Fig 2: nSMase2 regulates sEV secretion by counteracting V-ATPase-mediated endosomal acidification. Working model showing how nSMase2 regulates sEV secretion via counteracting V-ATPase-mediated endosome acidification. Ceramide generated by nSMase2 activity on the MVB drives inward membrane invagination to selectively incorporate the V-ATPase subunit ATP6V0A1 (V0A1) and other sEV cargoes into ILVs targeted for secretion. The sequestration of V0A1 into ILVs attenuates MVB lumen acidification to favor secretory MVB trafficking for sEV release. V1A, ATP6V1A; V1E1, ATP6V1E1.

Fig 3: nSMase2 regulates endolysosomal acidification by modulating V-ATPase assembly. (A) Confocal microscopy images of HeLa cells treated with DMSO or GW4869 (GW), co-stained for ceramide and ATP6AP2. Nuclei were stained with Hoechst. (B) Dot plots of mean fluorescence intensity of ceramide puncta (left) and mean fluorescence intensity of ceramide that colocalized with ATP6AP2 (right) from A. Data are presented as mean±s.d.; n>50 cells from two biological replicates. ****P<0.0001 (unpaired two-tailed Student's t-test). (C) Confocal microscopy images of the cytoplasmic distribution of ATP6V1E1 (V1E1) in HeLa cells upon SMPD3 KD (siSMPD3) compared with that of control cells (siCtrl). Actin was stained using phalloidin, and nuclei were stained using Hoechst. Images are representative of three experiments. (D) Confocal microscopy images of DMSO- and GW4869-treated HeLa cells co-stained for ATP6V0A1 and ATP6V1A. Nuclei were stained with Hoechst. Arrows indicate ATP6V0A1 and ATPV1A colocalization. The scale bar in the magnified inset represents 5 µm. (E) Dot plots showing colocalization analysis of ATP6V1A with ATP6V0A1 from D. Data are presented as mean±s.d.; n>20 from two biological replicates. ****P<0.0001 (unpaired two-tailed Student's t-test). (F) Western blot analysis of ATP6V0A1 (V0A1), ATP6V1A (V1A) and ATP6V1E1 (V1E1), as well as GAPDH as loading control, in total cell lysates (CX) and sEV fractions prepared from control (siCtrl) and SMPD3 KD (siSMPD3) HeLa cells. (G) Quantification of signal intensity of V0A1 in sEV fractions from F normalized to the corresponding loading control GAPDH in the CX before normalization to the respective control. Data are presented as mean±s.d. of three biological replicates. **P<0.001 (unpaired two-tailed Student's t-test). (H) Western blot analysis of V0A1, V1A and V1E1 levels in organelle or cytosolic fractions upon SMPD3 KD in HeLa cells. Vinculin and GOSR2 were used as cytosolic or organelle markers, respectively. (I) Quantifications of signal intensity of V1A and V1E1 in organelle and cytosolic fractions, and V0A1 in the organelle fraction from H. Cytosolic and organelle signals for each subunit were normalized to their respective loading control, vinculin or GOSR2, before normalization to the control condition. Data are presented as mean±s.d. of three biological replicates. *P<0.01 (left, unpaired two-tailed Student's t-test; middle and right, one-way ANOVA followed by Sidak's multiple comparisons test). Molecular masses in D and H are indicated in kDa.

Fig 4: TNFa regulates sEV secretion through nSMase2 activation. (A) Confocal microscopy images of control (Ctrl) and TNFa-treated HeLa cells co-stained for ceramide and HRS. Nuclei were stained with Hoechst. The scale bar in the magnified inset represents 5 µm. (B) Dot plots of mean fluorescence intensity of ceramide puncta (left) and mean fluorescence intensity of ceramide that colocalized with HRS (right) from A. Data are presented as mean±s.d.; n>50 cells from two biological replicates. **P<0.001; ****P<0.0001 (unpaired two-tailed Student's t-test). (C) Western blot analysis of Alix, CD63, syntenin and CD81 in sEV fractions prepared from equal amounts of control or TNFa-treated HeLa cells. (D) Signal intensity quantifications of Alix, CD63, syntenin and CD81 in sEV fractions from C, calculated by normalizing the signal to loading control GAPDH levels in the corresponding cell lysates in E before normalization to their respective control. Data are presented as mean±s.d. of four biological replicates. **P<0.001; ***P<0.0004; ****P<0.0001; ns, not significant (unpaired two-tailed Student's t-test). (E) Western blot analysis of the indicated exosomal markers, GAPDH and calnexin in the corresponding cell lysates (CX) from C. GAPDH was probed as a loading control. (F) Quantifications of intracellular Alix, CD63, syntenin and CD81 signal intensity from E normalized to loading control GAPDH before normalization to the respective control. Data are presented as mean±s.d. of three biological replicates. **P<0.001 (unpaired two-tailed Student's t-test). (G) Western blot analysis of Alix, CD63 and syntenin in an sEV fraction prepared from an equal amount of control untreated cells (Ctrl) or control siRNA-treated (siCtrl) and SMPD3 KD (siSMPD3) HeLa cells treated with TNFa. (H) Signal intensity quantifications of Alix and syntenin in sEV fractions from G, calculated by normalizing the signal to loading control GAPDH levels in the corresponding cell lysates before normalization to their respective control. Data are presented as mean±s.d. of five (Alix) and four (syntenin) biological replicates. *P<0.01, **P<0.001, ***P<0.0004, ****P<0.0001 (one-way ANOVA followed by Tukey's multiple comparisons test). (I) NTA quantification of sEV concentration isolated from control and SMPD3 KD HeLa cells with or without TNFa treatment. Data are presented as mean±s.d. of three biological replicates. *P<0.01 (one-way ANOVA followed by Tukey's multiple comparisons test). (J) Representative size distribution of sEV analyzed by NTA from I. (K) Western blot analysis of ATPV0A1 (V0A1), ATP6V1A (V1A) and ATP6V1E1 (V1E1), as well as GAPDH as loading control, in the total cell lysates (CX) and corresponding sEV fractions prepared from untreated HeLa cells (Ctrl) and from control and SMPD3 KD HeLa cells treated with TNFa. (L) Quantifications of signal intensity of V0A1 in the sEV fraction from K normalized to the corresponding loading control GAPDH in the CX before normalization to the respective control. Data are presented as mean±s.d. of four biological replicates. (one-way ANOVA followed by Tukey's multiple comparisons test). Molecular masses in C,E,G and K are indicated in kDa.

Fig 5: SidK-AL568 detects the V-ATPase in acidic intracellular organelles. (A) HeLa cells were stained with SidK-AL568 (magenta). Nuclei labeled with DAPI (blue). Arrowheads mark peripheral, likely endocytic, vesicles (open) and juxtanuclear vesicles and cisternae, likely including the Golgi compartment (closed). (B–F) V-ATPase in HeLa cells labeled with SidK-AL568 (magenta) and costained for (B) LAMP1, (C) trans-Golgi, (D) cis-Golgi, (E) ER, and (F) mitochondria (green). Nuclei labeled with DAPI (blue). Side panels show individual SidK-AL568 and organelle channels in area denoted by dashed square, at 1.1, 0.9, 1.1, 1.7, and 1.5× magnification, respectively. (G and H) HeLa cells were immunostained using the V1A antibody, a-ATP6V1A (Abnova; catalog no. H00000523-M02; magenta) in cells costained for LAMP1 (G) or trans-Golgi (H), in green. Nuclei labeled with DAPI (blue). Side panels show individual a-V1A and organelle channels in dotted square, at 1.1 and 1.4× magnification, respectively. (A–H) are extended focus images representative of =30 fields from three or more experiments. Colocalization and Mander’s coefficients (M) between the organellar marker and V-ATPase probe shown in orange. All scale bars: 5 µm.