G P Mannaerts

Docosahexaenoic acid deficit is not a major pathogenic factor in peroxisome-deficient mice

Docosahexaenoic acid (DHA), a major component of membrane phospholipids in brain and retina, is profoundly reduced in patients with peroxisome biogenesis disorders (Zellweger syndrome). Supplementing newborn patients with DHA resulted in improved muscular tone and visual functions. The purpose of this study was to investigate (a) whether DHA levels were also reduced in newborn PEX5 knockout mice, the mouse model of Zellweger syndrome that we recently generated; (b) whether these levels could be normalized by supplying DHA; and (c) whether this results in longer survival. The DHA concentration in brain of newborn PEX5−/− mice was reduced by 40% as compared with levels in normal littermates; in liver, no differences were noticed. The daily administration of 10 mg of DHA-ethyl ester (EE) to pregnant heterozygous mothers during the last 8 days of gestation resulted in a normalization of brain DHA levels in Zellweger pups. However, no clinical improvement was observed in these pups, and the neuronal migration defect was unaltered. These data suggest that the accretion of DHA in the brain at the end of embryonic development is not only supported by the maternal supply but also depends on synthesis in the fetal brain. Furthermore, the DHA deficit does not seem to be a major pathogenic factor in the newborn Zellweger mice.

The role of 2-hydroxyacyl-CoA lyase, a thiamin pyrophosphate-dependent enzyme, in the peroxisomal metabolism of 3-methyl- branched fatty acids and 2-hydroxy straight-chain fatty acids

2-Hydroxyphytanoyl-CoA lyase (abbreviated as 2-HPCL), renamed to 2-hydroxyacyl-CoA lyase (abbreviated as HACL1), is the first peroxisomal enzyme in mammals that has been found to be dependent on TPP (thiamin pyrophosphate). It was discovered in 1999, when studying alpha-oxidation of phytanic acid. HACL1 has an important role in at least two pathways: (i) the degradation of 3-methyl-branched fatty acids like phytanic acid and (ii) the shortening of 2-hydroxy long-chain fatty acids. In both cases, HACL1 catalyses the cleavage step, which involves the splitting of a carbon-carbon bond between the first and second carbon atom in a 2-hydroxyacyl-CoA intermediate leading to the production of an (n-1) aldehyde and formyl-CoA. The latter is rapidly converted into formate and subsequently to CO(2). HACL1 is a homotetramer and has a PTS (peroxisomal targeting signal) at the C-terminal side (PTS1). No deficiency of HACL1 has been described yet in human, but thiamin deficiency might affect its activity.

Thiamine pyrophosphate: An essential cofactor for the α - oxidation in mammals - implications for thiamine deficiencies?

Abstract.The identification of 2-hydroxyphytanoyl-CoA lyase (2-HPCL), a thiamine pyrophosphate (TPP)-dependent peroxisomal enzyme involved in the α-oxidation of phytanic acid and of 2-hydroxy straight chain fatty acids, pointed towards a role of TPP in these processes. Until then, TPP had not been implicated in mammalian peroxisomal metabolism. The effect of thiamine deficiency on 2-HPCL and α-oxidation has not been studied, nor have possible adverse effects of deficient α-oxidation been considered in the pathogenesis of diseases associated with thiamine shortage, such as thiamine-responsive megaloblastic anemia (TRMA). Experiments with cultured cells and animal models showed that α-oxidation is controlled by the thiamine status of the cell/tissue/organism, and suggested that some pathological consequences of thiamine starvation could be related to impaired α-oxidation. Whereas accumulation of phytanic acid and/or 2-hydroxyfatty acids or their α-oxidation intermediates in TRMA patients given a normal supply of thiamine is unlikely, this may not be true when malnourished.

Activation of 3-methyl-branched fatty acids in rat liver.

1. Subcellular fractionation of rat liver revealed that 3-methylmargaric acid, a monobranched phytanic acid analogue, can be activated by mitochondria, endoplasmic reticulum and peroxisomes. 2. Indirect data (effects of pyrophosphate and Triton X-100) suggested that the peroxisomal activation of 3-methylmargaric, 2-methylpalmitic and palmitic acid is catalyzed by different enzymes. 3. Despite many attempts, column chromatography of solubilized peroxisomal membrane proteins so far did not provide more conclusive data. On various matrices, lignoceroyl-CoA synthetase clearly eluted differently from the synthetases acting on 3-methylmargaric, 2-methylpalmitic and palmitic acid. The latter three however, tended to coelute together, although often not in an identical manner.

Peroxisomal localization of α-oxidation in human liver

It was found that α-oxidation in rat liver is a peroxisomal process, consisting of an activation, a 2-hydroxylation, and a reaction leading to the production of formate. α-Oxidation of 3-methyl-substituted fatty acids was quantified by measuring the production of formate and CO2, and the production of a 2-hydroxy-3-methylacyl-CoA-intermediate was demonstrated. We wanted to extend these findings to human liver, in view of the controversy over the subcellular localization of α-oxidation in man.In homogenates from human liver, rates of α-oxidation were highest when measured in the presence of ATP, CoA, Mg2+, 2-oxoglutarate, ascorbate and Fe2+. In subcellular fractions prepared by differential centrifugation and in fractions obtained after subfractionation of a peroxisome-enriched fraction on a Percoll gradient, production of formate and of a 2-hydroxy-3-methylacyl-CoA intermediate coincided with the peroxisomal marker catalase. In broken fractions, production of CO2 was almost negligible as compared to formate production.We conclude from our findings that in human liver, as in rat liver, α-oxidation of 3-methyl-substituted fatty acids is a peroxisomal process, consisting of an activation reaction, a 2-hydroxylation reaction and a reaction or reactions leading to the generation of formate as a primary product which is subsequently converted to CO2. Furthermore activation, 2-hydroxylation and generation of formate appear to be coupled (see Casteels et al 1996). These data demonstrate that Refsum disease should indeed be classified as a peroxisomal disease.

Identification of a Novel Human Peroxisomal 2,4-Dienoyl-CoA Reductase Related Protein Using the M13 Phage Protein VI Phage Display Technology

Recently, we reported the successful use of the gVI-cDNA phage display technology to clone cDNAs coding for novel peroxisomal enzymes by affinity selection using immobilized antisera directed against peroxisomal subfractions (Fransen, M.; Van Veldhoven, P.P.; Subramani, S. Biochem. J., 1999, 340, 561-568). To identify other unknown peroxisomal enzymes, we further exploited this promising approach. Here we report the isolation and cloning of another novel human cDNA encoding a protein ending in the tripeptide AKL, a C-terminal peroxisomal targeting signal (PTS1). Primary structure analysis revealed that this molecule shared the highest sequence similarity to members of the 2,4-dienoyl-CoA reductase (DCR) family. However, functional analysis indicated that a recombinantly expressed version of the novel protein did not possess DCR activity with either 2-trans,4-trans-hexadienoyl-CoA or 2-trans,4-trans-decadienoyl-CoA as a substrate. The recombinant protein interacted with HsPex5p, the human PTS1-binding protein. Binding was competitively inhibited by a PTS1-containing peptide and was abolished when the last amino acid of the PTS1 signal was deleted. Transfection of mammalian cells with gene fusions between green fluorescent protein (GFP) and the human cDNA confirmed a peroxisomal localization and, therefore, the functionality of the PTS1. These results further demonstrate the suitability of the gVI-cDNA phage display technology for cDNA expression cloning using an antibody as a probe.

Permeability of the Peroxisomal Membrane

Purified peroxisomes are permeable to sucrose and to a variety of other small molecules, including the cofactors NAD, CoA, ATP and carnitine. The permeability of the organelles is not mediated by the presence of a number of specific translocases, since cofactor diffusion into isolated peroxisomes was rapid even at low temperature, not saturable and not inhibited by analogs. Reconstitution of the peroxisomal integral membrane protein fraction into liposomes made the vesicles permeable to sucrose and to other molecules that entered the intact organelle. Separation of the integral membrane proteins on sucrose density gradients and reconstitution of the gradient fractions into liposomes suggested that the permeabilizing activity resides with a 22 kDa integral membrane protein. The observations indicate that the permeability of the peroxisomal membrane to small water-soluble molecules is based on the presence of a nonselective poreforming protein.

The growing family of peroxisomal thiolases

Thiolases catalyze the reversible thiolytic cleavage of different fatty 3-oxoacyl-CoA. Animal tissues contain two classes of thiolase: acetoacetyl-CoA specific thiolases and 3-oxoacyl-CoA thiolases that possess a broad substrate specificity and act on medium and long straight-chain 3-oxoacyl-CoA. Thiolases participate in different catabolic (fatty acid oxidation and bile acid formation) and anabolic (cholesterogenesis, ketone body synthesis, fatty acid elongation) processes. Peroxisomes play a role in most of these pathways and harbor different thiolases. The aim of the present study was to elucidate the functional significance of multiple thiolases in mammalian peroxisomes. Methods for the purification of peroxisomal thiolases were based on the alteration in their chromatographic properties in presence or absence of low concentrations of CoA. In peroxisomes from normal rat liver, four main thiolases were detected: 3-oxoacyl-CoA thiolase A (thiolase A), sterol carrier protein 2/3-oxoacyl-CoA thiolase (SCP-2/thiolase), and two new enzymes: acetoacetyl-CoA thiolase and a long chain 3-oxoacyl-CoA thiolase (LC-thiolase). Purified preparations of thiolase A and 3-oxoacyl-CoA thiolase B (thiolase B, the enzyme was isolated from clofibratetreated rat liver) reacted with antibodies raised against thiotase A, but could be distinguished by their isoelectric points, their N-terminal amino acid sequences and their stability, the latter being higher for thiolase A. Thiotase A displayed a substrate specificity that is roughly similar to that of thiolase B. Both enzymes reacted with short, medium, and long straightchain 3-oxoacyl-CoA (1). Purified preparation of SCP-2/thiolase consisted of a 58 and a 46 kDa polypeptide. Internal peptide sequencing and immunoblot analysis revealed that the 46 kDa polypeptide is the N-terminal (thiolase) domain of SCP-2/thiolase. The enzyme exists in vivo as a mixture of three isoforms consisting of homoand heterodimeric combinations of the 58 and 46 kDa subunits. SCP-2/thiolase is active with medium and long straight-chain 3-oxoacyl-CoA but also with 2-methylbranched 3-oxoacyl-CoA and the 24-oxoacyl-CoA derivative of trihydroxycoprostanic acid (2). Acetoacetyl-CoA thiolase is a homotetrameric protein with a subunit molecular mass of 42 kDa. The enzyme cleaves only acetoacetyl-CoA as substrate and is also active in the reverse direction (condensation of two molecules of acetylCoA). Kinetic analysis, internal peptide microsequencing, immunological, and some other data indicate that the peroxisomal, cytoplasmic, and mitochondrial acetoacetyl-CoA thiolases are three closely related but distinct proteins. Acetoacetyl-CoA thiolase from peroxisomes most probably catalyzes the initial step of cholesterol synthesis in these organelles. Peroxisomal LC-thiolase has an unusual substrate specificity that distinguishes it from other known mammalian thiolases. After subfractionation of peroxisomes, the enzyme activity was recovered in the membrane fraction.

ALPHA -OXIDATION OF 3-METHYL-BRANCHED FATTY ACIDS : A REVISED PATHWAY CONFINED TO PEROXISOMES

c~-Oxidation of phytanic acid (3,7,11,15-tetramethylhexadecanoic acid), a 3-methyl-branched isoprenoid-derived fatty acid, was generally believed to consist of a hydroxylation in the 2-position, followed by an oxidative decarboxylation generating CO 2 and a fatty acid shortened by one carbon. Until recently, however, the (z-oxidation process has remained a matter of debate as to its subcellular localization, the reaction intermediates, and the cofactors involved (1). The observation that formate is formed during o~-oxidation of phytanic acid in human skin fibroblasts (2) has been the basis for the recent unraveling of the cz-oxidation process. For our study, we used synthetic 3-methyl-branched fatty acids that were shown to be valid substitutes for phytanic acid in the study of the (z-oxidation process. As in human fibroblasts, formate was formed during a-oxidation of 3-methyl-branched fatty acids in intact rat hepatocytes (3) and was shown to be the first product which is then converted to CO 2, presumably in an NAD+-dependent cytosolic process (4). Consequently, o~-oxidation was assessed by the sum of formate and CO 2 in all further experiments. When studied in permeabilized rat hepatocytes and rat liver homogenates, the a-oxidation process required ATE CoA, Mg 2+, Fe 2+, ascorbate, and 2-oxoglutarate (4). Subsequent experiments in subcellular fractions of rat liver demonstrated that this set of cofactors reflects two steps: first, activation by an acyl-CoA synthetase requiring ATR CoA, and Mg 2÷ and then 2-hydroxylation of the resulting 3-methylacylCoA by a 3-methylacyl-CoA hydroxylase requiring Fe 2÷, ascorbate, and 2-oxoglutarate (4). The whole 0~-oxidation pathway up to the formation of formate was clearly a peroxisomal process (4). Mihalik et al. (5) came to the same conclusion using a different approach. Furthermore, our results suggest that the 2-hydroxylation is a membrane-bound process (4). The 2-hydroxy-3-methylacyl-CoA is cleaved into formyl-CoA and a 2-methyl-branched fatty aldehyde by a 2hydroxy-3-methylacyl-CoA lyase, probably located in the peroxisomal matrix and requiring no additional cofactors. In peroxisomes formyl-CoA is actively converted to formate (6), Lipids 34, S159 (1999). *To whom correspondence should be addressed at Katholieke Universiteit Leuven, Campus Gasthuisberg, Afdeling Farmakologie, Herestraat 49, B3000 Leuven, Belgium. while the 2-methyl-branched fatty aldehyde is converted to a 2-methyl-branched fatty acid in the presence of NAD + (7). After activation to its CoA-ester, the 2-methyl-branched fatty acid can be degraded by peroxisomal ~-oxidation. Our studies show that in rat liver the whole c~-oxidation process up to the formation of formate and a 2-methylbranched fatty acid consists of five enzymatic steps, which are all confined to peroxisomes.