Targeting neutrophils extracellular traps (NETs) reduces multiple organ injury in a COVID-19 mouse model

K18-hACE2 humanized mice (B6·Cg-Tg(K18-ACE2)2Prlmn/J) were obtained from The Jackson Laboratory and were bred in the Centro de Criação de Animais Especiais (Ribeirão Preto Medical School/University of São Paulo). This mouse strain has been previously used as the model for SARS-CoV-2-induced disease and it presents signs of diseases, and biochemical and lung pathological changes compatible with the human disease [22]. Mice had access to water and food ad libitum. The manipulation of these animals was performed in Biosafety Levels 3 (BSL3) facility and the study was approved by the Ethics Committee on the Use of Animals of the Ribeirão Preto Medical School, University of São Paulo (#066/2020).

Male K18-hACE2 mice, aged 8 weeks, were infected with 2 × 10⁴ PFU of SARS-CoV-2 (in 40 µL) by the intranasal route. Uninfected mice (n = 5) were given an equal volume of PBS through the same route. On the day of infection, 1 h before virus inoculation, animals were treated with DNase I (10 mg/kg, s.c., Pulmozyme, Roche) (n = 6) or vehicle (PBS, s.c.) (n = 6). DNase I was also given once a day until 5 days post-infection. Body weight was evaluated on the baseline and all days post-infection. The right lung was collected, harvested, and homogenized in PBS with steel glass beads. The homogenate was added to TRIzol reagent (1:1), for posterior viral titration via RT-qPCR, or to lysis buffer (1:1), for ELISA assay, and stored at − 70 °C. The left lung was collected in paraformaldehyde (PFA, 4%) for posterior histological assessment.

Five μm lung, heart, and kidney slices were submitted to Hematoxylin and Eosin staining. A total of 10 photomicrographs in 40X magnification per animal were randomly obtained using a microscope ScanScope (Olympus) and Leica. Morphometric analysis was performed by the protocol established by the American Thoracic Society and European Thoracic Society (ATS/ERS) [23].

Plasma or homogenate from the lung was incubated overnight in a plate pre-coated with anti-MPO antibody (Thermo Fisher Scientific; cat. PA5-16672) at 4 °C. The plate was washed with PBS-T (Phosphate-Buffered Saline with Tween 20). Next, samples were incubated overnight at 4 °C. Finally, the plate with samples was washed and over MPO-bound DNA was quantified using the Quant-iT PicoGreen kit (Invitrogen; cat. P11496).

Lung homogenate was added to the RIPA buffer solution (Sigma-Aldrich, cat. R0278) and centrifuged at 10,000 g at 4 °C for 10 min. The supernatant was collected. The ELISA method was performed to detect the concentration of cytokines and chemokines using kits from R&D Systems (DuoSet), according to the manufacturer’s instructions. The following targets were evaluated: TNF-α, IL-6, IL-10, CXCL1, CCL2, and CCL4.

Lungs were harvested and fixed with PFA 4%. After dehydration and paraffin embedding, 5 μm sections were prepared. The slides were deparaffinized and rehydrated by immersing the through Xylene and 100% Ethanol 90% for 15 min, in each solution. Antigen retrieval was performed with 1.0 mM EDTA, 10 mM Trizma-base, pH 9·0 at 95 °C for 30 min. Later, endogenous peroxidase activity was quenched by incubation of the slides in 5% H₂O₂ in methanol for 15 min. After blocking with IHC Select Blocking Reagent (Millipore, cat. 20773-M) for 2 h at room temperature (RT), the following primary antibodies were incubated overnight at 4 °C: goat polyclonal anti-myeloperoxidase (anti-MPO, R&D Systems, cat. AF3667, 1:100) and rabbit polyclonal, anti-histone H3 (H3Cit; Abcam; cat. ab5103; 1:100). The slides were then washed with TBS-T (Tris-Buffered Saline with Tween 20) and incubated with secondary antibodies donkey anti-goat IgG Alexa Fluor 488 (Abcam, cat. ab150129) and alpaca anti-rabbit IgG AlexaFluor 594 (Jackson ImmunoReseacher; Cat. 611-585-215; 1:1000). Autofluorescence was quenched using the TrueVIEW Autofluorescence Quenching Kit (Vector Laboratories, cat. SP-8400-15). Slides were then mounted using Vectashield Antifade Mounting Medium with DAPI (Vector Laboratories, Cat# H-1200-10). Images were acquired by Axio Observer combined with LSM 780 confocal microscope (Carl Zeiss) at 63X magnification at the same setup (zoom, laser rate) and tile-scanned at 4 fields/image. Images were analyzed with Fiji by Image J.

Renal dysfunction was assessed by the levels of blood creatinine, and creatine kinase-MB was used as an index of cardiac lesions. The determinations were performed using a commercial kit (Bioclin).

Peripheral blood samples were collected from healthy controls by venipuncture and the neutrophil population was isolated by Percoll density gradient (GE Healthcare; cat. 17-5445-01). Isolated neutrophils (1.5 × 10⁷ cells) were stimulated with 50 nM of PMA (Sigma-Aldrich; cat. P8139) for 4 h at 37 °C. The medium containing NETs was centrifuged at 450 g to remove cellular debris for 10 min, and NETs-containing supernatants were collected and centrifuged at 18,000 g for 20 min. Supernatants were removed, and DNA pellets were resuspended in PBS. NETs were then quantified with a GeneQuant (Amersham Biosciences Corporation).

Lung tissue was harvested for detection of apoptotic cells in situ with Click-iT Plus TUNEL Assay Alexa Fluor 488, according to the manufacturer’s instructions (Thermo Fisher Scientific; cat. C10617). Human alveolar basal epithelial A549 cells (5 × 10⁴) were maintained in DMEM and cultured with purified NETs (10 ng/ml) pretreated, or not, with DNase I (0·5 mg/ml; Pulmozyme, Roche). The cultures were then incubated for 24 h at 37 °C. Viability was determined by flow cytometric analysis of Annexin V staining.

Lung tissue was harvested and digested with type 2 collagenase to acquire cell suspensions. Cells were then stained with Fixable Viability Dye eFluor 780 (eBioscience; cat. 65-0865-14; 1:3000) and monoclonal antibodies specific for CD45 (BioLegend; clone 30-F11; cat. 103138; 1:200), CD11b (BD Biosciences; clone M1/70; cat. 553311) and Ly6G (Biolegend; clone 1A8; cat. 127606) for 30 min at 4 °C. A549 cells (1 × 10⁵) were stained with FITC ApoScreen Annexin V Apoptosis Kit (SouthernBiotech; cat. 10010-02), according to the manufacturer’s instructions. Data were collected on a FACSVerse (BD Biosciences) and analyzed with FlowJo (TreeStar).

Statistical significance was determined by either two-tailed unpaired Student t-test, and by one-way or two-way ANOVA followed by Bonferroni’s post hoc test. P < 0·05 was considered statistically significant. Statistical analyses and graph plots were performed and built using GraphPad Prism 9·3·1 software.

To investigate the role of NETs in COVID-19 pathophysiology we used K18-hACE2 mouse intranasally infected with SARS-CoV-2; [22, 24]. Infected mice were treated with saline or DNAse I (10 mg/kg; s.c.) administered 1 h before virus infection and once daily up to 4 days after infection (Fig. 1a). We evaluated body weight loss from baseline measure (basal) and clinical scores, as read out of disease progression (Additional file 1: Table S1). Our data showed that DNase I treatment attenuates both body weight loss and clinical score caused by SARS-CoV-2 infection compared to vehicle-treated mice (Fig. 1b, c).

We observed an increase in NETs concentration (Fig. 1f, g) and the number of neutrophils in the lung of infected mice upon induction of the COVID-19 model (Fig. 1d, e). These data corroborate some findings published by either our group or other authors, showing the presence of systemic levels of NETs, as well as localized in pulmonary tissue from COVID-19 patients [8‐10].

DNase I treatment did not alter the number of neutrophils that infiltrated into the lung tissue of SARS-CoV-2 infected mice. The FACS analysis showed that the frequency and absolute numbers of neutrophils (Ly6G + CD11b + cells; Fig. 1d, e) were not altered by the treatment. Similar results were observed when analyzing CD45 + cell populations (Additional file 1: Fig. S1). However, we demonstrate that DNAse I treatment of SARS-CoV-2-infected mice resulted in a substantial reduction of NETs production that associates with a lower disease score (Fig. 1h). These data suggest that NETs play a crucial role in SARS-CoV-2 infection.

Lung inflammation is the primary cause of life-threatening respiratory disorders in critical and severe forms of COVID-19 [3, 4, 25]. NETs have been identified in the lung tissue of SARS-CoV-2 infected patients and lung injury experimental animal models [8, 9, 26]. Thus, we next investigated the effect of NETs degradation by DNase I treatment of the COVID-19 mouse model.

We noticed an extensive injury in the lung tissue of K18-hACE2-infected mice with interstitial leukocyte infiltration. The alveolar units showed architectural distortion compromising the alveolar-capillary barrier. DNase I treatment was able to prevent lung damage caused by SARS-CoV-2 infection (Fig. 2a). Moreover, analyzing the area fraction, as a score of septal thickness, DNase I-treatment reduced the area fraction in SARS-CoV-2-infected mice and was associated with a lower concentration of NETs released (Fig. 2b). The reduction in lung pathology was associated with a reduction of pro-inflammatory cytokines/chemokines, especially IL-6 and CXCL-1 in the lung tissue of mice treated with DNase I (Fig. 2c). These results indicate that pharmacological inhibition of NETs could be a novel approach to inhibit SARS-CoV-2-induced lung pathology.

COVID-19 is a systemic viral disease that can affect other vital organs besides the lungs, such as the heart and kidney [27‐30]. To investigate this context, we harvested the heart and kidneys of animals 5 days post-infection and evaluated whether NETs inhibition could change this phenotype. We observed that the heart of mice infected with SARS-CoV-2 showed pathological changes in cardiac tissue with diffuse and sparse cardiac inflammatory cell infiltration and perivascular injury (Fig. 3a). Moreover, after treatment, there is a clear reduction of inflammatory cells in heart tissue, leaving “injured” (fine) fibers and interstitial edema, similar to what is seen in autoimmune myocarditis. In addition, plasma Creatine Kinase MB (CK-MB) concentration in the infected group was higher than in control mice (Fig. 3b). Interestingly, DNase I treatment attenuated the pathology and CK-MB concentration.

Finally, we aimed to investigate whether SARS-CoV-2 infection induces kidney damage in K18-hACE2 mice. Consistent with lung and heart pathological changes, we found the presence of ischemic tubulointerstitial nephritis in the COVID-19 model, mimicking acute tubular necrosis with cellular glomerulitis (Fig. 3c). Furthermore, the levels of creatinine in the blood of infected mice were higher in comparison with uninfected animals (Fig. 3d). As expected, DNase I treatment reduced kidney injury (Fig. 3c, d). Collectively, these findings indicate that pharmacological inhibition of NETs with DNase I prevents multi-organ dysfunction in COVID-19.

To understand the possible role of NETs in the pathophysiology of COVID-19, we explored the hypothesis that NETs could be involved in lung damage accompanied by the disease, as previously described [9, 31].

To this end, lung analysis of k18-hACE2 mice infected with SARS-CoV-2 reveals a massive presence of apoptotic cells after 5 days of infection. DNase I treatment was able to prevent lung apoptosis caused by the COVID-19 mouse model (Fig. 4a). Our analysis showed a reduction of approximately 60% in TUNEL-positive cells after NETs degradation (Fig. 4b).

We found that exposure of A549 cells to purified NETs significantly increased the percentage of apoptotic cells in comparison with untreated cells. Importantly, the pretreatment of purified NETs with DNase I prevented NETs-induced epithelial (Fig. 4c, d) apoptosis to similar levels observed in untreated cells. Taken together, these results are evidence that DNase I exhibits a protective effect on NETs deleterious functions in lung tissue and could be a potential strategy to block organ injury during COVID-19.