Chromosomal insertions of Tn917 in Bacillus subtilis

We describe 46 insertions of the Streptococcus faecalis transposon Tn917 into the chromosome of Bacillus subtilis. These insertion mutations were mapped genetically. Some caused auxotrophic requirements, and others were cryptic. These insertions were scattered around the B. subtilis chromosome. The mutant strains were useful in several ways for mapping and cloning B. subtilis genes and were added to the Bacillus Genetic Stock Center collection. Among the auxotrophic markers were a new serine auxotrophy and deletion-insertions that caused auxotrophy in one case for homoserine and threonine, in another case for uracil and either cysteine or methionine, and in a third case for leucine, isoleucine, and valine.

Youngman et al. (13) have described methods for "shotgunning" the Streptococcus faecalis transposon Tn917 into the Bacillus subtilis chromosome. We used these methods, with some variations, to isolate a group of insertion mutations of various sorts. Some were cryptic; others caused auxotrophic requirements. As a group, these insertion mutations covered most of the B. subtilis chromosome and could be used in several ways for studying the genome of B. subtilis. For example, any insertion mutation can be exchanged for an altered transposon that carries the Escherichia coli plasmid pBR322 within it, and nearby genes can be cloned from such an insertion (10,11). Alternatively, any Tn917 insertion can be replaced by different altered transposon (12) that contains a promoterless E. coli lacZ gene preceded by a good Bacillus ribosome binding site; if a B. subtilis chromosomal promoter causes transcription into the transposon in the proper orientation, P-galactosidase will be expressed. Control of the B. subtilis promoter can then be studied by monitoring P-galactosidase concentrations.
We collected a number of such Tn917 insertions and (with the help of the Cornell University Biological Sciences 487 classes of 1982, 1983, 1984, and 1985) located them on the chromosome map of B. subtilis; in a few cases we characterized them further. Because we think that these insertion mutations will be useful to people researching the B. subtilis genome, we submitted strains carrying the mutations to the Bacillus Genetic Stock Center (Ohio State University, Columbus), which will supply the strains to persons requesting them. We include here a brief description of the strains.

MATERIALS AND METHODS
Bacterial strains and transposon vector. All bacteria except strain W23 were derivatives of B. subtilis 168. Strain W23 was isolated independently (8). Most of the Tn917 insertions were originally isolated in strain CU3601. This strain was constructed as follows. Strain CU1064 metB5 attsp5 (15) was transduced via phage PBS1 to become a carrier of pTV1, the transposon vector described by Youngman et al. (13). The resulting strain, CU3270, was subjected to the transposition regimen described below, and a glutamate auxotroph carrying gltAB::Tn9J7 was isolated. This strain, CU3278, had lost pTV1; its insertion was the source of the transposon in strain CU4132. CU3278 was transduced to Met' by a PBS1 phage grown in CU1065 trpC2 attsPl (15) attspp gltAB+ (pTV1).
All of the Tn917 insertions isolated in CU3601 were eventually transferred by transformation or by PBS1 transduction into strain CU1147 trpC2 (SPf3 c2) (7). The selection was for resistance to the macrolides-lincosamidesstreptogramin B (MLS) antibiotics, a trait of the transposon. CU1147 is lysogenic for the heat-inducible mutant of phage SPP. As far as we know, the strains are all isogenic with CU1147 except for the transposon insertion that each carries.
The transposon vector pTV1 (13) is a plasmid that carries the 5.3-kilobase transposon Tn917, a chloramphenicol resistance marker, and a heat-sensitive replication system; pTV1 can replicate well in B. subtilis at 30 or 37°C, but cannot replicate at 50°C. The transposon encodes resistance to the MLS antibiotics. We used 1 ,g of erythromycin ml-' and 25 ,ug of lincomycin ml-' a5 selective conditions for MLS resistance (MLS9. A subinhibitory level of erythromycin (20 ng ml-1) induces transposition (13). Because we found that low concentrations of mitomycin C also induce transposition, we sometimes also added 10 ng of mitomycin C ml-1.
Selection of transposon insertions. Most of the transposon insertions described below were isolated by the protocol presented here. We grew CU3601 in antibiotic medium no. 3 (Difco Laboratories) supplemented with chloramphenicol (10 ,ug ml-') and erythromycin (20 ng ml-') at 37°C for 18 h with aeration. Transpositions of Tn917 from pTV1 to the bacterial chromosome were selected by plating samples of these cultures on tryptose-blood agar plates (Difco) containing selective concentrations of erythromycin and lincomycin and incubating the samples at 50°C, a temperature that is restrictive for pTV1 replication. The resulting colonies arose from cells in which Tn9O7 had been transposed from the plasmid to the chromosome. We confirmed this by scoring for the loss of the plasmid marker chloramphenicol resistance. The MLSr clones were then tested for their ability to grow on minimal medium containing tryptophan and (because we were not interested in insertions into the gltA and glItB genes) glutamate. The requirements of the auxotrophic mutations were determined with a grid analogous to that described by Davis et  the chromosome by using phage PBS1 transduction (9), DNA transformation (9), and derivatives of the B. subtilis map kit strains (2).
We were able by a second method to isolate Tn917 insertions linked to a known auxotrophic marker. We produced lysates of the generalized transducing phage PBS1 by growing the phage in mixed cultures of bacteria carrying random Tn9J7 insertions on their chromosomes. The lysates were prepared as follows. Several cultures of strain CU3601 were grown for 18 h in antibiotic medium no. 3 containing 20 ng of erythromycin ml-' and 10 ng of mitomycin C ml-'. Each culture was then enriched for transposition events by diluting 1:100 with antibiotic medium no. 3 containing 1 jxg of erythromycin ml-' and 25 gxg of lincomycin ml-' and by incubating the cultures at 50°C overnight with aeration. These cultures were then diluted appropriately with fresh antibiotic medium no. 3, and PBS1 lysates were prepared from them (9). For example, to place transposition insertions near the purA gene of B. subtilis, we used these lysates to transduce a purAJ6 auxotroph simultaneously to prototrophy and to MLSr. Essentially all of the transductants then carried cryptic insertions of Tn9J7 linked to purA.

RESULTS
Observations on transposition events. We produced a strain of B. subtilis, CU3870 trpC2 attsp5, carrying the transposition vector pTV1. When we used 10 ng of erythromycin ml-' to induce transposition as described above, we found that about 1 cell in 104 could form MLS' colonies at 50°C. The frequency of auxotrophs among these colonies was 0.5 to 3% in different experiments. More than 90o of the auxotrophs were Glt-; that is, their requirements could be satisfied by either glutamate or aspartate, and the transposon was located in the gltAB region, very close to the B. subtilis terminus of replication. We replaced the gItAB region (which had originated in B. subtilis 168) with the homologous region from strain W23, as described in Materials and Methods. When the newly constructed strain CU3601 was similarly tested, the frequency of Gltmutants among the auxotrophs was only 25 to 30%. We therefore used CU3601 for the isolation of the insertion mutants described here. We found that a large fraction (30 to 50%o) of the cryptic insertions were linked (sometimes very closely) to gltAB in strain CU3601.
Several of the insertions of Tn9J7 were accompanied by deletions of nearby-presumably adjacent-DNA. Two independently isolated insertions into the ilvBC-leu region (liv-J::Tn9J7 and liv-2::Tn917) resulted in apparently identical deletions of 2.7 kilobases of DNA, including the right end of ilvB and most or all of ilvC, as determined both by our failure to recover transformation recombinants with point mutations in the region and by Southern analysis (manuscript in preparation). The urc-83::Tn917 insertion, which caused auxotrophy for uracil and for either cysteine or methionine, was also accompanied by a deletion. DNA extracted from this strain could not repair point mutations in cysC (with a requirement for cysteine or methionine) or in various pyr genes (with requirements for uracil), all of which were clustered on the B. subtilis chromosome. Similarly, the mth-84::Tn9J7 insertion required threonine and either methionine or homoserine. DNA extracted from a strain carrying this marker could not correct a number of point mutations in the thrA gene, nor could it correct a hom mutation which itself was possibly a deletion.
We did not detect a precise excision of Tn917 from a chromosomal location. We plated more than 109 cells of each e Requires threonine and either homoserine or methionine (see the text). of more than 40 auxotrophic insertion strains to search for revertants, but we found none that had lost the transposon. Most of these auxotrophs did not have detectable deletions.
On the other hand, a few Tn917-generated auxotrophic insertions could revert to prototrophy. The pyr-82::Tn9J7 insertion caused uracil auxotrophy, but prototrophic revertants were recovered at a frequency of 10-5 to 10-6. The revertants retained Tn917 or at least its MLSr marker. We suspect that the transposon lies in a regulatory region of the pyr operon and that mutations or small deletions in the region can suppress the Pyr-phenotype. Similarly, the mth-83: :Tn9J7 insertion resulted in auxotrophy for threonine and either homoserine or methionine. The insertion was very closely linked to thrA and hom mutations. Mutations could be isolated from strains carrying this insertion that were either Met' or Thr+ or else Met' and Thr+; all retained the transposon. Again, we suspect that the insertion lies in a regulatory region, possibly between genes encoding homoserine synthetase and homoserine dehydrogenase. Auxotrophic insertion mutations. Most of the auxotrophic insertion mutations (Table 1) corresponded to the phenotypes described in the most recent edition of the B. subtilis chromosome map (6). A few additional insertion mutations are described below, proceeding clockwise around the chromosome.
Insertion mutations labeled arg-342 (4) have the same phenotype as argO mutations; that is, they require arginine, citrulline, or ornithine. They differ from the argO insertion mutations in their chromosome location. The serA84::Tn9J7 insertion mutation (and several similar VOL. 167, 1986 532 VANDEYAR AND ZAHLER such cells glycine (200 ,ug ml-'), serine (40 ,uag ml-'), and threonine (40 ,ug ml-'). We are not certain that this class of insertion mutations differs from the class labeled glyA on the CU4148 1A628 B. subtilis chromosome map. The liv-3::Tn9J7 insertion mutation lies between the ilvB and the ilvC regions, as determined by transformation map-CU4149 1A629 ping. Strains carrying liv-3::Tn9J7 were prototrophic at 30 and 37°C but were Livat 50°C (manuscript in preparation). The serC insertion mutation caused auxotrophy for serine, CU4150 1A630 which we used at 4 ,g ml-1. The serC gene is located clockwise to the sdh genes and was 98% cotransduced with CU4151 1A631 the latter. Better growth was achieved with 40 ,ug each of serine and threonine ml-'.
Cryptic insertions. Nineteen strains carrying cryptic inser-CU4152 1A632 tions of Tn917 in the B. subtilis chromosome are listed in Table 2. Their genetic map positions were indicated by the method of Hong and Ames (5) as amended by Davis et al. CU4153 1A633 (1). The insertion is designated by three letters, the first of CU4154 1A634 which is z. The second and third letters indicate the approximate map position of an insertion in one-tenth and onehundredth segments, respectively, of the total genetic map, CU4155 1A635 starting at the origin of replication (which equals 0 or 100) and proceeding clockwise. The second letter corresponds to the one-tenth map segments, which are lettered consecu- prophage. We therefore included strain CU2111, received from P. J. Youngman (his strain PY37), which is lysogenic for a derivative of SP1 c2 into which Tn917 has been inserted CU4162 1A642 (Table 2). When CU2111 is induced by heat shock or by mitomycin C, it releases the SP, c2 phage carrying Tn917; the phage particles are capable of forming plaques and of CU4163 1A643 lysogenizing sensitive or lysogenic B. subtilis cells, which they convert to MLSr (14). The transposon is located within  gives the chromosome locations of genes to which the cryptic insertions were linked. ined in the text. All 111 are lysogenic for DISCUSSION (6).
The 47 strains included in this study carry Tn917 inserd-Selection was for tions in most regions of the B. subtilis chromosome. Two major gaps remain. One lies between 11 and 540 on the 3600 ,Sr.  (6). and 12% of the B. subtilis map. We did not search specifically for Tn917 insertions in these regions of the chromosome by the PBS1 linkage method. We have no reason to think that such a search would fail, although there may be clusters of essential genes that would make the search difficult. The region between 182 and 1930 is covered by a transposon inserted into the SP1 prophage.
The strains are useful for cloning nearby genes and for studying the regulation of genes carrying the transposon. For these reasons we deposited the strains in the Bacillus Genetic Stock Center, which will supply cultures upon request. Some of the strains carry transposons that have adjacent deletions associated with them. These include strains CU4131, CU4139, and CU4145. Other strains may have undetected deletions associated with them. We never detected a precise excision of Tn917 which would allow us to return an auxotrophic insertion to the prototrophic state.
We have evidence that Tn917 is polar in one orientation (manuscript in preparation), the orientation in which chromosomal transcription enters the right end of the transposon (12). In the other orientation, readthrough transcription can occur at low temperatures (30 to 37°C), but a rather weak transcription terminator prevents most readthrough transcription at 50°C. The phenotype of CU4140 (liv-3::Tn917), which is prototrophic at 37°C (although its growth is stimulated by branched-chain amino acids) and Livat 50°C (requiring leucine, isoleucine, and valine), is an example of the evidence leading to this conclusion.
Some of the auxotrophic strains included in this study gave rise to prototrophic mutations that retained the transposon in its original insertion site. We did not examine these strains at length, but they may be useful in investigations of the pathways involved. They include strain CU4129 (pyr-82::Tn917) and strain CU4144 (mth-83::Tn917). Because strain CU3601 carries DNA from B. subtilis W23 in its gitAB region, three of the strains in this collection (CU4154, CU4155, and CU4156) may contain DNA of W23 origin.
The Mth (requiring threonine and either homoserine or methionine) and Urc (requiring uracil and either cysteine or methionine) phenotypes of B. subtilis described here have not been reported before. The serC mutation is unique. In our experience, spontaneous Liv mutations do not occur in B. subtilis, and the ath (adenine and thiamine), cym (cysteine or methionine, located near cysA), and ala (L-alanine) auxotrophic mutations are quite rare among mutations induced either by N-methyl-N'-nitro-N-nitrosoguanidine or by 6chloro-9([3-(2-chlorethyl)-amino]propyl)-2-methoxyacridine .