New 3D human model finds TB host-directed drug candidates

Tuberculosis remains a major global health threat, complicated by drug-resistant strains and long, toxic treatment courses. Scientists are exploring Host-directed therapies (HDT) as a way to boost the body’s own defenses and improve outcomes alongside antibiotics. But many promising HDT candidates identified in simple laboratory tests fail when pushed into animals or human trials, and standard macrophage cultures do not recreate the complex architecture of a TB lesion. To address this gap, Suraj B. Sable and colleagues developed a practical, scalable platform that grows three-dimensional (3D) human tuberculoma-like structures. These tuberculoma-emulating assemblies are built by co-culturing human cells with pathogenic mycobacteria in a format designed for high-throughput screening (HTS). The goal was to make an accessible system that reproduces the critical features of tuberculous granulomas — the crowded, layered, and sometimes necrotic tissue environments where Mycobacterium tuberculosis hides — so researchers can test HDT candidates in settings that better mimic human disease.

The researchers combined several modern tools to characterize their model and run screens. Using high-content imaging alongside immunological and transcriptomic approaches, the team showed the 3D structures develop classic tuberculoma attributes and undergo solid, necrotic, and cavitary transformations seen in real lesions. The HTS-compatible format allowed screening of antibody biosimilars and HDT compounds directly in the 3D human in vitro tuberculomas. Results diverged sharply from simpler tests: many agents that worked in two-dimensional (2D) cultures were ineffective once granulomas formed in the 3D model. Conversely, several candidates reduced mycobacterial burdens and shrank granuloma-like lesions in the 3D system. Mechanistic readouts implicated innate immune pathways, and several effective compounds induced rapid autophagy flux in the 3D environment. The team went on to validate the effectiveness of one such compound, the multi-kinase inhibitor AT9283, in a mouse model, demonstrating the platform’s ability to move hits toward in vivo testing.

This work offers a practical bridge between simple lab assays and animal studies by providing a human-derived, lesion-like test bed for TB therapeutics. An HTS-ready 3D tuberculoma bioplatform can help prioritize HDT candidates that truly work in granuloma milieus, reducing time and resources spent on false leads from 2D tests. The ability to screen antibody biosimilars and small molecules and to read out immune and transcriptomic effects in situ means researchers can begin to unravel how promising compounds change host responses inside realistic tissue contexts. By highlighting compounds that trigger autophagy flux and validating AT9283 in mice, the study points to specific repurposing opportunities and immune mechanisms worth following. Beyond tuberculosis, the platform provides a framework for advancing human in-vitro granuloma models and for testing therapies against other granulomatous diseases as tissue-like culture systems evolve.