Chapter One - Selection and validation of orthogonal tRNA/synthetase pairs for the encoding of unnatural amino acids across kingdoms

Abstract

As an increasing number of protein structures are resolved at atomic and near-atomic resolution, conventional amino acid mutagenesis may be insufficient to test many mechanistic hypotheses. As a result, the development of new tRNA/aminoacyl-tRNA synthetase (aaRS) pairs has become an important tool for determining intricate molecular interactions and understanding protein structures. This chapter describes in detail the directed evolution of new tRNA/aaRS pairs in Escherichia coli for the incorporation of non-canonical amino acids (ncAA). Section 1 describes the selection of new tRNA/aaRS pairs in E. coli. Section 2 details the use of a synthetase to incorporate an ncAA into a mammalian cell line, and 1 Introduction, 2 Screening of the pyrrolysine synthetase library in both include methods on the determination of synthetase efficacy and fidelity.

1. Introduction

The release of increasing numbers of atomic and near-atomic structures of clinically relevant proteins has led to hypotheses that are often best answered via structure-guided mutagenesis. In this regard, conventional amino acids are limited in terms of the chemical precision with which specific attributes of side chain and main chain chemistry can be modified. This shortcoming ultimately limits the mechanistic resolution of laboratory hypothesis testing. To address this, several methods enabling the encoding of synthetic amino acids have been pursued, including nonsense suppression. Within this general method, variations have been developed including chemical misacylation and co-evolution of tRNA/synthetase pairs (Leisle, Valiyaveetil, Mehl, & Ahern, 2015). Of these, the latter utilizes standard DNA transfection of cDNA plasmids that encode the tRNA and synthetase. Therefore, this method is accessible to most investigators competent in standard molecular biology methods, and in theory it can be applied to any transducible cell type. The primary technical hurdles for the application of tRNA/synthetase pairs relate to the initial identification of synthetase/tRNA/AA combinations that are efficient and orthogonal in the biological system of interest. Ideally, the synthetase will enable the efficient encoding of a particular non-canonical amino acid (ncAA) or set of ncAAs with negligible spurious incorporation of natural amino acids.

Here, we describe in technical detail the approach we use to identify orthogonal synthetases that specifically and efficiently encode a broad range of ncAAs into proteins in both bacteria and mammalian cells. Section 1 gives step by step details relating to identification of candidate synthetases from a randomized pyrrolysine (Hao et al., 2002; Srinivasan, James, & Krzycki, 2002) based-library (Arbely, Torres-Kolbus, Deiters, & Chin, 2012) and the validation and comparative evaluation of the synthetases in prokaryotic expression systems. Section 2 describes the evaluation and verification of candidate synthetases in mammalian cells. This article thus provides the reader with a conceptual and technical roadmap to both discover as well as evaluate the efficacy and fidelity of future orthogonal synthetase/tRNA pairs.

2. Screening of the pyrrolysine synthetase library in E. coli

An overall depiction of the screening workflow (Nguyen, Garcia Alai, Kapadnis, Neumann, & Chin, 2009; Wang, Brock, Herberich, & Schultz, 2001), is shown in Fig. 1. It begins by transforming bacteria with a Pyl aminoacyl-tRNA synthetase (aaRS) library and selection plasmid containing a chloramphenicol resistance gene with a premature TAG codon (Fig. 1). The transformants are plated in the presence of the ncAA and chloramphenicol. Three fates are possible: (1) some synthetases within the library enable charging of Pyl tRNA with the ncAA and thus bacterial survival; (2) still other synthetases enable charging of Pyl tRNA with natural amino acids and thus bacterial survival; and (3) remaining synthetases are incompetent to charge Pyl tRNA with either the ncAA of interest or natural amino acids and are thus selected against. Plasmids are purified from the cells and retransformed into a bacterial cell line with a negative selection plasmid containing an arabinose inducible barnase lethality gene with a premature TAG codon. In the presence of arabinose but without ncAA, the surviving synthetases are those that do not encode for canonical amino acids. Plasmid containing the selected ncAA-RS is purified from these cells and retransformed into a fluorescent reporter plasmid containing a TAG codon at site 150 in superfolder GFP (sfGFP) (Miyake-Stoner et al., 2010; Pedelacq, Cabantous, Tran, Terwilliger, & Waldo, 2006). The cells are plated on an inducing agar that includes the selected for ncAA. Those synthetases that have successfully encoded the ncAA will show as individual green fluorescent colonies.

Fig. 1.

Workflow for screening of orthogonal tRNA/synthetase pairs in E. coli. Outline of the selection screen workflow. The process involves positive and negative survival screening, followed by nonsense suppression within a fluorescent reporter.

After selection, the performance of the isolated ncAA-RS/tRNA pairs is subsequently evaluated in E. coli using several assays. A fluorescence assay is used to test efficiency, or the ability of the synthetase to suppress TAG codons in sfGFP and produce ncAA-sfGFP, as well as fidelity, or the inability to suppress TAG codons in the absence of ncAA (Section 2.3.4). The concentration dependence of charging of an ncAA by the synthetase is indirectly measured via an uptake assay (Section 2.3.5), which also allows determination of working concentrations of ncAA in E. coli (Rauch, Porter, Mehl, & Perona, 2016). Permissivity, or the ability of one synthetase to encode multiple distinct ncAAs, is also explored via fluorescence assay (Section 2.3.6). Finally, ncAA-sfGFP is expressed and purified in order to attain gold-standard mass spectrometry confirmation of stable in vivo encoding (Section 2.3.7).

Note: when specific sources are recommended, they are listed in parentheses.

2.1. Equipment

1. Water bath

2. Open air incubator

3. Incubator shaker

4. Fluorescence plate reader (Biotek Synergy or similar)

5. DNA agarose gel apparatus

6. Refrigerated centrifuge

7. Microfluidizer

8. Spectrophotometer (Nanodrop or similar)

9. Electroporator

2.2. Buffers and reagents

1. SOC broth

2. Commercial mini-prep kit (Macherey-Nagel or similar)

3. Commercial gel extraction kit (Macherey-Nagel or similar)

4. Electro-competent cells

   1. E. coli dh10b
5. Plasmids (available from the Unnatural Protein Facility at Oregon State University)

   1. pBK
   2. pREP
   3. pYOBB2
   4. pALS_GFP
6. Chemicals and solutions

   1. Triethylammonium bicarbonate (TEAB) Buffer (Thermo Scientific #90114)
   2. Luria-Bertani (LB) Medium
   3. Kanamycin stock solution (50 mg/mL in dH₂O)
   4. Tetracycline stock solution (50 mg/mL in DMF)
   5. Chloramphenicol (40 mg/mL in ethanol)
   6. LB agar
   7. 20% Arabinose
   8. 5% Aspartate
   9. 25 × Mineral Salts (625 mM Na₂HPO₄, 625 mM KH₂PO₄, 125 mM Na₂SO₄, 1.25 M NH₄Cl)
   10. 25 × 18 amino acid mix (5 g/L each of: Glutamic Acid Na, Aspartic Acid, Lysine HCl, Arginine HCl, Histidine HCl, Alanine, Proline, Glycine, Threonine, Serine, Glutamine, Asparagine, Valine, Leucine, Isoleucine, Phenylalanine, Tryptophan, Methionine)
   11. Glycerol
   12. Glucose
   13. 1 M MgSO₄
   14. 5000 × Trace Metals (20 mM CaCl₂ 2H₂O, 10 mM MnCl₂ 4H₂O, 10 mM ZnSO₄ 7H₂O, 2 mM CoCl₂ 6H₂O, 2 mM CuCl₂, 2 mM NiCl₂, 2 mM Na₂SeO₃, 2 mM Na₂MoO₄ 2H₂O, 2 mM H₃BO₃, 50 mM FeCl₂)
   15. Sterile water
   16. Imidazole
   17. Tris-buffered saline (20 mM Tris, 500 mM NaCl, pH 7.5)

7. TALON resin

8. Disposable gravity flow column

9. Spin concentrator (VWR Centrifugal Filter Cat# 82031–344 or similar)

2.3. Protocol

2.3.1. Library propagation

1. Transform 1 μg of library DNA into 500 μL of dH10b cells containing the pREP positive selection plasmid by electroporation. Transformations should be done as 10 reactions with 50 μL of cells each.

2. After electroporation, recover each reaction in 1 mL of Super Optimal broth with Catabolite repression (SOC) for 1 h at 37 °C and 250 rpm.

3. Pour 6, 10 cm LB agar plates containing kanamycin and tetracycline at 50 μg/mL and allow to completely dry. Create a serial dilution of 100 μL of recovered cells from 10^{− 1} to 10^{− 6}. Plate 100 μL of each recovery on a corresponding LB agar plate, spread completely, and allow cells to dry onto the plate. Incubate at 37 °C for 16 h. Use the remaining 9.9 mL of recovery to inoculate 500 mL of LB medium containing 50 μg/mL Kan and Tet and grow for 16 h at 37 °C and 250 rpm.

4. After 16 h, calculate the library coverage from the serially diluted plates using the following formula: number of cells amount plated(μL) • dilution factor • recovery volume (μL) = total transformants

2.3.2. Positive selection

1. Using the 500 mL overnight culture from the library propagation, inoculate 500 mL of fresh LB medium containing antibiotics with 3–5 mL of culture. Grow new library culture until OD₆₀₀ reaches the mid-log phase.

2. Pour 11, 15 cm LB agar plates. Of these, 10 plates should contain 50 μg/mL of Kan and Tet, 1 mM of selection ncAA, and 40 μg/mL of chloramphenicol (Cm). The ncAA is usually prepared as a 100 × stock in equimolar NaOH. If NaOH does not solubilize the ncAA, HCl may be used. One control plate should contain all but the chloramphenicol. Plate 100 μL of the library culture at mid-log phase onto plates, allowing them to fully dry before incubating for 16 h at 37 °C.

3. After 16 h, remove the colonies from the 10 plates containing Cm by using 5 mL of LB medium per plate and gently scraping cells from agar surface with a cell spreader. Pool scraped cells and recover for 1 h at 37 °C and 250 rpm. Miniprep the recovered cells and elute in 50 μL of dH₂O.

4. To isolate the pBK plasmid, run all 50 μL of mini-prepped DNA on a 1% agarose gel at 120 V constant for 30 min. Isolate the band near 1700 bp and purify by gel extraction, with the final elution volume between 30 and 35 μL in dH₂O.

2.3.3. Negative selection

1. Transform 10 ng of the plasmid-separated pBK from the positive selection into 50 μL of electro-competent dH10b cells containing the pYOBB2 negative selection plasmid. Recover the cells in 1 mL SOC and incubate for 1 h at 37 °C and 250 rpm.

2. Pour 4, 15 cm LB agar plates. Of these, 3 plates should contain 50 μg/mL of Kan, 25 μg/mL of Cm, and 0.2% arabinose. The additional plate should contain both Kan and Cm but no arabinose. Plate 100 μL of recovery on each plate and allow to dry fully before incubating for 16 h at 37 °C.

3. After 16 h, remove the individual cells from the 3 plates containing arabinose by using 5 mL of LB medium per plate and gently scraping cells from the agar with a cell spreader. Pool scraped cells in a 15 mL culture tube and recover for 1 h at 37 °C and 250 rpm. Miniprep the recovered cells and elute in 50 μL of dH₂O. To isolate the pBK plasmid, run all 50 μL of mini-prepped DNA on a 1% agarose gel at 120 V constant for 30 min. Isolate the band near 1700 bp and purify by gel extraction, with the final elution volume between 30 and 35 μL in dH₂O.

2.3.4. Isolation of synthetases and evaluation via fluorometry

1. Transform 10 ng of the plasmid-separated pBK from the negative selection into 50 μL of electro-competent dH10b cells containing the pALS GFP plasmid. Recover the cells in 1 mL SOC and incubate for 1 h at 37 °C and 250 rpm.

2. Pour 3, 15 cm auto-induction plates from a 500 mL mixture containing: 407 mL minimal agar, 25 mL 5% aspartate, 25 mL 10% glycerol, 20 mL 25 × M Salts, 20 mL 25 × 18 amino acid mix, 1.25 mL 20% arabinose, 1 mL MgSO₄, 0.125 mL 40% glucose, and 0.1 mL 5000 × trace metals. Of these, 2 plates should contain 50 μg/mL of Kan and Tet and 1 mM ncAA. The additional plate should contain both Kan and Tet but no amino acid. Plate 100 μL of recovery on each plate and allow to dry fully before incubating for 16–24 h at 37 °C.

3. The following day, visibly green colonies should be present. If there are none, allow the plates to develop in the dark at 25 °C for another 24 h.

4. Prepare a 96 well block by making 50 mL of non-inducing media consisting of: 43 mL sterile H₂O, 2.5 mL 5% aspartate, 2 mL 25 × M Salts, 2 mL 25 × 18 amino acid mix, 0.625 mL 40% glucose, 0.1 mL MgSO₄, and 0.01 mL 5000 × trace metals with 50 μg/mL of Kan and Tet. Fill each well with 500 μL of non-inducing media. Pick 96 individual green colonies with sterile toothpicks or small pipette tips and place one per well. Cover the block with a Breathe-Easier membrane and incubate for 16–24 h at 37 °C and 300 rpm.

5. After sufficient growth in non-inducing media, prepare 100 mL of auto-inducing media made up of: 82 mL sterile H₂O, 5 mL 5% aspartate, 5 mL 10% glycerol, 4 mL 25 × M Salts, 4 mL 25 × 18 amino acid mix, 0.25 mL 20% arabinose, 0.2 mL MgSO₄, 0.125 mL 40% glucose, 0.02 mL 5000 × trace metals with 50 μg/mL of Kan and Tet. Add 480 μL per well of auto-inducing media into a sterile 96 well block. With the additional 50 mL of auto-inducing media, add 1 mM of ncAA from a 100 × stock. Add 480 μL per well of the auto-inducing media containing ncAA in a sterile 96 well block. Inoculate both blocks with 20 μL of non-inducing culture in their respective wells. Cover with Breathe-Easier membranes and incubate for 48 h at 37 °C and 300 rpm.

6. After 24 and 48 h, fluorescence readings are taken on a BioTek Synergy2 or similar plate reader with emission and excitation filters at 528 and 485 nm, respectively. Fluorescence is normalized by OD₆₀₀ to account for discrepancies in growth. All individual readings with a fluorescence value above 2000 rfu and a minimum fold change of 10 × of ncAA block over control are considered successful synthetase hits. Individual synthetase hits are then sequenced to determine new active site mutations.

2.3.5. Uptake assay

1. Prepare a sterile 96 well block with enough wells to include all synthetase hits of interest and a range of ncAA concentrations, such as 0–2 mM. Add 480 μL of auto-inducing media to each well of the block. Add ncAA to desired concentration range. Inoculate wells with 20 μL of the non-inducing media culture (Section 2.3.4, step 4) with the respective synthetase hit. Cover with a Breath-Easier membrane and incubate for up to 48 h at 37 °C and 300 rpm. Readings are commonly taken at 24 and 48 h on a BioTek Synergy2 or similar plate reader with emission and excitation filters for GFP.

2. The concentration of ncAA at which half maximal fluorescence is achieved is known as the UP₅₀. The resulting curves are fit to a polynomial equation to identify at which point the ncAA concentration allows for half maximal ncAA-sfGFP expression. The TAG site in the pALS_GFP reporter plasmid allows for a correlation between ncAA concentration and RS functional ability and thus serves as a surrogate for direct enzyme kinetic measurement. Ideally, this experiment is done in triplicate to obtain an accurate margin of error.

2.3.6. Permissivity profile of synthetase hits

1. Prepare a sterile 96 well block and calculate the number of wells needed to include all synthetase hits and ncAAs of interest. This may take multiple blocks. Prepare a 100 × stock of each ncAA at a desired concentration. This may be based on the ncAA concentration that achieved optimal normalized fluorescence in the uptake measurements, or a standard concentration of 1 mM may be used.

2. Add ncAA stocks to auto-inducing media to make a 1 × final concentration. Add 480 μL of auto-inducing media to each well corresponding to the desired ncAA and synthetase combination. Use 20 μL of the previously used non-inducing cell culture (Section 2.3.4, step 4) to inoculate each respective well. Cover with a Breathe-Easier membrane and incubate for up to 48 h at 37 °C and 300 rpm. Fluorescence readings are commonly taken at 24 and 48 h on a BioTek Synergy2 or similar plate reader. This is best done in triplicate.

2.3.7. Purification of sfGFP_HIS for mass spectrometry

1. Add 5 mL of non-inducing media to 15 mL culture tubes, one per synthetase to be expressed. Inoculate each culture tube with 20 μL of respective non-inducing culture from the 96 well block previously used (Section 2.3.4, step 4).

2. After sufficient growth is achieved, prepare 250 mL culture flasks with 50 mL of auto-inducing media for each synthetase/ncAA combination being expressed. The ncAA concentration can be determined from uptake measurements or a standard concentration of 1 mM may be used. Dilute ncAA into the auto-inducing media from a 100 × stock. Also prepare a 250 mL culture flask with 50 mL of auto-inducing media without ncAA for each synthetase to observe background incorporation as well as one expressing WT sfGFP_HIS as a control for mass spectrometry.

3. Inoculate each 50 mL culture with 500 μL of a 5 mL synthetase culture grown 16–24 h before. Incubate 50 mL cultures for 16–26 h at 37 °C and 250 rpm.

4. After sufficient cell growth is achieved, spin cultures down in 50 mL conicals for 10 min at ~ 5500 rcf. Decant excess media and keep the cell pellet. Resuspend the cells in 10 mL of phosphate or tris-buffered saline (pH 7.5) containing 5 mM imidazole.

5. Lyse cells. The preferred method is mechanical lysis by microfluidization at 18,000 psi for bacterial cells. Lyse in the same buffer used to resuspend the cell pellet. Spin lysed cells at 10,500 rcf for 30 min. Separate lysate from cell debris by moving supernatant into a 50 mL conical or equivalent receptacle. Add approximately 200 μL of Talon or Nickel resin slurry that has been washed and suspended in the same buffer used in cell lysis. Bind with constant gentle agitation for 1 h at 4 °C.

6. Prepare clean gravity flow columns for each condition. Agitate bound resin and lysate, and pour slowly through the gravity flow column. Continue until there is no longer any lysate. Wash resin with the same buffer used for lysis at 50 × the anticipated elution volume, twice. Elute the protein from the resin using tris- or phosphate buffered saline with 200 mM imidazole as the elution buffer. Add the elution buffer slowly in multiple segments to ensure optimal protein concentration.

7. To prepare the samples for mass spectrometry, exchange buffers from the elution buffer to tetra-ethyl ammonium bicarbonate or equivalent buffer. This can be done via spin concentrating. To equilibrate the concentrator, spin through a volume of the new buffer equivalent to the volume of the protein to be concentrated. Then concentrate the protein by spinning, decanting the separated saline buffer, and reconstituting the remaining protein to the original volume with the new buffer. This should be done three times, or enough to ensure adequate desalting of protein.

8. Quantify the protein concentration via A₂₈₀, on a nanodrop or similar spectrophotometer. Concentrate protein to ~ 0.5 mg/mL and flash freeze in liquid nitrogen.

9. Mass spectrometry results should be interpreted as delta of mass (in Daltons) in comparison to the WT sfGFP_HIS submitted in parallel. We have observed the mass of the dominant LC peak of WT sfGFP_HIS as 27826.4 Da (Fig. 2A). A second mass is commonly observed corresponding to the loss of methionine (− 131 Da) at position 1 (M1).

Fig. 2.

Example deconvoluted ESI mass spectra from sfGFP purified via HIS tag. (A) Spectrum from dominant liquid chromatography peak from sfGFP_HIS purified out of E. coli with observed masses (including common variant wherein methionine 1 (M1) has been lost to hydrolysis.) (B) Spectrum from dominant liquid chromatography peak from sfGFP_V5_HIS purified out of mammalian HEK 293T cells with observed mass of 29140.6. The spectra were collected by Novatia, Inc. (Newtown, PA) from proteins expressed and purified by the authors.

3. Validation and evaluation of identified synthetases in mammalian cells

This section details the adaptation of the synthetases selected in E. coli to the HEK cell derivative 293T, which is a workhorse cell line for expression and functional characterization of recombinant human proteins. This work parallels that of the verification steps performed in E. coli above. Specifically, a similar fluorescence plate reader assay is used to quantify efficiency, fidelity, and permissivity of the synthetase in transfected HEK cells (Section 3.3.1). Additionally, for the ncAA/synthetase combinations of most interest, ncAA-sfGFP is expressed and purified in order to attain gold-standard mass spectrometry confirmation of stable in vivo encoding (Section 3.3.2).

Note: when specific sources are recommended, they are listed in parentheses.

3.1. Equipment

1. Water bath

2. Sterile biosafety cabinet

3. CO₂ incubator

4. Refrigerated centrifuge

5. Fluorescence plate reader (Biotek Synergy or similar)

3.2. Buffers and reagents

1. Cells

   1. 293T HEK cells (ATCC CRL_3216)
   2. NEB stable cells (New England Biolabs)
2. Plasmids (available from the Unnatural Protein Facility at Oregon State University)

   1. sfGFP_N150TAG_V5_HIS + PylT
   2. sfGFP_WT_V5_HIS + PylT
   3. pAcBac1_WT_PylRS + PylT
3. Chemicals and solutions

   • c. DMEM high glucose (Gibco)
   • d. 100 × penicillin/streptomycin (PS) (Gibco)
   • e. 100 × l-glutamine (Gibco)
   • f. Fetal Bovine Serum (Sigma-Aldrich)
   • g. Trypsin EDTA 0.25% (Gibco)
   • h. Sterile DMSO (Sigma-Aldrich)
   • i. PolyJET transfection reagent (SignaGen Laboratories)
   • j. Triethylammonium bicarbonate (TEAB) Buffer (Thermo Scientific #90114)
   • k. DPBS lacking Ca^{2 +} and Mg^{2 +} (Gibco)
   • l. Roche Mini cOmplete EDTA-free protease inhibitor (Millipore-Sigma)
   • m. Ultrapure water (Invitrogen)
   • n. RIPA buffer (Sigma-Aldrich)
   • o. Pierce Universal Nuclease (Thermo Fisher)

4. Nickel-NTA resin (Qiagen)

5. Disposable flow columns

6. Amicon ultra filter 10K MWCO 4 mL

3.3. Protocol

3.3.1. Transfection and quantitation of incorporation into sfGFP_N150TAG_V5_HIS

1. Obtain 293T cells from ATCC (CRL-3216). Maintain cells in complete HEK media, defined as DMEM high glucose (Gibco) supplemented with 10% FBS (Sigma-Aldrich), 1% PS (Gibco) and 1% l-glutamine (Gibco). Passage cells using 0.25% Trypsin EDTA (Gibco) every 24–48 h to ratios 1:3–1:8, taking particular care to always dissociate to single cells. Do not allow cells to become over-confluent. Perform experiments after 5 but no more than 35 passages. Freeze cells in complete HEK media supplemented with 10% sterile DMSO.

2. Subclone appropriate synthetase mutations into pAcBac1_WT_PylRS + PylT, a synthetase/tRNA combination plasmid for expression in mammalian cells. Propagate this plasmid in NEB “stable” cells (NEB) at 30 °C, according to the manufacturer’s instructions.

3. One day before transfection, seed 293T cells to 6 well dishes so that they will be approximately 80% confluent at time of transfection. Dissolve the amino acids as 100 × stocks (200 mM) in equimolar NaOH. Dilute the amino acids to 1 × in HEK cell media and correct the pH of the media with 6 M HCl. If NaOH solubilization is suboptimal, dissolve in HCl, correcting with NaOH. Exchange media approximately 30 min before transfection.

4. Transfect cells using PolyJET reagent (SignaGen) or similar according to manufacturer’s instructions. If using PolyJET, the total amount of DNA per well will be 1 μg, consisting of 0.85 μg synthetase/tRNA plasmid and 0.15 μg sfGFP_150TAG_V5_HIS plasmid. Use 3 μL of PolyJET. To maximize consistency, we recommend designing master mixes of DNA/PolyJET complexes that are distributed to individual wells after mixing. For comparisons of efficiency, transfect in parallel pAcBac1_WT PylRS + PylT and sfGFP_150TAG_V5_HIS + PylT into wells with 2 mM Boc-l-Lysine amino acid (Sigma), prepared identically to the above.

5. Change media approximately 16 h later with or without added unnatural amino acid as appropriate for the experimental condition. Image cells approximately 24 h post transfection, using standard epifluorescence for GFP.

6. Place plates on ice, aspirate media and wash 2 × with ice-cold DPBS (Gibco). Ensure that phenol red from the media is completely removed. Per well, add 200 μL of ice-cold RIPA buffer (Sigma) which has been previously supplemented with protease inhibitors (1 Roche cOmplete Mini tablet per 10 mL buffer) and Pierce universal nuclease (0.5 μL per mL buffer). Rock the plate to cover all cells with the buffer. Incubate on ice for 10 min.

7. At the end of the incubation, triturate and collect lysates. Pellet cell debris by spinning at > 10,000 × g for 20 min at 4 °C. Collect supernatants. Compare GFP fluorescence among conditions on a plate reader. Typically, 40 μL per sample is required for each well of a 96 well plate.

3.3.2. Affinity purification of sfGFP_N150TAG_V5_HIS from HEK cells for mass spectrometry

1. Scale the transfection from pilot experiments so that expectation of sfGFP in lysate is well above 1 μg, which is the minimum required for ESI mass spec. A reasonable expectation of yield is 1–10 μg per 10 cm dish (0.1 g cell pellet) of sfGFP with unnatural amino acids encoded. This being said, the efficiency of the sfGFP purification is enhanced by starting with sufficient material. We recommend purifying from at least 1.0 g of HEK cell pellet for a given synthetase/amino acid combination (~ 10 × 10 cm dishes).

2. At 24–48 h post transfection, wash cells in 10 mL ice-cold DPBS per dish. Scrape cells on ice in cold 2.5 mL per dish of DBPS (Gibco) which has been previously supplemented with protease inhibitors (1 Roche cOmplete Mini tablet per 10 mL buffer). Pellet cells via centrifugation at 1000 × g at 4 °C. Remove supernatant and flash freeze cell pellet in liquid nitrogen. Pellets can be stored at − 80 °C, or one can proceed directly to lysis.

3. Triturate cell pellet on ice in 10 mL hypotonic lysis buffer (10 mM Tris-HCl, pH 8.0 plus 0.1% Triton X-100) per gram of wet pellet weight. Incubate for 10 min on ice. Following incubation, disrupt cells in a tight-fitting homogenizer with 20 strokes on ice. Spin lysates at 20,000 × g for 25 min at 4 °C. During centrifugation, equilibrate the Nickel-NTA resin (Qiagen) by washing in 3 × 5 slurry volumes with bind/wash buffer (150 mM NaCl, 25 mM Tris-HCl, 20 mM Imidazole, pH 8.0). Prepare 0.1 mL resin per mL of lysate.

4. Collect supernatant (cleared lysate) and dilute 1:2 with bind/wash buffer which has been previously supplemented with protease inhibitors (1 Roche cOmplete Mini tablet per 10 mL buffer). Allow to bind to Ni-NTA resin for 1 h at 4 °C with gentle rotation.

5. At the end of 1 h, load into an empty gravity flow column and allow to flow through. Wash column 3 × with 10 column volumes of bind/wash buffer. Elute protein with successive 1 column volumes of bind/wash buffer supplemented with 230 mM additional imidazole (250 mM total) and pH-corrected to 8.0. Assay elutions for presence of GFP (plate reader) or protein (standard Bradford assay).

6. Pool GFP-containing elutions and concentrate and buffer exchange via three rounds of dilution and concentration (10K MWCO 4 mL Amicon ultra filter) with tetra ethyl ammonium buffer, prepared by 10-fold dilution of 1 M TEAB pH 8.5 (Thermo) in ultrapure water (Invitrogen).

7. Concentrate sfGFP to ~ 0.5 μg/mL and flash freeze in LN2. Store at − 80 °C until ready to submit for mass spectrometry. Interpretation of the deconvoluted spectra is as described in Section 2.3.7. We have observed the mass of the dominant LC peak of WT sfGFP_V5_HIS expressed in 293T cells as 29140.6 Da (Fig. 2B).

4. Discussion

Unnatural amino acid mutagenesis via nonsense suppression is a powerful chemical biology approach that is currently limited by the availability and applicability of tools. Here we have described a rationale and step by step workflow to identify and verify robust tRNA/synthetase pairs for novel amino acids. Efficiencies may vary widely for a given synthetase/amino acid combination. To the extent possible, the efficiency and fidelity of encoding of the unnatural amino acid should be confirmed in each specific context using biochemical or functional assays. Good information on the relative performance of an evolved synthetase can be gleaned from comparison with the WT synthetase for a given system. In this case, we describe the use of the WT Pyl RS synthetase in combination with Boc-lysine as one such standard.

Moving forward, once a synthetase/tRNA/ncAA combination is validated using model proteins such as sfGFP, its application can be extended to membrane proteins, such as ion channels. These proteins are very large and often extensively post-translationally modified. As such they may require tryptic digestion for verification of encoding by mass spectrometry.