Platelets impair the resolution of inflammation in atherosclerotic plaques in insulin-resistant mice after lipid lowering

Abstract

Insulin resistance impairs benefits of lipid-lowering treatment, as evidenced by higher cardiovascular disease risk in individuals with type 2 diabetes versus those without. Because platelet activity is higher in insulin-resistant patients and promotes atherosclerosis progression, we questioned whether platelets impair inflammation resolution in plaques during lipid lowering. In mice with obesity and insulin resistance, we induced advanced plaques and then implemented lipid lowering to promote atherosclerotic plaque inflammation resolution. Concurrently, mice were treated with either platelet-depleting or control antibodies for 3 weeks. Platelet activation and insulin resistance were unaffected by lipid lowering. Both antibody-treated groups showed reduced plaque macrophages, but plaque cellular and structural composition differed. In platelet-depleted mice, single-cell RNA-seq revealed dampened inflammatory gene expression in plaque macrophages and an expansion of a subset of Fcgr4+ macrophages having features of inflammation-resolving, phagocytic cells. Necrotic core size was smaller and collagen content greater, resembling stable human plaques. Consistent with the mouse results, clinical data showed that patients with lower platelet counts had decreased proinflammatory signaling pathways in circulating nonclassical monocytes after lipid lowering. These findings highlight that platelets hinder inflammation resolution in atherosclerosis during lipid-lowering treatment. Identifying novel platelet-targeted therapies following lipid-lowering treatment in individuals with insulin resistance may be a promising therapeutic approach to promote atherosclerotic plaque inflammation resolution.

Keywords: Atherosclerosis; Cardiology; Diabetes; Inflammation; Platelets; Vascular biology.

Figure 1. Platelet activation and impaired glucose tolerance are sustained during the lipid-lowering treatment period in obese Ldlr^–/– male mice.

(A) Study design. Eight-week-old Ldlr^–/– male mice fed a high-fat, high-cholesterol (HFHC) diet for the whole duration of the study (23 weeks). After the atherosclerosis progression period (obese baseline, Ob), mice were injected with ApoB ASO to lower lipids (lipid-lowering period). All lipid-lowered mice were split into 2 groups, injected either with isotype control antibody (IgG) or platelet depletion antibody (αCD42b) every 3 days for a total of 3 weeks. (B) Platelet counts in circulating blood of Ob, IgG, and αCD42b 3 days before harvest. (C) Blood glucose measurements (lean n = 16, Ob n = 47) and (D) area of curve (AOC) quantification of glucose tolerance tests (GTTs) from lean, Ob, IgG, and αCD42b mice after 16 weeks of HFHC diet during atherosclerosis progression. (E) Blood glucose measurements (Ob n = 3, IgG n = 4, αCD42b n = 4) and (F) AOC quantification of GTTs from Ob, IgG, and αCD42b mice after 22 weeks of HFHC diet during atherosclerosis resolution. Mean fluorescent intensity (MFI) of platelet (G) JonA and (H) P-selectin in nonstimulated (NS) samples and upon 100 μM PAR4-AP agonist stimulation of lean and Ob circulating blood samples. (I) Mean platelet volume (MPV) from circulating blood of lean, Ob, and IgG-treated mice, 3 days before harvest. In C and E, error bars represent SD. Data were analyzed by Kruskal-Wallis with Dunn’s post hoc test (B and F), unpaired Mann-Whitney test (C, D, and H), unpaired Student’s t test (G), or ordinary 1-way ANOVA with Tukey’s multiple-comparison test (I). P values are shown in graphs.

Figure 2. Platelet deficiency alters plaque composition after lipid lowering.

(A) Representative images (scale bars: 0.5 mm) and (B) quantification of immunohistochemical staining for CD68 content (%) in aortic root plaques of Ob, IgG, and αCD42b mice. (C) Representative images (scale bars: 0.5 mm) of polarized light and (D) quantification of collagen content (%) in aortic root plaques of Ob, IgG, and αCD42b mice. Data in B and D were analyzed by 1-way ANOVA with Šídák’s post hoc test. P values are shown in graphs. Dotted lines in C outline the plaques.

Figure 3. CD45⁺ scRNA-seq shows that loss of platelets changes the phenotype of plaque myeloid cells into a more proresolving state after lipid lowering.

(A) UMAP embedding of plaque leukocytes showing 20 identified cell types. Μacs, macrophages. (B) Percentages of macrophage subtypes identified in IgG and αCD42b macrophages. *P = 0.01 (foamy macrophages), *P = 0.02 (Fcgr4⁺ macrophages), using the t-test function in the R package propeller. (C) Volcano plot showing results of a pseudobulk differential expression analysis of αCD42b relative to IgG samples in the Fcgr4⁺ macrophage cluster. Significantly differentially expressed genes (DEGs) are highlighted in the plot. (D) Overrepresented Gene Ontology (GO) terms (q < 0.05) found among significantly downregulated genes in αCD42b relative to IgG samples in Fcgr4⁺ macrophage cluster. (E) Normalized scaled expression of genes that are associated with overrepresented GO terms shown in D. All genes are significantly downregulated in αCD42b-treated mice relative to IgG-treated mice in Fcgr4⁺ macrophage cluster. (F) Volcano plot showing results of a pseudobulk differential expression analysis of αCD42b relative to IgG samples in foamy macrophage 1 cluster. Significantly DEGs are highlighted in the plot. (G) Overrepresented GO terms (q < 0.05) found among significantly downregulated genes in αCD42b relative to IgG samples in foamy macrophage 1 cluster. (H) Normalized scaled expression of genes that are associated with overrepresented GO terms shown in G. All genes are significantly downregulated in αCD42b-treated mice relative to IgG-treated mice in foamy macrophage 1 cluster. In C, E, F, and H, P < 0.05 and |log₂(fold change)| ≥ 0.6 were used.

Figure 4. In plaques after lipid lowering, platelet depletion leads to increased abundance of FCGR4⁺ macrophages and reduces necrotic core size.

(A) Representative images (scale bars: 100 μm) of portions of the aortic root and (B) quantification of FCGR4⁺ content (%) in total aortic root plaques of IgG- and αCD42b-treated mice. (C) Quantification of normalized FCGR4⁺ integrated density of IgG- and αCD42b-treated mice. (D) Representative images (scale bars: 0.5 mm); note that this is the same representative image as shown in Figure 2A, but now with examples of necrotic areas shown (dotted outlines), and (E) quantification of necrotic core content (%) of IgG- and αCD42b-treated mice. (F) Negative (r = –0.7247) correlation analysis between FCGR4⁺ area and necrotic core content from IgG- (blue circles) and αCD42b-treated (pink squares) mice. Data in B and E were analyzed by an unpaired Student’s t test. Data in C were analyzed by unpaired Mann-Whitney test. Data in F were analyzed by a simple linear regression (R² = 0.5250). P values shown in graphs.

Figure 5. Individuals with T2D and lower platelet counts undergoing lipid-lowering treatment have decreased proinflammatory signaling pathways in circulating nonclassical monocytes.

(A) CHORD study design. scRNA-seq was performed in PBMCs at follow-up. (B) Platelet counts at follow-up from individuals with T2D following lipid-lowering treatments, stratified into the top and bottom third tertile. (C) Top 15 downregulated KEGG pathways of nonclassical monocytes in individuals with low (n = 3) compared to high (n = 3) platelet counts from the CHORD study. Data in B were analyzed by an unpaired Student’s t test.