ABSTRACT
RATIONALE: Acute respiratory distress syndrome (ARDS) results from injury to the alveolar epithelial cell (AEC) barrier leading to pulmonary edema;recovery requires epithelial regeneration. However, the specific defect resulting in ongoing barrier permeability in fatal early ARDS is unknown. In mouse models of ARDS, we found that AEC2s assume a transitional state characterized by transient cell cycle arrest during differentiation towards AEC1s. Transitional cells persist and are senescent in human idiopathic pulmonary fibrosis (IPF), leading some to speculate that persistence of transitional cells is pathognomonic of fibrosis in humans. We hypothesized that transitional cells also arise early in human ARDS and that incomplete AEC1 differentiation from the transitional state underlies barrier permeability and death from respiratory failure in early ARDS. We speculated that in contrast to IPF, transitional cells in early ARDS are in transient cell cycle arrest but not senescent and maintain capacity for an AEC1 fate. METHODS AND RESULTS: Lung tissue was obtained from patients who died within two weeks of hospitalization of ARDS due to COVID-19 or other etiologies, and from patients with IPF. Histology revealed diffuse alveolar damage without fibrosis in patients with ARDS. Immunostaining demonstrated AEC damage and abundant transitional cells in both ARDS and IPF. In ARDS, transitional cells existed in a monolayer on alveolar septa, filling gaps denuded of AEC1s and displaying spread morphologies without AEC1 marker expression, suggesting ongoing but incomplete differentiation. In fibrosis, transitional cells existed on a background of architectural distortion and fibrosis. Meta-analysis of single cell RNA sequencing (scRNAseq) datasets demonstrated that transitional cells were transcriptionally highly similar. However, the senescence marker p16 was expressed in transitional cells in human IPF but not in mouse models. Immunostaining confirmed that transitional cells in IPF but not human ARDS expressed p16. CONCLUSION: We conclude that transitional cells arise in early human ARDS without fibrosis. We propose that incomplete AEC1 differentiation from the transitional state is the specific defect in epithelial regeneration underlying barrier permeability and respiratory failure. We speculate that in early human ARDS, as in mouse models, transitional cells retain the capacity to differentiate into AEC1s, restoring alveolar architecture without fibrosis. However, in IPF and fibroproliferative ARDS, transitional cells become senescent, lose capacity for AEC1 differentiation, and fibrosis ensues. Evolution of transitional cells from a transient cell cycle arrest to a permanent cell cycle arrest (senescence) may be the key defect driving the pathogenesis of fibrosis after injury.
ABSTRACT
COVID-19 is a multiorgan disease that most prominently causes respiratory failure and the acute respiratory distress syndrome (ARDS). Epidemiologic studies of patients with COVID-19 have shown a high percentage of hospitalized patients requiring intensive care and prolonged invasive mechanical ventilation and high mortality. The cellular mechanisms that underlie prolonged ventilator dependence and high mortality in these patients remain unknown. Previous studies in alveolar epithelial regeneration have identified the presence and a possible pathogenic role of a persistent transitional cell state that cuboidal type 2 cells (AEC2s) adopt during differentiation into squamous type 1 cells (AEC1s). This transitional state has a unique gene expression signature that includes strong expression or keratin 8 (Krt8). We hypothesized that ARDS secondary to COVID-19 was associated with the persistence of the transitional state of alveolar regeneration, perhaps contributing to prolonged ventilator dependence and mortality. Institutional review board approval was obtained for relevant clinical data. Briefly, the patients included in this study developed progressively worsening ARDS despite lung protective ventilation and adjunctive therapies, with progressive diffuse alveolar infiltrates on serial imaging. All patients died within two weeks of hospitalization, and all had persistent hypoxemia despite maximally tolerable ventilator support near the end of life. Lung tissue from autopsy specimens from these patients was utilized for immunofluorescence, hematoxylin and eosin, and trichrome staining. Histology was consistent with diffuse alveolar damage. Staining revealed extensive damage to AEC1s (Figure 1, stained for HTI-56) and AEC2s (not shown) and abundant Krt8hi transitional cells (Figure 1) in lung tissue from COVID-19 patients (Figure 1) without significant fibrosis. Our findings establis h the presence of Krt8hi transitional cells in conjunction with extensive alveolar epithelial damage and failure to regenerate AEC1s. We speculate that prolonged barrier dysfunction from dysregulated alveolar repair, perhaps mediated by a failure of Krt8hi transitional cells to differentiate into mature AEC1s, contributes to ventilator dependent respiratory failure and mortality. Though prior studies have suggested a role of persistent Krt8hi transitional cells in fibrosis, our findings suggest that prevalence of Krt8hi transitional cells alone is not sufficient to cause fibrosis, suggesting that that the extensive epithelial damage observed in COVID-19 ARDS is potentially reversible. We conclude that in severe COVID-19 ARDS, alveolar regeneration is incomplete, characterized by persistence of the Krt8hi transitional state but that differentiation of these cells into mature AEC1s resulting in barrier restoration and clinical recovery may still be possible.