Functional binding of E-selectin to its ligands is enhanced by structural features beyond its lectin domain.

Selectins are key to mediating interactions involved in cellular adhesion and migration, underlying processes such as immune responses, metastasis, and transplantation. Selectins are composed of a lectin domain, an epidermal growth factor (EGF)-like domain, multiple short consensus repeats (SCRs), a transmembrane domain, and a cytoplasmic tail. It is well-established that the lectin and EGF domains are required to mediate interactions with ligands; however, the contributions of the other domains in mediating these interactions remain obscure. Using various E-selectin constructs produced in a newly developed silkworm-based expression system and several assays performed under both static and physiological flow conditions, including flow cytometry, glycan array analysis, surface plasmon resonance, and cell-rolling assays, we show here that a reduction in the number of SCR domains is correlated with a decline in functional E-selectin binding to hematopoietic cell E- and/or L-selectin ligand (HCELL) and P-selectin glycoprotein ligand-1 (PSGL-1). Moreover, the binding was significantly improved through E-selectin dimerization and by a substitution (A28H) that mimics an extended conformation of the lectin and EGF domains. Analyses of the association and dissociation rates indicated that the SCR domains, conformational extension, and dimerization collectively contribute to the association rate of E-selectin-ligand binding, whereas just the lectin and EGF domains contribute to the dissociation rate. These findings provide the first evidence of the critical role of the association rate in functional E-selectin-ligand interactions, and they highlight that the SCR domains have an important role that goes beyond the structural extension of the lectin and EGF domains.

Total synthesis of a sialyl Lewis(x) derivative for the diagnosis of cancer.

The total synthesis of aminoethyl glycoside of sialyl Lewis(x) (sLe(x)) is described. A galactose donor was condensed with a diol of glucosamine to afford regioselectively a ß1,4 linked disaccharide, which was further stereoselectively fucosylated to provide a protected Lewis(x) trisaccharide. After chemical modification, the trisaccharide was sialylated to give regio- and stereoselectively an azidoethyl glycoside of sLe(x). Finally, deprotection and azide reduction afforded the target compound. This compound will be coupled with protein and then be used to conduct further preclinical studies for the diagnosis of cancer.

Expression, purification, and refolding of active recombinant human E-selectin lectin and EGF domains in Escherichia coli.

Attempts to obtain active E-selectin from Escherichia coli (E. coli) have not yet been successful. In this study, we succeeded in expressing the recombinant lectin and epidermal growth factor domain fragments of human E-selectin (rh-ESLE) in E. coli on a large-scale. The rh-ESLE protein was expressed as an inactive form in the inclusion bodies. The inactive form of rh-ESLE was denatured and solubilized by 6 M guanidine hydrochloride and then purified by Ni(2+) affinity chromatography under denaturing conditions. Denatured rh-ESLE was then refolded by a rapid-dilution method using a large amount of refolding buffer, which contained arginine and cysteine/cystine. The refolded rh-ESLE showed binding affinity for sLe(X) (K(d) = 321 nM, B(max) = 1.9 pmol/µg protein). This result suggests that the refolded rh-ESLE recovered its native and functional structure.

Lipid raft localization of cell surface E-selectin is required for ligation-induced activation of phospholipase C gamma.

E-selectin, an endothelial cell surface adhesion receptor for leukocytes, also acts as a signaling receptor. Upon multivalent ligation, E-selectin transduces outside-in signals into the endothelium leading to changes in intracellular Ca(2+) concentration and activation of the mitogen-activated protein kinase signaling pathway. In addition, following leukocyte engagement, E-selectin associates via its cytoplasmic domain with components of the actin cytoskeleton and undergoes alterations in phosphorylation state that result in changes in gene expression. In this study, we show that E-selectin is localized in cholesterol-rich lipid rafts at the cell surface, and that upon ligation E-selectin clusters and redistributes in the plasma membrane colocalizing with a fraction of caveolin-1-containing rafts. In addition, we demonstrate that leukocyte adhesion via E-selectin results in association with and activation of phospholipase Cgamma (PLCgamma). Moreover, we show that disruption of lipid rafts with the cholesterol-depleting drug methyl-beta-cyclodextrin disrupts the raft localization of E-selectin as well as the ligation-induced association of E-selectin with PLCgamma, and subsequent tyrosine phosphorylation of PLCgamma. In contrast, cholesterol depletion has no effect on E-selectin-dependent mitogen-activated protein kinase activation. Thus, these findings demonstrate that the presence of E-selectin in lipid rafts is necessary for its association with, and activation of, PLCgamma, and suggest that this subcellular localization of E-selectin is related to its signaling function(s) during leukocyte-endothelial interactions.

In vitro Bartonella quintana infection modulates the programmed cell death and inflammatory reaction of endothelial cells.

Bartonella quintana is an epicellular bacterium, which in vivo as well as in vitro, invades endothelial cells and develops within them inducing proliferative effects that play a pivotal role in neovascular manifestation of this disease. We investigated the effect of live Bartonella quintana and its LPS on apoptosis and inflammatory response in HUVEC-C, an endothelial cell line. The kinetics of the programmed cell death of Bartonella quintana-infected HUVEC-C showed a peculiar course. Even if early during infection apoptosis reached a peak after 6 h, later on apoptosis was inhibited. Such apoptosis inhibition was not observed during Bartonella quintana lipopolysaccharide treatment because LPS-stimulated HUVEC-C did progress to cell death. Evaluation of multiple cell signal transduction pathways revealed an overexpression of Apaf 1 and caspase 8 in HUVEC-C after 2 h of infection, and of bcl-2 starting from 10 h post Bartonella quintana infection. Moreover, Bartonella quintana and its LPS showed a different effect on the activation of genes involved in inflammatory response as revealed by molecular analysis of host cells. Bartonella quintana appears to be able to inhibit programmed cell death, inducing intracellular signals leading to survival and proliferation through the bcl-2 gene, despite the early increase of inflammatory status induced in endothelial cells. This mechanism, together with a poor endotoxin ability to stimulate strong inflammatory response, could contribute to the capability of the bacteria to persist intracellularly, causing chronic disease and producing neovascular manifestations.

Heparan sulfate and chondroitin sulfate proteoglycans inhibit E-selectin binding to endothelial cells.

E-selectin is a cell adhesion molecule involved in the initial rolling and adhesion of leukocytes to the endothelium during inflammation. In addition, in vitro studies have suggested that an interaction between E-selectin and binding sites such as sialyl Lewis X-containing oligosaccharides on endothelial cells may be important for angiogenesis. In order to investigate the binding of E-selectin to endothelial cells, we developed an ELISA assay using chimeric E-selectin-Ig molecules and endothelial cells fixed on poly-L-lysine coated plates. Our results indicate that E-selectin-Ig binds to both bovine capillary endothelial cells and human dermal microvascular endothelial cells in a calcium-dependent and saturable manner. The binding is inhibited markedly by heparin and by syndecan-1 ectodomain, and moderately by chondroitin sulfate, but not by sialyl Lewis X-containing oligosaccharides. These results suggest that heparan sulfate and chondroitin sulfate proteoglycans on endothelial cells are potential ligands for E-selectin.

Antibodies eluted from acutely rejected renal allografts bind to and activate human endothelial cells.

This study was designed to investigate how antiendothelial antibodies (EAbs) are involved in acute irreversible renal graft rejection. Eluates from 25 renal allografts, lost by irreversible rejection (n = 22) and by renal vein thrombosis (controls n = 3), were tested against a panel of cultured human umbilical vein endothelial cells (HUVEC). All patients were under immunosuppression at the time of nephrectomy. EAbs binding and membrane expression of adhesion molecules ELAM-1 and VCAM-1 were analyzed by flow cytometry (FACS) and by semiquantitative RT-PCR for mRNAs coding for those molecules. The absence of anti-HLA antibodies against the donor was ascertained at transplant, and before and after nephrectomy by the negativity of specific crossmatches performed using the most sensitive techniques. EAbs eluted from eight rejected kidneys bound to HUVEC. They did not induce any cytotoxicity, but their incubation with HUVEC (4 h at 37 degrees C; 2.5 mg/ml) led to upregulation of mRNAs coding for VCAM-1 (35- to 60-fold increases) and ICAM-1 (8- to 12-fold increases) as compared with control EAbs. Membrane expression of adhesion molecules was also strikingly increased, with 80% of the cells expressing VCAM-1 and 65% expressing ELAM-1 upon incubation. EAbs were detected in eight out of nine (88.8%) eluates from kidneys lost from acute vascular rejection, but in none of the 13 (0.0%) kidneys lost from other types of rejection (p < 0.0001). We conclude that EAbs, capable of activating human endothelial cells, can be recovered from acutely rejected kidneys and may play a direct role in the pathogenesis of acute rejection.

Comparison of E-selectin-binding glycoprotein ligands on human lymphocytes, neutrophils, and bovine gamma delta T cells.

We compared E-selectin-binding cell surface ligands on bovine gamma delta T cells and human leukocytes using an E-selectin/Ig chimera. The chimera worked well in flow cytometric studies and showed that bovine gamma delta T cells were the only lymphocyte population in newborn animals that bound E-selectin chimera. Furthermore, the chimera blocked gamma delta T cell rolling on E-selectin. Chimera reacted with four potential glycoprotein ligands of 180, 200, 250, and 300 kDa in Western blot analysis of gamma delta T cell detergent lysates, and it specifically precipitated at least two of these E-selectin ligands (200 and 250 kDa) from lysates of cell surface biotinylated gamma delta T cells. Preclearing bovine gamma delta lysates of GD3.5 Ag and workshop cluster 1 did not abrogate E-selectin ligand precipitation, suggesting that these surface markers do not represent E-selectin ligands. Human neutrophils possessed three E-selectin-binding ligands of approximately 80 to 90, 130, and 230 kDa, while human lymphocytes variably possessed three ligands of 120, approximately 220 to 240, and 260 kDa. Cross-precipitation experiments confirmed the results of others that neutrophil L-selectin serves as the 80 to 90-kDa E-selectin ligand. The human lymphocyte approximately 220 to 240-kDa and 260-kDa ligands may be analogous to the bovine gamma delta T cell molecules, whereas the 120-kDa is unique to human cells. The identities of the human and bovine lymphocyte E-selectin ligands are unknown. Finally, E-selectin ligand-1 apparently may have a minimal role, if any, in lymphocyte/E-selectin interactions, since a polyclonal anti-E-selectin ligand-1 serum stained a minimal number of cutaneous lymphocyte-associated Ag-positive human lymphocytes and bovine gamma delta T cells.

Liver endothelial E-selectin mediates carcinoma cell adhesion and promotes liver metastasis.

E-selectin is a cytokine-inducible endothelial cell adhesion receptor which is involved in the process of leukocyte rolling, the first in a cascade of interactions leading to leukocyte transmigration. Several studies have implicated this receptor in carcinoma cell adhesion to the endothelium, an interaction thought to be required for tumor extravasation during metastasis. To study the role of this receptor in the process of metastasis, we utilized a murine carcinoma line H-59 which is highly metastatic to the liver in vivo. When adhesion of H-59 cells to primary cultures of murine hepatic endothelial cells was measured, it was found that the tumor cells had a low basal level of adhesion to the sinusoidal endothelial cells, which could be significantly and specifically augmented by pre-activation of the endothelial cells with rTNF alpha. This incremental increase in adhesion to the activated endothelium could be completely and specifically abolished by a neutralizing monoclonal antibody to murine E-selectin (MAb 9A9). Similar results were obtained with 2 highly metastatic human colorectal carcinoma lines, HM 7 and CX-1, but not with a second murine subline, M-27, which is poorly metastatic to the liver. To assess the role of E-selectin in metastasis to the liver in vivo, the effect of MAb 9A9 on experimental liver metastasis was evaluated using the syngeneic H-59 model. We show here that this antibody caused a marked, specific and Fc-independent inhibition of experimental liver metastasis, reducing the median number of metastases by 97% relative to the control groups. Our results provide evidence that endothelial E-selectin is a mediator of carcinoma metastasis to the liver.

Biological and biophysical characterization of recombinant soluble human E-selectin purified at large scale by reversed-phase high-performance liquid chromatography.

A first step in the development of a high-throughput screening assay for antagonists of human E-selectin is the purification and characterization of the selectin. In the present paper we describe a single-step, rapid, reversed-phase HPLC purification protocol for the recombinant, soluble form of human E-selectin (rshE-selectin) produced in Chinese hamster ovary cells. The procedure resulted in high protein yields with recoveries of greater than 98%. Characterization of the reversed-phase purified rshE-selectin showed this product to be analogous to rshE-selectin purified using conventional chromatographic techniques with respect to biological activity and molecular shape. However, the carbohydrate composition of reversed-phase purified rshE-selectin, which had been variable with conventionally purified material, was found to be constant across several isolations. The protocol described herein eliminated the high mannose component associated with previously purified rshE-selectin and provided a uniform carbohydrate composition for additional experimental studies, such as NMR. This fact, coupled with the high yield and simplicity of the present purification scheme are distinct advantages over those previously published. It is expected that other mammalian selectins, such as P-selectin and L-selectin, would also be amenable to reversed-phase HPLC purification.