Outwitting EF-Tu and the ribosome: translation with d-amino acids

Key components of the translational apparatus, i.e. ribosomes, elongation factor EF-Tu and most aminoacyl-tRNA synthetases, are stereoselective and prevent incorporation of d-amino acids (d-aa) into polypeptides. The rare appearance of d-aa in natural polypeptides arises from post-translational modifications or non-ribosomal synthesis. We introduce an in vitro translation system that enables single incorporation of 17 out of 18 tested d-aa into a polypeptide; incorporation of two or three successive d-aa was also observed in several cases. The system consists of wild-type components and d-aa are introduced via artificially charged, unmodified tRNAGly that was selected according to the rules of ‘thermodynamic compensation’. The results reveal an unexpected plasticity of the ribosomal peptidyltransferase center and thus shed new light on the mechanism of chiral discrimination during translation. Furthermore, ribosomal incorporation of d-aa into polypeptides may greatly expand the armamentarium of in vitro translation towards the identification of peptides and proteins with new properties and functions.

. EMSAs with mutant EF-Tu and L-aa-or D-aa-tRNA Gly u . Red and yellow arrows highlight clearly visible and barely visible ternary complexes with D-aa-tRNA, respectively. In the main text, we have assigned the lower band the 15-mer peptide fMSKAKFARTKPHANA and the upper band the full-length product fMSKAKFARTKPHANAxHHHHHH. This is corroborated by the following evidence: If we translate template 'O' in presence of RF2 to produce a defined stop after 15 amino acids, we find only the lower band ( Figure S4A and B, lane 5), while we find exclusively the upper band if we translate template 'G 1 ' in presence of glycine and GlyRS (lane 4). Translating template 'G 1 ' in absence of glycine/GlyRS without or with added deacyl-tRNA Gly u yields the lower band and a weak upper band (lanes 1 and 2, respectively). The weak upper band observed under these conditions most likely arises from unspecific readthrough of the hungry codon at position 16 during the prolonged incubation time (3 hours at 37°C); we observed the same effect with a hungry tyrosine-specific codon at position 16 and with template 'O' in absence of RF2 (not shown). If we additionally omit L-histidine, we find no band at all (lane 3), suggesting that the histidine residue at position 12 is essential for co-purification of the lower band. The LC-MS data shown in Figure 6 demonstrated the formation of full-length products in response to the addition of L-Trp-tRNA Gly or D-Trp-tRNA Gly to the translation reaction. In the negative control reaction of that experiment, which contained only deacyl-tRNA Gly , we were able to detect an abundant mass that precisely matches the calculated mass of the proposed 15-mer peptide (m calc = 1684.88 Da) and heavier isotopes of it ( Figure S4C).  For comparison: the peak height of 50 ng L-Trp-or D-Trp-Peptide in the EIC with the symmetric peak detection settings applied here is 1×10 4 .
* GlyRS consists of two only loosely coupled chains, tagging only one entails loss of the other during affinity chromatography. Thus, both encoding cistrons were first cloned into pASG-IBA103 to fuse the -subunit to a C-terminal TwinStrep-tag and the fused gene was subcloned into pASG-IBA105 to fuse the -subunit to an Nterminal TwinStrep-tag. PheRS also consists of two subunits that are tightly coupled, so one tag is sufficient.
The gene lepA encoding EF4 was synthesized and subcloned into a StarGate-compatible vector by GeneArt. The gene encoding RF2 harbors an inframe stop codon, expression of the protein requires a frameshift. This necessity was eliminated by site directed mutagenesis. All mutageneses were performed using the QuikChange Lightning site directed mutagenesis kit (Agilent). For verification, all plasmids were sequenced (LGC Genomics).
Expression plasmids were transformed into E.coli NEB Express (New England Biolabs). Cells were grown for 20 h at 30°C in 50-200 ml of EnPresso medium (BioSilta), expression was then induced for 24 h by the addition of anhydrotetracycline to 200 ng/ml, the temperature was increased to 37°C. For RF3 expression, cells were grown to 0.7 OD 600 /ml at 37°C in 2YT-medium and induced for 4 h. Cells were harvested by centrifugation (20 min, 4°C, 6,000 x g), resuspended in Buffer W (100 mM Tris-HCl pH 8.0, 150 mM NaCl, 1 mM EDTA) and lysed by two or three passages at 500 bar through a French®Press (Thermo Electron). Cell debris was pelleted by centrifugation (30 min, 4°C, 15,500 x g). The supernatant was filtered through 0.8/0.2 µm syringe filters (PALL) and subjected to affinity chromatography (AC) over StrepTrap HP 5 ml columns using an ÄKTA Express instrument (GE Healthcare). Buffer W was used for binding and washing, elution was done in the same buffer supplemented with 2.5 mM desthiobiotin (IBA GmbH). For all except IF1 and IF3, fractions were pooled and diluted in 20 mM Tris-HCl pH 8.0, 10 mM KCl and subjected to anion exchange chromatography over HiTrap Q HP 5 ml columns (GE healthcare) in the same buffer with a gradient from 10 mM to 500 mM KCl during 20 column volumes (CV). Purification of both EF-Tu and EF-Ts required a 40 CV gradient, as these proteins co-elute from AC and have a near-identical pI.
IF2 co-elutes from AC with 30S ribosomal subunits. 16S rRNA was eliminated by anion exchange. Ribosomal proteins were subsequently eliminated by a second AC. mM HEPES-KOH pH 7.6 @ 0°C, 100 mM KCl, 10 mM MgCl 2 , 7 mM -mercaptoethanol, 30% glycerol) by buffer addition and re-concentration, and were finally flash-frozen in liquid nitrogen and stored in aliquots at -80°C.
To determine specific protein concentrations, dilutions of the proteins and serial dilutions of bovine serumalbumine (Pierce) were subjected to 10% Bis-Tris SDS-PAGE (novex). Gels were stained with SyproRed (Molecular Probes) according to the manufacturer's instructions and scanned using a MolecularImager FX (Bio-Rad). Band intensities were quantified using ImageLab software (Bio-Rad). This method is independent of the amino acid composition of the proteins, inert to any components of the protein storage buffer and does not co-quantify contaminating proteins of different size if any are present.
Solution 2 (16.67x) contains 5 mM each of the amino acids required for the intended purpose. The pH was adjusted to 7.60 on ice using acetic acid or KOH (depending on the composition). Solution 3 (20x) contains 20.4 µM IF1, 8 µM IF2, 9 µM IF3, 12.6 µM EF-G, 31.6 µM EF-Ts, 4.8 µM RF-1, 3.2 µM RF-3, 9.4 µM RRF, 80 µg/mL creatine kinase (Roche), 2.4 µM ADK, 1.3 µM NDK, 280 µM iPPase, 100 µg/ml T7 RNAP (Stratagene), and 24 µM reassociated 70S ribosomes. For assembly, proteins were mixed in an Amicon Ultra 0.5 ml centrifugal filter column (3 kDa MWCO, Millipore) and concentrated. Stock buffer (see above) was added to 450 µl and concentrated again. Ribosomes were added last to make sure that they will not dissociate due to low magnesium conditions. Several cycles of concentration and addition of stock buffer were performed, until a buffer exchange of at least 99% was achieved. The desired final volume was adjusted with stock buffer. Single-use aliquots were flash-frozen in liquid nitrogen and stored at -80°C. Assembly and storage was done as described for solution 3 above. The aaRS present in translation experiments of templates G 1 , G 2 , G 3 and O are underlined.

Assembly of native tRNA Gly
Due to the special and mutually incompatible deprotection requirements of 4-thiouridine and dihydrouridine, native tRNA Gly was synthesized in three fragments, which were then ligated in two separate steps. Separate ligation steps were required to prevent an otherwise observed direct ligation of fragment 1 to fragment 3. Synthesis of native tRNA Gly fragment 1: Synthesis was commenced on 0.32g 1000A rA(Pac) CPG (41µmol/g). For coupling rG(Pac), rA(Pac) were used instead of rG(ibu) and rA(bz) amidites. The quantitative coupling of the 4-thiouridine building block was achieved by double coupling of the amidite. Oxidization of phosphit triesters was achieved using 0.02M iodine in pyridine/H 2 O (9/1, v/v). Tac 2 O served as capping reagent (Proligo fast deprotection capping reagent). After completed oligonucleotide assembly, the CPG was transferred into a glass bottle and dried under reduced pressure. For the removal of cyanoethyl protective groups 10 ml 1M DBU in ACN was added to the dry support at RT. After 2h the supernatant was discarded and the residue CPG was washed 5x with 20 ml ACN. To facilitate cleavage of the oligonucleotide from the CPG 15 ml of 50mM NaSH solution in tBuNH 2 /MeOH/H 2 O (1/1/2, v/v/v) were added and agitated for 3h at 55°C. After cooling, the supernatant was purified by size exclusion chromatography using NAP25 columns (GE-Healthcare) according to the manufacturer's instructions. The product containing fractions were pooled and freeze dried. Purification was done with IEX-chromatography Source15Q (GE-Healthcare) using 25 mM Tris buffer pH 7.5, 10 %ACN, 2M NaCl at 55°C followed by size exclusion chromatography (NAP10, GE-Healthcare). Yield: 60 OD, 2.40 mg, 798nmol.
Synthesis of native tRNA Gly fragment 2: Synthesis was commenced on 0.32g 1000A rA(Pac) CPG (41µmol/g). For coupling rG(Pac), rA(Pac) were used instead of rG(ibu) and rA(bz) amidites. The quantitative introduction of the dihydrouridine building block was achieved by double coupling of the amidite. Oxidization of phosphit triesters was achieved using 0.05M iodine in pyridine/H 2 O (9/1, v/v). Tac 2 O (Proligo fast deprotection capping reagent) was used as a capping reagent. After completed oligonucleotide assembly, the CPG was washed for 10 min with 10% Et 2 NH in ACN followed by a thorough 10 min wash with ACN for removal of the cyanoethyl protective goups. Finally the CPG was transferred into a glass bottle and dried under reduced pressure. To facilitate cleavage of the oligonucleotide from the CPG 20 ml of 50mM NaSH solution in 28% NH 3 (aq.) were added and agitated for 4h at 25°C. The supernatant was collected, concentrated and purified by size exclusion chromatography using NAP25 columns (GE-Healthcare) according to the manufacturer's instructions. The product containing fractions were pooled and freeze dried. Purification was done with IEX-chromatography Source15Q (GE-Healthcare) using 25 mM Tris buffer pH 7.5, 10 %ACN, 2M NaCl at 55°C followed by size exclusion chromatography (NAP10, GE-Healthcare). Yield: 35 OD, 1.40 mg, 392 nmol.
Synthesis of native tRNA Gly fragment 3: Synthesis was carried out using standard conditions as described above.
Ligation step 1: Triplicates of 200 µl of 10x T4 RNA Ligase 2 buffer (NEB), 15.51 nmol each of tRNA fragment 2 (nucleotides 10-20), fragment 3 (nucleotides 21-76), a complementary DNA strand spanning nucleotides 1-40 (sequence: AGCTTGGGAAGCTCTCGTTCTACCATTGAACTACGCCCGC) and H 2 O to 1973 µl were mixed and spread into 30 equal aliquots. The oligonucleotides were annealed by denaturing for 3 minutes at 95°C and cooling to 4°C at a rate of 0.2°C per second in a PCR cycler. To each aliquot, 30 µg T4 RNA ligase 2 per 1 nmol of fragment 3 were added (total reaction volume 2 ml). The reactions were incubated for 2 h at 37°C and precipitated by the addition of 0.1 volume of 3 M NaOAc pH 5.5 and 2.5 volumes of ethanol and subsequent centrifugation for 15 minutes at 21,500 x g. All pellets were dissolved in a total of 200 µl of 8 M urea and applied to an 8 % TBE-Urea maxi gel. Bands were visualized by UV shadowing. The band of the ligation product was cut out, recovered by electroelution and desalted over a NAP10 column (GE healthcare). This procedure was executed twice, yielding ~9.41 nmol of ligation product (~10%).
Ligation step 2: 150 µl of 10x T4 RNA Ligase buffer (NEB), 9.41 nmol of ligated fragment (2+3), 18.82 nmol of fragment 1, H 2 O to 1.5 ml were mixed and spread into eight equal aliquots of 185 µl. Annealing was done as decribed above. To each aliquot, 8 µg of T4 RNA ligase 2 per nmol of fragment 2+3 were added. The reactions were incubated for 2h at 37°C. Precipitation and purification of the ligation product was performed as described above. The final overall yield was ~520 pmol (~1.1%).
Judged by TBE-Urea-PAGE, the yield of both ligation reactions was about 90% (step 1) and 50% (step 2), but the recovery from the purification procedure was poor. Due to the low overall yield, native tRNA Gly could not be used for routine experiments.

Synthesis of (D)-aa-Flexizyme substrates
Synthesis of Flexizyme substrates was commenced from commercially available (D)-amino acid species. If necessary, -amino-as well as side chain protective groups were introduced by known protocols (7)(8)(9). Amino acid building blocks were then reacted to form ABT, DBE or CME esters. As amino acids can undergo racemization upon activation, the reaction conditions have to be selected carefully (10). In most cases, PyBOP (11,12)  The enantiopurity of the final products was determined by C.A.T. GmbH. The protective group pattern, the promoter group introduced to each amino acid and esterification method used as well as the enantiopurity of the final product is given in the table below.

Chemical synthesis procedures
Chemicals were obtained from commercial suppliers and used without further purification unless otherwise noted. PyBOP was purchased from Novabiochem. EDC was from Acros. HOBt, DMF (extra dry), 4M HCl in Dioxane were from Aldrich. Free and eventually N-Boc and/or side chain protected D-amino acid building blocks were purchased from Bachem and Iris Biotech. 1H-and 13C NMR-spectra were recorded on Bruker DPX 300 or Avance II 500 spectrometer. Chemical shifts (δ, ppm) for 1 H and 13 C are referenced to internal solvent resonances and reported relative to TMS. TLC was carried out on Merck DC Kieselgel 60 F254 aluminium sheets. Compounds were visualized under short-wavelength UV, with ninhydrin solution (300 mg ninhydrin, 3 ml acetic acid, 97 mL n-butanol) or Seebach-Reagent (2.5 g Molybdophosphoric acid, 1 g Ammonium cerium(IV) sulfate, 6 ml conc. sulfuric acid, 94 ml H 2 O). Flash column chromatography was carried out on Kieselgel 60 0.040-0.063 mm (Merck). General Procedure 1 (Thioesterification): At room temperature the Boc-and eventually side chain protected amino acid was dissolved in DMF (0.1 M), 0.95 eq. PyBOP and 2 eq. DÌPEA were added. After 2 min, 1 eq. H-ABT(Boc) was added and the resulting mixture was stirred for 30 min at room temperature. The reaction was quenched by addition of aq. sat. NaHCO 3 solution and extracted with EtOAc (3x). The organic phase was washed with water (2x), brine (1x), dried over Na 2 SO 4 and concentrated to dryness. Purification by flash column chromatography yielded the desired product.

General Procedure 2 (Boc-cleavage):
At room temperature the Boc-and eventually side chain protected amino acid ABT thioester was dissolved in DCM (2 ml/mmol) and 4 M HCl in dioxane (abs.) was added. After stirring for 2h, the reaction was concentrated to dryness and co-evaporated with toluene two times.

Synthesis of Boc-(D)-Ser(tBu)-ABT(Boc):
Following general procedure 1, 523 mg ( then added and incubation at 37°C was continued for another 2 minutes. Radiolabeled tRNA was purified immediately using the NucleoSpin RNA clean-up XS kit (Macherey-Nagel). To allow for purification of small RNAs, the supplied binding buffer was replaced by 0.1 volumes of 3M NaOAc pH 5.5 (Ambion) and 3 volumes of isopropanol. The tRNA was eluted twice with 30 µl of RNase-free water.

Misacylation of tRNA
For a typical reaction for downstream use in translation experiments, 10 µl of 500 mM HEPES-KOH pH 7.5, 3.75 µl of 1 mM Flexizyme and 2.5 µl of 1 mM tRNA and H 2 O to 60 µl were heated to 95°C for 2 min. and cooled to 20°C at a rate of 0.2°C/s. 20 µl of 3 M MgCl 2 were added, the reaction was incubated for 5 min. at room temperature and 2 min. on ice, then 20 µl of 25 mM Flexizyme substrate in DMSO were added and incubated for 3 h on ice. The RNA was precipitated by the addition of 0.1 volumes of 3 M NaOAc pH 5.5 (Ambion), 2.5 volumes of room temperature ethanol and centrifugation for 30 minutes at 21,500 x g and 25°C (to prevent co-precipitation of magnesium). The pellets were washed with 300 mM NaOAc pH 5.5 in 70% ethanol, again with 70% ethanol, air-dried, dissolved in 5 µl of H 2 O, and stored at -80°C until use within the same week.
For downstream use in EF-Tu electrophoretic mobility shift assays, tRNA was misacylated in batch for each series of experiments. The reaction volume per subsequent assay was scaled down to 20 µl and 400,000 cpm of 3'-32 P-radiolabeled tRNA were used instead of non-labeled tRNA. Misacylated tRNA was dissolved in 3 µl of H 2 O per subsequent assay and stored in aliquots of 3 µl at -80°C until use within the same week. One aliquot was generally used to determine the aminoacylation ratio following an earlier protocol (13): to one aliquot of 3 µl, 1 µl of S1 nuclease and 1 µl of 5x S1 nuclease buffer (Fermentas) were added and digestion was allowed to proceed for 10 min. at 37°C. The digests contained 32 P-AMP and aminoacyl-32 P-AMP, which were separated by thin layer chromatography (TLC) on polyethyleneimine-cellulose (Macherey-Nagel), the mobile phase was 5% acetic acid (v/v) with 100 mM NH 4 Cl. K-Screens (Kodak) were exposed to the TLC-sheets usually for 2-4 hours (as appropriate to obtain clear signals without saturated pixels) at -80°C and scanned with a Molecular Imager FX (Bio-Rad) at 50 µm resolution (508 dpi). The spots of aminoacyl-32 P-AMP and 32 P-AMP were quantified using ImageLab software (Bio-Rad) and the aminoacylation ratio was calculated as [aminoacyl-32 P-AMP / (aminoacyl-32 P-AMP + 32 P-AMP)].

EF-Tu electrophoretic mobility shift assay (EMSA)
To were added and the mix was spread into reaction vessels containing 3 µl of radiolabeled aminoacyl-tRNA or deacyl-tRNA in H 2 O. Incubation was continued at 37°C for 10 min. 2 µl of 50% glycerol with bromophenol blue were added and the samples applied to native 8% polyacrylamide gels (acrylamide:bisacrylamide 19:1, dimensions 1818 cm) that were cast in 1x running buffer. The running buffer was 10 mM MES-NH 4 OH pH 6.7; 10 mM Mg(OAc) 2 ; 65 mM NH 4 OAc; 1 mM Na-EDTA; 1 mM DTT, and 10 µM GTP. To reduce band smearing, samples were loaded under applied voltage. Gels were run for 220 min. at 150 V. K-Screens (Kodak) were exposed to the gels for 4 hours (for quantification purposes) or overnight (for qualitative analyses) at -80°C and scanned with a Molecular Imager FX (Bio-Rad) at 50 µm resolution (508 dpi). For evaluation using ImageLab software (Bio-Rad), three volumes per lane were defined: One that includes the ternary complex band, one that includes the free tRNA including the area between the distinct bands (thus including any smear due to TCs dissociated during gel electrophoresis) and one right above the TC band that defines the background for that lane.

Activity test of EF-Tu mutants by GFP translation
A mastermix was prepared comprising per reaction H 2 O to a final reaction volume of 20 µl, 2 µl of SUPERaseIn (Ambion), 2 µl of 120 mM Mg(OAc) 2 , 2 µl Solution 1, 1.2 µl Solution 2 (all canonical amino acids), 1 µl Solution 3, 1 µl Solution 4 (all aaRS), and 1 µl of 200 ng/µl pRSET/EmGFP (Intitrogen). The mastermix was spread to white twin.tec 96-well real-time PCR plates (Eppendorf). 2 µl of 2,000 ng/µl wild-type-EF-Tu or mutant EF-Tu in EF-Tu storage buffer (see above) were added and mixed by pipetting. The plates were sealed and incubated at 37°C in a PolarStar Optima microplate reader (BMG lab technologies). Fluorescence was measured by excitation at 485 nm and detection at 520 nm wavelength every ten min. until it reached a plateau. NaCl, 20 mM imidazole. 90 µl of the same buffer and one reaction of 10 µl were added and shaken at 750 rpm at 4°C for 1 h. The columns were spun dry, washed with 200 µl of water, and eluted twice with 100 µl each of 10% formic acid. The eluates were dried in a SpeedVac and subjected to 16% TRICINE-SDS PAGE (2 h at 125 V, gels and buffers from novex). Phosphorimager-Screens (K-Screens, Kodak) were exposed for 3-4 days and scanned using a MolecularImager FX (Bio-Rad) at 50 µm resolution (508 dpi).

Translation and detection of 35 S-Met labelled peptides
For relative quantification of bands using ImageLab software (Bio-Rad), free-hand volumes were defined precisely around the bands and background was subtracted globally. Care was taken not to underestimate bands of L-aa or control reactions as well as not to overestimate bands of D-aa reactions.

Diastereomeric peptide separation method and mass spectrometric detection.
The RP-HPLC method for separation of the D-and L-peptide isomers was established on synthetic peptides (formyl-MSKAKFARTKPHANA[(D-W or L-W)]HHHHHH), monoisotopic mass 2693.3 Da; Biosyntan). Furthermore, these peptides were used for translation sample spiking as well as a standard mixture for LC-MS performance control before and after sample analysis. As an internal standard for the entire LC-MS performance, the synthetic tripeptide VYV (Sigma-Aldrich, MW 379.2Da, detectable in a single charge state of M+H + with 380.2 m/z) was added to each sample. Separation has been achieved using an Agilent 1290 chromatographic system (Agilent Technologies) and accurate mass analysis using an online coupled ESI -QTOF 6520 (Agilent Technologies) mass spectrometer. Samples were injected into an Acquity UPLC BEH300 C18 column (2.1 x 100mm, 1.7µm particle size, 300Å pore size; Waters) and desalted for 0.7 minutes before switching the flow online to the MS. Elution was performed with a gradient of solvent A and B in 30 minute runs: after 3 min. of desalting with 100% solvent A, separation was performed with 0-15% solvent B in 17 min. followed by a column wash with 15-80% solvent B in 1 min. 80% solvent B was kept for 2 min. to elute the