The structure of procollagen I is similar to other fibrillar collagens, and it comprises three polypeptide α-chains, which form a unique triple-helical structure (Fig. 5.2). It is a heterotrimer of two α1 and one α2 chains. It contains an uninterrupted triple-helical domain flanked by short non-helical telopeptides. The telopeptides, which do not have a repeating Gly–X–Y structure and do not adopt a triple-helical conformation, account for 2 % of the molecule. Many macromolecules such as COMP, fibromodulin, matrilin, and decorin attach to collagen I (Fig. 5.3). Collagen I molecules form D-periodic (D  =  67 nm, the characteristic axial periodicity of collagen) cross-striated fibrils in the extracellular space, giving the tissue its mechanical strength and providing the major biomechanical scaffold for cell attachment and anchorage of macromolecules (Fig. 5.4).

Collagen II fibrils in the nucleus pulposus appear to be organized randomly (Fig. 5.1). In the adult, these collagen molecules are joined by hydroxypyridium cross-links (Eyre and Muir 1976). There is evidence that procollagen II can be expressed in two forms by alternative splicing of the primary gene transcript. The two mRNAs either include (IIA) or exclude (IIB) exon 2, so that procollagen II can be synthesized with or without exon 2 encoding the major portion of the amino propeptide (Sandell et al. 1991; Ryan and Sandell 1990). This polymorphism is thought to influence cell morphology, with the cells expressing procollagen IIA being narrow, elongated, and “fibroblastic” in appearance, while the cells expressing procollagen IIB are large and round. The expression of procollagen IIB appears to be correlated with abundant synthesis and accumulation of aggrecan (Sandell et al. 1991). Procollagen IIA may function in the annulus fibrosus and nucleus pulposus, particularly during development.

Collagen VI accounts for 10–20 % of the total collagen and is particularly prominent in the nucleus pulposus and the end plate cartilage (Roberts et al. 1991). It resembles collagens IX and X in that it is a shorter chain collagen than collagen II. It consists of α-chains that form a highly branched filamentous network based on the formation of tetramers but does not appear to form cross-links with other matrix molecules (Wu et al. 1987). The beaded filaments of collagen VI represent another important network in the disc matrix (Feng et al. 2006). It forms a tetramer of two pairs of antiparallel collagen VI molecules arranged such that two N-terminal ends are exposed at either end of the unit. Further assembly occurs both by ­end-to-end and side-to-side associations catalyzed by proteoglycans such as biglycan and decorin. These ligands (collagen VI tetramers) are also bound to molecules such as PRELP, fibronectin, and matrilin-1, -2, or -3, which in turn are bound to a collagen fiber, a procollagen molecule, or aggrecan.

Collagen IX is a “fibril-associated collagen with interrupted triple helices” (FACIT) collagen. Other FACIT members are collagen XII, XIV, XVI, and XXI; when compared with collagen IX, they are less characterized in terms of structure and function. Collagen IX is usually found in tissues containing collagen II. It is extensively cross-linked to fibrils of collagen II in an antiparallel orientation and may also be cross-linked to other collagen IX molecules (Eyre et al. 1988; van der Rest and Mayne 1988; Wu et al. 1992). It has a periodic distribution along the fibril of approximately 67 nm (Vaughan et al. 1988). The globular NC4 domain at the N-terminus of the α1(IX) chain extends out from the fibril (Vaughan et al. 1988) and may provide a molecular link between the fibrils and other interfibrillar matrix components, such as the proteoglycan aggrecan. The amino-terminal NC4 domain is very basic (Vasios et al. 1988; Muragaki et al. 1990). Some forms of collagen IX may lack the NC4 domain due to the use of an alternative promoter (Nishimura et al. 1989), and the expression of this variant is thought to be tissue specific and developmentally regulated. Collagen IX of the disc is distinct from that of hyaline cartilage in that the disc contains only the short form of α1(IX) that lacks the NC4 domain (Wu and Eyre 2003). Usage of the short α1(IX) transcript in disc tissue has no apparent effect on cross-linking behavior. The precise biological role(s) of collagen IX, its assembly in relationship to collagen II, and the mechanisms that regulate its synthesis and degradation remain unknown.

Collagen X is a non-fibrillar short-chain protein containing three identical α1 chains [α1(X)₃] with a low molecular weight of 59 kDa (Schmid and Linsenmayer 1985). Two exons encode the complete primary translation product, which contains the N-terminal region (NC2). The protein is comprised of 52 amino acids, 18 of which form the signal peptide (LuValle et al. 1988). The collagen X molecule also contains a relatively larger (162 amino acids) noncollagenous C-terminal domain (NC1) (Yamaguchi et al. 1989), which plays a key role in the intracellular assembly of the triple-helical collagen X molecules and may be necessary for aggregation to form extracellular networks. The presence of several functional RUNX2 binding sites within the promoter region of the COL10A1 genes suggests that it is a direct transcriptional target of RUNX2 during chondrogenesis (Zheng et al. 2003). It forms pericellular mats and is also closely associated with the fibrils of collagen II (Linsenmayer et al. 1998).

Collagen X is synthesized by hypertrophic chondrocytes during endochondral ossification in the growth plate (Mwale et al. 2000; Tchetina et al. 2003). It was shown to be present in human lumbar discs during aging and degeneration by Boos et al. (1997) and Xi et al. (2004). In addition, nucleus pulposus chondrocytes express collagen X in association with advanced age and degenerative disc lesions (Nerlich et al. 1997; Hristova et al. 2011). Aigner et al. (1998) studied the variation in the pattern of collagen X expression with age in normal human discs, particularly the cells of the inner annulus fibrosus and the nucleus pulposus. These workers showed that with age some cells from the inner annulus or the nucleus expressed a hypertrophic chondrocyte phenotype and secreted collagen X as a component of the disc matrix.

COMP is a 524-kD protein composed of five identical glycoprotein subunits each of 100–120 kD held together by a five-stranded coiled-coil domain in the N-terminal portion and exposing unique C-terminal globular domains. Each subunit has EGF-like and calcium-binding (thrombospondin-like) domains. It is present in the extracellular matrix of the nucleus pulposus and annulus fibrosus (Ishii et al. 2006; Lee et al. 2007). It exhibits a lamellar distribution pattern in the annulus fibrosus region. Higher nucleus pulposus levels of COMP are found in aging discs (Lee et al. 2007). COMP plays a role in regulating collagen fibril assembly.

Synthesis of fibronectin has been demonstrated in healthy disc tissue (Hayes et al. 2001; Anderson et al. 2010) and cartilage (Wurster and Lust 1984). Fibronectin is a minor component of the disc of young and older animals (Hayes et al. 2001). Its function in the discs is unknown, although it is well established that fibronectin can play a role in cell–matrix, matrix–matrix interactions as well as binding collagen and heparan sulfate proteoglycans through RGD sequences. As a result of alternative splicing, different isoforms of fibronectin exist. Disc cells synthesize fibronectin with either or both the ED-B and ED-A domains present (Anderson et al. 2010), the significance of which is unclear. Fibronectin is elevated in degenerated discs, and its fragments induce the cell to degrade the matrix (Oegema et al. 2000, Anderson et al. 2004).

Amyloid deposits in the intervertebral disc were first described by Bywaters and Dorling (1970). In the annulus, amyloid deposits are found lying between thick bundles of collagen fibers, while a diffuse nodular distribution is found in the nucleus pulposus, in proximity to cells (Ladefoged 1985). The incidence of amyloid deposition in discs increases with advancing age, although it is also found in the young disc (Ladefoged 1985; Yasuma et al. 1992). Amyloid deposits of different morphological types have been observed in the disc (Mihara et al. 1994). At the ultrastructu ral level, amyloid deposits are seen composed of 10-nm-wide nonbranching fibrils (Mihara et al. 1994). Changes in matrix glycosaminoglycans, particularly strongly sulfated glycosaminoglycans such as keratan sulfate, may play a role in the pathogenesis of localized and systemic amyloid deposition (Athanasou et al. 1995).

Tenascin has also been called hexabrachion (Erickson and Inglesias 1984) and myotendinous antigen (Chiquet and Fambrough 1984). The tenascins are a family of extracellular matrix glycoproteins with repeated structural domains homologous to epidermal growth factor (EGF), fibronectin type III, and the fibrinogens. It has been shown that aggrecan can interact with certain matrix proteins containing EGF-repeats, including tenascins, the fibrillins, as well as the fibulins, (Day et al. 2004) to form higher order networks. Tenascin is a large extended “octopus”-like molecule, in which six arms radiate out from a central point, with each arm having an Mr of 200 kDa and composed of a single polypeptide chain. The domain structure of the cloned molecule supports the finding that it can interact with multiple ligands (Nies et al. 1991). It is a hemagglutinin and can bind, directly or indirectly, with a variety of disc matrix molecules. In discs, it is present in young and adult annulus fibrosus and nucleus pulposus, where it is confined to the pericellular matrix (Gruber et al. 2002, 2006) and its expression can be modulated by mechanical strain (Benjamin and Ralphs 2004). It is thought to have a role in disc aging and degeneration, possibly by modulating fibronectin cell interactions and causing alterations in the shape of disc cells (Gruber et al. 2002).