Fig. 2

Diverse breadth and spatial location of non-neuronal olfactory subsystem-enriched genes. A tSNE plot of antennal single-cell transcriptomes highlighting non-neuronal classes in the antenna, as defined by expression of the indicated marker genes (based upon [29]). B–E tSNE plots of antennal single-cell transcriptomes and RNA FISH on whole-mount antennae of wild-type (Canton-S) animals illustrating non-neuronal expression patterns of various (B) Or subsystem-enriched and (C–E) Ir subsystem-enriched genes. Expression of a5 and a10 was previously described, but not related specifically to the Ir subsystem [80, 81]; the Jhedup RNA FISH expression pattern is consistent with observations of a transgenic promoter reporter for this gene [82]. The sacculus-specific Obp59a expression pattern—as previously described [54, 70]—is shown for comparison with novel, sacculus support cell-expressed genes. The bright-field channel is overlaid to reveal cuticle morphology; occasional fluorescence signal within sensillar hairs is likely artefactual. Scale bars, 20 μm. F Demarcation of antennal unannotated support cell clusters (I1-14, O1-7) through their expression of Ir and Or subsystem-enriched genes (shaded magenta and green boxes, respectively). Select genes illustrated in B–D are highlighted; see Additional file 5: Table S2 for the full dataset). Although expression was assessed qualitatively, not quantitatively, support cells could easily be categorized to each subsystem through their “fingerprint” of subsystem-specific gene expression. G tSNE plot of antennal single-cell transcriptomes in which antennal support cell clusters are assigned to the Or and Ir subsystems based upon their expression of subsystem-enriched genes. I11 likely represent sacculus support cells (see “Results” and D). Other antennal cell classes are also indicated (based upon [29]). sv-positive clusters (see A) labeled with “?” (e.g., near I10) were not reliably marked by Or or Ir subsystem-specific genes; these may represent support cells in the Johnston’s organ (or first antennal segment). As Johnston’s organ development is ato-dependent [83], we cannot exclude that there are shared markers between support cells of these two segments and that some of the “Ir subsystem” support cell clusters correspond to cells of this auditory organ