Permissive transmembrane helix heterodimerization is required for the ...

Biochem. J. (2008) 410, 495–502 (Printed in Great Britain)

495

doi:10.1042/BJ20071218

Permissive transmembrane helix heterodimerization is required for the expression of a functional integrin Ardcharaporn VARARATTANAVECH, Man-Li TANG, Hoi-Yeung LI, Chi-Hang WONG, S. K. Alex LAW, Jaume TORRES1 and Suet-Mien TAN1 School of Biological Sciences, Nanyang Technological University, 60 Nanyang Drive, Singapore 637551, Singapore

The current paradigm is that integrin is activated via insideout signalling when its cytoplasmic tails and TMs (transmembrane helices) are separated by specific cytosolic protein(s). Perturbations of the helical interface between the α- and βTMs of an integrin, as a result of mutations, affect its function. Previous studies have shown the requirement for specific pairing between integrin subunits by ectodomain-exchange analyses. It remains unknown whether permissive α/β-TM pairing of an integrin is also required for pairing specificity and the expression of a functionally regulated receptor. We performed scanning replacement of integrin β2-TM with a TM of other integrin βsubunits. With the exception of β4 substitution, others presented β2-integrins with modified phenotypes, either in their expression or ligand-binding properties. Subsequently, we adopted αLβ2 for follow-on experiments because its conformation and affinitystate transitions have been well defined as compared with other members of the β2-integrins. Replacement of β2- with β3-

TM generated a chimaeric αLβ2 of an intermediate affinity that adhered to ICAM-1 (intercellular adhesion molecule 1) but not to ICAM-3 constitutively. Replacing αL-TM with αIIbTM, forming a natural αIIb/β3-TM pair, reversed the phenotype of the chimaera to that of wild-type αLβ2. Interestingly, the replacement of αLβ2- with β3-TM showed neither an extended conformation nor the separation of its cytoplasmic tails, which are well-reported hallmarks of an activated αLβ2, as determined by reporter mAb (monoclonal antibody) KIM127 reactivity and FRET (fluorescence resonance energy transfer) measurements respectively. Collectively, our results suggest that TM pairing specificity is required for the expression of a functionally regulated integrin.

INTRODUCTION

β2-integrin heterodimer formation or led to precocious ligandbinding activity, suggesting that specific pairing between αand β-subunits is required for the expression of a functionally regulated receptor [8,9]. Whether specific pairing of integrin TMs is required to maintain integrin functionality has not been addressed. The TM maintains the connectivity between integrin ectodomain and its cytoplasmic tail. The disruption of α/β-TM association leads to integrin activation. For example, asparagine substitution studies have shown that TM mutations G972N [10] or G708N [11] at αIIb or β3 respectively led to activated αIIbβ3, concomitant with enhanced homo-oligomeric clustering. In a separate study, disulfide stabilization of the α/β-TMs abolished inside-out activation of αIIbβ3 [12]. Studies are carried out to define the mode of interaction and the orientation of the integrin TMs under resting and activated conditions. A model of α/β-TM interaction under resting conditions has been proposed previously, where interaction is mediated by a GXXXG-like motif located in both TMs [13], similar to that of the homodimer GpA (glycophorin A) [14]. However, others studies report differences in the orientation of the β-TM relative to the α-TM [12,15]. Instead the GpA-like model of integrin α/β-TM interaction may represent a transitory step that precedes the final separation of the TMs during integrin activation [12,16,17]. We have performed an exhaustive conformational search for these TM interactions and found two possible modes of α/β-TM association, which are consistent with those described above [18]. Curiously, many combinations of α/β-TM pairing are possible,

Integrins are heterodimeric type I membrane adhesion molecules formed by non-covalent association of an α- and β-subunits [1]. They are bona fide signalling receptors, which mediate bidirectional signal transduction. The array of signalling pathways emanating from integrins serve a variety of biological processes encompassing cell movement, growth and differentiation. Each integrin subunit consists of a large extracellular domain, a TM (transmembrane helix) and a cytoplasmic tail. Conformational changes of the integrin ectodomain are associated with its activation status that impinges on its ligand-binding capacity [2,3]. The trigger point for integrin inside-out activation is the separation of its α/β cytoplasmic tails, promoted by interaction with cytosolic protein talin [4]. Phosphorylations of the integrin cytoplasmic tails are also found to be important events that lead to integrin activation and cytoskeleton remodelling [5,6]. In humans, 24 specific combinations of integrin heterodimers have been identified [1]. Although structural studies of integrin ectodomain provide useful insights into integrin function regulation, it is still largely unknown how specific pairing between integrin α- and β-subunits is achieved [2,3,7]. This is intriguing, considering the fact that many different integrins can be expressed concomitantly in the same cell type but specific heterodimer formation is still maintained. Previously, we found that replacing regions of the integrin β2-subunit ectodomain with corresponding segments from integrin β1 or β7 either affected

Key words: αLβ2, cytoplasmic tail, heterodimerization, intercellular adhesion molecule (ICAM), integrin, transmembrane helix.

Abbreviations used: FRET, fluorescence resonance energy transfer; GpA, glycophorin A; HEK-293 cells, human embryonic kidney cells; HEK-293T cells, HEK-293 cells expressing the large T-antigen of SV40 (simian virus 40); HI-FBS, heat-inactivated fetal bovine serum; ICAM, intercellular adhesion molecule; mAb, monoclonal antibody; mCFP, monomeric cyan fluorescent protein; YFP, yellow fluorescent protein; mYFP, monomeric YFP; ROI, region of interest; SDM, site-directed mutagenesis; TM, transmembrane helix. 1 Correspondence may be addressed to either of these authors (email [email protected] or [email protected]). c The Authors Journal compilation c 2008 Biochemical Society

496

A. Vararattanavech and others

including non-native combinations of α/β-TMs [18]. However, these computational studies, which focused only on the TMs, do not provide information on the requirement of permissive TM pairing with respect to integrin functionality. To address the latter, we generated TM chimaeras of integrin αLβ2. Our results suggest that parallel evolution of the α- and β-TMs of an integrin heterodimer is required, not only for functional TM–TM interaction, but also for the proper biosynthesis of a functionally regulated receptor.

which mCFP or mYFP was tethered to the C-terminus of the integrin cytoplasmic tails. A 5-amino-acid linker (GPVAT) was introduced between αL or αL(R1096D) and mCFP, and a 6amino-acid linker (GGPVAT) was inserted between β2 or β23 and mYFP, based on the previous observation that αL-mCFP and β2-mYFP containing 5- and 6-residue linkers exhibited high FRET efficiency [28]. Constructs were verified by sequencing (Research Biolabs). Cell culture and transfection

MATERIALS AND METHODS Antibodies and reagents

The following mAbs (monoclonal antibodies) are gifts from different sources: KIM185 (a β2-specific and activating mAb) [19] and KIM127 (a β2-specific and activation reporter mAb) [20,21] were gifts from M. Robinson (CellTech, Slough, U.K.). MHM24 [22] (an αL-specific and function-blocking mAb) [22] was from A.J. McMichael (Institute of Molecular Medicine, Oxford, U.K.). IB4 (a β2-integrins heterodimer-specific and functionblocking mAb) [23] was obtained from A.T.C.C. (Manassas, VA, U.S.A.). H52 (a β2-specific mAb) was described previously [24]. Recombinant human ICAM-1 (intercellular adhesion molecule 1)/Fc and ICAM-3/Fc were prepared as described previously [25]. All general chemicals and reagents were from Sigma unless otherwise indicated.

HEK-293T cells [HEK-293 cells (human embryonic kidney cells) expressing the large T-antigen of SV40 (simian virus 40); A.T.C.C.] were cultured in DMEM (Dulbecco’s modified Eagle’s medium) containing 10 % (v/v) HI-FBS (heat-inactivated fetal bovine serum), 100 i.u./ml penicillin and 100 µg/ml streptomycin (Hyclone). The α- and β-integrin constructs were co-transfected into HEK-293T cells using the Polyfect transfection reagent (Qiagen). K562 cells were transfected with the integrin expression plasmids by electroporation using the Amaxa Nucleofector device and reagents as per the manufacturer’s instructions (Amaxa). Flow-cytometric analysis

Analyses of integrin expression on the transfectants were performed as described previously [25]. Primary mAb IB4 (20 µg/ ml) was used for each sample analysed. Cells were analysed on an FACSCaliburTM using the software CellQuest (Becton Dickinson Biosciences).

cDNA expression plasmids

The numbering of the integrin amino acids is based on Barclay et al. [26]. Integrin αL, αM, αX and β2 pcDNA3 expression plasmids were described previously [24]. The β2-TM was replaced with the corresponding TM of other integrin β-subunits by two consecutive procedures. First, SDM (site-directed mutagenesis), using the QuikChange® SDM kit (Stratagene) with relevant primers, was performed on wild-type β2 to generate β2HA having the HpaI and AflII sites encompassing the TM and part of the β2 cytoplasmic tail sequences. The first set of primers was designed to generate by PCR the TM of one of the integrin βsubunits TM. In addition, the 5 -end of the forward primer contained an HpaI site. The 5 -end of the reverse primer also contained complementary sequence to the β2 cytoplasmic tail. The second set of primers was designed to amplify the β2 cytoplasmic tail. The 5 -end of the forward primer also contained complementary sequence to the TM of one of the other β-subunits. The 5 -end of the reverse primer contained the AflII site. PCR products from both sets of primers were used for overlapping extension PCR to generate a chimaeric sequence containing the β2 cytoplasmic tail with a TM from another β-subunit. The product was digested with HpaI and AflII and subcloned into the similarly digested β2HA . Subsequently, the HpaI site in the chimaera construct was mutated by SDM for reversion to wild-type sequence. A similar approach was adopted for the replacement of αL-TM with that of αIIb. All TM chimaeras were subsequently verified by sequencing (Research Biolabs, Singapore). For FRET (fluorescence resonance energy transfer) experiments, the mammalian expression vectors pEYFP-N1 and pECFP-N1 (Clontech) were used. However, to inhibit their inherent tendency to form hetero- or homo-dimers, mCFP (monomeric cyan fluorescent protein) and mYFP (monomeric yellow fluorescent protein) were generated by replacing Leu221 , at the crystallographic dimer interface, with lysine [27]. mCFP and mYFP were subcloned into integrin expression vectors to generate αL-mCFP, αL(R1094D)-mCFP, β2-mYFP and β23-mYFP, in c The Authors Journal compilation c 2008 Biochemical Society

Cell-surface protein biotinylation and immunoprecipitation

Labelling of cell-surface proteins with biotin was performed essentially as described previously [29], with slight modifications. Briefly, HEK-293T transfectants were washed twice with PBS and cell-surface proteins labelled with sulfo-NHS (Nhydroxysuccinimido)-biotin (Pierce) at 0.5 mg/ml in PBS for 15 min at room temperature. Reaction was terminated by washing cells with PBS containing 10 mM Tris/HCl (pH 8.0) and 0.1 % (w/v) BSA. Labelled cells were incubated in medium containing 5 % HI-FBS and 10 mM Hepes with the relevant mAb (2 µg) at 37 ◦C for 30 min. Cells were washed twice with medium to remove unbound mAb and lysed by incubating in lysis buffer [10 mM Tris/HCl, pH 8.0, 150 mM NaCl and 1 % (v/v) Nonidet P40] containing appropriate protease inhibitors at 4 ◦C for 30 min. Immunoprecipitation was performed using rabbit anti-mouse IgG (Sigma) coupled with Protein A–Sepharose beads (Amersham Bioscience) as described previously [25]. Bound proteins were resolved on SDS/7.5 % PAGE under reducing conditions, and electroblotted on to Immobilon-P membrane (Millipore). Biotinylated protein bands were detected with streptavidin–HRP (horseradish peroxidase) followed by enhanced chemiluminescence detection using the ECL® -plus kit (Amersham Biosciences). Cell adhesion assays

Adhesion of transfectants to immobilized ICAMs (for αLβ2 studies) or BSA (for αMβ2 and αXβ2) studies was performed essentially as described previously [25,30,31]. Briefly, in the study using ICAM, each Polysorb microtitre well (Nunc) was first coated with 0.5 µg of goat anti-human IgG Fc-specific (Sigma) in 50 mM bicarbonate buffer (pH 9.2), blocked with 0.5 % BSA in PBS, and followed by 50 ng of ICAM/Fc in PBS. For adhesion to BSA, each well was coated with 0.01 % BSA in bicarbonate buffer, and non-specific sites blocked with 0.2 %

Functional integrin requires permissive transmembrane helix pairing

497

(w/v) polyvinylpyrrolidone 10 in PBS. BCECF [2 ,7 -bis-(2carboxyethyl)-5(6)-carboxyfluorescein; from Molecular Probes]labelled transfectants (∼ 2 × 104 per well) were allowed to adhere to the ligand-coated well in a medium containing 5 % HI-FBS and 10 mM Hepes (pH 7.4) (referred to as wash medium) at 37 ◦C for 30 min in a humidified 5 % CO2 incubator. After washing twice with wash medium, fluorescence signal, which corresponds to number of adherent cells, was measured with a fluorescent plate reader (FL600) (Bio-Tek Instruments). FRET analyses

K562 transfectants expressing a FRET fluorophore pair conjugated to integrin cytoplasmic tails were cytospun on to glass slides. FRET detection by acceptor photobleaching was performed on a Zeiss LSM510 confocal microscope (Carl Zeiss) to detect integrin cytoplasmic tails separation [32]. The following parameters were used for analyses: (i) mCFP: λex = 458 nm; emission filter BP, 470–500 nm; (ii) mYFP: λex = 514 nm; emission filter LP, 530 nm; and (iii) oil immersion × 63 objective. Photobleaching of mYFP of an entire cell within was achieved by scanning the region 20 times using the 514 argon laser line set at the maximum intensity. The cell membrane was selected as the ROI (region of interest). mCFP signals within the ROI pre- and post-mYFP bleaching were acquired. FRET efficiency (EF ) was calculated as a percentage using the equation EF = (I 6 − I 5 ) × 100/I 6 , where I n is the mCFP intensity at the nth time point. Bleaching was performed between the fifth and sixth time points. Similar analyses of unbleach cells using the equation CF = (I 6 − I 5 ) × 100/I 6 were made. The mean noise computed was N F = (I 5 − I 4 ) × 100/I 5 in which the mCFP signals at the 4th and 5th time points before the bleaching process were close to zero in all cases.

RESULTS Effect of β2-TM substitution on heterodimer formation with αL, αM and αX

Previously, we have examined integrin heterodimer pairing specificity via generation of chimaeras having integrin β2 ectodomain sequences replaced with corresponding sequences from β1- and β7-subunits [8]. To determine whether specific TM pairing is required for correct integrin heterodimer function, the TM of the β2-subunit was replaced with corresponding TM segments from other β-subunits (Figure 1A). Henceforth, β2-TM replaced with β3-TM, for example, is denoted as β23. The two chimaeric integrins αLβ23 and αIIbβ23 that serve as the focus of this investigation are shown (Figure 1B). HEK-293T cells were co-transfected with αL- and β2-TM chimaera expression plasmids, and the cell-surface expression of the heterodimer was examined by flow cytometry by using mAb IB4, which is specific for the β2-integrin heterodimer (Figure 1C). The level of surface expression for the heterodimers containing β2-TM chimaeras associated with αL was comparable with that of wild-type αLβ2, with the exception of the αLβ28 transfectant. To further verify the formation of integrin heterodimer, we performed biotin labelling of cell-surface proteins, followed by immunoprecipitation with mAb KIM185, which is β2-subunit-specific, in order to coprecipitate αL (Figure 1D). All β2 chimaeras, with the exception of β28, co-precipitated αL, which is consistent with the flow cytometry results. Unexpectedly, β27 showed slower migration as compared with wild-type β2 and the β2 chimaeras. This was due to additional glycosylation, as determined after peptide Nglycosidase F treatment (results not shown). The reason for the β27 aberrant glycosylation profile is not known at present. When

Figure 1

Analyses of heterodimer formation of αL with β2-TM chimaeras

(A) Sequence alignment of integrin TM of the β-subunits (upper panel) and α-subunits (lower panel). The residue number corresponding to β2 or αL is indicated at the top of the panel. (B) Schematic diagram showing the two chimaeric integrins αLβ23 and αIIbβ23 that are the molecules of focus in the present study. cyto. tail, cytoplasmic tail. (C) HEK-293T transfectants expressing the indicated integrins were stained with mAb IB4 (β2-specific and heterodimer-dependent), and analysed by flow cytometry. The lack of surface expression of αLβ28 was further verified with mAb H52 (β2-specific mAb) and MHM24 (αL-specific mAb). The unshaded area represents background staining with an irrelevant mAb. Non-transfected cells were also included as control. (D) Cell lysates of non-transfectants (–ve) and transfectants surface-labelled with biotin were subjected to immunoprecipitation using the β2-specific mAb KIM185. Proteins were resolved on an SDS/7.5 % PAGE under reducing conditions, and detected by ECL® .

the β2-chimaeras were expressed in the presence of αM and αX, the expression profiles of the respective heterodimers were similar to that of αL (results not shown). Effect of β2-TM substitution on β2 ligand binding

To test if the ligand-binding properties of integrins are affected by β2-TM substitutions, transfectants bearing β2 chimaeras with αL, αM or αX were allowed to adhere to relevant immobilized ligands. For adhesion assays of αLβ2, ICAM-1 was used as c The Authors Journal compilation c 2008 Biochemical Society

498


Figure 3 Constitutive activity of αLβ23 as determined by adhesion assays to ICAM-1 and ICAM-3 Figure 2 The effect of β2-TM chimaeras on the adhesion properties of αLβ2, αMβ2 and αXβ2 to immobilized ligands HEK-293T cells were transfected with the indicated α- and β-subunits. ICAM-1 was used for adhesion assays involving αL (A), and BSA was used for adhesion assays involving αM (B) and αX (C). The level of adhesion of αL, αM or αXβ2-TM variants to respective ligands was determined and expressed as relative adhesion to ligand with respect to that of wild-type integrin in the absence of activation. mAb KIM185 (10 µg/ml) was included as the activating agent. The variants β27 and β28 were not included due to aberrant post-translational modification and defective cell-surface expression respectively. w/o, without; w, with.

(A) Cell-surface expression levels of αLβ2 and αLβ23 were assessed by flow-cytometric analyses using IB4 (shaded area). The unshaded area represents background staining with an irrelevant mAb. (B) Transfectants bearing αLβ23 adhered to ICAM-1 constitutively in the absence of activating mAb KIM185 (10 µg/ml). Addition of KIM185 further increased the level of adhesion. (C) Transfectants bearing αLβ23 adhered to ICAM-3 in the presence of KIM185. By contrast, KIM185 and Mg/EGTA (ME; 5 mM MgCl2 and 1.5 mM EGTA) were required for wild-type αLβ2-mediated adhesion to ICAM-3. αLβ2-mediated adhesion specificity was demonstrated using IB4, which is also a function-blocking mAb. Adhesion level is presented as relative ICAM adhesion with respect to that of wild-type αLβ2 in the absence of activation. w/o, without; w, with.

Affinity modulation of chimaeric αLβ23

a ligand. αMβ2 and αXβ2 were reported to bind a wide number of proteins including denatured proteins [33], and BSA has been used for analyses of αMβ2 and αXβ2 adhesion assays [25,30,31]. Thus BSA was used for adhesion assays of αMβ2 and αXβ2 herein. The adhesion profiles that are expressed as fold adhesion relative to that of wild-type in the absence of mAb KIM185, which activates β2-integrin, are summarized (Figure 2). Although in one case, the ICAM-1 binding properties of αLβ24 transfectants were similar to that of the wild-type αLβ2, transfectants bearing αL in association with chimaeras β21, β23, β25 and β26 showed constitutive adhesion to ICAM-1, albeit at lower levels when compared with the adhesion profiles in the presence of activating mAb KIM185. The adhesion profiles of the β2 chimaeras in association with αM and αX were similar to those found with αL. c The Authors Journal compilation c 2008 Biochemical Society

Integrins transit through different affinity states under different conditions. Studies reveal the presence of low-, intermediate- and high-affinity integrins lacking the inserted (I) domain [34,35]. Similarly, the transition of αLβ2 from one affinity state to another and the requirement of these transitions for effective ICAM adhesion have been demonstrated [36]. The αLβ2 is considered to be in a high-affinity state when it adheres to both ICAM-1 and ICAM-3. When it adheres effectively to ICAM-1 but not ICAM-3, it is assigned an intermediate-affinity state [36]. The constitutive ligand-binding activities of the aforementioned β2 chimaeras in association with αL prompted us to determine the activation status of these chimaeric integrins. Because β3-TM has been relatively well studied with respect to β3-integrin affinity modulation and clustering, as compared with other β-subunit TMs [11,37], we focused our subsequent assays on β23. Transfectants expressing αLβ2 and αLβ23 were examined for their capacity to adhere to ICAM-1 and ICAM-3 (Figure 3).


499

Figure 5 Conformational analyses of αLβ23 and its affinity state with respect to a salt-bridge-disrupted mutant αLR1094Dβ2

Figure 4 Constitutively active αLβ23 reverting to wild-type phenotype by the replacement of αL-TM with αIIb-TM (A) The expression of αLβ2, αLβ23 and αLIIbβ23 of transfectants was assessed by flow-cytometric analyses using IB4 (shaded area). The unshaded area represents irrelevant mAb. (B) Adhesion of transfectants to ICAM-1. Whereas transfectants expressing αLβ23 showed constitutive adhesion to ICAM-1, albeit at a lower level as compared with that in the presence of KIM185, the adhesion profile of transfectants expressing αLIIbβ23 was similar to that of wild-type αLβ2. (C) The ICAM-3 adhesion profile of transfectants expressing αLIIbβ23 was similar to that of wild-type αLβ2. Both required KIM185 and ME for effective ICAM-3 adhesion as compared with αLβ23. αLβ2-mediated adhesion specificity was confirmed by using IB4. w/o, without; w, with.

Expression of αLβ2 and αLβ23 was comparable, as determined by mAb IB4 staining (Figure 3A). While the ICAM-1 adhesion profile of αLβ23 showed constitutive activity, this activity was lower than that observed in the presence of the activating mAb KIM185 (Figure 3B). αLβ2-mediated adhesion to ICAM-3 required a combination of two activating conditions, Mg2+ /EGTA and KIM185 (Figure 3C). Interestingly, transfectants bearing αLβ23 did not adhere to ICAM-3 without activation, but adhesion was detected in the presence of KIM185. This suggests that αLβ23 exhibits ligand-binding properties of an intermediateaffinity αLβ2. The precocious activity of αLβ23 is likely to originate from the altered interaction between the αL- and β23TMs. We tested this possibility by using a double chimaeric pair, αLIIb with β23. The TM of the α-subunit is that of αIIb, and the TM of the β-subunit is that of β3, thus forming a natural TM pair αIIbβ3 in the context of an αLβ2-integrin (Figure 4). The expression level of αLIIbβ23 was comparable with that of αLβ2 and αLβ23, as determined by IB4 staining (Figure 4A). Transfectants expressing αLβ23 adhered constitutively to ICAM1, but transfectants expressing αLIIbβ23 showed an ICAM-1 adhesion profile similar to that of wild-type αLβ2 (Figure 4B).

(A) Transfectants bearing the indicated integrins were surface-labelled with biotin followed by immunoprecipitation with mAb KIM127. Proteins were resolved on SDS/7.5 % PAGE under reducing conditions, and detected by the method of ECL® (Amersham). KIM185 was included as a control mAb. (B) ICAM-3 adhesion profiles of transfectants. The composite mutant αLR1094Dβ23 was included to determine whether salt-bridge disruption could further promote the αLβ23 to a high-affinity state with regard to ICAM-3 adhesion. The function-blocking mAb IB4 was used to demonstrate adhesion specificity mediated by αLβ2 in all cases. w/o, without; w, with.

Further, in the presence of a single activating agent, KIM185, transfectants expressing αLβ23 presented significant adhesion to ICAM-3, whereas transfectants bearing wild-type αLβ2 and αLIIbβ23 showed similar adhesion profiles with the need for two activating conditions, Mg2+ /EGTA and KIM185, for effective ICAM-3 adhesion (Figure 4C). In all cases, αLβ2-mediated adhesion specificity was demonstrated using IB4, which is a function-blocking mAb [23]. The conformation of the αLβ23 ectodomain

The αLβ2 extension reporter mAb KIM127 [21] was employed to test whether αLβ23, which shares similar ligand-binding properties of an intermediate affinity αLβ2, has an extended conformation. Transfectants expressing wild-type or chimaeric αLβ2-integrin were surface-labelled with biotin followed by immunoprecipitation with KIM127 (Figure 5A). Included as a control in the analysis was a mutant αLR1094Dβ2. The replacement of Arg1094 with an aspartic acid disrupts salt-bridge formation between αL Arg1094 and β2 Asp709 . Salt-bridge perturbation, which forces the cytoplasmic tails to separate, is known to activate αLβ2 [38]. In the presence of Mg2+ /EGTA, wild-type αLβ2 was precipitated by KIM127. By contrast, αLR1094Dβ2 was precipitated by KIM127 without the requirement of Mg2+ /EGTA, consistent with an extended conformation triggered by salt-bridge disruption. Noteworthy, αLβ23, which showed properties of an intermediate affinity receptor and was expected to display an extended conformation, did not react with KIM127 in the absence of Mg2+ /EGTA. The lack of KIM127 reactivity with αLβ23 was not due to the loss of KIM127 epitope as a result of β23 misfolding, because αLβ23 was precipitated c The Authors Journal compilation c 2008 Biochemical Society

500


by KIM127 when Mg2+ /EGTA was included. The ICAM-3 adhesion profiles of transfectants bearing αLβ23, αLR1094Dβ2 and αLR1094Dβ23 were nonetheless similar, and represent an intermediate-affinity phenotype (Figure 5B). Overall, these results suggest that although αLβ23 exhibits ligand-binding properties of an intermediate-affinity αLβ2, its conformation differs from that of αLβ2 treated with Mg2+ /EGTA or αLR1094Dβ2. The cytoplasmic tails of αLβ23 are not separated

Next, we performed FRET analyses to determine whether the cytoplasmic tails of αLβ23 are separated. mCFP and mYFP were fused to the C-terminus of the α- and β-subunits respectively of αLβ2, αLβ23 and αLR1094Dβ2 using a similar strategy to that adopted by others [28]. Separation of the α- and βcytoplasmic tails will lead to a poor or minimal FRET (Figure 6A). K562 cells were transfected with these expression constructs, and FRET analyses by the method of acceptor photobleaching were conducted [32]. Significant FRET was detected in wild-type αLβ2, whereas FRET efficiency was reduced in αLR1094Dβ2 as a result of forced separation of the cytoplasmic tails (Figure 6B). Interestingly, FRET efficiency in αLβ23 was comparable with that in wild-type αLβ2. Representative images of FRET for each FRET pair construct are shown (Figure 6C). DISCUSSION

A large amount of information concerning integrin regulation has been derived from structural and functional studies of integrin ectodomains and cytoplasmic tails [3]. However, it is the TM helices that connect these two domains, transferring the information from the separating cytoplasmic tails, after engagement of cytosolic interactors, to the integrin ectodomains. Integrin TMs have also been reported to play an important role in α/β-integrin dimerization [11,39,40], although it is not clear at present if TM specificity is required, and to what extent, for normal integrin function. Our results demonstrate that, in some cases, specificity of α/β-TM interaction is required for αLβ2-integrin biosynthesis or for the maintenance of its functional integrity. Replacement of β2- with β7-TM generated an αLβ2 chimaera that showed β2 aberrant glycosylation. When replaced with β8-TM, cellsurface receptor expression was abrogated. Further, replacement of β2- with β3-, β5- or β6-TM had no apparent effect on receptor expression, but showed altered ligand-binding properties. Indeed, further examination of the chimaera αLβ23 revealed a constitutively active receptor in an intermediate-affinity state, based on the adhesion properties to ICAMs. Although it is possible that this effect is due to intra-subunit conformational changes as a consequence of β2-TM exchange with β3-TM, as reported for other β2 mutational analyses, replacement of αL-TM with αIIbTM reverted the chimaera to a wild-type phenotype. This suggests that the interaction between the α- and β-TMs of integrins has considerable impact on its function and is consistent with reports where a point mutation at either TM generated an activated αIIbβ3 via disruption of the α/β-TM interface [37,41]. The fact that integrin is known to unbend during activation prompted us to examine the conformation of the constitutively active chimaera αLβ23. The lack of KIM127 reactivity on αLβ23 suggests that this chimaeric receptor is not fully extended. It is possible that the unfavourable association between αL- and β3TMs in αLβ23 leads to the splaying of the two subunits, which would still retain an overall bent conformation. Indeed, some reports suggest that bent αVβ3 and α4β1 can bind ligands [42,43]. There are also examples of engineered αLβ2 variants that lack c The Authors Journal compilation c 2008 Biochemical Society

Figure 6

Analyses of cytoplasmic tails separation in αLβ2-TM chimaera

(A) A diagram illustrating integrin α- and β-subunits with mCFP and mYFP fused to the C-terminus of their cytoplasmic tails respectively. The separation of the cytoplasmic tails when the integrin is activated will lead to a decrease in FRET efficiency. (B) Dot-plot of FRET efficiency measurements of wild-type αLmCFPβ2mYFP and mutants. Each dot represents one cell analysed. Horizontal bar represents the mean value of 80 cells analysed. (C) Representative images of K562 cells transfected with wild-type αLmCFPβ2mYFP and mutants. The entire cell was subjected to bleaching, but the cell membrane was chosen as the ROI for FRET measurements. The mCFP and mYFP signals pre- and post-bleach are presented as colour range from blue to red for low to high intensity respectively. Images were processed using the software LSM510, version 3.2. The fluorescence signal detected in the cell derives from proteins found in the Golgi that was not included in the ROI.


reactivity with KIM127 despite their inherent propensity to bind ICAM. An αLβ2 variant with an open conformation I domain that bound avidly to ICAM-1 had poor reactivity with KIM127 [29,44]. Recently, we showed that an αLβ2 variant locked in a bent conformation via an engineered disulfide failed to react with KIM127 even in the presence of Mg2+ /EGTA despite its constitutive ICAM-1-binding property [45]. These studies suggest that a non-fully extended αLβ2 could bind ligand. Based on recent electron microscopy results, integrins may adopt different ‘bent’ conformations [46]. These conformations could be transient, as they may represent snapshots of wild-type integrin conformations during activation. We reasoned that the replacement of β2-TM with β3-TM generates an αLβ23 chimaera that is entrapped in one of these bent conformations that allows ICAM-1 binding but has poor reactivity with KIM127 because it is not fully extended. The FRET results also suggest that the cytoplasmic tails of αLβ23 are not significantly separated as a result of the mismatch αLand β3-TMs. We conjectured that the TMs of αLβ23 are not separated as this would also lead to cytoplasmic tail separation when the connectivity between the TM and the cytoplasmic tail is maintained. Instead, it is tempting to speculate that the nonpermissive pair of αL/β3-TMs are juxtaposed in an orientation that does not culminate in their complete separation. As the issue of integrin affinity against valency regulation during ligand binding has sparked much debate [47,48], it may be argued that enhanced ICAM binding shown by αLβ23 is due to receptor clustering, rather than to a conformational change. Although we cannot rule out this possibility at present, others have shown [41] that perturbation of the αIIbβ3-TM interface by leucine scan point mutations had a significant effect on its conformation and ligand-binding affinity but not on receptor clustering. Thus there remains much to be learned from studying TM–TM interactions of integrins. This study highlights the importance of permissive TM–TM interactions of an integrin, as most non-cognate TM pairs investigated herein generated integrins with anomalies in function. Therefore it is apparent that the expression of a functionally regulated integrin requires parallel evolution of its α- and β-TMs, failure of which may lead to deleterious adhesive events in vivo, as elegantly demonstrated in transgenic mice, expressing a constitutively active αLβ2, with defective immune responses [49]. This work was supported by the Singapore Agency for Science, Technology, and Research (A*STAR) BMRC (Biomedical Research Council) grant 04/1/22/19/358. A. V. is supported by A*STAR BMRC grant 03/1/22/19/238. We thank M. Cooray for her technical assistance.

REFERENCES 1 Hynes, R. O. (2002) Integrins: bidirectional, allosteric signaling machines. Cell 110, 673–687 2 Arnaout, M. A., Mahalingam, B. and Xiong, J. P. (2005) Integrin structure, allostery, and bidirectional signaling. Annu. Rev. Cell Dev. Biol. 21, 381–410 3 Luo, B. H., Carman, C. V. and Springer, T. A. (2007) Structural basis of integrin regulation and signaling. Annu. Rev. Immunol. 25, 619–647 4 Tadokoro, S., Shattil, S. J., Eto, K., Tai, V., Liddington, R. C., de Pereda, J. M., Ginsberg, M. H. and Calderwood, D. A. (2003) Talin binding to integrin β tails: a final common step in integrin activation. Science 302, 103–106 5 Fagerholm, S. C., Hilden, T. J., Nurmi, S. M. and Gahmberg, C. G. (2005) Specific integrin α and β chain phosphorylations regulate LFA-1 activation through affinity-dependent and -independent mechanisms. J. Cell Biol. 171, 705–715 6 Nurmi, S. M., Autero, M., Raunio, A. K., Gahmberg, C. G. and Fagerholm, S. C. (2007) Phosphorylation of the LFA-1 integrin β2 chain on Thr-758 leads to adhesion, Rac-1/Cdc42 activation, and stimulation of CD69 expression in human T cells. J. Biol. Chem. 282, 968–975 7 Luo, B. H. and Springer, T. A. (2006) Integrin structures and conformational signaling. Curr. Opin. Cell Biol. 18, 579–586

501

8 Hyland, R. H., Douglass, W. A., Tan, S. M. and Law, S. K. (2001) Chimeras of the integrin β subunit mid-region reveal regions required for heterodimer formation and for activation. Cell Commun. Adhes. 8, 61–69 9 Tng, E., Tan, S. M., Ranganathan, S., Cheng, M. and Law, S. K. (2004) The integrin αLβ2 hybrid domain serves as a link for the propagation of activation signal from its stalk regions to the I-like domain. J. Biol. Chem. 279, 54334–54339 10 Li, W., Metcalf, D. G., Gorelik, R., Li, R., Mitra, N., Nanda, V., Law, P. B., Lear, J. D., Degrado, W. F. and Bennett, J. S. (2005) A push–pull mechanism for regulating integrin function. Proc. Natl. Acad. Sci. U.S.A. 102, 1424–1429 11 Li, R., Mitra, N., Gratkowski, H., Vilaire, G., Litvinov, R., Nagasami, C., Weisel, J. W., Lear, J. D., DeGrado, W. F. and Bennett, J. S. (2003) Activation of integrin αIIbβ3 by modulation of transmembrane helix associations. Science 300, 795–798 12 Luo, B. H., Springer, T. A. and Takagi, J. (2004) A specific interface between integrin transmembrane helices and affinity for ligand. PLoS Biol. 2, e153 13 Schneider, D. and Engelman, D. M. (2003) GALLEX, a measurement of heterologous association of transmembrane helices in a biological membrane. J. Biol. Chem. 278, 3105–3111 14 MacKenzie, K. R., Prestegard, J. H. and Engelman, D. M. (1997) A transmembrane helix dimer: structure and implications. Science 276, 131–133 15 Adair, B. D. and Yeager, M. (2002) Three-dimensional model of the human platelet integrin αIIbβ3 based on electron cryomicroscopy and x-ray crystallography. Proc. Natl. Acad. Sci. U.S.A. 99, 14059–14064 16 Gottschalk, K. E. (2005) A coiled-coil structure of the αIIbβ3 integrin transmembrane and cytoplasmic domains in its resting state. Structure 13, 703–712 17 Gottschalk, K. E., Adams, P. D., Brunger, A. T. and Kessler, H. (2002) Transmembrane signal transduction of the αIIbβ3 integrin. Protein Sci. 11, 1800–1812 18 Lin, X., Tan, S. M., Law, S. K. and Torres, J. (2006) Unambiguous prediction of human integrin transmembrane heterodimer interactions using only homologous sequences. Proteins 65, 274–279 19 Robinson, M. K., Andrew, D., Rosen, H., Brown, D., Ortlepp, S., Stephens, P. and Butcher, E. C. (1992) Antibody against the Leu-CAM β chain (CD18) promotes both LFA-1- and CR3-dependent adhesion events. J. Immunol. 148, 1080–1085 20 Stephens, P., Romer, J. T., Spitali, M., Shock, A., Ortlepp, S., Figdor, C. G. and Robinson, M. K. (1995) KIM127, an antibody that promotes adhesion, maps to a region of CD18 that includes cysteine-rich repeats. Cell Commun. Adhes. 3, 375–384 21 Beglova, N., Blacklow, S. C., Takagi, J. and Springer, T. A. (2002) Cysteine-rich module structure reveals a fulcrum for integrin rearrangement upon activation. Nat. Struct. Biol. 9, 282–287 22 Hildreth, J. E., Gotch, F. M., Hildreth, P. D. and McMichael, A. J. (1983) A human lymphocyte-associated antigen involved in cell-mediated lympholysis. Eur. J. Immunol. 13, 202–208 23 Wright, S. D., Rao, P. E., Van Voorhis, W. C., Craigmyle, L. S., Iida, K., Talle, M. A., Westberg, E. F., Goldstein, G. and Silverstein, S. C. (1983) Identification of the C3bi receptor of human monocytes and macrophages by using monoclonal antibodies. Proc. Natl. Acad. Sci. U.S.A. 80, 5699–5703 24 Al-Shamkhani, A. and Law, S. K. (1998) Expression of the H52 epitope on the β2 subunit is dependent on its interaction with the α subunits of the leukocyte integrins LFA-1, Mac-1 and p150,95 and the presence of Ca2+ . Eur. J. Immunol. 28, 3291–3300 25 Tan, S. M., Hyland, R. H., Al-Shamkhani, A., Douglass, W. A., Shaw, J. M. and Law, S. K. (2000) Effect of integrin β2 subunit truncations on LFA-1 (CD11a/CD18) and Mac-1 (CD11b/CD18) assembly, surface expression, and function. J. Immunol. 165, 2574–2581 26 Barclay, A. N., Lawrson, B. M. H., Law, S. K., McKnight, A. J., Tomlinson, M. G. and van der Merwe, P. A. (1997) The Leucocyte Antigen Facts Book, Academic Press, London 27 Zacharias, D. A., Violin, J. D., Newton, A. C. and Tsien, R. Y. (2002) Partitioning of lipid-modified monomeric GFPs into membrane microdomains of live cells. Science 296, 913–916 28 Kim, M., Carman, C. V. and Springer, T. A. (2003) Bidirectional transmembrane signaling by cytoplasmic domain separation in integrins. Science 301, 1720–1725 29 Cheng, M., Foo, S. Y., Shi, M. L., Tang, R. H., Kong, L. S., Law, S. K. and Tan, S. M. (2007) Mutation of a conserved asparagine in the I-like domain promotes constitutively active integrins αLβ2 and αIIbβ3. J. Biol. Chem. 282, 18225–18232 30 Tang, M. L., Kong, L. S., Law, S. K. and Tan, S. M. (2006) Down-regulation of integrin αMβ2 ligand-binding function by the urokinase-type plasminogen activator receptor. Biochem. Biophys. Res. Commun. 348, 1184–1193 31 Shaw, J. M., Al-Shamkhani, A., Boxer, L. A., Buckley, C. D., Dodds, A. W., Klein, N., Nolan, S. M., Roberts, I., Roos, D., Scarth, S. L. et al. (2001) Characterization of four CD18 mutants in leucocyte adhesion deficient (LAD) patients with differential capacities to support expression and function of the CD11/CD18 integrins LFA-1, Mac-1 and p150,95. Clin. Exp. Immunol. 126, 311–318 c The Authors Journal compilation c 2008 Biochemical Society

502


32 Karpova, T. S., Baumann, C. T., He, L., Wu, X., Grammer, A., Lipsky, P., Hager, G. L. and McNally, J. G. (2003) Fluorescence resonance energy transfer from cyan to yellow fluorescent protein detected by acceptor photobleaching using confocal microscopy and a single laser. J. Microsc. 209, 56–70 33 Davis, G. E. (1992) The Mac-1 and p150,95 β2 integrins bind denatured proteins to mediate leukocyte cell–substrate adhesion. Exp. Cell Res. 200, 242–252 34 Takagi, J., Petre, B. M., Walz, T. and Springer, T. A. (2002) Global conformational rearrangements in integrin extracellular domains in outside-in and inside-out signaling. Cell 110, 599–611 35 Xiao, T., Takagi, J., Coller, B. S., Wang, J. H. and Springer, T. A. (2004) Structural basis for allostery in integrins and binding to fibrinogen-mimetic therapeutics. Nature 432, 59–67 36 Tang, R. H., Tng, E., Law, S. K. and Tan, S. M. (2005) Epitope mapping of monoclonal antibody to integrin αLβ2 hybrid domain suggests different requirement of affinity states for intercellular adhesion molecules (ICAM)-1 and ICAM-3 binding. J. Biol. Chem. 280, 29208–29216 37 Litvinov, R. I., Vilaire, G., Li, W., DeGrado, W. F., Weisel, J. W. and Bennett, J. S. (2006) Activation of individual αIIbβ3 integrin molecules by disruption of transmembrane domain interactions in the absence of clustering. Biochemistry 45, 4957–4964 38 Li, Y. F., Tang, R. H., Puan, K. J., Law, S. K. and Tan, S. M. (2007) The cytosolic protein talin induces an intermediate affinity integrin αLβ2. J. Biol. Chem. 282, 24310–24319 39 Li, R., Babu, C. R., Lear, J. D., Wand, A. J., Bennett, J. S. and DeGrado, W. F. (2001) Oligomerization of the integrin αIIbβ3: roles of the transmembrane and cytoplasmic domains. Proc. Natl. Acad. Sci. U.S.A. 98, 12462–12467 40 Lin, X., Tan, S. M., Law, S. K. and Torres, J. (2006) Two types of transmembrane homomeric interactions in the integrin receptor family are evolutionarily conserved. Proteins 63, 16–23 Received 5 September 2007/13 November 2007; accepted 21 November 2007 Published as BJ Immediate Publication 21 November 2007, doi:10.1042/BJ20071218

c The Authors Journal compilation c 2008 Biochemical Society

41 Luo, B. H., Carman, C. V., Takagi, J. and Springer, T. A. (2005) Disrupting integrin transmembrane domain heterodimerization increases ligand binding affinity, not valency or clustering. Proc. Natl. Acad. Sci. U.S.A. 102, 3679–3684 42 Adair, B. D., Xiong, J. P., Maddock, C., Goodman, S. L., Arnaout, M. A. and Yeager, M. (2005) Three-dimensional EM structure of the ectodomain of integrin αVβ3 in a complex with fibronectin. J. Cell Biol. 168, 1109–1118 43 Chigaev, A., Buranda, T., Dwyer, D. C., Prossnitz, E. R. and Sklar, L. A. (2003) FRET detection of cellular α4 integrin conformational activation. Biophys. J. 85, 3951–3962 44 Lu, C., Shimaoka, M., Zang, Q., Takagi, J. and Springer, T. A. (2001) Locking in alternate conformations of the integrin αLβ2 I domain with disulfide bonds reveals functional relationships among integrin domains. Proc. Natl. Acad. Sci. U.S.A. 98, 2393–2398 45 Shi, M., Foo, S. Y., Tan, S. M., Mitchell, E. P., Law, S. K. and Lescar, J. (2007) A structural hypothesis for the transition between bent and extended conformations of the leukocyte β2 integrins. J. Biol. Chem. 282, 30198–30206 46 Nishida, N., Xie, C., Shimaoka, M., Cheng, Y., Walz, T. and Springer, T. A. (2006) Activation of leukocyte β2 integrins by conversion from bent to extended conformations. Immunity 25, 583–594 47 Bazzoni, G. and Hemler, M. E. (1998) Are changes in integrin affinity and conformation overemphasized? Trends Biochem. Sci. 23, 30–34 48 Carman, C. V. and Springer, T. A. (2003) Integrin avidity regulation: are changes in affinity and conformation underemphasized? Curr. Opin. Cell Biol. 15, 547–556 49 Semmrich, M., Smith, A., Feterowski, C., Beer, S., Engelhardt, B., Busch, D. H., Bartsch, B., Laschinger, M., Hogg, N., Pfeffer, K. and Holzmann, B. (2005) Importance of integrin LFA-1 deactivation for the generation of immune responses. J. Exp. Med. 201, 1987–1998