Identification and analysis of odorant receptors expressed in the two main olfactory organs, antennae and palps, of Schistocerca americana

Abstract

Locusts depend upon their sense of smell and provide useful models for understanding olfaction. Extending this understanding requires knowledge of the molecular and structural organization of the olfactory system. Odor sensing begins with olfactory receptor neurons (ORNs), which express odorant receptors (ORs). In insects, ORNs are housed, in varying numbers, in olfactory sensilla. Because the organization of ORs within sensilla affects their function, it is essential to identify the ORs they contain. Here, using RNA sequencing, we identified 179 putative ORs in the transcriptomes of the two main olfactory organs, antenna and palp, of the locust Schistocerca americana. Quantitative expression analysis showed most putative ORs (140) are expressed in antennae while only 31 are in the palps. Further, our analysis identified one OR detected only in the palps and seven ORs that are expressed differentially by sex. An in situ analysis of OR expression suggested ORs are organized in non-random combinations within antennal sensilla. A phylogenetic comparison of OR predicted protein sequences revealed homologous relationships among two other Acrididae species. Our results provide a foundation for understanding the organization of the first stage of the olfactory system in S. americana, a well-studied model for olfactory processing.

Figure 1

Statistical characterization of SameORs identified in the unified transcriptome assembly. (a) Frequency distribution of CDS length (bp). (b) Relation between CDS length and number of TMDs predicted from the deduced amino acid sequences of the 179 putative ORs. Putative OR sequences were divided into database A (dark blue) and database B (light blue) for further analysis. (c) Length of full and partial CDSs. (d) Blastx results of putative SameOR CDSs against a recently posted S. americana genome assembly. Shown are the percentage of sequences with a significant match (E < 1E⁻⁷) to either a predicted OR, an uncharacterized (Unch.) or a none-OR loci, and those that did not match any genomic region (None). (e) Schematic highlighting of SameOR63 CDS region amplified by RT-PCR and sequenced using Sanger. The Sanger chromatogram at the bottom is an example of the sequencing results. (f) Frequency distributions of CDS length in Database A (dark blue) and Database B (light blue).

Figure 2

Expression of putative SameORs by tissue type and sex. RNAseq reads were aligned against identified putative SameORs and quantified. All data shown are based on DESeq2 normalized read counts. Quantification was performed in three replicates per condition. (a) Expression heat map of the 159 putative SameORs contained in Database A. Data are sorted according to antenna expression from highest to lowest normalized counts. Each row corresponds to the same putative SameOR in all columns. Colored circles next to each named SameOR indicate expression level: black, 10 < average <  ~ 4300; orange, 5 < average < 10; green, 0 < average < 5 normalized counts. Red: none detected. Light gray lines indicate grouped OR labels; see Supplementary Fig. 3 for an enlarged version. Asterisks: statistically significant differences in expression between female and male in antenna or palp tissues from panel (c). (b) Expression of the 10 most expressed putative SameORs in female (orange bars) and male (gray bars) tissue. Expression level is given relative to Orco. Error bars: standard deviation. (c) Differential expression between female and male in antennae (left panel) and palp (right panel) tissues. Red: differentially expressed putative SameORs padj < 0.1.

Figure 3

Expression of SameOrco, SameOR1, and SameOR2 in antenna and palp tissue sections, respectively. RNAscope in situ hybridization was used to characterize the spatial distribution of these putative ORs. (a–c) Probes against SameOrco (green) in antenna sections (a, b) and palp sections (c). Orco⁺ cells in trichoid (Tr) and basiconic (Ba) sensilla are highlighted by arrows in panel a. Trichoid sensilla containing one, two, and three ORNs are depicted by arrows in panel (b). (d, e) SameOrco probe with the probe against the second most expressed putative SameOR in each tissue type were used. SameOR1 (antenna, panel d) and SameOR2 (palp, panel e) are present only in Orco⁺ cells. Blue label in all images:DAPI. Scale bars: 10 μm in panels a–c; 15 μm in panels d–e.

Figure 4

Characterization of SameOR71, SameOR58, and SameOR160 expression in trichoid sensilla. RNAscope in situ hybridization was used to determine the spatial distribution of these putative ORs. (a) Examples of antenna sections showing different combinations of probes against SameOrco and SameOR58 + SameOR160 (top), SameOR58 + SameOR71 (middle), and SameOR160 + SameOR71 (bottom), respectively. (b) Examples of individual trichoid sensilla containing one, two, or three ORNs and different combinations of the SameORs tested and indicated by gray arrows. Note that the images showing one and two ORNs are the same as in Fig. 3b but with additional color channels. Middle panel: red arrows: a cluster of three Orco⁺ cells that do not express either of the two tested putative SameORs. (c) Summary of the configurations observed in trichoid sensilla. Scale bars: 20 and 10 μm in panel a and b, respectively.

Figure 5

OR protein family phylogeny including putative SameORs and two other members of the Acaridae family (S. gregaria—Sgre and L. migratoria—Lmig). The maximum likelihood tree was generated from amino acid sequences deduced from the 159 putative ORs contained in Database A, and published ORs from S. gregaria and L. migratoria (see Methods). (a) Enlarged view of the co-receptor Orco subgroup shown in panel (b). Confidence bootstrapping values are indicated. (b) The tree was rooted using Orco protein family members, including D. melanogaster (Dmel), A. gambiae (Agam), M. sexta (Msex), Sgre, Lmig, and putative SameOrco (shown in panel a) protein’s amino acid sequences. Confidence bootstrapping values above 80 are shown (1000 iterations and a correlation coefficient of 0.99 were used). For descriptive purposes the ORs were divided into eight groups based on the first seven nodes. The length of each predicted protein’s amino acid sequence is included in the labels.