University of Edinburgh
Multimodal profiling to monitor the response to epigenetic and xenobiotic perturbation in distal lung epithelial in vitro models
Abstract
dc:description.abstractDistal lung epithelial cells are a primary site for environmental exposure, xenobiotic injury, and pulmonary tissue repair. Perturbation of these cells may result in cytotoxicity, growth arrest, or repair-associated epithelial-mesenchymal plasticity. In chronic lung disease and cancer, this plasticity can become maladaptive. Cytotoxic injury, cytostasis, epigenetic dysregulation and partial epithelial-mesenchymal transition (EMT)-associated remodelling can often produce overlapping phenotypes. Traditional single-endpoint toxicological assays may therefore fail to distinguish transient damage from more nuanced and persistent disease-associated epithelial-state changes. Here, I show that a phenotype-anchored multimodal approach, structured around reproducible baseline, EMT and mesenchymal-epithelial transition (MET)-like reference states, can resolve the depth, persistence and reversibility of repair-associated plasticity in distal lung epithelial models. I used adenocarcinoma-derived alveolar-like A549 cells and non-transformed hTERT-immortalised small-airway-like HSAEC1-KT cells to establish multimodal baseline, EMT-stimulated and MET-like reference states. Baseline cells were maintained under unperturbed culture conditions, EMT was induced using TGF-β1 plus EGF stimulation, and MET-like re-entry was modelled by complete cytokine withdrawal. Quantitative morphometry, live-cell proliferation tracking, RNA sequencing and label-free proteomics showed that both models entered EMT-associated states, but differed in MET-like recovery towards baseline. Together, these reference states established multimodal signatures along a baseline-EMT-MET axis, allowing deviations from these signatures to be interpreted as xenobiotic-induced changes in distal lung epithelial state. Perturbing DNA methyltransferase 1 (DNMT1)-dependent maintenance methylation with GSK-3484862 caused widespread CpG hypomethylation but did not induce canonical EMT alone. It produced a distinct hypomethylation-associated state, deepened EMT-associated remodelling during EMT stimulation and impaired MET-like epithelial re-entry after cytokine withdrawal. Applying this phenotype-anchored multimodal approach to a chemically diverse panel of known and suspected pneumotoxins, anti-fibrotic compounds, and EMT-modulating compounds revealed strong model and readout dependence. Bleomycin and PHMG-p were injury-dominant, while amiodarone showed injury-dominant features in A549 and stress-associated EMT-like morphology in HSAEC1-KT. Paraquat and methotrexate produced clear EMT-like morphologies, yet generated only a partial EMT-associated transcriptional pattern in HSAEC1-KT cells. Anti-fibrotic nintedanib achieved mild attenuation of A549 EMT-associated morphology during EMT co-treatment. Ultimately, my findings demonstrate that individual readouts are insufficient to classify distal lung epithelial perturbations reliably along the baseline-EMT-MET axis. Multimodal assessment separated compounds that primarily injure cells or suppress growth from those that alter the depth, persistence or reversibility of epithelial-state change. The proposed multimodal workflow provides a practical framework for nuanced toxicological evaluation of chemical effects on pulmonary epithelial health.
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Peychev, Mihail D.
- Advisors dc:contributor.advisor
-
- Adams, Ian
- Pennings, Sari
Subjects
dc:subject × 6Rights
- Language dc:language.iso
- en
Identifiers
dc:identifier.*- Identifier URI
- https://doi.org/10.7488/era/7803