Myeloid-Specific Deletion of Peptidylarginine Deiminase 4 Mitigates Atherosclerosis

Abstract

Increasing evidence suggests that neutrophil extracellular traps (NETs) may play a role in promoting atherosclerotic plaque lesions in humans and in murine models. The exact pathways involved in NET-driven atherogenesis remain to be systematically characterized. To assess the extent to which myeloid-specific peptidylarginine deiminase 4 (PAD4) and PAD4-dependent NET formation contribute to atherosclerosis, mice with myeloid-specific deletion of PAD4 were generated and backcrossed to Apoe^-/- mice. The kinetics of atherosclerosis development were determined. NETs, but not macrophage extracellular traps, were present in atherosclerotic lesions as early as 3 weeks after initiating high-fat chow. The presence of NETs was associated with the development of atherosclerosis and with inflammatory responses in the aorta. Specific deletion of PAD4 in the myeloid lineage significantly reduced atherosclerosis burden in association with diminished NET formation and reduced inflammatory responses in the aorta. NETs stimulated macrophages to synthesize inflammatory mediators, including IL-1β, CCL2, CXCL1, and CXCL2. Our data support the notion that NETs promote atherosclerosis and that the use of specific PAD4 inhibitors may have therapeutic benefits in this potentially devastating condition.

Keywords: atherosclerosis; inflammation; macrophages; neutrophil extracellular traps; peptidylarginine deiminase 4.

Figure 1

Neutrophil extracellular traps (NETs), but not macrophage extracellular traps (METs) are present in atherosclerotic lesions and are associated with the development of atherosclerosis. (A) WT mice were fed with high-fat chow (HFC) for indicated weeks, and aortic root sections were stained for NETs markers and observed by confocal immunofluorescence microscopy. Blue: DAPI, green: MPO, red: Ly-6G, and cyan: Cit-H3. Data are representative of four mice in each time point. NETs are indicated by white arrows. (B) WT mice were fed with HFC for 5 weeks, and aortic root sections were stained for METs markers (blue: DAPI, green: MPO, red: F4/80, and cyan: Cit-H3) and observed by confocal immunofluorescence microscopy. Lower panel represents higher magnification of the NETs area (white square) in the upper panel. Macrophages are indicated by white arrows. (C) Quantification of NETs formation from (A) (n = 4 per time points). Results represent the mean ± SEM. (D) Higher magnification of (B) for characterizing macrophages. Right panel represents overlay channels of DAPI (blue) and F4/80 (red) of the NETs area (white square) of left panel. Macrophages are indicated by white arrows.

Figure 2

Myeloid peptidylarginine deiminase 4 (PAD4)-deficient neutrophils display impaired ability to generate neutrophil extracellular traps (NETs). (A) CD45⁺CD11b⁺Ly6G⁺ peritoneal neutrophils from WT and PAD4 KO mice sorted and lysed for western blot for PAD4. (B) CD45⁺CD11b⁺Ly6G⁺ peritoneal neutrophils from WT and PAD4 KO mice were stimulated with A23187 (10 µM) for 5 h or left unstimulated (UN). (B) Quantification of NETs formation (n = 3). (C) Representative immunofluorescence microscopy depicts NETs. DAPI: blue; green: neutrophil markers (CD45⁺CD11b⁺Ly6G⁺); red: Cit-H3. Bar: 100 µm. Lower panels are the 10× magnification of upper panels. Data are representative of three independent experiments. While neutrophils from WT mice display prominent NET formation with A23187, neutrophils from PAD4 KO mice do not form NETs ****p <0.001.

Figure 3

Lack of peptidylarginine deiminase 4 (PAD4) in myeloid lineage reduces atherosclerosis burden in association with diminished neutrophil extracellular trap (NET) formation in the artery. WT and PAD4 KO mice were fed high-fat chow (HFC) for 10 weeks. (A) Representative images of aortic root sections from WT and PAD4 KO mice stained with Oil Red O (red) and hematoxylin. n = 13–14/group in two independent experiments. (B) Quantitation of plaque area relative to the aortic lumen area of aortic root sections from WT and PAD4 KO mice placed on HFC for 10 weeks. (C) Representative images of en face preparations of intact aortas from WT and PAD4 KO mice placed on HFC for 10 weeks and stained with Sudan IV (Red) (n = 8–9/group in two independent experiments). (D) Quantitation of plaque area from the en face preparations of intact aortas from WT and PAD4 KO placed on HFC for 10 weeks. (E) Representative confocal immunofluorescence microscopy images of aortic root sections from WT and PAD4 KO mice placed on HFC for 10 weeks and stained for DNA (DAPI, blue), MPO (green), Cit-H3 (red), and Ly-6G (white). Areas of netting neutrophils are observed in the WT but not in the PAD4 KO mice. White arrows indicate NETs area. Data are representative of six mice in each group in two independent experiments. *p < 0.05, **p < 0.01.

Figure 4

Lack of peptidylarginine deiminase 4 (PAD4) in myeloid lineage reduces proinflammatory responses in the aorta. (A) mRNA levels of IL-1β, TNF-α, CCL2, and IL-17A in the aortas from WT and PAD4 KO mice placed on high-fat chow (HFC) for 10 weeks. mRNA levels were normalized to GAPDH utilizing delta delta CT methods, and expressed relative to levels measured in one of the WT mice Results represent the mean ± SEM (n = 16–22/group from four independent experiments). Aortas from WT and PAD4 KO mice placed on HFC for 10 weeks were enzymatically digested, and total live aortic leukocytes (B), macrophages (C), neutrophils (D), IL-17A-producing aortic αβT cells (E), and IL-17A-producing aortic γδ T cells (F) were quantified by fluorescence-activated cell sorting. Results represent the mean ± SEM (n = 10–11/group from two independent experiments). *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 5

Lack of peptidylarginine deiminase 4 (PAD4) in myeloid lineage did not affect peripheral immune responses. WT and PAD4 KO mice were fed high-fat chow (HFC) for 10 weeks. (A) Percentages of CD4⁺ cells, CD8⁺ T cells, B cells, neutrophils, dendritic cells, proinflammatory monocytes (n = 6–7/group), macrophages (n = 6–7/group), Th1 cells, Th17 cells, and Tregs from spleens of WT and PAD4 KO mice (n = 13–14/group from two independent experiments) p = NS. (B) Percentages of CD4⁺ cells, CD8⁺ T cells, B cells, Th1 cells, Th17 cells, and Tregs from lymph nodes (LN) of WT and PAD4 KO mice (n = 9–10/group from two independent experiments), p = NS.

Figure 6

Neutrophil extracellular traps (NETs) present in atherosclerotic lesions stimulate inflammatory responses in arterial macrophages. (A) Bone marrow (BM)-derived neutrophils were stimulated in the absence (UN) or presence (A23187) of A23187 for 4 h. Half the UN-NETs or A23187-NETs were digested by deoxyribonuclease (DNase) I. NETs were quantified by measuring Cit-H3-DNA complexes on ELISA. (B) BM-derived macrophages were stimulated with UN-NETs (BMN-UN), UN-NETs treated with DNase I (BMN-UN-DNase I), A23187-NETs (BMN-A23), or A23187-NETs treated with DNase I (BMN-A23-DNase I) for 4 h. Gene expression levels of IL-1β, CCL2, CXCL1, and CXCL2 were determined. mRNA levels were normalized to GAPDH and expressed relative to levels measured in one of the BMN-UN conditions (C). WT and peptidylarginine deiminase 4 (PAD4) KO mice were fed high-fat chow (HFC) for 10 weeks, and aortic root sections were stained for indicated markers and observed by confocal immunofluorescence microscopy. Lower panel represents enlarged area of the white squares in upper panels. Blue: DAPI, green: F4/80, red: IL-1β, and magenta: Cit-H3. Data are representative of four mice in two independent experiments. (D) WT and PAD4 KO mice were fed HFC for 10 weeks, and aortic root sections were stained for indicated markers and observed by confocal immunofluorescence microscopy. Lower panel represents enlarged area of the white squares in upper panels. Blue: DAPI, green: F4/80, red: CCL2, and magenta: Cit-H3. Data are representative of four mice in two independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 7

Deoxyribonuclease (DNase) I treatment abolished neutrophil extracellular traps (NETs) formation and ameliorated atherosclerotic burden. WT and peptidylarginine deiminase 4 (PAD4) KO mice were fed on high-fat chow (HFC) for 6 weeks, starting at 3-week HFC, 400 U of DNase I or vehicle control (PBS) was intravenously administered three times weekly until the end of experiments. (A) Representative confocal immunofluorescence microscopy images of aortic root sections stained for DAPI (blue), MPO (green), Ly-6G (red), and Cit-H3 (cyan). Data are representative of five mice in each group. (B) Quantification of NETs from (A) (n = 5/group). (C) Representative images of aortic root sections stained for lipid (Oil Red O, red) and hematoxylin (n = 5/group). (D) mRNA levels of IL-1β, TNF-α, CCL2, CXCL1, and CXCL2 in the aorta from WT and PAD4 KO mice placed on HFC for 6 weeks and administered with DNase I or vehicle control (PBS). mRNA levels were normalized to the GAPDH and expressed relative to levels measured in one of the vehicle control-treated WT mice (n = 5/group). *p < 0.05.