Discovery and Optimization of Peptide Macrocycles

Discovery and Optimization of Peptide Macrocycles Abstract Introduction: Macrocyclic peptides are generally more resistant to proteolysis and often have higher potency than linear peptides and so they are excellent leads in drug design. Their study is significant because they offer the potential as a new generation of drugs that are potent and specific, and thus might have fewer side effects than traditional small molecule drugs. Areas covered: This article covers macrocyclic drug leads based on nature-derived cyclic peptides as well as synthetic cyclic peptides and close derivatives. The natural peptides include cyclotides, sunflower-derived peptides, theta-defensins and orbitides. Technologies to make engineered cyclic peptides covered here include cyclization via amino acid linkers, CLIPS, templates, and stapled peptides. Expert opinion: Macrocyclic peptides are promising drug leads and several are in clinical trials. In our opinion, they offer key advantages over traditional small molecule drugs, as well as some advantages over protein-based ‘biologics’ such as antibodies or growth factors. These include the ability to penetrate cells and attack intracellular targets such as protein-protein interactions as well as to hit extracellular targets. Some macrocyclic peptides such as cyclotides offer the potential for production in plants, thus reducing manufacture costs and potentially increasing opportunities for their distribution to developing countries at low cost. Keywords: Cyclic peptides, cyclotides, drug design, kalata B1, peptide macrocycles, stapled peptides 1 Abbreviations: CCK, cyclic cystine knot; SFTI, sunflower trypsin inhibitor; MALDI-TOF, matrix-assisted laser desorption/ionization time of flight; MCoTI, Momordica cochinchinensis trypsin inhibitor; RTD, rhesus theta-defensin; AEP, asparaginyl endopeptidase; BTD, baboon theta-defensin; KLK4, kallikrein-related peptidase 4; PawS, Preproalbumin with SFTI-1; PDP, PawS derived peptides; SICLOPPS, split intein-mediated curcular ligation of peptide and proteins; MOrPHs, macrocyclic organo-peptide hybrids; CLIPS, chemical ligation of peptides onto scaffolds; TABA, N,N′,N′′-(benzene-1,3,5-triyl)-tris(2-bromoacetamide); GHRH, Growth hormone-releasing hormone. 1. Introduction Peptide macrocycles have stimulated much interest in the field of drug discovery due to their unique structural and biopharmaceutical properties. Cyclization of their linear parent peptides can be achieved via a range of connectivities, including head-to-tail, sidechain-to-sidechain or termini- to-sidechain linkages; it is this variety that results in a structurally diverse class of molecules, both from natural and synthetically derived sources. Macrocycles have captivated the imagination of medicinal chemists because of their diverse design applications, including the incorporation of bioactive motifs into cyclic peptide scaffolds, or the stabilization and optimization of bioactive structures using synthetic linkages (Figure 1). There are currently ~30 peptide macrocycles registered or in clinical development, with cyclosporin A the only one to be administered orally [1]. -Insert Figure 1 here Despite their promise, peptide macrocycles are an underrepresented class within marketed drugs, consistent with Lipinski’s “rule-of-five” [2], which notes a preference for a molecular weight of lower than 500 Da amongst approved drugs. Notwithstanding this size rule for drug-like 2 properties, many large biologics such as engineered antibodies have larger molecular weights (>5000 Da) or violate other aspects of Lipinski’s rules. These biologics typically have the advantage of exquisite target specificity but lack the oral bioavailability of ‘small molecule’ drugs, thus requiring alternative administration methods such as injection. An example of a high molecular weight blockbuster injectable drug is insulin, for the treatment of diabetes and adalimumab for arthritis. Molecular weight alone does not preclude oral activity, and in an extension of the Lipinski rules Veber et al. noted that, in some cases, oral bioactivity can be achieved even for molecular weights of greater than 500 Da, provided other biophysical properties are suitable. In particular, molecular flexibility and polar surface area are important considerations in the development of orally bioavailable peptide macrocycles [3]. Peptide macrocycles fill a gap between small molecule drugs and biologics, with their sizes typically ranging from 500–5000 Da. They can exhibit high target specificity and, in some cases, cellular permeability and/or oral bioavailability, and their size is sufficient to target intracellular protein-protein interactions. This is important because protein-protein interfaces have previously been perceived generally as ‘undruggable’, i.e. not able to be targeted by either small molecule drugs or biologics, so the availability of technologies to address them represents a major advance in the field of drug design. This article describes recent advances in the optimization of naturally and synthetically derived peptide macrocycles in the development of leads for a range of therapeutic targets. We focus on studies published over the last few years, as earlier background literature in this field is covered in a number of excellent reviews [4-6]. 2. Natural peptide macrocycles in drug discovery 3 Naturally occurring peptide macrocycles are structurally diverse and have been isolated from bacteria [7], fungi [8], plants [8] and mammals [9]. Their diversity spans a range from small pentapeptides with no disulfide bonds to large polypeptides with multiple disulfide bonds, as schematically illustrated in Figure 2. In addition to the interest generated from their structures, these peptides have been extensively investigated for their biological functions and applications in drug design, based largely on their intrinsically high stability and ability to be engineered to optimize function. In particular, certain classes of naturally occurring peptide macrocycles have been utilized as molecular scaffolds, whereby a new or modified biological function is introduced into the molecule whilst maintaining the intrinsic stability of the macrocycle. Amongst the various classes of natural peptide macrocycles, we focus here on ribosomally synthesized disulfide-rich peptides of fewer than 50 residues as these are a particularly well studied target class. Bacterial and fungal peptide macrocycles typically do not include disulfide bonds and will not be covered herein; readers are referred to other recent literature on these microorganism-derived macrocycles [7, 8, 10]. -Insert Figure 2 here 2.1 Cyclotides The most abundant group of ribosomally synthesized peptide macrocycles are head-to-tail cyclized peptides known as cyclotides, found predominantly in the Violaceae (violet) and Rubiaceae (coffee) plant families [11]. The prototypic cyclotide, kalata B1, was first discovered as the active ingredient in a tea made from the plant Oldenlandia affinis, used in a traditional African medicine to accelerate childbirth. Gran isolated the bioactive peptide from O. affinis and partially characterized its amino acid content in the early 1970s [12], but it was not until some 20 years later that the macrocyclic structure was elucidated [13]. Since then, more than 360 cyclotides have been 4 characterized, as documented in CyBase [14], and it is likely that this number will increase, with estimates of tens of thousands of cyclotides existing within the Rubiaceae family alone [11]. Cyclotides comprise ~30 amino acids and exhibit a unique structure known as the cyclic cystine knot (CCK). In this structure two disulfide bonds and their connecting backbone segments form a ring that is threaded by a third disulfide bond, with this knot motif embedded within a cyclic peptide backbone, as illustrated in Figure 3. This conserved structural feature engenders cyclotides with ultra-stable properties, including resistance to thermal, chemical or enzymatic breakdown [15]. Three cyclotide subfamilies have been defined: the Möbius, bracelet and trypsin inhibitor cyclotides. The Möbius and bracelet families differ mainly in the cis or trans configuration of an X-Pro peptide bond in one of the six backbone ‘loops’ of the cyclotide framework [16], with kalata B1 being the prototypic Möbius cyclotide and cycloviolacin O2 the most well studied bracelet cyclotide. The trypsin inhibitor subfamily currently comprises eight peptides (MCoTI-I–VIII) isolated from the tropical vine Momordica cochinchinensis. Members of this subfamily have very different sequences from Mobius or bracelet cyclotides yet maintain the CCK structure [17, 18]; of these MCoTI-I and MCoTI-II are the most well studied examples. -Insert Figure 3 here Natural cyclotides have a diverse range of biological activities, including anti-HIV [16, 19], anti-microbial [20], hemolytic [20-23], and cancer cell toxicity [24, 25], as well as pesticidal activities, with the latter presumed to be their natural function [26-29]. However, they have gained most attention for their applications as molecular scaffolds in drug design. With synthetic protocols for cyclotides now well established [30-32], various groups have ‘grafted’ foreign bioactive epitopes into cyclotide scaffolds to introduce novel bioactivities. This grafting concept was first explored by Clark et al., who introduced non-native hydrophobic residues into loop 5 of kalata B1 5 and demonstrated the plasticity of the CCK framework to accommodate changes in the primary

Load more