During February 1, 2010–January 31, 2012, a total of 2,898 respiratory specimens were tested for M. pneumoniae and Chlamydophila pneumoniae by using a multiplex testing real-time assay (ProPneumo-1 Assay; Gen-Probe Inc., San Diego, CA, USA). A total of 96 specimens were positive for M. pneumoniae, and 16 specimens were positive for C. pneumoniae. Among M. pneumoniae–positive specimens, 67 (70%) and 29 (30%) were from the upper and lower respiratory tract, respectively. Six (6.0%) specimens were collected from children < 4 years of age, 48 (50%) from persons 5–20 years of age, 19 (20%) from persons 21–40 years of age, 19 (20%) from persons 41–60 years of age, and 23 (24%) from persons >65 years of age. All M. pneumoniae–PCR positive specimens were cultured and 42 (44%) of the 96 primary specimens yielded positive isolates.

Nested PCR amplification and DNA sequencing of the partial 23S rRNA gene were performed to detect mutations at nucleotide positions 2063, 2064, 2067, 2617 in the 23S rRNA gene, which are associated with macrolide resistance (2,8). In addition to macrolide resistance, molecular determinants of fluoroquinolones (gyrA and parC) and tetracycline (16S rRNA) resistance were also analyzed (10,11).

For macrolide resistance, 91 (95%) of 96 specimens were amplified and analyzed for mutations. Mutations that have been associated with macrolide resistance were found in 11 (12.1%) of the 91 specimens (Table 1). Of the 11 isolates with a mutant genotype, 10 (90.9%) contained a mutation at nucleotide position 2063 (A2063G), and 2 (18.2%) specimens had a mutation at position 2064 (A2064G). In 4 isolates, a mixed population of wild type and mutant at position 2063 were identified on sequence chromatograms. One specimen had wild type and co-mutations at positions 2063 and 2064. None of the specimens contained any mutations at positions A2067 or C2617.

In addition to macrolide resistance, molecular determinants of fluoroquinolone and tetracycline resistance in M. pneumoniae were examined. A previous report showed that substitutions at position 99 (83 for Escherichia coli) of gyrA and positions 81, 83, and 87 (78, 80, and 84 for E. coli) of parC were associated with fluoroquinolone resistance (10). In our study, none of the isolates contained any mutations that have been associated with fluoroquinolones resistance. Similarly, amplification and sequencing of 16S rRNA gene regions encompassing the tetracycline binding site did not show any mutations at positions 968 (T968C) and 1193 (G1193A), which have been shown to be associated with tetracycline resistance among M. pneumoniae (11).

Typing of M. pneumoniae isolates (42/96) by amplification and Sanger sequencing of almost the entire P1 adhesion gene was performed by using primer pairs ADH1/2, ADH3/4, and ADH2BF/R, which amplify 3 fragments of ≈2,280, 2,580 and 767 bp, respectively (13). Sequencing reactions were performed in both directions by using the Big Dye Terminator Cycle Sequencing Ready Reaction DNA Sequencing Kit in an ABI 3730 or 3750 automated sequencer (Applied Biosystems, Foster City, CA, USA).

P1 gene sequencing identified 6–11 variable number tandem repeat AGT sequences. Using sequence typing and comparing homology of nucleotide and amino acid sequences to reference sequences, we found that 16 (38%) isolates belonged to type 1 and shared high homology at nucleotide (99.7%–99.9%) and amino acid (99.5%–99.9%) levels. Seven of the P1 type 1 isolates had a point mutation in E179K, and 3 of these isolates also had a second mutation Q1232E compared with the type 1 reference strain (M129). Twenty-six (62%) isolates were characterized as type 2 and were differentiated within 3 variants (Table 2). Among 5 resistant isolates that were typed, 3 belonged to type 1 and 2 belonged to type 2.

In this study, 12.1% of M. pneumoniae–positive specimens contained mutations that are associated with macrolide resistance, and most (90.9%) specimens had a mutation at nt 2063. This finding is not surprising because this mutation has been shown to be predominant among macrolide-resistant M. pneumoniae and has been associated with high-level resistance (erythromycin MIC >64 mg/L) (2,3,14). None of the specimens contained any mutations at positions 2067 or 2617 because mutations at these positions are rare. Previous studies have shown reduced efficacy rate of macrolides for treating infections with M. pneumoniae isolates containing 3 mutations (7,9).

Typing of the P1 gene showed no clear association between macrolide-resistant isolates and specific subtype because 3 of the 5 resistant isolates belonged to type 1 and the remaining 2 belonged to type 2; there was no evidence of clonality. This finding is consistent with reports in which macrolide-resistant M. pneumoniae isolates of both types were found (2,5,8). However, according to a recent study in China, where macrolide-resistant M. pneumoniae is highly prevalent (>90%), most isolates were type 1 (15).

None of the M. pneumoniae isolates contained any gyrA gene mutations associated with fluoroquinolone resistance or any 16S rRNA gene mutations associated with tetracycline resistance. These findings are consistent with those of a study in which none of the isolates were resistant to fluoroquinolones or minocycline (2).