Recognition between tRNASer and archaeal seryl-tRNA synthetases monitored by suppression of bacterial amber mutations


  • Editor: Marco Moracci

Correspondence: Ivana Weygand-Durasevic, Department of Chemistry, Faculty of Science, University of Zagreb, Horvatovac 102a, 10000 Zagreb, Croatia. Tel.: +385 1 460 6230; fax: +385 1 460 6401; e-mail:


Two dissimilar seryl-tRNA synthetases (SerRSs) exist in Methanosarcina barkeri: one of bacterial type (bMbSerRS) and the other resembling SerRSs present only in methanogenic archaea (mMbSerRS). While the expression of the archaeal bMbSerRS gene in Escherichia coli complements the function of thermolabile SerRS at a nonpermissive temperature, mMbSerRS does not. Our recent X-ray structural analysis of mMbSerRS revealed an idiosyncratic N-terminal domain and a catalytic zinc ion in the active site, identifying methanogenic-type SerRSs as atypical members of the SerRS family. To shed further light on substrate discrimination by methanogenic-type SerRS, we developed an in vivo system in E. coli to study tRNA serylation by mMbSerRS variants. We show that coexpression of the M. barkeri SerRS gene, encoding either bacterial- or methanogenic-type SerRS, with the gene for cognate archaeal suppressor tRNA leads to suppression of bacterial amber mutations, implying that the E. coli translation machinery can use serylated tRNA from methanogenic archaea as a substrate in protein synthesis. Furthermore, because serylation of M. barkeri serine-specific tRNA by endogenous E. coli SerRS is negligible, suppression is entirely dependent on recognition between archaeal partners (mMbSerRS/suppressor tRNASer). Thus, the efficiency of suppression by mMbSerRS variants quantified in the described β-galactosidase-based reporter system, accurately reflects enzymes' serylation propensity obtained by in vitro kinetic measurements.


In general, aminoacyl-tRNA synthetases (aaRSs) must discriminate among a population of tRNA species to precisely aminoacylate their cognate tRNAs (Ibba & Söll, 2000). The mode of recognition of tRNA depends on identity elements, as well as the number of tRNA molecules that carry partially overlapping determinants.

tRNA recognition by seryl-tRNA synthetases (SerRSs) is of particular interest primarily because there are two distinct serine-charging enzymes: a standard or a bacterial type of SerRS was found in the majority of organisms (prokaryotes, eukaryotes and archaea), while a highly diverged SerRS (methanogenic-type) is confined to the methanogenic archaea (Methanococcales). We have previously shown cross-species replacement of bacterial-type SerRSss in several systems: Saccharomyces cerevisiae SerRS or Zea mays organellar SerRS complemented a temperature-sensitive gene of the bacterial enzyme in Escherichia coli (Weygand-Durasevic et al., 1993; Rokov et al., 1998); Z. mays organellar SerRS complemented the S. cerevisiae strain bearing a disrupted mitochondrial SerRS gene (Rokov-Plavec et al., 2002); and Z. mays cytosolic SerRS replaced the function of the disrupted cytosolic SerRS gene in S. cerevisiae (Mocibob & Weygand-Durasevic, 2008). Because methanogenic-type enzymes are characterized by a unique fold of the tRNA-binding region, which significantly differs from the corresponding region found in bacterial-type SerRSs, and contain a catalytically important zinc ion in the active site (Bilokapic et al., 2006), we examined the cross-domain serylation ability of these atypical SerRSs. Our experiments revealed that methanogenic-type SerRSs recognize eukaryotic and bacterial tRNAsSerin vitro, in addition to their homologous tRNASer and tRNASec substrates (Bilokapic et al., 2004; Gruic-Sovulj et al., 2006). This shows that some serine determinants are conserved in all three domains of life.

We have recently focused on seryl-tRNASer formation in Methanosarcina barkeri, because this archaeon comprises two dissimilar SerRSs: one of a bacterial and the other of a methanogenic type (bMbSerRS and mMbSerRS, respectively). In order to investigate their requirements for tRNASer recognition, variant transcripts of M. barkeri tRNASer were kinetically analyzed by both enzymes. Although two synthetases successfully serylate transcripts of all three archaeal tRNASer isoacceptors, the two SerRSs do not possess a uniform mode of tRNASer recognition (Korenčićet al., 2004). In order to shed further light on the nature of serylation in methanogenic archaea, we set out to develop an in vivo system suitable for monitoring the interaction between archaeal SerRSs and their cognate tRNAs. Although a number of methanogenic-type SerRSs from various archaeal species failed to complement the function of bacterial thermolabile SerRS, our results revealed that coexpression of M. barkeri bacterial- (bMbSerRS) or methanogenic-type SerRS (mMbSerRS) with the cognate tRNA leads to suppression of bacterial amber mutations. This implies that the E. coli translation machinery can use serylated tRNAs from methanogenic archaea as substrates in protein synthesis. Furthermore, suppression is predominantly dependent on recognition between archaeal partners, because serylation of archaeal tRNASer by endogenous bacterial synthetase is negligible, which is in accordance with previous in vitro experiments (Kim et al., 1998; Rokov-Plavec et al., 2004; Bilokapic et al., 2008). A number of mMbSerRS mutants carrying mutations in the active site were tested as well, revealing a much lower suppression efficiency, fully in agreement with our recent in vitro results (Bilokapic et al., 2008).

Materials and methods


Oligonucleotides were synthesized and DNAs were sequenced by Invitrogen. [14C]Serine (160 μCi mmol−1) was from Perkin Elmer Life Sciences Inc. Restriction enzymes were from New England Biolabs. The strains and plasmids are listed in Table 1.

Table 1.   Strains and plasmids used in this study
Strains of E. coli usedGenotype and descriptionReferences
 KL229FserS15(ts) thyA6 rpsL120Low et al. (1971)
 JR104F′trpA(UAG)211/glyV55Δ(tonB-trpAB) argE(UAG) rpoBWeygand-Durasevic et al. (1993)
 XACΔ14F′ara argE(UAG) rpoB gyrAΔlac pro/F′lacI–Z pro AB+Miller & Albertini (1983)
 XAC-A24F′ara argE(UAG) rpoB gyrAΔlac pro/F′lacI(UAG)–Z proABMiller & Albertini (1983)
Plasmids used
 pBAD24Contains an inducible BAD promoterGuzman et al. (1995)
 pBADEcSerRSPlasmid pBAD24 containing the gene for E. coli SerRS cloned behind the BAD promoterThis work
 pBADScSerRSPlasmid pBAD24 containing the gene for Saccharomyces cerevisiae SerRS cloned behind the BAD promoterThis work
 pBADmMbSerRSPlasmid pBAD24 containing the gene for methanogenic-type M. barkeri SerRS cloned behind the BAD promoterThis work
 pBADbMbSerRSPlasmid pBAD24 containing the gene for bacterial-type M. barkeri SerRS cloned behind the BAD promoterThis work
 pBADSerRS mutantsPlasmid pBAD24 containing the gene for mMbSerRS mutants cloned behind the BAD promoterThis work
 pET15bMbSerRSPlasmid pET15b containing the gene for bacterial-type M. barkeri SerRS cloned behind the lac promoterKorencic et al. (2004)
 pET15mMbSerRSPlasmid pET15b containing the gene for methanogenic-type M. barkeri SerRS cloned behind the lac promoterKorencic et al. (2004)
 pET15SerRS mutantsSame as above, but containing various mutated mMbSerRS genesBilokapic et al. (2006, 2008)
 pTechContains a constitutive lpp promoterMin et al. (2003)
 pTechsupSMbPlasmid pTech containing the gene for M. barkeri suppressor tRNASer cloned behind the lpp promoterThis work
 pACYCsupSIS. pombe supSI gene cloned between EcoRV and SalI sites of pACYC, distal to the tet promoterWeygand-Durasevic et al. (1994)
 pACYCsupSHHuman supSH gene cloned between EcoRV and SalI sites of pACYC, distal to the tet promoterWeygand-Durasevic et al. (1994)
 pTechsupFPlasmid pTech containing the gene for E. coli suppressor tRNATyr cloned behind the lpp promoterThis work
 pTechsupFG73Plasmid pTech containing the mutant gene for E. coli suppressor tRNATyr (A73G) cloned behind the lpp promoterThis work

SerRS constructs

Methanogenic and bacterial SerRS genes were identified in the M. barkeri genomic DNA sequence ( They were amplified by PCR using Expand High Fidelity polymerase, and cloned into pET15b vector, yielding pET15bMbSerRS and pET15mMbSerRS expression plasmids (Korenčićet al., 2004) encoding bacterial- and methanogenic-type SerRSs, respectively. In order to prepare mMbSerRS variants, site-directed mutagenesis was carried out using a Quik change mutagenesis kit (Stratagene, La Jolla, CA) as described (Bilokapic et al., 2008). NcoI–XhoI fragments containing synthetase genes with the N-terminal His6-tag were then recloned into the pBAD24 vector, where the SerRS expression is under control of an arabinose-inducible promoter.

Construction of synthetic suppressor tRNA gene

tRNASer genes were identified from the M. barkeri genomic DNA sequence at the Joint Genome Institute (JGI), using tRNAscan-SE ( The sequence for the isoacceptor with the CGA anticodon (Korenčićet al., 2004) was converted into one comprising the anticodon complementary to the amber STOP codon. A synthetic gene for M. barkeri suppressor tRNA was constructed from two overlapping synthetic oligonucleotides and inserted between BamHI and PstI sites of the pTech plasmid, placing transcription under control of the lpp promoter and the rrnC terminator.

Complementation assay in E. coli

Escherichia coli strain KL229 was transformed with plasmids carrying SerRS genes from E. coli and M. barkeri. To test for complementation, transformants were plated on M9 minimal plates (containing 200 μg mL−1 thymine) supplemented with 100 mg L−1 ampicillin and, when necessary, with 0.2% arabinose and grown overnight at 30 or 37 °C.

Suppression of E. coli amber mutations

Escherichia coli strains XAC-A24 and JR 104 were described in Weygand-Durasevic et al. (1994). The genotype is given in Table 1. Suppression of argE or trpA amber mutations in strains JR104 and XAC-A24 was tested by plating E. coli cells on selective M9 minimal glucose plates. The efficiency of suppression was determined in the E. coli strain XAC-A24 by measuring the β-galactosidase activity produced from a lacI–lacZ fusion harboring a nonsense mutation in the lacI portion (Coulondre & Miller, 1977). The truncated protein resulting from premature termination of protein synthesis at the lacI in-frame stop codon is unable to degrade chromogenic 2-nitrophenyl β-d-galactopyranoside, while suppression of the nonsense mutation results in the synthesis of functional β-galactosidase in E. coli (Miller, 1972 and Supporting Information). Suppression is defined as 100% from the β-galactosidase activity of strain XACΔ14, which contains a lacI–Z fusion with no amber mutation. Values are the average of triplicate measurements.

Results and discussion

Complementation of SerRS function in E. coli strain KL229

Bacterial and methanogenic M. barkeri serS genes were tested for their ability to complement E. coli strain KL229 encoding thermolabile SerRS. Transformants of KL229 cells with arabinose-inducible pBAD plasmids, carrying M. barkeri or E. coli serS genes, were grown at permissive (30 °C) and nonpermissive (37 °C) temperatures (Low et al., 1971; Weygand-Durasevic et al., 1993). While the transformants carrying bacterial M. barkeri and E. coli serS genes showed efficient growth at both temperatures, the methanogenic-type M. barkeri serS was unable to complement the function of thermolabile E. coli serS at a nonpermissive temperature (Korenčić, 2004 and Fig. 1).

Figure 1.

 Complementation of SerRS function in Escherichia coli strain KL229. Individual colonies of KL229 transformants, carrying bacterial or Methanosarcina barkeri SerRS genes on pBAD plasmids at 30°C, were restreaked to fresh plates [M9glu containing 200 μg mL−1 thymine, 0.2% arabinose and ampicillin (100 μg mL−1)] and grown at 30 and 37°C.

Given the fact that methanogenic-type SerRSs efficiently serylate E. coli tRNASerin vitro (Bilokapic et al., 2004), lack of complementation was not expected. The experiment was repeated by varying the SerRS expression level in the presence of different arabinose concentrations or by introducing the synthetase gene on different plasmids (not shown). Although the protein extract from transformed KL229 cells grown in the presence of arabinose showed a 60-fold higher serylation activity than the extract from noninduced transformants (not shown), complementation was not achieved. Besides M. barkeri SerRS, we tested for complementation of a number of methanogenic-type SerRSs from other archaeal species (Supporting Information, Table S1), which were all unable to replace bacterial SerRS function in vivo. It may be that toxic effects (Bedouelle et al., 1990; Gagnon et al., 1996) caused by the overexpression of methanogenic-type SerRSs in E. coli preclude detection of complementation. Toxicity could be explained as a result of misincorporation of amino acids into proteins mediated by erroneously aminoacylated tRNAs. On the other hand, the lack of complementation may be caused by the inability of methanogenic-type SerRSs to serylate all E. coli tRNASer isoacceptors.

Thus, in our search for in vivo assays that would allow monitoring the specificity of recognition between archaeal partners (mMbSerRS/tRNASer), we used the suppression assay.

Suppression of E. coli amber mutations by archaeal tRNAsSer requires coexpression of active archaeal SerRS

Bacterial strains JR104 and XAC-A24 (Table 1) carry several amber mutations, whose suppression may reflect recognition and aminoacylation levels of suppressor tRNA by selected aaRSs. Both strains have an argE(UAG) mutation suppressible by any amino acid. Amber mutation in strain JR104, at the position 211 in the trpA gene, can be suppressed with only a limited set of amino acids, including serine (but not tyrosine) (Murgola, 1985), and is therefore also useful for testing misacylation of noncognate tRNAs with serine. The other suppressible marker in strain XAC-A24, in which a UAG in-frame codon has been inserted in to the lacI part of a lacI–lacZ fusion gene, was used for quantification of suppression efficiency (Normanly et al., 1986; McClain & Foss, 1988; Jahn et al., 1991; Polycarpo et al., 2006).

We have previously shown that coexpression of eukaryotic SerRS with cognate serine-specific tRNAs leads to efficient suppression of bacterial amber mutations (Weygand-Durasevic et al., 1994). Serine-specific tRNAs are especially suitable to be used in such assays, because the anticodon is not a recognition element for interaction with the cognate synthetase, and its alteration does not change the tRNA identity (Asahara et al., 1994). There are three serine-specific isoacceptors in M. barkeri (Korenčićet al., 2004) and all are successfully aminoacylated by both homologous SerRSs (Korenčićet al., 2004), one of methanogenic and the other of bacterial type, although with different efficiencies (Korenčićet al., 2004). tRNASerCGA and tRNASerGGA isoacceptors were similar in their kinetic properties towards both SerRSs; however, tRNASerGCT seemed to be a preferable substrate for the bacterial-type enzyme in vitro. Because we were primarily interested in the recognition properties of methanogenic-type SerRS in vivo, we converted the tRNASerCGA isoacceptor sequence into the tRNASer suppressor sequence (supSMb) and placed it behind the lpp promoter in pTech plasmid (pTechsupSMb). No suppression of bacterial mutation argE(UAG) was achieved upon expression of M. barkeri suppressor tRNASer in the JR104 tester strain (Fig. 2). In accordance with the in vitro results (Kim et al., 1998; Rokov-Plavec et al., 2004; Bilokapic et al., 2008), this experiment confirms that serine-specific tRNAs from methanogenic archaea are poor substrates for E. coli SerRSs (Fig. 2). In the second experiment, the tester strains were cotransformed with a pair of compatible plasmids, one of which carried the gene for a methanogenic-type synthetase, while the second carried the M. barkeri suppressor tRNASer sequence. Archaeal partners suppressed all tested bacterial amber mutations and behaved practically as an orthogonal tRNA/synthetase pair in our reporter system (Fig. 2). This is because serylation of archaeal tRNASer by endogenous bacterial synthetase is negligible (Kim et al., 1998; Rokov-Plavec et al., 2004; Bilokapic et al., 2008), while heterologous serylation of cognate bacterial tRNA isoacceptors observed in vitro (Bilokapic et al., 2004, 2008) cannot be monitored in the suppressor assay described above. In accordance with in vitro studies (Kim et al., 1998; Rokov-Plavec et al., 2004; Bilokapic et al., 2008), overexpression of E. coli SerRS from the pBAD vector together with pTechsupSMb results in very low suppression (Fig. 2). We also show that coexpression of eukaryotic serine-specific tRNAs from pACYCsupSI and pACYCsupSH (encoding Schizosaccharomyces pombe and Homo sapiens tRNASer suppressors, respectively) and mMbSerRS did not result in suppression of bacterial amber mutations (Fig. 2). Our earlier studies revealed that both pACYCsupSI and pACYCsupSH are active suppressors that led to suppression of bacterial nonsense mutation after coexpression with S. cerevisiae SerRS (Weygand-Durasevic et al., 1994). Furthermore, supSI, S. pombe-derived suppressor tRNA, does not appear to be a substrate for mMbSerRS in vivo, further supporting the importance of specific tRNA synthetase recognition for efficient suppression. The results obtained in vitro show that eukaryotic tRNAsSer are poor substrates for M. barkeri methanogenic-type SerRS. The S. cerevisiae tRNASer is an even less preferred substrate for M. barkeri SerRS, then for other methanogenic-type SerRSs (S. Bilokapic, pers. commun., Bilokapic et al., 2004; Gruic-Sovulj et al., 2006), while human tRNASer could not be serylated by M. barkeri SerRS (S. Bilokapic, pers. commun.).

Figure 2.

 Suppression of Escherichia coli amber trpA(UGA) mutation in strain JR104. Efficient suppression by archaeal tRNASer suppressor requires expression of active archaeal SerRS. Escherichia coli strain JR104 was cotransformed with a vector carrying the amber suppressor tRNA gene (Methanosarcina barkeri, Schizosaccharomyces pombe or Homo sapiens tRNASer,) and a compatible vector carrying a serine-specific synthetase gene (methanogenic-type M. barkeri SerRS gene and its inactive mutants, bacterial M. barkeri, Saccharomyces cerevisiae and E. coli SerRS genes). Suppression of trpA(UGA) was checked by streaking the transformants on M9 minimal glucose plates supplemented with arabinose, arginine and required antibiotics and incubating at 37°C for 20 h.

To confirm that active mMbSerRS is needed for efficient tRNA aminoacylation and suppression, inactive synthetase mutants were constructed (Fig. 3) and assayed in vivo. In accordance with the observation that the activity of M. barkeri SerRS depends on the active site zinc ion involved in serine binding (Bilokapic et al., 2006), no suppression was obtained when mMbSerRS variants, bearing altered serine ligands Cys306 and Cys461, were coexpressed with archaeal suppressor tRNA (Fig. 2). Expression of inactive mMbSerRS mutant proteins in the JR104 suppressor strain was confirmed by a Western blot (Fig. S1).

Figure 3.

 View on the active site of the mMbSerRS with the bound seryl-adenylate analogue and modeled CCA end of tRNA. Active site zinc ion is in cyan. Indicated residues have been mutated, and the serylation propensity of mMbSerRS variants was analyzed in vitro (Bilokapic et al., 2006, 2008) and in vivo (this work). Cys306 and Cys461 function (together with Glu355) as zinc-binding residues. Trp396 determines the size of the active site, and Glu338 and Arg347 are presumably involved in both steps of the aminoacylation reaction by interacting with ATP and the 3′-end of tRNASer. Two consecutive glycines (Gly340 and Gly341) assure conformational flexibility of the motive 2 loop, which is required for efficient serylation. Arg267 is located in the idiosyncratic helix-turn-helix motif of mSerRSs and presumably contributes to the interactions with the acceptor end of the tRNA.

The efficiency of suppression by practically orthogonal supSMb/mMbSerRS pair reveals the serylation propensity of archaeal synthetase variants in vivo

In order to analyze the contribution of individual amino acids to substrate binding and catalysis, mutations were introduced at various positions of the mMbSerRS gene, either according to the crystal structure of the enzyme complexed with its cognate aminoacyl-adenylate or on the basis of modeling the mMbSerRS:tRNA complex (Bilokapic et al., 2006). Mutated residues are indicated in Fig. 3. Enzyme variants were characterized by determination of aminoacylation kinetic parameters (Bilokapic et al., 2008) and the ability to serylate the suppressor tRNA in vivo (this work).

The XAC-A24 β-galactosidase-based reporter system was used to quantify the suppression obtained upon coexpression of mMbSerRS variants and supSMb. Expression of synthetases or their mutated variants was either from pBAD24, where it is dependent on induction with arabinose (Guzman et al., 1995), or from pET15b plasmids, which enables constitutive expression by inefficient recognition of T7 promoter by bacterial RNA polymerase (Weygand-Durasevic et al., 1993). Because mMbSerRS expression was quite low from the pET15b plasmid, several synthetase variants with the altered active site did not sustain the growth of transformed XAC-A24 cells (Table 2). Their genes were recloned to the pBAD vector. Higher expression of mutated synthetases led to the production of higher quantities of serylated suppressor tRNA, which was required for rescuing the amber mutation (Table 2). The results presented are fully in agreement with our recent kinetic analyses of mMbSerRS variants in amino acid activation and/or tRNA serylation reactions (Bilokapic et al., 2006, 2008). Alteration of zinc-binding residues (Cys306 and Cys461) resulted in an enzyme that was inactive in vivo (Table 2) and in vitro (Bilokapic et al., 2006), while the mutant with the modified serine-binding pocket (W396A) displayed significant reductions in suppression efficiency (Table 2) and decreased catalytic efficiency (Kcat/Km) for serine (Bilokapic et al., 2008) compared with the wild-type mMbSerRS. Furthermore, results obtained both in vivo (Table 2) and in vitro (Bilokapic et al., 2008) are consistent with the involvement of several motif 2 residues (Glu338, Arg347, Gly340 and Gly341) in the serylation reaction. We have also confirmed the importance of Arg267 in the aminoacylation reaction. This residue is located in the idiosyncratic helix-turn-helix fold of mMbSerRS and presumably contributes to the interactions with the acceptor end of tRNA (Bilokapic et al., 2008). The growth of the transformants on selective plates (the result of argE suppression) and the suppression efficiency, measured after suppression of lacI–lacZ, faithfully reflect the serylation specificity constant calculated from in vitro determined kinetic parameters (Table 2). Thus, the system is useful for monitoring the serylation potential of methanogenic-type archaeal SerRS in bacterial cells in vivo. The expression of all mMbSerRS variants in XAC-A24 strain has been verified by a Western blot (Fig. S2).

Table 2.   Suppression efficiency of Methanosarcina barkeri SerRS variants
EnzymetRNARel Kcat/Km*,†Expression from pET15bExpression from pBAD24
w/o Arg
w/o Arg
  1. Relative specificity constants (Kcat/Km)wt/(Kcat/Km)mut for Methanosarcina barkeri tRNA and serine* determined in aminoacylation reaction with wild-type (wt) and mutant (mut) mMbSerRS enzymes have been taken from Bilokapic et al. (2008).

  2. Growth without arginine (Arg) was tested on M9 minimal plates (incubated for 18 h at 37°C), while suppression efficiency was measured by assaying β-galactosidase activity in Escherichia coli strain XAC/A24 as described in Materials and methods. An activity of 100% corresponds to the β-galactosidase activity of strain XACΔ14, which contains a lacI–Z fusion with no amber mutation. Results were also reported as the percentage of mutant enzyme suppression activity relative to that of the wild-type enzyme.

  3. ND, not determined.

mMbSerRSsupMb1+++25.97 ± 3.12100 ± 0.12+++32.18 ± 4.13100 ± 0.13
bMbSerRSsupMb +++10.53 ± 1.5640.54 ± 0.06+++14.41 ± 3.5244.78 ± 0.11
mR267AsupMb16.12*+5.28 ± 1.4520.33 ± 0.06NDNDND
mE338AsupMb5.59*+5.14 ± 0.8919.79 ± 0.03NDNDND
mW396AsupMb134.80+2.05 ± 0.956.36 ± 0.04+2.92 ± 0.689.06 ± 0.02
mG340V/G341AsupMb887.90*NDND+1.63 ± 0.565.05 ± 0.02
mR347AsupMb377.23*NDND+1.31 ± 0.894.07 ± 0.03
mC306A/C461AsupMbNDND0.23 ± 0.170.72 ± 0.01
mC306supMbNDND0.27 ± 0.150.89 ± 0.01
supMb 0.24 ± 0.130.76 ± 0.010.29 ± 0.190.90 ± 0.01
 0.26 ± 0.140.78 ± 0.010.26 ± 0.200.88 ± 0.01

Discrimination of tRNAs with overlapping identity elements in vivo and in vitro

Identity elements required for serylation have been studied in a number of organisms, providing insights into tRNASer recognition in different domains of life (rewieved in Korenčićet al., 2004; Weygand-Durasevic & Cusack, 2005). While the length and the position of the variable arm were identified as crucial determinants for serylation, the importance of the discriminator base varies in different organisms. For example, G73 serves only as an antideterminant in Bacteria (Asahara et al., 1994) and lower eukaryotes (Himeno et al., 1997), but it is an essential identity requirement for human tRNASer (Breitschopf & Gross, 1994; Lenhard et al., 1999). Furthermore, it is known that tRNA recognition is contextually dependent, and identity requirements for a given system are influenced by characteristics of other isoaccepting systems with similar tRNA substrates. While in bacteria, three families of tRNA isoacceptors carry a long extra arm (Ser, Leu and Tyr), there are only two such tRNAs in archaea and eukaryotes (specific for Ser and Leu) (Nazarenko et al., 1992; Breitschopf & Gross, 1994; Lenhard et al., 1999). We supposed therefore that M. barkeri SerRSs would have developed a natural manner of discrimination against tRNALeu, but that archaeal enzymes could mischarge E. coli tRNATyr.

We have expressed from pTech plasmid the supF (Sherman et al., 1992), an efficient tyrosine-specific suppressor tRNA, comprising an adenine at the position of the discriminator base and its mutated variant carrying G73 [supF(G73)]. Suppression was tested in strain JR104 after coexpression (with supF) of either mMbSerRS or bMbSerRS from pBAD plasmids (Fig. 4) Mischarging of supF and supF(G73) by endogenous E. coli SerRS is very low, but increases proportionally with the level of E. coli enzyme, which is in accordance with the previously reported promiscuity of those suppressor tRNAs (Sherman et al., 1992). Coexpression of supF(G73), but not supF, with bMbSerRS, leads to efficient suppression. This indicates a strong preference of bacterial-type M. barkeri SerRS for G at the position of the discriminator base. Furthermore, the preference is stronger than for E. coli SerRS, which, when overexpressed, weakly recognized the original supF suppressor (Fig. 4). Thus, in accordance with our in vitro data (Korenčićet al., 2004), our in vivo results further suggest that the bacterial-type SerRS from M. barkeri adopted G73 as one of the major elements of serine identity. In contrast, no increase in the suppression level relative to the endogenous EcSerRS control was observed after coexpression of methanogenic-type SerRS with either supF or supFG73, implying a higher dependence of the bacterial-type SerRS on G73 identity element relative to the methanogenic-type archaeal SerRS. The same trend in tRNA recognition pattern exists in vitro: mutational analysis shows a dramatic decline of aminoacylation efficiencies for variants of G73 for bMbSerRS and a less severe decrease of the specificity constant for mMbSerRS (Korenčićet al., 2004). The discriminator base and bases from the first three pairs of the acceptor stem are comparatively most conserved in M. barkeri tRNAsSer, which suggests their involvement in specific tRNASer recognition. Accordingly, our in vivo assay sensitively reports the changes at the position of the discriminator base.

Figure 4.

 Mischarging of suppressor tRNATyr from Escherichia coli. Dependence on G73 is more pronounced for bMbSerRS than for mMbSerRS. Suppression of trpA(UGA) was checked by streaking the transformants on M9 minimal glucose plates supplemented with arabinose, arginine and required antibiotics and incubating at 37°C for 20 h.


We thank Davor Katusic for technical help and Silvija Bilokapic for discussing the mMbSerRS structure and preparing the related figure. This work was supported by grants from the Ministry of Science, Education and Sports of the Republic of Croatia (project 119-0982913-1358) and the Unity Through Knowledge Fund (project 10/07).