Figure 2

Analysis of the evolution of cis -elements upstream of matK -2 and matK -3. A) Schematic representation of the genomic region encompassing the matK-2 and matK-3 editing sites. Edited Cs and corresponding codon transitions are shown in blue; other bases and corresponding codons at the editing site are shown in red. Numbers above refer to the nucleotide position relative to the first base of the matK reading frame in Arabidopsis. This sequence interval was used to generate matK alignments. B) Alignment of the sequence interval from -30 to +10 around both matK editing sites. Green = species that shows editing at respective matK site = "editors" (see Additional file 6); red = species with no detectable editing = "non editors" or with no C at editing site. A consensus sequence was generated based on all edited sequences for each site. Deviations from this consensus are marked in white. Sequences are ordered according to phylogenetic position (Soltis et al, 2000). (Eur = eurosids; V = vitaceae; Sax = saxifragales; Car = cayophyllids; Ast = asterids; M = magnolids.) Third-codon positions are marked with asterisks. C) Analysis of sequence conservation in sequences containing a C at the editing site (C-element; blue border) and in sequences without a C (T-element; red border). Sequences from n different genera were aligned and analyzed using the WebLogo software. Note that n includes one species from each genus in the matK trees shown in Additional files 1 and 2, and not just those analyzed in B. Residues exhibiting differential conservation are marked with blue arrows. The two most variable residues are marked with bold arrows. Third-codon positions are marked with asterisks.