Plant resistance genes (R genes), especially the nucleotide binding site le
ucine-rich repeat (NBS-LRR) family of sequences, have been extensively stud
ied in terms of structural organization, sequence evolution and genome dist
ribution. These studies indicate that NBS-LRR sequences can be split into t
wo related groups that have distinct amino-acid motif organizations, evolut
ionary histories and signal transduction pathways. One NBS-LRR group, chara
cterized by the presence of a Toll/interleukin receptor domain at the amino
terminal end, seems to be absent from the Poaceae. Phylogenetic analysis su
ggests that a small number of NBS-LRR sequences existed among ancient Angio
sperms and that these ancestral sequences diversified after the separation
into distinct taxonomic families. There are probably hundreds, perhaps thou
sands, of NBS-LRR sequences and other types of R gene-like sequences within
a typical plant genome. These sequences frequently reside in 'mega-cluster
s' consisting of smaller clusters with several members each, all localized
within a few million base pairs of one another. The organization of R-gene
clusters highlights a tension between diversifying and conservative selecti
on that may be relevant to gene families that are unrelated to disease resi
stance.