Superior cellular activities of azido- over amino-functionalized ligands for engineered preQ1 riboswitches in E.coli

ABSTRACT For this study, we utilized class-I and class-II preQ1-sensing riboswitches as model systems to decipher the structure-activity relationship of rationally designed ligand derivatives in vitro and in vivo. We found that synthetic preQ1 ligands with amino-modified side chains that protrude from the ligand-encapsulating binding pocket, and thereby potentially interact with the phosphate backbone in their protonated form, retain or even increase binding affinity for the riboswitches in vitro. They, however, led to significantly lower riboswitch activities in a reporter system in vivo in E. coli. Importantly, when we substituted the amino- by azido-modified side chains, the cellular activities of the ligands were restored for the class-I conditional gene expression system and even improved for the class-II counterpart. Kinetic analysis of ligand binding in vitro revealed enhanced on-rates for amino-modified derivatives while they were attenuated for azido-modified variants. This shows that neither high affinities nor fast on-rates are necessarily translated into efficient cellular activities. Taken together, our comprehensive study interconnects in vitro kinetics and in vitro thermodynamics of RNA-ligand binding with the ligands’ in vivo performance and thereby encourages azido- rather than amino-functionalized design for enhanced cellular activity.


N-(5-Hydroxypentyl)phthalimide
Aminopentanol (2.820 g, 27.3 mmol) was dissolved in tolune (41 ml) and phthalic anhydride (4.049 g, 27.3 mmol) and triethylamine (273 mg, 2.7 mmol) were added. The reaction mixture was heated to 130°C and refluxed for 4 hours. The solvents were evaporated, and the oily residue was dissolved in ethyl acetate. The organic phase was washed with 1 M HCl, saturated NaHCO3 and saturated NaCl, dried over Na2SO4 and evaporated.

4-Iodobutyl benzoate 1
To a cooled solution of sodium iodide (18.4 g, 123 mmol) in tetrahydrofuran (10 ml, 123 mmol) and acetonitrile (5 ml), benzoyl chloride (14.8 ml, 123 mmol) was added in one portion. The reaction mixture was stirred in the dark overnight. After dilution of the reaction mixture with water and ether, the organic layers were separated and the aqueous layer was extracted three times with ether. The combined organic layers were washed with saturated NaHSO3, Na2CO3 and dried over MgSO4. The solvent was removed in vacuo to give the product as colorless oil.

4-Azidobutanol 1
Lithium hydroxide (3.10 g, 74 mmol) was added to a solution of 4-azidobutyl benzoate (13.5 g, 61.6 mmol) in tetrahydrofuran (40 ml), water (16 ml) and methanol (4.4 ml). The reaction mixture was stirred overnight and subsequently diluted with water and ether. The layers were separated and the aqueous layer was further extracted with ether three times. The combined organic layers were washed with saturated NaCl and dried over MgSO4. The solvent was removed in vacuo to give the product as yellow oil.

Deprotection of oligonucleotides
The solid support was treated each with ammonium hydroxide (28-30%, 0.

Determination of ligand binding affinities (KD)
All fluorescence experiments were carried out on a Cary Eclipse spectrometer (Varian, Palo Alto, USA) equipped with a peltier block, a magnetic stirring device, and a RX2000 stoppedflow apparatus (Applied Photophysics Ltd., Leatherhead, UK). 2-Aminopurine labeled RNA samples (0.5 µM) were prepared in a total volume of 1000 µl of buffer (50 mM MOPS, 100 mM KCl, 2 mM MgCl2, pH 7.5) (4). The samples were heated to 90°C for 2 minutes, allowed to cool to room temperature and held at 20°C in the peltier controlled sample holder in a quartz cuvette equipped with a small stir bar. Then, ligand was manually pipetted in 1μL aliquots in a way not to exceed a total volume increase of 2%. The solution was stirred during each titration step and allowed to equilibrate for at least 10 minutes before data collection. Spectra were recorded from 320 to 500 nm using the following instrumental parameters: excitation wavelength, 308 nm; increments, 1 nm; scan rate, 120 nm/min; slit widths, 10 nm. The apparent binding

Determination of ligand binding kinetics (kobs and kon)
Rate constants k for individual riboswitch variants were measured under pseudo-first-order conditions with ligand in excess over RNA (5). Stock solutions were prepared for each 2-

Cellular assays (IC50)
Synthetic class I and class II preQ1 riboswitch encoding DNAs with EcoRI and SphI overhangs were purchased from IDT (Integrated DNA Technologies) and cloned into EcoRI/SphI sites of a pQE-70 bacterial expression vector (Qiagen) upstream to an eGFP reporter gene. Thus, the introduced riboswitch sequences replace the vector-encoded ribosomal binding site and provide the Shine-Dalgarno sequence required for translation. The plasmids were transformed into preQ1-deficient E.coli B105 cells (6). Bacterial cultures were grown to an OD600 of ~0.5 in LB medium (1% tryptone, 0.5% yeast extract, 0.5% NaCl, 10 mM MgSO4, pH 7) before induction of reporter transcription by addition of 0.8 mM isopropyl-β-D-thiogalactopyranoside (IPTG). At the same time, ligands were added at different concentrations (0.1 µM -7.5 mM). Supporting Figure S2. Comparison of in vitro and in vivo performance of preQ1 class-I and -II riboswitches with modified ligand derivatives. A) Exemplary fluorescence time traces of Ap-labeled preQ1 RNAs in response to Mg 2+ and ligand analog 2 (conditions: 0.5 μM RNA, 100 mM KCl, 50 mM MOPS, pH 7.5, 293 K. Ligands: 2 mM MgCl2, 5 μM compound 2); affinities KD were obtained from plots of normalized AP fluorescence intensities plotted as a function of ligand concentrations (for details see Supporting Figure S2); rate constants kon(293) of the Tte preQ1 class-I riboswitch were obtained from plots of observed rates kobs vs ligand concentrations (for details see Supporting Figure S3); binding to the Spn preQ1-II riboswitch was independent of ligand concentration; for details of kobs determination see Supporting Figure S4 Supporting Figure S4. Stopped-flow fluorescence spectroscopy was used to monitor the kinetics of ligand preQ1 binding to the Spn preQ1 class-II riboswitch. Exemplary real time Ap fluorescence time traces of the Spn A11Ap variant in response to preQ1 1 and preQ1 analogs 3 to 9 (solid line represents single-exponential curve fits); the observed rates kobs were independent of ligand concentrations in the range tested (2 to 14-fold excess of ligand). c(RNA) = 0.3 µM, c(MgCl2) = 2 mM, 100 mM KCl, 50 mM MOPS, pH 7.5, 293 K. Ligand concentration c(ligand) = 1.8 µM.