Tb³⁺-tRNA for LRET studies of Protein Synthesis

Abstract

When suitably labeled bulk tRNAs are transfected into cells they give rise to FRET (fluorescence resonance energy transfer) signals via binding to ribosomes that provide a measure of total protein synthesis. Application of this approach to monitoring rates of specific protein synthesis requires achieving very high signal-to-noise ratio. Such high ratios may be attainable using LRET (luminescence resonance energy transfer) in place of FRET. Lanthanide complexes containing an antenna chromophore are excellent LRET donors. Here we describe the synthesis of a Phe-tRNA^Phe labeled with a Tb³⁺ complex, denoted Tb³⁺-Phe-tRNA^Phe that, notwithstanding the bulkiness of the Tb³⁺ complex, is active in protein synthesis.

INTRODUCTION

We and our collaborators recently have demonstrated that when fluorescently-labeled tRNAs (fl-tRNAs), both FRET (fluorescence resonance energy transfer) donors and acceptors, are transfected into cells, they give rise to FRET signals.¹ Such signals are expected to arise when donor-labeled and acceptor-labeled tRNAs come into close proximity (< 6 nm), as they do on the ribosome during the course of polypeptide elongation.^2,3 In accord with this expectation, the intensity of the FRET signal was found to correlate with the number of ribosomes engaged in protein synthesis, providing a real-time, live-cell assay for measuring rates of protein synthesis. In this prior work, fl-tRNAs were synthesized by condensing fluorescent weak-base amines, including hydrazides, with the tetrahydrouridine (THU) residues in bulk tRNA that are formed by specific reduction of dihydrouridine (DHU) residues within the D-loops of tRNA, as described.^4–6 The FRET signal generated using bulk fl-tRNAs provided a measure of the rate of total protein synthesis within transfected cells.¹

Current efforts are focused on applying this approach to measure the intracellular rates of synthesis of specific proteins, which will be facilitated by higher signal-to-background ratios. Background interference from unbound donor and acceptor fl-tRNAs, autofluorescent compounds and light scattering, while minimized by excitation wavelength selection and emission filters, limits the sensitivity and precision of the FRET approach. Substituting gated-detection of LRET (luminescence resonance energy transfer) for FRET offers a plausible solution to the background problem. LRET uses long-lifetime lanthanide (Ln³⁺) donors (e. g., Tb³⁺ or Eu³⁺) to temporally separate the delayed (>100 µs) emission of acceptors from the prompt (ns) emission of fluorescent species that are unrelated to resonance energy transfer, with reported increases in signal-to-background ratios of ≥ 500-fold.⁷

Ln³⁺ complexes suitable for LRET studies are composed of an “antenna” chromophore, necessary to overcome the inherently low absorbance of Ln³⁺, attached to a multidentate chelating site for the Ln³⁺.^7,8 Here we report the synthesis of a hydrazide derivative of such a Ln³⁺-complex, its attachment to tRNA^Phe, and the activity of this bioconjugate, charged with Phe, both in forming a ternary complex with EF-Tu·GTP that is required for Phe-tRNA^Phe uptake by the ribosome, and in supporting synthesis of the full-length protein emerald green fluorescent protein (EmGFP).

EXPERIMENTAL PROCEDURES

Chemicals and Instrumentation

The following reagents were purchased from Aldrich: 1,3-phenylenediamine, diethyl-1,3-acetonedicarboxylate, zinc chloride, tert-butyl carbazate, N-(3-dimethylaminopropyl)-N’-ethylcarbodiimide hydrochloride (EDAC), diethylenetriaminepentaacetic acid (DTPA), terbium (III) chloride, europium (III) chloride. E. Coli tRNA^Phe was obtained from Chemical Block (Moscow). ³H-Phenylalanine (Phe) was obtained from Perkin Elmer.

Assaying the translational activity of the Tb³⁺-tRNA using cell-free protein synthesis (CFPS)

The CFPS kit (5prime) was utilized to synthesize EmGFP according to the manufacturer’s manual, with some modifications.⁹ Briefly, the cell lysate was supplemented with protease inhibitors and an inhibitor of tRNA^Phe-aminoacyl-synthetase (PheRS) and then dialyzed in order to deplete all amino acids while preventing Phe regeneration and tRNA^Phe aminoacylation. Following dialysis, the other CFPS components were added according to the manufacturer’s manual, excluding the amino acid Phe. The total ribosome concentration was estimated as 0.3 µM based on A₂₆₀. The ability of this CFPS^−Phe mixture to produce EmGFP was dependent on the addition of exogenous Phe, supplied in the form of either Phe-tRNA^Phe or Tb³⁺-Phe-tRNA^Phe (2 µM). The reaction was initiated by addition of 130 ng plasmid DNA per 10 µl reaction volume and fluorescence was recorded using a plate reader (EnVision - Perkin-Elmer) at 30 °C using excitation and emission wavelengths of 486 and 535 nm, respectively.

Spectra

¹H NMR spectra were recorded in a Bruker DMX 300 spectrometer in deuterated solvents. ¹H chemical shifts were referenced to internal solvent resonances and reported relative to SiMe₄. Low-resolution electrospray (ESI) mass spectra were obtained at the Waters SQD equipped with an Acquity UPLC. UV-visible (UV-vis) absorption spectra were recorded using a Hewlett Packard 8452A Diode Array Spectrophotometer. Fluorescence emission spectra were recorded using a Fluorolog-3 spectrofluorometer (Horiba Jobin Yvon). Time-resolved luminescence measurements were performed on a FS900 spectrofluorometer (Edinburgh Instruments, UK), equipped with R2658P photomultiplier tube (Hamamatsu) and a xenon flash lamp as the excitation source. Kinetic experiments were performed on a KinTek SF-300X stopped-flow spectrofluorometer.

Synthesis (Fig. 1)

Fig. 1. Synthetic Scheme for VI.

Compounds I and III were prepared as described.^10,11

Compound II. To a suspension of I (70 mg, 0.32 mmol) in DCM (2 mL) and DMF (0.8 mL) was added tert-butyl carbazate (85 mg, 0.64 mmol) and N-(3-dimethylaminopropyl)-N’-ethylcarbodiimide hydrochloride (EDAC) (80 mg, 0.41 mmol). The reaction was maintained for 20 h at room temperature. Solvent was removed by evaporation in vacuo. The residue was dissolved in ethyl acetate (15 mL) and extracted with a solution of saturated aqueous NaHCO₃ (3 × 10 mL). The organic layer was dried over anhydrous Na₂SO₄ and evaporated. The product was purified by column chromatography on silica gel using a chloroform/ethanol mixture (8:2) as eluent. Yield 62%. ESIMS (m/z) 333 [M+H]⁺. ¹H NMR (300 MHz, DMSO d₆): 1.39 (s, 9H, Boc), 3.52 (s, 2H, -CH₂CO-), 5.74 (broad, 2H, 7 amino), 6.07 (s, 1H, 3H), 6.37 (d, 1H, 8H, J= 2.1), 6.43 (dd, 1H, 6H, J₁= 8.7 Hz, J₂= 2.1 Hz), 7.41 (d, 1H, 5H, J= 8.7), 9.85 (1H, broad, NH), 11.20 (1H, broad, amide). ¹³C NMR (300 MHz, DMSO-d₆): 28.99, 44.06, 80.77, 97.40, 110.68, 111.01, 116.84, 125.92, 140.41, 146.01, 151.35, 156.50, 162.62, 170.35.

Compound III. Pyridine (1.2 mL) and acetonitrile (1.2 mL) warmed to 50 °C were mixed with a solution of DTPA (490 mg, 1.37 mmol) in acetic anhydride (0.8 mL, 8.60 mmol). After 24 h, solvent was removed by evaporation in vacuo and the residue was washed with acetic anhydride and diethyl ether. The solid was dried under vacuum. Yield 92%. ESIMS (m/z) 356 [M−H]⁻. ¹H NMR (300 MHz, DMSO-d₆): 3.71 (s, 8H, terminal -NCH₂CO₂), 3.30 (s, 2H, central -NCH₂CO₂), 2.74 (t, 4H, J= 5.7 Hz, -NCH₂CH₂N-), 2.59 (t, 4H, J= 5.7 Hz, -NCH₂CH₂N-).

Compound IV. To a suspension of DTPA dianhydride III (54 mg, 0.15 mmol) in DMF (1 mL) was added triethylamine (0.08 mL, 0.6 mmol) under nitrogen. After 10 min a solution of II (15 mg, 0.04 mmol) in DMF (0.5 mL) was added and the reaction was maintained at room temperature for 3 h. A few drops of water were added to quench the reaction. The product was purified by preparative thin layer chromatography, using acetonitrile/water (7:3) as eluent. Yield 30%. ESIMS (m/z) 706 [M−H]⁻. ¹H NMR (300 MHz, D₂O): 1.39 (s, 9H, Boc), 2.60-3.90 (m, 20H), 6.47 (s, 1H, 3H), 7.16 (m, 1H, 8H), 7.67-7.45 (m, 2H, 5H-6H).

Compound V. A mixture of DCM:TFA (1:1) (100 uL) was added to IV (2 mg, 0.003 mmol) at 0 °C and the reaction was maintained at room temperature for 2 h. V was used without further purification in the preparation of VI. The solvent was evaporated in vacuo. Quantitative yield. ESIMS (m/z) 608 [M+H]⁺. 606 [M−H]⁻; ε_max= 14800 M⁻¹ cm⁻¹ at 340 nm.¹²

Compound VI. Derivatization of Phe-tRNA^Phe with the hydrazide V was based on a previously described protocol^5,6 involving nucleophilic substitution at THU residues. Selective reduction of the two DHU residues at positions 16 and 20 in E. coli tRNA^Phe to THU residues was carried out by incubating tRNA^Phe (2.5 mg/mL), NaBH₄ (10 mg/mL, added from 100 mg/mL in 10 mM KOH) in 40 mM Tris-HCl (pH=7.5), at 0 °C for 60 min, in a total volume of 390 µL. The reaction was quenched by lowering the pH to 4–5 by gradually adding 6 M acetic acid. Unreacted NaBH₄ was removed via three ethanol precipitations of tRNA^Phe (reduced). Aminoacylation of tRNA^Phe(reduced) (0.5 mg/mL) was performed in a buffer containing 75 µM ³H-phenylalanine (370 cpm/pmol), 100 mM Tris-HCl (pH 8.0), 4 mM ATP (pH 7.8), 20 mM MgCl₂, 1 mM EDTA, 7 mM 2-mercaptoethanol and 100 µM S100 crude synthetase in a total volume of 1.7 mL. The reaction was incubated at 37 °C during 30 min. Following phenol-chloroform extraction and ethanol precipitation, Phe-tRNA^Phe(reduced) was loaded onto a FPLC monoQ column. Chromatography with a gradient of 0.5–0.9 M NaCl in 50 mM NaAc (pH 5.0), yielded a preparation of Phe-tRNA^Phe(reduced) in which the tRNA was 30% charged with Phe. This preparation (0.7 mg) was resuspended in 11 µL of 0.2 M NaCOOH (pH 3.7) and 10 µL of DMSO. 2 µL of a DMSO solution of V (290 mM) was added and incubation was carried out for 150 min at 37 °C. During this period, additional 1 µL portions of V in DMSO solution were added at 45 and 90 min. Vacuum drying and removal of excess V by three rounds of ethanol precipitation gave an overall stoichiometry of labeling of 0.4 V/tRNA. Unlabeled Phe-tRNA^Phe (reduced) was removed from V-labeled Phe-tRNA^Phe by RP-HPLC using a C18/15 µm column (Waters) with a gradient of 6–21% ethanol in buffer containing 20 mM NH₄Ac (pH 5.0), 10 mM MgAc₂ and 400 mM NaCl, with absorption monitored at 260 and 340 nm. In this gradient, the labeled tRNA elutes after the unlabeled tRNA. This procedure yielded preparations, designated as VI, containing 1.3 V/Phe-tRNA^Phe; i. e., mixtures of Phe-tRNA^Phes containing 1 or 2 V/tRNA^Phe.

Synthesis of Tb³⁺-Phe-tRNA^Phe. VI was dissolved in water and 1.5 eq of a 300 µM TbCl₃ solution, based on the stoichiometry of V incorporated into VI, was added, with vigorous stirring for 30 min at 0 °C. Excess TbCl₃ was removed by 4 rounds of ethanol precipitation. An analogous procedure was used to prepare Eu³⁺-Phe-tRNA^Phe. Suitable control experiments demonstrated that incubation with TbCl₃ of tRNA^Phe not labeled with V afforded no luminescence at 550 nm on excitation at 345 nm (see Fig. S1).

Cell-free expression of EmGFP using the CFPS^−Phe kit

The assay was performed essentially as previously described.⁹ E. coli Phe-tRNA^Phes (2 µM, measured by the amount of charged tRNA) either unlabeled or labeled with V and Tb³⁺, were added to a CFPS^−Phe kit and EmGFP synthesis was monitored by fluorescence intensity at 535 nm.

Ternary Complex (TC) formation

TCs were formed by combining elongation factor Tu (EF-Tu) and Phe-tRNA^Phe in buffer A (50 mM Tris-HCl [pH7.5], 70 mM NH₄Cl, 30 mM KCl, 7 mM MgCl₂ and 1 mM DTT, 1 mM GTP, 1.5 mM phosphoenolpyruvate and 0.5 mg/L of pyruvate kinase) as previously described¹³ except that the ratio of EF-Tu/tRNA employed was 4:1 unless otherwise specified. Time-dependent TC formation was performed at 25 °C by mixing 0.1 uM of Tb³⁺-Phe-tRNA^Phe with 0.4 uM of EF-Tu^Cy5 with excitation of 345 nm. Static fluorescence was recorded with excitation of 345 nm.

RESULTS

Synthesis of the probe, V

Compound I was obtained by condensing 1,3-phenylenediamine and diethyl-1,3 acetonedicarboxylate, followed by saponification of the ethyl ester according to previously reported procedure.¹⁰ We introduced the Boc-protected hydrazide function by treatment of I with tert-butyl carbazate and N-(3-dimethylaminopropyl)-N’-ethylcarbodiimide hydrochloride (EDAC), yielding II. An excess of DTPA dianhydride III was reacted with II in Et₃N and DMF at room temperature for 3 hours, yielding IV, which was purified by thin layer chromatography. Deprotection of IV with TFA:DCM (1:1) afforded the DTPA-based chelate hydrazide (V).

Labeling of Phe-tRNA^Phe(reduced) with V

E. coli tRNA^Phe has DHU residues at positions 16 and 20. Synthesis of charged, VI labeled tRNA^Phe for use in functional studies was carried out using the three-step procedure previously employed to label DHU positions, involving 1) NaBH₄ reduction of the DHU residues to THU to form tRNA^Phe(reduced); 2) charging of tRNA^Phe(reduced) with Phe by Phe-RS and 3) labeling of Phe-tRNA^Phe(reduced) with a large molar excess of hydrazide.^5,6 Such labeling proceeds via a substitution at the THU residues to yield N-substituted tetrahydrocytidine (THC) residues within the tRNA. Previous work⁶ showed that Cy3 hydrazide labeled Phe-tRNA^Phe(reduced) at both positions 16 and 20 to near equal extents (relative labeling 1.0:1.2, respectively) and we assume a similar distribution is obtained for labeling of tRNA^Phe(reduced) by VI. In the present case, the overall labeling stoichiometry was 0.4 V/tRNA^Phe. A more enriched preparation containing 1.3 V/tRNA^Phe, denoted VI, was obtained by separation of derivatized tRNA from underivatized tRNA by RP-HPLC. Within VI, 24–30% of the tRNA^Phe was charged with Phe.

Formation and characterization of the Tb³⁺-VI Complex

Treatment of VI with aqueous lanthanide trichlorides afforded Tb³⁺ and Eu³⁺ complexes. It was also possible to label Phe-tRNA^Phe with a preformed Ln³⁺-V complex. However, resolution of the resulting reaction mixture by RP-HPLC was less successful than the procedure described in Experimental above. We denote the charged form of Tb³⁺-VI as Tb³⁺-Phe-tRNA^Phe. On excitation of the 7-aminoquinolone moiety at 345 nm, Tb³⁺-VI shows the multiple luminescence bands characteristic of Tb³⁺ complexes in the range 485 – 620 nm, corresponding to the ⁵D₄ to ⁷F_n (n = 3 – 6) transitions indicated (Fig. 2A). The following luminescence lifetimes were measured: Tb³⁺-VI: 1.06 ms, Eu³⁺-VI: 0.6 ms, Tb³⁺-Phe-tRNA: 1.08 ms. All lifetime measurements were well fit by single exponential decays.

Fig. 2. TC formation between EF-Tu^Cy5·GTP and Tb³⁺-Phe-tRNA^Phe.

A. Luminescence spectrum of Tb³⁺-Phe-tRNA^Phe (0.2 µM) with varying amounts of EF-Tu^Cy5 on excitation at 345 nm. B. Fluorescence spectrum of EF-Tu^Cy5 (0.2 µM) with varying amounts of Tb³⁺-Phe-tRNA^Phe on excitation at 345 nm. C. Decrease in Tb³⁺-Phe-tRNA^Phe luminescence (550 ± 10 nm) and increase in EF-Tu^Cy5 fluorescence (680 ± 10 nm) on excitation at 345 nm following rapid mixing of Tb³⁺-Phe-tRNA^Phe (0.1 µM, measured as Phe), containing 0.24 Phe/tRNA, and EF-Tu^Cy5·GTP (0.4 µM) (final concentrations). In parts A. and B. excitation frequency doubling in the Fluorolog-3 spectrofluorometer resulted in a very large peak at 690 nm. This artifact was not a problem for the KinTek SF-300X stopped-flow spectrofluorometer employed in part C.

Functional assays

Ternary complex (TC) formation

Binding of aminoacyl-tRNA (aa-tRNA) to the ribosome occurs via the ternary complex aa-tRNA·EF-Tu·GTP, where EF-Tu is a translation elongation factor. We have previously shown that an EF-Tu mutant labeled with Cy5 at position 348, denoted EF-Tu^Cy5, retains functional activity in protein synthesis.¹⁴ Within the TC, this position is within 34 Å and 24 Å of positions 16 and 20, respectively, of aminoacyl-tRNA [calculated from pdb1OB2], making it suitable for LRET studies with Tb³⁺-VI. Formation of the TC between Tb³⁺-Phe-tRNA^Phe and EF-Tu^Cy5·GTP is accompanied by a decrease in Tb³⁺-Phe-tRNA^Phe luminescence as EF-Tu^Cy5 is increased (Fig. 2A) and an increase in EF-Tu^Cy5 fluorescence as Tb³⁺-Phe-tRNA^Phe is increased (Fig. 2B). The K_d value estimated from the results in Figs. 2A, B is approximately 0.3 µM. Additional evidence for LRET between Tb³⁺ and Cy5 within the TC comes from the difference in luminescence lifetimes of the TC complexes made from Tb³⁺-Phe-tRNA^Phe and either EF-Tu^Cy5·GTP (0.45 ms, t_L) or unlabeled EF-Tu·GTP (1.08 ms, t⁰_L). An estimated LRET efficiency, ϕ, equal to 0.58, could be calculated from the measured luminescence lifetime values via eq 1.

$$\mathrm{f}=({{\mathrm{t}}_{\mathrm{L}}}^{-1}-{{{\mathrm{t}}^0}_{\mathrm{L}}}^{-1})/{{\mathrm{t}}_{\mathrm{L}}}^{-1}$$ (1)

The transient changes in Tb³⁺-Phe-tRNA^Phe luminescence and EF-Tu^Cy5 fluorescence (Fig. 2C) on rapid mixing of Tb³⁺-Phe-tRNA^Phe and EF-Tu^Cy5·GTP allows calculation of a 2^nd order rate constant for TC formation of 1.8 × 10⁶ M⁻¹s⁻¹.

EmGFP Synthesis

We recently described a real-time coupled transcription–translation assay that continuously monitors formation of the fluorescent protein Emerald GFP (EmGFP) and can be manipulated so that EmGFP formation is totally dependent on exogenously added Phe-tRNA^Phe, through use of a so-called CFPS^−Phe kit.⁹ Here we use this assay to demonstrate that Tb³⁺-Phe-tRNA^Phe displays activity in supporting EmGFP synthesis that is roughly comparable to that of native Phe-tRNA^Phe (Fig. 3). The assay was carried out under conditions in which the extent of EmGFP synthesis is roughly proportional to added Phe-tRNA^Phe concentration. It is possible that the lower amount of EmGFP synthesized with Tb³⁺-Phe-tRNA^Phe reflects a functional heterogeneity of this sample. For example, the labeling stoichiometry is 1.3 VI/tRNA, and it is possible that the 2:1 labeled tRNA is less active than either of the 1:1 complexes.

Fig. 3. Phe-tRNA^Phe-dependent Cell-Free Protein Synthesis of EmGFP.

Experiments were performed using the CFPS^−Phe mixture, supplemented with 2 µM of the indicated tRNAs, initiated with 130 ng plasmid DNA. Recordings of EmGFP fluorescence were performed every 45 seconds using the plate reader, set to 30 °C using excitation and emission wavelengths of 486 and 535 nm, respectively.

DISCUSSION

This work demonstrates that Phe-tRNA^Phe derivatized with the rather bulky luminescent Tb³⁺ complex VI at THU positions formed by selective NaBH₄ reduction of DHUs at residues 16 and 20 within the D-loop retains high functionality in protein synthesis (Fig. 3). Such retention of functionality is not completely unexpected. Positions 16 and 20 are both quite far from the site of EF-Tu interaction with charged tRNAs, accounting for the readiness with which Tb³⁺-Phe-tRNA^Phe reacts with EF-Tu to form a ternary complex (Fig. 2C), and proximal to the variable loop of tRNA, within which insertions of up to 19 nucleotides are well-tolerated by the ribosome during protein translation.

The success of the current work opens up the prospect that Tb³⁺-Phe-tRNA^Phe, or molecules like it, can be used to monitor LRET signals in live cells that are generated by juxtaposition on the ribosome of a tRNA labeled with VI and a suitable fl-tRNA, labeled, for example, with Cy5, that serves as an LRET acceptor. The antenna in VI, an aminooxoquinoline, was chosen based on prior work showing it to have excellent optical and stability properties for LRET experiments.¹⁵ More recently, other antennas have been described that have superior optical properties,^16,17 but are somewhat larger than 7-aminoquinolones. Further work will determine whether tRNAs containing Ln³⁺ complexes of such antennas will demonstrate functionality comparable to that found with Tb³⁺-Phe-tRNA^Phe.