ENHANCED PLATFORMS FOR UNNATURAL AMINO ACID INCORPORATION IN MAMMALIAN CELLS

Abstract

The present invention involves the ability to 1) use a virus assisted directed evolution platform to significantly improve the activity of engineered nonsense-suppressor tRNAs in mammalian cells, 2) provide mutants of archaeal pyrrolysyl and E. coli leucyl tRNAs that show remarkably improved Uaa incorporation efficiency in mammalian cells, and 3) use these tRNAs to express recombinant proteins in mammalian cells incorporating Uaas at significantly improved yields.

Claims

1. A composition comprising a variant archaeal or bacterial suppressor tRNA, wherein the variant tRNA has increased activity to incorporate an unnatural amino acid into a mammalian protein relative to its wild type counterpart tRNA.

2. The composition of claim 1, wherein the activity of the variant tRNA is increased over the wild type tRNA by about 2.5 to 80-fold.

3-4. (canceled)

5. The composition of claim 2, wherein the variant tRNA is a pyrrolysyl tRNA (tRNA.sup.Pyl) derived from SEQ ID NO: 1.

6. The composition of claim 5, wherein the variant tRNA.sup.Pyl comprises a sequence selected from the group consisting of: SEQ ID NOS: 2-27, or a nucleic acid sequence with at least 90% sequence identity with the full-length sequence of any of SEQ ID NOS: 2-27.

7. (canceled)

8. The composition of claim 2, wherein the variant tRNA is a leucyl tRNA (tRNA.sup.Leu) derived from SEQ ID NO: 28.

9. The composition of claim 8, wherein the variant tRNA.sup.Leu comprises a SEQ ID selected from the group consisting of SEQ ID NOS: 29-45, or a nucleic acid sequence with at least 90% sequence identity with the full-length sequence of any of SEQ ID NOS: 29-45.

10-11. (canceled)

12. A viral vector comprising the composition of claim 1.

13-23. (canceled)

24. A cell comprising the viral vector of claim 12.

25. The cell of claim 24, wherein the cell is a mammalian cell.

26. The mammalian cell of claim 25, wherein the cell further comprises plasmids encoding: a) a protein essential for viral replication, wherein a nonsense codon is inserted into the protein sequence rendering viral replication dependent on the activity of the variant suppressor tRNA; b) a cognate Uaa RNA Synthetase (UaaRS); and c) genetic components required for viral replication.

27-29. (canceled)

30. A method of virus-assisted directed evolution of orthogonal suppressor tRNA variants of interest with increased biological activity relative to the wild type suppressor tRNA, the method comprising the steps of: a) encoding a library of suppressor tRNA variants of interest in a virus genome; b) infecting a population of mammalian host cells with the virus vectors at low multiplicity of infection (MOI) and maintaining the population of cells under conditions suitable for virus replication in the cells, wherein virus replication in mammalian cells requires expression of an essential protein dependent on the activity of the tRNA variant of interest; and c) harvesting and selectively amplifying the virus progeny encoding active tRNA variants to remove cross-reactive tRNA molecules, whereby orthogonal suppressor tRNA variants with increased biological activity are recovered.

31-38. (canceled)

39. A method of producing a protein in a mammalian cell with one, or more, amino acid analogs at specified positions in the protein, the method comprising, a. culturing the mammalian cell in a culture medium under conditions suitable for growth, wherein the cell comprises a nucleic acid that encodes a protein with one, or more, selector codons, wherein the cell further comprises a variant archaea-derived pyrrolysyl tRNA with increased biological activity that recognizes the selector codon and its cognate aminoacyl-RNA Synthetase, and b. contacting the cell culture medium with one, or more, lysine analogs under conditions suitable for incorporation of the one, or more, lysine analogs into the protein in response to the selector codon, thereby producing the protein with one, or more lysine analogs.

40-46. (canceled)

47. A method of site-specifically incorporating one, or more, pyrrolysyl residues into a protein or peptide in a cell, the method comprising, a. culturing the cell in a culture medium under conditions suitable for growth, wherein the cell comprises a nucleic acid that encodes a protein or peptide of interest with one, or more, amber, ochre or opal selector codons at specific sites in the protein or peptide, wherein the cell further comprises a variant archaea-derived pyrrolysyl-tRNA.sup.Pyl with increased biological activity that recognizes the selector codon, and further comprises an archaea PylRNA Synthetase; b. contacting the cell culture medium with one, or more, pyrrolysl residues under conditions suitable for incorporation of the one, or more, pyrrolysyl residues into the protein or peptide at the sites of the selector codon(s), thereby producing the protein or peptide of interest with one, or more site-specifically incorporated pyrrolysyl residues.

48. The method of claim 47, wherein the variant pyrrolysyl tRNA (tRNA.sup.Pyl) is derived from SEQ ID NO: 1.

49-50. (canceled)

51. A method of site-specifically incorporating one, or more, leucine analog residues into a protein or peptide in a cell, the method comprising, a. culturing the cell in a culture medium under conditions suitable for growth, wherein the cell comprises a nucleic acid that encodes a protein or peptide of interest with one, or more, amber, ochre or opal selector codons at specific sites in the protein or peptide, wherein the cell further comprises a variant E. coli-derived tRNA.sup.Leu with increased biological activity that recognizes the selector codon, and further comprises an E. coli LeuRNA Synthetase; b. contacting the cell culture medium with one, or more, leucine analog residues under conditions suitable for incorporation of the one, or more, leucin analog residues into the protein or peptide at the sites of the selector codon(s), thereby producing the protein or peptide of interest with one, or more site-specifically incorporated leucine analog residues.

52. The method of claim 51, wherein the variant tRNA is a leucyl tRNA (tRNA.sup.Leu) derived from SEQ ID NO: 28.

53-61. (canceled)

62. A mammalian cell with a stably integrated variant tRNA-Pyl or tRNA-Leu for Uaa incorporation.

63-66. (canceled)

67. An engineered mammalian cell that comprises less than 250, 200, 150, 100, 75, 50 copies of a gene encoding a variant suppressor tRNA capable of incorporating an unnatural amino acid into a protein of interest.

68-70. (canceled)

71. The cell of claim 67, wherein the variant tRNA-Leu is selected from the group consisting of SEQ ID NOS: 29-45 and the Uaa is a leucine analog.

72. The cell of claim 71, wherein the leucine analog is any of structures 7-12.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In the accompanying drawings, reference characters refer to the same parts throughout the different views. The drawings are not necessarily to scale; emphasis has instead been placed upon illustrating the principles of the invention. The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawings(s) will be provided by the Office upon request and payment of the necessary fee. Of the drawings:

[0040] FIG. 1a-b shows the VADER selection scheme. a, Mammalian cells are infected with AAV2 encoding the tRNA library at low MOI. Plasmids encoding TAG-mutant of Cap, other genetic components needed for AAV replication, and the cognate aaRS are provided in trans by transfection in the presence of a suitable azido-Uaa. Active and orthogonal tRNA mutants facilitate generation of packaged progeny AAV2 incorporating the Uaa into their capsid, which are isolated by chemoselective biotin conjugation followed by streptavidin pulldown. b, Two AAV2 vectors, encoding i) E. coli tRNA.sup.Tyr and EGFP (Tyr-EGFP), and ii) tRNA.sup.Pyl and mCherry (Pyl-mCherry), were mixed in 10.sup.4:1 ratio and subjected to the VADER selection scheme using MbPylRS and its substrate AzK. FACS analysis of the surviving population show >30,000 fold cumulative enrichment of PylmCherry. Data shown as mean±s.d. (n=3 independent experiments).

[0041] FIG. 2a-b shows Directed evolution of tRNACUA.sup.Pyl. a, The sequences randomized to create four different libraries (A1, A2, T1, T2) of tRNACUA.sup.Pyl are highlighted in four different colors. (FIG. 2a discloses SEQ ID NO: 1) b, Analysis of tRNA sequences emerging from the selection of each library. c, Efficiency of TAG suppression for each of the unique fully base-paired tRNA.sup.Pyl selectants measured using an EGFP-39TAG reporter. The tRNA encoded in the pAAV plasmid (also harboring a wild-type mCherry reporter) was cotransfected into HEK293T cells with MbPylRS and EGFP-39TAG in the presence or absence of 1 mM AzK. Expression of EGFP-39TAG facilitated by each tRNA.sup.Pyl mutant was measured in cell-free extract, normalized relative to wild-type mCherry expression and plotted as a percentage of the reporter expression facilitated by wild-type tRNA.sup.Pyl. Data shown as mean±s.d. (n=3 independent experiments).

[0042] FIG. 3a-c shows improved efficiency of A2.1 tRNACUA.sup.Pyl. a, Sequences of wildtype (WT) and A2.1. b, Expression of EGFP-39TAG using the WT or A2.1 tRNA, with WT or AcK-selective MbPylRS, in the presence (+) and absence (−) of the appropriate Uaa. c, Expression of EGFP-39TGA and EGFP-39TAA using tRNAUCA.sup.Pyl and tRNAUUA.sup.Pyl (for both WT and A2.1 mutant), respectively, and MbPylRS in the presence or absence of AzK. Expression of EGFP-39TAG is measured in HEK293T cell-free extract and reported relative to its wild-type counterpart. Data shown as mean±s.d. (n=3 independent experiments).

[0043] FIG. 4a-b shows that single AAV2-encoded tRNA gene can facilitate the expression of TAG-inactivated capsid gene (Cap) and the production of progeny virus. a, Scheme of the experiment. HEK293T cells are infected with AAV2 encoding a tRNACUA.sup.Pyl and a wild-type EGFP gene at a very low MOI, then further transfected with plasmids encoding: i) AAV2 Rep and Cap-454-TAG genes, ii) MbPylRS, and iii) AdHelper in the presence or absence of 1 mM AzK. The feasibility of packaging AAV2 incorporating AzK at the 454 position of Cap, and that it does not perturb the virus, have been previously demonstrated Suppression of the TAG codon at 454 position of Cap leads to AzK incorporation into all three overlapping capsid proteins, VP1, VP2 and VP3 (60 total copies), at a surface exposed site. An identical experiment in which Cap-454-TAG is replaced by a wild-type Cap, was also performed. After 48 hours, the progeny virus was harvested from these cells and titered by infecting freshly seeded HEK293T cells, followed by their FACS analysis. b, AzK-dependent production of progeny virus is observed when Cap-454-TAG is used (magnified in the inset); the efficiency is significantly lower than the identical experiment where wild-type Cap is used instead. Data shown as mean±s.d. (n=3 independent experiments).

[0044] FIG. 5a-c shows AAV2-454-AzK can be isolated by bioorthogonal attachment of a photo-cleavable DBCO-biotin conjugate followed by streptavidin binding and photorelease. a, Structure of the photocleavable DBCO-biotin conjugate. b, AM/2-454-AzK was treated with different concentrations of the DBCO-biotin for 1 hr, the reaction was quenched using excess AzK, and the small molecules were removed by dialysis. The biotin-labeled virus was captured using streptavidin-agarose, then released by 365 rim irradiation. The infective titer of the virus was measured (by infecting HEK293T cells followed by FACS) before treatment, after biotin modification, and after photo-release, then normalized for volume change. Infectivity relative to untreated virus was plotted. We find that a low degree of biotinylation (at 5 μM reagent; each virion harbors 60 AzK residues) does not affect the AAV2 infectivity, but increased modification of the capsid (at higher DBCO-biotin concentrations) does. Also, using 5 μM DBCO-biotin, modified virus is recovered from streptavidin resin with good efficiency (˜30%), whereas the yield is poor when higher reagent concentration is used. c, Using this optimized labeling/capture strategy, AAV2-454-AzK encoding an mCherry reporter can be enriched from its mixture with an EGFP-encoding AAV2 with wild-type capsid (no azide). Fluorescence microscopy images of HEK293T cells infected with the mixed virus population before and after the selection are shown.

[0045] FIG. 6 shows Schematic maps of AAV2 cargoes containing various tRNAs and fluorescent proteins used in this study.

[0046] FIG. 7 shows representative microscopy images of cells infected with the mixed. virus population (Tyr-EGFP:Pyl-mCherry) before selection, after step 1, and step 2. Merged images from the EGFP and mCherry channels are shown for each. The ratios of the two viruses (Pyl-mCherry:Tyr-EGFP) as measured by FACS for these experiments are shown below.

[0047] FIG. 8a-c shows representative fluorescence microscopy images of HEK293T cells expressing nonsense-inactivated EGFP reporters, suppressed using the wild-type tRNA.sup.Pyl or its most efficient evolved mutant, A2.1. a, Co-transfection of wild-type or A2.1 tRNA.sup.Pyl, encoded in the pAAV plasmid (also encoding a wild-type mCherry reporter) with MbPylRS and EGFP-39-TAG reporter in the presence or absence of 1 mM AzK. Expression of mCherry is shown as a control in each experiment. b, Expression of EGFP-39TGA and EGFP-39TAA using tRNAUCAPyl and tRNAUUA.sup.Pyl (for wild-type and A2.1 mutant), respectively. The tRNAs encoded in the pIDTsmart vector were co-transfected with the appropriate EGFP mutant and MbPylRS in the presence or absence of 1 mM AzK. c, Incorporation of AcK into EGFP-39TAG using wild-type and A2.1 tRNA.sub.CUA.sup.Pyl. The tRNAs encoded in the pIDTsmart vector was co-transfected with EGFP-39TAG mutant and MbPylRS-AcKRS3 mutant, in the presence or absence of 5 mM AcK.

[0048] FIG. 9 shows sequences of selected tRNAs that are fully base-paired (including G-2U wobble pairing) from each of the four libraries. See FIG. 2c for the characterization of their activity. The numbering scheme for the tRNA is shown below.

[0049] FIG. 10a-c shows the improved activity of a mutant leucyl tRNA obtained through the VADER selection scheme. a, shows the sequence of wild-type TAG-suppressor E. coli leucyl tRNA (EcLtR) (SEQ ID NO: 47-which is identical to SEQ ID NO:28 but with T not U), b, shows the sequence of one of the improved mutants of EcLtR (EcLtRh1) (SEQ ID NO: 48-identical to SEQ ID NO:30 but with T not U) identified by the methods described herein, c, shows the activity of EcLtR and EcLtR-h1 in HEK293T cells, when co-transfected with an engineered EcLeuRS mutant that selectively charges Uaas. Expression of a fill-length EGFP-39-TAG reporter used to measure the activity of the tRNAs. EcLtR-h1 shows remarkably high efficiency.

[0050] FIG. 11 shows the sequence and the secondary structure of the wild-type pyrrolysyl tRNA (SEQ ID NO:1), and further shows the custom randomization targeted to each position in the acceptor stem to produce.

[0051] FIG. 12 shows the results of using next-generation Illumina DNA sequencing (NGS) to characterize the degree of enrichment for each mutant in a tRNA library, when subjected to VADER selection scheme. The resulting enrichment factors can be used to estimate the efficiency of the corresponding tRNA.

[0052] FIG. 13 shows the results of evaluation assays for improved activity of selected tRNA-Pyl mutants demonstrated using the expression of EGFP-39TAG reporter in HEK293T cells (as described earlier), normalized relative to wild-type mCherry expression, and plotted as a percentage of the reporter expression facilitated by wild-type tRNA-Pyl.

[0053] FIG. 14a-b The top panel (14a) shows the compiled sequences of the acceptor stem region of 120 different bacterial leucyl-tRNAs in the Weblogo format (https://weblogo.berkeley.edu/). In this format, the relative abundance of a particular nucleotide found at a particular position within this set of tRNA sequences is represented by the relative height of the corresponding letter code. The bottom panel (14b) shows the sequence and the secondary structure of the wild-type E. coli leucyl-tRNA (SEQ ID NO: 49/SEQ ID NO:28), and further shows the custom randomization targeted to each position in the acceptor stem guided by the sequence alignment.

[0054] FIG. 15 shows the results of evaluation assays for improved activity of selected tRNA-Leu mutants demonstrated using the expression of EGFP-39717AG reporter in HEK293T cells (as described earlier), normalized relative to wild-type mCherry expression, and plotted as a percentage of the reporter expression facilitated by wild-type tRNA-Leu.

[0055] FIG. 16a-b shows the results of comparing the activities of WT tRNA-Pyl and tRNA-Pyl-2 (SEQ :ID NO: 2) under controlled tRNA expression through baculovirus delivery. HEK293T cells were transduced with a baculovirus encoding MbPylRS and EGFP-39-TAG with a fixed MOI of 1, along with an increasing MOI (0.3 to 3) of a second baculovirus encoding an mCherry reporter and either the WT or engineered (SEQ ID NO: 2) tRNA-Pyl. All expressions except the no-AzK control are performed in the presence of 1 mM AzK, an Uaa substrate for MbPylRS. Expression of the mCherry is shown in the bottom panel, which increases linearly as increasing MOI of tRNA-mCherry virus is used and is comparable for WT and engineered tRNA-Pyl virus at the same MOI. The top panel shows EGFP-39-TAG expression relative to an identical virus encoding wild-type EGFP at the same MOI. The engineered tRNA-Pyl facilitates EGF-39-TAG expression at a much lower expression level (lower MOI) relative to the WT tRNA-Pyl.

[0056] FIG. 17a-b shows the results of comparing the activities of WT tRNA-Leu and tRNA-Leu-30 (SEQ ID NO: 30) under controlled tRNA expression through baculovirus delivery. HEK293T cells were transduced with a baculovirus encoding EcLeuRS and EGFP-39-TAG with a fixed MOI of 1, along with an increasing MOI (3 to 15) of a second baculovirus encoding an mCherry reporter and either the WT or engineered (SEQ ID NO: 30) tRNA-Leu. All expressions except the no-Uaa control are performed in the presence of 1 mM Cap, a Uaa substrate for EcLeuRS. Expression of the mCherry reporter is shown in the bottom panel, which increases linearly as increasing MOI of tRNA-mCherry virus is used and is comparable for WT and engineered tRNA-Leu virus at the same MOI. The top panel shows EGFP-39-TAG expression relative to an identical virus encoding wild-type EGFP at the same MOI. The engineered tRNA-Leu facilitates EGF-39-TAG expression at a much lower expression level (lower MOI) relative to the WT tRNA-Leu.

[0057] FIG. 18 shows structures of Uaas used in the methods described herein.

[0058] FIG. 19a-b shows the results of assays evaluating the improved tRNA-Pyl (SEQ ID NO: 2; top panel) and tRNA-Leu (SEQ ID NO: 30; bottom panel) facilitates more efficient incorporation of various Uaas shown FIG. 18.

[0059] FIG. 20 shows the amino acid sequence (SEQ ID NO: 46) of the AAV2 VPI capsid protein.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0060] The present invention describes a novel strategy, virus-assisted directed evolution of tRNA (VADER) in mammalian cells, to obtain highly efficient active, orthogonal tRNA variant molecules. The tRNA variants as described herein, produced by the methods described herein, are characterized by increased (enhanced) biological activity of nonsense codon suppression and incorporation of unnatural amino acids in a site-specific manner in proteins of interest. The methods to select for these highly efficient tRNA variants couple the activity of the suppressor tRNA to the replication of a human virus (e.g., Adeno-associated virus, or AAV). In certain embodiments, the method comprises: i) encoding the library of tRNA variants in the virus genome to enable its controlled delivery to mammalian cells; ii) inserting a nonsense codon in an essential virus protein to render viral replication dependent on the activity of the suppressor tRNA, facilitating selective amplification of virions encoding active tRNA variants; and iii) the enriched tRNA sequences can be readily retrieved by isolating and sequencing the genome of the freshly amplified virus.

[0061] The methods of the present invention specifically demonstrate the ability to enrich an AAV population encoding an active suppressor tRNA relative to AAV population encoding an inactive tRNA in the range of about 10,000 to 50,000-fold and is typically about >30,000 fold, thus providing a powerful selection scheme to enrich active mutants from a naïve tRNA library. In particular, there is a 2.5-fold to 80-fold increase in activity in the variant tRNAs identified and isolated by the methods described herein.

[0062] Next generation sequencing (such techniques are known to those of skill in the art-see for example, the kits/reagents commercially available from Illumina) of the virus-encoded tRNA library before and after the VADER selection method described herein can be performed to evaluate and confirm enrichment of each possible mutant/variant in the library.

[0063] The technology of the present invention can be further applied to evolve different suppressor tRNAs commonly used for Uaa incorporation in mammalian cells, including, for example, the archaea-derived pyrrolysyl tRNA, and the E. coli derived leucyl tRNA, Subjecting synthetic mutant libraries of both tRNAs resulted in the identification of variants that demonstrate significantly improved activity for Uaa incorporation in mammalian cells. The mutants show particularly improved efficiency relative to their wild-type counterparts when expressed at a lower level, further confirming their enhanced intrinsic efficiency.

[0064] As a result of the present invention, a general strategy to evolve the efficiency of any engineered suppressor tRNA for Uaa incorporation in mammalian cells is now available. Encoding the tRNA library in a viral genome and subjecting the resulting library to the VADER selection scheme will enable selective enrichment of those that encode active tRNA mutants.

[0065] As a result of the present invention, methods are now available that can also be used to evolve the efficiency of other biological parts in mammalian cells, if its activity can be coupled to the expression of AAV capsid proteins. Such biological parts include, but are not limited to, promoter elements, internal ribosomal entry sites (IRES), novel transcription factors, receptor proteins (e.g., GPCR), gene or mRNA editing proteins (e.g., Cas/CRISPR), mammalian two-hybrid systems, etc.

[0066] The mutant suppressor tRNAs (e.g., pyrrolysyl and leucyl) generated through the VADER selection scheme as described herein enable highly efficient Uaa incorporation in mammalian cells. These tRNA mutants can be used to improve the yields of Uaa-incorporated protein in mammalian cells (e.g., antibodies and other therapeutically related proteins).

[0067] The improved suppressor tRNAs (e.g., pyrrolysyl and leucyl) generated through the VADER selection scheme can be used to create improved expression vectors (e.g., viral vectors) that deliver the genetic machinery for Uaa incorporation into mammalian cells and tissues. Importantly, as these tRNAs are more efficient, fewer copies of tRNA variant need be encoded per genome. Currently, including multiple tRNA copies (to achieve high enough expression of tRNAs) often leads to genome instability of expression vectors (e.g., viral vectors).

[0068] The improved suppressor tRNAs (e.g., pyrrolysyl and leucyl) generated through the VADER selection scheme can be used to create stable cell lines for protein expression incorporating Uaas. Currently, the requirement of encoding a large number of tRNA copies per genome makes it challenging to encode the UNA-incorporation machinery stably in the mammalian genome. The increased efficiency of new tRNAs will allow the use of much fewer copies.

[0069] The present invention establishes a unique virus-assisted directed evolution platform in mammalian cells capable of improving the activity of tRNAs and other biological parts for biotechnology applications. It also describes suppressor tRNA variants that demonstrate significantly improved activity in mammalian cells.

[0070] While this invention has been particularly shown and described with references to preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention encompassed by the appended claims.

[0071] Without further elaboration, it is believed that one skilled in the art can, based on the above description, utilize the present invention to its fullest extent. The following specific embodiments and examples are, therefore, to be construed as merely illustrative, and not limitative of the remainder of the disclosure in any way whatsoever.

EXAMPLES

[0072] The following examples are provided to illustrate embodiments of the present invention but are by no means intended to limit its scope.

[0073] The examples described herein will be understood by one of ordinary skill in the art as exemplary protocols. One of ordinary skill in the art will be able to modify the below procedures appropriately and as necessary.

[0074] Materials and Methods

[0075] Cell culture. HEK293T cells (ATCC) were maintained at 37° C. and 5% CO2 in DMEM-high glucose (HyClone) supplemented with penicillin/streptomycin (HyClone, final concentration of 100 U/mL penicillin and 100 μg/mL streptomycin) and 10% fetal bovine serum (Coming). All references to DMEM below refer to the complete medium described here.

[0076] General cloning. For all cloning, the E. coli TOP10 strain was used for transformation and plasmid propagation and bacteria were grown using LB for solid and liquid culture. All PCR reactions were carried out using Phusion Hot Start II DNA Polymerase (Thermo Scientific) according to the manufacturer's protocol. Restriction enzymes and T4 DNA ligase were from New England Biolabs (NEB). All DNA oligos were purchased from Integrated DNA Technologies (IDT). Sanger sequencing was performed by Eton Bioscience.

[0077] Unnatural amino acids Azido-lysine (AzK) was purchased from Iris Biotech GMBH (Germany). Nε-acetyllysine (AcK) was purchased from Bachem.

[0078] Packaging and titration of mock and library tRNAs into AAV (wild-type capsid). To package various cargo into AAV-2, 8 million HEK293T cells were seeded in a 10 cm tissue culture dish. The following day, the cells were transfected with 8 μg each of the appropriate cargo plasmid (pAAV-ITR-tRNA-fluorescent protein), pHelper, and pAAV-RC2 using polyethylenimine (PEI) (Sigma). Media was exchanged for fresh DMEM 24 hours after transfection. 72 hours after transfection, the cells were resuspended, pelleted, and lysed by freeze/thawing as previously described. 1 Virus was concentrated and semi-purified by PEG precipitation, 1 resuspended in 1 mL DMEM with FBS and flash frozen.

[0079] Sequences Described in the Examples

[0080] Wild type and derived sequences described in the Examples and throughout the application are listed in the Table:

TABLE-US-00001 tRNA- SEQ gGAAACCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggGGUUU Pyl WT ID 1 Ccgcca Pyl hits SEQ. gGGCGGCugaucauguagaucgaacggacucuaaauccguucagecggguuagauucccggGCUGC ID 2 Ccgcca SEQ gGGUGACugaucauguagaucgaacggacucuaaauccguucagecggguuagauucccggGUUG ID 3 CCcgcca SEQ gGGGGGCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggGCUCC ID 4 Ccgcca SEQ gGGCGGCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggGUUGC ID 5 Ccgcca SEQ gGGCGCCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggGGCGC ID 6 Ccgcca SEQ gGGGAGGugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggCCUCC ID 7 Ccgcca SEQ gGGGACCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggGGUCC ID 8 Ccgcca SEQ gGCCGGGugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggCCUGG ID 9 Ccgcca SEQ gGGGACCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggGGUCC ID 10 Ucgcca SEQ gGGGCCCugaucauguagaucgaacggacucuaaauccguucagecggguuagauucccggGGGUC ID 11 Ccgcca SEQ gGGGGCCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggGGCUC ID 12 Ccgcca SEQ gGGGUCCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucceggGGAUC ID 13 Ccgcca SEQ gGGGACCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggGGUUC ID 14 Ccgcca SEQ gGGGAGGugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggUCUU ID 15 CCcgcca SEQ gGGGGGGugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggCCCCC ID 16 Ucgcca SEQ gGUGGGGugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggCCCUA ID 17 Ccgcca SEQ gGGGGUCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggGACUC ID 18 Ccgcca SEQ gGGUCCCugaucauguagaucgaacggacucuaaauccguucagecggguuagauucccggGGGGU ID 19 Ccgcca SEQ gGGGACCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggGGUUC ID 20 Ucgcca SEQ gGGCGGCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggGCCGC ID 21 Ccgcca SEQ gAGCACCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggGGUGC ID 22 Ucgcca SEQ gGGGGGGugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggCCCCC ID 23 Ccgcca SEQ gAGGGGGugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggCCCCC ID 24 Ucgcca SEQ gGGAGCCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggGGUUC ID 25 Ccgcca SEQ gGGAGCCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggGGUUC ID 26 Ccgcca SEQ gAGGACCugaucauguagaucgaacggacucuaaauccguucagccggguuagauucccggGGUUC ID 27 Ucgcca LeuWT SEQ GCCCGGAugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguuc ID 28 aagucccgcUCCGGGUacca Leu hits SEQ GCCCGGAugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguuc ID 29 aagucccgcUCCGGGCacca SEQ GCCCGGAugguggaaucgguagacacaagggaCucuaaaucccucggcguucgcgcugugcggguuc ID 30 aagucccgcUCCGGGCacca SEQ GGGCGUGugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguuc ID 31 aagucccgcCGCGCCCacca SEQ GGGCGCGugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguuc ID 32 aagucccgcCGCGCCCacca SEQ GGGCAUGugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguuc ID 33 aagucccgcCAUGCCCacca SEQ GGGCACGugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguuc ID 34 aagucccgcCGUGCCCacca SEQ GGGGGUGugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguu ID 35 caagucccgcCGCCCCCacca SEQ GGGGGCGugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguuc ID 36 aagucccgcCGUGCCCacca SEQ GGGGAUGugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguu ID 37 caaguccegcCGUCCCCacca SEQ GGGGACGugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguuc ID 38 aagucccgcCGUGCCCacca SEQ GCCCGUAugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguuc ID 39 aagucccgcUGCGGGCacca SEQ GGGAUAGugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguuc ID 40 aagucccgcCUAUCCCacca SEQ GGGCAUGugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguuc ID 41 aagucccgcCGUGCCCacca SEQ GGGCAGAugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguuc ID 42 aagucccgcUCUGCCCacca SEQ GGGCGUAugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguuc ID 43 aagucccgcUGCGCCCacca SEQ GGGCAAGugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguuc ID 44 aagucccgcCGUGCCCacca SEQ GCACACAugguggaaucgguagacacaagggauucuaaaucccucggcguucgcgcugugcggguuc ID 45 aagucccgcUGUGUGCacca

Example 1: Positive Selection

[0081] 8 million HEK293T cells each were seeded in three 10 cm tissue culture dishes. The next day, the cells were infected with virus containing a tRNAPyl library at an apparent MOI of 5 (the actual MOI is substantially reduced in the presence of PEI, the transfection reagent). Four hours after infection, the cells were transfected with 22 μg of pHelper and 10 μg of pIDTSmart-RC2(T454TAG)-PylRS per dish using PEI. 1 mM AzK was also added at this point. One day after transfection the culture media was exchanged with fresh DMEM containing 1 mM AzK. Cells were harvested three days after transfection and lysed as for virus isolation. The culture media was saved and recombined with clarified lysate, and this mixture was treated with 500 U universal nuclease (Thermo Scientific) for 30 minutes. Virus was recovered by PEG precipitation using 11% polyethylene glycol (Fisher) as previously described 1 and resuspended in 3 mL PBS. The small-scale mock positive selections were carried out in 12-well plates. 0.7 million cells per well were seeded and infected the next day with AAV carrying a tRNAPyl-mCherry cargo. Four hours later the cells were transfected as described for the above selections, but with the transfection mix and AzK scaled down by a factor of 15. For PEI only wells, cells received a comparable amount of transfection reagent but no plasmid. Media was changed the day after transfection for fresh DMEM containing 1 mM AzK. Virus was harvested three days post-transfection and PEG-precipitated as described for the selections above. Confluent cells in a 12-well plate were infected with the entire output of one mock selection well and analyzed by flow cytometry.

Example 2: Negative Selection

[0082] The virus from positive selection (3 mL) was labeled with photocleavable DBCO-sulfo-biotin (Jena Biosciences) at a concentration of 5 μM for one hour in the dark with mixing. Immediately after labeling, excess DBCO-biotin was quenched with AzK (1 mM final concentration) and the reactions were dialyzed overnight using Slide-A-Lyzer 100 kDa MWCO devices (Thermo Scientific) against 1 L PBS at 4° C. The dialyzed virus mixtures were split into three 2 mL tubes and each rotated overnight with 400 μL streptavidin agarose resin (Thermo Scientific) at 4° C. The next day, each tube of beads was washed eight times with 1 mL PBS containing additional NaCl (final concentration 300 mM) with mixing between washes. Finally, the washed beads were resuspended in 8 mL PBS (300 mM NaCl) and the virus was eluted from the resin via four 30-second irradiations using a 365 nm UV diode array (Larson Electronics), with mixing between irradiations.

Example 3: Viral DNA Recovery, Amplification, and Cloning

[0083] The eluted virus was concentrated from 3 mL to 300 μL using Amicon Ultra-4 100 kDa MWCO centrifugal concentrators (Millipore). This mixture was heated to 100° C. for 10 minutes in order to denature the viral capsid proteins and expose the DNA. Viral DNA was then cleaned up and concentrated by ethanol precipitation using yeast tRNA (Ambion) and resuspended in a final volume of 50 μL. 20 μL of this mixture was added to a 200 μL PCR reaction and amplified with tRNAAmp-F and R primers. The resulting DNA was digested with KpnI and NcoI and cloned into the library cloning vector using the same protocol as for original library generation.

Example 4: Mock Selections Using Pyl-mCherry and Tyr-GFP.

[0084] The mock selections shown in FIG. 1 followed the same protocol as above, except that the starting library virus was a 1:10,000 mixture of virus made from pAAV-ITR-PytR-mCherry to virus made from pAAV-ITR-EcYtR-GFP. Mock selection results were analyzed by flow cytometry as described for virus titering, but here cells in a 12-well plate were infected with 200 μL of the virus pool after either positive or negative selection. Red and green fluorescent cells were counted to determine the virus ratio.

Example 5: Hit Sequencing and Characterization

[0085] For each library, 30-50 colonies were picked from the transformation plates generated above and sent for Sanger sequencing (Eton Bioscience). All sequences in which all randomized bases were paired were treated as potential hits, and these tRNAs were subcloned into pAAV-ITR-PytR-mCherry for analysis. Initial hit analysis was conducted by transfecting HEK293T cells in 24-well plates with 0.5 μg each of a potential hit pAAV-ITR-PytR-mCherry plasmid, pIDTSmart-MbPylRS, and pAcBacl-GFP(39TAG) in the presence and absence of 1 mM AzK. Two days after transfection, cells were lysed with CelLytic M buffer (Sigma) and EGFP and mCherry fluorescence were measured on aMolecular Devices SpectraMax M5 microplate reader. Values for an untransfected well were subtracted, and EGFP-fluorescence was normalized to mCherry fluorescence for each well. The best hit, Ac2.1 (GGG/CCU), was selected for further analysis with other stop codons and a different synthetase and Uaa, AcKRS3 and AcK. HEK293T cells in a 12-well plate were transfected with 0.375 μg pIDTSmart-PytR containing either the wild-type or evolved tRNA, 0.375 μg pIDTSmartaaRS containing the appropriate synthetase, and 0.75 μg pAcBacl-EGFP containing one or two of the appropriate stop codons. A wild-type EGFP control well used pIDTSmart-PytR(TAG, wild-type), pIDTSmart-MbPylRS, and pAcBacl-EGFP(wild-type) in the same ratios. Two days after transfection, cells were lysed and EGFP fluorescence was measured by microplate reader. Values from an untransfected well were subtracted.

Example 6: Further Evolution of the Pyrrolysyl-tRNA Using Custom-Randomized Mutant Libraries

[0086] In the first-generation VADER experiments, only short segments of the tRNA (3 base pairs at one time) were randomized at one time to create small mutant libraries, which were subjected to selection. While it led to the identification of improved mutants, we surmised that the ability to randomize and select a larger sequence space may lead to the identification of even more efficient mutants. However, randomizing a larger segment of the tRNA exponentially increases the size of the library. For example, randomizing one additional base pair in the stem region increases the number of library members by 16 fold. Because of technical limitations, it is currently challenging to use our VADER platform to process a library size larger than 10.sup.5, while ensuring complete coverage of all possible mutants. This size limit restricts us to the complete randomization of no more than 4 base pairs in the stem region of a tRNA for engineering its activity. However, when a base pair in a tRNA-stem region is completely randomized, only 6 out of the 16 resulting mutants can still maintain the base pairing interaction (either A:T, G:C, or G:U), which is essential for the stability of the stem region. The majority of mutants that cannot base pair result in an unpaired ‘bubble’ in the middle of the tRNA stem, which typically compromises tRNA performance. A different way of synthesizing the tRNA library was envisioned, where each base pair is only randomized to the desirable base-paired sequences. This approach takes advantage of the recent advances in DNA synthesis technology, enabling the synthesis of a large number of distinct DNA oligonucleotides of significant length (up to 300 nucleotides). This enables the synthesis of a DNA library, encoding the entirety of the tRNA gene, where each position of each library member can be specified, making it possible to only include mutants that base pair and avoid those which do not.

[0087] Using the DNA synthesis service provider TWIST bioscience, a pyrrolysyl-tRNA library was created as depicted in the FIG. 11, where six base pairs in the acceptor stem were randomized to desired combinations of base-pairing sequences. The resulting library was packaged in AAV2 and was subjected to the VADER selection scheme in duplicate as described above. The AAV2-packaged library was sequenced using Illumina platform for next generation sequencing before and after subjecting it to VADER selection (FIG. 12). The enrichment of each mutant was calculated using its abundance before and after the selection step and the mutants were ranked based on the degree of enrichment. The mutants that exhibited the highest degree of enrichment upon selection were resynthesized and their activities were benchmarked using the EGFP-39-TAG expression assay as described before (FIG. 13). As shown in FIG. 13, 26 tRNA-Pyl mutants demonstrated activity that was at least 250% higher relative to wild-type tRNA-Pyl, with the most active mutant demonstrating 540% activity relative to WT-tRNA-Pyl.

Example 7: Evolution of E. coli Leucyl-tRNA for Enhanced Nonsense Suppression Activity in Mammalian Cells

[0088] Application of the VADER selection scheme for engineering tRNA activity in mammalian cells is not restricted only to the pyrrolysyl tRNA. It can be used to also improve the activity of other tRNAs that are suitable for Uaa. incorporation in mammalian cells. The use of the VADER methodology described herein was also used to improve the activity of E. coli leucyl tRNA (tRNA-Leu), which along with its cognate E. coli leucyl-tRNA synthetase (EcLeuRS) has been previously used for Uaa incorporation in mammalian cells (J. Am. Chem. Soc. 2004, 126, 14306; Blochemistry 2018, 57, 441). As shown with the work described herein on the tRNA-Pyl, engineering the acceptor stem is often very attractive, as this region interfaces with many components of the translation system. To design a ‘smart’ library, the sequences of 120 known bacterial tRNA sequences were aligned to generate a consensus sequence of the acceptor stem (FIG. 14). This consensus sequence could be used as a guide to predict which parts of the acceptor stem may be important in tRNA-aaRS interaction (identity element), and which regions offer room for alteration. Based on this approach, a custom-randomization library of the tRNA-Leu acceptor stem (FIG. 14) was designed. This library was packaged in AAV2 and subjected to VADER selection scheme as described above. A previously developed polyspecific EcLeuRS mutant (Biochemistry 2018, 57, 441), that can charge the azido-containing (azido-modified) Uaa AzK was used in the VADER scheme to charge the tRNA mutants. Next-generation Illumina DNA sequencing was used to measure the enrichment of each library member before and after the selection, as described above, and the ones exhibiting the most enrichment were resynthesized and characterized using the previously described EGFP-39-TAG reporter expression assay. As shown in FIG. 15, seventeen tRNA-Leu mutants demonstrated activity that was at least 1,000% higher relative to wild-type tRNA-Leu, with the most active mutant demonstrating approximately 13,000% enhanced activity relative to WT-tRNA-Leu. These tRNA sequences, including WT-tRNA-Leu, contains U at the 33 position, the first nucleotide in the anticodon loop. Mutation of this U to C can result in enhanced nonsense suppression activity, as it typically provides a better context for the nonsense suppressor anticodon. To find out if this is the case, the 33-U to C in mutant 29 (SEQ ID NO: 29) was mutated to create mutant 30 (SEQ ID NO: 30). Indeed, this mutant shows significantly improved suppression activity relative to 29 (FIG. 15). Introducing this mutation to other identified tRNA-Leu mutants (SEQ ID NOS: 31-45) should also result in an further improvement of their activity.

Example 8: The Engineered tRNA Mutants Show Further Improvement Relative to Their Wild-Type Counterparts when Their Expression Levels are Controlled Limited

[0089] So far, all evaluation of all tRNA activities were performed by transient transfection of plasmids encoding the tRNA, aaRS, and the reporter into mammalian cells. It is well-established that transient transfection of mammalian cell culture results in an uncontrolled and heterogeneous level of DNA delivery, such that some of the cells uptake and overexpress the associated plasmids at a very high level, while others do not. It was previously demonstrated that the resulting overexpression of the encoding tRNA and aaRS can compensate for their poor intrinsic activity and inflate the estimate of their inherent efficiency (ACS Synth. Bial. 2017, 6, 13). It is further demonstrated as described herein, hat, as a result, comparing two different Uaa incorporation systems by transient transfection may yield an inaccurate estimate, where the efficiency of the weaker system is overestimated. It was surmised that the difference in efficiency observed using the transient-transfection assay might be underestimating the actual degree of improvement of inherent efficiency of different tRNA mutants relative to their wild-type counterparts.

[0090] To overcome this challenge, a previously developed a baculovirus vector that facilitates controlled and more homogeneous delivery of transgenes to mammalian cells (ACS Synth. Biol. 2017, 6, 13) was used. The expression level of the transgene can be simply controlled by systematically altering the virus-to-cell ratio. Using this delivery system, it is possible to compare two different genetic systems for Uaa incorporation across a large spectrum of different expression levels, which more accurately reveals differences in their intrinsic performance. To compare the activity of the engineered tRNA-Pyl and tRNA-Leu mutants relative to their wild-type counterparts using this approach, baculovirus vectors were constructed that encode a wild-type mCherry reporter, as well as one of the four tRNAs: WT tRNA-Pyl, WT-tRNA-Leu, tRNA-Pyl-2 (SEQ ID 2), or tRNA-Leu-30 (SEQ ID 30). A second baculovirus was developed to deliver an EGFP-39-TAG reporter, as well as the necessary aaRS (MbPylRS for tRNA-Pyl, or EcLeuRS for tRNA-Leu). HEK293T cells were transduced with the aaRS/EGFP-39-TAG baculovirus with a fixed MOI (multiplicity of infection, or number of infective virus particles added per cell) of 1, along with an increasing MOI (0.3 to 15) of either the WT or engineered tRNA virus. Expression of the mCherry (confirming the delivery of the tRNA-baculovirus at desired level) and the EGFP-39-TAG (representing the Uaa incorporation efficiency in response to TAG) were recorded 48 hours post-transfection using their characteristic fluorescence in cell-free extract. As shown in FIG. 16, the engineered tRNA-Pyl facilitates EGF-39-TAG expression at a much lower expression level (lower MOI) relative to the WT tRNA-Pyl. For example, when WT tRNA-Pyl virus is used at MOI 1, expression of EGFP-39-TAG is only 0.5% with respect to the wild-type EGFP control; while the virus encoding engineered tRNA-Pyl (SEQ ID NO: 2) affords 9.3% EGFP-39-TAG expression at the same MOI, indicating a >18 fold higher efficiency of the latter at this expression level. At a higher MOI of 3, the engineered tRNA show approximately 14 fold higher activity relative to wild-type, underscoring how higher expression can underestimate true differences in intrinsic activity. We also compared the activities of wild-type tRNA-Leu and one of its engineered counterparts (SEQ ID NO: 30) in the same manner. As shown in FIG. 17, the engineered tRNA provide nearly 29-fold improved EGFP-39-TAG expression at the highest MOI tested (15). When the tRNAs were expressed at a lower level, the difference was even more stark; e.g., at MOI 5, the WT tRNA-Leu affords no detectable EGFP-39-TAG expression, while the tRNA-Leu-30 allows its expression at a 16% level relative to the wild-type reporter. That the engineered tRNAs provide significantly higher efficiency at lower expression levels is highly significant, since this will make it significantly easier to generate stable mammalian cell-lines with genomically integrated aaRS/tRNA that provide high Uaa incorporation efficiency.

Example 9: The Engineered tRNA Mutants More Efficient Incorporation of Numerous Uaas

[0091] Without being hound by theory, although the improved activity of the engineered tRNAs is not fully understood, it is likely that these interface with the mammalian translation system much better than their wild-type counterparts, which are borrowed from a different domain of life. Consequently, the improved activity of these tRNAs should enable more efficient incorporation of all Uaas which can be incorporated by an engineered mutant of its cognate aaRS. In order to demonstrate this hypothesis, several Uaas that can be incorporated using engineered MbPylRS (structures 1-6; FIG. 18) or EcLeuRS (structures 7-12; FIG. 18) were evaluated for their incorporation efficiency using the aforementioned EGFP-39-TAG expression assay. Indeed, each of these Uaas were incorporated by the engineered tRNAs at a significantly higher efficiency into the reporter by the two engineered tRNAs relative to their wild-type counterparts (FIG. 19).

[0092] While this invention has been particularly shown and described with references to preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention encompassed by the appended claims.