Single-cell RNA-sequencing has transformed our understanding of alveolar epithelial type 2 (AT2) cells and alveolar lipofibroblasts (LIFs) during lung injury and repair. Both cell types undergo dynamic transitions through intermediate states that determine whether the lung proceeds toward regeneration or fibrosis. Emerging evidence highlights reciprocal paracrine signaling between AT2/AT1 transitional cells and LIF-derived myofibroblasts (aMYFs) as a key regulatory axis. Among these, amphiregulin (AREG)–EGFR signaling functions as a central profibrotic pathway whose inhibition can restore alveolar differentiation and repair. The human WI-38 fibroblast model provides a practical platform to study the reversible LIF–MYF switch and screen antifibrotic and pro-regenerative compounds. Candidate therapeutics including metformin, haloperidol and FGF10 show promise in reprogramming fibroblast and epithelial states through metabolic and signaling modulation. Integrating WI-38-based assays, alveolosphere co-cultures, and multi-omics profiling offers a translational framework for identifying interventions that halt fibrosis and actively induce lung regeneration. This review highlights a unifying framework in which epithelial and mesenchymal plasticity converge to define repair outcomes and identifies actionable targets for promoting alveolar regeneration in chronic lung disease.

Idiopathic pulmonary fibrosis (IPF) is a progressive and fatal interstitial lung disease with limited treatment options. Depression and anxiety are common comorbidities in patients with IPF, and emerging evidence suggests that neuroactive pathways may also influence fibrotic remodeling. On this basis, we investigated desipramine, a tricyclic antidepressant, as a potential modulator of fibroblast state in lung fibrosis. Desipramine was identified in an FDA-approved compound screen as a pro-lipogenic hit in TGF-β-stimulated fibroblasts and was subsequently evaluated across a stepwise validation pipeline of increasing biological complexity. In WI-38 fibroblasts, desipramine was well tolerated at 10 μM and reduced myofibroblast-associated features while increasing lipid-associated staining. In a fibroblast-supported alveolosphere assay, desipramine altered qualitative organoid clustering and changed the transcript levels of specific mesenchymal markers under profibrotic stimulation, whereas direct treatment of MLE-12 epithelial cells did not elicit a consistent response. While desipramine demonstrated pro-lipogenic and anti-myofibroblastic phenotypic shifts in reductionist 2D cultures, these effects failed to translate robustly into complex 3D human lung tissue slices or in vivo disease models. Ultimately, our findings highlight the critical necessity of utilizing complex translational pipelines to rigorously validate early screening hits before therapeutic efficacy is assumed.