Inactivation of the adenosine A_2A receptor protects apolipoprotein E–deficient mice from atherosclerosis

>Mice

A_2AR^−/− mice in C57BL/6J background ¹⁰ were bred with apoE^−/− (C57BL/6J background) mice to generate Apoe^−/−/A_2AR^−/− mice and their littermate controls. Chimeric mice with or without A_2AR in their bone marrow-derived cells were produced by bone marrow transplantation, as described ¹¹. Mice were fed a western diet for 3 months or 6 months and then euthanized for collection of aortas. All animal experiments and care were approved by the University of Minnesota Animal Care and Use Committee, in accordance with AAALAC guidelines.

Blood lipid and leukocyte analysis

Blood lipid was determined via an automated enzymatic technique (Boehringer Mannheim GmbH). Blood leukocytes were quantified using an automated blood cell counter (Hemavet 850FS, CDC Technologies, Oxford, CT).

Measurement of plasma cytokines

Cytokine levels were determined by multiplex assay on the Luminex (Austin, TX) platform using BioPlex software (Bio-Rad, Hercules, CA) and mouse-specific bead sets according to the manufacturer's instructions (R&D Systems, Minneapolis, MN).

Preparation of mouse aortas and quantification of atherosclerosis

Aortas of atherosclerotic mice were collected and processed for oil red O staining using either en face preparation of whole aortas or cross-sections of aortic sinuses ¹².

Histological analysis of atherosclerotic lesions

Red oil O staining was performed on frozen sections (5 µm thick) of atherosclerotic aortic sinuses. Using specific antibodies, immunostaining to detect expression of macrophage F4/80 (Accurate, Westbury, NY) and phospho-NF-κB p65 (Cell signaling, Danvers, MA) was performed. Slides were examined under a light microscope (Carl Zeiss, Thornwood, NY), and images were digitized into a Macintosh computer. Samples from ten mice were analyzed per group. Quantification was done by dividing the area of positive staining by the total measured lesion area in digitized images.

TUNEL assays

Thioglycollate-elicited peritoneal macrophages were recovered from wt or A_2AR^−/−mice, plated in 8-well culture slides (BD, Bedford, MA) at 0.5 × 10⁶ cells/well, and cultured in DMEM/10% fetal bovine serum for 16 h. The cells were then incubated with 100 µg/mL of human oxidized LDL (ox-LDL) (Biomedical Technologies, Stoughton, MA) with or without the p38 inhibitor SB203580 at 20 µM (Calbiochem, La Jolla, CA) for 20 h. Cells were fixed in 4% paraformaldehyde and apoptosis determined using the Dead End Fluorometric TUNEL System (Promega, Madison, WI) following the manufacturer’s instructions. The same TUNEL staining was also conducted on frozen sections of mouse aortic sinuses to examine the apoptotic cells in atherosclerotic lesions.

Real-time PCR

Total RNA from atherosclerotic arteries was extracted using Trizol reagent (Invitrogen, Carlsbad, CA), and cDNA was synthesized using a first-strand cDNA synthesis kit (Fermentas). PCR was performed with a LightCycler 2.0 thermal cycler (Roche, Indianapolis, IN) using SYBR Green as a double-stranded DNA–specific dye. The relative amount of each gene in each sample was estimated by the ΔΔC_T method. Supplemental table 4 lists the sequences of primers for cytokines.

Western blotting

Mouse peritoneal macrophages were lysed and transferred to a polyvinylidene fluoride membrane. Antibodies against p38, phospho-p38, caspase-3 and GAPDH (Cell Signaling, Danvers, MA) were applied. The blots were incubated with alkaline phosphatase–conjugated secondary antibodies, developed with a chemifluorescence reagent, and scanned by Storm 860 (GE Healthcare).

Electrophoretic mobility shift assay

Nuclear protein was extracted using NucBuster Protein Extraction kit (Novagen, San Diego, CA). A biotin end-labeled double-stranded oligonucleotide (5′-biotin-GGAGAGTGGGGACTACCCCCTCTGCT-3′) and a non-labeled oligonucleotide containing the NF-κB consensus sequence were incubated with the extracted nuclear protein. The samples were subjected to SDS-PAGE and transferred to a nylon membrane. The biotin-labeled DNA was detected with the LightShift Chemiluminescent Electrophoretic Mobility Shift Assay kit (Pierce, Rockford, IL).

Statistical analysis

Statistical analysis was performed with Instat software (GraphPad). Data are presented as mean ± SEM. Data were analyzed with either a one-way ANOVA followed by a Bonferroni correction post-hoc test or a Student’s t-test to evaluate two-tailed levels of significance. The null hypothesis was rejected at P < 0.05.

Atherosclerosis in Apoe^−/−/A_2AR^−/− mice

To determine the role of A_2AR in the development of atherosclerotic lesions in vivo, Apoe^−/−/A_2AR^−/− mice and their littermate Apoe^−/− mice were fed a chow diet or western diet for three months. These mice exhibited no differences in blood pressure, number of circulating leukocytes, differential counts, or blood glucose (Suppl. Table 1, 2, 3). The level of blood alanine aminotransferase (ALT) in Apoe^−/−/A_2AR^−/− mice was four time higher than that in Apoe−/− mice on western diet (Suppl. Table 4). The weight of Apoe^−/−/A_2AR^−/− mice fed a Western diet was 23% higher for males and 12% higher for females compared with sex-matched Apoe^−/− mice fed the same diet. Total blood cholesterol was 45% higher in male Apoe^−/−/A_2AR^−/− mice and 25% higher in females compared with Apoe^−/− mice on both chow and western diets; this increase was due solely to increased LDL cholesterol (Table 1 and Suppl. Table 3). Interestingly, lipid profiles were similar in A_2AR^−/− and wild type mice on western diet. In A_2AR^−/− and wild type mice on both chow and western diet, blood IL-6 levels were not detectable. In contrast, blood IL-6 levels were detectable and much higher in Apoe^−/−/A_2AR^−/− than in Apoe^−/− mice (25 ± 8.2 versus 20 ± 5.8 pg/mL, as shown in Suppl. Table 5). Despite the higher body weight, blood cholesterol, and proinflammatory cytokine levels, atherosclerotic lesion size in the aortas of Apoe^−/−/A_2AR^−/− mice was decreased by 26% in females and 20% in males compared to Apoe^−/− mice (Fig. 1b). In addition, the aortic sinuses displayed much smaller lesions in Apoe^−/−/A_2AR^−/− mice than in Apoe^−/− mice (Fig. 1c).

To examine whether A_2AR deficiency could also protect mice from advanced atherosclerosis, Apoe^−/−/A_2AR^−/− mice and their littermate Apoe^−/− mice were placed on a Western diet for six months. In accordance with the changes in mice fed a Western diet for three months, Apoe^−/−/A_2AR^−/− mice gained more body weight and had a much higher level of blood total cholesterol than Apoe^−/− mice (Table 1). Aortic atherosclerotic lesions in female Apoe^−/−/A_2AR^−/− mice were 51% smaller compared with those in female Apoe^−/− mice, and the lesions in male Apoe^−/−/A_2AR^−/− mice were 55% smaller compared with controls (Fig. 1a). These results confirmed the data obtained from mice fed a Western diet for three months, and demonstrated even greater protection against atherosclerosis in Apoe^−/−/A_2AR^−/− mice during a longer period of atherosclerotic challenge.

The cellular components of atherosclerotic lesions in cross-sections of the aortic sinus were also compared. Macrophages and foam cells were mainly located in the cap and shoulders of lesions in Apoe^−/− and Apoe^−/−/A_2AR^−/− mice, but the total number of macrophages and foam cells in lesions of Apoe^−/−/A_2AR^−/− mice was greatly diminished (Fig. 1d). This was further supported by the lower levels of mRNA encoding the monocyte marker CD68 in lesions of Apoe^−/−/A_2AR^−/− mice than those of Apoe^−/− mice (Fig. 1e).

Atherosclerosis in chimeric mice lacking A_2AR and apoE in bone marrow–derived cells (BMDCs)

To determine the influence of leukocyte A_2AR in the formation of atherosclerotic lesions, we studied atherosclerosis in Apoe^−/− chimeric mice fed a Western diet for three months. Apoe^−/− mice lacking A_2AR in their BMDCs did not differ from Apoe^−/− mice in body weight or blood cholesterol level (data not shown). In the aortic sinuses, a 30% reduction was observed in the average size of lesions in chimeric mice lacking A_2AR in their BMDCs compared to that in controls (Fig. 1f), suggesting that protection against atherosclerosis in Apoe^−/−/A_2AR^−/− mice was mainly due to A_2AR deficiency in BMDCs.

The presence of macrophages in atherosclerotic lesions of chimeric mice was also assessed; Apoe^−/− mice lacking A_2AR in their BMDCs demonstrated significantly fewer macrophages in lesions compared with Apoe^−/− mice (Fig. 1g).

Inflammatory status of atherosclerotic lesions in Apoe^−/−/A_2AR^−/− mice

Atherosclerosis is a chronic inflammatory disease, and disease progression is usually accompanied by increased inflammation ². A_2AR^−/− mice and A_2AR-deficient macrophages exhibited increased inflammatory phenotype following inflammatory stimulation ^{6, 7}. Since Apoe^−/−/A_2AR^−/− mice developed small atherosclerotic lesions, we speculated that A_2AR-deficient macrophages might react to modified LDL differently from their response to other inflammatory stimuli. To test this possibility, we examined the inflammatory response of A_2AR-deficient macrophages to ox-LDL in an in vivo peritonitis model. On the third day of thioglycollate-induced peritonitis, mice were injected intraperitoneally with ox-LDL. Peritoneal macrophages were collected 30 minutes after the ox-LDL injections. As shown by an electrophoretic mobility shift assay, both wt and A_2AR-deficient macrophages displayed significant levels of nuclear P65/P50 binding to the NF-κB consensus sequence, indicating activation of the NF-κB pathway in thioglycollate-elicited macrophages. Compared with wt macrophages, A_2AR-deficient macrophages showed increased NF-κB activation before and after ox-LDL treatment (Fig. 2a).

To determine the level of NF-κB activation in foam cells present in atherosclerotic lesions of Apoe^−/−/A_2AR^−/− mice, sections of atherosclerotic lesions were immunostained with phosphor-p65-specific antibody. Phosphorylation of p65 is an indicator of NF-κB activation. Among the macrophages/foam cells present in lesions, many more cells demonstrated positive staining for phospho-p65 in lesions of Apoe^−/−/A_2AR^−/− mice than in those of Apoe^−/− mice (Fig. 2b). In addition to NF- κB signaling, we also determined the mRNA levels of proinflammatory cytokines in lesions by real-time RT-PCR. The levels of IL-1b and IL-6 mRNA were much higher in atherosclerotic lesions of Apoe^−/−/A_2AR^−/− mice than in those of Apoe^−/− mice (Fig. 2c). These results indicate that, in an atherosclerotic environment, A_2AR-deficient macrophages exhibited an inflammatory phenotype. Notably, the mRNA level of IL-10, an anti-inflammatory cytokine, was also increased in lesions of Apoe^−/−/A_2AR^−/− mice.

Apoptotic foam cells in atherosclerotic lesions of Apoe^−/−/A_2AR^−/− mice

Apoptosis of macrophages or foam cells during the early stages of atherosclerosis decreases atherosclerosis ^13–15. To investigate whether this was the mechanism responsible for suppressed atherosclerosis in Apoe^−/−/A_2AR^−/− mice, we first performed TUNEL-staining to detect apoptotic cells on cross-sections of atherosclerotic lesions. In lesion areas containing F4/80-positive macrophages, many more cells were positive for TUNEL-staining in lesions of Apoe^−/−/A_2AR^−/− mice than in those of Apoe^−/− mice (Fig. 3a).

Macrophages in the peritoneal cavities of atherosclerotic mice with thioglycollate-induced peritonitis differentiate into foam cells ¹⁶. Using this in vivo foam cell formation model, wt and A_2AR-deficient foam cells were generated and assayed by flow cytometry. Among the F4/80-positive foam cells, the percentage of annexin V-positive but PI-negative cells was 12% for foam cells from Apoe^−/−/A_2AR^−/− mice and 5% for foam cells from Apoe^−/− mice (Fig. 3b). Similar results were obtained by TUNEL-staining (Fig. 3c). Caspase-3 is a critical executioner of apoptosis, and the cleaved p17 fragment represents its active form. The p17 fragment of caspase-3 was detected in foam cells by western blot. The level of p17 was much higher in A_2AR-deficient foam cells than wt cells (Fig. 3d).

Activation of p38 MAPK in A_2AR-deficient macrophages

Activation of A_2AR increases intracellular cAMP ⁴, which, in turn, inhibits activation of the intracellular signaling molecule p38 MAPK via the cAMP response element–binding protein–induced dynein light chain ¹⁷. In an in vitro assay using isolated peritoneal macrophages, p38 MAPK activation in response to ox-LDL stimulation was much more robust in A_2AR-deficient than in wt macrophages (Fig. 4a). Furthermore, the level of ox-LDL-induced active caspase-3 was much higher in A_2AR-deficient macrophages than wt macrophages (Fig. 4b). To determine whether p38 activation was a possible mechanism for the increased apoptosis of A_2AR-deficient macrophages, A_2AR-deficient macrophages were first pretreated with the p38 inhibitor SB203580, followed by incubation with ox-LDL for induction of apoptosis. Incubation with ox-LDL elicited apoptosis in 20% of A_2AR-deficient macrophages and 9% of wt macrophages. SB203580 pretreatment decreased ox-LDL-mediated apoptosis in both cases, but this decrease was more pronounced for A_2AR-deficient macrophages than wt cells. The percentage of apoptotic A_2AR-deficient macrophages was reduced almost to the level measured for wt macrophages, indicating that increased p38 activation is the underlying mechanism for apoptosis of A_2AR-deficient macrophages (Fig. 4c).