Identification of pseudouridine methyltransferase in Escherichia coli

Abstract

In ribosomal RNA, modified nucleosides are found in functionally important regions, but their function is obscure. Stem-loop 69 of Escherichia coli 23S rRNA contains three modified nucleosides: pseudouridines at positions 1911 and 1917, and N3 methyl-pseudouridine (m(3)Psi) at position 1915. The gene for pseudouridine methyltransferase was previously not known. We identified E. coli protein YbeA as the methyltransferase methylating Psi1915 in 23S rRNA. The E. coli ybeA gene deletion strain lacks the N3 methylation at position 1915 of 23S rRNA as revealed by primer extension and nucleoside analysis by HPLC. Methylation at position 1915 is restored in the ybeA deletion strain when recombinant YbeA protein is expressed from a plasmid. In addition, we show that purified YbeA protein is able to methylate pseudouridine in vitro using 70S ribosomes but not 50S subunits from the ybeA deletion strain as substrate. Pseudouridine is the preferred substrate as revealed by the inability of YbeA to methylate uridine at position 1915. This shows that YbeA is acting at the final stage during ribosome assembly, probably during translation initiation. Hereby, we propose to rename the YbeA protein to RlmH according to uniform nomenclature of RNA methyltransferases. RlmH belongs to the SPOUT superfamily of methyltransferases. RlmH was found to be well conserved in bacteria, and the gene is present in plant and in several archaeal genomes. RlmH is the first pseudouridine specific methyltransferase identified so far and is likely to be the only one existing in bacteria, as m(3)Psi1915 is the only methylated pseudouridine in bacteria described to date.

FIGURE 1.

Secondary structure of E. coli 23S rRNA stem–loop 69 and the structural formula of m³Ψ. (Left) Stem–loop 69 contains three post-transcriptional modifications: two pseudouridines (Ψ) and one 3-methylpseudouridine (m³Ψ) located at position 1915 according to standard E. coli 23S rRNA numeration. (Right, bold) The methyl group of the m³Ψ-modified base.

FIGURE 2.

Analysis of 1915 region of E. coli 23S rRNA by primer extension. rRNA was isolated from various putative RNA methyltransferase gene knockout strains from (lanes 1–11) the Keio collection, (lanes 12,17) E. coli wild-type MG1655 strain, (lane 18) the MGΔybeA strain, and (lane 19) the ΔybeA/pQE30-ybeA strain. MGΔybeA is the ybeA gene knockout in the MG1655 genetic background, and the ΔybeA/pQE30-ybeA strain is the ΔybeA strain transformed with a plasmid (pQE30-ybeA) expressing YbeA protein. A strong stop signal indicates the presence of methylation at position 1915. (Lanes 13–16) The sequence of the corresponding region of the E. coli 23S rRNA gene.

FIGURE 3.

HPLC analysis of nucleoside composition of 23S rRNA fragment corresponding to positions 1778–1921. The 23S rRNA fragment was isolated from 70S ribosomes of E. coli strains (A) MG1655, (B) MGΔybeA, (C) MGΔybeA/pQE30-YbeA, and (D) MGΔybeA strain ribosomes treated with YbeA in vitro. Peaks corresponding to three standard nucleosides (C, U, and G) and two modified nucleosides, pseudouridine (Ψ) and 3-methylpseudouridine (m³Ψ), are indicated.

FIGURE 4.

Methyltransferase activity of YbeA in vitro. 70S ribosomes were isolated from E. coli wild-type MG1655, MGΔybeA, and ΔybeA/ΔrluD strains. Twenty-four picomoles of 70S ribosomes or 50S subunits was incubated with 100 μM [¹⁴C]-SAM, 3 μg of YbeA, and 1 μg of RluD (if indicated). Incorporation of [¹⁴C]-methyl groups into TCA insoluble material was determined. The data presented are the average of three experiments.

FIGURE 5.

Sequence alignment of proteins belonging to the COG1576 cluster of SPOUT methyltransferases superfamily. The domains of organisms are indicated on the left as follows: (B) Bacteria; (A) Archaea; (E) Eukaryota. Each class of organisms is represented by one sequence, denoted by species name and NCBI gene identification (GI) number. Columns of residues are shaded according to the percent of identity: (dark gray) 100% identity; (light gray) 90% identity. Secondary structure elements derived from E. coli YbeA protein crystal structure (1NS5) are shown on top of the alignment, together with predicted SAM binding motif (Anantharaman et al. 2002). The alignment was performed using MUSCLE (Edgar 2004), and the figure was generated with Jalview (Clamp et al. 2004).