Abstract

Much of the remarkable ability of Rhizobium sp. strain NGR234 to nodulate at least 110 genera of legumes, as well as the nonlegume Parasponia andersonii, stems from the more than 80 different Nod factors it secretes. Except for nodE, nodG, and nodPQ, which are on the chromosome, most Nod factor biosynthesis genes are dispersed over the 536,165-bp symbiotic plasmid, pNGR234a. Mosaic sequences and insertion sequences (ISs) comprise 18% of pNGR234a. Many of them are clustered, and these IS islands divide the replicon into large blocks of functionally related genes. At 6 kb, NGRRS-1 is a striking example: there is one copy on pNGR234a and three others on the chromosome. DNA sequence comparisons of two NGRRS-1 elements identified three types of IS, NGRIS-2, NGRIS-4, and NGRIS-10. Here we show that all four copies of NGRRS-1 probably originated from transposition of NGRIS-4 into a more ancient IS-like sequence, NGRIS-10. Remarkably, all nine copies of NGRIS-4 have transposed into other ISs. It is unclear whether the accumulation of potentially mutagenic sequences in large clusters is due to the nature of the IS involved or to some selection process. Nevertheless, a direct consequence of the preferential targeting of transposons into such IS islands is to minimize the likelihood of disrupting vital functions.