Loss of the SNX1/2–SNX5/6 complex leads to a pronounced retrograde sorting defect of the CI-MPR. (A) Degradation assay of endogenous CI-MPR. HeLa cells were transfected with nontargeting siRNA, siRNA against VPS35, or a pool of siRNAs against SNX1, SNX2, SNX5, and SNX6. 72 h after transfection, cells were incubated with 10 µg/ml cycloheximide and lysed at different time points as indicated. The level of endogenous CI-MPR was analyzed by quantitative Western blotting. Molecular masses are given in kilodaltons. (A, top right) n = 4 independent experiments (one-way ANOVA compared with nontargeting control). (A, bottom right) n = 4 independent experiments (two-way ANOVA compared with nontargeting control). (B) CI-MPR–CD8 uptake assays in retromer and retromer-linked SNX-BAR–knockdown cells show a retrograde trafficking defect only in SNX1/2–SNX5/6–depleted cells. HeLa cells were transfected with nontargeting siRNA, siRNA against VPS35, or a pool of siRNAs against SNX1, SNX2, SNX5, and SNX6. 72 h after transfection, cells were incubated with αCD8 antibody, and its trafficking was followed for 60 min. The retrograde transport to the TGN was assayed through measuring colocalization of CD8 signal with the TGN marker TGN46. n = 3 independent experiments; 0 min, ≥72 cells per condition; 10 min, ≥74 cells per condition; 20 min, ≥78 cells per condition; 60 min, ≥69 cells per condition (means ± SEM; two-way ANOVA compared with nontargeting control; *, P < 0.05; **, P < 0.01; ****, P < 0.0001). Bars: (main images) 20 µm; (insets) 5 µm.