Random replication and random assortment model for plasmid incompatibility in bacteria

A mathematician's guide to plasmids: an introduction to plasmid biology for modellers

Plasmids, extrachromosomal DNA molecules commonly found in bacterial and archaeal cells, play an important role in bacterial genetics and evolution. Our understanding of plasmid biology has been furthered greatly by the development of mathematical models, and there are many questions about plasmids that models would be useful in answering. In this review, we present an introductory, yet comprehensive, overview of the biology of plasmids suitable for modellers unfamiliar with plasmids who want to get up to speed and to begin working on plasmid-related models. In addition to reviewing the diversity of plasmids and the genes they carry, their key physiological functions, and interactions between plasmid and host, we also highlight selected plasmid topics that may be of particular interest to modellers and areas where there is a particular need for theoretical development. The world of plasmids holds a great variety of subjects that will interest mathematical biologists, and introducing new modellers to the subject will help to expand the existing body of plasmid theory.

Fixation dynamics of beneficial alleles in prokaryotic polyploid chromosomes and plasmids

Theoretical population genetics has been mostly developed for sexually reproducing diploid and for monoploid (haploid) organisms, focusing on eukaryotes. The evolution of bacteria and archaea is often studied by models for the allele dynamics in monoploid populations. However, many prokaryotic organisms harbor multicopy replicons—chromosomes and plasmids—and theory for the allele dynamics in populations of polyploid prokaryotes remains lacking. Here, we present a population genetics model for replicons with multiple copies in the cell. Using this model, we characterize the fixation process of a dominant beneficial mutation at 2 levels: the phenotype and the genotype. Our results show that depending on the mode of replication and segregation, the fixation of the mutant phenotype may precede genotypic fixation by many generations; we term this time interval the heterozygosity window. We furthermore derive concise analytical expressions for the occurrence and length of the heterozygosity window, showing that it emerges if the copy number is high and selection strong. Within the heterozygosity window, the population is phenotypically adapted, while both alleles persist in the population. Replicon ploidy thus allows for the maintenance of genetic variation following phenotypic adaptation and consequently for reversibility in adaptation to fluctuating environmental conditions.

Plasmid co-infection: linking biological mechanisms to ecological and evolutionary dynamics

As infectious agents of bacteria and vehicles of horizontal gene transfer, plasmids play a key role in bacterial ecology and evolution. Plasmid dynamics are shaped not only by plasmid-host interactions but also by ecological interactions between plasmid variants. These interactions are complex: plasmids can co-infect the same cell and the consequences for the co-resident plasmid can be either beneficial or detrimental. Many of the biological processes that govern plasmid co-infection-from systems that exclude infection by other plasmids to interactions in the regulation of plasmid copy number-are well characterized at a mechanistic level. Modelling plays a central role in translating such mechanistic insights into predictions about plasmid dynamics and the impact of these dynamics on bacterial evolution. Theoretical work in evolutionary epidemiology has shown that formulating models of co-infection is not trivial, as some modelling choices can introduce unintended ecological assumptions. Here, we review how the biological processes that govern co-infection can be represented in a mathematical model, discuss potential modelling pitfalls, and analyse this model to provide general insights into how co-infection impacts ecological and evolutionary outcomes. In particular, we demonstrate how beneficial and detrimental effects of co-infection give rise to frequency-dependent selection on plasmid variants. This article is part of the theme issue 'The secret lives of microbial mobile genetic elements'.

Mathematical Models of Plasmid Population Dynamics

With plasmid-mediated antibiotic resistance thriving and threatening to become a serious public health problem, it is paramount to increase our understanding of the forces that enable the spread and maintenance of drug resistance genes encoded in mobile genetic elements. The relevance of plasmids as vehicles for the dissemination of antibiotic resistance genes, in addition to the extensive use of plasmid-derived vectors for biotechnological and industrial purposes, has promoted the in-depth study of the molecular mechanisms controlling multiple aspects of a plasmids' life cycle. This body of experimental work has been paralleled by the development of a wealth of mathematical models aimed at understanding the interplay between transmission, replication, and segregation, as well as their consequences in the ecological and evolutionary dynamics of plasmid-bearing bacterial populations. In this review, we discuss theoretical models of plasmid dynamics that span from the molecular mechanisms of plasmid partition and copy-number control occurring at a cellular level, to their consequences in the population dynamics of complex microbial communities. We conclude by discussing future directions for this exciting research topic.

Quantifying plasmid dynamics using single-cell microfluidics and image bioinformatics

Multicopy plasmids play an important role in bacterial ecology and evolution by accelerating the rate of adaptation and providing a platform for rapid gene amplification and evolutionary rescue. Despite the relevance of plasmids in bacterial evolutionary dynamics, evaluating the population-level consequences of randomly segregating and replicating plasmids in individual cells remains a challenging problem, both in theory and experimentally. In recent years, technological advances in fluorescence microscopy and microfluidics have allowed studying temporal changes in gene expression by quantifying the fluorescent intensity of individual cells under controlled environmental conditions. In this paper, we will describe the manufacture, experimental setup, and data analysis pipeline of different microfluidic systems that can be used to study plasmid dynamics, both in single-cells and in populations. To illustrate the benefits and limitations of microfluidics to study multicopy plasmid dynamics, we will use an experimental model system consisting on Escherichia coli K12 carrying non-conjugative, multicopy plasmids (19 copies per cell, in average) encoding different fluorescent markers and β-lactam resistance genes. First, we will use an image-based flow cytometer to estimate changes in the allele distribution of a heterogeneous population under different selection regimes. Then we will use a mothermachine microfluidic device to obtain time-series of fluorescent intensity of individual cells to argue that plasmid segregation and replication dynamics are inherently stochastic processes. Finally, using a microchemostat, we track thousands of cells in time to reconstruct bacterial lineages and evaluate the allele frequency distributions that emerge in response to a range of selective pressures.

Evolutionary Rescue and Drug Resistance on Multicopy Plasmids

Bacteria often carry "extra DNA" in the form of plasmids in addition to their chromosome. Many plasmids have a copy number greater than one such that the genes encoded on these plasmids are present in multiple copies per cell. This has evolutionary consequences by increasing the mutational target size, by prompting the (transitory) co-occurrence of mutant and wild-type alleles within the same cell, and by allowing for gene dosage effects. We develop and analyze a mathematical model for bacterial adaptation to harsh environmental change if adaptation is driven by beneficial alleles on multicopy plasmids. Successful adaptation depends on the availability of advantageous alleles and on their establishment probability. The establishment process involves the segregation of mutant and wild-type plasmids to the two daughter cells, allowing for the emergence of mutant homozygous cells over the course of several generations. To model this process, we use the theory of multitype branching processes, where a type is defined by the genetic composition of the cell. Both factors-the availability of advantageous alleles and their establishment probability-depend on the plasmid copy number, and they often do so antagonistically. We find that in the interplay of various effects, a lower or higher copy number may maximize the probability of evolutionary rescue. The decisive factor is the dominance relationship between mutant and wild-type plasmids and potential gene dosage effects. Results from a simple model of antibiotic degradation indicate that the optimal plasmid copy number may depend on the specific environment encountered by the population.

Defective Interference in the Killer System of Saccharomyces cerevisiae

The K(1) killer virus (or plasmid) of Saccharomyces cerevisiae is a noninfectious double-stranded RNA genome found intracellularly packaged in an icosahedral capsid. This genome codes for a protein toxin and for resistance to that toxin. Defective interfering virus mutants are deletion derivatives of the killer virus double-stranded RNA genome; such mutants are called suppressive. Unlike strains carrying the wild-type genome, strains with these deletion derivatives are neither toxin producers nor toxin resistant. If both the suppressive and the wildtype virus are introduced into the same cell, most progeny become toxin-sensitive nonkillers (J. M. Somers, Genetics 74:571-579, 1973). Diploids formed by the mating of a killer with a suppressive strain were grown in liquid culture, and RNA was extracted from samples taken up to 41 generations after the mating. The ratio of killer RNA to suppressive RNA decreased with increasing generations; by 41 generations the killer RNA was barely detectable. The copy numbers of the suppressive genome and its parental killer were virtually the same in isogenic strains, as were the growth rates of diploid strains containing either virus alone. Therefore, suppressiveness, not being due to segregation or overgrowth by faster growing segregants, is likely due to preferential replication or maintenance of the suppressive genome. Three suppressive viruses, all derivatives of the same killer virus (T. K. Sweeney et al., Genetics 84:27-42, 1976), did not coexist stably. The evidence strongly indicates that the largest genome of the three slowly suppressed both of the smaller genomes, showing that larger genomes can suppress smaller ones and that suppression can occur between two suppressives. Of 48 isolates of strains carrying the suppressive viruses, 5 had newly detectable RNA species, all larger than the original suppressive genomes. At least seven genes necessary for maintenance of the wild-type killer virus (MAK genes) were needed by a suppressive mutant. No effect of ski mutations (affecting regulation of killer virus double-stranded RNA replication) on suppressiveness was observed.

Rate of segregation due to plasmid incompatibility

We calculated the rates of segregation due to plasmid incompatibility under several simple models. A common feature of all the models that we considered is that incompatibility is caused by the inability of the segregation mechanism to distinguish between two incompatible plasmids.

We measured the rate of segregation due to incompatibility of a pair of ColE1 derivatives under two conditions: (1) One plasmid was introduced into cells carrying the other by conjugation. (2) Cells carrying both plasmids were maintained by selection and then selection was released.

Interpretation of the results was made more difficult by effects of the Plasmids on the host cell's growth rate. These experiments gave results in agreement with the predictions of a random pool replication model. Published results were also in reasonable agreement with this model.

A kinetic model for plasmid curing

A simple mathematical model of drug-induced plasmid elimination (curing) considering density-dependent growth rates and plasmid transfers is presented. It describes nonlinear population dynamics of conjugative plasmids during in vitro curing experiments in batch culture. The model was tested on kinetics of acridine orange curing of F'lac plasmid. Effects of density dependence, plasmid elimination, selection for plasmidless segregants, conjugation, initial and maximal population density, and postsegregational killing on curing kinetics are simulated and discussed.

Identification and classification of bacterial plasmids

Images

Non-mendelian inheritance of mitochondrial DNA and ribosomal DNA in the myxomycete, Didymium iridis

The inheritance of both the mitochondrial DNA (mtDNA) and the nuclear-encoded extrachromosomal ribosomal DNA (rDNA) has been studied in the myxomycete, Didymium iridis, by DNA-DNA hybridization of labeled probes to total DNA at various stages of the life cycle. Both the mtDNA and rDNA populations rapidly become homogeneous in individuals, but there is a qualitative difference in the patterns of inheritance of these two molecules. One parental rDNA type was preferentially inherited in all crosses; selective replication of this molecule is tentatively proposed as the mechanism of inheritance. In contrast, either parental mtDNA type could be inherited. Since the inherited population of parental mtDNA molecules are not partitioned into cells in this coenocytic organism, no known mechanism of inheritance can explain the rapid and apparently random loss of one parental mtDNA type in individuals.

Coexistence of incompatible plasmids in a bacterial population living under a feast and famine regime

A model is formulated to examine the possibility of (co)existence of plasmids of the same incompatibility and surface exclusion group in a bacterial population living under a feast-and-famine regime. The condition is given under which a growth rate decreasing plasmid can invade a bacterial population. It appears that in case only one plasmid type is present, the frequency of plasmid bearers will tend to a stable equilibrium if the food supply at each growth site gets exhausted and if both plasmid-free and plasmid-bearing bacteria need an equal quantity of food per cell division. If these two conditions are not satisfied, the frequency of plasmid-bearers might oscillate. Two plasmids will sometimes be able to coexist, but only if they follow different survival strategies; one with a high conjugational transfer rate and a lower fitness of its host, and the other with a low transfer rate and a higher host fitness. Coexistence of three plasmids of the same surface exclusion group is impossible.

Effect of the deletion of a fragment dispensable for the autonomous maintenance of plasmid pT181 on the competition between incompatible plasmids

The deletion of the 560-bp HindIII C fragment from pT181 derivatives does not change the stability or copy number of the plasmid but affects its ability to compete with undeleted, incompatible plasmids for maintenance in the host cell. The disadvantage of the deleted plasmids seems to be manifested at the level of replication. It results that for plasmid pT181 a sequence dispensable for autonomous maintenance and replication control could affect the outcome of the competition between autonomous, incompatible plasmids.

Double-stranded RNAs that encode killer toxins in Saccharomyces cerevisiae: unstable size of M double-stranded RNA and inhibition of M2 replication by M1

The sizes of M1 and M2 (but not L) change rapidly with growth, varying by perhaps as much as 33%. Size variation is seen within 76 generations. In addition, the exclusion of M2 by M1 or L-A-E [( EXL]) is mediated by inhibition of replication or segregation, not by enhanced degradation of preexisting molecules.

Recombination between repeated DNA sequences occurs more often in plasmids than in the chromosome of Bacillus subtilis

Directly repeated pBR322 sequences 3.7-3.8 kb long recombine with a frequency of about 10% per generation when carried on plasmids related to pC194 and 0.01% per generation when carried on the Bacillus subtilis chromosome. Recombination is therefore 1,000 times more efficient in plasmids than in the chromosome of this organism.

Double-stranded RNAs that encode killer toxins in Saccharomyces cerevisiae: unstable size of M double-stranded RNA and inhibition of M2 replication by M1

The sizes of M1 and M2 (but not L) change rapidly with growth, varying by perhaps as much as 33%. Size variation is seen within 76 generations. In addition, the exclusion of M2 by M1 or L-A-E [( EXL]) is mediated by inhibition of replication or segregation, not by enhanced degradation of preexisting molecules.

Partition of unit-copy miniplasmids to daughter cells. I. P1 and F miniplasmids contain discrete, interchangeable sequences sufficient to promote equipartition

Hybrids formed by insertion of the plasmid maintenance regions of P1 or F into a lambda delta att vector form stable unit-copy plasmids in their Escherichia coli host. They must therefore both be substrates for an accurate cellular partition apparatus that ensures that all daughter cells inherit a plasmid copy. Analysis of deletion mutants of both types of hybrid showed that, although the P1 and F plasmid maintenance regions differ in sequence and specificity, they are similar in general organization. Each contains an approximately 3 X 10(3) base-pair region that is essential for replication (rep) and an adjacent but separable 3 X 10(3) base-pair region that is essential for the stability of plasmid maintenance (par). Each par region is thought to specify the recognition of the plasmid as a substrate for equipartition. The deletion mutants provide sources of isolated rep and par sequences from both P1 and F DNA. These elements were then used to construct composite plasmids with novel combinations and arrangements of rep and par sequences. Heterologous constructions containing P1 rep and F par or F rep and P1 par sequences were maintained faithfully. We conclude that par regions are both necessary and sufficient to promote equipartition of replicating plasmid DNA. This activity is exerted only in cis but otherwise seems to be independent of the position or orientation of the par sequences within the DNA. Both P1 and F par regions include DNA sequences (incB of P1, incD of F) that we propose are analogues of the centromeres of eukaryotic chromosomes. The remaining portions of the par regions are known to encode protein products that, we believe, act at the inc sites. Extra copies of these inc sites appear to exert incompatibility by competition for the cellular partition apparatus.

Staphylococcus aureus chromosomal mutation specifically affecting the copy number of Inc3 plasmids

A chromosomal mutation leading to an important increase in the copy number of plasmid pT181 and its derivatives has been isolated from Staphylococcus aureus strain 8325. The amplification effect in the mutant strain SA1350 was found to be specific for plasmids of the Inc3 group, to which belongs pT181. There are some other differences in the behavior of Inc3 plasmids between SA1350 and 8325, including stable maintenance in SA1350 at high copy number of temperature-sensitive replication mutants at restrictive temperatures, and altered incompatibility properties. Derivatives of SA1350 carrying only Inc3 plasmid mutants with high copy numbers (Cop mutants) could not be obtained, suggesting a lethal runaway plasmid replication in this situation. SA1350 expressed also a temperature-sensitive phenotype. The relationship of this character to the plaC1 mutation determining the amplification of Inc3 plasmids has not yet been elucidated.

The yeast plasmid 2mu circle encodes components required for its high copy propagation

The yeast plasmid 2mu and certain hybrid plasmids constructed from it are maintained stably and at high copy number in yeast cells. By examining various mutant hybrid 2mu plasmids, we show that these properties require the integrity of four plasmid loci. Two of these, designated REP1 and REP2, are active in trans and correspond to two open coding regions of 2mu. The other two loci are active only in cis and correspond to the origin of replication and to a region, designated REP3, located several hundred bp away from the origin and consisting of direct repeats of a 62 bp sequence. We propose that the REP loci constitute a copy control system that overrides normal cellular restriction on plasmid replication and amplifies the plasmid when copy number is low.

Replication control mutations of plasmid R6-5 and their effects on interactions of the RNA-I control element with its target

Nine high copy number mutations of plasmid R6-5, representing five phenotypically distinct groups, have been identified by DNA sequencing. In each mutant plasmid examined, a single nucleotide change was found. The effects of the mutations on possible gene products, and DNA-RNA secondary structure, were analyzed and compared with the observed phenotypes. The results of this study exclude the possibility that the primary plasmid replication control element, the product of the copA gene, is a polypeptide, and they are consistent with a model of plasmid replication control by the copA product which has the following features: (i) RNA-I, a short untranslated RNA molecule, is the product of the copA gene and regulates the frequency of initiation of plasmid replication, (ii) the hexanucleotide single-strand loop of the major hairpin of RNA-I is its active site, (iii) this active site functions by base pair interactions with its "target," its DNA template strand, or its complementary sequence on RNA-II, a transcript of opposite polarity that is the message of the repA gene, and (iv) the sequence and size of the loop, and the stability of the stem of the hairpin, are all critical factors that govern the functioning of RNA-I.

Some stochastic models for plasmid copy number

Some stochastic models for the copy number of plasmids in a cell line are studied. When considering the behavior of copy number in the whole cell line, the theory of multitype branching processes is appropriate. Attention is paid to the cure rate in the cell line, and the asymptotic fractions of cells containing a given number of plasmids. These quantities are used to compare the models numerically.

Partition mechanism of F plasmid: two plasmid gene-encoded products and a cis-acting region are involved in partition

Plasmids that replicate using the replication origin (oriC) of the E. coli chromosome are not stably inherited through cell division, but can be stabilized by joining with a particular segment of F plasmid that presumably provides the partition function. The segment necessary for stabilization has been located within a 3.0 kb segment outside of the region essential for autonomous replication of the F plasmid. This segment contains three functionally distinct regions: two of them (designated sopA and sopB) specify gene products that act in trans, whereas the third region (sopC) acts in cis. All three functions seem to be essential for normal partition of the plasmid into daughter cells during cell division. The cis-acting region also specifies plasmid incompatibility.

Chloramphenicol acetyltransferase: enzymology and molecular biology

Naturally occurring chloramphenicol resistance in bacteria is normally due to the presence of the antibiotic inactivating enzyme chloramphenicol acetyltransferase (CAT) which catalyzes the acetyl-S-CoA-dependent acetylation of chloramphenicol at the 3-hydroxyl group. The product 3-acetoxy chloramphenicol does not bind to bacterial ribosomes and is not an inhibitor of peptidyltransferase. The synthesis of CAT is constitutive in E. coli and other Gram-negative bacteria which harbor plasmids bearing the structural gene for the enzyme, whereas Gram-positive bacteria such as staphylococci and streptococci synthesize CAT only in the presence of chloramphenicol and related compounds, especially those with the same stereochemistry of the parent compound and which lack antibiotic activity and a site of acetylation (3-deoxychloramphenicol). Studies of the primary structures of CAT variants suggest a marked degree of heterogeneity but conservation of amino acid sequence at and near the putative active site. All CAT variants are tetramers composed in each case of identical polypeptide subunits consisting of approximately 220 amino acids. The catalytic mechanism does not appear to involve an acyl-enzyme intermediate although one or more cysteine residues are protected from thiol reeagents by substrates. A highly reactive histidine residue has been implicated in the catalytic mechanism.

Replication and incompatibility properties of plasmid pE194 in Bacillus subtilis

pE194, a 3.5-kilobase multicopy plasmid, confers resistance to the macrolide-lincosamide-streptogramin B antibiotics in Bacillus subtilis. By molecular cloning and deletion analysis we have identified a replication segment on the physical map of this plasmid, which consists of about 900 to 1,000 base pairs. This segment contains the replication origin. It also specifies a trans-acting function (rep) required for the stable replication of pE194 and a negatively acting copy control function which is the product of the cop gene. The target sites for the rep and cop gene products are also within this region. Two incompatibility determinants have been mapped on the pE194 genome and their properties are described. One (incA) resides within the replication region and may be identical to cop. incB, not located in the replication region, expresses incompatibility toward a copy control mutant (cop-6) but not toward the wild-type replicon.

Genetic and physical map of a P1 miniplasmid

The prophage form of bacteriophage P1 is a unit-copy plasmid which is maintained with great fidelity in its Escherichia coli host. The plasmid maintenance functions of P1 are clustered in one region of the genome. An 11.5-kilobase fragment from this region has been cloned into a lambda delta att vector and promotes stable unit-copy plasmid maintenance. The properties of the lambda vector facilitated the isolation of deletion mutants affecting the P1 DNA. Twenty-eight deletion mutants were isolated, and their lesions were mapped by physical techniques. The genetic properties of the mutants with respect to plasmid replication, stability of plasmid maintenance, and ability to exert incompatibility effects against P1 and P7 plasmids were determined. These properties, along with those of several subfragments of the P1 insert cloned into high-copy-number plasmid vectors, allow the construction of an unambiguous genetic and physical map of the maintenance functions. A region of less than 3 kilobases, the rep region, is essential for plasmid replication and contains the incA incompatibility determinant within an 800-base-pair segment. Immediately adjacent to rep is a second region of approximately 3 kilobases which is required for stable plasmid maintenance, but not replication. This region, par, contains a second incompatibility element incB which is approximately 1 kilobase in size. The par region appears to specify equipartition of plasmid copies to daughter cells during cell division.

Temperature sensitive replication plasmids are passively distributed during cell division at non-permissive temperature: a new model for replicon duplication and partitioning

To see whether plasmid molecules in bacteria are equally partitioned or randomly distributed at cell division, the segregation properties of temperature sensitive replication mutants of the E. coli plasmid pSC101 were tested at non-permissive temperature. The results support the idea that at least unreplicated molecules are passively distributed and thus the Equipartition Model is unlikely even under physiological conditions if plasmids replicate randomly. Therefore, we developed a new model which involves the Random Replication Hypothesis and assumes that only the two products of the last plasmid replication event are actively partitioned into two daughter cells and the others are randomly distributed. Mathematical studies revealed that the incompatibility segregation rate predicted by this model fits the experimental data.

Maintenance of the bacteriocinogenic plasmid Clo DF13 in Escherichia coli cells. I. Localisation and mutual interactions of four Clo DF13 incompatibility regions

The incompatibility properties of the bacteriocinogenic plasmid Clo DF13 have been examined. By using Clo DF13, Clo DF13 deletion, and transposon insertion mutants as well as compatible R plasmids into which Clo DF13 fragments have been cloned, we could identify and localise four different incompatibility regions on the Clo DF13 genome. These regions, designated incA, incB, incC, and incD are located in the following positions: incA about 32%, incB between 45% and 50%; incC about 97% and incD between 1.8% and 9% of the Clo DF13 genome. We studied the contribution of each of the four inc regions, separately and/or in combination with each other, to the incompatibility between two plasmid replicons. Two types of incompatibility can be distinguished: Type I evoked by incD, that overlaps the replication control area of Clo DF13 and type II, caused by incA, B and C. From our observations we present a model for plasmid incompatibility based on a combination of the existing repressor dilution and membrane attachment models.

Co-curing of plasmids affecting killer double-stranded RNAs of Saccharomyces cerevisiae: [HOK], [NEX], and the abundance of L are related and further evidence that M1 requires L

We describe two sets of plasmid-plasmid interactions in the yeast Saccharomyces cerevisiae. [HOK], [EXL], [NEX], and [KIL-k1] are genetically defined plasmids, and M1 and L are biochemically defined double-stranded RNA plasmids. We show that (i) [HOK], [NEX], and the abundance of L are related, and (ii) under submaximal curing conditions, all colonies retaining M1 also retain L. There are three pieces of evidence that either [NEX] required [HOK] for replication or [NEX] and [HOK] are on the same plasmid. The evidence is as follows. (i) The great majority of strains containing [HOK] also contain [NEX]. However, two [HOK] [NEX-o] strains do exist. (ii) Growth at 39 degrees C or growth at 34 degrees C with 3% ethanol or 2-propanol cures [HOK] and [NEX]. In a [HOK] [NEX] strain, the two plasmids are always co-cured. (iii) [HOK] and [NEX] are both maintained in mak4, mak6, and mak27 strains (mak = maintenance of [KIL-k1]), but not in mak3, mak10, and pet18 strains. Strains containing [HOK] and [NEX] have about fourfold more L double-stranded RNA than their isochromosomal, cured derivatives. In addition, a cytoductant which has acquired [HOK] and [NEX] has fourfold more L than its parent. These results are consistent with either [HOK] being a form of L or [HOK] increasing the copy number of L. Using a K1 killer strain in which L, as well as M1, could be cured by growth at 38 degrees C, we examined the distribution of loss of M1 and L under conditions giving 98% M-o colonies and at least 50% L-o colonies. No M1L-o colonies were observed, supporting the previous suggestion by others that M1 requires L.

Regular segregation of composite plasmid Rms201

Copy number mutants Rms201ts15 and Rms201ts16 were isolated at 30 degrees C from a temperature-sensitive replication mutant (Rms201ts14) of the conjugative plasmid Rms201. The numbers of plasmids per chromosome of ML1410(Rms201ts14), ML1410(Rms201ts15), and ML1410(Rms201ts16) grown at 30 degrees C were 2.2, 7.4, and 20, respectively. The synthesis of covalently closed circular plasmid deoxyribonucleic acid stopped in Rms201ts14, Rms201ts15, and Rms201ts16 immediately after a "shift-up" in temperature (42 degrees C). At 42 degrees C, antibiotic-sensitive derivatives appeared after a certain lag time: the lag times of ML1410(Rms201ts14), ML1410(Rms201ts15), and ML1410(Rms201ts16) were 2.5, 5, and 6.8 generations, respectively. After these times, plasmid-positive cells in the populations decreased at a rate of about 50% per generation in all of the mutants. From these results we conclude that plasmid segregation (partition) of Rms201 occurs by regular segregation (partition).

Unstable amplification of an altered dihydrofolate reductase gene associated with double-minute chromosomes

We have studied a line of 3T6 mouse fibroblasts grown in progressively increasing concentrations of methotrexate. Initially, drug resistance results from amplification of the gene encoding the normal dihydrofolate reductase. Growth of these methotrexate-resistant populations at higher methotrexate concentrations results in the emergence of cells expressing high levels of dihydrofolate reductase with a reduced methotrexate affinity. Using the fluorescence-activated cell sorter, we demonstrate that the variant gene is not present in the population of cells resistant to lower levels of methotrexate, and hence we postulate that the mutational event occurred in cells already containing multiple normal dihydrofolate reductase genes. Growth of the variant cells in the absence of selection is associated with the permanent loss of the altered genes and the disappearance of double-minute chromosomes, on which these genes reside. The pattern of accumulation and loss of double-minute chromosomes is reproduced following transformation of methotrexate-sensitive cells with the altered genes. Our results are consistent with autonomous replication of double-minute chromosomes and a selective advantage of cells with the smallest number of extrachromosomal elements necessary for survival at a given methotrexate concentration.

A mutant defective in partitioning of composite plasmid Rms201

Escherichia coli harboring mutant plasmids defective in maintenance stability (from the conjugative plasmid Rms201) showed a wide distribution of ampicillin resistance levels, as well as increased frequency of plasmid loss from the cell. The amounts of covalently closed circular deoxyribonucleic acid of mutant plasmid Rms268 and parental plasmid Rms201 per chromosome were 5.3 and 6.1%, respectively. The beta-lactamase activities of strains W3630(Rms268) and W3630(Rms201) were 0.56 and 0.44 U/mg of protein, respectively. Frequency of plasmid loss from W3630(Rms268) was about 0.8 to 1.2% per cell generation, 100 times more than that of the wild-type strain. Ampicillin resistance levels of the colonies harboring the mutant plasmid showed a wide distribution, from low (100 micrograms/ml) to high (1,600 micrograms/ml). A miniplasmid (pMS268) with a mass of 7 X 10(6) daltons and encoding ampicillin resistance was isolated from Rms268. Frequency of pMS268 loss from W3630(pMS268) was about 0.8 to 1.9% per cell generation. W3630(pMS268) also showed a wide range of distribution in the levels of ampicillin resistance. These results indicated that the copies of Rms268 in E. coli did not segregate evenly between daughter cells at cell division and that the gene involved was located on the miniplasmid.

Replication control and switch-off function as observed with a mini-F factor plasmid

Mini-F is a fragment of the F plasmid, consisting of 9,000 base pairs, which carries all of the genes and sites required for replicon maintenance and control. Its copy number is one to two per chromosome. This plasmid is joined to ColE1, whose copy number is 16 to 20. Under normal circumstances the composite plasmid replication exhibited ColE1 characteristics, maintaining a high copy number. However, when ColE1 replication was inhibited by deoxyribonucleic acid polymerase I inactivation, its replication exhibited mini-F characteristics, maintaining a low copy number. These observations are in complete agreement with those of Timmis et al. (Proc. Natl. Acad. Sci. U.S.A. 71:4556-4560, 1974), who examined the behavior of a recombinant plasmid formed between pSC101 and ColE1. The transition from high to low copy number allowed us to examine the control system acting in cells carrying plasmids exhibiting intermediate copy numbers. The initiation of the mini-F replication system as represented by deoxyribonucleic acid synthesis of the composite plasmid was completely blocked when there were multiple copies of mini-F in a cell. It was not restored until the copy number was lowered to one to two, after which replication was first detected. ppF, a mini-F replicon packaged in a phage lambda head behaved similarly: its replication was completely shut off when the resident mini-F genome copy number was high and was inhibited partially when the resident mini-F genome copy number was low. These experiments clearly demonstrate that there is a switch-off mechanism acting on deoxyribonucleic acid synthesis (initiation) in a cell carrying mini-F, and its intensity is related to the plasmid copy number. This result supports the "inhibitor dilution model" proposed by Pritchard et al. (Symp. Soc. Gen. Microbiol. 19:263-297, 1969). The nature of the hypothetical inhibitor is discussed.

Incompatibility properties of Col E1 and pMB1 derivative plasmids: random replication of multicopy replicons

The incompatibility properties of Col E1-like plasmids have been examined in Rec+ and RecA- bacteria. Two Col E1- (or two pMB1-) derivative plasmids coreplicated in the same clone for many cell doublings, irrespective of the rec genotype of host bacteria. Their kinetics of segregation were found to be consistent with models that assume a random choice of template molecule for each plasmid replication event, but with models based on a single (master) template molecule per cell. In contrast, minimal coreplication of a Col E1- and a pMB1-derivative plasmid occurred, with the latter type rapidly excluding the former. We suggest here that the pMB1 derivatives, pMB9 and pBR322, are less sensitive than Col E1 derivatives to the putative inhibitor that regulates plasmid replication, due to base sequence differences in their target for the inhibitor, and consider one mechanism whereby the duplication of Col E1-like plasmids might be regulated.

Plasmid replication functions: two distinct segments of plasmid R1, RepA and RepD, express incompatibility and are capable of autonomous replication

The genetic determinants for replication and incompatibility of plasmid R1 were investigated by gene cloning methods, and three types of R1 miniplasmid derivatives were generated. The first, exemplified by plasmid pKT300, consisted of a single BglII endonuclease-generated deoxyribonucleic acid fragment derived from the R1 region that is located between the determinants for conjugal transfer and antibiotic resistance. Two types of miniplasmids could be formed from PstI endonuclease-generated fragments of pKT300. One of these, which is equivalent to miniplasmids previously generated from plasmids R1-19 and R1-19B2, consisted of two adjacent PstI fragments that encode the RepA replication system of plasmid R1. The other type contained a segment of R1, designated the RepD replication region, that is adjacent to the RepA region and that has not been identified previously as having the capacity for autonomous replication. Plasmid R1, therefore, contained two distinct deoxyribonucleic acid segments capable of autonomous replication. The RepA-RepD miniplasmid pKT300 had a copy number about eightfold higher than that of R1 and hence lacked a determinant for the regulation of plasmid copy number. Like R1, it was maintained stably in dividing bacteria. RepA miniplasmids had copy numbers which were two- to fourfold higher than that of R1 (i.e., which were lower than that of pKT300) and were maintained slightly less stably than those of pKT300 and R1. The RepD miniplasmid was not maintained stably in dividing bacteria. Previous experiments have shown that incompatibility of IncFII group plasmids is specified by a plasmid copy control gene. Despite the fact that RepA miniplasmids of R1 were defective in copy control, they nevertheless expressed incompatibility. This suggests that two genes are responsible for plasmid copy control, one that specifies incompatibility and is located on RepA miniplasmids and another that is located outside of, but adjacent to, the RepA replication region. Hybrid plasmids composed of pBR322 and one PstI fragment from the RepA region, P-8, exhibited incompatibility towards R2 and RepA miniplasmids but not the RepD miniplasmid, whereas hybrids composed of pBR322 and the PstI fragment of the RepD region, P-3, exhibited incompatibility towards R1 and the RepD miniplasmid but not RepA miniplasmids. These results indicate that the two replication systems are functionally distinct and that, although the RepA system is the principal replication system of R1, the RepD system also plays a role in the maintenance of this plasmid.