A novel unanticipated type of pseudouridine synthase with homologs in bacteria, archaea, and eukarya

Abstract

Putative pseudouridine synthase genes are members of a class consisting of four subgroups that possess characteristic amino acid sequence motifs. These genes have been found in all organisms sequenced to date. In Escherichia coli, 10 such genes have been identified, and the 10 synthase gene products have been shown to function in making all of the pseudouridines found in tRNA and ribosomal RNA except for tRNA(Glu) pseudouridine13. In this work, a protein able to make this pseudouridine was purified by standard biochemical procedures. Amino-terminal sequencing of the isolated protein identified the synthase as YgbO. Deletion of the ygbO gene caused the loss of tRNA(Glu) pseudouridine13 and plasmid-borne restoration of the structural gene restored pseudouridine13. Reaction of the overexpressed gene product, renamed TruD, with a tRNA(Glu) transcript made in vitro also yielded only pseudouridine13. A search of the database detected 58 homologs of TruD spanning all three phylogenetic domains, including ancient organisms. Thus, we have identified a new wide-spread class of pseudouridine synthase with no sequence homology to the previously known four subgroups. The only completely conserved sequence motif in all 59 organisms that contained aspartate was GXKD, in motif II. This aspartate was essential for in vitro activity.

FIGURE 1.

MonoS column purification of tRNA^Glu Ψ13 synthase. (A) Activity (□) was tested as described in Materials and Methods and units were expressed as described. Protein was measured as A₂₈₀ (▪). (B) Equal volumes of the column fractions indicated were electrophoresed in SDS gels. (St) MW standards with values as indicated. (Arrow) protein band taken for sequencing.

FIGURE 2.

Ψ sequence analysis of the truD deletion strain (ΔtruD). Deletion of the ygbO (renamed truD) gene in MG1655, construction of wild-type-containing pTrc99A, transformation, and growth of the transformed cells, isolation of RNA, and Ψ sequencing were all performed as described in Materials and Methods. The primer was complementary to tRNA^Glu residues 17–33. The ΔtruD strain was transformed with pTrc99A with no insert (0) and with the wild-type truD structural gene (D80D). As a control, wild-type MG1655 was transformed with pTrc99A with no insert (0). Cells at an A₆₀₀ of 0.6 were induced with 2 mM IPTG for 3 h at 37°C before phenol extraction of the RNA. Reaction with (+) or without (−) CMC followed the standard sequencing protocol. (A, G, C, U) RNA sequencing lanes. The arrow shows the position of the stop one residue 3′ to Ψ13.

FIGURE 3.

In vitro activity of the wild-type (D80D) and two mutant TruD synthases (D80N, D80T). (A) ³H release activity of wild-type and mutant TruD. A 5-[³H]uracil-containing tRNA^Glu transcript was prepared and pseudouridine formation measured as release of ³H as described in Materials and Methods. Substrate concentration was 100 nM in a reaction volume of 0.1 mL. Wild-type and mutant recombinant synthases were overexpressed from pET28a plasmids and purified (see Materials and Methods). A total of 0.05 μg of D80D (♦), no enzyme (⋄), and 0.54 μg of D80N and D80T (triangles and circles, respectively) were added. A total of 0.15 μg of wild-type or mutant enzyme was added to the reaction at 120 min (arrow). (B) Ψ sequence analysis of the site of in vitro Ψ formation on tRNA^Glu. RNAs that had been reacted with the recombinant wild-type and mutant enzyme to completion (200 min) were obtained by phenol extraction of the mixture and analyzed by the sequencing technique described in Materials and Methods. RNA was reacted with (+) or without (−) CMC following the standard sequencing protocol. (A, G, C, U) RNA sequencing lanes. The arrow shows the position of the stop one residue 3′ to Ψ13.

FIGURE 4.

Sequence alignment of TruD homologs from other genomes. The alignment was constructed as described in the text. The identity of each homolog is given by the gi number followed by the organism code taken from the Swiss-Prot site (http://www.expasy.org/cgi-bin/speclist), except that MICDE is Microbulbifer degradans 2-40; MAGS1 is Magnetococcus sp. MC-1; PLAYY is Plasmodium yoelii yoelii. There are six motifs numbered from the amino terminus to the carboxyl terminus as I to VI. The numbers between motifs II and VI are the number of amino acid residues between each motif. The numbers to the left of motif I are the number of residues to the amino terminus. Those to the right of motif VI are the number to the carboxyl terminus. Amino acids conserved 100% are white on black, amino acids with a >80% consensus are black on gray. If there is no single amino acid residue with a >80% consensus, amino acids are classified as polar (p: KRHDEQNST), hydrophopic (h: ALICVMYFW) or aliphatic as a subset of hydrophopic (a: ALICVM) as shown in the CONSENSUS line. If any column is not p, h, or a with a >80% consensus, then the column is classified according to side-chain size as tiny as a subset of small (t: GAS), small (s: GASCDNPSTV), or big (b: QERKYMFWLI). Residues that fit into a column classification are shown in black. Those that do not fit into the classification of its corresponding column are shown in gray. The secondary structure at the top of the alignment is the output of the secondary structure prediction program PHD (Rost and Sander 1993, 1994) using TruD as the query with (e) β-trand, (c) coil, and (h) helix. The most confident structure predictions (>82% accuracy) are shown in uppercase letters. The consensus sequence and a summary of the motifs are shown below the alignment.