(LFA-1) 2Β L Α Integrin Subunit I Domain in Regulation of 2Β The

The Role of the CPNKEKEC Sequence in the β2 Subunit I Domain in Regulation of Integrin αLβ2 (LFA-1)

This information is current as Tetsuji Kamata, Kenneth Khiem Tieu, Takehiko Tarui, of September 27, 2021. Wilma Puzon-McLaughlin, Nancy Hogg and Yoshikazu Takada J Immunol 2002; 168:2296-2301; ; doi: 10.4049/jimmunol.168.5.2296

http://www.jimmunol.org/content/168/5/2296 Downloaded from

References This article cites 43 articles, 27 of which you can access for free at: http://www.jimmunol.org/content/168/5/2296.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 27, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2002 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. ␤ The Role of the CPNKEKEC Sequence in the 2 Subunit I ␣ ␤ 1 Domain in Regulation of Integrin L 2 (LFA-1)

Tetsuji Kamata,2* Kenneth Khiem Tieu,* Takehiko Tarui,* Wilma Puzon-McLaughlin,* Nancy Hogg,† and Yoshikazu Takada2* ␣ ␣ ␤ The L I (inserted or interactive) domain of integrin L 2 undergoes conformational changes upon activation. Recent studies show ␣ ␤ that the isolated, activated L I domain is sufﬁcient for strong ligand binding, suggesting the 2 subunit to be only indirectly ␣ ␤ involved. It has been unclear whether the activity of the L I domain is regulated by the 2 subunit. In this study, we demonstrate ␤ ␤ that swapping the disulﬁde-linked CPNKEKEC sequence (residues 169–176) in the 2 I domain with a corresponding 3 sequence, 2؉ ␤ ␣ 174 or mutating Lys to Thr, constitutively activates L 2 binding to ICAM-1. These mutants do not require Mn for ICAM-1 2؉

binding and are insensitive to the inhibitory effect of Ca . We have also localized a component of the mAb 24 epitope (a reporter Downloaded from ␤ 173 175 ␤ of 2 integrin activation) in the CPNKEKEC sequence. Glu and Glu of the 2 I domain are identiﬁed as critical for mAb ␤ 24 binding. Because the epitope is highly expressed upon 2 integrin activation, it is likely that the CPNKEKEC sequence is ␣ exposed or undergoes conformational changes upon activation. Deletion of the L I domain did not eliminate the mAb 24 epitope. ␣ This conﬁrms that the L I domain is not critical for mAb 24 binding, and indicates that mAb 24 detects a change expressed in ␤ ␤ part in the 2 subunit I domain. These results suggest that the CPNKEKEC sequence of the 2 I domain is involved in regulating

␣ http://www.jimmunol.org/ the L I domain. The Journal of Immunology, 2002, 168: 2296–2301. ␣ ␤ ␣␤ ␣ he integrin L 2 (LFA-1, CD11a/CD18) is an het- domain of the integrin subunits undergoes conformational ␤ ␣ ␤ erodimeric receptor of the 2 integrin family. L 2 is ex- changes on activation. The two different conformations of the in- T pressed on all leukocytes, is crucial to the inflammatory tegrin ␣ subunit I domain (open and closed) have recently been process, and mediates adhesion to ligands ICAM-1, ICAM-2, and defined, and it has been proposed that these two structures repre- ␣ ␤ ICAM-3 (reviewed in Refs. 1Ð4). The adhesiveness of L 2 can sent the high-affinity and low-affinity conformations, respectively be dynamically regulated by intracellular signals (inside-out sig- (11Ð13). Recently, Kallen et al. (14) found that a chemical, lova- 2ϩ ␣ ␣ ␤ naling) (5). Activation from the outside of the cell with Mg and statin, binds to the L I domain and blocks L 2-ligand interac-

␣ ␤ by guest on September 27, 2021 EGTA results in the formation of a high-affinity form of L 2,as tion. It is likely that lovastatin blocks the conformational change shown by an increased ability to bind to soluble ICAM-1, and in that occurs when the closed (inactive) form alters to the open (ac- ␣ ␣ the expression of an activation reporter epitope recognized by tive) form. The M and L I domain, with an open and closed mAb 24 (2, 6, 7). mAb 24 was originally proposed to recognize an conformation, has been generated by site-directed mutagenesis ␤ ␣ ␣ ␣ ␣ ␣ epitope common to all 2 integrin subunits ( L, M, X) (6Ð9). (15, 16). The isolated L I domain with locked open conformation ␣ 3 ␤ The L subunit has an I (inserted or interactive) domain of is sufficient for ligand binding, suggesting that the 2 subunit may ϳ200 amino acid residues that is critically involved in ligand bind- be only indirectly involved in ligand binding (17). ing. The I domain consists of a central ␤ sheet, surrounded by It has been proposed that the integrin ␤ subunit also has an I seven ␣ helices, which is folded as a globular domain (Rossmann- domain structure within the N-terminal region, which has been ␣ ␤ fold). At the top of the globular domain, the I domain has a metal validated in the recent v 3 crystal structure (11, 18). It is unclear ion-dependent adhesive site (MIDAS) that is involved in coordi- ␤ how and whether the 2 subunit I domain is involved in the con- nating cations Mg2ϩ or Mn2ϩ and in binding ligands (10). The I ␣ ␤ formational changes that L 2 undergoes upon activation. We have reported that the disulfide-linked predicted loop of the inte- ␤ grin 3 I domain (the CYDMKTTC sequence) is critically in- *Department of Cell Biology, The Scripps Research Institute, La Jolla, CA 92037; volved in the ligand binding and specificity of the non-I domain † and Leukocyte Adhesion Laboratory, Imperial Cancer Research Fund, London, integrin ␣ ␤ (19). Also, it has been proposed that the loop is United Kingdom v 3 localized within the putative ligand-binding pocket in the non-I Received for publication June 20, 2001. Accepted for publication December 13, 2001. domain integrin ␣ ␤ (20). The recent ␣ ␤ crystal structure The costs of publication of this article were defrayed in part by the payment of page IIb 3 v 3 charges. This article must therefore be hereby marked advertisement in accordance shows that the CYDMKTTC sequence is actually exposed to the ␤ with 18 U.S.C. Section 1734 solely to indicate this fact. surface in the headpiece of the 3 subunit (18) (Fig. 1). In the 1 This work was supported by National Institutes of Health Grant GM49899 (to Y.T.) present study, we designed mutagenesis experiments to determine and by Department of the Army Grant DAMD17-97-1-7105 (to T.K.). This is pub- the potential function of the disulfide-linked CPNKEKEC se- lication 14111-VB from The Scripps Research Institute. ␤ quence in the 2 subunit I domain. We show that mutation of the 2 Address correspondence and reprint requests to Dr. Yoshikazu Takada, Department ␣ ␤ of Cell Biology, The Scripps Research Institute, 10550 North Torrey Pines Road, La CPNKEKEC sequence constitutively activates L 2. We propose ␤ Jolla, CA 92037. E-mail address: [email protected], or Dr. Tetsuji Kamata at the that this sequence in the 2 I domain is critically involved in reg- current address: Department of Anatomy, Keio University School of Medicine, 35 ␣ ulating the L subunit I domain. We found that the CPNKEKEC Shinanomachi, Shinjuku-ku, Tokyo 160, Japan. E-mail address: [email protected]. ␤ keio.ac.jp sequence of 2 is part of the mAb 24 epitope, indicating that mAb ␤ 3 Abbreviations used in this paper: I, inserted or interactive; CHO, Chinese hamster 24 detects a change in the 2 I domain and/or in interdomain ovary; MIDAS, metal ion-dependent adhesive site. interaction on activation.

Materials and Methods Materials ␤ mAb 24 was generated as previously described (6). mAbs IB-4 (anti- 2) ␣ (21) and TS 1/22 (anti- L) (22) were obtained from American Type Cul- ␤ ture Collection (ATCC, Manassas, VA). mAbs MEM-48 (anti- 2) and ␣ MEM-83 (anti- L) (23) were provided by V. Horejsi (Institute of Molec- ular Genetics, Academy of Sciences of the Czech Republic, Prague, Czech Republic). ICAM-1/mouse C␬ fusion protein was obtained from G. Weitz ␤ (Novartis, Basel, Switzerland). Rat anti-mouse 2 mAb (C71/16) was pur- ␤ chased from BD PharMingen (San Diego, CA). Mouse 2 cDNA was obtained from ATCC. ␣ ␤ cDNA construct and expression of L 2 ␣ ␤ ␤ Human L, human 2, and mouse 2 cDNAs were subcloned into pBJ-1 (24), and site-directed mutagenesis was conducted using unique restriction site elimination (25). The presence of mutations was confirmed by DNA ␣ ␤ sequencing. L and 2 cDNAs in the pBJ-1 vector were transfected into Chinese hamster ovary (CHO) cells by electroporation. Flow cytometry was conducted as described (26). In some experiments, 0.5 mM Mn2ϩ was added to induce mAb 24 binding. Downloaded from Adhesion assays ␣ ␤ Adhesion of CHO and K562 cells expressing L 2 to ICAM-1 was assayed as described (27). Briefly, wells of 96-well Immulon-2 plates were coated with goat anti-mouse C␬ chain polyclonal Ab (Caltag Laboratories, South San Francisco, CA; 0.4 ␮g/well in 100 ␮l of PBS), and then with ICAM- 1/mouse C␬ fusion protein (8 ␮g/ml), unless otherwise specified. Unoc-

cupied protein binding sites were blocked by incubating the wells with 1% http://www.jimmunol.org/ heat-denatured BSA. Cells (105/well) were added and incubated for1hat 37¡Cin100␮l of Tyrode/5 mM HEPES buffer, pH 7.4, in the presence of 2ϩ 2ϩ 0.1% BSA and 2 mM MgCl2 or 0.1 mM Mn . In some experiments, Ca was added at indicated concentrations. Bound cells were quantiﬁed by assaying endogenous phosphatase activity (27). Data are shown as means Ϯ SD of triplicate experiments. Results ␤ Effect of mutating conserved residues in the 2 I domain on

␤ by guest on September 27, 2021 mAb 24 epitope, a reporter for 2 integrin activation, on ICAM-1 binding It has been reported that mutating Asp112, Ser114, Asp209, and 212 ␤ ␤ Glu residues in 2 significantly blocks ligand binding (28, 29). FIGURE 1. Effect of mutations in the 2 I domain on ICAM-1 and mAb 209 ␤ 24 binding. Point mutants in the ␤ I domain were transiently expressed in Also, it has been reported that mutating Asp of 2 affects bind- 2 112 ␣ ing of mAb 24 (29). Several residues (which correspond to Asp , CHO cells together with wild-type L. Cells were tested for their ability to Ser114, Asn207, Asp209, Asp216, Asp250, and Asp278 in ␤ ) in the I bind to ICAM-1 (A). The transfected cells were stained with mAb 24 in the 2 presence of 0.5 mM Mn2ϩ (B), or with MEM-48 (anti-␤ ) to monitor the domain of the ␤ subunit are critical for fibronectin binding to 2 1 expression of ␣ ␤ , followed by FITC-labeled anti-mouse IgG, and ana- non-I domain integrin ␣ ␤ (30, 31). It has not been established, L 2 5 1 lyzed by flow cytometry. Data are expressed as percentage of bound cells however, whether these residues that are critical for ligand binding or percentage of positive cells in flow cytometry. relate to the conformational alteration of the I domain during activation. To address this question, we tested whether the conserved ␤ ␤ residues in the 2 subunit I domain are critical for both exposure (residues 169Ð176), of the 2 subunit with the corresponding se- ␤ ␤ of the mAb 24 epitope and binding of ICAM-1 on activation. We quence of the 3 subunit (designated the 2-3-2 mutant). The cor- ␤ ␤ mutated several amino acid residues of the 2 subunit that are responding sequence of the 3 subunit dictates ligand binding and ␤ ␣ ␤ ␣ ␤ conserved among integrin subunits to Ala and tested their ability specificity in v 3 and IIb 3 (19, 20) and is exposed to the sur- ␤ ␣ ␤ to bind to ICAM-1 and mAb 24 using CHO cells transiently ex- face at the top of the I domain in v 3 (18) (Fig. 2). We tran- ␣ ␤ ␤ pressing L 2 mutants. We found that mutating several conserved siently expressed the 2-3-2 mutant into K562 cells together with 112 114 116 120 151 207 209 ␣ residues (Asp , Ser , Ser , Asp , Asp , Asp , Asp , wild-type L and tested its ability to bind to ICAM-1 (Fig. 3A). We 212 216 250 278 ␤ ␣ ␤ Glu , Asp , Asp , and Asp ) blocks ICAM-1 binding (Fig. found that the 2-3-2 mutant showed activation of L 2.Wein- ␤ 1A). Interestingly, we found that all of these mutations block the troduced point mutations within residues 170Ð175, in which the 2 ␤ exposure of the mAb 24 epitope on activation (Fig. 1B). These residues are changed to the corresponding 3 residues, and studied ␤ 174 results suggest that there is a correlation between the 2 residues their ability to bind to ICAM-1. We found that the Lys to Thr ␣ ␤ that are critical for binding to ICAM-1 and those allowing expo- (K174T) mutant also activated L 2, but none of the other mutants ␣ ␤ sure of the mAb 24 epitope. affected ICAM-1 binding to L 2. We then stably expressed the ␤ and the K174T mutants in ␤ ␤ 2-3-2 Effect of mutating the CPNKEKEC B- C loop sequence in the CHO cells together with wild-type ␣ and further sorted the cells ␤ ␣ ␤ L 2 subunit on ICAM-1 binding to L 2 ␣ ␤ to obtain high expressers (designated L 2-3-2-CHO and the ␤ ␣ ␤ ␣ ␤ To study the potential role of the 2 subunit I domain in regulation L 2K174T-CHO cells, respectively). Wild-type L 2 and the ␣ ␤ ␤ ␣ ␤ ␣ ␤ of L 2, we generated another 2 subunit mutant, in which we L 2-3-2 and L 2K174T mutants were expressed in CHO cells at replaced the ␤B-␤C disulfide-linked loop sequence, CPNKEKEC comparable levels (Fig. 3B). The divalent cation Mg2ϩ is required ␤ ␣ ␤ 2298 THE CPNKEKEC SEQUENCE OF THE 2 SUBUNIT AND L 2 ACTIVATION

mAb 24, indicating that Glu175 is critical for mAb 24 binding. These results indicate that Glu173 and Glu175 are critically in- ␤ volved in mAb 24 binding. We also found that the 2-3-2, E173K, ␤ E175T, and E175A mutations similarly blocked binding of anti- 2 ␣ ␤ mAb IB-4 to L 2. This suggests that the IB-4 and mAb 24 epitopes overlap within the predicted loop 170Ð175. 175 ␤ We introduced mutation of Ala to Glu in the mouse 2 subunit and tested whether this mutation generates the mAb 24 ␤ ␤ epitope in mouse 2. We expressed the A175E mouse 2 mutant ␣ on CHO cells together with wild-type human L. We found that ␣ ␤ the human L/mouse 2(A175E) bound to mAb IB-4 but did not bind to mAb 24 (data not shown). These results suggest that the A175E mutation is enough for generating the IB-4 epitope but is ␤ not enough for generating the mAb 24 epitope in the mouse 2 subunit. It has been reported previously that mAb 24 recognizes an ␤ ␣ epitope common to several 2 integrin subunits (6Ð9). We tested ␣ whether the L I domain is required for mAb 24 binding. We found that deletion of the whole I domain did not block mAb 24 Downloaded from ␣ binding (Fig. 5). Several residues in the MIDAS of the L I domain (Asp137, Thr206, and Asp239) have been reported to be critical for ␣ ␤ L 2-ICAM interaction (26, 32). We found that mutating these ␣ MIDAS residues in the L I domain had only a minimal effect on mAb 24 binding (Thr208 was used as a control). These results

suggest that mAb 24 binding does not require the ␣L I domain. http://www.jimmunol.org/ ␤ FIGURE 2. The structure of the I-like domain sequence of the subunit This confirms previous findings showing that mAb 24 does not ␤ emphasizing the disulfide-linked loop C169-C176. The structure of the 3 ␣ ␤ recognize the isolated I domain (33) and is still expressed by L 2 I domain-like structure was taken from a recent study (18). The position of ␣ ␤ ␤ when the I domain is deleted in Jurkat cell L 2 transfectants (34). the loop (residues 159Ð182 in 2) is shown by an arrow and aligned with ␤ ␤ the corresponding 1 and 3 sequences. The diverse disulfide-linked se- ␤ ␤ ␤ ␤ Discussion quences (residues 169Ð176 in 2)in 1, 2, and 3 are boxed. We swapped ␤ ␤ residues 169Ð176 of 2 with the corresponding sequence of 3 (designated In the present study, we establish that the CPNKEKEC sequence ␤ ␤ ␣ ␤ 2-3-2 mutant). of 2 may be involved in the regulation of L 2 ligand-binding ␤ activity. We have shown that the 2-3-2 or the K174T mutation ␣ ␤ by guest on September 27, 2021 within the CPNKEKEC sequence constitutively activates L 2. ␣ ␤ 2ϩ ␣ ␤ ␣ for the adhesiveness of L 2, and Ca bound to L 2 may serve This is the first study to show that the L I domain can be activated ␤ to maintain an inactive state (7). We determined the capacity of by mutation of the 2 I domain. This observation is consistent with ␣ ␤ ␣ ␤ ␤ ␣ L 2-3-2- and L 2K174T-CHO cells to adhere to a range of the idea that the 2 I domain can indirectly regulate the L Ido- 2ϩ ␣ ␤ ICAM-1 concentrations in the presence of Mg to activate L 2 main. It is interesting to speculate that the CPNKEKEC sequence ␣ ␤ ␣ ␤ ␤ ␣ (Fig. 3C). We found that the L 2-3-2 and L 2K174T showed of the 2 subunit may make direct contact with the L subunit (the ␣ ␤ much higher adhesion to ICAM-1 than wild-type L 2. These re- I domain or the ␤-propeller) and that this interaction would keep ␤ ␣ ␤ sults suggest that the 2-3-2 and K174T mutations constitutively the I domain in an inactive conformation. When L 2 is activated, ␣ ␤ 2ϩ ␣ activated L 2. We next tested whether Ca was able to suppress the sequence would be detached from the L subunit and the mAb ␤ ␣ the activating effects of the 2-3-2 and the K174T mutation on 24 epitope would be expressed. It has been reported that an L 2ϩ ␣ ␤ ␣ ligand binding, because Ca has an inhibitory effect on L 2- peptide (residues 238Ð254 of L) blocks the appearance of the 2ϩ ␣ ␤ ICAM-1 interaction (7). Although Ca has an inhibitory effect on mAb 24 epitope even when L 2 is activated (35). Because this ␣ ␤ ␣ wild-type L 2, it did not show any inhibitory effect on the two peptide sequence is exposed on the surface of the L subunit, this mutants (Fig. 3D). sequence may be a potential contact site for the CPNKEKEC sequence. As the K174T mutation also causes constitutive activation ␤ ␤ The epitope for mAb 24 is located in the B- C disulfide-linked of LFA-1, as described in this study, it may be particularly im- ␤ loop sequence of the 2 subunit portant in maintaining the “off” contact of the CPNKEKEC loop. ␤ Expression of the mAb 24 epitope is associated with 2 integrin This mutation has been found as a missense mutation in a leuko- ␣ ␤ activation (2, 6, 7). The L 2-3-2 mutant was tested for its reac- cyte adhesion deficiency-1 patient (36). ␣ ␤ tivity with anti- L 2 Abs by flow cytometry. We found that re- Another possible mechanism of the constitutive activation of the ␤ ␤ ␤ ␣ ␤ placement of the 2 subunit I domain B- C disulfide-linked loop L I domain by the 2-3-2 and K174T mutations is that these mu- ␤ ␣ ␤ with the homologous loop from the 3 subunit eliminates reactiv- tations change the divalent cation-binding properties of L 2.It ity with mAb 24 (Fig. 4). This result suggests that this mutation has been proposed that extracellular Ca2ϩ can regulate the function ␣ ␤ 2ϩ may have destroyed the mAb 24 epitope. To test this possibility, of integrins, including L 2. To date two distinct classes of Ca ␤ ␤ we examined the reactivity of the 2-to- 3 mutants with mAb 24. binding sites that differ in their affinity for the metal ion have been We found that the E173K and E175T mutations block mAb 24 characterized (reviewed in Ref. 37). It has been proposed that a binding (Fig. 3). We also introduced human-to-mouse mutations high-affinity Ca2ϩ binding site promotes binding to ligands and ␤ 2ϩ within the swapped region of 2. There is only one residue dif- that a low-affinity site appears to compete with a Mg occupied ␤ ␤ ference between human and mouse 2 at position 175 (Glu in site (37Ð39). The homologous 3 I domain contains the high-af- human and Ala in mouse) within the loop 170Ð175. We found that finity Ca2ϩ binding site that promotes ligand binding and the low- the Glu175 to Ala mutation (E175A) obliterated the binding of affinity Ca2ϩ binding site that is inhibitory for ligand binding (37, The Journal of Immunology 2299

␤ FIGURE 4. Effect of the 2-3-2 mutation and point mutations within residues 170Ð175 on the reactivity of activation-dependent mAb 24. The ␤ 2-3-2 mutant was transiently transfected into CHO cells together with ␣ ␤ wild-type L. In addition, individual residues of the 2 subunit were

␤ Downloaded from changed to the corresponding 3 residues (the P170Y, N171D, K172 M, E173K, K174T, and E175T mutations), or changed to the corresponding ␤ residue in mouse 2 (the E175A mutation). The point mutants were tran- ␣ siently expressed on CHO cells together with wild-type L. The transfected cells were stained with mAb 24 in the presence of 0.1 mM Mn2ϩ, or with ␤ mAbs IB-4 and MEM-48 (anti- 2), followed by FITC-labeled anti-mouse IgG, and analyzed by ﬂow cytometry. The data are expressed as percentage of positive cells using ﬂow cytometry. http://www.jimmunol.org/

40). It has been reported that mutations in the MIDAS residues in ␤ 2ϩ 2ϩ the 3 I domain, which affect Mg /Mn binding, leave the inhibitory Ca2ϩ binding site intact (41), suggesting that the inhibitory low-affinity Ca2ϩ binding site is distinct from the MIDAS-like ␣ ␤ motif. The recent v 3 crystal structure shows that there is an ␤ additional cation-binding site within the 3 I domain adjacent to MIDAS (designated ADMIDAS) (20). The Ca2ϩ-binding affinity by guest on September 27, 2021 ␤ of the ADMIDAS is unclear. In the present study, the 2-3-2 and ␣ ␤ the K174T mutations make L 2 insensitive to the inhibitory effect of Ca2ϩ. One possible explanation is that these mutations block 2ϩ 2ϩ ␤ access of Ca to the low-affinity Ca site of the 2 I domain. The CPNKEKEC sequence and the ADMIDAS-like motif in the ␤ ␤ 2 subunit are both in the upper face of the 2 I domain and in close proximity to each other (20). It is interesting to speculate that removal of Ca2ϩ from this site and corresponding addition of Mg2ϩ to the MIDAS site may be the basis of the Mg2ϩ/EGTA- induced activation of LFA-1 (which also correlates with mAb 24 expression (7)). A dynamic relationship between the ␤B-␤C loop sequence CPNKEKEC and ADMIDAS may be an essential part of the allosteric control of LFA-1, leading to conformational change associated with increased ICAM-1 binding. The present study establishes that the CPNKEKEC sequence of ␤ ␤ the 2 subunit is a part of the mAb 24 epitope (a reporter for 2

␣ ␤ ␤ FIGURE 3. Constitutive activation of L 2 by the 2-3-2 and the K174T ␤ mutations. A, Several 2 mutants within residues 170Ð175 were transiently ␣ expressed on K562 cells together with wild-type L and tested for their ability to bind to ICAM-1 in adhesion assays. The transfected cells were ␣ stained with TS1/22 (anti- L), followed by FITC-labeled anti-mouse IgG, ␣ ␤ and analyzed by ﬂow cytometry to monitor expression of L 2 mutants. B, ␣ ␤ Expression of wild-type and mutant L 2 in cloned CHO cell lines was ␣ ␤ determined using ﬂow cytometry. C, CHO cells expressing the L 2 mutants were tested for their ability to bind to ICAM-1 in the presence of2mMMg2ϩ as a function of ICAM-1-coating concentrations. D, The 2ϩ ␣ ␤ effect of Ca on ICAM-1 binding to the L 2 mutants on CHO cells was tested by adding various concentrations of Ca2ϩ to the medium containing 2 mM Mg2ϩ. ␤ ␣ ␤ 2300 THE CPNKEKEC SEQUENCE OF THE 2 SUBUNIT AND L 2 ACTIVATION

2. Stewart, M., and N. Hogg. 1996. Regulation of leukocyte integrin function: affinity vs. avidity. J. Cell. Biochem. 61:554. 3. Lub, M., Y. van Kooyk, and C. G. Figdor. 1995. Ins and outs of LFA-1. Immunol. Today 16:479. 4. Harris, E. S., T. M. McIntyre, S. M. Prescott, and G. A. Zimmerman. 2000. The leukocyte integrins. J. Biol. Chem. 275:23409. 5. Diamond, M. S., and T. A. Springer. 1994. The dynamic regulation of integrin adhesiveness. Curr. Biol. 4:506. 6. Dransfield, I., and N. Hogg. 1989. Regulated expression of Mg2ϩ binding epitope on leukocyte integrin ␣ subunits. EMBO J. 8:3759. 7. Dransfield, I., C. Cabanas, A. Craig, and N. Hogg. 1992. Divalent cation regulation of the function of the leukocyte integrin LFA-1. J. Cell Biol. 116:219. 8. Dransfield, I., A. M. Buckle, and N. Hogg. 1990. Early events of the immune response mediated by leukocyte integrins. Immunol. Rev. 114:29. 9. Dransfield, I., C. Cabanas, J. Barrett, and N. Hogg. 1992. Interaction of leukocyte integrins with ligand is necessary but not sufficient for function. J. Cell Biol. 116:1527. 10. Qu, A., and D. Leahy. 1995. Crystal structure of the I-domain of the CD11a/ ␣ ␤ CD18 (LFA-1, L 2) integrin. Proc. Natl. Acad. Sci. USA 92:10277. 11. Lee, J. O., P. Rieu, M. A. Arnaout, and R. Liddington. 1995. Crystal structure of the A domain from the ␣ subunit of integrin CR3 (CD11b/CD18). Cell 80:631. 12. Lee, J. O., L. A. Bankston, M. A. Arnaout, and R. C. Liddington. 1995. Two ␤ ␣ conformations of the integrin A-domain (I-domain): a pathway for activation? FIGURE 5. Effect of mutations in the 2 and L I domains that are ␣ Structure 3:1333. critical for ligand binding on the reactivity to mAb 24. Several L point 13. Emsley, J., C. G. Knight, R. W. Farndale, M. J. Barnes, and R. C. Liddington. Downloaded from ␣ ␣ ␤ mutants in the MIDAS of the L I domain that are critical for ligand 2000. Structural basis of collagen recognition by integrin 2 1. Cell 101:47. binding (26) were transiently expressed in CHO cells together with wild- 14. Kallen, J., K. Welzenbach, P. Ramage, D. Geyl, R. Kriwacki, G. Legge, ␤ 2ϩ S. Cottens, G. Weitz-Schmidt, and U. Hommel. 1999. Structural basis for LFA-1 type 2. Cells were stained with mAb 24 in the presence of 0.5 mM Mn , ␣ ␤ inhibition upon lovastatin binding to the CD11a I-domain. J. Mol. Biol. 292:1. MEM-83 (anti- L), or IB-4 (anti- 2), as described above. Data are ex- 15. Shimaoka, M., J. M. Shifman, H. Jing, J. Takagi, S. L. Mayo, and T. A. Springer. pressed as percentage of positive cells in flow cytometry. 2000. Computational design of an integrin I domain stabilized in the open high affinity conformation. Nat. Struct. Biol. 7:674. 16. Lu, C., M. Shimaoka, Q. Zang, J. Takagi, and T. A. Springer. 2001. Locking in 173 ␣ ␤ http://www.jimmunol.org/ integrin activation) and that, within this sequence, Glu and alternate conformations of the integrin L 2 I domain with disulfide bonds re- 175 veals functional relationships among integrin domains. Proc. Natl. Acad. Sci. Glu are particularly critical. These findings are consistent with USA 98:2393. a recent report by Lu et al. (16) and clearly indicate that exposure 17. Lu, C., M. Shimaoka, M. Ferzly, C. Oxvig, J. Takagi, and T. A. Springer. 2001. ␣ ␤ An isolated, surface-expressed I domain of the integrin ␣ ␤ is sufficient for of the mAb 24 epitope upon L 2 activation reflects conforma- L 2 ␤ ␤ strong adhesive function when locked in the open conformation with a disulfide tional change of the 2 I domain. In the I domain (20), the bond. Proc. Natl. Acad. Sci. USA 98:2387. CPNKEKEC sequence is in the loop protruding from the upper 18. Xiong, J.-P., T. Stehle, B. Diefenbach, R. Zhang, R. Dunker, D. L. Scott, J. Andrzej, S. L. Goodman, and M. A. Arnaout. 2001. Crystal structure of the face of the globular domain. We propose that the mAb 24 epitope ␣ ␤ ␣ ␤ extracellular segment of integrin v 3. Science 294:339. may be located at the boundary between the L and 2 subunits. 19. Takagi, J., T. Kamata, J. Meredith, W. Puzon-McLaughlin, and Y. Takada. 1997. ␣ ␤ ␣ ␤ Changing ligand specificity of v 1 and v 3 integrins by swapping a short However, it is unclear whether the exposure of the mAb 24 epitope by guest on September 27, 2021 ␤ is due to changes in the domain-domain interaction or due to con- diverse sequence of the subunit. J. Biol. Chem. 272:19794. ␤ ␤ 20. Puzon-McLaughlin, W., T. Kamata, and Y. Takada. 2000. Multiple discontinuous formational changes in the 2 I domain. It is possible that the 2 ligand-mimetic antibody binding sites define a ligand binding pocket in integrin ␣ ␤ I domain undergoes conformational changes on activation. We IIb 3. J. Biol. Chem. 275:7795. 21. Wallace, J. L., K. E. Arfors, and G. W. McKnight. 1991. A monoclonal antibody have previously reported that activating and inhibiting mAbs against the CD18 leukocyte adhesion molecule prevents indomethacin-induced ␤ against the homologous integrin 1 subunit recognize overlapping gastric damage in the rabbit. Gastroenterology 100:878. ␤ 22. Sanchez-Madrid, F., A. M. Krensky, C. F. Ware, E. Robbins, J. L. Strominger, epitopes within residues 207Ð218 of the 1 I domain (42). These ␤ ␤ S. J. Burakoff, and T. A. Springer. 1982. Three distinct antigens associated with anti- 1 mAbs induce conformational changes in the 1 I domain human T lymphocyte-mediated cytolysis: LFA-1, LFA-2 and LFA-3. Proc. Natl. ␤ that either activate or inactivate the 1 integrins. Thus, changes in Acad. Sci. USA 79:7489. ␤ 23. Bazil, V., I. Stefanova, I. Hilgert, H. Kristofova, S. Vanek, and V. Horejsi. 1990. domain-domain interaction and in the conformation of the 2 I ␣ ␤ Monoclonal antibodies against human leucocyte antigens. IV. Antibodies against domain may also occur simultaneously on L 2 activation. subunits of the LFA-1 (CD11a/CD18) leukocyte-adhesion glycoprotein. Folia The present study establishes that a number of conserved resi- Biol. 36:41. ␤ ␣ ␤ 24. Lin, A. Y., B. Devaux, A. Green, C. Sagerstrom, J. F. Elliott, and M. Davis. 1990. dues in the 2 I domain are critical for ICAM-1 binding to L 2 ␣ ␤ Expression of T cell antigen receptor heterodimers in a lipid-linked form. Science and the exposure of the mAb 24 epitope on activation of L 2. 249:677. These residues include several that have not been tested in other ␤ 25. Deng, W. P., and J. A. Nickoloff. 1992. Site-directed mutagenesis of virtually any 116 120 151 ␤ plasmid by eliminating a unique site. Anal. Biochem. 200:81. subunits (Ser , Asp , and Asp in 2). Interestingly, all of the 26. Kamata, T., R. Wright, and Y. Takada. 1995. Critical Thr and Asp residues within ␤ ␤ ␤ ␤ critical 2 residues are clustered in the 2 I domain in a recent the I domains of 2 integrins for interactions with ICAM-1 and C3bi. J. Biol. ␤ Chem. 270:12531. I domain structure (20). If the 2 subunit I domain is indirectly ␤ 27. Prater, C. A., J. Plotkin, D. Jaye, and W. A. Frazier. 1991. The properdin-like involved in ICAM-1 binding (43), these 2 residues may contrib- type I repeats of human thrombospondin contain a cell attachment site. J. Cell ute to ligand binding through possible domain-domain interaction, Biol. 112:1031. ␤ ␤ 28. Bajt, M., T. Goodman, and S. McGuire. 1995. 2 (CD18) mutations abolish a possible conformational change, and/or cation binding in the 2 ␣ ␤ ligand recognition by I domain integrins LFA-1 ( L 2, CD11a/CD18) and I domain, rather than direct interaction with ICAM-1. Further stud- ␣ ␤ MAC-1 ( M 2, CD11b/CD18). J. Biol. Chem. 270:94. ies will be required to determine how the conformation of the ␣ 29. Goodman, T. G., and M. L. Bajt. 1996. Identifying the putative metal ion-de- ␤ ␣ ␤ ␣ ␤ subunit I domain is regulated by the ␤ subunit I domain on pendent adhesion site in the 2 (CD18) subunit required for L 2 and M 2 ligand interactions. J. Biol. Chem. 271:23729. activation. 30. Takada, Y., J. Ylanne, D. Mandelman, W. Puzon, and M. Ginsberg. 1992. A point ␤ ␣ ␤ mutation of integrin 1 subunit blocks binding of 5 1 to fibronectin and invasin but not recruitment to adhesion plaques. J. Cell Biol. 119:913. Acknowledgments 31. Puzon-McLaughlin, W., and Y. Takada. 1996. Critical residues for ligand binding ␤ We thank V. Horejsi and G. Weitz for valuable reagents. in the integrin 1 subunit. J. Biol. Chem. 271:20438. 32. Edwards, C. P., M. Champe, T. Gonzalez, M. E. Wessinger, S. A. Spencer, L. G. Presta, P. W. Berman, and S. C. Bodary. 1995. Identification of amino acids References in the CD11a I-domain important for binding of the leukocyte function-associated 1. Springer, T. A. 1994. Traffic signals for lymphocyte recirculation and leukocyte antigen-1 (LFA-1) to intercellular adhesion molecule-1 (ICAM-1). J. Biol. Chem. emigration: the multistep paradigm. Cell 76:301. 270:12635. The Journal of Immunology 2301

␤ ␣ ␤ 33. Randi, A. M., and N. Hogg. 1994. I domain of 2 integrin lymphocyte function ing of collagen by the integrin 2 1 in human platelets. J. Biol. Chem. 275: associated antigen-1 contains a binding site for ligand intercellular adhesion mol- 24560. ecule-1. J. Biol. Chem. 269:12395. 39. Labadia, M. E., D. D. Jeanfavre, G. O. Caviness, and M. M. Morelock. 1998. 34. Leitinger, B., and N. Hogg. 2000. Effects of I domain deletion on the function of Molecular regulation of the interaction between leukocyte function-associated ␤ antigen-1 and soluble ICAM-1 by divalent metal cations. J. Immunol. 161:836. the 2 integrin lymphocyte function-associated antigen-1. Mol. Biol. Cell 11:677. 35. McDowall, A., B. Leitinger, P. Stanley, P. A. Bates, A. M. Randi, and N. Hogg. 40. Cierniewski, C. S., T. Byzova, M. Papierak, T. A. Haas, J. Niewiarowska, L. Zhang, M. Cieslak, and E. F. Plow. 1999. Peptide ligands can bind to distinct 1998. The I domain of integrin leukocyte function-associated antigen-1 is in- ␣ ␤ volved in a conformational change leading to high affinity binding to ligand sites in integrin IIb 3 and elicit different functional responses. J. Biol. Chem. intercellular adhesion molecule 1 (ICAM-1). J. Biol. Chem. 273:27396. 274:16923. 41. Lin, E. C. K., B. I. Ratnikov, P. M. Tsai, E. R. Gonzalez, S. McDonald, 36. Arnaout, M. A., N. Dana, S. K. Gupta, D. G. Tenen, and D. M. Fathallah. 1990. ␤ ␤ ␤ A. J. Pelletier, and J. W. Smith. 1997. Evidence that the integrin 3 and 5 Point mutations impairing cell surface expression of the common subunit subunits contain a metal ion-dependent adhesion site-like motif but lack an I (CD18) in a patient with leukocyte adhesion molecule (Leu-CAM) deficiency. domain. J. Biol. Chem. 272:14236. J. Clin. Invest. 85:977. 42. Takada, Y., and W. Puzon. 1993. Identification of a regulatory region of integrin 37. Leitinger, B., A. McDowall, P. Stanley, and N. Hogg. 2000. The regulation of ␤ subunit using activating and inhibiting antibodies. J. Biol. Chem. 268:17597. 2ϩ 1 integrin function by Ca . Biochim. Biophys. Acta 1498:91. 43. Huang, C., Q. Zang, J. Takagi, and T. A. Springer. 2000. Structural and func- ␤ 38. Onley, D. J., C. G. Knight, D. S. Tuckwell, M. J. Barnes, and R. W. Farndale. tional studies with antibodies to the integrin 2 subunit: a model for the I-domain. 2000. Micromolar Ca2ϩ concentrations are essential for Mg2ϩ-dependent bind- J. Biol. Chem. 275:21514. Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021