Overlapping, but not identical, sites are involved in the recognition of C3bi, neutrophil inhibitory factor, and adhesive ligands by the alphaMbeta2 integrin.

The αMβ2 (CD11b/CD18, Mac-1) integrin receptor binds numerous ligands, including neutrophil inhibitory factor (NIF), C3bi, and certain immobilized protein substrates, represented by denatured ovalbumin. These ligands share no obvious structural similarities, yet their interactions with receptor are inhibited by NIF and involve the I domain, a stretch of ∼200 amino acids in the αM subunit. Recombinant wild-type and mutant forms of αMβ2 have been used to compare the recognition requirements of these ligands. The various constructs were expressed efficiently on the surface of human embryonic kidney 293 cells and formed α·β heterodimeric complexes. The wild-type transfectants bound the three ligands in a similar fashion to naturally occurring αMβ2. NIF inhibited these interactions, and deletion of the D248PLGY from within the I domain abolished binding of all three ligands, suggesting an overlapping recognition specificity. A single point mutation of Ser138 to Ala in the β2 subunit abolished C3bi binding and cell adhesion but did not affect NIF binding. A switch of the R281QELNTI sequence in helix 6 of the αM I domain to the corresponding sequence in the I domain of the αL (QETLHKF) subunit completely abrogated adhesion while not affecting C3bi and NIF binding. The two mutant receptors also did not support activation-dependent adhesion to fibrinogen. Thus, the contact sites for NIF, C3bi, and adhesive proteins, represented by denatured ovalbumin and fibrinogen, in αMβ2 are overlapping but not identical.

We (21) and others (22) have shown that the I domain of ␣ M ␤ 2 (␣ M I-domain) is capable of binding NIF. I(A) domains, inserted domains of ϳ200 amino acids, are found in several integrin ␣-subunits as well as in other proteins and have been implicated in mediating a variety of protein-protein interactions, including ligand binding to integrins. Recently, the structure of ␣ M I domain has been solved by x-ray crystallography (23) and was shown to contain a novel cation binding motif. Mutations of residues involved in coordination of cations within the ␣ M I domain destroyed C3bi binding to the ␣ M ␤ 2 receptor (24). Using blocking mAbs which map to the I domain or recombinant ␣ M I domain itself, this region of ␣ M ␤ 2 has been implicated in the binding of intracellular adhesion molecule-1, Fg (25), and C3bi (20), as well as NIF. Taken together, these data suggest that the ␣ M I domain is an independent structural unit, capable of interacting with many different proteins. Nevertheless, the ligand binding functions of ␣ M ␤ 2 are not merely a property of its I domain. Bajt et al. (26) have shown that mutations of Asp 134 or Ser 136 in ␤ 2 abrogated the binding of ␣ M ␤ 2 to C3bi, suggesting that, in addition to I domain, portions of the ␤ subunit also are involved in ligand recognition by ␣ M ␤ 2 .
In this study, we examined the structural requirements for ␣ M ␤ 2 to bind to C3bi, NIF, and representative protein-coated surfaces. Based upon the ligand binding properties of recombinant wild-type ␣ M ␤ 2 and selected mutants, we conclude that the contact regions in ␣ M ␤ 2 for each of these three ligands are overlapping but not identical. These analyses demonstrate the complexity of ligand binding functions of ␣ M ␤ 2 and identify specific regions of the receptor involved in selective recognition of individual ligands. Site-directed Mutagenesis-All mutations were created by oligonucleotide-directed mutagenesis using uracil-containing single stranded M13mp18 DNA (27). Two unique restriction sites, ClaI at position 535 (Ile 139 ), and NheI, at position 1141 (Ala 332 ), flanking the I domain of ␣ M , were introduced into ␣ M using the following mutagenic primers, 5Ј-CAGAGCCATCgATgAgGAAGGCAATGTCACTATC-3Ј for ClaI (the underlined nucleotides are the ClaI recognition sequence and lower case nucleotide for the mismatched base) and 5Ј-AGAGGTGATGctAGCGCT-GAAGCCTTC-3Ј for NheI. No amino acid residue was changed by the former mutation, whereas Ala 332 (at nucleotide position 1141) was changed to Ser, which is present in mouse ␣ M . This mutation had no effect on binding of the three ligands of interest by comparison to wild-type ␣ M ␤ 2 (data no shown). To mutate R 281 QELNTI in helix 6 (nomenclature according to the published crystal structure (23)) in the I domain of ␣ M to its counterpart in ␣ L (QETLHKF), a 51-mer primer 5Ј-AGGCGGCTTGGATGCGAattTgTgAAGtgtTTcctGGGATTTCTCA-CTGCG-3Ј was used, and resultant plaques were screened with XmnI (GAANNNNTTC). This mutant I domain was transferred back into ␣ M using the ClaI and NheI sites, creating the ␣ M mutant, ␣ M (H6). The Ser 138 to Ala mutation of ␤ 2 subunit, ␤ 2 (S138A), was prepared in a similar fashion, using the single-stranded DNA of ␤ 2 in M13mp19 and the following mutagenic primer, 5Ј-CTCTCCTACgCCATGCTTGA-3Ј. To create ␣ M (⌬D 248 PLGY), the primer 5Ј-AGGGATGACATCCTCGC-CAAACTTTTCTCC-3Ј (30-mer) was used to loop out the five amino acid residues. DNA sequencing of the entire I domain was conducted confirming the presence of the mutations and the correctness of the rest of the I domain. The cDNA of ␣ M and ␤ 2 were inserted separately in the pCIS2M expression vector employing XbaI and XhoI sites; expression of ␣ M or ␤ 2 was under the control of the human cytomegalovirus promoter and enhancer.

Materials-Human
Expression of ␣ M ␤ 2 in 293 Cells-The expression vectors containing wild-type and mutated ␣ M (pCIS2M-␣ M ) and ␤ 2 (pCIS2M-␤ 2 ) were purified using CsCl gradients and transfected, together with pRSVneo (neomycin-resistant gene), into 293 cells according to our established procedures (27). G418 (600 g/ml)-resistant colonies were pooled, and ␣ M ␤ 2 -expressing cells were sorted by FACS (FACStar, Becton-Dickinson, San Jose, CA), using ␣ M -specific mAb, OKM1, which recognizes an epitope outside of the I domain (28). To generate a ␤ 2 (S138A) cell line that has equivalent expression level as the wild-type ␣ M ␤ 2 receptor, a established cell line of ␤ 2 (S138A) receptor was re-sorted with OKM1, and the cell population that has the same fluorescent intensity as the wild-type was pooled and cultured.
Immunoprecipation Analyses-␣ M ␤ 2 -expressing 293 cells were 125 Isurface-labeled with lactoperoxidase according to a published method (29). The cells were solubilized with 20 mM Tris, pH 7.4, containing 150 mM NaCl, 1% Triton X-100, 1 mM phenylmethylsulfonyl fluoride, 1 mM CaCl 2 , 10 mM benzamidine, 10 mM leupeptin, and 2.5 g/ml soybean trypsin inhibitor. After 30 min at room temperature, cell lysates were centrifuged at 14,000 rpm in an Eppendorf microcentrifuge for 10 min at 4°C, and the supernatants were removed for immunoprecipitation. Lysates from 10 6 cells were incubated with 10 g of normal mouse IgG and 50 l Zysorbin-G (Zymed Laboratory, San Francisco, CA) for 2 h at 4°C. After centrifugation, the supernatant was mixed with 10 g of mAb and kept at 4°C overnight. The ␣ M ␤ 2 heterodimer was captured by incubating with 50 l of Protein G-Sepharose (Pharmacia Biotech Inc.) for 2 h at 4°C. The resin was washed three times with TBS (20 mM Tris, 150 mM NaCl, pH 7.4), containing 1 mM Ca 2ϩ , and eluted with SDS-PAGE loading buffer (50 mM Tris, pH 6.8, 10% glycerol, 1% SDS, 0.1 mg/ml bromphenol blue). Immunoprecipitates were analyzed on 7% SDS-PAGE, and the gels were dried and then exposed to Kodak XAR-5 x-ray film for 48 h at 22°C.
NIF Binding to ␣ M ␤ 2 -expressing Cells-125 I-labeled NIF (0.05 g), radiolabeled as described previously (21), was added to 200 l of 10 6 transfected cells in the presence or absence of 1 g of unlabeled NIF for 30 min at 22°C. The buffer employed was HBSS containing 2.5 mM Ca 2ϩ . Binding reactions were terminated by centrifugation of the cell suspension through a 20% sucrose solution at 12,000 ϫ g for 5 min. Radioactivity in the cell pellet was measured with a gamma counter. Each data point is the average of two independent experiments. Dissociation constants, K d , were calculated from regression analysis of nonlinear least squares fits of the binding data using a program in Sig-maPlot (Jandel Co., San Rafael, CA), and the equation, C3bi Binding to ␣ M ␤ 2 -expressing Cells-C3bi binding was performed with slight modification of the method of Bilsland (30). Sheep erythrocytes coated with C3bi (EC3bi) were prepared using anti-sheep erythrocyte IgM antibody, M1/87 (ATCC, Rockville, MD), and human C5deficient serum (Sigma). Briefly, 7 ϫ 10 8 sheep erythrocytes (Colorado Serum Company, Denver, CO) were washed twice in HBSS, containing 5 mM HEPES and 1 mM Mg 2ϩ , and coated with IgM as described (30). The coated erythrocytes were surface-labeled with biotin using 1 mg of sulfosuccinimidyl 6-(biotinamido) hexanoate (Pierce) at 37°C for 20 min. The biotinylated cells were resuspended in 0.9 ml of HBSS with 5 mM HEPES, 1 mM Ca 2ϩ , and 1 mM Mg 2ϩ , mixed with 100 l of C5deficient serum, and incubated at 37°C for 60 min. After washing twice, the resulting EC3bi were resuspended in 10 ml of the above solution.
To perform the EC3bi binding assays, a total of 2 ϫ 10 5 ␣ M ␤ 2expressing cells were seeded onto polylysine (50 g/ml)-coated 24-well non-tissue culture polystyrene plates (Becton Dickinson, Franklin Lakes, NJ) for 15 min at 37°C, followed by addition of 2 ϫ 10 7 EC3bi. After 60 min at 37°C, unbound EC3bi were removed by washing with PBS. Bound EC3bi were fixed with 2% paraformaldehyde overnight, and excess paraformaldehyde was neutralized with 1% bovine serum albumin at 37°C for 2 h. Bound EC3bi were quantitated by addition of 300 l of avidin-alkaline phosphatase conjugate (1:2000 dilution) (Zymed Laboratory, San Francisco, CA) at 37°C for 90 min, followed by washing three times with PBS, and addition of 250 l of 3 mg/ml p-nitrophenyl phosphate. After 10 min incubation at 37°C, the absorbance at 405 nm was determined.
Adhesion of ␣ M ␤ 2 -expressing Cells to Denatured Ovalbumin, Fibrinogen, or NIF-Denatured ovalbumin (dOva) was prepared according to the method of Davis (15). Briefly, ovalbumin was denatured with 8 M urea, reduced with 10 mM dithiothreitol, and alkylated with 60 mM iodoacetamide. The mixture was dialyzed extensively against PBS and frozen at Ϫ80°C until use. For adhesion assay, 100 l of 20 g/ml dOva, 50 g/ml Fg, or 200 g/ml NIF was deposited at the center of wells of 24-well non-tissue culture polystyrene plates. After 90 min at 22°C, 400 l of blocking buffer (0.05% polyvinylpyrrolidone in PBS, which effectively prevented adhesion of ␣ M ␤ 2 -transfected and mock-transfected 293 cells to noncoated plates) was added. After washing twice with PBS, 2 ϫ 10 6 cells in 400 l of HBSS containing 5 mM HEPES, 1 mM CaCl 2 , and 1 mM MgCl 2 were added to each well. The plate was incubated at 37°C for 25 min in humidified CO 2 incubator, followed by three washes with PBS. To measure the number of adherent cells, 150 l of lysis buffer (50 mM sodium acetate, pH 5.0, 1% Triton X-100) and 3 mg/ml p-nitrophenyl phosphate was added and incubated at 37°C for 1 h. Substrate cleavage by the cellular acid phosphatase was stopped by addition of 50 l of 1 N NaOH, and the absorption was measured at 405 nm with a microplate reader (Molecular Devices, Menlo Park, CA).

RESULTS
Wild-type Recombinant ␣ M ␤ 2 Is Similar to Naturally Occurring ␣ M ␤ 2 -To examine the ligand binding functions of ␣ M ␤ 2 , the receptor was expressed as a recombinant protein in 293 cells. The wild-type recombinant and naturally occurring ␣ M ␤ 2 behaved similarly by the following criteria. First, the wild-typetransfected but not mock-transfected cells bound to NIF, EC3bi, and dOva-coated surfaces. These represent known ligands for ␣ M ␤ 2 on neutrophils and ␣ M ␤ 2 -bearing cell lines (15,19). Data for NIF binding are shown in Fig. 1A. Whereas mock-transfected cells failed to bind this ␣ M ␤ 2 -specific ligand, the cell transfected with wild-type ␣ M ␤ 2 bound NIF in a specific and saturable manner. The binding isotherm in Fig. 1A yielded a K d of 7 nM with a B max of 4.2 ϫ 10 5 sites per cell. This K d value is similar to that for NIF binding to neutrophils, 5 nM (19). Second, the activating antibody KIM185 enhanced the ligand binding functions of the ␣ M ␤ 2 transfectants as it does for ␣ M ␤ 2 on neutrophils (31). As shown in Fig. 1B, adhesion of the transfectants to Fg was markedly enhanced by KIM185 in an ␣ M ␤ 2 -dependent manner, i.e. adhesion was blocked by mAb 2LPM19c (Fig. 1B), specific for ␣ M , as well as by mAb MHM23 (see Fig. 5B), specific for ␤ 2 . Adhesion to dOva also was ␣ M ␤ 2mediated although maximal interaction did not require stimulation with mAb KIM185. Third, the recombinant wild-type ␣ M ␤ 2 formed a heterodimeric complex. Shown in Fig. 2, lane 1, is the immunoprecipation of cell surface-labeled wild-type ␣ M ␤ 2 transfectants. The estimated molecular masses for the recombinant ␣ M and ␤ 2 subunits were 165 and 95 kDa, similar to those of the naturally occurring subunits.
The Recognition Sites for NIF, C3bi, Fg, and dOva Are Overlapping-Evidence that the recognition of NIF, EC3bi, Fg, and dOva by ␣ M ␤ 2 is overlapping is derived from two observations. First, NIF inhibited the binding of all three test ligands to the wild-type ␣ M ␤ 2 transfectants. At a concentration of 10 g/ml, NIF not only blocked the binding of 125 I-NIF to the ␣ M ␤ 2transfectants but also blocked the interactions of these cells with EC3bi and dOva by Ͼ90%. Second, the mutant ␣ M ␤ 2 , in which residues D 248 PLGY in the ␣ M I-domain were deleted, failed to bind all three ligands. Compared with wild-type, NIF, EC3bi binding and adhesion to dOva was reduced by 92, 91, and 97%, respectively. This mutant is very similar to one recently described by McGuire and Bajt (32) in which residues F 245 GDPLGY were deleted. These investigators demonstrated that this mutant maintained native ␣ M ␤ 2 conformation as it retained the capacity to bind a conformational-sensitive mAb. Taken together, these results suggest binding of these three ligands to ␣ M ␤ 2 has common structural requirements.
Nonidentical Binding Sites for NIF, C3bi, Fg, and dOva in ␣ M ␤ 2 -Evidence that the recognition requirements for these three ␣ M ␤ 2 ligands are distinct was derived from the characterization of two other mutant ␣ M ␤ 2 receptors. In the first mutant, designated ␤ 2 (S138A), Ser 138 residue in ␤ 2 subunit was changed to Ala, and, in the second mutant, designated ␣ M (H6), a segment in helix 6 of the I domain of ␣ M (R 281 QELNTI) was switched to its counterpart sequence in the I domain of ␣ L (QETLHKF). Both mutants were stably expressed in 293 cells. By FACS analysis with mAbs to the ␣ M subunit (OKM1, LM2/1, 2LPM19c, 44, M1/70) or ␤ 2 (TS1/18, MHM23) subunits, wild-type and the two mutants were expressed at similar levels on the cell surface; the mean fluorescent intensities with OKM1 for mutant ␤ 2 (S138A) was the same as for wild-type and was 2-fold higher for ␣ M (H6). Immunoprecipitations of 125 I-surface-labeled cells are shown in Fig.  2. OKM1, a mAb to the ␣ M subunit, precipitated bands of similar mobility on SDS-PAGE from the two mutant ␣ M ␤ 2 transfectants and from cells expressing the wild-type receptor. TS1/18, a mAb to the ␤ 2 subunit, also immunoprecipitated both subunits. Similar results also were obtained with mAbs LM2/1 to the ␣ M subunit and MHM23 to the ␤ 2 subunit (data not shown). Taken together, the FACS and immunoprecipitation data indicate that both mutant receptors were cell-surfaceexpressed at similar levels and formed heterodimeric complexes.  (lanes 1-3), TS1/18, a mAb to ␤ 2 (lanes 4 and 5), or IV.3, an irrelevant mAb (lane 6), overnight at 4°C. After washing, the immunoprecipitates were subjected to SDS-PAGE (7% gels under nonreducing conditions) and exposed to Kodak XAR-5 film for 48 h . Lanes 1, 4, and 6, wild-type; lane 2, ␤ 2 (S138A); lanes 3 and 5, ␣ M (H6).
As shown in Fig. 3A, both mutants bound NIF similarly to wild-type ␣ M ␤ 2 . With each transfectant, a 20-fold excess of unlabeled NIF inhibited 125 I-NIF binding by more than 99%. The interactions of all three ␣ M ␤ 2 transfectants with NIF depended upon divalent cations, as addition of 1 mM EDTA completely abolished their NIF binding activities. NIF binding was not affected by MHM23, a mAb to the ␤ 2 subunit which is subsequently shown to block cell adhesion to dOva and Fg (see below). At a concentration of 10 g/ml, 125 I-NIF binding to all three ␣ M ␤ 2 transfectants was inhibited by less than 2%. Moreover, NIF, at a concentration as high as 175 nM, did not inhibit binding of this mAb to the wild-type and mutant cells as assessed by FACS (not shown). NIF binding isotherms, similar to those shown in Fig. 1A, were constructed with each cell line. The K d values derived from these analyses are summarized in Table I. These values were in a similar range although the affinity of the ␣ M (H6) mutant for NIF was slightly lower than those of the other two mutants. The preservation of NIF binding function to the wild-type and mutant receptors also was demonstrable when NIF was presented as an immobilized ligand. When NIF was deposited onto a plastic surface, and cell adhesion was measured as described above for dOva and Fg, the two mutant receptors supported adhesion as well as wildtype ␣ M ␤ 2 (Fig. 3B). The adhesion of the cells to NIF was completely inhibited by 10 g/ml soluble NIF and was not affected by mAb MHM23. Two conclusions can be drawn from these results. 1) The mutations, ␤ 2 (S138A) and ␣ M (H6), did not affect the structural integrity of ␣ M ␤ 2 as both mutants maintained a conformation compatible with high affinity NIF binding, whether NIF was presented as a soluble ligand or as an adhesive substrate. 2) The NIF binding site does not involve the mutated areas of ␣ M or ␤ 2 .
EC3bi binding to the two mutant receptors is shown in Fig.  4. Mutation of the H6 segment did not alter recognition of this ligand. Several controls were performed to verify the specificity of the EC3bi interaction with this mutant as well as wild-type ␣ M ␤ 2 . No detectable binding to wild-type or the ␣ M (H6) was observed with uncoated erythrocytes or IgM-coated erythrocytes; mock-transfected 293 cells did not rosette with EC3bi. As anticipated, EC3bi binding to both ␣ M (H6) and wild-type cell lines required divalent cations, as no binding was observed in the presence of 1 mM EDTA (Fig. 4). In contrast, the ␤ 2 (S138A) transfectant lost more than 80% of its EC3bi binding activity. Thus, this mutant bound NIF, but not C3bi, indicating that the binding requirements for these two ligands are distinct.
The adhesion of wild-type and the two mutant receptors to dOva is shown in Fig. 5A. Wild-type ␣ M ␤ 2 readily supported adhesion to dOva-coated surfaces in a divalent cation-dependent interaction, as inclusion of 1 mM EDTA abolished adhesion by 99%. Addition of 10 g/ml NIF or 5 g/ml MHM23 blocked the cell adhesion by more than 95%. Mutation of the H6 segment of ␣ M or Ser 138 of ␤ 2 , however, completely abrogated the adhesive activity of the receptors. A similar pattern was observed when the adhesive substrate was Fg. As shown in Fig.  5B, wild-type ␣ M ␤ 2 adhered to Fg upon activation with KIM185, and this adhesion was blocked by mAb MHM23 or NIF. However, the ␣ M H6 and ␤ 2 Ser 138 transfectants failed to adhere to Fg. In separate experiments, we verified that the activating mAb still bound to these mutant ␣ M ␤ 2 -bearing cells by FACS. Thus, regions in both the ␣ M and ␤ 2 subunits are critical for cell adhesion to dOva and Fg, and the ␣ M (H6) mutant, which was capable of binding NIF and C3bi, could not adhere to either substratum. DISCUSSION One of the defining characteristics of integrins is the capacity of its family members to bind multiple ligands (1). ␣ M ␤ 2 has one of the broadest ligand repertoires, recognizing no fewer than 12 ligands. To date, analyses of the ligand binding functions have emphasized that many ␣ M ␤ 2 ligands share common binding requirements. However, as shown in this study, certain mutations, both in the I domain and in ␤ 2 , can selectively inactivate the recognition of individual ligands. Thus, we propose that the binding sites for the ligand analyzed (NIF, EC3bi, and the adhesive substrates, Fg and dOva) are not identical.
Two regions of ␣ M ␤ 2 have been implicated in providing the overlapping ligand contact sites, the I domain of ␣ M and the ␤ 2 134 -138 region. The role of the I domain has been demon-   (20 -22, 25, 33-35), that several function blocking mAbs map to the I domain (28), and that NIF blocks intracellular adhesion molecule-1 (22) and EC3bi binding (20). The crystal structure of the I domain shows that cation coordinating residues are provided from within noncontiguous loops, which connect ␣ helices and ␤ strands (23). As cations are intimately involved in the binding of most ligands to integrins (36), sequences within these loops also may provide several of the overlapping as well as ligand-selective contact sites. While the ␤ 2 134 -138 is not required for NIF binding, the single point mutation of S138A abolished EC3bi, Fg, and dOva binding. Bajt et al. (26) had previously shown that Asp 134 and Ser 136 of ␤ 2 are required for C3bi binding to ␣ M ␤ 2 , and these mutations in our study also block ␣ M ␤ 2 -mediated cell adhesion to dOva (data not shown). In contrast to their experience with the S138A mutation in COS cells, we were able to express this mutant (as well as S134A and S136A) on the cell surface of 293 cells. Therefore, this ␤ 2 region, which also has been implicated in ligand and cation binding to ␤ 3 integrins (37), is involved in binding multiple ligands to ␣ M ␤ 2 .
The nonidentical nature of the binding sites for NIF, EC3bi, and adhesive ligands (Fg and dOva) is demonstrated by two of the recombinant ␣ M ␤ 2 receptors analyzed. ␤ 2 (S138A) bound NIF with a similar affinity as wild-type receptor but failed to bind EC3bi, Fg, or dOva. Hence, the NIF binding site cannot be identical to those of the other three ligands. NIF binds to the expressed I domain of ␣ M with high affinity, and it may be that the entire binding pocket for NIF resides in the ␣ M I domain.
The ␣ M (H6) mutant bound EC3bi and NIF but not Fg or dOva. Thus, the binding sites for Fg and dOva cannot be identical to those of the other two ligands. In the crystal structure, helix 6 is placed at some distance from the cation binding site in the ␣ M I domain. Nevertheless, ␣ M ␤ 2 -mediated adhesion to Fg or dOva is still divalent cation-dependent. Either the binding pocket for Fg and dOva extends over a broad region of the I domain to include both helix 6 and the cation binding site or one of the other cation binding sites (26) is required for interaction with these ligands.
To explain these data, we hypothesize that the binding sites for the ligands analyzed are overlapping but not identical. An alternative explanation of our data is that NIF, C3bi, and adhesive ligands share the same binding site but interact with different affinities. This possibility seems unlikely, since 1) the mutant ␤ 2 (S138A) bound NIF with the same, if not higher affinity than the wild-type ␣ M ␤ 2 ( Table I), suggesting that the mutations did not grossly alter the affinity of ␣ M ␤ 2 toward its ligands. Therefore, the failure of ␤ 2 (S138A) to interact with C3bi, dOva, and Fg cannot simply be caused by a decrease in binding affinity. Moreover, the ␤ 2 (S138A) mutant still bound NIF when it was presented in the same format, as an immobilized substrate, as dOva or Fg. 2) If these ligands all shared the same binding site, the same mAb should block interactions with all ligands. This was not the case. A mAb specific for ␤ 2 subunit, MHM23, blocked ␣ M ␤ 2 interactions with C3bi (38), Fg, and dOva but had no effect on NIF binding.
In summary, at least three functionally distinct sites in ␣ M ␤ 2 mediate adhesion and its interactions with C3bi and NIF. One site in ␤ 2 , which includes Asp 134 , Ser 136 , and Ser 138 , is required for C3bi binding and adhesion to protein-coated surfaces. The second site, located in helix 6 of ␣ M I domain, is important for adhesion to Fg and dOva but not for C3bi and NIF binding. A third class of contact site, also in I domain, is shared by all three ligands, thereby accounting for the ability of NIF to block ␣ M ␤ 2 binding of the other test ligands. Thus, ␣ M ␤ 2 can be viewed as a mosaic in which certain regions of the receptor provide common recognition sites for NIF, EC3bi, Fg, and FIG. 4. EC3bi binding to wild-type and mutant ␣ M ␤ 2 transfectants. Biotinylated EC3bi (2 ϫ 10 7 ) were added to 2 ϫ 10 5 ␣ M ␤ 2expressing cells that were preseeded onto polylysine-coated 24-well plates. After 60 min at 37°C, bound EC3bi were determined using avidin-alkaline phosphatase and p-nitrophenyl phosphate, measuring the absorbance at 405 nm. The value for wild type was taken as 100%.  ). B, transfected cells (2 ϫ 10 6 ) were added to each well in the presence of buffer (solid bar) or 5 g/ml KIM185 (open bar), 5 g/ml KIM185 plus 5 g/ml MHM23 (hatched bar), 5 g/ml KIM185 plus 10 g/ml NIF (crossed bar). For both substrates, the nonadherent cells were removed by three washes with PBS, and the adherent cells were quantitated by measuring acid phosphatase activity. Each assay was performed in duplicate, and the data are representative of 2-3 independent experiments. dOva, whereas other regions mediate selective recognition of these ligands. As these molecules represent three general categories-soluble (NIF), cell-surface (EC3bi), and matrix-deposited (Fg and dOva)-of ␣ M ␤ 2 ligands, this model may apply to other ␣ M ␤ 2 ligands.