Laboratoire de Génétique Microbienne, Institut National de la Recherche Agronomique, 78352Jouy-en-Josas Cedex, France1
Author for correspondence: Maarten van de Guchte. Tel: +33 1 34 65 25 28. Fax: +33 1 34 65 25 21. e-mail: guchte{at}biotec.jouy.inra.fr
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Keywords: gene regulation, T-box, tRNA, Lactococcus lactis
Abbreviations: CDM, chemically defined medium
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The non-translated leader transcripts of these genes or operons share strikingly similar secondary structures in which the specifier codon occupies a crucial, conserved position (Fig. 1; Grundy & Henkin, 1993
; Chopin et al., 1998
; Luo et al., 1998
). This codon corresponds to the amino acid which the operon is responsible for the biosynthesis of, or in the case of aminoacyl-tRNA-synthetase-encoding genes, to the amino acid that is coupled to a tRNA by that particular synthetase, and is thought to bind to the corresponding tRNA in a codonanticodon interaction. A second conserved position in the secondary structure is reserved for the anti-acceptor which is situated in the actual terminator and is complementary to the four 3' terminal nucleotides of the acceptor arm of the tRNA to which this sequence is thought to bind (Grundy & Henkin, 1993
; Grundy et al., 1994
; Van de Guchte et al., 1998
). The latter interaction explains the regulatory potential of this system. An uncharged tRNA can bind to the leader transcript, maintain the antiterminator conformation and thus promote the expression of the operon or gene. A tRNA that is charged with an amino acid on the acceptor arm, however, would not be able to bind and stabilize the antiterminator in the leader transcript. The latter adopts the energetically more favourable terminator conformation and expression of the operon or gene is suppressed.
|
|
These results showed that when acceptor and anti-acceptor are compatible, specifier codonanticodon complementarity is neither sufficient for an efficient response to occur nor strictly indispensable for a low level response, and consequently suggested the existence of additional means of recognition between the trp leader transcript and tRNATrp. Here we present experimental results that strongly suggest the implication of the tRNA D- and T-arms in the recognition of conserved stemloop structures in the trp leader transcript.
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
DNA manipulations and transformation.
E. coli and L. lactis were transformed by electroporation (Dower et al., 1988 ; Holo & Ness, 1995
). Plasmid DNA was isolated by the method of Birnboim & Doly (1979)
. Restriction enzymes and T4 DNA ligase were purchased from Boehringer or New England Biolabs and used according to the instructions of the suppliers. Taq DNA polymerase and Vent DNA polymerase for use in PCR were purchased from Promega and New England Biolabs, respectively. Polynucleotide kinase was purchased from New England Biolabs. PCR-mediated DNA amplification was performed using a Perkin Elmer Cetus DNA thermal cycler 2400 or 9600. DNA sequence analysis was performed using an Applied Biosystems 373 automated DNA sequencer. Oligonucleotides were synthesized using a Beckman Oligo 1000M DNA synthesizer.
PCR-directed mutagenesis and overexpression of tRNAs.
Mutations in the lactococcal trp leader region were generated by PCR using mutagenic primers as described previously (Van de Guchte et al., 1998 ). Mutated leader fragments (EcoRIHindIII) were subsequently cloned in the delivery vector pIL931 and integrated upstream of the E. coli lacZ gene in L. lactis IL56012 to reconstitute a functional trpE-lacZ fusion on the chromosome. The fusion is located within the his operon, which is transcriptionally silent in the presence of histidine in the culture medium (Delorme et al., 1999
). Recombinant tRNA sequences were generated by PCR using two partially overlapping oligonucleotides and subsequently cloned under control of the lactococcal promoter P59 (Van der Vossen et al., 1987
) in pNuc9 as described previously (Van de Guchte et al., 1998
).
RNA manipulations.
For RNA extraction, cultures were grown in the presence of Trp, Asn and erythromycin to an OD600 of 0·6. The cells were collected by centrifugation and resuspended in an equal volume of fresh medium containing or lacking Asn or Trp. After incubation for 30 min in these media at 30 °C, the cells were harvested and RNA was extracted as described by Raya et al. (1998) . Aliquots of 20 µg were subjected to electrophoresis in 1% agarose and blotted onto Nytran membranes (NY12N; Schleicher & Schuell). Northern hybridizations were performed using either of the following
-32P labelled oligonucleotide probes: 5'-CTGGTCACTTCGATGGTT-3' (complementary to the E. coli lacZ gene) or 5'-GGCGTATGTCACGCGGTGCCACCG-3' (complementary to lactococcal trp leader transcripts). Transcripts were quantified using a PhosphorImager (Storm; Molecular Dynamics). The relative readthrough was determined as described previously (Van de Guchte et al., 1998
). Briefly, the relative amount of full-length transcript was determined using a lacZ-specific probe and a correction for the Trp-independent variation in the level of transcription initiation (Raya et al., 1998
) and possible gel charging errors was made after rehybridization with a trp leader-specific probe. Comparison of the signals obtained with the lacZ probe for different lanes on the same Northern blot allowed, after correction based on the signals obtained with the trp leader probe, calculation of a relative antitermination value for each lane (relative to a given lane on the same blot). The presented relative antitermination values are mean values from two to five experiments.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
|
Sequence complementarity between tRNATrp and the trp leader transcript
The above observations prompted a renewed analysis of the trp leader. The lactococcal tRNATrp sequence, which has recently become available (P. Renault, unpublished), was used to search for mRNAtRNA complementarity that could be indicative of additional recognition mechanisms. As shown in Fig. 4(a), there is extensive sequence complementarity between the trp leader and tRNATrp. A stretch of 21 nt in stemloop structure IV of the trp leader (Fig. 1
) is complementary to the D-arm and the anticodon arm of tRNATrp, whereas another stretch of 18 nt is complementary to the T-arm. Moreover, parts of the latter sequence are repeated several times upstream of the 18 nt sequence in stemloop structure II of the trp leader (Fig. 1
). This large extent of mRNAtRNA complementarity suggests a functional role for stemloop structures II and IV in the establishment or maintenance of the mRNAtRNA interaction. If these complementary regions were important for an efficient mRNAtRNA interaction, it would be expected that tRNAAsn would show less, or differently structured, complementarity to the trp leader than tRNATrp, since the specifier (AAC) mutant showed only a weak response to Asn depletion. As shown in Fig. 4(b)
, tRNAAsn (P. Renault, unpublished) does indeed show substantially lower complementarity than tRNATrp to the trp leader. In addition, tRNAAsn differs markedly from tRNATrp in the size of its D-stem, D-loop and extra arm.
Effect of chimeric tRNAs on transcription antitermination in the specifier Asn mutant
To examine whether the above differences between tRNAAsn and tRNATrp could be responsible for the weak antitermination response in the specifier (AAC) mutant, chimeric tRNAs were constructed. Based on the L. lactis tRNAAsn gene, recombinant genes were created in which (i) the D-arm, (ii) the T-arm and extra arm or (iii) both sequences were replaced by the corresponding arms of tRNATrp (Fig. 5). In addition, a gene was created based on the tRNATrp gene in which the anticodon CCA (anti-TGG, Trp) was replaced by GTT (anti-AAC, Asn). The resulting tRNA can be regarded as a tRNAAsn in which all but the anticodon is replaced by the corresponding parts of tRNATrp. The different chimeric tRNAs all contain the identity elements of tRNAAsn, the anticodon GUU and the discriminator base G, and therefore are likely to be chargeable by asparaginyl-tRNA synthetase, even if tRNATrp and tRNAAsn are normally charged by synthetases that belong to two different classes (Giegé et al., 1998
).
tRNAAsn and the chimeric tRNAs were overproduced in the specifier (AAC) mutant and their effect on transcription antitermination in the trpE-lacZ fusion was studied. Under conditions of Asn limitation, the overproduction of tRNAAsn affected antitermination only a little or not at all (Fig. 3, lanes 2 and 5). In contrast, overproduction of tRNAAsn containing either the D-arm or the T-arm and extra arm of tRNATrp, resulted in a clearly enhanced level of antitermination which was increased even further when all these tRNATrp arms were present (Fig. 3
, lanes 6, 7, 8). Expression of the latter chimera resulted in a level of antitermination approaching that realized by the overproduction of tRNATrp in which the anticodon had been replaced by the anticodon GUU (anti-AAC, Asn) (Fig. 3
, lane 9). These results suggest an important role for the tRNA D- and T-arms in the mRNAtRNA interaction.
Since it is the ratio between the amount of uncharged and charged tRNA that determines whether, and to what extent, antitermination occurs (Grundy et al., 1994 ; Putzer et al., 1995
; Van de Guchte et al., 1998
), the possibility remained that the increased antitermination in the presence of chimeric tRNAs was due to an impaired chargeability of these tRNAs. To rule out this possibility, the experiments were repeated in the presence of an excess of Asn in the culture medium. Under these conditions, substitution of the D- or T-arm and extra arm in tRNAAsn did not induce antitermination (Table 2
, compare chimeric tRNAs 1 and 2 to tRNAAsn), indicating that these chimeric tRNAs can be charged efficiently and that enhanced antitermination during Asn starvation using these tRNAs thus results from an improved interaction with the trp leader. The overexpression of tRNAAsn in which both the D-arm and the T-arm were substituted, or tRNATrp with anticodon Asn resulted in antitermination to approximately 20% of the level obtained under conditions of Asn limitation (compare chimeric tRNAs 3 and 4 in Table 2
to the same tRNAs in Fig. 3
). This result indicates that these chimeric tRNAs can also be charged with Asn, albeit with reduced efficiency. The enhanced antitermination using these tRNAs thus appears to be due essentially to an improved interaction with the trp leader.
|
Effect of mutations in chimeric tRNAs on transcription antitermination
The lactococcal tRNATrp and tRNAAsn, and the chimeric tRNAs not only differ in the nucleotide sequences of their D-, T- and extra arms, but also in the sizes of the D-stem and D-loop, as well as the size of the extra arm, features that may also affect mRNAtRNA interaction. To address the question whether the antitermination-enhancing effects of the chimeric tRNAs could be ascribed to either an improved mRNAtRNA sequence complementarity or to an enhanced structural compatibility because of the sizes of the D- and extra arm, the effects of small mutations in tRNAAsn were studied.
To allow for maximum interaction between mRNA and tRNA, the chimeric tRNA containing the D-arm of tRNATrp was used to create mutations in the T-arm or in the extra arm, whereas the chimeric tRNA containing the T-arm of tRNATrp was used to create mutations in the D-arm. The resulting tRNAs are shown in Fig. 5. In tRNAs 5 and 7 the complementarity of the T- or the D-arm, respectively, to the trp leader has been improved, without changing the size of these arms or the extra arm. In tRNA 6 the size of the extra arm has been reduced, thereby also slightly improving the complementarity to the trp leader. In tRNA 8 the size of the D-loop has been reduced without improving mRNA complementarity.
The results of overproduction of these modified tRNAs, presented in Table 3, show that it was not possible to separate sequence complementarity and arm size effects. Only the overexpression of tRNA 6 resulted in a slightly enhanced antitermination. These results suggest that a combination of primary sequence and structure information is responsible for the correct interaction of mRNA and tRNA.
|
mRNAtRNA complementarity in analogous systems
On the basis of the conserved structural features of their leader transcripts, several aminoacyl-tRNA synthetase genes and amino acid biosynthesis operons in Bacillus subtilis and L. lactis have been postulated to be regulated by tRNA-mediated transcription antitermination (Grundy & Henkin, 1993 , 1994
; Henkin, 1994
; Van de Guchte et al., 1998
; Luo et al., 1998
; Chopin et al., 1998
; Delorme et al., 1999
). Taking the L. lactis trp leader as a model, we looked for the presence of mRNAtRNA sequence complementarity in the systems listed in Fig. 6
, focusing on complementarity between stemloop structure II of the leader transcript and the T-arm of the cognate tRNA and between stemloop structure IV and the D-arm of the tRNA.
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
In this specifier mutant of the lactococcal trp leader, antitermination can still be induced by Trp limitation, suggesting that interactions between tRNA and leader transcript other than the codonanticodon and acceptoranti-acceptor interactions actively contribute to the stabilization of the antiterminator conformation of the leader transcript. As overexpression of lactococcal tRNAAsn under conditions of Asn limitation had no inducing effect in the wild-type trp leader, these interactions appear to be part of a recognition mechanism allowing the trp leader to specifically interact with tRNATrp.
Asn starvation induced a low level of antitermination in the specifier (AAC) mutant. The overexpression of tRNAAsn did not affect this induction level. In contrast, the overexpression of chimeric tRNAs, consisting of tRNAAsn in which either the D-arm or the T-arm and extra arm were replaced by the corresponding sequences from tRNATrp, gave rise to increased antitermination. The overexpression of a chimeric tRNA containing both sequences from tRNATrp augmented antitermination still further, as did the overexpression of tRNATrp in which the anticodon had been replaced by the anti-Asn (anti-AAC) anticodon. The fact that in the single-arm replacement mutants antitermination was only observed in the absence of Asn from the culture medium indicates that these chimeric tRNAs can be charged efficiently and that, therefore, the enhanced antitermination is not due to an impaired charging of the tRNA, but to an intrinsic property of the replacement arm. In the multi-arm replacement mutants, enhanced antitermination largely depended on the absence of Asn, suggesting that here too the intrinsic arm properties rather than the partially impaired chargeability of these chimeric tRNAs are responsible for the enhancement.
Together, these results evoke an important role for the tRNA D- and T-arms in the specificity and efficacy of the mRNAtRNA interaction. The exceptional degree of sequence complementarity between these tRNA arms and stemloop structures IV and II, respectively, of the trp leader transcript strongly suggests the latter structures, to which until now no specific function had been assigned, as the counterparts in the interaction. The fact that mutations in tRNAAsn or the trp leader (with specifier Asn), aimed at improving the complementarity between the two sequences, did not result in enhanced antitermination shows that sequence complementarity per se is not sufficient to allow an efficient mRNAtRNA interaction. Most likely, a combination of sequence and structural context is important. These observations lead us to propose a model of tRNA-mediated transcription antitermination according to which the cognate tRNA would be recognized by the following identity elements of the leader transcript: the specifier codon, the nucleotide complementary to the tRNA discriminator and stemloop structures II and IV of the leader transcript.
The analysis of other leader transcripts from L. lactis and B. subtilis that are subject to tRNA-mediated antitermination showed the presence of short sequences that are complementary to the D- or T-arms of the respective cognate tRNAs in nearly all cases. Although none of these systems showed sequence complementarity as exhaustive as the lactococcal trp system, a general model of mRNAtRNA recognition may hold true since in the majority of cases the mRNAtRNA complementary sequences involve a number of nucleotides situated in loop structures or in other single-stranded regions of the respective RNAs. Scarabino et al. (1999) used an in vitro molecular selection approach to show that a tRNA preferentially binds RNA sequences complementary to its single-stranded anticodon, D- or T-loop sequence. The complementary sequences appeared to be present in a single-stranded loop or adjacent to a stemloop structure in the selected RNAs, while the same sequences in an RNA molecule without secondary structure did not bind to the tRNA. Earlier studies by Grosjean et al. (1976)
showed that complementary triplets in hairpin loops have a much higher affinity than free trinucleotides. In agreement with these results, long-range RNARNA interactions involved in E. coli ColEI replication control have been shown to be initiated by the interaction between complementary kissing loops of the folded RNA structures (Tomizawa, 1984
). The proposed interactions between the tRNA D- and T-loops and stemloop structures IV and II of the leader transcripts could fit a similar model in which very short complementary sequences placed in the correct structural context suffice for an interaction to take place.
Alternatively, one could conceive a model in which the nascent mRNA, after the establishment of a specifier codonanticodon interaction, is guided in its folding by a dynamic interaction with the tRNA. In the lactococcal trp leader, the fact that parts of the 18 nt sequence complementary to the tRNATrp T-arm are repeated several times upstream of this sequence (Fig. 1) may be indicative of such a process.
In conclusion, our results clearly show the importance of the D- and T-arms of tRNA in the establishment or maintenance of an efficient, specific mRNAtRNA interaction in the L. lactis trp leader. These tRNA sequences presumably interact with complementary sequences in the leader transcript, notably in stemloop structures II and IV. Although the short tRNA-complementary sequences in the conserved stemloop structures of other T-box leaders are less suggestive, these interactions may represent a general principle in tRNA-mediated transcription antitermination.
![]() |
ACKNOWLEDGEMENTS |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Birnboim, H. C. & Doly, J. (1979). A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res 7, 1513-1523.[Abstract]
Chopin, A., Biaudet, V. & Ehrlich, S. D. (1998). Analysis of the Bacillus subtilis genome sequence reveals nine new T-box leaders. Mol Microbiol 29, 662-664.[Medline]
Delorme, C., Ehrlich, S. D. & Renault, P. (1999). Regulation of expression of the Lactococcus lactis histidine operon. J Bacteriol 181, 2026-2037.
Dower, W. J., Miller, J. F. & Ragsdale, C. W. (1988). High efficiency transformation of E. coli cells by high voltage electroporation. Nucleic Acids Res 16, 6127-6145.[Abstract]
Frenkiel, H., Bardowski, J., Ehrlich, S. D. & Chopin, A. (1998). Transcription of the trp operon in Lactococcus lactis is controlled by antitermination in the leader region. Microbiology 144, 2103-2111.[Abstract]
Giegé, R., Sissler, M. & Florentz, C. (1998). Universal rules and idiosyncratic features in tRNA identity. Nucleic Acids Res 26, 5017-5035.
Grosjean, H., Söll, D. G. & Crothers, D. M. (1976). Studies of the complex between transfer RNAs with complementary anticodons. I. Origins of enhanced affinity between complementary triplets. J Mol Biol 103, 499-519.[Medline]
Grundy, F. J. & Henkin, T. M. (1993). tRNA as a positive regulator of transcription antitermination in B. subtilis. Cell 74, 475-482.[Medline]
Grundy, F. J. & Henkin, T. M. (1994). Conservation of a transcription antitermination mechanism in aminoacyl-tRNA synthetase and amino acid biosynthesis genes in Gram-positive bacteria. J Mol Biol 235, 798-804.[Medline]
Grundy, F. J., Rollins, S. M. & Henkin, T. M. (1994). Interaction between the acceptor end of tRNA and the T box stimulates antitermination in the Bacillus subtilis tyrS gene: a new role for the discriminator base. J Bacteriol 176, 4518-4526.[Abstract]
Henkin, T. M. (1994). tRNA-directed transcription antitermination. Mol Microbiol 13, 381-387.[Medline]
Holo, H. & Nes, I. F. (1995). Transformation of Lactococcus lactis by electroporation. Methods Mol Biol 47, 195-199.[Medline]
Le Loir, Y., Gruss, A., Ehrlich, S. D. & Langella, P. (1998). A nine-residue synthetic propeptide enhances secretion efficiency of heterologous proteins in Lactococcus lactis. J Bacteriol 180, 1895-1903.
Luo, D., Condon, C., Grunberg-Manago, M. & Putzer, H. (1998). In vitro and in vivo secondary structure probing of the thrS leader in Bacillus subtilis. Nucleic Acids Res 26, 5379-5387.
Miller, J. H. (1972). Experiments in Molecular Genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.
Putzer, H., Laalami, S., Brakhage, A. A., Condon, C. & Grunberg-Manago, M. (1995). Aminoacyl-tRNA synthetase gene regulation in Bacillus subtilis: induction, repression and growth-rate regulation. Mol Microbiol 16, 709-718.[Medline]
Raya, R., Bardowski, J., Andersen, P. S., Ehrlich, S. D. & Chopin, A. (1998). Multiple transcriptional control of the Lactococcus lactis trp operon. J Bacteriol 180, 3174-3180.[Abstract]
Rollins, S. M., Grundy, F. J. & Henkin, T. M. (1997). Analysis of cis-acting sequence and structural elements required for antitermination of the Bacillus subtilis tyrS gene. Mol Microbiol 25, 411-421.[Medline]
Scarabino, D., Crisari, A., Lorenzini, S., Williams, K. & Tocchini-Valentini, G. P. (1999). tRNA prefers to kiss. EMBO J 18, 4571-4578.
Simon, D. & Chopin, A. (1988). Construction of a vector plasmid family and its use for molecular cloning in Streptococcus lactis. Biochimie 70, 559-566.[Medline]
Terzaghi, B. E. & Sandine, W. E. (1975). Improved medium for lactic streptococci and their bacteriophages. Appl Microbiol 29, 807-813.
Tomizawa, J. (1984). Control of ColEI plasmid replication: the process of binding of RNA I to the primer transcript. Cell 38, 861-870.[Medline]
Van de Guchte, M., Ehrlich, S. D. & Chopin, A. (1998). tRNATrp as a key element of antitermination in the Lactococcus lactis trp operon. Mol Microbiol 29, 61-74.[Medline]
Van der Vossen, J. M. B. M., Van der Lelie, D. & Venema, G. (1987). Isolation and characterization of Streptococcus cremoris WG2-specific promoters. Appl Environ Microbiol 53, 2452-2457.[Medline]
Received 6 November 2000;
revised 3 January 2001;
accepted 16 January 2001.
HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
INT J SYST EVOL MICROBIOL | MICROBIOLOGY | J GEN VIROL |
J MED MICROBIOL | ALL SGM JOURNALS |