The TRAMP Complex Shows tRNA Editing Activity in S. cerevisiae

Molecular Biology and Evolution, May 2012

Transfer RNA (tRNA) editing is a widespread processing phenomenon that alters the sequence of primary transcripts by base substitutions as well as nucleotide deletions and insertions at internal or terminal transcript positions. In the corresponding tRNAs, these events are an important prerequisite for the generation of functional transcripts. Although many editing events are well characterized at the reaction level, it is unclear in most cases from which ancestral activities the modern editing enzymes evolved. Here, we show that in Saccharomyces cerevisiae, the noncanonical poly(A) polymerase Trf4p in the TRAMP complex can be recruited for such an editing reaction at an introduced tRNA transcript. As a distributive polymerase involved in RNA surveillance and quality control, it has a broad substrate spectrum and binds only transiently to the transcripts, limiting the number of added nucleotides at the editing position. These features exactly meet the criteria for an ancestral enzyme of a modern editing activity. Accordingly, our observations are a strong experimental support for the hypothesis that enzymatic promiscuity serves as an evolutionary starting point for the emergence of new functions and activities.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://mbe.oxfordjournals.org/content/29/5/1451.full.pdf

The TRAMP Complex Shows tRNA Editing Activity in S. cerevisiae

Helena Dickinson 0 Sandy Tretbar 0 Heike Betat 0 Mario Morl 0 0 Institute for Biochemistry, University of Leipzig , Leipzig, Germany Transfer RNA (tRNA) editing is a widespread processing phenomenon that alters the sequence of primary transcripts by base substitutions as well as nucleotide deletions and insertions at internal or terminal transcript positions. In the corresponding tRNAs, these events are an important prerequisite for the generation of functional transcripts. Although many editing events are well characterized at the reaction level, it is unclear in most cases from which ancestral activities the modern editing enzymes evolved. Here, we show that in Saccharomyces cerevisiae, the noncanonical poly(A) polymerase Trf4p in the TRAMP complex can be recruited for such an editing reaction at an introduced tRNA transcript. As a distributive polymerase involved in RNA surveillance and quality control, it has a broad substrate spectrum and binds only transiently to the transcripts, limiting the number of added nucleotides at the editing position. These features exactly meet the criteria for an ancestral enzyme of a modern editing activity. Accordingly, our observations are a strong experimental support for the hypothesis that enzymatic promiscuity serves as an evolutionary starting point for the emergence of new functions and activities. - Transfer RNA (tRNA) editing events play an important role in generating functional tRNA molecules that participate in translation, ensuring the viability of the cell. Such events are found in mitochondria of many organisms, ranging from single cells (Acanthamoeba, Spicellomyces) to higher eukaryotes (Schuster and Morl 2004). Besides internal base insertions and conversions, additions of nucleotides at 5#- as well as 3#-ends of tRNAs have been described. In these cases, the corresponding tRNA genes either encode mismatched or truncated acceptor stems that frequently result from overlapping gene organizations, where a tRNA gene shares 16 nt with its neighboring upstream or downstream gene encoded on the same strand. The observed editing events are required to correct or complete these transcripts (Schuster and Morl 2004). In the case of 5#editing, nucleotide additions occur in the nonconventional 3#- to 5#-direction, whereas all known nucleotidyl transfer reactions proceed in the opposite direction (Lonergan and Gray 1993a, 1993b; Laforest et al. 1997; Price and Gray 1999; Gott et al. 2010). Hence, the corresponding editing enzyme has close similarity to histidine tRNA guanylyltransferase that adds a single G residue to the 5#-end of tRNAHis (Price and Gray 1999; Bullerwell and Gray 2005; Abad et al. 2011). Insertional editing events at the 3#-end of tRNAs are also required for the removal of mismatches in acceptor stems (Lavrov et al. 2000; Leigh and Lang 2004). However, in most cases, 3#-editing restores truncated tRNA transcripts. In the majority of these events, the missing nucleotides are restored as A residues (Yokobori and Paabo 1995a, 1995b, 1997; Tomita et al. 1996; Reichert et al. 1998; Lavrov et al. 2000). In human mitochondria, for example, an overlap of one A residue was found between the tRNA genes for tyrosine and cysteine (Reichert et al. 1998). In the maturation of the primary transcript, the downstream tRNA is released as a complete transcript, whereas the upstream tRNA is missing the overlapping nucleotide at its 3#end. Subsequently, an unidentified nucleotide adding activity restores the missing A residue, converting the transcript into a functional tRNA. As this position represents an important identity element of tRNATyr (Fechter et al. 2000), it is crucial that the editing reaction takes place at high accuracy. Although overlapping tRNA genes are common in metazoans, Saccharomyces cerevisiae does not show such gene organizations. However, when tRNA genes overlapping for one position (comparable to the situation in human mitochondria) are introduced, yeast can process these transcripts and edit the resulting 3#-truncation of the upstream tRNA by inserting the missing A residue at high fidelity (Schuster et al. 2005). Hence, the situation in the yeast nucleus corresponds exactly to the evolutionary hypothesis of tRNA editing (Covello and Gray 1993; Cavalier-Smith 1997; Simpson et al. 2000): A preexisting promiscuous nucleotide inserting activity is recruited to catalyze an editing reaction on a new substrate (Schuster et al. 2005; Khersonsky and Tawfik 2010). Here, we show that this editing event is carried out by the noncanonical poly(A) polymerase Trf4p (originally identified as a topoisomerase 1 requiring function; Sadoff et al. 1995) as an integral part of the TRAMP4 complex. TRAMP stands for Trf4 Air2 Mtr4 polyadenylation (describing the individual The Author 2011. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. All rights reserved. For permissions, please e-mail: . components; LaCava et al. 2005), and the complex is involved in RNA surveillance, labeling incorrectly processed transcripts with poly(A) tails for degradation. It consists of one of the two poly(A) polymerases Trf4p (in the case of TRAMP4) or Trf5p (TRAMP5) as well as one of the RNAinteracting Zinc-knuckle proteins Air1p or Air2p (Arginine methyltransferaseinteracting RING finger protein 1 or 2) and the RNA helicase Mtr4p (mRNA transport) (Liang et al. 1996; Jensen and Moore 2005; LaCava et al. 2005; Vanacova et al. 2005; Wyers et al. 2005). In the case of the editing substrate, the Trf4p polymerase does not add a true poly (A) tail as a degradation tag (LaCava et al. 2005) but incorporates a very limited number of A residues that are compatible with the generation of a stable and mature tRNA. Thus, this involvement of the RNA surveillance complex TRAMP4 in tRNA 3#-editing in yeast represents a vivid example for molecular exaptation. Materials and Methods Identification of RNA-Specific Nucleotidyltransferases For the identification of yeast enzymes with nucleotidyltransferase activity, the following PFAM domains were used as a query at the InterPro web interface (http://www.ebi.ac .uk/interpro/): PF01909 (nucleotidyltransferase domain), PF00483 (nucleotidyltransferase), PF00136 (DNA-directed DNA polymerase, family B, multifunctional domain), PF00476 (DNA-directed DNA polymerase, family A, palm domain), PF00562 (DNA-directed RNA polymerase, subunit 2, domain 6), PF00623 (RNA polymerase, alpha subunit), and PF01743 (poly(A) polymerase, head domain). The resulting 119 S. cerevisiae entries were investigated for function using UniProtKB (http://www.uniprot.org/ help/uniprotkb). RNA editingunrelated proteins (DNA polymerases, template-dependent polymerases, metabolic enzymes, translation factors) were removed, leading to a list of four nucleotidyltransferases that were described as RNAspecific enzymes (Martin and Keller 2007). Other RNAdependent nucleotidyltransferases like terminal uridylyl transferase, poly(U) polymerase, or 2#-5#-oligo adenylate synthetase were not identified. Yeast Strains BY4741 (MAT a; his3D1; leu2D0; met15D0; ura3D0), BY4741 air1D (MAT a; his3D1; leu2D0; met15D0; ura3D0; YIL079c::kanMX4), and BY4741 air2D (MAT a; his3D1; leu2D0; met15D0; ura3D0; YDL175c::kanMX4) were purchased from EUROSCARF, Frankfurt, Germany. BY4741 trf4D (MAT a; his3D1; leu2D0; met15D0; ura3D0; YOL115W::nat), and BY4741 trf5D (MAT a; his3D1; leu2D0; met15D0; ura3D0; YNL299W::kanMX) were a gift by C. Rammelt, University of Basel, Switzerland. BY4741 air1D/air2D (MAT a; his3D1; leu2D0; lys2D0; ura3D0 YIL079c::kanMX4 YDL175c::natMX6) was a gift by David Tollervey, University of Edinburgh, Scotland. The TRF4 TAP-tagged yeast strain (endogenous promoter; MATa; his3D1; leu2D0; met15D0; ura3D0; YOR001w::kanMX4) was purchased from Open Biosystems (Thermo Fisher Scientific, England). Construction of Overlapping tRNA Precursor Overlapping tRNA genes for tyrosine and cysteine were generated and amplified by overlap extension polymerase chain reaction (PCR) (Schurer et al. 2002; Schuster et al. 2005). sup-tRNATyr(-1)/tRNACys: GGTGGGGTTCCCGAGCGGCCAAAGGGAGCAGACTCTA AATCTGCCGTCATCGACTTCGAAGGTTCGAATCCTTCC CCCACCGGCGCGTTAACAAAGCGGTTATGTAGCGGAT TGCAAATCCGTCTAGTCCGGTTCGACTCCGGAACGCG CCTCCA. Anticodons are presented in underlined italics, and the overlapping position is indicated in bold and underlined, representing the first nucleotide of tRNACys. The PCR product was inserted downstream of the galactose promoter of the plasmid Y352_gal1 and verified by sequencing. The correct plasmid was introduced in the indicated S. cerevisiae strains, and the overlapping tRNA cassette was expressed by galactose induction for 48 h at 30 C in uracil-dropout medium (Schuster et al. 2005). Preparation of RNA Cells were harvested by centrifugation at 2,800 rcf for 5 min and resuspended in 1 ml TRIzol Reagent (Invitrogen, Darmstadt, Germany). After addition of 0.6 g of Lysing Matrix C Bulk (MP Biomedicals, LLC, Illkirch, France), cells were homogenized in a FastPrep-24 instrument (MP Biomedicals, LLC) for 40 s at setting 6.0 m/s. Upon centrifugation, the upper phase was chloroform extracted. Following ethanol precipitation and washing with 70% ethanol, total RNA was resuspended in water and stored at 20 C until use. Sequence Analysis of tRNA 3#-termini tRNA was ligated to a DNA oligonucleotide (Purimex, Gottingen, Germany) carrying an RNA nucleotide at the phosphorylated 5#-end (tag oligo 5#-pUGG ATC GCG TAG CTC ATA CGA GTT-3#) (Betat et al. 2004). cDNA synthesis was carried out by annealing primer P1 (5#-AAC TCG TAT GAG CTA CGC GAT C-3#) to the tag oligonucleotide and using M-MLV Reverse Transcriptase (Fermentas Molecular Biology Tools; Thermo Fisher Scientific, St Leon-Rot, Germany) according to the manufacturers instructions. cDNA was amplified in a standard PCR reaction using primer P1 and a primer specific for Escherichia coli suppressor tRNATyr (5#-GGT GGG GTT CCC GAG CGG CC-3#) (Betat et al. 2004). PCR products were cloned using Clone JET PCR Cloning Kit (Fermentas Molecular Biology Tools; Thermo Fisher Scientific). DNA sequencing was performed on an ABI Prism 3700 automated sequencer (Amersham Pharmacia Biotech, Freiburg, Germany). The sequences were analyzed using the program SeqMan. TAP Tag Purification Trf4-TAPtagged strain was grown in 2 l YPD to OD600 3.0 3.5 and harvested by centrifugation, washed with H2O, and once with TAP buffer (50 mM TrisHCl, pH 7.5, 100 mM NaCl, 1.5 mM MgCl2, 0.15% NP40). For lysis, cells were resuspended in TAP buffer (containing 1 mM DTT, 1.3 lg/ml pepstatin A, 0.28 lg/ml leupeptin, 170 lg/ml PMSF, and 330 lg/ml benzamidine) and mixed with two volumes of glass beads. Lysis occurred while shaking in a bead mill (Fritsch, Idar-Oberstein, Germany). Glass beads were removed, and the lysate was cleared by centrifugation. The lysate was collected and incubated for 4 h at 4 C in the presence of 0.4 ml IgG sepharose (GE Healthcare, Germany). Beads were centrifuged and transferred to a mobicol column with 35 lm filter. Beads were washed with 10 ml TAP buffer (with 150 mM NaCl) containing 0.5 mM DTT. The protein complex was eluted by addition of 4 ll TEV protease (0.210 lg/ll) and 150 ll TAP buffer (0.5 mM DTT). The eluate was adjusted to a final glycerol concentration of 510% and stored at 80 C. Preparation of RNA Substrate The 3#-truncated version of the E. coli suppressor tRNATyr was prepared by standard in vitro transcription using T7 RNA polymerase and a PCR-derived DNA template. In the transcription unit, the sequence of the Hepatitis delta virus (HDV) ribozyme was located immediately downstream of the tRNA sequence. The ribozyme-mediated cleavage released the truncated tRNA with precise and homogeneous 3#-ends. The 2#,3# cyclic phosphate generated by the HDV ribozyme was removed by incubation with T4 polynucleotide kinase in 100 mM Tris/HCl pH 6.5, 100 mM magnesium acetate, and 5 mM b-mercaptoethanol in a reaction volume of 50 ll for 6 h at 37 C (Schurer et al. 2002). Polyadenylation Assay Polyadenylation reaction was carried out in the presence of the TAP-tag purified TRAMP4 complex and ATP (25 mM) in a reaction volume of 20 ll (Vanacova et al. 2005). Reactions were incubated at 37 C; products were separated by denaturing PAGE and visualized by autoradiography. Mass Spectrometric Analysis of TAP-TagPurified TRAMP4 Complex The complete preparation of TAP-purified TRAMP4 as shown in supplementary figure S1 (Supplementary Material online) was isolated from the sodium dodecyl sulfate polyacrylamide gel using an in-gel digestion protocol with porcine trypsin (Sigma-Aldrich, St Louis) in 10 mM NH4HCO3 at 37 C overnight according to Muller et al. (2010). Protein digestions were analyzed by C18 reversed-phase nanoscale liquid chromatography tandem mass spectrometry (MS) on a NanoAcquity UPLC system (Waters Corporation, Milford) connected to an LTQ-Orbitrap XL ETD hybrid mass spectrometer (Thermo Fisher Scientific, Waltham) as described (Muller et al. 2010). Briefly, samples were concentrated on a trapping column (nanoAcquity UPLC column, C18, 180 lm 2 cm, 5 lm; Waters Corporation, Milford) with 0.1% formic acid at flow rates of 15 ll/min. After 10 min, peptides were eluted onto the separation column (nanoAcquity UPLC column, C18, 75 lm 150 mm, 1.7 lm, Waters Corporation, Milford,). Chromatography was performed with 0.1% formic acid in solvents A (100% water) and B (100% acetonitrile), with peptides eluted over 90 min with a 240% solvent B gradient. Continuous scanning of eluted peptide ions was carried out in a mass range m/z 4001,600, automatically switching to CID-MS/MS mode on the ten most intensive ions exceeding an intensity of 3,000. For MS/MS CID measurements, a dynamic precursor exclusion of 3 min was applied. Mascot Daemon 2.2.2 was used for identification and quantification of proteins. Database searches were carried out against Swiss-Prot database entries for S. cerevisiae (www.expasy.org, UniProt Consortium 2011) using the Mascot search engine. For peptide identification, mass tolerance was set to 8 ppm for precursor ions and 0.5 Da for product ions. Two tryptic missed cleavages, methionine oxidation (optional modification) and cysteine carbamidomethylation (static modification), were considered. Relative quantitation was determined based on emPAI values (Ishihama et al. 2005). Results and Discussion Identification of Editing Candidates The described insertional editing reaction at the 3#-end of tRNAs requires an enzyme with nucleotide adding activity. Therefore, PFAM domains for nucleotidyltransferases and polymerases were used as a query at the InterPro website (http://www.ebi.ac.uk/interpro/; Hunter et al. 2009) in order to identify S. cerevisiae protein entries with corresponding activities. The resulting list of 119 protein entries was analyzed for protein function using UniProtKB (Magrane and Consortium 2011) and cleared from DNA polymerases, template-dependent RNA polymerases, as well as redundant and unrelated entries. As a result, four RNA-dependent enzymes were recognized which correspond to nucleotidyltransferases acting on RNA as described by Martin and Keller (2007). The CCA-adding enzyme (CCA1, systematic name YER168C, according to the Saccharomyces genome database) synthesizes the CCA triplet at the 3#-end of tRNAs, representing the site of aminoacylation. Poly(A) polymerase 1 (PAP1, YKR002W) generates long poly(A) tails on mRNAs, leading to an increased stability and translational efficiency (Wahle 1995). The noncanonical poly(A) polymerases Trf4p (TRF4, YOL115W) and Trf5p (TRF5, YNL299W) add short poly(A) tails to a partially overlapping spectrum of transcripts, tagging these RNAs for degradation (LaCava et al. 2005; Vanacova et al. 2005; Wyers et al. 2005). The CCA-adding enzyme could be excluded as a possible editing candidate, as it was shown in an in vitro study that this enzyme is not able to restore the missing A residue at the editing position (Reichert et al. 1998; Schuster et al. 2005). Instead, it directly adds the CCA triplet to the truncated tRNA. In contrast, poly(A) polymerase 1 represents a possible candidate for editing (Yokobori and Paabo 1997). However, it also had to be excluded from this analysis, as the corresponding yeast knockout strain is not viable (Engel et al. 2010). Hence, we focused on Trf4p and Trf5p as editing candidates. Editing Efficiency in S. cerevisiae trf4D and trf5D As yeast edits a recombinantly introduced 3#-terminally truncated tRNATyr at high efficiency (Schuster et al. 2005), this transcript was chosen to identify the responsible A-inserting activity. To distinguish between introduced and endogenous tRNA transcripts, a construct consisting of a suppressor tRNATyr (sup-tRNATyr) from E. coli overlapping for one residue with the E. coli tRNACys was used (Reichert et al. 1998; Schuster et al. 2005). The overlapping position is represented by a guanosine residue, corresponding to the first position of the downstream tRNACys, whereas the upstream sup-tRNATyr lacks an A residue at the 3#-end, denoted as sup-tRNATyr(-1). In the processing reaction analyzed in the BY4741 wild type strain, the downstream tRNACys is released as a complete transcript, carrying the G residue at position 1. The upstream sup-tRNATyr is produced in a truncated form, lacking the 3#-terminal position 73 (numbering according to Sprinzl et al. 1998) and has to be completed by the addition of an A residue (fig. 1) (Schuster et al. 2005). The construct was cloned into vector Y352_gal1 and expressed in BY4741 wt as well as trf4D and trf5D strains (Schuster et al. 2005). Individual transformed colonies from three to four independent experiments were grown in liquid culture, and total RNA was prepared. The suppressor tRNATyr was isolated from each individual transformant, ligated at the 3#-end to an RNA/ DNA oligonucleotide and amplified by RT-PCR (Schuster et al. 2005). The resulting PCR products were cloned, and individual clones were analyzed by sequencing. Three independent transformants of BY4741 wt were used as a control and showed an efficient editing reaction by A insertion at the 3#-end of the recombinant tRNA. On average, 55.9% of the transcripts carried the restored A, most of them followed by partial or complete CCA termini (fig. 2). Only very few clones showed misincorporation of C residues at the editing site or carried additional nucleotides different from the CCA sequence. Forty-anda-half percent of the analyzed clones represented the truncated suppressor tRNATyr(-1) as it was released from the overlapping construct. Efficiency and accuracy of this editing event are in good agreement with the previously described activity (Schuster et al. 2005). In the trf4D strain, the ratio was dramatically altered. Eighty-threepoint-three percent of the clones (resulting from four independent experiments) showed no editing, whereas only 5.4% carried an A residue at the editing site (again, some clones showed misincorporation of C residues). In contrast, deletion of TRF5 had little effect on the editing event. Here, the number of clones corresponding to correctly edited tRNA substrates with or without CCA ends (partial or complete) remained almost unchanged (58.7%). CTP misincorporation and addition of extra nucleotides were observed only in a very limited number of clones. Hence, these results indicate that the deletion of TRF5 has no effect on the editing reaction, whereas TRF4 is strongly involved and has a dramatic impact on the editing efficiency (fig. 2B). Air1p and Air2p Are Required for tRNA Editing Trf4p as well as Trf5p are bipartite poly(A) polymerases that do not have an RNA-binding domain. Instead, both enzymes are (mutually exclusive) parts of the TRAMP complex, interacting with one of the zinc-knuckle proteins Air1p or Air2p (Inoue et al. 2000; Vanacova et al. 2005). These proteins are described as RNA-binding subunits of TRAMP. Hence, it was tested whether these RNAinteracting proteins are required for an efficient tRNA 3#-end editing reaction. The overlapping tRNA construct was expressed in BY4741 strain lacking either AIR1 (air1D), AIR2 (air2D), or a combination of both (air1D/air2D). The editing substrate tRNATyr(-1) was analyzed as described above in three independent experiments for each strain. Deletion of AIR1 led to a moderate reduction of editing efficiency down to 38.4% (fig. 3). Again, the edited tRNA was further elongated by CCA addition as well as some rare extra nucleotide incorporations. Erroneous incorporation of C residues at the editing site was observed in very few cases. Hence, it can be concluded that Air1p only has a minor contribution to the editing event. In the case of air2D, the editing efficiency was dramatically reduced to 12.6% while 85.2% of the clones lacked the editing position (misincorporation of C residues was only observed very rarely). Accordingly, in air1D/air2D, editing was almost completely abolished (2.5% editing vs. 97.4% truncated tRNA), indicating that Air2p is an important component of the editing activity whereas Air1p (which interacts with Trf5p in the TRAMP5 complex; Houseley and Tollervey 2006) is involved to a much lesser extent (fig. 3B). In Vitro Analysis of TRAMP4-Catalyzed tRNA Editing As the in vivo data indicate Trf4p and Air2p as the responsible tRNA editing activity, this reaction is obviously catalyzed by the TRAMP4 complex. To verify this conclusion in vitro, a tandem affinity purification of this complex was conducted. An N-terminally TAP-tagged version of Trf4p was recombinantly expressed in yeast strain TRF4 TAP, and the resulting TRAMP4 complex was purified (supplementary fig. S1, Supplementary Material online). To determine the individual components of the preparation, a mass spectrometric analysis of the total isolated TAP-tagpurified fraction as shown in supplementary fig. S1 (Supplementary Material online) was performed. Neither the CCA-adding enzyme nor the canonical poly(A) polymerase or Trf5p were present while the standard components of the TRAMP4 complex Trf4p as well as Air1p, Air2p, and Mtr4p were identified as prominent components with high emPAI values (supplementary table S1, Supplementary Material online) (Ishihama et al. 2005). Therefore, a contamination of the extract with other RNA-specific and template-independent nucleotidyltransferases could be ruled out, and any tRNA editing reactions, catalyzed by the preparation, can be traced back to the purified TRAMP4 complex. The preparation was tested for editing activity using an internally labeled in vitro transcribed E. coli suppressor tRNATyr(-1) (fig. 4A). Interestingly, the TRAMP4 complex added only 12 A residues to the tRNA 3#-end but did not show any polyadenylation reaction as it was observed for the initiator tRNAMet of S. cerevisiae (fig. 4B) that is commonly used as a standard tRNA substrate for TRAMP4 in vitro (Kadaba et al. 2004; Vanacova et al. 2005; Jia et al. 2011). As missing modifications lead to a destabilized tertiary structure of this in vitro transcribed initiator tRNA, it was speculated that TRAMP4 has a sensing function to identify misfolded tRNA transcripts for polyadenylation and degradation (Kadaba et al. 2004; Vanacova et al. 2005). Correctly folded tRNAs, however, are not polyadenylated but seem to be accepted at least for the incorporation of a limited number of 12 A residues. The E. coli suppressor tRNATyr represents such a transcript, and it is frequently used as an in vitro substrate for various tRNA processing and aminoacylation reactions that rely on correctly structured tRNAs (Gaur et al. 1996; Fechter et al. 2000; Soderbom et al. 2005). Obviously, the TRAMP4 complex discriminates between the misfolded tRNAiMet, where it adds a poly(A) tail, and the editing substrate, where only individual A residues are incorporated. This is in perfect agreement with the recent finding that the TRAMP4 component Mtr4p, originally identified as an RNA helicase (Wang et al. 2008), has an additional function in FIG. 4. In vitro editing reaction catalyzed by a purified TRAMP4 complex. A radioactively labeled in vitro transcript of sup-tRNATyr (-1) was incubated with TRAMP4 for the indicated time points. Although only two to three A residues were added to the 3#truncated editing substrate (A), the aberrantly folded transcript of yeast tRNAiMet was readily polyadenylated (B). Obviously, TRAMP4 discriminates between a misfolded tRNA that is subjected to degradation and a 3#-truncated tRNA that can be restored by the incorporation of a limited number of A residues. modulating the polyadenylation activity of Trf4p, regulating the length of the added poly(A) tails in response to the nature of the bound RNA (Jia et al. 2011). Nevertheless, it is surprising that the blunt-ended acceptor stem of sup-tRNATyr(-1) is readily accepted for A addition, as Jia et al. (2011) observed that a reconstituted TRAMP4 complex had only a very weak activity on a 16 bp blunt-end duplex RNA oligonucleotide. It is conceivable that the short duplex RNA oligonucleotide and the tRNA substrate are recognized in different ways by the complex. Alternatively, it is also possible that in vitro reconstituted TRAMP4 differs in its activity and substrate specificity from the in vivo purified TAP-tagged complex used in our analysis. Evolution of tRNA Editing Activities It is frequently discussed that a progenitor activity for tRNA editing was a rather promiscuous nucleotide adding enzyme with a broad range of RNA substrates (Covello and Gray 1993; Simpson et al. 2000; Schuster et al. 2005; Khersonsky and Tawfik 2010). Once a mutation in a tRNA gene led to the occurrence of a truncated transcript, the promiscuous nucleotidyltransferase accepted this RNA as a new substrate and restored the missing part(s), leading to a 3#-terminal editing event. This activity would overcome the otherwise deleterious effect of such a nonfunctional tRNA gene, leading to a genetic fixation of the nucleotide incorporation as an editing event. Here, TRAMP4 meets this as well as other criteria for such an exaptation as an emerging editing enzyme, and it is likely that a similaryet unidentifiedenzymatic activity was recruited for the mitochondrial tRNA 3#-editing reactions in higher eukaryotes (fig. 5). The TRAMP4 complex is involved FIG. 5. TRAMP4 is an ideal candidate for the recruitment as a tRNAediting enzyme. Due to its function in RNA quality control (left), TRAMP4 has the required promiscuity for accepting a truncated tRNA transcript for nucleotide addition (right). Based on its distributive mode of action, TRAMP4 releases the edited tRNA for CCA addition and further maturation. in RNA surveillance and degradation, a function that requires a broad substrate spectrum in order to inspect the integrity of the cells RNA transcripts. This promiscuity might have already led to an additional function of the Trf4p polyadenylating activity in the 3#-end maturation of small nuclear RNAs U3 and U4, as it was shown that 3#-extended versions of these RNAs accumulate in a trf4D strain (Egecioglu et al. 2006). Second, the editing event does not require a processive polymerase, but an enzyme that adds nucleotides in a distributive manner. Such an activity allows the incorporation of a limited number of nucleotides at the editing siteideally, only a single residue is added, before the enzyme dissociates from the substrate. In the case of 3#-end editing, the CCA-adding enzyme takes over the tRNA and completes the transcript by synthesizing the CCA terminus. The obtained in vivo data indicate that such a substrate handover is actually taking place, as most of the tRNAs end with one single A residue at the editing position and a partial or complete CCA terminus (figs. 2 and 3). Again, TRAMP4 fulfills this prerequisite, as it has a distributive mode of polymerization, taking several rounds of RNA binding and dissociation events for the addition of individual nucleotides (Jia et al. 2011). For tRNAs, numerous editing events at 5#- as well as 3#ends are described (Lonergan and Gray 1993a; Yokobori and Paabo 1995b; Tomita et al. 1996; Laforest et al. 1997; Price and Gray 1999; Lavrov et al. 2000; Leigh and Lang 2004; Gott et al. 2010). The 5#-editing occurs in a 3#-5# direction opposite to the conventional RNA polymerization, similar to the reaction catalyzed by histidine tRNA guanylyltransferase (Price and Gray 1999; Bullerwell and Gray 2005), and there is convincing evidence that tRNAHis guanylyltransferaselike proteins (TLPs) found in all kingdoms of life are catalyzing these editing events (Jackman and Phizicky 2006; Abad et al. 2010; Rao et al. 2011). Indeed, for Dictyostelium discoideum, it was demonstrated that such TLP enzymes are responsible for tRNA 5#editing in mitochondria, representing the first identified activity for this type of editing reaction (Abad et al. 2011). In the case of tRNA 3#-editing, we show that TRAMP4 is the first example of a promiscuous and distributive nucleotidyltransferase that meets the criteria for an evolutionary progenitor of such an activity, and the presented data are a strong experimental support for the hypothesis that promiscuous enzyme activities serve as a starting point for the evolution of new enzymatic functions (Khersonsky and Tawfik 2010). Supplementary Material Supplementary figure S1 and table S1 are available at Molecular Biology and Evolution online (http:// www.mbe.oxfordjournals.org/). Acknowledgments We thank D. Tollervey (Edinburgh, Scotland) for providing strain air1D/air2D, C. Rammelt (Halle, Germany) for strains trf4D, trf5D, W. Keller for rabbit anti-Trf4p polyclonal antibody, and especially K. Breunig (Halle, Germany) for valuable discussions. Special thanks to S. Kalkhof and M. von Bergen (Leipzig, Germany) for mass spectrometric analysis and B. Coordes and K. Straer (Munich, Germany) for help with the TAP-tag purification of TRAMP4. We thank C. Polte and S. Bonin for excellent technical support. This work was funded by the Deutsche Forschungsgemeinschaft (Mo 634/5).


This is a preview of a remote PDF: https://mbe.oxfordjournals.org/content/29/5/1451.full.pdf

Helena Dickinson, Sandy Tretbar, Heike Betat, Mario Mörl. The TRAMP Complex Shows tRNA Editing Activity in S. cerevisiae, Molecular Biology and Evolution, 2012, 1451-1459, DOI: 10.1093/molbev/msr312