Autoantigen-Harboring Apoptotic Cells Hijack the Coinhibitory Pathway of T Cell Activation

Apoptosis is an important physiological process in development and disease. Apoptotic cells (ACs) are a major source of self-antigens, but ACs usually evade immune responses. The mechanism by which ACs repress T cell adaptive immune responses is poorly understood. T cell activation is finely regulated by a balance of costimulatory signaling (mediated by the costimulatory receptor CD28 on T cells) and coinhibitory signaling (mediated by the coinhibitory ligands CD80 and PD-L1 and -2 on Antigen-Presenting Cells). Here, we found that ACs specifically upregulated the coinhibitory ligand CD80 on macrophages. Conversely, ACs did not exhibit a robust regulation of the other coinhibitory ligands on macrophages or the costimulatory receptor CD28 on T cells. We show that the robust positive regulation of CD80 by ACs requires phagocytosis of ACs by macrophages. We also demonstrate that CD80 modulation by dead cells is a specific effect of ACs, but not necrotic cells (which stimulate immune responses). These results indicate that ACs modulate the coinhibitory pathway of T cell activation via CD80, and suggest a role for CD80 in suppressing T cell responses by ACs. Understanding a mechanism of regulating adaptive immune responses to ACs, which harbor an abundance of self-antigens, may advance our understanding of mechanisms of regulating autoimmunity and facilitate future therapy development for autoimmune disorders.


Results
Effect of ACs on expression of genes that suppress T cell functions. ACs are known to suppress adaptive immune responses and T cell activation. In order to understand how ACs achieve such an effect, we first hypothesized that ACs may be suppressing T cell activation and proliferation by a direct action on regulating genes required for T cell survival, proliferation or activation. To probe for such a direct effect on T cell genes, we arbitrarily tested two important genes that regulate T cell viability or activation, Vascular Endothelial Growth Factor-A (VEGF-A) and Arginase 2 (Arg2). VEGF, aside from its important roles in angiogenesis and tumor growth, has recently drawn much attention for its role in suppressing T cell functions and adaptive immune responses; and VEGF inhibition is currently being investigated as a therapeutic intervention to enhance anti-tumor immunity [23][24][25] . VEGF secreted from T cells themselves, or from dendritic cells, can act in an autocrine or paracrine manner, respectively, to activate VEGF tyrosine-kinase receptors (VEGFR-1 and -2) on T cells which leads to inhibition of T cell proliferation and inhibition of T cell receptor (TCR)-mediated T cell activation 26 . VEGF and its receptors, VEGFR-1 and -2, are inducibly upregulated upon T cell activation by anti-CD3 or anti-CD28 27,28 . VEGF can also inhibit dendritic cell functions [29][30][31] .
Since ACs regulate the expression of genes associated with innate immune responses, such as TNF (Tumor necrosis factor)-α, at the transcriptional level 4,5 , we first investigated the effect of ACs on transcriptional regulation of genes associated with adaptive immune responses, using quantitative real-time PCR (qRT-PCR) assays. Thus we incubated Jurkat 77 cells (a T cell line derived from human leukemia) with apoptotic S49 cells. After 3 hours we assessed the AC-induced changes in expression of VEGF-A using qRT-PCR. We found that ACs did not enhance the levels of the immunosuppressive VEGF-A, but actually slightly reduced them (Fig. 1a). Thus, we concluded that ACs do not suppress T cell activation via regulating VEGF-A.
We then tested whether ACs regulate Arg2. Arginase dramatically suppresses T cell proliferation and cytokine synthesis, by depleting arginine in the T cell environment, which leads to CD3ζ chain downregulation without affecting T cell viability per se [32][33][34][35][36] . Moreover, downstream products of the arginine metabolism might generate polyamines and toxic polycationic byproducts that have antiinflammatory properties 37 or induce apoptosis 38 . Thus, Jurkat cells were exposed to apoptotic S49 cells for 3 hours, and Arg2 expression was tested by qRT-PCR. We found that ACs, similarly to VEGF-A, induced a slight transcriptional downregulation of Arg2, a gene that normally suppresses T cell functions (Fig. 1b). To confirm changes in gene expression levels at the protein level, we performed immunoblotting of T cells exposed to different concentrations of ACs. We found that exposure of T cells to ACs significantly reduced protein levels of VEGF-A and Arg2 (Fig. 1c-e). To test whether ACs modulate VEGF-A or Arg2 production in APCs, which can either have a direct suppressive effect on APC functions or on T cell functions, we measured changes in VEGF-A or Arg2 in RAW264.7 macrophages exposed to ACs. Similarly to SCIentIFIC REPORTS | (2018) 8:10533 | DOI:10.1038/s41598-018-28901-0 T cells, macrophages also showed reduction in VEGF-A levels ( Fig. 2a,b) and Arg2 levels (Fig. 2c,d) upon exposure to ACs. Together, we concluded that ACs are unlikely to suppress adaptive immune responses via acting on general genes that regulate survival, or health of T cells or APCs.

Effect of ACs on the costimulatory receptor CD28.
Then we reasoned about other mechanisms that regulate T cell functions and activation, and hypothesized that ACs may regulate the costimulatory/coinhibitory signaling that modulates activation of T cells upon encountering APCs. Thus we exposed Jurkat T cells with ACs and measured CD28 levels on the surface of the cells cytofluorimetrically. We could not detect a significant change in CD28 levels on T cells after 6 hours of exposure to ACs (Fig. 3a,b). Even with extended time points (24 hours) and high ratios of AC:macrophage, we could only detect a slight decrease in CD28 protein levels that was barely statistically significant, but we did not observe a robust effect of ACs on CD28 levels (Fig. 3c). These results indicated that ACs do not exhibit robust regulation of CD28.
Effect of ACs on the coinhibitory ligands on macrophages, PD-L1, PD-L2 and CD80. Given the above result showing no robust regulation of CD28 by ACs and the fact that the effect of ACs on suppressing T cell activation was dominant even in presence of LPS which induces the CD86/costimulatory pathway 39 , we reasoned that ACs may be actively regulating the coinhibitory pathway in order to suppress T cell activation (as coinhibition overrides costimulation). Furthermore, considering that ACs are phagocytosed by macrophages first and possibly regulate T cell activation through macrophages, we decided to investigate whether ACs may regulate the macrophages' coinhibitory signaling.
Dendritic cells ingesting tumor ACs which activated T cells showed upregulation of the costimulatory ligand CD86 secondary to AC phagocytosis 40,41 . Thus, we found it plausible to propose that T cell-suppressing ACs may upregulate the coinhibitory ligands (CD80, PD-L1 and/or PD-L2). Consistent with our proposal is the fact that these coinhibitory ligand signals are indispensable for suppression of T cell activation 42 . Moreover, lamina propria Jurkat 77 human T cells were exposed to ACs (mouse S49 cells) at a ratio of 10 ACs per T cell for 0 or 3 hours. RNA was then extracted from the T cells and qRT-PCR performed for the indicated genes. Shown are relative gene expression levels for Arg2 or VEGF-A, normalized to GAPDH, at 0 or 3 hours post-exposure to the ACs. (c) Jurkat E6-1 human T cells were incubated with apoptotic HeLa cells at the indicated ratios for 6 hours and then immunoblotted for VEGF-A or Arg2; α-tubulin used as a loading control. (d,e) Quantification of relative protein levels from multiple independent experiments as in (c). *p < 0.05, **p < 0.01 (Student's t-test).
We first considered the possibility that ACs regulate PD-L1 and PD-L2. PD-L1 and -2 are coinhibitory ligands that bind to and activate the coinhibitory receptor PD-1 on T cells in a high affinity binding similar to CD80/CTLA-4 binding 17 , and are both important negative regulators of T cell functions. For example, PD-L1 is upregulated in exhausted CD8+ T cell of lymphocytic choriomeningitis virus (LCMV)-chronically infected mice, and blockade of PD-L1 signaling with anti-PD-L1 antibodies restored CD8+ T cell function and triggered viral clearance 44,45 . PD-L2 −/− APCs were also more effective in inducing T cell responses than wild-type APCs, and PD-L2 −/− mice showed increased T cell activity over wild-type mice 46 . Double knockout of both PD-L1 and PD-L2, which completely blocks coinhibitory signaling through PD-1, led to stronger T cell activation than single knockout of either gene 47 . This suggested to us that the PD-L1/2 coinhibitory pathway may also be taken advantage of by other immunosuppressive stimuli such as ACs.
Since PD-L2 is inducibly expressed on dendritic cells and macrophages, we tested whether ACs regulate PD-L1/2 mRNA levels. Thus we incubated RAW264.7 macrophages with apoptotic Jurkat 77 cells, and assessed changes in mRNA expression levels of PD-L1 and -2. We could not detect significant changes in PD-L mRNA levels ( Fig. 4a-c) even with increasing the AC:macrophage ratio (Fig. 4b,c). Therefore we concluded that ACs do not significantly regulate PD-L1 and -2.
We then tested the effect of ACs on CD80 expression. RAW264.7 macrophages were exposed to apoptotic Jurkat 77 cells. After 3 hours, we measured changes in CD80 expression levels. Interestingly, we found that ACs significantly enhanced CD80 mRNA levels (Fig. 4d). We then attempted to determine the minimum ratio of AC:macrophage sufficient to produce such an effect. We found that even a ratio of one AC per macrophage was sufficient to significantly induce CD80 expression (Fig. 4e). Having seen an effect of ACs on CD80 transcriptional levels, we wanted to confirm the effect of ACs on CD80 protein expression levels on the surface of macrophages. Thus, we incubated RAW264.7 macrophages with ACs, or with a positive control known to induce CD80 expression, lipopolysaccharide (LPS) 48,49 , or the combination of ACs and LPS, and then performed a cytofluorimetric assay for CD80. We found that exposure of macrophages to ACs for 16 hours led to significant upregulation of CD80 expression on macrophages ( Fig. 5a-e). Combining ACs with LPS led to an additive effect on CD80 expression ( Fig. 5a-e).
To investigate the in-vivo relevance of this result, we used primary murine macrophages as model APCs. Thus, primary macrophages were stimulated by exposure to apoptotic cells or a positive control (LPS + IFNγ (interferon γ) combination). Similarly to RAW264.7 cells, primary macrophages also showed a substantial effect of ACs on upregulating CD80 levels on macrophages ( Fig. 5f-h). Taken together, these data confirm that ACs induce CD80 expression levels on macrophages.

In-depth characterization of the effect of ACs on CD80. Effect of ACs on CD80 expression on mac-
rophages is specific to ACs, but not necrotic cells (NCs). Next, we wanted to investigate whether the effect of ACs on CD80 expression is an effect specific to ACs or a nonspecific effect shared by all dead corpses (apoptotic or necrotic). Thus we incubated RAW264.7 macrophages with LPS, dead cells (either apoptotic or necrotic), or a combination of LPS plus dead cells. We then measured macrophages' CD80 surface expression using cytofluorimetry. While ACs dramatically enhanced CD80 levels, NCs caused no increase in CD80 expression levels ( Fig. 6a-g). Thus we concluded that the observed upregulation of CD80 expression on macrophages upon encountering ACs is a specific effect of ACs, suggesting that CD80 upregulation is important for suppressing T cell activation and adaptive immune responses, which is a specific response to ACs not shown by NCs that induce immune responses. Time-course of CD80 upregulation by ACs. To further characterize the effect of ACs on CD80, we performed a time-course determination of CD80 expression after encountering ACs. RAW264.7 macrophages were incubated with ACs for various durations. At each time point, CD80 expression was assessed using cytofluorimetry. The positive control, LPS, showed no significant effect on CD80 levels after 5 or 10 hours, but showed a modest effect after 20 hours (Fig. 7a-d). Conversely, ACs showed a significant effect on CD80 upregulation as early as 5 hours after incubation with macrophages. Longer durations of exposure to ACs gave further increase in CD80 expression, which always showed an additive effect to LPS when macrophages were exposed to a combination of ACs and LPS (Fig. 7a-d).

Figure 5.
ACs strongly upregulate expression of CD80 on RAW264.7 macrophages and primary murine macrophages. (a-e) RAW264.7 murine macrophages were exposed to ACs (human Jurkat 77 cells) at a ratio of 10 ACs per macrophage, for 16 hours, and CD80 expression was analyzed using flow cytometry. 10 6 RAW264.7 cells were plated per well of a 6-well plate 24 hours before ACs or LPS (500 ng/ml) addition ("Unstim" denotes unstimulated cells, exposed to no treatment). 10 7 ACs (Jurkat77 cells induced to apoptose by 200 ng/mL Actinomycin D treatment for ~12 hours) were added per well. The macrophages were harvested after 16 hours, stained with anti-CD80-FITC and analyzed with flow cytometry. (e) The experiment was repeated five independent times, and average CD80 levels were plotted. (f-h) Primary murine macrophages were exposed to no treatment, LPS + IFNγ or ACs as in (a-e) and were processed similarly for flow cytometric analysis of CD80 expression as in (a-e). *p < 0.05, **p < 0.01, ***p < 0.001 (Student's t-test). Concentration and cell-type dependency of the effect of ACs on CD80. To further understand the effect of ACs on CD80, we asked whether the concentration of ACs could influence the magnitude or robustness of CD80 upregulation in macrophages. We exposed RAW264.7 macrophages to varying concentrations of ACs, and measured CD80 expression on macrophages using flow cytometry. We found that gradually increasing concentrations of ACs (1, 5 or 10 ACs per macrophage) all upregulated CD80 expression; and the increasing AC concentrations showed subtle, but not statistically significant, differences in CD80 upregulation (Fig. 8a-e). These data indicate that the effect of ACs on CD80 upregulation is independent of the AC concentration. This suggests that any mild production of ACs has a significant effect on upregulating the coinhibitory ligand CD80, and thus possibly suppressing T cell responses and evading autoimmune responses to self-antigens carried on the minimal amount of ACs.
Furthermore, we wanted to investigate if the effect of ACs is dependent on the cell-type of ACs. We thus incubated RAW264.7 macrophages with two different AC types, Jurkat 77 or HeLa cells, and assessed CD80 expression via cytofluorimetry. We found that all AC types tested induced similar effects on upregulating CD80 levels on macrophages (Fig. 8f). This data suggests that upregulation of CD80 by ACs is independent of the cell type of ACs, suggesting that upregulation of CD80 could be a universal mechanism used by all AC types to evade adaptive immune recognition and possibly autoimmune responses to self-antigens carried on/in the ACs. . ACs (and not NCs) specifically upregulate expression of CD80 on macrophages. (a-g) RAW264.7 murine macrophages were exposed to ACs (human Jurkat 77 cells) or NCs at a ratio of 10 ACs per macrophage, for 16 hours, and CD80 expression was analyzed using flow cytometry. 10 6 RAW264.7 cells were plated per well of a 6-well plate 24 hours before ACs or NCs or LPS (500 ng/ml) addition. (g) The experiment was repeated five independent times, and average CD80 levels were plotted. *p < 0.05, **p < 0.01 (Student's t-test). Effect of AC recognition, or phagocytosis, by macrophages on CD80 expression. To fully characterize the regulation of CD80 by ACs, we finally desired to investigate whether upregulation of CD80 required phagocytosis of the ACs by the macrophages or whether only recognition of the ACs by macrophages is sufficient to induce CD80 upregulation. Thus, we tested the effect of ACs on macrophages' CD80 in presence or absence of the actin polymerization inhibitor, cytochalasin D, which significantly blocks phagocytosis 50,51 . Interestingly, while recognition only of ACs (under blocking of phagocytosis) induced a significant but mild CD80 upregulation, phagocytosis of ACs by macrophages induced a dramatic, much stronger, upregulation of CD80 on macrophages (Fig. 9a-e). This is consistent with the findings in dendritic cells, where phagocytosis of ACs by DC decreased antigen-specific (a-d) RAW264.7 murine macrophages were exposed to ACs (human Jurkat 77 cells) at a ratio of 10 ACs per macrophage, for 5, 10 or 20 hours, and CD80 expression was analyzed using flow cytometry. 10 6 RAW264.7 cells were plated per well of a 6-well plate 24 hours before ACs or LPS (500 ng/ml) addition. (d) Time-course plot of CD80 expression. *p < 0.05, **p < 0.01, ***p < 0.001 (Student's t-test).
SCIentIFIC REPORTS | (2018) 8:10533 | DOI:10.1038/s41598-018-28901-0 activation of T cell proliferation 8,10 . The requirement of phagocytosis seems to be a specific phenomenon for AC-mediated regulation of adaptive immunity, as AC-mediated regulation of innate immunity and cytokine secretion could be recapitulated by recognition only, but not necessarily phagocytosis, of ACs 5,9 .

Discussion
In this study we wanted to understand how ACs suppress T cell responses, focusing mostly on the effect of ACs on macrophages. In most cases ACs were shown to dampen T cell responses 14,[52][53][54][55] . However, sometimes tumor ACs engulfed by dendritic cells 40,41 or macrophages 10 could also elicit tumor-specific T cell responses and antitumor immunity. To account for the occasional paradoxical activation of T cells by ACs, two models could be proposed. Firstly, secondary signals form the microenvironment in which AC phagocytosis takes place may work in concert with ACs to favor a certain response. Microenvironments that also contain immunostimulatory stimuli such as  57 . However, we focused on uncovering the effects triggered by ACs per se in a 'neutral' microenvironment, to understand how AC clearance in physiological contexts mechanistically regulates immune responses to ACs and the AC-associated self-antigens. Secondly, different APC types (dendritic cells or macrophages) in distinct locales possess differential locale-specific activities (activation or inhibition of T cells). Co-ordination of these activities may give rise to a dominant context-specific response. For example, in the intestinal lamina propria there are two types of APCs presenting commensal microbe and dietary antigens to T cells: dendritic cells induce the differentiation of the effector Th17 T cells, whereas macrophages induce the differentiation of Tregs 43 . These Tregs showed increased production of anti-inflammatory cytokines, and decreased production of immunostimulatory cytokines, and were much less proliferative (i.e. became anergic) upon restimulation with the antigen. Since macrophages are much more abundant than dendritic cells in the lamina propria, T cell tolerance is the predominant response in this context. Indeed, that fact further encouraged us to investigate the AC-mediated regulation of T cell activation by macrophages, leading to comprehensive understanding of the immune regulation performed by ACs. Figure 9. CD80 upregulation by ACs under conditions of functional or blocked phagocytosis. (a-e) RAW264.7 murine macrophages were exposed to ACs (human Jurkat 77 cells) at a ratio of 10 ACs per macrophage, for 16 hours in presence of vehicle, "Veh" (a,b) or cytochalasin D, "CytoD" (c,d), and CD80 expression was analyzed using flow cytometry. The cells were treated with vehicle or 2 μM cytochalasin D starting at 1 hour before addition of the ACs and continuing throughout the incubation period of the macrophages with the ACs. (e) Quantification of CD80 relative levels (expressed as ΔMFI (change in mean fluorescence intensity) relative to the Unstimulated control "Unstim") of the CD80 histogram) upon exposure to ACs in presence of vehicle or cytochalasin D. *p < 0.05, ***p < 0.001 (Student's t-test).
SCIentIFIC REPORTS | (2018) 8:10533 | DOI:10.1038/s41598-018-28901-0 Moreover, the idea that ACs may suppress T cell activation through macrophages is plausible, given the fact that macrophage-mediated immunosuppression can reinforce dendritic cell-mediated immunosuppression or predominate the dendritic cell-mediated immunostimulation occasionally observed, macrophages being more abundant than dendritic cells, leading to the overall effect of T cell tolerization to ACs and their associated self-antigens.
Our attempts at unraveling the mechanism by which ACs suppress T cell responses through macrophages demonstrated that ACs take control of the coinhibitory pathway, by inducing a substantial upregulation of the coinhibitory ligand CD80. Importantly, we observed that this effect is specific to AC, while necrotic cells (which stimulate immune responses) failed to produce such an effect on CD80 expression, suggesting that CD80 upregulation may be a potential mechanism-at least in part-used by ACs specifically to suppress T cell activation. Additional mechanisms, however, cannot be completely ruled out, but they are unlikely to play the major role played by CD80 upregulation. For example, the possibility that ACs suppress T cell responses through downregulating the costimulatory ligands such as CD86 is less likely, because CD86 levels in dendritic cells did not change significantly upon uptake of ACs 8,39 . Still, secondary mechanisms, including cytokine-mediated actions, might possibly complement the effect on enhancing coinhibition, as another layer of tightening the regulation of T cells adaptive immune responses to ACs.
We have confirmed the robust effect of ACs in regulating CD80 on macrophages, but how that regulation translates into an actual effect on T cells in vivo is an interesting topic for future investigations. Naïve T cells differentiate upon antigen recognition into effector (immunostimulating) T cell subsets such as Th1, Th2 and Th17, or into immunosuppressive Tregs. Dendritic cells ingesting ACs enhanced the development of Tregs, but suppressed the development of the effector Th17 T cells. Although conditioned medium of these dendritic cells, containing their secreted cytokines and soluble factors, could recapitulate such an effect 57 , presence of costimulatory/coinhibitory signaling that takes place during dendritic cell-T cell interaction markedly enhanced Treg proliferation 58 . That indicates that AC-mediated regulation of T cell activation is dependent on direct APC-T cell interaction. Some reports suggest that ACs modulate the maturation of dendritic cells via expression of various surface molecules recognized by T cells which regulate T cell activation and response 8,59 , further highlighting the significance of APC-T cell interaction. Another way in which ACs were thought to regulate T cell responses is through regulating the migration ability of dendritic cells. Dendritic cells migrate upon antigen ingestion, as they mature expressing molecules necessary to prime T cells, to secondary lymphoid tissues such as lymph nodes (LNs) and activate T cells 60,61 . However, such a model of AC action is unlikely, given the fact that AC-ingesting and -noningesting dendritic cells showed comparable migration to the draining LNs 40 . Whether the same is true for macrophages ingesting ACs remains to be determined in in-vivo models.
The implications of our results for diseases are abundant, as understanding the mechanisms used by ACs to regulate adaptive immunity will enlighten our understanding of mechanisms that the body uses to regulate immunity in physiological conditions, whose disruption may cause diseases. Firstly, immunosuppression by ACs may serve as a mechanism to prevent auto-immunity to self-antigens carried in ACs; failure of that mechanism might lead to development of auto-immunity. For example, Xia et al. 62 studied the effect of apoptotic β-cell infusion on β-cell antigen-specific CD4+ T cell proliferation and showed that suppression of T cell activation by ACs delayed the onset of diabetes in the autoimmune diabetes-prone (NOD) mice. Thus, our work proposing a mechanism of immunoinhibition by ACs may help facilitate the development of novel effective therapies for autoimmune disorders, such as rheumatoid arthritis, systemic lupus erythematosus and Type I diabetes mellitus. Secondly, immunosuppression exerted by ACs may decrease effectiveness of anticancer chemotherapy, as tumor chemotherapeutic treatment increasingly produces ACs that negatively regulate T cell functions and adaptive immunity. Thus, it is plausible to hypothesize that decreased effectiveness of chemotherapeutic treatments may arise, at least partly, from the progressive inhibition of T cell immune responses by ACs, the by-product of these treatments. Developing mechanisms of functional blockade of the chemotherapy-produced ACs might be a promising key to maintaining effectiveness of chemotherapeutic agents and minimizing their undesirable side effects on the immune system. In fact, some therapies targeting the coinhibitory molecules, which regulate adaptive immune responses by ACs, have been designed 63,64 and are now in clinical trials 65 .
Overall, our results demonstrate that ACs specifically regulate CD80 levels on macrophages. Inducing coinhibitory signaling through CD80-CTLA-4 binding could either override costimulatory signals, counteracting initiation of T cell activation, or enhance differentiation of immunosuppressive Tregs. Our results highlight the importance of the coinhibitory pathway in suppressing an immune response to ACs, and suggest a potential mechanism of immune regulation that may be used by the body to control reactivity to self-antigens carried by ACs and evade autoimmune responses.
Cells and cell culture. The following cell lines were used. RAW264.7 is a macrophage cell line derived from an adult BALB/c male mouse. Jurkat 77 and Jurkat E6-1 are T lymphocyte cell lines derived from a human T cell leukemia. HeLa is an epithelial cell line derived from a human cervical epithelium adenocarcinoma. S49 is a murine T lymphocyte cell line derived from lymphoma of BALB/c mouse. Cell lines were obtained from ATCC and were grown at 37 °C in a humidified, 5% (v/v) CO 2 incubators, as per standard cell-culture procedures. Primary macrophages (from CellBiologics, Cat # C57-6032TF) were isolated as previously described 66 : briefly, mouse peritoneal macrophages were induced by injecting C57BL/6 mice with sterile thioglycollate, and then collected by peritoneal lavage on day 3. Peritoneal macrophages were seeded at a density of 10 6 cells/well of a 12-well plate ~24 hours before addition of the ACs. After 12 hours, they were observed to have strongly adhered to the plate, then they were washed with PBS and then with media. 12 hours later, cells were re-washed immediately before ACs addition.
Preparation of Apoptotic Cells. Cells were induced to undergo apoptosis with 200 ng/mL Actinomycin D added in the medium ~12 hours. Apoptosis was verified by flow cytometry per standard procedures.
Preparation of Necrotic Cells. Cells were induced to undergo necrosis by incubation at 56 °C for 30 minutes (necrosis verified by loss of membrane integrity indicated by trypan blue uptake), immediately before they were added to the macrophages.