Expression of peroxisome proliferator-activated receptors alpha and gamma in differentiating human colon carcinoma Caco-2 cells

The expression of peroxisome proliferator-activated receptors alpha (PPARalpha) and gamma (PPARgamma) was studied in the human adenocarcinoma Caco-2 cells induced to differentiate by long term culture (15 days). The differentiation of Caco-2 cells was attested by increases in the activities of sucrase-isomaltase and alkaline phosphatase (two brush border enzymes), fatty acyl-CoA oxidase (AOX) and catalase (two peroxisomal enzymes), by an elevation in the protein levels of villin (a brush border molecular marker), AOX, peroxisomal bifunctional enzyme (PBE), catalase and peroxisomal membrane protein of 70 kDa (PMP70). and by the appearance of peroxisomes. The expression of PPARalpha and PPARgamma was investigated by Western blotting, immunocytochemistry, Northern blotting and S1 nuclease protection assay during the differentiation of Caco-2 cells. The protein levels of PPARalpha, PPARgamma, and PPARgamma2 increased gradually during the time-course of Caco-2 cell differentiation. Immunocytochemistry revealed that PPARalpha and gamma were localized in cell nuclei. The PPARgamma1 protein was encoded by PPARgamma3 mRNA because no signal was obtained for PPARgamma1 mRNA using a specific probe in S1 nuclease protection assay. The amount of PPARgamma3 mRNA increased concomitantly to the resulting PPARgamma1 protein. On the other hand, the mRNA of PPARalpha and PPARgamma2 were not significantly changed, suggesting that the increase in their respective protein was due to an elevation of the translational rate. The role played by the PPAR subtypes in Caco-2 cell differentiation is discussed.

Fibrate induction of the adrenoleukodystrophy-related gene (ABCD2): promoter analysis and role of the peroxisome proliferator-activated receptor PPARalpha

X-linked adrenoleukodystrophy (X-ALD) is a neurodegenerative disease due to a defect in the ABCD1 (ALD) gene. ABCD1, and the two close homologues ABCD2 (ALDR) and ABCD3 (PMP70), are genes encoding ATP-binding cassette half-transporters of the peroxisomal membrane. As overexpression of the ABCD2 or ABCD3 gene can reverse the biochemical phenotype of X-ALD (reduced beta-oxidation of very-long-chain fatty acids), pharmacological induction of these partially redundant genes may represent a therapeutic approach to X-ALD. We previously reported that the ABCD2 and ABCD3 genes could be strongly induced by fibrates, which are hypolipidaemic drugs and peroxisome-proliferators in rodents. We provide evidence that the induction is dependent on peroxisome proliferator-activated receptor (PPARalpha) as both genes were not induced in fenofibrate-treated PPARalpha -/- knock-out mice. To further characterize the PPARalpha pathway, we cloned and analysed the promoter of the ABCD2 gene, the closest homologue of the ABCD1 gene. The proximal region (2 kb) of the rat promoter displayed a high conservation with the human and mouse cognate sequences suggesting an important role of the region in regulation of the ABCD2 gene. Classically, fibrate-induction involves interaction of PPARalpha with a response element (PPRE) characterized by a direct repeat of the AGGTCA-like motif. Putative PPRE motifs of the rat ABCD2 promoter were studied in the isolated form or in their promoter context by gel-shift assay and transfection of COS-7 cells. We failed to characterize a functional PPRE, suggesting a different mechanism for the PPARalpha-dependent regulation of the ABCD2 gene.

Rat adrenoleukodystrophy-related (ALDR) gene: full-length cDNA sequence and new insight in expression

X-linked adrenoleukodystrophy (X-ALD) is an inherited demyelinating disorder due to mutations in the ALD gene, which encodes a peroxisomal ABC half-transporter (ALDP). It has been suggested that ALDP assembles with ALDRP (adrenoleukodystrophy-related protein), a close homologous half-transporter, to form a functional heterodimer. For the first time full-length ALDRP cDNA (5.5 kb) was cloned, and 5' and 3' RACE analysis revealed that alternative usage of polyadenylation sites generates the two transcripts of 3.0 and 5.5 kb observed in the rat in Northern blot analysis. Southern blotting and chromosomal mapping demonstrated one ALDR locus in the rat genome. Characterisation of the 3' flanking region suggested that an ID sequence might be responsible for high expression of the 5.5 kb ALDRP transcript in rat brain. ALDR gene expression was found to be high in the liver of rats before weaning and very low in adult rats; the reverse developmental regulation was observed in the brain. Fenofibrate, which is a potent inducer of the ALDR gene in the liver of adult rats, could not induce the ALDR gene in suckling rats. The exact significance of this result with regard to development of an efficient pharmacological gene therapy for X-ALD is discussed.

The four murine peroxisomal ABC-transporter genes differ in constitutive, inducible and developmental expression

Four ATP-binding cassette (ABC) half-transporters have been identified in mammalian peroxisomes: adrenoleukodystrophy protein (ALDP), adrenoleukodystrophy-related protein (ALDRP), 70-kDa peroxisomal membrane protein (PMP70) and PMP70-related protein (P70R). Inherited defects in ALDP cause the neurodegenerative disorder X-linked adrenoleukodystrophy (X-ALD). By comparative Northern blot analyses we found each of the four murine peroxisomal ABC transporter mRNA species at maximum abundance only in a few tissues, which differed for each family member. The four genes were also regulated differentially during mouse brain development: ALDP mRNA was most abundant in embryonic brain and gradually decreased during maturation; ALDRP and P70R mRNA accumulated in the early postnatal period; and the amount of PMP70 transcript increased slightly during the second and third postnatal week. The different expression patterns could explain why beta-oxidation is defective in X-ALD, although ALDRP and PMP70 can replace ALDP functionally in fibroblasts. Dietary fenofibrate had no effect on the ALD and P70R genes, but strongly increased expression of the ALDR and PMP70 genes in mouse liver. However, in P-glycoprotein Mdr1a-deficient mice fenofibrate treatment increased ALDR gene expression also in the brain, suggesting that the multidrug-transporter P-glycoprotein restricts entry of fenofibrate to the brain at the blood-brain barrier. Analysis of the promoter sequences revealed a cryptic nuclear hormone receptor response element of the DR+4 type in the ALDR promoter and a novel 18-bp sequence motif present only in the 5' flanking DNA of the ALDR and PMP70 genes. The mouse ALDR gene uses a single transcription start site but alternative polyadenylation sites. These data are of importance for the use of ALDP-deficient mice as a model in pharmacological gene therapy studies.

Copurification of dihydroxyacetone-phosphate acyl-transferase and other peroxisomal proteins from liver of fenofibrate-treated rats

Dihydroxyacetone-phosphate acyl-transferase (DHAP-AT), a peroxisomal membrane-bound enzyme that catalyzes the first step of ether-glycerolipid synthesis, was purified from liver of rats treated with fenofibrate, a peroxisome proliferator. The protocol first included isolation of peroxisomes, their purification through a discontinuous gradient and solubilization of membranes in CHAPS. DHAP-AT was further purified by four chromatographic steps, namely low-pressure size-exclusion, cation-exchange, hydroxylapatite and chromatofocusing. The chromatofocusing step led to a 4000-fold increase in the specific activity of DHAP-AT with respect to the liver homogenate with a yield of about 0.2%. Trypsin digestion of a 64-kDa protein band upon SDS-PAGE resulted in a peptide sequence unknown in databases. A corresponding degenerated oligonucleotide was used as a probe in Northern blotting, and a transcript of 3.3 kb was detected in some rat tissues. Moreover, the overall procedure allowed co-purification of four major peroxisomal enzymes: urate-oxidase, catalase, multifunctional enzyme and palmitoyl-CoA oxidase, respectively.

Fenofibrate differently alters expression of genes encoding ATP-binding transporter proteins of the peroxisomal membrane

The 70-kDa peroxisomal membrane protein (PMP 70), adrenoleukodystrophy protein (ALDP) and adrenoleukodystrophy-related protein (ALDRP) belong to the ATP-binding transporter family, share a structure of half-transporters and are localized in the peroxisomal membrane of mammals. It was suggested that these proteins may heterodimerize to form functional transporters. The expression of the three genes was examined in various tissues of control or fenofibrate (a peroxisome proliferator)-treated rats using Northern and immuno-blotting techniques. The patterns of tissue expression were distinct for the three genes. Upon treatment, expression of the ALD gene was not altered while that of the PMP 70 and ALDR genes was strongly increased in intestine and liver, respectively. The absence of coordinated expression excludes that the three transporters function as exclusive and obligatory partners. We also report for the first time that the ALDR gene is inducible in rodents and that the corresponding mRNA is different in length in rat (3.0 and 5.5 kb) and in mouse and human (4.2 kb).

Differential effects of nonhydroxylated flavonoids as inducers of cytochrome P450 1A and 2B isozymes in rat liver

Flavanone, flavone, and tangeretin differentially affected the activities of cytochrome P540 1A and 2B isozymes in rat liver. Flavone and, to a lesser extent, tangeretin, increased activities of ethoxyresorufin O-deethylase, methoxyresorufin O-demethylase, and pentoxyresorufin O-dealkylase (PROD), whereas flavanone mainly enhanced PROD activity. Immunoblot analysis indicated that flavone and tangeretin increased cytochrome P450 1A1, 1A2, and 2B1,2 forms, whereas flavanone only enhanced the cytochrome P450 2B isozymes. Northern blot study showed that flavone and tangeretin increased the level of the cytochrome P450 1A2 mRNAs. The concentration of the other mRNAs were slightly or not affected by flavonoids. These results suggest that the induction of P450 1A2 by flavone and tangeretin might involve a transcriptional and/or post-transcriptional mechanism.

Effects of two peroxisome proliferators (ciprofibrate and fenofibrate) on peroxisomal membrane proteins and dihydroxyacetone-phosphate acyl-transferase activity in rat liver

The effects of ciprofibrate and fenofibrate, which are more potent peroxisome proliferators than clofibrate, on the activities of dihydroxyacetone-phosphate acyl-transferase (DHAP-AT) and glycerol-3-phosphate acyl-transferase (G3P-AT) were studied at the two pH optima 5.5 and 7.4 in subcellular fractions of rat liver, and in solubilized peroxisomal membranes (PMP) as well. Protein was also analyzed by gel electrophoresis. 1) Under the conditions of the specific activity of peroxisomal acyl-CoA oxidase (CN(-)-ACO) being increased (8 to 9-fold), there was no specific induction of the DHAP-AT activity when measured at pH 5.5 in purified peroxisomes and PMP. However, the total activities of DHAP-AT in these two fractions were increased by 6 to 11 times, as a result of hepatomegaly and peroxisome proliferation. In contrast, they were only slightly enhanced (x 1.1 to 2.2-fold) when determined at pH 7.4. The magnitude of the effects of a fibrate treatment was, therefore, dependent on the pH of the incubation medium. 2) Experiments of reversibility of enzyme induction reinforced the finding that the peroxisomal DHAP-AT activity is not specifically induced by ciprofibrate and fenofibrate. 3) Our results suggest the existence of a peroxisomal G3P-AT, non-inducible by fibrates, in the rat liver. 4) Induction of peroxisomal membrane-associated polypeptides with apparent molecular masses of 26- and 36-kDa was evidenced in stained electrophoretic gels of protein.

Proteins and enzymes of the peroxisomal membrane in mammals

Proteins of the peroxisomal membrane can be schematically divided into two groups, one being made up of more or less characterized proteins with generally unknown functions and the other consisting of enzyme activities of which the corresponding proteins have not been characterized. In the present report, these proteins and enzymes are described with the addition of unpublished results regarding their induction by peroxisome proliferators at the post-transcriptional level. Integral membrane proteins (IMPs) can be isolated using an alkaline solution of sodium carbonate. A dozen of preponderant IMPs can be seen on sodium dodecyl sulfate polyacrylamide gel electrophoresis, and the major band corresponds to a 70 kDa IMP, of which the corresponding rat cDNA is known. Some IMPs have been characterized by immunoblot analysis. Recently, a cDNA has been cloned for a peroxisome assembly factor (35 kDa IMP). Functions have also been proposed for some IMPs but are not yet firmly settled. Some IMPs (450/520, 70 and 26 kDa) are strongly induced by peroxisome proliferators. Our results extend to cipro- and fenofibrate the observation that the 70 kDa IMP mRNA level is strongly increased in di(2-ethylhexyl)phtalate-treated rats. All the enzyme activities associated with the peroxisomal membrane are involved in lipid metabolism: activation of substrates (fatty acids), ether lipid biosynthesis, and formation of precursors (fatty alcohols). It is believed that the same long-chain acyl-CoA synthetase occurs in the peroxisome as well as in the outer mitochondrial membrane and the endoplasmic reticulum. However, two highly homologous but different cDNAs encoding rat liver and brain long-chain acyl-CoA synthetases have been isolated recently. Evidence has been accumulated for a distinct synthetase that specifically activates very-long chain fatty acids. The first two steps of ether lipid biosynthesis require dihydroxyacetone-phosphate (DHAP) acyltransferase and alkyl-DHAP synthetase, the active sites of which are located on the inner surface of the membrane. In contrast, the catalytic site of the acyl/alkyl-DHAP reductase, which generates sn-1-alkyl-glycerol-3-phosphate, is located on the outer surface. Long-chain fatty alcohols, which are obligate precursors of ether lipids and wax esters, are biosynthetized by the reduction of the corresponding acyl-CoAs via the action of an acyl-CoA reductase. Peroxisome proliferators do not appear to stimulate these enzyme activities specifically. However, we report that feno- and ciprofibrate treatments increase six-fold the palmitoyl-CoA synthetase mRNA level in the rat liver.

Unsaturated fatty acids esterified in 2-acyl-l-lysophosphatidylcholine bound to albumin are more efficiently taken up by the young rat brain than the unesterified form

The aim of the present study was to investigate whether unsaturated 2-acyl-lysophosphatidylcholine bound to plasma albumin is a relevant delivery form of unsaturated fatty acids to the developing brain. Twenty-day-old rats were perfused for 30 s with labeled palmitic, oleic, linoleic, and arachidonic acids in either their unesterified form or esterified in 2-acyl-lysophosphatidylcholine labeled on the choline and fatty acid moieties. Both forms were bound to albumin. Incorporation in brain lipid classes was followed within 1 h. The brain uptake of the unesterified fatty acids reached a plateau at 5-15 min and was maximal for arachidonic acid (0.45% of the perfused dose). The brain uptake of palmitoyl-lysophosphatidylcholine was similar to that of palmitic acid, whereas that of other lysophosphatidylcholines increased with the degree of unsaturation (rate and maximal uptake) and was six- to 10-fold higher than that of the corresponding unesterified fatty acid. 2-Acyl-lysophosphatidylcholines were taken up without prior hydrolysis and reacylated into doubly labeled phosphatidylcholine, which was the most labeled lipid class, whereas lipid distribution of the unesterified fatty acid was more diversified. Partial hydrolysis of 2-acyl-lysophosphatidylcholine occurred in the brain tissue, and redistribution of the fatty acyl moiety into other phospholipid classes was also observed and was the highest for arachidonic acid. In this case, the percentage of esterification of this fatty acid in phosphatidylinositol (expressed as a percentage of the total lipid fraction) was relatively lower than that observed when the unesterified form was used.(ABSTRACT TRUNCATED AT 250 WORDS)

Metabolic fate of sphingomyelin of high-density lipoprotein in rat plasma

The metabolic fate of high density lipoprotein (HDL) sphingomyelin in plasma was studied in rats over a 24-hr period after injection of HDL containing sphingomyelin which was 14C-labeled in the stearic (18:0) or lignoceric acid (24:0) moiety and 3H-labeled in the choline methyl groups. Decay of label in plasma followed three phases. The first two phases were similar for both isotopes and both types of sphingomyelin (t1/2 approximately 10 and 110 min). However, during the third phase (from 10 hr after injection), 3H label disappeared more slowly than 14C label from 18:0 sphingomyelin, whereas the 3H/14C ratio remained relatively constant when 24:0 sphingomyelin was used. Intact, doubly-labeled 18:0 sphingomyelin disappeared from HDL rapidly (t1/2 = 38 min) by tissue uptake and by transfer to very low density lipoprotein (VLDL). VLDL contained up to 12% of the sphingomyelin 1 hr after injection. This is the first demonstration of a transfer in vivo of sphingomyelin from HDL to VLDL. A similarly rapid transfer was also observed in vitro. Some nontritiated, [14C]18:0 or [14C]24:0 sphingomyelin was redistributed more slowly into HDL. Doubly-labeled phosphatidylcholine appeared in VLDL and HDL within 1 hr after injection and reached 1.8 and 2.1% of the injected 14C and 3H in VLDL at 1 hr, and 4.8 and 6.9% in HDL at 3 hr, respectively.

Time-course of utilization of [stearic or lignoceric acid]sphingomyelin from high-density lipoprotein by rat tissues

Utilization of stearic and lignoceric acids supplied by high-density lipoprotein (HDL) sphingomyelin to different tissues was followed for 24 h after rats were injected with HDL containing [[1-14C]stearic (18:0) or [1-14C]lignoceric (24:0) acid [Me-3H]choline]sphingomyelin. Both isotopes reached a maximum in tissue lipids 3-12 h after injection and were recovered mainly in the liver (30%) and small intestine (3%), whereas the other tissues contained approx. 1% or less of the injected dose. All the tissues were able to take up some intact sphingomyelin from HDL and hydrolyze it. In the lung and erythrocytes, the 3H:14C ratio of sphingomyelin remained unchanged throughout the studied period, while an increase in the isotopic ratio was observed in the kidney due to the 3H choline moiety re-used for synthesis of new sphingomyelin. Conversely, the isotopic ratio of sphingomyelin decreased in the liver, indicating a saving of the 14C-labelled fatty acids, especially 24:0. Furthermore, [24:0]ceramide in the liver remained at a high level (6% of the injected dose), whereas [18:0]ceramide decreased to 1%. When the tissues were examined 24 h after injection, the proportion of the 14C linked to sphingomyelin in the total 14C was always higher for both kinds of sphingomyelin than the molar proportion of sphingomyelin in the whole of lipid classes. However, in the majority of the extra-hepatic tissues, more [14C]18:0 than [14C]24:0 was recovered in sphingomyelin, and more 14C radioactivity from 18:0 than from 24:0 was redistributed in the other lipids. The choline moiety from both kinds of sphingomyelin was re-used to synthesize phosphatidylcholine, especially in the liver (up to 20% of the injected dose). All these results show that utilization of sphingomyelin from HDL by tissues normally occurs in vivo and that this phenomenon should be taken into account in the study of the phospholipid turnover of cell membranes. They also show that metabolism of sphingomyelin from HDL in the liver and other tissues is dependent on the sphingomyelin acyl moiety.

Uptake and utilization of double-labeled high-density lipoprotein sphingomyelin in isolated brain capillaries of adult rats

Isolated rat brain capillaries were incubated in the presence of high-density lipoprotein (HDL) containing [stearic acid-14C, (methyl-3H)choline]sphingomyelin. This double-labeled sphingomyelin was taken up in a concentration-dependent manner. Cerebral capillary-associated sphingomyelin had a 3H/14C ratio close to that of the incubation medium, a result indicating uptake of sphingomyelin without prior hydrolysis. TLC of lipid extracted from capillaries showed that part of the sphingomyelin (up to 40%) was hydrolyzed in the brain capillaries to ceramide and free fatty acids. The hydrolysis was proportional to the amount of incorporated sphingomyelin and reached a plateau when the HDL sphingomyelin concentration in the medium was 237 nmol/ml. The results of "pulse-chase" experiments showed that the choline moiety of sphingomyelin was recovered in the incubation medium after the chase period and that there was no redistribution of liberated choline in phosphatidyl-choline of capillaries.

Utilization of high-density lipoprotein sphingomyelin by the developing and mature brain in the rat

Utilization of very long chain saturated fatty acids by brain was studied by injecting 20-day-old and adult rats with high-density lipoprotein containing [stearic or lignoceric acid-14C, (methyl-3H)choline]sphingomyelin. Labeling was followed for 24 h. Very small amounts of 14C were recovered in the brain of all rats, and there was no preferential uptake of lignoceric acid. Approximately 20% of the entrapped 14C was located in the form of unchanged sphingomyelin 24 h after injection. This result shows that the rat brain utilizes very little very long chain fatty acids (greater than or equal to 20 C atoms) from high-density lipoprotein sphingomyelin, even during the myelinating period. The [3H]choline moiety from sphingomyelin was recovered in brain phosphatidylcholine in a higher proportion in comparison with the 14C uptake. The brain 3H increased throughout the studied period in all experiments, but was much higher in the myelinating brain than in the mature brain. From the radioactivity distribution in liver and plasma lipids, it is clear that the choline 3H in the brain originates from either double-labeled phosphatidylcholine of lipoproteins or tritiated lysophosphatidylcholine bound to albumin, both synthesized by the liver.

Turnover and uptake of double-labelled high-density lipoprotein sphingomyelin in the adult rat

Rat HDL containing [stearic acid-14C, (methyl-3H)choline]sphingomyelin was prepared by incubating labelled sphingomyelin liposomes with serum. HDL was then separated by ultracentrifugation and purified by gel-filtration chromatography. The maximum transfer was reached when 1.5 microliter sphingomyelin was incubated in the presence of 1 ml of serum at 37 degrees C for 1 h. When transfer was limited to a 5-7% increase in HDL mass, no significant change was observed in the HDL electrophoretic pattern, and rats could therefore be injected with this type of HDL under physiological conditions. Plasma radioactivity decay was followed for 24 h, and the recovery of both isotopes in 11 tissues was studied 24 h after the injection. The decay in plasma of both isotopes followed three exponential phases. During the first two phases, both isotopes disappeared with the same velocity (t1/2 = 12.8 and 98-105 min for the first and second phases, respectively). 10 h after injection, 3H had disappeared more slowly than 14C (t1/2 = 862 and 502 min for 3H and 14C, respectively) and 24 h after injection, only 1.5% of 14C and 2.5% of 3H remained in the plasma. This radioactivity was located mainly in HDL (80-85% for 3H and 14C), with a 3H/14C ratio close to that of injected sphingomyelin, and in VLDL, with the same isotopic ratio as that of liver lipids. Some 3H was associated with non-lipoprotein proteins. 17.5% of 3H and 23.4% of 14C were recovered in the liver, 1.6% of each isotope in erythrocytes, and 1.4% of 3H and 0.6% of 14C in kidney. Less than 1% of each isotope was recovered in each of the other tissues. Phosphatidylcholine was the lipid most labelled, and in several tissues sphingomyelin had a 3H/14C ratio close to that of injected sphingomyelin, showing an uptake without prior hydrolysis.