The DM3 plates were incubated at 37 °C for 3–4 days, and the colonies formed were replica plated onto LB plates containing

either Sp and Nm or Sp and Cm to count the number of colonies. The concentrations of the antibiotics used for the DM3 plates were 200, 150, 1, 5, and 10 μg mL−1 for Sp, kanamycin (Km), Em, Cm, and Tc, respectively. We define in this study the SGI-1776 cell line donor and the recipient as the strain that carries plasmids to be transferred and that acquires plasmids during cell fusion, respectively. Discrimination between the donor and the recipient among fusants was made possible by the recA mutation, which prevents recombination between the chromosomal DNAs originating from two host strains, and by selection with antibiotics to which find more the new plasmid carrier (the recipient) shows resistance. Plasmids pLS32neo (Kmr/Nmr) and pHV33 (Cmr) carry eight and no BsuM restriction sites (Table 1; Bron et al., 1988), respectively, which facilitates the examination of the effect of restriction and modification on plasmid transfer from the donor

to the recipient. To investigate the extent to which plasmid transfer by PEG-induced cell fusion is affected by the BsuM restriction, the protoplasts from strains R+ M+recA::Emr or R− M−recA::Emr carrying both pLS32neo and pHV33 (donors) were fused with those from plasmid-free R+ M+recA::Spr or R− M−recA::Spr strains (recipients) in various combinations. Regenerated colonies on DM3 plates Paclitaxel showing

either the Spr Kmr or Spr Cmr phenotype were further checked for resistance to Cmr or Nmr, respectively, to examine co-transfer of pHV33 and pLS32neo. When strain 168 recA::Emr carrying both pLS32neo (Kmr/Nmr) and pHV33 (Cmr) and strain 168 recA::Spr were used as the donor and the recipient, respectively, fusants resistant to either Spr Nmr or Spr Cmr were obtained at similar efficiencies (Table 2, line 1). The same experiment using the RM125 recA strain as both the donor of the plasmids and the recipient gave nearly identical fusion efficiencies (Table 2, line 2). These results show that both the R+ M+ and R− M− strains served as the donor and the recipient for plasmid transfer with similar efficiencies under the experimental conditions used. When the RM125 recA::Emr strain carrying the plasmid pair and strain 168 recA::Spr were used as the donor and the recipient, respectively, the number of Spr Nmr colonies was 0.23% of that obtained by the cross between the R+ M+ strains, whereas the efficiency of Spr Cmr colony formation was unaffected (Table 2, line 3). These results show that pLS32neo carrying eight BsuM restriction sites was subjected to BsuM restriction in the recipient cell, whereas the transfer of pHV33 was unaffected. It was also found that the transfer efficiency of pHV1401, a derivative of pHV33 carrying three BsuM sites (Bron et al., 1988), was reduced to the same level as that of pLS32neo (T. Maehara, M. Itaya, and T.