Omega-3 Fatty Acids Ameliorate Atherosclerosis by Favorably Altering Monocyte Subsets and Limiting Monocyte Recruitment to Aortic Lesions

Mice and diets

LDLr^−/− and apoE^−/− mice (C57BL/6 background) were purchased from The Jackson Laboratory (Bar Harbor, Maine, USA). Mice were housed in a specific pathogen-free facility with a 12-hour light/dark cycle. All protocols and procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of Wake Forest School of Medicine. At six to eight weeks of age, female mice were fed atherogenic diets containing 10% calories as PO and 0.2% cholesterol, supplemented with an additional 10% of calories as palm oil (PO), echium oil (EO; containing 18:4 n-3) or fish oil (FO) for 16 (LDLr^−/−) or 12 (apoE^−/−) weeks. See supplemental methods for detailed description of the oils. The synthetic diets were prepared by the diet kitchen in the Department of Pathology at Wake Forest School of Medicine as previously described²³. The fatty acid composition of the PO, EO and FO diets has been previously published²⁴. A standard chow diet, Prolab IsoPro RMH 3000 (LabDiet) was used as an additional control.

Quantification of aortic root lesion area

The aortic root was sequentially sectioned following an adaptation of the method of Daugherty et al²⁵. Once the aortic valves were visible, 8-μm aortic root sections were retained for the length of 640 μm. The sections were placed on a series of 10 slides. This resulted in 8 sections per slide, with a distance of 80 μm apart. Oil red O and hematoxylin staining was performed to assist in tissue visualization. Images were acquired with a Nikon DigitalSight DSFi1 camera using NIS Elements (Nikon) software. Intimal area measurements were obtained with ImagePro 6.2 software (Media Cybernetics). We defined intimal area as the lesion area between the internal elastic lamina and the luminal surface of the aorta. Three 8-μm aortic root sections were blindly quantified from each mouse, from a single slide. Sections were examined in the aortic root starting from the proximal region (when the three aortic valves appear) and proceeding distally, unless a section was deemed immeasurable, due to a fold, tear, etc. in which case the adjacent section was measured. The mean coefficient of variation for these measurements from individual mice on atherogenic diets was 10.7%.

Plasma lipid and lipoprotein analyses

Blood was collected by submandibular vein puncture in non-fasted mice in the presence of heparin. Plasma from these blood samples was isolated using a table top centrifuge (10,000 rpm for 15 min at 4°C). Plasma cholesterol or triglyceride concentrations were measured enzymatically using commercially available kits (Wako). Plasma lipoproteins were fractionated by fast protein liquid chromatography (FPLC) and cholesterol distribution among lipoproteins was measured by enzymatic assay²⁶.

Flow cytometry

Peripheral blood was obtained by submandibular vein puncture for circulating leukocyte analysis. For splenic cells, tissue was digested with an enzyme cocktail as published²⁷. The cell suspension was subsequently passed through a 70-μm nylon cell strainer (BD Falcon). Red blood cells were removed from flow cytometry preparations by treatment with ACK lysing buffer (Gibco). The remaining white blood cells were incubated with the following mAbs: CD11b-APC-Cy7 clone:M1/70 (BioLegend), CD115-APC clone:AFS98, CD62L-PerCP-Cy5.5 clone:MEL-14 (eBiosciences), Ly6G-PE clone:1A8, CD11c-PE-Cy7 clone:HL3, Ly6C-FITC or PE clone:AL-21 (BD Pharmingen). Data were acquired on a BD FACSCanto II instrument (BD Biosciences) and analyzed using FacsDiva software (BD Biosciences).

Monocyte labeling and quantification of monocyte recruitment

Monocytes were labeled using the Gr-1^lo method of Tacke et al²⁸. Briefly, 1-μm Fluoresbrite Yellow Green (YG) microspheres (2.5% solids [wt/vol], Polysciences, Inc.) were diluted 1:4 in PBS and injected retro-orbitally into anesthetized mice. Blood was drawn from recipient mice the day after labeling to evaluate the total number of bead-containing monocytes per mouse. This approach predominantly labels Ly6C^lo monocytes, although some transient labeling of Ly6C^hi monocytes has been observed²⁸.

Aortic root tissue sections were taken as described above and the number of beads confined to atherosclerotic lesions was quantified per section. A total of 8 sections, from a single slide, representative of the length of the aortic root, were analyzed per mouse. Beads were counted manually, at 20X magnification. To adjust for slight variations in labeling, data were normalized for individual mice based on the percentage of blood monocytes labeled. For example, if an individual mouse had 11% of blood monocytes labeled, and 15 bead-containing cells per section were counted, the calculation would be as follows 15/0.11 = 136. This adjustment gave a normalized value, which represented the actual number of monocytes expected to have entered the lesion.

Immunohistochemistry

Aortic roots were mounted in OCT and sectioned as described under the quantification of aortic root lesion area section of the methods. Aortic root cross-sections were fixed with acetone and incubated with an anti-CD68 antibody (1:75, overnight) from AbD Serotec (clone:FA11), followed by staining with an alkaline phosphatase-conjugated secondary antibody (goat anti-rat, 1:50, 30 minutes) from AbD Serotec; control slides contained no primary antibody. Four 8-μm aortic root sections from each mouse were blindly analyzed for the percent CD68-positive area.

Statistical analyses

All data were subjected to parametric analysis and passed the D’Agostino and Pearson omnibus normality test. Data are presented as mean ± SEM and were tested for significant differences (p< 0.05) using 1-way ANOVA with Tukey’s post-hoc analysis. Pearson’s correlation coefficient was calculated after the line of best fit was determined using linear regression analysis. All the analyses were performed with GraphPad Prism 5.01 (GraphPad Software Inc., San Diego, CA).

n-3 FAs reduced atherosclerosis in LDLr^−/−, but not apoE^−/− mice

As previously described^{23, 29}, n-3 FAs reduced atherosclerosis development in LDLr^−/− mice (Figure 1A and 1B). We observed a stepwise decrease in atherosclerosis that correlated with the degree of enrichment of long chain n-3 FAs following either EO or FO supplementation for 16 weeks²⁴. Standard rodent chow has negligible amounts of cholesterol, and as expected, this group had the lowest intimal lesion area. Interestingly, parallel studies in apoE^−/− mice resulted in no difference in atherosclerosis among diet groups after 12 weeks of diet consumption (Figure 1A and 1C). As expected, apoE^−/− chow-fed mice had reduced atherosclerosis compared to mice fed atherogenic diets. We also measured aortic cholesterol content in the entire aorta (arch, thoracic, abdominal) in LDLr^−/− and apoE^−/− mice and observed similar trends as with aortic root intimal area (Supplemental Figure 1A and 1B). Furthermore, aortic root intimal area positively correlated with aortic cholesterol ester content in mice of both backgrounds (Supplemental Figure 1C and 1D). Thus, dietary n-3 FAs reduced atherosclerosis in LDLr^−/−, but not apoE^−/− mice.

Differential plasma cholesterol responses to diet in LDLr^−/− and apoE^−/− mice

In agreement with previous reports^{23, 29}, n-3 FAs blunted hypercholesterolemia in the LDLr^−/− mice (Figure 2A and 2C), with equivalent reductions in EO and FO-fed mice compared to PO-fed mice. Plasma cholesterol measurements in apoE^−/− mice uncovered an unexpected finding (Figure 2B). The FO-fed mice had the highest exposure to cholesterol over time as illustrated using area under the curve (AUC) analysis for total plasma cholesterol, with all other groups exhibiting similar levels (Figure 2D). FPLC size fractionation of plasma revealed that the increase in total plasma cholesterol concentrations in the FO diet group was primarily due to an increase in VLDL-size particles (Supplemental Figure 2B). There is one other report of an increase in plasma cholesterol in n-3 FA-fed apoE^−/− mice²⁹. Plasma triglyceride concentrations increased in FO vs. PO-fed apoE^−/− mice (64±6 vs. 22±7 mg/dL, respectively; n=7–9 per group), consistent with previous reports^{11, 29–33}. Thus, differential plasma cholesterol and triglyceride responses to diet were observed in LDLr^−/− and apoE^−/− mice.

Plasma cholesterol does not correlate with disease burden in apoE^−/− mice

Atherosclerosis was similar among PO, EO, and FO-fed apoE^−/− mice (Figure 1C, Supplemental Figure 1B), despite two-fold higher plasma cholesterol AUC in the FO group (Figure 2D). Furthermore, linear regression analysis demonstrated a significant association between intimal area and plasma cholesterol in the LDLr^−/− (Figure 2E), but not in apoE^−/− mice (Figure 2F). These results suggest a unique property of FO that counterbalanced the hypercholesterolemia induced by FO feeding in apoE^−/− mice to promote atherosclerosis progression.

n-3 FAs favorably alter Ly6C^hi monocytes in atherosclerotic mice

Despite a substantial lowering of triglycerides^{6, 34}, n-3 FAs typically do not alter plasma cholesterol in humans and may, in fact, increase LDL cholesterol and lower HDL cholesterol in some individuals^{5, 6}. We investigated the possibility that n-3 FAs could reduce the Ly6C^hi monocyte subset, which infiltrate into atherosclerotic lesions²⁰, since reduced macrophage staining has been reported with n-3 fatty acid supplementation^10–12. We identified monocytes as CD11b⁺Ly6G⁻CD115⁺ (Supplemental Figure 3A). In agreement with other findings^{18, 20, 21}, Ly6C^hi cells preferentially expressed CD62L, whereas CD11c expression was found primarily on Ly6C^lo monocytes (Supplemental Figure 3B). We observed no significant difference in the percentage of monocytes in the circulation of LDLr^−/− mice during the 16-week atherosclerosis progression experiment based on atherogenic diet fat type, although chow-fed mice had significantly less monocytosis compared to PO and EO-fed mice (Supplemental Figure 4A and C). In agreement with another 12-week study in apoE^−/− mice¹⁹, the saturated fat-containing atherogenic diet (i.e., PO) did not significantly alter the percentage of monocytes in the circulation compared to chow-fed controls (Supplemental Figure 4B and D). However, using AUC analysis, we did observe an increase in percentage of monocytes in FO-fed mice compared to PO and chow-fed mice (Supplemental Figure 4D). With regard to subset distribution, the proportion of monocytes in the Ly6C^hi subset was not altered over 16 weeks in LDLr^−/− mice (Supplemental Figure 4E). However, after 12 weeks of experimental diet feeding both FO and chow-fed apoE^−/− mice showed a reduction in the Ly6C^hi subset compared to PO-fed mice, indicating a more favorable monocyte profile (Supplemental Figures 3A and 4F).

The inflammatory profile of splenic monocytes parallels atherosclerosis

The percentage of monocytes in the spleen, identified as a reservoir for circulating monocytes³⁵, was similar among diet groups in LDLr^−/− and apoE^−/− mice (Figure 3B and 3C). However, in the LDLr^−/−, n-3 FAs from EO and FO reduced the Ly6C^hi subset to levels of chow-fed mice (Figure 3A and 3D). In the apoE^−/− background, only the mice fed chow were different than the PO group (Figure 3E). In fact, the changes in the Ly6C^hi subset were strikingly similar to changes in intimal lesion area (compare Figures 3E and 1C). Furthermore, we found a significant positive association between percentage of Ly6C^hi splenic monocytes and intimal area in both LDLr^−/− and apoE^−/− mice (Figure 3F and Figure 3G).

n-3 FAs protect against neutrophilia in hypercholesterolemic mice

Recent advances in the field have also highlighted the importance of hypercholesterolemia-induced neutrophilia in atherosclerosis³⁶. We measured blood neutrophils in LDLr^−/− and apoE^−/− mice after 16 and 12 weeks on atherogenic diets, respectively (Supplemental Figure 5A and 5B). Just as n-3 FAs dissociated hypercholesterolemia from Ly6C^hi monocytosis, we also observed dissociation of hypercholesterolemia from neutrophilia after 16 weeks in the LDLr^−/− mice, where n-3 FAs reduced blood neutrophils to levels of chow-fed mice. In apoE^−/− mice, we detected a decrease in neutrophils with both EO and FO fed mice after 6 and 9 weeks on diet. Similar findings were observed in the spleen (Supplemental Figure 5C and 5D), a compartment that may reflect cumulative effects of dietary FAs on systemic neutrophil levels.

n-3 FAs reduce monocyte infiltration into atherosclerotic lesions

To determine whether n-3 FAs would decrease monocyte trafficking into atherosclerotic lesions, we performed monocyte-labeling experiments after maintaining LDLr^−/− or apoE^−/− mice on diet for 16 or 12 weeks, respectively. Using the procedure developed by Randolph and colleagues^{21, 28}, we phagocytically labeled monocytes with 1μm fluorescent beads. Within 24 hrs of injection, 0.7–1.2% of blood leukocytes were labeled in the LDLr^−/− diet groups, and 1.0–1.2% of blood leukocytes were labeled in the apoE^−/− groups. Of these bead⁺ cells, 63–81% (mean, 79.4%) were monocytes in LDLr^−/− mice, and 78–90% (mean, 85.3%) were monocytes in the apoE^−/−. This translated to 6.5–13.3% (mean, 9%) of monocytes carrying a fluorescent label in LDLr^−/− mice. In apoE^−/− mice, 6.0–11.6% (mean, 8%) of monocytes were bead⁺. The remainder of bead-labeled non-monocytic cells in blood were B cells (CD19⁺; 66%) and neutrophils (Ly6-G⁺; 34%), similar to previous findings using this technique²⁸.

The number of recruited monocytes (visualized as fluorescent beads, see Supplemental Figure 6) within the lesion area of the aortic root in mice fed PO, EO, or FO diets was compared to chow-fed controls in LDLr^−/− mice (Figure 4A) and apoE^−/− mice (Figure 4B). Because of slight variations in labeling (6.5–13.3% in LDLr^−/− mice, 6.0–11.6% in apoE^−/−), data were normalized for individual mice based on the percentage of blood monocytes labeled. This adjustment gave a normalized value, which represented the actual number of monocytes entering the lesion (i.e. normalized frequency). Normalization revealed a decrease in bead⁺ cells per intimal area per section in all groups compared to PO in LDLr^−/− mice (Figure 4C). In contrast, bead⁺ monocytes did not differ among diet groups in the apoE^−/− background (Figure 4D), although there was a trend towards a decrease in chow vs. PO-fed mice. Of interest, numbers of monocytes trafficking to lesions were higher in the apoE^−/− compared to the LDLr^−/− (compare scales Figure 4A and 4C versus 4B and 4D), perhaps reflecting the heightened inflammatory response in the apoE^−/− model^{37, 38}. Together these data indicate that n-3 FAs reduce infiltration of monocytes into atherosclerotic lesions to an extent related to the amount of atherosclerosis.

n-3 FAs decrease macrophage staining in atherosclerotic lesions

In concordance with our monocyte infiltration data, analysis of CD68 immunohistochemistry in LDLr^−/− mice (Figure 5A and 5B) revealed that n-3 FA-fed mice exhibit reduced (30–44%) macrophage staining compared to PO-fed mice, equivalent to chow-fed mice.