| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
,1
* Department of Medicine, Division of Gastroenterology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA;
Departments of Medicine and Molecular Pharmacology, Diabetes Research Center, and
Department of Physiology and Biophysics, Albert Einstein College of Medicine, Bronx, New York, USA;
Research Division, Joslin Diabetes Center, Harvard Medical School, Boston, Massachusetts, USA; and
|| Division of Health and Sciences and Technology, Harvard-Massachusetts Institute of Technology, Boston, Massachusetts, USA
2Correspondence: Brigham and Women’s Hospital, 75 Francis St., Boston, MA 02115. E-mail: dcohen{at}partners.org
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: fatty acid • triglyceride • glucose • respiratory quotient • phospholipid
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
. Whereas related proteins participate in intracellular lipid transport, metabolism, and cellular signaling (3
4
5)
, the biological function of PC-TP is uncertain. Mice lacking PC-TP are relatively resistant to atherosclerosis (6)
, and a coding region polymorphism in human PC-TP is predictive of larger, less atherogenic LDL particles (7)
.
The distribution of PC-TP expression is broad (8)
and appears as early as embryonic stem cells (9)
. Levels are highest in metabolically active tissues, including liver, heart, muscle, and kidney (8)
. PC-TP is distributed in both the cytoplasm and nucleus of cells (8)
. In response to clofibrate exposure, cytoplasmic PC-TP rapidly associates with mitochondria in some cell lines (10)
. Recently, yeast two-hybrid screening (11)
has demonstrated that PC-TP interacts with the mitochondrial-associated protein thioesterase superfamily member 2 (Them2; ref. 12
), suggesting the possibility that PC-TP plays a role in mitochondrial fatty acid metabolism (8
, 10)
.
We observed that Pctp–/– mice exhibit reduced fasting plasma glucose and free fatty acid (FFA) concentrations. Evaluation by glucose and insulin tolerance tests, as well as hyperinsulinemic-euglycemic clamp studies, revealed increased insulin sensitivity associated with profound decreases in hepatic glucose production in Pctp–/– mice in the absence of changes in peripheral glucose uptake from plasma. In addition to increased insulin-mediated Akt phosphorylation in cultured hepatocytes lacking PC-TP expression, marked increases in body fat composition of Pctp–/– mice were accompanied by increased plasma leptin and adiponectin and decreased tumor necrosis factor-
(TNF) concentrations. The changes in body fat could be linked to preferential fatty acid utilization for energy production and increased turnover of triglycerides. Collectively, these data support a key function for PC-TP in regulating energy substrate utilization.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
were backcrossed seven generations into FVB/NJ genetic background (14)
, and wild-type littermates were used as controls. All animals were housed in a standard 12 h alternate light-dark cycle facility and fed standard rodent diet 5001 (LabDiets, St. Louis, MO, USA) with free access to drinking water. Protocols for animal use and euthanasia were approved by the institutional committees of the Harvard Medical School or the Albert Einstein College of Medicine. Experiments were conducted using male mice at 8–12 wk of age.
Analytical techniques
Plasma triglyceride concentrations were determined using a serum triglyceride determination kit from Sigma (St. Louis, MO, USA). Plasma FFA concentrations were determined using a NEFA C free fatty acid kit from Wako Chemicals (Richmond, VA, USA). Blood glucose was determined using a OneTouch Ultra glucose monitor (LifeScan, Milpitas, CA, USA). Hepatic triglyceride concentrations were measured enzymatically after hepatic lipid extraction (15)
. Plasma insulin, leptin, adiponectin, and TNF concentrations were determined by ELISA as a service of the Joslin Diabetes and Endocrinology Research Center Specialized Assay Core (NIH 5P30 DK-36836, Joslin Diabetes Center, Boston, MA, USA). Plasma β-hydroxybutyrate concentrations were determined enzymatically (16)
. Hepatic concentrations and compositions of fatty acyl-CoAs were determined by HPLC (17)
.
Glucose and insulin tolerance tests
Glucose and insulin tolerance tests were performed after an overnight fast (18)
. Blood glucose was first measured at baseline. After an intraperitoneal injection of 2 mg/g body wt D-glucose (20% wt/vol) for glucose tolerance tests or 1 U/kg body wt insulin (HumulinR, Eli Lilly, Indianapolis, IN, USA) for insulin tolerance tests, blood glucose concentrations were measured periodically for up to 60 min (insulin tolerance tests) and 180 min (glucose tolerance tests).
Hyperinsulinemic-euglycemic clamp studies
Studies were preformed in conscious, unrestrained mice fitted with intravenous catheters (19)
. Briefly, food was removed 5 h before commencing the studies, and infusions lasted for 90 min. Mice received a constant infusion of HPLC-purified insulin (3.6 mU/min/kg body wt) and [3H-3]glucose (0.1 µCi/min; Perkin Elmer, Waltham, MA, USA). A glucose solution (10% wt/vol) was infused at a variable rate so that the plasma glucose concentration was constant at 8 mM throughout the experiment, as verified by measurements of plasma samples collected at 10 min intervals. Steady-state values of both plasma glucose concentration and specific activity were achieved by 40 min, so that [3H-3]glucose-specific activities were determined at 10 min intervals from 40 to 90 min. Ten minutes before the end of the experiment, [U-14C]lactate (5 µCi bolus, 0.4 µCi/min continuous infusion; Perkin Elmer) was administered in order to determine the contributions of gluconeogenesis, glycogenolysis, and glucose cycling to hepatic glucose output. At the end of the studies, mice were anesthetized, abdomens were quickly opened, and livers were freeze-clamped in situ with aluminum tongs that were precooled with liquid nitrogen. As described previously (20)
, plasma glucose concentrations and specific activity, as well as the specific activities in the liver of phosphoenolpyruvate and UDP-glucose, were measured and used to determine rates of glucose uptake, hepatic glucose production, total glucose output, glucose cycling, gluconeogenesis, and glycogenolysis.
Gene expression analysis
Relative mRNA expression levels were measured by quantitative real-time polymerase chain reaction (PCR). Briefly, total RNA was extracted from either liver or fat tissue using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. With the use of a Superscript III first-strand kit (Invitrogen), cDNA was synthesized from 2 µg of total RNA from liver and 100 ng of total RNA from fat tissue. Transcript levels were quantified using a Fast Start DNA Master SYBR Green I kit (Roche Applied Sciences, Indianapolis, IN, USA) in a LightCycler 2 apparatus (Roche Applied Sciences). The L32 ribosomal protein was used as an invariant control. Primer sequences were designed using Primer 3 (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi) with nucleotide sequences obtained from GenBank (National Center for Biotechnology Information, Bethesda, MD, USA). Gene-specific primer sequences are listed in Supplemental Table 1.
Primary hepatocyte cultures
Primary hepatocytes were isolated and cultured according to a standard procedure (21)
. Briefly, mice were anesthetized with an intraperitoneal injection of ketamine (87 mg/kg body wt; Webster Veterinary, Sterling, MA, USA) plus xylazine (13 mg/kg body wt; Webster Veterinary). Thereafter, the inferior vena cava was exposed, cannulated, and perfused for 5 min with liver perfusion media (Invitrogen), followed by a 10 min perfusion with liver digestion media (Invitrogen), each having been prewarmed to 37°C. The digested liver was diced in cold hepatocyte wash media (Invitrogen), passed through an 80 µm nylon mesh (Sefar-America, Depew, NY, USA), and washed an additional 3x. Cells were pelleted and resuspended in cold Williams E medium containing 10% FBS, 10–7 M dexamethasone, 10 µg/ml insulin, and 5 µg/ml transferrin. Viability was estimated by the trypan blue exclusion, and only preparations with viability in excess of 80% were used for experiments. Cells were plated overnight in 6-well Primaria plates (BD Biosciences, San Jose, CA, USA) at a density of 5 x 105 per well.
Insulin-mediated stimulation of Akt phosphorylation
After overnight incubation in culture, hepatocytes were washed 2x with warm PBS and incubated for a period of 12 h in Dulbecco modified Eagle medium (DMEM; Sigma) containing 20 mM lactate and 2 mM pyruvate (Sigma) but no glucose. Cells were then exposed to increasing insulin concentrations for a period of 30 min and then harvested in cold radioimmunoprecipitation assay buffer supplemented with protease inhibitors (Roche Diagnostics, Indianapolis, IN, USA) plus the phosphatase inhibitors NaF (50 mM), β-glycerophosphate (10 mM), activated sodium orthovanadate (2 mM), and EDTA (2 mM). Lysates were rotated slowly at 4°C for 30 min and then centrifuged at 12,000 g for 10 min to remove cellular debris. Proteins were fractionated by SDS-PAGE and subjected to Western blot analysis using primary rabbit polyclonal antibodies (Cell Signaling Technology, Danvers, MA, USA) to Akt and phosphorylated Akt (p-Akt; Ser-473) with detection by enhanced chemiluminescence (Western lightning chemiluminescence reagent, Perkin Elmer).
Rates of fatty acid synthesis
After overnight incubation in culture, primary hepatocytes were washed 2x with warm PBS and incubated with DMEM containing high glucose (Invitrogen) plus 0.5 µCi of [1-14C]acetate (57.5 mCi/mmol, Sigma) per well. After 16 h, cells were mixed with 2.5 ml of 10% KOH in methanol and 2 ml of distilled water. The mixture was heated to 90°C for 3 h and extracted 3x with 4 ml of petroleum ether. The lower aqueous phase was then mixed with 1 ml of 10 M H2SO4 and extracted 3x with petroleum ether. The organic phase was transferred to a scintillation vial, dried under nitrogen, resuspended in Ecoscint H (National Diagnostics, Atlanta, GA, USA), and counted in a scintillation counter.
Measurements of body fat
Body fat mass was assessed using two independent methods: nuclear magnetic resonance (NMR) measurements (EchoMRI, Echo Medical Systems, Houston, TX, USA) were utilized to determine the body fat mass of individual mice (22)
; the same cohort of mice was also analyzed by microcomputerized tomography (CT). A LaTheta LCT-100A MicroCT scanner (Aloka Instruments, Tokyo, Japan) was used to scan mice. While the mice were under isoflurane inhalation anesthesia, CT scanning was performed at 2 mm intervals (80 sections) under slow speed (18 s/slice) and high voltage conditions using medium view settings at a 37 mm diameter. CT scan data were stored using Aloka software (Aloka) for offline analysis using Amira image analysis software (Mercury Computer Systems, San Diego, CA, USA). The images were loaded into Amira 4.0 sequentially as a single entity with the proper order and slice separation. The fat tissue was selected and defined using the "Label Voxel" function in Amira with the properly defined intensity range. Subcutaneous fat and intraabdominal fat were then differentiated by the abdominal muscle wall and were defined separately. Finally, three-dimensional images of the subcutaneous fat and visceral fat were generated by the SurfaceGen function and the fat volume was calculated using the voxel size returned by the Tissue Statistics function in Amira.
Calorimetry
Indirect calorimetry was performed (19)
with mice individually housed with free access to food and water in Oxymax calorimeter cages (Columbus Instruments, Columbus, OH, USA). Mice were first acclimated for a period of 24 h, and then gas exchange, activity, and heat production values were collected periodically up to 24 h, with instrument settings and gas sensor calibrations as described previously (19)
. Rates of O2 consumption (VO2, ml/h/kg body wt) and of CO2 production (VCO2, ml/h/kg body wt) were measured using electrochemical and spectrophotometric sensors, respectively. Respiratory quotient was calculated as the ratio of VCO2 to VO2. Heat production was determined based on gas exchange according to the manufacturer’s algorithm. Activity of the mice was recorded by sensors built into the calorimeter cages. Gas exchange parameters were further analyzed by the percent relative cumulative frequency (PRCF) method (23)
. Data were transformed to create PRCF curves, which were fitted by nonlinear regression (Prism 4, GraphPad, San Diego, CA, USA) to the function PRCF = 100xH/(EC50H+XH), where X represents the gas exchange parameter, EC50 is the X value that corresponds to the 50th percentile, and H is the slope of the curve between 10 and 90% PRCF values. Curve fits were considered acceptable for R2 
0.98 (23)
.
Clearance and distribution of orally administered triglycerides
Mice were deprived of food for 16 h before oral triglyceride gavage. For measurements of clearance (24)
, blood was collected via a retro-orbital puncture at baseline and hourly up to 5 h for determination of plasma triglycerides. To determine tissue distribution, olive oil was supplemented with 2 µCi [1-14C]-oleic acid (54.6 mCi/mmol, Perkin Elmer). Four hours after gavage, mice were euthanized and tissues were immediately harvested. 14C content was determined by liquid scintillation counting and normalized to wet weight of tissues.
Hepatic triglyceride production rates
Hepatic triglyceride production rates were measured as described previously (25)
. Briefly, after 4 h of fasting, a baseline blood sample was obtained and mice were injected via tail vein with 20 mg Triton WR1339 (Sigma) dissolved in 200 µl PBS. Thereafter, blood was sampled periodically up to 120 min for determination of plasma triglyceride concentrations. Rates of triglyceride production (mg/kg/h) were calculated based on the slopes of linear increases in plasma triglyceride concentrations over time, assuming a plasma volume of 3.5% of body weight (25)
.
Nuclear expression of sterol response element binding protein 1c
Nuclear extracts were prepared from liver tissue (26)
, and proteins were fractionated by SDS-PAGE and subjected to Western blot analysis using mouse monoclonal antibodies to sterol response element binding protein 1c (Clone 2A4; Lab Vision Products, Fremont, CA, USA) with detection by enhanced chemiluminescence.
Lipoprotein and hepatic lipase activities
Total and hepatic lipase activities were determined by standard techniques (24
, 27)
. Lipoprotein lipase activity was calculated as the difference between total and hepatic lipase activity.
Statistics
Data are means ± SE. Differences between groups were analyzed using a two-tailed unpaired Student’s t test. A Mann-Whitney U test was utilized for comparison of activity measurements, which were not normally distributed.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
shows that the absence of PC-TP expression resulted in a 28% reduction in fasting plasma glucose concentrations (Fig. 1A
) and a 47% decrease in plasma concentrations of FFA (Fig. 1B
). These changes were accompanied by a nonsignificant 29% decrease in plasma insulin concentrations in Pctp–/– mice (Fig. 1C
). Glucose tolerance tests (Fig. 1D
) demonstrated reduced plasma glucose concentrations at each time point, as well as more rapid glucose clearance as evidenced by a reduced area under the curve (AUC, Fig. 1D
, top inset). The increase in glucose clearance was accompanied by reduced plasma insulin concentrations at 120 min after glucose administration (Fig. 1D
, bottom inset). In support of these findings, insulin tolerance tests (Fig. 1E
) demonstrated increased glucose clearance at 15 min after insulin administration, as well as a decrease in the AUC (inset).
|
To gain mechanistic insights into the increase in insulin sensitivity associated with the absence of PC-TP expression, we performed hyperinsulinemic-euglycemic clamp studies (Fig. 2
). These experiments were carried out according to a standardized experimental design (Fig. 2A
), which facilitated the measurements and calculations of the glucose fluxes that are illustrated schematically in Fig. 2B
and plotted in Fig. 2C
. In Pctp–/– mice, the glucose infusion rate was increased by 89%, whereas the rate of glucose uptake into tissues was unchanged. Consequently, hepatic glucose production was markedly decreased by 92%. Total hepatic glucose output was also reduced by 92%, which reflected decreases in hepatic glucose cycling by 92%, in hepatic gluconeogenesis by 85%, and in hepatic glycogenolysis by 96%. During the course of the clamp study, there were no genotype-specific changes in the concentrations of plasma insulin (wild type: 4.7±0.2 ng/ml; Pctp–/–: 4.5±0.2 ng/ml), glucagon (wild type: 31.5 ± 1.8 pM; Pctp–/–: 33.4±2.7 pM), or FFA (wild type: 0.5±0.1 mM; Pctp–/– 0.6±0.1 mM).
|
We next explored whether hepatic insulin signaling was influenced by PC-TP expression (Fig. 3
). Figure 3A
displays hepatic mRNA levels of insulin receptor substrate (IRS) proteins in the livers of Pctp–/– compared with wild-type mice. Whereas hepatic IRS1 mRNA levels remained unchanged, IRS2 mRNA levels were increased by 2-fold. To better ascertain whether PC-TP expression in liver per se might have influenced insulin signaling, we used primary hepatocyte cultures to measure the phosphorylation of Akt in response to a range of insulin concentrations in the media. Figure 3B
shows that there was an increase in Akt phosphorylation at 10–20 nM insulin in PC-TP-deficient hepatocytes, whereas a similar increase occurred at 10-fold higher insulin concentrations in hepatocytes isolated from wild-type mice (Fig. 3B
, inset).
|
Plasma modulators of insulin sensitivity and body fat mass are altered in Pctp–/– mice
We also examined whether modulators of insulin sensitivity may have contributed to the marked reduction in hepatic glucose production that was observed in Pctp–/– mice during the hyperinsulinemic-euglycemic clamp studies (Fig. 4
). Consistent with this possibility, in Pctp–/– compared with wild-type mice, the plasma concentrations of two insulin sensitizers were increased: leptin by 64% (Fig. 4A
) and adiponectin by 39% (Fig. 4B
). By contrast, the concentration of TNF, which reduces hepatic insulin sensitivity (28)
, was decreased by 31% (Fig. 4C
).
|
The substantial changes in plasma concentrations of leptin, adiponectin, and TNF suggested the possibility that body fat composition and distribution might be influenced by PC-TP expression, even though the body weights of the mice did not differ (Supplemental Fig. 1). This was assessed by both NMR and CT scanning (Fig. 5
). Figure 5A
demonstrates that the absence of PC-TP expression was associated with a 47% increase in body fat mass by NMR and a 70% increase by CT. Figure 5B
further demonstrates that NMR and CT measurements were highly correlated for individual mice. Whereas overall fat mass was increased in Pctp–/– mice, the increased plasma concentrations of both leptin and adiponectin suggested that the distribution of fat was also altered. Consistent with this possibility, Pctp–/– mice displayed a 25% reduction in epididymal fat pad weight (Fig. 5C
).
|
To further assess the distribution of increased adipose tissue mass in Pctp–/– mice, cross-sectional images obtained by CT scanning were utilized to reconstruct three-dimensional views of fat tissue that was distributed within and outside the abdominal cavity (Fig. 5D-F
). As depicted by coronal (Fig. 5D
), saggital (Fig. 5E
), and axial (Fig. 5F
) reconstructions, both the intra-abdominal and subcutaneous fat stores were increased markedly in Pctp–/– mice. Nevertheless, volumetric calculations revealed that the relative proportions of intra-abdominal and subcutaneous fat did not differ in Pctp–/– vs. wild-type mice (data not shown).
PC-TP regulates energy substrate utilization
Changes in insulin sensitivity and body composition prompted us to explore substrate utilization for energy production (Fig. 6
). During the 24 h period, time-dependent values of VO2 did not differ between Pctp–/– and wild-type mice (Fig. 6A
). This was the case during the nocturnal period designated period A, when mice typically eat more, and during the remainder of the 24 h (period B), when food consumption is lower. Indeed, the corresponding PRCF analysis (Fig. 6D
) revealed the same EC50 values. By contrast, VCO2 values shown in Fig. 6B
were significantly higher in wild-type mice during period A but not period B. Figure 6E
demonstrates that the EC50 was reduced in Pctp–/– mice. Figure 6C
plots average respiratory quotients during the entire 24 h period, as well as during periods A and B. Over the 24 h period of the experiment, the respiratory quotient was higher in wild-type than Pctp–/– mice, which was affirmed by the reduction in the EC50 value (Fig. 6F
). Figure 6C
also demonstrates that the respiratory quotients were further reduced in Pctp–/– mice during period A compared with the 24 h period. This is consistent with the observed reductions in VCO2 values (Fig. 6B
) during the same period. Figure 6C
also reveals that the average respiratory quotient during period B was also reduced in the absence of PC-TP expression, which occurred despite the absence of changes in mean VO2 and VCO2 values during period B (Fig. 6A, B
). By contrast, there were no changes in respiratory quotients for wild-type mice during period A, period B, or the 24 h time course of the experiment. Supplemental Fig. 2A–C demonstrates that heat production was also reduced in Pctp–/– mice. The EC50 was reduced to a similar degree as for values of VCO2 and respiratory quotients in the absence of PC-TP expression, and this was principally attributable to reductions that took place during the nocturnal period A. Whereas heat production in wild-type mice increased during the nocturnal period A, there was no change in Pctp–/– mice. There was also a trend toward decreased activity in Pctp–/– mice (Supplemental Fig. 2D, E), but this did not achieve statistical significance. Because activity values were not normally distributed, PRCF analysis was not applicable, and median values are presented in Supplemental Fig. 2E.
|
Altered triglyceride distribution in Pctp–/– mice
The decreased VCO2 and reduced respiratory quotients in Pctp–/– mice were indicative of preferential fatty acid utilization for energy production. Figure 7
characterizes the distribution of dietary and hepatic triglycerides. After a 16 h fast, triglyceride concentrations in plasma (Fig. 7A
) and liver (Fig. 7B
) of Pctp–/– mice were decreased by 20 and 43%, respectively. Figure 7C
demonstrates that plasma triglyceride concentrations after intragastric triglyceride administration were significantly reduced up to 1 h, after which values were consistently (albeit not significantly) lower throughout the course of the experiment. Consequently, the overall rate of clearance as reflected by the AUC (Fig. 7C
, inset) was more rapid in Pctp–/– compared with wild-type mice. Figure 7D
displays the tissue distribution of 14C radiolabel 4 h after intragastric administration of triglycerides plus 14C-labeled oleic acid. The distribution was very similar in Pctp–/– compared with wild-type mice, although there was a trend toward enrichment of radioactivity in subcutaneous adipose tissue (P<0.07). Figure 7E
shows that the hepatic triglyceride production rates in mice after a 4 h fast were increased 30% in the absence of PC-TP expression. This occurred without a concomitant increase in plasma triglycerides (wild type: 117±10 mg/dl; Pctp–/–: 125±16 mg/dl). Despite the more rapid clearance of triglycerides from the plasma of Pctp–/– mice, there were no changes in the activities of lipoprotein or hepatic lipase (data not shown).
|
Regulation of fatty acid metabolism by PC-TP in liver and adipose tissue
The expanded fat mass and preferential utilization of fatty acids in Pctp–/– mice suggested that PC-TP may play a role in the regulation of fatty acid transport and metabolism in liver and adipose tissue. Figure 8
A demonstrates a 51% reduction of SREBP1c mRNA levels in the absence of PC-TP expression, as well as marked down-regulation of mRNA levels of the SREBP1c target genes acetyl-CoA carboxylase- (ACC), fatty acid synthase (FAS), and stearoyl CoA desaturase 1 (SCD1) by 41, 66, and 68%, respectively. By contrast, there was up-regulation of the fatty acid transport protein (FATP) 2 but not FATP5 in the livers of Pctp–/– mice. Not shown is that mRNA levels of fatty acid oxidative genes CPT1, CPT2, and HMG-CoA synthase were not influenced by PC-TP expression nor was the plasma concentration of β-hydroxybutyrate. Figure 8B
confirms that nuclear SREBP1c protein levels were reduced in livers of Pctp–/– mice, and Fig. 8C
demonstrates a 62% reduction in rates of fatty acid synthesis in primary hepatocytes cultured from Pctp–/– compared with wild-type mice. Notwithstanding this marked reduction in fatty acid synthesis, Fig. 8D
shows that hepatic fatty acyl-CoA concentrations were not reduced and compositions were not altered.
|
Figure 8E
displays the mRNA levels of selected genes from subcutaneous (left panel) and epididymal fat (right panel). In subcutaneous fat, there was no change in FAS mRNA expression attributable to PC-TP expression. However, in Pctp–/– mice, FAS mRNA was decreased by 80% in epididymal fat. In subcutaneous fat of mice lacking PC-TP, there was a 50% increase in levels of CD36, a protein that mediates fatty acid uptake in adipose tissue. By contrast, in epididymal fat, there was a decrease in FATP1 expression that was not observed in subcutaneous fat. Not shown is that PC-TP expression did not influence mRNA expression of peroxisome proliferator-activated receptor-
(PPAR) or adiponectin.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 11)
, these data support a key function for PC-TP as a regulator of mitochondrial fatty acid utilization.
In the absence of PC-TP expression, we observed a moderate reduction in plasma glucose concentrations that was associated with a trend toward reduced concentrations of insulin and decreased AUC values for both glucose and insulin tolerance tests. However, under conditions of a hyperinsulinemia-euglycemic clamp, there was a profound reduction in hepatic glucose production that was attributable to decreases in glucose cycling, gluconeogenesis, and glycogenolysis. In this setting, the marked hepatic insulin sensitivity exhibited by Pctp–/– mice appeared to be the result of mechanisms that were both intrinsic and extrinsic to the liver. The finding in primary hepatocyte cultures that insulin-mediated Akt phosphorylation occurred at lower insulin concentrations than in hepatocytes cultured from wild-type mice suggests a potential role for PC-TP in modulating the insulin receptor signaling pathway. A potential limitation of this experiment is that primary hepatocytes may retain in vivo characteristics when cultured (29)
.
Adipose tissue secretes proteins that alter the response of the liver to insulin. Leptin and adiponectin are adipokines that enhance the response of the liver to insulin. Leptin, which circulates in plasma at concentrations that reflect body fat mass, increases hepatic glucose fluxes by centrally mediated mechanisms (30)
. Adiponectin promotes insulin sensitivity by direct action on the liver (31)
. Elevations in plasma concentrations of both leptin and adiponectin presumably contributed to enhanced insulin sensitivity in Pctp–/– mice.
The simultaneous elevations in plasma leptin and adiponectin concentrations led us to the observation that body fat mass was increased strikingly in Pctp–/– mice, as verified by NMR spectroscopic and CT imaging techniques. This change in composition occurred despite similar body weights and food intake as in wild-type control mice. Although a relative difference in intra-abdominal fat mass was not detected by CT, Pctp–/– mice exhibited a reduction in epididymal fat mass, suggesting that the proportion of visceral fat was decreased in the absence of PC-TP expression. Since PC-TP is not expressed in white adipose tissue (S. Shrestha and D. Cohen, unpublished results), a direct role in the accumulation of adipose tissue seems unlikely. This assertion is supported by the absence in Pctp–/– mice of a change in the expression levels of PPAR, a key transcription factor that promotes adipose tissue differentiation (32)
. In the absence of PC-TP, sustained expression of adiponectin mRNA expression in white adipose tissue presumably explains the increase in plasma adiponectin that accompanied the increased fat mass. An important limitation of our imaging and gene expression studies is that they did not examine the potential influence of PC-TP on brown fat, which is a key determinant of energy utilization in the mouse.
In the setting of insulin resistance associated with increased adiposity in humans and animal models, elevated plasma concentrations of the proinflammatory cytokine TNF reduce hepatic insulin sensitivity (28)
. Adipose tissue-derived TNF may originate from adipocytes or macrophages that are resident in fat tissue (33)
. Despite the marked increase in adiposity in Pctp–/– mice, TNF concentrations in plasma were substantially reduced. This may also have contributed to the increase in hepatic insulin sensitivity observed in the absence of PC-TP expression. Although macrophage infiltration of adipose tissues was not explored in the current study, the absence of PC-TP expression may have had a direct effect on their numbers or function. In this connection, we have shown that PC-TP is expressed in mouse peritoneal macrophages and that macrophages isolated from Pctp–/– compared with wild-type mice are more susceptible to cholesterol-induced apoptosis (34)
.
Considering the reduced hepatic glucose production in the setting of increased insulin sensitivity, we explored whether Pctp–/– mice might differentially utilize glucose and fatty acids for energy production. Indeed, reduced respiratory quotients in Pctp–/– mice suggested that the absence of PC-TP expression favors the utilization of fatty acids compared with glucose or amino acids for energy utilization. The difference was most pronounced during the nocturnal feeding period, when respiratory quotients in Pctp–/– mice averaged very close to 0.7, indicating the predominant utilization of fatty acids. In support of a switch to fatty acid utilization, fasting plasma and hepatic triglyceride concentrations were decreased in Pctp–/– compared with wild-type mice. Our data also suggested that clearance of plasma triglycerides after an oral triglyceride challenge was more rapid in the absence of PC-TP expression, as was hepatic triglyceride production during fasting. Whereas we cannot exclude the possibility that decreased triglyceride absorption might have accounted for the reduced AUC in Fig. 7C
after oral triglyceride challenge, a significant absorptive defect seemed unlikely because Pctp–/– mice grew normally and consumed chow at the same rate. In keeping with the observation that the activities of lipoprotein and hepatic lipases did not differ in Pctp–/– and wild-type mice, dietary triglycerides were distributed to tissues with similar efficiency irrespective of PC-TP expression. Our data do not explain the molecular mechanism for the increased rates of clearance of dietary and hepatic triglycerides from the plasma in Pctp–/– mice. However, it is possible that there were differences in the compositions of chylomicron and very low density lipoprotein (VLDL) particles attributable to PC-TP expression. In this connection, we previously demonstrated accumulation of small -migrating HDL particles in the plasma of chow-fed Pctp–/– mice (35)
.
In subcutaneous adipose tissue, the tendency toward preferential accumulation of fatty acids (Fig. 7D
), when taken together with increased expression of the fatty acid transporter CD36 (36)
and the absence of FAS down-regulation, may have contributed to the chronic accumulation of fat observed in Pctp–/– mice. By contrast, expression of the fatty acid transporter FATP1 (37)
was decreased in epididymal adipose tissue. This combined with a pronounced decrease in FAS expression may have accounted for the decrease in size of this fat pad in the absence of PC-TP expression.
Despite the increase in hepatic triglyceride secretion rates in Pctp–/– mice in the absence of dietary triglycerides, there was a marked reduction of hepatic fatty acid synthesis. This was most likely explained by the increase in hepatic insulin sensitivity. In support of this possibility, there was an increase in mRNA levels of IRS2, which is a negative regulator of insulin-mediated lipogenesis (38)
. In addition to phosphorylation of the IRS2 protein, insulin promotes its transcriptional up-regulation. SREBP1c, a key transcription factor that regulates lipogenic gene expression in the liver, is regulated by insulin (39)
. Consistent with their tendency toward lower serum insulin levels, Pctp–/– mice exhibited decreased expression of SREBP1c and its lipogenic targets together with decreased fatty acid synthesis (39
, 40)
. Since SREBP1c is also a negative regulator of IRS2, reduced levels of SREBP1c could have accounted for the increased levels of IRS2 mRNA (41)
. Alternatively, the lack of PC-TP expression could have led to an increase in the activity of the transcription factor TFE3, which promotes IRS2 expression and insulin signaling while suppressing lipogenesis (42)
. Leptin also down-regulates hepatic SCD1 expression (43)
, and its expression was increased in mice lacking PC-TP.
The increased utilization of fatty acids by extrahepatic tissues in the setting of down-regulation of fatty acid synthesis in the liver suggests that fatty acid turnover was increased in Pctp–/– mice. In support of this possibility, there was up-regulation of FATP2 mRNA (44)
, as well the lack of a reduction of in hepatic concentration of fatty acyl-CoAs despite the marked down-regulation of SCD1. Therefore, efficient fatty acid uptake and conversion to fatty acyl-CoAs most likely account for the increase in secretion of VLDL triglyceride secretion that was observed in food-deprived Pctp–/– mice (45)
. An expanded adipose compartment relative to muscle mass was sufficient to compensate for reduced hepatic glucose outputs.
Taken together, the current data suggest a role for PC-TP in determining substrate utilization for oxidative metabolism in tissues in which the protein is expressed. In the absence of PC-TP expression, there is preferential utilization of fatty acids for energy production. As a potential mechanism, we (11)
recently demonstrated that PC-TP interacts with the mitochondrial associated protein Them2 (12)
and stimulates its acyl-CoA thioesterase activity in vitro. Intracellularly, this could limit access of fatty acyl-CoA molecules to mitochondria by converting them to FFA, which are not taken up via CPT1. In the absence of PC-TP, more fatty acyl-CoAs would enter mitochondria as preferential substrates for energy production in oxidative tissues. In Pctp–/– mice, the combination of increased mitochondrial fatty acid consumption, decreased synthesis, and increased VLDL triglyceride secretion could explain, at least in part, the observed increase in hepatocellular insulin sensitivity (46)
. Importantly, our studies do not exclude the possibility that PC-TP expression influenced the function of peroxisomes, which are another major site for fatty acid oxidation in the cell.
Chronic adaptation to the absence of PC-TP expression involved an increase in the mass of adipose tissue and a corresponding decrease in lean body mass. In this connection, Kuriyama et al. (47)
demonstrated that decreased fatty acid synthesis in the mouse liver due to a conditional (i.e., short-term) deficiency in sterol cleavage-activating protein (SCAP) is compensated by fatty acid synthesis in adipose tissue. In SCAP-deficient mice, preferential uptake and conversion of glucose to fatty acids in adipocytes were attributable to increased insulin sensitivity in the fat tissue. Under these conditions, there was no net increase in total body fat. By contrast, decreased hepatic fatty acid synthesis in Pctp–/– mice was accompanied by an expansion of body fat mass. Although the hyperinsulinemic-euglycemic clamp studies demonstrated that peripheral glucose uptake was not influenced by PC-TP expression, there may have been a compensatory increase in glucose uptake into adipose tissue in response to preferential fatty acid oxidation in muscles. This could explain the lack of down-regulation of FAS mRNA expression in peripheral fat. By contrast, in epididymal fat pads, which were reduced in the absence of PC-TP expression, there was a marked down-regulation of FAS. In addition to the possibility that more glucose was converted to fatty acids, the up-regulation of CD36 in peripheral and down-regulation of FATP1 in epididymal fat suggested that the distribution of fatty acids in adipose tissue was altered in Pctp–/– mice.
When taken together with our prior observation that a lack of PC-TP expression attenuates atherosclerosis in Apoe–/– mice (6)
, the current finding of augmented hepatic insulin sensitivity in Pctp–/– mice implicates this highly specific lipid binding protein as a potential target for diabetic control and cardiovascular risk reduction. This possibility is supported by a recent genome-wide scan that revealed a coding region PC-TP polymorphism associated with large LDL particle size (7)
and suggested that the wild-type protein participates in the pathogenesis of small dense LDL particles (7)
, which are a manifestation of insulin resistance in humans (48)
.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication January 4, 2008. Accepted for publication February 21, 2008.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
; n=13) and Pctp–/– (•; n=10) mice were deprived of food for 16 h and injected intraperitoneally with 20% D-glucose (2 mg/g body wt). Top inset: AUC; bottom inset: plasma insulin concentrations. E) Insulin tolerance tests: wild-type (
