Fig 2: The assembly of the V-ATPase V1-V0 holoenzyme is dependent on microtubule acetylation A, BPerinuclear enrichment and colocalization of acetylated microtubules, lysosomes, and V-ATPase V1 subunits. RPE1 cells stably expressing LAMP1-GFP were treated with vehicle (VEH) or 10 µg/ml nocodazole (Noco) for 6 h before being fixed and subjected to immunofluorescence staining with antibodies against ATP6V1C1 and acetylated-a-tubulin (Ac-a-tubulin). The nucleus was stained with DAPI. Representative images of LAMP1-GFP, ATP6V1C1, and Ac-a-tubulin immunostaining are shown in (A). Representative distribution profiles of Ac-a-tubulin, LAMP1-GFP and ATP6V1C1 were plotted by Fiji software in (B).C–FThe assembly of V-ATPase V1-V0 holoenzymes after treatment with an a-tubulin deacetylase inhibitor or a microtubule depolymerizing agent. RPE1 cells were treated with VEH, tubacin (2.5 µM, 20 h), Noco (10 µg/ml, 6 h), or in combination. Cytosolic and membrane fractions were then extracted and analyzed by immunoblotting using antibodies against V-ATPase subunit V1A (ATP6V1A), V1C1 (ATP6V1C1), or V0D (ATP6V0D). LAMP1 and GAPDH were used as loading controls for membrane proteins and cytosolic proteins, respectively. The levels of ATP6V0D in the membrane fraction are shown in (D) after normalization against LAMP1 (n = 3 independent experiments, ns represents non-significant). The levels of assembled ATP6V1A (E) (n = 3 independent experiments, *P = 0.0118, **P = 0.0040) and ATP6V1C1 (F) (n = 3 independent experiments, *P = 0.0214, **P = 0.0013) are normalized against ATP6V0D in the membrane fraction.GRepresentative immunoblots for Ac-a-tubulin and GAPDH in the samples from (C).H–KThe assembly of V-ATPase V1-V0 holoenzymes in WT RPE1 cells over-expressing the a-tubulin deacetylase HDAC6 or EGFP as a control. Representative immunoblots for ATP6V0D, ATP6V1A, ATP6V1C1, LAMP1, and GAPDH are shown in (H). The levels of ATP6V0D in the membrane fraction (I) (n = 3 independent experiments, P = 0.2602) and those of assembled ATP6V1A (J) and ATP6V1C1 (K) (n = 3 independent experiments, *P = 0.0101, ***P = 0.0008) are quantified.LRepresentative immunoblots for HDAC6, Ac-a-tubulin and GAPDH in the WT RPE1 cells over-expressing HDAC6 or EGFP.M–OLoss of NDST3 enhances microtubule acetylation and the colocalization of the lysosomes with V-ATPase V1 subunits. WT and NDST3 KO RPE1 cells stably expressing LAMP1-GFP were fixed and analyzed by immunofluorescence staining with antibodies against ATP6V1C1 and Ac-a-tubulin. The relative levels of Ac-a-tubulin (N) (n = 75 cells in the WT group, n = 79 cells in the NDST3 KO group, ****P < 0.0001) and the colocalization of LAMP1-GFP with ATP6V1C1 (O) (n = 75 cells in the WT group, n = 79 cells in the NDST3 KO group, ****P < 0.0001) are quantified.P–SThe assembly of V-ATPase V1-V0 holoenzymes in NDST3 KO RPE1 cells over-expressing HDAC6 or EGFP as a control. Representative immunoblots for ATP6V0D, ATP6V1A, ATP6V1C1, LAMP1, and GAPDH are shown in (P). The levels of ATP6V0D in the membrane fraction relative to LAMP1 (Q) (n = 3 independent experiments, P = 0.9795) as well as the levels of assembled ATP6V1A (R) (n = 3 independent experiments, *P = 0.0316) and ATP6V1C1 (S) (n = 3 independent experiments, **P = 0.0027) are quantified.TRepresentative immunoblots for HDAC6, Ac-a-tubulin and GAPDH in the NDST3 KO RPE1 cells over-expressing HDAC6 or EGFP. Data information: Error bars represent ± standard deviation. Scale bar, 10 µm. Source data are available online for this figure.