Ribosomal incorporation of negatively charged d-α- and N-methyl-l-α-amino acids enhanced by EF-Sep

Abstract

Ribosomal incorporation of d-α-amino acids (dAA) and N-methyl-l-α-amino acids (^MeAA) with negatively charged sidechains, such as d-Asp, d-Glu, ^MeAsp and ^MeGlu, into nascent peptides is far more inefficient compared to those with neutral or positively charged ones. This is because of low binding affinity of their aminoacyl-transfer RNA (tRNA) to elongation factor-thermo unstable (EF-Tu), a translation factor responsible for accommodation of aminoacyl-tRNA onto ribosome. It is well known that EF-Tu binds to two parts of aminoacyl-tRNA, the amino acid moiety and the T-stem; however, the amino acid binding pocket of EF-Tu bearing Glu and Asp causes electric repulsion against the negatively charged amino acid charged on tRNA. To circumvent this issue, here we adopted two strategies: (i) use of an EF-Tu variant, called EF-Sep, in which the Glu216 and Asp217 residues in EF-Tu are substituted with Asn216 and Gly217, respectively; and (ii) reinforcement of the T-stem affinity using an artificially developed chimeric tRNA, tRNA^Pro1E2, whose T-stem is derived from Escherichia coli tRNA^Glu that has high affinity to EF-Tu. Consequently, we could successfully enhance the incorporation efficiencies of d-Asp, d-Glu, ^MeAsp and ^MeGlu and demonstrated for the first time, to our knowledge, ribosomal synthesis of macrocyclic peptides containing multiple d-Asp or ^MeAsp.

This article is part of the theme issue ‘Reactivity and mechanism in chemical and synthetic biology’.

1. Introduction

d-α-amino acids (dAA) and N-methyl-l-α-amino acids (^MeAA) are often found in natural bioactive peptides. dAA and ^MeAA contribute to biological and pharmacological properties of peptides such as high proteolytic stability, target affinity and membrane permeability [1–13], and therefore are practical components for development of novel peptide drugs. Although natural bioactive peptides bearing dAA and ^MeAA are generally synthesized by multienzyme clusters called non-ribosomal peptide synthetases [14,15] or post-translational modification enzymes [16,17], direct ribosomal incorporation of diverse dAA and ^MeAA into nascent peptides has also been demonstrated to date by using artificially charged dAA-transfer RNA (tRNA) or ^MeAA-tRNA [18–28]. However, incorporation of dAA and ^MeAA with negatively charged sidechains, such as d-Asp, d-Glu, ^MeAsp and ^MeGlu (figure 1a), is much less efficient than those of neutral or positively charged ones. Fujino et al. [20] evaluated the incorporation of 19 dAA derived from the 19 proteinogenic l-α-amino acids except for Gly using an Escherichia coli tRNA^Asn-based orthogonal tRNA, tRNA^AsnE2. Unfortunately, neither d-Asp nor d-Glu could be successfully incorporated into a model peptide, whereas the other dAA could be incorporated except for d-Pro, d-Lys and d-Ile. Later on, Achenbach et al. [21] showed a possibility of d-Asp incorporation into model peptides using tRNA^Gly, but identities of the peptides containing d-Asp were not verified in their study. Kawakami et al. [27] evaluated incorporation of 19 ^MeAA derived from the 19 proteinogenic l-amino acids, except for Pro, using an orthogonal tRNA, tRNA^AsnE1; however, neither ^MeAsp nor ^MeGlu could be introduced.

Figure 1.

Structures of negatively charged dAA and ^MeAA, EF-Tu, EF-Sep and tRNA used in this study. (a) Structures and abbreviations of negatively charged dAA and ^MeAA used in this study. (b) Structure of E. coli EF-Tu bound to Saccharomyces cerevisiae Phe-tRNA^Phe (PDB: 1OB2). (c) Structure of the amino acid binding pocket of E. coli EF-Tu (PDB ID: 1OB2). 3'-End of S. cerevisiae Phe-tRNA^Phe bound to the pocket is shown in orange. Mutations in EF-Sep are indicated by arrows. (d) Secondary structure of tRNA^Pro1E2_CGG and tRNA^Pro1_CGG used for incorporation of dAA and ^MeAA at the CCG codon. The T-stem motif derived from E. coli tRNA^Glu for improved EF-Tu binding is indicated by red. The D-arm motif for tight EF-P binding consists of a 9-nt D-loop closed by a stable 4 bp stem (indicated by blue). See also the electronic supplementary material, table S1 for the sequences. (Online version in colour.)

A reason for their low incorporation efficiency could be attributed to the low binding affinity of aminoacyl-tRNA to elongation factor-thermo unstable (EF-Tu), a translation factor responsible for accommodation of aminoacyl-tRNA onto the ribosome A site. It has been reported that EF-Tu recognizes two parts of aminoacyl-tRNA: the amino acid moiety and the T-stem (figure 1b) [29–31]. Since the amino acid binding pocket of EF-Tu has two negatively charged amino acids, E216 and D217, the affinity of negatively charged aminoacyl-tRNA to this site should be lower than those bearing neutral or positively charged ones owing to the electric repulsion. Uhlenbeck and colleagues showed that E. coli tRNA^Glu has the strongest binding affinity to EF-Tu among the 21 E. coli tRNAs tested in their study [30]. For instance, the binding energies (ΔG°) of Val-tRNA^Glu and Val-tRNA^Gln to EF-Tu were −11.7 and −8.3 kcal mol⁻¹, respectively. This is because the weak affinity of l-Glu to the amino acid binding pocket needs to be compensated for by the strong affinity of the T-stem region of tRNA^Glu [32–36]. In fact, mischarged Glu-tRNA^Gln is poorly accommodated onto the ribosome A site owing to the insufficient EF-Tu affinity, resulting in inefficient peptide synthesis [37]. Moreover, in the case of d-Asp, d-Glu, ^MeAsp and ^MeGlu, their d-configuration and N-methylation should further hinder their binding to EF-Tu because of their structural incompatibility. Iwane et al. [38] determined the binding affinities of Phe-tRNA^AsnE2#3 and ^MePhe-tRNA^AsnE2#3 to EF-Tu to be −9.3 and −7.9 kcal mol⁻¹, respectively, indicating that N-methylation significantly reduced the affinity of aminoacyl-tRNA (note that tRNA^AsnE2#3 is a variant of tRNA^AsnE2 whose T-stem was substituted with that of tRNA^Glu). These results indicate that compensation of the T-stem affinity is required for incorporation of ^MeAA in general. In fact, Iwane et al. [38] have succeeded in enhancing incorporation of various ^MeAA by optimizing T-stem structure to reinforce their affinity. Yet, the enhancement of incorporation of ^MeAsp and ^MeGlu remains elusive by such tRNA engineering.

In order to enhance the affinity of negatively charged amino acids, EF-Tu variants bearing mutations around the amino acid binding pocket have been developed in several previous studies. Park et al. [39] reported that an engineered EF-Tu, where six mutations around the binding pocket (figure 1c, H67R, E216N, D217G, F219Y, T229S and N274W) were introduced, was able to bind a suppressor tRNA charged O-phosphoserine (Sep). This EF-Tu mutant was referred to as EF-Sep. Using similar approaches introducing mutations to EF-Tu, EF-pY and EF-Sel1, containing three (E216V, D217G and F219G) and five mutations (H67R, Q98W, E216N, D217K and N274R), respectively, have also been developed for enabling the introduction of l-phosphotyrosine (pTyr) and selenocysteine (Sec), respectively [40,41]. All of these EF-Tu variants have substitutions of the negatively charged E216 and D217 with neutral or positively charged residues. Although these variants have been shown to enhance incorporation of Sep, pTyr and Sec, they were not applied to incorporation of other non-proteinogenic amino acids to the best of our knowledge. On the basis of these preceding studies, here we took advantage of EF-Sep in combination with the T-stem affinity reinforcement strategy to achieve the incorporation of negatively charged dAA and ^MeAA into nascent peptides.

2. Material and methods

(a) . In vitro transcription of transfer RNAs and flexizymes

DNA templates used for transcription of tRNAs and flexizymes (dFx and eFx) were prepared by extension and polymerase chain reaction (PCR) using the forward and reverse primer pairs summarized in the electronic supplementary material, table S1. Transcription of tRNAs and flexizymes was carried out overnight at 37°C in 4 ml of the following reaction mixture: 40 mM Tris–HCl (pH 8.0), 5 mM nucleoside triphosphate (NTP) mix, 22.5 mM MgCl₂, 1 mM dithiothreitol, 1 mM spermidine, 0.01% Triton X-100, 0.04 U µl⁻¹ RNasin RNase inhibitor (Promega) and 0.12 µM T7 RNA polymerase. In the case of tRNA preparation, 5 mM guanosine monophosphate was added to the above mixture and the concentration of NTP mix was reduced to 3.75 mM. The transcribed RNAs were treated with RQ1 DNase (Promega) for 30 min at 37°C and purified by polyacrylamide gel electrophoresis using 8% (tRNAs) or 12% (flexizymes) polyacrylamide gel containing 6 M urea.

(b) . Preparation of aminoacyl-transfer RNAs

d-Asp, d-Glu, ^MeAsp, ^MeGlu and d-Cys were pre-activated as 3,5-dinitrobenzylester (DBE) and ^ClAcd-Phe was pre-activated as a cyanomethyl ester according to the previously reported methods [20,42,43]. dFx was used for charging amino acids-DBE onto tRNA^Pro1E2; eFx was used for preparation of ^ClAcd-Phe-tRNA^fMet. The reactions were performed at 4°C for 2, 2, 24, 6, 6 and 2 h for aminoacylation of d-Asp, d-Glu, ^MeAsp, ^MeGlu, d-Cys and ^ClAcd-Phe, respectively, in the following reaction mixture: 50 mM HEPES-KOH (pH 7.5), 600 mM MgCl₂, 20% dimethyl sulfoxide, 25 µM eFx or dFx, 25 µM tRNA and 5 mM activated amino acid. The resulting aminoacyl-RNAs were recovered by ethanol precipitation and washed with 70% ethanol.

(c) . Ribosomal synthesis of model peptides

Ribosomal synthesis of model peptides was carried out at 37°C for 30 min in a custom-made flexible in vitro translation (FIT) system of the following composition [44]: 50 mM HEPES-KOH (pH 7.6), 100 mM potassium acetate, 12.8 mM magnesium acetate, 2 mM ATP, 2 mM GTP, 1 mM CTP, 1 mM UTP, 20 mM creatine phosphate, 0.1 mM 10-formyl-5,6,7,8-tetrahydrofolic acid, 2 mM spermidine, 1 mM dithiothreitol, 1.5 mg ml⁻¹ E. coli total tRNA, 1.2 µM E. coli ribosome, 0.6 µM methionyl-tRNA formyltransferase, 2.7 µM IF1, 3 µM IF2, 1.5 µM IF3, 0.1 µM EF-G, 20 µM EF-Tu, 5 µM EF-Sep, 20 µM EF-Ts, 5 µM EF-P, 0.25 µM RF2, 0.17 µM RF3, 0.5 µM RRF, 4 µg ml⁻¹ creatine kinase, 3 µg ml⁻¹ myokinase, 0.1 µM inorganic pyrophosphatase, 0.1 µM nucleotide diphosphate kinase, 0.5 µM DNA template, 0.1 µM T7 RNA polymerase, 0.13 µM AspRS, 0.09 µM GlyRS, 0.11 µM LysRS, 0.03 µM MetRS, 0.02 µM TyrRS, 0.5 mM Asp, 0.5 mM Gly, 0.5 mM Lys, 0.5 mM Met, 0.5 mM Tyr and 20 µM each pre-charged aminoacyl-tRNAs. The concentrations of EF-Tu and EF-Sep were modified accordingly as mentioned in the results section. DNA templates were transcribed into messenger RNA (mRNA) by T7 RNA polymerase included in the translation system, and then translated into peptides. Preparation of the DNA templates were performed by extension and PCR using the corresponding forward and reverse primer pairs (electronic supplementary material, table S1). In the case of translation of macrocyclic peptide P3, 10-formyl-5,6,7,8-tetrahydrofolic acid and Met were removed from the above reaction mixture, where 12.5 µM ^ClAcd-Phe-tRNA^fMet was added for reprogramming the initiator AUG codon.

(d) . Matrix-assisted laser desorption/ionization-time of flight mass spectrometric analysis of ribosomally synthesized model peptides

Of the above translation reaction mix, 7.5 µl was diluted with 7.5 µl of 2× tris-buffered saline (TBS) buffer (100 mM Tris–HCl (pH 7.6) and 300 mM NaCl), mixed with 5 µl slurry of anti-Flag M2 affinity gel (Sigma), and incubated with gentle rotation for 15 min at room temperature. The peptides on the anti-Flag gel were washed twice with 40 µl of 1× TBS buffer (50 mM Tris–HCl (pH 7.6) and 150 mM NaCl) and eluted from the gel by adding 30 µl of 0.2% trifluoroacetic acid. The eluent was desalted with SPE C-tip (Nikkyo Technos) and co-crystallized with α-cyano-4-hydroxycinnamic acid on a sample plate. The samples were analysed by UltrafleXtreme (Bruker Daltonics) in a reflector/positive mode, where peptide calibration standard II (Bruker Daltonics) was used for external mass calibration.

(e) . Quantification of ribosomally synthesized peptides by tricine sodium dodecyl sulfate-polyacrylamide gel electrophoresis and autoradiography

For quantification of model peptides, the translation reaction was carried out in the presence of 0.05 mM [¹⁴C]-Asp instead of 0.5 mM cold [¹²C]-Asp so that the peptides were radiolabelled. Then, 2.5 µl of the translation reaction mixture was mixed with the same volume of stop solution (0.9 M Tris–HCl (pH 8.45), 8% SDS, 30% glycerol and 0.001% xylene cyanol) and incubated at 95°C for 3 min. The samples were then subjected to 15% tricine sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS–PAGE) and analysed by autoradiography using a Typhoon FLA 7000 (Cytiva). Expression levels of peptides were normalized by the intensity of the [¹⁴C]-Asp band.

3. Results

(a) . Single incorporation of negatively charged d-α- and N-methyl-l-α-amino acids into a model peptide

d-Asp, d-Glu, ^MeAsp and ^MeGlu were introduced at a CCG codon of an mRNA, mR1, to synthesize peptide P1 (figure 2a). These amino acids were chemically esterified with 3,5-dinitrobenzyl alcohol, charged onto tRNA^Pro1E2 _CGG using a flexizyme variant, called dFx [45] and added to the translation system. tRNA^Pro1E2 is an artificially developed chimeric tRNA based on E. coli tRNA^Pro1 whose T-stem is replaced with that of E. coli tRNA^Glu [23] (figure 1d). Therefore, we can expect high binding affinity of the T-stem to EF-Tu and EF-Sep to promote accommodation of aminoacyl-tRNA^Pro1E2. Note that the previously reported binding energy values (ΔG°) of Val-tRNA^Glu and Val-tRNA^Pro to EF-Tu were −11.7 and −9.3 kcal mol⁻¹, respectively [30]. In addition, the D-arm motif derived from tRNA^Pro1 efficiently recruits EF-P and thereby promotes the peptidyl transfer reaction [46]. We have previously shown that the use of tRNA^Pro1E2 enables efficient incorporation of not only l-Pro but also various non-proteinogenic substrates such as dAA, α,α-dialkyl-amino acids, β-amino acids, γ-amino acids, α-aminoxy acids and α-hydrazino acids in the presence of EF-P [23,47–51].

Figure 2.

Single incorporation of negatively charged dAA or ^MeAA into a model peptide. (a) Sequences of an mRNA (mR1) and the corresponding peptide (P1). Negatively charged dAA or ^MeAA, indicated by Xaa, were introduced at the CCG codon using pre-charged aminoacyl-tRNA. The amino acid sequence of ‘flag’ is Asp-Tyr-Lys-Asp-Asp-Asp-Asp-Lys. (b) MALDI-TOF MS of ribosomally synthesized P1 peptides, P1-Xaa, containing d-Asp, d-Glu, ^MeAsp or ^MeGlu. tRNA^Pro1E2_CGG was used for incorporation of these amino acids at the CCG codon. Translation was carried out in the presence of 5 µM EF-Sep. Red arrows indicate monovalent ions ([M + H]⁺) of the desired products. ‘obsd.’ and ‘calcd.’ indicate observed and calculated m/z values, respectively. (c) Expression levels of P1-Xaa containing d-Asp, d-Glu, ^MeAsp or ^MeGlu. tRNA^Pro1E2_CGG or tRNA^Pro1_CGG was used for incorporation of these amino acids at the CCG codon. Translation reactions were carried out in the presence or absence of 5 µM EF-Sep. Numbers above the bars represent the ratio of the two bars. p-values were estimated by Student's t-test and indicated by asterisks (n = 3). See the electronic supplementary material, figure S1a for the raw data of tricine SDS–PAGE. (Online version in colour.)

Translation of the peptide was conducted in a reconstituted E. coli in vitro translation system, referred to as the FIT system [44], customized for inclusion of five proteinogenic amino acids (Met, Tyr, Lys, Asp and Gly) and the corresponding aminoacyl-tRNA synthetases (ARSs); other proteinogenic amino acids and their ARSs were omitted. We have previously reported that an optimization of IF2, EF-G, EF-Tu and EF-P concentrations is critical for efficient incorporation of dAA, such as d-Ala, d-Ser, d-Cys and d-His [22,23]. Therefore, we applied the same concentrations of them as the previous reports (3 µM IF2, 0.1 µM EF-G, 20 µM EF-Tu and 5 µM EF-P). Note that the concentration of EF-Sep in this system was 5 µM unless otherwise indicated; both EF-Tu and EF-Sep were present in this FIT system. The resulting products were analysed by matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry (MS), showing that all substrates could be successfully incorporated into the full-length desired peptides (figure 2b, P1-d-Asp, P1-d-Glu, P1-^MeAsp and P1-^MeGlu).

To evaluate the expression level, peptide P1 was translated in the presence of [¹⁴C]-labelled Asp in place of cold [¹²C]-Asp, separated by tricine SDS–PAGE and quantified by autoradiography (figure 2c; electronic supplementary material, figure S1a). In order to confirm the significance of the reinforced T-stem affinity, tRNA^Pro1E2 was compared with tRNA^Pro1 in incorporation of d-Asp, d-Glu, ^MeAsp and ^MeGlu in the absence of EF-Sep (figure 2c, red striped and blue striped bars for tRNA^Pro1E2 and tRNA^Pro1, respectively; electronic supplementary material, figure S1a). The expression levels of P1-d-Asp, P1-d-Glu, P1-^MeAsp and P1-^MeGlu were 0.04 µM, 0.10 µM, 0.14 µM and 0.04 µM, respectively, when using tRNA^Pro1E2. These values were 7.1-fold, 1.7-fold, 1.7-fold and 2.2-fold higher than those synthesized using tRNA^Pro1. These results indicate that the use of tRNA^Pro1E2 significantly improved the translation efficiency compared to the use of tRNA^Pro1 (p < 0.05), showing the importance of the T-stem motif of tRNA^Pro1E2 derived from tRNA^Glu. Then, we compared the expression levels in the presence (5 µM) and absence (0 µM) of EF-Sep using tRNA^Pro1E2. Consequently, the expression levels of P1-d-Asp, P1-d-Glu, P1-^MeAsp and P1-^MeGlu were 0.07 µM, 0.16 µM, 0.18 µM and 0.06 µM, respectively, in the presence of EF-Sep, which were 1.9-fold, 1.5-fold, 1.3-fold and 1.7-fold higher than those in the absence of EF-Sep (figure 1c, red solid and striped bars for 5 µM and 0 µM EF-Sep; electronic supplementary material, figure S1a). Because the p-values for the fold changes were smaller than 0.05 for P1-d-Asp, P1-d-Glu and P1-^MeGlu, we concluded that EF-Sep significantly improved the incorporation efficiencies of these amino acids. On the other hand, the enhancement of P1-^MeAsp expression by EF-Sep was not statistically significant in this analysis (1.3-fold, p-value > 0.05).

(b) . Consecutive incorporation of ^MeAsp into a model peptide

We assumed that the single incorporation of these amino acids is efficient enough when using tRNA^Pro1E2 in the presence of EF-P, and therefore we could not observe a large enhancement effect of expression levels by addition of EF-Sep, particularly in the case of ^MeAsp incorporation. Therefore, we next evaluated the effect of EF-Sep on consecutive ‘double’ incorporation of ^MeAsp into a model peptide P2 that has consecutive CCG codons (figure 3a). Since the consecutive incorporation of ^MeAsp requires more frequent accommodation of ^MeAsp-tRNA, we assumed that the enhancement effect of EF-Sep would be more clearly observed. As a comparison, incorporation of ^MeAla, an N-methyl-amino acid with a neutral sidechain, into P2 was also evaluated. Clean expression of P2-^MeAsp and P2-^MeAla was confirmed by MALDI-TOF MS in the presence of 5 µM EF-Sep (figure 3b). Then, we quantified their expression with titration of EF-Sep or EF-Tu concentration (figure 3c). The expression levels of P2-^MeAsp and P2-^MeAla in the absence of EF-Sep were 0.02 µM and 0.8 µM, respectively, showing less efficiency in consecutive incorporation of ^MeAsp than ^MeAla. Expression of P2-^MeAsp was significantly enhanced by 1.9-fold, 2.7-fold, 3.1-fold and 3.4-fold (p < 0.01 for all values) in the presence of 1, 2, 4 and 8 µM EF-Sep, respectively, whereas that of P2-^MeAla was not improved by EF-Sep. We also confirmed that addition of the same amounts of EF-Tu did not enhance the expression level of P2-^MeAsp. Since the basal EF-Tu concentration is 20 µM in this system, the addition of 1, 2, 4 and 8 µM EF-Tu makes their final concentrations 21, 22, 24 and 28 µM, respectively.

Figure 3.

Consecutive incorporation of ^MeAsp or ^MeAla into a model peptide. (a) Sequences of an mRNA (mR2) and the corresponding peptide (P2). ^MeAsp or ^MeAla was introduced at the CCG codon using pre-charged aminoacyl-tRNA^Pro1E2_CGG. (b) MALDI-TOF MS of P2-^MeAsp and P2-^MeAla. Translation was carried out in the presence of 5 µM EF-Sep. Red arrows indicate the monovalent ion ([M + H]⁺) of the desired product. ‘obsd.’ and ‘calcd.’ indicate observed and calculated m/z values, respectively. (c) Expression levels of P2-^MeAsp and P2-^MeAla. Translation was carried out in the presence of 0, 1, 2, 4 and 8 µM EF-Sep or EF-Tu. The basal EF-Tu concentration is 20 µM and the concentrations of additional EF-Sep and EF-Tu are indicated. Numbers indicate the ratio of the expression levels in the presence of EF-Sep or EF-Tu to those in the absence. p-values were estimated by Student's t-test and indicated by asterisks (n = 3). See the electronic supplementary material, figure S1b for the raw data of tricine SDS–PAGE. (Online version in colour.)

(c) . Ribosomal synthesis of model macrocyclic peptides containing multiple d-Asp or ^MeAsp

As a showcase, we next demonstrated expression of a model macrocyclic peptide, P3, containing multiple d-Asp or ^MeAsp in the presence of EF-Sep (figure 4a). N-chloroacetyl-d-phenylalanine (^ClAcd-Phe) and d-Cys were assigned at the initiator and elongator AUG codons using pre-charged ^ClAcd-Phe-tRNA^fMet_CAU and d-Cys-tRNA^Pro1E2_CAU, respectively. Note that Met was removed from the translation system in order to reprogramme the initiator and elongator AUG codons. The N-terminal ClAc and the downstream sulfhydryl groups of these d-amino acids spontaneously formed a thioether bond to give a macrocyclic structure [52]. Between these cyclizing residues, two d-Asp or ^MeAsp residues were introduced at CCG codons by using tRNA^Pro1E2_CGG. Expression of the macrocyclic peptides containing d-Asp or ^MeAsp residues were confirmed by MALDI-TOF MS (figure 4b,c, P3-d-Asp and P3-^MeAsp).

Figure 4.

Ribosomal synthesis of model macrocyclic peptides containing d-Asp or ^MeAsp. (a) Sequence of mRNA (mR3) and the corresponding peptide (P3). The thiol group of d-Cys attacks the N-terminal chloroacetyl group to form a thioether bond. (b) MALDI-TOF MS of P3-d-Asp and P3-^MeAsp. Translation was carried out in the presence of 5 µM EF-Sep. d-Asp or ^MeAsp was introduced at the CCG codon using tRNA^Pro1E2_CGG. ^ClAcd-Phe and d-Cys were introduced at the initiator and elongator AUG codons using tRNA^fMet_CAU and tRNA^Pro1E2_CAU, respectively. Red arrows indicate monovalent ions ([M + H]⁺) of the desired peptides. ‘obsd.’ and ‘calcd.’ indicate observed and calculated m/z values, respectively. (c) Structures of P3-d-Asp and P3-^MeAsp. (Online version in colour.)

4. Discussion

In this study, we showed that EF-Sep is able to promote ribosomal incorporation of dAA and ^MeAA with negatively charged sidechains, d-Asp, d-Glu, ^MeAsp and ^MeGlu, into model peptides. This is the first report, to the best of our knowledge, of application of EF-Sep to incorporation of these substrates. In addition, the reinforcement of the T-stem affinity of tRNA to EF-Tu/EF-Sep also contributed to enhancing their incorporation efficiency. Even though the observed enhancement of the single incorporation by EF-Sep was rather modest (approx. 2-fold), greater than 3-fold enhancement was observed for the consecutive incorporation. Since the E216N and D217G of EF-Sep were designed for improving the Sep incorporation, EF-Sep might not be yet optimal for the dAA and ^MeAA incorporation. Therefore, the appropriate optimization of residues in the binding pocket may further improve their incorporation efficiencies. Other EF-Tu variants, such as EF-pY and EF-Sel1, could also be applicable for the same purpose.

Our group has previously succeeded in preparation of random macrocyclic peptide libraries that have multiple dAA and/or ^MeAA and applied them to in vitro selection of inhibitor peptides against various target proteins by means of the RaPID (random non-standard peptides integrated discovery) system [44,53,54]. Here we have demonstrated the ribosomal synthesis of a model macrocyclic peptide bearing multiple d-Asp or ^MeAsp. This work adds d-Asp and ^MeAsp to the reprogrammed genetic code, increasing the diversity of available chemical space in macrocyclic peptides. This enables us to perform discovery campaigns of potent ligands containing d-Asp and ^MeAsp against protein targets of interest. Such efforts are underway in our laboratory.

Data accessibility

All the data that support the findings in this study are available within the article and the electronic supplementary material [55].

Authors' contributions

T.K.: conceptualization, funding acquisition, investigation, project administration, writing—original draft; H.S.: funding acquisition, supervision, writing—review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

The authors declare no competing financial interests.

Funding

This work was supported by the Japan Society for the Promotion of Science (JSPS), Grant-in-Aid for Scientific Research (B) (grant no. 18H02080), Grant-in-Aid for Scientific Research (A) (grant no. 22H00439), Grant-in-Aid for Challenging Research (Exploratory) (grant no. 21K18233) to T.K. and Grant-in-Aid for Specially Promoted Research (grant no. 20H05618) to H.S.