The other possibility is that the BsuM enzyme remained in the cytoplasmic portion of the donor R+ M+ cell in the fusant and that, upon division of the fusants, various quantities of the enzyme were distributed LY2835219 nmr to the progeny cells depending on where the cell division took place. It has been demonstrated that L-form colonies of B. subtilis arise among those of the bacillary form

after PEG-induced cell fusion (Hauser & Karamata, 1992), which indicates that cell division can occur while the fusant is still without the cell wall. As the L-form B. subtilis cell divides by an extrusion–resolution mechanism that is independent of FtsZ (Leaver et al., 2009), it is possible that a similar mechanism was involved in the division of the fused cells that were produced by random collision between the donor and the recipient protoplasts. This may result in incorporation of various parts and quantities of the cytoplasm of the R+ M+ cell into that of the

R− M− protoplast. Thus, upon cell fusion and subsequent cell division, different proportions of the donor and the recipient cell cytoplasms will constitute progeny cells. Cell death by restriction of the recipient chromosome will occur if a higher proportion of the cytoplasm from the R+ M+ donor cell occupies the progeny cell after division. There must be a critical level of the BsuM restriction enzyme under which the recipient DNA is saved from the restriction attack, and this may account for the reduced but significant efficiency of plasmid transfer from the R+ M+ donor to the R− M− recipient cell. We note in this check details respect that the cotransfer efficiencies of pLS32neo and pHV33 from the R+ M+ to R− M− cells were 1/9 to 1/7 of the levels observed for the

fusion between the homologous pairs (see ‘Results’). As the chromosomal DNA in this fraction of the fused cells escaped the restriction attack, it is tempting Glutathione peroxidase to speculate that the cytoplasmic space in the donor R+ M+ cell containing both plasmids but not enough quantities of the BsuM restriction enzyme to destroy the recipient chromosomal DNA corresponds to 1/9 to 1/7 of the space that brings about transfer of both plasmids in the case of homologous pairs. The heterospecific cell fusion between B. subtilis RM125 and either B. stearothermophilus or B. circulans was successful but the efficiencies were relatively low (see ‘Results’). The restriction system(s) in B. stearothermophilus CU21 or B. circulans BM used here has not been studied yet, but it is known that by a rough estimate, most bacteria carry one or more restriction systems (Wilson & Murray, 1991), which may have affected the fusion efficiency in this study. Another possible reason for the low efficiency is that the compositions of the membrane constituents in those bacteria are different from that of B. subtilis RM125, causing inefficient membrane fusion between the heterospecific bacteria.