**3.1 Direct transfer of guest BACs**

As commercially- or laboratory-prepared BAC clones carried no antibiotic resistance markers for *B. subtilis*, the first BAC-BGM required the pre-installation of a counter-selection system to stimulate the integration process (see Fig. 3). In our initial experiments on the integration of mouse genomic DNA carried by BACs (Kaneko et al., 2003, 2005, 2009; Itaya et al., 2000), we observed no structural disorder during the integration process despite the short repeats generally present in the mouse genome (Itaya et al., 2000; Kaneko et al., 2003, 2009). In addition, the BGM stably carried the 25-kbp-long inverted repeats present in the rice chloroplast genome (Itaya et al., 2008). Consequently, we thought that the BGM could accommodate not only very large-sized DNA but also a wide range of sequence variations from other genomes.

Fig. 3. Cloning of BACs carrying a non-marker for *B. subtilis.*  Top: The present counter selection system is shown. The *c*I repressor gene and the neomycin-resistance gene under the Pr prompter result in the positive selection of markerless BACs for integration.

Bottom: BAC clones in the new BAC vectors, p108BGMC or p108BGME, carrying an antibiotics marker for *B. subtilis* can be cloned directly in BAC-BGM.

common cloning locus in a manner reminiscent of the integration by homologous recombination illustrated in Fig 1b. Before the creation of the BAC vector, pBR322 and its derivatives were widely used for various gene-cloning experiments as they offered several advantages, e.g. a small size, a medium-sized copy number, and an ability to carry DNA up to 30 kbp. DNA cloned in pBR322 via the *E. coli* molecular cloning system immediately became guest DNA in the pBR322-based BGM. Integration required only two homologous sequences and appropriate selection markers for the bacterium. The DNA flow from *E. coli* BACs to the BGM vector is shown in Fig. 1b; it is similar to the pBR322-based system but

As commercially- or laboratory-prepared BAC clones carried no antibiotic resistance markers for *B. subtilis*, the first BAC-BGM required the pre-installation of a counter-selection system to stimulate the integration process (see Fig. 3). In our initial experiments on the integration of mouse genomic DNA carried by BACs (Kaneko et al., 2003, 2005, 2009; Itaya et al., 2000), we observed no structural disorder during the integration process despite the short repeats generally present in the mouse genome (Itaya et al., 2000; Kaneko et al., 2003, 2009). In addition, the BGM stably carried the 25-kbp-long inverted repeats present in the rice chloroplast genome (Itaya et al., 2008). Consequently, we thought that the BGM could accommodate not only very

large-sized DNA but also a wide range of sequence variations from other genomes.

Fig. 3. Cloning of BACs carrying a non-marker for *B. subtilis.* 

antibiotics marker for *B. subtilis* can be cloned directly in BAC-BGM.

less BACs for integration.

Top: The present counter selection system is shown. The *c*I repressor gene and the

Bottom: BAC clones in the new BAC vectors, p108BGMC or p108BGME, carrying an

neomycin-resistance gene under the Pr prompter result in the positive selection of marker-

requires major modifications.

**3.1 Direct transfer of guest BACs** 

After BAC clones were regularly used both in *E. coli* and *B. subtilis*, the next step was to elaborate the engineering/manipulation of the DNA. Figure 4 illustrates the design and modification of nucleotide sequences inside the guest DNA. Examples are detailed below; various size ranges in section 3-2, connecting two overlapping BAC clones in 3-3, applying genome techniques developed for *B. subtilis* in 3-4, and the unique preservation of designed BACs in BGM for prolonged storage in the absence of special facilities in 3-5.

Fig. 4. Manipulation of the guest DNA.

A giant mouse genomic region integrated, for example, into BGM can be re-designed to the indicated structures by relying on the high fidelity of homologous recombination in *B. subtilis*.
