If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
* This work was supported in part by INSERM, the Association pour la Recherche sur le Cancer, by European Community Grants QLG1-CT-1999-00674, “Eurodiabetesgenes,” and QLK3-CT-2000-01038, and by Aventis (Frankfurt, Germany). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ‡ Both authors contributed equally to this work. § Recipient of a doctoral fellowship from the Ministère de l'Enseignement Supérieur et de la Recherche (France). ∥ Supported by European Community Grant QLG1-CT-1999-00674. ** Supported by Fondation pour la Recherche Médicale. ‡‡ Supported by Association pour la Recherche Contre le Cancer.
Phosphatidylinositol 3-kinase signaling regulates the expression of several genes involved in lipid and glucose homeostasis; deregulation of these genes may contribute to insulin resistance and progression toward type 2 diabetes. By employing RNA arbitrarily primed-PCR to search for novel phosphatidylinositol 3-kinase-regulated genes in response to insulin in isolated rat adipocytes, we identified fatty aldehyde dehydrogenase (FALDH), a key component of the detoxification pathway of aldehydes arising from lipid peroxidation events. Among these latter events are oxidative stresses associated with insulin resistance and diabetes. Upon insulin injection, FALDH mRNA expression increased in rat liver and white adipose tissue and was impaired in two models of insulin-resistant mice, db/db and high fat diet mice. FALDH mRNA levels were 4-fold decreased in streptozotocin-treated rats, suggesting that FALDH deregulation occurs both in hyperinsulinemic insulin-resistant state and hypoinsulinemic type 1 diabetes models. Moreover, insulin treatment increases FALDH activity in hepatocytes, and expression of FALDH was augmented during adipocyte differentiation. Considering the detoxifying role of FALDH, its deregulation in insulin-resistant and type 1 diabetic models may contribute to the lipid-derived oxidative stress. To assess the role of FALDH in the detoxification of oxidized lipid species, we evaluated the production of reactive oxygen species in normal versus FALDH-overexpressing adipocytes. Ectopic expression of FALDH significantly decreased reactive oxygen species production in cells treated by 4-hydroxynonenal, the major lipid peroxidation product, suggesting that FALDH protects against oxidative stress associated with lipid peroxidation. Taken together, our observations illustrate the importance of FALDH in insulin action and its deregulation in states associated with altered insulin signaling.
is a key component of the intracellular insulin signaling machinery. PI3K activation occurs after binding of the Src homology 2 domains of its p85 regulatory subunit to specific tyrosine-phosphorylated sites of the insulin receptor substrates (
). By activating these kinases, PI3K generates several insulin-dependent actions on metabolism such as glucose transport, glycogen synthesis, glycolysis, and lipogenesis. Moreover, PI3K acts by modulating the expression of a number of genes involved in lipid and glucose homeostasis, including phosphoenolpyruvate carboxykinase (
), two enzymes that control key steps of gluconeogenesis in liver and the expression of which is repressed by insulin via activation of PI3K. In contrast, insulin up-regulates hexokinase-2 in skeletal muscle and glucokinase in liver in a PI3K-dependent manner (
). Finally, recent studies suggest that PI3K up-regulates the expression of the glucose transporter Glut-4 in muscle and in brown adipose tissue and of p85α in muscle, suggesting a positive feedback effect of PI3K on its own expression (
) have described a decrease in expression and/or activation of PI3K in response to insulin in different insulin-resistant rodent models as well as in patients with type 2 diabetes. We hypothesized that defects in the regulation of gene expression controlled by PI3K may participate in the pathogenesis of type 2 diabetes and insulin resistance.
To identify genes involved in glucose homeostasis, which are specifically controlled by PI3K in response to insulin, we performed RNA arbitrarily primed (RAP)-PCR on freshly isolated rat adipocytes in the presence or absence of insulin and wortmannin, a pharmacological inhibitor of PI3K. By using this approach, we isolated several clones corresponding to both previously identified and novel genes. One RAP-PCR product corresponds to fatty aldehyde dehydrogenase (FALDH), a member of the aldehyde dehydrogenase family that oxidizes aliphatic and aromatic aldehydes to the corresponding carboxylic acids (
). In the present work, we characterized the regulation of FALDH gene expression by insulin, and we addressed the role of FALDH in insulin action in normal, insulin-resistant, and type 1 diabetes conditions.
Materials—Recombinant human insulin was from Novo-Nordisk (Copenhagen, Denmark). Recombinant human epidermal growth factor was from Strathmann Biotec AG (Hamburg, Germany). LY294002 was from Calbiochem. Wortmannin and methylglyoxal (MGO) were from Sigma. Chloromethyl-2′,7′-dichlorofluorescein diacetate (CM-DCFDA) was from Molecular Probes (Eugene, OR). 4-Hydroxynonenal (4-HNE) was from Merck.
Antibodies to phospho-Ser-473 PKB were from New England Biolabs (Beverly, MA), and anti-Myc was from Santa Cruz Biotechnology (Santa Cruz, CA). Secondary anti-mouse or anti-rabbit antibodies conjugated to horseradish peroxidase were purchased from Jackson Immuno-Research (Copenhagen, Denmark).
Animal Treatment and Protocols—Care of animals was performed in accordance with institutional guidelines. Male Wistar rats (150-200 g), 8-10 weeks of age, were purchased from Janvier Laboratory (Laval-LeGenest, France). Male C57BL/6J and db/db mice were obtained at 6 weeks of age from Janvier Laboratory. Animals were maintained in a temperature-controlled facility (22 °C) on a 12-h light/dark cycle and were given free access to food (standard laboratory chow diet from UAR, Epinay-S/Orge, France) unless otherwise indicated. In some experiments, rats or mice were deprived of food for 20 h. Fasted animals were given an intraperitoneal injection with insulin at 1 IU/kg (Actrapid 100 IU/ml, Novo-Nordisk, Denmark) for 6 h or refed with chow diet for 8 h. Animals were euthanized by cervical dislocation, and tissues were rapidly harvested, weighed, and processed for preparation of total RNA or cell culture. High fat diet male C57BL/6J mice were generated by giving high fat and high sucrose diets (D12327, Research Diet, New Brunswick, NJ) over 15 weeks. To generate type 1 diabetic animals, Sprague-Dawley rats were given an intraperitoneal injection of streptozotocin (65-70 mg/kg) as described previously (
Adipose Tissue Culture—Epididymal fat pads from Wistar rats were dissected, minced finely, and digested with Liberase Blendzyme 3 (Roche Applied Science) by shaking at 37 °C for 45 min. Isolated adipocytes were separated from stromal-vascular cells (S-V) by filtration through a 30-μm nylon mesh. The cells were then washed twice with Krebs-Ringer bicarbonate Hepes pH 7.4 (KRBH) buffer and processed for RNA preparation. For primary adipocyte culture, isolated adipocytes were placed in Falcon 2059 tubes in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% (v/v) fetal calf serum, 50 units/ml penicillin, 50 μg/ml streptomycin for primary culture. The cells were incubated at 37 °C, 5% CO2. Cells were depleted overnight in DMEM supplemented with 5% (w/v) bovine serum albumin (BSA) before insulin treatment. The samples were collected, and total RNA was isolated and was analyzed by RT-PCR. For adenoviral infection, isolated adipocytes in culture were infected for 12 h, and the medium was then replaced for 12 h before starvation.
Isolation and Culturing of Hepatocytes—Hepatocytes were isolated from male Wistar rats by collagenase dissociation of the liver as described by Fehlmann et al. (
). Freshly isolated hepatocytes were incubated in Krebs-Ringer bicarbonate (KRb) buffer, pH 7.4, containing defatted BSA (10 mg/ml), gentamycin (0.05 mg/ml) and were gassed with a mixture of 5% CO2, 95% O2. Primary cultures of hepatocytes were performed essentially as described by Morin et al. (
). Cells were plated at a final concentration of 106 cells/ml in Waymouth's medium supplemented with 10% (w/v) fetal calf serum. After 4 h at 37 °C, the medium was replaced with serum-free Waymouth's medium containing defatted BSA (2 mg/ml), penicillin, and streptomycin, and cells were incubated overnight before insulin treatment. For adenoviral infection, hepatocytes were infected with adenovirus for 12 h, and culture medium was replaced, and cells were incubated for 24 h at 37 °C.
RAP-PCR—RAP-PCR was performed with the RAP-PCR kit (Stratagene, Amsterdam, The Netherlands), according to the method of Welsh et al. (
). Isolated adipocytes were treated or not with insulin (10-7m) or cotreated with insulin and wortmannin (100 nm) for 6 h. After extraction of total RNA, poly(A+) RNA was purified using a Poly(A) Tract mRNA Isolation System (Promega, Charbonnières, France). The RT reaction was performed using 100 ng of RNA poly(A+) and different arbitrary primers (1.25 μm) for 1 h at 37 °C by adding to the First Strand Buffer 125 μm of each dNTP, 40 units of RNase (Stratagene), and 25 units of Moloney murine leukemia virus-reverse transcriptase (Stratagene). The reaction was stopped by heating to 90 °C for 5 min.
To perform RAP-PCRs, 1× Taq Reaction Buffer, 1 μm arbitrary primers, 50 μm of each dNTP, 1 unit of Taq polymerase, and 0.2 μCi/ml of [α-33P]ATP were added to 1 μl of the cDNA. The PCR was started by incubation at 37 °C for 5 min and 72 °C for 5 min, followed by 40 cycles at 94 °C for 1 min, 60 °C for 2 min, and 72 °C for 2 min.
To visualize the RAP-PCR products, 5 μl from each reaction was mixed with 10 μl of stop buffer containing 80% (v/v) formamide, 50 mm Tris-HCl, pH 8.8, 1 mm EDTA, 0.1% (w/v) xylene cyanol, 0.1% (w/v) bromphenol blue, and heated at 80 °C for 2 min. 4 μl of each reaction was loaded on a 4% acrylamide, 7 m urea gel. Electrophoresis was performed at 55 watts in 1× TBE buffer, and the gel was autoradiographed. The RAP-PCR gel and the autoradiogram were aligned, and individual bands representing differentially expressed products were cut and removed from the gel. Each isolated band was incubated in 70 μl of elution buffer at 60 °C for 1 h and at room temperature for 12 h. Eluted samples were collected and reamplified by PCR with the same primer used for the RAP-PCR. The RAP-PCR products were then sub-cloned in pCR 2.1 TOPO (Invitrogen) and sequenced using M13 reverse and forward primers.
FALDH Assay—Hepatocytes were washed with PBS and homogenized in 2 ml of homogenization buffer (25 mm Tris-HCl, pH 8.0, 250 mm sucrose) with a Teflon glass Potter homogenizer. Final protein concentration was assayed by bicinchoninic acid technique (BCA protein assay kit, Interchim, Montluçon, France). FALDH assay was performed as described previously by Kelson et al. (
) using 100 μg of crude homogenate with a mixture containing 193 μl of deionized water, 100 μl of 200 mm glycine-NaOH buffer, pH 9.5, 0.4% v/v Triton X-100, 40 μl of 100 mm pyrazole, 20 μl of 10 mg/ml bovine serum albumin (fatty acid free), 24 μl of 25 mm NAD+. Each sample was incubated at 37 °C, and the reaction was initiated by adding 3 μl of 160 μm dodecanol. FALDH activity was monitored by measuring the increase in absorbance at 340 nm at different times. Control reactions were run for each sample by substituting 3 μl of 100% v/v ethanol for the aldehyde solution and running the reaction as above. FALDH-dependent enzyme activity was calculated by subtracting the absorbance measured in the absence of aldehyde from that seen in its presence and dividing by 60 min to express the final result as units/min.
Adipocyte Differentiation—3T3-L1 preadipocytes (American Type Culture Collection, Manassas, VA) were grown in DMEM supplemented with 10% (v/v) fetal calf serum, 50 units/ml penicillin, 50 μg/ml streptomycin and allowed to reach confluence as described previously. At 2 days post-confluence (day 0), differentiation was initiated by the addition of 100 nm insulin, 1 μm dexamethasone, and 0.25 mm isobutylmethylxanthine in DMEM with 10% (v/v) fetal bovine serum. Three days later (day 3), the induction medium was replaced by DMEM supplemented with 10% (v/v) fetal bovine serum and insulin only and was changed every 2 days. Adipogenesis was assessed by analysis of the expression of adipocyte-specific genes (aP2 or PPARγ) and by lipid accumulation using microscopic analysis.
Total RNA Extraction and Northern Blot Analysis—Total cellular RNA from tissues (liver, muscle, or white adipose tissue) or cells was isolated using the Trizol reagent (Invitrogen) following the manufacturer's instructions. For Northern blot analysis, 10 μg of total RNA was denatured in formamide and formaldehyde at 60 °C and separated by electrophoresis on 1.2% w/v agarose gels. RNA was then transferred to positively charged Hybond-N membranes (Amersham Biosciences) and cross-linked to the membrane by heating to 80 °C. Specific cDNA probes were labeled with [α-32P]dCTP by random priming using the Rediprime kit (Amersham Biosciences) and purified with the Probequant kit (Amersham Biosciences). Blots were hybridized with labeled cDNA probes overnight at 42 °C in NorthernMax hybridization buffer (Ambion, Inc., Austin, TX). Membranes were then washed in 1× SSC, 0.5% (w/v) SDS, and exposed to PhosphorImager for 4-24 h. The signals were scanned (Storm 840) and quantified using ImageQuant 5.0 software (Amersham Biosciences). Blots were stripped and rehybridized with a 5′-32P-labeled 18 S oligonucleotide probe in order to normalize the signal.
Real Time Quantitative PCR—Total RNA was treated with DNase (Ambion), and 1 μg was reverse-transcribed for 60 min at 42 °C using the Reverse Transcription System kit (Promega) in the presence of random primers and oligo(dT)15. Quantitative PCR was performed by monitoring in real time the increase in fluorescence of the SYBR Green dye on an ABI PRISM 7000 Sequence Detector System (Applied Biosystems, Courtaboeuf, France) according to the manufacturer's instructions. PCR primers for each gene were designed using Primer Express software (Applied Biosystems, Courtaboeuf, France), with a melting temperature at 58-60 °C and a resulting product of ∼100 bp. Each PCR was carried out in triplicate in a 20-μl volume using SYBR Green I Master Mix Plus (Eurogentec, Seraing, Belgium) for 15 min at 95 °C for initial denaturing, followed by 40 cycles of 95 °C for 30 s and 60 °C for 30 s in the ABI Prism 7000 sequence Detector System (Applied Biosystems). To exclude the contamination of nonspecific PCR products such as primer dimers, melting curve analysis was applied to all final PCR products after the cycling protocols. Values for each gene were normalized to expression levels of HPRT mRNA in rat tissue and 36B4 mRNA in mouse tissue. Each RT-PCR quantification experiment was performed in triplicate.
Relative quantification of FALDH gene was calculated by using 2-ΔCt formula, as recommended by the manufacturer (Applied Biosystems). Results were expressed relative to the control condition, which was arbitrary assigned a value of 1.
Primers sequences used to quantify FALDH mRNA by real time RT-PCR were designed by using the Primer Express software from Applied Biosystems. Oligonucleotides used were as follows: rat FALDH sense, 5′-AGCCCAGCTACATTGACAGAGA-3′, and antisense, 5′-ACACAGGATATAGTCAGAGCAATACA-3′; mouse FALDH sense, 5′-CAGCATTTCCTGGAGCAATG-3′, and antisense, 5′-AGCTTGGAATTACCCTTTCGTTCT-3′; HPRT sense, 5′-AGCCTGGTCATGTTGCCTTT-3′, and antisense, 5′-AAAGAACTTATAGCCCCCCTTGA-3′; and 36B4 sense, 5′-CTTTATCAGCTGCACATCACTCAGA, and antisense, 5′-TCCAGGCTTTGGGCATCA-3′.
Western Blot—Western blotting was performed on whole cell lysates from isolated adipocytes. Protein extracts were obtained by mixing 200 μl of fat cell suspension with 200 μl of Laemmli buffer (3% (w/v) SDS, 70 mm Tris-HCl, pH 7, 11% (v/v) glycerol). The samples were incubated for 5 min at 90 °C, and protein concentration was assayed with the bicinchoninic acid technique (BCA protein assay kit, Interchim, Monluçon, France). Proteins were separated by SDS-PAGE, transferred to nitrocellulose membranes (Hybond C, Amersham Biosciences), and analyzed by immunoblotting. Immunoreactive proteins were detected by enhanced chemiluminescence (Amersham Biosciences).
Generation of Recombinant Adenoviruses—Adenoviruses expressing FALDH were generated by homologous recombination in Escherichia coli BJ 5183 as described previously (
). Briefly, co-transformation of E. coli BJ 5183 led to recombination between FALDH (cloned in pCDNA3) and a viral vector recombinogenic with the pCDNA3 cytomegalovirus promoter and poly(A) sequence (VmcDNA, provided by S. Rusconi, University of Fribourg, Switzerland) (
). Recombinants were screened by PCR analysis using a pair of primers that annealed to the viral vector and to the cytomegalovirus promoter sequence, respectively. A positive clone harboring FALDH was further amplified in E. coli DH5-α, digested with PacI, and transfected by the calcium phosphate method into helper 293 cells to produce viral particles. Adenoviruses were stored in 0.1 m Tris, 0.25 m NaCl, 1 mg/ml BSA, 50% (v/v) glycerol, pH 7.5, at -20 °C. Adenoviruses expressing p110α CAAX (Ad p110α CAAX, where AA is aliphatic amino acid) and GFP (Ad-GFP) were described previously (
). Starved cells were treated with MGO or 4-HNE at times and concentrations indicated below. The cells were washed two times in PBS and exposed to 5 × 10-5m CM-DCFDA in phosphate-buffered saline for 1 h at 37 °C. The cells were washed two times in PBS, solubilized in water, and sonicated. The fluorescence was determined at 488/525 nm, normalized on a protein basis, and expressed as percentage of control.
Identification of Differentially Expressed Genes by Insulin-induced PI3K Activation in Isolated Rat Adipocytes—To identify novel genes controlled specifically by PI3K in response to insulin, we performed RAP-PCR on freshly isolated rat adipocytes treated or not with insulin (10-7m) and with or without exposure to wortmannin (100 nm), an inhibitor of PI3K. Briefly, poly(A) RNA was extracted for each condition and reverse-transcribed. RAP-PCR was performed using several combinations of arbitrary primers. RAP-PCR products were then separated by electrophoresis. We found 27 clones showing PI3K-dependent modulation in mRNA expression. Subsequent Northern blot analysis identified and confirmed 10 differentially expressed genes in isolated adipocytes, corresponding to both previously identified and unidentified genes. Among these clones, we isolated a RAP-PCR product corresponding to FAS cDNA. The transcriptional control of this gene by insulin via PI3K activation has been described previously (
Characterization of FALDH mRNA Expression in Isolated Adipocytes—Another RAP-PCR product, recognizing a 4-kb mRNA by Northern blot analysis, corresponds to the FALDH gene. By using real time quantitative PCR, we observed a 4-fold increase in FALDH gene expression after stimulation of isolated adipocytes with insulin (10-7m, 6 h) (Fig. 1A). When adipocytes were incubated simultaneously with insulin and wortmannin (100 nm), the increase in expression was abolished, suggesting a PI3K-specific control on the FALDH gene. In parallel, we tested another growth factor, epidermal growth factor (EGF), which also activates the PI3K pathway in adipocytes (Fig. 1B). As expected, we observed the same increase in FALDH gene expression in cells stimulated with EGF and insulin, suggesting that both factors are able to control FALDH expression via a PI3K-dependent pathway. To confirm PI3K activation, a phosphoserine PKB Western blot was performed (Fig. 1B, bottom). Both insulin and EGF increase the phosphorylation of PKB on serine 473, which is compatible with PI3K-dependent PKB activation. Finally, to confirm the role of PI3K in the control of FALDH expression, we performed an adenovirus infection experiment (Fig. 1C) in which isolated adipocytes were infected with Ad-GFP as negative control or with Ad-p110α CAAX, a constitutively active form of PI3K (
). As expected, FALDH expression was increased upon insulin stimulation of adipocytes expressing GFP (4.19-fold). When cells were infected with Ad-p110α CAAX, we observed an insulin-independent increase in FALDH mRNA expression (3.32-fold). Moreover, FALDH expression was not modified in adipocytes ectopically expressing p110α CAAX and treated with the PI3K inhibitor, LY294002. These results confirm that FALDH expression is controlled by a PI3K-dependent pathway.
Effect of Insulin on FALDH Activity in Isolated Hepatocytes—To determine whether insulin increased FALDH activity, we performed FALDH activity measurements on isolated hepatocytes after insulin stimulation (10-7m, 16 h) as described previously (
). As a positive control for FALDH activity, hepatocytes were infected with adenovirus expressing FALDH, and the associated FALDH activity was measured as described under “Experimental Procedures.” We observed an increase in FALDH activity in cells infected with Ad-FALDH (+58%) in comparison to cells infected with Ad-GFP alone (Fig. 1D). Likewise, when cells were stimulated with insulin, we found a significant increase in FALDH activity in comparison to non-stimulated cells.
Tissue Distribution of FALDH Gene—To analyze the tissue distribution of the FALDH gene, we performed Northern blot analysis using total RNA from different rat tissues. As presented in Fig. 2, FALDH mRNA is expressed in all tissues tested except in skeletal muscle and spleen. In addition, the highest FALDH mRNA level was observed in the liver, which is in keeping with the major detoxification function of this organ.
Effects of Insulin Administration on FALDH mRNA Expression in Insulin-responsive Tissues—To determine whether insulin increases FALDH gene expression in vivo, we examined FALDH expression in liver, white adipose tissue, and skeletal muscle of fasted rats after insulin administration (1 IU/kg). RNA was extracted from fat pads, liver, and skeletal muscle 6 h following insulin injection. FALDH expression was examined by Northern blot (Fig. 3A). As a positive control for insulin-induced gene expression, we analyzed the expression of the FAS gene (data not shown). Insulin induced a 2.5-fold increase in the expression of FALDH in the liver. In contrast, by Northern blot, FALDH was not detected in white adipose tissue and in skeletal muscle of either control or insulin-treated animals. By using the more sensitive real time quantitative PCR assay, we examined the amount of FALDH mRNA in WAT and in skeletal muscle. In adipose tissue, we also found a 2.5-fold increase in FALDH mRNA level (Fig. 3B). In contrast, we did not detect expression of FALDH in skeletal muscle in control or insulin-treated animals. Together, our data suggest that insulin plays a role in vivo as a stimulator of FALDH gene expression in the liver and in white adipose tissue.
Effect of Food Intake on FALDH mRNA Expression—Because food intake increases plasma insulin concentration, we also examined the effect of fasting and refeeding on FALDH gene expression. After fasting overnight, mice were divided into the following three groups: one group served as a control fasting group, a second group was treated for 6 h with insulin (1 IU/kg), and the third group was refed a standard diet for 8 h. FALDH expression was analyzed by Northern blot in liver and by real time quantitative PCR in fat pads. Fig. 4A shows the increase in the expression of the FALDH transcript in the liver after insulin stimulation as well as after refeeding. Likewise, by real time PCR we observed a 2-fold increase in FALDH gene expression after refeeding in white adipose tissue (Fig. 4B). To summarize, both insulin injection and food intake increase the expression of the FALDH gene in rat liver and white adipose tissue.
Regulation of FALDH Gene Expression in Type 1 Diabetic Rats—To further confirm the role of insulin in the induction of the FALDH gene, Sprague-Dawley rats were treated or not with streptozotocin (65-70 mg/kg), a drug that specifically destroys beta cells of the endocrine pancreas leading to type 1 diabetes. Untreated control and streptozotocin (STZ)-treated rats were euthanized 15 days after injection. RNA extracted from the liver was analyzed by Northern blot. As shown in Fig. 5A, basal FALDH expression was decreased by ∼75% in STZ rats compared with control rats. The expression of FALDH in white adipose tissue was studied by real time quantitative PCR. As shown on Fig. 5B, basal expression of the FALDH was also decreased (by ∼50%) in white adipose tissue of STZ rats. The finding that FALDH expression is decreased in the liver and white adipose tissue of rats with STZ-induced type 1 diabetes confirms the regulation of the FALDH gene by insulin.
Expression of the FALDH Gene in Insulin-resistant Rodent Models—Several insulin-resistant conditions found in type 2 diabetes or obesity both in humans and rodent models are characterized by a deterioration of the PI3K cascade upon challenge with insulin (
). We thus studied whether FALDH expression is affected by insulin resistance in db/db mice and in mice on a high fat diet (HFD). 8-Week-old db/db mice and control mice C57BL/6 (control) were injected or not with insulin. RNA extracted from white adipose tissue was isolated, and gene expression levels were detected using real time quantitative PCR. In control mice, we observed a 5.7-fold increase in the level of transcript expression in response to insulin (Fig. 6). This variation in gene expression was much less marked (2.5-fold increase approximately) in the db/db mice as compared with control mice. Thus, the positive regulation of the FALDH gene in response to insulin is impaired in this insulin-resistant murine model. We then studied a mouse model of type 2 diabetes induced by HFD, more representative of the situation found in humans. We determined the expression of FALDH gene in C57BL/6 mice fed for 15 weeks with a diet enriched in fat and sugar, as compared with mice fed with standard chow diet (control). We followed body weight and glycemia of the mice for 15 weeks. After this period, HFD mice presented an increase in body weight of ∼11 g compared with control mice, and had a higher fasting glycemia of ∼2 mm. Moreover, measurement of glycemia 90 min after insulin injection (1 IU/kg) indicated that injected HFD mice are insulin-resistant because their reduction in blood glucose is marginal as compared with control mice (data not shown). 6 h after the insulin injection, mice were euthanized, and the expression of FALDH transcript was analyzed by real time quantitative PCR using 1 μg of RNA from white adipose tissue (Fig. 6B). The results obtained show that expression of the FALDH transcript is increased 2-fold in control mice in response to insulin, although it does not vary significantly in HFD mice. To conclude, FALDH gene expression is deregulated in genetic and acquired insulin-resistant models, db/db mice, and mice on a HFD.
FALDH Gene Expression during Adipocyte Differentiation—To determine whether the expression of FALDH gene is modulated during adipocyte differentiation, we performed real time quantitative PCR on the stroma vascular fraction (S-V) and on isolated adipocytes obtained from rats. FALDH gene expression was 5-fold higher in isolated adipocytes compared with the S-V fraction (Fig. 7A), suggesting that FALDH expression is induced during adipocyte differentiation. As a control, we also examined the expression of the adipocyte marker aP2 by Northern blot analysis. As expected, aP2 transcript was weakly expressed in the S-V fraction and significantly expressed in the adipocyte fraction. To confirm that the FALDH gene is induced during differentiation, we monitored FALDH mRNA levels in 3T3-L1 cells during adipocyte differentiation (Fig. 7B). FALDH expression was significantly increased 5 days after the addition of the differentiation mix and reached the maximal level at day 7 (an ∼2-fold increase). As a control for differentiation, the expression of peroxisome proliferator-activated receptor γ (PPARγ), a master regulator of adipocyte gene expression, was also analyzed. As expected, PPARγ expression markedly increased in 3T3-L1 cells throughout differentiation and was maximal at day 7 postinduction. To summarize, expression of FALDH gene increases during differentiation of rat adipocytes and mouse 3T3-L1 cells.
Effect of FALDH on Reactive Oxygen Species Induced by Lipid Peroxidation—ROS produced by lipid peroxidation are very toxic for mammalian cells. 4-HNE is the major toxic α,β-unsaturated aldehyde formed during lipid peroxidation in biological membranes exposed to oxidative stress (
). As FALDH is involved in cell detoxification of lipid peroxidation products, we measured ROS formation in 3T3-L1 adipocytes ectopically expressing FALDH and treated or not with 4-HNE or methylglyoxal (MGO), another unsaturated aldehyde (Fig. 8). As a control of FALDH expression, a Western blot with anti-Myc antibody was performed. We found that FALDH-myc was expressed only in cells infected with the FALDH adenovirus construct. As expected, we observed an increase in the amount of ROS (by ∼40%) in cells treated with MGO or 4-HNE. Ectopic expression of FALDH significantly decreased ROS production in non-treated cells and in cells treated by 4-HNE, suggesting that FALDH protects against the major lipid peroxidation product, 4-HNE. In contrast, FALDH does not modify the amount of ROS in cells treated by MGO, suggesting that MGO is not a substrate of the enzyme.
The PI3K cascade is a key pathway mediating the metabolic effects of insulin. Moreover, altered insulin action, often characterized by a deregulation of the PI3K cascade, is found in both diabetic and obese subjects (
). We thus hypothesized that many genes important in glucose and lipid homeostasis and controlled by insulin through PI3K are likely to exist. Therefore, we believe that the development of insulin resistance is partly due to modifications of gene expression modulated by PI3K. By using RAP-PCR, we identified several genes, the expression of which is controlled by PI3K in response to insulin. Briefly, isolated adipocytes were exposed or not to insulin and, in order to inhibit the activation of endogenous PI3K, were treated or not with wortmannin. RAP-PCR was carried out using RNA extracted from these cells. An advantage of the RAP-PCR approach is that it permits the simultaneous analysis of differential gene expression in more than two experimental conditions. For two differentially expressed genes, the sequences corresponded to cDNA coding for previously identified proteins. One corresponded to fatty-acid synthase, which is involved in insulin action (
). The identification of this protein confirms the validity of our approach.
A second clone corresponded to the cDNA coding for another known protein, FALDH, for which a role in insulin action has not, to our knowledge. been reported previously. FALDH belongs to the family of aldehyde dehydrogenases, which are particularly important for the detoxification of exogenous and endogenous aldehydes such as those resulting from lipid peroxidation (
) that fatty aldehydes accumulate in the diabetic rat liver due to failure of the detoxification systems involving FALDH. Detoxification enzymes such as FALDH may be important for protection against oxidative stresses, which have been linked to the development of diabetic complications, such as atherosclerosis and retinopathy (
Inhibition of PI3K by wortmannin abolished the insulin-induced stimulation of endogenous FALDH mRNA in isolated rat adipocytes. Expression of the constitutively active p110α catalytic subunit of PI3K resulted in an increase in FALDH expression independent of insulin, thereby confirming that FALDH gene expression is controlled by PI3K activity.
EGF is also able to activate PI3K in different cell types including adipocytes. In accord with this, we found that EGF increases FALDH expression to a level similar to that of insulin seen in isolated adipocytes. This result is not surprising because at least one other gene controlled by insulin, plays a role in glucose homeostasis; FAS, has been found to be modulated by EGF (
) have shown that control of FAS gene expression by insulin or EGF was mediated by the transcription factor, SREBP-1c. This protein belongs to a family of three basic helix-loop-helix transcription factors that play a key role in the coordination of lipogenic gene expression and in cellular lipid homeostasis (
). It is thus tempting to speculate that the control of FALDH expression by insulin and EGF is mediated by SREBP-1c and that FALDH may play a role in lipid homeostasis.
To determine whether FALDH expression augmented by insulin was associated with an increase in FALDH activity in hepatocytes, we performed FALDH activity measurement on isolated hepatocytes after exposure to the hormone. We found that insulin treatment of cells increases FALDH activity suggesting that insulin enhances the expression of a functional and active enzyme. However, we did not observe a perfect correlation between the insulin-induced increase in FALDH mRNA expression (∼2.5-fold) and an increase in FALDH enzymatic activity (∼1.5-fold). This could at least be explained by the following hypotheses. First, FALDH activation could be controlled by different mechanisms, as yet unidentified, such as phosphorylation/dephosphorylation and/or post-transcriptional modifications linked to the presence of specific metabolites and independent of the expression of the enzyme, for example. A second possible explanation could be that the increase in FALDH mRNA does not lead to a proportional increase in FALDH protein expression at least in the time frame we analyzed.
Next we studied FALDH gene regulation in response to insulin in vivo. Our results show that in normal rats, insulin specifically increases FALDH gene expression in liver and white adipose tissue. In contrast, gene expression is very low and apparently not modified by insulin in skeletal muscle. We also found that food intake leads to an increase in FALDH mRNA expression in liver and white adipose tissue. This increase could be due to post-prandial insulin secretion. However, in this context, we cannot attribute these effects solely to insulin as other hormones regulated by food intake can act on FALDH gene expression. To fully confirm the regulation of this gene by insulin, we studied the FALDH expression in rats treated with STZ, which specifically destroys pancreatic beta cells and severely reduces insulin production (
). We showed a concomitant reduction in FALDH gene expression compared with control rats. Injection of STZ rats with insulin restored normal FALDH expression in white adipose tissue (data not shown). These results confirm the control of FALDH gene expression by insulin. Moreover, we observed that the expression of FALDH was increased during adipocyte differentiation (rat adipocytes and 3T3-L1 cells), similar to several other genes involved in insulin action. Hence it will be interesting to determine whether FALDH plays a role in adipocyte differentiation.
Next we focused our attention on FALDH gene expression in insulin-resistant states. We studied the following two murine models: (i) db/db mice, which represent a model of monogenic obesity associated with a severe insulin-resistant state (
). Our results show that expression of FALDH mRNA is not modulated in response to insulin in white adipose tissue in these two models. By taking into account the role of FALDH in detoxification, we believe that a reduction in its expression could contribute to the cellular damage found in insulin resistance. Indeed, type 2 diabetes is known to be associated with high oxidative stress, leading to lipid peroxidation. This elevated stress may be due to the accumulation of stress substrates and/or a reduction in detoxification systems. It has been shown recently that lipid peroxidation products increase in the liver of diabetic rats (
). The accumulation of these products, which are very toxic for the cell, is correlated with a reduction in the activity of detoxification enzymes (Fig. 9).
In conclusion, the reduction of FALDH expression that we observed in both insulin-resistant murine models and in type 1 diabetic rats may have an important implication for our understanding of the pathophysiology of the complications associated with diabetes. In addition, our data suggest a role for this enzyme in cell protection against lipid peroxidation under normal conditions.
To confirm this hypothesis, we measured the ROS produced by 3T3-L1 adipocytes as a marker of oxidative stress. To induce oxidative stress, the cells were treated with 4-HNE, the major toxic product of lipid peroxidation (
). Expression of FALDH protected cells from oxidative stress induced by 4-HNE but not from oxidative stress induced by MGO. Indeed, contrary to MGO, 4-HNE appears to be a preferential substrate of FALDH. These results are not surprising because previous studies showed that aldehydes with long carbon chains constitute preferential substrates of the enzyme (
). Thus, FALDH may have an important biological role in oxidizing and detoxifying long chain aldehydes produced during the peroxidation of microsomal membranes.
The control of FALDH gene expression by insulin is particularly intriguing. Indeed, because FALDH is involved in cellular detoxification, we expect that the loss of insulin regulation of gene expression in insulin resistance and in type 1 diabetes may be related to the elevated oxidative stress described in diabetic patients. A better understanding of insulin action on the enzymes involved in detoxification will help to unravel the pathophysiology of diabetes and its associated complications. It is possible that these detoxification enzymes will provide new targets for the prevention and treatment of the metabolic disturbance of diabetes and its complications.
We thank Georges Manfroni for the expertise in animal care; Marie-Noëlle Monthouël for assistance in real time PCR analysis; Claire Chaussade for assistance in adenovirus experiments, Audrey Riboulet for assistance in ROS experiments; and Sarah Longnus, Anne Johnston, and Jean Giudicelli for discussion and critical reading of the manuscript. We also thank Johan Auwerx (IGBMC, Strasbourg, France) and Paul Grimaldi (INSERM U470, Nice, France) for providing cDNAs and William Rizzo (University of Nebraska Medical Center) for the FALDH assay protocol.