**Background:** Developing effective treatments for COVID-19 requires comprehensive understanding of the pathology caused by SARS-CoV-2, which primarily targets the respiratory system and lung tissue. Traditional 2D cultures do not resemble the biochemical microenvironment or proper physiology of the human body, while 3D cell culture systems can better mimic the internal microenvironment, leading to higher accuracy in viral and antiviral compound screening studies. This review surveys the components and factors that should be considered during generating lung spheroids as models for respiratory disease, including cell composition, matrix type, vasculature system, and implementation in COVID-19 studies.
**Methods:** The authors conducted a narrative review of the literature on 3D lung models, covering the history of lung spheroid generation, cell sources (including primary epithelial cells, stem cells, and pluripotent stem cells), natural and synthetic polymer scaffolds, vascularization approaches, and applications in COVID-19 research. The review summarizes findings from multiple studies that developed various 3D lung models including spheroids, organoids, and organ-on-a-chip systems.
**Key Results:** The first lung model with a 3D structure was generated by Benali and collaborators in 1992 using human primary epithelial respiratory cells cultured on collagen matrix. Multiple methods for spheroid generation exist including hanging drop, hydrogel matrices, microfluidic systems, and bioreactors. Cell sources for lung spheroids include A549 adenocarcinoma cells, primary basal cells, alveolar epithelial type II (AEC2) cells, adult stem cells (basal cells, AEC2, pulmonary neuroendocrine cells, bronchioalveolar stem cells), and human pluripotent stem cells. Key culture conditions include supplementation with epidermal growth factor (EGF), retinoic acid, and Rho-associated protein kinase inhibitors (ROCKi).
For COVID-19 applications, multiple 3D lung models have been developed. Han et al. developed lung organoids from human pluripotent stem cells (hPSC-LOs) using 100% Matrigel, showing susceptibility of alveolar type-II-like cells to SARS-CoV-2 infection with chemokine induction patterns similar to COVID-19 patients. Salahudeen et al. used basement membrane extract II to develop distal lung organoids, identifying AT2 cells, basal cells, and SCGB1A1+ club cells as targets for viral infection, while no infection was observed in ciliated cells. Youk et al. established a matrix-free model of 3D human alveolar stem cells (hAT2) showing rapid viral replication and high expression of proinflammatory and interferon-associated genes. Katsura et al. developed alveospheres from human AT2 cells on Matrigel-coated plates, finding that 40% of AT2s expressed ACE2 and about 80% were positive for TMPRSS2, with pre-treatment with interferons showing prophylactic effectiveness. Suzuki et al. developed human bronchial organoids (hBO) on Matrigel showing high expression of ACE2 and TMPRSS2 after SARS-CoV-2 infection with increased cytotoxicity and intracellular viral genome.
Natural biomaterials for scaffolds include collagen (forming about 15% of dry weight of human lung), silk fibroin (ultimate tensile strength of 7.25 MPa), fibrin, Matrigel (extracted from murine Engelbreth-Holm-Swarm sarcomas), glycosaminoglycans, hyaluronic acid, and chitosan. Synthetic polymers include polyethylene glycol (PEG), poly(N-isopropylacrylamide) (PNIPAm), polydimethylsiloxane (PDMS), and poly-ϵ-caprolactone (PCL). For vascularization, endothelial cells (ECs), endothelial progenitor cells (EPCs), and mesenchymal stromal cells (MSCs) are commonly incorporated. Co-culture of HUVEC with other cell types has shown capillary-like network formation. Wörsdörfer and colleagues generated vascularized neural and tumor organoids by co-culture of mesodermal progenitor cells with neural spheres and tumor cells.
**Clinical Implications:** 3D lung models represent promising biomimetic systems for drug screening and in vitro studies for SARS-CoV-2. They can be used to understand mechanisms of COVID-19 infection, test efficacy of potential antiviral drugs, identify new drug targets, and study long-term effects of the virus on lung tissue including post-acute sequelae of COVID-19. However, technical challenges remain including determining optimal cell types, inadequate vascularization leading to core necrosis, and limitations of both natural polymers (lack of mechanical properties and reproducibility) and synthetic polymers (absence of ECM components and inertness). The authors suggest that integrating bioactive peptides into synthetic polymers and combining natural and synthetic scaffolds in composite polymers represent valid solutions to overcome these limitations.