Nomenclature of plasmid-mediated 16S rRNA methylases responsible for panaminoglycoside resistance.

Production of 16S rRNA methylase has recently drawn attention as a novel aminoglycoside resistance mechanism in pathogenic gram-negative bacteria (1). It confers very-highlevel resistance to all aminoglycosides that are currently available for parenteral formulation. Six distinct genes, rmtA, rmtB, rmtC, rmtD, armA, and npmA, encoding their respective enzymes have been identified in clinical and veterinary strains from various geographic areas, including East Asia, Europe, and the Americas, since 2003 (1, 10). NpmA is the only enzyme among them that methylates residue A1408, whereas the others methylate residue G1405, both within the aminoacyl site (A site) of the 16S rRNA (7, 10). All six genes are confirmed to be or are likely to locate on plasmids (3, 4, 10, 11, 12, 14). Recent findings also indicate that some of these genes are capable of crossing the barrier between glucose-fermenting and nonfermenting species. For instance, armA has been identified in both members of the family Enterobacteriaceae and in Acinetobacter baumannii (5, 13), and rmtD has been identified in Klebsiella pneumoniae and Pseudomonas aeruginosa (unpublished data). We will likely see an increasing number of reports about this resistance mechanism, including identification of genes encoding new 16S rRNA methylases. Historically, the nomenclature of genes and enzymes for many resistance mechanisms has become complicated and nonsystematic (6). An extreme example is that of aminoglycoside acetyltransferases, where new gene names are arbitrarily assigned from one of the two coexisting nomenclature systems (9). The situation is somewhat better with -lactamases and macrolide resistance genes, due to a registry and guidelines, respectively (http://www.lahey.org/Studies/) (8). To prevent confusion over the nomenclature of 16S rRNA methylases, we would like to propose practical rules for the nomenclature of these enzymes, which shall apply to any relevant enzymes to be identified in the future. Currently, the highest and lowest identities of amino acid sequences among the G1405 16S rRNA methylases are 81.7% between RmtA and RmtB and 25.8% between ArmA and RmtD, respectively (2, 3). On the other hand, identities lower than 10% are observed between the G1405 16S rRNA methylases and the NpmA that methylates A1408 (10) (Table). Thus, we propose the following rules. A gene that has an amino acid identity greater than 95% with the closest known 16S rRNA methylase gene will be assigned a variant number starting from two in the order of the dates on which the sequences are deposited in the GenBank/EBML/DDBJ, e.g., rmtA2 and then rmtA3, analogous to the nomenclature of the qnr genes. A gene that has between 50 and 95% amino acid identity with the closest known 16S rRNA methylase gene will be assigned a new alphabet letter according to the closest existing gene name, e.g., rmtE, rmtF, armB, or armC, provided that the gene is shown to confer a consistent aminoglycoside resistance profile. A gene that has either an amino acid identity of less than 50% with the closest known 16S rRNA methylase gene or that is proven to methylate a new residue of 16S rRNA may be assigned a brand new gene name, like npmA, contingent upon demonstration of 16S rRNA methylation activity of the gene product and attributable resistant phenotype. Data regarding 16S rRNA methylase genes in pathogenic bacteria will be accumulated and provided at the following website http://www.nih.go.jp/niid/16s_database/index.html.