The protocols were approved by the Cedars-Sinai Institutional Review Board (IRB). Peripheral blood mononuclear cells (PBMCs) were isolated from blood collected from 10 patients with ACS within 72 h of admission to the Cedars-Sinai Cardiac Intensive Care Unit. Patients were consented under the approved IRB protocol Pro48880. Exclusions were inability to give informed consent, age less than 18 years old, active cancer treated with chemotherapy or radiation, patients taking immune-suppressive drugs, and pregnant women. PBMCs from 10 stable CAD patients were isolated from blood collected on the day of coronary angiography, consented under the approved IRB protocol Pro50839 with the same exclusions. Consented data use was limited to age, sex, LDL levels, and use/non-use of cholesterol-lowering medication. PBMCs were isolated using Ficoll density gradient centrifugation and cryo-preserved in commercially available cryogenic solution (Immunospot) in liquid nitrogen. Cryo-preserved PBMCs from self-reported controls (N = 15) were purchased from a commercial source (Immunospot).

Cryo-preserved PBMCs were thawed, rinsed in anti-aggregation solution (Immunospot), and seeded in culture plates at a density of 3 × 10⁶cells per ml of RPMI 1640 medium supplemented with 10% heat-inactivated pooled human serum and 1× antibiotic/antimycotic. Peptides (LifeTein) used for stimulation corresponded to LL-37 and the truncated cathelin domain of hCAP-18 [cat-hCAP-18 (aa 39-136)]. Cells were stimulated with one of the following: 20 μg/ml LL-37 or cat-hCAP-18 peptide, 0.5× T cell stimulation cocktail containing PMA and ionomycin (Thermo Fisher). Culture medium was added at ¹/₃ of the starting volume 48 h later to replenish the nutrients in the medium. Cells were harvested 72 h after seeding, stained for viability (LIVE/DEAD Fixable Aqua Dead Stain Kit, Thermo Fisher), and subjected to cell surface staining for flow cytometry using the following antibodies: CD3, CD4, CD8, CD45RA, CD45RO, CD62L, and CD197 (CCR7). Isotypes were used as staining control. CD4+ or CD8+ T Effector cells were gated on CD45RO+CD62L(−)CD197(−). T Effector Memory cells were CD45RO+CD45RA(−) CD62L(−)CD197(−), T Effector Memory RA+ cells were CD45RO+CD45RA(+)CD62L(−)CD197(−). Results were tabulated as Response Index using the following calculation (20):

The results are expressed as Response Index to account for inherent variations introduced by culturing cells in vitro over time, controlled for by assessing response relative to baseline cell phenotype (% no stimulation) and maximal stimulation (% cocktail stimulation) for each subject PBMC. Each data point represents one subject.

In samples with sufficient cell numbers, stimulated PBMCs cultured for 72 h were collected for RNA extraction using Trizol (Thermo Fisher) and subjected to qRT-PCR with SYBR green and a primer pair for human PDCD1. Cyclophilin A served as the reference gene. Results were expressed as fold-change relative to non-treated cells of each sample using the Ct_ΔΔ method.

Conditioned medium was collected after 72 h of LL-37 stimulation of PBMCs and IFN-γ was measured using a commercially available ELISA kit (Abcam) according to manufacturer’s instructions.

PBMCs from control samples were stimulated with LL-37 alone or in the presence of either anti-human HLA-A, B, C (clone W632) or anti-human HLA-DR (clone L243) monoclonal antibody (Biolegend; 30 μg/ml) for 72 h and stained for flow cytometry as described above.

The studies were approved by the Cedars-Sinai Institutional Animal Care and Use Committee. Male apoE−/−mice were purchased from Jackson Laboratory at 6 weeks of age and housed in a specific pathogen-free facility, kept on a 12-h day/night cycle, and had unrestricted access to water and food.

The mCRAMP peptide was purchased (AnaSpec) with >95% purity according to the manufacturer. Donor mice were subcutaneously immunized with either 100 μg mCRAMP in adjuvant [Adju-Phos (Brenntag, 12.5 μl of a 2% solution) and monophosphoryl lipid A (MPLA-SM VacciGrade, InvivoGen, 10 μg)] or adjuvant alone in a final volume of 200 μl using PBS as vehicle. One group of mice was fed normal chow and immunized at 7, 10, and 12 weeks of age. Mice were then euthanized and splenocytes isolated to assess T cell response at 13 weeks of age using flow cytometry. Another group of mice was fed normal chow and immunized at 7, 10, and 12 weeks of age. The mice were then euthanized as T cell donors at 13 weeks of age. Recipient mice were fed high fat diet consisting of 21% fat, 0.15% cholesterol (TD.88137, Envigo) starting at 7 weeks of age until euthanasia at 23 weeks of age.

After collection of donor mouse splenocytes, red blood cells were lysed using RBC lysis buffer (BioLegend), washed with PBS, and cells from the same group were pooled. Splenocytes were enriched for T cells using a commercially available mouse T cell magnetic isolation kit (ThermoFisher) according to the manufacturer’s protocol. T cells were enriched to ∼90% assessed by flow cytometry. After counting the isolated T cells, a suspension of 2 x10⁶ T cells in PBS was injected by tail vein injection into 18 week-old recipient mice fed with high fat diet for 11 weeks. High fat diet feeding in recipient mice continued for an additional 5 weeks and mice were euthanized at 23 weeks of age.

At 23 weeks of age, recipient mice were euthanized and spleens, aortas and hearts were collected. Serum was collected for cholesterol level measurement. Spleens were collected for flow cytometric staining and analysis. A small portion of the spleen was used for RNA extraction. The aortas were dissected free of connective tissue and fat, and stained with Oil Red O for en face lipid staining (19). Aortic size and plaque content were quantified by a blinded observer using image analysis software (ImagePro Plus version 4.0, Media Cybernetics Inc., Rockville, Maryland). Heart bases were imbedded in OCT (Optimum Cutting Temperature, Tissue-Tek) and frozen for cryo-sectioning. Ten-micron cryosections of the aortic sinus were collected. Lipid was stained using Oil-Red-O. Staining for macrophage was performed using CD68 antibody. Collagen area was assessed using Masson trichrome stain. Three slides of approximately 0.1-millimeter intervals for each animal were used for each stain and averaged. Image analysis was performed using ImagePro. Tissue calcification in aortic sinus plaques was confirmed by Alizarin Red-S stain.

Splenocytes were stained for viability and surface markers CD3, CD4, CD8, CD44, and CD62L. Intracellular staining for FoxP3 was performed after surface marker staining followed by fixation and permeabilization (eBioscience). For intracellular cytokine staining, cells were resuspended in 10% heat inactivated FBS-RPMI medium containing 1× Monensin (Invitrogen) and cultured at 37°C and 5% CO₂ for 4 h. After viability and surface marker staining, cell fixation and permeabilization was performed followed by intracellular staining for IL-10 and IFN-γ for flow cytometry.

For degranulation assay to measure CD8+ T cell cytolytic activity, splenocytes were incubated with 2.5 μg/ml fluorescent conjugated CD107a in RPMI for 1 h followed by the addition of Monensin and incubation for another 4 h. Cells were collected and stained for surface markers for flow cytometry.

Splenic RNA was isolated using Trizol and subjected to qRT-PCR using primer pairs for mouse IL-1β, Pdcd1, Ctla4, Wnt10b, Runx2, RANKL, and osteocalcin. GAPDH was used as reference gene. Data were analyzed using the Ct_ΔΔ method with one Adjuvant sample as calibrator. Results are expressed as fold change relative to Adjuvant.

Mouse serum levels of soluble RANKL and undercarboxylated osteocalcin, the active form of osteocalcin, were measured using commercially available ELISA kits (ThermoFisher and MyBioSource, respectively) according to manufacturers’ instructions.

Statistical analysis was performed using R software version 3.5 (R Core Team, 2018) and GraphPad Prism version 7 (GraphPad Software, La Jolla, California). Data are presented as mean ± standard deviation. For multiple group analysis, significance for normally distributed samples was tested using ANOVA followed by Holm-Sidak’s multiple comparisons test. Significance for non-normally distributed samples was tested using Kruskal-Wallis test followed by Dunn’s multiple comparisons test. Correlation was analyzed using the Spearman test with a two-tailed P-value. For two-group analysis, following normality testing, differences between groups were performed using t-test or Wilcoxon test, accordingly. A P-value < 0.05 was considered significant, but data trends were also noted.

Given the role of anti-microbial peptides as potential self-antigens in atherosclerosis, and the possible association with acute events, we tested if the cleaved fragment of hCAP-18, the cationic antimicrobial peptide LL-37, would induce differential T cell immune responses in patients with ACS. Peripheral blood mononuclear cells (PBMCs) from self-reported healthy controls (Controls), patients with stable CAD (Stable) or ACS were stimulated with LL-37 for 72 h. PBMCs stimulated with the cathelin domain of hCAP-18 (cat-hCAP-18) served as control. Baseline characteristics of the subjects are detailed in Table 1. Stimulation with LL-37 resulted in reduced CD8+ Effector T cell response, in both T Effector Memory (TEM) and T Effector Memory RA+ (TEMRA) cells in PBMCs from controls and patients with stable CAD, while PBMCs from patients with ACS were resistant to this reduction (Supplementary Figure 1 and Figures 2A–C). CD4+ Effector T cell responses were trending similar to CD8+ Effector T cells (Figures 2D–F). There was no significant difference in CD8+ Effector T cell response when cells were stimulated with cat-hCAP-18 (Figure 3). There was reduced expression of programmed cell death protein 1 (PDCD1) mRNA in PBMCs from patients with ACS stimulated with LL-37 compared to PBMCs from controls (Figure 4A), but no difference was noted between groups in PDCD1 mRNA expression when cells were stimulated with cat-hCAP-18 (Figure 4B).

The nature of the self-reactive responses to cat-hCAP-18 and LL-37 were investigated further by testing the relationship between the T cell response to both antigens in each subject. There was significant correlation in CD4+ Effector T cell response (Figure 4D) to cat-hCAP-18 and LL-37, but not in CD8+ Effector response (Figure 4C). The results suggest that there is potentially shared antigenic reactivity to cat-hCAP-18 and LL-37 in CD4+ T cell responses. IFN-γ measured in conditioned medium from PBMCs stimulated with LL-37 showed qualitative difference in ACS compared to control and stable CAD patients (Supplementary Figure 2). The results suggest that LL-37 reactive T cells may be involved in the acute event. Additionally, Effector T cell responses in control PBMCs stimulated with LL-37 was partially reversed by blocking with HLA Class-I antibody (Supplementary Figure 3) suggesting HLA Class-I mediated response.

The T cell response in ACS patients was further subdivided into patients with their first ACS event and those who had a recurrent event. The CD8+ T Effector response to LL-37 was consistent between first and recurrent ACS patients, compared to patients with Stable CAD (Supplementary Figures 4A–C). On the other hand, CD4+ T Effector response was significantly higher only in the recurrent ACS patients but not in the first ACS event compared to Stable patients (Supplementary Figure 4D). This suggests that CD8+ T cell response to LL-37 persists in immunologic memory and is a common response in both first event ACS and recurrent ACS. CD4+ T cell Effector response on the other hand seems to have evolved and is more prominent in the recurrent ACS. No significant differences were observed in cat-hCAP-18 response between first ACS event and recurrent ACS patients in CD8+ T Effector (Supplementary Figures 5A–C) and CD4+ T Effector (Supplementary Figures 5D–F) responses.

To investigate the potential role of T cells that are self-reactive to the cationic antimicrobial peptide in atherosclerosis, we used the apoE−/− mouse model of atherosclerosis and immunization with the mouse ortholog of LL-37 called mCRAMP.

To investigate the potential role of T cells reactive to the mouse cationic antimicrobial peptide mCRAMP in the male apoE−/− mouse model of atherosclerosis, we first tested if T cells were reactive to mCRAMP. ApoE−/− mice fed with normal chow were immunized with mCRAMP at 7, 10, and 12 weeks of age and euthanized 1 week later for assessment of T cell response. Mice injected with adjuvant alone served as control. There was no difference in CD8+ Effector Memory T cells (Supplementary Figure 6 and Figure 5A) but a significant increase in CD8+ Central Memory (CM) T cells (Figure 5B) in splenocytes from apoE−/− mice immunized with mCRAMP. Additionally, there was decreased CD8+FoxP3+ cells (Supplementary Figure 6 and Figure 5C) and increased cytolytic activity, assessed by CD8+CD107a+ T cells (Supplementary Figure 7 and Figure 5D), in splenocytes of mice immunized with mCRAMP compared to adjuvant. No differences were observed in CD4+ memory T cell subsets or in CD4+FoxP3+ Treg cells (Figures 5E–G).

To assess whether mCRAMP-primed T cells are functionally involved in modifying atherosclerosis, adoptive transfer of donor T cells from apoE−/− mice immunized with mCRAMP or adjuvant alone was performed on apoE−/− recipient mice that had been fed high fat diet for 11 weeks prior to transfer. The 11-week feeding with high fat diet assured that the recipient mice were already primed for atherosclerosis. Recipient mice were euthanized 5 weeks after cell transfer.

Recipients of T cells from mCRAMP immunized mice had significantly reduced aortic plaque area compared to recipients of T cells from adjuvant mice (28% reduction, P < 0.05, Figures 6A,B). There was no significant difference between groups in mean body weight (mCRAMP = 42 ± 6 gr; Adjuvant = 43 ± 6 gr) or mean serum cholesterol (mCRAMP = 1500 ± 294 mg/dL; Adjuvant = 1270 ± 366 mg/dL). Thus, atherosclerosis progression was reduced in the mCRAMP T cell recipient mice without differences in weight or serum cholesterol compared to adjuvant T cell recipient mice.

Aortic sinus plaque size, lipid content (Oil Red O staining; Figures 7A–C), macrophage (CD68 staining; Figures 7D,E), and collagen area (Masson’s trichrome staining; Figures 7F,G) were not different between the T cell recipient groups. There were also no differences in IL-1β (Figure 8A), PDCD1 (Figure 8B) or cytotoxic T-lymphocyte-associated protein 4 (Ctla4, Figure 8C) splenic mRNA expression between the T cell recipient groups.

Focal staining of hematoxylin was observed in several aortic sinus sections from recipient mice, suggesting calcification in the plaque. The presence of plaque calcification was assessed with the calcium-specific Alizarin Red S staining. Calcification in atherosclerotic plaques occurred in 56% of adjuvant T cell recipient control mice, compared to none in the mCRAMP T cell recipient mice (Figures 9A,B and Table 2; Fisher’s exact test P = 0.003).

Immune pathways associated with T cell mediated tissue calcification were then investigated further. Recipients of T cells from mCRAMP immunized mice had increased IL-10 and IFN-γ expression in splenic CD8+ T cells (Supplementary Figure 8 and Figures 9C,D), but not in CD4+ T cells (Figures 9E,F), compared to adjuvant T cell recipient control mice. There was increased expression of Wnt10b mRNA in splenocytes of mCRAMP T cell recipients, when compared to adjuvant T cell recipient control mice (Figure 10A). However, there was no difference in the expression of Runx2 mRNA (Figure 10B). Furthermore, splenocytes from mCRAMP T cell recipients had increased RANKL and osteocalcin mRNA expression when compared to adjuvant (Figures 10C,D). However, no difference between groups was noted in serum RANKL and undercarboxylated osteocalcin concentration (Figures 10E,F), which is the known active state of osteocalcin.