Skip to main content
NCBI home page
RNA logoLink to RNA
. 2011 Oct;17(10):1846–1857. doi: 10.1261/rna.2628611

Plasticity and diversity of tRNA anticodon determinants of substrate recognition by eukaryotic A37 isopentenyltransferases

Tek N Lamichhane 1, Nathan H Blewett 1,3, Richard J Maraia 1,2,4
PMCID: PMC3185917  PMID: 21873461

Transfer RNAs are subject to a wide variety of modifications. Once such modification is N6-(isopentenyl)adenosine. This paper examines the substrate specificity of modifying enzymes from budding and fission yeast. Complex patterns of substrate determinants are uncovered. These determinants differ between the budding and fission yeast in enzymes. This study demonstrates previously unappreciated molecular plasticity and biological diversity of the tRNA-isopentenyltransferase system in eukaryotes.

Keywords: codon; anticodon loop, tRNA modification, tRNATrp, wobble base, i6A, (isopentenyl)adenosine

Abstract

The N6-(isopentenyl)adenosine (i6A) modification of some tRNAs at position A37 is found in all kingdoms and facilitates codon-specific mRNA decoding, but occurs in different subsets of tRNAs in different species. Here we examine yeasts' tRNA isopentenyltransferases (i.e., dimethylallyltransferase, DMATase, members of the Δ2-isopentenylpyrophosphate transferase, IPPT superfamily) encoded by tit1+ in Schizosaccharomyces pombe and MOD5 in Saccharomyces cerevisiae, whose homologs are Escherichia coli miaA, the human tumor suppressor TRIT1, and the Caenorhabditis elegans life-span gene product GRO-1. A major determinant of miaA activity is known to be the single-stranded tRNA sequence, A36A37A38, in a stem–loop. tRNATrpCCA from either yeast is a Tit1p substrate, but neither is a Mod5p substrate despite the presence of A36A37A38. We show that Tit1p accommodates a broader range of substrates than Mod5p. tRNATrpCCA is distinct from Mod5p substrates, which we sort into two classes based on the presence of G at position 34 and other elements. A single substitution of C34 to G converts tRNATrpCCA to a Mod5p substrate in vitro and in vivo, consistent with amino acid contacts to G34 in existing Mod5p-tRNACysGCA crystal structures. Mutation of Mod5p in its G34 recognition loop region debilitates it differentially for its G34 (class I) substrates. Multiple alignments reveal that the G34 recognition loop sequence of Mod5p differs significantly from Tit1p, which more resembles human TRIT1 and other DMATases. We show that TRIT1 can also modify tRNATrpCCA consistent with broad recognition similar to Tit1p. This study illustrates previously unappreciated molecular plasticity and biological diversity of the tRNA-isopentenyltransferase system of eukaryotes.

INTRODUCTION

The greatest diversity of tRNA modifications occurs on nucleotides 34 and 37 in the anticodon loop (ACL), which optimize codon:anticodon fit in the ribosome and promote translational fidelity (Gefter 1969; Vacher et al. 1984; Ericson and Bjork 1991; Jenner et al. 2010). Here we examine yeasts' tRNA isopentenyltransferases (i.e., dimethylallyltransferase, DMATase, members of the Δ2-isopentenylpyrophosphate transferase, IPPT superfamily) encoded by tit1+ in Schizosaccharomyces pombe and MOD5 in Saccharomyces cerevisiae, whose functional homologs are Escherichia coli miaA, human tumor suppressor TRIT1, and the Caenorhabditis elegans life-span gene product GRO-1, which convert A37 to N6-(isopentenyl)adenosine (i6A37) in their target tRNAs (Dihanich et al. 1987; Soderberg and Poulter 2000; Lemieux et al. 2001; Spinola et al. 2005). A major determinant of miaA activity is a single-stranded sequence A36A37A38 in a stem–loop (Motorin et al. 1997; Soderberg and Poulter 2000).

In E. coli, miaA creates i6A37 that may be further modified, whereas eukaryotic tRNAs contain i6A37 without further modification. The i6A37 modification occurs on different subsets of tRNAs in different model organisms, largely consistent with the presence or absence of A36A37A38 in the tRNA. In bacteria, i6A37 has been found on all tRNAs for codons with U in the first position—Cys, Leu, Phe, Ser, Trp, and Tyr (Bjork 1995; Persson et al. 1994), whereas in eukaryotes i6A37 has been found only on a subset of those tRNAs, consistent with their lack of A36A37A38 in tRNAs for Leu, Phe, and sometimes Cys. Specifically, yeasts' tRNAPhe and tRNALeu contain G37 and are therefore not modified with i6A37. tRNACysGCA carries i6A37 in S. cerevisiae (Holness and Atfield 1976) but not S. pombe, whose tRNACysGCA has G37 (Table 1). However, in apparent discord is S. cerevisiae tRNATrpCCA, which contains A36A37A38 but not i6A37 (Keith et al. 1971; Juhling et al. 2009; this study). Thus, one cannot confidently predict which tRNAs will be substrates for i6A37 modification based only on sequence, at least in yeast and maybe other eukaryotes. Better understanding of the molecular determinants of substrate specificity for DMATases may inform us about the evolution of this ancient activity. In addition, knowing which tRNAs are modified will be important toward deciphering codon-specific effects during translation in different model organisms. Yet, the basis for tRNATrpCCA discrimination in yeast is unknown, and the specificity of any eukaryotic DMATase has not been reported.

TABLE 1.

Differences in i6A37 among tRNAs in S. cerevisiae and S. pombe

graphic file with name 1846tbl1.jpg

Open in a new tab

Synthetic RNA minihelices used to examine MiaA in vitro indicate that sequences required for activity reflect a consensus derived from natural tRNA substrates (Motorin et al. 1997; Soderberg and Poulter 2000). In addition to A36A37A38, miaA requires an anticodon stem of at least 5 bp (Motorin et al. 1997; Soderberg and Poulter 2000). MiaA is 316 amino acids in length, very similar to other bacterial DMATases, whereas eukaryotic DMATases contain an additional C-terminal domain of ∼100 amino acids that includes a Zn-finger motif. Crystal structures of Mod5p bound to tRNACysGCA reveal interaction of the Zn-finger motif with the top of the anticodon stem, somewhat distant from the catalytic site and adjacent region that contact other nucleotides in the ACL (Zhou and Huang 2008). In addition to A37, which is flipped out into a catalytic pocket similar to MiaA-tRNAPheGAA (Chimnaronk et al. 2009; Seif and Hallberg 2009), the structures also reveal Mod5p interactions with U33 and C35, as well as contacts to the G34 base (Zhou and Huang 2008).

Our interest in specificity arose as we identified tRNATrpCCA as a substrate of Tit1p in S. pombe and confirmed that tRNATrpCCA is apparently unmodified at N6 of A37 in S. cerevisiae despite the presence of A36A37A38. Pursuing the basis of this, our data argued that differential tRNA trafficking in the two yeasts was not responsible. We then performed modification assays in vitro and in vivo for Tit1p and Mod5p and found that discrimination occurs at the level of substrate recognition. Moreover, discrimination by Mod5p can be reversed by changing a single base of tRNATrpCCA, C34 to G. These observations are consistent with extensive differences in the amino acid sequences of Mod5p and Tit1p in the region of Mod5p that contacts the G34 base of tRNA seen in existing crystal structures. Mutation of Mod5p in this region led to selective debilitation of activity toward its G34-containing natural substrates. Multiple sequence alignments reveal significant diversity in this region of DMATases. Accordingly, the human TRIT1 protein also has broad substrate recognition, more similar to the S. pombe enzyme. Thus, insight into the molecular plasticity and biological diversity of the isopentenyltransferase-tRNA substrate system, which affects codon-specific translation, is revealed.

RESULTS

Mutants in tit1+, formerly known as sin1+, derived from a genetic screen after NaNO2-induced mutagenesis (Thuriaux et al. 1976), exhibit loss of codon-specific tRNA-mediated suppression (TMS) by sup3-e (and sup9-e), which encodes a serine-inserting tRNASerUCA (Kohli et al. 1979; Rafalski et al. 1979; Egel et al. 1980; Janner et al. 1980). The multiple previously described sin1 mutants were mapped genetically but were not sequenced, and the precise mutations remain unknown (Thuriaux et al. 1976; Kohli et al. 1989). Strain yYH1 contains a weak suppressor tRNASerUCA allele (Huang et al. 2005, 2006) derived from sup3-e that suppresses a UGA nonsense codon in ade6-704 and the accumulation of red pigment. Due to a partial suppression phenotype, yYH1 can report either gain or loss of TMS (Huang et al. 2005, 2006). Our approach was to target tit1+ for deletion from yYH1. The resulting tit-Δ strain, designated yNB5, exhibited loss of TMS as reflected by more red pigment than yYH1 (Fig. 1B). Ectopic tit1+ rescued this phenotype (Fig. 1A,B, yNB5 + tit1+). Based on mutagenesis of threonine-19 of MiaA (Soderberg and Poulter 2001), which rendered it catalytically debilitated, and multiple sequence alignment, we mutated Tit1p threonine-12 to alanine. Indeed, tit1-T12A was inactive for TMS (Fig. 1A,B). Mid-Western blotting using anti-i6A antibody detects the physical presence of this modification on tRNA (Benko et al. 2000). This confirmed absence of the i6A modification in tit1-Δ and tit1-T12A tRNAs, while ectopic tit1+ restored the modification (Fig. 1C,D). The presence of at least two i6A-containing bands that migrated with the tRNAs suggests that substrates other than just the suppressor-tRNASerUCA, which is a minor tRNASer derived from the tRNASerUGA gene (Rafalski et al. 1979; Janner et al. 1980), are modified by Tit1p. Western blotting showed that Tit1p-T12A was expressed at a level comparable to Tit1p (Fig. 1E), providing evidence that tit1-T12A produces an inactive enzyme.

FIGURE 1.

FIGURE 1.

Open in a new tab

Deletion of tit1+ causes loss of tRNA i6A and tRNA-mediated suppression. Various strains growing in EMM with nonlimiting (A) or limiting (B) adenine, the latter reflecting tRNA-mediated suppression (TMS) activity. yYH1 (WT, tit1+), the parent strain of yNB5, was transformed with empty vector (row 1); this strain exhibits partial TMS activity (Huang et al. 2005). yNB5 (tit1-Δ) was transformed with empty vector, tit1+, tit1-T12A, or MOD5, as indicated for rows 2–5. (C,D) Analysis of S. pombe strains in A and B for i6A in tRNA by mid-Western blotting (Benko et al. 2000). (C) Ethidium bromide–stained total RNA in TBE-urea gel. (D) Blot of C after incubation with anti-i6A antibody and secondary processing for chemiluminescence (only the tRNA region is shown). (E) The tit1+ (lane 1), tit1-T12A (lane 2), and empty vector (lane 3) transformed strains from above were analyzed by immunoblotting using anti-HA antibody.

Substrate specificity of Tit1p

To set up an in vitro modification assay similar to that used for MiaA (Soderberg and Poulter 2000, 2001), we used purified recombinant Tit1p and 17-nt or 19-nt synthetic oligo-RNAs representing ASLs as minihelix analogs of tRNAs (Soderberg and Poulter 2000, 2001), 14C-DMAPP (Dihanich et al. 1987) (dimethylallyl pyrophosphate), and 250 nM Tit1p, which was determined to robustly modify one of our best substrate minihelices, SerAGA-19. We examined stem lengths for TyrGUA and SerAGA ASLs as depicted in Figure 2A. Of these, only the ASLs with a predicted stem of 6 bp were efficiently modified by Tit1p (Fig. 2B). Although natural tRNA substrates have only a 5-bp AC stem, these data suggest that the eukaryotic enzyme requires more contact with the upper part of the AC stem than does MiaA (Motorin et al. 1997; Soderberg and Poulter 2000).

FIGURE 2.

FIGURE 2.

Open in a new tab

Stem length of anticodon stem–loop (ASL) substrates of Tit1p. (A) Predicted structures of ASL oligos representing tRNASer and tRNATyr of various lengths; A37 is circled. Nucleotide positions 28 and 33 are indicated for SerAGA-17; numbering is the same for all others. For some, an extra closing G-C base pair was added (lowercase). For TyrGUA-19, an A-U base pair was added for comparison to SerAGA-19. (B) In vitro modification of the ASLs in A using recombinant Tit1p and 14C-DMAPP. (C) SerAGA-19 (lane 1) and its derivative ASLs with the substitutions indicated above lanes 2–7 were assayed as in B.

To examine specificity for A36A37A38, we assayed ASLs with substitutions as in Ser-G36, Ser-G38, Ser-G37, and Ser-G36,G38 (Fig. 2C). Position 36 and 37 substitutions greatly diminished activity (Fig. 2C, lanes 2,3). However, Ser-G38 was modified by Tit1p (Fig. 2C, lane 4), unexpected because MiaA exhibits much reduced activity on similar substrates (Soderberg and Poulter 2000). Ser-C38 and Ser-U38 were not modified (Fig. 2C, lanes 6,7).

Upon finding that Tit1p can modify substrates with G38 ASL, we surveyed the S. pombe tRNA genome database (Lowe 2011). Significantly, no tRNAs that contain A36A37G38 were found among the 186 S. pombe tRNA sequences. Thus, although Tit1p may be able to modify substrates containing A36A37G38, no such natural tRNA exists. We were unable to determine if Tit1p could modify a tRNASerUCA-G38 in vivo because the mutant tRNASerUCA-G38 failed to accumulate in S. pombe presumably due to rapid decay or counterselection (data not shown).

The C-terminal Zn finger of Tit1p is important for activity in vitro and in vivo

To try to understand the apparent difference between MiaA and Tit1p in ASL stem length, we focused on the eukaryote-specific C-terminal Zn motif. We deleted 55 amino acids from the C terminus of Tit1p to make Tit1p(1-379) and also mutated both cysteines of the C2H2 Zn finger to alanine (Tit1p-ZnAA). We examined these purified Tit1p proteins by in vitro modification of ASL minihelices, in vitro modification of hypomodified tRNAs isolated from tit1-Δ cells, and in vivo by TMS. The mutated proteins were inactive on all of the ASL minihelices tested (data not shown).

The purified proteins were examined for activity on tRNAs isolated from tit-Δ yNB5 cells (Fig. 3A,B). Wild-type Tit1p produced two bands (Fig. 3A, lane 1) that reflect tRNAsSer and tRNATyr based on their sizes of 82 and 74 nt, respectively, and other characteristics described below. Tit1p(1-379) and Tit1-ZnAA were inactive (Fig. 3A, lanes 2,4). Tit1p suffered no loss of activity when mixed with Tit1p(1-379) or Tit1-ZnAA, indicating that the mutated proteins did not contain an inhibitor (Fig. 3A, lanes 3,5).

FIGURE 3.

FIGURE 3.

Open in a new tab

The Tit1p extended C terminus including Zn finger is critical for activity in vitro and in vivo. (A) Equal amounts of purified recombinant proteins were tested for in vitro activity using 14C-DMAPP and total RNA from yNB5 (tit1-Δ) as substrate, with protein(s) as indicated above the lanes. (B) The gel in A stained with ethidium bromide. (C) tRNA-mediated suppression assay of the strains transformed with the expression plasmids indicated. (D) Proteins from the yNB5 strains in C transformed with HA-tit1+ (lane 1), HA-tit1(1-379) (lane 2), and HA-tit1-ZnAA (lane 3) were examined by immunoblotting using anti-HA antibody.

For in vivo analysis, the proteins were expressed with HA tags. The mutated proteins were inactive for TMS (Fig. 3C) even though they accumulated in S. pombe as did Tit1p-HA (Fig. 3D). The cumulative results with these mutants revealed very good agreement of data obtained with ASLs and cellular tRNAs in vitro, and TMS in vivo.

tRNATrpCCA is a substrate of Tit1p in vitro

Although tRNATrpCCA's from S. pombe and S. cerevisiae contain A36A37A38, S. cerevisiae tRNATrpCCA was found to contain unmodified A37 (Keith et al. 1971; Juhling et al. 2009). Two 19-nt ASLs representing tRNATrpCCA from S. cerevisiae and S. pombe (Fig. 4A), which migrated with different mobilities, were modified by Tit1p (Fig. 4B, lanes 2,3). These ASLs were resynthesized and again migrated with the relative mobilities seen in Figure 4B even before modification (data not shown).

FIGURE 4.

FIGURE 4.

Open in a new tab

Identification of tRNATrpCCA as a substrate of Tit1p. (A) Nineteen-nucleotide ASL oligos representing tRNATrpCCA from S. pombe and S. cerevisiae used in B for in vitro modification by Tit1p; SerAGA-19 (lane 1) from S. pombe is a control; Trp-19 from S. pombe (lane 2) and Trp-19 from S. cerevisiae (lane 3). The S. pombe and S. cerevisiae ASLs Trp-19 run differently on TBE-urea gel (see text). (C) In vitro modification assay after elimination of specific tRNAs by RNase H, to identify S. pombe substrates of Tit1p. (Lanes 1,2) In vitro modification by Tit1p of cellular RNA isolated from yYH1 (tit1+, lane 1) and yNB5 (tit1-Δ, lane 2). (Lanes 3–10) RNA from yNB5 was pre-incubated with oligo-DNA(s) complementary to the anticodon loop of the tRNA(s) targeted for elimination indicated above the lanes, followed by RNase H. The RNA was then purified prior to the Tit1p modification assay. (D–H) Positive hybridization in the absence of i6A37 (PHA6) assay to identify in vivo substrates of Tit1p in S. pombe. (D) The EtBr-stained gel from which the blot hybridizations below it were derived. Lanes 1–4 and 5–8 contain duplicate samples of 5 μg and 10 μg of RNA from yYH1 (tit1+) and yNB5 (tit1-Δ) as indicated. The same blot was probed, stripped, and reprobed sequentially with the probes indicated to the right. (E,G) The 32P-labeled probes were complementary to the anticodon loop (ACL) region of the tRNAs indicated to the right. (F,H) The 32P-labeled probes were complementary to the TΨC region of the tRNAs indicated to the right. The relative positions of the ACL and TΨC probes are indicated below the blots.

We next performed the in vitro 14C modification assay on tRNAs isolated from tit1-Δ cells before and after antisense oligo-DNA annealing followed by RNase H treatment to identify, by elimination, individual bands on the gel as specific tRNA isoacceptor substrates of Tit1p (Fig. 4C). tRNA from yYH1 (tit1+) was not a substrate (Fig. 4C, lane 1), as expected if the tRNAs already contain i6A37 or another modification to the N6 of A37, whereas tRNA from yNB5 (tit1-Δ) was modified (Fig. 4C, lane 2). The electrophoresis used for Figure 4C resolved the yNB5 tRNA into three bands of Tit1p products (Fig. 4C, lane 2). In Figure 4C, lanes 3–10, antisense oligo-DNAs spanning the anticodon loop (ACL) of the specific tRNAs targeted for degradation were incubated with the cellular tRNA and subsequently treated with RNase H, then purified prior to addition of 14C-DMAPP and Tit1p. The S. pombe genome harbors seven genes for tRNASerAGA, three genes for tRNASerUGA, and one gene for tRNASerCGA, all of which contained A36A37A38 (Lowe 2011). Targeting tRNASerAGA with antisense oligo and RNase H reduced the amount of upper band but not the two lower bands (Fig. 4C, lane 3). Targeting tRNASerUGA or tRNASerCGA alone had negligible effect on the amount of upper band (Fig. 4C, lanes 4,5). When tRNASerAGA and tRNASerUGA or tRNASerCGA were targeted simultaneously, the upper band was most diminished (Fig. 4C, lanes 6,7). Thus, as expected based on size, the three potential tRNAsSer substrates of Tit1p migrate as the single upper band. The four genes encoding S. pombe tRNATyr are of identical sequence (Lowe 2011). Targeting tRNATyrGUA caused disappearance of only the lowest band (Fig. 4C, lane 8). The three genes encoding S. pombe tRNATrpCCA are of identical sequence (Lowe 2011). Targeting tRNATrpCCA led to disappearance of the middle band only (Fig. 4C, lane 9). Targeting both tRNATyrGUA and tRNATrpCCA left only the upper band (Fig. 4C, lane 10). These data suggest that tRNAsSer, tRNATyrGUA, and tRNATrpCCA are substrates for Tit1p in vitro and in S. pombe, and that they exhibit the relative mobilities indicated to the right of Figure 4C.

tRNATrpCCA is a substrate of Tit1p in vivo

Indirect evidence, based on altered chromatographic mobility, had suggested that S. pombe tRNATrpCCA was a substrate of tit1 (then sin1), although detection of i6A37, which was confirmed for tRNAsSer and tRNATyr, was not reported for tRNATrpCCA (Janner et al. 1980). To examine substrates of Tit1p in vivo, we developed a Northern blot assay that we refer to as positive hybridization in the absence of i6A37 (PHA6) to detect i6A37 hypomodified tRNA. The PHA6 assay works with the expectation that the bulky isopentenyl group on N6 of A37 would interfere with stringent hybridization with a 32P-oligo ACL probe that spanned the 37 position. Hybridization would be greater in the absence of i6A37 modification as in tit1-Δ cells and less so in yYH1 cells in which i6A37 was present. By this approach, for a Mod5p substrate tRNA we would expect significantly more 32P-oligo ACL probe signal in tit1-Δ than in tit1+ cells, after calibration to correct for loading using a 32P-oligo probe complementary to the TΨC loop of the same tRNA. This was readily observed using an ACL 32P-probe to tRNASerAGA on the blot (Fig. 4E) derived from the gel shown in Figure 4D. The membrane was stripped and hybridized with a 32P-DNA probe complementary to the TΨC loop of tRNASerAGA, which showed no difference between yYH1 and yNB5 (Fig. 4F). Having validated this assay for a known substrate, we next examined tRNATrpCCA. The ACL probe hybridized with tRNATrpCCA from yNB5 but not yYH1 (Fig. 4G). Control hybridization with a TΨC probe revealed comparable hybridization with tRNATrpCCA from both strains (Fig. 4H), suggesting that tRNATrpCCA is modified in yYH1 but not yNB5 that has no Tit1p, and that lack of i6A37 has no significant effect of the tRNA levels.

In an attempt to understand the apparent discrepancy in the substrate activity of tRNATrpCCA in S. cerevisiae and S. pombe, we first wanted to confirm (or not) that tRNATrpCCA is not modified in S. cerevisiae. The PHA6 assay did not reveal a significant difference in ACL hybridization to tRNATrpCCA in MOD5 (ABL8) and mod5-Δ cells, similar to the tRNATrpCCA TΨC probe (Fig. 5A–C), supporting the previous data (Keith et al. 1971; Juhling et al. 2009). We note that this approach cannot rule out the possibility that a small fraction of tRNATrpCCA may be modified by Mod5p in vivo.

FIGURE 5.

FIGURE 5.

Open in a new tab

S. cerevisiae tRNATrp is not a substrate of Mod5p in vitro or in vivo. (A–E) PHA6 assay of S. cerevisiae RNAs from MOD5 replete and mod5-Δ cells as described in Figure 4 but using probes complementary to the ACL and TΨC regions of the S. cerevisiae tRNAs schematically depicted in F and indicated to the right; the same blot was probed, stripped, and reprobed sequentially. (G) In vitro modification by purified recombinant Tit1p, of S. cerevisiae tRNAs from mod5-Δ cells after elimination of specific tRNAs by RNase H, to identify in vivo substrates of MOD5. In vitro modification using Tit1p of RNA from ABL8 (MOD5) cells after mock preincubation (lane 1) or preincubation with oligo-DNA antisense to S. cerevisiae tRNATrp, both followed by RNase H treatment and purification prior to the 14C-DMAPP modification assay (lane 2). (Lane 3) Tit1p-mediated modification of S. cerevisiae RNA isolated from MT8 (mod5-Δ) after mock preincubation. (Lanes 4–10) RNA from MT8 (mod5-Δ) was preincubated with an oligo-DNA(s) antisense to the anticodon loop of the tRNA(s) targeted for elimination indicated above the lanes followed by RNase H and purification, prior to the Tit1p 14C-DMAPP modification assay. The tRNAs assigned to the bands are summarized to the right. (H) Comparison of Tit1p and Mod5p for in vitro modification of tRNAs from S. cerevisiae. (Lanes 1,2) ABL8 (MOD5) tRNA; (lanes 3,4) MT8 (mod5-Δ) tRNA modification by Tit1p and Mod5p. The tRNAs assigned to the bands are summarized to the right. (I) Ethidium-stained gel of the assay in H.

As discussed below, tRNA introns reside between positions 37 and 38, and in such cases splicing is a prerequisite for i6A37 modification. We considered this potentially relevant because all of the tRNATrpCCA genes in S. cerevisiae contain introns, whereas none of the tRNATrpCCA genes in S. pombe do (Lowe 2011). To see if tRNAs with introns may be precluded from i6A37 modification, we examined S. cerevisiae tRNASerCGA, whose gene also contains an intron at the same position, using the PHA6 blot assay. The tRNASerCGA ACL probe showed efficient hybridization with tRNASerCGA from mod5-Δ but not from MOD5 cells (Fig. 5D,E), consistent with tRNASerCGA containing i6A37 (Etcheverry et al. 1979; Johansson and Bystrom 2005). Furthermore, S. cerevisiae tRNATyr contains i6A37 (Madison and Kung 1967) despite introns in all of its genes (Fig. 5G; Lowe 2011). Thus, it seemed unlikely that the difference in tRNATrpCCA modification in S. cerevisiae and S. pombe was due to differential presence/absence of introns in their genes.

We next subjected RNA from S. cerevisiae ABL8 (MOD5) and MT8 (mod5-Δ) cells to in vitro modification by Tit1p. A single band from ABL8 (MOD5) was produced (Fig. 5G, lane 1) that was absent after S. cerevisiae tRNATrpCCA was targeted by antisense DNA and RNase H (Fig. 5G, lane 2). This provides evidence that S. cerevisiae tRNATrpCCA is unmodified at the N6 of A37 in vivo since it is available for isopentenylation by Tit1p. When RNA from mod5-Δ cells was subjected to Tit1p, additional bands were observed as expected (Fig. 5G, lane 3). These additional bands were not observed in Figure 5G, lane 2 presumably because they were already modified in the MOD5 replete cells, ABL8. Examination by antisense DNA and RNase H treatment identified these as tRNATrp, tRNACys, tRNATyr, and tRNAsSer (Fig. 5G, lanes 4–8). The data indicate that the fastest band from mod5-Δ cells is a combination of tRNATrpCCA and tRNACysGCA, both at 72 nt in length. S. cerevisiae tRNASer and tRNATyr are 82 and 75 nt, respectively.

Purified Mod5p and Tit1p differ in specificity for tRNATrpCCA

We used purified recombinant Mod5p for in vitro modification in parallel with Tit1p (Fig. 5H,I). Tit1p but not Mod5p modified the lowest band corresponding to tRNATrpCCA in RNA from MOD5 replete cells (Fig. 5H, lanes 1,2). Mod5p produced three bands from mod5-Δ RNA (Fig. 5H, lane 4), although with preference for the upper two, as compared to Tit1p (Fig. 5H, lane 3), consistent with selective inactivity for tRNATrpCCA. The data in Figure 5G,H suggests that Mod5p does not recognize tRNATrpCCA as a substrate while it does recognize tRNACys, tRNATyr, and tRNAsSer.

Inactivity of tRNATrpCCA for Mod5p can be reversed by a C34G substitution

To gain insight into the determinants of recognition by Mod5p, we compared the ASL regions of its five natural substrates with the tRNATrpCCA ASL (Fig. 6A). We examined Mod5p activity on the ASL minihelices and their substituted derivatives. Mod5p modified Ser 19 (Fig. 6B, lane 1) but not the ASL derived from S. cerevisiae tRNATrp (Fig. 6B, lane 2, scTrp-19) or S. pombe tRNATrp (spTrp-19) (data not shown). Our approach was to make substitutions in scTrp-19 that would convert it to a Mod5p substrate, guided by Ser 19 as an end point. We first examined the stem. Changing the C29-G41 base pair in scTrp-19 to A-U to match all other Mod5p substrates did not produce activity (Fig. 6B, lane 3), nor did other base-pair changes in the stem (data not shown). Neither did a triple-base-pair substitution at 27-43, 28-42, and 29-41 in scTrp-19 that made it identical to the Ser 19 stem produce activity (Fig. 6B, lane 4). We then focused on the anticodon.

FIGURE 6.

FIGURE 6.

Open in a new tab

C34G substitution activates tRNATrpCCA as a Mod5p substrate. (A) Comparison of the ASL regions of the two groups of natural Mod5p substrates and the nonsubstrate tRNATrpCCA, all of which contain A36A37A38. (Upper row) The three tRNAsSer substrates that comprise the non-G34 group and contain G35; (lower row) the first two substrates represent the G34 group, composed of tRNATyrGUA and tRNACysGCA; (lower right) the nonsubstrate tRNATrpCCA. (B–D) In vitro modification of ASLs: (B) Mutations to the stem of ASLs representing tRNATrpCCA do not activate it as a Mod5p substrate. (Lane 1) Ser-AGA-19; (lane 2) scTrp-19; (lane 3) scTrp-19-mut2-C29A•G41U; (lane 4) scTrp-19-mut6: U27A•A43U, U28C•A42G, C29A•G41U; the stem is identical to the Ser-AGA-19 stem in lane 1. (C) C34G substitution activates ASL representing tRNATrpCCA as a Mod5p substrate. Mod5p modification of Ser-AGA-19 (lane 1) or the scTrp-19 ASLs with the anticodon sequences indicated above the lanes (upper panel). (Lower panel) The ASLs gel in upper panel after staining with EtBr. (D) Parallel in vitro modification of the same set of ASLs by Mod5p (lanes 1–5) and Tit1p (lanes 6–10). (Lanes 1,6) Ser-AGA-19; (lanes 2,7) scTrp-ICA; (lanes 3,8) scTrp-CAA; (lanes 4,9) scTrp-CUA; (lanes 5,10) scTrp-UCA. (E–I) C34G substitution activates tRNATrpCCA as a Mod5p substrate in vivo. MOD5 and mod5-Δ cells were transformed with empty vector (lanes 1–4) or vector expressing an S. pombe gene encoding tRNATrpCCA (lanes 5–8) or the point mutated tRNATrpGCA (lanes 9–12) as indicated above the lanes. RNA from the transformed cells was fractionated and blotted, and the membrane was sequentially hybridized, stripped, and rehybridized with probes indicated to the right of the panels according to the PHA6 Northern blot assay.

The natural substrates can be arranged into two groups with distinct characteristics. That is, the tRNAsSer (Fig. 6A, upper row) all have a G in position 35; an A, C, or U in position 34; and a long variable arm common to type 2 (or class II) tRNAs, whereas the other group members, tRNATyrGUA and tRNACysGCA, have pyrimidine in position 35, G at 34, and a short variable arm common to type 1 (class I) tRNAs. In contrast, the nonsubstrate tRNATrpCCA has pyrimidines in both the 34 and 35 positions and a short variable arm. Indeed, a single substitution of scTrp-19 ASL that replaced C34 with G34 activated it as a substrate for Mod5p (Fig. 6C, lane 4). Substituting C34 with A, or C35 with G, did not convert the scTrp-19 ASL to a substrate (Fig. 6C, lanes 3,5). Thus, the scTrp-19 ASL minihelix with a single substitution converting it to the GCA anticodon (normally found on tRNACysGCA) was active for modification by Mod5p. This seems significant since the crystal structures of Mod5p-tRNACysGCA (Zhou and Huang 2008) show three side chains of Mod5p (H86, K127, K181) in close contact with the N6 oxy group of G34, and one side chain (Y84) in close contact to N2 of G34 (Supplemental Fig. S2; see PDB files 3EPH, 3EPJ, 3EPK, and 3EPL).

We compared the ASL minihelices for Tit1p and Mod5p in parallel (Fig. 6D). Since the only Mod5p substrate whose gene encodes A in position 34, tRNASerAGA, contains inosine (I) at this position (Zachau et al. 1966; in Johansson and Bystrom 2005), we also examined ASLs with I34. The ASL with I34 was relatively inactive for both proteins (Fig. 6D, lanes 2,7). While the scTrp-19 CAA and CUA ASLs showed little if any substrate activity for Mod5p, these were more active as Tit1p substrates (Fig. 6D). scTrp-19 UCA showed significant substrate activity for Mod5p and relatively more for Tit1p although less than Ser 19 (Fig. 6D). Relative substrate activities of the ASLs are summarized in Table 2.

TABLE 2.

Relative activities (0–5) of Mod5p and Tit1p toward different ACL analogs of tRNATrp and tRNASer

graphic file with name 1846tbl2.jpg

Open in a new tab

Activation of tRNATrpCCA as a Mod5p substrate by C34G substitution in vivo

We next asked if a single C-to-G substitution at position 34 would be sufficient to convert tRNATrpCCA from an inactive to an active Mod5p substrate in vivo. For this we cloned the wild-type S. pombe tRNATrpCCA gene and a version containing a single C-to-G substitution at position 34, used them to transform S. cerevisiae ABL8 (MOD5) and MT8 (mod5-Δ) cells, and performed the PHA6 Northern blot assay (Fig. 6E–I). Six samples were examined, each with 1× and 2× concentrations of RNA, as indicated above the lanes of Figure 6E, the EtBr-stained gel from which the blot for Figure 6F–I was derived. Transformation with the empty vector showed no hybridization with any of the sptRNATrpCCA-specific probes (Fig. 6F–H, lanes 1–4) as expected, demonstrating specificity of hybridization. sptRNATrpCCA isolated from MOD5 and mod5-Δ showed comparable hybridization with the ACL probe although with slightly more signal in lane 8 than in lane 6 (Fig. 6F, lanes 5–8). In contrast, the sptRNATrpGCA bearing the C34-to-G34 substitution showed significantly more hybridization with the ACL probe in mod5-Δ than in MOD5 (Fig. 6F, lanes 9–12), providing evidence that tRNATrpGCA was modified by MOD5. The sptRNATrpCCA TΨC probe detected comparable amounts of sptRNATrpCCA and sptRNATrpGCA (Fig. 6H, lanes 5–12). As a further control, we also probed for endogenous sctRNASerCGA using a TΨC probe (Fig. 6I). Quantitation with calibration using the sptRNATrpCCA TΨC probe suggested that ∼10%–15% of the sptRNATrpCCA may be modified by Mod5p; in contrast, ∼85% of the sptRNATrpGCA was modified by Mod5p (data not shown). The data in Figure 6E–I provided evidence that a C34G substitution was sufficient to activate tRNATrpCCA as a Mod5p substrate in vivo. The data obtained with ACLs in vitro and tRNA substrates in vivo were largely in agreement.

Mutations to the anticodon binding loop of Mod5p alter tRNA substrate preference

As alluded above, the Mod5p side chains of H86, K127, K181, and Y84 are in close contact to the oxy group at position 6 and N2 amine groups of G34 in the Mod5p-tRNACysGCA structure (PDBs 3EPH, 3EPJ, 3EPK, and 3EPL) (Zhou and Huang 2008). Sequence alignment (Supplemental Fig. S1) reveals that these are not conserved by Tit1p and furthermore that K127 follows a loop in the Mod5p-tRNACysGCA structure (colored white in Supplemental Fig. S2) that helps shape the G34 binding pocket. Notably, this loop is foreshortened by three or more residues in S. cerevisiae and related yeasts Zygosaccharyomyces and Kluyveromyces, relative to S. pombe, Schizosaccharomyces japonicus, and other species (Supplemental Fig. S3), suggesting that it may contribute to differential substrate recognition. We mutated this region of Mod5p and examined activities on RNA isolated from MOD5 and mod5-Δ cells (Fig. 7A,B).

FIGURE 7.

FIGURE 7.

Open in a new tab

The anticodon binding loop of Mod5p alters tRNA substrate preference. (A) In vitro modification of S. cerevisiae mod5-Δ RNA by wild-type Mod5p (lane 1) and various mutated Mod5 proteins (lanes 2–5). Mod5p-loop+K127D (lane 2) contains an 8-amino-acid insertion after position 120 as well as the K127D mutation; the Mod5 proteins used in lanes 3–5 are single point mutations as described above the lanes. (B) Quantitative scanning of each lane of the gel shown in A using a Fuji PhosphorImager. For normalization, the total counts observed in tRNAsSer in Mod5p (wild-type, lane 1) were considered as 100%, and the others were compared accordingly. Note that the black tracing (Mod5p-K127D) for tRNAsSer is not visible because it was completely masked (overlaid) by the red tracing (Mod5p wild type). (C) In vitro modification of ASL substrates indicated above the lanes by Mod5p wild type and Mod5p-loop+K127D; (upper panel) EtBr staining of gel autoradiogram shown in lower panel. (D) In vitro modification of S. cerevisiae MOD5 (ABL8) and mod5-Δ (MT8) RNA by purified recombinant human TRIT1.

The S. cerevisiae tRNAsSer type 2 substrates of Mod5p migrate as one band with the slowest gel mobility due to their large variable arm, while the G34 group, comprised of tRNATyrGUA and tRNACysGCA, are shorter and each exhibits distinct faster mobilities (Fig. 5G,H). In S. cerevisiae there are 16 genes for the tRNAsSer substrates, eight genes for tRNATyrGUA, and four genes for tRNACysGCA. Thus, the relative intensities of the three bands observed after Mod5p-mediated in vitro modification should reflect the net effect of their differential abundances and efficiencies as substrates.

Wild-type Mod5p produced three bands comprising both groups of substrates with the pattern observed in Figure 7A, lane 1, in which the middle band is the most intense. In contrast, the mutated Mod5p-loop-K127D protein exhibited selective decrease in the lower two bands, tRNACysGCA and tRNATyrGUA, relative to the upper tRNAsSer group band (Fig. 7A, lane 2). The single point mutated proteins Mod5p-K127D and Mod5p-Y84S produced patterns nearly indistinguishable from wild-type Mod5p (Fig. 7A, lanes 1,3,4) and thereby serve as controls for the loop-mutated Mod5p (Fig. 7A, lane 2). Mod5p-K181H consistently showed slight reduction of the two lower bands relative to the upper band (see legend for Fig. 7B; Supplemental Fig. S4). Quantitative scanning of each of the lanes of the gel in Figure 7A is shown in Figure 7B. The same differential patterns were reproducibly observed for the loop-mutated Mod5p and -K181H proteins in multiple experiments including when different preparations of the mod5-Δ RNA were used (Supplemental Fig. S3; data not shown). The Mod5p-loop mutated protein also exhibited selective disproportionate decrease in activity toward a Tyr-19-GUA relative to Ser-AGA-19 and the wild-type Mod5p (Fig. 7C). These data support the idea that for the G34 group of Mod5p substrates (G34, Y35, and short variable arm; tRNACysGCA and tRNATyrGUA), G34 is a determinant of activity. Mutation of the G34 binding loop of Mod5p selectively inhibits activity toward these substrates more than the non-G34 substrates, tRNAsSer. Although we hoped that point mutations to Mod5p making it more similar to Tit1p would activate it for recognition of tRNATrpCCA, this was not observed.

Human TRIT1 appears similar to Tit1p in its anticodon binding loop length and exhibits analogous substrate specificity for tRNATrpCCA

According to multiple sequence alignment, the loop in Mod5p containing K127 appears to be foreshortened by internal deletions in the Mod5 proteins from S. cerevisiae and its close homologs in Kluyveromyces lactis and Zygosaccharyomyces rouxii, relative to S. pombe and S. japonicus and other species (Supplemental Fig. S3). Similar loop length and lack of conservation of other residues of Tit1p and TRIT1 relative to Mod5p prompted us to examine TRIT1 substrate specificity for tRNATrpCCA. Figure 7D shows that purified recombinant human TRIT1 can indeed modify tRNATrpCCA. We conclude that hTRIT1 and Tit1p exhibit an apparently similar substrate recognition that enables tRNATrpCCA modification, but differ from Mod5p, which is more restricted.

DISCUSSION

The major new finding reported here is that activity of the DMATase, Mod5p, of the model organism, S. cerevisiae, is quite sensitive to the base identity of wobble position 34. Specifically, Mod5p does not recognize tRNATrpCCA as a substrate despite the presence of A36A37A38. A single base substitution that changes tRNATrpCCA to tRNATrpGCA activates it for modification by Mod5p, consistent with crystal structure contacts between G34 of tRNACysGCA and Mod5p. This is surprising since examination of MiaA substrates would not have predicted this (Motorin et al. 1997; Soderberg and Poulter 2000). As noted in the Introduction, tRNAs that carry i6A in prokaryotes include the full set of those that decode all UNN codons, whereas in eukaryotes the subset of tRNAs with i6A is more restricted. Several eukaryotic tRNAs that decode UNN codons lack the A36A37A38 motif and would therefore not be expected to be DMATase substrates. However, the work here shows that i6A restriction extends to tRNATrpCCA in S. cerevisiae despite the presence of A36A37A38. This restriction can be overcome by substitution of tRNATrpCCA to tRNATrpGCA. Why eukaryotes restrict i6A37 to only a subset of tRNAs is unknown.

Molecular plasticity

It was helpful to sort the natural substrates of Mod5p into two groups, the type 2 group composed of tRNAsSer that contain A, U, or C at position 34, G at position 35, and a large variable arm; and the type 1 group that differs in that its members, tRNACysGCA and tRNATyrGUA, has G at position 34, pyrimidine at 35, and a small extra arm. The cumulative data suggest that G34 is an important determinant of Mod5p activity for the tRNACysGCA and tRNATyrGUA of the type 1 group. Consistent with this, tRNATrpCCA, which resembles the type 1 group with pyrimidine at position 35 and a small variable arm but lacks G at 34 and is inactive as substrate, can be converted to a Mod5p substrate by a single C-to-G substitution at position 34. However, G34 is clearly not required in the natural group 2 substrates, which all have G at position 35 and a large variable arm. The results suggest that Mod5p and other eukaryotic DMATases can recognize determinants in addition to the major one, A36A37A38, in substrate tRNAs that otherwise differ in anticodon sequence and other features. The data suggest molecular plasticity in the recognition of different types of tRNA substrates that differ, among other features, in the sequence at positions 34 and 35 of their anticodons.

Using synthetic ASLs, we found that Tit1p required a longer stem length than was reported for MiaA. The longer length requirement by Tit1p is consistent with the interpretation that the eukaryotic DMATase prefers more structure at the top of the stem because this is where the eukaryote-specific Zn-finger-containing C-terminal region interacts, as seen in the published crystal structure.

A second conclusion is that the homologous DMATases of two distant yeasts overlap in substrate specificity but do not recognize all of the same tRNAs. Specifically, S. cerevisiae Mod5p does not modify tRNATrpCCA, while S. pombe Tit1p clearly does. Examination of the existing cocrystal structure of Mod5p-tRNACysGCA provides insight into a structural basis for this since Tit1p has not conserved the amino acid side chains in Mod5p that make base-specific contacts to G34. Therefore, this study illustrates previously unappreciated biological diversity of the tRNA-isopentenyltransferase system of eukaryotes. These findings advance understanding of the functional and mechanistic basis of differential tRNA i6A37 modification.

Multiple sequence alignment and mutagenesis of the Mod5p G34 binding loop that contains K127 and comprises a region of significant difference from Tit1p and other DMATases suggested that human TRIT1 may exhibit activity for tRNATrpCCA, which we then demonstrated. The alignment would also suggest that the pathogenic Candida species may also exhibit broad specificity, although this remains to be determined.

We have shown that Tit1p catalyzes isopentenylation on tRNAs in vitro and in S. pombe. Tti1-Δ cells show loss of TMS that is complemented by ectopic Tit1p but not by a catalytic mutant. This reflects importance of i6A37 in the function of suppressor-tRNA in codon-specific TMS. This is consistent with anticodon loop modification effects on ASL structure and activity (Agris 2008) as the isopentenyl group appears to limit anticodon loop width and increase stacking with codon:anticodon bases affecting base-pairing and codon-specific translation (Bjork 1995; Agris et al. 2007). The Tit1p constructs with deletion of the extended C terminus or substitution of two cysteines of the Zn finger greatly reduced activity in vitro and for TMS in vivo. The Mod5p Zn finger interacts with the top of the anticodon stem (Zhou and Huang 2008). We suspect that the 15-nt and 17-nt ASLs were not modified by Tit1p because of inefficient interaction with the Zn finger due to their short stem, although other interpretations are possible.

Species-specific tRNA systems

Reliable genome-scale tRNA gene prediction indicates that tRNA gene copy number and corresponding codon usage vary between related species (Lowe 2011), suggesting that dynamics in decoding systems accompanies genome evolution and speciation. Results reported here, that S. cerevisiae and S. pombe differ in their subsets of i6A37-containing tRNAs (Table 1) and that this is in part due to differential recognition of tRNATrpCCA by their otherwise homologous DMATases, is consistent with the idea that alterations in the dynamics of tRNA-related decoding systems accompany genome evolution and/or speciation.

Codon-specific translation effects of anticodon loop modification

While i6A37 is carried by tRNAsSerAGA, UGA, and CGA, as well as tRNATyr in both S. cerevisiae and S. pombe, each has another i6A37-containing tRNA not found in the other yeast. tRNACysGCA carries i6A37 in S. cerevisiae but not in S. pombe, in which tRNACysGCA contains G37, as summarized in Table 1. Since i6A37 affects codon-specific translation, the mRNAs that are sensitive to it may differ in organisms with different subsets of i6A37 tRNAs, including perhaps mRNA subsets with differential synonymous codon use bias (Begley et al. 2007). Indeed, in S. cerevisiae the anticodon U34 base modification mediated by Trm9p appears to have been “keyed” to the translation of a specific subset of mRNAs with large bias in their use of the cognate codons (Begley et al. 2007). Somewhat similar perhaps are the effects of the N6-threonylcarbamoyl adenosine at position 37 (t6A37) of tRNAs decoding ANN codons (El Yacoubi et al. 2009, 2011; Srinivasan et al. 2011). Phenotypes attributable to lack of this modification are pleiotropic (El Yacoubi et al. 2011; Srinivasan et al. 2011), presumably related at least in part to mRNA-specific effects. Remarkably, deficiency of t6A37 promotes de-repression of specific mRNAs that bear upstream ORFs whose translation requires t6A37-modified tRNA (Daugeron et al. 2011). Determining whether or not defects in codon-specific translation of different subsets of i6A37-hypersensitive mRNAs can account for the phenotypic differences observed for S. pombe tit1-Δ mutants and S. cerevisiae mod5-Δ mutants will be a challenge for the future.

MATERIALS AND METHODS

RNA and DNA oligos

RNA and DNA oligos were purchased from IDT (Integrated DNA Technologies).

Strains

yYH1 is h− leu1-32:: [tRNAmSer7T-leu1+] ura4-D ade6-704. yNB5 is h− leu1-32:: [tRNAmSer7T-leu+] ura4-D tit1Δ-Kan+ ade6-704. yAS99 is h− ade6-704, ura4-D, leu1-32. ABL8 and MT8 were obtained from Anita Hopper (Gillman et al. 1991). The medium used for S. pombe was YES (3% glucose, 0.5% yeast extract + 225 mg/L adenine, histidine, leucine, uracil, and lysine). Edinburgh Minimal Media (EMM) lacking leucine and uracil was used with adenine at 10 mg/L for tRNA-mediated suppression (TMS) as previously described (Huang et al. 2005, 2006).

Plasmids

tit1+ including 600 bp of upstream DNA was cloned into pREP42X from which the NMT promoter was excised so that Tit1p was expressed from its own promoter with an HA tag on the C terminus. Transformed cells were grown in EMM lacking uracil to an O.D.600 of 0.5. Mutagenesis was carried out by QuickChange XL (Stratagene). All constructs were verified by sequencing.

S. pombe tRNATrp was cloned into pRS316. Transformed S. cerevisiae cells were grown in SC medium lacking uracil, to an O.D.600 of 0.6.

Northern blotting including PHA6 assay

S. pombe cells were grown to an O.D.600 of 0.5 in liquid media. RNA was extracted with hot acidic phenol:chloroform. Five micrograms of total RNA was electrophoresed in Novex 10% TBE-urea polyacrylamide gels and transferred to GeneScreen Nylon membranes using Invitrogen iBlot transfer apparatus, cross-linked by UV, and baked under vacuum. Blots were incubated overnight at highly stringent hybridization temperatures with 32P-labeled oligo-DNA complementary to the RNA species indicated.

Mid-Western blotting

Total RNA was electrophoresed, transferred to nitrocellulose, and processed as above. The membrane was blocked for 1 h at room temperature in 5% skim milk/1× PBS/0.05% Tween 20. Anti-i6A antibody (provided by Anita Hopper) was added at 1/50 dilution. After washing three times for 5 min each in 1× PBS/0.05% Tween 20, a secondary anti-rabbit HRP antibody was incubated for 1 h in 1% skim milk/1× PBS/0.05% Tween 20 at 1/500 dilution and then processed for chemiluminescence. The membrane was washed three times for 5 min each with 1× PBS/0.05% Tween 20 and exposed to a PhosphorImager screen.

Protein extraction and immunoblotting

Cells were grown to an O.D.600 of 0.5. Protein was extracted in 150 mM NaCl, 50 mM Tris-Cl (pH 7.5), 1 mM EDTA, 0.5% NP-40, and 0.1 mM PMSF (Phenyl Methylsulfonyl Fluoride) and added fresh, using glass beads and a BioSpec Products MiniBead Beater. Debris was pelleted, and the supernatant was adjusted to 10% glycerol and frozen at −80°C. Protein was analyzed on 4%–12% Bis-Tris PAGE gels (Novex) and stained with Simply Blue SafeStain (Invitrogen). To examine tagged proteins, the gel contents were transferred to PVDF membrane using iBlot apparatus (Invitrogen). Anti-HA antibody was used followed by secondary Ab, and the membrane was processed for chemiluminescence according to standard methods.

Protein purification

DNAs encoding Tit1p and its derivatives Mod5p and TRIT1 were cloned into pET15b (Invitrogen) and expressed in E. coli BL21(DE3)pLysS (Invitrogen), each with a His6 tag. Induced protein was purified by Talon metal affinity (Clontech) and the yield quantified by Bradford assay (Bio-Rad).

In vitro tRNA isopentenylation assay

Twenty micrograms of total RNA purified from yNB5 (tit1-Δ) or yYH1 (wild-type) or synthetic ASL at 5 μM final concentration was resuspended in 100 μL of reaction buffer: 50 mM Tris-Cl (pH 7.5), 5 mM MgCl2, 0.1 mM 2-mercaptoethanol, 1 nmol of 14C-DMAPP (10 μM, dimethylallyl pyrophosphate), and 100 U of Superasein (Ambion AM2694). Recombinant Tit1p at 250 nM was used for each reaction. After incubation, RNA was extracted with acidic phenol–chloroform. The organic phase was re-extracted with 250 μL of TES (10 mM Tris-Cl at pH 7.5, 10 mM EDTA, 0.5% SDS). The aqueous phase (RNA) was precipitated and resuspended in formamide RNA loading buffer, dried, and electrophoresed on a 12% TBE/urea gel. After drying the gel was exposed to a PhosphorImager screen. For RNase H experiments, a 25-mer antisense oligo DNA complementary to positions 23 to 47 of the tRNA was used.

SUPPLEMENTAL MATERIAL

Supplemental material is available for this article.

ACKNOWLEDGMENTS

We thank V. Cherkasova for insight, assistance, and comments; M. Bayfield, D. Hatfield, M. Ibba, and D. Söll for critical comments; Amanda Crawford for reagents and assistance; and Anita Hopper (Ohio State University, Columbus) for anti-i6A antibody and the MOD5 and mod5-Δ strains. This work was supported by the Intramural Research Program of the NICHD, NIH.

Footnotes

Article published online ahead of print. Article and publication date are at http://www.rnajournal.org/cgi/doi/10.1261/rna.2628611.

REFERENCES

  1. Agris PF 2008. Bringing order to translation: the contributions of transfer RNA anticodon-domain modifications. EMBO Rep 9: 629–635 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Agris PF, Vendeix FA, Graham WD 2007. tRNA's wobble decoding of the genome: 40 years of modification. J Mol Biol 366: 1–13 [DOI] [PubMed] [Google Scholar]
  3. Begley U, Dyavaiah M, Patil A, Rooney JP, DiRenzo D, Young CM, Conklin DS, Zitomer RS, Begley TJ 2007. Trm9-catalyzed tRNA modifications link translation to the DNA damage response. Mol Cell 28: 860–870 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Benko AL, Vaduva G, Martin NC, Hopper AK 2000. Competition between a sterol biosynthetic enzyme and tRNA modification in addition to changes in the protein synthesis machinery causes altered nonsense suppression. Proc Natl Acad Sci 97: 61–66 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bjork GR 1995. Biosynthesis and function of modified nucleosides. In tRNA: Structure, biosynthesis, and function (ed. D Soll, UL Rajbhandary), pp. 165–205 ASM Press, Washington, DC [Google Scholar]
  6. Chimnaronk S, Forouhar F, Sakai J, Yao M, Tron CM, Atta M, Fontecave M, Hunt JF, Tanaka I 2009. Snapshots of dynamics in synthesizing N6-isopentenyladenosine at the tRNA anticodon. Biochemistry 48: 5057–5065 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Daugeron MC, Lenstra TL, Frizzarin M, El Yacoubi B, Liu X, Baudin-Baillieu A, Lijnzaad P, Decourty L, Saveanu C, Jacquier A, et al. 2011. Gcn4 misregulation reveals a direct role for the evolutionary conserved EKC/KEOPS in the t6A modification of tRNAs. Nucleic Acids Res doi: 10.1093/nar/gkr178 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dihanich ME, Najarian D, Clark R, Gillman EC, Martin NC, Hopper AK 1987. Isolation and characterization of MOD5, a gene required for isopentenylation of cytoplasmic and mitochondrial tRNAs of Saccharomyces cerevisiae. Mol Cell Biol 7: 177–184 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Egel R, Kohli J, Thuriaux P, Wolf K 1980. Genetics of the fission yeast Schizosaccharomyces pombe. Annu Rev Genet 14: 77–108 [DOI] [PubMed] [Google Scholar]
  10. El Yacoubi B, Lyons B, Cruz Y, Reddy R, Nordin B, Agnelli F, Williamson JR, Schimmel P, Swairjo MA, de Crecy-Lagard V 2009. The universal YrdC/Sua5 family is required for the formation of threonylcarbamoyladenosine in tRNA. Nucleic Acids Res 37: 2894–2909 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. El Yacoubi B, Hatin I, Deutsch C, Kahveci T, Rousset JP, Iwata-Reuyl D, Murzin AG, de Crecy-Lagard V 2011. A role for the universal Kae1/Qri7/YgjD (COG0533) family in tRNA modification. EMBO J 30: 882–893 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Ericson JU, Bjork GR 1991. tRNA anticodons with the modified nucleoside 2-methylthio-N6-(4-hydroxyisopentenyl)adenosine distinguish between bases 3′ of the codon. J Mol Biol 218: 509–516 [DOI] [PubMed] [Google Scholar]
  13. Etcheverry T, Colby D, Guthrie C 1979. A precursor to a minor species of yeast tRNASer contains an intervening sequence. Cell 18: 11–26 [DOI] [PubMed] [Google Scholar]
  14. Gefter ML 1969. The in vitro synthesis of 2′-omethylguanosine and 2-methylthio 6N (γ,γ, dimethylallyl) adenosine in transfer RNA of Escherichia coli. Biochem Biophys Res Commun 36: 435–441 [DOI] [PubMed] [Google Scholar]
  15. Gillman EC, Slusher LB, Martin NC, Hopper AK 1991. MOD5 translation initiation sites determine N6-isopentenyladenosine modification of mitochondrial and cytoplasmic tRNA. Mol Cell Biol 11: 2382–2390 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Holness NJ, Atfield G 1976. The nucleotide sequence of cysteine transfer ribonucleic acid from baker's yeast. Products of complete digestion with pancreatic ribonuclease and ribonuclease T1. Biochem J 153: 437–446 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Huang Y, Intine RV, Mozlin A, Hasson S, Maraia RJ 2005. Mutations in the RNA polymerase III subunit Rpc11p that decrease RNA 3′ cleavage activity increase 3′-terminal oligo(U) length and La-dependent tRNA processing. Mol Cell Biol 25: 621–636 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Huang Y, Bayfield MA, Intine RV, Maraia RJ 2006. Separate RNA-binding surfaces on the multifunctional La protein mediate distinguishable activities in tRNA maturation. Nat Struct Mol Biol 13: 611–618 [DOI] [PubMed] [Google Scholar]
  19. Janner F, Vogeli G, Fluri R 1980. The antisuppressor strain sin1 of Schizosaccharomyces pombe lacks the modification isopentenyladenosine in transfer RNA. J Mol Biol 139: 207–219 [DOI] [PubMed] [Google Scholar]
  20. Jenner LB, Demeshkina N, Yusupova G, Yusupov M 2010. Structural aspects of messenger RNA reading frame maintenance by the ribosome. Nat Struct Mol Biol 17: 555–560 [DOI] [PubMed] [Google Scholar]
  21. Johansson MJ, Bystrom AS 2005. Transfer RNA modifications and modifying enzymes in Saccharomyces cerevisiae. In Topics in current genetics: Fine-tuning of RNA functions by modification and editing (ed. Grosjean H), pp. 88–120 Springer-Verlag, Berlin, Heidelberg [Google Scholar]
  22. Jühling F, Mörl M, Hartmann RK, Sprinzl M, Stadler PF, Pütz J 2009. tRNAdb 2009: compilation of tRNA sequences and tRNA genes. Nucleic Acids Res 37: D159–D162 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Keith G, Roy A, Ebel JP, Dirheimer G 1971. The nucleotide sequences of two tryptophane-tRNAs from brewer's yeast. FEBS Lett 17: 306–308 [DOI] [PubMed] [Google Scholar]
  24. Kohli J, Kwong T, Altruda F, Soll D, Wahl G 1979. Characterization of a UGA-suppressing serine tRNA from Schizosaccharomyces pombe with the help of a new in vitro assay system for eukaryotic suppressor tRNAs. J Biol Chem 254: 1546–1551 [PubMed] [Google Scholar]
  25. Kohli J, Munz P, Soll D 1989. Informational suppression, transfer RNA, and intergenic conversion. In Molecular biology of the fission yeast (ed. Nasim A et al. ), pp. 75–96 Academic Press, San Diego [Google Scholar]
  26. Lemieux J, Lakowski B, Webb A, Meng Y, Ubach A, Bussiere F, Barnes T, Hekimi S 2001. Regulation of physiological rates in Caenorhabditis elegans by a tRNA-modifying enzyme in the mitochondria. Genetics 159: 147–157 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lowe TM 2011. A genomic tRNA database. http://gtrnadbucscedu/
  28. Madison JT, Kung HK 1967. Large oligonucleotides isolated from yeast tyrosine transfer ribonucleic acid after partial digestion with ribonuclease T1. J Biol Chem 242: 1324–1330 [PubMed] [Google Scholar]
  29. Motorin Y, Bec G, Tewari R, Grosjean H 1997. Transfer RNA recognition by the Escherichia coli Δ2-isopentenyl-pyrophosphate:tRNA Δ2-isopentenyl transferase: Dependence on the anticodon arm structure. RNA 3: 721–733 [PMC free article] [PubMed] [Google Scholar]
  30. Persson BC, Esberg B, Olafsson O, Bjork GR 1994. Synthesis and function of isopentenyl adenosine derivatives in tRNA. Biochimie 76: 1152–1160 [DOI] [PubMed] [Google Scholar]
  31. Rafalski A, Kohli J, Agris P, Soll D 1979. The nucleotide sequence of a UGA suppressor serine tRNA from Schizosaccharomyces pombe. Nucleic Acids Res 6: 2683–2695 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Seif E, Hallberg BM 2009. RNA–protein mutually induced fit: Structure of Escherichia coli isopentenyl-tRNA transferase in complex with tRNA(Phe). J Biol Chem 284: 6600–6604 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Soderberg T, Poulter CD 2000. Escherichia coli dimethylallyl diphosphate:tRNA dimethylallyltransferase: Essential elements for recognition of tRNA substrates within the anticodon stem–loop. Biochemistry 39: 6546–6553 [DOI] [PubMed] [Google Scholar]
  34. Soderberg T, Poulter CD 2001. Escherichia coli dimethylallyl diphosphate:tRNA dimethylallyltransferase: Site-directed mutagenesis of highly conserved residues. Biochemistry 40: 1734–1740 [DOI] [PubMed] [Google Scholar]
  35. Spinola M, Galvan A, Pignatiello C, Conti B, Pastorino U, Nicander B, Paroni R, Dragani TA 2005. Identification and functional characterization of the candidate tumor suppressor gene TRIT1 in human lung cancer. Oncogene 24: 5502–5509 [DOI] [PubMed] [Google Scholar]
  36. Srinivasan M, Mehta P, Yu Y, Prugar E, Koonin EV, Karzai AW, Sternglanz R 2011. The highly conserved KEOPS/EKC complex is essential for a universal tRNA modification, t6A. EMBO J 30: 873–881 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Thuriaux P, Minet M, Hofer F, Leupold U 1976. Genetic analysis of antisuppressor mutants in the fission yeast Schizosaccharomyces pombe. Mol Gen Genet 142: 251–261 [DOI] [PubMed] [Google Scholar]
  38. Vacher J, Grosjean H, Houssier C, Buckingham RH 1984. The effect of point mutations affecting Escherichia coli tryptophan tRNA on anticodon–anticodon interactions and on UGA suppression. J Mol Biol 177: 329–342 [DOI] [PubMed] [Google Scholar]
  39. Zachau HG, Dutting D, Feldmann H 1966. The structures of two serine transfer ribonucleic acids. Hoppe Seylers Z Physiol Chem 347: 212–235 [DOI] [PubMed] [Google Scholar]
  40. Zhou C, Huang RH 2008. Crystallographic snapshots of eukaryotic dimethylallyltransferase acting on tRNA: Insight into tRNA recognition and reaction mechanism. Proc Natl Acad Sci 105: 16142–16147 [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from RNA are provided here courtesy of The RNA Society

RESOURCES