Fig 1.
Cell-type-specific dissection of the endogenous active zone scaffold.
(A) Schematic of split-GFP tagging system. GFP11 was inserted prior to the stop codon of brp, while GFP1–10 was expressed in specific cell types using the GAL4-UAS system. GFP11 and GFP1–10 reconstitute to emit fluorescence. (B) Visualization of Brp::rGFP in a cell-type-specific manner. Maximum intensity projections are shown. Immunostaining using antibody nc82 which labels all Brp in the brain, as a comparison to Brp::rGFP. Diagrams in each panel show the innervation pattern of neurons in the MB. GAL4 lines used: γ KCs (MB009B-GAL4), α/β KCs (MB008B-GAL4), α′/β′ KCs (MB370B-GAL4), PPL-1 (TH-GAL4), PAM (R58E02-GAL4), DPM (VT64246-GAL4), APL (GH146-GAL4). Scale bar, 20 μm. (C) Long-term live imaging of Brp::rGFP (green) in a growing motor neuron (magenta, CD4::tdTomato) from 48 to 96 h APF. OK371-GAL4 was used to express GFP1–10 and CD4::tdTomato. White boxes indicate zoomed-in areas shown in the lower panels. Scale bars: upper panels, 10 μm; lower panels, 5 μm. (D) Ex vivo live-imaging of Brp::rGFP in 3rd instar larval KCs. GFP1–10 was expressed using R13F02-GAL4. The left panel shows the MB calyx region. White boxes indicate zoomed-in areas shown on the right panels. Boxes 1 and 2 demonstrate Brp::rGFP fission and fusion event. White arrows indicate Brp::rGFP clusters undergoing fission or fusion. Scale bars: left panel, 10 μm; right panels, 200 nm.
Fig 2.
Intracellular active zone heterogeneity among MB compartments.
(A) Brp::rGFP visualized in KCs. Maximum intensity projection showing the horizontal lobe of the MB. Brp::rGFP was visualized using R13F02-GAL4. (B) Diagram showing compartments in the MB γ lobe and the intracellular Brp localization in a single γ KC. (C) Representative maximum intensity projection visualizing Brp::rGFP and CD4::tdTomato in a single γ KC. (D) Signal intensity of Brp::rGFP clusters of single KCs in different compartments. Individual Brp::rGFP clusters are measured. The graph shows median Brp::rGFP intensities of compartments, normalized to the average of the five compartments’ medians. n = 14 KCs (one KC per mushroom body, 9 brains). Repeated measures one-way ANOVA. γ1 vs. γ2: P = 0.0004; γ1 vs. γ3: P < 0.0001; γ1 vs. γ4: P < 0.0001; γ1 vs. γ5: P < 0.0001; γ2 vs. γ3: P < 0.0001; γ2 vs. γ4: P = 0.0001; γ2 vs. γ5: P < 0.0001. (E) Basal Ca2+ concentration near the active zones varies by compartments. The left panel shows the schematic of the Brp::mCherry::GCaMP6s ratiomatric sensor. GCaMP6s sensor is fused to mCherry and Brpshort. The sensor was expressed by R13F02-GAL4 and GCaMP6s signal is normalized by mCherry for analysis. Repeated measures one-way ANOVA. n = 8 brains. γ1 vs. γ3: P = 0.0317; γ1 vs. γ4: P = 0.0307; γ1 vs. γ5: P < 0.0001; γ2 vs. γ3: P = 0.0150; γ2 vs. γ4: P = 0.0307; γ2 vs. γ5: P = 0.0150; Scale bars, 10 μm. Error bars show S.E.M. Significant differences (P < 0.05) are indicated by distinct letters. The data underlying this Figure can be found in S1 Data.
Fig 3.
Acute starvation decreases the Brp::rGFP heterogeneity level among γcompartments.
(A) Graphical summary of the image analysis. 3D volumes were sampled from different γ compartments. Brp::rGFP clusters are then spatially detected using the 3D spot segmentation Fiji plugin. rGFP signal of individual clusters was measured and the median intensity is calculated for each compartment. (B) Median intensity of Brp::rGFP clusters in γ lobe compartments. R13F02-GAL4 was used. Repeated measures one-way ANOVA. n = 11 brains. γ1 vs. γ2: P < 0.0001; γ1 vs. γ3: P < 0.0001; γ1 vs. γ4: P < 0.0001; γ1 vs. γ5: P < 0.0001. γ2 vs. γ5: P = 0.0101; γ3 vs. γ5: P = 0.0162; γ4 vs. γ5: P = 0.0038. Error bars show S.E.M. (C) Standard deviation (SD) as an indicative of Brp::rGFP heterogeneity level among compartments. The median of Brp::rGFP intensity of clusters in each compartment was log-transformed and SD was calculated from five log-transformed medians for each brain sample. (D) Starvation for 48 hours reduced the Brp heterogeneity level in KCs. One-to-two-week-old flies were used for this experiment. Pseudo color in the images represents the value of log2 (pixel intensity/mean pixel intensity in the γ lobe). Pseudo color range: −1.0 to 1.0. The heterogeneity level (SD) was quantified for individual brain samples. One-way ANOVA on log-transformed data. Fed (n = 19 brains) vs. Starved (n = 17 brains): P = 0.0167; Starved vs. Refeed (n = 19 brains): P = 0.0167; Fed vs. Refeed: P = 0.8271; Scale bar, 10 μm. Box plots showing center (median), whiskers (Min. to Max.). Significant differences (P < 0.05) are indicated by distinct letters. The data underlying this Figure can be found in S1 Data.
Fig 4.
Octopamine controls the AZ structure in KCs.
(A) OANs in the adult brain. OANs were labeled by Tdc2-GAL4 driven CD4::tdTomato in the left panel. OA-VPM3/4 neurons were labeled with MB022B-GAL4 [10] driven CD8::GFP (right panel). Dashed lines indicate the MB. Scale bars, 20 µm. (B) AZ number and synaptic vesicle localization of OANs along the γ lobe. GFP1–10 and CD4::tdTomato were expressed using Tdc2-GAL4. nSyb::CLIP labeling synaptic vesicles and CD8::GFP were expressed using MB022B-GAL4. Scale bars, 2 µm. Images were cropped from the same MB and are identical in size. (C) AZ density of OANs across γ compartment. Brp::rGFP was visualized using Tdc2-GAL4, and cluster density was quantified. n = 6 brains. Repeated measures one-way ANOVA. γ1 vs. γ2: P < 0.0001; γ1 vs. γ3: P < 0.0001; γ1 vs. γ5: P < 0.0001; γ1 vs. γ5: P < 0.0001. γ2 vs. γ4: P = 0.0098; γ3 vs. γ5: P = 0.0407; γ4 vs. γ5: P = 0.0015. Error bars show S.E.M. (D) Schematic showing the innervation pattern of OANs along the γ lobe. (E) Reduced Brp compartmental heterogeneity level in Tβh mutants. Representative images of Brp::rGFP in control and TβhSK2-4 background are shown. Pseudo color in the images represents the value of log2 (pixel intensity/mean pixel intensity in the γ lobe). Pseudo color range: −1.0 to 1.0. Scale bars, 10 µm. Kruskal–Wallis test. Ctrl (n = 12 brains) vs. TβhSK1-8 (n = 11 brains) and TβhSK2-4 (n = 12 brains). Ctrl vs. TβhSK1-8: P = 0.0442; Ctrl vs. TβhSK2-4: P = 0.0228. *P < 0.05. Box plots showing center (median), whiskers (Min. to Max.). Significant differences (P < 0.05) are indicated by distinct letters or *. The data underlying this Figure can be found in S1 Data.
Fig 5.
Octβ2R and cAMP underlie the Brp compartmental heterogeneity.
(A) Knockdown of Octβ2R reduced the Brp heterogeneity level in γ KCs. Receptors are knocked-down in γ KCs specifically using RNAi with R13F02-GAL4. Kruskal–Wallis test. Ctrl (n = 13 brains) vs. OAMB (n = 11 brains), Octα2R (n = 10 brains) and Octβ2R (n = 12 brains). P = 0.0228 for Ctrl vs. Octβ2R. Ctrl (n = 20 brains) vs. TyrR (n = 22 brains), TyrR2 (n = 22 brains) and OctTyrR (n = 15 brains). *P < 0.05; ns = not significant. (B) Knockdown of Rut significantly decreased the Brp heterogeneity level in γ KCs. Representative images showing Brp::rGFP in control and in Rut knockdown group. Pseudo color in the images represents the value of log2(pixel intensity/mean pixel intensity in the γ lobe). Pseudo color range: −1.2 to 1.2. Mann–Whitney U-test. Ctrl (n = 10 brains) vs. Rut-RNAi (n = 10 brains): P = 0.0011. Box plots showing center (median), whiskers (Min. to Max.). *P < 0.05 and ns = not significant. The data underlying this Figure can be found in S1 Data.
Fig 6.
Internal states adjust the Brp compartmental heterogeneity through octopamine.
(A) Graphical summary of the working model. Localized innervation of OANs onto γ KCs regulates the Brp accumulation and forms the Brp compartmental heterogeneity via Octβ2R and cAMP pathway. Nutritional states may modulate OAN activity, thereby adjusting the Brp heterogeneity. (B) Starvation for 48 hours did not affect the Brp::rGFP heterogeneity level in TβhSK2-4 mutant. GFP1−10 was expressed in KCs using R13F02-GAL4 in control, or TβhSK2-4 background flies. 1−2 weeks old flies were used. Pseudo color in the images represents the value of log2 (pixel intensity/mean pixel intensity in the γ lobe). Pseudo color range: −1.2 to 1.2. For the wild-type (WT) group, Fed (n = 16 brains) vs. Starved (n = 17 brains): P = 0.0187; For the TβhSK2-4 group, Fed (n = 16 brains) vs. Starved (n = 17 brains): P > 0.05. Mann–Whitney U-test. (C) Sleep deprivation for 12 h decreases Brp::rGFP heterogeneity level in a Tβh-dependent manner. For the Ctrl group, Undisturbed (n = 13 brains) vs. Sleep deprived (n = 14 brains): P = 0.0222; For the TβhSK2-4 group, Undisturbed (n = 14 brains) vs. Sleep deprived (n = 12 brains): P > 0.05. Mann–Whitney U-test. Pseudo color range: −1.1 to 1.1. Box plots showing center (median), whiskers (Min. to Max.). *P < 0.05 and ns = not significant. The data underlying this Figure can be found in S1 Data.
Table 1.
Key resources used in this study.