The ccd operon of plasmid F produces three proteins, CcdA, CcdB, and RepD. Prior research has established that the operon is autorepressed and that at least CcdB, but not RepD, is required for autorepression. A role for CcdA in autorepression was suggested but not clearly shown. We now present a series of biochemical experiments which show that both CcdA and CcdB proteins are required for maximal formation of protein-ccd operator complexes. We also show that CcdA and CcdB are present in a complex whether or not ccd operator is present. The clear implication is that autorepressor is a complex of CcdA and CcdB. We also map the start site of the ccd transcript thus providing the first experimental evidence for the location of the ccd promoter.

A major class of replicons in procaryotes is typified by low copy number, nonrandom intracellular distribution, and stable inheritance. Included in this class are chromosomes of gram-positive and gram-negative bacteria as well as a number of plasmids from these organisms. Replicons in this major class have remarkable structural and functional similarities in the genes that effect and control replication. In the present work a review of plasmid F is presented as a paradigm for many aspects of this group's maintenance features.

The thermosensitivity of dnaA(Ts) mutations can be suppressed by integration of plasmid F (integrative suppression). In the light of the recent finding that F requires DnaA protein for both establishment and maintenance, integrative suppression of 11 dnaA(Ts) mutations by a mini-F, pML31, integrated near oriC was examined. The plating efficiency of integratively suppressed strains was dnaA(Ts) allele-dependent and medium-dependent. The initiation capability of suppressed dnaA(Ts) strains lacking the oriC site and their F- counterparts was determined at various temperatures between 30°C and 42°C. The degree of integrative suppression measured by the initiation capability varied in a dnaA(Ts) allele-dependent manner. F-directed DNA replication was most affected by the dnaA(Ts) mutations mapping in the middle of the gene whereas oriC-dependent replication was most thermosensitive in strains carrying mutations mapping in the carboxy-terminal half of the gene. The results indicated that the integrative suppression by F plasmid is a DnaA-dependent process and suggested that the requirements for DnaA protein in the oriC-dependent replication and F replication processes are qualitatively different.

Chromosomal mutants were isolated in which, for several small plasmids, there was an increased amount of either covalently closed circular plasmid DNA or total plasmid DNA or both. The mutations were mapped to recD, which has been shown to affect exonuclease V activity and a variety of plasmid maintenance and replication functions. Our results suggest that rolling-circle plasmid replication can occur in recD mutants and that site-specific recombination can resolve the resulting linear multimers into covalently closed circular plasmid forms.

Determination of plasmid concentration (plasmids/unit volume) is a tedious, time-consuming undertaking when classical procedures are used. Measurement of relative plasmid concentration (plasmids/chromosome equivalent) is quicker but quite inaccurate for very small, low copy number plasmids such as mini-F. Also, one cannot compare validly the plasmid/chromosome equivalents at different growth rates because the amount of chromosome per unit cell volume varies. Therefore, we have devised a method that simultaneously avoids both of these difficulties. We have done this by adapting the Grunstein-Hogness technique [Grunstein, M. and Hogness, D. (1975) Colony hybridization: a method for the isolation of cloned DNAs that contain a specific gene. Proc. Natl. Acad. Sci. U.S.A. 72 3961–3965] of in situ colony lysis with known cell volumes applied to a filter membrane with the aid of a filtration manifold. The amount of specific gene sequence was then determined by hybridization with a radioactive probe. We have succesfully used the technique to measure the concentration of mini-F and native F plasmids that vary in size from 5.0 to 95.0 kb. This method offers several advantages over currently employed techniques. These include: sensitivity, speed, and cost and equipment that is required.

We have previously described a monoclonal antibody (BU-1) to 5-bromo-2-deoxyuridine (BrdUrd) that is useful for measurement of cell cycle S-phase. BU-1 hybridoma supernatant reacted with incorporated BrdUrd after the cells had been ethanol fixed; without a requirement for acid or base denaturation. We have found that this reactivity is lost if purified antibody is used, if the culture supernatants are heated, or if a mycoplasma-free hybridoma line is isolated. The supernatant contained endogenous DNase activity that was a result of mycoplasma infection of the cell line. This DNase activity was required for staining the cells with BU-1 in the absence of other denaturation steps. The endogenous DNase could be substituted for by the addition of bovine pancreatic DNase I. The disruption of the double stranded DNA structure with an enzyme rather than with harsh chemical or heat treatments does not affect protein structure or cellular morphology and allows the detection of incorporated BrdUrd of morphologic or antigenic cell subsets. DNase pre-treatment may also be useful for detection of other 'hidden' DNA antigens.

We previously reported the existence of a series of chemically induced trans recessive copy-number mutations (cop) for mini-F plasmids and the existence of a similar series of cop mutations induced by insertion of the ampicillin resistance transposon Tn3. In this paper we describe the experiments showing that these two series of mutations are in different genes. Briefly, the experiments show that the one mutant series can complement the other, that the mutations map in distinct but adjacent regions, that the copy numbers of double mutants are the products of the copy numbers determined by the single mutations, and that Tn3 does not elevate copy number by a polar effect on the adjacent cop gene defined by chemical mutagenesis. We term the latter gene copA and the gene mutated by Tn3, copB. We also demonstrate here that copB mutations are recessive to the wild type allele. Further, we have characterized copB by deletion and recombinational analysis as the series of five 19- to 22-base-pair directly repeated sequences that had previously been designated incC-that is, one of the incompatibility genes. The evidence for this conclusion is that plasmids lacking two, three or five direct repeats have their copy number elevated proportionately. Possible mechanisms for copB control of replication are discussed.

We have taken advantage of two situations in which the incompatibility barrier between F plasmids is overcome to show that wild-type genes controlling F copy number (cop+) are dominant in trans over mutant genes. The simplest interpretation of our findings is that the cop mutations have inactivated a repressor gene that controls F replication. Since the cop. mutations all map in a region that others have shown by sequence analysis to theoretically encode four proteins, a strong possibility exists that one of these proteins is the repressor.

Certain derivative mini-F plasmids were found to segregate into Escherichia coli minicells, in contrast to the intact mini-F plasmid which does not. Segregation was not related to the presence or absence of the normal origin of vegetative replication, but appeared to be affected by regions of F which encode replication, incompatibility, copy number control, and partitioning functions. Segregation of mini-F plasmids into minicells was not random; the plasmid concentration in minicells did not correlate with the plasmid concentration in cells. Genes, or gene products, of F from the region spanning the sequences 44.1–49.3F appeared to affect the ability of mini-F plasmids to segregate into minicells. Segregation of mini-F plasmids into minicells was not directly related to stable plasmid inheritance. These results argue for the sequestration of mini-F plasmids in host cells.

We have analysed as a function of cell doubling times the control of R6K plasmid replication in rep+ and rep strains of Escherichia coli. The rep mutation results in an alteration or loss of an enzyme that unwinds helical DNA. We found in rep+ bacteria that R6K relative dosage (plasmids per genome equivalent) remained nearly constant as growth rates increased. From this we concluded that the average plasmid concentration (plasmids per unit cell mass or volume) fell relative to the average concentration of chromosome origins when growth rates increased. In this context, the control of R6K replication is similar to that of other plasmids as seen by different workers. We also found that the relative dosage of R6K in rep mutants is greater than in rep+ bacteria when both strains were grown at fast growth rates. This finding was expected since at fast growth rates the number of genome equivalents per unit mass is expected to be lower in rep mutants. Unexpectedly, however, we found the effect of the rep mutation on R6K relative dosage had occurred in a step-like manner at a slow growth rate of about 120 min per generation. This implies that both the relative dosage and concentration of R6K had increased in a step-like manner. We also found that the effect of the rep mutation on R6K concentration was lost at fast growth rates while the effect of the mutation on R6K relative dosage was not lost.

By use of recombinant DNA techniques, we have inserted the lac+ operon into a transposon (Tn3). We constructed the recombinant in such a way that the essential step in assaying for transposition consisted of screening for bacteria with a thermostable Lac+ phenotype. Our results showed that transposition of the Tn3[lac+] element occurred and that its frequency was derepressed compared to frequencies reported by others for wild-type Tn3 transposition.

A number of plasmid systems have been examined for the ability of their covalently closed circular deoxyribonucleic acid (CCC DNA) forms to cosediment in neutral sucrose gradients with the folded chromosomes of their respective hosts. Given that cosedimentation of CCC plasmid and chromosomal DNA represents a bound or complexed state between these replicons, our results can be expressed as follows. (i) All plasmid systems complex, on the average, at least one plasmid per chromosomal equivalent. (ii) Stringently controlled plasmids exist predominantly in the bound state, whereas the opposite is true for plasmids that exist in multiple copies or are under relaxed control of replication. (iii) The degree to which a plasmid population binds to host chromosomes appears to be a function of plasmid genotype and not of plasmid size. (iv) For the colicin E1 plasmid the absolute number of plasmids bound per folded chromosome equivalent does increase as the intracellular plasmid/chromosome ratio increases in cells starved for required amino acids or in cells treated with chloramphenicol; however, the ratio of bound to free plasmids remains constant during plasmid copy number amplification.