Protein Kinase Cβ Deficiency Increases Fatty Acid Oxidation and Reduces Fat Storage*

Metabolic syndrome is common in the general population, but there is little information available on the underlying signaling mechanisms regulating triglyceride (TG) content in the body. In the current study, we have uncovered a role for protein kinase Cβ (PKCβ) in TG homeostasis by studying the consequences of a targeted disruption of this kinase. PKCβ-/- mutant mice were considerably leaner and the size of white fat depots was markedly decreased compared with wild-type littermates. TG content in the liver and skeletal muscle of PKCβ-/- mice was also significantly low. Interestingly, mutant animals were hyperphagic and exhibited higher food intake and reduced feed efficiency versus wild type. The protection from obesity involves elevated oxygen consumption/energy expenditure and increased fatty acid oxidation in adipose tissue with concurrent increased mitochondria genesis, up-regulation of PGC-1α and UCP-2, and down-regulation of perilipin. The ability of PKCβ deficiency to promote fat burning in adipocytes may suggest novel therapeutic strategies for obesity and obesity-related disorders.

intermediate in TG biosynthesis), and has been particularly implicated in the pathogenesis of obesity and insulin resistance (2)(3)(4). PKC consists of several isoforms (5,6), and evidence from knock-out studies has indicated that several PKC isoforms can modulate glucose or energy metabolism in an isoform-specific manner. For example, inactivation of PKC␣ did not alter either basal glucose or energy metabolism (7), whereas inactivation of PKC slightly reduced liver TG contents, without affecting muscle TG levels, and caused increased overall insulin sensitivity (8). On the other hand, inactivation of PKC led to increased susceptibility to obesity and dietary insulin resistance in mice (9). Unlike many PKC isoforms, PKC␤ is expressed as a major isoform in a variety of tissues, and knock out of PKC␤ caused subtle changes in glucose homeostasis but significantly increased insulin-stimulated glucose uptake in adipocytes and muscle (10). We have previously shown involvement of PKC␤ in regulating expression of low density lipoprotein receptor gene in cultured cells (11)(12)(13). Moreover, PKC␤ has been linked with many vascular abnormalities in retinal, renal, and cardiovascular tissues (14,15).
How does PKC␤ deficiency affect glucose metabolism? As several manifestations of type 2 diabetes are associated with alterations in intracellular lipid partitioning, one plausible mechanism is the modulation of tissue TG metabolism. We hypothesized that PKC␤ Ϫ/Ϫ mice would have reduced levels of tissue TG, causing increased sensitivity to insulin. To test this hypothesis, we studied the metabolism of mice with targeted disruption of the PKC␤ gene (PKC␤ Ϫ/Ϫ ), with special attention to lipid metabolism. We found that mice lacking PKC␤ show decreased fat in adipose tissue, liver, and muscle. These mice consumed 20 -30% more food than did the wild type (WT), yet lost body weight. Finally, PKC␤ Ϫ/Ϫ mice exhibit increased fatty acid oxidation with concurrent upregulation of PGC-1␣ and UCP-2 genes. These results raise the possibility that pharmacological manipulation of PKC␤ may lead to loss of body fat in the context of normal caloric intake.

EXPERIMENTAL PROCEDURES
Production and Genotyping of PKC␤ Ϫ/Ϫ Mice in C57BL/6 Background-Generation of PKC␤ Ϫ/Ϫ mice was described earlier (16). They were crossed back ten times to the C57BL/6 background. Mice were housed in a pathogen-free barrier facility, 25°C, 12-h dark/12-h light cycle, and fed regular rodent chow food. All procedures on mice followed the guidelines established by The Ohio State University College of Medicine Animal Care Committee. We developed a PCR method for genotyping using tail tips collected when the mice were weaned at 3 weeks of age. For genotyping, tail DNA was extracted, and polymerase chain reaction (PCR) test was performed using the following primers: PKC␤ forward primer,  5Ј-TGTGCTTTTACAGGGGCTTC-3Ј; PKC␤ reverse  primers, 5Ј-ATTCAGGCTGCGCAACTG T-3Ј and 5Ј-CT-TTCCTGGATGGTGCATCT-3Ј. Polymerase chain reaction conditions were 35 cycles at 95°C for 30 s, 55°C for 45 s, and 72°C for 50 sec. Products were separated by electrophoresis on a 5% polyacrylamide gel and visualized with ethidium bromide staining. Most of the experiments were performed on animals starved for 16 h.
Blood Chemistries-We determined plasma TG and cholesterol concentrations by colorimetric kit assays (Roche Diagnostics). Plasma glucose levels were determined with a kit assay (Sigma). Plasma leptin levels were determined by LINCO, Inc.
Tissue Lipids-We measured tissue TG with a TG 320A kit (Sigma) as described. For qualitative analysis of tissue lipids, lipids were extracted from tissues and separated on ALSILG Silica Gel TLC plates (Whatman) using hexane:ethyl ether:acetic acid (83:16:1).
Histology-Tissues were fixed by immersion or perfusion in neutral buffered formalin, dehydrated in ethanols, transitioned into xylene, and embedded in paraffin. We stained sections with hematoxylin and eosin.
Diet, Feeding, and Weighing-At 32 weeks, the mice were housed individually and allowed to acclimatize for 2 weeks. At 34 weeks of age, the food was weighed three times weekly over a period of 10 weeks, and food intake was determined over the time period of 34 -44 weeks of age (n ϭ 6 of each genotype).
Gene Expression Analysis-Total RNA was isolated from white adipose tissue (WAT) and brown adipose tissue (BAT) of WT and PKC␤ Ϫ/Ϫ mice, using the TRIzol method as described previously (17) (Invitrogen). All RNA samples were digested with DNase I to eliminate any contaminating DNA. Equal amounts of total RNA (1 g/assay) were reverse-transcribed using the SuperScript First-Strand Synthesis Kit (Invitrogen, Life Technology, Inc.) and random hexamer primers. The resulting cDNA (reverse transcription mixture) was subjected to real-time PCR using SYBR Green PCR master mix kit (Applied Biosystems). To ensure the validity of the SYBR Green-based mRNA quantifications, most of the mRNA was also quantified using 32 P-labeled dCTP. This alternative method and the SYBR Green method gave similar results. Forward and reverse primers for amplification of genes are described in supplemental Table 1. All data are expressed as the -fold induction relative to each control value.
Electron Microscopy-Epididymal fat pads and brown adipose tissue from WT and PKC␤ Ϫ/Ϫ mice were fixed in 2% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.4. The samples were postfixed with OsO 4 , dehydrated with alcohol, and embedded in epoxy embedding medium. Thin sections were cut at 70-nm thickness with a Leica EM UC6 ultramicrotome. Samples were stained with 2% uranyl acetate followed by Reynold's lead citrate. Electron microscope imaging of all the prepared materials was done with a Tencai G2 Spirit (FEI, Hillsboro, OR) operated at 80 kV.
Production of 14 CO 2 from [1-14 C]Palmitic Acid and [1-14 C]Oleic Acid-Palmitate and oleate oxidation was measured by the production of 14 CO 2 from [1-14 C]palmitic acid (0.2 Ci/ml) and [1-14 C]oleic acid (0.2 Ci/ml) with unlabeled palmitate and oleate present in the medium. Cells were incubated for 1 h in 20-ml plastic scintillation flasks. The flasks had a centered isolated well containing a loosely folded piece of filter paper moistened with 0.2 ml of 2-phenylethylamine/ methanol (1:1, v/v). After the 1-h incubation period, the medium was acidified with 0.25 ml of H 2 SO 4 (5N), and the flasks were maintained sealed at 37°C for an additional 30 min. At the end of this period, filter papers were carefully removed and transferred into scintillation vials for radioactivity counting.
Statistical Analysis-All results are presented as means Ϯ S.E. Statistical comparisons were by Student's t tests or analysis of variance, as indicated in the text and figure legends. Statistical significance was set at p Ͻ 0.05, where NS indicates not significant.
Histological analysis revealed greater numbers of larger adipocytes in WT than PKC␤ Ϫ/Ϫ mice. This difference in frequency distribution was reflected in a ϳ2.5-fold reduction in mean surface area of adipocytes from PKC␤ Ϫ/Ϫ mutant mice, suggesting that decreased cell size partly contributed to fat mass decrease. The smaller adipocyte size in WAT of PKC␤ Ϫ/Ϫ mice may reflect a certain population sensitive to lipid accumulation. Adipocytes of WT and PKC␤ Ϫ/Ϫ BAT pads were filled with multilocular lipid droplets that, overall, appeared larger in WT mice (Fig. 1G). Thickness of the adipose tissue beneath the dermis in PKC␤ Ϫ/Ϫ mutant mice was also clearly reduced compared with that in the WT mice (Fig. 1H). The mean thickness of adipose tissues of WT and PKC␤ Ϫ/Ϫ mice was 53.1 Ϯ 13.7 and 19.7 Ϯ 3.1 m (mean Ϯ S.E.; n ϭ 10 regions for each genotype of mice; p Ͻ 0.01), respectively.
Although the body weight of PKC␤ Ϫ/Ϫ mice was less than WT (Fig. 1B), these mice appeared to be mildly hyperphagic and consistently consumed a greater caloric load (21.6 Ϯ 0.7 PKC␤ Ϫ/Ϫ versus 17.5 Ϯ 0.4 calories/day WT, a 20% increase for PKC␤ Ϫ/Ϫ versus WT, n ϭ 6, p Ͻ 0.05) (Fig. 3A). When energy intake was calculated relative to an increase in body weight (feed efficiency) over the 10-week period, the PKC␤ Ϫ/Ϫ mice had a lower feed efficiency as compared with the WT mice (0.52 Ϯ 0.02 WT versus 0.73 Ϯ 0.1 calories/g of body weight gained/day PKC␤ Ϫ/Ϫ , 40% increase, p Ͻ 0.001) (Fig. 3B). The plasma leptin levels were significantly reduced in PKC␤ Ϫ/Ϫ mice compared with WT (3.7 Ϯ 0.15 WT versus 0.85 Ϯ 0.07 ng/ml, PKC␤ Ϫ/Ϫ , n ϭ 8, p Ͻ 0.001) (Fig. 3C). It is clear that the decreased body weight is not a result of decreased food intake and lower plasma leptin level may be sufficient to increase appetite. The interesting phenomenon that the PKC␤ Ϫ/Ϫ mice ate more food than WT mice daily, but gained less weight, suggests that  there are important alterations in energy expenditure and disposition. To examine energy expenditure, we compared the weight loss of WT and PKC␤ Ϫ/Ϫ mice after a 16-h fast. Because energy intake is eliminated, the weight loss under fasting conditions provides a simple observation of energy expenditure. Fig. 4A shows that fasting-induced weight loss was higher in PKC␤ Ϫ/Ϫ mice than in the WT (11.7 Ϯ 0.9 PKC␤ Ϫ/Ϫ versus 7.8 Ϯ 0.3% WT, n ϭ 7, p Ͻ 0.003). These results suggested that the decreased adiposity of the PKC␤ Ϫ/Ϫ was due to increased energy expenditure.
We next carried out indirect calorimetry to investigate whether the resistance to weight gain was caused by increased energy expenditure. The PKC␤ Ϫ/Ϫ mice consistently exhibited a 31% increase in oxygen consumption versus WT littermates (5.1 Ϯ 0.83 liter/kg of body weight of VO 2 WT versus 7.4 Ϯ 0.7 liter/kg of body weight of PKC␤ Ϫ/Ϫ during the entire 3 h; n ϭ 6; Fig. 4B). The respiratory exchange ratio of 0.727 Ϯ 0.004 for WT mice and 0.713 Ϯ 0.003 for PKC␤ Ϫ/Ϫ mice showed that both animals were largely using fatty acids as an energy source. Thus, the failure to accumulate fat with age in PKC␤ Ϫ/Ϫ mice seems to stem from a sharp increase in metabolic rates.
To understand the molecular basis of fat loss in PKC␤ Ϫ/Ϫ mice, we analyzed the expression levels of several key genes involved in adipose tissue energy homeostasis. Results are presented in Table 1. Neither peroxisome proliferator-activated receptor ␣ (PPAR␣), which targets enzymes of fatty acid oxida-tion (18), nor PPAR␦, recently shown to activate fat burning (19), were altered in PKC␤ Ϫ/Ϫ mice. The expression of other adipogenic genes, such as CCAAT enhancer-binding protein ␣ (C/EBP␣) and CD36, were also unaltered in mutant mice. There was a significant increase in the expression of UCP-2 in BAT. UCP-2 is widely expressed (20) and was also increased in WAT. PPAR␥-coactivator ␣ (PGC-1␣), a key regulator of mitochondrial biogenesis and respiration (21), was strikingly elevated in both BAT and WAT. This is in agreement with a recent study showing an inverse relationship between PGC-1␣ protein expression and TG accumulation in rodent skeletal muscle (22). The increase in PGC-1␣ may also account for increased expression of UCP-2 (23). Consistent with an increase in PGC-1␣, expression of medium chain acyl coenzyme A dehydrogenase, a nuclearly encoded mitochondrial fatty acid ␤ oxidation enzyme, regulated in parallel with fatty acid oxidation rates, was also increased. There was also an increase in carnitine palmitoyltransferase 2, which is involved in fatty acid transport to mitochondria. The increase in lipogenic SREBP-1c expression is possibly due to lower endogenous TG levels. The decreased expression of the lipid droplet protein perilipin can promote increased lipolysis in the PKC␤ Ϫ/Ϫ adipocytes. Perilipin has recently been shown to play a key role in determining body habitus as its absence leads to a lean and obesity-resistant phenotype in mice (24).
To test whether PKC␤ deficiency affected oxidation of fatty acids in BAT and WAT, production of 14 CO 2 from [1-14 C]palmitate and [1-14 C]oleate was measured. It is significantly increased in adipocytes of WAT (205 ϩ 15% palmitate and 190 ϩ 10% oleate) and BAT (187 ϩ 10% palmitate and 165 ϩ 9% oleate) from PKC␤ Ϫ/Ϫ mice compared with WT littermates (Fig. 5). This appears to be accompanied by increase in number and size of mitochondria containing numerous cristae in both WAT and BAT of PKC␤ Ϫ/Ϫ mice (Fig. 6). The above results are consistent with a recent demonstration that AMPK activation inhibits adipose fatty acid oxidation (25). Leptin is a known activator of AMPK, and low plasma leptin levels in PKC␤ Ϫ/Ϫ mice compared with WT ( Fig. 3) are expected to promote adipose fatty acid oxidation in the PKC␤ Ϫ/Ϫ mice.
It is interesting to note that disruption of PKC␤ in mice achieves several outcomes that are typically associated with the inactivation of key components of adipose tissue storage capacity (26 -29), resulting in enhanced lipolysis or a reduced ability to store TG. Importantly, this reduced energy storage consistently results in increased energy expenditure due to repartitioning of energy substrates, to the point where increased food intake cannot compensate. Increased expression of the mitochondrial genes (PGC-1␣ or UCPs) in PKC␤ Ϫ/Ϫ mice could promote systemic energy expenditure because of thermogenic effects, i.e. generation of heat (reviewed in Ref. 30) and a concurrent decrease in oxidative phosphorylation (31). Potential  mechanisms for elevated fat loss in PKC␤ Ϫ/Ϫ mice may be related to an increase in PGC-1␣ expression resulting in enhanced mitochondrial capacity and "internal combustion," as forced expression of this coactivator in human fat cells enhances fatty acid oxidation (32). Increase in PGC-1␣ would thus expand the oxidative machinery required for enhanced oxidation, while increased UCP-2 protein would dissipate the unneeded energy as heat. Involvement of PKC␤ in regulating mitochondrial biogenesis is supported by a recent observation that linked PKC␤ with mitochondrial survival pathways (33). Taken together, it appears that one of the consequences of PKC␤ deficiency is to partition fat toward increased oxidation and the elevation in fatty acid oxidation capacity is powerful enough to overcome increased food intake. It is not clear whether the increased fatty acid oxidation in adipose tissue is sufficient to significantly promote redistribution of fat from liver and muscle into the adipocytes, much as fat transplantation does in fat-deficient lipodystrophic mice (34). It is also conceivable that central regulatory actions of PKC␤ in the brain, especially hypothalamus, may impact body energy expenditure and metabolic control. Various central nervous system-mediated gene targets result in a lean phenotype, often with increased metabolic rate (35). In any event, PKC␤dependent increased oxidative capacity may have important therapeutic implications for the treatment of obesity and obesity-related disorders. PKC␤ antagonists are currently undergoing clinical trials to reduce diabetes-linked complications (14,15). Our data raise an interesting possibility that inhibition of PKC␤ may also prevent or delay the development of obesity and obesity-related disorders.

TABLE 1 Expression of various mRNAs in WAT and BAT of 8 -11-month-old WT and PKC␤ ؊/؊ mice
Total RNA was isolated from three WT and three PKC␤ Ϫ/Ϫ mice. The RNA was pooled, and 10-ng cDNA aliquots were subjected to real-time PCR. Data are representative of four separate experiments. The -fold change is calculated as the ratio of knockout/control expression and is an average of two experiments. ND, not determined.

Gene
Relative