Digoxin reduces atherosclerosis in apolipoprotein E‐deficient mice

Abstract

Background and Purpose

Numerous in vitro studies have suggested that digoxin suppresses inflammation and alters lipid metabolism. However, the effect of dioxin on atherosclerosis is poorly understood. The present study was conducted to determine whether digoxin affects the development of atherosclerosis in a murine model of atherosclerotic disease.

Experimental Approach

Apolipoprotein E‐deficient mice maintained on a Western‐type diet were administered PBS (control), low‐dose digoxin (1 mg·kg⁻¹· day⁻¹) or high‐dose digoxin (2 mg·kg⁻¹ · day⁻¹) via i.p. injection for 12 weeks.

Key Results

Digoxin dose‐dependently reduced atherosclerotic lesion formation and plasma lipid levels (reductions of 41% in total cholesterol, 54% in triglycerides and 20% in low‐density lipoprotein cholesterol in the high‐dose digoxin‐treated group). Moreover, treatment with digoxin markedly attenuated IL‐17A expression and IL‐17A‐related inflammatory responses and increased the abundance of regulatory T cells (Tregs).

Conclusions and Implications

Our data demonstrate that digoxin acts as a specific antagonist of retinoid‐related orphan receptor‐γ to decrease atherosclerosis by suppressing lipid levels and IL‐17A‐related inflammatory responses.

Abbreviations

Cyp8b1

sterol 12α‐hydroxylase

Elovl3

ELOVL fatty acid elongase 3

Insig2a

insulin‐induced gene 2a

ROR

retinoid‐related orphan receptor

WD

Western‐type diet

Tables of Links

TARGETS
Nuclear hormone receptors a	Enzymes b
RORα (NR1F1)	Cyp8b1
RORβ (NR1F2)
RORγ (NR1F3)

LIGANDS
Digoxin	MCP‐1 (CCL2)
IFN‐γ	MMP‐2
IL‐1β	MMP‐9
IL‐6	Ranitidine
IL‐10	TNF‐α
IL‐17A

These Tables list key protein targets and ligands in this article which are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Pawson et al., 2014) and are permanently archived in the Concise Guide to PHARMACOLOGY 2015/16 (^a,bAlexander et al., 2015a, 2015b).

Introduction

Coronary artery disease (CAD) arising from atherosclerosis is a leading cause of morbidity and mortality worldwide and has severe consequences for both individuals and society. The underlying pathology of CAD is chronic inflammation resulting from an imbalance in lipid metabolism and from maladaptive immune responses (Moore et al., 2013). Early atherosclerotic lesions form due to local endothelial dysfunction that leads to infiltration of inflammatory blood cells and increased uptake or subendothelial retention of lipid‐rich plasma (Simionescu et al., 1986; Simionescu, 2007). The accumulation of normal and modified lipoproteins transforms them into cholesterol‐laden foam cells that promote the formation of early atherosclerotic lesions. Therefore, decreasing circulating cholesterol levels and reducing inflammation are known to be effective measures in preventing the development of atherosclerosis (Grundy, 1998; Charo and Taub, 2011).

The nuclear receptor gene superfamily consists of cytosolic and nuclear transcription factors that regulate gene expression during development and maintain systemic homeostasis. Members of this superfamily include steroid receptors, retinoid receptors, thyroid receptors and orphan receptors. Retinoid‐related orphan receptor (ROR) nuclear receptors are members of the orphan nuclear receptor class. The ROR subfamily consists of three members, RORα (NR1F1), RORβ (NR1F2) and RORγ (NR1F3). The tissue expression patterns of RORs are diverse, and each ROR protein exhibits tissue‐specific expression patterns reflecting its biological functions (Jetten et al., 2001). RORγ is most highly expressed in muscle, liver, kidney, mammary gland, thymus and heart tissues (Hirose et al., 1994; Medvedev et al., 1996). RORγ, which plays a critical role in lipid/glucose homeostasis and various immune functions, has been implicated in metabolic syndrome and several inflammatory diseases. The RORγ gene generates two isoforms, RORγ1 and RORγ2 (RORγt). RORγt plays a critical role in several immune processes (Huh and Littman, 2012). Recently, studies identified a critical role of RORγt in the lineage specification of uncommitted CD4⁺ T helper cells into Th17 cells (Park et al., 2005), and inhibiting RORγt activity regulates not only the expression of Th17 effector cytokines but also the trafficking and expansion of Th17 cells. It has been reported that RORγ directly regulates the transcription of several lipid metabolic genes (Huang et al., 2007; Takeda et al., 2014). Therefore, RORγ antagonists have emerged as important drug targets for the treatment of various diseases, such as multiple sclerosis and rheumatoid arthritis.

Digoxin has long been used as an effective treatment for heart failure. Additionally, a previous study showed that digoxin acts as a specific antagonist of RORγ without affecting RORα (Fujita‐Sato et al., 2011; Huh et al., 2011). Therefore, the aim of the present study was to determine whether digoxin affects atherosclerotic lesions in apolipoprotein E‐deficient (ApoE^−/−) mice.

Methods

Animals

ApoE^−/− mice on a C57BL/6 background were purchased from the Jackson Laboratory (Bar Harbor, Maine). Animals (8 weeks old, 19 to 22 g) were randomly separated into three groups (n = 8 mice per group). Each mouse in the first (control) group was injected with the same volume (0.1 mL) of PBS containing 0.1% DMSO. The second and third groups received i.p. injections of approximately 20 μg (1 mg·kg⁻¹ · day⁻¹; low dose) or 40 μg of digoxin (2 mg·kg⁻¹ · day⁻¹; high dose) and were fed an atherogenic Western‐type diet (WD) containing 0.15% cholesterol and 21% fat for 12 weeks. The mice were weighed every day, and the drug dose was adjusted according to the daily weight. At the end of the 12‐week treatment period, on the day of killing, all animals were anaesthetized via i.p. injection of sodium pentobarbital (50 mg·kg⁻¹) and were exsanguinated via retro‐orbital venous puncture under general anaesthesia. The mice were subsequently killed via cervical dislocation. The blood samples obtained from the animals were stored at −80°C until further analysis. After the animals had been killed, their aortas were isolated. All animals were handled, and all experiments were conducted in accordance with the NIH Guidelines for the Care and Use of Laboratory Animals (Science and Technology Department of Hubei Province, China) and were approved by the Institutional Animal Care and Use Committee at Tongji Medical College (Huazhong University of Science and Technology; IACUC number: 436). The animals were housed in cages (n = 1 per cage) located in a well‐ventilated holding room at constant humidity of 55 ± 5% and temperature of 24 ± 1°C under a 12 h light–dark cycle and received water and food ad libitum. Animal studies are reported in compliance with the ARRIVE guidelines (Kilkenny et al., 2010; McGrath and Lilley, 2015).

Randomization

Randomization was conducted by an individual other than the operator. The ApoE^−/− mice were sequentially numbered from 1 to 24, and reproduced 24 numbers from the random number table, each number shall be in addition to the 3, and the remainder 0, 1, 2 on behalf of the control, low‐dose digoxin and high‐dose digoxin groups.

Blinding

Data analysis was conducted in a blinded manner (single‐blind study design).

Mouse hepatocyte isolation

Eight‐week‐old ApoE^−/− mice were anaesthetized via i.p. injection of 50 mg·kg⁻¹ pentobarbital. Mouse primary hepatocytes were isolated using a two‐step in situ collagenase perfusion method. The liver was removed and perfused with buffer A [10 mM HEPES, pH 7.4, gentamicin sulphate (1 mg·mL⁻¹ medium) and 0.5 mM EGTA in calcium/magnesium‐free HBSS] for 15 min, followed by perfusion with 0.5 mg·mL⁻¹ type IV collagenase dissolved in Earle's balanced salt solution (EBSS) at 37°C until the liver capsule was incised. After perfusion, the thick fibrous connective tissue was discarded, and filtered cell suspensions were harvested. Primary hepatocytes were collected via centrifugation and used in the subsequent experiments.

Flow cytometry

After administration of the WD for 12 weeks, spleen lymphocytes were isolated with Ficoll‐Paque Plus, and red blood cell lysis buffer (RCLB) was used to remove the erythrocytes. Flow cytometry was used to analyse the subpopulations of lymphocytes in the spleen. Cells were stained with anti‐CD4‐FITC, anti‐CD25‐APC, anti‐IL‐17A‐PE, anti‐IFN‐γ‐PE, anti‐IL‐4‐PE and anti‐Foxp3‐PE antibodies and their respective isotype controls (eBioscience) according to the manufacturer's instructions. Flow cytometry was performed using a FACSCalibur system (Becton‐Dickinson, San Jose, CA, USA). The data were analysed using flowjo software (Treestar Inc.).

Quantification of aortic atherosclerotic lesions

Aortas and aortic valves were prepared as follows. In brief, en face preparations of entire aortas were dissected, fixed, opened longitudinally and pinned on black wax plates. Then, lesions were visualized by staining with Oil Red O (Sigma). The aortic roots were fixed in 4% formaldehyde, processed and embedded in optimum cutting temperature compound. The resultant aortic sinus cryosections (7 μm) were stained with Oil Red O and haematoxylin. The mean atherosclerotic areas were calculated from eight different mice. Ten serial cryosections/tissue sections for each mouse were evaluated. The total atherosclerotic area for each plaque area measurement from each mouse was used for this calculation. image‐pro plus 6.703 software (Media Cybernetics) was used for statistical analysis.

Immunostaining and immunofluorescence

For histological analysis, the aortic roots were sliced into 5 μm serial cryostat sections in the aortic valve plane. Cryosections were fixed in 4% paraformaldehyde for 30 min and rinsed in the tris‐buffered saline (TBS). Non‐specific binding sites were blocked using an avidin/biotin blocking kit, followed by incubation of the sections in 1% BSA (Sigma) and 5% normal goat serum in PBS. The slides were incubated overnight at 4°C with an anti‐mouse SMA antibody (1:200) for smooth muscle cells, an anti‐CD68 antibody (1:200) for macrophages, an anti‐Foxp3 antibody (1:50) for regulatory CD4⁺ T cells (Tregs) and anti‐IL‐17A and anti‐CD4 antibodies (1:50) for T cells. The slides were rinsed and incubated with secondary antibodies. The processed sections were visualized using an Olympus microscope (IX71; Olympus Corporation, Tokyo, Japan) and a fluorescence microscope (Olympus Microscope BX‐51; Olympus Corporation) or a confocal microscope (Nikon, Tokyo, Japan). The means from 10 serial cryosections/tissue from eight samples per group were recorded. Macrophages and smooth muscle cells were quantified by assessing the percentage of the total plaque area that was positive for each marker. CD4⁺ T cells and Tregs were assessed by counting the number of positively stained cells.

Lipid measurement

The plasma levels of total cholesterol, triglycerides, LDL cholesterol and HDL cholesterol were measured using chemically modified enzyme‐based assay kits (Kyowa Medex, Tokyo, Japan) according to the manufacturer's instructions. Cholesterol and triglycerides were extracted from liver tissues as described. After removal of tissue debris via centrifugation at 12 000× g, the supernatant was dried under nitrogen. Total cholesterol content was determined using an Amplex® Red Cholesterol Assay Kit (Life Technologies). Liver tissues were prepared and analysed for triglycerides using a Triglyceride Colorimetric Assay Kit (Item No. 10010303; Cayman, Ann Arbor, MI, USA) according to the manufacturer's instructions. Protein concentrations were determined using the Lowry assay (Lowry et al., 1951).

RNA extraction and real‐time RT‐PCR

Total RNA was prepared using TRIzol Plus (Takara) according to the manufacturer's instructions. RNA purity and concentration were measured using a spectrophotometer. One microgram of total RNA was reverse transcribed into cDNA using an RNA PCR kit (TaKaRa Bio Inc, Seta, Japan). The cDNA was used as a template for RT‐PCR. The sequences of the PCR primers used are shown in Supporting Information Table S1. The PCR reaction mixture included SYBR Green (Takara). Real‐time PCR was performed using the Applied Biosystems Step One system (Applied Biosystems, Carlsbad, CA, USA), and the housekeeping gene GAPDH was used as an internal control for mRNA abundance. Fold changes in the levels of target gene mRNAs were determined using the formula 2^−△△Ct.

Measurements of cytokines

Mouse serum was collected and stored frozen at −80°C until the cytokine levels were determined via elisa according to the manufacturer's instructions. The levels of mouse IL‐17A, IL‐10 and IL‐6 were measured using a mouse multi‐cytokine detection kit (Millipore, Billerica, MA, USA).

Measurement of the plasma digoxin levels

Twenty‐four hours after the final injection, blood was collected from the retro‐orbital vein, mixed and drawn into a capillary tube. Plasma was separated from the blood samples via low‐speed centrifugation (3000× g for 10 min at 4°C), collected and frozen at −80°C until analysis. The plasma digoxin levels were measured using a commercially available kit (Monobind Inc. Lake Forest, CA 92630, USA) according to the manufacturer's instructions.

Statistical analysis

All data were first evaluated for normal distribution using the Kolmogorov–Smirnov test. The results of normally distributed data are expressed as the means ± SEM. The data that did not pass the test for normality are presented as the medians with the 25th and 75th percentiles. For normally distributed data with equivalent group variances, one‐way ANOVA followed by the Holm–Sidak test was used for multiple comparisons between ≥3 groups. Post hoc tests were performed only if P < 0.05, and there was no significant inhomogeneity of variance. When group data were not normally distributed or if the group variances were unequal, the Kruskal–Wallis test followed by the Dunn post hoc test was used. P < 0.05 was considered statistically significant. All statistical analyses were performed using spss software (version 17.0, SPSS Inc., Chicago, IL, USA). The data and statistical analysis comply with the recommendations on experimental design and analysis in pharmacology (Curtis et al., 2015).

Reagents and antibodies

Mouse hepatocytes were cultured in DMEM (Gibco, Life Technologies, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FCS) (Gibco, Life Technologies, Grand Island, NY, USA) and 100 U·mL⁻¹ streptomycin/penicillin. Digoxin was obtained from Sigma (D6770; St. Louis, MO, USA). A purified anti‐CD4 antibody (clone RM4‐5) was obtained from BD Systems (Franklin Lakes, NJ, USA). An anti‐CD68 antibody (ab955) was obtained from Abcam. An anti‐smooth muscle actin (SMA) antibody (lot no. BM0002) was obtained from Boster (Wuhan, China). Anti‐CD4‐FITC (clone GK1.5), anti‐forkhead box P3 (Foxp3)‐PE (clone FJK‐165), anti‐CD25‐APC (clone PC61.5), anti‐interferon (IFN)‐γ‐PE (clone XMG1.2), anti‐IL‐4‐PE (clone 11B11) and anti‐IL‐17A‐PE antibodies (clone eBio17B7) were obtained from eBioscience (San Diego, CA, USA).

Results

Digoxin attenuates atherosclerotic lesion development in ApoE^−/− mice

To investigate the potential effects of digoxin on atherosclerosis, we evaluated the severity of atherosclerosis based on the morphological and histological changes that occurred in digoxin‐treated and PBS (0.1% DMSO)‐treated mice. Plasma digoxin levels were measured 24 h after the final injection (low‐dose digoxin‐treated group: 1.1 ± 0.25 ng·mL⁻¹; high‐dose digoxin‐treated group: 2.12 ± 0.20 ng·mL⁻¹; Supporting Information Fig. S1). Treatment with digoxin resulted in a reduction in the severity of atherosclerosis, along with a significant trend towards a reduction in the lesion area with increasing doses of digoxin (Figure 1A–D). Treatment with high‐dose digoxin (2 mg·kg⁻¹ day⁻¹) resulted in a 45% reduction in the severity of atherosclerosis compared with the control treatment (Figure 1A). Quantification of Oil Red O staining of the entire aorta revealed decreased atherosclerosis in mice treated with either dose of digoxin compared with the PBS‐treated control mice (Figure 1B). These findings were in line with the amounts of atherosclerotic lesions observed in the aortic sinuses of mice treated with low‐dose digoxin, high‐dose digoxin or PBS (Figure 1D).

Figure 1.

Digoxin inhibits the development of atherosclerosis in ApoE^−/− mice. (A) Representative images of Oil Red O staining of en face preparations of aortas from the different treatment groups (injected with a low‐dose digoxin at 1 mg·kg⁻¹ day⁻¹, high‐dose digoxin at 2 mg·kg⁻¹ day⁻¹ or 0.1 mL of PBS containing 0.1% DMSO for 12 weeks). (B) Quantitative analysis of the atherosclerotic surface area of the entire aorta. (C) Cryosections of the aortic sinus stained with Oil Red O; haematoxylin was used as a counterstain. (D) The measured lesion sizes in the aortic roots were averaged to determine the average lesion size of 10 sections of the aortic sinus. The data are expressed as the means ± SEM. One‐way ANOVA followed by the Holm–Sidak test, *P < 0.05; PBS group, n = 8; low‐dose digoxin group (Dig‐LD), n = 7; high‐dose digoxin group (Dig‐HD), n = 8.

Metabolic effects of digoxin

Lipid metabolism critically influences atherosclerosis. The total cholesterol, HDL cholesterol and LDL cholesterol and triglyceride levels were measured in ApoE^−/− mice after 12 weeks of WD feeding to determine whether digoxin affects any of these parameters. As shown in Figure 2, the total plasma cholesterol, triglyceride and LDL cholesterol levels were significantly different between the PBS‐treated mice and the mice treated with the high dose of digoxin (Figure 2A–C); however, the HDL cholesterol levels were slightly, but not significantly, lower in the digoxin groups than in the control group (Figure 2D). Treatment with digoxin dose‐dependently reduced mouse body weight gain (Figure 2E). These data indicate that digoxin protects against atherosclerosis in ApoE^−/− mice and that this effect may be attributed to alterations in serum lipid metabolism and body weight gain.

Figure 2.

To determine the effects of digoxin treatment on plasma cholesterol levels, we measured the concentrations of total cholesterol (A), triglycerides (B), LDL cholesterol (C) and HDL cholesterol (D). Body weight was recorded at the indicated time points (E). The body weights are expressed as the average weights of eight mice (*a comparison with the PBS‐treated group). (A–D) The data are expressed as the means ± SEM. One‐way ANOVA followed by the Holm–Sidak test. (E) Repeated‐measures ANOVA followed by a post hoc test from five independent experiments, *P < 0.05; PBS group, n = 8; low‐dose digoxin group (Dig‐LD), n = 7; high‐dose digoxin group (Dig‐HD), n = 8.

To assess whether digoxin reduced lipid levels, we examined the concentrations of total cholesterol and triglycerides in the liver. Our results show that the cholesterol and triglyceride levels were reduced by digoxin (Figure 3A, B). Next, we tested the expression of known RORγ target genes. The expression levels of insulin‐induced gene 2a (Insig2a), ELOVL fatty acid elongase 3 (Elovl3) and sterol 12α‐hydroxylase (Cyp8b1) were reduced in the liver in a digoxin dose‐dependent manner (Figure 3C). Identical results were observed in hepatocytes at the indicated concentrations of digoxin (Figure 3D–F). These results suggest that digoxin regulates lipid levels by down‐regulating RORγ target genes.

Figure 3.

Effects of digoxin treatment on the levels of cholesterol and triglycerides in the liver. We measured the concentrations of total cholesterol (A) and triglycerides (B) in the liver. To determine the effects of digoxin treatment on the expression of known RORγ target genes, the mRNA levels of Insig2a, Elovl3 and Cyp8b1 in the liver were determined via RT‐PCR analysis (C). Hepatocytes were incubated in the indicated concentrations of digoxin for 24 h. (D–F) The Insig2a, Elovl3 and Cyp8b1 expression levels were determined via RT‐PCR analysis (*a comparison with the PBS‐treated group). The data are expressed as the means ± SEM from five independent experiments. One‐way ANOVA followed by the Holm–Sidak test, *P < 0.05; PBS group, n = 8; low‐dose digoxin group (Dig‐LD), n = 7; high‐dose digoxin group (Dig‐HD), n = 8.

Effect of digoxin on the stability of atherosclerotic plaques

We discovered that digoxin dose‐dependently decreased atherosclerotic plaque formation. Moreover, the proportion of lesion cross‐sectional area containing CD68⁺ macrophages was reduced in both digoxin‐treated groups compared with the PBS‐treated group (Figure 4A, B). More importantly, the abundance of α‐SMA⁺ vascular smooth muscle cells (VSMCs) at the fibrous cap was significantly increased in a digoxin dose‐dependent manner (Figure 4A, C). Furthermore, the collagen⁺ area was increased in the atherosclerotic lesions of mice that received digoxin compared with those that received PBS (Figure 4A, D). Additionally, the abundance of CD4⁺ cells was significantly decreased in the mice administered digoxin (especially high‐dose digoxin) compared with the mice treated with PBS (Figure 4A, E). These data suggest that digoxin dose‐dependently accelerates the stability of atherosclerotic lesions, which were composed of fewer macrophages and CD4⁺ T cells and more frequently expressed α‐SMA at the fibrous cap and collagen due to digoxin treatment.

Figure 4.

Effect of digoxin on the stability of atherosclerotic plaques in ApoE^−/− mice. (A) Representative photomicrographs of aortic root sections stained with CD68, α‐smooth muscle actin, collagen and CD4. (B–E) Quantitative analysis of the data is shown. (B and D) The data are expressed as the means ± SEM. One‐way ANOVA followed by the Holm–Sidak test. Other data (C and E) are expressed as the medians with 25th and 75th percentiles. The Kruskal–Wallis test followed by the Dunn post hoc test, *P < 0.05; PBS group, n = 8; low‐dose digoxin group (Dig‐LD), n = 7; high‐dose digoxin group (Dig‐HD), n = 8.

Digoxin treatment changes the levels of cytokines and frequencies of Th17 cells and Tregs in ApoE^−/− mice

To assess the effect of digoxin on Th17 cell differentiation, flow cytometry analysis was performed to determine splenocyte numbers. Compared with the control treatment, high‐dose digoxin treatment significantly decreased the proportion of Th17 cells (Figure 5A–B). To our surprise, the percentage of Foxp3‐expressing Tregs was greatly increased in the high‐dose digoxin‐treated group (Figure 5C–D), and we obtained the same results in atherosclerotic plaques (Supporting Information Fig. S4). Low‐dose digoxin treatment also decreased the abundance of Th17 cells and increased the Treg count (Figure 5A–D). In addition, in the atherosclerotic plaques, double staining for IL‐17A and CD4 showed that the IL‐17A protein was expressed in CD4⁺ T cells within the lesions (Supporting Information Fig. S3). However, no marked changes were observed in the abundance of CD4⁺ T cells in the spleen expressing IFN‐γ or IL‐4, markers of Th1 and Th2 cells, respectively (Figure 6A–D).

Figure 5.

Effect of digoxin on the balance of Th17 cells/Tregs. (A–D) Digoxin changed the percentage of Th17 cells and CD4⁺ forkhead box P3 (Foxp3)⁺ Tregs among the mouse splenocytes. The concentrations of IL‐17A, IL‐10 and IL‐6 were measured by elisa. The data are expressed as the means ± SEM. One‐way ANOVA followed by the Holm–Sidak test from five independent experiments, *P < 0.05; PBS group, n = 8; low‐dose digoxin group (Dig‐LD), n = 7; high‐dose digoxin group (Dig‐HD), n = 8.

Figure 6.

Effect of digoxin on the balance of Th1/Th2 cells. (A–D) CD4⁺ IFN‐γ⁺ cells and CD4⁺ IL‐4⁺ cells were detected via intracellular staining for IFN‐γ and IL‐4. The quantities of CD4⁺ IFN‐γ⁺ cells and CD4⁺ IL‐4⁺ cells were determined by analysing the FACS data using flowjo software. (E) The mRNA expression levels of relevant cytokines [TNF‐α, IFN‐γ, IL‐1β, MCP‐1(CCL‐2), MMP‐2 and MMP‐9] in the aortic root were determined via RT‐PCR analysis. (F) The mRNA levels of T‐bet, RORγt and Foxp3 in the aortic root were determined via RT‐PCR analysis. The data are expressed as the means ± SEM. One‐way ANOVA followed by the Holm–Sidak test from five independent experiments, NS indicates P > 0.05; *P < 0.05; PBS group, n = 8; low‐dose digoxin group (Dig‐LD), n = 7; high‐dose digoxin group (Dig‐HD), n = 8.

Furthermore, at both doses of digoxin, the serum levels of IL‐17A were significantly decreased. In contrast, only the high dose of digoxin markedly increased the IL‐10 levels and reduced IL‐6 secretion compared with PBS (Figure 5E–G). In addition to the microscopic assessment, the aortic tissue was harvested after 12 weeks. We then determined the expression of Th17‐related cytokines in the aorta samples to address whether the expression of these cytokines was reduced at the lesion site. RT‐PCR was used to determine the mRNA expression of TNF‐α, IL‐1β, MCP‐1 (CCL‐2), IFN‐γ, IL‐1β, MMP‐2 and MMP‐9. Digoxin treatment dose‐dependently reduced the expression of TNF‐α, IL‐1β, MCP‐1 (CCL‐2), IFN‐γ, MMP‐2 and MMP‐9 (Figure 6E). Although there were no significant differences in the RORγt or T‐bet mRNA levels, the Foxp3 mRNA levels were dose‐dependently significantly higher in the digoxin‐treated groups. These findings suggest that digoxin attenuates atherosclerotic lesions and that this effect may contribute to the attenuation of IL‐17A‐related inflammatory responses and the enhancement of Treg expansion rather than the imbalance in the Th1/Th2 ratio.

Discussion and conclusions

In the present study, we demonstrated that digoxin significantly inhibits the development of atherosclerotic lesions in ApoE^−/− mice fed a WD for 12 weeks (Figure 1). These observations serve as direct evidence of the atheroprotective effects of digoxin. The anti‐atherogenic effects of digoxin (especially at the high dose) appear to involve a marked reduction in the plasma levels of total cholesterol, LDL cholesterol and triglycerides (Figure 2A–C). In addition, digoxin, which acts as a potent antagonist of RORγt, markedly decreased the expression of IL‐17A and IL‐17A‐related inflammatory mediators and increased the proportion of Tregs (Figure 4).

The levels of total cholesterol, triglycerides and LDL cholesterol significantly affect the incidence of major vascular events and the development of atherosclerosis in humans (Heart Protection Study Collaborative, 2002; National Cholesterol Education Program Expert Panel on Detection et al., 2002; Group et al., 2014). Lowering total cholesterol and LDL cholesterol level can prevent the progression of atherosclerosis. Additionally, many studies have demonstrated that an elevated triglyceride level is also an independent risk factor of coronary heart disease (Hokanson and Austin, 1996; Schulte et al., 1999; Sarwar et al., 2007; Talayero and Sacks, 2011). Most importantly, we demonstrated that digoxin, an RORγ antagonist, markedly decreased the levels of total cholesterol, triglycerides and LDL cholesterol by 41, 54 and 20%, respectively, in ApoE^−/− mice treated with high‐dose digoxin compared with control mice (Figure 2A–C). We obtained the same results from liver tissue (Figure 3A, B). Subtle changes in this lipoprotein fraction may have profound effects on the development of atherosclerosis.

To fully explore our hypothesis, we measured the hepatic expression of several lipid metabolic genes. Yukimasa Takeda et al. (2014) found that RORγ functions as an important linker between the circadian clock and regulation of lipid metabolism, as evidenced by the target genes of RORγ (Insig2a, Elovl3 and Cyp8b1). In this study, we demonstrated that the expression levels of Insig2a, Elovl3 and Cyp8b1, but not RORγ (Fig S2), were markedly reduced by digoxin compared with PBS (Figure 3C). Our results also showed that Insig2a, Elovl3 and Cyp8b1 were significantly down‐regulated by digoxin in hepatocytes (Figure 3D, E). These results suggest that digoxin down‐regulates RORγ target genes by inhibiting RORγ activity rather than regulating RORγ expression. Insig2a and Cyp8b1 play critical roles in the regulation of lipid metabolism (Cervino et al., 2005). As reported previously, Elovl3 is involved in the regulation of the progression of adipogenesis. Elovl3‐ablated mice exhibit impaired formation of triglycerides and attenuated lipid accumulation in liver and adipose tissue. The expression levels of adipogenic, lipolytic and lipogenic genes were also reduced in Elovl3‐ablated mice, and these mice are resistant to diet‐induced weight gain (Zadravec et al., 2010; Kobayashi and Fujimori, 2012). Our results also showed that mice treated with either dose of digoxin exhibited significant reductions in body weight gain (Figure 2E). Suppression of Insig2a, Elovl3 and Cyp8b1 expression may have contributed to the reductions in the total cholesterol, triglyceride and LDL cholesterol levels and the inhibition of body weight gain in ApoE^−/− mice. However, lipid metabolism and weight gain are regulated by complex processes, and further studies are needed to elucidate the mechanisms involved.

Atherosclerosis is a chronic inflammatory disease involving T lymphocytes (Libby, 2002). Innate and adaptive immune responses have been shown to modulate local and systemic inflammation during all stages of atherosclerosis. The Th17 cell/Treg balance controls inflammation and plays a key role in the development and progression of atherosclerosis (Xie et al., 2010). Th17 cells are a subset of CD4⁺ cells that secrete IL‐17A, IL‐6 and TNF‐α. Functional blockade of IL‐17A results in decreased local inflammation and reduced atherosclerotic lesion development (Stockinger and Veldhoen, 2007; Erbel et al., 2009; Smith et al., 2010; Taleb et al., 2010). In our study, we demonstrated that high‐dose digoxin treatment markedly reduced the proportion of IL‐17A⁺ CD4⁺ T cells in the spleen and the plasma levels of IL‐17A and IL‐6 compared with PBS treatment (Figure 5). Moreover, we showed that digoxin decreased the mRNA expression of the inflammatory cytokines TNF‐α, MCP‐1, IFN‐γ, MMP‐2 and MMP‐9 in the aorta (Figure 6E). These findings are consistent with those of previous studies suggesting that IL‐17 deficiency inhibits the expression of inflammatory cytokines such as MCP‐1, IL‐1β, IL‐6 and TNF‐α (Jones and Chan, 2002; Iwakura and Ishigame, 2006; Usui et al., 2012). As reported by Jinping Liu in 2014, digoxin treatment significantly reduced the expression of IL‐17A, MCP‐1 (CCL‐2), IFN‐γ and MMP‐2 (Wei et al., 2014). We also found that digoxin dose‐dependently attenuated inflammatory cell infiltration (Figure 4A, E). In addition, reduced plaque vulnerability was indicated by increased VSMC accumulation and by the elevated collagen content of the fibrous cap (Figure 4A–D). Previous studies showed that vulnerable plaques are rich in inflammatory cells and exhibit a substantial loss of smooth muscle cells and collagen content (Virmani et al., 2000; Clarke et al., 2008). Our results were consistent with the concept that digoxin reduced plaque vulnerability by increasing the α‐SMA⁺ and collagen⁺ areas and decreasing inflammatory cell infiltration (Figure 4). Down‐regulation of inflammatory cytokines, such as IL‐17A, MMP‐2 and MMP‐9, may enhance the stability of lesions. Although several publications have independently supported the hypothesis that IL‐17A protects against atherosclerosis, the precise role of IL‐17 in atherosclerosis remains controversial. Several recent studies by various investigators have shown additional and, in some cases, more direct evidence supporting the role of IL‐17 in atherosclerosis. It appears that the role of IL‐17 in this disease is context‐dependent and may vary according to the cytokine profile of the local microenvironment in which IL‐17 functions (Taleb et al., 2015). Th17 cell/Treg imbalance plays a crucial role in inflammatory and autoimmune diseases such as diabetes, rheumatoid arthritis and multiple sclerosis. Th17 cells promote various aspects of immune activation, whereas Tregs can dampen inflammatory processes. Previous studies showed that CD4⁺ CD25⁺ Tregs significantly reduced plaque development in an ApoE‐knockout mouse model (Ait‐Oufella et al., 2006; Mor et al., 2007). In the present study, digoxin (especially at the high dose) significantly increased the proportion of Tregs and increased the expression of IL‐10 (Figure 5C, D, and F). Wu et al. demonstrated that digoxin increased the proportion of Tregs in vivo and inhibited the process of IL‐6‐mediated conversion of Tregs into Th17 cells in vitro (2013). The attenuation of the conversion of Tregs into Th17 cells by digoxin may be associated with the increased abundance of Tregs observed in atherosclerotic plaques (Supporting Information Fig. S4) and the spleen (Figure 5C). We also determined the expression of the transcription factors T‐bet, RORγt and Foxp3 in the aortic root. Our results showed that digoxin increased the expression of Foxp3 but not T‐bet or RORγt (Figure 6F). This result supports the hypothesis that digoxin treatment inhibits the activity of RORγt instead of its mRNA or protein expression. Our findings indicate that the balance of Th17 cells/Tregs was disrupted by digoxin treatment, suggesting that an increase in the abundance of Tregs might be involved in the protective effect of digoxin against the development and progression of atherosclerosis.

Taken together, our results provide convincing evidence that digoxin exerts protective effects against the development of atherosclerosis. These effects appear to be mediated by antagonizing RORγ activity, thereby decreasing the plasma levels of total cholesterol, triglycerides and LDL cholesterol in ApoE^−/− mice. Moreover, digoxin suppresses IL‐17A expression and IL‐17A‐related inflammation and increases the abundance of Tregs. The plasma levels of the cardiac glycoside digoxin (Supporting Information Fig. S1) after the final injection (1 mg·kg⁻¹ day⁻¹; low dose) into mice were at or below the therapeutic range for humans (0.5–2.0 ng·mL⁻¹). However, there are important differences between animals and humans; further studies are necessary. Our findings also indicate that antagonizing the activity of RORγ may represent a novel strategy for atherosclerosis treatment.

Author contributions

H.R.S., X.B.M. and Q.T.Z. designed the experiments. Y.C.Z., Y.Z.L. and X.Q.Z. performed the experiments. K.W.Y., R.R.Z., and Y.Z.W. analysed the data. H.T.S. and Y.M. prepared the figures. H.R.S. wrote the main text. All authors reviewed the manuscript.

Conflict of interest

The authors declare no conflicts of interest.

Declaration of transparency and scientific rigour

This Declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research recommended by funding agencies, publishers and other organizations engaged with supporting research.

Supporting information

Figure S1 Plasma levels of digoxin. Plasma levels of digoxin were detected in mice treated via i.p. injection of Dig‐LD (n = 7 each) or Dig‐HD (n = 8 each). The data are expressed as the means ± SEMs at three independent experiments. Dig‐LD (low‐dose digoxin), Dig‐HD (high‐dose digoxin).

Figure S2 Effect of digoxin on expression of RORγ. The mRNA levels of RORγ in liver were determined by RT‐PCR analysis. The data are expressed as the means ± SEMs at five independent experiments. NS indicates P > 0.05; PBS, n = 8; Dig‐LD, n = 7; Dig‐HD, n = 8; Dig‐LD (low‐dose digoxin), Dig‐HD (high‐dose digoxin).

Figure S3 The expression of IL‐17A and CD4⁺ cells in atherosclerotic plaques. IL‐17A was expressed in CD4⁺ cells. Anti‐CD4 (green) and anti‐IL‐17A (red) DAPI (blue) stained for immunofluorescence. Areas of co‐localization are shown in yellow in the merged image (arrows). PBS, n = 5; Dig‐LD, n = 5; Dig‐HD, n = 5; Dig‐LD (low‐dose digoxin), Dig‐HD (high‐dose digoxin).

Figure S4 The expression of Tregs in atherosclerotic plaques. (A) Representative sections of aortic sinus stained with anti‐Foxp3 for Tregs. Black arrows indicate examples of Foxp3‐positive cells. (B) Quantitative analysis of data was shown. The data are expressed as the means ± SEMs. One‐way ANOVA followed by the Holm–Sidak test method. *P < 0.05; PBS, n = 5; Dig‐LD, n = 5; Dig‐HD, n = 5; Dig‐LD (low‐dose digoxin), Dig‐HD (high‐dose digoxin).

Table S1 Real‐time RT‐PCR primer sequences.

Supporting info item

Shi, H. , Mao, X. , Zhong, Y. , Liu, Y. , Zhao, X. , Yu, K. , Zhu, R. , Wei, Y. , Zhu, J. , Sun, H. , Mao, Y. , and Zeng, Q. (2016) Digoxin reduces atherosclerosis in apolipoprotein E‐deficient mice. British Journal of Pharmacology, 173: 1517–1528. doi: 10.1111/bph.13453.