Aberrant neutrophil degranulation in hospitalized patients with COVID‐19 partially remains for 6 months

Neutrophils are important players in COVID‐19, contributing to tissue damage by release of inflammatory mediators, including ROS and neutrophil elastase. Longitudinal studies on the effects of COVID‐19 on neutrophil phenotype and function are scarce. Here, we longitudinally investigated the phenotype and degranulation of neutrophils in COVID‐19 patients (28 nonhospitalized and 35 hospitalized patients) compared with 17 healthy donors (HDs). We assessed phenotype, degranulation, CXCL8 (IL‐8) release, and ROS generation within 8 days, at one or 6 month(s) after COVID‐19 diagnosis. For degranulation and ROS production, we stimulated neutrophils, either with ssRNA and TNF or granulocyte‐macrophage colony‐stimulating factor and N‐Formylmethionyl‐leucyl‐phenylalanine. During active COVID‐19, neutrophils from hospitalized patients were more immature than from HDs and were impaired in degranulation and ROS generation, while neutrophils from nonhospitalized patients only demonstrated reduced CD66b+ granule release and ROS production. Baseline CD63 expression, indicative of primary granule release, and CXCL8 production by neutrophils from hospitalized patients were elevated for up to 6 months. These findings show that patients hospitalized due to COVID‐19, but not nonhospitalized patients, demonstrated an aberrant neutrophil phenotype, degranulation, CXCL8 release, and ROS generation that partially persists up to 6 months after infection.


Introduction
The COVID-19 pandemic arose at the beginning of 2020 and the virulence of SARS-CoV-2 still remains enigmatic.In one out of five unvaccinated, naive individuals the infection can result in severe disease, while others are barely affected by the virus or experience mild symptoms only [1].COVID-19-related critical illness is associated with a cytokine storm, lung injury, viral sepsis, acute respiratory distress syndrome, shock, organ failure, and potential fatality [2,3].An elevated neutrophil-to-lymphocyte ratio has been identified as a biomarker for early risk stratification for severe COVID-19 [4,5].Neutrophilic infiltration of the lungs is increased in patients with COVID-19 pneumonia compared with patients with non-COVID-related pneumonia [6].Moreover, neutrophils can drive the onset of the cytokine storm which contributes to severe COVID-19 [6][7][8].
Neutrophils are the most abundant peripheral blood leukocytes, acting as first responders to infection [9].They employ several mechanisms to kill microbes, including phagocytosis, degranulation, generation of ROS, and the formation of neutrophil extracellular traps (NETs), a process called NETosis [10].Additionally, neutrophils secrete an array of cytokines and chemokines, amongst which CXCL8 (IL-8), by which more neutrophils are recruited [9].Neutrophils are a prominent source of granular protease as well and combined with NETosis and ROS generation, uncontrolled neutrophil activation can lead to collateral tissue damage [9,10].In mice, it was shown, for example, that neutrophil elastase (NE), a protease of primary granules, causes alveolar damage in the lungs, which decreases host tolerance to a lung infection [11].NE also decorates NETs and in deceased COVID-19 patients high levels of NETs were found in the bronchi and alveolar spaces of the lower respiratory tract, which could contribute to acute respiratory distress syndrome [12].NE was also shown to be an independent predictor of lung and multi-organ damage in COVID-19 patients [13].Although neutrophils are vital for infection clearance [10,14], their contribution to tissue damage and decreased host tolerance to infection appoints them as major players in severe COVID- 19.
Both an increased frequency of immature neutrophils and a hyperactivated status of neutrophils have been described in COVID-19 patients [6,8,[15][16][17][18][19]. Here, we investigated neutrophil phenotype, degranulation, CXCL8 release, and ROS generation in COVID-19 patients experiencing either mild or moderate to severe disease, distinguished in this study by hospitalization.We longitudinally studied the neutrophil compartment in patients for up to 6 months as it has been postulated that COVID-19 leaves an immunological imprint on leukocytes [20,21].Therefore, we assessed neutrophil phenotype, degranulation, ROS generation, and CXCL8 production when COVID-19 was diagnosed, as well as 28 and 180 days after infection.We show that neutrophils from hospitalized COVID-19 patients produce elevated CXCL8 and are impaired in both degranulation and ROS generation.At 6 months (day 180) after infection, CXCL8 production and activation status (CD63 expression) of neutrophils from hospitalized COVID-19 patients were still increased.Hence, these data support the hypothesis that COVID-19 imprints the innate immune system, which could underlie long-term symptoms, that is, severe fatigue [22], after SARS-CoV-2 infection.

Characteristics of COVID-19 patients
Patient characteristics and COVID-19 severity and symptoms of both nonhospitalized and hospitalized patients are shown in Table 1.Three patients admitted to the ICU are among the hospitalized patient group.Nonhospitalized patients were significantly younger than hospitalized patients (p = 0.0045), whereas no differences in sex, BMI, and smoking were found between groups.

Neutrophils from hospitalized COVID-19 patients show an immature phenotype which is reversed within 1 month after diagnosis
As it was previously shown that neutrophils from hospitalized COVID-19 patients show a more immature phenotype, known as a left-shift of the neutrophil compartment, with decreased expression of CD10 (metalloprotease) and CD16 (FcγRIIIb) maturation markers [8,17,23,24], we first determined the neutrophil phenotype at day 0 (within 8 days after COVID-19 diagnosis) of nonhospitalized and hospitalized patients in comparison to healthy donors (HDs).Expression of CD10, CD11b, CD16, CD32, CD88, and CD177 was analyzed on CD15 + CD16 + neutrophils (gating strategy shown in Supporting information Fig. S1A and B).Neutrophil CD10 expression at day 0 in hospitalized patients was significantly lower compared with nonhospitalized and HDs, as shown in representative histograms (Fig. 1A) and as combined data from all donors (Fig. 1B).CD16 expression on neutrophils from hospitalized patients also tended to be lower; however, this was not statistically significant (Fig. 1A, B).Furthermore, an increased population of CD10 -CD16 low neutrophils was observed in hospitalized patients (Fig. 1B), which is indicative of active release from bone marrow [25].Neutrophil expression of CD16 correlated with that of CD10 (R 2 = 0.47) and BMI (R 2 = 0.11) in hospitalized patients (Supporting information Fig. S2A, B).No other correlations between neutrophil phenotype with age or BMI of patients were found (Supporting information Fig. S2B).
Reduced surface expression of CD10 and CD16 on neutrophils is a predictor of a poor outcome in sepsis patients [26], similar to a reduced expression of CD88, the complement receptor C5aR [27].Since sepsis is a potential complication of COVID-19 [2], we analyzed CD88 expression on neutrophils in our cohort and we observed a significant reduction of CD88 expression at day 0 in hospitalized patients but not nonhospitalized patients compared with HDs (Fig. 1A,B).Furthermore, we analyzed the expression of integrin CD11b, CD32 (FcγRII), and CD177 (NB1 antigen; Supporting information Fig. S1A and B).CD11b and CD32 are indicative of in vivo primed neutrophils, that is, by TNF [28,

29]
. Interestingly, neutrophils of nonhospitalized patients were partially primed, with 1.5-fold higher median CD11b expression and 0.8-fold lower median CD32 expression compared with HDs (Supporting information Fig. S2C).CD11b and CD32 both significantly and positively correlated with CD10 expression (Supporting information Fig. S2D).CD177 is also a specific marker of neutrophil activation, indicative of secondary granule release [19,30].This glycosylphosphatidylinositol-anchored protein is also severely dysregulated in sepsis and has been postulated as a prognostic marker for COVID-19 disease severity [15].Since a clear bimodal expression of CD177 was observed, we analyzed the percentage of CD177 + cells.The median frequency of CD177 + cells was higher in hospitalized patients, although the effect was not statistically significant (Supporting information Fig. S2E).
Next, we performed longitudinal analyses of the neutrophil phenotype and we demonstrated that expression of CD10 and CD16 was significantly elevated at day 28 compared with day 0 and CD10 was restored to HD expression level in hospitalized patients (ns to HD; Fig. 1C).Similarly, the proportion of CD10 -CD16 low neutrophils in hospitalized patients normalized at day 28 to levels seen in HD (Fig. 1C).CD88 expression was increased at day 28 (p = 0.063) and at day 180 (p = 0.052) compared with day 0 and was not significantly different from HDs anymore (Fig. 1C).The expression of CD11b and CD32 on neutrophils from nonhospitalized patients was also restored to HD level at day 28 (ns to HD) (Supporting information Fig. S2F).Taken together, patients who were hospitalized because of COVID-19 presented with an immature neutrophil phenotype and reduced neutrophil CD88 expression at the onset of the disease, which was restored to HD level within 1 month.Nonhospitalized patients showed a partially primed neutrophil phenotype with an increased CD11b and slightly reduced CD32 expression compared with HDs.

Decreased neutrophil degranulation in hospitalized patients
The observed differences in neutrophil phenotype between nonhospitalized and hospitalized patients raised the question of whether this was associated with differences in neutrophil functionality.Neutrophils release the contents of preformed granules upon degranulation, which contain an array of proteases and antimicrobial proteins [10].We previously showed that efficient neutrophil activation requires a double stimulus and that the response to different stimuli varies [31].To mimic SARS-CoV-2 RNA sensing by neutrophils, we stimulated neutrophils with ssRNA, a TLR7/8 ligand, as neutrophils express TLR8 [32], in combination with the inflammatory cytokine TNF.Alternatively, we used a combination of GM-CSF and fMLP to stimulate neutrophils, as different stimuli may induce variable neutrophil responses.Upon activation CD16 is cleaved from the neutrophil membrane [30,33], while CD63 and CD66b, markers of primary granule and secondary granule release, respectively, are upregulated [30,31,34].The gating strategy of neutrophils is shown in Supporting information Fig. S3A and B. Neutrophils of hospitalized patients expressed higher levels of CD16, indicating reduced CD16 cleavage, compared with neutrophils of HDs (Fig. 2A,B), irrespective of the combination of stimuli used (ssRNA and TNF or GM-CSF and fMLP), while nonhospitalized patients demonstrated similar CD16 cleavage as HDs.Aberrant CD16 expression in hospitalized patients was restored to HD level at day 28 and still at day 180 (ns to HDs; Fig. 2C).
Hospitalized patients also demonstrated less CD63 and CD66b membrane expression upon stimulation of neutrophils compared with HDs.CD63 expression of hospitalized patient neutrophils was only significantly different from HDs when neutrophils were stimulated with ssRNA and TNF (Fig. 2D), while CD66b expression was less increased upon stimulation with either combination of stimuli in both nonhospitalized and hospitalized patients (Fig. 2F).The basal level of CD63 and CD66b expression of unstimulated neutrophils was higher in hospitalized patients than in HDs, while only the MFI of CD63 was significantly increased compared with HDs (Fig. 2H, Supporting information Fig. S4A).Baseline level of CD63 + granule release from hospitalized patient neutrophils was still significantly increased at day 180 compared with HD level (Fig. 2I).Additionally, fold induction of CD63 expression upon ssRNA and TNF stimulation was still significantly decreased at day 180 compared with HD (Fig. 2E) and median absolute MFI values of stimulated neutrophils were lower than median of HD neutrophils at day 180, although not significantly (Fig. 2I).Interestingly, CD66b expression was lowest in nonhospitalized patients upon stimulation with either stimuli combinations (Supporting information Fig. S4B).Fold induction of CD66b expression at day 28 and day 180 was significantly different from day 0 and similar to HD level for both hospitalized and nonhospitalized patients (Fig. 2G, Supporting information Fig. S4C).In a subset of patients' elastase release upon stimulation was also measured (Supporting information Fig. S5).Similar to neutrophil CD63 expression, baseline elastase release was increased in COVID-19 patients as compared with HDs (Supporting information Fig. S5A).Upon stimulation with ssRNA and TNF, elastase release was decreased in hospitalized patients (Supporting information Fig. S5A), however, in contrast to CD63 expression (Fig. 2E), restored ssRNA and TNF-induced elastase produc-tion was observed at day 28 and day 180 (Supporting information Fig. S5B).Our data indicate that neutrophil primary and secondary degranulation was impaired in hospitalized patients, whereas only secondary granule release was reduced in nonhospitalized patients.Secondary granule release and CD16 shedding were restored at 1 month postdiagnosis, whereas primary granule release was at least in part still affected up to 6 months after infection.

Persistently elevated CXCL8 release by neutrophils from hospitalized patients
CXCL8 is the main chemoattractant for neutrophils, while it is also produced by neutrophils, thereby contributing to the inflammatory neutrophil loop and consequent tissue damage.Therefore, we assessed the production of CXCL8 by neutrophils after one day of culture.Resting (unstimulated) neutrophils of COVID-19 patients showed elevated spontaneous CXCL8 release compared with HD neutrophils, with median (IQR) production of 35.0 (17.5-50.0)pg/mL by hospitalized patient neutrophils and 11.0 (10.0-14.5)pg/mL by HD neutrophils (Fig. 3A).When neutrophils were stimulated with ssRNA and TNF, CXCL8 release was significantly higher by neutrophils from hospitalized patients at 2.97 (1.35-4.56)ng/mL than by HD neutrophils at 0.44 (0.37-0.57) ng/mL, whereas CXCL8 release by neutrophils from nonhospitalized patients did not differ from HDs (Fig. 3A).A slight increase of CXCL8 production by neutrophils from hospitalized patients was observed with GM-CSF and fMLP stimulation, although the effect was not statistically significant.Hospitalized patient neutrophils still significantly produced more CXCL8 at day 180 after diagnosis compared with HDs either resting or upon ssRNA and TNF stimulation, and a trend toward HD level was observed in stimulated neutrophils with significantly diminished CXCL8 secretion at day 180 in comparison to day 0 (Fig. 3B).Hence, CXCL8 production by neutrophils from hospitalized COVID-19 patients is elevated at day 0 and remains increased at 6 months after diagnosis.

ROS generation by stimulated neutrophils from COVID-19 patients is impaired
ROS production is an antimicrobial feature of neutrophils in the defense against bacteria and fungi and in addition, ROS destroys components required for viral transmission [10,14].We assessed intracellular ROS production by flow cytometry using the indicator DHR-123 that emits green fluorescence upon oxidation by ROS and the gating strategy is shown in Supporting information Fig. S3C.Baseline intracellular ROS levels were increased in COVID-19 patients compared with HD at day 0, as shown in Fig. 4A and B. Homeostatic ROS production by resting (unstimulated) neutrophils of both nonhospitalized and hospitalized patients was not significantly different from HD neutrophils anymore at day 180 (Fig. 4C).In contrast to the elevated baseline  intracellular ROS level at day 0, ROS generation upon stimulation with either ssRNA and TNF or GM-CSF and fMLP was decreased in both nonhospitalized and hospitalized patients compared with HDs (Fig. 4D, F and Supporting information Fig. S6A, C).At 28 days post-COVID-19 diagnosis the capacity to produce intracellular ROS was restored to HD level when neutrophils were stimulated with the strongest ROS stimulant used in this study, namely ssRNA and TNF (Fig. 4E and Supporting information Fig. S6B).When stimulated with GM-CSF and fMLP, median ROS production by neutrophils of both nonhospitalized and hospitalized patients was still lower than HDs at 180 days post-COVID-19, albeit not significantly (Fig. 4G and Supporting information Fig. S6D).Overall, neutrophils of COVID-19 patients showed impaired ROS production in response to stimuli, which was restored at day 28 in response to ssRNA and TNF, whereas the baseline level of intracellular ROS was increased at the onset of disease.

Discussion
In this study, we show that peripheral neutrophils of hospitalized COVID-19 patients demonstrate a more immature phenotype.Furthermore, neutrophil degranulation and intracellular ROS production upon stimulation were decreased in hospitalized patients and patients with mild disease (nonhospitalized), whereas CXCL8 release was increased upon stimulation only in hospitalized patients.Neutrophil maturation, CD16 and CD66b expression, and intracellular ROS production were restored within a month after diagnosis.In contrast, elevated baseline CD63 expression, spontaneous CXCL8 release, and impaired CD63 + (primary) granule release were observed up to 180 days after diagnosis.
Neutrophils from hospitalized COVID-19 patients with moderate disease severity demonstrate a more immature phenotype characterized by lower CD10 and CD16 expression and increased CD10 -CD16 low population.The abundance of immature neutrophils in peripheral blood of hospitalized COVID-19 patients with reduced CD10 and CD16 expression on neutrophils has also been observed by other research groups [8,17,24,35].Neutrophil release from bone marrow underlies neutrophilia, as observed in COVID-19 [5], and CD10 -CD16 low neutrophils are indicative of active release [25].This left shift is likely the result of elevated neutrophil recruitment, for example by neutrophil attractant CXCL8, which is elevated in plasma of hospitalized COVID-19 patients [8,17].Indeed, we observed an increase in spontaneous CXCL8 release from neutrophils of nonhospitalized and hospitalized patients compared with HDs.Moreover, neutrophils from hospitalized patients produced significantly more CXCL8 than neutrophils from HDs upon stimulation up to 6 months postdiagnosis, whereas neutrophil maturation was restored at 28 days postdiagnosis.The transient nature of the immature neutrophil signature was also observed by others [8,19,24,35].
Interestingly, we found a significant reduction in neutrophil CD88 (C5aR) expression, a proposed predictor of sepsis severity, in hospitalized patients, which was also reported by Metzemaekers et al. [17] in severe patients.In addition, the frequency of CD177 + (NB1 antigen) neutrophils was highest in hospitalized COVID-19 patients, albeit not significantly different from HDs, as described before [19,23,35].Our observation that CD11b expression was increased in nonhospitalized patients at diagnosis was in agreement with Jukema et al. [36], who demonstrate, similar to our data, that CD11b expression restored to normal levels at 3-6 months after active COVID-19 disease.In contrast, we observed reduced CD32 expression on neutrophils from nonhospitalized patients.Both CD11b and CD32 are markers indicative of primed neutrophils and it was shown that human neutrophils upregulate both CD11b and CD32 upon stimulation with fMLP [28,29,37].However, we observed that CD10 expression on neutrophils from hospitalized patients was correlated with CD11b and CD32 expression, with CD10 low immature neutrophils demonstrating the lowest CD11b and CD32 expression, indicative of naive (nonprimed) neutrophils.Strikingly, the aberrant expression of CD11b and CD32 is exclusive to nonhospitalized COVID-19 patients.
The expression of functional granule markers CD63 and CD66b, indicative of primary and secondary granule release, were elevated on unstimulated neutrophils of hospitalized patients, as previously observed [8,19].Lourda et al. [35] showed that CD16 dim immature neutrophils express higher levels of CD63 and CD66b than mature neutrophils with the highest expression of both granule markers on immature neutrophils of COVID-19 hospitalized patients.This suggests that the more immature neutrophil phenotype observed in hospitalized patients could underlie the increased expression of CD63 on resting neutrophils, although we did not find a correlation between CD10 and CD63 baseline expression (data not shown).Moreover, we found that baseline intracellular ROS levels were increased in COVID-19 patients confirming previous observations (38).These data suggest a more activated state of neutrophils in COVID-19 patients, based on a higher CD63 baseline expression and on increased spontaneous ROS production and CXCL8 release.In contrast, Spijkerman et al. [24] report no signs of increased activation in the neutrophil compartment directly measured in peripheral blood of COVID-19 patients upon hospital admission versus healthy controls.However, the lack of mild patients and age-matched controls could perhaps explain why no differences were found, as aging affects neutrophil function [39].Their analysis method was applied given that neutrophils get easily activated ex vivo when present for a prolonged period (>30 min) in a blood collection tube [40].In our study, we isolated neutrophils 16-24 h after blood was drawn, meaning that neutrophils could potentially have been activated by this procedure.Evidently, this procedure was also followed for HDs and we did find phenotypical and functional differences between neutrophils from patients and HDs.
Neutrophils from both nonhospitalized and hospitalized COVID-19 patients demonstrated significantly impaired ROS generation and CD66b + secondary granule release upon stimulation.CD66b expression on stimulated neutrophils and CD16 shedding recovered within a month after diagnosis, while CD63 + primary granule release did not.Recovery of stimulated ROS production also occurred within a month postdiagnosis, especially when stimulated with ssRNA and TNF.The impaired ROS generation by patient neutrophils upon stimulation was also observed by others [16].The impeded ROS production at onset of disease is most likely related to neutrophil immaturity, as low-density neutrophils, proposedly immature neutrophils, found in sepsis or systemic inflammatory response syndrome produced less ROS in response to E. coli [41].However, we did not find a correlation of CD10 expression with ROS production (data not shown).
Our study has several limitations.We determined the phenotype of neutrophils and measured their degranulation capacity by CD63 and CD66b expression, elastase release, CXCL8 release, and ROS generation, whereas other neutrophil functions, like extracellular trap and phagocytosis, were not measured.Only three critical patients who were treated in the ICU were included in our cohort and thus the entire disease severity spectrum is not covered in our study.However, for the neutrophil phenotype and degranulation capacity assessed here, the ICU patients were within the range of other hospitalized patients and therefore both groups were combined.We analyzed the phenotype and degranulation capacity of neutrophils obtained from COVID-19 patients using different stimuli; however, we did not combine these stimuli with SARS-CoV-2.Previous reports have shown that different components of SARS-CoV-2, like ssRNA and the spike protein, activate neutrophils isolated from healthy donors and induce the production of CXCL8, ROS, and NETs [42][43][44][45].No data on phenotype or degranulation capacity of neutrophils from individuals prior to SARS-CoV-2 infection was available.Therefore, we cannot investigate whether congenital or pre-existing differences in the neutrophil compartment between nonhospitalized and hospitalized COVID-19 patients underlie a more severe disease course of COVID-19 in hospitalized patients.Also, we did not include control patients diagnosed with other (respiratory) infectious diseases, and therefore the observed effects on neutrophil phenotype, degranulation, CXCL8 release, and ROS generation cannot be attributed specifically to SARS-CoV-2 infection.Previous studies compared COVID-19 with various infectious diseases and showed that bacterial or other viral infections cause a similar decrease in CD10 expression on neutrophils [24].Additionally, intracellular ROS production was reduced in other pneumonia patients as well [16] and high quantity of NETs were also observed in respiratory tract samples of influenza or varicella pneumonia patients [12].Collectively, it appears that aberrant neutrophil activation in COVID-19 is not unique to this disease.
Here, we show using prospective data that neutrophils of hospitalized COVID-19 patients have a more immature and activated phenotype versus HDs, while they are partially refractory to stimulation since we observed reduced CD16 shedding, degranulation, and ROS production upon stimulation.Furthermore, we show that stimulation of neutrophils from hospitalized patients potently induces CXCL8 release, which likely corresponds to the reported higher CXCL8 plasma levels and the left-shift of the neutrophil compartment [6,8,17].Elevated CXCL8 release and baseline CD63 expression as well as impaired CD63 + granule release upon stimulation are seen up to 6 months postdiagnosis of hospitalized patients, which suggests a low-grade ongoing inflammation.Our data suggest that the persistent abnormality of neutrophil function in hospitalized COVID-19 patients may contribute to the development of chronic COVID-19 symptoms like chronic severe fatigue symptoms beyond 6 months after illness onset [22,46,47].

Neutrophil isolation
Prior to Ficoll-Isopaque density centrifugation, heparin collection tubes containing blood were kept at room temperature for 16-24 h due to study logistics.Granulocytes were isolated after density gradient centrifugation and erythrocyte lysis, as previously described [31].Neutrophils were then resuspended in IMDM (Gibco; Thermo Fischer Scientific Inc.) supplemented with 10% heat-inactivated (HI) FBS (HyClone, Thermo Fischer Scientific Inc.) and gentamycin (86 μg/mL, Duchefa Biochemie B.V.) and immediately used in different assays.

CXCL8 and neutrophil elastase ELISA
CXCL8 release by neutrophils was determined in culture supernatant by ELISA according to the manufacturer's instructions (Invitrogen Life Technologies).Neutrophil elastase concentration was analyzed in culture supernatants by ELISA as described previously [31] using the following antibodies for coating and detection: polyclonal rabbit IgG directed against elastase (1.5 ng/mL, Sanquin), biotinylated rabbit anti-human elastase (1 ng/mL, Sanquin), and streptavidin-peroxidase (Amersham Life Science).Absorbance was measured at 450 nm with reference at 655 nm by using a VersaMax microplate reader (Molecular Devices).

Figure 1 .
Figure 1.Neutrophils from hospitalized COVID-19 patients show an immature phenotype and reduced CD88 expression.Neutrophils were isolated from patients at different time points and analyzed for expression of CD10, CD16, and CD88.(A) Representative histogram overlays of CD10, CD16, and CD88 expression on neutrophils isolated from a healthy donor (HD), a nonhospitalized or hospitalized patient at day 0 are shown.(B) Expression of CD10, CD16, and CD88 on neutrophils and the proportion of CD10 -CD16 low neutrophils of HDs (n = 17), nonhospitalized (i = 28), or hospitalized patients (n = 35) is shown as mean fluorescence intensity (MFI) at day 0. Lines indicate the median value for each study group.(C) Marker expression on neutrophils from hospitalized patients is shown at day 0 (n = 35), day 28 (n = 21), or day 180 (n = 23) as compared with HDs.Line indicates the medium value for each study group.Dotted line indicates median value of HDs.Symbols in red indicate ICU patients.Significant differences are indicated as *p < 0.05, **p < 0.01 ***p < 0.001, ****p < 0.0001.

Figure 2 .
Figure 2. Reduced degranulation of neutrophils from hospitalized COVID-19 patients in response to stimuli.Neutrophils were stimulated for 2 h with either ssRNA and TNF or GM-CSF and fMLP or cultured without stimuli.(A) Representative histograms are shown of an HD or hospitalized patient at day 0 for unstimulated or ssRNA + TNF stimulated neutrophils for markers CD16, CD63, and CD66b.(B) CD16 expression of ssRNA + TNF or GM-CSF + fMLP stimulated neutrophils is shown as fold change to unstimulated HDs (n = 17), nonhospitalized (n = 28), or hospitalized patients (n = 35) at day 0. (C) CD16 expression on neutrophils from hospitalized patients (fold change to unstimulated) is shown at day 0 (n = 35), day 28 (n = 23), or day 180 (n = 23) as compared with HDs (D) CD63 expression (fold change), indicating primary granule release, is shown of stimulated neutrophils at day 0 of all three groups.(E) CD63 expression (fold change) is depicted at day 0, 28, and 180 in hospitalized patients as compared with HDs.(F) CD66b expression (fold change) is shown of stimulated neutrophils at day 0 of all three groups.(G) CD66b expression (fold change) is depicted at day 0, 28, and 180 in hospitalized patients as compared with HDs.(H) CD63 and CD66b expression (MFI) on unstimulated neutrophils from HDs (n = 17), nonhospitalized (n = 28), or hospitalized patients (n = 35) at day 0 is depicted.(I) CD63 expression (MFI) on neutrophils at day 0 (n = 35), day 28 (n = 23), or day 180 (n = 23) of hospitalized patients as compared with HDs is shown.Line indicates the medium value for each study group.Dotted line indicates the median value of HDs.Symbols in red indicate ICU patients.Significant differences are indicated as *p < 0.05, **p < 0.01 ***p < 0.001, ****p < 0.0001.

Figure 4 .
Figure 4. Impaired intracellular ROS generation by neutrophils from COVID-19 patients upon stimulation.Neutrophils were stimulated for 1 h with ssRNA and TNF or GM-CSF and fMLP or cultured without stimuli.(A) Representative histograms of DHR-123 stained neutrophils are depicted of an HD, a nonhospitalized, and a hospitalized patient at day 0. (B) Intracellular ROS level is depicted of unstimulated neutrophils from HDs (n = 17), nonhospitalized (n = 28), or hospitalized patients (n = 35) at day 0. (C) ROS level of unstimulated neutrophils is shown at day 0 (n = 28), day 28 (n = 32) or day 180 (n = 30) of nonhospitalized patients and at day 0 (n = 35), day 28 (n = 23), or day 180 (n = 23) of hospitalized patients as compared with HDs.(D) Intracellular ROS is depicted as fold change to unstimulated ssRNA + stimulated neutrophils of HDs (n = 17), nonhospitalized (n = 28), or hospitalized patients (n = 35) at day 0. (E) ROS production in neutrophils (fold change to unstimulated) is shown at day 0 (n = 28), day 28 (n = 32), or day 180 (n = 30) of nonhospitalized patients, and day 0 (n = 35), day 28 (n = 23), or day 180 (n = 23) of hospitalized patients as compared with HDs.(F) Neutrophils were stimulated with GM-CSF + fMLP and intracellular ROS production (fold change to unstimulated) is shown at day 0 of all three groups.(G) ROS production in neutrophils (fold change to unstimulated) is shown at day 0, 28, or 180 of non-and hospitalized patients as compared with HDs.Lines indicate the median value for each study group.Dotted line indicates the median value of HDs.Symbols in red indicate ICU patients.Significant differences are indicated as *p < 0.05, **p < 0.01 ***p < 0.001, ****p <0.0001.
* Symptoms registered within 2 weeks after the onset.