The single-cell immune atlas of bacterial pneumonia: from inflammatory cell infiltration to treatable cell states

Abstract

Bacterial pneumonia remains a major cause of infectious mortality, and antimicrobial resistance has intensified interest in host-directed therapy (HDT). Recent bronchoalveolar lavage fluid (BALF) and peripheral blood atlases (2024–2025) provide single-cell evidence from human disease, but the main translational gap is prioritization rather than catalog expansion. In this review, we reorganize published datasets from bacterial pneumonia and related lung infection cohorts (>500,000 cells; 2020–2026) into a state-prioritization scheme termed “treatable cell states.” Evidence is currently strongest for severity-linked expansion of indoleamine 2,3-dioxygenase 1–positive/programmed death-ligand 1–positive (IDO1+/PD-L1+) macrophages and pathological neutrophils; exhausted-like CD8+ T-cell states show emerging human transcriptomic and translational support but still require protein-level, functional, and longitudinal validation, whereas evidence for injured epithelium and repair-failed stromal states in bacterial pneumonia remains partly inferential and draws on adjacent lung-injury literature. We therefore treat the six proposed state categories as a ranked hypothesis set for macrophage-, neutrophil-, T-cell-, epithelial-, and stromal-directed interventions, not as validated clinical targets. Most proposed interventions still require longitudinal and functional validation before routine bedside use.