key: cord-0992030-sxe5ydon authors: Nicolai, Leo; Leunig, Alexander; Brambs, Sophia; Kaiser, Rainer; Joppich, Markus; Hoffknecht, Marie‐Louise; Gold, Christoph; Engel, Anouk; Polewka, Vivien; Muenchhoff, Maximilian; Hellmuth, Johannes C.; Ruhle, Adrian; Ledderose, Stephan; Weinberger, Tobias; Schulz, Heiko; Scherer, Clemens; Rudelius, Martina; Zoller, Michael; Keppler, Oliver T.; Zwißler, Bernhard; von Bergwelt‐Baildon, Michael; Kääb, Stefan; Zimmer, Ralf; Bülow, Roman D.; von Stillfried, Saskia; Boor, Peter; Massberg, Steffen; Pekayvaz, Kami; Stark, Konstantin title: Vascular neutrophilic inflammation and immunothrombosis distinguish severe COVID‐19 from influenza pneumonia date: 2020-12-20 journal: J Thromb Haemost DOI: 10.1111/jth.15179 sha: 7030bff509a4291c59f856064a7fdbf51f434e63 doc_id: 992030 cord_uid: sxe5ydon OBJECTIVE: Infection with severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) can lead to severe pneumonia, but also thrombotic complications and non‐pulmonary organ failure. Recent studies suggest intravascular neutrophil activation and subsequent immune cell–triggered immunothrombosis as a central pathomechanism linking the heterogenous clinical picture of coronavirus disease 2019 (COVID‐19). We sought to study whether immunothrombosis is a pathognomonic factor in COVID‐19 or a general feature of (viral) pneumonia, as well as to better understand its upstream regulation. APPROACH AND RESULTS: By comparing histopathological specimens of SARS‐CoV‐2 with influenza‐affected lungs, we show that vascular neutrophil recruitment, NETosis, and subsequent immunothrombosis are typical features of severe COVID‐19, but less prominent in influenza pneumonia. Activated neutrophils were typically found in physical association with monocytes. To explore this further, we combined clinical data of COVID‐19 cases with comprehensive immune cell phenotyping and bronchoalveolar lavage fluid scRNA‐seq data. We show that a HLADR(low) CD9(low) monocyte population expands in severe COVID‐19, which releases neutrophil chemokines in the lungs, and might in turn explain neutrophil expansion and pulmonary recruitment in the late stages of severe COVID‐19. CONCLUSIONS: Our data underline an innate immune cell axis causing vascular inflammation and immunothrombosis in severe SARS‐CoV‐2 infection. Since its animal-human transmission in late 2019, a novel coronavirus termed severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has spread globally, infecting millions within months. [1] [2] [3] Coronavirus disease 2019 (COVID-19) is characterized by respiratory failure in severe cases, but is also associated with non-pulmonary organ failure and a systemic prothrombotic state. 4 We and others have linked intravascular neutrophil and platelet activation, neutrophil extracellular trap formation (NETosis), and subsequent activation of the coagulation cascade, a process termed "immunothrombosis," to COVID-19 progression, providing a possible explanation for multi-organ involvement. [5] [6] [7] However, neutrophil activation and NETosis have been identified as a common effector function in a range of inflammatory disorders, calling into question the specificity of immunothrombosis in contributing to SARS-CoV-2-associated acute respiratory distress syndrome (ARDS). [8] [9] [10] In addition, upstream regulation of immunothrombosis, which might be amenable to pharmacological treatment, remains poorly understood. 11 Detailed methodology is provided in supporting information. Details of the analyzed cohorts are stated in Tables S1 and S2 in supporting information. COVID- 19 Approach and results: By comparing histopathological specimens of SARS-CoV-2 with influenza-affected lungs, we show that vascular neutrophil recruitment, NETosis, and subsequent immunothrombosis are typical features of severe COVID-19, but less prominent in influenza pneumonia. Activated neutrophils were typically found in physical association with monocytes. To explore this further, we combined clinical data of COVID-19 cases with comprehensive immune cell phenotyping and bronchoalveolar lavage fluid scRNA-seq data. We show that a HLADR low CD9 low monocyte population expands in severe COVID-19, which releases neutrophil chemokines in the lungs, and might in turn explain neutrophil expansion and pulmonary recruitment in the late stages of severe COVID-19. Our data underline an innate immune cell axis causing vascular inflammation and immunothrombosis in severe SARS-CoV-2 infection. COVID-19, immunopathology, immunothrombosis, monocytes, neutrophils, SARS-CoV-2 • It remains unclear if COVID-19-related immunothrombosis is a general feature of (viral) pneumonia, and how it is regulated in severe SARS-CoV-2 infection. • We compared histopathological specimens of influenza and COVID-19 lungs and utilized comprehensive immune cell phenotyping, functional in vitro assays, and bronchoalveolar lavage fluid scRNA-seq data to investigate this further. • We identified vascular neutrophil recruitment, NETosis, and immunothrombosis to be key features of COVID-19 as compared to a less prominent role in influenza. • CD9 low HLA-DR low monocytes are a prominent feature of severe COVID-19 and release neutrophil-attracting chemokines in the lungs, causing pulmonary neutrophil recruitment. • Histopathological comparison of COVID-19 lungs with influenza specimens shows immunothrombotic vascular occlusions. Table S1 were used for all analyses except histopathology, for which a total of 13 patients detailed in Table S2 were used. For flow cytometry and longitudinal analysis of clinical data patients with pre-existing diseases like severe kidney or liver failure, immunosuppressive therapy, autoimmune disease, chronic inflammation, or acute extracorporeal membrane oxygenation therapy were excluded. COVID-19 patients consisted of CoV-Sev patients requiring intensive care treatment and hospitalized CoV_int patients treated on the ward. We included age-matched control patients without infectious disease and control non-COVID pneumonia cases (Table S1 ). Immunohistochemistry stainings were performed on formalin-fixed paraffinized lung specimens as previously described. 12 Specimens were from six COVID-19 and H1N1 (n = 4) or seasonal influenza (n = 3) patients obtained via the DeRegCOVID register. Preparation and analysis of patient blood were performed as previously described 5 (Table S3 in Count matrices for the scRNA-seq dataset were downloaded from GEO (Accession GSE145926) 14 and analyzed using Seurat. Gene-set enrichment analyses were conducted on Biological Process subset of GeneOntology (GO-BP). 15 Values of individual patients are represented as dots in graphs, and data in main text and figures is mean ± standard error of the mean unless otherwise noted. Unpaired, two tailed t-tests, or in case of significant F-test, Mann-Whitney U tests were used. For regression analysis, the black line represents best-fit line, shaded area with dashed lines is 95% confidence interval, r 2 and P value (slope non-zero) are shown in plots. Excel (Microsoft) and Prism (GraphPad) software were used for data analysis, Illustrator (Adobe Inc) for visualization. To better understand the immunopathology of severe SARS-CoV-2 infection, we phenotyped peripheral blood leukocytes in controls (Ctrl) and non-COVID-19 pneumonia patients (Ctrl_pneu) compared to COVID-19 patients on normal wards (CoV_int) and requiring intensive care treatment (ARDS cohort, CoV_sev; Figure S1A -C in supporting information). While CoV_int patients showed comparable leukocyte counts to healthy controls, a hallmark of CoV_sev was an expansion of the neutrophil granulocyte compartment, translating into elevated leukocyte counts and an increased neutrophil-lymphocyte ratio (NLR; Figure 1A and Figure S1A ). The NLR also correlated with disease severity, as measured by oxygenation index (PaO 2 /FiO 2 , Horowitz index; Figure 1B , Figure S1C ). In fact, neutrophil counts were the only quantitative immune cell parameter correlating positively with disease severity ( Figure 1C , Figure S1D ). Longitudinal sampling in the CoV_sev cohort revealed a rise in peripheral neutrophil counts just before manifestation of ARDS requiring mechanical ventilation ( Figure 1D ). In contrast, CoV_int patients not requiring mechanical ventilation showed stable neutrophil counts throughout the disease course ( Figure 1E ). This implicates an important role of neutrophils in disease progression. To further examine this link, we histopathologically compared lung specimens of deceased COVID-19 (n = 6) and influenza (H1N1 or seasonal) pneumonia cases (n = 7). Total neutrophil recruitment as well as activation did not differ between SARS-CoV-2 and influenza pneumonia ( Figure 1F is also a vascular disease, and show that pulmonary immunopathology in SARS-CoV-2 infection might differ from other viral infections, requiring novel treatment strategies. 18, 19 On the other hand, the identified correlation with disease severity and involvement in vascular inflammation suggests neutrophils to be causative in the development of organ damage and mortality in COVID-19. 11 In kidney and heart specimens we could indeed see similar trends in immunothrombotic occlusions ( Figure S1G, H) . Therefore, it is of paramount importance to understand neutrophil recruitment and mobilization, especially to the lungs. As neutrophils are mobilized and recruited to the lungs at ad- Figure 3A , Figure S2F ). In contrast, we identified a classical monocyte subpopulation, MS9, which was robustly upregulated in severe COVID-19. In line with recent scRNA-seq data, this population showed particularly low CD9 expression levels with upregulation of complement C1q-receptor (CD93), scavenger-receptor CD36, and pattern-recognition receptor CD14, indicating enhanced phagocytic potential ( Figure 3B and Figure S2C -G). [23] [24] [25] Indeed, this CD14 hi CD9 low MS9 cluster correlated significantly with disease severity and represented one of the largest subpopulations comprising approximately 20% of all monocytes in severe COVID-19 ( Figure 3C ). To examine the role of this monocyte subpopulation in more detail we performed in vitro assays of sorted monocyte subsets from healthy individuals (see supporting information). Indeed, supernatant of stimulated CD14 hi CD16 lo monocytes prompted increased migration by neutrophils in a Boyden chamber migration assay ( Figure S2H ). Furthermore, neutrophils incubated with supernatant specifically from the CD9 lo HLA-DR lo monocyte subset showed increased expression of neutrophil activation marker CD163, compared to supernatant from CD16 lo CD9 int/hi classical monocytes ( Figure S2I ). To gain mechanistic insight into monocytes/monocytic macrophages recruited to the lungs in COVID-19, we re-analyzed publicly analyses. 26 Indeed, processes involving neutrophil chemotaxis were robustly upregulated in FABP4pulmonary monocytes/ macrophages in severe COVID-19, specifically CXCL8 (IL8), CCL2, CCL3, CCL4, CCL7, CXCL3, CCL3L1, and CCL4L2 ( Figure 4A -B, Figure S3B ). By analyzing chemokine transcriptomes across all discovered cell types in BALF, we were able to confirm that recruited monocytic macrophages were the key source of neutrophil-attracting chemokines in the failing lungs ( Figure 4C ). In line with the proinflammatory CD9 low peripheral blood monocyte subset MS9 identified in CoV_sev patients, CD9 expression by monocytic macrophages in the BALF dropped in the severe group. This suggests that blood-monocyte subset MS9 may give rise to a pulmonary monocytic-macrophage subset, which in turn contributes to neutrophil activation ( Figure S3C ). Clinically, COVID-19 presents heterogeneously with lung involvement, but also central nervous system and gastrointestinal symptoms. 4 In severe disease, however, this seems to converge into severe immunopathology, i.e., host damage by a dysregulated immune response. 11 The mechanisms and regulation of immunopathology in SARS-CoV-2 infection are so far incompletely understood. Our data, in agreement with prior studies, The HLADR low CD9 low monocyte population in severe COVID-19 releases neutrophil chemokines in the lung. In silico reanalysis of publicly available single cell RNA sequencing data published by Liao et al. 14 A, Top eight up-and downregulated GO-BP pathways of severe versus mild cases for monocytic macrophages in bronchoalveolar lavage. Fold enrichment is relative to mild cases. B, Volcano plot of relative fold change of severe versus mild cases for monocytic macrophages. CXCL8, CCL2, CCL3, CCL4, CCL7, CXCL3, CCL4L2, and CCL3L1, when significantly increased, are marked in red and annotated. C, Dot plots of average and percentage expression for different cell groups counts preceding respiratory failure. Neutrophil activation and NETosis have been implicated in a wide range of diseases ranging from atherosclerosis to cancer. 9, 28 To better understand the involvement of these cells in SARS-CoV-2 infection we compared histopathological specimens of COVID-19 lungs with lethal viral pneumonia caused by H1N1 or seasonal influenza virus. Our data underline neutrophil-driven immunothrombosis as a key element of severe COVID-19 as immunothrombotic vessel occlusion and NETosis were strongly elevated compared to influenza pneumonia. While NETosis and innate immunity have also been implicated in influenza our data point to a substantially increased contribution to immunopathology in SARS-CoV-2 infection. 29, 30 In addition, elevated vWF activity in severe COVID-19 points to increased endothelial activation and has also been implicated as a marker of NETosis. 16 This data might therefore link endothelii- CoV-2 infection. 19 So how are neutrophils recruited and activated? Our study points to an innate immune cell axis consisting of CD9 low monocytes that release proinflammatory, neutrophil-attracting chemokines in the failing lungs of severe COVID-19 patients. Interestingly, this HLA-DR low population of monocytes has also been identified in other COVID-19 studies using single cell RNA-sequencing of peripheral blood mononuclear cells, highlighting this subset as a general feature of this disease. 31, 32 Principal limitations of the presented study are the limited number of patients and histological specimens analyzed. However, our findings are in line with work from several other groups addressing vascular inflammation in COVID-19. 24, 31 In summary, our data provide evidence for an innate immune cell axis in vascular inflammation and immunothrombosis observed in severe SARS-CoV-2 infection. Targeting neutrophil-monocyte partnership might be a valuable therapeutic approach to dampen disease progression in COVID-19. All employed code for scRNA-seq analysis is available from GitHub. (https://github.com/mjopp ich/Covid Immune, https://github.com/ mjopp ich/scrna seq_cellt ype_predi ction). We thank Anna Titova for technical assistance. We would like to thank all CORKUM investigators and staff. The authors thank the patients and their families for their participation in the CORKUM registry. The study was supported by the German Registry of COVID-19 Autopsies (DeRegCOVID). Open access funding enabled and organized by Projekt DEAL. A novel coronavirus from patients with pneumonia in China Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia A new coronavirus associated with human respiratory disease in China Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study Immunothrombotic dysregulation in COVID-19 pneumonia is associated with respiratory failure and coagulopathy Neutrophil extracellular traps (NETs) as markers of disease severity in COVID-19. medRxiv Neutrophil extracellular traps (NETs) contribute to immunothrombosis in COVID-19 acute respiratory distress syndrome Platelets induce neutrophil extracellular traps in transfusion-related acute lung injury Neutrophil extracellular traps license macrophages for cytokine production in atherosclerosis Neutrophil extracellular traps enriched in oxidized mitochondrial DNA are interferogenic and contribute to lupus-like disease COVID-19: immunopathology and its implications for therapy Eosinophil-platelet interactions promote atherosclerosis and stabilize thrombosis with eosinophil extracellular traps Data-driven phenotypic dissection of AML reveals progenitor-like cells that correlate with prognosis Single-cell landscape of bronchoalveolar immune cells in patients with COVID-19 Gene ontology: tool for the unification of biology. The Gene Ontology Consortium Neutrophil activation and NETosis are the major drivers of thrombosis in heparin-induced thrombocytopenia Elevated von Willebrand factor antigen is an early plasma predictor of acute lung injury in nonpulmonary sepsis syndrome Association of COVID-19 inflammation with activation of the C5a-C5aR1 axis Endothelial cell infection and endotheliitis in COVID-19 In vivo two-photon imaging reveals monocyte-dependent neutrophil extravasation during pulmonary inflammation Monocytes are potent facilitators of alveolar neutrophil emigration during lung inflammation: role of the CCL2-CCR2 axis Monocytes control second-phase neutrophil emigration in established lipopolysaccharide-induced murine lung injury Human monocyte subsets and phenotypes in major chronic inflammatory diseases Elevated calprotectin and abnormal myeloid cell subsets discriminate severe from mild COVID-19 Severe COVID-19 is marked by a dysregulated myeloid cell compartment Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources Pulmonary vascular endothelialitis, thrombosis, and angiogenesis in Covid-19 Cxcr1 and cxcr2 chemokine receptor agonists produced by tumors induce neutrophil extracellular traps that interfere with immune cytotoxicity The role of platelets in mediating a response to human influenza infection Neutrophils-related host factors associated with severe disease and fatality in patients with influenza infection A single-cell atlas of the peripheral immune response in patients with severe COVID-19 Complex immune dysregulation in COVID-19 patients with severe respiratory failure The authors declare no conflicts of interest. Data is available upon reasonable request from authors.