A series of mutations of the P1 plasmid prophage that lead to increased copy number was isolated and analyzed. The copy number of the mutants was elevated at least fiveto eightfold relative to wild-type P1, as determined by single-cell resistance to antibiotics, activity of enzymes, content of superhelical DNA, and reassociation kinetics. The copy number of two of the mutants was temperature dependent. Based on dominance tests, the mutants fell into two classes, cis specific and recessive. The

more »

... atter class included a temperature-sensitive copy mutant. The existence of a class of recessive mutants suggests that the replication of the Pl plasmid is negatively regulated. Because of their small size and dispensable nature, plasmids are ideal systems for studying control of DNA replication. The plasmid we have chosen is the prophage Pl, which is unusual in that it is the prophage state of a bacteriophage. When present as a plasmid, Pl exhibits properties similar to those of the conjugative R and F plasmids. These include autonomous replication (8), maintenance at a low number of copies per cell (8), a segregation pattern that rarely results in loss from a lysogenic cell (22; J. R. Scott, unpublished data), and integrative suppression of dnaA mutants of Escherichia coli (2, 4, 16) . Since it has the alternative of replicating in a lytic cycle and of forming infectious phage, Pl has unique advantages for studies of genetic control functions of plasmids. Genetic mapping of most plasmids is time consuming and difficult and usually relies on deletion analysis. Genetic mapping of Pl, on the other hand, is relatively easy because it can be done by lytic phage crosses and because a map with many markers that can be scored by spot tests is currently available (32). The existence of a lytic cycle that results in the release of a burst of several hundred progeny allows the simple analysis of a very large number of recombination events from a sinde cross; a similar simple technique is not avaiable for other plasmids. Deletion mapping ca also be performed with PI, usually by the use of spot tests (3, 23, 32). Defective prophage containing deletions and restriction nucleaset Present address: AmGen, Newbury Park, CA 91320. generated fragments of PI DNA cloned into various vectors are currently available for deletion mapping. A genetic approach to the study of control of DNA replication involves the isolation of mutations which affect plasmid copy number. Such mutations may be in either the host or the plasmid. To test these mutations for complementation, dominance, and effect on incompatibility, cells carrying two differently marked derivatives of the same plasmid (heteroplasmid cells) must be constructed. In the case of plasmids such as Rl or ColEl, conjugation or transformation followed by selection must be used to form heteroplasmid cells. Since selection must be used, early events cannot be studied in these systems. However, P1 can be introduced into Pl-carrying strains by infection, so these early events can be readily followed (4a). Because incompatibility between two wild-type P1 plasmids is strong enough to preclude isolation of colonies carrying both (4a), and we did not know whether Pt copy mutants would show less incompatibility, we used another member of the Y incompatibility group, prophage P7, for our dominance test experiments. P1 and P7 share 90% homology (33) and are mutually incompatible, but incompatibility between Pt and P7 is much less strong than between two Pls or two P7s (4a). This reduced incompatibility aBlows determinaion of dominance relationships. In this paper, we describe the isolation, genetic mapping, copy number determination, and dominnce relationships of a series of P1 mutants altered in copy control.