b), radial distance (c) and along the vertical z-axis (d). The resulting distributions of relative laminar height (from 0, most basal to 1, most apical, i.e. bordering to the lumen), relative radius (from 0, innermost to 1, outermost) and relative height within the organ (from 0, top section to 1, bottommost section) are shown unbinned as empirical cumulative distribution function (ECDF). Color of ECDF graphs corresponds to the color employed in panel a. KS-test with a p-value cutoff of < 0.01 was used to evaluate the significance of differences between the distributions, if any. Significance is indicated by asterisks, **, p ≤ 0.01; ****, p ≤ 0.0001

Fig. 7

Comparison of spatial distribution parameters between OlfC genes. Schematic representation of spatial distributions for different OlfC genes and two marker genes, omp and trpc2, by ellipses ranging from the 1st to the 3rd quartile value for x and y parameter. Color code for OlfC genes as before. a Radius (x axis) is depicted vs. laminar height (y axis). Note the distributions for all OlfC genes and trpc2, the marker for microvillous neurons, center on rather apical positions within the lamina (large values for laminar height), clearly segregated from the much more basal positions for omp-positive, ciliated neurons. No correlation is apparent between radius and laminar height values. b Radius (x axis) is depicted vs. height within the organ (y axis). No correlation is apparent between radius and height within organ values