Extracellular traps derived from macrophages, mast cells, eosinophils and neutrophils are generated in a time‐dependent manner during atherothrombosis

Abstract Extracellular traps generated by neutrophils contribute to thrombus progression in coronary atherosclerotic plaques. It is not known whether other inflammatory cell types in coronary atherosclerotic plaque or thrombus also release extracellular traps. We investigated their formation by macrophages, mast cells, and eosinophils in human coronary atherosclerosis, and in relation to the age of thrombus of myocardial infarction patients. Coronary arteries with thrombosed or intact plaques were retrieved from patients who died from myocardial infarction. In addition, thrombectomy specimens from patients with myocardial infarction were classified histologically as fresh, lytic or organised. Neutrophil and macrophage extracellular traps were identified using sequential triple immunostaining of CD68, myeloperoxidase, and citrullinated histone H3. Eosinophil and mast cell extracellular traps were visualised using double immunostaining for eosinophil major basic protein or tryptase, respectively, and citrullinated histone H3. Single‐ and double‐stained immunopositive cells in the plaque, adjacent adventitia, and thrombus were counted. All types of leucocyte‐derived extracellular traps were present in all thrombosed plaques, and in all types of the in vivo‐derived thrombi, but only to a much lower extent in intact plaques. Neutrophil traps, followed by macrophage traps, were the most prominent types in the autopsy series of atherothrombotic plaques, including the adventitia adjacent to thrombosed plaques. In contrast, macrophage traps were more numerous than neutrophil traps in intact plaques (lipid cores) and organised thrombi. Mast cell and eosinophil extracellular traps were also present, but sparse in all instances. In conclusion, not only neutrophils but also macrophages, eosinophils, and mast cells are sources of etosis involved in evolving coronary thrombosis. Neutrophil traps dominate numerically in early thrombosis and macrophage traps in late (organising) thrombosis, implying that together they span all the stages of thrombus progression and maturation. © 2018 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of Pathological Society of Great Britain and Ireland.


Introduction
Extracellular traps (ETs) are thread-like structures of decondensed DNA decorated with proteins from cytoplasmic granules [1,2]. Neutrophil extracellular traps (NETs) were the first that were identified, and it was found that these structures are involved in the elimination of pathogens [3]. The process of ET formation is called 'etosis', which is a specific type of cell death [4]. Nowadays, there is increasing interest in the role of NETs also in autoimmune and cardiovascular diseases [5]. Several studies have documented their presence in human atherosclerotic plaques and thrombosis [6][7][8][9][10]. NETs are believed to promote endothelial dysfunction [1]; to stimulate thrombus formation on disrupted plaque, mainly by providing scaffolds for fibrin deposition [11]; and to stabilise clot formation through the activation of coagulation cascade and thrombin generation, via their bearing of tissue factor (TF) [9,12].
Recent research has shown that ETs can also be generated by cells other than neutrophils [13], such as macrophages [14][15][16], mast cells [17,18], and eosinophils [19,20]; these are termed macrophage extracellular traps (METs), mast cell extracellular traps (MCETs), and eosinophil extracellular traps (EETs), respectively. METs, MCETs, and EETs have been identified in infectious [15,18,21,22] and autoimmune diseases [17,19]. Given the abundant presence of 506 KR Pertiwi et al macrophages [23][24][25], and to a lesser extent also mast cells [26,27] and eosinophils in atherosclerotic plaques [28,29], we hypothesised that these structures may contribute to atherothrombosis, and in particular to the process of thrombosis resulting from plaque rupture or erosion. Therefore, we investigated the presence and relative extent of ETs released by macrophages (METs), mast cells (MCETs), and eosinophils (EETs) in intact and thrombosed (eroded and ruptured) coronary atherosclerotic plaques of autopsied patients who had died of acute myocardial infarction (MI). In addition, we evaluated whether the occurrence of different types of etosis also relates to the age of a coronary thrombus in MI patients. For this purpose, we used coronary thrombectomy materials, in which the age of the thrombus may vary substantially from fresh (representing recent-onset thrombus) to cell-rich organised masses (representing thrombus that is weeks old) [30]. For quantitative comparison of the extent of different ET types within the plaques and the thrombi, we also included neutrophils and NETs in this study, as previously reported [6,31].

Autopsy materials
Paraffin blocks containing coronary plaques obtained at autopsy from patients with acute MI were derived from the pathology archives of the Academic Medical Center, Amsterdam. For the purpose of this study, six plaques with intact endothelial surface without thrombus and 12 thrombosed plaques were selected, of which the latter had either a fibrous cap rupture (n = 6) or plaque erosion underlying the thrombus (n = 6).

Thrombectomy specimens
Paraffin blocks containing thrombus aspiration materials derived from MI patients were retrieved from the pathology archives of the Academic Medical Center, Amsterdam. The retrieved thrombus blocks were cut into 5-μm-thick sections and histomorphologically graded on H&E-stained sections according to the age of thrombus into three categories -fresh, lytic, and organised -as previously described [30,32,33]. Fresh thrombus (up to 1 day) was composed of intact platelets, erythrocytes, and/or granulocytes; lytic thrombus (1-5 days) was identified by the presence of colliquation necrosis and karyorrhexis of granulocytes; and organised thrombus (> 5 days) was marked by the appearance of (myo)fibroblasts and extracellular matrix deposits. Thrombus materials with a mixed composition of different ages were separately graded. From the total file of archived specimens, we randomly selected 48 specimens, resulting in 24 fresh, 26 lytic, and 18 organised thrombi for further immunohistochemistry in this study.
Criteria for the proper secondary use of human tissue in The Netherlands were met and accordingly the AMC Medical Ethical Board grants a waiver for the use of 'left-over materials' that are used anonymously.

Quantification of immunostaining results
Digitised immunostained sections were downloaded from the Philips image management system. Using the downloaded images, we further investigated the areas of interest for quantification. For plaque specimens, the number of immunopositive cells was counted in different topographic locations: plaques (intact plaques), thrombus and adjacent plaque (thrombosed plaques), and adventitia (both plaque types); whereas for coronary thrombi, these were quantified in areas of interest according to the distinct features of each histological thrombus age. The surface areas of the selected regions were measured in mm 2 . The separate downloaded images were then digitally aligned by a non-linear registration as described previously [6], creating image stacks to calculate the immunopositive cell-specific (MPO, CD68, tryptase or EMBP) antibodies and their co-localisation with CitH3. NETs were determined as MPO + CD68 − CitH3 + cells, while METs were either CD68 + MPO − CitH3 + or CD68 + MPO + CitH3 + cells (see supplementary material, Figure S1). In terms of MCETs and EETs, they were defined as tryptase + CitH3 + cells or EMBP + CitH3 + cells, respectively. All measurements were expressed as number of cells per mm 2 .

Statistics
Statistical analysis was conducted using SPSS 24.00 (IBM Corporation, Armonk, NY, USA). According to the normal or non-normal distribution of the data, we performed either a t-test or a Mann-Whitney test for comparison between thrombosed and intact plaques, and either Kruskal Wallis or ANOVA for different thrombus ages with a post hoc test when the results were significant (a value of p < 0.05 was considered significant).

Extracellular traps in intact and thrombosed atherosclerotic plaques at autopsy
The occurrence of NETs, METs, MCETs, and EETs was noticed in all 12 coronary segments with thrombosed plaques. They were present in different locations: not only in the plaque but also in adjacent thrombus and in surrounding perivascular fat (see Figure 1 for representative examples of NETs, METs, MCETs, and EETs in plaques). METs were the most numerous type of extracellular trap inside the intact (not thrombosed) plaques (Figure 2A), and were located mostly around the lipid core of lesions. MCETs and EETs were only sparsely present in this location (Figure 2A). In the adjacent adventitia and perivascular fat of these intact plaques, not only NETs but also METs and MCETs were low in number, and EETs were almost never encountered ( Figure 2B). When compared with these intact atherosclerotic plaques, the average number of NETs, METs, MCETs, and EETs was significantly increased in the plaques with thrombotic complications, either in the (thrombo)plaque or in the adventitia and perivascular fat (p < 0.05, Figure 2A,B). In addition, the majority of ETs in thrombosed plaques were neutrophil-derived, followed by macrophage-, mast celland eosinophil-derived traps.

Extracellular traps in coronary thrombus aspirates
Next, we investigated the presence and extent of different types of ETs in relation to the age of the coronary thrombus in thrombus aspirates. In line with the findings on autopsy specimens, all types of ETs could be detected. Cell-specific differences in the source of traps were noticed when thrombus specimens of different ages were compared (see Figure 3 and supplementary material, Figure S3 for representative examples of NETs, METs, MCETs, and EETs in coronary thrombi). Neutrophils were the major source of extracellular traps ( Figure 4) in fresh (151 per mm 2 ) and lytic thrombi (136 per mm 2 ), followed by METs in fresh (45 per mm 2 ) and in lytic thrombi (99 per mm 2 ). In organised thrombi, on the other hand, METs exceeded NETs in number (30 per mm 2 versus 20 per mm 2 , Figure 4). The numbers of mast cell-and eosinophil-derived traps were much lower, with most MCETs present in organised thrombi (6 per mm 2 , Figure 4) and EETs in lytic thrombi (7 per mm 2 , Figure 4).

Discussion
In this study, we showed that not only neutrophils but also other types of leucocytes form extracellular traps (ETs) in atherosclerotic plaques and to a much larger extent in the coronary thrombus of MI patients. NETs and METs are the two most prominent contributors to etosis in coronary atherothrombosis, although low numbers of mast cell-and eosinophil-derived traps (MCETs and EETs) can also be encountered in the lesions. When the age of a thrombus was taken into consideration, NETs appeared to dominate in the early (fresh) stages, whereas METs are significantly more numerous in the late (organising) stages of thrombus formation (see Figure 4 and supplementary material, Figure S3G-L). These findings imply that once etosis is considered as a process involved in the progression and maturation of coronary thrombus after its onset, different types of leucocytes participate in a time-dependent manner.

Macrophage extracellular traps (METs) in atherothrombosis
Recent studies have shown that macrophages also are capable of releasing extracellular DNA in response to micro-organisms and inflammatory mediators [13,34]. In vitro, METs were visualised with the citrullinated histone H4 in IFN-γ-induced human monocyte-derived macrophages upon Mycobacterium tuberculosis infection [15] and with citrullinated histone H3 in human peripheral monocytes infected with Candida albicans [35]. In vivo, a histological study on kidney biopsies from patients with ANCA-associated vasculitis reported the presence of METs (identified as MPO + CD68 + elastase − macrophages co-expressing CitH3) in glomerular MPO-containing macrophages. In all those studies, similar antibodies were applied to those we used in the present sudy, to visualise histones as the key elements in the structure of extracellular traps [36].
In intact plaques, we observed significantly higher numbers of METs compared with the other types of ETs; they were mostly located around and inside the lipid core, which also contained immunostainable remnants of dead macrophages. Macrophage cell death, generally attributed to necrosis and apoptosis, is considered an important contributor to the growth of lipid cores of atherosclerotic plaques. Death of (foamy) macrophages leads to extracellular spill of lipids and, as a consequence, volume expansion of the lipid core (also termed the 'graveyard of macrophages') [37]. Our study reveals that also 'metosis', as shown by the presence of METs, contributes to macrophage cell death and, as can be anticipated, to the expansion of the lipid-rich atheroma of plaques which typifies the vulnerable type of lesions.
Still, in thrombosed coronary atherosclerotic plaques (be they ruptured or eroded), the numbers of METs were significantly increased compared with intact plaques, not only in the plaque or overlying thrombus but also in the periadventitial tissues surrounding these complicated plaques. As yet, it remains unknown whether METs contribute to the onset of thrombus formation or otherwise aggravate the process of progression of a thrombus, for example by forming a scaffold for fibrin network or stimulating other pro-thrombotic and pro-coagulant mediators. The significant number of METs that we found in the perivascular fat tissue around thrombosed plaques follows the pattern that we reported also for NETs previously [6]. We interpreted the abundant presence of NETs in the adventitia and perivascular fat tissue as a sequential effect of acute plaque complications in which the periadvential inflammation is triggered by neutrophil activation and cell death. It could be that the same process applies for the formation of METs at this location. However, at present, we cannot say whether the formation of METs can be considered beneficial or detrimental in atherothrombosis, as macrophage functions are considerably dynamic according to their micro-environment, differentiation, and polarisation states [23,38].

Mast cell and eosinophil extracellular traps (MCETs and EETs) in atherothrombosis
Several in vitro and in vivo studies have described the presence of MCETs and EETs mainly in infectious, allergic, and autoimmune diseases. In vitro, MCETs were first recognised on human mast cell-1 (HMC-1) infected with several bacteria [18], while EETs were noticed in human purified eosinophils primed with IL-5 and IFN-γ under the influence of lipopolysaccharide (LPS), complement (C5a) or eotaxin [39]. In vivo, MCETs were identified in psoriatic lesions and in skin explant cultures treated with IL-23 and/or IL-1β by using immunofluorescence with tryptase and DAPI [17], while EETs were detected in the intestines of patients with Crohn's disease [20], in skin diseases [19], and in bronchial biopsies of asthmatic patients [40] with the use of DNA and EMBP immunostaining. Overall, the use of cell-specific antibodies in those studies -tryptase for mast cells and EMBP for eosinophils -was also implemented in our study.
In thrombosed coronary atherosclerotic plaques (be they ruptured or eroded), we found only a relatively small number of MCETs and EETs, in line with the usually small number of mast cells and eosinophils, respectively, that have been reported in plaques, including the lesions of MI patients [41,42]. Still, despite their low number, mast cells attract interest because of the role that they play in the process of plaque destabilisation via their release of pro-inflammatory cytokines and vasoactive mediators (histamine, chymase, and tryptase). It could be that the formation of traps by mast cells facilitated these functions [11,13]. Although to a lesser extent compared with thrombosed plaques, most mast cells have been reported in the adventitia of intact plaques around microvessels [43,44], which is also the site where we observed most MCETs. Therefore, it is conceivable that their traps stimulate pro-inflammatory functions in the periadventitial fat [45][46][47].
Furthermore, investigations into the underlying mechanisms of etosis other than netosis are currently less numerous [13,34]. The formation of ETs by neutrophils, macrophages, mast cells, and eosinophils may share some resemblances as well as differences. For example, the generation of reactive oxygen species (ROS) [13] and the involvement of enzymes (MPO, elastase or tryptase) in ET generation [18,34] will likely be the same in neutrophils, macrophages, and mast cells. On the other hand, cytokines and chemokines involved in the generation of traps can be similar or different for the various types of leucocytes: for example, IFN-γ and C5a in NETs and EETs; IL-8 in NETs; and IL-23 and IL-1β in MCETs [13].
In conclusion, not only neutrophils but also macrophages, and to a lesser extent also mast cells and eosinophils, generate ETs after the onset of coronary plaque complications. In turn, the formation of ETs spans all stages of coronary thrombosis evolution. Different types of leucocytes and their ETs participate in orchestrating thrombus organisation and maturation towards stability in a time-dependent manner. More knowledge about the role of ETs during atherothrombotic disease is important, since it may provide new strategies in the treatment of cardiovascular disease.