Background
Sepsis is frequently encountered in the intensive care unit (ICU) and has a high mortality rate [
1], and so the development of more efficient methods for diagnosis and treatment is required. Methods of assessing the severity of sepsis include the Acute Physiological and Chronic Health Evaluation (APACHE) II score [
2] and the Sepsis-Related Organ Failure Assessment (SOFA) score [
3]. In addition, numerous biomarkers [
4] are frequently used to predict morbidity and mortality in patients with sepsis, including acute-phase proteins such as procalcitonin, C-reactive protein, inflammatory cytokines and chemokines, cell surface proteins of inflammatory cells, and coagulation markers. While these physiological scores and biomarkers are useful, novel biomarkers that can achieve more reliable early diagnosis and assist therapeutic decision making are urgently needed.
Neutrophil extracellular traps (NETs) are a potential biomarker for sepsis because neutrophils are the most abundant of the leukocytes and play a central role in the pathogenesis of sepsis. Neutrophils attack extracellular microbes by releasing toxic proteins and enzymes, including myeloperoxidase (MPO), neutrophil elastase (NE), and defensins, from their granules through the process of degranulation [
5].
Recently, detection of DNA fragments circulating in the bloodstream known as cell-free DNA (cf-DNA) has received increasing attention as a prognostic marker. The plasma level of cf-DNA has been examined in various acute and chronic disorders, including trauma [
6,
7], sepsis [
8], cancer [
9], stroke [
10], and myocardial infarction [
11]. However, plasma cf-DNA levels are not only increased as a result of release from NETs but are also elevated by cellular necrosis and apoptosis. Thus, only detecting an increase in cf-DNA is insufficient evidence to verify NET formation [
12].
NET remnants are complexes formed between DNA and neutrophil-derived proteins, including MPO and NE. Myeloperoxidase-conjugated DNA (MPO-DNA) and neutrophil elastase-conjugated DNA (NE-DNA) can be detected in body fluids by enzyme-linked immunosorbent assay (ELISA) as an objective, quantitative, and specific marker of NET formation [
13‐
15]. In this study, circulating levels of soluble NET remnants (MPO-DNA and cf-DNA) were investigated in patients with septic shock to evaluate the extent of NET formation during sepsis.
The aim of this study was to compare the plasma levels of MPO-DNA and cf-DNA between septic shock patients and healthy volunteers, as well as to investigate the relationship of circulating NET levels with 28-day mortality, organ dysfunction, and known parameters of the severity of sepsis.
Discussion
This study assessed the profile of plasma MPO-DNA and cf-DNA in patients with septic shock, resulting in three main findings. First, both plasma MPO-DNA and cf-DNA levels were significantly higher in septic shock patients on admission than in healthy volunteers. Second, the plasma MPO-DNA level was closely related to the severity of organ dysfunction, including parameters such as the P/F ratio, MAP, and SOFA score, whereas the MPO-DNA level was not related to the DIC score. Third, a high plasma MPO-DNA levels on days 3 and 7 of sepsis were associated with 28-day mortality.
Elevation of the plasma cf-DNA level in our septic shock patients on admission is consistent with previous reports of a several-fold to hundred-fold increase of cf-DNA in septic patients compared with healthy volunteers [
8,
21‐
23]. Plasma cf-DNA has several potential sources in patients with septic shock, including release from NETs, cellular necrosis or apoptosis, and destruction of pathogens [
24]. Thus, circulating cf-DNA in septic shock patients is derived from dead cells or from neutrophils that have undergone NETosis. On the other hand, circulating MPO-DNA is more specific for NETs than cf-DNA [
25] because these web-like structures released by activated neutrophils are composed of DNA associated with neutrophil granule proteins such as NE, cathepsin G, and MPO [
5]. Elevated plasma levels of MPO-DNA have been reported in patients with transfusion-related acute lung injury [
26], ANCA-associated small vessel vasculitis [
27], severe coronary atherosclerosis [
28], and thrombotic microangiopathy [
29], and excessive NET formation is related to the pathogenesis of these conditions. Our finding of an increase in MPO-DNA indicates that NET formation is accelerated in the early stage of septic shock. This is consistent with a recent report by Kaufman et al. who measured plasma NE-DNA complexes to assess circulating NETs and identified an increase in patients with sepsis or burns [
30]. They concluded that NET formation is augmented in patients with SIRS regardless of whether the inflammatory insult is infectious or traumatic. The elevated plasma levels of IL-6 and IL-8 observed in patients with septic shock in this study may have facilitated NET formation since various cytokines (including tumor necrosis factor (TNF), IL-1, IL-8, and IL-6) are known to accelerate the production of NETs [
31,
32].
While this study revealed a significant increase in both cf-DNA and MPO-DNA levels on day 1 of sepsis, the cf-DNA level declined rapidly and MPO-DNA decreased more slowly. The different changes in these two markers over time can be explained by differing levels of resistance to digestion by DNases [
33], which selectively digest the DNA threads of NETs. While cf-DNA is rapidly digested by DNases, MPO-DNA appears to be far more stable and consequently persists for longer in the circulation [
34]. However, further investigation will be needed to clarify the detailed mechanisms involved in clearance of circulating NET remnants.
There have been some previous reports that the highest plasma cf-DNA levels are observed in septic shock patients who eventually die, suggesting that cf-DNA could be a useful prognostic marker [
8,
20‐
22,
35]. It has also been reported that elevated plasma levels of cf-DNA predict the development of multiple organ dysfunction in trauma patients [
7] and septic patients [
8]. However, we found no correlation between the plasma cf-DNA level and mortality, with MPO-DNA being superior to cf-DNA as a prognostic marker in patients with septic shock. The plasma MPO-DNA level was also closely related to several parameters of organ dysfunction, including the P/F ratio, MAP, and SOFA score. These differences in the relation with organ dysfunction or mortality between MPO-DNA and cf-DNA can be explained by the differing specificity of these two markers for NETs. Hamaguchi et al. studied a mouse model of sepsis due to cecal ligation and puncture (CLP), reporting that plasma cf-DNA is not derived from NETs released by activated neutrophils and is mainly from other host cells [
36]. In addition, a study performed in sepsis patients showed that cf-DNA predominantly has a low molecular weight (150–250 bp) corresponding to the size of apoptotic nucleosomal DNA [
23]. Thus, an increase in cf-DNA is probably related to cellular damage and apoptosis that may occur in various tissues by several mechanisms. Moreover, the influence of impaired renal and hepatic function on clearance of cf-DNA seems to be complex [
35]. Taken together, further prospective studies are needed to clarify the origin of cf-DNA and its usefulness as an accurate and sensitive biomarker for sepsis.
At present, it is unclear whether NET formation has an essential role in host defenses against bacterial invasion [
37]. In a mouse CLP model, Czaikoski et al. showed that degradation of circulating DNA (a major constituent of NETs) by systemic treatment with recombinant human DNase (rhDNAse) led to earlier death than in the control group, possibly due to an increase in the bacterial load [
38]. They concluded that NETs have a beneficial role in killing pathogens and that depletion of NETs leads to aggravation of polymicrobial sepsis. On the other hand, it has been reported that NETs are not required to control bacterial proliferation because congenital absence of NETs in peptidylarginine deiminase 4 knockout mice or treatment with rhDNAse to prevent NET formation did not increase the bacterial load in animals with sepsis [
39,
40]. Lefrançais et al. showed that an increase in NETs, as assessed from the plasma level of NE-DNA complexes, was associated with the severity of ARDS, while lower plasma DNase I levels were associated with development of sepsis-induced ARDS [
15]. They concluded that strategies to reduce NET levels could have a favorable effect on lung function. The inverse correlation between the plasma NET level and the P/F ratio observed in the present study corresponds to the findings reported by Lefrançais. Although an essential role of NETs in killing pathogens has not been demonstrated at present, our results confirm that excessive NET formation contributes to the development of multiple organ dysfunction and mortality in patients with septic shock.
According to the immunothrombosis hypothesis, NET formation is linked to platelet aggregation and hypercoagulation associated with septic shock [
41]. NETs can support immunothrombosis through binding to von Willebrand factor (vWF) [
42] and by activation of the platelet Toll-like receptor (TLR)4-dependent interaction between platelets and neutrophils [
43]. However, we found no correlation between the DIC score and the plasma MPO-DNA level in the present study. Consistent with our result, Kaufman et al. reported that neither platelet TLR4 nor plasma vWF levels were correlated with plasma NE-DNA in sepsis patients, suggesting that these two factors are not involved in the acceleration of NETosis in this condition [
30]. Very recently, Delabranche et al. showed that circulating NET levels (assessed by measuring plasma MPO-DNA) are elevated in sepsis patients with DIC compared with sepsis patients without DIC [
44], and they suggested that NETs may play a critical role in the onset of DIC accompanying sepsis. The discrepancies between these studies may be attributable to differences in the methods of measuring circulating NETs, as well as differences in the patients enrolled, statistical analysis, and treatment (including anticoagulant therapy for DIC). Anticoagulant therapy, including antithrombin, recombinant thrombomodulin (rTM), and protease inhibitor supplementation, is strongly recommended in the Japanese guidelines for management of DIC [
45,
46], and these therapies may affect the formation of NETs. Recently, administration of antithrombin was reported to reduce NET formation in the lungs during lipopolysaccharide (LPS)-induced endotoxemia [
47]. Serine protease inhibitors [
48] and rTM [
49] are also known to inhibit formation of NETs in vitro.
Some limitations of the present study should be considered. We only examined a limited number of patients, which means that our results require validation on a larger scale. Another limitation is that our age-matched control group consisted of healthy volunteers and not critically ill patients with or without sepsis. Therefore, our findings may have been due to acute illness or underlying comorbidities and not necessarily related to septic shock per se.