The genome sequence of Bacillus subtilis subsp. spizizenii W23: insights into speciation within the B. subtilis complex and into the history of B. subtilis genetics The genome sequence of Bacillus subtilis subsp. spizizenii W23 has been determined. The sequence strongly suggests that W23 is a direct descendant of B. subtilis ATCC 6633. W23 shares a 3.6 Mb core genome with the intensively studied model organism B. subtilis subsp. subtilis 168, and gene order within this core has been strongly conserved. Additionally, the W23 genome has 157 accessory (that is, non-core) genome segments that are not found in 168, while the 168 genome has 141 segments not found in W23. The distribution of sequences similar to these accessory segments among other genomes of the B. subtilis species complex shows that those sequences having entered into the phylogeny of the complex more recently tend to be larger and more AT-rich than those having entered earlier. A simple model can account for these observations, in which parasitic or symbiotic DNAs are transferred into the genome and then are reduced in size and modified in base composition during speciation.
INTRODUCTIONStrain W23 holds an important place in the history of Bacillus subtilis genetics. The development of B. subtilis as a bacterial model system began over 60 years ago with the pioneering mutagenesis studies of Burkholder & Giles (1947). Strain 168, a tryptophan auxotrophic mutant isolated in those experiments, was subsequently found to be naturally competent for genetic transformation (Spizizen, 1958). This discovery quickly spawned a research community focused on exploring the genetics and development of this Gram-positive sporeformer and on exploiting its biotechnological potential.In addition to strain 168, a handful of other Burkholder and Giles mutants were also preserved (Spizizen, 1984). One of them, strain 23, was a threonine auxotroph that can readily revert to prototrophy (D. R. Zeigler, unpublished results). Strain W23, believed to be one such revertant, was widely viewed as a 'wild-type' counterpart to strain 168. For several years, W23 served as a 'universal donor' and 168 derivatives as 'universal recipients' for strain construction and genetic mapping (Anagnostopoulos & Spizizen, 1961;Nester & Lederberg, 1961). Evidence gradually accumulated, however, that W23 and 168 were fundamentally different in cell wall composition, resident prophages and genome content (Zeigler et al., 2008). It is now clear that 168 is closely related to -and probably a descendant of -the B. subtilis Marburg strain, which is stocked in the American Type Culture Collection (ATCC) as 6051 T and in the National Collection of Industrial, Marine and Food Bacteria (NCIMB) collection as 3610 T (Srivatsan et al., 2008; Zeigler et al., 2008). It is equally clear that W23 must have arisen from an independent isolate, although its exact origin has remained a mystery (Hemphill & Whiteley, 1975).Strain W23 continues to have relevance for the B. subtilis research community. For example, its c...