Involvement of Apolipoprotein E in Excess Fat Accumulation and Insulin Resistance

Animal studies were conducted in accordance with the institutional guidelines for animal experiments at Tohoku University. ApoE^+/−;Ay/+ mice were obtained by mating male KKAy mice (Nippon CLEA, Shizuoka, Japan) and female apoE-deficient mice with the C57BL/6J background (15) (The Jackson Laboratory, Bar Harbor, ME). Male apoE^+/−;Ay/+ mice were mated with female apoE^+/− mice with the C57BL/6J background. Mice of three genotyopes, apoE^+/+;Ay/+, apoE^+/−;Ay/+, and apoE^−/−;Ay/+, were selected. Littermates were used in each experiment. These mice were housed individually and started on a high-fat diet consisting of 15.3% (wt/wt) fat (Quick Fat; Nippon CLEA, Shizuoka, Japan) at 6 weeks of age. Experiments were performed 5 weeks after high-fat loading. Viruses were administered intravenously at a dose of 2 × 10⁸ plaque-forming units 4 weeks after high-fat loading. For pair-feeding experiments, apoE^+/+;Ay/+ mice were allotted into two groups at 4 weeks of age. One group was given their daily food allotments based on the previous days’ consumption by apoE^−/−;Ay/+ littermate mice.

Recombinant adenovirus, containing the human apoE2, E3, E4, or bacterial β-galactosidase gene (Adex1CAlacZ) cDNA under the CAG promoter, was prepared as described previously (16).

Oxygen consumption was measured with an O₂/CO₂ metabolism measuring system (model MK-5000RQ; Muromachikikai, Tokyo, Japan) as described previously (4).

Livers and adipose tissues were removed, fixed with 10% formalin, and embedded in paraffin. Tissue sections were stained with hematoxylin-eosin or 0.1% (wt/vol) Oil Red O. Total adipocyte areas were traced manually and analyzed. Brown and white adipocyte areas were measured in 100 or more cells per mouse in each group.

Frozen livers or adipose tissues were homogenized, and triglycerides were extracted with CHCl₃:CH₃OH (2:1 vol:vol), dried, and resuspended in 2-propanol. Triglyceride contents were measured using Lipidos liquid (Toyobo, Osaka, Japan).

Mice that had been fasted for 16 h were injected with 100 μl normal saline (0.9% NaCl), with or without 10 units/kg body wt of insulin via the tail vein. Hindlimb muscles were removed 300 s later and immediately homogenized. After centrifugation, the resulting supernatants were used for immunoprecipitation with anti–insulin receptor substrate 1 (IRS-1) antibody (17). Immunoprecipitates were subjected to SDS-PAGE and then immunoblotted using antiphosphotyrosine antibody (4G10) as described previously (17).

Blood glucose, plasma insulin, leptin, adiponectin, tumor necrosis factor (TNF)-α, total cholesterol, triglyceride, and free fatty acid (FFA) concentrations were determined as described previously (4). Plasma lipoproteins were analyzed by high-performance liquid chromatography (HPLC) using molecular sieve columns (Skylight Biotech, Akita, Japan) (18).

Glucose and insulin tolerance tests were performed as described previously with slight modification (19). Glucose tolerance tests were performed on fasted (10 h) mice. Mice were given oral glucose (1 g/kg body wt), and blood glucose was assayed immediately before and at 15, 30, 60, and 120 min postadministration. Insulin tolerance tests were performed on fed mice. Mice were injected with human regular insulin (1.5 units/kg body wt; Eli Lilly, Kobe, Japan) into the intraperitoneal space, and blood glucose was assayed immediately before and at 20, 40, 60, and 80 min postinjection.

Hyperinsulinemic-euglycemic clamp studies were performed as described previously (20). Chronically cannulated, conscious and unrestrained mice were fasted for 6 h before the study. Insulin (500 mU · kg⁻¹ · min⁻¹) was infused throughout the clamp study. Blood glucose was monitored every 5 min via carotid arterial catheter samples. Glucose was infused at a variable rate to maintain blood glucose at 120 mg/dl. The glucose infusion rate and endogenous glucose production were calculated as described (20).

Total RNA was isolated from 0.1 g mouse adipose tissue with Isogen (Wako Pure Chemical, Osaka, Japan), and cDNA was synthesized with a Cloned AMV First Strand Synthesis Kit (Invitrogen, Rockville, MD) using 5 μg total RNA. cDNA synthesized from total RNA was evaluated using real-time quantitative PCR (Light Cycler Quick System 350S; Roche Diagnostics, Mannheim, Germany). The relative amount of mRNA was calculated with 28S rRNA as the invariant control. The primers used are described in Supplemental Table 1 (online appendix [available at http://diabetes.diabetesjournals.org]).

Murine 3T3-L1 preadipocytes were maintained and differentiated into adipocytes as described previously (21). β-VLDL (d < 1.006 g/ml) was purified from the blood of apoE^−/− mice with high-fat loading and labeled with fluorescent lipid DiI (1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate) as described previously (22). After labeling with DiI, β-VLDL was incubated with an equal amount of recombinant human apoE3 for 1 h at 37°C (23). DiI-labeled β-VLDL (4 μg protein/ml) with or without apoE had been incubated with fully differentiated 3T3-L1 adipocytes for 8 h at 37°C. Cells were rinsed twice with PBS and solubilized with 0.1 N NaOH/1% SDS. Fluorescent intensities of cell lysates were measured with a fluorescent spectrophotometer (F-2000; Hitachi, Tokyo, Japan) (24).

To assess hepatic triglyceride secretion, 500 mg/kg Triton WR-1339 (Sigma), which blocks lipolysis of triglycerides in peripheral tissue, was injected into the tail veins of the 6 h–fasted apoE^+/+;Ay/+ mice, and blood samples were taken immediately before and 30, 60, and 90 min after injection (25).

Male C57BL/6N mice at 6 weeks of age were injected via tail vein with fluorescent β-VLDL (5 μg/mouse) with or without apoE in 0.2 ml PBS. Thirty minutes after injection, the mice were killed and their livers excised for lipid extraction and fluorescence measurement as described previously (26).

All data were expressed as means ± SEM. The statistical significance of differences was assessed by the unpaired t test and one-factor ANOVA.

This study was prompted by the unexpected observation that aged apoE^−/− mice are leaner and more insulin sensitive than apoE^+/+ mice of the same age with the C57BL/6J background. Therefore, we weighed apoE^−/− and apoE^+/+ mice from 6 to 42 weeks of age, while maintaining the animals on high-fat chow. Body weights were similar through 10 weeks of age, but apoE^−/− mice had significantly lower body weights after 14 weeks of age (Fig. 1A). Glucose tolerance in 30-week-old apoE^−/− mice was moderately better than that in apoE^+/+ littermate mice (Fig. 1B). Thus, apoE deficiency prevents the development of obesity and associated glucose intolerance, while its effects on glucose tolerance are small.

Next, to elucidate the roles of apoE in development of the metabolic disorders associated with obesity, we established apoE^+/+;Ay/+, apoE^+/−;Ay/+, and apoE^−/−;Ay/+ mice. These mice were fed a 15% fat chow from 6 weeks of age. First, we compared body weights. As shown in Fig. 1C, mice gained weight in an apoE gene dosage-dependent manner; apoE deficiency significantly suppressed weight gain. Several organs and tissues were weighed at 11 weeks of age. Weights of the liver and both brown and white adipose tissues were significantly lower in an apoE gene dosage-dependent manner (Fig. 1D). In contrast, weights of other organs, which are minimally involved in lipid accumulation, were similar in apoE^+/+;Ay/+ and apoE^−/−; Ay/+ mice (Supplemental Table 2 [online appendix]). Triglyceride contents of the liver and mesenteric adipose tissue were significantly smaller in apoE^−/−;Ay/+ than in apoE^+/+;Ay/+ mice (apoE^+/+;Ay/+ vs. apoE^−/−;Ay/+ mice: liver, 0.467 ± 0.069 vs. 0.365 ± 0.036 mg/protein, P = 0.02; white adipose tissue, 12.4 ± 1.4 vs. 5.8 ± 1.3 mg/protein, P = 0.002). Histological analyses revealed that apoE deficiency inhibited hepatic fat accumulation, while abundant lipid droplets were present in the livers of apoE^+/+;Ay/+ mice (Fig. 1E and F). In addition, brown (Fig. 1G) and white (Fig. 1H) adipose tissues of apoE^−/−;Ay/+ mice were significantly smaller than those of apoE^+/+;Ay/+ mice. ApoE deficiency also decreased cell diameters in both brown (Fig. 1I) and white (Fig. 1J) adipose tissues. These findings suggest that apoE deficiency results in resistance to obesity via suppression of fat accumulation in the liver and fat tissues under obesity-inducing conditions.

As described previously (15), lipid metabolism was markedly impaired with apoE deficiency (Supplemental Fig. 1A [online appendix]). Plasma cholesterol, triglyceride, and FFA levels were markedly higher in apoE^−/−;Ay/+ than in apoE^+/+;Ay/+ mice. HPLC analysis of plasma lipid profiles revealed that chylomicron, VLDL, and LDL fractions were markedly increased in apoE^−/−;Ay/+ compared with apoE^+/+;Ay/+ mice (Supplemental Fig. 1B [online appendix]).

Next, we examined the effects of apoE deficiency on glucose metabolism as well as insulin sensitivity in genetically obese Ay mice. From age 8 weeks onward, fasting blood glucose was markedly increased in apoE^+/+;Ay/+ mice on high-fat chow, but this glucose elevation was significantly inhibited by apoE deficiency (Fig. 2A). Fasting blood insulin levels at 11 weeks of age were lower, by 78%, in apoE^−/−;Ay/+ than in apoE^+/+;Ay/+ mice (Fig. 2B). Glucose (Fig. 2C) and insulin (Fig. 2D) tolerance tests documented that ApoE^−/−;Ay/+ mice were more glucose tolerant and insulin sensitive. We further examined insulin sensitivity using hyperinsulinemic-euglycemic clamp procedures. ApoE^+/+;Ay/+ mice fed high-fat chow exhibited severe insulin resistance, while in apoE^−/−;Ay/+ mice, glucose infusion rates were markedly higher, by 54-fold (Fig. 2E). Thus, despite hyperlipidemia, especially increased plasma FFA levels, apoE deficiency markedly improves glucose tolerance and insulin sensitivity. In addition, insulin-stimulated tyrosine phosphorylation of IRS-1 in muscle was enhanced in apoE^−/−;Ay/+ compared with apoE^+/+;Ay/+ mice (Fig. 2F). Thus, apoE deficiency apparently prevents surplus fat accumulation in adipose tissue and insulin resistance in muscle, resulting in better glucose tolerance.

We next determined plasma adipocytokine profiles (Fig. 3A). In apoE^−/−;Ay/+ mice, plasma leptin levels were slightly decreased and TNF-α levels were markedly lower than those in apoE^+/+;Ay/+ mice, while plasma adiponectin levels were significantly higher. Thus, apoE deficiency improved obesity-induced alterations in adipocytokine profiles. In addition, quantitative RT-PCR revealed that expressions of F4/80, monocyte chemoattractant protein-1, inducible nitric oxide synthase, and interleukin-6 in mesenteric adipose tissue were significantly lower in apoE^−/−;Ay/+ than in apoE^+/+;Ay/+ mice (Fig. 3B), suggesting inhibition of inflammation and macrophage invasion into adipose tissue. Obesity is reportedly associated with macrophage infiltration of adipose tissue, which is likely to promote insulin resistance (27,28). Inhibition of macrophage invasion of adipose tissue may be involved in the higher insulin sensitivity and glucose tolerance observed in apoE^−/−;Ay/+ mice.

As described above, plasma adiponectin levels were increased in apoE^−/−;Ay/+ mice, suggesting normal adipocyte differentiation in vivo. Furthermore, mRNA expressions for adipocyte-related proteins, such as CCAAT/enhancer binding protein-α, peroxisome proliferator–activated receptor-γ, and aP2, were similar among adipose tissues from apoE^+/+;Ay/+, apoE^+/−;Ay/+, and apoE^−/−;Ay/+ mice (Fig. 3C). mRNA expressions levels of these three genes were also similar in adipose tissues from younger apoE^+/+;Ay/+ and apoE^−/−;Ay/+ mice, 4 weeks of age, when body weights were not significantly different (data not shown). In addition, apoE deficiency did not alter sterol regulatory element–binding protein 1 (SREBP1) expression (Fig. 3C). These findings indicate that adipocytes are normally differentiated in apoE^−/−;Ay/+ mice in vivo.

To confirm that the metabolic phenotypes observed in apoE^−/−;Ay/+ mice were, in fact, mediated by apoE deficiency, we examined metabolic effects of adenovirus-mediated apoE expression in the livers of apoE^−/−;Ay/+ mice. Replenishment of apoE protein (human apoE3) resulted in markedly decreased plasma cholesterol, triglyceride, and FFA levels (Supplemental Fig. 2A [online appendix]), indicating functional expression of apoE. HPLC analyses of plasma lipid profiles revealed that adenoviral replenishment of apoE protein in apoE^−/−;Ay/+ mice markedly decreased the chylomicron and VLDL fractions (Supplemental Fig. 2B [online appendix]).

Increases in body weight for 7 days after adenoviral administration were significantly greater with the apoE adenovirus than with the LacZ control adenovirus (Fig. 4A). Liver weights tended to be increased (Fig. 4B), and those of brown adipose (Fig. 4C) and epididymal, mesenteric, and retroperitoneal white adipose (Fig. 4D) tissues were significantly increased with apoE adenoviral administration. Histological analyses revealed that apoE replenishment increased sizes of brown (Fig. 4E and F) and white (Fig. 4G and H) adipocytes. In addition, glucose tolerance tests revealed that apoE replenishment worsened glucose tolerance in apoE^−/−;Ay/+ mice (Fig. 4I). These findings show clearly that circulating apoE contributes to increased adiposity and the glucose intolerance associated with obesity. Furthermore, plasma leptin levels were significantly increased on day 7 after adenoviral administration. TNF-α and adiponectin levels tended to be increased and decreased, respectively (Fig. 4J), suggesting that circulating apoEs are involved in obesity-induced alterations in adipocytokine levels.

ApoE occurs in three major isoforms (apoE2, -E3, and -E4) in humans. ApoE3, the most common isoform, is considered to be the wild type. To compare the roles of the three human apoE isoforms in obesity and diabetes, recombinant adenoviruses encoding human apoE2 and -E4 as well as apoE3 were injected into apoE^−/−;Ay/+ mice. Administration of these apoE adenoviruses resulted in similar expression amounts of apoE proteins (data not shown), and similar increases in body weights (Supplemental Fig. 3A [online appendix]) and blood glucose levels (Supplemental Fig. 3B [online appendix]). These findings suggest that the three apoE isoforms contribute similarly to fat accumulation and glucose tolerance.

Why does apoE deficiency inhibit obesity in genetically obese mice? We next examined the uptake of β-VLDL, with or without apoE, into fully differentiated 3T3-L1 adipocytes. β-VLDL obtained from apoE^−/− mice was labeled with DiI, followed by incubation with or without recombinant human apoE3. As shown in Fig. 5A, uptake of apoE-deficient VLDL was markedly lower, by 85%, than that of apoE-positive (after incubation with human apoE3) VLDL. These findings suggest that impaired VLDL uptake into adipocytes contributes to decreased adiposity in apoE^−/−;Ay/+ mice. Thus, VLDL uptake into adipocytes is likely to play a role in excess fat deposition and, thereby, in the development of diabetes associated with obesity.

ApoE deficiency reportedly reduces hepatic VLDL secretion, resulting in fatty liver findings (11). In our model as well, hepatic triglyceride secretion was inhibited in apoE^−/−;Ay/+ mice compared with apoE^+/+;Ay/+ mice, by 48% (Fig. 5B). However, interestingly, apoE^−/−;Ay/+ mice displayed less fat accumulation in the liver than apoE^+/+;Ay/+ mice (Fig. 1D–F). To elucidate the underlying mechanism, we examined β-VLDL uptake into the liver. Fluorescence-labeled β-VLDL, with or without apoE, was intravenously injected into wild-type C57BL/6 mice. Fluorescence values in the liver were then measured. Uptake of apoE-deficient β-VLDL was markedly lower, by 49%, than that of apoE-positive VLDL (Fig. 5C). Thus, despite decreased secretion, decreased β-VLDL uptake with apoE deficiency may contribute to prevention of hepatic steatosis. Therefore, apoE is likely to be involved in excess fat uptake into hepatocytes as well as adipocytes. Taken together with the findings that adenoviral apoE replenishment decreased the VLDL fraction (Supplemental Fig. 2B [online appendix]), our results indicate that apoE-dependent VLDL transport into tissues, including the liver and adipose tissue, is involved in the development of obesity, resulting in glucose intolerance and insulin resistance.

Next, to elucidate the systemic mechanism underlying the obesity prevention associated with apoE deficiency, we first measured food intakes in apoE^+/+;Ay/+ and apoE^−/−;Ay/+ mice. Interestingly, apoE deficiency in Ay mice significantly suppressed food intake (Fig. 5D). Then, to eliminate the secondary effects of reduced food intake in apoE^−/−;Ay/+ mice, ApoE^+/+;Ay/+ mice were allotted the same amounts of food consumption as apoE^−/−;Ay/+ mice, followed by weight measurement and glucose tolerance testing. Pair-feeding blunted the body weight increments in apoE^+/+;Ay/+ mice, while the weights of pair-fed apoE^+/+;Ay/+ mice were significantly greater than those of apoE^−/−;Ay/+ mice (Fig. 5E). ApoE^−/−;Ay/+ mice exhibited better glucose tolerance than pair-fed apoE^+/+;Ay/+ mice (Fig. 5F). Thus, the inhibition of obesity and glucose intolerance by apoE deficiency is not attributable solely to decreased food intake.

Next, we measured basal metabolic rates. As shown in Fig. 5G, resting oxygen consumption in both the light and the dark phase at 5 weeks of age was significantly greater in apoE^−/−;Ay/+ than in apoE^+/+;Ay/+ mice. Taken together with the pair-feeding experiment results, these findings show that decreased food intake and increased energy expenditure both contribute to the prevention of obesity and insulin resistance with apoE deficiency and that apoE is involved in regulation of energy metabolism.

To examine the effects of apoE deficiency on insulin resistance associated with obesity, apoE^−/− mice were interbred with KK-Ay mice. ApoE^−/−;Ay/+ mice showed resistance to the development of obesity and glucose intolerance. Insulin sensitivity was markedly greater in apoE^−/−;Ay/+ than in apoE^+/+;Ay/+ mice. Recently, several attempts to induce obesity in apoE^−/− mice have been reported, but the results have been somewhat controversial (12–14). In the present study, in addition to inhibition of adiposity and insulin resistance with apoE deficiency, adenoviral apoE replenishment reversed inhibition of obesity and glucose intolerance. These findings directly demonstrate apoE involvement in the development of obesity and obesity-related disorders of glucose metabolism and insulin sensitivity. Chiba et al. (14) previously reported that ob/ob;apoE^−/− mice are also resistant to obesity. They attributed decreased adiposity in ob/ob;apoE^−/− mice to impaired adipocyte differentiation based on in vitro findings that expression levels of aP2 and peroxisome proliferator–activated receptor-γ were markedly lower when bone marrow stromal cells and 3T3-L1 cells were cultured with apoE-less VLDL. However, in the present study, adipocytes in apoE^−/−;Ay/+ mice expressed these adipocyte-related proteins normally in vivo. Furthermore, apoE^−/−;Ay/+ mice showed better insulin sensitivity and less hepatic lipid accumulation, accompanied by improved adipocytokine profiles, than apoE^+/+;Ay/+ mice. Thus, decreased adiposity and improved insulin sensitivity in apoE^−/−;Ay/+ mice can be explained by factors other than adipocyte differentiation.

While body weight and adiposity were similar in young apoE^−/− and apoE^+/+ mice, apoE deficiency ameliorated obesity and insulin resistance under obesity-inducing conditions, such as aging, genetic susceptibility, and dietary loading, suggesting that apoE is involved in obesity development, i.e., excess fat accumulation, while being less involved in basal fat accumulation. Lack of apoE in β-VLDL markedly impaired β-VLDL transport into adipocytes. ApoE is an important component of VLDL. There are several receptors, including VLDL receptor (VLDLR) and LDL receptor–related protein, which recognize VLDL in an apoE-dependent manner (22). Among them, VLDLRs reportedly have similar affinities for apoE2, -E3, and -E4 isoforms (29). In addition, VLDLR-deficient mice reportedly exhibit obesity resistance with high-fat chow loading (30). Taken together with our findings that replenishment of apoE2, -E3, and -E4 isoforms contributes similarly to fat accumulation and glucose tolerance in apoE^−/−;Ay/+ mice, the apoE-VLDLR interaction plays an important role in the development of obesity. Furthermore, it is well known that high levels of plasma VLDL are associated with obesity and type 2 diabetes (31). Thus, receptor-mediated VLDL transport into adipocytes in an apoE-dependent manner is involved mainly in excess lipid uptake into adipocytes. Lipid uptake, required for adipocyte differentiation and metabolic activities, might be mediated mainly by apoE-independent lipid transport pathways or de novo lipid synthesis in adipocytes.

In addition, transport of apoE-deficient β-VLDL into the liver was markedly impaired compared with that of apoE-positive β-VLDL. Despite decreased triglyceride secretion, apoE deficiency decreased hepatic fat accumulation in Ay, but not in wild-type, mice (11). Hepatic expressions of SREBP1c and fatty acid synthase were similar in apoE^−/−;Ay/+ and apoE^+/+;Ay/+ mice (data not shown), suggesting no apparent decrease in hepatic fatty acid synthesis in apoE^−/−;Ay/+ mice. These findings suggest that apoE-dependent uptake of β-VLDL into hepatocytes is involved in the development of hepatic steatosis in Ay mice. The machinery that transports β-VLDL into the liver, including the receptor(s) playing a major role, is a potential target for elucidating the mechanism underlying hepatic steatosis.

Intriguingly, in apoE^−/−;Ay/+ mice, food intake was decreased and energy expenditure enhanced compared with apoE^+/+;Ay/+ mice. Pair-feeding experiments revealed that both these phenomena result in obesity resistance in apoE^−/−;Ay/+ mice. There appear to be several possible explanations for these alterations in energy metabolism. First, hyperlipidemia induced by apoE deficiency might contribute to decreased food intake and increased energy expenditure. However, LDLR^−/− mice, which also exhibit hyperlipidemia, reportedly become more obese and diabetic in response to high-fat and high-carbonate diets than wild-type mice (13). In addition, food intake was similar in LDLR^−/− mice and LDLR^+/+ mice (13). Therefore, although hyperlipidemia is more severe in apoE^−/− than in LDLR^−/− mice, involvement of hyperlipidemia in food intake regulation might be unlikely in our model. Second, we speculate that leptin sensitization is involved in decreased food intake and increased energy expenditure. Obesity is well known to be associated with poor responses to leptin despite hyperleptinemia, a state defined as leptin resistance (32). Lower plasma leptin levels with lower body weight in apoE^−/−;Ay/+ mice compared with apoE^+/+;Ay/+ mice suggests greater leptin sentivity in the former. Therefore, decreased food intake and increased energy expenditure in apoE^−/−;Ay/+ mice might be explained by leptin sensitization. We have recently reported that alterations in metabolism in adipose tissue affect food intake amounts (19). However, ob/ob;apoE^−/− mice are also reportedly resistant to obesity (14). Since ob/ob mice are leptin deficient, the obesity resistance in ob/ob;apoE^−/− mice is not mediated by leptin signaling, e.g., leptin sensitization, although food intake and energy expenditure were not measured in the earlier study (14). Thus, mechanisms other than leptin sensitization might be involved in decreased food intake and increased energy expenditure in apoE^−/−;Ay/+ mice. Third, direct effects of apoE on neurons are also possible. ApoE, produced by glial cells, is a major apolipoprotein in the brain and mediates the transport of cholesterol and phospholipids, and its receptors are abundantly expressed on neurons (33). Furthermore, numerous studies have shown that apoE plays multiple roles in the nervous system. In the central and peripheral nervous systems, apoE promotes neurite outgrowth and regeneration (34). ApoE protects neurons from oxidative injury (35) and modulates amyloid-β deposition (36), interactions with Alzheimer amyloid precursor protein (37), and transmission of signals to neurons (38). In this context, modulation of neurons by apoE might be involved in energy metabolism. ApoE is reportedly expressed in tissues other than the liver, including the brain (33) and adipose tissue (39). ApoE deficiency in these tissues may affect the metabolic phenotypes of apoE^−/−;Ay/+ mice observed herein. Intensive research, including tissue-specific disruption of apoE or its receptor, is required to examine this hypothesis.

ApoE is involved in surplus fat accumulation and energy metabolism, including regulation of food intake and energy expenditure. Thus, excess fat accumulation via an apoE-dependent pathway might play a role in development of the metabolic syndrome. In addition to dissipation of surplus energy (4), apoE-dependent excess lipid transport is a potentially novel therapeutic target for the metabolic syndrome.

This work was supported by a Grant-in-Aid for Scientific Research (B2, 15390282) to H.K.; a Grant-in-Aid for Scientific Research (17790599) to Y.I. from the Ministry of Education, Science, Sports and Culture of Japan; and a Grant-in-Aid for Scientific Research (H16-genome-003) to Y.O. from the Ministry of Health, Labor and Welfare of Japan. This work was also supported by the 21st Century COE Programs “CRESCENDO” to H.K. and “the Center for Innovative Therapeutic Development for Common Diseases” to Y.O. from the Ministry of Education Science, Sports and Culture.

We thank I. Sato, J. Fushimi, K. Kawamura, and M. Hoshi for technical support.