Similar to mechanisms of recruitment of activated leukocytes to inflamed tissues, selectins mediate adhesion and extravasation of circulating cancer cells. Our objective was to determine whether sialyl Lewis X modified core 2 O-glycans (C2-O-sLeX) present on colon and hepatic carcinoma cells promote their adhesion and invasion. We examined membrane expression of C2-O-sLeX, selectin binding, invasion of human colon and hepatic carcinoma cell lines, and mRNA levels of alpha-2,3 fucosyltransferase (FucT-III) and core 2 beta-1,6 N-acetylglucosaminyltransferase (C2GnT1) genes, necessary for C2-O-sLeX synthesis, by quantitative reverse-transcriptase (RT) PCR. Synthesis of core 2 branched O-glycans decorated by sLeX is dependent on C2GnT1 function and thus we determined enzyme activity of C2GnT1. The cell lines that expressed C2GnT1 and FucT-III mRNA by quantitative RT-PCR were highly positive for C2-O-sLeX by flow cytometry, and colon carcinoma cells possessed highly active C2GnT1 enzyme. Cells bound avidly to E-selection but not to P- and L-selectin. Gene knock-down of C2GnT1 in colon and hepatic carcinoma cells using short hairpin RNAs (shRNA) resulted in a 40–90% decrease in C2-O-sLeX and a 30–50% decrease in E-selectin binding compared to control cells. Invasion of hepatic and colon carcinoma cells containing C2GnT1 shRNA was significantly reduced compared to control cells in Matrigel assays and C2GnT1 activity was down-regulated in the latter cells. The sLeX epitope was predominantly distributed on core 2 O-glycans on colon and hepatic carcinoma cells. Our findings indicate that C2GnT1 gene expression and the resulting C2-O-sLeX carbohydrates produced mediate the adhesive and invasive behaviors of human carcinomas which may influence their metastatic potential.
Citation: St. Hill CA, Baharo-Hassan D, Farooqui M (2011) C2-O-sLeX Glycoproteins Are E-Selectin Ligands that Regulate Invasion of Human Colon and Hepatic Carcinoma Cells. PLoS ONE 6(1): e16281. https://doi.org/10.1371/journal.pone.0016281
Editor: Donald Gullberg, University of Bergen, Norway
Received: October 4, 2010; Accepted: December 9, 2010; Published: January 19, 2011
Copyright: © 2011 St. Hill et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported in part by funds from the National Institutes of Health, National Cancer Institute grant 5KO8CA111829-04. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Recognition and binding of selectins to sialyl Lewis X (sLeX) and related oligosaccharides are crucial interactions that regulate leukocyte adhesion to blood vessels and extravasation into tissues in an inflammatory response . Similar mechanisms are used by circulating tumor cells during metastasis to enter target organs , . Selectin adhesion molecules: L-selectin constitutively expressed on most leukocytes , P-selectin expressed in activated platelets and endothelial cells , , and E-selectin induced on cytokine stimulated endothelial cells , mediate these processes. High affinity binding of selectins to sLeX on human leukocytes is greatly enhanced when sLeX is terminally displayed on core 2 based O-linked glycans (C2-O-sLeX) , , . On leukocytes, C2-O-sLeX decorates the mucin P-selectin glycoprotein ligand-1 (PSGL1, CD162) and is particularly important for high-strength binding interactions with P-selectin , , , .
Synthesis of carbohydrates with sialyl Lewis structures increases upon neoplastic transformation and are useful markers for cancers . Several carcinomas including colon, gastric, lung, pancreatic, prostate, and urinary bladder highly express the selectin ligand sLeX and expression is significantly correlated with advanced disease and a poor prognosis , , , , , . Up-regulation of vascular E-selectin in cancers with high expression of sialyl Lewis structures is thought to be a risk factor for hematogenous metastasis . In colon and hepatocellular carcinomas, a high content of sLeX antigens is associated with increased metastatic potential ,  but the molecular mechanisms involving sLeX that regulate metastasis are not well understood. It is important to understand the role of glycans in cancer progression because the altered N- or O-glycosylation status of tumors may predict their metastatic potential and promote invasion and metastasis . We have previously demonstrated that a colon carcinoma cell line expressing C2-O-sLeX carbohydrates binds to E-selectin , and that C2-O-sLeX is a tumor-associated antigen in colon cancer tissues . In this report, we investigated the impact of disrupting C2-O-sLeX synthesis on the distribution of sLeX on core 2 O-glycans versus N-glycans on colon and hepatic carcinoma cell lines, their E-selectin binding capacity, and the influence on their invasive properties as initial steps to elucidate the function of these carbohydrates in metastasis.
C2-O-sLeX biosynthesis is complex and occurs by a series of enzymatic steps involving activities of glycosyltransferases (Fig. 1). Briefly, N-acetylgalactosamine is added to serine or threonine residues at the protein backbone, followed by galactose in a β1,3 linkage, creating the core 1 extension. The enzyme core 2 β1,6-N-acetylglucosaminyltransferase (C2GnT1) initiates the core 2 extension by adding N-acetylglucosamine to N-acetylgalactosamine in a β1,6 linkage and is the key branching enzyme in core 2 O-glycan biosynthesis. Core 2 branches are elongated by several glycosyltransferases and terminated by the addition of sialic acid in an α2,3 linkage to galactose. Finally, α(1,3/1,4) fucosyltransferase (FucT-III) catalyzes the addition of fucose to the chain in an α1,3 linkage, resulting in the formation of C2-O-sLeX . Functional activity of both C2GnT1 and FucT-III enzymes is necessary for C2-O-sLeX synthesis , , .
A simplified version of C2-O-sLeX biosynthesis is shown. Core 1 O-glycans are synthesized by addition of β1,3 galactose to N-acetylgalactosamine. The core 2 β1,6-N-acetylglucosaminyltransferase (C2GnT1) enzyme catalyzes the addition of a β1,6 branch to a core 1 O-glycan to form a core 2 β1,6 branched O-glycan. The core 2 O-glycan structure is further modified by a series of enzymatic reactions (omitted for clarity) including modification by α2,3 sialyltransferase and α1,3 fucosyltransferase (FucT-III) to form a sLeX terminus (dotted box). These modifications result in the synthesis of the sLeX-modified core 2 β1,6 O-glycan (C2-O-sLeX) structure. GalNAc, N-acetylgalactosamine; Gal, galactose; GlcNAc, N-acetylglucosamine; NeuAc, sialic acid; Fuc, fucose.
We targeted the C2GnT1 gene to disrupt C2-O-sLeX biosynthesis in carcinoma cells. We chose this gene because high C2GnT1 expression in carcinomas has previously been correlated with vessel invasion, depth of tumor invasion, and metastasis , , . Using the CHO-131 monoclonal antibody (mAb) which specifically detects C2-O-sLeX , we directly examined the ability of tumor cells transduced with C2GnT1 shRNA and expressing low levels C2-O-sLeX to bind to selectins. We have previously shown that C2-O-sLeX is a tumor-associated ligand that is abundant on the invasive front of human colon carcinoma tissues, and mediates E-selectin binding when expressed on a colon carcinoma cell line , . In this report, we directly demonstrated for the first time that expression and activity of the C2GnT1 gene responsible for C2-O-sLeX synthesis in colon and hepatic carcinoma cell lines regulated invasion of tumor cells, a key property that facilitates metastasis. We found that the sLeX epitope was predominantly distributed on O-glycans compared to N-glycans and was an important E-selectin ligand in these cell lines.
Human carcinoma cell lines highly express genes involved in C2-O-sLeX synthesis and bind to E-selectin
We examined mRNA expression of C2GnT1 and FucT-III genes in colon (LS174T) and hepatic (HepG2) cell lines by RT-PCR. Both cell lines expressed C2GnT1 and FucT-III mRNA necessary for C2-O-sLeX synthesis (Fig. 2A). To investigate whether mRNA expression of the C2GnT1 and FucT-III genes resulted in cell surface expression of C2-O-sLeX, we tested the carcinoma cell lines for reactivity with CHO-131 mAb by flow cytometry. Approximately 75% of un-manipulated LS174T colorectal adenocarcinoma cells (MFI, 5866) and 55% of HepG2 hepatic carcinoma cells (MFI, 1659) reacted positively with CHO-131 mAb but did not react with an isotype control mAb (Fig. 2B).
(A) Endogenous C2GnT1 and FucT-III mRNA transcripts in LS174T colorectal adenocarcinoma (lane 1) and HepG2 hepatic carcinoma (lane 2) cell lines were detected by RT-PCR. (B) Flow cytometric analysis of positive staining of LS174T and HepG2 cells labeled with CHO-131 mAb (anti-C2-O-sLeX). A representative example of 3 experiments is shown for (A-B). (C) LS174T (*p = 0.008) and HepG2 cells (#p = 0.001) substantially bind to E-selectin in the presence of calcium ions (Ca2+). Binding is impeded in the presence of EDTA, a calcium ion chelator. The average of two experiments is shown.
The binding of the sLeX epitope on colon carcinoma cells to E-selectin has been associated with a greatly increased metastatic potential and a poor prognosis . We therefore examined the participation of C2-O- sLeX in E-selectin binding of LS174T and HepG2 cells by detecting the ability of these carcinoma cell lines expressing C2-O-sLeX to bind to a mouse E-selectin Fc chimera by flow cytometry. Selectin binding interactions are calcium dependent and are dissociated in the presence of EDTA, a chelator of calcium ions. LS174T and HepG2 cells with high endogenous C2-O-sLeX expression bound avidly to E-selection (98% and 95% cellular binding for LS174T and HepG2 cells respectively compared to binding in the presence of 10 mM EDTA) (Fig. 2C).
Knock-down of the C2GnT1 gene results in decreased E-selectin binding
Currently, a specific inhibitor of C2GnT1 expression and activity is not available. To directly evaluate the influence of C2GnT1 activity on the generation of C2-O-sLeX epitopes on tumor cells and the binding to E-selectin, we instead used short hairpin RNA (shRNA) gene silencing techniques to target the C2GnT1 gene in HepG2 and LS174T cells. The four lentiviral pGIPZ-C2GnT1-shRNA vectors, the PLKO.1-C2GnT-shRNA vector, the scrambled pGIPZ-shRNA vector, and the empty PLKO.1 control vector were tested for silencing of the C2GnT1 gene in LS174T colon carcinoma and HepG2 hepatic carcinoma cells. All shRNA sequences achieved consistent gene knockdown in the cell lines tested indicating that our observed effects were specifically due to loss of the C2GnT1 gene. The clones that were most efficient at gene knockdown were selected for subsequent experiments. Thus, LS174T cells were stably transduced with one of the lentiviral pGIPZ-C2GnT1-shRNA vectors or with a scrambled pGIPZ-shRNA vector. HepG2 cells were transiently transfected with the lentiviral PLKO.1-C2GnT-shRNA vector or with an empty PLKO.1 control vector. The C2GnT1 shRNA sequences were homologous to only the human C2GnT1 gene when compared to the human genome using NCBI Nucleotide BLAST. The scrambled shRNA vectors were not complementary to any gene in the human genome.
As shown in Fig. 3A, mRNA levels of the C2GnT1 gene were considerably lower in LS174T cells transduced with C2GnT1 shRNA compared to cells transduced with the scrambled shRNA vector by densitometry. Similarly, mRNA levels of C2GnT1 were lower in HepG2 cells transfected with C2GnT1 shRNA compared to cells transfected with the control vector (Fig. 3B). In contrast, in both cell lines, FucT-III mRNA levels were not affected by silencing of the C2GnT1 gene, indicating that the shRNA clones did not have off-target effects. The C2GnT1 protein was detected in LS174T cells transduced with the scrambled shRNA vector but not in cells transduced with C2GnT1 shRNA after immunoprecipitation and Western blotting with an anti-C2GnT1 antibody (Fig. 3C and D). The percentage of cells that was reactive with CHO-131 mAb, indicating expression of cell surface C2-O-sLeX before and after C2GnT1 gene knock-down, was assessed by flow cytometry. In multiple experiments using LS174T and HepG2 cells, we consistently achieved approximately 30–40% reduction of C2-O-sLeX expression after C2GnT1 gene knock-down and a representative experiment is shown in Fig. 3D. We verified silencing of the C2GnT1 gene by examining C2GnT1 enzyme activity in LS174T cells and observed a significant ∼40% decrease in enzyme activity of LS174T cells transduced with C2GnT1 shRNA compared to those cells transduced with scrambled shRNA (p = 0.02, Fig. 3E). For all groups of cells, viability was assessed by trypan blue exclusion and was greater than 95%.
(A) C2GnT1 suppression by shRNA resulted in decreased mRNA levels of C2GnT1 but not FucT-III in LS174T cells, (scrambled shRNA vector, lane 1, and C2GnT1 shRNA vector, lane 2). (B) Similar effects on C2GnT1 mRNA but not FucT-III mRNA were observed in HepG2 cells after transient transfection with C2GnT1 shRNA (empty control vector, lane 1, and targeting shRNA vector, lane 2). Densitometry values are normalized to β-actin. (C) The C2GnT1 glycoprotein (50 kDa) was present in scrambled shRNA transduced LS174T cells (lane 1) but absent in C2GnT1 shRNA transduced cells (lane 2) as detected by immunoprecipitation and Western blotting using a polyclonal anti-C2GnT1 antibody. (D) The procedures were repeated using the C2GnT1 antibody for immunoprecipitation and a negative control IgG antibody for Western blotting. Note the absence of a detected C2GnT1 protein 50 kDa band. (D) We observed a 30–40% decrease in reactivity with CHO-131 mAb by flow cytometry for LS174T and HepG2 cells that contained C2GnT1 shRNA. Representative data from multiple experiments are shown. (E) C2GnT1 enzyme activity, measured as nmols of N-acetylglucosamine (GlcNAc) transferred per mg protein per hour in a glycosyltransferase assay, was also significantly decreased in C2GnT1 shRNA transduced LS174T cells, *p = 0.02. Mean values ± standard deviations of two independent experiments are presented.
For LS174T and HepG2 cells in which the C2GnT1 gene was silenced and for control cells, reactivity with a mouse E-selectin/Fc chimera was assessed by flow cytometry. We did not observe a significant decrease in E-selectin binding of LS174T cells containing C2GnT1 shRNA compared to un-manipulated cells and cells transduced with scrambled shRNA (Fig. 4A). However, we observed a significant 40% decrease in E-selectin binding for C2GnT1 shRNA-transfected HepG2 cells compared to those cells transfected with the control vector, (p = 0.02, Fig. 4B).
(A) Significant differences in E-selectin binding were not observed among LS174T un-manipulated cells, cells transduced with scrambled shRNA, or with C2GnT1 shRNA by flow cytometry. (B) E-selectin binding was significantly decreased for HepG2 cells transfected with C2GnT1 shRNA compared to cells transfected with the control vector, *p = 0.02. (C) Shear flow assays were performed at an E-selectin/Fc chimera concentration of 1 µg/ml and shear stresses ranging from 0.5 – 1.5 dynes/cm2 in the presence of 10 µg/ml of an IgG isotype control mAb. At a shear stress of 0.5 dynes/cm2, significantly fewer C2GnT1 shRNA transduced cells accumulated on E-selectin than un-manipulated LS174T cells, *p = 0.04. (D) In the same run of experiments, at a shear stress of 0.5 dynes/cm2, significantly fewer un-manipulated LS174T cells treated with a 10 µg/ml of a blocking anti-E-selectin mAb accumulated on E-selectin than those cells treated with an isotype control mAb, #p = 0.004. Significantly fewer LS174T cells transduced with C2GnT1 shRNA and treated with a blocking anti-E-selectin mAb accumulated on E-selectin than those cells treated with an isotype control mAb, &p = 0.01. The same assay as shown in (C) at 0.5 dynes/cm2 is included for comparison. For both groups of cells treated with an isotype control mAb, significantly fewer un-manipulated LS174T cells accumulated on E-selectin than LS174T cells transduced with C2GnT1 shRNA, *p = 0.04. Each continuous shear flow assay was performed in duplicate for each shear stress and the bars represent the mean ± standard deviations. Representative data from three independent experiments are shown.
To further assess the influence of C2GnT1 expression on E-selectin binding of LS174T cells, we performed more stringent hydrodynamic shear flow assays that closely simulate forces occurring in the microvasculature. The adhesive behavior of cells was examined at a range of shear stresses between 0.5 and 1.5 dynes/cm2 and chimera concentrations between 0.1 and 5 µg/ml in the presence of 10 µg/ml of an anti-E-selectin function-blocking mAb or an isotype-matched control mAb. We observed optimal E-selectin binding of isotype control mAb treated un-manipulated LS174T cells and LS174T cells transduced with C2GnT1 shRNA at a concentration of 1 µg/ml of the E-selectin/Fc chimera and a shear stress of 0.5 dynes/cm2 (Fig. 4C). Under these conditions, accumulation of LS174T cells transduced with C2GnT1 shRNA was reduced by 36% compared to accumulation of un-manipulated LS174T cells, p = 0.04. For both groups of cells, fewer cells accumulated on the E-selectin chimera as the shear stress was increased to 1.5 dynes/cm2. In order to assess the specificity of the interaction of cells with E-selectin, both groups of cells were exposed to an anti-E-selectin function-blocking mAb (10 µg/ml) or an isotype-matched control mAb at the same concentration. Accumulated cells at a shear stress of 0.5 dynes/cm2 are shown in Fig. 4D and binding of both groups of cells exposed to the isotype control mAb is shown for comparison. The presence of an E-selectin blocking mAb greatly diminished the numbers of un-manipulated LS1704T (p = 0.004) and C2GnT1 transduced cells bound to E-selectin (p = 0.0001). The addition of 20 mM EDTA to perfusates prevented the binding of cells to E-selectin (data not shown).
LS174T and HepG2 cells predominantly express sLeX on core 2 O-glycans
In order to determine the nature of the carbohydrate structure to which the sLeX epitope is attached in un-manipulated LS174T and HepG2 cells, we treated cells with agents to inhibit O-glycosylation (BGN) and N-glycosylation (SWN) because these glycans are the main carriers of sLeX. Treated and untreated cells were labeled with CSLEX1 mAb that detects sLeX on any structure to assess the percentage of sLeX positive cells on O-glycans versus N-glycans by flow cytometry. For both LS174T and HepG2 cells in the presence of BGN but not SWN, a significant decrease in the percentage of cells reactive with CSLEX1 mAb was observed compared to untreated cells (Fig. 5A and B, p = 0.0004 for LS174T cells and p<0.0001 for HepG2 cells) indicating that sLeX was mainly expressed on O-glycans in these cell types. Treatment of LS174T and HepG2 cells with both BGN and SWN inhibitors did not cause a further reduction in cells reactive with CSLEX1 mAb but the decrease was significant when compared to untreated cells, (p = 0.002 and p<0.0001 respectively), and was likely due to the effects of BGN.
(A) The percentage of sLeX positive LS174T cells was significantly decreased after treatment with BGN (an O-glycosylation inhibitor) *p = 0.0004, or with BGN and SWN (an N-glycosylation inhibitor, **p = 0.002) compared to untreated cells (None). (B) For HepG2 cells, a reduction in the percentage of sLeX positive cells was observed after treatment with BGN *p < 0.0001, or with BGN and SWN **p < 0.0001, compared to untreated cells. (C) The cells used in (A) and (B) before and after treatment with BGN were also labeled with CHO-131 mAb to specifically detect C2-O-sLeX. Note the significant reduction in the percentage of C2-O-sLeX positive LS174T and HepG2 cells after treatment with BGN compared to untreated cells, * and **p < 0.0001. (D) A significant decrease in E-selectin binding was observed for LS174T cells after treatment with BGN *p = 0.003, and after BGN and SWN **p = 0.0005, compared to untreated cells. (E) Similar results were observed for HepG2 cells after BGN, #p = 0.0005, and after BGN and SWN, ##p<0.0001, compared to untreated cells.
LS174T and HepG2 cells, either untreated or treated with BGN, were subsequently labeled with CHO-131 mAb that specifically detects C2-O-sLeX positive cells to determine the percentage of the subset of core 2 O-glycans that carried sLeX. We observed that reactivity with CHO-131 mAb was significantly reduced by 40% in BGN-treated LS174T and by 60% in HepG2 cells compared to untreated cells (Fig. 5C, p<0.0001). Binding to E-selectin was also evaluated in untreated cells and in cells treated with the inhibitors. As shown in Fig. 5D and E, BGN treatment significantly decreased E-selectin binding of LS174T cells by 38% (p = 0.003) and of HepG2 cells by 67% (p = 0.0005) compared to untreated cells. A combination of BGN and SWN treatment did not result in further reductions in E-selectin binding of LS174T and HepG2 cells but significant differences were observed when compared to untreated cells, p = 0.0005 and p<0.0001 respectively, indicating the effects of BGN treatment.
C2-O-sLeX mediates the invasion of carcinoma cells
We examined the influence of silencing of the C2GnT1 gene on the invasive properties of LS174T and HepG2 carcinoma cells. Using a Matrigel invasion assay that mimics active invasion of tumor cells across a basement membrane, we found that after 48 hours, the number of C2GnT1 shRNA-transduced LS174T cells that invaded across the Matrigel membrane was significantly decreased compared to LS174T cells transduced with scrambled shRNA (mean number of cells/field = 27 versus 162, p<0.0001) (Fig. 6A, C, and D). Similarly, we observed a significant decrease in invasion of HepG2 cells transfected with C2GnT1 shRNA compared to cells transfected with a control vector, (mean number of cells/field = 14 versus 99, p<0.0005) (Fig. 6B, E, and F). Our results indicated that C2GnT1 gene expression mediated the invasive properties of LS174T and HepG2 cells by an unknown mechanism.
(A) LS174T cells transduced with C2GnT1 shRNA were significantly less invasive than cells transduced with scrambled shRNA, *p<0.0001, in Matrigel transwell invasion assays. (B) Similarly, decreased invasion was observed for HepG2 cells that were transfected with C2GnT1 shRNA compared to transfection with a control vector, *p = 0.0005. (C) Representative photomicrographs from two separate experiments of invasion of LS174T cells transduced with the scrambled shRNA vector compared to (D) cells transduced with C2GnT1 shRNA and (E) HepG2 cells transfected with the control vector compared to (F) HepG2 cells transfected with C2GnT1 shRNA. The invading cells were stained and counted in five separate fields of view at 100X magnification. Arrows indicate invasive cells.
In the metastatic process, circulating cancer cells in the bloodstream arrest and adhere to endothelium, extravasate and invade into organs distant from the primary tumor, survive, proliferate, and promote angiogenesis. These multistep events require complex adhesive interactions between ligands on cancer cells and cellular adhesion molecules and other substrates present in the tumor microenvironment , . Many types of cancers over-express altered carbohydrates that profoundly impact tumor growth, adhesion, invasion, and metastasis. Alterations in glycosylation, including the production of sLeX epitopes, are generated by glycosyltransferases, yet surprisingly the role of glycosyltransferases in cancer malignancy is still not well defined.
In this report, we have directly shown that the glycosyltransferase C2GnT1 and its product cell membrane expressed C2-O-sLeX regulate E-selectin binding and invasion of colon and hepatic carcinoma cells. After gene silencing of C2GnT1 and treatment with O- and N-glycosylation inhibitors, we observed significant reductions in the high levels of C2-O-sLeX present on cells, binding to E-selectin, and cellular invasiveness into Matrigel. Binding of carcinoma cells to E-selectin was specific as evidenced by the lack of appreciable binding in the presence of a function blocking anti-E-selectin mAb. Our findings are novel because they indicate that targeting of the C2GnT1 gene and subsequent down-regulation of C2-O-sLeX expression, a selectin ligand on colon and hepatic carcinoma cells, altered key properties of metastasis that are likely occur in vivo: decreased attachment of circulating tumor cells to endothelium and invasion into tissues. The possibility of the cellular toxicity of the gene silencing clones introduced into cells was excluded because the viability of all groups of cells was similar.
The observations reported here agree with our previous findings that C2GnT1 gene expression results in synthesis of C2-O-sLeX that binds to E-selectin , . We have found that C2-O-sLeX is predominantly expressed at the advancing edge of invasive colorectal adenocarcinomas and high mRNA levels of C2GnT1 are present in these carcinomas compared to normal colonic tissues . Our results support previous studies that suggest that C2GnT1 plays an important role in cancer. Earlier reports found that C2GnT1 activity is increased in human leukemia cells and metastatic murine tumor cell lines , , . C2GnT1 gene expression is more highly associated with progression of colon and lung cancers than expression of sialyl Lewis structures ,  and also promotes prostate cancer progression . The cell surface sialyl Lewis ligands on cancer cells generated by C2GnT1 expression increase the binding efficiency of clonal cancer cell lines to E-selectin on endothelial cells which is directly proportional to their metastatic potential .
In this study, we did not further examine the possible mechanisms by which down-regulation of C2GnT1 and C2-O-sLeX expression led to reduced invasiveness of cancer cells but we are currently investigating potential processes. We speculate that changes in C2GnT1 and C2-O-sLeX expression may alter the expression and function of other adhesion molecules on cancer cells or regulate components of the extracellular matrix known to participate in invasive events. C2GnT1 may regulate the expression of tumor-associated epitopes such as cell surface associated mucins which may favor tumor progression . For instance, mucins such as MUC1 and MUC4 present on cancer cells participate in signal transduction pathways that contribute to the invasive and metastatic activities of adenocarcinomas . Over-expression of MUC1 is associated with invasive and metastatic colon cancer , , and MUC1 is also expressed on liver cancers . Oligosaccharide and protein structures on MUC1 can mediate adhesion events by binding to E-selectin . Conceivably, C2-O-sLeX may be one of the oligosaccharides attached to MUC1 or another mucin present on cancer cells that may modulate their invasive properties.
Matrix metalloproteinases (MMPs) are a large group of proteins that degrade the extracellular matrix and facilitate the invasion and metastasis of malignant cells into distant tissue sites. MMPs promote motility, migration, and invasion of both colon and hepatocellular carcinomas , . The membrane type MMPs, and in particular MT1-MMP bound to tumor cells promote cancer cell motility and invasion. MT1-MMP has also been reported to cleave MUC1 , and the cell adhesion molecule CD44: an E-selectin ligand that participates in adhesion and invasion . MT1-MMP also participates in processing the αv subunit of integrins, which are heterodimeric transmembrane glycoprotein receptors that function as cell anchoring and signaling molecules and are also known to be involved in invasion and tumor progression , . It is possible that C2GnT1 and/or C2-O-sLeX expression in tumors may modulate the activities of MMPs and consequently their targets such as MUC1, CD44, or integrins and thereby indirectly regulate invasion.
Other enzymes can compete with C2GnT1 for the core 1 substrate and prevent the formation of core 2 O-glycans, including C2-O-sLeX , and therefore it is important to determine the structure to which C2-O-sLeX is attached. Differences in this structure may considerably influence the adhesion, invasion, and metastatic potential of cancer cells. Several scaffold molecules containing complex core 2 O-glycans, other O-glycan structures, N-glycans, or glycolipids are also terminally decorated with sLeX , , . We have shown here that in a human colon carcinoma and hepatic carcinoma cell line, sLeX was primarily distributed on core 2 O-glycans forming C2-O-sLeX. Several reports indicate that high expression of C2GnT1 mRNA transcripts in carcinoma cells positively correlates with metastasis , , . This report suggests that C2GnT1 activity, resulting in the generation of C2-O-sLeX carbohydrates on tumor cells, in addition participates in cancer invasion. Thus, C2-O-sLeX may have distinct functions depending on its site of expression: on circulating tumor cells it could enhance attachment of tumor cells to endothelium and C2-O-sLeX could facilitate invasion on malignant cells that have extravasated and invaded into distant metastatic sites from the bloodstream.
Abnormal glycosylation is a prominent feature of malignant transformation and cancer progression . Although our studies indicate that abundant C2-O-sLeX structures present on colon and hepatic carcinoma cells mediate invasion, this moiety is not the only participant in E-selectin binding and invasion because knock-down of the C2GnT1 gene and treatment with O- and N-glycan inhibitors did not abrogate selectin binding and invasion, nevertheless, C2-O-sLeX may act in concert with other molecules to promote these processes.
C2GnT1, also known as C2GnT-L (leukocyte type), is a key branching enzyme that controls mammalian core 2 O-glycan synthesis , . Huang et al. recently reported that gene expression of another C2GnT isotype, C2GnT-M (mucin type), is downregulated in colorectal carcinomas and that C2GnT-M expression suppressed cell growth, adhesion, motility, invasion, and colony formation ability . C2GnT1 and C2GnT-M may have alternate functions in regulating the progression of carcinomas at various stages of malignancy of the tumor. C2GnT-M may act predominantly as an inhibitor of early stages of colorectal carcinoma growth. As the tumor progresses to more advanced stages, C2GnT1 activity may dramatically increase resulting in a corresponding increase in C2-O-sLeX expression, and increased invasion of cancer cells.
C2-O-sLeX is specifically recognized by the CHO-131 mAb that has been extensively characterized in cancer and immune cells , , , , . Other antibodies such as CSLEX1 mAb recognize the sLeX epitope alone that can be present on several types of glycan structures including core 2 O-glycans and thus, these antibodies have broader recognition of glycan structures . The reactivity of CHO-131 mAb is unique in that it requires the functional activity of the glycosyltransferases C2GnT1, α2,3-sialyltransferase, and α1,3-fucosyltransferase , , . CHO-131 mAb is not a function blocking antibody, however, and at high concentrations does not inhibit interactions of C2-O-sLeX terminally displayed on leukocyte PSGL1 with P-selectin .
In summary, our results indicate that the C2GnT1 enzyme, that regulates C2-O-sLeX synthesis of cancer cells, participates in invasion of colon and hepatic carcinomas. Our findings are novel because we have identified a previously un-described function of the carbohydrate enzyme C2GnT1 and its associated product C2-O-sLeX in mediating the invasive properties of colon and hepatic carcinoma cells. It is well established that inflammatory responses contribute to tumor development including the invasive behavior of tumor cells. Invasion requires extensive proteolysis of the extracellular matrix at the invasive front. Our study highlights the complex interactive roles that molecules in the tumor microenvironment play in neoplastic development. C2GnT1 and C2-O-sLeX appear to contribute to tumor invasion but are not likely to be the only candidates responsible for invasive events. On carcinoma cells, C2-O-sLeX may act in concert with other adhesion molecules or proteases to promote invasion and ultimately metastasis. Our results inspire additional questions about the specific mechanisms involving C2GnT1 and C2-O-sLeX that mediate invasion and we are currently designing studies to directly investigate these events. We expect that our findings will impact and support the design of therapies aimed at preventing the formation of specific carbohydrates or interrupting their tumor-promoting effects in colon and hepatic cancer.
Materials and Methods
The human cell lines LS174T (colorectal adenocarcinoma) and HepG2 (hepatocellular liver carcinoma) were obtained from American Type Culture Collection (Manassas, VA) and authentication testing, including confirmation of a negative status for mycoplasma, bacteria, fungi contamination, confirmation of species identity, and cytogenetic analysis, was performed prior to purchase. Cell lines were passaged for less than 6 months after thaw for use in these studies. LS174T and HepG2 cells were grown in Minimal Essential Medium (MEM) supplemented with 1.5 g/L sodium bicarbonate, 4.5 g/L glucose, 10 mM HEPES, 1.0 mM sodium pyruvate, 1mM non-essential amino acids, 10% fetal bovine serum, and a combination of 100 U/ml penicillin/100 µg/ml streptomycin.
Flow cytometric analysis
A FACSCanto instrument and BD FACSDiva software (Becton Dickinson Biosciences, San Jose, CA) were used for flow cytometric analyses as previously described , . Labeling of cells (5×106 cells per sample) was performed with CHO-131 mouse anti-human IgM monoclonal antibody (mAb) that detects C2-O-sLeX (10 µg/ml) and was generously provided by Dr. Bruce Walcheck, University of Minnesota, St. Paul, MN or with mouse anti-human CSLEX1 IgM mAb (BD Biosciences (San Jose, CA). For the flow cytometric adhesion assays, cells were tested for binding to 10 µg/ml of a mouse E-selectin/human Fc chimera (R & D Systems, Minneapolis, MN) in buffer containing 2mM CaCl2 because binding is calcium dependent, and with and without 2 mM EDTA (a calcium ion chelator), or with 10 mM EDTA alone as controls. Bound antibodies were detected with a phycoerythrin (PE) conjugated F(ab')2 goat anti-mouse IgM secondary antibody and bound chimera was detected with a PE-conjugated F(ab')2 goat anti-human-IgG secondary antibody (Jackson Immunoresearch Laboratories, West Grove, PA). To evaluate the levels of background staining, cells were labeled with only the secondary antibody.
Hydrodynamic shear flow assays
Cancer cell adhesion mechanisms were examined under hydrodynamic shear stress with a parallel plate flow chamber system (Glycotech, Gaitherburg, MD) as previously described . A functional site density was defined as the minimum concentration of chimeric molecules in the correct orientation for binding required for attachment of ∼90% cancer cells at a hydrodynamic shear stress of 0.5 dynes/cm2 and was determined to be 0.5 µg/ml for an E-selectin/Fc chimera. LS174T colon carcinoma cells were perfused over 1 µg/ml of the E-selectin/Fc chimera under shear stresses ranging from 0.5 to 1.5 dynes/cm2 that were within the range previously described to occur in post-capillary venules . Each assay was performed in triplicate in the presence of 10 µg/ml of an E-selectin function blocking mAb (R & D Systems, Minneapolis, MN) or the same concentration of an isotype-matched control mAb (Invitrogen, Carlsbad, CA, USA). Attached cells were enumerated at 100X magnification after 3 minutes of perfusion.
Quantitative Reverse-Transcriptase-PCR (RT-PCR) analysis
Total RNA was extracted from 2×106 cells of each cell line using Trizol reagent (Invitrogen) and DNA was removed using DNase I treatment. Reverse transcription was performed to synthesize cDNA using a First-Strand Synthesis System (Invitrogen) and 20 pmols each of primers specific for human C2GnT1, human FucT-III, and human β-actin. Quantitative real-time PCR was performed on an icycler thermocycler (Bio-Rad Laboratories, Hercules, CA) using Taq DNA polymerase (Continental Lab Products, San Diego, CA, USA) according to the manufacturer's instructions. Reactions were performed for 30 cycles with the following parameters: 94°C, 50 sec; 58°C, 50 sec; 72°C, 50 sec.
Gene silencing of C2GnT1
A pGIPZ-C2GnT1-short hairpin (sh)RNAmir vector engineered to silence human C2GnT1 gene expression, a control pGIPZ vector expressing scrambled shRNA, four lentiviral C2GnT1-shRNA vectors in a pLK01 vector system, and an empty pLK01 vector were purchased from Open Biosystems (Huntsville, AL). All vectors were tested for silencing of the C2GnT1 gene in human hepatic carcinoma cells (HepG2) and human colon carcinoma cells (LS174T) and the vector that was most efficient at gene silencing was used in subsequent assays with each cell line. Thus, HepG2 cells were transiently transfected with a C2GnT1 shRNA-pLKO1 vector or with the empty pLK01 vector. LS174T cells were stably transduced with the pGIPZ-C2GnT1-shRNAmir vector or with the scrambled pGIPZ-shRNA vector. C2GnT1 gene silencing was confirmed in transfected or transduced cells by RT-PCR, Western blot, and flow cytometry to detect cell surface expression of C2-O-sLeX. The specificity of gene silencing was assessed by the use of more than one shRNA vector, by comparing the C2GnT1 and scrambled shRNA sequences to the complete human genome using NCBI Nucleotide BLAST, and by investigation of the effects of the C2GnT1 shRNA vectors on the human FucT-III gene.
Immunoprecipitations and Western blot analyses
Whole cell lysates of LS174T cells were prepared with lysis buffer (SolObuffer-200, FabGennix International Inc, Frisco, TX) according the manufacturer's instructions. Protein concentration was quantified by the Bradford assay (Bio-Rad, Hercules, CA). Approximately 500 µg of protein was immunoprecipitated with 4 µl of a rabbit polyclonal anti-C2GnT1 primary antibody in immunoprecipitation buffer (50mM Tris-HCl, pH 8.0, 150mM NaCl, 1% Igepal, and protease inhibitor). Protein complexes were precipitated with 40 µl of Protein G beads. Products were subjected to 10% SDS-PAGE and transferred to an Immobilon-P transfer membrane (Millipore Corporation, Medford, MA), blocked with 5% blotting grade blocker non-fat dry milk, and probed with the C2GnT1 antibody or with a rabbit IgG negative control antibody and subsequently incubated with a HRP-conjugated goat anti-rabbit-IgG antibody. Protein bands were visualized with an ECL system (Pierce, Rockford, IL).
C2GnT1 glycosyltransferase assay
The activity of C2GnT1 glycosyltransferase in LS174T cells was evaluated as previously described by Prorok-Hamon et al . Cells were lysed in buffer containing 120mM NaCL, 40mM Tris, 1% Triton X-100 and 20% glycerol (Fisher, Pittsburgh, PA), and 1% Protease inhibitor (Roche Diagnostics Corporation, Indianapolis, IN). The substrate was 10 mM Gal β1,3 GalNAcα-O-paranitrophenyl (Toronto Research Chemicals, Downsview ON, Canada) and 10 mg/ml of proteins were used as the source of transferases.
Treatment of cells with O-glycan and N-glycan inhibitors
HepG2 and LS174T cells (3×105 cells/well) were treated with either 2mM benzyl 2-acetamido-2-deoxy-alpha-D-galactopyranoside (BGN, an O-glycosylation inhibitor), 10 ug/ml of swainosine (SWN, an N-glycosylation inhibitor), or a combination of both agents for 3–5 days. Using untreated and treated cells, the presence of membrane-associated sLeX was detected with CSLEX1 mAb and membrane-associated C2-O-sLeX was detected with CHO-131 mAb. For both cell types, binding to an E-selectin/Fc chimera was assessed by flow cytometry.
Matrigel invasion assays
Twenty-four well BioCoat™ Matrigel™ Invasion Chambers (Becton-Dickinson, Bedford, MA, USA) were used for the tumor cell invasion assays according to the manufacturer's instructions. Cell suspensions (2.5×104 cells/chamber) in 0.5 ml serum-free medium were added to the upper chamber of each well and 0.5 ml complete medium containing 10% fetal bovine serum was added to each bottom chamber. Chambers were incubated for 24 hours in a humidified tissue culture incubator at 37°C and 5% CO2. The invading cells were stained and counted in five separate fields of view at 100X magnification. Digital images were captured using a Zeiss Axiovert 200 inverted microscope and an AxioCam MRc camera using AxioVision 4.1 software (Carl Zeiss Inc., Germany).
We thank Drs. Bruce Walcheck and Kaylee Schwertfeger for review of the manuscript and Katherine Krieser for technical assistance and review of the manuscript.
Conceived and designed the experiments: CAS. Performed the experiments: MF DBH. Analyzed the data: CAS DBH MF. Contributed reagents/materials/analysis tools: CAS. Wrote the paper: CAS MF.
- 1. Berg EL, Magnani J, Warnock RA, Robinson MK, Butcher EC (1992) Comparison of L-selectin and E-selectin ligand specificities: the L-selectin can bind the E-selectin ligands sialyl Le(x) and sialyl Le(a). Biochem Biophys Res Commun 184: 1048–1055.
- 2. Lafrenie RM, Buchanan MR, Orr FW (1993) Adhesion molecules and their role in cancer metastasis. Cell Biophys 23: 3–89.
- 3. Mannori G, Crottet P, Cecconi O, Hanasaki K, Aruffo A, et al. (1995) Differential colon cancer cell adhesion to E-, P-, and L-selectin: role of mucin-type glycoproteins. Cancer Res 55: 4425–4431.
- 4. Lewinsohn DM, Bargatze RF, Butcher EC (1987) Leukocyte-endothelial cell recognition: evidence of a common molecular mechanism shared by neutrophils, lymphocytes, and other leukocytes. J of Immunol 138: 4313–4321.
- 5. Larsen E, Palabrica T, Sajer S, Gilbert GE, Wagner DD, et al. (1990) PADGEM-dependent adhesion of platelets to monocytes and neutrophils is mediated by a lineage-specific carbohydrate, LNF III (CD15). Cell 63: 467–474.
- 6. Geng JG, Bevilacqua MP, Moore KL, McIntyre TM, Prescott SM, et al. (1990) Rapid neutrophil adhesion to activated endothelium mediated by GMP-140. Nature 343: 757–760.
- 7. Bevilacqua MP, Pober JS, Mendrick DL, Cotran RS, Gimbrone MAJ (1987) Identification of an inducible endothelial-leukocyte adhesion molecule. Proc Natl Acad Sci USA 84: 9238–9242.
- 8. Lowe JB (2002) Glycosylation in the control of selectin counter-receptor structure and function. Immunol Rev 186: 19–36.
- 9. Ellies LG, S. T, Petryniak B, Lowe JB, Fukada M, et al. (1998) Core 2 oligosaccharide biosynthesis distinguishes between selectin ligands essential for leukocyte homing and inflammation. Immunity 9: 881–890.
- 10. Sperandio M, Thatte A, Foy D, Ellies LG, Marth JD, et al. (2001) Severe impairment of leukocyte rolling in venules of core 2 glucosaminyltransferase-deficient mice. Blood 97: 3812–3819.
- 11. Leppänen A, Mehta P, Ouyang YB, Ju T, Helin J, et al. (1999) A novel glycosulfopeptide binds to P-selectin and inhibits leukocyte adhesion to P-selectin. J Biol Chem 274: 24838–24848.
- 12. Li F, Wilkins PP, Crawley S, Weinstein J, Cummings RD, et al. (1996) Post-translational modifications of recombinant P-selectin glycoprotein ligand-1 required for binding to P- and E-selectin. J Biol Chem 271: 3255–3264.
- 13. Ramachandran V, Nollert MU, Qui H, Liu WJ, Cummings RD, et al. (1999) Tyrosine replacement in P-selectin glycoprotein ligand-1 affects distinct kinetic and mechanical properties of bonds with P- and L-selectin. Proc Natl Acad Sci USA 96: 13771–13776.
- 14. Epperson TK, Patel KD, McEver RP, Cummings RD (2000) Noncovalent association of P-selectin glycoprotein ligand-1 and minimal determinants for binding to P-selectin. J Biol Chem 275: 7839–7853.
- 15. Kannagi R, Izawa M, Koike T, Miyazaki K, Kimura N (2004) Carbohydrate-mediated cell adhesion in cancer metastasis and angiogenesis. Cancer Sci 95: 377–384.
- 16. Irimura T, Nakamori S, Matsushita Y, Taniuchi Y, Todoroki N, et al. (1993) Colorectal cancer metastasis determined by carbohydrate-mediated cell adhesion: role of sialyl-Lex antigens. Semin Cancer Biol 4: 319–324.
- 17. Nakamori S, Kameyama M, Imaoka S, Furukawa H, Ishikawa O, et al. (1993) Increased expression of sialyl Lewisx antigen correlates with poor survival in patients with colorectal carcinoma: clinicopathological and immunohistochemical study. Cancer Res 53: 3632–3637.
- 18. Renkonen J, Paavonen T, Renkonen R (1997) Endothelial and epithelial expression of sialyl Lewis(x) and sialyl Lewis(a) in lesions of breast carcinoma. Int J Cancer 74: 296–300.
- 19. Haier J, Nasralla M, Nicolson G (2000) Cell surface molecules and their prognostic values in assessing colorectal carcinomas. Ann Surg 231: 11–24.
- 20. Renkonen J, Heinävaara S, Sankila R, Paavonen T (2000) Endothelial and epithelial expression of sialyl Lewis x in squamous carcinoma of the tongue. APMIS 108: 705–712.
- 21. Pauli BU, Augustin-Voss HG, el-Sabban ME, Johnson RC, Hammer DA (1990) Organ-preference of metastasis. The role of endothelial cell adhesion molecules. Cancer Metastasis Rev 9: 175–189.
- 22. Ito K, Ye CL, Hibi K, Mitsuoka C, Kannagi R, et al. (2001) Paired tumor marker of soluble E-selectin and its ligand sialyl Lewis A in colorectal cancer. J Gastroenterol 36: 823–829.
- 23. Zhang BH, Chen H, Yao XP, Cong WM, Wu MC (2002) E-selectin and its ligand-sLeX in the metastasis of hepatocellular carcinoma. Hepatobiliary Pancreat Dis Int 1: 80–82.
- 24. Hakomori S (1996) Tumor malignancy defined by aberrant glycosylation and sphingo(glyco)lipid metabolism. Cancer Res 56: 5309–5318.
- 25. St. Hill CA, Bullard KM, Walcheck B (2005) Expression of the high-affinity selectin glycan ligand C2-O-sLeX by colon carcinoma cells. Cancer Lett 217: 105–113.
- 26. St. Hill CA, Farooqui M, Mitcheltree G, Gulbahce HE, Jessurun J, et al. (2009) The high affinity selectin glycan ligand C2-O-sLex and mRNA transcripts of the core 2 beta-1,6-N-acetylglucosaminyltransferase (C2GnT1) gene are highly expressed in human colorectal adenocarcinomas. BMC Cancer 9: 79.
- 27. Mitsakos A, Hanisch FG, Uhlenbruck G (1988) Biosynthesis of the cancer-associated sialyl-Lex determinant in human amniotic fluid. Biol Chem Hoppe-Seyler 369: 661–665.
- 28. Walcheck B, Leppänen A, Cummings RD, Knibbs RN, Stoolman LM, et al. (2002) The monclonal antibody CHO-131 binds to a core 2 O-glycan terminated with sialyl-Lewis x, which is a functional glycan ligand for P-selectin. Blood 99: 4063–4069.
- 29. Yago T, Leppanen A, Carlyon JA, Akkoyunlu M, Karmakar S, et al. (2003) Structurally distinct requirements for binding of P-selectin glycoprotein ligand-1 and sialyl Lewis x to Anaplasma phagocytophilum and P-selectin. J Biol Chem 278: 37987–37997.
- 30. Smith MJ, Smith BR, Lawrence MB, Snapp KR (2004) Functional analysis of the combined role of the O-linked branching enzyme core 2 beta1-6-N-glucosaminyltransferase and dimerization of P-selectin glycoprotein ligand-1 in rolling on P-selectin. J Biol Chem 279: 21984–21991.
- 31. Shimodaira K, Nakayama J, Nakamura N, Hasebe O, Tsutomu K, et al. (1997) Carcinoma-associated expression of core 2 β-1,6-N-acetylglucosaminyltransferase gene in human colorectal cancer: role of O-glycans in tumor progression. Cancer Res 57: 5201–5206.
- 32. Machida E, Nakayama J, Amano J, Fukuda M (2001) Clinicopathological significance of core 2 β1,6-N-acetylglucosaminyltransferase messenger RNA expressed in the pulmonary adenocarcinoma determined by in situ hybridization. Cancer Res 61: 2226–2231.
- 33. Renkonen J, Räbinä J, Mattila P, Grenman R, Renkonen R (2001) Core 2 β1,6-N-acetylglucosaminyltransferases and α1,3-fucosyltransferases regulate the synthesis of O-glycans on selectin ligands on oral cavity carcinoma cells. APMIS 109: 500–506.
- 34. Weiss L, Orr FW, Honn KV (1989) Interactions between cancer cells and the microvasculature: a rate-regulator for metastasis. Clin Exp Metastasis 7: 127–167.
- 35. Brockhausen I, Kuhns W, Schachter H, Matta KL, Sutherland DR, et al. (1991) Biosynthesis of O-glycans in leukocytes from normal donors and from patients with leukemia: increase in O-glycan core 2 UDP-GlcNAc:Gal beta 3 GalNAc alpha-R (GlcNAc to GalNAc) beta(1-6)-N-acetylglucosaminyltransferase in leukemic cells. Cancer Res 51: 1257–1263.
- 36. Saitoh O, Piller F, Fox RI, Fukuda M (1991) T-lymphocytic leukemia expresses complex, branched O-linked oligosaccharides on a major sialoglycoprotein, leukosialin. Blood 77: 1491–1499.
- 37. Yousefi S, Higgins E, Daoling Z, Pollex-Krüger A, Hindsgaul O, et al. (1991) Increased UDP-GlcNAc:Gal beta 1-3GaLNAc-R (GlcNAc to GaLNAc) beta-1, 6-N-acetylglucosaminyltransferase activity in metastatic murine tumor cell lines. Control of polylactosamine synthesis. J Biol Chem 266: 1772–1782.
- 38. Hagisawa S, Ohyama C, Takahashi T, Endoh M, Moriya T, et al. (2005) Expression of core 2 β1,6-N-acetylglucosaminyltransferase facilitates prostate cancer progression. Glycobiology 15: 1016–1024.
- 39. Sawada R, Tsuboi S, Fukuda M (1994) Differential E-selectin-dependent adhesion efficiency in sublines of human colon cancer exhibiting distinct metastatic potentials. J Biol Chem 269: 1425–1431.
- 40. Beum PV, Singh J, Burdick M, Hollingsworth M, Cheng PW (1999) Expression of core 2 β-1,6-N-acetylglucosaminyltransferase in a human pancreatic cancer cell line results in altered expression of MUC1 tumor associated epitopes. J Biol Chem 274: 24641–24648.
- 41. Singh PK, Hollingsworth MA (2006) Cell surface-associated mucins in signal transduction. Trends Cell Biol 16: 467–476.
- 42. Nakamori S, Ota DM, Cleary KR, Shirotani K, Irimura T (1994) MUC1 mucin expression as a marker of progression and metastasis of human colorectal carcinoma. Gastroenterology 106: 353–361.
- 43. Hiraga Y, Tanaka S, Haruma K, Yoshihara M, Sumii K, et al. (1998) Immunoreactive MUC1 expression at the deepest invasive portion correlates with prognosis of colorectal cancer. Oncology 55: 307–319.
- 44. Lee MJ, Lee HS, Kim WH, Choi Y, Yang M (2003) Expression of mucins and cytokeratins in primary carcinomas of the digestive system. Mod Pathol 16: 403–410.
- 45. Zhang K, Baeckström D, Brevinge H, Hansson GC (1996) Secreted MUC1 mucins lacking their cytoplasmic part and carrying sialyl-Lewis a and x epitopes from a tumor cell line and sera of colon carcinoma patients can inhibit HL-60 leukocyte adhesion to E-selectin-expressing endothelial cells. J Cell Biochem 60: 538–549.
- 46. Leeman MF, Curran S, Murray GI (2003) New insights into the roles of matrix metalloproteinases in colorectal cancer development and progression. J Pathol 201: 528–534.
- 47. Harada T, Arii S, Mise M, Imamura T, Higashitsuji H, et al. (1998) Membrane-type matrix metalloproteinase-1(MT1-MTP) gene is overexpressed in highly invasive hepatocellular carcinomas. J Hepatol 28: 231–239.
- 48. Thathiah A, Carson DD (2004) MT1-MMP mediates MUC1 shedding independent of TACE/ADAM17. Biochem J 382(Pt 1): 363–373.
- 49. Napier SL, Healy ZR, Schnaar RL, Konstantopoulos K (2007) Selectin ligand expression regulates the initial vascular interactions of colon carcinoma cells: the roles of CD44V and alternative sialofucosylated selectin ligands. J Biol Chem 282: 3433–3441.
- 50. Ratnikov BI, Rozanov DV, Postnova TI, Baciu PG, Zhang H, et al. (2002) An alternative processing of integrin alpha(v) subunit in tumor cells by membrane type-1 matrix metalloproteinase. J Biol Chem 277: 7377–7385.
- 51. Bates RC, Bellovin DI, Brown C, Maynard E, Wu B, et al. (2005) Transcriptional activation of integrin beta6 during the epithelial-mesenchymal transition defines a novel prognostic indicator of aggressive colon carcinoma. J Clin Invest 339–347.
- 52. Dalziel M, Whitehouse C, McFarlane I, Brockhausen I, Gschmeissner S, et al. (2001) The relative activities of the C2GnT1 and ST3Gal-I glycosyltransferases determine O-glycan structure and expression of a tumor-associated epitope on MUC1. J Biol Chem 276: 11007–11015.
- 53. De Graaf TW, Van der Stelt ME, Anbergen MG, van Dijk W (1993) Inflammation-induced expression of sialyl Lewis X-containing glycan structures on alpha 1-acid glycoprotein (orosomucoid) in human sera. J Exp Med 177: 657–666.
- 54. Katnik I, Goodarzi MT, Turner GA (1996) An improved ELISA for the determination of sialyl Lewis(x) structures on purified glycoconjugates. Glycoconj J 13: 1043–1047.
- 55. Brinkman-van der Linden EC, de Haan PF, Havenaar EC, van Dijk W (1998) Inflammation-induced expression of sialyl LewisX is not restricted to alpha1-acid glycoprotein but also occurs to a lesser extent on alpha1-antichymotrypsin and haptoglobin. Glycoconj J 15: 177–182.
- 56. Kannagi R, Izawa M, Koike T, Miyazaki K, Kimura N (2004) Carbohydrate-mediated cell adhesion in cancer metastasis and angiogenesis. Cancer Sci 95: 377–384.
- 57. Williams D, Schachter H (1980) Mucin synthesis. I. Detection in canine submaxillary glands of an N-acetylglucosaminyltransferase which acts on mucin substrates. J Biol Chem 255: 11247–11252.
- 58. Bierhuizen MF, Fukuda M (1992) Expression cloning of a cDNA encoding UDP-GlcNAc:Gal beta 1-3-GalNAc-R (GlcNAc to GalNAc) beta 1-6GlcNAc transferase by gene transfer into CHO cells expressing polyoma large tumor antigen. Proc Natl Acad Sci USA 89: 9326–9330.
- 59. Huang M-C, Chen H-Y, Huang H-C, Huang J, Liang J-T, et al. (2006) C2GnT-M is downregulated in colorectal cancer and its re-expression causes growth inhibition of colon cancer cells. Oncogene 25: 3267–3276.
- 60. Descheny L, Gainers ME, Walcheck B, Dimitroff CJ (2006) Ameliorating skin-homing receptors on malignant T cells with a fluorosugar analog of N-acetylglucosamine: P-selectin ligand is a more sensitive target than E-selectin ligand. J Invest Dermatol 126: 2065–2073.
- 61. Ni Z, Campbell JJ, Niehans G, Walcheck B (2006) The monoclonal antibody CHO-131 identifies a subset of cutaneous lymphocyte-associated antigen T cells enriched in P-selectin-binding cells. J Immunol 177: 4742–4748.
- 62. Urdal DL, Brentrall TA, Bernstein ID, Hakomori S-l (1983) A granulocyte reactive monoclonal antibody, 1G10, identifies the Gal/3-4 (Fuc a 1-3) GlcNAc (X determinant) expressed in HL-60 cells on both glycolipid and glycoprotein molecules. Blood 62 1022-1026:
- 63. St. Hill CA, Alexander SR, Walcheck B (2003) Indirect capture augments leukocyte accumulation on P-selectin in flowing whole blood. J Leukoc Biol 73: 464–471.
- 64. Lawrence MB, Springer TA (1991) Leukocytes roll on a selectin at physiological flow rates: distinction from and prerequisite for adhesion through integrins. Cell 65: 1–20.
- 65. Prorok-Hamon M, Mathieu S, El-Battari A (2006) Visualizing intracellular distribution and activity of core2 β (1,6)-Nacetylglucosaminyltransferase-I in living cells. In: Brockhausen I, editor. Methods in Molecular Biology. Totowa, NJ: Human Press Inc. pp. 171–187.