Airway immune signatures predict tuberculosis progression

Tuberculosis (TB) infects millions, but most people who carry Mycobacterium tuberculosis control the bug and never get sick; only about 5–10% go on to develop active disease. Current tests such as antigen-specific interferon (IFN)-γ release assays (IGRA) can tell if a person has been exposed, but they cannot predict who will progress to active TB. Imaging with 18 F-Fluorodeoxyglucose positron emission-computed tomography (PET-CT) can reveal increased immune activity in the lungs and lymph nodes, suggesting where the battle is happening, yet the early local immune signals that decide protection or progression are poorly understood. To address that gap, a team led by William J. Branchett studied people who had recently lived with TB patients and were carefully followed to see who stayed healthy and who developed TB. The researchers sampled the airways using bronchoalveolar lavage (BAL) and analyzed those airway cells at single-cell resolution to look for patterns that differentiate controllers from progressors. Their goal was to capture the early, local immune events in the lung that might determine whether infection is contained or becomes disease.

The study used unbiased analysis of bulk and scRNA-seq of BAL samples to profile the airway immune landscape. By examining single cells from the lung washes, the team identified distinct neutrophil signatures that were either type I IFN-dependent or type I IFN-independent in people with active TB and in household contacts who later progressed to disease. They also found an inverse relationship between airway neutrophils and T cells: samples with many neutrophils had fewer T cells, and those T cells bore molecular signs of exhaustion, cytotoxicity and cell death in progressors and TB patients with a neutrophil-dominated airway profile. In contrast, T cells from contacts who remained healthy showed gene patterns linked to regulation, quiescence and a stem-like profile. Crucially, both the inflammatory neutrophil signature linked to progression and the protective stem-like T cell signature were recapitulated in scRNA-seq data from non-human primate (NHP) granulomas, where they associated with disease or immune protection respectively. These methods tied airway cell states to outcomes in humans and validated them in an animal model.

The findings point to specific cell states and pathways in the lung that appear to steer infection toward either control or disease. A neutrophil-dominated, inflammatory airway picture coupled with exhausted, dying T cells marks progression, while a balanced airway with stem-like, regulated T cells marks protection. Because these patterns showed up early in household contacts and were mirrored in non-human primate granulomas, they could serve as targets for new diagnostics, host-directed therapies, or vaccine strategies that aim to boost protective T cell states or blunt harmful neutrophil-driven inflammation. Importantly, the study suggests that sampling the airway directly with bronchoalveolar lavage (BAL) and applying bulk and scRNA-seq can reveal signals missed by blood tests alone, offering a route to more predictive biomarkers and interventions that act at the site of infection in the lung.