COVID-19-Associated White Lung Correlates With the Dysfunctional Neutrophil Response Revealed by Single-Cell Immune Profiling

Abstract

Most individuals with COVID-19, caused by SARS-CoV-2 infection, experience asymptomatic or mild-to-moderate symptoms, while a minority of patients may deteriorate to severe illness or fatal outcomes [1]. Severe COVID-19 can lead to critical complications, including respiratory distress and increased mortality rates [2]. One such complication is the development of “white lung” on chest radiographs (e.g., X-ray), characterized by extensive inflammation and fluid accumulation affecting 70%–80% of the lung area [3]. The appearance of white lung signals a critical stage in COVID-19 patients, profoundly impairing lung function, often requiring mechanical ventilation and ICU admission, and substantially increasing mortality risk [1, 2]. Despite extensive research into the pathophysiology of COVID-19, the mechanisms underlying “white lung” remain poorly understood.

Here, we performed single-cell RNA sequencing analysis of bronchoalveolar lavage fluid (BALF) to characterize the pathophysiology of “white lung” in COVID-19 (Figure 1A). BALF samples were collected from 16 patients with moderate (MO, n = 3), severe (SE, n = 6), and “white lung” (WL, n = 7) syndrome, as well as from 3 healthy controls (HC) (Figure 1A). After quality control filtering (Supporting Information S1: Figure S1A–C), we obtained transcriptome data sets from 136,015 cells (mean = 7159 cells/sample). Using uniform manifold approximation and projection (UMAP), we identified 7 major cell types (Supporting Information S1: Figure S1D) and, through sub-clustering, 44 distinct cell states representing diverse respiratory cell types (Supporting Information S1: Figure S1E). UMAP visualization (Supporting Information S1: Figure S1F) revealed substantial inter-group heterogeneity. The distribution of seven major clusters was portrayed through R_O/E (Supporting Information S1: Figure S1G) [1]. We observed an obvious expansion of NK and neutrophils in COVID-19 patients with “white lung” (Supporting Information S1: Figure S1G–J, Figure 1B). However, NK cells comprised < 0.5% of the total cell population in these patients (Supporting Information S1: Figure S1I), implying that their expansion is unlikely to be the primary driver of this complication. In contrast, neutrophils constituted up to 85% of BALF cells in COVID-19 patients with “white lung,” whereas this proportion did not exceed 25% in any other group (Figure 1B, Supporting Information S1: Figure S1H). PCA analysis clearly distinguished neutrophils from “white lung” patients from those in controls and patients with moderate and severe COVID-19 (Supporting Information S1: Figure S2A,B). Among BALF immune cells, neutrophils exhibited a significant association with “white lung” patients (Supporting Information S1: Figure S2C). These results suggested that neutrophil infiltration may be a key driver of “white lung” development in COVID-19.

Sub-clustering of neutrophils revealed 11 transcriptionally distinct subtypes: 2 immature, 2 mature, 3 aged, and 5 homeostatic subsets (Figure 1C). All neutrophil subsets were enriched in COVID-19 patients with “white lung,” further supporting the key role of neutrophil infiltration in this severe complication (Supporting Information S1: Figure S2D). Partition-based graph abstraction (PAGA) analysis identified two different neutrophil differentiation trajectories culminating in aged subsets (Supporting Information S1: Figure S2E). Homeostatic neutrophils seemed to represent transitional stages bridging immature and aged subsets (Supporting Information S1: Figure S2E), potentially offering therapeutic targets.

BALF Neutrophils in COVID-19 patients with “white lung” harbored two LDNs (Low-density neutrophils) clusters designated Neu_Immature_01/02 (Figure 1C). LDNs, primarily produced during pathological conditions (e.g., severe infection and sepsis during emergency myelopoiesis), are linked to dysfunctional immune responses characterized by both immunosuppression and inflammation [4]. The Neu_Immature_01/02 clusters highly expressed multiple ISGs (ISG15, IFITM1, IFITM3, and RSAD2) as well as genes involved in neutrophil extracellular trap (NET) formation (MPO, ELANE, and PRTN3), which are implicated in severe infection (Figure 1D). These clusters also expressed PADI4, a key NETosis (NETs) regulator (Figure 1D). NETs have been implicated in the pathogenesis of severe infectious disease [5]. Additionally, Neu_Immature_01/02 expressed CD24, OLFM4, LCN2, and BPI genes, previously correlated with poor outcomes in severe infection [4]. The other neutrophil subsets also highly expressed NET-related genes, highlighting systemic dysregulation of neutrophil responses (Figure 1D). In addition to NET formation, neutrophils from “white lung” patients released pro-inflammatory molecules (e.g., S100A8/9/12, CCL3/4, and CXCL8), known to trigger cytokine storms in COVID-19 (Figure 1D) [1]. S100A8/A9/A12, key drivers of the COVID-19 cytokine storm, were significantly upregulated in neutrophils from “white lung” patients (Supporting Information S1: Figure S2G). Using previously defined inflammatory and cytokine scores [1], we determined that neutrophils were the primary source of inflammation in “white lung” patients (Supporting Information S1: Figure S2F). These results highlighted that exaggerated inflammatory response driven by neutrophils contribute to lung immunopathology and are likely a key factor in the development of “white lung” in COVID-19.

PD-L1 (CD274) and arginase 1 (ARG1), which inhibit T cell activation [4], were f highly expressed in neutrophils from COVID-19 patients with “with white” (Figure 1E). PDL1⁺ neutrophils (All subsets; Figure 1E) have been shown to exert suppressive functions in cancer, HIV-1 infection, and lymphoid tissue (lymph nodes, spleen, and blood) after LPS exposure. ARG1⁺ neutrophils (Immature neutrophils; Figure 1E) deplete arginine and impede T cell function in severe infection [4]. The ARG1⁺ cells, mainly immature neutrophils, overlapped with PDL1-expressing cells (Figure 1E), suggesting the existence of dysfunctional, potentially suppressive neutrophils in “white lung” patients. Specifically, strong interactions were identified between immature neutrophils and effector and exhausted T cells (e.g., CD4_Exhaustion, CD8_Effector_GZMA) (Figure 1F). Multiple ligand-receptor (L-R) pairs, including HLA-E_KLRD1, HLA-E_KLRC1/2, and HLA-E_KLRK1, exhibited strong interaction potential in “white lung” patients (Supporting Information S1: Figure S2J). The HLA-E_KLRD1/C1/C2/K1 signaling pathway is implicated in T cell dysfunction and viral persistence during chronic viral infections [1]. These findings provide a basis for future functional studies investigating the immunopathogenesis and potential therapeutic strategies for “white lung” in COVID-19.

To investigate functional differences among neutrophils, we analyzed differentially expressed genes (DEGs) in each patient group. Compared to healthy controls, we identified 4682, 758, and 1131 upregulated genes in neutrophils from patients with moderate, severe, and “white lung” COVID-19, respectively (Supporting Information S1: Figure S2H). Of these, 471 DEGs were exclusive to “white lung” COVID-19 (Supporting Information S1: Figure S2H), and enriched in pathways linked to neutrophil activation and degranulation (Supporting Information S1: Figure S2I). Overactive neutrophil response and degranulation can promote NET formation, exacerbate inflammation and tissue damage, as well as contribute to the pathology. Accordingly, neutrophils from “white lung” patients showed higher neutrophil activation and degranulation scores, further implicating dysregulated neutrophil responses in the pathogenesis of “white lung” (Figure 1H, Supporting Information S1: Figure S2I). One homeostatic subtype (Neu_Homeostatic_04) and two aged subtypes (Neu_Aged_02/03) were fully activated (Supporting Information S1: Figure S3A), indicating their contribution to lung damage in “white lung” patients. These subsets consistently expressed high levels of activation- and degranulation-related genes (Supporting Information S1: Figure S3B–D). CXCL8 (IL-8) is crucial for neutrophil recruitment and activation at sites of infection, and its interaction with its receptor CXCR2 on neutrophil triggers priming, activation, and subsequent tissue damage. CXCL8 and CXCR2 expression was significantly higher in neutrophils from “white lung” patients compared to those from patients with moderate and severe cases, as well as healthy controls (Figure 1H). Hence, blocking the CXCL8-CXCR2 axis might provide a potential therapeutic target for controlling COVID-19-related white lung complication. Particularly, representative chest CT images from patients exhibiting “white lung” were presented in Supporting Information S1: Figure S4.

In summary, our study reveals substantial neutrophil infiltration in the lungs of patients with COVID-19–associated “white lung.” These infiltrated neutrophils contribute to immunopathological damage by promoting NET formation, exacerbating inflammation, suppressing T-cell responses, and undergoing excessive activation and degranulation. Our findings provide insights into the immunopathology of “white lung” in COVID-19, highlighting the detrimental role of neutrophils and suggesting potential therapeutic targets.

Yi Wang and Guogang Xu conceived the study. Yi Wang, Guogang Xu, and Xiong Zhu designed the study. Yi Wang, Guogang Xu, and Xiong Zhu supervised this project. Xiaoxia Wang performed the experiments. Yi Wang, Guogang Xu, and Xiong Zhu founded the study and contributed the reagents and materials. Yi Wang contributed to the analysis tools. Yi Wang performed the software. Yi Wang and Xiaoxia Wang analyze the data. Yi Wang drafted the original paper. Yi Wang and Guogang Xu revised and edited this paper. Yi Wang, Guoagang Xu, and Xiong Zhu reviewed the paper. All authors have read and approved the final manuscript.

Ethics approval for this study was granted by the Ethics Committee of Sanya People's Hospital (Ethical approval No. SYPH-2023-03) and was conducted in compliance with the Declaration of Helsinki for medical research involving human subjects. Written informed consent was obtained from all participants.

The authors declare no conflicts of interest.