Structure of methanogen genes

Abstract

An analysis of the structure and presumed regulatory signals in sequenced methanogen genes is presented. The evidence for polycistronic transcriptional units is extended by inclusion of a DNA sequence, cloned from Methanobrevibacter smithii , which precedes the M. smithii pro C gene ( Hamilton and Reeve , 1985a). Comparison of DNA sequences indicates that a consensus core ribosome binding sequence for methanogens would be 5 'AGGTGA and that many of the ribosome binding sequences are, in fact, longer than the core sequence containing 7 bases with the potential to hybridize to 16SrRNA. The majority, but not all, of the sequenced methanogen genes conform to the rule in which RNY codons (R=purine, Y=pyrimidine, N=purine or pyrimidine) occur most frequently in the translated reading frame ( Shepherd 1981,1983). Codon usages by different methanogens reflect the need to accommodate genomes with very different overall %mol G+C contents. Codons such as AUA, AGA, and AGG, rarely used by E. coli are frequently used by methanogens. The dinucleotide, CG, which occurs very infrequently in eucaryotic DNAs ( Subak-Sharpe et al., 1967; Lennon and Faser , 1983; Nussinov , 1984) is also rarely found in methanogen genes. Methylation and demethylation of the cytosine residue in CG sequences has been implicated in regulation of eucaryotic gene expression ( Weisbrod , 1982) and this role used as an argument for the infrequent occurrence of CG-containing sequences. The rarity of CG in the DNA sequences so far obtained from methanogens may indicate a similar regulatory usage of CG-containing sequences in archaebacterial methanogens.