A new general approach to purify proteins from complex mixtures

The selection of chromatography media and their sequential use represent a major difficulty to isolate a single protein from very crude protein extracts. The process described here consists of two main steps: (i) a rational selection of few media from a relatively large collection and (ii) the definition of the sequence of columns to get the best purity of the target protein. From the first step, one sorbent is selected for its properties to capture the protein to purify, regardless whether other protein impurities are also co-adsorbed; then 5-7 other complementary sorbents are identified to remove impurities but without interacting with the target protein under the same buffering conditions. The second step consists in superimposing sorbents under a cascade manner with the sorbent in charge to capture the target protein located in the last position. Non-adsorbed proteins are eliminated in the flowthrough; other impurities are progressively removed by the sorbent sequence and the target protein is finally desorbed and isolated from the last sorbent using an optimized gradient. All operations are performed with a single adsorption buffer for all columns and all monitoring performed by means of mass spectrometry associated with ProteinChip arrays and polyacrylamide gel electrophoresis. Examples of protein isolation/identification from human serum are described namely thyroxin-binding proteins and transferrin. The first is isolated thanks to a series of dye chromatography media, the second (transferrin) using current chromatographic media. In both cases the target proteins were purified at a level estimated of about 95% and 85%, respectively. Isolated proteins were pure enough for the purpose of formal identification by either peptide fingerprinting or sequencing.