Cytokine exocytosis and JAK/STAT activation in the Drosophila ovary requires the vesicle trafficking regulator α-Snap

Abstract

How vesicle trafficking components actively contribute to regulation of paracrine signaling is unclear. We genetically uncovered a requirement for α-soluble NSF attachment protein (α-Snap) in the activation of the Janus kinase/signal transducer and activator of transcription (JAK/STAT) pathway during Drosophila egg development. α-Snap, a well-conserved vesicle trafficking regulator, mediates association of N-ethylmaleimide-sensitive factor (NSF) and SNAREs to promote vesicle fusion. Depletion of α-Snap or the SNARE family member Syntaxin1A in epithelia blocks polar cells maintenance and prevents specification of motile border cells. Blocking apoptosis rescues polar cell maintenance in α-Snap-depleted egg chambers, indicating that the lack of border cells in mutants is due to impaired signaling. Genetic experiments implicate α-Snap and NSF in secretion of a STAT-activating cytokine. Live imaging suggests that changes in intracellular Ca²⁺ are linked to this event. Our data suggest a cell-type specific requirement for particular vesicle trafficking components in regulated exocytosis during development. Given the central role for STAT signaling in immunity, this work may shed light on regulation of cytokine release in humans.

Keywords: Cell fate specification; Drosophila oogenesis; Exocytosis; JAK/STAT signaling.

Fig. 1.

α-Snap or Syx1A depletion in polar cells prevents border cell specification. Shown are egg chambers of indicated genotypes and stages. Egg chambers in all panels are oriented with anterior to the left, posterior to the right; scale bars: 20 µm. Nuclei were stained using DAPI (blue). Arrowheads indicate border cell clusters. (A,B) c306-Gal4-driven expression of UAS-mCD8-GFP (green; the ‘>’ symbol in labels denotes drives expression of) in polar cells of egg chambers at stage 2 (A), and in polar cells and border cells (BCs) of egg chambers at stages 8 and 9 (A) and stage 10 (B). Expression of Eya (red) marks follicle cells. (C,D) c306-Gal4-driven α-Snap RNAi (C) or Syx1A RNAi (D) results in no BC specification. Compare red staining of Arm+Eya (C) and Arm (D) with that shown in panel L. (E) Stage 8 wild-type anterior follicle cells, showing cytoplasmic α-Snap (green) apically enriched; cortical F-actin (red). The asterisk indicates the area shown magnified in insets. (F,G) upd-Gal4-driven GFP expression in polar cells at stages 2 (F) to 10 (G). (H,I) upd-Gal4-driven α-Snap RNAi (H) or Syx1A RNAi (I) results in no BC specification. (J,K) fruitless-Gal4 drives GFP (green) in anterior cells, excluding polar cells (FasIII, red), before (J) and after (K) BC specification (magnified in insets). (L) fruitless-Gal4-driven α-Snap RNAi produces wild-type BCs. (M) Penetrance of the BC phenotype for the indicated genotypes. A lack of bar (1^st, 5^th, 8^th and 9^th position) in the graph means that all egg chambers contained border cells. ***P<0.0005 (two-tailed Fisher's exact test).

Fig. 2.

Blocking apoptosis in α-Snap-depleted polar cells does not rescue border cell specification. (A,B) upd-Gal4 driving α-Snap RNAi results in egg chambers that have polar cells early (FasIII-positive, red; arrow in A), but lack polar and border cells later (B; Arm, red; DAPI, blue). (C) p35 expression in polar cells yields normal border cells (arrowhead). (D,E) p35 expression in α-Snap-depleted polar cells rescues polar cell maintenance but not border cell specification (arrow, magnified in insets). All scale bars: 20 µm. (F) Penetrance of the no-polar (PC) and/or no-border cells (BC) phenotypes by genotype (line JF03266).

Fig. 3.

α-Snap functions upstream of Upd to promote JAK/STAT signaling and cell specification. (A,B) Border cells (arrow; Eya- and Arm-positive, red) form and migrate correctly in α-Snap^G8/+ (A) or Stat92E³⁹⁷/+ (B) heterozygotes. (C) An α-Snap^G8/Stat³⁹⁷ egg chamber that contains migratory-defective border cells. (D) Penetrance of border cell migration defects by genotype. (E) Penetrance of the no border cell phenotype by genotype. A lack of bar (4^st and 5^th position) in the graph indicates that all egg chambers contained border cells. *P<0.05, ***P<0.0005, two-tailed Fisher's exact test. (F) upd overexpression in polar cells yields a wild-type phenotype.

Fig. 4.

Ca²⁺ signaling in polar cells prior to border cell specification. (A,B,C) Movie stills of Movies 1–3, showing transient Ca²⁺-dependent GFP fluorescence in polar cells of egg chambers. The transgenes used are indicated. White arrowheads mark fluorescence in polar cells; outlined arrowheads indicate the same cells at a different time without fluorescence. In C, white stars indicate fluorescence in polar cells of a younger egg chamber; outlined stars indicate the same cell at a different time without fluorescence. GCaMP6-GFP (green; the ‘>’ symbol after Gal4 in labels denotes ‘drives expression of’); FM4-64 lipophilic dye, red. (A’,B’,C’) Quantifications of the changes in fluorescence intensity (ΔF/F) of the respective cells shown in Movies 1–3 over the timeframe of each movie. C' shows the fluorescence intensity of the starred cell (ROI*, blue) and the cell marked by the arrow at two different positions due to shifting of the egg chamber (ROIs orange and yellow). (D–G) Egg chambers at stages 9 and 10 of the indicated genotypes; Apt (red) marks anterior follicle cells, FasIII (green) polar cells and DAPI (blue) nuclei. upd-Gal4-driven expression of UAS-GFP (green) in polar cells results in at least six migratory cells (see square bracket in D). upd-Gal4-driven expression of UAS-PV (green) in polar cells to sequesters free Ca²⁺ results in fewer migratory cells (three cells, triangles), and poor migration (E,G). The average number of clustered migratory cells in upd-Gal4; UAS-PV, 4.8 (n=23), was significantly smaller than the average of 5.8 cells observed in upd-Gal4; UAS-GFP, (n=28); P<0.002; Mann-Whitney U test. (H) A proposed model of Upd cytokine secretion from polar cells. (1) Upd is loaded into apically targeted vesicles that then associate with Syx1A at the plasma membrane and dock. (2) α-Snap and NSF/NSF2 associate with Syx1 and the vesicle SNARE complex. In the presence of Ca²⁺, membranes fuse and Upd is released (3). α-Snap and NSF are required to reset the vesicle fusion machinery for another round; Syx1A is recycled back to the plasma membrane. α-Snap also mediates the release of a polar cell maintenance factor(s) (X).