BlnI or AvrII (5'-CCTAGG) sites are very rare in the Salmonella typhimurium LT2 genome. BlnI was used to construct a physical map which was correlated with the genetic map by using three methods. First, TnlO carries BlnI sites, and the extra restriction sites produced by 34 genetically mapped TnlO insertions were physically mapped by using pulsed-field gel electrophoresis. Second, six genetically mapped Mud-P22 prophage insertions were used to assign BlnI fragments. Integration of Mud-P22 introduces 30 kb of DNA that can easily be detected by a "shift up" in all but the largest BlnI fragments. Finally, induced Mud-P22 insertions package more than 100 kb of genomic DNA adjacent to one side of the insertion. Some of the smaller BlnI fragments were localized by hybridization to a dot blot array of 52 lysates from induced Mud-P22 insertions. Of the 10 BlnI sites mapped, 6 probably occur in or near the 16S rRNA genes at about 55, 71, 83, 86, 88.5, and 89.5 A number of restriction maps of bacterial genomes have been constructed by using endonucleases that cleave infrequently (22) and then separating the resulting fragments by pulsed-field gel electrophoresis (PFGE) (2,5,7,8,10,17,19,20,(25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(36)(37)(38)(39). The physical order of these fragments has usually been determined by using Southern blotting with genetically mapped probes such as cloned genes (34) or transposons (38) or by hybridization of fragments from one restriction digest to fragments from a different digest (1, 4). An alternative strategy employs a genetically mapped transposon integration that carries a rare restriction site that can be used to cleave and thus locate a restriction fragment in the genetic map (22,35). In this paper, we took advantage of the serendipitous occurrence of two BlnI sites in the transposon TnlO. BinI sites are very rare in enterobacterial genomes (22). We correlated genetically mapped TnlO insertions with the BlnI physical map of Salmonella typhimurium LT2. In addition, Mud-P22 prophage insertions (42), which lack BlnI sites, introduce 30 kb and were used to monitor changes in the length of BlnI fragments. This correlated genetically mapped Mud-P22 insertions with the BlnI physical map.While the methods mentioned above can quickly locate the largest restriction fragments, smaller fragments are less likely to contain a transposon integration or to be detected by probing with cloned genes. The position of a small fragment is likely to be only crudely mapped by blotting to pulsed-field digests. We had available a panel of 52 genetically mapped Mud-P22 integrations. These can be induced to amplify and package 100 kb or more in one direction from the site of integration (42). We used these induced lysates to produce a dot blot array which could then be probed with the smaller BlnI restriction fragments.
MATERUILS AND METHODSPhage and bacterial strains. Most of the strains used are listed in Tables 2 and 3. Strain MS1017 (12) lacks the * Corresponding author. mitomycin C-inducible Fels-2 prophage...