MDC-L, a Novel Metalloprotease Disintegrin Cysteine-rich Protein Family Member Expressed by Human Lymphocytes*

The metalloprotease disintegrin cysteine-rich (MDC) proteins are a recently identified family of transmembrane proteins that function in proteolytic processing of cell surface molecules and in cell adhesion. Since lymphocytes must interact with a constantly changing environment, we hypothesized that lymphocytes would express unique MDC proteins. To identify MDC proteins expressed in human lymph node, a polymerase chain reaction-based strategy combined with degenerate oligonucleotide primers was employed. We report here the identification of MDC-L (ADAM 23), a novel member of the MDC protein family. The results obtained from cDNA cloning and Northern blot analysis of mRNA isolated from various lymphoid tissues indicate that a 2.8-kilobase mRNA encoding a transmembrane form, MDC-Lm, and a 2.2-kilobase mRNA encoding a secreted form, MDC-Ls, are expressed in a tissue-specific manner. MDC-L mRNA was shown to be predominantly expressed in secondary lymphoid tissues, such as lymph node, spleen, small intestine, stomach, colon, appendix, and trachea. Furthermore, immunohistochemical staining with an anti-MDC-L antibody demonstrated that cells with typical lymphocyte morphology are responsible for expression of the MDC-L antigen in these lymphoid tissues. MDC-Lm was found to be expressed on the surface of human peripheral blood lymphocytes and transformed B- and T-lymphocyte cell lines as an 87-kDa protein. Thus, we have identified a novel lymphocyte-expressed MDC protein family member.

MDC, 1 or ADAMs, are a family of transmembrane glycoproteins with a unique domain structure (1,2). The prototypical MDC protein comprises pro, metalloprotease, disintegrin, cys-teine-rich, EGF, transmembrane, and cytoplasmic domains. The metalloprotease domain is homologous to the reprolysins, members of the zinc binding metzincin superfamily that also includes the matrix metalloproteases (MMPs), the astacins, and serralysins (3). The prodomain regulates the activity of the metalloprotease by blocking access to the zinc ion; removal of the prodomain or inactivation of a conserved cysteine thiol group results in a gain of proteolytic activity. The disintegrin domains of MDC proteins are homologous to small nonenzymatic peptides isolated from the venom of snakes. Snake venom disintegrin peptides interfere with platelet aggregation by inhibiting binding of fibrinogen to the integrin ␣ IIb ␤ 3 (4). The functions previously attributed to the individual domains suggests roles for MDC proteins in cell adhesion and the proteolysis of extracellular proteins.
The first mammalian MDC proteins identified were the fertilins (5). These proteins function in the attachment and fusion of the sperm and egg during fertilization. A similar cell fusion function for the MDC protein meltrin-␣ (MDC-12) was shown for the formation of multinucleated myotubes (6). These MDC proteins have, in addition to the domains described above, a fusion peptide-like sequence not found in all members of the MDC protein family. However, the interaction of the fertilin disintegrin domain on the sperm surface with the integrin ␣ 6 ␤ 1 on the surface of the egg is also required for membrane adhesion and fusion (7)(8)(9). Additional evidence for the functionality of the disintegrin domain comes from studies using recombinantly expressed metargidin (MDC-15) disintegrin domain, which was shown to specifically interact with the integrin ␣ v ␤ 3 (10). Thus, the disintegrin domains of MDC proteins appear to be a new family of integrin ligands.
More recently, MDC proteins have been shown to function in ectodomain shedding of several cell surface proteins. Tissue necrosis factor-␣ converting enzyme (TACE; MDC-17) was the first mammalian MDC protein shown to have secretase activity (11)(12)(13). Additionally, meltrin-␥ (MDC-9) is involved in ectodomain shedding of membrane anchored heparin-binding EGFlike growth factor, and considerable evidence supports the proteolytic processing of notch by the Drosophila MDC protein Kuzbian (14 -17). Many other proteins are shed from cell surfaces by metalloproteases, implicating other MDC proteins in proteolytic processing (18). Thus, MDC proteins may function in a variety of processes. First, they may control the release of ligands from the cell surface as in the case of tissue necrosis factor-␣. Second, MDC proteins may rapidly regulate adhesive events by cleaving receptors or counter receptors from the cell surface, such as in the shedding of L-selectin (19,20). Third, the similarity between membrane-bound MMPs and MDC proteins suggests a role in extracellular matrix proteolysis. Finally, MDC protein themselves may function as ligands for integrin receptors.
* This work was supported by National Institutes of Health Grant GM51616 and the Oklahoma Center for the Advancement of Science and Technology. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The work presented here describes MDC-L, a novel member of the MDC protein family, expressed by lymphocytes in the tissues examined. Two transcripts, one encoding a prototypical MDC protein and the other encoding a secreted form, demonstrated tissue specific regulation. The transmembrane form of MDC-L was expressed as an 87-kDa protein on the surface of peripheral blood lymphocytes and both B-and T-lymphocyte cell lines. We hypothesize that MDC-L may play a role in the adhesive and proteolytic events that occur during lymphocyte emigration or function in ectodomain shedding of lymphocyte proteins such as FasL, CD40L, or an as yet unidentified lymphocyte protein (21,22).
Nested PCR-Manipulation of recombinant DNA was by standard techniques (24). Restriction enzymes, T4 DNA ligase, and Taq polymerase were purchased from Roche Molecular Biochemicals (Indianapolis, IN). Oligonucleotides were synthesized by the Molecular Biology Resource Facility at the University of Oklahoma Health Sciences Center. The degenerate oligonucleotide primers used were B1 (GARGGNG-ARGAYTGYGAYTG), B2 (GGNGARGAYTGYGAYTGYGG), C1 (CART-AYTCNGGNARRTCRCA), and C2 (TAYTCNGGNARRTCRCAYTC) in which N signifies any deoxynucleotide, R signifies either A or G, and Y signifies either T or C (25). Nested PCR was performed as follows. For reaction 1, 500 ng of human lymph node cDNA (CLONTECH, Palo Alto, CA) was combined with 0.5 g of primers B1 and C1 in a hot start PCR using the AmpliWax (Perkin Elmer) protocol. The PCR products were then separated from the oligonucleotides by centrifugal filtration. For reaction 2, the reaction 1 products were combined with 0.5 g of primers B2 and C2 in a hot start PCR protocol. The PCR products obtained from reaction 2 were subcloned into pCRII-TOPO (Invitrogen, Carlsbad, CA) and the nucleotide sequence of both strands determined.
cDNA Cloning-A 190-bp EcoRI fragment from a pCRII-TOPO MDC-L subclone was electroeluted from a 5% polyacrylamide gel (PAGE) and labeled with [␣-32 P]dCTP (Amersham Pharmacia Biotech) by the random primer method. After removal of the unincorporated nucleotides, the radiolabeled 190-bp EcoRI fragment was used to probe a gt10 human lymph node cDNA library (CLONTECH). Positive plaques were isolated, plated at a lower density, and probed until all the plaques on a plate were positive. The cDNA from the positive phage were subcloned into pCRII-TOPO vector after PCR of a phage lysate with gt10 forward and reverse primers (26). The nucleotide sequence of both strands was determined using universal and gene specific primers. The 5Ј rapid amplification of cDNA ends (RACE) was performed using Marathon-Ready cDNA (CLONTECH) as recommended by supplier with the primers AP1 and the gene-specific primer MET12 (CT-GTCCATCCCGATGTATGGGGC). The PCR products obtained were separated from the primers by centrifugal filtration and subjected to a second PCR with the primers AP2 and MET12. The 600-bp product obtained was subcloned into pCRII-TOPO and both strands sequenced using universal and gene-specific primers.
Messenger RNA Analysis-The multiple tissue northern and mRNA dot blot were purchased from CLONTECH. Hybridization of the mRNA dot blot was performed with the same probe used for cDNA cloning. The MDC-Lm-specific probe was a 790-bp BamHI/SalI (vector site) fragment from the 3Ј end of MDC-Lm cDNA that had been extracted from an agarose gel. The MDC-Ls-specific probe was a HindIII fragment from the 3Ј-untranslated region of MDC-Ls cDNA. All probes were radiolabeled with [␣-32 P]dCTP by the random primer method to a specific activity greater than 1 ϫ 10 9 cpm/g. Polyadenylated RNA transferred to nylon membranes were hybridized with the specified probe as described (27). The nylon membranes were exposed to x-ray film at Ϫ80°C for 3 days. After exposure of the nylon membranes to film, the bound probe was stripped by heating at 95°C for 10 min in 0.5% SDS, rinsing in H 2 O, and storing at Ϫ20°C.
Purification of MDC-L Fusion Proteins-The MDC-L disintegrin domain (residues Gly 418 -Glu 476 ) was PCR-amplified from the gt10 subclone 2.2.5 with the oligonucleotides MDC-01 (CGGGATCCGGTGAGG-ACTGCGACTGCGGG) and MDC-02 (AACTGCAGTTATTCAGGCAG-GTCGCACTC), digested with BamHI and PstI, subcloned into the hexahistidine (H6) expression vector pQE30 (Qiagen Inc., Valencia, CA), and transformed into Eschericia coli M15[pREP4]. Cells harboring the pQE30 MDC-L disintegrin domain recombinant plasmid were grown in SB media plus 100 g/ml ampicillin and 50 g/ml kanamycin at 37°C to an A 600 nm of 0.5-0.7. Isopropyl-␤-D-thiogalactopyranoside was added to a final concentration of 2 mM, and the cells were further incubated at 37°C for 2 h. The cell culture was centrifuged at 4,000 ϫ g for 10 min, and the pellet was resuspended in sonication buffer (50 mM sodium phosphate, pH 7.8, 300 mM NaCl, 5 mM 2-mercaptoethanol) and stored at Ϫ80°C. The thawed cell suspension was sonicated, centrifuged at 11,000 ϫ g for 20 min at 4°C, the supernatant diluted 1:3 in sonication buffer, and affinity-purified on Ni-NTA agarose. Material bound to the Ni-NTA-agarose was washed with 50 mM sodium phosphate, pH 6.0, 300 mM NaCl, 5 mM 2-mercaptoethanol, 10% glycerol and then eluted with the same buffer at pH 4.0. The purified fusion protein was dialyzed in phosphate-buffered saline (PBS) and examined by 15% Tris-Tricine PAGE. The correct amino acid composition was verified by electrospray mass spectrometry. The MDC-Lm EGF domain (residues Thr 627 -Ser 666 ) was PCR amplified from the gt10 subclone 4.2.2 with the oligonucleotides MDC-E1 (TCAGAATTCACCAATTGCTCATCCA-AG) and MDC-E2 (CAATCTAGACTAGGAGAAGTGGAAGACCAC), digested with EcoRI and XbaI, and subcloned into the malE gene fusion vector pMALc2 (New England Biolabs, Beverly, MA). Maltose-binding protein-MDC-L fusion proteins were purified and analyzed as described (28).
Polyclonal Antiserum-For the production of polyclonal antibodies, two rabbits were hyperimmunized with either purified H6-MDC-L disintegrin domain or maltose-binding protein-MDC-L EGF domain fusion proteins following standard techniques (29). Purified immunoglobulin was obtained by protein A-Sepharose affinity chromatography (Amersham Pharmacia Biotech). Affinity-purified anti-MDC-L disintegrin domain antibody was obtained by incubating 5 mg of purified immunoglobulin with the disintegrin domain affinity resin overnight at 4°C, washing extensively with PBS, and eluting with 0.1 M glycine, pH 3.0, 0.5 M NaCl. Eluted antibody was immediately neutralized with 0.1 M Tris, pH 8.0, and then dialyzed in PBS.
Immunoprecipitations were performed by surface biotinylation of 2 ϫ 10 7 cells with 0.5 mg of sulfo-NHS-biotin (Pierce) in PBS for 2 h.
The cells were then washed three times in PBS and lysed on ice for 30 min in 0.5 ml of radioimmune precipitation buffer (50 mM Tris, pH 8.0, 150 mM NaCl, 1% Triton X-100, 0.5% deoxycholic acid, 0.1% SDS). After centrifugation at 16,000 ϫ g for 10 min, 100 l of the supernatant was incubated with 5 l of protein A-Sepharose pre-armed with the specified antiserum for 1 h on ice. The resin was washed three times with radioimmune precipitation buffer, boiled in SDS-PAGE loading buffer, centrifuged, subjected to SDS-PAGE, transferred to nitrocellulose, and biotinylated proteins detected as described above.
Immunohistochemistry-Human tissues were obtained from either normal portions of surgical specimens or from autopsy specimens at University Hospital and Children's Hospital of the University of Oklahoma Health Sciences Center and were procured according to institutional guidelines. Tissues were fixed in 10% paraformaldehyde, processed, paraffin-embedded, and sectioned using conventional techniques. Snap-frozen tonsil and stomach were used to evaluate possible antigen loss when studying paraformaldehyde-fixed, paraffin-embedded tissues. Tissue sections were treated as described (31). Briefly, deparaffinized sections were treated with 1.25% H 2 O 2 in methanol for 30 min to block endogenous peroxidase activity. Immunohistochemical staining for MDC-L antigen was performed using standard strepavidin biotin peroxidase methodology at room temperature. The slides were incubated with PBS containing 5% swine serum to inhibit nonspecific antibody binding, followed by addition of the primary affinity-purified antibody diluted in PBS containing 1% bovine serum albumin for 60 min. Biotin-conjugated swine anti-rabbit secondary antibody was applied for 20 min, followed by strepavidin-horseradish peroxidase(Dako Corp., Carpinteria, CA) for 30 min. Diaminobenzidine (Sigma) was used as chromogen. Hematoxylin was used for nuclear counterstaining.
General Procedures-Protein concentrations were determined by the BCA assay (Pierce). The MDC-L disintegrin domain affinity column was generated by cross-linking 2.6 mg of H6-MDC-L disintegrin domain in 0.1 M NaHCO 3 , pH 8.0, 0.5 M NaCl to 0.3 g of CNBr-activated Sepharose. The affinity matrix was then incubated in 0.1 M Tris, pH 8.0, 500 mM NaCl for 2 h at room temperature, washed several times with alternating 0.1 M glycine, pH 3.0, 0.5 M ,NaCl and 0.1 M Tris, pH 8.0, 0.5 M NaCl, and stored in PBS at 4°C. Data base searches were performed using the National Center for Biotechnology Information's BLAST sequence similarity searching program. Molecular modeling were performed using the SWISS-MODEL protein modeling server (32).

Identification of a Novel Disintegrin Domain from Human
Lymph Node-To determine whether lymphocytes express any MDC protein family members, human lymph node cDNA was subjected to two consecutive rounds of PCR with two sets of internally nested degenerate oligonucleotide primers. These degenerate primers were based on the coding sequence of conserved regions within the disintegrin domains of MDC proteins (25). A 165-171-bp product was obtained after the secondary PCR (Fig. 1a). The PCR products were subcloned and the nucleotide sequences determined. Identified among the PCR subclones were sequences encoding the disintegrin domains of human MDC-9 (6, 34) and MDC-20 (35) (Fig. 1b). However, 8 of the 10 inserts examined possessed an open reading frame encoding a unique polypeptide with significant homology to many of the cellular and snake venom disintegrins (Fig. 1b). Data base searches using GenBank BLAST (33) indicated that this novel disintegrin domain was greater than 60% identical to the disintegrin domains of human sperm maturation-related glycoprotein GP-83 and the macaque epididymal apical protein I (GenBank accession nos. AF090327 and X66139, respectively). These results suggest that a novel MDC family member, which we have designated MDC-L (ADAM 23), may be expressed within the human lymph node.
Cloning of MDC-L cDNA-Using the isolated cDNA encoding the putative MDC-L disintegrin domain as a probe, a human lymph node cDNA library (6 ϫ 10 5 plaques) was screened and 29 positive phage were isolated. The cDNA from eight phage that contained inserts larger than 1 kb were subcloned and the nucleotide sequence determined. None of these overlapping partial cDNAs contained the 5Ј end of the MDC-L cDNA. Therefore, the missing MDC-L cDNA was acquired by 5Ј-RACE reactions using human lymph node cDNA. The cDNA clones obtained resulted in the reconstruction of two distinct full-length cDNAs with identical 5Ј nucleotide sequences through bp 1608; however, the two cDNAs diverged beyond this point (Fig. 2). Neither cDNA sequence was found in data base searches using GenBank BLAST. The 2,786-bp cDNA form contained an open reading frame encoding a polypeptide with the typical domain structure of other MDC family members, including a signal peptide, as well as pro, metalloprotease, disintegrin, cysteine-rich, EGF, and transmembrane domains (1,25,36,37) (Fig. 2). We have designated this putative transmembrane protein MDC-Lm (Fig. 3). The 2,087-bp cDNA form contains an open reading frame encoding a protein with a signal peptide, pro, metalloprotease, and disintegrin domains identical to the deduced MDC-Lm protein; however, the cysteine-rich domain differed in sequence and contained a stop codon midway through (Fig. 2).The putative secreted form of this protein has been designated MDC-Ls (Fig. 3). Both MDC-L cDNA forms contain consensus Kozak (38) and polyadenylation signal sequences (39), suggesting that these mRNAs are synthesized and translated (Fig. 2).
Examination of the deduced MDC-L amino acid sequences revealed that the metalloprotease domain has the sequence H 339 EMGHNFGMFHD 350 and C 361 VMDK 365 , which fits the consensus sequences for zinc binding, catalytic activity, and the structurally important Met turn motif (3). Molecular modeling demonstrated that the histidines required for zinc ion coordination, catalytic glutamic acid, and Met turn could all be placed correctly within the tertiary structure (data not shown). The prodomain also possessed a "cysteine switch" pattern, S 167 TCGM 171 , typically found in MMPs (40). The sequence K 195 DRK 198 exists at the predicted boundary between the prodomain and the metalloprotease domain. This sequence does not fit with the furin-like enzyme cleavage site identified in some MDC and MMP family members (40). The disintegrin domain has the sequence K 469 DEC 472 within the putative integrin recognition loop (2) (Fig. 2). Additionally, the deduced MDC-Lm protein has six potential N-linked glycosylation sites, while the MDC-Ls form has three potential N-linked glycosylation sites. The motifs identified suggest that MDC-L has a catalytically active metalloprotease and a non-RGD integrin recognition site.
MDC-L mRNA Expression-Messenger RNA from several lymphoid tissues were examined for expression of MDC-Lm and MDC-Ls (Fig. 4). As shown in Fig. 4a, an mRNA of 2.8 kb in lymph node and spleen hybridized to the MDC-Lm probe. Longer exposures also revealed a 2.8-kb mRNA band in the peripheral blood leukocyte mRNA (data not shown). When the same blot was screened with a probe specific for MDC-Ls, a 2.2-kb band was seen in spleen mRNA, but not in lymph node or the other tissues examined (Fig. 4a). Since the MDC-Ls cDNA was originally isolated from lymph node, mRNA from this tissue was used in a more sensitive RT-PCR analysis using one primer common to both cDNAs and another specific for each MDC-L mRNA. As shown in Fig. 4b, PCR products of the expected size were produced in both reactions, and their identity was verified by cleavage at restriction sites unique to each MDC-L cDNA (results not shown). As predicted from the Northern blot results, the MDC-Lm mRNA appeared to be more prevalent. These results indicate that two forms of MDC-L mRNA are expressed in a tissue-specific manner.
The results obtained above prompted us to further examine the tissue specificity of MDC-L expression. Polyadenylated RNA isolated from 50 human tissues were analyzed by mRNA dot blotting with a probe common to both MDC-L forms. MDC-L mRNA was predominantly expressed in secondary lymphoid tissues (Fig. 5). No significant expression was seen in any of the fetal tissues or controls, with the exception of the E. coli DNA control. A comparison of the MDC-L cDNA sequences with the E. coli genome identified a 23-bp region of untranslated E. coli DNA that identically matched 23 bp from the MDC-L probe used. Thus, we believe that the hybridization of the MDC-L probe with the E. coli DNA was artifactual. Since MDC-L mRNA appeared to be specific for secondary lymphoid tissues, we suspected that it was the lymphocyte population in these tissues that express MDC-L.
Lymphocyte Expression of MDC-L-Peripheral blood leukocytes were further analyzed for the presence of MDC-Lm and MDC-Ls mRNA. Biomagnetic separation was used to isolate anti-CD2, CD19, CD14, and CD16 positive cells from normal whole blood. Each of these was then analyzed by nested RT-PCR for the presence of MDC-Lm and MDC-Ls mRNA (Fig.  6a). Both T-and B-lymphocytes, CD2 and CD19 positive cells, respectively, expressed MDC-Lm and MDC-Ls mRNA. MDC-L mRNA was also seen in the CD14 and CD16 positive populations. Although Northern blot analysis suggested that MDC-Lm mRNA was synthesized in peripheral blood leukocytes at low levels, results from RT-PCR indicate that mRNA from both forms of MDC-L are expressed.
MDC-L protein expression in peripheral blood lymphocytes was examined by immunoblotting whole cell lysates (200 g/ lane) with an anti-MDC-L disintegrin domain polyclonal serum. Under non-reducing conditions, two bands of 87 and 67 kDa were specifically recognized (Fig. 6b). Similar results were typically obtained under reduced conditions except the 87-kDa species migrated at 92 kDa. These molecular sizes approximate those predicted from the cDNA (MDC-Lm ϭ 85,359 Da and MDC-Ls ϭ 59,427 Da). However, these results could not distinguish whether the 67-kDa protein was the MDC-Ls protein or a processed form of the MDC-Lm protein lacking the prodomain. These results did indicate that the MDC-Lm, and possibly the MDC-Ls, proteins are expressed by peripheral blood lymphocytes.
To determine which cell types within the lymphoid tissue were expressing the MDC-L protein, the anti-MDC-L disintegrin domain antiserum was used for immunohistochemical staining of sections from human lymph node, small intestine, colon, stomach, spleen, thymus, and appendix. The results for lymph node, small intestine, and stomach are shown in Fig. 7. Cells with typical lymphocyte morphology were specifically recognized by the anti-MDC-L antibody in all the tissues examined. The majority of the lymph node MDC-L positive (MDC-L ϩ ) lymphocytes were seen in the cortex and paracortex; only a few MDC-L ϩ cells were localized within the follicles (Fig.  7a). MDC-L ϩ lymphocytes were also found in white pulp of the spleen and thymus, although the percentage of lymphocytes that were MDC-L ϩ appeared to be much lower in these tissues compared with lymph node and appendix (data not shown). MDC-L ϩ lymphocytes of the small intestine and colon were found intraepithelially and within the lamina propria (Fig. 7b). Examination of stomach showed MDC-L ϩ staining of lympho-cytes in the lamina propria, muscularis mucosa, and lymphoid aggregates (Fig. 7, c and d). These results demonstrate that a fraction of the lymphocyte population was specifically recognized by the anti-MDC-L.

Expression of MDC-L by T-and B-lymphocyte Cell Lines-
Since the RT-PCR and immunohistochemical studies indicated that both T-and B-lymphocytes were MDC-L ϩ , the T-cell lines HuT78 (␣␤ T-cell receptor) and A139.1 (␥␦ T-cell receptor), as well as the B-cell line RD105U were examined for the expression of MDC-L. A protein of 87 kDa was specifically recognized in all three cell lines by an anti-MDC-L disintegrin domain antiserum in immunoblots of whole cell lysates (Fig. 8a). A single protein of 87 kDa was also specifically immunoprecipitated from surface-labeled cells (Fig. 8b). The specificity of the anti-disintegrin domain antibody was verified by immunoblotting and immunoprecipitation of an 87-kDa protein with a polyclonal antibody that recognizes the EGF domain of MDC-L (Fig. 8, c and d). Neither antiserum specifically recognized any protein in Western blots of the erythroleukemic cell line K562 or human umbilical vein endothelial cells (data not shown). These findings indicate that, at least in the transformed state, both T-and B-lymphocytes express the transmembrane form of MDC-L on the cell surface.

DISCUSSION
This study describes a novel member of the MDC protein family, which we have named MDC-L (ADAM 23) because of its expression in human lymphocytes. Two distinct transcripts that encoded a transmembrane, MDC-Lm, and potentially secreted, MDC-Ls, form of the protein were identified. MDC-L mRNA was predominantly localized to secondary lymphoid tissues, and further examination demonstrated that the lymphocyte population was responsible for MDC-L expression in these tissues.
A PCR approach using degenerate oligonucleotides was used to identify MDC protein family members expressed in human lymph node. Disintegrin domain sequences encoded by MDC-9 and MDC-20 were found; however, the overwhelming majority of the inserts encoded the disintegrin domain of MDC-L. Although it is tempting to propose that MDC-L is the major MDC protein expressed by lymphocytes, PCR approaches such as this may tend to select for a particular cDNA species. Therefore, other disintegrin domain containing proteins may not have shown up in our study. Two MDC family members have been shown to be expressed within lymph node: decysin and TACE (41,42). Decysin, which is expressed by germinal center dendritic cells, possesses a truncated disintegrin domain and does not have a typical MDC protein domain structure. The disintegrin domain of TACE lacks the carboxyl-terminal conserved residues incorporated in the design of the degenerate oligonucleotides. Thus, TACE would not have been expected to produce a PCR product under the experimental conditions used in this study (11,12). It is not surprising MDC-9 was identified in this study, since this protein is a widely expressed secretase (14). Interestingly, MDC-20 was previously reported as a testisspecific protein (35). The presence of MDC-20 observed here may be due to genomic DNA contamination of the cDNA library or expression of MDC-20 at low levels within other tissues such as human lymph node.
MDC-Lm possessed a prototypical MDC or ADAM family domain structure. Consensus sequences for zinc binding and catalytic activity were present, suggesting that MDC-L may have a proteolytic function. Several MDC proteins have been shown to have an active metalloprotease. MDC-9, TACE, and KUZ function in ectodomain shedding (11,12,14,16), MADM (MDC-10) has type IV collagenase activity (43,44), and meltrin-␣ (MDC-12) recognizes ␣ 2 -macroglobulin (45). The metalloprotease domain of MDC-L was most homologous to the hemorrhagic metalloproteases fibrolase and jararhagin found in the venom of Agkistrodon contortrix and Bothrops jararaca, respectively (46,47). Since both of these snake venom metalloproteases have fibrinolytic activity, the MDC-L metalloprotease may have a MMP-like activity. However, a role in ectodomain shedding of a lymphocyte surface target proteins, such as FasL or CD40L, has not been ruled out. MDC-L also possessed a consensus sequence for a cysteine switch found to regulate proteolytic activity in some MDC and MMP protein family members. However, MDC-L did not contain a consensus cleavage sequence for a furin-type endopeptidase at the prodomainmetalloprotease boundary. Several proteolytically active MMPs also lack a consensus sequence for furin cleavage, suggesting that another mechanism for removal of the prodomain may exist for MDC-L (40). The sequence motifs noted here suggest that MDC-L may be synthesized as a zymogen that becomes enzymatically active upon maturation.
Snake venom disintegrin peptides function as antagonists of integrin receptor function (4,48). The integrin recognition site in several of these disintegrin peptides has been localized to an RGD sequence within an extended loop of the peptide. Integrin recognition of mammalian MDC disintegrin domains has only been demonstrated for MDC-15 and the fertilins (7, 9, 10). In the case of MDC-15, an RGD sequence located within a region similar to that of the snake venom peptides is recognized by the integrin ␣ v ␤ 3 (10). A synthetic peptide derived from the same region within the mouse fertilin-␤ disintegrin domain, AQDE, has been shown to inhibit sperm-egg adhesion and cell fusion (7,8). Evidence suggests that this peptide is blocking sperm binding to the mouse egg integrin ␣ 6 ␤ 1 (9). MDC-L contains a sequence at this site, AKDE, similar to that found in fertilin-␤, suggesting the possibility that MDC-L may also be recognized by the integrin ␣ 6 ␤ 1 .
Two distinct MDC-L transcripts were identified in this study. These may arise from alternative splicing or gene duplication. The precedent for alternative splicing of another MDC protein family member exists. Similar to the results shown here for MDC-L, MDC-12 is also expressed as transmembrane and secreted forms (49). MDC-L and MDC-12 differ with respect to the point of divergence between the transmembrane and secreted forms. The two forms of MDC-L diverge midway through the cysteine-rich domain, whereas the two MDC-12 forms diverge at the start of the transmembrane domain. MDC-Lm and MDC-Ls were observed to differ in tissue expression. MDC-Ls was preferentially expressed in spleen, although it could also be detected in lymph node using RT-PCR techniques. This suggests that each of the forms of MDC-L has unique functions.
Northern, RT-PCR, and immunoblot results indicated that MDC-Lm is expressed by peripheral blood lymphocytes at low levels. Although none of these results are quantitative, the increase in mRNA expression observed in lymphoid tissues suggests that MDC-L expression may be up-regulated during or after lymphocyte extravasation. The RT-PCR results demonstrated the presence of MDC-Lm and MDC-Ls transcripts in peripheral blood T-and B-lymphocytes. In addition, MDC-Lm mRNA was present in the CD14 and CD16 positive populations. These results suggest that other leukocytes may express MDC-L. Alternatively, the CD14 and CD16 populations, which are predominantly made up of monocytes and neutrophils, respectively, may also contain B-lymphocytes and NK cells, respectively. Further studies are required to clearly quantitate MDC-Lm expression on the surface of specific peripheral blood leukocyte lineages.
MDC-Lm was expressed as an 87-kDa protein on the surface of both T-and B-lymphocyte cell lines and peripheral blood lymphocytes. The identity of MDC-Lm was confirmed by immunoblot and immunoprecipitation with a polyclonal anti-MDC-L EGF domain antiserum. Since this domain was not present in MDC-Ls, the anti-MDC-L EGF antiserum would be expected to specifically recognize the MDC-Lm protein. Only a single protein of 87 kDa was immunoprecipitated with both anti-MDC-L disintegrin domain and anti-MDC-L EGF antiserum, suggesting that MDC-Lm does not covalently associate with other lymphocyte cell surface proteins. Attempts to immunoprecipitate a secreted form of MDC-L from conditioned media or from human plasma were unsuccessful (results not shown); thus, we were unable to demonstrate translation of the MDC-Ls transcript. Peripheral blood lymphocytes did express a 62-kDa anti-MDC-L disintegrin domain reactive protein.
Whether this band was a processed form of MDC-Lm lacking the prodomain or cell-associated MDC-Ls was not determined.
At present there are 22 members of the MDC or ADAM protein family. While the in vivo function of only a few have been determined, the roles established are both diverse and important to the development and life of an organism. One of the first identified roles for MDC proteins was in sperm-egg fusion. These MDC proteins all possess a characteristic fusion peptide, which was not present in either of the MDC-L cDNA sequences (5). However, this does not rule out a role in cell adhesion for MDC-L.
The identification of an MDC protein expressed by lymphocytes suggests that MDC-L may have an important immunological function. FasL is shed from the surface of cytotoxic T-cells by an unknown metalloprotease, and sFasL has been proposed to function in protection of neighbor cells from apoptosis (21). An unknown metalloprotease is also implicated in the shedding of CD40L from the surface of T-lymphocytes (22). MDC-L may also function in a similar fashion to MMPs, which are required for emigration of lymphocytes from the vasculature. The expression pattern observed in this study indicates that MDC-L is expressed by both mature T-and B-lymphocytes. With the aid of anti-MDC-L monoclonal antibodies and the many well characterized lymphocyte surface markers, it should be possible to determine the temporal and leukocyte lineage expression pattern of MDC-L. This information should provide valuable insight into the in vivo function of this novel MDC protein.