**3. Discussion**

This study investigated the molecular mechanisms involved in the progressive worsening of BLM-induced pulmonary fibrosis in three genotypes of mice carrying stepwise variations of p38 activity in AEC II. We demonstrated that BLM-induced severe inflammation and fibrosis that was correlated with increased p38 activity in the lungs. Transcriptome analysis of this model provided a connection between the progression of pulmonary fibrosis and genes driving ER functions, ECM-cell interaction, and the immune system. Moreover, we identified candidate genes associated with IPF progression in comparison to a publicly available IPF dataset. The results of these comprehensive analyses sugges<sup>t</sup> that the progression of pulmonary fibrosis occurs concurrently with increased p38 activity in AEC II, which provokes the enhancement of inflammation and immune systems. Therefore, this novel model of pulmonary fibrosis serves as a tool for understanding IPF progression.

Addressing the mechanisms contributing to IPF progression can lead to improved prognosis as this complex multi-pathway disease shows heterogeneity in its clinical course [1,4]. The present study applied a severe model of BLM-induced pulmonary fibrosis to study disease progression. Although the BLM-induced pulmonary fibrosis model is insufficient to mimic the pathogenesis of IPF, it has shown high reproducibility and the important mechanisms of pulmonary fibrosis, such as

epithelial-mesenchymal crosstalk and TGF-β signaling pathway [25,26]. BLM-induced pulmonary fibrosis shows a transition from the inflammatory to the fibrotic phase at around 7 dpi, establishment of fibrosis at 14 dpi, and subsequent formation of reversible lesions [27,28]. In this context, the analyses were mainly conducted at 8 dpi as the optimal timing to evaluate the progression of pulmonary fibrosis. Additionally, the transcriptome profiling approach enables us to reveal the molecular mechanisms regulating fibrosis in this model and compare them to the profiles in the lung samples of IPF patients. Recent studies have shown that variations in gene expression of BLM-induced pulmonary fibrosis were correlated with changes in IPF severity [23,24]. Notably, the genes di fferentially expressed in BLM-induced pulmonary fibrosis are most abundant in the active fibrotic phase (7–14 dpi), which shows the highest correlation with IPF lung samples [23,24]. This finding explains the rationale that the gene expression profiles of BLM-treated lungs are altered before remarkable changes in morphology and function occur. Taken together, these findings are compatible with our study, which has implications for the development of pulmonary fibrosis in transcriptome analysis.

Although apoptosis and reprogramming of lung epithelial cells play a prominent role in IPF, the molecular details remain uncertain [20,29]. p38 is required for maintaining AEC II homeostasis as a physiological function, whereas extracellular stimuli-mediated enhancement of p38 is attributed to lung inflammation and immune responses and is associated with apoptosis in AEC II [30,31]. We focused on p38 activity in AEC II to examine pulmonary fibrosis progression and performed transcriptome analysis. The results showed distinct expression of p38 MAPK pathway genes that was positively correlated with stepwise changes in intrinsic p38 activity in the lungs and the contribution of the immune system and ER functions to the development of pulmonary fibrosis mediated by activation of the p38 MAPK pathway. Regarding lung inflammation, the exacerbation correlated with increased p38 activity in the AEC II manifested as increases in inflammatory cells and pro-inflammatory cytokines in BALF and the enrichment of genes facilitating immune cell infiltration and cytokine interaction pathways. Augmentation of pro-fibrotic cytokines and immune response arising from inflammation leads to progression of tissue remodeling and fibrosis in the lungs [32]. In particular, the TGF-β signaling pathway driven by p38 induces EMT and fibroblast proliferation and activation through epithelial–mesenchymal crosstalk [33–35]. In this study, TGF-β1 was included in the 137 overlapping genes that showed correlations with intrinsic p38 activity in the lungs and upregulation among the three mouse groups treated with BLM (Figure 4 and Supplementary Tables S5 and S6). Additionally, IL-13 and IL-17, which in BALF were upregulated with a concomitant increase in intrinsic p38 activity in the lungs, can promote TGF-β signaling pathway-dependent EMT and fibroblast proliferation and resistance to apoptosis [36,37]. These cytokines originate from immune cells such as T cells, suggesting an association between the immune system and fibrosis [38,39]. A previous study showed that increased immune cells and aberrant regenerating epithelial cells express inflammatory mediators, including IL-17, in active fibrotic lesions of IPF lungs [40]. Furthermore, single-cell RNA-seq analysis of epithelial cells displaying atypical phenotypes in IPF lungs showed that these epithelial cells modulated the expression of inflammatory response- and TGF-β signaling pathway-related genes, leading to fibrotic remodeling [41]. Collectively, these findings strongly sugges<sup>t</sup> that inflammation and immune response enhanced by increased p38 activity in AEC II may contribute to the fibrotic process in the lungs.

Another possible explanation for the mechanism a ffecting fibrosis is that maladaptive ER stress response and its mediated apoptosis occurred concomitantly with increased p38 activity in AEC II. ER functions to retain cellular homeostasis by conducting posttranslational modification of proteins, with an adaptive process called unfold protein response under various stress conditions, although an excess of ER stress disrupting this adaptation elicits apoptosis [42]. The enrichment analysis in our study revealed that the ER protein processing pathway activated by BLM-induced reactive oxygen species was correlated with increased p38 activity in the AEC II. This result is consistent with the fundamental principle that ER stress can function in concert with the p38 MAPK pathway [43]. Simultaneously, AEC II homeostasis is sustained by its interaction with ECM, and p38 may participate in it [44]. A recent study using human fibroblasts revealed that the p38 MAPK pathway mediated the acquired resistance of ER stress modified by ECM metabolism through cell-to-ECM interaction [45]. In our study, ECM-receptor interaction was also a pathway enriched in accord with increased p38 activity in AEC II. Moreover, upregulation of ECM-related genes in enrichment analysis and matrix metalloproteases (MMPs; MMP-2, -3, and -9) in BALF (Supplementary Table S1), in addition to higher amounts of collagen, were connected to increased p38 activity in AEC II. These results are consistent with the study that MK2, a downstream substrate of p38, engaged in fibroblast activation and ECM production potentiated by fibroblast activation [46]. In addition, MMPs, which degrade all components of the ECM, are regulated by p38, while their upregulation leads to apoptosis and abnormal regeneration of lung epithelial cells [47]. Hence, ER stress in AEC II can be augmented by not only BLM-induced cytotoxicity, but also by the accumulation of ECM and AEC II to ECM interaction, controlled by the p38 MAPK pathway. These results emphasize the importance of p38 activity in AEC II and its related molecules in the progression of pulmonary fibrosis. In contrast, this murine model created by the intratracheal administration of BLM was not followed up after the establishment of pulmonary fibrosis. Therefore, further studies are required to determine whether p38 activity in AEC II influences the restoration of BLM-induced pulmonary fibrosis.

We validated four therapeutic target genes by comparing our data with publicly available data from IPF patients. First, *EPHA3* is expressed predominantly in lymphocytes and encodes a receptor tyrosine kinase implicated in regulating cell adhesion and cellular motility [48]. A recent study demonstrated that a novel epithelial cell population derived from IPF lungs co-expressed EPHA3 and CC chemokine receptor (CCR)-10 and facilitated the development of lung remodeling [49]. CCL28, a chemokine ligand for CCR10, was upregulated in BALF in accordance with increased p38 activity (Supplementary Table S1) [50]. These findings sugges<sup>t</sup> that coordination between CCL28-CCR10 chemokine signaling and the p38 MAPK pathway has important implications for reprogramming of epithelial cells, a speculation that warrants further investigation. Second, *POU2AF1* encodes a transcriptional coactivator that regulates B cell maturation and humoral immunity and is expressed in both airway epithelial and B cells [51,52]. A prior study using IPF lungs documented that transcriptome analysis identified *POU2AF1* as a promoter of pulmonary fibrosis and it is highly expressed in aggregates of B cells [53]. Third, *EDA2R* regulates ectodermal tissue development; its expression in lung epithelial cells such as AEC II was ascertained by single-cell RNA-seq data set in IPF lungs [41,54]. Genome-wide association study using human lung tissue identified *EDA2R* as a candidate gene involved in lung aging [55]. In addition, this gene accelerates the apoptotic process in two di fferent types of epithelial cells by activation of p53 signaling and caspase cascade [56,57]. These findings sugges<sup>t</sup> the involvement of *EDA2R* in AEC II senescence and apoptosis. Lastly, *SAMSN1* is expressed in healthy lung epithelial cells but not in lung cancer cells [58]. Although its functions in the lungs remain unknown, this gene is pivotal in regulating B cell activation and di fferentiation [59]. Thus, changes in the expression levels of these candidate genes by p38 activity may be involved in promoting fibrosis through molecular interactions between epithelial and immune cells in the IPF lung. This hypothesis is supported by two previous reports showing an association of lymphocytes and epithelial cells with progressive fibrosis in transcriptome analysis of IPF lungs [60,61]. Therefore, the interplay between these genes and the p38 MAPK pathway may be key to understanding the immunological mechanisms underlying IPF progression. However, further studies are needed to confirm the clinical significance of these genes in the patients with rapidly progressive IPF.

#### **4. Materials and Methods**
