A prolonged innate systemic immune response in COVID-19 – Scientific Reports

Patient characteristics

Twenty-seven hospitalized patients with PCR-verified SARS-CoV-2 infection participated in the study. The median age of the covid patients was 66 years (range 48-90 years). There were 17 male (63%) and 10 female (37%) participants. The control group consisted of 6 male (50%) and 6 female (50%) healthy participants verified by PCR negative test for SARS-CoV-2. The median age of the covid patients was 33 years (range 25-58 years). Demographic and clinical characteristics of enrolled patients and healthy controls are summarized in Table 1†

COVID-19 is characterized by neutrophil hyperreactivity and high production of pro-inflammatory cytokines

Neutrophils from patients with COVID-19, compared to healthy subjects, were characterized by significantly higher expression intensity of neutrophil activation marker CD11b (mean fluorescence intensity (MFI) ± SD of 5448 ± 2071 and 3475 ± 1035, respectively; p< 0.01; Fig. 1A). The intensity of fluorescence and the percentage of neutrophils positive for granulocyte activation marker CD66b were also significantly increased in COVID-19 patients compared to controls (1257 ± 408 and 605 ± 115.7, respectively; p< 0.0001; Fig. 1B and median and interquartile range of 6.220 (4.6-10.3) respectively compared to 2.380 (1.83-3.98); p< 0.0001; Fig. 1C). Furthermore, we found that plasma levels of the pro-inflammatory cytokines, TNF-α, IL-6 and IL-8, – central cytokines in antiviral responses and cytokines that trigger a cytokine storm phenomenon – were significantly higher in patients with COVID-19 compared with the healthy donors (TNF, median and interquartile range of 1.545 (1.165-2.076) compared to 0.6383 (0.513-0.730); p< 0.0001; Fig. 2A; IL-6, median and interquartile range of 8.749 (3.79-21.79) compared to 0.4955 (0.01-0.819); p< 0.0001); Fig. 2b; IL-8, median and interquartile range of 20.19 (17.47-24.07) compared to 15.07 (13.62-16.30); p< 0.0001); Fig. 2C).

Neutrophils in COVID-19 patients show increased survival compared to neutrophils from healthy donors

Neutrophil populations detected in blood from COVID-19 patients showed significantly higher expression of CD47, compared to healthy donors, which is a molecular “don’t eat me” signal that inhibits phagocytosis of the expressing cell and prolongs survival (MFI ± SD of 2204 ± 631 and 1323 ± 231, respectively; p< 0.0001; Fig. 3A). On average, 32% (±13%) of neutrophils from COVID-19 patients showed CD47 compared to 25% (±15%) in healthy donors (p< 0.05; Fig. 3b). The expression of CD36, which is an “eat me” signal, was not significantly different between groups (Fig. 3CD).

Low expression of CD49 on neutrophils in COVID-19 patients indicates increased migration to lung tissue

CD49^† neutrophils accumulate in the lungs during viral infections¹²† Expression of CD49 is also increased on the surface of senescent neutrophils. COVID-19 patients had about 50% CD49^† neutrophils in the blood in contrast to healthy controls, where more than 80% of the neutrophils in the blood expressed CD49 (49.3 ± 21.4 vs. 81.9 ± 14.6; p< 0.0001; Fig. 4A). MFI of CD49 on neutrophils was also lower compared to healthy controls. However, this finding did not reach statistical significance (p= 0.051, fig. 4b). Low expression of CD49 on neutrophils in the blood indicates that neutrophils highly migrate to peripheral tissues, including lungs, during SARS-CoV-2 infection, and that they are of young age.

COVID-19 patients have high levels of NET markers in their plasma

Consistent with previous research, plasma levels of circulating cell-free DNA (cf-DNA) and cell-free nucleosomes were significantly higher in patients with COVID-19 compared to healthy subjects (median (IQR) of 98.2 (88.7-120.3) ) vs. 78.9 (75.6-83.3); p< 0.0001 and 0.47 (0.29-0.95) vs. 0.16 (0.12-0.26); p< 0.0001; Fig. 5A–B). Expression of those markers indicates formation of NETs in the blood of affected patients.

High-dimensional cluster analysis reveals neutrophil clusters characterized by prolonged hyperreactivity

To extend the findings, we processed data as described in the Methods section and mapped the neutrophil populations on t‐SNE composite plots, which revealed clear localizations of neutrophil populations in COVID-19 patients and healthy controls. Phenograph analysis revealed 17 unique clusters in t‐SNE space. The detailed characterization of clusters is attached as Supplementary Table 1† figure 6 shows the distribution and localization of neutrophils in all study participants (Fig. 6A), COVID-19 patients only (Fig. 6B) and healthy controls only (Fig. 6C).

As can be seen in fig. 6, the neutrophil populations in COVID-19 patients are characterized by completely opposite phenotypes than the populations seen in healthy controls. Neutrophils from healthy controls were grouped into clusters 2, 4, 6, 8, 9, 13, 15, and 17 (Fig. 6b). COVID-19-associated neutrophils were mainly located in clusters 1, 3, 5, 7, 10-12, 14 and 16 (Fig. 6C). Clusters of SARS-CoV-2 infected patients include neutrophils characterized by high expression of activation markers (CD66b, CD11b, CD62L) and “don’t eat me” markers (CD47).

Neutrophil populations from COVID-19 patients were plotted on the same t-SNE patterns with respect to gender (Fig. 7AC). The analysis found that compared to female patients, male patients had significantly more neutrophils in cluster 3 (p< 0.01) including neutrophils with a high expression of "don't eat me" and activation markers (CD47, CD11b). On the other hand, females had significantly more neutrophils plotted within cluster 1 that includes adult non-activated neutrophils (CD16^HighCD62L^High†

In addition, patients with peripheral blood markers of hyperinflammation (high levels of C-reactive protein CRP and high ferritin levels) were characterized by over-representation of neutrophils within clusters 5 (representing premature unactivated neutrophils) and 1 (representing mature unactivated neutrophils). ) activated neutrophils). On the other hand, those patients with normal level of CRP had significantly more neutrophils that can be classified as aged and pre-apoptotic (Fig. 8A).

Another analysis, comparing patients with severe disease, defined as hospitalization greater than 7 days, and milder disease, hospitalization less than 7 days, revealed that cluster 3, including hyperactivated neutrophils containing the “don’t eat me expressing markers are overrepresented in patients with severe disease (Fig. 8B) as well as in elderly patients (over 70 years of age) (Fig. 8C).