The Influence of Neighboring Base Composition on Substitutions in Plant Chloroplast Coding Sequences

The bias of nucleotide substitutions in noncoding sequences of the plant chloroplast genome is strongly dependent on the composition of the two flanking bases. Substitutions in G+C-rich environments have a strong bias toward transitions, while in A+T-rich environments, transversions outnumber transitions. In this study, the correlation between the composition of neighboring bases and nucleotide substitution bias is examined in a variety of coding sequences in the plant chloroplast genome. The analyses include both multiple sequence alignments, for the rbcL and n&F genes, and a pairwise comparison of rice and maize for 22 chloroplast genes. Substitution bias, measured in terms of transversion proportion, is found to be significantly correlated with the composition of the two neighboring bases, as in noncoding sequences. However, the influence of the two neighboring bases on substitutions in the &CL gene is different than what is observed in n&F and other coding sequences, where context dependency is the same as in noncoding sequences. This difference is examined and appears to be the result of an influence of nucleotides other than the two immediate neighbors on substitution bias in conjunction with a lower overall A+T content in the &CL gene. The results indicate that the synonymous substitution process in the chloroplast genome is strongly context dependent in a complex manner.