The secondary structures of the eubacterial RNase P RNAs are being elucidated by a phylogenetic comparative approach. Sequences of genes encoding RNase P RNA from each of the recognized subgroups (a, I, y, and 8) of the proteobacteria have now been determined. These sequences allow the refinement, to nearly the base pair level, of the phylogenetic model for RNase P RNA secondary structure. Evolutionary change among the RNase P RNAs was found to occur primarily in four discrete structural domains that are peripheral to a highly conserved core structure. The new sequences were used to examine critically the proposed similarity (C. Guerrier-Takada, N. Lumelsky, and S. Altman, Science 246:1578-1584) between a portion of RNase P RNA and the "exit site" of the 23S rRNA of Escherichia cofi. Phylogenetic comparisons indicate that these sequences are not homologous and that any similarity in the structures is, at best, tenuous.RNase P cleaves leader sequences from precursor tRNAs to generate the mature 5' end of tRNA (see references 1 and 19 for reviews). In the eubacteria Bacillus subtilis and Escherichia coli, RNase P is composed of a small protein (119 amino acids) and a large RNA (-400 nucleotides). Although RNase P functions as a ribonucleoprotein in vivo, the RNA alone is an efficient and accurate catalyst at elevated concentrations of salt in vitro. RNase P contains the only RNA known to act catalytically in vivo, in the sense that each molecule of RNase P RNA acts on many substrate molecules.An accurate model for the structure of eubacterial RNase P RNA is a prerequisite for understanding the mechanism of substrate recognition, catalytic activity, and evolution of this ribozyme. A model for the secondary structure of RNase P RNA in eubacteria was originally derived by phylogenetic comparative analysis of the sequences of genes encoding RNase P RNA from several members of the y subdivision of the proteobacteria and four species of gram-positive eubacteria of the genus Bacillus (12). On the basis of the criteria used for analysis of ribosomal RNA structure (8), covariation of two base pairings within a potential continuous helix was considered proof of the occurrence of that helix. In order to increase the resolution of this model, we have cloned and sequenced the genes for RNase P RNA from a phylogenetically diverse sampling of the following additional proteobacteria: the a proteobacteria Agrobacterium tumefaciens and Rhodospirillum rubrum, the P proteobacterium Alcaligenes eutrophus, the y proteobacterium Chromatium vinosum, and the 8 proteobacterium Desulfovibrio desulfuricans (Fig. 1). Together with the previously obtained gene sequences for RNase P RNA, these analyses complete a sampling of each of the major phylogenetic branches of proteobacteria.* Corresponding author. Nucleic acid extraction. Extracts for the isolation of genomic DNAs were prepared by resuspending the cells in STE (10 mM Tris-Cl [pH 91, 100 mM NaCl, 1 mM EDTA) (15 ml/g of cells [wet weight]), treating them with lysozyme (1 mg/ml) for 5 min ...