Hyperglycemia promotes myelopoiesis and impairs the resolution of atherosclerosis

Abstract

Diabetes is a major risk factor for atherosclerosis. Although atherosclerosis is initiated by deposition of cholesterol-rich lipoproteins in the artery wall, the entry of inflammatory leukocytes into lesions fuels disease progression and impairs resolution. We show that diabetic mice have increased numbers of circulating neutrophils and Ly6-C(hi) monocytes, reflecting hyperglycemia-induced proliferation and expansion of bone marrow myeloid progenitors and release of monocytes into the circulation. Increased neutrophil production of S100A8/S100A9, and its subsequent interaction with the receptor for advanced glycation end products on common myeloid progenitor cells, leads to enhanced myelopoiesis. Treatment of hyperglycemia reduces monocytosis, entry of monocytes into atherosclerotic lesions, and promotes regression. In patients with type 1 diabetes, plasma S100A8/S100A9 levels correlate with leukocyte counts and coronary artery disease. Thus, hyperglycemia drives myelopoiesis and promotes atherogenesis in diabetes.

Figure 1

Leukocytosis develops in response to hyperglycemia via expansion and proliferation of BM progenitor cells. Chow fed non-diabetic WT (C57BL/6J), STZ-diabetic and Akita-diabetic mice were treated with a SGLT2i (5mg/kg; ISIS) in the drinking water for 4 wks. Representative flow cytometry plots of blood leukocyte subsets from (A) STZ-diabetic and (B) Akita-diabetic mice. (C and D) Quantification of monocyte subsets and neutrophils in (C) STZ-diabetic and (D) Akita-diabetic mice. (E-H) HSPC, CMP and GMP analysis in the BM: The percentage of the respective populations in (E) STZ-diabetic and (F) Akita-diabetic mice, and cell cycle (G₂M phase) analysis in (G) STZ-diabetic and (H) Akita-diabetic mice was performed by flow cytometry. All experiments n=10-12/group. *P<0.05 vs. all groups and ^P<0.05 vs. WT+STZ or Akita respectively. All values are means ± SEM.

Figure 2

Neutrophil-derived S100A8/A9 promotes leukocytosis via increased proliferation of BM progenitor cells. (A) Plasma levels of S100A8/A9 in STZ-diabetic mice treated with SGLT2i. n=6. (B) mRNA expression of S100a8, S100a9 and Hmgb1 in FACS isolated neutrophils. n=6, *P<0.05 vs. WT, ^P<0.05 vs. WT+STZ group. (C) Total BM cells were isolated from WT mice and cultured in HG (25mM, 16hrs) and stimulated with S100A8/A9 complex. GMP proliferation was assessed by measuring EdU incorporation via flow cytometry. n=4 independent experiments, *P<0.05 vs. vehicle. (D) Monocyte levels and (E) BM progenitor cells in WT mice in response to vehicle or S100A8/A9 complex (20μg/kg/mouse, i.v, twice daily, 3 days) *P<0.05 vs. vehicle. (F-I) S100a9^−/− BMT: (F) Experimental overview: WT mice were transplanted with BM from either WT or S100a9^−/− mice and made diabetic with STZ. (G) Blood leukocyte levels after 4 weeks of diabetes. (H) Percentage of HSPCs, CMPs and GMPs in the BM and (I) percentage of HSPCs, CMPs and GMPs in the G₂M phase of the cell cycle. D-I, n=5/group. *P<0.05 vs. all groups. All values are means ± SEM.

Figure 3

RAGE on myeloid progenitor cells mediates S100A8/A9-stimulated leukocytosis in diabetic mice. (A) GMP proliferation in response to S100A8/A9 (2μg/ml) as measured by EdU incorporation. n=4 independent experiments. *P<0.05 vs. vehicle in each respective genotype. (B) Blood leukocyte levels in WT and Rage^−/− mice with and without STZ-diabetes. n=5-6/group. *P<0.05 vs. all groups. (C) Surface expression of RAGE on CMPs (histogram and quantification) in WT and STZ-diabetic mice treated with SGLT2i. n=6, *P<0.05 vs. all groups, *P<0.05 vs. WT+STZ. (D) Experimental overview: WT mice were transplanted with BM from WT or Rage^−/− mice and made diabetic with STZ. (E-G) After 4wks of diabetes, (E) blood leukocytes, (F) BM HSPC, CMP, GMP numbers and (G) proliferation measured by flow cytometry. n=5/group. *P<0.05 vs. all groups. (H-L) Competitive BMT: (H) Experimental overview: Equally mixed portions of CD45.1 and CD45.2 BM from the respective genotypes were transplanted into WT CD45.2 recipients and made diabetic with STZ. (I) Numbers of CD45.1 and CD45.2 monocytes and neutrophils from the respective donor BM. (J) Percentage of CD45.1 and CD45.2 CMPs and GMPs and (K) percentage of CD45.1 and CD45.2 CMPs and GMPs in the cell cycle from each respective donor BM. Data are means ± SEM, n=5-6/group. Numbers in parentheses indicate ratio of CD45.1:CD45.2. *P<0.05 vs. W/W. (L) Scheme summarizing the cell extrinsic proliferative pathway induced by S100A8/A9-RAGE signaling. (M) GMP proliferation in response to S100A8/A9 ± the NF-κB inhibitor (SN50, 20μM). (N) mRNA of M-CSF, GM-CSF and G-CSF as quantified by qRT-PCR. n=4 independent experiments. *P<0.05 vs. all groups, ^P<0.05 vs. HG. (O) GMP proliferation in response to S100A8/A9 ± neutralizing antibodies to M-CSF and/or GM-CSF or isotype controls (ISO) (all 30μg/ml). n=4 independent experiments. *P<0.05 vs. ISO vehicle, ^P<0.05 vs. respective ISO control. All values are means ± SEM.

Figure 4

Neutrophils drive hyperglycemia-mediated leukocytosis in diabetes. (A-F) Neutrophils in WT and STZ-diabetic mice were depleted by injecting anti-Ly6G antibody (clone 1A8, 1mg/mouse, i.p injection) every 3 days for 3 weeks. (A) Neutrophil and (B) monocyte levels in WT and STZ mice treated with anti-Ly6-G or an isotype control. (C) Plasma levels of S100A8/A9. (D) CMP and GMP cell populations in the BM and (E) CMP and GMP cell proliferation assessed by DAPI staining and represented as percentage of cells in the G₂M phase of the cell cycle. (F) Surface expression of RAGE on CMPs. All experiments n=5/group. *P<0.05 indicated diabetes effect, ^P<0.05 indicated anti-Ly6G effect. All values are means ± SEM.

Figure 5

Defective lesion regression in diabetic mice is improved by normalizing plasma glucose. (A) Experimental overview: Ldlr^−/− mice were fed a HCD (0.15%) for 16 weeks to develop atherosclerotic lesions. At this time point a group of mice was sacrificed to determine baseline (pre-regression) lesion characteristics. The remaining mice were divided into 3 groups; Reg, Reg+STZ and Reg+STZ+SGLT2i and placed on chow diet for 6 weeks (n=10-11/group). (B) Blood glucose and total cholesterol levels at baseline and 6 weeks of regression. *P<0.05 vs. all groups. (C) Blood leukocyte levels at baseline and after 6 weeks of regression as assessed by flow cytometry. n=9-11/group. *P<0.05 vs. all groups, ^P<0.05 vs. Reg+STZ. (D) Quantification of mean lesion areas. (E) Representative ORO stained lesions and quantification of ORO stain as percent of lesion area. (F) Representative CD68⁺ stained lesions and quantification as CD68⁺ area/lesion. *P<0.05 vs. Reg and Reg+STZ+SGLT2i, ^P<0.05 vs Reg+STZ. (G-H) FACS isolated Ly6-C^hi monocytes were (G) allowed to adhere to cultured human aortic endothelial cells under static conditions or (H) assessed for their ability to migrate towards CCL2 (in a transwell chamber). *P<0.05 vs. all groups, ^P<0.05 vs. Reg+STZ n=4-6/group. (I) Representative images and quantification of lesions stained with an anti-Ly6-C antibody (FITC; green) and Hoechst dye (blue; nuclei) using confocal microscopy. Arrows indicate Ly6-C⁺ cells (Ly6-C^hi monocytes). *P<0.05 vs. all groups, ^P<0.05 vs. Reg+STZ. All values are means ± SEM.

Figure 6

Impaired lesion regression in diabetic mice is due to increased monocyte recruitment. (A) Ldlr^−/− mice were fed a HCD for 16wks to develop atherosclerotic lesions. Aortic transplantation model overview: All donor Ldlr^−/− mice were injected with EdU (1 mg, i.v) to label newly formed monocytes. 48 hrs post-injection, aortas were dissected from these mice and either processed for baseline measurement of EdU⁺ cells or transplanted into chow fed Akita mice ± SGLT2i. 48hrs prior to termination, mice were injected with green fluorescent microspheres to determine monocyte entry. (B) Monocyte entry determined by fluorescent beads. (C-D) Monocyte egress: (C) Representative images showing EdU stain (red) in the aortic arch sections of donor mice (baseline) and grafts from Akita and Akita+SGLT2i. (D) Quantification of monocyte egress as represented by the number of EdU⁺ cells per section. *P<0.05 vs. baseline, n=5/group. (E) Diet induced regression model overview: Ldlr^−/− mice fed with a HCD for 16 wks were divided into 3 groups; regression (reg), reg+STZ and reg+STZ+SGLT2i and placed on chow diet. At the end of 4 weeks, all mice were injected (i.v) with clodronate liposomes (CLO, 250μl) to deplete the circulating monocytes. 48 hrs later they were injected with green fluorescent microspheres. 4 days later, a portion for mice from each group were sacrificed to determine the baseline measurement of beads in the atheroma and to assess Ly6-C^hi monocyte entry. A second group of mice was assessed 14 days later (6 weeks of regression) for quantification of labeled macrophages remaining in plaques. 48 hrs prior to sacrificing the final group of mice they were injected with EdU to also assess Ly6-C^hi monocyte entry. (F) Ly6-C^hi monocyte entry as determined by EdU⁺ cells in the lesion. (G) Ly6-C^hi monocyte entry and monocyte/macrophage retention in the lesion as assessed by number of beads/section. (H) Percentage of monocyte/macrophage egress compared to baseline. *P<0.05 vs. all groups, ^P<0.05 final vs. baseline, n=6/group. All values are means ± SEM.

Figure 7

S100A8/A9 correlates with leukocytes in T1DM patients with CAD and stimulates myelopoiesis in human CD34+ progenitor cells. (A) Leukocyte and S100A8/A9 levels in T1DM patients with and without CAD. (B-C) Regression analysis: (B) S100A8/A9 vs. WBCs and (C) S100A8/A9 vs. neutrophils. n=49. (D) Proliferation of CD34⁺ progenitor cells to increasing doses of S100A8/A9 was measured by EdU incorporation. (E) Production of CD14⁺ monocytes from CD34⁺ progenitor cells to increasing doses of S100A8/A9. D, E, n=4 independent experiments, data are means ± SEM, *P<0.05 vs. control. (F) Schematic overview: In response to hyperglycemia, (1) neutrophils produce S100A8/A9, which is (2) sensed by RAGE on BM CMPs and macrophages to signal via NF-κB to induce M-CSF, GM-CSF production. M-CSF and GM-CSF in turn acts in an autocrine/paracrine fashion to stimulate CMP and GMP proliferation, producing monocytes and neutrophils. (3) Ly6-C^hi monocytes in the circulation become activated, adhere to the endothelium and readily enter lesions leading to increased lesional macrophages.