Absence of SNX5 reduces macropinocytosis in BMMs. (A) SNX5 is localised on newly formed macropinosomes. Quiescent SNX5^+/+ BMMs were incubated with 500 µg/ml of FITC-conjugated 70 kDa dextran in the presence of 50 ng/ml CSF-1 for 15 min at 37°C. Monolayers were fixed, permeabilised and stained with rabbit anti-SNX5 antibodies (red). Macropinosomes are indicated by arrows. Macropinosomes indicated by arrows are magnified in the inset. Scale bars: 10 µm (main panel); 1 µm (inset). (B,C) Quiescent BMMs from SNX5^+/+ and SNX5^−/− mice were incubated with 500 µg/ml FITC-conjugated 70 kDa dextran and 50 ng/ml CSF-1 for 15 min and fixed and processed. Scale bar: 10 μm. In C, the level of FITC–dextran uptake was analysed using Metamorph software. The mean is indicated. (D) Analysis of FITC–dextran uptake as in B by flow cytometry. 20,000 events were collected per sample. Dashed lines, cells incubated at 4°C on ice, solid lines, cells incubated at 37°C. (E,F) Quiescent BMMs from SNX5^+/+ and SNX5^−/− mice were pulsed with 50 µg/ml DQ-ovalbumin in the presence of CSF-1 for 3 min, washed and chased for 30 min at 37°C in the presence of CSF-1 or incubated with DQ-ovalbumin in the presence of CSF-1 for 30 min on ice (4°C). Scale bar: 10 µm. In F, level of DQ-ovalbumin uptake was analysed using Metamorph software; the mean is indicated. ***P<0.001 (Student's t-test).

Absence of SNX5 does not affect macropinocytosis in immature splenic dendritic cells. (A) Immunoblot of BMMs and purified splenic dendritic cells (DC) from SNX5^+/+ mice. BMMs and dendritic cells were lysed in a reducing sample buffer and samples were resolved in a 10% NuPAGE gel. Proteins were then transferred onto a PVDF membrane and probed with anti-SNX5 (top panel) and anti-α-tubulin (bottom panel) antibodies as a loading control. (B) Immunofluorescence images and (C) flow cytometry histogram of Nycodenz-enriched dendritic cells from SNX5^+/+ and SNX5^−/− mice following incubation with 1 mg/ml of FITC-conjugated 70 kDa dextran at 37°C for 15 min and live cells stained with (B) biotinylated hamster anti-mouse CD11c antibody followed by streptavidin conjugated with Alexa Fluor 647 and DAPI prior to cytospin and microscopy or (C) stained with anti-CD11c antibody conjugated to eFluor^® 450, anti-CD86 antibody conjugated to APC and anti-MHCII antibody conjugated to PerCP-Cy5.5. Shown are flow cytometric plots gated on CD11c⁺, MHCII dendritic cells and analysed for CD86 expression and FITC–dextran update. Solid lines, cells incubated with dextran at 37°C; dashed lines, cells incubated with dextran at 4°C. Scale bar: 10 µm.

Absence of SNX5 does not affect phagocytosis or retromer transport of CD8–M6PR. (A,B) BMMs from SNX5^+/+ and SNX5^−/− mice were incubated with Alexa-Fluor-488-labelled zymosan A bioparticles at 37°C for 60 min. Non-phagocytosed zymosan particles were removed with ice-cold PBS washes. BMMs were (A) processed for immunofluorescence and stained with TRITC–phalloidin (red) followed by DAPI (blue). Scale bar: 10 μm. In B, the mean±s.d. of fluorescent particles per cell were scored for triplicate experiments of 50 cells each. Phagocytosis was expressed as the mean of particles per cell. (C,D) Expression of CD8–M6PR was introduced into BMMs by recombinant adenovirus. Prior to labelling with anti-CD8, BMMs was incubated with mouse BD Fc Block™. The internalisation and transport of the antibody–CD8–M6PR complex from the plasma membrane to the Golgi was tracked by incubation at 37°C for the indicated time. Monolayers were then fixed, permeabilised and stained with Alexa-Fluor-488-conjugated anti-mouse-IgG, anti-GRASP65 (where indicated) (red) and DAPI (blue). Scale bar: 10 µm. (D) Percentage of antibody-CD8-M6PR at the Golgi region after 60 min incubation at 37°C, determine by calculating the percentage of total pixels which overlapped with GRASP65 using Metamorph software. The mean is indicated.

SNX5 is localised to actin-rich plasma membrane ruffles in response to CSF-1 stimulation. (A) Endogenous SNX5 is localised on actin-rich regions of plasma membrane. Quiescent SNX5^+/+ BMMs were stimulated with CSF-1 for 3 min at 37°C, and monolayers then fixed, permeabilised and stained with either anti-SNX5 antibody followed by Alexa-Fluor-488-conjugated anti-rabbit-IgG antibody and TRITC–phalloidin (red). Arrows indicate regions of actin-rich plasma membrane ruffles. The boxed region is magnified on the far right. Scale bars: 10 µm. (B) SNX5 is localised with CSF-1 receptor at the plasma membrane. GFP-SNX5 was introduced into BMMs by recombinant adenovirus. Transduced BMMs were rendered quiescent overnight followed by 3 min stimulation with CSF-1 at 37°C, and monolayers then fixed, permeabilised and stained for CSF-1 receptor (red). Arrows indicate regions of plasma membrane where colocalisation is observed between GFP–SNX5 and CSF-1 receptor. Boxed region is magnified on the far right. Scale bars: 10 µm. (C) Effect of GW2580 on macropinocytosis in BMMs from SNX5^+/+ and SNX5^−/− mice. Quiescent BMMs from SNX5^+/+ and SNX5^−/− mice were pre-treated with 5 µM of GW2580 or DMSO carrier control for 1 h at 37°C. Cells were then incubated with 500 µg/ml of FITC-conjugated 70 kDa dextran and 50 ng/ml CSF-1 in the absence or presence of GW2580 (as indicated) for 15 min at 37°C prior to processing for immunofluorescence microscopy. Level of FITC-dextran uptake was analysed using Metamorph software. The mean is indicated. **P<0.01, ***P<0.001 (Student's t-test). (D) Level of cell surface CSF-1 receptor in SNX5^+/+ and SNX5^−/− BMMs. Quiescent BMMs were gently removed from petri dishes, stained for CSF-1 receptor followed by Alexa-Fluor-488-conjugated anti-rabbit-IgG antibody (solid lines) and 20,000 live cells were analysed by flow cytometry. Dashed lines, cells stained with conjugate alone.

Absence of SNX5 prevents the formation of actin-rich dorsal ruffles but not actin-rich peripheral ruffles in response to CSF-1 stimulation. (A) Peripheral ruffles. Quiescent BMMs from SNX5^+/+ (top panels) and SNX5^−/− (bottom panels) mice were either stimulated with CSF-1 for 3 min at 37°C or left unstimulated, as indicated. Monolayers were then fixed, permeabilised and stained using TRITC–phalloidin (red) and DAPI (blue). Scale bar: 10 µm. (B) Dorsal ruffles. Quiescent BMMs from SNX5^+/+ (top panels) and SNX5^−/− (bottom panels) mice were either stimulated with CSF-1 for 15 min at 37°C or left unstimulated, as indicated. Images are x-z cross-sections of fixed BMMs stained with TRITC–phalloidin. Images are a series of x-z cross-sections of fixed BMMs stained with TRITC–phalloidin across the width of the cell (number refers to the position of the cross section as a fraction of the total cell width). Arrows highlight the actin-stained dorsal ruffles on the SNX5^+/+ macrophage (top) compared to the lack of dorsal ruffles on SNX5^−/− macrophage (bottom).

Ultrastructural analysis of dorsal plasma membrane ruffles in the presence and absence of SNX5. BMMs from SNX5^+/+ and SNX5^−/− mice were deprived of CSF-1 for 18 h and then either left unstimulated or stimulated with CSF-1 for 3 min (A,B) or 15 min (B). The cells were fixed with 0.4% glutaraldehyde and 4% paraformaldehyde in 0.1 M phosphate buffer for 16–18 h at 4°C, washed and dehydrated in graded ethanol solutions in water. The coverslips were dried and mounted on specimen stubs and coated with gold–palladium alloy. CSF-1-induced membrane ruffles were imaged using SEM (3500× magnification). Scale bar: 20 µm (top two rows), in magnified image, 2 µm. (B) Membrane ruffling was estimated by manually tracing the perimeter of the ruffles and measuring the area. The percentage ruffling for each cell was obtained by dividing the ruffling area by the sum of ruffling area and cell area (see Materials and Methods). The results shown are mean±s.d. of 30 cells from each population per time point over two independent experiments. ***P<0.001 (non-parametric Student's t-test).