We continue to study (1) replication control, (2) partition and (3) plasmid addiction of the low copy number plasmid prophage P1 and their contribution to its remarkable stability. (1) Studies of strand opening at the origin of the P1 replicon established that the bacterial initiator, DnaA, in relieving local superhelical tension, permits the two strands to assume very different structures. The difference may be relevant to the unidirectionality of replication. Studies of how E. coli chaperone proteins are involved in plasmid replication were undertaken with mutants of the plasmid initiator, RepA, selected for chaperone independence. The evidence suggests that the pairing of DNA-bound RepA monomers is decreased in the mutants, resulting in a copy number increase, in strong support of our previous proposal that the pairing of DNA-bound monomers controls copy number. Apparently the pairing domain of RepA is critical for both the proper folding of that protein and its interactions with chaperones. (2) The P1-encoded system that normally assures plasmid partitioning and consequent stabilization was found to cause plasmid destabilization in a number of situations. Critical variables were shown to be the context of the centromere analog parS and the concentration of the ParB protein, variables expected to influence the fraction of parS bound to ParB. We propose that a context that does not lead to appropriate cyclic variations in the affinity of parS for ParB may lead to destabilization rather than stabilization. Mutations in parS, parB and in vector sequences adjacent to parS, each of which suppresses destabilization, have been characterized. (3) Genes of P1 have been identified that encode an antidote/poison protein pair capable of inducing severe withdrawal symptoms in cells that lose P1. Features of this addiction system that are responsible for an excess of antidote over poison during plasmid retention have been elucidated. Genetic evidence has been obtained that the reversal of this ratio following plasmid loss is due to the selective degradation of antidote by ClpXP protease.
