Partitioning of plasmid R1 in Escherichia coli: II. Incompatibility properties of the partitioning system

Abstract

A theoretical as well as an experimental study of the effect of the partitioning system on plasmid R1 drd-19 incompatibility was performed. The theoretical numerical analysis (by computer) was based upon the following assumptions: (i) The partitioning ( par) mechanism is independent of the replication ( rep) and replication control ( cop) mechanism. (ii) A par mutation causes random (binomial) distribution of plasmid copies between the daughter cells at cell division. (iii) In the par + case, the plasmid copies are equipartitioned and selected randomly for partitioning. (iv) Selection of plasmid copies for replication is random. (v) Two different replication control systems were considered: Model 1 assumes that the plasmid copy number is set to exactly 2n before cell division, whereas in Model 2 exactly n copies are synthesized per cell per cell cycle. Numerical analysis was performed for the n values 2–8. The result was that in all cases ( par +/ par +, par +/ par, par/par), steady states with respect to copy number distribution within the heteroplasmid population were rapidly (within five or six generations) established, giving constant loss rates. The rate of loss was slightly higher in the par/par case than in the other two. The two replication control models gave almost identical loss rates. In the par +/ par case, the par + plasmid had an advantage over the par plasmid. The experimental approach was to create heteroplasmid populations of Escherichia coli by introducing two genetically marked derivatives of plasmid R1 drd-19 and then follow the reduction in the relative size of this population during exponential growth in LB medium. The loss rate was essentially the same in the par +/ par + and par +/ par combination and slightly higher in the par/par case, suggesting that plasmid incompatibility mainly is caused by the replication and copy number control system. In the par +/ par situation, the par + plasmid had a pronounced advantage over the par plasmid. The par region of plasmid R1 (without the basic replicon) was cloned onto the vector pSF2124. A par (deletion) mutation was not complemented by par +. Plasmid pSF2124, which does not seem to carry a par system of its own, could use the R1 par system, adding to the conclusion that par is independent of rep and cop. Plasmids pSF2124 and R1 drd-19 are completely compatible, whereas plasmid pSF2124 carrying the R1 par system and plasmid R1 drd-19 showed a weak incompatibility although the copy numbers of the two plasmids were not affected in the heteroplasmid cells. Hence, the partitioning system causes incompatibility, but the effect is weak compared to that of the cop system. The result is consistent with some sort of assortment of plasmids before partitioning.