Trna-based methods and related compositions

Abstract

The invention relates to methods of determining the effect of agents on the acylation of tRNAs, methods of determining the acylation status or efficiency of acylation of tRNAs, and methods of making polypeptides. The invention also relates to split tRNAs, fusion RNA molecules, and labelled tRNAs or portions of tRNAs. The invention further relates to nucleic acid constructs, including nucleic acids encoding circularly permutated transcripts. Furthermore, the invention relates to proteins, enzymes, and synthetases obtained by the methods or by using the products disclosed herein.

Description

[0001] tRNA-BASED METHODS AND RELATED COMPOSITIONS FIELD OF THE INVENTION The invention relates to methods of determining the effect of agents on the acylation of tRNAs, methods of determining the acylation status or efficiency of acylation of tRNAs, and methods of making polypeptides. The invention also relates to split tRNAs, fusion RNA molecules, and labelled tRNAs or portions of tRNAs. The invention further relates to nucleic acid constructs, including nucleic acids encoding circularly permutated transcripts. Furthermore, the invention relates to proteins, enzymes, and synthetases obtained by the methods or by using the products disclosed herein. BACKGROUND OF THE INVENTION The genetic code of living cells has been reprogrammed to enable the site-specific incorporation of hundreds of non-canonical amino acids (ncAAs) into proteins1,2, and the encoded synthesis of non-canonical polymers and macrocyclic peptides and depsipeptides3-6. Despite remarkable progress, the monomers that can be site- specifically incorporated into proteins in cells are essentially limited to α-L amino acids with variant side chains, and closely related hydroxy acids. While a wider range of monomers have been incorporated in in vitro translation reactions7-10– primarily into short peptides – these in vitro approaches cannot be extended to living cells. (S)β3- paraBromo-homophenylalanine ((S)β3-pBrhF) has been incorporated at very low levels in competition with phenylalanine (Phe) at Phe codons in E. coli, using forcing conditions of Phe starvation11; this approach leads to a mixture of amino acids at all Phe codons and does not enable the site -specific incorporation of (S)β3-pBrhF at a single position in response to a single codon, as required to reprogram the genetic code. The encoded, site specific, incorporation of a non-canonical monomer (ncM) via cellular translation requires both the acylation of an orthogonal tRNA with the ncM by an orthogonal synthetase, and ribosomal polymerization of the ncM into a polymer chain (Fig.1). Current methods for engineering aminoacyl-tRNA synthetases that acylate new monomers rely on translational readouts12,13and therefore require the monomers to be ribosomal substrates for incorporation, often at specific sites in proteins. Since many ncMs of interest are poor ribosomal substrates10,11,14-18, this creates an evolutionary deadlock in cells; an orthogonal synthetase cannot be evolved to acylate an orthogonal tRNA with ncMs that are poor ribosomal substrates, and ribosomes cannot be evolved to polymerize ncMs that cannot be acylated onto orthogonal tRNAs. The inventors previously described tRNA extension (tREX), a rapid and scalable method to determine the aminoacylation status of user-defined tRNAs from cells19. In this approach total tRNA is isolated from cells and the 2’,3’ diol on the ribose at the 3’ end of non-acylated tRNAs is selectively oxidized to a dialdehyde, while acylated tRNAs are protected from oxidation of the diol. A DNA probe bearing a fluorophore is then annealed to the 3’ end of the tRNA of interest, under conditions that facilitate deacylation of acylated tRNAs, to reveal the free diol at their 3’ ends. This enables the polymerase-mediated extension of non-oxidized tRNAs (that were acylated). The resulting difference in mass between oxidized, non-extended, and non-oxidized, extended tRNAs, is resolved by gel electrophoresis, allowing acylated and free tRNAs to be distinguished. Another previous method is disclosed in Saito et al. (the EMBO Journal, Vol.20, No.7, pp 1797-1806, 2001). This method relies on a biotinylated substrate and is performed in vitro. SUMMARY OF THE INVENTION In a first aspect, there is provided a method of determining the effect of a polypeptide-of-interest or nucleic-acid- of-interest on acylation of a tRNA or acylation status of a tRNA, the method comprising: i) incubating the polypeptide-of-interest or nucleic-acid-of-interest, the tRNA, and a substrate with which the tRNA could be acylated, under conditions conducive to acylation of the tRNA, wherein the tRNA is split into at least two portions, and one of said tRNA portions is present as part of a fusion RNA molecule also comprising a sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest; ii) exposing the tRNA to conditions capable of labelling tRNAs that have been acylated; and iii) identifying whether the polypeptide-of-interest or nucleic-acid-of-interest is associated with a labelled tRNA. In a second aspect, there provided a split tRNA comprising a first chain and a second chain, wherein the first chain comprises a first tRNA portion and comprises a first stem region, and the second chain comprises a second tRNA portion and comprises a second stem region. In an embodiment, there is provided a split tRNA comprising a first chain and a second chain, wherein: the first chain comprises a first tRNA portion and comprises a first stem region, and the second chain comprises a second tRNA portion and comprises a second stem region; the first tRNA portion corresponds to the 5’ portion of a parental tRNA split at the anticodon and the second tRNA portion corresponds to the 3’ portion of the parental tRNA split at the anticodon; the first stem region is located at the 3’ end of the first tRNA portion and the second stem region is located at the 5’ end of the second tRNA portion; and the first stem region and the second stem region are complementary. In a third aspect, there is provided an RNA molecule comprising a sequence encoding a polypeptide-of-interest or nucleic-acid-of-interest and comprising a portion of a tRNA. In a fourth aspect, there is provided a method of determining the acylation status of a tRNA or efficiency of acylation of a tRNA, the method comprising: i) incubating the tRNA and a substrate with which the tRNA could be acylated, under conditions conducive to acylation of the tRNA; ii) exposing the tRNA to conditions capable of blocking the 3’ end of free tRNAs; and iii) exposing the tRNA to conditions that lead to the addition of nucleotides to the 3’ end of tRNAs that are not blocked, wherein at least one nucleotide comprises a label. In a fifth aspect, there is provided a tRNA comprising additional nucleotides at the 3’ end, wherein at least one nucleotide comprises a label. In a sixth aspect, there is provided a nucleic acid encoding any split tRNA of the second aspect or any RNA molecule of the third aspect. In an embodiment, the nucleic acid comprises, from 5’ to 3’: a second stem-region-encoding sequence, a second tRNA-portion-encoding sequence, a sequence encoding a loop, a first tRNA-portion-encoding sequence, and a first stem-region-encoding sequence, wherein the first stem-region-encoding sequence and second stem-region-encoding sequence encode complementary stem region sequences. In a seventh aspect, there is provided a method of making a polypeptide, wherein the method comprises: i) providing a sequence of a polypeptide-of-interest identified by any suitable method disclosed herein, and ii) producing a polypeptide according to said sequence. In an embodiment, there is provided a method of making a polypeptide or nucleic acid, wherein the method comprises: i) providing a library comprising a plurality of sequences encoding polypeptides-of-interest or nucleic- acids-of-interest, wherein each sequence within the library is linked to a portion of a tRNA; ii) incubating, under conditions conducive to acylation, each polypeptide-of-interest or nucleic-acid-of-interest with a tRNA comprising the tRNA portion linked to the sequence encoding the respective polypeptide-of-interest or nucleic-acid-of-interest, and wherein the incubation includes a substrate with which the tRNAs could be acylated; iii) exposing the tRNAs to conditions capable of labelling tRNAs that have been acylated; and iv) identifying whether each polypeptide-of- interest or nucleic-acid-of-interest is associated with a labelled tRNA; and v) making a polypeptide according to the sequence of an identified polypeptide-of-interest or nucleic-acid-of-interest. Also provided is use of an acyl-tRNA synthetase as disclosed herein in a method of genetically incorporating a monomer into a polymer. Also provided is a method of making a polymer, wherein the method comprises: i) use of an acyl-tRNA synthetase as disclosed herein to acylate a tRNA with a monomer, and ii) incorporation of the monomer into a polymer chain. BRIEF DESCRIPTION OF THE DRAWINGS Fig.1: Encoded cellular incorporation of non-canonical monomers into proteins and into non-canonical polymers requires both tRNA acylation and ribosomal polymerization. The encoded, site specific, incorporation of a non-canonical monomer (ncM, yellow star) via cellular translation requires both the acylation of an orthogonal tRNA with the ncM by an orthogonal synthetase, and ribosomal polymerization of the ncM into a polymer chain. Current methods for engineering aminoacyl-tRNA synthetases that acylate new monomers rely on translational readouts and therefore require the monomers to be ribosomal substrates. For ncMs that are poor ribosomal substrates this co-dependence creates an evolutionary deadlock in cells; an orthogonal synthetase cannot be evolved to acylate an orthogonal tRNA with ncMs that are poor ribosomal substrates, and ribosomes cannot be evolved to polymerize ncMs that cannot be acylated onto orthogonal tRNAs. To break this deadlock, we develop direct selections for orthogonal synthetases to aminoacylate their cognate orthogonal tRNAs with ncMs, independent of whether the ncMs are ribosomal substrates. Fig.2: Acylation-dependent incorporation of modified nucleotides in tRNA extension enables the sensitive detection, and selective isolation of acylated tRNAs. a, Fluorescent tRNA extension (fluoro-tREX) and biotin- tRNA extension (bio-tREX). tRNAs are isolated from cells using a mild phenol lysis and oxidised by sodium periodate. The diol functionality of the 3’ ribose on non-acylated tRNAs is oxidized to the dialdehyde. The acyl group of charged tRNAs protects the diol functionality of the 3’ ribose and prevents the oxidation to a dialdehyde. A Cy3 labelled DNA probe complementary to the 3’ end of a target tRNA is annealed, and Klenow exo (-) and modified nucleotides are added. This results in the selective extension and labelling of tRNAs that were acylated. For fluoro-tREX, Cy5 labelled nucleotides are incorporated, and, following gel electrophoresis, Cy3 and Cy5 fluorescence is visualized. Acylated tRNAs lead to a Cy5 and Cy3 signal, whereas non-acylated tRNAs only give a Cy3 signal. For bio-tREX, biotinylated nucleotides are incorporated and the resulting mixture is bound to streptavidin beads. The beads are washed, and then biotinylated tRNAs are released by heating in formamide; selective isolation of tRNAs that were acylated may be visualized by SYBR gold staining following gel electrophoresis. b, Fluoro-tREX detected the acylation of tRNAPylCUAin the presence of its cognate aminoacyl- RNA synthetases (PylRS) and its ncAA substrate, BocK (1). Cells harbouring tRNAPylCUA were grown in the presence (+) and absence (-) of PylRS and BocK (1). The Cy3 signal resulted from specific annealing of the Cy3 labelled probe to tRNAPylCUA and detected the presence of this tRNA. The Cy5 signal resulted from the extension of the previously acylated tRNA with Cy-5-dCTP, and was dependent on PylRS and BocK (1). c, Bio-tREX permits the selective isolation of previously acylated tRNAs. Cells harbouring tRNAPylCUAwere grown in the presence (+) and absence (-) of PylRS and BocK (1). Isolation of the tRNA and associated probe was visualized by SYBR gold staining, for RNA, and Cy3 fluorescence, for the probe. Fig.3: Transcription, assembly, maturation and acylation of split tRNAs expressed from split or circularly permuted genes. a, Schematic for producing split tRNAs in trans from two genes. The tRNA gene is split at the anticodon loop and the anticodon stem sequence is extended for optimal assembly of the transcribed RNA in vivo; this creates two genes: one for the 5’ half and one for the 3’ half of the split tRNA. The gene for each tRNA half is transcribed and the split tRNA is assembled, matured, and acylated in cells. b, Schematic for producing split tRNAs in cis from a single gene. The tRNA sequence is circularly permutated by connecting the 3’ half with a ‘loop’ sequence to the 5’ half, splitting the sequence at the anticodon, and extending the anticodon stem. Transcription, assembly in cis, and maturation leads to a functional split tRNA. c, The in vivo transcription, assembly, maturation and acylation of split tRNAPylproduced from genes for the 5’ half and 3’ half. Cells were grown in the presence of PylRS, the presence or absence of BocK (1), and the genes for either one or both tRNA halves. Only the expression of both tRNAPylhalves led to a BocK (1) dependent acylation signal, as judged by fluoro-tREX. Note that the purification conditions used to isolate these stRNAs were different from the conditions used to isolate intact tRNAs or in cis transcribed stRNAs. Under these conditions we do not observe the Cy3 probe. d, The choice of loop region is pivotal for the efficient expression, maturation and acylation of circularly permutated split tRNAs in cells. Six circularly permutated split tRNAPylwith different loop sequences were assayed by fluoro-tREX. For the argY-argZ and leuP-leuV loops (derived from the intergenic regions of pairs of tRNA genes in E. coli) the fluoro-tREX signal for split tRNA production (Cy3) and acylation (Cy5) was comparable to the corresponding signal for intact tRNAPyl(Fig.11). Fig.4: stmRNAs couple acylation phenotype to aminoacyl-tRNA synthetase genotype, enabling efficient and selective isolation of PylRS variants with a wide range of activities. a, Schematic representation of the cis split tRNA-mRNA fusion (stmRNA) gene. Within the cell, this gene is transcribed and processed to create an stmRNA in which the 3’ half of the split tRNA is covalently linked to the mRNA of PylRS. The mRNA of PylRS is translated to generate the PylRS enzyme which, in the presence of its substrate (yellow star), acylates the stmRNA. This creates a physical link between the acylation phenotype and the genotype of the PylRS mRNA. b, BocK (1) dependent acylation of the stmRNA visualized by fluorescent mRNA extension (Fluoro-mREX). Cells harbouring the stmRNA gene encoding wild type (wt) PylRS (stmRNAwt), and cells harbouring the stmRNA gene encoding an attenuated PylRS (stmRNAat), were grow in presence and absence of the PylRS substrate BocK (1). Fluoro-mREX led to a BocK (1) dependent Cy5 fluorescent band of the expected length for stmRNAwtbut not for stmRNAat. These results demonstrated that stmRNA genes were functionally expressed in cells, converted into mature stmRNAs, and led to the production of active PylRS enzymes, which acylated the stmRNA. The 16S (1.5 kb) and 23S rRNA (2.9 kb) were used as loading controls and size markers. The fusion between the 3’ half of the tRNA and the mRNA is 1.5 kb. The Fluoro-mREX signal was visualized on denaturing gels under conditions where the non-covalent binding of the Cy3 labelled probe is disrupted, as a result only Cy5 fluorescence was visualized in Fluoro-mREX. c, Schematic of biotin mRNA extension (bio-mREX). Bio-mREX is based on the same principles as bio-tREX. Biotinylated stmRNAs – resulting from extension, of non-oxidized, previously acylated stmRNAs with biotinylated dNTPs – are bound to streptavidin beads and stringently washed. The mRNA of PylRS within is then reverse transcribed on the beads and quantified by qPCR. Formerly acylated stmRNAs should lead to high molecular counts for the cDNA, whereas non-acylated stmRNAs are expected to only result in background signal. d, Efficient and selective isolation of the cDNA of active PylRS variants via bio-mREX. Cells harbouring the stmRNAwtconstruct or stmRNAatwere grown in presence and absence of BocK (1) and bio-mREX was performed. The number of cDNA molecules from the wt stmRNA constructs in presence of BocK (1) was a hundred-fold higher than in the absence of BocK (1) and a 300-fold higher than for stmRNAatin the presence of Bock (1). There was a minimal difference in number of cDNA molecules from stmRNAatwith and without BocK (1). Additionally, a sample of the isolated RNA prior to each pulldown was reverse transcribed to measure the total number of stmRNA molecules used as input for bio-mREX (Fig.12b). In presence of BocK (1), 2.5 % of the input stmRNA molecules were recovered. The dashed line represents 2.5% of input (calculated from the average of all input samples). e, Tailoring the 5’UTR of PylRS mRNAs within stmRNAs leads to a stronger correlation between the acylation of stmRNAs by the PylRS enzyme variants encoded within them and the read through of an amber stop codon by the same PylRS enzyme variants when paired with tRNAPylCUA (original RBS: R-squared value= 0.4694, p value= 0.3148; RBS_2: R-squared value= 0.9742, p value=0.013). Cells harbouring stmRNA genes encoding four distinct PylRS variants (PylRS(CbzK1-4)) with either the original 5’UTR (Orig. RBS) or a designed 5’ UTR (RBS2 – stmRNAvol2) were grown in presence of CbzK (2) and bio-mREX was performed. The measured cDNA molecules were plotted against the fluorescence intensity of GFP(150CbzK)His6, resulting from read-through of the amber codon in GFP(150TAG) Pyl His6 by each PylRS variant paired with tRNA CUA in cells provided with CbzK (2). Fig.5. tRNA display enables the direct selection of orthogonal aminoacyl-tRNA synthetases that aminoacylate their cognate orthogonal tRNAs with ncAAs. a, Schematic representation of tRNA display, a translation independent strategy for selecting aaRS enzymes that aminoacylate a specific tRNA with a desired monomer. In tRNA display a library of stmRNAs is transformed into cells and grown in presence and absence of non-canonical monomers of interest (represented by the yellow star) in multiple replicates. The library contains PylRS variants that are active and selective for the ncM (yellow), PylRS variants that are neither active with the ncMs or canonical amino acids in the cell (light blue) and PylRS variants that are active with one or more canonical amino acid in the cell and are not selective for the ncM (dark blue). Bio-mREX is performed for each replicate and the cDNA is submitted for next generation sequencing (NGS). The results are analysed by plotting the selectivity (the ratio of the relative abundance of a particular sequence in the positive samples (+ncM), divided by the relative abundance in the negative sample (-ncM)), against the enrichment (the ratio of the same sequence in the positive samples, divided by the relative abundance in the input library) of all observed sequences; this results in a spindle-shaped plot (henceforth referred to as spindle plot). Active and selective PylRS variants are expected to be highly selective and enriched (yellow dot – upper right quadrant), active but nonselective variants are expected to be non-selective but highly enriched (dark blue dot), and inactive PylRS variants are expected to be non-selective and non-enriched (light blue dot). b, The structures of non-canonical α-alpha-amino acids used in this study. N6-((benzyloxy)carbonyl)-L-lysine (CbzK) (2), N6-((prop-2-yn-1-yloxy)carbonyl)-L-lysine (AlkyneK) (3), N6-benzoyl-L-lysine (BenzK) (4), 3-([2,2'-bipyridin]-5-yl)-2-aminopropanoic acid (BiPyA) (5), Nτ- methyl-L-histidine (NτmH) (6), (S)-2-amino-3-(thiophen-3-yl)propanoic acid (3-ThiA) (7) ), (S)-2-amino-3- (pyridin-3-yl)propanoic acid (PyA) (8), (S)-2-amino-3-(4-iodophenyl)propanoic acid (pIF) (9), (S)-2-amino-3-(4- bromothiophen-2-yl)propanoic acid (BrThiA) (10), (2S)-2-amino-3-(((2-((1-(6-nitrobenzo[d][1,3]dioxol- 5yl)ethyl)thio)ethoxy)carbonyl)amino)propanoic acid (pcDAP) (11). c, tRNA display with PylRS library (stmRNAvol2-lib1), this library mutates three positions to all other canonical amino acids (Y306X, L309X, and N346X). The spindle plot shows the result of the tRNA display selection against CbzK (2) using one step of parallel selection as outlined in (a). Samples were run in triplicate and data processed as described in the Methods. Red dots indicate 65 clones that were further characterized. d, ln(Enrichment, +2) of PylRS mutants derived from tRNA display (red dots in (c)) vs. the GFP fluorescence measured in cells containing the corresponding PylRS mutant / tRNAPylCUA pair, GFP(150TAG)His6 and CbzK (2). The dotted line represents the linear regression for the displayed data points; R-squared = 0.6611, p < 0.0001. e-j, (left) white bar: GFP fluorescence from cells containing GFP(150TAG) , the in Pyl His6 dicated PylRS variant / tRNA CUA pair, and the indicated ncAA. Grey bar: the wtPylRS / tRNAPylCUA pair with the same ncAA. Fluorescence is shown as a fraction of the fluorescence generated by the wt PylRS / tRNAPylCUApair with 2 mM BocK (1) and GFP(150TAG)His6. (right) ESI-MS of GFP(150X)His6, where X is the indicated ncAA. f, found mass: 27922.0 Da, expected mass 27923.3 Da; g, found mass: 27944.8 Da, expected mass 27945.5 Da; h, found mass: 27867.6 Da, expected mass 27866.4 Da; i, found mass: 27862.0 Da, expected mass 27861.4 Da; j, found mass: 27986.4 Da, expected mass 27986.2 Da; k, found mass: 27945.6 Da, expected mass 27944.3 Da. Fig.6. tRNA display enables selection of orthogonal aminoacyl-tRNA synthetases that charge non- canonical monomers. a, Structures of non-canonical monomers used in this study. (S)-3-amino-3-(3- bromophenyl)propanoic acid ((S)β3mBrF) (12), (S)-3-amino-6-(((benzyloxy)carbonyl)amino)hexanoic acid ((S)- β3CbzK) (13), (S)-6-acetamido-3-aminohexanoic acid ((S)-β3AcK) (14), BocAhx (15), 6- (((benzyloxy)carbonyl)amino)hexanoic acid (CbzAhx) (16), 3-amino-2-((1-ethyl-1H-imidazol-5- yl)methyl)propanoic acid (β2NeH (17) ), 3-amino-4-(4-bromophenyl)butanoic acid (β3pBrhF) (18), 2-benzyl-3- hydroxypropanoic acid (β2OH-F) (19), 3-amino-3-phenylpropanoic acid (β3F) (20). b, Fluoro-tREX for the indicated PylRS variants. Experiments were performed in triplicate, using tRNAs extracted from cells harboring a pMB1 plasmid encoding each PylRS variant and TRNAPylCUAin presence and absence of 4 mM 15. c-d, Fluoro- tREX for the indicated PylRS variants from the primary selection (panel c) and after further evolution (panel d). Experiments were performed in triplicate, using tRNA extracted from cells harboring a pMB1 plasmid encoding each PylRS variant and tRNAPylCUAin presence and absence of 4 mM 12. e, Selected PylRS variants acylate tRNAPylCUA with 12. LC-MS traces (scanning ion mode on AQC adduct of substrate 12) on AQC-derivatized eluates from tRNA-pull downs from cells expressing the indicated PylRS variants. Cells harbouring a pMB1 plasmid encoding the corresponding PylRS variant and tRNAPylCUA(or only tRNA (-) as a control were grown in the presence of substrate 12, and tRNA pulldowns were performed using a biotinylated probe against TRNAPylCUA. f, Quantification of relative acylation of tRNAPylCUAby selected PylRS variants. The integrated area under the peak for of LC-MS traces shown in e. g, GFP fluorescence from cells containing GFP(150TAG)His6, the PylRS(12_1) or PylRS(12_1evol1), and tRNAPylCUA, and grown in the presence or absence of substrate 12. Fluorescence is shown as a fraction of the fluorescence generated by the wt PylRS / tRNAPylCUA pair with 2 mM BocK (1) and GFP(150TAG)His6. h, intact ESI-MS of GFP(150(S)β3mBrF)His6purified from cells harbouring PylRS(12_1evol1), tRNAPylCUA, and GFP(150TAG)His6grown in the presence of 4 mM 12. Found mass: 27939.0 Da, predicted mass: 27,938.2 Da. i, Close up on residue 150 of GFP(150(S)β3mBrF)His6 from a crystal structure determined at 1.5 Å. The 2Fo-Fc map is shown at contour level of sigma = 2 (PDB code 8OVY). The electron density (blue) confirms the incorporation of 12 at position 150, clearly demonstrating the extension of the peptide backbone by one methylene group and the stereochemistry of the β-amino acid in the protein. Fig.7. The DNA probe used throughout this work is specific for MmtRNAPylCUAand, when fluorescently labeled, provides a loading control for the relative concentration of MmtRNAPylCUA in fluro-tREX experiments. a, Northern blot showing the relative amount of MmtRNAPylCUAused in the fluoro-tREX experiments shown in panel b. We isolated tRNAs from DH10β cells harboring a pMB1 plasmid encoding MmtRNAPylCUAor no cells harboring no plasmid. tRNA samples isolated from cells containing MmtRNAPylCUA were used undiluted and diluted into tRNAs isolated from cells without MmtRNAPylCUAat ratios of 1:4, 1:16, 1:64 and northern blots were run on the tRNA samples, using the previously verified (ref Cervettini et al.) MmtRNAPylCUAspecific probe 5’(Btn)- TGGCGGAAACCCCGGGAATCTAACCCGGCT-3’ (SEQ ID NO: 103). The data shows a dilution series of MmtRNAPylCUAand confirms that the northern blot signal is dependent on the presence of MmtRNAPylCUAin the tRNA sample. We carried out the experiments in biological duplicates, with similar results. b, The Cy3 signal in fluro-tREX decreases with dilutions of MmtRNAPylCUA. Fluoro-tREX was run on the tRNA samples characterized in a. In brief, the fluorescently labeled DNA probe 5’GGGCCCATTAACATCACCTGGCGGAAACCCCGGGAATCTAACCCGGCT-3’Cy3 (SEQ ID NO: 104) was annealed to MmtRNAPylCUAand the probe was extended by Klenow (exo-) in the presence of Cy5-dCTPs. We ran a urea PAGE gel and visualized Cy3 as well as Cy5 fluorescence. c, correlation of Cy3 signal and fluorescence signal determined by northern blot. Band intensities were determined by densitometry using ImageJ. The two signals showed a strong and significant correlation (R2value=0.9935, p value=0.0032). d, correlation between the observed Cy3 signal and the theoretical quantity of MmtRNAPylCUA; this is the relative amount of MmtRNAPylCUA predicted to be in each sample on the basis of the dilution series (1, 1:4, 1:16, 1:64) . Cy3 fluorescence robustly reflects the amount of loaded tRNA (R2value=0.9657, p value=0.0173). Experiments were carried out in two biological replicates producing similar results. Fig.8. Characterizing the acylation activity of PylRS variants which support amber suppression activity over two orders of magnitude, by fluoro-tREX. a, Production of GFP150CbzKHis6 from GFP150TAGHis6 from cells harboring a pMB1 plasmid encoding one of four different N6-((benzyloxy)carbonyl)-L-lysine (CbzK) 2 variants of PylRS (RS1: Y306G, L305G; RS2: Y306G, N346G; RS3: Y306S; RS4: Y306G) and MmtRNAPylCUAand a p15A plasmid encoding GFP150TAGHis6in the presence and absence of 2 mM CbzK 2. PylRS variants 1-4 lead to amber suppression activity over two orders of magnitude, when measured by fluorescence of GFP150CbzKHis6. Dots represent the mean of three biological replicates, error bars show ± s.d. b, The signal of fluoro-tREX can resolve the aminoacylation activity of all CbzK PylRS variants characterized in panel a. We isolated total tRNA from DH10β cells harboring a pMB1 plasmid encoding one of the four PylRS variants (RS1, RS2, RS3, or RS4) and MmtRNAPylCUAand performed fluoro-tREX. c, Cy5 signal vs GFP fluorescence plot for CbzK-RS 2, 3, and 4. The Cy5 fluorescent signal for these three variants corresponded well to the observed levels of GFP expression, permitting the identification and distinction of low as well as high activity PylRS variants via fluoro-tREX. Experiments were carried out in two biological replicates producing similar results. Fig.9. Following oxidation, the deacylation of tRNAs under alkaline conditions increases the acylation signal in fluro-tREX and thereby permits the robust detection of acylation by hydroxy acids, and carboxylic acids. A, Chemical structure of N6-(tert-butoxycarbonyl)-L-lysine (BocK) 1, (S)-6-((tert-butoxycarbonyl)amino)-2- hydroxyhexanoic acid (OH-BocK) 21 as well as 6-((tert-butoxycarbonyl)amino)hexanoic acid (BocAhx) 15. MmPylRS is highly active with BocK 1, OH-BocK121 and, and shows acylation activity with BocAhx 152. B, The free acid of 1, 21 and 15 span a range of the estimated pKas. The rate constant for the alkaline hydrolysis of esters, to give a fixed alcohol and a variable carboxylic acid, increases as the pKaof the resulting carboxylic acid decreases. We therefore expect the rate of hydrolysis for acylated tRNAs to be slower when the acylating monomers are α-hydroxy acids, simple carboxylic acids (and β-amino acids), than when the acylating monomers are α-amino acids c, MmPylRS acylates MmtRNAPylin cells with BocK 1, OH-BocK 21, and BocAhx 15 respectively. Northern blot of MmtRNAPylfrom tRNAs isolated from cells harboring a pMB1 plasmid encoding the MmPylRS / MmtRNAPylpair in presence and absence of BocK, OH-BocK, or BocAhx. The experiments were carried out in three biological replicates producing similar results. D, After oxidation, a 7iacylation step under alkaline conditions is necessary to robustly detect acylation activity of MmPylRS by fluoro-tREX with non-alpha amino acid substrates. Fluoro-tREX was performed with the tRNA samples described in panel c with and without an incubation of the tRNAs for 45 minutes with 50 mM bicine at pH 9.6 post oxidation. The acylation signals from OH-BocK as well as BocAhx were dependent on the 7iacylation of the tRNAs before the Klenow (exo-) extension step of the protocol. The experiments were carried out in three biological replicates producing similar results. Fig.10. Split MmtRNAPyls require a base-pairing stem region of ten bases to be efficiently acylated by MmPylRS in cells when both tRNA halves are expressed in trans. A-c, Acylation of split MmtRNAPylin cells was dependent on the presence of both tRNA halves and a base-pairing region of ten base-pairs. tRNAs were isolated from DH10β cells harboring split tRNA constructs with base-pairing stems of eight, ten or fourteen base-pairs where the 3’MmtRNAPylhalf was encoded on a pMB1 plasmid, and the 5’ MmtRNAPylhalf on a p15A plasmid and the cells were grown in presence and absence of BocK. For the split tRNA construct with a ten base-pair long stem, cells were also grown with pMB1, or p15A plasmids lacking either the 3’, or 5’ tRNA half respectively. Fluoro- tREX was performed with isolated tRNAs. For stems of eight base-pairs length, a weak acylation signal was observed by fluoro-tREX, and for stems which were fourteen base-pair long, cleavage products were predominantly observed by fluoro-tREX. The experiments were carried out in three biological replicates producing similar results. D, The same as for a-c but with a split tRNA construct with a twelve base-pair long stem region, leading to a weaker acylation signal when compared to a ten base-pair long stem and to the observations of multiple bands, which are likely to result from stem cleavage. For split tRNAs produced in trans we purified and concentrated the extension reaction before loading, following the general procedure for fluro- tREX B. Under these conditions we do not observe the Cy3 signal for the probe associated with the extension product. Fig.11. The sequence of the loop region of circularly permutated split tRNAs is crucial for the robust expression and acylation of split tRNAs in cells. Split tRNAs were isolated from DH10β cell harboring a pMB1 plasmid encoding one of seven circularly permutated MmtRNAPylconstructs, each with a different loop region (Cm. gln, Cm. gly, E coli (Ec.) argY-argZ, Ec. alaW-alaX, Ec. leuP-leuV, Ec. glnW-metL) in presence and absence of 4 mM BocK 1. Intact MmtRNAPylCUA was produced as a control. Split tRNAs were isolated and fluoro-tREX was performed. Split MmtRNAPylconstructs with loops constituting of the E. coli intergenic regions argY-argZ, or leuP-leuV led to high acylation and split tRNA expression levels as judged by Cy5, or Cy3 fluorescence respectively, which were comparable to the signals observed for intact MmtRNAPylCUA. Mean Cy3 signal for MmtRNAPyl10185 ± 1102; for argZ-argY 12063 ± 630; for leuP-leuV 13467 ± 465. Mean Cy5 signal for MmtRNAPyl1656 ± 62; for argZ-argY 1397 ± 66; for leuP-leuV 1442 ± 67 Cy3 and Cy5 signals were determined by densionmetry. The experiments were carried out in three biological replicates producing similar results. Fig.12. A Protein production, via read through of an amber stop codon, by wild type (wt) and attenuated (at) PylRS / tRNAPylCUA pairs. Production of GFPAllocKHis6 from GFP150TAGHis6 from cells harboring a pMB1 plasmid encoding either wild type (wt) MmPylRS or an attenuated (at) mutant (H338A, F342A, M344A, E396A, S399A) and MmtRNAPylCUA and a p15A plasmid encoding GFP150TAGHis6 in the presence and absence of 2 mM AllocK. Dots represent the mean of three biological replicates, error bars show ± s.d. B. Input samples for bio- mREX of wt and attenuated stmRNA. A fraction of purified stmRNA was reverse transcribed, and then quantified via qPCR in the same manner as the samples subjected to extension with biotinylated nucleotides and pull down. Addition of BocK did not lead to a significant difference in isolated stmRNA levels (attenuated PylRS: p value= 0.1648, wt PylRS: p value= 0.6390). Fig.13. Relationship between the acylation signal measured, by bio-mREX, for stmRNAs and the GFP fluorescence signal measured for intact, translation-competent tRNAs. For stmRNAs, active aminoacyl-tRNA synthetases (aaRS) lead to the acylation of their encoding stmRNAs, which by bio-mREX get extended, separated and ultimately reverse transcribed. This results in the cDNA of the active synthetase, which can be quantified by qPCR. In the case of an inactive aaRS the stmRNAs is not acylated and no cDNA produced in bio-mREX experiments. Therefore, the activity of a synthetase in bio-mREX correlates with the number of cDNA molecules measured by qPCR. In canonical translation an active aaRS enzyme leads to an acylated, intact, cognate tRNACUAwhich is used in protein translation. Inactive aaRS enzymes lead to non-acylated tRNAs, which are not used in protein translation. The production of GFP protein from GFP150TAGHi6, as measured by GFP fluorescence, reports on the acylation of tRNACUA, as well as the other steps in the production of protein. Fig.14. Resolving the activity of PylRS variants in bio-mREX via 5’UTR tuning. a, Nucleotide sequences of 5’UTR sequences RBS1-3 with translation initiation rates given. Sequences were generated and initiation rates predicted by DeNovo DNA. b, Cells harboring a plasmid encoding stmRNAs bearing a PylRS CbzK mutant (see Fig.8) under the control of either the initial 5’UTR, or one of three designed 5’UTR sequences were grown in presence of 2 mM CbzK 2, stmRNA transcription was induced for 20 minutes, total RNA isolated by phenol chloroform extraction, and bio-mREX was performed. For the stmRNA construct with the initial 5’UTR region (stmRNAvol1) low activity PylRS mutants (e.g. CbzK-RS2) led to a saturation of the acylation signal. For all designed 5’UTR sequences, which were designed to lead to low translation levels, low activity PylRS variants could be resolved by bio-mREX. The stmRNA under the control of RBS2 (stmRNAvol2) led to a good correlation of PylRS activity as measured by amber suppression of GFP150TAGHis6and the acylation data measured by bio- mREX (see Fig.4e) and was used for all future experiments. Dots represent the mean of three biological replicates, error bars show ± s.d. Fig.15. PylRS libraries used in this work. a, Overview of the seven libraries designed and created. These libraries target a total of 11 amino acid residues in the PylRS active site and employed several types of degenerate codons. NNK codons are depcited as dark red, DBK codons (+ lysine codon) as blue, NDT codons as dark green, NRT codons as yellow. For certain sites, custom residue mixes encompassing the most commonly observed mutations were used (1-7 mixes, depcited as grey spheres). All libraries were created with at least 109independent transformants. N = A, T, G, C; K = G, T; D = G, A, T; B = G, T, C; R = G, A. The custom mixes are described in the methods. b, The eleven amino acid residues targeted for mutageneesis in the PylRS active site are shown in red. Image was rendered using Pymol, based on the PDB structure 2ZIN. Fig.16. tRNA display identifies active and selective orthogonal aaRS variants from an stmRNA library. a, Spindle plot from the tRNA display selection using stmRNA library 1 and ncAA 1. b, Identifying the region of the spindle plot enriched in active and selective clones. We expect the top right quadrant of the spindle plot to be enriched in active and selective clones. Since selectivity is derived from the ratio of sequence counts + ncAA and -ncAA, enriched clones with negative selectivity values would correspond to specific enrichment of a clone in the -ncAA condition with respect to the +ncAA condition. We postulated that most apparent enrichments of this type were spurious, and therefore that regions of the spindle plot where positive selectivity values were mirrored by negative selectivity values of the same magnitude may contain substantial noise. Based on this postulate we expected the active and ncAA selective clones to be most enriched in the region of the plot where, for a given positive enrichment value, the selectivity becomes asymmetric. To enrich for this asymmetric population we binned mean selectivity values: 5350 points in the spindle plot (Fig.5 c) were divided into 500 equal bins along the enrichment+1dimension (163 bins contained data). The mean selectivity of each bin was plotted against the natural logarithm of enrichment+1. From this, a threshold of ca 7.4 (corresponding to a logarithmic score value of 2) was used to define the enrichment value at which the spindle plot is asymmetric along the selectivity axis. c, Experimental GFP fluorescence values for 100 clones plotted against the natural logarithm of enrichment score in the presence of 1. GFP was expressed from GFP150TAGHis6 in the presence of the MmPylRS variant clone, the cognate MmtRNAPylCUAand the ncAA (1). Points above the symmetry threshold are colored in red (65 points), points below are colored in blue (21 points). There is a strong positive and significant correlation between the tRNA display sequence data and experimental expression data for the red points (R-square value=0.6611, p<0.0001 value), but no significant correlation for the blue points (R2value=0.0392, p value=0.397), this is consistent with our postulate. The red points are also shown in Fig.5c. In subsequent selections, on the basis of this analysis, we primarily focused on identifying clones on the right-hand side of the spindle plot where, for a given enrichment value, the magnitude of the positive selectivity value for a clone is of greater magnitude than negative selectivity values for clones with the same enrichment value. d, tRNA display identifies ncAA specific PylRS variants. Plot shows the experimental selectivity vs selectivity from the spindle plot for the clones show in panel b, color coding as in panel b. The experimental selectivity is derived from GFP expression experiments, as in panel b, but + / - ncAA. Fig.17. Schematic representation of tRNA display based strategy for selecting PylRS variants that direct the incorporation of ncAAs into proteins. In the first round naïve stmRNAvol2libraries 2, 13, 14, 3D, 4D, and 5D were transformed into BL21 cells and grown overnight. Libraries 3D and 4D were combined to generate library 3D4D. The five libraries were grown to OD600 of 0.3-0.4 and 2.6 mL of the cell culture from each library was added to a stock solution of each ncAA; this resulted in 50 samples (five libraries x ten ncAAs). Cells were grown for 40 min, stmRNAs induced, and cells grown for another 20 min. Bio-mREX was performed on the isolated RNA for each of the 50 samples. For each reaction, cDNA was amplified with primers suitable for Golden Gate assembly. Then all amplicons of the libraries selected for the same ncAA were combined at equimolar ratios (resulting in ten combined libraries in total) and cloned into a fresh ColE1vector backbone. This created ten pre-selected libraries. The ten pre-selected libraries were transformed into BL21 cells and grown over night. The preselected libraries for ncAAs 2 - 6 were combined to create a single cluster library. Similarly, the preselected libraries for ncAAs 7 – 11 were combined to create a second cluster library.2.6 mL of the first cluster library was added to solutions of ncAAs 2 – 6.2.6 mL of the first cluster library was also added to a sample without ncAA, as a control. Cells were grown for 40 min, stmRNAs induced and cells grown for another 20 min, and the RNA isolated. Three RNA samples were converted to cDNA as bio-mREX input control. Bio-mREX was performed on the isolated RNA. This generated seven samples (bio-mREX input, -ncAA control for bio-mREX, and five bio-mREX samples for ncAAs 2 – 6. The experiment was performed in triplicates, generating 21 samples. The cDNA of each sample was sequenced by NGS and analyzed to generate spindle plots and sequence tables. The second cluster was treated analogously to the first cluster, unsing ncAAs 7 – 11 in place of 2 – 6. Fig.18. Schematic representation of selection strategy for non-canonical monomers. Library 14 was transformed into BL21 cells and grown overnight. Cells were grown to OD600of 0.3-0.4.4 mL of the library culture was added into stock solutions of each ncM.4 mL of the library culture was also added to a well without ncM. The cells were grown for 40 min, stmRNAs induced, and cells grown for another 20 min and the RNA was isolated. Bio-mREX was performed on the isolated RNA for each sample.The experiment was performed in 4 replicates, leading to 40 cDNA samples. An additional 6 cDNA samples were generated for 6 of the RNA inputs to bio-mREX. The resulting 46 cDNA samples were sequenced by NGS and analyzed to generate spindle plots and sequence tables. Fig.19. Schematic of tRNA pulldown followed by LC-MS analysis to determine the identity of the monomer on the target tRNA. tRNAs are extracted from cells expressing the tRNA of interest and the cognate orthogonal aaRS, grown in the presence of the ncM. A biotinylated probe is annealed, and the targeted tRNA is pulled down. After washing, the ncM is eluted by alkaline deacylation, derivatised with AQC, and detected using LC-MS. Fig.20. Schematic representation of selection strategy for random mutagenesis selections by tRNA display. We performed an error prone PCR reaction across the active site sequence of PylRS variants 12_1 and 12_2 using the GeneMorph II (Agilent) kit. The diversified PCR amplicons were cloned into a fresh ColE1 plasmid backbone by Golden Gate assembly and the error prone stmRNA libraries transformed into BL21 cells and grown overnight. Cells were grown to OD600of 0.3-0.4.2.6 mL of the library culture was added into a stock solution of 12 to a concentration of 4 mM.2.6 mL of the library culture was also added to a well without ncM. The cells were grown for 40 min, stmRNAs induced, and cells grown for another 20 min and the RNA was isolated. Bio-mREX was performed on the isolated RNA for each sample. The experiment was performed in four replicates, leading to eight cDNA samples. An additional four cDNA samples were generated for four of the RNA inputs to Bio-mREX. The resulting twelve cDNA samples were sequenced by NGS and analyzed to generate spindle plots and sequence tables. Fig.21. Schematic representation of two step selection strategy for non-canonical monomers. Library 14 was transformed into BL21 cells and grown overnight. Cells were grown to OD600of 0.3-0.4.4 mL of the library culture was added into stock solutions of each ncM. The cells were grown for 40 min, stmRNAs induced, and cells grown for another 20 min and the RNA was isolated. Bio-mREX was performed on the isolated RNA for each sample. The experiment was performed in four replicates. For each replicate the cDNA was amplified with primers suitable for Golden Gate assembly. Then all amplicons of the libraries selected for the same ncM were combined at equimolar ratios and cloned into a fresh ColE1vector backbone. This created one preselected library of each ncM. The pre-selected libraries were transformed into BL21 cells and grown over night. Cells were grown to OD600of 0.3-0.4. For each ncM, 4 mL of the respective preselected library culture was added into stock solutions of the ncM. For each preselected library 4mL of the library culture was also added to a well without ncM. The cells were grown for 40 min, stmRNAs induced, and cells grown for another 20 min and the RNA was isolated. Bio-mREX was performed on the isolated RNA for each sample. The experiment was performed in three replicates, leading to 6 cDNA samples per ncM. An additional three cDNA samples were generated for each ncM using three RNA inputs to bio-mREX of the respective preselected libraries. For each ncM the resulting nine cDNA samples were sequenced by NGS and analyzed to generate spindle plots and sequence tables. Fig.22. Screening of intergenic regions (IGRs) for 1R26, deltaClos, and Nitra PylRS stmRNAs. a) 1R26 stmRNA IGR screening. b) Validation of best IGRs from a). c) deltaClos stmRNA IGR screening. d) Replicate for best IGRs from c) and additional assessment of the two best-performing IGRs for the MmPylRS stmRNA. e) Nitra stmRNA IGR screening. X axis labels (other than the Mm control) are the names of the natural or synthetic IGRs. Cultures inoculated 1:20 in presence or absence of 4 mM ncM (Mm and deltaClos: AllocK, 1R26: CbzK, Nitra: NMH). stmRNA expression was induced for 20 min at OD600 = 0.8. RNA isolation, acylation-dependent pulldown, reverse transcription, and qPCR performed as described in the methods. Points indicate biological replicates (means of qPCR triplicates), columns their means, and error bars the standard deviation. Y axes are scaled consistently to highlight different stmRNA acylation levels between the systems. Note that the Mm control has a strongly reduced RBS strength. Data in b), d) and e) is from the same experiment (one negative Mm sample was lost). Fig.23. Discovery of attenuated 1R26 and Nitra PylRS sets. a) Screening for CbzK-specific 1R26, Nitra, and deltaClos PylRS based on active site transplants from MmPylRS variants and rationally designed derivatives of these. b) and c) Chosen sets of CbzK-selective 1R26 and Nitra variants cover a broad activity range up to their respective wt level (separate experiment from a). a-c) Cells were transformed with plasmids encoding a GFP(150- TAG) reporter, the respective unsplit tRNA underlying the above stmRNA constructs, and either the respective wt PylRS or a mutant carrying the residues indicated in brackets at the position homologous to MmPylRS residues 306, 309, and 346. Main cultures were inoculated 1:50 and grown for 22 h under induction of the reporter in either the presence or absence of a ncM at 2 mM (AllocK for wt PylRS, CbzK for the active site mutants), followed by measurement in a plate reader. For details refer to the methods section. Points indicate biological replicates, columns their means, and error bars the standard deviation. Fig.24. tRNA display by greedy1-1R26 stmRNAs resolves broad range of acylation activity. a) qPCR stmRNA quantification after acylation-dependent pulldown. Experiment as described above (2 mM CbzK), means of three biological replicates (two for GLG -CbzK), error bars: SD. b) +Cbzk data from a) plotted against corresponding GFP data from Fig.23ab. Error bars: SD. Fig.25. Split and circularly permuted tRNAAlamatures and maintains its activity and orthogonality. a) Structure of the split tRNAAlaconstructs. b) Establishment of a tRNAAla-specific fluoro-tREX to detect tRNA acylation. c) Screening of split tRNAAlawith three different intergenic regions. d) Validation of leuP-leuV-based tRNAAlasplit. a-d) Cells were transformed with constructs encoding tRNA (PylT or AlaT) or tRNA and the cognate aaRS. Cultures were inoculated 1:50, cells harvested at OD 0.6 and RNA isolated by acidic chloroform / phenol extraction. Samples were oxidised (or when indicated in b) alternatively also either deacylated+oxidised or untreated to generate negative and positive controls) and concentrations adjusted to a common denominator. Cy3- labelled probes were annealed, a 3’-oxidation-sensitive tRNA elongation carried out in the presence of Cy5-dCTP, urea PAGE performed, and gels imaged for Cy3 and Cy5 fluorescence. Nucleic acids were then stained with SYBR gold and the gels imaged again. Fig.26. Identification of tRNA positions that may be canonically permissive to tRNA splits. a) Structural analysis reveals four regions not participating in secondary or tertiary structure interactions. Figure based on alignment of PDB entries 3JXE and 2AKE. b) Conceivable stem-stabilised split tRNA classes for systems with anticodon recognition. c) Integration of tRNA identity element landscape led to ten canonical splitting sites (black bars) that minimally interfere with tRNA structure and identity elements for all 17 isoacceptor classes exhibiting anticodon recognition. tRNAs are presented by aaRS class and subclass. Invariant positions presented in gray. Positions in red are conserved or nearly conserved strong identity elements for the recognition by the cognate aaRS. Positions in yellow are weak identity determinants or determinants not conserved in the three domains of life. Red dashed lines indicate tertiary interactions involved in identity. Gray lines in the central canonical map indicate tertiary interactions. Figure c) in large parts adapted from Giegé and Eriani (2023- The tRNA identity landscape for aminoacylation and beyond. Nucleic acids research 51, 1528-1570. https: / / doi.org / 10.1093 / nar / gkad007). Fig.27. Design of split tRNAs for a TrpRS / tRNATrpand a ProRS / tRNAPropair under maintenance of tRNA identity elements including the anticodon. a) Assessing activity and tRNA orthogonality of the model pairs by an amber suppression assay with a GFP(150-TAG) reporter. Experiment performed as described above and in the methods. tRNAPyl(Mm), cognate synthetase: MmPylRS (wt; AllocK 2 mM). tRNATrp, TrpTRS (5-OH-Trp 2 mM). tRNATrp, cognate synthetase: ProRS (wt substrate specificity; no ncM used). Points indicate biological replicates, columns their means, and error bars the SD. b) General structures of the split tRNA constructs. c) From 18 possible split site combinations (including stem size compensatory deletions of one or two nucleotides between neighbouring splitting sites) a subset of 10 splits was chosen (black split names), which covered all split classes (i.e. all tRNA loops: D, A, V, T), all possible splitting sites, and within each split class all possible compensatory deletion sizes. Split nomenclature designates the split class (D, A, V, or T), then the two canonical nucleotide position numbers from which the stem protrudes. Two non-consecutive numbers in a name accordingly indicate a compensatory deletion of the nucleotides between these positions. d) Mapping of the chosen splitting sites onto the Trp and Pro tRNA identity element map. Note that the Trp and Pro system belong each to a different of the two existing aaRS classes, which diverge in structure and tRNA binding mode. Figure c) and d) modified from Giegé and Eriani (2023). Fig.28. Screening of diverse tRNA splitting sites reveals functional splitting adjacent to the anticodon and in the D loop for both a tRNATrpand a tRNAPro. Establishment of a fluoro-tREX to detect acylation of a) tRNATrpor b) tRNAProby their cognate aaRS. c) and d) Screening of tRNA splits indicated above the lanes for the tryptophanyl system (c) and the prolyl system (d) by fluoro-tREX. Experiment performed as described before and in the method section. Substrates: 2 mM 5-OH-Trp for the Trp system, no ncM for the Pro pair. Biological replicates of Figures a) and b) were performed, just as the equivalent to c) with the IGR leuW-glnU. Fig.29. Different anticodon adjacent and D loop tRNA splits maintain acylation at varying degrees, while generally improving on orthogonality towards cognate aaRS. a) Validation of identified splitting sites for tRNATrpand b) tRNAPro. Fluoro-tREX performed as described above and in the methods. Substrates: 2 mM 5-OH-Trp for TrpRS, no ncM for ProRS. Neighbouring lanes of identical experimental conditions are biological replicates. Fig.30. Attenuated variant screening in p15a. Fig.31. Second replicate of Fig.28 a) b). Fig.32. Screening of diverse tRNA splitting sites as in Fig.28. DETAILED DESCRIPTION The inventors provide herein methods of measuring the extent of acylation of a tRNA. These methods allow the effect of an agent on the acylation status of a tRNA to be measured. For instance, the methods may be used to: i) determine whether an agent can charge a tRNA, ii) determine the extent or efficiency with which an agent can charge a tRNA, iii) determine whether an agent has an indirect effect on the charging of a tRNA, or iv) whether an agent can chemically alter a substrate in a manner that affects charging or deacylation. The agent may be any polypeptide, nucleic acid, or condition that has or is suspected of having an effect on the acylation of a tRNA. The methods disclosed herein do not require the charged tRNA to be active in a ribosome. Hence, the methods allow the measurement of tRNA acylation with substrates that are not compatible with ribosomal translation, with substrates for which compatible ribosomes have not yet been identified, or with substrates for which compatible ribosomes have not yet been developed. The inventors also provide molecules, tRNAs, nucleic acid constructs, and labelled tRNAs for use with such methods. Method comprising split tRNAs One aspect of the present disclosure relates to the inventors’ demonstration that tRNAs may be split into two or more portions, may be functionally expressed in this format, and may be acylated in this split format. One advantage of splitting tRNAs in this manner is that the inventors demonstrate that additional sequences may be covalently linked to at least one of the tRNA portions. This then enables methods where the identity of an individual agent capable of effecting charging of a tRNA may be measured, even in large parallel libraries. As an example, the inventors fuse acyl-tRNA synthetase genes to a portion of a tRNA. Under conditions where the tRNA is expressed and the acyl-tRNA synthetase is expressed as a polypeptide, the acylation of the tRNA can then be measured and a direct link between the acylation status and the acyl-tRNA synthetase in question can be made. Thus, in a first aspect, there is provided a method of determining the effect of a polypeptide-of-interest or nucleic- acid-of-interest on acylation of a tRNA or acylation status of a tRNA, the method comprising: i) incubating the polypeptide-of-interest or nucleic-acid-of-interest, the tRNA, and a substrate with which the tRNA could be acylated, under conditions conducive to acylation of the tRNA, wherein the tRNA is split into at least two portions and one of said tRNA portions is present as part of a fusion RNA molecule also comprising a sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest; ii) exposing the tRNA to conditions capable of labelling tRNAs that have been acylated; and iii) identifying whether the polypeptide-of-interest or nucleic-acid-of-interest is associated with a labelled tRNA. The polypeptide-of-interest or nucleic-acid-of-interest may be any agent that has or is suspected of having an effect on the acylation status of a tRNA. For instance, the polypeptide-of-interest may be an acyl-tRNA synthetase and the methods may be used to determine if the synthetase can charge a tRNA or the extent or efficiency with which the synthetase can charge a tRNA. In other examples, the nucleic-acid-of-interest may be a ribozyme that is capable of acylating or assisting with the acylation of a tRNA, and this catalysis may be measured. In other examples, the methods may be used to measure deacylation of a tRNA. Thus, an agent suspected of having deacylation activity may be assayed. Such approaches also extend to determining the activity of agents suspected of protecting acylated-tRNA from deacylation. An example of such an agent would be EF-Tu or variants of EF-Tu. EF-Tu is involved in the transport of acylated-tRNAs into the ribosome and is likely important for hard-to-translate monomers. EF-Tu is expected to increase the stability of acylated tRNAs. Hence, the methods disclosed herein may be used to select for EF-Tu variants that protect acylated tRNAs from deacylation. As discussed further herein, such embodiments may include a conditional deacylation step that is distinct from any deacylation performed as a part of tRNA labelling. The agent may not directly charge the tRNA itself and may assist with the charging or may be anywhere upstream of the actual charging. For instance, the agent may be part of a biosynthetic pathway for the creation of a substrate for a tRNA, such as an enzyme for the generation of an amino acid or an unnatural amino acid, and hence the activity of the biosynthetic pathway may be assayed by measuring the extent of acylation of a tRNA. The substrate with which the tRNA is charged may be suitable for incorporation into a polymer by a ribosome, or it may be desired for the substrate to be incorporated into a polymer by a ribosome. Alternatively, the substrate may be suitable for charging the tRNA but not relevant to downstream incorporation into a polymer. Such embodiments are useful, for instance, when evolving enzymes for the generation of substrates that are not relevant to protein production. In such cases, the tRNA display is used as a readout but improved charging of the tRNA is not the ultimate goal. Alternatively, the agent may act upon the substrate after the tRNA has been charged. For instance, the agent may be an enzyme that chemically alters the substrate. The agent may alter the substrate to be more or less resistant to deacylation. Examples of alterations that would alter the stability of the acyl bond include the conversion of an alcohol moiety to halogen moiety (e.g. an Appel reaction), the removal of an amine, or the addition of a C-C bond. In an embodiment, the polypeptide-of-interest is an enzyme capable of altering the substrate in a manner that affects its propensity to deacylation. As discussed further herein, such embodiments may include a conditional deacylation step that is distinct from any deacylation performed as a part of tRNA labelling. The substrate with which the tRNA is acylated may be any suitable. For instance, in some examples, the tRNA may be charged with one of the 22 naturally occurring amino acids or one of the 20 canonical amino acids. In other examples, the tRNA may be charged with an unnatural amino acid, for instance an α-amino acid with a non- natural side chain. The substrate may be an α,α-disubstituted-amino-acid. The substrate may be a non-α-amino acid. The substrate may not be an amino acid at all, for instance the substrate may be a hydroxy acid, optionally the hydroxy acid may be an α-hydroxy acid or a β-hydroxy acid. The methods of the present disclosure are compatible with substrates that may not be suitable for incorporation into a polymer by a ribosome, and so the methods of the present disclosure are particular advantageous when studying acylation of the tRNA with a non- amino-acid substrate. The tRNA may be based on or derived from a naturally occurring or engineered tRNA. For instance, tRNA may be engineered to be compatible with particular synthetases and may have a modified or no anticodon. In preferred embodiments, particularly where the method is carried out in a cell, the tRNA does not comprise an anticodon. The inventors have noted that tRNAs without anticodons are not ribosomal substrates and the experimental tRNAs lacking anticodons interfere with endogenous cellular processes to a lesser extent. The tRNA is split into at least two portions. This means that the tRNA is made up of at least two separate RNA chains that associate to form the tRNA. This may be referred to as a “split” tRNA, which is where a parent tRNA is modified such that part of the tRNA is present on one RNA chain and the other part of the tRNA is present on another RNA chain or chains. In particular examples, the tRNA has been split into two, so one portion is present on one RNA chain and the other portion is present on a second RNA chain; this may be referred to as a first tRNA chain and a second tRNA chain. In some embodiments, the tRNA is split into two portions and no more. In such embodiments, the two portions form a complete tRNA that is capable of being charged and may comprise all sequences of the parental tRNA other than those at the split site. The location of the split should be at a site that is not recognised by the agent that acylates the tRNA. This is so that the split does not, itself, affect the acylation of the tRNA. In an example, the tRNA is split at the anticodon. Thus, the tRNA may be split at the location of the anticodon in the parental tRNA; the parental tRNA is the tRNA from which the split tRNA is derived. The parental tRNA may be a wild type tRNA or an engineered tRNA upon which the split tRNA is based. In some examples of embodiments where the tRNA is split at the anticodon, the tRNA may be a tRNA that can be charged by an acyl-tRNA synthetase that does not recognise the anticodon. An example of a synthetase that does not recognise the anticodon is pyrrolysyl-tRNA synthetase, and so the tRNA may be compatible with a pyrrolysyl- tRNA synthetase or compatibility with such a synthetase may be desired or not desired. The tRNA portions may be portions of a tRNA suitable for being charged, suspected of being suitable for being charged, for which charging is desirable, or for which charging is undesirable by an agent to be tested. The tRNA portions may be portions of a tRNAPyl. In examples, the tRNAPylis a Methanosarcina mazei (Mm) tRNAPyl, Candidatus Methanomethylophilus sp.1R26 (1R26) tRNAPyl, Clostridiales bacterium (deltaClos) I2B72 tRNAPyl, or Nitrososphaeria archaeon (Nitra) Int6C10 tRNAPyl. Purely illustrative examples of sequence encoding portions of a split tRNA are TCCGTTCAGCCGGGTTAGATTCCCGGGGTTTCCGCCA (SEQ ID NO: 1), which is the 3’ half of the Methanosarcina mazei (Mm) tRNAPyland GGAAACCTGATCATGTAGATCGAATGGA (SEQ ID NO: 2), which is the 5’ half of the MmtRNAPyl. Other examples include: The 3’ portion ACCTGTAAGCGGGGTTCGACCCCCCGGCCTTTCGCCA (SEQ ID NO: 23), which encodes the 3’ half of a 1R26 tRNAPyland the 5’ portion GGAGGGCGCTCCGGCGAGCAAACGGGT (SEQ ID NO: 24), which encodes the 5’ half of a 1R26 tRNAPyl. The 3’ portion TCCGCGAACAACGGGTGAAACTCCCGTACACCTCGCCA (SEQ ID NO: 25), which encodes the 3’ half of a deltaClos tRNAPyland the 5’ portion GGGGTGTAGATCGGATTGATCGCGTGGA (SEQ ID NO: 26), which encodes the 5’ half of a deltaClos tRNAPyl. The 3’ portion GCCACGGTTAGCCGGGTTCAACTCCCGGGTTCATCGCCA (SEQ ID NO: 27), which encodes the 3’ half of a Nitra tRNAPyland the 5’ portion GGTGAACTGGTCCGGGACCACCAGGC (SEQ ID NO: 28), which encodes the 5’ half of a Nitra tRNAPyl. The tRNA may be split at the anticodon and may be a tRNALeu, tRNAAla, or tRNASer. The tRNA may be from any domain, may be prokaryotic, or may be from a bacterial species or an archaeal species. A purely illustrative example of a non-PylRS tRNA split at the anticodon is below. This example is representative for any of tRNALeu, tRNAAla, or tRNASerbut is an example of a tRNA that is ordinarily charged by Ala. AGGCGGAGACGAGGGTTCAATTCCCTCCCGGACCACCA (SEQ ID NO: 46), which encodes the 3’ half of a tRNAAlaand GGTCCGGTAGATCAGTGGAAGATCGCCGCTT (SEQ ID NO: 47), which encodes the 5’ half of a tRNAAla. In other examples, the tRNA is not split at the anticodon. This is particularly relevant to embodiments where the tRNA’s anticodon is, at least in part, recognised by a relevant acyl-tRNA synthetase. In some embodiments, the tRNA is split in the D loop, in the anticodon loop and to the 5’ side of the anticodon, in the variable loop, or in the T loop. In particular, embodiments, the tRNA is split at one of the sites illustrated in Fig.27c. The tRNA may be split between residues D15 & D16, D15 & D17, D15 & D18, D16 & D17, D16 & D18, D17 & D18, A31 & A32, A31 & A33, A32 & A33, V45 & V46, V45 & V47, V45 & V48, V46 & V47, V46 & V48, V47 & V48, T56 & T57, T56 & T58, or T57 & T58 (as illustrated in Fig.27c and as defined in Giegé and Eriani (2023 - The tRNA identity landscape for aminoacylation and beyond. Nucleic acids research 51, 1528- 1570. https: / / doi.org / 10.1093 / nar / gkad007). The tRNA may be split between residues D15 & D17, D15 & D18, D16 & D17, A31 & A33, A32 & A33, V45 & V48, V46 & V48, V47 & V48, T56 & T57, or T56 & T58. The tRNA may be from any domain, may be prokaryotic, or may be from a bacterial species or an archaeal species. In some embodiments, the tRNA is a tRNATrpor tRNAPro. tRNATrpor tRNAPromay be split in the D loop, in the anticodon loop and to the 5’ side of the anticodon, in the variable loop, or in the T loop. In particular, embodiments, the tRNATrpor tRNAProis split at one of the sites illustrated in Fig.27c. The tRNATrpor tRNAPromay be split between residues D15 & D16, D15 & D17, D15 & D18, D16 & D17, D16 & D18, D17 & D18, A31 & A32, A31 & A33, A32 & A33, V45 & V46, V45 & V47, V45 & V48, V46 & V47, V46 & V48, V47 & V48, T56 & T57, T56 & T58, or T57 & T58. The tRNATrpor tRNAPromay be split between residues D15 & D17, D15 & D18, D16 & D17, A31 & A33, A32 & A33, V45 & V48, V46 & V48, V47 & V48, T56 & T57, or T56 & T58. The tRNATrpmay be split between A32 & A33, D15 & D17, or D16 & D17. The tRNAPromay be split between A31 & A33, A32 & A33, D15 & D17, or D15 & D18. The tRNA may be from any domain, may be prokaryotic, or may be from a bacterial species or an archaeal species. Examples of sequences encoding a tRNAProsplit outside of the anticodon include the following. The 3’ portion CTAAACCACGCGGTTATGGGTTCAAATCCCATCTTCTCAACCA (SEQ ID NO: 48) and the 5’ portion GAGAAGTAGCACAATTTGGTAGTGCACGTGGT (SEQ ID NO: 49), this is an A31 / A32 split. The 3’ portion CTAAACCACGCGGTTATGGGTTCAAATCCCATCTTCTCAACCA (SEQ ID NO: 50) and the 5’ portion TGAGAAGTAGCACAATTTGGTAGTGCACGTGGTT (SEQ ID NO: 51), this is an A32 / A33 split. The 3’ portion TGGTAGTGCACGTGGTTCTAAACCACGCGGTTATGGGTTCAAATCCCATCTTCTCAACCA (SEQ ID NO: 52) and the 5’ portion TGAGAAGTAGCACAAT (SEQ ID NO: 53), this is a D15 / D17 split. The 3’ portion GGTAGTGCACGTGGTTCTAAACCACGCGGTTATGGGTTCAAATCCCATCTTCTCAACCA (SEQ ID NO: 54) and the 5’ portion TGAGAAGTAGCACAAT (SEQ ID NO: 55), this is a D15 / D18 split. Examples of sequences encoding a tRNATrpsplit outside of the anticodon include the following. The 3’ portion TCTAGACGCGTAGAACCCCGTTCGAGCCGGGGAGCCCTCACCA (SEQ ID NO: 56) and the 5’ portion GGGGGCTTAGTGAAACTGGCATCACGACGCGC (SEQ ID NO: 57), this is an A32 / A33 split. The 3’ portion TGGGATCACGACGAGCTCTAGACTCGTAGAACCCCGTTCGAGCCGGGGAGCCCTCACCA (SEQ ID NO: 58) and the 5’ portion GGGGGCTTAGTGAAA (SEQ ID NO: 59), this is a D15 / D17. The 3’ portion GGCATCACGACGAGCTCTAGACTCGTAGAACCCCGTTCGAGCCGGGGAGCCCTCACCA (SEQ ID NO: 60) and the 5’ portion GGGGGCTTAGTGAAA (SEQ ID NO: 61), this is a D15 / D18 split. The 3’ portion TGGCATCACGACGAGCTCTAGACTCGTAGAACCCCGTTCGAGCCGGGGAGCCCTCACCA (SEQ ID NO: 62) and the 5’ portion GGGGGCTTAGTGAAAC (SEQ ID NO: 63), this is a D16 / D17 split. The splitting of the tRNA provides a site at which further nucleic acid sequence may be included. Thus, one or more of the tRNA portions may be fused to additional nucleic acid sequence. Expressed another way, a fusion RNA molecule may comprise one of the tRNA portions and further sequence. The further sequence may comprise sequence encoding any agent capable of effecting charging of a tRNA or acylation status of a tRNA, as discussed further herein. As such, the further sequence may encode the polypeptide-of-interest or the nucleic-acid-of- interest. In this manner, the sequence of the agent is physically linked to the tRNA for which charging is being assayed. In embodiments where the tRNA is split at the anticodon, each portion may comprise a stem region, wherein the stem region of each portion is complementary. The inventors demonstrate that the presence of such a stem region assists with assembly and stability of the tRNA. The stem region is positioned such that it provides an extension to the anticodon stem loop of the tRNA. The stem region may be from 8 to 14, 8 to 12, or 10 to 12 nucleotides in length. This means that a region of 8 to 14, 8 to 12, or 10 to 12 complementary base pairs is attached to each portion of the tRNA at the anticodon split site. The stem region may be 10 nucleotides in length. The stem region may be 12 nucleotides in length. The stem region may include or be according to a sequence encoded by TAGCGACGTAGC (SEQ ID NO: 3) on one strand and GCTACGTCGCTA (SEQ ID NO: 4) on the other strand. For example, in embodiments where the tRNA is split at the anticodon, a sequence according to SEQ ID NO: 3 may be attached to the 3’ end of the tRNA portion from the 5’ side of the anticodon, and a sequence according to SEQ ID NO: 4 may be attached to the 5’ end of the tRNA portion from the 3’ side of the anticodon. In a preferred embodiment, the stem region may include or be according to a sequence encoded by TAGCGACGTA (SEQ ID NO: 5) on one strand and TACGTCGCTA (SEQ ID NO: 6) on the other strand. For example, in embodiments where the tRNA is split at the anticodon, a sequence according to SEQ ID NO: 5 may be attached to the 3’ end of the tRNA portion from the 5’ side of the anticodon, and a sequence according to SEQ ID NO: 6 may be attached to the 5’ end of the tRNA portion from the 3’ side of the anticodon. In another embodiment, the stem region may include or be according to a sequence encoded by GTACGACCCA (SEQ ID NO: 7) on one strand and TGGGTCGTAC (SEQ ID NO: 8) on the other strand. For example, in embodiments where the tRNA is split at the anticodon, a sequence according to SEQ ID NO: 7 may be attached to the 3’ end of the tRNA portion from the 5’ side of the anticodon, and a sequence according to SEQ ID NO: 8 may be attached to the 5’ end of the tRNA portion from the 3’ side of the anticodon. In another embodiment, the stem region may include or be according to a sequence encoded by TAGCGACGTAG (SEQ ID NO: 64) on one strand and TTATGTCGCTA (SEQ ID NO: 65) on the other strand. For example, in embodiments where the tRNA is split at the anticodon, a sequence according to SEQ ID NO: 64 may be attached to the 3’ end of the tRNA portion from the 5’ side of the anticodon, and a sequence according to SEQ ID NO: 65 may be attached to the 5’ end of the tRNA portion from the 3’ side of the anticodon. Embodiments where the tRNA is split outside of the anticodon may also comprise a stem region attached to each portion. The stem region of each portion may be complementary. The stem region may be from 8 to 25, 10 to 23, 12 to 22, 13 to 21, 14 to 20, 15 to 19, or 16 to 18 nucleotides in length. This means that a region of 8 to 25, 10 to 23, 12 to 22, 13 to 21, 14 to 20, 15 to 19, or 16 to 18 base pairs is attached to each portion of the tRNA at the split site. The stem region may be 17 nucleotides in length. For instance, the stem region may include or be according to a sequence encoded by GGCGGATAGCGACGTAG (SEQ ID NO: 66) on one strand and TTATGTCGCTATCCGCC (SEQ ID NO: 67) on the other strand. For example, a sequence according to SEQ ID NO: 66 may be attached to the 3’ end of the tRNA portion from the 5’ side of the split, and a sequence according to SEQ ID NO: 67 may be attached to the 5’ end of the tRNA portion from the 3’ side of the split. The sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest may be attached to the tRNA via the stem region. Thus, the fusion RNA molecule may comprise the sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest, a stem region, and a tRNA portion, wherein the stem region is situated between the sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest and the tRNA portion. A linker sequence may be present in-between the sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest and the stem region. The linker may be a region of nucleic acids that provide flexibility. The experimental data disclosed herein demonstrate that the linker is not essential. An illustrative sequence encoding a linker is provided below: GCTTAATTAGCTGACCTACTAGTCGGCCGGCGGATGAGAGAAGATTTTCAGCCTGATAC (SEQ ID NO: 9). The sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest may be attached to the 5’ portion of the tRNA or the 3’ portion of the tRNA. In an embodiment, the sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest may be part of a fusion RNA that comprises, from 5’ to 3’, the sequence encoding the polypeptide-of-interest / nucleic-acid-of-interest, a stem region, and a portion of the tRNA from 3’ to the split. In particular embodiments where the tRNA is split at the anticodon, the sequence encoding the polypeptide-of- interest or nucleic-acid-of-interest may be attached to the portion of the tRNA that is 3’ to the anticodon in the parent molecule. Thus, the sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest may be part of a fusion RNA that comprises, from 5’ to 3’, the sequence encoding the polypeptide-of-interest / nucleic-acid-of- interest, a stem region, and a portion of the tRNA from 3’ to the anticodon. In particular, the fusion RNA may comprise, from 5’ to 3’, the sequence encoding the polypeptide-of-interest / nucleic-acid-of-interest, a linker, a stem region, and a portion of the tRNA from 3’ to the anticodon. Illustrative examples of stem regions attached to a tRNA portions are provided below. In these sequences, the underlined portions represent the stem region and the tRNA portions are in plain text. The first example illustrates a sequence encoding the 3’ half of the MmtRNAPylfused to a stem region. The second example illustrates a sequence encoding the 5’ half of the MmtRNAPylfused to a stem region. TACGTCGCTATCCGTTCAGCCGGGTTAGATTCCCGGGGTTTCCGCCA (SEQ ID NO: 11) GGAAACCTGATCATGTAGATCGAATGGATAGCGACGTA (SEQ ID NO: 12) Other illustrative examples are below. In these sequences, the underlined portions represent the stem region and the tRNA portions are in plain text. The first example illustrates a sequence encoding the 3’ half of a 1R26 tRNAPylfused to a stem region. The second example illustrates a sequence encoding the 5’ half of a 1R26 tRNAPylfused to a stem region. TACGTCGCTAACCTGTAAGCGGGGTTCGACCCCCCGGCCTTTCGCCA (SEQ ID NO: 38) GGAGGGCGCTCCGGCGAGCAAACGGGTTAGCGACGTA (SEQ ID NO: 39) Further examples include those below. The underlined portions represent the stem region and the tRNA portions are in plain text. The first of each pair is the 3’ portion of a tRNA fused to a stem region and the second of each pair is the 5’ portion of the tRNA fused to a stem region. TACGTCGCTATCCGCGAACAACGGGTGAAACTCCCGTACACCTCGCCA (SEQ ID NO: 68 - deltaClos tRNAPyl) GGGGTGTAGATCGGATTGATCGCGTGGATAGCGACGTA (SEQ ID NO: 69 - deltaClos tRNAPyl) TACGTCGCTAGCCACGGTTAGCCGGGTTCAACTCCCGGGTTCATCGCCA (SEQ ID NO: 70 - Nitra tRNAPyl) GGTGAACTGGTCCGGGACCACCAGGCTAGCGACGTA (SEQ ID NO: 71 - Nitra tRNAPyl) TTATGTCGCTAAGGCGGAGACGAGGGTTCAATTCCCTCCCGGACCACCA (SEQ ID NO: 72 - tRNAAla) GGTCCGGTAGATCAGTGGAAGATCGCCGCTTTAGCGACGTAG (SEQ ID NO: 73 - tRNAAla) TTATGTCGCTATCCGCCCTAAACCACGCGGTTATGGGTTCAAATCCCATCTTCTCAACCA (SEQ ID NO: 74 - tRNAProA31 / A32 split) GAGAAGTAGCACAATTTGGTAGTGCACGTGGTGGCGGATAGCGACGTAG (SEQ ID NO: 75 - tRNAProA31 / A32 split) TTATGTCGCTATCCGCCCTAAACCACGCGGTTATGGGTTCAAATCCCATCTTCTCAACCA (SEQ ID NO: 76 - tRNAProA32 / A33 split) TGAGAAGTAGCACAATTTGGTAGTGCACGTGGTTGGCGGATAGCGACGTAG (SEQ ID NO: 77 - tRNAProA32 / A33 split) TTATGTCGCTATCCGCCTGGTAGTGCACGTGGTTCTAAACCACGCGGTTATGGGTTCAAATCCCATCT TCTCAACCA (SEQ ID NO: 78 - tRNAProD15 / D17 split) TGAGAAGTAGCACAATGGCGGATAGCGACGTAG (SEQ ID NO: 79 - tRNAProD15 / D17 split) TTATGTCGCTATCCGCCGGTAGTGCACGTGGTTCTAAACCACGCGGTTATGGGTTCAAATCCCATCTT CTCAACCA (SEQ ID NO: 80 - tRNAProD15 / D18 split) TGAGAAGTAGCACAATGGCGGATAGCGACGTAG (SEQ ID NO: 81 - tRNAProD15 / D18 split) TTATGTCGCTATCCGCCTCTAGACGCGTAGAACCCCGTTCGAGCCGGGGAGCCCTCACCA (SEQ ID NO: 82 - tRNATrpA32 / A33 split) GGGGGCTTAGTGAAACTGGCATCACGACGCGCGGCGGATAGCGACGTAG (SEQ ID NO: 83 - tRNATrpA32 / A33 split) TTATGTCGCTATCCGCCTGGGATCACGACGAGCTCTAGACTCGTAGAACCCCGTTCGAGCCGGGGAG CCCTCACCA (SEQ ID NO: 84 - tRNATrpD15 / D17 split) GGGGGCTTAGTGAAAGGCGGATAGCGACGTAG (SEQ ID NO: 85 - tRNATrpD15 / D17 split) TTATGTCGCTATCCGCCGGCATCACGACGAGCTCTAGACTCGTAGAACCCCGTTCGAGCCGGGGAGC CCTCACCA (SEQ ID NO: 86 - tRNATrpD15 / D18 split) GGGGGCTTAGTGAAAGGCGGATAGCGACGTAG (SEQ ID NO: 87 - tRNATrpD15 / D18 split) TTATGTCGCTATCCGCCTGGCATCACGACGAGCTCTAGACTCGTAGAACCCCGTTCGAGCCGGGGAG CCCTCACCA (SEQ ID NO: 88 - tRNATrpD16 / D17 split) GGGGGCTTAGTGAAACGGCGGATAGCGACGTAG (SEQ ID NO: 89 - tRNATrpD16 / D17 split) The tRNA and / or fusion RNA may be expressed from a construct as disclosed for the sixth aspect of the invention. The tRNA and / or fusion RNA may be expressed from a nucleic acid construct encoding a circularly permutated transcript. Two purely illustrative examples of constructs encoding split tRNAs are provided below. In these sequences, the underlined portions represent the stem region, the tRNA portions are in plain text, and the loop regions are in bold. The loop region illustrated for SEQ ID NO: 13 is SEQ ID NO: 19 and is discussed further in relation to the sixth aspect. The loop region illustrated for SEQ ID NO: 14 is SEQ ID NO: 17 and is discussed further in relation to the sixth aspect. TACGTCGCTATCCGTTCAGCCGGGTTAGATTCCCGGGGTTTCCGCCAAACGAGGCGATATCAAAAA AAGTAAGATGACTGTGGAAACCTGATCATGTAGATCGAATGGATAGCGACGTA (SEQ ID NO: 13) TACGTCGCTATCCGTTCAGCCGGGTTAGATTCCCGGGGTTTCCGCCATCTCTTACTTGATATGGCTT TAGTAGCGGTATCAATATCAGCAGTAAAATAAATTTCCCGATGGAAACCTGATCATGTAGATCGA ATGGATAGCGACGTA (SEQ ID NO: 14) Other purely illustrative examples of constructs encoding split tRNAs are provided below. In these sequences, the underlined portions represent the stem region, the tRNA portions are in plain text, and the loop regions are in bold. The loop regions are discussed further in relation to the sixth aspect. TACGTCGCTAACCTGTAAGCGGGGTTCGACCCCCCGGCCTTTCGCCAAATTAGACAGCTATACAAT CGGAGGGCGCTCCGGCGAGCAAACGGGTTAGCGACGTA (SEQ ID NO: 40) TACGTCGCTATCCGCGAACAACGGGTGAAACTCCCGTACACCTCGCCATCTCTTACTTGATATGGC TTTAGTAGCGGTATCAATATCAGCAGTAAAATAAATTTCCCGATGGGGTGTAGATCGGATTGATC GCGTGGATAGCGACGTA (SEQ ID NO: 90) TACGTCGCTAGCCACGGTTAGCCGGGTTCAACTCCCGGGTTCATCGCCATCTCTTACTTGATATGG CTTTAGTAGCGGTATCAATATCAGCAGTAAAATAAATTTCCCGATGGTGAACTGGTCCGGGACC ACCAGGCTAGCGACGTA (SEQ ID NO: 91) TTATGTCGCTAAGGCGGAGACGAGGGTTCAATTCCCTCCCGGACCACCAAACGAGGCGATATCAAA AAAAGTAAGATGACTGTGGTCCGGTAGATCAGTGGAAGATCGCCGCTTTAGCGACGTAG (SEQ ID NO: 92) TTATGTCGCTATCCGCCCTAAACCACGCGGTTATGGGTTCAAATCCCATCTTCTCAACCAAACGAGG CGATATCAAAAAAAGTAAGATGACTGTGAGAAGTAGCACAATTTGGTAGTGCACGTGGTGGCGGA TAGCGACGTAG (SEQ ID NO: 93) TTATGTCGCTATCCGCCCTAAACCACGCGGTTATGGGTTCAAATCCCATCTTCTCAACCAAACGAGG CGATATCAAAAAAAGTAAGATGACTGTTGAGAAGTAGCACAATTTGGTAGTGCACGTGGTTGGCG GATAGCGACGTAG (SEQ ID NO: 94) TTATGTCGCTATCCGCCTGGTAGTGCACGTGGTTCTAAACCACGCGGTTATGGGTTCAAATCCCATCT TCTCAACCAAACGAGGCGATATCAAAAAAAGTAAGATGACTGTTGAGAAGTAGCACAATGGCGG ATAGCGACGTAG (SEQ ID NO: 95) TTATGTCGCTATCCGCCGGTAGTGCACGTGGTTCTAAACCACGCGGTTATGGGTTCAAATCCCATCTT CTCAACCAAACGAGGCGATATCAAAAAAAGTAAGATGACTGTTGAGAAGTAGCACAATGGCGGA TAGCGACGTAG (SEQ ID NO: 96) TTATGTCGCTATCCGCCTCTAGACGCGTAGAACCCCGTTCGAGCCGGGGAGCCCTCACCAAACGAG GCGATATCAAAAAAAGTAAGATGACTGTGGGGGCTTAGTGAAACTGGCATCACGACGCGCGGCGG ATAGCGACGTAG (SEQ ID NO: 97) TTATGTCGCTATCCGCCTGGGATCACGACGAGCTCTAGACTCGTAGAACCCCGTTCGAGCCGGGGAG CCCTCACCAAACGAGGCGATATCAAAAAAAGTAAGATGACTGTGGGGGCTTAGTGAAAGGCGGA TAGCGACGTAG (SEQ ID NO: 98) TTATGTCGCTATCCGCCGGCATCACGACGAGCTCTAGACTCGTAGAACCCCGTTCGAGCCGGGGAGC CCTCACCAAACGAGGCGATATCAAAAAAAGTAAGATGACTGTGGGGGCTTAGTGAAAGGCGGAT AGCGACGTAG (SEQ ID NO: 99) TTATGTCGCTATCCGCCTGGCATCACGACGAGCTCTAGACTCGTAGAACCCCGTTCGAGCCGGGGAG CCCTCACCAAACGAGGCGATATCAAAAAAAGTAAGATGACTGTGGGGGCTTAGTGAAACGGCGG ATAGCGACGTAG (SEQ ID NO: 100) Step i) of the method of the first aspect comprises incubating the polypeptide-of-interest / nucleic-acid-of-interest, the tRNA, and the substrate with which the tRNA could be acylated, under conditions conducive to acylation of the tRNA. Such conditions may be within a cell. Thus, the polypeptide-of-interest / nucleic-acid-of-interest and the tRNA may be expressed within a cell, for instance from one or more vectors. In some embodiments, the tRNA and the mRNA of the polypeptide-of-interest are expressed from the same DNA construct, and the polypeptide-of- interest mRNA is then translated. In other embodiments, a portion of the tRNA and the polypeptide-of-interest mRNA are expressed from the same DNA construct, and the polypeptide-of-interest mRNA is then translated. As discussed, the mRNA of the polypeptide-of-interest may be directly or indirectly fused to a portion of the tRNA. The substrate may be produced within the cell or may be provided exogenously. The cell may be cultured in conditions that, in normal circumstances, would enable the charging of a tRNA with a substrate. In some examples, the cell is a prokaryotic cell, a bacterial cell, or an E. coli cell. As discussed herein, the polypeptide-of-interest or nucleic-acid-of-interest may be an agent that can act upon the substrate after the tRNA has been acylated to alter the susceptibility of the substrate-tRNA to deacylation. In such embodiments, step i) may comprise first incubating the tRNA and the substrate under conditions conducive to acylation of the tRNA and then incubating the acylated tRNA with the polypeptide-of-interest or nucleic-acid-of- interest; these incubations may be concurrent or performed in a cell expressing the relevant molecules. In such embodiments, the methods may comprise a step of exposing the tRNAs to deacylation conditions in-between step i) and step ii), where the deacylation conditions are such that only substrates that have been rendered more stable are retained or are such that only substrates that have been rendered less stable are removed. This step may be referred to as a conditional deacylation step. In an example, the polypeptide-of-interest encodes an enzyme, and it is desired to determine whether the enzyme can act upon a substrate in a substrate-tRNA complex to increase the stability of the substrate-tRNA complex. The tRNA is charged with the substrate, exposed to the enzyme, and then exposed to conditions that would deacylate substrates for which the stability has not been increased. The method may then proceed to step ii), allowing the identification of whether the enzyme is capable of increasing the stability of the substrate because such enzymes will be associated with tRNAs that retain substrates and so are labelled. As discussed herein, labelling step ii) may also comprise a deacylation step. This second deacylation step should remove all substrates in order to allow the labelling process to proceed. Step ii) of the method of the first aspect comprises exposing the tRNA to conditions capable of labelling tRNAs that have been acylated. This step may be performed by any suitable method, for instance the substrate with which the tRNA is acylated may be labelled, and hence step ii) may be performed concurrently with step i). The substrate may be labelled, directly or indirectly, with a label. The label may be an optically detectable label, such as a fluorescent label, or may be a physically detectable label, such as biotin or a magnetic bead. The physically detectable label may be capable of being bound to an immobilised or magnetic moiety, to allow the separation of labelled molecules. The substrates may be labelled with a first label comprising a ligand or receptor of a ligand- receptor pair into the substrate, for instance biotin. The first label may then be further labelled by a second label capable of binding the first label, such as streptavidin. For example, a biotinylated substate may be further labelled by a fluorescent molecule attached to streptavidin. Alternatively, immobilised streptavidin may be used to separate biotinylated substrate-tRNA complexes from other components. In some embodiments, the label is not a part of the substrate or attached to the substrate. The methods of labelling the tRNAs may be as disclosed for the fourth aspect of the invention. In preferred embodiments, the tRNA is labelled by extension of the tRNA itself. For instance, additional nucleotides may be added to the 3’ end of tRNAs that have previously been acylated. To ensure that only tRNAs that have been acylated are labelled, the tRNA may be exposed to conditions that block the 3’ end of the free tRNAs but do not block the 3’ end of acylated tRNAs. Any conditions preventing the subsequent extension of the tRNA by the addition of nucleotides to the 3’ end would be suitable, and an example is exposure to sodium periodate under oxidising conditions. Such conditions will not block acylated tRNAs because they are protected by the substrate acylated to the tRNA. The tRNA may then be deacylated. The tRNA may be exposed to conditions suitable for the removal of substrates that have acylated the tRNA, such that tRNAs that have formerly been acylated have an intact 3’ ribose to which additional nucleotides may be added whereas tRNAs that were not acylated are blocked. As discussed herein, some embodiments relate to polypeptides-of-interest or nucleic-acids-of-interest that are tested to determine whether they can alter the substrate in a manner that alters the propensity of the substrate- tRNA complex to deacylation. Such embodiments can include a first deacylation step which is a conditional deacylation step. For example, where the conditions are chosen to not deacylate substrate-tRNA complexes that have been rendered more stable or where the conditions are chosen to only deacylate substrate-tRNA complexes that have been rendered less stable. In such embodiments, the conditions that block the 3’ end of the free tRNAs but do not block the 3’ end of acylated tRNAs are applied after the first deacylation step. After the blocking step a second deacylation step is performed to remove all substrates that have acylated the tRNA. The addition of nucleotides may be achieved by the annealing of an oligonucleotide to the 3’ end of the tRNA, where the oligonucleotide overhangs the 3’ end of the tRNA, followed by the extension of the tRNA by polymerisation based on the template provided by the oligonucleotide. The oligonucleotide may be fluorescently labelled. The extension of the 3’ end of the tRNA may be in accordance with the method tREX as described in Cervettini, et al. (Rapid discovery and evolution of orthogonal aminoacyl-tRNA synthetase–tRNA pairs. Nature Biotechnology 38, 989-999, doi:10.1038 / s41587-020-0479-2 (2020); incorporated herein by reference). In embodiments where the tRNA is labelled by extension with additional nucleotides, labelled tRNAs may be identified based on the length of the tRNA, such as by the use of gel electrophoresis. Alternatively, the tRNA may be labelled by extension of the tRNA with a particular nucleotide sequence that can be identified. For instance, a particular nucleic acid sequence may be identified by the binding of a fluorescently labelled complementary probe or may be identified by subsequently sequencing the tRNAs and identifying the additional sequence. The additional sequence may comprise a barcode. The additional nucleic acid sequence may comprise a primer binding site, hence allowing the amplification of tRNAs that have been acylated. The tRNA may be labelled by the extension of the tRNA with at least one nucleotide comprising a label. The label may be directly detectable, for instance by optical methods. The label may be a fluorescent label. The label may be a first label comprising a ligand or receptor of a ligand-receptor pair. For instance, the first label may be bound by a second label comprising the respective receptor or ligand and an optically detectable moiety such as a fluorescent moiety. The first label may be bound by a second label that is immobilised, allowing the capture of the tRNA. The first label may be bound by a second label that comprises a physically detectable moiety, such as a magnetic bead allowing pull-down of the tRNA. An example of a suitable ligand-receptor is biotin-streptavidin, but other ligand-receptor pairs would be suitable for use with the invention. Also encompassed are interactions with an intermediate molecule or molecules between the first label and the second label, such as an adapter molecule or molecules. The method of the first aspect may comprise the capture of labelled tRNAs or RNA fusion molecules via the first label or second label bound to the first label. The capture of the tRNA may only capture one portion of the tRNA and does not need to capture all portions. In an embodiment, the capture comprises the capture of the RNA fusion molecule comprising a portion of the tRNA and the sequencing encoding the polypeptide-of-interest or nucleic- acid-of-interest. Examples of suitable labels for capture are discussed herein, and include biotin, magnetic moieties, or other ligand-receptor pairs where one moiety is immobilised. “Capture” as used for any aspect herein, is where the molecule is physically separated from other components. For instance, the captured molecules may be held in position while other components are removed. Or the captured molecules may be removed and retained, e.g. by magnetism, while other components are left in place. The method of the first aspect comprises identifying whether the polypeptide-of-interest or nucleic-acid-of- interest is associated with a labelled tRNA. As such, it can be determined whether a particular agent is capable of affecting the charging status of a tRNA or whether a particular agent is not capable of affecting the charging status of a tRNA. This information may include whether the agent leads to an increase or decrease in the level of charging compared to a control or a comparator. The information can also include whether the agent leads to an increase or decrease the susceptibility of the substrate-tRNA complex to deacylation. In embodiments where the polypeptide-of-interest is an acyl-tRNA synthetase, it may be determined that the synthetase is capable of charging a tRNA with a particular substrate or it may be determined whether a variant has increased or decreased charging functionality. It may be determined if the synthetase is unable to charge a tRNA with a particular substrate. In some examples, a library of polypeptides-of-interest or nucleic-acids-of-interest may be tested in parallel to identify variants with particular properties. In other examples, multiple synthetases may be tested to identify synthetases that have activity to a particular tRNA and substrate but do not have activity to other tRNAs and substrates. Thus, sets of synthetases may be identified that do not cross-react and hence may be used in concert. As such, the methods of the present disclosure may be used to identity sets of orthogonal synthase-tRNA pairs. The identification of the polypeptide-of-interest or nucleic-acid-of-interest may be via sequencing a barcode associated with tRNAs that are labelled. In examples, the barcode is within the tRNA, is a part of additional nucleotides added as a part of the labelling reaction, or is comprised within sequences fused to the split-site. Alternatively, the sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest itself may be, at least in part or completely, sequenced. As such, any polypeptides-of-interest / nucleic-acids-of-interest that are associated with a tRNA that was previously charged can be directly identified. As discussed, the tRNA may be labelled in a manner that allows the physical separation of the tRNA if it was previously charged. Sequences associated with the physically separated tRNAs may be sequenced, hence identifying sequences associated with tRNAs that were charged. The physical separation may only be of the portion of the tRNA fused to the sequence encoding the polypeptide-of-interest / nucleic-acid-of-interest. Hence, the physical separation may be of the fusion RNA molecule comprising a tRNA portion and a sequence encoding the polypeptide-of-interest / nucleic-acid-of-interest. Alternatively, sequences associated with all tRNAs, whether labelled or not, may be sequenced and the label may be used to identify sequences that are associated with tRNAs. For instance, all of the fusion RNA molecules may be sequenced, in part or in their entirety, and additional nucleic acid sequence added as a label may be used to identify sequences associated with charged tRNAs. Such sequencing may comprise sequencing all or a part of the sequence encoding the polypeptide-of-interest / nucleic-acid-of-interest. Split tRNAs As discussed in relation to the first aspect, the inventors have developed split tRNAs that find utility in methods of determining the acylation status of a tRNA or efficiency of acylation of a tRNA. Thus, in a second aspect of the invention, there is provided a split tRNA comprising a first chain and a second chain, wherein the first chain comprises a first tRNA portion and comprises a first stem region, and the second chain comprises a second tRNA portion and comprises a second stem region. The split tRNA may be any as discussed in relation to the first aspect. In a particular embodiment, there is provided a split tRNA comprising a first chain and a second chain, wherein: the first chain comprises a first tRNA portion and comprises a first stem region, and the second chain comprises a second tRNA portion and comprises a second stem region; the first tRNA portion corresponds to the 5’ portion of a parental tRNA split at the anticodon and the second tRNA portion corresponds to the 3’ portion of the parental tRNA split at the anticodon; the first stem region is located at the 3’ end of the first tRNA portion and the second stem region is located at the 5’ end of the second tRNA portion; and the first stem region and the second stem region are complementary. As discussed in relation to the first aspect, a split tRNA is a tRNA that has been split into at least two portions, meaning that the tRNA is made up of at least two separate RNA chains that associate to form the tRNA. In particular examples, the tRNA is split into two portions and no more. In such embodiments, the two portions form a complete tRNA that is capable of being charged and may comprise all sequences of the parental tRNA other than those at the split site. In some embodiments of the second aspect, the tRNA is split at the anticodon. Thus, the tRNA is split at the location of the anticodon in the parental tRNA; the parental tRNA is the tRNA from which the split tRNA is derived. The parental tRNA may be a wild type tRNA or an engineered tRNA upon which the split tRNA is based. The split tRNA may be compatible with a pyrrolysyl-tRNA synthetase or compatibility with such a synthetase may be desired or not desired. The split tRNA may comprise portions of a tRNA suitable for being charged, suspected of being suitable for being charged, for which charging is desirable, or for which charging is undesirable by an agent to be tested. The split tRNA may comprise portions of a tRNAPyl. In examples, the tRNAPylis an mmtRNAPyl, 1R26 tRNAPyl, deltaClos tRNAPyl, or Nitra tRNAPyl. In other examples, the tRNA may be split at the anticodon and may be a tRNALeu, tRNAAla, or tRNASer. The tRNA may be from any domain, may be prokaryotic, or may be from a bacterial species or an archaeal species. In other embodiments, the tRNA is not split at the anticodon. This is particularly relevant to embodiments where the tRNA’s anticodon is, at least in part, recognised by a relevant acyl-tRNA synthetase. In some embodiments, the tRNA is split in the D loop, in the anticodon loop and to the 5’ side of the anticodon, in the variable loop, or in the T loop. In particular, embodiments, the tRNA is split at one of the sites illustrated in Fig.27c. The tRNA may be split between residues D15 & D16, D15 & D17, D15 & D18, D16 & D17, D16 & D18, D17 & D18, A31 & A32, A31 & A33, A32 & A33, V45 & V46, V45 & V47, V45 & V48, V46 & V47, V46 & V48, V47 & V48, T56 & T57, T56 & T58, or T57 & T58 (as illustrated in Fig.27c and as defined in Giegé and Eriani (2023)). The tRNA may be split between residues D15 & D17, D15 & D18, D16 & D17, A31 & A33, A32 & A33, V45 & V48, V46 & V48, V47 & V48, T56 & T57, or T56 & T58. The tRNA may be from any domain, may be prokaryotic, or may be from a bacterial species or an archaeal species. In some embodiments, the tRNA is a tRNATrpor tRNAPro. tRNATrpor tRNAPromay be split in the D loop, in the anticodon loop and to the 5’ side of the anticodon, in the variable loop, or in the T loop. In particular, embodiments, the tRNATrpor tRNAProis split at one of the sites illustrated in Fig.27c. The tRNATrpor tRNAPromay be split between residues D15 & D16, D15 & D17, D15 & D18, D16 & D17, D16 & D18, D17 & D18, A31 & A32, A31 & A33, A32 & A33, V45 & V46, V45 & V47, V45 & V48, V46 & V47, V46 & V48, V47 & V48, T56 & T57, T56 & T58, or T57 & T58. The tRNATrpor tRNAPromay be split between residues D15 & D17, D15 & D18, D16 & D17, A31 & A33, A32 & A33, V45 & V48, V46 & V48, V47 & V48, T56 & T57, or T56 & T58. The tRNATrpmay be split between A32 & A33, D15 & D17, or D16 & D17. The tRNAPromay be split between A31 & A33, A32 & A33, D15 & D17, or D15 & D18. The tRNA may be from any domain, may be prokaryotic, or may be from a bacterial species or an archaeal species. Illustrative examples of tRNA portions are encoded by SEQ ID NO: 1 & SEQ ID NO: 2, SEQ ID NO: 23 & SEQ ID NO: 24, SEQ ID NO: 25 & SEQ ID NO: 26, SEQ ID NO: 27 & SEQ ID NO: 28, SEQ ID NO: 46 & SEQ ID NO: 47, SEQ ID NO: 48 & SEQ ID NO: 49, SEQ ID NO: 50 & SEQ ID NO: 51, SEQ ID NO: 52 & SEQ ID NO: 53, SEQ ID NO: 54 & SEQ ID NO: 55, SEQ ID NO: 56 & SEQ ID NO: 57, SEQ ID NO: 58 & SEQ ID NO: 59, SEQ ID NO: 60 & SEQ ID NO: 61, and SEQ ID NO: 62 & SEQ ID NO: 63. These sequences are further discussed in relation to the first aspect. The first chain of the split tRNA comprises the first stem region located at the 3’ end of the first tRNA portion and the second chain of the split tRNA comprises the second stem region located at the 5’ end of the second tRNA portion. The first and second stem regions may form a complementary extension to the anticodon stem-loop of the tRNA. The first and the second stem region may be from 8 to 14, 8 to 12, or 10 to 12 nucleotides in length. The first and second stem region may be 10 nucleotides in length. The first and second stem region may be 12 nucleotides in length. The first stem region may include or be according to a sequence encoded by SEQ ID NO: 3 and the second stem region may include or be according to a sequence encoded by SEQ ID NO: 4. In a preferred embodiment, the first stem region may include or be according to a sequence encoded by SEQ ID NO: 5 and the second stem region may include or be according to a sequence encoded by SEQ ID NO: 6. The first stem region may include or be according to a sequence encoded by SEQ ID NO: 7 and the second stem region may include or be according to a sequence encoded by SEQ ID NO: 8. The first stem region may include or be according to a sequence encoded by SEQ ID NO: 64 and the second stem region may include or be according to a sequence encoded by SEQ ID NO: 65. Embodiments where the tRNA is split outside of the anticodon may also comprise a stem region attached to each portion. The stem region of each portion may be complementary. The stem region may be from 8 to 25, 10 to 23, 12 to 22, 13 to 21, 14 to 20, 15 to 19, or 16 to 18 nucleotides in length. This means that a region of 8 to 25, 10 to 23, 12 to 22, 13 to 21, 14 to 20, 15 to 19, or 16 to 18 base pairs is attached to each portion of the tRNA at the split site. The stem region may be 17 nucleotides in length. For instance, the stem region may include or be according to a sequence encoded by GGCGGATAGCGACGTAG (SEQ ID NO: 66) on one strand and TTATGTCGCTATCCGCC (SEQ ID NO: 67) on the other strand. For example, a sequence encoded by SEQ ID NO: 66 may be attached to the 3’ end of the tRNA portion from the 5’ side of the split, and a sequence encoded by SEQ ID NO: 67 may be attached to the 5’ end of the tRNA portion from the 3’ side of the split. In a particular embodiment, the first and / or second chain comprises additional sequence, which may be attached to the stem region. In a particular embodiment, the second chain comprises additional sequence that is located 5’ to the second stem region. The additional sequence may encode a barcode or any relevant sequence. In particular embodiments, the additional sequence may comprise sequence encoding a polypeptide-of-interest or nucleic-acid- of-interest. The polypeptide-of-interest or nucleic-acid-of-interest may be any as described for the first aspect of the present disclosure. In a particular embodiment, the sequence encodes an acyl-tRNA synthetase. The sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest may be attached directly to the stem region. Alternatively, a linker sequence may be present in-between the sequence encoding the polypeptide-of- interest or nucleic-acid-of-interest and the stem region. The linker may be a region of nucleic acids that provide flexibility. The linker may be as discussed for the first aspect. Illustrative examples of split tRNAs are provided in relation to the first aspect. In particular, examples of tRNA portions include those encoded by SEQ ID NO: 1 and SEQ ID NO: 2. Illustrative examples of tRNA portions fused to stem regions are provided in SEQ ID NOs: 11 and 12. Illustrative examples of constructs encoding split tRNAs are provided in SEQ ID NO: 13 and SEQ ID NO: 14. Other illustrative examples encoding tRNA portions include SEQ ID NO: 23 and SEQ ID NO: 24, which are illustratively fused to stem regions in SEQ ID NO: 38 and SEQ ID NO: 39. And an exemplary construct encoding this split tRNA is presented in SEQ ID NO: 40. These examples are purely illustrative and the other examples disclosed herein are also relevant. RNA fusion molecules As discussed in relation to the first aspect, the inventors have developed split tRNAs that can tolerate additional fused sequence and still be capable of being charged. For instance, the split tRNAs may comprise additional nucleic acid sequence encoding proteins or nucleic acids. These split RNAs, as discussed for the second aspect, may comprise an RNA molecule that includes a portion of a tRNA and a sequence encoding a polypeptide-of- interest / nucleic-acid-of-interest. Thus, in a third aspect of the invention, there is provided an RNA molecule comprising a sequence encoding a polypeptide-of-interest or nucleic-acid-of-interest and comprising a portion of a tRNA. The polypeptide-of-interest or nucleic-acid-of-interest may be any as described for the first aspect of the present disclosure. The portion of a tRNA is a part of a tRNA and is not an intact tRNA. The portion may be the portion to the 5’ side of split site in the parental tRNA from which the portion was derived or the portion may be the portion to the 3’ side of a split site in the parental tRNA from which the portion was derived. The portion may be a portion of a tRNA suitable for being charged, suspected of being suitable for being charged, for which charging is desirable, or for which charging is underivable by an agent to be tested. The portion may be the portion to one side of an anticodon site in the parental tRNA from which the portion was derived. The parental tRNA may be a wild type tRNA or an engineered tRNA. The portion may be the portion to the 5’ side of an anticodon site in the parental tRNA from which the portion was derived or the portion may be the portion to the 3’ side of an anticodon site in the parental tRNA from which the portion was derived. In a particular embodiment, the portion is the portion to the 3’ side of an anticodon site in the parental tRNA from which the portion was derived. The portion may be a portion of a tRNA suitable for being charged, suspected of being suitable for being charged, for which charging is desirable, or for which charging is underivable by an agent to be tested. The portion may be a portion of a tRNA that is compatible with a pyrrolysyl-tRNA synthetase, for which compatibility with a pyrrolysyl-tRNA synthetase is desired, or for which compatibility with a pyrrolysyl-tRNA synthetase. The portion may be a portion of a tRNAPyl. In examples, the tRNAPylis an mmtRNAPyl, 1R26 tRNAPyl, deltaClos tRNAPyl, or Nitra tRNAPyl. In other examples, the portion may be from a tRNA split at the anticodon which is a tRNALeu, tRNAAla, or tRNASer. The tRNA may be from any domain, may be prokaryotic, or may be from a bacterial species or an archaeal species. An illustrative portion is encoded by the sequence provided as SEQ ID NO: 1. Other examples include SEQ ID NO: 23, SEQ ID NO: 25, SEQ ID NO: 27, and SEQ ID NO: 46. In other examples, the portion is from a tRNA that is not split at the anticodon. This is particularly relevant to embodiments where the tRNA’s anticodon is, at least in part, recognised by a relevant acyl-tRNA synthetase. In some embodiments, the portion is from a tRNA split in the D loop, in the anticodon loop and to the 5’ side of the anticodon, in the variable loop, or in the T loop. In particular, embodiments, the portion is from a tRNA split at one of the sites illustrated in Fig.27c. The portion may be from a tRNA split between residues D15 & D16, D15 & D17, D15 & D18, D16 & D17, D16 & D18, D17 & D18, A31 & A32, A31 & A33, A32 & A33, V45 & V46, V45 & V47, V45 & V48, V46 & V47, V46 & V48, V47 & V48, T56 & T57, T56 & T58, or T57 & T58 (as illustrated in Fig.27c and as defined in Giegé and Eriani (2023)). The portion may be from a tRNA split between residues D15 & D17, D15 & D18, D16 & D17, A31 & A33, A32 & A33, V45 & V48, V46 & V48, V47 & V48, T56 & T57, or T56 & T58. The portion may be from a tRNA from any domain, may be prokaryotic, or may be from a bacterial species or an archaeal species. In some embodiments, the portion is from a tRNA that is a tRNATrpor tRNAPro. The portion may be from a tRNATrpor tRNAProsplit in the D loop, in the anticodon loop and to the 5’ side of the anticodon, in the variable loop, or in the T loop. In particular, embodiments, the portion is from a tRNATrpor tRNAProsplit at one of the sites illustrated in Fig.27c. The portion may be from a tRNATrpor tRNAProsplit between residues D15 & D16, D15 & D17, D15 & D18, D16 & D17, D16 & D18, D17 & D18, A31 & A32, A31 & A33, A32 & A33, V45 & V46, V45 & V47, V45 & V48, V46 & V47, V46 & V48, V47 & V48, T56 & T57, T56 & T58, or T57 & T58. The portion may be from a tRNATrpor tRNAProsplit between residues D15 & D17, D15 & D18, D16 & D17, A31 & A33, A32 & A33, V45 & V48, V46 & V48, V47 & V48, T56 & T57, or T56 & T58. The portion may be from a tRNATrpsplit between A32 & A33, D15 & D17, or D16 & D17. The portion may be from a tRNAProsplit between A31 & A33, A32 & A33, D15 & D17, or D15 & D18. The portion may be from a tRNA may be from any domain, may be prokaryotic, or may be from a bacterial species or an archaeal species. Illustrative portions are encoded by the sequences provided as SEQ ID NOs: 48, 50, 52, 54, 56, 58, 60, and 62. The RNA molecule of the third aspect may comprise a stem region. The stem region may be as disclosed for the first aspect. The stem region may be from 8 to 14, 8 to 12, or 10 to 12 nucleotides in length. The stem region may be 10 or 12 nucleotides in length. The stem region may include or be according to a sequence encoded by SEQ ID NO: 4, SEQ ID NO: 6, SEQ ID NO: 8, or SEQ ID NO: 64. In a particular embodiment, the stem region is according to a sequence encoded by SEQ ID NO: 6. The stem region may be positioned such that it provides an extension to the anticodon stem loop. Embodiments where the tRNA portion is from a tRNA split outside of the anticodon may also comprise a stem region. The stem region may be from 8 to 25, 10 to 23, 12 to 22, 13 to 21, 14 to 20, 15 to 19, or 16 to 18 nucleotides in length. The stem region may be 17 nucleotides in length. For instance, the stem region may include or be according to a sequence encoded by SEQ ID NO: 66. The stem region may be positioned in-between the sequence encoding polypeptide-of-interest / nucleic-acid-of- interest and the portion of a tRNA. In a particular embodiment, the RNA molecule of the third aspect comprises a stem region connected to a portion of a tRNA that is the portion to the 3’ side of split site in a parental tRNA. As such, the RNA molecule comprises the 3’ half of a tRNA that is split and is attached to an extension of the stem region. Purely illustrative examples are provided as the RNAs encoded by SEQ ID NO: 11, SEQ ID NO: 38, SEQ ID NO: 68, SEQ ID NO: 70, SEQ ID NO: 72, SEQ ID NO: 74, SEQ ID NO: 76, SEQ ID NO: 78, SEQ ID NO: 80, SEQ ID NO: 82, SEQ ID NO: 84, SEQ ID NO: 86, and SEQ ID NO: 88. In a particular embodiment, the RNA molecule of the third aspect comprises a stem region connected to a portion of a tRNA that is the portion to the 3’ side of an anticodon site in a parental tRNA. As such, the RNA molecule comprises the 3’ half of a tRNA that is split at the anticodon and is attached to an extension of the stem region. Purely illustrative examples are provided as the RNAs encoded by SEQ ID NO: 11, SEQ ID NO: 38, SEQ ID NO: 68, SEQ ID NO: 70, and SEQ ID NO: 72. The sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest may be attached directly to the stem region. Alternatively, a linker sequence may be present in-between the sequence encoding the polypeptide-of- interest or nucleic-acid-of-interest and the stem region. The linker may be a region of nucleic acids that provide flexibility. The linker is optional and suitable examples are discussed for the first and second aspects. In a particular embodiment, the RNA molecule may comprise, from 5’ to 3’, sequence encoding a polypeptide-of- interest or nucleic-acid-of-interest, optionally a linker, a stem region, and a portion of a tRNA, wherein the portion of a tRNA corresponds to the 3’ portion of a split parental tRNA. In a particular embodiment, the RNA molecule may comprise, from 5’ to 3’, sequence encoding a polypeptide-of-interest or nucleic-acid-of-interest, optionally a linker, a stem region, and a portion of a tRNA, wherein the portion of a tRNA corresponds to the 3’ portion of a parental tRNA split at the anticodon. Methods of determining the acylation status of a tRNA or efficiency of acylation of a tRNA The inventors have further developed methods of labelling tRNAs that have been acylated. These methods enable the identification of tRNAs that have been charged and are particularly sensitive. An embodiment of these methods is referred to herein as fluoro-tREX. This method enables a specific acylated tRNA, from a pool of tRNAs isolated from cells, to be labelled by primer extension with fluorescent dNTPs. Another embodiment of these methods is referred to herein as bio-tREX. Bio-tREX enables the selective isolation of a specific acylated tRNA, by using primer extension with biotinylated dNTPs and streptavidin pulldown. Thus, in a fourth aspect of the invention, there is provided a method of determining the acylation status of a tRNA or efficiency of acylation of a tRNA, the method comprising: i) incubating the tRNA and a substrate with which the tRNA could be acylated, under conditions conducive to acylation of the tRNA; ii) exposing the tRNA to conditions capable of blocking the 3’ end of free tRNAs; and iii) exposing the tRNA to conditions that lead to the addition of nucleotides to the 3’ end of tRNAs that are not blocked, wherein at least one nucleotide comprises a label. The tRNA may be a naturally occurring or engineered tRNA. The tRNA may be an intact tRNA. The tRNA may be engineered to be compatible with particular synthetases and may have a modified or no anticodon. In preferred embodiments, particularly where the method is carried out in a cell, the tRNA does not comprise an anticodon. The inventors have noted that tRNAs without anticodons are not ribosomal substrates and the experimental tRNAs lacking anticodons interfere with endogenous cellular processes to a lesser extent. The tRNA may be any as described in relation to the first or second aspect of the present disclosure or may be formed by any RNA molecule of the third aspect of the present disclosure. In examples, the tRNA may be a split tRNA, the tRNA may be split at the anticodon, the tRNA may be derived from a pyrrolysyl-tRNA synthetase compatible tRNA, the tRNA may be a tRNAPylor derived from a tRNAPyl. In examples, the tRNAPylis an mmtRNAPyl, 1R26 tRNAPyl, deltaClos tRNAPyl, or Nitra tRNAPyl. In other examples, the portion may be from a tRNA split at the anticodon which is a tRNALeu, tRNAAla, or tRNASer. The tRNA may be from any domain, may be prokaryotic, or may be from a bacterial species or an archaeal species. The substrate may be any as described in relation to the first aspect of the invention. Step i) may be as described for the first aspect. For instance, the conditions may be within a cell. Thus, the tRNA may be expressed within a cell, for instance from one or more vectors. The substrate may be produced within the cell or may be provided exogenously. The cell may be cultured in conditions that, in normal circumstances, would enable the charging of a tRNA with a substrate. In some examples, the cell is a prokaryotic cell, a bacterial cell, or an E. coli cell. Step i) may comprise exposure to an agent that has or is suspected of having an effect on the acylation status of a tRNA. The agent may be any polypeptide, nucleic acid, or condition that has or is suspected of having an effect on the acylation status of a tRNA. In particular, the agent may be any polypeptide-of-interest or nucleic-acid-of-interest as discussed for the first aspect. As discussed herein, step i) may comprise charging the tRNA with the substrate and exposing the substrate-tRNA complex to an agent capable of chemically altering the substrate. For instance, the agent may increase or decrease the susceptibility of the substrate-tRNA complex to deacylation. Thus, in-between step i) and step ii), there may be a conditional deacylation step. In step ii) of the fourth aspect, the tRNA is exposed to conditions capable of blocking the 3’ end of free tRNAs. The 3’ end of acylated tRNAs would not be blocked under such conditions. Any conditions preventing the subsequent extension of the tRNA by the addition of nucleotides to the 3’ end would be suitable, and an example is exposure to sodium periodate under oxidising conditions. Such conditions will not block acylated tRNAs because they are protected by the substrate acylated to the tRNA. The tRNA may then be exposed to conditions suitable for the removal of substrates that have acylated the tRNA, such that tRNAs that have formerly been acylated have an intact 3’ ribose to which additional nucleotides may be added, whereas tRNAs that were not acylated are blocked. Thus, subsequent to step ii), the method may comprise the deacylation of the tRNA. In some embodiments the method comprises a conditional deacylation step prior to step ii) and, in such embodiments, the deacylation step subsequent to step ii) is a second deacylation step. In step iii) of the fourth aspect, the tRNA is exposed to conditions that lead to the addition of nucleotides to the 3’ end of tRNAs that are not blocked, wherein at least one nucleotide comprises a label. As discussed in relation to the first aspect, the addition of nucleotides may be achieved by the annealing of an oligonucleotide to the 3’ end of the tRNA, where the oligonucleotide overhangs the 3’ end of the tRNA, and the extension of the tRNA by polymerisation based on the template provided by the oligonucleotide. The oligonucleotide may be labelled, for instance fluorescently labelled. The extension of the 3’ end may be in accordance with the method tREX (Cervettini, et al., 2020). The nucleotide or at least one of the nucleotides with which the 3’ end of the tRNA is extended comprises at least one label. This enables the detection of tRNAs that have formerly been acylated. Prior art methods detect acylated tRNAs by the use of a labelled substrate (Saito et al., 2001), which has disadvantages. For instance, the synthetase would need to be compatible with the labelled substrate and it would not be possible to optimise specifically for the unlabelled substrate. Other prior art methods detect the tRNA based on the size of the tRNA (Cervettini, et al., 2020), rather than by direct detection of a label attached to at least one nucleotide. Such methods are less sensitive. The label may be directly detectable, for instance by optical methods. The label may be a fluorescent label. The label may be a first label comprising a ligand or receptor of a ligand-receptor pair. For instance, the first label may be bound by a second label comprising an optically detectable moiety such as a fluorescent moiety. The first label may be bound by a second label that is immobilised, allowing the capture of the tRNA. The first label may be bound by a second label that is attached to a physically detectable moiety, such as a magnetic bead allowing pull-down of the tRNA. An example of a suitable ligand-receptor is biotin-streptavidin, but other ligand-receptor pairs would be suitable for use with the invention. Also encompassed are interactions with an intermediate molecule or molecules between the first label and the second label, such as an adapter molecule or adapter molecules. In an embodiment, at least one nucleotide comprises a fluorescent label or comprises a first label that can be bound by a second label comprising a fluorescent moiety. In another embodiment, at least one nucleotide comprises a physically detectable moiety or comprises a first label that can be bound by a second label comprising a physically detectable moiety. Thus, the method may comprise a step, after step iii), of contacting the tRNA with a second label capable of binding to the first label. The method of the fourth aspect may comprise the capture of labelled tRNAs via the first label or second label bound to the first label. In embodiments where the tRNA is a split tRNA, the capture of the tRNA may only capture one portion of the tRNA and does not need to capture all portions. Examples of suitable labels for capture are discussed herein, and include biotin, magnetic moieties, or other ligand-receptor pairs where one moiety is immobilised. Labelled tRNAs As discussed in relation to the fourth aspect, the inventors provide labelled tRNAs. Thus, in a fifth aspect of the invention, there is provided a tRNA comprising additional nucleotides at the 3’ end, wherein at least one nucleotide comprises a label. The tRNA may be any as discussed in relation to the fourth aspect. The labels may be any as discussed in relation to the fourth aspect. Nucleic acid constructs In a sixth aspect of the invention, there is provided a nucleic acid encoding any split tRNA of the second aspect or RNA molecule of the third aspect. The nucleic acid may be a DNA construct. The nucleic acid may be part of a vector. Hence, there is provided a vector comprising a nucleic acid of the sixth aspect. The vector may be present within a cell, such as a prokaryotic, bacterial, or E. coli cell. The inventors demonstrate herein that split tRNAs can be produced from a single gene, in which the two halves of the tRNA are circularly permuted and linked by an intervening sequence. The primary transcript of this gene is processed in cells to yield a functional split tRNA, which is acylated by an acyl-tRNA synthetase. Thus, in a particular embodiment, there is provided a nucleic acid comprising, from 5’ to 3’: a second stem- region-encoding sequence, a second tRNA-portion-encoding sequence, a sequence encoding a loop, a first tRNA- portion-encoding sequence, and a first stem-region-encoding sequence, wherein the first stem-region-encoding sequence and second stem-region-encoding sequence encode complementary stem region sequences. In a particular embodiment, there is provided a nucleic acid comprising, from 5’ to 3’: i) a sequence encoding a polypeptide-of-interest or nucleic-acid-of-interest, optionally ii) a sequence encoding a linker, iii) a second stem- region-encoding sequence, iv) a second tRNA-portion-encoding sequence, v) a sequence encoding a loop, vi) a first tRNA-portion-encoding sequence, vii) a first stem-region-encoding sequence, wherein the first stem-region- encoding sequence and second stem-region-encoding sequence encode complementary stem region sequences. The loop is suitable for the generation of a split RNA from a circularly permutated transcript. The loop may be a loop that is removed, spliced, or cleaved during RNA processing. The resultant RNA may be split into two chains, one comprising features i), optionally ii), iii), and iv) and the other comprising features vi) and vii). The loop may be, or may be derived from, an intergenic region from a naturally occurring tRNA operon. The tRNA operon may be derived from a tRNA operon of a cell in which a method of the present disclosure is performed. The tRNA operon may be an E. coli tRNA operon. The intergenic region may be identified by the tRNA operon generator disclosed in Dunkelmann et. al. (“A 68-codon genetic code to incorporate four distinct non-canonical amino acids enabled by automated orthogonal mRNA design. Nature Chemistry, AOP, doi: 10.1038 / s41557-021-00764-5 (2021); herein incorporated by reference). The loop may be, or may be derived from, the intergenic region in the E.coli tRNA operon between glyW-cysT, argY-argZ, or leuP-leuV. Examples of suitable loop regions include: TCGTCCT TAAATCGTGCATGGATTCACACAATTATAAA (SEQ ID NO: 15) AATTTTGCACCCAGCAAACTTGGTACGTAAACGCATCGT (SEQ ID NO: 16) TCTCTTACTTGATATGGCTTTAGTAGCGGTATCAATATCAGCAGTAAAATAAATTTCCCGAT (SEQ ID NO: 17) AGATTGTTTCTTCG (SEQ ID NO: 18) AACGAGGCGATATCAAAAAAAGTAAGATGACTGT (SEQ ID NO: 19) GTTTAAAAGACATCGGCGTCAAGCGGATGTCTGGCTGAAAGGCCTGAAGAATTT (SEQ ID NO: 20) AATTAGACAGCTATACAATC (SEQ ID NO: 21) TAATTCACCACAAAAACGCAGTGCTGCCGCTAAGT (SEQ ID NO: 22) In preferred embodiments, the loop region is SEQ ID NO: 17, SEQ ID NO: 19, or SEQ ID NO: 21. The encoded polypeptide-of-interest or nucleic-acid-of-interest may be any as disclosed herein. In a particular example, the nucleic acid encodes an acyl-tRNA synthetase. The encoded linker may be any as disclosed herein. The encoded first stem region and second stem region may be any as disclosed herein. The encoded first and second tRNA portions may be any disclosed herein. For instance, the second tRNA portion may be a 3’ portion of a tRNA split at the anticodon and the first tRNA portion may be a 5’ portion of the tRNA split at the anticodon. The tRNA portions may be from a tRNAPyl. In examples, the tRNAPylis an mmtRNAPyl, 1R26 tRNAPyl, deltaClos tRNAPyl, or Nitra tRNAPyl. The tRNA portions may be from a tRNALeu, tRNAAla, or tRNASer. Illustrative examples are SEQ ID NO: 1 and SEQ ID NO: 2, respectively. Other examples are SEQ ID NO: 23 & SEQ ID NO: 24, SEQ ID NO: 25 & SEQ ID NO: 26, SEQ ID NO: 27 & SEQ ID NO: 28, and SEQ ID NO: 46 & SEQ ID NO: 47. In other examples, the tRNA portions are from a tRNA not split at the anticodon. Such embodiments are fully explained elsewhere (e.g. see the first aspect). Illustrative examples are SEQ ID NO: 48 & SEQ ID NO: 49, SEQ ID NO: 50 & SEQ ID NO: 51, SEQ ID NO: 52 & SEQ ID NO: 53, SEQ ID NO: 54 & SEQ ID NO: 55, SEQ ID NO: 56 & SEQ ID NO: 57, SEQ ID NO: 58 & SEQ ID NO: 59, SEQ ID NO: 60 & SEQ ID NO: 61, and SEQ ID NO: 62 & SEQ ID NO: 63. The construct may comprise a 5’ untranslated region (UTR). In a preferred embodiment, the 5’ UTR is CTCTGTCTGCCCCCCTACCGAAG (SEQ ID NO: 10). Other illustrative examples include: TTGTCCACCCACTCAAGGC (SEQ ID NO: 33) or GTCAGAGTATAACAATCATACCCCGAG (SEQ ID NO: 34). Another 5’ UTR is CCTCTAGAAGGAGATGGAAAA (SEQ ID NO: 35). The 5’ UTR is positioned 5’ to the sequence encoding a polypeptide-of-interest or nucleic-acid-of-interest. The nucleic acids of the sixth aspect are applicable to any other aspect of the present disclosure featuring split tRNAs. As such, the nucleic acids of the sixth aspect may be used to encode the split tRNAs for any of the other aspects. Exemplary constructs are provided below. The italicised portions are the 5’ UTR, the regions of “N”s encode the polypeptide-of-interest or nucleic-acid-of-interest (which may be of any length), the bold and italicised portions encode the optional linker, the underlined portions encode the stem regions, the plain text portions encode the tRNA portions, and the bold portions encode the loop. CTCTGTCTGCCCCCCTACCGAAGNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNGCTTAATTAGCTGACCTACTAGTCGGCCGGCGGATGAGAGAAGATTTTCAGCC TGATACTACGTCGCTATCCGTTCAGCCGGGTTAGATTCCCGGGGTTTCCGCCAAACGAGGCGATATCAAAAAAAGT AAGATGACTGTGGAAACCTGATCATGTAGATCGAATGGATAGCGACGTA (SEQ ID NO: 29 – not included in listing due to undefined gap (ST.26 paragraph 37)) CTCTGTCTGCCCCCCTACCGAAGNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNGCTTAATTAGCTGACCTACTAGTCGGCCGGCGGATGAGAGAAGATTTTCAGCCTGATACTA CGTCGCTAACCTGTAAGCGGGGTTCGACCCCCCGGCCTTTCGCCAAATTAGACAGCTATACAATCGGAGGGCGCTC CGGCGAGCAAACGGGTTAGCGACGTA (SEQ ID NO: 30– not included in listing due to undefined gap (ST.26 paragraph 37)) CTCTGTCTGCCCCCCTACCGAAGNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNGCTTAATTAGCTGACCTACTAGTCGGCCGGCGGATGAGAGAAGATTTTCAGC CTGATACTACGTCGCTATCCGCGAACAACGGGTGAAACTCCCGTACACCTCGCCATAATTCACCACAAAAACGCAG TGCTGCCGCTAAGTGGGGTGTAGATCGGATTGATCGCGTGGATAGCGACGTA (SEQ ID NO: 31– not included in listing due to undefined gap (ST.26 paragraph 37)) CTCTGTCTGCCCCCCTACCGAAGNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNGCTTAATTAGCTGACCTACTAGTCGGCCGGCGGATGAGAGAAG ATTTTCAGCCTGATACTACGTCGCTAGCCACGGTTAGCCGGGTTCAACTCCCGGGTTCATCGCCATCTCTTACTTG ATATGGCTTTAGTAGCGGTATCAATATCAGCAGTAAAATAAATTTCCCGATGGTGAACTGGTCCGGGACCACCAGG CTAGCGACGTA (SEQ ID NO: 32– not included in listing due to undefined gap (ST.26 paragraph 37)) CCTCTAGAAGGAGATGGAAAA_polypeptide-of-interest / nucleic-acid-of- interest_GCTTAATTAGCTGACCTACTAGTCGGCCGGCGGATGAGAGAAGATTTTCAGCCTGATACTACGTCGCTAA CCTGTAAGCGGGGTTCGACCCCCCGGCCTTTCGCCAAATTAGACAGCTATACAATCGGAGGGCGCTCCGGCGAGCA AACGGGTTAGCGACGTA (SEQ ID NO: 41– not included in listing due to undefined gap (ST.26 paragraph 37)) CCTCTAGAAGGAGATGGAAAA_polypeptide-of-interest / nucleic-acid-of- interest_GCTTAATTAGCTGACCTACTAGTCGGCCGGCGGATGAGAGAAGATTTTCAGCCTGATACTACGTCGCTAT CCGCGAACAACGGGTGAAACTCCCGTACACCTCGCCATCTCTTACTTGATATGGCTTTAGTAGCGGTATCAATATC AGCAGTAAAATAAATTTCCCGATGGGGTGTAGATCGGATTGATCGCGTGGATAGCGACGTA (SEQ ID NO: 101– not included in listing due to undefined gap (ST.26 paragraph 37)) CCTCTAGAAGGAGATGGAAAA_polypeptide-of-interest / nucleic-acid-of- interest_GCTTAATTAGCTGACCTACTAGTCGGCCGGCGGATGAGAGAAGATTTTCAGCCTGATACTACGTCGCTAG CCACGGTTAGCCGGGTTCAACTCCCGGGTTCATCGCCATCTCTTACTTGATATGGCTTTAGTAGCGGTATCAATAT CAGCAGTAAAATAAATTTCCCGATGGTGAACTGGTCCGGGACCACCAGGCTAGCGACGTA(SEQ ID NO: 102– not included in listing due to undefined gap (ST.26 paragraph 37)) The construct may be: a 5’UTR, any sequence encoding a polypeptide-of-interest or nucleic-acid-of-interest, optionally a linker as disclosed herein, a stem region as disclosed herein or shown above, a tRNA portion as disclosed herein or shown above, a loop as disclosed herein or shown above, a tRNA portion as disclosed herein or shown above, and a stem region as disclosed herein or shown above. Methods of making polypeptides or nucleic acids In a seventh aspect of the invention, there is provided a method of making a polypeptide or nucleic acid, wherein the method comprises: providing a sequence of a polypeptide-of-interest or nucleic-acid-of-interest identified by any screening method of the first aspect or a labelling method of the fourth aspect and producing a polypeptide or nucleic acid according to said sequence. In an embodiment, the method may comprise performing a method of the first aspect or fourth aspect of the invention and then producing an identified polypeptide or nucleic acid. As such, in an embodiment, the method is a method of making a polypeptide or nucleic acid, wherein the method comprises: i) providing a library comprising a plurality of sequences encoding polypeptides-of-interest or nucleic- acids-of-interest, wherein each sequence within the library is linked to a portion of a tRNA; ii) incubating, under conditions conducive to acylation, each polypeptide-of-interest or nucleic-acid-of- interest with a tRNA comprising the tRNA portion linked to the sequence encoding the respective polypeptide-of- interest or nucleic-acid-of-interest, and wherein the incubation includes a substrate with which the tRNAs could be acylated; iii) exposing the tRNAs to conditions capable of labelling tRNAs that have been acylated; iv) identifying whether each polypeptide-of-interest or nucleic-acid-of-interest is associated with a labelled tRNA; and v) making a polypeptide or nucleic acid according to the sequence of an identified polypeptide-of-interest or nucleic-acid-of-interest. Any of the features of the first, second, third, fourth, fifth, or sixth aspects of the invention are applicable to the seventh aspect. The tRNA may be any as disclosed herein. For instance, the tRNA may be a tRNA split at the anticodon and comprising a stem region that is an extension of the anticodon stem loop. The tRNA may be split outside of the anticodon. The sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest may be attached to the stem region optionally via a linker. The stem region and linker may be any as disclosed herein. The constructs encoding the split tRNAs may be according to the sixth aspect. Step iii) may comprise any method or technique of labelling tRNAs that have been acylated as disclosed herein. Step iv) may comprise a quantification of the number of tRNAs associated with a particular sequence-of-interest that are labelled. Hence, step iv) may determine the efficiency of acylation or acylation status. Step iv) may comprise any methods of identifying a sequence-of-interest as disclosed herein. For instance, sequence information may be obtained, completely or in part, for a polypeptide-of-interest or nucleic-acid-of-interest associated with a labelled tRNA. In some examples, tRNA portions linked to sequences encoding polypeptides-of- interest or nucleic-acids-of-interest may be captured if they are associated with labelled tRNAs, and sequence information may be obtained. Step v) may comprise making a polypeptide or nucleic acid that can charge a tRNA with a particular substrate. Step v) may comprise making a polypeptide or nucleic acid that can charge a tRNA with a particular substrate at the highest efficiency, specificity, or combination of efficiency and specificity within the library. The method may be performed multiple times, and step v) may comprise making a polypeptide or nucleic acid that can charge a tRNA with a particular substrate but cannot charge other tRNAs or charge with other substrates, and so can be used in orthogonal systems. Step v) may comprise making a polypeptide or nucleic acid that can change the chemical nature of a substrate in a desired manner. For instance, the polypeptide or nucleic acid may be identified by embodiments where the charged tRNA is exposed to the polypeptide / nucleic acid and it is determined whether the polypeptide / nucleic acid can alter the substrate in a manner that affects the susceptibility of the substrate- tRNA to deacylation. The polypeptide-of-interest or nucleic-acid-of-interest may be any as disclosed herein. In a particular embodiment, the polypeptide-of-interest is an acyl-tRNA synthetase. The library comprising a plurality of sequences encoding polypeptides-of-interest may be a library of variant acyl-tRNA synthetases. For instance, acyl-tRNA synthetases comprising random mutations. The acyl-tRNA synthetases may be pyrrolysyl-tRNA synthetases. The acyl-tRNA synthetases may be variants of or based on pyrrolysyl-tRNA synthetases. As such, the library may be a library of pyrrolysyl-tRNA synthetase variants. Alternatively, the acyl-tRNA synthetases may be suitable for use with tRNALeu, tRNAAla, or tRNASer, or modified versions thereof. In yet further examples, the acyl-tRNA synthetases are suitable for use with tRNATrpor tRNAPro. Thus, in an embodiment, there is provided a method of making an acyl-tRNA synthetase, wherein the method comprises: i) providing a library comprising a plurality of sequences encoding variant acyl-tRNA synthetases, wherein each sequence within the library is linked to a portion of a tRNA; ii) incubating, under conditions conducive to acylation, each acyl-tRNA synthetase with a tRNA comprising the tRNA portion linked to the sequence encoding the respective acyl-tRNA synthetase, and wherein the incubation includes a substrate with which the tRNAs could be acylated; iii) exposing the tRNAs to conditions capable of labelling tRNAs that have been acylated; iv) identifying whether each acyl-tRNA synthetase is associated with a labelled tRNA; and v) making an acyl-tRNA synthetase according to the sequence of an identified acyl-tRNA synthetase. Beta-amino-acid-acyl-tRNA synthetases The inventors have made use of the methods and tools disclosed herein and have identified acyl-tRNA synthetases that are capable of acylating tRNAs with beta amino acids. This demonstrates that the methods disclosed herein are functional and can be used for the intended purpose. Further details of these experiments are provided in the documents to which priority is claimed (EP2306393.6, filed 28 April 2023 and EP2400299.0, filed 9 January 2024, each of which is incorporated herein by reference). α,α-disubstituted-amino-acid-acyl-tRNA synthetases The inventors have made use of the methods and tools disclosed herein and have identified acyl-tRNA synthetases that are capable of acylating tRNAs with α,α-disubstituted-amino acids. This demonstrates that the methods disclosed herein are functional and can be used for the intended purpose. Further details of these experiments are provided in the documents to which priority is claimed (EP2306393.6, filed 28 April 2023 and EP2400299.0, filed 9 January 2024, each of which is incorporated herein by reference). Beta-hydroxy-acid-acyl-tRNA synthetases The inventors have made use of the methods and tools disclosed herein and have identified acyl-tRNA synthetases that are capable of acylating tRNAs with beta-hydroxy acids. This demonstrates that the methods disclosed herein are functional and can be used for the intended purpose. Further details of these experiments are provided in the documents to which priority is claimed (EP2306393.6, filed 28 April 2023 and EP2400299.0, filed 9 January 2024, each of which is incorporated herein by reference). All of the features described herein (including any accompanying claims, abstract and drawings), and / or all of the steps of any method or process so disclosed, may be combined with any of the above aspects in any combination, except combinations where at least some of such features and / or steps are mutually exclusive. For a better understanding of the invention, and to show how embodiments of the same may be carried into effect, reference will now be made to the Examples, which are not intended to limit the invention in any way. EXAMPLES Summary of Examples The genetic code of living cells has been reprogrammed to enable the site-specific incorporation of hundreds of non-canonical amino acids (ncAAs) into proteins, and the encoded synthesis of non-canonical polymers and macrocyclic peptides and depsipeptides. Current methods for engineering orthogonal (O)-aminoacyl-tRNA synthetases to acylate new monomers rely on translational readouts and therefore require the monomers to be ribosomal substrates. O-synthetases cannot be evolved to acylate O-tRNAs with non-canonical monomers (ncMs) that are poor ribosomal substrates (and ribosomes cannot be evolved to polymerize ncMs that cannot be acylated onto O-tRNAs); this co-dependence creates an evolutionary deadlock that has restricted the scope of translation in living cells to alpha-L-amino acids and closely related hydroxy acids. Here, we break this deadlock by developing a direct selection for O-synthetases that acylate cognate O-tRNAs with ncMs, independent of whether the ncMs are ribosomal substrates. We develop split tmRNAs, composed of a non-covalent assembly between the 5’ half of an O-tRNA and a fusion between the 3’ half of the O-tRNA and the mRNA that encodes the cognate synthetase; this couples the synthetase genotype to split tmRNA acylation in cells. We also develop an approach to specifically isolate and enrich acylated split tmRNAs. We combine these advances into tRNA display, which enables the direct, rapid and scalable selection for O-synthetases that acylate their cognate O-tRNA; this approach uses 50-times less ncM than translation-based selections and we parallelize tRNA display to select efficient synthetases for 8 ncAAs. Using tRNA display, we directly select O-synthetases that acylate their cognate O-tRNA with a β-amino acid. We build on this advance to demonstrate the genetically encoded, site-specific cellular incorporation of a β-amino acid into a protein, and thereby expand the chemical scope of the E. coli genetic code. Introduction Here we develop derivatives of tREX in which specific acylated tRNAs, isolated from cells, are labelled with dNTP analogs by primer extension; this enables fluorescent imaging (fluoro-tREX) or capture (bio-tREX) of acylated tRNAs. We create Methanosarcina mazei pyrrolysine tRNA Pyl CUA (MmtRNA CUA, henceforth referred to as tRNAPylCUA) genes that are circularly permuted (which we refer to as cis split (s)tRNAPylgenes) by joining the 5’ and 3’ ends through an intervening sequence and creating new 5’ and 3’ ends at the anticodon. The intervening sequence is processed out of the transcript to generate a split stRNA, composed of a 5’ half and a 3’ half, which is acylated by Mmpyrrolysyl-tRNA synthetase (henceforth referred to as PylRS), the cognate aminoacyl-tRNA synthetase of the parent tRNA. We connect the genotype responsible for acylation to the acylation itself, by fusing the gene for PylRS to the cis stRNAPylgene, creating cis stmRNAPyl. We demonstrate that we can selectively enrich – by more than 300-fold – stmRNAs, encoding active PylRS variants, with respect to attenuated activity variants, using bio-mREX (a variation of bio-tREX applied to the stmRNA). We increase the dynamic range of PylRS enrichment by maximizing the transcription of stmRNAs, while minimizing the translation of the PylRS mRNA encoded within the stmRNA. We generate stmRNA libraries with combinations of mutations in the PylRS gene and use parallel bio-mREX based enrichments, in the presence and absence of ncAAs, followed by reverse transcription and NGS to rapidly and scalably define active and selective PylRS variants, through a process we term tRNA display. We use tRNA display to select orthogonal aminoacyl-tRNA synthetases (aaRSs) that specifically acylate their cognate, orthogonal tRNAs with a carboxylic acid, that cannot function in translation, and with a β-amino acid, a class of ncMs generally considered to be poor ribosomal substrates10,11,14-18. Moreover, using the β-amino acid pair we demonstrate the site-specific co-translational incorporation of a β-amino acid in a recombinant protein. Example 1 - Sensitive detection & efficient isolation of acylated tRNAs We demonstrated that we could determine the aminoacylation status of a specific tRNA isolated from cells by periodate oxidation followed by selective, primer mediated, extension of non-oxidized tRNAs with nucleotide derivatives bearing a fluorophore (Cy5) or biotin (Fig.2a). We isolated tRNAs from cells expressing tRNAPylCUA in presence and absence of PylRS and (Nε-((tert- butoxy)carbonyl)-L-lysine (BocK, 1), a known and efficient substrate of PylRS. We oxidized the isolated tRNAs with sodium periodate and annealed a DNA probe, containing a 3’ Cy3 and a 5’poly G stretch, to the 3’ end of tRNAPyl. We extended the free 3’ end of tRNAPylCUA, resulting from aminoacylation-mediated protection from periodate oxidation, using Klenow exo (-) and a dNTP mix in which dCTP was replaced with Cy5 labelled dCTP. We visualized the fluorescent signals following gel electrophoresis. We detected a strong Cy5 labelled band, corresponding to extended tRNAPylCUA, from cells expressing PylRS and provided with BocK (1); in contrast, tRNAs from cells that were not provided with PylRS or BocK (1) yielded minimal Cy5 signals (Fig.2b). The Cy3 signal resulting from the tRNA-DNA probe hybrid provided a measure of tRNA abundance (Fig.7). Note that under our electrophoresis conditions we do not resolve the non-extended primer from the labelled extension product, this was necessary for tREX (where the extension product and non-extended product were indistinguishable by fluorescence) but is not necessary for our new approach (where the extension reaction generates new fluorescence). These experiments demonstrate that our approach, which we named fluorescent tREX (fluoro-tREX), allows the acylation of a tRNA to be followed through the generation of a fluorescent signal. To assess the dynamic range of fluoro-tREX, we used four PylRS variants which direct the, tRNAPylCUA-mediated, incorporation of Nε-(carbobenzyloxy)-L-lysine (CbzK, 2) in response to the amber codon in GFP(150TAG)His6; these PylRS variants led to GFP(150CbzK)His6fluorescence levels spanning 200-fold, reflecting their distinct activities. The Cy5 signal for these PylRS variants in fluoro-tREX corresponded to expression levels of GFP(150CbzK)His6 (Fig.8). We conclude that fluoro-tREX has a wide dynamic range and can detect the activity of synthetase variants that give rise to low protein yields, and which may have low activities. As we sought to use tRNA extension-based methods to detect tRNAs that were acylated with monomers beyond α-L- amino acids we wanted to ensure that, following periodate oxidation, a range of monomers could be cleaved from the tRNA, as required for the 3’ extension step that generates the signal in tREX-based approaches. The cleavage of monomers from the tRNA is an ester hydrolysis, for which the rate of cleavage is expected to be inversely related to the pKa of the monomer that acylates the tRNA20. We confirmed, by northern blot, that the α- L- amino acid BocK (1), its hydroxy acid analog, and its desamino carboxylic acid analog can be attached to tRNAPylCUAby PylRS (Fig.9)21. The pKaof a carboxylic acid in an α-L-amino acid is about 2.3, the pKaof a hydroxy acid is about 3.7, and the pKa of a desamino carboxylic acid is about 4.6. While we could detect acylation by the α-amino acid with our initial fluoro-tREX protocol, we could not observe the acylation of tRNAs by the hydroxy acid or simple carboxylic acid by fluoro-tREX. By treating tRNAs with base, following oxidation, we improved the detection of acylation with α-L-amino acids, hydroxy acids and carboxylic acids. As the pKaof β- amino acids lies between the pKaof hydroxy acids and carboxylic acids we expected our revised protocol to also detect β-amino acids attached to tRNAs. Next, we replaced Cy5-dCTP with biotinylated dCTP (bio-dCTP) in the extension step of fluoro-tREX, thereby creating biotin-tREX (bio-tREX). We selectively captured the biotinylated extension product (resulting from tRNA molecules that were protected from periodate oxidation by their aminoacylation) on magnetic streptavidin beads, washed the beads, and eluted bound tRNA extension products. The presence of the tRNAPylCUA extension product in the eluate was dependent on the presence of PylRS, BocK (1) in cells, and the addition of bio-dCTP to the extension reaction (Fig.2c). We conclude that biotin-tREX enabled the selective capture of tRNA extension products from tRNAs that were aminoacylated. Example 2 - Split tRNAs: in vivo assembly, maturation, and acylation. Next, we asked whether we could split the tRNAPylCUA gene at the anticodon to create a split tRNAPyl(stRNAPyl). We envisioned generating a system in which the 5’ and 3’ halves of a split tRNA gene were transcribed, assembled in vivo via non-covalent interactions (including base-pairing), matured by the cellular tRNA processing machinery, and recognized and efficiently acylated by PylRS (Fig.3a,b). We first designed a series of constructs in which we split the tRNA gene into two at the anticodon. We replaced the sequence of the anticodon stem loop in each half of the tRNAPylCUAgene with an extension of eight to fourteen nucleotides in length; the extensions within each pair of tRNA halves were designed to base pair with each other and form a stem that may stabilize the split (s)tRNAPyl. We hypothesized that a minimal length of the stem would aid efficient stRNA assembly and that stems that were too long might lead to degradation of the stRNAPylby cellular RNAses that target double stranded RNA. We expressed each pair of tRNA halves in trans (from two different plasmids) in the presence or absence of PylRS and BocK (1), and isolated stRNAPyls. We analyzed the acylation of stRNAPyls by fluoro-tREX, as a measure of stRNAPylassembly and function. For the stRNAPylwith a stem length of eight base pairs we observed attenuated acylation, and this construct may not stably assemble in cells. For stRNAPyls with stems above twelve base pairs we observed gel bands consistent with degradation products (Fig.10). Stems of either ten or twelve base pairs resulted in robust aminoacylation, which was dependent on the presence of both tRNA halves, PylRS, and BocK (1) (Fig.3c). We concluded that a ten base pair stem is sufficient to facilitate the association of the two tRNA halves, minimize degradation, and enable robust aminoacylation of the assembled tRNA. We therefore performed all subsequent experiments with stRNAPyls bearing a ten base pair stem. To the best of our knowledge this data constitutes the first example of a split tRNA being assembled, matured and acylated in cells. Next, we turned our attention to designing an optimal expression system for stRNAPyls, which would: (1) ensure equimolar stoichiometry of both tRNA halves, and (2) facilitate spatial proximity, and thereby the assembly, of both tRNA halves into a tRNA body that can be acylated. Certain organisms, such as the archaea Cyanidioschyzon merolae, possess tRNAs that are expressed as circular permutants22, where the tRNA genes are split at either the anticodon stem loop-, T-, or D-loop, and joined together at the acceptor stem through intervening sequence; in these organisms, posttranscriptional maturation through splicing and processing leads to a topologically regular tRNA. Inspired by this natural expression strategy, we hypothesized that we could produce the two tRNA halves in cis from one transcript by inserting an intervening sequence between them, thereby creating a circular permutation of the parent tRNA that may be processed to yield the stRNA (Fig.3 b). We noted that the intergenic regions of polycistronic tRNA operons in Escherichia coli (E. coli) connect the 3’ half of a tRNA to the 5’ half of the following tRNA23; therefore these intergenic regions have a similar topology to the intervening sequences of C. merolae. Using a tRNA operon generator24, we selected three E. coli intergenic regions (glyW-cysT, leuP-leuV, and argY-argZ), which are found between tRNAs with high sequence similarity to tRNAPylCUA. We used these intergenic regions, and two intervening sequences from circularly permuted archaeal tRNAs22, as the intervening sequences in tests of our circular permutation strategy. We created cis stRNAPylgenes in which the 3’ half of the stRNAPylgene was connected through one of five ‘loop’ sequences (two archaeal intervening sequences, three E. coli intergenic regions) to the 5’half of the split tRNA gene. We grew cells harboring the cis stRNAPylconstructs, with PylRS in the presence and absence of BocK (1). Isolation of tRNAs followed by fluoro-tREX revealed efficient expression of stRNAs from cis stRNAPylgenes (Cy3 signal), and substantial, BocK-dependent, acylation of the stRNAPyls produced from the cis stRNAPylgenes (Cy5 signal) (Fig.3d). cis stRNAPylgenes with the intergenic regions from E. coli produced stRNAPylat levels comparable to tRNAPylCUA, and stRNAPylwas acylated at comparable levels to tRNAPylCUA(Fig.12A). We proceeded to use the E. coli intergenic region leuP-leuV for all further experiments. We conclude that tRNAPylCUA can be split, expressed from a single transcript in cis, and that the split tRNA functions as an efficient substrate for acylation by PylRS. Example 3 - Covalently linking acylation phenotype and genotype Next, we fused the PylRS coding sequence and a linker sequence to the 5’ end of the 3’ half of the cis stRNAPylexpression cassette, creating the stmRNAPylcassette (Fig.4a). To maximize expression, we put the entire stmRNAPylcassette under the control of an inducible T7 promoter and terminator. We hypothesized that transcription, processing and maturation of this construct would lead to a split tRNA in which the mRNA encoding the synthetase was covalently linked to the 5’ end of the 3’ half of the tRNA (the 3’ end of the 3’ half is acylated by the synthetase). Translation of the synthetase mRNA within the stmRNAPylwould generate the synthetase protein which, in the presence of its cognate ncM, would acylate stmRNAPyl. This would generate a covalent link between the monomer attached to the tRNA and the mRNA of the synthetase gene that catalyzed the attachment. To test our hypothesis, we grew cells harboring wt stmRNA in the presence and absence of BocK (1), isolated total RNA, and performed fluoro-mREX, a modified version of fluoro-tREX which was optimized for larger RNAs (by using an RNA isolation method suited for mRNAs, as well as agarose gels for the analysis). Remarkably, for the wt stmRNA we observed a BocK (1) dependent fluorescent (Cy5) band, in fluoro-mREX, with an electrophoretic mobility consistent with the length of the stmRNA construct. This is consistent with the encoded synthetase aminoacylating the stmRNA with its substrate. To further confirm that the fluoro-mREX signal results from the activity of the synthetase encoded in the stmRNA construct, we generated a new construct (stmRNAat) containing a PylRS variant with severely attenuated (at.) activity, as judged by GFP expression assays. We did not observe a fluorescent signal for stmRNAatin fluoro- mREX (Fig.4b); this provides additional evidence that the fluoro-mREX signal we observed for the wt stmRNA results from aminoacylation of the split tRNA by the synthetase encoded on the same construct. We conclude that our stmRNA construct is functional. The construct is transcribed and processed to generate a split tRNA in which the 3’ half is fused to the synthetase mRNA. The synthetase gene is translated, and the resulting protein catalyzes the acylation of the 3’ end of the 3’ half of the stmRNA. In fluoro-mREX the acylation is converted into a fluorescent ‘phenotype’ via the addition of fluorescent nucleotides to the RNA strand that contains the mRNA of the synthetase. This creates a physical linkage between the mRNA sequence of the synthetase, its genotype, and the fluorescent ‘phenotype’, generated as a result of the activity of the synthetase. Example 4 - Efficient acylation-specific enrichment of PylRS genotypes. Next, we aimed to selectively isolate acylated stmRNAs with respect to non-acylated stmRNAs and, reverse transcribe the isolated PylRS mRNA within stmRNAs to directly yield the cDNA of the PylRS gene responsible for acylation (Fig.4c). We envisioned that the resulting cDNA could then either be used as a direct readout in a quantitative (q)PCR, converted to DNA for further selection and directed evolution, or sequenced. To selectively isolate acylated stmRNAs with respect to non-acylated stmRNAs we created bio-mREX (Fig.4c), an adaptation of bio-tREX for stmRNAs. In this approach we used the same methods to isolate RNA that we used for fluoro-mREX. Following periodate oxidation of the isolated RNA, we add the DNA probe, Klenow exo (-) and a dNTP mix in which dCTP is replaced with bio-dCTP. The stmRNAs that had been aminoacylated are tagged with biotin and then captured on the streptavidin beads, and the beads are washed to remove non-specific binders. PylRS mRNAs are then directly reverse transcribed, from the stmRNAs captured on the beads, to create PylRS cDNA. The PylRS cDNA is released by RNase H treatment and heating, before it is quantified by a qPCR- based method. To test our approach, we grew E. coli cells harboring the wt stmRNA construct in the presence and absence of BocK (1) and performed bio-mREX. Strikingly, we recovered over a hundred-fold more cDNA molecules in the presence of BocK (1) (Fig.4d). To further demonstrate that bio-mREX selectively recovers active PylRS variants, we compared bio-mREX with stmRNAatand wt stmRNA from cells with and without BocK (1). We recovered 300-fold more DNA molecules from wt stmRNA with BocK (1), than from stmRNAat(Fig.4d, Fig.12B). Furthermore, in the presence of BocK (1) we recovered approximately 2.5 % of the wt stmRNA molecules that we put into the pulldown, consistent with a minimal loss of acylation signal. We conclude that bio-mREX enables the efficient and selective recovery of stmRNAs, and the genes for synthetases that acylate the split tRNAs within them. To test the dynamic range of bio-mREX, we created stmRNA constructs for a set of four PylRS (CbzK) variants (PylRS mutants that direct the incorporation of CbzK, 2). When paired with tRNAPylCUAand 2, these PylRS variants lead to GFP fluorescence from GFP(150TAG)His6, and the level of GFP fluorescence generated by the PylRS variants spans two orders of magnitude (Fig.4e). Assuming that the GFP signal primarily reflects the efficiency with which the PylRS(CbzK) variants aminoacylate their cognate tRNAPylCUA, we would ideally see a correlation between GFP fluorescence and the number of molecules recovered in bio-mREX (Fig.13). However, when assayed via bio-mREX, we observed a similarly high recovery of cDNA for the three most active PylRS variants; this suggested that above a threshold activity the first generation of bio-mREX could not effectively differentiate between different acylation activities (Fig.14, Fig.4e). The maximum potential dynamic range of bio-mREX will be defined by the number of stmRNAs in a cell, as this defines the number split tRNA substrates and therefore the number of acylation events that can occur before all stmRNAs in a cell are acylated. We expressed the stmRNAs from a strong T7 promoter to maximize the stmRNA transcription and therefore the potential dynamic range. However, in stmRNAs the PylRS mRNA is on the same transcript as the split tRNA substrate of the PylRS enzyme; therefore, increasing the transcription of the stmRNA is likely to increase the concentration of the PylRS enzyme. At high PylRS enzyme concentrations, intrinsically less active enzymes may acylate the split tRNA substrate to completion (through mass action) leading to a compression in the dynamic range of bio-mREX. These considerations suggested that high stmRNA levels and low PylRS enzyme levels would maximize the experimental dynamic range of bio-mREX. The production of PylRS is a function of the abundance of the stmRNA transcript and the efficiency with which the PylRS mRNA within this transcript is translated. Since translational efficiencies in E. coli can be controlled by ribosome binding site (RBS) efficiencies we set out to selectively tune the concentration of the PylRS enzyme through RBS engineering. We designed 5’UTR sequences with attenuated translation rates using a de novo DNA RBS calculator25and introduced them in to the mRNAs for the PylRS(CbzK) variants within stmRNAs. For all predicted RBS sequences, the correlation between GFP expression, resulting from amber suppression, and the number of molecules recovered by bio-mREX improved with respect to the original construct (Fig.14, Fig.4e)). RBS2 displayed a good correlation and we proceeded to use the stmRNA utilizing this RBS, termed stmRNAvol2, for all subsequent experiments. In conclusion, by combining the stmRNAvol2construct with bio-mREX we developed a pulldown with which we could selectively isolate cDNA of active PylRS variants over inactive variants over a large dynamic range. Example 5 - Direct selection for cellular acylation by tRNA display. Next, we set up the first method for directly selecting synthetase enzymes on the basis of their tRNA acylation activity. We envisaged that we could efficiently detect active and selective PylRS variants by running parallel bio- mREX-based selections coupled with deep sequencing by next generation sequencing (NGS) (Fig.5a). In such an experiment, PylRS active site libraries are cloned into stmRNAvol2constructs and transformed into E. coli cells. Overnight cultures are then diluted into media in the presence (positive samples) and absence (negative samples) of the monomer of choice and bio-mREX is performed. The experiments are performed in multiple replicates, and cDNA of the positive and negative samples, as well as the cDNA reverse transcribed from the input library, are subsequently barcoded and sequenced by NGS (Fig.5a). We termed this selection approach tRNA display. NGS data from the selections of the samples enables the calculation of two key parameters: (1) the enrichment – defined as the average abundance of a particular sequence in the positive samples, over the abundance of the same sequence in the input RNA and (2), the selectivity – defined as the ratio of the abundance of a sequence in the positive sample, over the abundance of the same sequence in the negative sample. Desired PylRS variants correspond to sequences that are highly enriched as well as highly selective. Plotting the natural logarithm of the enrichment against the natural logarithm of the selectivity results in a spindle-shaped distribution. We expect any desired sequences to be in the top right quadrant of the spindle plots (Fig.5a) (highly enriched, and highly selective). To test tRNA display we generated a small PylRS library, in which we expected many sequences to be active, as a stmRNAvol2construct. The library targeted positions Y306, L309, and N346 in PylRS, where mutations had previously been identified that enable the efficient incorporation of CbzK (2)26,27(Fig.5b and Fig.15). We transformed the library into E. coli cells and performed a single round of tRNA display. We observed a large population of highly enriched and selective variant PylRS sequences in the spindle plot derived from the sequencing of this experiment (Fig.5c). To assess the predictive power of tRNA display, we measured the in vivo production of GFP(150CbzK)His6 from GFP(150TAG) i Pyl His6 n the presence of tRNA CUA and 65 PylRS variants, which appeared to be potential hits on the basis of their position on the spindle plot (Fig.16). We observed a positive correlation between the enrichment derived from NGS data and the translation activity derived from GFP production for these hits (Fig.5d), and the vast majority of these hits were selective (Fig.16). We concluded that tRNA display permits the direct, translation independent, identification of active and selective PylRS enzymes, from a library of PylRS sequences. Example 6 - High throughput selection for ncAAs by tRNA display. To further validate the utility of tRNA display, we ran parallel selections (Fig.17) using six independent, highly diverse, PylRS active site libraries (Fig.15), and ten ncAAs, 2-11 (Fig.5b). After two rounds of selection, we analyzed the spindle plots derived from the NGS data and identified individual PylRS mutants for ncAAs 2,3,4,7,8,9,10,11 that were enriched and selective; the enriched and selective mutants for each of these ncAAs showed convergent sequence motifs (data are illustrated in the Supplementary Figs.12-22 of the present priority document EP2400299.0, filed 9 January 2024 – N.B. the following references to EP2400299.0 are to this document). We demonstrated the incorporation of ncAAs 3,4,7,8,9,10,11 in response to the amber codon in GFP(150TAG)His6, in cells containing tRNAPylCUAand the corresponding PylRS mutants identified by tRNA display. GFP production was ncAA dependent and ESI-MS confirmed the incorporation of each ncAA in GFP (Fig.5e-i - also illustrated in Supplementary Figs.13-22 of EP2400299.0). We note that our results include an aminoacyl-tRNA synthetase for 7, which enables the first incorporation of this thiophene containing ncAA into a protein. Out of the 30 characterized variants with a selectivity score greater than or equal to ten and an enrichment score greater than or equal to five, 27 were active with their cognate ncAA in protein expression (20 variants had activities at least 50% of the activity of PylRS with BocK (1)). All 27 variants selectively incorporated their ncAA substrate, as judged by mass spectrometry (Fig.5 e-i, - also illustrated in Supplementary Figs.13-22 of EP2400299.0). In summary, we demonstrated the parallel, scalable and rapid selection of PylRS variants, using diverse libraries and ncAAs, through tRNA display. We note that the amount of ncAA used in each tRNA display selection is 50- 100 times lower than used in current methods for synthetase selection. Example 7 - Discovery of a β-amino acid specific PylRS variant by tRNA display Next, we challenged tRNA display to discover synthetases for classes of ncMs which either cannot be translated or are known – from in vitro studies – to be poor ribosomal substrates (Fig.6)10,11,14-18. We focused on nine ncMs (12-20, Fig.6a), these monomers include β2and β3amino acids with variable stereochemistry, β-hydroxy acids, and carboxylic acids.6-((tert-butoxycarbonyl)amino)hexanoic acid (BocAhx, 15), is a known substrate for wt PylRS21, and provided a control for our ncM selections. To select PylRS variants for these monomers we ran parallel tRNA display selections (Fig.18) using a highly diverse library, which mutated residues 300, 302, 305, 306, 309, 344, 346, 348, and 401 in the PylRS active site (Fig.15), and each ncM. We analyzed the spindle plots derived from the NGS data for each ncM to identify PylRS variants which were enriched and selective (data are illustrated in the Supplementary Fig.24-32 of EP2400299.0). For 12 and 15 we identified enriched and selective sequences, and the most selective hits converged on a distinct sequence pattern for each of these ncM. We identified sequences similar to the wild-type PylRS sequence for ncM 15, 6-((tert- butoxycarbonyl)amino)hexanoic acid (BocAhx), indicating that the selection had converged on active synthetases for this monomer. The four most selective and active hits from the selection, PylRS(15_1) to PylRS(15_4), differ from the wild-type sequence by point mutations of V401; we showed, by fluoro-tREX, that all of these selected sequences were active and selective for 15 (Fig.6b – also illustrated in Supplementary Fig.27 of EP2400299.0). As 15 is expected to be more challenging to cleave from the tRNA than other monomers, including β-amino acids (Fig.9), this result provided confidence that tRNA display would enable the selection of synthetases for a wide range of other monomers, including β-amino acids. Remarkably, two PylRS variants, PylRS(12_1) and PylRS(12_2), identified by tRNA display selection, directed (S)-3-amino-3-(3-bromophenyl)propanoic acid (β3-mBrF, 12)-dependent acylation of tRNAPylCUA, as judged by fluoro-tREX (Fig.6c – also illustrated in Supplementary Fig.24 of EP2400299.0). To verify the identity of the monomer attached to the tRNA by PylRS(12_1) and PylRS(12_2) we captured the acylated tRNAPylCUAon streptavidin beads via a biotinylated probe for tRNAPylCUA, washed the beads, and eluted the ncM by heating under alkaline conditions; we then derivatized the free ncM and analyzed the sample by LC-MS (Fig.19). Using this approach, we confirmed that both PylRS variants charged tRNAPylCUAwith the ncM 12 (Fig.6e,f). To the best of our knowledge PylRS(12_1) / tRNAPyland Pyl Pyl CUA RS(12_2) / tRNA CUA are the first β-amino acid specific orthogonal aminoacyl-tRNA synthetase / tRNA pairs described to date. Next we increased the activity of PylRS(12_1) and PylRS(12_2), by random mutagenesis of the active site region of the PylRS gene within the stmRNA construct followed by tRNA display-based selection (Fig.20 and Supplementary Figs.34, 35 of EP2400299.0). From the resulting spindle plot, we identified sequences, carrying one to three additional mutations with respect to the parental clones, which were enriched and selective. The three most enriched and selective hits, PylRS(12_1evol1-3, were derivatives of PylRS(12_1). We confirmed the specificity of PylRS(12_1evol1-3) for acylating tRNAPylCUA with 12, by both fluoro-tREX and our LC-MS based assay (Fig.6d,f,g, Fig.19 – also indicated in Supplementary Fig.36 of EP2400299.0). PylRS(12_1evol1-3) were notably more active in acylating tRNAPylCUAwith 12 than PylRS(12_1) (Fig.6c,d,e,f). Next, we combined PylRS(12_1 or PylRS(12_1evol1) with tRNAPylCUAand GFP150TAGHis6in cells. While β-amino acids are expected to be poor ribosomal substrates10,11,14-18, we found that production of GFP was dependent upon addition of 12 to cells. GFP fluorescence was 3-fold higher with PylRS(12_1evol1) than with PylRS(12_1), consistent with the higher acylation activity of PylRS(12_1evol1). We confirmed the incorporation of 12 at position150 in GFPHis6 purified from cells with PylRS(12_1), or PylRS(12_1evol1), by ESI-MS and MS / MS. In the absence of 12 we observe some GFP production resulting from the incorporation of Phe. However, in the presence of 12 (Fig.6g,h – also illustrated in Supplementary Fig.37 of EP2400299.0), we produce more GFP, and we only detect incorporation of 12 by intact MS and MS / MS. We conclude that in the presence of 12 the background incorporation of Phe observed in the absence of 12 is effectively outcompeted. Similar observations have previously been made for efficient and selective ncAA incorporation systems28and the fidelity of the natural code is also known to rely on competition29. We note that we did not observe incorporation of 12 at position 3 of GFP in experiments using GFP3TAGHis6(also illustrated in Supplementary Fig.37 of EP2400299.0), consistent with β-amino acids being poor ribosomal substrates10,11,14-18and the fact that β-amino acids may not be tolerated at all positions in a protein. Next, we performed two rounds of selection with lib14 and ncMs A1, A2, A3, A4, A5 (β-amino acids with varied side chains), A6 (an α,α-disubstituted-amino acid) and A7 (a β-hydroxy acid) (illustrated in Fig.45a, Fig.47 of EP2400299.0). From the resulting spindle plots (illustrated in Fig.48-54 of EP2400299.0), we identified enriched and selective sequences, and the most selective hits converged on a distinct sequence pattern for each ncM. We note that the synthetases selected for all six β-amino acids differ in sequence, but contain common mutations M300D and A302H (see Figs.29, 48-52 of EP2400299.0). The sequence pattern observed for the β-hydroxy acid A7 is similar to the one observed for β-amino acids. However, the residue at position 300 – which may be in direct proximity to the amine- / hydroxy- group – is changed from aspartic acid to asparagine (see Fig.54 of EP2400299.0). The PylRS variants, identified by tRNA display selection, directed the specific acylation of tRNAPylCUAby their cognate monomer, as judged by fluoro-tREX and our LC-MS based assay (see Fig.45d-q of EP2400299.0). We quantified the fraction of acylation as a function of ncM concentration see (see Fig.55 of EP2400299.0). To the best of our knowledge, we have discovered the first specific aminoacyl-tRNA synthetase / tRNA pairs for three distinct classes of ncM: β-amino acids, α,α-disubstituted-amino acids and, β- hydroxy acids. Example 8 - Enoding ncMs in proteins To investigate the incorporation of the β-amino acids, α,α-disubstituted-amino acids, and β-hydroxy acids into proteins we combined the orthogonal synthetase / orthogonal tRNA pairs we have discovered for eight ncMs, 12, A1, A2, A3, A4, A5, A6, A7, with GFP150TAGHis6 in cells, and added the cognate ncM. We observed ncM dependent GFP production for 12, A2, A5 and A6, with isolated yields ranging from 3 to 35 mg per litre of culture, and mass spectrometry confirmed the incorporation of these β-amino acids and α,α-disubstituted-amino acids in GFP (Fig.6h, - also illustrated in Figs.42, 45r, 45s, 56, and 57 of EP2400299.0). In the absence of 12 we observe some GFP production resulting from the incorporation of natural amino acids. However, in the presence of 12, we produce more GFP, and we only detected incorporation of 12 by intact MS and MS / MS (Fig. 6h – also illustrated in Figs.42, 56 of EP2400299.0). We conclude that in the presence of 12 the background incorporation, observed in the absence of 12, was effectively outcompeted; we made similar observations for A2, A5 and A6 (Fig.6h – also illustrated in Figs.45s, 42, 56 of EP2400299.0). Similar observations have previously been made for efficient and selective ncAA incorporation systems and the fidelity of the natural code is also known to rely on competition. We conclude that 12, A2, A5 and A6 are site-specifically incorporated with high- fidelity. We did not observe incorporation of ncMs 12, A2, A5 or A6 at position 3 of GFP (from GFP3TAGHis6) (illustrated in Fig.57 of EP2400299.0), indicating that these ncMs are not tolerated at all positions in a protein; similar site-dependent incorporation efficiency has previously been observed for ncAAs. We note that we did not observe ncM-dependent increase in production of GFP from GFP150TAGHis6or GFP3TAGHis6 with A1, A3, A4 or A7 when cells were provided with these ncMs and their cognate orthogonal synthetase / orthogonal tRNA pairs (illustrated in Figs.45r, 57 of EP2400299.0). These observations are consistent with these ncMs being poor substrates for ribosomal polymerization. For A4 and A7 we observe a decrease in GFP production upon addition of ncM (illustrated in Fig.45r of EP2400299.0); this is consistent with these ncMs, once acylated onto the orthogonal tRNA, inhibiting read-through of the amber codon. The discovery of orthogonal synthetases that are specific for these ncMs for provides a starting point for selecting ribosomes that polymerize them. Example 9 - Structure of β-amino acid protein To further characterize the incorporation of 12 at position 150 of GFP, we solved the structure of GFP(150(S)β3mBrFHis6 at 1.5 Å by X-ray crystallography (the protein was purified from cells harboring PylRS(12_1) and tRNAPylCUA (Fig.6i). The electron density shows two consecutive carbon atoms (C2 and C3) in the protein backbone at position 150; the meta-bromo phenyl substituent is attached to C3 of the β-amino acid, and the stereochemistry at C3 corresponds to the expected (S) stereoisomer. Our structure confirms the site- specific incorporation of the expected (S)β3mBrF-amino acid in the protein. The introduction of (S)β3mBrF leads to a notable kink in the beta barrel of GFP, when compared to the wildtype protein (illustrated in Supplementary Fig.38 of the priority document). Interestingly, the hydrogen bonding networks of the residues immediately preceding and following the β-amino acid in the polypeptide chain remain essentially unperturbed; this indicates that this beta barrel can accommodate the β3-amino acid at this position (illustrated in Supplementary Fig.38 of EP2400299.0). To the best of our knowledge, this represents the first structure for a protein produced in vivo that contains a β-amino acid. Taken together, our data demonstrate the site-specific incorporation of a β-amino acid in a protein produced in cells. Example 10 - Discussion The in vivo incorporation of backbone modified monomers is a long-standing, unaddressed challenge in expanding the scope of encoded cellular polymer synthesis beyond α-L amino acids and their close analogs4,30. The co-dependence of tRNA acylation and ribosomal polymerization31in the translational readouts commonly used to select for tRNA acylation12,13or ribosomal polymerization creates an evolutionary deadlock; this deadlock has hitherto limited the range of monomers that can be used to acylate specific orthogonal tRNAs in vivo. tRNA display breaks this co-dependence-based deadlock. It achieves this by creating a selectable acylation phenotype and coupling this phenotype to the sequence of the corresponding synthetase gene; this enables the direct selection for orthogonal aminoacyl-tRNA synthetases that acylate their cognate orthogonal tRNAs with ncMs, without any requirement for the ncM to be a ribosomal substrate or function in translation. Using tRNA display we have rapidly and scalably selected efficient aaRS systems for eight ncAAs; these selections consume orders of magnitude less compound than previous approaches. Moreover, we have used tRNA display to select, and improve, tRNA acylation systems for ncMs; these systems could not have been discovered using previous approaches. Because tRNA display generates a large number of synthetase sequences for which we can predict the activity and selectivity, the data generated from our approach may enable the de novo design of active sites for new ncMs. We note that it may be possible to extend tRNA display to many other aaRS / tRNA pairs, immediate targets include those pairs with minimal anticodon recognition – including quintuply orthogonal PylRS / tRNA pairs32, and pairs for Ala, Leu, and Ser. It may also be possible to extend the approach to discover active sites that recognize one ncM but not others, as required for the encoded synthesis of non-canonical heteropolymers5. We note that it may be possible to use extensions of tRNA display to select for genes that direct the biosynthesis of ncMs, or that bind to and protect tRNAs acylated with ncMs (e.g.: ncM-specific EF-Tu variants). Having established that tRNAPylCUAwas acylated with the ncM 12 in cells, we were able to identify a site in GFP where incorporation of this ncM was tolerated to realize the genetically encoded incorporation of a β-amino acid in a protein. In future work we will leverage the cellular acylation of tRNAs with ncMs to enable translation-based selections for orthogonal ribosomes28,31,33,34that can polymerize ncMs at a wider range of sites. Translation-based selections31may also generate orthogonal ribosomes that facilitate the encoded cellular synthesis of polymers composed of more diverse ncMs. We anticipate that the repertoire of ncMs that can be used to construct modified proteins and non-canonical polymers may be further expanded by combining approaches for directly encoding ncMs with elegant approaches for increasing the chemical scope of peptides and proteins via post-translational modification and protein ligation35-39. The genetic encoding of β-amino acids may enable the creation of protease-resistant proteins40-42in cells. Future developments may combine strategies for encoding non-canonical polymers in cells5with an expansion in the range of monomers that can be encoded to enable the encoded biosynthesis of foldamers43, composed of β-amino acids and other ncMs, which can form secondary, tertiary and quaternary structures44,45. We anticipate that it may be possible to complement and augment the canonical functions of cells through the design and directed evolution of genetically encoded foldamers. In addition to ncM 12, using tRNA display we have discovered orthogonal synthetase / orthogonal tRNA pairs that are selective for eight new ncMs, including β-amino acids, α,α-disubstituted-amino acids, β-hydroxy acids, and thereby directly facilitated the genetic encoding, and site-specific incorporation of β-amino acids, α,α- disubstituted-amino acids into proteins in a living organism. Example 11 – Confirming tRNA display feasibility using in non-mazei PylRS systems To confirm the feasibility of tRNA display in non-mazei PylRS systems, a set comprising MmPylRS, 1R26 PylRS, Nitra PylRS, the N-terminal domain-deficient Clos PylRS, and the corresponding quadruply orthogonal tRNAs were chosen. The three non-mazei PylRS will subsequently be named “1R26”, “deltaClos”, and “Nitra”. To extend tRNA display to the 1R26, deltaClos, and Nitra PylRS, their previously identified splitable tRNAs g1 (1R26), I2B72 (deltaClos), and Int6C10 (Nitra) were screened for suitable intergenic regions (IGRs) to generate functional stmRNA fusion constructs (Fig.22). Relevant sequences are SEQ ID NOs: 23 & 24, 25 & 26, and 27 & 28, respectively. The intergenic regions assessed were either a) proposed, based on tRNA homology, by a tRNA operon generator that was built for the design of non-split-tRNA operons (Dunkelmann et al., 2021 or WO2023285596 ), b) designed in-silico, or were c) the best intergenic regions for the MmPylRS-stmRNA. Screening was carried out by assessing the enrichment of stmRNA in the presence versus in the absence of a model substrate. To this end, the quantity of pulled-down stmRNA was measured via qPCR following reverse transcription. While for deltaClos its wt sequence was used, an 1R26 PylRS active site mutant specific for Carbobenzyloxy-L-lysine (CbzK), and a Nitra mutant with specificity for N-Methyl-L-hisitidine (NMH) were used, as the latter two respectively exhibit higher activity or less misacylation than their wt counterpart. Several highly active intergenic regions were identified for 1R26, one of which (the synthetic IGR greedy1 – SEQ ID NO: 21) even greatly surpassed the enrichment facilitated by the optimised Mm stmRNA (138-fold and 359- fold) in two separate experiments (360- and 678-fold; Fig.22ab). For deltaClos and Nitra several active intergenic regions were found as well, although the enrichments achieved by any screened intergenic region (including the intergenic regions highly active in the context of the 1R26 and / or MmPylRS) were significantly lower for these two systems than for the Mm stmRNA control (Fig.22c-e). The highest observed enrichment for deltaClos was 22-fold and 12-fold for Nitra. Taking together the four experiments for deltaClos and Nitra the IGR argY-argZ (SEQ ID NO: 17) seems to be the most promising candidate for both systems. Note that the term “enrichment” in this set of experiments relates to the comparison of - ncM and + ncM conditions and is therefore rather a proxy for close-to-optimal tRNA display selectivity values achievable with these systems, and less for the tRNA display enrichment metric, which is quantitatively intertwined but functionally distinct. Another interesting observation in the intergenic region screening were the large differences in acylated molecule counts for the assessed systems. While they were 2 to 3 orders of magnitude higher for the highly active 1R26 stmRNAs when compared to the MmPylRS stmRNA (itself based on the IGR leuP-leuV – SEQ ID NO: 19), this difference is likely owed to the fact that the final Mm construct featured a greatly attenuated RBS strength (in- silico predicted to be 5 a.u., contrasting 831 a.u. for the RBS currently driving 1R26 stmRNA translation). The qPCR acylated stmRNA counts for deltaClos were however still slightly lower than for the Mm control, although the deltaClos stmRNA constructs supported a 20-fold stronger RBS (105 a.u.). This observation was even more severe for Nitra, albeit the assessed Nitra constructs possessed a 50-fold stronger RBS (251 a.u.). Optimisation of the RBS in the MmPylrRS stmRNA has been carried out to ensure that the acylation-dependent pulldown could resolve different PylRS variant activities over a wide range of activities by differential enrichments of the encoding stmRNAs. For the original strong RBS, which was present in early MmPylrRS stmRNA iterations, no substantial enrichment differences were observable for PylRS variants of medium to high activity, as the expressed PylRS amount seemingly saturated the stmRNA population available for acylation. Due to the thousandfold higher acylation for the greedy1-based 1R26 stmRNA when compared to the final Mm construct, it was reasoned that a similar saturation effect, which would hamper tRNA display evolutions, may occur. In order to address this possibility especially for 1R26, but also for the other systems, it was necessary to obtain PylRS variants for these three systems that would cover their activity range over several orders of magnitude. To this end active sites from MmPylRS variants from a previous CbzK selection (more specifically residues 306, 309 and 346) and rationally designed derivatives were transplanted into the homologous positions of the (non- stmRNA) wt 1R26, deltaClos, and Nitra PylRS based on a sequence alignment. The activity of these variants with the (non-split) tRNAs underlying the respective stmRNA constructs was then assessed by a GFP amber suppression assay (Fig.23a). For deltaClos no useful set of attenuated variants was identified in our standard pMB1 system or in the context of a p15a backbone that by experience often facilitates elevated signal levels in this assay. For both 1R26 and the Nitra PylRS systems, however, CbzK-specific variants were found that cover a broad activity range up to their respective wt level (Fig.23bc). The GFP amber suppression data also supports that the Nitra PylRS is significantly less active than 1R26 and especially the Mm PylRS system as the stmRNA qPCR experiment suggested before. This conflicts with our other data, a conflict that was resolved by a repetition of the active site transplant screening with PylRS expression from p15a (Fig.30). All assessed variants for all three systems show far stronger signals when expressed from p15a up to saturation at around 20000 a.u. (equivalent to the signal level for the Mm wt and the most active 1R26 variants in pMB1). This suggests that the GFP assay driven from p15a cannot resolve PylRS amber suppression activities above 20% (or even less) of the wt MmPylRS signal due to saturation phenomena. Combining this observation with the IGR screening qPCR data, where Nitra and deltaClos exhibited numbers of acylated stmRNAs considerably lower than 1R26 and even lower than the MmPylRS stmRNA with its weak RBS, it seems likely that the poor enrichments observed before for Nitra and deltaClos are the consequence of far lower PylRS activity when compared to the 1R26 and Mm PylRS. To assess whether tRNA display by a 1R26 stmRNA can resolve widely different acylation activities, the identified CbzK-selective active sites were cloned into the greedy1-based stmRNA. Then the acylation-dependent pulldown experiment was repeated as before. An 830-fold enrichment was reached for the non-attenuated active site, again far surpassing the enrichments observed before with the Mm system. The comparison with the most strongly attenuated variant (5-fold enrichment) emphasizes the importance of highly active synthetases for the setup of tRNA display (Fig.24a). Importantly, no saturation of acylated molecule counts could be detected for the most active 1R26 variants and instead a strong correlation with the corresponding amber suppression activity data was observed (Fig.24b). Thus, the greedy1-1R26 stmRNA should be ready for tRNA display selections. Example 12 – Extending tRNA display to non-PylRS aaRS / tRNA pairs While the establishment of tRNA display for additional PylRS may offer the easiest access to the evolution of orthogonal active sites, the extension of tRNA display to non-pyrrolysyl systems promises to broaden the evolvable ncM substrate scope to a significantly larger extent. Currently, tRNA display takes advantage of the absence of anticodon recognition by PylRS to establish a mRNA fusion in the tRNAPylanticodon loop. Thus, it seems plausible that tRNA display may be extended to aaRS, which do also not rely on anticodon binding to recognize the identity of their cognate tRNA. This is the case for leucyl-, alanyl-, and seryl-aaRS, which therefore lend themselves as next steps in the tRNA display journey. Split constructs, where the tRNA was split under removal of the anticodon loop and concurrent circular permutation, were made for an in-silico designed tRNAAla(Fig.25a – relevant sequences include SEQ ID NOs: 46 & 47, SEQ ID NOs: 72 & 73, and SEQ ID NO: 92). A Cy3-labelled DNA probe complementary to the 3’ end of the alanyl-tRNA was designed as described by Cervettini et al. (2020), validated for its specificity, and a fluoro-tREX assay set up based on this probe. The alanyl-tRNA was produced at high level, the designed probe selectively bound the unsplit alanyl-tRNA, and it was possible to detect acylation of this tRNA (Fig.25b). Therefore, it was next assessed with this fluoro-tREX assay, whether the tRNAAlacould be split. Three different IGRs were screened for this purpose: argY-argZ, leuP-leuV, and ileV-alaV (the latter of which was proposed by the operon generator, while the other two had previously proven robust with the pyrrolysyl systems). All three facilitated successful maturation of the split tRNAAlaas shown by the acylation-dependent probe-mediated Cy5 incorporation at the tRNA 3’ end (Fig.25c). The leuP-leuV-based split tRNAAlamaintained the activity and orthogonality of the unsplit tRNA (Fig.25d), though some variability was observed between the replicates. The results of these experiments suggest that tRNA display can be extended to aaRS / tRNA pairs without anticodon recognition. Example 13 – Extending tRNA display to aRS / tRNA pairs with anticodon recognition We disclose herein a broadly generalisable strategy to establish the necessary phenotype-genotype linkage, thereby making it possible to extend tRNA display to aRS / tRNA pairs with anticodon recognition. In a structure guided approach, ten positions were identified that may be canonically permissive to tRNA splits due to a lack of participation in secondary or tertiary structure interactions, while interfering minimally with tRNA identity elements for any of the 17 tRNA isoacceptor classes exhibiting anticodon recognition (Fig.26). These splitting sites were also deemed suitable, as all of them have at least one more adjacent site that may allow for a split (based on nucleotide positions important for structure and recognition), thus allowing for deletions between these proximal sites to compensate for the steric demand of the stabilising stem. Including such compensatory deletions of one or two nucleotides between adjacent splitting sites, the ten identified positions thus amount to a total of 18 possible splits falling into four classes: 6 D loop splits (class D), 3 splits to the side of the anticodon (class A), 6 variable loop splits (class V), and 3 T loop splits (class T). It should be noted that across all 17 isoacceptor classes these 10 positions only neighbour identity elements in 5 cases (tRNAIleand tRNAGlyposition 16, tRNAProposition 37, tRNAPheposition 31, and tRNAGluposition 46), for all of which the availability of multiple splitting sites and potential compensatory deletions makes the existence of non-perturbant splits more likely. To assess the splitting permissiveness of these sites and get an impression of their generalisability, two systems were chosen as models. These two cover a class I aaRS (Trp), as well as a class II aaRS (Pro), thus representing two structurally distant classes with different tRNA binding modes. The activity and tRNA orthogonality of both systems was assessed by a amber suppression assay and deemed sufficient, as both, the aRS / tRNA pairs showed activities close to the MmPylRS system, and considerable tRNA activity (Fig.27a). Fig.27b-d summarise the rationale behind the split constructs: The 11 bp stabilising stem of the pyrolysyl- and alanyl-tRNA splits was elongated by three GC pairs to compensate for the reduced continuous stem (upon removal of the anticodon loop the stabilising stem protruded directly from the anticodon stem there). The IGR leuP-leuV (SEQ ID NO: 19) was chosen for both tRNAs based on its previous robust functionality across several systems. For one system a second IGR (leuW-glnU), which was proposed by the tRNA operon generator, was also assessed. Of the 18 possible splitting site combinations a subset of 10 splits was chosen, which covered all split classes (i.e. all tRNA loops), all possible splitting sites, and within each split class all possible compensatory deletion sizes (Fig.27cd). D16 / 17 was not cloned for tRNAProas RNA secondary structure predictions anticipated a collapse of the D loop. For tRNATrpthe variant A31 / 33 could not be cloned for either IGR, suggesting toxicity to the endogenous translational system. Additionally, a control that faithfully recapitulated the design of split tRNAs for the pyrrolysyl systems and tRNAAla– i.e. a split that removed the entire anticodon loop – was also designed, to investigate the possibility of setting up tRNA display in its established form for these systems irrespective of their synthetases anticodon recognition. Fluoro-tREX assays based on probes specific for the respective tRNAs were set up as before (Fig.28ab; Fig.31 shows a second biological replicate). The probes were specific for their tRNAs and tRNA acylation was detectable. Both tRNAs seemed to be orthogonal (compare lane 11 and 12), although tRNAProto a substantially higher extent, while the tRNATrporthogonality observed in the amber suppression assay did not carry over at the expected degree, arguably due to kinetic phenomena that are not well reflected in the GFP assay, which is an endpoint measurement. While the activity of both synthetases remained behind the Mm system, this was not regarded problematic for the purpose of the experiment. The tRNAProprobe seems to dissociate during urea PAGE as no Cy3 signal was observed, but the probe dependent Cy5 signal was detected, hence screening of the split tRNAs by fluoro-tREX was pursued next for both systems (Fig.28ab). Neither the tryptophanyl, nor the prolyl system yielded an active tRNA when split by the current tRNA display approach, where the anticodon loop was completely removed. This suggests, that different strategies are necessary to extend tRNA display to any aaRS / tRNA pair where the anticodon is a tRNA identity element. For the tRNATrpsplits, transcripts were detected for most of the constructs (the lack of T loop split transcripts may be an artifact as the probes reach from the tRNA 3’ end to the variable loop). While no acylation was observed for any Trp construct permuted by leuW-glnU (Fig.32), three splits (A32 / 33, D15 / 17, and D16 / 17) retained their ability to be acylated for leuP-leuV (Fig.28c). Intriguingly, tRNAProclosely resembled this observation and also facilitated acylation for splitting sites adjacent to the anticodon (A31 / A33 and A32 / 33), and in the D loop (D15 / 17 and, albeit only weakly, D15 / 18). Except for tRNAPro-D15 / 18 these hits were therefore investigated more closely. To investigate, whether the splitable tRNATrpand tRNAProalso maintain their orthogonality, or the observed acylation is caused by endogenous synthetases, the identified hits were analysed in presence and absence of their respective cognate synthetase (Fig.29). It was confirmed that the splitting sites on the 5’ side of the anticodon and in the D loop can facilitate tRNA maturation and acylation for both tRNAs. For the tryptophanyl system a significant decrease in acylation was observed, while especially the (anticodon preserving) splits in the anticodon loop of tRNAProkept high acylation signals. Interestingly, no split led to an orthogonality decrease, but instead all splits increased on the orthogonality of the split tRNAs towards their aaRS. This observation, combined with the identification of similar splitting sites across two tRNA isoacceptor classes belonging to aaRS from both class I and II, while screening only a subset of the above proposed splitting site combinations, suggests that tRNA display may actually be extended to orthogonal pairs with anticodon recognition based on these design principles. Example 14 – Methods Buffers Resuspension buffer (RB): 100 mM NaOAc, 50 mM NaCl, 0.1 mM EDTA, pH 5.0 Deacylation buffer (DB): 50 mM Bicine pH 9.6, 1 mM EDTA Buffer D (D): 50 NaOAc pH 5, 150 NaCl, 10 mM Mg2Cl, 0.1 mM EDTA Hybridization buffer (HB): 10 mM Tris-HCl, 25 mM NaCl, pH 7.4 Klenow (exo-) master mix with Cy5-11-dCTP (KMM-Cy5): 17 μL water, 5 μL 10x NEBuffer 2.0, 1 μL Klenow exo(-), 1 dNTPS-dCTP (10 mM), 1 μL Cy5-11-dCTP (20 μM). Orange loading dye (OLD): 8 M urea, Orange G Klenow (exo-) master mix with Cy5-11-dCTP mini (KMM-Cy5-mini): 1.1 μL water, 1.2 μL 10x NEBuffer 2.0, 0.2 μL Klenow exo(-), 0.5 dNTPS-dCTP (5 mM), 1 μL Cy5-11-dCTP (5 μM). Klenow (exo-) master mix with Bio-11-dCTP (KMM-Bio): 17 μL water, 5 μL 10x NEBuffer 2.0, 1 μL Klenow exo(-), 1 dNTPS-dCTP (10 mM), 1 μL Bio-11-dCTP (20 μM). Washing buffer (WB): 10 mM Tris-HCl, 150 mM LiCl, 1 mM EDTA, 0.05% v / v Tween20, pH 7.5 Binding buffer (BB): 20 mM Tris-HCl, 1 M LiCl, 2 mM EDTA, 0.05% v / v Tween20, pH 7.5 Formamide loading buffer (FMB): 90% formamide SDS Lysis Buffer (SLB): 100 mM NaOAc, 50 mM NaCl, 0.1 mM EDTA, pH 5.0, 1% (m / w) SDS Alkaline washing buffer (AWB): 25 mM NaOH, 4 mM EDTA, 0.05% Tween20 Reverse transcription hybridization mix (RHM): 1 μL DNA primer (2 μM), 1 μL 10 mM dNTPs, 1 μL 10x HB, 10 μL water RT master mix (RMM): 4 μL SSIV buffer, 1 μL RNAse Out, 1 μL SSIV RT, 10.1 M dithiothreitol (DTT) Acidic washing buffer 1(aWB1): 100 mM NaOAc pH 5 Acidic washing buffer 1 plus Tween20 (aWB1-T): 100 mM NaOAc pH 5, 0.01% (v / v) Tween-20 Acidic washing buffer 2 (aWB2): 20 mM NaOAc pH 5 Media SOC: Super Optimal Broth plus glucose 2xYT-s: Yeast Extract Tryptone supplemented with 75 μg / mL spectinomycin 2xYT-s-t: Yeast Extract Tryptone supplemented with 75 μg / mL spectinomycin and 10 μg / mL tetracycline 2xYT-s-ap: Yeast Extract Tryptone supplemented with 75 μg / mL spectinomycin and 50 μg / mL apramycin 2xYT-am: Yeast Extract Tryptone supplemented with 75 μg / mL ampicillin Chemicals NcM 1 and 2 were purchased from Bachem. NcM 4 was purchased from Fluorochem. NcM 5 was purchased from Ambeed. NcMs 6, 8, 13, 17 and 18 were purchased from Enamine. NcM 7 was purchased from aaBlocks. NcMs 9, 10, 15 and 16 were purchased from Merck. NcMs 12 and 20 were purchased from BLD. NcM 14 was purchased from Advanced ChemBlock. NcM 19 was purchased from AstaTech. NcMs 3 was synthesized as previously described (Spinck, M. et al. s. Nature Chemistry 15, 61-69 (2023)) and ncM 11 was custom synthesized as previously described (Tang, S. et al. Nature 602, 701-707 (2022)). NcMs 13 and 18 were Boc deprotected in concentrated HCl in dioxane. DNA constructs cloning Standard cloning was performed by Gibson assembly using NEBuilder® HiFi DNA Assembly Master Mix (NEB) according to manufacturers guidelines. Libraries were generated by enzymatic inverse PCR, as previously described. Briefly, a template plasmid was amplified by PCR using two primers (see primer list) containing degenerate codons at desired mutagenesis sites and a BsaI cleavage site. In the case of custom mixes, primers containing different codons were manually mixed and used for PCR reactions. PCR products were gel purified and digested using BsaI and DpnI. Subsequently, samples were purified, ligated using T4 Ligase, and transformed into electrocompetent E. coli DH10ß cells ensuring a minimal transformation efficiency of 109. Individual colonies (>10) were evaluated using sanger sequencing for quality control of the library assembly. Total plasmid DNA was prepared from the resulting culture, sequenced as a bulk using Sanger sequencing and used for subsequent experiments. General Protocols Isolation and oxidation of tRNAs (Protocol A) This protocol was used to isolate tRNAs from 1-10 mL of cell culture. Chemically competent DH10ß cells were transformed with a pMB1 plasmid encoding a PylRS variant and a tRNA, or a circularly permutated split tRNA, and rescued in 1 mL of SOC for 1 h at 37°C, 700-1000 r.p.m. Cells were transferred into selective 2xYT-s medium and grown over night. Overnight cultures were diluted in a ratio between 1:20 and 1:40 and grown to OD600of 0.5-1. Cells were centrifuged at 4200 rcf at 4˚C for 12 min, taken up in 200 µL RB, transferred to a 96- well plate and centrifuged at 4200 rcf at 4˚C for 12 min. Cells were resuspended in 135 µL RB and 15 µL liquid phenol was added. Cells were lysed by shaking at 650 r.p.m. for 20 min, and then centrifuged at 4200 rcf at 4˚C for 20 min; the cell lysate was added to 40 µL chloroform, and the resulting suspension mixed by pipetting up and down. The mixture was centrifuged at 4200 rcf at 4˚C for 10 min, and 115 µL of the aqueous layer were transferred into 6 µL 0.1 M NaIO4. The isolated RNA was oxidized for 1 h on ice, and the oxidation reaction was quenched by addition of 8 µL of 0.1 M DTT. tRNAs were purified using the ZR-96 Oligo Clean & Concentrator from Zymo Research. In brief, 250 µL oligo binding buffer was added to the oxidation reaction, subsequently 400 µL isopropanol was added, and the mixture transferred to a 96-well silica column plate. The plate was centrifuged for 2 min, 4200 rcf at room temperature, and 800 µL of oligo wash buffer was added. The plate was centrifuged for 2 min, 4200 rcf at room temperature, aerated, and centrifuged for another 4 min, 4200 rcf at room temperature. Finally, the RNA was eluted in either 14 µL water, when the samples were not processed, or in 50 µL water, when the samples were further deacylated, by centrifugation for 4 min, 4200 rcf at room temperature. Isolation and oxidation of tRNAs (Protocol B) The volumes described in this protocol were used to isolate tRNAs from 5-25 mL of cell culture as previously described. In brief, cells were grown as described in protocol A, washed with 800 µL RB and transferred to a 1.5 mL Eppendorf tube. Cells were taken up in 225 µL RB and 25 µL liquid of phenol was added. Cells were lysed by vortexing for one minute and incubation, with head over tail rotation, for 20 min. Lysed cells were centrifuged for 15 min, 20000 rcf. at room temperature, the cell lysate was added to 250 µL chloroform, the samples were vortexed for one minute and then centrifuged, 10 min, 20000 rcf. at room temperature.200 µL of the aqueous layer was transferred into 10 µL 0.1 M NaIO4and the RNA was oxidized for 1 h on ice. Finally, the oxidized RNA was added to 440 µL EtOH and precipitated for at least 20 min at -20 °C. The samples were centrifuged for 25 min, 20000 rcf. at 4 °C and aspirated. RNA pellets were dried for 10 min at room temperature and dissolved in water or buffer D. tRNA deacylation 45 µL of isolated RNA was added to 5 µL 10 x DB and tRNAs were deacylated for 36 min at 42˚C. The deacylation reaction was quenched by addition of 6 µL 3 M NaOAc, and tRNAs were purified using the ZR-96 Oligo Clean & Concentrator from Zymo Research, as described in protocol A for the isolation and oxidation of tRNAs (with the exception of using 100 µL of oligo binding buffer, instead of 250 µL). Deacylated tRNAs were eluted in 14 µL water. Fluoro-tREX This protocol weas used to run the experiments shown in Fig.3c and Fig.10. RNA concentrations were adjusted to the lowest common denominator and 10 µL of RNA was added to 2.5 µL 10 x HB, 11.5 µL water and 1 µL extension primer (2 µM). The DNA primer was hybridized at 65°C for 5 min, before addition of 25 µL KMM- Cy5 and extension at 37°C for 6 min. Samples were purified using the 10 µg NEB Monarch RNA clean-up Kit (NEB) and eluted in 12 µL water.12 µL of OLD was added, the samples were loaded onto a Novex TBE 6M urea 10 or 15% PAGE gel (Invitrogen) and run for 36 minutes in 0.5x Tris-borate-EDTA (TBE) buffer at 270 V. Gels were imaged on an Amersham Typhoon Biomolecular Imager (GE) using the Cy3 and Cy5 emission filters. Then gels were stained with SYBR Gold (Invitrogen) and imaged again using the same filters. Mini-fluoro-tREX Unless stated otherwise, all fluoro-tREX experiments were run with the mini-fluoro-tREX protocol. RNA concentrations were adjusted to match the lowest concentration in the samples being compared.6 µL of RNA was added to 0.5 µL 10 x HB, and 0.5 µL extension primer (2 µM). The DNA primer was hybridized at 65 °C for 5 min, before addition of 5 µL KMM-Cy5-mini and extension at 37 °C for 6 min. Samples were analyzed as described for fluoro-tREX. Bio-tREX RNA concentrations were adjusted to match the lowest concentration in the samples being compared.10 µL of RNA was added to 2.5 µL 10 x HB, 11.5 µL water and 1 µL extension primer (2 µM). The DNA probe was hybridized at 65 °C for 5 min, before addition of 25 µL KMM-bio and extension at 37 °C for 6 min.10 µL of Dynabeads MyOne Streptavidin C1 magnetic beads (Invitrogen) per reaction were washed three times with 200 µL WB, resuspended in 50 µL BB; the beads were added to the extension reaction and binding was performed for at least 30 min at 4°C, with head over tail rotation. The supernatant was removed, and the beads were washed four times with 200 µL WB. The washed beads were resuspended in 10 µL FLB and heated to 98°C for three minutes to release the tRNAs. Beads were removed and the supernatant was directly loaded onto a Novex TBE 6M urea, 10- or 15%-PAGE gel (Invitrogen) and run for 36 minutes in 0.5x TBE at 270 V. Gels were stained with SYBR Gold (Invitrogen) and imaged on an Amersham Typhoon Biomolecular Imager (GE) using the Cy2 emission filter. Northern blotting tRNAs were isolated following the general protocol A or B, omitting the oxidation by NaIO4.2 to 3 µg of RNA was loaded onto acidic urea PAGE gel (9% acrylamide (19:1), 100 mM sodium acetate pH 5, 8 M urea) and the gel was run for 12-16 , using 100 mM NaOAc as running buffer, at 6 Watt constant power. The gel was stained with SYBR gold (Invitrogen) to identify the tRNAs and an appropriate section of the gel was cut and blotted using iBlot DNA Transfer Stack (Invitrogen) with the iBlot Dry Blotting System. The tRNAs were cross-linked to the membrane (Stratalinker UV Crosslinker 2400), which was subsequently blocked in Ambion ULTRAhyb-Oligo buffer (Invitrogen) for 30 min. The biotinylated DNA probe was added to a final concentration of 0.2 µg / mL and hybridized overnight at 37°C and 160 r.p.m. The membrane was washed three times with 20 mL 0.5 x TBE buffer and transferred into 15 mL Odyssey blocking buffer for 20 min before addition of IRDye® 800CW Streptavidin (LI-COR) to a final concentration of 0.2 µg / mL. Finally, the membrane was washed three times with 20 mL 0.5x TBE and imaged on an Amersham Typhoon Biomolecular Imager (GE) using the IR long range emission filter. mRNA extraction and oxidation (A) The volumes given are suited for 2 to 3 mL of cell culture and were adjusted proportionally when required. Chemically competent BL21 cells were transformed with a pColE1 plasmid encoding the stmRNA construct, which was under the control of a T7 promoter and T7 terminator, rescued in SOC, shaken at 220 r.p.m., 37˚C for 1 h, diluted into 2xYT-am and grown overnight. The overnight cultures were diluted in a ratio of 1:20 into 2xYT- am in absence or presence of the ncM and grown at 37˚C, 220 r.p.m. to an OD600of 0.5-0.8. PylRS production was induced by addition of Isopropyl β-D-1-thiogalactopyranoside (IPTG) to a final concentration of 1 mM and cells were grown for 20 min, 220 r.p.m. at 37˚C. Cells were centrifuged at 4200 rcf at 4˚C for 12 min, resuspended in 800 µL RB, transferred to a 96-well plate and centrifuged at 4200 rcf, 4˚ C for 12 min. Subsequently, the procedure outlined in the user manual of the Agencourt® RNAdvance™ Cell v2 RNA isolation kit (Beckman) was followed. In brief, 200 µL LBE containing 10 µL proteinase K were used to lyse cells at room temperature for 1 h.244 µL BBC beads were mixed with 266 µL isopropanol, added to the lysate, and the RNA was bound to the beads for 10 min at room temperature. Beads were washed three times with 200 µL 80% EtOH, after the final wash the beads were carefully dried, and the RNA was eluted in 80 µL water. 70 µL of the RNA solution was added to 40 µL RB and 7.5 µL of 0.1 M NaIO4. The oxidation was run on ice for 1 h and quenched with 10.5 µL 100 mM DTT. To the oxidized RNA, a mix of 1.5 µL 1.6 M Na2CO3 and 16.5 µL DNAse I buffer (Ambion I DNAse Thermofisher) was added and the samples were resuspended. Subsequently, 18 µL of DNAse I were added, and the RNA incubated at 37˚C for 30 min. The digestion was cooled on ice and 300 µL Agencourt RNAClean XP beads (Beckman) were added and the RNA bound for 10 min at rt. The beads were washed three times with 80% EtOH. After the final wash the beads were carefully dried, and the RNA was eluted in 25 µL water. mRNA extraction and oxidation (B) A similar protocol as outlined in procedure A was followed, but the RNA was isolated using acid phenol / chloroform extraction. In brief, BL21 E. coli harboring stmRNAs were grown in 5 mL of 2xYT-am in the presence or absence of the ncM until an OD6000.5-0.9. At this point stmRNA expression was induced by addition of 1 M IPTG to a final concentration of 1 mM. After 20 min, cells were harvested by centrifugation at 4200 rcf for 12 min at 4°C. Cell pellets were resuspended in 800 µL RB, transferred into 1.5 mL Eppendorf tubes, and centrifuged for 3 min at 4200 rcf at room temperature. The supernatant was removed, and the pellets were resuspended in 500 µL SLB.500 µL acid-phenol was rapidly added and the tubes were vortexed for 1 min and centrifuged for 6 min, 21, 000 rcf at room temperature.450 µL of the aqueous layer was recovered and 50 µL of 2.5 M KCl was added. Acid-phenol chloroform extraction was repeated, retrieving 400 µL of the aqueous layer. 400 µL chloroform was added and the samples were vortexed for 1 min followed by centrifugation at 21,000 rcf 6 min at room temperature.300 µL of the aqueous layer was recovered, and chloroform extraction was repeated with 300 µL chloroform.200 µL of the aqueous phase were transferred to new 1.5 mL Eppendorf tubes and 10 µL of 0.1 M NaIO4was added. The oxidation reaction was run for 1 h on ice.440 µL ethanol was added and RNA was precipitated at -20 °C for at least 20 min. The RNA was pelleted by centrifugation at 21,000 rcf at 4°C for 30 min. The supernatant was removed, and the pellets air-dried for 10 min. The RNA was resuspended in 50 µL water.30 µL of each RNA sample was digested in 120 µL 1x Ambion DNAse I buffer including 12 µL Ambion DNAse I enzyme. Samples were purified with 50 µg NEB Monarch RNA clean-up Kit (NEB) and eluted into 20 µL water. We note, that when the RNA is isolated by acid-phenol chloroform extraction, an additional deacylation step in deacylation buffer (DB) is required, to measure acylation of stmRNAs with non-α amino acid monomers. Fluoro-mREX RNA concentration for all samples were adjusted to match the lowest concentration in the samples being compared.6-12 µg of RNA was added to a mixture of 1 µL DNA probe (2 µM), 2.5 µL 10x HB and water (added to a final volume of 25 µL). The primer was annealed at 65˚ C for 5 min.25 µL KMM-Cy5 was added. The primer was extended for 6 min, 37˚ C. Samples were purified using the 10 µg NEB Monarch RNA clean-up Kit (NEB) and eluted in 12 µL water.12 µL GLB was added to each sample. Gel electrophoresis was conducted using 1% agarose gels (cast using NorthernMax MOPS running buffer) in NorthernMax MOPS running buffer at 135 V, 42 minutes. Gels were stained with SYBR Gold (Invitrogen) and imaged on an Amersham Typhoon Biomolecular Imager (GE) using the Cy2 and Cy5 emission filter. Bio-mREX The RNA concentration of all samples was adjusted to match the lowest concentration in the samples being compared.6-12 µg of RNA was added to a mixture of 1 µL DNA probe (2 µM), 2.5 µL 10x HB and water (added to a final volume of 25 µL). The primer was annealed at 65˚ C for 5 min.25 µL KMM-bio were added. The primer was extended for 6 min,37˚C.10 µL of Dynabeads MyOne C1 streptavidin beads (Invitrogen) were washed two times with WB, added in 50 µL BB to the extension reaction, and the biotinylated stmRNAs were bound to the beads for 1 h, 4˚C, with head over tail rotation. The beads were washed on a magnetic stand with three times 200 µL WB, two times 200 µL AWB, one time 200 µL WB, one time 200 µL water and were finally resuspended in 13 µL RHM. After the AWB wash and after the final wash with WB, the beads were transferred into new plastic tubes. The primer was annealed at 65˚C for 5 min.7 µL of RMM was added and the RNA reverse transcribed at 50˚C for 10 min.1 µL of RNAseH was added and the mixture heated to 37˚C for 15 min and 98˚C for 3 min to release the cDNA from the beads. Finally, the cDNA was separated from the beads and either used for quantification by qPCR, NGS, or as a template for further cloning. qPCR of cDNA from bio-tREX qPCR reactions were run in triplicate for each bio-mREX sample and were composed of 2 µL of cDNA, 10 µL PowerUp SYBR Green Master Mix (Applied Biosystems), 0.4 µL of each primer and 7.2 µL water. A standard was generated by PCR of the MmPylRS gene and quantified using a Qubit 2 Fluorometer (Life Technologies) and the Qubit 1x dsDNA HS Assay Kit (Invitrogen). A five step fivefold serial dilution was used to generate a qPCR standard curve. This allowed calculation of qPCR efficiency and the number of molecules in each sample. qPCR was run on a ViiA 7 Real-Time PCR System (Applied Biosystems) using the standard supplier protocol for SYBR Green (Invitrogen). Preparation of cDNA from bio-tREX for NGS Half of the cDNA from the 20µL reverse transcription reaction from bio-tREX was added into a PCR mix containing 25 µL Q5® High-Fidelity 2X Master Mix, 12 µL water and 2 µL of a 10 µM predefined mix of indexing primers. A standard PCR program with 29 amplification cycles and an annealing temperature of 60 °C was used. Extension times were adapted to the amplicon length according to the manufacturer’s guidelines. DNA was bound to 100 µL of Agencourt AMPure XP (Beckman) for 10 min and the beads were washed three times with 200 µL 80% EtOH. Beads were dried and DNA was eluted in 25 µL water. DNA concentrations were measured using Qubit 2 Fluorometer (Life Technologies) and the Qubit 1x dsDNA HS Assay Kit (Invitrogen) and 80 ng of each amplicon were combined into the NGS library. The combined library was diluted in HT1 Hybridization Buffer (Illumina) to a concentration of 2 nM. PhiX (Illumina) was added to increase the diversity of the library at a 20% molar ratio.12 µL of the library was added to 18 µL HT1 Hybridization Buffer (Illumina) and 20 µL of the diluted mixture was used for NGS analysis. Cloning of cDNA from bio-tREX for further evolution Half of the cDNA from the µL 20 reverse transcription reaction, from bio-tREX, was added into a PCR mix containing 25 µL Q5® High-Fidelity 2X Master Mix, 12 µL water and 2 µL of a 10 µM predefined mix of golden gate assembly primers. A touchdown PCR program was used. The initial annealing temperature of 65 °C was decreased over 10 cycles by 0.5 °C per cycle. Subsequently 20 regular cycles using an annealing temperature of 58 °C were performed. Extension times were adapted to the amplicon length according to the manufacturer’s guidelines. DNA was bound to 100 µL of Agencourt AMPure XP (Beckman) for 10 min and the beads were washed three times with 200 µL 80% EtOH. Beads were dried and DNA was eluted in 25 µL water. The amplicon was then cloned into a new pColE1 backbone, previously amplified by golden gate primers, by two-piece Golden Gate assembly according to NEB (New England Biolabs) guidelines. NGS data analysis NGS was performed on a MiSeq system (in the case of the test evolution with library 1 and substrate 1) or a NextSeq2000 system (in all other cases). The resulting cDNA from tRNA display was amplified using oligos NGS A(1-8) and NGS_B(1-8) containing different combination of Nextera sequencing barcodes via PCR. Samples were purified, quantified, and combined in equimolar amounts. Paired end reads were first paired using PEAR7, and aligned to a reference sequence of MmPylRS using Bowtie28. The relevant library positions were extracted and translated to amino acids, and resulting variants were counted using R script. Subsequent operations were performed using the frequency of each variant in each library which was computed as the count value divided by the total number of counts of that library. Using R script, enrichment and selectivity scores were calculated for all variants as follows. First, variants that were only present in all positive replicates were considered (tables were merged using AND operator). Assuming that highly enriched sequences could potentially not be covered in the negative and the input samples but may still be of interest, the negative and the naïve replicates were merged to the positive table using an OR operator. A placeholder value of 0.95 counts was adopted in cases in which a replicate was not covering a specific variant. The resulting data set was used to calculate mean enrichments in the presence and in the absence of the respective substrate, computed as the quotient of the mean frequencies in one condition and the input condition. The resulting positive and negative enrichments were used to calculate the selectivity value for each variant (equivalent to the quotient of positive and negative frequencies). For further analysis, variants were filtered using an empirically determined threshold value for the normalized standard deviation of the positive frequency (dispersion error in the plus substrate condition). tRNA pulldown and ncM identification by LC-MS tRNAs were isolated from 8 mL of cells following the general protocol B omitting the oxidation by NaIO4. The RNA pellet was resuspended in 90 µL buffer D and RNA concentrations adjusted to match the lowest concentration in the samples being compared.0.5 µL of biotinylated DNA probe (100 µM) was added to the RNA and the DNA probe was hybridized at 65˚C for 5 min.40 µL Streptavidin Dynabeads MyOne C1 (Invitrogen), were washed twice with buffer D-T, and added into 10 µL buffer D. The beads were added to hybridization reaction and the probe was bound to the beads for minimally 30 min at 4 ˚C with head over tail rotation. The samples were washed three times with 200 µL of acW1-T, twice with 200 µL of acW1, three times with 200 µL of acW2 and once with 200 µL water, all on a magnetic stand.24 µL of DB was added and the beads, which were incubated at 42 ˚C for one hour.12 µL of the deacylation mix was added to 3 µL 6-Aminoquinolyl-N- hydroxysccinimidyl carbamate (AQC - 3 mg / mL in acetonitrile) and the reaction incubated at 55 ˚C for 15 min. Samples were analyzed on an Agilent Technologies 6130 Quadrupole LC / MS using single ion monitoring. GFP(150X)His6 and GFP(3X)His6, where X stands for any ncM, activity assay Chemically competent DH10ß cells harboring a p15A plasmid encoding GFP150TAGHis6or GFP3TAGHis6(two versions of the p15A plasmid with either a tetracycline or apramycin resistance cassette, which led to similar results, were used interchangeably) were transformed with a pMB1 plasmid encoding MmPylRS, or a mutant thereof, and MmtRNAPylCUA, rescued in SOC and grown over night in 2xYTs-t or 2xYTs-ap.20 µL of the overnight culture was diluted into 480 µL 2xYTs-t or 2xYTs-ap containing 0.2% L-arabinose in presence and absence of 2 to 4 mM of the respective ncM in a 96-well-plate format. Cells were grown for 16-20 h at 37˚C at 700 r.p.m. The plates were centrifuged for 12 min, 4200 rcf at 4˚C and the cells resuspended in 150 µL PBS.100 µL of the resuspended cells were transferred into a Costar 96-well flat bottom plate and the OD600, and GFP fluorescence was measured using a PHERAstar FS plate reader. GFP(150X)His6 isolation for MS analysis Three replicates of the protein produced as described above were combined in a 1.5 mL Eppendorf tube, centrifuged at 4200 rcf for 3 min, frozen at -20 ˚C, thawed and resuspended in 150 µL BugBuster (Millipore). Cells were lysed for 1 h with head over tail rotation. Lysed cells were centrifuged for 20 min, 20000 rcf, at 4 ˚C and the lysate was added to 20 µL of NiNTA beads. GFP(150)Hi6 was bound to the beads for 20 min at room temperature with head over tail rotation. The beads were washed six times with 60 µL 30 mM imidazole in PBS, and the protein was eluted with five times 30 µL 300 mM imidazole in PBS. For low activity mutants of ncM 125-15 mL of cell culture was used for protein production. The volumes of Bug Buster were adjusted proportionally, all other volumes were kept the same. Mass spectroscopy ESI-MS as well as MS / MS were obtained as previously described24, 27. Protein expression, purification, and crystallization Chemically competent DH10ß cells harboring a p15A plasmid encoding GFP150TAGHis6 and a pMB1 plasmid encoding MmPylRS(12_1) and MmtRNAPylwere transformed, rescued in SOC and grown over night in 2xYTs-ap. 10 mL of the overnight culture was diluted into 1 L 2xYTs-ap containing 0.2% L-arabinose in presence of 5 mM 12. Bacterial pellet of 1 L expression culture of GFP150(S)β3mBrF-His6 was lysed by sonication, centrifuged at 142,000 rcf for 30 minutes and supernatant bound to Ni-NTA beads (Qiagen). Beads were washed three times before protein was eluted and further purified by gel filtration using a Superdex 75 HiLoad 26 / 60 pg column (GE Healthcare) in 25 mM Tris pH 7.4, 200 mM NaCl and 0.06 % NaN3. The purified protein was concentrated using Vivaspin 20, 10,000 MWCO (Sartorius) to 6 mg / mL. Sample was Trypsin digested with Sequencing Grade Modified Trypsin (Promega) in a 50:1 ratio. Sample was incubated for 1 hour at 37 ^C, centrifuged at 21,000 rcf for 10 minutes before plating in crystal trays. Crystallization trials with multiple commercial crystallization kits were performed in 96-well sitting-drop vapor diffusion plates (Molecular Dimensions) at 18°C and set up with a Mosquito HTS robot (TTP Labtech). Drop ratios of 0.2 μL protein solution plus 0.2 μL reservoir solution were used for screening. The only useful dataset was collected from a crystal harvested from the Fusion screen (Molecular Dimensions) with following composition: 37.5 % PEG 3350 / PEG 1K / MPD (1:1:1), 0.1 M Bicine / Trizma pH 8.5, 0.8 % (w / v) Morpheus III Alkaloids and 0.12 M Morpheus Alcohols. Crystals were harvested and flash frozen in liquid nitrogen. Diffraction data collection, processing, and structure solution. Diffraction data were collected at the ESRF on beamline ID23-2 at an energy of 14.2 keV. Data were processed with XDS via the pipeline autoProc (Global Phasing ltd.). The structure was solved by molecular replacement with MolRep using the homologue model PDB 2B3P. Interactive building was performed with Coot, refinement with REFMAC5, and validation with Molprobity. Figures of the structure were prepared with PyMOL (PyMOL Molecular Graphics System, Schrödinger, LLC). Selection for ncAAs The selection was performed as described in Fig.17. RNA was isolated and oxidized as described in general procedure A. Bio-mREX was performed as specified in the general procedure. After the first round of selection the new libraries were assembled from the amplified cDNA as described above. After the second round of selection the NGS samples were prepared from the isolated cDNA as described above, the NGS run using a P2 600 cycles cartridge, and the data was analyzed as specified above. Selection for ncMs The selection was performed as described in Fig.18. RNA was isolated and oxidized as described in general procedure A. Bio-mREX was performed as specified in the general procedure. The NGS samples were prepared from the isolated cDNA as described above, the NGS run using a P1600 cycles cartridge, and the data was analyzed as specified above. Selection for substrate 12 using a random mutagenesis library The concentrations of the pMB1 plasmids encoding PylRS hits 12_1 and 12_2 were measured by Qubit 2 Fluorometer (Life Technologies) and the Qubit 1x dsDNA HS Assay Kit (Invitrogen) and the plasmids combined in equimolar amounts. The combined plasmids were used for an error prone PCR of the active site of PylRS using golden gate primers and the GeneMorph II kit (Agilent) at conditions leading to the maximal number of random mutations. The amplicons were cloned into a new pColE1 backbone by two-piece Golden Gate assembly according to NEB (New England Biolabs) guidelines. The selection was performed as outlined in Fig.20. RNA was isolated and oxidized as described in general procedure A. Bio-mREX was performed as specified in the general procedure. The NGS samples were prepared from the isolated cDNA as described above, the NGS run using a P2600 cycles cartridge, and the data was analyzed as specified above. Code availability The code for analysing tRNA display NGS data will be available upon publication at https: / / github.com / JWChin- Lab / tRNA_display.

[0002] References 1 Dumas, A., Lercher, L., Spicer, C. D. & Davis, B. G. Designing logical codon reassignment - Expanding the chemistry in biology. Chem Sci 6, 50-69, doi:10.1039 / c4sc01534g (2015). 2 Young, D. D. & Schultz, P. G. Playing with the molecules of life. ACS chemical biology 13, 854-870 (2018). 3 Chin, J. W. Expanding and reprogramming the genetic code. Nature 550, 53-60, doi:10.1038 / nature24031 (2017). 4 De La Torre, D. & Chin, J. W. Reprogramming the genetic code. Nature Reviews Genetics 22, 169-184, doi:10.1038 / s41576-020-00307-7 (2021). 5 Robertson, W. E. et al. Sense codon reassignment enables viral resistance and encoded polymer synthesis. Science 372, 1057-1062 (2021). 6 Spinck, M. et al. Genetically programmed cell-based synthesis of non-natural peptide and depsipeptide macrocycles. Nature Chemistry 15, 61-69 (2023). 7 Ellman, J. A., Mendel, D. & Schultz, P. G. Site-specific incorporation of novel backbone structures into proteins. Science 255, 197-200 (1992). 8 Mendel, D., Cornish, V. W. & Schultz, P. G. Site-directed mutagenesis with an expanded genetic code. Annual reviewof biophysics and biomolecular structure 24, 435-462 (1995).9 Hecht, S. M. Expansion of the genetic code through the use of modified bacterial ribosomes. Journal of molecular biology 434, 167211 (2022). 10 Katoh, T. & Suga, H. In vitro genetic code reprogramming for the expansion of usable noncanonical amino acids. Annual Review of Biochemistry 91, 221-243 (2022). 11 Melo Czekster, C., Robertson, W. E., Walker, A. S., Soll, D. & Schepartz, A. In Vivo Biosynthesis of a beta-Amino Acid- Containing Protein. J Am Chem Soc 138, 5194-5197, doi:10.1021 / jacs.6b01023 (2016). 12 Santoro, S. W., Wang, L., Herberich, B., King, D. S. & Schultz, P. G. An efficient system for the evolution of aminoacyl- tRNA synthetase specificity. Nature biotechnology 20, 1044-1048 (2002). 13 Chin, J. W., Martin, A. B., King, D. S., Wang, L. & Schultz, P. G. Addition of a photocrosslinking amino acid to the genetic code of Escherichia coli. Proceedings of the National Academy of Sciences 99, 11020-11024 (2002). 14 Tan, Z., Forster, A. C., Blacklow, S. C. & Cornish, V. W. Amino Acid Backbone Specificity of the Escherichia c oli Translation Machinery. Journal of the American Chemical Society 126, 12752-12753 (2004). 15 Pavlov, M. Y. et al. Slow peptide bond formation by proline and other N-alkylamino acids in translation. Proceedings of the National Academy of Sciences 106, 50-54 (2009).16 Katoh, T., Tajima, K. & Suga, H. Consecutive Elongation of D-Amino Acids in Translation. Cell Chem Biol 24, 46-54,doi:10.1016 / j.chembiol.2016.11.012 (2017). 17 Katoh, T. & Suga, H. Ribosomal Incorporation of Consecutive beta-Amino Acids. J Am Chem Soc 140, 12159-12167, doi:10.1021 / jacs.8b07247 (2018). 18 Dedkova, L. M. et al. beta-Puromycin selection of modified ribosomes for in vitro incorporation of beta-amino acids. Biochemistry 51, 401-415, doi:10.1021 / bi2016124 (2012). 19 Cervettini, D. et al. Rapid discovery and evolution of orthogonal aminoacyl-tRNA synthetase-tRNA pairs. Nat Biotechnol, doi:10.1038 / s41587-020-0479-2 (2020). 20 Barton, P., Laws, A. P. & Page, M. I. Structure–activity relationships in the esterase-catalysed hydrolysis and transesterification of esters and lactones. Journal of the Chemical Society, Perkin Transactions 2, 2021-2029 (1994). 21 Kobayashi, T., Yanagisawa, T., Sakamoto, K. & Yokoyama, S. Recognition of non-alpha-amino substrates by pyrrolysyl-tRNA synthetase. J Mol Biol 385, 1352-1360, doi:10.1016 / j.jmb.2008.11.059 (2009). 22 Soma, A. et al. Permuted tRNA genes expressed via a circular RNA intermediate in Cyanidioschyzon merolae. Science 318, 450-453 (2007). 23 El Yacoubi, B., Bailly, M. & de Crécy-Lagard, V. Biosynthesis and function of posttranscriptional modifications oftransfer RNAs. Annual review of genetics 46, 69-95 (2012).24 Dunkelmann, D. L., Oehm, S. B., Beattie, A. T. & Chin, J. W. A 68-codon genetic code to incorporate four distinct non- canonical amino acids enabled by automated orthogonal mRNA design. Nat Chem, doi:10.1038 / s41557-021-00764-5 (2021). 25 Salis, H. M., Mirsky, E. A. & Voigt, C. A. Automated design of synthetic ribosome binding sites to control protein expression. Nature biotechnology 27, 946-950 (2009). 26 Yanagisawa, T. et al. Structural Basis for Genetic-Code Expansion with Bulky Lysine Derivatives by an Engineered Pyrrolysyl-tRNA Synthetase. Cell Chem Biol 26, 936-949 e913, doi:10.1016 / j.chembiol.2019.03.008 (2019). 27 Dunkelmann, D. L., Willis, J. C. W., Beattie, A. T. & Chin, J. W. Engineered triply orthogonal pyrrolysyl-tRNA synthetase / tRNA pairs enable the genetic encoding of three distinct non-canonical amino acids. Nat Chem 12, 535-544, doi:10.1038 / s41557-020-0472-x (2020).28 Wang, K., Neumann, H., Peak-Chew, S. Y. & Chin, J. W. Evolved orthogonal ribosomes enhance the efficiency ofsynthetic genetic code expansion. Nat Biotechnol 25, 770-777, doi:10.1038 / nbt1314 (2007). 29 Fredens, J. et al. Total synthesis of Escherichia coli with a recoded genome. Nature 569, 514-518, doi:10.1038 / s41586- 019-1192-5 (2019). 30 Arranz-Gibert, P., Vanderschuren, K. & Isaacs, F. J. Next-generation genetic code expansion. Current Opinion inChemical Biology 46, 203-211 (2018).31 Schmied, W. H. et al. Controlling orthogonal ribosome subunit interactions enables evolution of new function. Nature 564, 444-448, doi:10.1038 / s41586-018-0773-z (2018). 32 Beattie, A. T., Dunkelmann, D. L., Chin, J. W. Quintuply orthogonal pyrrolysyl-tRNA synthetase / tRNAPyl pairs. Nature Chemistry In press (accepted) (2023). 33 Rackham, O. & Chin, J. W. A network of orthogonal ribosome x mRNA pairs. Nat Chem Biol 1, 159-166, doi:10.1038 / nchembio719 (2005). 34 Neumann, H., Wang, K., Davis, L., Garcia-Alai, M. & Chin, J. W. Encoding multiple unnatural amino acids via evolution of a quadruplet-decoding ribosome. Nature 464, 441-444, doi:10.1038 / nature08817 (2010). 35 Morinaka, B. I. et al. Natural noncanonical protein splicing yields products with diverse β-amino acid residues. Science359, 779-782 (2018).36 Lakis, E., Magyari, S. & Piel, J. In Vivo Production of Diverse β-Amino Acid-Containing Proteins. Angewandte Chemie 134, e202202695 (2022). 37 Camarero, J. A. & Muir, T. W. Native chemical ligation of polypeptides. Current Protocols in Protein Science 15, 18.14. 11-18.14.21 (1999). 38 Niquille, D. L. et al. Nonribosomal biosynthesis of backbone-modified peptides. Nature chemistry 10, 282-287 (2018). 39 Arnison, P. G. et al. Ribosomally synthesized and post-translationally modified peptide natural products: overview and recommendations for a universal nomenclature. Natural product reports 30, 108-160 (2013). 40 Seebach, D. et al. β-Peptides: Synthesis by Arndt-Eistert homologation with concomitant peptide coupling. Structure determination by NMR and CD spectroscopy and by X-ray crystallography. Helical secondary structure of a β-hexapeptide in solution and its stability towards pepsin. Helvetica Chimica Acta 79, 913-941 (1996). 41 Hintermann, T. & Seebach, D. The Biological Stability of?-Peptides: No Interactions between?-and?-Peptidic Structures? Chimia 51, 244-244 (1997). 42 Seebach, D. et al. Biological and pharmacokinetic studies with β-peptides. Chimia 52, 734-734 (1998). 43 Gellman, S. H. Foldamers: a manifesto. Accounts of chemical research 31, 173-180 (1998). 44 Horne, W. S., Price, J. L. & Gellman, S. H. Interplay among side chain sequence, backbone composition, and residue rigidification in polypeptide folding and assembly. Proceedings of the National Academy of Sciences 105, 9151-9156 (2008). 45 Wang, P. S. & Schepartz, A. β-Peptide bundles: Design. Build. Analyze. Biosynthesize. Chemical Communications 52, 7420-7432 (2016).

Claims

CLAIMS 1. A method of determining the effect of a polypeptide-of-interest or nucleic-acid-of-interest on acylation of a tRNA or acylation status of a tRNA, the method comprising: i) incubating the polypeptide-of-interest or nucleic-acid-of-interest, the tRNA, and a substrate with which the tRNA could be acylated, under conditions conducive to acylation of the tRNA, wherein the tRNA is split into at least two portions, and one of said tRNA portions is present as part of a fusion RNA molecule also comprising a sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest; ii) exposing the tRNA to conditions capable of labelling tRNAs that have been acylated; and iii) identifying whether the polypeptide-of-interest or nucleic-acid-of-interest is associated with a labelled tRNA.

2. The method of claim 1, wherein the tRNA does not comprise an anticodon.

3. The method of claim 1 or claim 2, wherein the tRNA is split at the location of an anticodon in a parental tRNA from which the tRNA portion is derived.

4. The method of any preceding claim, wherein the tRNA is a tRNAPyl, tRNALeu, tRNAAla, or tRNASer.

5. The method of any preceding claim, wherein the tRNA is split into a first tRNA chain and a second tRNA chain, the first tRNA chain comprises a tRNA portion from the 5’ side of the location of anticodon in a parental tRNA from which the tRNA portion is derived, and the second tRNA chain comprises a tRNA portion from the 3’ side of the location of anticodon in a parental tRNA from which the tRNA portion is derived.

6. The method of claim 5, wherein: a) the first tRNA chain comprises a sequence encoded by SEQ ID NO: 2 and the second tRNA chain comprises a sequence encoded by SEQ ID NO: 1; b) the first tRNA chain comprises a sequence encoded by SEQ ID NO: 24 and the second tRNA chain comprises a sequence encoded by SEQ ID NO: 23; c) the first tRNA chain comprises a sequence encoded by SEQ ID NO: 26 and the second tRNA chain comprises a sequence encoded by SEQ ID NO: 25; d) the first tRNA chain comprises a sequence encoded by SEQ ID NO: 28 and the second tRNA chain comprises a sequence encoded by SEQ ID NO: 27; or e) the first tRNA chain comprises a sequence encoded by SEQ ID NO: 47 and the second tRNA chain comprises a sequence encoded by SEQ ID NO:

46.

7. The method of claim 1, wherein the tRNA is split in the D loop, in the anticodon loop and to the 5’ side of the anticodon, in the variable loop, or in the T loop.

8. The method of claim 1 or claim 7, wherein tRNA is split between residues D15 & D16, D15 & D17, D15 & D18, D16 & D17, D16 & D18, D17 & D18, A31 & A32, A31 & A33, A32 & A33, V45 & V46, V45 & V47, V45 & V48, V46 & V47, V4640 & V48, V47 & V48, T56 & T57, T56 & T58, or T57 & T58.

9. The method of claim 8, wherein tRNA is split between residues D15 & D17, D15 & D18, D16 & D17, A31 & A33, A32 & A33, V45 & V48, V46 & V48, V47 & V48, T56 & T57, or T56 & T58.

10. The method of any one of claims 7-9, wherein the tRNA is a tRNATrpor tRNAPro.

11. The method of any one of claims 1 or 7-10, wherein the tRNA is split into a first tRNA chain and a second tRNA chain, the first tRNA chain comprises a tRNA portion from the 5’ side of the split in a parental tRNA from which the tRNA portion is derived, and the second tRNA chain comprises a tRNA portion from the 3’ side of the location of the split in a parental tRNA from which the tRNA portion is derived.

12. The method of claim 11, wherein the first tRNA chain and the second tRNA chain comprise, respectively, sequences encoded by: SEQ ID NO: 49 & SEQ ID NO: 48, SEQ ID NO: 51 & SEQ ID NO: 50, SEQ ID NO: 53 & SEQ ID NO: 52, SEQ ID NO: 55 & SEQ ID NO: 54, SEQ ID NO: 57 & SEQ ID NO: 56, SEQ ID NO: 59 & SEQ ID NO: 58, SEQ ID NO: 61 & SEQ ID NO: 60, or SEQ ID NO: 63 & SEQ ID NO:

62.

13. The method of claim 5 or claim 6, wherein the 3’ end of the first tRNA chain is connected to a first stem region, wherein the first stem region comprises 8-14, 8-12, or 10-12 nucleotides, and the 5’ end of the second tRNA chain is connected to a second stem region, wherein the second stem region comprises 8-14, 8-12, or 10-12 nucleotides and is complementary to the first stem region.

14. The method of claim 13, wherein the first stem region and the second stem region each consist of a region that is 10 nucleotides in length.

15. The method of claim 13, wherein: a) the first stem region is encoded by the sequence SEQ ID NO: 3 and the second stem region is encoded by SEQ ID NO: 4; or b) the first stem region is encoded by the sequence SEQ ID NO: 5 and the second stem region is encoded by SEQ ID NO: 6; c) the first stem region is encoded by the sequence SEQ ID NO: 7 and the second stem region is encoded by SEQ ID NO: 8; or d) the first stem region is encoded by the sequence SEQ ID NO: 64 and the second stem region is encoded by SEQ ID NO:

65.

16. The method of any preceding claim, wherein the tRNA is split into a first tRNA chain and a second tRNA chain, the first tRNA chain comprises a tRNA portion from the 5’ side of the split, and the second tRNA chain comprises a tRNA portion from the 3’ side of the split.

17. The method of claim 16, wherein the 3’ end of the first tRNA chain is connected to a first stem region and the 5’ end of the second tRNA chain is connected to a second stem region.

18. The method of any one of claims 5, 6, 11, 12, or 13-17, wherein the second tRNA chain is a part of the fusion RNA molecule also comprising the sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest.

19. The method of claim 18, wherein the connection between the second tRNA chain and the sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest is via the second stem region.

20. The method of any preceding claim, wherein the tRNA is split into a first tRNA chain and a second tRNA chain, and the fusion RNA molecule is expressed by a nucleic acid construct that comprises from 5’ to 3’: the sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest, a sequence encoding the second tRNA chain, and a sequence encoding the first tRNA chain.

21. The method of claim 20, wherein the tRNA is split into a first tRNA chain and a second tRNA chain, the first tRNA chain comprises a first stem region, and the second tRNA chain comprises a second stem region, the first stem region and the second stem region are complementary, wherein the nucleic acid construct comprises, from 5’ to 3’, the sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest, optionally a linker, a sequence encoding the second stem region, the sequence encoding the second tRNA chain, a loop, the sequence encoding the first tRNA chain, and a sequence encoding the first stem region.

22. The method of claim 21, wherein the loop is cleaved, removed, or spliced during RNA processing.

23. The method of claim 21 or claim 22, wherein expression of the nucleic acid construct results in: the fusion RNA molecule comprising, from 5’ to 3’, the sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest, optionally the linker, the second stem region, and the second tRNA chain; and a separate RNA molecule comprising, from 5’ to 3’, the first tRNA chain and the first stem region.

24. The method of any preceding claim, wherein the conditions capable of labelling tRNAs that have been acylated are conditions that comprise the following steps: a) exposure to conditions that block the 3’ end of free tRNAs but do not block the 3’ end of acylated tRNAs, b) exposure to conditions suitable to remove substrates that have acylated tRNAs; and c) exposure to conditions that attach a label to 3’ ends of tRNAs that are not blocked.

25. The method of claim 24, wherein step c) comprises conditions that lead to the addition of nucleotides to the 3’ ends of tRNAs that are not blocked, wherein at least one nucleotide comprises a label.

26. The method of any preceding claim, wherein the method comprises a step of capturing the fusion RNA molecule if it is associated with a labelled tRNA.

27. The method of any preceding claim, wherein the method comprises a step of sequencing a barcode associated with any labelled or captured fusion RNA molecule.

28. The method of any preceding claim, wherein the method comprises a step of obtaining sequence information for at least one sequence fused to any labelled or captured tRNA portion.

29. The method of any preceding claim, wherein the method comprises a step of obtaining sequence information for the sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest.

30. The method of any preceding claim, wherein the polypeptide-of-interest is an acyl-tRNA synthetase.

31. The method of any one of claims 1 to 29, wherein step i) comprises incubating the tRNA and the substrate under conditions conducive to acylation of the tRNA, and incubating the acylated tRNA with the polypeptide-of-interest or nucleic-acid-of-interest.

32. The method of claim 31, wherein, prior to step ii), the method comprises exposure to conditions suitable for conditionally deacylating acylated tRNAs.

33. The method of any one of claims 1-29, 31, and 32, wherein the polypeptide-of-interest or nucleic-acid-of- interest is capable of chemically altering the substrate or wherein the method comprises determining whether the polypeptide-of-interest or nucleic-acid-of-interest is capable of chemically altering the substrate.

34. A split tRNA comprising a first chain and a second chain, wherein:the first chain comprises a first tRNA portion and comprises a first stem region, and the second chain comprises a second tRNA portion and comprises a second stem region.

35. A split tRNA comprising a first chain and a second chain, wherein: the first chain comprises a first tRNA portion and comprises a first stem region, and the second chain comprises a second tRNA portion and comprises a second stem region; the first tRNA portion corresponds to the 5’ portion of a parental tRNA split at the anticodon and the second tRNA portion corresponds to the 3’ portion of the parental tRNA split at the anticodon; the first stem region is located at the 3’ end of the first tRNA portion and the second stem region is located at the 5’ end of the second tRNA portion; and the first stem region and the second stem region are complementary.

36. The method of claim 35, wherein a) the first tRNA portion comprises a sequence encoded by SEQ ID NO: 2 and the second tRNA portion comprises a sequence encoded by SEQ ID NO: 1; b) the first tRNA portion comprises a sequence encoded by SEQ ID NO: 24 and the second tRNA portion comprises a sequence encoded by SEQ ID NO: 23; c) the first tRNA portion comprises a sequence encoded by SEQ ID NO: 26 and the second tRNA portion comprises a sequence encoded by SEQ ID NO: 25; d) the first tRNA portion comprises a sequence encoded by SEQ ID NO: 28 and the second tRNA portion comprises a sequence encoded by SEQ ID NO: 27; or e) the first tRNA portion comprises a sequence encoded by SEQ ID NO: 47 and the second tRNA portion comprises a sequence encoded by SEQ ID NO:

46.

37. The split tRNA of any one of claims 34 to 36, wherein the first and the second stem region are 8-14, 8-12, or 10-12 nucleotides in length.

38. The split tRNA of any one of claims 34 to 37, wherein: a) the first stem region is encoded by the sequence SEQ ID NO: 3 and the second stem region is encoded by SEQ ID NO: 4; or b) the first stem region is encoded by the sequence SEQ ID NO: 5 and the second stem region is encoded by SEQ ID NO: 6; c) the first stem region is encoded by the sequence SEQ ID NO: 7 and the second stem region is encoded by SEQ ID NO: 8; or d) the first stem region is encoded by the sequence SEQ ID NO: 64 and the second stem region is encoded by SEQ ID NO:

65.

39. The split tRNA of any one of claims 34 to 38, wherein the first and second tRNA portion are from a tRNAPyl, tRNALeu, tRNAAla, or tRNASer.

40. The split tRNA of claim 34, wherein the first tRNA portion corresponds to the 5’ portion of a parental tRNA split at a split site and the second tRNA portion corresponds to the 3’ portion of the parental tRNA split at a split site; and the split site is in the D loop, in the anticodon loop and to the 5’ side of the anticodon, in the variable loop, or in the T loop.

41. The split tRNA of claim 40, wherein the split site is between residues D15 & D16, D15 & D17, D15 & D18, D16 & D17, D16 & D18, D17 & D18, A31 & A32, A31 & A33, A32 & A33, V45 & V46, V45 & V47, V45 & V48, V46 & V47, V4640 & V48, V47 & V48, T56 & T57, T56 & T58, or T57 & T58.

42. The split tRNA of claim 41, wherein the split site is between residues D15 & D17, D15 & D18, D16 & D17, A31 & A33, A32 & A33, V45 & V48, V46 & V48, V47 & V48, T56 & T57, or T56 & T58.

43. The split tRNA of any one of claims 40 to 42, wherein the first and second tRNA portion are from a tRNATrpor tRNAPro.

44. The split tRNA of any one of claims 34 to 43, wherein the second chain comprises a sequence, located 5’ to the second stem region and the second tRNA portion, encoding a polypeptide-of-interest or nucleic-acid-of- interest.

45. An RNA molecule comprising a sequence encoding a polypeptide-of-interest or nucleic-acid-of-interest and comprising a portion of a tRNA.

46. The RNA molecule of claim 45, wherein the portion of a tRNA corresponds to the 3’ portion of a parental tRNA split: at the anticodon; in the D loop, in the anticodon loop and to the 5’ side of the anticodon, in the variable loop, or in the T loop; between residues D15 & D16, D15 & D17, D15 & D18, D16 & D17, D16 & D18, D17 & D18, A31 & A32, A31 & A33, A32 & A33, V45 & V46, V45 & V47, V45 & V48, V46 & V47, V4640 & V48, V47 & V48, T56 & T57, T56 & T58, or T57 & T58; or between residues D15 & D17, D15 & D18, D16 & D17, A31 & A33, A32 & A33, V45 & V48, V46 & V48, V47 & V48, T56 & T57, or T56 & T58.

47. The RNA molecule of claim 45 or claim 46, wherein the portion of a tRNA is from a tRNAPyl, tRNALeu, tRNAAla, tRNASer, tRNATrp,or tRNAPro.

48. The RNA molecule of any one of claims 45 to 47, wherein the portion of a tRNA comprises a sequence encoded by any one of SEQ ID NOs: 1, 23, 25, 27, 46, 48, 50, 52, 54, 56, 58, 60, or 62.

49. The RNA molecule of any one of claims 45 to 48, wherein the RNA molecule comprises a stem region located at the 5’ end of the portion of a tRNA, and wherein the sequence encoding the polypeptide-of-interest or nucleic-acid-of-interest is located 5’ to the stem region.

50. The RNA molecule of claim 49, wherein the stem region is: 8-14, 8-12, or 10-12 nucleotides in length; or 8-25, 10-23, 12-22, 13-21, 14-20, 15-19, or 16-18 nucleotides in length.

51. The RNA molecule of claim 49 or claim 50, wherein the stem region is encoded by SEQ ID NO: 4, SEQ ID NO: 5, SEQ ID NO: 8, SEQ ID NO: 64, or SEQ ID NO:

66.

52. A method of determining the acylation status of a tRNA or efficiency of acylation of a tRNA, the method comprising: i) incubating the tRNA and a substrate with which the tRNA could be acylated, under conditions conducive to acylation of the tRNA; ii) exposing the tRNA to conditions capable of blocking the 3’ end of free tRNAs; and iii) exposing the tRNA to conditions that lead to the addition of nucleotides to the 3’ end of tRNAs that are not blocked, wherein at least one nucleotide comprises a label.

53. The method according to claim 52, wherein, prior to step iii), the method comprises exposing the tRNA to conditions capable of removing a substrate that has acylated a tRNA.

54. The method according to claim 52 or claim 53, wherein the label is, or is capable of being linked to, an optically detectable moiety or a physically detectable moiety.

55. The method according to claim 54, wherein the optically detectable moiety is a fluorescent moiety.

56. The method according to claim 54, wherein the physically detectable moiety is a magnetic moiety or is a ligand or receptor of a ligand-receptor pair.

57. The method according to any one of claims 52 to 56, wherein the method comprises capturing the tRNA if it has been labelled.

58. The method according to any one of claims 52 to 57, wherein the tRNA is a split tRNA according to any one of claims 34 to 44 or comprises an RNA molecule of any one of claims 45 to 51.

59. The method according to any one of claims 52 to 58, wherein the method is performed in a cell and the tRNA does not comprise an anticodon.

60. A tRNA comprising additional nucleotides at the 3’ end, wherein at least one nucleotide comprises a label.

61. The tRNA of claim 60, wherein the label is, or is capable of being linked to, an optically detectable moiety or a physically detectable label.

62. The tRNA of claim 61, wherein the optically detectable moiety is a fluorescent moiety.

63. The tRNA of claim 61, wherein the physically detectable moiety is a magnetic moiety or a ligand or receptor of a ligand-receptor pair.

64. The tRNA according to any of one claims 60 to 63, wherein the tRNA is a split tRNA according to any one of claims 34 to 44 or comprises an RNA molecule of any one of claims 45 to 51.

65. A nucleic acid encoding the split tRNA of any one of claims 34 to 44 or the RNA molecule of any one of claims 45 to 51.

66. A nucleic acid comprising, from 5’ to 3’: a second stem-region-encoding sequence, a second tRNA-portion-encoding sequence, a sequence encoding a loop, a first tRNA-portion-encoding sequence, and a first stem-region-encoding sequence, optionally wherein the first stem-region-encoding sequence and second stem-region-encoding sequence encode complementary stem region sequences.

67. The nucleic acid of claim 66, wherein the nucleic acid comprises, from 5’ to 3’: a sequence encoding a polypeptide-of-interest or nucleic-acid-of-interest, optionally a sequence encoding a linker, the second stem-region-encoding sequence, the second tRNA-portion-encoding sequence,the sequence encoding a loop, the first tRNA-portion-encoding sequence, and the first stem-region-encoding sequence.

68. The nucleic acid of claim 66 or claim 67, wherein: the second tRNA-portion-encoding sequence encodes a second tRNA portion corresponding to the 3’ portion of a parental tRNA split at the anticodon and the first tRNA-portion-encoding sequence encodes a first tRNA portion corresponding to the 5’ portion of the parental tRNA split at the anticodon; or the second tRNA-portion-encoding sequence encodes a second tRNA portion corresponding to the 3’ portion of a parental tRNA split at a split site and the first tRNA-portion-encoding sequence encodes a first tRNA portion corresponding to the 5’ portion of the parental tRNA split at a split site, wherein the split site is: in the D loop, in the anticodon loop and to the 5’ side of the anticodon, in the variable loop, or in the T loop; between residues D15 & D16, D15 & D17, D15 & D18, D16 & D17, D16 & D18, D17 & D18, A31 & A32, A31 & A33, A32 & A33, V45 & V46, V45 & V47, V45 & V48, V46 & V47, V46 40 & V48, V47 & V48, T56 & T57, T56 & T58, or T57 & T58; or between residues D15 & D17, D15 & D18, D16 & D17, A31 & A33, A32 & A33, V45 & V48, V46 & V48, V47 & V48, T56 & T57, or T56 & T58.

69. The nucleic acid of any one of claims 66 to 68, wherein the sequence encoding the loop comprises a sequence that is removed, spliced, or cleaved during RNA processing.

70. The nucleic acid of any one of claims 66 to 69, wherein the sequence encoding the loop comprises or is according to any one of SEQ ID NOs: 15 to 22 or TCGTCCT.

71. The nucleic acid of any one of claims 66 to 70, wherein the sequence encoding the loop comprises or is according to SEQ ID NO: 17 or SEQ ID NO:

19.

72. A method of making a polypeptide or nucleic acid, wherein the method comprises: i) providing a library comprising a plurality of sequences encoding polypeptides-of-interest or nucleic- acids-of-interest, wherein each sequence within the library is linked to a portion of a tRNA; ii) incubating, under conditions conducive to acylation, each polypeptide-of-interest or nucleic-acid-of- interest with a tRNA comprising the tRNA portion linked to the sequence encoding the respective polypeptide-of- interest or nucleic-acid-of-interest, and wherein the incubation includes a substrate with which the tRNAs could be acylated; iii) exposing the tRNAs to conditions capable of labelling tRNAs that have been acylated; iv) identifying whether each polypeptide-of-interest or nucleic-acid-of-interest is associated with a labelled tRNA; and v) making a polypeptide according to the sequence of an identified polypeptide-of-interest or nucleic-acid- of-interest.

73. A method of making a polypeptide, wherein the method comprises: i) providing a sequence of a polypeptide-of-interest identified by the screening method of any one of claims 1 to 33 or claims 52 to 59, and ii) producing a polypeptide according to said sequence.

74. The method of claim 72 or claim 73, wherein the polypeptide is an acyl-tRNA synthetase.

Images