A facile and reversible method to decrease the copy number of the ColE1-related cloning vectors commonly used in Escherichia coli.

We report a technique which uses the cointegrate intermediate of transposon Tn1000 transposition as a means to lower the copy number of ColE1-type plasmids. The transposition of Tn1000 from one replicon to another is considered a two-step process. In the first step, the transposon-encoded TnpA protein mediates fusion of the two replicons to produce a cointegrate. In the second step, the cointegrate is resolved by site-specific recombination between the two transposon copies to yield the final transposition products: the target replicon with an integrated transposon plus the regenerated donor replicon. Using in vitro techniques, the DNA sequence of the Tn1000 transposon was altered so that cointegrate formation occurs but resolution by the site-specific recombination pathway is blocked. When this transposon was resident on an F factor-derived plasmid, a cointegrate was formed between a multicopy ColE1-type target plasmid and the conjugative F plasmid. Conjugational transfer of this cointegrate into a polA strain resulted in a stable cointegrate in which replication from the ColE1 plasmid origin was inhibited and replication proceeded only from the single-copy F factor replication origin. We assayed isogenic strains which harbored plasmids encoding chloramphenicol acetyltransferase to measure the copy number of such F factor-ColE1-type cointegrate plasmids and found that the copy number was decreased to the level of single-copy chromosomal elements. This method was used to study the effect of copy number on the expression of the fabA gene (which encodes the key fatty acid-biosynthetic enzyme beta-hydroxydecanoylthioester dehydrase) by the regulatory protein encoded by the fadR gene.     

Effect of Growth Conditions on the Formation of the Relaxation Complex of Supercoiled ColE1 Deoxyribonucleic Acid and Protein in Escherichia coli

Colicinogenic factor E1 (ColE1) is present in Escherichia coli strain JC411 (ColE1) cells to the extent of about 24 copies per cell. This number does not appear to vary in situations which give rise to twofold differences in the amount of chromosomal deoxyribonucleic acid (DNA) present per cell. If cells are grown in the absence of glucose, approximately 80% of the ColE1 molecules can be isolated as strand-specific DNA-protein relaxation complexes. When glucose is present in the medium, only about 30% of the plasmid molecules can be isolated as relaxation complexes. Medium shift experiments in which glucose was removed from the medium indicate that within 15 min after the shift the majority (>60%) of the plasmid can be isolated as relaxation complex. This rapid shift to the complexed state is accompanied by a two- to threefold increase in the rate of plasmid replication. The burst of replication and the shift to the complexed state are both inhibited by the presence of chloramphenicol. Inhibition of protein synthesis in log cultures by the addition of chloramphenicol or amino acid starvation allows ColE1 DNA to continue replicating long after chromosomal replication has ceased. Under these conditions, noncomplexed plasmid DNA accumulates while the amount of DNA that can be isolated in the complexed state remains constant at the level that existed prior to treatment. In the presence of chloramphenicol, there appears to be a random dissociation and association of ColE1 DNA and “relaxation protein” during or between rounds of replication.     

pUb6060: a broad-host-range, DNA polymerase-I-independent ColE2-like plasmid

A ColE2-like, cryptic plasmid, pUB6060, of 5.8 kb has been found in a clinical isolate of Plesiomonas shigelloides. The complete sequence of pUB6060 has been determined and reveals a number of interesting features about the plasmid. The ColE2-like replication locus is linked to a functional ColE1-like mobilization locus. Replication is, unusually for ColE2 replicons, DNA polymerase-I-independent and may involve two, rather than the usual one, plasmid-encoded functions. Additionally, it carries two ORFs encoding products of unknown function. The pUB6060 replicon maintains a moderate plasmid copy number (10 per chromosome copy) and permits replication in diverse gram-negative bacteria.     

Nature of the compatible inheritance of hybrid plasmid pAS8 and its deletion mutants with replicon ColE1]

Hybrid plasmid pAS8, that consists of RP4 and ColE1 replicons, is incompatible with RP4 but co-exists with ColE1 replicon. The deletion mutants of pAS8, that replicates under the control of ColE1 replicon only, are incompatible with ColE1 derivatives, although the copy number of pAS8 and its deletion mutants in the cell is the same (1-3 per the chromosome). Incompatibility effect of plasmids is expressed in a sharp decrease in transformant's yield during selection of incoming and resident plasmids markers. In this case incompatibility is a very fast process, that leads to the elimination of resident or incoming plasmid just before plating on the selective medium. On the base of negative control theory, the repressors yield, synthesized in the presence of 1-3 copies of the deletion mutant is enough for the expression of ColE1-specific incompatibility. This ColE1 incompatibility is probably connected with the functional activity of ColE1 replicon. When ColE1 factor replicates under the control of RP4 replicon the expression of ColE1-specific incompatibility does not occur. Possibly, the incompatibility effect takes place when pAS8 deletion mutants cause the synthesis of ColE1-specific repressor. Also, the replicons of ColE1 may competein the membrane attachment site.     

Analysis of mini-F plasmid replication by transposition mutagenesis

Derivatives of a mini-F plasmid in which Tn3 is inserted in F deoxyribonucleic acid were obtained, and the sites of insertion for 40 of the derivatives were mapped. Tn3 was found to insert at many sites within mini-F, but most insertions were within the 43.0- to 43.7-kilobase (kb), 44.2- to 44.7-kb, and 45.9- to 46.3-kb segments. Hence, these segments are unnecessary for mini-F replication. Most of the Tn3 derivatives were similar to their parent miniplasmid with respect to copy number, stability, and incompatibility. Insertions at 45.15 kb and near 46.0 kb caused a moderate disruption of copy number control, and insertion at 47.6 kb resulted in unstable maintenance. Deletion derivatives lacking deoxyribonucleic acid between 40.3 and 44.76 kb and between 45.92 and 49.4 kb were obtained. This observation suggests either that mini-F contains a third origin, in addition to those already reported to be at 42.6 and 44.4 kb, or that the reported position of the secondary origin, 44.4 kb, is incorrect and that this origin is between 44.76 and 45.92 kb.     

Nonintegrated plasmid-folded chromosome complexes: genetic studies on formation and possible relationship to plasmid replication.

When folded chromosomes are purified from plasmid-containing bacteria, a reproducible fraction of the host's covalently closed, circular (CCC) plasmid DNA copurifies with the chromosomes. From this copurification, we infer the existence of nonintegrative plasmid-chromosome (NPC) complexes. Previously, we noted that plasmids dependent on DNA polymerase III and with stringent control of replication complex to a greater extent than plasmids dependent on DNA polymerase I and with relaxed control of replication. We have examined this subject in more depth and find that: (i) The composite plasmids formed by in vitro recombination of a “stringent” with a “relaxed” replicon complex to chromosomes at the frequency of the component replicon which directs replication; (ii) all of the detectable replicative intermediates, but only 25% of the CCC forms, of plasmid ColE1 complex to chromosome; and (iii) when a mini-F plasmid is deleted for the DNA sequences which include the primary origin of replication, the complexing frequency decreases 30 to 40%. We conclude from these findings that NPC complexes either indirectly or directly relate to plasmid replication. Further, we find that the EcoRI kan⁺ fragments of pML31 and the ampicillin resistance transposon, Tn3, promote complexing of both ColE1 and mini-F plasmids to host chromosomes. The biological significance of this latter complexing is unknown. However, we conclude from these studies and from point (iii) that complexing is determined in part by unique plasmid sequences.     

Selection and characterization of ColE1 plasmid mutants that exhibit altered stability and replication.

This report describes a method for isolating mutants of plasmid ColE1 that exhibit unstable maintenance and altered replication characteristics. It also describes the initial characterization of four mutants isolated by that method. A chimeric plasmid, pHSG124, containing a ColE1 derivative and a temperature-sensitive replication derivative of pSC101 was mutagenized in vitro, using hydroxylamine. By adjusting the growth conditions of transformants containing the mutagenized chimeric deoxyribonucleic acid, it was possible to rapidly screen colonies and identify those that had a high probability of carrying ColE1 mutants that exhibit unstable maintenance. Of those mutants, some exhibited altered copy number or accumulated catenated structures. Evidence is presented which suggests that the mutations in three of the mutants are probably located in the HaeII A fragment of ColE1.     

Incompatibility and transforming efficiency of ColE1 and related plasmids

Summary Replicons derived from the ColE1 plasmid are incompatible with one another, but are compatible with their naturally occurring relatives ColK and CloDF13. The incompatibility results in loss, by segregation, of one or the other ColE1 plasmid. In most cases, the smaller derivatives tend to displace the larger ones, and the rate of displacement depends on the difference in size. One mini-plasmid retains only 19% of the sequences of ColE1, yet it exrrts strong incompatibility: other ColE1 plasmids are rapidly lost when it is introduced into the host. The region essential for ColE1 incompatibility is deduced to lie within 700 base pairs of the origin of replication.The transforming efficiency of any ColE1 plasmid is markedly lowered when another incompatible replicon is resident in the competent cells, even when the transforming plasmid is much smaller than the resident.A model of incompatibility is proposed to account for these effects.     

Participation of the lytic replicon in bacteriophage P1 plasmid maintenance.

P1 bacteriophage carries at least two replicons: a plasmid replicon and a viral lytic replicon. Since the isolated plasmid replicon can maintain itself stably at the low copy number characteristic of intact P1 prophage, it has been assumed that this replicon is responsible for driving prophage replication. We provide evidence that when replication from the plasmid replicon is prevented, prophage replication continues, albeit at a reduced rate. The residual plasmid replication is due to incomplete repression of the lytic replicon by the c1 immunity repressor. Incomplete repression was particularly evident in lysogens of the thermoinducible P1 c1.100 prophage, whose replication at 32 degrees C remained almost unaffected when use of the plasmid replicon was prevented. Moreover, the average plasmid copy number of P1 in a P1 c1.100 lysogen was elevated with respect to the copy number of P1 c1+. The capacity of the lytic replicon to act as an auxiliary in plasmid maintenance may contribute to the extraordinary stability of P1 plasmid prophage.     

Trimethoprim-Produced F-Specific Insertion Mutations in Escherichia coli K-12

Besides producing thymine-requiring mutants (thy), trimethoprim (TMP) cured the mini-ColE1 replicon pML21 at an appreciable frequency. The cured Escherichia coli K-12 cells behaved like polA mutants by failing to support the stable maintenance of the ColE1 plasmid. The mini-F replicon pSC138, which was lacking all three insertion sequences (IS3, gammadelta, and IS2) normally used for F-specific integration and excision, was not cured by TMP. Instead, it integrated into specific regions of the E. coli chromosome and thus caused auxotrophic mutations in operons which were always localized on either side of oriC (origin of chromosomal replication). The incompatibility and replication functions of the integrated plasmid in auxotrophs were retained, and the plasmid DNAs recovered from spontaneously occurring revertants did not show any alterations in their contour lengths as determined by electron microscopy. The F replicon (fragment 5) contained in plasmid pSC138 carried two origins of replication, the primary origin, oriV(1) at 42.6F and the secondary origin, oriV(2), at 44.1F. Another mini-F plasmid pMF21, deleted of the primary origin of replication (oriV(1)), was still capable of autonomous replication but failed to integrate onto the chromosome after TMP treatment. Furthermore, the composite plasmid pRS5, which normally uses only the replication origin and functions of the pSC101 component, was also insensitive to TMP. On the basis of these results, we propose a new scheme of F integration via the functional oriV(1) and suggest the involvement of a similar mechanism in the formation of Hfr strains by integrative suppression.     

Characterization of a mini-F plasmid derived from an F mutant expressing incompatibility in the autonomous but not in the integrated state

Plasmid DNA from Escherichia coli F' ser/MA219 harboring an altered F' factor, which expressed incompatibility in the autonomous but not in the integrated state (DeVries and Maas, 1973, J. Bacteriol. 115, 213-220), was digested with the restriction endonuclease EcoRI and ligated to a nonreplicating trpED fragment. A miniplasmid was obtained containing a 5.7-kb EcoRI fragment capable of self-replication. This plasmid, designated pRE300, was incompatible with mini-F as well as with ColE1 derivatives. It represents a cointegrate formed in vivo between a 2.2-kb segment of the F replication region and a ColE1-type replicon of unknown derivation. The F-derived component of pRE300 corresponds to a minimalized F replicon (43.85-46.05 kb F) retaining oriII and the incB locus but missing the incC and incD functions. It is postulated that the Inc- mutation resulted from the insertion of a transposable DNA sequence into the incC locus of the parent F plasmid.     

Regulation of DNA replication by iterons: an interaction between the ori2 and incC regions mediated by RepE-bound iterons inhibits DNA replication of mini-F plasmid in Escherichia coli.

In bacteria, plasmids and some DNA viruses, DNA replication is initiated and regulated by binding of initiator proteins to repetitive sequences. To understand the control mechanism we used the plasmid mini-F, whose copy number is stringently maintained in Escherichia coli, mainly by its initiator protein RepE and the incC region. The monomers of RepE protein bound to incC iterons, which exert incompatibility in trans and control the copy number of mini-F plasmid in cis. Many incompatibility defective mutants carrying mutations in their incC iterons had lost the affinity to bind to RepE, while one mutant retained high level binding affinity. The mutated incC mini-F plasmids lost the function to control the copy number. The copy number of the wild-type mini-F plasmid did not increase in the presence of excess RepE. These results suggested that the control of replication by incC iterons does not rely on their capacity to titrate RepE protein. Using a ligation assay, we found that RepE proteins mediated a cross-link structure between ori2 and incC, for which the dimerization domain of RepE and the structure of incC seem to be important. The structure probably causes inhibition of extra rounds of DNA replication initiation on mini-F plasmids, thereby keeping mini-F plasmid at a low copy number.     

Roles of Escherichia coli heat shock proteins DnaK, DnaJ and GrpE in mini-F plasmid replication

Summary A subset of Escherichia coli heat shock proteins, DnaK, DnaJ and GrpE were shown to be required for replication of mini-F plasmid. Strains of E. coli K12 carrying a missense mutation or deletion in the dnaK, dnaJ, or grpE gene were virtually unable to be transformed by mini-F DNA at the temperature (30° C) that permits cell growth. When excess amounts of the replication initiator protein (repE gene product) of mini-F were provided by means of a multicopy plasmid carrying repE, these mutant bacteria became capable of supporting mini-F replication under the same conditions. However, the copy number of a high copy number mini-F plasmid was reduced in these mutant bacteria as compared with the wild type in the presence of excess RepE protein. Furthermore, mini-F plasmid mutants that produce altered initiator protein and exhibit a very high copy number were able to replicate in strains deficient in any of the above heat shock proteins. These results indicate that the subset of heat shock proteins (DnaK, DnaJ and GrpE) play essential roles that help the functioning of the RepE initiator protein in mini-F DNA replication.     

Isolation and characterization of replication-deficient mutants of ColE1 plasmids.

Replication-defective mutants of plasmid ColE1 were isolated from a chimeric plasmid formed by ligating a temperature-sensitive replication derivative of pSC101, pHSG1, with a ColE1-Tn3-containing plasmid. The replication-defective ColE1 mutants isolated were all spontaneous deletion mutants that had lost the ColE1 replication origin and regions adjacent to it. The extent of a deletion was determined by analyzing restriction endonuclease-generated deoxyribonucleic acid fragments of the ColE1 plasmid component of the chimeras by both agarose and polyacrylamide gel electrophoresis. None of the chimeras containing the replication-defective ColE1 mutants was able to replicate in the presence of chloramphenicol. The expression of ColE1 incompatibility was either markedly reduced or not detectable in the replication mutants isolated.     

Plasmids for recombination-based screening.

To facilitate recombination-based screening, we constructed the ColE1-based plasmid, pi G4, that confers chloramphenicol resistance, contains a polylinker with multiple unique restriction enzyme recognition sequences, and contains the genetic marker, supF. To facilitate recombination-based screening followed by rapid DNA sequencing, we inserted the selectable marker, supF, into each of 20 high-copy-number (hcn) pUC-derived NoC plasmids that were designed for multiplex DNA sequencing. To facilitate recombination-based screening of common cDNA libraries that often contain ColE1 sequences, we constructed a supF-carrying plasmid whose replication was driven from an R6K replicon that does not share sequence homology with ColE1. Furthermore, we incorporated a useful polylinker and increased the copy number of this plasmid to create the 4.4-kb hcn plasmid, pMAD1. Thus, these plasmids allow: (1) background-free transformation of cells by a supF plasmid carrying an antibiotic-resistance marker; (2) simultaneous performance of the recombination-based assay and DNA sequencing; and (3) screening bacteriophage cDNA libraries that contain ColE1 sequences by recombination with a supF plasmid that is not homologous to ColE1 derivatives.     

Effect of Inhibitors of Ribonucleic Acid and Protein Synthesis on the Cyclic Adenosine Monophosphate Stimulation of Plasmid ColE1 Replication

Addition of cyclic adenosine 3'-5'-monophosphate (c-AMP) to growing Escherichia coli cells, colicinogenic for the plasmid ColE1, results in a fourfold stimulation in the rate of synthesis of the plasmid deoxyribonucleic acid (DNA). The stimulation is transient (30 min) and is succeeded by a brief period (30 min) of cessation of plasmid DNA replication. The stimulation of ColE1 DNA replication also occurs in chloramphenicol-treated cells. Rifampin inhibits ColE1 DNA replication in the presence or absence of c-AMP. Employing thymine starvation conditions to stop ColE1 DNA synthesis, it was found that c-AMP, added during the period of thymine starvation, effected a stimulation in the amount of subsequent replication which took place when replicating conditions were restored. The stimulatory effect of c-AMP under these conditions was not prevented by chloramphenicol but was completely eliminated when rifampin was present. Under these conditions, when rifampin was added after the effect of c-AMP was allowed to occur, subsequent replication of the plasmid could take place, but only one round of replication occurred. A model to account for the c-AMP effects is presented.