Fig 1.
CI-MPR directly interacts with the PX domain of SNX5 or SNX6, but not retromer or the SNX3-retromer complex.
(A) GST-CI-MPR (aa1–164) or GST pull-down of purified retromer VPS35/VPS26/VPS29, and SNX1ΔN/SNX6. Shown is a Coomassie blue–stained SDS-PAGE gel of purified proteins used (left) and bound samples (right). The SUMO fusion of SNX6 was uncleaved in order to distinguish the SNX6 protein bands from those of SNX1ΔN or GST-CI-MPR. (B) GST-CI-MPR or GST pull-down of purified SNX1ΔN/SNX6, SNX1ΔN/SNX6-F149D, MBP-SNX5PX. Shown is a Coomassie blue–stained SDS-PAGE gel of purified proteins used (left) and bound samples (right). (C) GST-CI-MPR pull-down of purified MBP-SNX5PX. Shown are the CI-MPR constructs used (top) and a Coomassie blue–stained gel of purified proteins used and bound samples (bottom). (D) Isothermal titration calorimetry of CI-MPR (aa21–48, blue) or DMT1-II (green) titrated into SNX5PX in a buffer containing 100 mM Hepes (pH 7.5), 300 mM NaCl, 2 mM βME at 25°C. Top and bottom panels show raw and integrated heat from injections, respectively. The black curve in the bottom panel represents a fit of the integrated data to a single-site binding model. Experiments were triplicated, and the numerical data are included in S1 Data. (E) Isothermal titration calorimetry of CI-MPR (aa21–48, blue) or DMT1-II (green) titrated into in the SNX3-retromer complex under the conditions identical to (D). Top and bottom panels show raw and integrated heat from injections, respectively. The black curve in the bottom panel represents a fit of the integrated data to a single-site binding model. (F) Identification of the CI-MPR-binding residues of SNX5PX by NMR. Residues showing significant change of chemical shift or line broadening with the addition of the CI-MPR peptide (aa21–48) are displayed as sticks. CI-MPR, cation-independent mannose 6-phosphate receptor; DMT1-II, divalent metal transporter 1; GST, glutathione-S-transferase; MBP, maltose binding protein; MW, molecular weight; NMR, nuclear magnetic resonance; PX, phox-homology; SNX, Sorting Nexin family; SNX5PX, PX domain of SNX5; VPS, vacuolar protein sorting.
Fig 2.
Cooperative binding of an SBM within the CI-MPR tail and PtdIns(3)P recruits SNX-BAR to a membrane.
(A) Sequence alignment of the known SNX-BAR interactors. Red color indicates residues conserved among the interactors; blue color indicates a His residue with the third position of INSR, which does not bind to the SNX-BAR proteins. The number before the first residue of each sequence indicates the position of this residue in the cytoplasmic tail. Secondary structure of IncE is listed below the sequences. The AP-2 binding motif within CI-MPR (YSKV) is indicated with a box. (B) Isothermal titration calorimetry of CI-MPR (aa21–48) WT or mutants titrated into SNX5PX in a buffer containing 100 mM Hepes (pH 7.5), 300 mM NaCl, 2 mM βME at 25°C. Raw data are listed at the top; integrated normalized data are listed at the bottom. Colors indicate the peptides used in the experiment. Experiments were triplicated, and the numerical data are included in S1 Data. (C) Steady-state localization of Venus-CI-MPR WT and mutants in HeLa cells. Cells were transfected with Venus-CI-MPR constructs (green) and stained with phalloidin (red). Representative images are shown. Scale bar: 10 μm. (D) Colocalization of Venus-CI-MPR and phalloidin in cells in D. Each dot represents Pearson’s correlation coefficients (r) from one cell. Experiments were triplicated, and the numerical data are included in S1 Data. ****P < 0.0001. P values were compared with WT using one-way ANOVA, Tukey's multiple comparisons test, throughout the paper unless otherwise indicated. (E) Binding of the SNX1ΔN/SNX6 complex to liposomes in a liposome flotation assay. Liposomes with or without PtdIns(3)P were incubated with purified SNX1ΔN (aa140–C)/SNX6 complex in the presence or absence of His-CI-MPR tail and then centrifuged in a Histodenz gradient. Samples that floated to the top of the gradient were collected and subjected to SDS-PAGE. Gel image from one of three independent experiments is shown. Total intensity of SNX1 and SNX6 protein bands was quantified using ImageJ and normalized to the level of sample 4 in each experiment. Error bars represent standard deviation. The numerical data are included in S1 Data. (F) Binding of the SNX1ΔN/SNX6 complex WT or mutants to liposomes in a liposome flotation assay. Liposomes with PtdIns(3)P were incubated with purified SNX1ΔN (aa140–C)/SNX6 complex WT or mutants in the presence of His-CI-MPR tail. Shown is the CBB-stained SDS-PAGE for samples floated to the top of the gradient. Intensity of protein bands is normalized to the level of WT proteins. Statistic data represent the results from n = 3 independent experiments and are expressed as mean ± SD. The numerical data are included in S1 Data. AP-2, adaptor protein complex 2; BAR, Bin/Amphiphysin/Rvs; CBB, Coomassie brilliant blue; CI-MPR, cation-independent mannose 6-phosphate receptor; IGF1R, Insulin-like growth factor 1 receptor; INSR, insulin receptor; MW, molecular weight; PtdIns(3)P, phosphatidylinositol 3-phosphate; SBM, SNX-BAR-binding motif; SEMA4C, semaphorin 4C; SNX, Sorting Nexin family; WT, wild type.
Fig 3.
Dual recognition by SNX-BAR and SNX27 mediates transport of SEMA4C.
(A) Model showing dual recognition of SEMA4C by SNX-BAR and the SNX27-retromer complex. The PX domains of SNX1/SNX2 and SNX5/SNX6 interacts with membrane-bound PtdIns(3)P and ligands like SEMA4C, respectively. SNX27 harbors a PDZ domain, followed by a PX and FERM domain. Its PDZ and FERM domain bind to the PDZbm within SEMA4C, and SNX1/SNX2, respectively. For simplicity, the PX domain of SNX27 is omitted. (B) GST, GST-SEMA4C-tail (aa1–149), or GST-SEMA4C-Δ4 (aa1–145) pull-down of purified MBP-SNX5PX, or SNX27PDZ, or the mixture of MBP-SNX5PX and SNX27PDZ. Shown are Coomassie blue–stained SDS-PAGE gels of purified proteins used (left) and bound samples (right). (C) GST, GST-SEMA4C (aa47–71) WT or mutants (Y3Y5, L21I23), or GST-SEMA4C (aa47–65) pull-down of purified MBP-SNX5PX. Shown are Coomassie blue–stained SDS-PAGE gels of purified proteins used (left) and bound samples (right). (D) Control, SNX1+2-KO, SNX5+6-KO, and SNX27-KO HeLa cells were transiently transfected with plasmids encoding CD8A-SEMA4C. Cells were incubated with monoclonal anti-human CD8A antibody on ice for 30 min. Unbound antibodies were removed, and the internalization of antibody-bound CD8A was carried out in DMEM at 37°C for 3 h. The internalized CD8A–antibody was detected using Alexa-488 secondary antibodies, with lysosomes stained with LAMP1 (red). Scale bar: 10 μm. (E) Quantification of CD8A/LAMP1 colocalization in cells in D. Each dot represents Pearson’s correlation coefficients from one cell. N = 3 independent experiments. P values were calculated using one-way ANOVA, Tukey's multiple comparisons test. *P < 0.05; ***P < 0.001. The numerical data are included in S1 Data. (F) HeLa cells were transiently transfected with plasmids encoding CD8A-SEMA4C WT, L21I23, Δ4, and Δ4-L21I23. Cells were incubated with monoclonal anti-human CD8A antibody on ice for 30 min. Unbound antibodies were removed, and the internalization of antibody-bound CD8A was carried out in DMEM at 37°C for 1 h. The internalized CD8A–antibody was detected using Alexa-488 secondary antibodies, with early endosomes stained with EEA1 (red). Scale bar: 10 μm. (G) Quantification of CD8A/EEA1 colocalization in cells in F. Each dot represents Pearson’s correlation coefficients from one cell. N = 3 independent experiments. P values were calculated using one-way ANOVA, Tukey's multiple comparisons test. **P < 0.01; ****P < 0.0001. The numerical data are included in S1 Data. BAR, Bin/Amphiphysin/Rvs; DMEM, Dulbecco’s modified Eagles medium; EEA1, early endosome antigen 1; GST, glutathione-S-transferase; KO, knockout; LAMP1, lysosomal-associated membrane protein 1; MBP, maltose binding protein; PDZbm, PDZ-binding motif; PtdIns(3)P, phosphatidylinositol 3-phosphate; PX, phox-homology; SEMA4C, semaphorin 4C; SNX, Sorting Nexin family; SNX5PX, PX domain of SNX5; WT, wild type; VPS, vacuolar protein sorting.
Fig 4.
Identification and experimental verification of additional cargo proteins of SNX-BARs.
(A) Schematic representation of different functions of SBM-containing proteins. Seventy-one human SBM-containing proteins were manually classified into six categories according to literature, including receptor (GPCR and non-GPCR), enzyme, channel, transporter, cadherin, and others (top), or were analyzed by GO and classified by their biological processes. (B) GST-CI-MPR, PTHR, TRAILR1, STIMA, CED-1, JAM3, or GST pull-down of purified MBP-SNX1/His6-sumo-SNX5. Shown are a Coomassie blue–stained SDS-PAGE gel of purified proteins (bottom) and immunoblot using anti-MBP antibody for the same sample (top). CED-1 does not have a SBM and is labeled as gray. (C) GST-TRAILR1 and GST-PTHR WT or mutants pull-down of purified MBP-SNX1//His6-sumo-SNX5. Shown are a Coomassie blue–stained SDS-PAGE gel of purified proteins (bottom) and immunoblot using anti-MBP antibody for the same sample (top). (D) Immunofluorescence analysis of internalized TRAILR1 (green) in control, SNX1+2-KO, SNX5+6-KO, and SNX27-KO HeLa cells, with endosomes stained with an anti-EEA1 antibody (red). Cells were incubated with antibody against the ectodomain of TRAILR1 and the lysosomal protease inhibitor leupeptin for 6 h. Cells were then fixed, permeabilized, and stained with antibodies against EEA1 and the internalized TRAILR1 antibody. (E) Colocalization of internalized TRAILR1 antibody and EEA1 in cells in D. Each dot represents Pearson’s correlation coefficients from one cell. ****P < 0.0001. Experiments were triplicated, and the numerical data are included in S1 Data. (F) Immunofluorescence analysis of internalized TRAILR1 (green) in control, SNX1+2-KO, SNX5+6-KO, and SNX27-KO HeLa cells, with lysosome stained with an anti-LAMP1 antibody (red). Cells were incubated with antibody against the ectodomain of TRAILR1 and the lysosomal protease inhibitor leupeptin for 6 h. Cells were then fixed, permeabilized, and stained with antibodies against LAMP1 and the internalized TRAILR1 antibody. (G) Colocalization of internalized TRAILR1 antibody and LAMP1 in cells in F. Each dot represents Pearson’s correlation coefficients from one cell. *P < 0.05; ****P < 0.0001. Experiments were triplicated, and the numerical data are included in S1 Data. (H) Steady-state localization of Venus-TRAILR1 WT and mutants in HeLa cells. Cells were transfected with Venus-TRAILR1 constructs (green) and stained with phalloidin (red). Representative images are shown. (I) Colocalization of Venus-TRAILR1 and phalloidin in cells in H. Each dot represents Pearson’s correlation coefficients from one cell. ****P < 0.0001. Experiments were triplicated, and the numerical data are included in S1 Data. BAR, Bin/Amphiphysin/Rvs; CED-1, Cell death abnormality protein 1; CI-MPR, cation-independent mannose 6-phosphate receptor; EEA1, early endosome antigen 1; GO, Gene Ontology; GPCR, G protein–coupled receptor; GST, glutathione-S-transferase; KO, knockout; LAMP1, Lysosomal-associated membrane protein 1; MBP, maltose binding protein; PTHR, Parathyroid hormone/parathyroid hormone-related peptide receptor; SBM, SNX-BAR-binding motif; SNX, Sorting Nexin family; STIMA, STIM Activating Enhancer; TRAILR1, TNF-related apoptosis-inducing ligand receptor 1; WT, wild type.
Fig 5.
Depletion of SNX-BARs or SNX27 attenuates TRAIL-induced apoptosis.
(A) Cell morphology of control, SNX1+2-KO, SNX5+6-KO, and SNX27-KO HeLa cells treated with different concentrations of GST-Trail (0, 50, and 100 ng/mL) for 24 h. (B) Cells were treated with GST-Trail (0 or 100 ng/mL) for 24 h, subjected to Annexin V/PI staining, and analyzed by flow cytometry. (C) Quantification of increased apoptotic cells upon the treatment of GST-TRAIL (apoptotic cell % with GST-TRAIL − apoptotic cell % without GST-TRAIL), as shown in B. Statistic data represent the results from n = 5 independent experiments and are expressed as mean ± SEM *P < 0.05; ***P < 0.001. The numerical data are included in S1 Data. (D) Protein lysates from control, SNX1+2-KO, SNX5+6-KO, and SNX27-KO KO HeLa cells treated with different concentrations of GST-Trail. Representative blot from n = 4 independent experiments. (E-F) Quantification of c-PARP upon the treatment of GST-TRAIL at a concentration of 50 ng/mL (E) or 100 ng/mL (F). Percentage of c-PARP was calculated using the band intensity of c-PARP divided by the total bond intensity of cleaved and uncleaved PARP. Statistic data represent the results from n = 4 independent experiments and are expressed as mean ± SD. *P < 0.05; ***P < 0.001. The numerical data are included in S1 Data. BAR, Bin/Amphiphysin/Rvs; c-PARP, cleaved PARP; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; GST, glutathione-S-transferase; KO, knockout; ns, not significant; PARP, Poly(ADP-Ribose) Polymerase; PI, propidium iodide; SNX, Sorting Nexin family; TRAIL, TNF-related apoptosis-inducing ligand.
Fig 6.
Model showing how sorting nexins mediate cargo transport.
SNX-BARs and SNX3-retromer mediate the endosome-to-TGN transport of CI-MPR and DMT1/wntless, respectively. SNX17 and retriever mediate the recycling of transmembrane cargo proteins, such as ITGB1, to the plasma membrane. Finally, both SNX27-retromer and SNX-BARs are involved in the endocytic transport of a variety of cargo, likely through cooperating with each other, although definitive evidence is still lacking. BAR, Bin/Amphiphysin/Rvs; CI-MPR, cation-independent mannose 6-phosphate receptor; DMT1, divalent metal transporter 1; GLUT1, Glucose Transporter Type 1; ITGB1, Integrin Subunit Beta 1; PTHR, Parathyroid hormone/parathyroid hormone-related peptide receptor; SEMA4C, semaphorin 4C; SNX, Sorting Nexin family; TGN, trans-Golgi network; TRAILR1, TNF-related apoptosis-inducing ligand receptor 1.