Collagen glycosylation

Collagens are the most abundant proteins in animals, but the fact that collagens are O-linked glycoproteins is often overlooked. Collagen carry a specific type of O-glycans on the amino acid hydroxylysine (Hyl). Collagen glycosylation is strongly conserved among animals as it is found from simple sponges up to mammals.

Collagens are the main component of the extracellular matrix, often building fibrillar meshes interacting with multiple adhesion proteins such as laminin, fibronectin and integrins. Collagen domains are characterized by repeats of the triplet sequence Gly-X-Y, where X and Y can be any amino acid. The repeated occurrence of glycine is essential for the formation of the right-handed triple-helix conformation typical of collagens. The thermal stability of collagens is increased by the vitamin C-dependent hydroxylation of proline, which is abundant in collagen sequences. Another frequent amino acid of collagen, lysine in position Y, is also hydroxylated. The hydroxylation of proline and lysine and the glycosylation of Hyl take place in the ER before the formation of triple-helix collagen.

FIG: COLLAGEN MATURATION (ER)

Selected Hyl residues on collagen are glycosylated by the addition of (β1-O)Gal, which is often extended by α1-2 linked Glc. Collagen glycosylation was first described in 1957 by Grassmann and collaborators and the structure of the glycan chain has been elucidated by Robert Spiro in the late ’60. Thereafter, the glycosyltransferase activity involved in hydroxylysine glycosylation have been partially characterized but the genes encoding the collagen glycosyltransferases have only been identified recently with the description of two β1-O Gal-transferases encoded by the GLT25D1 and GLT25D2 genes.

FIG: GLC-GAL-HYL

The extent of Hyl glycosylation varies according to the type of collagens and to the tissues expressing collagens. For example, collagen type IV is more glycosylated than fibrillar collagen types I, II and III. The functional significance of collagen O-glycans is presently unknown. Given that the carbohydrates face the outside of the triple-helix, glycans likely mediate interactions between collagen fibrils and adhesion proteins. In this context, glycosylation has been suggested to regulate the organization of collagen fibrils, as an inverse relationship between Hyl glycosylation and fibril diameter has been observed. Interestingly, glycosylated Hyl also occurs in the collagen domains of non-fibrillar soluble proteins such as members of the collectin family encompassing the complement protein C1q, the mannose-binding lectin MBL and the surfactant proteins A and D. Since the collagen domain of collectins is involved in the oligomerization process, it is likely that the glycan chains participate in this process, too.

The importance of lysine hydroxylation in collagen integrity has been established by showing that mutations of the PLOD1 (lysine hydroxylase-1) gene leads to a connective tissue disorder classified as Ehlers-Danlos syndrome type VI (EDS VI; OMIM 225400). Patients display severe muscular hypotonia at birth with progressive crippling kyphoscoliosis. A subgroup of EDS VI, referred to as EDS VIB (OMIM 229200), yet express normal levels of lysine hydroxylation in collagens, thus suggesting that a modification downstream of the hydroxylation may be responsible for the clinical manifestations. This hypothesis however remains speculative as long as cases of collagen glycosylation deficiency have not been demonstrated.

Collagen glycosylation has been related to rheumatoid arthritis, an autoimmune disease characterized by joint inflammation, which affects about 1% of the human population. Several studies have shown that a glycosylated collagen type II peptide is recognized by the T-cells of rheumatoid arthritis patients. Accordingly, changes in collagen glycosylation may lead to the creation of new epitopes, which are recognized by the immune system, hence initiating an inflammatory response towards collagen. An explanation for the development of an anti-collagen immune response may be linked to the exposure to microorganisms expressing collagen-like (glyco)proteins. In fact, the collagen motif Gly-X-Y is also found in several bacterial and viral proteins, for which an exact function of the collagen domain still remains unknown. Whereas the collagen domain of most viral proteins are quite short with stretches of 4-5 Gly-X-Y repeats, some bacterial and fungal proteins are very similar to animal collagens. The presence of proline and lysine within the collagen domain of bacterial proteins, such as the SclB protein of Streptococcus pyrogenes, points to potential acceptor sites for collagen hydroxylation and glycosylation. However, so far no collagen hydroxylase or collagen glycosyltransferase has been described in bacteria and fungi.