Fig 1: Effect of retromer and WIPI1 on human endosomes Colocalization of WIPI1 with Vps26A. The indicated WIPI1mCherry variants and VPS26EGFP were expressed for 18 h in HK2 cells, from which endogenous WIPI1 had been deleted (WIPI1-KO). The cells were analyzed by confocal microscopy. Scale bars: 10 µm. Insets show enlargements of the outlined areas.Colocalization between WIPI1 variants and Vps26 was quantified in the cells from A using the Manders’ colocalization coefficient M2. The red line indicates the mean; n = 120 cells per condition, pooled from three independent experiments. P values were calculated by t-test (analysis performed with 99% confidence ***P < 0.0001).Expression levels of WIPI1mCherry variants. Lysates (50 µg of protein per sample) from the cells in A were analyzed by SDS–PAGE and Western blot against WIPI1 and tubulin.Depletion of VPS35. HK2 cells expressing the indicated WIPI1eGFP variants were transfected with siRNA against VPS35 (VPS35 KD) or a control siRNA pool. Lysates (50 µg per sample) from the cells were analyzed by SDS–PAGE and Western blot against Vps35 and tubulin.Blots from D were quantified on a LICOR Odyssey fluorescence imager. n = 3. Red lines indicate the means, n = 3 independent experiments, using a Welch’s t-test statistical analysis. Bars represent the mean and error bars the SD.Cells from D were analyzed by confocal microscopy 18 h after transfection. Scale bar: 10 µm.Quantification of tubule length in the cells from F. Data are means ± SD. of n = 210 tubules per condition, from three independent experiments. P values were calculated by unpaired Student’s t-test. 99% confidence: ***P < 0.0001.
Fig 2: Interaction of Atg18 and CSC Vps17 labilizes the Atg18-Vps26 interaction. Wild-type or vps17? cells expressing ATG18HA3-yEGFP from a centromeric plasmid were logarithmically grown in YPD. Genomically tagged Vps26yomCherry was pulled down from whole-cell extracts and analyzed for associated Atg18HA3-yEGFP by SDS–PAGE and Western blotting. Glucose-6-phosphate dehydrogenase (Zwf1) serves as a loading control. The intensity of the interacting Atg18HA3-yEGFP was quantified on a LICOR fluorescence imager and normalized to the amount of Vps26yomCherry. Values of the wild-type interaction were used as the reference and set to 1. n = 4 independent experiments were analyzed using an unpaired Student's t-test. Bars represent the mean and errors bars the SEM, **P < 0.01).Epistasis of ATG18 and retromer genes concerning vacuolar morphology. The indicated cells were logarithmically grown in YPD medium, stained with FM4-64 and calcofluor white as in Fig 2A and analyzed by confocal microscopy. Scale bar: 5 µm.Quantification of vacuole morphology. The number of vacuoles per cell was measured in the cells from B. The graph shows the fractions of cells displaying the indicated numbers of vacuolar vesicles (n = 3 biological experiments with at least 100 cells per condition and experiment were analyzed using an unpaired Student's t-test. Bars represent the mean, and errors bars represent the SEM. **P < 0.01.
Fig 3: SNX27 is required for activity-dependent membrane delivery of the GluN2A-containing NMDARs(A) At DIV12, primary hippocampal neurons were co-transfected with plasmids encoding SEP-GluN2A, either WT, or the phospho-deficient S1459A (SA), or phospho-mimetic S1459D (SD) mutant, with various pRK5-myc-SNX27 constructs, including WT, the retromer-associated VPS26 binding-deficient mutant (L65A), and the PDZ-dead mutant (H112A). Representative images of surface and total SEP-GluN2A in a neuron from each group, together with enlarged images of the boxed regions, are shown. Scale bars, 50 µm and 10 µm (enlarged images).(B) Quantification of the surface/total GluN2A ratio normalized to the value of control neurons co-expressing SEP-GluN2A WT and myc-SNX27 WT. Data are presented as mean ± SEM (WT-WT, n = 39 neurons; WT-L65A, n = 20; WT-H112A, n = 38; S1459A-WT, n = 40; S1459A-L65A, n = 21; S1459A-H112A, n = 39; S1459D-WT, n = 37; S1459D-L65A, n = 25; and S1459D-H112A, n = 36; from three independent cultures). ****p < 0.0001 using one-way ANOVA with a Tukey’s multiple comparison test.(C) Primary cortical neurons were transduced with lentiviral particles expressing GFP alone (control) or SNX27 shRNA at DIV9. At DIV15, surface biotinylation assays were performed in SNX27 knockdown and control neurons following 5 min of glycine stimulation (cLTP). The relative amounts of surface and total proteins were assessed by western blotting using specific antibodies against GluN2A, SNX27, and ß-actin.(D) Quantification of the surface/total ratio of GluN2A in SNX27 knockdown and control neurons following glycine stimulation. Data represent mean ± SEM of band intensities relative to their respective control values (dashed line; n = 7; from three independent experiments). *p < 0.05 using a Mann-Whitney’s test.(E) Hippocampal neurons expressing either GFP-SNX27 or GFP alone were stimulated with glycine and subjected to a PLA assay. Representative images showing the PLA signals (top) and GFP fluorescence (bottom) in a 50-µm segment of a primary dendrite of a neuron in each group. Scale bar, 10 µm.(F and G) Quantification of the number of PLA puncta in the primary dendrite of neurons expressing GFP-SNX27 (F) orGFP alone (G) following glycine stimulation. Data are presented as mean ± SEM (GFP-SNX27, control, n = 31 neurons, cLTP, n = 30; GFP, control, n = 16, cLTP, n = 14; from four independent cultures). ***p < 0.001 using a Mann-Whitney test.
Fig 4: USP32 promotes retrograde trafficking by way of the retromer complex. a Effects of VPS35 and VPS26 depletion on the size and intracellular distribution of late endosomes (LEs). Representative confocal images of fixed MelJuSo cells transfected as indicated and immunostained against major histocompatibility class II (MHC-II) (white) are shown with the corresponding immunoblot analyses; targeting small interfering RNA (siRNA): (+); control siRNA: (-). Cell and nuclear boundaries are depicted in dashed magenta and white lines, respectively. b Percent cells harboring dispersed (gray bars) and/or enlarged (white bars) MHC-II-positive vesicles in response to VPS35 and VPS26 depletion, n = 3 independent experiments. c Effects of USP32 depletion on cellular abundance of endogenous VPS35 and VPS26, as assessed by immunoblot. d, e Effects of Rab7 ubiquitylation status on the retromer compartment. d Representative confocal images of fixed MelJuSo cells transferred with the indicated siRNAs and immunostained for endogenous VPS35 (white). Boxed perinuclear (PN) and peripheral (PP) region overlays of VPS35 (magenta) with MHC-II (green) highlight retromer/LE interactions. Cell and nuclear boundaries depicted in dashed magenta and white lines, respectively. e Colocalization between VPS35 and MHC-II in control cells (white bars) vs. those depleted of USP32 (gray bars). Plots report Mander’s overlap calculated from multicell images (black circles) taken from n = 3 independent experiments. See also Supplementary Fig. 8a–e. f–h Effect of Rab7 ubiquitylation status on its interaction with VPS35 as measured by proximity-based labeling with biotin. f Biotinylation of RFP-VPS35 in the presence of free GFP-BirA (-), GFP-BirA-Rab7 (WT) vs. GFP-BirA-2KR (2KR) assayed in HEK293T cells. g Biotinylation of RFP-VPS35 by GFP-BirA-Rab7 (white bar) vs. GFP-BirA-2KR (gray bar) above GFP-BirA background control, n = 3 independent experiments. h Quantification of endogenous VPS35 and RFP-VPS35 biotinylation (combined) by GFP-BirA-Rab7 above GFP-BirA control in control HeLa cells (siCtrl, white bar) vs. those depleted of USP32 (siUSP32_2, black bar), n = 5 independent experiments. Bar graphs report mean, error bars reflect ±s.d. See also Supplementary Fig. 8f. All significant values were calculated using Student’s t test: **p < 0.05, ***p < 0.001, NS = not significant. Sale bars = 10 µm
Fig 5: GluN2A Ser-1459 phosphorylation enhances SNX27-retromer binding(A) ITC experiments comparing the binding of GluN2A C-terminal peptides, either WT (blue), or the S1459A phospho-deficient (orange), S1459E phospho-mimetic (green), or S1462E PDZ binding-deficient (purple, negative control) mutant, to the purified SNX27 PDZ domain protein. The thermodynamic parameters of binding are given in Table S1.(B) GST pull-down experiments from HEK293T cells co-transfected with plasmids encoding myc-SNX27 and GST alone or GST-GluN2A C-tails including WT, S1459A (SA), S1459E (SE), and the V1464E PDZbm mutant (PDZd). Total cell lysates (input) and bound proteins (pull-down) were resolved by SDS-PAGE and analyzed by western blotting with specific antibodies against myc, VPS35, VPS26, and GST.(C–E) Quantification of the levels of myc-SNX27 (C), VPS35 (D), and VPS26 (E) binding to GST-GluN2A C-tails. Data represent mean ± SEM of band intensities normalized to WT values (n = 3–4; from three independent experiments). *p < 0.05, ***p < 0.001, ****p < 0.0001 using one-way ANOVA with a Tukey’s multiple comparison test.(F) GST pull-down experiments revealed that the H112A mutation in myc-SNX27, which disrupts binding to PDZbm, failed to bind both WT and the S1459E phospho-mimetic GST-GluN2A C-tails in HEK293T cells.(G) HEK293T cells were co-transfected with plasmids encoding myc-SNX27, GST-GluN2A C-tails (WT or S1459A), together with pEGFP or pEGFP-tCaMKIIa (constitutively active truncated CaMKIIa). Cells were lysed and pulled down with GSH-Sepharose. Bound proteins and total lysates were analyzed by immunoblotting with specific antibodies against myc, GST, GluN2A pS1459, and GFP.(H) Quantification of the level of myc-SNX27 binding to GST-GluN2AC-tails. Data represent mean ± SEM of band intensities normalized to WT values(n = 3; from three independent experiments). *p < 0.05 using two-way ANOVA with a Sidak’s multiple comparison test.
Supplier Page from Abcam for Anti-VPS26 antibody [EPR13456]