An exception to the generally conservative nature of plastid gene evol
ution is the gene coding for the beta'' subunit of RNA polymerase, rpo
C2. Previous work by others has shown that maize and rice have an inse
rtion in the coding region of rpoC2, relative to spinach and tobacco.
To assess the distribution of this extra coding sequence, we surveyed
a broad phylogenetic sample comprising 55 species from 17 angiosperm f
amilies by using Southern hybridization. The extra coding sequence is
restricted to the grasses (Poaceae). DNA sequence analysis of 11 speci
es from all five subfamilies within the grass family demonstrates that
the extra sequence in the coding region of rpoC2 is a repetitive arra
y that exhibits more than a twofold increase in nucleotide substitutio
n, as well as a large number of insertion/deletion events, relative to
the adjacent flanking sequences. The structure of the array suggests
that slipped-strand mispairing causes the repeated motifs and adds to
the mechanisms through which the coding sequence of plastid genes are
known to evolve. Phylogenetic analyses based on the sequence data from
grass species support several relationships previously suggested by m
orphological work, but they are ambiguous about broad relationships wi
thin the family.