The CC′ and DE Loops in Ig Domains 1 and 2 of MAdCAM-1 Play Different Roles in MAdCAM-1 Binding to Low- and High-affinity Integrin α4β7*

Lymphocyte homing is regulated by the dynamic interaction between integrins and their ligands. Integrin α4β7 mediates both rolling and firm adhesion of lymphocytes by modulating its affinity to the ligand, mucosal addressin cell adhesion molecule-1 (MAdCAM-1). Although previous studies have revealed some mechanisms of α4β7-MAdCAM-1 binding, little is known about the different molecular bases of the low- and high-affinity α4β7-MAdCAM-1 interactions, which mediate rolling and firm adhesion of lymphocytes, respectively. Here, we found that two loops in immunoglobulin domains 1 and 2 (D1 and D2) of MAdCAM-1 played different roles in MAdCAM-1 binding to low-affinity (inactive) and high-affinity (activated) α4β7. The Asp-42 in the CC′ loop of D1 was indispensable for MAdCAM-1 binding to both low-affinity and high-affinity α4β7. The other CC′ loop residues except for Arg-39 and Ser-44 were essential for MAdCAM-1 binding to both inactive α4β7 and α4β7 activated by SDF-1α or talin, but not required for MAdCAM-1 binding to Mn2+-activated α4β7. Single amino acid substitution of the DE loop residues mildly decreased MAdCAM-1 binding to both inactive and activated α4β7. Notably, removal of the DE loop greatly impaired MAdCAM-1 binding to inactive and SDF-1α- or talin-activated α4β7, but only decreased 60% of MAdCAM-1 binding to Mn2+-activated α4β7. Moreover, DE loop residues were important for stabilizing the low-affinity α4β7-MAdCAM-1 interaction. Thus, our findings demonstrate the distinct roles of the CC′ and DE loops in the recognition of MAdCAM-1 by low- and high-affinity α4β7 and suggest that the inactive α4β7 and α4β7 activated by different stimuli have distinct conformations with different structural requirements for MAdCAM-1 binding.

Lymphocyte homing from circulation to lymphoid tissues and sites of inflammation is regulated by the dynamic interaction between lymphocyte integrin receptors and endothelial immunoglobulin superfamily cell adhesion molecules (1)(2)(3).
Integrin ␣ 4 ␤ 7 is an important lymphocyte homing receptor that can mediate both rolling and firm adhesion of lymphocytes, two of the critical steps in lymphocyte migration and tissue-specific homing (4). Its ligand, MAdCAM-1, 2 is preferentially expressed on high endothelial venules of gut-associated lymphoid organs and on lamina propria venules, helping lymphocyte traffic to mucosal organs (5). The interaction between MAdCAM-1 and integrin ␣ 4 ␤ 7 is the key step in lymphocyte homing to gut and plays vital roles in both gut mucosal immune homeostasis and intestinal inflammation (6,7).
Human MAdCAM-1 is a multidomain molecule consisting of two Ig-like domains followed by mucin-like sequences (8). Domain swapping experiments with MAdCAM-1 and VCAM-1 have demonstrated the requirement of the two Ig domains of MAdCAM-1 for efficient integrin ␣ 4 ␤ 7 binding (9). The crystal structure of MAdCAM-1 highlights two protruding loops from the two Ig domains, the CCЈ loop in D1 and DE loop in D2, which are important for the interaction between MAdCAM-1 and ␣ 4 ␤ 7 (10,11). The essential integrin-binding motif (LDTS) resides in the CCЈ loop of MAdCAM-1 D1, and the Asp-42 in the LDTS motif serves as the primary ␣ 4 ␤ 7 -binding site by directly interacting with the metal ion at the metal ion-dependent adhesion site (MIDAS) in the integrin ␤ 7 I domain (8,(12)(13)(14). Mutational studies suggest that some other residues in the CCЈ loop of MAdCAM-1 are also involved in MAdCAM-1-␣ 4 ␤ 7 binding (9,12,13). The DE loop in D2 is predominated by negatively charged residues, which have been reported to play an important role in determining integrin binding specificity (15). Mutagenesis study has shown that some residues in the DE loop are important for MAdCAM-1 binding to integrin ␣ 4 ␤ 7 (9).
Integrins are ␣/␤ heterodimeric cell adhesion molecules that mediate cell-cell, cell-extracellular matrix, and cell-pathogen interactions and transmit signals bidirectionally across the plasma membrane (16 -18). Cell adhesion through integrin is dependent on the dynamic regulation of integrin affinity. The low-affinity integrin ␣ 4 ␤ 7 mediates rolling adhesion of lymphocytes. Upon activation, ␣ 4 ␤ 7 converts to a high-affinity state, which mediates firm cell adhesion. Early studies on integrin structure have revealed that integrin extracellular domains exist in at least three distinct global conformational states: bent with a closed headpiece, extended with a closed headpiece, and extended with an open headpiece, which correspond to the low-affinity, intermediate-affinity, and high-affinity states, respectively (19 -21). The equilibrium among these different affinity states is regulated by integrin inside-out signaling and certain extracellular stimuli, such as divalent cations (22,23). When compared with the low-affinity state in Ca 2ϩ ϩ Mg 2ϩ , removal of Ca 2ϩ or the addition of Mn 2ϩ strikingly increases ligand binding affinity and adhesiveness of almost all integrins (14, 24 -26). Crystal structures of ␣ V ␤ 3 and ␣ IIb ␤ 3 integrins revealed three interlinked metal ion-binding sites in integrin ␤ I domain (27,28). The central MIDAS is flanked by two metal ion-binding sites, the adjacent to MIDAS (ADMIDAS) site and the synergistic metal ion-binding site. The divalent cation at MIDAS directly coordinates the acidic side chain shared by all integrin ligands and is essential for integrin-ligand binding (14,29). The synergistic metal ion-binding site and ADMIDAS function as positive and negative regulatory sites, respectively (14, 25, 30 -32). Upon activation, integrin with the bent conformation converts to the extended conformation coupled with a series of global and local conformational changes, including separation of cytoplasmic tails, extension of integrin ectodomains, swing-out of the hybrid domain, ␤ I domain ␣7 helix downward movement, and conformational rearrangement at the integrin ligand-binding site around MIDAS and ADMIDAS.
Despite the above advances in understanding of ␣ 4 ␤ 7binding hotspots on MAdCAM-1 and integrin conformational rearrangement during activation, the molecular basis for the recognition of MAdCAM-1 by low-and high-affinity ␣ 4 ␤ 7 remains elusive. The mechanism of rolling and firm adhesion of lymphocyte mediated by ␣ 4 ␤ 7 on MAdCAM-1 is not well understood.
In this study, we found that the CCЈ loop in D1 and DE loop in D2 of MAdCAM-1 exerted different functions in MAdCAM-1 binding to low-and high-affinity ␣ 4 ␤ 7 . In addition, we demonstrated that the inactive ␣ 4 ␤ 7 and ␣ 4 ␤ 7 activated by different stimuli might have distinct conformations with different structural requirements for MAdCAM-1 binding.
Isolation of Peripheral Blood Lymphocytes (PBLs)-Peripheral venous blood from normal donors was collected using anticoagulant citrate dextrose as an anticoagulant. Human peripheral blood mononuclear cells were isolated from buffy coats using a Ficoll density gradient, washed, and suspended in RPMI 1640 (Invitrogen), supplemented with 100 g/ml penicillin, 100 g/ml streptomycin, 2 mM glutamine (all from Invitrogen), and 10% (v/v) FBS (Biocherom AG, Germany). Monocytes were depleted by incubation on tissue culture plastic for 30 min at 37°C. The lymphocytes were rich in the supernatant fluid.
Flow Chamber Assay-The flow chamber assay was performed as described (25). A polystyrene Petri dish was coated with a 5-mm diameter, 20-l spot of 10 g/ml purified huMAdCAM-1/Fc in coating buffer (PBS, 10 mM NaHCO 3 , pH 9.0) for 1 h at 37°C followed by 2% BSA in coating buffer for 1 h at 37°C to block nonspecific binding sites. Cells were washed twice with Ca 2ϩ -ϩ Mg 2ϩ -free HEPES-buffered saline (20 mM Hepes, pH 7.4, 5 mM EDTA, 0.5% BSA), resuspended at 1 ϫ 10 7 /ml in buffer A (Ca 2ϩ -ϩ Mg 2ϩ -free HEPES-buffered saline, 0.5% BSA), and kept at room temperature. Cells were diluted to 1 ϫ 10 6 /ml in buffer A containing different divalent cations immediately before infusion in the flow chamber using a Harvard apparatus programmable syringe pump. Cells were allowed to accumulate for 30 s at 0.3 dyne/cm 2 and 10 s at 0.4 dyne/cm 2 . Then, shear stress was increased every 10 s from 1 dyne/cm 2 up to 16 dynes/cm 2 in 2-fold increments. The number of cells remaining bound at the end of each 10-s interval was determined. Rolling velocity at each shear stress was calculated from the average distance traveled by rolling cells in 3 s. To avoid confusing rolling with small amounts of movement due to tether stretching or measurement error, a velocity of 2 m/s, which corresponds to a movement of one-half cell diameter during the 3-s measurement interval, was the minimum velocity required to define a cell as rolling instead of firmly adherent.
SDF-1␣ Stimulation Assay under Flow-A polystyrene Petri dish was coated with a 5-mm diameter, 20-l spot of 10 g/ml purified huMAdCAM-1/Fc with SDF-1␣ (2 g/ml) or huMAdCAM-1/Fc alone in coating buffer for 1 h at 37°C followed by 2% BSA in coating buffer for 1 h at 37°C to block nonspecific binding sites. The PBLs (1 ϫ 10 6 /ml) were infused into the chamber, and then cells were allowed to settle for 2 min and to accumulate for 30 s at 0.3 dyne/cm 2 and 10 s at 0.4 dyne/cm 2 . Then, shear stress was increased every 10 s from 1 dyne/cm 2 up to 16 dynes/cm 2 , in 2-fold increments. Cells remaining bound under the wall shear stress of 1 dyne/cm 2 were counted. The PBLs (1 ϫ 10 6 /ml) preincubated with ␣ 4 ␤ 7blocking mAb Act-1 (2 g/ml) or treated with 5 mM EDTA were used as control.

RESULTS
To study the molecular basis for the recognition of MAd-CAM-1 by low-and high-affinity integrin ␣ 4 ␤ 7 , we generated soluble human MAdCAM-1 protein (from Val-1 to Pro-315), which contains domain 1, domain 2, and mucin-like domain with C-terminal fused Fc1 and two regions of human IgG1. Because the CCЈ loop in D1 of MAdCAM-1 possesses the LDTS motif, which has been implicated as the primary integrin-binding site, we first investigated the function of the CCЈ loop in the binding of low-and high-affinity integrin ␣ 4 ␤ 7 to MAdCAM-1 by introducing a series of single point mutations in the CCЈ loop based on the MAdCAM-1 crystal structure ( Fig. 1) (10, 11).
Asp-42 in CCЈ Loop of MAdCAM-1 D1 Is Essential for MAdCAM-1 Binding to Both Low-affinity and High-affinity ␣ 4 ␤ 7 -Adhesive behavior of 293T cells stably expressing human integrin ␣ 4 ␤ 7 in shear flow was characterized in a parallel wall flow chamber by allowing them to adhere to MAdCAM-1 adsorbed to the lower wall. The shear stress was incrementally increased, and the velocity of the cells remaining bound at each increment was determined. Human ␣ 4 ␤ 7 293T transfectants behaved as described previously for lymphoid cells expressing ␣ 4 ␤ 7 (26). In 1 mM Ca 2ϩ ϩ 1 mM Mg 2ϩ , about 90% of the bound ␣ 4 ␤ 7 transfectants rolled at the shear stress of 1 dyne/cm 2 ( Fig. 2A). By contrast, cells were firmly adherent in 0.5 mM Mn 2ϩ (Fig. 2B). Rolling and firm adhesion represent the low-and high-affinity interactions of ␣ 4 ␤ 7 with MAdCAM-1, respectively. As control, ␣ 4 ␤ 7 transfectants treated with ␣ 4 ␤ 7 blocking antibody Act-1 or with 5 mM EDTA did not accumulate on MAdCAM-1 substrates (Fig. 2, A and B). In contrast to the robust cell adhesion on WT MAdCAM-1, substitution of Asp-42 with Ala abolished both rolling and firm adhesion on MAdCAM-1 mediated by the low-affinity ␣ 4 ␤ 7 in 1 mM Ca 2ϩ ϩ 1 mM Mg 2ϩ and high-affinity ␣ 4 ␤ 7 in 0.5 mM Mn 2ϩ , suggesting its essential role in integrin-ligand binding (Fig. 2, A and B).
In Addition to Asp-42, other CCЈ Loop Residues Except for Arg-39 and Ser-44 Are Essential for MAdCAM-1 Binding to Low-affinity ␣ 4 ␤ 7 -When compared with the efficient rolling cell adhesion on WT MAdCAM-1, single amino acid substitution of most residues in the MAdCAM-1 CCЈ loop with Ala abolished the rolling cell adhesion on MAdCAM-1 mediated by low-affinity ␣ 4 ␤ 7 in 1 mM Ca 2ϩ ϩ 1 mM Mg 2ϩ (Fig. 2A). Arg-39  and Ser-44 to Ala mutations showed less effect, which led to 70 and 80% decrease of cell adhesion to MAdCAM-1, respectively.
CCЈ Loop Residues Other than Asp-42 Are Not Important for MAdCAM-1 Binding to High-affinity ␣ 4 ␤ 7 Activated by Mn 2ϩ -In contrast to rolling cell adhesion mediated by low-affinity ␣ 4 ␤ 7 , the firm cell adhesion mediated by Mn 2ϩ -activated ␣ 4 ␤ 7 was only slightly affected by the same mutations, except for Asp-42 (Fig.  2B). More than 70% of cell adhesion was retained when Leu-41, Thr-43, Leu-45, and Gly-46 were mutated to Ala, respectively. The rest of the mutations only caused less than 10% loss of cell adhesion in Mn 2ϩ . Thus, except for the primary MAdCAM-1-␣ 4 ␤ 7 -binding site Asp-42, the rest of the CCЈ residues other than Arg-39 and Ser-44 in the CCЈ loop are essential for the interaction between low-affinity ␣ 4 ␤ 7 and MAdCAM-1, but not required for the binding of Mn 2ϩ -activated ␣ 4 ␤ 7 to MAdCAM-1.
CCЈ Loop Residues Other than Arg-39 and Ser-44 Are Crucial for MAdCAM-1 Interaction with High-affinity Integrin ␣ 4 ␤ 7 Activated by Talin or SDF-1␣-Besides the unphysiological strong activation by Mn 2ϩ , integrin can be activated by more physiological pathways such as overexpression of intracellular talin or SDF-1␣ stimulation (34). Talin is a cytoskeletal protein that can interact with the cytoplasmic tail of the integrin ␤ subunit and activate integrin. To investigate the influence of the CCЈ loop mutations on the interaction between MAdCAM-1 and talin-activated ␣ 4 ␤ 7 , we overexpressed the GFP-talin-head in ␣ 4 ␤ 7 293T transfectants (supplemental Fig. S1) and examined the cell adhesion behavior to WT and mutant MAdCAM-1 under flow (Fig. 2C). The firmly adherent ␣ 4 ␤ 7 transfectants on WT MAdCAM-1 increased from 11 to 37% of total bound cells under the shear stress of 1 dyne/cm 2 after co-transfection with GFP-talin-head, suggesting the activation of integrin ␣ 4 ␤ 7 by talin. Surprisingly, unlike the mild effects of the CCЈ mutations on the cell adhesion mediated by Mn 2ϩactivated ␣ 4 ␤ 7 to MAdCAM-1, the cell adhesion mediated by talin-activated ␣ 4 ␤ 7 was greatly disrupted by most CCЈ loop mutations. GFP-talin-head overexpression only slightly increased cell adhesion on R39A, G40A, S44A, L45A, G46A, and V48A mutants. R39A and S44A showed less effect, which led to 76 and 54% decrease of cell adhesion mediated by talinactivated ␣ 4 ␤ 7 to MAdCAM-1. Thus, the above results suggest that the CCЈ loop residues other than Arg-39 and Ser-44 are crucial for the recognition of MAdCAM-1 by talin-activated ␣ 4 ␤ 7 and that integrin ␣ 4 ␤ 7 activated by Mn 2ϩ and talin could have different conformations with different structural requirements for MAdCAM-1 binding.
Because the CCЈ loop might play different roles in MAdCAM-1 interaction with Mn 2ϩ -and talin-activated ␣ 4 ␤ 7 , we next tested its function in the interaction between MAdCAM-1 and ␣ 4 ␤ 7 activated by SDF-1␣. SDF-1␣ can induce integrin activation through the PI3 kinase pathway by binding to CXCR4, the G protein-coupled receptor of SDF-1␣ (35,36). Human PBLs were used that express high levels of integrin ␣ 4 ␤ 7 (37) and CXCR4 (38). In contrast to the robust cell adhesion to WT MAdCAM-1 in 1 mM Ca 2ϩ ϩ 1 mM Mg 2ϩ , PBL adhered weakly to most CCЈ loop mutants and did not adhere to D42A and T43A mutants at all (Fig. 2D). R39A and S44A mutations showed less effect, which led to 86 and 68% decrease of cell adhesion to MAdCAM-1, respectively. Activa-tion of ␣ 4 ␤ 7 by SDF1-␣ stimulation notably increased the number of PBLs bound to WT MAdCAM-1, but not to the MAdCAM-1 CCЈ loop mutants, suggesting the importance of the CCЈ loop residues in the interaction between MAdCAM-1 and ␣ 4 ␤ 7 activated by SDF1-␣ (Fig. 2D). Interestingly, cell adhesion to the MAdCAM-1 CCЈ loop mutants could be increased by talin but not SDF1-␣, indicating the subtle difference between integrins activated by talin and SDF-1␣ (Fig. 2, C  and D).
Taken together, the above data demonstrate that Asp-42 in the CCЈ loop is essential for MAdCAM-1 binding to both lowaffinity and high-affinity ␣ 4 ␤ 7 , and the rest of the CCЈ loop residues other than Arg-39 and Ser-44 are crucial for MAdCAM-1 binding to ␣ 4 ␤ 7 activated by physiological stimuli as SDF-1␣ or talin, but not required for MAdCAM-1 binding to ␣ 4 ␤ 7 activated by Mn 2ϩ . Thus, integrins activated by distinct stimuli might have different high-affinity conformations with diverse structural requirements for MAdCAM-1 binding.
CCЈ Loop Is Required for the Stable Interaction between MAdCAM-1 and Low-affinity Integrin ␣ 4 ␤ 7 -To test the effect of the CCЈ loop on the strength of ␣ 4 ␤ 7 -mediated cell adhesion to MAdCAM-1, we examined the resistance of adherent cells to detachment by increasing shear stress (Fig. 3). MAdCAM-1 R39A and S44A mutants were chosen for studies on low-affinity ␣ 4 ␤ 7 in Ca 2ϩ ϩ Mg 2ϩ and high-affinity ␣ 4 ␤ 7 activated by talin or SDF-1␣ because they were the only CCЈ loop mutants supporting clearly detectable cell adhesion under those conditions. In 1 mM Ca 2ϩ ϩ 1 mM Mg 2ϩ , ␣ 4 ␤ 7 293T transfectants detached more rapidly from the R39A and S44A mutants than from the WT MAdCAM-1 (Fig. 3A), suggesting the less stable interaction between integrin and the mutant ligands. In contrast, adhesion of cells bearing high-affinity ␣ 4 ␤ 7 activated by Mn 2ϩ , talin, or SDF-1␣ to the MAdCAM-1 CCЈ loop mutants was much less susceptible to the increased shear stress (Fig. 3, B-D).
The Intact DE Loop Is Essential for MAdCAM-1 Binding to Low-affinity ␣ 4 ␤ 7 , but Not for Its Binding to High-affinity ␣ 4 ␤ 7 Activated by Mn 2ϩ -To further define the different structural requirements for MAdCAM-1 binding to inactive and activated integrin ␣ 4 ␤ 7 , we generated a series of single amino acid substitutions and a partial deletion (⌬DE, from Glu-152 to Asp-158) of the DE loop in MAdCAM-1 D2. MAdCAM-1 with deletion of the whole DE loop (from Glu-149 to Asp-158) could not be expressed. All of the single amino acid substitutions by Ala in the DE loop (residue 149 -158) decreased the cell adhesion mediated by both low-affinity and high-affinity ␣ 4 ␤ 7 on MAdCAM-1, but to a different extent (Fig. 4A). For the rolling cell adhesion mediated by low-affinity ␣ 4 ␤ 7 on MAdCAM-1 in 1 mM Ca 2ϩ ϩ 1 mM Mg 2ϩ , substitution of Glu-152 and Glu-153 by Ala led to an ϳ70 and 60% decrease of adherent cells, respectively. E150A, E154A, and D158A showed less effect, which resulted in an ϳ50% decrease (Fig. 4A). Other point mutations had even milder impact, retaining from 60 to 80% of adherent cells. For the firm cell adhesion mediated by high-affinity ␣ 4 ␤ 7 in Mn 2ϩ , most DE loop point mutations showed much milder effects, except for Glu-150 and Glu-154, which led to 50 and 40% decrease of bound cells, respectively (Fig. 4B). Notably, the DE loop residues that mostly affected MAdCAM-1 interaction with low-affinity and high-affinity ␣ 4 ␤ 7 are different, suggesting the different roles of the DE loop residues in the recognition of integrin ␣ 4 ␤ 7 before and after activation. Different from single amino acid substitutions, the partial deletion of the DE loop abolished the interaction between low-affinity ␣ 4 ␤ 7 and MAdCAM-1, but only caused about 60% loss of cell adhesion to MAdCAM-1 mediated by the Mn 2ϩ -activated ␣ 4 ␤ 7 . The data demonstrate that the intact DE loop is essential to MAdCAM-1 binding to low-affinity ␣ 4 ␤ 7 , but not to highaffinity ␣ 4 ␤ 7 activated by Mn 2ϩ .
The Intact DE Loop Is Required for MAdCAM-1 Binding to Integrin ␣ 4 ␤ 7 Activated by Talin and SDF-1␣-To further study the function of the DE loop in MACAM-1 binding to activated ␣ 4 ␤ 7 , we examined the impact of DE loop deletion on the interaction between MAdCAM-1 and ␣ 4 ␤ 7 activated by talin and SDF-1␣. Opposite to the partial rescued cell adhesion to ⌬DE MAdCAM-1 after ␣ 4 ␤ 7 was activated by Mn 2ϩ , the activation of ␣ 4 ␤ 7 by either talin or SDF-1␣ did not rescue the abolishment of cell adhesion by DE loop deletion in 1 mM Ca 2ϩ ϩ 1 mM Mg 2ϩ (Fig. 5). Thus, the intact DE loop is important for MAdCAM-1 interaction with both low-affinity and high-affinity integrin ␣ 4 ␤ 7 activated by talin or SDF-1␣. On the other hand, ␣ 4 ␤ 7 activated by Mn 2ϩ could support decent cell adhesion to MAdCAM-1 in the absence of the intact DE loop, suggesting that the conformation of Mn 2ϩ -activated ␣ 4 ␤ 7 may be different from those of the low-affinity and high-affinity ␣ 4 ␤ 7 activated by more physiological stimuli. The overexpression of GFP-talin-head augmented the firm adhesion of ␣ 4 ␤ 7 293T transfectants on both WT and DE loop single residue mutant MAdCAM-1 (Fig. 5A). In contrast, SDF-1␣ stimulation increased the PBL adhesion only to the WT MAdCAM-1, but not to the DE loop mutants (Fig. 5B). These data suggest that the residues in the DE loop might be involved in distinguishing the subtle difference between ␣ 4 ␤ 7 activated by talin and ␣ 4 ␤ 7 activated by SDF-1␣.
The DE Loop Is Required for the Stable Interaction between MAdCAM-1 and Low-affinity Integrin ␣ 4 ␤ 7 -Next, we investigated the function of the DE loop in the strength of ␣ 4 ␤ 7 -mediated cell adhesion to MAdCAM-1 (Fig. 6). In 1 mM Ca 2ϩ ϩ 1 mM Mg 2ϩ , the DE loop mutations significantly decreased the shear resistance of adherent cells bearing low-affinity ␣ 4 ␤ 7 (Fig.  6A). In contrast, the same mutations showed little effect on the stability of adhesion mediated by high-affinity ␣ 4 ␤ 7 activated by Mn 2ϩ , talin, or SDF-1␣ (Fig. 6, B-D). Thus, the residues in the DE loop of MAdCAM-1 are important for stabilizing the interaction between low-affinity ␣ 4 ␤ 7 and MAdCAM-1.

DISCUSSION
Lymphocyte homing to gut is dependent on the interaction between integrin ␣ 4 ␤ 7 and MAdCAM-1. The resting (low-affinity) and activated (high-affinity) integrin ␣ 4 ␤ 7 can mediate rolling and firm adhesion of lymphocytes, respectively, which are two of the critical steps in lymphocyte homing. Previous studies have shown that integrin undergoes global and local conformational changes upon activation, resulting in the distinct conformations of low-affinity and high-affinity integrins. Thus, it is tempting to speculate that the low-affinity and highaffinity ␣ 4 ␤ 7 binds MAdCAM-1 differently, which might play a fundamental role in supporting the rolling and firm cell adhesion. The integrin ␣ 4 ␤ 7 -MAdCAM-1 interaction is dependent on a conserved acidic peptide motif in the first Ig-like domain of MAdCAM-1, which is present as a surface-exposed structure. The Asp-42 in this motif forms the primary interaction with the divalent cation at ␤ 7 MIDAS. Because the primary interaction between Asp-42 and the MIDAS metal ion is shared by both low-affinity and high-affinity ␣ 4 ␤ 7 -MAdCAM-1 binding, there should be other interactions between MAdCAM-1 and ␣ 4 ␤ 7 that determine rolling or firm adhesion.
Although previous studies have revealed that some residues in MAdCAM-1 are important for MAdCAM-1-␣ 4 ␤ 7 binding (9,12,13,15), the structural basis for supporting MAdCAM-1-␣ 4 ␤ 7 -mediated rolling and firm cell adhesion remains elusive because the static cell adhesion assay used in those studies is unable to distinguish rolling and firm cell adhesion. In this study, we used a flow chamber assay to screen the critical residues in MAdCAM-1, which are important for supporting rolling and firm cell adhesions, respectively. Our results demonstrate that the CCЈ and DE loops play distinct roles in the recognition of MAdCAM-1 by low-and high-affinity ␣ 4 ␤ 7 and suggest that the inactive ␣ 4 ␤ 7 and ␣ 4 ␤ 7 activated by different stimuli have distinct conformations with different structural requirements for MAdCAM-1 binding (Fig. 7).
The Asp-42 in the CCЈ loop is required for both low-affinity ␣ 4 ␤ 7 -mediated rolling adhesion and high-affinity ␣ 4 ␤ 7 -mediated firm adhesion. In addition to Asp-42, most other CCЈ loop residues other than Arg-39 and Ser-44 are essential for the lowaffinity ␣ 4 ␤ 7 -MAdCAM-1 interaction, suggesting the potential binding sites at the CCЈ loop for low-affinity ␣ 4 ␤ 7 . In contrast, the same CCЈ loop mutations only slightly decreased cell adhesion mediated by Mn 2ϩ -activated ␣ 4 ␤ 7 , suggesting the different inter-