Skip to main content
. 2018 Dec 14;7:e41050. doi: 10.7554/eLife.41050

Figure 1. Combinatorial expression of δ-Pcdhs in mouse olfactory sensory neurons (OSNs).

(A) Representative image of a double label RNA in situ hybridization with Pcdh19 (red) and Pcdh10 (green) in E17.5 olfactory epithelium. Both probes are co-expressed in a subset of neurons (arrowhead). Scale bar, 50 μm. (B) Heat map showing the percentage of co-expression among δ-Pcdhs and OSNs expressing one of five different odorant receptors. The color intensity indicates the percent of co-expression for any one δ-Pcdh with a given receptor. (C) Representative confocal images of Olfr124 positive OSNs co-expressed with Pcdh1 (top row) but not Pcdh10 (bottom row). Arrowhead indicates location of Olfr124 positive OSN. Scale bar, 50 μm. (D) Heat map of log2 transformed NanoString counts. (E) Constrained gamma-normal mixture modeling analysis shows individual, randomly selected OSNs express zero to seven δ-Pcdhs. (F) qRT-PCR of randomly selected single OSNs shows a mosaic pattern of δ-Pcdh expression similar to the NanoString data.

Figure 1 —source data 1. NanoString codeset and primer sequences.
elife-41050-fig1-data1.xlsx (14.6KB, xlsx)
DOI: 10.7554/eLife.41050.005
Figure 1 —source data 2. NanoString nCounter data.
elife-41050-fig1-data2.xlsx (19.8KB, xlsx)
DOI: 10.7554/eLife.41050.006

Figure 1.

Figure 1—figure supplement 1. Expression of δ-Pcdhs in OSNs.

Figure 1—figure supplement 1.

(A) Single color RNA in situ hybridization of Ncam1 (a marker of OSNs), Notch2 (a marker of non-neuronal sustentacular cells), and δ-Pcdhs in P7 olfactory epithelium. Note punctate expression of δ-Pcdhs. Scale bar, 100 μm. Pcdh11x and Pcdh18 could not be detected. (B–G) Single color RNA in situ hybridization of Pcdh1, Pcdh9, and Pcdh17 in E17.5 (B,D,F) and P7 (C,E,G) olfactory epithelia. Arrowheads indicate areas of enriched regional expression. Scale bar, 400 μm for (B,D,F) and 500 μm (for C,E,G). (H) Confocal analysis of a round robin double label RNA in situ hybridization series from E17.5 olfactory epithelia. Values indicate percent overlap in OSNs for any given pair. Pcdh11x and Pcdh18 could not be detected with this approach. (I) Constrained gamma-normal mixture modeling was used to determine if expression of a given gene was ‘on’ or ‘off’ within a given cell. Each line represents a density plot from the model for a single cell. Blue curves represent the lowly expressed component (e.g. ‘off’), which was allowed to vary in relative proportion but with constant mean and variance parameters. Red curves represent the highly expressed component (e.g. ‘on’) as a normal distribution with variable mean and variance parameters. The dashed curve represents the sample density of all cells across all genes. (J) Ribbon plot comparing percentage of OSNs expressing a given δ-Pcdh as determined by NanoString (red line) and quantification of RNA in situ hybridization signal (blue line). Similar trends were observed for both methods, suggesting enzymatic dissociation during OSN isolation did not greatly alter δ-Pcdh expression. Shaded regions represent 95% CI. (K) Ribbon plot comparing δ-Pcdh expression in single OSNs as detected by NanoString and three different single OSN RNA-seq studies. Data from RNA-seq studies were re-analyzed using the constrained gamma-normal mixture modeling approach. Cells were first filtered based on positive gene expression of Ncam1 to parallel the selection of Ncam1 positive OSNs used in this study. The three single OSN RNA-seq studies follow similar distributions, with the majority of OSNs expressing zero or one δ-Pcdh. In contrast, the NanoString dataset detects more δ-Pcdhs per cell. Ribbons represent standard deviation following repeated bootstrapping of samples. (L) Mean number of δ-Pcdhs per OSN from single cell RNA-seq datasets and NanoString. Numbers above bars represent the number of Ncam1 positive cells in each study.