Pex7p is a WD40-containing protein involved in peroxisomal import of proteins containing an N-terminal peroxisome-targeting signal (PTS2). The interaction of human recombinant Pex7p expressed in different hosts/systems with its PTS2 ligand and other peroxins was analysed using various experimental approaches. Specific binding of human Pex7p to PTS2 could be demonstrated only when Pex7p was formed in vitro by a coupled transcription/translation system or synthesized in vivo in Chinese hamster ovary K1 cells transfected with a construct coding for a Pex7p-green fluorescent protein (GFP) fusion protein. Apparently, no cofactors are required and only monomeric Pex7p binds to PTS2. The interaction is reduced upon cysteine alkylation and is impaired upon truncation of the N-terminus of Pex7p. Interaction of Pex7p with other peroxins could not be demonstrated in bacterial or yeast two-hybrid screens, or in pull-down binding assays. The GFP fusion proteins, tagged at either the N- or C-terminus, were able to restore PTS2 import in rhizomelic chondrodysplasia punctata fibroblasts, and Pex7p-GFP was located both in the lumen of peroxisomes and in the cytosol.

Based on peroxin protein 5 (Pex5p) homology searches in the expressed sequence tag database and sequencing of large full-length cDNA inserts, three novel and related human cDNAs were identified. The brain-derived cDNAs coded for two related proteins that differ only slightly at their N-terminus, and exhibit 39.8% identity to human PEX5p. The shorter liver-derived cDNA coded for the C-terminal tetratricopeptide repeat-containing domain of the brain cDNA-encoded proteins. Since these three proteins specifically bind to various C-terminal peroxisome-targeting signals in a manner indistinguishable from Pex5p and effectively compete with Pex5p in an in vitro peroxisome-targeting signal 1 (PTS1)-binding assay, we refer to them as ‘Pex5p-related proteins’ (Pex5Rp). In contrast to Pex5p, however, human PEX5Rp did not bind to Pex14p or to the RING finger motif of Pex12p, and could not restore PTS1 protein import in Pex5 −/− mouse fibroblasts. Immunofluorescence analysis of epitope-tagged PEX5Rp in Chinese hamster ovary cells suggested an exclusively cytosolic localization. Northern-blot analysis showed that the PEX5R gene, which is localized to chromosome 3q26.2–3q27, is expressed preferentially in brain. Mouse PEX5Rp was also delineated. In addition, experimental evidence established that the closest-related yeast homologue, YMR018wp, did not bind PTS1. Based on its subcellular localization and binding properties, Pex5Rp may function as a regulator in an early step of the PTS1 protein import process.

The ontogeny of the following peroxisomal metabolic pathways was evaluated in mouse liver and brain: α-oxidation, β-oxidation and ether phospholipid synthesis. In mouse embryos lacking functional peroxisomes (PEX5 -/- knock-out), a deficiency of plasmalogens and an accumulation of the very-long-chain fatty acid C 26:0 was observed in comparison with control littermates, indicating that ether phospholipid synthesis and β-oxidation are already active at mid-gestation in the mouse. Northern analysis revealed that the enzymes required for the β-oxidation of straight-chain substrates are present in liver and brain during embryonic development but that those responsible for the degradation of branched-chain substrates are present only in liver from late gestation onwards. The expression pattern of transcripts encoding enzymes of the α-oxidation pathway suggested that α-oxidation is initiated in the liver around birth and is not active in brain throughout development. Remarkably, a strong induction of the mRNA levels of enzymes involved in α-oxidation and β-oxidation was observed around birth in the liver. In contrast, enzyme transcripts that were expressed in brain were present at rather constant levels throughout prenatal and postnatal development. These results suggest that the defective ether phospholipid synthesis and/or peroxisomal β-oxidation of straight-chain fatty acids might be involved in the pathogenesis of the prenatal organ defects in peroxisome-deficient mice and men.

To elucidate unknown mammalian peroxisomal enzymes and functions, we subjected M13 phage expressing fusions between the gene encoding protein VI and a rat liver cDNA library to an immunoaffinity selection process in vitro (biopanning) with the use of antibodies raised against peroxisomal subfractions. In an initial series of biopanning experiments, four different cDNA clones were obtained. These cDNA species encoded two previously identified peroxisomal enzymes, catalase and urate oxidase, and two novel proteins that contained a C-terminal peroxisomal targeting signal (PTS1). A primary structure analysis of these novel proteins revealed that one, ending in the tripeptide AKL, is homologous to the yeast peroxisomal 2,4-dienoyl-CoA reductase (EC 1.3.1.34; DCR), an enzyme required for the degradation of unsaturated fatty acids, and that the other, ending in the tripeptide SRL, is a putative member of the short-chain dehydrogenase/reductase (SDR) family, with three isoforms. Green fluorescent protein (GFP) fusions encoding GFP-DCR-AKL, GFP-DCR, GFP-SDR-SRL and GFP-SDR were expressed in mammalian cells. The analysis of the subcellular location of the recombinant fusion proteins confirmed the peroxisomal localization of GFP-DCR-AKL and GFP-SDR-SRL, as well as the functionality of the PTS1. That the AKL protein is indeed an NADPH-dependent DCR was demonstrated by showing DCR activity of the bacterially expressed protein. These results demonstrate at the molecular level that mammalian peroxisomes do indeed contain a DCR. In addition, the results presented here indicate that the protein VI display system is suitable for the isolation of rare cDNA clones from cDNA libraries and that this technology facilitates the identification of novel peroxisomal proteins.

The functionality of the C-terminus (Ser-Asn-Leu; SNL) of human d -aspartate oxidase, an enzyme proposed to have a role in the inactivation of synaptically released d -aspartate, as a peroxisome-targeting signal (PTS1) was investigated in vivo and in vitro . Bacterially expressed human d -aspartate oxidase was shown to interact with the human PTS1-binding protein, peroxin protein 5 (PEX5p). Binding was gradually abolished by carboxypeptidase treatment of the oxidase and competitively inhibited by a Ser-Lys-Leu (SKL)-containing peptide. After transfection of mouse fibroblasts with a plasmid encoding green fluorescent protein (GFP) extended by PKSNL (the C-terminal pentapeptide of the oxidase), a punctate fluorescent pattern was evident. The modified GFP co-localized with peroxisomal thiolase as shown by indirect immunofluorescence. On transfection in fibroblasts lacking PEX5p receptor, GFP–PKSNL staining was cytosolic. Peroxisomal import of GFP extended by PGSNL (replacement of the positively charged fourth-last amino acid by glycine) seemed to be slower than that of GFP–PKSNL, whereas extension by PKSNG abolished the import of the modified GFP. Taken together, these results indicate that SNL, a tripeptide not fitting the PTS1 consensus currently defined in mammalian systems, acts as a functional PTS1 in mammalian systems, and that the consensus sequence, based on this work and that of other groups, has to be broadened to ( S /A/C/K/N)-( K /R/H/Q/N/S)- L .

Ceramide has been suggested to be a potent bioactive lipid involved in cell growth, differentiation and apoptosis. Its precursor, dihydroceramide, does not affect these processes. The truncated dihydroceramide analogues N -hexanoyl-[4,5- 3 H]- D - erythro- sphinganine and N -[1- 14 C]-hexanoyl- D - erythro -sphinganine were used to study the conversion of dihydroceramide into ceramide by rat hepatocytes. The formation of tritiated water after the addition of the tritiated substrate to intact and permeabilized rat hepatocytes was followed to measure enzyme activity. Desaturation was severely depressed in permeabilized hepatocytes, suggesting loss of cofactors. Of a variety of cofactors tested in the permeabilized cells, NADPH appeared to be stimulatory, pointing to the involvement of a desaturase. In agreement with this, the addition of inhibitors and redox effectors known to affect δ 9 -stearoyl-CoA desaturase and δ 1 -plasmanylethanolamine desaturase to intact cells resulted in severe inhibition of the desaturation. When added to permeabilized cells fortified with NADPH, these compounds counteracted the NADPH stimulation. The enzyme system was further studied in broken cells. On cell fractionation, the activity was recovered in the microsomal fraction. The results indicate that the conversion of dihydroceramide into ceramide is catalysed by a desaturase and not by a dehydrogenase or an oxidase as was generally believed.

In the rat, 2-methyl branched fatty acids and the bile acid intermediates di- and tri-hydroxycoprostanic acids are desaturated by pristanoyl-CoA oxidase and trihydroxycoprostanoyl-CoA oxidase respectively. In the human, these compounds are oxidized by a single enzyme, branched-chain acyl-CoA oxidase, which according to its amino acid sequence is the human homologue of rat trihydroxycoprostanoyl-CoA oxidase. Pristanoyl-CoA oxidase is apparently absent from human tissues as indicated by immunoblot analysis [Van Veldhoven, Van Rompuy, Fransen, de Béthune and Mannaerts (1994) Eur. J. Biochem. 222 , 795–801] and Northern-blot analysis [Vanhooren, Fransen, de Béthune, Baumgart, Baes, Torrekens, Van Leuven, Mannaerts and Van Veldhoven (1996) Eur. J. Biochem. 239 , 302–309] of human tissues. In this paper we present evidence, however, that at least the gene for pristanoyl-CoA oxidase is present in the human. A human liver cDNA encoding a protein of 700 amino acids, showing 75% amino acid identity with rat pristanoyl-CoA oxidase and harbouring a peroxisomal C-terminal-targeting signal (SKL), was isolated. Bacterial expression of the cDNA resulted in a fusion protein that was cross-reactive with antibodies directed against rat pristanoyl-CoA oxidase and the C-terminal SKL sequence. Screening of a genomic library with the isolated cDNA as a probe resulted in a genomic clone in which four introns were localized. By means of fluorescence in situ hybridization the gene for human pristanoyl-CoA oxidase was mapped at chromosome position 4p15.3. We conclude that a gene for pristanoyl-CoA oxidase is present in the human genome. The gene appears to be expressed to such a low extent in liver that its mRNA cannot be detected by routine Northern-blot analysis and that its product remains undetected by standard immunoblotting or by enzyme activity measurements. We speculate that the gene may be expressed under special (e.g. certain developmental stages) conditions or in certain specialized tissues not examined thus far.

The second (enoyl-CoA hydratase) and third (3-hydroxyacyl-CoA dehydrogenase) steps of peroxisomal β-oxidation are catalysed by two separate multifunctional proteins (MFPs), MFP-1 being involved in the degradation of straight-chain fatty acids and MFP-2 in the β-oxidation of the side chain of cholesterol (bile acid synthesis). In the present study we determined which of the two MFPs is involved in the peroxisomal degradation of pristanic acid by using the synthetic analogue 2-methylpalmitic acid. The four stereoisomers of 3-hydroxy-2-methylpalmitoyl-CoA were separated by gas chromatography after hydrolysis, methylation and derivatization of the hydroxy group with ( S )-2-phenylpropionic acid, and the stereoisomers were designated I–IV according to their order of elution from the column. Purified MFP-1 dehydrated stereoisomer IV but dehydrogenated stereoisomer III, so by itself MFP-1 is not capable of converting a branched enoyl-CoA into a 3-ketoacyl-CoA. In contrast, MFP-2 dehydrated and dehydrogenated the same stereoisomer (II), so it is highly probable that MFP-2 is involved in the peroxisomal degradation of branched fatty acids and that stereoisomer II is the physiological intermediate in branched fatty acid oxidation. By analogy with the results obtained with the four stereoisomers of the bile acid intermediate varanoyl-CoA, stereoisomer II can be assigned the 3 R -hydroxy, 2 R -methyl configuration.

In this study we attempted to determine the number of 2-enoyl-CoA hydratases involved in peroxisomal β-oxidation. We therefore separated peroxisomal proteins from rat liver on several chromatographic columns and measured hydratase activities on the eluates with different substrates. The results indicate that rat liver peroxisomes contain two hydratase activities: (1) a hydratase activity associated with multifunctional protein 1 (MFP-1) (2-enoyl-CoA hydratase/Δ 3 ,Δ 2 -enoyl-CoA isomerase/ l -3-hydroxyacyl-CoA dehydrogenase) and (2) a hydratase activity associated with MFP-2 (17β-hydroxysteroid dehydrogenase/ d -3-hydroxyacyl-CoA dehydrogenase/2-enoyl-CoA hydratase). MFP-1 forms and dehydrogenates l -3-hydroxyacyl-CoA species, whereas MFP-2 forms and dehydrogenates d -3-hydroxyacyl-CoA species. A portion of MFP-2 is proteolytically cleaved, most probably in the peroxisome, into a 34 kDa 17β-hydroxysteroid dehydrogenase/ d -3-hydroxyacyl-CoA dehydrogenase and a 45 kDa d -specific 2-enoyl-CoA hydratase. Finally, the results confirm that MFP-1 is involved in the degradation of straight-chain fatty acids, whereas MFP-2 and its cleavage products seem to be involved in the degradation of the side chain of cholesterol (bile acid synthesis)

The composite trihydroxycoprostanoyl-CoA oxidase cDNA sequence, derived from overlapping clones isolated via screening of two different rat liver expression libraries, consisted of 2509 bases and contained an open reading frame of 2046 bases, encoding a protein of 681 amino acids with a calculated molecular mass of 76711 Da. The reading frame and identity of the trihydroxycoprostanoyl-CoA oxidase cDNA were confirmed by the location of various tryptic peptides, obtained from the purified enzyme, in the deduced amino acid sequence. The C-terminus (His-Lys-Met) of trihydroxycoprostanoyl-CoA oxidase did not seem to interact with the C-terminal peroxisomal targeting signal 1 (PTS1) import receptor, although the tripeptide fits the rule of conserved PTS1 variants for targeting of proteins to glycosomes of Trypanosomatidae. At the protein level, trihydroxycoprostanoyl-CoA oxidase showed 45% identical amino acids with rat palmitoyl-CoA oxidase, whereas the identity with pristanoyl-CoA oxidase was much lower (22%). Northern analysis of multiple rat tissues revealed a signal (approx. 2.6 kb) only in liver and (although much weaker) in kidney. Dot-blot analysis of total liver RNA revealed that the mRNA for trihydroxycoprostanoyl-CoA oxidase is not induced after treatment of rats with structurally unrelated peroxisome proliferators and indicates that highly similar mRNAs are present in other mammals, including man. Immunocytochemistry showed a decrease in trihydroxycoprostanoyl-CoA oxidase protein in individual liver peroxisomes (‘diluting-out effect’) after treatment of rats with bezafibrate, whereas the palmitoyl-CoA oxidase labelling was significantly increased.