A distinct monocyte transcriptional state links systemic immune dysregulation to pulmonary impairment in long COVID

nature immunology Article https://doi.org/10.1038/s41590-025-02387-1 A distinct monocyte transcriptional state links systemic immune dysregulation to pulmonary impairment in long COVID Received: 17 March 2025 Accepted: 3 December 2025 Published online: 14 January 2026 Check for updates Saumya Kumar 1,2, Chaofan Li 3,4, Liang Zhou1,2, Qiuyao Zhan1,2, Ahmed Alaswad1,2, Sonja Volland 5, Bibiana Costa 1,2, Simon Alexander Krooss1,2,6,7,8,9, Isabel Klefenz 10, Hagen Schmaus1,2,11, Antonia Zeuzem1,2,9,11, Dorothee von Witzendorff1,2,11,12, Helena Lickei1,2,9,11,12, Lea Pueschel 9, Anke R. M. Kraft 1,2,9,11,12, Markus Cornberg 1,2,9,11,12, Andreas Rembert Koczulla7,13,14,15, Isabell Pink16,17, Marius M. Hoeper 16,17, Cheng-Jian Xu1,2,18, Susanne Häussler12,19,20,21, Miriam Wiestler 9, Mihai G. Netea 18,22, Thomas Illig5,17,23,24, Jie Sun 3,4,24 & Yang Li 1,2,12,17,18,23,24 The mechanisms driving immune dysregulation in long COVID disease remain elusive. Here we integrated single-cell multiome data, immuno logical profiling and functional assays to investigate immune alterations across multiple cohorts. A transcriptional state in circulating monocytes (LC-Mo) was enriched in individuals with mild–moderate acute infection and accompanied by persistent elevations of plasma CCL2, CXCL11 and TNF. LC-Mo showed TGFβ and WNT–β-catenin signaling and correlated with fatigue severity. Protein markers of LC-Mo were increased in individuals with pronounced fatigue or dyspnea, and those with severe respiratory symptoms showed higher LC-Mo expression. Epigenetically, LC-Mo exhibited AP-1- and NF-κB1-driven profibrotic programs. LC-Mo-like macrophages in bronchoalveolar lavage samples from individuals with severe respiratory symptoms displayed a profibrotic profile, and individuals with a high LC-Mo transcriptional state showed impaired interferon responses after stimulation. Collectively, our findings define a pathogenic monocyte transcriptional state linking systemic immune dysfunction to persistent long COVID disease, providing mechanistic insights and potential therapeutic targets. Long COVID affects 10–20% of individuals after severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, with symptoms ranging from mild discomfort to severe, long-lasting impairments such as fatigue, respiratory issues and neurological problems. These symptoms can persist for over 3 years (refs. 1–5), representing a substantial health burden and prompting efforts to better characterize long COVID (LC), including biomarker discovery for improved diagnosis6–10. LC presents with diverse symptoms reflecting multiorgan system abnormalities11–13. The evidence suggests multiple possible A full list of affiliations appears at the end of the paper. causes, including persistence of viral remnants or reactivation of latent viruses7,14–17. Yet, persistent immune dysregulation is a consistent finding in LC studies10,11,14,16–19. Although most LC cases follow mild-to-moderate acute illness, many studies do not stratify individuals by acute infection (AI) severity6–8, which is crucial because severe cases, especially those treated in the intensive care unit, develop immune changes due to intensive medical interventions20,21. Failing to account for these differences may confound LC-associated molecular signatures, highlighting the importance of refined patient grouping. e-mail: Nature Immunology | Volume 27 | February 2026 | 200–212 200 Article To address this gap, we stratified individuals with LC by acute COVID-19 severity to better resolve immune heterogeneity and identify molecular features underlying chronic symptoms. We applied single-cell multiomics profiling of peripheral blood mononuclear cells (PBMCs) and measured plasma cytokines from individuals with LC with fatigue and respiratory symptoms using longitudinal and cross-sectional samples. We identified a distinct circulating CD14⁺ monocyte state associated with LC (‘LC-Mo’), which was enriched in individuals with mild-to-moderate AI. This state coincided with persistent elevation of circulating cytokines, indicating systemic inflammation. In two independent cohorts of individuals with LC with severe respiratory symptoms and abnormal lung function, LC-Mo expression was increased in circulating CD14⁺ monocyte subsets. In broncho alveolar lavage (BAL) myeloid cells from individuals with severe respiratory symptoms, LC-Mo-like macrophages showed a profibrotic gene expression profile. Functionally, CD14⁺ monocytes from individuals with LC-Mo enrichment showed dysregulated responses to ex vivo stimulation, indicating impaired immune regulation. Together, these findings provide systemic insight into the cellular and molecular basis of LC and highlight potential therapeutic targets. https://doi.org/10.1038/s41590-025-02387-1 Individuals presenting with headache, dyspnea or fatigue to the pneumology outpatient clinic at Hannover Medical School (MHH) were recruited according to the German S1 guidelines22 and the Delphi Consensus Criteria21 for LC (4–12 weeks) and post-COVID-19 syndrome (>12 weeks). These criteria included symptoms persisting beyond the acute phase of SARS-CoV-2 infection or its treatment, new symptoms emerging after recovery and attributed to prior infection or worsening of pre-existing conditions. Because heterogeneity in LC molecular profiles may be shaped by acute disease severity and treatment, we stratified individuals with acute SARS-CoV-2 infection (AI) and LC into those with mild-to-moderate (WHO score of 1–5) AI (AIM and LCAM) and those with severe (WHO scores 6–9) AI (AIS and LCAS). Cohort 1 included 45 individuals recruited between April 2020 and August 2021 at MHH, of which 9 gave longitudinal samples and 36 gave cross-sectional samples (n = 78 total samples), including 11 donors with AI categorized as AIM (n = 7 donors, 42.8% women, median age = 52, range 23–66 years of age, WHO score range 1–5) and AIS (n = 4 donors, 50% women, median age = 37, range 32–54, WHO score range 6–9), 37 donors with LC categorized as LCAM (n = 29 donors, 8 longitudinal donors with two to three time points and 21 single-time-point donors, 58% women, median age = 49, range 31–84 years) and LCAS (n = 8 donors, 3 with two to four time points, 5 single-time-point donors, 25% women, median age = 46, range 19–75) and 8 donors who had recovered after 4–8 months of LC (RLC; 1 longitudinal donor with two time points and 7 single-time-point donors, 37.5% women, median age = 38, range 19–65), in addition to 6 prepandemic noninfected control individuals (NI; 50% women, median age = 40, range 24–61). LC and RLC samples were collected 1.7–10.2 months after infection. Cohort 2 included 117 LCAM donors (24 donors with two to four time points, 93 single-time-point donors, 58.9% women, median age = 48, range 19–83) and 25 LCAS donors (12 longitudinal donors, 13 single-time-point donors, 20% women, median age = 53, range 18–81), recruited between May 2020 (...truncated)