ABSTRACT
The genetic and phenotypic structure of sympatric populations of wild bacteria traditionally identified as Bacillus subtilis and B. licheniformis was analyzed. Small soil samples were taken from a single, tiny site in the Sonoran Desert of Arizona, USA, to provide a true population analysis, in contrast to many analyses of genetic structure using bacterial strain collections of widely heterogeneous origin. Genetic analyses of isolates used multilocus enzyme electrophoresis, mismatches in restriction fragment length polymorphism, and variants from Southern hybridization with B. subtilis DNA probes. Phenotypic analyses of isolates used the API test system for detection of growth and acid production on specific carbon sources. The two species were distinct both phenotypically and genetically, despite their known potential for genetic exchange in laboratory experiments. Genic and genotypic diversity were high in both species, and only 16% of observed allozyme variants might possibly be common to both species. Hence, there is probably modest genetic exchange, if any, between the species in nature. Clear hierarchies of population-genetic structure were found for both species. Different types of genetic data yield concordant population structures for B. subtilis. For both species, two-locus and multilocus statistical analyses of linkage demonstrated modest to strong disequilibrium at the species level but truly panmictic subunits within each species. The evidence for extensive genetic recombination within these fine-scale subdivisions is unequivocal, indicating that the sexuality of these bacteria can be well expressed in nature. The relation of these results to processes of bacterial evolution and speciation is discussed.
ABSTRACT
Experiments employing both broth and soil cultures demonstrated the capacity for bidirectional genetic exchange between the eubacterial species Bacillus subtilis and Bacillus licheniformis. The process was studied using standard laboratory strains and wild isolates of these species. The genetic exchange in soil occurs spontaneously. The interspecific recombination involved markers for antibiotic resistance and for the use of specific carbon sources (API characters). Hybrids frequently had unstable phenotypes, i.e., lacked a consistent expression of foreign genes over repeated transfer and growth. This instability often involved a "correction" back toward the phenotype of one or the other of the parental species for many differentiating characters; the final phenotype was always that of the more probable or actually known recipient species. This "correction" process is reminiscent of phenomena associated with the instability of artificial fusion protoplasts or noncomplementing diploids of B. subtilis, as well as the merodiploids formed by intergeneric crosses with enteric bacteria. The hybrids observed here must also be diploid, in some manner, because they sequentially express traits of both parental species at rates well above the frequency of mutation. Among the unstable changes in hybrids of the wild strains there was a 3:1 bias in favor of "correction." The dynamics of the hybridization process in soil are described. It appears that the hybrids are formed most rapidly following outgrowth from spores and during the early growth of parental vegetative cell populations. Later on, the hybrids are much less frequent in the soil cultures, suggesting that they are competitively inferior to the parental species. It is argued that the capacity for recombination found between B. subtilis and B. licheniformis could locally erase their distinctness, even though they possess only about 15% DNA sequence homology. Yet they remain distinct in the wild. The methods and results of these experiments prepare the way for detailed studies of the nature of species and species boundaries throughout the genus Bacillus.
