Article Figures & Data
Figures
- Figure 1.
Population of nonspiking LNs has patchy morphology and lacks voltage-gated sodium current. Innervation patterns of a single LNs labeled by the (A) R32F10-Gal4 and (C) R70A09-Gal4 driver lines obtained by stochastic labeling using SPARC2 stochastic labeling (red and blue, respectively). Single cells are stochastically labeled from the Gal4 populations. B, D, Cartoon examples of the patchy and pan-glomerular LN morphologies, respectively. Sample LN responses to 10−4 pentyl acetate for (E) R70A09-Gal4 and (F) R32F10-Gal4 lines. Horizontal bar denotes the timing of the odor pulse as well as scale (500 ms). Other odors tested include methyl acetate and ethyl acetate at 10−4 and 10−2 dilutions, respectively. G, Current step stimulus applied to sample traces in H–J. H, Spiking LN voltage response to current clamp steps from G. I, Same as in H, except in the presence of 1 μm TTX. J, Voltage response to steps in G for a sample nonspiking LN. K, Voltage clamp responses of a single spiking LN (blue) and multiple nonspiking LNs (red, n = 9) during a voltage step from −60 mV holding potential to −30 mV for a duration of 50 ms. Horizontal bars denote time, vertical bars denote voltage or current.
- Figure 2.
Voltage-gated sodium channel gene transcript, para, is detected in adult nonspiking LNs but conditional tagging reveals lack of translation. A, Larval central nervous system (CNS) expression of para transcript revealed through HCR. White boxes indicate regions of interest expanded in B and C and also represent scale at 63 μm along each side. Red indicates Bacchus protein which labels most CNS neuron nuclei. Transcripts of para are indicated in green. B, C, Expanded views of brain and ventral nerve cord, respectively, as indicated in A. Transcript was rarely detected in brain regions labeled with Bacchus while overlap was common in the ventral nerve cord (n = 4). D, HCR straining for GFP transcript (red) in R32F10-GFP>UAS-mCD8::GFP (GFP in green). GFP transcript stain was only detected in GFP-positive cells. Scale bar indicates 10 μm. E, F, Sample GFP (left), para transcript (middle), and merge (right) of (E) nonspiking and (F) spiking LN populations in the AL. Transcripts were stained using hybridization chain reaction (Molecular Instruments), and images were masked to emphasize co-labeling with GFP. G, Violin plot of para transcript stain intensities in the AL, normalized by volume. n = 14 for R32F10-Gal4, 12 for R70A09. White circles denote means. Difference is not statistically significant (Student’s t test, p = 0.18). Sample images for (H) R32F10-Gal4 and (I) R70A09-Gal4 of UAS-Flp-driven para-FlpTag-GFP (red and blue, respectively). Dashed lines indicate measured AL regions; arrowheads indicate examples of non-AL staining. Background stain (anti-NCAD) is shown in gray. J, Violin plot of GFP labeling intensity, normalized by AL volume. n = 12 for R32F10-Gal4, 14 for R70A09. White circles denote means. Difference is significant (Student’s t test, p = 0.000035).
- Figure 3.
Nonspiking LNs labeled by R32F10 are a distinct population of neurons. Full brain expression patterns of (A) R32F10-Gal4, (B) R70A09-Gal4, and (C) NP3056-Gal4 expressing GFP (green). Background stain is anti-NCAD (magenta). Sample images of R32F10-LexA > LexAop-mCherry (D1, F1), NP3056-Gal4 > UAS-GFP (D2), R70A09-Gal4 > UAS-GFP (F2), and merges (E3, G3). White dashed lines denote the location of R32F10-LexA labeled somas. Quantification indicates effectively zero overlap between R32F10-Lexa and (E) NP3056-Gal4 (n = 5) and (G) R70A09-Gal4 (n = 6).
- Figure 4.
Sample odor activation patterns of LN lines reveal specific responses in nonspiking cells. NP3056-Gal4 (A1–A4 and B1–B4), R70A09-Gal4 (C1–C4 and D1–D4), and R32F10-Gal4 (E1–E4 and F1–F4) > UAS-GCAMP7b response patterns and temporal ΔF/F traces to pentyl acetate, benzaldehyde, cVA, and valeric acid. Images are an average of the three frames around the peak of the ΔF/F in response to a given odor and are normalized to the same scale within a given fly. Colorbars indicate the range of ΔF/F values in percent. Traces represent the average of three trials ± SEM. Vertical bars denote ΔF/F value in percent. Horizontal bar denotes the timing of the odor pulse as well as scale (500 ms). Odors were presented at 10−4 dilution in the odor vial except for cVA, which was pure. Scale bar in A1 represents 10 μm. Images and traces each correspond to the same fly, e.g., one fly was used to generate A1–A4 and B1–B4. G, Response latency, half rise time, half decay time, and response duration of odor responses average across flies and odors. Central lines indicate means while the top and bottom edges of the boxes indicate 75th and 25th percentiles, respectively. The whiskers represent the range from minimum to maximum, excluding outliers. Outliers were excluded from plotting but included in analysis. R32F10-Gal4 was significantly different from the other lines for half decay time and response duration (denoted by *, ANOVA with Tukey’s post hoc test, p = 8.8e-4 and 6.2e-4, respectively).
- Figure 5.
Nonspiking LN odor response patterns requires more PCs to explain comparable variance. Odor response patterns in nonspiking LNs are decorrelated. PCA was run across odor response patterns for individual flies and results for a single fly are shown here. Score outputs are projected back onto the AL to visualize sample PCs in the context of odor response spatial patterns. PC1 shows substantial contribution to the variance explanation in NP3056-Gal4 (A1) and R70A09-Gal4 (B1), while PCs 2–4 show little patterning (A2–A4 and B2–B4). All PCs display scores in distinct spatial patterns for R32F10-Gal4 (C1–C4). D, Mean variance explained by each PC (% ± SD), n = 8 for NP3056-Gal4, 15 for R32F10-Gal4, and 10 for R70A09-Gal4 lines. PCA was performed on each fly individually, and explained variances were pooled for plotting and statistical analysis. In each PC, the explained variance for R32F10-Gal4 is significantly different from the other lines (denoted by *, ANOVA with Tukey’s post hoc test, p = 1.6e−8, 5.8e−6,1.1e−7,1.4e−6 for PCs 1–4, respectively). Correlation coefficients were calculated between odor responses for (E) NP3056-Gal4, (F) R70A09-Gal4, and (G) R32F10-Gal4 lines. Coefficients were calculated between odors for individual flies and are presented as averages. * denotes statistical significance in the difference between a given odor pair and the corresponding odor pairs of the other two LN lines (Kruskal–Wallis with Dunn’s post hoc test, p = 2.9e−4 for pentyl acetate and benzaldehyde, p = 3.7e−3 for pentyl acetate and cVA, p = 0.52 for pentyl acetate and valeric acid, p = 6.5e−5 for benzaldehyde and cVA, p = 2.9e−3 for benzaldehyde and valeric acid, and p = 0.011 for cVA and valeric acid).
- Figure 6.
Sizes of activated regions increase with odor concentration in nonspiking LNs. A–C, Same as Figure 4, but for increasing concentrations of pentyl acetate. Odor concentrations were 10−10, 10−8, 10−6, and 10−4. Scale bar in A1 denotes 10 μm. Colorbars indicate the range of ΔF/F values in percent. D, Mean percent of the AL activated by the odor, ±SEM, as determined by the number of pixels above an intensity equal to the 40th percentile of the dynamic range of all AL pixels for a given fly. Samples of this region are outlined in white dashes in A4, B4, and C4. Dashed lines denote statistical significance in the difference between the indicated R32F10-Gal4 odor concentration pairs (ANOVA, p = 0.00005, with Tukey’s post hoc test, p values: 10−10 and 10−8 = 0.9677; 10−10 and 10−6 = 0.0379; 10−10 and 10−4 = 0.0013; 10−8 and 10−6 = 0.1047; 10−8 and 10−4 = 0.0049; 10−6 and 10−4 = 0.5954). Concentration pairs for R70A09-Gal4 and NP3056-Gal4 were not significant (ANOVA, p = 0.6383 and 0.3374, respectively). n = 11, 12, and 10 for NP3056-Gal4, R70A09-Gal4, and R32F10-Gal4, respectively.
Tables
Fly Line Source Identifier GMR32F10-Gal4 Bloomington RRID:BDSC_49725 GMR70A09-Gal4 Bloomington RRID:BDSC_47720 UAS-mCD8::GFP Bloomington RRID:BDSC_32194 GMR32F10-LexA Bloomington RRID:BDSC_53565 LexAop-GFP Bloomington RRID:BDSC_52266 UAS-FLP Bloomington RRID:BDSC_4539 para-flptag-GFP Bloomington RRID:BDSC_92146 NP3056-Gal4 DGRC RRID:113080 UAS-jGCAMP7b Bloomington RRID:BDSC_79029 UAS-mCherry Bloomington RRID:BDSC_52268 LexAop-mCherry Bloomington RRID:BDSC_52271 Transgenic Drosophila lines used in this study from the Bloomington Drosophila Stock Center and the Drosophila Genomics Resource Center.
Figure Odorant Dilution Solvent Source Identifier 1 Pentyl acetate 10−2, 10−4 Parraffin oil Sigma-Aldrich CAS: 628-63-7 4, 5 Pentyl acetate 10−4 Parraffin oil Sigma-Aldrich CAS: 628-63-7 4, 5 Benzaldehyde 10−4 Parraffin oil Sigma-Aldrich CAS: 100-52-7 4, 5 cVA Pure None Pherobank CAS: 6186-98-7 4, 5 Valeric acid 10−4 Water Sigma-Aldrich CAS: 109-52-4 6 Pentyl acetate 10−10, 10−8, 10−6, 10−4 Parraffin oil Sigma-Aldrich CAS: 628-63-7 Chemicals, dilutions, solvents, and sources of odors used during odor exposure experiments.
Reagent Dilution Source Identifier Mouse anti-Bruchpilot 1:50 Developmental Studies Hybridoma Bank, Iowa Catalog #nc82; RRID: AB_2314866 Rat anti-NCAD 1:500 Developmental Studies Hybridoma Bank, Iowa Catalog #DN-Ex #8; RRID: AB_528121 Chicken anti-GFP 1:1000 Invitrogen Catalog #A-10262; RRID: AB_2534023 Rabbit anti-DsRed 1:400 Clontech Catalog #632496; RRID: AB_10013483 Streptavidin Alexa Fluor 568 1:250 Invitrogen Catalog #S11226; RRID: AB_2315774 Goat anti-chicken Alexa Fluor 488 1:250 Invitrogen Catalog #A-11039; RRID: AB_2534096 Goat anti-rabbit Alexa Fluor 633 1:250 Invitrogen Catalog #A-21071; RRID: AB_2535732 Goat anti-rat Alexa Fluor 568 1:250 Invitrogen Catalog #A-11077; RRID: AB_2534121 Goat anti-mouse Alexa Fluor 633 1:250 Invitrogen Catalog #A-21050; RRID: AB_2535718 HCR fruitfly-para-B1 1:250 Molecular Instruments N/A HCR transgenic-EGFP-B3 1:250 Molecular Instruments N/A HCR amplifier B1 Alexa Fluor 647 1:50 Molecular Instruments N/A HCR amplifier B3 Alexa Fluor 546 1:50 Molecular Instruments N/A Biocytin Hydrazide 13 mm Life Technologies Catalog #B-1603 Primary and secondary antibodies and in situ hybridization reagents and sources.
Figure Distribution Test Sample size Statistical data 2G Normal Unpaired t test (two tailed) R32F10 n = 14, R70A09 n = 12 t = −1.3664, df = 24, p = 0.1845 2J Normal Unpaired t test (two tailed) R32F10 n = 12, R70A09 n = 14 t = 5.0714, df = 24, p = 0.000035 4G Normal ANOVA (Tukey’s post hoc test) NP3056 n = 8, R32F10 n = 15, R70A09 n = 11 Latency p = 0.3484 half rise p = 0.1807 half decay p = 8.8e-4 response duration p = 6.2e-4 5D Normal ANOVA NP3056 n = 8, R32F10 n = 15, R70A09 n = 10 F = [35.5101, 18.5213, 28.5431, 21.8342];
p = [1.6e-8, 5.8e-6, 1.1e-7, 1.4e-6]5G Nonparametric Kruskal–Wallis NP3056 n = 8, R32F10 n = 15, R70A09 n = 11 Descending, left to right:
X2 = [16.3083, 11.1783, 1.3186, 19.2841, 11.6920,
9.0253];
p = [2.9e-4, 0.0037, 0.5172, 6.5e-5, 0.0029, 0.011]6D Normal ANOVA (Tukey’s post hoc test) NP3056 n = 11, R32F10 n = 10, R70A09 n = 12 R70A09: F = 0.5698, p = 0.6383;
NP3056: F = 1.1598, p = 0.3374;
R32F10: F = 7.4735, p = 0.00005;
R32F10 concentration pair p values:
−10 and −8 = 0.9677; −10 and −6 = 0.0379;
−10 and −4 = 0.0013; −8 and −6 = 0.1047;
−8 and −4 = 0.0049; −6 and −4 = 0.5954Summary of statistics used in each figure
In this issue
