CD151, a member of the tetraspanin family of proteins, forms a stable complex with integrin α3β1 and regulates integrin-mediated cell-substrate adhesion. However, the molecular basis of the stable association of CD151 with integrin α3β1 remains poorly understood. In the present study, we show that a panel of anti-human CD151 mAbs (monoclonal antibodies) could be divided into three groups on the basis of their abilities to co-immunoprecipitate integrin α3: Group-1 mAbs were devoid of sufficient activities to co-precipitate integrin α3 under both low- and high-stringency detergent conditions; Group-2 mAbs co-precipitated integrin α3 under low-stringency conditions; and Group-3 mAbs exhibited strong co-precipitating activities under both conditions. Group-1 mAbs in particular exhibited increased reactivity toward integrin α3β1-unbound CD151, indicating that the binding sites for Group-1 mAbs are partly blocked by bound integrin α3β1. Epitope mapping using a series of CD151 mutants with substitutions at amino acid residues that are not conserved between human and mouse CD151 revealed that Gly176/Gly177, Leu191 and Gln194 comprise epitopes characteristic of Group-1 mAbs. Replacement of short peptide segments, each containing one of these epitopes, with those of other tetraspanins lacking stable interactions with integrin α3β1 demonstrated that the segment from Cys185 to Cys192, including Leu191, was involved in the stable association of CD151 with integrin α3β1, as was the Gln194-containing QRD peptide. Taken together these results indicate that two consecutive segments including two Group-1 epitopes, Leu191 and Gln194, comprise an interface between CD151 and integrin α3β1, and, along with the epitope including Gly176/Gly177, are concealed by bound integrin.
- basement membrane
- cell adhesion
- integrin α3β1
The tetraspanin superfamily consists of at least 32 members in mammals, including CD9, CD63, CD81, CD82 and CD151 [1–6]. Tetraspanins are present in different combinations on almost all types of cell and tissue, and have been implicated in diverse cellular functions involving cell–cell and cell–substratum interactions. Tetraspanins associate with each other and with other transmembrane proteins, for example, integrins and immunoglobulin superfamily proteins, forming multimolecular membrane microdomains, often referred to as a ‘tetraspanin web’ or ‘tetraspanin-enriched microdomain’ [4,7–9]. Tetraspanins possess four transmembrane domains, a small extracellular loop (EC1), a larger extracellular loop (EC2), a small intracellular loop and N- and C-terminal short cytoplasmic tails [4,10]. EC2 is located between the third and fourth transmembrane domains, and contains a conserved CCG motif and two other conserved cysteine residues, which together form two intramolecular disulfide bonds. Some tetraspanins, for example, CD63 and CD151, possess two additional cysteine residues that introduce an extra disulfide bond. EC2 is divided into constant and variable regions on the basis of conservation of secondary structures [11,12]. The variable region is a site of hypervariability among family members, and is thought to mediate specific protein–protein interactions [3,10].
CD151 forms a stable complex with the laminin-binding integrins α3β1, α6β1, α6β4 and α7β1 to regulate their functions, including cell attachment, spreading and migration [13,14]. Among these interactions, association of CD151 with integrin α3β1 is especially strong and highly stoichiometric . Indeed, nearly all integrin α3β1 is bound by CD151, and no integrin α3β1 has been found in cell/tissue types that do not express CD151 [13,15]. However, CD151 is not required for cell-surface expression of integrin α3β1, as shown by studies with CD151-knockout mice and CD151-knocked-down cells [16–18]. The molecular basis of the stable interaction between CD151 and integrin α3 remains unclear, although it has been proposed that a stretch consisting of three amino acid residues (Gln194–Asp196) in the variable region of EC2 is involved in the stable association of CD151 with integrin α3β1 .
The biological functions of CD151–integrin complexes have been addressed by experiments using a CD151 mutant that is incapable of binding to integrins: the CD151 mutant inhibited integrin α3- and α6-dependent cable formation in COS7 cells on Matrigel . A recent in vivo study showed that CD151-knockout mice exhibit strain-dependent renal defects similar to those found in podocyte-specific conditional knockout mice for integrin α3 . We previously found that an anti-CD151 mAb (monoclonal antibody), 8C3, has the ability to dissociate CD151 from integrin α3β1, thereby attenuating the binding activity of integrin α3β1 to laminin-10/11 (511/521) . These observations indicate the functional importance of the interactions between CD151 and integrins.
mAbs with defined epitopes are valuable probes for dissecting structure–function relationships among proteins. Many mAbs against CD151 have been generated and used as probes for CD151 functions and its associations with integrins [13,14,19,21–26]. The associations of CD151 with integrins are responsible, at least in part, for the different patterns of reactivity for CD151 observed with different mAbs, in cells and tissues [14,22,23]. TS151r, one of the best-characterized anti-CD151 mAbs, has been shown to bind only to CD151 that is not occupied by integrin α3β1 . The epitopes of TS151r are thought to be masked by bound integrin α3β1 and to overlap with an integrin-binding site . We previously reported that the mAb 8C3 is capable of dissociating the CD151–integrin α3β1 complex, whereas another mAb, 11G5, is not . In the present study, we performed epitope mapping of ten mouse anti-human CD151 mAbs, including 8C3, 11G5 and TS151r. We also demonstrate that these mAbs showed distinct capabilities of co-immunoprecipitating integrin α3. In addition, we examined their reactivities toward integrin α3-knocked-down and integrin α3-overexpressing cells, demonstrating that a group of mAbs that showed little activity to co-immunoprecipitate integrin α3 exhibited increased reactivities toward CD151 uncomplexed with integrin α3. We also explored the binding sites of CD151 to integrin α3β1 by mutational analysis, to correlate the reactivities of the mAbs toward the CD151–integrin α3β1 complex with their epitope maps.
Cell culture, antibodies and cDNA
A549 human lung adenocarcinoma, NIH 3T3 mouse fibroblast and HEK (human embryonic kidney)-293T cells were maintained in 10-cm dishes in DMEM (Dulbecco's modified Eagle's medium; Sigma) supplemented with heat-inactivated 10% (v/v) FBS (fetal bovine serum; JRH Biosciences). Cells were cultured at 37 °C in a humidified atmosphere containing 5% CO2.
mAbs against human CD151 (8C3), human integrin α3 (3G8) and human integrin α5 (8F1), and a pAb (polyclonal antibody) against human integrin α3, were produced as described previously [21,27–29]. The anti-human CD151 mAbs SFA1.2B4, TS151, TS151r, LIA1/1, VJ1/16, 11B1 and 14B5 have also been described previously [9,14,23,24,30]. 11G5 and 14A2.H1 were purchased from Serotec and BD Pharmingen respectively. Anti-actin pAb and R-phycoerythrin-conjugated anti-mouse IgG antibody were purchased from Sigma. Anti-GFP (green fluorescent protein) mAb JL-8 was obtained from Clontech. Anti-EGFR (epidermal-growth-factor receptor) pAb was obtained from Santa Cruz Biotechnology. Normal mouse IgG was obtained from Santa Cruz Biotechnology or Southern Biotech. Alexa Fluor® 488-conjugated anti-mouse IgG antibody was purchased from Molecular Probes. Peroxidase-conjugated anti-mouse IgG and anti-rabbit IgG antibodies were obtained from Jackson ImmunoResearch.
Construction of the expression vector for human CD151–EGFP (enhanced GFP) fusion protein has been described previously . The expression vector for human integrin α3 was a gift from Dr Tsutomu Tsuji (Hoshi University School of Pharmacy and Pharmaceutical Sciences, Tokyo, Japan) . The point mutations in CD151 were introduced using the Altered Sites II in vitro Mutagenesis System (Promega) or the QuikChange® site-directed mutagenesis kit (Stratagene). The sequences of the primers are listed in Supplementary Tables S1 and S2 (at http://www.BiochemJ.org/bj/415/bj4150417add.htm).
siRNA (small interfering RNA) and DNA transfections
In RNA interference experiments, A549 cells were grown on 10-cm dishes in DMEM containing 10% (v/v) FBS (DMEM/FBS). The cells were transfected with 900 pmol of siRNA using Lipofectamine™ 2000. At 3 days after transfection, cells were subjected to flow cytometry, cell-surface labelling and immunoblot analyses, as described below. The target sequence of integrin α3 siRNA was 5′-GCUACAUGAUUCAGCGCAA-3′. A non-specific control siRNA had the following sequence: 5′-AAUACGCCGGCGGAACUAC-3′.
For exogenous expression of CD151–EGFP fusion proteins for two-colour flow-cytometric analyses, NIH 3T3 cells were grown on six-well plates in DMEM/FBS. Cells were transfected with 1.5 μg of expression vectors using FuGENE™ 6 (Roche), according to the manufacturer's protocol. At 24 h after transfection, cells were subjected to flow-cytometric analysis as described below.
For exogenous expression of integrin α3 for two-colour flow-cytometric analysis with anti-CD151 mAbs, HEK-293T cells were grown on 10-cm dishes in DMEM/FBS. Cells were transfected with 25 μg of integrin α3 or pcDNA3 empty vector together with 2.5 μg of EGFP expression vector using Lipofectamine™ 2000, according to the manufacturer's protocol. At 24 h after transfection, cells were subjected to flow-cytometric analysis as described below.
For exogenous expression of CD151–EGFP fusions and integrin α3 for co-immunoprecipitation assays, HEK-293T cells were cultured on 6-cm dishes in DMEM/FBS. Cells were co-transfected with expression vectors for CD151–EGFP fusion (6 μg) and integrin α3 (10 μg) using Lipofectamine™ 2000, according to the manufacturer's protocol. At 24 h after transfection, cells were subjected to the assay described below.
Flow-cytometric analysis was performed as described previously . For the analyses with integrin α3-knocked-down cells, suspended cells were incubated with normal mouse IgG, anti-integrin α3 mAb or a series of anti-human CD151 mAbs for 30 min on ice. After being washed with PBS, cells were incubated with Alexa Fluor® 488-conjugated anti-mouse IgG antibody for 30 min on ice. For two-colour flow-cytometric analysis of NIH 3T3 cells expressing CD151–EGFP fusion proteins and HEK-293T cells co-expressing integrin α3 and EGFP, cells were incubated with normal mouse IgG, anti-integrin α3 mAb or a panel of anti-human CD151 mAbs for 30 min on ice. Cells were then incubated with R-phycoerythrin-conjugated anti-mouse IgG antibody for 30 min on ice. Stained cells were analysed using a FACScalibur flow cytometer (Becton Dickinson).
Immunoprecipitation and immunoblotting
Cells were washed three times with ice-cold PBS and lysed in a lysis buffer containing 1% (w/v) Triton X-100 or Brij 97 in addition to 5% (v/v) glycerol, 150 mM NaCl, 1 mM EDTA, 20 mM Tris/HCl (pH 7.5), 1 mM PMSF, 5 μg/ml leupeptin and 5 μg/ml aprotinin. Lysates were centrifuged at 36000 g at 4 °C for 20 min; then, the protein concentrations of the clarified lysates were determined using BCA (bicinchoninic acid; Pierce) or Bradford (Bio-Rad) protein assay kits. Equal protein amounts of the clarified lysates were subjected to immunoblotting or immunoprecipitation. For detection of CD151 and CD151–EGFP by immunoblotting, SDS/PAGE was performed under non-reducing conditions; for detection of integrin α3 and actin, SDS/PAGE was performed under reducing conditions. For immunoprecipitation, antibodies were added to the lysates and incubated at 4 °C for 1–3 h. Protein G–Sepharose beads were then added, and rotated at 4 °C for 30–60 min. The immune complexes were pelleted by centrifugation at 16000 g at 4 °C for 30 s, and then washed three times with lysis buffer. The immune complexes were boiled for 3 min in sample buffer [0.125 M Tris/HCl (pH 6.8), 20% (v/v) glycerol and 4% SDS] in the presence (for detection of integrin α3) or absence (for detection of CD151 and CD151–EGFP) of 10% (v/v) 2-mercaptoethanol, and then subjected to SDS/PAGE. The proteins separated by SDS/PAGE were transferred on to PVDF membranes (Millipore) in 0.1 M Tris base, 0.192 M glycine and 20% (v/v) methanol using a semi-dry electrophoretic transfer cell (Bio-Rad). Membranes were treated with blocking buffer containing 10% non-fat dried skimmed milk powder in T-TBS [Tris-buffered saline (20 mM Tns/HCl, pH 7.5, and 150 mM NaCl) containing 0.1% Tween 20), at room temperature (25 °C) for at least 1 h, and then incubated with antibodies in blocking buffer, at room temperature for 1 h. After being washed three times with T-TBS, membranes were incubated with peroxidase-coupled secondary antibodies in blocking buffer, at room temperature for 1 h. Membranes were then washed four times with T-TBS, and visualized using the ECL® (enhanced chemiluminescence) system (GE Healthcare).
Cells transfected with siRNAs were detached from dishes using PBS containing 0.025% trypsin and 1 mM EDTA, suspended in DMEM/FBS, and kept in suspension for 1 h at 37 °C. Cells were washed twice with biotinylation buffer containing 150 mM NaCl and 20 mM Hepes/NaOH (pH 8.0), and surface-labelled at room temperature for 15 min with 2 mg/ml sulfo-NHS-LC-biotin [sulfo-succinimidyl-6-(biotinamido)hexanoate; Pierce]. After being washed with DMEM and PBS, cells were lysed in a buffer containing 1% (w/v) Triton X-100, 5% (v/v) glycerol, 150 mM NaCl, 1 mM EDTA, 20 mM Tris/HCl (pH 7.5), 1 mM PMSF, 5 μg/ml leupeptin and 5 μg/ml aprotinin. Protein concentrations of lysates were determined using the BCA protein assay kit, and equal protein amounts of the lysates were subjected to immunoprecipitation with anti-CD151 mAb (8C3), anti-integrin α3 mAb (3G8), anti-integrin α5 mAb (8F1) and normal mouse IgG. Immunoprecipitates were blotted with peroxidase-conjugated streptavidin (Zymed).
Differences in the abilities of mouse anti-human CD151 mAbs to co-precipitate integrin α3
We examined a panel of anti-human CD151 mAbs, namely, 8C3, TS151r, 14A2.H1, LIA1/1, VJ1/16, 14B5, SFA1.2B4, 11B1, 11G5 and TS151, for their capability to co-precipitate integrin α3 from lysates prepared from human A549 cells with Triton X-100-containing buffer, to assess their reactivities toward the CD151–integrin α3β1 complex (Figure 1A). Integrin α3 was co-immunoprecipitated with CD151 by 11G5 and TS151, but not by other mAbs, although SFA1.2B4 showed faint co-immunoprecipitation activity. When immunoprecipitation was performed on lysates prepared with lysis buffer containing Brij 97, a milder detergent than Triton X-100, not only 11G5 and TS151, but also 11B1 and SFA1.2B4 were capable of strongly co-precipitating integrin α3 bound to CD151 (Figure 1B). Other mAbs showed only weak co-precipitating activities under the less-stringent detergent condition, compared with 11G5, TS151, 11B1 and SFA1.2B4. When immunoprecipitates obtained with anti-CD151 mAbs were immunoblotted with anti-EGFR pAb as a control for unrelated proteins, EGFR was not detected (Figures 1A and 1B), confirming the specificity of co-immunoprecipitation of integrin α3 by anti-CD151 mAbs from both Triton X-100 and Brij 97 lysates. In addition, we also performed immunoprecipitation with the anti-integrin α3 mAb 3G8 from both Triton X-100 and Brij 97 lysates, followed by immunoblotting with anti-CD151 mAb. CD151 was co-precipitated with anti-integrin α3 mAb from both types of lysate (Figures 1C and 1D), demonstrating that the complex between CD151 and integrin α3 was stable in the presence of these detergents. Taken together, these findings suggest that anti-CD151 mAbs can be classified into three groups based on their abilities to co-precipitate integrin α3 under two detergent conditions with different stringencies (Figure 1E). The mAbs in the first group (Group-1) do not effectively coimmunoprecipitate integrin α3 under either condition; these included 8C3, TS151r, 14A2.H1, LIA1/1, VJ1/16 and 14B5. Those in the second group (Group-2) strongly co-immunoprecipitate integrin α3 only under the low-stringency detergent condition (that is, in buffer containing Brij 97), and include 11B1 and SFA1.2B4. Those in the third group (Group-3) potently co-immunoprecipitate integrin α3, even under the high-stringency detergent condition (that is, in buffer containing Triton X-100), and include 11G5 and TS151.
Differences in the reactivities of anti-human CD151 mAbs toward integrin α3-knocked-down and integrin α3-overexpressing cells
Increased occupancy of CD151 by integrin α3β1 has been shown to reduce the reactivity of the mAb TS151r towards CD151 , indicating that the epitope recognized by this antibody is concealed upon association of CD151 with integrin α3β1. To explore this ‘masking effect’ of integrin α3β1 on the reactivities of the anti-CD151 mAbs of different groups, we knocked down the expression of integrin α3 in A549 cells by RNA interference, and examined how the reactivities of the mAbs to CD151 were altered in the integrin α3-knocked-down cells. Flow-cytometric analysis of control and integrin α3 siRNA-treated A549 cells labelled with anti-integrin α3 antibodies confirmed a marked (∼90%) reduction in the level of integrin α3 expression in integrin α3-knocked-down cells (Table 1 and Figure 2A). Immunoblot analysis also showed that the expression of integrin α3 was markedly diminished (>95%) following treatment with integrin α3 siRNA (Figure 2B). Notably, CD151 expression was also reduced (∼40%) in integrin α3-knocked-down cells. A similar reduction in the level of CD151 expression was observed in integrin α3-knocked-down cells upon biotinylation of cell-surface proteins, although the level of integrin α5 expression was unchanged (Figure 2C). Since exogenous expression of integrin α3 has been shown to elevate the cell-surface expression of CD151 in human myeloid leukaemia K562 cells [13,15,23], it seems likely that the cell-surface expression of CD151 is stabilized by integrin α3 and, therefore, that knockdown of integrin α3 leads to turnover of CD151 within cells.
When integrin α3-knocked-down cells were subjected to flow-cytometric analysis using a panel of anti-CD151 mAbs, Group-2 (11B1 and SFA1.2B4) and Group-3 (11G5 and TS151) mAbs showed reduced reactivities to integrin α3-knocked-down cells, consistent with the results obtained in immunoblot and cell-surface biotinylation analyses of total and surface-expressed CD151 respectively (Table 1 and Supplementary Figure S1 at http://www.BiochemJ.org/bj/415/bj4150417add.htm). By contrast, there was no significant decrease in the reactivities of Group-1 mAbs toward integrin α3-knocked-down cells. 14B5, a Group-1 mAb, exhibited increased reactivity toward integrin α3-knocked-down cells. Given the ∼40% reduction in the surface expression level of CD151 in integrin α3-knocked-down cells, these results indicated that the reactivities of Group-1 mAbs were increased in these cells. Since the knockdown of integrin α3 was expected to unmask epitopes concealed by the association of CD151 with integrin α3β1, the increased reactivities of Group-1 mAbs indicated that the epitopes they recognize were sterically concealed by bound integrin α3β1, as has been shown for TS151r .
To confirm the masking effect of integrin α3 on the binding of Group-1 mAbs to CD151 on the cell surface, integrin α3 was transiently overexpressed in HEK-293T cells, which express integrin α3 at low levels. The expression vector encoding integrin α3 was co-transfected with one-tenth of the amount of expression vector encoding EGFP, so that almost all EGFP-positive cells were positive for exogenously expressed integrin α3. The reactivities of Group-1 mAbs toward integrin α3-overexpressing HEK-293T cells were significantly reduced, but those of Group-3 mAbs remained unaltered (Table 2 and Supplementary Figure S2 at http://www.BiochemJ.org/bj/415/bj4150417add.htm). These results are consistent with previous reports showing that overexpression of integrin α3 interferes with the binding of Group-1 mAbs (TS151r, 14A2.H1 and 14B5), but not that of a Group-3 mAb (11G5), to CD151 on cell surfaces [14,23], and support our conclusion that Group-1, but not Group-3, mAbs exhibit increased reactivities toward CD151 that is unoccupied by integrin α3. It should be noted that two Group-2 mAbs, SFA1.2B4 and 11B1, differed in their reactivities toward integrin α3-overexpressing HEK-293T cells: the reactivities of SFA1.2B4 to HEK-293T cells decreased upon overexpression of integrin α3, whereas the reactivity of 11B1 did not. Although 11B1 and SFA1.2B4 were both classified as Group-2 mAbs based on their differential activities to co-precipitate integrin α3 from low- and high-stringency detergents, they may not be closely related to each other in other properties, as is evident from the clear distinction in their epitopes (see below).
Epitope mapping of anti-human CD151 mAbs
To explore the molecular basis of the distinct reactivities among different groups of anti-CD151 mAbs in more detail, we set out to map the amino acid residues within CD151 that are critical for antigen recognition by these mAbs. To this end, human CD151 with or without amino acid substitutions was transiently expressed in mouse NIH 3T3 cells in fusion proteins with EGFP, followed by determination of the reactivities of those fusion proteins toward a panel of anti-CD151 mAbs by two-colour flow-cytometric analysis. Bound anti-CD151 mAbs were detected by R-phycoerythrin-conjugated anti-mouse IgG antibody, whereas the expression levels of exogenous CD151 were quantified by EGFP fluorescence. When the CD151–EGFP fusion proteins are fully capable of binding to the mAbs being tested, the amounts of the mAbs bound to the transfected cells should be proportional to the expression levels of the fusion proteins detected by EGFP fluorescence (Figure 3A). We expressed 11 different human CD151–EGFP fusion proteins, each containing a substitution at one or two amino acid residues within the EC2 extracellular domain that differ between human and mouse CD151 (Figure 3B). When the mutated residue(s) that comprises the epitope of the mAb is tested, the reactivity of the mAb toward the transfected cells should be compromised. Two-colour flow-cytometric analysis revealed decreases in the reactivities of all Group-1 mAbs (8C3, TS151r, 14A2.H1, LIA1/1, VJ1/16 and 14B5) to R165Q, G176D/G177S, L191G and Q194K mutants, although LIA1/1 and VJ1/16 also showed reduced reactivities to the Q173G mutant (Table 3 and Supplementary Figure S3 at http://www.BiochemJ.org/bj/415/bj4150417add.htm). The reduction in the binding to Q194K mutant was most prominent in Group-1 mAbs. These results indicate that Arg165, Gly176, Gly177, Leu191 and Gln194 comprise the epitopes within CD151 that are recognized by Group-1 mAbs. Group-2 mAbs (11B1 and SFA1.2B4) were clearly distinct from Group-1 mAbs in their binding specificities toward CD151 mutants. The reactivity of mAb 11B1 was not compromised by mutations that provoked a significant reduction in binding to Group-1 mAbs, except for Q194K, whereas the reactivity of SFA1.2B4 was compromised by R165Q, Q173G and GG176/177DS mutations. Group-3 mAbs, 11G5 and TS151, were also distinct from Group-1 mAbs: their reactivities were not affected by G176D/G177S, L191G and Q194K mutations, whereas they were sensitive to Q173G, P137S and R165Q mutations. The reactivities of all mAbs toward the 11 mutant forms of CD151–EGFP are summarized in Figure 4(B). These results indicate that Gln173 comprises a common part of the epitopes recognized by Group-3 mAbs, whereas no common epitopes were found for Group-2 mAbs.
Gly176/Gly177 and Gln194, which comprise the epitopes recognized by all Group-1 mAbs, were also involved as part of the epitopes recognized by Group-2 mAbs, implying that these residues are involved in the inabilities of Group-1 and -2 mAbs to co-immunoprecipitate integrin α3 under the high-stringency detergent conditions. In addition, Leu191 was specifically recognized by Group-1 mAbs, indicating that recognition of Leu191 results in the weak activities of Group-1 mAbs to co-immunoprecipitate integrin α3 under both low- and high-stringency detergent conditions and the increased reactivities of Group-1 mAbs to integrin α3β1-unbound CD151. It should also be noted that almost all recognition sites revealed by epitope mapping were mapped to the variable region of EC2.
Exploration of regions within CD151 responsible for its binding to integrin α3β1
We designated the three segments in the EC2 variable region, which are divided by three disulfide bonds, as Sites-1, -2 and -3 (Figure 4A). The epitope mapping of anti-CD151 mAbs, particularly that of Group-1 mAbs, raised the possibility that not only the region including Gln194 in Site-3, which has been reported to participate in the interaction between CD151 and integrin α3β1 , but also the regions around Gly176/Gly177 and Leu191 in Site-1 and Site-2 respectively, are involved in the formation of the complex between CD151 and integrin α3β1 (Figure 4B). To explore this possibility, we expressed CD151–EGFP mutants (Site-1 SW and Site-2 SW) in which the amino acid residues around Gly176/Gly177 and Leu191 were replaced with the corresponding residues of CD37 and NAG2 respectively, neither of which are capable of forming a stable complex with integrin α3β1 [14,33], in HEK-293T cells, together with exogenous integrin α3 (Figure 5). Immunoprecipitation with the anti-integrin α3 mAb demonstrated that the Site-1 SW mutant with a substitution of five amino acid residues, including Gly176/Gly177, retained the ability to be co-precipitated with integrin α3 under the high-stringency detergent conditions, whereas the Site-3 SW mutant, in which three amino acid residues including Gln194 were replaced with the corresponding residues of CD63 , did not (Figures 5A–5C). Other CD151 mutants in which the amino acid residues surrounding Gly176/Gly177 were replaced with alanine were also found to retain the ability to co-precipitate with integrin α3 (Figures 5B and 5D), indicating that the region around Gly176/Gly117 was not directly involved in the stable association of CD151 with integrin α3β1.
By contrast, the Site-2 SW mutant in which six amino acid residues (amino acids 186–191) within Site-2, including Leu191, were swapped with the corresponding residues of NAG2, failed to co-immunoprecipitate with integrin α3, as did the Site-3 SW mutant (Figures 5E and 5F). These results indicate that Site-2 containing Leu191 is involved in the complex formation between CD151 and integrin α3β1. Two-colour flow-cytometric analysis also demonstrated that this mutant showed as strong a reactivity to Group-2 and -3 mAbs as wild-type CD151, whereas it exhibited reduced reactivity to Group-1 mAbs (Supplementary Figure S4 at http://www.BiochemJ.org/bj/415/bj4150417add.htm), similar to the reactivity of the L191G mutant to these mAbs revealed by epitope mapping analysis (Table 3 and Supplementary Figure S3). In support of these results, the Site-2 SW mutant was recognized by 11G5, a Group-3 mAb, but not by the Group-1 mAb 8C3 (Figure 5G) in Western blot analysis, consistent with our conclusion that Leu191 comprises part of the epitopes recognized by 8C3. Taken together, the unaltered reactivities toward Group-2 and -3 mAbs, including 11G5, are indicative of the efficient expression of the Site-2 SW mutant protein in transfected cells, and suggest that the swap mutation in Site-2 did not impose any significant alteration to the gross conformation of CD151. Thus the failure of the Site-2 SW mutant to co-precipitate with integrin α3 does not seem to be due to conformational inactivation, but results from the loss of the residues involved in the association between CD151 and integrin α3β1.
To further investigate the involvement of Site-2 and Site-3 in the formation of the complex between CD151 and integrin α3, we synthesized two peptides, KTVVAL (amino acids 186–191) and GQRDHA (amino acids 193–198), corresponding to the substituted regions at Site-2 and Site-3 respectively. We examined whether these peptides could dissociate the complex between CD151 and integrin α3, thereby inhibiting co-immunoprecipitation of CD151 with integrin α3. However, these peptides did not show any inhibitory effects on the co-immunoprecipitation of CD151 with integrin α3, even when both of the peptides were simultaneously added (Supplementary Figure S5 at http://www.BiochemJ.org/bj/415/bj4150417add.htm). Since the three-dimensional structure around Site-2 and Site-3 might be required for stable complex formation between CD151 and integrin α3, synthetic peptides that were free from the constraint stabilizing the native conformation of peptide sequences could not exert their inhibitory effects.
In the present study, we performed epitope mapping of a series of mouse anti-human CD151 mAbs that have different abilities to co-immunoprecipitate integrin α3 and distinct reactivities toward CD151 unoccupied by integrin α3β1. We divided these antibodies into three groups, Groups-1, -2 and -3, based on their abilities to co-immunoprecipitate integrin α3 under different detergent conditions. Given that 8C3, a Group-1 mAb incapable of effectively co-immunoprecipitating integrin α3 with CD151, has been shown to dissociate CD151 from integrin α3β1 , the differences in the co-precipitating activities among the three groups of anti-CD151 mAbs may reflect their abilities to dissociate the complex between CD151 and integrin α3β1. Thus Group-1 mAbs are most potent at disrupting the interaction of CD151 with integrin α3β1, whereas Group-2 mAbs are less potent and Group-3 mAbs are devoid of disrupting activity. Our results also show that Group-1 mAbs have increased reactivities toward CD151 in integrin α3-knocked-down cells, but they have decreased reactivities in integrin α3-overexpressing cells, indicating that the epitopes for these mAbs are concealed by the integrin α3β1 bound to CD151. Since TS151r, another Group-1 mAb, has been shown to preferentially recognize CD151 unoccupied by integrin α3β1 , it is also possible that Group-1 mAbs may selectively bind to CD151 uncomplexed with integrin α3β1, thereby failing to co-precipitate integrin α3. In either case, our results indicate that Group-1 mAbs bind at and/or near the integrin α3β1-binding site on CD151, and competitively and/or sterically interfere with the interaction between CD151 and integrin α3β1.
The complex between CD151 and integrin α3β1 has been proposed to form through two distinct interactions, namely, Triton X-100-resistant and Triton X-100-sensitive interactions [19,34]. Both of these interactions contribute to the complex formation between CD151 and integrin α3 under the low-stringency detergent conditions (that is, in the presence of the Brij series of detergents), although only the Triton X-100-resistant interaction secures complex formation under the high-stringency detergent conditions (that is, in the presence of Triton X-100). Since Group-1 mAbs showed weak activities to co-precipitate integrin α3 under the low-stringency detergent conditions and were totally devoid of co-precipitating activity under the high-stringency detergent conditions, we reasoned that Group-1 mAbs could destabilize the Triton X-100-resistant, but not the Triton X-100-sensitive, interaction, thereby resulting in a complete abrogation of co-precipitation of integrin α3 under the high-stringency detergent conditions. Under the low-stringency detergent conditions, however, the Triton X-100-sensitive interaction that plays an auxiliary role in stabilizing the CD151–integrin α3 complex under the low-stringency detergent conditions is not compromised by Group-1 mAbs, leading to residual co-immunoprecipitation of integrin α3 under the low-stringency detergent conditions.
Our results show that the anti-CD151 mAbs differ in their accessibilities to CD151 in integrin α3-knocked-down A549 cells. Among the mAbs tested, Group-2 and -3 mAbs, but not Group-1 mAbs, exhibited clear reductions in their reactivities toward integrin α3-knocked-down cells. Given the partial decrease in CD151 expression in integrin α3-knocked-down cells, the sustained reactivities of Group-1 mAbs are indicative of their increased reactivities towards CD151 unoccupied by integrin α3β1. Furthermore, in flow-cytometric analyses using HEK-293T cells overexpressing integrin α3, Group-1 mAbs exhibited a significant reduction in their reactivities toward integrin α3-overexpressing cells, whereas Group-2 (11B1, but not SFA1.2B4) or Group-3 mAbs did not. These results are consistent with previous reports showing that overexpression of integrin α3 reduces the cell-surface binding of TS151r, 14A2.H1 and 14B5 (Group-1 mAbs), but not 11B1 (Group-2 mAb) or 11G5 (Group-3 mAb), in K562 and HeLa cells [14,23], confirming that Group-1 mAbs preferentially bind to CD151 unoccupied by integrin α3β1.
Epitope mapping of the anti-CD151 mAbs revealed that Arg165, Gly176/Gly177, Leu191 and Gln194 are within epitopes recognized by all Group-1 mAbs (8C3, TS151r, 14A2.H1, LIA1/1, VJ1/16 and 14B5), whereas Arg165 was also recognized by TS151, a Group-3 mAb that has a strong ability to co-immunoprecipitate integrin α3. It seems likely, therefore, that a surface of CD151 containing Gly176/Gly177, Leu191 and Glu194, but not Arg165, may comprise an interface involved in the formation of the complex between CD151 and integrin α3β1. This view is consistent with a previous report showing that a short stretch in Site-3 containing Gln194 participates in the strong association of CD151 with integrin α3β1 . Furthermore, we found that Site-2, which includes Leu191, is also required for the formation of the CD151–integrin α3β1 complex, whereas the region around Gly176/Gly177 in Site-1 is not. It is conceivable, therefore, that Site-1, containing Gly176/Gly177, does not directly contribute to the formation of the complex between CD151 and integrin α3β1, but that this site may be situated near the interface and its recognition by Group-1 mAbs may sterically interfere with the association of CD151 with integrin α3β1.
Among Group-2 mAbs that exhibit a moderate activity to co-immunoprecipitate integrin α3, SFA1.2B4 recognizes Arg165 and Gln173 together with Gly176/Gly177, but not Leu191 or Glu194. Arg165 and/or Gln173 are also recognized by 11G5 and TS151, mAbs that have a strong ability to co-precipitate integrin α3, implying that recognition of Arg165 and Gln173 by SFA1.2B4 does not compromise the co-immunoprecipitation of integrin α3 by SFA1.2B4. By contrast, recognition of Gly176/Gly177 by SFA1.2B4 may lead to disruption of the CD151–integrin α3β1 complex by SFA1.2B4 under the high-stringency detergent conditions.
Gln194 is part of the epitope recognized by the Group 2 mAb 11B1, but the binding of 11B1 to CD151 was only partially compromised by replacement of this residue with lysine (the Q194K mutation), in striking contrast with Group-1 mAbs. Furthermore, this mAb did not recognize Gly176/Gly177 or Leu191, which, together with Gln194, comprise the epitopes recognized by Group-1 mAbs. These distinctions between 11B1 and Group-1 mAbs may explain the difference in their capabilities to co-precipitate integrin α3 under the low-stringency detergent conditions. In our epitope mapping analysis, we could not find any crucial epitope for recognition by 11B1, except that Gln194 was partially involved in the recognition of CD151 by 11B1. An unidentified epitope(s) recognized by 11B1 may be involved in the strong co-precipitation activity of 11B1 under the low-stringency detergent conditions.
A Gln194-containing short segment has been shown to comprise the epitopes recognized by TS151r, probably being localized within the interface between CD151 and integrin α3β1 . Gln194 also comprises a part of the epitopes recognized by 11B1, whose reactivity toward CD151 either complexed or uncomplexed with integrin α3 was unchanged, as demonstrated by the experiments with integrin α3-knocked-down and integrin α3-overexpressing cells. This latter observation may argue against the conclusion that Gln194 is concealed by bound integrin α3β1. However, it should again be noted that the impact of the Q194K mutation on the reactivity of 11B1 to CD151 was rather small (see Table 3 and Figure 4), and therefore the contribution of Gln194 to CD151 recognition by 11B1 may be only marginal, resulting in a lack of a masking effect of integrin α3 on the binding of 11B1 to CD151, even if Gln194 is occupied by integrin α3β1.
In summary, Gly176/Gly177, Leu191 and Gln194 comprise the epitopes characteristically recognized by Group-1 mAbs that only weakly co-immunoprecipitate integrin α3. The subsites Site-2 and Site-3 within the EC2 variable region, which contain Leu191 and Gln194 respectively, are functionally involved in the association of CD151 with integrin α3β1; hence, the association of integrin α3β1 with CD151 leads to the blockage of the binding of Group-1 mAbs to CD151. Site-1, containing Gly176/Gly177, does not seem to be directly involved in the formation of the complex between CD151 and integrin α3β1, yet the recognition of this subsite by Group-1 mAbs may be sterically compromised by the association of integrin α3β1 with CD151. Consistent with this scenario, Gly176/Gly177 and Gln194 comprise part of the epitopes recognized by Group-2 mAbs that fail to co-precipitate integrin α3 under the high-stringency detergent conditions. Given the importance of the mAbs that dissociate and/or differentially recognize the CD151–integrin α3β1 complex as probes for CD151 functions on the cell surface [13,21–26], our results are informative with respect to the mechanism of molecular recognition used by these mAbs, and provide insight into the molecular basis of complex formation between CD151 and integrin α3β1 and their assembly into multimolecular membrane microdomains.
We thank Dr Tsutomu Tsuji (Hoshi University School of Pharmacy and Pharmaceutical Sciences, Tokyo, Japan) for kindly providing the human integrin α3 expression vector. This work was supported by Grant-in-Aid for Scientific Research (to K. S. and M. Y.) from Japan Society for Promotion of Science and by Grant-in Aids for Young Scientists (to M. Y.) and for Scientific Research on Priority Areas (to K. S.) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.
Abbreviations: BCA, bicinchoninic acid; DMEM, Dulbecco's modified Eagle's medium; EGFP, enhanced green fluorescent protein; EGFR, epidermal-growth-factor receptor; FBS, fetal bovine serum; GFP, green fluorescent protein; HEK, human embryonic kidney; mAb, monoclonal antibody; pAb, polyclonal antibody; siRNA, small interfering RNA
- © The Authors Journal compilation © 2008 Biochemical Society