YibK is the 2'-O-methyltransferase TrmL that modifies the wobble nucleotide in Escherichia coli tRNA(Leu) isoacceptors

Abstract

Transfer RNAs are the most densely modified nucleic acid molecules in living cells. In Escherichia coli, more than 30 nucleoside modifications have been characterized, ranging from methylations and pseudouridylations to more complex additions that require multiple enzymatic steps. Most of the modifying enzymes have been identified, although a few notable exceptions include the 2'-O-methyltransferase(s) that methylate the ribose at the nucleotide 34 wobble position in the two leucyl isoacceptors tRNA(Leu)(CmAA) and tRNA(Leu)(cmnm5UmAA). Here, we have used a comparative genomics approach to uncover candidate E. coli genes for the missing enzyme(s). Transfer RNAs from null mutants for candidate genes were analyzed by mass spectrometry and revealed that inactivation of yibK leads to loss of 2'-O-methylation at position 34 in both tRNA(Leu)(CmAA) and tRNA(Leu)(cmnm5UmAA). Loss of YibK methylation reduces the efficiency of codon-wobble base interaction, as demonstrated in an amber suppressor supP system. Inactivation of yibK had no detectable effect on steady-state growth rate, although a distinct disadvantage was noted in multiple-round, mixed-population growth experiments, suggesting that the ability to recover from the stationary phase was impaired. Methylation is restored in vivo by complementing with a recombinant copy of yibK. Despite being one of the smallest characterized α/β knot proteins, YibK independently catalyzes the methyl transfer from S-adenosyl-L-methionine to the 2'-OH of the wobble nucleotide; YibK recognition of this target requires a pyridine at position 34 and N⁶-(isopentenyl)-2-methylthioadenosine at position 37. YibK is one of the last remaining E. coli tRNA modification enzymes to be identified and is now renamed TrmL.

FIGURE 1.

(A) The global network of shared genomic context for tRNA-modification proteins. Genes are represented as spheres, which are colored according to their functional role. Lines linking the spheres represent instances of shared genomic context between the linked genes, including shared gene clustering, co-occurrence in genomes, and gene-fusion events. Strong and weak interactions are marked as red or blue links, respectively. (Orange spheres) Genes coding for tRNA modification enzymes used as baits; (white spheres) the chosen candidate genes. As can be observed, genes coding for tRNA modification enzymes and proteins involved in other translation processes form a densely connected network (i.e., they tend to share the same genomic contexts). (B) Details of YibK and YfiF subnetworks. Networks were projected graphically using Biolayout Express 3D (Freeman et al. 2007).

FIGURE 2.

MALDI-MS spectra of RNase T1 oligonucleotides from bulk E. coli tRNA. The theoretical m/z values of fragments are shown in Table 3 and match well with the empirical values shown above the peaks.

FIGURE 3.

(A) Expanded region of the RNase T1 MALDI-MS spectra. Fragments from tRNA^Leu_CmAA and tRNA^Leu_cmnmUmAA with m/z values of 4933.5 and 4974.5 are seen in the wild-type and ΔyfiF samples, and the corresponding peaks are shifted to masses that are 14 Da smaller in the ΔyibK mutant. For all spectra, the 2′–3′-cyclic forms are apparent; these are 18 Da smaller and seen to the left of the linear phosphate forms, which are indicated with their m/z values. (B) In vivo complementation of BW25113 ΔyibK cells by recombinant 6His-YibK. (C) Secondary structures of tRNA^Leu_CmAA and tRNA^Leu_cmnmUmAA. (Gray) Unique fragments resulting from T1 digestion.

FIGURE 4.

MALDI-MS spectra of RNase A oligonucleotides from bulk E. coli tRNA. Empirical m/z values of fragments are indicated above the peaks, and match well with the theoretical values (Table 3). (A) The m/z 2074.2 and 2162.1 peaks correspond to fragments from tRNA^Leu_CmAA and tRNA^Leu_cmnm5UmAA. Both fragments are missing in the ΔyibK strain. (B) Enlargement of the region containing the AAms²i⁶AACp fragment from tRNA^Trp at monoisotopic m/z of 1754.2 and the AAms²i⁶AAUp fragments at monoisotopic m/z of 1755.2 that arise from RNase A digestion of the ΔyibK strain tRNA^Leu isoacceptors. Although the naturally occurring ¹²C:¹³C ratio (∼99:1) in all the samples makes it impossible to distinguish unambiguously between these two fragments, the proportionally higher peak in the ΔyibK sample at m/z 1755.2 is consistent with the presence of the AAms²i⁶AAUp fragments.

FIGURE 5.

In vitro methylation by YibK of the tRNA^Leu_CAA chimera. (A) Expression and purification of the tRNA^Leu_CAA chimera. Bulk tRNA (first four lanes) and chimera tRNA^Leu_CAA purified from ΔyibK cells (fifth lane) were run on a 3% agarose gel. (B) HPLC analysis of the YibK activity with (left) or without (right) SAM on the chimera tRNA^Leu_CAA purified from ΔyibK cells. Absorbance was monitored at 270 nm. mAU, absorbance units × 10⁻³.

FIGURE 6.

Identity determinants in tRNA^Leu_CAA for recognition by YibK. YibK activity in vitro on wild-type and mutant versions of the tRNA^Leu_CAA chimera was monitored by HPLC analysis. Chromatogram views at top (35–42 min) show the Cm production (percent of RNA molecules methylated by YibK) for wild-type and mutant versions of the tRNA^Leu_CAA chimera extracted from yibK or miaA/yibK strains. Chromatogram views at bottom (56–62 min) show the proportion (percent) of tRNA substrates modified with ms²i⁶A.