The universal YrdC/Sua5 family is required for the formation of threonylcarbamoyladenosine in tRNA.

Threonylcarbamoyladenosine (t(6)A) is a universal modification found at position 37 of ANN decoding tRNAs, which imparts a unique structure to the anticodon loop enhancing its binding to ribosomes in vitro. Using a combination of bioinformatic, genetic, structural and biochemical approaches, the universal protein family YrdC/Sua5 (COG0009) was shown to be involved in the biosynthesis of this hypermodified base. Contradictory reports on the essentiality of both the yrdC wild-type gene of Escherichia coli and the SUA5 wild-type gene of Saccharomyces cerevisiae led us to reconstruct null alleles for both genes and prove that yrdC is essential in E. coli, whereas SUA5 is dispensable in yeast but results in severe growth phenotypes. Structural and biochemical analyses revealed that the E. coli YrdC protein binds ATP and preferentially binds RNA(Thr) lacking only the t(6)A modification. This work lays the foundation for elucidating the function of a protein family found in every sequenced genome to date and understanding the role of t(6)A in vivo.

INTRODUCTION

Found in all sequenced genomes to date, the YrdC (COG0009) family has been ranked in the top 10 conserved hypothetical proteins to be considered as high priority targets for experimental study (1). The crystal structures of two Escherichia coli COG0009 family members YrdC and YciO (2,3) reveal an α/β twisted open-sheet structure with a positively charged cleft in its center that was proposed to form a nucleic-acid-binding site (2). Binding to double-stranded RNA (including tRNA) was experimentally verified for the E. coli YrdC protein (YrdC) (2). The study of the COG0009 encoding gene from Saccharomyces cerevisiae, SUA5, preliminarily addressed the function of this gene family. A mutant allele, sua5-1 (Ser107 → Phe replacement) was identified as a suppressor of a translation initiation defect in the leader region of the cyc1-1019 allele of the iso-1-cytochrome c encoding gene CYC1. In a strain carrying the cyc1-1019 allele, initiation was shown to occur at an aberrant up-stream and out of frame ATG (due to a single base-pair substitution) and resulted in expression of ∼2% of the normal Cyc1 protein amount. Interestingly, suppression did not occur at the transcriptional level as observed for the other sua suppressors identified in the same screen (4). A strain carrying a sua5 null mutation sua5::LEU2 displayed pleiotropic phenotypes that included; slow growth, cytochrome a.a3 deficiency, decreased cytochrome b levels and inability to grow on lactate or glycerol (4). The Sua5 protein was implicated in post-transcriptional modulation on gene expression but its biochemical function remained uncharacterized (4). In humans, the YrdC ortholog (YrdCH) was identified as a retinoblastoma binding protein 10 (RBBP10) interactor and a possible role in translation was proposed based on its high and ubiquitous expression profile in several tissues including those in which high levels of protein synthesis occurs (5). The YrdC ortholog was also identified as a new ischemia/reperfusion-inducible protein (IRIP) that indirectly regulates the activity of a variety of transporters probably through its interaction with RSC1A1, a modulator of membrane transport with unclear mode of action (6).

In E. coli, a small deletion of 12 nucleotides including the initiation codon of yrdC (ΔyrdC) was found to suppress the temperature-sensitive phenotype caused by prfA1 (an allele of the prfA gene encoding release factor 1). These mutations were found to be inseparable and a function in ribosome maturation was suggested for yrdC based on the observation that the double mutant displayed slightly higher amounts of 17S rRNA, an immature form of 16S rRNA (7,8).

In summary, members of the COG009 family seem to act as cellular regulators, with pleiotropic effects often linked to translation that vary with the experimental system. Overall the role at the molecular level of the YrdC/Sua5 proteins remained unclear. The essentiality of the gene family also remained and un-resolved issue, as it appeared to vary with the organism considered and the approach utilized to construct mutations. In E. coli, no yrdC deletion mutant was obtained in the systematic single-gene knockout mutant collection, whereas the E. coli homolog yciO was successfully deleted in the same study (9). Also, a yrdC mutant could be isolated in a prfA1 background, but the two mutations were not dissociable, suggesting again that gene yrdC might be essential in wild-type E. coli (8). However, conflicting results have been reported when an insertion of a Tn5-araPBAD construct within the first 15 bp of yrdC did not lead to lethality on rich medium supplemented with glucose (10).

In yeast, even though it was isolated as a suppressor, the sua5::LEU2 allele was dissected away from the original mutation, and deletion of the gene was successfully achieved in homozygous backgrounds suggesting that the gene was not essential (4). However, in the yeast systematic gene deletion collection, a single SUA5 copy could be deleted in the diploid background BY4743, and no deletion was obtained in the haploid background BY4741 (11).

N6-threonylcarbamoyl adenosine (t6A) is an anticodon-loop modification found at position 37 of tRNAs decoding ANN codons. Although considerable biochemical and biophysical information exists on the function of this hypermodified base (12–21), the t6A biosynthesis pathway has only been partially biochemically characterized and shown to be an ATP-dependent process requiring threonine and carbonate although none of the genes involved in its biosynthesis have been identified (22).

In the present study we show by combining a comparative genomic approach with experimental validation that the YrdC/Sua5 family is involved in t6A biosynthesis. We also investigated the essentiality of this gene family in two biological systems, E. coli and yeast. In light of the role of this gene family in the formation of the critically important tRNA modification t6A37, hypotheses to explain some of the pleiotropic effects observed in previous studies are presented.

MATERIALS AND METHODS

Bioinformatics

Analysis of the COG009 family was performed in the SEED database (23). We also used the Blast tools and resources at NCBI (24). Annotations for paralog families were made using physical clustering on the chromosome when possible, by building phylogenetic trees using the ClustalW tool (25) integrated in SEED or by deriving specific protein motifs (http://espript.ibcp.fr/ESPript/ESPript/).

Strains and growth conditions

All E. coli and S. cerevisiae strains used in this study are listed in Table 1. Bacteria were routinely grown on LB medium (BD Diagnostics Systems) at 37°C unless otherwise stated. Bacterial growth media were solidified with 15 g/l agar (BD Diagnostics Systems) for the preparation of plates. Transformation and P1 transductions were performed following standard procedures (26). The sensitivity to P1 phage of all recipient strains used was verified. Ampicillin (Amp, 100 µg/ml), Kanamycin (Km, 50 µg/ml), l-Arabinose (Ara, 0.02–0.2%), were used when appropriate. Bacterial chromosomal DNA was prepared using the Qiagen kit DNeasy® Tissue Kit and yeast chromosome DNA was prepared using Blood & Tissue Kit (Qiagen, Valencia, CA).

Table 1.

Strains and plasmids used

Strain name	Relevant characteristics	Reference
Yeast strains
BY4741	MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 (S288C)	(59)
BY4743	MATa/MATα his3-Δ1/his3-Δ1 leu2-Δ0/leu2-Δ0 lys2-Δ0/lys2-Δ0 ura3-0/ura3-0	(59)
VDC5118	YGN63 transformed with plasmid pYES-DEST52; Ura+, Leu+	This study
VDC5119	YGN63 transformed with plasmid pYES-DEST52; Ura+, Leu+	This study
VDC5120	YGN63 transformed with plasmid pBY135; Ura+, Leu+	This study
VDC5121	YGN63 transformed with plasmid pBY135; Ura+, Leu+	This study
VDC5122	YGN63 transformed with plasmid pBY136; Ura+, Leu+	This study
VDC5123	YGN63 transformed with plasmid pBY136; Ura+, Leu+	This study
VDC5124	YGN63 transformed with plasmid pBY137; Ura+, Leu+	This study
VDC5125	YGN63 transformed with plasmid pBY137; Ura+, Leu+	This study
VDC5126	YGN63 transformed with plasmid pBY138; Ura+, Leu+	This study
VDC5127	YGN63 transformed with plasmid pBY138; Ura+, Leu+	This study
VDC5300	Derivative of BY4743; SUA5/sua5-Δ1::LEU2	This study
VDC5309	Derivative of BY4741 transformed with pYES-DEST52::yrdCEc; Ura+	This study
VDC5416	Derivative of VDC5309; sua5-Δ1::LEU2; Ura+ Leu+	This study
VDC5417	Derivative of VDC5309; SUA5; Ura+ Leu+	This study
VDC5440	Derivative of VDC5416 after FOA treatment; sua5-Δ1::LEU2;Ura− Leu+ (cured of pYES-DEST52::yrdCEc), clone 14	This study
VDC5441	Derivative of VDC5416 after FOA treatment; sua5-Δ1::LEU2;Ura− Leu+ (cured of pYES-DEST52::yrdCEc), clone 15	This study
YJN63	MATα CYC I cyc7-67 ura3-52 leu2-3,112 cyh2 sua5::LEU2	(4)
YMH13	MATα cyc1–1019 cyc7-67 ura3–52 leu2–3,112 cyh2	(4)
E. coli strains
BL21(DE3)	F− ompT hsdSB(rB−, mB−) gal dcm (DE3)
E. coli TOP10	F− mcrA Δ(mrr−hsdRMS−mcrBC) ϕ80lacZΔM15 ΔlacX74 recA1 araD139 Δ(ara-leu) 7697 galU galK rpsL (StrR) endA1 nupG	Invitrogen
MG1655		Coli Genetic Stock Center
NM1100	IN(rrnD−rrnE)1, rph-1, att PLCI857 Red αβγTetR	(60)
VDC5173	Derivative of NM1100 ΔyrdCEc-1::KmR/pBAD24::yrdCEc; KmR/AmpR	This study
VDC5174	Derivative of NM1100 ΔyrdCEc-1::KmR/pBAD24::yrdCEc; KmR/AmpR	This study
VDC5217	MG1655 transformed with plasmid pBY125; AmpR	This study
VDC5218	MG1655 transformed with plasmid pBY125; AmpR	This study
VDC5219	MG1655 transformed with plasmid pBY126; AmpR	This study
VDC5220	MG1655 transformed with plasmid pBY126; AmpR	This study
VDC5221	MG1655 transformed with plasmid pBY128; AmpR	This study
VDC5222	MG1655 transformed with plasmid pBY128; AmpR	This study
VDC5223	MG1655 transformed with plasmid pBY130; AmpR	This study
VDC5224	MG1655 transformed with plasmid pBY130; AmpR	This study
VDC5227	MG1655 transformed with plasmid pBAD24; AmpR	This study
VDC5228	MG1655 transformed with plasmid pBAD24; AmpR	This study
VDC5241	MG1655 transformed with plasmid pBY166; AmpR	This study
VDC5333	MG1655 transformed with plasmid pYC101.1; AmpR	This study
VDC5333	MG1655 transformed with plasmid pYC101.4; AmpR	This study
VDC5335	MG1655 transformed with plasmid pYC102.3; AmpR	This study
VDC5336	MG1655 transformed with plasmid pYC102.3; AmpR	This study
VDC5337	MG1655 transformed with plasmid pYC103.1; AmpR	This study
VDC5337	MG1655 transformed with plasmid pYC103.2; AmpR	This study
VDC5339	MG1655 transformed with plasmid pYC104.2; AmpR	This study
VDC5340	MG1655 transformed with plasmid pYC104.4; AmpR	This study
VDC5342	MG1655 transformed with plasmid pBY167; AmpR
VDC5350	Derivative of VDC5217 obtain from P1 transduction of the ΔyrdCEc-1::KmR allele of VDC5173; KmR, AmpR	This study
VDC5351	Derivative of VDC5219 obtain from P1 transduction of the ΔyrdCEc-1::KmR allele of VDC5173; KmR, AmpR	This study
VDC5354	Derivative of VDC5217 obtain from P1 transduction of the ΔyrdCEc-1::KmR allele of VDC5173; KmR, AmpR	This study
VDC5355	Derivative of VDC5219 obtain from P1 transduction of the ΔyrdCEc-1::KmR allele of VDC5173; KmR, AmpR	This study
VDC5358	Derivative of VDC5223 obtain from P1 transduction of the ΔyrdCEc-1::KmR allele of VDC5173; KmR, AmpR	This study
VDC5359	Derivative of VDC5223 obtain from P1 transduction of the ΔyrdCEc-1::KmR allele of VDC517; KmR, AmpR	This study
VDC5362	Derivative of VDC5242 obtain from P1 transduction of the ΔyrdCEc-1::KmR allele of VDC5173; KmR, AmpR	This study
VDC5363	Derivative of VDC5242 obtain from P1 transduction of the ΔyrdCEc-1::KmR allele of VDC5173; KmR, AmpR	This study
VDC5366	Derivative of VDC5333 obtain from P1 transduction of the ΔyrdCEc-1::KmR allele of VDC5173; KmR, AmpR	This study
VDC5367	Derivative of VDC5333 obtain from P1 transduction of the ΔyrdCEc-1::KmR allele of VDC5173; KmR, AmpR	This study
VDC5370	Derivative of VDC5337 obtain from P1 transduction of the ΔyrdCEc-1::KmR allele of VDC5173; KmR, AmpR	This study
VDC5371	Derivative of VDC5337 obtain from P1 transduction of the ΔyrdCEc-1::KmR allele of VDC5173; KmR, AmpR	This study
Plasmids
pBN200	pET21a:: yrdCEc	This Study
pBN204	pYES::SUA5; URA3	This study
pBY125	pBAD24:: ywlCBs cloned as NcoI/XhoI fragment; AmpR	This study
pBY126	pBAD24:: yrdCEc cloned as NcoI/XhoI fragment; AmpR	This study
pBY128	pBAD24:: yciOEc cloned as NcoI/XhoI fragment; AmpR	This study
pBY130	pBAD24:: SUA5 cloned as NcoI/PstI fragment; AmpR	This study
pBY135.1	pYesDEST52:: yrdCEc; AmpR URA3	This study
pBY136.1	pYesDEST52:: yciOEc; AmpR URA3	This study
pBY137.1	pYesDEST52:: ywlCBs; AmpR URA3	This study
pBY138.1	pYesDEST52:: yrdCMm; AmpR URA3	This study
pBY140	pCR2.1::UpyrdCEc; AmpR (999 base pairs of upstream region of yrdCEc cloned in pCR2.1)	This study
pBY141.6	pCR2.1::UpyrdCEc::DnyrdCEc; AmpR (999 base pairs of upstream and downstream region of yrdCEc cloned in pCR2.1)	This study
pBY146.1	pGEM-T easy:: UpyrdCEc::DnyrdCEc; AmpR	This study
pBY152.1	pGEM-T easy::UpyrdCEc::KmR::DnyrdCEc (ΔyrdCEC-1::KmR cassette)
pBY166	pBAD24:: ycdCMm; AmpR	This study
pBY167	pBAD24:: ycdCMm; AmpR	This study
pYC101.1	Derived from site directed mutagenesis of pBY126; yrdCEcArg110→Ala; AmpR	This study
pYC101.4	Derived from site directed mutagenesis of pBY126; yrdCEcArg110→Ala; AmpR	This study
pYC102.2	Derived from site directed mutagenesis of pBY126; yrdCEcArg52→Ala; AmpR	This study
pYC102.3	Derived from site directed mutagenesis of pBY126; yrdCEcArg52→Ala; AmpR	This study
pYC103.1	Derived from site directed mutagenesis of pBY126; yrdCEcLys50→Ala; AmpR	This study
pYC103.2	Derived from site directed mutagenesis of pBY126; yrdCEcLys53→Ala; AmpR	This study
pYC104.2	Derived from site directed mutagenesis of pBY126; yrdCEcLys50→Ala; AmpR	This study
pYC104.4	Derived from site directed mutagenesis of pBY126; yrdCEcLys50→Ala; AmpR	This study

Yeast strains were routinely grown on YPD (DIFCO Laboratories) at 30°C unless otherwise stated. Synthetic minimal, with or without Agar, SD base or SD base Gal/Raf with or without dropout (-uracil, -ura; -leucine, -leu; -histidine, -his) were purchased from Clontech (Palo Alto, CA) and prepared as recommended by the manufacturer. Glucose (Glu, 2%), Glycerol (Gly, 4%), 5-fluoro-orotic acid (5-FOA, 0.1%) and G 418 (300 µg/ml) were used when appropriate. Transformations were carried out as described in the pYES-DEST52 manual (Invitrogen, Carlsbad, CA).

Cloning and strains constructions

All plasmids constructed are listed in Table 1. Sequences of all primers used are given in Table 2. The E. coli wild-type genes yrdC (NP_417741) and yciO (NP_415783) were amplified by PCR from genomic DNA prepared from strain E. coli K12 MG1655. In a similar fashion the wild-type gene ywlC (NP_391576) was amplified from Bacillus subtilis 168 genomic DNA and the wild-type yeast gene SUA5 (NP_011346) was amplified from strain BY4743 genomic DNA. The wild-type gene yrdC (NP 987306) was amplified from Methanococcus maripaludis S2 genomic DNA. For subcloning in pBAD24, all forward primers (ol1) contained an NcoI sequence and the reverse primers (ol2) contained a HindIII sequence, while the reverse primer to amplify SUA5 contained a PstI site, and are BADEcyrdCol1/BADEcyrdCol2; BADEcyciOol1/BADEcyciOol2; BADBsywlCol1/BADBsywlCol2; BADMmyrdCol1/BADMmyrdCol2; BADsua5ol1/BADsua5ol2. PCR fragments were first inserted into pCR2.1 TOPO (Invitrogen) and transformed into Top 10 cells (Invitrogen) according to the manufacturer protocol. Sequences were verified before subcloning into the appropriate sites of pBAD24 (27).

Table 2.

Primers used in this study

Primer name	Primer sequence
BADBsywlCol1	CCATGGAAACGAAAAGATGGTTT
BADBsywlCol2	AAGCTTTCAGCGAATCACTCTTCCTCC
BADEcyciOol1	CCATGGGCCAGTTTTTTTATATTCAT
BADEcyciOol2	AAGCTTTTATAAGAAAGGCTTCACATCAC
BADEcyrdCol1	CCATGGATAATAACCTGCAAAGAGA
BADEcyrdCol2	AAGCTTTTACCCCTGTCGAAACAGTTC
BADMmyrdCol1	CCATGGAAACTTTCGAGATTTCAGAAA
BADMmyrdCol2	AAGCTTTTATTCATTTACTGCTTTTAAAATCTC
BADsua5ol1	CCATGGACCTTGGACGACATTTTTT
BADsua5ol2	CTGCAGTTAAAACTGTATACAATTATTTGCAG
YcMmyrdcol1	CACCATGAAAACTTTCGAGATTTCAGAAAGT
YcMmyrdCol2	TCATTATTCATTTACTGCTTTTAAAATCTC
YcBsywlCol1	CACCATGAAAACGAAAAGATGGTTTGTG
YcBsywlCol2	TCAGCGAATCACTCTTCCTCCGG
YcEcyciOol1	CACCATGAGCCAGTTTTTTTATATTCATC
YcEcyciOol2	TCATTATAAGAAAGGCTTCACATCAC
YcEcyrdCol1	CACCATGAATAATAACCTGCAAAGAGA
YcEcyrdCol3	TCATTACCCCTGTCGAAACAGTTCA
CENupsau5.ol1	GGTACCTATTTCGTTGAAAAATTCAGGC
CENupsau5.ol2	GGTACCGTATGGGCGACTTTTCGTAT
BN1	CGGCGCTCCAAUAATGTACCTTGGACGACATTTTTTGGCA
BN2	CGACCAATGCTCGAGTTAAAACTGTATACAATTATTTGCAGC
BN3	CAG CCG CAT ATG AAT AAT AAC CTG CAA AGA GAC GC
BN4	GCC TCG CTC GAG CCC CTG TCG AAA CAG TTC ACC CG
US71_ol1	ACA ATG CCC TGC TAT GGC TG
US979_ol2	GTT ATT ATT CCG CCG AAA CCG
DS1561_ol1	TTA GCG GCC GCC GAC AGG GGT AAC ATA ATG G
DS2552_ol2	TTA CTC GAG GCG ATG CTG ACG AAA ACT CG
FRTKnol1	GTGTAGGCTGGAGCTGCTTC
FRTknol2	CTTAGTTCCTATTCCGAAGTTC
sua5::Leu.ol1	TTT CGT TGA AAA ATT CAG GC
sua5::Leu.ol3	GGG CGA CTT TTC GTA TAT ACA
Upleu2_ol1	AGA TCT CAC ACA GGG GCG CTA TCG C
Dsleu2_ol2	AGA TCT TAT AAA GTT TAT GTA CAA ATA TCA
Chksua5leuol4	TTG CCA ATA CAA CAT AAC CG
Chksua5leuol5	GTT ATT GCT CAT CAG CAG TA
EcyrdCLys50Ala.ol1	ACTGTTGGAGTTAGCGCAGCGTCCGGTTG
EcyrdCLys50Ala.ol2	CAACCGGACGCTGCGCTAACTCCAACAGT
EcyrdCLys56Ala.ol1	CAGCGTCCGGTTGATGCGGGGCTGATTTTAATC
EcyrdCLys56Ala.ol2	GATTAAAATCAGCCCCGCATCAACCGGACGCTG
EcyrdCArg52Ala.ol1	GAGTTAAAACAGGCGCCGGTTGATAAGGGGC
EcyrdCArg52Ala.ol2	GCCCCTTATCAACCGGCGCCTGTTTTAACTC
EcyrdCArg110Ala.ol1	CTGGTTGACGGGCGCGTTTGATTCGCTTGCTG
EcyrdCArg110Ala.ol2	CAGCAAGCGAATCAAACGCGCCCGTCAACCAG
Leucassol3	AGT CAT CGA ATT TGA TTC TG

For subcloning into pYES, the PCR amplified products corresponding to the yrdC, yciO, yrdC, ywlC were first cloned into the entry vector pENTR/D-TOPO (Invitrogen), using YcEcyrdCol3/YcEcyrdCol2; YcEcyciOol1/YcEcyciOol2; YcMmyrdcol1/YcMmyrdCol2; YcBsywlCol1/YcBsywlCol2; then sequenced before recombination into pYES-DEST52 (carrying the URA3 gene for selection on –ura media) (Invitrogen) using the Gateway LR Clonase Enzyme Mix according the manufacturer protocol. Plasmids were transformed into yeast strains according to the manufacturer protocol (Invitrogen, Carlsbad, CA). Transformants were screened on SD -ura plates and checked for presence of the orthologous gene in trans by PCR using the gene-specific primers used for cloning.

For cloning into pRS313 (28), primers CENupsua5.ol1 and CENdnsua5.ol2 were used to amplify the SUA5 gene from yeast as well as 272 bp of upstream region and 124 bp of downstream region (UpSUA5Dn). The upstream region was designed based on previous data showing that the presence of a 190 bp fragment upstream of SUA5 is sufficient to drive gene expression (4). The PCR product was cloned into pCR-Blunt TOPO (Invitrogen, Carlsbad, CA) yielding pBY168.6. After checking by sequencing, the UpSUA5Dn was cloned into pRS313 using the EcoRI restriction site flanking the insert in pCR-Blunt and the EcoRI site of pRS313. The resulting plasmid, pBY176.3 (pRS313::UpSUA5Dn), allows SUA5 to be expressed under its native promoter and was used for complementation purposes.

For construction of pBN204, the yeast SUA5 gene was amplified using forward primer BN1 and reverse primer BN2 from plasmid pM249 (4). The forward primer included the yeast initiation sequence AAUA and a BamHI site. The reverse primer included an XhoI site. These sites were used to ligate this DNA fragment into pYES2 (URA3) (Invitrogen) to generate pBN204. For construction of pBN200, yrdC was amplified from genomic DNA using primers BN3 and BN4 before cloning into pET21b (Novagen, Inc., Madison, WI) using NdeI and XhoI.

Construction of the E. coli ΔyrdC::KmR allele

Fragments upstream (908 bp) and downstream (991bp) of yrdC were PCR amplified from E. coli MG1655 genomic DNA using US71_ol1 and UP979_ol2, and DS1561_ol1 and DS2552_ol2, respectively. For subcloning purposes, DS1561_ol1 and DS2552_ol2 contained NotI and XhoI restriction sites respectively. The yrdC upstream fragment (UpyrdC) was cloned into pCR2.1-TOPO (Invitrogen, Carlsbad, CA). The resulting plasmid pCR2.1::UpyrdC (pBY140) was checked for orientation. The yrdC downstream fragment (DnyrdC) was digested with NotI and XhoI and ligated into pBY140 digested with the same enzymes to give plasmid pBY141.6. The Upstream::Downstream fragment from pBY141.6 was then subcloned into pGEM-T easy vector (Promega) using the ApaI and SacI sites, and the resulting plasmid pBY146.1 was then digested with the EcoRV site present between the upstream and the downstream fragments and used to clone the KmR cassette (containing FRT sites) that had been amplified from pKD4 (29) using the FRTKn_ol1 and FRTKn_ol2 primers. The resulting plasmid pGEM-T easy::UpyrdC::KmR::DnyrdC (pBY152.1) was then used as a template to generate by PCR the UpyrdC::KmR::DnyrdC deletion cassette used for the replacement of yrdC. Induction of λ recombination functions and preparation of NM1100/pBAD24::yrdC, NM1100/pBAD24::yciO, and NM1100/pBAD24 competent cells was done as described in (30). After electroporation, cells were plated on LB Km supplemented with 0.02% arabinose. Km resistant colonies were re-isolated twice and checked for gene replacement using Km-specific primers [k1 and k2, (29)] in combination with upstream and downstream primers (UP71_ol1 and DN2552_ol2) as described in the text.

Construction of the sua5-Δ1::LEU2 allele

Primers sua5leu.ol1 and sua5leu.ol3 were used to amplify a 1671 bp fragment encompassing SUA5. This fragment was cloned into pGEMT easy (Promega, Madison, WI) yielding pBY154. A BglII site exists within the amplified fragment and when used deletes about 1000 bp from the amplified fragment, while retaining, 225 bp of homology region on the 5′ side (56 of which are the 56 first bp of SUA5) and 323 bp of homology on the 3′ side. A leucine cassette was amplified from wild-type BY4743 genomic DNA using primers UPleu.ol1 and DNleu.ol2 that contained BglII sites and ligated into pBY154 after digestion of both vector and insert by BglII. The resulting plasmid, pBY164 was used as template to amplify a Up:sua5-Δ1::Dn::LEU2 cassette using sua5leu.ol1 and sua5leu.ol3 primers. The Up::sua5-Δ1::Dn::LEU2 PCR product was used to transform yeast cells. Transformants were plated on SD Gal/Raf -leu and verified by PCR. To cure the URA3 marker containing plasmids (plasmid derived from pYES-DEST52 and pYES2), 10 μl of serial dilutions of yeast liquid cultures grown in SD Gal/Raf -leu were spotted onto two different types of plates. Plates containing SD Gal/Raf -ura -leu medium to check for the presence of expected markers and 5-FOA medium (SD Gal/Raf supplemented with 50 μg/ml of uracil and 0.1% 5-FOA) to select yeast cells that did not retain the URA3 carrying plasmid [in presence of the URA3 gene 5-FOA becomes toxic (31)].

Construction of point mutantions in yrdC by site-directed mutagenesis

Plasmid pBY126 (pBAD24::yrdC) was used as a template for mutagenesis. Primer pairs EcyrdCLys50Ala.ol1/EcyrdCLys50Ala.ol2, EcyrdCArg52Ala.ol1/EcyrdCArg52Ala.ol2, EcyrdCLys53Ala.ol1/EcyrdCLys53Ala.ol2 and EcyrdCArg110Ala.ol1/EcyrdCArg110Ala.ol2, were used to mutate Lysine at position 50, Arginine at position 52, Lysine at position 53 and, Arginine at position 110 into Alanine. Mutagenesis was performed as described in the QuickChanges™ Site Directed Mutagenesis Kit from (Stratagene, La Jolla, CA). All plasmids were checked by sequencing.

Growth curves

Growth curves were performed using a Bioscreen C MBR (Oy Growth Curves AB Ltd, Finland) at 30° and at maximum shaking. 300 µl of culture was used in each well, and 10 replicates were used for each condition. Yeast cultures were grown on SD Gal/Raf -his to saturation diluted 200 times in SD Gal/Raf -his before loading on the Bioscreen. The growth curves presented are representative of 10 replicates.

Purification of bulk tRNA

For E. coli derivatives, 1 liter of cultures were grown overnight at 37°C in LB media supplemented with ampicillin (50 μg/ml) and arabinose when required, harvested, and then stored at –20°C. The cell pellets were defrosted and resuspended in 20 ml of buffer (10 mM Tris, 10 mM MgCl2, pH 7.4). For yeast cultures, 2 l of cell cultures were grown to saturation at 30°C, harvested and pellets immediately resuspended in 20 ml of buffer (10 mM Tris, 10 mM MgCl2, pH 7.4). In both cases, the homogenate was treated with buffer-saturated phenol, pH 7.4 (Sigma), the aqueous layer was collected after the phenol extraction step, and the RNA was precipitated with ethanol. The bulk tRNA was then purified on Nucleobond AX-400 columns (Clontech Laboratories, Mountain View, CA) (according to the manufacturer's protocol) and precipitated with isopropanol.

HPLC separation coupled with electrospray tandem mass spectrometry analysis of digested bulk tRNA

To compare the nucleoside constituents of the bulk tRNA purified from the different strains, samples of each purified bulk tRNA preparation were enzymatically hydrolyzed as described by Pomerantz et al. (32) Approximately 100 μg of bulk tRNA was digested with 10 units of Nuclease P1 (Sigma), 0.01 units of Phosphodiesterase I (Sigma) and 3 μl of E. coli Alkaline Phosphatase (Sigma P4252) in a total volume of 113 μl. The digested extracts were lyophilized and resuspended in 20 μl of water. The resuspended tRNA degradation extracts were injected in a LC–MS/MS system for separation and identification of the nucleosides. A C18-RP (Gemini 5μ, 250 × 4.6 mm, Phenomenex, Torrance, CA, 14 USA) analytical column was used, with a C18 cartridge to prolong its lifetime. The mobile phase consisted of two solvents: 250 mM of ammonium acetate pH6.0 for solvent A and 40% acetonitrile for solvent B. The gradient used for nucleoside separation was performed as previously described (32). It consisted in an elution gradient of several steps: 0–5% B in 7 min, 5–10% B in 15 min, 25–50% B in 5 min, 50–75% B in 4 min and stay at 75% B for 3 min, then 75–100% B in 8 min. The column was then washed at 100% B for 7 min and re-equilibrated at 100% A for 10 min. The flow rate used was 1 ml/min. Chromatography was performed at room temperature. The elution of nucleosides was followed by UV detection at 254 nm. The HPLC was coupled to a hybrid triple quadrupole-ion trap (4000 Q-TRAP, Applied Biosystems, Foster City, CA, USA) mass spectrometer equipped with a TurboIonSpray (TIS) interface operated in the positive ion mode. The ESI parameters were: curtain gas (CUR) at 30 psi, the ion source (IS) at 5000V, nebulizer gas (GS1) at 45 psi, TIS gas (GS2) at 45 psi and the TIS probe temperature at 300°C. The information-dependent acquisition (IDA) mode of operation was employed in which a survey scan from m/z 100–600 was acquired followed by collision-induced dissociation (CID) of the two most intense ions. Survey and MS/MS spectra for each five IDA cycle were accumulated for 1 s and 3 s, respectively. tRNA extractions and analysis were performed at least twice independently.

HPLC analysis

After digestion, samples were immediately used for analysis on HPLC. For that, 100 µl of digested sample was added to a glass HPLC micro-vial and run using 250 mM ammonium acetate pH 6 (Organic Solvent A) and 40% Acetonitrile (Aqueous Solvent B). Samples run for 60 min at a flow rate of 1.5 ml/min in a Suppleco Discovery C-18 Reverse Phase column (15 cm × 4.6 mm× 5 um) (Suppleco 504955) detected by a Waters 2487 Dual Wavelength UV/Vis detector. Samples were injected by a Waters 717 plus Autosampler, and pumped by a Waters 1525 Binary Pump with a Waters Inline Degasser. Analysis was done by Waters Breeze Software v3.20. The method used was a linear gradient consisting of 98% solvent A and 2% solvent B at start with a flow rate of 1.5 ml/min, decreasing to 80% A and 20% B at 60 min with no change in flow rate.

Preparation of tRNAThr

The plasmid containing the E. coli tRNAThrCGU gene thrW was a kind gift of Dr Waas (The Scripps Research Institute). tRNAThrCGU was overexpressed in MRE600 cells induced with isopropyl-β-d-thiogalactoside (IPTG, 0.5 mM, 37°C, 12 h) in LB-Amp (100 µg/ml). The harvested cells were lysed in buffer (0.3 M NaOAc, pH 5.2, 10 mM EDTA) and subjected to phenol extraction (pH 4.5) and ethanol precipitation. tRNAThr was partially purified from total nucleic-acid extract on a Nucleobond column (Clontech Laboratories, Mountain View, CA) using manufacturer protocols, precipitated with isopropylalcohol (100%, room temperature), washed with ethanol (80%, –20°C), and dissolved in water, at which point it was 80% pure. tRNAThr was further purified on a C4 reverse-phase preparative column pre-equilibrated with buffer A (10 mM sodium phosphate pH 5.5, 1 M sodium formate, 8 mM MgCl2) and eluted with a 60-min 0–40% gradient of buffer B (10 mM sodium phosphate pH 5.5, 10% methanol). The tRNA eluted in two equal-sized peaks. Upon assaying the fractions for threonine acceptance from E. coli threonyl-tRNA synthetase using a standard aminoacylation assay as described in (33), both peaks were found to contain tRNAThr at >97% purity. Treatment with base (pH 8.5) of a pooled sample from both peaks yielded the same two separate peaks by analytical reverse-phase chromatography, indicating that the two species do not correspond to different aminoacylation states. Fractions for the two peaks were pooled separately and tRNA was precipitated, washed twice with cold ethanol, and re-dissolved in water. MALDI-TOF mass spectrometry analysis of final tRNA from the two peaks revealed molecular masses of 24 540 and 24 686 Da. The mass difference is accounted for by the molecular mass of carbamoyl-threonine (145.12 Da).

The tRNAThr transcript was synthesized by standard T7 polymerase transcription (34) of a linearized DNA template [pUC18; (35)] containing the E. coli tRNAThrCGU gene thrW (kind gift of Dr K. Beebe, The Scripps Research Institute). After phenol–chloroform extraction of the reaction, the RNA was ethanol precipitated, re-dissolved in water and purified by anion-exchange high-performance liquid chromatography on a heated (90°C) preparative DNA-Pac column (Dionex, Sunnyvale, CA) according to standard protocols (34). The single tRNAThr eluant was pooled, ethanol precipitated, re-suspended in water and assayed for threonine acceptance. Final purity was >95% by 10% urea PAGE analysis. All final tRNA samples were prepared for binding studies by heating to 90°C and slow-cooling after addition of 5 mM MgCl2 to ensure proper folding.

Overexpression and purification of the E. coli YrdC protein (YrdCEc)

E. coli BL21(DE3) cells were transformed with a pET21 derivative (Novagen) vector containing the E. coli yrdC gene (pBN200). The protein was overexpressed by induction with 1 mM IPTG in LB medium containing 100 μg/ml Amp. Cells were harvested and lysed in buffer (20 mM Tris, 500 mM NaCl, 28 mM imidazole, 1 mM β-mercaptoethanol and protease inhibitor cocktail). The protein was purified by Ni-NTA affinity chromatography, dialyzed against phosphate-buffered saline and concentrated using a YM-3 Centriprep filter (Millipore Corp., Billerica, MA).

Measurements of YrdCEc binding to tRNA by Trp fluorescence quenching

We used steady-state techniques to monitor changes in the intrinsic fluorescence of the two tryptophan side chains (at positions 89 and 106) of E. coli YrdC upon addition of tRNA as previously described (2). Using a Jobin Yvon FluoroMax-3 spectrofluorometer (Edison, NJ), with a cell compartment kept at a constant temperature of 21°C, the fluorescence signal (λex = 295 nm; λem = 345 nm) of free YrdC (2 μM in 20 mM HEPES, 100 mM NaCl, pH 7.5) was taken. Defined volumes of a concentrated aqueous solution of tRNA were then added in steps to final tRNA concentrations of 0.05–3.50 μM and the fluorescence signal was taken after a 3-min incubation period at each step to allow for complete binding. To correct for dilution effects, the same titration was repeated with the addition of the same incremental volumes of water. All readings were corrected for buffer absorption without need for inner filter correction (absorbance at 295 nm by tRNA at its highest concentration was less than 0.05). Theoretical curves were fitted to the experimental data and dissociation constants were determined according to the following equation (36): where I0 and I are the corrected fluorescence intensities of protein and protein–tRNA complex, respectively. P0 and N0 are the total concentrations of protein and tRNA, respectively, f is the fluorescence intensity ratio of bound and free protein, and Kd is the dissociation constant.

Detection of ATP binding by STD-NMR

NMR experiments were performed at 278K on a Varian INOVA 600 NMR spectrometer equipped with a 5 mm triple resonance inverse detection probe. Samples were prepared in 500 µl of 99% D2O PBS buffer at pH 7.5 (uncorrected) containing 10 mM phosphate, 137 mM NaCl, 2.7 mM KCl, 0.5 mM MgCl2 and 0.02% NaN3. Concentration of the NMR sample was 20 µM protein and 1 mM ligand. The STD pulse sequence was implemented as suggested in (37), with minor variations. The excitation train was composed of 50 ms G4 composite pulses with 1 ms delay between pulses. On-resonance and off-resonance irradiations were centered at –1 ppm and 30 ppm, respectively, and the total saturation time was 3 s. Initial FIDs were multiplied by a line broadening function of 1 Hz and zero filled to twice the original data size prior to Fourier transformation. A total of 512 scans with 16 dummy scans were acquired, with a recycle delay of 2 s.

RESULTS

Involvement of the Sua5/YrdC family of proteins in t6A formation

Bioinformatic identification

Analysis of all sequenced tRNAs (38) suggested that all organisms should have a t6A biosynthesis pathway. Using the MGDB database (http://mbgd.genome.ad.jp/), a list of orthologous families present in E. coli K12, S. cerevisiae, B. subtilis 168, Buchnera aphidicola APS, Wigglesworthia glossinidia, Wollbachia endosymbiot mel and all the mollicutes in the database (16 genomes) was generated. This analysis revealed the existence of 95 orthologous protein families, mostly related to translation (ribosomal protein, tRNA synthetases and tRNA modification). Only nine of these families did not have an experimentally verified function when this work was initiated but many have been functionally characterized since. The YrdC/Sua5 (Figure 1A) family was the most probable candidate for an enzyme involved in t6A biosynthesis for the following reasons: (i) all organisms sequenced to date have a homolog of the YrdC/Sua5 proteins, some organisms, such as E. coli, have two representatives (YrdC and YciO), but most have only one; (ii) the YrdC domain is also found in the enzyme family HypF (2) [involved in the maturation of the metal center of the Ni-Fe hydrogenase HypE (39)] which catalyzes chemistries similar to the ones expected in t6A biosynthesis (22); (iii) YrdC and YciO of E. coli were found to bind double-stranded RNA and tRNA (2); (iv) Sua5 in yeast was proposed to be involved in translation (4).

Figure 1.

(A) Schematic representation of the YrdC/Sua5 family. YrdC from E. coli is 192 amino acids, yeast Sua5 is 420 amino acids and the human IRIP is 279 amino acids with 1–55 being a mitochondrial signal peptide. The position of the KxR(∼50)SxN conserved motif is given for the E. coli sequence. (B) Sequence alignment of the YrdC domain of YrdC and Sua5 family proteins The E. coli YrdC, S. tokodaii Sua5 and E. coli YciO sequences are aligned based on 3D alignment of the crystal structures. Secondary structure elements from the crystal structure of E. coli YrdC are shown above the sequences. Red boxes indicate conserved residues in YrdC and YciO families. Residues found to interact with bound AMP in the S. tokodaii Sua5 crystal structure are indicated by red and black stars. K50 and R52, mutated in the present study and used to distinguish between the YciO and the YrdC families, are indicated by red stars. S107, the residue mutated in the yeast sua5-1 mutant is indicated by a #. Ec: E. coli, Mm: M. maripaludis, Bs: B. subtilis, St: Sulfolobus tokodaii, Sc: S. cerevisiae.

In vivo experimental validation

In order to examine the t6A phenotype (i.e. presence or absence of t6A modification in RNA) of strains lacking members of the SUA5/yrdC gene family an existing strain, YJN63 (sua5::LEU2) (4), which carries a deletion allele of SUA5 was analyzed. Bulk tRNA was extracted from both YGN63 and a near isogenic wild-type strain YMH13 (SUA5) (4) and processed for LC–MS/MS analysis as described in the ‘Materials and Methods’ section. This analysis revealed that the 25.5 min peak detected on the UV trace in the parent strain, corresponding to the protonated molecular weight of t6A (MH+= 413m/z), disappeared in the strain carrying the sua5::LEU2 allele (Figure 2A). The t6A peak was recovered when this strain was transformed with a plasmid carrying the wild-type SUA5 gene (plasmid pBN204) confirming that the t6A− phenotype (absence of t6A peak) was due to the loss of SUA5 (Figure 2A).

Figure 2.

(A) LC–MS/MS analysis of yeast tRNA extracted from different strains linking the disappearance of the t6A peak to the deletion of sua5. UV traces and ion extraction chromatograms for 413 m/z (small windows) are shown for WT (YMH13) upper panels, sua5::LEU2 (YJN63) middle panels and YGN63 transformed with pBN204 (SUA5 URA3) lower panels. (B) Complementation of the t6A minus phenotype followed by HPLC analysis of bulk yeast tRNA extracted from various strains. The position of the t6A peak was confirmed by running a t6A standard and spiking the WT sample (data not shown). Strain YGN63 was transformed with URA3 plasmids carrying SUA5 (as a control) or carrying wild-type genes yrdC, yciO and ywlC, respectively and the recovery of the t6A peak was monitored. (C) LC–MS/MS analysis of yeast tRNA extracted from the sua5::LEU2 yrdC strain. UV traces and ion extraction chromatograms for 413 m/z (small windows) are shown. Positive and negative controls were run as presented in (A) (data not shown).

To determine whether sua5 homologs were actual functional orthologs, E. coli, B. subtilis and M. maripaludis yrdC-like genes were tested for their ability to complement the t6A− phenotype of the sua5::LEU2 strain. YGN63 was transformed with URA3 plasmids carrying the following wild-type genes: yrdC (pBY135.1), yciO (pBY136.1), ywlC (pBY137.1) or yrdC (pBY138.1). Bulk tRNA was then extracted and analyzed to check for the presence of t6A in the transformed strains. All genes with the exception of yciO, were able to complement the t6A− phenotype of YJN63 (Figure 2B and C).

Essentiality of yrdC orthologs

Essentiality of S. cerevisae SUA5 wild-type gene

In order to clarify the contradictory data on the essentiality of SUA5, the replacement of the wild-type allele by a sua5-Δ1::LEU2 cassette was attempted in three different backgrounds (see ‘Material and Methods’ section for construction details): the diploid, BY4743 (SUA5/SUA5 ura3Δ0/ura3Δ0 leu2-Δ0/leu2-Δ0), the haploid BY4741 (SUA5 ura3Δ0 leu2-Δ0) and VDC5309 (SUA5 yrdC) which is strain BY4741 transformed with a yrdC carrying plasmid (pBY135.1). Leu+ clones were obtained in all three backgrounds, however, when analyzed by PCR, only clones derived from the diploid BY4743 and the merodiploid VDC5309 carried the sua5-Δ1::LEU2 marker in the right locus, yielding SUA5/sua5-Δ1::LEU2 and sua5-Δ1::LEU2 yrdC derivatives, respectively (Figure 3A). In the haploid background BY4741, ectopic insertion of the cassette occurred yielding SUA5 LEU2 strains, but no correct insertion of the deletion cassette could be obtained. Thus, SUA5 could only be disrupted in backgrounds where a functional copy remained present in the genome, suggesting that SUA5 is an essential gene in yeast, and in agreement with the result of the systematic yeast gene deletion study (11).

Figure 3.

(A) PCR amplification to confirm the presence and correct location of the sua5-Δ1::LEU2 allele in various yeast strains. Lanes 1, 3 and 5: primers Chksua5leu.ol4 and Chksua5leu.ol5 and lanes 2, 4 and 6: primers Chksua5leu.ol4 and Leucassol3. Strains VDC5417 (BY4741 derivative SUA5 LEU2 URA3 yrdC) lanes 1/2; VDC5300 (BY4743 derivative (SUA5/sua5-Δ1::LEU2) lanes 3/4; VDC5416 (BY4741 derivative sua5-Δ1::LEU2 yrdC) lanes 5/6. Strains were tested with both primer combinations. In VDC5417 the cassette has inserted ectopically, i.e. Leu+ and negative in PCR tests using the Chksua5leu.ol4/Leucassol3 primer set. For both VDC5033 and VDC5416 insertion of the cassette at the correct location was confirmed. Expected sizes are: 750 bp for sua5-Δ1::LEU2 allele at the right locus using the Chksua5leu.ol4/Leucassol3 combination and 2300 bp for the wild-type allele SUA5 and 2800 bp for the sua5-Δ1::LEU2 allele using the Chksua5leu.ol4/Chksua5leu.ol5 primer combination. Annealing positions for primers used are shown at the bottom of panel A. Only relevant genotypes are shown, refer to Table 1 for completeness. (B) FOA curing of pBY135.1 (URA3 yrdC) from strains VDC5300 (SUA5/sua5-Δ1::LEU2), VDC5417 (SUA5 LEU2 URA3 yrdC) and VDC5416 (sua5-Δ1::LEU2 yrdC). 20 μl of serial dilutions of saturated cultures of VDC5416, VDC5417 and VDC5300 were replica-plated on SD Gal/Raf -ura -leu and FOA plates and analyzed after 5 days. Strains carrying the URA3 marker are unable to grow on 5-FOA. VDC5300 was used as a control for auxotrophies. (C) Verification of cured strains by PCR using the plasmid specific primers DEST52.ol1 and DEST52.ol2: lane 1, VDC5417; lane 2, VDC5416; lane 3 and 4 the corresponding respective FOA treated derivatives.

In order to further demonstrate the essentiality of SUA5, attempts were made to dissociate the sua5-Δ1::LEU2 allele from the URA3 yrdC plasmid pBY135.1. If SUA5 is essential then the plasmid should not be curable from VDC5416 (sua5-Δ1::LEU2 pBY135.1), while it should be readily lost from VDC5417 (SUA5 LEU2 pBY135.1). Since the URA3 allele carried on the plasmid confers sensitivity to 5-FOA, strains were tested for their ability to grow in the presence of this compound. Culture dilutions of VDC5416 were plated on 5-FOA medium and the resulting growth phenotype compared to that of two control strains: VDC5417 carrying an ectopic insertion of the deletion cassette and the heterozygote VDC5300 (SUA5/sua5-Δ1::LEU2 ura3Δ0/ura3Δ0). Derivatives from both control strains grew in presence of 5-FOA as expected. After prolonged incubation, 5-FOAR derivatives of sua5-Δ1::LEU2 pBY135.1 were also recovered suggesting that pBY135.1 could be eliminated from this background (Figure 3B). Both the curing of pBY135.1 and the sua5-Δ1::LEU2 genotype were confirmed by PCR (Figure 3C) and by plating on SD Gal/Raf agar plates lacking either uracil/leucine or leucine (data not shown). These results showed that the SUA5 gene is not essential in yeast confirming results of a previous study (4). Further analysis of the growth phenotype of sua5-Δ1::LEU2 strains confirmed the growth defect phenotypes observed on FOA plates. Growth experiments on YPD show a dramatic decrease in growth rate of cured strain VDC5440 compared to the haploid parent strain BY4741 (Figure 4A). A plasmid carrying SUA5 under the control of its endogenous promoter region (pBY176.3) complemented both the t6A deficiency of VDC5440 and the slow growth phenotype (Figure 4B and C), confirming that it was not due to a secondary mutation. These results show that neither the gene nor the modification is essential in yeast. The failure to obtain the haploid mutant by direct transformation is most likely due to its reduced growth phenotype.

Figure 4.

(A) Growth phenotype of the VDC5440 (sua5-Δ1::LEU2 ura3Δ0) compared to the WT parent BY4741 (B) Growth phenotype of VDC5440 transformed with plasmid pRS313 carrying SUA5 under the control of its endogenous promoter region (pBY176.3) or empty plasmid control pRS313 on SD Gal/Raf -his. All growth experiments were performed at 30°C. Shown here are representative plots out of ten replications. (C) LC–MS/MS analysis of tRNA extracted from yeast strains showing the absence or presence of the t6A peak on the UV trace (left side) and extraction chromatogram (right side) for VDC5440 transformed with pBY176.3 (upper panels) and VDC5440 transformed with pRS313 (lower panels).

Essentiality of E. coli yrdC

To test whether the yrdC gene is essential or not in E. coli, a similar approach as that employed for yeast was followed. Homologous replacement of the wild-type yrdC allele by a ΔyrdC::KmR allele was attempted by direct transformation of a UpyrdC::KmR::DnyrdC PCR generated fragment into three different backgrounds: E. coli NM1100, NM1100/pBAD24 and NM1100/pBAD24::yrdC (pBY126) (see ‘Material and Methods’ section), the latter strain contains a functional copy of yrdC on a plasmid. Km resistant clones with insertion of the marker in the expected location (as verified by PCR) were obtained in NM1100/pBAD24::yrdC but not in the two other backgrounds suggesting that the chromosomal copy of the gene can only be eliminated if its function is insured by a gene is trans (Figure 5A).

Figure 5.

(A) PCR amplification to confirm the presence of the ΔyrdC::KmR allele in kanamycin-resistant NM1100/pBAD24::yrdC derivatives. Two independent KmR strains (VDC5173, lanes 1, 4 and 7 and VDC5174 lanes, 2, 5 and 8) obtained from transformation of NM1100/pBAD24::yrdC with the PCR generated ΔyrdC::KmR cassette and the NM1100/pBAD24::yrdC background strain (lanes, 3, 6 and 9) were subjected to PCR analysis using the primer combinations UpyrdC/K2 (lanes 1, 2 and 3), DsyrdC/K1 (lanes 4, 5 and 6) and UpyrdC/DnyrdC (lanes 7, 8 and 9). The expected sizes for the UpyrdC/DsyrdC amplicons in WT and mutant strains are 2340 and 3113 bp respectively, DsyrdC/K1 and UpyrdC/K2 should give no amplification for the NM1100 control. (B) Growth phenotype of MG1655 ΔyrdC::KmR P1 transductants. Strains were streaked on LB Amp plates supplemented or not with 0.02% arabinose and incubated 24 h at 37°C. MG1655/pBAD24 was used as control.

The essentiality of yrdC was further demonstrated by P1 transduction experiments. Strain VDC5173 (NM1100 ΔyrdC::KmR/pBAD24::yrdC) was used as donor for P1 transduction with the following strains used as recipients: MG1655 transformed with pBAD24 derivatives carrying either functional yrdC homologs, yrdC, ywlC, yrdC or SUA5 (pBY126, pBY125, pBY167 and pBY130 respectively) or carrying the non-functional homolog yciO (pBY128). P1 transductions were carried at least five times in two independent MG1655 backgrounds for each ortholog, Km-resistant colonies were recovered in all backgrounds (with a variation of 120–500 colonies per transduction) except MG1655/pBAD24 and MG1655/pBAD24::yciO that never yielded any colonies (while non-essential markers were readily transduced in these strains). These data confirm that yrdC is an essential gene in the E. coli strains MG1655 and NM1100, and supports the result obtained in yeast, namely that SUA5, yrdC, ywlC are functional orthologs but yciO.is not. The pBAD construct places the genes under the control of an arabinose inducible promoter and among the orthologs, SUA5 was the least proficient in replacing the yrdC gene. In the absence of the inducer, the E. coli strain carrying ΔyrdC1/pBY130 (pBAD24::SUA5) showed a delayed growth phenotype not observed for strains carrying the three other genes (Figure 5B).

Analysis of bulk tRNA extracted from all four classes of P1 transductants obtained showed that all contained t6A indicating that these orthologs complement both the essentiality and tRNA modification phenotypes (data not shown).

Direct involvement of YrdC in t6A biosynthesis

Preferential binding of YrdC to apomodified tRNAThr

To investigate whether members of the YrdC/Sua5 family act directly on tRNA in t6A biosynthesis, YrdC binding to a tRNA species known to contain t6A in E. coli, tRNAThr, was tested. Three forms of E. coli tRNAThrCGU were generated; two in vivo expressed forms and one in vitro transcribed form that lacked all post-transcriptional modifications. As seen previously by other investigators (40), overexpression of tRNAThr produced two forms that differ by the exact molecular mass of carbamoyl-threonine suggesting that they differ by the presence or absence of this modification at nucleotide 37. Both forms were aminoacylated with ThrRS to similar degrees indicating that they were properly folded, functional tRNAs. Fluorescence quenching data show that YrdC binds to fully modified tRNAThr and to partially modified tRNAThr (lacking only the threonyl-carbamoyl modification) with estimated dissociation constants of 0.62 ± 0.13 and 0.11 ± 0.05 μM, respectively, indicating high binding affinity (Figure 6A). Only weak and apparently non-specific binding between YrdC and unmodified in vitro transcript of tRNAThr could be detected. These results suggest that YrdC is involved in t6A biosynthesis by interacting directly with tRNAs, consistent with the previously reported binding of YrdC to total tRNA (2).

Figure 6.

(A) YrdC binding to tRNAThr measured by the intrinsic fluorescence quenching assay. Relative Trp fluorescence intensity from YrdC as a function of partially modified in vivo expressed tRNAThr lacking the t6A37 modification (filled square), fully modified in vivo expressed tRNAThr (filled circle) and an unmodified tRNAThr transcript (filled triangle). Sample conditions: 2 µM protein, 20 mM HEPES, pH 7.5, 100 mM NaCl, 20°C. λex = 295 nm, λem = 345 nm. (B) STD-NMR spectra of nucleotides in the presence of E. coli YrdC. For the nucleotides ATP, ADP, CTP and UTP two traces are shown, where the lower trace is the reference 1D NMR spectrum, and the upper trace is the STD spectrum obtained by irradiation of the aliphatic region of YrdC. Ligand binding is indicated by positive signals in the STD spectrum, which is particularly strong for the H8, H2 and H1 resonances of ATP and ADP in the 6–8 ppm region. Little or no binding is observed for CTP or UTP.

Binding of YrdC to ATP

t6A biosynthesis in crude enzyme preparations has been shown to require ATP (22) and the Sua5St ortholog from Sulfolobus tokodaii has been recently found to bind and hydrolyze ATP (41). To investigate whether YrdCEc can specifically bind ATP, we screened various nucleotides for their potential to bind YrdC by saturation-transfer difference NMR (STD-NMR) spectroscopy. A strong STD-NMR signal was seen with ATP and ADP but not with CTP and UTP, indicating adenine-specific binding (Figure 6B), and pointing to the presence of one or more ATP-binding sites in YrdCEc.

Essentiality of E. coli YrdC conserved residues Arg52 and Lys50

Arg52 and Lys50 residues are strictly conserved among YrdC proteins but are absent from the YciO proteins (Figure 1B). Both residues are located in the center of a positively charged cleft that is the putative nucleic-acid-binding site, and based upon the crystal structure of the S. tokodaii protein Sua5St, their side chains (Arg59 and Lys57 using S. tokodaii residue numbering) interact with and stabilize an AMP derived phosphate group found bound in the cleft. (41). This suggested that nucleotide binding by YrdC and Sua5 (41) is critical for function. To test the functional significance of these two residues in vivo, site-directed mutagenesis was used to construct E. coli strain MG1655 carrying pBAD24 plasmids harboring yrdC Lys50 → Ala or Arg52→Ala, and yrdC Arg110→Ala or Lys53→Ala for control (pYC102.3, pYC104.2, pYC101.1 and pYC103.1, respectively). These were used in P1 experiments as recipient strains as described earlier. More than a hundred clones were recovered from P1 experiments when the control strain MG1655pyrdC, MG1655pyrdC Arg110→Ala or MG1655pyrdC Lys53→Ala were used as recipients, but no transductants were recovered when the MG1655/pYC104.2 and MG1655/pYC102.3 were used, implying that Arg52 and Lys56 are essential for YrdC function. All receiver strains were tested for sensitivity to P1. These results suggest that nucleotide binding is essential for the function of these proteins and are consistent with the known ATP requirement for t6A biosynthesis (22).

DISCUSSION

Our results indicate that the universally conserved YrdC/Sua5 family is involved in the biosynthesis of N6-threonylcarbamoyl adenosine. Deletion of SUA5 led to the loss of t6A in two independent yeast backgrounds. E. coli YrdC preferentially bound tRNAThr containing all modifications but t6A as compared to in vitro (unmodified) transcribed tRNA. Two invariant residues involved in nucleotide binding were characterized, strongly indicating that the role of YrdC/Sua5 in t6A biosynthesis is mediated by their ATP binding and hydrolyzing functions, and further suggesting a direct role for this protein in the biosynthetic pathway, since ATP is required for t6A formation (22).

The YrdC/Sua5 proteins have been assigned to COG009 and most organisms have only one member of this family. Exceptions include plants and several bacteria that have two homologs. Expression of the second COG009 encoding gene of E. coli, yciO, failed to complement the t6A minus phenotype of yeast strains lacking a functional SUA5 gene. The two subfamilies can also be distinguished by sequence analysis as residues K50 and R52 are found at the bottom of a positively charged concave surface in the crystal structure of YrdC and are strictly conserved in the YrdC subfamily but absent in YciO sequences (Figure 1B). These residues are essential for YrdC function as shown by site directed mutagenesis and complementation experiments. Therefore COG009 should be split into two sub-families as the results presented here suggest distinct functions for YciO and YrdC.

Like m1G37 and Ψ55, t6A37 is one of the few tRNA modifications present in all organisms for which the whole set of tRNA sequences is available (38,42). It is one of the few universal modifications believed to have been present in the last common ancestor (43,44) and every sequenced genome (>600 to date) has a YrdC ortholog.

In this study the essentiality of this gene family has been resolved. In E. coli, yrdC appears essential, however whether the observed lack of viability is due to the absence of the t6A modification still remains to be established. In apparent contradiction to our results, insertion of the PBAD promoter within the first 15 nucleotides of yrdC was reported to be lethal under repressing conditions, namely on minimal but not rich medium (10). The phenotype on minimal medium was attributed to polar repression of the downstream aroE gene (aromatic amino-acid biosynthetic gene), which is essential for growth on minimal medium, whereas on LB neither aroE nor yrdC was considered essential. However, since expression from the pBAD promoter is not completely repressed by glucose (27,45), expression under pBAD repression may yield sufficient YrdC protein to fulfill essential function(s) in the cell.

In yeast, neither the gene nor the modification is essential, as two SUA5 deletion strains lacking t6A have been constructed [this study and (4)]. However, t6A− strains display a marked slow growth phenotype [this study and (4)] and possess defective mitochondria (4). Since mitochondria are considered of prokaryotic origin, this result provides additional evidence that this modification plays an essential role in prokaryotic translation.

The t6A modification has been shown to be a determinant for efficient aminoacylation by bacterial Isoleucyl-tRNA synthetase (IleRS), whereas it is not for yeast cytoplasmic IleRS (46,47) and this might be the molecular basis for the requirement of YrdC and Sua5 in bacteria and the yeast mitochondria, respectively.

Four mitochondrial tRNAs contain t6A (Lys, Ile, Arg2 and Met) however only one yrdC/SUA5 homolog has been identified in yeast, and this protein is predicted to be cytoplasmic (48). In the absence of a mitochondrial specific homolog, Sua5 must also be targeted to the mitochondria, possibly by alternative translational start site usage as has been described for proteins such as yeast Trm5, HTS1, VAS1 and Arabidopsis thaliana alanyl-, threonyl- and valyl-tRNA synthetases (49–52).

Strains lacking a member of the YrdC/Sua5 family resulting in a defective pool of tRNAs, will be instrumental in studying the in vivo role of this modification. Accurate codon–anticodon recognition is critical in insuring the accuracy of translation and the t6A37 modification has been shown to affect this process in vitro, mainly by preventing the formation of U33-A37 across-the-loop base pairing interaction, and allowing cross-strand stacking of A38 and t6A37 with the first position of the codon (15,16,18–20). The t6A modification is found in seven E. coli and 12 S. cerevisiae tRNAs, four of which are mitochondrial (38). As the effects of anticodon-surrounding modifications in adjusting the thermodynamic properties of anticodon/codon interactions for cognate and none-cognate recognition by the ribosome are context dependent, it is expected that the absence of t6A in these tRNAs will affect translation in several ways. This is reflected by the pleiotropic phenotypes of strains lacking a functional YrdC or Sua5 (4–8). In yeast, for example, the sua5-1 mutation was isolated as a suppressor of an initiation defect affecting the CYC1 gene expression (4). CYC1 is a nuclear gene encoding the iso-1 form of the electron carrier protein cytochrome c which is localized to the mitochondria (53). The cycl-1019 mutant expresses 2% of the normal amount of iso-l-cytochrome c due to a single base-pair substitution that creates an upstream ATG codon that is out of frame with the normal CYCl translational start. The sua5-1 point mutation suppressed this translational defect, restoring Cyc1 levels to 60% of the normal levels (4). In this sua5 suppressor strain lacking t6A (this study), Cyc1 protein expression is possibly due to high levels of frameshift errors during the translation initiated at the upstream and out of frame ATG, thus restoring the original reading frame. The absence of other modifications at position 37 such as Wyeosine or m1G has also been shown to drastically increase frame-shifting in yeast (54,55). However, the fact that the Cyc1 protein is also expressed at nearly of wild-type levels in a wild-type CYC1 sua5::LEU2 strain (4), i.e. in the absence of the aberrant ATG found in the cyc1-1019 allele, suggests yet another hypothesis. In Eukaryotes, the cytoplasmic initiator tRNA contains a t6A at position 37, and is restricted to recognition of AUG alone, while its prokaryotic counterpart is neither modified nor restricted to recognizing AUG, suggesting that this modification may play an important role in selecting the appropriate translational start site. In the absence of t6A, the scanning ribosome could initiate at the natural ATG start codon, but also promiscuously at other ATG or non-ATG codons, as interactions between initiator tRNAs and their codons would be less stringent. This would account for the situation in prokaryotes, where the initiator tRNA is not modified at position 37 and the ribosome can initiate at either ATG, GTG or TTG (56). For yeast, the lack of stringent translational start site selection would lead to several N-terminal variants, including active forms of the Cyc1 protein and would account for the similar levels of Cyc1 activity seen in both the sua5::LEU2 strain and the cyc1-1019 sua5-1 suppressor strain.

Finally, other pleiotropic phenotypes linked to the absence of YrdC/Sua5 members have been observed. In E. coli, yrdC has been described as a possible ribosome maturation factor (RimN) needed for correct processing of 16S ribosomal RNA, and for maintenance of normal amounts of translating ribosomes (7,8). The ribosome maturation phenotype may reflect an indirect consequence of the yrdC mutation since tRNA defects due to loss of modifications(s) or other cellular malfunctions have been shown to cause ribosome maturation defects (57).

In conclusion, we propose that the diverse phenotypes of YrdC/Sua5 mutant strains are due to the central and primordial role of the t6A modification that includes: insuring accurate decoding, maintaining the reading frame, selecting the appropriate translation initiation codon (mainly in Eukaryotes) and, determining recognition and aminoacylation of tRNAs by prokaryotic and mitochondrial IleRSs.

Although a biochemical function for the YrdC/Sua5 proteins (i.e. catalytic activity) aside from tRNA and ATP binding (this study) and ATPase activity (41), has not been determined, appropriately modified tRNA substrates appear critical for interacting with these proteins. We were unable to show threonine binding by YrdC, under ATP binding conditions, nor incorporation of threonine in a tRNAThr transcript in the presence of YrdC (data not shown). Further experimentation using apomodified tRNA as a substrate is needed to determine the role of YrdC in t6A biosynthesis. It is also likely that other protein partners interact with YrdC especially since the biosynthesis of t6A is proposed to require two ATP-dependent steps (58).

FUNDING

National Science Foundation (grant no MCB-05169448); U.S. Department of Energy (grant no. DE–FG02–07ER64498) [to V.D.C.]; National Institutes of Health (GM-23562) [to P.S.]. Funding for open access charge: U.S. Department of Energy (grant no. DE–FG02–07ER64498).

Conflict of interest statement. None declared.