This knowledge is required to evaluate the effect of differential flagellum expression on competition for nodulation. To this end, we obtained site-directed mutants in each of the flagellin-encoding Akt inhibitor gene clusters of B. japonicum, both in the background of the LP 3004 and LP 3008 strains, and tested their competitiveness. Strains are summarized in Supporting Information, Table S1. Bradyrhizobium japonicum was grown in Götz medium (Quelas et al., 2006) or HM salts with 0.1% yeast extract, 0.1%l-arabinose, and 0.1% sodium gluconate (Kanbe et al., 2007). For conjugation, PSY medium (Regensburger & Hennecke, 1983) was used. Swimming assays were performed in Götz agar (0.3% w/v) (Althabegoiti

et al., 2008). Escherichia coli was grown in Luria–Bertani (Sambrook UK-371804 mouse & Russell, 2001). Antibiotics were at the following concentrations (μg mL−1): streptomycin (Sm), 400 (B. japonicum) or 100 (E. coli); spectinomycin (Sp), 200; kanamycin, 150 (B. japonicum) or 25 (E. coli); ampicillin, 200; and gentamicin, 100 (B. japonicum) or 10 (E. coli). Deletion mutants were obtained and checked as described (Quelas et al., 2010) using the primers and plasmids indicated in Table S1. Strains LP 5843 and LP5844 (ΔfliC1-4) carried the nptII cassette in the replacement of bases 6 410 133–6 418 950, thus removing 8817 bp between bll5843 and bll5846 coding regions (Kaneko et al., 2002). Strains LP6865 and LP 6866 (ΔfliCI-II) carried the

Ω-Sm-Sp-interposon between bases 7 560 766 and 7 563 627, thus replacing 2861 bp of bll6865 and bll6866 coding regions. The double mutants LP6543 and LP 6644 had nptII between bases 6 410 133 and 6 418 950

of LP6865 and LP 6866, respectively. Rhizobia grown in liquid HM salts were vortexed for 5 min and centrifuged at 10 000 g for 30 min at 4 °C. The supernatant was incubated with 1.3% polyethylene glycol 6000 and 166 mM PtdIns(3,4)P2 NaCl for 2 h at 4 °C. Afterwards, this suspension was centrifuged at 11 000 g for 40 min at 4 °C and the resulting pellet was resuspended in phosphate-buffered saline. For analysis, the samples were boiled in Laemmli (1970) loading buffer for 10 min and then separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (Laemmli, 1970). Light microscopy was performed using a Nikon Eclipse E 200 microscope. Videos were recorded using a Nikon 518CU digital camera coupled to the microscope. Electron microscopy was performed as described elsewhere (Althabegoiti et al., 2008). Competitiveness was assayed using mixtures of LP 3004 or LP 3008 with the indicated mutant (Fig. S1). Each strain was at a concentration of approximately 106 rhizobia mL−1 in a modified N-free Fåhraeus plant nutrient solution contained in vermiculite pots (Lodeiro et al., 2000a; López-García et al., 2001, 2002). The pots were allowed to drain the excess solution through holes at the bottom to achieve 100% field capacity, and one plantlet was aseptically planted in each pot.