Fig. 3: GAPDH regulates exosome biogenesis and clustering in Drosophila secondary cells.

a Schematics show male fruit fly and its accessory gland (AG) containing main cells and secondary cells (SCs), which are only found at the distal tip of the gland. Exosomes can be visualized at the AG lumen as fluorescent puncta. A schematic of a secondary cell expressing a GFP-tagged form of Breathless (Btl-GFP; green) is also shown. The Rab11 compartments, which contain intraluminal vesicles (ILVs; green) and dense-core granules (DCGs; dark grey), and the late-endosomes and lysosomes (magenta) are marked. b Basal wide-field fluorescence and differential interference contrast (‘Merge’) views of living secondary cells (SCs) expressing GFP-tagged form of Breathless (Btl-GFP; green) with no other transgene (control); or also expressing the open reading frame of the human GAPDH protein (hGAPDH), an RNAi construct targeting Drosophila GAPDH1 (gapdh1–RNAi #1), or an RNAi construct targeting Drosophila GAPDH2 (gapdh2–RNAi #1). SC outline approximated by dashed white circles, and acidic compartments are marked by LysoTracker Red (magenta). Btl-GFP-positive intraluminal vesicles (ILVs; green in ‘Merge’; grey in ‘Zoom’) are apparent inside compartments, surrounding dense-core granules (DCGs; asterisk in ‘Zoom’) and connecting DCGs to the limiting membrane of the compartment (yellow arrowheads, except in GAPDH2 knockdown, where ILVs only surround peripheral small DCGs). DCG compartment outline is approximated by white circles. c Confocal transverse images of fixed accessory gland (AG) lumens from the same genotypes, containing Btl-GFP fluorescent puncta (yellow arrows). d Bar chart showing percentage of DCG compartments per cell containing clustered Btl-GFP-positive ILVs that are in contact with DCGs (hGAPDH, p = 0.5382; gapdh1–RNAi #1, p > 0.9999; gapdh2–RNAi #1, p < 0.0001; Kruskal–Wallis and Dunn correction). e Bar chart showing the percentage of DCG compartments per cell containing single spherical DCG (hGAPDH, p > 0.9999; gapdh1–RNAi #1, p = 0.0816; gapdh2–RNAi #1, p < 0.0001; Kruskal–Wallis and Dunn correction). f Bar chart showing number of Btl-GFP fluorescent puncta in the lumen of AGs for the different genotypes (hGAPDH, p = 0.0067; gapdh1–RNAi #1, p = 0.0238; gapdh2–RNAi #1, p = 0.0298; Kruskal–Wallis and Dunn correction). All data are from six-day-old male flies shifted to 29 °C at eclosion to induce expression of transgenes. Genotypes are: w; P[w⁺, tub-GAL80^ts]+; dsx-GAL4, P[w⁺, UAS-btl-GFP]/+ with no expression of other transgenes (control) (n = 11 glands/n = 31 cells) (n = 19 AG lumens for ILV secretion), or with UAS-hGAPDH (n = 10 glands/n = 33 cells)(n = 13 AG lumens), UAS-gapdh1-RNAi (n = 10 glands/n = 32 cells)(n = 11 AG lumens) or UAS-gapdh2-RNAi overexpression (n = 10 glands/n = 33 cells)(n = 11 AG lumens). Scale bars in (b) (5 µm), in ‘Zoom’ (1 µm), and in c (20 µm). ***p < 0.001, **p < 0.01, *p < 0.05 and ns non-significant relative to control, Kruskal–Wallis followed by Dunn’s multiple comparison test. Data shown in (d–f) as mean ± SEM (n = 3 independent biological experiments). All experiments shown were independently repeated three times. Source data are provided as a Source Data file.