situations (Humphries, 2000; Hynes, 2002).Adhesion that is mediated by integrins iscontrolled dynamically to allow cell positioningand migration and to prevent abnormaltrafficking and anchorage. Integrin signalling inresponse to ligand binding is achieved by acombination of receptor clustering andconformational changes. Both of theseprocesses can be elicited from the inside of thecell (resulting in the acquisition of a high-affinity receptor conformation and priming ofthe receptor for ligand binding) and from theoutside (to mediate ligand-induced activationand signalling).

The study of changes in integrin shape hasbeen greatly aided by the availability ofmonoclonal antibodies (mAbs) that detectconformation-dependent epitopes. ThesemAbs have not only helped to pinpointthe intramolecular changes that determine theintegrin activation state, but have also provenuseful for regulating function. In recent years,

major advances have been made in ourunderstanding of the mechanisms that regulateintegrin affinity (Arnaout et al., 2005; Luoet al., 2007) and, accordingly, we now also havean improved knowledge of the mechanisms ofaction of function-regulating mAbs. As thesemAbs work in different ways, there is a dangerthat researchers might select the wrong reagentfor their studies and/or misinterpret the datathat they obtain. In this Cell Science at a Glancearticle, we have therefore attempted to explainbriefly the mechanisms of antibody regulationof integrins. The accompanying poster liststhree classes of key reagents: those that inhibitligand engagement; those that stimulate ligandengagement or report a high-affinity integrinstate (‘activation specific’); and those thatserve as equally important negative controls.Partly owing to space constraints and partlyowing to a lack of available information, wehave restricted our selection of mAbs to thosethat recognise human integrins. Furthermore,

4009Cell Science at a Glance

(See poster insert)

Anti-integrin monoclonalantibodiesAdam Byron, Jonathan D.Humphries, Janet A. Askari, SueE. Craig, A. Paul Mould and MartinJ. Humphries*Wellcome Trust Centre for Cell-Matrix Research,Michael Smith Building, Faculty of Life Sciences,University of Manchester, Manchester M13 9PT, UK*Author for correspondence(martin.humphries@manchester.ac.uk)

Journal of Cell Science 122, 4009-4011Published by The Company of Biologists 2009doi:10.1242/jcs.056770

IntroductionIntegrins are a family of 24 heterodimerictransmembrane receptors that support cell-celland cell-ECM (extracellular matrix) interactionsin a multitude of physiological and disease

© Journal of Cell Science 2009 (122, pp. 4009-4011)

Anti-integrin Monoclonal AntibodiesAdam Byron, Jonathan D. Humphries, Janet A. Askari, Sue E. Craig, A. Paul Mould and Martin J. Humphries

Abbreviations: Cyto, cytoplasmic; EGF, epidermal growth factor; mAb, monoclonal antibody; ND, not determined; PSI, plexin-semaphorin-integrin homology; TM, transmembrane.

Low-affinity integrin High-affinity integrin

ααA-domain-containingintegrin (alternative form)

Inhibitory Stimulatory oractivation-specific

mAb 15

D57 23C6

15F11 E7P66.4B4CSβ6

PL98DF6 6F8LM142

Epitope Antibody Epitope

FB12

12F1Gi9

JA218

P1E6

ASC-6P1B5

IA3ND

HP2/1

P4C2

PS/2

JBS5

mAb 16

P1D6

GoH3

6A11

Y9A2

ND

ND

ND

17E6L230

αE7-1

αE7-2

Ber-ACT8

TS1/22mAb 38

ND

ND

2LPM19c

CBRM1/5

M1/70ND

ND

ND

ND

ND

ND

ND

ND

ND

3.9496K

BU15

217I

240I †Ligand-mimetic antibody

J1B5 TS2/7 16B431H4

17C6J143

mAb 11VC5

3C12 7A5 A9A1 396214 44H68F2

LF61 TS2/4YTH81.5

CBRM1/20OKM1

CBR-p150/2E1

Integrin Epitope

ND

ND

212D

β3

αIIb αV

α2

α4

α5

αIIb

β-subunit

α-subunit

Cytodomain

TMdomain

β-taildomain

αL

αX

β1

β2

β3

β7

α6 α1 α2 α3 α5 α7 α8 α9 α10 α11 α4 αE αL αM αX αD

β5 β6 β8 β1 β7 β2β4

11D1 5C4 14E5 3E1 K20

Cytodomain

Calf-2domain

Calf-1domain

Thighdomain

β-propeller TMdomain

EGF-likerepeats 1-4

GenuPSIdomain

Hybriddomain

βAdomain

αAdomain

Integrin

αD

αX

αM

αL

αE

αV

α9

α7

α6

α5

α4

α3

α2

α1

Integrin

β1

β2

β3

β4

β5

β6

β7

β8

αIIbβ3

αVβ3

αVβ5

αVβ6

Antibody

4B4

mAb 13

P4C10

JB1A

CLB LFA-1/1

TS1/18

7E4

MHM23

7E3

SZ-21

ASC-8

ALULA

6.3G9

FIB504FIB27

37E1

PAC-1†

10E52G12

WOW-1†

LM609

6.8G6†

10D5

Antibody

JBS2

HP1/3

SNAKA51

PT25-2

PMI-1

MEM-83

NKI-L16

496B

12G108A2TS2/16

15/7HUTS-4

8E3N29

9EG7

mAb 24

MEM-148

KIM127

CBR LFA-1/2

MEM-48

KIM185

AP3

AP5

LIBS6

LIBS2

10F82B82G3

Genu

Integrin heterodimers

Non-functional

CBR LFA-1/7

Key

Stimulatory

Inhibitory

P1F6P3G2

αLβ2 YTA-1

Jour

nal o

f Cel

l Sci

ence

4010

we only report mAbs that either affect ligandbinding and cell adhesion or, to our knowledge,are non-functional controls. Although wehave not catalogued mAbs for use inimmunoprecipi tation, immunoblotting, flowcytometry or ELISA-type experiments, manyof the reagents listed are suitable for thesepurposes. The poster is not intended torecommend one mAb in favour of another, butto represent some of the best-characterisedexamples. The researcher must thereforedetermine the most appropriate reagent fortheir specific purpose.

Integrin structure and conformationalchangesThe first crystal structure of an integrin (V3)was solved in 2001 (Xiong et al., 2001). Thestructure revealed a ‘head’ region, which wasthe main point of contact between the twosubunits and comprised a -propeller fold in the-subunit and a von Willebrand factor-Adomain in the -subunit (the A domain). Thehead was supported by two rod-like ‘legs’.The -subunit leg comprised three -sandwichdomains (termed thigh, calf-1 and calf-2) andthe -subunit leg included a PSI (plexin-semaphorin-integrin) domain, animmunoglobulin fold (termed the hybriddomain), four epidermal growth factor (EGF)-like repeats and a cystatin-like fold (termed the-tail domain). A soluble form of the integrinwas used for the crystallisation studies, but it isnow well established that both integrin legs linkto transmembrane domains and then to shortcytoplasmic domains that can interact with eachother or with cytoskeletal and signallingproteins (Wegener et al., 2007).

Interestingly, the initial crystal structurerevealed a bent molecule, with articulationpoints in both integrin legs at the so-called genu.This form of the integrin is now thought torepresent the conformation with low affinity fora ligand. The adoption of a high-affinityconformation involves a series of shapechanges, including unbending of the receptorand various inter-module and intra-modulemovements, such as swing-out of the hybriddomain away from the -subunit and -helicalmovements in the A domain. Most evidencepoints to a separation of the cytoplasmic andtransmembrane domains as a key step in theacquisition of the high-affinity conformation. Itis currently unclear how many classes ofintegrin conformation exist, but primed andligand-bound integrins have similarconformations, and these are distinctly differentfrom low-affinity receptors. The postertherefore contains two general representationsof integrins (bent and extended). As yet, thereare no mAbs that are able to distinguish primed

from ligand-bound integrins, although therelative expression of different epitopes variesbetween these states.

Using the prototypic peptide ligand arginine-glycine-aspartate (RGD) in co-crystallisationstudies, the ligand-binding pocket inV3 integrin was located at the junction of the-propeller and the A domain in the head(Xiong et al., 2002). The aspartate carboxylgroup of RGD was found to coordinate adivalent cation directly in a so-called metal-ion-dependent adhesion site or MIDAS. This is nowaccepted as a common mode of ligand bindingby integrins. Half of all integrin -subunitscontain an A-domain inserted into the -subunit-propeller, and for those dimers that containthis A domain, the module has evolved toincorporate the main ligand-binding site througha very similar mechanism to that of theA domain. Intriguingly, ligand binding tothe A domain causes engagement of theA domain by an intramolecular pseudo-ligand(a glutamate residue at the base of theA domain) and therefore relays ligand bindingto the -subunit (Alonso et al., 2002). Thus, theconformational changes that underpin changesin integrin activation state are common to allreceptors.

Mechanism of action of regulatorymAbsStimulatory or activation-specific mAbsWhereas all stimulatory or activation-specificmAbs appear to increase ligand-binding affinityby reducing off-rate (i.e. the rate ofdissociation), the reagents fall into twosubclasses (Humphries, 2000). One subclassrecognises epitopes that are regulated by ligandand cation binding. The epitopes for these mAbsare frequently termed LIBS (ligand-inducedbinding sites). A second subclass is not affectedby ligand or cation binding (this includes thewidely used anti-1 mAb, TS2/16). It isprobable that cation- and ligand-regulated mAbsrecognise conformations of integrins that arefound in either the primed or ligand-occupiedstate, and that they therefore shift aconformational equilibrium in favour of thoseforms. The epitopes might be created de novo asa result of movement of secondary structureelements within protein modules (e.g. within theA domain) or by exposure of masked epitopes(e.g. as a result of leg separation, unbending ordomain movement). Inevitably, any mAb thatbinds preferentially to the high-affinity form ofthe receptor will activate integrin by skewing theconformational equilibrium. However, somemight appear to lack stimulatory activity eitherbecause the equilibrium might already befully displaced towards the high-affinity stateor because the cell type tends to exhibit

all-or-none-type activation responses (e.g.mAb 24 and leukocyte 2 integrins).

A special subclass of cation- and ligand-regulated mAbs are those that containligand-mimetic peptide sequences within theircomplementarity-determining loops (e.g.PAC-1 and WOW-1, which recognise 3; and6.8G6, which binds 6). These reagents areparticularly useful for detecting high-affinityconformations of integrins because theyeffectively act as ligands. However, these mAbsalso block binding because they competedirectly for ligand occupancy. The binding ofnon-ligand-regulated mAbs probably induces aprimed conformation in the integrin by forcingshape changes rather than by stabilisingnaturally occurring conformations. Moststimulatory mAbs recognise sites throughout the-subunit, implying that major changes takeplace in this region of the receptor. Key sites thatare recognised by ligand-regulated mAbsinclude the PSI domain (which is partly buriedin the bent form), the A domain (in whichmovement of its 1 and 7 helices create newepitopes), the hybrid domain (in which swing-out exposes epitopes), the genu (which ismasked in the bent form) and the EGF-likerepeats (which are partly covered in the bentform). Although most activating mAbsrecognise the -subunit, activating anti--subunit mAbs do exist, with examples thataffect conformation of the A domain or detectleg separation. Thus, the choice of stimulatorymAbs for research purposes should be informedby the mechanism of action of the mAb, and thisis generally determined by the location of itsepitopes. The acquisition of a high-affinity stateis best achieved using mAbs that stabilise anextended conformation or that stabilise the high-affinity form of the A domain.

Inhibitory mAbsIt is generally assumed that inhibitory mAbssterically interfere with ligand binding andtherefore act as competitive inhibitors.Surprisingly, of those anti-integrin mAbs onwhich detailed analyses have been performed,most act as allosteric inhibitors (Mould et al.,1996). In this case, mAbs appear to prevent theconformational changes that are needed forligand binding to occur, and they mighttherefore stabilise the unoccupied state of thereceptor. Although many inhibitory mAbs areallosteric inhibitors, their epitopes are usuallyvery close to ligand contact sites. Similarly tostimulatory mAbs, some of these epitopes areregulated by ligand and cation binding and aretermed LABS (ligand-attenuated binding sites).There are also a few examples of mAbs thatrecognise the high-affinity conformation ofintegrins, but their binding blocks ligand

Journal of Cell Science 122 (22)

Jour

nal o

f Cel

l Sci

ence

4011

engagement (e.g. 12F1, which recognises 2;and CBRM1/5, which binds M).

Non-functional mAbsThe use of the most appropriate non-functionalintegrin mAb is of paramount importance forthe correct interpretation of data relating to thestudy of integrin regulation and function. Thesereagents therefore constitute the best type ofcontrol and, where available, a mAb directedagainst the same subunit and domain of theintegrin of interest should be used. There arerelatively few non-functional mAbs reported,and the choice of reagents can therefore belimited. The reason for this paucity is probablythe conformationally dynamic nature ofintegrins, with stabilisation of manyconformations, even distal from the main sitesof ligand binding, having a resultant effect onintegrin function.

TherapeuticsIntegrin-based cell adhesion contributes to thepathogenesis of a wide spectrum of humandisorders. mAbs that block integrin functionshave been shown to be of clinical benefit for thetreatment of thrombotic and autoimmunedisorders, with total sales of anti-integrinantibodies for therapeutic use exceeding$1 billion in 2008 (La Merie, 2009). Anti-integrin mAbs that are currently in clinical useare ReoPro (abciximab) and Tysabri(natalizumab). ReoPro is a humanised versionof a function-blocking mAb (7E3) thatrecognises the 3 subunit and blocks the plateletfibrinogen receptor IIb3; it is used in thetreatment of unstable angina and as an adjuvantto percutaneous coronary interventions. Theepitope of 7E3 includes residues on the top faceof the 3 subunit A-domain, close to the MIDASsite (Artoni et al., 2004). Tysabri is an inhibitorymAb that recognises the 4 subunit and is usedin the treatment of relapsing forms of multiple

sclerosis. Blockade of T-cell 41 integrinprevents these cells from entering the centralnervous system, thereby slowing the destructionof nerve sheaths. The epitope of Tysabri lies inthe -propeller domain of the 4 subunit (Hurynet al., 2004). A number of other function-blocking antibodies are currently at preclinicalstages of development for the treatment ofcancer and fibrosis. In the future, it is likely thatantibodies that stimulate integrin function willalso find clinical uses, such as enhancing repairin the central nervous system (Andrews et al.,2009).

PerspectivesIntegrin mAbs have contributed immensely toour understanding of integrin activation-stateregulation, integrin biology and integrin-basedtherapeutics. Nonetheless, much is still to belearned, for example regarding whatdifferentiates conformation-specific signals.The development of new mAbs that candiscriminate between these signals wouldundoubtedly aid progress in this area and mightalso reveal new avenues for therapeuticintervention. It is noteworthy that mAbs thatrecognise certain integrin subunits orheterodimer activation states (e.g. ligand-mimetic 1 mAbs or activation-specific-subunit mAbs) are at present missing orunder-represented in the arsenal of availablemAbs, and their production would shed furtherlight on many areas of integrin biology. Movingforward, it is apparent that global systems-basedanalyses of the complexity of integrin signallingis both required and informative, and the use ofintegrin mAbs in such studies might help toelucidate activation-state-specific signallingcomplexes and pathways.

Work from the authors’ laboratory that contributed tothis article was funded by grants from the WellcomeTrust (045225 and 074941). We are grateful to thefollowing colleagues for suggesting mAbs: Michael

Brenner, Filippo Giancotti, Mark Ginsberg, SimonGoodman, Donald Gullberg, Jonathan Higgins, NancyHogg, Richard Hynes, John Marshall, Mark Morgan,Lou Reichardt, Dean Sheppard, Arnoud Sonnenberg,Tim Springer, Mark Travis, Klaus von der Mark andKen Yamada.

ReferencesAlonso, J. L., Essafi, M., Xiong, J.-P., Stehle, T. andArnaout, M. A. (2002). Does the integrin A domainact as a ligand for its A domain? Curr. Biol. 12, R340-R342.Andrews, M. R., Czvitkovich, S., Dassie, E., Vogelaar, C.F., Faissner, A., Blits, B., Gage, F. H., ffrench-Constant,C. and Fawcett, J. W. (2009). 9 integrin promotes neuriteoutgrowth on tenascin-C and enhances sensory axonregeneration. J. Neurosci. 29, 5546-5557.Arnaout, M. A., Mahalingam, B. and Xiong, J.-P. (2005).Integrin structure, allostery, and bidirectional signaling.Annu. Rev. Cell Dev. Biol. 21, 381-410.Artoni, A., Li, J., Mitchell, B., Ruan, J., Takagi, J.,Springer, T. A., French, D. L. and Coller, B. S. (2004).Integrin 3 regions controlling binding of murine mAb 7E3:Implications for the mechanism of integrin IIb3activation. Proc. Natl. Acad. Sci. USA 101, 13114-13120.Humphries, M. J. (2000). Integrin structure. Biochem. Soc.Trans. 28, 311-339.Huryn, D. M., Konradi, A. W., Ashwell, S., Freedman,S. B., Lombardo, L. J., Pleiss, M. A., Thorsett, E. D.,Yednock, T. and Kennedy, J. D. (2004). The identificationand optimization of orally efficacious, small molecule VLA-4 antagonists. Curr. Top. Med. Chem. 4, 1472-1484.Hynes, R. O. (2002). Integrins: bidirectional, allostericsignaling machines. Cell 110, 673-687.La Merie, S. L. (2009). Top 20 biologics 2008. R&DPipeline News Special Edition 1, 1-25.Luo, B. H., Carman, C. V. and Springer, T. A. (2007).Structural basis of integrin regulation and signaling. Annu.Rev. Immunol. 25, 619-647.Mould, A. P., Akiyama, S. K. and Humphries, M. J.(1996). The inhibitory anti-1 integrin monoclonal antibody13 recognizes an epitope that is attenuated by ligandoccupancy: evidence for allosteric inhibition of integrinfunction. J. Biol. Chem. 271, 20365-20374.Wegener, K. L., Partridge, A. W., Han, J., Pickford, A.R., Liddington, R. C., Ginsberg, M. H. and Campbell, I.D. (2007). Structural basis of integrin activation by talin.Cell 128, 171-182.Xiong, J.-P., Stehle, T., Diefenbach, B., Zhang, R.,Dunker, R., Scott, D. L., Joachimiak, A., Goodman, S. L.and Arnaout, M. A. (2001). Crystal structure of theextracellular segment of integrin V3. Science 294, 339-345.Xiong, J.-P., Stehle, T., Zhang, R., Joachimiak, A., Frech,M., Goodman, S. L. and Arnaout, M. A. (2002). Crystalstructure of the extracellular segment of integrin V3 incomplex with an Arg-Gly-Asp ligand. Science 296, 151-155.

Journal of Cell Science 122 (22)

Jour

nal o

f Cel

l Sci

ence