The construct for hexahistidine (His)-tagged N-protein of SARS-CoV-2 was kindly provided by Dr. Feng Cong from Guangdong Laboratory Animals Monitoring Institute. The whole sequence of N-protein was prepared via PCR, confirmed via sequencing, and inserted into pET28a vector for expression as a His-tagged fusion protein. The protein was purified via nickle affinity chromatography (BBI Life Sciences, Shanghai, China) and followed by dialysis to remove extra salt and imidazole. To remove the contaminated endotoxin, the recombinant protein was processed with ToxinEraser™ Endotoxin Removal Kit (GenScript, Piscataway, NJ, USA). The final endotoxin concentration was examined by ToxinSensor™ Chromogenic LAL Endotoxin Assay Kit (GenScript). The endotoxin levels in different batches were consistently < 0.1 EU/μg protein (<0.01 ng/μg). The final recombinant protein was aliquoted and stored at -80°C freezer. To neutralize the possible protein-bound LPS in the recombinant protein, N-protein was pretreated with polymyxin B (250 µg/ml, MilliporeSigma, St. Louis) for 1 h at 37°C before all experiments. For denaturation of N-protein, N-protein samples were incubated in a heating block at 80°C for 1 h. To block the effect of N-protein in vitro, N-protein (1 mg/ml) was preincubated with an anti-N-protein antibody (Genscript, 1 mg/ml) for overnight at 4°C before the stimulation experiment.

C57BL/6 mice (6-8 weeks old) were purchased from Shanghai Laboratory Animal Center (Shanghai, China). C3H/HeJ (resistant to LPS) and C3H/HeN (normal response to LPS) mice, 6-8 weeks old, were acquired from Beijing Vital River Laboratory Animal Technology (Beijing, China) and Shanghai Experimental Animal Center (Shanghai, China), respectively. All animals were housed in a temperature-controlled environment at the Zhejiang University Laboratory Animal Center. Protocols for animal research were preapproved by the Institutional Animal Care and Use Committee of Zhejiang University School of Medicine. Mice within each cage were simultaneously randomized into treatment groups in random order. N-protein was pretreated with polymyxin B (250 µg/ml) at 37°C for 1 h for endotoxin neutralization, while bovine serum albumin (BSA) was preincubated at 37°C for 1 h as control. Mice were instilled intratracheally with N-protein (3 μg, 15 μg, and 75 μg per mouse in 50 µL), BSA (75 μg in 50 µL), polymyxin B (50 µl of 250 µg/ml), or phosphate-buffered saline (PBS, 50 µL). To explore the role of NF-ĸB in N-protein-induced lung injury, mice were treated with intraperitoneal injection of NF-ĸB inhibitor pyrrolidine dithiocarbamate (PDTC, 50 mg/kg, MilliporeSigma) 1 h before N-protein treatment. Based on our experience, the optimal duration for N-protein to induce significant acute lung injury was 24 h. Animals were sacrificed at 24 h after N-protein, BSA, polymyxin B, PDTC, or PBS treatment. Lungs and bronchioalveolar lavage (BAL) samples were collected for further analysis.

Intratracheal administration of solution into mice was modified from Helms et al. (17). Mice were anesthetized with phenobarbital (50 mg/kg) via intraperitoneal injection. The fur covering the tracheal region was shaved with an electric razor. A layer of hair removal cream was applied to the shaved area and cleaned with a damp towel after 3 minutes of application. The surgical area was wiped with alcohol pads. The anesthetized mice were positioned onto an angled restraining stand. A small skin incision was made in the tracheal region. The anterior tracheal muscles were dissected with blunt incision to visualize and access the tracheal rings. Fifty µl of solution was administered into trachea using a 3/10 mL insulin syringe with 30 gauge x ½ inch needle. The solution was injected slowly with the needle bevel up and parallel with the trachea. To confirm the success of instillation, air was introduced through the syringe and expansion of lung was observed. The needle was removed from the trachea. Mice were then allowed to recover in room air for 30 min and returned to the cages with free access to food and water.

To obtain BAL cells from mice, the tracheas were canulated and the lungs were lavaged three times with 0.5 ml of cold PBS. Total white blood cell counts in the BAL were directly examined using a hemocytometer. BAL was then labeled with FITC-conjugated anti-mouse Ly-6G (Gr-1) antibody (Thermo Fisher Scientific, Waltham, MA) and analyzed for the percentage of neutrophils via flow cytometry. Neutrophil counts in the BAL were calculated by multiplying the percentage of neutrophils by the total number of white blood cells. BAL samples were then centrifuged at 400 x g for 5 min at 4°C to collect supernatant. Protein concentration in BAL supernatant, which is an index of capillary leakage during lung inflammation, was determined using a BCA Protein Assay Kit (Thermo Fisher Scientific). Cytokine levels of interleukin (IL)-1β, IL-6, tumor necrosis factor (TNF)-α, and interferon (IFN)-γ in the BAL were examined by ELISA (R&D Systems, Minneapolis, MN) per manufacturer’s protocol.

BAL collection and lung tissue pathology were performed on separate mice. To harvest the lungs for histological analysis, the tracheas were cannulated with a 18G needle and the lungs inflated with 0.6 ml of chilled 4% paraformaldehyde. After fixation for 24 h, lung tissue was embedded in paraffin wax and sectioned at 5 μm in thickness. Lung sections were then stained with hematoxylin and eosin (H&E) to assess cellular and morphological structures. Images of lung sections were taken with an Olympus VS120 microscope (Shinjuku, Tokyo, Japan).

To determine macrophage recruitment and phenotype in BAL, cells were labelled with the following surface markers: PE anti-mouse F4/80 (BioLegend, San Diego, CA), BV650 anti-mouse CD86 (BD, Franklin Lakes, NJ), PE-cy7 anti-mouse iNOS (Thermo Fisher Scientific), APC anti-mouse CD206 (Thermo Fisher Scientific) for 20 min at 4°C. Cells were then washed twice with cold flow cytometry buffer (PBS, 0.5% BSA) and incubated for 30 min at 4°C. Cells were subsequently washed and resuspended with cold flow cytometry buffer for analysis. Isotype-matched control antibodies were used to distinguish the cut-off between negative and positive fluorescent populations. Data were collected by a BD LSRFortessa™ flow cytometer and analyzed using FlowJo V10 software.

For BMDMs, C57BL/6 mice (6-8 weeks old) were sacrificed via cervical dislocation to harvest femurs and tibias. Bone marrow was flushed out of the ends of the bones by inserting a 25-gauge needle attached to a 10 ml syringe filled with PBS, until the color of the bone turned from red to white. Bone marrow was further dispersed by passing through an 19-gauge needle attached to a 10 ml syringe filled with PBS. The cell suspension was collected and centrifugated at 250 g for 5 min at room temperature. Red blood cells in the pellet were lysed by incubating with 1x lysis buffer (10 mM KHCO3, 155 mM NH4Cl, 0.1 mM EDTA) for 3 min. Cells were pelleted again, washed with PBS, and cultured in medium containing DMEM supplemented with GM-CSF (20 ng/ml, PeproTech, Rock Hill, NJ), 10% fetal bovine serum (Biological Industries, Cromwell, CT), and 1% penicillin plus streptomycin at 37°C with 5% CO2 overnight. Non-adherent cells were harvested and seeded on 6-well culture plates (2 x 10⁶ cells/well). Fresh culture medium was fed every 3 days. Adherent BMDMs were harvested with lidocaine/EDTA (0.4% lidocaine, 5 µM EDTA) after 7 days and quantitated for further experiments. BMDMs (5 x 10⁵ cells) were cultured in a standard 12-well plate and stimulated with N-protein (5 µg/ml), LPS (10 ng/ml), or PBS for different time periods. Then, BMDMs were harvested for Western blot, qRT-PCR, and phenotype analyses. To determine phenotypes of BMDMs, cells were labelled with the following surface markers: PE anti-mouse F4/80 and BV650 anti-mouse CD86 for 20 min at 4°C. BMDMs were then washed twice with cold flow cytometry buffer (PBS, 0.5% BSA) and incubated for 30 min at 4°C. Then, cells were washed and resuspended with cold flow cytometry buffer for analysis. Data were collected by a BD LSRFortessa™ flow cytometer and analyzed using FlowJo V10 software.

Whole-cell lysates of BMDMs were prepared in lysis buffer (10 mM Tris-HCl, pH 7.4, 1 mM EDTA, 150 mM NaCl, 0.5% Nonidet P-40, 1 mM Na₃VO₄, and 1 mM PMSF). The total protein concentration was determined using a BCA protein assay kit, exhibiting strong absorbance at 562 nm. Equal amount of protein extracts (30 μg) or N-protein (0.5 μg) were run on 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), transferred to polyvinylidenefluoride membranes (Millipore, Billerica, MA), blocked with 5% non-fat milk in tris-buffered saline with Tween-20 (TBST), and probed with the primary antibodies to N-protein (Genscript), phosphor-NF-ĸB p65 (#3033, Cell Signaling Technology, Danvers, MA), and NF-ĸB p65 (#8242, Cell Signaling Technology). After washing with TBST, membranes were incubated with specific secondary antibodies conjugated to horseradish peroxidase. The binding was detected via enzyme-linked chemiluminescence using the EZ-ECL kit (Biological Industries, Kibbutz Beit-Haemek, Israel).

Total RNA was isolated from the cells with the use of Trizol reagent (Thermo Fisher Scientific). The RNA concentration was determined by a Nanodrop Spectrophotometer (ND-2000; Thermo Fisher Scientific). RNA (1 µg) was reversely transcribed to cDNA using the Superscript III First-strand kit (Thermo Fisher Scientific). The resulting cDNA was amplified by Power SYBR Green RT-PCR reagent kit (Takara Bio, Kusatsu, Japan). mRNA values were normalized for differences in sample concentration to the levels of β-actin mRNA and expressed as relative fold change compared with the reference of control samples. PCR primer sequences were listed as follows: IL-1β forward 5’GAAATGCCACCTTTTGACAGTG3’, reverse 5’TGGATGCTCTCATCAGGACAG3’; IL-6 forward 5′CTGCAAGAGACTTCCATCCAG3′, reverse 5′AGTGGTATAGACAGGTCTGTTGG3′; TNF-α forward 5’CAGGCGGTGCCTATGTCTC3’, reverse 5’CGATCACCCCGAAGTTCAGTAG3’; inducible nitric oxide synthase (iNOS) forward 5’CAGGCTGGAAGCTGTAACAAAG3’, reverse 5’GAAGTCATGTTTGCCGTCACTC3’; YM-1 forward 5’CAGGTCTGGCAATTCTTCTGAA3’, reverse 5’GTCTTGCTCATGTGTGTAAGTGA3’; CD206 forward, 5’CTCTGTTCAGCTATTGGACGC3’, reverse 5’TGGCACTCCCAAACATAATTTGA3’; β-actin forward 5’CGTTGACATCCGTAAAGACC3’, reverse 5’AACAGTCCGCCTAGAAGCAC3’.

Data are presented as mean ± standard deviation of the mean (SD). Student’s t test, one-way analysis of variance (ANOVA) with Tukey’s post hoc analysis, or two-way AVOVA with Tukey’s post hoc analysis ( Figure 4 only) was performed for parametric data with n ≥ 8. The Mann-Whitney test or Kruskal-Wallis test with Dunn’s post hoc analysis was performed for small sample size (n < 8) and non-parametric data with n ≥ 8. Data were tested for normality by using the D’Agostino-Pearson Omnibus normality test, the Shapiro-Wilk normality test, and Kolmogorov-Smirnov test with Dallal-Wilkinson-Lilliefor corrected P value. Statistical analysis was carried out using the GraphPad Prism 7. Results were considered significant if p < 0.05.

To study the effects of N-protein on lung injury, recombinant N-protein was prepared as a His-tagged fusion protein ( Figure 1A ). The molecular size of the protein was approximately 50 kDa. Coomassie blue staining showed greater than 95% purity of affinity-purified N-protein ( Figure 1A ). The endotoxin level was <0.1 EU/μg protein (<0.01 ng/μg) after processed with an endotoxin removal kit. The identity of the purified protein was confirmed by Western blot analysis with an anti-N-protein antibody ( Figure 1B ). To address the possibility that protein-bound LPS might escape detection, N-protein was pretreated with polymyxin B (250 µg/ml) at 37 °C for 1 h for all in vivo and in vitro experiments. Incubation of N-protein with polymyxin B did not result in protein aggregation ( Supplemental Figure 1 ). To determine the effects of N-protein on pulmonary inflammation, C57BL/6 mice were treated intratracheally with N-protein (75 µg/mouse) or PBS. Treatment of N-protein resulted in an acute inflammatory response, with thickened alveolar septa and increased cellularity compared to control at 24 h ( Figure 1C ). There was a significant increase in the leukocyte infiltration score in the lung of N-protein group compared with the control. Similar doses of proinflammatory stimuli, such as HMG-1 and SARS S protein, were reported to induce lung inflammation in animal studies (18, 19).

C57BL/6J mice were treated with N-protein at three dose levels intratracheally (3, 15, and 75 µg/mouse in 50 µl). BAL fluids were harvested at 24 h after N-protein exposure and analyzed for pulmonary endothelial permeability, cellular infiltration, and concentrations of proinflammatory cytokines. Pulmonary endothelial permeability, as determined by total protein levels in the BAL, was significantly elevated in the medium dose (15 µg/mouse) and high dose (75 µg/mouse) groups compared with PBS control ( Figure 2A ). High dose of N-protein also increased the total number of cells in the BAL, while neutrophils were raised in all 3 dose groups compared with control. In addition, the concentrations of proinflammatory cytokines IL-1β, IL-6, TNF-α, and IFN-γ were significantly higher in the high dose group than the control ( Figure 2B ). Medium dose of N-protein only raised the TNF-α level, while low dose had no effect on all four cytokines. In contrast, intratracheal administration of bovine serum albumin (75 µg/mouse) or polymyxin B (50 µl of 250 µg/ml) did not result in any significant change in the above parameters for lung injury.

To study the effects of N-protein on macrophage polarization, mice were subjected to N-protein (75 µg/mouse) or PBS intratracheally. BAL was harvested at 24 h and analyzed phenotypically via flow cytometry ( Supplemental Figure 2 for gating strategy). Flow cytometry analysis demonstrated that N-protein increased the number of macrophages (F4/80+) in the BAL compared with the control ( Figure 3A ). The results also showed that N-protein significantly elevated the expression of F4/80+CD86+ and F4/80+iNOS+ cells, favoring the orientation toward M1 macrophages ( Figure 3B ). In addition, N-protein decreased the levels of F4/80+CD206+ cells, a marker for M2 macrophages ( Figure 3C ).

To substantiate that N-protein-induced lung injury is not the result of LPS contamination, N-protein-induced lung injury was compared between C3H/HeJ (TLR4 mutation) and C3H/HeN (control) mice. Previous studies have reported that mice with TLR4 mutation are resistant to LPS-induced lung injury (20). Mice were administered with N-protein (75 µg/mouse) or PBS, and then examined for lung injury at 24 h. TLR-4 mutation did not alter the effects of N-protein on total protein concentration, total cell count, and neutrophil infiltration in the BAL compared with control mice, indicating N-protein is responsible for the lung injury ( Figure 4 ).

To examine the signal transduction pathways involved, BMDMs were exposed to polymyxin B-treated N-protein (5 µg/ml) and activation of NF-ĸB was examined by Western blot, using a phospho-specific p65 antibody. N-protein triggered phosphorylation of NF-ĸB p65, beginning at 10 min after the exposure, reaching a peak between 30 min to 4 h, and then returning to the baseline at 24 h ( Figures 5A, G ). In addition, the NF-ĸB activation was lost after N-protein denaturation via heat inactivation (80°C, 1 h) ( Figure 5B ). Furthermore, incubation with an anti-N-protein antibody blocked the effect of N-protein on NF-ĸB p65 phosphorylation ( Figure 5C ). As a control, LPS (10 ng/ml) had a relatively short duration for NF-ĸB activation ( Figures 5D, G ), with a peak between 30 min to 1 h. Additionally, LPS was not inactivated at the temperature and duration applied to denature N-protein as the p65 phosphorylation pattern was unaltered ( Figure 5E ). Moreover, LPS induced NF-ĸB signaling was efficiently inhibited by polymyxin B treatment ( Figure 5F ).

Prior studies have demonstrated that activation of NF-κB promotes M1 macrophage polarization, resulting in proinflammatory responses (21). To study whether N-protein modulates macrophage phenotypes in vitro, BMDMs were cultured with polymyxin B-treated-N-protein (5 µg/ml) or PBS for 24 h. Flow cytometry results showed that N-protein enhanced the level of M1 macrophages (F4/80+CD86+) compared with the control ( Figure 6A ). N-protein also elevated the mRNA expression of M1 macrophage marker iNOS while simultaneously decreased the levels of M2 markers CD206 and YM-1 ( Figure 6B ). The effect of N-protein on macrophage polarization was abolished by N-protein denaturation via heating. Meanwhile, pretreatment with N-protein antibody displayed similar effects with the exception of elevating YM-1 expression ( Figures 6A, B ). In addition, mRNA levels of proinflammatory IL-1β, IL-6, TNF-α were raised by N-protein treatment ( Figure 6C ). Conversely, N-protein denaturation and N-protein antibody blocked the impacts of N-protein on proinflammatory cytokines.

To assess the effect of N-protein on NF-ĸB activation in the lungs, C57BL/6 mice were treated with N-protein intratracheally (75 µg/mouse in 50 µl) for 0 (immediate before treatment), 2, 4, or 12 h. Levels of phospho-specific NF-ĸB p65 were measured by Western blot of lysates from BAL cells ( Figure 7A ). NF-ĸB p65 phosphorylation was undetectable at time 0, dramatically elevated at 2 h after the exposure, and gradually decreased at 4 and 12 h. To establish that NF-κB mediates N-protein-induced lung injury, mice were treated with PBS, N-protein (75 µg/mouse), or N-protein + NF-κB inhibitor PDTC (50 mg/kg). PTDC alleviated the effects of N-protein on total protein concentration, total cell count, and neutrophil infiltration in the BAL at 24 h, indicating that NF-κB activation is at least partially responsible for the lung injury ( Figure 7B ).