A new tunable 3D alveolospheres model from human alveolar epithelial type 2 cells (AEC2) with reduced heterogeneity for studying cigarette smoke extract exposure

Three-dimensional (3D) organoid models, such as alveolospheres, are unique tools for investigating the mechanisms underlying emphysema. However, high inter-organoid heterogeneity hampers consistent results in emphysema research and drug testing. To develop a tunable 3D alveolosphere derived from human primary type II alveolar epithelial cells (AEC2) for modeling alterations linked to cigarette smoke exposure. AEC2 (HTII-280+) were isolated from 52 lung samples from both COPD and non-COPD patients, then cultured in 3D, comparing Matrigel to preformed photopolymerized hydrogel microwells of adjustable size and stiffness. Topological and phenotypic characterization were performed on days (D)1, 7, and 14. Lamellar bodies (LBs) were quantified using artificial intelligence (AI) analysis of transmission electron microscopy (TEM) serial block-face images. Chronic exposure to 1% or 5% cigarette smoke extract (CSE) was performed for 5 consecutive days. Compared to Matrigel-based spheroid cultures, alveolospheres generated in microwells display reduced heterogeneity in size. Such alveolospheres were maintained in culture for 14 days and exhibited central lumen formation from D7 to D14. Across different hydrogel stiffness, a stiffness of 5 kPa was found to best support long-term organoid maintenance. The presence of tight junctions (TEM, ZO-1 immunostaining) suggested an auto-organization. AEC1 markers (P2XR4, PDPN) increased from D1 to D14 while AEC2 markers (ABCA3, SFTPA, SFTPC) persisted over time, in qPCR. TEM indicated surfactant synthesis, and AI-driven LB quantification revealed a decrease in LB-containing cells over time. CSE exposure resulted in cell death, architectural disorganization, oxidative stress, and inflammation. Similarly, alveolospheres derived from COPD patients showed increased expression of inflammatory and cell death markers. This standardized and adjustable 3D alveolosphere model, derived from human primary AEC2, successfully reproduces key native alveolar features. Exposure to CSE provides a relevant platform for studying responses to cigarette smoke exposure.