Impaired balance between neutrophil extracellular trap formation and degradation by DNases in COVID-19 disease

Abstract

Background: Thrombo-inflammation and neutrophil extracellular traps (NETs) are exacerbated in severe cases of COVID-19, potentially contributing to disease exacerbation. However, the mechanisms underpinning this dysregulation remain elusive. We hypothesised that lower DNase activity may be associated with higher NETosis and clinical worsening in patients with COVID-19.

Methods: Biological samples were obtained from hospitalized patients (15 severe, 37 critical at sampling) and 93 non-severe ambulatory cases. Our aims were to compare NET biomarkers, functional DNase levels, and explore mechanisms driving any imbalance concerning disease severity.

Results: Functional DNase levels were diminished in the most severe patients, paralleling an imbalance between NET markers and DNase activity. DNase1 antigen levels were higher in ambulatory cases but lower in severe patients. DNase1L3 antigen levels remained consistent across subgroups, not rising alongside NET markers. DNASE1 polymorphisms correlated with reduced DNase1 antigen levels. Moreover, a quantitative deficiency in plasmacytoid dendritic cells (pDCs), which primarily express DNase1L3, was observed in critical patients. Analysis of public single-cell RNAseq data revealed reduced DNase1L3 expression in pDCs from severe COVID-19 patient.

Conclusion: Severe and critical COVID-19 cases exhibited an imbalance between NET and DNase functional activity and quantity. Early identification of NETosis imbalance could guide targeted therapies against thrombo-inflammation in COVID-19-related sepsis, such as DNase administration, to avert clinical deterioration.

Trial registration: COVERAGE trial (NCT04356495) and COLCOV19-BX study (NCT04332016).

Fig. 1

Plasma NETs markers correlate with COVID 19 clinical severity, biological markers of disease severity and predict hospitalization in non-severe patients. A–C We compared three NET biomarkers, MPO-DNA (A), H3cit (B), and H3cit-DNA complexes (C) between non-severe (oupatients) (n = 93), severe (without needs of oxygen therapy or with standard oxygen therapy) (n = 15) and critical COVID-19 patients (with high flow nasal oxygen, invasive intubation or prone ventilation) (n = 37). Results are expressed as fold change compared to healthy donors (n = 21). Statistical evaluation was performed using a Compound Poisson-Gamma model adjusted for age, sex and BMI. D–I Spearman correlation between three NET biomarkers, MPO-DNA (D, G), H3cit (E, H), and H3cit-DNA complexes (F, I) and two biological disease severity markers, i.e. CRP (D–F) and neutrophils/lymphocytes ratios (G–I) across all COVID-19 patients (n = 145). Statistical analysis are adjusted for age, gender and BMI, after log transformation of CRP variable. J–K Receiver-operating characteristic curve for prediction of hospitalization in COVID-19 outpatients using MPO–DNA, H3cit and H3cit-DNA complexes levels (J)

Fig. 2

The level of functional DNase is lower in hospitalized patients compared to outpatients, resulting in an imbalanced NET markers/DNase ratio. A We examined the level of functional DNase (DNase) between non-severe (n = 93), severe (n = 15) and critical COVID-19 patients (n = 37). Results are expressed as fold change compared to healthy donors (n = 21). B, C Spearman correlation explores the relationship between the level of functional DNase and two disease severity markers, CRP (B), and neutrophils/lymphocytes ratio (C) across all patients (n = 145). D–F We compared the ratio between NET biomarkers [MPO-DNA (D), H3cit (E), and H3cit-DNA (F)] and DNase between non-severe (n = 93), severe (n = 15), and critical COVID-19 patients (n = 37) Results are expressed as a fold change ratio. G–L Spearman correlation investigates the correlation between the ratio of NET markers [MPO-DNA (G, J), H3cit (H, K), and H3cit-DNA (I, L)] over DNase and two severity markers, CRP (G–I), and neutrophils/lymphocytes ratio (J–L) in all COVID-19 patients (n = 145). Comparisons between groups were performed using a Compound Poisson-Gamma model adjusted for age, sex and BMI. Correlation were realized using Spearman method, adjusted for age, gender and BMI, after log transformation of CRP variable

Fig. 3

The equilibrium between NET biomarkers and DNase1 or DNase1L3 antigens is disrupted in severe patients. A–D Comparison of the quantity of DNase1 protein (A) and the ratios between NET biomarkers [MPO-DNA (B), H3cit (C), H3cit-DNA (D)] and DNase1 antigen among non-severe (n = 32), severe (n = 15), and critical COVID-19 patients (n = 37). Results are expressed as fold change in comparison to healthy donors (n = 7) and fold change ratio. E–H Comparison of the quantity of DNase1L3 protein (E) and the ratios between NET biomarkers [MPO-DNA (F), H3cit (G), and H3cit-DNA (H)] and DNase1L3 antigen among non-severe (n = 32), severe (n = 15), and critical (n = 37) COVID-19 patients. Results are expressed as fold change in comparison to healthy donors (n = 7) and fold change ratio. Statistical analyses were conducted using a Compound Poisson-Gamma model adjusted for age, sex and BMI

Fig. 4

Defective dendritic cells and plasmacytoid dendritic cells in severe COVID-19 patients. A, B Violin plot displaying normalized gene expression levels of DNASE1 (A) and DNASE1L3 (B) in blood cells from healthy donors (n = 24) (data analysis from Stephenson E. and al (19)). Means are indicated on the right. C–F Compound Poisson-Gamma model adjusted for age, sex and BMI compared the count of CD123 + pDCs (C), CD11c + DCs (D), CD141 + cDC1 (E), and CD1c + cDC2 (F) between severe (n = 27) and critical (n = 37) patients. Results are expressed as frequency of alive lineage negative cells. The horizontal line defines median value. (G) Violin plot showing DNASE1L3 normalized gene expression in pDC from healthy donor (n = 24), non-severe (n = 25), severe (n = 15) and critical (n = 16) COVID-19 patient (data analysis from Stephenson et al. [19]). Statistical analyses were conducted using the Kruskal–Wallis test with Dunn’s multiple comparisons