First printed in R&D Systems' 2001 Catalog.
The ADAM (A Disintegrin And Metalloprotease) family includes proteins containing disintegrin-like and metalloprotease-like domains.1 They are also referred to as MDC (Metalloprotease, Disintegrin, Cysteine-rich) proteins.2 ADAMs are involved in diverse processes such as development, cell-cell interactions and protein ectodomain shedding (refer to references 2-7 for reviews on ADAMs). TACE (TNF-a Converting Enzyme, ADAM17) is an example of a member of the ADAM family.8,9
Gene symbols, alternative names, selected characteristics and references for the 40 members of the ADAM family identified to date are presented in Table 1. The gene symbols recommended by the Human Gene Nomenclature Committee distinguish individual members and represent the approximate order of discovery. For example, ADAM1/fertilin a and ADAM2/fertilin ß are the first two members of the family (cloned in 1992).10,11 ADAMTS# gene symbols designate a subset of ADAM proteins that contain a thrombospondin (TS) motif. For the same gene identified by different symbols, the lowest number is listed and the other symbols used are included as aliases. This is the case for ADAM18 and ADAMTS5, which also have the aliases ADAM27 and ADAMTS11, respectively.
The characteristics listed within Table 1 highlight known physiological functions and unique features of the ADAMs. Some of these characteristics, however, may be restricted to ADAMs from certain species. For example, ADAM1/fertilin a is involved in sperm-egg fusion in rodents, while the human counterpart is a pseudo gene.12 For more detailed descriptions of each ADAM, please refer to the references cited.
|Fig. 1. Domain structures of the ADAM proteins. Four domains are common in ADAM proteins: pro, metalloprotease, disintegrin-like (DisInt) and cysteine-rich (Cys-rich). The majority of ADAMs are type I integral membrane proteins, containing EGF-like, transmembrane (TM), and cytosolic domains. ADAMTS proteins are secreted, containing spacer and unique regions as well as one or more copies of the thrombospondin type I (TS) motifs..
Domain Structure and Function
All of the deduced amino acid (aa) sequences of ADAMs predict multi-domain structures (see Figure 1). A signal peptide at the N-terminus targets proteins for the secretory pathway and is removed prior to secretion from the cell or anchoring on the cell surface. Starting with the pro domain, ADAM proteins contain between 800 and 1200 aa residues. The pro domain consists of about 200 residues that separate from the metalloprotease domain by one or more furin cleavage sites.13 Furin or furin-like proprotein convertases can cleave the pro domain from several ADAM precursors including 9, 12, 15 and ADAMTS1.14-17 This process does not seem to depend on the proteolytic activity of these ADAMs because the mutation of the catalytic Glu residue does not prevent removal of the pro domain. In contrast, removal of the pro domain from ADAM28 is, at least in part, autocatalytic.18 In the pro domain of ADAMs, there is a sequence context similar to the cysteine switch motif of matrix metalloproteases (MMPs) and other reprolysins. The cysteine switch motif keeps MMPs and reprolysins in a latent form and disruption of its interaction with zinc in the active site is required in order to activate MMPs and reprolysins.19,20 The cysteine switch motif in ADAMs may play a similar role during ADAM biosynthesis, preventing them from autocatalysis and self-destruction. This conjecture is supported by the observations that the pro domain of TACE and ADAM12 both act as inhibitors of the catalytic domain and are required for secretion of functional proteases.15,21
The metalloprotease domain of ADAMs consists of about 200 aa. The structural relationship of ADAMs to other zinc MMPs is shown in Figure 2.22 ADAMs are reprolysin-like proteins, which together with MMPs, astacins and serralysins constitute the metzincin superfamily. The zinc-binding site of the reprolysin family is within the consensus sequence of HExxHxxGxxHD; three His residues and a water molecule tetrahedrally coordinate the zinc and the Glu residue acts as a catalytic base.23 The presence of an Asp residue after the third His is one of the major features that distinguish reprolysins from other metzincins that contain a Met-turn structure.24 Unlike other zinc protease families, however, ADAM proteins have two distinguishing characteristics. First, the intact zinc-binding site is absent in several ADAMs, including 2-7, 11, 14, 18, 22, and 29. Therefore, these ADAMs are not considered zinc proteases. Second, the metalloprotease domain is not retained in several mature ADAMs. For example, both a and ß subunits of fertilin (ADAMs 1 and 2) lack the protease domain in their mature forms.10 Since cleaved ADAM protease domains have not been isolated, it is unknown whether these protease domains have additional functions. Several snake venom reprolysins such as mature atrolysins, however, consist of only the protease domain.25, 26
|Fig. 2. Relationship of ADAMs to zinc metalloproteases.
The disintegrin domain of ADAMs consists of 60 to 90 aa with 6 to 15 Cys residues showing sequence similarity to the disintegrins (peptides generated from reprolysin precursors and isolated from snake venom).25 Disintegrins have an RGD integrin-binding site, bind to the platelet integrin GPIIb/IIIa (aIIb/ß3) and inhibit platelet aggregation. Specific interaction of the disintegrin domain of human ADAM15 with integrin aVß3 is RGD-dependent, indicating that this domain functions as an adhesion molecule and may be involved in aVß3-mediated cell-cell interactions.27,28 Although most ADAMs do not contain an RGD sequence in the disintegrin domain, they can still bind integrins. For example, ADAM2 binds through its disintegrin domain to a6ß1 integrin which, in turn, interacts with the tetraspan protein CD9; this process is important in sperm passage into oviducts and sperm-egg interactions.29 The conserved ECD motif of the ADAM2 disintegrin loop (especially the Asp residue) is involved in cell-cell adhesion during fertilization.30,31 A similar peptide of ADAM3/cyritestin and an antibody generated against the peptide inhibit both sperm-egg adhesion and fusion.32 The disintegrin domain of ADAMs 12 and 15 specifically binds to integrin a9ß1 in an RGD-independent manner and this binding supports cell-cell interaction.33 Cyclic and linear peptides from the disintegrin domain of ADAM16 and ADAM9 inhibit fertilization in Xenopus laevis.34 A short aa sequence (lacking any RGD motif) of the disintegrin loop in ADAM23 mediates the interaction between ADAM23 and the aVß3 integrin, indicating that ADAM23 may be important in aVß3-mediated cell-cell interactions occurring in normal and pathological processes.35
The functions of the Cys-rich and EGF-like domains of ADAMs are not well understood. Structurally, the Cys-rich and EGF-like domains consist of about 160 aa with 10 to 14 Cys residues and about 40 aa with 6 Cys residues, respectively. The two domains may be important for interactions of ADAMs with other proteins such as chaperons involved in biosynthesis and/or other partners on the cell surface. The protein-protein interactions could in turn warrant the correct targeting and efficient transport of the ADAMs and also regulate ADAM biological activities on the cell surface. This conjecture is supported by the following observations. The Cys-rich domain of TACE may play a role in the release of the pro domain and may be required for the shedding of interleukin 1 receptor type II.21,36 The secreted ADAM12, which is highly expressed in placental tissues, binds to insulin-like growth factor-binding protein-3 (IGFBP-3) through its Cys-rich domain and has IGFBP-3 protease activity.37 Within the Cys-rich domain of ADAMs 1, 9, 11 and 12, there is a hydrophobic stretch of about 23 aa that has been termed as the potential fusion peptide because its predicted secondary structure is similar to that of viral fusion proteins.5
Many ADAMs are type I membrane proteins that are anchored through a transmembrane (TM) domain near the C-terminus. Several of these ADAMs also have an alternatively spliced form that diverges before the TM domain, leading to the production of a soluble, secreted form. These ADAMs include 11, 12, 17, and 28.38-41 It is not known whether all the membrane-anchored ADAMs have a soluble counterpart generated through either alternative splicing or shedding from the cell surface. All of the ADAMTSs lack a TM domain and are secreted proteases. Having both soluble and membrane-anchored forms allows ADAMs to regulate events not only on or near the cell surface, but also at a distance from cells.6
The cytosolic portion of ADAMs is variable in length (between 40 to 250 aa). Due to its considerable size and noticeable motifs, the cytosolic domain of ADAM may transmit signals between the interior and exterior of cells. For example, ADAM9 and ADAM15 interact with two SH3 domain-containing proteins, endophilin I and SH3PX1, which may have a role in regulating the function of ADAM9 and ADAM15 by influencing their intracellular processing, transport or final subcellular localization.42 Binding of the cytosolic domain of ADAM12, a marker of skeletal muscle regeneration, to the muscle-specific actin-binding protein, a-actinin-2, is required for myoblast fusion.43 The cytosolic and TM domains of ADAM12 also contain a signal for retention in the trans-golgi network.44 Similar signals may also be present in other ADAMs, such as ADAM15, that have been shown to reside at intracellular compartments in addition to the cell surface.16
ADAMTSs contain one or more copies of the thrombospondin (TS) type 1 motif that are conserved in thrombospondin 1 (TS1) and TS2. TS1 and TS2 are two multifunctional extracellular matrix (ECM) proteins that influence cell adhesion, motility, and growth. The TS domain of ADAMTS4 is critical for aggrecan substrate recognition and cleavage.45 The TS domain of ADAMTS1 is capable of binding heparin, and together with the spacer region is important for a tight interaction of ADAMTS1 with the ECM.46, 47 The spacer region of ADAMTS1 is also necessary to degrade aggrecan.48 However, the functions of the spacer and unique regions have not been studied for the majority of ADAMTSs.
Substrates and Inhibitors
ADAMs 1, 8-10, 12, 13, 15-17, 19-21, 24-26, 28, 30, 31 and ADAMTSs 1-9 contain the intact zinc-binding site in the metalloprotease domain. These ADAMs (or their processed forms) most likely function as zinc proteases. The majority of these proteases were initially identified through molecular cloning or genetic analysis rather than biochemical characterizations. As a result, the natural substrates and inhibitors for these enzymes remain to be identified. Natural substrates are presently known for only a handful of the ADAMs. They include: TNF for ADAM17/TACE, NOTCH for ADAM10/KUZ, procollagens I and II for ADAMTS2/procollagen N-proteinase (PNPI), aggrecan for ADAMTS4/aggrecanase-1 and ADAMTS5/aggrecanase-2.
This limited list of known natural substrates reflects broad and diverse activities of ADAMs. TACE cleavage of the membrane-anchored TNF precursor generates soluble TNF-a, an important cytokine involved in inflammation.49 KUZ processing of the cell surface receptor NOTCH separates the extracellular and intracellular portions of the receptor, a step required for NOTCH-mediated lateral inhibition during Drosophila and vertebrate neurogenesis.50 PNPI cleavage of procollagens I and II is essential for the formation of collagen fibers. Deficiencies of PNPI activity in vivo cause several connective tissue disorders characterized by a severe skin fragility such as dermatosparaxis in cows and sheep and type VIIC Ehlers-Danlos in humans due to the accumulation of procollagen and formation of abnormal collagen fibers.51 Aggrecanase processing of aggrecan, the major proteoglycan of cartilage, is fundamental for maintaining the compressibility and stiffness of cartilage.52
Several ADAM family members have proteolytic activity in vitro. ADAM9 cleaves the insulin B chain and several synthetic peptides.14 ADAM10 is able to process pro-TNF-a and type IV collagen.53,54 ADAM12 can cleave a2-macroglobulin.55 ADAM17 processes the extracellular portion of the NOTCH 1 receptor.56 ADAMs 9, 10 and 17 are also potential a-secretases that can cleave amyloid precursor protein (APP).6,57 The autocatalytic activity of ADAM28 is, at least in part, responsible for its prodomain removal and protein maturation.18 ADAMTS1 processes aggrecan and forms a covalent-binding complex with a2-macroglobulin that is dependent on the zinc-binding site.17,48 In addition to aggrecan, brevican is also processed by ADAMTS4/aggrecanase-1. The different cleavage sites in brevican by ADAMTS4/aggrecanase-1 and MMPs are similar to those enzyme-specific sites in aggrecan, indicating that aggrecanases and MMPs may work in concert for the physiological turnover of both brevican and aggrecan.58
The crystal structure of the protease domain of human TACE shows that in comparison to MMPs, the active site cleft is similar but secondary structure differs.59 The structural similarities and differences between TACE and MMPs may explain why some MMP inhibitors such as TIMP-3 inhibit TACE, while others such as TIMP-1 do not. Similarly, the proteolytic activity of ADAM10 is inhibited by TIMP-1, -3 and hydroxamate inhibitors, but not by TIMP-2 and -4.60 It is important to note that contribution of additional domains of ADAMs to interactions with substrates or inhibitors cannot be predicted from these studies in which the protease domains instead of the full-length molecules were used.
Cell Fusion and Adhesion
The roles of fertilin and cyritestin during fertilization illustrate the importance of ADAM proteins in cell-cell fusion and adhesion. Sperm from fertilin ß knockout mice is deficient in sperm-egg membrane adhesion, sperm-egg fusion, migration from the uterus into the oviduct, and binding to egg zona pellucida.61 Mouse egg integrin a6ß1 functions as a sperm receptor that interacts with the disintegrin domain of fertilin ß in a sequence-specific manner.31,62 Cyritestin functions with fertilin ß in sperm-egg plasma membrane adhesion and fusion.32
The specific functions of the following ADAMs may reflect the combination of cell adhesion and protease activities. ADAM10/Kuzbanian is essential for the partitioning of neural and non-neuronal cells during development of both the central and peripheral nervous systems in Drosophila and is required for cells to receive signals inhibiting the neural fate.63 SUP-17, the potential ADAM10 in C. elegans, facilitates LIN-12/NOTCH signaling by acting on or in concert with the extracellular domain of LIN-12.64 In Xenopus, ADAM13 (most closely related to ADAM12) may be involved in neural crest cell adhesion and migration as well as myoblast differentiation.65 ADAM12/Meltrin a is required for and provokes myogenesis (myoblast fusion).39,66 ADAM19/ Meltrin ß and a may play roles in osteoblast differentiation and/or function but are not likely to be involved in osteoclast fusion.67 Human ADAM19 is an important marker for the differentiation and characterization of dendritic cells and the distinction between macrophages and dendritic cells.68 MIG-17, a C. elegans gene that has sequence similarity to ADAMTS1 yet lacks a TS domain, directs migration of distal tip cells by remodeling the basement membrane.69 Expression of ADAMTS1 is closely associated with acute inflammation.46 ADAMTS1 mRNA is up-regulated by progesterone during the ovulation process.70 ADAMTS1 is essential for normal growth, fertility, organ morphology and function based on knockout mice.71 ADAMTS1 (METH1) and ADAMTS8 (METH2) suppress FGF-2-induced vascularization in the cornea pocket assay and inhibit VEGF-induced angiogenesis in the chorioallantoic membrane assay.72
A tremendous advance in the understanding of structure and function of the ADAM family of proteins has been made within the past five years. These results have demonstrated the importance of the ADAM proteins in diverse biological processes. Studies have also raised many interesting questions that remain to be answered. For example, what are the substrate specificities of ADAMs that are functional proteases? What are the physiological regulators that activate or inhibit these proteases? How are the protease, adhesion and signaling activities of the ADAMs regulated in response to developmental, physiological and pathological stimuli? In order to address these questions and perform the necessary biochemical and genetic analyses of ADAMs, purified proteins, antibodies, cell culture and animal models will be required.
- Wolfsberg, T.G. et al. (1995) Dev. Biol. 169:378.
- Blobel, C.P. (1997) Cell 90:589.
- Wolfsberg, T.G. & J.M. White (1996) Dev. Biol. 180:389.
- Black, R.A. & J.M. White (1998) Curr. Opin. Cell Biol. 10:654.
- Wolfsberg, T.G. & J.M. White (1998) in Handbook of Proteolytic Enzymes (Barrett, A.J., Rawlings, N.D. & J.F. Woessner, ed.), pp.1310-1313, Academic Press, San Diego.
- Schlondorff, J. & C.P. Blobel (1999) J. Cell Sci. 112:3603.
- Kaushal, G.P. & S.V. Shah (2000) J. Clin. Invest. 105:1335.
- Black, R.A. et al. (1997) Nature 385:729.
- Moss, M.A. et al. (1997) Nature 385:733.
- Blobel, C.P. et al. (1992) Nature 356:248.
- Wolfsberg, T.G. et al. (1993) Proc. Natl. Acad. Sci. USA 90:10783.
- Jury, J.A. et al. (1997) Biochem. J. 321:577.
- Hurskainen, T.L. et al. (1999) J. Biol. Chem. 274:25555. (1999) J. Biol. Chem. 274:13427.
- Lum, L. et al. (1998) J. Biol. Chem. 273:26236.
- Kuno, K. et al. (1999) J. Biol. Chem. 274:18821.
- Howard, L. et al. (2000) Biochem. J. 348:21.
- Van Wart, H.E. & H. Birkedal-Hansen (1990) Proc. Natl. Acad. Sci. USA 87:5578.
- Grams, F. et al. (1993) FEBS Lett. 335:76.
- Milla, M.E. et al. (1999) J. Biol. Chem. 274:30563.
- Hooper, N.M. (1996) in Zinc Metalloproteases in Health and Disease (Hooper, N.M. ed), pp. 1-21, Taylor & Francis, Bristol, PA.
- Jiang, W. & J.S. Bond (1992) FEBS Lett. 312:110.
- Bode, W. et al. (1993) FEBS Lett. 331:134.
- Fox, J.W. & J.B Bjarnason (1996) in Zinc Metalloproteases in Health and Disease (Hooper, N. M. ed), pp. 47-81, Taylor & Francis, Bristol, PA.
- Bjarnason, J.B. & J.W. Fox (1998) in Handbook of Proteolytic Enzymes (Barrett, A.J., Rawlings, N.D., & J.F. Woessner, ed.), Academic Press, San Diego, pp.1247-1254.
- Zhang, X.-P. et al. (1998) J. Biol. Chem. 273:7345.
- Kratzschmar, J. et al. (1996) J. Biol. Chem. 271:4593.
- Chen, M.S. et al. (1999) Proc. Natl. Acad. Sci. USA 96:11830.
- Zhu, X. et al. (2000) J. Biol. Chem. 275:7677.
- Bigler, D. et al. (2000) J. Biol. Chem. 275:11576.
- Yuan, R. et al. (1997) J. Cell. Biol. 137:105.
- Eto, K. et al. (2000) J. Biol. Chem. Paper in Press.
- Shilling, F. M. et al. (1997) Dev. Biol. 186:155.
- Cal, S. et al. (2000) Mol. Biol. Cell. 11:1457.
- Reddy, P. et al. (2000) J. Biol. Chem. 275:14608.
- Shi, Z. et al. (2000) J. Biol. Chem. 275:18574.
- Emi, M. et al. (1993) Nat. Genet. 5:151.
- Gilpin, B.J. et al. (1998) J. Biol. Chem. 273:157.
- Cerretti, D.P. et al. (1999) Cytokine 11:541.
- Roberts, C. M. et al. (1999) J. Biol. Chem. 274:29251.
- Howard, L. et al. (1999) J. Biol. Chem. 274:31693.
- Galliano, M.-F. et al. (2000) J. Biol. Chem. 275:13933.
- Hougaard, S. et al. (2000) Biochem. Biophys. Res. Comm. 275:261.
- Tortorella, M.D. et al. (2000) J. Biol. Chem. 275:25791.
- Kuno, K. et al. (1997) J. Biol. Chem. 272:556.
- Kuno, K. & K. Matsushima (1998) J. Biol. Chem. 273:13912.
- Kuno, K. et al. (2000) FEBS Lett. 478:241.
- Bazzoni, F. & B. Beutler (1996) New Engl. J. Med. 334:1717.
- Pan, D. & G.M. Rubin (1997) Cell 90:271.
- Colige, A. et al. (1997) Proc. Natl. Acad. Sci. USA 94:2374.
- Tortorella, M.D. et al. (1999) Science 284:1664.
- Rosendahl, M.S. et al. (1997) J. Biol. Chem. 272:24588.
- Millichip, M.I. et al. (1998) Biochem. Biophys. Res. Comm. 245:594.
- Loechel, F. et al. (1998) J. Biol. Chem. 273:16993.
- Brou, C. et al. (2000) Mol. Cell 5:207.
- Duxbaum, J.D. et al. (1998) J. Biol. Chem. 273:27765.
- Nakamura, N. et al. (2000) J. Biol. Chem. Paper in Press.
- Maskos, K. et al. (1998) Proc. Natl. Acad. Sci. USA 95:3408.
- Amour, A. et al. (2000) FEBS Lett. 473:275.
- Cho, C. et al. (1998) Science 281:1857.
- Almeida, E.A.C. et al. (1995) Cell 81:1095.
- Rooke, J. et al. (1996) Science 273:1227.
- Wen, C. et al. (1997) Development 124:4759.
- Alfandari, D. et al. (1997) Dev. Biol. 182:314.
- Yagami-Hiromasa, T. et al. (1995) Nature 377:652.
- Inoue, D. et al. (1998) J. Biol. Chem. 273:4180.
- Fritsche, J. et al. (2000) Blood 96:732.
- Nishiwaki, K. et al. (2000) Science 288:2205.
- Robker, R. L. et al. (2000) Proc. Natl. Acad. Sci. 97:4689. <
- Shindo, T. et al. (2000) J. Clin. Invest. 105:1345.
- Vazquez, F. et al. (1999) J. Biol. Chem. 274:23349.
- Perry, A. C.F. et al. (1995) Biochem. J. 307:843.
- Adham, I. M. et al. (1998) DNA Cell Biol. 17:161.
- Linder, B. & U.A.O. Heinlein (1997) Dev. Growth Diff. 39:243.
- Yashida, S. et al. (1990) Int. Immunol. 2:585.
- Yoshiyama, K. et al. (1997) Genomics 41:56.
- Weskamp, G. et al. (1996) J. Cell Biol. 132:717.
- Glynn, P. (1998) in Handbook of Proteolytic Enzymes (Barrett, A.J., Rawlings, N.D., & J.F. Woessner,ed.), pp.1314-1315, Academic Press, San Diego.
- Podbilewicz, B. (1996) Mol. Biol. Cell 7:1877.
- Peschon, J.J. et al. (1998) Science 282:1281.
- Lum, L. et al. (1999) J. Biol. Chem. 274:13613.
- Rio, C. et al. (2000) J. Biol. Chem. 275:10379.
- Frayne, J. et al. (1998) Mol. Hum. Reprod. 4:429.
- Van Huijsduijnen, R.H. (1998) Gene. 206:273.
- Poindexter, K. et al. (1999) Gene 237:61.
- Zhu, G.-Z. et al. (1999) Gene 234:227.
- Liu, L. & J.W. Smith (2000) Endocrinol. 141:2033.
- Nagase, T. et al. (1997) DNA Res. 4:141.
- Matthews, R.T. et al. (2000) J. Biol. Chem. 275:22695.
- Abbaszade, I. et al. (1999) J. Biol. Chem. 274:23443.
- Clark, M.E. et al. (2000) Genomics 67:343.
- Mueller, C.G.G. et al. (1997) J. Exp. Med. 186:655.