Discovery of de novo peptides by mRNA and RaPID ...

Introduction

Peptides are emerging therapeutics, filling the gap between small molecular drugs and antibodies 1, 2. Historically, peptide drugs are developed from naturally occurring bioactive peptides such as peptide hormones and peptidic natural products [3]. The structure, which is composed of proteinogenic amino acids (pAAs), of native peptide hormones has resulted in poor pharmacokinetic properties, such as low metabolic stability and limited routes of administration. Structural modifications are therefore generally required in developing these molecules into therapeutics, as illustrated by the well-established drugs such as insulin detemir and semaglutide 4, 5. On the other hand, natural product peptides often consist of various non-proteinogenic structures in terms of sidechains, backbones, and even peptide scaffolds [6]. These structural features not only contributed to their potent bioactivities but also confer favorable pharmacokinetic properties, including proteolytic resistance and oral availability.

Aside from the serendipitous discovery, de novo target-directed bioactive peptide discovery has been of growing interest recently. Advances in high-throughput peptide screening technologies have enabled the discovery of de novo peptides against biologically relevant protein targets, including the previously undruggable ones [7]. A growing number of peptide leads are identified via representative technologies, including chemically synthesized one-bead, one-compound libraries, DNA-encoding libraries (DELs), split-intein circular ligation of peptides and proteins, and peptide display methods 8, 9, 10, 11, 12. Among the de novo peptide discovery approaches, message RNA (mRNA) display technology is featured by its tremendous library size (typically more than 1012 molecules) from randomized template mRNA 13, 14, 15. Facilitated by the mRNA-templated translation in the in vitro reconstituted system, non-proteinogenic building blocks could be incorporated into peptides via the direct addition of pre-acylated tRNAs while omitting the corresponding occupied codons (Figure 1a). Notably, Suga and coworkers have developed the random nonstandard peptides integrated discovery (RaPID) system, which integrates in vitro genetic code reprogramming and mRNA display technology [16]. In the RaPID system, non-proteinogenic amino acids (npAAs) are introduced by acyl-tRNA prepared by flexizyme, an aminoacylating ribozyme, which drastically increases the structural diversity of the peptide library [17]. The versatility of the RaPID system is reflected in its broad use in the rapid and facile discovery of de novo nonstandard peptide relevant in drug development.

Advances in technologies such as the RaPID system and orthogonal aminoacyl-tRNA synthetases (ARSs) have enabled the incorporation of various unnatural building blocks into displayed peptides 18, 19. These include, but are not limit to, the incorporation of N-methyl- [20], N-alkyl- [21], β- 22, 23, 24, γ- [25], D-amino acids 26, 27, as well as hydroxy 28, 29, amino carbothioic [30], and aminobenzoic acids 31, 32. Despite these developments, the chemical space accessible through mRNA display remains limited due to two major challenges. First, the tRNA charged with the intended building block is not always obtainable using current methods such as flexizyme, engineered ARSs, or chemical synthesis. For example, γ-amino acid-charged tRNA undergoes rapid intramolecular lactam formation via nucleophilic attack of the γ-amino group, preventing its ribosomal incorporation unless the γ-amino group is masked [33]. Second, even when the desired acyl-tRNA is available, incorporation of npAAs or other modifications is constrained by the translation machinery itself (Figure 1b). The ribosome must catalyze condensation of these unnatural building blocks, which often occurs with limited success. While translation initiation can tolerate certain unnatural building blocks, such as aromatic oligoamide foldamers, elongation is far less permissive. Although engineered ribosomes have occasionally enabled translation of specific npAAs, their incorporation typically suffers from modest or poor efficiency [34].

Overcoming these challenges is critical, as expanding the chemical space in de novo peptide discovery not only broadens the accessible sequence diversity but also imparts more drug-like properties to target peptide ligands. Although mRNA display technology enables rapid identification of high-affinity peptide ligands against therapeutic targets, the resulting peptides often require extensive structural optimization, as exemplified by MK-0616 [35] and zilucoplan [36]. To address these limitations, the displayed peptides can be chemically and/ or enzymatically derivatized after the translation event to introduce nonstandard building blocks and backbone structures that are inaccessible to ribosomal machinery. Those post-translational modification (PTM) approaches have emerged as a promising alternative to ribosomal incorporation, offering access to an expanded chemical space (Figure 1c). The combination of such approaches with the in vitro displays, particularly the RaPID display that enables the encoding of npAAs , expand chemical diversity with a nearly infinite possibility of choice of building blocks in the format of phenotype-genotype conjugates. Besides mRNA and RaPID display, such methods have also been applied to nucleotide-encoded mass libraries peptide selections, including phage display, to generate highly diversified libraries 11, 37, 38. In this review, we summarize recent advances in enzymatic, chemical, and synergistic PTM strategies applied to mRNA display, and we discuss emerging technologies that hold potential for integration with this platform.

Access through your organization

Check access to the full text by signing in through your organization.

Access through your organization

Section snippets

Integration of mRNA display with post-translational modifying enzymes

PTM of proteins, such as phosphorylation, glycosylation, and prenylation, play essential roles in several cellular processes. In secondary metabolism, peptidic natural products exhibit remarkable structural diversity, often incorporating non-proteinogenic elements. In the biosynthesis of ribosomally synthesized and post-translationally modified peptides (RiPPs), the translation machinery produces a precursor peptide that undergoes extensive enzymatic tailoring to introduce exotic structural

Investigation on substrate scopes of post-translational modifying enzymes for mRNA display

The remarkable substrate promiscuity of PTM enzymes, while advantageous, introduces technical challenges for mRNA display. Ideally, every member of a designed library should undergo quantitative modification; therefore, accurately assessing enzymatic activity across large libraries is critical for effective design.

For many PTM enzymes, overall peptide sequence exerts minimal influence on modification efficiency, which is primarily determined by residues flanking the modification site. To

De novo discovery of pseudo-natural product ligands in the mRNA display integrated with post-translational modifying enzymes

A comprehensive understanding of the substrate scope of enzymatic PTMs has enabled the construction of large libraries of highly modified peptides for mRNA display workflows. Among these modifications, prenylation is particularly valuable because it enhances molecular lipophilicity — a property typically lacking in unmodified peptides. Goto, Suga, and colleagues characterized the substrate scope of KgpF, a Trp-C3-prenyltransferase from the kawaguchipeptin biosynthetic pathway, and used it to

Integration of chemical post-translational modifications with mRNA display

In addition to enzymatic PTMs, chemical and chemo-enzymatic strategies serve as complementary tools to diversify the chemical space available to mRNA display libraries. The chemical PTM strategies offer greater versatility in substrate scope (Table 1). In principle, these methods can be applied to virtually any peptide library and even combined with other chemical approaches to expand structural diversity. However, because the mRNA and cDNA components in mRNA display systems are highly

Chemical post-translational modifications of in vitro expressed non-proteinogenic amino acid-containing peptide

Ribosomal translation requires acyl-tRNA, which can be generated either in situ using an aminoacyl-tRNA synthetase (AARS)/tRNA pair or prepared separately through tools such as flexizyme. However, in some cases, desired acyl-tRNAs are difficult to access using the current methods, or even if acyl-tRNAs are accessible, some of them are inefficiently utilized by the ribosomal machinery.

To overcome these limitations, one strategy involves incorporating npAAs via genetic code reprogramming as

Chemo-enzymatic post-translational modifications and their applications in mRNA display

While enzymatic PTM coupled with mRNA display enables the discovery of diverse de novo pseudo-natural product peptides, it requires a precise definition of substrate scope and careful optimization of enzyme cocktails. In certain cases, combining chemical and enzymatic strategies offers a more versatile approach than relying on a single PTM method.

Goto, Suga, and colleagues utilized the cyclodehydratase PatD, a key enzyme in azoline-containing cyanobactin biosynthesis, to synthesize a

Outlook

mRNA, particularly RaPID, display has emerged as a powerful platform for the rapid discovery of de novo peptides targeting a wide range of therapeutically relevant proteins, including those previously considered ‘undruggable’. However, its reliance on ribosomal translation imposes significant constraints on the chemical diversity of accessible peptides. In this review, we highlight strategies that leverage PTMs to overcome these limitations. The successful identification of numerous complex

Funding

This work was supported by the Fundamental Research Funds for the Central Universities (590226004 to Y.Z.), the Start-up Foundation of Hubei Hongshan Laboratory, China (105-11020132 to Y.Z.), and the Japan Society for the Promotion of Science (JSPS) Grant-in-Aid for Specially Promoted Research (JP20H05618 to H.S.).

CRediT authorship contribution statement

Yuchen Zhang: Writing – reviewing & editing. And Hiroaki Suga: Writing – reviewing & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

© 2026 Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.