Hyperhomocysteinemia promotes inflammatory monocyte generation and accelerates atherosclerosis in transgenic cystathionine β-synthase deficient mice

Abstract

Background

Hyperhomocysteinemia (HHcy) is an independent risk factor for cardiovascular diseases. Monocytes display inflammatory and resident subsets, and commit to specific functions in atherogenesis. In this study, we examined the hypothesis that HHcy modulates monocyte heterogeneity and leads to atherosclerosis.

Methods and Results

We established a novel atherosclerosis susceptible mouse model with both severe HHcy and hypercholesterolemia, in which the mouse cystathionine β-synthase (CBS) and apolipoprotein E (apoE) genes are deficient, and an inducible human CBS transgene is introduced to circumvent the neonatal lethality of the CBS deficiency (Tg-hCBS apoE^−/− Cbs^−/− mice). Severe HHcy accelerated atherosclerosis and inflammatory monocyte/macrophage accumulation in lesions and increased plasma TNFα and MCP-1 levels in Tg-hCBS apoE^−/− Cbs^−/− mice fed a high fat diet. Furthermore, we characterized monocyte heterogeneity in Tg-hCBS apoE^−/− Cbs^−/− mice and another severe HHcy mouse model (Tg-S466L Cbs^−/−) with a disease relevant mutation (Tg-S466L) that lacks hyperlipidemia. HHcy increased monocyte population and selective expansion of inflammatory Ly-6C^hi and Ly-6C^mid monocyte subsets in blood, spleen and bone marrow of Tg-S466L Cbs^−/− and Tg-hCBS apoE^−/− Cbs^−/− mice. These changes were exacerbated in Tg-S466L Cbs^−/− mice with aging. Addition of L-homocysteine (100–500 μM), but not L-cysteine, maintained the Ly-6C^hi subset and induced the Ly-6C^mid subset in cultured mouse primary splenocytes. Homocysteine-induced differentiation of Ly-6C^mid subset was prevented by catalase plus SOD, and the NAD(P)H oxidase inhibitor, apocynin.

Conclusions

HHcy promotes differentiation of inflammatory monocyte subsets and their accumulation in atherosclerotic lesions via NAD(P)H oxidase-mediated oxidant stress.

Introduction

Despite advances in our understanding of the pathogenesis of cardiovascular disease (CVD), the established risk factors do not fully account for its occurrence. Abundant clinical and epidemiological studies have revealed the close relationship between hyperhomocysteinemia (HHcy) and the increased frequency of CVD.^{1, 2} Meta-analysis has shown that an increase of 5 μM in plasma homocysteine (Hcy) levels enhances the risk of CVD by 1.6- to 1.8-fold, which is similar to that seen with an increase of 20 mg/dL (0.52mM) in cholesterol concentration.³ Therefore, HHcy has been identified as a potent risk factor for CVD, equivalent to smoking and hyperlipidemia.⁴

During the last decade, profound atherogenic effects of HHcy have been established in experimental models. We and others have demonstrated that HHcy promotes atherogenesis by stimulating vascular smooth muscle cell proliferation,⁵ inhibiting endothelial cell growth and post-injury reendothelialization, impairing endothelial relaxation, accelerating neointimal formation,^6–10 and inhibiting high density lipoprotein biosynthesis.¹¹ It has been suggested that autoimmune response induced by HHcy may be intimately related to CVD.¹² However, the fundamental basis of HHcy induced immune response and its contribution to atherosclerosis remains unknown.

Support for atherosclerosis as a systemic inflammatory process has grown significantly over the recent years. Monocytes migrate from the bloodstream and give rise to tissue macrophages and dendritic cells. Macrophages accumulate in the vessel wall, take up modified lipoprotein, and participate in atherogenesis. We now know that circulating monocytes, in both humans and mice, are heterogeneous and that different monocyte subsets exhibit distinct pathophysiological roles.¹³ The inflammatory monocyte subset is recruited into inflammatory tissues, while the resident monocyte subset enters normal tissues. Emerging new evidence supports the notion that monocyte heterogeneity bridges the pathogenic mechanism of dyslipidemia and the inflammatory response in the progression of atherosclerosis.^{13, 14} How HHcy influences monocyte heterogeneity has not been investigated.

Cystathionine β-synthase (CBS) is a key enzyme in the transsulfuration pathway that catabolizes Hcy. The S466L mutation of the CBS gene was identified in homocystinuric patients and is associated with thrombosis and elevated levels of plasma Hcy (~167μM).¹⁵ These findings support the clinical applicability of using genetic animal models to study the effect of HHcy on CVD.

Meada and colleagues created a mouse model for Cbs^−/−, in which mice have plasma Hcy levels of about 200 μM. However, most Cbs^−/− mice have a short life span and die before weaning. We have generated the apoE^−/− Cbs^−/− mice by cross-breeding apoE^−/−mice with Cbs^−/+ mice, and are the first to demonstrate that severe HHcy accelerates atherosclerosis independent of dietary manipulation in apoE^−/− Cbs^−/− mice.¹⁶ These studies were difficult and relied on breeding a large colony of mice. To circumvent neonatal lethality problem in the Cbs^−/− mice, the Kruger laboratory created a transgenic mouse (Tg-hCBS), in which the human CBS cDNA is under the control of a zinc inducible metallothionein promoter.¹⁷ By crossing these animals to Cbs mice and supplying zinc in the drinking water during pregnancy and lactation, Tg-hCBS was able to rescue the neonatal lethal phenotype and generated Tg-hCBS Cbs^−/− animals. At weaning, zinc was removed and the expression of Tg-hCBS was stopped, which caused the animals to develop severe HHcy (plasma Hcy >150μM). A second mouse model was developed in which a mutant human CBS transgene (Tg-S466L) was used in a similar manner.¹⁸ In this study, we also generated an atherosclerosis susceptible HHcy model (Tg-hCBS apoE^−/− Cbs^−/−). Using these newly established models of severe HHcy, we now report our findings testing the novel hypothesis that HHcy modulates monocyte heterogeneity and accelerates atherogenesis.

Materials and methods

Gene-targeted mice and diet

All mice are in a C57B/L6 strain background. The transgenic mice, Tg-hCBS and Tg-S466L, were established as described previously.^{17, 18} Pups were genotyped at day 10 for both the human transgene and the mouse Cbs deficiency as previously described.¹⁶ The Tg-hCBS apoE^−/− Cbs^−/− mouse line was generated by crossbreeding Tg-hCBS Cbs^−/− mice with apoE^−/− mice (Jackson Laboratory, Bar Harbor, ME). These animals were all born to mothers drinking zinc water to induce transgene expression. ZnCl₂ was withdrawn after weaning at one month of age.

Animals were fed standard rodent chow diet (0.43% methionine, TD 2018SX, Harlen Teklad, Madison, WI) before dietary intervention. Age matched littermates with evenly distributed sex were selected for each experimental group. Control mice for all experiments were sibling transgene positive mice that were in either Cbs^−/+ or Cbs^+/+ background. The Cbs^−/+ and Cbs^+/+ mice in the control groups have similar plasma Hcy levels (data not shown).^{17, 18} This might be related to the human CBS transgene which provides partial function to metabolize Hcy in both lines. For lesion analysis, Tg-hCBS apoE^−/− Cbs^−/− mice and their sibling control mice were switched to a high fat (HF) diet (0.2% cholesterol, 21.2% fat, 0.75% methionine, 1.43 mg/g choline, 2.23mg/kg folate, 32.5μg/kg B12, 22.25 mg/kg B6; TD88137, Harlan Teklad, Madison, WI) at 8 weeks of age and were maintained on this diet for an additional 8 weeks. The mouse protocols were approved by the Temple University Institutional Animal Care and Use Committee.

Mouse primary splenocyte culture and chemical treatments

Mouse primary splenocytes from two month old wild-type mice (C57B/L6) were isolated as described in the supplemental methods. Splenocyte suspensions were plated at a density of 1.5×10⁶ cells/well in 24-well plates in RPMI1640 medium (Hyclone, Logan, UT) supplemented with 10% FCS, 2mM L-glutamine,10mM HEPES, 0.1mM nonessential amino acid, 1mM sodium pyruvate, 50μM 2-mercaptoethanol and 1% antibiotics (penicillin, streptomycin). To examine the effect of Hcy on Ly-6C monocyte subset survival, cells were treated with L-Hcy (100–500μM), L-cysteine (L-Cys, 100–500μM) as the sulfhydryl containing amino acid control, or recombinant mouse interferon-γ (rIFNγ, 100U/ml, R&D Systems Inc., Minneapolis, MN) as a classic inducer of Ly-6C expression, at 0 hr and continued for 24 hr. To determine the effect of Hcy on Ly-6C monocyte subset differentiation, splenocytes were treated with L-Hcy or L-Cys at 24 hr after plating for an additional 48 hr.

To illustrate the mechanism mediating HHcy-promoted inflammatory phenotypic shift in monocytes, splenocytes were cultured for 24 hr and then treated with the antioxidants catalase-polyethylene glycerol (CAT-PEG, 250U/mL) plus superoxide dismutase-PEG (SOD-PEG, 250U/mL), the peroxynitrite inhibitor uric acid (60μg/mL), the NAD(P)H oxidase inhibitor apocynin (100μM), the xanthine oxidase inhibitor allopurinol (30μg/mL), or the nitric oxide synthase (NOS) inhibitor N (G)-nitro-L- arginine methyl ester (L-NAME, 1mM) 1 hr before exposure to L-Hcy (500μM). To test hypomethylation mechanism, cells were treated with 50 μM L-Hcy in the presence of 25μM adenosine and 10 μM erythro-9-(2-hydroxy-3-nonyl)-adenine hydrochloride, an adenosine deaminase inhibitor to stabilize adenosine as we reported previously.^{6, 19} Adenosine is a normal constituent of all body fluids and is not stable in the cell culture medium. The addition of adenosine and its stabilizer facilitate the conversion of Hcy to S-adenosylhomocysteine, a potent inhibitor of methyltransferase, and sensitize a hypomethylation mechanism.^{6, 19}

L-Hcy was freshly prepared by reducing L-homocystine with a 2-fold molar excess of dithiothreitol for 30 minutes at 37°C, pH 8.0 as described.²⁰ All chemicals, if not specified above, were purchased from Sigma-Aldrich (St. Louis, MO).

The following methods are described in Supplemental Methods:

1. Body and organ weight measurements

2. Hcy and methionine (Met) measurements

3. Aortic sinus cross-sectioning and lesion analysis

4. Plasma tumor necrosis factor-α and monocyte chemoattractant protein-1 analysis

5. Peripheral blood, spleen, and bone marrow cell isolation

6. Flow cytometry analysis

7. Statistical Analysis

Results

Severe HHcy reduced body weight in Tg-hCBS apoE^−/− Cbs mice

In Tg-S466L Cbs^−/− mice at 6 months of age, plasma Hcy was 213.9±67.8μM and this did not change at 8 months of age. In contrast, control animals had plasma Hcy of 4.5μM at 6 months and 4.1μM at 8 months of age (Figure 1A). Methionine (Met) levels were slightly elevated in Tg-S466L Cbs^−/− mice and slightly increased with aging (56.6±13.9μM vs 42.3±1.2μM at 6 months and 68.2±5μM vs 47.9±11.3μM in the controls at 8 months of age). Hcy levels in Tg-hCBS apoE^−/− Cbs^−/− mice fed a HF diet were increased to 179.5±24.1 μM (p<0.001, Figure 1B). In this strain, Met levels were not significantly different from that of the control mice. In Tg-hCBS apoE^−/− Cbs^−/− mice fed a regular diet, plasma levels of Hcy and Met were increased to 200.7±19.4μM (p<0.001) and 107.7±16.1μM (p=0.004), respectively (Figure 1C).

Figure 1. Plasma levels of Hcy and Met, and organ weight in transgenic CBS deficient mice.

Tg-hCBSapoE^−/− Cbs mice were fed a HF diet at 8 weeks of age for an additional 8 weeks. The Tg-S466L Cbs and Tg-hCBS apoE^−/−Cbs mice were fed a rodent chow. A, B & C, Plasma levels of Hcy and Met. D, Body weight, E, F, G and H. heart, liver, spleen and kidney weights relative to tibia lengths, respectively. Values are mean±SD. P values from independent t test. n=10–12, Hcy, homocysteine; Met, methionine, HF, high fat.

In Tg-hCBS apoE^−/− Cbs^−/− mice fed a HF diet, severe HHcy was associated with a 15.6% reduction in the body weights (p=0.04, Figure 1D). However, the weights of heart, liver and kidney (Figure 1E, 1F and 1H) were not changed in Tg-hCBS apoE^−/− Cbs^−/− mice. Spleen weight had a trend to decrease that did not reach statistical significance (Figure 1G).

Severe HHcy accelerated atherosclerosis and enhanced Ly-6C monocyte/macrophage accumulation in lesions in Tg-hCBS apoE^−/− Cbs mice

To confirm the atherogenic effects of HHcy, Tg-hCBS apoE^−/− Cbs^−/− and their sibling control mice were fed a HF diet and examined for atherosclerotic lesions by cross-section analysis of the aortic sinus (Figure 2A). Atherosclerotic lesions were significantly increased in Tg-hCBS apoE^−/− Cbs^−/− mice (63.8±14.6×10⁴μm², p=0.004) compared with those of their sibling controls (39.2±15×10⁴μm²). The lesion percentage of aortic sinus was increased from 29.7±8.7% in the controls to 39.7±6.5% (p=0.019) in Tg-hCBS apoE^−/− Cbs^−/− mice. The atherosclerotic lesion areas were positively correlated with plasma Hcy levels (r=0.620, p=0.008) (Figure 2C).

Figure 2. HHcy accelerated atherosclerotic lesions, enhanced MC/Mφ and Ly-6C positive MC/Mφ accumulation in the lesion in Tg-hCBS apoE^−/− Cbs mice.

Mice were fed a HF diet at 8 weeks of age for an additional 8 weeks. A. Photomicrographs of mouse aortic sinus cross sections stained with oil red O and counterstained with hematoxylin. B. Photomicrographs of mouse aortic sinus cross sections immunostained with MOMA-2 (MC/Mφ marker, green), Ly6C (red), DAPI (blue). Merge (yellow to orange) shows the accumulation of Ly-6C positive MC/Mφ in the lesion. C, D, E & F. Quantitative analysis of lesions in the aortic sinuses. Atherosclerotic lesion area was defined as the neointimal region between the lumen and IEL (C), MC/Mφ area (D), Ly-6C positive area (E), and Ly-6C positive MC/Mφ area (F). Correlation analysis of Hcy levels with lesion area, MC/Mφ area, Ly-6C positive area, and Ly-6C positive MC/Mφ area were shown in scattered dot graphs. Each data point represents one mouse. Values are mean ± SD, n=8–9. MC, monocytes; Mφ, macrophages; IEL, internal elastic lamina.

We further examined the involvement of monocytes in HHcy-related atherosclerosis. Aortic sinus cross sections were stained with MOMA-2 antibody, a marker for monocyte/macrophage, and Ly-6C, a marker for the inflammatory monocyte subset (Figure 2B). Monocyte/macrophage area was significantly increased in the lesion of Tg-hCBS apoE^−/− Cbs^−/− mice (47.7±7.6×10⁴μm², p<0.001) compared with that in their controls (18.5±10.7×10⁴μm²). The percentage of monocyte/macrophage area in the lesion was increased from 46.9±20.7% in the controls to 76.3±12.3% (p=0.005) in Tg-hCBS apoE^−/− Cbs^−/− mice. Lesion monocyte/macrophage area was positively correlated with plasma Hcy levels (r=0.687, p=0.003) (Figure 2D).

The Ly-6C positive area (23.6 ± 4.6×10⁴ μm², p=0.001) and the percentage of Ly-6C positive area in the lesion (38.7±11.4%, p<0.001) were elevated in Tg-hCBS apoE^−/−Cbs^−/− mice, compared with that of their controls (6.8±4.1 × 10⁴μm², 17.9±5.1%). Ly-6C positive areas were positively related to plasma Hcy levels (r=0.722, p=0.002).

The Ly-6C and MOMA-2 double positive areas represented the Ly-6C inflammatory monocyte/macrophage subset, which was increased in the lesion of Tg-hCBS apoE^−/−Cbs^−/− mice (10.2±3.8×10⁴μm² and 17.5 ± 8.4% of the lesion, p=0.001 and p=0.004) from their controls (1.6±0.8 × 10⁴μm², 4.8±1.9% of the lesion). Ly-6C positive monocyte/macrophage accumulation in the lesion was positively correlated with Hcy levels (r=0.678, p=0.005) (Figure 2F).

Severe HHcy increased plasma TNF-α and MCP-1 levels in Tg-hCBS apoE^−/− Cbs mice

Severe HHcy increased plasma levels of the proinflammatory cytokine TNF-α (182.8±23.39 vs 146.7±9.36pg/mL, p=0.007) and of the chemotactic factor MCP-1 (682.9±320.32 vs 265.9±50.14pg/mL, p=0.011) in Tg-hCBS apoE^−/− Cbs^−/− mice after 8-weeks of HF diet, compared with their sibling controls on the same diet. Further, we identified significant positive correlations between the concentrations of TNF-α (r=0.706, p=0.005) and MCP-1 (r=0.745, p=0.008) with the Hcy levels in these mice (Figure 3).

Figure 3. HHcy increased plasma TNF-α and MCP-1 levels in Tg-hCBS apoE^−/− Cbs mice.

EDTA anticoagulant plasma was harvested from mice fed a HF diet at 8 weeks of age for an additional 8 weeks. TNF-α and MCP-1 levels in plasma were assessed by ELISA. Correlation analysis of Hcy levels with plasma TNF-α and MCP-1 levels were shown in scattered dot graphs. Each data point represents one mouse. Values are mean ± SD. P values from independent t test or correlation analysis. n=10–12.

Severe HHcy increased monocyte population in both Tg-S466L Cbs and Tg-hCBS apoE^−/− Cbs mice

We studied the bone marrow (BM) and spleen because BM produces monocytes, and spleen serves as a reservoir for circulating monocytes. As shown in Figure 4A, mononuclear cells (MNC) were gated in gate ii based on granularity and cell size because of their lower granular content and relative larger in size compared with granulocytes and lymphocytes. The gate ii MNC were then further grouped based on the expression levels of CD11b, a monocyte marker. Monocytes were defined as CD11b⁺ MNC and described in the histogram of Figure 4A. The percentage of CD11b⁺ MNC, monocytes, in nucleated cells (gate i) were presented in Figure 4B. Severe HHcy significantly increased monocyte population in blood, spleen and BM in 6 month old Tg-S466L Cbs^−/− mice. Interestingly, HHcy-induced monocyte expansion appeared to be enhanced with aging, as monocyte populations were further increased in all three tissues isolated from 8 months old Tg-S466L Cbs^−/− mice.

Figure 4. HHcy increased monocyte population in peripheral blood, spleen and BM in transgenic Cbs mice independent of hyperlipidemia.

Peripheral blood, spleen and BM cells were isolated from 6 and 8 month old Tg-S466L Cbsmice (A&B) and 15 month old Tg-hCBS apoE^−/− Cbs mice (C&D), stained with antibodies against CD11b and Ly6C, and analyzed by flow cytometry. A & C. Representative dot plots depicting nucleated cells (gate i, eliminated red blood cell content) and MNC cells (gate ii). Histograms describe the percent of CD11b⁺ MNC/monocytes in gate ii. Shaded and open curve area represent data from the non-antibody negative control and CD11b antibody staining groups, respectively, B & D. CD11b⁺ MNC/monocyte percentage in gate i. The percent of CD11b⁺ MNC/monocytes in gate i was calculated by multiplying CD11b⁺ MNC/monocyte event percent in gate ii with the MNC percent in gate i. Values are mean ± SD. n=6. Independent t test was used. BM, bone marrow, SSC, side-scatter light; FSC, forward-scatter light; MNC, mononuclear cells.

Further, we examined monocyte population in 15 month old Tg-hCBS apoE^−/− Cbs mice without dietary manipulation (Figure 4C&D). Severe HHcy significantly promoted monocyte expansion in blood, spleen and BM in Tg-hCBS apoE^−/− Cbs^−/− mice.

Severe HHcy elevated CD11b⁺Ly-6C^hi and CD11b⁺Ly-6C^mid inflammatory monocyte subsets in both Tg-S466L Cbs and Tg-hCBS apoE^−/− Cbs mice

We further characterized monocyte subsets by flow cytometry analysis using anti-CD11b and anti-Ly-6C antibodies. Monocytes expressing high levels of Ly-6C have previously been shown to represent the inflammatory subset.¹⁴ CD11b⁺ MNC within gate ii, defined in Figure 4, were further divided into three groups based on their Ly-6C expression levels (low, middle and high) (Figure 5A). The CD11b⁺Ly-6C^hi inflammatory monocyte subset was the most abundant group in blood and increased with aging in all three tissues. Tg-S466L Cbs^−/− mice at 6 months of age, compared with their controls, had increased percentages of CD11b⁺Ly-6C^hi cells in blood (42.3± 6.9% vs 22.4±10.4%, p=0.019), spleen (5.4±1.5% vs 3.1±0.7%, p=0.01), and BM (12.9±1.6 vs 6.4±0.8%, p=0.001). This was further increased in blood and spleen and stabilized in BM at 8 months of age. The CD11b⁺Ly-6C^mid monocytes were increased in all three tissues in Tg-S466L Cbs^−/− mice compared to control mice at 6 months of age: 24.5±5% vs 17.5±1.5% (p=0.022) in blood, 24.8 ± 0.3% vs 18.5±2.3% (p=0.027) in the spleen, and 5.9±0.8% vs 3.9±0.6% (p=0.047) in BM. This difference disappeared at 8 months of age in all three tissues. The CD11b⁺Ly-6C^low resident monocyte subset in the spleen and in the BM increased between 6 and 8 months of age in both Tg-S466L Cbs^−/− and control animals. In blood and spleen, control animals had a larger CD11b⁺Ly-6C^low subset percentage compared to Tg-S466L Cbs^−/− at 6 months of age, but this difference was not present at 8 months of age.

Figure 5. Characterization of monocyte subsets in transgenic Cbs mice.

Peripheral blood, spleen and BM cells were isolated from 6 and 8 month old Tg-S466L Cbs mice (A&B) and 15 month old Tg-hCBS apoE^−/− Cbs mice (C&D), stained with antibodies against CD11b and Ly6C, and analyzed by flow cytometry. MNC were gated in gate ii as described in Figure 4, and divided into three subsets based on CD11b and Ly-6C expression: CD11b⁺Ly-6C^hi, CD11b⁺Ly-6C^mid and CD11b⁺Ly-6C^low. A & C. Representative dot plots depicting Ly-6C and CD11b expression in gate ii MNC. B & D. Ly-6C subsets in gate ii CD11b⁺ MNC/monocyte. Values are mean ± SD. n=6. Independent t test was used. SSC, side-scattered light; FSC, forward-scatter light; MNC, mononuclear cells.

Moreover, we analyzed monocyte subsets in 15 month old Tg-hCBS apoE^−/− Cbs mice without dietary manipulation (Figure 5C&D). CD11b⁺Ly-6C^hi and CD11b⁺Ly-6C^mid subsets were significantly increased in three compartments in Tg-hCBS apoE^−/− Cbs^−/− mice.

L-Hcy maintained CD11b⁺Ly-6C^hi subset and induced CD11b⁺Ly-6C^mid subset in cultured mouse primary splenocytes

To verify the effect of HHcy on monocyte differentiation we established a primary mouse splenocyte culture model. The spleen was used as an enriched source of mononuclear cells because of the limited number of monocytes in blood. A 24 hr culture of primary mouse splenocytes have been used as an in vitro model of monocyte subset survival.¹³ Using this model, we found that the CD11b⁺Ly-6C^hi inflammatory subset made up 4.5±0.1% of monocytes in freshly isolated splenocytes from 2 months old C57B/L6 mice. The CD11b⁺Ly-6C^hi subset was decreased to 3.5±0.5% after 24 hr and to 0.95±0.27% after 72 hr in cultured control cells. (Figure 6A & B). rIFNγ(100U/mL) significantly induced Ly-6C expression and increased the CD11b⁺Ly-6C^hi inflammatory subset at 24 hr. These are in agreement with other observations on Ly6C monocyte subset survival in cultured spleen¹³ and monocytic cell lines.²¹ L-Hcy (200 and 500μM) maintained the survival of CD11b⁺Ly-6C^hi population after 24 hr in culture to the levels found in freshly isolated splenocytes (4.5%), which were higher than that of the 24 hr non-stimulation control group (3.5%).

Figure 6. Ly-6Cmonocyte survival and differentiation in cultured primary mouse splenocytes.

Splenocytes were isolated from 2 month old C57B/L6 wild-type mice and cultured. Cells were treated with rIFNγ, L-Hcy or L-Cys at indicated doses for 24 hr for survival study, or 48 hr during a 72 hr culture for differentiation study. Cells were stained with antibodies against CD11b and Ly6C, and analyzed by flow cytometry. MNC were defined by gate ii and further divided into three subsets based on Ly-6C expression levels. A. Representative dot plots depicting the distribution of monocyte subsets of CD11b⁺Ly-6C^low, CD11b⁺Ly-6C^mid and CD11b⁺Ly-6C^hi cells. B. CD11b⁺Ly-6Cmonocyte subset distribution. Schematic designs are inserted to describe the experimental strategies. Data are representative of three separate experiments and are shown as mean ± SEM. * compared with the control group at the same time point. ^# compared with the L-Hcy treatment with identical concentrations. MNC, mononuclear cells.

To assess monocyte differentiation using Ly-6C induction as an indicator, we treated splenocytes with L-Hcy or L-Cys 24 hr after culturing for additional 48 hr and then examined Ly-6C expression, as others used in BM cells and monocytic cell lines for monocytes differentiation studies.^22,21 We found that the CD11b⁺Ly-6C^mid subset was significantly reduced with culturing time in the controls. The CD11b⁺Ly-6C^mid subset comprised 15.9% of the MNC when freshly isolated (0 hr), which was reduced to 10.6% and 8.7% following 24 hr and 72 hr of culture, respectively. L-Hcy (100, 200 and 500 μM) significantly increased the CD11b⁺Ly-6C^mid subset after 72 hr of culture in a dose-dependent manner, from 8.7% (control) to 11.2% (p=0.001), 12.2% (p<0.001) and 15% (p<0.001), respectively. In contrast, L-Cys, did not change the proportions of CD11b⁺Ly-6C^hi and CD11b⁺Ly-6C^mid. The significant differences between L-Hcy and L-Cys treatments on CD11b⁺Ly-6C^hi and CD11b⁺Ly-6C^mid subset differentiation confirmed the specific effects of Hcy. Finally, both L-Hcy and L-Cys reduced CD11b⁺Ly-6C^low subset after 72 hr in culture.

Hcy-induced CD11b⁺Ly-6C^mid subset formation was prevented by catalase (CAT) plus SOD, and a NAD(P)H oxidase inhibitor in cultured mouse primary splenocytes

To elucidate the mechanism underlying Hcy-induced inflammatory monocyte differentiation, we preincubated mouse splenocytes with various inhibitors prior to Hcy treatment (Figure 7A). The combination of adenosine and L-Hcy (50μM), a condition we established to sensitize intracellular hypomethylation,⁶ did not change CD11b⁺Ly-6C^mid subset differentiation compared with adenosine control (Figure 7A&B). CAT, a hydrogen peroxide scavenger, together with SOD, a superoxide anion scavenger, reversed Hcy effects and brought the CD11b⁺Ly-6C^mid population to 116% (#p=0.002) of the no L-Hcy control. Similarly, NAD(P)H oxidase inhibitor apocynin completely prevented Hcy-induced CD11b⁺Ly-6C^mid subset differentiation. In contrast, Hcy maintained the capacity to induce CD11b⁺Ly-6C^mid subset differentiation in the presence of the xanthine oxidase inhibitor allopurinol (Figure 7C), peroxylnitrite scavenger uric acid and the NOS inhibitor L-NAME. SOD/CAT and apocynin slightly increased the baseline of CD11b⁺Ly-6C^mid, whereas allopurinol, L-NAME and uric acid did not. Interestingly, uric acid and L-NAME reduced CD11b⁺Ly-6C^mid subset formation in the presence of Hcy, but not as much as SOD/CAT or apocynin.

Figure 7. Anti-oxidants abolished Ly-6C^mid monocyte differentiation induced by Hcy in cultured primary mouse splenocytes.

Splenocytes were isolated from 2 month old C57B/L6 wild-type mice. Cells were cultured for 24 hr and then treated with SOD-PEG (250U/mL) plus CAT-PEG (250U/mL), uric acid (60μg/mL), apocynin (100μM), allopurinol (30μg/mL), or L-NAME (1mM) for 1 hr before exposure to L-Hcy (500μM), or treated with 50 μM L-Hcy in the presence of 25μM adenosine for an additional 48 hr. Cells were then stained with antibodies against CD11b and Ly6C, and analyzed by flow cytometry. MNC were defined by gate ii and further divided into three monocyte subsets based on CD11b and Ly-6C expression. A. Representative dot plots depicting the distribution of monocyte subsets of CD11b⁺Ly-6C^low, CD11b⁺Ly-6C^mid and CD11b⁺Ly-6C^hi cells. B & C. CD11b⁺Ly-6C^mid monocyte subsets. Data are representative of 3 independent experiments. Data are expressed as mean ± SEM. *compared with the parallel no L-Hcy control. ^#compared with500 μM L-Hcy treatment without other inhibitors. PEG, Polyethylene glycol; SOD, superoxide dismutase; CAT, Catalase; MNC, mononuclear cells.

Discussion

In the present study, we generated a novel genetic mouse model with severe HHcy and hypercholesterolemia, Tg-hCBS apoE^−/−Cbs^−/− mice, and examined the effects of HHcy on atherosclerotic lesion formation and monocyte accumulation. The Tg-hCBS apoE^−/−Cbs mice described here is a useful model for HHcy related atherosclerosis research. This model has a survival advantage over our previous apoE^−/−Cbs mice.¹⁶ The Tg-hCBS apoE^−/−Cbs^−/− mice retained severe HHcy and survived normally. The control Tg-hCBSapoE^−/−Cbs mice fed a HF diet have plasma Hcy levels of 13.9 μM, which are similar to what we have observed in the apoE^−/−Cbs^−/+ mice fed the same diet (14.2 μM).¹⁶ Plasma Hcy levels in the Tg hCBS apoE^−/−Cbs^−/− mice on regular diet is 200μM, which is higher than 169 μM in the Tg hCBS Cbs^−/− mice.²³ These data support the notion that hyperlipidemia increases Hcy levels.

The Tg-hCBS apoE^−/− Cbs^−/− mice were slightly smaller than their littermate controls. This was also found in Tg-S466L Cbs^−/− mice (data not shown). These results are in accordance with our previous reports.²⁴ We have observed a trend in decreasing spleen weight in severe HHcy Tg-hCBS apoE^−/− Cbs^−/− mice (Figure 1G), suggesting that HHcy causes growth retardation in general and reduces spleen weight, potentially creating a negative impact on immune system development.

We previously reported that HHcy accelerates atherosclerosis and increases LDL uptake by peritoneal macrophages in apoE^−/−/Cbs^−/+ and apoE^−/−/Cbs^−/− mice with and without dietary perturbation.¹⁶ The current study is in agreement with our previous findings, and provides further evidence demonstrating that severe HHcy accelerates atherosclerosis, and increases Ly6C positive monocyte/macrophage inflammatory subset accumulation in the atherosclerotic lesion (Figure 2). These findings led to the new hypothesis that severe HHcy contributes to atherosclerosis via promoting inflammatory monocytes.

It has been suggested that inflammatory processes play a pivotal role throughout atherosclerotic progression. We found that plasma TNFα and MCP-1 levels were significantly elevated in Tg-hCBS apoE^−/− Cbs^−/− mice, and were positively correlated with plasma Hcy levels (Figure 3). TNFα is a proinflammatory cytokine that activates the endothelium. MCP-1 is a chemokine that mediates monocyte adhesion and infiltration into inflammatory sites. It was suggested that Hcy-induced MCP-1 expression in aortic endothelial cells may also contribute to leukocyte recruitment.²⁵ Our findings support the hypothesis that HHcy promotes systemic inflammation and monocyte infiltration into the vessel wall.

Clinical criteria usually consider >8% circulating MNC, based on their mononuclear morphological feature, as monocytosis in humans. The control Tg-S466LCbs mice animals have 3–5% MNC (gate ii) among the peripheral nucleated cells (gate i) (Figure 4A). Severe HHcy induced monocytosis (8.6% MNC) at 8 months of age. Monocytosis was also observed in the control Tg-hCBS apoE^−/−Cbs mice (10.3% MNC) and exacerbated by HHcy in Tg-hCBS apoE^−/−Cbs^−/− mice (11.6% MNC) (Figure 4C). Using CD11b as a monocyte molecular marker, we found that severe HHcy significantly promoted monocyte expansion in peripheral blood, spleen and bone marrow in both mouse lines independent of hyperlipidemia. (Figure 4B&D).

Recent studies show that circulating monocytes display heterogeneity, which commit to specific functions in atherogenesis. Using Ly-6C as a marker, monocytes can be divided into three subsets: Ly-6C^hi, Ly-6C^mid and Ly-6C^low fractions.^{22, 26} The Ly-6C^hi monocytes are recognized as the inflammatory subset, which can differentiate into macrophages in atheromas of apoE^−/− mice.¹³ The Ly-6C^mid monocytes are considered as the intermediate inflammatory subset based on their ability to uptake inflammatory particles and migrate to sites of inflammation.²⁶ We found that severe HHcy was associated with increases in both Ly-6C^hi and Ly-6C^mid monocyte subsets in peripheral blood, spleen and BM in both mouse models, which were further increased with aging (Figure 5). Therefore, we have demonstrated, for the first time, that HHcy selectively promotes the expansion of inflammatory monocyte subsets in the periphery and BM in mice independent of hyperlipidemia.

Our data indicate that mouse primary splenocytes are a relevant in vitro model of monocyte survival and differentiation.¹³ The mouse spleen is an hematopoietic organ and was used to study monocyte/macrophage heterogeneity and in vitro monocyte differentiation.^{13, 14} Here, we have demonstrated an Hcy specific effect on promoting Ly-6C monocyte differentiation that is not mimicked by Cys. The distinct effects of Hcy and Cys on monocyte differentiation may be related to the different biochemical activities of these amino acids. Our study is the first to demonstrate that HHcy induces monocyte expansion and selectively promotes differentiation of Ly-6C^hi and Ly-6C^mid inflammatory monocytes in the diseased models of severe HHcy. The modulation of HHcy on monocyte heterogeneity may enhance inflammatory responses during atherogenesis and contribute to the increased risk of CVD in HHcy.

Finally, we demonstrated that Hcy-induced inflammatory monocyte differentiaiton was prevented by SOD/CAT and NAD(P)H oxidase inhibitor apocynin and partially reduced by uric acid and L-NAME. These data suggest that oxidant stress, especially superoxide anion, mostly due to NAD(P)H oxidase activation, is the major mechanism mediating Hcy-induced Ly-6C^mid inflammatory monocyte differentiation. Because Hcy did not induce the Ly-6C^mid subset in the presence of adenosine, a condition we established to sensitize intracellular hypomethylation,⁶ we ruled out a hypomethylation mechanism. It have been reported that Hcy induced MCP-1 secretion in human monocytes via oxidant stress.²⁷ Moreover, Hcy-stimulated superoxide anion production is regulated by protein kinase C activation dependent of NAD(P)H oxidase in the human monocytic cell line THP-1.²⁸ Future studies examining NAD(P)H oxidase regulation should provide important insight into the understanding of Hcy-induced monocyte differentiation and atherosclerosis.

In summary, our studies provide the first evidence demonstrating that severe HHcy induces systematic inflammation and accelerates atherosclerosis in a novel model of severe HHcy and hypercholestorolemia. We found that HHcy induces inflammatory monocyte subset differentiation in mice independent of hyperlipidemia, and in cultured splenocytes mostly via NAD(P)H oxidase-mediated oxidant stress. Our studies indicate that HHcy-induced inflammatory monocyte subset differentiation may be responsible for the increased risk of CVD in HHcy.