We established two distinguishing mouse models of ACO that exhibited features of asthma and COPD using PPE with OVA and papain, and subjected them to comprehensive analysis. The mouse models were acceptable and clinically suggestive of ACO. They will be useful for the investigation of the pathogenesis of ACO and development of diagnostic markers and therapeutic targets.
Prior experimental models of ACO involved exposure to allergens and cigarette smoke (CS) (19–21), which does not reflect the pathogenesis of ACO and there are remain unmet needs for experimental models of ACO because of some barriers. To set up appropriate animal models that best reflect the pathogenesis of ACO, combination of the most relevant features from both experimental model of asthma and COPD will enable us to make the acceptable animal model for ACO. Exposure to OVA or house dust mites is used to induce airway inflammation and remodeling in asthma models (22). The classical mouse model of COPD makes use of lipopolysaccharide, PPE, elastase, and cigarette smoke extract (CSE) (23). The intraperitoneal injection or inhalation of CSE can induce airway inflammation (24, 25). However, exposure to CSE and allergens does not yield consistent airway inflammation and airway resistance (21, 26–28). In one study, male surfactant protein-D gene deletion in C57BL/6J mice aged 8–10 weeks exposed to OVA and CS was used; type-2 inflammation did not differ between the OVA and OVA + CS groups, and CS-exposed mice failed to show emphysematous changes (19). Thus, further studies are needed to develop a CSE exposure protocol and clarifying the role of CS exposure for development of ACO model.
Our pulmonary disease mouse model required only 3 weeks to establish after the administration of OVA with PPE and papain, and thus reflects early-onset airway inflammation. Emphysema developed in the COPD and ACO models, but macrophage/neutrophil-associated inflammatory cytokines were not consistently elevated in these models. This factor constitutes a limitation of our models, related to the consideration that emphysema in the ACO and COPD models was not developed over a long period or associated with smoking exposure; furthermore, emphysema alone is not representative of COPD. However, our two ACO models might reflect the heterogeneity of ACO. In two nationwide COPD cohorts, the proportions of current smokers in the ACO groups were smaller than or similar to those in the COPD groups (12). Although CS exposure is a risk factor for COPD, approximately 25–40% of COPD cases are not associated with smoking (29). Our ACO models encompassed emphysema, but not airway inflammation consistent with COPD. The reflection of all COPD endotypes in an ACO model is difficult. However, our models are simpler to establish and more reproducible than are those based on CSE exposure.
Type 2 inflammation mediated by IL-4 or IL-13 and eosinophilic inflammation in blood or sputum are markers of an asthmatic endotype (30) and can be suppressed by targeted therapy (31). We observed the elevation of eosinophils in BALF and type 2 cytokine expression in the ACO-a model; indeed, the levels were comparable to or higher than those in the asthma model. Some studies reported NGAL is a promising ACO-specific marker (32, 33). The NGAL level in BALF was higher in both ACO models than in the COPD model, and the level of NGAL in serum was higher in the ACO-a model than in the COPD model. The BALF and serum NGAL levels were similar in the asthma and ACO-a models, suggesting that this biomarker can distinguish ACO from COPD. The blood eosinophil count is predictive of the response to inhaled corticosteroids (ICSs) in COPD (34–36). Since the 2019 GOLD guidelines recommend that ICSs be used in combination with a long-acting bronchodilator for patients with COPD whose blood eosinophil counts ≥ 300/µL (2). However, caution must be used with ICS inhaler treatment of COPD because of the variability in blood eosinophil counts and relevance to exacerbations (37). Moreover, some patients with ACO do not benefit from ICS treatment (38), possibly because of the heterogeneity of ACO itself, such as asthma and COPD.
R was significantly decreased in the COPD model compared with the asthma and ACO models. E was significantly lesser and C was significantly greater in the COPD and ACO-b models than in the asthma model. E and C trends in the ACO-a model were similar to those in the COPD and ACO-b models. R increased with the methacholine concentration in the asthma and both ACO models. Thus, the physiological responses in the two ACO models reflect features of asthma (through AHR) and COPD (through simultaneously increased C and decreased E).
We established two murine ACO models with the pathological features of asthma and COPD and confirmed their physiological characteristics, facilitating further research on airway diseases. However, this study has several limitations. First, because the models were established 3 weeks after the administration of PPE with OVA and papain, they reflect early-onset inflammation. Follow-up studies on the long-term stability and reproducibility of the models are needed. Second, emphysema was induced rapidly and independent of smoke inhalation, and emphysema alone is not representative of COPD. However, our two ACO models reflect the heterogeneity of ACO. Third, although AHR occurred in both ACO models, the magnitude of macrophage/neutrophilic and/or eosinophilic airway inflammation was not analyzed, especially in ACO-b models. Thus, efforts to develop a model that reflects neutrophilic inflammation and the clinical features of COPD beyond emphysema are warranted.