How many replicons make a nodule?

Allan Downie, in his commendable News and Views discussion1 of our work2, suggests that a second, non-chromosomal symbiotic replicon could be present in the bacterial Rhizobium species NGR234. So far, our as-yet unpublished work on physical mapping and random sequencing of the NGR234 genome by V. Viprey, C. Freiberg and X. P. has produced no evidence of another plasmid. Further, the electrophoretic methods we used could visualize the twin symbiotic plasmids (relative molecular mass (Mr) 1×109) of R. meliloti, but in the event consistently demonstrated only a single plasmid of Mr3.1×108± 2×107 in NGR234 (refs 3, 4). Thus, although Morrison et al.5 reported that NGR234 contains two plasmids of Mr3×108 and 4.5×108, it seems unlikely that a second plasmid exists. It is true, however, that some nodulation and other symbiotic genes map to the chromosome6. (Incidentally, the fixABC cluster is present on pNGR234a.)