Slc7a11‐Mediated Cystine/Glutamate Antiport Reprograms Macrophage Polarization and Ameliorates Atherosclerosis

TL;DR

This study shows that the Slc7a11 protein helps reprogram macrophages to reduce inflammation and atherosclerosis, offering a new therapeutic approach for heart disease.

Contribution

The study reveals that Slc7a11-mediated amino acid metabolism can reprogram macrophages and serve as a novel therapeutic strategy for atherosclerosis.

Findings

Slc7a11 overexpression in macrophages reduces atherosclerotic lesions and increases plaque stability.

Slc7a11 inhibits M1 macrophage polarization and promotes M2 polarization through glutathione synthesis.

Targeting Slc7a11 with ferrostatin-1 enhances antioxidant capacity and ameliorates atherosclerosis.

Abstract

Atherosclerotic cardiovascular diseases (ASCVDs) remain the primary cause of morbidity and mortality. Macrophages are involved in the progression and regression of atherosclerosis, and macrophage amino acid metabolism is important during this process. Here, we identified that the expression of cystine/glutamate antiporter Slc7a11 was upregulated by oxidized low‐density lipoprotein, and specifically enhanced in the macrophages of atherosclerotic plaques. Macrophage‐specific Slc7a11 overexpression in ApoE null mice (ApoE–

/–
Slc7a11MOE
) attenuated atherosclerotic lesions and increased the plaque stability under a 16‐week western diet. ApoE–

/–
Slc7a11MOE
displayed unchanged blood lipids, decreased inflammatory cytokines, and increased antioxidant capacity. Mechanistically, Slc7a11‐mediated cystine uptake and glutathione synthesis inhibited the classically activated macrophage (M1) polarization via reducing Stat1 phosphorylation, and promoted alternatively activated macrophage (M2) polarization via enhancing Stat6 phosphorylation. Macrophage‐targeting lipid nanoparticles loading with ferrostatin‐1, an antioxidant reagent, promotes Slc7a11‐mediated glutathione synthesis, also enhanced plaque stability and ameliorated the progression of atherosclerosis. These findings reveal a critical role of Slc7a11 in the phenotypic switch of macrophage and indicate that Slc7a11‐mediated amino acid metabolism could be utilized as a novel therapeutic strategy in the prevention of ASCVDs.

Slc7a11‐mediated cystine uptake and glutathione synthesis inhibits M1 macrophage polarization by reducing Stat1 phosphorylation and markers expressions and promotes M2 macrophage polarization by enhancing Stat6 phosphorylation and markers expressions. Conversely, glutamate treatment or glutathione depletion reverses this macrophage phenotype. Macrophage‐specific Slc7a11 overexpression or macrophage‐targeting lipid nanoparticles loading with ferrostatin‐1 promotes M2 macrophage polarization and attenuates the development of atherosclerosis.

Figures

Slc7a11 expression is enhanced in the macrophages of atherosclerotic lesions. (A) Pathway enrichment analysis of RNA‐sequencing data from foamy and nonfoamy macrophages isolated from atherosclerotic intima. (B) Heatmap presenting differentially expressed genes in cellular response to oxidative stress, ferroptosis, glutathione metabolic process, and amino acid transmembrane transport pathways. (C) Real‐time PCR analysis of Slc7a11 mRNA level in bone marrow‐derived macrophage (BMDMs) treated with oxidized low‐density lipoprotein (oxLDL, 50 or 100 µg/mL) for 0, 6, 12, and 24 h (n = 3 biological replicates per group). (D and E) Western blot analysis of Slc7a11 protein expression level in BMDMs treated with oxLDL (100 µg/mL) for 0, 12, 24, and 36 h (n = 3 biological replicates per group). (F and G) Western blot analysis of Slc7a11 protein expression level in BMDMs treated with 0, 50, 100, and 200 µg/mL oxLDL for 24 h (n = 3 biological replicates per group). (H) Quantification of flow cytometric analysis for Slc7a11 expression in BMDMs treated with oxLDL (100 or 200 µg/mL) for 36 and 48 h, respectively. (I) Slc7a11 mRNA expression level in BMDMs treated with 50 µg/mL oxLDL for 24 h in the presence or absence of Nrf2‐inhibitor Nrf‐In‐1 (2 µM) (n = 3 biological replicates per group). (J) Representative images of immunofluorescence staining for Slc7a11 (red), Mac‐3 (macrophage marker, green), and 4′,6‐diamidino‐2‐phenylindole (DAPI, cell nuclei, blue) in the atherosclerotic plaques from ApoE–/– mice fed a western diet for 16 weeks. Images are representative of n = 3 independent experiments. Scale bar, 50 µm. Student's t‐test was used to compare the two groups. One‐way ANOVA test was used to compare multiple groups. *p < 0.05, p < 0.01, *p < 0.001. The molecular weight (in kDa) was indicated to the right of each band.

Macrophage‐specific Slc7a11 overexpression attenuates atherosclerotic lesions and increases the plaque stability. Eight‐week‐old ApoE– /– and ApoE–/–Slc7a11MOE mice were fed a western diet for 16 weeks. (A) Hematoxylin and eosin (H&E) staining of aortic plaques from ApoE– /– and ApoE–/–Slc7a11MOE mice. Scale bar, 200 µm. (B) Lesion area of aortic plaques across a 400 µm distance of the aortic root from ApoE– /– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group). (C) Necrotic core areas of aortic plaques across a 400 µm distance of the aortic root from ApoE– /– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group). (D) The proportion of early, moderate, and advanced plaques from ApoE– /– and ApoE–/–Slc7a11MOE mice based on the histological staining (n = 10 mice in each group). (E) Representative images of immunofluorescence staining for smooth muscle cells (α‐SMA, red) and 4′,6‐diamidino‐2‐phenylindole (DAPI, cell nuclei, blue) in the aortic plaques from ApoE– /– and ApoE–/–Slc7a11MOE mice. Scale bar, 50 µm. (F) Quantification of the smooth muscle cells area in the aortic plaques from ApoE– /– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group, and each mouse had two sections). (G) Representative images of immunofluorescence staining for Mac‐3 (macrophage marker, green) and DAPI (cell nuclei, blue) in the aortic plaques from ApoE– /– and ApoE–/–Slc7a11MOE mice. Scale bar, 50 µm. (H) Quantification of macrophages numbers in the aortic plaques from ApoE– /– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group, and each mouse had two sections). (I) Representative images of Masson trichrome staining of aortic plaques from ApoE– /– and ApoE–/–Slc7a11MOE mice. Scale bar, 200 µm. (J) Quantification of Masson trichrome staining in aortic plaques from ApoE– /– and ApoE–/– Slc7a11MOE mice (n = 10 mice in each group, and each mouse had two sections). (K, L, and M) Glutathione (GSH), glutathione peroxidase (GSH‐PX), and malondialdehyde (MDA) levels in serum from ApoE– /– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group). (N) Representative images of immunofluorescence staining for 4‐hydroxynonenal (4‐HNE, lipid oxidation product, red), Mac‐3 (macrophage marker, green), and DAPI (cell nuclei, blue) in the aortic plaques from ApoE– /– and ApoE–/–Slc7a11MOE mice. Scale bar, 50 µm. (O) Quantification of 4‐HNE fluorescence intensity in the aortic plaques from ApoE– /– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group, and each mouse had two sections). Student's t‐test was used to compare the two groups. *p < 0.05, p < 0.01, *p < 0.001.

Slc7a11 overexpression in macrophages inhibits classical M1 polarization and promotes alternative M2 polarization. (A) Classical M1 marker gene expressions in the aorta arteries from ApoE– /– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group). (B) Alternative M2 marker gene expressions in the aorta arteries from ApoE– /– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group). (C) Western blot analysis of p‐Stat1 and Stat1 in bone marrow‐derived macrophage (BMDMs) from ApoE– /– or ApoE–/–Slc7a11MOE mice treated with liposaccharides (LPS, 1 µg/mL) and interferon‐γ (IFN‐γ, 50 ng/mL) for 0, 6, 12, and 24 h (n = 3 biological replicates in each group). (D) Classical M1 marker gene expressions in BMDMs from ApoE– /– or ApoE–/–Slc7a11MOE mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h (n = 3 biological replicates in each group). (E) Western blot analysis of p‐Stat1 and Stat1 in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 0, 6, 12, and 24 h (n = 3 biological replicates in each group). (F) Classical M1 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h (n = 3 biological replicates in each group). (G) Western blot analysis of p‐Stat1 and Stat1 in BMDMs treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 0, 6, 12, and 24 h in the presence and absence of erastin (15 µM) (n = 3 biological replicates in each group). (H) Classical M1 marker gene expressions in BMDMs treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h in the presence and absence of erastin (15 µM) (n = 3 biological replicates in each group). (I) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from ApoE– /– or ApoE–/–Slc7a11MOE mice treated with interleukin‐4 (IL‐4, 30 ng/mL) for 0, 15, 30, 60, 120, and 180 min (n = 3 biological replicates per group). (J) Alternative M2 marker gene expressions in BMDMs from ApoE– /– or ApoE–/–Slc7a11MOE mice treated with IL‐4 (30 ng/mL) for 24 h (n = 3 biological replicates per group). (K) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL‐4 (30 ng/mL) for 0, 15, 30, 60, 120, and 180 min (n = 3 biological replicates in each group). (L) Alternative M2 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL‐4 (30 ng/mL) for 24 h (n = 3 biological replicates in each group). (M) Western blot analysis of p‐Stat6 and Stat6 treated with IL‐4 (30 ng/mL) for 0, 15, 30, 60, 120, and 180 min in the presence and absence of erastin (15 µM) (n = 3 biological replicates in each group). (N) Alternative M2 marker gene expressions in BMDMs treated with IL‐4 (30 ng/mL) for 24 h in the presence and absence of erastin (15 µM) (n = 3 biological replicates in each group). Student's t‐test was used to compare the two groups. One‐way ANOVA test was used to compare multiple groups. *p < 0.05, p < 0.01, *p < 0.001. The molecular weight (in kDa) was indicated to the right of each band.

Slc7a11‐mediated cystine uptake and glutathione synthesis regulate classical M1 macrophages polarization. (A) Western blot analysis of p‐Stat1 and Stat1 in bone marrow‐derived macrophage (BMDMs) treated with liposaccharides (LPS, 1 µg/mL) and interferon‐γ (IFN‐γ, 50 ng/mL) for 0, 6, 12, and 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (B) Western blot analysis of p‐Stat1 and Stat1 in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (C) Western blot analysis of p‐Stat1 and Stat1 in BMDMs from ApoE– /– or ApoE–/–Slc7a11MOE mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (D) Classical M1 marker gene expressions in BMDMs treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (E) Classical M1 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (F) Classical M1 marker gene expressions in BMDMs from ApoE– /– or ApoE–/–Slc7a11MOE mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (G) Western blot analysis of p‐Stat1 and Stat1 in BMDMs treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 0, 6, 12, and 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (H) Western blot analysis of p‐Stat1 and Stat1 in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (I) Western blot analysis of p‐Stat1 and Stat1 in BMDMs from ApoE– /– or ApoE–/–Slc7a11MOE mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (J) Classical M1 marker gene expressions in BMDMs treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (K) Classical M1 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (L) Classical M1 marker gene expressions in BMDMs from ApoE– /– or ApoE–/–Slc7a11MOE mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (M) Western blot analysis of p‐Stat1 and Stat1 in BMDMs treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 0, 6, 12, and 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (N) Western blot analysis of p‐Stat1 and Stat1 in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (O) Western blot analysis of p‐Stat1 and Stat1 in BMDMs from ApoE– /– or ApoE–/–Slc7a11MOE mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (P) Classical M1 marker gene expressions in BMDMs treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (Q) Classical M1 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (R) Classical M1 marker gene expressions in BMDMs from ApoE– /–or ApoE–/–Slc7a11MOE mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). Student's t‐test was used to compare the two groups. One‐way ANOVA test was used to compare multiple groups. *p < 0.05, p < 0.01, *p < 0.001. The molecular weight (in kDa) was indicated to the right of each band.

Slc7a11‐mediated cystine uptake and glutathione synthesis control alternative M2 macrophages polarization. (A) Western blot analysis of p‐Stat6 and Stat6 in bone marrow‐derived macrophage (BMDMs) treated with interleukin‐4 (IL‐4, 30 ng/mL) for 0, 15, 30, 60, 120, and 180 min after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (B) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL‐4 (30 ng/mL) for 30 min after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (C) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from ApoE– /– or ApoE–/–Slc7a11MOE mice treated with IL‐4 (30 ng/mL) for 30 min after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (D) Alternative M2 marker gene expressions in BMDMs treated with IL‐4 (30 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (E) Alternative M2 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL‐4 (30 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (F) Alternative M2 marker gene expressions in BMDMs from ApoE– /– or ApoE–/–Slc7a11MOE mice treated with IL‐4 (30 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (G) Western blot analysis of p‐Stat6 and Stat6 in BMDMs treated with IL‐4 (30 ng/mL) for 0, 15, 30, 60, 120, and 180 min after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (H) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL‐4 (30 ng/mL) for 30 min after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (I) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from ApoE– /– or ApoE–/–Slc7a11MOE mice treated with IL‐4 (30 ng/mL) for 30 min after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (J) Alternative M2 marker gene expressions in BMDMs treated with IL‐4 (30 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (K) Alternative M2 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL4 (30 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (L) Alternative M2 marker gene expressions in BMDMs from ApoE– /– or ApoE–/–Slc7a11MOE mice treated with IL‐4 (30 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (M) Western blot analysis of p‐Stat6 and Stat6 in BMDMs treated with IL‐4 (30 ng/mL) for 0, 15, 30, 60, 120, and 180 min after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (N) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL‐4 (30 ng/mL) for 30 min after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (O) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from ApoE– /– or ApoE–/–Slc7a11MOE mice treated with IL‐4 (30 ng/mL) for 30 min after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (P) Alternative M2 marker gene expressions in BMDMs treated with IL‐4 (30 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (Q) Alternative M2 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL‐4 (30 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (R) Alternative M2 marker gene expressions in BMDMs from ApoE– /–or ApoE–/–Slc7a11MOE mice treated with IL‐4 (30 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). Student's t‐test was used to compare the two groups. One‐way ANOVA test was used to compare multiple groups. *p < 0.05, p < 0.01, *p < 0.001. The molecular weight (in kDa) was indicated to the right of each band.

Macrophage‐targeting lipid nanoparticles (LNP) loading with ferrostatin‐1 (Fer‐1) attenuate atherosclerotic plaque in ApoE–/– mice. (A) Construction of the bilayer lipid nanoparticles modified with galactose loading with Fer‐1 (LNP–Fer‐1). The lipid nanoparticles have the drug sustained release effects in vivo. The LNP–Fer‐1 nanoparticles were composed of cholesterol, hydrogenated soybean phosphatidylcholine (HSPC), 1,2‐distearoyl‐sn‐glycero‐3‐phosphoethanolamine‐N‐[methoxy(polyethylene glycol)‐2000] (DSPE–mPEG2000), and DSPE–mPEG2000–galactose. (B) Schematic illustration of macrophage‐targeting lipid nanoparticles (LNP–Fer‐1) therapy in atherosclerotic ApoE–/– mice. Eight‐week‐old ApoE– /– mice received the tail vein injection of saline, LNP, Fer‐1, and LNP–Fer‐1 (twice for 1 week; each time 1 mg/kg calculated by Fer‐1; other groups with equal volume) for 6 weeks after 12 weeks of atherosclerotic modeling (n = 5 mice in each group). (C) Hematoxylin and eosin (H&E) staining of aortic plaques in saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group). Scale bar, 200 µm. (D) Lesion area of aortic plaques across a 400 µm distance of the aortic root in saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group). (E) Necrotic core areas of aortic plaques across a 400 µm distance of the aortic root in saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group). (F) The proportion of early, moderate, and advanced plaques in saline, Fer‐1, LNP, and LNP–Fer‐1 groups based on the histological staining (n = 5 mice in each group). (G) Quantification of the smooth muscle cells (α‐SMA) area for the aortic plaques in saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group, at least 2 sections per mouse). (H) Quantification of Masson trichrome staining for aortic plaques in saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group, at least 2 sections per mouse). (I, J, and K) Glutathione (GSH), glutathione peroxidase (GSH‐PX), and malondialdehyde (MDA) levels of serum in saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group). (L) Quantification of 4‐hydroxynonenal (4‐HNE) fluorescence intensity in the aortic plaques from saline, Fer‐1, LNP, and LNP–Fer‐1 groups. (n = 5 mice in each group, at least 2 sections per mouse). (M) Quantification of macrophages numbers (Mac‐3) for the aortic plaques in saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group, at least 2 sections per mouse). (N) Classical M1 marker gene expressions from aorta arteries of saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group). (O) Alternative M2 marker gene expressions from aorta arteries of saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group). One‐way ANOVA test was used to compare multiple groups. *p < 0.05, p < 0.01, *p < 0.001.

Slc7a11‐mediated cystine/glutamate antiport regulates macrophage polarization and atherosclerosis development. Slc7a11‐mediated cystine uptake and glutathione synthesis inhibits the classically activated macrophage polarization (M1) by reducing the Stat1 phosphorylation and classical M1 gene expressions (Tnfa, Nos2, Il6, Il1b), and promotes alternatively activated macrophage polarization (M2) by enhancing Stat6 phosphorylation and alternative M2 gene expressions (Arg1, Mrc1, Irf4, Retnla). In contrast, glutamate treatment or glutathione depletion reverses the phenotypic switch of macrophage from M1 to M2. Macrophage‐specific Slc7a11 overexpression or macrophage‐targeting lipid nanoparticles loading with ferrostatin‐1 (LNP–Fer‐1) promotes M2 macrophage polarization, decreases atherosclerotic macrophage content, and attenuates the development of atherosclerosis.

Full text

Introduction

1

Atherosclerosis is a typical chronic inflammation disease characterized by the presence of cholesterol‐engorged macrophages in the arterial plaques [1]. Atherosclerotic cardiovascular diseases (ASCVD) remains the primary cause of morbidity and mortality in the world [2]. Progression of atherosclerosis involves the activation of numerous cell types, leading to a localized inflammatory response [3]. Macrophage metabolism alterations and cell death promote the formation of a necrotic core, whereas reduced efferocytosis levels contribute to the growth of the core [4, 5], which further triggers inflammation, necrosis, and thrombosis. Classically activated macrophages (M1) secrete proinflammatory cytokines (e.g., TNF‐α, IL‐1β), promoting tissue damage and plaque rupture. In contrast, alternatively activated macrophages (M2) release anti‐inflammatory mediators that facilitate tissue repair and contribute to plaque stabilization [6]. With the high heterogeneity and plasticity of macrophages in plaques, the macrophage phenotypes can influence the progression and regression of atherosclerosis [7, 8]. Given the complex and diverse microenvironment in the plaques, it is critical to develop new therapeutic strategies for antiatherosclerosis.

Changes in amino acid metabolism influence immune cell polarization and proliferation, resulting in localized inflammation and increased cell‐to‐cell interactions in atherosclerosis [9]. Glycine‐based therapy attenuated atherosclerosis by inducing glutathione (GSH) production [10]. Oral l‐arginine reduced the incidence rate of cardiovascular death, reinfarction, and recurrent myocardial ischemia in patients with ST‐segment elevation myocardial infarction [11]. However, the effect of branched‐chain amino acid (leucine, isoleucine and valine) homeostasis on cardiovascular health was still controversial. Leucine supplementation in drinking water improved the plasma lipid profiles and systemic inflammation and attenuated atherosclerosis in ApoE null mice [12]. In contrast, high‐protein diet‐induced plasma leucine elevation inhibited the monocyte/macrophage autophagy mediated by mTORC1 and promoted atherogenesis [13]. Therefore, amino acid metabolism plays a complicated role in macrophage proliferation and polarization. The roles of other amino acids and their metabolism such as cystine and glutamate in atherogenesis remain unclear.

SLC7A11, the light chain subunit of amino acid transporter system Xc–, imports the extracellular cystine and exports the intracellular glutamate for GSH biosynthesis [14]. Emerging evidence suggested an association between Slc7a11 and cardiovascular diseases (CVDs). For example, overexpressing Slc7a11 in cardiomyocytes raised GSH level and reduced ferroptosis‐induced cardiomyopathy [15]. Inhibition of Slc7a11 promoted the rat vascular smooth muscle cell calcification [16]. The two amino acids involved in Slc7a11 transport system are closely related to cardiovascular health. Plasma glutamate levels displayed a positive correlation with the risk of CVDs, particularly stroke events [17, 18]. Elevated circulating glutamate levels were associated with subclinical atherosclerosis in a cross‐sectional study [18]. It suggested that the amino acids mediated by Slc7a11 may play a role in the atherosclerosis process. However, the function of Slc7a11 and its mediated amino acids in atherosclerosis remains unclear.

In the present study, we reveal that Slc7a11 expression is specifically enhanced in the macrophages of atherosclerotic plaques from *ApoE^–/–^

• mice. Therefore, we aim to investigate how macrophage *Slc7a11‐*mediated amino acid transport modulates the development of atherosclerosis. We found that Slc7a11‐mediated cystine/glutamate exchange modulates the phenotypic switch of macrophage and influences the progress of atherosclerosis. Genetically macrophage‐specific Slc7a11 overexpression or a lipid nanoparticle (LNP) targeting macrophage cystine metabolism suppresses atherosclerosis development in *ApoE^–/–^
• mice.

Results

2

Slc7a11 Expression is Specifically Enhanced in the Macrophages of Atherosclerotic Lesions

2.1

Foam cells generation contributes to the onset and development of atherosclerosis, and oxidized low‐density lipoprotein (oxLDL) is typically used to induce macrophage to form foam cells [19, 20]. To explore the difference between foamy macrophages and nonfoamy macrophages, we analyzed the RNA‐sequencing data from atherosclerotic plaque in mice [21]. Transcriptome analyses revealed the participation of pathways in oxidant stress, macrophage changes, ferroptosis, and amino acid transport (*p *< 0.05) (Figure 1A). Slc7a11 gene expression level was top‐rated in signaling pathways involved in the development of atherosclerosis, including cellular response to oxidative stress, ferroptosis, GSH metabolic process, and amino acid transmembrane transport (Figure 1B). To further investigate whether macrophage Slc7a11 was involved in the atherosclerosis progression, we used oxLDL to stimulate the expression of Slc7a11. In bone marrow‐derived macrophages (BMDMs), Slc7a11 mRNA expression was upregulated by oxLDL, and displayed a dose‐dependent manner (Figure 1C). Consistent with the mRNA regulatory patterns, oxLDL also increased Slc7a11 protein level in a time‐ and dose‐dependent manner (Figure 1D,G). The oxLDL‐induced protein expression of Slc7a11 was verified by using flow cytometry (Figures 1H and S1). The nuclear factor erythroid 2‐related factor 2 (Nrf2) transcribes Slc7a11 mRNA expression under stress condition [22]. Therefore, we investigated whether Nrf2 contributed to the oxLDL‐induced Slc7a11 expression. Notably, the oxLDL‐induced mRNA expression of Slc7a11 was inhibited in BMDMs treated with Nrf2 inhibitor (Nrf2‐In‐1) (Figure 1I). To further confirm the involvement of Slc7a11 in atherosclerosis, we explored macrophagic Slc7a11 expression in the atherosclerotic plaques of arteries samples from *ApoE–^/–^

• mice receiving 16‐week western diet. Immunofluorescence staining showed that membrane transporter Slc7a11 colocalized with Mac‐3 macrophages in the atherosclerotic plaques of *ApoE^–/–^
• mice receiving 16‐week western diet (Figures 1J and S2). Collectively, these results suggested that Slc7a11 expression was enhanced in the macrophages of atherosclerotic plaques.

*Slc7a11 expression is enhanced in the macrophages of atherosclerotic lesions. (A) Pathway enrichment analysis of RNA‐sequencing data from foamy and nonfoamy macrophages isolated from atherosclerotic intima. (B) Heatmap presenting differentially expressed genes in cellular response to oxidative stress, ferroptosis, glutathione metabolic process, and amino acid transmembrane transport pathways. (C) Real‐time PCR analysis of Slc7a11 mRNA level in bone marrow‐derived macrophage (BMDMs) treated with oxidized low‐density lipoprotein (oxLDL, 50 or 100 µg/mL) for 0, 6, 12, and 24 h (n = 3 biological replicates per group). (D and E) Western blot analysis of Slc7a11 protein expression level in BMDMs treated with oxLDL (100 µg/mL) for 0, 12, 24, and 36 h (n = 3 biological replicates per group). (F and G) Western blot analysis of Slc7a11 protein expression level in BMDMs treated with 0, 50, 100, and 200 µg/mL oxLDL for 24 h (n = 3 biological replicates per group). (H) Quantification of flow cytometric analysis for Slc7a11 expression in BMDMs treated with oxLDL (100 or 200 µg/mL) for 36 and 48 h, respectively. (I) Slc7a11 mRNA expression level in BMDMs treated with 50 µg/mL oxLDL for 24 h in the presence or absence of Nrf2‐inhibitor Nrf‐In‐1 (2 µM) (n = 3 biological replicates per group). (J) Representative images of immunofluorescence staining for Slc7a11 (red), Mac‐3 (macrophage marker, green), and 4′,6‐diamidino‐2‐phenylindole (DAPI, cell nuclei, blue) in the atherosclerotic plaques from ApoE–/– mice fed a western diet for 16 weeks. Images are representative of n = 3 independent experiments. Scale bar, 50 µm. Student's t‐test was used to compare the two groups. One‐way ANOVA test was used to compare multiple groups. *p < 0.05, **p < 0.01, **p < 0.001. The molecular weight (in kDa) was indicated to the right of each band.

Macrophage‐Specific Slc7a11 Overexpression Attenuates Atherosclerotic Lesions and Increases Plaque Stability

2.2

To investigate the pathophysiological role of macrophage Slc7a11 in atherosclerosis development, we generated *ApoE^–/–^

• mice with macrophage‐specific Slc7a11 overexpression (*ApoE^–/–^Slc7a11^MOE^
• mice). Male *ApoE^–^

^/–^ and *ApoE^–/–^Slc7a11^MOE^

• mice were fed a western diet for 16 weeks, then the pathological changes of atherosclerosis were evaluated. Body weight (Figure S3A) or serum lipid profiles (triglyceride [TG], total cholesterol [TC], low‐density lipoprotein cholesterol [LDL‐C], and high‐density lipoprotein cholesterol [HDL‐C]) presented no significant difference between *ApoE^–/–^Slc7a11^MOE^
• and *ApoE^–^

^/–^ mice (Figure S3B–E). The atherosclerotic lesion area in *ApoE–^/–^Slc7a11^MOE^

• mice was regressed by 61% in the aortic roots when compared with *ApoE^–/–^
• mice (Figure 2A,B). Cross‐sectional analysis of aortic roots further demonstrated a remarkable reduction of the plaque necrotic core area in *ApoE^–/–^Slc7a11^MOE^
• mice (Figure 2C). The classification of aortic plaque by Stary method showed that *ApoE^–/–^Slc7a11^MOE^
• mice had more plaques in the early and moderate stages, whereas *ApoE^–/–^
• mice had more severe atherosclerotic plaques (Figure 2D). *ApoE^–/–^Slc7a11^MOE^
• mice had higher ratios of α‐smooth muscle actin (α‐SMA) (Figures 2E,F and S3F) and lower Mac‐3 positive cells (Figure 2G,H), indicating more smooth muscle cells within the fibrous cap and fewer macrophage infiltration in *ApoE^–/–^Slc7a11^MOE^
• mice compared with *ApoE^–/–^
• mice. Masson trichrome staining of aortic sinus plaques revealed more collagen accumulation in *ApoE^–/–^Slc7a11^MOE^
• mice compared with *ApoE^–/–^
• mice (Figure 2I,J). In addition, the oxidative stress level was also evaluated by detecting circulating antioxidants (GSH and GSH peroxidase [GSH‐PX]) and lipid peroxidation products (malondialdehyde [MDA] and 4‐hydroxynonenal [4‐HNE]). Compared with *ApoE^–/–^
• mice, *ApoE^–/–^Slc7a11^MOE^
• mice displayed higher plasma GSH and GSH‐PX levels (Figure 2K,L), but lower plasma MDA level (Figure 2M). Costaining of Mac‐3 and 4‐HNE in atherosclerotic plaques showed a reduced macrophage lipid peroxidation level in *ApoE^–/–^Slc7a11^MOE^
• when compared with *ApoE^–/–^
• mice (Figure 2N,O). Taken together, these results indicated that macrophage‐specific Slc7a11 overexpression remarkably alleviated atherosclerosis development and enhanced the stability of atherosclerotic lesions by decreasing macrophage infiltration and oxidative stress.

*Macrophage‐specific Slc7a11 overexpression attenuates atherosclerotic lesions and increases the plaque stability. Eight‐week‐old ApoE–

/– and ApoE–/–Slc7a11MOE mice were fed a western diet for 16 weeks. (A) Hematoxylin and eosin (H&E) staining of aortic plaques from ApoE–

/– and ApoE–/–Slc7a11MOE mice. Scale bar, 200 µm. (B) Lesion area of aortic plaques across a 400 µm distance of the aortic root from ApoE–

/– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group). (C) Necrotic core areas of aortic plaques across a 400 µm distance of the aortic root from ApoE–

/– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group). (D) The proportion of early, moderate, and advanced plaques from ApoE–

/– and ApoE–/–Slc7a11MOE mice based on the histological staining (n = 10 mice in each group). (E) Representative images of immunofluorescence staining for smooth muscle cells (α‐SMA, red) and 4′,6‐diamidino‐2‐phenylindole (DAPI, cell nuclei, blue) in the aortic plaques from ApoE–

/– and ApoE–/–Slc7a11MOE mice. Scale bar, 50 µm. (F) Quantification of the smooth muscle cells area in the aortic plaques from ApoE–

/– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group, and each mouse had two sections). (G) Representative images of immunofluorescence staining for Mac‐3 (macrophage marker, green) and DAPI (cell nuclei, blue) in the aortic plaques from ApoE–

/– and ApoE–/–Slc7a11MOE mice. Scale bar, 50 µm. (H) Quantification of macrophages numbers in the aortic plaques from ApoE–

/– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group, and each mouse had two sections). (I) Representative images of Masson trichrome staining of aortic plaques from ApoE–

/– and ApoE–/–Slc7a11MOE mice. Scale bar, 200 µm. (J) Quantification of Masson trichrome staining in aortic plaques from ApoE–

/– and ApoE–/– Slc7a11MOE mice (n = 10 mice in each group, and each mouse had two sections). (K, L, and M) Glutathione (GSH), glutathione peroxidase (GSH‐PX), and malondialdehyde (MDA) levels in serum from ApoE–

/– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group). (N) Representative images of immunofluorescence staining for 4‐hydroxynonenal (4‐HNE, lipid oxidation product, red), Mac‐3 (macrophage marker, green), and DAPI (cell nuclei, blue) in the aortic plaques from ApoE–

/– and ApoE–/–Slc7a11MOE mice. Scale bar, 50 µm. (O) Quantification of 4‐HNE fluorescence intensity in the aortic plaques from ApoE–

/– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group, and each mouse had two sections). Student's t‐test was used to compare the two groups. *p < 0.05, **p < 0.01, **p < 0.001.

Slc7a11 Overexpression in Macrophages Inhibits M1 Polarization and Promotes M2 Polarization

2.3

Studies have shown that increasing macrophage polarization toward an M1 phenotype or decreasing polarization toward an M2 phenotype accelerates the progression of atherosclerotic plaques [23, 24]. To determine how Slc7a11 attenuated the atherosclerosis progression in macrophages, we measured the mRNA expression of macrophages polarization markers in the aortic arteries from *ApoE^–/–^Slc7a11^MOE^

• and *ApoE^–/–^
• mice. M1 markers (Tnfa, Nos2, Il6, Il1b) were less expressed, whereas M2 markers (Arg1, Mrc1, Irf4, Retnla) were more expressed in the aortas from *ApoE^–/–^Slc7a11^MOE^
• mice when compared with *ApoE^–/–^
• mice (Figure 3A,B). It suggested that Slc7a11 overexpression might attenuate atherosclerosis via modulating macrophages polarization in *ApoE^–/–^
• mice.

*Slc7a11 overexpression in macrophages inhibits classical M1 polarization and promotes alternative M2 polarization. (A) Classical M1 marker gene expressions in the aorta arteries from ApoE–

/– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group). (B) Alternative M2 marker gene expressions in the aorta arteries from ApoE–

/– and ApoE–/–Slc7a11MOE mice (n = 10 mice in each group). (C) Western blot analysis of p‐Stat1 and Stat1 in bone marrow‐derived macrophage (BMDMs) from ApoE–

/– or ApoE–/–Slc7a11MOE mice treated with liposaccharides (LPS, 1 µg/mL) and interferon‐γ (IFN‐γ, 50 ng/mL) for 0, 6, 12, and 24 h (n = 3 biological replicates in each group). (D) Classical M1 marker gene expressions in BMDMs from ApoE–

/– or ApoE–/–Slc7a11MOE mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h (n = 3 biological replicates in each group). (E) Western blot analysis of p‐Stat1 and Stat1 in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 0, 6, 12, and 24 h (n = 3 biological replicates in each group). (F) Classical M1 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h (n = 3 biological replicates in each group). (G) Western blot analysis of p‐Stat1 and Stat1 in BMDMs treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 0, 6, 12, and 24 h in the presence and absence of erastin (15 µM) (n = 3 biological replicates in each group). (H) Classical M1 marker gene expressions in BMDMs treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h in the presence and absence of erastin (15 µM) (n = 3 biological replicates in each group). (I) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from ApoE–

/– or ApoE–/–Slc7a11MOE mice treated with interleukin‐4 (IL‐4, 30 ng/mL) for 0, 15, 30, 60, 120, and 180 min (n = 3 biological replicates per group). (J) Alternative M2 marker gene expressions in BMDMs from ApoE–

/– or ApoE–/–Slc7a11MOE mice treated with IL‐4 (30 ng/mL) for 24 h (n = 3 biological replicates per group). (K) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL‐4 (30 ng/mL) for 0, 15, 30, 60, 120, and 180 min (n = 3 biological replicates in each group). (L) Alternative M2 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL‐4 (30 ng/mL) for 24 h (n = 3 biological replicates in each group). (M) Western blot analysis of p‐Stat6 and Stat6 treated with IL‐4 (30 ng/mL) for 0, 15, 30, 60, 120, and 180 min in the presence and absence of erastin (15 µM) (n = 3 biological replicates in each group). (N) Alternative M2 marker gene expressions in BMDMs treated with IL‐4 (30 ng/mL) for 24 h in the presence and absence of erastin (15 µM) (n = 3 biological replicates in each group). Student's t‐test was used to compare the two groups. One‐way ANOVA test was used to compare multiple groups. *p < 0.05, **p < 0.01, **p < 0.001. The molecular weight (in kDa) was indicated to the right of each band.

We investigated the underlying mechanism how Slc7a11 modulates macrophage polarization. The M1 macrophage phenotype is controlled by signal transducer and activator of transcription 1 (Stat1), which is activated through its phosphorylation in response to interferons and other immunological signals [25, 26]. Lipopolysaccharide (LPS) and interferon‐γ (IFN‐γ) were used to activate M1 phenotype in BMDMs from wildtype, *Slc7a11^–/–^

• and *Slc7a11^MOE^
• mice, and then Stat1 phosphorylation and the mRNA levels of M1 markers (Tnfa, Nos2, Il6, Il1b) were measured to evaluate the M1 polarization. Compared with wildtype BMDMs, Slc7a11 overexpression decreased LPS/IFN‐γ‐induced M1 polarization (Figure 3C,D), while Slc7a11‐deficient macrophages promoted LPS/IFN‐γ‐induced M1 polarization (Figure 3E,F). Similar to *Slc7a11^–/–^
• macrophages, pharmacological inhibition of Slc7a11 via erastin also remarkably increased LPS/IFN‐γ‐induced Stat1 phosphorylation and M1 marker expressions (Figure 3G,H).

Next, the influence of Slc7a11 on M2 macrophage polarization was investigated. IL‐4 was used to activate M2 macrophages phenotype, and Stat6 phosphorylation and the mRNA level of M2 markers (Arg1, Mrc1, Irf4, Retnla) were measured. Compared with wildtype BMDMs, Slc7a11 overexpression promoted IL‐4‐induced M2 polarization (Figure 3I,J), while either Slc7a11 deficiency or erastin treatment suppressed IL‐4‐induced M2 polarization (Figure 3K–N). Collectively, these results demonstrated that Slc7a11 inhibited LPS/IFN‐γ‐induced M1 polarization and promoted IL‐4‐induced M2 polarization.

Slc7a11‐Mediated Cystine Uptake and GSH Synthesis Promote the Phenotypic Switch of Macrophage From M1 to M2

2.4

Slc7a11 mediates the uptake of extracellular cystine in exchange for intracellular glutamate to synthesize GSH [14, 27]. We further investigated whether Slc7a11‐mediated macrophage polarization was through cystine uptake and GSH synthesis. GSH treatment could significantly decrease LPS/IFN‐γ‐induced M1 polarization in BMDMs from wildtype, *Slc7a11^–/–^ *, *ApoE^–/–^ *, and *ApoE^–/–^Slc7a11^MOE^

• mice (Figure 4A–F), suggesting GSH suppressed macrophage M1 polarization. GSH treatment decreased Stat1 phosphorylation (Figure 4A–C) and M1 markers mRNA expressions (Figure 4D–F) in BMDMs from wildtype, *Slc7a11^–/–^ *, *ApoE^–/–^
• and *ApoE^–/–^Slc7a11^MOE^
• mice. Specifically, in Slc7a11‐deficient macrophages, GSH treatment reverted Stat1 phosphorylation and M1 markers to wildtype macrophage level (Figure 4B,E). In *Slc7a11^MOE^
• macrophages, GSH further inhibited M1 polarization when Slc7a11 overexpressed (Figure 4C,F).

*Slc7a11‐mediated cystine uptake and glutathione synthesis regulate classical M1 macrophages polarization. (A) Western blot analysis of p‐Stat1 and Stat1 in bone marrow‐derived macrophage (BMDMs) treated with liposaccharides (LPS, 1 µg/mL) and interferon‐γ (IFN‐γ, 50 ng/mL) for 0, 6, 12, and 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (B) Western blot analysis of p‐Stat1 and Stat1 in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (C) Western blot analysis of p‐Stat1 and Stat1 in BMDMs from ApoE–

/– or ApoE–/–Slc7a11MOE mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (D) Classical M1 marker gene expressions in BMDMs treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (E) Classical M1 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (F) Classical M1 marker gene expressions in BMDMs from ApoE–

/– or ApoE–/–Slc7a11MOE mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (G) Western blot analysis of p‐Stat1 and Stat1 in BMDMs treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 0, 6, 12, and 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (H) Western blot analysis of p‐Stat1 and Stat1 in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (I) Western blot analysis of p‐Stat1 and Stat1 in BMDMs from ApoE–

/– or ApoE–/–Slc7a11MOE mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (J) Classical M1 marker gene expressions in BMDMs treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (K) Classical M1 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (L) Classical M1 marker gene expressions in BMDMs from ApoE–

/– or ApoE–/–Slc7a11MOE mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (M) Western blot analysis of p‐Stat1 and Stat1 in BMDMs treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 0, 6, 12, and 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (N) Western blot analysis of p‐Stat1 and Stat1 in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (O) Western blot analysis of p‐Stat1 and Stat1 in BMDMs from ApoE–

/– or ApoE–/–Slc7a11MOE mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (P) Classical M1 marker gene expressions in BMDMs treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (Q) Classical M1 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (R) Classical M1 marker gene expressions in BMDMs from ApoE–

/–or ApoE–/–Slc7a11MOE mice treated with LPS (1 µg/mL) and IFN‐γ (50 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). Student's t‐test was used to compare the two groups. One‐way ANOVA test was used to compare multiple groups. *p < 0.05, **p < 0.01, **p < 0.001. The molecular weight (in kDa) was indicated to the right of each band.

Next, we used cystine‐deprived medium to abolish GSH synthesis to evaluate its influence on macrophage polarization. Cystine deprivation increased Stat1 phosphorylation (Figure 4G–I) and M1 markers mRNA expressions (Figure 4J>–L) in BMDMs from wildtype, *Slc7a11^–/–^ *, *ApoE^–/–^ *, and *ApoE^–/–^Slc7a11^MOE^

• mice. Consistent with cystine deprivation, glutamate treatment, which disrupts cystine uptake via Slc7a11, also increased Stat1 phosphorylation (Figure 4M–O) and M1 markers mRNA expressions (Figure 4P–R) in BMDMs from wildtype, *Slc7a11^–/–^ *, *ApoE^–/–^ *, and *ApoE^–/–^Slc7a11^MOE^
• mice. These data suggested that Slc7a11 inhibited macrophage M1 polarization through cystine uptake and GSH synthesis.

The effects of GSH, cystine, and glutamate on M2 polarization were measured by measuring Stat6 phosphorylation and M2 markers mRNA expressions in BMDMs treated with IL‐4. GSH treatment significantly increased IL‐4‐induced M2 polarization in BMDMs from wildtype, *Slc7a11^–/–^ *, *ApoE^–/–^ *, and *ApoE^–/–^Slc7a11^MOE^

• mice (Figure 5A–F), suggesting GSH promoted macrophage toward the M2 phenotype. Compared with wildtype macrophage, GSH treatment significantly reverted Stat6 phosphorylation and M2 marker expressions in *Slc7a11‐*deficient macrophages (Figure 5B,E). GSH treatment further promoted M2 polarization in *Slc7a11^MOE^
• macrophages (Figure 5C,F). On the other hand, either cystine deprivation or excess glutamate treatment decreased Stat6 phosphorylation and M2 markers mRNA expressions in BMDMs from wildtype, *Slc7a11^–/–^ *, *ApoE^–/–^ *, and *ApoE^–/–^Slc7a11^MOE^
• mice (Figure 5G–R). Taken together, these data suggested that Slc7a11 inhibited macrophage M1 polarization but promoted M2 polarization through modulating cystine‐mediated GSH synthesis.

*Slc7a11‐mediated cystine uptake and glutathione synthesis control alternative M2 macrophages polarization. (A) Western blot analysis of p‐Stat6 and Stat6 in bone marrow‐derived macrophage (BMDMs) treated with interleukin‐4 (IL‐4, 30 ng/mL) for 0, 15, 30, 60, 120, and 180 min after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (B) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL‐4 (30 ng/mL) for 30 min after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (C) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from ApoE–

/– or ApoE–/–Slc7a11MOE mice treated with IL‐4 (30 ng/mL) for 30 min after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (D) Alternative M2 marker gene expressions in BMDMs treated with IL‐4 (30 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (E) Alternative M2 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL‐4 (30 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (F) Alternative M2 marker gene expressions in BMDMs from ApoE–

/– or ApoE–/–Slc7a11MOE mice treated with IL‐4 (30 ng/mL) for 24 h after incubation with or without glutathione (1 mM) for 6 h (n = 3 biological replicates per group). (G) Western blot analysis of p‐Stat6 and Stat6 in BMDMs treated with IL‐4 (30 ng/mL) for 0, 15, 30, 60, 120, and 180 min after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (H) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL‐4 (30 ng/mL) for 30 min after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (I) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from ApoE–

/– or ApoE–/–Slc7a11MOE mice treated with IL‐4 (30 ng/mL) for 30 min after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (J) Alternative M2 marker gene expressions in BMDMs treated with IL‐4 (30 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (K) Alternative M2 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL4 (30 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (L) Alternative M2 marker gene expressions in BMDMs from ApoE–

/– or ApoE–/–Slc7a11MOE mice treated with IL‐4 (30 ng/mL) for 24 h after cystine deprivation or none deprivation for 6 h (n = 3 biological replicates per group). (M) Western blot analysis of p‐Stat6 and Stat6 in BMDMs treated with IL‐4 (30 ng/mL) for 0, 15, 30, 60, 120, and 180 min after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (N) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL‐4 (30 ng/mL) for 30 min after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (O) Western blot analysis of p‐Stat6 and Stat6 in BMDMs from ApoE–

/– or ApoE–/–Slc7a11MOE mice treated with IL‐4 (30 ng/mL) for 30 min after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (P) Alternative M2 marker gene expressions in BMDMs treated with IL‐4 (30 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (Q) Alternative M2 marker gene expressions in BMDMs from wildtype or Slc7a11 deficiency mice treated with IL‐4 (30 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). (R) Alternative M2 marker gene expressions in BMDMs from ApoE–

/–or ApoE–/–Slc7a11MOE mice treated with IL‐4 (30 ng/mL) for 24 h after incubation with or without glutamate (600 µM) for 6 h (n = 3 biological replicates per group). Student's t‐test was used to compare the two groups. One‐way ANOVA test was used to compare multiple groups. *p < 0.05, **p < 0.01, **p < 0.001. The molecular weight (in kDa) was indicated to the right of each band.

Macrophage‐Targeting LNP Loading With Fer‐1 Attenuate Atherosclerotic Plaque in ApoE–/–

Mice

2.5

Ferrostatin‐1 (Fer‐1) therapy has been used to restore cellular GSH content by scavenging lipid reactive oxygen species (ROS) [28, 29]. Fer‐1 treatment suppressed classical M1 polarization while promoting alternative M2 polarization in BMDMs (Figure S4). To achieve macrophage‐specific targeting and sustained release in mice, we constructed a LNP modified with galactose encapsulating Fer‐1 (LNP–Fer‐1). Macrophages possess the natural high phagocytic capability to internalize nanoparticles, while the galactose on the surface of LNPs can promote the receptor‐mediated endocytosis of macrophage within atherosclerotic lesions [30]. The LNPs encapsulating Fer‐1 were composed of cholesterol, hydrogenated soybean phosphatidylcholine (HSPC), 1,2‐distearoyl‐sn‐glycero‐3‐phosphoethanolamine‐N‐[methoxy(polyethylene glycol)‐2000] (DSPE–mPEG2000), and DSPE–mPEG2000–galactose using a thin‐film hydration method according the previous studies (Figure 6A) [31]. High‐performance liquid chromatography (HPLC) analysis showed that the encapsulation efficacy of Fer‐1 in the LNPs was 80.8%, and the drug loading efficacy (DLE%) was 5.4%, indicating the lipophilic Fer‐1 was efficiently embedded within the LNPs. Transmission electron microscopy inspection showed that LNP–Fer‐1 was nanometer‐sized spherical particles and uniformly dispersed in aqueous solution (Figure S5A,B). LNP–Fer‐1 displayed an absorption peak in the UV–vis spectrum (Figure S5C), while the zeta potential of LNP–Fer‐1 also showed a significant decrease when compared with an empty LNP (Figure S5D). The dynamic light scattering analysis showed that the mean diameter of LNP–Fer‐1 and LNP was 133.3 ± 5.6 and 111.5 ± 4.3 nm, respectively (Figure S5E). To verify the macrophage‐targeting capacity of LNP–Fer‐1, BMDMs from wildtype mice was coincubated with LNP–Fer‐1 labeled with Cy5.5 (LNP–Fer‐1–Cy5.5) for 6 h. Fluorescence signal from LNP–Fer‐1–Cy5.5 could be observed within BMDMs (Figure S5F).

To evaluate the therapeutic efficacy of LNP–Fer‐1 on atherosclerosis, LNP, Fer‐1, LNP–Fer‐1, or saline was respectively administered to male *ApoE^–/–^

• mice via tail vein injections for 6 weeks (twice for 1 week; each time 1 mg/kg calculated by Fer‐1; other groups with equal volume) after 12 weeks of atherosclerotic modeling by high‐fat diet (Figure 6B). No significant difference was observed in the final body weight in each group (Figure S6A). Serum TG, TC, and LDL‐C were decreased in mice receiving Fer‐1 or LNP–Fer‐1 (Figure S6B–E). No cardiac, hepatic, and renal injury was also found in groups receiving LNP, Fer‐1, or LNP–Fer1 administration (Figure S6F–M). Mice receiving Fer‐1 or LNP–Fer‐1 displayed alleviated pathological parameters of atherosclerosis. In particular, LNP–Fer‐1 treatment further attenuated the atherosclerotic lesions and necrotic areas in the progression of atherosclerosis compared with mice receiving Fer‐1, LNP, or saline (Figure 6C–F). LNP–Fer‐1 also stabilized the atherosclerotic plaques by increasing the vascular smooth muscle cell ratio (Figures 6G, S6N, and S7A) and collagen content (Figures 6H and S7B), when compared with mice receiving Fer‐1, LNP, or saline. Moreover, LNP–Fer‐1 treatment also showed increased GSH level and GSH‐PX activity, but decreased lipid peroxidation, reflected by lower MDA and 4‐HNE levels than other treatments (Figures 6I–L and S7C).

*Macrophage‐targeting lipid nanoparticles (LNP) loading with ferrostatin‐1 (Fer‐1) attenuate atherosclerotic plaque in ApoE–/– mice. (A) Construction of the bilayer lipid nanoparticles modified with galactose loading with Fer‐1 (LNP–Fer‐1). The lipid nanoparticles have the drug sustained release effects in vivo. The LNP–Fer‐1 nanoparticles were composed of cholesterol, hydrogenated soybean phosphatidylcholine (HSPC), 1,2‐distearoyl‐sn‐glycero‐3‐phosphoethanolamine‐N‐[methoxy(polyethylene glycol)‐2000] (DSPE–mPEG2000), and DSPE–mPEG2000–galactose. (B) Schematic illustration of macrophage‐targeting lipid nanoparticles (LNP–Fer‐1) therapy in atherosclerotic ApoE–/– mice. Eight‐week‐old ApoE–

/– mice received the tail vein injection of saline, LNP, Fer‐1, and LNP–Fer‐1 (twice for 1 week; each time 1 mg/kg calculated by Fer‐1; other groups with equal volume) for 6 weeks after 12 weeks of atherosclerotic modeling (n = 5 mice in each group). (C) Hematoxylin and eosin (H&E) staining of aortic plaques in saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group). Scale bar, 200 µm. (D) Lesion area of aortic plaques across a 400 µm distance of the aortic root in saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group). (E) Necrotic core areas of aortic plaques across a 400 µm distance of the aortic root in saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group). (F) The proportion of early, moderate, and advanced plaques in saline, Fer‐1, LNP, and LNP–Fer‐1 groups based on the histological staining (n = 5 mice in each group). (G) Quantification of the smooth muscle cells (α‐SMA) area for the aortic plaques in saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group, at least 2 sections per mouse). (H) Quantification of Masson trichrome staining for aortic plaques in saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group, at least 2 sections per mouse). (I, J, and K) Glutathione (GSH), glutathione peroxidase (GSH‐PX), and malondialdehyde (MDA) levels of serum in saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group). (L) Quantification of 4‐hydroxynonenal (4‐HNE) fluorescence intensity in the aortic plaques from saline, Fer‐1, LNP, and LNP–Fer‐1 groups. (n = 5 mice in each group, at least 2 sections per mouse). (M) Quantification of macrophages numbers (Mac‐3) for the aortic plaques in saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group, at least 2 sections per mouse). (N) Classical M1 marker gene expressions from aorta arteries of saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group). (O) Alternative M2 marker gene expressions from aorta arteries of saline, Fer‐1, LNP, and LNP–Fer‐1 groups (n = 5 mice in each group). One‐way ANOVA test was used to compare multiple groups. *p < 0.05, **p < 0.01, **p < 0.001.

Next, we investigated whether Fer‐1 suppressed atherogenesis by affecting macrophage infiltration and polarization. To evaluate macrophages infiltration in aortic plaque, immunofluorescence staining of Mac‐3 showed that LNP–Fer‐1 treatment effectively reduced the number of macrophages compared with mice receiving Fer‐1, LNP, or saline (Figures 6M and S7D). To evaluate macrophages polarization, the mRNA expressions of M1/M2 markers in aorta root were measured. LNP–Fer‐1 treatment reduced the M1 markers expression (Tnfa, Nos2, Il6, and Il1b; Figure 6N) and increased the M2 markers expression (Arg1, Mrc1, Irf4, Retnla; Figure 6O) in aorta root when compared with mice receiving Fer‐1, LNP, or saline. Taken together, these results indicate that macrophage‐targeting LNP–Fer‐1 alleviate atherosclerosis development through modulating macrophage polarization.

Discussion

3

Emerging evidence demonstrated the important role of amino acid metabolism in CVDs [10, 11, 12]. The primary function of cystine/glutamate antiporter Slc7a11 is the maintenance of redox homeostasis via balancing glutamate and cystine transport and GSH synthesis [14, 27]. However, the function of Slc7a11 in atherosclerosis development is largely unknown. Here, we provide an innovative perspective that Slc7a11 acts as a modulator of macrophage polarization and ameliorates atherosclerosis. We found that Slc7a11 expressions were upregulated by oxLDL, and specifically enhanced in the macrophages of atherosclerotic plaques. Macrophage‐specific Slc7a11 overexpression attenuates atherosclerotic lesions and increases the plaque stability. Mechanistically, Slc7a11 expression inhibits the classically activated macrophage (M1) polarization by reducing Stat1 phosphorylation, and promotes alternatively activated macrophage (M2) polarization by enhancing Stat6 phosphorylation. Slc7a11‐mediated cystine uptake and GSH synthesis promote macrophage polarization from M1 to M2. Macrophage‐targeting LNPs loading with Fer‐1, an antioxidant reagent promotes GSH synthesis, also ameliorates the progression of atherosclerosis. These findings reveal a critical role of Slc7a11 in the phenotypic switch of macrophage and atherosclerosis progression.

The connection between Slc7a11, as well as cystine and glutamate, and CVDs has been mentioned in previous studies performed in human and animal models. For example, coronary artery disease patients displayed increased circulating glutamate concentration [32]. Slc7a11 overexpression in cardiomyocytes improved ejection fraction, fractional shortening, and cardiomegaly caused by Fth‐deficiency via preventing cardiac ferroptosis [15]. The influence of ferroptosis on atherosclerosis remains to be investigate. It is regarded that ferroptosis mainly occurred in advanced atherosclerotic plaques [33]. However, the role of ferroptosis in early atherosclerotic plaques is not clear. Therefore, different from Slc7a11‐mediated ferroptosis in advanced atherosclerotic plaques, Slc7a11‐mediated amino acid metabolism and macrophage polarization may prevent the development of atherosclerosis. Emerging evidence highlights the critical role of amino acid metabolism during atherogenesis [9]. Reduced glutamine uptake in macrophages was reported to accelerate atherosclerotic progression, with a concurrent expansion of inflammatory macrophage subsets [34]. Mice receiving homoarginine displayed reduced atherosclerosis in the aortic root and brachiocephalic trunk via modulating actin cytoskeleton in CD4^+^ T cells [35].

Mechanistically, Slc7a11‐mediated amino acid metabolism modulates plaque microenvironment, which influences macrophage phenotypic plasticity and controls atherosclerosis progression and regression [8, 36]. Slc7a11 overexpression significantly decreased macrophage infiltration in the atherosclerotic plaques. In addition, mice with macrophage‐specific Slc7a11 overexpression attenuated atherosclerotic plaques and increased plaque stability through reprogramming macrophage polarization from M1 to M2. M2 macrophages decline with the growth of lesions, but M1 macrophages and inflammatory markers continue to increase [37, 38]. Slc7a11 overexpression raised the M2 markers expression and decreased the M1 markers expression in the mice aortic artery tissues. Slc7a11‐mediated cysteine transport and GSH synthesis inhibited Stat1 phosphorylation and M1 activation, but promote Stat6 phosphorylation and M2 polarization. M2 macrophages are assumed to secrete collagen and clear the apoptotic cells to stabilize the atherosclerotic lesions in the early stage [4]. ROS in macrophages triggers GSH synthesis to buffer intracellular oxidative stress [39]. Emerging evidence suggests that ROS function as pivotal secondary messengers modulating M1/M2 macrophages polarization [40, 41]. Previous studies have shown that macrophage‐specific Slc7a11 knockout promoted systemic inflammatory response of macrophages [42, 43]. In the aortic artery tissues from *ApoE^–/–^Slc7a11^MOE^

• mice, higher early plaques and less advanced plaques rather than severely calcified late plaques were seen, supporting previous findings that Slc7a11 is a primary protective factor against vascular calcification [16, 44].

Changes in the metabolism of specific amino acids are among the most significant and early traits that distinguish different subsets of macrophages [45]. Metabolism alterations are intimately linked to macrophage activation [46]. Arginine metabolism promoted M2 protumor polarization of TAMs, thereby suppressing antitumor immunity and enhancing breast cancer progression [47]. Sarcosine treatment upregulated the expression of anti‐inflammatory macrophage markers through the activation of GCN2–ATF4 signaling pathway in IL‐4‐stimulated BMDMs [48]. However, the precise role of Slc7a11‐mediated cystine/glutamate metabolism in macrophage polarization is still poorly understood. In contrast to our findings that Slc7a11‐mediated cystine uptake and GSH synthesis promoted the phenotypic switch of macrophage from M1 to M2, a previous study reported that GSH promoted LPS‐induced proinflammatory IL‐1β production in peritoneal macrophages [49]. It suggested a tissue‐specific macrophage activation pattern may exist, that is, macrophage phenotype switch was influenced by oxidative environment, energy metabolism, or mitochondrial function [50].

Macrophage‐targeted nanodelivery systems have emerged as promising strategies for atherosclerosis treatment [51, 52]. In our study, macrophage‐specific Slc7a11 overexpression or LNPs boosting GSH suppressed atherosclerosis development by inhibiting inflammation, independent of changes in blood lipid levels. According to current clinical guidelines, even when LDL levels with statin treatment are lower than clinical practice guideline thresholds, the residual risk of atherosclerosis remains [53]. Residual inflammatory risk appears to be more strongly associated with future cardiovascular events than residual cholesterol risk [54]. Low‐dose colchicine, combined with modern antiplatelet medications and lipid‐lowering therapy, improved both composite and individual cardiovascular outcomes in individuals with coronary disease [55, 56]. It suggests that a combination of active lipid‐lowering and anti‐inflammation strategies could further reduce atherosclerosis risk. Macrophage‐specific Slc7a11 overexpression or a LNP has the potential to be a novel anti‐inflammatory strategy for ASCVD. Furthermore, our findings identify Slc7a11 as a promising novel therapeutic target for inflammatory diseases. Enhanced Slc7a11‐mediated amino acid transport might alleviate inflammation‐driven diseases. Future research should focus on screening for potent Slc7a11 activators and evaluating their efficacy and potential side effects in relevant preclinical models.

In conclusion, this study highlights the significant role of Slc7a11‐mediated amino acid metabolism in the development of atherosclerosis (Figure 7). Targeting macrophage Slc7a11 or modulating Slc7a11‐mediated amino acid metabolism is a novel therapeutic strategy for the treatment of ASCVDs.

Slc7a11‐mediated cystine/glutamate antiport regulates macrophage polarization and atherosclerosis development. Slc7a11‐mediated cystine uptake and glutathione synthesis inhibits the classically activated macrophage polarization (M1) by reducing the Stat1 phosphorylation and classical M1 gene expressions (Tnfa, Nos2, Il6, Il1b), and promotes alternatively activated macrophage polarization (M2) by enhancing Stat6 phosphorylation and alternative M2 gene expressions (Arg1, Mrc1, Irf4, Retnla). In contrast, glutamate treatment or glutathione depletion reverses the phenotypic switch of macrophage from M1 to M2. Macrophage‐specific Slc7a11 overexpression or macrophage‐targeting lipid nanoparticles loading with ferrostatin‐1 (LNP–Fer‐1) promotes M2 macrophage polarization, decreases atherosclerotic macrophage content, and attenuates the development of atherosclerosis.

Materials and Methods

4

Mice

4.1

C57BL/6J wildtype mice and apolipoprotein E‐deficient mice (*ApoE^–/–^ *, C57BL/6J background) were obtained from Beijing Vital River Laboratory Animal Technology (Beijing, China). Slc7a11 gene knockout mice were obtained from GemPharmatech (*Slc7a11^–/–^ *, C57BL/6J background, Cat. No. T015610, China). Slc7a11 conditional overexpression mice (*Slc7a11^LoxP‐STOP‐LoxP^ *) were reported previously [15, 57]. *Slc7a11^LoxP‐STOP‐LoxP^

• mice were bred with the LysM‐Cre transgenic mice to generate the macrophage‐specific Slc7a11 overexpressing mice (*Slc7a11^LoxP‐STOP‐LoxP^ *; LysM‐Cre, *Slc7a11^MOE^ *). Then, the *Slc7a11^MOE^
• mice were bred with the *ApoE^–/–^
• mice to generate *ApoE^–/–^ Slc7a11^MOE^
• mice. Mice were housed in the ventilated cages under the specific‐pathogen‐free conditions (temperature: 20–26°C; humidity: 40–70%; pressure: 45 Pa; animal illumination: 15–20 Lux; light: 12 h/12 h light/dark cycle) and had access to food and water ad libitum. Mice were fed a western diet (Research Diets; Cat. No. D12079B, USA) containing 41% fat, 43% carbohydrates, and 17% protein with an energy density of 4.7 kcal/g. To establish the atherosclerosis model, 8‐week‐old male *ApoE^–^

^/–^ and *ApoE^–^ * ^/–^ *Slc7a11^MOE^

• mice were fed a western diet (Research Diets; Cat. No. D12079B) for 16 weeks (10 mice per group). Mice were sacrificed using tribromoethanol (250 mg/kg).

Gene Ontology and Kyoto Encyclopedia of Genes and Genomes Enrichment Analysis

4.2

Genes that met the cutoff criterion of fold change > 1.25 or fold change < 0.8 with p value < 0.05 were identified as differentially expressed genes between foamy macrophages and nonfoamy macrophages. The differentially expressed genes name list was imported into the Database for Annotation, Visualization, and Integrated Discovery (https://davidbioinformatics.nih.gov/) and limited the species to Mus musculus. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment analysis were performed to enrich the critical signaling pathways. p Value < 0.05 was set as the cutoff criteria. A heatmap was drawn using GraphPad Prism software (version 8.0, USA) to display differentially expressed genes in the cellular response to oxidative stress, ferroptosis, GSH metabolic process, and amino acid transmembrane transport pathways.

Induction of Primary BMDMs

4.3

Mouse BMDMs were isolated and induced as previously described [58]. Eight‐week‐old mice were euthanized using tribromoethanol (250 mg/kg), and the leg bones were washed with 75% alcohol and cold phosphate‐buffered saline (PBS) in sequence. Bone marrow flushed from mouse femurs and tibias were dispersed in cold DMEM. Cell suspension was filtered through a 70 µm cell strainer, and then centrifuged at 300*×g* for 10 min. Bone marrow cells were resuspended in the conditioned DMEM (10% FBS, 1% penicillin–streptomycin solution) supplemented with 30% (vol/vol) supernatants L929 mouse fibroblasts medium, which provides macrophage colony‐stimulating factor. Next, cell suspension was plated onto 10 cm plates and cultured at 37°C in a 5% CO_2_ incubator. On Day 7, the induced BMDMs were digested and plated onto multiple well plates overnight with the fresh DMEM (10% FBS, 1% penicillin–streptomycin solution).

Cell Treatments

4.4

For the experiments of oxLDL stimulation, the BMDMs were treated with oxLDL (Yiyuan Biotech; Cat. No. YB‐002, China) at indicated times. For the experiment of Nrf2 inhibitor, the BMDMs were treated with 50 µg/mL oxLDL for 24 h in the presence of 2 µM Nrf2‐inhibitor Nrf‐IN‐1 (TargetMol; Cat. No. 1610022‐76‐8, China).

For the classically activated macrophages (M1) activation, the attached BMDMs were stimulated with LPS (1 µg/mL; Solarbio; Cat. No. L8880, China) and murine IFN‐γ (50 ng/mL; PeproTech, Cat. No. 315–05–100 µg, USA). For the alternatively activated macrophages (M2) activation, the attached BMDMs were stimulated with murine interleukin‐4 (IL‐4, 30 ng/mL; PeproTech; Cat. No. 214–14–20 µg).

For cystine deprivation, cells were cultured with DMEM containing no cystine, methionine, and glutamine (Gibco; Cat. No. 21013024, USA) with 10% FBS and 1% penicillin–streptomycin solution. Meanwhile, the above culture medium was added with the additional methionine (final concentration is 200 µM; Solarbio; Cat. No. M0010) and glutamine (final concentration is 4 mM; Solarbio; Cat. No. G8230).

Following a 6‐h cystine starvation, BMDMs were stimulated to induce M1 or M2 polarization.

For glutamate uptake, cells were cultured with DMEM containing glutamate (600 µM; Sigma–Aldrich; Cat. No. G8415, USA) with 10% FBS and 1% penicillin–streptomycin solution. Following a 6‐h glutamate treatment, BMDMs were stimulated to induce M1 or M2 polarization.

For GSH uptake, cells were cultured with DMEM containing GSH (1 mM; Solarbio; Cat. No. G0010) with 10% FBS and 1% penicillin–streptomycin solution. Following a 6‐h GSH treatment, BMDMs were stimulated to induce M1 or M2 polarization.

For BMDMs treated with Fer‐1 (2 µM; Selleck; Cat. No. 347174‐05‐4, USA), cells were cultured with DMEM with 10% FBS and 1% penicillin–streptomycin solution. Following a 1‐h Fer‐1 treatment, BMDMs were stimulated to induce M1 or M2 polarization.

The treatment time and reagent dosages were detailed in corresponding figure legends.

Flow Cytometry

4.5

Primary BMDMs were treated with oxLDL for 36 and 48 h. After oxLDL treatment, BMDMs were centrifugated at 60 ×g for 5 min. Suspension cells were incubated with Alexa‐488 labeled Slc7a11 antibody (Novus; Cat. No. NB300‐318AF488, USA) in the dark for 30 min at 4°C. After incubation with primary antibody, cells were incubated with a secondary antibody conjugated to fluorescein isothiocyanate for 1 h. Fluorescence intensity was detected by the FACSFortessa or FACSAira III flow cytometer (BD, USA). Results were analyzed using the FlowJo software (version 10, USA).

RNA Extraction and Real‐Time Quantitative PCR

4.6

Total RNA was extracted from BMDMs or mice aortas using Trizol reagent (Invitrogen; Cat. No. 15596026CN, USA). BMDMs were repeatedly blown using a Trizol reagent and collected in the centrifuge tube. Mice aortas in the Trizol reagent were pulverized using an automatic freeze‐grinding device, and centrifugated at 6000 ×g for 10 min. Tissue supernatant was used to extract total RNA. Then, the collected cells from BMDMs or mice aortas were vortexed for 1 min with chloroform and centrifuged at 1.3 × 10^4^ ×g for 15 min at 4°C. Next, the supernatant was precipitated with isopropanol for 10 min and centrifuged at 1.3 × 10^4^ ×g for 10 min. Exacted RNA was washed with 75% (vol/vol) ethanol twice and dissolved in the nuclease‐free water. Extracted RNA (100–1000 ng) was reverse‐transcribed using the HiScript III RT SuperMix kit (Vazyme; Cat. No. R323‐01, China). Real‐time quantitative PCR (qPCR) was performed using Taq Pro Universal SYBR qPCR Master Mix (Vazyme; Cat. No. Q712‐02) on the QuantStudio 5 real‐time PCR system (Applied Biosystems, USA). The relative gene expression was calculated using the 2 (−DeltaDeltaC(T)) method. Real‐time qPCR primer sequences were provided in Table S1.

Western Blotting

4.7

Primary BMDMs cultured in six‐well plates were washed with cold PBS, and then were lysed using 200 µL per well of sodium dodecyl sulfate polyacrylamide gel electrophoresis loading buffer (with dithiothreitol; Solarbio; Cat. No. P1040). Next, lysed cells were mechanically removed using cell scrapers, and were heated at 4°C for 15 min to denature protein. Protein concentration was measured using the BCA Protein Assay Kit (Solarbio; Cat. No. PC0020). BMDMs protein was fractioned in the 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis and was transferred to 0.45 µm polyvinylidene difluoride membranes (Millipore; Cat. No. IPFL00010, USA). The membranes were blocked with 5% skimmed milk (Yili, China) for 2 h, and then incubated with corresponding antibodies at 4°C overnight. After being washed with tris‐buffered saline containing 0.1% Tween‐20 (Solarbio; Cat. No. T1081; Tween‐20; Solarbio; Cat. No. T8220), the membranes were probed with the horseradish peroxidase‐conjugated secondary antibodies for 1 h. Blots on membranes were detected by an ECL system (ChemiScope 6100; Qinke, China). Western blotting images were quantified by using Image J software (version 1.8.0).

The following primary antibodies were diluted at 1:1000: Slc7a11 (Abcam; Cat. No. ab37185, UK), Stat1 (Cell Signaling Technology; Cat. No. 9172S, USA), p‐Stat1 (Invitrogen; Cat. No. 44–376G), Stat6 (Abcam; Cat. No. ab32520), p‐Stat6 (Abcam; Cat. No. ab263947), and β‐actin (ABclonal; Cat. No. AC004, China).

Serum Biochemical Measurements

4.8

Whole blood was acquired through heart puncture and collected in a coagulant tube for serum separation. The serum was transferred to a new centrifuge tube after centrifugation at 800 ×g for 15 min at room temperature. Serum TG, TC, LDL‐C, HDL‐C, creatine kinase (CK), CK‐myocardial band, lactate dehydrogenase, lactate dehydrogenase 1, aspartate aminotransferase, alkaline phosphatase, creatinine, and uric acid were measured using the automated clinical analyzer (Roche; cobas 8000, Germany).

Serum Oxidative Indicators Measurements

4.9

Serum MDA and GSH concentrations were measured using a colorimetric assay according to the commercial kits (Jiancheng Bio; Cat. No. A003‐1‐1, Cat. No. A006‐2‐1, China). Serum GSH‐PX was measured using the corresponding commercial kits (Jiancheng Bio, Cat. No. A005‐1‐2).

Histological Analysis of Atherosclerotic Lesions

4.10

Mice aortic tissues were fixed in the 4% paraformaldehyde (Solarbio; Cat. No. P1110) for 48 h. Fixed tissue samples were dehydrated in a graded series of ethanol and embedded in the paraffin. Then, the aortic paraffin block was serially sectioned at 5 µm thickness. Mice aortic sections were deparaffinized in xylene, rehydrated with ethanol, and then used for subsequent analysis.

To observe the atherosclerotic lesions, the aortic sections were stained using the commercial hematoxylin and eosin staining kit (H&E Solarbio; Cat. No. G1120). To further measure the collagen deposits in plaques, the aortic sections were stained using the commercial Masson staining kit (Solarbio; Cat. No. G1340). To quantify the lesion area and necrotic area in the aortic root, H&E staining images of the aortic sinus within 400 µm were analyzed by Image J software (version 1.8.0). To classify the stage of the atherosclerotic plaques, H&E staining images were divided into the early, moderate, and advanced stages according to the Stary method [59]. Early‐stage plaques were characterized by typical fatty streaks and fibrosis. Moderate plaques exhibited foam cells and cholesterol clefts. Advanced plaques were defined by extensive foam cells, cholesterol clefts, multifocal necrosis, and mineralization.

Immunofluorescence Staining

4.11

Deparaffinized and rehydrated sections underwent microwave‐mediated antigen retrieval in citrate buffer solution (pH 6.0; ZSGB‐BIO; Cat. No. ZLI‐9065, China). Blocking with immunol staining blocking buffer (Beyotime; Cat. No. P0102, China), the sections were incubated with the primary antibodies overnight at 4°C. Indirect immunofluorescence was conducted by incubation with the specific fluorescence‐labeled secondary antibodies under humidified conditions for 1 h. After incubation, cell nuclei were counterstained with 4,6‐diamidino‐2‐phenylindole (DAPI; Solarbio; Cat. No. S2110). At the end of experiments, immunofluorescence staining images were acquired under a LSM780 confocal laser scanning microscope (Zeiss, Germany).

The primary and secondary antibodies were as follows: α‐SMA (1:100 dilution; Cell Signaling Technology; Cat. No. 48938), Mac3 (1:100 dilution; Biolegend; Cat. No. 108502, USA), CD68 (1:100 dilution; Thermo Fisher Scientific; Cat. No. 14‐0688‐82, USA), 4‐HNE (1:100 dilution; Abcam; Cat. No. ab46545), goat anti‐mouse IgG H&L (Alexa Fluor 488) (1:300 dilution; Abcam; Cat. No. ab150113, USA), goat anti‐rabbit IgG H&L (Alexa Fluor 488) (1:300 dilution; Abcam; Cat. No. ab150007), goat anti‐mouse IgG H&L (Alexa Fluor 594) (1:300 dilution; Abcam; Cat. No. ab150116), and goat anti‐rabbit IgG H&L (Alexa Fluor 594) (1:300 dilution; Abcam; Cat. No. ab150080).

Preparation of LNPs

4.12

LNPs loading with the Fer‐1 were prepared using a thin‐film hydration method [31, 60]. Briefly, Fer‐1 was prepared as a 10 mg/mL ethanol storage solution. Next, the cholesterol (Avanti; Cat. No. 57‐88‐5, USA), HSPC (Avanti; Cat. No. 97281‐48‐6), DSPE–mPEG2000 (Avanti; Cat. No. 474922‐26‐4), and DSPE–mPEG2000–galactose at a molar ratio 1:3:0.5:0.5 were mixed with Fer‐1 solution in a round‐bottom flask. After the mixture was dried at 40°C for 15 min using a vacuum rotary evaporator, a thin film was formed at the bottom of the flask. The thin film was hydrated with 1 mL PBS at 40°C for 20 min. Subsequently, the dissolved solution was repeatedly extruded overnight at 4°C using an Avanti extruder (61000; Avanti) after sonication, forming LNPs loading with Fer‐1 (LNP–Fer‐1). LNP–Fer‐1 nanoparticles were rapidly dissolved and stored at 4°C after dialyzed and lyophilized.

Characterization of LNPs

4.13

The morphology of LNP–Fer‐1 was characterized by a transmission electron microscope (H‐7500; Hitachi, Japan). Briefly, 10 µg samples were put onto copper grids. After deposition, the excess fluid was removed with filter paper. Next, 10 µL of 1% uranyl acetate dihydrate was applied to the grids for 1 min. The LNPs images were captured by the transmission electron microscope. UV–vis spectra from 200 nm to 600 nm of LNPs were measured by a UV spectrophotometer (UV‐1780; Shimadzu, Japan). The distribution and zeta potentials of LNPs were measured by dynamic light scattering (Zetasizer Pro, Malvern, UK).

Determination of the Encapsulation Efficiency and DLE of LNPs

4.14

To determine the concentration of Fer‐1 in the nanoparticles, LNP–Fer‐1 nanoparticles were dissolved in the triton X‐100, and Fer‐1 content was measured using HPLC (1260 Infinity II; Agilent, USA) with a ZORBAX Eclipse Pluse C18 column (3.5 µm, 4.6 mm × 100 mm). Encapsulation efficiency was calculated by the following formula: encapsulation efficiency (%) = amount of Fer‐1 in nanoparticles/total amount of Fer‐1 used × 100%. To measure the release of Fer‐1 from nanoparticles, 1 mL samples were added into the dialysis device and immersed in the 48 mL of releases medium in the acceptor chamber. The dialysis device was maintained at 37°C with the shaking at 1.0 × 10^−4^ ×g. The aliquots were frozen and lyophilized. After exacting Fer‐1 with triton X‐100, the concentration was measured using HPLC. DLE was calculated by the following formula: DLE (%) = weight of Fer‐1 in nanoparticles/total weight of nanoparticles containing Fer‐1 × 100%.

Macrophages‐Targeting Capacity of LNPs

4.15

Primary BMDMs were grown on chamber slides in six‐well plate. To determine the macrophages‐targeting capacity of LNPs, 100 µL of Cy5.5 labeled LNP–Fer‐1 was incubated with the BMDMs for 6 h. After incubation, BMDMs were blocked with immunol staining blocking buffer (Beyotime; Cat. No. P0102, China). Then, cells were incubated with Mac‐3 antibody (1:100 dilution; Biolegend; Cat. No. 108502). Subsequently, cells were incubated with Alexa‐488 (1:300 dilution; Abcam; Cat. No. ab150113) for immunofluorescence detection. The antifading mounting medium with DAPI was dropped onto the chamber slides. Immunofluorescence images were acquired using a confocal laser scanning microscope.

Statistical Analysis

4.16

The results were presented as mean ± standard deviation. The data were statistically analyzed and displayed using GraphPad Prism software (version 8.0, USA). Data with multiple comparisons were analyzed by the one‐way analysis of variance. Data with two groups comparison were analyzed by two‐tailed unpaired t‐test. Differences with a p value below 0.05 were considered as statistically significant. The specific statistical information was presented in each figure legend.

Author Contributions

Fudi Wang, Hao Wang, Peng An, Yongting Luo, and Junjie Luo designed the study and directed the research. Shuaishuai Zhou, Siyue Li, Baixue Liu, Wen Shao, and Jin Tao carried out experiments. Yongting Luo, Junjie Luo, Jingyi Qi, Chang Fan, and Jiaxin Shi conducted the statistical analyses. Shuaishuai Zhou, Peng An, and Hao Wang drafted the manuscript. Fudi Wang, Hao Wang, Peng An, Yongting Luo, and Junjie Luo revised the manuscript. All authors have read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (32571359 and 32371229 to P.A., 32171171 to H.W, 82470442 to Y.L., and 32570908 to J. L.), the Henan Natural Science Foundation (242300421096 to H.W.), the State Key Laboratory of Cardiovascular Disease, Fuwai Hospital, Chinese Academy of Medical Sciences (2024GZkf‐05 to J.L.), the Beijing Natural Science Foundation (7262078 to Y.L.), the Pinduoduo‐China Agricultural University Research Fund (PC2023B01014 to J. L.), and the support of the 2115 Talent Development Program of China Agricultural University.

Conflicts of Interest

The authors declare no conflicts of interest.

Ethics Statement

All animal experimental procedures were performed in accordance with Guide for the Care and Use of Laboratory Animal and were approved by the Committee on the Ethics of Animal Experiments of China Agricultural University (No. AW92503202‐5‐1).

Supporting information

Figure S1: Oxidized low‐density lipoprotein (OxLDL) upregulates Slc7a11 protein in macrophages (related to Figure 1). Figure S2: Representative immunofluorescence images of Slc7a11 negative control staining in the atherosclerotic plaques from *ApoE^–/–^

• mice fed a western diet for 16 weeks (related to Figure 1). Figure S3: Macrophage‐specific Slc7a11 overexpression had no effect on the body weight and serum lipid profiles (related to Figure 2). Figure S4: Ferrostatin‐1 (Fer‐1) treatment inhibits classical M1 polarization and promotes alternative M2 polarization in bone marrow‐derived macrophages (BMDMs) (related to Figure 6). Figure S5: Characterization of macrophage‐targeting lipid nanoparticles loading with ferrostatin‐1 (LNP–Fer1) (related to Figure 6). Figure S6: The effect of lipid nanoparticles loading with Ferrostatin‐1 (LNP‐Fer‐1) on serum parameters (related to Figure 6). Figure S7: Lipid nanoparticles loading with Ferrostatin‐1 (LNP‐Fer‐1) increases the plaque stability (related to Figure 6). Table S1: Primers for real‐time quantitative PCR analysis.