SEARCH

SEARCH BY CITATION

Keywords:

  • macropinocytosis;
  • EEA1;
  • endocytosis;
  • immunocytochemistry;
  • A431 cells

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Association of early endosomal autoantigen 1 (EEA1) with macropinosomes was examined in EGF-stimulated A431 cells by dual labeling with immunofluorescence of anti-EEA1 and FITC-dextran (FDx), a fluid-phase endocytic marker. Addition of EGF to A431 cells drastically enhanced macropinosome formation. Newly formed macropinosomes labeled with 5-min pulse of FDx were located at the cell periphery and labeled weakly for EEA1. After a 5-min chase, these macropinosomes aggregated and frequently fused with each other. Immunofluorescence showed that EEA1 appeared on the membrane of FDx-labeled macropinosomes at that time, suggesting that EEA1 functioned in homotypic macropinosome fusion. With longer chase (30–60 min), macropinosomes decreased in number and size, indicating that FDx was largely exocytosed via recycling compartments. A small amount of FDx-labeled macropinosomes remained in the perinuclear region even at 60 min after pulse labeling. They were EEA1-positive but negative for cathepsin D, a lysosomal enzyme. This indicates that macropinosomes do not mature to late endosomes or fuse with lysosomes. Instead, EEA1 continuously mediates homotypic fusion as long as the macropinosomes persist. Anat Rec Part A 277A:298–306, 2004. © 2004 Wiley-Liss, Inc.

Endocytosis is involved in cellular multifunction such as regulation of channels and growth factor receptors, uptake of nutrients, degradation of internalized materials, recycling of synaptic vesicles, and antigen processing. In receptor-mediated endocytosis, specific ligands bound to cell surface receptors are selectively taken up into early endosomes through clathrin-coated vesicles. Some internalized receptors move to late endosomes and finally to lysosomes, where they are degraded. Others recycle to the plasma membrane (Mukherjee et al., 1997; Leung et al., 2000). The mechanism by which endocytosed material is transported through receptor-mediated endocytosis is explained with two models: the vesicle shuttle model and the maturation model (Helenius et al., 1983; Murphy et al., 1992).

In macropinocytosis, extracellular solutes and a variety of nutrients or antigens are taken up nonselectively into macropinosomes. These organelles mature differently in different cell types (Hewlett et al., 1994; Swanson and Watts, 1995; Nobes and Marsh, 2000; Maniak, 2001). In macrophages, macropinosomes primarily move centripetally, shrink, and merge with tubular lysosomes (Racoosin and Swanson, 1992, Racoosin and Swanson, 1993; Araki et al., 1996). They have been reported neither to concentrate receptors nor to contain any discernible coats except for F-actin at the early stage of formation (Racoosin and Swanson, 1993; Swanson and Watts, 1995; Mukherjee et al., 1997; Araki et al., 2000). The macropinocytic pathway in macrophages and dendritic cells is important for MHC class I and class II antigen presentation (Nobes and Marsh, 2000). Although most other cells do not exhibit macropinocytic activity under normal condition, some types of cells show remarkable macropinocytosis after the stimulation with cytokines. Most prominently, epidermal growth factor (EGF) induces macropinocytosis in the human epidermoid carcinoma cell line A431 (West et al., 1989; Hewlett et al., 1994). Unlike in macrophages, most materials macropinocytosed by EGF-stimulated A431 cells are recycled back to the extracellular medium (West et al., 1989; Hewlett et al., 1994; Swanson and Watts, 1995). Compared with macropinocytosis in antigen presentation cells, the transport pathway and physiological significance of the macropinocytosis in A431 cells are less well delineated.

Early endosomal autoantigen 1 (EEA1) is an effector of the small GTPase Rab5, which localizes both to cytosol and to membranes of early endosomes (Mu et al., 1995; D'Arrigo et al., 1997; Patki et al., 1997; Simonsen et al., 1998; Christoforidis et al., 1999; Pfeffer, 1999; Leung et al., 2000; Rubino et al., 2000; Wilson et al., 2000; Mills et al., 2001). EEA1 plays a key role in tethering and fusion of early endosomes in receptor-mediated endocytosis (Christoforidis et al., 1999; McBride et al., 1999; Dumas et al., 2001; Mills et al., 2001; Lawe et al., 2002). It is recruited from cytosol to early endosomes via its C-terminal FYVE finger and N-terminal Rab5-binding site. The FYVE domain interacts with phosphatidylinositol 3-phosphate (PI(3)P), suggesting that recruitment of EEA1 to early endosomes depends on dual interaction with Rab5 and PI(3)P (Simonsen et al., 1998; Rubino et al., 2000; Dumas et al., 2001; Lawe et al., 2002). McBride et al. (1999) hypothesized a model of molecular events leading to homotypic fusion of early endosomes, consisting of Rab5 activation and EEA1 recruitment to the membrane of early endosomes, EEA1 oligomer assembly with N-ethylmaleimide-sensitive factor and Rabaptin5/Rabex5 on endosomal membranes, Rab effector-mediated vesicle docking, and EEA1-syntaxin13 interaction for fusion. Although knowledge about the molecular mechanism of EEA1 function is advancing rapidly, the implication of EEA1 in endocytic pathways other than the receptor-mediated endocytosis has been less thoroughly investigated.

Recently, EEA1 recruitment to phagosomes was demonstrated by immunofluorescence microscopy (Fratti et al., 2001; Gillooly et al., 2001). This fact indicated that EEA1 might mediate fusion not only of early endosomes but also of other endocytic compartments. However, it is still unknown whether EEA1 is associated with the fluid-phase macropinocytic pathway and what role it plays in A431 cells. In this study, we have shown that EEA1 localizes not only to early macropinosomes but also to long-lived macropinosomes in the perinuclear region of the cell, suggesting that EEA1 continuously mediates the macropinosome fusion for longer time than in the receptor-mediated endocytic pathway.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Reagents

Texas Red-conjugated transferrin, lysine-fixable FITC-dextran, MW 10 kD (FDx), and Texas Red-dextran, MW 10 kD (TRDx), were purchased from Molecular Probes (Eugene, OR). Purified mouse monoclonal anti-EEA1 antibody was from Transduction Laboratories (San Diego, CA). A primary antibody, rabbit anticathepsin D serum, was a gift from Dr. S. Yokota, Yamanashi Medical School. As a secondary antibody, Texas Red-labeled antirabbit IgG was from Vector Lab (Burlingame, CA). Dulbecco's modified essential medium (DMEM), fetal bovine serum (FBS), and normal goat serum were from Gibco-BRL (Grand Island, NY). Paraformaldehyde (PFA), bovine serum albumin (BSA, fraction V), Triton X-100, saponin, HEPES, and human recombinant EGF were from Sigma Chemical (St. Louis, MO). Circular glass coverslips of 12 mm in diameter (no. 1 thickness) were from Matsunami Glass (Osaka, Japan). Multiwell plates were from Sumitomo Bakelite (Tokyo, Japan).

Cell Culture

A431 cells were maintained in DMEM containing 10% FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin. Before experiments, they were incubated for at least 1 hr at 37°C in serum-free Ringer's buffer consisting of 155 mM NaCl, 5 mM KCl, 1 mM MgCl2, 2 mM Na2HPO4, 10 mM glucose, 10 mM HEPES at pH 7.2, and 0.5 mg/ml BSA. The cells were plated on 12 mm diameter coverslips in multiwell plate and cultured to obtain sufficient confluency.

Macropinocytosis and Receptor-Mediated Endocytosis

Fluorescent dextran and fluorescent transferrin were used as probes of fluid-phase macropinocytosis and receptor-mediated endocytosis, respectively. The serum-starved cells were stimulated with 100 ng/ml EGF in the presence of 1 mg/ml FITC-labeled, lysine-fixable dextran (FDx) or 25–50 μg/ml fluorescent-transferrin (FITC or Texas Red-labeld Tf) for 5-min pulse labeling, and then chased in Ringer's buffer without EGF, FDx, or fluorescent-Tf for 0, 5, 25, and 55 min, yielding macropinosomes aged 5, 10, 30, and 60 min, respectively. The cells were fixed with 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4, containing 6% sucrose for 30 min at room temperature, rinsed three times with phosphate buffer saline (PBS) for 5 min each, then either observed by fluorescence microscopy (Nikon TE300) or further processed for immunofluorescence microscopy.

Immunofluorescence Microscopy

After rinsing with PBS, fixed A431 cells were aldehyde-quenched by 0.25% NH4Cl in PBS for 10 min, and pretreated with a permeabilizing/blocking solution consisting of 0.25% saponin, 2% BSA, and 2% normal goat serum in PBS for 10–20 min. Then the cells were incubated with mouse monoclonal anti-EEA1 antibody (1:250) and/or rabbit polyclonal anticathepsin D antibody (1:500) diluted in the permeabilizing/blocking solution for 90 min at room temperature. After rinsing with PBS, the cells were incubated with Alexa 594-labeled antirabbit IgG, Alexa 488-labeled antimouse IgG, or Alexa 594-labeled antimouse IgG (all 1:500 dilutions). As negative controls, normal mouse IgG or rabbit serum was substituted for the specific antibody at the same concentration. The specimen coverslips were mounted on glass slides using Fluorogard (Bio-Rad Lab) and observed by a confocal laser scanning microscope (Olympus, GB200). Laser intensity, aperture size, gain, and black level of the confocal laser scanning microscope were carefully determined so that each signal was not detected in the other channel. Furthermore, we observed single labeling specimens to confirm that the other channel signal did not encounter.

Video Microscopy

A431 cells were cultured onto 25 mm circular coverslips and assembled in Leiden chamber on the thermo-controlled stage of the inverted phase-contrast microscope (Nikon TE300). Time-lapse images of living A431 cells after EGF stimulation were collected using MetaMorph Imaging System as previously described (Araki et al., 2003).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Macropinocytosis and Receptor-Mediated Endocytosis in A431 Cells

Immediately after the addition of EGF to A431 cells, phase-bright vacuoles formation was markedly induced at the cell periphery where membrane ruffling was active (Fig. 1A and B). By FDx pulse-labeling for 5 min, these vacuoles were labeled with FDx (Fig. 1C), suggesting that these vacuoles were macropinosomes newly formed from peripherally located ruffles. It was previously reported that fluid-phase incorporation by macropinocytosis utilizes a pathway distinct from receptor-mediated endocytosis in EGF-stimulated A431 cells (Hewlett et al., 1994). To confirm that this macropinocytic pathway does not cross the receptor-mediated endocytic pathway at any stage, we observed the cells similarly pulse-labeled with FDx, Texas Red-Tf, or FITC-Tf and chased for various times (Fig. 2). After 5-min pulse labeling, FDx was in large macropinosomes predominantly located at the cell periphery. In contrast, Tf was in punctate vesicles and relatively small vacuoles, presumably early endosomes distributed throughout the cells. Thus, the size and distribution pattern of FDx-labeled compartments were apparently different from Tf-labeled endosomes. With increasing chase time, fluorescence of both probes diminished, indicating that they were recycled back to extracellular medium by exocytosis. After 30 min, FDx-labeled macropinosomes decreased in size and number, whereas Tf-labeled endosomes decreased only in number. Tubular extensions from FDx-labeled macropinosomes and FITC-Tf-labeled endosomes were occasionally observed. After 60 min, a small mount of both probes remained in cells, though most of them seemed to have recycled. In the cells doubly labeled with FDx and Texas Red-Tf, the two probes labeled distinct compartments and scarcely mixed at any stage (Fig. 2I–L). Consistent with earlier reports (West et al., 1989; Hewlett et al., 1994), FDx labeled only the fluid-phase macropinocytic pathway, and this did not intersect the receptor-mediated endocytic pathway.

thumbnail image

Figure 1. Close relationship between macropinosomes and peripheral ruffles. A431 cells were pulsed with 1 mg/ml FITC-dextran (FDx) and 100 ng/ml EGF at 37°C for 5 min, fixed with 4% paraformaldehyde for 30 min, then observed simultaneously by phase-contrast (A and B) and fluorescence (C) microscopy. At an upper focal plane of phase-contrast microscopy (A), phase-dark membrane ruffles were notable in the cell periphery. Many phase-bright macropinosomes were formed in the cytoplasm beneath the active ruffling area of cell periphery (B). These macropinosomes (arrows) were labeled with FDx, a fluid-phase probe (C). Scale bar = 10 μm.

Download figure to PowerPoint

thumbnail image

Figure 2. Comparison of endocytic pathways for macropinocytosis of FDx and receptor-mediated endocytosis of transferrin. The cells were incubated with 1 mg/ml FITC-dextran (FDx) and/or 25 μg/ml fluorescent-transferrin (Tf) for 5-min pulse labeling under EGF stimulation and then chased in probe-free Ringer's buffer at 37°C for 0, 5, 25, and 55 min, yielding endosomes aged 5 (A, E, I, M), 10 (B, F, J, N), 30 (C, G, K, O), and 60 min (D, H, L, P), respectively. FDx-labeled large macropinosomes predominantly formed at the cell periphery (A). With chase time, FDx-labeled macropinosomes decreased in size and number (A–D). FITC-Tf-labeled punctate vesicles and relatively small early endosomes distribute throughout the cell (E). FITC-Tf labeling also markedly decreased with chase time (E–H). I–L: The EGF-stimulated A431 cells were simultaneously dual-labeled with FDx and Texas Red-Tf. These two probes were taken up by distinct endocytic components and scarcely mixed at any stages. M–P: The cells labeled by receptor-mediated endocytosis of Texas Red-Tf were immunostained for EEA1 (green). EEA1 localized on Texas Red-Tf-loaded early endosomes at 5- to 10-min age (M, N) but not later (O, P). n, nuclei. Scale bar = 10 μm.

Download figure to PowerPoint

In the receptor-mediated endocytosis of Texas Red-Tf, immunoreaction for EEA1 was detected on 5- to 10-min-aged early endosomes (Fig. 2M and N). However, after 30- to 60-min chase, Texas Red-Tf-loaded compartments were scarcely immunolabeled for EEA1 (Fig. 2O and P).

EEA1 Is Associated With Macropinosomes

Although EEA1 is required for tethering and fusion between early endosomes in receptor-mediated endocytosis, it is unclear at present whether EEA1 is necessary for such fusion between macropinosomes. To clarify if EEA1 is associated with any stages of macropinocytosis, the cells were pulsed with FDx and EGF and chased in FDx-free Ringer's buffer. When observed just after 5-min FDx pulse-labeling, newly formed FDx-filled macropinosomes were located at the cell periphery and somewhat heterogeneous in size and shape. Immunofluorescence for EEA1 showed that EEA1 localized on the smaller endosomes, but scarcely localized on FDx-labeled macropinosomes earlier than 5-min age (Fig. 3A–C and J). A small population of macropinosomes was faintly positive for EEA1. At 10 min, EEA1 localized on the membrane of macropinosomes, which tended to tether/dock and sometimes to fuse with each other (Fig. 3D–F and K). In some cells, EEA1-positive macropinosomes partially elongated to form FDx-filled tubular extensions. EEA1 rarely localized to such tubular profiles.

thumbnail image

Figure 3. Dual labeling with macropinocytosis of FDx and immunofluorescence for EEA1. After 5-min pulse with FDx, the cells were chased at 37°C for 0 (AC, J), 5 (DF, K), 25 (GI), and 55 min (L), then processed for immunofluorescence using anti-EEA1 primary antibody and Alexa 594-labeled secondary antibody. Individual terms involved at 5-min (A–C, J), at 10-min (D–F, K), at 30-min (G–I), and 60-min macropinosomes (L). A, D, and G are FDx images. B, E, and H show immunolocalization of EEA1. C, F, and I show merged images of FDx and EEA1 immunofluorescence. J, K, and L are lower magnification of dual labeling images at 5-, 10-, and 60-min macropinosomes. EEA1 was not associated with the peripherally located macropinosomes just after 5-min pulse (A–C, J). EEA1 localized on the membrane of macropinosomes, which tethered/fused together at 10-min ages (D–F, K). Although macropinosomes decreased in size and number, FDx seemed to be more concentrated in centrally located macropinosomes at 30–60 min. EEA1 was still positive on these macropinosomes (G–I). Some smaller vesicles labeled with FDx were negative for EEA1 (I, L). n, nuclei. Scale bar = 10 μm.

Download figure to PowerPoint

After 30 min, macropinosomes had shrunk and clustered in the central region. Surprisingly, even after 30 or 60 min, FDx-labeled macropinosomes in the cell central region remained EEA1-positive. EEA1 appeared to be more concentrated on the centrally clustered shrunken macropinosomes at 30–60 min than on the large round macropinosomes at 10 min. Smaller vesicles or tubules filled with FDx were sometimes shown around macropinosomes, but they were negative for EEA1 (Fig. 3G–I and L). Examination of 100 FDx-labeled macropinosomes in each age category showed 24% EEA1-positive macropinosomes at 5 min, 65% at 10 min, and 97% at 30 min. In time-lapse video microscopic observations, the cell centrally located macropinosomes made repetitive transient contacts and detachments just like kiss-and-run fusion during chase times over 60 min, although they sometimes fused into one macropinosomes, resulting in a decrease in number (Fig. 4). QuickTime movie is available on our Web site (www.kms.ac.jp/∼anatomy2/SupplFig.4.mov).

thumbnail image

Figure 4. Time-lapse phase-contrast video microscopy of living A431 cells after EGF stimulation. Numbers indicate times (min) after the EGF addition. It is apparent that many macropisomes are formed by active ruffling at the cell periphery immediately after EGF addition. These macropinosomes gradually decrease in number and size with time. In the movie assembled from these images with 10-sec intervals, it is notable that macropinosomes continuously show transit fusion and fission (kiss-and-run) for a long time (60 min). n, nuclei. Scale bar = 10 μm

Download figure to PowerPoint

Macropinosomes Do Not Mature Into Lysosomes in A431 Cells

Cathepsin D, a lysosomal protease, can be used as a marker for late endosomes and lysosomes, since this enzyme is delivered from trans Golgi network to late endosomes and finally lysosomes by mannose-6-phosphate receptors. When the A431 cells were doubly immunolabeled for anti-EEA1 and anticathepsin D, EEA1 was never associated with cathepsin D-positive structures (Fig. 5). Furthermore, FDx-labeled macropinosomes, including their tubular extensions, were cathepsin D-negative at all stages.

thumbnail image

Figure 5. Relationships between EEA1-positive structures and lysosomes. AD: The cells were stimulated with 100 ng/ml EGF. After 5 (A), 10 (B), 30 (C), and 60 min (D) at 37°C, these cells were fixed and doubly immunostained for EEA1 (green) and cathepsin D (red). There was no colocalization between EEA1 (green) and cathepsin D (red) at any stages (A–D). E and F: Cells pulse-labeled with FDx were chased for 25 (E) and 55 min (F), fixed, and processed for immunofluorescence using a primary antibody against cathepsin D and Alexa 594-conjugated secondary antibody. Even at 30- and 60-min age, FDx-labeled macropinosomes (green) did not fuse with cathepsin D-positive structures (red). n, nuclei. Scale bar = 10 μm.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Macrophages, dendritic cells, osteoclasts, and Dictyostelium amoebae are capable of macropinocytosis (Swanson and Watts, 1995; Hacker et al., 1997; Maniak, 2001) and many other cultured cells perform macropinocytosis after stimulation with growth factors. EGF stimulates macropinocytosis in human A431 cells with no effect on endocytosis of transferrin or of EGF itself (West et al., 1989; Sandvig and Deurs, 1990; Hewlett et al., 1994). Macropinosomes arise from the deformation of ruffles in protrusions at cell margins (Swanson and Watts, 1995; Araki et al., 2000; Nobes and Marsh, 2000). This tendency is also supported by our findings on A431 cells in this study (Fig. 1). At 5-min pulse labeling, newly formed macropinosomes were roughly spherical and heterogeneous in size (Figs. 1–3). Subsequently, macropinosomes gathered in some places of the cells and fused with each other, resulting in large round macropinosomes.

The present study emphasized the subcellular localization of EEA1 in the macropinocytic pathway, in comparison with the receptor-mediated endocytic pathway. Wilson et al. (2000) demonstrated the subcellular localization of EEA1 by immunoelectron microscopy; EEA1 was located near cytoplasmic surfaces of early endosomal membranes, but was not localized on plasma membrane or clathrin-coated vesicles in the MDCK cells at early stages of receptor-mediated endocytosis. Likewise, our observations in A431 cells revealed that EEA1 localized neither to newly formed macropinosomes nor to plasma membrane. Filamentous actin (F-actin) is enriched in dynamic protrusions of the cell surface, circular ruffles, and nascent macropinosomes at an initial stage of macropinocytosis (Dowrick et al., 1993; Swanson and Watts, 1995; Hacker et al., 1997; Araki et al., 2000). Actin cytoskeleton reorganization is necessary for active ruffling and macropinosome formation, as most F-actin quickly dissociates from the membrane after macropinosome internalization (Hartwell and Weinert, 1989; Swanson and Watts, 1995; Araki et al., 2000). It is likely that F-actin removal occurs before the fusion of the macropinosomes with other macropinosomes or vesicles. Recruitment of EEA1 to macropinosome membranes could occur just after the dissociation of F-actin. Accordingly, EEA1 appearance on macropinosomes temporally corresponded to the beginning of the macropinosome fusion.

Ten-minutes-old macropinosomes tended to tether or fuse together into large round structures. Such tethering/fusion has been reported in early endosomes after receptor-mediated endocytosis (Dumas et al., 2001; Lawe et al., 2002). Tethering of endocytic vesicles is thought to represent a general intermediate preceding the docking, priming, and fusion of vesicles with target membranes of early endosomes (Christoforidis et al., 1999; Dumas et al., 2001). Furthermore, it was demonstrated that EEA1 played a role in the docking or fusion between incoming endocytic vesicles and adjoining early endosomes. EEA1 was identified as an early endosomal Rab5 effector protein by biochemical and histochemical analysis (McBride et al., 1999; Wilson et al., 2000; Dumas et al., 2001). Endosome tethering depends on the binding of EEA1 to PI(3)P through the FYVE domain at its extreme COOH terminus. In addition, EEA1 interacts with Rab5 through a motif adjacent to the FYVE domain to regulate subsequent fusion dynamics (Simonsen et al., 1998; Raiborg et al., 2001; Virbasius et al., 2001; Lawe et al., 2002; Murray et al., 2002). The PI 3-kinase inhibitor wortmannin causes dissociation of EEA1 from early endosomal membranes, suggesting that the PI 3-kinase product PI(3)P is required for binding of EEA1 to the membrane (Patki et al., 1997). A common mechanism could regulate macropinosome membrane fusion. In our previous study, PI 3-kinase inhibitors perturbed macropinosome formation from circular ruffles (Araki et al., 1996). However, this perturbation seemed not to result from the inhibition of EEA1-mediated fusion processes. Therefore, PI3-kinase products such as PI(3)P, PI(3,4)P2, and PI(3,4,5)P3 may regulate different processes in macropinocytosis by binding to specific domains of target proteins. Further studies are required to clarify the roles of different PI3-kinase products in the temporal and spatial regulation of macropinocytosis.

Using a technique of merged image of dual fluorescence using FDx and TRDx in A431 cells, Hewlett et al. (1994) reported that 64% of macropinosomes at 13-min age could fuse with the adjoining ones. Our immunofluorescence study revealed that EEA1 associated with 65% of macropinosomes at 10-min age. The rate of increase in EEA1-positive macropinosomes seemed to be almost identical with the frequency of macropinosome fusion. Taken together, these results suggest that EEA1 plays a pivotal role tethering molecules in homotypic fusion of macropinosomes, similar to the case of early endosome fusion.

FDx-labeled tubular extensions arose from macropinosomes, most frequently at 5- to 10-min age. Similar tubular profiles from macropinosomes were also observed in the MDCK cells, PtK2 cells, and amoebae and were thought to be recycling compartments for exocytosis (Dowrick et al., 1993; Hewlett et al., 1994; Maniak, 2001; Clarke et al., 2002). FDx-labeled smaller vesicles budding from macropinosomes were also seen at 30–60 min. These tubular profiles and smaller budding vesicles shown in A431 cells were negative for EEA1. It is possible that these EEA1-negative structures are recycling compartments, because EEA1 is associated with Rab5-positive early endosomes, but not rab11-positive recycling compartments in the receptor-mediated endocytic pathway of MDCK cells (Leung et al., 2000). The transport by vesicles and tubules with smaller diameters may differentially recycle membrane and content, since the surface/volume ratio of compartments increases as their diameter decreases.

Macropinosomes older than 30 min drastically decreased in number and size, but remained in the central regions of A431 cells. Unexpectedly, EEA1 did not dissociate from macropinosomes, even after longer chase. In contrast, EEA1 associated more frequently with macropinosomes located in the central region of the cells. Also, FDx appeared to be more condensed with increasing age of macropinosomes. It was noteworthy that EEA1 localized on macropinosomes as long as macropinosomes were present, while EEA1 localization on early endosomes was transient. This suggests that macropinosomes continue to fuse each other and to maintain features of early endosomes. In our preliminary observations by video microscopy, macropinosomes repetitively contacted and dissociated each other just like kiss-and-run, the continuous fusion and fission process of adjoining vesicles seen in phagolysosome biogenesis (Storie and Desjardins, 1996). The kiss-and-run homotypic fusion of macropinosomes continued over 1 hr, while we could see macropinosomes in the cell. The kiss-and-run mechanism is considered to contribute to efficient membrane turnover by exchanging some constituents by incomplete fusion via transient small pores (Storie and Desjardins, 1996; Ales et al., 1999; Verstreken et al., 2002). If exchange between macropinosomes occurs by kiss-and-run, then exchange should be more rapid for small than for large contents (Storie and Desjardins, 1996; Ducolus et al., 2002). It is conceivable that prolonged localization of EEA1 on centrally located vacuoles contributes to the continuous kiss-and-run fusion event of macropinosomes, resulting in size-dependent sorting of constituents of the macropinosome content.

FDx-labeled macropinosomes did not merge with cathepsin D-positive profiles in A431 cells, in contrast to macrophages (Racoosin and Swanson, 1993) and immature dendritic cells (Lutz et al., 1997). Moreover, EEA1 did not associate with cathepsin D-positive structures in A431 cells, indicating that macropinosomes never mature or fuse to late endosomes or lysosomes. The physiological role of macropinocytosis in A431 cells is still unclear; however, it is now apparent that this pathway is not for degradation of endocytosed materials. Although most macropinocytosed fluid is recycled back to extracellular medium, some solutes in macropinosomes would be concentrated and retained in the cell for a while. EEA1-mediated fusion of macropinosomes may play an important role of condensation and sorting of some compositions of the macropinosome content. EEA1 has been known to be one of the most specific early endosomal marker. However, Wilson et al. (2000) pointed out that EEA1 did not associate with entire early endosomes but predominantly associated with sorting subdomain of early endosomes from basolateral surface in polarized MDCK cell. The present study revealed that EEA1 localized on macropinosomes for a long time. Now, we should mind the possibility that EEA1 may present on some other endocytic compartments than usual early endosomes.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

We thank Prof. Joel A. Swanson, University of Michigan, for reading the article and for helpful suggestions. We are also grateful to Mrs. Y. Iwabu and Mr. K. Yokoi for technical and secretarial assistance.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED