Temporal Relationships Between Circulating Levels of CC and CXC Chemokines and Developing Atherosclerosis in Apolipoprotein E*3 Leiden Mice
Objectives— CC and CXC chemokines are implicated in leukocyte recruitment during development of atherosclerotic lesions, suggesting circulating levels of chemokines may be useful serum markers of atherogenesis. Serum chemokine concentrations were measured in apolipoprotein (apo) E*3 Leiden mice and their nontransgenic littermates and related to the differing rates of atherogenesis in these animals.
Methods and Results— Mice were fed a high-fat, high-cholesterol/cholate (HFC/C) diet for 18 weeks. Circulating levels of JE/monocyte chemotactic protein-1 increased (P<0.05) after 2 to 4 weeks, coincident with development of diet-induced hypercholesterolemia, and remained elevated throughout the study. Circulating KC concentrations increased (P<0.05) after consumption of HFC/C diet; however, unlike JE, serum KC concentrations increased more rapidly in apoE*3 Leiden mice than their nontransgenic littermates. Hepatic expression of JE and KC mRNA were detected by in situ hybridization in all mice fed HFC/C diet. Aortic expression of JE mRNA was seen only in apoE*3 Leiden mice within macrophage-rich atherosclerotic lesions. By contrast, no aortic expression of KC mRNA was detected by in situ hybridization.
Conclusions— Increases in serum chemokine concentrations did not reflect temporal aortic production of these molecules and proved less predictive than serum cholesterol of the markedly different extent of atheroma in apoE*3 Leiden and nontransgenic mice.
Atherosclerosis is typified by a chronic inflammatory response involving the recruitment and activation of mononuclear leukocytes.1 In part, this process is controlled by the production of chemoattractant cytokines (chemokines) by cells of the artery wall. Chemokines are members of a superfamily of small, secreted proteins (8 to 16 kDa) that mediate migration and activation of leukocytes and arterial cells by interacting with G-protein-coupled cell-surface receptors.2 Chemokines are divided into families according to the arrangement of the first 2 of 4 conserved cysteine residues, the 2 largest subfamilies being designated CC and CXC.2
Members of the CC (or β) chemokine subfamily predominantly chemoattract monocytes and T-lymphocytes but not neutrophils, and extensive experimental evidence supports a role for the prototypical CC chemokine monocyte chemotactic protein-1 (MCP-1) in atherogenesis. Expression of MCP-1 is upregulated in human atherosclerotic plaque3,4⇓ and found in arteries of primates5 and transgenic mice consuming hypercholesterolemic diets.6–10⇓⇓⇓⇓ Overexpression of the murine homologue of MCP-1, JE, transgene by leukocytes accelerates lesion formation in apolipoprotein (apo) E−/− mice,7 whereas absence of JE reduces atherosclerosis in LDL receptor−/− mice consuming a high-fat and high-cholesterol diet.8 Similarly, absence of CCR-2, the major monocyte receptor for JE, reduces monocyte migration9 and inhibits lesion formation in apoE−/− mice.10,11⇓ However, although the absence of MCP-1 and CCR-2 can delay and diminish atherosclerosis, it does not completely arrest the progression of the disease.8–11⇓⇓⇓
Chemokines of the CXC (or α) family, particularly those that contain the Glu-Leu-Arg motif, principally induce the migration of neutrophils and not monocytes.12 Despite this apparent selectivity and the scarcity of neutrophils within atherosclerotic lesions,13 recent data implicate members of the CXC chemokine subfamily in atherogenesis. These include interleukin-8 (IL-8) and growth-related oncogene (GRO) chemokines that bind to CXCR-1/-2 and CXCR-2 receptors, respectively.12,14⇓ The murine ligands KC (GRO-α) and macrophage inflammatory protein-2 (MIP-2; GRO-β/γ) bind to the murine CXCR-2 receptor homolog mIL-8RH.15 Increased expression of IL-8 has been shown in macrophages isolated from human atheromatous lesions,16,17⇓ and immunoreactive IL-8, GRO-α, and murine KC have been reported within macrophage-rich areas of human and murine atherosclerotic lesions.17,18⇓ Macrophages in human and murine atherosclerotic lesions express CXCR-2 and mIL-8RH, respectively.18 Bone marrow transplantation studies demonstrated that the absence of leukocyte mIL-8RH resulted in fewer monocytes and reduced lesion size in the LDL receptor−/− model of atherosclerosis.18 Both IL-8 and GRO-α are capable of inducing monocyte adhesion in vitro, particularly under flow conditions19; however, the effects of receiving mIL-8RH−/− bone marrow were more pronounced at later stages of lesion development in vivo, indicating a role in retention and expansion of intimal macrophages.18,20⇓
Circulating levels of chemokines are increased in several acute and chronic inflammatory and immune-related disorders,21,22⇓ suggesting that their levels may reflect enhanced tissue expression and secretion of these molecules during disease. Little is known about circulating chemokine concentrations during the development of atherosclerotic lesions, but elevated plasma CC23,26⇓ or CXC chemokines24–26⇓⇓ have been reported in patients with congestive heart failure,23,24⇓ ischemia-reperfusion injury,25 or aortic aneurysms.26 Importantly, serum chemokine concentrations were associated with degree of disease severity.23,24⇓
This study examines the hypothesis that circulating levels of CC or CXC chemokines may prove useful serum markers of atheroma in transgenic mice expressing the human apoE*3 Leiden gene.27,28⇓ These mice have impaired hepatic clearance of chylomicron and VLDL remnant lipoproteins and, when fed an atherogenic diet, develop marked hypercholesterolemia and accelerated atherosclerosis compared with their nontransgenic littermates. We measure circulating levels of CC (JE) and CXC (KC and MIP-2) chemokines in apoE*3 Leiden mice and their nontransgenic littermates consuming an atherogenic diet. We define the temporal relationships between increases in serum chemokines (JE and KC) and changes in serum lipids and compare hepatic chemokine expression and lipid accumulation with chemokine expression, macrophage content, and lesion development in the aortic root.
Please see the online data supplement, available at http://atvb.ahajournals.org.
The effects of feeding high-fat/high-cholesterol/cholate (HFC/C) diet on serum CC (JE) and CXC (KC and MIP-2) chemokine concentrations (Figure 1) were analyzed using terminal samples from mice culled at the time points indicated.
At the start of the study, serum JE concentrations were higher in apoE*3 Leiden mice than in nontransgenic mice (P<0.05, Figure 1A). Consumption of HFC/C diet caused rapid, and approximately equivalent, increases from baseline in serum JE in apoE*3 Leiden and nontransgenic mice. Serum JE concentrations were 3- to 5-fold higher at 4, 8, and 18 weeks in apoE*3 Leiden mice and 3- to 4-fold higher at 4, 12, and 18 weeks in nontransgenic mice consuming HFC/C diet compared with transgenic animals fed normal diet (P<0.05).
Serum KC levels were also higher at the start of the study in apoE*3 Leiden mice than in nontransgenic animals (P<0.05, Figure 1B). Consumption of HFC/C diet increased serum KC significantly in apoE*3 Leiden and nontransgenic mice, but this increase was much more rapid in the transgenic animals (Figure 1B). At 2 and 4 weeks, serum KC was 3.4- and 2.9-fold higher, respectively, in apoE*3 Leiden than in nontransgenic mice consuming the atherogenic diet. However, by week 12, serum KC levels were approximately equivalent in these 2 groups of animals and remained so at week 18. Serum KC levels were also significantly higher at weeks 4, 8, and 18 in apoE*3 Leiden mice and at weeks 4 and 18 in nontransgenic mice fed HFC/C diet than in apoE*3 Leiden mice consuming normal diet.
Concentrations of the other mIL-8RH ligand, MIP-2, exhibited a more complex, almost biphasic profile and did not remain elevated throughout the study (Figure 1C). At the start of the study, serum MIP-2 levels were higher (P<0.05) in nontransgenic than in apoE*3 Leiden mice. Increases from baseline (P<0.05) in serum MIP-2 concentrations were observed at weeks 2, 4, 8, and 12 in apoE*3 Leiden mice fed HFC/C diet, but by week 18, MIP-2 levels fell below baseline values. At weeks 2, 4, and 18, levels of MIP-2 in apoE*3 Leiden and nontransgenic mice consuming HFC/C diet were significantly higher than those in the group of apoE*3 Leiden mice consuming normal diet. However, at no time were serum levels of MIP-2 higher in apoE*3 Leiden than in nontransgenic mice consuming the same diet.
Lipids and Lipoproteins
The effects of feeding HFC/C diet on serum cholesterol and triglyceride concentrations in apoE*3 Leiden and nontransgenic mice are shown in Figure I (see http://atvb.ahajournals. org). As expected,28 consumption of HFC/C diet for up to 18 weeks resulted in serum cholesterol levels that were higher (4- to 8-fold) in apoE*3 Leiden mice compared with nontransgenic mice (P<0.00001) and 10- to 18-fold higher than those in transgenic animals fed normal diet.
Serum triglyceride concentrations were higher throughout the study in both groups of apoE*3 Leiden mice than in nontransgenic mice. At all time points, serum triglyceride levels in nontransgenic animals were between 20% and 38% of the levels seen in apoE*3 Leiden mice consuming the same HFC/C diet. Serum triglyceride levels in the apoE*3 Leiden mice fed HFC/C diet were significantly lower (56%) at week 4 of the study than in apoE*3 Leiden mice fed normal diet.
Cholesterol lipoprotein profiles were analyzed at weeks 2, 4, 8, 12, and 18 using pooled serum samples from animals in the same experimental group. After 4 weeks of HFC/C diet, the lipoprotein profiles of the animals did not change substantially, and representative data from this time point are shown in Figure IC. As expected,28 HFC/C diet increased VLDL/LDL cholesterol in nontransgenic animals, with much larger increases in this fraction in the apoE*3 Leiden mice.
Livers of mice from both groups consuming HFC/C diet for 18 weeks contained more cholesteryl ester but 3- to 4-fold lower triglyceride than those from apoE*3 Leiden mice consuming normal diet (Table). No significant differences in hepatic phospholipid or free cholesterol content were evident between the groups.
Hepatic Chemokine Expression
The earliest time points at which JE or KC mRNA could be detected by in situ hybridization were investigated using histological sections of liver and aorta from at least 4 animals per group. Representative sections are shown. Expression of KC or JE mRNA were not detected in hepatic sections from apoE*3 Leiden mice fed normal diet at any time (Figure IIA, see http://atvb.ahajournals.org).
Hepatic expression of both KC and JE mRNA were detected as early as week 2 in both apoE*3 Leiden and nontransgenic mice consuming HFC/C diet (Figures IIB, IID, IIF, and IIH); more extensive expression of both chemokines were evident after 18 weeks on the same diet (Figures IIC, IIE, IIG, and III).
Aortic Chemokine Expression, Macrophage Staining, and Lesion Development
JE and KC mRNA expression patterns in the aortic root (Figure 2) were compared with serial sections stained for the presence of macrophages using the macrophage-specific antibody MOMA-2 (Figure 3). Lesion cross-sectional areas were 26- to 43-fold larger throughout the study in apoE*3 Leiden mice compared with nontransgenic mice fed HFC/C diet (Figure III, see http://atvb.ahajournals.org).
After 4 weeks on atherogenic diet, early intimal lesions and MOMA-2-positive staining were evident in aortae from all (4 of 4) of the apoE*3 Leiden mice probed (Figures 3A and III). However, no expression of KC mRNA was found, and expression of JE mRNA was detected in aortic sections from only 1 of 4 apoE*3 Leiden mice (data not shown). Minimal evidence of macrophage staining, JE mRNA expression (data not shown), or lesion development could be detected in nontransgenic controls (Figures 3B and III).
After 8 weeks, larger intimal lesions had developed in apoE*3 Leiden mice, particularly in the aortic root (Figure III), and this coincided with the sites of JE mRNA expression (3 of 4) (Figure 2B); however, KC mRNA remained undetected (Figure 2C).
After 18 weeks on HFC/C diet, more advanced lesions were seen in apoE*3 Leiden mice, accompanied by extensive positive staining for macrophages within the core of the lesion in the aortic root (Figure 3C). Again, more extensive expression of JE (Figure 2D), but not KC mRNA, accompanied lesion development. No JE mRNA was detected when sections were incubated with sense probe (Figure 2E). Early fatty streak lesions were seen in nontransgenic mice at week 18 (Figure III), but no JE (Figure 3F) or KC mRNA expression was detected in these sections.
Correlations Between Serum Cholesterol and Chemokine Concentrations and Lesion Area
The value of serum JE and KC measurements in predicting the extent of murine atheroma were compared with serum cholesterol concentrations. Previously,28 it was established that total exposure to serum cholesterol, calculated as area under the curve (AUC) (see Methods), accurately predicted the extent of atheroma in apoE*3 Leiden mice. Here, serum cholesterol exposure again correlated closely with lesion area (R=0.93) in apoE*3 Leiden mice but not their nontransgenic littermates. Lower positive correlations between lesion area and calculated AUC for serum JE (R=0.67), KC (R=0.62), and MIP-2 (R=0.62) were seen in apoE*3 Leiden mice, but no such correlations were seen in nontransgenic animals consuming the same atherogenic diet. Significantly, highly positive correlations between serum cholesterol AUC and serum JE (R=0.86 and 0.86), KC (R=0.89 and 0.78), and MIP-2 (R=0.84 and 0.82) were seen in apoE*3 Leiden and nontransgenic mice, respectively, consuming HFC/C diet.
Atherosclerosis is a major cause of morbidity and mortality in Western societies. A serum marker reflecting clinically silent arterial pathogenesis would be a useful tool in assessing risk of developing coronary artery disease. We investigated the potential utility of systemic CC and CXC chemokines in this regard, because prototypical members of each of these chemokine subfamilies have been implicated in the recruitment or retention of monocyte-macrophages during atherogenesis. Consumption of an atherogenic diet induced sustained increases in serum CC and CXC chemokines, JE, and KC in both apoE*3 Leiden and nontransgenic mice. Serum levels of JE increased and remained approximately equivalent in both groups of animals, but serum KC increased much more rapidly in apoE*3 Leiden mice compared with their nontransgenic littermates (Figure 1B). We therefore investigated the temporal relationships between circulating chemokine concentrations, hepatic and aortic chemokine expression, lipid accumulation in the liver and serum, and development of atherosclerotic lesions in these animals.
ApoE*3 Leiden mice consuming an atherogenic diet develop atherosclerosis much more rapidly than nontransgenic mice fed the same diet (Figure III).27,28⇓ In apoE*3 Leiden mice, lesion development seems directly related to the duration of hypercholesterolemia induced in these animals (above and Reference 28). Maximal increases in serum cholesterol concentrations were detected in both strains after 4 weeks on atherogenic diet (Figure IA), although serum cholesterol levels remained much lower in nontransgenic than in apoE*3 Leiden mice. However, despite the very differing rates of atherogenesis and degree of hypercholesterolemia induced by HFC/C diet, apoE*3 Leiden mice and their nontransgenic littermates had similar levels of hepatic lipids, particularly cholesteryl ester, after 18 weeks (Table).
Temporal increases in serum JE correlated with onset of hypercholesterolemia in both groups of animals consuming the atherogenic diet; however, the eventual extent of this rise was similar in both groups, suggesting that the absolute concentration of serum lipids was not a key determinant of serum JE levels. This contrasts with the reported effect of hypercholesterolemia on the expression of the JE/MCP-1 receptor, CCR-2, by circulating monocytes.30 Rather, serum levels of JE may be more closely associated with hepatic inflammation, possibly because of lipid accumulation. Indeed, hepatic expression of JE and KC mRNA were detected in both groups of mice after 2 weeks of consumption of the atherogenic HFC/C diet (Figure II), suggesting that the liver may be a primary source of serum chemokines in this study. Consumption of an atherogenic diet by C57BL mice has previously been shown to be associated with enhanced hepatic oxidative stress, activation of nuclear factor-κB, and increased expression of JE, KC, serum amyloid A, and heme-oxygenase mRNA.31,32⇓ Indeed, circulating levels of serum amyloid A were increased and approximately equivalent in the 2 groups of animals consuming HFC/C diet in this study (data not shown).
The contribution of JE from developing aortic lesions to serum JE concentrations does not seem substantial, because serum JE levels in apoE*3 Leiden mice did not exceed those in the nontransgenic controls. Aortic expression of JE mRNA was also not seen until much later (8 weeks) in apoE*3 Leiden mice consuming HFC/C diet (Figure 2). Interestingly, positive macrophage staining could be detected within early lesions in both apoE*3 Leiden (4 weeks) and nontransgenic (4 weeks) mice (Figure 3) before detection of JE mRNA by in situ hybridization. Thus, apoE*3 Leiden mice seem to differ from apoE and LDL receptor-deficient models of atherogenesis, in which JE/MCP-1 expression seems to coincide directly with monocyte infiltration.29,33,34⇓⇓ This could be explained by the relative sensitivities of the techniques or probes used; alternatively, other chemokines may be involved in the very early stages of monocyte recruitment to the vessel wall in this murine model.
Systemic levels of ligands for mIL-8RH, KC, and MIP-2 exhibited markedly differing profiles. The almost biphasic profile of serum MIP-2, seen in all the groups of animals, is intriguing, and the reasons underlying this finding are unclear. Because serum MIP-2 levels could not provide a reliable estimate of inflammatory status, they were not investigated further. By contrast, measurement of circulating levels of KC revealed differing inflammatory responses in the 3 groups of animals in this study. Maximal levels of serum KC were achieved by 4 weeks in apoE*3 Leiden mice consuming HFC/C diet and by 12 weeks in the nontransgenic controls and did not increase significantly in the apoE*3 Leiden mice fed normal diet. Previous reports of aortic CXC chemokine expression12–20⇓⇓⇓⇓⇓⇓⇓⇓ suggested intralesional expression of KC could be a factor contributing to the enhanced serum KC levels seen in apoE*3 Leiden mice. However, the apparent absence of KC mRNA within developing lesions in this study does not support this explanation. It seems more likely that the observed increases in serum KC are attributable to enhanced production of KC from the liver (above and References 31 and 32) and other tissues. For example, plasma IL-8 concentrations correlate directly with IL-8 expression by blood mononuclear cells,35–37⇓⇓ and patients with hypercholesterolaemia35 or congestive heart failure24 also exhibit enhanced IL-8 expression.
In conclusion, circulating levels of chemokines JE and KC are elevated and sustained during the chronic inflammatory response elicited by feeding an atherogenic diet to apoE*3 Leiden and nontransgenic mice. We propose that serum concentrations of these molecules may reflect their output from 1 or more sites of inflammation, assuming that clearance of these molecules from the circulation remains constant. Consumption of an atherogenic diet seems to trigger hepatic inflammation and chemokine expression, indicating that the liver may be an important source of the observed increases in systemic CC and CXC chemokine levels. Temporal increases in systemic JE and KC did not seem to reflect lesional expression of these molecules or, indeed, differentiate the markedly different extent of atheroma in apoE*3 Leiden mice compared with the nontransgenic controls.
This study was supported by a Biotechnology and Biological Sciences Research Council/SmithKline Beecham Industrial CASE studentship to N.M.
- Received May 28, 2003.
- Accepted June 2, 2003.
- ↵Yla-Herttuala S, Lipton BA, Rosenfeld ME, Goldberg IJ, Steinberg D, Witztum JL. Expression of monocyte chemoattractant protein-1 in macrophage-rich areas of human and rabbit atherosclerotic lesions. Proc Natl Acad Sci U S A. 1991; 88: 5252–5256.
- ↵Neiken NA, Coughlin SR, Gordon D, Wilcox JN. Monocyte chemoattractant protein-1 in human atheromatous plaques. J Clin Invest. 1991; 88: 1121–1127.
- ↵Yu X, Dluz S, Graves DT, Zhang L, Antoniades HN, Hollander W, Prusty S, Valente AJ, Schwartz CJ, Sonsenshine GE. Elevated expression of monocyte chemoattractant protein-1 by vascular smooth muscle cells in hypercholesterolaemic primates. Proc Natl Acad Sci U S A. 1992; 89: 6953–6957.
- ↵Aiello RJ, Bourassa PAK, Lindsey S, Weng WF, Natoli E, Rollins BJ, Milos PM. Monocyte chemoattractant protein-1 accelerates atherosclerosis in apolipoprotein E-deficient mice. Arterioscler Thromb Vasc Biol. 1999; 19: 1518–1525.
- ↵Jonasson L, Holne T, Skalli O, Bondjers G, Hansson GK. Regional accumulations of T-cells, macrophages and smooth muscle cells in the human atherosclerotic plaque. Arteriosclerosis. 1986; 6: 131–138.
- ↵Bozic CR, Gerard NP, von Uexkull-Guldenband C, Kolakowski LF, Conklyn MJ, Breslow R, Showell HJ, Gerard C. The murine interleukin-8 type B receptor homologue and its ligands. J Biol Chem. 1994; 269: 29355–29358.
- ↵Wang N, Tabas I, Winchester R, Ravalli S, Rabbani LE, Tall A. Interleukin-8 is induced by cholesterol loading of macrophages and expressed by macrophage foam cells in human atheroma. J Biol Chem. 1996; 271: 8837–8842.
- ↵Apostolopoulos J, Davenport P, Tipping PG. Interleukin-8 production by macrophages from atheromatous plaques. Arterioscler Thromb Vasc Biol. 1996; 16: 1007–1012.
- ↵Bossink AWJ, Paeman L, Jansen PM, Thijs LG, van Damma J. Plasma levels of the chemokines monocyte chemotactic proteins-1 and -2 are elevated in human sepsis. Blood. 1995; 86: 3841–3847.
- ↵Koch AE, Kunkel SL, Harlow LA, Mazarakis DD, Haines GK, Burdick MD, Pope RM, Strieter RM. Macrophage inflammatory protein-1α: a novel chemokine cytokine for macrophages in rheumatoid arthritis. J Clin Invest. 1994; 93: 921–928.
- ↵Aukrust P, Ueland T, Muller F, Andreassen AK, Nordoy I, Aas H, Kjekskhus J, Simonsen S, Froland SS, Gullestad L. Elevated circulating levels of CC chemokines in patients with congestive heart failure. Circulation. 1998; 97: 1136–1143.
- ↵Damas JK, Gullestad L, Ueland T, Solum NO, Simonsen S, Froland SS, Aukrust P. CXC-chemokines, a new group of cytokines in congestive heart failure: possible role of platelets and monocytes. Cardiovasc Res. 2000; 45: 421–436.
- ↵Van den Maagdenberg AMJM, Hofgker MH, Kripenfort PJA, de Brujin I, van Vlijmen B, van der Boom H, Havekes LN, Frants RR. Transgenic mice carrying the apolipoprotein E*3 Leiden gene exhibit hyperlipoproteinaemia. J Biol Chem. 1993; 268: 10540–10545.
- ↵Groot PHE, van Vlijmen BJM, Benson GM, Hofker MH, Schiffelers R, Vidgeon-Hart M, Havekes LM. Quantitative assessment of aortic atherosclerosis in APOE*3 Leiden transgenic mice and its relationship to serum cholesterol exposure. Arterioscler Thromb Vasc Biol. 1996; 16: 923–933.
- ↵Han KH, Tangirala RK, Green SR, Quehenberger O. Chemokine receptor CCR2 expression and monocyte chemoattractant protein-1-mediated chemotaxis in human monocytes. Arterioscler Thromb Vasc Biol. 1991; 18: 1983–1991.
- ↵Liao F, Andalibi A, deBeer FC, Fogelman AM, Lusis AJ. Genetic control of inflammatory gene induction and NF-κB-like transcription factor activation in response to an atherogenic diet in mice. J Clin Invest. 1993; 91: 2572–2579.
- ↵Liao F, Andalibi A, Qiao, J-H, Allayee H, Fogelman AM, Lusis AJ. Genetic evidence for a common pathway mediating oxidative stress, inflammatory gene induction, and aortic fatty streak formation in mice. J Clin Invest. 1994; 94: 877–884.
- ↵Reckless J, Rubin EM, Verstuyft JB, Metcalfe JC, Grainger DJ. Monocyte chemoattractant protein-1, but not tumour necrosis factor-α, is correlated with monocyte infiltration in mouse lipid lesions. Circulation. 1999; 99: 2310–2316.
- ↵Kostulas N, Pelidou SH, Kvisakk P, Kostulas V, Link H. Increased IL-1β, IL-8 and IL-17 mRNA expression in blood mononuclear cells observed in a prospective ischaemic stroke study. Stroke. 1999; 30: 2174–2179.
- ↵Kostulas N, Kvisakk P, Huang Y, Matsuevicius D, Kostulas V, Link H. Ischaemic stroke is associated with a systemic increase of blood mononuclear cells expressing interleukin-8 mRNA. Stroke. 1998; 29: 462–466.