Biochemical diversity in a phage display library of random decapeptides

Detailed knowledge and understanding of the structure and function of biologically important macromolecules is frequently insufficient to permit rational, de novo design of recognition molecules and therapeutics. Traditional drug discovery has, thus, focused on screening and identifying native molecules as drugs or as templates for genetic engineering or organic synthesis. The number and novelty of lead compounds for drug discovery might be expanded significantly, however, by the ability to express and screen large libraries of peptide structures with phage display technologies [Scott and Smith, Science 249 (1990) 386-390]. The significance of such libraries as sources of novel biological ligands will depend in part on the depth and degree of biochemical diversity they comprise. We have prepared a phage display library of greater than 2 x 10(6) individual decapeptides produced as N-terminal fusions to the pIII surface protein of fd filamentous phage. The decapeptides were expressed from a degenerate DNA insert sequence that was chemically synthesized with an equal mixture of all four nucleotide bases at the three positions in each of the ten codons. Fifty-two clones were picked randomly and without prior selection from the population and the sequences of their peptide inserts were determined. Our results confirm and document the broad representation at the primary amino acid sequence level that is expected in a library expressed from random DNA inserts. More significantly, biochemical characterization shows these insert sequences correspond to structures comprising a wide range and combination of isoelectric, hydropathic, and biochemical properties necessary in drug discovery to access a significant percentage of the repertoire of possible peptide structures by affinity or activity screening.