Upregulated expression and function of integrin adhesive receptors in systemic lupus erythematosus patients with vasculitis

Upregulated expression and function of integrin

adhesive receptors in systemic lupus

erythematosus patients with vasculitis.

T Takeuchi, … , J Koide, T Abe

J Clin Invest.

1993;

92(6)

:3008-3016.

https://doi.org/10.1172/JCI116924

.

Upregulation of integrin adhesive receptors has been implicated in various pathological

conditions. We examined expression and function of integrin adhesive receptors on

peripheral blood lymphocytes from patients with systemic lupus erythematosus (SLE),

particularly those with the complication of vasculitis, and found that VLA-4 and LFA-1

expression was increased in SLE patients with vasculitis, while LFA-1 but not VLA-4

expression was increased in those without vasculitis. These results suggested a role of

VLA-4 in the pathogenesis of vasculitis in SLE. Functional studies further demonstrated that

adhesion to cytokine-activated human umbilical cord vein endothelial cells and to the CS-1

alternatively spliced domain of fibronectin was significantly increased in SLE patients with

vasculitis. Analysis of the functional epitopes on the alpha 4 chain demonstrated that

antigen densities of all the functional epitopes were increased in those with vasculitis,

indicating that the increased expression of VLA-4 resulted from the increased number of

VLA-4 molecules, and was not secondary to an increase in one particular functional

epitope. Immunoprecipitation studies further support these results. Interestingly, high

molecular weight bands associated with VLA-4 were observed in about half of the SLE

patients with vasculitis. These results introduce a possibility that upregulation of integrin

adhesive receptors has a potential role in the pathogenesis of vasculitis in SLE.

Research Article

Find the latest version:

Upregulated Expression

and

Function of Integrin

Adhesive

Receptors

in

Systemic Lupus Erythematosus Patients

with

Vasculitis

TsutomuTakeuchi, Kouichi Amano, Hiromi Sekine, Jun Koide, and TohruAbe DepartmentofInternalMedicine, SaitamaMedicalCenter,Saitama MedicalSchool,

Saitama 350,Japan

Abstract

Upregulation of integrin adhesive receptors has been

impli-cated in variouspathologicalconditions. We examined

expres-sion and function ofintegrinadhesive receptorsonperipheral

blood lymphocytes from patientswithsystemic lupus

erythe-matosus(SLE), particularly those with the complication of vasculitis, and found that VLA4 and LFA-1 expression was

increasedinSLE patientswithvasculitis, whileLFA-1butnot VLA-4expression was increased inthose without vasculitis. These resultssuggestedarole ofVLA4 inthe pathogenesis of vasculitis in SLE. Functional studies further demonstrated that adhesiontocytokine-activatedhuman umbilical cord vein endo-thelial cells and to the CS-1 alternatively spliced domain of

fibronectin was significantly increased in SLE patients with

vasculitis. Analysis ofthefunctional epitopes onthea4chain

demonstrated that antigendensities of all the functional

epi-topeswereincreasedinthose withvasculitis, indicatingthat the

increased expression ofVLA4 resulted from the increased numberof VLA4molecules,and wasnotsecondarytoan

in-creasein oneparticular functionalepitope.

Immunoprecipita-tion studies further support these results. Interestingly, high molecular weight bands associatedwith VLA4wereobserved inabout halfof theSLEpatients with vasculitis. These results introducea possibility that upregulation ofintegrin adhesive

receptors hasapotential roleinthepathogenesisof vasculitis inSLE. (J. Clin. Invest. 1993.92:3008-3016.) Key words: sys-temiclupus erythematosus * integrins-VLA4_*LFA-1_*

vascu-litis

Introduction

Systemic lupus erythematosus (SLE) is a multi-system dis-order characterizedbytheproductionofavariety of autoanti-bodies(1).Some autoantibodiesarepathogenic (2)andcause

tissuedamage by directly reactingwith the surface ofplatelets

anderythrocytes (3),orby formingimmunecomplexesupon

Address correspondencetoDr. Tsutomu Takeuchi, Department of In-ternal Medicine, Saitama Medical Center, Saitama Medical School,

1981Tsujido-machi,Kamoda,Kawagoe-shi, Saitama 350,Japan.

Receivedfor publication15April1993andinrevisedform12July

1993.

1.Abbreviationsusedinthispaper:ECM, extracellularmatrix;

HU-VEC, human umbilicalcordveinendothelial cell;KLH, keyhole

lim-pethemocyanin; MFR,meanfluorescence intensity.

activation of complement (4). Anti-DNA and DNA immune

complexeshavebeen eluted from the affected kidneys in SLE

patients(5),andcan causelupus glomerulonephritis, indicat-ing that immune complex-mediated tissue injuries are of

cen-tral importance in the pathogenesis of lupus nephritis. This is the best understood scenario for the pathogenesis of SLE. Al-though the deposition of immune complexes in the tissue has been implicated in the pathogenesis of SLE, it is still possibleto

assume additional and alternative mechanisms for the tissue injuries in the disease. In this regard, immunopathological find-ings bySavage (6) showed that immune complexes can rarely be found in the blood vessels from biopsy specimens of SLE patients with vasculitis.

Recently, Phillip et al. suggested that an alternative mecha-nism of vascular damage mayoccurin SLEinvolving intravas-cular activation and aggregation of neutrophils in the absence ofimmune complex deposition (7). They introduced the

possi-bility that the Schwartzman reaction isamodel of this type of vascularinjury, and that it is not mediated by immune

com-plexdeposition.Cytokine-primed neutrophil-dependent vascu-lardamage has thus been demonstratedtoparticipateina rab-bit model of hemorrhagic vasculitis in the absence of immune complexdeposition (8). If this is the case for the mechanism of tissue injury in SLE, interaction between leukocytes and endo-thelium or substitium in connective tissues may haveamajor role inpathogenesis of the disease.

Recent progress in thestudy of adhesion molecules has re-vealed that several sets of surface molecules participate in cell-cell and cell-cell-substitium adhesions (9). Members ofthe

_#

I inte-grin family of adhesion molecules, which comprises nine dif-ferentachains and a noncovalently associated

_#

1 subunit, are major adhesive receptors for the extracellular matrix

(ECM)'

proteins ( 10). On the other hand, the (32integrin family

medi-atescell-cell adhesion. In the (31 integrin subfamily, VLA-4, and now VLA-3, are exceptionally involved in both cell-cell andcell-ECM adhesion. Thefunctionandexpressionof these adhesive receptors are properly regulated (9-11). Upregula-tion ofthese adhesive receptors has been found in several

patho-logicalconditions suchasatherosclerosis(12),and bronchial asthma, as well as in the synovial T cells of rheumatoidarthritis

( 13). We alsoexperiencedacaseof SLEcomplicated by

necro-tizingvasculitis of themesenteric artery, which resulted in per-foration of the ileum. In thisparticular patient,theexpression

and function of VLA-4 was increased during the vasculitic event,indicatingarole of theintegrinadhesive receptors in the pathogenesis of vasculitis in SLE.2 A similar ideacame from

2.Takeuchi,T.,K.Aoki,J.Koide,H.Sekine,andT. Abe. 1993.

Upre-gulated expressionandfunction of SM-27(VLA-4) antigen on T cells

from a patient with lupus erythematosuscomplicated by vasculitis.

Manuscript submitted for publication.

3008 T. Takeuchi,K. Amano, H.Sekine,J. Koide, andTAbe

J.Clin. Invest.

© The American Society for Clinical Investigation,Inc.

0021-9738/93/12/3008/09 $2.00

thefollowing studies in which we conducted a series of

experi-ments toidentify the surface structures important for the patho-genesis of SLE, and attempted to develop mAbs directed against the surface antigen on lymphocytes from SLE patients

onthebasis ofincreased reactivity with the patients' lympho-cytescompared to normal individuals. Among those mAbs, we found that a clone, SM-27, recognizing the surface antigen whoseexpression on SLE lymphocytes was greater than nor-mallymphocytes, identifies the VLA-4 antigen ( 14).

These studies, however, were performed only in a case of SLE complicated by vasculitis, and there has not yet been a determination of the expression and function of integrin adhe-sive receptors in substantial numbers of SLE patients with vas-culitis to permit us to draw some conclusions.

Inthisstudy weattempted to delineate the expression and function of these adhesive receptors of the integrin family on peripheral blood lymphocytes in patients with SLE. We demon-strated that expression andfunctionof VLA-4 as well as LFA- 1 wereupregulated in SLE patients with vasculitis, while LFA- 1 but not VLA-4expressionwasincreased in SLE patients with-out vasculitis. Moreover, the function of VLA-4 in terms of adhesiveness tofibronectinandcytokine-activatedendothelial

cells isincreased inthesepatients.Immunoprecipitation stud-iesshowing increased140-, 80-,and 70-kD bands ofthe VLA-4

achain inpatients withvasculitisconfirmed the above results.

Surprisingly, wealsofoundthat 1 80-kD extra-bands were ob-served by immunoprecipitation of VLA-4 from peripheral bloodlymphocytesofthevasculitis patients.These results in-troduce a possible role of integrin adhesive receptors in the

pathogenesis of vasculitis inSLE.

Methods

Patients. Thesamplepopulation consisted of 27 patientswith SLE

satisfyingtheclassification criteriaproposed bytheAmerican Rheu-matism Association ( 15). Diseaseactivity wasdetermined for each

patientatthetimeof blooddrawing bytwophysiciansonthebasis of multi-system disease manifestations, as described previously (16). Briefly,manifestations includedfever, rash,arthritis,serositis,

vasculi-tis,activenephritis,andactive centralnervoussystem disease. Patients with twoor moremanifestationswereconsideredtohavesubstantially

activedisease. Diseasedurationwasdetermined from the date of

diag-nosistothatof blooddrawing.

Patients with known renal diseasewerethosewith hematuria(2 10 redblood cells perhigh powerfield), proteinuria(. Ig per 24h),RBC casts,orcellular casts, eithercurrentlyorin the past. Sixpatientswere

judgedto havevasculitisbydemonstrationofnecrotizingvasculitis in thebiopsyspecimens.Their clinicalmanifestationsattributableto

vas-culitisweredigitalinfarction,skinulcers,opticnerveneuritis,

perfora-tionof theileum,andmononeuritismultiplex.Ninepatientswere

un-treated(eight patientswithoutvasculitis andonewithvasculitis). 18 patientswerereceivinglowtointermediate doses ofprednisolone(5to

30mg/d),butnoneof thepatientshadreceivedcytotoxic drugsduring

thepreceding6mo atthe time of blooddrawing.

Thenormalcontrolpopulation consisted of42healthy individuals withnosignificantillness.

mobs.Spleen cells fromBALB/Cmiceimmunizedbythree

intra-peritonealinjectionsof 1 xI07PBLsfromapatientwith SLE

compli-catedby vasculitis (caseS.M., 38-yr-oldfemale)werefused with NS-l myeloma cellsat aratio of10:1 byastandard procedure, and were

distributedinto 96 wells with normalspleencellsasfeeders.

Hybrid-omasproducingantibodiesreactive withPBLfrom the SLEpatient

wereselectedbymeansof indirectimmunofluorescence witha

cytoflu-orometerand usedfor furtheranalysisasdescribedpreviously ( 14).

The isotypeof SM-27 wasIgG1.The otheranti-VLA-4 antibodies are

asfollows: HP2/1 (IgGl; Immunotech S.A., Luminy, France); L25

(IgG2b; Becton Dickinson Immunocytometry Systems, San Jose, CA); andB-5G10 (IgGl; Upstate Biotechnology, Inc., LakePlacid,

NY). Leu 5b (CD2), Leu 4 (CD3), Leu 3 (CD4), Leu 2 (CD8), G25.2 (CDlla), Leu15(CDl lb), andL 130(CD18)werepurchased from Becton Dickinson Immunocytometry Systems. Tl 1 (CD2), T3

(CD3), T4 (CD4), T8 (CD8), 2H4 (CD45RA), and 4B4(CD29)

were purchased from Coulter Immunology (Hialeah, FL). T2/7 (VLA-1) was purchased from T Cell Sciences (Cambridge, MA). BU 15(CDlIc), Gi-9 (VLA-2), and SAM-l (VLA-5) were obtained from Immunotech S.A. M-kid2 (VLA-3) and GoH3 (VLA-6)were

obtained from Upstate Biotechnology, Inc. Anti-ICAM-l and anti-VCAM-1 were obtained from British Bio-technology Ltd. (Abingdon, OX, England).

Immunofluorescence analysis. Cells were incubated with mAbs for 30 min onice, then incubated with a F(ab')2 fragment of goat anti-mouse antibodyconjugated with FITC (Tago, Inc., Burlingame, CA) asdescribed previously (16).

The flow cytometer (Becton Dickinson Immunocytometry Sys-tems) was run on fixed laser power (600 mW at 488 nm) over the periodof the study. Flow cytometric standardization was achieved by running 4.05-imfluorescent beads (Becton Dickinson Immunocyto-metry Systems), which exhibited constant light scatter and

fluores-cenceproperties. Background fluorescence reactivitywasdetermined with control ascitic fluid obtained from mice immunized with a nonse-creting hybridoma. Control and patient samples were run with the FITCphotomultiplier adjusted to a high enough voltage such that the fluorescencehistogram derived from the forward angle vs. side

scatter-gatedlymphocyte population yielded 2-7% background autofluores-cence intensity in the range reserved for positively FITC-stained cells. Allsamples were, then, analyzed using the fixed photomultiplier set-ting. For each sample, 10,000 cells were analyzed on a log fluorescence scale on the flow cytometer. For determination of the surface antigen expression, the mean log FITC-fluorescence channel of the positively-stained cells wasdetermined from a single-parameter histogram.

Isolation ofcells. PBL from normal individuals and patients with SLEwereseparatedfromheparinized venous blood by Ficoll-Hypaque density gradient centrifugation (Pharmacia LKBBiotechnology, Upp-sala, Sweden) as described previously(16). PBL were suspended in medium containing 10% FCS (Hazelton Biologics, Inc., Lenexa, KS)

andwereincubated on 1% serum-coated dishes, and passed over a nylon-wool column to yield a T cell-enriched population. This T cell population was more than 92% reactive with anti-T3 antibody.

ECMproteinsandtheirfragments. Fibronectin, laminin, and colla-gentypesI, II, III,IV werepurchasedfrom Sigma Chemical Co. (St. Louis, MO). Oligopeptides corresponding to the VLA-4-binding site of fibronectin were synthesized byan amino acid synthesizer (Biolynx Peptide Synthesizer; Pharmacia LKB Biotechnology), then purified with an FPLA system with a mono-Q column. The oligopeptides GPEILDVPST (CS-lA) and TLPHPNLHGP (CS-1B) were

conju-gated tokeyholelimpet hemocyanin (KLH) by the maleimid method

as described in the manufacture's brochure (Pierce Chemical Co., Rockford, IL).

Assaysofcell attachmenttoECM proteinsand HUVECs.Forcell attachmenttoECMproteins,flat-bottomed 96-well microtiter plates (Costar Corp., Cambridge, MA)werecoated with 100y1 of 0.1 M

NaHCO3 containing 10 ,ug offibronectin, laminin, collagen, CS-1A,

andCS-I Boligopeptides for2 h at37°C and thencoatingwasblocked with RPMI 1640medium in the presence of 1 % BSA (Sigma Chemical Co.) for 1 h at 37°C. For attachment tocytokine-activated human

umbilical cord vein endothelial cells (HUVECs), confluent

mono-layers of HUVECs activated with recombinantIL-1,Band TNFafor 24

h(10U/well; Genzyme, Boston, MA)wereprepared in 96-wellplates. Cellsweresuspended inRPMI1640mediumcontaining1%BSA,and

plated in triplicate on ECMprotein-coated plates (100 ,l final

vol-ume). After 30-min incubation at 37°C for ECM bindingassay of 15-minincubation atroomtemperaturefor monolayerbindingassay, unbound cellswereremoved by threewashings with attachment

dium. Bound cellswerefixedwith 1.25% glutaraldehyde in PBS (0.01 M,pH 7.4) at4VCovernight,andstained with 4% trypan blue. Bound cells werecounted underamicroscope,and the resultswereexpressed asbound cells per mm2orlysed with 10% SDS and theOD590was measured byaTitertek Multiscan plus (FlowLaboratories,Horsham, PA).Forantibodyblockingassay, mAbswereadded 30 minbefore the adhesion assay.

Cellsurface labeling and immunoprecipitation. Cells were surface labeledwithsulfo-NHS-biotin (PierceChemicalCo.)asdescribed in themanufacture'sbrochure,andlysedin the presence of 1% NP-40as

describedpreviously (17).Celllysateswerepreclearedwith insoluble proteinA(SigmaChemical Co.), and incubated with mAbs. The im-mune complexes were precipitated with protein G-agarose beads (Pharmacia LKB Biotechnology), and washed five times withlysis buffer. Samplebufferwasadded,and themixturewasboiled for 5 min. Thesamples were run on6%polyacrylamidegels.After

electrophore-sis, the gelswereequilibratedfor 10 min inatransferbuffer(25mM Tris-HCl, 192 mMglycine,and 20%methanol) andproteinswere

elec-trophoretically blotted onto polyvinylidene difluoride membranes

(MilliporeCorp.,Bedford,MA). The membranesweresoakedat37°C

for 1hinblockingagents(Blockace;DainipponPharmaceuticals, To-kyo), and streptoavidin-horseradish peroxidase conjugate

(Amer-sham International, Amersham, UK) was added at a dilution of 1:1,000 with PBScontaining0.1% Tween-20 and 10% Blockace. After incubationat37°C for 1 h, the membraneswerewashed three times and thedetection of the horseradishperoxidase-labeledproteinswas

carriedoutwithchemiluminescence-enhancingreagentsasdescribed in themanufacture's brochure(AmershamInternational).The treated membraneswerevisualizedonECL x-ray films.

Results

Expression ofintegrinadhesive receptors in SLE. To delineate

arole ofintegrinadhesive receptors in thepathogenesis ofSLE withvasculitis,weexaminedexpression ofintegrinson PBL in SLEpatientswithvasculitis,thosewithoutvasculitis, and

nor-mal individuals. The age, sex, disease activity, diseasetype,

diseaseduration, clinical manifestations of the

vasculitis,

and

dose of prednisolone in SLE patients studied are summarized in Table I. Themean ageandsexwerecomparable between the SLE patients and normal individuals. All patients without

vas-culitishad active disease,exceptfortwo, while all six patients

with vasculitiswerejudgedtobeactive. Diseasetypeand

dis-easedurationwerecomparable between SLE patients without and with vasculitis. The dose of prednisolonewas not statisti-cally significantbetween these two populations of SLE.

The representative cytofluorographic pattern in Fig. 1 shows that CD29, CDl la, and CD1 8 antigens haveabimodal distribution in fluorescence intensityon alog scale, indicating that therearetwodistinctfractions expressing highorlow lev-els of CD29, CD I Ia,andCD 18. Therefore, thepercent

expres-sion and mean fluorescence intensity (MFI) of the population expressing high levels of antigens are shown as CD29 high,

CDl la high, and CD1 8 high. As seen in Table II, there areno

significant changes in percent expression of CD3, VLA- 1, VLA-2, VLA-3, VLA-4, VLA-5, and VLA-6, whereas the

per-centexpression of the CD29 high population was increased in the SLE patients with vasculitis, supporting the previous

obser-vation.With respect to,32integrins,the percentexpressions of CD I la high, CD 1 I c, and CD 18 high are significantly higher in the vasculitis patients than in normal individuals (56±29%, 45±6%, and 57±29%, respectively), while CD1 la high and CD 18 higharealso increased in the SLE patients without

vascu-litis (49±18%, and 47±19%, respectively). When we deter-mined theantigen densities bycomparingthe MFI of the sur-face antigens, that ofVLA-4 a was significantly increased in the SLEpatients withvasculitis,compared to normal individuals (101±56,and61±22,respectively),whereas nosignificant in-crease wasobserved in the SLEpatients without vasculitis.

Sim-ilarly,the MFIofCDI Ia aswell asCD18wasincreased inthe

vasculitis patients. These results clearlyindicate that expres-sion of VLA-4 and LFA- 1 antigens wasupregulatedinthe vas-culitis patients. It is well accepted that after activation it is the CD4+Tcells that express adhesion molecules such as CD29,

Table I.CharacteristicsofSLEPatientsStudied

Disease Disease Clinical manifestation

Subject Number Age Sex Activity subset duration of thevasculitis Prednisolone

(F/M) (active/inactive) (R/NR) (mg/d)

Normal 42 31.8±6.4 42/0

(20to48)

SLE 21 35.5±12.6 21/0 19/2 4/17 21±12 8.8±9.0

(14to59) (3-48) (0to30)

SLEwith vasculitis 6 31.2±7.9 6/0 6/0 1/5 25±11 9.8±4.7

(23to41) (9-37) (0-15)

S.M. 41 F Active NR Digitalnecrosis 10

Opticneuritis

S.A. 37 F Active NR Healperforation due to 10

mesenteric vasculitis

F.S 24 F Active NR Digital necrosis 12

N.Y. 23 F Active R Skinulcers 15

Y.K. 38 F Active NR Digital necrosis 0

T.K. 24 F Active NR Skin ulcers 12

Digitalnecrosis

Resultswereexpressedasmean±SD. Disease subset was shown as R (renal) or NR (non-renal). Disease duration was expressedasmean months±SD.

114

(58) : ..

IW--E

152

(42)

A.

.0.

A-VLA-4p

(CD29)

LFA-1a (CD1la)

LFA-1p (CD18)

Figure1. Flowcytometricanalysis

for the integrin adhesivereceptors onPBLsfromrepresentativecases

ofanormalindividual,aSLE

pa-tient without vasculitis and thatwith

vasculitis. The ordinaterepresents

the forwardlightscatterand the ab-scissarepresentsthelog fluorescence intensity.

PgP-1, and LFA-3 accompanied by anisoform change from

CD45RAtoCD45RO(18). Accordingly,we nextdetermined

theexpression of these

integrins

onCD4+ T cells in SLE. As

shown in Table

III,

the percentage of CD4+CD45RA+ cells

decreasedandthatof

CD4+CD29high+

cells increased in SLE both with and without vasculitis.Althoughthepercent

expres-sion ofVLA-6was

significantly

high

inthe SLE

patients

with

vasculitis,thatofVLA-5 andVLA-6wasalso

high

inthe SLE

patientswithoutvasculitis, whencomparedtothat of normal

individuals. Antigen densitiesasdetermined

by

the MFIona

flow cytometer have shown thatexpressionof VLA-4(a4)was

TableII. ExpressionofIntegrin Adhesive ReceptorsonPBLs

from SLE and NormalIndividuals

SLE

Normal Vasculitis(-) Vasculitis(+)

Sufaceantigen (n=42) (n=21) (n=6)

%expression

CD2 81±4 82±10 85±9

CD3 72±6 79±11 75±15

CD4 41±5 38±13 31±12§

CD8 26±6 35±12 35±10$

VLA-1 (al) 6±5 8±10 16±14

VLA-2 (a2) 4±3 9±11 9±2

VLA-3(a3) 16±15 20±8 22±9

VLA-4(a4) 92±8 93±7 96±4

VLA-5(a5) 54±14 48±14 58±18

VLA-6 (a6) 30±17 22±17 31±12

CD29high 51±12 58±15 65±16*

CD1lahigh 36±9 49±18§ 56±29$

CD1lb 24±7 30±15 26±6

CDllc 24±8 29±12 45±6$

CD18high 36±10 47±19* 57±29*

HML-1 (aE//37) 0 0 0

Meanfluorescence

intensity

VLA-4(a4) 61±22 68±24 101±56§

CD29high 125±78 120±57 120±57

CD1lahigh 186±52 204±40 259±42*

CD 18high 218±46 217±42 272±47*

*P<0.05.

*P<0.01.

§P<0.005.

increased inMFIin the SLE patients with vasculitis, while that of CD1 la was increased in the SLE patients both with and

withoutvasculitis.

The results showed that VLA-4 antigen expressed on PBL and CD4+ T cells wassignificantly high in the SLE patients with vasculitis, while LFA-1 expression was increased in the SLEpatientsbothwithandwithout vasculitis.

Adherence of T cells to cytokine-activated HUVECs in SLE.Theaboveresults clearly show that the integrin adhesive receptors wereupregulated on the PBL as well as on CD4+ T cellsfromSLE patients with vasculitis. Next, we attempted to

determinethe function ofthese adhesive receptors. First, we

studied adhesiontoendothelial cells, since this process should beof primary importancefor eliciting vascular inflammation

(9, 19). Throughthe use of IL- 1 ,- and TNFa-stimulated

HU-VECs,whichhave already been shown to express ICAM- 1 and VCAM- 1 antigens (20), the ligands for LFA- 1 and VLA-4, respectively, adhesion of peripheral blood T cells from normal

individuals,and patients without, or with vasculitis, was

exam-ined. Asshownin Fig. 2, T cells from patients with vasculitis

showedasignificant increase inadhesion to cytokine-activated

endothelial cells( 189±47), whereas T cells from patients

with-outvasculitisshowed aslight but significant increase in

adhe-Table III. Expression ofIntegrin Adhesive Receptors

onCD4+ TCellsinSLE Patients with or without Vasculitis

SLE

Normal Vasculitis(-) Vasculitis(+)

Subset of CD4 (n=42) (n=21) (n=6)

%of CD4+ subset

CD45RA+ 39±10 29±18* 29±10*

CD29high 46±10 61±26* 71±31*

VLA-4(A4) 95±7 89±23 97±12

VLA-5(aS) 49±7 68±11t 65±32

VLA-6(a6) 44±5 74±28t 71±34*

MFH

VLA-4(A4) 66±18 66±78 114±35t

CD29high 115±12 120±16 125±20

CD11ahigh 195±42 235±48* 270±81*

CD18high 242±61 266±52 259±29

*P<0.05.

*P<0.01.

IntegrinsinSystemic Lupus Erythematosuswith Vasculitis 3011 Normal

SLE

SLE with vasculitis It

L-I

NEG

70

.,

126

.t

VLA-4a

.$~~~~~~

(CD49d)

125

(62)

li'

6188

256

1

(84) 188 (68)

is- id- lp

'i

.ja.0

."

T cells adhered to cytokine-activated HUVECs (cells/mm2)

Figure 2. Adhesion of peripheralblood Tcellsfrom

sex-andage-matched normal individuals (black bar;n=27),

SLEpatients withoutvasculitis(graybar;n=21 ),and

those withvasculitis(shaded bar;n=6)to

cytokine-ac-tivatedHUVECswasexamined inthepresenceofmAbs directed against ICAM-1 and VCAM-l.*P<0.05, **P

<0.01.

sion (62±40) and it was only minimal in normal controls

(28±9). Sincewedetectnotonly VLA-4/VCAM-1interaction

butalsoLFA- /ICAM-1 interaction inthisassaysystem,we

determinedtheindividualpathway by blocking with mAbsto

ICAM- and/or VCAM 1. In thepresenceof anti-ICAM-1

ad-hesion of T cells from the SLE patients with vasculitis was

decreasedto - 75%, but it_wasstillpronounced (142±46). In

contrast, adhesion ofT cells from the SLE patients without vasculitis was decreased almost the same amount as thatof

normal individuals (30±21 vs. 26±13). In the presence of

anti-VCAM- 1, inhibition of adhesion of T cells from the SLE patients with vasculitiswas58% ofthatinthe absence of

block-ing antibody, while little inhibition wasobserved inthe SLE

patients withoutvasculitis. The resultssuggestthat the

path-ways involving VLA-4/VCAM-I and LFA-1/ICAM-1 were

functionally upregulated in the SLEpatients with vasculitis, whereas in those without vasculitis only the LFA-1/ICAM-1

pathway was functionally upregulated, supporting the above

results obtained with the expression of integrins. Although

therewassignificantinhibition of T cell adhesionto

cytokine-activated HUVECs inthepresenceofanti-ICAM-1and

anti-VCAM-1,it isworth noting that the inhibitionwasonly 80% in thepresence ofboth anti-ICAM-1 and anti-VCAM-1. This

implies that additional and/or undefined pathwaysmay con-tributeto theincreasedadhesion of T cells of these patients. Nevertheless, VLA-4/VCAM-1 and LFA-1/ICAM-1 seemed

tobemajor pathways for increased adhesiontoactivated HU-VECsinthesepatients.

Adhesion ofperipheral TcellstoECMproteins. Itis well knownthatamong (1 integrins VLA-4 isauniquemolecule that mediates both cell-cell and cell-ECM protein adhesion (21 ). Therefore,weexamined alternative functions of adhesive

of VLA-4toECMproteinsinthese patients. For thispurpose, wesynthesized oligopeptides correspondingtothe

VLA-4-bind-BSA Fibronectin CS-IA

Normal

SLE

SLE+ vasculitis

ing site of fibronectin,which has been shown tobetheCS-I

alternatively spliced domain. By using oligopeptides desig-nated CS-lA (GPEILDVPST)as wellasnonbinding control

peptides designatedCS-IB(TLPHPNLHGP), weexamined,

asdepictedinFigs.3 and4,adhesion of T cellstothese oligo-peptidesandtothreeECMproteins.AsshowninFig. 4,T cells

from the representative patient with vasculitis showed

pro-nounced adhesionnotonlytofibronectin but alsotothe CS-lA

oligopeptides,whiletheyadheredonly slightlytolaminin,and

adherenceto the negative control, BSA, wasminimal. Such enhanced adhesionwasnot observedinTcells from normal

individualsand from the SLEpatientwithout vasculitis.Fig.3 confirmed the abovefindingthat T cells from thepatientswith

vasculitis showedasignificantincrease in adhesionto fibronec-tin and CS-lAoligopeptides,while adhesion to CS- lB and

la-minin wasonly slightlyincreased. On the otherhand,we

ob-servedaslightincrease in adhesiontofibronectin,CS- A,and

CS-lB inSLE withoutvasculitis,but the increasewasnot

signif-icant in comparison with that in normal individuals. These

resultssuggestthat increased adhesiontofibronectinaswellas

toCS- lAoligopeptides byTcells from thepatientswith

vascu-litiscanbe ascribedtotheincreased function ofVLA-4. More-over, weshowed that anti-VLA-4antibodyalmostcompletely

inhibited this adhesion (data not shown), confirming the

above results.

Clinicalcourseoftherepresentative patientsand the

expres-sion andfunction of integrin adhesive receptors. We found upregulated expressionandfunction of theseintegrinadhesive receptors in the SLEpatientwith vasculitis.Thenextquestion

is whether such alteration of the integrin receptors may be

correlated with the clinicalcourse of the disease. To address thisquestion,weexamined forexpressionofintegrinadhesive receptorson the PBL frompatientswithvasculitiswhomwe

could followuponserially.Asshown in the leftpanelofFig. 5,

CS-lB Laminin Collagen

0 0.2 0.4 0.2 0.4 0.2 0.4 0.2 0.4 0.2 0.4 0.2 0.4

Adhesion to ECM proteins (OD5ss)

Figure 3. Adhesion of peripheral blood T cells from

sex-andage-matchednormalindividuals(n=27),SLE

pa-tientswithoutvasculitis(n=21), and thosewith

vascu-litis(n=6)toa seriesofextracellularmatrixproteins

andsynthetic oligopeptides correspondingtothe VLA-4

binding regionwasexamined.Adherent cellswerefixed

with formaldehyde, stained withtrypanblue, and lysed with 10%SDSasdescribed in theMethods. Resultswere

expressedasOD5sg ofindividual wells.

3012 T. Takeuchi,K. Amano, H.Sekine,J.Koide, and T. Abe

0 100 200

antibody

(-)

anti-ICAM-1

anti-VCAM-1

anti-ICAM- 1

+

anti- VCAM- 1

::L

1*~~~~~~~~~~~~~I*

ml~~~/

7/7'l llV

BSA Fibronectin CS-IA Laminin

Figure4.Adhesionofperipheral bloodT cells from anormal individual(upper column), a SLEpatientwithoutvasculitis (middle column),and thatwith vasculitis(lower column) to negative control (BSA), fibronectin,CS-IAoligopeptides, and laminin was shown.Afteradhesion to the plates, the adherent cells were fixed with formaldehyde, and stained with trypan blue. Photography was taken at 1:100.

the MFIofVLA-4(closed circle)wassignificantly elevated in

patientF.S. Soonafterthat, sheexperienceddigital necrosisof

theleftfourth fingerand was admitted to the hospital. She was

successfullytreated withincreaseddosesofprednisolone and

surgical debridement ofthenecroticfinger, andexpression of VLA-4 returned to within normal limits. Also in patient Y.K.,

expression ofVLA-4and the percentageoftheCD29 high

pop-ulation (opencircles)wereincreasedwhen she experienced a

digital infarctdue tonecrotizing vasculitis,whereas expression

of these moleculesreturned to normalinJuly 1992. However, when digital necrosis recurred shortly after that, theMFIofthe VLA-4 antigen increased again. Although the sample size is rather small and we may have towaittodraw definitive

conclu-sions from thesetwo cases, the correlation of upregulated ex-pression of VLA-4 antigen and the vasculitic events observed

herefurthersupporttheidea that such adhesive receptors may

participateinthepathogenesisofvasculitis inpatientswith SLE.

Functionalepitopesof the VLA-4antigen inSLE patients with vasculitis. Theabove results suggestthat upregulated ex-pression andfunctionof VLA-4 occurs on PBL from the SLE patients with vasculitis, and it has been reported that there are three distinct and independently inhibitable functional

epi-topes onVLA-4 molecules (22). Since the increased expres-sion of VLA-4 obtained in the above study was derived from MFIbySM-27, it remains to bedetermined whetherepitopes other than that recognized by SM-27 areupregulated or not. Totest this, we used aseries ofanti-VLA-4 mAbsand exam-ined PBL from thevasculitispatients forexpressionof the dis-tinct functionalepitopes.SinceSM-27 could block the aggrega-tion inducedbyL-25definingthe B2epitope,adhesiontoCS-I

92'11 93'2

Vasculitis and date -100

t digital necrosis tdigital

necrosis

90 11 91'11 92'7 93'2

of flow cytometric analysis

100

0

0

0 .o

50

.0

CY) csJ

.O.

0

0)

CL

0o

Figure 5. Timecourseof theexpression of VLA-4 alpha and beta chains in thetworepresentative SLEpatientscomplicatedbyvasculitis. Percent expression ofCD29(VLA-4betachain)high+

populationwasshownasopen circles.

IntegrinsinSystemic LupusErythematosus with Vasculitis 3013

~1'-

200

cn

0

0 ₁₀₀ C 0 0

+ O

c 0 0 0 0

0

0

Case F. S.

tdigital necrosis

mean fluorescence intensity on a log scale

50 100 150

antibody specificity

4B4 ,B,

SM-27 a4:B1

HP 2/1 a4:B1

L25 a4:B2

B-5G10 a4: C

offibronectinandtoVCAM-1(datanotshown),SM-27seems

todefine BI functional epitopes. Using HP2/1 (defines B1),

SM-27(Bi), L-25 (B2), B-5G10(C),we observedasimilar

increase in the MR of all the epitopes recognized by these

mAbs.As shownin Fig. 6,the MFI ofHP2/ 1, SM-27, L-25,

and B-5G10wasincreasedinPBL fromthevasculitispatients

and theMFI ratiosof vasculitispatientsand normal individ-ualsare 1.75, 1.72, 1.77,and 1.67, respectively.

These resultssuggest thatthe increase in MFI of VLA-4

antigendidnotresultfromanincreaseofoneparticular

func-tional epitope but was due to an increase in the number of VLA-4molecules expressedonPBL.

Structureofthe VLA-4antigeninSLE patients with vascu-litis. To furtherexamine thestructural basis of VLA-4

mole-culesexpressedon PBL, wecarried outan

immunoprecipita-tionanalysisofthe VLA-4antigenonPBLfromthese patients. AsseeninFig. 7,weclearly showed 140-kD bands

correspond-ingtothe intacta4chain ofVLA-4antigenaswellas80- and

70-kD fragments, from normalindividuals, theSLEpatients without vasculitis and those with vasculitis. The 80- and 70-kD

bands hadmoreintensesignalsinthepatients with vasculitis

than those ofnormalindividuals andthepatients without

vas-culitis, supportingthe aboveobservations that the number of

VLA-4molecules expressedon PBL in the SLE patientswith

vasculitiswasgreaterthan those of normal individualsand SLE

patientswithoutvasculitis. Surprisingly,weobservedanextra

band migrating around 180kD along with the 80-, 70-,and 140-kDa4bands inthreeof the six patients with vasculitis.

Discussion

Inthisstudy,wedemonstrated upregulation in expressionand functionof, 1 aswellasfl2 integrinsonPBLfrom SLE patients

with vasculitis. More specifically, expression of VLA-4 and

LFA-1wassignificantlygreaterinpatients withvasculitis than innormal individuals, but LFA-1wasalso upregulated inSLE

patientswithoutvasculitis. Function intermsof adhesiveness

tocytokine-activated endothelialcells and the CS-1domain of

fibronectinwasenhanced in SLE patients withvasculitis. More-over, amongthose patients with upregulation ofintegrin adhe-sive receptors, halfof the patients with vasculitis had the high

molecularmassform (180-kD) ofVLA-4.

Itis wellaccepted that CD4+Tcellsaredistinguished by their differential expression of isoforms of CD45, namely

CD45RA andCD45RO, with distinct functions (23-25). In

addition,it hasbeenshown thatexpression ofthe 131 and ofa4,

Figure 6. MFIoffunctionalepitopesonthe VLA-4alpha chainrecognized byvarious anti-VLA-4 antibodies ex-pressedonPBLfromsex-andage-matchednormal indi-viduals (black bar;n=27), SLEpatients without vascu-litis(graybar;n=21), and those with vasculitis(shaded bar;n=6)wasshown. *P<0.05.

a5,anda6ontheCD45RO subsetof CD4+ Tcells isseveral

timeshigher thanontheCD45RAsubset of CD4+ T cells(1 1, 21). Similar observations werereportedforexpressionof 32

integrins (9, 18). These findings suggest that the subsets of CD4+ T cellsmayrepresenttwostages ofTcelldifferentiation,

memory and naive (26-28). The present study showingthe decreased percentage of CD4+CD45RA+ cellsinSLEpatients

both with and without vasculitis supports theprevious observa-tions( 16, 23).Onemaywonder whether the increased

expres-sion of these adheexpres-sion molecules could be the result of acceler-ated conversion from CD45RAtoCD45RO in these vasculitic

patients. However,theincreasedexpressionofVLA-4antigen

wasobserved only in thepatients with vasculitis, not inthe

patientswithoutvasculitis,whileboth kinds ofpatientshada

p

p 180 e

140

w^l

lo

804~ ~ ~ ~ ~ ~~..

p704

K.T.

Nm11

Normal

qp... 4-p 180

(-) (+)

vasculitis

SLE

Figure7.Immunoprecipitation analysisofVLA-4expressedonPBL

fromanormal individual,tworepresentativecasesof SLEpatients

withoutvasculitis,and those with vasculitis. VLA-4antigenwas im-munoprecipitated bySM-27 andprotein G-sepharosebeads,run on

6%homogeneous acrylamide gels. Afterelectrophoresis, the_proteins

wereblottedontoPVDF membranes and detectedby

chemilumines-cence-enhancingreagentsasdescribed.

decrease inCD4+CD45RA+ cells withanormal level ofVLA-5,which arguesagainst thispossibility.The patternof expres-sion in

_f3I

integrins in whicha 4isincreased, but

a'

is normal hasalso beenfound in synovialTcells inpatientswith rheuma-toidarthritis (29). Such differentialexpression ofa4 anda5has been demonstrated in human PBL after activation with ConA and PMA(30). Given the evidence that thestructureof thea14

gene promoter isquite different from that of the a5 gene(31, 32),thedifferentialexpression of integrinsinthe SLEpatients

withvasculitiscould bedeterminedatthe message level. Theconstitutive expressionofintegrinsdoesnot

necessar-ilyindicatethatthey arefunctionally active (9,11, 21). Accord-ingly, wecarriedoutfunctional studiestodeterminethe adhe-sive functionofthesemolecules, and foundthat VLA-4and LFA- 1 werefunctionally upregulatedaswell. Thestudy

ofinhi-bition by various anti-VLA-4 antibodies showing that

en-hancedadhesionof the patients'TcellstotheCS-1 domain of fibronectinwascompletelyinhibitedbyanti-VLA-4 antibod-ies,indicatesthat VLA-4 hasthemajorreceptors forbindingto

the CS- 1domain offibronectin in thesepatients.On the other hand,enhancedadhesion ofthepatients'Tcellsto cytokine-activated endothelialcellswasinhibited- _80%bya

combina-tionof anti-VCAM-1andanti-ICAM-1.Thus,theresults

indi-catean alternative adhesion pathwayinthis adhesionsystem.

Sinceotherpathwayssuchas

CD44/hyaluronic

acidand

car-bohydrates/ELAM-1arepostulatedtoinvolve adhesionto

en-dothelialcells(33-35),the

possibility

shouldbeexaminedin the future. Nevertheless,

VLA-4/VCAM-

1 and LFA- 1/

ICAM- 1 interactions are main contributorsto the enhanced

adhesion tocytokine-activated endothelial cells. Since it has been demonstrated thatLFA-l

/ICAM-l

and

VLA-4/VCAM-1 pathways havedifferentroles inadhesion and

transendothe-lial migration ofT cells (36), upregulated function ofboth pathways may berequiredtoinduce vascularinflammation.

Upregulation ofLFA- 1expressionwasalso observed in

pa-tients withvasculitisaswellasthose without

vasculitis,

suggest-ing that VLA-4, in addition to increasing the expression of

LFA-1, seems tohave akey role in thepathogenesis of vasculi-tis. Therefore,we focusedonthe VLA-4 molecule and delin-eatedfurther detailed mechanismsfor the increased

expression

andfunction ofVLA-4.

Theimmunoprecipitation ofthe VLA-4antigen fromthe

peripheralblood

lymphocytes

showed that the

80-,

and 70-kd

fragments of

aC4

occurred with more

intensity

in the SLE

pa-tients with vasculitis than normalindividuals and those

with-outvasculitis. These results furthersupport theevidence

ob-tained by indirect immunofluorescence

analysis, using

a

vari-etyof anti-VLA4antibodies

recognizing

different functional

epitopes of thea'4chain. Severallines of

evidences, therefore,

suggest the upregulated

expression

ofVLA-4

molecules,

but notofaparticular

epitope

ofthea'4 chain.

Surprisingly,

the

highmolecular massbandswereidentifiedin three ofthe six SLEpatients with vasculitis. We proposetwo

possible

sources

of this 180-kD band:(a)the

high

molecularmassform of the VLA-4 a chain, and (b) another molecule associated with VLA-4.Sinceweusedconditions for

preparing

the cell

lysate

in which the

a3

heterodimer

dissociates,

it isless

likely

that the

180-kD bandistheotherachainofVLA, for

example

a2. aIEL with molecular massof 180-kD has been shownto associate with

f7

and display an adhesion function similar to thatof

VLA-4.However,wedidnotobservethe

expression

ofHML- 1

antigenonPBLfromthe

patients

with

vasculitis,

whichargues

against that possibility. These results, therefore, favor the former possibility. In this regard, Parker et al. recently

de-scribedanovel form of theintegrin

a'

subunit witha molecu-lar massof 1 80-kD on the PEER T cellline,and demonstrated thata!4180 anda4/150representdifferent conformations of the

same a'1polypeptide(37).Althoughthefunction and

signifi-cance ofa14/180 awaits furtherinvestigation, study is now in progress todeterminewhetherthe 180-kD bandisa4/180 or not.

Faucireported that manypatientswith active vasculitisdo nothave demonstrablecirculating or deposited immune com-plexes (38), and Savage stated that the mere presence of im-munedepositsdoes notnecessarily result in tissue injury with

an inflammatory infiltrate and fibrinoid necrosis, since

de-positsmay bedemonstrated in vessel wallswithout

accompany-ingvasculitic lesions (6). It has been suggested that an

alterna-tive mechanismsuch as theSchwartzmanreaction, but not the Arthusreaction, participates in the pathogenesis of vasculitis (7). Inthismodel,intravascular LFA- I /ICAM- 1

(leukocyte-endothelium) adhesionwasshown to be necessaryfor thistype

ofcytokine-primed, neutrophil-dependent vasculitisto occur

(8). Given theevidence thatVLA-4mediates homotypic

ag-gregationoflymphocytes,andadhesion oflymphocytes to

en-dothelialcellsvia VCAM-1 moleculesaswellas tothe

alterna-tively spliced domain of fibronectin (21, 39, 40),the upregu-latedfunction ofVLA-4inaddition to that of LFA- 1 could be themechanismsresponsible forthedevelopment ofthe

charac-teristicpathological features of vasculitisinSLE.Furthermore,

as suggested in the RAsynovialTlymphocytes(13, 29, 41 ), VLA-4/VCAM- 1 and LFA- 1 /ICAM- 1 pathways not only

me-diateadhesion ofleukocyte-leukocyteand leukocyte-endothe-lium, butbydelivering signals intothe cells toproliferateby

interacting with their ligands could be important mechanisms for the infiltration ofthe inflammatory cells into the vessel

walls andinterstitium, and the maintenance of these cells to perpetuate the vascularinflammation.

The presentstudyprovidesthefirst evidencethat

expres-sion and function ofVLA-4antigenwas upregulated in SLE

patients with vasculitis,introducingapotential role ofVLA-4

antigen in the pathogenesis of vasculitis in SLE. At present,

however, it isnotclearwhether theupregulated expressionof integrins isthe consequenceofthevasculitis events oris

di-rectlyinvolvedin thepathogenesis.Toaddresstheissue, itmay

be useful to examine the effect of anti-VLA-4

antibody

in model animals forvasculitis. Ithas beenreportedthat induc-tion of

adjuvant

arthritis in Lewis ratscould be

successfully

inhibited by anti-ICAM-1 antibody (42)andratcardiac

allo-graft survivalwasextendedbyadministration of anti-ICAM-1

and/oranti-LFA-1 antibodyinheterotropicratcardiac

trans-plantation (43). If it isthe case in modelanimals for vasculitis by anti-VLA-4

antibody,

the intervention with the VLA-4

pathwayinadditionto

LFA-l

/ICAM-l

interactionmay

con-tributetodevelopinga newandideal

therapeutic

strategyfor

theintractabilityandsometimes

fatality

ofSLE causedby

vas-culitis.

Acknowledaments

We would like to thank Drs. Martin E. Hemler and Takami

Matsuyamafor valuable discussions.

This study was_{supported in part by} a

grant-in-aid

for scientific research (C) from the

Ministry

of

Education,

Science and

Culture,

Japan.

References

1.Tan,E. M. 1989.Diagnosticmarkers forautoimmune disease andprobes

for cell_biology.Adv.Immunol.14:93-151.

2._Ebling,F.M.,and B. H.Hahn. 1989.Pathogenicsubsets of antibodies to

DNA.Int. Rev. Immunol.5:79-95.

3. Tan, E. M. 1991. Autoantibodies in pathologyand cell biology. Cell.

67:841-842.

4.Dixon,F.J.,J. D.Feldman,andJ. J.Vazquez.1961.Experimental glomer-ulonephritis:thepathogenesisof alaboratorymodelresemblingthespectrumof

humanglomerulonephritis. J.Exp.Med. 113:899-920.

5.Winfield,J. B.,I.Faiferman,and D.Koffler. 1977.Avidityof anti-DNA antibodies in serum and_{IgG glomerular}eluatesfrompatientswith SLE.J. Clin.

Invest. 59:90-96.

6._Savage,C.0. S. 1991._Pathogenesisofsystemicvasculitis. InSystemic

Vasculitides.A.Churg,J.Churg,andG. G.Hunder,editors.Igaku Shoin,New

York. 7-30.

7.Philips,M.R.,S. B.Abramson,and G. Weissman.1989.Neutrophil

adhe-sionand autoimmunevascularinjury.Clin.Asp. Autoimmunity.3:6-15.

8. _Argenbright,L. W.,and R.W. Barton. 1992.Interactionsofleukocyte integrinswith intercellularadhesionmolecule1intheproductionof

inflamma-toryvascular_injuryinvivo. J.Clin. Invest. 89:259-272.

9._Springer,T.A. 1990.Adhesionreceptorsoftheimmunesystem.Nature (Lond).346:425-434.

10.Hynes,R.0.1992._{Integrins:versatility,}modulation,andsignalingin cell adhesion. Cell. 69:11-25.

11. Shimizu, Y.,G. A. VanSeventer, K.J. _Horgan, and S. Shaw. 1990.

Regulatedexpressionand_bindingof threeVLA(B 1)integrin receptorson T cells. Nature(Lond.).345:250-253.

12._Cybulsky,M.I.,and J. Gimbrone M. A. 1990. Endothelial_expressionof a mononuclear_leukocyteadhesion molecule_{during atherogenesis.}Science(Wash. DC).251:788-791.

13.Laffon, A.,V. R.Garcia,A.Humbria,A.A._Postigo,A. L.Corbi,L. M.de,

and M. F. Sanchez. 1991._Upregulatedexpressionand functionof VLA-4 fibro-nectinreceptorson human activated Tcells in rheumatoid arthritis. J.Clin. Invest.88:546-52.

14.Takeuchi, T.,and T. Abe. 1993.Possible role of VLA-4antigenin

vasculi-tisof thepatientswith_{systemic lupuserythematosus (SLE).}InVasculitis

Syn-dromes.T.Tanaber,editor.Hokkaido_UniversityPress,Sapporo. 187-196. 15.Tan,E.M.,A.S.Cohen,J. F.Fries,A. T.Masi,D. J.McShane,N. F.

Rothfield,J.G.Schaller,N.Talal,and R. J. Winchester.1982. The 1982 revised

criteria forthe classificationofsystemic lupus erythematosus.Arthritis Rheum.

25: 127 1-1277.

16.Takeuchi, T.,S.Tanaka,A. D.Steinberg,T.Matsuyama,J.Daley,S. F.

Schlossman,and C.Morimoto. 1988. Defective_expressionof the 2H4 molecule afterautologousmixedlymphocytereactionactivation insystemic lupus

erythe-matosus_patients.J.Clin. Invest.82:1288-1294.

17. Takeuchi, T.,C. E. Rudd,S. F.Schlossman,and C. Morimoto. 1987.

Induction of_{suppressionfollowing autologous}mixedlymphocytereaction: role

of a novel2H4antigen.Eur.J. Immunol. 17:97-103.

18. Sanders, M. E., M. W.Makgoba, S.0. _Sharrow,D. Stephany,T. A. Springer, H. A. _Young,and S.Shaw. 1988. Human memoryTlymphocytes

expressincreasedlevels of three cell adhesionmolecules(LFA-3, CD2,and

LFA-I)and three othermolecules(UCHL-1,CDw29,andPgp- I )andhave enhanced IFN-'y production. J.Immunol. 140:1401-1407.

19. Pober, J.S., and R.S. Cotran. 1990. The role ofendothelialcells in inflammation._{Transplantation (Baltimore).}50:537-544.

20.Carlos,T._M.,R.B.Schwartz,N. L.Kovach,E.Yee,M.Rosso, L. Osborn,

G.Chi-Rosso, B. Newman, R. Lobb,and J. M. Harlan. 1990. Vascular cell

adhesionmolecule-Imediateslymphocyteadherence tocytokine-activated

cul-turedhumanendothelialcells.Blood.76:965-970.

21.Hemler,M.E.,M.J.Elices,C.Parker,andS.Takada.1990.Structure of

the_integrinVLA-4anditscell-cellandcell-matrixadhesion.Immunol. Rev.

114:45-65.

22.Pulido,R.,M.J.Elices,M. R.Campanero,L.Osborn, S. Schiffer,P. A. Garcia,R.Lobb,M.E.Hemler,and M. F.Sanchez.1991.Functional evidence

forthreedistinctandindependentlyinhibitableadhesionactivities mediatedby

the human integrin VLA-4. Correlation with distinctalpha4epitopes.J.Biol.

Chem. 266:10241-10245.

23.Morimoto,C.,A.D.Steinberg,N. L.Letvin,M.Hagan,T.Takeuchi,J.

Daley,H.Levin,and S. F.Schlossman. 1987.Adefect ofimmunoregulatory T

cell subsets insystemic lupus erythematosuspatients demonstrated with anti-2H4antibody.J. Clin.Invest. 79:762-768.

24.Smith,S.H.,M. H.Brown,D.Rowe,R. E.Callard, andP.C.Beverley.

1986. Functional subsetsof human helper-inducer cells defined byanew mono-clonalantibody, UCHL-1.Immunology.63:58-62.

25. Thomas,M.L., andL. Lefrancois.1988.Differentialexpression of the leukocyte-commonantigen family.Immun. Today. 9:320-326.

26. Akbar, A. N., L. Terry, A.Timms,P.C.L. Beverley,and G. Janossy.1988. Lossof CD45R andgain of UCHL-Ireactivity isafeature of promedTcells. J.

Immunol. 140:2171-2178.

27. Serra,H.M.,J. F. Krowka, J. A. Ledbetter, and L. M. Pilarsky. 1988. Loss ofCD45R(Lp220)represents apost-thymicT celldifferentiationevent.J.

Im-munol. 140:1435-1441.

28.Sanders,M.E., M. W. Makgoba, and S. Shaw. 1988.Human naive and

memoryT cells: reinterpretationof helper-inducer and suppressor-inducer

sub-sets.Immunol. Today. 9:195-199.

29. Garcia-Vicuna,R.,A.Humbria,A. A.Postigo,C.Lopez-Elzaurdia,M. 0. DeLandazuri,F.Sanchez-Madrid,and A.Laffon.1992. VLAfamilyin

rheuma-toid arthritis: evidence for in vivoregulated adhesion ofsynovialfluid T cellsto

fibronectinthroughVLA-5integrin. Clin. Exp.Immunol.88:435-441.

30. Kohn, F. R., M. E. Grigg, and H.-G. Kingemann. 1991. Differential regulation of fibronectin receptorsubunitgene and cell surfaceexpression in humanperipheral bloodTlymphocytes. J.Immunol. 146:1484-1489.

31.Birkenmeier, T. M., J. J. McQuillan, E. D. Boedeker, W. S.Argraves,E.

Ruoslahti, andD. C. Dean. 1991. TheaSbl fibronectin receptor.J.Biol. Chem. 266:20544-20549.

32. Rosen, G. D.,T. M.Birkenmeier, andD. C. Dean. 1991. Characterization of thea4 integrin gene promoter. Proc. Natl. Acad. Sci. USA. 88:4094-4098.

33.Shimizu, Y.,W.Newman,T. V.Gopal,K. J.Horgan,N.Graber,L.D. Beall, G. A. van Seventer, and S. Shaw. 1991. Four molecularpathwaysof T cell

adhesion to endothelial cells: roles of LFA-1, VCAM-1, and ELAM- 1 and

changesinpathway hierarchy under different activation conditions.J.Cell Biol. 113:1203-1212.

34. Picker,L.J.,T. K.Kishimoto,C. W.Smith, R. A. Warnock, and E. C. Butcher. 1991. ELAM- Iisanadhesion molecule forskin-homingT cells. Nature

(Lond.). 349:796-799.

35.Aruffo,A., 1.Stamenkovic,M.Melnick,C. B.Underhill,and B. Seed.

1990. CD44 is theprincipalsurface receptor forhyaluronate.Cell. 61:1303-1313. 36.Oppenheimer-Marks, N.,L. S.Davis,D. T.Bogue,J.Ramberg,and P. E.

Lipsky. 1991.Differential utilizationofICAM-I and VCAM- Iduringthe

adhe-sion andtransendothelial migration ofhuman T lymphocytes. J. Immunol.

147:2913-2921.

37. Parker, C.M.,C.P.Pujades,M. B. Brenner, and M. E. Hemler. 1993. a4180, a novelform oftheintegrina4subunit.J.Biol. Chem. 268:7028-7035.

38.Fauci,A. S. 1987. The vasculiticsyndromes.InPrinciplesof Internal

Medicine.E.Braunwald,K.Isselbacher,R.Petersdorf,J.Wilson, J. Martin,and A.S.Fauci, editors. McGraw-Hill,NewYork. 1438-1445.

39.Bednarczyk,J.L.,and B. W.McIntyre. 1990.Amonoclonalantibodyto VLA-4a-chain (CDw49d) induceshomotypic lymphocyte aggregation. J.

Im-munol. 144:777-784.

40. Elices, M.J.,L.Osborn, Y.Takada,C. Crouse,S.Luhowskyj, M. E.

Hemler,and R. R. Lobb.1990. VCAM-I onactivated endothelium interactswith

theleukocyteintegrinVLA-4 at a site distinct from theVLA-4/fibronectin

bind-ingsite.Cell.60:577-84.

41.Postigo,A.A., R.Garcia-Vicuna,F.Diaz-Gonzalez,A.G. Arroyo, M.0.

DeLandazuri, G. Chi-Rosso,R. R.Lobb,A.Laffon, andF.Sanchez-Madrid.

1992.Increasedbinding of synovialTlymphocytes from rheumatoid arthritisto

endothelial-leukocyte adhesionmolecule- I(ELAM-1) and vascularcell adhesion

molecule-l (VCAM-I). J. Clin.Invest.89:1445-1452.

42.Iigo,Y., T.Takashi,and T.Tamatani.1991. ICAM- 1dependentpathway

iscritically involvedin thepathogenesis of adjuvant arthritis inrats.J. Immunol. 147:4167-4171.

43. Isobe, M.,H. Yagita,and K. Okumura. 1992.Specificacceptanceof cardiacallograft aftertreatmentwithantibodiestoICAM-1andLFA- 1. Science (Wash.DC). 255:1125-1127.