First printed in R&D Systems' 1997 Catalog.


Figure 1. VCAM-1/ICAM-1/CD31 (PECAM-1) structural comparison. With the isolation and sequencing of the immunoglobulin light and heavy chains, it was realized that 1) repeating regions or domains of similar amino acid (aa) sequences existed in these chains; 2) each domain was flanked by two cysteine residues; and 3) each domain could be generally classified as belonging to one of two types, either a “V” (or variable) or “C” (or constant) type. Common to both domain types is the existence of two beta-sheets, with each sheet stabilized in space by a disulfide bond that links the flanking cysteines (represented in 2-dimensions by a closed “loop”). What is different between the V and C domains, however, is the makeup of the beta-sheets. The V domain beta-strands (linear aa arrays), while C-type domains have one four-beta-strand, and one three-beta-sheet. Within C-type domains, a further distinction can also be made. While C-type beta-strands have little aa sequence homology to V-type beta-strands, modest aa homology can be seen between select C2 beta-strands and analogous V beta-strands. Thus, C2-type domains seem to represent an intermediate between C1 and V-type domains. It should be noted that while the “variable” region of the immunoglobulin molecule is classified as a V-type region, this is based only on the makeup of its beta-sheets and not on the variable nature of its aa sequences.1

Additions to the immunoglobulin superfamily (IgSF), have forced reconsideration of the general function of the group, and it is perhaps instructive to consider the IgSF as a group of molecules that mediate cell surface recognition, and through such recognition, control cell behavior and activity.1 To date, there are at least 12 major subdivisions and over 100 individual molecules in the IgSF.1, 2 Within these subdivisions or categories, considerable variability exists relative to the prototypical member, the immunoglobulin molecule. In some molecules, certain V-type domains lack disulfide linkages, while many Ig-like domains co-exist with fibronectin or thrombospondin motifs and general molecular architecture can range from glycosylphosphatidylinositol (GPI)-linked molecules to molecules with single transmembrane insertion or to molecules with multiple transmembrane segments.2 Molecules associated with cell-to-cell interactions are most often characterized by the presence of C2-type domains (see Fig. 1). This category includes the cell adhesion molecules (CAMs).2

Vascular cell adhesion molecule (VCAM-1, CD106, or INCAM-110), platelet endothelial cell adhesion molecule (PECAM-1/CD31) and intercellular adhesion molecules 1, 2 & 3 (ICAM-1, 2 & 3) are five functionally related CAM/IgSF molecules that are critically involved in leukocyte-connective tissue/endothelial cell interactions.2-5 Expressed principally on endothelial cells, these molecules in general regulate leukocyte migration across blood vessel walls and provide attachment points for developing endothelium during angiogenesis.3, 5, 6

Structural Information

Human VCAM-1, is a 100-110 kDa, 715 aa residue, type I (extracellular N-terminus) transmembrane glycoprotein characterized by the presence of seven C2-type immunoglobulin domains.7-11 Approximately 80 kDa in predicted molecular weight, human VCAM-1 contains a 674 aa residue extracellular segment, a 22 aa residue transmembrane domain, and a 19 aa residue cytoplasmic tail. There are multiple N-linked glycosylation sites, and each C2 domain is associated with a pair of cysteines that form disulfide linkages, stabilizing the overall domain.1, 2 In addition, two extra cysteines are found in association with C2 domains number 1 and number 4, and it appears that disulfide bonds also form between these cysteines.2, 8 Rat and mouse VCAM-1 sequences have been published and each molecule has been found to be approximately 75% identical to the human molecule.8 While the significance of this degree of homology is unclear, this is in marked contrast to ICAM-1 where human and mouse sequences share only 50% aa sequence identity.12

Although VCAM-1 with seven domains is considered the predominant form of VCAM-1,7, 8, 10 alternatively spliced forms are known to occur. In rabbits, an eight domain variant has been reported7 and, in humans, a six domain form (the 4th domain being absent) occurs with some regularity.9, 10 In mouse, the situation is somewhat more complex. Mouse cells stimulated with endotoxin or IL-1 beta are known to produce a novel form of VCAM that exhibits the first three C2 domains and terminates in a 36 aa residue tail that arises from the use of an alternate exon 5 (exon 5 normally codes for domain 47).13, 14 Notably, the tail does not represent a transmembrane segment and the molecule is anchored to the cell membrane through a GPI-link. While truncated, this 43 kDa form of VCAM still demonstrates ligand-receptor binding.13, 14 Cells known to express VCAM-1 include activated neurons,15 endothelial cells,16, 17 smooth muscle cells,18 fibroblasts,19, 20 macrophages (Kupffer cells),21 dendritic cells,22 oocytes23 and Sertoli cells.24 Cell lines known to express VCAM-1 include the HS20 human bone marrow stromal cell line25 and LAN-1 neuroblastoma cells.14

Finally, soluble forms of VCAM-1 have been identified in tissue culture supernatants and in blood.11, 26-29 The exact mechanism that generates soluble molecules is unknown, but may involve proteolytic cleavage and/or alternative splicing.11, 14 Blood levels appear elevated in diseases as diverse as acute myelomonocytic leukemia,26 bronchial asthma,27 acute phase multiple sclerosis28 and sepsis.29


The ligands (or co-receptors) for VCAM-1 have been identified and found to be the alpha4beta1 (or VLA-4) and alpha4beta7 integrins.30-33 Integrins are non-covalently linked heterodimers composed of one large alpha subunit (120-180 kDa) and one small beta subunit (90-120 kDa). To date, there are at least eight subfamilies of integrins known, each distinguished by the presence of a distinct beta1-8 subunit. Of the better known integrins, beta2-type integrins are noted for their binding to the ICAMs plus their association with GPI-linked proteins, while beta1 integrins (VLA) bind extracellular matrix and VCAM-1.3, 34, 35 The overall complexity of the integrin family is demonstrated by the fact that 1) certain alpha subunits can associate with multiple beta subunits (e.g., alpha4 with beta1 and beta7); 2) many individual beta subunits associate with multiple alpha subunits (e.g., alphaL/CD11a, alphaM/CD11b, and alphaX/CD11c combine with beta2/CD18); 3) each individual integrin can bind to ICAM-1, -2, and -3); and 4) individual integrin co-receptors can bind many integrins (e.g., fibronectin binds alphavbeta3, alphavbeta6, alpha5beta1, and alpha4beta1.3

The principal ligand or co-receptor for VCAM-1 is alpha4beta1/VLA-4.32, 36 The human a4 subunit is a 150 kDa, 999 aa residue type I (extracellular N-terminus) transmembrane glycoprotein.37 It contains a 944 aa residue extracellular segment, a 23 aa residue transmembrane domain, and a 32 aa residue cytoplasmic tail. The extracellular segment is characterized by seven homologous repeating domains, with each domain approximately 30 - 40 aa residues in length, separated from other domains by approximately 30 aa residues.37 Classically, VLA complexes possess a chains that show either a 180 aa residue insert between domains 2 and 3, or evidence of proteolytic processing in the C-terminus of the extracellular region that results in two subunits (130 and 30 kDa) that are linked together via a disulfide bond.40 In the case of the alpha4 molecule, neither of the circumstances apply. Instead, the alpha4 chain shows occasional cleavage mid-chain, yielding two nearly equivalent (80 and 70 kDa), noncovalently linked fragments.37, 38 While the significance of this is unclear, it may represent a mechanism that regulates functional VLA-4 expression. Mouse alpha4 has also been cloned and found to be 84% identical to human alpha4 at the amino acid level. Remarkably, there is only one aa difference in the 32 aa residue cytoplasmic tail.39 The human beta1 molecule has also been isolated and found to be a 130-140 kDa, 778 aa residue type I (extracellular N-terminus) transmembrane glycoprotein.40 It shows 47 aa residues in its cytoplasmic tail, 23 aa residues in its transmembrane segment, and 708 aa residues in the extracellular domain. Within the extracellular segment is a four-fold repeat of cysteine-rich amino acids.40 Mouse beta1 has also been isolated and found to be 92% identical to human beta1, all extracellular potential N-linked glycosylation sites and cysteine residues are conserved.41 In addition to binding to VCAM-1 (via the first and fourth Ig-like domains),42, 43 VLA-4 is also known to bind to fibronectin,44-46 and to itself via homotypic binding.31, 47 Cells known to express VLA-4 include eosinophils,48, 49 basophils,50, 51 NK cells,36, 52 monocytes, T cells, B cells, thymocytes,38 and myeloma cells.53 Neutrophils have not been shown to express VLA-4.54

Although alpha4beta7 has been identified as binding to VCAM-1,33 it appears that the primary receptors for alpha4beta7 are MadCAM-155 and fibronectin.33 Following activation, both eosinophil56 and NK cell57 binding to VCAM-1 is enhanced, and alpha4beta7 probably plays a role in the transendothelial migration of these cells.

Biological Functions

VCAM-1 has a modest number of activities, all of which seem to be related to its expression as a membrane-bound adhesion molecule. The most studied activity for VCAM-1 is related to leukocyte extravasation.1, 3, 58, 59 At this time, the model that best describes general leukocyte migration from the circulation is based on three (or four) sequential steps that constitute an adhesion cascade.3, 58 In the first step, leukocytes under the influence of fluid shear forces transiently bind to adjacent endothelium. This binding is mediated by endothelially-expressed CD34, P- and E-selectin, and leukocyte-expressed L-selectin plus complex carbohydrates containing sialic acid and fucose.58 In the presence of bloodstream shear forces, this transient binding results in a series of leukocyte-endothelial engagements/disengagements that cause leukocytes to "roll" along the endothelium. During some periods of engagement, bound or tethered leukocytes come into contact with an "activating" molecule such as a chemokine or CD31, and this interaction results in leukocyte adhesive activation. This activation is somewhat unusual, in that a transduced signal from the ligated leukocyte chemokine receptor directly or indirectly stimulates the cytoplasmic tails of otherwise inactive, membrane-bound leukocyte integrins. This results in a reconfiguration of the extracellular integrin domain that greatly facilitates a strong interaction with ICAM-1, ICAM-2, and VCAM-1, allowing leukocytes to bind firmly to endothelium and to begin a search for endothelial edges that will direct their migration.3, 58, 59

Normally, VCAM-1 has a low to nominal expression on unstimulated endothelium.60, 61 It is, however, inducible by a number of cytokines (IL-1, TNF-alpha, IL-4 and IL-13) plus ECM molecules.60, 62, 63 When induced, VCAM-1 has the potential to play a significant role in migration for leukocytes that express VLA-4 (e.g., lymphocytes, monocytes, eosinophils, basophils). However, the actual contribution that VCAM-1 makes may depend upon the expression and activation state of other related adhesion molecules. TNF-alpha is a molecule produced by mast cells and macrophages in response to microbial agents.64, 65 Upon exposure to TNF-alpha, endothelial cells upregulate the expression of ICAM-1, E-selectin and VCAM-1.66, 67 These are molecules utilized by both neutrophils (E-selectin and ICAM-1)34, 68 and monocytes (E-selectin, ICAM-1, and VCAM-1)69, 70 during extravasation. Therefore, in the context of an early infection, any leukocytic infiltrate would be expected to be mixed and predominantly neutrophilic in nature with VCAM-1 playing a minor role in migration. If, however, an antigenic challenge is "allergic" in nature and involves IgE antibody, mast cells could again be involved, but this time release IL-4.71 Although both TNF-alpha and IL-4 induce endothelial VCAM-1 expression, IL-4, unlike TNF-alpha, does not upregulate E-selectin or ICAM-1.54, 72 This would remove adhesion molecule support for almost all neutrophil and some monocyte extravasation, and result in a predominantly eosinophilic infiltration. Finally, VCAM-1 also plays an important role in lymphocyte homing and migration.

VCAM-1 has also been suggested to play a role in cardiovascular development. In particular, mice deficient in VCAM-1 fail to develop a pericardium and this is attributed to the lack of a functional VLA-4/VCAM-1 system. It is hypothesized that the absence of VCAM-1 on myocardial cells fails to provide VLA-4-expressing mesothelial cells with their natural ligand, resulting in an inability to migrate into proper position during development.73, 74


  1. Williams, A.F. & A.N. Barclay (1988) Annu. Rev. Immunol. 6:381.
  2. Brummendorf, T. & F.G. Rathjen (1995) Prot. Profile 2:963.
  3. Imhof, B.A. & D. Dunon (1995) Adv. Immunol. 58:345.
  4. Griffiths, C.E.M. et al. (1995) Br. J. Dermatol. 133:823.
  5. Watt, S.M. et al. (1995) Leuk. Lymph. 17:229.
  6. Patey, N. et al. (1996) Am. J. Pathol. 148:465.
  7. Cybulsky, M.I. et al. (1991) Proc. Natl. Acad. Sci. USA 88:7859.
  8. Hession, C. et al. (1992) Biochem. Biophys. Res. Commun. 183:163.
  9. Osborn, L. et al. (1989) Cell 59:1203.
  10. Hession, C. et al. (1991) J. Biol. Chem. 266:6682.
  11. Pigott, R. et al. (1992) Biochem. Biophys. Res. Commun. 187:584.
  12. Horley, K.J. et al. (1989) EMBO J. 8:2889.
  13. Moy, P. et al. (1993) J. Biol. Chem. 268:8835.
  14. Terry, R.W. et al. (1993) Proc. Natl. Acad. Sci. USA 90:5919.
  15. Birdsall, H.H. et al. (1992) J. Immunol. 148:2717.
  16. Rice, G.E. & M.P. Bevilacqua (1989) Science 246:1303.
  17. Sano, H. et al. (1995) Int. Arch. Allergy Immunol. 107:533.
  18. Ardehali, A. et al. (1995) Circulation 92:450.
  19. Meng, H. et al. (1995) J. Invest. Dermatol. 105:789.
  20. Schmitz, B. et al. (1995) Acta Haematol. 94:173.
  21. van Oosten, M. et al. (1995) Hepatology 22:1538.
  22. Huang, M-J. et al. (1995) Leuk. Lymph. 18:259.
  23. Campbell, S. et al. (1995) Hum. Reprod. 10:1571.
  24. Riccioli, A. et al. (1995) Proc. Natl. Acad. Sci. USA 92:5808.
  25. Van der Velde-Zimmermann, D. et al. (1996) Int. J. Cancer 66:225.
  26. Sudhoff, T. et al. (1996) Leukemia 10:682.
  27. Koizumi, A. et al. (1995) Clin. Exp. Immunol. 101:468.
  28. Matsuda, M. et al. (1995) J. Neuroimmunol. 59:35.
  29. Boldt, J. et al. (1996) Crit. Care Med. 24:385.
  30. Elices, M.J. et al. (1990) Cell 60:577.
  31. Pulido, R. et al. (1991) J. Biol. Chem. 266:10241.
  32. Chan, B.M.C. et al. (1992) J. Biol. Chem. 267:8366.
  33. Ruegg, C. et al. (1992) J. Cell Biol. 117:179.
  34. Hynes, R.O. (1992) Cell 69:11.
  35. Petty, H.R. & R.F. Todd (1996) Immunol. Today 17:209.
  36. Rott, L.S. et al. (1996) J. Immunol. 156:3727.
  37. Takada, Y. et al. (1989) EMBO J. 8:1361.
  38. Hemler, M.E. (1990) Annu. Rev. Immunol. 8:365.
  39. Neuhaus, H. et al. (1991) J. Cell Biol. 115:1149.
  40. Argraves, W.S. et al. (1987) J. Cell Biol. 105:1183.
  41. Holers, V.M. et al. (1989) J. Exp. Med. 169:1589.
  42. Vonderheide, R.H. et al. (1994) J. Cell Biol. 125:215.
  43. Vonderheide, R.H. & T.A. Springer (1992) J. Exp. Med. 175:1433.
  44. Wayner, E.A. et al. (1989) J. Cell Biol. 109:1321.
  45. Mould, A.P. et al. (1990) J. Biol. Chem. 265:4020.
  46. Irie, A. et al. (1995) EMBO J. 14:5550.
  47. Altevogt, P. et al. (1995) J. Exp. Med. 182:345.
  48. Weller, P.F. et al. (1991) Proc. Natl. Acad. Sci. USA 88:7430.
  49. Walsh, G.M. et al. (1991) J. Biol. Chem. 146:3419.
  50. Bochner, B.S. et al. (1996) J. Immunol. 157:844.
  51. Lavens, S.E. et al. (1996) Am. J. Respir. Cell Mol. Biol. 14:95.
  52. Allavena, P. et al. (1991) J. Exp. Med. 173:439.
  53. Tatsumi, T. et al. (1996) Jpn. J. Cancer Res. 87:837.
  54. Bochner, B.S. et al. (1991) J. Exp. Med. 173:1553.
  55. Berlin, C. et al. (1993) Cell 74:185.
  56. Walsh, G.M. et al. (1996) Immunology 89:112.
  57. Perez-Villar, J.J. et. al. (1996) Immunology 89:96.
  58. Springer, T.A. (1994) Cell 76:301.
  59. Bevilacqua, M.P. (1993) Annu. Rev. Immunol. 11:767.
  60. Swerlick, R.A. et al. (1992) J. Immunol. 149:698.
  61. Meerschaert, J. & M.B. Furie (1994) J. Immunol. 152:1915.
  62. Sironi, M. et al. (1994) Blood 84:1913.
  63. Morisaki, N. et al. (1995) Biochem. Biophys. Res. Commun. 214:1163.
  64. Malaviya, R. et al. (1996) Nature 381:77.
  65. Mattson, E. et al. (1993) FEMS Immunol. Med. Microbiol. 7:281.
  66. Modur, V. et al. (1996) J. Biol. Chem. 271:13094.
  67. Meng, H. et al. (1995) J. Cell. Physiol. 165:40.
  68. Luscinskas, F.W. et al. (1989) J. Immunol. 142:2257.
  69. Carlos, T. et al. (1991) Blood 77:2266.
  70. Meerschaert, J. & M.B. Furie (1995) J. Immunol. 154:4099.
  71. Bradding, P. et al. (1992) J. Exp. Med. 176:1381.
  72. Schleimer, R.P et al. (1992) J. Immunol. 148:1086.
  73. Yang, J.T. et al. (1995) Development 121:549.
  74. Kwee, L. et al. (1995) Development 121:489.