Collagen is a ubiquitous biomolecule that provides the tensile strength in the connective tissues of all multicellular organisms. It is the most abundant protein in mammals, constituting a quarter of their total weight. It is the major fibrous element of skin, bone, tendon, cartilage, blood vessels, and teeth.
Osteogenesis imperfecta, also known as OI or brittle bone disease, is a genetically transmitted disease of type I fibrillar collagen. The disease has a range of severities, from lethal in the perinatal period (either before or shortly after birth) to nearly asymptomatic.
The cause of the disease is mutations in the COL1A1 or COL1A2 genes. These genes code for the a1 and a2 chains of collagen type I. At this time, over 150 different point mutations in COL1A1 or COL1A2 are known to cause OI, 75 of which are lethal and 75 of which are not. Despite knowing the structure of the collagen molecule and being able to identify mutations that are both lethal and non-lethal, the precise relationship between the features of a mutation and its lethality remain obscure. For example, there are several mutations to a particular amino acid (such as from glycine to serine) that are lethal in one location and non-lethal in another. The problem of understanding the relationship between a mutation and its phenotype is a common subtask of finding structure-to-function relationships. The goal of this collaborative research is to use molecular modeling techniques to discover the factors that govern the lethality of collagen mutations.
Today, there is a tremendous need for biomaterials that can perform intended functions when implanted in the human body and that are biocompatible. Collagen is a rigid, rod-like molecule which aggregates into a fibril. It is an excellent candidate for experimental modification to create novel biomaterials, due to its biocompatible ability to form strong fibers. Increased understanding of both the molecular structure of collagen molecules and alterations to native structure in the mutants will provide the necessary information to design useful collagen-based biomaterials.
In order to understand OI and to design collagen-based biomaterials, it is essential to develop three-dimensional atomic models that provide detail about the structural and energetic properties of both native and mutated collagen molecules. Three-dimensional coordinates for collagen models are obtained either from the Brookhaven Protein Data Bank or are generated programatically from idealized collagen structures and amino acid side-chain rotamer libraries using the program gencollagen. Refinement and molecular simulations on these models are either performed locally or at the San Diego Supercomputer Center.
Currently, our collaborators view static images of the models because their interactive modelling facilities are limited. During telephone conversations the static images are frequently inadequate because questions arise regarding molecular interactions that were not captured in the image set. The introduction of a molecular modeling collaboratory available on a wide variety of platforms -- particularly low-cost platforms -- would alleviate these problems.