∆3-∆2-ENOYL-CoA ISOMERASE FROM THE YEAST SACCHAROMYCES CEREVISIAEMolecular and structural characterization

ANUMURSULA

Department of Biochemistry andBiocenter Oulu,

University of Oulu

OULU 2002

ANU MURSULA

∆3-∆2-ENOYL-CoA ISOMERASE FROM THE YEAST SACCHAROMYCES CEREVISIAEMolecular and structural characterization

Academic Dissertation to be presented with the assent ofthe Faculty of Science, University of Oulu, for publicdiscussion in Kajaaninsali (Auditorium L6), Linnanmaa, onApril 19th, 2002, at 2 p.m.

OULUN YLIOPISTO, OULU 2002

Copyright © 2002University of Oulu, 2002

Reviewed byProfessor Juha RouvinenProfessor Reijo Lahti

ISBN 951-42-6657-9 (URL: http://herkules.oulu.fi/isbn9514266579/)

ALSO AVAILABLE IN PRINTED FORMATActa Univ. Oul. A 380, 2002ISBN 951-42-6656-0ISSN 0355-3191 (URL: http://herkules.oulu.fi/issn03553191/)

OULU UNIVERSITY PRESSOULU 2002

Mursula, Anu, 3-2-Enoyl-CoA isomerase from the yeast Saccharomyces cerevisiaeMolecular and structural characterizationDepartment of Biochemistry, University of Oulu, P.O.Box 3000, FIN-90014 University of Oulu,Finland, Biocenter Oulu, University of Oulu, P.O.Box 5000, FIN-90014 University of Oulu, Finland Oulu, Finland2002

Abstract

The hydratase/isomerase superfamily consists of enzymes having a common evolutionary origin butacting in a wide variety of metabolic pathways. Many of the superfamily members take part in β-oxidation, one of the processes of fatty acid degradation. One of these β-oxidation enzymes is the ∆3-∆ 2-enoyl-CoA isomerase, which is required for the metabolism of unsaturated fatty acids. It catalyzesthe shift of a double bond from the position C3 of the substrate to the C2 position.

In this study, the ∆ 3-∆ 2-enoyl-CoA isomerase from the yeast Saccharomyces cerevisiae wasidentified, overexpressed as a recombinant protein and characterized. Subsequently, its structure andfunction were studied by X-ray crystallography.

The yeast ∆ 3-∆ 2-enoyl-CoA isomerase polypeptide contains 280 amino acid residues, whichcorresponds to a subunit size of 32 kDa. Six enoyl-CoA isomerase subunits assemble to form ahomohexamer. According to structural studies, the hexameric assembly can be described as a dimerof trimers. The yeast ∆ 3-∆ 2-enoyl-CoA isomerase is located in peroxisomes, the site of fungal β-oxidation, and is a necessary prerequisite for the β-oxidation of unsaturated fatty acids; the enoyl-CoA isomerase knock-out was unable to grow on such carbon sources.

In the crystal structure of the yeast ∆ 3-∆ 2-enoyl-CoA isomerase, two domains can be recognized,the N-terminal spiral core domain for catalysis and the C-terminal α-helical trimerization domain.This overall fold resembles the other known structures in the hydratase/isomerase superfamily. Site-directed mutagenesis suggested that Glu158 could be involved in the enzymatic reaction. Structuralstudies confirmed this, as Glu158 is optimally positioned at the active site for interaction with thesubstrate molecule. The oxyanion hole stabilizing the transition state of the enzymatic reaction isformed by the main chain NH groups of Ala70 and Leu126.

The yeast ∆ 3-∆ 2-enoyl-CoA isomerase hexamer forms by dimerization of two trimers, as in theother superfamily members. An extensive comparison of the five known structures of this familyshowed that the mode of assembly into hexamers is not a conserved feature of this superfamily, sincethe distance between the trimers and the orientation of the trimers with respect to each other varied.Marked differences were also detected between the two yeast enoyl-CoA isomerase crystal formsused in this study, one being crystallized at low pH and the other at neutral pH. The results suggestthat the yeast ∆ 3-∆ 2-enoyl-CoA isomerase could occur as a trimer at low pH.

Keywords: b-oxidation, enoyl-CoA isomerase, enoyl-CoA hydratase, hydratase/isomerasesuperfamily, X-ray crystallography

Acknowledgements

This work was carried out in the Department of Biochemistry and Biocenter Oulu at Universtity of Oulu during years 1998-2001.

I want to warmly thank professor Kalervo Hiltunen for his excellent supervision and constant encouragement during this study. His optimism and enthusiasm as well as his enormous knowledge of protein chemistry have helped me a lot. I am grateful to professor Rik Wierenga for taking me into his group when my project began to orient more towards structural studies, for giving me the valuable opportunity to learn new things, both practical and theoretical, and especially for his patient guidance during this learning process. This has opened a completely new world to me. All the other professors and group leaders in the Department of Biochemistry are acknowledged for creating such excellent research facilities and an enjoyable working atmosphere.

I am grateful to my co-authors, Dr. Aner Gurvitz, Dr. Hanspeter Rottensteiner, Prof. Andreas Hartig, Seppo Kilpeläinen, M.Sc., Dr. Daan van Aalten, Dr. Yorgo Modis and Dr. Arie Geerlof, for their contribution to our joint publications. Daan van Aalten also deserves my warmest thanks for his important help at the beginning of my crystallographic studies.

Professors Juha Rouvinen and Reijo Lahti are gratefully acknowledged for their valuable comments on the manuscript of this thesis, and Sirkka-Liisa Leinonen, Lic. Phil., for the careful revision of the language.

I wish to thank all the members of the RW and KH groups for everyday collaboration and numerous cheerful moments. Special thanks go to Inari and Petri Kursula, Antti Haapalainen, Tuomo Glumoff and Jukka Taskinen for all the helpful discussions and good advice.

Anneli Kaattari, Virpi Hannus, Tuula Koret, Maila Konttila, Jaakko Keskitalo and the rest of the staff of our department deserve many thanks for making things always run so smoothly. I thank Seppo Kilpeläinen for setting up and maintaining the Unix computer system. Ville Ratas, Kyösti Keränen, Tuomo Glumoff and Petri Kursula are gratefully acknowledged for taking care of the crystallization laboratory and the X-ray diffractometer.

My warmest thanks go to the ”Mafia-sisters”: Mira, Mari, Pia, Tanja and especially Anna for their friendship. As somebody wisely said: Life would be nothing without

friends and calories! I also wish to thank the incredible people from Teekkaritorvet for their relaxed company and for keeping music in my life.

I cannot thank enough my mother and father. I thank them for their never-ending love and support, for always believing in me, even though I did not always believe in myself, and for encouraging me to make my own choices in life. I also thank my brother Aki just for being there and for often cheering me up with his good sense of humour. Finally, I wish to thank my dear Janne for making my life wonderful.

This work was financially supported by the Academy of Finland and the Sigrid Jusélius Foundation.

Oulu, March 2002 Anu Mursula

Abbreviations

ABC ATP-binding cassette ACBP acyl-CoA-binding protein ADP adenosine diphosphate ATP adenosine triphosphate B-factor temperature factor 4-CBA 4-chlorobenzoyl coenzyme A cDNA complementary deoxyribonucleic acid ChcB 2-cyclohexenylcarbonyl-CoA isomerase CoA coenzyme A CPT carnitine palmitoyltransferase C-terminus carboxyl terminus DNA deoxyribonucleic acid ECI1 S. cerevisiae gene YLR284c Eci1p ECI1 gene product, yeast ∆3-∆2-enoyl-CoA isomerase E. coli Escherichia coli ER endoplasmic reticulum EMSA electrophoretic mobility shift assay FAD flavin adenine dinucleotide (oxidized form) GBP gastrin-binding protein GFP green fluorescent protein 4-HBA 4-hydroxybenzoyl coenzyme A HMPHP 4-hydroxy-3-methoxyphenyl-β-hydroxypropionyl-CoA kDa kilodalton MAD multiwavelength anomalous dispersion MFE-1, -2 multifunctional enzyme type 1, type 2 MES 2-(N-morpholino)ethanesulfonic acid MTP mitochondrial trifunctional protein NAD+ nicotinamide adenine dinucleotide (oxidized form) NCS noncrystallographic symmetry N-terminus amino terminus ORE oleate response element

ORF open reading frame PCR polymerase chain reaction PECI peroxisomal ∆3-∆2-enoyl-CoA isomerase PTS peroxisomal targeting signal RNA ribonucleic acid S. cerevisiae Saccharomyces cerevisiae SCP sterol carrier protein SDR short-chain alcohol dehydrogenase/reductase SDS-PAGE sodium dodecyl sulphate polyacrylamine gel electrophoresis TEA triethanol amine Å ångström, 10-10 m

List of original articles

This thesis is based on the following articles, which are referred to in the text by their Roman numerals: I Gurvitz A, Mursula AM, Firzinger A, Hamilton B, Kilpeläinen SH, Hartig A, Ruis

H, Hiltunen JK & Rottensteiner H (1998) Peroxisomal ∆3-cis-∆2-trans-enoyl-CoA isomerase encoded by ECI1 is required for growth of the yeast Saccharomyces cerevisiae on unsaturated fatty acids. J Biol Chem 273: 31366-31374.

II Mursula AM, van Aalten DMF, Modis Y, Hiltunen JK & Wierenga RK (2000)

Crystallization and X-ray diffraction analysis of peroxisomal ∆3-∆2-enoyl-CoA isomerase from Saccharomyces cerevisiae. Acta Cryst D56: 1020-1023.

III Mursula AM, van Aalten DMF, Hiltunen JK & Wierenga RK (2001) The crystal

structure of ∆3-∆2-enoyl-CoA isomerase. J Mol Biol 309: 845-853. IV Mursula AM, Geerlof A, Hiltunen JK & Wierenga RK (200X) Structural studies on

∆3-∆2-enoyl-CoA isomerase: the variable mode of assembly of the trimeric disks of the crotonase superfamily. Manuscript.

Contents

Acknowledgements Abbreviations List of original articles Contents 1 Introduction................................................................................................................... 13 2 Review of the literature................................................................................................. 14

2.1 Uptake and activation of fatty acids prior to β-oxidation....................................... 14 2.2 The β-oxidation cycle ............................................................................................ 16

2.2.1 β-Oxidation in mitochondria........................................................................... 16 2.2.2 β-Oxidation in peroxisomes ............................................................................ 18 2.2.3 β-Oxidation in the yeast Saccharomyces cerevisiae ....................................... 20

2.3 The requirement for auxiliary enzymes in β-oxidation .......................................... 21 2.3.1 β-Oxidation of (poly)unsaturated fatty acids .................................................. 21 2.3.2 β-Oxidation of α-methyl branched-chain fatty acids ...................................... 23

2.4 Hydratase/isomerase superfamily of enzymes ....................................................... 25 2.4.1 Members of the superfamily ........................................................................... 25

2.4.1.1 Trans-2-enoyl-CoA hydratase-1 ............................................................... 28 2.4.1.2 ∆3-∆2-Enoyl-CoA isomerase..................................................................... 29 2.4.1.3 ∆3,5-∆2,4-Dienoyl-CoA isomerase ............................................................. 30 2.4.1.4 4-Chlorobenzoyl-CoA dehalogenase........................................................ 32 2.4.1.5 Methylmalonyl-CoA decarboxylase......................................................... 32 2.4.1.6 Other members ......................................................................................... 32

2.4.2 Structure and function of the hydratase/isomerase proteins ............................ 34 2.4.2.1 Structural aspects ..................................................................................... 34 2.4.2.2 Catalytic mechanisms............................................................................... 36

3 Aims of the present study.............................................................................................. 42 4 Materials and methods .................................................................................................. 43

4.1 Strains and plasmids............................................................................................... 43 4.2 Saccharomyces genome database........................................................................... 43 4.3 Gene disruption (I) ................................................................................................ 43 4.4 Analysis of transcriptional regulation (I) ............................................................... 44

4.5 Subcellular localization (I) ..................................................................................... 44 4.6 Construction of the ECI1 expression vector (I)...................................................... 44 4.7 Expression of the recombinant protein (I).............................................................. 44 4.8 Protein purification (I, II) ....................................................................................... 45 4.9 Enzyme assays (I) .................................................................................................. 45 4.10 Site-directed mutagenesis (III) ............................................................................. 45 4.11 Determination of native molecular mass (I, III, IV)............................................. 46 4.12 Crystallization (II, III, IV).................................................................................... 46 4.13 X-ray diffraction data collection and processing (II-IV)...................................... 47

4.13.1 Data from unliganded crystals (II) ................................................................ 47 4.13.2 Multiwavelength anomalous dispersion (MAD) data (III)........................... 47

4.13.2.1 Preparation of heavy-atom derivative crystals ....................................... 47 4.13.2.2 MAD data collection .............................................................................. 48

4.13.3 Data from octanoyl-CoA complexed crystals (IV)........................................ 48 4.14 Structure determination by the MAD method and refinement of the structure (III) ................................................................................................................ 48 4.15 Structure determination by the molecular replacement method and refinement of the structure (IV) ................................................................................... 49 4.16 Structure analysis and validation (III-IV) ............................................................ 49 4.17 Comparison of the structures belonging to the hydratase/isomerase superfamily (IV)........................................................................................................... 49

5 Results........................................................................................................................... 51 5.1 Characterization of ECI1 and its gene product, Eci1p (I) ...................................... 51 5.2 Crystallization of yeast ∆3-∆2-enoyl-CoA isomerase (II, IV)................................. 52 5.3 Structure of the unliganded yeast ∆3-∆2-enoyl-CoA isomerase (III) ...................... 54 5.4 Structure of ∆3-∆2-enoyl-CoA isomerase complexed with octanoyl-CoA (IV)...... 56

6 Discussion ..................................................................................................................... 58 6.1 ECI1 encodes for a monofunctional peroxisomal ∆3-∆2-enoyl-CoA isomerase (I)................................................................................................................. 58 6.2 The structures of the unliganded and the octanoyl-CoA-complexed yeast ∆3-∆2-enoyl-CoA isomerase and the differences between them (III, IV) ............................... 61 6.3 Comparison of the yeast ∆3-∆2-enoyl-CoA isomerase structures with the other structures within the hydratase/isomerase superfamily (III, IV) .................................. 64

6.3.1 Comparison at the monomer level .................................................................. 64 6.3.2 Comparison at the hexamer level .................................................................... 65

7 Conclusions................................................................................................................... 67 8 Future perspectives ....................................................................................................... 68 9 References..................................................................................................................... 69

1 Introduction

Fatty acids are an important source of metabolic energy. When stored in adipose tissue as triacylglycerols, they are the most concentrated form of energy storage in animals. Fatty acids and their derivatives also have other physiological functions. They can act as components of biological membranes and serve as hormones and intracellular messengers.

For energy metabolism, fatty acids, activated into their coenzyme A esters, are degraded by sequential removal of acetyl-CoA units during a process called β-oxidation. Acetyl-CoA molecules are further targeted to the citric acid cycle for complete oxidation to CO2 and water or to ketone body formation in liver or cholesterolgenesis. In mammalian cells, β-oxidation occurs in two subcellular organelles, mitochondria and peroxisomes (Lazarow & De Duve 1976). In the yeast Saccharomyces cerevisiae (S. cerevisiae), however, fatty acids are degraded only in peroxisomes (Kunau et al. 1988). One cycle of β-oxidation of a saturated fatty acid consists of four subsequent enzymatic reactions: dehydrogenation/oxidation, hydration, second dehydrogenation and thiolytic cleavage. For the metabolism of double bonds in unsaturated fatty acids, auxiliary enzymes, such as ∆3-∆2-enoyl-CoA isomerase, 2,4-dienoyl-CoA reductase and ∆3,5-∆2,4-dienoyl-CoA isomerase, are needed.

Proteins can be classified into superfamilies according to their amino acid sequence similarity. Most enzymes participating in the hydration and isomerization reactions of β-oxidation belong to the low-similarity hydratase/isomerase superfamily of enzymes (Müller-Newen et al. 1995). In addition to β-oxidation, members of this superfamily also participate in many other metabolic pathways. Hydratase/isomerase superfamily members are thought to have evolved from a common ancestor and, despite the large variability of reactions they are able to catalyze, they appear to be both structurally and mechanistically related.

In this study, the gene encoding the S. cerevisiae ∆3-∆2-enoyl-CoA isomerase, a member of the hydratase/isomerase superfamily, was identified, the gene product was purified, characterized and crystallized, and its three-dimensional structure with and without an active site ligand was determined.

2 Review of the literature

2.1 Uptake and activation of fatty acids prior to ββββ-oxidation

For fatty acids to be utilizable as an energy source, they need to be released from triacylglycerols by lipases (for a review, see Gibbons et al. 2000), transported bound to albumin in blood to the target tissues and taken up by the cells either by diffusion or by transport proteins in the plasma membrane (for reviews, see van Nieuwenhoven et al. 1996, Hamilton 1998). In cytosol, fatty acids are bound to fatty acid binding proteins and delivered to appropriate subcellular sites (reviewed by Bernlohr et al. 1997).

Before fatty acids can enter the β-oxidation cycle, they must be activated to their CoA esters by the ATP-driven acyl-CoA synthetase (Fig. 1). Acyl-CoA synthetases are present in mitochondria, peroxisomes and the endoplasmic reticulum (ER), and they vary in substrate specificity (reviewed by Van Veldhoven & Mannaerts 1999). For mitochondrial β-oxidation, long-chain fatty acids are activated by a long-chain acyl-CoA synthetase located on the outer mitochondrial membrane with its active site exposed to the cytosolic side (Aas 1970, Hesler et al. 1990). Long-chain acyl-CoAs cannot readily traverse the inner mitochondrial membrane, while instead, the acyl moiety is coupled to carnitine by the malonyl-CoA-sensitive carnitine acyltransferase I on the outer mitochondrial membrane, then shuttled across the inner mitochondrial membrane by a carnitine acylcarnitine translocase in exchange for a carnitine molecule from the mitochondrial matrix (Pande 1975, Pande & Parvin 1976), after which the acyl moiety is linked back to a CoA molecule by a carnitine acyltransferase II located on the matrix side of the inner mitochondrial membrane (Fig. 1) (for a review, see Kerner & Hoppel 2000). Carnitine acyltransferases are commonly also called carnitine palmitoyltransferases (CPT) because of their substrate chain length specificity. Short- and medium-chain fatty acids do not require such a transport system for mitochondrial import, and they are activated in the mitochondrial matrix by short- and medium-chain acyl-CoA synthetases (Webster et al. 1965, Aas & Bremer 1968, Aas 1970, Eaton et al. 1996).

15

OOATP CoASCRC HS CoAR

O-

CH3 HCH3 H

N+ COOH-N+ CH3C CH2 CH2COOH-CH3C CH2 CH2

CH3 OCH3 OH

OCHS-CoA

R

OMM CPT I

IMM

CPT II

HS-CoA

+acyl-CoAsynthetase

acyl-CoAfatty acid

carnitine acyl-carnitine

translocase

acyl-CoAcarnitine

cytosol

mitochondrial matrix

Fig. 1. Activation of long-chain fatty acids and their transport into mitochondria for ββββ-oxidation (modified from Eaton et al. 1996, Kerner & Hoppel 2000). The long-chain acyl-CoA synthetase, which activates long-chain fatty acids to long-chain acyl-CoAs in an ATP-dependent manner, is located on the outer mitochondrial membrane (OMM). The carnitine palmitoyltransferase I (CPT I), acylcarnitine carnitine translocase (translocase) and CPT II couple the acyl moiety to carnitine, shuttle the acylcarnitine through the inner mitochondrial membrane (IMM) and regenerate acyl-CoA, respectively.

For peroxisomal β-oxidation, fatty acids are activated at different subcellular locations. Long-straight-chain and 2-methyl-branched-chain fatty acids are activated by acyl-CoA synthetases on the cytoplasmic side of the peroxisomal membrane, on the outer mitochondrial membrane and in ER (Krisans et al. 1980, Mannaerts et al. 1982, Vanhove et al. 1991, Wanders et al. 1992). The same long-chain acyl-CoA synthetase is probably also responsible for the activation of branched-chain fatty acids (Wanders et al. 1992, Watkins et al. 1996). Very-long-chain acyl-CoAs (>C20) are generated only in peroxisomes and ER (Singh & Poulos 1988, Lazo et al. 1990) by a very-long-chain fatty acyl-CoA synthetase (Uchida et al. 1996, Uchiyama et al. 1996). The peroxisomal very-long-chain acyl-CoA synthetase is located on the matrix side of the peroxisomal membrane, in contrast to the peroxisomal long-chain acyl-CoA synthetase (Mannaerts et al. 1982), and, in addition to the straight-chain fatty acids, it also activates branched-chain fatty acids, such as pristanic acid (Steinberg et al. 1999). The peroxisomal very-long-chain acyl-CoA synthetase could thus have an important role in the intraperoxisomal

16

reactivation of pristanic acid (Steinberg et al. 1999), which is the α-oxidation product of phytanic acid (see “β-Oxidation of α-methyl-branched-chain fatty acids”). Derivatives of fatty acids oxidized in peroxisomes, namely dicarboxylic fatty acids, prostaglandins and the carboxylic side chains of bile acid intermediates are activated to their CoA esters by ER enzymes (Vamecq et al. 1985, Prydz et al. 1988, Schepers et al. 1988, Schepers et al. 1989). Peroxisomes do not use the carnitine-coupled transport system present in mitochondria for the import of acyl-CoA esters. It is not completely clear how activated fatty acids enter the peroxisomal matrix for degradation by β-oxidation. Long-chain and very-long-chain fatty acyl-CoAs probably reach the matrix via a membrane-bound transporter containing an ATP-binding cassette (ABC) motif, as shown in S. cerevisiae (Hettema et al. 1996). The homologous protein in human is affected in adrenoleukodystrophy, a peroxisomal disease, in which the metabolism of very-long-chain fatty acids is impaired (Aubourg et al. 1993, Mosser et al. 1993, 1994). How CoA esters of dicarboxylic fatty acids, prostaglandins and bile acid intermediates reach peroxisomes is not known yet.

In the yeast S. cerevisiae, medium-chain fatty acids first enter the peroxisome, the exclusive site of β-oxidation in yeast, and then get activated by an intraperoxisomal acyl-CoA synthetase, whereas long-chain fatty acids are activated in the cytosol and transported via Pat1p and Pat2p, the ABC transporter proteins (Hettema et al. 1996, Trotter 2001).

2.2 The ββββ-oxidation cycle

In this section, the β-oxidation of saturated fatty acids in mammalian mitochondria and peroxisomes as well as in yeast peroxisomes is described. The metabolism of unsaturated fatty acids is discussed in the chapter titled ”The requirement for auxiliary enzymes of β-oxidation”.

2.2.1 ββββ-Oxidation in mitochondria

The bulk of dietary long-chain fatty acids are metabolized by the β-oxidation enzymes of mitochondria. Mitochondrial β-oxidation degrades fatty acids completely to acetyl-CoA, which is then oxidized by the citric acid cycle or, during starvation, condensed into ketone bodies. In mitochondria, β-oxidation enzymes are organized into two separate functional systems: the inner membrane-bound complex for long-chain fatty acid oxidation and the soluble matrix system for the degradation of medium- and short-chain fatty acids (for reviews see Eaton et al. 1996, Hiltunen & Qin 2000, Liang et al. 2001, Reddy & Hashimoto 2001). One cycle of β-oxidation consists of four subsequent reactions (Fig. 2). First, the acyl-CoA molecule is dehydrogenated by an acyl-CoA dehydrogenase leading to the generation of trans-2-enoyl-CoA and FADH2. Second, 2-enoyl-CoA hydratase-1 (hydratase-1, crotonase) adds water to the trans-2 double bond and L-3-hydroxyacyl-CoA is formed. Third, another dehydrogenase, L-3-hydroxyacyl-CoA dehydrogenase catalyzes the formation of 3-ketoacyl-CoA and NADH. As the fourth

17

step, 3-ketoacyl-CoA thiolase cleaves 3-ketoacyl-CoA to acetyl-CoA and an acyl-CoA molecule shortened by two carbon atoms, which can re-enter the β-oxidation cycle.

In rat liver mitochondria, four straight-chain acyl-CoA dehydrogenases with different chain length specificities have been characterized. The very-long-chain acyl-CoA dehydrogenase is a homodimer located on the inner mitochondrial membrane and active towards acyl-CoA molecules up to 24 carbon atoms in length (Izai et al. 1992). The mitochondrial matrix contains separate short-, medium- and long-chain acyl-CoA dehydrogenases (Ikeda et al. 1985). These enzymes are homotetramers with substrate specificities of <C8, C6-C12 and C10-C22, respectively. For 2-methyl-branched fatty acyl-CoAs, a separate acyl-CoA dehydrogenase exists (Ikeda et al. 1983).

SCoA

O

R

SCoA

O

R

SCoA

O

R

OH

SCoA

O

R

O

SCoA

OSCoA

O

R

FAD

FADH2

H2O

NAD+NADH + H+

CoA

1.

2.

3.

4.

subsequent cyclesof beta-oxidation

Fig. 2. The ββββ-oxidation cycle. The intermediates of the pathway are shown. The four reactions are catalysed by 1. acyl-CoA dehydrogenase in mitochondria or acyl-CoA oxidase in peroxisomes, 2. 2-enoyl-CoA hydratase, 3. 3-hydroxyacyl-CoA dehydrogenase, and 4. 3-ketoacyl-CoA thiolase. The acyl-CoA shortened by two carbon atoms can enter subsequent cycles of ββββ-oxidation.

The very-long and long-chain 2-enoyl-CoAs formed by the respective dehydrogenases

are fed into the long-chain-specific inner membrane-bound trifunctional β-oxidation complex for the second, third and fourth steps of the β-oxidation cycle (Carpenter et al. 1992, Uchida et al. 1992). The mitochondrial trifunctional enzyme (MTP) is an oligomer formed of four 79-kDa α-subunits and four 48-kDa β-subunits. The α-subunit contains

18

the long-chain 2-enoyl-CoA hydratase-1 and long-chain L-3-hydroxyacyl-CoA dehydrogenase activities, whereas the β-subunit catalyzes the long-chain 3-ketoacyl-CoA thiolase reaction (Kamijo et al. 1993).

Once the long-chain acyl-CoAs have been shortened enough by the membrane-bound complex, the acyl-CoAs are further oxidized by the monofunctional medium- and short-chain-specific enzymes in the mitochondrial matrix. The short-chain 2-enoyl-CoA hydratase-1, also called crotonase, is a homohexamer and most active towards C4 substrates, but can also metabolize, at a much lower rate, substrates up to C16 (Hass & Hill 1969, Furuta et al. 1980). Another matrix-associated 2-enoyl-CoA hydratase-1, which is most active towards medium- and long-chain substrates, has been detected in human and pig mitochondria (Jackson et al. 1995).

The soluble L-3-hydroxyacyl-CoA dehydrogenase in the matrix acts on substrates with chain lengths from C4 to C16, but prefers substrates with shorter chain length (Osumi & Hashimoto 1980, He et al. 1989, Eaton et al. 1996). For the final step of the β-oxidation cycle, the mitochondrial matrix contains two short- and medium-chain-specific 3-ketoacyl-CoA thiolases. One is specific for acetoacetyl-CoA and 2-methylacetoacetyl-CoA and the other for substrates ranging from C6 to C16 (Eaton et al. 1996).

2.2.2 ββββ-Oxidation in peroxisomes

The presence of β-oxidation in peroxisomes was first discovered by Cooper and Beevers in 1969. They found that fatty acids could be oxidized in glyoxysomes, which are plant organelles closely related to peroxisomes, of germinating castor bean seedlings. Lazarow and De Duve (1976) were able to further show that mammalian peroxisomes were also capable of fatty acid β-oxidation. The peroxisomal β-oxidation cycle consists principally of the same four reactions as the mitochondrial pathway: dehydrogenation/oxidation, hydration, another dehydrogenation and thiolytic cleavage (Fig. 2). The main differences lie in the substrate and stereo specificities of these pathways (reviewed by Hiltunen & Qin 2000, Reddy & Hashimoto 2001, Van Veldhoven et al. 2001, Wanders et al. 2001). Peroxisomes degrade fatty acids and fatty acid derivatives that cannot be oxidized by mitochondrial enzymes. The main role of peroxisomal β-oxidation is to shorten or otherwise convert fatty acids into a form that can be accepted by the mitochondrial enzymes. The substrates of peroxisomal β-oxidation include very-long-chain fatty acids, dicarboxylic fatty acids, branched-chain fatty acids (e.g. pristanic acid), intermediates of C27 bile acid synthesis (di- and trihydroxycholestanoic acid), prostaglandins, leucotrienes and some xenobiotics. Very-long-chain fatty acids (>C20) are not oxidized effectively in mitochondrial β-oxidation for two reasons. First, mitochondria do not contain very-long-chain specific acyl-CoA synthetase (Uchiyama et al. 1996) and very-long-chain fatty acids are not substrates for the mitochondrial long-chain acyl-CoA synthetase (Lazo et al. 1990). Second, very-long-chain fatty acids are not substrates for the human carnitine palmitoyltransferase required for the mitochondrial import of long-chain fatty acids (Wanders et al. 2001). The reason why many of the branched-chain fatty acids, such as phytanic and pristanic acid, are oxidized mainly in peroxisomes is the fact that carnitine palmitoyltransferase does not accept branched-chain fatty acids as substrates, either

19

(Singh et al. 1994). After being chain-shortened or sufficiently modified in peroxisomes, the acyl moieties are linked to carnitine inside peroxisomes and transported into mitochondria for complete oxidation. Also the acetyl-CoA molecules formed in peroxisomes after every β-oxidation cycle are taken into mitochondria for oxidation in the citric acid cycle.

The first step of peroxisomal β-oxidation is catalyzed by acyl-CoA oxidase instead of acyl-CoA dehydrogenase in mitochondria. Acyl-CoA oxidase, being linked to FAD, donates the electrons obtained from the oxidation reaction directly to molecular oxygen, thus generating H2O2. The product of the oxidation reaction is the same as the product of mitochondrial dehydrogenation, trans-2-enoyl-CoA. The oxidation is catalyzed by multiple acyl-CoA oxidases using either straight-chain or branched-chain substrates. Two acyl-CoA oxidases have been characterized in human peroxisomes (Casteels et al. 1990, Vanhove et al. 1993) and three in rat peroxisomes (Osumi et al. 1980, Schepers et al. 1990, Van Veldhoven et al. 1991, 1992, 1994).

Mammalian peroxisomes contain two distinct multifunctional enzymes, type 1 and type 2, both of which catalyze the second and third reactions of β-oxidation, but with opposite chirality. No monofunctional enzymes catalyzing these reactions in mammalian peroxisomes exist. Multifunctional enzyme type 1 (MFE-1) was first purified and characterized from rat liver peroxisomes by Osumi and Hashimoto (1979). The purified 77-kDa polypeptide contained the 2-enoyl-CoA hydratase-1 and L-3-hydroxyacyl-CoA dehydrogenase activities (Osumi & Hashimoto 1979), the hydratase-1 activity being located in the amino terminal (N-terminal) domain and the dehydrogenase activity in the carboxyl terminal (C-terminal) domain (Ishii et al. 1987). Later, however, it was found that instead of being a bifunctional protein, MFE-1 is trifunctional and also catalyzes the isomerization of cis-3-enoyl-CoAs to trans-2-enoyl-CoA (Palosaari & Hiltunen 1990). The hydratase-1 and isomerase reactions occur in the same catalytic domain in the N-terminal part of MFE-1 (Palosaari et al. 1991). The MFE-2 of rat liver peroxisomes is a 79 kDa polypeptide catalyzing two reactions, hydration of trans-2-enoyl-CoA to D-3-hydroxyacyl-CoA and dehydrogenation of D-3-hydroxyacyl-CoA to 3-ketoacyl-CoA (Dieuaide-Noubhani et al. 1997a, Qin et al. 1997a). The D-3-hydroxyacyl-CoA dehydrogenase activity is located in the N-terminal part and the 2-enoyl-CoA hydratase-2 activity in the middle portion. The C-terminal part is similar to the rat sterol carrier protein 2 (Mori et al. 1991, Seedorf & Assman 1991). The amino acid sequences of MFE-1 and MFE-2 are not homologous and they use different substrates in peroxisomal β-oxidation. MFE-1 catalyzes the hydration and dehydrogenation of very-long-chain and other straight-chain fatty acids, such as dicarboxylic acids and eicosanoids, whereas MFE-2 is more active towards branched-chain substrates, such as pristanoyl-CoA and intermediates of bile acid synthesis (Dieuaide-Noubhani et al. 1997a, Dieuaide-Noubhani et al. 1997b, Qin et al. 1997a, Qin et al. 1997b).

The final reaction of the β-oxidation cycle is catalyzed by a 3-ketoacyl-CoA thiolase, which creates acetyl-CoA and an acyl-CoA chain shortened by two carbon atoms. For this, multiple enzyme isoforms also exist in mammalian peroxisomes. In rat, two closely related straight-chain ketoacyl-CoA thiolases could be found, one being inducible by peroxisome proliferators and the other not (Miyazawa et al. 1981, Hijikata et al. 1990). In humans, only one gene is similar to the rat genes (Bout et al. 1988). In human and rat peroxisomes, there is yet another 3-ketoacyl-CoA thiolase called peroxisomal thiolase 2

20

or SCPx. In addition to the thiolase domain it also contains a sterol carrier protein (SCP) domain (Seedorf et al. 1994). SCPx is capable of metabolizing branched-chain fatty acids and bile acid intermediates, whereas the original ketoacyl-CoA thiolase cleaves only straight-chain ketoacyl-CoAs (Antonenkov et al. 1997, Wanders et al. 1997).

The straight-chain-metabolizing L-specific β-oxidation pathway and the D-specific route that also metabolizes branched-chain fatty acids similarly differ in their inducibility. The expression of the “classical” enzymes is strongly induced in rodents by peroxisome proliferators, such as clofibrate, whereas the enzymes with branched-chain specificity are not (Lazarow & De Duve 1976, Van Veldhoven et al. 1992).

2.2.3 ββββ-Oxidation in the yeast Saccharomyces cerevisiae

In the yeast S. cerevisiae, β-oxidation is restricted to peroxisomes (Kunau et al. 1988). The enzymes of the β-oxidation cycle consist of an acyl-CoA oxidase, a multifunctional enzyme and a thiolase, as in mammalian peroxisomes (for a review see Trotter 2001). These enzymes are encoded by the yeast genes POX1 (Dmochowska et al. 1990), FOX2 (Hiltunen et al. 1992) and FOX3/POT1 (Einerhand et al. 1991, Igual et al. 1991), respectively. The disruption of any of these genes results in the inability of yeast cells to grow on fatty acids as their sole carbon source (Dmochowska et al. 1990, Igual et al. 1991, Hiltunen et al. 1992), indicating that there are no isoforms for these enzymes in S. cerevisiae and that the enzymes must be able to metabolize substrates of all chain lengths. The enzyme encoded by FOX2 catalyzes the same D-specific reactions as the mammalian MFE-2. In fact, fox2p was the first multifunctional enzyme characterized as catalyzing the D-specific route of hydration and dehydrogenation (Hiltunen et al. 1992). This finding ruled out the previous assumption of yeast MFE catalyzing three reactions: L-specific hydration, L-specific dehydrogenation and epimerization, which would convert L-3-hydroxyacyl-CoA to D-3-hydroxyacyl-CoA and vice versa (Hiltunen et al. 1992). Thus, S. cerevisiae completely lacks the L-specific 2-enoyl-CoA hydratase-1 and 3-hydroxyacyl-CoA dehydrogenase. In vivo studies have shown, however, that the L-specific route can also be introduced into yeast cells (Filppula et al. 1995). This was accomplished by expressing the rat MFE-1 in yeast cells devoid of the endogenous D-specific MFE and then testing the complementation of growth on fatty acids. Indeed, although MFE-1 and yeast MFE are not related in terms of the amino acid sequence, they can functionally complement each other in S. cerevisiae (Filppula et al. 1995).

The NADH formed in the 3-hydroxyacyl-CoA dehydrogenation reaction is recycled back to NAD+ by malate dehydrogenase (van Roermund et al. 1995). The other end product of β-oxidation, acetyl-CoA, is transported to the mitochondria as an acetyl-carnitine derivative and oxidized by the citric acid cycle. Alternatively, acetyl-CoA may be metabolized by the glyoxylate cycle that produces isocitrate and succinate, which can be further metabolized in mitochondria (van Roermund et al. 1995).

21

2.3 The requirement for auxiliary enzymes in ββββ-oxidation

2.3.1 ββββ-Oxidation of (poly)unsaturated fatty acids

Since trans-2 is the only unsaturated intermediate in the classical β-oxidation cycle and most of the double bonds in the dietary fatty acids are in cis-configuration, the additional double bonds in (poly)unsaturated fatty acids must be converted to the trans-2 configuration in order for the β-oxidation to proceed. This is accomplished by the auxiliary enzymes of β-oxidation (Eaton et al. 1996, Hiltunen & Qin 2000). The cis-double bonds at even-numbered positions in fatty acids are degraded via the reductase-dependent route (Fig. 3A). When the double bond reaches position cis-4, a trans-2 double bond is added by acyl-CoA dehydrogenase/oxidase, creating 2,4-dienoyl-CoA. This compound is reduced in a NADPH-dependent manner by 2,4-dienoyl-CoA reductase, resulting in eucaryotes in 3-enoyl-CoA, which is further isomerized by ∆3-∆2-enoyl-CoA isomerase to trans-2-enoyl-CoA, the substrate of the β-oxidation spiral (Wang & Schulz 1989).

The metabolism of cis-5 and other unsaturated fatty acids with double bonds at odd-numbered positions was initially thought to occur via the isomerase-dependent pathway: chain shortening in β-oxidation would result in cis-3-enoyl-CoA, which would be converted to trans-2-enoyl-CoA by the ∆3-∆2-enoyl-CoA isomerase (Fig. 3B). Tserng and Jin (1991), however, showed that the cis-5 double bonds are removed by NADPH-dependent reduction in mitochondria. This alternative route was found to contain a novel enzyme, ∆3,5-∆2,4-dienoyl-CoA isomerase, which acts together with 2,4-dienoyl-CoA reductase and ∆3-∆2-enoyl-CoA isomerase to produce trans-2 double bonds (Fig. 3C) (Smeland et al. 1992). This pathway involves the isomerization of trans-2-cis-5-dienoyl-CoA to ∆3-cis-5-dienoyl-CoA by the ∆3-∆2-enoyl-CoA isomerase and a second isomerization to ∆2-∆4-dienoyl-CoA by the novel dienoyl-CoA isomerase followed by the reductase-dependent route. The rat mitochondrial ∆3,5-∆2,4-dienoyl-CoA isomerase was subsequently purified from rat liver (Chen et al. 1994, Luo et al. 1994). Shortly after this, it was discovered that peroxisomes also contain a ∆3,5-∆2,4-dienoyl-CoA isomerase and are thus capable of metabolizing fatty acids with double bonds in odd-numbered carbons (He et al. 1995, Luthria et al. 1995). Two studies, using different experimental approaches, were published to discuss whether the isomerase-dependent pathway (Fig. 3B) or the route requiring dienoyl-CoA isomerase, enoyl-CoA isomerase and dienoyl-CoA reductase (Fig. 3C) (referred to as the reductase-dependent pathway) is the predominant one in metabolizing cis-5 double bonds in mitochondria. Tserng and co-workers (1996) found that cis-5-decenoate was completely degraded via the reductase-dependent pathway in liver mitochondria, while when the chain length was elongated to cis-5-tetradecenoate, 86 % of the fatty acid was metabolized via the reductase-dependent route. Shoukry and Schulz (1998) reported an opposite finding, stating that 80 % of 2,5-octadienoyl-CoA is oxidized via the isomerase-dependent pathway. They also suggested that a small portion of trans-2-trans-4-dienoyl-CoA formed by the dienoyl-CoA isomerase could enter directly the process of β-oxidation at the site of hydratase-1 action without reduction by the dienoyl-CoA reductase.

22

SCoAR

O

SCoAR

O

SCoAR R SCoA

R SCoA

R SCoA

OH O

O

SCoAR

O

RSCoA

R SCoA

O O

R SCoA

O

SCoAR

SCoAR

SCoAR

SCoAR

Even-numbered double bonds

4

4 2

3

2

5

5

2

5

5

3

2

5 3

24

3

2

Odd-numbered double bonds

O

O

O

O

O

O

ODECR

DECR

ECI

ECI ECI

ECI

DECI

B CA

Fig. 3. The routes for the degradation of double bonds in unsaturated fatty acids. The dashed arrows indicate reactions of the classical ββββ-oxidation cycle and the solid arrows reactions catalysed by auxiliary enzymes; DECR, 2,4-dienoyl-CoA reductase; ECI, ∆∆∆∆3-∆∆∆∆2-enoyl-CoA isomerase; DECI, ∆∆∆∆3,5-∆∆∆∆2,4-dienoyl-CoA isomerase. The even-numbered double bonds are degraded via the reductase-dependent route (A). The odd-numbered double bonds can be oxidized via either the isomerase-dependent pathway (B) or the route that also requires dienoyl-CoA isomerase and dienoyl-CoA reductase (C).

23

All the auxiliary enzymes required for the metabolism of unsaturated fatty acids exist in both the peroxisomes and the mitochondria of mammalian cells. The molecular characterization of the ∆3,5-∆2,4-dienoyl-CoA isomerase and the ∆3-∆2-enoyl-CoA isomerase will be described in more detail in the chapter “Hydratase/isomerase superfamily of enzymes”. The existence of a mitochondrial 2,4-dienoyl-CoA reductase was first demonstrated by Kunau and Dommes (1978). Since then, reductases from various sources have been characterized, including dienoyl-CoA reductases from bovine (Dommes et al. 1982), rat (Hakkola & Hiltunen 1993) and human (Koivuranta et al. 1994) mitochondria as well as rat (Dommes et al. 1981, Kimura et al. 1984), human (De Nys et al. 2001) and mouse (Geisbrecht et al. 1999a) peroxisomes. As an example, the human mitochondrial 120 kDa 2,4-dienoyl-CoA reductase is composed of four 36-kDa subunits, each containing a N-terminal mitochondrial targeting sequence (Koivuranta et al. 1994). At the amino acid sequence level, it shows 82.7 % similarity to the corresponding rat mitochondrial dienoyl-CoA reductase (Hirose et al. 1990). All the characterized eucaryotic 2,4-dienoyl-CoA reductases belong to the functionally heterologous short-chain alcohol dehydrogenase/reductase (SDR) superfamily with a characteristic NADPH-binding site called the Rossmann fold (Hiltunen & Qin 2000).

Bacteria also contain the machinery for the β-oxidation of unsaturated fatty acids, but it will not be discussed here. The β-oxidation of the yeast S. cerevisiae has been subject to intensive investigations, including this study. The pathways for the metabolism of unsaturated fatty acids in yeast will be described in the “Discussion” section.

2.3.2 ββββ-Oxidation of αααα-methyl branched-chain fatty acids

Pristanic acid (2,6,10,14-tetramethylpentadecanoic acid) can be derived from two sources, either directly from the diet or as the α-oxidation product of phytanic acid (3,7,11,15-tetramethylhexadecanoic acid) (reviewed by Hiltunen & Qin 2000, Wanders et al. 2001) (Fig. 4). Phytanic acid cannot be directly β-oxidized because the 3-methyl group inhibits the second dehydrogenation step. In order for phytanic acid to be suitable for β-oxidation, it has to be shortened by a process called α-oxidation. In this pathway, phytanoyl-CoA is first hydroxylased to 2-hydroxyphytanoyl-CoA, then cleaved to pristanal and formyl-CoA and finally oxidized to pristanic acid, which can be activated to pristanoyl-CoA and metabolized by peroxisomal β-oxidation (Mihalik et al. 1995, Croes et al. 1997, Verhoeven et al. 1997). Pristanoyl-CoA undergoes three cycles of β-oxidation in peroxisomes, after which it is transported into mitochondria for complete oxidation (Verhoeven et al. 1998). Pristanic acid is a racemic mixture containing the (2S,6R,10R) and (2R,6R,10R) diastereomers. The branched-chain acyl-CoA oxidase of β-oxidation, however, can only act on the 2S-isomer (Van Veldhoven et al. 1996). An enzyme overcoming this problem was purified from rat liver by Schmitz and co-workers (1994). The enzyme was named α-methylacyl-CoA racemase, and it catalyzed the conversion of the 2R-isomers to the 2S-conformation suitable for β-oxidation (Fig. 4). In addition to pristanoyl-CoA acid, other α-methylacyl-CoAs, the branched side chains of bile acid intermediates, namely di- and trihydroxycholestanoyl-CoA, and arylpropionic acids also serve as substrates for α-methylacyl-CoA racemase (Schmitz et al. 1994). During bile

24

acid biosynthesis, (25R)-di- and (25R)-trihydroxycholestanoyl-CoA are produced from cholesterol by mitochondrial 27-hydroxylation (Shefer et al. 1978, Batta et al. 1983). Because of the S-stereospecificity of the branched-chain acyl-CoA oxidase, α-methylacyl-CoA racemase has to convert di- and trihdydroxycholestanoyl-CoAs to their 25S-stereoisoforms, after which the side chain can be shortened by peroxisomal β-oxidation (Fig. 4) (Van Veldhoven et al. 1996) leading to the formation of choloyl-CoA.

In addition to rat liver, racemase has also been purified from human liver, and the cDNA sequences of the rat and mouse racemases have been determined (Schmitz et al. 1995, 1997). Also, the gene structure of the mouse racemase has been resolved (Kotti et al. 2000). The rat α-methylacyl-CoA racemase cDNA encodes for a polypeptide of 39.7 kDa, in good agreement with the molecular mass of the purified enzyme (Schmitz et al. 1994, 1997). All the α-methylacyl-CoA racemases studied have dual subcellular localization, both in mitochondria and in peroxisomes (Van Veldhoven et al. 1997, Kotti et al. 2000). At least in mouse, the same gene encodes both mitochondrial and peroxisomal racemases and the gene product is targeted to two different locations without modifications (Kotti et al. 2000). The physiological function of the peroxisomal racemase is clear: to convert the (R)-α-methyl groups of pristanoyl-CoA and trihydroxycholestanoyl-CoA to the S-conformation. In mitochondria, the degradation product of pristanoyl-CoA, 2,6-dimethylheptanoyl-CoA, is also a substrate for racemase, since both of its methyl groups occur in the R-conformation (Ferdinandusse et al. 2000).

HO

O

O

O

SCoA

SCoA

SCoA

OHHOH HO

H

HO

OH

O

O

SCoA

O

SCoA

SCoA

(3R)-phytanoyl-CoA

(2R)-pristanoyl-CoA

(3S)-phytanoyl-coA

(2S)-pristanoyl-CoA

cholesterol

(25R)-trihydroxycholestanoyl-CoA (25S)-trihydroxycholestanoyl-CoA

α-oxidation

racemase

racemase

peroxisomalββββ-oxidation

Fig. 4. αααα-Methylacyl-CoA racemase is required for the ββββ-oxidation of pristanoyl-CoA, which is the αααα-oxidation product of phytanoyl-CoA, and for bile acid synthesis. The peroxisomal ββββ-oxidation steps include the action of branched-chain acyl-CoA oxidase, MFE-2 and SCPx. The figure was modified from Ferdinandusse et al. 2000.

25

2.4 Hydratase/isomerase superfamily of enzymes

The hydratase/isomerase superfamily, also called the crotonase superfamily, was first defined to consist of 2-enoyl-CoA hydratases and ∆3-∆2-enoyl-CoA isomerases participating in β-oxidation (Müller-Newen & Stoffel 1993). The first members of the superfamily included the rat peroxisomal MFE-1 (Osumi et al. 1985), the rat monofunctional mitochondrial 2-enoyl-CoA hydratase-1 (Minami-Ishii et al. 1989), the α-subunit of the fatty acid degradation multienzyme complex from Escherichia coli (E. coli) (Dirusso 1990) and the rat mitochondrial monofunctional ∆3-∆2-enoyl-CoA isomerase (Müller-Newen & Stoffel 1991, Palosaari et al. 1991). These enzymes have low but significant similarity in their amino acid sequence, about 25-27 %, and they share at least one common catalytic amino acid (Müller-Newen & Stoffel 1993, Müller-Newen et al. 1995). Since then, the hydratase/isomerase superfamily has expanded to consist of over 30 members acting in a wide range of metabolic pathways but still possessing an amino acid sequence pattern typical of the superfamily (Müller-Newen & Stoffel 1993, Wu et al. 1997, Holden et al. 2001). The reactions catalyzed by the hydratase/isomerase superfamily members nowadays include dehalogenation, hydration/dehydration, isomerization, decarboxylation, formation/cleavage of carbon-carbon bonds, and hydrolysis of thioesters (Table 1). What is common for the enzymes is that, with one exception, they all use coenzyme A esters as substrates and their catalytic mechanisms involve the stabilization of an oxyanion intermediate (Babbitt & Gerlt 1997, Holden et al. 2001). Because of the similarity of their amino acid sequences, the hydratase/isomerase superfamily members are thought to have evolved from a common ancestor and, therefore, to be both mechanistically and structurally related. This assumption has also been confirmed by the five crystal structures solved so far within the hydratase/isomerase superfamily (Benning et al. 1996, Engel et al. 1996, Modis et al. 1998, Benning et al. 2000, Kurimoto et al. 2001). In this section, the members of the hydratase/isomerase superfamily will be described and their structure-function relationships will be discussed, the emphasis being on the enzymes for which the structure is known.

2.4.1 Members of the superfamily

Table 1 summarizes the hydratase/isomerase superfamily members with known sequence and function. Since there is a vast amount of sequence information available in genome databases due to the genome sequencing projects, this is not a complete list. In addition, the list does not contain hydratase/isomerase proteins with unknown function.

26

Table 1. The hydratase/isomerase superfamily members with known sequence and function.

Enzyme

Reference

Monofunctional 2-enoyl-CoA hydratases rat mitochondrial Minami-Ishii et al. 1989 human mitochondrial Kanazawa et al. 1993 Clostridium acetobutylicum Boynton et al. 1996 Rhizobium meliloti Margolin et al. 1995 E. coli CaiD, crotonobetainyl-CoA hydratase

Eichler et al. 1994, Elssner et al. 2001

2-enoyl-CoA hydratases as a part of a multifunctional proteina

rat peroxisomal MFE-1 Osumi et al. 1985 human peroxisomal MFE-1 Chen et al. 1991, Hoefler et al. 1994 guinea pig peroxisomal MFE-1 Caira et al. 1996 α-subunit of rat MTPb Kamijo et al. 1993 α-subunit of human MTP Kamijo et al. 1994 α-subunit of pig MTP Yang et al. 1994a E. coli fadB gene product Dirusso 1990, Yang et al. 1991 Pseudomonas fragi faoA gene product Sato et al. 1992 human AU-specific RNA-binding protein Nakagawa et al. 1995 mouse AU-specific RNA-binding protein Brennan et al. 1999

∆3-∆2-Enoyl-CoA isomerases

rat mitochondrial Müller-Newen & Stoffel 1991, Palosaari et al. 1991

human mitochondrial Janssen et al. 1994, Kilponen et al. 1994 mouse mitochondrial Stoffel et al. 1993 rat peroxisomal MFE-1c Osumi et al. 1985 human peroxisomal Geisbrecht et al. 1999b mouse peroxisomal Geisbrecht et al. 1999b E. coli fadB gene product Dirusso 1990, Yang et al. 1991 Streptomyces collinus ChcBd Patton et al. 2000

∆3,5-∆2,4-Dienoyl-CoA isomerases

rat peroxisomal/mitochondrial FitzPatrick et al. 1995, Filppula et al. 1998 S. cerevisiae peroxisomal Gurvitz et al. 1999

4-Chlorobenzoyl-CoA dehalogenases

Pseudomonas sp. CBS-3 Babbitt et al. 1992 Arthrobacter sp. SU Schmitz et al. 1992

27

Table 1. Continues

Enzyme

Reference

Methylmalonyl-CoA decarboxylase E. coli Haller et al. 2000

3-Hydroxyisobutyryl-CoA hydrolases

Human Hawes et al. 1996 Arabidopsis Zolman et al. 2001

Naphthoate synthases

E. coli Sharma et al. 1992 Bacillus subtilis Driscoll & Taber 1992

Others

Rhodopseudomonas palusris 2-ketocyclohexanecarboxyl-CoA hydrolase

Pelletier & Harwood 1998

Rhodococcus sp 6-oxocamphor hydrolasee

Grogan et al. 2001

Pseudomonas fluorescens feryloyl-CoA hydratase/HMPHP-CoA lyasef

Gasson et al. 1998

a In MFE-1, the α-subunit of MTP and the fadB gene product the 2-enoyl-CoA hydratase-1 activity is located in the N-terminal part of the polypeptide. b MTP, mitochondrial trifunctional protein c The enoyl-CoA isomerase activity is located in the same domain as the 2-enoyl-CoA hydratase-1 activity. d ChcB, 2-cyclohexenylcarbonyl-CoA isomerase e Not CoA-dependent f HMPHP, 4-hydroxy-3-methoxyphenyl-β-hydroxypropionyl-CoA

28

2.4.1.1 Trans-2-enoyl-CoA hydratase-1

Trans-2-enoyl-CoA hydratase-1 (hydratase-1) can occur either as a monofunctional enzyme or as an integral part of a multifunctional enzyme, and these vary in substrate chain length specificity, as described in the section “The β-oxidation cycle”. Hydratase-1 catalyzes the second step of β-oxidation, the addition of water to the trans-2 double bond generating L-3-hydroxyacyl-CoA (Figs 2, 6) (Willadsen & Eggerer 1975). The mitochondrial short-chain-specific, monofunctional 2-enoyl-CoA hydratase-1, also called crotonase, is the most intensively studied hydratase-1, and it has been purified from rat, bovine and pig (Hass & Hill 1969, Fong & Schulz 1977, Furuta et al. 1980). All these hydratases have similar properties. In solution, they are hexamers formed of six identical subunits and they have the highest catalytic activity towards crotonyl-CoA (trans-2-butyryl-CoA) as substrate. The cDNA sequences of rat and human hydratase-1s have been determined (Minami-Ishii et al. 1989, Kanazawa et al. 1993). In fact, the rat hydratase-1 was the first monofunctional member of the hydratase/isomerase superfamily characterized at the molecular level, and it is also the most thoroughly investigated one. The cDNA of rat hydratase-1 encodes a polypeptide of 290 amino acid residues, including a 29-residue N-terminal mitochondrial targeting sequence. The calculated molecular mass of the mature polypeptide is 28.3 kDa, which is in agreement with the size of the polypeptide isolated from rat liver, 26 kDa (Furuta et al. 1980, Minami-Ishii et al. 1989). Rat hydratase-1 is a very fast and efficient enzyme with a reaction rate close to being diffusion-controlled; its kcat towards crotonyl-CoA is 2100-5700 sec-1 and kcat/Km is 2.8 x 108 sec-1M-1 (Furuta et al. 1980, Fersht 1999). The reaction rate, however, decreases almost linearly with increasing substrate chain length (Furuta et al. 1980).

Hydratase-1 activity as part of a multifunctional protein has been characterized in peroxisomal MFE-1 (Osumi et al. 1979), the α-subunit of MTP (Kamijo et al. 1993) and the α-subunit of the E. coli fatty acid degradation complex, encoded by the fadB gene (Dirusso 1990, Yang et al. 1991). In all of these enzymes, the hydratase-1 activity is located in the N-terminal part of the polypeptide, since this region shows significant sequence similarity to rat hydratase-1 (Ishii et al. 1987, Minami-Ishii et al. 1989, Dirusso 1990, Kamijo et al. 1993). The rate of the hydratase-1 reaction catalyzed by multifunctional enzymes is considerably lower than that of rat 2-enoyl-CoA hydratase-1. For example, the kcat of rat MFE-1 towards crotonyl-CoA is 895 sec-1, i.e. up to six-fold lower when compared to hydratase-1 (Furuta et al. 1980). The C-terminal parts of the multifunctional proteins are also similar in sequence with each other, and they contain 3-hydroxyacyl-CoA dehydrogenase activity. Peroxisomal MFE-1 has, in addition, ∆3-∆2-enoyl-CoA isomerase activity sharing the active site with hydratase-1 (Palosaari & Hiltunen 1990, Palosaari et al. 1991), and the fadB gene product contains the ∆3-∆2-enoyl-CoA isomerase and 3-hydroxyacyl-CoA epimerase activities (Yang et al. 1988). The multifunctional proteins with hydratase-1 activity characterized from other sources are shown in Table 1. All these proteins share significant sequence similarity.

A 78-kDa gastrin-binding protein (GBP) purified from porcine gastric mucosal membranes (Baldwin et al. 1986) was suggested to be a component of the gastrin receptor controlling gastrin-dependent acid secretion in parietal cells (Mu et al. 1987). Its subsequent sequence determination surprisingly showed it to have significant sequence

29

similarity with MFE-1 in both the N-terminal and the C-terminal parts (Mantamadiotis et al. 1993). The molecular characterization of the α-subunit of the pig MTP, however, showed its sequence to be identical to the sequence reported to belong to GBP (Yang et al. 1994a). This makes the reported GBP unlikely to serve as a gastrin receptor. The ability of the assumed GBP to bind gastrin was explained by the finding that gastrin and the GBP antagonist benzotrip bind to the trifunctional protein and inhibit its enzyme activities (Hashimoto et al. 1996).

The messenger RNA of some lymphokines and proto-oncogenes are rapidly degraded via a signal provided by an AU-rich element within the 3´-untranslated region of the transcripts. Nakagawa and co-workers (1995) purified and cloned a human 32-kDa protein named AUH that binds specifically to AU-rich transcripts. Surprisingly, the protein had sequence homology to 2-enoyl-CoA hydratase-1, but not to the known RNA-binding proteins, and the recombinant protein possessed a low degree of hydratase-1 activity. The physiological substrate of the protein is unlikely to be crotonyl-CoA because the ability of the enzyme to use it as a substrate was about 1000-fold less than that of bovine hydratase-1. The corresponding mouse AUH has also been characterized, and it was shown to be 94 % identical with its human counterpart (Brennan et al. 1999). Interestingly, AUH was found to be located in mitochondria, suggesting that it could provide a link between the mitochondrial metabolic pathways and RNA stability (Brennan et al. 1999).

The CaiD gene of E. coli was previously suggested to encode for a carnitine racemase in the carnitine pathway (Eichler et al. 1994). Recent studies have shown, however, that, in addition to having a 30 % identical amino acid sequence compared to 2-enoyl-CoA hydratase-1, CaiD also catalyzes the hydratase-1 reaction. More specifically, it hydrates crotonobetainyl-CoA to L-carnitinyl-CoA, and it was thus renamed crotonobetainyl-CoA hydratase or carnitinyl-CoA dehydratase (Elssner et al. 2001).

2.4.1.2 ∆3-∆2-Enoyl-CoA isomerase

The ∆3-∆2-Enoyl-CoA isomerase is required in the β-oxidation of unsaturated fatty acids, and it catalyzes the reaction converting cis-3-enoyl-CoA or trans-3-enoyl-CoA into trans-2-enoyl-CoA, the substrate of 2-enoyl-CoA hydratase-1 in the β-oxidation cycle (Figs 3, 6, see “β-oxidation of (poly)unsaturated fatty acids”) (Hiltunen & Qin 2000). Similarly to 2-enoyl-CoA hydratase-1, the ∆3-∆2-enoyl-CoA isomerase can occur either as a monofunctional enzyme or as part of a multifunctional enzyme, as described above. Monofunctional isomerases have been purified from several sources, including the mitochondria of rat (Stoffel & Grol 1978, Palosaari et al. 1990, Müller-Newen & Stoffel 1991), bovine (Euler-Bertram & Stoffel 1990) and human (Kilponen & Hiltunen 1993) as well as the peroxisomes of cucumber seedlings (Engeland & Kindl 1991), human and mouse (Geisbrecht et al. 1999b).

The rat mitochondrial enoyl-CoA isomerase is a basic protein with a subunit size of 29 kDa and pI 9.7 (Palosaari et al. 1990, Müller-Newen & Stoffel 1991, Palosaari et al. 1991). The subunit sizes and the amino acid sequences of the rat and human mitochondrial enoyl-CoA isomerases are highly similar with 74 % sequence identity

30

(Kilponen et al. 1994). Interestingly, however, the pI of the human enzyme is three pH units more acidic compared to its rat counterpart. This difference is explained by the fact that many basic amino acid residues in the rat enoyl-CoA isomerase have been changed to neutral of basic ones in the human enzyme (Kilponen et al. 1994). Another difference is that the human enoyl-CoA isomerase was reported to be a homodimer with a native molecular mass of 70 kDa (Kilponen & Hiltunen 1993), whereas the rat enzyme was found to be a trimer (Palosaari et al. 1990, Zeelen et al. 1992). Recent studies in our laboratory have suggested, however, that the human enoyl-CoA isomerase is also most probably a trimer. The rat and human enoyl-CoA isomerases also differ in their substrate specificity, the rat enzyme being most active towards short-chain 3-enoyl-CoAs (Palosaari et al. 1991) and the human isomerase having, in contrast, no clear-cut substrate chain length specificity (Kilponen & Hiltunen 1993). A finding showing that there is also a long-chain-specific enoyl-CoA isomerase in rat mitochondria (Kilponen et al. 1990) could explain why the rat enoyl-CoA isomerase described here metabolizes only short-chain substrates. In human, no such enzyme has been characterized, suggesting that the mitochondrial enoyl-CoA isomerase should be able to use 3-enoyl-CoAs of all chain lengths as substrates (Kilponen & Hiltunen 1993).

Recently, a monofunctional ∆3-∆2-enoyl-CoA isomerase has also been characterized from the peroxisomes of human and mouse (Geisbrecht et al. 1999b). This enzyme, called PECI, has a somewhat higher molecular mass, 39.4 kDa, compared to the other monofunctional members of the hydratase/isomerase superfamily. It was found that, in addition to the hydratase/isomerase-like sequence, PECI has an extra domain about 80 amino acids in length at its N-terminus. This domain has significant sequence identity compared to the acyl-CoA binding protein (ACBP). ACBP is believed to be involved in the intracellular acyl-CoA transport and pool formation as well as in the control of fatty acid metabolism. ACBP exists in many isoforms and can also occur as a subdomain in acyl-CoA metabolizing enzymes, as in the case of PECI (Knudsen et al. 2000).

A novel ∆3-∆2-enoyl-CoA isomerase has been recently characterized from Streptomyces collinus. This enzyme, named 2-cyclohexenylcarbonyl-CoA isomerase (ChcB), is involved in the production of a polyketide antifungal antibiotic called ansatrienin A by catalyzing the isomerization of 2-cyclohexenylcarbonyl-CoA to 1-cyclohexenylcarbonyl-CoA (Patton et al. 2000). ChcB has wide substrate specificity, being able also to metabolize straight-chain 3-enoyl-CoAs and even with higher specific activity than 2-cyclohexenylcarbonyl-CoA. This could imply that 2-cyclohexenyl-carbonyl-CoA is not the physiological substrate for ChcB. However, the disrupted chcB mutant was able to grow on unsaturated fatty acids, whereas ansatrienin biosynthesis was blocked (Patton et al. 2000).

2.4.1.3 ∆3,5-∆2,4-Dienoyl-CoA isomerase

∆3,5-∆2,4-Dienoyl-CoA isomerase takes part in the metabolism of double bonds at odd-numbered positions in unsaturated fatty acids together with 2,4-dienoyl-CoA reductase and ∆3-∆2-enoyl-CoA isomerase (see “β-oxidation of (poly)unsaturated fatty acids”). It catalyzes the conversion of 3,5-dienoyl-CoA to 2,4-dienoyl-CoA, the substrate of

31

dienoyl-CoA reductase (Figs 3, 8). Only one mammalian ∆3,5-∆2,4-dienoyl-CoA isomerase, namely the rat one, has been characterized at the molecular level. The cDNA was isolated by FitzPatrick and colleagues (1995) by screening peroxisome proliferator-induced genes in rat liver. The cDNA was proposed to encode for a 2-enoyl-CoA hydratase-1 on the basis of its sequence similarity with rat hydratase-1, and it was subsequently named rECH1 (FitzPatrick et al. 1995). Filppula and co-workers (1998) expressed rECH1 as a recombinant protein, starting from residue 54, and purified and characterized the gene product. It was discovered that rECH1 does not encode for a hydratase-1 but, instead, for a ∆3,5-∆2,4-dienoyl-CoA isomerase having a specific activity of 0.12 µmol/min/mg towards 3,5,8,11,14-eicosapentaenoyl-CoA as substrate. Low but detectable levels of hydratase-1 activity (5nmol/min/mg) were also recorded, but no ∆3-∆2-enoyl-CoA isomerase activity was found (Filppula et al. 1998). An antibody raised against the rat ∆3,5-∆2,4-dienoyl-CoA isomerase recognized two polypeptides with molecular sizes of 32 kDa and 36 kDa, respectively, the smaller one being in the mitochondrial and the larger in the peroxisomal fraction. Dual distribution of the same gene product was further verified by immunoelectron microscopy. The amino acid sequence of ∆3,5-∆2,4-dienoyl-CoA isomerase contains both the N-terminal mitochondrial targeting sequence and the C-terminal tripeptide targeting it to peroxisomes (Filppula et al. 1998). Thus, it seems likely that, upon mitochondrial import, the targeting sequence is cleaved off, giving rise to the 32-kDa species, and that the full-length 36-kDa polypeptide is taken to the peroxisomes without modifications. A recent study has argued, however, that the peroxisomal form of dienoyl-CoA isomerase is also 32 kDa in size and has a truncated N-terminus (Zhang et al. 2001). The occurrence of the same gene product in two different organelles also explains the finding by He and co-workers (1995) suggesting that both the peroxisomal and the mitochondrial ∆3,5-∆2,4-dienoyl-CoA isomerase share similar chain length specificities and that the antibody against the mitochondrial isoenzyme also recognizes the peroxisomal one.

The presence of a novel enzyme, ∆3,5,7-∆2,4,6-trienoyl-CoA isomerase, acting in the β-oxidation of unsaturated fatty acids with conjugated double bonds was detected in rat and pig mitochondria (Liang et al. 1999). Zhang and co-workers (2001) found, however, that the trienoyl-CoA isomerase is actually the same enzyme as the ∆3,5-∆2,4-dienoyl-CoA isomerase. To study its enzymatic properties in more detail, the mature mitochondrial form of dienoyl-CoA isomerase, starting from residue 35, was expressed as a recombinant protein and its properties were characterized (Zhang et al. 2001). A specific ∆3,5-∆2,4-dienoyl-CoA isomerase activity of 2450 µmol /min/mg was measured, which was considerably higher than that detected previously (Filppula et al. 1998). In addition, a ∆3-∆2-enoyl-CoA isomerase activity of 0.034 µmol/min/mg was detected, which had not been recorded earlier (Filppula et al. 1998). Furthermore, the enzyme catalyzed the isomerization of ∆3,5,7- trienoyl-CoA to ∆2,4,6-trienoyl-CoA at a rate of 48 µmol/min/mg (Zhang et al. 2001).

In S. cerevisiae, a gene encoding the ∆3,5-∆2,4-dienoyl-CoA isomerase has been identified (Gurvitz et al. 1999). The gene product, Yor180cp/Dci1p, catalyzes the dienoyl-CoA isomerase reaction but is dispensable for the β-oxidation of unsaturated fatty acids with odd-numbered double bonds. Apparently, the degradation of odd-numbered double bonds occurs via the isomerase-dependent route in S. cerevisiae (Gurvitz et al. 1999).

32

2.4.1.4 4-Chlorobenzoyl-CoA dehalogenase

Halogenated hydrocarbons, such as chloroaromatics, which are used widely as industrial and agricultural agents, constitute a major class of environmental pollutants, which are inefficiently detoxified by conventional methods. A new strategy for the disposal of these pollutants is the use of microbial biodegradation. One of the bacterial strains capable of metabolizing chlorinated hydrocarbons is Pseudomonas sp. strain CBS-3, which can use 4-chlorobenzoate (4-CBA) as its sole carbon source. 4-CBA is converted into 4-hydroxybenzoate (4-HBA) by an enzyme system that catalyzes the aromatic substitution of the chloride with a hydroxyl group from a water molecule (Müller et al. 1984). The enzyme complex, called the 4-CBA dehalogenase complex, consists of three polypeptides, namely the 57-kDa 4-CBA:CoA ligase activating 4-CBA to 4-CBA-CoA, the 30-kDA 4-CBA-CoA dehalogenase catalyzing the aromatic substitution reaction (Fig. 7) and the 16-kDa thioesterase cleaving 4-HBA-CoA into 4-HBA and CoA (Scholten et al. 1991). The cloning and sequencing of the components of the enzyme complex revealed that the sequence of the 30-kDa 4-CBA-CoA dehalogenase is similar to 2-enoyl-CoA hydratase-1 (Minami-Ishii et al. 1989) and the other members of the hydratase/isomerase superfamily characterized so far (Babbitt et al. 1992). The 4-CBA-CoA dehalogenase and 2-enoyl-CoA hydratase-1 were suggested to be both evolutionary and also mechanistically related, since they both require the activation of water for addition across a carbon-carbon bond in catalysis (Babbitt et al. 1992). Similarly to the rat ∆3-∆2-enoyl-CoA isomerase, the 4-CBA-CoA dehalogenase forms homotrimers (Benning et al. 1996).

2.4.1.5 Methylmalonyl-CoA decarboxylase

The genome-sequencing project of E. coli revealed that its genome contains seven paralogues of the hydratase/isomerase superfamily, four with unknown function. One of the unknown genes was ygfG, a member of a four-gene operon also containing the genes sbm, ygfD and ygfH (Haller et al. 2000). This operon was found likely to encode enzymes forming a metabolic cycle which decarboxylates succinate to propionate. The metabolic context of this pathway was, however, left unresolved (Haller et al. 2000). The reaction catalyzed by YgfG was determined to be the decarboxylation of methylmalonyl-CoA to propionyl-CoA, a new reaction in the hydratase/isomerase superfamily (see also Fig. 9). YgfG was thus renamed methylmalonyl-CoA decarboxylase. Similarly to 2-enoyl-CoA hydratase-1 and dienoyl-CoA isomerase, methylmalonyl-CoA decarboxylase is also a hexamer consisting of six identical 29 kDa subunits (Benning et al. 2000, Haller et al. 2000).

2.4.1.6 Other members

Dihydroxynaphthoate synthase catalyzes a ring closure reaction forming 1,4-dihydroxy-2-naphthoic acid from o-succinylbenzoyl-CoA (Meganathan & Bentley 1979). This is a

33

reaction involved in the biosynthesis of a bacterial electron carrier menaquinone (vitamin K2). The gene encoding dihydroxynaphthoate synthase, menB, has been sequenced from E. coli (Sharma et al. 1992) and Bacillus subtilis (Driscol & Taber 1992).

In contrast to the ring closure reaction of dihydroxynaphthoate synthase, another hydratase/isomerase family member, 2-ketocyclohexanecarboxyl-CoA (2-ketochc-CoA) hydrolase catalyzes a hydrolytic ring cleavage of 2-ketochc-CoA to pimelyl-CoA, which is required in the anaerobic benzoate degradation in Rhodopseudomonas palustris (Egland et al. 1997, Pelletier & Harwood 1998). The 2-ketochc-CoA hydrolase is encoded by the badI gene located in a cluster of anaerobic benzoate degradation genes (Egland et al. 1997) and highly induced when grown on benzoate. Interestingly, 2-ketochc-CoA hydrolase shares the highest amino acid sequence identity, about 45 %, with dihydroxynaphthoate synthase, which reflects the similar nature of the reactions they catalyze, involving either ring closure or cleavage steps (Pelletier & Harwood 1998).

3-Hydroxyisobutyryl-CoA (HIB-CoA) hydrolase is a vital enzyme catalyzing the hydrolytic cleavage of HIB-CoA to free CoA and 3-hydroxyisobutyrate in the valine catabolic pathway. This is an interesting reaction, since the destruction of an activated intermediate is rare in metabolic pathways, especially if the later steps of the pathway also require CoA esters as substrates, as in this case. HIB-CoA hydrolase is, however, essential for the removal of a toxic intermediate of the pathway, methacrylyl-CoA, which is detoxified by two subsequent reactions, hydration to HIB-CoA by hydratase-1 and cleavage by HIB-CoA hydrolase (Shimomura et al. 1994). The toxicity of methacrylyl-CoA is probably due to its ability to react with free thiol groups of proteins, thereby inactivating them (Brown et al. 1982). HIB-CoA hydrolase has been purified from rat liver (Shimomura et al. 1994), and the cDNA sequence of the human enzyme has been determined (Hawes et al. 1996). The enzyme is specifically active with a very high turnover rate towards HIB-CoA and lesser turnover towards 3-hydroxypropionyl-CoA (Shimomura et al. 1994, Hawes et al. 1996). These powerful catalytic properties and the strict substrate specificity are required for efficient removal of methacrylyl-CoA during valine catabolism and for prevention of the interference of HIB-CoA hydrolase with the catabolism of fatty acids, leucine and isoleucine. Surprisingly, the cDNA of HIB-CoA hydrolase had no similarity to other thioesterases but, instead, showed striking homology to the hydratase/isomerase superfamily members (Hawes et al. 1994). Thus, the hydratase/isomerase proteins catalyze two subsequent reactions in the valine catabolic pathway. Recently, a plant HIB-CoA hydrolase encoded by an Arabidopsis gene CHY1 has also been characterized (Zolman et al. 2001). The chy1 mutant is resistant to a plant hormone indole-3-butyric acid and also has impaired β-oxidation (Zolman et al. 2001). CHY1 is located in the peroxisomes, where the β-oxidation of plants also occurs, and shares 43 % sequence identity with the mammalian mitochondrial HIB-CoA hydrolase. The metabolic function of CHY1 was confirmed by expressing it as a recombinant protein and by showing that the expressed protein had specific HIB-CoA hydrolase activity (Zolman et al. 2001). The impaired β-oxidation and hormone sensitivity in the mutant were suggested to be due to the inactivation of the enzymes of the respective pathways by accumulating methacrylyl-CoA (Zolman et al. 2001).

Yet another different reaction catalyzed by an enzyme belonging to the hydratase/isomerase superfamily was characterized from Pseudomonas fluorescens. This novel enzyme was found to be involved in the metabolism of ferulic acid (4-hydroxy-3-

34

methoxy-trans-cinnamic acid) to vanillin by catalyzing both the hydration of feruloyl-CoA to 4-hydroxy-3-methoxyphenyl-β-hydroxypropionyl-CoA (HMPHP-CoA) and its subsequent cleavage to vanillin and acetyl-CoA (Gasson et al. 1998). The enzyme, feruloyl-CoA hydratase/HMPHP-CoA lyase could have potential use in the biotechnological production of vanillin, one of the principal flavouring compounds in the world. Since feruloyl-CoA is a breakdown product of lignin and plant cell wall material, plant waste could be used as starting material for vanillin production (Gasson et al. 1998, Priefert et al. 2001).

The enzyme 6-oxocamphor hydrolase from Rhodococcus makes an interesting exception to the rule that hydratase/isomerase proteins only accept CoA esters as substrates. It catalyzes the asymmetric hydrolysis of 6-oxocamphor to (2R,4S)-α-campholinic acid as well as the hydrolysis of some other bicyclic β-diketones (Grogan et al. 2001). The characterization of 6-oxocamphor hydrolase adds a new enzyme to the group of biocatalysts that could be employed in the synthesis of chemical intermediates (Grogan et al. 2001).

2.4.2 Structure and function of the hydratase/isomerase proteins

2.4.2.1 Structural aspects

As mentioned above, the crystal structures of five members of the hydratase/isomerase superfamily are known, those being the rat 2-enoyl-CoA hydratase-1 (Engel et al. 1996, 1998), the 4-CBA-CoA dehalogenase from Pseudomonas (Bennning et al. 1996), the rat ∆3,5-∆2,4-dienoyl-CoA isomerase (Modis et al. 1998), the methylmalonyl-CoA decarboxylase from E. coli (Benning et al. 2000) and the very recently published human AUH protein, a RNA-binding homologue of 2-enoyl-CoA hydratase-1 (Kurimoto et al. 2001). All these structures have a similar overall fold consisting of an N-terminal core domain, containing the first 200-250 amino acid residues, and a C-terminal trimerization domain (Fig. 5). The N-terminal domain has a spiral fold formed of four turns, each turn being shaped by two β-strands and an α-helix. The β-strands form two β-sheets, which are almost at right angles with respect to each other and connected by the α-helices. The C-terminal domain contains four α-helices that are mainly used for subunit-subunit contacts and also for shaping the acyl-CoA-binding pocket. In the structures of hydratase-1, dienoyl-CoA isomerase, 4-CBA-CoA dehalogenase and the AUH protein, the C-terminal helices H9 and H10 protrude out of the core domain and cover the active site of the adjacent subunit. In the methylmalonyl-CoA decarboxylase structure, however, the active site is fully contained within one subunit since the C-terminal domain folds over the active site of the same subunit (Benning et al. 2000). All the known hydratase/isomerase structures assemble into disk-like trimers via contacts by the C-terminal domain; in addition, in all except 4-CBA-CoA dehalogenase, two trimers bind together to form a hexamer. Thus, the hexamers are described to be dimers of trimers.

The human AUH protein differs from the other hydratase/isomerase proteins in that its physiological function is not likely to be enzymological but to bind the AU-rich elements

35

of single-stranded RNA (Nakagawa et al. 1995, Kurimoto et al. 2001). The AUH protein is able to catalyze the hydratase-1 reaction, but the reaction rate is so low that the activity is rather residual than of any physiological significance (Nakagawa et al. 1995, Kurimoto et al. 2001). The binding of RNA is accomplished by the clefts between the trimers in the hexamer, the edges of the clefts being formed by the α-helix H1 of each subunit. The 20-residue segment, the R-peptide, which has been found to bind to the AU-rich elements of RNA (Nakagawa & Moroni 1997), is mostly contained within the helix H1. This segment is rich in positively charged lysine residues, which form the “lysine comb” that could bind the negatively charged phosphate groups of RNA. The mutation of the lysines to negatively charged residues abolishes the RNA binding. In addition, the cleft between the trimers is wide enough to accommodate a RNA molecule (Kurimoto et al. 2001). In contrast, in 2-enoyl-CoA hydratase-1, the helix H1 is negatively charged and no RNA-binding activity for it has been detected (Kurimoto et al. 2001).

Of 2-enoyl-CoA hydratase-1, 4-CBA-CoA dehalogenase and methylmalonyl-CoA decarboxylase also structures with an active site ligand exist (Engel et al. 1996, 1998; Benning et al. 1996, 2000). The acyl-CoA molecule is bound in such a way that the 3´-phosphate ADP and pantothenic acid moieties are bound in a curved conformation on the outside of the core domain against a β-sheet, whereas the acyl part slides into the substrate-binding pocket inside the protein (Fig. 5). The shape of the binding pocket is mainly determined by the conformation of the α-helix H2 of the core domain and the helices H9 and H10 of the trimerization domain. The variability in these regions enables the binding of different substrates at the active sites. More detailed discussion on the structures of the hydratase/isomerase superfamily will follow in the ”Discussion” section.

Fig. 5. The fold of the rat 2-enoyl-CoA hydratase-1 (pdb-entry code 1DUB) monomer. The N- and C-termini are marked, as are also the αααα-helices H1-H10. The ββββ-strands are shown as arrows. The ligand acetoacetyl-CoA bound at the active site is presented as a ball-and-stick model.

36

2.4.2.2 Catalytic mechanisms

The first sequenced members of the hydratase/isomerase superfamily, mono- and multifunctional 2-enoyl-CoA hydratases and ∆3-∆2-enoyl-CoA isomerases, were suggested to be, in addition to their evolutionary relationship, also mechanistically related, since both the hydratase-1 and the isomerase reactions involve protonation or deprotonation at the α-carbon (C2) of the substrate (Fig. 6) (Müller-Newen & Stoffel 1993). The hydratase-1 reaction consists of an addition of a hydroxyl group from an activated water molecule to the β-carbon (C3) and the protonation of the α-carbon (Fig 6). The reaction follows the syn stereochemistry, in which both the proton and the hydroxyl group are added to the same side of the planar α,β-double bond (Willadsen & Eggerer 1975). In the reaction catalyzed by the ∆3-∆2-enoyl-CoA isomerase, a proton is abstracted from the α-methylene group of the substrate and subsequently donated to the γ-carbon (C4), resulting in a shift of the double bond from the β,γ-position to the α,β-conformation (Fig. 6) (Müller-Newen & Stoffel 1993). The initial sequence comparisons and mutagenesis studies showed that Glu164 of hydratase-1 was the only protic amino acid residue completely conserved among the characterized hydratase/isomerase superfamily members (Müller-Newen & Stoffel 1993, Müller-Newen et al. 1995). Glu164 (Glu165 in the rat enoyl-CoA isomerase) was confirmed to be essential for the catalytic activity of both the rat 2-enoyl-CoA hydratase-1 and the enoyl-CoA isomerase by site-directed mutagenesis, suggesting that it is probably involved in the protonation/deprotonation step (Müller-Newen & Stoffel 1993, Müller-Newen et al. 1995). The determination of the crystal structure of rat hydratase-1 showed another protic residue, Glu144, to be also involved in catalysis (Engel et al. 1996). This residue is also completely conserved in the proteins catalyzing the hydratase-1 reaction, including the multifunctional proteins, but not in monofunctional ∆3-∆2-enoyl-CoA isomerases. It was proposed that Glu144 acts as a catalytic base and activates the water molecule for the attack at C3, whereas Glu164 is the catalytic acid protonating C2 (Engel et al. 1996). The water molecule needed for activation is seen to be bound between Glu144 and Glu164 in the crystal structure of 2-enoyl-CoA hydratase-1 (Engel et al. 1996). Further mutational studies (Kiema et al. 1999) confirmed the necessity of Glu144 for the hydratase-1 reaction, the Glu144Ala variant having 2000-fold lower hydratase-1 activity compared to the wild type. Kiema and co-workers (1999) further found that rat hydratase-1 also possesses slight residual ∆3-∆2-enoyl-CoA isomerase activity. Glu144Ala and Glu164Ala mutants were tested for isomerase activity, and the mutation of Glu164 was found to totally abolish isomerase activity, while the activity in the Glu144Ala variant was only lowered 10-fold. These findings are in line with the suggestion that Glu164 is needed for both the hydratase-1 and the enoyl-CoA isomerase activities, whereas Glu144 is only essential for the water activation reaction of hydratase-1 (Kiema et al. 1999). The amino acid residue responsible for the protonation of the γ-carbon in the isomerization reaction is not known, the proton could be donated by a water molecule or the deprotonation and the protonation could be accomplished by the same residue. A crystal structure is needed to determine the active site architecture of the ∆3-∆2-enoyl-CoA isomerase. Interestingly, in the mammalian monofunctional peroxisomal ∆3-∆2-enoyl-CoA isomerase, the glutamate at position 164 (hydratase-1 numbering) is not conserved (Geisbrecht et al.

37

1999b), nor is it in the novel bacterial isomerase, 2-cyclohexenylcarbonyl-CoA isomerase (Patton et al. 2000).

Spectroscopic studies have shown that, in 2-enoyl-CoA hydratase-1, the substrate is activated for catalysis by forming a strong hydrogen bond from a protein atom to the oxygen of the thioester carbonyl group of the substrate (D´Ordine et al. 1994). In fact, crystallographic and mutation studies have indicated that this hydrogen bond in hydratase-1 is made by the main chain NH group of Gly141, and it is essential for the catalysis to occur (Engel et al. 1996, Bell et al. 2001). Hydrogen bonding from Gly141 together with the NH group of Ala98 form the oxyanion hole, which also stabilizes the negative charge forming on the thioester oxygen during the transition state of the enzymatic reaction (Engel et al. 1996, 1998). The ability to stabilize the intermediate seems to be a common requirement in the hydratase/isomerase superfamily, since the locations of the oxyanion hole residues are conserved in all of the structurally characterized members (Babbitt & Gerlt 1997, Holden et al. 2001).

O-

Glu144

OH

H

R

O

SCoA

Glu164

OH

R

O-

SCoA

Glu164

OH

Glu144

OH

R

O

SCoA

Glu164

O-

Glu144

OHOH

R

O

SCoAR

O-

SCoAR

H

B1-B1

H

B2

H

B2

H H

B2-

H

B1

O

SCoA

A

B

H2O

Fig. 6. The reaction mechanisms of rat 2-enoyl-CoA hydratase-1 (A) and ∆∆∆∆3-∆∆∆∆2-enoyl-CoA isomerase (B) (adopted from Kiema et al. 1999). The reversible reaction of 2-enoyl-CoA hydratase-1 is shown in the direction of dehydration. In the reverse reaction, the hydration, Glu144, the general base activates a water molecule for an attack at the C3 of the trans-2-enoyl-CoA. The oxyanion hole residues Ala98 and Gly141 stabilize the negative charge forming on the thioester carbonyl oxygen during the intermediate state of the reaction. Glu164 acts as a catalytic acid and protonates C2 to form L-3-hydroxyacyl-CoA. In the reaction mechanism of enoyl-CoA isomerase, a double bond is shifted from C3 to C2. The catalytic base B1, in the case of the rat mitochondrial enoyl-CoA isomerase Glu165, abstracts a proton from C2, leading to an anionic transition state. Another unknown residue (B2) donates a proton to the C4, completing the reaction. The intermediates of the reactions are shown in brackets.

38

The dehalogenation of 4-chlorobenzoyl-CoA (4-CBA-CoA) to 4-hydroxybenzoyl-CoA (4-HBA-CoA) by 4-CBA-CoA dehalogenase occurs via three reaction intermediates (Fig. 7) (Yang et al. 1994b). First, an active site carboxylate residue attacks the C4 of the benzoyl ring of the substrate, forming a covalent Meisenheimer complex. Second, the chloride ion leaves, rearomatizing the benzoyl ring and generating the aryl enzyme intermediate. Third, an activated water molecule adds to the acyl carbonyl carbon, leading to a tetrahedral intermediate, from which the product 4-HBA-CoA is formed and the catalytic carboxylate generated (Yang et al. 1994b, Yang et al. 1996). Mutagenesis and structural studies revealed that the active site nucleophilic carboxylate forming the covalent complex is Asp145 (Benning et al. 1996, Yang et al. 1996), a residue not conserved in 2-enoyl-CoA hydratases. The catalytic base activating the water molecule was found to be His90, which is not conserved in the previously characterized hydratase/isomerase family members, either (Benning et al. 1996, Yang et al. 1996). In addition, Trp137 is important in providing a hydrogen bond to Asp145 (Benning et al. 1996, Yang et al. 1996) and thereby correctly positioning the catalytic residue with respect to the substrate (Lau & Bruice 2001). In the Meisenheimer intermediate, a negative charge develops on the thioester carbonyl oxygen. As in 2-enoyl-CoA hydratase-1, this charge is stabilized by the hydrogen bonds in the oxyanion hole, which is formed by the NH groups of Phe64 and Gly114 (Benning et al. 1996). Similarly, these hydrogen bonds are also involved in the activation of the substrate for catalysis by polarizing the 4-CBA-CoA. In addition, the α-helix terminating in Gly114 provides a dipolar electrostatic component, which contributes to the polarization of the thioester carbonyl oxygen and the whole benzoyl moiety, thus facilitating the attack of Asp145 to C4 of the ring (Clarkson et al. 1997, Taylor et al. 1997).

O SCoA

Cl

C

O-

C

O

Asp Asp

O

O

O-

Cl

SCoA

Asp

Cl-

C

O

O

OO

H

SCoA

His

N

H

H2O

Asp

O-

C

O

O

OH

His

NH+

SCoA

Asp C

O

O

OHO-

SCoA

His

N

1. 2. 3.

Fig. 7. The dehalogenation of 4-chlorobenzoyl-CoA to 4-hydroxybenzoyl-CoA by the 4-CBA-CoA dehalogenase from Pseudomonas (modified from Benning et al. 1996 and Taylor et al. 1997). The reaction involves three intermediates: 1. The Meisenheimer complex, where Asp145 attacks the C4 of the benzoyl ring. The negative charge on the thioester carbonyl oxygen is stabilized at the oxyanion hole formed by Phe64 and Gly114. 2. Chloride leaves the complex and an arylated enzyme is generated. 3. His90 activates a water molecule for attack, and a tetrahedral intermediate is formed. The product 4-HBA-CoA leaves this complex and the catalytic residues are regenerated. The charged intermediates are shown in brackets.

39

∆3,5-∆2,4-Dienoyl-CoA isomerase catalyzes the simultaneous shift of two double bonds from the ∆3,5 position to the ∆2,4 conformation during the β-oxidation of unsaturated fatty acids with odd-numbered double bonds (Luo et al. 1994). According to the crystal structure of the rat ∆3,5-∆2,4-dienoyl-CoA isomerase, there are three protic amino acid residues at the active site of the enzyme, Asp176, Glu196 and Asp204 (Modis et al. 1998), of which Asp176 and Glu196 are equivalents to the catalytic residues of rat hydratase-1, Glu144 and Glu164 (Engel et al. 1996), and Asp204 corresponds to the active site aspartate-145 of 4-CBA-CoA dehalogenase (Benning et al. 1996, Yang et al. 1996). Glu196 and Asp204 are positioned correctly at the active site of the dienoyl-CoA isomerase for proton exchange from the carbons C2 and C6, whereas Asp176 is probably not directly involved in catalysis but, instead, optimizes the catalytic properties of Glu196 by hydrogen bonding to it. The importance of Asp204 for catalysis was also shown by mutagenesis (Modis et al. 1998). The following reaction mechanism was therefore proposed (Fig. 8): Glu196 would abstract the proton from C2 followed by rearrangements of the double bonds. To prevent the accumulation of charge in the substrate, Asp204 would donate a proton to the C6 of the substrate and the product would be released. As in the other superfamily members, the negative charge on the thioester oxygen forming during the transition state of the reaction is stabilized in the oxyanion hole residues Ile117 and Gly173 (Modis et al. 1998). In a recent mechanistic study on the ∆3,5-∆2,4-dienoyl-CoA isomerase, Asp176, Glu196 and Asp204 were mutated separately to nonprotic residues, and the characteristics of the mutants were investigated (Zhang et al. 2001). The mutation of either Glu196 or Asp204 lowered the dienoyl-CoA isomerase activity 105-fold compared to the wild type, whereas the Asp176Ala variant had only 10-fold lower activity. Moreover, Glu196 could catalyze the ∆2,∆3-isomerization of 2,5-dienoyl-CoA to 3,5-dienoyl-CoA when Asp204 was inactivated, and Asp204 was able to catalyze the ∆5,∆4-isomerization of 2,5-dienoyl-CoA to 2,4-dienoyl-CoA when Glu196 was mutated. However, these monoene isomerizations occurred at a much lower rate than the simultaneous diene isomerization in the wild type (Zhang et al. 2001). The results confirmed the suggestion by Modis and co-workers (1998) that Glu196 and Asp204 are involved in proton exchange from carbons C2 and C6, respectively. The ability of the ∆3,5-∆2,4-dienoyl-CoA isomerase to catalyze the isomerization of 2,5-dienoyl-CoA to 2,4-dienoyl-CoA without the 3,5-intermediate could imply that a separate monofunctional ∆3-∆2-enoyl-CoA isomerase is not necessarily needed in the pathway metabolizing odd-numbered double bonds. Zhang and colleagues (2001) also found that the ∆3,5-∆2,4-dienoyl-CoA isomerase is a ∆3,5,7-∆2,4,6-trienoyl-CoA isomerase and that Asp204 is the residue also catalyzing the shift of the double bond from ∆7 to ∆6 during the metabolism of fatty acids with conjugated double bonds.

40

R

O

SCoA

H

Glu

O-

Asp

OH

R

H

O-

Asp

OH

Glu

O

SCoA

Fig. 8. The reaction catalyzed by ∆∆∆∆3,5-∆∆∆∆2,4-dienoyl-CoA isomerase (modified from Modis et al. 1998). Glu196 acts as a catalytic base in the rat enzyme and abstracts a proton from the C2 of the substrate, 3,5-dienoyl-CoA. The acid Asp204 donates the proton to C6, and the product 2,4-dienoyl-CoA is released. The transition state of the reaction can be inferred from the ∆∆∆∆3-∆∆∆∆2-enoyl-CoA isomerase reaction in Fig. 6.

In the amino acid sequence of E. coli, methylmalonyl-CoA decarboxylase Glu144 of rat hydratase-1 is conserved as Glu113 (Haller et al. 2000). This is the first case where only Glu144 but not Glu164 is conserved. The crystal structure of methylmalonyl-CoA decarboxylase showed, surprisingly, that Glu113 is probably not directly involved in catalyzing the decarboxylation of methylmalonyl-CoA to propionyl-CoA, since its location is not appropriate with respect to the substrate and, in addition, it is hydrogen-bonded to Arg86 (Benning et al. 2000). Instead, Tyr140 is positioned in such a way that it could orient the carboxylate group of methylmalonyl-CoA properly for decarboxylation; therefore, Tyr140 is suggested to be the catalytic residue (Fig. 9). The nonpolar environment would destabilize the negatively charged carboxylate group and enhance the decarboxylation (Benning et al. 2000). The reaction also requires activation of the substrate by polarization of the thioester carbonyl oxygen and stabilization of the anionic transition state, which is accomplished by hydrogen bonding by the oxyanion hole residues, His66 and Gly110 (Benning et al. 2000). It is currently not known which residue serves as the general acid delivering a solvent-derived proton to the α-carbon to generate the product, propionyl-CoA.

The Glu144 of rat hydratase-1 is also conserved in the 6-oxocamphor hydrolase (Grogan et al. 2001) and the Glu164 in the HIB-CoA hydrolase (Hawes et al. 1996) and the feruloyl-CoA hydratase/lyase (Gasson et al. 1998). In the absence of structural and mutational data, the possible role of these residues in the reaction mechanisms can only be speculated. Similarly, nothing is known about the catalytic residues of the 2-ketochc-CoA hydrolase (Pelletier & Harwood 1998) and the dihydroxynaphthoate synthase (Sharma et al. 1992), since none of the catalytic residues identified in the other members are conserved in them.

41

H3C

O

SCoA

CO2

CO2-H

Tyr

OH

Tyr

H3C

OH

H

O-

H

A

SCoAH3C

Tyr

OH

H H

O

A-

SCoA

Fig. 9. The decarboxylation of methylmalonyl-CoA to propionyl-CoA catalyzed by the methylmalonyl-CoA decarboxylase from E. coli (adapted form Benning et al. 2000). Tyr140 orients the carboxylate group in such a way that the decarboxylation process is facilitated. The anionic intermediate is stabilized by hydrogen bonding to His66 and Gly110. It is unclear which residue (A) serves as a catalytic acid and protonates the intermediate so that propionyl-CoA can form.

As a conclusion, within the hydratase/isomerase superfamily, polypeptides with considerably low sequence identities (20-30 %) still fold into strikingly similar three-dimensional structures with the same active site design. Three conserved sites for catalytic amino acids have been described so far, the Glu144 and Glu164 of 2-enoyl-CoA hydratase-1 and the Asp145 of 4-CBA-CoA dehalogenase. Some members of the family use none of these residues for catalysis, indicating that more catalytic residues are yet to be identified. The common active site template with multiple catalytic “stations” supports the variety of catalytic reactions to occur within the hydratase/isomerase superfamily, and the evolution of enzymes catalyzing new reactions has probably taken place by mutating residues within the active site template (Xiang et al. 1999). To provide experimental evidence in support of this hypothesis, Xiang and co-workers (1999) engineered the active site residues of 4-CBA-CoA dehalogenase in such a way that the residues corresponding to the Glu144 and Glu164 of hydratase-1 were incorporated. Indeed, the engineered variant of 4-CBA-CoA dehalogenase was able to catalyze the 2-enoyl-CoA hydratase-1 reaction, the syn addition of water to a α,β-double bond. Above all, the evolution of this superfamily has probably been dominated by the need to stabilize the anionic transition state of the enzymatic reaction (Holden et al. 2001). This is highlighted by the fact that the two peptidic NH groups forming the oxyanion hole are conserved in all the structurally characterized members of the superfamily. Furthermore, the latter of the oxyanion hole residues is always at the end of an α-helix. This helix dipole adds to the polarizing effect of the oxyanion hole. Thus, all the proteins in this family have a common structural solution, the oxyanion hole, to the mechanistic problem of how to lower the free energy of the transition state in order for the reaction to proceed (Babbitt & Gerlt 1997, Holden et al. 2001).

3 Aims of the present study

The complete sequencing of the genome of baker’s yeast, S. cerevisiae, revealed that it has three genes encoding possible members of the hydratase/isomerase superfamily. The initial goal of the study was to start identifying the functions of these gene products and, especially, to search for the ∆3-∆2-enoyl-CoA isomerase or the ∆3,5-∆2,4-dienoyl-CoA isomerase of S. cerevisiae. To start the search, the first aim was:

1. To subclone the yeast gene YLR284c into a bacterial expression vector, to

overexpress it and to determine the enzymatic activity of the gene product.

Once the gene YLR284c had been found to encode for a ∆3-∆2-enoyl-CoA isomerase, more specific aims for the study were set:

2. To purify and characterize the recombinant yeast ∆3-∆2-enoyl-CoA isomerase. 3. To determine the crystal structure of the ∆3-∆2-enoyl-CoA isomerase in the

absence and presence of an active site ligand. 4. To compare the ∆3-∆2-enoyl-CoA isomerase structures with each other and with

the other known structures belonging to the hydratase/isomerase superfamily.

4 Materials and methods

Most of the methods are described in more detail in the original articles referred to by their Roman numerals.

4.1 Strains and plasmids

All the bacterial and yeast strains as well as the construction of the plasmids used in the experiments are described in the original articles.

4.2 Saccharomyces genome database

The sequence of the genome of S. cerevisiae is available (Cherry et al. 1997), and it was found to contain three open reading frames (ORF) encoding proteins with amino acid sequence similarity to 2-enoyl-CoA hydratase-1 and other members of the hydratase/isomerase superfamily of enzymes. These genes are YDR036c, YOR180c and YLR284c. YOR180c was later found to encode a ∆3,5-∆2,4-dienoyl-CoA isomerase and was renamed DCI1 (Gurvitz et al. 1999). The physiological function of the polypeptide encoded by YDR036c still remains unknown. YLR284c and its gene product Ylr284cp are the topic of this study and will be referred to as ECI1 and Eci1p, respectively.

4.3 Gene disruption (I)

The ECI1 knockout yeast strain was constructed by the short flanking homology method (Wach et al. 1994) based on polymerase chain reaction (PCR) -targeting following the European Functional Analysis Network guidelines. In this method, the target gene is replaced by a kanMX cassette, which gives the knockout strain resistance to the drug geneticin. To test the ability of the ECI1-disrupted strain to metabolize saturated and unsaturated fatty acids, the mutant strain was grown on palmitic (C16:0), oleic [cis-

44

C18:1(9)], octadecenoic [cis-C18:1(12)] or arachidonic acid [cis-C20:4(5, 8, 11, 14)] as the sole carbon source. The experiments were made in the Institute of Biochemistry and Molecular Cell Biology, University of Vienna, Austria.

4.4 Analysis of transcriptional regulation (I)

Northern blot analysis and electrophoretic mobility shift assay (EMSA) (Sambrook et al. 1989) were used to study whether the expression of ECI1 is up-regulated when cells are grown on oleic acid and whether this induction is mediated via a promoter region called oleate response element (ORE) (Einerhand et al. 1993, Filipits et al. 1993). The experiments were carried out in the Institute of Biochemistry and Molecular Cell Biology, University of Vienna, Austria.

4.5 Subcellular localization (I)

Although Eci1p has a modified peroxisomal targeting sequence type 1 (PTS1, HisArgLeuCOOH) in its carboxyl terminus, its subcellular localization was confirmed. ECI1 was fused to the 3’ end of the gene encoding the green fluorescent protein (GFP) in a plasmid, and the linearized plasmid was integrated into the trp1 locus of wild-type yeast cells (strain BJ1991) and cells lacking peroxisomes (BJ1991pex6∆) (Voorn-Brouwer et al. 1993). A fluorescence microscopy method was used to detect the localization of GFPs in yeast cells, as described earlier (Görner et al. 1998). The experiments were made in the Institute of Biochemistry and Molecular Cell Biology, University of Vienna, Austria.

4.6 Construction of the ECI1 expression vector (I)

To express Eci1p as a recombinant protein in E. coli, ECI1 was amplified from S. cerevisiae genomic DNA using PCR and gene-specific primers. Genomic DNA was isolated according to Ausubel and co-workers (1989) from BJ1991 cells. The PCR product was ligated into pUC plasmid (Pharmacia) and subsequently sequenced (Sanger et al. 1977) to verify the correct open reading frame. ECI1 was further subcloned into the pET3a expression vector (Novagen), and pET3a-ECI1 was transformed into the expression host, BL21(DE3)pLysS E. coli cells (Novagen).

4.7 Expression of the recombinant protein (I)

The over-expression of full-length Eci1p using the pET expression system (Studier 1991) was performed according to the manufacturer’s instructions. The induction of protein production was carried out at 30°C. The cells were separated, suspended in lysis buffer (20 mM potassium phosphate, pH 7.6, 100 mM KCl) and stored at -70°C.

45

4.8 Protein purification (I, II)

The bacterial cells were lysed by thawing the frozen cell suspension. Homogenation of the lysate was enhanced by adding DNase, RNase and lysozyme to the suspension and by incubating the mixture at +35°C for 15-30 minutes. The soluble protein-containing supernatant was separated by centrifugation. For the characterization of Eci1p (I), the recombinant protein was purified using three subsequent chromatographic steps. The columns used were DEAE Sephacel anion exchange, Resource S cation exchange and Superdex 200 HR 10/30 size exclusion columns. For the crystallization of Eci1p (II), the purification protocol was modified in order to obtain larger amounts of pure protein. Butyl sepharose was used as the first column, followed by hydroxyapatite and Poros SP columns. The purity of the protein was analyzed by sodium dodecyl sulphate polyacrylamine gel electrophoresis (SDS-PAGE) (Laemmli 1970) using 12% polyacrylamide gels and Coomassie Blue staining (Ausubel et al. 1989). The purified protein was stored at +4°C in 20 mM potassium phosphate, pH 7.2, 450 mM KCl, 1 mM EDTA, 1 mM EGTA, 0.5 mM benzamidine-HCl.

4.9 Enzyme assays (I)

The ∆3-∆2-Enoyl-CoA isomerase and 2-enoyl-CoA hydratase-1 activity measurements were performed as described by Palosaari and Hiltunen (1990) using 60 µM trans-3-hexenoyl-CoA and crotonyl-CoA as substrates, respectively. The reaction was monitored by following the generation of a magnesium complex of 3-ketoacyl-CoA at 303 nm using a Shimadzu UV 3000 spectrophotometer. The substrates were synthesized by Anna-Leena Hietajärvi and Werner Schmitz using the mixed anhydride method (Rasmussen et al. 1990).

For the determination of ∆3,5-∆2,4-dienoyl-CoA isomerase enzyme activity, the substrate, 3,5,8,11,14-eicosapentenoyl-CoA, was generated by incubating 60 µM arachidonoyl-CoA (5,8,11,14-eicosapentenoyl-CoA) (Sigma) with 0.2 U yeast acyl-CoA oxidase (Sigma) as described earlier (Filppula et al. 1998). The formation of the ∆2-∆4

conjugated double bond caused by ∆3,5-∆2,4-dienoyl-CoA isomerase activity was detected spectrophotometrically at 300 nm.

To determine kinetic parameters, 5, 10, 20, 40, 60, 100 and 200 µM of trans-3-hexenoyl-CoA were used, and the Km and Vmax values were calculated using GraFit software (Sigma).

4.10 Site-directed mutagenesis (III)

Multiple amino acid sequence alignments of Ecip1 with other hydratase/isomerase superfamily members were performed by CLUSTALX (Thompson et al. 1997) to predict the catalytic residues of Eci1p (I, III, IV). The active site amino acid of the rat mitochondrial enoyl-CoA isomerase, Glu165 (Müller-Newen & Stoffel 1993), is not

46

conserved in Eci1p, being replaced by a phenylalanine (Phe150). A protic side chain two residues before that, Tyr148, was suggested to be a possible candidate for a catalytic residue. Another option was Glu158, which corresponds to the active site residues of ∆3,5-∆2,4-dienoyl-CoA isomerase (Modis et al. 1998) and 4-CBA-CoA dehalogenase (Benning et al. 1996, Yang et al. 1996), Asp204 and Asp145, respectively. Both Tyr148 and Glu158 were mutated to alanine by using the QuickChangeTM mutagenesis kit (Stratagene) following the manufacturer’s instructions. For the mutagenesis, oligonucleotides which contained the desired mutation were used in the PCR reaction: 5’-CAAGGTTTATTTGCTAGCGCCCTTTGCTAACTTAG-3’ and 5’-CTAAGTTAGCAAAGGGCGCTAGCAAATAAACCTTG-3’ for the Tyr148Ala mutation and 5’-CTTAGGACTAATTACCGCAGGTGGTACAACGGTC-3’ and 5’-GACCGTTGTACCACCTGCGGTAATTAGTCCTAAG-3’ for the Glu158Ala mutation. The ECI1 gene ligated into the pET3a-expression vector was used as the template and pfu as the DNA polymerase. The mutated proteins were expressed and purified as described, and their ∆3-∆2-enoyl-CoA isomerase activity was measured.

4.11 Determination of native molecular mass (I, III, IV)

The elution volume of Eci1p from the Superdex 200 HR10/30 gel filtration column was compared to that of dienoyl-CoA isomerase, a hexameric protein of 170 000 Da. Gel filtration was also used to find out whether the purified E158A mutant Eci1p has the same native molecular mass as the wild-type enzyme. The native size of Eci1p was also analyzed using the DynaPro dynamic light scattering device (Protein Solutions).

The native size of Eci1p at acidic and neutral pH was analyzed by sedimentation velocity experiments using an analytical ultracentrifuge (Beckman XL-A) by Dr. Arie Geerlof at the EMBL, Heidelberg, Germany.

4.12 Crystallization (II, III, IV)

Prior to crystallization, the purified Eci1p was concentrated to either 1, 2.3 or 2.5 mg/ml by centrifugal filtration in the following buffer: 20 mM potassium phosphate, pH 7.2, 450 mM KCl, 1 mM EDTA, 1 mM EGTA, 0.5 mM benzamidine-HCl. N-octyl-β-D-glucoside (10mM) was used as an additive in the 2.3 mg/ml protein solution. The crystallization conditions for the 1 and 2.3 mg/ml protein solutions were determined by the sparse-matrix screen (Jancarik & Kim 1991) using Crystal Screen and Crystal Screen II kits (Hampton Research). The crystallization was performed by the sitting and hanging-drop vapour diffusion methods by mixing equal volumes (2µl) of the protein and precipitant solutions. The screens were performed at +22°C and +4°C. The pH and the precipitant concentration of the promising conditions were further optimized.

To obtain crystals complexed with an active site ligand, octanoyl-CoA was added to a final concentration of 2 mM to protein solutions containing 1 and 2.5 mg/ml of Eci1p prior to crystallization. For the 1 mg/ml protein solution, Crystal Screens (Hampton Research) were used as described above. To obtain crystals from the 2.5 mg/ml Eci1p

47

solution, the automated screening method described by Zeelen and co-workers (1994) using a crystallization robot was utilized. The crystallization was carried out in hanging drops as described above, and the crystallization conditions producing crystals were optimized.

4.13 X-ray diffraction data collection and processing (II-IV)

Crystal quality was tested and the heavy atom test data sets were collected using a rotating-anode X-ray generator (Nonius) and a MAR345 image-plate detector (MAR research) at the Department of Biochemistry, University of Oulu. The data sets required for structure determination were collected at synchrotrons either in Hamburg, Germany, or in Lund, Sweden, using a MAR345 or MARCCD detector. All data sets were processed and scaled using programs DENZO and SCALEPACK (Otwinowski & Minor 1997).

4.13.1 Data from unliganded crystals (II)

Unliganded crystals were mounted in quartz capillaries and exposed to X-rays for the determination of crystal quality, the unit cell parameters and the space group. Due to the radiation sensitivity of the crystals, the conditions for cryo-cooling had to be determined. Crystals belonging to the space group P6322 with a monomer in the asymmetric unit were soaked for about 30 s in mother liquor containing an excess of ammonium sulphate (2.25 M), to prevent the crystals from solubilizing and, in addition, 20 % (v/v) ethylene glycol as cryo protectant. The crystals were then immediately flash-frozen in a stream of liquid nitrogen. The crystals were kept at 100 K during the collection of the X-ray diffraction data. The native, unliganded data set at 2.5Å resolution was collected using synchrotron radiation at beamline X11 at DESY, Hamburg.

4.13.2 Multiwavelength anomalous dispersion (MAD) data (III)

4.13.2.1 Preparation of heavy-atom derivative crystals

Crystals complexed with heavy atoms were required for the determination of the crystal structure of Eci1p because attempts to use molecular replacement yielded no solutions. Heavy-atom compounds were dissolved in crystallization mother liquor in a final concentration of 1-5 mM. Crystals belonging to the space group P6322 were soaked in the heavy-atom solutions for 2-20 hours, after which X-ray diffraction data were collected in cryogenic conditions and processed as described above. The search resulted in one suitable heavy-atom derivative forming peaks in the Patterson map, potassium perrhenate

48

(KReO4), which could be used in the MAD experiment. Prior to the MAD data collection, the crystal was soaked in 1 mM KReO4 for four hours and subsequently frozen at 100K using 20% ethylene glycol as cryo protectant.

4.13.2.2 MAD data collection

The MAD data were collected using four wavelengths at beamline BW7A at DESY, Hamburg. The wavelengths chosen were (1) the peak of anomalous scattering of rhenium (Re), (2) the inflection point of anomalous scattering, (3) the high-energy remote and (4) the low-energy remote. The datasets were processed and scaled separately with the DENZO and SCALEPACK programs from the HKL package (Otwinowski & Minor 1977). Further data processing was done with programs of the CCP4 package (Collaborative Computational Project Number 4 1994).

4.13.3 Data from octanoyl-CoA complexed crystals (IV)

Small crystals co-crystallized with octanoyl-CoA belonged to a tetragonal space group, such as P422, and diffracted at best to 3 Å resolution. There was a trimer in the asymmetric unit. X-ray diffraction data at 3.3 Å were collected at beamline 711 at Max-Lab, Lund. Before data collection, the crystal was frozen in a stream of liquid nitrogen using 20% glycerol in mother liquor as cryo protectant. The cryo solution also contained 1 mM octanoyl-CoA.

4.14 Structure determination by the MAD method and refinement of the structure (III)

The MAD datasets collected at the peak and remote wavelengths were scaled to the dataset collected at the inflection point wavelength with SCALEIT (Collaborative Computational Project Number 4 1994). The initial heavy-atom positions were located with RSPS (Collaborative Computational Project Number 4 1994) and SOLVE (Terwilliger & Berendzen 1999). The two heavy-atom positions found were refined and the phases to 2.5 Å resolution were calculated with the program MLPHARE (Collaborative Computational Project Number 4 1994). The initial maps were improved by solvent-flattening calculations and the phases were extended to 2.15 Å resolution with the DM program (Cowtan & Main 1996), after which the maps were easily interpretable with the O program (Jones et al. 1991). The model of Eci1p monomer was built manually in O. Subsequently, the model was subjected to restrained refinement, using the maximum-likelihood target, with REFMAC (Murshudov et al. 1997). Water molecules were added automatically by the solvent-building mode of the program ARP/wARP (Perrakis et al. 1999). Ethylene glycol and perrhenate molecules were also added into the

49

refinement. The pdb entry code for the perrhenate-complexed yeast enoyl-CoA isomerase is 1HNU.

The unliganded structure (not soaked in perrhenate) of the yeast enoyl-CoA isomerase was solved by rigid body refinement with REFMAC using the perrhenate-complexed structure as the starting model. Further restrained refinement to 2.5 Å resolution was also carried out with REFMAC. The pdb entry code is 1HNO.

4.15 Structure determination by the molecular replacement method and refinement of the structure (IV)

The 3.3 Å structure of yeast enoyl-CoA isomerase complexed with octanoyl-CoA was solved by the molecular replacement method with the AmoRe program (Navaza 1994) using data from 8 to 4 Å. The 2.15 Å structure of the yeast ∆3-∆2-enoyl-CoA isomerase monomer solved by the MAD method was used as the search model. Since the space group of the crystal was not certain, the solution was searched for by using the space groups P422, P4212, P4122, P4322, P4222, P42212, P41212 and P43212. The molecular replacement calculations only yielded solutions in the space group P41212. The coordinates of the octanoyl-CoA-complexed Eci1p trimer were first refined by rigid body refinement of each monomer with REFMAC (Murshudov et al. 1997) and then subjected to further restrained refinement with REFMAC by using loose non-crystallographic symmetry (NCS) restraints. The ligand, octanoyl-CoA, was also built into the active sites of the subunits and then refined using NCS restraints. The coordinates of octanoyl-CoA were taken from the 2-enoyl-CoA hydratase-1 structure (2DUB, Engel et al. 1998). In addition, two water molecules and three phosphate ions were added to the model. The coordinates were submitted to the pdb with the entry code 1K39.

4.16 Structure analysis and validation (III-IV)

The structures of Eci1p were analyzed and their quality was evaluated using the programs O (Jones et al. 1991), WHATIF (Vriend 1990), PROCHECK (Laskowski et al. 1993), DSSP (Kabsch & Sander 1983) and ICM (Abagyan & Totrov 1994).

4.17 Comparison of the structures belonging to the hydratase/isomerase superfamily (IV)

The structure of the yeast ∆3-∆2-enoyl-CoA isomerase was compared by superposition studies with the other known structures within the hydratase/isomerase superfamily. For these comparisons, the following PDB entries were used: 2-enoyl-CoA hydratase-1 (liganded with acetoacetyl-CoA, 1DUB; liganded with octanoyl-CoA, 2DUB), 4-CBA-CoA dehalogenase (1NZY), methylmalonyl-CoA decarboxylase (unliganded, 1EF8;

50

liganded, 1EF9) and ∆3,5-∆2,4-dienoyl-CoA isomerase (1DCI). All superpositions were done with the lsq option in O (Jones et al. 1991), using the Cα-atoms of 25 residues from β-strands as well as their equivalents in the other structures.

5 Results

5.1 Characterization of ECI1 and its gene product, Eci1p (I)

Northern analysis revealed that ECI1 transcripts accumulate when the cells are grown on oleate, whereas no detectable gene expression could be seen when glucose was used as the carbon source. Furthermore, according to EMSA studies, the regulation is mediated by the binding of the transcription factors Pip2p and Oaf1p (Luo et al. 1996, Rottensteiner et al. 1996) to the ORE in the promoter region of the gene.

To study the potential role of Eci1p in fatty acid degradation, an ECI1-deleted yeast strain was grown on saturated and unsaturated fatty acids as the sole carbon source. The ECI1 knock-out was able to utilize the saturated fatty acid, palmitic acid, suggesting that Eci1p does not participate in the β-oxidation of straight-chain fatty acids. Instead, the ECI1-disrupted cells did not grow on any of the three unsaturated fatty acids tested. This indicated that Eci1p is involved in the metabolism of double bonds in both odd- and even-numbered positions of fatty acids, possibly acting as a ∆3-∆2-enoyl-CoA isomerase. The function of Eci1p was tested further by supplying the mutant yeast cells with the rat peroxisomal ∆3-∆2-enoyl-CoA isomerase, which is part of MFE-1, also containing 2-enoyl-CoA hydratase-1 and L-specific 3-hydroxyacyl-CoA dehydrogenase activities (Palosaari & Hiltunen 1990). The ∆3-∆2-enoyl-CoA isomerase activity of MFE-1 could restore the growth of ECI1-deleted cells on oleic acid, suggesting that ECI1 encodes for an ∆3-∆2-enoyl-CoA isomerase.

For further characterization, the full-length Eci1p, consisting of 280 amino acid residues, was expressed as a recombinant protein in E. coli and purified. Analysis on SDS-PAGE showed one single protein band corresponding to a molecular mass of 32 kDa, which is in agreement with the molecular mass calculated on the basis of the amino acid sequence of Eci1p, 31.7 kDa. Eci1p eluted from the gel filtration column in the same elution volume as the rat ∆3,5-∆2,4-dienoyl-CoA isomerase, which has a native mass of 170 kDa (Filppula et al. 1998). In addition, according to dynamic light scattering analysis, the approximate molecular mass is 151 kDa. From these results, it could be concluded that, in solution, Eci1p is a hexameric protein formed of six identical 32-kDa subunits. The purified Eci1p was assayed for ∆3-∆2-enoyl-CoA isomerase, ∆3,5-∆2,4-

52

dienoyl-CoA isomerase and 2-enoyl-CoA hydratase-1 activities. No dienoyl-CoA isomerase or hydratase-1 activity could be detected, whereas the purified protein was found to possess a specific ∆3-∆2-enoyl-CoA isomerase activity of 11.2 µmol/min/mg (kcat 6.0 sec-1) and a Km value of 21.5 µM for trans-3-hexenoyl-CoA. Eci1p is thus a monofunctional enoyl-CoA isomerase and will be referred to as the yeast ∆3-∆2-enoyl-CoA isomerase.

The amino acid sequence of the yeast enoyl-CoA isomerase was aligned with the sequences of some other hydratase/isomerase superfamily members, hydratase-1 (Minami-Ishii et al. 1989), mitochondrial enoyl-CoA isomerase (Müller-Newen & Stoffel 1991, Palosaari et al. 1991), 4-CBA-CoA dehalogenase (Babbitt et al. 1992) and ∆3,5-∆2,4-dienoyl-CoA isomerase (FitzPatrick et al. 1995, Filppula et al. 1998), in order to make suggestions for the possible catalytic residues of the yeast enoyl-CoA isomerase. Surprisingly, the active site amino acid of hydratase-1, Glu164 (Müller-Newen et al. 1995), which is also present in the mitochondrial enoyl-CoA isomerase, was not conserved in the yeast enoyl-CoA isomerase, but was replaced by a phenylalanine. Instead, we suggested that a protic residue close to that, Tyr148, could be involved in the proton exchange. Another possibility was determined to be Glu158, which corresponds to the catalytic residues of the dienoyl-CoA isomerase and the 4-CBA-CoA dehalogenase, Asp204 (Modis et al. 1998) and Asp145 (Benning et al. 1996, Yang et al. 1996), respectively. Both Tyr148 and Glu158 were separately exchanged to alanine, and the enoyl-CoA isomerase activity of the mutated enzymes was measured. The enoyl-CoA isomerase activity of Tyr148Ala was comparable to the activity of the wild type (unpublished results), whereas the Glu158Ala variant lacked isomerase activity completely, suggesting that Glu158 is involved in the reaction mechanism of the yeast enoyl-CoA isomerase.

For the determination of its subcellular localization, Eci1p was tagged with GFP, and the location in both wild-type cells and pex6∆ cells lacking peroxisomes was examined by fluorescence microscopy. In the wild-type cells, GFP-Eci1p was detected as punctate fluorescence indicating compartmentalization. In the pex6∆ cells, instead, the fluorescence was diffuse due to the location of GFP-Eci1p in the cytosol. DNA staining, which showed intact mitochondria in both yeast cell types, excluded the localization of the fusion protein in the mitochondria. This led to the conclusion that Eci1p is located in yeast peroxisomes.

5.2 Crystallization of yeast ∆∆∆∆3-∆∆∆∆2-enoyl-CoA isomerase (II, IV)

The initial crystallization conditions for yeast enoyl-CoA isomerase were searched for by using Hampton Screen and Hampton Screen II (Hampton Research). The most promising conditions were optimized with respect to pH and the precipitant concentration. Two unliganded crystal forms were obtained at room temperature. Crystals of tetragonal shape grew in 0.1 M sodium acetate pH 4.8, 2.2 M (NH4)2SO4 when 1 mg/ml protein solution was used (Fig. 10A). The crystals had maximum dimensions of 0.3 x 0.2 x 0.2 mm3, and they diffracted to about 3 Å resolution when the rotating anode X-ray source was used.

53

Fig. 10. Yeast ∆∆∆∆3-∆∆∆∆2-enoyl-CoA isomerase crystal forms. (A) A tetragonal unliganded crystal. The longest dimension of the crystal is 0.3 mm. (B) A hexagonal crystal belonging to the space group P6322. This crystal form was grown at pH 5.5 and was used in the determination of the perrhenate-complexed and the unliganded yeast isomerase structures. The longest dimension of the crystal is 0.3 mm. (C) The octanoyl-CoA-complexed crystal form crystallized at pH 7.0. The longest dimension of the crystals is 0.15 mm. The crystal belonged to a primitive tetragonal space group and the unit cell dimensions were a=b=118.1 Å, c=222.3 Å, α=β=γ= 90°. This crystal form was not analyzed further. The other crystal form was obtained in 0.1 MES pH 5.5, 5 % 1,4-dioxane, 1.4 M (NH4)2SO4 using a protein concentration of 2.3 mg/ml and 10 mM N-octyl-β-D-glucoside as an additive (Fig 10B). The crystals had hexagonal shape and they belonged to the space group P6322 with unit cell dimensions of a=b=116.1 Å, c=123.3 Å, α=β= 90°, γ=120° and one monomer per asymmetric unit. The diffraction limit of the hexagonal crystals was better than 3 Å when using the home X-ray source and 2.15 Å when synchrotron radiation was used at DESY, Hamburg. These crystals were easy to obtain and could be frozen in the presence of 20 % ethylene glycol, in addition to which they had higher symmetry than the tetragonal crystals. Thus, the hexagonal unliganded crystal form was chosen to be used in the structure determination of the yeast enoyl-CoA isomerase.

The first attempts to obtain crystals liganded with various fatty acyl-CoA esters were made by soaking unliganded hexagonal crystals in solutions containing fatty acyl-CoAs. This did not lead to ligand binding, however, possibly because of the low pH of the hexagonal crystals. The next step was to use co-crystallization. Prior to the crystallization experiments, the enoyl-CoA isomerase was concentrated to 2.5 mg/ml and octanoyl-CoA was added to the solution in a concentration of 2 mM. Octanoyl-CoA is an inhibitor of the enoyl-CoA isomerase that can bind to the active site but cannot be used as a substrate. In order to facilitate ligand binding, crystallization conditions with neutral or slightly basic pH were screened and further optimized. The best defined crystals were obtained in 0.1 M TEA, pH 6.5-7.0, 1.2-1.4 M sodium citrate (Fig. 10C). The crystals were quite small, about 0.15 mm in their longest dimension, and diffracted only to 3 Å resolution using synchrotron radiation at Max-Lab in Lund. Prior to data collection, the crystals were frozen using 20 % glycerol as cryo protectant. The crystals belonged to the primitive tetragonal space group such as P422 and had a trimer in the asymmetric unit. The unit cell dimensions were a=b=116.7 Å, c=216.8 Å, α=β=γ= 90°. Another octanoyl-CoA-liganded, hexagonal crystal form was obtained in 0.1 M sodium acetate pH 5.1, 1.7 M (NH4)2SO4, but because of its low crystallization pH, it was not used in the structure determination.

54

5.3 Structure of the unliganded yeast ∆∆∆∆3-∆∆∆∆2-enoyl-CoA isomerase (III)

The structure of the yeast ∆3-∆2-enoyl-CoA isomerase was solved by the MAD method using potassium perrhenate (KReO4) in determining the phases. The heavy atoms were incorporated into the hexagonal unliganded isomerase crystals by soaking the crystals in crystallization solution containing KReO4 prior to data collection at 100 K. The MAD data were 99.5 % complete with 5-fold redundancy. The resolution limit of the data was 2.15 Å. Perrhenate was found to be bound at two sites in the isomerase subunit, one of the sites being the active site (Fig. 11). The yeast enoyl-CoA isomerase structure complexed with perrhenate was refined to the R and free R-factors of 21.5 % and 25.7 %, respectively. Another unliganded data set without heavy atoms was collected at 2.5 Å to determine whether the bound perrhenate ion changes the conformation of the active site residues. The unliganded structure was solved by rigid body refinement (Murshudov et al. 1997) using the perrhenate-complexed structure as the model and refined to the R-factor of 20.1 % and the free R-factor of 25.1 %. The structures were found to be virtually identical, and they will therefore not be described individually.

The structure of the yeast enoyl-CoA isomerase consists of two domains, the N-terminal core domain and the C-terminal trimerization domain (see “Discussion”, Fig. 13A). The N-terminal domain contains the active site, and it is folded in the spiral-fold topology, each turn of the spiral being formed of two β-strands and an α-helix. The four turns of the core domain are followed by the trimerization domain, which consists of four α-helices, namely H7, H8 H9 and H10. The C-terminal domain folds over the core domain and covers the active site. A salt bridge between Lys233 in helix H8 and Asp135, which comes just after helix H3, anchors the domains together. This salt bridge is conserved in the known structures of hydratase/isomerase superfamily. There are three undefined regions in the structure of the yeast enoyl-CoA isomerase: the residues 1 to 7 in the N-terminus, the residues 271 to 280 in the C-terminus and the residues 74 to 88 near helix H2. These regions could not be built because of the lack of any features in the electron density map. Due to the disorder in the C-terminus, the PTS1 targeting sequence could not be detected, either. Other parts of the structure are well defined.

Three enoyl-CoA isomerase monomers, related by a crystallographic three-fold axis, form a disk-like trimer. The contacts between the subunits are extensive. The main contact region concerns the helix H8 of one subunit fitting into a complementary docking site shaped by residues from the helices H4, H5 and H9 of the adjacent subunit. In the crystal, two trimers are packed in such a way that they form hexamers with 32 crystallographic symmetry. The interaction between the trimers is, however, loose and there are only 30 intertrimer contacts.

In the unliganded structure, the mode of substrate binding of the yeast enoyl-CoA isomerase could be inferred by superimposing the ligand of the 4-CBA-CoA dehalogenase structure (Benning et al. 1996) onto the active site of the enoyl-CoA isomerase. The active site architecture is presented in Figure 11. Site-directed mutagenesis studies suggested that Glu158 would be the catalytic amino acid residue. From the ligand superposition, it could be seen that Glu158 is, indeed, situated so that it could exchange protons with the ligand molecule. In addition, there are no other protic

55

Re

O2

O3 O1

O4

W390

W231

W223

W126

W94

Glu158

Asn101

ND2

OD1

W409

W365

W302

W23W19

W15

W423

W35

N

Leu126

N

Ser127

N Ala70

OH

OD1

Glu251

N

Thr157

Thr105

Asn248

ND2

Ile156

Tyr38

O O

O

ON

OG1

2.8 2.9

2.63.0

3.7

3.1

2.74.0

2.7

3.7

2.8

3.2

3.5

2.8

2.6

3.4

2.5

3.1

2.4

2.52.6

2.6

4.3

3.63.1

2.9

2.9

2.8

3.5

2.4

2.7

OE1OE2

OE1

OE2

Fig. 11. The hydrogen-bonding network at the active site of the yeast enoyl-CoA isomerase. The distances between the hydrogen-bonded atoms are shown. A perrhenate ion is bound at the active site contacting the catalytic residue Glu158. The bindings of W390 and the ReO4- ion are mutually exclusive, as they compete for the same binding site. The water molecules W365, W302, W23, W35, W19, W423, W15 are part of the buried water cluster. W126 is in contact with the bulk solvent. W390 is in the oxyanion hole formed by N(Ala70) and N(Leu126). In the octanoyl-CoA-complexed enoyl-CoA isomerase structure, W390 is replaced by the thioester oxygen atom of the ligand. Asn248 and Glu251 extend from the helix H9 of the C-terminal domain, and the residues Ile156, Thr157 and the catalytic Glu158 are located just before the helix H4 (see original article III, Fig. 2). Tyr38 leads to the empty apolar cavity and is only weakly hydrogen-bonded to Leu126. In the octanoyl-CoA complex, the position of Tyr38 has not changed and the empty apolar cavity behind it still exists.

56

side chains in the immediate vicinity. The Glu158 side chain extends from a loop between the β-strand B4 and the helix H4. This region is part of the tight subunit-subunit interface. The side chain of Glu158 is hydrogen-bonded to only one protein atom, ND2 of Asn101. The thioester oxygen atom of the superimposed acyl-CoA points towards the NH groups of Ala70 and Leu126, which are thus suggested to form the conserved oxyanion hole (Fig. 11) (Holden et al. 2001). The catalytic site is accessible for acyl-CoA via an entrance lined by residues from the β-strands B2 and B4 as well as the C-terminal α-helix H10. Molecular-surface calculations with ICM (Abagyan & Totrov 1994) detected two large cavities near the active site. One of them is near Glu158 and is filled with water molecules (Fig. 11). The other cavity is mainly lined by apolar side chains, and no solvent or other molecules could be detected in this hydrophobic pocket. The apolar cavity is separated from the active site pocket by the side chain of Tyr38.

5.4 Structure of ∆∆∆∆3-∆∆∆∆2-enoyl-CoA isomerase complexed with octanoyl-CoA (IV)

The liganded structure of the yeast enoyl-CoA isomerase was determined from crystals co-crystallized with octanoyl-CoA, an inhibitor of the enoyl-CoA isomerase. This structure is also referred to as the neutral pH form on the basis of its crystallization condition. Molecular replacement calculations detected the orientations and positions of two subunits of the isomerase trimer in the space group P41212. The solution for the third subunit could not be found using AmoRe (Navaza 1994). Instead, the coordinates were obtained by superimposing the unliganded enoyl-CoA isomerase trimer, also referred to as the low pH form, onto the two assembled AMoRe solutions with the lsq option in O (Jones et al. 1991) and by extracting the coordinates of the superimposed third subunit. The rigid body refinement (Murshudov et al. 1997) of the new trimer resulted in a model with R- and Rfree-factors of 40.1 % and 41.0 %, respectively. The octanoyl-CoA molecules were built into the active sites of the subunits, and the structure was refined at 3.3 Å resolution to the final R-factor of 21.8 % and free R-factor of 29.2 % with good geometry statistics. During refinement, the residues 74-88, which were disordered in the unliganded structure (the low pH form), became visible in the electron density map, although they still continued to refine with high B-factors. This region, located just before helix H2, adopted a helical conformation and was named helix H2A. Part of helix H2 also had a different, slightly upward bent conformation compared to the low pH form of isomerase. In addition to the structural changes in helix H2, H9 and H10 also adopted somewhat different conformations when the two structures were compared (see “Discussion”, Fig. 13B). The N- and C-termini were also disordered in the liganded structure. A few more residues could, however, be built in each terminus. In the current structure, the subunits A, B and C consist of the residues 4-272, 4-274 and 5-270, respectively, leaving the peroxisomal targeting sequence still disordered.

The monomers of the liganded, neutral pH form, the yeast enoyl-CoA isomerase, assemble into a tight trimer similarly to the unliganded, low pH form. In addition, both of the pH forms crystallize as hexamers with 32 symmetry. However, the mode of assembly is rather different. In contrast to the loose intertrimer contacts in the low pH form, the

57

trimers in the neutral pH form interact closely with each other, having 159 contacts across the intertrimer space. The subunits are also 5.6 Å closer to each other. Moreover, the second trimeric disk appears to have rotated 25° around its threefold axis when the two pH forms are compared. The different hexameric packing and the loose intertrimer contacts in the low pH form suggested that the yeast isomerase could be a trimer at low pH. The ultracentrifugation sedimentation velocity runs indeed suggested that, at pH 5.6, the yeast enoyl-CoA isomerase could occur as a trimer, whereas at pH 7.2, a hexamer was found in good agreement with the gel filtration and dynamic light scattering as well as crystallographic studies.

The mode of octanoyl-CoA binding to enoyl-CoA was very much like expected on the basis of the superposition studies. The CoA moiety is bound in a bent conformation against the β-sheet formed by regions B1, B2, B3 and B4. The entrance to the catalytic site is covered by the helices H2A and H10. The octanoyl moiety is bound near the catalytic residue Glu158 and points upwards towards Gly35, Tyr38, Phe97 and Arg100 (Figs 12, 13B). The thioester oxygen is hydrogen-bonded to the main chain NH groups of Ala70 and Leu126, thus forming an oxyanion hole (Fig. 12), as speculated on the basis of the superposition studies. The only structural difference near the catalytic site in a comparison of the unliganded and octanoyl-CoA-complexed crystal forms is a small shift of approximately 1 Å in the Glu158 side chain towards the ligand.

Fig. 12. Octanoyl-CoA bound at the active site of the yeast ∆∆∆∆3-∆∆∆∆2-enoyl-CoA isomerase. The catalytic residue Glu158 as well as the residues forming the oxyanion hole, Ala70 and Leu126, are shown. The main chain NH groups of Ala70 and Leu126 are hydrogen-bonded to the thioester oxygen of the octanoyl-CoA molecule. The electron density map drawn around the ligand was calculated after an omit refinement. For this, the atoms of octanoyl-CoA were omitted from the model and 10 cycles of refinement with REFMAC (Murshudov et al. 1997) were performed, after which a difference electron density map was calculated. The map is contoured at 2σσσσ.

6 Discussion

6.1 ECI1 encodes for a monofunctional peroxisomal ∆∆∆∆3-∆∆∆∆2-enoyl-CoA isomerase (I)

At the beginning of this project, it was suggested that the ECI1 gene product could be involved in the β-oxidation pathway for three reasons: [1] its amino acid sequence is similar to the hydratase/isomerase superfamily members often involved in acyl-CoA ester metabolism, [2] the promoter region of ECI1 contains a sequence resembling OREs (Einerhand et al. 1993, Filipits et al. 1993), which are involved in upregulating the gene expression when S. cerevisiae cells are grown on fatty acids, such as oleic acid, and [3] ECI1 ends with nucleotides encoding a modified PTS1, which could target the gene product into peroxisomes (Gould et al. 1989), where the β-oxidation in yeast exclusively occurs (Kunau et al. 1988). S. cerevisiae genes encoding peroxisomal β-oxidation enzymes, such as POX1 (Dmochowska et al. 1990), FOX2 (Hiltunen et al. 1992), FOX3 (Einerhand et al. 1991) and SPS19 (Gurvitz et al. 1997a, 1997b), are also upregulated when cells are grown on oleic acid or other fatty acids, and this upregulation is mediated by OREs (Einerhand et al. 1993, Filipits et al. 1993). The transcription factors Oaf1p and Pip2p (Luo et al. 1996, Rottensteiner et al. 1997) bind to the OREs and induce the transcription. Transcriptional analysis of ECI1 showed that the amount of ECI1 mRNA was strongly increased in the presence of oleic acid, whereas no transcription could be detected when cells were grown on glucose or ethanol. Likewise, no ECI1 transcripts could be observed in PIP2- and OAF1-deficient cells grown on oleic acid, indicating that the oleic acid induction is mediated by ORE-bound Pip2p and Oaf1p. This was further verified by EMSA experiments.

The ability of the ECI1-deleted strain to utilize various fatty acids was investigated by culturing mutant cells on both saturated and unsaturated fatty acids. Eci1p was found to be dispensable for the metabolism of saturated fatty acids, but it was required for the degradation of double bonds at both odd- (oleic and arachidonic acid) and even-numbered (octadecenoic and arachidonic acid) positions of unsaturated fatty acids. For the metabolism of even-numbered double bonds, an additional auxiliary enzyme, dienoyl-

59

CoA reductase, sps19p (Gurvitz et al. 1997a) is needed, whereas in the degradation of odd-numbered double bonds, Eci1p can act either directly on cis-3-double bonds or via an alternative route also requiring the dienoyl-CoA isomerase, Dci1p, and the dienoyl-CoA reductase (Fig. 3). In yeast, most of the odd-numbered double bonds are probably metabolized via the isomerase-dependent route, leaving the pathway including Dci1p dispensable (Gurvitz et al. 1999, for a review, see Trotter 2001). The fact that the ECI1-deleted strains could not utilize any of the unsaturated fatty acids tested implies that Eci1p is likely to be the only ∆3-∆2-enoyl-CoA isomerase in S. cerevisiae.

For characterization, Eci1p was expressed in E. coli and purified to homogeneity. On SDS-PAGE, the subunit size was determined to be 32 kDa, and the native molecular mass was found to correspond to that of a hexamer by gel filtration and dynamic light scattering. Hexameric assembly is a common feature in the hydratase/isomerase superfamily. 2-Enoyl-CoA hydratase-1 (Furuta et al. 1980), dienoyl-CoA isomerase (Filppula et al. 1998) and methylmalonyl-CoA decarboxylase (Benning et al. 2000) are also hexamers in solution. In addition, their subunit sizes are comparable to that of Eci1p.

The active sites of hydratase/isomerase proteins commonly display residual catalytic activities that are related to the actual enzyme activity. For example, the rat ∆3,5-∆2,4-dienoyl-CoA isomerase contains traces of hydratase-1 activity (Filppula et al. 1998) and the rat hydratase-1 contains some isomerase activity (Kiema et al. 1999). This is likely to be due to the common evolutionary origin of these enzymes. Eci1p was, however, found only to have ∆3-∆2-enoyl-CoA isomerase activity but no detectable hydratase-1 or dienoyl-CoA isomerase activities. Eci1p is thus a monofunctional ∆3-∆2-enoyl-CoA isomerase with a specific activity of 11.2 µmol/min/mg (kcat 6.0 s-1) and a Km value of 21.5 µM when trans-3-hexenoyl-CoA is used as substrate. The results obtained in this study were also confirmed by another independent study on ECI1 published almost simultaneously (Geisbrecht et al. 1998). Geisbrecht and co-workers (1998) also found that Eci1p is required for the degration of unsaturated fatty acids and that it possesses only ∆3-∆2-enoyl-CoA isomerase activity. The enzyme activity measurements were done using cis-3-octenoyl-CoA as substrate. The resulting specific isomerase activity was 16 µmol/min/mg (kcat 8.5 sec-1), which is slightly higher than that obtained with the substrate shorter by two carbon atoms in our experiments. In neither of the studies, however, was substrate specificity with different chain lengths determined. Nevertheless, it can be speculated that because Eci1p is most probably the only ∆3-∆2-enoyl-CoA isomerase in S. cerevisiae, it has to be able to utilize unsaturated fatty acids of all chain lengths. The results of these two studies also showed that Eci1p can use both cis- and trans-3-enoyl-CoAs as substrates, as was expected. The cis-isoform is more abundant in the unsaturated fatty acids occurring in nature, but the trans-isoform is the product of the 2,4-dienoyl-CoA reductase reaction (Dommes & Kunau 1984), which means that both cis- and trans-3-enoyl-CoAs must be acceptable substrates for enoyl-CoA isomerases. The kinetic parameters of the yeast enoyl-CoA isomerase are compared to those of other monofunctional isomerases (Euler-Bertram & Stoffel 1990, Palosaari et al. 1990, Müller-Newen & Stoffel 1993, Geisbrecht et al. 1999b) and peroxisomal MFE-1 (Palosaari & Hiltunen 1990) in table 2. The kcat values of all the monofunctional enoyl-CoA isomerases, with the exception of the bovine mitochondrial one, are more or less comparable with each other, the activity of the yeast enoyl-CoA isomerase, however, being slightly lower. Nevertheless, it should be noted that variable substrates were used in

60

Table 2. The kinetic parameters of ∆3-∆2-enoyl-CoA isomerases.

Source of the ∆3-∆2-enoyl-CoA isomerase

Substrate kcat (sec-1)

Km

(µM) Reference

Rat mitochondrial trans-3-C6-CoA 19 n.r. a Palosaari et al. 1990 cis-3-C12-CoA 31 37 Müller-Newen & Stoffel 1993 Bovine mitochondrial cis-3-C12-CoA 128 32 Euler-Bertram & Stoffel 1990 Yeast peroxisomal trans-3-C6-CoA 6.0 21.5 this study cis-3-C8-CoA 8.5 n.r. Geisbrecht et al. 1998 Human peroxisomal cis-3-C8-CoA 18 n.r. Geisbrecht et al. 1999b Rat MFE-1 trans-3-C6-CoA 0.41 23.6 Kiema, Taskinen, Pirilä,

Koivuranta, Wierenga & Hiltunen, unpublished

a n.r., not reported

the experiments. Interestingly, the isomerase activity of peroxisomal MFE-1 is considerably lower when compared to the monofunctional enzymes.

Multiple-sequence alignments (see original articles I, III and IV) were used to compare the yeast enoyl-CoA isomerase with the other hydratase/isomerase superfamily members and to elucidate the possible active site amino acid residues. Site-directed mutagenesis was used to test whether the candidates acted as catalytic residues. On the basis of sequence alignments, the catalytic residue of yeast isomerase was suggested to be either Tyr148 or Glu158. It was found that Glu158 is possibly the residue involved in the catalytic mechanism of yeast isomerase, since its mutation to alanine made the enzyme totally inactive, whereas the Tyr148Ala mutation had no effect. Glu158 corresponds to the catalytic residues of the rat ∆3,5-∆2,4-dienoyl-CoA isomerase and 4-CBA-CoA dehalogenase. Neither of the mutations affected the expression level or solubility of the protein. Although yeast isomerase has only low sequence identity compared to the other hydratase/isomerase proteins, often less than 20 %, it was predicted to have the same elements of secondary structure and thus similar overall structure as 2-enoyl-CoA hydratase-1 (Engel et al. 1996). The structure would thus consist of a spiral core domain for catalysis and an α-helical trimerization domain. Like hydratase-1, yeast isomerase would form disk-like trimers that would dimerize into hexamers.

Proteins are targeted into peroxisomes by either PTS1 (Gould et al. 1989) or PTS2 sequences (Swinkels et al. 1991). ECI1 has both a modified PTS1, HisArgLeuCOOH, in its carboxy terminus and a possible PTS2 in the amino-terminal part (Yang et al. 2001). In our study, only the presence of PTS1 was taken into consideration, and its ability to direct yeast isomerase into peroxisomes was studied by tagging the isomerase with GFP at its N-terminus. The fusion protein was indeed found to be located in peroxisomes. The same result was also obtained by Geisbrecht and co-workers (1998). Recent studies have indicated, however, that the peroxisomal localization process of Eci1p is more complicated than it seemed at first (Yang et al. 2001). Yang and co-workers (2001) also discovered the existence of a PTS2, but found that it is not used in peroxisomal targeting. The PTS1 receptor, pex5, however, is indispensable for the peroxisomal import of Eci1p, although the actual PTS1 of the yeast enoyl-CoA isomerase is not (Yang et al. 2001). This leads to the suggestion that Eci1p forms a complex with another peroxisomal protein

61

and uses the PTS1 of the partner for pex5 binding and peroxisomal import. This partner was found to be Dci1p (Yang et al. 2001), which has 45 % sequence identity to Eci1p and is the yeast dienoyl-CoA isomerase (Gurvitz et al. 1999). If Dci1p is deleted from the yeast strain, Eci1p is able to use, although inefficiently, its own PTS1 for peroxisomal import (Yang et al. 2001). In a normal situation, yeast isomerase possibly forms a hetero-oligomer with Dci1p, which is able to enter the peroxisomes.

6.2 The structures of the unliganded and the octanoyl-CoA-complexed yeast ∆∆∆∆3-∆∆∆∆2-enoyl-CoA isomerase and the differences

between them (III, IV)

The structure of the yeast ∆3-∆2-enoyl-CoA isomerase determined in this study is the fifth published structure within the hydratase/isomerase superfamily. The other previously known structures are those of the rat 2-enoyl-CoA hydratase-1 (Engel et al. 1996, Engel et al. 1998), the 4-chlorobenzoyl-CoA dehalogenase from Pseudomonas sp. strain CBS-3 (Benning et al. 1996), the rat ∆3,5-∆2,4-dienoyl-CoA isomerase (Modis et al. 1998) and the methylmalonyl-CoA decarboxylase from E. coli (Benning et al. 2000). Very recently, the structure of the human AUH protein, the RNA-binding homologue of hydratase-1, has also been published (Kurimoto et al. 2001). Liganded structures are known for all of them, except the dienoyl-CoA isomerase and the AUH protein. The overall structure of the yeast ∆3-∆2-enoyl-CoA isomerase was found to resemble closely the other structures within the hydratase/isomerase superfamily, having a similar spiral core domain for catalysis and a helical C-terminal domain for trimerization.

In both the unliganded and the liganded yeast isomerase structures, the C-terminal PTS1-like sequence, HisArgLeuCOOH, is disordered. This is considered to be a common feature of peroxisomal proteins, the only exceptions being the rat dienoyl-CoA isomerase, whose PTS1 is buried in a pocket between the two dienoyl-CoA isomerase trimers (Modis et al. 1998), and the SCP-2-like domain of the human MFE-2, whose PTS1 is solvent-exposed (Haapalainen et al. 2001). In the unliganded yeast isomerase (the low pH form), part of the region preceding helix H2, i.e. the residues 74-88, is disordered (Fig. 13A). In the liganded structure (the neutral pH form), however, this region becomes ordered and appears to adopt a helical conformation. This new region, helix H2A, is still quite flexible, since the B-factors are high and no side chain density can be seen for all residues. In addition, H2 takes a different conformation compared to the low pH form (Fig. 13B). Conformational changes are also seen in the helices H9 and H10 in a comparison of the two pH forms.

In both the low and neutral pH forms of the yeast enoyl-CoA isomerase, three monomers are packed into a tight trimer. The mode of assembly into hexamers is, however, rather different. In the neutral pH form, the two trimers interact closely with 159 atom-atom contacts within a contact distance of 3.5 Å, whereas in the low pH form, there are only 30 intertrimer contacts. The distance between the trimers is longer in the low pH form, and the trimers are also rotated 25° with respect to each other. The structurally most variable regions, i.e. the helices H2 and H9, form the contact surface in

62

the intertrimer space and the changes in their conformations thus correlate with the differences in the mode of assembly of the trimer disks into hexamers in the two pH forms. In the neutral pH form, residues from H2 and H9 make up most of the 159 intertrimer contacts and both regions are involved in forming 10 salt bridges to the other trimer (see original article IV, table 2). The importance of these salt bridges rationalizes the possible pH dependency of the mode of assembly of the trimers into hexamers. Ultracentrifugation sedimentation velocity studies also suggest that the different mode of hexamer assembly could be pH-dependent, since at pH 5.6 the yeast isomerase has a molecular mass corresponding to a trimer and at pH 7.2 it is clearly a hexamer. This agrees with the crystallographic data showing that, at low pH, the isomerase trimers are only loosely assembled into a hexamer. Additional ultracentrifugation experiments to confirm these results could, however, not be performed because, at low pH, the yeast enoyl-CoA isomerase does not stay soluble and aggregates rapidly.

Fig. 13. The overall fold of the low pH form (A) and the neutral pH form (B) of the yeast ∆∆∆∆3-∆∆∆∆2-enoyl-CoA isomerase monomer. The most variable αααα-helix H2 and the C-terminal helices H7, H8, H9 and H10 are shown. In addition, in the neutral pH form structure (B) the new αααα-helix H2A and the bound octanoyl-CoA molecule are depicted.

There are no major conformational differences at the catalytic site between the low and neutral pH forms, and the mode of binding of octanoyl-CoA into the active site of the neutral pH form is much like expected on the basis of superposition studies. Octanoyl-CoA could be built in all of the three subunits, but the occupancy of the ligand seems to be relatively low as judged from the high B-factors (unpublished results). The mean B-factor for the ligand atoms is 86 Å2, whereas for the protein atoms it is 37 Å2. The ADP and octanoyl moieties of the ligand were well defined in the electron density map, but the pantothenic acid part was mostly disordered (Fig. 12). The considerably higher B-factors for the octanoyl-CoA atoms in comparison to the protein atoms are also seen in the structure of 2-enoyl-CoA hydratase-1 complexed with octanoyl-CoA (Engel et al. 1998). Near the active site, the side chain of Tyr38 leads to a large empty apolar cavity. It is possible that this cavity could be able to bind the long fatty acid chain in catalysis. The octanoyl chain, unfortunately, is too short to reach this hydrophobic pocket, leaving its

63

role still unravelled. Binding studies with longer fatty acid substrates are needed to determine whether this cavity can be used for binding. Moreover, because of the low resolution of the liganded structure (3.3 Å), no precise and reliable analysis of the protein-ligand interactions and the active site architecture can be made.

The side chain of the catalytic residue, Glu158, is positioned so that it contacts both C2 and C4 of the octanoyl-CoA molecule (Fig. 12). The conformation of the actual substrate molecule, cis/trans-3-enoyl-CoA, is suggested to be similar, making deprotonation of C4 and protonation of C2 possible. Both cis- and trans-3-enoyl-CoA are thought to bind quite rigidly in the same conformation up to carbon C4, in order for the reaction to proceed and the right product, trans-2-enoyl-CoA, to form (Fig. 14). The trans-isoforms bind in a similar, straight fashion as saturated fatty acyl-CoAs, such as octanoyl-CoA. Instead, the cis isoform needs more space in the active site pocket because the cis double bond produces a bend in the acyl chain (Fig. 14). Crystallographic binding studies with different substrates are needed to determine how the substrate-binding pocket adopts to the space requirements of kinked acyl chains. The reaction mechanism of the yeast ∆3-∆2-enoyl-CoA isomerase can be speculated to occur as follows: Glu158 in its deprotonated form abstracts a proton from C2 of 3-enoyl-CoA. This leads to a transition state with double bonds at C1 and C3 and a negative charge on the thioester oxygen. This negative charge is stabilized by hydrogen bonds in the oxyanion hole formed by the main chain NH groups of Ala70 and Leu126 (Figs 11, 12). Subsequently, the protonated Glu158 donates the proton to C4 of the substrate and the product, trans-2-enoyl-CoA, is released (original article III, Fig. 1). The stabilization of the intermediate state of the reaction by hydrogen bonding in the oxyanion hole is a conserved feature in the hydratase/isomerase family (Holden et al. 2001), and also according to the sequence alignment, the position of the residues forming the oxyanion hole is conserved (see original article IV, Fig. 1).

R

O

SCoA

SCoA

O

R

R

O

SCoA12

34

1234

5

12

34

5

∆3-∆2-enoyl-CoAisomerase

trans-3-enoyl-CoA

cis-3-enoyl-CoA

trans-2-enoyl-CoA

Fig. 14. The substrates of ∆∆∆∆3-∆∆∆∆2-enoyl-CoA isomerase, trans- and cis-3-enoyl-CoA and the product, trans-2-enoyl-CoA. R is the remainder of the acyl chain.

64

The role of Glu158 as the only catalytic residue is consistent with the experimental data: [1] Glu158Ala mutant is totally inactive, [2] in the crystal structures Glu158 is optimally positioned for catalysis and no other protic residues are in the vicinity, and [3] the sequence alignment of the mitochondrial and peroxisomal enoyl-CoA isomerases (Müller-Newen & Stoffel 1991, Kilponen et al. 1994, Geisbrecht et al. 1999b) (see original article III, Fig. 2) shows that, in all peroxisomal isomerases, a glutamate at position 158 is conserved. In the mitochondrial isomerases, the conserved catalytic glutamate (Glu165 in the rat mitochondrial isomerase, Müller-Newen & Stoffel 1993) is at the position of Phe150 of the yeast isomerase. Still, it has to be emphasized that a better resolution structure with an active site ligand is needed for an accurate analysis of the protein-ligand interactions as well as the reaction mechanism.

6.3 Comparison of the yeast ∆∆∆∆3-∆∆∆∆2-enoyl-CoA isomerase structures with the other structures within the hydratase/isomerase superfamily

(III, IV)

6.3.1 Comparison at the monomer level

The members of the hydratase/isomerase protein family have a common evolutionary origin and they are seen to be both structurally and mechanistically similar, although they are involved in different metabolic pathways. The similar overall structure is seen in all the known structures within the hydratase/isomerase superfamily. The second spiral turn of the core domain and especially helix H2 are the most variable regions, both in the two isomerase structures determined in this study and in all the other members of the hydratase/isomerase superfamily. The marked structural variability of this region correlates with the sequence variability shown by multiple sequence alignments (see article IV, Fig. 1). For example, the yeast enoyl-CoA isomerase and the rat dienoyl-CoA isomerase (FitzPatrick et al. 1995) have a 13-residue insertion when compared to 2-enoyl-CoA hydratase-1 (Minami-Ishii et al. 1989). Since H2 is involved in shaping the substrate-binding pocket, this variability in its structure enables the binding of very different acyl-chains in the active sites of the enzymes and also facilitates the involvement of hydratase/isomerase family members in so many different pathways. In 2-enoyl-CoA hydratase-1 (Engel et al. 1996) and the dienoyl-CoA isomerase (Modis et al. 1998), helix H2 is split into two parts, as in the neutral pH form of the yeast isomerase. As in the yeast enoyl-CoA isomerase, part of this region in hydratase-1 is flexible and has high B-factors. In the octanoyl-CoA complexed structure of hydratase-1 (Engel et al. 1998), this flexible region (residues 114-118) becomes completely disordered and allows the fatty acyl chain to reach towards the intertrimer space. 2-enoyl-CoA hydratase-1 thus does not have a hydrophobic cavity in the core domain for acyl chain binding, as suggested for the yeast enoyl-CoA isomerase.

When the subunits of these enzymes are superimposed, conformational differences are seen, in addition to helix H2, also in H1, H7, H9 and H10. Helix H9 is the most variable

65

helix in the decarboxylase which also has the lowest sequence identity with the yeast enoyl-CoA isomerase. The changes in H2 and H9 influence the mode of assembly into hexamers.

The most notable difference in the yeast isomerase structure when compared to the structures of 2-enoyl-CoA hydratase-1 (Engel et al. 1996), dienoyl-CoA isomerase (Modis et al. 1998) and 4-CBA-dehalogenase (Benning et al. 1996) is the structural switch of the helices H9 and H10 of the C-terminal domain. This is an example of domain swapping, first described for the seminal ribonuclease dimer (Piccoli et al. 1992) and the diphtheria toxin dimer (Bennett et al. 1994). In the yeast enoyl-CoA isomerase, H9 and H10 are positioned so that they fold over the core domain and cover the active site of the same subunit (Fig. 13). In the structures of hydratase-1, dienoyl-CoA isomerase and dehalogenase, however, H9 and H10 protrude away from the core domain and cover the active site of the neighbouring subunit (Fig. 5). The similar structural switch of H9 and H10 as in isomerase is also seen in the structure of methylmalonyl-CoA decarboxylase (Benning et al. 2000). Despite the domain swapping, in the trimers of the hydratase/isomerase enzymes, the positions of H9 and H10 are equivalent, and the active sites are always covered by the C-terminal domain of either the same or the adjacent subunit.

6.3.2 Comparison at the hexamer level

All the enzymes of the hydratase/isomerase superfamily for which the structure is known form crystallographic hexamers, although the 4-CBA-CoA dehalogenase is a trimer in solution (Benning et al. 1996). In addition, all these hexamers are formed by dimerization of two trimeric disks (Engel et al. 1996, Modis et al. 1998, Benning et al. 2000, Kurimoto et al. 2001).

The variability of the helices H2 and H9, which are close to the intertrimer space, correlates with interesting variability in the mode of assembly of trimers into hexamers in the hydratase/isomerase superfamily. In order to compare the hexameric assemblies, the crystallographic packing information was used to construct hexamers of all the known structures, and the assembly was subsequently analyzed by superposition on the neutral pH form of the yeast enoyl-CoA isomerase. In this analysis, the structure of the AUH protein (Kurimoto et al. 2001) was not used, since it was published after the preparation of the original article IV. The difference in assembly can be described by two parameters: a rotation (Κ) of the trimers with respect to each other and a shift (∆) along the axis of rotation, indicating the difference in the distance of the two trimers compared to the neutral pH form isomerase hexamer (original article IV, table 2, Fig. 4). It was found that the orientation of the trimeric disks with respect to each other is not conserved, nor is the distance between the disks. The trimers in hydratase-1 (Engel et al. 1996) and the dienoyl-CoA isomerase (Modis et al. 1998) have the same relative orientation with respect to each other, but the distance between the trimers is not the same. This was also found for the low pH form of isomerase and dehalogenase (Benning et al. 1996). A relatively short distance between the trimers is observed for the dehalogenase, although it occurs as a trimer in solution, whereas the longest distance is observed for the

66

decarboxylase, although this enzyme is a hexamer in solution (Benning et al. 2000). The neutral pH form of the yeast isomerase is the most compact structure with the largest number of intertrimer contacts and the shortest distance between the trimeric disks. The low pH form isomerase, however, has the lowest number of contacts. Only a few contacts are also found for the trimeric dehalogenase. Interestingly, the variable helices H2 and H9 are always involved in intertrimer contacts, in many cases also H1. Although, at the monomer level, the overall fold of the enzymes of the hydratase/isomerase superfamily resemble each other and the secondary structure elements are mainly conserved, the mode of assembly of trimer disks into hexamers is clearly not a conserved feature of this enzyme superfamily.

7 Conclusions

The ∆3-∆2-enoyl-CoA isomerase (Eci1p) from S. cerevisiae, encoded by the ECI1 gene, is a monofunctional, peroxisomal enzyme catalyzing the isomerization of trans/cis-3-enoyl-CoA to trans-2-enoyl-CoA. Eci1p was shown to be, at neutral pH, a hexameric protein formed of six identical 32-kDa subunits. According to the structural studies, the subunits are assembled into a hexamer as a dimer of trimers.

The unliganded structure of yeast isomerase was determined using a crystal obtained at low pH, and the structure with an active site ligand from a crystal at neutral pH. The overall fold of the yeast ∆3-∆2-enoyl-CoA isomerase was found to be typical for the hydratase/isomerase superfamily in that it had a N-terminal spiral core domain for catalysis and a C-terminal α-helical trimerization domain. Some conformational differences could, however, be detected between the two yeast isomerase crystal forms. These rearrangements correlated with the much tighter assembly of the isomerase trimers into hexamers at neutral pH compared to the loosely interacting trimers in the low pH form, suggesting that the assembly could be pH-dependent. A comparison with the other known structures belonging to the hydratase/isomerase superfamily further showed that the mode of assembly of the trimers into hexamers is not a conserved feature.

Site-directed mutagenesis and structural data indicated that Glu158 is the only catalytic residue of Eci1p involved in shuttling the proton from the C2 of the substrate to the C4 of the product. We suggest that the negatively charged transition state of the enzymatic reaction is stabilized by hydrogen bonding to the peptide NH groups of the conserved oxyanion hole residues, Ala70 and Leu126.

8 Future perspectives

Future studies on the yeast ∆3-∆2-enoyl-CoA isomerase will focus on crystallographic binding studies with substrates, both with cis- and trans-3 double bonds, and substrate analogues of variable chain length and with better resolution, in order to more accurately determine the catalytic mechanism of the yeast enoyl-CoA isomerase and to elucidate the role of the empty apolar cavity in the binding of longer fatty acyl chains. Also, the substrate chain length specificity will be studied by enzyme assays. In addition, structural studies on the human mitochondrial ∆3-∆2-enoyl-CoA isomerase and the rat MFE-1 are ongoing in our laboratory. These enzymes also belong to the hydratase/isomerase superfamily but have different catalytic residues for the enoyl-CoA isomerase reaction compared to the yeast enoyl-CoA isomerase and also occur in different oligomeric forms, the mitochondrial isomerase being a trimer and the MFE-1 a monomer. The crystal structures of these enzymes are expected to give invaluable information about the importance of different catalytic residues at the active site architecture.

9 References

Aas M (1970) Organ and subcellular distribution of fatty acid activating enzymes in the rat. Biochim Biophys Acta 231: 32-47.

Aas M & Bremer H (1968) Short-chain fatty acid activation in rat liver. A new assay procedure for the enzymes and studies on their intracellular localization. Biochim Biophys Acta 164: 157-166.

Abagyan RA & Totrov MM (1994) Biased probability Monte Carlo conformational searches and electrostatic calculations for peptides and proteins. J Mol Biol 235: 983-1002.

Antonenkov VD, Van Veldhoven PP, Waelkens E & Mannaerts GP (1997) Substrate specificities of 3-oxoacyl-CoA thiolase A and sterol carrier protein 2/3-oxoacyl-CoA thiolase purified from normal rat liver peroxisomes. Sterol carrier protein 2/3-oxoacyl-CoA thiolase is involved in the metabolism of 2-methyl-branched fatty acids and bile acid intermediates. J Biol Chem 272: 26023-26031.

Aubourg P, Mosser J, Douar AM, Sarde CO, Lopez J & Mandel JL (1993) Adrenoleukodystrophy gene: unexpected homology to a protein involved in peroxisome biogenesis. Biochimie 75: 293-302.

Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA & Struhl K (1989) Current Protocols in Molecular Biology. John Wiley & Sons, Inc. New York, USA.

Babbitt PC & Gerlt JA (1997) Understanding enzyme superfamilies. Chemistry as the fundamental determinant in the evolution of new catalytic activities. J Biol Chem 272: 30591-30594.

Babbitt PC, Kenyon GL, Martin BM, Charest H, Sylvestre M, Scholten JD, Chang KH, Liang PH & Dunaway-Mariano D (1992) Ancestry of the 4-chlorobenzoate dehalogenase: analysis of amino acid sequence identities among families of acyl:adenyl ligases, enoyl-CoA hydratases/isomerases, and acyl-CoA thioesterases. Biochemistry 31: 5594-5604.

Baldwin GS, Chandler R, Scanlon DB & Weinstock J (1986) Identification of a gastrin binding protein in porcine gastric mucosal membranes by covalent cross-linking with iodinated gastrin. J Biol Chem 261: 12252-12257.

Batta AK, Salen G, Shefer S, Dayal B & Tint GS (1983) Configuration at C-25 in 3α, 7α, 12α-trihydroxy-5β-cholestan-26-oic acid isolated from human bile. J Lipid Res 24: 94-96.

Bell AF, Wu J, Feng Y & Tonge PJ (2001) Involvement of glycine 141 in substrate activation by enoyl-CoA hydratase. Biochemistry 40: 1725-1733.

Bennett MJ, Choe S & Eisenberg D (1994) Domain swapping: entangling alliances between proteins. Proc Natl Acad Sci USA 91: 3127-3131.

Benning MM, Haller T, Gerlt JA & Holden HM (2000) New reactions in the crotonase superfamily: structure of methylmalonyl CoA decarboxylase from Escherichia coli. Biochemistry 2000 39: 4630-4639.

Benning MM, Taylor KL, Liu RQ, Yang G, Xiang H, Wesenberg G, Dunaway-Mariano D & Holden HM (1996) Structure of 4-chlorobenzoyl coenzyme A dehalogenase determined to 1.8 Å resolution: an enzyme catalyst generated via adaptive mutation. Biochemistry 35: 8103-8109.

70

Bernlohr DA, Simpson MA, Hertzel AV & Banaszak LJ (1997) Intracellular lipid-binding proteins and their genes. Annu Rev Nutr 17: 277-303.

Bout A, Teunissen Y, Hashimoto T, Benne R & Tager JM (1988) Nucleotide sequence of human peroxisomal 3-oxoacyl-CoA thiolase. Nucleic Acids Res 16: 10369.

Boynton ZL, Bennet GN & Rudolph FB (1996) Cloning, sequencing, and expression of clustered genes encoding β-hydroxybutyryl-coenzyme A (CoA) dehydrogenase, crotonase, and butyryl-CoA dehydrogenase from Clostridium acetobutylicum ATCC 824. J Bacteriol 178: 3015-3024.

Brennan LE, Nakagawa J, Egger D, Bienz K & Moroni C (1999) Characterisation and mitochondrial localisation of AUH, an AU-specific RNA-binding enoyl-CoA hydratase. Gene 228: 85-91.

Brown GK, Hunt SM, Scholem R, Fowler K, Grimes A, Mercer JF, Truscott RM, Cotton RG, Rogers JG & Danks DM (1982) β-hydroxyisobutyryl coenzyme A deacylase deficiency: a defect in valine metabolism associated with physical malformations. Pediatrics 70: 532-538.

Caira F, Cherkaoui-Malki M, Hoefler G & Latruffe N (1996) Cloning and tissue expression of two cDNAs encoding the peroxisomal 2-enoyl-CoA hydratase/3-hydroxyacyl-CoA dehydrogenase in the guinea pig liver. FEBS Lett 378: 57-60.

Carpenter K, Pollitt RJ & Middleton B (1992) Human liver long-chain 3-hydroxyacyl-coenzyme A dehydrogenase is a multifunctional membrane-bound β-oxidation enzyme of mitochondria. Biochem Biophys Res Commun 183: 443-448.

Casteels M, Schepers L, Van Veldhoven PP, Eyssen HJ & Mannaerts GP (1990) Separate peroxisomal oxidases for fatty acyl-CoAs and trihydroxycoprostanoyl-CoA in human liver. J Lipid Res 31: 1865-1872.

Chen GL, Balfe A, Erwa W, Hoefler G, Gaertner J, Aikawa J & Chen WW (1991) Import of human bifunctional enzyme into peroxisomes of human hepatoma cells in vitro. Biochem Biophys Res Commun 178: 1084-1091.

Chen LS, Jin SJ & Tserng KY (1994) Purification and mechanism of ∆3,∆5-t-2,t-4-dienoyl-CoA isomerase from rat liver. Biochemistry 33: 10527-10534.

Cherry JM, Ball C, Weng S, Juvik G, Schmidt R, Adler C, Dunn B, Dwight S, Riles L, Mortimer RK & Botstein D (1997) Genetic and physical maps of Saccharomyces cerevisiae. Nature 387(66632 Suppl): 67-73.

Clarkson J, Tonge PJ, Taylor KL & Dunaway-Mariano D, Carey PR (1997) Raman study of the polarizing forces promoting catalysis in 4-chlorobenzoate-CoA dehalogenase. Biochemistry 36: 10192-10199.

Collaborative Computational Project, Number 4. (1994) The CCP4 Suite: Programs for Protein Crystallography. Acta Cryst D50: 760-763.

Cooper TG & Beevers H (1969) β-Oxidation in glyoxysomes from castor bean endosperm. J Biol Chem 244: 3514-3520.

Cowtan KD & Main P (1996) Phase combination and cross validation in iterated density-modification calculations. Acta Cryst D52: 43-48.

Croes K, Casteels M, Asselberghs S, Herdewijn P, Mannaerts GP & Van Veldhoven PP (1997) Formation of a 2-methyl-branched fatty aldehyde during peroxisomal α-oxidation. FEBS Lett 412: 643-645.

De Nys K, Meyhi E, Mannaerts GP, Fransen M & Van Veldhoven PP (2001) Characterisation of human peroxisomal 2,4-dienoyl-CoA reductase. Biochim Biophys Acta 1533: 66-72.

Dieuaide-Noubhani M, Asselberghs S, Mannaerts GP & Van Veldhoven PP (1997b) Evidence that multifunctional protein 2, and not multifunctional protein 1, is involved in the peroxisomal β-oxidation of pristanic acid. Biochem J 325: 367-373.

Dieuaide-Noubhani M, Novikov D, Vandekerckhove J, Veldhoven PP & Mannaerts GP (1997a) Identification and characterization of the 2-enoyl-CoA hydratases involved in peroxisomal β-oxidation in rat liver. Biochem J 321: 253-259.

Dirusso CC (1990) Primary sequence of the Escherichia coli fadBA operon, encoding the fatty acid-oxidizing multienzyme complex, indicates a high degree of homology to eucaryotic enzymes. J Bacteriol 172: 6459-6468.

71

Dmochowska A, Dignard D, Maleszka R & Thomas DY (1990) Structure and transcriptional control of the Saccharomyces cerevisiae POX1 gene encoding acyl-coenzyme A oxidase. Gene 88: 247-252.

Dommes V, Baumgart C & Kunau WH (1981) Degradation of unsaturated fatty acids in peroxisomes. Existence of a 2,4-dienoyl-CoA reductase pathway. J Biol Chem 256: 8259-8262.

Dommes V & Kunau WH (1984) 2,4-Dienoyl coenzyme A reductases from bovine liver and Escherichia coli. Comparison of properties. J Biol Chem 259: 1781-1788.

Dommes V, Luster W, Cvetanovic M, Kunau WH (1982) Purification by affinity chromatography of 2,4-dienoyl-CoA reductases from bovine liver and Escherichia coli. Eur J Biochem 125: 335-341.

D'Ordine RL, Tonge PJ, Carey PR & Anderson VE (1994) Electronic rearrangement induced by substrate analog binding to the enoyl-CoA hydratase active site: evidence for substrate activation. Biochemistry 33: 12635-12643.

Driscoll JR & Taber HW (1992) Sequence organization and regulation of the Bacillus subtilis menBE operon. J Bacteriol 174: 5063-5071.

Eaton S, Bartlett K & Pourfarzam M (1996) Mammalian mitochondrial β-oxidation. Biochem J 320: 345-357.

Egland PG, Pelletier DA, Dispensa M, Gibson J & Harwood CS (1997) A cluster of bacterial genes for anaerobic benzene ring biodegradation. Proc Natl Acad Sci USA 94: 6484-6489.

Eichler K, Bourgis F, Buchet A, Kleber HP & Mandrand-Berthelot MA (1994) Molecular characterization of the cai operon necessary for carnitine metabolism in Escherichia coli. Mol Microbiol 13: 775-786.

Einerhand AW, Kos WT, Distel B, & Tabak HF (1993) Characterization of a transcriptional control element involved in proliferation of peroxisomes in yeast in response to oleate. Eur J Biochem 214: 323-331.

Einerhand AW, Voorn-Brouwer TM, Erdmann R, Kunau WH & Tabak HF (1991) Regulation of transcription of the gene coding for peroxisomal 3-oxoacyl-CoA thiolase of Saccharomyces cerevisiae. Eur J Biochem 200: 113-122.

Elssner T, Engemann C, Baumgart K & Kleber HP (2001) Involvement of coenzyme A esters and two new enzymes, an enoyl-CoA hydratase and a CoA-transferase, in the hydration of crotonobetaine to L-carnitine by Escherichia coli. Biochemistry 40: 11140-11148.

Engel CK, Kiema TR, Hiltunen JK & Wierenga RK (1998) The crystal structure of enoyl-CoA hydratase complexed with octanoyl-CoA reveals the structural adaptations required for binding of a long chain fatty acid-CoA molecule. J Mol Biol 275: 847-859.

Engel CK, Mathieu M, Zeelen JP, Hiltunen JK & Wierenga RK (1996) Crystal structure of enoyl-coenzyme A (CoA) hydratase at 2.5 angstroms resolution: a spiral fold defines the CoA-binding pocket. EMBO J 15: 5135-5145.

Engeland K & Kindl H (1991) Purification and characterization of a plant peroxisomal ∆2,∆3-enoyl-CoA isomerase acting on 3-cis-enoyl-CoA and 3-trans-enoyl-CoA. Eur J Biochem 196: 699-705.

Euler-Bertram S & Stoffel W (1990) Purification and characterization of bovine liver 3-cis-2-trans-enoyl-CoA isomerase. Biol Chem Hoppe Seyler 371: 603-610.

Ferdinandusse S, Denis S, IJlst L, Dacremont G, Waterham HR & Wanders RJ (2000) Subcellular localization and physiological role of α-methylacyl-CoA racemase. J Lipid Res 41: 1890-1896.

Fersht A (1999) Structure and mechanism in protein science. A guide to enzyme catalysis and protein folding. W. H. Freeman and Company, New York, USA.

Filipits M, Simon MM, Rapatz W, Hamilton B & Ruis H (1993) A Saccharomyces cerevisiae upstream activating sequence mediates induction of peroxisome proliferation by fatty acids. Gene 132: 49-55.

Filppula SA, Sormunen RT, Hartig A, Kunau WH & Hiltunen JK (1995) Changing stereochemistry for a metabolic pathway in vivo. Experiments with the peroxisomal β-oxidation in yeast. J Biol Chem 270: 27453-27457.

Filppula SA, Yagi AI, Kilpelainen SH, Novikov D, FitzPatrick DR, Vihinen M, Valle D & Hiltunen JK (1998) ∆3,5-∆2,4-dienoyl-CoA isomerase from rat liver. Molecular characterization. J Biol Chem 273: 349-355.

72

FitzPatrick DR, Germain-Lee E & Valle D (1995) Isolation and characterization of rat and human cDNAs encoding a novel putative peroxisomal enoyl-CoA hydratase. Genomics 27: 457-466.

Fong JC & Schulz H (1977) Purification and properties of pig heart crotonase and the presence of short chain and long chain enoyl coenzyme A hydratases in pig and guinea pig tissues. J Biol Chem 252: 542-547.

Furuta S, Miyazawa S, Osumi T, Hashimoto T & Ui N (1980) Properties of mitochondrial and peroxisomal enoyl-CoA hydratases from rat liver. J Biochem (Tokyo) 88: 1059-1070.

Gasson MJ, Kitamura Y, McLauchlan WR, Narbad A, Parr AJ, Parsons EL, Payne J, Rhodes MJ & Walton NJ (1998) Metabolism of ferulic acid to vanillin. A bacterial gene of the enoyl-SCoA hydratase/isomerase superfamily encodes an enzyme for the hydration and cleavage of a hydroxycinnamic acid SCoA thioester. J Biol Chem 273: 4163-4170.

Geisbrecht BV, Liang X, Morrell JC, Schulz H & Gould SJ (1999a) The mouse gene PDCR encodes a peroxisomal ∆2,∆4-dienoyl-CoA reductase. J Biol Chem 274: 25814-25820.

Geisbrecht BV, Zhang D, Schulz H & Gould SJ (1999b) Characterization of PECI, a novel monofunctional ∆3,∆2-enoyl-CoA isomerase of mammalian peroxisomes. J Biol Chem 274: 21797-21803.

Geisbrecht BV, Zhu D, Schulz K, Nau K, Morrell JC, Geraghty M, Schulz H, Erdmann R & Gould SJ (1998) Molecular characterization of Saccharomyces cerevisiae ∆3,∆2-enoyl-CoA isomerase. J Biol Chem 273: 33184-33191.Gibbons GF, Islam K & Pease RJ (2000) Mobilisation of triacylglycerol stores. Biochim Biophys Acta 1483: 37-57.

Gould SJ, Keller G-A, Hosken N, Wilkinson J & Subramani S (1989) A conserved tripeptide sorts proteins to peroxisomes. J Cell Biol 108: 1657-1664.

Grogan G, Roberts GA, Bougioukou D, Turner NJ & Flitsch SL (2001) The desymmetrization of bicyclic β-diketones by an enzymatic retro-Claisen reaction. A new reaction of the crotonase superfamily. J Biol Chem 276: 12565-12572.

Gurvitz A, Mursula AM, yagi AI, Hartig A, Ruis H, Rottensteiner H & Hiltunen JK (1999) Alternatives to the isomerase-dependent pathway for the β-oxidation of oleic acid are dispensable in Saccharomyces cerevisiae. Identification of YOR180c/DCI1 encoding peroxisomal ∆3,5-∆2,4-dienoyl-CoA isomerase. J Biol Chem 27: 24514-24521.

Gurvitz A, Rottensteiner H, Hiltunen JK, Binder M, Dawes IW, Ruis H & Hamilton B (1997b) Regulation of the yeast SPS19 gene encoding peroxisomal 2,4-dienoyl-CoA reductase by the transcription factors Pip2p and Oaf1p: β-oxidation is dispensable for Saccharomyces cerevisiae sporulation in acetate medium. Mol Microbiol 26: 675-685.

Gurvitz A, Rottensteiner H, Kilpelainen SH, Hartig A, Hiltunen JK, Binder M, Dawes IW & Hamilton B (1997a) The Saccharomyces cerevisiae peroxisomal 2,4-dienoyl-CoA reductase is encoded by the oleate-inducible gene SPS19. J Biol Chem 272: 22140-22147.

Görner W, Durchschlag E, Martinez-Pastor MT, Estruch F, Ammerer G, Hamilton B, Ruis H & Schüller C (1998) Nuclear localization of the C2H2 zinc finger protein Msn2p is regulated by stress and protein kinase A activity. Genes Dev 12: 586-597.

Haapalainen AM, van Aalten DM, Merilainen G, Jalonen JE, Pirila P, Wierenga RK, Hiltunen JK & Glumoff T (2001) Crystal structure of the liganded SCP-2-like domain of human peroxisomal multifunctional enzyme type 2 at 1.75 Å resolution. J Mol Biol 313: 1127-1138.

Hakkola EH & Hiltunen JK (1993) The existence of two mitochondrial isoforms of 2,4-dienoyl-CoA reductase in the rat. Eur J Biochem 215: 199-204.

Haller T, Buckel T, Retey J & Gerlt JA (2000) Discovering new enzymes and metabolic pathways: conversion of succinate to propionate by Escherichia coli. Biochemistry 39: 4622-4629.

Hamilton JA (1998) Fatty acid transport: difficult or easy? J Lipid Res 39: 467-481. Hashimoto T, Shindo Y, Souri M & Baldwin GS (1996) A new inhibitor of mitochondrial fatty acid

oxidation. J Biochem (Tokyo) 119: 1196-1201. Hass GM & Hill RL (1969) The subunit structure of crotonase. J Biol Chem 244: 6080-6086. Hawes JW, Jaskiewicz J, Shimomura Y, Huang B, Bunting J, Harper ET & Harris RA (1996)

Primary structure and tissue-specific expression of human β-hydroxyisobutyryl-coenzyme A hydrolase. J Biol Chem 271: 26430-26434.

He XY, Shoukry K, Chu C, Yang J, Sprecher H & Schulz H (1995) Peroxisomes contain ∆3,5,∆ 2,4-dienoyl-CoA isomerase and thus possess all enzymes required for the β-oxidation of unsaturated

73

fatty acids by a novel reductase-dependent pathway. Biochem Biophys Res Commun 215: 15-22.

He XY, Yang SY & Schulz H (1989) Assay of L-3-hydroxyacyl-coenzyme A dehydrogenase with substrates of different chain lengths. Anal Biochem 180: 105-109.

Hesler CB, Olymbios C & Haldar D (1990) Transverse-plane topography of long-chain acyl-CoA synthetase in the mitochondrial outer membrane. J Biol Chem 265: 6600-6605.

Hettema EH, van Roermund CW, Distel B, van den Berg M, Vilela C, Rodrigues-Pousada C, Wanders RJ & Tabak HF (1996) The ABC transporter proteins Pat1 and Pat2 are required for import of long-chain fatty acids into peroxisomes of Saccharomyces cerevisiae. EMBO J 15: 3813-3822.

Hijikata M, Wen JK, Osumi T, Hashimoto T (1990) Rat peroxisomal 3-ketoacyl-CoA thiolase gene. Occurrence of two closely related but differentially regulated genes J Biol Chem 265: 4600-4606.

Hiltunen JK & Qin Y-M (2000) β-oxidation – strategies for the metabolism of a wide variety of acyl-CoA esters. Biochim Biophys Acta 1484: 117-128.

Hiltunen JK, Wenzel B, Beyer A, Erdmann R, Fosså A & Kunau WH (1992) Peroxisomal multifunctional β-oxidation protein of Saccharomyces cerevisiae. Molecular analysis of the fox2 gene and gene product. J Biol Chem 267: 6646-6653.

Hirose A, Kamijo K, Osumi T, Hashimoto T & Mizugaki M (1990) cDNA cloning of rat liver 2,4-dienoyl-CoA reductase. Biochim Biophys Acta 1049: 346-349.

Hoefler G, Forstner M, McGuinness MC, Hulla W, Hiden M, Krisper P, Kenner L, Ried T, Lengauer C, Zechner R, Chen GL & Moser HW (1994) cDNA cloning of the human peroxisomal enoyl-CoA hydratase: 3-hydroxyacyl-CoA dehydrogenase bifunctional enzyme and localization to chromosome 3q26.3-3q28: a free left Alu Arm is inserted in the 3' noncoding region. Genomics 19: 60-67.

Holden HM, Benning MM, Haller T & Gerlt JA (2001) The crotonase superfamily: divergently related enzymes that catalyze different reactions involving acyl coenzyme A thioesters. Acc Chem Res 34: 145-157.

Igual JC, Matallana E, Gonzalez-Bosch C, Franco L & Perez-Ortin JE (1991) A new glucose-repressible gene identified from the analysis of chromatin structure in deletion mutants of yeast SUC2 locus. Yeast 7: 379-389.

Ikeda Y, Dabrowski C & Tanaka K (1983) Separation and properties of five distinct acyl-CoA dehydrogenases from rat liver mitochondria. Identification of a new 2-methyl branched chain acyl-CoA dehydrogenase. J Biol Chem 258: 1066-1076.

Ikeda Y, Okamura-Ikeda K & Tanaka K (1985) Purification and characterization of short-chain, medium-chain, and long-chain acyl-CoA dehydrogenases from rat liver mitochondria. Isolation of the holo- and apoenzymes and conversion of the apoenzyme to the holoenzyme. J Biol Chem 260: 1311-1325.

Ishii N, Hijikata M, Osumi T & Hashimoto T (1987) Structural organization of the gene for rat enoyl-CoA hydratase:3-hydroxyacyl-CoA dehydrogenase bifunctional enzyme. J Biol Chem 262: 8144-8150.

Izai K, Uchida Y, Orii T, Yamamoto S & Hashimoto T (1992) Novel fatty acid β-oxidation enzymes in rat liver mitochondria. I. Purification and properties of very-long-chain acyl-coenzyme A dehydrogenase. J Biol Chem 267: 1027-1033.

Jackson S, Schaefer J, Middleton B & Turnbull DM (1995) Characterisation of a novel enzyme of human fatty acid β-oxidation: a matrix-associated, mitochondrial 2-enoyl-CoA hydratase. Biochem Biophys Res Commun 214: 247-253.

Jancarik J & Kim S-H (1991) Sparse matrix sampling: a screening method for crystallization of proteins. J Appl Cryst 24: 409-411.

Janssen U, Fink T, Lichter P & Stoffel W (1994) Human mitochondrial 3,2-trans-enoyl-CoA isomerase (DCI): gene structure and localization to chromosome 16p13.3. Genomics 23: 223-228.

Jones TA, Zou JY, Cowan SW & Kjeldgaard M (1991) Improved methods for binding protein models in electron density maps and the location of errors in these models. Acta Cryst A47: 110-119.

74

Kabsch W & Sander C (1983) Dictionary of protein secondary structure: Pattern recognition of hydrogen-bonded and geometrical features. Biopolymers 22: 2577-2637.

Kamijo T, Aoyama T, Komiyama A & Hashimoto T (1994) Structural analysis of cDNAs for subunits of human mitochondrial fatty acid β-oxidation trifunctional protein. Biochem Biophys Res Commun 199: 818-825.

Kamijo T, Aoyama T, Miyazaki J & Hashimoto T (1993) Molecular cloning of the cDNAs for the subunits of rat mitochondrial fatty acid β-oxidation multienzyme complex. Structural and functional relationships to other mitochondrial and peroxisomal β-oxidation enzymes. J Biol Chem 268: 26452-26460.

Kanazawa M, Ohtake A, Abe H, Yamamoto S, Satoh Y, Takayanagi M, Niimi H, Mori M & Hashimoto T (1993) Molecular cloning and sequence analysis of the cDNA for human mitochonrial short-chain enoyl-CoA hydratase. Enzyme & Protein 47: 9-13.

Kerner J & Hoppel C (2000) Fatty acid import into mitochondria. Biochim Biophys Acta 1486: 1-17.

Kiema TR, Engel CK, Schmitz W, Filppula SA, Wierenga RK & Hiltunen JK (1999) Mutagenic and enzymological studies of the hydratase and isomerase activities of 2-enoyl-CoA hydratase-1. Biochemistry 38: 2991-2999.

Kilponen JM & Hiltunen JK (1993) β-Oxidation of unsaturated fatty acids in humans. Isoforms of -∆3,∆2-enoyl-CoA isomerase. FEBS Lett 322: 299-303.

Kilponen JM, Häyrinen HM, Rehn M & Hiltunen JK (1994) cDNA cloning and amino acid sequence of human mitochondrial ∆3,∆2-enoyl-CoA isomerase: comparison of the human enzyme with its rat counterpart, mitochondrial short-chain isomerase. Biochem J 300: 1-5.

Kilponen JM, Palosaari PM & Hiltunen JK (1990) Occurrence of a long-chain ∆3,∆2-enoyl-CoA isomerase in rat liver. Biochem J 269: 223-226.

Kimura C, Kondo A, Koeda N, Yamanaka H & Mizugaki M (1984) Studies on the metabolism of unsaturated fatty acids. XV. Purification and properties of 2,4-dienoyl-CoA reductase from rat liver peroxisomes. J Biochem (Tokyo) 96: 1463-1469.

Knudsen J, Neergaard TB, Gaigg B, Jensen MV & Hansen JK (2000) Role of acyl-CoA binding protein in acyl-CoA metabolism and acyl-CoA-mediated cell signaling. J Nutr 130: 294S-298S.

Koivuranta KT, Hakkola EH & Hiltunen JK (1994) Isolation and characterization of cDNA for human 120 kDa mitochondrial 2,4-dienoyl-coenzyme A reductase. Biochem J 304: 787-792.

Kotti TJ, Savolainen K, Helander HM, Yagi A, Novikov DK, Kalkkinen N, Conzelmann E, Hiltunen JK & Schmitz W (2000) In mouse=α-methylacyl-CoA racemase, the same gene product is simultaneously located in mitochondria and peroxisomes. J Biol Chem 275: 20887-20895.

Krisans SK, Mortensen RM & Lazarow PB (1980) Acyl-CoA synthetase in rat liver peroxisomes. Computer-assisted analysis of cell fractionation experiments. J Biol Chem 255: 9599-9607.

Kunau WH, Buhne S, de la Garza M, Kionka C, Mateblowski M, Schultz-Borchard U & Thieringer R (1988) Comparative enzymology of β-oxidation. Biochem Soc Trans 16: 418-420.

Kunau WH & Dommes P (1978) Degradation of unsaturated fatty acids. Identification of intermediates in the degradation of cis-4-decenoly-CoA by extracts of beef-liver mitochondria. Eur J Biochem 91: 533-544.

Kurimoto K, Fukai S, Nureki O, Muto Y & Yokoyama S (2001) Crystal structure of human AUH protein, a single-stranded RNA binding homolog of enoyl-CoA hydratase. Structure 9: 1253-1263.

Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685.

Laskowski RA, MacArthur MW, Moss DS & Thornton JM (1993) PROCHECK: a program to check the stereochemical quality of protein structures. J Appl Cryst 26: 283-291.

Lau EY & Bruice TC (2001) The active site dynamics of 4-chlorobenzoyl-CoA dehalogenase. Proc Natl Acad Sci USA 98: 9527-9532.

Lazarow PB & De Duve C (1976) A fatty acyl-CoA oxidizing system in rat liver peroxisomes; enhancement by clofibrate, a hypolipidemic drug. Proc Natl Acad Sci USA 73: 2043-2046.

75

Lazo O, Contreras M, Yoshida Y, Singh AK, Stanley W, Weise M & Singh I (1990) Cellular oxidation of lignoceric acid is regulated by the subcellular localization of lignoceroyl-CoA ligases. J Lipid Res 31: 583-595.

Liang X, Le W, Zhang D & Schultz H (2001) Impact of the intramitochondrial enzyme organization on fatty acid oxidation. Biochem Soc Trans 29: 279-282.

Liang X, Zhu D & Schulz H (1999) ∆3,5,7,∆2,4,6-trienoyl-CoA isomerase, a novel enzyme that functions in the β-oxidation of polyunsaturated fatty acids with conjugated double bonds. J Biol Chem 274: 13830-13835.

Luo MJ, Smeland TE, Shoukry K & Schulz H (1994) ∆3,5,∆2,4-dienoyl-CoA isomerase from rat liver mitochondria. Purification and characterization of a new enzyme involved in the β-oxidation of unsaturated fatty acids. J Biol Chem 269: 2384-2388.

Luo Y, Karpichev IV, Kohanski RA & Small GM (1996) Purification, identification, and properties of a Saccharomyces cerevisiae oleate-activated upstream activating sequence-binding protein that is involved in the activation of POX1. J Biol Chem 271 :12068-12075.

Luthria DL, Baykousheva SP & Sprecher H (1995) Double bond removal from odd-numbered carbons during peroxisomal β-oxidation of arachidonic acid requires both 2,4-dienoyl-CoA reductase and ∆3,5,∆2,4-dienoyl-CoA isomerase. J Biol Chem 270: 13771-13776.

Mannaerts GP, Van Veldhoven P, Van Broekhoven A, Vandebroek G & Debeer LJ (1982) Evidence that peroxisomal acyl-CoA synthetase is located at the cytoplasmic side of the peroxisomal membrane. Biochem J 204: 17-23.

Mantamadiotis T, Sobieszczuk P, Weinstock J & Baldwin GS (1993) Nucleotide sequence encoding a novel member of the hydratase/dehydrogenase family. Biochim Biophys Acta 1170: 211-215.

Margolin W, Bramhill D & Long SR (1995) The dnaA gene of Rhizobium meliloti lies within an unusual gene arrangement. J Bacteriol 177: 2892-2900.

Meganathan R & Bentley R (1979) Menaquinone (vitamin K2) biosynthesis: conversion of o-succinylbenzoic acid to 1,4-dihydroxy-2-naphthoic acid by Mycobacterium phlei enzymes. J Bacteriol 140: 92-98.

Mihalik SJ, Rainville AM & Watkins PA (1995) Phytanic acid α-oxidation in rat liver peroxisomes. Production of α-hydroxyphytanoyl-CoA and formate is enhanced by dioxygenase cofactors. Eur J Biochem 232: 545-551.

Minami-Ishii N, Taketani S, Osumi T & Hashimoto T (1989) Molecular cloning and sequence analysis of the cDNA for rat mitochondrial enoyl-CoA hydratase. Structural and evolutionary relationships linked to the bifunctional enzyme of the peroxisomal β-oxidation system. Eur J Biochem 185: 73-78.

Miyazawa S, Furuta S, Osumi T, Hashimoto T & Ui N (1981) Properties of peroxisomal 3-ketoacyl-coA thiolase from rat liver. J Biochem (Tokyo) 90: 511-519.

Modis Y, Filppula SA, Novikov DK, Norledge B, Hiltunen JK & Wierenga RK (1998) The crystal structure of dienoyl-CoA isomerase at 1.5 Å resolution reveals the importance of aspartate and glutamate sidechains for catalysis. Structure 6: 957-970.

Mori T, Tsukamoto T, Mori H, Tashiro Y & Fujiki Y (1991) Molecular cloning and duduced amino acid sequence of nonspecific lipid transfer (sterol carrier protein 2) of rat liver: a higher molecular mass (60 kDa) protein contains the primary sequence of nonspecific lipid transfer protein as its C-terminal part. Proc Natl Acad Sci USA 88: 4338-4342.

Mosser J, Douar AM, Sarde CO, Kioschis P, Feil R, Moser H, Poustka AM, Mandel JL & Aubourg P (1993) Putative X-linked adrenoleukodystrophy gene shares unexpected homology with ABC transporters. Nature 361: 726-730.

Mosser J, Lutz Y, Stoeckel ME, Sarde CO, Kretz C, Douar AM, Lopez J, Aubourg P & Mandel JL (1994) The gene responsible for adrenoleukodystrophy encodes a peroxisomal membrane protein. Hum Mol Genet 3: 265-271.

Mu FT, Baldwin G, Weinstock J, Stockman D & Toh BH (1987) Monoclonal antibody to the gastrin receptor on parietal cells recognizes a 78-kDa protein. Proc Natl Acad Sci USA 84: 2698-2702.

76

Müller R, Thiele J, Klages U & Lingens F (1984) Incorporation of [18O]water into 4-hydroxybenzoic acid in the reaction of 4-chlorobenzoate dehalogenase from pseudomonas spec. CBS 3. Biochem Biophys Res Commun 124: 178-182.

Müller-Newen G, Janssen U & Stoffel W (1995) Enoyl-CoA hydratase and isomerase form a superfamily with a common active-site glutamate residue. Eur J Biochem 228: 68-73.

Müller-Newen G & Stoffel W (1991) Mitochondrial 3-2trans-Enoyl-CoA isomerase. Purification, cloning, expression, and mitochondrial import of the key enzyme of unsaturated fatty acid β-oxidation. Biol Chem Hoppe Seyler 372: 613-624.

Müller-Newen G & Stoffel W (1993) Site-directed mutagenesis of putative active-site amino acid residues of 3,2-trans-enoyl-CoA isomerase, conserved within the low-homology isomerase/hydratase enzyme family. Biochemistry 32: 11405-11412.

Murshudov GN, Vagin AA & Dodson EJ (1997) Refinement of Macromolecular Structures by the Maximum-Likelihood Method. Acta Cryst D 55: 247-255.

Nakagawa J & Moroni C (1997) A 20-amino-acid autonomous RNA-binding domain contained in an enoyl-CoA hydratase. Eur J Biochem 244: 890-899.

Nakagawa J, Waldner H, Meyer-Monard S, Hofsteenge J, Jeno P & Moroni C (1995) AUH, a gene encoding an AU-specific RNA binding protein with intrinsic enoyl-CoA hydratase activity. Proc Natl Acad Sci USA 92: 2051-2055.

Navaza J (1994) AmoRe: an automated package for molecular replacement. Acta Cryst A50: 157-163.

van Nieuwenhoven FA, Van der Vusse GJ & Glatz JF (1996) Membrane-associated and cytoplasmic fatty acid-binding proteins. Lipids 31: S223-S227.

Osumi T & Hashimoto T (1979) Peroxisomal β-oxidation system of rat liver. Copurification of enoyl-CoA hydratase and 3-hydroxyacyl-CoA dehydrogenase. Biochem Biophys Res Commun 89: 580-584.

Osumi T & Hashimoto T (1980) Purification and properties of mitochondrial and peroxisomal 3-hydroxyacyl-CoA dehydrogenase from rat liver. Arch Biochem Biophys 203: 372-383.

Osumi T, Hashimoto T & Ui N (1980) Purification and properties of acyl-CoA oxidase from rat liver. J Biochem (Tokyo) 87: 1735-1746.

Osumi T, Ishii N, Hijikata M, Kamijo K, Ozasa H, Furuta S, Miyazawa S, Kondo K, Inoue K, Kagamiyama H & Hashimoto T (1985) Molecular cloning and nucleotide sequence of the cDNA for rat peroxisomal enoyl-CoA:hydratase-3-hydroxyacyl-CoA dehydrogenase bifunctional enzyme. J Biol Chem 260: 8905-8910.

Otwinowski Z & Minor W (1997) Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol 276: 307-326.

Palosaari PM & Hiltunen JK (1990) Peroxisomal bifunctional protein from rat liver is a trifunctional enzyme possessing 2-enoyl-CoA hydratase, 3-hydroxyacyl-CoA dehydrogenase, and ∆3,∆2-enoyl-CoA isomerase activities. J Biol Chem 265: 2446-2449.

Palosaari PM, Kilponen JM, Sormunen RT, Hassinen IE & Hiltunen JK (1990) ∆3,∆2-enoyl-CoA isomerases. Characterization of the mitochondrial isoenzyme in the rat. J Biol Chem 265: 3347-3353.

Palosaari PM, Vihinen M, Mantsala PI, Alexson SE, Pihlajaniemi T & Hiltunen JK (1991) Amino acid sequence similarities of the mitochondrial short chain ∆3,∆2-enoyl-CoA isomerase and peroxisomal multifunctional ∆3,∆2-enoyl-CoA isomerase, 2-enoyl-CoA hydratase, 3-hydroxyacyl-CoA dehydrogenase enzyme in rat liver. The proposed occurrence of isomerization and hydration in the same catalytic domain of the multifunctional enzyme. J Biol Chem 266: 10750-10753.

Pande SV (1975) A mitochondrial carnitine acylcarnitine translocase system. Proc Nat Acad Sci 72: 883-887.

Pande SV & Parvin R (1976) Characterization fo carnitine acylcarnitine translocase system of heart mitochondria. J Biol Chem 251: 6683-6691.

Patton SM, Cropp TA & Reynolds KA (2000) A novel ∆3,∆2-enoyl-CoA isomerase involved in the biosynthesis of the cyclohexanecarboxylic acid-derived moiety of the polyketide ansatrienin A. Biochemistry 39: 7595-7604.

77

Pelletier DA & Harwood CS (1998) 2-Ketocyclohexanecarboxyl coenzyme A hydrolase, the ring cleavage enzyme required for anaerobic benzoate degradation by Rhodopseudomonas palustris. J Bacteriol 180: 2330-2336.

Perrakis A, Morris RJH & Lamzin VS (1999) Automated protein model building combined with iterative structure refinement. Nature Struct Biol 6: 458-463.

Piccoli R, Tamburrini M, Piccialli G, Di Donato A, Parente A & D'Alessio G (1992) The dual-mode quaternary structure of seminal RNase. Proc Natl Acad Sci USA 89: 1870-1874.

Priefert H, Rabenhorst J & Steinbuchel A (2001) Biotechnological production of vanillin. Appl Microbiol Biotechnol 56: 296-314.

Prydz K, Kase BF, Bjorkhem I & Pedersen JI (1988) Subcellular localization of 3α,7α-dihydroxy- and 3α,7α,12α-trihydroxy-5β-cholestanoyl-coenzyme A ligase(s) in rat liver. J Lipid Res 29: 997-1004.

Qin YM, Haapalainen AM, Conry D, Cuebas DA, Hiltunen JK, Novikov DK (1997b) Recombinant 2-enoyl-CoA hydratase derived from rat peroxisomal multifunctional enzyme 2: role of the hydratase reaction in bile acid synthesis. Biochem J 328: 377-382.

Qin YM, Poutanen MH, Helander HM, Kvist AP, Siivari KM, Schmitz W, Conzelmann E, Hellman U & Hiltunen JK (1997a) Peroxisomal multifunctional enzyme of β-oxidation metabolizing D-3-hydroxyacyl-CoA esters in rat liver: molecular cloning, expression and characterization. Biochem J 321: 21-28.

Rasmussen JT, Boerchers T & Knudsen J (1990) Comparison of the binding affinities of acyl-CoA-binding protein and fatty-acid-binding protein for long-chain acyl-CoA esters. Biochem J 265: 849-855.

Reddy JK & Hashimoto T (2001) Peroxisomal β-oxidation and peroxisome proliferator-activated receptor α: an adaptive metabolic system. Annu Rev Nutr 21 : 193-230.

van Roermund CTW, Elgersma Y, Singh N, Wanders RJA & Tabak HK (1995) The membrane of peroxisomes in Saccharomyces cerevisiae is impermeable to NAD(H) and acetyl-CoA under in vivo conditions. EMBO J 14: 3480-3486.

Rottensteiner H, Kal AJ, Filipits M, Binder M, Hamilton B, Tabak HF & Ruis H (1996) Pip2p: a transcriptional regulator of peroxisome proliferation in the yeast Saccharomyces cerevisiae. EMBO J 15: 2924-2934.

Sambrook J, Fritsch EF & Maniatis T (1989) Molecular Cloning: A Laboratory Manual, 2nd Ed., Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

Sanger F, Nicklen S & Coulson AR (1977) DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci U S A 74: 5463-5467.

Sato S, Hayashi M, Imamura S, Ozeki Y & Kawaguchi A (1992) Primary structures of the genes, faoA and faoB, from Pseudomonas fragi B-0771 which encode the two subunits of the HDT multienzyme complex involved in fatty acid β-oxidation. J Biochem (Tokyo) 111: 8-15.

Schepers L, Casteels M, Vamecq J, Parmentier G, Van Veldhoven PP & Mannaerts GP (1988) β-Oxidation of the carboxyl side chain of prostaglandin E2 in rat liver peroxisomes and mitochondria. J Biol Chem 263: 2724-2731.

Schepers L, Casteels M, Verheyden K, Parmentier G, Asselberghs S, Eyssen HJ & Mannaerts GP (1989) Subcellular distribution and characteristics of trihydroxycoprostanoyl-CoA synthetase in rat liver. Biochem J 257: 221-229.

Schepers L, Van Veldhoven PP, Casteels M, Eyssen HJ & Mannaerts GP (1990) Presence of three acyl-CoA oxidases in rat liver peroxisomes. An inducible fatty acyl-CoA oxidase, a noninducible fatty acyl-CoA oxidase, and a noninducible trihydroxycoprostanoyl-CoA oxidase. J Biol Chem 265: 5242-5246.

Schmitz A, Gartemann KH, Fiedler J, Grund E & Eichenlaub R (1992) Cloning and sequence analysis of genes for dehalogenation of 4-chlorobenzoate from Arthrobacter sp. strain SU. Appl Environ Microbiol 58: 4068-4071.

Schmitz W, Albers C, Fingerhut R & Conzelmann E (1995) Purification and characterization of an α-methylacyl-CoA racemase from human liver. Eur J Biochem 231: 815-822.

Schmitz W, Fingerhut R & Conzelmann E (1994) Purification and properties of an α-methylacyl-CoA racemase from rat liver. Eur J Biochem 222: 313-323.

78

Schmitz W, Helander HM, Hiltunen JK & Conzelmann E (1997) Molecular cloning of cDNA species for rat and mouse liver α-methylacyl-CoA racemases. Biochem J 326: 883-889.

Scholten JD, Chang KH, Babbitt PC, Charest H, Sylvestre M & Dunaway-Mariano D (1991) Novel enzymatic hydrolytic dehalogenation of a chlorinated aromatic. Science 253: 182-185.

Seedorf U & Assmann G (1991) Cloning, expression, and nucleotide sequence of rat liver sterol carrier protein 2 cDNAs. J Biol Chem 266: 630-636.

Seedorf U, Brysch P, Engel T, Schrage K & Assmann G (1994) Sterol carrier protein X is peroxisomal 3-oxoacyl coenzyme A thiolase with intrinsic sterol carrier and lipid transfer activity. J Biol Chem 269: 21277-21283.

Sharma V, Suvarna K, Meganathan R & Hudspeth ME (1992) Menaquinone (vitamin K2) biosynthesis: nucleotide sequence and expression of the menB gene from Escherichia coli. J Bacteriol 174: 5057-5062.

Shefer S, Cheng FW, Batta AK, Dayal B, Tint GS, Salen G & Mosbach EH (1978) Stereospecific side chain hydroxylations in the biosynthesis of chenodeoxycholic acid. J Biol Chem 253: 6386-6392.

Shoukry K & Schulz H (1998) Significance of the reductase-dependent pathway for the β-oxidation of unsaturated fatty acids with odd-numbered double bonds. Mitochondrial metabolism of 2-trans-5-cis-octadienoyl-CoA. J Biol Chem 273: 6892-6899.

Shimomura Y, Murakami T, Fujitsuka N, Nakai N, Sato Y, Sugiyama S, Shimomura N, Irwin J, Hawes JW & Harris RA (1994) Purification and partial characterization of 3-hydroxyisobutyryl-coenzyme A hydrolase of rat liver. J Biol Chem 269: 14248-14253.

Singh H, Beckman K & Poulos A (1994) Peroxisomal β-oxidation of branched chain fatty acids in rat liver. Evidence that carnitine palmitoyltransferase I prevents transport of branched chain fatty acids into mitochondria. J Biol Chem 269: 9514-9520.

Singh H & Poulos A (1988) Distinct long chain and very long chain fatty acyl CoA synthetases in rat liver peroxisomes and microsomes. Arch Biochem Biophys 266: 486-495.

Smeland TE, Nada M, Cuebas D & Schulz H (1992) NADPH-dependent β-oxidation of unsaturated fatty acids with double bonds extending from odd-numbered carbon atoms. Proc Natl Acad Sci U S A 89: 6673-6677.

Steinberg SJ, Wang SJ, Kim DG, Mihalik SJ & Watkins PA (1999) Human very-long-chain acyl-CoA synthetase: cloning, topography, and relevance to branched-chain fatty acid metabolism. Biochem Biophys Res Commun 257: 615-621.

Stoffel W, Duker M & Hofmann K (1993) Molecular cloning and gene organization of the mouse mitochondrial 3,2-trans-enoyl-CoA isomerase. FEBS Lett 333: 119-122.

Stoffel W & Grol M (1978) Purification and properties of 3-cis-2-trans-enoyl-CoA isomerase (dodecenoyl-CoA ∆-isomerase) from rat liver mitochondria. Hoppe Seylers Z Physiol Chem 359: 1777-1782.

Studier FW (1991) Use of bacteriophage T7 lysozyme to improve an inducible T7 expression system. J Mol Biol 219: 37-44.

Swinkels BW, Gould SJ, Bodnar AG, Rachubinski RA & Subramani S (1991) A novel, cleavable peroxisomal targeting signal at the amino-terminus of the rat 3-ketoacyl-CoA thiolase. EMBO J 10: 3255-3262.

Taylor KL, Xiang H, Liu RQ, Yang G & Dunaway-Mariano D (1997) Investigation of substrate activation by 4-chlorobenzoyl-coenzyme A dehalogenase. Biochemistry 36: 1349-1361.

Terwilliger TC & Berendzen J (1999) Automated structure solution for MIR and MAD. Acta Cryst D55: 849-861.

Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F & Higgins D (1997) The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucl Acids Res 24: 4876-4882.

Trotter PJ (2001) The genetics of fatty acid metabolism in Saccharomyces cerevisiae. Annu Rev Nutr 21: 97-119.

Tserng KY & Jin SJ (1991) NADPH-dependent reductive metabolism of cis-5 unsaturated fatty acids. A revised pathway for the β-oxidation of oleic acid. J Biol Chem 266: 11614-11620.

79

Tserng KY, Jin SJ & Chen LS (1996) Reduction pathway of cis-5 unsaturated fatty acids in intact rat-liver and rat-heart mitochondria: assessment with stable-isotype-labelled substrates. Biochem J 313: 581-588.

Uchida Y, Izai K, Orii T & Hashimoto T (1992) Novel fatty acid β-oxidation enzymes in rat liver mitochondria. II. Purification and properties of enoyl-coenzyme A (CoA) hydratase/3-hydroxyacyl-CoA dehydrogenase/3-ketoacyl-CoA thiolase trifunctional protein. J Biol Chem 267: 1034-1041.

Uchida Y, Kondo N, Orii T & Hashimoto T (1996) Purification and properties of rat liver peroxisomal very-long-chain acyl-CoA synthetase. J Biochem (Tokyo) 119: 565-571.

Uchiyama A, Aoyama T, Kamijo K, Uchida Y, Kondo N, Orii T & Hashimoto T (1996) Molecular cloning of cDNA encoding rat very long-chain acyl-CoA synthetase. J Biol Chem 271: 30360-30365.

Vamecq J, de Hoffmann E, Van Hoof F (1985) The microsomal dicarboxylyl-CoA synthetase. Biochem J 230: 683-693.

Vanhove GF, Van Veldhoven PP, Fransen M, Denis S, Eyssen HJ, Wanders RJ & Mannaerts GP (1993) The CoA esters of 2-methyl-branched chain fatty acids and of the bile acid intermediates di- and trihydroxycoprostanic acids are oxidized by one single peroxisomal branched chain acyl-CoA oxidase in human liver and kidney. J Biol Chem 268: 10335-10344.

Vanhove G, Van Veldhoven PP, Vanhoutte F, Parmentier G, Eyssen HJ & Mannaerts GP (1991) Mitochondrial and peroxisomal β-oxidation of the branched chain fatty acid 2-methylpalmitate in rat liver. J Biol Chem 266: 24670-24675.

Van Veldhoven PP, Casteels M, Mannaerts GP & Baes M (2001) Further insights into peroxisomal lipid breakdown via α- and β-oxidation. Biochem Soc Trans 29: 292-298.

Van Veldhoven PP, Croes K, Asselberghs S, Herdewijn P & Mannaerts GP (1996) Peroxisomal β-oxidation of 2-methyl-branched acyl-CoA esters: stereospecific recognition of the 2S-methyl compounds by trihydroxycoprostanoyl-CoA oxidase and pristanoyl-CoA oxidase. FEBS Lett 388: 80-84.

Van Veldhoven PP, Croes K, Casteels M & Mannaerts GP (1997) 2-methylacyl racemase: a coupled assay based on the use of pristanoyl-CoA oxidase/peroxidase and reinvestigation of its subcellular distribution in rat and human liver. Biochim Biophys Acta 1347: 62-68.

Van Veldhoven PP & Mannaerts GP (1999) Role and organization of peroxisomal β-oxidation. Current views of fatty acid oxidation and ketogenesis. From organelles to point mutations. Adv Exp Med Biol 466: 261-272.

Van Veldhoven PP, Van Rompuy P, Vanhooren JC & Mannaerts GP (1994) Purification and further characterization of peroxisomal trihydroxycoprostanoyl-CoA oxidase from rat liver. Biochem J 304: 195-200.

Van Veldhoven PP, Vanhove G, Assselberghs S, Eyssen HJ & Mannaerts GP (1992) Substrate specificities of rat liver peroxisomal acyl-CoA oxidases: palmitoyl-CoA oxidase (inducible acyl-CoA oxidase), pristanoyl-CoA oxidase (non-inducible acyl-CoA oxidase), and trihydroxycoprostanoyl-CoA oxidase. J Biol Chem 267: 20065-20074.

Van Veldhoven PP, Vanhove G, Vanhoutte F, Dacremont G, Parmentier G, Eyssen HJ & Mannaerts GP (1991) Identification and purification of a peroxisomal branched chain fatty acyl-CoA oxidase. J Biol Chem 266: 24676-24683.

Verhoeven NM, Roe DS, Kok RM, Wanders RJ, Jakobs C & Roe CR (1998) Phytanic acid and pristanic acid are oxidized by sequential peroxisomal and mitochondrial reactions in cultured fibroblasts. J Lipid Res 39: 66-74.

Verhoeven NM, Schor DS, ten Brink HJ, Wanders RJ & Jakobs C (1997) Resolution of the phytanic acid α-oxidation pathway: identification of pristanal as product of the decarboxylation of 2-hydroxyphytanoyl-CoA. Biochem Biophys Res Commun 237: 33-36.

Voorn-Brouwer T, van der Leij I, Hemrika W, Distel B & Tabak HF (1993) Sequence of the PAS8 gene, the product of which is essential for biogenesis of peroxisomes in Saccharomyces cerevisiae. Biochim Biophys Acta 1216: 325-328.

Vriend G (1990) WHAT IF: a molecular modeling and drug design program. J Mol Graph 8: 52-56. Wach A, Brachat a, Pohlmann R & Philippsen P (1994) New heterologous modules for classical or

PCR-based gene disruptions in Saccharomyces cerevisiae. Yeast 10: 1793-1808.

80

Wanders RJ, Denis S, van Roermund CW, Jakobs C & ten Brink HJ (1992) Characteristics and subcellular localization of pristanoyl-CoA synthetase in rat liver. Biochim Biophys Acta 1125: 274-279.

Wanders RJ, Denis S, Wouters F, Wirtz KW & Seedorf U (1997) Sterol carrier protein X (SCPx) is a peroxisomal branched-chain β-ketothiolase specifically reacting with 3-oxo-pristanoyl-CoA: a new, unique role for SCPx in branched-chain fatty acid metabolism in peroxisomes. Biochem Biophys Res Commun 236: 565-569.

Wanders RJA, Vreken P, Ferdinandusse S, Jansen GA, Waterham HR, van Roermund CWT & Van Grunsven EG (2001) Peroxisomal fatty acid α- and β-oxidation in humans: enzymology, peroxisomal metabolite transporters and peroxisomal disesases. Biochem Soc Trans 29: 250-267.

Wang HY & Schulz H (1989) β-oxidation of polyunsaturated fatty acids with conjugated double bonds. Mitochondrial metabolism of octa-2,4,6-trienoic acid. Biochem J 264: 47-52.

Watkins PA, Howard AE, Gould SJ, Avigan J & Mihalik SJ (1996) Phytanic acid activation in rat liver peroxisomes is catalyzed by long-chain acyl-CoA synthetase. J Lipid Res 37: 2288-2295.

Webster LT Jr., Gerowin LD & Rakita L (1965) Purification and characteristics of a butyryl coenzyme A synthetase from bovine heart mitochondria. J Biol Chem 240: 29-33.

Willadsen P & Eggerer H (1975) Substrate stereochemistry of the enoyl-CoA hydratase reaction. Eur J Biochem 54: 247-252.

Wu WJ, Anderson VE, Raleigh DP & Tonge PJ (1997) Structure of hexadienoyl-CoA bound to enoyl-CoA hydratase determined by transferred nuclear Overhauser effect measurements: mechanistic predictions based on the X-ray structure of 4-(chlorobenzoyl)-CoA dehalogenase. Biochemistry 36: 2211-2220.

Xiang H, Luo L, Taylor KL & Dunaway-Mariano D (1999) Interchange of catalytic activity within the 2-enoyl-coenzyme A hydratase/isomerase superfamily based on a common active site template. Biochemistry 38: 7638-7652.

Yang G, Liang PH & Dunaway-Mariano D (1994b) Evidence for nucleophilic catalysis in the aromatic substitution reaction catalyzed by (4-Chlorobenzoyl)coenzyme A dehalogenase. Biochemistry 33: 8527-8531.

Yang G, Liu RQ, Taylor KL, Xiang H, Price J & Dunaway-Mariano D (1996) Identification of active site residues essential to 4-chlorobenzoyl-coenzyme A dehalogenase catalysis by chemical modification and site directed mutagenesis. Biochemistry 35: 10879-10885.

Yang SY, He XY, Styles J, Luo MJ, Schulz H & Elzinga M (1994a) Primary structure of the large subunit of trifunctional β-oxidation complex from pig heart mitochondria. Biohem Biophys Res Commun 198: 431-437.

Yang SY, Li JM, He XY, Cosloy SD & Schulz H (1988) Evidence that the fadB gene of the fadAB operon of Escherichia coli encodes 3-hydroxyacyl-coenzyme A (CoA) epimerase, ∆3-cis-∆2-trans-enoyl-CoA isomerase, and enoyl-CoA hydratase in addition to 3-hydroxyacyl-CoA dehydrogenase. J Bacteriol 170: 2543-2548.

Yang X, Purdue PE & Lazarow PB (2001) Eci1p uses a PTS1 to enter peroxisomes: either its own or that of a partner, Dci1p. Eur J Cell Biol 80: 126-138.

Yang XY, Schulz H, Elzinga M & Yang SY (1991) Nucleotide sequence of the promoter and fadB gene of the fadBA operon and primary structure of the multifunctional fatty acid oxidation protein from Escherichia coli. Biochemistry 30: 6788-6795.

Zeelen JP, Hiltunen JK, Ceska TA & Wierenga RK (1994) Crystallization Experiments with 2-Enoyl-CoA Hydratase, Using an Automated 'Fast-Screening' Crystallization Protocol. Acta Cryst D50: 443-447.

Zeelen JP, Pauptit RA, Wierenga RK, Kunau WH & Hiltunen JK (1992) Crystallization and preliminary X-ray diffraction studies of mitochondrial short-chain ∆3,∆2-enoyl-CoA isomerase from rat liver. J Mol Biol 224: 273-275.

Zhang D, Liang X, He XY, Alipui OD, Yang SY & Schulz H (2001) ∆3,5,∆2,4-dienoyl-CoA isomerase is a multifunctional isomerase. A structural and mechanistic study. J Biol Chem 276: 13622-13627.

81

Zolman BK, Monroe-Augustus M, Thompson B, Hawes JW, Krukenberg KA, Matsuda SP & Bartel B (2001) chy1, an Arabidopsis mutant with impaired β-oxidation, is defective in a peroxisomal β-hydroxyisobutyryl-CoA hydrolase. J Biol Chem 276: 31037-31046.