The transcription factor carbohydrate response element-binding protein (ChREBP) mediates insulin-independent, glucose-stimulated gene expression of multiple liver enzymes responsible for converting excess carbohydrate to fatty acids for long-term storage. To investigate ChREBP's role in the development of obesity and obesity-associated metabolic dysregulation, ChREBP-deficient mice were intercrossed with ob/ob mice. As a result of deficient leptin expression, ob/ob mice overeat, become obese and resistant to insulin, and display marked elevations in hepatic lipogenesis, gluconeogenesis, and plasma glucose and triglycerides. mRNA expression of all hepatic lipogenic enzymes was significantly lower in ob/ob-ChREBP−/− than in ob/ob mice, resulting in decreased hepatic fatty acid synthesis and normalization of plasma free fatty acid and triglyceride levels. Overall weight gain in addition to adiposity was reduced in the doubly deficient mice. The former was largely attributable to decreased food intake and may result from decreased hypothalamic expression of the appetite-stimulating neuropeptide agouti-related protein. mRNA expression and activity of gluconeogenic enzymes also was lower in the doubly deficient mice, contributing to significantly lower blood glucose levels. The results of this study suggest that inactivation of ChREBP expression not only reduces fat synthesis and obesity in ob/ob mice but also results in improved glucose tolerance and appetite control.
- carbohydrate response element-binding protein
overeating may be the single most significant cause of the current epidemic of obesity and obesity-associated diseases occurring in the United States. Mechanisms that result in the complex of metabolic derangements that specifically increase risk for type 2 diabetes, heart disease, nonalcoholic fatty liver disease, and other obesity-associated disorders are not completely understood, however, impeding development of effective strategies to combat the pathological consequences of caloric excess. The tendency of obesity to disrupt normal mechanisms of appetite control exacerbates the problem (30).
When carbohydrate intake exceeds short-term requirements for energy or repleting glycogen, metabolizable sugars are converted to fatty acids for long-term storage as fat by the combined activities of the glycolytic and lipogenic pathways in liver. Eating a diet that is high in carbohydrates induces gene transcription of over a dozen enzymes in liver that are involved in the metabolic conversion of carbohydrate to storage fat (8, 12, 25, 26). Insulin, which is secreted by the pancreas in response to elevated blood glucose levels when carbohydrates are eaten, is well known to promote lipogenesis and to increase transcription of many lipogenic enzyme genes. However, nutrients, especially carbohydrates, also play an important role in the transcriptional regulation of many of the same lipogenic genes by mechanisms that are independent of hormone signaling.
The mechanism by which excess carbohydrate activates a glucose-signaling pathway to induce transcription of lipogenic enzyme genes independently of insulin became clearer with the discovery of the transcription factor carbohydrate response element-binding protein (ChREBP) (29). When glucose availability is low, a phosphorylated, inactive pool of ChREBP is maintained in the hepatocyte cytosol (16). When glucose availability increases, glucose metabolism through the pentose shunt pathway increases, resulting in an increased concentration of the metabolite xyulose 5-phosphate. Xyulose 5-phosphate activates a specific protein phosphatase that dephosphorylates ChREBP (15), which then is translocated to the nucleus where it binds to carbohydrate response elements in the promoters of lipogenic enzyme genes (14) and the gene encoding the glycolytic enzyme liver pyruvate kinase [LPK (29)] to coordinately activate their transcription.
In mice with a targeted disruption of the ChREBP gene (ChREBP−/−), de novo fatty acid synthesis and overall adiposity are decreased compared with wild-type (WT) mice (13). ChREBP−/− mice do, however, store modest amounts of fat obtained directly from the diet. Liver triglyceride storage is reduced significantly in ChREBP−/− mice ad libitum fed a high-carbohydrate diet compared with similarly fed WT mice (13). To investigate the effect of ChREBP deficiency on development of obesity and metabolic dysregulation resulting from caloric excess in the studies reported here, ChREBP−/− and ob/ob mice were intercrossed to produce doubly deficient ob/ob-ChREBP−/− mice. The obesity gene (ob) encodes leptin, an adipocyte hormone that regulates the neuroendocrine response to nutrients (4, 7). As the result of a mutation in the ob gene, ob/ob mice do not express leptin, their food intake is increased, and thermogenesis is decreased (4). ob/ob mice become obese and insulin resistant and develop fatty livers. Hyperinsulinemia in ob/ob mice occurs prior to marked hyperphagia or obesity (1, 11) and may result from a direct effect of leptin on pancreatic β-cells to suppress insulin secretion (17).
RESEARCH DESIGN AND METHODS
Standard laboratory chemicals and reagents were purchased from Sigma-Aldrich (St. Louis, MO), unless specified otherwise.
All studies were approved by the University of Texas Southwestern Medical Center Institutional Animal Care and Use Committee. All mice used in these studies were male. The generation of ChREBP−/− mice was described previously (13). Mice were maintained on a 12:12-h light-dark cycle and had ad libitum access to standard laboratory chow (Harlan-Teklad Mouse/Rat Diet 7002; Harlan-Teklad Premier Laboratory Diets). Lepob/+ mice (C57BL6/6J; Jackson Laboratories, Bar Harbor, ME) and ChREBP−/− mice were intercrossed to produce double heterozygous (ChREBP−/+-ob/+) mice, which were intercrossed to produce double homozygous (ob/ob-ChREBP−/−) mice. Because of the low fertility of ob/ob mice, Lepob/+ mice were intercrossed to produce ob/ob offspring for study. ChREBP and Lep genotyping were done by PCR. Genotyping for the ChREBP mutation was described previously (13). The genotyping strategy for the intercrossed mice was based on the fact that the ob mutation generates a DdeI restriction site. A 160-bp region spanning the site of the ob mutation was amplified by polymerase chain reaction using oligonucleotides 5′-TGTCCAAGATGGACCAGACTC-3′ and 5′-ACTGGTCTGAGGCAGGGAGCA-3′, digested with DdeI, and the products resolved on a 3% agarose gels.
Blood and tissue sampling.
Plasma insulin, free fatty acids, and triglyceride concentrations were measured using Grazyme Insulin EIA and NEFA-C Test kits (Wako Pure Chemical, Osaka, Japan) and Triglyceride Reagent (Sigma-Aldrich). Liver triglyceride and glycogen contents were determined as described previously (13).
Hepatocytes were isolated using a collagenase perfusion method with some modifications to our previously described procedure (14). Perfusate solutions were maintained at 32°C to decrease hepatic O2 consumption during perfusion, and the bilateral renal artery and vein and celiac artery were ligated to maximize perfusion of the liver, following the suggestions of Y. Tochino (Dept. of Internal Medicine and Molecular Science, Osaka University).
De novo lipogenesis in hepatocytes.
Hepatocytes were allowed to attach to culture wells in DMEM culture medium containing 100 nM dexamethasone, 10 nM insulin, and 100 nM 3,3′,5-triiodothyronine for 3 h. Nonadherent cells were removed, and hepatocyte monolayers were incubated an additional 3 h in the same medium containing [U-14C]glucose (Amersham, Piscataway, NJ) and a total glucose concentration of 5.5 or 25 mM. At the end of the incubation period, the cells were harvested, total lipids were extracted by the Bligh-Dyer method (3), and incorporated 14C radioactivity was measured by liquid scintillation counting. The water exchange ratio was measured using 3H-labeled water (Amersham) and was ∼60%.
Quantitative real-time RT-PCR.
Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA), and cDNA was synthesized using the SuperScript II System (Invitrogen). Quantitative PCR was performed using the I-cycler system (Bio-Rad). Primer pairs for liver enzymes were those described previously (13). Primer pairs for neuropeptide mRNA quantitation were as follows: proopiomelanocortin (POMC) F 5′-cctggcaacggagatgaac, R 5′-ccacaccgcctcttcctc; agouti-related protein (AgRP) F 5′-tggcggaggtgctagatc, R 5′-cattgaagaagcagcggcagtag; cocaine- and amphetamine-regulated transcript (CART) F 5′-ctactctgccgtggatgat, R 5′-tcttgagcttcttcaggacttc; and neuropeptide Y (NPY) F 5′-ctcgtgtgtttgggcattc, R 5′-gattgatgtagtgtcgcagag.
Experimental mRNA levels were normalized to cyclophilin mRNA in the same sample, and the comparative cycle threshold method was used to compare the normalized mRNA values between animals. Insulin receptor substrate (IRS)-2/IRS-1 mRNA ratios were calculated without correction for any minor differences in SYBR Green binding to the respective PCR products. Taq polymerase was purchased from Promega (Madison, WI).
Liver metabolites and enzyme activities.
Animals were killed by cervical dislocation between 8:00 and 9:00 AM; their livers were removed immediately, freeze-clamped between aluminum blocks cooled in liquid nitrogen, and ground to powder while frozen. Perchloric acid extracts of a portion of the frozen liver powder were prepared and neutralized with KHCO3, and the resulting supernatant solutions were used for metabolite measurements in standard assays (2a). By use of enzymes purchased from Roche Diagnostics (Indianapolis, IN), glucose and glucose 6-phosphate were determined using hexokinase and glucose-6-phosphate dehydrogenase, pyruvate and phosphoenolpyruvate were determined using lactate dehydrogenase and pyruvate kinase, and lactate and malate were determined using lactate dehydrogenase and malate dehydrogenase, respectively. Fructose-2,6-bisphosphate was determined using pyrophosphate:fructose-6-phosphate phosphotransferase (Sigma). Standard assays were used to measure hepatic enzyme activities in additional portions of the same liver samples (2a).
Insulin resistance test.
Fed mice were injected intraperitoneally with 1 U/kg body wt human insulin (Sigma), and plasma glucose concentrations were measured using AACU-Chek Active (Roche Diagnostics).
Phenotypic characteristics of ob/ob-ChREBP−/− mice.
Gross phenotypic characteristics of WT, ChREBP−/−, ob/ob, and ob/ob-ChREBP−/− mice are compared in table 1. Weight gain was significantly reduced in ob/ob-ChREBP−/− mice compared with ob/ob mice. The body weight of ob/ob-ChREBP−/− animals at 6 wk of age was just 60% that of age-matched ob/ob mice and not different from ChREBP−/− mice. Liver weights of ob/ob-ChREBP−/− mice, however, were significantly higher even than those of ob/ob mice, but hepatomegaly in ob/ob-ChREBP−/− mice resulted from high liver glycogen content and not fat storage. The glycogen concentration in livers from ob/ob-ChREBP−/− mice was 2.4- and 3.0-fold that in ob/ob and WT mice, respectively. Liver triglyceride levels in ob/ob-ChREBP−/− mice in contrast were not different from those in WT mice. Plasma triglyceride and free fatty acid levels in ob/ob-ChREBP−/− mice were similar to those in WT and ChREBP−/− mice, and epididymal and brown adipose tissue weights in ob/ob-ChREBP−/− mice were significantly lower than those in ob/ob animals.
Plasma glucose levels in ob/ob-ChREBP−/− mice also were lower than those in ob/ob mice, by ∼30% and only somewhat higher than those in ChREBP−/− mice. Nonetheless, plasma insulin levels were not significantly different between ob/ob and ob/ob-ChREBP−/− mice and were over 10-fold higher than those in WT mice. Systemic insulin resistance in ob/ob-ChREBP−/− mice was comparable to that of ob/ob mice (Fig. 1), suggesting that lower plasma glucose levels in the doubly deficient mice do not result from an increase in insulin-mediated glucose disposal. As observed previously (13), ChREBP−/− mice are modestly insulin resistant (Fig. 1), consistent with small but significant increases in their plasma insulin and glucose levels compared with WT littermates (Table 1).
Liver enzyme and transcription factor expression.
Insulin promotes lipogenesis by activating the transcription factor sterol response element-binding protein (SREBP)-1c (6, 9, 10, 18). However, ChREBP is required for maximal induction of lipogenic enzyme gene expression in liver, as well as that of LPK (13). In ob/ob mice, the level of ChREBP mRNA is twice that in WT animals (Table 2), consistent with their elevated blood glucose. To investigate whether ChREBP deficiency in ob/ob-ChREBP−/− mice results in decreased gene expression of key glycolytic and lipogenic enzymes, levels of mRNA encoding these enzymes in livers of the different mouse strains were determined (Table 2). In liver from ob/ob mice fed a standard diet, all glycolytic and lipogenic enzyme mRNA levels measured were 4–15 times higher than those in WT mice. In ob/ob-ChREBP−/− mouse liver, mRNA levels for lipogenic enzymes, i.e., acetyl-CoA carboxylase (ACC), fatty acid synthase (FAS), and stearoyl-CoA dehydrogenase (SCD), were significantly lower than in ob/ob mice and were comparable with those in ChREBP−/− mice. SREBP-1c mRNA, however, was significantly higher in both ob/ob and ob/ob-ChREBP−/− mice than in either WT or ChREBP−/− mice. mRNA expression of peroxisome-proliferator-activated receptor-γ (PPARγ), a minor PPAR isoform in liver implicated in fatty liver development in ob/ob mice (19), as well as that of another minor isoform, PPARδ, was high in liver of ob/ob mice. In ob/ob-ChREBP−/− mice, mRNA expression of both minor isoforms was slightly lower than in ob/ob mice but still considerably higher than in WT mice. mRNA expression for the predominant liver isoform, PPARα, which regulates gene expression of enzymes involved in fatty acid oxidation (2), was slightly higher in ob/ob than in WT mice and slightly lower in ob/ob-ChREBP−/− mice. These results suggest that transcriptional activation by ChREBP is critical for induction of lipogenic enzyme gene expression in ob/ob mice. Of the glycolytic enzyme mRNAs examined, only that for LPK was lower in ob/ob-ChREBP−/− than in ob/ob mice, consistent with identification of ChREBP as the major regulator of LPK gene transcription.
Glucose-6-phosphatase (G-6-Pase) catalyzes the final step in glycogenolysis and gluconeogenesis and is needed for glucose release from liver. G-6-Pase mRNA levels were elevated in ob/ob mice compared with WT and ChREBP−/− mice (Table 2), consistent with high hepatic glucose output in ob/ob mice (5). mRNA expression of the gluconeogenic enzyme phosphoenolpyruvate carboxykinase (PEPCK), however, was modestly reduced compared with WT mice. In ob/ob-ChREBP−/− mice, PEPCK mRNA was even lower than in ob/ob mice, and G-6-Pase mRNA levels also were lower than those in WT or ChREBP−/− mice, suggesting a decrease in hepatic gluconeogenesis and glucose output in the doubly deficient mice. mRNA expression of PPARγ coactivator-1α (PGC-1α), a coactivator of PEPCK gene transcription, paralleled that of PEPCK mRNA in all strains examined.
The insulin signaling protein IRS-2 is implicated in insulin suppression of hepatic glucose output (21, 27). In ob/ob mice, IRS-2 mRNA levels tended to be lower, whereas IRS-1 mRNA levels tended to be higher, resulting in a significant decrease in IRS-2/IRS-1 mRNA ratio (Table 2). ob/ob-ChREBP−/− mice expressed WT levels of both mRNAs.
To investigate the extent to which observed differences in glycolytic, gluconeogenic, and lipogenic enzyme mRNA expression are reflected by differences in enzymatic activities in these mouse strains, total activity of selected enzymes (Table 3) and intermediate metabolite content (Table 4) in liver were determined. The latter measurements are particularly important for determining the net in vivo activity of liver glycolysis because of the extensive regulation of many glycolytic enzymes by allosteric effectors and the concomitant activity of the opposing gluconeogenic pathway.
Glucose phosphorylation, primarily by glucokinase (GK) in liver, is required for glucose entry into the glycolytic pathway or for storage as glycogen. GK activity in both ob/ob-ChREBP−/− and ob/ob mice was nearly threefold that in WT mice and was consistent with their higher liver GK mRNA levels. In ob/ob mice, however, activity of the opposing enzyme G-6-Pase also was elevated more than twofold compared with the WT value. Consistent with increased activities of both enzymes, liver glucose 6-phosphate content in ob/ob mice was only modestly elevated (Table 4). In contrast, in livers of ob/ob-ChREBP−/− mice, G-6-Pase activity was only one-tenth the WT value. Consistent with a low G-6-Pase/GK ratio, glucose 6-phosphate content in ob/ob-ChREBP−/− mouse liver was nearly 10-fold that in WT mice. The low G-6-Pase/GK ratio is likely to contribute to elevated liver glycogen storage in ob/ob-ChREBP−/− mice.
Pyruvate, lactate, and malate levels are significantly altered in ChREBP−/− mice. The redox state in liver of ChREBP−/− mice is markedly reduced on the basis of the calculation of NAD/NADH and NADP/NADPH couples using these values. We are currently investigating the reason for this unusual change in the redox potential in these mice.
As predicted by decreased mRNA expression of lipogenic enzymes in ob/ob-ChREBP−/− mice, enzymatic activities of the lipogenic enzymes ACC and FAS, and that of malic enzyme, which provides NADPH for lipogenesis, were all decreased significantly in ob/ob-ChREBP−/− compared with ob/ob mouse liver (Table 3). Consistent with the 75–85% decreases measured in the individual enzyme activities, fatty acid synthesis from glucose in hepatocytes isolated from ob/ob-ChREBP−/− mice was reduced by over 70% compared with that in hepatocytes from ob/ob animals under conditions of both low and high glucose concentrations (Table 3). Fatty acid β-oxidation in hepatocytes isolated from ob/ob-ChREBP−/− mice, however, was even lower than in hepatocytes isolated from ob/ob mice (Table 3). Thus decreased fatty acid synthesis in liver of ob/ob-ChREBP−/− mice, and not an increase in fatty acid β-oxidation, appears to account for their lower levels of liver and plasma triglycerides.
To investigate whether a decrease in food consumption might contribute to the overall decrease in weight gain in ob/ob-ChREBP−/− mice, mice of the different strains were fed a standard rodent chow diet without restriction for 15 wk. Food consumption was measured daily, and the mice were weighed every week. ob/ob-ChREBP−/− mice showed a dramatic reduction in weight gain compared with ob/ob littermates at all time points (Fig. 2). Differences in weight gain between the groups were directly related to food intake. ob/ob mice consumed an average of 7.5 ± 0.6 g of chow daily, over one and one-half times the 4.4 ± 0.1 and 4.7 ± 0.1 g consumed daily by WT and ChREBP−/− mice, respectively. ob/ob-ChREBP−/− mice averaged 5.8 ± 0.2 g of chow daily, which was 25% less than the ob/ob animals, although still more than was eaten by either ChREBP−/− or WT mice. To investigate the effect of restricted food intake on weight gain, 9-wk-old mice of all strains were fed 4.4 g/day of standard chow, the amount naturally consumed by WT mice, for 20 days, and their weights were measured daily. Only ob/ob mice lost significant weight, ∼20% of their initial weight (Table 5), by caloric restriction. Weight loss in the ob/ob mice occurred almost entirely in the first week of restricted food intake and then stabilized, whereas the other groups maintained their initial weights throughout (data not shown). Liver triglyceride content was strikingly reduced in ob/ob mice after 20 days of restricted food intake and modestly reduced in WT mice, whereas there was essentially no change in the liver triglyceride content of either ChREBP−/− or ob/ob-ChREBP−/− mice.
Leptin modulates food consumption largely through its effects on appetite-controlling neuropeptides (22). To investigate whether reduced food intake in ob/ob-ChREBP−/− mice might result from altered expression of these neuropeptides by leptin-independent mechanisms, hypothalamic mRNA levels encoding appetite-stimulating or orexic neuropeptides, NPY and AgRP, and the anorexic or appetite-inhibiting peptides CART and POMC-derived α-melanocyte-stimulating hormone were determined (Fig. 3). In ChREBP−/− mice, mRNA levels of all of the neuropeptides were not significantly different than those in WT mice. In ob/ob mice, consistent with their high food intake, NPY and AgRP mRNAs were much higher than in WT mice, whereas CART and POMC mRNA levels were much lower. In ob/ob-ChREBP−/− mice, mRNA expression of anorexic neuropeptides did not differ from that in ob/ob mice. However, NPY mRNA was somewhat higher, whereas mRNA for AgRP was lower, ∼30% less, suggesting that this hormone might be responsible for reduced food intake in ob/ob-ChREBP−/− mice. Consistent with this possibility, intracerebroventricular injection of AgRP to ob/ob-ChREBP−/− mice increased their food intake to levels comparable to those of ob/ob mice (data not shown).
Previously, we demonstrated (13) that ChREBP−/− mice exhibit phenotypic effects that indicate that ChREBP contributes to both basal and carbohydrate-induced expression of all lipogenic enzymes and several key glycolytic enzymes essential for coordinated control of glucose metabolism and the synthesis of fatty acids and triglycerides in vivo. The results presented herein from ChREBP-deficient ob/ob mice extend the important roles of ChREBP to the regulation of carbohydrate utilization and fat storage under conditions of caloric excess.
In liver, ob/ob-ChREBP−/− mice show phenotypic effects quite similar to those in ChREBP−/− mice (13) Decreased induction of LPK and lipogenic enzyme gene expression resulted in markedly decreased hepatic fatty acid synthesis. The near normalization of liver triglyceride content and plasma free fatty acid and triglyceride levels in ob/ob-ChREBP−/− mice suggests that de novo synthesis of fatty acids from carbohydrate, rather than an increase in total intake of dietary fat, is the predominant cause of elevated plasma triglycerides and fatty liver development in ob/ob mice. Caloric restriction in ob/ob-ChREBP−/− mice did not result in any additional decrease in hepatic liver triglycerides, in marked contrast to the greater than 50% decrease in liver triglyceride content in ob/ob mice, providing further evidence of the primary role of ChREBP-regulated fatty acid synthesis in hepatic fat deposition. Epididymal and brown fat weight also were decreased in ob/ob-ChREBP−/− mice. It is not clear whether these effects are secondary to decreased hepatic fatty acid synthesis or result from a direct role of ChREBP in regulating fat synthesis in these tissues. ChREBP expression is detectable in a number of different tissues, albeit at much lower levels than in liver (29).
As observed previously, ChREBP deficiency leads to increased storage of liver glycogen (13). Decreased LPK expression in ChREBP-deficient mice results in lower glycolytic activity and consequently higher liver glucose 6-phosphate content; this ultimately leads to increased glycogen synthesis. The much greater increase in liver glycogen content in ob/ob-ChREBP−/− mice compared with mice deficient solely in ChREBP is likely to result both from greater reduction in G-6-Pase expression, which limits glycogenolysis, and from greater food intake in the doubly deficient mice.
Hyperglycemia in type 2 diabetes and other insulin resistance syndromes results in part from inappropriately high rates of hepatic glucose synthesis and release. The loss of insulin-sensitive repression of key gluconeogenic enzyme gene transcription as the result of downregulated IRS-2 expression is one mechanism suggested to result in this selective hepatic insulin resistance (21, 23, 27). In ob/ob mice in the studies reported here, liver IRS-2 mRNA levels were lower and G-6-Pase but not PEPCK mRNA levels were higher than those in WT mice. ChREBP deficiency in ob/ob mice was associated with normalization of liver IRS-2 and IRS-1 mRNA levels, suppression of liver G-6-Pase mRNA to levels lower than those in WT or ChREBP−/− mice, and a further lowering of PEPCK mRNA levels compared with modestly decreased levels in ob/ob mice. Notably, the apparent effect of ChREBP deficiency in ob/ob mice to improve insulin-sensitive regulation of liver gluconeogenesis occurred despite markedly elevated plasma insulin levels that were not different from those in ob/ob mice. Insulin represses IRS-2 gene transcription through an insulin response element identical to that of the PEPCK gene, and this mechanism was suggested as a potential cause of decreased liver IRS-2 mRNA levels in chronic hyperinsulinemia (31). Results from the studies reported here indicate that ChREBP activity is required for suppression of IRS-2 but not PEPCK mRNA levels in chronically hyperinsulinemic ob/ob mice. IRS-2 and IRS-1 mRNA levels in mice genetically deficient only in ChREBP were not different from those in WT mice. Taken together, these results suggest that downregulation of IRS-2 expression in ob/ob mice occurs as a consequence of hepatic metabolic derangements that are prevented by ChREBP deficiency.
Increased mRNA expression of liver G-6-Pase but not PEPCK was observed previously in ob/ob mice (26), whereas elevated mRNA expression of both PEPCK and G-6-Pase was found in an earlier study (30). Differences in study conditions that may account for these divergent results are not apparent, but it is noteworthy that liver PGC-1α mRNA levels in ob/ob mice in the latter study were also elevated compared with control mice. In contrast, in ob/ob mice described in this report, liver PGC-1α mRNA levels were modestly decreased, and PEPCK but not G-6-Pase mRNA expression paralleled PGC-1α mRNA levels in all mice examined. Elevated G-6-Pase expression in ob/ob mice is thus not dependent on elevated PGC-1α expression, as was suggested previously (30).
Despite significantly lower plasma glucose levels in ob/ob-ChREBP−/− mice compared with ob/ob mice, there was no corresponding decrease in plasma insulin levels. This observation is consistent with previous suggestions that hyperglycemia is not the predominant cause of hyperinsulinemia in ob/ob mice (1, 17). Severe peripheral insulin resistance in ob/ob and ob/ob-ChREBP−/− mice is likely to result, in large part, from persistent hyperinsulinemia. It is perhaps not surprising, then, that the apparent restoration of insulin-sensitive suppression of liver gluconeogenic enzyme activity in ob/ob-ChREBP−/− mice is not sufficient to fully normalize plasma glucose levels.
An additional role of ChREBP uncovered in this study was that of appetite control. The elevated food intake of ob/ob mice was significantly reduced by ChREBP deficiency and was accompanied by reduced expression of the appetite-stimulating neuropeptide AgRP and reversed by AgRP administration. Expression of another appetite-stimulating hormone, NPY, was somewhat increased in ob/ob-ChREBP−/− compared with ob/ob mice. However, although AgRP administration increases cumulative food intake in mice, NPY administration acutely increases food intake without altering cumulative consumption (24). Thus the decrease in AgRP expression with ChREBP deficiency in ob/ob mice is a likely mechanism contributing to decreased food consumption and overall reduction in weight gain in ob/ob-ChREBP−/− mice. Whether ChREBP decficiency decreases AgRP expression in ob/ob mice indirectly or ChREBP directly contributes to increased AgRP gene transcription in ob/ob mice will be the subject of future investigation.
In summary, ChREBP deficiency overcomes obesity in ob/ob mice, reduces specific risk factors for obesity-associated diseases, and improves appetite control, suggesting that reduction of ChREBP activity may have beneficial effects in the treatment of obesity. The current findings suggest that the physiological roles of ChREBP are broad and not limited to liver. Continued investigation into ChREBP function in other tissues as well as liver may uncover new and interesting mechanisms in response to hormones and dietary nutrients.
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 © 2006 by American Physiological Society