Memory B lymphocytes from secondary lymphoid organs interact with E-selectin through a novel glycoprotein ligand

Recirculation of B lymphocytes through the secondary lymphoid organs is key for recognition and response to foreign antigen. B lymphocytes within secondary lymphoid organs comprise a heterogeneous population of cells at distinct differentiation stages. To ascribe a particular adhesive behavior to discrete B-cell subsets within secondary lymphoid organs, we investigated their functional interaction with endothelial selectins under flow. We describe herein the characterization of a subset of human tonsillar B cells that interact with E-selectin but not P-selectin. E-selectin–interacting B cells had a phenotype of non–germinal center (CD10^–, CD38^–, CD44⁺), memory (IgD^–) cells. Furthermore, FucT-VII was expressed selectively in CD44⁺ E-selectin–adherent B lymphocytes. B-cell rolling on E-selectin required sialic acid but was independent of previously described selectin ligands. A novel glycoprotein ligand of 240 kDa carrying N-linked glycans was isolated from B-cell membranes by an E-selectin immunoadhesin. Binding of this protein was strictly Ca²⁺ dependent, was inhibited by a cell adhesion–blocking mAb against E-selectin, and required the presence of sialic acid but not N-linked carbohydrates. Our results enable us to assign to resident memory B lymphocytes a novel adhesion function, the rolling on E-selectin, that provides insights on the adhesion pathways involved in homing of memory B cells to tertiary sites.

B lymphocytes continuously recirculate through the blood and secondary lymphoid organs, such as peripheral lymph nodes, spleen, Peyer’s patches, or tonsils, thus allowing them to encounter and react to antigen localized at these sites. Naive B cells enter the tonsils and will either continue the recirculation surveillance process in case they do not encounter antigen, or they will specifically recognize and respond to antigen located in the tonsils. B cells that encounter antigen become activated and undergo clonal expansion and somatic hypermutation that leads to differentiation into memory and plasma cells (1, 2). These cells return to the circulation, and some are directed to tertiary sites where they will exert their effector functions (3). B cells within secondary lymphoid organs can thus be classified, depending on their differentiation stage, into discrete subpopulations that are located at different compartments within these organs. B-cell subpopulations, including naive, germinal center (GC), memory, or plasma cells, can be phenotypically characterized by their differential expression of several cell surface markers (1).

The selectin family of adhesion molecules has been classically implicated in the initial steps of the interaction of lymphocytes with the endothelium during lymphocyte homing. L-selectin, expressed by most circulating lymphocytes, is the principal receptor implicated in the interaction of lymphocytes with high endothelial venules (HEV), which is the first step for their entry into secondary lymphoid organs. This molecule also mediates lymphocyte recruitment to inflammatory sites (4). P-selectin, expressed by platelets and endothelial cells shortly after activation with inflammatory mediators, has been implicated in early leukocyte rolling during the inflammatory response (4). E-selectin is also an inducible endothelial adhesion molecule, although its expression follows slower kinetics, reaching its maximum level after 4–6 hours of stimulation with inflammatory mediators. This molecule is involved in leukocyte rolling during acute and chronic inflammation (5, 6). Moreover, E-selectin has been found in endothelium undergoing proliferation and tube formation, thus implicating this molecule in angiogenesis (7, 8). Thus, the role of E-selectin is not restricted to mediating leukocyte-endothelial interactions during inflammation. Accordingly, constitutive expression of E-selectin has been reported at low levels in venules of tonsils, adenoids, skin, and hematopoietic tissues (9–13). Furthermore, expression of E-selectin has been reported on the surface of astrocytes and hepatocytes (14, 15).

Carbohydrate structures, such as Le^x, sLe^x, or Le^a, decorate glycoproteins that have been identified as the high-affinity selectin ligands. Expression of functional selectin ligands requires the action of α(1-3)-fucosyltransferases (FucTs). Several of these enzymes have been described, although only FucT-IV and -VII have been involved in the synthesis of E-selectin ligands in leukocytes (4). Mucin-like glycoprotein ligands of selectins were identified by affinity isolation using antibody-like selectin fusion proteins. P-selectin glycoprotein ligand-1 (PSGL-1) was first identified as a P-selectin ligand (16, 17), although it has also been reported to bind E-selectin (16, 18) and L-selectin (19–21). This molecule mediates rolling on P-selectin (19–22), but it is unable to mediate rolling on E-selectin on certain cell types (22, 23). Moreover, L-selectin has recently been proposed as a high-affinity ligand for E-selectin (24, 25). Other E-selectin–binding proteins have been identified, including E-selectin ligand-1 (ESL-1), which was described in myeloid cells as a 150-kDa glycoprotein binding via N-linked oligosaccharides to E-selectin (26, 27). A 250-kDa glycoprotein on bovine γ/δ T cells was described as a new E-selectin ligand (24, 28).

Several studies have addressed the presence of selectin ligands on B cells. We reported previously that in vitro–activated tonsillar B cells interact with E- and P-selectin under static conditions (29). More recently, Yago et al. (30) showed that peripheral blood (PB) B cells bind well to P-selectin, but not to E-selectin, under flow conditions. To our knowledge, the present work provides the first evidence that a subset of tonsillar B cells interacts with E-selectin under defined laminar flow conditions. This functional approach has allowed us to ascribe a novel adhesive behavior to the subset of resident memory B cells from secondary lymphoid organs. Furthermore, the B-cell rolling on E-selectin was not mediated by any of the previously identified E-selectin ligands. Using a recombinant antibody-like form of E-selectin, we have found a novel functional ligand of 240 kDa for E-selectin on tonsillar B cells.

E-selectin–mediated rolling of tonsillar B lymphocytes on endothelium. To investigate whether different B-cell subpopulations in secondary lymphoid organs can be distinguished based on their differential selectin-mediated cell interactions, we performed adhesion assays of purified nonactivated tonsillar B cells to HUVECs under flow conditions. Very few adhesive contacts took place between B cells and the control unactivated HUVEC monolayer. In contrast, many resting tonsillar B cells were able to contact, roll, and subsequently arrest on 4-hour TNF-α–activated HUVEC monolayers. Examination of B cells in contact with the endothelium after 5 minutes of perfusion revealed that 23% of the B cells were rolling on the surface of the monolayer and that the remaining cells had arrested after the rolling period (Figure 1a). The number of rolling B cells was highest at a shear rate of 1 dyn/cm² and decreased when the shear was increased to 1.5 dyn/cm² or higher (Figure 1b). The B-cell rolling velocities were 40 ± 11 at 1.8 dyn/cm² and decreased to 10 ± 4 as the shear stress level was lowered to 1 dyn/cm² (Figure 1c). Based on these data, the flow rate of 1 dyn/cm² was used for all subsequent flow adhesion experiments.

Figure 1

Adhesion of B lymphocytes to HUVEC monolayers under defined laminar flow conditions. Confluent HUVEC monolayers grown on glass coverslips were activated or not (control) with TNF-α for 4 hours, and then the coverslips were inserted in a parallel-plate flow chamber. B lymphocytes were perfused through the chamber at an estimated wall shear stress of 1 dyn/cm² (0.54 mL/min) (a), or B cells were perfused across 4-hour TNF-α–activated HUVEC monolayers at 1, 1.5, or 1.8 dyn/cm² (b and c). The number of rolling and adherent cells per field of view was determined by analysis of videotapes as described in Methods. Values represent the average number of interacting cells (a), arithmetic mean ± SEM number of rolling cells (b), or rolling velocity (c) calculated from 3 independent experiments after 5 minutes of perfusion at each level of flow.

To identify the molecules on the endothelial surface involved in these adhesive interactions, endothelial cells were treated with blocking mAb’s directed to different adhesion molecules. An anti–E-selectin blocking mAb virtually abrogated these interactions, showing that E-selectin is mediating initial attachment and rolling of B cells on the endothelial monolayer. Neither P-selectin nor vascular cell adhesion molecule-1 (VCAM-1) was implicated, as blocking mAb’s against these molecules did not affect the rolling (Figure 2).

Figure 2

B-lymphocyte rolling on activated endothelium is mediated by E-selectin. Four-hour TNF–α-activated HUVEC monolayers were grown on glass coverslips and treated for 20 minutes at room temperature with 10 μg/mL of mAb directed to HLA class I (W6/32), E-selectin (H18/7), P-selectin (HDPG 2/3), or VCAM-1 (4B9); then the coverslip was assembled in a parallel-plate flow chamber. B cells were perfused across the HUVEC monolayers at an estimated shear stress of 1 dyn/cm² for 5 minutes, and the number of rolling cells was determined in 4 separate fields at the different time points of each independent experiment. Values represent arithmetic means ± SEM number of rolling cells calculated from 3 independent experiments.

To further study B-cell rolling on endothelial selectins, flow adhesion experiments were performed using monolayers of CHO cells stably transfected with human P-selectin and E-selectin. B lymphocytes did not attach or roll on parental untransfected CHO cells (CHO-DUKX) (Figure 3). P-selectin transfectants (CHO-P) were also unable to support B-cell rolling (Figure 3), although they consistently supported rolling of neutrophils and HL60 in experiments performed in parallel (not shown). In contrast, B cells readily interacted with E-selectin transfectants (CHO-E) and rolled at low velocities of 17 ± 3 μm/s. The specificity of these adhesive interactions was further demonstrated by treatment of CHO-E monolayers with a blocking mAb against E-selectin, which completely inhibited rolling of B cells (Figure 3). Rolling interactions were abolished using the Ca²⁺ chelator EDTA (data not shown).

Figure 3

E-selectin, but not P-selectin, transfectants support rolling of B lymphocytes. CHO-DUKX, CHO-P, and CHO-E monolayers were grown on glass coverslips. The latter were treated for 20 minutes with 10 μg/mL of mAb directed to HLA class I (W6/32) or E-selectin (H18/7). B cells were then perfused across the monolayers for 5 minutes at 1 dyn/cm². The number of rolling cells was calculated as the mean of 4 different fields at every time point of each independent experiment. Values represent arithmetic mean ± SEM number of rolling cells calculated from 3 independent experiments.

The subset of tonsillar B lymphocytes interacting with E-selectin bears the memory phenotype. To characterize phenotypically the subset of B cells that roll on E-selectin, we performed adhesion assays to the CHO-E monolayer under rotation conditions, which resembles the flow shear, enabling us to recover and study E-selectin–adherent cells. The expression of different B-cell markers on adherent cells, as well as in the unfractionated population of B cells, was analyzed. Staining with the pan–B-cell marker CD19 showed that both the total and adherent populations comprised only B lymphocytes with no contamination of CHO cells. The analysis of CD38, CD44, and CD10 expression reveals that unfractionated cells consist of a heterogeneous population that express different levels of these surface markers (Figure 4a). In contrast, the majority of the cells in the adherent population was negative for the GC markers CD38 and CD10, have low levels of the activation markers CD5 and CD23, and have high levels of CD44 (Figure 4a). Thus, E-selectin–adherent B cells bear the CD38^–, CD44⁺⁺, CD10^– phenotype, described for B cells populating compartments other than the GC.

Figure 4

The E-selectin–binding subpopulation consists of CD19⁺, CD5^–, CD44⁺⁺, CD23^–, CD38^–, CD10^–, and IgD^– memory B lymphocytes. B cells purified from human tonsils were allowed to attach to CHO-E monolayers in C6 wells under rotation conditions (72 rpm) at 37°C; then wells were washed, and adherent cells were collected and analyzed by FACS as described in Methods. (a) Histograms showing expression of CD19, CD5, CD44, CD23, CD38, and CD10 on the total and adherent B-cell populations. The percentage of positive cells (highly positive for CD44, ++) is shown in the upper right corner of each histogram box. (b) Double staining of total and adherent subpopulations was carried out. Staining with mAb directed to CD38, CD44 in FL-2, and IgD in FL-1 is shown.

Resident B cells outside the GC of secondary lymphoid organs comprise naive and memory lymphocytes, which can be distinguished by the expression of cell surface IgD (1). We therefore decided to investigate the expression of IgD in the total and E-selectin–adherent subpopulation. Double-staining analysis showed that adherent CD38^– cells were mainly IgD^– (Figure 4b), thus displaying a phenotype coincident to that observed for secondary lymph node memory B cells. As already described here, the subset of CD38⁺ IgD^– that corresponded to GC cells was absent in the adherent population (Figure 4b). In addition, non-GC CD44⁺⁺ cells in the unfractionated population consisted of 2 discrete cell populations, IgD⁺ and IgD^–, whereas CD44⁺⁺ cells in the adherent population were uniformly IgD^– memory B cells (Figure 4b). Hence, cells capable of interacting with E-selectin were mainly memory CD38^-, CD44⁺⁺, IgD^– B lymphocytes.

Further analyses of the expression of different adhesion molecules on E-selectin–adherent cells was performed. The level of LFA-1 integrin was similar in both the total and the adherent subpopulations. In contrast, adherent cells expressed higher levels of the integrin subunits α4, α5, β1, αx, as well as other adhesion molecules such as L-selectin (Table 1).

Table 1

Expression of adhesion molecules in the total and E-selectin–adherent subpopulations of B cells

B-cell rolling on E-selectin is independent of previously described selectin ligands but requires sialic acid on the lymphocyte surface. Different carbohydrate determinants have been shown to decorate the selectin ligands; thus, expression of functional selectin ligands requires glycosylation that is carried out by different α(1-3)-fucosyltransferases (FucTs) (4). We therefore investigated the expression of FucTs, as well as carbohydrate determinants, in B lymphocytes. We analyzed the presence of different (α1,3)-FucTs involved in the glycosylation of the E-selectin ligands. The expression of mRNA transcripts for FucT-VII, FucT-IV, and C2GnT was measured by semiquantitative RT-PCR in total RNA isolated from B-cell samples and compared with data obtained using HL60 cells as a positive control. B cells were found to endogenously express FucT-VII, FucT-IV, and C2GnT (Figure 5a). To determine whether these enzymes were selectively expressed in the adherent population, B cells were fractionated into CD44⁺ and CD38⁺ cells, and the subpopulations were again analyzed by semiquantitative RT-PCR using 2-fold serial dilutions of input cDNA to assess more accurately the relative differences in mRNA expression. As shown in Figure 5b, the adherent population (CD44⁺) of B cells expressed at least 32-fold more FucT-VII mRNA than GC (CD38⁺) nonadherent cells, whereas FucT-IV and C2GnT levels were represented at easily detectable levels in both B-cell subpopulations. These results strongly suggest the involvement of FucT-VII in the glycosylation of the E-selectin ligand on B cells.

Figure 5

Expression of glycosyltransferase mRNA in human tonsillar B lymphocytes. Expression of FucT-VII, FucT-IV, and C2GnT in purified tonsillar B cells was carried out as described in Methods. pgk1 is a housekeeping enzyme included as an internal control. HL60, a human myeloid cell line, was included as a positive control. (a) The presence or absence of RT in the reaction is indicated by a “+” or “–”, respectively. C corresponds to control HL60 cells, and B corresponds to B-cell samples from the total population. (b) B lymphocytes were fractionated into CD38⁺ and CD44⁺ cells and were analyzed by semiquantitative RT-PCR, using 2-fold serial dilutions going from left to right. The last lane in each set is the negative control of no RT (–RT).

Previous flow cytometry studies showed that tonsillar B lymphocytes failed to express CLA, sLe^x, Le^x, or Le^a. B cells also lack expression of CD57, CDw65 CD35, CD66, and sulfatides (29). We have extended this analysis by determining the expression of other additional carbohydrate determinants on tonsillar B cells. The epitope recognized by mAb 2H5, but not SNH3 and FH6, was detected in these cells (Figure 6). The mAb’s HECA452, CSLEX, and 2F3 recognize sLe^x or variants of this structure that are FucT-VII–dependent epitopes (6, 37, 39, 45) and have been suggested to bind directly to E-selectin. These epitopes were not found to be expressed in resting B lymphocytes (Figure 6). High-affinity E-selectin ligands, such as PSGL-1, ESL-1, or L-selectin, have been described (16–28). The expression of PSGL-1 was analyzed using 2 specific mAb’s, PSL 275 and KPL-1, the latter specifically recognizing the sulfated NH₂-terminal domain of this molecule (21). Neither anti–PSGL-1 mAb recognized B lymphocytes (Figure 6). L-selectin was weakly expressed, although the level of expression was higher in the adherent population (Table 1).

Figure 6

Surface expression of carbohydrate determinants on B lymphocytes. Total B-cell samples readily purified from tonsils were analyzed by FACS as described in Methods. Histograms representing cell surface expression of CD19 (Bu12), different carbohydrate epitopes (SNH3, FH6, 2F3, 2H5, HECA452, and CSLEX), PSGL-1 (KPL-1 and PSL 275), as well as nonbinding control antibody (X63) are shown.

Rolling of B cells on CHO-E monolayers was not affected by treatment with adhesion-blocking antibodies against PSGL-1 or L-selectin (Figure 7). The effect of blocking polyclonal anti–ESL-1 antibody was also explored using affinity-purified polyclonal antibodies. Neither of these antibodies had an inhibitory effect on B-cell rolling (Figure 7), thus indicating that interaction of B cells with E-selectin is not mediated by any of the previously described high-affinity E-selectin ligands.

Figure 7

ESL-1, PSGL-1, and L-selectin do not mediate rolling of B cells on CHO-E. B cells were treated for 20 minutes at 4°C with 10 μg/mL of controls anti–HLA class I (W6/32) and anti-IgG, or with blocking anti–PSGL-1 (KPL1), L-selectin (LAM 1/3), or polyclonal anti–ESL-1 (89060). Cells were then perfused across CHO-E monolayers at 1 dyn/cm², and the number of rolling cells was calculated. The arithmetic mean ± SEM number of rolling cells calculated from 3 independent experiments is shown.

To investigate the biochemical characteristics of the E-selectin ligand(s) expressed by tonsillar B cells, treatment with different enzymes affecting sialylation or glycosylation of the cell surface molecules was carried out. Neuraminidase treatment completely blocked rolling on CHO-E, showing an absolute dependence on sialic acid for this interaction (Figure 8). OSGE, a protease that acts specifically on O-sialomucin–like proteins, did not affect B-cell rolling (Figure 8), whereas it completely blocked rolling of HL60 on CHO-P (data not shown). These results suggest that sialic acid–bearing glycans are responsible for the interaction of B lymphocytes with E-selectin.

Figure 8

Neuraminidase treatment of B cells greatly reduces rolling on CHO-E monolayers. B cells were treated for 1 hour at 37°C with or without 40 μg/mL OSGE or 0.1 U/mL neuraminidase, and were then washed and perfused across CHO-E monolayers. The number of rolling cells was determined as described previously. Values represent arithmetic mean ± SEM number of rolling cells calculated from 3 independent experiments.

E-selectin/IgG specifically binds a single 240-kDa glycoprotein from B lymphocytes. Soluble recombinant selectin/IgG fusion proteins have been used to identify high-affinity selectin ligands such as PSGL-1 or ESL-1 by affinity precipitation assays (16, 17, 26, 27). To characterize the ligand for E-selectin on B lymphocytes, surface biotinylated lysates obtained from the total population of tonsillar B cells were precipitated with the antibody-like fusion protein E-selectin/IgG. This chimeric protein, but not human IgG, precipitated a band of approximately 240 kDa under reducing conditions (Figure 9a). Antibodies against ESL-1 precipitated a band of 150 kDa that was not precipitated by the E-selectin chimera, thus indicating that ESL-1 is present on the surface of B cells but is unable to bind E-selectin. Immunoprecipitation using polyclonal antibodies against PSGL-1 confirmed its absence from the B-cell surface. As a control for the selective surface labeling with biotin, an antibody against Hsp90 was unable to precipitate this intracellular protein from the surface-labeled lysates (Figure 9a), whereas it did recognize Hsp 90 from metabolically intracellular labeled 32dcl3 cells (Figure 9b). Thus, ESL-1 and E-selectin–binding protein of 240 kDa were present on the surface of B cells. The 240-kDa band was completely removed from the E-selectin/IgG matrix by washing the beads in the presence of 3 mM EDTA (Figure 10a). In addition, the fusion protein VE-cadherin/IgG was unable to precipitate the 240-kDa band (Figure 10b). The Ca²⁺-dependent binding of the 240-kDa protein to E-selectin and the failure to bind human IgG or VE-cadherin/IgG suggest that the selectin part in the E-selectin/IgG fusion protein is responsible for binding to the 240-kDa protein. Furthermore, preincubation of the fusion protein with the cell adhesion–blocking antibody UZ4 against mouse E-selectin, but not with a control antibody, partially inhibited the precipitation of the 240-kDa ligand (Figure 10d). Similar levels of inhibition were observed for the precipitation of ESL-1 from 32Dcl3 cells, using the same E-selectin fusion protein (34).

Figure 9

Affinity isolation of E-selectin ligands from human tonsillar B cells. B cells were isolated from human tonsils and surface biotinylated, as described in Methods (a). The 32Dcl3 cells were metabolically labeled with [³⁵S]methionine/[³⁵S]cysteine as described in Methods (b). Detergent extracts of the cells were immunoprecipitated with human IgG as control (human IgG), E-selectin/IgG (E-selectin/IgG), rabbit control IgG (co), affinity-purified rabbit antibodies against ESL-1 (αESL-1) or PSGL-1 (αPSGL-1), an mAb against CD45 (αCD45) or Hsp90 (αHSP90), or a control mAb (co). Immunoprecipitated antigens were isolated by SDS-PAGE under reducing conditions, transferred to nitrocellulose, and detected with peroxidase-conjugated streptavidin and ECL (a). In b, electrophoresed antigens were viewed by fluorography. Molecular mass markers (in kilodaltons) are indicated on the left.

Figure 10

The binding of the 240-kDa ligand to the E-selectin/IgG fusion protein is E-selectin–specific. This binding requires sialic acid but is independent of N-linked carbohydrate side chains. B lymphocytes were surface biotinylated, and detergent extracts were immunoprecipitated with human IgG as control (hIgG), E-selectin/IgG, VE-cadherin/IgG, or affinity-purified rabbit anti–ESL-1 (αESL-1). (a) E-selectin/IgG matrix was washed in the presence or absence of 3 mM EDTA. (c) B lymphocytes were treated with or without 1 U of neuraminidase before surface biotinylation. (d) Before incubation with the E-selectin/IgG affinity matrix with the cell extracts, the matrices were incubated with PBS (–), 400 μg of the control mAb 28AG6 against human IgG₁ Fc part, or 400 μg of the mAb UZ4 against mouse E-selectin. (e) The 240-kDa ligand was affinity isolated with E-selectin/IgG and eluted with EDTA. The eluted material was either directly electrophoresed (lane 1) or treated with (lanes 2 and 4) or without (lanes 3 and 5) 0.5 U of endoglycosidase F for 12 hours at 37°C. A third of the treated samples was directly electrophoresed (lanes 2 and 3) or reprecipitated with E-selectin/IgG (lanes 4 and 5). Labeled proteins were electrophoresed on an 6% polyacrylamide gel under reducing conditions, transferred to nitrocellulose, and detected with peroxidase-conjugated streptavidin. Molecular mass markers (in kilodaltons) are indicated. In d and e, the arrows indicate the 240-kDa E-selectin ligand.

To confirm the requirement of sialic acid for the binding of the 240-kDa ligand to the E-selectin, B lymphocytes were treated with or without neuraminidase and precipitated with the E-selectin/IgG or an antibody against ESL-1 (Figure 10c). Removal of sialic acid impaired the interaction of the 240-kDa ligand to E-selectin. In contrast, the sialidase treatment did not block the precipitation of ESL-1, which only suffered a slight decrease in the apparent molecular weight, thus showing that the disappearance of the 240-kDa band was due to lack of interaction with the E-selectin/IgG and not to unspecific degradation.

To determine the involvement of N-linked glycans in the binding of the 240-kDa glycoprotein ligand to E-selectin/IgG, we treated the biotin-labeled ligand that had been eluted with EDTA from the E-selectin/IgG affinity matrix with endoglycosidase F. This treatment caused a slight but clearly detectable decrease in the apparent molecular weight of the glycoprotein ligand (Figure 10e, lane 2), showing that this molecule carries N-linked carbohydrate side chains . Aliquots of both the endoglycosidase F and the mock-treated samples were reprecipitated with E-selectin/IgG, thus indicating that the binding of the 240-kDa ligand to E-selectin does not require N-linked glycans.

Purified human tonsillar B cells are composed of different subpopulations, each displaying a well-defined phenotype (1). Our results show that a subset of B lymphocytes purified from human tonsils interact with E-selectin, but not P-selectin, under flow conditions and that this interaction was independent of known E-selectin binding proteins. The cells that bound E-selectin were studied further by performing adhesion assays under rotation conditions to recover adherent B cells. The adherent cells expressed low levels of the GC markers CD10 and CD38, whereas they expressed high levels of CD44. Thus, GC cells were unable to interact with E-selectin. Accordingly, other studies have shown that GC B cells lack expression of the homing receptor L-selectin and are unresponsive to stromal cell–derived factor-1α (SDF-1α) and TNF-α (46–48). Altogether, these results suggest a nonmigratory phenotype of GC cells most likely related to their differentiation state. This is consistent with previous studies showing that these cells were unable to recirculate through lymphoid organs (49). Non-GC B cells (CD38^–, CD44⁺⁺) include cells of the naive and memory phenotypes that can be identified by the expression of surface IgD (1). The subset of cells that we found to be capable of interacting with E-selectin bears the memory phenotype (CD44⁺⁺, CD38^–, IgD^–). On the basis of our observations, we can assign a novel adhesive function for memory B cells, rolling on E-selectin, that might be important for directing the migration of these cells.

We have reported previously that phorbol-activated, but not resting, tonsillar B lymphocytes interact with E- and P-selectins under static conditions (29). The apparent differences in the results obtained in the previous study compared with the current work are most likely due to the adhesion conditions used in each study. Adhesion through all 3 selectins has been shown to require a minimum level of fluid shear stress to support rolling interactions (50, 51). Thus, studies under flow are the most appropriate technique to study selectin interactions in vitro (3, 4). On the other hand, the use of flow assays for studying adhesion of tonsillar B lymphocytes, which are not subjected to flow in their physiologic localization, might seem contradictory. However, the subpopulation of memory and plasma B cells generated in the tonsils will return to the circulation and will interact with the endothelium under flow conditions (2). Thus, it seems reasonable to investigate the migratory capacities of these cells under flow in order to learn about their functional behavior.

It is of interest to note the differences in the ability of resident and PB B cells to interact with endothelial selectins. Although we have found that memory tonsillar B cells can interact with E-selectin but not P-selectin, others have reported that PB B cells interact with P-selectin, and only weakly with E-selectin, under flow conditions (30). The differences regarding P-selectin binding can be simply explained by the fact that PSGL-1 is expressed by circulating blood B cells but not tonsillar B cells (refs. 21, 30; Figures 6 and 9). These results suggest that PSGL-1 expression is downregulated from the B-cell surface upon entering the tonsils during the recirculation process. Presumably, the small number of circulating blood B cells capable of binding to E-selectin may correspond to the percentage of memory cells present in the PB pool that were originated in the secondary lymphoid organs, although formal demonstration of this notion needs further investigation.

The expression of different carbohydrate epitopes has been studied in the present and previous works (29). We have found no expression of the CLA epitope recognized by HECA452, which mediates skin homing of memory T cells by interacting with E-selectin (6) and has been shown to decorate PSGL-1 (52). Although B cells express the epitope recognized by mAb 2H5, this carbohydrate epitope does not seem to be involved in binding to E-selectin because it is broadly expressed in the total population of B cells, whereas only a subpopulation binds to E-selectin. This is in agreement with previous works showing that 2H5 binds structures other than sLe^x or other related sialylated or fucosylated structures (45). In addition, although this mAb has been shown to block adhesion of HL60 to P- and E-selectin (53), it was unable to block B-cell rolling on CHO-E. Furthermore, 2H5 precipitated an 80- to 90-kDa band under reducing conditions on the B-cell lysates (not shown) that does not correspond to any of the protein ligands precipitated with the E-selectin/IgG chimera.

E-selectin has been shown to interact with PSGL-1, ESL-1, and L-selectin in certain cell types. However, none of these ligands seem to mediate interaction of tonsillar memory B cells with E-selectin, because specific blocking antibodies do not affect B-cell rolling. The antibodies used for blocking the L-selectin molecule are directed to the lectin domain protein region and not to the carbohydrates decorating this molecule. Thus, our results indicate that the lectin domain is not involved in the rolling of B cells on E-selectin. However, it has been reported previously that L-selectin from lymphocytes is unable to bind E-selectin, in contrast to neutrophil L-selectin (54). The inability of L-selectin from B cells to interact with E-selectin was further confirmed in affinity precipitation assays using a recombinant form of E-selectin, which showed that PSGL-1, L-selectin, and ESL-1 were not precipitated by the E-selectin chimera. Instead, we have found that the E-selectin/IgG fusion protein precipitates a band of approximately 240 kDa from the surface of tonsillar B cells. The precipitation of the 240-kDa band was shown to be E-selectin–specific by several criteria. First, no binding of the protein occurs to human IgG or VE-cadherin/IgG fusion protein. Second, the binding is Ca²⁺ dependent. Third, the binding is partially inhibited by the anti-mouse E-selectin mAb, but not by an anti-human IgG (Fc part)–specific mAb. Fourth, the binding is drastically reduced if sialic acid is removed from the B-cell surface by treatment with neuraminidase. Whether the latter molecule is related to the 250-kDa ligand of E-selectin found in bovine γ/δ T cells (24, 28) is not known. It is intriguing that the E-selectin ligands expressed by tonsillar B cells appear to be sensitive to neuraminidase but not to O-glycoprotease treatment, as has also been reported for the γ/δ T-cell ligand (55). Furthermore, other putative E-selectin ligands found in human T lymphocytes share similar molecular weights with those found in B cells (24). Whether E-selectin ligands expressed by B cells are the same molecules as those detected on T lymphocytes is a matter that deserves further investigation.

It is surprising that ESL-1, expressed at high levels on the surface of tonsillar B cells, is unable to interact with E-selectin, despite the fact that these cells express mRNA of FucT-IV and FucT-VII. The expression of these FucTs in CHO cells is sufficient to convert ESL-1 into an E-selectin–binding glycoform (56). Similarly, it is surprising that the analyzed tonsillar B cells do not express the sLe^x-like epitope HECA452, which is normally expressed on human cells that express these FucTs. This could argue for differences in the glycosylation machinery of B cells. In this respect, it is interesting to note that human lymphocytes, in contrast to human neutrophils (25), are unable to express an E-selectin–binding glycoform of L-selectin (54), although they can express E- and P-selectin–binding glycoforms of PSGL-1, which require the expression of FucT-VII. Nevertheless, the selective expression of FucT-VII by CD44⁺ E-selectin–adherent B cells is in accordance with the rolling of tonsillar memory B cells on E-selectin and with data suggesting that FucT-VII is responsible for the generation of functional ligands in vitro (45, 56, 57) and in vivo (58). The presence of C2GnT, an enzyme involved in the generation of functional PSGL-1 (4), might be related to the presence of functional PSGL-1 in PB B cells, which are the cells that populate the tonsils and lose the expression of PSGL-1. Our results suggest that the downregulated expression of PSGL-1 from the surface membrane of B cells after they enter the tonsils is uncoupled to the expression of the enzymes involved in the functional modification of this molecule.

The homing of T cells to different effector sites has been well studied, although the mechanisms by which memory T cells are formed is less known. By contrast, differentiation of B cells leading to the formation of memory and plasma cells has been extensively studied, although much less effort has been directed toward understanding the pathways of recirculation and homing of B cells as well as the molecules implicated in these processes. It is well known that memory B cells from secondary lymphoid organs home to secretory sites (2, 59). There is an increasing body of evidence supporting the homing of tonsillar B cells to secretory sites, such as nasal and bronchial mucosa as well as salivary and lacrimal glands (60–62), but the mechanisms regulating these trafficking pathways are largely unknown. Moreover, the constitutive expression of E-selectin in HEV of normal tonsils, adenoids (9, 10), and skin (6, 12, 13) has been reported, suggesting that E-selectin might be implicated in the recirculation of B cells through the peripheral secondary lymphoid organs and the mucosae. On the basis of the observations reported herein, we postulate that E-selectin plays a key role in the homing of memory B cells to secretory mucosae and in their recirculation through the tonsils and other local secondary lymphoid organs. This issue is supported by the fact that mice deficient in both E- and P-selectins show an increased susceptibility to mucocutaneous infections (63–65). On the other hand, B cells have also been found forming clusters of tightly packed cells at restricted sites in the inflamed tissue (66). However, the phenotypic characterization of B cells found in inflamed tissues has not been reported. Given that E-selectin is mainly an endothelial adhesion molecule found in inflamed tissues, it could be involved in the extravasation of B cells through the endothelial barrier at sites of inflammation. Our data provide insights for elucidating what molecules are implicated in the homing of resident memory B cells to tertiary sites.

We are indebted to R.T. Camphausen for the generous gift of CHO-DUKX, CHO-P, and CHO-E that made this work possible. We also acknowledge the kind gift of mAb’s by R. Hallmann, J.M. Harlan, G.D. Johnson, E.C. Butcher, P.I. Teraski, R. Kannagi, R.T. Camphausen, and T.F. Tedder. We want to thank M.A. del Pozo, M. Vicente, M.D. Gutierrez, A. Batista, and M. Viton for insightful advice and excellent technical support. We acknowledge Servicio de Otorrinolaringologia del Hospital Niño Jesus (Madrid, Spain) and Rudack Klinik und Poliklinik für Hals-Nasen und Ohrenheilkunde Muenster (Muenster, Germany) for generous provision of human tonsils. This work was supported by grants PM950162 from the Ministerio de Educación y Ciencia (to M.O. de Landázuri), grant FPI from the Comunidad Autonoma de Madrid (to M.C. Montoya), grant SFB293 from the Deutsche Forschungsgemeinschaft, Germany (to D. Vestweber), grant CB204/RPG-96-097-03 from the American Cancer Society (to G. Kansas), and grants HL-53993 and HL-36028 from the National Institutes of Health (to F.W. Luscinskas). G. Kansas is an Established Investigator of the American Heart Association.

1. Liu, Y-J, Arpin, C. Germinal center development. Immunol Rev 1997. 156:111-126.
   View this article via: PubMed CrossRef Google Scholar
2. Sprent, J. Immunological memory. Curr Opin Immunol 1997. 9:371-379.
   View this article via: PubMed CrossRef Google Scholar
3. Salmi, M, Jalkanen, S. How do lymphocytes know where to go: current concepts and enigmas of lymphocyte homing. Adv Immunol 1997. 64:139-218.
   View this article via: PubMed Google Scholar
4. Kansas, G. Selectins and their ligands: current concepts and controversies. Blood 1996. 88:3259-3287.
   View this article via: PubMed Google Scholar
5. Olofsson, AM, et al. E-selectin mediates leukocyte rolling in interleukin-1 treated mesentery venules. Blood 1994. 84:2749-2758.
   View this article via: PubMed Google Scholar
6. Picker, LJ, Kishimoto, TK, Smith, CW, Warnock, RA, Butcher, EC. ELAM-1 is an adhesion molecule for skin homing T cells. Nature 1991. 349:796-799.
   View this article via: PubMed CrossRef Google Scholar
7. Nguyen, M, Strubel, N, Bischoff, J. A role for sialyl Lewis -X/A glycoconjugates in capillary morphogenesis. Nature 1993. 365:267-269.
   View this article via: PubMed CrossRef Google Scholar
8. Kralig, BM, et al. E-selectin is present in proliferating endothelial cells in human hemangiomas. Am J Pathol 1996. 148:1181-1191.
   View this article via: PubMed Google Scholar
9. Andoh, N, et al. Expression of E- and P-selectins by vascular endothelial cells in human tonsils. Acta Otolaryngol (Stockh) Suppl 1996. 523:52-54.
10. Perry, ME, Niall, W, Kirkpatrik, A, Happerfield, LC, Gleeson, MJ. Expression of adhesion molecules on the microvasculature of the pharyngeal tonsil (adenoid). Acta Oto-Laryngol Suppl 1996. 523:47-51.
11. Schweitzer, KM, et al. Constitutive expression of E-selectin and vascular cell adhesion molecule-1 on endothelial cells of hematopoietic tissues. Am J Pathol 1996. 148:165-175.
    View this article via: PubMed Google Scholar
12. Norton, J, Sloane, JP, Al-Saffar, N, Haskard, DO. Vessel associated adhesion molecules in normal skin and acute graft-versus-host disease. J Clin Pathol 1991. 44:586-591.
    View this article via: PubMed CrossRef Google Scholar
13. Uccini, S, et al. Molecular mechanisms involved in intraepithelial lymphocyte migration: a comparative study in skin and tonsil. J Pathol 1993. 169:413-419.
    View this article via: PubMed CrossRef Google Scholar
14. Hurwitz, AA, Lyman, WD, Guida, MP, Calderon, TM, Berman, JW. Tumor necrosis factor α induces adhesion molecule expression on human fetal astrocytes. J Exp Med 1992. 176:1631-1636.
    View this article via: PubMed CrossRef Google Scholar
15. Essani, NA, McGuire, GM, Manning, AM, Jaesche, H. Endotoxin-induced activation of the nuclear transcription factor kB and expression of E-selectin messenger RNA in hepatocytes, Kupffer cells and endothelial cells in vivo. J Immunol 1996. 156:2956-2963.
    View this article via: PubMed Google Scholar
16. Moore, KL, et al. Identification of a specific glycoprotein ligand for P-selectin (CD62) on myeloid cells. J Cell Biol 1992. 118:445-456.
    View this article via: PubMed CrossRef Google Scholar
17. Sako, D, et al. Expression cloning of a functional glycoprotein ligand for P-selectin. Cell 1993. 75:1179-1186.
    View this article via: PubMed CrossRef Google Scholar
18. Asa, D, et al. The P-selectin glycoprotein ligand functions as a common human leukocyte ligand for P- and E-selectins. J Biol Chem 1995. 270:11662-11670.
    View this article via: PubMed CrossRef Google Scholar
19. Walcheck, B, Moore, KL, McEver, RP, Kishimoto, TK. Neutrophil-neutrophil interactions under hydrodynamic shear stress involve L-selectin and PSGL-1. A mechanism that amplifies initial leukocyte accumulation on P-selectin in vitro. J Clin Invest 1996. 98:1081-1087.
    View this article via: JCI PubMed Google Scholar
20. Tu, L, et al. L-selectin binds to P-selectin glycoprotein ligand-1 on leukocytes. Interactions between the lectin, epidermal growth factor and consensus repeat domains of the selectins determine ligand binding specificity. J Immunol 1996. 157:3995-4004.
    View this article via: PubMed Google Scholar
21. Snapp, KR, et al. A novel P-selectin glycoprotein ligand-1 monoclonal antibody recognizes an epitope within the tyrosine sulfate motif of human PSGL-1 and blocks recognition of both P- and L-selectin. Blood 1998. 91:154-164.
    View this article via: PubMed Google Scholar
22. Snapp, KR, Wagers, AJ, Craig, R, Stoolman, LM, Kansas, GS. P-selectin glycoprotein ligand-1 (PSGL-1) is essential for adhesion to P-selectin but not E-selectin in stably transfected hematopoietic cell lines. Blood 1996. 89:896-901.
    View this article via: PubMed Google Scholar
23. Steegmaier, M, Blanks, JE, Borges, E, Vestweber, D. P-selectin glycoprotein ligand-1 mediates rolling of mouse bone marrow-derived mast cells on P-selectin but not efficiently on E-selectin. Eur J Immunol 1997. 27:1339-1345.
    View this article via: PubMed CrossRef Google Scholar
24. Jones, WM, Watts, GM, Robinson, MK, Vestweber, D, Jutila, MA. Comparison of E-selectin-binding glycoprotein ligands on human lymphocytes, neutrophils, and bovine γδ T cells. J Immunol 1997. 159:3574-3583.
    View this article via: PubMed Google Scholar
25. Zollner, , et al. L-selectin from human, but not mouse neutrophils binds directly to E-selectin. J Cell Biol 1997. 136:707-716.
    View this article via: PubMed CrossRef Google Scholar
26. Lenter, M, Levinovitz, A, Isenmann, S, Vestweber, D. Monospecific and common glycoprotein ligands for E- and P-selectin on myeloid cells. J Cell Biol 1994. 125:471-481.
    View this article via: PubMed CrossRef Google Scholar
27. Steegmaier, M, et al. The E-selectin-ligand ESL-1 is a variant of a receptor for fibroblast growth factor. Nature 1995. 373:615-620.
    View this article via: PubMed CrossRef Google Scholar
28. Walcheck, B, Watts, G, Jutila, MA. Bovine γ/δ T cells bind E-selectin via a novel glycoprotein receptor: first characterization of a lymphocyte/E-selectin interaction in an animal model. J Exp Med 1993. 178:853-863.
    View this article via: PubMed CrossRef Google Scholar
29. Postigo, AA, Marazuela, M, Sanchez-Madrid, F, de Landázuri, MO. B lymphocyte binding to E- and P-selectins is mediated through de novo expression of carbohydrates on in vitro and in vivo activated human B cells. J Clin Invest 1994. 94:1585-1596.
    View this article via: JCI PubMed Google Scholar
30. Yago, T, et al. Analysis of initial attachment of B cells to endothelial cells under flow conditions. J Immunol 1997. 158:707-714.
    View this article via: PubMed Google Scholar
31. Vachino, G, et al. P-selectin glycoprotein ligand-1 is the major counter-receptor for P-selectin on stimulated T cells and is widely distributed in non-functional form on many lymphocytic cells. J Biol Chem 1995. 270:21966-21974.
    View this article via: PubMed CrossRef Google Scholar
32. Bevilacqua, MP, Pober, JS, Mendrick, DL, Cotran, RS, Gimbrone (Jr), MA. Identification of an inducible endothelial-leukocyte adhesion molecule. Proc Natl Acad Sci USA 1987. 84:9238-9242.
    View this article via: PubMed CrossRef Google Scholar
33. Hallmann, A, Zimmerman, U, Sorokin, LM, Needham, L, Mark, Kvd. Adhesion of leukocytes to the inflamed endothelium. Scand J Rheumatol 1995. 24(Suppl. 101):107-109.
    View this article via: CrossRef Google Scholar
34. Levinovitz, A, Mühlhoff, J, Isenman, S, Vestweber, D. Identification of a glycoprotein ligand for E-selectin on mouse myeloid cells. J Cell Biol 1993. 121:449-459.
    View this article via: PubMed CrossRef Google Scholar
35. Postigo, AA, Sánchez-Mateos, P, Lazarovits, AI, Sánchez-Madrid, F, de Landázuri, MO. α4β7 integrin mediates B cell binding to fibronectin and vascular cell adhesion molecule-1. Expression and function of α4 integrins on human B lymphocytes. J Immunol 1993. 151:2471-2483.
    View this article via: PubMed Google Scholar
36. Carlos, T, et al. Vascular cell adhesion molecule-1 (VCAM-1) mediates lymphocyte adhesion to cytokine-activated cultured human endothelial cells. Blood 1990. 76:965-970.
    View this article via: PubMed Google Scholar
37. Fukushima, K, et al. Characterization of sialosylated Lewis x as a new tumor associated antigen. Cancer Res 1984. 44:5279-5285.
    View this article via: PubMed Google Scholar
38. Sawada, M, et al. Specific expression of a complex sialyl Lewis X antigen on high endothelial venules of human lymph nodes: possible candidate for L-selectin ligand. Biochem Biophys Res Commun 1993. 193:337-347.
    View this article via: PubMed CrossRef Google Scholar
39. Ohmori, K, et al. A distinct type of sialyl Lewis X antigen defined by a novel monoclonal antibody is selectively expressed on helper memory T cells. Blood 1993. 82:2797-2805.
    View this article via: PubMed Google Scholar
40. Spertini, O, Kansas, GS, Reiman, KA, Mackay, CR, Tedder, TF. Function and evolutionary conservation of distinct epitopes on the leukocyte adhesion molecule-1 (LAM-1) that regulate leukocyte migration. J Immunol 1991. 147:942-949.
    View this article via: PubMed Google Scholar
41. Montoya, MC, Luscinskas, FW, del Pozo, MA, Aragonés, J, de Landázuri, MO. Reduced intracellular oxidative metabolism promotes firm adhesion of human polymorphonuclear leukocytes to vascular endothelium under flow conditions. Eur J Immunol 1997. 27:1942-1951.
    View this article via: PubMed CrossRef Google Scholar
42. Hahne, M, Jöger, U, Isenmann, S, Hallmann, R, Vestweber, D. Five tumor necrosis factor-inducible cell adhesion mechanisms on the surface of mouse endothelioma cells mediate the binding of leukocytes. J Cell Biol 1993. 121:655-664.
    View this article via: PubMed CrossRef Google Scholar
43. Gotsch, U, et al. VE-cadherin antibody accelerates neutrophil recruitment in vivo. J Cell Sci 1997. 110:583-588.
    View this article via: PubMed Google Scholar
44. Borges, E, et al. The binding of T cell–expressed P-selectin glycoprotein ligand-1 to E- and P-selectin is differentially regulated. J Biol Chem 1997. 272:28786-28792.
    View this article via: PubMed CrossRef Google Scholar
45. Wagers, AJ, Stoolman, LM, Kannagi, R, Craig, R, Kansas, GS. Expression of leukocyte fucosyltransferases regulates binding to E-selectin. Relationship to previously implicated carbohydrate epitopes. J Immunol 1997. 159:1917-1929.
    View this article via: PubMed Google Scholar
46. Munro, JM, Briscoe, DM, Tedder, TF. Differential regulation of leukocyte L-selectin (CD62L) expression in normal lymphoid and inflamed extralymphoid tissues. J Clin Pathol 1996. 49:721-727.
    View this article via: PubMed CrossRef Google Scholar
47. Bleul, CC, Schultze, JL, Springer, TA. B lymphocyte chemotaxis regulated in association with microanatomic localization, differentiation state, and B cell receptor engagement. J Exp Med 1998. 187:753-762.
    View this article via: PubMed CrossRef Google Scholar
48. Corcione, A, et al. Recombinant tumor necrosis factor enhances the locomotion of memory and naive lymphocytes from human tonsils through the selective engagement of the type II receptor. Blood 1997. 90:4493-4501.
    View this article via: PubMed Google Scholar
49. Reichert, RA, Gallatin, WM, Weissman, IL, Butcher, EC. Germinal center B cells lack homing receptors necessary for normal lymphocyte recirculation. J Exp Med 1983. 157:813-827.
    View this article via: PubMed CrossRef Google Scholar
50. Finger, EB, et al. Adhesion through L-selectin requires a threshold of hydrodynamic shear. Nature 1996. 379:266-269.
    View this article via: PubMed CrossRef Google Scholar
51. Lawrence, MB, Kansas, GS, Kunkel, EJ, Ley, K. Threshold levels of fluid shear promote leukocyte adhesion through selectins (CD62L, P, E). J Cell Biol 1997. 136:717-727.
    View this article via: PubMed CrossRef Google Scholar
52. Fuhlbrigge, RC, Kieffer, JD, Armerding, D, Kupper, TS. Cutaneous lymphocyte antigen is a specialized form of PSGL-1 expressed on skin-homing T cells. Nature 1997. 389:978-981.
    View this article via: PubMed CrossRef Google Scholar
53. Tamatani, T, et al. Recognition of consensus CHO structure in ligands for selectins by novel antibody against sialyl Lewis X. Am J Physiol 1995. 269:H1282-H1287.
    View this article via: PubMed Google Scholar
54. Picker, LJ, et al. The neutrophil selectin LECAM-1 presents carbohydrate ligands to the vascular selectins ELAM-1 and GMP-140. Cell 1991. 66:921-933.
    View this article via: PubMed CrossRef Google Scholar
55. Jutila, MA, Kurk, S. Analysis of bovine γδ T cell interactions with E-, P-, and L-selectin. Characterization of lymphocyte on lymphocyte rolling and the effects of O-glycoprotease. J Immunol 1996. 156:289-296.
    View this article via: PubMed Google Scholar
56. Zollner, O, Vestweber, D. The E-selectin ligand-1 is selectively activated in Chinese hamster ovary cells by the α(1,3)-fucosyltransferases IV and VII. J Biol Chem 1996. 271:33002-33008.
    View this article via: PubMed CrossRef Google Scholar
57. Knibbs, RN, et al. α(1,3)-fucosyltransferase VII-dependent synthesis of P-and E-selectin ligands on cultured T lymphoblasts. J Immunol 1998. 161:6305-6315.
    View this article via: PubMed Google Scholar
58. Mály, P, et al. The α(1, 3) fucosyltransferase Fuc-TVII controls leukocyte trafficking through an essential role in L- E-, and P-selectin ligand biosynthesis. Cell 1996. 86:643-653.
    View this article via: PubMed CrossRef Google Scholar
59. Brandtzaeg, P. The B-cell development in tonsillar lymphoid follicles. Acta Oto-Laryngol Suppl 1996. 523:55-59.
60. Liu, Y-J, et al. Memory B cells from human tonsils colonize mucosal epithelium and directly present antigen to T cells by rapid up-regulation of B7-1 and B7-2. Immunity 1995. 2:239-248.
    View this article via: PubMed CrossRef Google Scholar
61. Brandtzaeg, P. The role of humoral mucosal immunity in the induction and maintenance of chronic airway infections. Am J Respir Crit Care Med 1995. 151:2081-2087.
    View this article via: PubMed Google Scholar
62. Brandtzaeg, P. Humoral immune response patterns of human mucosae: induction and relation to bacterial respiratory tract infections. J Infect Dis 1992. 165(Suppl. 1):S167-S176.
    View this article via: PubMed Google Scholar
63. Bullard, DC, et al. Infectious susceptibility and severe deficiency of leukocyte rolling and recruitment in E-selectin and P-selectin double mutant mice. J Exp Med 1996. 183:2329-2336.
    View this article via: PubMed CrossRef Google Scholar
64. Frenette, PS, Mayadas, TN, Rayburn, H, Hynes, RO, Wagner, D. Susceptibility to infection and altered hematopoiesis in mice deficient in both P- and E-selectins. Cell 1996. 84:563-574.
    View this article via: PubMed CrossRef Google Scholar
65. Muñoz, FM, Hawkins, EP, Bullard, D, Beaudet, A, Kaplan, SL. Host defense against systemic infection with Streptococcus pneumoniae is impaired in E-, P-, and E-/P-selectin–deficient mice. J Clin Invest 1997. 100:2099-2106.
    View this article via: JCI PubMed Google Scholar
66. Meeusen, E, Lee, S, Brandon, M. Differential migration of T and B cells during an acute inflammatory response. Eur J Immunol 1991. 21:2269-2272.
    View this article via: PubMed CrossRef Google Scholar