Control of replication of plasmid R1: structures and sequences of the antisense RNA, CopA, required for its binding to the target RNA, CopT.

Abstract

The replication frequency of plasmid R1 is determined by the availability of the RepA protein, which acts at the origin of replication to promote initiation. Synthesis of RepA is negatively regulated both at the transcriptional and post‐transcriptional levels. Post‐transcriptional control is exerted through the action of an antisense RNA, CopA RNA. The target of CopA RNA, CopT RNA, is located in the leader region of the RepA mRNA. Binding between CopA and CopT inhibits repA expression. We have previously presented an in vitro analysis of the binding reaction between CopA and CopT RNAs. In this communication, we extend the in vitro analysis by determining the regions of CopA required for binding, and also demonstrate that binding occurs in at least two steps. The first step is the formation of an initial, transient complex; stem‐loop II is the structure in CopA necessary and sufficient for this step. The subsequent step(s), resulting in the formation of a complete duplex, requires a stretch of single‐stranded nucleotides located 5′ to stem‐loop II in CopA, and its counterpart in CopT. We show that the single‐stranded region can be positioned on either side of stem‐loop II provided that there is a complementary stretch of nucleotides in CopT, indicating that the second step(s) is not sequence‐specific. Furthermore, the effects of salt concentration and temperature on the binding reaction indicate that duplex formation occurs through a mechanism of gradual intra‐strand breaking and inter‐strand formation of hydrogen bonds.

An unusual structure formed by antisense-target RNA binding involves an extended kissing complex with a four-way junction and a side-by-side helical alignment

“…We showed previously that full CopA-CopT duplexes are only slowly formed in vitro, and identified a stable complex using double-strand-specific enzyme RNase III-and Pb 2ϩ -catalyzed cleavages (Malmgren et al+, 1997)+ To obtain detailed structural information about the CopA-CopT complex, we used a range of enzymes and chemical reagents+ To study the extended kissing complex, CopA was replaced by the truncated CopA variant CopI, which contains only the major stem-loop structure (Persson et al+, 1990a) (Chen et al+, 1993) and DEPC (Weeks & Crothers, 1993) are known to be very sensitive to stacking of base rings+ Therefore, position N7 of purines within a helix is not reactive unless the deep groove is widened+ RNAs were also subjected to Pb 2ϩ -induced cleavage, which has proven exquisitely sensitive to subtle structural variations (Malmgren et al+, 1997 CopA-CopT complexes were formed at 37 8C for 2-15 min and subjected to enzymatic cleavages or chemical modifications+ We previously showed that within 15 min of incubation, at given concentrations, most RNAs were in the form of stable complexes, but only a minor fraction was converted to a full duplex (Malmgren et al+, 1997 Structure mapping of the complexes formed by CopICopT and CopA-CopT supported the presence of two intermolecular helices, B and B9, as modules in the four-helix junction structure (Fig+ 6)+ To delimitate the number of base pairs in the two intermolecular helices, we used site-directed mutagenesis of the copA/copT gene sequence+ Base-pair inversions were introduced at three positions within stem II+ These mutations affect neither the stability nor the structure of the RNAs in their free state (data not shown), but were expected to affect formation of helices B and B9 (Fig+ 6)+ Polymerase chain reaction (PCR)-generated templates were used for transcription of mutated CopI, CopA, and CopT Cleavages by RNase V1 was used to obtain a signature of the extended kissing complex: new or enhanced cleavages were induced in CopA or CopI at consecutive positions U52-C56, and in CopT at C104-A105+ Furthermore, in the CopA-CopT complex, the formation of the intermolecular helix C that greatly enhances complex stability was characterized by several V1 cuts in CopT at U162-C163 and U167-A168, as well as by the disappearance of V1 cleavages in CopA at positions G7-A9 and G16-A19, resulting from melting of helix I+ Therefore, we used RNase V1 to probe the structure of the homologous and heterologous complexes formed with the H1, H2, and H3 RNA variants (Fig+ 5)+ In all these experiments, we used conditions (RNA concentrations, time of incubation, buffers) under which the stable CopA-CopT complex is almost quantitatively formed+ RNase V1 probing of end-labeled CopA RNA variants indicated almost identical cleavage patterns of free mutant and wild-type RNAs+ Hence, no major structural changes were caused by the mutations+ The three homologous CopA/CopT and the heterologous CopA-H3/ CopT-wt complexes showed the characteristic RNase V1 patterns, that is, the appearance of cleavages at positions U52-C56 in CopA (Fig+ 5A)+ Interestingly, the heterologous complexes formed between CopT-wt and either CopA-H1 or CopA-H2 were different: significant protections were observed at U49-C50 of CopA-H1 and CopA-H2 (Fig+ 5A), whereas positions U52-C56 remained uncleaved, that is, a pattern resembling that of free RNAs+ The same experiments were also performed using end-labeled CopT variants (Fig+ 5B)+ Again, no major structural changes occurred in CopT mutant RNAs, and identical cleavage patterns were observed for all homologous CopA-CopT complexes, as well ...…”

“…The overall topology structure of the extended kissing complex is similar to some other RNA four-way junctions (Krol et al+, 1990;Walter et al+, 1998a;Nowakowski et al+, 1999) and DNA Holliday junctions (Duckett et al+, 1992)+ In both DNA and RNA four-way junctions, divalent ions are required for the formation and stabilization of antiparallel X-shaped structures (Duckett et al+, 1992;Walter et al+, 1998b)+ The particularity of the proposed CopA-CopT structure is the crossing over of the strands at the junction under the constraints imposed by the two loops connecting intermolecular helices B and B9+ This forces a side-by-side alignment of the two helical domains that brings the 59 tail of CopA in close proximity to the complementary region of CopT (Fig+ 7)+ The formation of intermolecular helix C, which clamps the two long helical domains, greatly enhances the stability of the complex (Persson et al+, 1990a;Malmgren et al+, 1997)+ Crystallographic analysis of a group I ribozyme domain revealed a similar organization (Cate et al+, 1996): a sharp bend induced by an internal loop allows a side-by-side alignment of two helical domains that is additionally stabilized by metal-and ribosemediated backbone contacts and two long-range tertiary interactions+ A side-by-side configuration was also proposed for the hairpin ribozyme, here stabilized by interactions between two internal loops (Earnshaw et al+, 1997)+ The formation of a stable RNA-RNA complex is not unique to CopA-CopT, and is also a key feature in the replication control of plasmids belonging to the IncB and IncIa groups (Siemering et al+, 1994; plasmids+ In these systems, the antisense RNAs inhibit the formation of a pseudoknot structure that activates rep translation (Wilson et al+, 1993;)+ All these antisense and target RNAs are characterized by stable hairpins with identical loop sequences and bulged residues in the upper stem regions+ Enzymatic probing performed on (antisense) RNAI in pMU720 plasmid bound to its target indicated that a full duplex was not rapidly formed in vitro+ Instead, binding resulted in an extended kissing complex stabilized by 59 tail interactions (Siemering et al+, 1994)+ One may therefore speculate that, in all these systems, the final product of the binding reaction in vitro is characterized by an overall topology very similar to that reported here, except that the lengths of helices B and B9, if formed in the IncB/IncIa cases, could be different+…”

“…CopA binds to the leader region of the repA mRNA (CopT), located about 80 nt upstream of the repA start codon (Fig+ 1)+ Binding prevents tap translation and thereby repA expression (Blomberg et al+, , 1994Malmgren et al+, 1996)+ The CopA-CopT binding process is viewed as a series of reactions leading to progressively more stable complexes (Persson et al+, 1988(Persson et al+, , 1990a(Persson et al+, , 1990bMalmgren et al+, 1997)+ CopA and CopT are fully complementary and both RNAs contain a major stem-loop structure (II/II9 in Fig+ 1) that is essential for high pairing rates and control (Öhman & Wagner, 1989;Hjalt & Wagner, 1992, 1995+ The initial step involves a transient looploop interaction (kissing complex) between the complementary hairpin loops (Persson et al+, 1990a(Persson et al+, , 1990b)+ Indeed, a truncated CopA (CopI, Fig+ 1), lacking the 59 proximal 30 nt and consisting only of the major stemloop, does not form stable duplexes with CopT, but is capable of competing with CopA for binding (Persson et al+, 1990b)+ It was recently shown that in both CopICopT and CopA-CopT complexes, initial kissing is rapidly followed by more extended intermolecular interactions (Malmgren et al+, 1997)+ Subsequently, the single-stranded region in the 59 tail of CopA pairs with its complement in CopT to yield the stable, inhibitory CopA-CopT complex+ This complex is the dominant product of binding in vitro (Malmgren et al+, 1996(Malmgren et al+, , 1997+ Complete duplex formation is very slow and has been proposed to be irrelevant for control (Malmgren et al+, 1996(Malmgren et al+, , 1997Wagner & Brantl, 1998)+ Different pairing pathways that result in rapid formation of stable antisense-target RNA complexes have been described (Kittle et al+, 1989;Persson et al+, 1990b;Tomizawa, 1990;Siemering et al+, 1994;Thisted et al+, 1994)+ A common feature is the use of a restricted single-stranded region in each interacting RNA for the initial step+ In most cases, binding initiates between two loops, in some cases between a loop and a singlestranded RNA segment+ Subsequently, more stable complexes are either formed by the pairing of distal RNA segments or, in the latter case, by extension of the first helix+…”

“…In the plasmids of the ColE1-family, control of replication is mediated by an antisense RNA, RNAI, that interacts with the preprimer, RNAII, via initial and transient base pairing between complementary loops + NMR studies were performed on two RNA hairpins carrying sevenmembered complementary loops derived from RNAI/ RNAII (Marino et al+, 1995;Lee & Crothers, 1998)+ These studies indicated that all seven loop bases were paired in the loop-loop helix, and continuous stacking of the loop nucleotides on the 39 side of their respective stems was observed+ Loop-loop interactions in plasmid R1 (IncFII plasmid; Persson et al+, 1990aPersson et al+, , 1990bMalmgren et al+, 1997), pMU720 (IncB plasmid; Siemering et al+, 1994), and ColIb-P9 (IncIa plasmid; Asano et al+, 1998) FIGURE 1. Sequences and secondary structures of the antisense RNA (CopA) and of the leader segment of the repA mRNA+ Structures are based on chemical and enzymatic probing (Öhman & Wagner, 1989; this work)+ The target sequence (CopT) of the antisense RNA (CopA) is shown in brackets+ The Shine and Dalgarno (SD), stop codon of tap, and start codons of tap and repA are indicated+ Binding of CopA prevents translation of the tap reading frame by occluding ribosome binding at tap initiation site (Ϫ)+ Under these conditions, the stable RNA stem-loop that sequesters the repA ribosome binding site (RBS) prevents translation of the repA reading frame (Blomberg et al+, 1994)+ If CopA fails to bind, ribosomes translate the tap reading frame, terminate at the tap stop codon, and reinitiate at the repA RBS by a direct translational coupling (Blomberg et al+, , 1994)+ The sequence of the truncated antisense RNA (CopI) sufficient to inhibit tap translation is boxed Malmgren et al+, 1996)+ are clearly similar+ In all these systems, it was suggested that an initial loop-loop interaction is rapidly converted to an extended kissing complex+ This requires partial melting of the upper stem regions, most probably facilitated by the presence of bulged residues (Siemering et al+, 1994;Hjalt & Wagner, 1995)+ Interestingly, the extended kissing complex suffices for inhibition in vivo Wilson et al+, 1993)+ In the case of plasmid R1, the extended kissing complex is also capable of blocking ribosome binding at the tap translation initiation site in vitro (Malmgren et al+, 1996), suggesting the existence of a bulky structure+…”

Trans-acting RNA inhibits tRNA suppressor activity in vivo

“…Secondary structures of tRNA leu amber suppressor (A), aptamer 1 (B), and aptamer 2 (C), represented as a tRNA-like molecule+ Under each aptamer the linear sequences corresponding to the anticodon loop of the suppressor tRNA are aligned with the corresponding sequences of the aptamer+ Complementary sequences are indicated in bold+ tRNA suppressor inhibition 905 complementary to the anticodon loop of the suppressor are flanked by sequences one would expect not to pair; the 5 nt at the 39 end are also complementary to the 59 half of the anticodon stem of the suppressor (Fig+ 1C)+ When the loops of aptamer 2 and of the suppressor interact, the 39 part of the aptamer's stem is free to base pair with the 59 part of the suppressor's stem; the aptamer-suppressor transitory coupling can lead to the formation of a more stable 12-bp interaction (Persson et al+, 1990;Zeiler & Simons, 1998)+ We cloned both aptamer 1 and 2 into pGFIB-I (Normanly et al+, 1986(Normanly et al+, , 1990)+ In this high-copy number plasmid that encodes an ampicillin-resistant gene, each aptamer is constitutively expressed from a strong synthetic promoter based on the constitutive promoter sequence of the E. coli lipoprotein gene lpp+ Distal to the promoter and restriction site polylinker is a transcription terminator, corresponding to the termination sequence of the ribosomal RNA operon rrnC+…”