CD43 Interacts With Moesin and Ezrin and Regulates Its Redistribution to the Uropods of T Lymphocytes at the Cell-Cell Contacts

THE POLARIZATION OF cell surface and intracellular structures is important during differentiation for immune response induction and cognate interactions between immune and other cells.1 T lymphocytes migrate through lymphocyte organs and extravasate into inflammatory sites to accomplish effector functions by sequential engagement of adhesion and signaling receptors.2-4 Effector activity of these cells involves plasma membrane clustering of LFA-1 and CD4 and cytoskeletal reorganization of microfilaments, microtubules, and talin.5-7 Migrating cells also display other polarized features, such as the localization of chemokine receptors towards the leading edge of the cell,8,9 an area of the cell that undergoes dynamic changes in the organization of the actin cytoskeleton,10 and the uropod at the opposite pole where intercellular adhesion molecules (ICAMs), CD44, and CD43 molecules are clustered.11 

CD43 (leukosialin, sialophorin) is a transmembrane glycoprotein that bears a heavily O-glycosylated extracellular N-terminal region and that is expressed by hematopoietic cells.12-15 In humans, low and high molecular weight (113 to 120 and 125 to 135 kD, respectively) glycoforms of CD43 are generated by differential O-glycosylation.16 Although its high negative charge may provide a repulsive barrier that interferes with cellular adhesion phenomena, this molecule was reported to be a counter-receptor for ICAM-1 and major histocompatibility complex (MHC) class I molecules.17,18 In this regard, CD43 has been described as promoting leukocyte aggregation.19,20 However, the functional analysis of T cells from CD43-deficient mice showed an in vitro increase in their proliferative response to mitogens and an enhanced homotypic cell aggregation capability, suggesting that CD43 could act as an antiadhesive molecule.21 In accordance, CD43 expression in target cells diminishes its susceptibility to cytolysis by T cytotoxic and natural killer cells.22,23Thus, the role of CD43 as an adhesion receptor still remains controversial. However, it is now clear that CD43 plays a role in the regulation of T-lymphocyte adhesiveness and cellular activation.24,25 In this regard, we previously described the regulatory role of CD43 on integrin-mediated T-cell adhesion to endothelium and proteins of extracellular matrix.26 On the other hand, the existence of a putative endothelial cell ligand for CD43 has recently been proposed, suggesting a role for this molecule in the selective traffic of T cells through lymphoid tissues.27 

Interaction of cell adhesion receptors with cytoskeletal proteins may propagate signals for cellular reorganization.1 Previous studies have demonstrated that CD43 colocalized with the ezrin-radixin-moesin (ERM) family of plasma membrane-cytoskeleton cross-linkers implicated in the formation of membrane dynamic structures such as filopodia, microvilli, and microspicules.28-30 During cytokinesis of cell division, these proteins interact directly or indirectly with the cytoplasmic domain of CD43 in the cleavage furrow.31 In this report, we explored the cellular localization of ERM proteins and CD43 in polarized motile lymphocytes as well as in cell aggregates induced by anti-CD43 antibodies. In addition, we studied the possible participation of CD43 in the formation of cell-cell contacts. Evidence is presented for the association of moesin and ezrin with CD43 and its coredistribution to the cellular uropods during T-cell aggregation and cell recruitment induced through CD43.

The anti-ICAM-3 HP2/19 (IgG2a), anti-CD43 HP2/21 (IgM) and TP1/36 (IgG1), anti-CD44 HP2/9 (IgG1), anti-CD45 RP2/21 (IgG1), anti-CCR2 MCP-1R03 (IgG1), antimoesin/radixin 38/87 (IgG1), and anti-HLA-A,B W6/32 (IgG1) monoclonal antibodies (MoAbs) have been described.9,32-36 The fluorescein isothiocyanate (FITC)-conjugated 2D1 anti-CD45 (IgG1) MoAb was purchased from Becton Dickinson (San Jose, CA). The moesin-specific polyclonal antiserum (pAb) 95/2 was raised in rabbits by immunization with recombinant human moesin and purified by affinity chromatography.37 The ezrin-specific pAb 90/3 was also raised in rabbits by immunization with purified human ezrin.38 The affinity-purified polyclonal antibodies 464 and 457, raised against unique peptides from murine ezrin and radixin, respectively,39 were kindly provided by Dr F. Solomon (Department of Biology and Center for Cancer Research, MIT, Cambridge, MA). Recombinant human (rh) moesin was obtained as described.38 Butanedione monoxime was purchased from Sigma Chemical Co (St Louis, MO). The 80-kD fibronectin fragment (FN80) was a generous gift of Dr A. Garcı́a-Pardo (Centro de Investigaciones Biológicas, Madrid, Spain). The chimeric ICAM-1-Fc, consisting of the extracellular domains of ICAM-1 and the Fc region of IgG₁, was obtained as described.11 

Human T lymphoblasts were prepared from peripheral blood mononuclear cells by treatment with phytohemagglutinin (PHA) 0.5% (Pharmacia, Uppsala, Sweden) for 48 hours. Cells were then washed and cultured in RPMI 1640 (Flow Laboratories, Irvine, Scotland) containing 10% fetal calf serum (FCS; Flow Laboratories) and 50 U/mL interleukin-2 (IL-2) kindly provided by Eurocetus (Madrid, Spain). T lymphoblasts cultured by 10 to 15 days were used in all experiments. These cells were analyzed by flow cytometry and their phenotype was 98% CD3 and 99% CD45RO; also, they show a heterogeneous expression of chemokine receptors CCR2 and CCR5.9,40 Melanoma cells and CD8⁺tumor-infiltrating lymphocytes (TIL) were isolated from melanoma specimens from patients with metastatic melanoma (Department of Pathology, Hospital General Universitario Gregorio Marañón, Madrid, Spain). These cells were cultured in AIM-V medium (Flow Laboratories) containing 10% HY-ultroser serum (Pharmacia) and 5,000 U/mL IL-2.40 Jurkat cells were grown in complete medium without the addition of IL-2.

Immunofluorescence experiments were performed as described.11 Briefly, 1 × 10⁶ T lymphoblasts were incubated in flat-bottommed, 24-well plates (Costar Corp, Cambridge, MA) in a final volume of 400 μL complete medium on coverslips coated with FN80 at 20 μg/mL. The polarization-inducing anti-CD43 HP2/21 (supernatant) and TP1/36 at 4 μg/mL were added and cells were allowed to remain in a cell incubator at 37°C and 5% CO₂ atmosphere. After 30 minutes, cells were fixed with 3.7% (wt/vol) paraformaldehyde in phosphate-buffered saline (PBS), pH 7.4, at room temperature and then rinsed in TBS. Fixed cells were incubated with specific MoAb or pAb. After washing, cells were stained with an FITC-labeled rabbit F(ab)′₂ antimouse IgG, donkey antirabbit IgG (Pierce, Rockford, IL), or Cy3-tagged goat antimouse IgG (Amersham, Pittsburgh, PA). For double immunofluorescence analysis, cells were saturated with 10% normal mouse serum in TBS. Coverslips were then incubated with biotinylated MoAb, followed by washing and labeling with either FITC-extrAvidin (Sigma) or Cy3-streptavidin.

Homotypic cell aggregation was performed as previously described.33 Briefly, T lymphoblasts were incubated in flat-bottommed, 24-well microtiter plates (Costar Corp) at 4 × 10⁶/mL in a final volume of 0.5 mL of complete medium. Then, 1:20 dilution of anti-CD43 HP2/21 supernatant was added and cells were incubated at 37°C and 5% CO₂ atmosphere. Aggregation was then determined at different periods of time by direct visualization of the plate with an inverted microscope and counting the free cells of at least five randomly chosen areas of 0.025 mm². Results were expressed as a percentage of aggregated cells.

For immunofluorescence staining, after either 10 minutes (small cell aggregates) or 30 minutes (large cell aggregates) of incubation, cells were fixed with 3.7% paraformaldehyde in PBS for 10 minutes at room temperature and rinsed in TBS. To directly visualize CD43, cells were stained with a 1:50 dilution of an FITC-labeled rabbit F(ab)′₂ antimouse IgG (Pierce). To visualize moesin, after cell permeabilization with 0.1% Triton X-100, cell aggregates were incubated with the biotinylated antimoesin 38/87 MoAb followed by washing and labeling with 1:1,000 dilution of Cy3-streptavidin. Cells were observed using a Nikon Eclipse-800 photomicroscope with 60× and 100× oil immersion objectives. Preparations were photographed with Ektachrome 400 film (Eastman Kodak Co, Rochester, NY).

Video microscopy analysis was performed using a Nikon Diaphot 300 inverted microscopy equipped with a Sony SSC-M350CE CCD black and white videocamara coupled to a Sony SVT-5000P time lapse videocassette recorder and a Sony PVM-1453MD video monitor (Japan).

T lymphoblasts were allowed to attach for 20 minutes at 37°C to 35-mm plastic petri dishes (Costar Corp) previously coated with ICAM-1Fc (10 μg/mL) in the presence of anti-CD43 HP2/21. After the addition of a second layer of T lymphoblasts, cell-cell interactions were recorded for 1 hour under phase contrast using a 20× objective. Images were acquired every 3.2 seconds, and sequential frames were digitalized by using Optimas software (Optimas Corp, Bothell, WA).

T lymphoblasts either untreated or treated with the anti-CD43 HP2/21 MoAb adhered to FN80-coated dishes were lysed by incubation for 20 minutes in 1.5 mL RIPA buffer containing 0.1% sodium dodecyl sulfate (SDS), 0.5% deoxycholate, 1% Nonidet P-40, 150 mmol/L NaCl, 50 mmol/L Tris (pH 8), 1 mmol/L p-amidino phenylmethyl sulfonyl fluoride (PMSF), and 15 μg/mL leupeptin. Cell lysate was removed from the dish with a rubber policeman and then precleared by centrifugation at 10,000g for 20 minutes. Immunoprecipitations were performed with 45 μL of TP1/36 anti-CD43 or RP2/21 anti-CD45 (control) MoAb directly coupled to Sepharose 4B beads (Pharmacia) at 1 mg/mL. Proteins bound to Sepharose beads were eluted by boiling in sample buffer, subjected to SDS 7.5% polyacrylamide gel electrophoresis (PAGE) under reducing conditions, and transferred onto a nitrocellulose membrane (Millipore, Bedford, MA) in Tris-Glycine-Methanol buffer during 25 minutes at 17 V using a Transfer-Blot SD Semi-Dry Transfer Cell (Bio-Rad Laboratories, Hercules, CA). To detect moesin, membranes were soaked overnight in TBS containing 3% bovine serum albumin (BSA) and washed three times with TBS-0.1% Tween 20 during 15 minutes, followed by 1 hour of incubation with a 1:1,000 dilution of rabbit antimoesin 95/2 pAb. After three washes, blots were incubated with a peroxidase-conjugated goat antirabbit IgG (Pierce) and developed using an enhanced chemiluminiscence system (Amersham Corp). To detect the presence of ezrin coprecipitated with CD43, the 95/2 and secondary antibodies were removed from the membrane by treating it with a stripping buffer (100 mmol/L 2-mercaptoetanol, 2% SDS, 62.5 mmol/L Tris-HCl, pH 6.7) at 50°C for 30 minutes. Then, blots were reprobed with the antiezrin 90/3 pAb.

The cytoplasmic region of CD43 without the nucleotides coding for the first seven residues plasma membrane-proximal was amplified by polymerase chain reaction (PCR) and cloned as Sal I-NotI fragments into pGEX-4T (Pharmacia LKB Biotechnology, Uppsala, Sweden). For amplification of the CD43 cytoplasmic region (from residues T294 to P400), the primers SalI43T8 (5′-CAATTTGTCGACAACTGGGGCCCTCGTGCTGAGC-3′) and NotCD43 (5′-ATAAGAATGCGGCCGCCGACACTTAAGGGGCAGCC-3′) were used.

Expression of GST-fusion proteins in DH1OB cells and purification were performed following the manufacturer's instructions. The proteins were stored at 4°C on glutathione Sepharose 4B beads as a 50% slurry in 10 mmol/L Tris, pH 7.4, 140 mmol/L NaCl, 0.5% Triton X-100, 0.02% azide, and protease inhibitors (Complete Protease Inhibitors; Boehringer Mannheim Corp, Indianapolis, IN). The amount of fusion protein was stimated by coomassie blue staining of SDS-PAGE.

Jurkat cells (2 × 10⁷ cells/mL) were labeled overnight with a mixture of [³⁵S]methionine/cysteine (500 μCi) in methionine/cysteine-free RPMI 1640 medium supplemented with 10% dialyzed fetal bovine serum. ³⁵S-labeled Jurkat cells were disrupted in 1 mL lysis buffer (10 mmol/L Tris/HCl, pH 7.6, 150 mmol/L NaCl, 1% Nonidet P-40, 5 mmol/L EDTA, 50 mmol/L sodium fluoride, 0.4 mmol/L sodium orthovanadate, 10 mmol/L iodoacetamide, 5 mmol/L sodium pyrophosphate, 1 mmol/L penoxymethyl sulphofluoride, and 10 mg/mL aprotinin, leupeptin, pepstatin A, chymostatin, and α1-antitrypsin) on ice for 10 minutes, followed by 10 minutes of microcentrifugation. The supernatant was mixed with approximately 50 μg of GST-fusion proteins attached to glutathione Sepharose 4B beads for 12 hours at 4°C. The beads were collected by centrifugation and washed twice with lysis buffer alone, twice with lysis buffer plus 0.1% SDS, twice with lysis buffer plus 0.65 mol/L NaCl, and twice with lysis buffer alone. The beads were then subjected to SDS-PAGE 8% under reducing conditions.

We have previously reported the regulatory role of CD43 in cell morphology and its redistribution towards the uropods of polarized T lymphocytes.26 Herein, we investigated whether the engagement of CD43 is also capable to redistribute other membrane receptors as well as cytoskeletal components to different specialized compartments of the cell. Fluorescence microscopy analysis of T lymphoblasts adhered to FN80 and stimulated with anti-CD43 MoAb showed that, during cell polarization, the chemokine receptor CCR2 redistributed to the leading edge of the cell (Fig 1i, A and B, and 1ii), whereas CD43 was concentrated at the cellular uropod (Fig 1i, A). In contrast, CD45 did not redistribute during CD43-induced cell polarization (Fig 1i, C).

Immunofluorescence staining of ERM cytoskeletal proteins using specific monoclonal and polyclonal antibodies showed that, in anti-CD43–treated T lymphocytes, CD43, ezrin, and moesin (Fig2i, A, B, and D, respectively) redistributed to the cellular uropods, whereas radixin showed a staining pattern at the uropod neck (Fig 2i, C). Furthermore, double immunofluorescence staining showed that CD43 (Fig 2ii, A and C) colocalized with moesin (Fig 2ii, B) and ezrin (Fig2ii, D) at the uropod tips. In addition, a weak signal of ezrin was also detected at the leading edge (Fig 2ii, D).

To further analyze the spatial codistribution of CD43 and moesin, we performed immunofluorescence analysis on CD8⁺ TIL adhered to FN80. These in vivo activated cells (CD45RO⁺, CD69⁺) were isolated from patients with melanoma. In these cells, CD43 and moesin were also colocalized at its pronounced uropods (Fig 3i). When TIL were allowed to adhere to melanoma cells, they displayed the polarized morphology typical of motile cells with CD43 clustered in the cellular uropods (Fig 3ii).

To determine whether moesin and ezrin physically interact with CD43, cell lysates from T lymphoblasts either nonstimulated or stimulated with the anti-CD43 MoAb HP2/21 were immunoprecipitated with the TP1/36 anti-CD43 MoAb or with the RP2/21 anti-CD45 MoAb as control, transferred to nitrocellulose membranes, and blotted with pAb against either moesin (Fig 4A) or ezrin (Fig 4B). Coprecipitation of moesin and ezrin was observed in immunoprecipitates from CD43 (Fig 4A and B, lanes 3 and 4, respectively), whereas neither moesin nor ezrin were detected in the CD45 immunoprecipitates (Fig 4A and B, lanes 1 and 2). An increase in the amount of moesin associated to CD43 was detected in immunoprecipitates from cells pretreated with the polarization-inducing HP2/21 anti-CD43 MoAb compared with untreated cells (Fig 4A, lanes 3 and 4). In contrast, comparable amounts of ezrin were found associated to CD43 in treated and untreated cells (Fig 4B, lanes 3 and 4).

To further confirm the interaction of CD43 with moesin and ezrin, we performed precipitation studies with a GST fusion protein containing the CD43 cytoplasmic region (Fig 5). All attempts to induce the expression in DH1OB cells of a GST-fusion protein containing the full-length cytoplasmic region of CD43 failed. However, we were able to induce the expression of a GST-fusion protein bearing the cytoplasmic region of CD43 lacking the first seven N-terminal residues, designated as GST-CyD7CD43 (Fig 5A). We found that the GST-CyD7CD43 fusion protein specifically precipitated two bands of 78 and 80 kD from metabolically labeled Jurkat T-cell lysates, whereas no protein band was observed in precipitates from GST control protein (Fig 5B). Western blot analysis performed in parallel showed that the 78- and 80-kD proteins bound to GST-CyD7CD43 corresponded to moesin and ezrin, respectively (Fig 5C). These results further indicate that the cytoplasmic tail of CD43 interacts with moesin and ezrin.

The induction of homotypic leukocyte aggregation by anti-CD43 MoAb has been described.20,41,42 In this regard, we have previously selected a number of anti-CD43 MoAb for their ability to promote rapid and strong homotypic aggregation of normal T lymphoblasts and U937 myelomonocytic cells.26 However, little is known about the dynamics of aggregation and the nature of cell-cell interactions established during this phenomenon. In an attempt to understand this issue, we performed time-lapse videomicroscopy studies of homotypic T-lymphoblast aggregate formation induced by the anti-CD43 MoAb HP2/21 (Fig 6i). Five minutes after addition of anti-CD43, T cells initiated locomotion on the substratum, displayed a polarized morphology, and established early cell-cell interactions with other polarized and spherical cells, mainly through their uropods, forming motile small aggregates (Fig 6i). One hour later, motile small cell aggregates contacted with other small aggregates through externally exposed cell uropods generating large cellular aggregates (Fig 6i). Immunofluorescence studies performed in parallel showed that CD43 as well as moesin were concentrated at the uropods of small cell aggregates that were mediating cell-cell interactions (Fig 6ii, A and B, and iii, A, respectively). On the other hand, in large cell aggregates, CD43 and moesin were located in the uropods projected towards the outer milieu (Fig 6iii, C, and iii, B, respectively) as well as in cell-cell contacts. These results strongly suggested that the clusters of CD43 projected by the prominent uropods could facilitate the interaction of this molecule with a putative ligand to allow formation of cell aggregates.

Polarized T cells express linear arrays of myosin at the uropod neck,43 and we have described that the disruption of this structure by the butanedione monoxime prevents uropod formation and redistribution of both CD43 and moesin.26,44 To investigate the role of myosin in the induction of T lymphoblast aggregation triggered through CD43, we performed time-lapse videomicroscopy of T-cell aggregation in the presence of butanedione monoxime. At short times, although cell aggregation was observed in the absence of this agent (Fig 6i), the addition of this drug prevented the acquisition of polarized morphology of T cells adhered to FN80 and the formation of cell clusters (Fig 7A). At longer times, when the butanedione effect start to vanish, some cells become polarized participating in small aggregates through its uropods (Fig 7A and B). We have also tested the role of colchicine and cytochalasin D on cell aggregation induced by the anti-CD43 HP2/21 MoAb. Interestingly, whereas cytochalasin D inhibit cell aggregation, no significant inhibition was observed under colchicine treatment (Fig7B). These results indicate that myosin plays a central role in the establishment of cell-cell interactions required for aggregate formation induced by CD43.

We have recently described the involvement of T-cell uropods in the transport and recruitment of other lymphocytes as well as the role of ICAM-3 and chemokines in the induction of this phenomenon.40 To investigate whether similar phenomena take place upon CD43 engagement, we studied this issue by time-lapse videomicroscopy. A first layer of T cells was allowed to attach to ICAM-1-Fc–coated surface plates and then induced to develop uropods with the HP2/21 anti-CD43 MoAb. A second layer of T cells was then added and cellular interactions were filmed. Interestingly, cells of the second layer (phase-bright cells) were contacted, trapped, and transported by cells of the first layer (phase-dark cells) locomoting on the ICAM-1-Fc–coated surface (Fig 8A). Quantification of cells of the second layer recruited and transported by the adhered cells of the first layer showed that the treatment with anti-CD43 induced a 10-fold increase in cell recruitment compared with unstimulated cells (Fig 8B).

CD43 is a heavily charged transmembrane sialomucin that provides stimulatory signals inducing lymphocyte proliferation and IL-2 secretion24,45 and costimulates T lymphoblasts in a CD28-independent manner.25 In contrast, a negative regulation of T-cell activation and adhesion through CD43 has also been reported.21,22 However, it has been described that anti-CD43 MoAb are able to trigger homotypic leukocyte aggregation in both CD18-dependent and CD18-independent manner.19,20,41,46Because cell aggregation and polarization studies have become useful assays for analyzing leukocyte motility and activation in vitro, we decided to study the role of CD43 in these assays to further understand the biological function of this molecule.

Previously, we have reported that proaggregatory and activating anti-CD43 MoAbs enhance β1 and β2 integrin-mediated T-cell adhesion to both endothelial and ECM proteins and induce T-cell polarization with redistribution of CD43 to cellular uropods.26 We demonstrate herein that CD43 plays an important regulatory role in T-cell polarization, inducing the appearance of specialized cellular domains with redistribution of the CCR2 chemokine receptor to the leading edge and ERM actin-binding proteins to the uropod of polarized T lymphocytes. For the ERM proteins analyzed, ezrin and moesin colocalized with CD43 at the uropod tips, whereas radixin was detected at the uropod neck. It is worth mentioning that the data regarding redistribution of CD43 and ezrin contrast with our previous results showing no colocalization of ezrin with ICAM-3 at the uropods of T cells treated with chemokines.44 This discrepancy seems to be due to the different polyclonal antibodies used in both studies; whereas 90/3 pAb was generated against purified human ezrin, the affinity-purified pAb 464 was raised against an unique peptide of mouse ezrin.44 Therefore, it is evident that the bright immunostaining of cell uropods produced by the 90/3 pAb indeed corresponds to the redistribution of ezrin to this structure. Furthermore, the CD43/ezrin codistribution found by us agrees with previous data from other investigators regarding the presence of ezrin in the uropods of mouse T cells and with the colocalization of ERM proteins and CD43 in the cleavage furrows during cytokinesis of mouse thymocytes in division.47,48 Interestingly, our data demonstrate that ezrin and moesin physically interact with CD43 in unstimulated T lymphocytes, yet only moesin/CD43 interaction was increased after CD43 stimulation. Moreover, the coprecipitation studies with the GST-CyD7CD43 fusion protein demonstrate a direct interaction of moesin and ezrin with the cytoplasmic region of CD43. In this regard, other adhesion molecules, such as CD44 and ICAM-3, have been reported to interact with moesin in motile T cells, and this association is increased under conditions that induce cell polarization such as the treatment with chemokines.44 Thus, it appears that the process of moving membrane molecules towards the trailing edge of the cell is directed by the submembranous cytoskeleton.49 By linking adhesion molecules to submembranous cytoskeleton, moesin could promote the redistribution of membrane complexes towards the cellular uropod.

It has been described that cell aggregation can be induced in T lymphoblasts upon engagement of a large array of cell adhesion molecules, including CD43.33,50,51 We have undertaken a dynamic assessment of cell aggregation by time-lapse video recording of CD43-stimulated T lymphocytes. Upon treatment with anti-CD43 MoAb, the formation of large cell aggregates takes place by congregation of small ones. In small aggregates, cells contact each other mainly through their uropods, where moesin and CD43 colocalize, suggesting that moesin/CD43 interaction may have a role in the initial phase of the establishment of cell-cell contacts, as it has been described for ICAM-3.33 In accordance, a similar mechanism of cell-cell interactions has been described in NK-resistant cells, which after ezrin transfection, display an uropod where ICAM-2 is redistributed, allowing cellular interactions and becoming NK-sensitive.47Similarly, uropod formation, moesin/CD43 interaction, and CD43 redistribution during T-cell polarization induced through CD43 would increase molecular density and accessibility of CD43 in the cellular uropod enabling interactions with putative ligands, facilitating homotypic cell aggregation. In this regard, the existence of cellular ligands for CD43 involved in leukocyte endothelial cell interactions has recently been postulated on the basis of in vivo inhibition of lymphoid cell homing with an anti-CD43 MoAb.27 

In previous studies using the myosin-disrupting drug butanedione monoxime, we found that myosin is involved in uropod formation, cell aggregation induced by anti-ICAM-3 MoAb, and adhesion receptors and moesin redistribution to the uropod.11,26,43,44 Our data indicate that myosin is also necessary for the cell-cell interactions that occur during the homotypic T-cell aggregation induced through CD43, reinforcing the issue that uropod formation is required to optimize cell-cell interactions.

We previously reported that cellular uropods induced by both chemokines and activating anti-ICAM-3 MoAb are able to mediate the recruitment of lymphocytes and transports them through extracellular matrix layers or endothelial cells specialized in lymphocyte extravasation.40 Likewise, CD43-mediated cell polarization and CD43 and moesin redistribution towards uropods are involved in the migration of T lymphoblasts on ICAM-1 as well as in the capture and transport of additional bystander cells. Because the phenotype of T lymphoblasts used in these studies is similar to memory and effector T cells (95% CD45RO, 98% CD3, and 96% CD43^high), our work agrees with a previous report showing that the high expression of CD43 by memory T cells facilitate the extravasation of these cells.20 In this regard, it is conceivable that the highly polarized morphology of TIL, with a pronounced uropod where most of CD43 molecules are redistributed, facilitates the motility of these lymphocytes through cellular layers amplifying the recruitment of leukocytes as well as its interaction with tumoral cells. Altogether, our findings on cell aggregation induced through CD43 suggest that this molecule is a proadhesive rather than antiadhesive molecule and provide evidence for the role of CD43 as a membrane receptor able to interact with putative counter-receptor(s), mediating cell-cell contacts through uropods in leukocyte intercellular interactions.

Recently, it has been postulated the involvement of CD43 in the selective binding and recruitment of T lymphocytes to lymphoid organs as well as in the existence of an unidentified vascular ligand for CD43.27 In neutrophils, the interaction of P-selectin glycoprotein ligand-1 (PSGL-1) with P-selectin is uncoupled upon cell activation with the concomitant redistribution of PSGL-1 towards the cellular uropod.52 On the other hand, PSGL-1 redistribution towards the uropod of neutrophils after chemotactic activation could be responsible for cell-cell interactions between neutrophils and activated platelets.53 Because CD43, as PSGL-1, is a sialylated mucin, it is feasible that, in T lymphocytes, the redistribution of CD43 to the cellular uropod by its interaction with ERM-mediated cytoskeleton would facilitate transendothelial migration and recruitment of other T lymphocytes by interaction through their uropods.

The authors thank Dr P. Lauzurica for critical readings of the manuscript. We are greatfully indebted to M.C. Montoya and E. Fernández-Villareal for their help and advice with different techniques and to R. Tejedor for preparing recombinant ICAM-1-Fc.