Breakdown of 2-Hydroxylated Straight Chain Fatty …Breakdown of 2-Hydroxylated Straight Chain Fatty Acids via Peroxisomal 2-Hydroxyphytanoyl-CoA Lyase A REVISED PATHWAY FOR THE -OXIDATION

Breakdown of 2-Hydroxylated Straight Chain Fatty Acids viaPeroxisomal 2-Hydroxyphytanoyl-CoA LyaseA REVISED PATHWAY FOR THE �-OXIDATION OF STRAIGHT CHAIN FATTY ACIDS*

Received for publication, November 29, 2004, and in revised form, January 6, 2005Published, JBC Papers in Press, January 11, 2005, DOI 10.1074/jbc.M413362200

Veerle Foulon‡§, Mieke Sniekers‡§, Els Huysmans, Stanny Asselberghs, Vincent Mahieu,Guy P. Mannaerts, Paul P. Van Veldhoven¶, and Minne Casteels�

From the Afdeling Farmacologie, Departement Celbiologie, Katholieke Universiteit Leuven, Campus Gasthuisberg,3000 Leuven, Belgium

2-Hydroxyfatty acids, constituents of brain cerebro-sides and sulfatides, were previously reported to be de-graded by an �-oxidation system, generating fatty acidsshortened by one carbon atom. In the current study weused labeled and unlabeled 2-hydroxyoctadecanoic acidto reinvestigate the degradation of this class of lipids.Both in intact and broken cell systems formate was iden-tified as a main reaction product. Furthermore, the gen-eration of an n � 1 aldehyde was demonstrated. In per-meabilized rat hepatocytes and liver homogenates,studies on cofactor requirements revealed a depend-ence on ATP, CoA, Mg2�, thiamine pyrophosphate, andNAD�. Together with subcellular fractionation data andstudies on recombinant enzymes, this led to the follow-ing picture. In a first step, the 2-hydroxyfatty acid isactivated to an acyl-CoA; subsequently, the 2-hydroxyfatty acyl-CoA is cleaved by 2-hydroxyphytanoyl-CoAlyase, to formyl-CoA and an n � 1 aldehyde. The severeinhibition of formate generation by oxythiamin treat-ment of intact fibroblasts indicates that cleavagethrough the thiamine pyrophosphate-dependent 2-hy-droxyphytanoyl-CoA lyase is the main pathway for thedegradation of 2-hydroxyfatty acids. The latter proteinwas initially characterized as an essential enzyme in theperoxisomal �-oxidation of 3-methyl-branched fatty ac-ids such as phytanic acid. Our findings point to a newrole for peroxisomes in mammals, i.e. the breakdown of2-hydroxyfatty acids, at least the long chain 2-hydroxy-fatty acids. Most likely, the more abundant very longchain 2-hydroxyfatty acids are degraded in a similarmanner.

In mammals, 2-hydroxyfatty acids (2-OH-FA)1 are present inseveral tissues but are most abundant in brain, where theyrepresent �6% of the total fatty acids. The 2-hydroxy deriva-tives of C18 to C26 straight chain saturated and/or mono-unsaturated fatty acids appear to be present exclusively incerebrosides, cerebroside sulfates, and ceramides, most ofwhich are found in myelin (1). Furthermore, in brain cerebro-sides even more than half of the fatty acids are 2-OH-FA, andalso odd-numbered fatty acids are present in an unusuallylarge proportion (2). The ratio of 2-OH-FA to normal fatty acidsincreases during myelination, whereas the percentage of odd-numbered fatty acids continues to increase up to the age of10–15 years (3).

In 1964 Levis and Mead (4) reported for the first time theexistence of an �-oxidation system for the degradation of theC20 to C26 straight chain fatty acids of rat brain sphingolipids.It was postulated that this pathway would consist of two steps,generating first 2-hydroxy even-numbered fatty acids, and sub-sequently odd-numbered fatty acids one carbon atom shorter.Later, it was reported that in rat brain the decarboxylationreaction was performed by a microsomal enzyme and that a2-keto fatty acid, found only in small amounts, was formed asan intermediate (5). �-Oxidation of straight chain fatty acidshas also been studied in plants, and the formation of an n � 1aldehyde was reported by several authors (6–8). �-Oxidationhas also been described in yeast (9) and in protozoa (10).

More recently, the involvement of peroxisomes in the �-oxi-dation of cerebronic acid (2-hydroxytetracosanoic acid) was de-scribed. The decarboxylation of 2-hydroxytetracosanoic acidwas apparently independent of the preceding formation of anacyl-CoA and was supposed to be distinct from the �-oxidationof 3-methyl-branched fatty acids such as phytanic acid (11).The latter pathway is currently thought to proceed as follows;1) activation to a CoA ester, 2) hydroxylation of carbon 2 byphytanoyl-CoA hydroxylase (PAHX), and 3) cleavage of thehydroxylated CoA-ester by 2-hydroxyphytanoyl-CoA lyase (2-HPCL) to formyl-CoA (12) and a 2-methyl-branched fatty alde-hyde (13) in a TPP-dependent manner (14). Both PAHX and2-HPCL are peroxisomal enzymes.

According to our data PAHX does not act on straight chainfatty acids or their CoA esters (15, 16) and, hence, cannot beinvolved in the formation of 2-hydroxyfatty acids; others claim,however, that PAHX can hydroxylate straight chain acyl-CoAs

* This work was supported by Geconcerteerde Onderzoeksacties vande Vlaamse Gemeenschap Grants GOA 99/03-09 and GOA 2004/08,Fonds voor Wetenschappelijk Onderzoek-Vlaanderen Grant G.0115.02,and European Union project RDDPT, Grant QLG3-CT-2002-00696. Thecosts of publication of this article were defrayed in part by the paymentof page charges. This article must therefore be hereby marked “adver-tisement” in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

‡ Both authors contributed equally to this work.§ Supported by a fellowship from the Fonds voor Wetenschappelijk

Onderzoek-Vlaanderen.¶ To whom correspondence may be addressed: Departement Celbiolo-

gie, Afdeling Farmacologie, Faculteit Geneeskunde, Katholieke Univer-siteit Leuven, Herestraat 49, Box 601, 3000 Leuven, Belgium. Tel.:32-16-345802; Fax: 32-16-345699; E-mail: [email protected].

� To whom correspondence may be addressed: Departement Celbiolo-gie, Afdeling Farmacologie, Faculteit Geneeskunde, Katholieke Univer-siteit Leuven, Herestraat 49, Box 601, 3000 Leuven, Belgium. Tel.:32-16-345816; Fax: 32-16-345699; E-mail: [email protected].

1 The abbreviations used are: 2-OH-FA, 2-hydroxyfatty acids;2-HPCL, 2-hydroxyphytanoyl-CoA lyase; �, substrate/BSA ratio; PAHX,phytanoyl-CoA hydroxylase; TPP, thiamine pyrophosphate; Mops,4-morpholinepropanesulfonic acid; BSA, bovine serum abumin; HPLC,high performance liquid chromatography.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 280, No. 11, Issue of March 18, pp. 9802–9812, 2005© 2005 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

This paper is available on line at http://www.jbc.org9802

by guest on February 10, 2020http://w

ww

.jbc.org/D

ownloaded from

(17). Regardless of this discrepancy, a recently described fattyacid 2-hydroxylase, highly abundant in brain and encoded bythe FA2H gene (18), is likely responsible for the formation of2-hydroxyfatty acids in man.

The current study was undertaken to elucidate the degrada-tion of 2-hydroxyfatty acids and to highlight a possible role of2-HPCL in this process. Hereby, we made use of 2-hydroxyocta-decanoic acid, labeled and unlabeled, and its unlabeled CoAester. These substrates are easier to synthesize and manipu-late than the more abundant very long chain 2-hydroxyfattyacids. Although 2-hydroxyoctadecanoic acid is a less abundant2-hydroxyfatty acid, it forms a more than negligible fraction ofthe 2-hydroxyfatty acids in brain cerebrosides (4% in newborn,1% in adult brain) and sulfatides (15 and 1%, respectively) (3).

EXPERIMENTAL PROCEDURES

Fatty Acids and Derivatives—[1-14C]Hexadecanoic acid was fromPerkinElmer Life Sciences. 2-Hydroxyhexadecanoic and 2-hydroxyeico-sanoic acid were from Larodan. 2-Keto-octanoic acid was obtained fromServa. The synthesis of most fatty acids/derivatives has been describedbefore: 2-hydroxy-3-methyl[1-14C]hexadecanoic acid, labeled and unla-beled 3-methylhexadecanoic acid, 3-methylhexadecanoyl-CoA, and2-hydroxy-3-methylhexadecanoyl-CoA (19); 2-methylhexadecanoic acidand 2-methylhexadecanoyl-CoA (20); 3-hydroxy-3-methylhexadecanoyl-CoA (14), 3-hydroxy-2-methylhexadecanoyl-CoA (21); tetradecanal,2-methylpentadecanal, and heptadecanal (13).

2-Hydroxyoctadecanoic acid was purchased from Larodan or synthe-sized as follows (adapted from Sandhir et al. (11) and Ashton et al. (22)).Stearic acid treated with PBr3 was brominated in dry dichloromethanein a closed screw-capped thick-wall vial for 72 h at 70 °C (Hell-Volhard-Zelinski reaction; molar ratio acid/PBr3/Br2, 1/12/36). After removal ofthe solvent, water was carefully added to hydrolyze the 2-bromoocta-decanoylbromide, and the 2-bromooctadecanoic acid (Rf � 0.50; Silicagel 60; hexane/diethyl ether/acetic acid, 60/40/1, v/v; stearic acid Rf �0.59) was extracted into diethyl ether. The bromo group was convertedto a hydroxyl group in two steps. The dried diethyl ether extract wasdissolved in acetic acid containing potassium acetate and placed underreflux at 120 °C for 24 h. After dilution with water, the 2-acetoxyocta-decanoic acid (Rf � 0.26) was extracted into diethyl ether, dried, andhydrolyzed in 1.5 N NaOH/methanol (15/85, v/v) for 2 h under reflux.The mixture was acidified, and the formed 2-hydroxyoctadecanoic acidwas extracted into diethyl ether and further purified by preparativeTLC (Rf � 0.14). Overall yield was 35%. 2-Hydroxy[1-14C]octadecanoicacid and 2-hydroxynonadecanoic acid were synthesized in a similarmanner, starting from [1-14C]octadecanoic acid (Moravek Biochemicals,Inc.) and nonadecanoic acid (Fluka), respectively. The CoA esters of2-OH-FA were prepared by transesterification of their thiophenol ester,prepared with N,N�-dicyclohexylcarbodiimide, in 0.5 M NaHCO3/tetra-hydrofuran/ethanol (1/5/2, v/v) and purified on a C18-SPE cartridge (1 g;Supelco). Yield was 66% based on CoA input.

Animals and Cell Lines—Male Wistar rats weighing �200 g weremaintained on a standard laboratory diet and a constant light-darkcycle and were fasted overnight before sacrifice. All studies were ap-proved by the University Ethics committee. Control human skin fibro-blasts and fibroblasts from X-linked adrenoleukodystrophy patientswere kindly provided by Dr. G. Matthys (Center for Human Genetics,Leuven, Belgium). Skin fibroblasts from patients affected with multipleacyl-CoA dehydrogenase deficiency, medium chain acyl-CoA dehydro-genase deficiency, and Zellweger syndrome were provided by Dr. J. VanHove (University Hospitals, Leuven). Rhizomelic chondrodysplasiapunctata type 1 fibroblasts and rat C6 glial cells were obtained fromATCC (Manassas, VA). The preparation of fibroblasts from Pex5�/�and Pex5�/� mice has been described before (23).

Cells were cultured at 37 °C and 5% CO2 in Dulbecco’s modifiedEagle’s medium containing Glutamax (Invitrogen) and a mixture ofantibiotics and antimycotics (Invitrogen) and supplemented with 10%fetal calf serum (Invitrogen) or 2% Ultroser G (BioSepra).

Rat hepatocytes were isolated as described by Mannaerts et al. (24)and permeabilized as described by Croes et al. (25). Mouse hepatocyteswere prepared as slightly modified from Honkakoski and Negishi (26),plated in collagen-coated T25 culture flasks at 0.5 � 106 cells, and usedafter an overnight recovery period.

Preparation of Homogenates, Subcellular Fractions, and Cell Ly-sates—Homogenates of rat liver and brain were prepared in 0.25 M

sucrose containing 5 mM Mops-NaOH, pH 7.2, and 0.1% (v/v) ethanol

(homogenization medium). Subcellular fractionation of rat liver andbrain was done essentially as described previously (27) for rat liver,with some slight modifications for brain; the pellet obtained after cen-trifugation of a rat brain postmitochondrial supernatant at 13,000 � g(L-fraction) was resuspended in 0.85 M sucrose containing 5 mM Mops-NaOH, pH 7.2, and 0.1% (v/v) ethanol, overlaid with an equal volume ofhomogenization medium, and centrifuged for 45 min at 108,000 � g(Beckman SW55) to remove most of the myelin (28). Peroxisomal mem-branes were obtained after subfractionation of the peroxisomal fractionover a Nycodenz gradient and subsequent separation of matrix andmembranes by sonication and centrifugation. Marker enzymes andprotein were measured as described previously (27, 29). Cells, grown toconfluence in 175 cm2 flasks, were harvested by trypsinization, pelleted,washed twice with phosphate-buffered saline, and homogenized by soni-cation (Branson Sonifer; output 4) in 1 ml of 0.25 M Tris-HCl, pH 7.2.

Fatty Acid Oxidation by Intact and Permeabilized Cells—For meas-urement of fatty acid oxidation in intact rat hepatocytes, 1.25 � 106

cells were incubated for 10 min in 0.5 ml of Krebs-Henseleit buffer, pH7.4, containing 100 �M defatted albumin, 20 mM Hepes buffer, and 50�M concentrations of the appropriate substrate (standard conditions).Incubations with permeabilized and washed hepatocytes were startedby adding a 100-�l cell suspension (1.25 � 107 hepatocytes/ml) to 400 �lof incubation mixture containing 50 �M labeled substrate and the ap-propriate cofactors. Adherent cells (mouse hepatocytes, human andmouse fibroblasts, and rat C6 glial cells) were grown to near confluencein 25-cm2 flasks and incubated (mouse hepatocytes for 6 h (30), othercells for 24 h) with 4 �M labeled substrate in culture medium in thepresence of 0.2% Ultroser (31).

Oxidation rates in homogenates and subcellular fractions were meas-ured in 20 mM Hepes-NaOH, pH 7.2, 25 �M BSA, 50 �M radioactivesubstrate, and the appropriate cofactors. All reactions were terminatedafter 10 min by adding HClO4 to a final concentration of 2% unlessotherwise specified. The released 14CO2 and the amounts of [14C]for-mate and 14C-labeled acid-soluble material were determined as de-scribed before (19).

Enzyme Activity Measurements—The production of acyl-CoA esterswas measured in 20 mM Hepes-NaOH, pH 7.2, 12.5 �M BSA, 4 mM ATP,0.5 mM CoA, and 50 �M 1-14C-labeled substrate (250 �l final volume).Reactions were terminated by the addition of 2 ml of isopropanol/0.1 N

HCl (1/1, v/v); fatty acids were extracted with 4 ml of heptane, and analiquot of the water layer containing the CoA-esters was counted in aliquid scintillation counter (PerkinElmer Life Sciences).

When using 1-14C-labeled substrates 2-hydroxyphytanoyl-CoA lyaseactivity was quantified by measuring [14C]formate and its oxidationproduct ,14CO2 (32), since the primary 2-HPCL product, [14C]formyl-CoA, is quickly hydrolyzed to formate (12). Incubations (37 °C) wereperformed in a final volume of 250 �l containing 50 mM Tris-HCl, pH7.5, 6.6 �M BSA, 0.8 mM MgCl2, and 20 �M TPP (referred to as standardconditions) with 40 �M substrate (2-hydroxy[1-14C]octadecanoic acid,2-hydroxy-3-methyl[1-14C]hexadecanoyl-CoA, 2-hydroxy-3-methyl[1-14C]-hexadecanoic acid, 3-methyl[1-14C]hexadecanoyl-CoA, 3-methyl[1-14C]-hexadecanoic acid, [1-14C]hexadecanoyl-CoA, or [1-14C]hexadecanoicacid). For the competition experiments a substrate concentration of 10�M was used, and a 50 �M concentration of a related compound wasadded. When using unlabeled substrates (2-hydroxyoctadecanoyl-CoA,2-hydroxy-3-methylhexadecanoyl-CoA), 2-HPCL activity was quanti-fied by measuring the formation of the n � 1 aldehyde.

Lipid Analysis—The identification of the CoA esters was based onthe formation of fluorescent acyl etheno-CoA derivatives by bromoacet-aldehyde (prepared by refluxing bromodiethylacetal under acidic con-ditions and brought to pH 4.6 with sodium acetate, up to a finalconcentration of 100 mM acetate) (adapted from Larson and Graham(33)). Briefly, standard incubations (250 �l final volume) containing 50�M unlabeled substrate were stopped by the addition of 25 �l of 1 N

H2SO4. After the addition of 25 nmol of the appropriate internal stand-ard, the samples were extracted with 1.2 ml of isopropanol/heptane (4/1,v/v), and the upper phase was evaporated under N2. The residue wasreconstituted in 100 �l of water and transferred to a derivatization vial.Bromoacetaldehyde reagent (200 �l; �250 mM, pH 4.6) was added, thesamples were kept in the dark, placed at 80 °C for 15 min, and imme-diately afterward put on ice. An aliquot of the derivatization mixturewas injected onto a C18 column (Symmetry; 150 � 4.6 mm; 5 �m; 100Å, Waters) on a Waters 1525 HPLC. The acyl-CoA esters were elutedwith a gradient of acetonitrile in 0.25 M ammonium acetate buffer, pH5.0: linear gradient 10–66%, 15 min; linear gradient 66–80%, 2 min;isocratic 80%, 2 min; linear gradient 80–10%, 2 min; isocratic 10%, 6min. Detection was performed on a Waters 2475 fluorescence detector(excitation 230 nm; emission 420 nm).

�-Oxidation of Straight Chain Fatty Acids 9803

by guest on February 10, 2020http://w

ww

.jbc.org/D

ownloaded from

For the identification and analysis of aldehydes, reactions werestopped with 125 �l of 2 N HCl. After the addition of 2.5 nmol of theappropriate internal standard and 125 �l of a solution of 2,4-dinitro-phenylhydrazine (3.75 mg in 5 ml of 2 N HCl), the samples wereincubated for 30 min at 50 °C (in this and subsequent steps sampleswere protected from light). After adding 0.5 ml of methanol, the hydra-zones were extracted into 2 ml of hexane. 1.75 ml of the upper phasewas evaporated, the residue was dissolved in 88 �l of acetonitrile, and50 �l (1.25 nmol of internal standard) was injected on a C18 column(Symmetry 150 � 4.6 mm; 5 �m, 100 Å Waters; Waters 1525 HPLCsystem) eluted with acetonitrile under isocratic conditions. Detectionwas done at 360 nm (Waters 484 tunable absorbance detector).

Generation and Purification of Recombinant Hs 2-HPCL—Human2-HPCL cDNA was amplified from a human liver cDNA library withprimers Hs 2-HPCL-F8 (5�-CGCGGATCCGATGCCGGACAGTAAACT-TCGC) and Hs 2-HPCL-R8 (5�-AATGCATGCTTACATATTAGAGCGG-GTC), and after digestion with BamHI and SphI, the correspondingPCR product was subcloned in pJR233 (34) (construct yVF3). PlasmidyVF3 was transformed into competent Saccharomyces cerevisiae CB80cells. Transformants were selected and grown on minimal essentialmedium containing 0.67% (w/v) yeast nitrogen base without aminoacids, 2% (w/v) glucose, and a supplement of bases and amino acids(20–150 �g/ml) as required (Sc-ura). Positive colonies were grown at

33 °C in 50 ml Sc-ura containing 0.5% glucose as the sole carbon source.After 12–15 h, cells from 15 ml of culture were harvested by centrifu-gation, washed, and broken with glass beads in 400 �l of lysis buffer (50mM potassium phosphate buffer, pH 7.4, containing 1 mM EDTA, 10 mM

MgCl2, 1 mM TPP, and a mix of protease inhibitors). Cell debris wasremoved by centrifugation, and lyase activity was measured on 50-�lsamples.

To generate a polyhistidine fusion product of Hs 2-HPCL, two oligos(Hs 2-HPCL-H1, 5�-CAAGATGGGCATCATCATCATCATCATCGG,and Hs 2-HPCL-H2, 5�-ATCCCGATGATGATGATGATGATGCCCCAT-CTTGGTAC) were allowed to hybridize to generate a small linker. Theresulting adaptor sequence was cloned between the KpnI-BamHI sitesof yVF3 (yVF7), and S. cerevisiae CB80 cells, transformed with yVF7,were grown for 18 h in 400 ml of Sc-ura containing 0.5% (w/v) glucose.The fusion protein was purified from cell lysate (prepared as describedbefore in 8 ml of lysis buffer) on nickel nitrilotriacetic acid-agaroseessentially as described for the purification of phytanoyl-CoA hydroxy-lase (16). Analysis of the purified fraction by SDS-PAGE (12% poly-acrylamide, w/v) and subsequent immunoblotting with anti-His anti-body (Clontech) revealed one single polyhistidine-tagged protein with amolecular mass of �63 kDa. The yield was 77–400 �g of protein with aspecific activity of about 26 milliunits/mg of protein (substrate, 40 �M

2-hydroxy-3-methyl[1-14C]hexadecanoyl-CoA).

TABLE IOxidation of 2-hydroxyoctadecanoic acid in intact cells

CO2, formate, and acid-soluble material (ASM, including formate, which is retained in the water phase) production was measured in theindicated cells incubated with 2-hydroxy[1-14C]octadecanoic acid, as described under “Experimental procedures.” Results are expressed as themean � S.E. of two or three experiments.

Cell type Activity CO2 Formate ASM

Intact rat hepatocytes (n � 2) nmol/108 cells/min 6.36 0.67 2.11Confluent mouse hepatocytes (n � 3) nmol/mg protein/6 h 2.68 � 0.48 0.11 � 0.10 3.20 � 1.30Confluent fibroblasts (n � 3) nmol/mg protein/24 h 2.60 � 0.56 17.43 � 4.09 16.56 � 1.30Confluent C6 glial cells (n � 2) nmol/mg protein/24 h 0.01 5.49 6.24

TABLE IIOxidation of 2-hydroxyoctadecanoic acid in human and murine fibroblasts

Production of CO2 and formate was measured in fibroblasts incubated with 4 �M 2-hydroxy[1-14C]octadecanoic acid, as described under“Experimental Procedures.” Results were calculated as the mean values of two determinations per cell line. MADD, multiple acyl-CoA dehydro-genase deficiency; MCAD, medium chain acyl-CoA dehydrogenase deficiency; RCDP, rhizomelic chondrodysplasia punctata; X-ALD, X-linkedadrenoleukodystrophy.

CO2 Formate CO2 � formate

nmol/mg protein/24 h

HumanControl fibroblasts (n � 3) 2.60 � 0.56 17.43 � 4.09 20.03 � 4.00MADD fibroblasts (n � 1) 2.60 17.54 20.15MCAD fibroblasts (n � 1) 2.70 13.98 16.68X-ALD fibroblasts (n � 3) 3.13 � 1.30 19.90 � 2.91 23.03 � 1.52RCDP fibroblasts (n � 1) 0.67 16.10 16.80Zellweger fibroblasts (n � 1) 1.05 10.34 11.39

MousePex5�/� fibroblasts (n � 1) 2.35 9.61 11.96Pex5�/� fibroblasts (n � 1) 1.36 4.11 5.48

FIG. 1. Breakdown products of 2-hydroxy[1-14C]octadecanoic acid, 3-methyl[1-14C]hexadecanoic acid, and [1-14C]hexadecanoicacid in isolated intact rat hepatocytes. Isolated rat hepatocytes were incubated with the indicated fatty acids (200 �M for 2-hydroxy[1-14C]octadecanoic acid and 3-methyl[1-14C]hexadecanoic acid; 50 �M for [1-14C]hexadecanoic acid) in the presence of increasing concentrations ofunlabeled formate. Incubations were stopped after 10 min with HClO4, and the amount of 14CO2 (white bars), [14C]formate (black bars), and [14C]acid-soluble material (gray bars) was measured.

�-Oxidation of Straight Chain Fatty Acids9804

by guest on February 10, 2020http://w

ww

.jbc.org/D

ownloaded from

RESULTS

Degradation of 2-OH-FA in Intact and PermeabilizedCells; Indication for a Role of Peroxisomes and 2-HPCL

Analysis of the medium of control human fibroblasts and ratC6 glial cells incubated with 2-hydroxy[1-14C]octadecanoic acidfor different possible labeled degradation products revealedformate as the major acid-soluble oxidation product (Table I).In intact isolated rat and cultured mouse hepatocytes, on theother hand, the amount of acid-soluble material was markedlyhigher than the amount of formate generated (Table I), sug-gesting that in these cells, apart from formate, another metab-olite was generated as the primary reaction product or thatformate/CO2 was subsequently converted into another acid-soluble metabolite. When intact rat hepatocytes were incu-bated in the presence of increasing concentrations of unlabeledformate, the production of 14CO2 from 2-hydroxy[1-14C]octadec-anoic acid decreased dramatically (Fig. 1). This decrease was

accompanied by a compensatory increase in the generation of[14C]formate (Fig. 1), as was shown earlier for 3-methylhexa-decanoic acid (Ref. 35; see also Fig. 1), whereas the �-oxidationof straight chain fatty acids was unaffected throughout thewhole formate concentration range (Fig. 1). Hence, these re-sults support the contention that 2-hydroxyfatty acids areshortened by a process resembling the �-oxidation of phytanicacid whereby not CO2, but formyl-CoA/formate is the primaryoxidation product that is subsequently converted into CO2 andpossibly other metabolites.

To obtain a first insight in the organelle(s)/enzymes involved,the oxidation of 2-hydroxy[1-14C]octadecanoic acid was ana-lyzed in fibroblasts from patients affected with various fattyacid oxidation disorders. In patients with the mitochondrialfatty acid oxidation disorders medium chain acyl-CoA dehydro-genase deficiency or multiple acyl-CoA dehydrogenase defi-ciency, characterized by a deficiency in the medium chain or

TABLE IIICofactor requirements for oxidation of 2-hydroxyoctadecanoic acid in permeabilized hepatocytes

CO2 and formate production was measured in permeabilized rat hepatocytes incubated with 50 �M 2-hydroxy[1-14C]octadecanoic acid. Concen-trations of the cofactors tested: ATP, 4 mM; CoA, 0.5 mM; Mg2�, 2.4 mM; TPP, 0.02 mM; NAD�, 2 mM. Results are expressed as % of the ratesobserved under standard conditions (ATP, CoA, Mg2�, and TPP; CO2, 5.10 nmol/108 cells/min, and formate, 25.48 nmol/108 cells/min) and are meanvalues of two experiments, except for ATP, CoA, Mg2�; ATP, CoA, Mg2�, NAD�; and NAD�, which was a single measurement.

Additions CO2 Formate CO2 � formate

% of standard conditions

None 13.26 11.83 11.83ATP, Mg2�, TPP 100.08 16.49 30.22CoA, Mg2�, TPP 7.63 6.24 6.64ATP, CoA, Mg2� 59.71 101.71 71.56ATP, CoA, Mg2�, TPP 100 100 100ATP, CoA, Mg2�, TPP, NAD� 141.78 118.23 123.64ATP, CoA, Mg2�, NAD� 70.28 151.11 93.08NAD� 23.99 78.45 39.35

FIG. 2. Identification of 2-hydroyx-octadecanoyl-CoA (A–C) and hepta-decanal (D–F) in incubations of ratliver homogenate with 2-hydroxy-octadecanoic acid and 2-hydroxy-octadecanoyl-CoA, respectively. HPLCanalysis of acyl etheno-CoA esters (leftpanel) and aldehydes (right panel) formedduring incubations of rat liver homogenatewith 2-hydroxyoctadecanoic acid (100 �M)and 2-hydroxyoctadecanoyl-CoA (40 �M),respectively. Homogenates were incubatedfor 0 min (A and D) and 10 min (B and E),and reaction products were derivatized af-ter the addition of the appropriate internalstandard (IS) as described under “Experi-mental Procedures.” Elution of standardsis shown in C and F.

�-Oxidation of Straight Chain Fatty Acids 9805

by guest on February 10, 2020http://w

ww

.jbc.org/D

ownloaded from

multiple acyl-CoA dehydrogenases, respectively, no decrease inoxidation rates was observed (Table II). Hence, these dataexclude a significant role of these mitochondrial dehydroge-nases in the degradation of 2-hydroxylated straight chain fattyacids. Furthermore the oxidation rates of the 2-hydroxyfattyacids were not reduced either in cells from X-linked adrenoleu-kodystrophy patients, characterized by a deficiency in the per-oxisomal activation/import of very long chain fatty acids, or inthose from a rhizomelic chondrodysplasia punctata type 1 pa-tient with a proven deficiency of Pex7p, the import receptor forperoxisomal matrix proteins containing a peroxisomal target-ing signal-2 (Table II). Only in fibroblasts from a patient withZellweger syndrome, with a demonstrated defect in the perox-isomal import of proteins containing a peroxisomal targetingsignal-1, was a decrease (43%) in oxidation rate noticed. Asimilar decrease was seen in fibroblasts from Pex5�/� mice, amouse model for Zellweger syndrome (Table II). These datasuggest that peroxisomes are important and that peroxisomalenzymes with a peroxisomal targeting signal-1 signal are prob-ably involved in the degradation of 2-OH fatty acids. The ad-dition to the culture medium of etomoxir or tetradecylglycidicacid, inhibitors of carnitine palmitoyl-CoA transferase and,hence, of fatty acyl-CoA entry into the mitochondria, did notsignificantly affect the oxidation of 2-hydroxy[1-14C]octade-canoic acid (results not shown), another indication that mito-chondria are likely not involved.

To study the necessary cofactors, rat hepatocytes, permeabi-lized with �-toxin to deplete the cytosolic cofactors (25), wereincubated with different sets of cofactors. In these hepatocytesthe oxidation of 2-hydroxy[1-14C]octadecanoic acid was strictlydependent on ATP, Mg2�, and CoA and was enhanced by TPP(Table III). Extra addition of NAD� resulted in a further 20%increase of total oxidation.

The dependence on TPP was further investigated by cultur-

ing fibroblasts and C6 glial cells in the presence of the thiamineantimetabolite oxythiamin. The addition of this compound (1mM) to the culture medium for several cell divisions reducedthe overall oxidation rates of 2-hydroxyoctadecanoic acid by 44,39, and 60% in control fibroblasts, rhizomelic chondrodysplasiapunctata fibroblasts, and C6 glial cells, respectively. Thesedata confirm the involvement of a TPP-dependent enzyme andare similar to the effect of oxythiamin on �-oxidation of 3-meth-ylhexadecanoic acid. The addition of oxythiamin had no effecton the �-oxidation of long chain fatty acids.2

The finding that ATP and CoA are essential suggests thatthe oxidation of 2-hydroxyfatty acids involves an activationreaction. The dependence of the pathway on TPP points to areaction similar to that of the TPP-dependent cleavage of 2-hy-droxy-3-methylacyl-CoA esters during the �-oxidation of3-methyl-branched fatty acids.

Degradation of 2-OH-FA in Homogenates; Identification ofIntermediates and End Products, Subcellular

Localization, and Enzymes Involved

In rat liver homogenates similar results as those in perme-abilized hepatocytes were obtained (results not shown): break-down of 2-hydroxy[1-14C]octadecanoic acid in whole rat liverhomogenates was dependent on the presence of CoA, ATP,Mg2�, and TPP. Although the oxidation rates in the presence ofADP were about 65% of those obtained with ATP, no oxidationwas detected upon the addition of GTP or AMP. Furthermore,CoA could not be replaced by desulfo-CoA or dephospho-CoA.

In homogenates the main product was formate (�90% oftotal products). The addition of KCN did not affect the oxida-

2 V. Foulon, M. Sniekers, M. Casteels, and P. P. Van Veldhoven,manuscript in preparation.

FIG. 3. Kinetics of the activation of 2-hydroxyoctadecanoic acid in rat liver homogenates. Generation of labeled acyl-CoA wasmeasured in incubations of 2-hydroxy[1-14C]octadecanoic acid (50 �M in A and B; varying substrate concentrations are in C and D) with rat liverhomogenates. Reactions were stopped after 3 min except in panel A. The velocity (V, expressed in nmol/min/g) versus substrate concentration (S,expressed in �M) curve, as shown in panel C, was linearly transformed according to the method of Lineweaver-Burk as shown in D.

�-Oxidation of Straight Chain Fatty Acids9806

by guest on February 10, 2020http://w

ww

.jbc.org/D

ownloaded from

tion of 2-hydroxy[1-14C]octadecanoic acid (results not shown),again ruling out mitochondria as a key player.

Identification of Intermediates and End Products of 2-Hy-droxyoctadecanoic Acid Oxidation—When rat liver homoge-nates were incubated with 2-hydroxy[1-14C]octadecanoic acidin the presence of ATP, CoA, and Mg2�, labeled acyl-CoA esterswere generated. HPLC analysis of the etheno-CoA esters re-vealed a peak with significant fluorescence at 18.7 min coelut-

ing with the 2-hydroxyoctadecanoyl-CoA standard (Fig. 2,A–C). The amount of 2-hydroxyoctadecanoyl-CoA (0.630nmol/mg of protein/min) based on the internal standard, agreedwell with the total amount of labeled CoA ester formed, basedon extraction procedures (0.746 nmol/mg of protein/min) (meanof two measurements). Additionally, the generation of a 2-hy-droxyoctadecanoyl-CoA intermediate was also demonstrated inrat brain homogenates and in lysates from rat C6 glial cells.

FIG. 4. Subcellular distribution of acyl-CoA synthetase activity (Fractionation A) and the production of heptadecanal from2-hydroxyoctadecanoyl-CoA (Fractionation B) in rat liver. In two separate experiments fresh liver homogenates were fractionated bydifferential centrifugation into nuclear (N), heavy mitochondrial (M), light mitochondrial (L), microsomal (P), and cytosolic (S) fractions.Fractionation A, fractions were incubated with 50 �M 2-hydroxy[1-14C]octadecanoic acid, and the production of the CoA ester (synthetase) wasmeasured as described under “Experimental Procedures.” Fractionation B, the production of heptadecanal from 2-hydroxyoctadecanoyl-CoA (lyase)was measured in each fraction as described under “Experimental Procedures.” Marker enzymes were determined in each fraction: carboxylesterase(endoplasmic reticulum), glutamate dehydrogenase (GDH; mitochondria), catalase (peroxisomal matrix), and lactate dehydrogenase (LDH;cytosol). Results are expressed as relative specific activities versus percentage of total protein. Relative specific activity is defined as the percentageof total recovered activity present in a particular fraction divided by the corresponding percentage of protein. Recovery for synthetase activity was89%; recovery for heptadecanal production was 109%; recoveries for marker enzymes were between 86 and 119%.

�-Oxidation of Straight Chain Fatty Acids 9807

by guest on February 10, 2020http://w

ww

.jbc.org/D

ownloaded from

In the presence of Mg2� and TPP, heptadecanal was formedby rat liver homogenates (Fig. 2, D–F) incubated with 2-hy-droxyoctadecanoyl-CoA at rates of 0.014 nmol/mg of protein/min (n � 2). Overall, these data indicate that 2-hydroxyacyl-CoA esters are intermediates in the degradation of 2-hydroxystraight chain fatty acids. Unfortunately, due to the presence ofinterfering peaks, the formation of the n � 1 aldehyde could notbe demonstrated in incubations with brain homogenate.

Characterization of the acyl-CoA Synthetase, Catalyzing theFormation of 2-Hydroxyacyl-CoA Esters—In rat liver homoge-nates the activation of 2-hydroxyoctadecanoic acid was shownto be linear for up to 3 min (Fig. 3A). The formation of 2-hy-droxyoctadecanoyl-CoA reached an optimum at a substrate/albumin ratio (�) of 4 (Fig. 3B). Measuring the generation of2-hydroxyoctadecanoyl-CoA at increasing substrate concentra-tions (at � � 4) resulted in a plateau from 100 �M onward;transformation of the data according to Lineweaver-Burk al-lowed the calculation of an apparent Km of 19.5 �M (Fig. 3,C–D). Subsequent experiments in rat liver homogenates andsubcellular fractions were, therefore, performed at a substrateconcentration of 50 �M and a � ratio of 4 and were terminatedafter 3 min. For measurement of synthetase activity in lysatesof rat C6 glial cells, optimum conditions were 100 �M substrate,a � ratio of 2, and an incubation time of 10 min.

The activation of 2-hydroxyoctadecanoic acid in subcellularfractions of rat liver showed a bimodal distribution coincidingwith the light mitochondrial fraction, enriched in peroxisomes,and the microsomal fraction (Fig. 4, Fractionation A). In thelight mitochondrial fraction the activity appeared to be pre-dominantly membrane-associated. Further separation of thelight mitochondrial fraction on a Nycodenz gradient revealedan association with peroxisomes (data not shown). In rat braina bimodal distribution was observed as well, the microsomallocalization being most prominent (results not shown).

To further characterize the acyl-CoA synthetase responsiblefor the activation of 2-hydroxy straight chain fatty acids, incu-bations of lysates from C6 glial cells with 2-hydroxy[1-14C]octa-decanoic acid were performed with and without the addition ofa range of unlabeled fatty acids. Whereas 3-methylhexade-canoic acid, 2-methylhexadecanoic acid, eicosatetraenoic, andtetracosanoic acid had no major inhibitory effect on the activa-tion of 2-hydroxyoctadecanoic acid, hexadecanoic acid and oc-tadecanoic acid markedly reduced the activation rates (resultsnot shown), indicating that the substrate spectrum of the acti-vating enzyme involved certainly covers long chain fatty acids.

Identification of 2-HPCL as the Enzyme Catalyzing the Cleav-age Reaction—Investigation of the formation of heptadecanalfrom 2-hydroxyoctadecanoyl-CoA in subcellular fractions of ratliver indicated that the responsible enzyme has a peroxisomallocalization (Fig. 4, Fractionation B). Both this subcellular dis-tribution and the marked thiamine dependence point toward2-HPCL as the enzyme catalyzing the cleavage reaction.

In the presence of ATP, CoA, Mg2�, and TPP, incubations ofthe purified recombinant polyhistidine-fused human 2-HPCLwith 2-hydroxyoctadecanoyl-CoA yielded heptadecanal, inanalogy with the formation of 2-methylpentadecanal from2-hydroxy-3-methylhexadecanoyl-CoA. The reaction kinetics ofrecombinant 2-HPCL toward these substrates were compared bymeasuring the generation of 2-methylpentadecanal and heptade-canal. The cleavage rate of the branched substrate was linear forup to 10 min, whereas for the straight chain substrate it waslinear for up to at least 30 min (results not shown). At increasingsubstrate concentrations a plateau from 40 �M onward wasreached for 2-hydroxy-3-methylhexadecanoyl-CoA, whereas for2-hydroxyoctadecanoyl-CoA a plateau was reached from 20 �M

onward (Fig. 5, A and B); apparent Km values of 15.8 and 6.3 �M,

respectively, could be calculated. Furthermore, the amount ofaldehyde generated with both substrates increased linearly withthe amount of recombinant protein up to at least 10 �g.

The substrate spectrum of recombinant, polyhistidine-fusedhuman 2-HPCL was further investigated using labeled sub-strates. Incubation of the recombinant 2-HPCL with 2-hy-droxy-3-methyl[1-14C]hexadecanoyl-CoA under standard con-ditions resulted in production of [14C]formyl-CoA/[14C]formateat rates of 26.06 � 1.17 milliunits/mg of protein (mean � S.E.;n � 4), but the lyase showed no activity toward [1-14C]-labeled3-methylhexadecanoyl-CoA, hexadecanoyl-CoA, 3-hydroxyhexa-decanoyl-CoA, octadecanoyl-CoA, or toward 2-hydroxy-3-meth-ylhexadecanoic acid, 3-methylhexadecanoic acid, hexadecanoicacid, and 2-hydroxyhexadecanoic acid. Furthermore, the pro-duction of [14C]formyl-CoA/[14C]formate from 2-hydroxy-3-methyl[1-14C]hexadecanoyl-CoA was not reduced in the presenceof 2-keto-octanoate, 2-hydroxyhexadecanoic acid, 2-methylhexa-decanoyl-CoA, 2-methylhexadecanoic acid, 3-methylhexade-canoyl-CoA, 3-hydroxy-3-methylhexadecanoyl-CoA, or 3-hy-droxy-2-methylhexadecanoyl-CoA (Fig. 6). The addition of 50 �M

2-hydroxyhexadecanoyl-CoA or 2-hydroxyoctadecanoyl-CoA toincubations with 10 �M 2-hydroxy-3-methyl[1-14C]hexadecanoyl-CoA, however, reduced the cleavage rates to a major extent (Fig.6), indicating that 2-hydroxy straight chain acyl-CoA esters com-pete with 2-hydroxy-3-methyl-branched acyl-CoAs for cleavageby 2-HPCL.

DISCUSSION

In the current study 2-hydroxyoctadecanoic acid was used asa substrate to document the degradation of 2-hydroxy straight

FIG. 5. Characterization of the cleavage reaction of 2-hydroxy-3-methylhexadecanoyl-CoA and 2-hydroxyoctadecanoyl-CoA byrecombinant human 2-HPCL. Aldehyde production was measured inincubations of 2-hydroxy-3-methylhexadecanoyl-CoA (A) or 2-hy-droxyoctadecanoyl-CoA (B) (varying substrate concentrations; � � 6)with recombinant human 2-HPCL (2 �g of protein). Reactions werestopped after 10 min. The velocity (V, expressed in nmol/assay) versussubstrate concentration (S, expressed in �M) curves were linearly trans-formed according to the method of Lineweaver-Burk.

�-Oxidation of Straight Chain Fatty Acids9808

by guest on February 10, 2020http://w

ww

.jbc.org/D

ownloaded from

chain fatty acids. This compound is less abundant than 2-hy-droxytetracosanoic acid (cerebronic acid) in brain cerebrosidesand sulfatides (3), but it offers advantages over the very longchain substrates both in synthesis and solubility. The fact that2-hydroxy[1-14C]octadecanoic acid gives rise to the generationof labeled formate in intact as well as in broken cell systemswas a first indication that 2-hydroxyfatty acids can be de-graded via a process resembling the �-oxidation of 3-methyl-branched fatty acids. This was further supported by the re-quirement of ATP, Mg2�, CoA, and TPP in permeabilized cellsand by the observation that 2-hydroxyoctadecanoyl-CoA servesas a substrate for recombinant 2-HPCL, thereby being con-verted to heptadecanal. The in vivo significance of the 2-HPCL-dependent pathway was highlighted by the marked inhibitoryeffect of the thiamine antimetabolite oxythiamin in culturedcells.

Overall, our data led to the following picture (Fig. 7). A2-hydroxy fatty acid is first activated to its CoA ester in anATP/Mg2�/CoA-dependent reaction. The generated 2-hy-droxyacyl-CoA is then cleaved by 2-HPCL into an n � 1 alde-hyde and formyl-CoA. The latter is converted to formate andsubsequently to CO2. As for the 2-methyl n � 1 aldehydesformed during �-oxidation of 3-methyl-branched fatty acids,the n � 1 aldehyde is most probably dehydrogenated to thecorresponding odd-numbered fatty acid, which can then befurther degraded via �-oxidation. Although we cannot excludethat 2-hydroxyfatty acids can be degraded by other pathways,the inhibition by oxythiamin suggests that the main pathway,at least in the cells and tissues we studied, is via the TPP-de-pendent 2-HPCL. In addition, we did not find evidence for theformation of a 2-keto fatty acid, formed as an intermediate viaan �-OH acid oxidase (36). In kidney, but not in liver, anL-2-hydroxy long chain acid oxidase has been described (37).

2-Hydroxyacyl-CoAs as substrate for �-oxidation have notbeen described in plants or bacteria, but in plants n � 1aldehydes as a product of �-oxidation have been reported (6).The relevance of 2-HPCL to the �-oxidation of 2-hydroxystraight chain fatty acids in these organisms is unknownat present.

With regard to the stereochemistry of the process we studied,not all details are yet available. The naturally occurring 2-hy-droxyfatty acids possess a D-configuration, but so far racemic2-hydroxyoctadecanoic acid has been used. The finding that in

C6 glial cells within 10 h �70% of the substrate was oxidizedsuggests that racemization can occur or that both isomers aredegraded. Because 2-HPCL is able to cleave all four possibleisomers of 3-methyl-2-hydroxyhexadecanoyl-CoA (38), the lat-ter possibility is more likely.

Measurement of the activation of 2-hydroxyoctadecanoic acidin subcellular fractions from rat liver revealed that the activityis associated both with the peroxisomal membrane and withthe endoplasmic reticulum. This distribution together with thefact that, among the fatty acids tested, only hexadecanoic acidand octadecanoic acid competitively inhibited the activation of2-hydroxyoctadecanoic acid, point toward the long chain acyl-CoA synthetase as the activating enzyme. However, because itis well known that the substrate spectra of long chain and verylong chain acyl-CoA synthetases overlap (39, 40), no definitiveconclusions can be drawn as to the specific enzyme involved.Moreover, it might well be that the chain length of the sub-strate determines the involved isoform. Hence, long chain 2-hy-droxyfatty acids (e.g. 2-hydroxyoctadecanoic acid) might be ac-tivated by a long chain acyl-CoA synthetase and very longchain 2-hydroxyfatty acids might be activated by a very longchain acyl-CoA synthetase.

The contention of 2-HPCL being involved in the degradationof 2-hydroxyfatty acids was strengthened by the increased rel-ative specific activities of heptadecanal formation from 2-hy-droxyoctadecanoyl-CoA in subcellular fractions enriched inperoxisomes, the organelles that harbor 2-HPCL (14). Themarked inhibition of 2-hydroxyoctadecanoic acid degradationby the addition of oxythiamin to cultured cells further corrob-orates the thiamine dependence of the pathway in the intactcell. The finding that in rat liver homogenates and permeabi-lized hepatocyte stimulation by the addition of TPP was onlylimited can be explained by the fact that in these preparationsTPP is still partly bound to the enzyme. As shown earlier,2-HPCL only gradually loses its activity during purificationdue to the release of bound TPP (14).

The fact that 2-hydroxyfatty acid degradation is decreasedby only 50% in fibroblasts from Zellweger patients andPex5�/� mice, an animal model for Zellweger syndrome (23),might suggest that 2-HPCL, a peroxisomal matrix enzyme (14),remains partially active in the cytosol under conditions whereperoxisomal protein import is deficient. This is in contrast withmost other peroxisomal matrix enzymes, which are labile in the

FIG. 6. Effect of competitors on the production of 14C-formate/14C-formyl-CoA from 2-hydroxy-3-methyl[1-14C]hexadecanoyl-CoAin incubations with recombinant human 2-HPCL. The production of [14C]formate/[14C]formyl-CoA from 2-hydroxy-3-methyl[1-14C]hexade-canoyl-CoA (10 �M) with 4.7 �g of recombinant human 2-HPCL was measured with and without the addition of 50 �M unlabeled substrates. PanelA, 2-hydroxyhexadecanoyl-CoA, 2-hydroxyhexadecanoic acid, 2-methylhexadecanoyl-CoA, 2-methylhexadecanoic acid, or 2-keto-octanoate meas-ured in the presence of 1.67 �M BSA. Panel B, 2-hydroxy-3-methylhexadecanoyl-CoA, 3-methylhexadecanoyl-CoA, 2-hydroxyhexadecanoyl-CoA,2-hydroxyoctadecanoyl-CoA, 3-hydroxy-3-methylhexadecanoyl-CoA, or 3-hydroxy-2-methylhexadecanoyl-CoA measured in the presence of 10 �M

BSA. Results, calculated from single determinations, are represented as the percentage of control activity.

�-Oxidation of Straight Chain Fatty Acids 9809

by guest on February 10, 2020http://w

ww

.jbc.org/D

ownloaded from

cytosol when their import is impaired, but in agreement withour previous measurements of 2-HPCL activity in liver homo-genates of Zellweger patients and Pex5�/� mice using 2-hy-droxy-3-methylhexadecanoyl-CoA as substrate (41). Whetherthe partial impairment of 2-hydroxyfatty acid degradationleads to an accumulation of 2-hydroxyfatty acids in tissuesfrom Zellweger patients is unknown. Moderate increases incerebrosides and/or gangliosides have been described in fibro-blasts or CHO cells lacking peroxisomes, but their 2-hydroxy-fatty acid content was not documented (42, 43). When consid-ering the identification of this second substrate for2-hydroxyphytanoyl-CoA lyase, one could expect that an iso-lated deficiency of 2-HPCL, although hitherto not identified,might lead not only to an impaired �-oxidation of phytanic acidbut also to an accumulation of 2-hydroxyfatty acids.

The identification of 2-HPCL as the cleavage enzyme in thedegradation of 2-hydroxyfatty acids is also of interest for itsreaction mechanism. It demonstrates that the methyl branchat position 3 is not necessary but that both the hydroxy group

at position 2 and the CoA moiety are important. A 2-hydroxycarboxyl compound (instead of a 2-keto compound) is an un-usual substrate for TPP-dependent enzymes. Only one otherenzyme, N2-(2-carboxyethyl)-L-arginine synthase, catalyzingthe condensation of L-arginine and D-glyceraldehyde in thepresence of TPP (the first step in the biosynthesis of clavulanicacid), has been reported to show activity toward compoundswith a hydroxy group at position 2 (44). In all TPP-dependentcleavage reactions described so far decarboxylation involvesthe activation of the C2-H of the thiazole ring of TPP to anintermediate carbanion. This is followed by a nucleophilic at-tack at the carbonyl atom of the substrate (carbon 2) (45). Mostlikely, the formation of a carbanion is also required for thecleavage of 2-hydroxy-(3-methyl)acyl-CoA esters by 2-HPCL.However, this carbanion will then attack carbon 1 of the sub-strate, which is highly reactive due to the presence of thethioester bond. Ultimately, this will lead to the formation offormyl-CoA and an n � 1 fatty aldehyde (Fig. 8).

The observation that extra addition of NAD� to the incuba-

FIG. 7. �-Oxidation of 2-hydroxyfatty acids and of 3-methyl-branched fatty acids. The scheme represents the proposed �-oxidationpathway for 2-hydroxy straight chain fatty acids next to the �-oxidation pathway for 3-methyl-branched fatty acids. *FA2H, fatty acid 2-hydrox-ylase (18).

�-Oxidation of Straight Chain Fatty Acids9810

by guest on February 10, 2020http://w

ww

.jbc.org/D

ownloaded from

tion mixture resulted in a further 20% increase of total oxida-tion (Table III) might be explained by a diminished productinhibition through NAD�-dependent dehydrogenation of the gen-erated aldehyde. Fatty aldehyde dehydrogenase activity is foundin peroxisomes and endoplasmic reticula (13, 46). It is currentlyunknown whether fatty aldehydes generated in the peroxisomeare dehydrogenated exclusively by the peroxisomal dehydrogen-ase or also partially by the endoplasmic reticulum enzyme.

In summary, our present work shows that (the bulk of)2-hydroxyfatty acids undergo an initial degradation that ap-parently shares three reactions (activation, cleavage of theC1-C2 bond, aldehyde dehydrogenation) with the �-oxidationsequence of 3-methyl-branched fatty acids, giving rise to n � 1odd-numbered fatty acids, which can subsequently be degradedvia �-oxidation. It is of interest to note that the second enzymeof the �-oxidation sequence of 3-methyl-branched fatty acids(the peroxisomal PAHX, which hydroxylates 3-methyl-acyl-CoAs to 2-hydroxy-3-methyl-acyl-CoAs) is likely not involved inthe synthesis of 2-hydroxy straight chain fatty acids (see theIntroduction). One of the underlying reasons might be that an(imaginary) hydroxylation of straight chain acyl-CoAs by

PAHX would lead to the immediate further breakdown of the2-hydroxylated acyl-CoAs by 2-HPCL within the peroxisome.

To control the levels of 2-hydroxyfatty acids in brain cere-brosides and sulfatides, supposed to play a role in myeliniza-tion, a strategy relying on different sets of enzymes for theirsynthesis and degradation, located at different subcellularsites, might be more beneficial. Moreover, as was reported inolder literature and recently discussed by Alderson et al. (18),hydroxylation of straight chain fatty acids might occur onlyafter incorporation in sphingolipids, which would further ruleout peroxisomes as a key player in this hydroxylation process.In this context it is of interest to note that 2-hydroxyoctade-canoic acid was hardly incorporated into complex lipids whengiven to intact cells.3 Overall, the �-oxidation of straight chainfatty acids, as has been described especially for brain, appearsto proceed as follows; 1) hydroxylation of the fatty acid by afatty acid 2-hydroxylase (see the Introduction), 2) activation ofthe 2-hydroxyfatty acid to a 2-hydroxyacyl-CoA, 3) cleavage of

3 M. Sniekers, V. Foulon, P. P. Van Veldhoven, and M. Casteels,unpublished data.

FIG. 8. Generation of a carbanion inenzyme-bound TPP and the proposedreaction mechanism for 2-HPCL. Re-action of 2-HPCL with the substrate 2-hy-droxyoctadecanoyl-CoA involves a nucleo-philic attack of the C2-atom of TPP at thecarbon 1 of the substrate. A key functionherein is the interaction of a conservedglutamate with the N1� atom of the coen-zyme, resulting in an increased basicity ofits 4� amino group, which facilitates thedeprotonation of the C2 (Ref. 45 and ref-erences cited herein).

�-Oxidation of Straight Chain Fatty Acids 9811

by guest on February 10, 2020http://w

ww

.jbc.org/D

ownloaded from

the CoA ester into formyl-CoA and an n � 1 fatty aldehyde, and4) dehydrogenation of the aldehyde to the corresponding n � 1odd-numbered fatty acid.

Acknowledgments—We thank Dr. Myriam Baes for providingPex5�/� mouse fibroblasts, Ruud Dirkx for providing mouse hepato-cytes, and Wendy Geens and Luc Govaert for expert technicalassistance.

REFERENCES

1. Kishimoto, Y., and Radin, N. S. (1963) J. Lipid Res. 58, 139–1432. Hajra, A. K., and Radin, N. S. (1963) J. Lipid Res. 58, 270–2783. Svennerholm, L., and Stallberg-Stenhagen, S. (1968) J. Lipid Res. 9, 215–2254. Levis, G. M., and Mead, J. F. (1964) J. Biol. Chem. 239, 77–805. Lippel, K., and Mead, J. F. (1968) Biochim. Biophys. Acta 152, 669–6806. Gerhardt, B. (1992) Prog. Lipid Res. 31, 417–4467. Andersen Borge, G. I., Slinde, E., and Nilsson, A. (1998) Biochim. Biophys.

Acta 1394, 158–1688. Koeduka, T., Matsui, K., Akakabe, Y., and Kajiwara, T. (2002) J. Biol. Chem.

277, 22648–226559. Fulco, A. J. (1967) J. Biol. Chem. 242, 3608–3613

10. Kaya, K., Ramesha, C. S., and Thompson, G. A., Jr. (1984) J. Biol. Chem. 259,3548–3553

11. Sandhir, R., Khan, M., and Singh, I. (2000) Lipids 35, 1127–113312. Croes, K., Van Veldhoven, P. P., Mannaerts, G. P., and Casteels, M. (1997)

FEBS Lett. 407, 197–20013. Croes, K., Casteels, M., Asselberghs, S., Herdewijn, P., Mannaerts, G. P., and

Van Veldhoven, P. P. (1997) FEBS Lett. 412, 643–64514. Foulon, V., Antonenkov, V. D., Croes, K., Waelkens, E., Mannaerts, G. P., Van

Veldhoven, P. P., and Casteels, M. (1999) Proc. Natl. Acad. Sci. U. S. A. 96,10039–10044

15. Croes, K., Foulon, V., Casteels, M., Van Veldhoven, P. P., and Mannaerts, G. P.(2000) J. Lipid Res. 41, 629–636

16. Foulon, V., Asselberghs, S., Geens, W., Mannaerts, G. P., Casteels, M., andVan Veldhoven, P. P. (2003) J. Lipid Res. 44, 2349–2355

17. Mukherji, M., Kershaw, N. J., Schofield, C. J., Wierzbicki, A. S., and Lloyd,M. D. (2002) Chem. Biol. 9, 597–605

18. Alderson, N. L., Rembiesa, B. M., Walla, M. D., Bielawska, A., Bielawski, J.,and Hama, H. (2004) J. Biol. Chem. 279, 48562–48568

19. Croes, K. (1998) �-Oxidation of 3-Methyl-branched Fatty acids: a RevisedPathway, Ph.D. thesis, Acta Biomedica Lovaniensa, Leuven UniversityPress, Leuven

20. Vanhove, G., Van Veldhoven, P. P., Vanhoutte, F., Parmentier, G., Eyssen,H. J., and Mannaerts, G. P. (1991) J. Biol. Chem. 266, 24670–24675

21. Novikov, D. K., Vanhove, G. F., Carchon, H., Asselberghs, S., Eyssen, H. J.,Van Veldhoven, P. P., and Mannaerts, G. P. (1994) J. Biol. Chem. 269,27125–27135

22. Ashton, R., Robinson, R., and Smith, J. C. (1936) J. Chem. Soc. 283–285

23. Baes, M., Gressens, P., Baumgart, E., Carmeliet, P., Casteels, M., Fransen, M.,Evrard, P., Fahimi, D., Declercq, P. E., Collen, D., Van Veldhoven, P. P., andMannaerts, G. P. (1997) Nat. Genet. 17, 49–57

24. Mannaerts, G. P., Debeer, L. J., Thomas, J., and De Schepper, P. J. (1979)J. Biol. Chem. 254, 4585–4595

25. Croes, K., Casteels, M., Van Veldhoven, P. P., and Mannaerts, G. P. (1995)Biochim. Biophys. Acta 1255, 63–67

26. Honkakoski, P., and Negishi, M. (1998) Biochem. J. 330, 889–89527. Declercq, P. E., Haagsman, H. P., Van Veldhoven, P., Debeer, L. J., Van Golde,

L. M., and Mannaerts, G. P. (1984) J. Biol. Chem. 259, 9064–907528. Kovacs, W. J., Faust, P. L., Keller, G. A., and Krisans, S. K. (2001) Eur.

J. Biochem. 268, 4850–485929. Van Veldhoven, P. P., Croes, K., Asselberghs, S., Herdewijn, P., and Man-

naerts, G. P. (1996) FEBS Lett. 388, 80–8430. Dirkx, R., Vanhorebeek, I., Martens, K., Schad, A., Grabenbauer, M., Fahimi,

D., Declercq, P. E., Van Veldhoven, P. P., and Baes, M. (2005) Hepatology,in press

31. Van Veldhoven, P. P., Huang, S., Eyssen, H. J., and Mannaerts, G. P. (1993)J. Inherited Metab. Dis. 16, 381–391

32. Casteels, M., Croes, K., Van Veldhoven, P. P., and Mannaerts, G. P. (1994)Biochem. Pharmacol. 48, 1973–1975

33. Larson, T. R., and Graham, I. A. (2001) Plant J. 25, 115–12534. Brocard, C., Lametschwandtner, G., Koudelka, R., and Hartig, A. (1997)

EMBO J. 16, 5491–550035. Croes, K., Casteels, M., de Hoffmann, E., Mannaerts, G. P., and Van Veld-

hoven, P. P. (1996) Eur. J. Biochem. 240, 674–68336. Jones, J. M., Morrell, J. C., and Gould, S. J. (2000) J. Biol. Chem. 275,

12590–1259737. Belmouden, A., Le, K. H., Lederer, F., and Garchon, H. J. (1993) Eur. J. Bio-

chem. 214, 17–2538. Croes, K., Casteels, M., Dieuaide-Noubhani, M., Mannaerts, G. P., and Van

Veldhoven, P. P. (1999) J. Lipid Res. 40, 601–60939. Coleman, R. A., Lewin, T. M., Van Horn, C. G., and Gonzalez-Baro, M. R.

(2002) J. Nutr. 132, 2123–212640. Steinberg, S. J., Morgenthaler, J., Heinzer, A. K., Smith, K. D., and Watkins,

P. A. (2000) J. Biol. Chem. 275, 35162–3516941. Foulon, V. (2001) �-Oxidation of 3-Methyl-branched Fatty Acids: Study of the

Enzymes Involved in the Reaction Sequence, Ph.D. thesis, Acta BiomedicaLovaniensa, Leuven University Press, Leuven

42. Saito, M., Iwamori, M., Lin, B., Oka, A., Fujiki, Y., Shimozawa, N., Kamoshita,S., Yanagisawa, M., and Sakakihara, Y. (1999) Biochim. Biophys. Acta1438, 55–62

43. Tatsumi, K., Saito, M., Lin, B., Iwamori, M., Ichiseki, H., Shimozawa, N.,Kamoshita, S., Igarashi, T., and Sakakihara, Y. (2001) Biochim. Biophys.Acta 1535, 285–293

44. Townsend, C. A. (2002) Curr. Opin. Chem. Biol. 6, 583–58945. Tittmann, K., Mesch, K., Pohl, M., and Hubner, G. (1998) FEBS Lett. 441,

404–40646. Verhoeven, N. M., Jakobs, C., Carney, G., Somers, M. P., Wanders, R. J., and

Rizzo, W. B. (1998) FEBS Lett. 429, 225–228

�-Oxidation of Straight Chain Fatty Acids9812

by guest on February 10, 2020http://w

ww

.jbc.org/D

ownloaded from

Guy P. Mannaerts, Paul P. Van Veldhoven and Minne CasteelsVeerle Foulon, Mieke Sniekers, Els Huysmans, Stanny Asselberghs, Vincent Mahieu,

-OXIDATION OF STRAIGHT CHAIN FATTY ACIDSα2-Hydroxyphytanoyl-CoA Lyase: A REVISED PATHWAY FOR THE

Breakdown of 2-Hydroxylated Straight Chain Fatty Acids via Peroxisomal

doi: 10.1074/jbc.M413362200 originally published online January 11, 20052005, 280:9802-9812.J. Biol. Chem. 

  10.1074/jbc.M413362200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/280/11/9802.full.html#ref-list-1

This article cites 42 references, 19 of which can be accessed free at

by guest on February 10, 2020http://w

ww

.jbc.org/D

ownloaded from