Figures and data

Effect of WIPI2 knockdown on EGFR degradation.
A. B. Control and WIPI2KD cells were serum-starved for 24 h and then supplemented with EGF (100 ng/ml). After the indicated periods of time, cells were fixed, permeabilized, DAPI-labeled (blue) and decorated with antibodies to EGFR (magenta), EEA1 or LAMP1 (green). Scale bars: 10 μm. C. Colocalization of EGFR with EEA1 or LAMP1 (white) was quantified over time using the images from A and B and Manders’ correlation coefficients were calculated. M1 indicates the fraction of magenta pixels overlapping with green pixels. Values are the mean ± s.d.; (n = 3). 150 cells, stemming from 3 independent biological experiments, were quantified per sample. D. EGFR degradation. Control and WIPI2KD cells were stimulated with EGF for the indicated periods of time, lysed and subjected to SDS-PAGE and Western blot analysis for EGFR. α-Tubulin served as a loading control. E. Quantification of EGFR from (D), using the value of control cells at time 0 (cells starved for 24 h) as 100% reference. Data are means ±s. d., from three independent experiments. Data points form these three experiments are indicated for each time point.

GLUT1 expression and localization upon WIPI2 knockdown.
A. GLUT1 cell surface exposure. Control and WIPI2KD cells were fixed and stained with antibody against GLUT1 and with DAPI. Where indicated, cells had been permeabilized with 0.05% before staining to reduce plasma membrane staining and provide better access to GLUT1 inside the cell. Scale bars: 10 μm. B. Quantification of GLUT1-immunofluorescence in cells from A. Regions of interest (ROIs) corresponding to each cell and in some regions outside the cells (background) were manually defined using image J software. Total cell fluorescence was integrated, and the background fluorescence was subtracted. The resulting total cell fluorescence was divided by the area of the cell. This value is shown in the graph. 150 cells per condition, stemming from three independent experiments, were quantified. Values of individual cells and the means are presented by smaller and larger circles, respectively, coloured according to the independent experiment that generated them. P values were calculated applying an unpaired Student’s t-test with unequal variances. The analysis was performed with 99% confidence. NS = not significant (P > 0.05).

Impact of WIPI2 knockdown on β1-Integrin localization.
A. β1-Integrin cell surface expression. Control and WIPI2KD cells were fixed and stained with antibody to β1-Integrin (green) and with DAPI (blue), without detergent permeabilization. Scale bars: 10 μm. B. Quantification of β1-Integrin immunofluorescence in cells from A. Regions of interest (ROIs) corresponding to each cell and to areas outside the cells (background) were manually defined using ImageJ software. Total cell fluorescence was integrated, and the background fluorescence was subtracted. The resulting total cell fluorescence was divided by the area of the cell. This value is shown in the graph. 150 cells per condition, stemming from three independent biological experiments, were quantified. Individual cells and means are presented by smaller and larger circles, respectively, coloured according to the independent experiment that they stem from. P values were calculated through an unpaired Student’s t-test with unequal variances.The analysis was performed with 99% confidence. ***P < 0.001. C.D. Immunofluorescence staining of intracellular β1-Integrin (green) and EEA1 (magenta) or LAMP1 (magenta) in control and WIPI2KD cells. Overlaps are marked in white. Cells had been fixed, permeabilized with 0.05% saponin and stained with antibodies to the indicated proteins. Scale bars: 10 μm. E.G. Colocalization of β1-Integrin with LAMP1 or EEA1 was measured in cells from C, using Manders’ colocalization coefficient M2, calculated in Image J software. It indicates the fraction of green pixels overlapping with the magenta pixels. Colocalization was quantified from three independent experiments with a total of 150 cells per condition. Values of individual cells and means are presented by smaller and larger circles, respectively, coloured according to the independent experiment that they stem from. An unpaired Student’s t-test with unequal variances was used to calculate P-values. The analysis was performed with 99% confidence. ***P < 0.001. F.H. Quantification of LAMP1- or EEA1-immunofluorescence in cells from C or D, respectively, was performed as in B. 120 cells per condition, stemming from three independent biological experiments, were quantified. Individual cells and means are presented by smaller and larger circles, respectively, coloured according to the independent experiment that they stem from. Data are means ± s.d. P-values were calculated applying an unpaired Student’s t-test with unequal variances. The analysis was performed with 99% confidence. NS = not significant (P > 0.05).

Role of the amphipathic α-helix of WIPI2 in β1-Integrin sorting.
A. The amphipathic α-helix. Helical wheel projections showing the CD-loop on blade 6 of the wildtype sequence, WIPI2WT, and WIPI2Sloop. Coloured arrows indicate the two pairs of amino acids that have been swapped in WIPI2Sloop. The magnitude and direction of the hydrophobic moment of the helices was predicted using the online tool Heliquest (Gautier et al., 2008). It is indicated by the vector in the centre of the wheels. Sequences of the hydrophobic loop region of WIPI2WT and WIPI2Sloop are shown, and predicted α-helices are plotted in magenta. The two pairs of hydrophobic/hydrophilic amino acids that are swapped in WIPI2Sloop are highlighted by rectangles in the sequences. B.C. β1-Integrin localisation. WIPI2KD (B) and control (C) cells were transfected with a plasmid carrying siRNA-resistant EGFPWIPI2WT or EGFPWIPI2Sloop. After 18h of viral transfection, cells were fixed, permeabilized with 0.05% saponin and stained with DAPI and antibodies to β1-Integrin. Dashed lines indicate the circumference of untransfected cells, while transfected cells are indicated by asterisks. Scale bars: 10 μm. D. Quantification of β1-Integrin immunofluorescence in cells from B and C. Regions of interest (ROIs) corresponding to cells expressing the indicated WIPI2 variants, and some regions outside the cells (background), were manually defined using Image J software. Total cell fluorescence was integrated and the background fluorescence was subtracted. The resulting total cell fluorescence was divided by the area of the cell. This value is shown in the graph. 105 cells per condition, stemming from three independent biological experiments, were quantified. Values of individual cells and means are presented by smaller and larger circles, respectively, coloured according to the independent experiment that generated them. P-values were calculated applying an unpaired Student’s t-test with unequal variances. The analysis was performed with 99% confidence: ****P < 0.00001. NS: not significant (P>0.05)

Effects of the WIPI2 FSSS motif on β1-Integrin recycling.
A.B. Influence of WIPI2 variants on β1-Integrin. WIPI2-knockdown (WIPI2KD, A) and control (B) HK2 cells were transfected with a plasmid expressing siRNA-resistant wildtype or the indicated FSSS variants of EGFPWIPI2. After 18 h of viral transfection, cells were fixed (without detergent permeabilization) and stained with DAPI and antibodies to β1-Integrin (red). Dashed lines indicate the circumference of untransfected cells, while transfected cells are indicated by asterisks. Scale bars: 10 μm. C. D . Quantification of β1-Integrin immunofluorescence in cells from A and B was performed as in Figure 4 D. 105 cells per condition, stemming from three independent biological experiments, were quantified. Individual cells and means are presented by smaller and larger circles, respectively, coloured according to the independent experiment that they stem from. P-values were calculated applying an unpaired Student’s t-test with unequal variances. The analysis was performed with 99% confidence: ****P < 0.00001.

Integration of WIPI2 with coat subunits.
A. Interaction of WIPI2HA with the CCC complex subunit CCDC93. HK2 cells and HK2 cells stably expressing WIPI2HA were detergent solubilised, the total cell extracts were incubated with anti-HA beads and washed. Adsorbed protein was analysed by SDS-PAGE and Western blotting using the indicated antibodies. The intensity of the interacting CCDC93 was quantified with a LICOR Odyssey fluorescence imager. The background from the corresponding position in the sample from cells without HA-tag was subtracted. The resulting intensity is shown relative to the intensity of WIPI2HA signal on the beads. N = 3 biological replicates from independent experiments. Red bars show the means. Error bars represent the SEM. P values were calculated applying an unpaired Student’s t-test with unequal variances. *P < 0.01. B. Retriever-dependent interaction of WIPI2HA and SNX17. HK2 cells stably expressing WIPI2HA were treated with siRNA to VPS26C (VPS26CKD), or with non-specific siRNA (VPS26CCT), and lysed. The total cell extracts were incubated with anti-HA beads and adsorbed proteins were analysed using the indicated antibodies as in A. Values shown stem from 3 independent experiments. C. D. Selectivity for Retriever versus Retromer. The co-immunoadsorption experiments were performed as in A and analysed for co-adsorbed (C) VPS26C (Retriever) or (D) VPS26 (Retromer). N=3 independent experiments. *P < 0.01.**P < 0.001. NS=not significant.

Interaction of WIPI1 with VPS26 and SNX27, but not VPS26C
A. Interaction of WIPI1HA with VPS26. HK2 cells and HK2 cells stably expressing WIPI1HA were detergent-solubilised, the total cell extracts were incubated with anti-HA beads and washed. Adsorbed protein was analysed by SDS-PAGE and Western blotting using the indicated antibodies. The intensity of the interacting VPS26 was quantified with a LICOR Odyssey fluorescence imager. The background from the corresponding position in the sample from cells without HA-tag was subtracted. The resulting intensity is shown relative to the intensity of the WIPI1HA signal on the beads. N = 3 biological replicates from independent experiments. Red bars show the means and error bars represent the SEM. P values were calculated applying an unpaired Student’s t-test with unequal variances. B. Lack of interaction of WIPI1HA with VPS26C. HK2 cells stably expressing WIPI1HA were used for co-immunoadsorption experiments as in A and decorated with the indicated antibodies. N = 3. C. Interaction of WIPI1HA with SNX27. HK2 cells stably expressing WIPI1HA were used for co-immunoadsorption experiments as in A and decorated with the indicated antibodies. N = 3. **P < 0.001. NS: not significant.

Impact of the FSSS motif on the WIPI2-Retriever interaction.
A. HK2 cells and HK2 cells stably expressing the indicated WIPI2HA variants were detergent solubilised. Anti-HA beads were incubated with the total cell extracts, washed, and adsorbed proteins were analysed by SDS-PAGE and Western blotting using the indicated antibodies. B. Band intensities from the blots in A were quantified with a LICOR Odyssey infrared fluorescence imager and plotted as the ratio of VPS26C over WIPI2HA. N=3 independent experiments were quantified. Red bars show the means and error bars represent the SEM. P values were calculated applying an unpaired Student’s t-test with unequal variances. **P < 0.001.

Role of the WIPI2 FSSS motif for recruiting WIPI2 and Retriever to Rab11 endosomes.
WIPI2-knockdown HK2 cells were transfected with plasmids expressing the indicated siRNA-resistant EGFPWIPI2 variants and mCherry-RAB5, mCherry-RAB7, or DsRed-Rab11. Cells were fixed, permeabilized, immuno-stained for VPS35L and analyzed by confocal microscopy. Examples of the quantified images are presented in Supplementary Figures S9 to S11. Colocalization with EGFPWIPI2 and VPS35L was assessed for: A. RAB11 B. RAB5 C. RAB7. Colocalization was quantified using Manders’ colocalization coefficient M2, calculated in ImageJ. M2 refers to the fraction of VPS35L colocalising with the different RAB-proteins. For the triple colocalizations WIPI2/VPS35L/RAB, M2 indicates the fraction of green pixels (WIPI2) overlapping with the pixels positive for the VPS35L/RAB colocalization. Colocalization was quantified from three independent biological experiments in a total of 120 cells. Individual cells. Individual cells and means are presented by smaller and larger circles, respectively, coloured according to the independent experiment that they stem from. P-values were calculated by an unpaired Student’s t-test with unequal variances. The analysis was performed with 99% confidence. **P < 0.001; ***P < 0.0001. NS = not significant (P>0.05).

Efficiency of WIPI2 knockdown.
Lysates of HK2 cells (50 μg per sample) were treated with non-target siRNA (Control) or siRNA against WIPI2 (WIPI2KD). HK2 cells were analyzed by SDS-PAGE and Western blot against the indicated proteins. α-tubulin was used as a loading control. A representative blot is shown. The signals were quantified on a LICOR Odyssey infrared fluorescence image WIPI2/a-Tubulin ratios were calculated. Red bars indicate the mean and error bars the SEM; n=3 independent biological experiments. P-values were calculated by an unpaired Student’s t-test with unequal variances. ***P < 0.0001.

Surface expression of β1-Integrin in cells depleted of WIPI1.
A. Control and WIPI1KD cells were fixed and stained with DAPI and with antibody to β1-Integrin. Scale bars: 10 μm. B. Quantification of β1-Integrin-immunofluorescence in cells from A. Regions of interest (ROIs) corresponding to each cell and in some regions outside the cells (background) were manually defined using ImageJ software. Total cell fluorescence was integrated, and the background fluorescence was subtracted. The resulting total cell fluorescence was divided by the area of the cell. This value is shown in the graph. 150 cells per condition, stemming from three independent biological experiments, were quantified. Individual cells and means are presented by smaller and larger circles, respectively, coloured according to the independent experiment that they stem from. P values were calculated applying an unpaired Student’s t-test with unequal variances The analysis was performed with 99% confidence. NS = not significant (P>0.05).

Expression levels of WIPI2 variants.
A. FSSS variants. Cell lysates from Control and WIPI2KD cells were transfected for 18 h with siRNA resistant constructs for EGFPWIPI2WT, EGFPWIPI2S67A and EGFPWIPI2S67E. The cells were analyzed by SDS–PAGE and Western blotting with antibodies to WIPI2 and α-Tubulin. A representative blot is shown. B. Quantification of the signals for transiently expressed EGFPWIPI2 variants from A and for endogenous WIPI2. Blots from three independent biological experiments were quantified on a Licor Odyssey infrared scanner. Graphs on the right side show the mean and SEM. P-values were calculated using an unpaired Student’s t-test with unequal variances. NS: not significant. C. SLoop variant. EGFPWIPI2 and EGFPWIPI2SLoop were expressed and analyzed by Western blotting as in described in A. D. Blots from C were quantified as in B.

EGFPWIPI2 variants are preferentially associated with recycling endosomes
A.B. Colocalization with RAB proteins. HK2 cells were transiently transfected with plasmids carrying the indicated EGFPWIPI2 variants with mCherry-RAB5 or RFP-RAB4. After 18h transfection, cells were analysed by live cell confocal microscopy. Arrows point to tubular structures that are readily detectable in cells expressing the SLoop and S67 variants of WIPI2. Scale bars: 10 μm. Insets show enlargements of the outlined areas with a white dashed line. Scale bars: 2 μm. C. Colocalization analysis of the images from (A-B). Quantification was carried out for the tubular structures and the entire cell using the Manders’ colocalization coefficient M2, indicating the fraction of green pixels overlapping with magenta pixels, calculated in Image J software. Colocalization was quantified from three independent experiments with a total of 90 cells per condition. Values of individual cells and means are presented by smaller and larger circles, respectively, coloured according to the independent experiment that they stem from. An unpaired Student’s t-test with unequal variances was used to calculate P-values. The analysis was performed with 99% confidence. NS = not significant (P>0.05).

WIPI2 colocalization with Retriever, CCC complex and Retromer
A. WIPI2 colocalization with SNX17, CCDC93, VPS26C and VPS26. HK2 cells were fixed 18 h after transient transfection with EGFPWIPI2, stained with the indicated antibodies and imaged by confocal microscopy. Scale bars: 10 μm. Insets show enlargements of the outlined areas with a white dashed line. Scale bars: 2 μm. B. Quantification. Using the images from (A), colocalization was assessed using Manders’ colocalization coefficient M2, calculating the fraction of green pixels overlapping with magenta pixels. 75 cells per condition, stemming from three independent biological experiments, were quantified. Values of individual cells and means are presented by smaller and larger circles, respectively, coloured according to the independent experiment that generated them. P values were calculated applying an unpaired Student’s t-test with unequal variances. The analysis was performed with 99% confidence. ****P < 0.00001.

WIPI1 colocalizes with VPS26, but not with VPS26C
A. WIPI1 colocalization with VPS26, SNX27, and VPS26C. HK2 cells were fixed 18 h after transient transfection with EGFPWIPI1, stained with the indicated antibodies and imaged by confocal microscopy. Scale bars: 10 μm. Insets show enlargements of the outlined areas with a white dashed line. Scale bars: 2 μm. B. Quantification. Using the images from (A), colocalization was assessed using Manders’ colocalization coefficient M2, indicating the fraction of green pixels overlapping with magenta pixels, calculated in Image J software. 75 cells per condition, stemming from three independent biological experiments, were quantified. Individual cells and means are presented by smaller and larger circles, respectively, coloured according to the independent experiment. P values were calculated applying an unpaired Student’s t-test with unequal variances. The analysis was performed with 99% confidence. ****P < 0.00001.

WIPI2 colocalizes with Integrin-β1 on early endosomal compartments.
A. WIPI2 colocalization with Integrin-β1. HK2 cells were fixed 18 h after transient transfection with EGFPWIPI2, stained with the indicated antibodies and imaged by confocal microscopy. Scale bars: 10 μm. Insets show enlargements of the outlined areas with a white dashed line. Scale bars for these are 2 μm. B. Quantification. Using the images from (A), colocalization was assessed using Manders’ colocalization coefficient M2. 75 cells per condition, stemming from three independent biological experiments, were quantified. Individual cells and means are presented by smaller and larger circles, respectively, coloured according to the independent experiment. P values were calculated applying an unpaired Student’s t-test with unequal variances. The analysis was performed with 99% confidence. ****P < 0.00001.

Impact of the FSSS motif on colocalization of WIPI2 and Retriever.
A. HK2 cells were fixed 18 h after transient transfection with the indicated EGFPWIPI2 variants, stained with antibodies to VPS26C and imaged by confocal microscopy. Scale bars: 10 μm. Insets show enlargements of the outlined areas with a white dashed line. Scale bars: 2 μm B. Quantitative colocalization analysis of the images from (A). Quantification was assessed using the Manders’ colocalization coefficient M2, indicating the fraction of green pixels overlapping with magenta pixels, calculated in Image J software. 75 cells per condition, stemming from three independent biological experiments, were quantified. Values of individual cells and means are presented by smaller and larger circles, respectively, coloured according to the independent experiment that generated them. P values were calculated applying an unpaired Student’s t-test with unequal variances. The analysis was performed with 99% confidence. ****P < 0.00001.

WIPI2 colocalization with RAB5 and Retriever.
HK2 cells expressing mCherry-RAB5 and EGFPWIPI2 wild-type (A) or its indicated variants (B.C), were fixed, permeabilized and stained with antibodies to VPS35L (blue). Scale bars: 10 μm. Insets show enlargements of the outlined areas. Scale bars: 2 μm. Quantifications from multiple experiments are shown in Figure. 9

WIPI2 colocalization with RAB7 and Retriever.
HK2 cells expressing mCherry-RAB7 and wild-type EGFPWIPI2 (A) or its indicated variants (B.C) were fixed, permeabilized and stained with antibodies to VPS35L (blue). Scale bars: 10 μm. Insets show enlargements of the outlined areas. Scale bars: 2 μm. Quantifications from multiple experiments are shown in Figure. 9

WIPI2 colocalization with RAB11 and Retriever.
HK2 cells expressing mCherry-RAB7 and wild-type EGFPWIPI2 (A) or its indicated variants (B.C) were fixed, permeabilized and stained with antibodies to VPS35L (blue). Scale bars: 10 μm. Insets show enlargements of the outlined areas. Scale bars: 2 μm. Quantifications from multiple experiments are shown in Figure. 9

