Molecular Events in Transmembrane Signaling via E-selectin

E-selectin is a cytokine-inducible adhesion molecule that is expressed by activated endothelial cells at sites of inflammation. In addition to supporting rolling and stable arrest of leukocytes, there is increasing evidence that E-selectin functions in transmembrane signaling into endothelial cells during these adhesive interactions. We have previously shown that adhesion of HL-60 cells (which express ligands for E-selectin), or antibody-mediated cross-linking of E-selectin, results in formation of a Ras/Raf-1/phospho-MEK macrocomplex, extracellular signal-regulated protein kinase (ERK1/2) activation, and c-fosup-regulation. All of these downstream signaling events appear to require an intact cytoplasmic domain of E-selectin. Here we demonstrate that tyrosine 603 in the cytoplasmic domain of E-selectin is required for the E-selectin-dependent ERK1/2 activation. Tyrosine 603 plays an important role in mediating the association of E-selectin with SHP2, and the catalytic domain of SHP2 is, in turn, critical for E-selectin-dependent ERK1/2 activation. An adapter protein complex consisting of Shc·Grb2·Sos bridges between SHP2 and the Ras·Raf·phospho-MEK macrocomplex. These molecular events thus outline a mechanism by which cross-linking of E-selectin by engagement of ligands on adherent leukocytes can initiate a multifunctional signaling pathway in the activated endothelial cell at sites of inflammation.

Cultured Cells-Human umbilical cord vein endothelial cells (HU-VEC) were isolated and established in culture as previously described (18). Primary cultures were serially passaged (1:3 split ratio) and maintained in Medium 199 buffered with 25 mmol/liter HEPES buffer and supplemented with 20% fetal bovine serum (FBS), endothelial cell growth factor (25 g/ml), and porcine intestinal heparin (50 g/ml). For experimental use, subcultured (passage 2 or 3) endothelial cells were plated on gelatin-coated 35-or 100-mm tissue culture dishes (Difco Laboratories, Detroit, MI). HL-60, a human promyelocytic leukocyte cell line, was obtained from the American Type Culture Collection (ATCC) and grown in RPMI 1640 medium supplemented with 10% FBS, 100 units/ml penicillin, 100 g/ml streptomycin, and 20 mM L-glutamine. JY human lymphocytic cells, kindly provided by Dr. T. A. Springer (Center for Blood Research, Boston, MA), were maintained in RPMI 1640 medium supplemented with 10% FBS, 100 units/ml penicillin, 100 g/ml streptomycin, and 20 mM L-glutamine. COS-7 cells, a monkey kidney fibroblast cell line, was obtained from the American Type Culture Collection (ATCC) and grown in Dulbecco's modified Eagle's medium supplemented with 10% FBS, 100 units/ml penicillin, 100 g/ml streptomycin, and 20 mM L-glutamine.
Generation of E-selectin Tyrosine Mutants-A PCR-directed mutagenesis approach was used to generate specific mutants of E-selectin. Two tyrosine mutants (Tyr 603 3 Phe, Tyr 608 3 Phe) were generated as follows. A forward primer F2 (5Ј-GGTTTGGTGAGGTGCTCCATTC-3Ј) and two different reverse primers, R8 (5Ј-GCGTTAACTTAAAGGAT-GAAAGAAGGCTTTTGG-3Ј) and R19 (5Ј-GCGTTAACTTAAAGGATG-TAAGAAGGCTTTTGGAAGCTTCC-3Ј) were used to amplify two fragments using wild-type E-selectin cDNA as the template. Each reverse primer (R8 and R19) contained both a stop codon and a HpaI restriction site at its 3Ј end in order to make mutant E-selectin cytoplasmic domain fragments. In both cases, the amplified fragments were ligated into the Topo-TA cloning vector (Invitrogen Corp., San Diego, CA), and the mutated sequences were confirmed by sequence analysis. The fragments were gel purified, and prepared by sequential digestion with EcoRI and HpaI. They were then used to replace the corresponding EcoRI/HpaI fragments of the wild-type E-selectin cDNA in the expression vector, pCDM8 (Invitrogen).
Immunoprecipitation-After treatment, cells were rinsed with icecold Dulbecco's phosphate-buffered saline and lysed in either Triton lysis buffer (for E-selectin and SHP2 association) (1% Triton X-100, 10 mM Tris, pH 7.5, 100 mM NaCl, 50 mM NaF, 1 mM Na 3 VO 4 , 2 mM phenylmethlysulfonyl fluoride, 21 g/ml aprotinin) (19) or Nonidet lysis buffer (for adapter proteins) (25 mM HEPES, 0.1% SDS, 0.5% sodium deoxycholate, 3% Nonidet P-40, 125 mM NaCl, 10 mM NaF, 10 mM Na 3 VO 4 , 10 mM sodium pyrophosphate, 20 g/ml aprotinin, 20 g/ml leupeptin) (20) on ice for 5 min. Cells were then scraped off the plates and collected in microcentrifuge tubes. The total cell lysates were centrifuged at 14,000 rpm for 15 min at 4°C and the supernatant was pre-cleared with protein A/G-agarose for 1 h at 4°C. Aliquots of the supernatant were subjected to immunoprecipitation with various antibodies overnight at 4°C. Twenty-five l of protein A/G-agarose beads were added to the incubation for an additional 1 h at 4°C. The immunocomplex was then washed three times with the Triton lysis buffer or Nonidet lysis buffer and resuspended in 50 l of the same buffer. Twenty-five microliters of 3 ϫ sample buffer was added and samples were boiled at 100°C for 5 min. Each tube was then vortexed and centrifuged at 14,000 rpm for 2 min.
In Vitro Immunocomplex Phosphatase Assay-After treatment, cells were lysed on ice with Triton buffer for 5 min and centrifuged at 14,000 rpm for 15 min at 4°C. The supernatant was then pre-cleared with protein A/G-agarose for 1 h at 4°C and the resultant supernatant was subjected to immunoprecipitation with H18/7 overnight at 4°C. Twenty-five microliters of protein A/G-agarose beads were added to the incubation for an additional 1 h at 4°C. The immunocomplex was then washed three times with the Triton lysis buffer and re-suspended in 50 l of the phosphatase assay buffer (80 mM MOPS, 10 mM EDTA, 10 mM dithiothreitol, 1 mM Na 3 VO 4 , pH 7.0) (19) and incubated at 37°C for 30 min. After the reaction, 25 l of 3 ϫ SDS buffer were added to the beads and the proteins were eluted from the beads by boiling for 5 min.
Immunoblotting-Aliquots (25 l) of immunoprecipitates were separated on a 7.5 or 12% SDS-polyacrylamide electrophoresis gel and then transferred to a nylon membrane (Millipore, Bedford, MA). Membranes were blocked with either 5% bovine serum albumin (when anti-phosphorylated tyrosine antibody was used) or 5% nonfat milk in TTBS (20 mM Tris, 138 mM NaCl, 0.5% Tween 20, pH 7.6) for 1 h at room temperature and then incubated with various primary antibodies (1: 1000 diluted in blocking buffer), including RC20, SHP2, PY99, Shc, Grb2, and Sos, for an additional 1 h at room temperature. After three washes with TTBS, membranes were incubated with a horseradish peroxidase-conjugated polyclonal goat anti-mouse or goat-anti-rabbit antibody (1:1000) (Santa Cruz Biotechnology, Inc.) in TTBS for an additional hour at room temperature, and again washed three times in TTBS. The labeled proteins were visualized using an enhanced chemiluminscence kit (Amersham Bioscience, Inc.).

Cross-linking Cell Surface E-selectin Molecules Induce Tyrosine Phosphorylation on E-selectin and Its
Association with SHP2-There are two tyrosine residues on the cytoplasmic domain of E-selectin that could potentially play a role in mediating cytosolic events associated with transmembrane signals. To examine whether these tyrosine residues actually become phosphorylated as a consequence of perturbation of the extracellular domain of E-selectin, we compared the state of tyrosine phosphorylation of E-selectin following antibody-mediated cross-linking of cell surface E-selectin (Fig. 1A, upper FIG. 1. E-selectin becomes tyrosine phosphorylated and associated with SHP2 upon antibody-mediated cross-linking. A, HUVEC monolayers were activated with IL-1␤ (10 units/ml, 4 h, 37°C) and treated with H4/18 (a mouse mAb against E-selectin, 10 g/ml, 30 min, 4°C) or W6/32 (a mouse mAb against HLA class I molecules, 10 g/ml, 30 min, 4°C), followed by a goat anti-mouse IgG (GAM-IgG) (1:200, 37°C) for the times indicated (0, 2, 15, and 30 min) to induce selective cell surface molecular cross-linking. Tyrosine phosphorylation of E-selectin was detected by sequential immunoprecipitation with a second E-selectin-specific mouse mAb, H18/7, and immunoblotting with a phosphotyrosine monoclonal antibody (RC20) (upper panel). The same blot was stripped and blotted with a goat anti-mouse E-selectin antibody (lower panel). B, HUVEC surface E-selectin molecules were crosslinked with mAb H4/18, followed by a goat anti-mouse IgG for the times (0, 2, 15, and 30 min) indicated, and SHP2 association with E-selectin detected by sequential immunoprecipitation and immunoblotting. C, COS-7 cells were transfected with WT-E-selectin, Tyr 603 3 Phe mutant, or Tyr 608 3 Phe mutant. Following selective cross-linking (H4/18ϩIgG or HϩG) of cell surface E-selectin for 2 min, SHP2 association with E-selectin was detected by sequential immunoprecipitation and immunoblotting. C, COS-7 cells; Mock, mock transfected COS-7 cells. panel). Tyrosine phosphorylation of E-selectin was detectable 2-30 min after cross-linking of cell surface E-selectin (2 min, 1-fold; 15 min, 4-fold; 30 min, 3-fold increase over IL-1␤ alone, densitometry results), but not as much after cross-linking of HLA class I molecules (2 min, no change; 15 min, 2-fold; 30 min, 1-fold increase over IL-1␤ alone). Comparable amounts of E-selectin were present in these immunocomplexes (Fig. 1A, lower panel).
It has been shown that the protein-tyrosine phosphatase SHP2 becomes associated with tyrosine-phosphorylated receptors and regulates signaling from cell surface receptors (21,22). Therefore we examined the association between E-selectin and SHP2 following antibody-mediated cross-linking of cell surface E-selectin. Co-immunoprecipitation experiments revealed that SHP2 became associated with E-selectin, as early as 2 min after cross-linking and was still detectable at 30 min (Fig. 1B). The time difference between the E-selectin tyrosine phosphorylation (15 min) and SHP2 association with E-selectin (2 min) could have been attributed to that the anti-SHP2 antibody was more sensitive to detect SHP2 than the anti-phosphotyrosine antibody to detect E-selectin phosphorylation. To determine which tyrosine residue(s) on the cytoplasmic domain of Eselectin were responsible for the observed association between SHP2 and E-selectin, we generated two single tyrosine point mutants, Tyr 603 3 Phe and Tyr 608 3 Phe, and transfected these mutants into COS-7 cells. Forty-eight to 72 h post-transfection, cell surface E-selectin molecules were cross-linked using a monoclonal antibody specific for E-selectin (H4/18), and the association between SHP2 and E-selectin was again examined using co-immunoprecipitation. As seen in Fig. 1C, mutation of Tyr 603 essentially ablated the association of SHP2 with E-selectin induced by the cell surface E-selectin cross-linking. The Tyr 608 mutant exhibited no difference in SHP2 association under either basal or cross-linking conditions. A fluorescence immunoassay documented that equivalent amounts of E-selectin were expressed at the cell surface in both the Tyr 603 -and Tyr 608 -transfected cells (data not shown).
Tyrosine Residue 603 Is Essential for E-selectin-dependent ERK1/2 Activation-We had previously shown that cross-linking cell surface E-selectin resulted in the activation of ERK1/2 (7). To determine whether the two tyrosine residues in the cytoplasmic domain of E-selectin play any role in the subsequent ERK1/2 activation, we transfected these single tyrosine point mutants into COS-7 cells, cross-linked cell surface Eselectin and measured the level of phosphorylated ERK1/2. As seen in Fig. 2, the mutation, Tyr 603 3 Phe resulted in a complete loss of ERK1/2 activity, while the Tyr 608 3 Phe mutant had no impact on the activation of ERK1/2 (upper panel). Equal amounts of ERK1/2 protein were present in each sample as determined by stripping and re-probing using an antibody that recognizes total ERK1/2 (lower panel). A fluorescence immunoassay documented that comparable levels of cell surface Eselectin were expressed on both types of transfected COS-7 cells (data not shown). Therefore, tyrosine residue 603 in the cytoplasmic domain of E-selectin appears to play an essential role in the E-selectin-dependent activation of ERK1/2.
The Catalytic Domain of SHP2 Plays an Important Role in the E-selectin-dependent ERK1/2 Activation-To determine whether SHP2 functions in E-selectin-dependent ERK1/2 activation, we co-transfected WT-E-selectin together with either wild type-SHP2 (wSHP2) (Fig. 3, lanes 5 and 6) or a mutant SHP2 (muSHP2) (lanes 7 and 8), which lacks catalytic activity, into COS-7 cells. As control, we co-transfected the empty vector, PJ3, with WT-E-selectin (lanes 3 and 4). Cell surface Eselectin molecules then were selectively cross-linked using an E-selectin mAb (H4/18) and phosphorylated ERK1/2 was measured as an index of the ERK1/2 activity. Fig. 3 shows that co-transfection of muSHP2 with WT-E-selectin resulted in an inhibition of cross-link-mediated ERK1/2 activation. Comparable amounts of both wSHP2 and muSHP2 were detected by Western blot (data not shown). Thus, this dominant negative experiment indicated that the catalytic property of SHP2 plays an important role in downstream E-selectin-dependent ERK1/2 activation.
SHP2 Associated with E-selectin Is Enzymatically Active and Becomes Tyrosine Phosphorylated-SHP2 is a protein-tyrosine phosphatase. To exam whether the SHP2 associated with the cytoplasmic domain of E-selectin is enzymatically active, Eselectin was immunoprecipitated from IL-1␤-activated, mAb H18/7 cross-linked HUVEC and subjected to immunocomplex phosphatase assays (Fig. 4). The longer cell surface E-selectin was cross-linked, the lower the resultant level of tyrosine phosphorylation of E-selectin in the immunocomplex. Comparable amounts of E-selectin and SHP2 were detectable in the immunocomplex (lower panels). These data suggest that proteintyrosine phosphatase SHP2 in the E-selectin-containing immunoprecipitates is catalytically active, and that E-selectin can be one of the substrates for SHP2 in vitro.
There are two tyrosine residues on the COOH terminus of SHP2 (8). We therefore examined if SHP2 became tyrosine phosphorylated in an E-selectin-dependent manner. After the adhesion of fixed HL-60 cells, or mAb cross-linking of cell surface E-selectin, SHP2 was immunoprecipitated from the total cell lysates and immunoblotted with the phosphotyrosine antibody, PY99. As seen in Fig. 5, SHP2 showed significantly increased tyrosine phosphorylation after only 2 min of fixed HL-60 cell adhesion or cell surface E-selectin cross-linking. This suggested that SHP2 not only could act as a tyrosine phosphatase, but also could provide a possible docking site via its phosphorylated tyrosine residues for the binding of other adapter proteins.
Shc, Grb2/Sos, and Ras Form an E-selectin-dependent Signaling Complex-It has been shown that SHP2 itself can act as  5 and 6) or with WT-E-selectin and mutant SHP2 (catalytically inert) (mSHP2) (lanes 7 and 8). Lanes 3,5,and 7, no cell surface E-selectin cross-linking. Lanes 2, 5, and 8, cell surface E-selectin molecules were cross-linked with H4/18, followed by GAM-IgG, as described in the legend to Fig. 1. Phospho-ERK1/2 was measured as an index of activated ERK1/2. an adapter protein, via its phosphorylated tyrosine residues as docking sites for downstream adapter proteins, such as Shc and Grb2 (22). Therefore we utilized co-immunoprecipitation/immunoblotting to examine the association between SHP2, Shc, and other adapter proteins. As seen in Fig. 6A, SHP2 became associated with Shc after 2 min of cell surface E-selectin crosslinking, but was no longer associated after 15 or 30 min. Increased Shc/Grb2 association was evident after 2 min of the adhesion of fixed HL-60, but not of JY cells (a control for non-E-selectin-dependent adhesion), and mAb-induced crosslinking of cell surface E-selectin, but not HLA class I molecules (Fig. 6B). As reported for the other cell types, Grb2 and Sos formed a constitutive complex in IL-1␤-activated HUVEC (23-25) (Fig. 6C).
We previously had reported that Ras⅐Raf-1⅐Phospho-MEK form an E-selectin-dependent macrocomplex (7). It has been described that Sos, a guanine nucleotide-releasing factor, can recruit Ras to the cytoplasmic membrane to carry out Ras/Raf-1/MEK signaling (20,25). To determine whether the events we describe above were related to Ras, we immunoprecipitated Sos from total HUVEC cell lysates and immunoblotted for Ras (Fig.  6D). Increased Ras association with Sos was apparent following either adhesion of HL-60 or antibody-induced cross-linking of cell surface E-selectin. This suggests that Ras is associated with upstream adapter proteins in an E-selectin-dependent manner.

DISCUSSION
There is ample evidence that transmembrane adhesion molecules can both physically bridge and biochemically transduce signals among interacting cells. The selectins, a family of adhesion molecules involved in leukocyte adhesion to activated vascular endothelium, are no exception. Binding to L-selectin has been shown to generate both inside-out and outside-in signals, such as the activation of the Ras pathway and potentiation of the oxidative burst of human neutrophils (20, 26 -31). P-selectin also has been shown to transduce signals across the plasma membrane, resulting in the induction of tyrosine phosphorylation of focal adhesion kinase and transient increases in intracellular Ca 2ϩ (4,32). Our laboratory has demonstrated that adhesion of HL-60 cells, which expresses ligands for Eselectin, or monoclonal antibody (mAb)-mediated cell surface E-selectin cross-linking, can induce cytoskeletal linkage of Eselectin (5), dephosphorylation of serine residues in the cytoplasmic domain of E-selectin (6), and activation of ERK1/2, as well as the formation of a Ras⅐Raf⅐MEK macrocomplex (7).
In this study, we used E-selectin single-point mutants to demonstrate that one tyrosine residue, Tyr 603 , in the cytoplasmic domain of E-selectin plays an essential role in the association between E-selectin and SHP2, and in mediating the ERK1/2 activation induced by cross-linking cell surface E-selectin. A consensus sequence for SHP2 binding, Y(I/V)X(V/I/L/ P), has been defined in several proteins, such as platelet-derived growth factor ␤-receptor and platelet endothelial cell adhesion molecule)-1 (33,34). The cytoplasmic domain of Eselectin does contain similar motifs around the Tyr 603 (YQKP) and Tyr 608 (YIL) residues. Upon cross-linking of cell surface E-selectin, SHP2 becomes associated with E-selectin in a timedependent (Fig. 1B) and dose-dependent (data not shown) manner; however, only Tyr 603 appeared to play an essential role in this association. Furthermore, Tyr 603 plays an important role in E-selectin-dependent ERK1/2 activation (Fig. 2). Although the basal level of the association between SHP2 and E-selectin mutated at Tyr 608 appears to be slightly higher than the wildtype E-selectin (Fig. 1C), this increased association does not result in any E-selectin-dependent ERK1/2 activation (Fig. 2), suggesting a nonspecific effect. Cross-linking another endothelial cell surface heterodimer, the HLA class I molecule, which is FIG. 4. SHP2 associated with E-selectin is catalytically active.
Cell surface E-selectin molecules were selectively cross-linked with a mouse mAb H4/18 and by GAM-IgG (see Fig. 1, legend), for the times (1, 2, and 5 min) indicated. E-selectin phosphorylation on tyrosine residues was detected by immunoprecipitation/immunoblotting. Comparable amounts of E-selectin (middle panel) and SHP2 (lower panel) were detectable in these immunocomplexes.
FIG. 5. SHP2 is tyrosine phosphorylated following HL-60 adhesion or cell surface E-selectin cross-linking. All samples were treated with IL-1␤ (10 units/ml, 4 h, 37°C). Paraformaldehyde-fixed HL-60 cells (2 ϫ 10 6 /ml) were allowed to adhere to activated HUVEC monolayers at 4°C for 30 min, and then the monolayers were warmed to 37°C for the times (2, 15, and 30 min) indicated on the figure. Alternatively, cross-linking of cell surface E-selectin with H4/18, followed by goat-anti-mouse IgG, was performed as described in the legend to Fig. 1, for the times indicated.
FIG. 6. The formation of an E-selectin-dependent adapter protein complex. HUVEC were treated with IL-1␤ (10 units/ml, 4 h, 37°C) as control (C, control). A, cell surface E-selectin molecules were cross-linked by the mouse mAb, H18/7 (10 g/ml, 30 min, 4°C), followed by a goat anti-mouse IgG (1:200, 37°C) for the times (0, 2, 15, and 30 min) indicated. The amounts of Shc associated with SHP2 were detected by co-immunoprecipitation. B, paraformaldehyde-fixed HL-60 (2 ϫ 10 6 /ml) or JY cells (non-E-selectin-dependent adhesion) were allowed to adhere to activated HUVEC monolayers for 30 min at 4°C and then transferred to 37°C for 2 min. Alternatively, cell surface E-selectin or HLA class I molecules were selectively cross-linked with mAb H18/7 (HϩG) or W6/32 (WϩG), respectively, as described in the legend to Fig. 1, legend. C, cell surface E-selectin molecules were cross-linked with H18/7 and a goat anti-mouse IgG for the times (0, 2, 15, and 30 min) indicated and the association of Grb2 and Sos was detected by co-immunoprecipitation. D, paraformaldehyde-fixed HL-60 and JY cells (2 ϫ 10 6 /ml) were allowed to adhere to activated HUVEC for 30 min at 4°C and warmed to 37°C for 2 min. Cell surface E-selectin and HLA class I molecules were cross-linked using a mAb H18/7 (HϩG) or W6/32 (WϩG), respectively, for 2 min. The association of Ras and Sos was detected by co-immunoprecipitation. present at comparable density on the surface of IL-1␤-activated HUVEC, did not generate any changes in tyrosine phosphorylation on E-selectin (Fig. 1A). Our previously studies have indicated that antibody-mediated cross-linking can be utilized to mimic the specific clustering of cell surface E-selectin that presumably occurs during leukocyte adhesion, and results in activation of ERK1/2 (5,7). Taken together, these data strongly suggest the existence of an E-selectin-specific signaling pathway leading to the activation of ERK1/2 that includes Tyr 603 in the cytoplasmic domain of E-selectin. Previously, our laboratory had shown that serine residues on the cytoplasmic domain of E-selectin became de-phosphorylated upon HL-60 adhesion (6). There have been reports that serine/threonine dephosphorylation may be involved in the subsequent tyrosine phosphorylation (35). We are in the process of determining whether cross-linking-induced dephosphorylation on serine residues is required for the observed phosphorylation on tyrosine residues in the cytoplasmic domain of E-selectin.
It has been shown that overexpression of a catalytically inactive SHP2 can block ERK1/2 activation in response to insulin, platelet-derived growth factor (36,37), epidermal growth factor (38), and fibroblast growth factor (13), via a dominant-negative effect. In our system, COS-7 cells co-transfected with WT-E-selectin and catalytically inert SHP2 (dominant negative SHP2), showed significantly blunted E-selectindependent ERK1/2 activation (Fig. 3). This suggests that SHP2 is a positive effector upstream of E-selectin-dependent ERK1/2 activation. We have further observed that E-selectin-associated SHP2 can de-phosphorylate tyrosine-phosphorylated E-selectin in vitro (Fig. 4), however; the exact substrate(s) of SHP2 in the context of E-selectin-dependent signaling in vivo are yet to be determined.
In other systems, tyrosine-phosphorylated SHP2 functions as an adapter protein with positive effects for downstream signaling (12,16,22,39). In our system, we demonstrate by co-immunoprecipitation that SHP2 itself becomes tyrosine phosphorylated and associates with Shc upon cross-linking of cell surface E-selectin (Figs. 5 and 6A). This further results in the formation of an E-selectin-dependent signaling complex, which includes Shc, Grb2, and Sos (Fig. 6, B and C). These adapter proteins form a macromolecular complex to bridge SHP2 to downstream signals. Previously we had described, that upon cross-linking cell surface E-selectin, Ras was activated and became associated with Raf-1 and phospho-MEK (7). Here, we show that Sos becomes associated with Ras in an E-selectin-dependent manner (Fig. 6D). This suggests that the E-selectin-dependent signaling complex, which includes SHP2, Shc, Grb2, and Sos, is physically related to the Ras/Raf-1/MEK/ ERK1/2 pathway.
In summary, we have shown that E-selectin can transduce signals across the endothelial cell cytoplasmic membrane, via the phosphorylation of a specific tyrosine residue in its cytoplasmic domain. This results in the association of the cytoplasmic domain of E-selectin with SHP2, a protein-tyrosine phosphatase. SHP2, in turn, acts as both a positive effector to the downstream E-selectin-dependent ERK1/2 activation and an adapter protein to bridge between E-selectin and a downstream adapter complex, comprised of Shc, Grb2, and Sos. These events thus outline a molecular mechanism by which crosslinking of E-selectin during leukocyte adhesive interactions on the surface of an activated endothelial cell, can initiate a transmembrane signaling cascade within that endothelial cell. The downstream consequences of these signaling events, including the modulation of endothelial gene expression, may have multifunctional implications for the pathology of inflammation.