Infection with SARS-CoV-2 Is Associated with Elevated Levels of IP-10, MCP-1, and IL-13 in Sepsis Patients

Immunothrombosis, an excessive inflammatory response with simultaneous overactivation of the coagulation system, is a central pathomechanism in sepsis and COVID-19. It is associated with cellular activation, vascular damage, and microvascular thrombosis, which can lead to multiple organ failure and death. Here, we characterized factors related to immunothrombosis in plasma samples from 78 sepsis patients. In the course of routine clinical testing, SARS-CoV-2 was detected in 14 of these patients. Viral infection was associated with a higher mortality. Both, COVID-19 negative and COVID-19 positive sepsis patients showed increased levels of effectors of immunothrombosis, including platelet factor 4, D-dimer, nucleosomes, citrullinated histone H3, high mobility group box-1 protein, as well as phosphatidylserine-expressing platelet-derived extracellular vesicles, compared to healthy controls (n = 25). Using a 27-plex cytokine bead array, we found that Interleukin (IL)-1ra, IL-6, IL-8, IL-13, tumor necrosis factor (TNF)-α, interferon inducible protein (IP)-10, monocyte chemotactic protein (MCP)-1, macrophage inflammatory protein (MIP)-1α, and granulocyte-colony stimulating factor (G-CSF) were elevated in both, COVID-19 negative and COVID-19 positive sepsis patients, as compared to healthy controls. SARS-CoV-2 infection was associated with elevated levels of IP-10, MCP-1, and IL-13, while all other mediators widely overlapped between COVID-19 negative and COVID-19 positive patients.


Introduction
Sepsis is a life-threatening systemic inflammatory syndrome triggered by an inadequate host response to infection [1]. COVID-19, caused by the severe acute respiratory syndrome coronavirus SARS-CoV-2, shares many pathophysiological and clinical features with sepsis, including coagulopathy and vascular microthrombosis, as well as multiple organ failure [2][3][4], which accounts for most of the deaths in COVID-19 [5].

Data Collection
Clinical data collected from the electronic medical record included the date of sampling, gender, age, ward, the result of routine blood culture (BC), clinical diagnosis, death within one month after sampling, antimicrobial treatment, puncture site, C-reactive protein (CRP), procalcitonin (PCT), red blood cell, white blood cell, monocyte, neutrophil, lymphocyte, and platelet counts, as well as neutrophil-to-lymphocyte ratio (NLR). The 14 COVID-19 patients were classified according to the WHO severity criteria (moderate, n = 6; severe, n = 5; critical, n = 3) [25].

Blood Culture
Routine blood culture testing was performed on the wards by inoculating at least one set of BACTEC™ Plus Aerobic/F and BACTEC™ Lytic/10 Anaerobic/F culture bottles (both BD Diagnostics, Vienna, Austria) with the patient's whole blood. Culture bottles were immediately transported to the routine microbiological laboratory and incubated at 37 • C using the BACTEC FX Packaged continuous BC monitoring system (BD Diagnostics) until growth or for 5 days post-inoculation. Positive BCs were further used for pathogen identification after cultivation on agar plates using matrix-assisted laser desorption/ionization time of flight mass spectrometry (MALDI-TOF MS) based on the reference Biotyper library v4.1 (both Bruker Daltonik, Bremen, Germany) and antimicrobial susceptibility testing (AST). Disk diffusion was used for the latter and was performed in accordance with the current EUCAST guidelines at the time point of sampling [26].

Coagulation-Related Parameters
In addition to platelet counts, plasma concentrations of PF4 and D-dimer were quantified by ELISA (R&D Systems, Minneapolis, MN, and Technozym, Vienna, Austria, respectively). The cell death detection ELISA (Roche, Mannheim, Germany) was used to quantify nucleosomes in plasma. Results are indicated as arbitrary units (AU), which represent the ratio of the maximum measured absorbance and the absorbance of the individual samples. Citrullinated histone H3 and HMGB-1 were quantified by ELISA (Cayman Chemical, Ann Arbor, MI, and IBL International, Hamburg, Germany).

Characterization of Extracellular Vesicles
EVs were characterized using a CytoFLEX LX flow cytometer (Beckman Coulter, Brea, CA, USA) equipped with 405 nm, 488 nm, 561 nm, and 631 nm lasers. For staining, samples were diluted 1:100 in sterile-filtered Annexin V (Anx5) binding buffer (0.1 µm Minisart syringe filter, Sartorius Stedim Biotech, Goettingen, Germany). Aliquots (100 µL) of the diluted samples were incubated for 15 min at room temperature in the dark with APC-conjugated Anx5 (BD Biosciences, San Jose, CA, USA) as a marker for EVs exposing phosphatidylserine in combination with PE-conjugated anti-hCXCL4 (R&D Systems) and PC7-conjugated anti-CD41 or with FITC-conjugated anti-CRP (Abcam, Cambridge, UK). All fluorochrome conjugates and the respective antibody clones are listed in Supplementary  Table S1. To remove potential precipitates, fluorochrome conjugates were centrifuged at 18,600× g for 10 min at 4 • C prior to use. Stained samples were diluted 1:5 in sterile-filtered Anx5 binding buffer (BD Biosciences) prior to analysis. Calibration of the flow cytometer was performed with fluorescent silica beads (1 µm, 0.5 µm, 0.1 µm; excitation/emission 485/510 nm; Kisker Biotech, Steinfurt, Germany). The triggering signal for EVs was set to the violet side scatter (405 nm), and the EV gate was set below the 1 µm bead cloud (Supplementary Figure S1) as previously described [27,28]. Acquisition was performed for 2 min at a flow rate of 10 µL/min. Data were analyzed using the Kaluza Software (Beckman Coulter, version 2.1). EVs were identified as Anx5 + events in the EV gate.

Statistical Analysis
Statistical analysis was carried out using R (version 4.1.3). We reported the biascorrected and accelerated (BCa) bootstrap 95% confidence intervals (CI) of the median. If the bootstrapping confidence interval returned the warning "extreme order statistics used as endpoints", we used the percentage bootstrap confidence interval. Calculation of confidence intervals was performed by nonparametric bootstrapping using the boot package with 100,000 iterations [31,32]. Correlation analysis was performed using the Spearman correlation test and heat map analyses of the correlation coefficients were performed.

Patient Characteristics
Samples from 78 sepsis patients were analyzed in this study. Fourteen of these patients were diagnosed with COVID-19 during routine clinical testing in the course of the pandemic, with mean Ct values of 27.9 for the orf1a/b gene and 28.3 for the E gene. Twenty-five healthy individuals served as controls. The patient characteristics are summarized in Table 1. The in-hospital mortality was 11% for COVID-19 negative and 43% for COVID-19 positive patients. Bacteremia was confirmed by blood culture in 9 of the 78 sepsis patients (COVID-19 negative, n = 8; COVID-19 positive, n = 1). The panel of identified pathogens included Staphylococcus aureus (n = 4), Pseudomonas aeruginosa (n = 2), Escherichia coli (n = 2), Klebsiella pneumoniae (n = 1), Streptococcus pneumoniae (n = 1), Enterobacter cloacae (n = 1), and Staphylococcus epidermidis (n = 1). Hereby, two patients suffered from co-infections with up to three different pathogens, and Pseudomonas aeruginosa was isolated from one COVID-19 positive patient. Blood culture-positive patients exhibited higher PCT levels than blood culture-negative patients, 16 Table 2. The parameters presented in Table 2 were grouped into factors related to coagulation and complement activation (Figure 1), as well as cytokines, chemokines, and growth factors ( Figure 2). Mediator profiles of blood culture-negative and blood culture-positive patients are summarized in Supplementary Table S2, and mediator profiles of survivors and non-survivors are compiled in Supplementary Table S3   Data are represented as mean ± standard deviation and counts (percentage). Laboratory parameters are reported as median (lower; upper confidence interval). a n = 10; n.a., not applicable; ICU, intensive care unit; CRP, C-reactive protein; PCT, procalcitonin.

Coagulation-and Complement-Related Parameters
Coagulation-and complement-related parameters included platelet counts, PF4, D-dimer, nucleosomes, citrullinated histone H3, HMGB-1, EV counts, as well as the terminal complement complex C5b-9. EVs were further analyzed regarding their platelet origin (CD41 + ) and their association with CRP (CRP + ). Both patient groups showed a trend towards lower platelet counts, however, the confidence intervals for both groups overlapped with the control group. D-dimer was elevated in both patient groups. PF4 was increased in COVID-19 negative patients, while the confidence intervals overlapped with the control for COVID-19 positive patients.
We also measured markers of NET formation and observed increased nucleosome and citrullinated histone H3 levels in both patient groups. Likewise, HMGB-1, a known inducer of NET formation, as well as C5b-9 were elevated in both patient groups.

Discussion
Blood culture confirmed bacterial infection in 11.5% of the 78 sepsis patients included in our study (13% of the COVID-19 negative and 7% of the COVID-19 positive sepsis patients). Viral co-infection was associated with a higher prevalence of ICU admission and a higher mortality. Our data confirm recent meta-analyses reporting bacterial coinfections in 7-8% of COVID-19 patients, most frequently in those with critical illness [34,35]. The close connection between bacterial and viral infection is highlighted by recent work on mechanisms leading to gut dysbiosis in COVID-19, showing that SARS-CoV-2 can induce dysbiosis by triggering intestinal inflammation, by dysregulating the angiotensinconverting enzyme 2 (ACE2), which is essential for amino acid metabolism, and by infection of intestinal bacteria [36,37].
In line with previous studies, our data suggest that inflammatory mediator profiles in sepsis patients who tested positive for SARS-CoV-2 do not substantially differ from those of critically ill patients suffering from sepsis or acute respiratory distress syndrome [48]. We found increased levels of IL-1ra, IL-6, IL-8, IL-13, TNF-α, IP-10, MCP-1, MIP-1α, and G-CSF in both, COVID-19 negative and COVID-19 positive sepsis patients, as compared to healthy controls. Three out of the 27 factors included in the cytokine bead array, IP-10, MCP-1, and IL-13, were higher in COVID-19 positive than in COVID-19 negative sepsis patients. The chemokine IP-10 is mainly released by lymphocytes, neutrophils, and endothelial cells in response to IFNs and lipopolysaccharide to mediate the recruitment of immune cells [49]. Elevated IP-10 levels, which increased with disease progression, have been described in sepsis patients [50]. Elevation of IP-10 was also observed in COVID- 19, where several studies demonstrated an association of IP-10 with increased viral titers and disease severity [39,51]. Along with IP-10, MCP-1 has been associated with disease progression in COVID-19 patients in previous studies, as well [39,52,53]. Consistent with our findings, Donlan and co-workers detected significantly increased IL-13 levels in COVID-19 patients [54]. IL-13 is a type-2 cytokine for the recruitment of eosinophils and M2 macrophages to the lung and is implicated in smooth muscle cell proliferation as well as in mucus production and fibrosis [55]. Its elevation has been closely related to respiratory failure requiring mechanical ventilation, while its blockade showed the potential to ameliorate disease severity [54].
The inflammatory responses in sepsis and in COVID-19 are highly dynamic, resulting in pronounced heterogeneity of patient cohorts. Under this aspect, it is essential to consider that most studies, including ours, are limited to the analysis of single time points rather than providing time courses. Therefore, they can only yield snapshots of the dynamic inflammatory response. From this point of view, a recent prospective study by Loftus and co-workers is particularly valuable. They compared inflammatory and immunosuppressive profiles of 30 critically ill patients with SARS-CoV-2 infection with ten critically ill bacterial sepsis patients at five-time points over a period of 21 days. According to this study, bacterial sepsis was rather associated with early, severe inflammation and profound immunosuppression, while SARS-CoV-2 patients had less severe early inflammation, but persistent immunosuppression, making them particularly vulnerable to secondary infection [56].
The activation and interaction of platelets, neutrophils, and monocytes are major drivers of immunothrombosis. Elevated levels of PF4 have been linked to enhanced platelet activation [57][58][59] and to a hypercoagulable state associated with increased D-dimer levels in sepsis and COVID-19 [9,18,60]. Our data support these observations, since both, COVID-19 negative and COVID-19 positive sepsis patients had increased PF4 and Ddimer levels compared to healthy individuals. However, we could not confirm previous findings [46,61] on distinct D-dimer levels among sepsis and COVID-19 patients.
Activated neutrophils exacerbate immunothrombosis by releasing NETs, DNA fibers decorated with antimicrobial peptides and histones. Exaggerated NET formation has been correlated with disease severity and poor clinical outcome in sepsis as well as COVID-19 [9,[62][63][64][65]. NETosis is induced by a range of proinflammatory mediators, including IL-1β and IL-8, while the increased generation of reactive oxygen species is a crucial intracellular process triggering NETosis. Platelet activation represents another route of NET generation, and platelets enhance NETosis by interacting with neutrophils through TLR4, PF4, and extracellular vesicle-dependent processes. NET-associated histones induce further platelet activation and contribute to vascular damage due to their cytotoxicity.
We determined nucleosomes and citrullinated histone H3 as markers of NET formation. Both were elevated in COVID-19 positive and COVID-19 negative sepsis patients, supporting previous findings by Traby and co-workers, who reported elevated levels of citrullinated histone H3, D-dimer, CRP, IL-6, as well as an increased neutrophil-to-lymphocyte ratio in COVID-19 patients [60].
Extracellular vesicles are increasingly considered as amplifiers of immunothrombosis, particularly the subset of EVs exposing phosphatidylserine on their surface, which propagates thrombin generation [23,24]. In line with previous studies [60,66], we observed elevated levels of phosphatidylserine-exposing EVs in COVID-19 negative as well as COVID-19 positive sepsis patients, the majority of which were platelet-derived. While the levels of platelet-derived EVs did not differ between COVID-19 negative and COVID-19 positive sepsis patients, the concentration of CRP + EVs was higher in COVID-19 positive sepsis patients. The association of CRP with EVs may further support inflammation, as there is increasing evidence that functionally inert pentameric CRP dissociates to proinflammatory monomeric CRP upon binding to phosphocholine-enriched membranes [67].

Conclusions
Our study provides evidence that the profiles of inflammatory mediators largely overlap between COVID-19 negative and COVID-19 positive sepsis patients. Our data support the notion that activated platelets and platelet-derived EVs are potential mediators of immunothrombosis in sepsis and COVID-19.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/diagnostics13061069/s1, Figure S1: Calibration and controls for the flow cytometric characterization of extracellular vesicles.; Figure S2: Spearman-based correlation plot of all inflammatory mediators analyzed in COVID-19 negative sepsis patients.; Figure S3: Spearman-based correlation plot of all inflammatory mediators analyzed in COVID-19 positive sepsis patients.; Table S1: Antibodies and fluorochrome conjugates used for flow cytometric characterization of extracellular vesicles.; Table S2: Mediator profiles in sepsis patients with unknown and confirmed bacteremia.; Table S3: Inflammatory mediator profiles in survivors and non-survivors.  Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Data Availability Statement: All authors confirm that all relevant data are included in the article. Additional statistical data are available from the corresponding author upon request.