Plant disease resistance genes encode members of an ancient and diverse protein family within the nucleotide-binding superfamily

The nucleotide binding site (NBS) is a characteristic domain of many plant resistance gene products. An increasing number of NBS-encoding sequences are being identified through gene cloning, PCR amplification with degenerate primers, and genome sequencing projects. The NBS domain was analyzed from 14 known plant resistance genes and more than 400 homologs, representing 26 genera of monocotyledonous, dicotyle-donous and one coniferous species. Two distinct groups of diverse sequences were identified, indicating divergence during evolution and an ancient origin for these sequences. One group was comprised of sequences encoding an N-terminal domain with Toll/Interleukin-1 receptor homology (TIR), including the known resistance genes, N, M, L6, RPP1 and RPP5. Surprisingly, this group was entirely absent from monocot species in searches of both random genomic sequences and large collections of ESTs. A second group contained monocot and dicot sequences, including the known resistance genes, RPS2, RPM1, I2, Mi, Dm3, Pi-B, Xa1, RPP8, RPS5 and Prf. Amino acid signatures in the conserved motifs comprising the NBS domain clearly distinguished these two groups. The Arabidopsis genome is estimated to contain approximately 200 genes that encode related NBS motifs; TIR sequences were more abundant and outnumber non-TIR sequences threefold. The Arabidopsis NBS sequences currently in the databases are located in approximately 21 genomic clusters and 14 isolated loci. NBS-encoding sequences may be more prevalent in rice. The wide distribution of these sequences in the plant kingdom and their prevalence in the Arabidopsis and rice genomes indicate that they are ancient, diverse and common in plants. Sequence inferences suggest that these genes encode a novel class of nucleotide-binding proteins.