We isolated mutant strains of the facultative anaerobe Salmonella typhinuinum that grow either aerobically or anaerobically. Strict anaerobic mutants contained a defective DNA topoisomerase I gene (topi), while strict aerobic mutants contained a defective DNA gyrase subunit A gene (gyrA, also naLA). Topoisomerase I activity was detected in cell-free extracts of strict aerobic mutants but not of strict anaerobic mutant strains, whereas gyrase activity was detected in extracts of strict anaerobic mutants but not of strict aerobic mutants. Furthermore, extracts of wild-type cells, cultured under vigorous aerobic condition, contain topoisomerase I activity but no significant gyrase activity. In contrast, the extracts of anaerobically cultured wild-type cells contain gyrase activity but no significant topoisomerase I activity. Sucrose gradient centrifugation with ethidium bromide showed that chromosomal DNA in strict aerobic mutants and aerobically grown wild-type cells was relaxed, while the chromosomal DNA of strict anaerobic mutants and anaerobically grown wild-type cells was more supercoiled. Aerobic cultures of wild type and strict aerobic mutants produced both superoxide dismutase and catalase, whereas anaerobic cultures of wild type and strict anaerobic mutants did not. These results lead us to conclude that activity of topoisomerase I, associated with relaxation of chromosomal DNA, is necessary for expression of genes required for aerobic growth, whereas activity of gyrase, associated with supercoiling of chromosomal DNA, is necessary for expression of genes required for anaerobic growth.

No abstract

A prolonged incubation of Escherichia, Salmonella or Pseudomonas at 48 degrees C with nalidixic acid selected mutants (T48) able to grow at 48 degrees C. A prolonged incubation at 54 degrees C of the T48 mutants selected mutants (T54) able to grow at 54 degrees C. These mutants were susceptible to the same bacteriophages as the original mesophilic strains. Auxotrophic phenotypes of Escherichia coli and Salmonella typhimurium mesophilic parents were demonstrated by these mutants if they were cultivated on minimal agar with cellobiose at 48 degrees C or 54 degrees C or on a minimal agar with glucose at 37 degrees C. The T48 alleles mapped in the gyrA region of E. coli or S. typhimurium chromosome. In S. typhimurium the T54 alleles, which permit growth at 54 degrees C, were shown by cotransductional analysis to be linked to gyrA.

~ ~ ~~Thermophilic mutants were isolated from mesophilic Bacillus subtilis and Bacillus pumilus by plating large numbers of cells and incubating them for several days at a temperature about 10 "C above the upper growth temperature limit for the parent mesophiles. Under these conditions we found thermophilic mutant strains that were able to grow at temperatures between 50 "C and 70 "C at a frequency of less than 10-'O. The persistence of auxotrophic and antibiotic resistance markers in the thermophilic mutants confirmed their mesophilic origin. Transformation of genetic markers between thermophilic mutants and mesophilic parents was demonstrated at frequencies of to lo-' for single markers and about lo-' for two unlinked markers. With the same procedure we were able to transfer the thermophilic trait from the mutant strains of Bacillus to the mesophilic parental strains at a frequency of about lo-', suggesting that the thermophilic trait is a phenotypic consequence of mutations in two unlinked genes. I N T R O D U C T I O NEvery organism has its own optimal temperature for growth, being either a mesophile, a thermophile or a psychrophile. The factors that dictate the optimum growth temperature have not been established. Identification of genes controlling growth temperatures of these different classes of organisms by genetic means, e.g. transformation, should provide information that will improve our understanding of the mechanisms that determine the range of temperatures over which different organisms can grow. McDonald & Matney (1 963) claimed the transformation of a characteristic which permitted a strain of Bacillus subtilis to grow at temperatures as high as 55 "C to other B. subtilis strains whose upper temperature limit for growth was 50 "C. Furthermore, Friedman & Mojica (1978) and Lindsay & Creaser (1975) both claimed transformation of the thermophilic trait from thermophilic 'Bacillus caldolyticus' or Bacillus stearoth~r~ic~philus to mesophilic B. subtilis. These observations have not been substantiated. The major difficulty in characterizing the thermophilic trait has been a lack of genetic evidence, because the organisms used for these studies were not genetically characterized.In this communication we report the isolation of thermophilic mutants from two genetically marked mesophilic species (B. suhtilis and Bacillus purnilus) and transformation of the thermophilic trait to mesophilic strains. Differences in their gene expression and behaviour over a range of temperatures are discussed. M E T H O D S5nc.torItr. The B. .stihtili,s strains used were 168 DB-1 from Dr Roy Doi, University of Texas, Arlington, Tx., USA : 1 A348 irpC,? rpoBSOO, resislant to 50 1.18 rifampicin ml-' , and I E4(pLS I3), resistant to 50 pg lincomycin ml-'. from the Bacillus Genetic Stock Center, Ohio State University, USA ; and 168 BR I5 I 1j.s-3 trpC2, requiring Iysine and tryptophan. The B. p r t t i i / t / s strains used were Bp203 Arg-, requiring arginine. and wild-type BpB

In a strict anaerobic environment, lack of expression of bacterial recBC function and recA regulatory functions for the SOS repair system is demonstrated by the use of the carcinogenesis-mutagenesis assay and Salmonella phage P22 mutants requiring these host functions for replication. Therefore, we suggest that error-prone repair (SOS repair) is confined to aerobic environments in facultative anaerobes.