mdc-9 (adam-9/meltrin γ) functions as an adhesion molecule by binding the αvβ5 integrin

MM

MD

R

mfmibctidsibtiFctimA

t

rfcpdmfepcthpch

Biochemical and Biophysical Research Communications 280, 574–580 (2001)

doi:10.1006/bbrc.2000.4155, available online at http://www.idealibrary.com on

0CA

DC-9 (ADAM-9/Meltrin g) Functions as an Adhesionolecule by Binding the avb5 Integrin

in Zhou, R. Graham, G. Russell, and Peter I. Croucherivision of Genomic Medicine, University of Sheffield Medical School, Sheffield, United Kingdom

eceived December 11, 2000

cellular matrix components (11–13). However, less iskA

dwnHmtdwcphtmtcMtomcbaisMtedsimo

M

apL

MDC-9 is a widely expressed member of theetalloproteinase/disintegrin/cysteine-rich protein

amily. The disintegrin domain of MDC-9 lacks an RGDotif but has recently been reported to bind the a6b1

ntegrin; however, it is unclear whether MDC-9 canind other integrins. In the present study myelomaells, but not lymphoblastoid cells, were shown to bindo immobilised, recombinantly expressed MDC-9 dis-ntegrin domain (A9dis). Binding was divalent cation-ependent, being supported by Mn21 and Ca21. Adhe-ion of myeloma cells to A9dis was completelynhibited by an antibody to the avb5 integrin but noty antibodies to other subunits. RGD-containing pep-ides had no effect on binding, suggesting that MDC-9nteracts with avb5 in an RGD-independent manner.low cytometric analyses demonstrated that myelomaells, but not lymphoblastoid cells, expressed avb5 onhe cell membrane. These data indicated that the dis-ntegrin domain of MDC-9 can function as an adhesion

olecule by interacting with an avb5 integrin. © 2001

cademic Press

Key Words: MDC-9; ADAM-9; meltrin g; adhesion; in-egrin; avb5; disintegrin.

The MDC (metalloproteinase/disintegrin/cysteine-ich) or ADAM (a disintegrin and metalloproteinase)amily is a recently described group of proteins with aharacteristic domain structure (1). This includes theresence of a pro-region, a metalloproteinase domain, aisintegrin-like region, a cysteine-rich domain, a trans-embrane region and a short cytoplasmic tail (1). The

unctions of members of this family are unclear. How-ver, their domain organisation suggest that they maylay roles in a number of important biological pro-esses, including proteolysis, cell/cell or cell/matrix in-eractions and cell signalling (2–4). Recent studiesave focused on their role in proteolysis and have im-licated members of the family in the processing ofytokines, the shedding of cytokine receptors and ad-esion molecules(5–10), and the remodelling of extra-

574006-291X/01 $35.00opyright © 2001 by Academic Pressll rights of reproduction in any form reserved.

nown of the function of other domains of MDC/DAMs, including the disintegrin-like domain.Studies have demonstrated that the disintegrin-like

omain of the sperm protein fertilin b (ADAM-2),hich mediates sperm–egg fusion (14–16), may recog-ise the a6b1 integrin on the egg membrane (17–19).uman metargidin (ADAM-15), which is the onlyember of the MDC family to contain the RGD tripep-

ide motif in the disintegrin-like domain (20), has beenemonstrated to mediate cell adhesion via interactionsith either avb3 or a5b1 integrins (21, 22). More re-

ently, murine metargidin, which contains TDD inlace of RGD and ADAM-12, which also lacks RGD,ave been reported to bind the integrin a9b1 (23). Fur-hermore, ADAM-23, another non-RGD containingember of the family has recently been shown to in-

eract with the avb3 integrin (24). However, it is un-lear whether the non-RGD containing members of theDC/ADAM family interact with a single specific in-

egrin or whether they have the capacity to bind tother integrin heterodimers, as is the case with humanetargidin (22). MDC-9, a widely expressed, non-RGD

ontaining, member of this family (25, 26), has recentlyeen shown to bind to the a6b1 integrin on fibroblasts,lthough it is not clear whether it can also bind to otherntegrins on different cells. Therefore in the presenttudy we have expressed the disintegrin-like domain ofDC-9 as a GST fusion protein and demonstrate that

his protein can also mediate binding to human my-loma cells. Furthermore, we provide evidence that theistingrin-like domain of MDC-9 mediates this adhe-ive interaction by binding to the avb5 integrin suggest-ng that this non-RGD containing MDC/ADAM family

ember has the capacity to interact with more thanne integrin.

ATERIALS AND METHODS

Antibodies and cell lines. The mouse IgG1 anti-human avb5 mAbscites (P1F6), mouse IgG1 anti-human a2 mAb ascites (P1E6), RGDeptide (GRGDSP) and control peptide (GRGESP) were from GIBCOife-Technologies (Paisley, UK) and the mouse IgG1 anti-human b3

antibody (PM6/13) was purchased from Serotec (Oxford, UK). TherammtHo(bRL

as

cecpGGfmtpdlacstarSAgsumbuab

ttsprw

hfcpacopflce

bfc0N

25 mM Hepes, pH 7.4, were added to A9dis-coated plates and incu-bCpbommcpma1flapd

iflBwaobb

msct

R

E

wpssfrpaIbmp4pevwcwpgidi

Vol. 280, No. 2, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

at anti-human b1 (mAB 13) and rat anti-human a5 (mAB 16) werekind gift from Prof. K. Yamada (NIH, U.S.A.). Isotype controlouse IgG1 and rat IgG were from Sigma (Poole, UK). Rabbit antiouse F(ab9)2 IgG and HRP conjugated anti-rabbit IgG were ob-

ained from DAKO (Ely, UK). The human myeloma cell lines NCI-929 and RPMI-8226, and lymphoblastic cell line, ARH-77, were

btained from the European Collection of Animal Cell CulturesSalisbury, UK). The human myeloma cell line, JJN-3, was providedy Professor I. Franklin (Glasgow, UK). Cells were maintained inPMI 1640 supplemented with 10% fetal calf serum (FCS), 2 mM-glutamine, 1 mM sodium pyruvate, 13 MEM non-essential aminocids and 5 mM 2-mercaptoethanol at 37°C in a humidified atmo-phere of 95% air and 5% CO2.

Generation of GST/MDC-9 disintegrin domain fusion protein. ADNA encoding the entire disintegrin domain, and containing anqual number of cysteine residues, was generated by polymerasehain reaction (PCR) from a JJN-3 cDNA library using the followingrimers, 59-primer, 59-GGA-TCC-AGT-GCA-GAG-GAC-TTT-AG-AA and 39-primer, 59-G-AAT-TCT-GCC-GTT-GTA-GCA-ATA-GC. The 59-primer contained a BamHI restriction site introduced

or subcloning and encodes the most C9-terminal 7 amino acids of theetalloproteinase domain. The 39 primer contained an EcoRI restric-

ion site and was complementary to the cDNA sequence of MDC-9,redicted to encode the start (N-terminal region) of the cysteine-richomain between residues 508–515. The cDNA fragment bearing fullength MDC-9 disintegrin domain was subcloned into the BamHInd EcoRI site of a pGEX-2T expression vector (Amersham Pharma-ia Biotech, Little Chalfont, UK) and transformed into the XL-1 Bluetrain of E. coli (Stratagene, Cambridge, UK). Plasmids containinghe disintegrin domain of MDC-9 (pGEX-2T-A9dis) were sequencednd transformed into the BL21 strain of E. coli for production ofecombinant protein (A9dis). Synthesis of the glutathione-transferase (GST)/MDC-9 disintegrin domain fusion protein (GST/9dis) was induced in BL21 cells by 0.1 mM isopropyl-1-thio-b-D-alactopyranoside (IPTG) (Amersham Pharmacia Biotech). The fu-ion protein was extracted by mild sonication in PBS and purifiedsing glutathione Sepharose 4B (Pharmacia Biotech) affinity chro-atography. A9dis was released from GST by cleavage with throm-

in (Pharmacia Biotech). The concentration of A9dis was quantitatedsing the BCA protein assay. The quality of the purified GST/A9disnd A9dis was determined by SDS–PAGE with Coomassie brilliantlue staining.

Cell adhesion assays. 96-well microtiter plates were coated withhe purified recombinant A9dis, diluted to the specified concentra-ion in PBS (pH 7.2). As a control, protein was also prepared byubjecting the empty pGEX-2T vector to an identical expression,urification and coating protocol as the pGEX-2T-A9dis vector. Theemaining protein binding sites were blocked by incubating wellsith bovine serum albumin (10 mg/ml, BSA, Sigma) in PBS.To assess whether A9dis could function as an adhesion molecule,

uman myeloma cells or human lymphoblastoid cells were harvestedrom tissue culture, washed and resuspended in RPMI 1640 mediumontaining 1 mg/ml BSA. 1 3 105 cells were added to each well ofrecoated 96-well plates and incubated at 37°C for 1.5 h. The non-dherent cells were removed by gentle washing with PBS and boundells were then fixed with 4% formaldehyde. Wells were then washednce in 70% ethanol, once with PBS and air-dried. 100 ml of 5 mMropidium iodide were added to each well to stain the nuclei. Theuorescence associated with each well was measured on a fluores-ence plate reader at an excitation wavelength of 530 nm and anmission wavelength of 620 nm.In experiments to examine the effect of divalent cations on cell

inding, human myeloma cells (1 3 105), containing one of theollowing divalent ions, Mg21, Mn21, Ca21, Zn21 and Ni21, at a con-entration of 1 mM in Hanks’ balanced salt solution (HBSS) (KCl,.40 g/L, KH2PO4, 0.06 g/L, NaCl, 8.00 g/L, NaHCO3, 0.35 g/L,a2HPO4, 0.48 g/L, D-glucose, 1 g/L) containing 1 mg/ml BSA and

575

ated at 37°C for 1.5 h. To determine whether the concentration ofa21 affects cell adhesion to A9dis binding was performed in theresence of Ca21 at a concentration of 0.05, 0.5, 1 and 2.5 mM. Inlocking experiments, cells were preincubated with 1:5000 dilutionf anti-avb5 mAb, 10 mg/ml anti-b3, 1:150 dilution of anti-a2 mAb, 10g/ml anti-b1, 15 mg/ml anti-a5, or appropriate control in RPMI 1640edium containing 1 mg/ml BSA for 45 min. The RGD peptide and

ontrol peptide, at a concentration of 10 mg/ml, were also used tore-incubate cells prior to introduction into the adhesion assay. Theyeloma cells (1 3 105/well), containing blocking agents, were then

dded to the A9dis or control coated wells and allowed to adhere for.5 h. Adherent cells were stained with propidium iodide and theuorescence associated with each well was assessed as describedbove. The fluorescence associated with wells coated with controlrotein was subtracted from that of those coated with A9dis toetermine levels of specific binding.

Flow cytometric analysis. To assess the level of expression of thentegrin avb5 on RPMI-8226, NCI-H929, JJN-3 and ARH-77 cells,ow cytometric analysis was performed using standard protocols.riefly, 1 3 106 cells were removed from culture and washed. Cellsere incubated with mouse anti-avb5 antibody or isotype controlntibody, followed by FITC-conjugate F(ab9)2 rabbit anti-mouse sec-ndary antibody. PI was added at a final concentration of 8 mg/mlefore samples were analysed on a Becton–Dickinson FACSortenchtop cytometer.

Statistics. Data are presented at the mean 6 SD of triplicateeasurements. Experiments were repeated on a minimum of three

eparate occasions and a representative experiment is shown in eachase. Comparisons between groups were performed by Student’s test.

ESULTS

xpression of the MDC-9 Disintegrin Domain Protein,A9dis

A cDNA encoding the MDC-9 disintegrin domainas amplified by PCR and introduced into theolylinker of pGEX-2T. Automated fluorescence DNAequencing confirmed the MDC-9 disintegrin domainequence was located 59 and immediately downstreamrom the thrombin cleavage site, and in the correcteading frame (Fig. 1A). The pGEX-A9dis construct, orGEX-2T alone, were transformed into E. coli BL21nd expression induced using the lactose analoguePTG. The GST-A9dis fusion protein was expressed atetween 2.5 and 8 mg per litre of culture, under opti-al induction conditions. Figure 1B shows the overex-

ression of the fusion protein, at the expected size of5kDa, by the cells carrying the pGEX-2T-A9dis ex-ression construct (Fig. 1B, lane 2) and of GST, at thexpected size of 29 kDa, by the cells bearing the controlector pGEX-2T (Fig. 1B, lane 1). The fusion proteinas purified by Glutathione-Sepharose 4B affinity

hromatography and liberated from GST by cleavageith thrombin. SDS–polyacrylamide gel electro-horetic analysis showed that the fusion protein mi-rated at a position of 45 kDa (Fig. 1B, lane 3). Follow-ng cleavage with thrombin the MDC-9 disintegrinomain (A9dis) was shown to be approximately 16 kDan size (Fig. 1B, lane 5), whereas the GST that was

el

T

M8tepNc

nificant binding of each of the myeloma cell lines toAms

e0smb0ceNonsNtssc

mslamC

ccawcip6mms

dtPmmbb01sLicr


luted from the affinity matrix was 29 kDa (Fig. 1B,ane 4).

he Disintegrin Domain of MDC-9 Functionsas an Adhesion Molecule

To determine whether the disintegrin domain ofDC-9 could function as an adhesion molecule, RPMI-

226, JJN-3, NCI-H929 and ARH-77 cells were allowedo adhere to microtitre plates coated with A9dis, atither 5, 25 or 100 mg/ml, or an appropriate controlrotein. The myeloma cell lines RPMI-8226, JJN-3 andCI-H929 were each able to bind to A9dis protein in a

oating concentration dependent manner (Fig. 2). Sig-

FIG. 1. Expression and purification of the MDC-9 disintegrinomain protein. (A) Schematic representation of the domain struc-ure of MDC-9 and the GST-A9dis fusion protein. S, signal sequence;ro, pro-region; MP, metalloproteinase domain; DI, disintegrin do-ain; Cys, cysteine-rich region; EGF, EGF-like domain; TM; trans-embrane domain; and Cyto, cytoplasmic tail. The amino acid num-

ers are indicated. (B) Expression of the A9dis protein was analysedy 12.5% SDS–polyacrylamide gel and visualised by staining with.1% Coomassie brilliant blue R-250. Lane M, protein markers. Lane, supernatant of lyzed cells carrying the control GST vector. Lane 2,upernatant of lyzed cells carrying the GST-A9dis expression vector.ane 3, GST-A9dis protein purified by Glutathione-Sepharose affin-

ty chromatography. Lane 4, GST separated from A9dis followingleavage with thrombin. Lane 5, Purified MDC-9dis protein sepa-ated from GST following thrombin cleavage.

576

9dis was seen at concentrations of protein as low as 5g/ml. In contrast, the lymphoblastic cell line, ARH-77,howed little ability to bind to A9dis.Ca21 (1 mM) increased binding of each of the my-

loma cell lines examined to A9dis by 8- to 10-fold (P ,.005 in each case) (Fig. 3). Mn21 (1 mM) was alsohown to promote an increase (4- to 8-fold) in A9dis-ediated myeloma cell adhesion, although the level of

inding varied between cells lines (NCI-H929, P ,.05; JJN-3, P , 0.01; and RPMI-8226, P , 0.005). Inontrast, the presence of Zn21 or Ni21 had only modestffects on adhesion to A9dis. The presence of Zn21 ori21 were able to promote a small increase in adhesion

f JJN-3 cells to A9dis (P , 0.05); however, Ni21, butot Zn21, was able to significantly increase the adhe-ion of RPMI-8226 cells (P , 0.05). Neither Zn21 nori21 were able to promote adhesion of NCI-H929 cells

o the A9dis protein. Mg21, which has been reported toupport a number of integrin ligand interactions, washown to have no effect on A9dis-mediated myelomaell adhesion.Although Ca21 was shown to promote adhesion ofyeloma cells to A9dis, previous studies have demon-

trated that Ca21 can both support and inhibit integrinigand interactions depending on the specific ligandnd the Ca21 concentration. In the present studyyeloma cells were shown to bind to A9dis in aa21 concentration dependent manner (Fig. 3B). Al-

FIG. 2. Myeloma cells adhere to the A9dis protein. 96-well mi-rotiter plates were coated with 5, 25, and 100 mg/ml of A9dis, orontrol protein. RPMI-8226, JJN-3, and NCI-H929 myeloma cellsnd ARH-77 lymphoblastoid cells were incubated in the coated 96-ell plates for 1.5 h at 37°C. After washing to remove unbounded

ells, the nuclei of cells adhered to plate were stained with propidiumodide and levels of fluorescence was measured with a fluorescencelate reader. Data are expressed in fluorescence units measured at20 nm. The data shown represent one of four independent experi-ents. Each value represents the mean 6 SD of triplicate measure-ents after the corresponding values of control protein had been

ubtracted.

tehA

The Integrin avb5 Mediates Adhesion of Myeloma

acgaiit(thwbimp

svdmtattt

M

lircaiRcc(

D

bcamtipia

pwNdmK0pttnHcsiopc


hough at low concentrations (0.05 mM) Ca21 had noffect on cell adhesion, concentrations of 0.5 mM origher promoted a significant increase in binding to9dis.

FIG. 3. Adhesion of myeloma cells to the MDC-9 disintegrinrotein is cation dependent. 96-well microtiter plates were coatedith 100 mg/ml of A9dis, or control protein. (A) RPMI-8226, JJN-3 orCI-H929 myeloma cells (1 3 105), containing one of the followingivalent ions: Mg21, Mn21, Ca21, Zn21 or Ni21, at a concentration of 1M, in Hanks’ balanced salt (HBSS) solution (KCl, 0.40 g/L,H2PO4, 0.06 g/L, NaCl, 8.00 g/L, NaHCO3, 0.35 g/L, Na2HPO4,.48g/L, D-glucose, 1 g/L) containing 1 mg/ml BSA and 25 mM Hepes,H 7.4, were added and incubated for 1.5 h at 37°C. After washing,he cells were staining with PI and levels of fluorescence were de-ermined. Data are expressed in fluorescence units measured at 620m. (B) RPMI-8226, JJN-3 or NCI-H929 myeloma cells (1 3 105), inBSS containing increasing concentrations Ca21, were added to

oated plates and incubated for 1.5 h at 37°C. Adhesion was mea-ured as described previously. The data shown represent one of threendependent experiments and each value represents the mean 6 SD.f triplicate determination after the corresponding values of controlrotein had been subtracted. *P ,0.05, **P , 0.01, ***P , 0.005ompared to the no ion control.

577

Cells to A9dis

To define the molecules that may be involved indhesion to the A9dis protein, myeloma cells were in-ubated with antibodies raised against specific inte-rins, or integrin subunits, including, b1, b3, a2, a5 andvb5, prior to binding to immobilised A9dis. The bind-

ng of myeloma cells to A9dis was almost completelynhibited by the anti-avb5 antibody, when compared tohe effect of an isotype control antibody (P , 0.001)Fig. 4A). The anti-b1 antibody was also able to inhibithe adhesion of myeloma cells to A9dis (P , 0.05);owever, the magnitude of the response was smallhen compared to the inhibition achieved by the anti-ody raised against avb5. Anti-a2 and anti-a5 antibod-es had no effect on binding to A9dis. Interestingly,

yeloma cell adhesion to A9dis was increased in theresence of the anti-b3 antibody.The integrin avb5 has been reported to be an adhe-

ion receptor for vitronectin (27) and osteopontin (28)ia binding to their RGD-binding sites. However, theisintegrin domain of MDC-9 does not possess an RGDotif in the predicted integrin-binding site but con-

ains a TSE tripeptide sequence instead. Furthermore,RGD containing peptide (GRGDSP) had no effect on

he binding of myeloma cells to A9dis when comparedo binding in the absence of peptide (control) or a con-rol peptide (GRGESP) (Fig. 4B).

yeloma Cells, but Not Lymphoblastoid Cells,Express the avb5 Integrin

Myeloma cell lines, but not the lymphoblastoid celline, were shown to bind to A9dis. Furthermore bind-ng could be almost completed inhibited by an antibodyaised against the avb5 integrin; however, it was notlear whether this difference was a reflection of thebility of the different cell lines to express the avb5

ntegrin. Flow cytometric analysis confirmed thatPMI-8226, NCI-H929 and JJN-3 human myelomaells stained strongly for the avb5 integrin (Fig. 5). Inontrast, ARH-77 cells, showed little staining for avb5

Fig. 5).

ISCUSSION

In the present study myeloma cells were shown toind to the disintegrin-like domain of MDC-9 and thisould be inhibited by an antibody raised against thevb5 integrin. An antibody to the b1 subunit had aodest effect on adhesion. Although, this may suggest

hat binding could be mediated in part by a b1 contain-ng integrin, the level of inhibition was weak in com-arison to that seen with anti-avb5 antibody. Antibod-es to the b3 and a5 integrin subunits did not inhibitdhesion to the disintegrin domain of MDC-9 suggest-

imibssgaqttbAnptMpac

ntlbpcp

waccAfS*

hHchalaaTfl2tH


ng that different members of the MDC/ADAM familyay bind specifically to different integrins. Surpris-

ngly, the antibody to the b3 integrin promoted a smallut significant increase in binding. Although the rea-on for this is unclear, this may suggest that the b3

ubunit is involved in an interaction with the disinte-rin domain of MDC-9, possibly in conjunction with anv subunit, but that activation of this integrin is re-uired before binding can take place. These data con-rast with the data from Nath et al. that demonstratedhat the extracellular region of MDC-9 could mediateinding of fibroblasts by binding to a a6b1 integrin (29).lthough binding of MDC-9 to the avb5 integrin wasot examined in this study, binding was almost com-letely inhibited by antibodies to a6 and b1, suggestinghat in fibroblasts, other integrins do not mediateDC-9 adhesion(29). Taken together these data sup-

ort the suggestion that MDC-9 may be able to mediatedhesion by binding to different integrins on differentells.

The avb5 integrin is thought to be able to bind to aumber of extracellular matrix proteins including, vi-ronectin and osteopontin (27, 28). However, there areittle data to demonstrate binding to an integral mem-rane protein. Binding to these extracellular matrixroteins appears to be RGD-dependent, as peptidesontaining this tripeptide motif inhibit binding to theseroteins. In the present study RGD-containing pep-

FIG. 4. The effects of blocking antibodies (A) and an RGD containere coated with A9dis. RPMI-8226 myeloma cells were preincubate1:150 dilution of anti-a2 mAb, 10 mg/ml anti-b1, 15 mg/ml anti-a

ontaining peptide or control peptide at a concentration of 10 mg/ml, inells (1 3 105/well), containing blocking agents, were then added todhesion was measured as described in the legend to Fig. 2. The flu

rom that of the A9dis-coated wells. The data shown are a representD) represent adhesion as a percentage of control after the corresponP , 0.05, **P , 0.005, ***P , 0.001 compared to the appropriate

ing peptide (B) on the adhesion of myeloma cells to A9dis. 96-well platesd with either (A) a 1:5000 dilution of anti-avb5 mAb, 10 mg/ml anti-b3,5, or an appropriate species specific control antibody, or (B) an RGDRPMI 1640 medium containing 1 mg/ml BSA, for 45 min. The myelomathe A9dis-coated wells in triplicate and allowed to adhere for 1.5 h.

orescence associated with control protein coated wells was subtractedative experiment from three independent experiments. Data (mean 6ding values from wells coated with control protein had been subtracted.control.

578

FIG. 5. Flow cytometric analysis of avb5 integrin expression byuman myeloma cells and lymphoblastic cells. RPMI-8226 (A), NCI-929 (B), JJN-3 (C) myeloma cells and ARH-77 (D) lymphoblastoid

ells were stained with either a mouse anti-avb5 antibody (dark greyistograms) or an isotype control antibody (light grey histograms)nd then with a FITC-conjugated goat anti-mouse antibody. Theevel of avb5 upon the cell surface was determined by flow cytometry,s described under Materials and Methods. The data are representeds histograms of FITC-fluorescence (x-axis) versus counts (y-axis).he mean channel fluorescence values for the isotype control-FITC-uorescence data are RPMI-8226 5 3.17, NCI-H929 5 3.0, JJN-3 5.57 and ARH-77 5 2.47. The mean channel fluorescence values forhe anti-avb5-FITC-fluorescence data are RPMI-8226 5 28.01, NCI-929 5 9.82, JJN-3 5 23.11, and ARH-77 5 4.51.

tides had no effect on the adhesion of myeloma cells totsdi

dlwaddsvwopintdm(IcA(tbpssbredsdbifitf

maptwtmcidma

integrins. However, since the entire extracellular do-msimhatamsstttbois

vcsatmtaamt

A

Rmfat

R


he disintegrin domain of MDC-9. These data thereforeuggest that the binding site in avb5, for the disintegrinomain of MDC-9 may be distinct from the RGD bind-ng domain.

In the present study adhesion of myeloma cells to theisintegrin domain of MDC-9 was promoted by diva-ent cations. However, the profile of cation dependencyas different than that reported previously for othervb5-substrate interactions. Ca21 was able to promote aose-dependent increase in binding to the disintegrinomain of MDC-9. This is consistent with previoustudies that have demonstrated that binding of avb5 toitronectin can also be supported by Ca21 but contrastith the demonstration that avb5 mediated binding tosteopontin is insensitive to Ca21 (28). Mn21, whichromotes the ligand binding ability of a number ofntegrins, including avb5 dependent adhesion to vitro-ectin and osteopontin (28), was also able to promotehe ability of myeloma cells to bind to the disintegrinomain of MDC-9. However, Mg21which supports avb5

ediated adhesion to both vitronectin and osteopontin28) had no effect on myeloma cell adhesion to MDC-9.n support of this, Chen et al. have demonstrated thatation requirements for the binding of a6b1 integrin toDAM-2 differ from that required to bind laminin

30). Ca21 promotes and Mn21 inhibits a6b1 bindingo ADAM-2, whereas the converse is the case forinding to laminin, with Ca21 inhibiting and Mn21

romoting adhesion (30). These data support theuggestion that the binding sites, and/or activationtates, required for avb5 binding to MDC-9, and a6b1

inding of ADAM-2, are distinct from that previouslyeported for the binding of these integrins to specificxtracellular matrix proteins, as the cation depen-encies for these interactions are different. However,tudies from Eto et al. have suggested that the cationependency for ADAM-12 and murine ADAM-15inding to the a9b1 integrin do not differ from knownntegrin/extracellular matrix interactions (23). Thus,urther studies will be required to determine themportance of specific cations in ADAM/integrin in-eractions and whether the specificity relates to dif-erent ADAMs or to different substrates.

In the present study we demonstrate that humanyeloma cells bind to MDC-9 and they do so via the

vb5 integrin. Lymphoblastoid cells, which do not ex-ress avb5 do not bind to MDC-9. These data suggesthat at least in these cells MDC-9, interacts specificallyith the avb5 integrin. However, Nath et al. were able

o show that the adhesion of fibroblasts to MDC-9 wasediated by the a6b1 integrin suggesting that different

ell types may use different integrins to mediate bind-ng to MDC-9. In support of this Nath et al. have alsoemonstrated that haemopoietic cells can bind to hu-an metargidin using two different integrins, either

vb3 or a5b1, with different cell types utilising different

579

ain was used to investigate binding in each of thesetudies, the possibility that different integrins maybenteracting with the different extracellular domains

ust be considered. In support of this, recent studiesave suggested that peptides from the metalloprotein-se domain of jararhagin, a snake venom metallopro-einase that also contains a disintegrin domain, may beble to bind to the a2 integrin subunit (31). Further-ore, the cysteine-rich region of ADAM-12 has been

hown to support cell adhesion via interactions withyndecans and the b1 integrin subunit (32). Thus, al-hough in the present study the avb5 integrin bound tohe disintegrin-like domain of MDC-9 only, members ofhe MDC/ADAM family, or related molecules, may alsoe able to mediate cell adhesion via interactions withther domains. This could account for the differentntegrin requirements observed between the differenttudies.In conclusion, the data presented in this study pro-

ide evidence to suggest that the non-RGD containingellular disintegrin, MDC-9, mediates cell/cell adhe-ion. Adhesion to immobilised MDC-9 was mediated, innon-RGD dependent manner, via an interaction with

he avb5 integrin supporting the suggestion that thisember of the MDC/ADAM family can bind to more

han one integrin. These data also suggest that thevb5 integrin may be able to mediate cell to cell inter-ctions in addition to being able to bind to extracellularatrix proteins. Thus, MDC-9 must now be considered

o be a potential receptor for the avb5 integrin.

CKNOWLEGMENTS

This work was supported in part by grants from the Leukaemiaesearch Fund and Yorkshire Cancer Research. P.I.C. is a Leukae-ia Research Fund Bennett Senior Fellow. We are grateful to Pro-

essor K. Yamada for provision of antibodies mAB 13 and mAB 16,nd to Professor M. Humphries for expert advice during the course ofhese studies.

EFERENCES

1. Wolfsberg, T. G., Primakoff, P., Myles, D. G., and White, J. M.(1995) J. Cell Biol. 131, 275–278.

2. Huovial, A.-P. J., Almeida, E. A. C., and White, J. M. (1996) Curr.Opin. Cell Biol. 8, 692–699.

3. Blobel, C. P. (1997) Cell 90, 589–592.4. Black, R. A., and White, J. M. (1998) Curr. Opin. Cell Biol. 10,

654–659.5. Black, R. A., Rauch, C. T., Kozlosky, C. J., Peschon, J. J., Slack,

J. L., Wolfson, M. F., Castner, B. J., Stocking, K. L., Reddy, P.,Srinivasan, S., Nelson, N., Boiani, N., Schooley, K. A., Gerhart,M., Davis, R., Fitzner, J. N., Johnson, R. S., Paxton, R. J., March,C. J., and Cerretti, D. P. (1997) Nature 385, 729–733.

6. Moss, M. L., Jin, S.-L. C., Milla, M. E., Burkhart, W., Carter,H. L., Chen, W.-J., Clay, W. C., Didsbury, J. R., Hassler, D.,Hoffman, C. R., Kost, T. A., Lambert, M. H., Leesnitzer, M. A.,McCauley, P., McGeehan, G., Mitchell, J., Moyer, M., Pahel, G.,Rocque, W., Overton, L. K., Schoenen, F., Seaton, T., Su, J.-L.,

Warner, J., Willard, D., and Becherer, J. D. (1997) Nature 385,

1

1

1

1

1

1

1

1

A., Primakoff, P., Myles, D. G., and White, J. M. (1995) Cell 81,

11

2

2

2

2

2

2

2

2

2

2

3

3

3


733–736.7. Izumi, Y., Hirata, M., Hasuwa, H., Iwamoto, R., Umata, T.,

Miyado, K., Tamai, Y., Kurisaki, T., Sehara-Fujisawa, A., Ohno,S., and Mekada, E. (1998) EMBO J. 17, 7260–7272.

8. Peschon, J. J., Slack, J. L., Reddy, P., Stocking, K. L., Sunnar-borg, S. W., Lee, D. C., Russell, W. E., Castner, B. J., Johnson,R. S., Fitzner, J. N., Boyce, R. W., Nelson, N., Kozlosky, C. J.,Wolfson, M. F., Rauch, C. T., Cerretti, D. P., Paxton, R. J.,March, C. J., and Black, R. A. (1998) Science 282, 1281–1284.

9. Lum, L., Wong, B. R., Josien, R., Becherer, J. D., Erdjument-Bromage, H., Schlondorff, J., Tempst, P., Choi, Y., and Blobel,C. P. (1999) J. Biol. Chem. 274, 13613–13618.

0. Qi, H. L., Rand, M. D., Wu, X. H., Sestan, N., Wang, W., Rakic,P., Xu, T., and Artavanis-Tsakonas, S. (1999) Science 283, 91–94.

1. Millichip, M., Dallas, D., Wu, E., Dale, S., and McKie, N. (1998)Biochem. Biophys. Res. Commun. 245, 594–598.

2. Abbaszade, I., Liu, R. Q., Yang, F., Rosenfeld, S. A., Ross, O. H.,Link, J. R., Ellis, D. M., Tortorella, M. D., Pratta, M. A., Hollis,J. M., Wynn, R., Duke, J. L., George, H. J., Hillman, M. C.,Murphy, K., Wiswall, B. H., Copeland, R. A., Decicco, C. P.,Bruckner, R., Nagase, H., Itoh, Y., Newton, R. C., Magolda, R. L.,Trzasko, J. M., Hollis, G. F., Arner, E. C., and Burn, T. C. (1999)J. Biol. Chem. 274, 23443–23450.

3. Tortorella, M. D., Burn, T. C., Pratta, M. A., Abbaszade, I.,Hollis, J. M., Liu, R., Rosenfeld, S. A., Copeland, R. A., Decicco,C. P., Wynn, R., Rockwell, A., Yang, F., Duke, J. L., Solomon, K.,George, H., Bruckner, R., Nagase, H., Itoh, Y., Ellis, D. M., Ross,H., Wiswall, B. H., Murphy, K., Hillman, M. C., Hollis, G. F.,Newton, R. C., Magolda, R. L., Trzaskos, J. M., and Arner, E. C.(1999) Science 284, 1664–1666.

4. Blobel, C. P., Wolfsberg, T. G., Turck, C. W., Myles, D. G.,Primakoff, P., and White, J. M. (1992) Nature 356, 248–252.

5. Evans, J. P., Schultz, R. M., and Kopf, G. S. (1995) J. Cell Sci.108, 3267–3278.

6. Cho, C., O’Dell Bunch, D., Faure, J.-E., Goulding, E. H., Eddy,E. M., Primakoff, P., and Myles, D. G. (1998) Science 281, 1857–1859.

7. Almeida, E. A. C., Huovila, A.-P. J., Sutherland, A. E., Stephens,L. E., Calarco, P. G., Shaw, L. M., Mercurio, A. M., Sonnenberg,

580

1095–1104.8. Chen, H., and Sampson, N. S. (1999) Chem. Biol. 6, 1–10.9. Bigler, D., Takahashi, Y., Chen, M. S., Almeida, E. A. C., Os-

bourne, L., and White, J. M. (2000) J. Biol. Chem. 275, 11576–111584.

0. Kratzschmar, J., Lum, L., and Blobel, C. P. (1996) J. Biol. Chem.271, 4593–4596.

1. Zhang, X.-P., Kamata, T., Yokoyama, K., Puzon-McLaughlin, W.,and Takada, Y. (1998) J. Biol. Chem. 273, 7345–7350.

2. Nath, D., Slocombe, P. M., Stephens, P. E., Warn, A., Hutchin-son, G. R., Yamada, K. M., Docherty, A. J. P., and Murphy, G.(1999) J. Cell. Sci 112, 579–587.

3. Eto, K., Puzon-McLaughlin, W., Sheppard, D., Sehara-Fujisawa,A., and Takada, Y. (2000) J. Biol. Chem., in press.

4. Cal, S., Freije, M. P., Lopez, J. M., Takada, Y., and Lopenz-Otin,C. (2000) Mol. Biol. Cell 11, 1457–1469.

5. Weskamp, G., Kratzschmar, J., Reid, M. S., and Blobel, C. P.(1996) J. Cell Biol. 132, 717–726.

6. McKie, N., Dallas, D. J., Edwards, T., Apperley, J. F., Russell,R. G. G., and Croucher, P. I. (1996) Biochem. J. 318, 459–462.

7. Smith, J. W., Vestal, D. J., Irwin, S. V., Burke, T. A., andCheresh, D. A. (1990) J. Biol. Chem. 265, 11008–11013.

8. Hu, D. D., Lin, E. C. K., Kovach, N. L., Hoyer, J. R., and Smith,J. W. (1995) J. Biol. Chem. 270, 26232–26238.

9. Nath, D., Slocombe, P. M., Webster, A., Stephens, P. E., Do-cherty, A. J. P., and Murphy, G. (2000) J. Cell Sci. 113, 2319–2328.

0. Chen, M. S., Almeida, E. A. C., Huovila, A.-P. J., Takahashi, Y.,Shaw, L. M., Mercurio, A. M., and White, J. M. (1999) J. CellBiol. 144, 549–561.

1. Pentikainen, O., Hoffren, A.-M., Ivaska, J., Kapyla, J., Nyronen,T., Heino, J., and Johnson, M. S. (1999) J. Biol. Chem. 274,31493–31505.

2. Iba, K., Albrechtsen, R., Gilpin, B., Frohlich, C., Loechel, F.,Zolkiewska, A., Ishiguro, K., Koyima, T., Liu, W., Langford,J. K., Sanderson, R. D., Brakebusch, R., Fassler, R., and Wewer,U. M. (2000) J. Cell Biol. 149, 1143–1155.

mdc-9 (adam-9/meltrin γ) functions as an adhesion molecule by binding the αvβ5 integrin

Documents