Enkephalins stimulate leukemia cell migration and surface expression of CD9

Enkephalins stimulate leukemia cell migration

and surface expression of CD9.

W Heagy, … , K Duca, R W Finberg

J Clin Invest.

1995;

96(3)

:1366-1374.

https://doi.org/10.1172/JCI118171

.

Opioid peptides have been implicated in the regulation of tumor growth and biology;

however, little attention has been given to the mechanisms that are involved. In this study

we show that physiological concentrations of the endogenous opioid neuropeptide

methionine-enkephalin (MET-ENK) and the synthetic enkephalins D-Ala2, Me-Phe4,

Gly(ol)5 and D-Ala2, D-Leu5 are stimulants for the in vitro migration of pre-B acute

lymphoblastoid leukemia (ALL) cells. Activation of the human pre-B ALL cell lines NALM 6

and LAZ 221 with MET-ENK resulted in both an increase in their migration and an

augmentation in the surface expression of the leukemia cell marker CD9. The opiate

receptor antagonist naloxone reversed these enkephalin-induced effects on the leukemia

cells. When the pre-B ALL cells were preincubated with an anti-CD9 mAb before challenge

with MET-ENK their migration to the enkephalin was markedly reduced. These studies

show that endogenous and synthetic opioid peptides are stimulants for pre-B ALL cell

migration and suggest that CD9 is important in the regulation of leukemia cell motility.

Research Article

Find the latest version:

Enkephalins Stimulate Leukemia Cell Migration and Surface Expression of CD9

WyrtaHeagy,*§Karen Duca,* and Robert W.

Finberg*"

*Laboratory of Infectious Diseases, Dana-Farber Cancer Instituteand Departments ofMedicine and§Pathology, Harvard Medical

School,Boston, Massachusetts 02115

Abstract

Opioid peptides have been implicated in the regulation of

tumorgrowthandbiology; however,littleattentionhas been

giventothemechanismsthatareinvolved.Inthis studywe

show that physiological concentrations ofthe endogenous opioid neuropeptide methionine-enkephalin (MET-ENK)

and thesynthetic enkephalins

D-Ala2,Me-Phe4,Gly(ol)'

and

D-Ala2,D-Leu5 are stimulants forthe in vitro migration of

pre-B acute lymphoblastoid leukemia (ALL) cells.

Activa-tion of thehumanpre-B ALL celllines NALM6 and LAZ 221 with MET-ENK resulted in both an increasein their migrationand anaugmentation inthe surfaceexpressionof the leukemia cell markerCD9.Theopiatereceptor antago-nist naloxone reversed these enkephalin-induced effects on theleukemiacells. When thepre-BALL cellswere

preincu-bated with ananti-CD9 mAb before challenge with MET-ENK their migration to the enkephalin was markedly

re-duced. These studies show thatendogenous and synthetic opioidpeptidesarestimulants forpre-BALLcellmigration

andsuggest thatCD9isimportantintheregulation of

leuke-miacell

motility.

(J. Clin. Invest. 1995.96:1366-1374.) Key

words:enkephalins-opioidpeptides*cellmovement*

pre-Bcell * leukemialeukocyte-adhesion receptors

Introduction

First identified in brain tissues for theiranalgesic activity the

opioidfamilyofneuropeptidesare nowrecognizedas pleiotro-pichormones that modulate vital processes in theperipheryas

well as the CNS. These neuropeptides are important in the

regulationof moods,behaviors, higher cognitiveprocesses,

re-production,paincontrol,inflammatoryresponses, andavariety

of immunefunctions (1-5).Inhumans the intravenoususeof

opioid alkaloidshas been associated withanincreased riskfor infectious and oncogenic diseases (6). Evidence obtained in animal models has documented that use ofopioid peptides or

alkaloids diminishes resistance to infection (6-8). Studies

showing that stress stimulates the endogenous synthesis and releaseofopioid peptideshasfueledspeculationthat thestress induced-release of theseagents isimportantindisease processes

AddresscorrespondencetoWyrta Heagy,Endocrinology Office, Medi-cal Specialty Center-D3, Minneapolis Medical Research Foundation, 914 South Eighth Street, Minneapolis, MN 55404. Phone:

612-347-7629;FAX: 612-337-7372.

Receivedfor publication17February1994 andacceptedin revised

form 1June 1995.

( 1, 3). Within the past decade many reports have detailed the in vitroeffects of opioid peptides on human and rodent immuno-cytes;however, the precisemechanism(s) by which these neu-ropeptides influence healthordisease have not been defined.

Thebeststudiedactivity of opioid peptides is their charac-teristic effectonpain. This morphomimetic activity is dependent

ontheNH2 terminal sequence (Tyr-Gly-Gly-Phe-Met or Leu) of the analgesic peptides and is blocked by classical opiate receptorantagonists such as naloxone (4). Opioid peptides in-teract with a number of distinct opiate/opioid receptor types (e.g.,6,

_IL,

K) andsubtypes (e.g.,

ktl,

A2,

61, 61, K1, K2) that areexpressed in the CNS andperiphery (3, 4, 9, 10). Peripheral bloodleukocytes have been reported to express both classical (naloxone-sensitive) and nonclassical (naloxone-insensitive)

opiate/opioidrecognition sites (3, 9).

Opioid peptidesarepotentsignalsthatdirect leukocyte

mo-tility andoneof their key functions in the periphery may be in leukocyte recruitment and migration ( 11). Studies in rodent and large animal modelshave shown that the opioid pentapeptide methionine-enkephalin

(MET-ENK)'

enhances lymphocyte

trafficking(7, 11-13), stimulates transendothelial migrations, andcausestheredistribution ofcirculatingleukocytes ( 14, 15 ). Insheep the infusion of MET-ENK into the afferent lymphatics ofpopliteal nodes caused lymphocyte migrations from the node into the efferent lymph (16). When MET-ENK was injected into the cerebral ventricles of rodents both lymphocytes and macrophages were recruited to the injection site (14). Other studies carriedout in vitro haveconfirmed that MET-ENK is chemotactic for isolated humanperipheralblood T cells (11), monocytes (13, 14), andpolymorphonuclear leukocytes (17).

Opioid peptides, like other leukocyte chemoattractants, have been shown to trigger a number of biochemical and cellular

eventsincluding transmembranesignaling(18), cellflattening,

membrane spreading, and pseudopod extension (19). Despite this impressive number of studies documenting the effects of MET-ENKonleukocytemotilitythere islittle known

concern-ingthe molecular processes that regulate the enkephalin-medi-ated effects.

MET-ENKand otheropioid peptideshave been reportedto

bindtocertain kinds oftumorcells andtohaveantiproliferative

activity (20). Itisknown thatmanygrowth-modulating factors also regulate cellmovement (21-24); however, the effect of

opioid peptidesonthemotilityof leukemias, transformed leuko-cytes orothertypetumors has notbeen addressed. To gaina

better understanding of the actions of opioidpeptides wehave studied the effectsofMET-ENKand otherenkephalinson two humanpre-B acutelymphoblastoid leukemia (ALL) celllines,

NALM 6 and LAZ 221, and the EBV-transfonned mature B

1.Abbreviations used in thispaper:ALL,acutelymphoblastoid

leuke-mia; DADLE, [D-Ala, 2D-Leu5]-enkephalin; DAMGO, [D-Ala,2Me-Phe,4Gly(ol)5] -enkephalin; DesTyr-LEU-ENK, H-Gly-Gly-Phe-Leu-OH; MET-ENK, methionine-enkephalin.

J. Clin. Invest.

© The AmericanSocietyfor Clinical Investigation,Inc. 0021-9738/95/09/1366/09 $2.00

cell line LAZ 388. We have used a migration assayand moni-tored cell morphology (i.e., cell flattening and membrane spreading) using flow cytometric analysisofforward and right anglelight scatter forassessing the effects ofMET-ENKand

other opiate receptor ligands. Previously, other investigators have shown that chemotactic factors modulate the expression and function ofcell surface adhesion molecules (e.g., the

/31

and /2integrins) (22-24). Since MET-ENK has been shown tobeastimulus for lymphocyte migration ( 11)wehave investi-gated the possibility that this enkephalin also modulates the

expression of adhesion receptors. For this purpose we used

immunofluorescent staining and flow cytometrytomeasure the effects of MET-ENK on the surface expression of a panel ofB cell determinants.

In this report we show that MET-ENK and synthetic

en-kephalin analogs serve as potent signals for the migration of NALM 6 and LAZ 221 pre-B ALL cells. Exposure of these

leukemia cells to MET-ENK results inadiscrete and transient increase in CD9 surface expression aswell as an enhancement

inmigration. Preincubation of the leukemia cells with anti-CD9 mAb before challenge with MET-ENK blocks their migration to the enkephalin. These findings suggest that leukemia cell

migrationto MET-ENK is regulated, atleast inpart, via CD9 expression.Inprevious reports CD9 has been implicatedin the formation ofpre-B ALLcellhomotypicaggregations (25, 26)

and in thespontaneous locomotion ofsomekinds oftumor cells

(27, 28). Ourfindings provide the firstevidence thatCD9 is involved in the migration of lymphoid cells. These studies pro-vide new insights into the actions of opioid peptides and into

mechanisms that regulate lymphoid cell motility.

Methods

Cells. Thepre-BALLcell lines NALM6and LAZ 221(CD9', CDIO,

slg-)

and the EBV-transfonned B lymphoblastoidcell lineLAZ 388 (CDl9 +, CD20 +, CD9 -', CD10-) havebeen describedpreviously(29). The LAZ221 and LAZ 388lines were established from donorcells

harvested from apatientatthe Dana-FarberCancer Institute, Boston,

MA (29) and were obtained forthese studies fromDr. Jerome Ritz.

Cells were passaged (1-2 x 105/ml) every 2-4 d in RPMI 1640

medium (BioWhittaker, Inc., Walkersville, MD) supplemented with 2 mM glutamine, a solution of 100 U/ml-100 pg/ml of penicillin-streptomycin (Gibco Laboratories,GrandIsland, NY),and10% bovine

serum(Hyclone Laboratories, Inc., Logan, UT). Cultureswere main-tainedat37°Cina5%CO2 tissue culture incubator.

Reagents and antibodies. The naturally occurring pentapeptide

MET-ENK (Tyr-Gly-Gly-Phe-Met), and the synthetic peptides [D-Ala

2,D-Leu5]

-enkephalin (DADLE), [D-Ala2,Me-Phe

4,Gly

(ol)s]

-en-kephalin (DAMGO), H-Gly-Gly-Phe-Leu-OH (DesTyr-LEU-ENK),

and FMLP werepurchased fromSigma ChemicalCo.(St. Louis, MO).

Naloxone(Narcan)wasobtainedfromDuPontPharmaceuticals

(Ma-nati,PuertoRico).

Anti-CD9 and-CD10mAboftheIgG1 isotypewerefromBiodesign International(Kennebunkport,ME) and those of theIgG2aisotypewere

obtained from Amac, Inc. (Westbrook, ME). Anti-CD20 (IgGI), -CD14 (IgG2a), -CD29 (IgGI) and -CD54 (IgGI) mAb were from

Biodesign International. Anti-CD72 (IgG2a), and -CD23(IgGl) mAb

werefrom TheBinding Site,Inc.,(Birmingham,UnitedKingdom)and anti-CD22 (IgGI) was from Amac Inc., Anti-CD2 (IgGI), -CD3

(IgG2a), -CD1la (IgGl), -CD11b (IgGl), -CD18 (IgGI), -CD45

(IgG2a),-CD58(IgGI),-ClassI(HLA-A,B,C;IgG2a), and -Class II

(HLA-DR;IgG2b)wereprepared as ascites developed in Balb/c mice. Hybridomacellswereobtained forthe anti-Class II mAb(LB3.1 )from

Dr.Peter Knudsen and Dr. JackStrominger,Dana-Farber Cancer

Insti-tute, Boston, MA and for anti-CD2 (HB195), -CD3 (CRL 8001),

-CD1la (HB 202), -CDllb (HB204), -CD18 (HB203), -CD45

(HB196),-CD58 (HB205)and -Class I(HB 95)from the American Type Culture Collection (Rockville, MD). Anti-CD19 mAb (IgGl)

was agift from Dr. Ken Anderson, Dana-Farber Cancer Institute,

Bos-ton, MA. The mAbrecognizing VLA's 1-6 were from the V Human

Leukocyte Typing Workshop (subpanel 6)and the DE9Nanti-VLA /3

chain mAb(IgGl) wasprovided by Dr. JeffBergelson, Dana-Farber CancerInstitute, Boston, MA(30). The MOPC 21 myeloma protein (IgGI)that was used as a control waspurchasedfromOrganonTeknika

Cappel (Durham, NC).

Migration assays. Cellmigrationwasmeasuredusing

microchemo-taxischambers(NeuroProbe, Inc.,CabinJohn, MD)aspreviously de-scribed (11). Cells(5 x 106/ml)wereseparatedfrom peptides inthe

lower wells ofthechamber bynitrocellulose filters (Sartorious, 5

_gm

pore; Neuroprobe, Inc.). In some studies the cellswerepreincubated with lo-5 Mnaloxone for 15 hat37TCoranti-CD9,-CD1O,or-VLA 4mAb for 30 min at 4°C and thenincubated inthemicrochemotaxis chambers for 75-90min at 37°C. The nitrocellulose filters were fixed and stained withCongoreddye (SigmaChemical Co.). Cells withinthe

filterswereidentified byfluorescence microscopy andthenenumerated

utilizinganopticalimage analyzer (Optomax Vimageanalyzing sys-tem;Analytical Instruments, Hollis, NH).In somestudiesthe cells were

identified visually andphotographed with the use of a fluorescence microscope.

Forwardandrightangle lightscatteranalysis of the morphological responseto MET-ENK Theeffectsof MET-ENK on cellmorphology

were measured using flow cytometric analysis of forward and right angle lightscatter aspreviously described (31). For these studies 106

cellswereincubatedfor5 min at37°Cin a 1-mlvolof PBS,alone or

PBS containing MET-ENK. Some samples werepreincubated for 15

min oniceinPBScontainingi0-5Mnaloxone. Sampleswereanalyzed

using aFACScan' flowcytometer withexcitation at488nm (Becton Dickinson andCo., Mountain View, CA) (31). Histograms were

ob-tained from the analysisof 10,000cells and datawereanalyzedusing the LysysSystem software package (Becton Dickinson and Co.).

Fluo-rescence wasmeasuredasthearithmetic/linearmeanrelative

fluores-cence intensity; the range of values was between 1 and 10,000

(100_104).

Immunofluorescence staining and flow cytometric analysis ofcell

surface molecules. Cellswereresuspendedat106/mlin PBS and then

incubatedfor periods of 5-120min at37°C withMET-ENK(0-10-5 M, finalconcentration).Incubationswereterminated bytransferring the samplestoanice bath and then pelleting the cells by centrifugation at

1,800rpmfor5 min at4°C.

Immunofluorescence staining wascarried out in4-mltubes

main-tained inan ice waterbath. Cells (106) in 100-ILI vol of PBS were

incubated with the primary antibody for 30 min, washed, and then stainedbyasecondincubation with affinity isolated,human absorbed,

FITCconjugatedgoatF(ab')2 anti-mouse Ig (Tago Inc., Burlingame, CA). Cellswerefixedin a 2% solution offormalinandmaintainedin

thedark at10°C until analyzed.

Flowcytometric analysis wasperformed usingaFACScanD Flow Cytometer (Becton DickinsonandCo.) equippedwithaHP9000Series computer. Datawerecollectedashistograms depictingthefluorescence intensity for 5,000-10,000 cells per sample; arithmatic/linear mean

relativefluorescence intensitieswereobtainedusingtheLysysSystem

softwareanalysis package.Arithmatic/linear values(range, 1-10,000)

wereused for allsampletosample comparisons.Thenonbinding IgGl

MOPC myelomaproteinaswellasisotype matched nonreactivemAb

wereusedascontrols formeasuringtheautofluorescenceof B cells.

Results

Methionine-enkephalin stimulates analoxone-sensitive alter-ation in the morphology of NALM 6 cells. An initial set of

B

P

_Figure

1.

Methionine-enkephalin

10106 D_

C _A

C

inducesamorphologicalchange

C _{in NALM6 cells. NALM 6 cells}

110.4 \ 8510/ (106/ml)werepreincubated in

I 3 , /§ PBS alone\ or_{containing 10-5}M

IL 102 \ i; " / naloxone for

15

min

onice,

stimu-, lated with the indicated

concentra-.2 / n X /tions of MET-ENK, and then

in-is.

E

4 cubated for anadditional5

min

at

_ioo

37TC.

Analysis offorward and

s with_a L

_{._}

_. r

right angle lightscatterwas

per-g

/ _{formedusing a Becton Dickinson}

}L } @ X, /

_{~~~~~~~~~~~~~~~~~~~~FACScan't}

flow cytometeras de-=art#, , . . . _ " . " . . . scribedin Methods.Histogramsof

Q D0.01

O.;

1 O1 100 1000 0.0 0.01 0.1 I 10 100 1000

MET-ENK

(l0

iTENK(M) lightscatterwereobtainedby

ana-lyzing10,000 cells per sample. Data areexpressed as the arithmetic/linear meanfluorescence intensity (y-axis) versus concentration of MET-ENK (x-axis). (A) Right angle light

scatterfor samples preincubated in PBS alone(no-ni)orPBS withiO-5Mnaloxone(rn-rn).(B) forwardlightscatterfrom samples preincubated

inPBS alone(o0 o)orPBS with l0-5Mnaloxone(.*- ).

Bcells responded to thenaturally occurring opioid pentapeptide MET-ENK.Flowcytometric analysis offorward and right angle light scatter was usedformonitoringtheeffectsof MET-ENK

onthe morphology of NALM 6 cells. As indicated (Fig. 1),

treatment ofthe NALM 6 cells with MET-ENKprovoked a

rapid change in cellmorphology(i.e.,anincrease in right angle and forwardlight scatter). Physiologicaldoses(i.e., <0.1nM)

of MET-ENK initiated this morphological response since

en-kephalin doses as low as 0.01 nM elicited a near maximum

response (Fig. 1). These results provided evidence that the NALM 6 pre-B ALL cell linewas responsive toMET-ENK. Consistent with thespecificity expectedforaclassicalopiate/ opioid effect the morphological response to MET-ENK was

inhibited when the NALM 6 cells werepreincubated with the

opiatereceptorantagonistnaloxone(Fig. 1).Whereas the

appli-cation of MET-ENKtothe NALM 6 cells enhanced bothright angle and forwardlightscatter thepreincubation innaloxone,

alone caused a modest negative deflection in the baseline

(Fig. 1).

CD9 expression onpre-B acute lymphoblastoid leukemia cells is modulated by methionine-enkephalin. Our studies

showing that MET-ENK provoked physical changes in the NALM 6 cell membrane (Fig. 1) prompted us to investigate

thepossibilitythatthisenkephalinalsoinducedarearrangement in membrane proteins. Asindicated, (Figs. 2 and3 and Table I) apanelof mAbwasusedto assesstheeffects of MET-ENK

on the expression of B cell surface determinants; samples of NALM6,LAZ221, andLAZ388 cellswereincubatedwith and without the enkephalin and the expression of surface markers measured withthe useofimmunofluorescent staining and flow cytometry. The incubation of NALM6orLAZ 221 cells with MET-ENK resulted in atransient increase in theirexpression

of CD9(Figs. 2and3). Abrief incubation of the pre-BALL

cells with MET-ENK(i.e., c 10 min) was sufficienttocause theenhancement inCD9 expression (Fig. 3 A). Thisenhanced

levelof CD9 expression was sustained forperiods of 60 min butdeclinedto nearbaselineby 120 min(Fig.3A).The induc-tion of CD9 surface moleculesbyMET-ENK wasdose

depen-dent(Fig. 3 B). CD9expressionwasincreased,albeitmodestly,

by the lowMET-ENKdoseof 0.01 nM, further increased with the greater 100 nMdose, and still furtherincreased, with the maximum dose(10,000 nM)

(Fig.

3B). WhereasMET-ENK

causedanaugmentation in the CD9 molecules displayed on the surfaces ofpre-B ALL cells the enkephalin had no effect on their expression of CDIO (Figs. 2 and 3 A). These findings show that theMET-ENK-stimulated enhancement in CD9 ex-pression is specific and not solely the result of an enkephalin-mediated increase in cellsurfacearea.

When the pre-B ALLcells were incubated in amixture of MET-ENKand a100-foldexcessof naloxone the induction of CD9by MET-ENKwastotallyblockedindicatingthe specific-ity of the enkephalin response (Fig. 3 C).

Immunofluorescent staining for other adhesion molecules

0

._-E E

5

30

CD9

1 2 3 4 1

loglo

Fluorescence

CD10

2 3 4

Figure2.Methionine-enkephalincauses adiscrete,transient

up-regula-tion in theCD9 moleculesexpressed onthe surface of NALM 6 cells. NALM6cells(

106/ml)

wereincubatedat370Cwith MET-ENK(l0-7

M) for 0 (A and B)or30min(C and D). Cellswerethentransferred

toanicebath, pelletedbycentrifugation,andstainedasdescribed in Methods. Data are shownashistograms depictingthedistribution of 10,000cellslabeled with anti-CD9(Aand C)orCD-l0 (BandD)

i

3200 _A NALN6Ceb

2400l

i CDO0

< 200 I

0 20 40 60

0 100 120

Tim(Admnso)

16010

100

LAZ221Cofb

0.0`0.01

MET.E( (nm

io0oo

MET-EWK (nM)

Figure3. Methionine-enkephalincausesarapidanddose-dependent increase intheCD9but notCDIO molecules expressedonpre-B leukemic

cells. Data arepresentedasthe arithmetic/linearmeanrelativefluorescence intensities measuredfor 10,000 cells labeled with anti-CD9orCD1O

mAb. Cellsamples wereprepared and stained for immunofluorescent analysis asdescribed forFig. 2 andinMethods.InpanelA, NALM6 cells

wereincubated withMET-ENK(I0-7 M) for the times indicated. Inpanel B,LAZ221cells wereincubated withMET-ENK at the dosesindicated for5 min at370C. Inpanel C,LAZ221cellswereincubatedat370Cfor15minwithMET-ENK(10-7M)(.*-)orwith bothMET-ENK and

naloxone (10-5M)(o-o). The anti-CD9 andanti-CDIO mAb used for these experimentswereoftheIgG2a isotype.

Table I.Methionine-Enkephalin Effects onthe Surface ExpressionofLymphocyte Determinants

Fluorescence*

NALM6 LAZ 221 LAZ388 Antibody

Cell determinant isotype (-) (-) (+) (-) (+)

CD 9 IgGI 140.8 400.0 107.1 220.5 4.2 4.0

CD 10 IgGI 37.1 37.0 11.0 10.2 0.93 1.0

ClassII(HLA-DR) IgGI 14.0 14.8 41.9 43.0 29.0 30.0 ClassI(HLA-A,B,C) IgGI 225.9 227.0 156.7 157.4 29.9 31.2

CDla (LFAla) IgGI ND ND 1.0 1.1 12.4 11.4

CD1lb (Mac 1) IgGI 1.0 1.1 1.4 1.5 1.5 1.5

CD18(Integrin /3 chain) IgGI 1.1 1.4 1.0 1.1 11.9 10.1

CD54(ICAM-1) IgGI 9.3 10.5 1.5 1.5 18.7 16.6

CD44(homing receptor) IgGI 1.0 2.2 ND ND 79.0 84.6

CD19 IgGI ND ND 1.7 1.9 8.6 8.0

CD58 (LFA-3) IgGI 1.0 0.9 1.7 1.9 22.4 21.8

CD22 IgGI 2.3 3.1 ND ND 11.4 11.1

CD20 IgGI 1.5 1.3 1.0 1.1 1.7 1.7

CD72 IgG2a 1.6 2.8 ND ND 1.4 1.7

CD29

(VLA-,3

chain) IgGI 8.4 9.3 1.7 1.9 11.4 11.2

VLA-,/

chain;

DE9NII

IgGI 13.6 13.8 5.8 5.6 11.5 11.1

VLA-1 IgGI 1.1 0.6 ND ND 1.8 2.0

VLA-3 IgG 1.1 0.7 ND ND 1.0 1.0

VLA-5 IgGI 5.2 4.9 ND ND 1.0 1.0

CD49b(VLA-a2 chain) IgGI 1.0 0.9 ND ND 1.0 1.0

CD49f(VLA-a6chain) IgGI 1.8 1.3 ND ND ND ND

CD49d(VLA-a4 chain) IgGI 3.9 4.4 ND ND ND ND

CD23 IgGI 1.8 1.6 ND ND 68.2 70.7

CD45(T200) IgG2a 1.1 1.0 ND ND 27.4 27.3

*_{Cellswerelabeledforanalysisasdescribedfor Figs. 2 and 3 and in Methods. Fluorescence valuesweremeasured fromhistogramscollected for}

samplesof5,000or10,000cells. Dataareexpressedasratios:[(arithmetic-linear mean relative fluorescence intensity measured for indicatedmAb)]/ [(arithmetic-linearmeanrelative fluorescenceintensity measured for isotypematchedcontrol)]. *_{Cellswereincubated for 30minat37°Cin PBS}

alone (-)orPBScontaining10-7 M MET-ENK(+). I_{ND =notdone. 11Twoanti-VLA-f3I(CD29)mAb were usedforthesestudies;the data}

30- _{NALM 6 Cells 60 LAZ 221Cells Figure}4. NALM6 and LAZ221

cells

_migrate

tothe

_opioid

penta-50- Spontaneous locomotion peptidemethionine-enkephalin.

ILSz^T | ^ D

MET-ENK (0.01

nM) MET-ENKwas prepared in the

20 U MET-ENK

(100

nM)

assay bufferconsistingof

Dulbec-I0 i *0-*

*ET-EK

*1

*0

iM4

co's

modified PBS and 0.1%chick

eggalbumin and thenadded in

35-E 30- ul volumes to the lower wells of

I

******* thechemotaxis chamber. A

Sarto-10102

Ad**i*

,2 | a ^ | - ] If 204 s E rius(5-Impore) nitrocellulose

11

2-<filter

wasplacedbetweenthe up-i up-i E~EJ1IM M1 _{per and lower reservoirs and the}*****

_{** **}

101

cells(5 x

_106/ml)

werethen

EJ-dE~IE~EE~

__*-_ addedto the upper wells in60-II

0 0 portions of the assay buffer. After

10 20 30 40 50 60 10 20 30 40 50 60 a_90-minincubation at370Cthe

Distance

(gim)

Distance

_(gim)

filters were removed,fixed, and

stained with Congo red dye aspreviously described (11).Cells whichmigratedinto the filter were enumerated utilizingfluorescencemicroscopy and anOptimaxVimageanalyzingsystem. 10 highpowerfields (HPF)weremeasuredat10-pImdistancesthrough the filter.Data arepresented

as themeans±SDfor30 values(i.e., 10 readings X triplicate assays).Note: the scales for they-axisaredifferent inpanelsAand B.

including VLA 4(CD49d) or-5

(,6-1

integrins) andthe com-mon VLA

,61

chain (CD29)was equivalentonNALM 6cell samples incubated with (+)orwithout(-)MET-ENK(Table

I). With the pre-B ALL cells the fluorescence intensity

mea-sured for other p-1 integrins (VLA-1, VLA-3, CD49b, or

CD49f)orfor p-2integrins (CDlLa, CDllb, and CD18)was

near that oftheIgGI isotype control (i.e.,between 1 and 1.5 times the control value) regardless of whether the cells were

incubated with (+)orwithout(-) MET-ENK(Table I).

As expected, immunofluorescent staining of the LAZ 388 cellsrevealedaphenotype distinct from that of thepre-BALL lines andcharacteristicofmatureEBV-transformed B cells

(Ta-bleI).With the LAZ 388 cells the level of cell surfacestaining

for CD9 and the /3-1 and /-2integrinsaswellasother determi-nantswas equivalent for the MET-ENK treated (+)and

non-treated (-) samples(TableI).

Enkephalins stimulate analoxone-reversible enhancement inthe migrationofpre-Bacutelymphoblastoidleukemiacells.

MIET-ENK

was apotent stimulus for the

migration

of NALM 6 and LAZ 221 (pre-B ALL), butnot LAZ388 (EBV-trans-formed)cells (Fig. 4and TableII). The number ofNALM6

or LAZ221cells whichmigratedtoMET-ENKwasgreater than

that which responded to the assay buffer, alone (spontaneous locomotion) (P. <0.01) (Fig. 4 and Table II). The effect of MET-ENKonthepre-B ALL cellmigration was dose depen-dent(Fig.4). MET-ENKwas stimulatoryover afourlog dose range (0.01-100 nM); the low MET-ENK dose of 0.01 nM resulted in a modest response which was increased with the greater 1 nM dose and further increased with the maximum 100 nMenkephalin dose (Fig. 4).

Like MET-ENK the synthetic enkephalins DADLE and DAMGO enhanced the migration of pre-B ALL cells (P, 5 0.01); the potency of these synthetic enkephalins was near tothat measuredforthenaturally occurring pentapeptide MET-ENK(Table

II).

Unlike the enkephalins, the neutrophil chemo-tactic agent FMLP and the nonanalgesic peptide DesTyr-LEU-ENK had no effect onthe migration of pre-B ALL cells (P, NS - 0.05).

Themigration ofLAZ 388 cells toMET-ENK, DADLE, DAMGO, DesTyr-LEU-ENK,or

FMLP

wasequivalenttotheir spontaneous locomotion (response to assaybuffer, alone) (P, NS . 0.05)(Table

II).

Naloxone inhibited themigrationof NALM 6 cellsto

MET-ENK(Table III). Spontaneous locomotion resulting from

ran-Table II. ThePre-B AcuteLymphoblastoidLeukemiaCell Lines Nalm 6 andLaz221 MigratetoNatural andSynthetic Enkephalins*

Peptides(100nM) Spontaneous FMLP DesTyr-LEU-ENK

Cells locomotion (nonanalgesic) (nonanalgesic) MET-ENK DAMGO DADLE

NALM6 2190$ 3198 3085 6952 5070 6240

(P,NS; 2 0.05)1 (P.NS; 2 0.05) (P, 5 0.01) (P, 5 0.01) (P, 5 0.01)

LAZ 221 1886 2906 2010 6199 4602 4405

(P.NS; 20.05) (P, NS; 20.05) (P,50.01) (P, <0.01) (P,50.01)

LAZ 388 1938 1914 2244 1986 2043 2283

(P.NS;

(P.NS; 20.05) (P. NS; 20.05) (P. NS; 20.05) (P. NS; 20.05) (P.NS; 20.05) 2 0.05)

* Assays were as described forFig.4. $Migrationwasscoredasdescribed forFig.4. The numberspresentedarecellcountsobtainedby adding the valuesmeasuredat 10-ptm intervals throughthe filters. § Thestatisticalsignificanceof differencesbetween thetreatmentswasanalyzed bya

Table III. Naloxone Inhibits the Migration of NALM 6 Cells to Methionine-Enkephalin*

MEr-ENK MET-ENK

Spontaneouslocomotion (1nM) (100nM) Without Witht Without With Without With

Cellcounts 1,647 1,350 2,925 1,251 3,492 1,476

Pvalues"l (NS;20.05) (0.01) (0.01)

Percentresponse1 82 43 42

*_{NALM 6 cells were cultured in RPMI 1640 medium supplemented with 2%}

FBS, 100_Ag/mlofpenicillin/streptomycin,and 2 mMglutamineor in the culture

medium containing10-5Mnaloxone for 15 hat370C ina5% CO2 atmosphere. Viability ofthe cells incubatedwithorwithout naloxonewasequivalent(296%)

asjudged from vital staining withtrypanblue.NALM6cells treatedwithnaloxone were added to the upper wellsofthe migration chamber in assaybuffer that

contained naloxone(10-5M).Migrationwasallowed to proceed for 75 min at

37TC. tWithout=cells thatwereincubatedin assaybuffer alone;with=cells that were incubated in10-'M naloxone. 0_{Migrationwasscored as described}

for Fig. 4. The numbers presented are cell counts obtained by adding the values measured at10-psmintervals through the filters. 11 The statisticalsignificanceof

differencesbetween treatments was analyzed by a Student's t test. I_Percent

=[(Sum for cells incubatedwithnaloxone)/(sumfor cells incubatedwithout naloxone])X 100%.

dom movement wasequivalentfor cells incubated withor

with-outtheopiatereceptorantagonist(P. NS, 20.05) (TableHI).

Anti-CD9 monoclonalantibody blocksthe migrationof pre-Bacutelymphoblastoid leukemia cellsto

methionine-enkepha-lin. Themigration ofNALM6 cellstoMET-ENKwasinhibited

when the cells were preincubated with anti-CD9 (P = 0.02) butnotanti-CD10, anti-CD29, anti-Class IIorClass I mAb (P, NS 2 0.05)(Table IV). When the control cells (incubatedin assaybuffer, alone) orthosepretreatedwith anti-CD10,

anti-CD29, anti-Class II orClass I mAb were added tothe upper wells of themigration chamber the cells migrated toward MET-ENK in the lower chamber. On the other hand, most of the cells whichwerepretreated with the anti-CD9 mAb remained

onthe surface(i.e.,attheorigin)with fewmovingdown toward theenkephalin (Table IV). This effect was notdueto

differ-encesinantibody isotype,since the anti-CD9 mAb that blocked

migration and the anti-CD10 and -Class I mAb thatwere not

inhibitory wereofthe sameIgG2aisotype. Anti-CD9induced inhibition of cellmovementdidnotresultsolelyfromantibody

induced homotypic aggregates because anti-CD29 which like

anti-CD9 initiated pre-B cell homotypic aggregation (25)but hadnoeffectoncellmigration.

AsindicatedbythephotographsofFig. 5,cellspretreated

withanti-CD9wereimmobilizedatthesurface of the filter (Fig.

5).Thisimmobilizingeffectwas notduesolelytothe

antibody-induced clumping since many nonclumped, single cells also remained immobilizedonthetopside of thefilter(Fig. 5).

Discussion

Mostcancer-related deaths result from the dissemination and metastasis of tumor cells; therefore, it is important to identify thesignals that regulatethemigration of malignant cells. In this study factors which regulate the in vitro migration of leukemia cellswereidentified. The naturally occurring opioid pentapep-tide MET-ENK as well as the synthetic enkephalin analogs DADLE and DAMGO were shown to be stimulants for the

Table IV. Migration of NALM 6 and LAZ 221 Cells Is Inhibited

byPreincubation with Anti-CD9 MonoclonalAntibody*

Control Monoclonalantibody MET-ENK

(nM) (Number of cells at 40 KM)

(Buffer, anti-CD10 anti-CD29 anti-CD9

alone) IgG2a IgGI IgG2a Experiment 1

NALM6 cells 0.0 135 160 175 94

0.01 238 354 406 92

1.0 407 510 629 77 100 637 702 750 60

(Buffer, Class II ClassI anti-CD9

alone) IgG2b IgG2a IgG2a Experiment2

NALM 6 cells 0.0 200 173 183 171 100 463 367 448 103

LAZ222cells 0.0 74 72 65 60

100 280 231 324 92

* Cells

(107/ml)

werepreincubatedin themigration assay bufferor the

buffercontaining anti-CD10, -CD29(DE9N),or-Class II(LB 3.1), ClassI(W6/32)or-CD9mAbfor 30min at4°C,thenpelleted by centrifugation and washed twice in assay buffer before their additionto

the upper wells of the chemotaxis chamber. The mAb doses used for thesestudies(1/100 dilution of ascites)weresaturatingaspredetermined by immunofluorescent staining and flow cytometry. tAssayswere as

describedfor Fig. 4. Cells which migrated 40

_jim

from theoriginwere

identified and scoredvisually with theuseofafluorescence microscope. The numberspresentedarethe cellcountsdeterminedfor 10 high power fields X triplicatewells.

migration of NALM 6 and LAZ 221 pre-B ALL cells (Fig. 4 and TableII). Inaprevious study( 11)weshowed thatopioid

peptides werepotent signals for themigrationof normal

lym-phocytescausing both an enhancement in their directional

(che-motaxis) and random movement (chemokinesis). We found,

as previously reported for the normal lymphocytes ( 11), that activation of the pre-B ALL cells with MET-ENK caused an

increase in both their nondirectional and directional migration (data not shown). These findings identify opioid peptides as

signalsfor the in vitro migration of cultured human leukemia cells.

The pre-B ALL cells respondedtophysiological doses of MET-ENK(Figs. 1 and 4).Circulatingand local tissue levels of the enkephalinshave been difficulttoquantitatebecause of their rapid hydrolysis (4); however, in brain these peptides

have been shown to reach concentrations of at least 0.1 nM

(32).The levelmeasuredin brainsamples (0.1 nM)is 10 times the MET-ENK dose (0.01 nM) that induced near maximum activation of the morphological response (Fig. 1) and that

caused significant increases in the pre-B ALL cell migration

(Fig.

4).

It is ofnote that the LAZ 221 and NALM 6 pre-B ALL

cells express enkephalinase

(CD1O;

neutralendopeptidase)on

their surfaces (seeFig. 2). This enzyme hydrolyzes and

inacti-vatesMET-ENK(33) and may be important indown-regulating

responses to MET-ENK by either decreasing the enkephalin

Figure5. Anti-CD9 but notanti-CD1Omonoclonal antibody blocks the migration of NALM 6 cells to methionine-enkephalin. NALM 6 cells (107/ml)werepreincubated inthemigrationassaybuffer containinganti-CD9 oranti-CD1OmAb for 30min at40C.Migration assays were as

describedforFig. 4 andTable II. Cellswereidentified for photographywith the use of afluorescence microscope (x20objective). The cells showninpanel A (at theorigin, topsideof filter) and panel C (at

_60-Am

depth)depict cells preincubated with anti-CD9 mAb. The cells of panel B (atthe origin) and panelD(at60-pmdepth) were preincubated withanti-CDlOmAb.

221 cells to MET-ENK was dose dependent; whereas the 0.01 nMdosestimulatedamodest increase inmigration,the response was increased when the dose was titered to 10 nM, and still further increased with the maximum enkephalin dose of 100 nM(Fig. 4). This dose dependency may have resulted, at least in part, fromendogenous enkephalinaseactivityand MET-ENK hydrolysis.

In our studies we measured the effects of MET-ENK on

cellmorphology using flow cytometric analysis of forward and right angle lightscatter (Fig. 1). Inaprevious report (19) the effects of opioid peptides and other chemotactic factors on the shape of granulocytes were monitored by light and electron microscopy. The opioid peptides, like other chemotactic factors, were seen to activate a morphological response before the onset of cell migration (19). This response was characterized by membrane spreading as well as cell elongation and flattening. Our studies showed that only abrief exposure (- 5 min) of thepre-BALLcellstolowdoses ofMET-ENK(0.01 nM) was sufficient to cause themorphological response (i.e., change in forward andright angle light scatter) (Fig.1).Theseenkephalin

effects were probably related to the spreading and flattening that cells undergo in preparation for movement. The opiate

receptor antagonist naloxone inhibited the

enkephalin-stimu-lated reshaping of the pre-B ALL cells (i.e., the increases in light scatter) (Fig. 1).However, when applied alone, the antag-onist caused a modest deflection in the baseline light scatter (Fig. 1). The antagonist was not directly cytocidal because the viability that we measured (i.e., trypan blue exclusion) was equivalent (2 96%) for naloxone-treated and nontreated NALM6cells. Other investigators have reported that naloxone has growth inhibitory effects on some tumor cells (20, 34) and the naloxone effect on lightscatter that we observed may be relatedtosuch antiproliferative activity.

MET-ENK was a potent stimulant for the migration of

NALM6 and LAZ 221 pre-B ALL cells butnotfor the

EBV-transformed,mature LAZ388 cell line(Table II).The respon-siveLAZ 221 andunresponsive LAZ388cells havethe identi-calgenetic background since theywerederived fromonedonor

(33); however, as expected these cell lines werefound to be dissimilar with respecttophenotype(TableI).Theresponsive

LAZ221 cellsexhibited the characteristic phenotype of pre-B

ALLcells (e.g., CD9 +, CD10 +,CD18-) whereas the unrespon-sive LAZ 388 cells displayed the surface markers (e.g.,

CDlla+, CD18+, CD44+) of an EBV-transformed cell line (TableI).Otherinvestigators (9) have been unableto

receptors onEBV-transformed lymphocytes and it is likely that the LAZ 388cells are unresponsive to MET-ENK simply be-causeofadeficiency in enkephalin binding sites. It is known

that EBV transformation renders cells unresponsive to normal activation signals (35) and it is possible, therefore, that the LAZ 388 cells are insensitive to MET-ENKby virtue of their viral transformation.

Activation of NALM 6 or LAZ 221 cells with MET-ENK correlated with an increase in migration (Fig. 4) and an

en-hancement insurface expression of CD9 (Figs. 2 and 3) but not otheradhesion molecules (e.g., CD4, CD44,

fI,

or/2integrins) (Table I). Interestingly, a similar relationship has been de-scribedby Letarte et al. (25) for pre-B ALL cellaggregations and the expression of adhesion receptors. These investigators

(25)reported that the size and number of homotypic aggregates formed by various pre-B ALL cell lines correlated with the surface level of CD9, but not other adhesion molecules includ-ing CD1a, CD18, CD19, CD44, orCD54. Our studies with human pre-B ALL cells have linked CD9 expression with an

important leukocyte function (migration).

The precise mechanism(s) by which CD9 promotes or

strengthens adhesions has not,asyet, been identified. CD9 isa

24-27-kD glycoprotein thatbelongs to thenovel superfamily

of proteinshavingfour transmembrane domains (TM4SF) (36-39).CD9aswellassomeotherproteins belongingtothesame

superfamily (i.e., TAPA-1 and M33) form multimeric com-plexes with other adhesion molecules (40-44) and it ispossible

that such multimericcomplexesareimportantfor cell adhesion andmigration.

PreincubationofNALM6 orLAZ222cells with anti-CD9 butnotanti-CD1O, anti-CD29, anti-ClassI oranti-ClassH mAb inhibited theirmigrationto MET-ENK(Table IV).Miyake et

al. (28) showed that the spontaneous locomotion of cultured humanlung and gastriccancercellswasdecreased by exposure

to anti-CD9 mAb. Ourfindings thatanti-CD9 mAb block the migration of pre-B ALL cells toMET-ENK provide the first evidence forCD9involvement in themigrationof anylymphoid

cell.Further, they suggest thatCD9surface moleculesregulate

either by direct (e.g., CD9-mediated adhesions) or indirect (e.g., CD9-mediated effects on other adhesion molecules)

meanstheleukemia cellmigrationtotheenkephalin. Cell

move-ment overvarioussurfaces(e.g., membranes, culture dishes,or

other cells) isaccomplishedviaaseries of coordinated

attach-ments(adhesions) and detachmentsfrom the surface(22, 24).

Such attachmentsaredependentontheexpression of cell

sur-face adhesion molecules. If CD9-mediated adhesions play a

direct role in the leukemia cell migration then it islikely that the anti-CD9 mAb inhibits their migration by blocking CD9-mediated cellto substrate adhesion. Previously, Ikeyamaetal. (27) transfected various kinds of human and murinetumorcells with human CD9 cDNA and determined that theoverexpression

of surface CD9 correlated withadecrease in the spontaneous motility of the cells. If CD9-mediated adhesions play adirect role in cell motility then it may be that the overexpression of CD9 in the transfectedtumorcells resulted in astrengthening of cell-to-cell andcell-to-substrate adhesions(i.e., induced cell clumping and firm attachments to the substrate) which then secondarily diminished their locomotion. On the other hand, it isplausiblethatCD9regulates cell motility via indirect effects exertedonother molecules. For example, it is possible, atleast in theory, that CD9 delivers transmembrane signals (45) and such

signaling

may render cells

refractory

andunresponsiveto

factors that stimulate cell movement (e.g., MET-ENK). If this is the case then cell exposure to anti-CD9 mAb may result in the activation of inhibitory signals (e.g., protein-tyrosine phosphorylation) and such signals may thendown-regulate the cell's motility.

Exposure of NALM 6 and LAZ 221 pre-B ALL cells to MET-ENKresulted in an increase inmigration andan enhance-ment in the surface expression of CD9. When the pre-B ALL cells werepreincubated with ananti-CD9 mAb before challenge with theenkephalin, their migration was diminished. Our find-ings suggest that the cell surface expression of CD9 plays a

rolein the enkephalin-stimulatedmigration of pre-B ALL cells. These studies provide (a) clues to the mechanisms by which opioid peptides modulate normal and abnormal lymphocyte functions and (b) support for the notion that opioid peptides play a direct andimportant role in some malignantdiseases.

Acknowledgments

The authors thank Mr. LelandLeCuyer for hisexpertassistance with graphics and photography and Ms. EdnaTeng forexcellent technical

support.

This workwassupported bythe William F. MiltonFundof Harvard University (Boston, MA) and a National Institutes of Health Grant ROA129657.

References

1. Sibinga, N. E. S., and A. Goldstein. 1988. Opioid peptidesandopioid

receptors in cells of the immune system.Annu.Rev.ImmunoL 6:219-249.

2. Manzanares, J., K. J.Lookingland,and K. E.Moore. 1990.

Kappa-opioid-receptor-mediated regulation ofa-melanocyte-stimulating hormonesecretion

and tuberohypophyseal dopaminergic neuronal activity. Neuroendocrinology. 52:200-205.

3. Gilmore, W.,M.Moloney, and L. P. Weiner. 1990.The roleofopioid peptides in immunomodulation.Ann. NYAcad. Sci. 597:252-263.

4.Pasternak,G. W. 1993.Pharmacologicalmechanismsof opioid analgesics.

Clin.NeuropharmacoL 16:1-18.

5.Stein, C.,A. H. S.Hassan,K.Lehrberger,J.Giefing,and A. Yassouridis. 1993. Localanalgesic effectofendogenous opioidpeptides.Lancet342:321

-323.

6. Peterson, P. K.,T. W.Molitor, and C. C. Chao. 1993.Mechanisms of

morphine-induced immunomodulation. Biochem. Pharmacol.46:343-348. 7.Shavit, Y., G.W.Terman,F. C.Martin,J. W.Lewis,J. C.Lieberkind,and R. P.Gale. 1985. Stress, opioid peptides, the immune system, and cancer. J.

Immunol. 135(Suppl.):834.

8. Chao,C.C.,B.M.Sharp,C.Pomeroy,G. A. Filice,and P. K. Peterson.

1990.Lethality of morphinein miceinfected with Toxoplasma gondii.J. Pharma-coL Exp.Ther.252:605-609.

9.Borboni, P.,G. D. G.Sesti,M. A.Marini,P.D.Porto,M.S. G.Montani,

R. Lauro, and R. D. Pirro. 1989./3-endorphinreceptors on cultured andfreshly

isolatedlymphocytesfrom normal subjects. Biochem. Biophys. Res. Commun.

163:642-648.

10.Vaughn, L.K.,W. S.Wire,P. Davis,Y.Shimohigashi, G. Toth, R. J.

Knapp, V. J. Hruby,T. F. Burks, and H. I. Yamamura. 1990. Differentiation

betweenratbrain and mouse vas deferens6 opioidreceptors. Eur. J.Pharmacol.

177:99-101.

11.Heagy, W., M. Laurance, E. Cohen, and R. Finberg. 1990. Neurohormones

regulateTcell function. J.Exp. Med. 171:1625-1633.

12.Ye, S.,R. R.Applegreen,J. M.Davis,and H. T.Cheung.1989.Modulation oflymphocyte motility by ,B-endorphin and met-enkephalin. Immunopharma-cology.17:81-89.

13.Ruff,M.R.,and C. B. Pert. 1987. Human monocyte chemotaxis to

neuro-peptides.InHypothalamic DysfunctioninNeuropsychiatricDisorders. D.

Ner-rozzi,F.Goodwin, and E.Costa,editors. RavenPress, Ltd.,New York. 247-260.

14. Saland,L. C., D. E. V. Epps,E. Ortiz,and A. Samora. 1983. Acute

injectionsofopiate peptidesinto the rat cerebral ventricle: amacrophage-like

cellular response. Res.Bull 10:523-528.

15.Fischer,E. G.1988.Opioid peptidesmodulate immunefunctions.

16. Moore, T. C. 1984. Modification of lymphocyte traffic byvasoactive

neurotransmittersubstances.Immunology. 52:511-518.

17. Foris, G., G. A. Medgyesi, J. T. Nagy, and Z. Varga. 1987.

Concentration-dependent effect of met-enkephalin on human polymorphonuclear leukocytes.

Ann. NYAcad. Sci.496:151-157.

18. Heagy, W., M. A. Shipp, and R. W. Finberg. 1992. Opioid receptor agonists andCa2+ modulation in human B cell lines. J. Immunol.

149:4074-4081.

19. Falke, N. E., and E. G.Fischer. 1985.Cell shape ofpolymorphonuclear leukocytes isinfluenced by opioids.Immunobiology. 169:532-539.

20. Maneckjee,R., and J. D.Minna. 1992. Nonconventional opioid binding

sitesmediate growth inhibitoryeffectsofmethadone on humanlungcancer cells. Proc.Natl. Acad.Sci. USA.89:1169-1173.

21.Mohler,J.L.,E. N.Broskie,D. J.Ranparia,Y.Sharief,W. B.Coleman,

and G. J. Smith. 1992.Cancercell motility-inhibitory protein intheDunning adenocarcinoma model.Cancer Res.52:2349-2352.

22.Shimizu,Y., W. Newman, Y.Tanaka,andS. Shaw. 1992. Lymphocyte

interactions with endothelial cells. Immunol.Today. 13:106-112.

23.Hemler, M. 1990. VLA proteins in theintegrin family:structures,

func-tions, andtheir role in leukocytes. Annu. Rev.Immunol. 8:365-400.

24.Springer,T.A. 1990.Adhesionreceptorsoftheimmunesystem. Nature

(Lond.). 346:425-433.

25.Letarte, M., J. G.Seehafer,A.Greaver,A.Masellis-Smith,and A. R. E. Shaw. 1993.Homotypic aggregationofpre-Bleukemic celllines by antibodies

toVLAintegrinscorrelates with theirexpression.Leukemia(Baltimore).

7:93-103.

26. Masellis-Smith,A.1990. Anti-CD9 monoclonal antibodies induce

homo-typic adhesion of pre-Bcell linesbyanovel mechanism. J. Immunol.

144:1607-1613.

27.Ikeyama, S.,M.Koyama,M.Yamaoko,R.Sasada,and M.Miyake. 1993. Suppressionof cellmotilityandmetastasis by transfectionwith human motility-related protein(MRP-l/CD9)DNA. J.Exp. Med. 177:1231-1237.

28. Miyake, M.,M.Koyama,M.Seno,and S.Ikeyama. 1991. Identification ofthe motility-related protein (MRP-1), recognized bymonoclonal antibody M31-15,which inhibits cellmotility.J.Exp. Med. 174:1347-1354.

29. Shipp, M. A., J. Vijayaraghavan, E. V. Schmidt,E. L. Masteller, L.

D'Adamio,L. B.Hersh,and E. L.Tempel. 1989.Common acutelymphoblastic leukemiaantigen (CALLA)is active neutralendopeptidase24.11 ("enkephali-nase"): direct evidencebycDNAtransfectionanalysis.Proc. Natl. Acad.Sci.

USA.86:297-301.

30.Bergelson, J. M.,M. P.Shepley, B. M.C. Chan, M. E. Hemler,and R. W. Finberg. 1992. Identification of the integrin VLA-2 as areceptorfor Echovirusl. Science(Wash DC). 255:1718-1720.

31.Sklar,L.A.,Z.G.Oades,and D. A.Finney. 1984. Neutrophil degranula-tion detectedby right angle light scattering:spectroscopicmethods suitable for

simultaneousanalyses of degranulationorshape change,elastaserelease, andcell

aggregation.J.Immunol. 133:1483-1487.

32.Wardlaw, S. L.,W. B.Wehrenberg,M.Ferin,P.W.Carmel, andA.G.

Frantz. 1980.Highlevels of3-endorphininhypophyseal portal blood. Endocrinol-ogy. 106:1323-1326.

33. Shipp,M.A., G.B.Stefano,L.D'Adamio, S.N.Swetzer,F. D.Howard,J.

Sinisterra,B.Scharrer,and E. L. Reinherz. 1990. Downregulation of

enkephalin-mediatedinflammatoryresponsesby CD10/neutralendopeptidase24.11.Nature (Lond.). 347:394-396.

34. Zagon,I.S., and P. J.McLaughlin. 1987. Modulation ofmurine

neuro-blastomain nude mice byopioid antagonists. J.Natl.Cancer Inst. 78:141-147. 35. Nilsson, K., and G. Klein. 1982.Phenotypicand cytogeneticcharacteristics ofhumanB-lymphoidcell lines and their relevancefortheetiologyofBurkitt's lymphoma.Adv. Cancer Res. 37:319-380.

36.Rubinstein,E., P.Benoit,M.Billard,S.Plaisance,M.Prenant, G. Uzan,

and C.Boucheix.1993. Organizationof the human CD9 gene. Genomics. 16:132-138.

37. Ebener, U., S. Wehner, J. Cinatl, E. S. Gussetis, and B. Kornhuber. 1990. Expression ofmarkers sharedbetween humanhaematopoietic cells and neuroblastoma cells. Anticancer Res.10:887-890.

38. Komada, Y., H. Ochiai, K.Shimizu, E. Azuma, H. Kamiya, and M.

Sakurai. 1992. Shedding of CD9antigenintocerebrospinalfluidbyacute

lymph-oblastic leukemia cells.Blood. 76:112-116.

39.Zola,H., V.Furness,S.Barclay,H.Zowtyj,M.Smith,J. V.Melo,S. H.

Neoh,and J.Bradley. 1989.Thep24 leucocyte membrane antigen: modulation associatedwithlymphocyte activation anddifferentiation.Immunol. Cell Biol.

67:63-70.

40. Imai,T., and0.Yoshie. 1993. C33antigenand M38antigen recognized by monoclonal antibodies inhibitoryto syncytium formation byhuman T cell

leukemia virustype 1 are both membersofthe transmembrane 4superfamily and associatewith each other and with CD4 or CD8 in T cells. J. Immunol. 151:6470-6481.

41.Seehafer,J.G.,and A. R. Shaw.1991. Evidencethatthesignal-initiating

membraneprotein CD9is associated with smallGTP-binding proteins. Biochem. Biophys.Res. Commun. 179:401-406.

42. Slupsky, J. R., J. G. Seehafer, S. C. Tang, A. Masellis-Smith, and A. R.E.Shaw.1989.Evidencethat monoclonalantibodies againstCD9antigen induce specificassociation between CD9 and the plateletglycoproteinIlb-IIIa complex. J. Biol.Chem. 264:12289-12293.

43. Mitamura, T.,R.Iwamoto, T. Umata, T.Yomo, I. Urabe, andM.Tsuneoka.

1992.The 27-kDdiphtheriatoxinreceptor-associated protein (DRAP27) from

Vero cells is the monkey homologue of human CD9 antigen: expression of DRAP27 elevates the number ofdiphtheriatoxin receptors on toxin-sensitive

cells.J.CellBiol. 1992:1389-1399.

44.Bradbury,L. E., V. S.Goldmacher, andT. F.Tedder. 1993.TheCD19

signal transduction complexof Blymphocytes:deletionoftheCD19cytoplasmic domainalterssignaltransduction but notcomplex formationwithTAPA-1and Leu13. J.Immunol. 151:2915-2927.