Nature Immunology 6, 1096 - 1104 (2005)
Published online: 9 October 2005; | doi:10.1038/ni1259
N-acetylglucosamine-6-O-sulfotransferases 1 and 2 cooperatively control lymphocyte homing through L-selectin ligand biosynthesis in high endothelial venulesHiroto Kawashima1, Bronislawa Petryniak2, Nobuyoshi Hiraoka1, Junya Mitoma1, Valerie Huckaby1, Jun Nakayama3, Kenji Uchimura4, Kenji Kadomatsu4, Takashi Muramatsu4, John B Lowe2
& Minoru Fukuda11 Glycobiology Program, Cancer Research Center, The Burnham Institute, La Jolla 92037, California, USA. 2 Howard Hughes Medical Institute, Life Sciences Institute, University of Michigan Medical School, Ann Arbor, Michigan 48109, USA. 3 Department of Pathology, Shinshu University School of Medicine, Matsumoto 390-8621, Japan. 4 Department of Biochemistry, Nagoya University School of Medicine, Nagoya 466-8550, Japan.
Correspondence should be addressed to Minoru Fukuda minoru@burnham.org Lymphocyte homing is mediated by specific interactions between L-selectin on lymphocytes and sulfated carbohydrates restricted to high endothelial venules in lymph nodes. Here we generated mice deficient in both N-acetylglucosamine-6-O-sulfotransferase 1 (GlcNAc6ST-1) and GlcNAc6ST-2 and found that mutant mice had approximately 75% less homing of lymphocytes to the peripheral lymph nodes than did wild-type mice. Consequently, these mice had lower contact hypersensitivity responses than those of wild-type mice. Carbohydrate structural analysis showed that 6-sulfo sialyl Lewis X, a dominant ligand for L-selectin, was almost completely absent from the high endothelial venules of these mutant mice, whereas the amount of unsulfated sialyl Lewis X was much greater. These results demonstrate the essential function of GlcNAc6ST-1 and GlcNAc6ST-2 in L-selectin ligand biosynthesis in high endothelial venules and their importance in immune surveillance.Lymphocytes encounter antigens derived from foreign pathogens and initiate immune responses in secondary lymphoid organs such as lymph nodes and Peyer's patches. Therefore, homing of lymphocyte to secondary lymphoid organs is an essential process for immune surveillance. Lymphocyte homing is critically dependent on a specific interaction between the lymphocyte-homing receptor L-selectin and its ligands, whose expression is restricted to high endothelial venules (HEVs) in the lymph nodes1. This adhesive interaction mediates lymphocyte tethering and rolling on the surface of HEVs, which is a prerequisite for the subsequent lymphocyte chemokine-dependent activation, integrin-mediated firm attachment to the endothelium and transmigration across blood vessels2,
3,
4,
5.
L-selectin is a carbohydrate-binding protein that binds to its ligands on HEVs in a calcium-dependent way. HEV-restricted ligands for L-selectin include GlyCAM-1, CD34, podocalyxin-like protein, Sgp200, endoglycan and MAdCAM-1, all of which have mucin-like domains that act as scaffolding for O-glycans6. The ability of these ligands to function entirely depends on their 'decoration' with specific carbohydrate structures, including 6-sulfo sialyl Lewis X (sialyl- (2-3)-galactopyranosyl- (1-4)-[fucopyranosyl-(1-3)(sulfo-6)]-N-acetylglucosamine) which contains fucose, sialic acid and sulfate6.
The pivotal function of fucosylation of L-selectin ligands for their interaction with L-selectin has been established by genetic studies of mice. Mice deficient in fucosyltransferase VII (FucT-VII) have a reduction of nearly 80% in homing of lymphocytes to the peripheral lymph nodes, indicating that FucT-VII is the principal enzyme involved in fucosylation of L-selectin ligands on HEVs7. Studies using mice doubly deficient in both FucT-VII and FucT-IV have demonstrated an additional function for FucT-IV in HEV ligand fucosylation8. The essential involvement of sialylation of L-selectin ligands has also been demonstrated by experiments showing sialidase treatment of lymph node sections completely abrogates L-selectin-dependent binding of lymphocytes to HEVs in peripheral lymph nodes9.
Sulfation of L-selectin ligands is important in the interaction with L-selectin in vitro. An initial demonstration of such a requirement showed that treatment of lymph node organ culture with chlorate, an inhibitor of sulfation, abrogates interaction between recombinant L-selectin and its glycoprotein ligands10,
11. Subsequently, the use of a specific antibody identified the main capping group of L-selectin ligands on HEVs in human lymph nodes as 6-sulfo sialyl Lewis X12. The HEV-restricted sulfotransferase N-acetylglucosamine-6-O-sulfotransferase-2 (GlcNAc6ST-2, also called HEC-GlcNAc6ST or L-selectin ligand sulfotransferase) has been cloned and participates in the biosynthesis of 6-sulfo sialyl Lewis X and can reconstitute the L-selectin ligand in vitro13,
14. Furthermore, the 6-sulfo sialyl Lewis X structure is present in either the core 2 or extended core 1 branch or in both branches of L-selectin ligand O-glycans15. Moreover, the MECA-79 antibody16, which is widely used to detect HEVs in lymph nodes or HEV-like vessels at the sites of chronic inflammation, recognizes O-glycans containing 6-sulfo N-acetylglucosamine in the extended core 1 structure15. Although those studies have indicated a sulfation requirement of HEV-borne carbohydrates as L-selectin ligands in vitro, that requirement has not been demonstrated in vivo.
For assessment of the function of the sulfation of L-selectin ligands in vivo, mutant mice deficient in the HEV-restricted sulfotransferase GlcNAc6ST-2 have been generated17,
18. In these mice, the binding of MECA-79 to lymph node HEVs is substantially diminished, except for binding in the lining of HEVs on the surface away from the lumen, suggesting that the GlcNAc-6-O-sulfation in the extended core 1 branch is mediated mainly by GlcNAc6ST-2. Oligosaccharide structural analysis has indicated that GlcNAc-6-O-sulfation in the extended core 1 structure on GlyCAM-1 is substantially less in GlcNAc6ST-2-deficient mice than in wild-type mice17; however, the GlcNAc-6-O-sulfation in the core 2–branched O-glycan persists in GlcNAc6ST-2-deficient mice. Consistent with that observation, there is only partial disruption of lymphocyte homing (approximately 50%) in GlcNAc6ST-2-deficient mice17,
18.
Four mouse GlcNAc-6-O-sulfotransferases have been identified19,
20. GlcNAc6ST-1 is widely expressed in various tissues, including lymph node HEVs21, and functions in L-selectin ligand biosynthesis in vitro22 and in vivo23. GlcNAc6ST-1-deficient mice have a reduction of an approximately 20% in homing of lymphocytes to peripheral lymph nodes23, suggesting that GlcNAc6ST-1 and GlcNAc6ST-2 might have overlapping or complementary functions. To determine the sulfation requirement of L-selectin ligands in vivo in further detail, we generated mice deficient in both GlcNAc6ST-1 and GlcNAc6ST-2 and show here that these sulfotransferases cooperatively have a chief function in L-selectin ligand biosynthesis in HEVs. Systematic carbohydrate analysis of an HEV-specific glycoprotein showed that essentially no 6-sulfo sialyl Lewis X structure was synthesized in the HEVs of mice deficient in both GlcNAc6ST-1 and GlcNAc6ST-2 ('double-knockout' mice), but that the amount of the unsulfated sialyl Lewis X structure was up to sevenfold greater than that in wild-type mice. Moreover, contact hypersensitivity responses were substantially diminished in the double-knockout mice because of a reduction in lymphocyte trafficking to the draining lymph nodes. Our study provides a link between the structural alterations of carbohydrates in HEVs and lymphocyte recruitment in health and disease.
Results
Expression of L-selectin ligands in HEVs
To determine the expression of L-selectin-reactive carbohydrates in HEVs of mutant mice, we did immunofluorescence studies using MECA-79 and an L-selectin–immunoglobulin M (L-sel–IgM) chimeric protein. MECA-79 recognizes extended core 1 structures modified with GlcNAc-6-O-sulfate, whereas L-selectin interacts with various O-glycans modified with 6-sulfo sialyl Lewis X present in HEVs15 (Fig. 1). Binding of MECA-79 to HEVs in peripheral lymph nodes and mesenteric lymph nodes was completely abolished in double-knockout mice (Fig. 2a). Only expression of enhanced green fluorescent protein (EGFP), which replaced the catalytic domain and stem region of GlcNAc6ST-2 in the targeting vector17, was detectable in HEVs. In Peyer's patches of mice deficient in GlcNAc6ST-1 alone, MECA-79 staining was undetectable, consistent with published results23. In contrast, the binding of L-sel–IgM to HEVs of peripheral lymph nodes and mesenteric lymph nodes was not completely abolished in the double-knockout mice, although it was substantially reduced relative to that of wild-type mice or mice deficient in either GlcNAc6ST-1 or GlcNAc6ST-2 (Fig. 2b). These results indicated that GlcNAc-6-O-sulfation in the extended core 1 branch of O-glycans was abrogated in double-knockout mice but that L-selectin-reactive carbohydrates were still expressed in low abundance.
 | | |
Lymphocyte trafficking to lymph nodes
We next determined whether lymphocyte homing was affected in double-knockout mice. The number of lymphocytes in peripheral lymph nodes was reduced by 60% in double-knockout mice compared with that in wild-type mice (Fig. 3a). In a short-term homing assay, mice deficient in either GlcNAc6ST-1 or GlcNAc6ST-2 had a reduction of approximately 20% or 50%, respectively, in homing of lymphocytes to peripheral lymph nodes, whereas double-knockout mice had a reduction of approximately 75% (Fig. 3b). The residual homing of lymphocytes to peripheral lymph nodes in double-knockout mice was apparently mediated by L-selectin, as it was reduced to background by MEL-14, an antibody blocking L-selectin function (Fig. 3c). Lymphocyte homing to mesenteric lymph nodes and Peyer's patches was only partially blocked by MEL-14, possibly because interactions between 47 integrin and MAdCAM-1 are also involved in homing of lymphocytes to these secondary lymphoid organs24.
| | Figure 3. Lymphocyte trafficking to secondary lymphoid organs. | | |  | (a) Lymphocytes recovered from lymphoid organs of 7-week-old wild-type and null mice (n = 5–7 mice). *, P < 0.01, versus wild-type mice; **, P < 0.01, versus GlcNAc6ST-1-deficient mice; ***, P < 0.02, versus GlcNAc6ST-2-deficient mice. (b,c) Flow cytometry of lymphocytes in lymphoid organs. CMFDA-labeled lymphocytes (2.5  107 cells) were injected into tail veins of wild-type and null mice (b) or CMFDA-labeled lymphocytes (2.5 107 cells) incubated with 30  g of MEL-14 (rat IgG2a) or rat IgG were injected into tail veins of double-knockout mice (c). Then, 1 h later, fluorescent lymphocytes in lymphocyte suspensions from lymphoid organs were quantified. At least four recipient mice were tested in each experiment. (b) For PLN, *, P < 0.01, versus wild-type mice; **, P < 0.01, versus GlcNAc6ST-1-deficient mice; ***, P < 0.01, versus GlcNAc6ST-2-deficient mice. For MLN, *, P < 0.01, versus wild-type mice; **, P < 0.01, versus GlcNAc6ST-1-deficient mice; ***, P < 0.1, versus GlcNAc6ST-2-deficient mice. (c) *, P < 0.01, versus rat IgG-injected mice. (d–f) Lymphocyte rolling assay on GlyCAM-1 from various mouse lines captured by antibody to GlyCAM-1 at concentrations of 20 g/ml (d,e) and 50 g/ml (f). Data are representative of three (b,c) or two (d–f) independent experiments.
 Full Figure and legend (55K) |
| |
The additive reduction in lymphocyte homing in the double-knockout mice was consistent with results of an in vitro rolling assay, in which we applied normal lymphocytes to flow chambers coated with GlyCAM-1, an HEV-specific secreted glycoprotein, from the sera of wild-type or mutant mice (Fig. 3d–f). We noted no lymphocyte rolling when we used lymphocytes from L-selectin-deficient mice in this assay (data not shown), indicating that the rolling was mediated exclusively by L-selectin. GlyCAM-1 derived from mutant mice deficient in either GlcNAc6ST-1 or GlcNAc6ST-2 supported less lymphocyte rolling than that from wild-type mice (Fig. 3d). GlyCAM-1 derived from double-knockout mice did not support lymphocyte rolling when a low density of GlyCAM-1 was used. However, there were many transiently tethered lymphocytes in the same conditions (Fig. 3e), suggesting that the initial tethering of lymphocytes is less affected than the rolling velocity in double-knockout mice. GlyCAM-1 derived from double-knockout mice weakly supported rolling when we used a high density of GlyCAM-1 (Fig. 3d versus f), further supporting the idea that these mice have only weak L-selectin ligand activity.
Contact hypersensitivity response
To determine the relevance of GlcNAc6ST-1 and GlcNAc6ST-2 in immune responses in pathophysiological settings, we assessed contact hypersensitivity responses in mutant mice. Double-knockout mice showed significant reductions in ear swelling and leukocyte infiltration after sensitization and challenge with the hapten DNFB (2,4-dinitrofluorobenzene; Fig. 4a,b). HEV-like vessels reactive with MECA-79 or with L-sel–IgM were not detected in the ears of wild-type or mutant mice after the challenge (data not shown), suggesting that the reduction in the contact hypersensitivity response in double-knockout mice could have been due to qualitative or quantitative differences in the immune responses in the draining lymph nodes. In vitro proliferation of lymphocytes in response to the mitogen concanavalin A or to the water-soluble analog of DNFB, DNBS (2,4-dinitrobenzene sulfonic acid)25, was almost indistinguishable in wild-type versus mutant mice (Fig. 4c,d). These data indicate that antigen-specific lymphocytes were present at almost the same ratio in the draining lymph nodes of both mouse lines. Notably, however, the total number of lymphocytes decreased by 70% in double-knockout mice (Fig. 4e). Consistent with that observation, lymphocyte trafficking to the draining lymph nodes significantly decreased in double-knockout mice in a short-term homing assay (Fig. 4f). These results collectively suggest that the reduced contact hypersensitivity response of double-knockout mice could be attributed to the quantitative decrease in the total number of antigen-specific lymphocytes recruited to the draining lymph nodes.
| | Figure 4. Reduced contact hypersensitivity in the double-knockout mice. | | |  | (a) Ear swelling 24 h after challenge with DNFB or vehicle alone in wild-type and null mice (horizontal axis; n = 5). *, P < 0.01, versus wild-type mice; **, P < 0.01, versus GlcNAc6ST-1-deficient mice; ***, P < 0.01, versus GlcNAc6ST-2-deficient mice. (b) Hematoxylin-and-eosin staining of ear sections 24 h after DNFB challenge. *, ear cartilage; arrowheads, recruited leukocytes. Scale bar, 100 m. (c,d) Cell proliferation assay of lymphocytes from inguinal lymph nodes sensitized with DNFB and cultured in the presence of concanavalin A (Con A; 2 g/ml) or DNBS (200 g/ml; n = 3 mice). A490, absorbance at 490 nm. (e) Lymphocytes recovered from the inguinal lymph nodes (ILN) of DNFB-sensitized mice (horizontal axis; n = 6–8 mice). *, P < 0.01, versus wild-type mice; **, P < 0.01, versus GlcNAc6ST-1-deficient mice; ***, P < 0.01, versus GlcNAc6ST-2-deficient mice. (f) Trafficking of CMFDA-labeled lymphocytes to the inguinal lymph nodes of DNFB-sensitized mice (horizontal axis; n = 3 mice). *, P < 0.01, versus wild-type mice. Data (a–d,f) are representative of three independent experiments.
Full Figure and legend (97K) |
| |
Carbohydrate structural analysis of GlyCAM-1
To gain insight into the structural basis of the substantial reduction in lymphocyte homing in the normal and pathophysiological conditions described above, we undertook carbohydrate structural analysis of GlyCAM-1. We first radiolabeled lymph nodes from wild-type and mutant mice with [3H]galactose in organ culture and then released the O-glycans on GlyCAM-1 and did structural analysis. After removing sialic acid by mild acid hydrolysis, we fractionated oligosaccharides by gel filtration and anion-exchange column chromatography (Supplementary Fig. 1 online). We then analyzed unsulfated oligosaccharides by exoglycosidase treatment and gel filtration. The amount of unsulfated oligosaccharides, including those containing the sialyl Lewis X structure, was substantially increased in double-knockout mice (Fig. 5a). We also analyzed sulfated oligosaccharide fractions in detail by exoglycosidase treatment and high-performance liquid chromatography (HPLC; Supplementary Fig. 2 online). We found only very little oligosaccharide containing GlcNAc-6-O-sulfate in double-knockout mice (Fig. 5a). Conversely, the relative amount of sulfated oligosaccharides containing galactose-6-O-sulfate increased considerably in double-knockout mice (O-glycan structures, Supplementary Fig. 3 online).
| | |
We calculated the amount of sialylated oligosaccharides from the amount of [3H]galactose released by treatment of the total oligosaccharide fraction with a mixture of -galactosidase and -(1,3)- and -(1,4)-fucosidase before and after the removal of sialic acid (Supplementary Fig. 1 online). Sialic acid modification was present at the termini of 89.5%, 87.2%, 87.5% and 90.5% of the oligosaccharides in wild-type, GlcNAc6ST-1-deficient, GlcNAc6ST-2-deficient and double-knockout mice, respectively. Combined with the detailed structural analysis of the desialylated oligosaccharides, these results indicate that the 6-sulfo sialyl Lewis X structure was almost completely abrogated, whereas unsulfated sialyl Lewis X was overexpressed in the double-knockout mice (Fig. 5b).
Expression and substrate specificity of sulfotransferases
As shown above, some sulfated O-glycans were expressed in double-knockout mice. To determine which sulfotransferases were expressed and were involved in the biosynthesis of sulfated O-glycans in HEVs, we next did RT-PCR analysis using total RNA from peripheral lymph node HEV cells of wild-type mice, prepared by immunomagnetic selection with MECA-79 (Fig. 6a). After confirming expression of GlyCAM-1 and lack of expression of L-selectin in the MECA-79+ HEV cell preparations, we examined the expression of sulfotransferases in these cell preparations. In addition to GlcNAc6ST-1 and GlcNAc6ST-2, we detected GlcNAc6ST-4 (refs. 26,27). Expression of GlcNAc6ST-3 (ref. 28) was not detectable in these cell preparations, although it was readily detectable in the same conditions in mouse cornea, which is known to express this sulfotransferase29. All four mouse GlcNAc6STs efficiently transferred sulfate to core 2–branched O-glycans, as shown by [35S]Na2SO4 incorporation. In contrast, only GlcNAc6ST-2 and GlcNAc6ST-3 efficiently transferred sulfate to extended core 1 O-glycans, resulting in the generation of substantial MECA-79 epitope, as shown by immunoblot (Fig. 6b). GlcNAc6ST-1 inefficiently synthesized the MECA-79 epitope, whereas GlcNAc6ST-4 did not synthesize detectable amounts of this epitope, consistent with the finding that the MECA-79 epitope was abolished in the double-knockout mice (Fig. 2a). These results combined suggest that the small amount of core 2–branched O-glycans containing GlcNAc-6-O-sulfate identified in the double-knockout mice were synthesized by GlcNAc6ST-4. We also detected by RT-PCR expression of the galactose-6-O-sulfotransferase keratan sulfate sulfotransferase (KSST)30, which is probably involved in the biosynthesis of the O-glycans containing galactose-6-O-sulfate identified in our carbohydrate analysis.
| | Figure 6. Expression of sulfotransferases in HEVs and characterization of remaining lymphocyte homing in double-knockout mice. | | |  | (a) Expression of sulfotransferase genes. For this PCR, templates were single-stranded cDNAs constructed from MECA-79+ HEVs from peripheral lymph nodes (PLN 79+), whole-cell suspensions from peripheral lymph nodes (PLN whole) or corneas from wild-type mice (Cornea) or plasmid DNA containing full-length sulfotransferase cDNA (Plasmid). PCR was done in the presence (+) or absence (-) of reverse transcriptase (RT). Chst2, GlcNAc6ST-1; Chst4, GlcNAc6ST-2; Chst5, GlcNAc6ST-3; Chst7, GlcNAc6ST-4; Chst1, KSST; Glycam1, GlyCAM-1; Sell, L-selectin; Gapdh, glyceraldehyde-3-phosphate dehydrogenase. (b) Immunoblot of MECA-79. Lec2 cells were transiently transfected with expression vectors encoding GlyCAM-1-IgG without (N) or with Core2GlcNAcT-I (C2) and Core1-3GlcNAcT (C1), together with nothing else (Mock) or with GlcNAc6ST-1–GlcNAc6ST-4. Cells were metabolically labeled with [35S]Na2SO4. GlyCAM-1-IgG was purified with protein A–Sepharose and was analyzed by fluorography and immunoblot (IB) with MECA-79. The amount of GlyCAM-1-IgG was normalized by immunoblot with horseradish peroxidase–labeled antibody to human IgG (HRP-hIgG). (c) Expression of L-selectin ligands. Frozen sections (7 m in thickness) of peripheral lymph nodes from wild-type and mutant mice pretreated with sialidase (from Vibrio cholerae; 200 mU/ml) or not pretreated and stained with L-sel–IgM. Scale bar, 50 m. (d) Trafficking of lymphocytes to lymphoid organs of double-knockout mice. CMFDA-labeled lymphocytes were injected into double-knockout mice 1 h after injection of phosphate-buffered saline (PBS), AAA lectin (100 g) in PBS, or MAA lectin (40 g) in PBS. Trafficking is presented as a percentage of that in PBS-injected mice, which was set as 100% (n = 3). *, P < 0.01, and **, P < 0.1, versus PBS-injected mice. Data are representative of three independent experiments.
Full Figure and legend (94K) |
| |
Functions of unsulfated sialyl Lewis X
As described above, the main carbohydrate structural changes in the double-knockout mice were the almost complete absence of the 6-sulfo sialyl Lewis X structure and abundant expression of the unsulfated sialyl Lewis X structure. To determine whether the low but detectable L-selectin ligand activity in the double-knockout mice was mediated by the unsulfated sialyl Lewis X structure, we assessed L-sel–IgM binding (Fig. 6c). Staining of L-sel–IgM on HEVs was abrogated in frozen sections of peripheral lymph nodes from mice doubly deficient in both FucT-IV and FucT-VII, indicating that fucose modification was critically important for binding of L-sel–IgM to the HEVs of peripheral lymph nodes. In addition, binding of L-sel–IgM to HEVs was abolished by sialidase treatment of frozen sections of peripheral lymph nodes from double-knockout mice, indicating that sialic acid modification was also indispensable. In accordance with those observations, preinjection of the fucose-specific lectin Aleuria aurantia agglutinin (AAA), or the (2,3)-linked sialic acid–specific lectin Maackia amurensis agglutinin (MAA) into double-knockout mice inhibited homing of lymphocytes to peripheral lymph nodes by 85.2% or 78.7%, respectively (Fig. 6d). These results suggest that the residual L-selectin ligand activity in double-knockout mice was most likely mediated by the abundantly expressed unsulfated sialyl Lewis X structure.
Discussion
We have shown here that GlcNAc6ST-1 and GlcNAc6ST-2 cooperatively have a chief function in the biosynthesis of L-selectin ligand carbohydrates in HEVs. Mutant mice deficient in both sulfotransferases showed substantially lower lymphocyte trafficking to peripheral lymph nodes, as well as less robust inflammatory responses after sensitization and challenge with the hapten DNFB. We also did a systematic carbohydrate structural analysis of a HEV-derived glycoprotein and have provided a structural 'explanation' for the phenotypes of mutant mice.
We have shown here that more than 25% of the total carbohydrates on GlyCAM-1 are modified with the 6-sulfo sialyl Lewis X capping structure on HEVs in wild-type mice, indicating that this carbohydrate structure is indeed a principal capping structure of GlyCAM-1. Carbohydrate structural analysis of double-knockout mice indicated that GlcNAc6ST-1 and GlcNAc6ST-2 are required for biosynthesis of this capping structure. Although we detected another member of the GlcNAc-6-O-sulfotransferase family, GlcNAc6ST-4, in MECA-79+ HEV cells, structural analysis indicated that it has only a minor function in GlcNAc-6-O-sulfation of L-selectin ligands.
Although both GlcNAc6ST-1 and GlcNAc6ST-2 function in the GlcNAc-6-O-sulfation of L-selectin ligands, these sulfotransferases showed some differences in relation to L-selectin ligand biosynthesis in HEVs. GlcNAc6ST-2 seemed to be the main GlcNAc-6-O-sulfotransferase in peripheral lymph nodes and mesenteric lymph nodes, as staining with MECA-79 and L-sel–IgM binding was greatly diminished in the HEVs of these secondary lymphoid organs in the absence of GlcNAc6ST-2 but not in the absence of GlcNAc6ST-1. Consequently, homing of lymphocytes to these secondary lymphoid organs and the production of 6-sulfo sialyl Lewis X-containing O-glycans were more diminished in GlcNAc6ST-2-deficient mice than in GlcNAc6ST-1-deficient mice. Consistent with that finding, GlcNAc6ST-2 was expressed more abundantly in HEVs of peripheral lymph nodes than was GlcNAc6ST-1, as assessed by RT-PCR. In contrast, GlcNAc6ST-1 seemed to be the main GlcNAc-6-O-sulfotransferase in Peyer's patches, as MECA-79 antibody staining was abrogated in the absence of GlcNAc6ST-1. In accordance with that result, we found that expression of GlcNAc6ST-2 in Peyer's patches was undetectable, as judged by the fluorescence of the GlcNAc6ST-2–EGFP chimeric protein expressed under control of the promoter of the gene encoding GlcNAc6ST-2 (ref. 17).
Furthermore, GlcNAc6ST-2 seemed to be more relevant in the biosynthesis of fucosylated and sulfated O-glycans than GlcNAc6ST-1, because in mice deficient in GlcNAc6ST-2 alone, the percentage of O-glycans containing both fucose and sulfate was much lower, whereas the percentage of O-glycans containing only GlcNAc-6-O-sulfate but not fucose was slightly higher. The different ability in forming fucosylated and sulfated products may be due to the difference in the subcellular localizations of GlcNAc6ST-1 and GlcNAc6ST-2. GlcNAc6ST-1 is confined to the trans-Golgi network, whereas GlcNAc6ST-2 is distributed throughout the Golgi apparatus31; the latter may be more suitable to supply sulfated O-glycan acceptors to FucT-VII, a broadly Golgi-distributed glycosyltransferase32, although further experimental verification is required to clarify that point.
Moreover, the acceptor substrate specificity of GlcNAc6ST-1 and GlcNAc6ST-2 overlaps but differs. Both sulfotransferases efficiently transferred sulfate to core 2 O-glycans, whereas only GlcNAc6ST-2 efficiently transferred sulfate to extended core 1 O-glycans to form the MECA-79 epitope. These results collectively indicate that GlcNAc6ST-2 alone is not sufficient for full formation of 6-sulfo sialyl Lewis X despite its robust expression in HEVs of peripheral lymph nodes. The difference in the substrate specificities and possibly different subcellular localizations of GlcNAc6ST-1 and GlcNAc6ST-2 may contribute to the cooperative actions of these two sulfotransferases in the biosynthesis of L-selectin ligands in HEVs.
The homing of lymphocytes to peripheral lymph nodes in double-knockout mice was almost completely blocked by MEL-14. Although we detected binding of L-sel–IgM only on the lining of HEVs on the surface away from the lumen in double-knockout mice, it is conceivable that L-selectin ligands are weakly expressed on the lumenal surface of HEVs but below the detection level of immunofluorescence. The most likely candidate for supporting residual lymphocyte trafficking to peripheral lymph nodes in double-knockout mice is the unsulfated sialyl Lewis X structure, as it is abundantly expressed in the double-knockout mice and interacts with L-selectin in vitro33. In agreement with that hypothesis, homing of lymphocytes to peripheral lymph nodes is abolished in mice doubly deficient in both FucT-IV and FucT-VII (ref. 8). Similarly, our immunofluorescence study showed that L-sel–IgM did not bind to lymph node tissue sections of mice doubly deficient in both FucT-IV and FucT-VII. Moreover, preinjection of the fucose- and sialic acid–specific lectins AAA and MAA, respectively, inhibited homing of lymphocytes to peripheral lymph nodes and mesenteric lymph nodes in the double-knockout mice, indicating that both fucose and sialic acid are essential for residual lymphocyte homing in these mice.
Unsulfated sialyl Lewis X structures on extended core 1 as well as core 2–branched O-glycans efficiently support L-selectin-mediated lymphocyte rolling when the structures are present on P-selectin glycoprotein ligand 1, which contains sulfated tyrosine residues34. It has been reported that the P-selectin glycoprotein ligand 1–like molecule endoglycan, which is modified with the sialyl Lewis X structure and sulfated tyrosine residues, interacts efficiently with L-selectin35. It is possible, therefore, that the unsulfated sialyl Lewis X structure expressed in double-knockout mice supports lymphocyte rolling only when it is present on specific glycoproteins, such as endoglycan. However, our results have shown that GlyCAM-1 from double-knockout mice, which lacks tyrosine sulfation, weakly supported lymphocyte rolling, suggesting that abundant expression of unsulfated sialyl Lewis X in double-knockout mice may account at least in part for the residual lymphocyte homing. Indeed, GlyCAM-1 from double-knockout mice supported shear stress–dependent lymphocyte rolling, as expected given the L-selectin-dependent lymphocyte rolling36. The number of rolling lymphocytes on GlyCAM-1 from double-knockout mice decreased more than the number of lymphocytes with transient tethers. These results suggest that the rolling velocity is more affected than the initial tethering in double-knockout mice, consistent with the finding that 6-sulfation of sialyl Lewis X substantially decreases the rolling velocity of lymphocytes13,
17,
37. Intravital microscopy has shown the rolling velocity of lymphocytes increases considerably in the double-knockout mice, which results in a substantial reduction in the number of cells sticking on HEVs38. That same analysis has also shown that the rolling of B lymphocytes is diminished more substantially in HEVs than in medullary venules, consistent with the finding that rolling in HEVs is more dependent on sulfation than is rolling in medullary venules, as assessed by MECA-79 antigen expression39. Along with our carbohydrate structural analysis, those studies suggest that sialyl Lewis X can function as a substitute L-selectin ligand in the absence of 6-sulfo sialyl Lewis X, although the rolling velocity supported by the former is much faster than that supported by the latter.
In addition to the abundant unsulfated sialyl Lewis X in double-knockout mice, we have identified a previously unknown structure containing unsulfated sialyl Lewis X in the core 2 branch and galactose-6-O-sulfate in the core 1 branch; this carbohydrate structure represented 5.8% of the total carbohydrates on GlyCAM-1 derived from double-knockout mice. It is possible that this structure is also partially responsible for residual L-selectin ligand activity in these mice, as transfection of cDNA encoding KSST enhances L-selectin ligand activity in vitro14,
37. However, this structure should have minimal involvement in lymphocyte homing, because the substantial increase in its abundance in double-knockout mice was not associated with an increase in lymphocyte homing activity.
Although it has been reported that the 6'-sulfo sialyl Lewis X capping structure (sialyl-(2-3)-[sulfo-6]-galactopyranosyl-(1-4)-[fucopyranosyl-(1-3)]-N-acetylglucosaminyl-1-R) is expressed in HEVs40, we did not detect this carbohydrate structure in our analysis, possibly because this structure is not efficiently synthesized because of acceptor competition between FucT-VII and KSST. As described before, FucT-VII transfers few if any fucose residues to sialyl lactosamine modified with galactose-6-O-sulfate7. Conversely, KSST does not transfer sulfate to the galactose residue of the sialyl Lewis X structure41. Combined with our carbohydrate analysis, those findings suggest that KSST overexpression may negatively regulate L-selectin ligand activity by competing for acceptor substrates with FucT-VII, as fucosylation by FucT-VII is critically important for the biosynthesis of L-selectin ligands7. Preliminary results have indicated that KSST overexpression indeed inhibits sialyl Lewis X expression in FucT-VII transfectants (N.H., H.K. and M.F., unpublished observations). Thus, KSST may function both positively14,
37 and negatively in the biosynthesis of L-selectin ligand carbohydrates, depending on its abundance.
To determine the functions of GlcNAc6ST-1 and GlcNAc6ST-2 in immune responses in pathophysiological conditions, we assessed contact hypersensitivity responses in double-knockout mice and found that these mice had substantially less in ear swelling and leukocyte infiltration than did wild-type mice. We also found that the total number of lymphocytes was greatly reduced in the draining lymph nodes after sensitization with DNFB, although antigen-specific lymphocytes were present in the draining lymph nodes at the same ratio found in wild-type mice. Thus, it is likely that the quantitative reduction in antigen-specific lymphocytes resulted in the overall reduction in the contact hypersensitivity responses in these mice. This reduction occurred only in the double-knockout mice, indicating a critical threshold for the amount of L-selectin ligands to attain normal contact hypersensitivity responses. Our results are reminiscent of the phenotypes of L-selectin-deficient mice42 and mice doubly deficient in both FucT-IV and FucT-VII (ref. 43), in which contact hypersensitivity responses are impaired due in part to a lack of antigen-specific lymphocytes in the draining lymph nodes. In mice deficient in L-selectin or fucosyltransferase, however, not only homing of naive lymphocytes to the regional lymph nodes but also effector cell trafficking to the inflammatory sites is impaired, because the interaction between L-selectin and sialyl Lewis X–modified P-selectin glycoprotein ligand 1 is completely abolished. In contrast, only the former was impaired in our sulfotransferase double-knockout mice. Thus, we have shown that homing of naive lymphocytes to the regional lymph nodes is actually a key determinant of the contact hypersensitivity response.
HEV-like vessels reactive with MECA-79 are detected in several disorders associated with chronic inflammation44. MECA-79-reactive HEV-like vessels are detected in human gastric mucosa infected by Helicobacter pylori45. However, the physiological function of the MECA-79 epitope expressed in these specialized blood vessels is unclear, except for the demonstration that MECA-79 has a therapeutic effect in a sheep model of asthma46. As double-knockout mice completely lack the MECA-79 epitope, these mice will be useful for assessing the functions of this carbohydrate epitope at sites of chronic inflammation, such as the pancreas and salivary glands of nonobese diabetic mice47,
48, the hyperplastic thymus of AKR mice49, the inflamed joints of rheumatoid arthritis50 and the mouse model for inflammation induced by H. pylori (felis)51.
In conclusion, our findings have demonstrated the essential function of GlcNAc6ST-1 and GlcNAc6ST-2 in the biosynthesis of the 6-sulfo sialyl Lewis X structure in HEVs, which serves as a dominant ligand for the lymphocyte-homing receptor L-selectin. Our studies provide a link between carbohydrate structural changes and alterations in trafficking of lymphocytes to secondary lymphoid organs in normal and pathophysiological conditions. As many studies have linked carbohydrates to immune cell recruitment, studies linking carbohydrate structure and function should become increasingly important in this field.
Methods
Mice.
Double-knockout mice were generated by breeding of mice singly deficient in GlcNAc6ST-1 (ref. 23) and GlcNAc6ST-2 (ref. 17). Genomic DNA was isolated from mouse tails and was used for PCR genotyping. For the simultaneous detection of wild-type and mutant alleles encoding GlcNAc6ST-1, primers 5'-TCTATGAGCCTGTGTGGCACGT-3' (G6W5), 5'-GCATACCACCTTGTTAGTGGC-3' (G6W3) and 5'-TGACAACGTCGAGCACAGCTG-3' (NeoP5) were used. For detection of the alleles encoding GlcNAc6ST-2, primers F2W, R2W and R1M were used as described17. Mice doubly deficient in both FucT-IV and FucT-VII, generated as described8, were provided by the Functional Glycomics Consortium (La Jolla, California). Mice were treated in accordance with guidelines of the National Institutes of Health and the United States Department of Agriculture and experiments were approved by the Animal Research Committee of the Burnham Institute (La Jolla, California).
Lymphocyte homing assay.
This was done as described17. Lymphocytes from spleens and mesenteric lymph nodes of wild-type mice were labeled with chloromethyl fluorescence diacetate (CMFDA) and 2.5 107 cells were injected into the tail veins of 7- to 8-week-old wild-type and mutant mice. After 1 h, mice were killed and peripheral lymph nodes, mesenteric lymph nodes and Peyer's patches were collected. Lymphocyte suspensions were prepared from these organs and were analyzed by flow cytometry to determine the fractional content of fluorescent cells.
Lymphocyte rolling assay.
GlyCAM-1 was purified from serum obtained from mice of each strain52. The concentration of GlyCAM-1 in each preparation was almost equivalent, as assessed by immunoblot with antibody to GlyCAM-1 prepared in a similar way as described52. Purified GlyCAM-1 (10 l; equivalent to the amount of GlyCAM-1 purified from 10 l of serum) was applied twice (6 h and overnight) or three times (6 h, overnight, and 6 h) in parallel to polystyrene plates coated with antibody to GlyCAM-1 at concentrations of 20 or 50 g/ml, respectively. Lymphocytes were added to the flow chamber and were analyzed as described8. Based on velocity, rolling cells were classified into two groups: rolling lymphocytes that rolled at least two cell diameters for more than 0.5 s below critical velocity, and transiently tethered lymphocytes that had a high rolling velocity near that of free-floating cells with multiple pauses in less than 0.5 s.
Contact hypersensitivity.
Contact hypersensitivity responses were measured as described25. The dorsal skin of mice was shaved and 25 l of 0.5% DNFB (Sigma) in acetone/olive oil (4:1, volume/volume) was applied on days 0 and 1. On day 5, the right ear was treated with 20 l of 0.2% DNFB (10 l on the side of the pinna) and the left ear was treated with vehicle. Swelling was measured with a thickness gauge before and 24 h after treatment.
Lymphocyte proliferation assay.
Single-cell suspensions were prepared from inguinal lymph nodes of mice sensitized with DNFB on days 0 and 1 as described above and were cultured for 40 h in the presence or absence of 2 g/ml of concanavalin A (Sigma) or 200 g/ml of DNBS25 (MP Biomedicals). Cell proliferation was measured with a CellTiter 96AQueous One Solution Proliferation Assay kit (Promega). The absorbance at 490 nm obtained in the absence of concanavalin A or DNBS was subtracted from that obtained in the presence of these agents.
Structural analysis of O-glycans attached to GlyCAM-1.
Peripheral lymph node and mesenteric lymph node slices were metabolically labeled with 0.5 mCi/ml of [3H]galactose (American Radiolabeled Chemicals) as described15,
17. GlyCAM-1 was purified from the conditioned medium and its O-glycans were analyzed by Bio-Gel P-4 gel filtration (Bio-Rad), QAE-Sephadex A-25 anion-exchange column (Sigma) and HPLC before and after exoglycosidase treatment, as described13,
15,
17, except that the elution condition of the Asahipak NH2-bonded HPLC column was modified. Solvent A (64% acetonitrile and 36% H2O), solvent B (25 mM NaH2PO4 in solvent A) and solvent C (50 mM NaH2PO4 in solvent A) were used for HPLC as follows. Monosulfated tetrasaccharide core O-glycans were eluted with a linear gradient from 0% to 30% solvent B in solvent A for 10 min; from 30% to 65% for 40 min; and from 65% to 100% for 5 min, followed by 100% solvent B for 5 min. Monosulfated hexasaccharide core O-glycans were eluted with a linear gradient from 0% to 40% solvent B in solvent A for 10 min; from 40% to 80% for 40 min; and from 80% to 100% for 5 min, followed by 100% solvent B for 5 min. Disulfated O-glycans were eluted with a linear gradient from 0% to 40% solvent C in solvent A for 10 min; from 40% to 80% for 40 min; and from 80% to 100% for 5 min, followed by 100% solvent C for 15 min.
RT-PCR.
HEV cells were purified from peripheral lymph nodes of C57BL/6 mice by immunomagnetic selection with MECA-79 (ref. 53). Mouse corneas were excised surgically from C57BL/6 mice. Total RNA was purified from these preparations and was used for RT-PCR as described13. Primers used to detect genes encoding mouse GlcNAc6ST-1 and mouse GlcNAc6ST-2 have been described13. Primers used to detect genes encoding other molecules were as follows: 5'-TGCTGGTACTGTCCTCGTGG-3' and 5'-TGATGTTGCCACGAGCGAAGG-3' for mouse GlcNAc6ST-3; 5'-TCAACCTAAAGGTGGTGCAACT-3' and 5'-GGTTAAGAAGAAATCAGCGCGT-3' for mouse GlcNAc6ST-4; 5'-AAGCCCTACAACCTGGATGTG-3' and 5'-GAGTTGCGCACTGTGCTGTAT-3' for mouse KSST; 5'-CGGAATTCCCACCATGAAATTCTTCAC-3' and 5'-CGGGATCCAGTTCTTCCTCCACTGTC-3' for mouse GlyCAM-1; 5'-CTCTGCTACACAGCCTCTTGC-3' and 5'-AGGCTCACACTGGACCACTTG-3' for mouse L-selectin; and 5'-CCTGGCCAAGGTCATCCATGACA-3' and 5'-ATGAGGTCCACCACCCTGTTGCT-3' for mouse glyceraldehyde-3-phosphate dehydrogenase.
Transient transfection and metabolic cell labeling.
Lec2 cells54 were transiently transfected with an expression plasmid encoding GlyCAM-1–IgG13 in combination with those encoding core 2–N-acetylglucosaminyltransferase-I (Core2GlcNAcT-I) or core 1–1,3-N-acetylglucosaminyltransferase (Core1-3GlcNAcT) and one of the four GlcNAc6STs (GlcNAc6ST-1 through GlcNAc6ST-4). Cells were metabolically labeled with [35S]Na2SO4 (100 Ci/ml) as described13,
15. As a control, vectors without inserts were transfected into Lec2 cells. In Lec2 cells, MECA-79 antigen is efficiently synthesized, as Lec2 cells lack Golgi-sialylation54 and thus core 1 extension does not compete with 2,3-sialylation of core 1 structure (galactopyranosyl-(1-3)-N-acetylgalactosaminyl-1-serine (or threonine)).
Statistical analysis.
Student's t-test was used for statistical analysis.
Accession code.
BIND (http://bind.ca): 319185.
Note: Supplementary information is available on the Nature Immunology website.
Received 19 June 2005; Accepted 25 August 2005; Published online: 9 October 2005.
REFERENCES
-
Arbonés, M.L. et al. Lymphocyte homing and leukocyte rolling and migration are impaired in L-selectin-deficient mice. Immunity 1, 247–260 (1994). | Article | PubMed | ISI | ChemPort |
-
Springer, T.A. Traffic signals for lymphocyte recirculation and leukocyte emigration: the multistep paradigm. Cell 76, 301–314 (1994). | Article | PubMed | ISI | ChemPort |
-
Butcher, E.C. & Picker, L.J. Lymphocyte homing and homeostasis. Science 272, 60–66 (1996). | PubMed | ISI | ChemPort |
-
von Andrian, U.H. & Mempel, T.R. Homing and cellular traffic in lymph nodes. Nat. Rev. Immunol. 3, 867–878 (2003). | Article | PubMed | ISI | ChemPort |
-
Ley, K. & Kansas, G.S. Selectins in T-cell recruitment to non-lymphoid tissues and sites of inflammation. Nat. Rev. Immunol. 4, 325–335 (2004). | Article | PubMed | ISI | ChemPort |
-
Rosen, S.D. Ligands for L-selectin: homing, inflammation, and beyond. Annu. Rev. Immunol. 22, 129–156 (2004). | Article | PubMed | ISI | ChemPort |
-
Maly, P. et al. The (1,3)fucosyltransferase Fuc-TVII controls leukocyte trafficking through an essential role in L-, E-, and P-selectin ligand biosynthesis. Cell 86, 643–653 (1996). | Article | PubMed | ISI | ChemPort |
-
Homeister, J.W. et al. The (1,3) fucosyltransferases FucT-IV and FucT-VII exert collaborative control over selectin-dependent leukocyte recruitment and lymphocyte homing. Immunity 15, 115–126 (2001). | Article | PubMed | ISI | ChemPort |
-
Rosen, S.D., Singer, M.S. & Yednock, T.A. Involvement of sialic acid on endothelial cells in organ-specific lymphocyte recirculation. Science 228, 1005–1007 (1985). | PubMed | ISI | ChemPort |
-
Imai, Y., Lasky, L.A. & Rosen, S.D. Sulphation requirement for GlyCAM-1, an endothelial ligand for L-selectin. Nature 361, 555–557 (1993). | Article | PubMed | ISI | ChemPort |
-
Hemmerich, S., Butcher, E.C. & Rosen, S.D. Sulfation-dependent recognition of high endothelial venules (HEV)-ligands by L-selectin and MECA-79, an adhesion-blocking monoclonal antibody. J. Exp. Med. 180, 2219–2226 (1994). | Article | PubMed | ISI | ChemPort |
-
Mitsuoka, C. et al. Identification of a major carbohydrate capping group of the L-selectin ligand on high endothelial venules in human lymph nodes as 6-sulfo sialyl Lewis X. J. Biol. Chem. 273, 11225–11233 (1998). | Article | PubMed | ISI | ChemPort |
-
Hiraoka, N. et al. A novel, high endothelial venule-specific sulfotransferase expresses 6-sulfo sialyl Lewisx, an L-selectin ligand displayed by CD34. Immunity 11, 79–89 (1999). | Article | PubMed | ISI | ChemPort |
-
Bistrup, A. et al. Sulfotransferases of two specificities function in the reconstitution of high endothelial cell ligands for L-selectin. J. Cell Biol. 145, 899–910 (1999). | Article | PubMed | ISI | ChemPort |
-
Yeh, J.-C. et al. Novel sulfated lymphocyte homing receptors and their control by a core1 extension 1,3-N-acetylglucosaminyltransferase. Cell 105, 957–969 (2001). | Article | PubMed | ISI | ChemPort |
-
Streeter, P.R., Rouse, B.T.N. & Butcher, E.C. Immunohistologic and functional characterization of a vascular addressin involved in lymphocyte homing into peripheral lymph nodes. J. Cell Biol. 107, 1853–1862 (1988). | Article | PubMed | ISI | ChemPort |
-
Hiraoka, N. et al. Core 2 branching 1,6-N-acetylglucosaminyltransferase and high endothelial venule-restricted sulfotransferase collaboratively control lymphocyte homing. J. Biol. Chem. 279, 3058–3067 (2004). | Article | PubMed | ISI | ChemPort |
-
Hemmerich, S. et al. Sulfation of L-selectin ligands by an HEV-restricted sulfotransferase regulates lymphocyte homing to lymph nodes. Immunity 15, 237–247 (2001). | Article | PubMed | ISI | ChemPort |
-
Fukuda, M., Hiraoka, N., Akama, T.O. & Fukuda, M.N. Carbohydrate-modifying sulfotransferases: structure, function and pathophysiology. J. Biol. Chem. 276, 47747–47750 (2001). | PubMed | ISI | ChemPort |
-
Hemmerich, S. et al. Chromosomal localization and genomic organization for the galactose/N-acetylgalactosamine/N-acetylglucosamine 6-O-sulfotransferase gene family. Glycobiology 11, 75–87 (2001). | Article | PubMed | ISI | ChemPort |
-
Uchimura, K. et al. Molecular cloning and characterization of an N-acetylglucosamine-6-O-sulfotransferase. J. Biol. Chem. 273, 22577–22583 (1998). | Article | PubMed | ISI | ChemPort |
-
Kimura, N. et al. Reconstitution of functional L-selectin ligands on a cultured human endothelial cell line by cotransfection of 1–3 fucosyltransferase VII and newly cloned GlcNAc:6-sulfotransferase cDNA. Proc. Natl. Acad. Sci. USA 96, 4530–4535 (1999). | Article | PubMed | ChemPort |
-
Uchimura, K. et al. N-Acetylglucosamine 6-O-sulfotransferase-1 regulates expression of L-selectin ligands and lymphocyte homing. J. Biol. Chem. 279, 35001–35008 (2004). | Article | PubMed | ISI | ChemPort |
-
Berlin, C. et al. 47 integrin mediates lymphocyte binding to the mucosal vascular addressin MAdCAM-1. Cell 74, 185–195 (1993). | Article | PubMed | ISI | ChemPort |
-
Phanuphak, P., Moorhead, J.W. & Claman, H.N. Tolerance and contact sensitivity to DNFB in mice. I. In vivo detection by ear swelling and correlation with in vitro cell stimulation. J. Immunol. 112, 115–123 (1974). | PubMed | ISI | ChemPort |
-
Uchimura, K. et al. Diversity of N-acetylglucosamine-6-O-sulfotransferases: molecular cloning of a novel enzyme with different distribution and specificities. Biochem. Biophys. Res. Commun. 274, 291–296 (2000). | Article | PubMed | ISI | ChemPort |
-
Bhakta, S. et al. Sulfation of N-acetylglucosamine by chondroitin 6-sulfotransferase 2 (GST-5). J. Biol. Chem. 275, 40226–40234 (2000). | Article | PubMed | ISI | ChemPort |
-
Lee, J.K., Bhakta, S., Rosen, S.D. & Hemmerich, S. Cloning and characterization of a mammalian N-acetylglucosamine-6-sulfotransferase that is highly restricted to intestinal tissue. Biochem. Biophys. Res. Commun. 263, 543–549 (1999). | Article | PubMed | ISI | ChemPort |
-
Akama, T.O. et al. Human corneal GlcNAc 6-O-sulfotransferase and mouse intestinal GlcNAc 6-O-sulfotransferase both produce keratan sulfate. J. Biol. Chem. 276, 16271–16278 (2001). | Article | PubMed | ISI | ChemPort |
-
Fukuta, M. et al. Molecular cloning and characterization of human keratan sulfate Gal-6-sulfotransferase. J. Biol. Chem. 272, 32321–32328 (1997). | Article | PubMed | ISI | ChemPort |
-
de Graffenried, C.L. & Bertozzi, C.R. Golgi localization of carbohydrate sulfotransferases is a determinant of L-selectin ligand biosynthesis. J. Biol. Chem. 278, 40282–40295 (2003). | Article | PubMed | ISI | ChemPort |
-
Zerfaoui, M. et al. The cytosolic and transmembrane domains of the 1,6 N-acetylglucosaminyltransferase (C2GnT) function as a cis to medial/Golgi-targeting determinant. Glycobiology 12, 15–24 (2002). | Article | PubMed | ISI | ChemPort |
-
Foxall, C. et al. The three members of the selectin receptor family recognize a common carbohydrate epitope, the sialyl Lewisx oligosaccharide. J. Cell Biol. 117, 895–902 (1992). | Article | PubMed | ISI | ChemPort |
-
Mitoma, J. et al. Extended core 1 and core 2 branched O-glycans differentially modulate sialyl Lewis x-type L-selectin ligand activity. J. Biol. Chem. 278, 9953–9961 (2003). | Article | PubMed | ISI | ChemPort |
-
Fieger, C.B., Sassetti, C.M. & Rosen, S.D. Endoglycan, a member of the CD34 family, functions as an L-selectin ligand through modification with tyrosine sulfation and sialyl Lewis x. J. Biol. Chem. 278, 27390–27398 (2003). | Article | PubMed | ISI | ChemPort |
-
Finger, E.B. et al. Adhesion through L-selectin requires a threshold hydrodynamic shear. Nature 379, 266–269 (1996). | Article | PubMed | ISI | ChemPort |
-
Tangemann, K., Bistrup, A., Hemmerich, S. & Rosen, S.D. Sulfation of a high endothelial venule-expressed ligand for L-selectin: effects on tethering and rolling of lymphocytes. J. Exp. Med. 190, 935–941 (1999). | Article | PubMed | ISI | ChemPort |
-
Uchimura, K. et al. A major class of L-selectin ligands is eliminated in mice deficient in two sulfotransferases expressed in high endothelial venules. Nat. Immunol. advance online publication 9 October 2005(1038/ni1258). | Article |
-
M'Rini, C. et al. A Novel endothelial L-selectin ligand activity in lymph node medulla that is regulated by (1,3)-fucosyltransferase-IV. J. Exp. Med. 198, 1301–1312 (2003). | Article | PubMed | ChemPort |
-
Hemmerich, S., Leffler, H. & Rosen, S.D. Structure of the O-glycans in GlyCAM-1, an endothelial-derived ligand for L-selectin. J. Biol. Chem. 270, 12035–12047 (1995). | Article | PubMed | ISI | ChemPort |
-
Torii, T., Fukuta, M. & Habuchi, O. Sulfation of sialyl N-acetyllactosamine oligosaccharides and fetuin oligosaccharides by keratan sulfate Gal-6-sulfotransferase. Glycobiology 10, 203–211 (2000). | Article | PubMed | ISI | ChemPort |
-
Catalina, M.D. et al. The route of antigen entry determines the requirement for L-selectin during immune responses. J. Exp. Med. 184, 2341–2351 (1996). | Article | PubMed | ISI | ChemPort |
-
Smithson, G. et al. Fuc-TVII is required for T helper 1 and T cytotoxic 1 lymphocyte selectin ligand expression and recruitment in inflammation, and together with Fuc-TIV regulates naive T cell trafficking to lymph nodes. J. Exp. Med. 194, 601–614 (2001). | Article | PubMed | ISI | ChemPort |
-
Rosen, S.D. Endothelial ligands for L-selectin. From lymphocyte recirculation to allograft rejection. Am. J. Pathol. 155, 1013–1020 (1999). | PubMed | ISI | ChemPort |
-
Kobayashi, M. et al. Induction of peripheral lymph node addressin in human gastric mucosa infected by Helicobacter pylori. Proc. Natl. Acad. Sci. USA 101, 17807–17812 (2004). | Article | PubMed | ChemPort |
-
Rosen, S.D., Tsay, D., Singer, M.S., Hemmerich, S. & Abraham, W.M. Therapeutic targeting of endothelial ligands for L-selectin (PNAd) in a sheep model of asthma. Am. J. Pathol. 166, 935–944 (2005). | PubMed | ISI | ChemPort |
-
Hanninen, A. et al. Vascular addressins are induced on islet vessels during insulitis in nonobese diabetic mice and are involved in lymphoid cell binding to islet endothelium. J. Clin. Invest. 92, 2509–2515 (1993). | PubMed | ISI | ChemPort |
-
Faveeuw, C., Gagnerault, M-C. & Lepault, F. Expression of homing and adhesion molecules in infiltrated islets of Langerhans and salivary glands of nonobese diabetic mice. J. Immunol. 152, 5969–5978 (1994). | PubMed | ISI | ChemPort |
-
Michie, S.A., Streeter, P.R., Butcher, E.C. & Rouse, R.V. L-selectin and 47 integrin homing receptor pathways mediate peripheral lymphocyte traffic to AKR mouse hyperplastic thymus. Am. J. Pathol. 147, 412–421 (1995). | PubMed | ISI | ChemPort |
-
Michie, S.A., Streeter, P.R., Bolt, P.A., Butcher, E.C. & : Picker, L.J. The human peripheral lymph node vascular addressin. An inducible endothelial antigen involved in lymphocyte homing. Am. J. Pathol. 143, 1688–1698 (1993). | PubMed | ISI | ChemPort |
-
Fox, J.G. et al. Hypertrophic gastropathy in Helicobacter felis-infected wild-type C57BL/6 mice and p53 hemizygous transgenic mice. Gastroenterology 110, 155–166 (1996). | Article | PubMed | ISI | ChemPort |
-
Singer, M.S. & Rosen, S.D. Purification and quantification of L-selectin-reactive GlyCAM-1 from mouse serum. J. Immunol. Methods 196, 153–161 (1996). | Article | PubMed | ISI | ChemPort |
-
Girard, J.P. & Springer, T.A. Cloning from purified high endothelial venule cells of hevin, a close relative of the antiadhesive extracellular matrix protein SPARC. Immunity 2, 113–123 (1995). | Article | PubMed | ISI | ChemPort |
-
Deutscher, S.L., Nuwayhid, N., Stanley, P., Briles, E.I. & Hirschberg, C.B. Translocation across Golgi vesicle membranes: a CHO glycosylation mutant deficient in CMP-sialic acid transport. Cell 39, 295–299 (1984). | Article | PubMed | ISI | ChemPort |
Acknowledgments
We thank T. Akama, S. Chen and E. Lammar for critical reading of the manuscript, and A. Morse for organizing the manuscript. Supported by the National Institutes of Health (P01CA71932 to M.F. and J.B.L.; U54 GM62116 to the Functional Glycomics Consortium) and the Uehara Memorial Foundation, Japan (H.K.).
Competing interests statement:
The authors declare that they have no competing financial interests. |
