Global spread of multiple aminoglycoside resistance genes.

Emergence of the newly identified 16S rRNA methylases RmtA, RmtB, and ArmA in pathogenic gram-negative bacilli has been a growing concern. ArmA, which had been identified exclusively in Europe, was also found in several gram-negative pathogenic bacilli isolated in Japan, suggesting global dissemination of hazardous multiple aminoglycoside resistance genes.

Multidrug-resistant gram-negative super microbes have been emerging worldwide. Since carbapenems and fluoroquinolones are the last resort against infections caused by gram-negative bacilli (1,2), the proliferation and dissemination of such clinical isolates that produce metallo-β-lactamases and acquire mutations in gyrA and parC genes have become a global threat (3,4). Aminoglycosides, including amikacin and tobramycin, are still potent agents for use against resistant bacilli. One of the most common resistance mechanisms against aminoglycosides is the production of aminoglycoside-modifying enzymes, such as aminoglycoside acetyltransferases, aminoglycoside phosphorylases, and aminoglycoside adenyltransferases (5), which are mainly mediated by transferable large plasmids.

Recently, a series of special methylases that protect microbial 16S rRNA, the main target of aminoglycosides, was identified in several nosocomial pathogens, including Pseudomonas aeruginosa (6), Serratia marcescens (7), and Klebsiella pneumoniae (8). The newly identified 16S rRNA methylases RmtA and RmtB were reported from Japan in 2003 and 2004, respectively (6,7). The gene for ArmA was initially sequenced in Citrobacter freundii isolated in Poland (GenBank accession no. AF550415) and later characterized in K. pneumoniae isolated in France in 2003 (8). In 2004, nosocomial spread of ArmA- or RmtB-producing Escherichia coli and K. pneumoniae was reported from Taiwan (9).

These enzymes are capable of conferring an extraordinary high level of resistance (MIC >512 mg/L) against most clinically important aminoglycosides as was observed among aminoglycoside-producing actinomycetes, suggesting their probable phylogenic relationship with the intrinsic 16S rRNA methylases of actinomycetes (Figure). RmtA shared 82% amino acid identity with RmtB, but the amino acid sequence similarities between 16S rRNA methylases isolated from pathogenic gram-negative microbes and those from aminoglycoside-producing actinomycetes were relatively low (≤33%). From analyses of the genetic environments of genes encoding 16S rRNA methylases, the rmtA gene is likely associated with the mercury-resistant transposon Tn5041 (10); the rmtB gene was found in the flanking region of Tn3-like structure (7). The armA gene was found on a large plasmid which carries a type 1 integron (8) that mediates various gene cassettes responsible for multiple antimicrobial resistance. The structure of these genetic environments implied that the genes for these 16S rRNA methylases are mediated by mobile genetic elements carried by transferable large plasmids (7,8,10). In fact, the rmtA gene was transferred from P. aeruginosa strain AR-2 to an aminoglycoside-susceptible P. aeruginosa strain 105 by conjugation in vitro (6). The rmtB gene was also transferred from S. marcescens S-95 to E. coli by transformation (7). The armA gene was located on a composite transposon Tn1548 (11).

Figure

Phylogenic relationship among the 16S rRNA methylases. Each amino acid sequence was subjected to the analysis referred to the following sources: FmrO, accession no. JN0651; Kmr, accession no. AB164642; GrmA, accession no. M55520; GrmB, accession no. M55521; GrmO, accession no. AY524043; Kan, accession no. AJ414669; Sgm, accession no. A45282; KgmB, accession no. S60108; NbrB, accession no. AF038408; FMRO, Q08325; RmtA, (6); RmtB, (7); ArmA, (8); predicted enoyl-CoA hydratase/carnithine racemase of uncultured marine gamma proteobacterium EBAC20E09, accession no. AAS73112; putative methylase of Chlorobium tepidum, accession no. AAM72273; hypothetical protein of Nanoarchaeum equitans, accession no. AAR39385. The ClustalW program provided by the DNA Data Bank of Japan (http://www.ddbj.nig.ac.jp/search/clustalw-e.html) was used in this study.

Thus, the growing concern was that these newly identified aminoglycoside-resistance genes could easily spread and be further disseminated among the glucose-nonfermentative gram-negative bacilli, including P. aeruginosa and Acinetobacter spp. and the genera belonging to the family Enterobacteriaceae.

The Study

We conducted a preliminary screening of the 16S rRNA methylase-producing bacilli on our gram-negative microbial stock of 2,877 strains isolated from Japanese hospitals within the past several years. Arbekacin, a semisynthetic aminoglycoside belonging to the kanamycin group, requires 2 modifications at the (6´) aminogroup and the (2´´) hydroxyl group for inactivation, so this agent is not inactivated by known plasmid-mediated aminoglycoside-modifying enzymes. Therefore, a high-level arbekacin resistance (MIC >512 mg/L) was used as a marker for screening the 16S rRNA methylase-producing strains. All arbekacin-resistant strains were subjected to polymerase chain reaction (PCR) analysis to detect rmtA, rmtB, or armA, and all strains were PCR positive, except for a strain of Acinetobacter demonstrating a very high level of resistance to arbekacin (MIC 1,024 mg/L). This strain was later shown to produce both aminoglycoside 6´-acetyltransferase and 2´´-adenyltransferase (12), so arbekacin was inactivated in this strain by both 6´-acetylation and 2´´-adenylation. Each PCR primer set was used to detect rmtA and rmtB genes as in our previous reports (6,7). The PCR primers for amplification of armA were newly designed (forward: 5´-AGG TTG TTT CCA TTT CTG AG-3´, reverse: 5´-TCT CTT CCA TTC CCT TCT CC-3´), and the predicted size of the amplicon was 590 bp. These 3 sets of PCR primers were very reliable in detecting rmtA, rmtB, and armA genes, respectively. Each PCR amplicon was then subjected to sequencing analyses on both strands to confirm its nucleotide sequences for detecting mutations in the methylase genes.

As reported in our previous study, rmtA and rmtB genes had been found in P. aeruginosa isolates (6,10) and in 1 strain of S. marcescens (7), respectively. As shown in the Table, 5 P. aeruginosa strains isolated after our previous report (6) were rmtA positive. The rmtB gene was additionally identified in 4 K. pneumoniae, 2 E. coli, and 1 K. oxytoca strains in Japan. To our surprise, the armA gene, which had been found in various gram-negative microbial species belonging to the family Enterobacteriaceae exclusively in Europe as reported by Galimand et al. (13), was also identified in Japan in 1 strain each of E. coli, S. marcescens, and Acinetobacter sp. Notably, the armA and rmtB genes were also recently identified in K. pneumoniae and E. coli in Taiwan (9). Furthermore, the genetic environment of the armA gene found in C. freundii isolated in Poland was similar to that of K. pneumoniae isolated in France. The genetic environments of the armA gene found in the 3 Japanese microbial species, E. coli, S. marcescens, and Acinetobacter sp., (GenBank accession nos. AB116388 and AB117519), were also similar to those found in Europe (GenBank accession nos. AF550415 and AY220558). These findings suggest that the ArmA-producing gram-negative nosocomial microbes that harbor a very similar genetic environment carrying the armA gene have spread globally.

Table

Methylase-producing strains of 16S rRNA identified after previous study (6)

Species and strain	Type	Year of isolation	Hospital	Prefecture
Pseudomonas aeruginosa P122	RmtA	2002	A	Aichi
P. aeruginosa P340	RmtA	2002	B	Gifu
P. aeruginosa 02-386	RmtA	2002	C	Saitama
P. aeruginosa 03-29	RmtA	2003	D	Aichi
P. aeruginosa 03-230	RmtA	2003	E	Shizuoka
Escherichia coli 01-139	RmtB	2001	H	Yamanashi
Klebsiella pneumoniae 01-140	RmtB	2001	H	Yamanashi
Klebsiella oxytoca 01-141	RmtB	2001	H	Yamanashi
K. pneumoniae 01-142	RmtB	2001	H	Yamanashi
E. coli C316	RmtB	2002	F	Hyogo
Serratia marcescens S95	RmtB	2002	G	Kohchi
K. pneumoniae 03-252	RmtB	2003	H	Yamanashi
K. pneumoniae 03-518	RmtB	2003	H	Yamanashi
E. coli C316-2	ArmA	2003	F	Hyogo
S. marcescens ARS8	ArmA	2003	I	Tochigi
Acinetobacter sp. ARS6	ArmA	2003	J	Kanagawa

Conclusions

As described previously, arbekacin still shows a very broad antimicrobial spectrum from gram-positive to gram-negative nosocomial microbes and has been approved solely to treat methicillin-resistant Staphylococcus aureus (MRSA) infections in Japan since 1990 to ensure the prudent use of this agent. The emergence and presence of the 16S rRNA methylase-producing gram-negative bacilli, however, has not been well recognized in Japan to date; arbekacin has not been listed among the antimicrobial agents for daily antimicrobial susceptibility testing of gram-negative microbes.

The use of semisynthetic aminoglycosides, including arbekacin, in Japanese clinical settings for >10 years may have promoted the emergence and dissemination of the 16S rRNA methylase-producing gram-negative microbes in Japan. The large amount of various aminoglycosides used in livestock-farming environments could have also been a selective pressure for the emergence and spread of pathogenic microbes that harbor genetic determinants for the newly identified 16S rRNA methylases, as exemplified by recent isolation of ArmA-producing E. coli from swine in Spain (GenBank accession no. AY522431).

Since acquisition of multidrug resistance against clinically important antimicrobial agents such as carbapenems and fluoroquinolones has been developing rapidly worldwide, the acceleration of even greater aminoglycoside resistance among gram-negative bacilli promises to become an actual clinical concern in the near future, just as vancomycin-resistant enterococci (VRE) did in the 1990s (14). The emergence of gram-positive cocci including MRSA and VRE that acquire the 16S rRNA methylase could also be a grave clinical matter, although fortunately no such hazardous microbes have been identified. Thus, steps must be taken to further block proliferation of these multidrug-resistant gram-negative super microbes, including P. aeruginosa, K. pneumoniae, and Acinetobacter spp., as well as multidrug-resistant cocci such as MRSA and VRE, which have acquired an extraordinarily high level of resistance to various aminoglycosides through production of 16S rRNA methylases, especially in clinical environments.