Previous studies have demonstrated that macrophage-derived apolipoprotein E (apoE) reduces atherosclerotic lesion formation in lean apoE-deficient (−/−) mice. apoE has also been demonstrated to play a role in adipocyte differentiation and lipid accumulation. Because the prevalence of obesity has grown to epidemic proportions, we sought to determine whether macrophage-derived apoE could impact atherosclerotic lesion formation or adipose tissue expansion and inflammation in obese apoE−/− mice. To this end, we transplanted obese leptin-deficient (ob/ob) apoE−/− mice with bone marrow from either ob/ob;apoE−/− or ob/ob;apoE+/+ donors. There were no differences in body weight, total body adipose tissue, or visceral fat pad mass between recipient groups. The presence of macrophage-apoE had no impact on adipose tissue macrophage content or inflammatory cytokine expression. Recipients of apoE+/+ marrow demonstrated 3.7-fold lower plasma cholesterol (P < 0.001) and 1.7-fold lower plasma triglyceride levels (P < 0.01) by 12 wk after transplantation even though apoE was present in plasma at concentrations <10% of wild-type levels. The reduced plasma lipids reflected a dramatic decrease in very low density lipoprotein and a mild increase in high-density lipoprotein levels. Atherosclerotic lesion area was >10-fold lower in recipients of ob/ob;apoE+/+ marrow (P < 0.005). Similar results were seen in leptin receptor-deficient (db/db) apoE−/− mice. Finally, when bone marrow transplantation was performed in 4-mo-old ob/ob;apoE−/− and db/db;apoE−/− mice with preexisting lesions, recipients of apoE+/+ marrow had a 2.8-fold lower lesion area than controls (P = 0.0002). These results demonstrate that macrophage-derived apoE does not impact adipose tissue expansion or inflammatory status; however, even very low levels of macrophage-derived apoE are capable of reducing plasma lipids and atherosclerotic lesion area in obese mice.
- very low-density lipoproteins
- atherosclerotic lesions
obesity is a growing worldwide epidemic and is known to increase the risk of atherosclerotic cardiovascular disease (1, 12, 25). Thus it is imperative that studies on well-known anti-atherogenic targets be conducted in the context of obesity. There are many different obese mouse models available to study metabolic disease; however, most of these models are relatively resistant to high-fat diet-induced atherosclerosis. For example, leptin-deficient (ob/ob) and leptin receptor-deficient (db/db) mice have elevated high-density lipoprotein (HDL) levels (6, 26, 30, 31) and are therefore resistant to atherosclerotic lesion formation (27). To develop obese models that are susceptible to lesion formation, we have crossed the ob/ob and db/db mice onto both low-density lipoprotein receptor (LDLR) deficient and apoE−/− backgrounds (6, 11, 14). These obese mice develop hyperlipidemia and atherosclerotic lesions to a greater extent than their lean counterparts (11) and are useful for studying atherosclerosis in the context of obesity.
Macrophages are one of the key cell types involved in atherosclerotic lesion formation (9, 20, 23). They can promote lesion formation by accumulating in the arterial intima, engulfing oxidized lipoproteins within the lesion, and by secreting inflammatory cytokines and matrix metalloproteinases. Alternatively, certain proteins expressed by macrophages can actually protect against atherosclerosis by increasing cholesterol efflux from arterial macrophages. These proteins include membrane-bound lipid transporters such as ABC-AI, and apoproteins (apo) such as apoE (3, 7, 15, 21, 37) and apoAI (18, 24).
Although the liver is the major site of apoE synthesis, and is responsible for 90% of the apoE in circulation (22), other cell types such as macrophages (2) and adipocytes (39) also secrete significant amounts of apoE. Adipocyte apoE expression has been shown to be controlled by peroxisome proliferator-activated receptor γ, tumor necrosis factor (TNF)-α (38), and liver X receptor (19, 29, 33). The endogenous expression of apoE by adipocytes has been shown to be important for lipid accumulation and adipocyte differentiation (17). Because of this, lean apoE−/− mice have smaller adipocytes and lower levels of total triglyceride (TG) in their adipose tissue compared with apoE+/+ mice (17). Macrophages are present in adipose tissue of obese mice (35, 36); however, it is not known whether apoE secretion from these adipose tissue resident macrophages plays a role in adipose tissue physiology.
Using bone marrow transplantation (BMT), it has been shown that apoE−/− mice receiving marrow from apoE+/+ donors accumulate apoE in plasma at ∼10% of wild-type levels. This low level of apoE is sufficient to bring plasma cholesterol concentrations to near-normal levels and to dramatically reduce atherosclerotic lesion formation in lean mice (4, 21, 34). Even plasma levels of macrophage-derived apoE that are too low to influence plasma lipid levels are sufficient to reduce lesion formation (15, 37), indicating an atheroprotective role for macrophage secretion of apoE within the local milieu of the atherosclerotic lesion.
In the current study, we sought to determine whether macrophage-derived apoE influences adipose tissue physiology and whether low levels of macrophage-derived apoE could reduce lesion formation in obese apoE−/− mice.
MATERIALS AND METHODS
Leptin-deficient heterozygous (ob/+), leptin receptor deficient-heterozygous (db/+), and apoE−/− mice were originally purchased from Jackson Laboratories (Bar Harbor, ME) and were propagated within our mouse colony. Obese ob/ob;apoE−/− and db/db;apoE−/− mice were generated using standard breeding strategies. Mice were maintained on chow diet throughout the study and were kept in climate-controlled, 12:12-h light-dark cycles. Mice were fasted for 5 h and bled via the retroorbital venous plexus using heparinized capillary tubes for all blood collections. Mice were killed with isoflurane overdose followed by cervical dislocation. Mice were perfused, and visceral fat pads (the major abdominal fat taken from within the peritoneal cavity) were weighed at death. All animal care and experimental procedures were performed with prior approval from the Vanderbilt Institutional Animal Care and Usage Committee of Vanderbilt University Medical Center.
Recipient mice were lethally irradiated with 900 rads from a cesium gamma source. Postirradiation (4 h), mice were provided with 5 × 106 bone marrow cells via the retroorbital venous plexus. Bone marrow was isolated by flushing the long bones of donor mice with RPMI containing 10 U/ml heparin and 2% FBS. Recipient mice were provided with antibiotic water (5 mg neomycin and 25,000 units polymyxin B sulfate/l) for 1 wk before and 2 wk following transplantation. All transplants into ob/ob;apoE−/− mice were performed using donor marrow from ob/ob mice with or without apoE. Likewise, all transplants into db/db;apoE−/− mice were performed using donor marrow from db/db mice with or without apoE. Thus hematopoietic cell expression of leptin or leptin receptor was eliminated as additional variables in the BMT studies.
Total body fat was measured using a Bruker Minispec machine (Bruker Optics, Billerica, MA) available in the Vanderbilt Mouse Metabolic Phenotyping Center.
Plasma metabolic parameters.
Whole blood was centrifuged to isolate plasma. Plasma was immediately separated and aliquots were stored at −20°C. Plasma total cholesterol (TC) and TG were measured using kits from Raichem (San Diego, CA) adapted to microtiter plates. Nonesterified fatty acids (NEFA) were quantified using the NEFA-C kit from Wako (Richmond, VA). Blood glucose levels were quantified using a OneTouch glucometer from Johnson and Johnson (Northridge, CA), and plasma insulin levels were measured using an ELISA kit from Linco Research (St. Charles, MO).
Plasma lipoprotein analysis.
Fast-performance liquid chromatography (FPLC) was used to analyze lipoprotein profiles. Briefly, plasma samples taken at 12 wk post-BMT were pooled (3–4 mice/group). A 100-μl aliquot was separated over a Superose 6 column (Amersham Biosciences, Piscataway, NJ). Forty 500-μl fractions were collected, and the cholesterol content of fractions 11–40 was measured. Very low density lipoproteins (VLDL) elute in fractions 15–19, low-density lipoprotein (LDL) in fractions 20–25, and HDL in fractions 26–31. Untransplanted ob/ob and db/db mice have a unique lipoprotein pool termed “LDL/HDL1” that elutes in fractions 22–27 (10, 30, 31).
Western blot analysis.
Plasma samples were electrophoresed through 10% SDS gels (Invitrogen), transferred to nitrocellulose membranes, and probed with antibody to mouse apoE (1:25,000 dilution). Secondary horseradish peroxidase-labeled anti-rabbit was used in conjunction with the enhanced chemiluminescence kit from Amersham to detect immunoreactive signals. The primary anti-mouse apoE antibody was produced in the Protein and Immunology Core of the Clinical Research Unit at Vanderbilt University Medical Center. Plasma from C57BL/6 or ob/ob controls was diluted 1:10, whereas plasma from transplanted mice was undiluted. For Western blotting of FPLC samples, fractions 15–17, 18–20, 21–23, 24–26, 27–29, 30–32, 33–35, and 36–38 were pooled and processed as described for plasma samples.
RNA was isolated from visceral fat pads using the RNeasy mini kit from Qiagen (Valencia, CA) according to the manufacturer's instructions. cDNA was synthesized using the iScript cDNA synthesis kit from Bio-Rad (Hercules, CA). Real-time PCR reactions were performed using the iQSupermix from Bio-Rad on an iQ5 Thermocycler. Standards were created using a dilution series of cDNA (equal mix from all samples) concentrations ranging from 1:2 to 1:1,000. Samples were used at a 1:10 dilution. Relative expression level was compared using the delta delta Ct method normalized to 18S expression levels. Expression of macrophage markers F4/80 and CD68, chemokines macrophage chemoattractant protein 1 (MCP1) and macrophage inflammatory protein 1α (MIP-1α), as well as inflammatory proteins TNF-α and interleukin-6 (IL-6) were quantified.
At death, hearts were collected and frozen in optimal cutting temperature medium. Sections (10 μm) were collected beginning at the aortic root and extending for 300 μm according to the method of Paigen et al. (28). Lesions from 15 alternating sections were stained with Oil Red O, counterstained with hematoxylin, and quantified using imaging software from Kinetic Imaging (Durham, NC).
Statistical analyses of data were performed using unpaired Student's t-tests to compare recipients of apoE−/− and apoE+/+ marrow.
BMT in 2-mo-old ob/ob;apoE−/− mice.
We have previously reported that ob/ob;apoE−/− mice develop more extensive hyperlipidemia and atherosclerotic lesion formation compared with lean apoE−/− controls (11). Previous studies using BMT have demonstrated that macrophage-derived apoE is sufficient to bring plasma lipids to near-normal levels and to ameliorate atherosclerotic lesion formation in lean apoE−/− mice (4, 21, 34). To determine whether macrophage-derived apoE could induce similar effects in obese mice, 2-mo-old recipient ob/ob;apoE−/− mice were transplanted with bone marrow from ob/ob;apoE−/− or ob/ob;apoE+/+ donors (hereafter referred to as ob/ob;apoE−/−→ob/ob;apoE−/− and ob/ob;apoE+/+→ob/ob;apoE−/−, respectively).
To determine levels of apoE in plasma, Western blot analysis was performed. At 12 wk post-BMT, apoE was present in the plasma of apoE+/+ recipients at levels <10% of untransplanted C57BL/6 and ob/ob controls (Fig. 1). ApoE was not detected in recipients of apoE−/− marrow.
At baseline, there were no differences in body weight, total body fat, TC, TG, NEFA, or glucose levels between the two ob/ob;apoE−/− recipient groups (Table 1). At 12 wk post-BMT, all groups were heavier than at baseline. Mice were assessed for total body adipose tissue by nuclear magnetic resonance, and visceral fat pads were weighed. Male mice had more fat than female mice; however, the presence or absence of macrophage-derived apoE did not influence body weight, total body adipose tissue, or visceral fat pad weight. There was a significant reduction in both TC and TG levels in mice receiving apoE+/+ marrow (P < 0.0001 and P < 0.001, respectively) and a trend toward a reduction in NEFA levels. Glucose and insulin levels were not different between recipient groups at 12 wk post-BMT (Table 1).
Analysis of plasma lipoprotein profiles revealed a dramatic reduction in VLDL and LDL, with a concomitant increase in HDL levels in the ob/ob;apoE+/+→ob/ob;apoE−/− mice (Fig. 2A). Despite the improvement in lipoprotein profiles, they were not normalized to those of untransplanted ob/ob mice (Fig. 2B). The ob/ob;apoE+/+→ob/ob;apoE−/− mice retained a mild elevation in VLDL and LDL; however, the LDL/HDL1 peak typical of ob/ob mice (10, 30, 31) was not observed in the transplanted animals. Western blot analysis of apoE in the lipoprotein fractions of ob/ob;apoE+/+→ob/ob;apoE−/− plasma revealed that the apoE was shifted to LDL rather than LDL/HDL1 particles (Fig. 2C).
Atherosclerotic lesion area was analyzed at 12 wk post-BMT. In both male and female mice, lesion area was >10-fold lower in the ob/ob;apoE+/+→ob/ob;apoE−/− mice compared with ob/ob;apoE−/−→ob/ob;apoE−/− controls (P < 0.0001 and P = 0.0003 for males and females, respectively, Fig. 3).
Even though there were no differences in total body adipose tissue or in visceral adipose tissue mass, it is possible that macrophage apoE could influence macrophage recruitment to, or the overall inflammatory status of, adipose tissue. To test this, we isolated RNA and performed real-time RT-PCR analysis on visceral adipose tissue from six mice in each BMT group. Although apoE was expressed in adipose tissue of ob/ob;apoE+/+→ob/ob;apoE−/− mice, we detected no significant differences in expression of macrophage markers F4/80, or inflammatory cytokines TNF-α and IL-6, between the transplantation groups (data not shown).
BMT into 2-mo-old db/db;apoE−/− mice.
Leptin receptor-deficient mice lacking apoE expression (db/db;apoE−/−) are obese and hyperleptinemic (data not shown) and display hyperlipidemia similar to ob/ob;apoE−/− mice (compare Tables 1 and 2, baseline values). Two-month-old db/db;apoE−/− mice were transplanted with marrow from db/db;apoE−/− or db/db;apoE+/+ donor mice. At 12 wk post-BMT, plasma lipids and atherosclerotic lesion area were analyzed. Plasma lipids and lesion area were not different between male and female mice; thus, data from both genders were combined for these studies. At 12 wk post-BMT, the db/db;apoE+/+→db/db;apoE−/− mice demonstrated a dramatic reduction in plasma lipids compared with db/db;apoE−/−→db/db;apoE−/− controls (P < 0.01, Table 2). Atherosclerotic lesion area was also lower in the recipients of db/db;apoE+/+ marrow (P < 0.0001, Fig. 4).
BMT in 4-mo-old ob/ob;apoE−/− and db/db;apoE−/− mice.
To determine the effects of macrophage-derived apoE on preexisting lesions, 4-mo-old ob/ob;apoE−/− and db/db;apoE−/− mice were used as recipients. At 12 wk post-BMT, the recipients of apoE+/+ marrow displayed dramatic reductions in plasma TC levels (P < 0.0001, Table 3); however, TG levels were unchanged. Atherosclerotic lesion area was 2.8-fold lower in the recipients of apoE+/+ marrow compared with apoE−/− recipient controls (P = 0.0002, Fig. 5).
Our current data demonstrate that low levels of plasma apoE are able to dramatically reduce plasma lipids and delay atherosclerotic lesion formation in 2-mo-old and 4-mo-old obese apoE−/− mice. Thus low levels of apoE cannot only prevent the formation of new lesions, it can also delay the progression of preexisting lesions. As obesity trends continue to rise, it has become critical that therapeutic targets for atherosclerotic disease be studied in the context of obesity. Our data provide evidence that apoE may be an important target molecule to consider.
In previous reports, reconstitution of lean apoE−/− mice with bone marrow from apoE+/+ mice resulted in plasma levels of apoE between 3.5 and 10% of wild-type concentrations (21, 34). We have also shown that the threshold amount of apoE required for effects on TC levels in lean mice is 0.04 mg/dl or 1% of normal levels (16). Finally, even when apoE is expressed at levels too low to reduce plasma lipid levels, its secretion from macrophages within atherosclerotic lesions is sufficient to significantly reduce atherosclerotic lesion formation (15, 37). In the current report, we show that, in obese mice, plasma levels of apoE at <10% of normal are sufficient to bring plasma TC and TG concentrations to near-normal levels (Fig. 1). Thus, despite the higher initial TC levels in obese apoE−/− mice compared with lean apoE−/− mice (11), a similarly low level of plasma apoE can dramatically impact plasma lipoprotein clearance.
It has been reported previously that ob/ob mice exhibit a unique lipoprotein profile with an additional peak called LDL/HDL1 (30, 31). Interestingly, in the present study, ob/ob;apoE−/− mice transplanted with apoE+/+ bone marrow demonstrated a near-normalization of TC levels, but the distribution of cholesterol on lipoproteins did not normalize to that of untransplanted ob/ob mice (Fig. 2B). The ob/ob;apoE+/+→ob/ob;apoE−/− mice retained a slight elevation in VLDL and LDL but the LDL/HDL1 peak did not reappear. It has been shown that the accumulation of LDL/HDL1 in ob/ob mice is because of the defective uptake of these particles by the liver (31). Furthermore, we have previously shown that a functional axis of apoAI, hepatic lipase, and SR-BI is required for removal of these LDL/HDL1 particles (10). Although the reason for the absence of LDL/HDL1 in ob/ob;apoE+/+→ob/ob;apoE−/− mice is not clear, it can be postulated that hepatic production of apoE is required for the initial formation of LDL/HDL1 particles. It is also possible that low levels of apoE in plasma alter the way these lipoprotein particles are cleared by the liver. A provocative alternative explanation is that adipocyte-derived apoE is required for their initial formation. More studies are needed to distinguish between these possibilities.
The most likely explanation for the 10-fold decrease in atherosclerotic lesion area in 2-mo-old recipients is the amelioration of hyperlipidemia in the mice that received apoE+/+ bone marrow (Tables 1–3). However, local effects of apoE in the artery wall, as have been shown in lean mice (15, 37), cannot be ruled out. The mechanisms for this are not completely known but may include increased cholesterol efflux or decreased inflammation within the local milieu of the atherosclerotic lesion. It is also important to note that low levels of apoE were sufficient to cause 4-mo-old recipients of apoE+/+ marrow (with preexisting lesions) to have 2.8-fold lower lesion area than control mice receiving apoE-/- marrow. Because we did not quantify lesion area in a separate group of 4-mo-old mice, we cannot determine whether the apoE induced lesion regression. Liver-directed adenoviral delivery of apoE has been shown to cause regression of advanced atherosclerotic lesions related to reductions in plasma lipid levels in apoE−/− mice (13). It has also been shown that expression of apoE in the liver by gene transfer can reduce lesion progression in LDLR−/− mice in the absence of changes in plasma lipid levels (32). Thus apoE is a reasonable target for therapy of atherosclerosis in lean and obese conditions.
In our previous report (11), as well as in our current study, we noted no difference in body weight between ob/ob and ob/ob;apoE−/− mice. It has been reported that lean apoE−/− mice have lower adipose tissue mass than wild-type controls (17). In addition, obese agouti (Ay) mice that are apoE−/− are protected from diet-induced obesity compared with Ay;apoE+/+ controls (8). Furthermore, in high-fat/high-cholesterol diet feeding studies, ob/ob;apoE−/− mice have also been shown to gain less weight compared with ob/ob controls (5). Taken together, these studies suggest that apoE is important for adipocyte lipid accumulation during adipocyte expansion, as in the high-fat diet feeding studies, but that apoE is not required for adipose tissue maintenance in the absence of leptin. It is interesting to note that we detected no difference in body weight, total body adipose tissue, or visceral adipose tissue mass between the recipients of apoE+/+ and apoE−/− marrow in either the presence (db/db) or absence (ob/ob) of plasma leptin. Thus, in leptin-deficient mice that are already obese, the contribution of low levels of macrophage apoE did not alter adipose tissue lipid accumulation. Furthermore, gene expression analysis of macrophage markers F4/80 and CD68, as well as chemokines MCP-1 and MIP-1α and cytokines TNF-α and IL-6, demonstrates that macrophage apoE did not influence macrophage recruitment to or inflammation in adipose tissue in this model system.
In conclusion, reconstitution of obese apoE−/− mice with apoE+/+ marrow resulted in very low plasma levels of apoE (<10% of normal). Although the macrophage-derived apoE did not impact adipose tissue expansion or inflammation in mice that were already obese, these low levels of plasma apoE were sufficient to ameliorate hyperlipidemia and to delay atherosclerotic lesion formation. Furthermore, the protection from atherosclerotic lesion formation could be detected even in mice with preexisting lesions. These data indicate that apoE, and the pathways involved in systemic and local effects of apoE on lesion formation, may be useful targets for treatment of atherosclerotic disease in obesity.
This project was supported by an American Heart Association Scientist Development grant (0330011N). A. H. Hasty is also supported by a Career Development Award from the American Diabetes Association (1-07-CD-10). The Vanderbilt Mouse Metabolic Phenotyping Center is funded by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-59637.
We thank Bonnie Surmi and Saraswathi Viswanathan for careful reading of the manuscript.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 by American Physiological Society