We have recently re-examined the characteristic sharp denaturation temperature of the collagen molecule and fibre. It has been generally accepted for many years that denaturation is an equilibrium process involving the rupture of hydrogen bonds. We have now proposed that the process is an irreversible rate process, in which uncoupling of thea-chains initially occurs in a thermally labile domain devoid of hydroxyproline. The domain is located near the C-terminal and following alignment of the molecules in the quarter-stagger-end-overlap arrangement is located in the gap region of the fibre. The domain appears to be conserved in type I of several animal species, and is present in types II and III. Collagen molecules that co-polymerise to form fibres, types V and XI, do not possess this labile domain.

Ramachandran proposed that stabilisation of the triple helix occurred through hydrogen-bonded water-bridges involving the hydroxyl group of hydroxyproline. Recent studies have been equivocal, some questioning the role of water bridges and of hydroxyproline, whilst recent detailed X-ray studies of collagen-like peptides demonstrate the presence of a stabilising sheath of hydrogen-bonded water. Our findings support the proposal of hydrogen-bonded water-bridges stabilising the triple helix.