Table S1. siRNA-mediated knockdown of human phosphatase genes alters cellular PtdIns3P. U2OS EGFP−2xFYVE cells were transfected with siRNAs targeting human phosphatase genes for 48 hours (four siRNA sequences per gene per well; all four sequences displayed). Following knockdown, EGFP−2xFYVE signal and distribution was visualized by confocal microscopy and scored from −100 (decreased punctae from control cells) to +100 (increased punctae) and means determined. Select genes were validated using multiple unique siRNA sequences and their efficacy was incorporated into their scores (see the Materials and Methods). Genes were ranked based on their scores from 1 (most increased EGFP−2xFYVE) to 206 (most decreased).

Fig. S1. Target genes are effectively knocked down by siRNA. (A) Western blot analysis of whole cell extracts following transfection with control or VPS34 siRNA shows depletion of VPS34 protein levels. α-tubulin was probed as a loading control. (B) U2OS EGFP−2xFYVE cells were transfected with MTMR6 siRNA, fixed, and imaged at 60× by fluorescent microscopy as in Fig. 1. Scale bar: 10 m. (C) MTMR6 mRNA was depleted by 96% following transfection with MTMR6 siRNA for 48 hours. RNA extracted from cells transfected with control or MTMR6 siRNA was converted to cDNA and MTMR6 levels determined by qRT-PCR using gene-specific primers. Values were normalized to GAPDH. (D-I) U2OS EGFP−2xFYVE cells were transfected with control (D) or PTPRS siRNA E, siRNA-1; F, siRNA-2; G, siRNA-3; H, siRNA-4; I, siRNA-pool (1-4) for 48 hours, fixed, and imaged by fluorescent microscopy green: PtdIns3P, EGFP−2xFYVE; blue: nuclei). Insets are 2× magnifications of boxed regions, highlighting the abundant vesicles caused by PTPRS siRNA transfection. Scale bars: 10 m. (J) EGFP−2xFYVE punctae were quantified from cells following PTPRS knockdown with four unique siRNAs. Values represent means and bars represent s.e.m. (K) PTPRS mRNA knockdown following transfection for 48 hours with four unique siRNAs (individually or pooled) was determined by qRT-PCR using gene-specific primers and normalization to GAPDH, as described above.

Fig. S2. Knockdown of PTPsigma increases the abundance of autophagic, but not endocytic, vesicles. (A) U2OS cells were transfected with control or PTPRS siRNA, fixed and immunostained with anti-EEA1 antibodies (see Fig. 1H). EEA1-positive vesicles were quantified using image analysis software. Bars represent s.e.m. (B) U2OS cells transfected with control (left panels) or PTPRS (right panels) siRNAs were cultured for 1 hour with nutrient-rich medium (top panels) or 50 nM rapamycin (bottom panels). Cells were stained with anti-ATG12 antibodies and imaged by fluorescent microscopy at 60× (green: ATG12; blue: nuclei). Scale bars: 10 m. (C) U2OS cells transfected with control or PTPRS siRNA were cultured for 1 hour with normal growth medium (full nutrients; left), 25 M chloroquine in normal growth medium (middle) or 50 nM rapamycin and 25 M chloroquine in normal growth medium (right). Cells were fixed, immunostained with anti-LC3B antibodies, and imaged by fluorescent microscopy at 60×. LC3-positive punctae were quantified images using image analysis software (black, control siRNA; white, PTPRS siRNA). Bars represent s.e.m.; *P<0.05, **P<0.01.

Fig. S3. FL-PTPsigma colocalization with mRFP−LC3 and mock control for FL-PTPsigma immunofluorescence. (A) U2OS EGFP−2xFYVE cells were mock transfected (with transfection reagent but no DNA) for 24 hours, stained with PTPsigma (anti-D1) antibodies, and imaged as in Fig. 4A−C, Fig. 5. Absence of signal in the red channel demonstrates specificity of FL-PTPsigma expression captured in the above figures green, PtdIns3P; red, anti-PTPsigma (D1-targeted antibodies); blue, nuclei. Scale bars: 10 m. (B) FL-PTPsigma was transiently expressed in U2OS mRFP−LC3 cells and LC3 and PTPsigma imaged by fluorescent microscopy following 2 hour incubation with full nutrient medium (top panels) or amino acid starvation medium (lower panels) red: mRFP−LC3; green: PTPsigma (D1-targeted antibodies); blue: nuclei. Insets are 2× magnifications of boxed regions. White arrows indicate punctae positive for both PTPsigma and LC3.

Fig. S4. PTPsigma dephosphorylates phosphotyrosine, but not PtdIns3P, in vitro. (A) Recombinant GST fusions of PTP1B (left), PTPsigma (middle) and MTMR6 (right) were incubated with a phosphotyrosine peptide at the indicated concentrations for 15 minutes at 37°C and released phosphates measured by malachite green quenching and 650 nm absorbance. (B) Recombinant GST fusions of PTP1B (left), PTPsigma (middle) and MTMR6 (right) were incubated with water-soluble PtdIns3P substrate at the indicated concentrations and released phosphates measured by malachite green quenching and 650 nm absorbance. Bars represent s.d.

Movie 1. PTPsigma knockdown increases PtdIns3P vesicles abundance similarly to AA starvation. U2OS EGFP−2xFYVE cells cultured on number 1.5 coverglasses were imaged in the green (PtdIns3P; EGFP−2xFYVE) and DIC channels every 30 seconds for 15 minutes. Cells were transfected with control siRNA for 48 hours. 2.5 hours before imaging, cells were given fresh full nutrient medium.

Movie 2. PTPsigma knockdown increases PtdIns3P vesicles abundance similar to AA-starvation. Cells were transfected with control siRNA for 48 hours. 2.5 hours before imaging, cells were were starved of amino acids.

Movie 3. PTPsigma knockdown increases PtdIns3P vesicles abundance similar to AA-starvation. Cells were transfected with PTPRS siRNA for 48 hours. 2.5 hours before imaging, cells were given fresh full nutrient medium.