Flow Cytometry-Based Characterization of Mast Cells in Human Atherosclerosis

The presence of mast cells in human atherosclerotic plaques has been associated with adverse cardiovascular events. Mast cell activation, through the classical antigen sensitized-IgE binding to their characteristic Fcε-receptor, causes the release of their cytoplasmic granules. These granules are filled with neutral proteases such as tryptase, but also with histamine and pro-inflammatory mediators. Mast cells accumulate in high numbers within human atherosclerotic tissue, particularly in the shoulder region of the plaque. These findings are largely based on immunohistochemistry, which does not allow for the extensive characterization of these mast cells and of the local mast cell activation mechanisms. In this study, we thus aimed to develop a new flow-cytometry based methodology in order to analyze mast cells in human atherosclerosis. We enzymatically digested 22 human plaque samples, collected after femoral and carotid endarterectomy surgery, after which we prepared a single cell suspension for flow cytometry. We were able to identify a specific mast cell population expressing both CD117 and the FcεR, and observed that most of the intraplaque mast cells were activated based on their CD63 protein expression. Furthermore, most of the activated mast cells had IgE fragments bound on their surface, while another fraction showed IgE-independent activation. In conclusion, we are able to distinguish a clear mast cell population in human atherosclerotic plaques, and this study establishes a strong relationship between the presence of IgE and the activation of mast cells in advanced atherosclerosis. Our data pave the way for potential therapeutic intervention through targeting IgE-mediated actions in human atherosclerosis.


Introduction
Up to the present day, atherosclerosis, the main underlying pathology of acute cardiovascular syndromes like stroke or myocardial infarction, is the major cause of human mortality [1]. As in

Histology
The culprit part of atherosclerotic samples was placed in Kristensen's buffer for three to seven days for decalcification, after which the plaques were embedded in paraffin. Next, the plaques were sectioned in 5-µm thick sections using a microtome RM2235 (LEICA Biosystems, Amsterdam, The Netherlands). A Movat's pentachrome staining was routinely performed, and subsequently, the plaques were analyzed for histological parameters, as described in Table 1 (three sections/plaque), based on the semiquantitative scoring systems of the AtheroExpress biobank [23] and the Oxford Plaque Study [25]. In short, the plaques were assessed for the presence of unstable plaque features such as the presence of a necrotic core, inflammatory cells, and intraplaque hemorrhage, as well as stable plaque features such as smooth muscle cell (SMC)-rich extracellular matrix (ECM). To identify the mast cells in the lesion, atherosclerotic plaque sections were immunohistochemically stained for tryptase using an alkaline phosphatase-conjugated antibody directed against tryptase (1:250, clone G3, Sigma, Zwijndrecht, The Netherlands), after which nitro-blue tetrazolium and 5-bromo-4-chloro-3'-indolyphosphate were used as a substrate. Nuclear Fast Red was used as a counterstaining for the nuclei. For the morphologic analysis, slides were analyzed using a Leica DM-RE microscope (Leica Ltd., Cambridge, UK).

Flow Cytometry
The single plaque cells were stained with extracellular antibodies containing a fluorescent label, or were fixated and permeabilized (BD Biosciences, San Jose, CA, USA) for intracellular staining ( Table 2). The fluorescently labeled samples were measured on a FACS Canto II (BD Biosciences, San Jose (CA), USA) or Cytoflex (Beckman Coulter, Miami, FL, USA), and were analyzed using FlowJo software (v10).

Statistics
All of the data are depicted using GraphPad Prism 7.00. The values were tested for normalcy. Upon non-Gaussian distribution, an unpaired Mann-Whitney U-test was performed. In the case of more than two groups, a one way-ANOVA analysis was used. Differences lower than p < 0.05 were considered statistically significant.

Results
We analyzed the culprit part of the carotid and femoral plaques for its histology characteristics, based on the Movat's pentachrome staining ( Figure 1A). The characteristics of the individual plaques and the assessment of the plaque stability parameters are shown in Figure 1B. Overall, the presence of a necrotic core, inflammatory cells, and intraplaque hemorrhage establish that the majority of the plaques can be classified as advanced, as expected.
Next, we prepared single cell suspensions of the remainder of the individual plaques, and stained the cells for the surface markers to be analyzed using flow cytometry. In Figure 2A, we demonstrate the gating strategy that we followed in order to detect the human intraplaque immune cells. Specifically, we pre-selected all of the cells from the debris present in the human plaques based on their size (forward scatter, FSC) and granularity (side scatter, SSC). Of these, single cells were further separated according to their width (FSC-W) and area (FSC-A). In addition, the viability was detected according to the negative signal for a fluorescent viability dye (FVD − ). Viable white blood cells were identified according to the expression of the pan-leukocyte marker CD45. As the femoral plaques were generally bigger in size upon surgical removal compared with the carotid plaques, we were able to isolate more CD45 + immune cells from the femoral arteries (carotid: 12 × 10 5 ± 4.7 × 10 5 leukocytes vs. femoral: 76 × 10 5 ± 27 × 10 5 leukocytes; Figure 2B). Next, we prepared single cell suspensions of the remainder of the individual plaques, and stained the cells for the surface markers to be analyzed using flow cytometry. In Figure 2A, we demonstrate the gating strategy that we followed in order to detect the human intraplaque immune cells. Specifically, we pre-selected all of the cells from the debris present in the human plaques based on their size (forward scatter, FSC) and granularity (side scatter, SSC). Of these, single cells were further separated according to their width (FSC-W) and area (FSC-A). In addition, the viability was detected according to the negative signal for a fluorescent viability dye (FVD − ). Viable white blood cells were identified according to the expression of the pan-leukocyte marker CD45. As the femoral plaques were generally bigger in size upon surgical removal compared with the carotid plaques, we were able to isolate more CD45 + immune cells from the femoral arteries (carotid: 12 × 10 5 ± 4.7 × 10 5 leukocytes vs. femoral: 76 × 10 5 ± 27 × 10 5 leukocytes; Figure 2B).
Within these cells, we detected the population of intraplaque mast cells based on the high expression of their characteristic markers FcεRIα, the receptor for IgE [26], and of CD117-the receptor for stem cell factor, a growth factor required for the end-stage maturation of mast cells [27] ( Figure 2C). Accordingly, we observed that the percentage of mast cells out of all of the viable Within these cells, we detected the population of intraplaque mast cells based on the high expression of their characteristic markers FcεRIα, the receptor for IgE [26], and of CD117-the receptor for stem cell factor, a growth factor required for the end-stage maturation of mast cells [27] ( Figure 2C). Accordingly, we observed that the percentage of mast cells out of all of the viable leukocytes present inside the atherosclerotic plaques is 1.19% ± 0.31% for the carotid arteries (n = 9), and 1.32% ± 0.21% for the femoral arteries (n = 13) ( Figure 2D). Of note, a number of leukocytes may have been retained in the excluded debris material. Nonetheless, we enumerated the viable cells after manual quantification using Trypan Blue in relation to the percentage of viable CD45 + cells observed according to our gating strategy. We detected that the carotid plaques consisted of a mean 12 ± 5 × 10 3 mast cells, while the femoral plaques contained 94 ± 37 × 10 3 mast cells. Thus, because of the higher total number of leukocytes, we identified higher absolute mast cell numbers in the femoral plaques compared with the carotid plaques ( Figure 2E). have been retained in the excluded debris material. Nonetheless, we enumerated the viable cells after manual quantification using Trypan Blue in relation to the percentage of viable CD45 + cells observed according to our gating strategy. We detected that the carotid plaques consisted of a mean 12 ± 5 × 10 3 mast cells, while the femoral plaques contained 94 ± 37 × 10 3 mast cells. Thus, because of the higher total number of leukocytes, we identified higher absolute mast cell numbers in the femoral plaques compared with the carotid plaques ( Figure 2E). We proceeded to further characterize the human intraplaque cells with respect to their activation status. We therefore screened our cells for the expression of CD63, a lysosomal protein that fuses with the membrane upon the release of cellular content, and marks mast cell activation [28]. We detected 7.0 ± 3.5 × 10 3 CD63 + mast cells in carotid plaques, which indicates that about 56% of these cells are in an activated state, while the femoral plaques showed 57 ± 23 × 10 3 activated mast cells, at a total percentage of 61% ( Figure 3A,B). As the mast cell activation in, for example allergic reactions, predominantly occurs via IgE bound to its receptor (FcεR), we also stained the cells for the levels of IgE bound on their surface, and observed that inside the carotid arteries 7.8 ± 3.4 × 10 3 intraplaque mast cells, or approximately 63% of the total mast cell population, contained IgE, whereas in the femoral plaques 77 ± 36 × 10 3 mast cells, or 74% of all of the mast cells had IgE on their surface. Because IgE binding generally implies mast cell degranulation, we analyzed, within each arterial plaque sample, the population of mast cells that showed both bound IgE and CD63 expression (IgE + CD63 + ), as opposed to only IgE-binding (IgE + CD63 − ) and only a CD63 expression (IgE − CD63 + ) ( Figure 3B). We observed that 23.8% ± 3.7% of all human plaque mast cells We proceeded to further characterize the human intraplaque cells with respect to their activation status. We therefore screened our cells for the expression of CD63, a lysosomal protein that fuses with the membrane upon the release of cellular content, and marks mast cell activation [28]. We detected 7.0 ± 3.5 × 10 3 CD63 + mast cells in carotid plaques, which indicates that about 56% of these cells are in an activated state, while the femoral plaques showed 57 ± 23 × 10 3 activated mast cells, at a total percentage of 61% ( Figure 3A,B). As the mast cell activation in, for example allergic reactions, predominantly occurs via IgE bound to its receptor (FcεR), we also stained the cells for the levels of IgE bound on their surface, and observed that inside the carotid arteries 7.8 ± 3.4 × 10 3 intraplaque mast cells, or approximately 63% of the total mast cell population, contained IgE, whereas in the femoral plaques 77 ± 36 × 10 3 mast cells, or 74% of all of the mast cells had IgE on their surface. Because IgE binding generally implies mast cell degranulation, we analyzed, within each arterial plaque sample, the population of mast cells that showed both bound IgE and CD63 expression (IgE + CD63 + ), as opposed to only IgE-binding (IgE + CD63 − ) and only a CD63 expression (IgE − CD63 + ) ( Figure 3B). We observed that 23.8% ± 3.7% of all human plaque mast cells have IgE bound on their surface without expressing CD63, whereas the majority of mast cells, with 40.0% ± 3.9%, appeared to have IgE bound on their surface, and had also undergone degranulation. Interestingly, a proportion of mast cells, namely 19.6% ± 2.9%, appeared to be activated without showing any IgE-fragments bound on their surface, suggesting that this mast cell fraction had been activated via alternative mast cell activation pathways. The IgE-activated population was however significantly higher than the cells subjected to non-IgE mediated activation (p = 0.0005), and also higher than the mast cell population that showed a binding of IgE without being activated (p = 0.0067). When analyzing the carotid and femoral arteries separately, similar expression patterns were observed ( Figure 3E), suggesting that intraplaque mast cell activation mechanisms do not differ much between plaque locations.
40.0% ± 3.9%, appeared to have IgE bound on their surface, and had also undergone degranulation. Interestingly, a proportion of mast cells, namely 19.6% ± 2.9%, appeared to be activated without showing any IgE-fragments bound on their surface, suggesting that this mast cell fraction had been activated via alternative mast cell activation pathways. The IgE-activated population was however significantly higher than the cells subjected to non-IgE mediated activation (p = 0.0005), and also higher than the mast cell population that showed a binding of IgE without being activated (p = 0.0067). When analyzing the carotid and femoral arteries separately, similar expression patterns were observed ( Figure 3E), suggesting that intraplaque mast cell activation mechanisms do not differ much between plaque locations. As tryptase is the preferred marker to determine the presence of mast cells in tissue by means of immunohistochemistry ( Figure 4A), we also quantified the CD117 + FcεRI + mast cell populations that contained tryptase in a smaller sample of the same patients using flow cytometry (n = 7). We determined that 6.8 ± 1.4 × 10 3 of the carotid intraplaque mast cells, and 71 ± 27 × 10 3 of the femoral intraplaque mast cells were stained positive for tryptase ( Figure 4B). In Figure 4C, a flow cytometry plot of an atherosclerotic plaque with a low number of tryptase + mast cells (left panel) and one with a very high number of tryptase + mast cells (right panel) are shown. As not all intraplaque mast cells seem to stain positive for tryptase, these data indicate that one may underestimate the number of mast cells using immunohistochemistry. In addition, the amount of tryptase + mast cells in the plaque can apparently differ between patients, which may be caused by recent degranulation of the intraplaque mast cells. However, this remains to be further investigated in larger patient populations. As tryptase is the preferred marker to determine the presence of mast cells in tissue by means of immunohistochemistry ( Figure 4A), we also quantified the CD117 + FcεRI + mast cell populations that contained tryptase in a smaller sample of the same patients using flow cytometry (n = 7). We determined that 6.8 ± 1.4 × 10 3 of the carotid intraplaque mast cells, and 71 ± 27 × 10 3 of the femoral intraplaque mast cells were stained positive for tryptase ( Figure 4B). In Figure 4C, a flow cytometry plot of an atherosclerotic plaque with a low number of tryptase + mast cells (left panel) and one with a very high number of tryptase + mast cells (right panel) are shown. As not all intraplaque mast cells seem to stain positive for tryptase, these data indicate that one may underestimate the number of mast cells using immunohistochemistry. In addition, the amount of tryptase + mast cells in the plaque can apparently differ between patients, which may be caused by recent degranulation of the intraplaque mast cells. However, this remains to be further investigated in larger patient populations.

Discussion
In this study, we provide a novel flow cytometry-based approach to identify and characterize the mast cells in human atherosclerotic plaques. We classified ~1% of the total CD45 + leukocyte population obtained from the plaque tissue as mast cells, based on the CD117 and FcεRI expression. Previous data have established that these mast cells, although present in relatively low numbers in the plaque, can have a severe impact on plaque stability and on the risk of future cardiovascular events [15]. The majority of the mast cells present in arteries with advanced atherosclerosis are activated through IgE-binding, while a smaller fraction can undergo non-IgE-dependent activation. While IgE levels in the circulation [17] have been linked to increased incidence of acute cardiovascular events, and IgE fragments have been reported inside human atheromatic tissue [16], up until now, it was not clear to what extent this pathway affects mast cell activation in the area. Our data confirm that most mast cells present in the atherosclerotic plaques are activated [5,10], as it has been shown previously, specifically, for the shoulder region. We show here that the main activating mechanism is through classical antigen-sensitized IgE binding on their surface Fcεreceptors. In addition, we show that a small proportion of cells bind IgE without undergoing activation. The circulating IgE can thus bind on the surface of the intraplaque mast cells and sensitize them prior to antigen binding. The exact antigenic fragment that may cause intraplaque mast cell activation has not yet been identified, but the binding of lipid-specific antigenic fragments may be a possible mechanism [20]. Furthermore, the detection of IgE fragments inside the plaque tissue confirms that these fragments can surpass the endothelial wall and accumulate in the plaque area, which may explain why circulating IgE levels correlate with end-stage cardiovascular events

Discussion
In this study, we provide a novel flow cytometry-based approach to identify and characterize the mast cells in human atherosclerotic plaques. We classified~1% of the total CD45 + leukocyte population obtained from the plaque tissue as mast cells, based on the CD117 and FcεRI expression. Previous data have established that these mast cells, although present in relatively low numbers in the plaque, can have a severe impact on plaque stability and on the risk of future cardiovascular events [15]. The majority of the mast cells present in arteries with advanced atherosclerosis are activated through IgE-binding, while a smaller fraction can undergo non-IgE-dependent activation. While IgE levels in the circulation [17] have been linked to increased incidence of acute cardiovascular events, and IgE fragments have been reported inside human atheromatic tissue [16], up until now, it was not clear to what extent this pathway affects mast cell activation in the area. Our data confirm that most mast cells present in the atherosclerotic plaques are activated [5,10], as it has been shown previously, specifically, for the shoulder region. We show here that the main activating mechanism is through classical antigen-sensitized IgE binding on their surface Fcε-receptors. In addition, we show that a small proportion of cells bind IgE without undergoing activation. The circulating IgE can thus bind on the surface of the intraplaque mast cells and sensitize them prior to antigen binding. The exact antigenic fragment that may cause intraplaque mast cell activation has not yet been identified, but the binding of lipid-specific antigenic fragments may be a possible mechanism [20]. Furthermore, the detection of IgE fragments inside the plaque tissue confirms that these fragments can surpass the endothelial wall and accumulate in the plaque area, which may explain why circulating IgE levels correlate with end-stage cardiovascular events like atherothrombosis [29] and myocardial infarction [30]. Therefore, it is reasonable to acknowledge IgE as an important risk factor in cardiovascular episodes, even though it still remains to be elucidated whether it is a causative element [31]. In addition, our data raise an interesting question regarding patients who suffer from other syndromes with increased circulating IgE levels. The development of atherosclerosis is a chronic process, which spans from the formation of a fatty streak, during an individual's teenage years, and may result in an unfortunate acute event of an end-stage plaque rupture and vessel occlusion [32]. In the course of those years, a fraction of humans may be diagnosed with allergic inflammatory conditions, associating with high levels of circulating IgE [33]. This may mean that IgE has an increased chance to migrate through the endothelium and bind the intraplaque mast cells, raising the likelihood for a future cardiovascular event. In fact, there is compiling evidence that allergic asthma and atherosclerosis are linked [34], and mast cells are seemingly paramount in this [33]. In addition, patients with a genetic condition called hyper-IgE syndrome have recently been demonstrated to show signs of subclinical coronary atherosclerosis [35], and patients with systemic mastocytosis were recently described with a higher prevalence of cardiovascular disease events as compared to controls [36]. Interestingly, our group has demonstrated that the mast cell stabilizer cromolyn acts in a protective manner in atherosclerosis experimental studies [37]. Therefore, a mast cell stabilization approach may be an interesting preventive strategy in individuals who show high circulating IgE levels, but who have not yet been diagnosed with cardiovascular disease. In addition, we identified a group of mast cells that are activated without IgE binding, which suggests that this activation pathway is not the only way by which mast cells are activated in human atherosclerosis. We have previously established, in experimental atherosclerosis models, that mast cells in the vessel wall can be activated via neuropeptides such as Substance P [22] or Neuropeptide Y [38], or via complement components such as C5a [21]. These factors have also been shown to be present in human lesions [21,22]. It however remains to be established which of its receptors may induce mast cell activation in human atherosclerosis. The technical approach described in this study may be able to provide answers to these questions, by, for example, analyzing the expression of specific activating receptors on the intraplaque mast cells. It is still unclear what this implies in terms of mediator release. It would be interesting to further characterize the intracellular content of these cells to examine whether different activation pathways lead to the release of different proteases and cytokines, and how this may possibly affect the surrounding environment. Using this technique on a larger patient cohort may also reveal whether mast cell activation is affected by gender. A larger patient population will also allow for determining whether mast cell numbers or activation status relate to specific atherosclerotic plaque characteristics.
Our data provide an additional important point of attention when using tryptase-based immunohistochemistry to identify mast cells in tissue. Our flow cytometry data provide evidence that not all mast cells stain positive for tryptase, which can be explained by the fact that the recently degranulated mast cells may have released their tryptase content. One should therefore keep in mind that when using tryptase-immunohistochemistry only, the number of (activated) mast cells in the tissue may be an underestimation, and one may opt for a second quantification method, for example, a flow cytometry-based method using multiple markers.
In conclusion, in this study, we made use of flow cytometry to characterize mast cells in the advanced atherosclerotic plaques of human subjects. We confirm that the major pathway for the activation of mast cells inside the plaque tissue is IgE-mediated, and that the intraplaque mast cells are highly activated. This is particularly important, as it suggests that already available modes of therapy like mast cell stabilization agents [39] or anti-IgE treatment [40] may prove beneficial in patients suffering from atherosclerosis. We expect that larger scale patient studies and the analysis of more characterization markers, for instance through mass cytometry, will reveal new pathways via which mast cells may act in atherosclerosis, opening new ways to intervene in cardiovascular disease.