We describe an experimental setup to visualize with unprecedented high resolution phagosome formation and closure in three dimensions in living macrophages, using total internal reflection fluorescence microscopy. It allows monitoring of the base of the phagocytic cup, the extending pseudopods, as well as the precise site of phagosome scission.
Phagocytosis is a mechanism used by specialized cells to internalize and eliminate microorganisms or cellular debris. It relies on profound rearrangements of the actin cytoskeleton that is the driving force for plasma membrane extension around the particle. In addition, efficient engulfment of large material relies on focal exocytosis of intracellular compartments. This process is highly dynamic and numerous molecular players have been described to have a role during phagocytic cup formation. The precise regulation in time and space of all of these molecules, however, remains elusive. In addition, the last step of phagosome closure has been very difficult to observe because inhibition by RNA interference or dominant negative mutants often results in stalled phagocytic cup formation.
We have set up a dedicated experimental approach using total internal reflection fluorescence microscopy (TIRFM) combined with epifluorescence to monitor step by step the extension of pseudopods and their tips in a phagosome growing around a particle loosely bound to a coverslip. This method allows us to observe, with high resolution the very tips of the pseudopods and their fusion during closure of the phagosome in living cells for two different fluorescently tagged proteins at the same time.
Phagocytosis is a major cell function that starts with the recognition and binding of material to surface receptors, which then leads to the internalization and degradation of the ingested material. While single-celled eukaryotes such as the mold Dictyostelium discoideum and amoebae use phagocytosis for feeding on bacteria, higher organisms have evolved with professional cells. Macrophages or dendritic cells are the first line of defense against pathogens in various tissues and organs, and are crucial to activate the adaptive immune system through antigen presentation and cytokine production 1-4. Under certain circumstances phagocytosis can be performed by non-professional phagocytic cells, e.g., endothelial and epithelial cells. This process is important to maintain homeostasis during development and in adulthood for normal tissue turnover and remodeling. Finally, specialized phagocytes such as Sertoli cells in the testis or retinal pigment epithelial cells are extremely potent phagocytes 5.
The formation of a phagosome where degradation of microorganisms or cellular debris occurs starts with the clustering of phagocytic receptors on the surface of the phagocytic cell. Downstream signaling events following clustering of opsonic receptors such as Fc receptors (FcR) or complement receptors (CRs) have been well characterized. However, there are also numerous non-opsonic receptors including Toll-like receptors (TLRs), lectins, mannose receptors and scavenger receptors. These receptors recognize determinants on the particle surface such as mannose or fucose residues, phosphatidylserine, and lipopolysaccharides 1,6-9.
Pathogen or cell debris recognition involves binding to and clustering of several types of phagocytic receptor, which then lead to intense and transient actin remodeling. In parallel, focal exocytosis of intracellular compartments contributes to the release of membrane tension and is important for efficient phagocytosis of large particles. The signaling events leading to actin polymerization and membrane deformation were dissected in experimental models that triggered a single phagocytic receptor. During FcR-mediated phagocytosis, there is intense actin polymerization that is regulated by small GTPases (Rac, Cdc42). Among their downstream effectors, the Wiskott-Aldrich syndrome protein (WASP) leads to activation of the Actin-Related Protein 2/3 complex (Arp2/3) that nucleates actin filaments 1,2,4,10. Local production of phosphatidylinositol-4, 5-bisphosphate (PI(4,5)P2) is crucial for initial actin polymerization that drives pseudopod formation. Its conversion to PI(3,4,5)P3 is required for pseudopod extension and phagosome closure 11. Several pathways contribute to the disappearance of PI(4,5)P2. Firstly detachment of phosphatidylinositol phosphate kinases (PIPKIs) from the phagosome arrests PI(4,5)P2 synthesis. Secondly, it can be phosphorylated and consumed by class I PI3K kinases (PI3K) and converted in PI(3,4,5)P3 12. A role for phosphatases and phospholipases has also been implied in PI(4,5)P2 hydrolysis and F-actin removal during phagocytosis in mammalian cells and in Dictyostelium13,14. The phospholipase C (PLC)δ hydrolyzes PI(4,5)P2 into diacylglycerol and inositol-1,4,5-trisphosphate. The PI(4,5)P2 and PI(3,4,5)P3 phosphatase OCRL (oculocerebrorenal syndrome of Lowe) has also been implicated in phagosome formation. Precise local formation of F-actin and its depolymerization is tightly regulated in space and time and we have shown that recruitment of intracellular compartments is important to deliver locally the OCRL phosphatase, thus contributing to local actin depolymerization at the base of the phagocytic cup 13,15. For this, we used the experimental set up described here.
The mechanism and molecular players required for phagosome closure and membrane scission remain poorly defined because of the difficulties in visualizing and monitoring the site of phagosome closure. Until recently, phagocytosis was observed on fixed or living cells that internalize particles on their dorsal face or on their sides, making the timely visualization of the site of phagosome closure difficult. In addition, fixing methods could cause retraction of membranes and bias the results on pseudopodia extension and closure. By contrast, the assay that we have set up and describe here allows us to visualize pseudopod extension and the closure step of phagocytosis in living cells 13, based on total internal reflection microscopy (TIRFM) 16. This optical technique uses an evanescent wave to excite fluorophores in a thin area at the interface between a transparent solid (coverslip) and a liquid (cell culture medium). The thickness of the excitation depth is around 100 nm from the solid surface, allowing visualization of molecular events close to the plasma membrane. TIRFM allows a high signal-to-background ratio, and limits the out-of-focus fluorescence collected and the cytotoxicity due to illumination of cells.
Taking advantage of TIRFM, we developed the "phagosome closure assay" in which coverslips are activated with poly-lysine and then coated with IgG-opsonized red blood cells (IgG-SRBCs). Macrophages expressing transiently fluorescently tagged proteins of interest are then allowed to engulf the IgG-SRBCs. While cells detach the target particles that are non-covalently bound to the glass surface, the tips of the pseudopods can be observed and recorded in the TIRF mode. TIRF acquisitions are combined with acquisitions in the epifluorescence mode, after shifting the stage 3 µm above, which allows the visualization of the base of the phagocytic cup. Addition of pharmacological drugs such as those inhibiting actin or dynamin during the process is also possible to further dissect the process at the molecular level.
The protocol described in detail here is for a RAW264.7 murine macrophage cell line and opsonized particles, but virtually, it can be adapted to any other phagocytic cell and with other targets such as beads. This method will allow better characterization of the regulation in time and space of the molecular players involved in pseudopod extension and phagosome closure during various phagocytic processes.
Note: The plasmid used Lifeact-mCherry is a kind gift of Dr. Guillaume Montagnac, Institut Curie, Paris, generated after 17.
1. Cells and Transfection
Note: RAW264.7 macrophages are grown to sub confluency in complete medium (RPMI (Roswell Park Memorial Institute) 1640 medium, 10 mM HEPES, 1 mM sodium pyruvate, 50 µM β-mercaptoethanol, 2 mM L-Glutamine and 10% FCS (Fetal Calf Serum)) in a 100 mm plate. They are transfected with plasmids encoding fluorescently tagged proteins by electroporation. Routinely approximately 5 – 6 x 106 cells are transfected with 20 µg or 10 µg of plasmid for each transfection or co-transfection respectively. Note that other means of transfection based on electroporation or lipofection can be used as alternative approaches.
2. Opsonization of Red Blood Cells
Note: As a model of particle target for macrophages, sheep red blood cells (SRBCs) are used. Usually, around 7 x 106 SRBCs per 35 mm glass bottom dish is used.
3. Poly-lysine Coating of Coverslips
4. Non-covalent Fixation of SRBCs on Glass Bottom Dishes
5. Phagocytosis Visualized by TIRFM
The experimental system described in this manuscript is schematically represented in Figure 1. Transfected RAW264.7 macrophages expressing the proteins of interest fused to a fluorescent tag are placed into contact with IgG-opsonized sheep red blood cells (SRBCs) that were non-covalently fixed on the coverslip. The macrophages can detach the SRBC from the coverslip to engulf it. The TIRF microscope used allows concomitant acquisition of signals from the TIRF area corresponding to the tips of the pseudopods and signals in the epifluorescence mode after a Z shift 3 µm above.
It is essential to determine the critical angle for total internal reflection fluorescence as described in Figure 2. This ensures a clean TIRF signal from a 100 nm region below the plasma membrane. Figure 3 represents the development of acquisition parameters through a module called "Protocol Editor" included in the Live Acquisition software. This module allows users to create and manage workflows before they are sent to the microscope, which will execute the process. At the end of the acquisition, the user collects a TIFF stream.
Figure 4 shows, a representative live-cell TIRFM movie of a "phagosome closure assay" in RAW264.7 macrophages transfected with the plasmid (e.g., Lifeact-mCherry, a kind gift of Dr. Guillaume Montagnac, Institut Curie, Paris, generated after 17) to follow actin polymerization. Macrophages transiently expressing plasmid were allowed to engulf IgG-SRBCs bound to the coverslip. As the tips of the pseudopods were apposed to the glass coverslip around one SRBC, an F-actin ring was detected in the TIRF area (top panel) that progressively narrowed until it closed. In parallel, the blurry F-actin signal detected by epifluorescence after shifting the stage 3 µm above (bottom panel) corresponded to depolymerization at the base of the phagocytic cup. After 3 min of acquisition, the SRBC was totally internalized, as confirmed with transmitted light (panel on the right).
Therefore, this method can be used to properly distinguish the molecular events that take place at the very ends of the pseudopods and the molecular events occurring at the base of the phagocytic cup.
Figure 1. Schematic Representation of the "Phagosome Closure Assay" Analyzed by TIRFM. The phagosome closure assay is performed using macrophages transiently expressing one or two fluorescently tagged-protein. Macrophages are deposited on IgG-opsonized SRBCs non-covalently fixed on poly-lysine coated coverslips. Images are recorded in TIRF mode to detect the site of phagosome closure and in epifluorescence mode to detect the base of the phagocytic cup. Please click here to view a larger version of this figure.
Figure 2: Determination of the Critical Angle for TIRFM. (A) Using "Live acquisition", a cell expressing fluorescently tagged proteins is placed in the middle of the field (region 1 in red). With the TIRF Stack option, images were acquired at one excitation wavelength, with different angles starting from 0° up to 5°, with an increment of 0.01° (region 2 in red). (B) The sequence of images is opened using ImageJ Color Profiler software and the mean fluorescence intensity of a region of interest (ROI) is plotted with function of the angles on the x axis using "Stacks" and "PlotZ axis profile" (region 1 in red). (C) The x position of the peak of fluorescence on the plot corresponds to the critical angle: 1.98° (black dotted line). Any value after this angle can be used. As an example, 2.00° can be chosen (red line). Please click here to view a larger version of this figure.
Figure 3. Workflow Process using Live Acquisition Module: "Protocol Editor". (A) In the "Protocol Editor" window, a workflow canvas is created. (B) This protocol comprises a "Loop" of actions that the microscope will repeat the number of times decided by the user. As an example: 750 (1). One loop included: "Multi Channels" acquisition with laser excitation of fluorescent proteins of interest in TIRF mode. As example: Laser 491 nm intensity 50% -TIRF angle 2.00 (2); "Z move" of the objective 3 µm (3); "Multi Channels" acquisition in epifluorescence mode (4); "Z move" of the objective back to the TIRF region (5) and a "Snapshot" in transmitted light (6). Please click here to view a larger version of this figure.
Figure 4. F-actin is Accumulated as a Point at the site of Phagosome Closure. Phagosome closure assay was performed using RAW264.7 macrophages transiently expressing the plasmid Lifeact-mCherry (a kind gift of Dr. Guillaume Montagnac, Institut Curie, Paris, generated after 17). The plasmid signal was acquired in TIRF area (top) and in epifluorescence mode (bottom). The red arrowhead indicates the actin accumulation in the TIRF region. Transmitted light image at 3 min is presented, indicating an internalized SRBC. Scale bar, 10 µm. Please click here to view a larger version of this figure.
Movie 1: F-actin is Accumulated in the Tips of the Pseudopods and at the Site of Phagosome Closure, while it is Cleared from the Base of the Phagocytic Cup. Phagosome closure assay was performed using RAW264.7 macrophages transiently expressing the plasmid. Plasmid imaging in TIRF area (top) and in epifluorescence mode (bottom) using the TIRFM was performed alternatively every 50 msec during 3 min at 37 °C. Please click here to view this video.
The experimental protocol described here proposes an unprecedented method to follow in real time and in living cells, with high-resolution, the formation of a phagosome and in particular its closure. Several technical aspects have to be discussed. Firstly, the assay is very sensitive to temperature. It is very important to check that the heating chamber is at 37 °C and that all media, devices or cells are kept within the chamber to avoid temperature changes that could impair the efficiency of phagocytosis. We noticed that subtle temperature changes were sufficient to prevent phagosome formation. Secondly, the protocol described here allows a precise determination of the TIRF angle that will be used throughout the experiment. We believe that this constitutes a substantial advantage when performing TIRF experiments, allowing us to be confident when concluding about events near the plasma membrane. Thirdly, when working with transiently transfected cells, it is important to choose cells with moderate levels of expression of the proteins of interest. With the revolution of the genome-editing technology, it will be feasible to work with cells expressing proteins at native levels. Thus, quantitative fluorescence microscopy will be possible to explore the dynamics of major functions in living cells.
With this assay, it is now possible to analyze precisely in time and space the different players involved in pseudopod extension and fusion. The forced orientation of the target to engulf allows better monitoring of extending pseudopods and the closure site, compared with phagosome formation imaging using a confocal microscope and a different experimental set up. The only caveat is the low output. Not all cells chosen on the basis of their fluorescence will phagocytose during the acquisition time. A potential improvement would be to have a patterned surface coated with target particles to increase the number of concomitant acquisitions. Using this method, we have dissected the spatio-temporal organization of F-actin dynamics and recruitment of recycling (VAMP3-positive) vesicles during phagosome formation 13. We observed recruitment of F-actin at the very tips of closing pseudopods, but no VAMP3-positive vesicles in this closing zone of the phagocytic cup. VAMP3 was only detected as a plasmalemmal diffuse signal. In contrast, VAMP3-positive vesicles were detected at the base of the forming phagosome, as visualized in the epifluorescence images. These results showed that F-actin polymerization occurs in the extending pseudopods and in their tips, while there is depolymerization at the base where the vesicular traffic is concentrated. This spatio-temporal dynamic of F-actin polymerization was proposed in seminal work reviewed in 2 and recently also confirmed in 12.
Using the same experimental model, we recently showed that dynamin-2, which mediates the scission of endocytic vesicles, was recruited early and concomitantly with actin during phagosome formation and pseudopod extension. Dynamin-2 then accumulates at the site of phagosome scission. Acute treatment of cells with pharmacological inhibitors clearly showed that dynamin-2 plays a crucial role in phagosome scission 18. The unique possibility to monitor with high precision the closure step in phagosome formation allowed us to conclude the possible role of dynamin-2 in phagosome scission.
The phagosome closure assay can be in principle adapted to any other phagocytic cell (dendritic cells, neutrophils), provided that they express fluorescently tagged proteins of interest. In addition, the method can be extended to other target particles like beads or bacteria. The size and shape of the different targets could also be varied to get further insight into the mechanism of phagosome formation and in particular the forces involved in phagosome closure. Following the localization of different molecular players in living phagocytosing cells using the phagosome closure assay described here opens avenues for better definition of the molecular mechanisms and machineries important for phagosome formation and closure.
The authors have nothing to disclose.
We thank Dr. Alexandre Benmerah (Institut Imagine Necker, Paris, France) for initial discussions on the experimental approach and Dr. Jamil Jubrail for reading the manuscript. Nadège Kambou and Susanna Borreill are acknowledged for performing experiments with the method in our laboratory. This work was supported by grants from CNRS (ATIP Program), Ville de Paris and Agence Nationale de la Recherche (2011 BSV3 025 02), Fondation pour la Recherche Médicale (FRM INE20041102865, FRM DEQ20130326518 including a doctoral fellowship for FMA) to FN, and Agence Nationale de Recherche sur le SIDA et les hépatites virales” (ANRS) including a doctoral fellowship for JM.
Anti-sheep red blood cells IgG | MP Biomedicals | 55806 | |
Bovine Serum Albumin heat shock fraction, pH 7, ≥98% | Sigma | A7906 | |
Cell lifter | Corning | 3008 | |
Cuvettes 4mm | Cell project | EP104 | |
DPBS, no calcium, no magnesium | Thermo Fischer Scientific | 14190-094 | Room temperature |
Electrobuffer kit | Cell project | EB110 | |
100mm TC-Treated Cell Culture Dish | Corning | 353003 | |
Gene X-cell pulser | Biorad | 165-2661 | |
Gentamicin solution | Sigma | G1397 | |
Glass Bottom Dishes 35 mm uncoated 1.5 | MatTek corporation | P35G-1.5-14-C Case | |
iMIC | TILL Photonics | Oil-immersion objective (N 100x, NA1.49.), heating chamber with CO2, a camera single photon detection EMCCD ( Electron Multiplying Charge Coupled Device) and a 1.5X lens | |
Poly-L-Lysine Solution 0.1% | Sigma | P8920-100ml | Dilution at 0.01% in water |
RPMI 1640 medium GLUTAMAX Supplement | Life technologies | 61870-010 | |
RPMI 1640 medium, no phenol red (10×500 ml) | Life technologies | 11835-105 | Warm in 37°C water bath before use |
Sheep red blood cells (SRBCs) | Eurobio | DSGMTN00-0Q | Conserved in Alsever buffer at 4°C before use |