Regulatory RNAs as mediators of virulence gene expression in bacteria.

Bacteria exploit functional diversity of RNAs in a wide range of regulatory mechanisms to control gene expression. In last few years, small RNA molecules have been discovered at a staggering rate in bacteria, mainly in Escherichia coli. While functions of many of these RNA molecules are still not known, several of them behave as key effectors of adaptive responses, such as environmental cue recognition, stress response, and virulence control. Most fascinating, perhaps, is the discovery that mRNAs behave as direct sensors of small molecules or of environmental cues. The astonishing diversity of RNA-dependent regulatory mechanisms is linked to the dynamic properties and versatility of the RNA structure. In this review, we relate several recent studies in different bacterial pathogens that illustrate the diverse roles of RNA to control virulence gene expression.

Crystal structures of coaxially stacked kissing complexes of the HIV-1 RNA dimerization initiation site.

We describe the crystal structures of the RNA dimerization initiation sites (DIS) of HIV-1 subtypes A and B. Both molecules adopt a hairpin conformation, with loop sequences consisting of 272-AGGUGCACA-280 and 272-AAGCGCGCA-280, respectively. This includes a six-base self-complementary stretch (underlined) that allows homodimerization through the formation of a loop-loop, or 'kissing-loop', complex. The DISs for the two sequences have identical conformations, and the two interacting hairpins show a perfect coaxial alignment. The conserved purines, A272 and R273, are stacked in a bulged-out conformation and symmetrically join the upward and downward strands of each hairpin by crossing the helix major groove. There is good agreement between these structures and previous results from chemical probing in solution, as well as with differences in magnesium dependence for dimerization. The overall shape of the kissing-loop complex is very similar to that of the previously determined subtype A DIS duplex form. Unexpectedly, the purine R273 is the only base seen at a different position and is responsible for the difference in topology between the two forms. We propose that the transition from kissing-loop duplex could occur by a recombination mechanism based on symmetrical chain cleavage at R273 of each hairpin and subsequent cross-religation, rather than by base-pair melting and subsequent reannealing.

Isolation, crystallization, and investigation of ribosomal protein S8 complexed with specific fragments of rRNA of bacterial or archaeal origin.

The core ribosomal protein S8 binds to the central domain of 16S rRNA independently of other ribosomal proteins and is required for assembling the 30S subunit. It has been shown with E. coli ribosomes that a short rRNA fragment restricted by nucleotides 588-602 and 636-651 is sufficient for strong and specific protein S8 binding. In this work, we studied the complexes formed by ribosomal protein S8 from Thermus thermophilus and Methanococcus jannaschii with short rRNA fragments isolated from the same organisms. The dissociation constants of the complexes of protein S8 with rRNA fragments were determined. Based on the results of binding experiments, rRNA fragments of different length were designed and synthesized in preparative amounts in vitro using T7 RNA-polymerase. Stable S8-RNA complexes were crystallized. Crystals were obtained both for homologous bacterial and archaeal complexes and for hybrid complexes of archaeal protein with bacterial rRNA. Crystals of the complex of protein S8 from M. jannaschii with the 37-nucleotide rRNA fragment from the same organism suitable for X-ray analysis were obtained.

Dimerization of the 3'UTR of bicoid mRNA involves a two-step mechanism.

The proper localization of bicoid (bcd) mRNA requires cis-acting signals within its 3' untranslated region (UTR) and trans-acting factors such as Staufen. Dimerization of bcd mRNA through intermolecular base-pairing between two complementary loops of domain III of the 3'UTR was proposed to be important for particle formation in the embryo. The participation in the dimerization process of each domain building the 3'UTR was evaluated by thermodynamic and kinetic analysis of various mutated and truncated RNAs. Although sequence complementarity between the two loops of domain III is required for initiating mRNA dimerization, the initial reversible loop-loop complex is converted rapidly into an almost irreversible complex. This conversion involves parts of RNA outside of domain III that promote initial recognition, and dimerization can be inhibited by sense or antisense oligonucleotides only before conversion has proceeded. Injection of the different bcd RNA variants into living Drosophila embryos shows that all elements that inhibit RNA dimerization in vitro prevent formation of localized particles containing Staufen. Particle formation appeared to be dependent on both mRNA dimerization and other element(s) in domains IV and V. Domain III of bcd mRNA could be substituted by heterologous dimerization motifs of different geometry. The resulting dimers were converted into stable forms, independently of the dimerization module used. Moreover, these chimeric RNAs were competent in forming localized particles and recruiting Staufen. The finding that the dimerization domain of bcd mRNA is interchangeable suggests that dimerization by itself, and not the precise geometry of the intermolecular interactions, is essential for the localization process. This suggests that the stabilizing interactions that are formed during the second step of the dimerization process might represent crucial elements for Staufen recognition and localization.

The emergence of different resistance mechanisms toward nucleoside inhibitors is explained by the properties of the wild type HIV-1 reverse transcriptase.

Nucleoside reverse transcriptase inhibitors (NRTIs) represent one of the main drug families used against AIDS. Once incorporated in DNA, they act as chain terminators, due to the lack of a 3'-hydroxyl group. As for the other anti-human immunodeficiency virus type 1 drugs, their efficiency is limited by the emergence of resistant viral strains. Unexpectedly, previous studies indicated that resistance toward NRTIs is achieved via two distinct and generally exclusive mechanisms. Resistance mutations either decrease the efficiency of NRTIs incorporation or increase their excision from the extended primer. To understand the emergence of different resistance mechanisms toward a single inhibitor class, we compared the incorporation and the pyrophosphorolysis of several NRTIs using wild type reverse transcriptase (WT RT). We found that the efficiency of discrimination or excision by pyrophosphorolysis in the presence of nucleotides of a given NRTI is a key determinant in the emergence of one or the other resistance pathway. Indeed, our results suggest that the pathway by which RT become resistant toward a given NRTI can be predicted by studying the inhibition of WT RT, because the resistance mutations do not confer new properties to the mutant enzyme, but rather exacerbate pre-existing properties of the WT enzyme. They also help to understand the low cross-resistance toward d4T observed with the 3'-azido-3'-deoxythymidine (AZT or zidovudine)-resistant RT.

The fragile X mental retardation protein binds specifically to its mRNA via a purine quartet motif.

Fragile X syndrome is caused by the absence of protein FMRP, the function of which is still poorly understood. Previous studies have suggested that FMRP may be involved in various aspects of mRNA metabolism, including transport, stability and/or translatability. FMRP was shown to interact with a subset of brain mRNAs as well as with its own mRNA; however, no specific RNA-binding site could be identified precisely. Here, we report the identification and characterization of a specific and high affinity binding site for FMRP in the RGG-coding region of its own mRNA. This site contains a purine quartet motif that is essential for FMRP binding and can be substituted by a heterologous quartet-forming motif. The specific binding of FMRP to its target site was confirmed further in a reticulocyte lysate through its ability to repress translation of a reporter gene harboring the RNA target site in the 5'-untranslated region. Our data address interesting questions concerning the role of FMRP in the post-transcriptional control of its own gene and possibly other target genes.

Selection of genomic target RNAs by iterative screening.

A growing number of proteins are known to exert their regulatory or biological functions via RNA binding. In some cases genetic interactions allow us to infer candidate targets for RNA directed regulation, but in many other cases identification of potential regulatory targets is problematic. We have developed an in vitro biochemical screen, SETIS (SElection of <> Target RNAs by Iterative Screening) that allows screening of a major portion of the genome for identification of potential targets for RNA binding proteins.

Detailed analysis of RNA-protein interactions within the ribosomal protein S8-rRNA complex from the archaeon Methanococcus jannaschii.

The crystal structure of ribosomal protein S8 bound to its target 16 S rRNA from a hyperthermophilic archaeon Methanococcus jannaschii has been determined at 2.6 A resolution. The protein interacts with the minor groove of helix H21 at two sites located one helical turn apart, with S8 forming a bridge over the RNA major groove. The specificity of binding is essentially provided by the C-terminal domain of S8 and the highly conserved nucleotide core, characterized by two dinucleotide platforms, facing each other. The first platform (A595-A596), which is the less phylogenetically and structurally constrained, does not directly contact the protein but has an important shaping role in inducing cross-strand stacking interactions. The second platform (U641-A642) is specifically recognized by the protein. The universally conserved A642 plays a pivotal role by ensuring the cohesion of the complex organization of the core through an array of hydrogen bonds, including the G597-C643-U641 base triple. In addition, A642 provides the unique base-specific interaction with the conserved Ser105, while the Thr106 - Thr107 peptide link is stacked on its purine ring. Noteworthy, the specific recognition of this tripeptide (Thr-Ser-Thr/Ser) is parallel to the recognition of an RNA tetraloop by a dinucleotide platform in the P4-P6 ribozyme domain of group I intron. This suggests a general dual role of dinucleotide platforms in recognition of RNA or peptide motifs. One prominent feature is that conserved side-chain amino acids, as well as conserved bases, are essentially involved in maintaining tertiary folds. The specificity of binding is mainly driven by shape complementarity, which is increased by the hydrophobic part of side-chains. The remarkable similarity of this complex with its homologue in the T. thermophilus 30 S subunit indicates a conserved interaction mode between Archaea and Bacteria.

Structure and dimerization of HIV-1 kissing loop aptamers.

Dimerization of two homologous strands of genomic RNA is an essential feature of the retroviral replication cycle. In HIV-1, genomic RNA dimerization is facilitated by a conserved stem-loop structure located near the 5' end of the viral RNA called the dimerization initiation site (DIS). The DIS loop is comprised of nine nucleotides, six of which define an autocomplementary sequence flanked by three conserved purine residues. Base- pairing between the loop sequences of two copies of genomic RNA is necessary for efficient dimerization. We previously used in vitro evolution to investigate a possible structural basis for the marked sequence conservation of the DIS loop. In this study, chemical structure probing, measurements of the apparent dissociation constants, and computer structure analysis of dimerization-competent aptamers were used to analyze the dimers' structure and binding. The selected aptamers were variants of the naturally occurring A and B subtypes. The data suggest that a sheared base-pair closing the loop of the DIS is important for dimerization in both subtypes. On the other hand, the open or closed state of the last base-pair in the stem differed in the two subtypes. This base-pair appeared closed in the subtype A DIS dimer and open in subtype B. Finally, evidence for a cross-talk between nucleotides 2, 5, and 6 was found in some, but not all, loop contexts, indicating some structural plasticity depending on loop sequence. Discriminating between the general rules governing dimer formation and the particular characteristics of individual DIS aptamers helps to explain the affinity and specificity of loop-loop interactions and could provide the basis for development of drugs targeted against the dimerization step during retroviral replication.

Bulged residues promote the progression of a loop-loop interaction to a stable and inhibitory antisense-target RNA complex.

In several groups of bacterial plasmids, antisense RNAs regulate copy number through inhibition of replication initiator protein synthesis. These RNAs are characterized by a long hairpin structure interrupted by several unpaired residues or bulged loops. In plasmid R1, the inhibitory complex between the antisense RNA (CopA) and its target mRNA (CopT) is characterized by a four-way junction structure and a side-by-side helical alignment. This topology facilitates the formation of a stabilizer intermolecular helix between distal regions of both RNAs, essential for in vivo control. The bulged residues in CopA/CopT were shown to be required for high in vitro binding rate and in vivo activity. This study addresses the question of why removal of bulged nucleotides blocks stable complex formation. Structure mapping, modification interference, and molecular modeling of bulged-less mutant CopA-CopT complexes suggests that, subsequent to loop-loop contact, helix propagation is prevented. Instead, a fully base paired loop-loop interaction is formed, inducing a continuous stacking of three helices. Consequently, the stabilizer helix cannot be formed, and stable complex formation is blocked. In contrast to the four-way junction topology, the loop-loop interaction alone failed to prevent ribosome binding at its loading site and, thus, inhibition of RepA translation was alleviated.

Four-way junctions in antisense RNA-mRNA complexes involved in plasmid replication control: a common theme?

In several groups of bacterial plasmids, antisense RNAs regulate copy number through inhibition of replication initiator protein synthesis. In plasmid R1, we have recently shown that the inhibitory complex between the antisense RNA (CopA) and its target mRNA (CopT) is characterized by the formation of two intermolecular helices, resulting in a four-way junction structure and a side-by-side helical alignment. Based on lead-induced cleavage and ribonuclease (RNase) V(1) probing combined with molecular modeling, a strikingly similar topology is supported for the complex formed between the antisense RNA (Inc) and mRNA (RepZ) of plasmid Col1b-P9. In particular, the position of the four-way junction and the location of divalent ion-binding site(s) indicate that the structural features of these two complexes are essentially the same in spite of sequence differences. Comparisons of several target and antisense RNAs in other plasmids further indicate that similar binding pathways are used to form the inhibitory antisense-target RNA complexes. Thus, in all these systems, the structural features of both antisense and target RNAs determine the topologically possible and kinetically favored pathway that is essential for efficient in vivo control.

Role of conserved nucleotides in building the 16 S rRNA binding site for ribosomal protein S15.

Ribosomal protein S15 recognizes a highly conserved target on 16 S rRNA, which consists of two distinct binding regions. Here, we used extensive site-directed mutagenesis on a Escherichia coli 16 S rRNA fragment containing the S15 binding site, to investigate the role of conserved nucleotides in protein recognition and to evaluate the relative contribution of the two sites. The effect of mutations on S15 recognition was studied by measuring the relative binding affinity, RNA probing and footprinting. The crystallographic structure of the Thermus thermophilus complex allowed molecular modelling of the E. coli complex and facilitated interpretation of biochemical data. Binding is essentially driven by site 1, which includes a three-way junction constrained by a conserved base triple and cross-strand stacking. Recognition is based mainly on shape complementarity, and the role of conserved nucleotides is to maintain a unique backbone geometry. The wild-type base triple is absolutely required for protein interaction, while changes in the conserved surrounding nucleotides are partially tolerated. Site 2, which provides functional groups in a conserved G-U/G-C motif, contributes only modestly to the stability of the interaction. Binding to this motif is dependent on binding at site 1 and is allowed only if the two sites are in the correct relative orientation. Non-conserved bulged nucleotides as well as a conserved purine interior loop, although not directly involved in recognition, are used to provide an appropriate flexibility between the two sites. In addition, correct binding at the two sites triggers conformational adjustments in the purine interior loop and in a distal region, which are known to be involved for subsequent binding of proteins S6 and S18. Thus, the role of site 1 is to anchor S15 to the rRNA, while binding at site 2 is aimed to induce a cascade of events required for subunit assembly.

Identification of the in vitro HIV-2/SIV RNA dimerization site reveals striking differences with HIV-1.

Although their genomes cannot be aligned at the nucleotide level, the HIV-1/SIVcpz and the HIV-2/SIVsm viruses are closely related lentiviruses that contain homologous functional and structural RNA elements in their 5'-untranslated regions. In both groups, the domains containing the trans-activating region, the 5'-copy of the polyadenylation signal, and the primer binding site (PBS) are followed by a short stem-loop (SL1) containing a six-nucleotide self-complementary sequence in the loop, flanked by unpaired purines. In HIV-1, SL1 is involved in the dimerization of the viral RNA, in vitro and in vivo. Here, we tested whether SL1 has the same function in HIV-2 and SIVsm RNA. Surprisingly, we found that SL1 is neither required nor involved in the dimerization of HIV-2 and SIV RNA. We identified the NarI sequence located in the PBS as the main site of HIV-2 RNA dimerization. cis and trans complementation of point mutations indicated that this self-complementary sequence forms symmetrical intermolecular interactions in the RNA dimer and suggested that HIV-2 and SIV RNA dimerization proceeds through a kissing loop mechanism, as previously shown for HIV-1. Furthermore, annealing of tRNA(3)(Lys) to the PBS strongly inhibited in vitro RNA dimerization, indicating that, in vivo, the intermolecular interaction involving the NarI sequence must be dissociated to allow annealing of the primer tRNA.

In vivo selection of functional variations in essential sites of ribosomal RNA.

The technique of "in vivo selection of functional ribosomes" is a genetic approach to dissecting the link between the structure and function of critical sites of rRNA. This method proceeds through selection of functional variants among cells that express ribosomes from a pool of rRNA-containing randomized sites. The selection of bacterial clones with functional ribosomes is based on the use of a plasmid carrying a rRNA operon in which a site of interest has been randomized and a point mutation conferring an antibiotic resistance has been introduced. Cells expressing functional ribosomes are then selected on medium containing the antibiotic. With this approach one can isolate at once all the possible variations at a given rRNA site that are able to sustain normal ribosome function. The identification of covariations in between several nucleotides that maintain wild-type ribosome activity can thus help demonstrate the function of specific interactions in rRNA.

The crystal structure of UUCG tetraloop.

All large structured RNAs contain hairpin motifs made of a stem closed by several looped nucleotides. The most frequent loop motif is the UUCG one. This motif belongs to the tetraloop family and has the peculiarity of being highly thermodynamically stable. Here, we report the first crystal structure of two UUCG tetraloops embedded in a larger RNA-protein complex solved at 2.8 A resolution. The two loops present in the asymmetric unit are in a different crystal packing environment but, nevertheless, have an identical conformation. The observed structure is globally close to that obtained in solution by nuclear magnetic resonance. However, subtle differences point to a more detailed picture of the role played by 2'-hydroxyl groups in stabilising this tetraloop.

The crystal structure of HIV reverse-transcription primer tRNA(Lys,3) shows a canonical anticodon loop.

We have solved to 3.3 A resolution the crystal structure of the HIV reverse-transcription primer tRNA(Lys,3). The overall structure is exactly comparable to the well-known L-shape structure first revealed by yeast tRNA(Phe). In particular, it unambiguously shows a canonical anticodon loop. This contradicts previous results in short RNA fragment studies and leads us to conclude that neither frameshifting specificities of tRNA(Lys) nor tRNA(Lys,3) primer selection by HIV are due to a specific three-dimensional anticodon structure. Comparison of our structure with the results of an NMR study on a hairpin representing a nonmodified anticodon stem-loop makes plausible the conclusion that chemical modifications of the wobble base U34 to 5-methoxycarbonyl-methyl-2-thiouridine and of A37 to 2-methylthio-N-6-threonylcarbamoyl-adenosine would be responsible for a canonical 7-nt anticodon-loop structure, whereas the unmodified form would result in a noncanonical UUU short triloop. The hexagonal crystal packing is remarkable and shows tight dimers of tRNAs forming a right-handed double superhelix. Within the dimers, the tRNAs are associated head-to-tail such that the CCA end of one tRNA interacts with the anticodon of the symmetry-related tRNA. This provides us with a partial view of a codon-anticodon interaction and gives insights into the positioning of residue 37, and of its posttranscriptional modifications, relative to the first base of the codon.

NMR and biochemical characterization of recombinant human tRNA(Lys)3 expressed in Escherichia coli: identification of posttranscriptional nucleotide modifications required for efficient initiation of HIV-1 reverse transcription.

Reverse transcription of HIV-1 viral RNA uses human tRNA(Lys)3 as a primer. Some of the modified nucleotides carried by this tRNA must play a key role in the initiation of this process, because unmodified tRNA produced in vitro is only marginally active as primer. To provide a better understanding of the contribution of base modifications in the initiation complex, we have designed a recombinant system that allows tRNA(Lys)3 expression in Escherichia coli. Because of their high level of overexpression, some modifications are incorporated at substoichiometric levels. We have purified the two major recombinant tRNA(Lys)3 subspecies, and their modified nucleotide contents have been characterized by a combination of NMR and biochemical techniques. Both species carry psis, Ds, T, t6A, and m7G. Differences are observed at position 34, within the anticodon. One fraction lacks the 5-methylaminomethyl group, whereas the other lacks the 2-thio group. Although the s2U34-containing recombinant tRNA is a less efficient primer, it presents most of the characteristics of the mammalian tRNA. On the other hand, the mnm5U34-containing tRNA has a strongly reduced activity. Our results demonstrate that the modifications that are absent in E. coli (m2G10, psi27, m5C48, m5C49, and m1A58) as well as the mnm5 group at position 34 are dispensable for initiation of reverse transcription. In contrast, the 2-thio group at position 34 seems to play an important part in this process.

Progression of a loop-loop complex to a four-way junction is crucial for the activity of a regulatory antisense RNA.

The antisense RNA, CopA, regulates the replication frequency of plasmid R1 through inhibition of RepA translation by rapid and specific binding to its target RNA (CopT). The stable CopA-CopT complex is characterized by a four-way junction structure and a side-by-side alignment of two long intramolecular helices. The significance of this structure for binding in vitro and control in vivo was tested by mutations in both CopA and CopT. High rates of stable complex formation in vitro and efficient inhibition in vivo required initial loop-loop complexes to be rapidly converted to extended interactions. These interactions involve asymmetric helix progression and melting of the upper stems of both RNAs to promote the formation of two intermolecular helices. Data presented here delineate the boundaries of these helices and emphasize the need for unimpeded helix propagation. This process is directional, i.e. one of the two intermolecular helices (B) must form first to allow formation of the other (B'). A binding pathway, characterized by a hierarchy of intermediates leading to an irreversible and inhibitory RNA-RNA complex, is proposed.

Convergence of natural and artificial evolution on an RNA loop-loop interaction: the HIV-1 dimerization initiation site.

Loop-loop interactions among nucleic acids constitute an important form of molecular recognition in a variety of biological systems. In HIV-1, genomic dimerization involves an intermolecular RNA loop-loop interaction at the dimerization initiation site (DIS), a hairpin located in the 5' noncoding region that contains an autocomplementary sequence in the loop. Only two major DIS loop sequence variants are observed among natural viral isolates. To investigate sequence and structural constraints on genomic RNA dimerization as well as loop-loop interactions in general, we randomized several or all of the nucleotides in the DIS loop and selected in vitro for dimerization-competent sequences. Surprisingly, increasing interloop complementarity above a threshold of 6 bp did not enhance dimerization, although the combinations of nucleotides forming the theoretically most stable hexanucleotide duplexes were selected. Noncanonical interactions contributed significantly to the stability and/or specificity of the dimeric complexes as demonstrated by the overwhelming bias for noncanonical base pairs closing the loop and covariations between flanking and central loop nucleotides. Degeneration of the entire loop yielded a complex population of dimerization-competent sequences whose consensus sequence resembles that of wild-type HIV-1. We conclude from these findings that the DIS has evolved to satisfy simultaneous constraints for optimal dimerization affinity and the capacity for homodimerization. Furthermore, the most constrained features of the DIS identified by our experiments could be the basis for the rational design of DIS-targeted antiviral compounds.

Inhibition of the initiation of HIV-1 reverse transcription by 3'-azido-3'-deoxythymidine. Comparison with elongation.

Initiation of human immunodeficiency virus-1 reverse transcription requires formation of a complex containing the viral RNA, primer tRNA(3)(Lys), and reverse transcriptase. Initiation, corresponding to addition of the first six nucleotides to tRNA(3)(Lys), is distinguished from elongation by its high specificity and low efficiency (processivity). Here, we compared the inhibition of initiation and elongation of reverse transcription by 3'-azido-3'-deoxythymidine 5'-triphosphate (AZTTP), the active form of 3'-azido-3'-deoxythymidine. We report the first detailed study of nucleotide binding, discrimination, and pyrophosphorolysis by the authentic initiation complex. We showed that the initiation and elongation complexes bound AZTTP and dTTP with the same affinity, while the polymerization rates were reduced by 148-160-fold during initiation. The pyrophosphorolysis rate of dTTP was reduced by the same extent, indicating that the polymerization equilibrium is the same in the two phases. The efficient unblocking of the 3'-azido-3'-deoxythymidine 5'-monophosphate (AZTMP)-terminated primer by pyrophosphorolysis significantly relieved inhibition of DNA synthesis during elongation in the presence of physiological pyrophosphate concentrations. Remarkably, although pyrophosphorolysis of dTMP and AZTMP were equally efficient during elongation, reverse transcriptase was almost totally unable to unblock the AZTMP-terminated primer during initiation. As a result, inhibition of reverse transcription by AZTTP was more efficient during initiation than elongation of reverse transcription, despite a reduced selectivity of incorporation.