Control of ColE1 plasmid replication by antisense RNA

One of the two major classes of regulatory strategies that control plasmid copy number involves recognition via base pairing between two plasmid-encoded complementary RNAs. The detailed analysis of this control circuitry has revealed some features of regulatory mechanisms based on RNA-RNA interaction that distinguish them from those based on protein-nucleic acid interaction. These features provide a framework with which to understand other regulatory mechanisms based on RNA-RNA interaction, and will aid in the design of efficient artificial antisense RNA systems.     

Control of ColE1 plasmid replication. Intermediates in the binding of RNA I and RNA II.

Replication of plasmid ColE1 is regulated by a plasmid-specified small RNA (RNA I). RNA I binds to the precursor (RNA II) of the primer for DNA synthesis and inhibits primer formation. The process of binding of RNA I to RNA II that results in formation of a stably bound complex consists of a series of reactions forming complexes differing in the stability. Formation of a very unstable early intermediate that was previously inferred from the inhibition of stable binding caused by a second RNA I species was firmly established by more extensive studies. This complex is converted to a more stable yet reversible complex that was identified by its RNase sensitivity, which was altered from that of the earlier complex or from that of free RNA I or RNA II. In these complexes, most loops of RNA II interact with their complementary loops of RNA I. The kinetic and structural analyses of the binding process predict formation of a complex interacting at a single pair of complementary loops that precedes formation of these complexes. Thus the process of binding of RNA I to RNA II is seen to consist of a sequence of reactions producing a series of progressively more stable intermediates leading to the final product.     

Control of ColE1 replication: low affinity specific binding of Rop (Rom) to RNAI and RNAII.

We have studied the interactions between the three molecules Rop, RNAI and RNAII that are involved in the regulatory mechanism controlling the replication of ColE1 plasmids. We show that it is possible to purify the two RNA molecules by passing an RNA mixture through an affinity column containing Rop immobilized to a solid support. The dissociation constants of the Rop-RNAI and Rop-RNAII complexes are of the order of 10(-4) M, several orders of magnitude higher than dissociation constants of stable protein-nucleic acid complexes (10(-10) M in the lambda repressor system). Although complete RNAI molecules have higher affinity, stem-and-loop I alone can also bind Rop, suggesting that this structure plays an important role in the interaction. Rop protects the stems of RNAI and RNAII from digestion by RNases while the sensitivity of the loops to digestion by RNase T1 is not affected by high concentrations of Rop. We propose a model for Rop-RNAI/RNAII interaction in which the dimeric protein acts as an adaptor between stem structures to position the two RNAs in the correct position for loop interaction.     

The Rep protein binding elements of the plasmid ColE2-P9 replication origin

The plasmid ColE2-P9 (ColE2) origin (32bp) is specifically recognized by the plasmid-specified Rep protein that initiates DNA replication. The ColE2 origin is divided into at least three functional subregions (I, II, and III), and three sites (a, b, and c) found in subregions I and II play important roles in Rep protein binding. We performed SELEX experiments of plasmid ColE2 to determine the optimal sequences for specific binding of the Rep protein. From these experiments, we obtained a common 16-bp sequence (5'-TGAGACCANATAAGCC-3'), which corresponds to about one half of the minimal ColE2 origin and contains sites a and b. Gel mobility shift assays using single-point mutant origins and the Rep protein further indicated that high affinity sequence-specific recognition by the Rep protein requires sites a, b, and c, but that mutations in site c were less disruptive to this recognition than those in sites a and b.     

Indole inhibition of ColE1 replication contributes to stable plasmid maintenance

In the absence of active partitioning, strict control of plasmid copy number is required to minimise the possibility of plasmid loss at bacterial cell division. An important cause of multicopy plasmid instability is the formation of plasmid dimers by recombination and their subsequent proliferation by over-replication in a process known as the dimer catastrophe. This leads to the formation of dimer-only cells in which plasmid copy number is substantially lower than in cells containing only monomers, and which have a greatly increased probability of plasmid loss at division. The accumulation of dimers triggers the synthesis of the regulatory small RNA, Rcd, which stimulates tryptophanase and increases the production of indole. This, in turn, inhibits Escherichia coli cell division. The Rcd checkpoint hypothesis proposes that delaying cell division allows time for the relatively slow conversion of plasmid dimers to monomers by Xer-cer site-specific recombination. In the present work we have re-evaluated this hypothesis and concluded that a cell division block is insufficient to prevent the dimer catastrophe. Plasmid replication must also be inhibited. In vivo experiments have shown that indole, when added exogenously to a broth culture of E. coli does indeed stop plasmid replication as well as cell division. We have also shown that indole inhibits the activity of DNA gyrase in vitro and propose that this is the mechanism by which plasmid replication is blocked. The simultaneous effects of upon growth, cell division and DNA replication in E. coli suggest that indole acts as a true cell cycle regulator.     

The heat-shock DnaK protein is required for plasmid R1 replication and it is dispensable for plasmid ColE1 replication.

Plasmid R1 replication in vitro is inactive in extracts prepared from a dnaK756 strain but is restored to normal levels upon addition of purified DnaK protein. Replication of R1 in extracts of a dnaKwt strain can be specifically inhibited with polyclonal antibodies against DnaK. RepA-dependent replication of R1 in dnaK756 extracts supplemented with DnaKwt protein at maximum concentration is partially inhibited by rifampicin and it is severely inhibited at sub-optimal concentrations of DnaK protein. The copy number of a run-away R1 vector is reduced in a dnaK756 background at 30 degrees C and at 42 degrees C the amplification of the run-away R1 vector is prevented. However a runaway R1 vector containing dnaK gene allows the amplification of the plasmid at high temperature. These data indicate that DnaK is required for both in vitro and in vivo replication of plasmid R1 and show a partial compensation for the low level of DnaK by RNA polymerase. In contrast ColE1 replication is not affected by DnaK as indicated by the fact that ColE1 replicates with the same efficiency in extracts from dnaKwt and dnaK756 strains.     

The effect of dnaA protein and n' sites on the replication of plasmid ColE1

The role of the dna A protein in the replication of plasmid ColE1 and its derivatives was examined. Wild-type and mutant ColE1 plasmids were compared as to their ability to replicate in an in vitro replication system supplemented with ammonium sulfate fractionated extracts from a dnaA-overproducing strain. Synthesis on plasmid templates containing the wild-type origin of replication was stimulated 1.3-fold by addition of the dnaA-overproducing extract. A larger effect was observed after deletion of the primosome assembly site, the n' site, on the leading strand. On the latter template, synthesis was only about one-half that observed with the wild-type templates, but synthesis could be restored to normal levels by addition of the dnaA-overproducing fractions. When the n' site on the lagging strand of pBR322 was deleted, synthesis in the in vitro replication system was reduced to less than 10% of levels seen with intact templates. dnaA-overproducing extract did not restore activity since the dnaA site was also deleted on these plasmids. To verify that the observed stimulation of wild-type and leading strand n' site mutants was due to the dnaA protein, dnaA protein was purified to greater than 50% homogeneity, and antiserum was prepared. The purified protein stimulated synthesis on the plasmid templates to the same extent as the overproducing extracts, and dnaA antiserum blocked stimulation both by extracts and by the purified protein. Thus, dnaA protein, and, by inference, the dnaA recognition site at the ColE1 origin of replication seem to be important for ColE1 replication. The effect of dnaA protein is enhanced when the n'site is defective, suggesting that the dnaA protein plays a role similar to that of the proteins i, n, n', and n' in directing primosome assembly, as proposed by Seufert, W., and Messer, W. ((1987) Cell 48, 73-78).