Time-resolved NMR monitoring of tRNA maturation

Abstract

Although the biological importance of post-transcriptional RNA modifications in gene expression is widely appreciated, methods to directly detect their introduction during RNA biosynthesis are rare and do not easily provide information on the temporal nature of events. Here, we introduce the application of NMR spectroscopy to observe the maturation of tRNAs in cell extracts. By following the maturation of yeast tRNA^Phe with time-resolved NMR measurements, we show that modifications are introduced in a defined sequential order, and that the chronology is controlled by cross-talk between modification events. In particular, we show that a strong hierarchy controls the introduction of the T54, Ψ55 and m¹A58 modifications in the T-arm, and we demonstrate that the modification circuits identified in yeast extract with NMR also impact the tRNA modification process in living cells. The NMR-based methodology presented here could be adapted to investigate different aspects of tRNA maturation and RNA modifications in general.

Fig. 1

Schematic description of the method. a An isotope-labeled RNA (represented as red dots) is introduced in cell extracts in an NMR tube to yield an in extract NMR sample. Successive NMR measurements on a single sample incubated directly in the NMR spectrometer can then be performed. b RNA modifications are monitored on the RNA imino signals with 2D (¹H,¹⁵N) NMR spectra in a time-resolved fashion. Modifications cause the progressive disappearance of imino signals from unmodified nucleotides (signal M disappearing in blue) and the correlated appearance of new imino signals from modified nucleotide (signal M appearing in red). Modifications also cause disappearance and correlated appearance of imino signals of nearby nucleotides (signal P in green). Nucleotides far away from modification sites have unperturbed imino signals over time (signal U in black)

Fig. 2

Modification of yeast tRNA^Phe in yeast extract. a Sequence and cloverleaf representation of unmodified yeast tRNA^Phe (left) and modified tRNA^Phe (right). Main tertiary interactions are represented with thin lines. Modifications are in red. b (left) Imino (¹H,¹⁵N) correlation spectrum of a ¹⁵N-labeled unmodified tRNA^Phe measured in vitro in a buffer approaching physiological conditions (see “Methods” section). Assignments of the imino groups are reported on the spectrum. (right) Imino (¹H,¹⁵N) correlation spectrum of a ¹⁵N-labeled tRNA^Phe measured after 12 h of incubation at 30 °C in yeast extract. The most obvious changes as compared with the spectrum of the unmodified tRNA^Phe are highlighted with red patches

Fig. 3

The comparison of different maturation states gives the NMR signature of individual modifications. a Sequence and cloverleaf representation of unmodified yeast tRNA^Phe produced by in vitro transcription (left), yeast tRNA^Phe produced and modified in E. coli (middle), and yeast tRNA^Phe produced and modified in yeast (right). Main tertiary interactions are represented with thin lines. Modifications are in red. b Imino (¹H,¹⁵N) correlation spectra and imino group assignments of ¹⁵N-labeled tRNA^Phe measured in vitro: (left) unmodified yeast tRNA^Phe, spectrum in red; (middle) yeast tRNA^Phe produced and modified in E. coli, spectrum in cyan; (right) yeast tRNA^Phe produced and modified in yeast, spectrum in purple. c Superposition of imino (¹H,¹⁵N) correlation spectra of various tRNA^Phe samples and NMR signature of individual modifications: (left) unmodified (red) and E. coli-modified (cyan); (middle) unmodified (red) and yeast-modified (purple); (right) unmodified (red), E. coli-modified (cyan), and yeast-modified (purple). A selection of imino assignments is reported on the spectra. NMR signatures of modifications are reported with continuous line arrows (direct effects—see main text), or dashed arrows (indirect effects)

Fig. 4

Time-resolved NMR monitoring of RNA modifications in yeast tRNA^Phe. a Imino (¹H,¹⁵N) correlation spectra of a ¹⁵N-labeled tRNA^Phe measured in vitro (first top left spectrum—same as Fig. 2b left) and in a time-resolved fashion during a continuous incubation at 30 °C in yeast extract over 26 h (remaining eight spectra). Each NMR spectrum measurement spreads over a 2 h time period, as indicated. Modifications detected at the different steps are reported with continuous line arrows for direct effects, or dashed arrows for indirect effects (see main text for the distinction between direct and indirect effects). b Schematic view of the sequential order of the introduction of modifications in yeast tRNA^Phe as observed with NMR. Each modification is associated with a defined color also used in Figs. 5 and 6

Fig. 5

Complex modification circuits in yeast tRNA^Phe. a Imino (¹H,¹⁵N) correlation spectra of a ¹⁵N-labeled tRNA^Phe measured after 16 h of incubation at 30 °C in yeast extracts prepared from wild-type cells (top left spectrum) and in yeast extracts depleted of one modification enzyme at a time, namely dus1Δ, pus4Δ, trm1Δ, trm2Δ, trm4Δ, trm8Δ, and trm11Δ. Each NMR spectrum measurement spreads over a 2 h time period (t = 16–18 h). Complete series of time-resolved NMR monitoring of tRNA^Phe maturation in the different depleted yeast extracts are given in Supplementary Figs. 2–8. Modifications occurring at this specific step in each extract are reported with continuous line arrows for direct effects, or dashed arrows for indirect effects. Each strain is associated with a defined color also used in panel b. b Schematic view of the modification circuits revealed by the NMR monitoring of tRNA^Phe maturation in the different yeast extracts of panel a. Each modification is displayed on the cloverleaf structure with its associated color also used in Figs. 4 and 6. Arrows indicate stimulatory effects and blunted lines inhibitory effects. Thick and thin lines indicate strong and slight effects, respectively

Fig. 6

Quantitative analysis of nucleoside modifications in yeast tRNAs with LC–MS/MS. a Heat map depicting the relative comparison of normalized modification levels in total yeast tRNAs prepared from depleted strains (dus1Δ, pus4Δ, trm1Δ, trm2Δ, trm4Δ, trm8Δ, and trm11Δ) using the wild-type levels as reference. The nucleotides quantified by LC–MS/MS are listed on the left side of the map. The scale bar indicates the fold change in modification levels compared with the wild-type strain (increased levels shown as red, no change as white, and decreased levels as blue). b Histograms showing the relative abundance of T and m¹A modifications in total yeast tRNAs prepared from the depleted strains using the wild-type levels as reference. c Heat map depicting the relative comparison of normalized modification levels in specifically purified yeast tRNA^Phe prepared from the depleted strains using the wild-type levels as reference. The scale bar indicates the fold change as in panel a. d Histograms showing the relative abundance of T and m¹A modifications in specifically purified yeast tRNA^Phe prepared from the depleted strains using the wild-type levels as reference. In panels b and d, black dots represents individual measurements, data heights represent the mean of the biological replicates. Error bars correspond to the s.d. In all panels, significant changes compared to wild type are reported as *** for p < 0.001, ** for p < 0.01 and * for p < 0.05, n = 3. All other changes are not statistically significant. Statistical analyses of the variations compared to the wild-type strain were performed using a two-sided Student’s t-test. See also Supplementary Figs. 9 and 11 for similar analysis of all modified bases in total tRNA and purified tRNA^Phe. Modifications were quantified in three independent biological replicates, except for the modifications Ψ, m⁷G, and m¹A in the specifically purified tRNA^Phe of the trm4Δ strain, for which n = 2. Source data are provided as a Source Data file