Sequence comparison of release factors and structural model of RFH. A. Multiple alignment of RF sequences from bacteria with three RFs numbered according to the E. coli RF2 sequence. The N- and C-termini of RF1 and RF2 are not present in RFH and are excluded from the alignment. Abbreviations for organisms and gene bank accession numbers for complete genomes are: B.the – Bacteroides thetaiotaomicron VPI-5482 [NC_004663]; C.ace – Clostridium acetobutylicum [NC_003030]; Burk (or B.xen) – Burkholderia xenovorans (JGI, see text); E.car – Erwinia carotovora [NC_004547]; E.col – Escherichia coli CFT073 [NC_004431]; M.deg – Microbulbifer degradans (JGI, see text); P.aer – Pseudomonas aerugenosa [NC_002516]; R sol – Ralstonia solanacearum [NC_003296]; S.fle – Shigella flexneri 2a [NC_004337]; S.typ – Salmonella typhi CT18 [NC_003198]. Conserved residues are highlighted in color. The red color is used for those residues that are conserved in all three RF families. Green is used for residues that are specifically conserved for one type of factor, i.e. 100% conserved in RF1 and never appears in RFH or RF2. The remaining conserved residues are differentially shadowed in grey. The conserved deletion and insertion in RFH is marked in yellow and purple respectively. Boxes mark the occurrence of functionally important sequence motifs: the GG/GP motif contacting position one of the stop codon, the anticodon motifs and the GGQ-motif. Multiple alignment was produced using ClustalW . B. Cartoon representation of the model of RFH colored as in panel A. The model was made using the program Modeller , with pdb-files 2B9M and 2B64 (chain Y) as structural models and the above alignment as input. The figure has been produced using PyMol . Areas corresponding to the peptidyl transferase center (PTC) and mRNA positions are marked in light blue. The GP, GGQ and SXY motifs are marked with arrows. C. Cartoon representation of a superposition of the structural model of RFH (blue), A-site tRNA (green) and Thermus thermophilus RF2 (red). RF2 and tRNA are from pdb-files 2B9M and 1HIX. Only domains not present in RFH are shown from RF2 (residues 1–114 and 320–364, T. term numbering).