Novel Knowledge of Macrolide Resistance in Mycoplasma pneumoniae by Azithromycin Exposure

1. Introduction

Mycoplasma pneumoniae is the causative agent in community-acquired pneumonia, especially in children and young adults [1]. Macrolides are the first-line treatment for respiratory tract infections caused by M. pneumoniae [2]. However, macrolide-resistant M. pneumoniae (MRMP) was detected for the first time in 2001 in Japan [3]. The MRMP rate has been increased in many countries, especially in Asia [4]. The mechanism of macrolide resistance is the point mutation in domain V of the 23S rRNA sequence, and positions 2063, 2064, and 2067 are the main mutation sites [5]. Among these, the A2063G transition is the most common, with a high resistance level to 14- and 15-membered macrolides, such as erythromycin (ERY), clarithromycin (CLR), and azithromycin (AZM) [5]. Though the reason MRMP had not been detected until 2000 remains unclear, it was after >40 years since the first macrolide agent, ERY, was launched.

We have previously performed and are currently conducting multicenter collaborative epidemiological studies on M. pneumoniae infections since 2008 [6,7]. Furthermore, we reported that the MRMP rate has been decreasing and that P1 genotypes (type 1 and type 2 and its subtypes), which consist of P1 protein, important for M. pneumoniae to bind to the host epithelial cells, has changed regularly in their dominance approximately in 10 years [8], and the dominant genotype has recently been changed from type 1 to type 2 and its subtypes [9] in Japan.

Thus, we hypothesized that AZM, launched in 2000, may be related to the occurrence of MRMP because of short-term launches to detect MRMP and a long half-life. Thus, this study aimed to analyze more recent strains after P1 dominant genotype in Japan to prevent the MRMP rate from increasing in the future. To the best of our knowledge, this is the first study addressing the isolated occurrence of MRMP through exposure to low-concentration AZM.

2. Materials and Methods

2.1. Ethical Aspects

The study protocol was approved by the Ethics Committee of Kawasaki Medical School, Kurashiki, Japan, on September 8, 2021 (no. 3119-05).

2.2. Sample collection

M. pneumoniae samples used in this study were collected from pediatric patients with acute respiratory tract infections from 74 institutions located in eight areas across Japan (20 institutions in Kyushu, 25 in Chugoku, 3 in Shikoku, 11 in Kinki, 7 in Chubu, 3 in Kanto, 2 in Tohoku, and 3 in Hokkaido) from 2008, before MRMP pandemic in Japan.

2.3. Strains

Among our collected samples, seven macrolide-susceptible M. pneumoniae were selected ensuring that they had no point mutation related to macrolide-resistance and their minimum inhibitory concentrations (MICs) were susceptible with microdilution methods [10], including six strains of P1 type 2 or 2g2 isolated in 2019, and FH of standard strain (Table 1).

The six clinical strains were selected because the MRMP rate was decreasing recently and analyzing more recent isolates is deemed valuable for anticipating the potential re-emergence of MRMP. In addition, there has been a recent shift in the trends of P1 types from type 1 to type 2 or its subtypes. Therefore, type 2 or 2g2 is considered suitable for understanding the current trend.

2.4. Laboratory Tests and Statistical Analysis

These seven strains were obtained by cultivating specimens. The medium used for isolation and determination of the MIC was pleuropneumonia-like organism broth (PPLO; Oxoid, Hampshire, UK) supplemented with 0.5% glucose (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan), 20% mycoplasma supplement G (Oxoid), and 0.0025% phenol red (Sigma-Aldrich, St. Louis, MO, USA).

The MICs of AZM (LKT Labs, Inc., China) for these strains had been already determined with microdilution methods [10]. First, a medium containing 10⁵ to 10⁶ CFU/ml of M. pneumoniae was added to 96-well microplates and incubated at 37 °C for 6–8 days. MIC was defined as the lowest concentration of the antimicrobial agent, wherein the metabolism of the organism was inhibited, evidenced by the lack of a color change in the medium 3 days after the drug-free control first showed a color change.

Next, exposures of low-concentration AZM were performed as follows. As the first exposure, the PPLO broth including AZM at concentrations of 0.002–1.6 mg/mL, which had higher concentrations than these seven strain MICs of AZM was initially arranged. Then, these seven strains were cultivated in the PPLO broth including AZM at concentrations of 0.002–1.6 mg/m for 28 days. When strains were present after cultivations, these were identified in the mutation sites (2063, 2064, and 2617) in domain V of the 23S rRNA of M. pneumoniae using a direct sequencing method [11].

We performed the second exposure of AZM at higher concentrations than the ones that were able to be cultivated during the first AZM exposure for 28 days, and the strains that were able to be cultivated after the second AZM exposure were performed using direct sequencing at sites 2063, 2064, and 2617 in domain V of the 23S rRNA of M. pneumoniae The MICs of isolates, which detected the mutations by the direct sequencing after exposures of low-concentration AZM, were determined using the same way [10].

3. Results

Four out of these seven strains grew after the first AZM exposure in mediums including AZM 0.002–1.6 mg/mL Mutations were not detected in two of those, but the other two strains (M1601 and M1613) detected C2617G and C2617A mutations, respectively (Table 2).

These four strains were cultivated in mediums including AZM 0.16–128 mg/mL as the second exposure of low-concentration. Two out of these four strains grew after the second AZM exposure. Mutations were not detected in two of those after the second AZM exposure did not grow, but the other two strains (M1601 and M1613) grew until medium included 128, 16 mg/mL of AZM, respectively. These two strains were sequenced, and M1601 detected the A2063G mutation, which was changed before second AZM exposure. However, M1613 detected the C2617A mutation, which was not changed before the second AZM exposure (Table 3)

These two strains after the second AZM exposure were determined the MIC of AZM. The MIC of AZM in M1601 and M1613 was 128 and 0.0156 mg/mL, respectively (Table 4).

4. Discussion

Two previous studies reported that to induce the macrolide resistance in M. pneumoniae in vitro, exposure to macrolide agents were needed [3,12]. For example, clinical samples were used but only for ERY as exposure macrolide agent [3] Furthermore, the clinical samples used in their study were isolated for >20 years ago [3]. The other report only used reference strain of M. pneumoniae [12]. In addition, Okazaki et al. mentioned that 7.8% of the EM-sensitive isolates were detected in point mutations in the 23S rRNA of M. pneumoniae after grew in the medium containing 100μg/mL of EM during incubation for 10–28 days. The mutations detected after growth were A2063G, A2064G and A2064C. Pereyre et.al. reported that Macrolide-susceptible reference strain M129 of M. pneumoniae after 23–50 serial passages in subinhibitory concentrations of some kinds of macrolides detected C2611A mutation by ERY and AZM, A2063G mutation by Josamycin. Comparing our findings to these prior reports, our study aligns with the report by Okazaki et al. in detection period of mutations, that is, much shorter than that in Pereyre’s report. This difference may be attributed to the possibility of quasispecies in M. pneumoniae. Quasispecies, commonly used to describe sequence variants in heterogeneous virus populations, in the case of clinical isolates of M. pneumoniae, means they comprise mixed populations of drug-sensitive and drug-resistant molecular mutants. Chan et al. mentioned that 48.2% of clinical samples of M. pneumoniae detected Quasispecies by using pyrosequencing [13]. Thus, it is suggested that some clinical strains of M. pneumoniae which were diagnosed as macrolide-sensitive may already contain populations of macrolide-sensitive and macrolide -resistant molecular mutants. Consequently, the detection periods might be much shorter among clinical isolates than those in reference strain.

Naturally, bacteria are known to have certain mutation rates, and mutations related with antibiotic resistances are no exception [14]. Therefore, it is possible that macrolide resistant strains arise in isolates that already have populations of both macrolide-sensitive and macrolide -resistant molecular mutants after exposures to macrolides. However, MRMP was first detected in Japan in 2001 [3] marking the first instance since the launch of the first macrolide agents, EM in 1955.

Specifically, AZM exhibits a longer half-life and greater distribution into tissues and fluids compared to other macrolides [15,16]. This characteristic suggests that AZM can persist in the body at low concentrations for an extended duration, aligning with our study protocol. Moreover, it appears not to be coincidental that AZM was launched in 2000, immediately before the appearance of MRMP for the first time in 2001. A2063G mutation is the most prevalent point mutation associated with macrolide resistance in M. pneumoniae, as previously mentioned [9]. This may be because, among other mutations, including C2617A or C2617G detected in our study, their MICs of AZM are much lower than that of A2063G. Consequently, they are more susceptible to being eliminated by macrolide antibiotics, even at relatively lower concentrations.

Another novel aspect in our study was the detection of mutations related to macrolide resistance in recent clinical isolates of M. pneumoniae. As previously mentioned, the MRMP rate has been decreasing recently, and the major P1 genotype has shifted from type 1 to type 2 and its subtypes [9]. Since it has not been long since these isolates appeared and became predominant, they have had fewer opportunities for exposure to macrolide agents. Despite the rates of MRMP being much lower among strains of type 2 and its subtypes compared to type 1 [9], there is a possibility that these isolates acquired point mutations related to macrolide resistance through exposure to AZM. Therefore, our results serve as a reminder to exercise caution in antibiotic use to prevent the increase of MRMP.

This study had some limitations. First, only one type of macrolide agent was used. While using other macrolides could be considered, we chose AZM due to its long half-life, making it more likely to induce mutations upon exposure in vivo. Second, the sample size was small, raising uncertainty about how frequently mutations arise in macrolide-susceptible M. pneumoniae upon exposure to AZM. Thus, future studies should include a larger number of strains to address this limitation. Finally, there was uncertainty regarding whether the parent isolates used were originally a mix of macrolide-susceptible M. pneumoniae and MRMP mutants. Confirming this would require specialized methods, such as pyrosequencing [13].

Nevertheless, the crucial point remains that even recent clinical samples of macrolide-susceptible M. pneumoniae are prone to transition into MRMP, including the A2063G mutant with a high resistance mutation, upon exposure to AZM. It is important to note that these limitations do not compromise the conclusions drawn from this study.

5. Conclusions

Our study is the first to identify strains with the dominant A2063G mutation from recent macrolide-susceptible M. pneumoniae using AZM exposure. This highlights the potential emergence of MRMP with the use of macrolides, such as AZM, emphasizing the need for careful antibiotic management.