Phagocytosis alters immune surveillance.

a) Scanning electron micrographs of bone-marrow-derived macrophages (BMDM) 30-40 min after phagocytosis of 4 μm polystyrene microspheres coated with phospholipids including 5% phosphatidylserine (PS) or opsonized with IgG. b) Quantification of the ventral surface area of BMDM, determined by visualizing F-actin at the level of cell attachment to the coverslip. Only BMDM containing ≥3 particles were quantified. Here and elsewhere grey dots represent individual cells and colored dots represent means of separate experiments. n = 3. c-d) RAW264.7 cells were stained for F-actin before or 30-40 min after phagocytosis of IgG-opsonized sheep red blood cells (RBC). Representative confocal ventral images are shown in c. Inset, extended focus showing IgG-RBC (magenta). In d, the number of podosomes were quantified; cells from 3 independent experiments. e) Normalized ruffling cell area estimated over 10 min by acquiring images of LifeAct-expressing RAW 264.7 cells every 15 s as described in Methods and in S Fig 2. Data from 3 independent experiments. See also S Video 1. f-g) Uptake of 70 kDa TMR-dextran (orange) by BMDM over 10 min before and after the uptake of the indicated particles, shown in insets. Panel g shows quantification of dextran uptake in cells from 3 independent experiments. h-i) Chemokinesis of human neutrophils stimulated with 10 μM fMLP before and after phagocytosis of IgG-RBC. Trajectories from one video, tracking single cells every 10 s for 10 min. See also S Video 2. In i grey dots represent the average speed of human neutrophils from heathy donors; ▴ = female donor; ▾ = male donor. P values are determined using a one-way ANOVA (b, g) or Student’s t-test (d, e, i). All scale bars, 10 μm

Cytoskeletal remodeling is associated with a sustained shift toward Rho/formin activity.

a-b) In a, representative confocal images of the mid-section of BMDM stained with phalloidin (F-actin), 40 min after phagocytosis of PS-coated or IgG-opsonized beads. Insets are extended focus images showing PS-beads (cyan) or IgG-beads (magenta). In b, the normalized fluorescence intensity of the cortical F-actin was quantified in 3 independent experiments. c, f, i) Relative activity of indicated Rho GTPases measured by G-LISA in RAW 264.7 cells. In c and f, the cells were pretreated with the Rho activator, CN03, as specified in Methods. d-e) Confocal images of RAW 264.7 cells expressing dT-2xrGBD before and post-phagocytosis. In e the cortical dT-2xrGBD fluorescence, determined at a midsection plane, was normalized to the cytosolic fluorescence; data from 3 independent experiments. Where indicated, Rho was activated using CN03 or inactivated with CT04 as described in Methods. g-h) RAW 264.7 cells expressing PAK(PBD)-YFP before and after phagocytosis of indicated particles. In h the cortical PAK(PBD)-YFP signal was measured and normalized to the cytosolic fluorescence. j) Normalized cortical F- actin fluorescence intensity of RAW 264.7 cells containing ≥3 phagocytic targets in control cells or those treated with CN03, CT04, or the formin inhibitor SMIFH2 (5 μM); data from 3 independent experiments. P values are determined using a one-way ANOVA. All scale bars, 10 μm.

Loss of podosomes associated with Fc-mediated phagocytosis is the result of increased RhoA activation of formins.

a) Ventral sections of RAW 264.7 cells before and 30-40 min after phagocytosis of IgG-RBC. Cells were stained for F-actin (blue) and vinculin (orange). Inset, projected image of particles. Scale bars, 10 μm. b) Quantification of the number of podosomes per RAW 264.7 cell before and after phagocytosis of >3 IgG-RBC particles. Cells were either untreated or pretreated with CN03 or CT04, then either pretreated with the ROCK inhibitor Y-27632 (5 μM) or with the myosin inhibitor blebbistatin (5 μM) for 10 min before exposure to phagocytic targets for an additional 30 min in the continued presence of the inhibitors. The formin inhibitor SMIFH2 (5 μM), which impairs phagocytosis, was only added after exposure and ingestion of phagocytic targets, then incubated an additional 30 min; data from 3 independent experiments. P values are determined using a one-way ANOVA.

ERM proteins coordinate Rho/formin-mediated cortical cytoskeleton thickening, decreased membrane dynamics and podosome loss.

a-b) Fluorescence lifetime imaging microscopy (FLIM) of FlipperTM in RAW 264.7 cells before and after phagocytosis of IgG-opsonized RBC. CN03-treated (Rho-activated) cells are also shown. Scale bar, 10 μm. The mean lifetime fluorescence determined in 3 separate experiments is graphed in b. c) Immunoblot of phospho-ERM (p-ERM) and GAPDH, used as loading control. Lysates were made 40 min after cells were challenged with particles. d) Immunoblot of the 3 independent clones where ezrin, radixin and moesin (ERM) were sequentially knocked out in RAW 264.7 cells. e) qPCR of ezrin, radixin and moesin in wildtype (WT) and ERM triple-KO cells. Data are means ± SD of 3 separate experiments. f) Spontaneous blebbing of triple-KO ERM (ERM-KO) in minimal medium. See also S Video 3. g) Quantification of podosomes per cell in ERM-KO cells before and 30-40 min after phagocytosis of IgG-opsonized RBC. Data from 3 independent experiments. Compare these values to those of Figure 3b. h) Rho activity measured by G-LISA in ERM-KO. Data are means ± SD from 3 separate experiments. i) Ruffling cell area of ERM-KO expressing Life-Act-GFP, determined as in Fig. 1e in, n = 3. See also S Video 4. j) Ventral surface area of ERM-KO and its parental RAW 264.7 cell line before and 40 min after phagocytosis of IgG-opsonized RBC. Surface was determined as in Fig. 1b. Only cells that contained 3 or more beads were quantified. Data from 3 independent experiments. k) TRITC-70 kDa dextran uptake after a pulse of 10 min by WT or ERM-KO cells before or after phagocytosis; data from 3 independent experiments. P values determined using a one-way ANOVA. In f, g, and i, a Student’s t-test was used.

Reactive oxygen species generated during Fc-mediated phagocytosis trigger remodelling of the cytoskeleton.

a) Immunostaining of β-tubulin (cyan) in BMDM 30-40 min after undergoing Fc-mediated phagocytosis. b) β-tubulin and β-actin associated with the pellet (P) after lysis in cytoskeletal-stabilizing buffer before and after phagocytosis, as indicated. c) Representative bright field image of BMDM 40 min after phagocytosis of either PS-coated or IgG-opsonized beads in the presence of nitroblue tetrazolium (NBT). d) Representative bright field image of BMDM 40 min after phagocytosis of IgG-opsonized RBC the presence of NBT with or without 5 μM of the NOX2 inhibitor, GSK2795039. e) Quantitation of the formazan deposited per phagosome. Data from 3 independent experiments. f-g) F-actin staining of BMDM 30-40 min post-phagocytosis of IgG-opsonized RBC in the presence or absence of 5 μM GSK2795039, added at the time of phagocytosis. Insets show phagocytic particles (magenta). In g, the ventral surface area of BMDM was quantified as in Fig. 1; n = 3. h-i) Chemokinesis of human neutrophils stimulated with fMLP before and after phagocytosis of IgG-RBC in the presence of diphenyleneiodonium (DPI; 10 µM). In h, trajectories of single cells from a representative video acquiring images every 10 s for 10 min. in i the average PMN speed was measured in 3 independent experiments. ▴ = female donor; ▾ = male donor. j-k) RAW264.7 cells treated with DPI and imaged 40 min after phagocytosis of IgG-opsonized RBC. Cells were fixed and stained for F-actin (blue), Vinculin (orange) and IgG (magenta). The number of podosomes per cell is shown in k. Dots represent individual cells from 3 independent experiments. l-m) RAW264.7 cells expressing dT-2xrGBD and treated with or without GSK2795039 were imaged after phagocytosis. The cortical dT-2xrGBD fluorescence intensity normalized to cytosol fluorescence is graphed in m. Data from 3 independent experiments. n) Rho activity in RAW264.7 cells measured by G-LISA under the indicated conditions. P values calculated by one-way ANOVA (e, g, k, m) or Student’s t-test (i). All scale bars, 10 μm

Pathogens and their associated molecules cause podosome loss by generating reactive oxygen species.

a) BMDM differentiated in GM-CSF were incubated with or without heat-killed C. albicans expressing BFP (white). Cells were fixed and stained for concanavalin A to label extracellular C. albicans (green), then permeabilized and stained for F-actin (blue) and vinculin (orange). b-c) The number of podosomes per BMDM and the cortical F-actin intensity were determined as described in Fig. 1 in 3 independent experiments. d-e) Control wildtype or ERM-KO RAW 267.4 cells were treated with or without lipopolysaccharide (LPS) in the presence or absence of DPI, as indicated, then fixed and stained. In e, the number of podosomes per cell was determined in 3 independent experiments. P values are derived from one-way ANOVA (e) or Student’s t-test (b-c). All scale bars, 10 μm.