Dual lung models reveal compartment-specific tuberculosis drug effects

Tuberculosis (TB) remains a major global health challenge and researchers are seeking better ways to test drugs and host-directed strategies that might improve treatments. To address this, Caio César Barbosa Bomfim and colleagues built complementary human lung models in the lab that capture two distinct parts of the lower respiratory tract. One model uses alveolar macrophage-like (AML) cells to mimic the immune cells that sit inside air sacs, while the other is an airway air-liquid interface (ALI) culture that models the lining of the airways. The AMLs reproduced key morphological, transcriptional, and functional features of primary alveolar macrophages, including a CD16 + immunoregulatory phenotype that was highly permissive to Mycobacterium tuberculosis (Mtb) infection. The ALI cultures preserved epithelial barrier integrity and secretory functions, allowing researchers to apply bacteria to the airway surface in an apical Mtb infection setup. By putting these two models side-by-side, the team created a way to study how drugs and host-directed therapies act in different pulmonary compartments that together constitute the human lung.

The dual platform was used to evaluate standard-of-care antibiotics, host-directed therapies, and virulence-targeting agents under conditions that reflect each lung niche. AMLs captured macrophage-like responses while ALI cultures allowed drug penetration analysis and inflammatory profiling across an intact epithelial barrier. When the researchers benchmarked standard-of-care antibiotics they found compartment-specific activity: isoniazid, rifampicin, and moxifloxacin were effective in both AML and ALI systems, while pyrazinamide showed activity only in AMLs. The team also tested anti-inflammatory host-directed therapies and found that ibuprofen and doramapimod selectively reduced cytokine production without affecting bacterial load. These results show that some drugs have broad activity across lung compartments but others act primarily within macrophages, and that host-directed anti-inflammatory drugs can modulate inflammation independently of killing bacteria.

This work suggests a practical path toward more physiologically relevant preclinical testing. Because the dual-system is scalable and preserves distinct features of alveolar macrophages and airway epithelium, it can reveal where a candidate drug will actually work in the lung and whether a host-directed therapy will change inflammation without altering bacterial survival. That distinction matters for drug development and for combination strategies: a compound that looks potent in a macrophage assay may not reach or act in the airway epithelium, and anti-inflammatory agents like ibuprofen or doramapimod might reduce damaging inflammation but would not replace antimicrobial activity. By bridging the gap between conventional macrophage assays and the complex human lung, the platform could help researchers prioritize compounds, refine dosing and delivery studies, and better understand how different pulmonary niches contribute to treatment outcomes.