A sparse matrix approach to the solubilization of overexpressed proteins.

Many biophysical experiments depend on large amounts of pure, soluble protein. Indeed, the revolution in structural biology has depended on molecular biology's potential to make experiments possible by allowing the overexpression of normally rare proteins in a heterologous host. All too often, however, overexpressed proteins are poorly soluble in buffers that attempt to mimic physiological conditions. Often in such cases the overexpressed protein is assumed to be present in inclusion bodies and hopes of obtaining the desired sample from the overexpression vector are abandoned. We have developed a sparse matrix approach to the solubilization of such proteins that is often successful. This approach relies on well accepted theories of protein solubility and folding to build a sparse matrix that samples 'solubility space' effectively. The buffers of the sparse matrix are used to make crude extracts that are rapidly assayed for soluble protein using gel electrophoresis. We describe our approach and give examples of its application.