The synthesis of triazole peptidomimetics

Abstract

The natural synthesis of peptides and proteins occurs in an efficient iterative manner, and can be considered as a model for highly efficient chemical access to molecular complexity and functional materials. This process is mimicked by modern solid phase peptide synthesis, which can give good, controlled access to molecules with molecular weights above 5,000 Da. The reactions involved in efficient production of these large molecules are required to be high yielding and efficient processes. The Copper-catalysed Azide-Alkyne Ligation reaction (Cu-AAL) described simultaneously by the Sharpless and Meldal groups is also renowned as a high yielding and efficient reaction. Taking inspiration from these processes, we sought to produce triazole peptidomimetics, in which the amide bonds of peptides are replaced by Cu-AAL derived triazoles. While some approaches to this type of molecule have been published, these procedures are not truly general. We sought to devise a synthetic approach to the first cyclic variations of this class of compounds, as they may have interesting conformational properties, and catalytic, coordination, or materials applications. Peptides can be visualised as a sequence of amino acid monomer units, which have been systematically coupled together. In a similar fashion, triazole peptidomimetics can also be viewed as a series of discrete alkyne and azide-containing units which have been linked through Cu-AAL. In order to facilitate the production of the triazole peptidomimetics, it was of particular importance to explore the synthesis of these monomer units. One method, building on literature precedent, utilised the chiral pool offered by amino acids in a sequence of functional group transformation reactions. While the synthesis was found to be successful for a test case consisting of phenylalanine-derived alkynyl- and azido- monomer units, the approach was not applicable to a wide range of substrates. In addition, the number of functional transformations required resulted in poor yields of the monomer units. It was also investigated whether these molecules could be constructed from simple starting materials in a stereocontrolled fashion. A particularly promising reaction sequence utilised Ellman’s chiral sulfinamide auxiliary, which allowed for the synthesis of analogous phenylalanine monomer units in fewer steps and higher yields. With an acceptable quantity of the phenylalanine-like monomer units in hand, iterative chain extension to linear triazoles was demonstrated in a high-yielding 2-step sequence. The tetramer, pentamer and hexamer peptidomimetics were derived into a form whereby head-to-tail cyclisation could be attempted. Successful cyclisation was seen for both the tetrameric and pentameric compounds, although the cyclic hexameric species was not able to be isolated. With the knowledge that small cyclic triazole peptidomimetics were accessible via our reaction pathway, we were keen to pursue cyclic compounds that contained diverse chemical functionality. An existing biologically active cyclic pentapeptide was taken, and steps were taken towards the synthesis of its all-triazole analogue. To accomplish this goal, the monomer unit synthesis was adapted for the production of the relevant compounds containing side chains derived from arginine, glycine, aspartic acid and lysine. The 2-step chain extension sequence was performed using the diverse monomer units, giving a linear tetramer. Given the difficulties producing the intended monomer unit analogue of arginine, the synthesis was not completed. We were curious to determine whether our monomer units could be used in the process of introducing a single triazole, acting as an amide bond surrogate, in established peptide sequences. A serine protease named granzyme B, involved in the apoptosis pathway, was taken as a target enzyme. Inhibitors of the protease were formed by effectively swapping the scissile amide bond of substrates of granzyme B with a triazole linkage, thereby denying the enzyme the means of turning over the molecules. Two compounds were produced via this strategy, which bound to human granzyme B with low millimolar IC50’s in an isolated enzyme assay.