Chemical cross-linking restrictions on models for the molecular organization of the collagen fibre.

The nature of the precise packing of collagen molecules into a collagen fibril, producing the characteristic regular banding, is still debatable. The problem has been approached using electron microscopy and X-ray diffraction techniques, and several models have been proposed, including hexagonal packing, an octafibril structure, a two-strand rope and a five-strand rope (for review, see ref. 1). For the past decade the pentafibrillar model, originally proposed by Smith, has been widely accepted as the fundamental building unit. This model, based on the quarter-stagger end-overlap hypothesis of Hodge and Petruska, was supported by the X-ray diffraction data of Miller and Wray. These X-ray data have now been reinterpreted by Hulmes and Miller in terms of a quasi-hexagonal packing of collagen molecules. We argue here that until other characteristic parameters are taken into account, in particular the chemical cross-linking evidence, the packing problem is still unresolved.