Peptides are attractive as pharmaceutical agents, as they are typically more specific than small molecule drugs and less immunogenic and more bioavailable than larger proteins. The development of pharmacologically active peptides has relied on the scaffolds found in naturally occurring peptides, which is limiting in terms of the availability of wild-type peptides with a given shape and chemical characteristics.
Computational methods were developed for the accurate de novo design of ultra-stable peptides with precisely specified tertiary structures that could be used in a range of biotechnological applications, such as making new drugs [1]. A structurally diverse array of both genetically encoded and non-canonical peptides (containing 18-47 residues and disulfide crosslinks, some with heterochirality and/or N-C backbone cyclisation) were designed and experimentally produced. X-ray and NMR structures were nearly identical to the computational design models and the resulting peptides showed exceptional stability towards thermal and chemical denaturation.
In this work, we present the NMR analysis and structural characterization of these synthetically prepared heterochiral disulfide-constrained peptides and show that the design process can be successfully used to tailor-make peptides that adopt desired and useful shapes and characteristics. High-resolution NMR solution structures were determined for each of the designs, including topologies that contained two to three canonical secondary structural elements: HH, HHH, EEH, EHE, HEE and EE, along with HLHR, a cyclic topology with left- and right-handed helices. For example, the NMR structure ensemble and design model of the HLHR topology were well-matched (Ca r.m.s.d., 0.79 Å), demonstrating that these computational methods are sufficiently versatile and robust to design in a conformational space not explored by nature.