Post-translational modification of genetically encoded polypeptide libraries

The genetic encoding of polypeptides with biological display systems enables the facile generation and screening of very large combinatorial libraries of molecules. By post-translationally modifying the encoded polypeptides, chemically and structurally more diverse molecules beyond linear amino acid polymers can be generated. The first post-translational modification applied to encoded polypeptides, the oxidation of cysteine residues to form disulfide bridges, is a natural one and was used to cyclise short peptides soon after the invention of phage display. Recently a range of non-natural chemical strategies for the post-translational modification of encoded polypeptide repertoires were applied to generate optical biosensors, semisynthetic polypeptides, peptide–drug conjugates, redox-insensitive monocyclic peptides or multicyclic peptides, and these strategies are reviewed in this article.

Introduction

Different biological selection methods such as phage display, ribosome display, mRNA display, yeast display, bacterial display, or others have been developed over the last three decades to link large libraries of polypeptides (typically 10⁹ to 10¹² different random polypeptides) to their encoding DNA (or RNA) and to isolate ligands that bind to biological targets [1, 2, 3, 4]. These techniques in which genes are translated by the ribosomal machinery into linear polypeptides have been used for the isolation of short peptidic ligands [5] as well as large proteins such as antibodies [6, 7]. A first human antibody [8] and a protease inhibitor domain isolated with phage display [9] are already used as therapeutics in the clinic.

The vast majority of genetically encoded polypeptide libraries are based on the 20 natural amino acids and do not contain any post-translational modifications with the exception of disulfide bridges. Disulfide bonds are generated by the oxidation of two cysteine residues, that are in close spatial proximity, and are used by nature to stabilise the tertiary or quaternary structure of proteins as for example in antibodies. In genetically encoded libraries, disulfide bond formation was found to be a particularly effective means for the generation of small peptide ligands: short linear peptides that were isolated in the first phage panning experiments against antibodies or streptavidin in the late 1980s showed a relatively weak binding [10, 11, 12]. By conformationally constraining the peptides in the phage libraries through inserting two constant cysteine residues at both ends of the random amino acid sequence and oxidation of the thiols, O’Neil et al. [13] as well as McLafferty et al. [14] were able to isolate much better binders. By measuring the binding strength of phage-selected cyclic peptides to streptavidin, Giebel et al. showed that the cyclic peptides had 100-fold to 1000-fold higher affinities than linear peptides previously isolated to the same target [15]. The binding advantage of cyclic over linear peptides was underscored by the finding that selections with completely random peptide libraries often yielded peptide binders that contained two cysteines [11, 15, 16, 17]. In fact, the first disulfide cyclised peptide was isolated from such a library even before peptide libraries with cysteine residues at fixed position were created [11]. Following the successes with encoded disulfide cyclised peptides, numerous cyclic peptide binders were generated to proteases, kinases, receptors and other protein classes [5, 18, 19].

The good results achieved with disulfide cyclised peptide libraries demonstrated, in an impressive manner, the power of post-translationally modified encoded peptide libraries. In the last 15 years a number of non-natural chemical transformations were applied to genetically encoded peptide libraries. These approaches to generate chemically and structurally more diverse peptide libraries are discussed in the following three sections.