Mycoplasma pneumoniae infection impairs asthma control in pediatric patients and exacerbates allergic airway inflammation

Abstract

Objective

To analyze the effect of Mycoplasma pneumoniae (M. pneumoniae) infection on asthma control and its underlying characteristic mechanisms, and to provide evidence-based support for the clinical management of asthma control in children with asthma.

Methods

We enrolled children with asthma and performed a retrospective medical record review, M. pneumoniae detection and isolation from respiratory samples, in vitro antibiotic susceptibility testing, and metabolomic sequencing of bacteria. An ovalbumin-induced allergic asthma model was established using BALB/c mice, which were further divided into the M. pneumoniae-infected asthma group and the non-infected asthma group. Bronchoalveolar lavage fluid from the mice was subjected to detection of M. pneumoniae DNA and 16 S rRNA gene sequencing. Lung tissues were processed for hematoxylin–eosin staining to assess inflammatory infiltration. Periodic acid–Schiff staining was used to evaluate mucus hypersecretion. Metabolomic profiling and transcriptomic analysis were also performed.

Results

We included 145 children with asthma in the clinical cohort. The uncontrolled asthma rate in patients with asthma and M. pneumoniae co-infection was significantly higher than that in patients with asthma only. Among the respiratory specimens, 28 M. pneumoniae nucleic acid-positive samples were identified as genotype M4-5-7-2, and all of these strains carried an A2063G mutation in the 23 S rRNA gene. In the murine asthma model, M. pneumoniae infection significantly exacerbated allergic asthma. Specifically, M. pneumoniae infection led to aggravated allergic symptoms, increased airway hyperresponsiveness, elevated serum immunoglobulin E levels, and higher pathological scores in lung tissue as shown by hematoxylin–eosin staining and periodic acid–Schiff staining. Consistently, 16 S rRNA gene sequencing and transcriptomic and metabolomic analyses showed that M. pneumoniae infection was accompanied by changes in respiratory microbiota composition and altered the phosphatidylinositol 3-kinase/protein kinase signaling pathway, as well as variations in butanoate metabolism profiles. Notably, the changes of butanoate metabolism may be correlated with sophorose, which is a metabolite produced by M. pneumoniae.

Conclusion

This study shows that M. pneumoniae infection reduces the level of asthma control in pediatric patients with asthma. In the murine asthma model, M. pneumoniae infection exacerbates allergic airway inflammation, which may be correlated with altered butanoate metabolism induced by the infection.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12931-026-03755-7.

Introduction

Asthma is a heterogeneous disease characterized by chronic airway inflammation, airway hyperresponsiveness, and reversible airflow limitation, affecting approximately 300 million people worldwide, with an increasing incidence [1]. Childhood asthma is a common chronic disease that exerts a widespread effect on children’s health, families, and healthcare systems. Despite progress made in the prevention and control of childhood asthma, more 120 million children worldwide are still affected by asthma [1, 2]. Respiratory tract infections are important triggers for the onset and exacerbation of asthma, and these can greatly affect the disease course and risk of asthma [3, 4]. Among them, Mycoplasma pneumoniae (M. pneumoniae, MP) is a common pathogen that causes community-acquired pneumonia in children [5].

M. pneumoniae requires specific conditions and an in vitro culture period of 2–3 weeks [6]. M. pneumoniae infections are globally distributed, with a cyclic epidemic pattern occurring every 3–5 years. Children aged 5–15 years are the main population that is susceptible to M. pneumoniae [7, 8]. In recent years, the global detection rate of M. pneumoniae has rebounded, and the detection rate of macrolide-resistant Mycoplasma pneumoniae (MRMP) has also increased [9, 10]. From 2021 to 2023, the in vitro antibiotic resistance of M. pneumoniae isolates from Beijing to macrolides antibiotics increased [11]. However, no studies have reported the characteristics of M. pneumoniae isolates from children with asthma.

The incidence of M. pneumoniae infection in children with asthma and its effect on disease progression have become hot topics in the field of respiratory disease. Epidemiological data have shown that the detection rate of M. pneumoniae in school-age children with asthma is 13%–15%, which is higher than that in non-asthmatic populations [12, 13]. Studies have shown that M. pneumoniae infection accounts for 20% of pneumonia cases, and the risk of acute asthma exacerbation increases after infection [14, 15]. Regarding clinical outcomes, patients with asthma and M. pneumoniae antibodies have more severe conditions, increased airway obstruction, a reduced response to inhaled glucocorticoids, and require higher doses than those without M. pneumoniae antibodies [16]. Long-term follow-up data have shown that within 5 years, the lung function of children with asthma with insufficient treatment after M. pneumoniae infection is lower than that of children without asthma [13]. Therefore, research on the prevention of M. pneumoniae infection in children with asthma and its related mechanisms is important for the control of childhood asthma.

The molecular mechanism underlying M. pneumoniae-induced asthma exacerbation remains unclear. M. pneumoniae is hypothesized to aggravate the course of asthma through multiple pathways. Community-acquired respiratory distress syndrome toxin, the core virulence factor secreted by M. pneumoniae, can directly impair the airway barrier function by inducing vacuolization of epithelial cells. This process further triggers goblet cell metaplasia and basement membrane thickening, ultimately contributing to the development of persistent airway hyperresponsiveness [17, 18]. Additionally, M. pneumoniae releases extracellular vesicles during infection. These extracellular vesicles activate macrophages via the TLR2/NF-κB/JNK signaling pathway, which in turn increases interleukin (IL)-17 secretion, promotes Th2-type immune deviation, and facilitates type 2 inflammation [19].

Few mechanistic studies have focused on children with asthma complicated by M. pneumoniae infection. To further reveal the molecular mechanism underlying M. pneumoniae infection-induced exacerbation of symptoms in children with asthma, we aimed to study children with asthma and M. pneumoniae infection. Using M. pneumoniae-positive samples, multiple-locus variable-number tandem-repeat analysis (MLVA) and macrolide resistance mutation site detection were performed. Additionally, in vitro antibiotic sensitivity testing on isolated M. pneumoniae strains was performed. Furthermore, by comparing pulmonary and systemic inflammatory cytokine levels, coupled with lung tissue metabolomics and transcriptomics analysis in a mouse model of asthma, we aimed to investigate the effect of M. pneumoniae infection on acute asthmatic inflammation and identify the potential involved signaling pathway. This study explains the mechanism by which M. pneumoniae infection exacerbates bronchial asthma from multiple perspectives. Our findings could provide a reference for researchers regarding the interaction between M. pneumoniae and asthma, as well as offering a theoretical basis and technical support for preventing asthma and targeted therapeutic strategies.

Methods

Patient population and sample collection

Asthmatic children who visited the Capital centre for children's health Affiliated to Capital Medical University from July 2024 to March 2025 were enrolled. Inclusion criteria for the asthma group: in line with the diagnostic criteria for asthma specified in the Global Initiative for Asthma (GINA) guidelines, aged 3–18 years, excluded cases with other pathogen infections or chronic respiratory diseases (via multiplex PCR). For the asthma-M. pneumoniae co-infection (asthma-MP) group, in addition to fulling the aforementioned asthma diagnostic criteria, patients were required to have positive M. pneumoniae nucleic acid detection in respiratory specimens, using M. pneumoniae and Macrolide-Resistant Isolates Diagnostic Kit (PCR Fluorescence Probing; Mole BioScience Co., Ltd. Jiangsu, China). Asthma control level was assessed over the 4-week period preceding the date of respiratory specimen collection, using criteria based on GINA guidelines. This study was approved by the Ethics Committee of Capital Institute of Pediatrics (Approval No. SHERLL2023092).

Culture and molecular characteristics analysis of M. pneumoniae

M. pneumoniae-positive samples were inoculated into PPLO Broth Base liquid medium and cultured at 37 °C. Genomic DNA was extracted from respiratory samples (including pharyngeal swabs, sputum, and bronchoalveolar lavage fluid (BALF)) using TIANamp Bacteria DNA Kit (Tiangen Biotech Co., Ltd. Beijing, China). PCR was performed to amplify M. pneumoniae MLVA loci-Mpn13/14/15/16 and the sequence of the 23 S rRNA gene harboring A2063G/A2064G resistance mutation sites [20, 21]. The amplification conditions were pre-denaturation at 94 °C for 3 min, followed by 35 cycles (94 °C for 0.5 min, 55 °C for 0.5 min, 72 °C for 1 min), and 72 °C for 10 min. Amplified products were sent to SinoGenoMax Co., Ltd for sequencing.

Determination of minimum inhibitory concentration (MIC)

The microbroth dilution method was employed to determine the MIC of M. pneumoniae clinical isolates against selected antibiotics. Erythromycin, azithromycin, tetracycline, levofloxacin, and moxifloxacin were serially diluted twofold over a concentration range of 0.125 mg/L to 1024 mg/L. M. pneumoniae standard strain M129 was used as a positive control. Cultures were incubated at 37 °C, and the experiment was terminated upon color change in the positive control. Antimicrobial resistance was interpreted in accordance with the Clinical and Laboratory Standards Institute (CLSI) guidelines [22]. Each isolate was tested in triplicate to ensure result reproducibility.

Establishment of mouse models

Six- to eight-week-old male BALB/c mice (specific pathogen-free grade, Beijing Vital River Laboratory Animal Technology Co., Ltd.) were randomly assigned to 4 groups: control group (treated with 0.9% sodium chloride NaCl solution), asthma group (subjected to ovalbumin-OVA induced sensitization), M. pneumoniae (MP) group (infected via M. pneumoniae nasal instillation), and asthma-M. pneumoniae co-infection (MPA) group (subjected to OVA-induced sensitization combined with M. pneumoniae infection). Asthma model construction: On days 0 and 14, mice were intraperitoneally injected with OVA (20 µg per mouse) and aluminum hydroxide (2 mg per mouse) for sensitization. From days 21 to 23, mice were administered daily nasal instillations of OVA (100 µg per mouse) to induce asthmatic responses. Mice in the MP group and MPA group received daily nasal instillations of M. pneumoniae bacterial suspension (1 × 10⁷ color change units [CCU] per mouse) during the OVA challenge phase (days 21 to 23). Mice were sacrificed on days 1, 3, and 7 post-M. pneumoniae infection for subsequent analysis. All animal experiments were approved by the Ethics Committee of Capital Institute of Pediatrics (Approval No. DWLL2021005).

Non-invasive lung function testing

Airway resistance was evaluated using a BUXCO small animal non-invasive lung function system. Baseline respiratory parameters were recorded continuously for 3 min. Subsequently, mice were exposed to nebulized methacholine (MCH) at escalating concentrations (3.125, 6.25, 12.5, 25, and 50 mg/ml). Each MCH concentration was nebulized for 3 min, followed by a 3-minute recording period to capture the enhanced pause (Penh) and respiratory frequency.

Detection of peripheral blood cell subsets and serum IgE level

Full-automatic blood analyzer (Sysmex XN-1000) was used to determine the proportions of eosinophils, neutrophils, and lymphocytes in peripheral blood. The concentration of immunoglobulin E (IgE) in serum was measured via enzyme-linked immunosorbent assay (ELISA), using a commercial ELISA kit (Wuhan Elabscience Biotechnology Co., Ltd.).

Bronchoalveolar lavage fluid Collection and analysis

Mouse lungs were subjected to bronchoalveolar lavage (BAL) using 1 mL of sterile phosphate-buffered saline (PBS) per lavage, repeated three times. The collected BALF was centrifuged at 1,500 revolutions per minute (rpm) for 10 min to separate the supernatant and cellular pellet. Interleukin-4 (IL-4) and interferon-γ (IFN-γ) concentrations were quantified using ELISA. Following resuspension, the pellet was stained with Wright-Giemsa solution, and differential counting of inflammatory cell populations was performed.

Lung tissue histological staining and scoring

Hematoxylin-eosin (HE) staining and periodic acid-Schiff (PAS) staining were performed on mouse lung tissue sections for histological analysis. HE staining was used to evaluated the degree of inflammatory cell infiltration, scored on a 0–4 scale ( grade 0: no inflammatory cell infiltration; grade 1: occasional scattered inflammatory cells; grade 2: 1–3 layers of inflammatory cells surrounding the bronchus; grade 3: 4–5 layers of inflammatory cells surrounding the bronchus; grade 4: more than 5 layers of inflammatory cells surrounding the bronchus); PAS staining was used to quantify the ratio of goblet cell in bronchial epithelium, scored on a 0–4 scale (grade 0: none; grade 1: <25%; grade 2: 25%-50%; grade 3: 51%-75%; grade 4: >75%) [23].

16 S rRNA sequencing, metabolomic analysis, and transcriptomic sequencing

The 16 S rRNA sequencing of mice BALF was entrusted to Beijing Novogene Co., Ltd. Bacterial genomic DNA was extracted from BALF using TIANamp Bacteria DNA Kit (Tiangen Biotech Co., Ltd. Beijing, China). The V3-V4 regions of 16 S rRNA gene were amplified and were constructed for sequencing on the NovaSeq 6000 platform. Taxonomy was assigned against Silva 138.1 (16 S/18S) and Unite v9.0 (ITS). Alpha diversity and rarefaction curves were calculated in QIIME2. Beta diversity was analyzed using UniFrac distances, with group comparisons by Anosim and Adonis. Biomarkers were identified by LEfSe, and functional predictions were performed using PICRUSt2 and Tax4Fun.

Metabolomic profiling of clinical isolates and lung tissue was performed by Metabo-Profile (Shanghai, China) using Agilent 7890B-5977 A gas chromatography-mass spectrometry (GC-MS) system. Differential metabolites were screened via the XploreMET platform. Metabolites were identified using the JiaLib database. Data were preprocessed with ploreMET and analyzed by PCA, PLS-DA/OPLS-DA (VIP > 1.0), and t-test/Mann-Whitney U test (P < 0.05). KEGG metabolite ratios were calculated to identify biomarkers.

Transcriptomic sequencing of lung tissue was entrusted to Beijing Novogene Co., Ltd. Total RNA was extracted from lung tissues, strand-specific cDNA libraries were constructed. Sequencing was performed on the Illumina NovaSeq 6000 platform. Paired-end sequencing (150 bp) was performed on the Illumina platform. Reads were aligned using HISAT2, quantified as FPKM with featureCounts, and analyzed for differential expression using DESeq2 or edgeR (adjusted P ≤ 0.05). GO and KEGG enrichment were performed using clusterProfiler.

Statistical analysis

Statistical analysis was performed using SPSS software (30.0) and all graphs were generated with Graphpad Prism 9. For categorical data and continuous data not following a normal distribution, the independent-samples Kruskal-Wallis H test (for multiple groups) or Mann-Whitney U test (for two groups) was applied. For continuous data conforming to a normal distribution, the independent-samples t-test (for two groups) or one-way analysis of variance (ANOVA) followed by post-hoc tests (for multiple groups) was used. For two-factor experimental designs, two-way ANOVA was applied. Binary logistic regression was used to adjust for potential confounding factors. P < 0.05 represented statistically significant differences. The significance levels were denoted as follows: * P < 0.05, ** P < 0.01, and ***P < 0.001.

Results

Impaired control in children with asthma following M. pneumoniae infection

A total of 145 pediatric medical records were collected from July 2024 to March 2025, including 45 cases in the asthma-MP group and 100 cases in the asthma-only group. Among them, male patients accounted for 63.4%. The cohort had a mean age of 6.77 ± 3.14 years and a mean weight of 27.27 ± 12.36 kg. A total of 143 outpatient and inpatient records were obtained (2 records were excluded owing to missing data). The average course of the patients in the asthma-only group was 15.2 months compared with 8.3 months in the asthma-MP group. In the asthma-MP group, the incidence of wheezing was 47.7% and that of cough was 81.8%, which were significantly higher than those in the asthma-only group (P = 0.001). No significant differences in the prevalence of other allergic symptoms were observed (Table 1). Regarding the asthma status, 74.7% of children in the asthma-only group achieved good control, which was higher than that in children in the asthma-MP group (50%). The proportion of partial control of asthma in the asthma-MP group was significantly lower than that in the asthma-only group (P = 0.015) (Table S1, Fig. 1a). To adjust for potential confounders, multivariable logistic regression was performed with poor asthma control (partially controlled or uncontrolled) as the dependent variable. After adjusting for age and asthma duration, M. pneumoniae infection remained significantly associated with poorer asthma control (OR = 2.43, 95% CI: 1.05–5.60, P = 0.037). No significant association was observed for age or asthma duration (Table 2).

Table 1.

Clinical characteristics of asthma-only and asthma-M. pneumoniae co-infection patients

symptoms	asthma-only	asthma-MP	P value
asthma course (months)	15.2(4.0 ~ 24.0)	8.3(4.4 ~ 11.8)	< 0.001
wheezing	14(14.1%)	21(47.7%)	< 0.001
cough	39(39.4)	36(81.8%)	< 0.001
other lung symptoms	17(17.2%)	8(18.2%)	0.528
rhinitis	31(31.3%)	10(22.7%)	0.199
urticaria/eczema	6(6.06%)	5(11.4%)	0.22
allergic conjunctivitis	0(0.00%)	1(2.3%)	0.303

Fig. 1.

Comparison of asthma control levels, serum IgE, lung function, blood cell counts, and ICS use between asthma-only and asthma-MP co-infected children. a. Distribution of asthma control levels in children. Asthma control level classification criteria: daytime asthma symptoms >2 times/week; nighttime awakening due to asthma; use of reliever medication >2 times/week; activity limitation caused by asthma. Presence of 3-4 of the above indicates uncontrolled asthma, presence of 1-2 indicates partially controlled asthma, and absence of the above indicates well-controlled asthma. b. Serum IgE levels in children from the asthma-only and asthma-M. pneumoniae co-infection patients. c-g The counts and percentages of each blood cell in the peripheral blood of the patients. h-l Lung function indicators including FEV1, FVC, FEV1/FVC, PEF and FeNO. m Frequency of ICS used by patients per day

Table 2.

Multivariable analysis of the association between M. pneumoniae infection and asthma control

Variable	OR	95% CI	P value
M. pneumoniae infection	2.43	1.05–5.60	0.037
age	0.90	0.79–1.03	0.117
asthma course	1.00	0.98–1.03	0.864

Serum IgE levels were slightly higher in the asthma-MP group than that in the asthma-only group (Fig. 1b). The asthma-MP group also showed a lower eosinophil ratio (P = 0.03; Fig. 1d) and a higher lymphocyte ratio than the asthma-only group (P = 0.016; Fig. 1g). However, there was no significant difference in the eosinophil count (P = 0.07), white blood cell count (P = 0.73), or neutrophil ratio (P = 0.29) between the two groups (Fig. 1c and g).

Pulmonary function assessments showed that the forced expiratory volume in 1 s (FEV1; P = 0.032) and peak expiratory flow (PEF P = 0.030) were significantly lower in the asthma-MP group than in the asthma-only group (Fig. 1h and k). No significant difference was found in the forced vital capacity (FVC), FEV1/FVC ratio, or fractional exhaled nitric oxide (FeNO) levels between the two groups (Fig. 1i and l). Data on maintenance inhaled corticosteroid (ICS) use were collected from the two groups, comprising 117 cases. All patients received fixed-dose combination inhalers with age-standardized unit doses. Daily inhalation frequency was therefore used as the primary indicator of treatment intensity, as the dose per actuation was age-dependent and comparable across age groups. Six patients in the asthma-MP group and nine patients in the asthma-only group were not using ICSs. Approximately half of the enrolled patients used an ICS twice daily. No significant difference in the maintenance ICS dosage was observed between the asthma-only and asthma-MP groups (P = 0.26) (Fig. 1m).

Characteristics of M. pneumoniae related to exacerbation of childhood asthma

In this study, among the respiratory samples of 45 children, 31 were pharyngeal swabs, 1 was sputum, and 13 were BALF. There were 28 M. pneumoniae-positive samples. All (100%) of the positive samples were MLVA type M4-5-7-2, and all had macrolide resistance-related mutations (A2063G). Three M. pneumoniae isolates were isolated and cultured, and they were resistant to erythromycin and azithromycin. The MIC of erythromycin was 128 µg/ml, 256 µg/ml, and 256 µg/ml in these three isolates, and the MIC of azithromycin was 4 µg/ml, 32 µg/ml, and 4 µg/ml, which were consistent with the antibiotic resistance mutation detection. Additionally, these three M. pneumoniae isolates were sensitive to tetracycline, levofloxacin, and moxifloxacin.

Metabolic characteristics associated with exacerbation of childhood asthma

Gas chromatography-mass spectrometry, performed on in vitro cultured M. pneumoniae isolates was used to analyze metabolites of M. pneumoniae related to the exacerbation of childhood asthma. A total of 137 metabolites were identified, and the top 3 most abundant categories were amino acids (69.26%), organic acids (12.33%), and carbohydrates (9.63%) (Fig. 2a). A principal component analysis showed a significant difference in the metabolic profiles between the groups (Fig. 2b and c).

Fig. 2.

Metabolite classification, PCA, KEGG pathway enrichment, and top nine differential metabolites of M. pneumoniae isolates associated with childhood asthma. a Proportion of major metabolite categories in the M129 standard isolate and asthma-related M. pneumoniae isolate (MPA). b-c Principal component analysis (PCA) showing metabolic differences between M. pneumoniae isolates. d KEGG pathway enrichment bubble plot. e Box plot of the top 9 differential metabolites in asthma-related M. pneumoniae

A pathway enrichment analysis showed multiple differentially regulated metabolic pathways (P < 0.05), among which the glyoxylate and dicarboxylate metabolism pathway, beta-alanine metabolism pathway, and butanoate metabolism pathway was significantly enriched (Fig. 2d). Within these pathways, nine key differential metabolites, namely dihydroxyacetone, putrescine, glycolic acid, adenine, ribose, lyxose, glycero-3-phosphate, uracil, and sophorose, were screened out. These metabolites were also identified as potential target metabolites associated with M. pneumoniae-induced asthma exacerbation (Fig. 2e; Fig S1).

M. pneumoniae infection aggravates allergic airway inflammation in mice with asthma

The asthma mouse model with M. pneumoniae infection was successfully established (Fig. 3a). No significant difference in body weight was observed between any of the groups (Fig. 3c). The lung tissue weight (P < 0.001) and lung/body weight ratio (P < 0.05) were significantly higher in the MP group than in the control group. Additionally, lung tissue weight was significantly heavier in the asthma and MP group (P < 0.01) (Figs. 3d-e). During the nasal instillation challenge and nebulization period, the asthma and MPA groups showed conjunctival edema (more prominent in the MPA group), and increased nasal/oral secretions (Fig. 3b). Allergic symptom scores were significantly higher in the asthma group, MPA1 subgroup (1 day post-MP infection, P < 0.01), MPA3 subgroup (3 days post-MP infection, P < 0.001), and MPA7 subgroup (7 days post-MP infection, P < 0.0001) than in the control group and MP-only group (Fig. 3f).

Fig. 3.

Allergic symptoms, body/lung weight, lung function, blood inflammation, IgE, and lung histopathology in OVA-induced asthma with M. pneumoniae infection in mice. a OVA sensitization + M. pneumoniae nasal drip for mouse model construction. b Allergic symptoms in mice during challenge. c-e Mouse body weight, lung tissue weight, and lung weight ratio. f Allergic symptom scoring criteria: 0 points: No symptoms; 1 point: Scratching nose and head; 2 points: Swelling around eyes/nose, piloerection, reduced activity; 3 points: Dyspnea, perioral cyanosis; 4 points: No movement after needle prick/tremors/convulsions; 5 points: Death. g Waveforms of respiratory volume and airflow velocity in mice. h Lung function parameters (Penh, Frequency) curves under increasing MCH concentrations. i Routine blood test indicators. j Serum IgE levels measured by ELISA. k-m Lung tissue HE staining, PAS staining images, and pathological scores

Regarding respiratory waveforms and pulmonary function parameters, the Penh value of the asthma group was significantly higher than that of the control group (P < 0.001), which indicated the successful establishment of the mouse asthma model. At different post-infection time points, Penh values were significantly higher in the MPA1 (P < 0.001), MPA3 (P < 0.001), and MPA7 (P < 0.05) subgroups than in the control group. The respiratory frequency showed a trend to decrease with an increase in MCH concentrations in all of the groups. From baseline, the respiratory frequency in the MPA group was significantly lower than that in other groups (P < 0.001), with the difference remaining highly significant as the MCH concentration increased (P < 0.0001) (Figs. 3h).

The proportions of neutrophils, lymphocytes, and monocytes in the MPA3 subgroup were significantly different from those in the other groups (Control, asthma-only, MP-only, and MPA1 and MPA7 subgroups). In contrast, there was no significant difference in the proportion of eosinophils, absolute eosinophil count, or white blood cell count between the groups (Figs. 3i). Serum IgE levels were significantly higher in the asthma-only group and the MPA7 subgroup than in the control group (P < 0.05) (Fig. 3j).

There was positive detection of M. pneumoniae in BALF in the MP-only group and MPA group, which indicated successful M. pneumoniae infection in these groups. Gross examination of lung tissues showed obvious inflammatory changes in the MPA group, characterized by visible inflammatory infiltration, which was consistent with HE-stained lung tissue sections (Fig. 3k). HE staining scores (reflecting the degree of inflammatory cell infiltration) were significantly higher in the asthma-only group (P < 0.05), MPA1 subgroup (P < 0.01), and MPA3 subgroup (P < 0.001) than in the control group (Fig. 3l). PAS staining of lung sections from the MPA group showed prominent goblet cells aggregation around the airway, with a markedly high proportion of goblet cells (Fig. 3k). Among these groups, the MPA3 subgroup showed the highest PAS score (indicating the most severe goblet cell hyperplasia), with a significant difference (P < 0.0001) (Fig. 3m).

Changes in respiratory flora abundance in asthma with co-infection of M. pneumoniae

We performed 16 S rRNA gene sequencing to characterize the BALF microbiota of mice, with key findings at the genus level. The high relative abundance of Stenotrophomonas was consistent across all groups (Figure S2). Given that BALB/c mice have been reported to be non-permissive for Stenotrophomonas colonization, this finding may reflect background contamination rather than true biological colonization. Therefore, the 16 S rRNA sequencing results should be interpreted with caution, and no causal claims regarding the respiratory microbiota can be drawn from these data. Nevertheless, this did not affect the key finding that Mycoplasma was exclusively detected in the MPA1, MPA3, and MPA7 subgroups, which indicated M. pneumoniae colonization in the co-infected groups, with the highest abundance in the MPA7 subgroup. The abundance of Klebsiella in the MPA7 subgroup was significantly higher than that in the asthma group (P < 0.05). Prevotella and Rodentibacter showed a comparable abundance distribution pattern. Alterations in the microbial community structure were also reflected in Bacteroides, as well as in probiotic genera, such as Lactobacillus, of which its abundance in the lungs of the MPA group was lower than that in the control group (Figure S3).

Metabolomic profiling of lung tissue in mice with asthma and in those with asthma and co-infection of M. pneumoniae

A total of 151 metabolites were identified in the metabolomic analysis. In the asthma-only group and MPA group, the top three most abundant metabolite categories were consistent with minor differences. In the asthma group, amino acids, inorganic oxides, and organic acids accounted for 30%, 15%, and 12% compared with 28%, 15%, and 13% in the MPA group, respectively (Fig. 4a). The principal component analysis showed a trend of separation between the MPA group and the other groups. However, this separation did not reach significance (P = 0.177) (Fig. 4b).

Fig. 4.

Metabolomic analysis of mouse lung tissue. a Proportion of major metabolite categories in lung tissues of control, Asthma, and MPA mice. b PCA analysis illustrating metabolomic differences among the three groups. c Box plot of screened significantly differential metabolites. d KEGG pathway enrichment bubble plot. e Heat map of the combined analysis of metabolites in the lung tissue of MPA mice and metabolites of M. pneumoniae strain M129. Correlation analysis was performed using three independent biological replicates per group

Five significant differentially expressed metabolites were screened, namely 3-hydroxybutyric acid (3-HB), 2,5-diiodopyridine, uracil, glycolic acid, and fumaric acid (Fig. 4c). Two metabolic pathways, the arginine biosynthesis pathway and the butanoate metabolism pathway, were significantly enriched (P < 0.05) (Fig. 4d). Using Pearson relative analysis, we found that the key metabolite in this pathway, 3-HB, was significantly negatively correlated with sophorose, which is a metabolite of M129 (P < 0.01, R > 0.99) (Fig. 4e). This finding suggests an association between M. pneumoniae metabolism and host butanoate pathway suppression.

The PI3K-Akt signaling pathway is significantly enriched in the lungs of asthmatic mice infected with M. pneumoniae

Transcriptomic sequencing was performed on mouse lung tissues. A total of 12,637 co-expressed genes (shared across all experimental groups) were identified (Fig. 5a). A total of 768 differentially expressed genes were detected between the MPA3 subgroup and the asthma-only group, including 327 upregulated genes and 441 downregulated genes (Fig. 5b).

Fig. 5.

Transcriptomic analysis of mouse lung tissue. a Venn diagram of co-expressed gene numbers in transcriptomic analysis. b Volcano plot of differentially expressed genes between groups. c Downregulated KEGG pathway enrichment bubble plot

Compared with the findings in the asthma group, GO analysis of differentially expressed genes in the MPA3 subgroup showed that the top 30 most significantly enriched functional terms primarily involved bioactive lipid receptor activity and receptor ligand activity ( Figure S4). KEGG pathway analysis showed that the top 30 most significantly enriched pathways included the PI3K-Akt signaling pathway (P < 0.0001), FoxO signaling pathway, and mTOR signaling pathway (Fig. 5c, Figure S4).

Western blot experiments showed that the phosphorylation level of Akt (p-Akt) was highest in the MPA group and showed a trend to increase with a prolonged infection time. This trend was consistent with the gray value ratio of phosphorylated PI3K (p-PI3K) to total PI3K across the groups, which indicated activation of the PI3K-Akt pathway in the MPA group (P = 0.03, Fig. 6a and c). The MPA1 and MPA3 subgroups had a higher total inflammatory cell count in BALF than the asthma-only group (Figs. 6d-f). The number of eosinophils was significantly higher in the MPA and asthma groups, with the most prominent increase in the MPA3 and MPA7 subgroups (P < 0.0001) (Fig. 6g). The MPA3 subgroup had the highest number of neutrophils (P < 0.01) (Fig. 6h). No significant difference in the lymphocyte count was observed between the groups (Fig. 6i). IL-4 levels in BALF of the MPA group were significantly higher than those in other groups (P < 0.001), with the highest level in the MPA1 subgroup (Fig. 6j). No significant difference in INF-γ levels was found between the groups (Fig. 6k).

Fig. 6.

Western blot of PI3K/Akt phosphorylation, inflammatory cell counts, and IL-4/IFN-γ levels in BALF in mice. a The expressions of PI3K and p-Akt proteins in the lung tissue of mice were observed by WB. b-c The phosphorylation levels of PI3K and p-Akt proteins between groups were compared by measuring the gray values. d-i Microscopic images of BALF inflammatory cells (Wright-Giemsa staining) and corresponding cell counts. j-k IL-4 and IFN-γ levels in BALF measured by ELISA

Discussion

M. pneumoniae infection is a recognized risk factor for childhood asthma, with a reported wheezing incidence of up to 28.6% and an age-specific pattern [24, 25]. In particular, preschool children are prone to recurrent wheezing post-M. pneumoniae infection and have a high risk of progressing to persistent asthma [15, 26]. In this study, these clinical features were assessed at the time of M. pneumoniae detection, and may partly reflect acute respiratory infection symptoms. Based on the cross-sectional nature of the clinical data, the observed higher prevalence of cough and wheeze in the asthma-MP group than in the asthma-only group may have been confounded by acute infection symptoms. Prospective longitudinal studies are required to examine these possible effects. Additionally, the heterogeneity of respiratory specimen types may have affected the M. pneumoniae detection rate, and the use of pharyngeal swabs, which may have lower sensitivity than BALF, represents a limitation of the clinical study. To minimize bias, we ensured consistent nucleic acid extraction protocols and PCR detection conditions across all specimen types. Previous studies have shown that M. pneumoniae infection can elevate fractional exhaled nitric oxide levels and the peripheral blood eosinophil count [27]. This finding indicates a synergistic interaction between M. pneumoniae infection and type 2 inflammation, which is an important pathogenic driver of allergic asthma. Notably, the present study showed a decrease in eosinophil levels in asthma-MP group. During the acute phase of M. pneumoniae infection, neutrophils and monocytes often dominate the peripheral blood inflammatory response, leading to a relative reduction in the eosinophil count. Some children with asthma included in this study may have been classified as the non-eosinophilic asthma subtype. Patients with this subtype typically show low peripheral blood eosinophil levels and may have distinct inflammatory responses to M. pneumoniae infection compared with patients with eosinophilic asthma [28, 29].

The issue of M. pneumoniae drug resistance has become increasingly prominent since the widespread clinical use of macrolide antibiotics in the 1990s. Macrolides were originally the first choice for M. pneumoniae infection, but current global surveillance data (as of 2025) have shown considerable regional disparities in the prevalence of macrolide-resistant M. pneumoniae. In Asian regions, mutations in the V region of 23 S rRNA (primarily A2063G and A2064G mutations) are highly prevalent. Specifically, the MRMP detection rate has reached 93.7% in China and 89.4% in Japan. In Europe and the United States, the MRMP detection rates remain low (2.02% and 2.4%, respectively). However, the transmission rate of MRMP has accelerated since the coronavirus disease 2019 pandemic [30, 31]. Clinically, children with asthma infected with MRMP show worse outcomes, with a prolonged fever duration and a higher incidence of lung consolidation [32, 33]. This finding suggests that antibiotic resistance gene mutations may reduce the efficacy of macrolide treatment, thereby prolonging the M. pneumoniae infection cycle. The persistent infection, in turn, indirectly exacerbates asthmatic inflammation responses.

Molecular epidemiological studies have shown a close association between M. pneumoniae genotypes and macrolide resistance. The MLVA genotype M4-5-7-2 is prevalent in Asian regions, and M. pneumoniae genotypes show a considerable “periodic replacement” phenomenon (i.e., shifts in the dominant genotype over time). A Japanese study showed a striking genotype shift from 2018 to 2023, in which the proportion of the M3-5-6-2 genotype among M. pneumoniae increased from 12.6% to 51.8%. Concurrently, the macrolide resistance rate of this genotype rose from 43.1% to 79.4%, demonstrating a direct correlation between the genotype prevalence and resistance rates [34]. However, a study in Taiwan suggested that the M4-5-7-2 genotype was closely associated with macrolide resistance [35]. Despite the conflicting findings on specific genotypes (M3-5-6-2 vs. M4-5-7-2), both studies collectively indicate that differences in genotypes directly affect the occurrence of macrolide resistance mutations in M. pneumoniae.

Based on the established mouse asthma model with M. pneumoniae infection in this study, M. pneumoniae infection showed a significant time-dependent progression of pathophysiological changes. A mild elevation of inflammation was observed in the MPA1 subgroup. Peak airway resistance and peripheral blood inflammatory indicators were highest in the MPA3 subgroup. Additionally, the most severe allergic symptoms and peak pulmonary allergic inflammation were found in the MPA7 subgroup. Across the clinical course, M. pneumoniae infection consistently exacerbated core asthmatic phenotypes, increased airway hyperresponsiveness (reflected by Penh values), and enhanced goblet cell metaplasia. While Penh is a widely used and validated indicator of airway hyperresponsiveness in murine asthma models, it remains an indirect measure that may be affected by upper airway resistance. Future studies incorporating invasive measurements, such as FlexiVent, are necessary to further validate the physiological changes observed in this study. Additionally, clinical studies have reported that children with prolonged M. pneumoniae infection (> 3 months) have an increase incidence of bronchiectasis [36, 37]. Especially in children infected with MRMP, impairment of lung function persists, a reduced FEV1/FVC ratio is detectable 6 months post-infection, and small airway obstruction may still be observed 1–2 years later [38, 39]. These findings indicate that M. pneumoniae infection may drive airway remodeling in patients with asthma, transforming reversible airflow limitation into irreversible persistent dysfunction. This possibility highlights the urgency of preventing M. pneumoniae infection in children with asthma and ensuring timely treatment.

The metabolomic analysis of lung tissue from mice in the MPA group showed a significant reduction in 3-HB and butyrate. Studies have shown that butanoate deficiency can lead to a decrease in the proportion of regulatory T cells and an increase in the proportion of Th2 cells [40]. In this pathway, 3-HB links ketone body metabolism with the acetyl-CoA-dominated energy metabolism network and also inhibits NLRP3 inflammasome activation, reducing the release of pro-inflammatory cytokines such as IL-1β [41]. Therefore, we hypothesize that 3-HB and its associated butanoate metabolic pathway have a protective effect in children with asthma, and the suppression of this pathway may lead to exacerbation of asthma. Our integrated analysis of metabolites from M129 and lung tissue of M129-infected mice with asthma showed a significant negative correlation between sophorose, produced by M129 metabolism, and 3-HB, produced by lung tissue. This finding suggests that sophorose production downregulates 3-HB levels in the tissue. Furthermore, clinical isolates metabolically produced even higher levels of sophorose than the M129 strain. We speculate that infection of asthmatic mice with these clinical isolates would yield more pronounced results than using M129, indicating that epidemic strains could pose a greater threat to the health management of children with asthma. However, there is little evidence related to sophorose and inflammation, and further experimental verification is required. While our integrated multi-omics analyses showed a potential correlation between sophorose and reduced 3-HB levels, functional validation, such as sophorose treatment in airway epithelial cells or assessment of butanoate pathway enzyme activity, PI3K-Akt inhibition is also required to establish causality. This represents a key direction for future investigation.

Transcriptomic sequencing and Western blotting experiments showed upregulation of PI3K and Akt in the lung tissue of the MPA group, which indicated activation of this pathway. M. pneumoniae infection upregulates the miR-130b-3p pathway, which enhances PI3K-Akt signaling, activates NF-κB, and promotes IL-8 production ^[45], amplifying pro-inflammatory responses. Inhibiting the PI3K-Akt pathway reduces eosinophil infiltration in lung tissue, lowers IgE levels, decreases serum Th2 cytokine production, and alleviates murine allergic symptoms [42, 43]. This reverse validation confirms the PI3K-Akt pathway’s key role in immune deviation (e.g., Th2 polarization) caused by Mycoplasma infection. Notably, research on this pathway involves multiple Mycoplasma species; the specific signal transduction mechanisms unique to M. pneumoniae require further investigation. Subsequent studies will involve an integrated analysis of existing metabolomic and transcriptomic data to identify relevant molecular targets and further determine the pathogenic mechanism of M. pneumoniae in asthmatic mice.

This study still has several inherent limitations that should be noted. Due to the high relative abundance of Stenotrophomonas across all groups and the absence of negative controls in the 16 S rRNA sequencing run, the BALF microbiome results should be interpreted as exploratory only, and no causal or functional claims can be drawn from these data. In addition, the metabolomic, transcriptomic, and Western blot data presented in this study indicate notable correlations between M. pneumoniae infection, changes in butanoate metabolism, activation of the PI3K-Akt pathway, and exacerbation of allergic airway inflammation. However, these data do not directly establish a causal pathway. Functional validation is required to determine causality. Therefore, all mechanistic conclusions in this study should be interpreted as correlational rather than causal.

Conclusion

This study showed that children with asthma and concurrent Mycoplasma pneumoniae infection had a reduced asthma control status and markedly impaired respiratory condition (e.g., decreased FEV1 values and peak expiratory flow) compared with children with asthma without M. pneumoniae infection. Microbiologically, all M. pneumoniae isolates obtained from clinical samples were characterized as the genotype M4-5-7-2, with a 100% mutation rate at the A2063G site of the 23 S rRNA gene, which may contribute to persistent M. pneumoniae infection and subsequent exacerbation of asthmatic inflammation. Mechanistically, our findings suggest that M. pneumoniae infection exacerbate pulmonary allergic inflammation and elicits allergic symptoms are associated with downregulation of butanoate metabolism and upregulation of the PI3K-Akt signaling pathway. Functional validation is required to establish causality.

Abbreviations

ANOVA

one-way analysis of variance

asthma-MP

asthma-M. pneumoniae co-infection

BALF

Bronchoalveolar lavage fluid

ELISA

Enzyme-linked immunosorbent assay

HE

Hematoxylin-eosin

ICS

Inhaled corticosteroids

IFN-γ

Interferon-γ

IgE

Immunoglobulin E

IL-4

Interleukin-4

MIC

Minimum inhibitory concentration

MP

M. pneumoniae

MPA

asthma-M. pneumoniae co-infection

OVA

Ovalbumin

PAS

Periodic acid-Schiff

Penh

enhanced pause

PI3K-Akt

Phosphatidylinositol 3-kinase/protein kinase

Authors’ contributions

CY, XJ, XR have made substantial contributions to conception and design, and involved in drafting the manuscript. YC, XL, AS, BD, HZ, YF, GX, JC, YK, LG, JF, ZF, TF, ZX, ZY, YY, TZ, CL have made substantial contributions to acquisition of data, or analysis and interpretation of data. JY was involved in revising it critically for important intellectual content.

Funding

This work was financially supported by Beijing Natural Science Foundation (L232071 and 7232007), Beijing High-Level Public Health Technical Talent Project (2023-02-08), Beijing Hospitals Authority’s Ascent Plan (DFL20241301), Beijing Municipal Public Welfare Development and Reform Pilot Project for Medical Research Institutes (JYY2023-10), and Research Foundation of Capital Institute of Pediatrics (JHYJ-2023-05).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This research strictly adheres to scientific ethical guidelines. The use of all involved human patient case data was based on prior informed consent and underwent rigorous de-identification to protect patient privacy (Approval No. SHERLL2023092). The related animal experiments strictly comply with the UK’s amended Animals (Scientific Procedures) Act 1986. The experimental protocols were designed to minimize the number of animals used and alleviate their suffering as much as possible, and have received ethical approval for animal experimentation from the relevant institutional committee (Approval No. DWLL2021005). The entire research process upholds a responsible research attitude, respects life, and protects the rights and interests of all participants.

Consent for publication

not applicable.

Competing interests

The authors declare no competing interests.