SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

There is controversy regarding whether fructose in liquid beverages constitutes another dietary ingredient of high caloric density or introduces qualitative changes in energy metabolism that further facilitate the appearance of metabolic diseases. Central to this issue is the elucidation of the molecular mechanism responsible for the metabolic alterations induced by fructose ingestion. Fructose administration (10% wt/vol) in the drinking water of Sprague-Dawley male rats for 14 days induced hyperleptinemia and hepatic leptin resistance. This was caused by impairment of the leptin-signal transduction mediated by both janus-activated kinase-2 and the mitogen-activated protein kinase pathway. The subsequent increase in activity in the liver of the unphosphorylated and active form of the forkhead box O1 nuclear factor, which transrepresses peroxisome proliferator-activated receptor α activity, and a lack of activation of the adenosine monophosphate-activated protein kinase, led to hypertriglyceridemia and hepatic steatosis. These alterations are attributable to two key events: (1) an increase in the amount of suppressor of cytokine signaling-3 protein, which blocks the phosphorylation and activation of janus-activated kinase-2 and Tyr985 on the long form of the leptin receptor; and (2) a common deficit of phosphorylation in serine/threonine residues of key proteins in leptin-signal transduction pathways. The latter is probably produced by the early activation of protein phosphatase 2A, and further sustained by the accumulation in liver tissue of ceramide, an activator of protein phosphatase 2A, due to incomplete oxidation of fatty acids. Conclusion: Our data indicate that fructose ingestion as a liquid solution induces qualitative changes in liver metabolism that lead to metabolic diseases. (HEPATOLOGY 2008.)

Fructose makes up a significant proportion of energy in westernized diets, mainly due to the high intake of fructose-containing beverages. This situation has coincided with the growing prevalence of obesity and metabolic syndrome, recognized risk factors for cardiovascular diseases, over the past two decades.1, 2 Fructose in liquid diets induces hypertriglyceridemia to a greater extent than in solid diets.3 A 10% dietary energy increase in the form of carbohydrates is not sufficient to induce hypertriglyceridemia unless it is consumed as a liquid formula with a high monosaccharide content.4 We have recently reported that the administration of a fructose solution to male rats produces hyperleptinemia and partial impairment of the signal transducer and activator of transcription-3 (STAT-3)-related leptin-signal transduction pathway. These alterations could partially explain the metabolic changes induced by the ingestion of liquid fructose.5

There is intense debate on the subject of whether fructose incorporation as a component of westernized diets, especially in liquid beverages, represents just another dietary ingredient of high caloric density, or whether on top of this, fructose introduces qualitative changes in energy metabolism that further facilitate the appearance of obesity and metabolic syndrome.1, 6, 7 Here we demonstrate that fructose administration as a liquid diet to male rats induces impairment of the leptin-signal transduction mediated by both janus-activated kinase-2 (JAK-2) and the mitogen-activated protein kinase (MAPK) pathway. The lack of JAK-2 activity increases hepatic activity of the unphosphorylated and active form of the forkhead box O1 (FoxO1) nuclear factor. This in turn transrepresses peroxisome proliferator-activated receptor α (PPARα) activity, thus explaining our previous observation of a deficit of PPARα activity in livers of fructose-fed rats.5 The absence of activation of the MAPK pathway explains the lack of phosphorylation at serine (Ser)727 of STAT-3, and the absence of activation of the adenosine monophosphate–activated protein kinase (AMPK), which we have reported.5 These alterations result from two key events: an increase in suppressor of cytokine signaling-3 (SOCS-3) protein that blocks the phosphorylation and activation of JAK-2 and Tyr985 in the long form of the leptin receptor (ObRL); and a common deficit of phosphorylation in serine/threonine (Ser/Thr) residues of key proteins involved in leptin-signal transduction. This may be caused by the increase in liver protein phosphatase 2A (PP2A) activity induced by fructose.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Animals and Experimental Design

Male Sprague-Dawley rats purchased from Harlan Interfauna Ibérica (Barcelona, Spain) were maintained with water and food ad libitum at constant humidity and temperature with a light/dark cycle of 12 hours. The animals were randomly separated into a control group and a fructose-supplemented group (10 rats per group). Fructose was supplied as a 10 % (wt/vol) solution in drinking water for 2 weeks. Control animals received no fructose supplement. Food and fructose solution were removed at 8 AM, and the animals were killed by decapitation under isoflurane anesthesia at 10 AM. Eight additional rats (four control and four fructose-fed rats) were treated identically but the food and fructose solution were not removed before they were killed.

For experiments with 5-aminoimidazole 4-carboxamide 1-β-D ribofuranoside (AICAR), the animals were randomly separated into a control group, a fructose-supplemented group and a fructose-supplemented group treated with AICAR (Toronto Research Chemicals Inc., Ontario, Canada) (four animals per group). Experimental conditions were as described above. AICAR was administered as a subcutaneous dose of 500 mg/kg/day, for the last 3 days of the experiment.

All procedures were conducted in accordance with the guidelines established by the University of Barcelona Bioethics Committee, as stated in Law 5/1995 (21st July) of the autonomous government of Catalonia (Generalitat de Catalunya).

Sample Preparations

Blood and liver tissue samples were collected and stored as described.5 Total and nuclear extracts were isolated using the Helenius method.8 Protein concentrations were determined by the Bradford method.9

Cell Culture

FaO rat hepatoma cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% (vol/vol) fetal bovine serum. At 75% of confluence, the medium was changed to Dulbecco's modified Eagle's medium with 1% (vol/vol) fetal bovine serum, without (control) or with 25 mM fructose added (fructose-treated). Cellular and nuclear extracts were obtained after 6 hours of incubation.

Lipids, Glucose, Insulin, Adiponectin, and Leptin Analysis

Plasma triglycerides, nonesterified fatty acids, glucose, insulin, leptin, and adiponectin concentrations, as well as liver triglyceride content, were measured as described.5

Enzyme Activity Assays

Fatty Acid B-Oxidation Activity.

The activity of hepatic fatty acid β-oxidation was determined as described.10

Activity of 5′-AMPK.

AMPK was assayed in the 6% polyethylene glycol (molecular weight 8000; PEG 8000) fraction by following the incorporation of [32P] adenosine triphosphate into a substrate for AMPK (SAMS peptide; Upstate Biotechnology, Lake Placid, NY).11

Ceramide and Diacylglycerol Hepatic Contents

The content of ceramides in hepatic tissue was determined by the diacylglycerol (DAG) kinase method.12 For DAG determination, the same protocol was used, except for the alkaline hydrolysis at 37°C for 1 hour, which was omitted because it would have removed the DAG.

RNA Preparation and Analysis

Total RNA was isolated by using the Trizol reagent (Invitrogen Biotechnologies). The relative levels of specific messenger RNAs (mRNAs) were assessed by reverse-transcription polymerase chain reaction, as described.13 Adenosyl phosphoribosyl transferase was used as an internal control. The number of cycles, primer sequences, and resulting polymerase chain reaction products are listed in Table 1. The mRNA levels are expressed as ratios of the adenosyl phosphoribosyl transferase mRNA levels.

Table 1. Primers Used for the PCR Reaction
 GenBank numberPrimer SequencesPCR Product (bp)Amplification Cycles
  • *

    As reported by Hsieh et al.44

APRTL04970Forward: 5′-AGCTTCCCGGACTTCCCCATC-3′32923
  Reverse: 5′-GACCACTTTCTGCCCCGGTTC-3′  
c-fosX06769*Forward: 5′-CATCGGCAGAAGGGGCAAAGTAGAG-3′46331
  Reverse: 5′-TGCCGGAAACAAGAAGTCATCAAAG-3′  
FASM76767Forward: 5′-GTCTGCAGCTACCCACCCGTG-3′21420
  Reverse: 5′-CTTCTCCAGGGTGGGGACCAG-3′  
G6pcNM013098Forward: 5′-GATCGCTGACCTCAGGAACGC-3′19720
  Reverse: 5′-AGAGGCACGGAGCTGTTGCTG-3′  
L-PKM11709Forward: 5′-TATGGCGGACACCTTCCTGGA-3′25023
  Reverse: 5′-GCTGAGTGGGGAGGTTGCAAA-3′  
SCD1J02585Forward: 5′-GCTCATCGCTTGTGGAGCCCAC-3′52118
  Reverse: 5′-GGACCCCAGGGAAACCAGGAT-3′  
SOCS-3NM053565Forward: 5′-TTTTCGCTGCAGAGTGACCCC-3′25025
  Reverse: 5′-TGGAGGAGAGAGGTCGGCTCA-3′  

Western Blot Analysis

Thirty microgram aliquots of different protein fractions from rat livers were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Proteins were then transferred to Immobilon polyvinylidene difluoride transfer membranes (Millipore, Bedford, MA), blocked for 1 hour at room temperature with 5% nonfat milk solution in TBS-0.1% Tween-20, and incubated as described.5 Antibodies were obtained from Santa Cruz Technologies, except those for p-AKT, p-STAT-3, p-JAK2, p–extracellular signal-related activated kinase (ERK) 1/2, ERK 1/2, MAPK kinase (MEK1)/ERK activator kinase (MEK2), and p-MEK1/2, which were obtained from Cell Signaling (Danvers, MA).

Electrophoretic Mobility Shift Assays

Electrophoretic mobility shift assays were performed as described,5 using a binding site of carbohydrate response element binding protein (ChREBP) oligonucleotide with the following sequence: 5′-tcctgcatgtgccacaggcgtgtcacc-3′.

Coimmunoprecipitation

Coimmunoprecipitation assays were done with 50 μg of nuclear extracts as described.14, 15

Statistics

The results are expressed as the mean of n values ± standard deviation. Plasma samples were assayed in duplicate. Significant differences were established by the unpaired t test or the one-way analysis of variance test (with analysis a posteriori), using the computer program GraphPad InStat (GraphPad Software V2.03); when the number of animals was too small or the variance was not homogeneous, a nonparametric test was performed (Mann-Whitney or Kruskal-Wallis for comparing two or more than two groups, respectively). The level of statistical significance was set at P < 0.05.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Fructose Induced Hyperleptinemia, Hypertriglyceridemia, Hepatic Steatosis, and a Deficit in P-Ser727-STAT-3.

Fructose-supplemented rats showed increased ingestion of liquids (1.7-fold), with no significant changes in body weight and solid food consumption (Supporting Table 1). Fructose-fed rats were hypertriglyceridemic (1.3-fold) and presented high plasma concentrations of leptin (1.9-fold) and adiponectin (1.7-fold). Liver samples from fructose-fed rats had a higher triglyceride content (1.6-fold) and lower fatty acid β-oxidation activity (0.8-fold) than liver samples from control rats. Livers from fructose-fed rats had higher expression of ChREBP and genes related to fatty acid synthesis and glycolysis than livers from control rats, and showed impaired phosphorylation at Ser727 of STAT-3 and no significant changes in the expression of leptin-regulated genes (Supporting Fig. 1). These results are practically identical to those reported in our previous publication,5 confirming the induction of hyperleptinemia and the deficit in liver STAT-3 Ser phosphorylation in male rats after fructose consumption as a liquid diet.

Fructose Impaired Leptin Activation of AMPK Mediated by the MAPK Pathway.

Leptin, like adiponectin, controls fat catabolism and glucose production in liver tissue partly by increasing AMPK activity.16, 17 We have previously shown that fructose did not affect the phosphorylated, active form of AMPK (P-Thr-AMPK).5 Our present data confirmed our previous results; liver samples from fructose-fed rats showed similar levels of P-Thr-AMPK (data not shown), but also had similar AMPK activity to control (6,636 versus 7,974 cpm/μg protein/minute activity of a pooled sample of six fructose and control livers, respectively), when measured directly by using a SAMS peptide as a substrate.

AMPK phosphorylation and activation is induced by an upstream Ser/Thr kinase, LKB1 kinase.18, 19 LKB1 activation is also dependent on phosphorylation at Ser431 by the MAPK pathway. As leptin activates the MAPK pathway,20 we examined how fructose feeding influenced these activities. Despite having a 1.9-fold higher leptin plasma concentration than controls, fructose-fed rats showed no changes in Ser-phosphorylated and active forms of LKB1, mitogen activated kinase, or ERK (Fig. 1A). Although the expression of LKB (phosphorylated and total) was increased in livers of fructose-fed rats, the ratio between the phosphorylated and total forms of LKB was unchanged (0.43±0.23 versus 0.47±0.06, n = 6, for control and fructose-fed rats, respectively), indicating that fructose did not affect the activity of the enzyme. The expression of c-fos, a gene directly controlled by the activation of the MAPK pathway,20, 21 was also unchanged in liver samples from fructose-fed rats (Fig. 1B). Thus, the absence of increased AMPK activity in livers of fructose-fed rats can be attributed to impaired leptin activation of the MAPK pathway.

thumbnail image

Figure 1. Leptin signal transduction pathways in livers of fructose-supplemented rats. (A) Western blots of the total and active phosphorylated forms of MEK, ERK, and LKB1 in liver samples from control and fructose-supplemented rats. (B) Relative levels of c-fos mRNA in hepatic samples from control (empty bar), and fructose-supplemented rats (black bar). Each bar represents the mean ± standard deviation (SD) of values from six animals. A representative autoradiography at the top of the figure shows the bands corresponding to c-fos mRNA and that of the aprt gene, used as an internal control in the polymerase chain reaction (PCR) reaction to normalize the results, from liver samples of three animals from each treatment group. (C) Western blot of the active, phosphorylated form of JAK-2 in liver samples from control and fructose-supplemented rats. (D) Western blot of SOCS-3 protein in liver samples from control and fructose-supplemented rats. On the right of the figure, a bar plot shows the levels of SOCS-3 protein, expressed as the mean ± SD of values from six animals, in hepatic samples from control (empty bar) and fructose-supplemented (black bar) rats. *P < 0.05. (E) Western blots of the total and active, phosphorylated forms of Akt in liver samples from control and fructose-supplemented rats. For Western blots, the amount of protein loaded was confirmed by the Bradford method, and the uniformity of protein loading in each lane was assessed by determining the signal of β-actin, as a control-loading protein.

Download figure to PowerPoint

Leptin binding to the ObRL induces phosphorylation and activation of JAK-2, and phosphorylation of two tyrosine (Tyr) residues, Tyr985 and Tyr1138,20, 21 at the cytosolic tail of the receptor. While activation of the MAPK pathway by leptin can be directly mediated either by JAK-2 or by P-Tyr985-ObRL, initial activation of the STAT-3 pathway by phosphorylation at Tyr705 and dimerization is mediated by P-Tyr1138-ObRL.21 The expression of ObRL was increased in livers of fructose-fed rats (3.0-fold), as was the amount of P-Tyr1138-ObRL (1.9-fold) (Fig. 2), closely matching the increases in plasma leptin and liver Tyr705-STAT3 in these animals. In contrast, fructose supplementation had no effect on hepatic P-Tyr985-ObRL (Fig. 2), or P-JAK-2 (Fig. 1C), indicating that the lack of response of the MAPK pathway to the increased leptin concentration was due to a direct impairment of the leptin signal at the ObRL level. SOCS-3 protein acts by blocking the phosphorylation of JAK-2 and ObRL at Tyr985.22 Although levels of SOCS-3 mRNA were unmodified in livers of fructose-fed rats (see Supporting Fig. 1, and Ref.5), SOCS-3 protein increased 2.8-fold (Fig. 1D).

thumbnail image

Figure 2. ObRL protein levels in liver samples from fructose-supplemented rats. Bar plots showing the levels of P-Tyr985-ObRL, P-Tyr1138-ObRL, and total ObRL protein in hepatic samples from control (empty bar) and fructose-supplemented (black bar) rats. Each bar represents the mean ± standard deviation (SD) of values from six animals. On the right of each figure, a representative Western blot shows the ObRL bands corresponding to three different control and fructose-fed rats. The amount of protein loaded was confirmed by the Bradford method, and the uniformity of protein loading in each lane was assessed by determining the signal of β-actin, as a control-loading protein.

Download figure to PowerPoint

The Reduction of PPARα Activity Induced by Fructose Is Related to the Impairment of JAK-2 Stimulation.

Fructose reduces hepatic PPARα activity.5 Qu et al.23 have shown that in hamsters a high-fructose diet in solid form (60% fructose) increases the production of the active nuclear form of the FoxO1 nuclear factor. The activation of FoxO1 was related to decreased plasma triglyceride clearance in these animals, and was reversed by PPARα activation after fenofibrate administration to fructose-fed hamsters. The mechanism underlying those effects was the transrepressing activity of PPARα on FoxO1 transcriptional activity upon physical interaction between the activated forms of the two receptors.23 As transrepressing mechanisms can be bidirectional,14, 15 we sought to determine whether a similar situation was present in the livers of rats fed a 10% fructose liquid solution. Indeed, as shown in Fig. 3A, coimmunoprecipitation experiments indicate that fructose supplementation increased FoxO1-PPARα complexes, in comparison to control animals, pointing in our case to a transrepressing effect of FoxO1 on PPARα activity, which is partially responsible for the decreased activity of fatty acid oxidation observed in these animals. Fructose supplementation did not affect the level of Ser-phosphorylated Akt (Fig. 1E), reflecting the lack of activation of JAK-2 by leptin in these animals.20, 22, 24 When phosphorylated at a Ser residue by the protein kinase B or Akt kinase, FoxO1 is inactivated and migrates from the nucleus to the cytosolic compartment.25 Thus the excess nuclear activity of FoxO1 in the livers of fructose-fed animals is again related to a deficit in the leptin-signal pathway through JAK-2 activation.

thumbnail image

Figure 3. FoxO1-PPARα coimmunoprecipitation in livers of fructose-supplemented rats. (A) Liver nuclear extracts from three control (CTR) and three fructose-fed (FRC) rats were subjected to immunoprecipitation using anti-FoxO1 antibody coupled to protein-A agarose beads. Immunoprecipitates were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotted with an anti-PPARα antibody. The remaining animals from the control and fructose-fed groups yielded similar results. (B) Pooled samples of liver nuclear extracts from CTR, FRC, and FRC rats treated with AICAR (FRC + AICAR) during the last three days of fructose supplementation (500 mg AICAR/kg/day, subcutaneously [sc]) were subjected to immunoprecipitation using anti-FoxO1 antibody coupled to protein-A agarose beads. Immunoprecipitates were subjected to SDS-PAGE and immunoblotted with an anti-PPARα antibody.

Download figure to PowerPoint

The Lack of Stimulation of AMPK Activity Is Not the Main Mechanism Responsible for Decreased Fatty Acid Oxidation in Livers of Fructose-fed Rats.

Feeding fructose as a liquid diet to male rats reduces hepatic fatty acid oxidation, contributing to the fatty liver and hypertriglyceridemia in these animals; the mechanisms involved in this process are a lack of AMPK activity stimulation and decreased transactivating activity of PPARα,5, 13 both situations related to defective leptin signaling in liver. As it is not clear whether the two mechanisms are connected, we sought to directly activate AMPK by pharmacological means, thus bypassing the molecular defect in leptin signal transduction introduced by fructose feeding, to determine whether AMPK activity was sufficient to restore the deficit in hepatic fatty acid oxidation produced by fructose. To this end, to a group of fructose-fed male rats, we administered AICAR, an AMP analog that is known to directly activate AMPK.16, 26 Despite increasing hepatic AMPK activity by 1.3-fold, AICAR administration did not reverse the reduction in hepatic fatty acid oxidation or the hypertriglyceridemia produced by fructose feeding (Table 2). Furthermore, AICAR treatment did not modify the increased interaction between liver FoxO1 and PPARα shown by fructose-fed animals (Fig. 3B). Thus, AMPK activation seems to be unrelated to the PPARα system and is not sufficient to reverse the deficit in hepatic fatty acid oxidation induced by fructose feeding.

Table 2. AICAR Effects of Body Weight, Food and Liquid Ingestion, Plasma Analytes, Hepatic Triglyceride Content, and AMPK Activity
 CTRFRCFRC+AICAR
  • *

    P < 0.05 versus CTR. †Values obtained by assaying a pool of six different liver samples for each group.

  • Abbreviation: AUC, area under the curve.

AUC body weight (g/14 days/2 rats)2,690 ± 2032,740 ± 1092,876 ± 125
AUC ingested liquid (mL/14 days/2 rats)780 ± 51,090 ± 45* (×1.4)1,074 ± 106
AUC consumed diet (g/14 days/2 rats)557 ± 26513 ± 4525 ± 46
Fatty acid β-oxidation (nmol/minute/mg of liver protein)0.7 ± 0.050.5 ± 0.03* (×0.7)0.5 ± 0.07
Liver triglycerides (mg/g liver)4.8 ± 0.98.8 ± 2.7* (×1.8)6.1 ± 2.2
Plasma triglycerides (mg/dL)51 ± 884 ± 14* (×1.6)79 ± 12
Plasma glucose (mg/dL)159 ± 6169 ± 7172 ± 7
Plasma insulin (ng/mL)1.1 ± 0.11.0 ± 0.71.1 ± 0.6
Plasma leptin (ng/mL)2.7 ± 0.34.0 ± 0.6* (×1.5)4.7 ± 1.6
AMPK activity (cpm/μg protein/minute)†9,6769,78612,202 (×1.3)

Ceramide and the Active Subunit of PP2A Increased in Livers of Fructose-fed Rats.

The fatty acid palmitate is the precursor of de novo ceramide synthesis.27 Given the reduction in liver fatty acid catabolism introduced by fructose feeding, we determined whether the excess of unoxidized fatty acids was diverted to the synthesis of not only triglycerides, but other lipid species. Indeed, as shown in Fig. 4, despite showing similar amounts of DAG as livers from control rats (Fig. 4A), livers of fructose-fed rats presented a marked accumulation of ceramide (Fig. 4B). Furthermore, AICAR treatment did not affect the accumulation of ceramide in livers of fructose fed rats (data not shown).

thumbnail image

Figure 4. Diacylglycerol (DAG) and ceramide levels in livers of fructose-supplemented rats. Lipid extracts were prepared and assayed for DAG and ceramide as detailed in the Materials and Methods. (A) Phosphorimages of DAG and (B) ceramide present in a pool of liver samples from six control and six fructose-supplemented rats are shown.

Download figure to PowerPoint

Ceramides are known activators of PP2A28; fructose and other carbohydrates, thorough a common metabolite, xylulose-5-phosphate, are also potent PP2A activators.29 By using an antibody to the catalytic subunit of PP2A, we did not detect a clear increase of PP2A activity in livers of fructose-fed rats (data not shown), although the amount of ChREBP protein and the expression of liver pyruvate kinase (L-PK; see Supporting Fig. 1), a gene under direct transcriptional control of ChREBP,30 were markedly increased in liver samples from fructose-fed rats (2.2-fold and 1.9-fold, respectively).

As it is clearly established that the expression and transcriptional activity of ChREBP are directly controlled by the activity of PP2A,29, 30 we could have missed the increase in PP2A activity due to the absence of appropriate amounts of the PP2A activator, xylulose-5-phosphate, in the livers of our 2-hour fasted fructose-fed rats. To avoid this possibility, we determined PP2A activity and ChREBP binding activity in livers of unfasted fructose-fed rats and in FaO rat hepatoma cells incubated with 25 mM fructose for 6 hours. As can be seen in Fig. 5, fructose induced an increase in the active, catalytic subunit of PP2A (1.57-fold) in livers of unfasted fructose-fed rats (Fig. 5A). Incubation of liver nuclear extracts with a ChREBP response element oligonucleotide produced two bands that were specific for ChREBP, as they were competed in the presence of excess unlabeled oligonucleotide. Further, bands I and II contained ChREBP protein, as they were supershifted when nuclear extracts were preincubated with a specific ChREBP antibody (Fig. 5B). As can be seen in Fig. 5C, the intensity of band II was clearly increased (1.4-fold) in nuclear extracts from fructose-supplemented rats. FaO rat hepatoma cells incubated in the presence of 25 mM fructose for 6 hours showed an even greater increase in the expression of the active, catalytic subunit of PP2A (1.8-fold; Fig. 6A). Nuclear extracts from control FaO cells incubated with a ChREBP response element oligonucleotide also produced two bands (bands Ib and II) that were specific for ChREBP, as they were competed in the presence of excess unlabeled oligonucleotide (Fig. 6B). Incubation with fructose changed the pattern of the shifted bands, with almost total disappearance of band II, and the appearance of a new one, band Ia (Fig. 6C). Bands Ia and Ib contained ChREBP protein, as they were supershifted when nuclear extracts from FaO cells supplemented with fructose were preincubated with a specific ChREBP antibody. Overall, these results indicate a change in ChREBP binding activity related to PP2A activation.

thumbnail image

Figure 5. PP2A activation and ChREBP binding in rat liver of fructose-supplemented rats. (A) Western blot of the active, catalytic subunit of PP2A in liver samples from control and fructose-supplemented rats. On the left of the figure, a bar plot shows the levels of PP2A protein, expressed as the mean ± SD value from three control (empty bar) and three fructose-supplemented (black bar) samples. The amount of protein loaded was confirmed by the Bradford method, and the uniformity of protein loading in each lane was assessed by determining the signal of β-actin, as a control-loading protein.*P < 0.05. f.i., fold induction. (B) Electrophoretic mobility shift assay (EMSA) assay showing the binding of rat hepatic nuclear extracts (NEs) to a ChREBP response element oligonucleotide, forming two specific bands, as they disappeared in the presence of an excess of unlabeled oligonucleotide. Bands I and II contained ChREBP protein, as they were supershifted in the presence of a specific ChREBP antibody (IC, immunocomplex). Oct-1 antibody was used to demonstrate that the changes in the shifted bands were not due to unspecific interference by the presence of immunoglobulin proteins in the incubation medium. (C) EMSA assay showing the binding to a ChREBP oligonucleotide of control and fructose-supplemented rat hepatic NEs.

Download figure to PowerPoint

thumbnail image

Figure 6. PP2A activation and ChREBP binding in FaO rat hepatoma cells incubated with 25 mM fructose for 6 hours. (A) Western blot of the active, catalytic subunit of PP2A in samples from control and fructose-supplemented FaO cells. On the left of the figure, a bar plot shows the levels of PP2A protein, expressed as the mean ± standard deviation (SD) value from three control (empty bar) and three fructose-supplemented (black bar) samples. The amount of protein loaded was confirmed by the Bradford method, and the uniformity of protein loading in each lane was assessed by determining the signal of β-actin, as a control-loading protein. **P < 0.01. f.i., fold induction. (B) EMSA assay showing the binding of control FaO cell nuclear extracts (NEs) to a ChREBP response element oligonucleotide, forming two specific bands (bands Ib and II), as they disappeared in the presence of an excess of unlabeled oligonucleotide. Bands Ia and II were formed when NEs from FaO cells supplemented with fructose were used. Bands Ia and II contained ChREBP protein, as they were supershifted in the presence of a specific ChREBP antibody (IC, immunocomplex). Oct-1 antibody was used to demonstrate that the changes in the shifted bands were not due to unspecific interference by the presence of immunoglobulin proteins in the incubation medium. (C) EMSA assay showing the binding to a ChREBP oligonucleotide of control and fructose-supplemented FaO cell NEs.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We have previously shown that fructose administration as a liquid solution to rats induces hyperleptinemia and hepatic leptin resistance.5 Furthermore, the key difference between ingesting similar amounts of fructose and glucose as a liquid solution was the fact that, while both sugars similarly stimulated hepatic lipogenesis, only the former significantly reduced fatty-acid oxidation in the liver. We now describe a plausible molecular mechanism to explain these metabolic disturbances.

The reduction in hepatic fatty acid oxidation in fructose-fed rats is attributable to incomplete activation by leptin of two proteins involved in the control of fatty acid catabolism: the enzyme AMPK31 and the nuclear receptor PPARα.32 Nevertheless, the data we obtained by using AICAR, a direct activator of AMPK,16 indicates that activation of this enzyme does not completely restore fatty acid oxidation in the fructose-fed rat liver, nor does it directly stimulate the activity of PPARα. It is unclear whether there is a direct molecular relationship between the two systems.31, 33–35 Our data argue against this possibility, indicating that the main effect of AMPK on hepatic fatty acid oxidation may be due to the decrease in the concentration of malonyl-coenzyme A (CoA), as a result of the AMPK-dependent phosphorylation and inhibition of acetyl-CoA carboxylase.31 Given that the inhibitory constant (Ki) for malonyl-CoA of liver carnitine palmitoyltransferase I is very low,36 this mechanism may require proper PPARα activity to be relevant.

In any case, the lack of activation of AMPK and PPARα in the livers of fructose-fed rats is due to a generalized deficit of phosphorylation at Ser/Thr residues in key proteins involved in the signal transduction pathway of leptin.20 Leptin activates AMPK through a phosphorylation cascade initiated by the binding of leptin to ObRL. This binding phosphorylates Tyr residues in the same receptor (Tyr985 and Tyr1138) and in JAK-2.37 These changes activate a cascade of Ser/Thr phosphorylations mediated by the MAPK pathway, ultimately responsible for the activation of AMPK by the Ser/Thr kinase LKB1.19 All of these sequential Ser/Thr phosphorylations are lacking in the livers of fructose-fed rats, although they are hyperleptinemic. The lack of activation of the MAPK pathway may also explain the lack of phosphorylation of STAT-3 at Ser727.38 Moreover, by activating JAK-2, leptin can also Ser-phosphorylate and activate Akt.20, 22, 24 In turn, Akt activity can Ser-phosphorylate and inactivate FoxO1, enabling its migration from the cell nucleus to the cytosol.25 Again, defective leptin signaling through the JAK-2/Akt pathway in livers of fructose-fed rats leads to the transrepression of PPARα activity by its physical sequestration by the activated, unphosphorylated form of FoxO1.

The knowledge gathered about the metabolism of carbohydrates points to the Ser/Thr phosphatase PP2A as the main culprit for the above-mentioned deficit in Ser/Thr phosphorylation. The metabolism of fructose produces xylulose-5-phosphate, which activates PP2A.30 Although we did not detect an increase in hepatic PP2A activity in fructose-fed rats, this may have been attributable to their fasted state. By examining hepatic samples from unfasted fructose-fed rats and FaO rat hepatoma cells incubated with 25 mM of fructose, we showed that fructose indeed increased the expression of the active, catalytic subunit of PP2A and modified ChREBP binding activity. The finding that livers of fructose-fed rats showed a marked increase in the expression and activity of the transcription factor ChREBP, whose activation depends directly on PP2A,29, 30 indicates that PP2A activity may have increased early in the livers of fasted animals. The finding that ceramide, another activator of PP2A,28 accumulates in livers of fructose-fed rats also supports this assumption. Furthermore, saturated fatty acids, which may accumulate in livers of fructose-fed rats due to the increased synthesis and reduced oxidation, activate PP2A.39 PP2A activity has also been implicated in repression of PPARα gene expression,40 and inhibition of Akt and AMPK activities.39–41 All these effects could increase the hepatic metabolic derangements induced by fructose. The fact that fructose is mainly metabolized in liver, while glucose is distributed to and metabolized in other body tissues, can explain why, despite also being metabolized to xylulose-5-phosphate,30 glucose does not induce hyperleptinemia and hepatic leptin resistance when ingested at an equivalent amount of energy as fructose.5

Nevertheless, the activation of the Ser/Thr kinase PP2A does not explain the deficit of Tyr phosphorylation of JAK-2 and at Tyr985 of the ObRL, as seen in the livers of the hyperleptinemic fructose-fed rats. This lack of phosphorylation may be due to the increase in SOCS-3 proteins. These SOCS-3 proteins, whose expression is controlled by leptin through the activation of the STAT-3 pathway, act by blocking the Tyr phosphorylation of JAK-2 and ObRL at Tyr985, while sparing the phosphorylation at Tyr1138, which is responsible for the initial activation and dimerization of STAT-3 by phosphorylation at its Tyr705.22 This explains the closely matched values for the increases in plasma leptin (1.9-fold), P-Tyr1138-ObRL (1.9-fold), and P-Tyr705-STAT-3 (1.9-fold) observed in the livers of fructose-fed rats.

In summary, we propose that, by initially increasing the activity of the Ser/Thr phosphatase PP2A, fructose reduces the activity of the leptin signaling pathways, thus leading to a deficit in hepatic PPARα activity and fatty acid oxidation. The organism reacts in a similar way to that described for insulin resistance, by inducing hyperleptinemia to maintain the activity of the system. The overstimulation of the only functional leptin signaling pathway in the liver, that mediated by P-Tyr1138-ObRL and P-Tyr705-STAT-3, finally builds up enough SOCS-3 protein to further block the leptin signaling mediated by the JAK-2 and MAPK pathways. This blockage establishes a self-perpetuating loop that maintains and enhances the metabolic disturbances produced by fructose (Fig. 7). Given that the metabolic handling of fructose is similar in rats and humans,42 and the fact that fructose administration as liquid solution to healthy male volunteers results in hyperleptinemia,43 it is plausible that similar mechanisms may operate in humans, following excessive consumption of beverages containing fructose.

thumbnail image

Figure 7. Proposed effects of fructose ingestion on the liver leptin signal-transduction pathway. The increased burden of fructose in livers of fructose-supplemented rats activates protein phosphatase 2A (PP2A) by greatly increasing intrahepatocyte concentrations of xylulose-5-phosphate (X-5-P). PP2A dephosphorylates Ser/Thr residues on key enzymes participating in the transduction of leptin signaling, including janus-activated kinase-2 (JAK-2) and mitogen-activated protein kinase (MAPK) pathways, thus finally producing a deficit in fatty acid oxidation in the livers of fructose-fed rats by lack of activation of AMP-activated protein kinase (AMPK) and inhibition of peroxisome proliferator-activated receptor α (PPARα). The organism reacts against this state of partial liver leptin resistance by inducing hyperleptinemia. The overstimulation of the only leptin signaling pathway working in the liver, that mediated by P-Tyr1138-ObRL and P-Tyr705-STAT-3, finally builds up enough SOCS-3 protein to further block the leptin signaling mediated by the JAK-2 and MAPK pathways, in this way establishing an autoperpetuating loop that maintains and enhances the metabolic disturbances produced by fructose. L-PK, liver-pyruvate kinase; ChREBP, carbohydrate response element binding protein; STAT, signal transducer and activator of transcription; SOCS, suppressor of cytokine signaling; ObRL, long form of the leptin receptor.

Download figure to PowerPoint

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We have been nominated as a Consolidated Research Group by the Generalitat de Catalunya (SGR05-00833), with no financial aid.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
HEP_22523_SupportingData.doc35KSupporting Data
HEP_22523_SupportingFigure1.tif1198KSupporting figure

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.