Current Japanese clinical practice involves the usage of large amounts of new macrolides such as clarithromycin and roxithromycin for the treatment of diffuse panbronchiolitis, Helicobacter pylori and Mycobacterium avium complex infections. In this study, the phenotypes, genotypes, and macrolide resistance mechanisms of macrolide-inactivating Escherichia coll recovered in Japan from 1996 to 1997, were investigated.The isolation rate of erythromycin A highly-resistant E. coli (MIC > 1600 /xg/ml) in Japan slightly increased from 0.5% in 1986 to 1.2% in 1997. In six macrolide-resistant strains, recovered from the strains collected for this study during 1996 to 1997, the inactivation of macrolide could be detected with or without added ATPin the assay system. The appearance of erythromycin A-inactivating enzyme independent of ATPwas novel from Japanese isolates, and the ]H NMRspectra of oleandomycin hydrolyzed by the three ATP-independent isolates were examined. It was clearly shown that the lactone ring at the position of C-13 was cleaved as 13-H signal in aglycon of oleandomycin upper shifted. These results suggested the first detection of macrolide-lactone ring-hydrolase from clinical isolates in Japan. These results suggested the first detection of an ATP-independent macrolide-hydrolyzing enzyme from Japanese clinical isolates. Substrate specificity of the macrolide-hydrolyzing enzyme was determined with twelve macrolides including the newer membersof this group and it was found that not only erythromycin A but also the new macrolides, such as clarithromycin, roxithromycin, and azithromycin were inactivated. The NMRdata, broad spectrum of activity, and independence of co-enzymesupported our namingof the enzymeas a macrolide esterase. PCRmethodology was employed to detect an ereB-like gene from the 3 isolates producing macrolide esterase, and one of these was subsequently shown to contain both ereB-like and ermB-Wksgenes. It was also clearly shown that the other three isolates, which inactivated macrolide in the presence of ATP, had an mphA-liko gene.

Macrolide 2′-phosphotransferase [MPH(2′)] transfers the γ phosphate of ATP to the 2′-OH group of macrolide antibiotics. The role of aspartic acids in the putative ATP-binding site of MPH(2′)II was investigated through the substitution of alanine for aspartate by site-directed mutagenesis. D200A, D209A, D219A, and D231A mutant strains were unable to inactivate the substrate oleandomycin, while a D227A mutant retained 7% of the activity of the original enzyme.

Macrolide 2P-phosphotransferase (MPH(2P)) catalyzes the transfer of the Q-phosphate of ATP to the 2P-hydroxyl group of macrolide antibiotics. In this study, H198 and H205, conserved in the ATP-binding region motif 1 in the putative amino acid sequence of MPH(2P)II, were replaced by Ala to investigate their role. H205 was also subsequently replaced by Asn. H198A and H205N mutant enzymes retained more than 50% of the specific activity of the original enzyme to substrate oleandomycin. On the other hand, the specific activity of the H205A mutant enzyme was reduced to less than 1% of that of the wild enzyme. The results suggested that H205 is crucial for maintaining the catalytic activity of MPH(2P)II, and Asn can substitute for His at this position.

Macrolide is inactivated with ATP plus crude extract of Escherichia coli producing macrolide 2'-phosphotransferase (MPH(2')), but not by living cells. Therefore, a convenient method for detection of MPH(2') using intact cells is needed. In this report, we determine that the modified lysozyme-DNase-RNase (LDR) method (named ELDR method) is at least one hundred times more sensitive for the detection of MPH(2') activity than the LDR method and, in addition, highly sensitive for the detection of aminoglycoside-modifying enzymes. Therefore, three new MPH(2')-producing strains were found in clinically isolated E. coli in Japan in 1997 by this method. It suggests that MPH(2')-producing E. coli have been spread in Japanese clinical fields.