Emerging highly pathogenic H5N1 influenza triggers fibrotic remodeling in human airway organoids
- PMID: 40712003
- PMCID: PMC12302397
- DOI: 10.1080/22221751.2025.2532684 Abstract
- Discussion
The unpredictable evolution of IAVs makes future pandemics inevitable, highlighting the urgent need for preparedness to mitigate health, social, and economic impacts. H5N1 cases were first described in Hong Kong in 1997 when birds died from an outbreak of the disease, propagating to humans later in the same year [Citation33,Citation34]. Since the resurgence of H5N1 in 2002–2003, global efforts have been intensified to prevent its spread. The emergence of HPh/b-TX H5N1 represents a significant zoonotic threat, with its origin linked to infected dairy cows with highly pathogenic H5N1 from the clade 2.3.4.4b [Citation35]. This outbreak underscores the risk of cross-species transmission as the virus has spread within poultry, cattle, and humans, posing a significant public health threat [Citation36].
Infections with H5N1 cause severe pneumonia that can progress to lung fibrosis and respiratory failure [Citation31,Citation32]. While current FDA-approved antivirals reduce viral loads [Citation37], these antivirals are often ineffective in preventing lung injury and fibrosis, underscoring the need for novel therapeutic strategies targeting host responses. To better understand viral replication and pathogenesis, we employed HAO, a physiologically relevant model that recapitulates key aspects of the human lung environment.
We infected HAO with either rHPh-TX H5N1 or rHPb-TX H5N1 at MOI of 0.01 to assess viral replication efficiency. Our findings demonstrate that rHPh-TX H5N1 replicates more efficiently in HAO, serving as a surrogate for human airway epithelial tissues, compared to rHPb-TX H5N1.
The elevated levels of infectious rHPh-TX H5N1 particles detected in secreted mucus indicate a greater degree of viral adaptation to the human airway epithelium. This efficient replication and shedding into mucus may facilitate aerosolization and increase the potential for person-to-person transmission. Additionally, the robust replication of rHPh-TX H5N1 in HAO was associated with a markedly strong induction of the IFN response, reflecting heightened host recognition and immune activation.
Typically, H5N1 infection activates the RIG-I/MAVS signalling pathway, leading to IRF3 and NF-κB activation, which drives IFN-β production and ISG15 expression [Citation38–40].
Our data show that rHPh-TX H5N1 induces IRF3 and NF-κB expression, which increase IFN-β and ISG15 levels. While ISG15 contributes to antiviral defense, its dual role in immune regulation can either enhance or suppress IRF3/NF-κB signalling, potentially influencing disease severity.
In severe cases, H5N1 evades IFN pathways, leading to excessive NF-κB-driven inflammation (cytokine storm) and impaired IRF3-mediated responses, exacerbating lung damage [Citation41–43].
Cytokine profiling of infected HAO revealed that rHPh-TX H5N1 induced elevated IL-6, IL-1β, TNF, CCL5, and IP-10 (CXCL10) levels.
While these cytokines aid in viral clearance, prolonged inflammation may contribute to tissue injury and fibrosis (e.g. cytokine storm-induced lung injury).
Given the robust inflammatory response induced by rHPh-TX H5N1, we investigated its role in airway fibrogenesis during prolonged infection with rHPh-TX H5N1 compared to seasonal pH1N1. Our results indicate that prolonged infection of HAO with rHPh-TX H5N1 and pH1N1 induced fibroblast-like cells surrounding the infected area, associated with high cytokine response and α-SMA expression, indicating fibroblast-to-myofibroblast differentiation.
It has been reported that H1N1 infections can rapidly progress to ARDS and contribute to pulmonary fibrosis [Citation25].
Furthermore, severe IAV infections are associated with higher levels of TGF-β, indicating a relationship between disease severity and IAV-induced pulmonary fibrosis [Citation44]. This agrees with our findings showing that the rHPh-TX H5N1 infection induces higher TGF-β and ECM-associated gene expression than pH1N1. As such, the increased fibroblast activity in post-inflammatory repair pathways, with a pivotal role played by TGF-β and ECM, seems to be linked to IAV-induced pulmonary fibrosis.
We hypothesize that the EMT process plays a prominent role in fibrogenesis. During the EMT process, epithelial cells lose cell–cell attachment, polarity, and epithelial-specific markers, undergo cytoskeletal remodelling, and gain a mesenchymal phenotype [Citation45].
Our results indicate that rHPh-TX H5N1 infection induces more pronounced EMT and ECM remodelling than pH1N1, as evidenced by the significantly higher upregulation of TGF-β and ECM-associated genes. We also observed the spindle shape of fibroblasts with intensive expression of α-SMA. In this regard, activated fibroblasts are described as spindle or stellate morphology cells with intracytoplasmic stress fibres, a contractile phenotype, expression of various mesenchymal markers such as α-SMA, and collagen production [Citation27].
We also observed that at 10-DPI, pro-inflammatory (TNF, IL-6, IL-8, IL-1β) and pro-fibrotic (TGF-β) mediators were upregulated, along with ECM components (FN, COL1A, COL3A, MMP2, and MMP9).
Importantly, the upregulation of FN expression allows microbes to adhere to epithelial surfaces and contributes to the virulence of secondary bacterial infections [Citation46,Citation47].
Collectively, these factors contribute to a feedback loop of persistent lung damage and fibrosis [Citation26], emphasizing the need for targeted interventions to disrupt these pathways.
Our data showed that the ROCK pathway affects fibroblast foci formation and ECM deposition. The ROCK1 and ROCK2 pathways are key regulators of fibrosis with distinct roles in tissue remodeling [Citation20]. Our results indicate that the inhibition of ROCK1 activity significantly reduced fibroblast foci formation and ECM deposition, whereas ROCK2 inhibition had a lesser effect. This suggests a dominant role for ROCK1 in airway fibrosis during H5N1 infection.
Previous studies described the capacity of antifibrotic agents such as pirfenidone and nintedanib, FDA-approved to treat idiopathic pulmonary fibrosis (IPF), to reduce fibrosis through inhibiting key cytokines, including TGF-β and vascular endothelial growth factor (VEGF) [Citation48,Citation49]. Thus, unravelling the mechanisms of the ROCK signalling pathway in IAV-induced fibrogenesis may be essential for developing effective strategies to prevent and treat viral-induced pulmonary fibrosis, ultimately mitigating its long-term impact on respiratory health.
To our knowledge, this is the first study to characterize the profibrotic response triggered by the highly pathogenic HPh-TX H5N1 strain in human airway organoids. Our organoid model recapitulates the structural and functional complexity of the human airway epithelium, including multicellular composition and differentiation markers that are absent in traditional immortalized cell lines [Citation50]. Unlike animal models, this system provides a controlled, human-specific platform to study early epithelial responses to viral infection [Citation51].
The novelty of our approach is further underscored by using a recent zoonotic H5N1 isolate (HPh-TX H5N1), associated with the 2024 human and bovine outbreaks in the U.S., which has not previously been evaluated in this context.
While our study provides insights into H5N1 pathogenesis and triggering fibrogenesis, it also has limitations. Though physiologically relevant, the HAO model does not fully recapitulate in vivo immune responses or systemic factors that promote disease progression. HAO lacks immune cell components, limiting its ability to model immune cell recruitment and systemic cytokine amplification. While our findings demonstrate strong epithelial-intrinsic responses to rHPh-TX H5N1, they may underestimate the full extent of cytokine storm observed in vivo.
Future studies should evaluate the therapeutic potential of ROCK1 inhibitors in preclinical models of IAV-induced lung injury and fibrosis. Investigating the molecular links between IFN signalling, inflammation, and fibrotic remodelling will uncover additional therapeutic targets. Although our findings support the regulatory role of ROCK1, further investigation of downstream effectors, such as SMAD4 and p-MLC [Citation52,Citation53], is required to clarify the mechanisms underlying ROCK1-mediated fibrotic and inflammatory responses.
Future studies will include examining the fibrotic potential of other highly pathogenic IAV with zoonotic potential, such as H7N9. Additionally, we will assess the effects of currently approved antiviral therapies, such as neuraminidase inhibitors (e.g. Oseltamivir) and endonuclease inhibitors (e.g. Baloxavir), in combination with ROCK1 inhibitors, which could provide additive or synergistic benefits by simultaneously controlling virus-induced lung injury and mitigating fibrotic responses.
In conclusion, our findings highlight the efficient replication and innate immunity immune activation of H5N1 in HAO. H5N1 induces a robust inflammatory and fibrotic response, driven by NF-κB and TGF-β signalling, contributing to airway remodelling and fibrosis.
ROCK1 inhibition is a promising therapeutic strategy that warrants further investigation in appropriate in vivo models. These insights emphasize the importance of developing host-targeted therapies to prevent severe lung complications associated with influenza infections.