The Lubell laboratory applies the strength of organic synthesis to explore the chemical-biology of peptides through conformational restriction. Innovating methods for amino acid, polyamide and heterocycle synthesis, we study the medicinal chemistry of peptide structures by diversity-oriented approaches to create so called “peptidomimetic” prototypes for drug discovery. Our projects focus typically on unnatural peptides that possess novel mechanisms of action for modulating biological activity.

Figure 1. Peptide dihedral angles

Peptide backbone geometry is defined by the φ (phi), ψ (psi) and ω (omega) dihedral angle values (Figure 1). Side chain orientations are defined by χ​ torsion angle values. Although local restrictions may combine with hydrophobic effects, van der Waals interactions, and hydrogen-bonds to overcome peptide solvation and give rise to relatively stable folded protein three-dimensional structures from longer peptides, shorter linear peptides (i.e., <10 residues) exist normally in dynamic equilibria of interconverting conformations.  We explore the strategic use of stereoelectronic, covalent and steric constraints to restrict dihedral angles with consequences on global geometry that offset the dynamic equilibria, remove the entropy penalty for folding, and pre-organize peptide conformations for receptor binding affinity.

Azapeptides and azasulfurylpeptides

In the polyamide chain, the carbonyl and NH groups act respectively as Lewis base and Brønsted acid sites that interact within the peptide and with the environment through hydrogen bonds. Amino acid surrogates that maintain the register of the peptide sequence yet enhance the Lewis basicity and Brønsted acidity of the amide groups are being studied to favour secondary structures such as turn conformations, to enhance metabolic stability and to improve receptor selectivity and potency (Figure 2). For example, azapeptides are peptide analogs that employ a semicarbazide as an amino amide surrogate in which the backbone α-CH is replaced by nitrogen. Through electronic interactions, the semicarbazide favours backbone β-turn geometry due to a combination of urea planarity and hydrazine nitrogen lone pair – lone pair repulsion. Side-chains may also be preserved on aza-residues in optimal orientations due to adaptive chirality about the α-nitrogen. Linear azapeptides have become drugs, due in part to enhanced stability and protease resistance. In 2009, our laboratory introduced a submonomer method that advanced azapeptide synthesis by enabling addition of diverse side chains onto a common intermediate. For example, access to aza-proparglyglycine (azaPra) residues empowered azapeptide library construction by orthogonal chemistry in which the triple bond is modified selectively in the presence of the backbone and side chain functional groups. In addition, by employing amino-sulfamides as amino amide surrogates, we synthesize azasulfurylpeptides that have been shown by X-ray analysis to replicate the transition state of amide hydrolysis and to induce γ-turn conformations. Azasulfurylpeptides are being used in applications as enzyme inhibitors and allosteric modulators. 

Figure 2. Azapeptide and azasulfurylpeptides induce turns by electronic interactions

Prolines, amino lactams and azabicyclo[X.Y.0]alkanone amino acids

Figure 3. 5-tert­-butylprolyl, azapipecolyl, α-amino γ-lactam, α-amino β-hydroxy γ-lactam, hydroxyl indolizidinone and hydroxyl pyrrolizidinone residue targets.

Local constraints that restrict rotation about bond angles in a single amino acid or between two amino acids in a dipeptide can have significant consequences on the global conformation of the entire peptide (Figure 3).  For example, natural proline favors turn geometry and can order the alignment of amide bonds of distant amino acid residues in the peptide chain to form hairpin and β-sheet conformations. Studying various analogs of proline and its homolog pipecolic acid, we have employed the steric and electronic interactions of ring substituents to favour specific ring puckering and dihedral angle geometry. For example, 5-tert-butylproline was conceived to favour prolyl cis-amide isomers and shown to stabilize type VI β-turns in model peptides. Aza-pipecolic acid analogs are also being used as cis-amide surrogates, and have been employed to create mimics of the second mitochondria-derived activator of caspases (Smac) with potential anti-cancer activity.

In contrast to the proline heterocycle which possesses a covalent bridge between the nitrogen and α-carbon of the same amino acid, a so-called “Freidinger-Veber” amino lactam ring joins the α-carbon of one amino acid to the nitrogen of its C-terminal residue. Pioneering the applications of 5- and 6-membered cyclic sulfamidate electrophiles to make β- and γ-substituted amino acids, we have developed methods for introducing α- and β-amino γ-lactams, as well as α-amino β-hydroxy γ-lactams into peptides, and are currently exploring the impact of their configuration and substitution patterns on conformation and biology.  Studying aza variants of amino lactams in cyclic urea analogs, methods are being innovated for the introduction of side chains onto the heterocycle and the scanning of peptide structures. Conformational analysis of N-amino-imidazolidin-2-one model peptides has illustrated dynamic chirality about the α-nitrogen enabling mimicry of both type II and II’ conformers.  

Azabicyclo[X.Y.0]alkanone amino acids are constrained dipeptides in which the central amide nitrogen is bridged to the α-carbons of both its N- and C-terminal amino acid residues. Combining the attributes of proline and α-amino lactam structures described above, azabicycloalkanone amino acids can control peptide folding and serve as rigid platforms for orienting pharmacophores. Innovating synthetic methods to access these heterocycles with control of ring size, stereochemistry and ring substituents, we have conceived a stereo-controlled ring-closure metathesis–transannular lactam cyclization strategy that gives functionalized azabicyclo[X.Y.0]alkanone amino acids possessing different heterocycle ring sizes (e.g., 5,5-, 5,6-, 6,4-, 6,5-, 6,6-, and 7,5-fused systems) from their 8–10-membered dipeptide lactam precursors. This approach is currently being employed to constrain the backbone and side chain geometry of various peptides.

Privileged structures diazepine and triazepine mimics

Studying heterocycle synthesis using various methods including continuous flow chemistry and solid-phase synthesis, we are particularly interested in heterocycle frameworks that serve as ligands for diverse receptors, so called “privileged structures”.  Inspired by our recent crystallographic studies, we are studying diaze- and triazepinones as mimics of γ- and β-turn conformations. Employing these heterocycles in de novo designs to replicate the purported γ-turn conformer of the Trp-Lys-Tyr triad that is common to the two endogenous peptide ligands of the urotensin II receptor, we have conceived modulators that were shown in collaboration with Professor David Chatenet (INRS) to differentiate receptor-mediated vasoconstriction offering unique utility to study selectively physiological regulation of various organ systems, particularly the cardiovascular system.

Figure 4. Benzotriazepinone and pyrrolodiazepinone Urotensin receptor modulators

Macrocyclic peptides and azacyclopeptides

Macrocyclic peptides attract interest for drug development, because they can stabilize active conformers, and improve binding affinity, selectivity, stability and cell-membrane permeability. Diverse macrocyclic peptide chemistry is being explored to provide preclinical candidates with favourable properties. Developing synthetic methods to enhance macrocycle diversity with control over geometry, charge and functional groups to probe structure-activity relationships (SAR) and to improve biological activity, in 2017, we reported synthesis of azacyclopeptides by Cu-catalyzed Mannich addition on azaPra residues using the so-called 'A -coupling' of aldehyde, alkyne and amine components (Scheme 1). Cyclization was notably favored likely due to the aza-residue promoting turn geometry and the copper catalyst bringing together the acetylene and amine prior to ring closure. Suited for creating diversity, this multicomponent process provides an acetylene that can serve to make related macrocycles, and a tertiary amine that may improve bioavailability and enhance receptor recognition.

3

Scheme 1. A - Macrocyclization for azacyclopeptide synthesis

3

Allosteric modulation of receptors using novel peptides and peptidomimetics

Innovating methods for conceiving peptide and peptidomimetic drug candidates in collaborations with biochemists, pharmacologists and physicians, we have significant interest in allosteric modulators (Figure 4). By binding receptors at sites spatially distinct from the native ligand “orthosteric” binding site, allosteric ligands may reduce side effects, due to their potential to act selectivity on a subset of signaling pathways activated by the receptor without disruption of normal rhythms of endogenous ligand release and action.  In collaboration with Professor Sylvain Chemtob (Pharmacology), we are pioneering the design of allosteric modulators of membrane proteins based on their own structures.

For example, by utilizing the extra-cellular loop regions of G protein coupled receptors, we identified peptide leads that modulate the prostaglangin-F2α and vasopressin receptors (FP and V2) by allosteric mechanisms of action. Our methods led to the conception of the peptide PDC31 and small molecule peptidomimetics which delay labour for >24 hours in induced mice. Pursuing drugs to prevent premature birth, an unmet-medical need with the highest per patient cost, PDC31 is ready for phase II clinical trials, after demonstrating to be safe in reducing intrauterine pressure and pain associated with excessive uterine contractility in a phase Ib study of women patients with primary dysmenorrhea. Ongoing study of indolizidinone and related azapeptide peptide mimics of PDC31 has revealed allosteric mechanisms of action that implicate respectively negative and positive modulation of bias signalling on pathways leading to microtubule formation and FP expression in the presence of prostaglandin F2α.

Extending this method to cytokine receptors, we have identified peptides that modulate interleukin receptors. For example, the peptide 101.10A was conceived based on a loop of the accessory protein of the interleukin-1 receptor and shown to negatively modulate the latter by an allosteric mechanism featuring biased signalling. In various studies, 101.10A has exhibited promising anti-inflammatory activity and therapeutic potential for a variety of indications including preterm birth and ischemic retinopathy.  With 101.10A as lead candidate, we are studying methods to convert this peptide into peptidomimetics with improved potency and pharmacokinetic properties.

Modulators are also being studied to regulate the cluster of differentiation 36 receptor (CD36) which is expressed on macrophage plasma membranes and serves key roles in recognition and phagocytosis as a scavenger of multiple-ligands including oxidized long chain fatty acids and oxidized low density lipoproteins. Targeting CD36 to regulate macrophage-driven inflammation, selective linear and cyclic azapeptides are being pursued to ultimately find cures for age-related macular degeneration, the leading cause of adult blindness, and atherosclerosis, the main underlying cause of ischemic heart disease and related cardiovascular complications including acute myocardial infarction and stroke. In collaboration with Professor Huy Ong (Pharmacy), azapeptides have been identified that bind CD36 selectively, diminish oxidized lipid uptake, slow phagocytosis of photoreceptor disks, attenuate production of reactive oxygen species and inflammatory mediators, block hyper-proliferation of choroidal endothelial cells and curb angiogenesis. The azacyclopeptides have distinguished themselves as a next generation of first-in-class candidates because they exhibit unprecedented CD36 binding affinity and activity suppressing nitric oxide production, and pro-inflammatory cytokine and chemokine release induced by Toll-like receptor agonists. 

Figure 4. Peptide and peptidomimetic allosteric receptor modulators