A RaPID Response to SARS-CoV-2

1 Introduction

1.1 The RaPID Platform and Macrocyclic Peptide Drugs

Genetically encoded peptide libraries have become powerful resources for de novo drug discovery.^{1, 2} Embedded within state-of-the-art display technologies, they facilitate the identification of high-affinity peptide ligands from a vast sequence space.^1-3 Chemical modification of the library, including the incorporation of non-canonical amino acids, can greatly enhance the diversity and drug-likeness of the screened peptides.^4-6 Hence, most peptide displays favour the use of modified constrained peptides over their linear counterparts,⁷ as they possess beneficial pharmaceutical properties.⁸

Macrocyclic peptides are a diverse class of molecules characterised by their structural constraint.^{9, 10} Featuring a variety of topologies,^11-14 they possess architectures particularly suited to mimic and disrupt protein-protein interactions.¹⁵ Likewise, the inherent rigidity of constrained peptides enhances their target affinity and metabolic stability.^{16, 17} The remarkable targeting and exceptional binding affinity of constrained peptides have invited comparisons with antibodies, which are renowned for these properties.^{18, 19} With a comparatively low molecular weight, however, macrocyclic peptides maintain synthetic accessibility similar to small molecules.^{11, 17} Consequently, constrained peptides strike a balance that situates them in the ‘Goldilocks zone’ between small molecules and protein therapeutics (Figure 1),^{6, 20} rendering them highly relevant for future therapeutic development.^{8, 21}

Figure 1

Open in figure viewerPowerPoint

Constrained peptides occupy the ‘Goldilocks zone’ between small molecules and biologics^{8-10, 20} (molecules not to scale; structures obtained from PDB: 7Y4G, 4DGC, 5DK3).^22-24

The RaPID (Random Nonstandard Peptides Integrated Discovery) platform²⁵ (Figure 2) facilitates the identification of selective and high-affinity binding peptide macrocycles.²⁶ RaPID elegantly integrates the FIT system (flexible in vitro translation) with mRNA display technology.²⁷ Hence, flexizymes (flexible tRNA-aminoacylating ribozymes) are used to enable the incorporation of non-canonical amino acids.²⁸ In this manner, a growing range of non-canonical amino acids have been featured, including N-methyl-, d-, β- and γ-amino acids.²⁹ While various cyclisation chemistries are compatible with mRNA display,^{1, 29, 30} in most RaPID screenings the translation begins with chloroacetylated amino acids that form a thioether linkage upon reaction with a cysteine residue.²⁸ Due to the in vitro nature of the translation, it is possible to screen libraries of over >10¹² unique sequences against immobilised targets.²⁸ Information about the enriched peptides from the affinity-based selection can be recovered by sequencing, as puromycin is used to link the translated peptide to its genetic information.^{28, 29} The RaPID platform has yielded high-affinity ligands for an ever-growing range of targets,^{10, 26, 28, 31} and is therefore thought to have the capacity to produce macrocyclic ligands for virtually any given protein.³²

Figure 2

Open in figure viewerPowerPoint

Simplified overview of the RaPID platform for high-affinity cyclic peptide ligand discovery. Note that in some instances reverse transcription is performed prior to screening.^{26, 28, 32}

1.2 SARS-CoV-2 and its Key Viral Protein Targets

The emergence of SARS-CoV-2 and the resulting COVID-19 pandemic necessitated the urgent development of vaccines, drugs and diagnostic tools.³³ Despite similarities to other epidemic coronaviruses,³⁴ SARS-CoV-2 – being a novel pathogen³⁵ – required specific solutions tailored to its distinct virology.³⁶ A wide array of medically relevant viral targets were identified,^37-42 with the majority of vaccine- and drug-related research being conducted on the spike glycoprotein (S)³⁸ and the main protease (M^pro or 3CL^pro).^{39, 40} This mini-review intends to spotlight the role of the RaPID platform for the identification of ligands for these important SARS-CoV-2 proteins.^43-46

The spike glycoprotein (Figure 3) is the largest SARS-CoV-2 protein.³⁸ The homotrimer is embedded in the viral envelope, contributing to the characteristic structure of a coronavirus.⁴⁷ It is integral to the fusion of the virus and host cell, determining host range and tropism of SARS-CoV-2.^{48, 49} Fundamental to this process is the receptor-binding domain (RBD), which binds the ACE2 receptor for cell entry.⁵⁰ Before this central interaction occurs, the spike protein is primed by host proteases.^51-53 Its function as the mediator of viral entry combined with its immunogenicity made the spike protein a common element of successful vaccine development campaigns.⁵⁴ Furthermore, amino acid substitutions in the spike are known to alter the characteristics of the virus.^{38, 55} Viral evolution has spawned multiple SARS-CoV-2 variants primarily defined by their spike mutations,^56-58 with SARS-CoV-2 Omicron EG.5, a lineage circulating in late 2023, containing about 40 amino acid replacements in the spike.⁵⁹

Figure 3

Open in figure viewerPowerPoint

Cryo-EM structure of pre-fusion SARS-CoV-2 spike protein (PDB: 6VSB).⁶⁰ (a) Surface representation. (b) Ribbon diagram of the trimeric assembly. (c) Ribbon diagram of a single protomer showing both domains.

The main protease, M^pro (Figure 4), is one of the sixteen non-structural proteins of SARS-CoV-2.⁶¹ Together with the papain-like protease, PL^pro, it is involved in the viral polyprotein processing, making it indispensable for the viral infectious cycle.⁴⁰ M^pro exhibits unique substrate specificity, markedly different from host proteases, with a clear preference for glutamine in the P₁ position (Schechter-Berger⁶² nomenclature).³⁹ M^pro has therefore been the focal point of anti-coronaviral drug discovery,⁶³ yielding approved inhibitors including nirmatrelvir⁶⁴ and ensitrelvir.⁶⁵ Because there appears to be less evolutionary pressure on M^pro, the SARS-CoV-2 Omicron lineages contain only one amino acid substitution with marginal impact on its function.^66-68

Figure 4

Open in figure viewerPowerPoint

X-ray crystal structure of dimeric SARS-CoV-2 main protease (PDB: 7DQZ).⁶⁹ (a) Surface representation. (b) Ribbon diagram showing the individual protomers and catalytic dyad residues (H41, C145). (c) Ribbon diagram highlighting the three domains.