Search

Over recent years a number of cartilage and bone matrix molecules have been identified and characterized. These include major constituents such as collagens and proteoglycans as well as a number of less-abundant matrix proteins. In several cases these proteins have been characterized by cloning and sequence analysis of the corresponding cDNA. Some properties of the macromolecules have been studied and an understanding of their functions in the structure, assembly, and breakdown of connective tissue matrix is emerging. It appears that some of these molecules have structural roles whereas others participate in the assembly of the tissue. In this paper we attempt to give a current picture of the organization and role of the noncollagenous matrix macromolecules in cartilage and bone.

Comparison of NH2-terminal protein sequence from the rat OX-44 antigen with the sequence of the human CD37 antigen deduced from a cDNA clone shows that these antigens are species homologues. The CD37 sequence is 244 amino acids in length and lacks a conventional leader sequence. The molecule is likely to have an NH2-terminal cytoplasmic domain followed by three transmembrane sequences that lie within the first 110 amino acids. The rest of the molecule is hydrophillic and contains three sites for N-linked glycosylation.

Application of these methods to a number of macromolecular species shows, in general, that features which are only occasionally or marginally detected in bright field may be visualized clearly and consistently through the use of dark-field illumination. Given the minimal average thicknesses of contrasting metal required for dark-field examination (ca. less than 10(-7) g/cm2, or less than 0.4 nm thickness), distortions of molecular features by accumulation of coating metal, although not avoided altogether, are certainly minimized. Thus, application of minimal metal coating in conjunction with high-resolution dark-field electron microscopy should facilitate solution of structural problems in molecular biochemistry generally. The method is of particular current interest in the specific area of coagulation and adhesive protein conformation. The advantages of metal coating macromolecules for dark-field observation are summarized in Table I. In the case of macromolecules or macromolecular complexes which are of limited width (5 nm or less) even though extended lengthwise, contrast available for visibility tends to be limiting. It is for such structures that dark field offers a productive alternative. The dark-field method seems promising for study of certain macromolecular preparations of adhesive proteins vis-à-vis isolation of receptor loci. These include the complex between fibrinogen and GPIIb/IIIa; the complex between GPIb and von Willebrand protein, and the proposed complex between thrombospondin and fibrinogen. The use, for relating epitopic sites to topographical position, of Fab fragments directed against specific epitopes on coagulation macromolecules (and/or their complexes) of otherwise well-defined structure may also be extended by means of the dark-field approach.

Human blood basophils have been shown to form Charcot-Leyden crystals (CLC) and to contain quantities of CLC protein comparable to the eosinophil (Ackerman SJ, Weil GJ, Gleich GJ, J Exp Med 155:1597, 1982). We examined the subcellular localization of CLC protein in human basophils from peripheral blood using an ultrastructural postembedding immunogold method and affinity chromatography-purified polyclonal primary antibody to purified human eosinophil CLC protein. We found CLC protein to be uniquely associated with the main, large, particle-filled granule population of human basophils and particularly within intragranular Charcot-Leyden crystals in these granules. Rarely, CLC protein was localized within small, smooth perigranular vesicles. Neutrophils, lymphocytes, and monocytes present in preparations stained for the CLC protein were negative as were controls for the specificity of the immunogold staining. Mature eosinophils contained small numbers of positive crystalloid-free primary granules, as previously reported (Dvorak AM, Letourneau L, Login GR, Weler PF, Ackerman SJ, Blood 72:150, 1988). The presence of CLC protein in hexagonal and pyramidal crystals regularly present in activated types II and III basophils suggests two possibilities currently being investigated. Like eosinophils, basophils could synthesize CLC protein, or eosinophil-derived CLC protein could be internalized and stored in basophil granules. Regardless of the mechanism for acquisition of CLC protein (lysophospholipase) by basophils, the identification and subcellular localization of this enzyme in basophils requires that it be considered in basophil cell biology as well as in the pathobiology of basophil-rich reactions.