Discovery of Novel Fatty Acid Dioxygenases and Cytochromes P450

Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Pharmacy 176

Mechanisms of Oxylipin Biosynthesis in Pathogenic Fungi

INGA HOFFMANN

ACTA UNIVERSITATIS UPSALIENSIS ISSN 1651-6192 ISBN 978-91-554-8739-3 UPPSALA urn:nbn:se:uu:diva-206199 2013 Dissertation presented at Uppsala University to be publicly examined in B21, BMC, Husargatan 3, Uppsala, Friday, October 18, 2013 at 09:15 for the degree of Doctor of Philosophy (Faculty of Pharmacy). The examination will be conducted in English.

Abstract Hoffmann, I. 2013. Discovery of Novel Fatty Acid Dioxygenases and Cytochromes P450: Mechanisms of Oxylipin Biosynthesis in Pathogenic Fungi. Acta Universitatis Upsaliensis. Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Pharmacy 176. 67 pp. Uppsala. ISBN 978-91-554-8739-3.

Dioxygenase-cytochrome P450 (DOX-CYP) fusion enzymes are present in diverse human and plant pathogenic fungi. They oxygenate fatty acids to lipid mediators which have regulatory functions in fungal development and toxin production. These enzymes catalyze the formation of fatty acid hydroperoxides which are subsequently converted by the P450 activities or reduced to the corresponding alcohols. The N-terminal DOX domains show catalytic and structural homology to mammalian cyclooxygenases, which belong to the most thoroughly studied human enzymes. 7,8-Linoleate diol synthase (LDS) of the plant pathogenic fungus Gaeumannomyces graminis was the first characterized member of the DOX-CYP fusion enzyme family. It catalyzes the conversion of linoleic acid to 8R-hydroperoxylinoleic acid (HPODE) and subsequently to 7S,8S- dihydroxylinoleic acid by its DOX and P450 domains, respectively. By now, several enzymes with homology to 7,8-LDS have been identified in important fungi, e.g., psi factor-producing oxygenase (ppo)A, ppoB, and ppoC, of Aspergillus nidulans and A. fumigatus. By cloning and recombinant expression, ppoA of A. fumigatus was identified as 5,8-LDS. Partial expression of the 8R-DOX domains of 5,8-LDS of A. fumigatus and 7,8-LDS of G. graminis yielded active protein which demonstrates that the DOX activities of LDS are independent of their P450 domains. The latter domains were shown to contain a conserved motif with catalytically important amide residues. As judged by site-directed mutagenesis studies, 5,8- and 7,8-LDS seem to facilitate heterolytic cleavage of the oxygen-oxygen bond of 8R-HPODE by aid of a glutamine and an asparagine residue, respectively. Cloning and expression of putative DOX-CYP fusion proteins of A. terreus and Fusarium oxysporum led to the discovery of novel enzyme activities, e.g., linoleate 9S-DOX and two allene oxide synthases (AOS), specific for 9R- and 9S-HPODE, respectively. The fungal AOS are present in the P450 domains of two DOX-CYP fusion enzymes and show higher sequence homology to LDS than to plant AOS and constitute therefore a novel class of AOS. In summary, this thesis describes the discovery of novel fatty acid oxygenases of human and plant pathogenic fungi and the characterization of their reaction mechanisms.

Keywords: Fusion protein, Linoleate diol synthase, Allene oxide synthase, Cyclooxygenase, Oxygenase, HPLC, Mass spectrometry, Hydroperoxide isomerase, Aspergillus, Fusarium oxysporum

Inga Hoffmann, Uppsala University, Department of Pharmaceutical Biosciences, Box 591, SE-751 24 Uppsala, Sweden.

ISSN 1651-6192 ISBN 978-91-554-8739-3 urn:nbn:se:uu:diva-206199 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-206199) List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Hoffmann, I., Jernerén, F., Garscha, U., Oliw, E.H. (2011) Ex- pression of 5,8-LDS of Aspergillus fumigatus and its dioxygenase domain. A comparison with 7,8-LDS, 10-dioxygenase, and cyclooxygenase. Archives of Biochemistry and Biophysics, 506:216–22 II Hoffmann, I., Oliw, E.H. (2013) 7,8- And 5,8-linoleate diol synthases support the heterolytic scission of oxygen-oxygen bonds by different amide residues. Submitted manuscript III Hoffmann, I., Hamberg, M., Lindh, R., Oliw, E.H. (2012) Novel insights into cyclooxygenases, linoleate diol synthases, and lipoxygenases from deuterium kinetic isotope effects and oxidation of substrate analogs. Biochimica et Biophysica Acta – Molecular and Cell Biology of Lipids, 1821:1508-17 IV Jernerén, F., Hoffmann, I., Oliw, E.H. (2010) Linoleate 9R- dioxygenase and allene oxide synthase activities of Aspergillus terreus. Archives of Biochemistry and Biophysics, 495:67–73 V Hoffmann, I., Jernerén, F., Oliw, E.H. (2013) Expression of Fusion Proteins of Aspergillus terreus Reveals a Novel Allene Oxide Synthase. The Journal of Biological Chemistry, 288:11459-69 VI Hoffmann, I., Oliw, E.H. (2013) Discovery of a Linoleate 9S- Dioxygenase and an Allene Oxide Synthase in a Fusion Protein of Fusarium oxysporum. Submitted manuscript

Reprints were made with permission from Elsevier and the American Society for Biochemistry and Molecular Biology.

Additional Papers

• Jernerén, F., Garscha, U., Hoffmann, I., Hamberg, M., Oliw, E.H. (2010) Reaction mechanism of 5,8-linoleate diol synthase, 10R-dioxygenase, and 8,11-hydroperoxide isomerase of Asper- gillus clavatus. Biochimica et Biophysica Acta – Molecular and Cell Biology of Lipids, 1801:503-7 • Oliw, E.H., Jernerén, F., Hoffmann, I., Sahlin, M., Garscha, U. (2011) Manganese lipoxygenase oxidizes bis-allylic hydroperoxides and octadecenoic acids by different mechanisms. Bio- chimica et Biophysica Acta – Molecular and Cell Biology of Li- pids, 1811:138-47 • Oliw, E.H., Wennman, A., Hoffmann, I., Garscha, U., Hamberg, M., Jernerén, F. (2011) Stereoselective oxidation of regioisomeric octadecenoic acids by fatty acid dioxygenases. Journal of Lipid Research, 52:1995-2004

Contents

Introduction ...... 9 Oxylipins ...... 9 Mammalian oxylipins – Eicosanoids ...... 9 Plant oxylipins ...... 11 Fungal oxylipins ...... 12 Fatty acid oxygenases ...... 13 Fatty acid monooxygenases – Cytochrome P450 ...... 14 Fatty acid dioxygenases ...... 16 Pathogenic fungi ...... 22 Aims ...... 24 Comments on methodology ...... 25 Recombinant expression ...... 25 Expression in insect cells ...... 25 Expression in E. coli ...... 26 Analysis of protein expression ...... 26 Site-directed mutagenesis ...... 27 Enzyme activity assay ...... 27 Analysis of fatty acid metabolites ...... 27 High performance liquid chromatography ...... 27 Mass spectrometry ...... 28 Kinetic isotope effects ...... 29 Bioinformatics ...... 30 Results ...... 31 Dioxygenase activities of LDS (paper I) ...... 31 Hydroperoxide isomerase activities of LDS (paper II and V) ...... 32 Novel insights into the reaction mechanisms of COX, 8R-DOX, and LOX (paper III) ...... 34 Kinetic isotope effects of COX, 8R-DOX, 9-DOX, and LOX (paper III, V, and VI) ...... 37 Oxylipin profile of A. terreus (paper IV) ...... 38 Discovery of the first fungal AOS (paper V) ...... 39 Crucial amide residue for AOS activity of A. terreus (paper V) ...... 40 10-DOX and 5,8-LDS of A. terreus (paper V) ...... 40

Discovery of a 9S-DOX-AOS fusion enzyme of F. oxysporum (paper VI) ...... 41 Discussion ...... 42 LDS are fusion proteins ...... 42 DOX activities ...... 43 P450 activities ...... 44 Fusion enzymes with AOS activities ...... 47 Fungal 9-DOX activities ...... 47 AOS activities ...... 48 Fungal JA biosynthesis ...... 50 Conclusion ...... 52 Acknowledgements ...... 53 References ...... 55

Abbreviations

AA arachidonic acid AOS allene oxide synthase(s) ATEG Aspergillus terreus gene BLAST Basic local alignment search tool COX cyclooxygenase(s) CP chiral phase CYP cytochrome P450 DiHODE dihydroxyoctadecadienoic acid(s) (D-)KIE (deuterium) kinetic isotope effect DOX dioxygenase(s) DOX-CYP dioxygenase-cytochrome P450 EODE epoxyoctadecadienoic acid H(P)ODE hydro(pero-)xyoctadecadienoic acid(s) HPOTrE hydroperoxyoctadecatrienoic acid(s) HPETE hydroperoxyeicosatetraenoic acid HPLC high performance liquid chromatography JA jasmonic acid LA linoleic acid α-LA α-linolenic acid LDS linoleate diol synthase(s) LOX lipoxygenase(s) LT leukotriene MS mass spectrometry m/z mass-to-charge ratio NP normal phase NSAID non-steroidal anti-inflammatory drug OPDA oxophytodienoic acid P450 cytochrome(s) P450 PCR polymerase chain reaction PG prostaglandin PGHS prostaglandin H synthase(s) POX peroxidase ppo psi producing oxygenase psi precocious sexual inducer RP reversed phase VZPE vibrational zero-point energy

Introduction

Fatty acids are aliphatic monocarboxylic acids and are essential to all living organisms. They are the main components of biomembranes which facilitate the separation of a cell or a compartment from its environment and are therefore as old as life itself [1]. In addition to their structural importance for biomembranes, fatty acids and their oxygenation products act as intra- and extracellular signaling molecules in vertebrates, fungi, plants, and corals. Fatty acid derivatives and related enzymes are also important drugs and targets for the treatment of a wide range of pathophysiological conditions such as inflammation, pain, cardiovascular diseases, and fever. Fatty acid oxygenating enzymes are also involved in the development and mycotoxin production of human and plant pathogenic fungi. Several of these enzymes belong to the dioxygenase-cytochrome P450 (DOX-CYP) fusion enzyme family, as its prototype, 7,8-linoleate diol synthase (LDS) of Gaeumannomyces graminis. Related enzymes are present in fungi of medical, biological, and biotechnological importance. This thesis contributes to this research area and focuses on the biosynthesis of oxygenated fatty acids in pathogenic fungi.

Oxylipins Oxygenated fatty acids can be formed enzymatically or by autoxidation [2- 6]. The term oxylipin comprises oxygenated fatty acids and in a narrower sense only if, in at least one step, a mono- or dioxygenase (DOX) was involved in their formation [7]. Oxylipins are signaling molecules and are widely distributed in the Animalia, Fungi, and Plantae kingdoms [3, 8-11].

Mammalian oxylipins – Eicosanoids

Oxylipins which are derived from unsaturated C20 fatty acids, e.g., eicosapentaenoic, dihomo-γ-linoleic, or arachidonic acid (AA), are collectively termed eicosanoids and constitute the largest and most important group of mammalian oxylipins [12]. The research field around eicosanoids dates back to the early 1930s when studies on seminal plasma from sheep and humans led to the discovery of a novel group of biologically active compounds [13]. They were termed

9 prostaglandins since they were primarily detected in the sheep prostate-like vesicular gland and were shown to have remarkable physiological effects on the circulation and on numerous organs [14, 15]. In the following decades, their structures were elucidated and underlying enzymes were identified [13]. It was shown that prostaglandins are formed from AA via the cyclooxygenase (COX) pathway [16, 17] and that aspirin which has been used as an analgesic drug for more than a hundred years acts via inhibition of prostaglandin formation [18]. In 1982, Sune K. Bergström, Bengt I. Samuelsson, and John R. Vane were awarded the Nobel Prize “for their discoveries concerning prostaglandins and related biologically active substances”. AA is by far the most important precursor of eicosanoids and is present in phospholipids in cell membranes. It is released upon cellular stimulation by phospholipase A2 [19]. AA is oxygenated via three major pathways, the COX, lipoxygenase (LOX), and cytochrome P450 (P450) pathway yielding prostanoids, leukotrienes, and several hydroxyl and/or epoxy derivatives (Fig. 1), respectively, which are potent lipophilic mediators [20]. The latter stimulate G-protein-coupled receptors and are involved in a variety of physiological and pathophysiological processes, e.g., inflammation, immune response, allergy, thrombosis, reproduction, fever, pain, and control of blood pressure [8, 21]. Prostanoids are derived from AA via the COX pathway and include prostaglandins and thromboxane which contain a five- and six-membered ring structure, respectively [22]. The first step in their formation is the conversion of AA to prostaglandin (PG)H2 by the action of PGH2 synthases (PGHS). PGH2 is a transient key intermediate which is transformed to PGD2, PGE2, and PGF2α by the action of specific synthases and to PGI2 (syn. prostacyclin) and thromboxane A2 by specific P450 [23-25]. Prostanoids are involved in the maintenance of physiological conditions e.g., blood pressure, renal function, bronchial tonus, secretion of gastric juice, and when formed in excess also in the mediation of pathophysiological processes as pain, fever, and the development of cancer [22]. The human LOX pathway comprises six functional genes [26] and catalyzes the conversion of AA to hydroperoxyeicosatetraenoic acids (HPETE) which are further converted to leukotrienes, lipoxins, and hepoxilins. Lipoxins are trihydroxyeicosanoic acids containing 4 conjugated double bonds [27, 28]. Interest in these compounds is mainly due to their anti- inflammatory properties [29]. Hepoxilins are hydroxy epoxy metabolites generated by the action of 12-LOX on AA and further rearrangement [30]. However, the most thoroughly studied LOX is 5-LOX which transforms AA to 5S-HPETE and subsequently to leukotriene (LT)A4. The latter acts as a precursor for the biosynthesis of other leukotrienes, namely cysteinyl LT and LTB4. As indicated by their names, leukotrienes are primarily formed in different kinds of leukocytes and contain three conjugated double bonds (“- triene”). They participate e.g., in inflammatory, asthmatic, and allergic

10 reactions. The 5-LOX inhibitor zileuton and the cysteinyl LT-1 receptor antagonists montelukast and zafirlukast mediate dilatation of the bronchial smooth muscle and are therefore used in the control of asthma [8, 31]. Conversion of AA by P450 comprises the third pathway of AA-derived eicosanoids yielding hydroxyeicosatetraenoic and epoxyeicosatrienoic acids [32, 33]. These biologically active metabolites are formed by the action of P450 hydroxylases and epoxygenases, mainly in the kidney and liver, but also in other organs [33-35]. These enzymes are associated with the regulation of the vascular tone, volume homeostasis, and the release of peptide hormones [36].

Arachidonic acid COOH O COX COOH P450 O HETE and EET LOX OH PGH 15-LOX 2 COOH prostaglandin synthases 5-LOX CYP8A 12-LOX OOH 15S-HPETE CYP5A PGE2 OOH PGD2 COOH OOH COOH PGF2α 5S-HPETE COOH O 12S-HPETE COOH O O O COOH LTB4 OH OH OH LTC LTD LTE PGI2 TXA2 LTA4 4 4 4 Figure 1. Overview of eicosanoids derived from AA by the COX, LOX, and P450 pathways.

Plant oxylipins Plant oxylipins comprise a diverse group of hydroperoxy-, hydroxy-, epoxy-, and oxo-fatty acids, as well as divinyl ethers, volatile aldehydes, the plant hormones 12-oxophytodienoic acid (OPDA) and jasmonic acid (JA) and related derivatives [3]. These compounds are mainly derived from linoleic (LA, 18:2n-6), α-linolenic (α-LA, 18:3n-3), and roughanic acid (16:3n-3). Several oxylipins were shown to be involved in growth and development and in defense reactions against pathogens [37, 38]. Biosynthesis of plant oxylipins is mainly initiated by LOX or α-dioxygenases (α-DOX) generating hydroperoxides which are often further converted by various P450, e.g., allene oxide synthase (AOS), hydroperoxide lyase, desaturase, or epoxy alcohol synthase [37, 39]. Oxygenated fatty acids can also be generated by non-enzymatic reactions. The free-radical catalyzed conversion of α-LA yields C18-isoprostanoids [40]. These prostaglandin-like compounds are also called phytoprostanes and are formed during oxidative stress by reactive oxygen species. It is assumed that phyto-

11 prostanes are involved in defense reactions as a response to oxidative stress [4]. Among plant oxylipins, JA and its derivatives have attracted special interest. JA is formed sequentially from α-LA by several enzymes [41, 42]. At first, 13S-LOX catalyzes the formation of 13S-hydroperoxyoctadecatrienoic acid (13S-HPOTrE), which is transformed by an AOS and allene oxide cyclase to the cyclic intermediate cis-12-OPDA (Fig. 2). Reduction of the latter by the OPDA reductase 3 followed by three steps of chain shortening via β–oxidation finally yields JA. The latter is involved in plant growth and development and mediates plant responses to the environment, e.g., plant- plant communication [43].

OOH 13 ( ) 13S-LOX ( ) 6 COOH 13 6 COOH α-LA 13S-HPOTrE Allene oxide O synthase Allene oxide O COOH cyclase ( ) 6 COOH 12-OPDA OPDA reductase 3 O 3x O β-oxidation COOH COOH 10,11-Dihydro-OPDA JA Figure 2. Biosynthetic conversion of α-LA to JA, compiled from [41, 44].

Fungal oxylipins

In fungi, C18 fatty acids comprise the most abundant fatty acids [45, 46] and they are the most common precursors of fungal oxylipins, in particular LA and α-LA. The first fungal LOX was identified in Fusarium oxysporum in 1976 [47] and was shown to catalyze the conversion of LA to a mixture of 9- and 13- hydroxyoctadecadienoic acids (HODE) [48]. Some years later, a fungal oxylipin was identified which was generated via a novel pathway unrelated to lipoxygenases. This oxylipin was isolated in 1986 from the basidomycete Laetisaria arvalis and was therefore called laetisaric acid [49]. It was shown to act as a fungicide on the root pathogen Phythium ultimum [49]. This compound was identified as 8R-HODE which is obtained by reduction of its precursor, 8R-hydroperoxyoctadecadienoic acid (HPODE) [50]. 8R-HODE was also found to be present in Aspergillus nidulans along with 5,8-DiHODE and a lacton derivative of 5,8-DiHODE [51-53]. These LA-derived oxylipins were reported to possess hormone-like activities by acting as precocious sexual inducers (psi) of sporulation since they affect the

12 ratio of sexual (ascospores) and asexual spores (conidia) [54-56]. By now, 8R-HODE, 5,8-DiHODE, and/or related metabolites, e.g., 7,8- and 8,11-Di- HODE, have also been identified in several other fungi, e.g., Aspergillus fumigatus [57], the rice pathogen Magnaporthe grisea [58], Agaricus bisporus [59], and Leptomitus lacteus [60]. The physiological function of fungal oxylipins is mostly studied in the genus Aspergillus [61, 62] but is generally less well understood compared to mammalian or plant oxylipins [10]. The mechanism of biosynthesis of 8R-HODE and its subsequent conversion to 7,8-DiHODE by G. graminis was elucidated by Oliw and coworkers in the 1990s [63, 64]. They established that 8R-HODE is generated via 8R- HPODE in a typical DOX reaction which is unrelated to lipoxygenases. Several fungi have been reported to produce prostaglandins [65, 66], e.g., the pathogenic yeasts Candida albicans and Cryptococcus neoformans [67]. However, it is not conclusively established that eicosanoids can be formed in fungi, especially since these reports are numerically few and because fungi only contain small amounts of C20 fatty acids which are required for prostaglandin biosynthesis. Additionally, the identification of genes and enzymes responsible for fungal prostaglandin synthesis is still missing. Several fungi have been shown to produce JA and JA derivatives, e.g., Gibberella fujikuroi [68], Lasiodiplodia theobromae [69], and F. oxysporum [70]. A study from Thakkar et al. suggests a putative function of JA in the fungus’ infection process of plants [71]. The complete mechanism of fungal biosynthesis is still unknown and responsible enzymes have not been identified. However, important insight into fungal JA formation of L. theobromae was provided by Tsukada et al. in 2010 [72]. By using isotope-labeled substrates they showed that JA is likely formed by a similar biosynthetic pathway as described in plants, which is indicated by the presence of the key intermediate 12-OPDA during plant JA biosynthesis. At the same time, they revealed that the stereoselectivity during cycloolefin reduction of 12-OPDA differs in plant and L. theobromae, indicating separate biosynthetic pathways for JA formation in plant and fungi. As far known, the first steps of fungal oxylipin biosynthesis are all catalyzed either by heme or non-heme dioxygenases. Their mechanisms of action are described below.

Fatty acid oxygenases Fatty acid oxygenases can be divided into mono- and dioxygenases which catalyze the regio- and stereospecific insertion of one or two atoms of molecular oxygen into a fatty acid, respectively. For many of these enzymes heme serves as prosthetic group which is essential to most living organisms. Heme consists of an iron ion coordinated to four nitrogen atoms of the planar organic ring structure protoporphyrin IX (Fig. 3) [73]. The two remaining

13 axial ligands of the iron have great impact on the reactivity of the heme group [74]. Examples for heme-containing enzymes are P450, peroxidases, and catalases [75].

CH H2C 3 CH2 H3C NN Fe N N H C 3 CH3

HOOC COOH Figure 3. Chemical structure of heme.

Fatty acid monooxygenases – Cytochrome P450 Most fatty acid monooxygenases belong to the P450 superfamily. P450 are present in virtually all organisms and catalyze diverse reactions in the degra- dation and biosynthesis of exo- and endogenous compounds [76, 77]. P450 are hemoproteins and were named due to their absorption of light at 450 nm in the reduced state upon binding of carbon monoxide. P450 catalyze the transfer of molecular oxygen to X-H bonds (X = C, N, or S) accompanied by reduction of the other oxygen atom to water according to the following general formula [34, 78]:

+ + RH + O2 + NADPH + H  ROH + H2O + NADP

In humans, there are 57 P450, some of which are essential for the biosynthesis of e.g., steroids, bile acids, vitamins, and prostaglandins as well as the metabolism of drugs and toxins [79]. Especially plants are rich in P450 as illustrated by the model of higher plants, Arabidopsis thaliana, which contains at least 286 P450 genes [80]. By now, there are thousands of identified P450 proteins. They are divided into families and subfamilies [81] which comprise enzymes with more than 40 and 55% amino acid identities and which have assigned numbers and capital letters, respectively. However, the sequence identities of P450 are generally low, often below 20%. Never- theless, P450 have a conserved fold, in particular in the core of the protein close to the heme. Conserved domains include (i) the heme-binding loop with the absolutely conserved Cys residue which serves as proximal heme ligand and is present in the Phe-Xaa-Xaa-Gly-Xaa-Arg-Xaa-Cys-Xaa-Gly motif, (ii) the Glu-Xaa-Xaa-Arg motif in the K-helix on the proximal side of the heme, and (iii) the Ala/Gly-Gly-Xaa-Asp/Glu-Thr-Thr/Ser motif in the I- helix on the distal side [80]. The latter contains the famous acid-alcohol pair which is present in over 80% of all hydroxylases and contributes to distal hydrogen-bonding networks and active site water placements required during catalysis [82, 83].

14 P450 are famous for being able to catalyze a great diversity of reactions [84] which is due to the variability in the domains involved in substrate access and binding [85]. The prototype P450 reaction is though hydroxylation which occurs as illustrated in Fig. 4. A water molecule is replaced when the substrate binds to the enzyme. Then, the ferric heme is reduced to the ferrous state by an electron supplied by NADPH via a P450 reductase before binding of molecular oxygen which yields a ferrous P450 dioxygen complex. After another one-electron reduction and protonation of the Fe(III)-hydroperoxy complex, (P450[FeIII-O-OH] is obtained. In the next step, protonation and heterolytic cleavage of the dioxygen bond with elimi- nation of H2O forms a highly reactive iron-oxo intermediate (P450[FeIV=O]•+), called compound I. The latter performs hydrogen abstraction from the substrate yielding P450[FeIV-OH], also referred to as compound II, and a substrate radical. By a hydrogen abstraction-oxygen rebound mechanism [86] the oxygenated product is obtained which dissociates from the heme which returns to its resting state [82, 83]. The electrons necessary for oxygen cleavage and substrate hydroxylation are received from different redox partners [78], in most cases from NAD(P)H as cofactor via a cytochrome P450 reductase. L(OH)(OH) H O LO 2 RH ROH FeIII H2O OH Resting state R. oxygen RH rebound LOOH III FeIV Fe e- Compound II CYP74 RH

H2O2 O RH RH . LDS II FeIV + H+ Fe Compound I Peroxide shunt OH O2 H O O O2 2 + RH RH H II FeIII Fe Compound 0 + - H e Figure 4. Catalytic cycle of cytochromes P450 with molecular oxygen or hydroperoxides as oxygen donors, compiled from [39, 83]. Hydroxylation, the prototype P450 reaction, requires all steps of the circular route. In the peroxide shunt pathway, the same transformation is catalyzed but neither molecular oxygen nor reducing equivalents are needed. CYP74 and LDS use hydroperoxides as substrates and cir- cumvent several steps of the circular route with formation of allene oxides and diols, respectively, as indicated by grey and grey dashed arrows, respectively. The hydroxylation and peroxide shunt pathways are in black. Abbreviations: RH, substrate; ROH, hydroxylated RH; LOOH designates 8R-HPODE and HPODE in the LDS and CYP74 pathways, respectively; LO, allene oxide; L(OH)(OH), DiHODE.

15 P450 can be divided into four classes according to the electron transfer system which provides the catalytic site with electrons [80]. Class I comprises bacterial and mitochondrial P450 which require a flavin adenine dinu- cleotide (FAD)-containing reductase and an iron-sulfur protein (ferredoxin). In class II enzymes, as microsomal P450, electrons are transported from NADPH via a single membrane-bound enzyme, the FAD- and flavin mono- nucleotide (FMN)-containing reductase, usually designated cytochrome P450 reductase. Class III P450 are self-sufficient and do not require an external electron source and class IV enzymes “receive electrons directly from NAD(P)H” [80]. Prokaryotic P450 are soluble whereas eukaryotic P450 are mostly membrane-bound. P450 class III comprises various enzymes involved in oxylipin formation, e.g., CYP8A1, CYP5, and CYP74. They do not use molecular oxygen but hydroperoxides as oxygen donor and they neither need NADPH nor an external P450 reductase for catalysis [87], as indicated in Fig. 4. A conserved Asn residue which is present in the I-helices of CYP74 and CYP8A1 is assumed to facilitate oxygen-oxygen bond cleavage [39, 88]. Hydroperoxide isomerase activities of LDS catalyze the conversion of 8R-HPODE to diols by suprafacial hydrogen abstraction and oxygen insertion, typical of P450. In 2008, LDS were indeed suggested to contain P450 domains [89], which was experimentally confirmed for ppoA (5,8-LDS) of A. nidulans [90]. The latter enzyme was found to be a fusion enzyme of an N-terminal DOX and a C-terminal P450 domain. The former domain catalyzes the conversion of LA to 8R-HPODE which is transformed to 5,8- DiHODE by the P450 domain [90].

Fatty acid dioxygenases

Prostaglandin H2 synthases PGHS are hemoproteins which comprise COX and hydroperoxidase (POX) activities. They are often only referred to as COX and perform the first step in the formation of prostaglandins [91]. Research on PGHS has been performed since the 1960s and 70s. PGHS was purified from bull seminal vesic- les [92] and was identified as the major enzyme in the oxidative conversion of AA via PGG2 to PGH2 [93-95]. PGHS are integral membrane proteins present on the luminal site of the endoplasmic reticulum and of the inner and outer membranes of the nuclear membrane [96]. They exist in two isoforms, PGHS-1 and PGHS-2, which are both active as homodimers [95]. Each monomer consists of three structural domains, an N-terminal epidermal growth factor-like domain, a membrane binding domain and a C-terminal catalytic domain; the latter comprises the COX and POX sites [97]. The COX activity performs the two oxygenation steps yielding PGG2, and the POX activity reduces PGG2 to

16 PGH2 and conducts activation of the COX site. This activation is accom- plished in the POX site by trace amounts of fatty acid hydroperoxides from the environment [2] which oxidize the ferric heme to an oxyferryl heme radical cation (compound I) (Fig. 5). Compound I accepts an electron from Tyr385 in the COX site leading to formation of a tyrosyl radical (Tyr385•) and is in turn converted to intermediate II (oxyferryl iron plus a neutral protoporphyrin) [91].

O IV .+ . Fe + Tyr-385 Tyr-385 + AA ROH Compound I AA ROOH + 2 O2 POX O COX III IV . . Fe + Tyr-385 Fe + Tyr-385 Tyr-385 + PGG2 Resting state Intermediate II - O e e- IV PGG Fe + Tyr-385 2 . Compound II Tyr-385 + PGG 2 Figure 5. Scheme of POX and COX reaction cycles, compiled from Smith [91].

The COX reaction of PGHS is initiated by abstraction of the pro-S hydrogen at the bis-allylic C13 of AA by Tyr385• (Fig. 6). After double bond migration, a free radical at C11 is yielded which is capable to react with unacti- vated molecular oxygen. Antarafacial oxygen insertion at C11 forms an 11R- hydroperoxyl radical which attacks C9. This leads to two cyclization steps with formation of an endoperoxide and a radical at C15. This radical reacts with molecular oxygen and PGG2 is obtained which is the last step of the COX cycle [91]. Subsequently, the additional electron from the Tyr385• is transferred to intermediate II which becomes compound II (Fig. 5). The latter accepts an electron and is thus recovered to its initial state (ferric heme). The reaction cycle restarts by reduction of PGG2 to PGH2 and formation of compound I in the POX site [91]. Hydrogen abstraction from the bis-allylic C13 occurs with a deuterium kinetic isotope effect (D-KIE) of ~ 2 [98].

17 . Tyr-385 9 COOH COOH H H 13 . 11 15 Arachidonic acid + O2

O COOH COOH O . . OO + O2 O COOH O COOH O O

OOH OH Prostaglandin G Prostaglandin H 2 2

Figure 6. Biosynthetic conversion of AA to PGH2.

The first crystal structure was reported from ovine PGHS-1 in 1994 [99]. Since then, 3D structures of both isoforms of PGHS have been resolved from several species with different substrates and inhibitors [96, 100-103]. Struc- tural and site-directed mutagenesis studies facilitated the identification of crucial amino acids for catalysis and substrate positioning [104]. In the POX site, His388 serves as proximal heme ligand and His207 is located at the distal site of the heme and both histidines are required for full activity [105, 106]. AA binds to the active site of the COX site in an “L-shaped” conformation and Arg120 and Ser530 facilitate that C13 is located close to Tyr385 which performs hydrogen abstraction [96]. Additionally, Val349 and Tyr348 and a couple of other residue are also involved in proper substrate positioning [96, 104]. The two isoforms of PGHS display 60-65% sequence identity when derived from the same species [97] but their 3D structures are almost identical. PGHS-1 is constitutively expressed and necessary to maintain physiological conditions, whereas the inducible PGHS-2 is mostly associated with pathophysiological conditions as pain, inflammation, fever, and the development of cancer [21, 97]. However, PGHS-1 is also associated with thrombosis [96, 107]. Both isoforms are used as drug targets [95]. Aspirin, the first synthetic analgesic drug, has been on the market for over a hundred years. It is still a common drug due to its analgesic, anti-inflammatory, antipyretic, and an- tithrombotic properties and belongs to the group of nonsteroidal anti-inflammatory drugs (NSAID). It was not until 1971 that Sir John Vane could show that the action of these drugs depends on inhibition of prostaglandin formation [18]. NSAID act by inhibiting AA from binding to the COX site of PGHS in a reversible (e.g., ibuprofen and flurbiprofen) or irreversible way (aspirin)

18 [96]. Aspirin inhibits PGHS by acetylation of Ser530 [108]. The isoforms of PGHS differ in their active sites. The COX site of PGHS-2 is more spacious (~20%) compared to PGHS-1 and can therefore bind larger molecules [96]. This knowledge facilitated the discovery of PGHS-2-selective drugs, the so called coxibes, which due to their size mainly bind to PGHS-2. They have been on the market since 1999.

Linoleate diol synthases LDS are fusion proteins of N-terminal DOX and C-terminal P450 domains and the latter account for hydroperoxide isomerase activities. The prototype of this enzyme class is 7,8-LDS of G. graminis. Native 7,8-LDS which had been purified to homogeneity and the recombinant enzyme were studied thoroughly [109-113]. 7,8-LDS is a bifunctional heme-dependent enzyme which catalyzes the conversion of LA to 8R-HPODE and its subsequent isomerization to 7,8-DiHODE [63, 64, 109]. In addition to the latter reaction, 8R-HPODE can be transformed to its corresponding alcohol, 8R-HODE, by chemical reduction [63]. The enzymatic reaction starts with oxidation of the heme which forms a tyrosyl radical. This radical abstracts the pro-S hydrogen at C8 yielding a fatty acid radical. After antarafacial insertion of molecular oxygen, 8R-HPODE is obtained. Abstraction of the pro-S hydrogen at C7 and suprafacial oxygen insertion yield the end product 7,8-DiHODE [111] (Fig. 7). The N-terminal DOX domain shows catalytic and sequence homology to PGHS [114, 115]. Both enzymes are heme-dependent and initiate the oxygenation by formation of a tyrosyl radical, which abstracts a hydrogen atom from the fatty acid substrate, and catalyze antarafacial oxygen insertion. Recombinant expression and site-directed mutagenesis of catalytically active amino acids allowed further insights into the similarities between 7,8-LDS and PGHS [112, 113]. In 2003, Cristea and coworkers reported a homologous gene to 7,8-LDS in the rice blast fungus M. oryzae which also converts LA to 8-HODE and 7,8-DiHODE [58]. Subsequently and facilitated by the release of the genomes of A. nidulans and A. fumigatus [116, 117], Tsitsigiannis and coworkers identified three homologous genes to 7,8-LDS in the genomes of A. nidulans and A. fumigatus, and designated them psi producing oxygenase (ppo) A, ppoB, and ppoC since the corresponding enzymes affected the biosynthesis of psi-factors and the development of the fungus [56]. By recombinant expression, ppoA of A. nidulans and ppoC of A. fumigatus and A. nidulans could be linked to formation of 8-H(P)ODE and 5,8-DiHODE and to 10R-H(P)ODE, respectively [90, 118, 119]. The responsible enzymes were hence designated 5,8-LDS and 10R-dioxygenase (10R-DOX), respectively. The function of the ppoB gene still remains unknown but a recent report shows that formation of 8,11-DiHODE is not related to ppoB [120].

19 The hydroperoxide isomerase activity of LDS was suggested to be a P450 by Lee et al. in 2008 [89]. This was confirmed one year later by Brodhun and coworkers. They showed by spectroscopic analysis of the heme-carbon monoxide complex that the C-terminal domain of 5,8-LDS of A. nidulans is a P450 [90]. This was the proof that LDS are DOX-CYP fusion enzymes. 10R-DOX are also DOX-CYP fusion enzymes [119] and oxidize LA to 10R-HPODE by hydrogen abstraction at C8 followed by double bond migration and antarafacial oxygen insertion at C10 [57]. These enzymes lack the cysteine residue which serves as fifth heme ligand and their P450 domains are therefore inactive. 10R-HPODE is not further converted except for the non-enzymatic reduction of the hydroperoxide to the alcohol. Additionally to their role in fungal development [55, 56, 62, 121], ppo genes and corresponding enzymes might be involved in the formation of secondary metabolites since the double and triple mutants ΔppoAΔppoB and ΔppoAΔppoBΔppoC of A. nidulans exhibited overproduction of penicillin and produced only traces of the mycotoxin sterigmatocystin [122].

HOOC C H ( ) 10 5 11 + O 6 2 R HS HR OOH 10 -HPODE ( ) C H DOX HOOC 5 8 5 11 Linoleic acid OOH + O2 ( ) C H HOOC 3 8 5 11 HR HR HS HS 8R-HPODE

P450 OH OH C H ( ) C H ( ) 5 11 8 5 11 HOOC 3 8 HOOC 5 OH 5S,8R-DiHODE OH 7S,8S-DiHODE Figure 7. Catalytic mechanism of LDS and 10R-DOX, compiled from [57, 63, 64, 118].

Alpha-dioxygenases In 1998, a plant dioxygenase with homology to PGHS-1 and -2 of vertebrates was discovered in tobacco leaves. It was shown to be upregulated in response to pathogen infection and was therefore named pathogen-induced oxygenase (PIOX) [123]. Subsequent studies on this oxygenase and a homologous enzyme of Arabidopsis revealed that they catalyze oxidation at the α- carbon (C2) of fatty acids by abstraction of the pro-R hydrogen at C2 and suprafacial insertion of molecular oxygen. These enzymes were therefore designated α-DOX. The resulting 2R-hydroperoxy fatty acid derivatives are unstable and undergo spontaneous decarboxylation yielding chain-shortened

20 aldehydes [124-126]. By now, α-DOX have been discovered in various plants, e.g., cucumber, rice, pea, the green alga Ulva pertusa, and the moss Physcomitrella patens. α-DOX are divided into two subgroups, α-DOX-1 and α-DOX-2, due to sequence alignment and phylogenetic studies and differences in preferred substrates [125]. The homology of α-DOX and PGHS was judged from sequence align- ments, site-directed mutagenesis studies, and most importantly from the recently published crystal structure of α-DOX of A. thaliana establishing that the heme ligands, the catalytic tyrosine residue, and the overall structure of the catalytic domain are conserved [123, 125, 127-129]. The biological functions of α-DOX are associated with plant development and defense against pathogen attack [125].

Lipoxygenases LOX are non-heme metallodioxygenases and are widely distributed in plants, fungi, and mammals [10, 37, 130, 131]. They catalyze the regio- and stereospecific insertion of molecular oxygen into fatty acids which contain at least one 1Z,4Z-pentadiene structure. LOX consist of a single polypeptide chain folded into two domains, an N-terminal β-barrel and a large mostly α-helical domain which contains the catalytic center with the catalytic metal [2, 132]. The most thoroughly studied LOX are human 5-LOX, which initi- ates leukotriene biosynthesis by formation of 5S-HPETE [133], and soybean-LOX-1 (sLOX-1) [130]. The latter was already crystallized in 1947 [134] and catalyzes the conversion of LA to 13S-HPODE. By now, crystal structures are available for various LOX, e.g., human 5-LOX [135], coral 8R-LOX [136], arachidonate 15-LOX [137], sLOX-1 [138], soybean LOX-3 [139], and porcine leukocyte 12-LOX [140]. In most LOX, iron serves as catalytic metal which is coordinated by conserved ligands; in the case of sLOX-1 by three histidine residues, the car- boxylic group of the C-terminal isoleucine, an asparagine residue, and a water molecule [138]. In the course of the reaction, the metal redox cycles 2+ 3+ - between inactive and reduced Fe -OH2 and oxidized and active Fe -OH [141]. The latter is the catalytic base which in a process of proton-coupled electron transfer abstracts the bis-allylic hydrogen from a 1Z,4Z- pentadiene structure of the fatty acid. This yields a carbon-centered radical which is delocalized over the pentadiene structure. Molecular oxygen is inserted in an antarafacial way and regio- and stereospecifically at C1 or C5 of the pentadiene and a peroxyl radical is formed. The latter is reduced by the catalytic metal and becomes a peroxyl anion which after protonation yields the end product, a cis-trans conjugated hydroperoxide. sLOX-1 performs the initial hydrogen abstraction with an exceptionally high D-KIE of ~80 which is among the largest D-KIE ever seen in enzymatic reactions [142]. 13R-LOX of G. graminis has attracted special interest and is the most studied fungal LOX [143-145]. This LOX and the recently identified

21 9S-LOX of Magnaporthe salvinii [146] are the only known LOX with manganese as catalytic metal. In contrast to FeLOX, 13R-MnLOX catalyzes suprafacial oxygen insertion and oxygenates LA to ~80% 13R-HPODE and ~20% to the uncommon bis-allylic hydroperoxide 11-HPODE. The latter can be isomerized to 13R-HPODE [141]. 13R-MnLOX is secreted from G. graminis and might contribute to the oxidative damage of root cells described below [130]. Compared to mammalian and plant LOX little is known about the biological function of fungal LOX. However, there are indications that fungal LOX might be involved in quorum-sensing mechanisms as well as development [147, 148].

Pathogenic fungi The enzyme activities studied in this thesis mainly derive from A. fumigatus, A. terreus, G. graminis, M. oryzae, and F. oxysporum. They belong to the Ascomycota (sac fungi) which is the largest phylum within the kingdom Fungi [149]. As indicated by their name, all ascomycetes contain an ascus which is a sac-like structure that contains the ascospores after meiosis. There are more than 30.000 ascomycetes and they account for 75% of all described fungi. Some of them are famous for being serious human and plant pathogens [150], producers of secondary metabolites [151], or hosts for recombinant expression of proteins [152]. Aspergillus is a genus of filamentous fungi and comprises about 250 species [153]. Many of these moulds are of medical, biological, and industrial importance which is illustrated by the fact that the genomes of several aspergilli are sequenced [154]. Several species are dreaded due to their pathogenicity while the same or others species are invaluable for food fermentation processes or industrial production of secondary metabolites, e.g., penicillin, ascorbinic, or citric acid [155, 156]. A. fumigatus is infamous for being a potent human pathogen [157]. Its spores are ubiquitous in the atmosphere and enter the human body via the respiratory tract. In immunocompetent individuals, the spores are mostly eliminated by the innate immune system. In contrast, immunocompromized individuals are vulnerable to infection leading in the worst case to life- threatening invasive aspergillosis, the main cause of death of immunocompromized patients. The effects of the mostly exclusively applied antifungal drug amphotericin B vary considerably among different patient groups but are generally poor, leading to a mortality rate of up to 90% [158, 159]. Another important Aspergillus is A. terreus. It produces several secondary metabolites of economic importance, e.g., the cholesterol-lowering drugs lovastatin and simvastatin as well as itaconic acid [160, 161]. The latter is used as a monomer in the industrial synthesis of e.g., polyesters and plastics

22 [162]. A. terreus is also clinically relevant since it causes infections ranging from superficial infections to invasive aspergillosis in severely immunocompromized patients [163]. Due to limited response to the fungicidal drug amphotericin B, infections with A. terreus are associated with higher mortality rates compared to other aspergilli [164]. G. graminis and M. oryzae belong to the family Magnaporthaceae and are serious plant pathogens. G. graminis is a soilborne and root-infecting fungus which causes root rot of cereals and grasses [165]. This disease is also referred to as take-all and starts in the roots and extends to the whole plant causing nutrient- and water-deficiency symptoms which lead to pre- mature death. Most vulnerable to infection are wheat and barley. Take-all is the most important root disease of wheat worldwide and is favored by mono- culture. No effective and economic fungicide is available for disease control [166]. M. oryzae, previously known as M. grisea [167], was judged to be the sci- entifically and economically most important fungal pathogen by a fungal pathologist community [168]. It causes the rice blast disease, which is the most destructive disease of rice worldwide with devastating consequences. The fungus destroys 10-30% of the annual rice harvest [169]. A loss of 10% corresponds to the amount of rice that could feed 60 million people. Today, more than 50% of the world population rely on rice and it is expected that it will increase by 2 billion people by 2050 [167]. It will therefore be strongly required to restrain the losses of the rice harvest. The genome of M. oryzae is published and the fungus serves as model to study host-fungal pathogen interactions [170]. Despite the effort and progress to understanding the rice blast disease, little of this knowledge has led to practical consequences for disease control [167]. Another devastating pathogen is F. oxysporum. Its species complex comprises a wide range of soilborne pathogens which collectively infect more than 100 different hosts and account for severe crop losses of economically important plants, e.g., tomato, cotton, banana [168, 171]. Fusarium species also infect humans and cause a broad spectrum of infections, ranging from allergic reactions, mycotoxicosis, and external infections to (locally) invasive diseases. The latter are mainly restricted to severely immunocompromized patients and are frequently fatal [172]. The genome of an F. oxysporum isolate was released by the Broad Institute in 2007 which facilitates research on the pathogenic mechanism [171, 173].

23 Aims

Several fungi are serious plant and human pathogens which cause devastating losses of staple foods and account for fatal infections of immunocompromized patients, respectively. It is vital to identify targets to restrain fungal pathogenicity. In order to reach this goal, genes and proteins of relevant pathogens need to be studied and linked to biological effects. The overall goal of the present study was to obtain further insight into the mechanisms of oxylipin biosynthesis in human and plant pathogenic fungi and to identify enzyme activities of unknown oxygenase-like genes in their genomes. Many of the underlying enzymes are DOX-CYP fusion proteins which have been successively discovered and characterized since the 1990s. Their N-terminal DOX domains share structural and catalytic similarities with COX and catalyze the formation of hydroperoxides. The latter are then converted by the C-terminal P450 domains to diols.

Specific aims of the present thesis were to:

• Clone and express ppoA of A. fumigatus. • Express and characterize the 8R-DOX domains of LDS. • Determine crucial amino acids for the 8R-DOX and hydroperoxide isomerase activities of LDS. • Obtain further insights into the mechanistic actions of COX, LOX, and 8R-DOX by studying deuterium kinetic isotope effects and the oxidation of substrate analogs. • Determine the oxylipin profile of A. terreus. • Express putative DOX-CYP fusion proteins of A. terreus and F. oxysporum.

24 Comments on methodology

Recombinant expression The precondition to all kinds of enzymatic studies is the availability of ade- quate amounts of enzyme. While natural organisms under physiological conditions mostly produce enzymes at low concentrations, recombinant expression systems are widely used because they provide higher amounts of enzyme and allow protein modifications which facilitate further analysis or purification. Recombinant expression systems comprise living cells and take advantage of the cellular machinery to produce protein from a DNA template of interest. Bacterial, mammalian, and other eukaryotic cells are used for this purpose. These cells differ in their properties regarding their efficiency and ability to perform posttranslational modifications. The latter are often required to retain activity of eukaryotic enzymes. Expression was performed in insect cells and Escherichia coli, repre- senting eukaryotic and prokaryotic expression systems, respectively. E. coli was used more often due to higher yields, less laboratory effort, and lower costs.

Expression in insect cells In the beginning of the study, all recombinant enzymes were obtained by plasmid-driven expression in Sf21or Sf9 cells [174]. These lepidopteran cells lines stem from ovarian tissue of the fall army worm Spodoptera frugiperda (Sf). Sf9 is a subclone derived from Sf21 and these cells grow faster and at higher density. Protein expression is under control of the OpIE2 promoter from the Orgya pseudotsugata multicapsid nucleopolyhedrosis virus [175]. Cells were grown in adherent cultures. This is a convenient way to maintain cells, but growing as a monolayer of course limits the cell density and thereby the protein yield. Nevertheless, the amount of active enzyme obtained by insect cell expression was sufficient for enzyme activity assays. The open reading frames of the genes of interest were cloned into pIZ/V5- His expression vectors with disruption of the native stop codons. The expression constructs prolonged the protein sequences with a C-terminal V5- tag allowing Western blot analysis. Transfection of the cell lines was performed by a plasmid-liposome complex. Two days post transfection, the cells were harvested, disrupted by sonication, and cell debris was removed

25 by centrifugation. The crude extract was either used immediately for enzyme activity assays or stored at -80 °C until needed.

Expression in E. coli E. coli is one of the most thoroughly studied organisms and enables rapid and inexpensive overexpression of a large variety of proteins. The starting point for this was the work performed by Studier and colleagues in the 1980’s [176]. They managed to insert the open reading frame of the RNA polymerase of the T7 bacteriophage into the genome of E. coli under control of the lac repressor. Expression of the T7 RNA polymerase can thus be controlled by lactose or, more commonly, by isopropyl β-D-1-thiogalactopyranoside, which is a non-metabolizable mimic of lactose. The DNA coding for the protein of interest is inserted in a plasmid which is introduced into E. coli and which contains the T7 promoter sequence. The latter is famous for its extremely high affinity for the T7 RNA polymerase facilitating high expression levels. A common consequence of forcing E. coli to overexpression of a certain protein is the formation of insoluble inclusion bodies which often contain incorrectly folded and therefore inactive protein. Another disadvantage of E. coli is its disability to perform posttranslational modifications which are often required for activity of eukaryotic proteins. Although the largest parts of the proteins were indeed present in inclusion bodies, the amounts of soluble and active enzyme were sufficient for the present studies. Open reading frames of the genes of interest were extended with a 5’CAAC overhang by polymerase chain reaction (PCR) technology which facilitated direct blunt-end cloning into pET101D-TOPO [177]. This vector provides extension of the expressed protein with a V5-tag enabling detection of the protein by Western blot analysis. The vector was inserted into E. coli BL21(DE3) cells by heat shocking. Overexpression was induced by 0.1 mM isopropyl β-D-1-thiogalactopyranoside and lasted normally for 5 h at room temperature under moderate shacking. Cells were harvested by sonication, lyzed, and sonicated. After removal of cell debris by centrifugation, the crude extract was ready for enzyme activity assays.

Analysis of protein expression Recombinant enzymes were denatured with sodium dodecyl sulfate and separated by gel electrophoresis on an 8% polyacrylamide gel. After separation, protein expression was studied by Western blot analysis. This technique takes advantage of the highly specific binding of an antibody with its antigen and allows qualitative, and to a certain degree quantitative, analysis of a protein of interest. The protein sequences of all recombinant enzymes

26 were fused to a C-terminal V5-epitope tag which served as antigen during Western blot analysis. The separated enzymes were blotted on a nitrocellu- lose membrane which was subsequently incubated with two different anti- bodies. The primary antibody specifically bound to the V5-epitop of the recombinant protein, whereas the secondary was an anti-mouse immuno- globulin G conjugated to a horseradish peroxidase and bound to the primary antibody. Incubation with a substrate for the horseradish peroxidase yielded a luminescing product which could be detected by a photographic film, if a V5-taged protein was present on the membrane in sufficient amounts.

Site-directed mutagenesis This technique allows the replacement of single amino acids in a protein sequence and the study of the impact of single residues on catalysis. Site- directed mutagenesis is achieved by incorporation of point mutations in the DNA sequence by PCR technology. Essentially, the QuickChange protocol by Stratagene was followed [178] with Pfu or Phusion DNA polymerases. The primers were complementary to the DNA strands except for the desired mutation which was introduced in all amplicons during the PCR program. Methylated maternal DNA was digested with the restriction enzyme DpnI. Mutated plasmid was transformed into chemically competent E. coli cells (NEB5α) by heat shocking and mutagenesis was confirmed by sequencing.

Enzyme activity assay 300-500 µl of crude extracts from lyzed insect cells or E. coli were incubated with 100 µM fatty acid on ice for 30 min. Addition of 2-4 volumes of methanol stopped the reaction due to precipitation of the proteins. Fatty acid metabolites were purified with the aid of an octadecyl silica cartridge and eluted with ethyl acetate. Evaporation to dryness under a stream of nitrogen and dilution in ethanol reduced the sample volume to 40 µl whereof 10 µl were subject to analysis by high performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS). In some cases, hydroperoxy metabolites were reduced to the corresponding alcohols by addition of triphenylphosphine.

Analysis of fatty acid metabolites High performance liquid chromatography Chromatography is the most commonly applied method for the separation of substances in complex mixtures. Separation is based on physical and/or

27 chemical properties of the sample components, e.g., hydrophobicity and steric configuration. A sample is dissolved in a mobile phase which passes a stationary phase. Separation of the components occurs due to different affini- ties to the stationary phase which yields different retention times. In this study, fatty acid metabolites were separated by HPLC implying that a liquid mobile phase passes at high pressure (100-200 bar) through a column which contains the stationary phase. Analyses were either performed with octadecyl silica, silica, or chiral columns, also known as reversed, normal, and chiral phase (RP- / NP- / CP-) HPLC, respectively. Most often, mono- and dihydroxy metabolites were separated by RP-HPLC in isocratic mode with methanol/water/acetic acid (750/250/0.05 or 800/200/0.05) at a flow rate of 0.3 ml/min. Acetic acid served as additive to increase chromato- graphic separation. When higher resolution was required, metabolites were separated by NP-HPLC on silica columns with hexane/isopropyl alcohol/acetic acid. Separation of stereo enantiomers was achieved by CP-HPLC with Reprosil Chiral NR, Reprosil Chiral AM, or Chiralcel OB-H columns which contain chiral stationary phases. Eluted metabolites were subject to analysis by mass spectrometry.

Mass spectrometry This powerful method has been instrumental for the analysis of eicosanoids for decades. A mass spectrometer basically consists of an ion source, an analyzer, and a detector. Before entering the analyzer, the molecules have to be ionized and converted to the gas phase. Collision with an inert gas makes the ions break into fragments which can be separated by their mass-to-charge (m/z) ratios under the influence of electric and magnetic forces. The system is put under low pressure or vacuum which assures that only intramolecular fragmentations occur. This makes the m/z ratios and the relative intensities of fragments unique for every molecule. In the present study, ions of fatty acid metabolites were generated by electrospray ionization and analyzed in negative ion mode. During this process, the sample passes a nozzle which, at its end, has a strong negative volt- age. Evaporation of the solvent leads to a decrease of droplet size and an increase of the electric field density on the surface of the droplet. This process proceeds until only gaseous ions remain. The latter enter the analyzer, a linear quadrupol ion trap. This instrument enables trapping, monitoring, and fragmentation of the ions before they are detected. Identification of fatty acid metabolites was performed by in-line analysis of HPLC coupled to mass spectrometry and was thus based on both retention times and m/z ratios. This was important, in particular if no baseline separation was achieved during HPLC. One has to keep in mind that analysis by liquid chromatography-MS/MS may be hindered due to the matrix effect [179], which results into altered ion intensities detected during MS analysis

28 due to the co-elution of matrix compounds to the MS interface. Especially for analysis of metabolites at very low concentrations, this effect has to be considered. NP- and CP-HPLC effluents were combined with a mixture of isopropyl/water to facilitate ionization prior to MS/MS analysis.

Kinetic isotope effects Isotopes are chemical elements which have the same number of protons and electrons but different numbers of neutrons, leading to different mass numbers. Isotopes of the same element are chemically indistinguishable but be- have differently in a dynamic sense which can be explained by the fact that heavy objects move more slowly than lighter ones [180]. Kinetic isotope effects (KIE) describe differences in dynamic behavior between isotopes [180] and facilitate the identification of rate-determining steps during a reaction and thereby the elucidation of its mechanism [181]. For that purpose often deuterium-KIE (D-KIE) are determined since the mass ratio of the hydrogen isotopes protium and deuterium is much larger than for any other element [182]. Isotopes differ in their vibrational zero-point energies (VZPE) which describe the energy at their ground state at absolute zero. The VZPE is directly correlated to the mass of a particle in inverse proportion. Due to their lower VZPE, heavier isotopes require higher energies for bond cleavage, which in turn requires higher activation energies and results in lower reaction rates [180]. The D-KIE can therefore be estimated by the ratio of the reaction rates (k) for the protonated (H) and deuterated (D) substrate: D-KIE = kH/kD The difference of activation energy required for breaking a C-H and C-D bond is ~1 kcal/mol which typically leads to a seven fold reduction of the reaction rate at room temperature [183]. Larger D-KIE cannot be explained by differences in VZPE alone and are a strong indication of hydrogen tunneling. This effect describes the possibility of hydrogen to “tunnel” through the energy barrier without reaching the transition state which is required according to the semi-classical transition-state-theory (Fig. 8) [182]. Analyses of D-KIE were performed by comparison of the reaction rates during oxygenation of selectively deuterated and non-deuterated substrates by oxygen consumption assay, ultraviolet-spectroscopy, or by HPLC- MS/MS.

29 I II III E E E

ΔE a ΔEa ΔEa H H D D H, D, T T T

RC RC RC GS TS GS TS GS TS Figure 8. Three models depicting the activation energies (ΔEa) needed to reach the transition state (TS), compiled from [182]. I, Classical model, no difference in ΔEa for protium (H), deuterium (D), and tritium (T). II, Semiclassical model which allows for different VZPE of H, D, and T, implying different ΔEa for the isotopes. III, Semiclassical model with tunneling correction. In addition to differences in VZPE, this model takes into account that lighter isotopes tunnel the energy barrier at a lower energy under the top of the barrier leading to lower ΔEa. E, energy; GS, ground state; RC, reaction coordinate.

Bioinformatics Bioinformatics, also known as computational biology, comprises computer- based methods for the analysis and maintenance of biological data and has captured an irreplaceable position in many kinds of life science research. The amount of data, e.g., from sequencing projects, 3D protein structures, and high throughput analyses, grows constantly and the data are archived in public accessible databases. Bioinformatic tools include e.g., methods for structural predictions of proteins and most importantly for analysis and alignment of DNA or protein sequences. The study of sequence similarities may reveal conserved patterns and is of fundamental importance e.g., for evolutionary studies, since similarities of protein sequences or structures often imply similar functions. In this study, nucleotide and protein sequences were obtained from Gen- Bank (http://www.ncbi.nlm.nih.gov). The Basic Local Alignment Search Tool (BLAST) algorithm [184], which is the most widely used tool in bioinformatics [185], was applied for the search of homologs to a given nucleotide or protein sequence. The ClustalW algorithm [186] allowed alignment of sequences. The SWISS-MODEL server [187] provided 3D models of given protein sequences by automated structure homology modeling. The construction of phylogenetic trees was performed with the help of the Mo- lecular Evolutionary Genetics Analysis (MEGA) 4 or 5 software [188]. Se- quencing was performed by the Uppsala Genome Center (Rudbeck Labora- tory, Uppsala, Sweden).

30 Results

Dioxygenase activities of LDS (paper I) PpoA of A. fumigatus was cloned from total RNA by reverse-transcription PCR and expressed in insect cells and E. coli. The recombinant enzyme transformed LA mainly to 8-H(P)ODE and 5,8-DiHODE and was therefore identified as 5,8-LDS. LDS belong to the DOX-CYP fusion enzyme family and contain an N- terminal DOX and a C-terminal P450 domain (Fig. 9) [89, 90]. The latter contains the heme-thiolate Cys residue in the conserved motif Phe-Gly-Xaa- Gly-Phe-His-Xaa-Cys-Xaa-Gly which is typical for all P450 [80]. Replace- ments of the Cys residues of 5,8-LDS (Cys1006) and 7,8-LDS (Cys1080) of G. graminis with Ser retained DOX but abolished hydroperoxide isomerase activities. The DOX domains share sequence homology with COX [114, 115] which catalyze substrate activation by hydrogen abstraction performed by a tyrosyl radical in the conserved sequence Y(R/H)WH which is also present in DOX (Fig. 9). The Tyr374Phe mutant of 5,8-LDS of A. fumigatus abolished 8R- DOX activity, which is in homology with the Tyr376Phe mutant of 7,8-LDS of G. graminis [112], but retained hydroperoxide isomerase activity. This demonstrates that DOX and P450 activities of LDS are independent of each other. As mentioned above, 8R-DOX shares structural and catalytic homology to COX and so does 10R-DOX. 8R-DOX and 10R-DOX both catalyze hydrogen abstraction at C8, but the positions for oxygen insertion differ. Val349 and Ser353 are crucial for COX catalysis [104] and replacements of the corresponding residues of 10R-DOX (Leu384 and Val388) with less bulky and bulkier residues, respectively, increased the relative amounts of 8- HPODE up to five times [118]. Therefore, the corresponding residues of 5,8- LDS of A. fumigatus, Val328 and Leu332, were substituted with a series of different amino acids. The results suggest that neither Val328 nor Leu332 are involved in the correct direction of molecular oxygen since the relative formation of 8-H(P)ODE was not altered. This implies that the 8R-DOX domains are controlled by different residues compared to 10R-DOX and COX-1. Alignment of the predicted and known α-helices of 5,8- and 7,8-LDS and COX-1, respectively, led to the identification of the DOX domains. Expres-

31 sion constructs of 674 and 673 N-terminal amino acids of 5,8- and 7,8-LDS, covering one additional α-helix compared to COX-1, yielded protein with unchanged or slightly lower 8R-DOX activity. This was judged by comparison of Michaelis menten constants of truncated and full-length constructs. In the latter, the catalytic Cys residues were substituted with Ser. Constructs lacking these extra helices were expressed but inactive. This shows that the 8R-DOX domains are independent of their P450 domains and confirms that LDS are fusion enzymes.

5,8-LDS YRWH ANXXQ FGFGPHQCLG Tyr374 Gln890 Cys1006

1 125 250 375 500 625 750 875 1000 1079 H2N- -COOH NCBI: Peroxidase superfamily P450 1 - 1079 Km ~ 12 μM 1 - 674 active Km ~ 14 μM 1 - 633 inactive

7,8-LDS YRWH ANXXQ FGLGPHRCAG Tyr376 Asn938 Cys1080

1 125 250 375 500 625 750 875 1000 1165 H2N- -COOH NCBI: Peroxidase1 - 1165 superfamily P450 1 - 1165 Km ~ 11 μM 1 - 673 active K ~ 36 μM 1 - 643 m inactive Figure 9. Overview of conserved regions and partial constructs of LDS. Catalyti- cally important residues which were subject to site-directed mutagenesis are in bold. Km, Michaelis menten constant; NCBI, National Center for Biotechnology Infor- mation.

Hydroperoxide isomerase activities of LDS (paper II and V) The P450 domains of LDS contain a conserved sequence, Ala-Asn-Xaa- Xaa-Gln, which might be located within the putative I-helix close to the heme (Fig. 9). This was judged from a 3D homology model of 7,8-LDS of G. graminis based on the crystal structure of CYP8A1(Fig. 10) [88]. The model suggests that Ala934 and Asn938 of 7,8-LDS align with Ala283 and Asn287 of the I-helix of CYP8A1. The latter residue is located close to the endoperoxide of PGH2 and is therefore likely involved in the scission of the O-O bond [88].

32 Asn287 Asn938 Gln941

Ala283 Ala934

Cys441 Cys1080 Figure 10. Alignment of the I-helix and Cys ligand loop of CYP8A1 with a 3D homology model of 7,8-LDS of G. graminis. The heme groups and catalytically important residues of CYP8A1 (grey) and the corresponding residues of 7,8-LDS (black) are depicted as sticks (CYP8A1: Ala283, Asn287, and Cys441; 7,8-LDS: Ala934, Asn938, and Cys1080), as well as Gln941 of 7,8-LDS.

Replacements of Asn938 of 7,8-LDS of G. graminis with Leu, Asp, and Glu residues abolished and reduced the activities, respectively, suggesting that Asn938 is of catalytic importance (Table 1). Replacements of Gln941 had minor effects. Substitutions of the homologous amide residues of 5,8-LDS (Asn887 and Gln890) of A. fumigatus yielded the opposite results. Formation of 5,8-Di- HODE was retained as main metabolite when Asn887 was substituted with Leu, but was strongly reduced by the Gln890Leu replacement. Substitution of Gln890 with Glu retained 5,8-DiHODE as main metabolite, but had major impact on the product profile. Oxygenation was partly shifted to C11 and the amount of non-enzymatically formed epoxy alcohols was increased due to homolytic scission of the O-O bond. This is in agreement with the site-directed mutagenesis studies on the corresponding amide residues (Asn878 and Gln881) of 5,8-LDS of A. terreus. The Asn878Leu mutant supported heterolytic cleavage of 8R-HPODE as the native enzyme, but shifted oxygenation from C5 to C6 and C7. In contrast, exchanging Gln881 with Leu yielded a more rigorous effect. This replacement virtually abolished the hydroperoxide isomerase activity. Gln881 was also substituted with Asn, Glu, Asp and Lys yielding three effects. First, increased non-enzymatic formation of epoxy alcohols by homolytic scission of 8R-HPODE (Gln881Asn, Gln881Lys, Gln881Asp). Second, the product pattern was altered with increased formation of 6,8-DiHODE (Gln881Asn) and 8,11-DiHODE (Gln881Asp, Gln881Glu, Gln881Lys). Third, enzymatic formation of epoxy alcohols by homolytic cleavage of 8R-HPODE (Gln881Glu). The main results suggest that LDS catalyze hydroperoxide isomerase activities by two different amide residues. 7,8- And 5,8-LDS might use Asn and Gln residues, respectively. These residues seem to be involved in the

33 heterolytic scission of the oxygen-oxygen bond of 8R-HPODE. This is the first description that LDS might differ in catalytically important amino acids.

Table 1. Summary of the effects of the replacements of amide residues of the Ala- Asn-Xaa-Xaa-Gln motif on the hydroperoxide isomerase activities of LDS of G. graminis and 5,8-LDS of A. fumigatus and A. terreus, respectively. The catalytic activities of the native and mutated enzymes are schematically graded from +++ to – for simplicity. G. graminis, gg; A. fumigatus, af; A. terreus, at. 7,8-LDSgg 5,8-LDSaf 5,8-LDSat Native +++ Native +++ Native +++ N938L - N887L ++ N878L ++ N938Q + 6,8-DiHODE 6,8-DiHODE N938D + 8,11-DiHODE 7,8-DiHODE

Q941L ++ Q890L + Q881L - 6,8-DiHODE Q881D/E ++ 8,11DiHODE 8,11-DiHODE Epoxy alcohols Q941E ++ Q890E ++ Q881N/K ++ 5,8-DiHODE 8,11-DiHODE Epoxy alcohols 8,11-DiHODE Epoxy alcohols Other diols

Novel insights into the reaction mechanisms of COX, 8R-DOX, and LOX (paper III) The oxidation mechanisms of ovine COX-1, 8R-DOX of 5,8- and 7,8-LDS of A. fumigatus and G. graminis, and two LOX (13R-MnLOX and sLOX-1) were studied with cis-trans isomers of LA, selectively deuterated substrates, and fatty acids which differed in chain length and number of double bonds. The initial oxidation rates during the conversion of C18 and C20 mono- and polyunsaturated fatty acids by COX-1 were compared. C20 and C18 fatty acids were oxygenated in the following order of rates: 20:2n-6 > 20:1n-6 > 20:3n-9 > 20:1n-9 and 18:3n-3 > 18:2n-6 > 18:1n-6, respectively, whereas 18:1n-9 was not oxidized. LA, (9E,12Z)18:2 and (9Z,12E)18:2 were all oxidized by COX-1 by hydrogen abstraction at C11 and antarafacial oxygen insertion. The former two substrates mainly supported oxygenation at C9 and the latter at C13 (Fig. 11). Formation of the 13S-stereoisomer from (9Z,12E):18:2 was unexpected and was shown to involve bond rotation at C13 which yields the 1Z,4Z-pentadiene radical. These results demonstrate that a cis double bond in the n-6 position facilitates substrate positioning.

34 8R-DOX of 5,8- and 7,8-LDS oxygenated (9Z,12E)18:2 mainly at C8, as expected. Incubations with (9E,12Z)18:2 and (9E)18:1 revealed differences in substrate specificities, as only 8R-DOX of 5,8-LDS accepted these fatty acids as substrates. Unexpectedly, (9E,12Z)18:2 was mainly converted to the 9R-hydroxy metabolite which requires hydrogen abstraction at C11. This is the first report that 8R-DOX can catalyze hydrogen abstraction from bis- allylic carbon atoms.

H HS HR H C H11C5 11 5

12 9 912 ( ) ( ) (9E,12Z)18:2 6 6 COOH COOH O COX-1 2 9S-HPODE HS HR H COOH H11C5 COOH ( )6 ( ) 12 9 13 9 6 Z E H11C5 (9 ,12 )18:2 H

O2 13S-HPODE

H HR HS R C5H11 8 -DOX C5H11

9 12 9 12 ( ) ( ) ( 6 6 R COOH COOH O2 9 -HPODE (9E,12Z)18:2 Figure 11. Overview of hydrogen abstraction and formation of major oxygenation products of cis-trans isomers of LA by ovine COX-1 and 8R-DOX. Top, COX-1 oxidizes (9E,12Z)18:2 mainly by hydrogen abstraction at C11 and antarafacial oxygen insertion at C9. Middle, (9Z,12E)18:2 is oxidized by COX-1 mainly to the 13S- hydroperoxy metabolite which requires bond rotation at C13. Bottom, 8R-DOX catalyzes hydrogen abstraction at the bis-allylic C11 of (9E,12Z)18:2 with formation of 9R-HPODE.

At pH 9, sLOX-1 abstracts the pro-S hydrogen at C11 of LA with formation of 13S-HPODE and when the carboxyl group is uncharged due to a lower pH, the substrate binds in opposite direction and forms 9S-HPODE (Fig. 12) [189]. The oxygenation of cis-trans isomers of LA at neutral pH has been studied by Funk et al. [190]. We reinvestigated these oxidations at pH 9. Oxidation of (9E,12Z)18:2 mainly yielded 9-HPODE (95%), almost in racemic mixture (R/S ratio of 60/40), and small amounts of the trans isomer of 13-HPODE. In contrast to the reaction at neutral pH, we could not detect 13- HODE with 9Z,11E configuration which requires bond rotation at C9 [190].

35 The almost racemic formation of 9-HODE suggests that the substrate binds in both orientations (Fig. 12). Oxidation of (9Z,12E)18:2 yielded 9-HPODE (65%) and 13-HPODE (35%; R/S ratio, 64/36), essentially as previously reported at neutral pH [190]. Since [11R-2H](9Z,12E)18:2 was not oxidized at a detectable rate, (9Z,12E)18:2 likely binds to the enzyme in reverse orientation compared to LA with abstraction of the pro-R hydrogen of C11. Formation of the R- stereoisomer of 13-HPODE is likely achieved by bond rotation at C13.

LA pH 9 pH 6.6 HS HR HR HS C H H C ) ( ( ) 5 11 11 5 7 7 - 912 COO HOOC 9 12

S S O2 13 -HPODE O2 9 -HPODE

(9E,12Z)18:2 pH 9 pH 9

HS HR HR HS H11C5 C5H11

12 9 COO- - 9 12 ( ) OOC ) ( 7 7

O S 2 9 -HPODE O2 9R-HPODE

Z E (9 ,12 )18:2 pH 9 H

( ) 7 H H - R S OOC 9 12 ( ) C H H 7 5 11 C H - ( ) 5 11 OOC 9 12 7 C H 5 11 -OOC 9 12 O 13S-HPODE O2 2 S 13R-HPODE O 9 -HPODE 2 Figure 12. Overview of hydrogen abstraction and oxygen insertion of LA and its cis-trans isomers by sLOX-1. Top, At pH 9, sLOX-1 catalyzes abstraction of the pro-S hydrogen of C11, followed by oxygen insertion and formation of 13S- HPODE. At a lower pH, the substrate binds in inverse orientation with formation of 9S-HPODE. Middle, Oxidation of (9E,12Z)18:2 yields 9-HPODE, almost in racemic mixture, which suggests that the substrate binds in both orientations. Bottom, sLOX-1 oxidizes (9Z,12E)18:2 to a mixture of 9- and 13-HPODE. Formation of 13R-HPODE can be explained by bond rotation at C13.

36 Kinetic isotope effects of COX, 8R-DOX, 9-DOX, and LOX (paper III, V, and VI) The D-KIE of sLOX-1 and COX-1 during oxygenation of AA differ by a factor of greater than 10. While LOX are limited to hydrogen abstraction from bis-allylic carbon atoms, COX-1 is also capable to abstract hydrogen from allylic carbon atoms, a process which requires more energy. Our hypothesis was that the D-KIE might be higher during energetically more demanding reactions since their reaction rates should depend to a greater extent on the contribution of hydrogen tunneling. We therefore performed the following studies. (i) Comparison of the D- KIE of COX-1 during abstraction of hydrogen atoms from allylic and bis- allylic carbon atoms. (ii) Estimation of the D-KIE of hydrogen abstraction from allylic and bis-allylic carbon atoms by 8R- and 9-DOX, respectively. (iii) Comparison of the D-KIE of sLOX-1 and 13R-MnLOX which have different catalytic metals, Fe and Mn, respectively. The oxidation rates of non- and selectively deuterated LA and 20:1n-6 by COX-1 were compared by oxygen consumption assays. Oxygenation of LA and 20:1n-6 represent hydrogen abstractions at C11 (bis-allylic) and C13 (allylic), respectively. The results suggested a D-KIE of ~3 for LA and ~25 for 20:1n-6 (Fig. 13). In homology to COX-1, 8R-DOX catalyzed allylic hydrogen abstraction at C8 with a high D-KIE (>20), as judged by compari- 2 son of the reaction rates of during oxidation of [8,8- H2]18:1n-9 and 18:1n-9. These results suggest that oxidations at allylic carbon atoms by COX-1 and 8R-DOX are accompanied by large D-KIE which can be explained by hydrogen tunneling. The D-KIE of the 9R- and 9S-DOX activities of A. terreus and recombinant FOXB_01332 were estimated to be < 5, respectively. This was judged 2 by comparison of the oxidation rates of [11,11- H2]18:2n-6 with 18:2n-6 or 13 [ C18]18:2n-6. This implies that the D-KIE during hydrogen abstraction from bis-allylic carbon atoms by fungal 9-DOX activities and COX-1 are in the same order of magnitude. Oxygenation of [11S-2H]18:2n-6 by 13R-MnLOX occurred with a D-KIE which is in the same magnitude as for sLOX-1 (~53). This suggests that the catalytic metals, Fe and Mn, respectively, do not affect the D-KIE.

37 bis-allylic allylic

HH HH R R’ R’ R COX-1 ~3 ~25 COX-2 ~5 ND 8R-DOX ND >20 9R-DOX <5 ND 9S-DOX <5 ND sLOX-1 ~53 - MnLOX ~53 - Figure 13. Summary of apparent D-KIE during hydrogen abstraction from bis-allylic and allylic carbon atoms by different dioxygenases.

Oxylipin profile of A. terreus (paper IV) A BLAST search revealed the existence of five DOX-CYP homologues in the genome of A. terreus. According to a phylogenetic tree, two of these genes were likely to encode known enzyme activities, viz. 10R-DOX and 5,8-LDS activities, since their deduced protein sequences clustered with enzymes with known activities from other aspergilli. Since three genes remained orphans, the potential to find novel oxylipins encouraged us to study the oxygenation of fatty acids by A. terreus. In addition to the expected monohydroxy metabolites, 8R- and 10R- HPODE, 9(R)-hydroperoxy-10(E),12(Z)-octadecadienoic acid (9R-HPODE) was identified (Fig. 14). The latter was shown to be formed by hydrogen abstraction at C11, followed by double bond migration and suprafacial oxygen insertion at C9. This was judged by incubations with [11S-2H]18:2n-6 with retention of the deuterium label suggesting abstraction of the pro-R hydrogen of C11. Separation of the stereoisomers by CP-HPLC and comparison of the retention times with the one for 9S-HPODE obtained from potato stolons indicated that the 9R stereoisomer was formed with high stereoselectivity suggesting enzymatic formation. Besides the expected 5S,8R-DiHODE metabolite, oxidation of LA by A. terreus also yielded high amounts of two other polar metabolites, 9-hydroxy-10-oxo-12(Z)-octadecenoic acid (α-ketol) and 13-hydroxy-10-oxo- 11(E)-octadecenoic acid (γ-ketol) (Fig. 14). The latter two are most likely hydrolysis products of the unstable allene oxide 9(R),10-epoxy-10,12(Z)- octadecadienoic acid. The AOS activity of A. terreus seems to be specific for 9R-HPODE as substrate since neither 9S-HPODE, 13R/S-HPODE, nor 13R- HPOTrE were transformed to significant amounts of α-ketols. Neither 9R-HPODE nor the allene oxide had been described before to be enzymatic products from any other Aspergillus. This publication is the first report that linoleate 9R-DOX and allene oxide activities can be expressed by fungi.

38 ( ) C H HOOC 5 8 5 11 910 Linoleic acid

OOH ( ) C H ( ) C H HOOC 9 5 11 HOOC 5 8 5 11 7 8R-HPODE R C H OOH 9 -HPODE HOOC ( ) 10 5 11 6 OOH 10R-HPODE

OH ( ) 10 C5H11 HOOC 7 9 ( ) 5 C5H11 HOOC 3 8 O 9R(10)-EODE OH 5S,8R-DiHODE non-enzymatic hydrolysis C H ( ) C5H11 ( ) 5 11 HOOC 7 HOOC 7 HO O α-Ketol γ-Ketol O OH Figure 14. Overview of the metabolites formed from LA by A. terreus.

Discovery of the first fungal AOS (paper V) As reported in paper IV, A. terreus possesses 9R-DOX activity linked to an AOS. The genome of A. terreus contains five genes (ATEG) with homology to DOX-CYP fusion proteins which might account for these activities. Cloning and expression of all of them led to the identification of the first fungal AOS. The gene coding for this activity was shown to be located within the P450 domain of the DOX-CYP fusion gene ATEG_02036. Incu- bations of the recombinant enzyme with 9R-HPODE yielded mainly 9-hydroxy-10-oxo-12(Z)-octadecenoic acid (α-ketol) and trace amounts of 13- hydroxy-10-oxo-11(E)-octadecenoic acid (γ-ketol) which are spontaneous hydrolysis products of the unstable allene oxide 9(R),10-epoxy-10,12(Z)- octadecadienoic acid. The fungal AOS belongs to the CYP6003 family and shows higher sequence identity to 5,8-LDS (~37%) than to plant AOS (CYP74) or CYP8A1 (<15%). The low sequence homology to CYP74 and the absence of the characteristic insert of nine amino acids close to the cysteinyl ligand which is typical of other CYP74 [39] demonstrates that the fungal AOS sequence is unique and that it must have evolved independently from CYP74. In summary, this publication describes the discovery of the first allene oxide synthase of fungal origin.

39 Crucial amide residue for AOS activity of A. terreus (paper V) As the P450 domains of LDS, the AOS of A. terreus contains the conserved sequence Ala-Asn-Xaa-Xaa-Gln described above. The importance of the latter amide residues (Asn964 and Gln967) was studied by site-directed mutagenesis. Asn964 of the AOS was shown to be crucial for catalysis since replacements with Val and Asp almost abolished and strongly reduced the activity, respectively. In addition to small amounts of the α-ketol, incubations of the Asn964Asp mutant with 9R-HPODE also yielded epoxy alcohols. This indicates that the homolytic cleavage of the hydroperoxide is retained but that the desaturation step is disturbed which favors the formation of epoxy alcohols. These results suggest that Asn964 is important but not absolutely required for catalysis. This is in accordance with other members of P450 class III, e.g., CYP8A1 and CYP74, which also use Asn residues for formation of hydrogen bonds with the dioxygen of PGH2 and hydroperoxides, respectively, and facilitate homolytic scission of the oxygen-oxygen bonds [39, 88]. Substitution of Gln967 with Leu did not seem to affect catalysis. This was unexpected since the homologous Gln residue is important for the hydroperoxide isomerase activity of 5,8-LDS and since the fungal AOS shows higher homology to 5,8-LDS than to other P450 class III, as described above. In brief, Asn964 seems to be essential for the AOS of A. terreus.

10-DOX and 5,8-LDS of A. terreus (paper V) Cloning and expression of ATEG_04755 yielded a protein which supports 10R-DOX activity. The recombinant protein transformed LA and 18:3n-6 mainly to 10R-HPODE and 10-HPOTrE, respectively. In analogy with 10R- DOX of A. fumigatus and A. nidulans [118, 119], ATEG_04755 yielded 8- hydroperoxyoctadecamonoenoic acid as main metabolite when incubated with oleic acid. Cloning of ATEG_03992 yielded two different cDNA sequences. One coded for the same 1071 amino acids which were predicted by conceptual translation (GenBank accession number EAU35794) and supported 5,8-LDS activity. The other cDNA sequence is presumably the result of alternative splicing. In the splice variant, the last intron is part of the open reading frame, leading to a frameshift after the first 963 N-terminal amino acids and the introduction of a stop codon after 1064 residues. This implies that the splice variant lacks the heme thiolate residue which is absolutely required for all P450 activities.

40 The splice variant of ATEG_03992 supports 8R-DOX activity as judged by incubations with LA, α-LA, and oleic acid, but as expected, it does not possess any hydroperoxide isomerase activity. This is the first indication that the oxylipin profile might be influenced by alternative splicing events.

Discovery of a 9S-DOX-AOS fusion enzyme of F. oxysporum (paper VI) A BLAST search on AOS of A. terreus (ATEG_02036, GenBank accession number AGH14485) revealed the existence of a homologous protein in F. oxysporum Fo5176 with predicted DOX and P450 domains in the N- and C-terminal parts, respectively. The deduced protein sequences of FOXB_01332 (GenBank accession number EGU88194) and ATEG_02036 share about 50% sequence identity. We purchased the open reading frame of FOXB_01332 and expressed the protein in E. coli to study whether it could convert LA. The DOX and P450 domains were both expressed as active proteins. FOXB_01332 converted LA to 9S-HPODE by abstraction of the pro-R hydrogen, followed by double bond migration and antarafacial insertion of oxygen at C9, as judged by incubations with [11S-2H]18:2n-6 and [11R- 2H]18:2n-6, respectively. The hydroperoxide was transformed to an unstable allene oxide, 9S(10)-epoxy-10,12(Z)-octadecadienoic acid, (9S(10)-EODE), as indicated by the presence of its hydrolysis products 9-hydroxy-10-oxo- 12(Z)-octadecamonoenoic acid (α–ketol) and 13-hydroxy-10-oxo-11(E)- octadecamonoenoic acid (γ–ketol). FOXB_01332 is therefore designated 9S- DOX-AOS. The recombinant enzyme did not transform 9R-HPODE, 13S-, or 13R-HPODE to allene oxides which implies that 9S-DOX-AOS is not involved in the biosynthesis of allene oxides as precursors of JA. 9S-DOX-AOS is homologous to LDS, 10R-DOX, and to AOS of A. terreus, e.g., as indicated by conserved histidine ligands to the heme and the catalytically tyrosyl residue. In contrast to the latter enzymes, 9S-DOX- AOS does not have the conserved sequence Ala-Asn-Xaa-Xaa-Gln in its P450 domain which comprises catalytically important amide residues for the scission of the dioxygen bonds of hydroperoxides. It is therefore likely that fungal AOS constitute at least two different subfamilies. In conclusion, we report the first linoleate 9S-DOX. This enzyme is fused to a P450 domain with AOS activity specific for 9S-HPODE. This is the first description that allene oxides can be formed by a heme-dependent DOX- CYP fusion protein and without the need of a lipoxygenase.

41 Discussion

Oxylipins have regulatory functions in fungal development and mycotoxin production [61, 122]. Some fungi are severe plant and human pathogens. G. graminis and M. oryzae cause loss of wheat and rice crops which could feed millions of people while infections with Aspergilli account for fatal infections of immunocompromized patients being a severe problem in the clinic. The main goal is to find targets to contain fungal pathogenicity. We are though still quite far away from this goal. Therefore, basic research needs to be conducted to identify potential targets and to study genes and proteins and to link them to their biological effect. This may be facilitated by studying the oxygenation mechanisms of oxylipins and their signaling pathways. Fortunately, much of the knowledge gained from the eicosanoid field is applicable to fungal oxylipin research. The interest in research on fungal dioxygenases has increased considerably with the discovery that these enzymes share structural and catalytic similarities with COX [113, 114], enzymes which belong to the most thoroughly studied mammalian oxygenases due to their physiological and pharmaco- logical importance. The study of fatty acid oxygenases also allows insights into their evolutionary relationships and reveals similarities in plant, fungal, and animal pathogen responses. Furthermore, it is assumed that fungi comprise a rich source of still unknown secondary metabolites of potential value. Due to extensive sequencing projects [116], bioinformatic tools, and molecular biology, COX-related enzymes were identified in several important fungi [56, 57, 191]. The first fungal enzyme of this group was 7,8- LDS of G. graminis [109, 114]. The release of genomes of several aspergilli facilitated the identification of three 7,8-LDS-like genes, ppoA, B, and C, in A. nidulans and A. fumigatus [56, 57]. The cloning and recombinant expression of ppoA of A. fumigatus was the starting point for the present study.

LDS are fusion proteins LDS are members of the DOX-CYP fusion enzyme family. Their N- and C- terminal domains belong to the peroxidase and the P450 superfamilies, respectively. The first domain catalyzes the formation of a tyrosyl radical which abstracts the pro-S hydrogen of C8 with formation of 8R-HPODE. The latter

42 is transformed by the hydroperoxide isomerase activities of the C-terminal domains to 5,8- or 7,8-DiHODE, respectively. Fusion proteins have evolved by joining of two genes coding for separate proteins, most likely to ensure efficient progression of linked reactions. One example for a fusion protein is CYP102A1 (P450 BM3) from the bacterium Bacillus megaterium which catalyzes NADPH-dependent hydroxylation of fatty acids. Its FAD- and FMN-containing reductase is fused to the C-terminal end of the P450 enzyme ensuring efficient electron transport and one of the highest turnover numbers reported for P450 [192]. Another example is PGHS with its COX and POX activities described in the introduction.

DOX activities The 8R-DOX domains of LDS can be aligned with COX and α-DOX around the YRWH motif with the catalytically important tyrosine and the distal histidine. The identification of residues which are in control of the DOX reaction is based on the similarity to COX. This similarity also facilitated the identification of the 8R-DOX domains of LDS. Partial constructs of 5,8- and 7,8-LDS of A. fumigatus and G. graminis, respectively, corresponding in length to COX, yielded inactive protein, whereas those which contained one additional α-helix, yielded prominent 8R-DOX activities. The fact that the DOX domains could be expressed as active proteins and that they are structurally independent of their P450 domains confirms that LDS are fusion enzymes. Replacement of Tyr374 of 5,8-LDS of A. fumigatus abolished the DOX activity, whereas the P450 activity remained unaffected. This indicates that Tyr374 may have a homologous function as Tyr385 of COX-1 which is oxidized to a radical by the heme prior to hydrogen abstraction of the substrate. This is in agreement with corresponding mutations performed on other LDS, e.g., 7,8-LDS of G. graminis [112], 5,8-LDS of A. nidulans [90], and α-DOX of the plant Oryza sativa [127]. Other similarities of 7,8-LDS of G. graminis and 10R-DOX of A. fumigatus with COX were thoroughly studied by Garscha and Oliw, mainly by site-directed mutagenesis [118]. By these studies, Leu384 and Val388 were shown to control the regiospecificity during oxygen insertion by 10R-DOX. Leu384 corresponds to Val349 of ovine COX-1 which is known to be involved in controlling the dioxygenation at C11/15 [104, 108]. The corresponding residues of 5,8-LDS of A. fumigatus to Leu384 and Val388 are Val328 and Leu332. However, replacements of the latter two residues by site-directed mutagenesis did not imply any alteration of the relative formations of 8R- or 10R-H(P)ODE. These replacements were performed with the Cys1006Ser mutant as template to prevent further isomerization of 8-HPODE to 5,8-DiHODE to facilitate the analysis. Corresponding mutants of 5,8-LDS of A. nidulans (Val328Ala and Val328Leu) had intact

43 isomerase activities and had, if any, only minor impact on the oxygenation position [119]. It remains unknown which residues control of the insertion of oxygen in 5,8-LDS.

P450 activities The first prediction that LDS have a C-terminal P450 domain was reported by Lee and coworkers and was based on bioinformatic analysis [89]. One year later, Brodhun and coworkers confirmed this prediction by expression and characterization of 5,8-LDS of A. nidulans which included the record of the Soret peak at 450 nm characteristic for P450 [90]. Earlier attempts to identify the P450 domains of LDS failed due to instability of the carbon monoxide complex. As expected, replacement of the heme ligand Cys1006 of 5,8-LDS of A. fumigatus with Ser abolished the hydroperoxide isomerase activities. This is in accordance with the corresponding mutations performed on 7,8-LDS of G. graminis and 5,8-LDS of A. nidulans [90]. 10R-DOX are also DOX-CYP fusion enzymes, but they lack isomerase activities partly since the cysteinyl heme ligand is replaced by a Ser. Replacement of this residue to Cys was not sufficient to obtain P450 activity of 10R-DOX of A. fumigatus in analogy with 10-DOX of A. nidulans [119]. Partial expression of three different P450 domains of 5,8-LDS of A. fumigatus, which varied in length, yielded, in contrast to the DOX domains, inactive enzyme. Nevertheless, co-incubations of Tyr374Phe and Cys1006Ser of A. fumigatus with LA yielded both 8R-H(P)ODE and 5S,8R- DiHODE, showing that the two domains are catalytically independent of each other, which supports that LDS are fusion enzymes. In comparison to the DOX domains, little is known about the regiospecificity of the P450 activities. Although it is challenging to obtain information on the P450 domains by homology only, we could identify a conserved and catalytically important sequence, Ala-Asn-Xaa-Xaa-Gln, which might be located within the I-helix, as judged by a homology model of 7,8-LDS based on CYP8A1. The Asn residue of this sequence can be aligned with the catalytic Asn287 of CYP8A1 which in turn has a homologous function as the Thr residue of the conserved acid/alcohol pair of P450 class I and II which facilitates proton delivery for the hydroxylation reaction. Replacements of the amide residues of the Ala-Asn-Xaa-Xaa-Gln sequence were shown to affect hydroperoxide isomerase activities of 5,8-LDS of A. terreus and A. fumigatus and 7,8-LDS of G. graminis either with in loss of activity, shift of the oxygenation position, and/or (non-)enzymatic formation of epoxy alcohols. These residues might therefore have a similar function as the acid/alcohol pair and Asn residues of P450 class I and II and class III, respectively.

44 The shift to the enzymatic formation of epoxy alcohols is of special interest since it implies a change from heterolytic to homolytic cleavage of the oxygen-oxygen bond of 8R-HPODE (Fig. 15). It would be of great interest to understand the mechanistic differences favoring homo- and heterolytic O-O bond scissions. It is discussed that the active site microenvironment is influenced by water molecules, the electron density at the porphyrin iron, and by hydrogen bonds which facilitate charge separation to a higher or lower extend which promotes homo- or heterolytic scission [83]. The possibility of the enzyme to stabilize a radical also affects the scission process. Most likely, 3D structures with high resolution are inevitable to solve this issue.

Linoleic acid HOOC C8H17 LDS ( ) 5 O- Diols O 8R-HPODE •+ - Fe(III) - - Fe(IV) - HOOC C8H17 HOOC C H ( ) Heterolytic 8 17 5 O scission ( )5 O O Homolytic OH - Fe(III) - - Fe(III) - HOOC C H ( ) 8 17 Epoxy alcohols 5 O• - Fe(III) - O• • + - Fe(IV) -

...ANXXQ... Figure 15. Overview of the conversion of 8R-HPODE to diols and epoxy alcohols by heterolytic and homolytic scission of the oxygen-oxygen bond, respectively. Amide residues of the Ala-Asn-Xaa-Xaa-Gln motif of LDS seem to affect the scission.

Site-directed mutagenesis was the method of choice to study the impact of single amino acid residues on catalysis and the product pattern. Expression of protein was confirmed by Western blot analysis when required. Since LDS have dual enzyme activities, the enzyme activity of the unaffected domain served as internal control for expression, implying that Western blot analysis was only indicated when replacements in the DOX domain yielded inactive enzyme. It is tempting to overestimate the results obtained from site-directed mutagenesis studies. Therefore, one has to keep in mind that this technique has limitations. To interpret the loss of activity only based on site-directed mutagenesis is problematic since it may be a result of unspecific structural changes, disruption of the protein structure, or due to the loss of heme. The binding of the substrate may be affected or it may be a combination of these effects which is visible as the final effect (loss of activity or an altered pro-

45 duct pattern). Without 3D information from crystallographic studies, results from site-directed mutagenesis studies have to be interpreted with caution. However, even if 3D information is available, one has to consider that enzymes may undergo conformational changes by induced-fit mechanisms upon substrate binding [193]. Another fact important to allow for is that the results on catalysis and the product profile from overexpressed enzymes outside from their cellular environment may not be fully applicable to living organisms. There are indications that the hydroperoxide isomerase reactions are influenced by posttranslational modifications, alternative splicing events, and gene regulation. Eukaryotic enzymes are often subject to posttranslational modification. A study from Jernerén and coworkers suggest that the different oxygenation profiles of M. oryzae and its naturally occurring mutant (Δmac1 sum1-99), which possesses constitutively protein kinase A activity, might be due to phosphorylation [194]. Another indication that phosphorylation may affect enzyme activity was provided by Gilbert and coworkers with their recent study on human 5-LOX [195]. They mimicked phosphorylation at Ser-663 by replacing this residue with an Asp. This mutant showed robust 15-LOX but only traces of 5-LOX activity. However, replacements of predicted phosphorylation sites of 7,8- LDS of M. oryzae with Ser and Glu residues did not affect the product profile (Hoffmann and Oliw, unpublished observation). It would be interesting to co-express 7,8-LDS of M. oryzae with protein kinase A and to study the effect on the product profile although one has to keep in mind that there might be other mechanisms than phosphorylation which may affect the enzyme activity and the oxygenation products of M. oryzae. We found that alternative splicing events may influence the product pattern of A. terreus. Cloning of 5,8-LDS of A. terreus yielded besides the expected full-length cDNA a splice variant which encodes a protein that lacks the cysteinyl ligand. The splice variant of ATEG_03992 supports 8R-DOX activity as judged by incubations with LA, α-LA, and oleic acid, but as expected, it does not possess any hydroperoxide isomerase activity. This is the first indication that the oxylipin profile might be influenced by alternative splicing events. The oxylipin profile may also be affected by gene regulation as indicated by the fact that the oxygenation products of A. fumigatus and A. nidulans vary with the growth conditions [57]. Taken together it becomes obvious that oxylipin biosynthesis depends on a multitude of factors, e.g., posttranslational modification, alternative splicing, and gene regulation. Further research is needed to comprehend the overall picture.

46 Fusion enzymes with AOS activities We discovered the first fungal AOS in the C-terminal domains of the two DOX-CYP fusion proteins ATEG_02036 of A. terreus and FOXB_01332 of F. oxysporum, respectively. Unlike other AOS, these enzymes show about 30% homology to LDS and comprise therefore a novel group of AOS. This suggests that the group of heme-dependent DOX-CYP fusion proteins not only comprises LDS but also DOX-AOS. Fusion proteins with AOS activities have also been reported from the coral Plexaura homomalla [196, 197] and the cyanobacterium Acaryochloris marina [198]. These enzymes contain N-terminal catalase domains which account for AOS activities and C-terminal LOX domains. The independency of the 8R-LOX and AOS activities of the coral fusion enzyme was proved by separate recombinant expressions [197].

Fungal 9-DOX activities It has been known for a long time that LA can be oxygenated at C9 by the action of 9S-LOX of plants [199]. Our studies show that 9-HPODE also can be formed by a similar mechanism by fungal 9-DOX activities. A. terreus and FOXB_01332 comprise linoleate 9R- and 9S-DOX activities, respectively. In both cases, 9-HPODE serves as substrate for the subsequent formation of allene oxides, as it is the case in L. theobromae, which possesses 9R-DOX and AOS activities [200]. 8R-DOX of 5,8-LDS of A. fumigatus can also form 9R-HPODE by hydrogen abstraction at C11 when incubated with (9E,12Z)18:2. Although the latter substrate is not naturally occurring it shows that hydrogen abstraction and oxygen insertion can be altered by small steric changes. Whereas both 9-DOX abstract the bis-allylic pro-R hydrogen of C11, the activities differ in their direction of oxygen insertion (Fig. 16). 9S-DOX catalyzes antarafacial and 9R-DOX suprafacial oxygen insertion. However, it is known that single amino acid substitutions can change regio- and stereo- specificities of LOX [201, 202]. It remains to be determined which residues account for the difference of orientation of oxygen insertion of 9-DOX. In analogy with lipoxygenases, 9R-DOX does not oxygenate oleic acid, presumably due to the lack of bis-allylic carbon atoms. In contrast, 9S-DOX is capable to perform the energetically more demanding abstraction of hydrogen atoms from allylic carbon atoms, as seen by oxygenation of 18:1n-6 and 18:1n-9 at C11. 9-DOX and lipoxygenases show a characteristic difference. Whereas the latter are famous for high contribution of hydrogen tunneling to the reaction rate, 9-DOX perform bis-allylic hydrogen abstractions with a more than ten- fold lower D-KIE (<5) which is in the range of typical DOX reactions and virtually excludes hydrogen tunneling.

47 Comparison of partial sequences of ATEG_02036, 9S-DOX, and 8R- DOX reveal an interesting difference in the conserved YR/HWH motif of 8R-DOX and COX which comprise the putative catalytic tyrosine and distal heme. The Trp residue of this sequence is replace by a Phe in ATEG_02036 and 9S-DOX and in other related (>50% sequence identity) orphan sequences. However, replacement of this Phe of FOXB_01332 with Trp did not affect the oxygenation position. Nevertheless, the YR/HWH motif seems to define different subfamilies of dioxygenases. It is remarkable that recombinant ATEG_02036 does not possess any DOX activity, in particular 9R-DOX activity, since it shares about 50% sequence identity with 9S-DOX-AOS. This can be due to several reasons, e.g., cloning or expression artifacts, incorrect folding, instability, unsuitable choice of expression system. Stability problems when expressed in E. coli have been reported for 10R-DOX of A. nidulans [119]. Although 9R-DOX could be unrelated to ATEG_02036, the low D-KIE observed during formation of 9R-HPODE indicates the action of a protein of the peroxidase family. Cloning and expression of putative DOX-CYP fusion proteins of A. terreus did not yield protein with 9-DOX activity. It will be the issue of further investigations to elucidate the origin of the 9R-DOX activity.

9-Dioxygenase Allene oxide synthase

H2N- YRFH CXG/A -COOH

P450-FeIII O ( ) C5H11 . ( ) C5H11 HOOC 7 9R-DOX / HOOC 7 ATEG_ R FOXB_01332 OOH 9R-HPODE 02036 O 9 (10)-EODE 9R-DOX HRHS ( ) C5H11 + O HOOC 7 2 Linoleic acid FOXB_ 01332 C H ( ) C5H11 ( ) 5 11 HOOC 7 HOOC 7 FOXB_ HOO 9S-HPODE 01332 O 9S(10)-EODE Figure 16. Reaction mechanisms of fungal 9-DOX and AOS activities and catalytic residues of corresponding enzymes. The 9R- and 9S-DOX activities of A. terreus and FOXB_01332, respectively, catalyze abstraction of the pro-R hydrogen of C11 with formation of 9R- and 9S-HPODE by supra- and antarafacial oxygen insertion, respectively. The hydroperoxides are converted by the AOS activities of ATEG_02036 and FOXB_01332 to the corresponding allene oxides. The enzyme responsible for the 9R-DOX activity is unknown.

AOS activities The discovery of fungal AOS revealed that there are now at least three subgroups of AOS, e.g., CYP74, fungal, and catalase-related AOS. Fungal AOS were found to be present in the C-terminal domains of two DOX-CYP fusion proteins, e.g., ATEG_02036 and FOXB_01332. The sequence identities of

48 these enzymes to LDS and CYP74 is ~30 and below 15%, respectively. Fun- gal AOS show several similarities but differ in catalytic residues as well as substrate specificities. ATEG_02036 and FOXB_01332 selectively transform 9R- and 9S-HPODE, respectively, to the unstable allene oxides, 9R- and 9S(10)-EODE, as judged by the accumulation of their hydrolysis products. FOXB_01332 also converts 11S-hydroperoxides of C20 fatty acids. Importantly, 13S- and 13R-hydroperoxides were not transformed. CYP74C from tomato and homogenate preparations from potato stolons can form allene oxides from 9S-HPODE and 9S-HPOTrE which can undergo cyclization [39, 203, 204] but without further transformation to JA. Fungal AOS discovered in this study are most likely not involved in the formation of allene oxides as precursors of JA. The sequences of CYP74 have several characteristic features around the cysteinyl ligand [89], as reviewed by Brash [39]. In CYP74, the Phe residue in the typical Phe-Xaa-Xaa-Gly motif is replaced by Trp and these enzymes have a conserved Val residue four residues after the Cys-Xaa-Gly motif which is captured by Ala or Gly in conventional P450. Even more remarkable is the insert of nine amino acids in the conserved heme binding sequence of CYP74. Except for the Phe to Trp replacement in the Phe-Xaa- Xaa-Gly motif, which is present in ATEG_02036, none of these features characteristic of CYP74 is conserved in fungal AOS. In the course of the reaction, 9-HPODE undergoes homolytic cleavage by the AOS, presumably as reported for plant AOS, with formation of P450 compound II (P450[FeIVOH]) and an intermediate 9S(10R)-epoxy-12(Z)- octadecenoic acid with a radical at C11. Proton and electron transfer from C11 to the metal center results in formation of the allene oxide. In CYP74 and CYP8A1, homolysis of the oxygen-oxygen bonds of the substrates is facilitated by Asn residues which point from the I-helix towards the heme [39, 88] and have corresponding functions to Thr residues in the sequence Gly-Xaa-(Glu/Asp)-(Ser/Thr) of P450 hydroxylases [39]. The importance of Asn321 of CYP74A for catalysis is illustrated by the fact that the Asn321Gln mutant only exhibits 5% activity compared to the wild type enzyme [89]. The catalase-related coral AOS was also reported to contain an important Asn residue. Replacements of Asn137 of the coral AOS with Ala, Gln, Ser, Asp and His strongly reduced the reaction rate to 0.8 – 4% compared to the wt [205]. A crucial Asn has also been identified in the AOS of A. terreus. Asn964 of ATEG_02036 was shown to be catalytically important since replacements with Val and Asp abolished and reduced the activity, respectively. This residue is present in the Ala-Asn-Xaa-Xaa-Gln motif (Fig. 17), conserved in LDS, and aligns with Asn938 crucial for the hydroperoxide isomerase activity of 7,8-LDS. Interestingly, FOXB_01332 does not contain this sequence. At the homologous position of the amide residues in the Ala-Asn- Xaa-Xaa-Gln motif of LDS and the AOS of A. terreus, FOXB_01332 has an

49 acidic (Glu946) and alcoholic (Ser949) residue which both seem to lack catalytic importance as judged by site-directed mutagenesis studies. There are other amide residues which might be located in the I-helix of FOXB_01332 (Fig. 17). One of them (Asn921) is present in the sequence Asn-Val-Asp-Gln which resembles the Ala-Asn-Xaa-Xaa-Gln motif. However, replacement of Asn921 with Val did not alter the AOS activity. Additional AOS sequences would facilitate sequence homology studies to identify crucial residues. It will be the issue of further investigations to determine catalytically important residues involved in the homolytic cleavage of 9S-HPODE by FOXB_01332. In conclusion, these results indicate that previously known AOS are distinct from fungal AOS and that the latter can be divided into two subgroups.

918 Q G N N V D Q V T D N L W L T A F G G I G V P V T A 9S-DOX-AOS 918 Q G S S I D H V T D N L W L T A F G G I G V P V T A EKJ79444 919 Q G N S A E Q V V D N M W L T A F G G I G A P V T A EFQ27323 Plant 879 Q G N S V D Q V V D N M W L T A F G G I G V P - - - CCF39565 pathogens 930 Q G N S P D K V T D N M W L T A F G G I G V P V T A ENH82400 1019 S G K T P A E V A D I S W M N A V G G V G A T I G V BAE60972 937 M D M T A A E T A E V C W L T A V G G V G A P V G L EHA25900 936 M D M T A A E T A E V C W L T A V G G V G A P V G L GAA91201 Aspergilli 936 M G M S A E E V A D I C W L T A I G G V G T P S G V AOSat

944 F Y E V L S F F L R P E N E A I W A E V Q A I A Q K 9S -DOX-AOS 944 F Y E V L A F F L R P E N A S I W A E V Q A I A Q K EKJ79444 945 F Y E V L E Y F L R R E N A S I W A E V Q T L A Q K EFQ27323 Plant 902 F Y E V M E F F L R R E N A S I W A E V Q T L A Q A CCF39565 pathogens 956 F Y E V M D F F L R R D N E A V W A E V Q Q L A Q A ENH82400 1045 F T D V L N Y F L Q D E N S H H W E E I Q K L A A S BAE60972 963 V A D V L Q Y Y L R P E N I D H W K R I Q N L V S Q EHA25900 962 I A D V L Q Y Y L R P E N I H H W K Q I Q N L V S Q GAA91201 Aspergilli 962 V A N V L Q Y Y F R Y E N I G H W E E I Q K L V T Q AOSat “...A N X X Q...” Figure 17. Partial alignment of the P450 domains of protein sequences of ascomycetes with at least 50% sequence identity to AOS of A. terreus (AOSat) and/or 9S- DOX-AOS of F. oxysporum. The proteins derive from plant pathogens, 9S-DOX- AOS, EGU88194, F. oxysporum; EKJ79444, F. pseudograminearum; EFQ27323, Colletotrichum graminicola; CCF39565, C. higgensianum; ENH82400, C. orbicu- lare, and Aspergilli, BAE60972, A. oryzae; EHA25900, A. niger; GAA91201, A. kawachii; AOSat, AGH14485, A. terreus. Amide residues are in bold. AOSat contains the ANXXQ motif which is conserved in LDS and comprises catalytically important amide residues for LDS and AOSat.

Fungal JA biosynthesis Plants convert α-LA by a series of enzymes, e.g., AOS, to JA [206]. In contrast, the biosynthetic pathway of JA in fungi is not fully elucidated. As outlined in the introduction, plant and fungal JA biosynthesis appear to occur by similar but not identical mechanisms. Our results show that fungi can produce allene oxides by heme-dependent DOX-CYP fusion proteins, with and without an active DOX domain. This

50 indicates that the first two steps of fungal JA biosynthesis might be performed by a putative 13S-DOX-AOS fusion enzyme without the need for a lipoxygenase. A BLAST search on ascomycetous proteins revealed proteins with over 50% sequence identity to ATEG_02036 and/or FOXB which might encode 13S-DOX-AOS or 13S-AOS activities (Fig. 17). Although CYP74-like genes could be identified with the help of bioinformatics, it might be challenging to identify fungal AOS by bioinformatics alone due to the following reasons. (i) The computer-based identification of CYP74-like proteins was mostly due to the characteristic insert of nine amino acids close to the heme [39, 89] which is lacking in fungal AOS. (ii) Only two sequences of fungal AOS are available at present containing relatively few conserved elements. (iii) The fungal AOS seem to belong to two different subgroups as discussed above. Another approach to find AOS involved in fungal JA biosynthesis might be to express the open reading frames of candidate genes, provided that sequence information is available.

51 Conclusion

The main conclusions of the present thesis can be summarized as follows:

• PpoA of A. fumigatus was identified as 5,8-LDS by expression. • The 8R-DOX activities of LDS are independent of their hydroperoxide isomerase activities. • 8R-DOX domains are controlled by different residues compared to 10R- DOX and COX-1. • The hydroperoxide isomerase domains of LDS contain a conserved sequence which encloses amide residues which seem to facilitate heterolytic scission of the oxygen-oxygen bond of 8R-HPODE. • Hydrogen abstraction at allylic carbon atoms by 8R-DOX and COX-1 are accompanied by large D-KIE suggesting hydrogen tunneling. • A. terreus possesses linoleate 9R-DOX and AOS activities. • The P450 domain of the DOX-CYP fusion enzyme ATEG_02036 of A. terreus encodes an AOS specific for 9R-HPODE. • FOXB_01332 of F. oxysporum was identified as 9S-DOX-AOS. • Fungal AOS are present in the P450 domains of DOX-CYP fusion proteins. They show low homology to plant AOS (CYP74) and belong therefore to a novel class of AOS.

In summary, novel enzyme activities of pathogenic fungi and their reaction mechanisms were identified and characterized.

52 Acknowledgements

The present work was carried out at the Department of Pharmaceutical Bio- sciences, Division of Biochemical Pharmacology, Faculty of Pharmacy, Uppsala University, Sweden and was supported by the Swedish Research Council, the Knut and Alice Wallenberg Foundation, and Uppsala Univer- sity. Travel grants for participating in scientific workshops and conferences were obtained from the Swedish Academy of Pharmaceutical Sciences and IF’s Foundation and are gratefully acknowledged.

I would like to express my sincere gratitude to:

Prof. Ernst H. Oliw for being a fantastic supervisor. Thank you very much for your excellent guidance and support. Your enthusiasm for research is inspiring.

My co-supervisors, Prof. Helena Danielson and Dr. Maria Norlin, for good advice.

My co-authors, Prof. Mats Hamberg, Prof. Roland Lindh, Dr. Ulrike Garscha, and Dr. Fredrik Jernerén, for fruitful collaboration.

Ulrike Garscha, co-writer, former colleague, and friend, for introducing me to the lab and for the nice times we spent outside work.

Fredrik Jernerén, co-writer and former colleague. Thank you very much for all good advice, help in the lab, and company on conferences.

Anneli Wennman, thank you for being a great friend and colleague. I really appreciate working with you and I am very thankful for your support and the great times we have at work, when attending conferences, and outside work.

Saeid Karkehabadi, colleague, for bringing new expertise, knowledge, and spirit to the lab.

Christine Wegler, thank you for your support and friendship. I really enjoy our lunch breaks and when we meet outside work.

53 All exchange, master, and project students who have been in the Oliw lab for enriching our group. Special thanks to Anton Brännlund, Karen Chao, Fanny Chaumont, Anna Knecht, Lukas Löfling, Fredrik Malmqvist, Erik Melander, and Eva Röttele for valuable contributions to my work.

Past and present PhD students at the Department of Pharmaceutical Biosciences, especially Helén Andersson, Jonathan Alvarsson, Erika Brolin, Loudin Daoura, Ida Emanuelsson, Linnea Granholm, Anna Jonsson, Shima Momeni, Mats Nilsson, Tomas Nilsson, and Sara Palm; thanks to all of you for your support during teaching and for creating a good working environment. Alfhild Grönbladh and Jenny Johansson, thank you for your friendship and support.

Magnus Jansson, thanks for help with computer issues. I would also like to thank past and present members of the administrative and technical staff, especially Agneta Bergström, Ulrica Bergström, Marianne Danersund, Agneta Hortlund, Birgit Jansson, Erica Johansson, Elisabeth Jonsson, Marina Rönngren, Johanna Svensson, and Kjell Åkerlund.

I am very thankful for having wonderful friends in Germany and Sweden who supported and encouraged me a lot during my PhD-time. Every one of you brightens and colors my life which I truly appreciate! Special thanks to my “Marburger” friends. You are fantastic and I am sure that our friendship will never end! All adorable friends I met via the Kalmar Nation Choir; thank you so much for all the laughter, music, singing, playing, celebrations, dinners, excur- sions, etc. You helped me a lot to feel at home in Uppsala.

A big thank you to the Brennan family: Margareta, Paul, Sofia with Tomas and Veronica with Tom for the warm welcome into your family and all the happy moments we spend together.

I am deeply grateful that I am blessed with a wonderful family. My parents Eva and Rainer and sister Pernilla with Simon and Emma; your love, support, and loyalty are invaluable. My beloved Pernilla, you are the best sister and friend I could ever wish for!

Michael, I am so happy and grateful for your love, support, and all the joy you bring into my life!

54 References

1. Weber, H. (2002). Fatty acid-derived signals in plants. Trends Plant Sci. 7, 217-224. 2. Schneider, C., Pratt, D.A., Porter, N.A., and Brash, A.R. (2007). Control of oxygenation in lipoxygenase and cyclooxygenase catalysis. Chem. Biol. 14, 473-488. 3. Mosblech, A., Feussner, I., and Heilmann, I. (2009). Oxylipins: structurally diverse metabolites from fatty acid oxidation. Plant Physiol. Biochem. 47, 511-517. 4. Mueller, M.J. (2004). Archetype signals in plants: the phytoprostanes. Curr. Opin. Plant Biol. 7, 441-448. 5. Porter, N.A., Mills, K.A., and Carter, R.L. (1994). A mechanistic study of oleate autoxidation: competing peroxyl H-atom abstraction and rearrangement. J. Am. Chem. Soc. 116, 6690-6696. 6. Porter, N.A., Caldwell, S.E., and Mills, K.A. (1995). Mechanisms of free radical oxidation of unsaturated lipids. Lipids 30, 277-290. 7. Gerwick, W.H., Moghaddam, M., and Hamberg, M. (1991). Oxylipin metabolism in the red alga Gracilariopsis lemaneiformis: mechanism of formation of vicinal dihydroxy fatty acids. Arch. Biochem. Biophys. 290, 436-444. 8. Funk, C.D. (2001). Prostaglandins and leukotrienes: advances in eicosanoid biology. Science 294, 1871-1875. 9. Lagarde, M., Chen, P., Vericel, E., and Guichardant, M. (2010). Fatty acid- derived lipid mediators and blood platelet aggregation. Prostaglandins Leukot. Essent. Fatty Acids 82, 227-230. 10. Brodhun, F., and Feussner, I. (2011). Oxylipins in fungi. FEBS J. 278, 1047-1063. 11. Song, W.C., and Brash, A.R. (1991). Investigation of the allene oxide pathway in the coral Plexaura homomalla: formation of novel ketols and isomers of prostaglandin A2 from 15-hydroxyeicosatetraenoic acid. Arch. Biochem. Biophys. 290, 427-435. 12. Smith, W.L. (1989). The eicosanoids and their biochemical mechanisms of action. Biochem. J. 259, 315-324. 13. Baker, R.R. (1990). The eicosanoids: a historical overview. Clin. Biochem. 23, 455-458. 14. von Euler, U.S. (1936). On the specific vaso-dilating and plain muscle stimulating substances from accessory genital glands in man and certain animals (prostaglandin and vesiglandin). J. Physiol. 88, 213-234. 15. Bergstrom, S., Carlson, L.A., and Weeks, J.R. (1968). The prostaglandins: a family of biologically active lipids. Pharmacol. Rev. 20, 1-48.

55 16. Hamberg, M., and Samuelsson, B. (1967). On the mechanism of the biosynthesis of prostaglandins E-1 and F-1-alpha. J. Biol. Chem. 242, 5336-5343. 17. Hamberg, M., Svensson, J., Wakabayashi, T., and Samuelsson, B. (1974). Isolation and structure of two prostaglandin endoperoxides that cause platelet aggregation. Proc. Natl. Acad. Sci. U S A 71, 345-349. 18. Vane, J.R. (1971). Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nat New Biol. 231, 232-235. 19. Gijon, M.A., and Leslie, C.C. (1999). Regulation of arachidonic acid release and cytosolic phospholipase A2 activation. J. Leukoc. Biol. 65, 330- 336. 20. Cook, J.A., Geisel, J., Halushka, P.V., and Reines, H.D. (1993). Prostaglandins, thromboxanes, leukotrienes, and cytochrome P-450 metabolites of arachidonic acid. New Horiz. 1, 60-69. 21. Smyth, E.M., Grosser, T., Wang, M., Yu, Y., and FitzGerald, G.A. (2009). Prostanoids in health and disease. J. Lipid Res. 50 Suppl, S423-428. 22. Smith, W.L., Urade, Y., and Jakobsson, P.J. (2011). Enzymes of the cyclooxygenase pathways of prostanoid biosynthesis. Chem. Rev. 111, 5821-5865. 23. Kurumbail, R.G., Kiefer, J.R., and Marnett, L.J. (2001). Cyclooxygenase enzymes: catalysis and inhibition. Curr. Opin. Struct. Biol. 11, 752-760. 24. Wu, K.K., and Liou, J.Y. (2005). Cellular and molecular biology of prostacyclin synthase. Biochem. Biophys. Res. Commun. 338, 45-52. 25. Wang, L.-H., and Kulmacz, R.J. (2002). Thromboxane synthase: structure and function of protein and gene. Prostaglandins Other Lipid Mediat. 68– 69, 409-422. 26. Funk, C.D., Chen, X.-S., Johnson, E.N., and Zhao, L. (2002). Lipoxygenase genes and their targeted disruption. Prostaglandins Other Lipid Mediat. 68– 69, 303-312. 27. Serhan, C.N. (2005). Lipoxins and aspirin-triggered 15-epi-lipoxins are the first lipid mediators of endogenous anti-inflammation and resolution. Prostaglandins Leukot. Essent. Fatty Acids 73, 141-162. 28. Serhan, C.N., Hamberg, M., and Samuelsson, B. (1984). Lipoxins: novel series of biologically active compounds formed from arachidonic acid in human leukocytes. Proc. Natl. Acad. Sci. U S A 81, 5335-5339. 29. Romano, M. (2010). Lipoxin and Aspirin-Triggered Lipoxins. TheScientificWorldJOURNAL 10, 1048-1064. 30. Pace-Asciak, C.R. (1994). Hepoxilins: a review on their cellular actions. Biochim. Biophys. Acta 1215, 1-8. 31. Drazen, J.M., Israel, E., and O'Byrne, P.M. (1999). Treatment of Asthma with Drugs Modifying the Leukotriene Pathway. N. Engl. J. Med. 340, 197- 206. 32. Capdevila, J.H., and Falck, J.R. (2001). The CYP P450 arachidonic acid monooxygenases: from cell signaling to blood pressure regulation. Biochem. Biophys. Res. Commun. 285, 571-576. 33. Kroetz, D.L., and Zeldin, D.C. (2002). Cytochrome P450 pathways of arachidonic acid metabolism. Curr. Opin. Lipidol. 13, 273-283.

56 34. Oliw, E.H. (1994). Oxygenation of polyunsaturated fatty acids by cytochrome P450 monooxygenases. Prog. Lipid Res. 33, 329-354. 35. Spector, A.A. (2009). Arachidonic acid cytochrome P450 epoxygenase pathway. J. Lipid Res. 50 Suppl, S52-56. 36. Capdevila, J.H., Falck, J.R., and Harris, R.C. (2000). Cytochrome P450 and arachidonic acid bioactivation. Molecular and functional properties of the arachidonate monooxygenase. J. Lipid Res. 41, 163-181. 37. Andreou, A., Brodhun, F., and Feussner, I. (2009). Biosynthesis of oxylipins in non-mammals. Prog. Lipid Res. 48, 148-170. 38. Blee, E. (2002). Impact of phyto-oxylipins in plant defense. Trends Plant Sci. 7, 315-322. 39. Brash, A.R. (2009). Mechanistic aspects of CYP74 allene oxide synthases and related cytochrome P450 enzymes. Phytochemistry 70, 1522-1531. 40. Thoma, I., Krischke, M., Loeffler, C., and Mueller, M.J. (2004). The isoprostanoid pathway in plants. Chem. Phys. Lipids 128, 135-148. 41. Wasternack, C. (2007). Jasmonates: an update on biosynthesis, signal transduction and action in plant stress response, growth and development. Ann. Bot. 100, 681-697. 42. Vick, B.A., and Zimmerman, D.C. (1984). Biosynthesis of jasmonic Acid by several plant species. Plant Physiol. 75, 458-461. 43. Wasternack, C., and Kombrink, E. (2010). Jasmonates: structural requirements for lipid-derived signals active in plant stress responses and development. ACS Chem. Biol. 5, 63-77. 44. Hamberg, M., and Gardner, H.W. (1992). Oxylipin pathway to jasmonates: biochemistry and biological significance. Biochim. Biophys. Acta 1165, 1- 18. 45. Stahl, P.D., and Klug, M.J. (1996). Characterization and differentiation of filamentous fungi based on Fatty Acid composition. Appl. Environ. Microbiol. 62, 4136-4146. 46. Nemec, T., Jernejc, K., and Cimerman, A. (1997). Sterols and fatty acids of different Aspergillus species. FEMS Microbiol. Lett. 149, 201-205. 47. Matsuda, Y., Satoh, T., Beppu, T., and Arima, K. (1976). Purification and properties of Co2+ requiring heme protein having lipoxygenase activity from Fusarium oxysporum. Agric. Biol. Chem. 40, 963-976. 48. Matsuda, Y., Beppu, T., and Arima, K. (1978). Crystallization and positional specificity of hydroperoxidation of Fusarium lipoxygenase. Biochim. Biophys. Acta 530, 439-450. 49. Bowers, W.S., Hoch, H.C., Evans, P.H., and Katayama, M. (1986). Thallophytic allelopathy: Isolation and identification of laetisaric acid. Science 232, 105-106. 50. Brodowsky, I.D., and Oliw, E.H. (1993). Biosynthesis of 8R- hydroperoxylinoleic acid by the fungus Laetisaria arvalis. Biochim. Biophys. Acta 1168, 68-72. 51. Champe, S.P., and el-Zayat, A.A. (1989). Isolation of a sexual sporulation hormone from Aspergillus nidulans. J. Bacteriol. 171, 3982-3988.

57 52. Mazur, P., Meyers, H.V., Nakanishi, K., El-Zayat, A.A.E., and Champe, S.P. (1990). Structural elucidation of sporogenic fatty acid metabolites from Aspergillus nidulans. Tetrahedron Lett. 31, 3837-3840. 53. Mazur, P., Nakanishi, K., El-Zayat, A.A.E., and Champe, S.P. (1991). Structure and synthesis of sporogenic psi factors from Aspergillus nidulans. J. Chem. Soc., Chem. Commun. 1486-1487. 54. Champe, S.P., Rao, P., and Chang, A. (1987). An endogenous inducer of sexual development in Aspergillus nidulans. J. Gen. Microbiol. 133, 1383- 1387. 55. Tsitsigiannis, D.I., Zarnowski, R., and Keller, N.P. (2004). The lipid body protein, PpoA, coordinates sexual and asexual sporulation in Aspergillus nidulans. J. Biol. Chem. 279, 11344-11353. 56. Tsitsigiannis, D.I., Kowieski, T.M., Zarnowski, R., and Keller, N.P. (2005). Three putative oxylipin biosynthetic genes integrate sexual and asexual development in Aspergillus nidulans. Microbiology 151, 1809-1821. 57. Garscha, U., Jernerén, F., Chung, D., Keller, N.P., Hamberg, M., and Oliw, E.H. (2007). Identification of dioxygenases required for Aspergillus development. Studies of products, stereochemistry, and the reaction mechanism. J. Biol. Chem. 282, 34707-34718. 58. Cristea, M., Osbourn, A.E., and Oliw, E.H. (2003). Linoleate diol synthase of the rice blast fungus Magnaporthe grisea. Lipids 38, 1275-1280. 59. Wadman, M.W., van Zadelhoff, G., Hamberg, M., Visser, T., Veldink, G.A., and Vliegenthart, J.F. (2005). Conversion of linoleic acid into novel oxylipins by the mushroom Agaricus bisporus. Lipids 40, 1163-1170. 60. Fox, S.R., Akpinar, A., Prabhune, A.A., Friend, J., and Ratledge, C. (2000). The biosynthesis of oxylipins of linoleic and arachidonic acids by the sewage fungus Leptomitus lacteus, including the identification of 8R- Hydroxy-9Z,12Z-octadecadienoic acid. Lipids 35, 23-30. 61. Tsitsigiannis, D.I., and Keller, N.P. (2007). Oxylipins as developmental and host-fungal communication signals. Trends Microbiol. 15, 109-118. 62. Tsitsigiannis, D.I., Bok, J.W., Andes, D., Nielsen, K.F., Frisvad, J.C., and Keller, N.P. (2005). Aspergillus cyclooxygenase-like enzymes are associated with prostaglandin production and virulence. Infect. Immun. 73, 4548-4559. 63. Brodowsky, I.D., Hamberg, M., and Oliw, E.H. (1992). A linoleic acid (8R)-dioxygenase and hydroperoxide isomerase of the fungus Gaeumannomyces graminis. Biosynthesis of (8R)-hydroxylinoleic acid and (7S,8S)-dihydroxylinoleic acid from (8R)-hydroperoxylinoleic acid. J. Biol. Chem. 267, 14738-14745. 64. Hamberg, M., Zhang, L.Y., Brodowsky, I.D., and Oliw, E.H. (1994). Sequential oxygenation of linoleic acid in the fungus Gaeumannomyces graminis: stereochemistry of dioxygenase and hydroperoxide isomerase reactions. Arch. Biochem. Biophys. 309, 77-80. 65. Noverr, M.C., Toews, G.B., and Huffnagle, G.B. (2002). Production of prostaglandins and leukotrienes by pathogenic fungi. Infect. Immun. 70, 400-402.

58 66. Erb-Downward, J.R., and Huffnagle, G.B. (2007). Cryptococcus neoformans produces authentic prostaglandin E2 without a cyclooxygenase. Eukaryot. Cell 6, 346-350. 67. Noverr, M.C., Phare, S.M., Toews, G.B., Coffey, M.J., and Huffnagle, G.B. (2001). Pathogenic yeasts Cryptococcus neoformans and Candida albicans produce immunomodulatory prostaglandins. Infect. Immun. 69, 2957-2963. 68. Cross, B.E., and Webster, G.R.B. (1970). New metabolites of Gibberella fujikuroi. J. Chem. Soc. C., 1838-1842. 69. Aldridge, D., Galt, S., Giles, D., and Turner, W. (1971). Metabolites of Lasiodiplodia theobromae. J. Chem. Soc. C., 1623 - 1627. 70. Miersch, O., Bohlmann, H., and Wasternack, C. (1999). Jasmonates and related compounds from Fusarium oxysporum. Phytochemistry 50, 517- 523. 71. Thakkar, V.R., Subramanian, R.B., and Kothari, I.L. (2004). Culture filtrate of Lasiodiplodia theobromae restricts the development of natural resistance in Brassica nigra plants. Indian J. Exp. Biol. 42, 111-114. 72. Tsukada, K., Takahashi, K., and Nabeta, K. (2010). Biosynthesis of jasmonic acid in a plant pathogenic fungus, Lasiodiplodia theobromae. Phytochemistry 71, 2019-2023. 73. Layer, G., Reichelt, J., Jahn, D., and Heinz, D.W. (2010). Structure and function of enzymes in heme biosynthesis. Protein Sci. 19, 1137-1161. 74. Rydberg, P., Sigfridsson, E., and Ryde, U. (2004). On the role of the axial ligand in heme proteins: a theoretical study. J. Biol. Inorg. Chem. 9, 203- 223. 75. Hersleth, H.-P., Ryde, U., Rydberg, P., Görbitz, C.H., and Andersson, K.K. (2006). Structures of the high-valent metal-ion haem–oxygen intermediates in peroxidases, oxygenases and catalases. J. Inorg. Biochem. 100, 460-476. 76. Guengerich, F.P. (2008). Cytochrome p450 and chemical toxicology. Chem. Res. Toxicol. 21, 70-83. 77. Meunier, B., de Visser, S.P., and Shaik, S. (2004). Mechanism of oxidation reactions catalyzed by cytochrome p450 enzymes. Chem. Rev. 104, 3947- 3980. 78. Hannemann, F., Bichet, A., Ewen, K.M., and Bernhardt, R. (2007). Cytochrome P450 systems--biological variations of electron transport chains. Biochim. Biophys. Acta 1770, 330-344. 79. Nelson, D.R. (2013). A world of cytochrome P450s. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 368, 20120430. 80. Werck-Reichhart, D., and Feyereisen, R. (2000). Cytochromes P450: a success story. Genome Biol. 1, reviews3003.1-9 81. Nelson, D.R. (2009). The cytochrome p450 homepage. Hum. Genomics 4, 59-65. 82. Denisov, I.G., Makris, T.M., Sligar, S.G., and Schlichting, I. (2005). Structure and chemistry of cytochrome P450. Chem. Rev. 105, 2253-2277. 83. Makris, T.M., Denisov, I., Schlichting, I., and Sligar, S.G. (2005). Activation of Molecular Oxygen by Cytochrome P450, in P.R. Ortiz de Montellano (Ed.) Cytochrome P450. Structure, Mechanism, and Biochemistry. Kluwer Academic / Plenum Publishers, New York, 149-182.

59 84. Guengerich, F.P., and Munro, A.W. (2013). Unusual cytochrome p450 enzymes and reactions. J. Biol. Chem. 288, 17065-17073. 85. Li, H., and Poulos, T.L. (1994). Structural variation in heme enzymes: a comparative analysis of peroxidase and P450 crystal structures. Structure 2, 461-464. 86. Hamdane, D., Zhang, H., and Hollenberg, P. (2008). Oxygen activation by cytochrome P450 monooxygenase. Photosynth. Res. 98, 657-666. 87. Marnett, L.J. (2008). Biochemistry: Divergence from the superfamily. Nature 455, 300-301. 88. Chiang, C.W., Yeh, H.C., Wang, L.H., and Chan, N.L. (2006). Crystal structure of the human prostacyclin synthase. J. Mol. Biol. 364, 266-274. 89. Lee, D.S., Nioche, P., Hamberg, M., and Raman, C.S. (2008). Structural insights into the evolutionary paths of oxylipin biosynthetic enzymes. Nature 455, 363-368. 90. Brodhun, F., Göbel, C., Hornung, E., and Feussner, I. (2009). Identification of PpoA from Aspergillus nidulans as a fusion protein of a fatty acid heme dioxygenase/peroxidase and a cytochrome P450. J. Biol. Chem. 284, 11792-11805. 91. Smith, W.L. (2008). Nutritionally essential fatty acids and biologically indispensable cyclooxygenases. Trends Biochem. Sci. 33, 27-37. 92. Miyamoto, T., Ogino, N., Yamamoto, S., and Hayaishi, O. (1976). Purification of prostaglandin endoperoxide synthetase from bovine vesicular gland microsomes. J. Biol. Chem. 251, 2629-2636. 93. Smith, W.L., and Lands, W.E. (1972). Oxygenation of polyunsaturated fatty acids during prostaglandin biosynthesis by sheep vesicular gland. Biochemistry 11, 3276-3285. 94. Hamberg, M., Svensson, J., and Samuelsson, B. (1974). Prostaglandin endoperoxides. A new concept concerning the mode of action and release of prostaglandins. Proc. Natl. Acad. Sci. U S A 71, 3824-3828. 95. Simmons, D.L., Botting, R.M., and Hla, T. (2004). Cyclooxygenase isozymes: the biology of prostaglandin synthesis and inhibition. Pharmacol. Rev. 56, 387-437. 96. Garavito, R.M., Malkowski, M.G., and DeWitt, D.L. (2002). The structures of prostaglandin endoperoxide H synthases-1 and -2. Prostaglandins Other Lipid Mediat. 68-69, 129-152. 97. Smith, W.L., DeWitt, D.L., and Garavito, R.M. (2000). Cyclooxygenases: structural, cellular, and molecular biology. Annu. Rev. Biochem. 69, 145- 182. 98. Wu, G., Lu, J.M., van der Donk, W.A., Kulmacz, R.J., and Tsai, A.L. (2011). Cyclooxygenase reaction mechanism of prostaglandin H synthase from deuterium kinetic isotope effects. J. Inorg. Biochem. 105, 382-390. 99. Picot, D., Loll, P.J., and Garavito, R.M. (1994). The X-ray crystal structure of the membrane protein prostaglandin H2 synthase-1. Nature 367, 243- 249. 100. Luong, C., Miller, A., Barnett, J., Chow, J., Ramesha, C., and Browner, M.F. (1996). Flexibility of the NSAID binding site in the structure of human cyclooxygenase-2. Nat. Struct. Biol. 3, 927-933.

60 101. Kurumbail, R.G., Stevens, A.M., Gierse, J.K., McDonald, J.J., Stegeman, R.A., Pak, J.Y., Gildehaus, D., Miyashiro, J.M., Penning, T.D., Seibert, K., et al. (1996). Structural basis for selective inhibition of cyclooxygenase-2 by anti-inflammatory agents. Nature 384, 644-648. 102. Malkowski, M.G., Ginell, S.L., Smith, W.L., and Garavito, R.M. (2000). The productive conformation of arachidonic acid bound to prostaglandin synthase. Science 289, 1933-1937. 103. Malkowski, M.G., Thuresson, E.D., Lakkides, K.M., Rieke, C.J., Micielli, R., Smith, W.L., and Garavito, R.M. (2001). Structure of eicosapentaenoic and linoleic acids in the cyclooxygenase site of prostaglandin endoperoxide H synthase-1. J. Biol. Chem. 276, 37547-37555. 104. Thuresson, E.D., Lakkides, K.M., Rieke, C.J., Sun, Y., Wingerd, B.A., Micielli, R., Mulichak, A.M., Malkowski, M.G., Garavito, R.M., and Smith, W.L. (2001). Prostaglandin endoperoxide H synthase-1: the functions of cyclooxygenase active site residues in the binding, positioning, and oxygenation of arachidonic acid. J. Biol. Chem. 276, 10347-10357. 105. Seibold, S.A., Cerda, J.F., Mulichak, A.M., Song, I., Garavito, R.M., Arakawa, T., Smith, W.L., and Babcock, G.T. (2000). Peroxidase activity in prostaglandin endoperoxide H synthase-1 occurs with a neutral histidine proximal heme ligand. Biochemistry 39, 6616-6624. 106. Landino, L.M., Crews, B.C., Gierse, J.K., Hauser, S.D., and Marnett, L.J. (1997). Mutational analysis of the role of the distal histidine and glutamine residues of prostaglandin-endoperoxide synthase-2 in peroxidase catalysis, hydroperoxide reduction, and cyclooxygenase activation. J. Biol. Chem. 272, 21565-21574. 107. Patrono, C. (1994). Aspirin as an antiplatelet drug. N. Engl. J. Med. 330, 1287-1294. 108. Schneider, C., Boeglin, W.E., Prusakiewicz, J.J., Rowlinson, S.W., Marnett, L.J., Samel, N., and Brash, A.R. (2002). Control of prostaglandin stereochemistry at the 15-carbon by cyclooxygenases-1 and -2. A critical role for serine 530 and valine 349. J. Biol. Chem. 277, 478-485. 109. Su, C., and Oliw, E.H. (1996). Purification and characterization of linoleate 8-dioxygenase from the fungus Gaeumannomyces graminis as a novel hemoprotein. J. Biol. Chem. 271, 14112-14118. 110. Oliw, E.H., and Su, C. (1997). Catalytic and spectroscopic properties of linoleate diol synthase of the fungus Gaumannomyces graminis. Adv. Exp. Med Biol 433, 65-68. 111. Su, C., Sahlin, M., and Oliw, E.H. (1998). A protein radical and ferryl intermediates are generated by linoleate diol synthase, a ferric hemeprotein with dioxygenase and hydroperoxide isomerase activities. J. Biol. Chem. 273, 20744-20751. 112. Garscha, U., and Oliw, E. (2008). Pichia expression and mutagenesis of 7,8-linoleate diol synthase change the dioxygenase and hydroperoxide isomerase. Biochem. Biophys. Res. Commun. 373, 579-583. 113. Garscha, U., and Oliw, E.H. (2008). Critical amino acids for the 8(R)- dioxygenase activity of linoleate diol synthase. A comparison with cyclooxygenases. FEBS Lett. 582, 3547-3551.

61 114. Hörnsten, L., Su, C., Osbourn, A.E., Garosi, P., Hellman, U., Wernstedt, C., and Oliw, E.H. (1999). Cloning of linoleate diol synthase reveals homology with prostaglandin H synthases. J. Biol. Chem. 274, 28219- 28224. 115. Oliw, E.H., Su, C., and Sahlin, M. (1999). Catalytic properties of linoleate diol synthase of the fungus Gaeumannomyces graminis: a comparison with PGH synthases. Adv. Exp. Med. Biol. 469, 679-685. 116. Galagan, J.E., Calvo, S.E., Cuomo, C., Ma, L.J., Wortman, J.R., Batzoglou, S., Lee, S.I., Basturkmen, M., Spevak, C.C., Clutterbuck, J., et al. (2005). Sequencing of Aspergillus nidulans and comparative analysis with A. fumigatus and A. oryzae. Nature 438, 1105-1115. 117. Nierman, W.C., Pain, A., Anderson, M.J., Wortman, J.R., Kim, H.S., Arroyo, J., Berriman, M., Abe, K., Archer, D.B., Bermejo, C., et al. (2005). Genomic sequence of the pathogenic and allergenic filamentous fungus Aspergillus fumigatus. Nature 438, 1151-1156. 118. Garscha, U., and Oliw, E.H. (2009). Leucine/valine residues direct oxygenation of linoleic acid by (10R)- and (8R)-dioxygenases: expression and site-directed mutagenesis of (10R)-dioxygenase with epoxyalcohol synthase activity. J. Biol. Chem. 284, 13755-13765. 119. Brodhun, F., Schneider, S., Göbel, C., Hornung, E., and Feussner, I. (2010). PpoC from Aspergillus nidulans is a fusion protein with one active heme. Biochem. J. 425, 553-565. 120. Jernerén, F., and Oliw, E.H. (2012). The fatty acid 8,11-diol synthase of Aspergillus fumigatus is inhibited by imidazole derivatives and unrelated to PpoB. Lipids 47, 707-717. 121. Tsitsigiannis, D.I., Kowieski, T.M., Zarnowski, R., and Keller, N.P. (2004). Endogenous lipogenic regulators of spore balance in Aspergillus nidulans. Eukaryot. Cell 3, 1398-1411. 122. Tsitsigiannis, D.I., and Keller, N.P. (2006). Oxylipins act as determinants of natural product biosynthesis and seed colonization in Aspergillus nidulans. Mol. Microbiol. 59, 882-892. 123. Sanz, A., Moreno, J.I., and Castresana, C. (1998). PIOX, a new pathogen- induced oxygenase with homology to animal cyclooxygenase. Plant Cell 10, 1523-1537. 124. Hamberg, M., Sanz, A., and Castresana, C. (2002). α-Dioxygenase, a new enzyme in fatty acid metabolism. International Congress Series 1233, 307- 317. 125. Hamberg, M., Ponce de Leon, I., Rodriguez, M.J., and Castresana, C. (2005). Alpha-dioxygenases. Biochem. Biophys. Res. Commun. 338, 169- 174. 126. Hamberg, M., Sanz, A., and Castresana, C. (1999). alpha-oxidation of fatty acids in higher plants. Identification of a pathogen-inducible oxygenase (piox) as an alpha-dioxygenase and biosynthesis of 2-hydroperoxylinolenic acid. J. Biol. Chem. 274, 24503-24513. 127. Koeduka, T., Matsui, K., Akakabe, Y., and Kajiwara, T. (2002). Catalytic properties of rice alpha-oxygenase. A comparison with mammalian prostaglandin H synthases. J. Biol. Chem. 277, 22648-22655.

62 128. Liu, W., Rogge, C.E., Bambai, B., Palmer, G., Tsai, A.L., and Kulmacz, R.J. (2004). Characterization of the heme environment in Arabidopsis thaliana fatty acid alpha-dioxygenase-1. J. Biol. Chem. 279, 29805-29815. 129. Goulah, C.C., Zhu, G., Koszelak-Rosenblum, M., and Malkowski, M.G. (2013). The crystal structure of alpha-Dioxygenase provides insight into diversity in the cyclooxygenase-peroxidase superfamily. Biochemistry 52, 1364-1372. 130. Oliw, E.H. (2002). Plant and fungal lipoxygenases. Prostaglandins Other Lipid Mediat. 68-69, 313-323. 131. Haeggström, J.Z., and Funk, C.D. (2011). Lipoxygenase and leukotriene pathways: biochemistry, biology, and roles in disease. Chem. Rev. 111, 5866-5898. 132. Brash, A.R. (1999). Lipoxygenases: occurrence, functions, catalysis, and acquisition of substrate. J. Biol. Chem. 274, 23679-23682. 133. Radmark, O., and Samuelsson, B. (2009). 5-Lipoxygenase: mechanisms of regulation. J. Lipid Res. 50 Suppl, S40-45. 134. Theorell, H., Holman, R.T., and Akeson, A. (1947). Crystalline lipoxidase. Acta Chem. Scand. 1, 571-576. 135. Gilbert, N.C., Bartlett, S.G., Waight, M.T., Neau, D.B., Boeglin, W.E., Brash, A.R., and Newcomer, M.E. (2011). The structure of human 5- lipoxygenase. Science. 331, 217-219. 136. Neau, D.B., Gilbert, N.C., Bartlett, S.G., Boeglin, W., Brash, A.R., and Newcomer, M.E. (2009). The 1.85 A structure of an 8R-lipoxygenase suggests a general model for lipoxygenase product specificity. Biochemistry 48, 7906-7915. 137. Boyington, J.C., Gaffney, B.J., and Amzel, L.M. (1993). The three- dimensional structure of an arachidonic acid 15-lipoxygenase. Science 260, 1482-1486. 138. Minor, W., Steczko, J., Stec, B., Otwinowski, Z., Bolin, J.T., Walter, R., and Axelrod, B. (1996). Crystal structure of soybean lipoxygenase L-1 at 1.4 A resolution. Biochemistry 35, 10687-10701. 139. Skrzypczak-Jankun, E., Amzel, L.M., Kroa, B.A., and Funk, M.O., Jr. (1997). Structure of soybean lipoxygenase L3 and a comparison with its L1 isoenzyme. Proteins 29, 15-31. 140. Xu, S., Mueser, T.C., Marnett, L.J., and Funk, M.O., Jr. (2012). Crystal structure of 12-lipoxygenase catalytic-domain-inhibitor complex identifies a substrate-binding channel for catalysis. Structure 20, 1490-1497. 141. Oliw, E.H., Jernerén, F., Hoffmann, I., Sahlin, M., and Garscha, U. (2011). Manganese lipoxygenase oxidizes bis-allylic hydroperoxides and octadecenoic acids by different mechanisms. Biochim. Biophys. Acta 1811, 138-147. 142. Knapp, M.J., and Klinman, J.P. (2002). Environmentally coupled hydrogen tunneling. Linking catalysis to dynamics. Eur. J. Biochem. 269, 3113-3121. 143. Hamberg, M., Su, C., and Oliw, E. (1998). Manganese lipoxygenase. Discovery of a bis-allylic hydroperoxide as product and intermediate in a lipoxygenase reaction. J. Biol. Chem. 273, 13080-13088.

63 144. Su, C., and Oliw, E.H. (1998). Manganese lipoxygenase. Purification and characterization. J. Biol. Chem. 273, 13072-13079. 145. Hörnsten, L., Su, C., Osbourn, A.E., Hellman, U., and Oliw, E.H. (2002). Cloning of the manganese lipoxygenase gene reveals homology with the lipoxygenase gene family. Eur. J. Biochem. 269, 2690-2697. 146. Wennman, A., and Oliw, E.H. (2013). Secretion of two novel enzymes, manganese 9S-lipoxygenase and epoxy alcohol synthase, by the rice pathogen Magnaporthe salvinii. J. Lipid Res. 54, 762-775. 147. Horowitz Brown, S., Zarnowski, R., Sharpee, W.C., and Keller, N.P. (2008). Morphological transitions governed by density dependence and lipoxygenase activity in Aspergillus flavus. Appl. Environ. Microbiol. 74, 5674-5685. 148. Jensen, E.C., Ogg, C., and Nickerson, K.W. (1992). Lipoxygenase inhibitors shift the yeast/mycelium dimorphism in Ceratocystis ulmi. Appl. Environ. Microbiol. 58, 2505-2508. 149. Robbertse, B., Reeves, J.B., Schoch, C.L., and Spatafora, J.W. (2006). A phylogenomic analysis of the Ascomycota. Fungal Genet. Biol. 43, 715- 725. 150. Amaike, S., and Keller, N.P. (2011). Aspergillus flavus. Annu. Rev. Phytopathol. 49, 107-133. 151. Shwab, E.K., and Keller, N.P. (2008). Regulation of secondary metabolite production in filamentous ascomycetes. Mycol. Res. 112, 225-230. 152. Ward, O.P. (2012). Production of recombinant proteins by filamentous fungi. Biotechnol. Adv. 30, 1119-1139. 153. Mabey Gilsenan, J.E., Atherton, G., Bartholomew, J., Giles, P.F., Attwood, T.K., Denning, D.W., and Bowyer, P. (2009). Aspergillus genomes and the Aspergillus cloud. Nucleic Acids Res. 37, D509-514. 154. Geiser, D.M., Klich, M.A., Frisvad, J.C., Peterson, S.W., Varga, J., and Samson, R.A. (2007). The current status of species recognition and identification in Aspergillus. Stud. Mycol. 59, 1-10. 155. Sanchez, J.F., Somoza, A.D., Keller, N.P., and Wang, C.C. (2012). Advances in Aspergillus secondary metabolite research in the post-genomic era. Nat. Prod. Rep. 29, 351-371. 156. Papagianni, M. (2007). Advances in citric acid fermentation by Aspergillus niger: biochemical aspects, membrane transport and modeling. Biotechnol. Adv. 25, 244-263. 157. Dagenais, T.R., and Keller, N.P. (2009). Pathogenesis of Aspergillus fumigatus in Invasive Aspergillosis. Clin. Microbiol. Rev. 22, 447-465. 158. Latge, J.P. (1999). Aspergillus fumigatus and aspergillosis. Clin. Microbiol. Rev. 12, 310-350. 159. Denning, D.W. (1996). Aspergillosis: diagnosis and treatment. Int. J. Antimicrob. Agents 6, 161-168. 160. Barrios-Gonzalez, J., and Miranda, R.U. (2009). Biotechnological production and applications of statins. Appl. Microbiol. Biotechnol. e-Pub 10 October, 2009.

64 161. Bonnarme, P., Gillet, B., Sepulchre, A.M., Role, C., Beloeil, J.C., and Ducrocq, C. (1995). Itaconate biosynthesis in Aspergillus terreus. J. Bacteriol. 177, 3573-3578. 162. Okabe, M., Lies, D., Kanamasa, S., and Park, E.Y. (2009). Biotechnological production of itaconic acid and its biosynthesis in Aspergillus terreus. Appl. Microbiol. Biotechnol. 84, 597-606. 163. Balajee, S.A. (2009). Aspergillus terreus complex. Med. Mycol. 47 Suppl 1, S42-46. 164. Walsh, T.J., Petraitis, V., Petraitiene, R., Field-Ridley, A., Sutton, D., Ghannoum, M., Sein, T., Schaufele, R., Peter, J., Bacher, J., et al. (2003). Experimental pulmonary aspergillosis due to Aspergillus terreus: pathogenesis and treatment of an emerging fungal pathogen resistant to amphotericin B. J. Infect. Dis. 188, 305-319. 165. Besi, M.I., Tucker, S.L., and Sesma, A. (2009). Magneporthe and its relatives. In eLS. (Chichester: John Wiley & Sons, Ltd), 1-9. 166. Cook, R.J. (2003). Take-all of wheat. Physiol. Mol. Plant Pathol. 62, 73-86. 167. Skamnioti, P., and Gurr, S.J. (2009). Against the grain: safeguarding rice from rice blast disease. Trends Biotechnol. 27, 141-150. 168. Dean, R., Van Kan, J.A., Pretorius, Z.A., Hammond-Kosack, K.E., Di Pietro, A., Spanu, P.D., Rudd, J.J., Dickman, M., Kahmann, R., Ellis, J., et al. (2012). The Top 10 fungal pathogens in molecular plant pathology. Mol. Plant Pathol. 13, 414-430. 169. Talbot, N.J. (2003). On the trail of a cereal killer: Exploring the biology of Magnaporthe grisea. Annu. Rev. Microbiol. 57, 177-202. 170. Dean, R.A., Talbot, N.J., Ebbole, D.J., Farman, M.L., Mitchell, T.K., Orbach, M.J., Thon, M., Kulkarni, R., Xu, J.R., Pan, H., et al. (2005). The genome sequence of the rice blast fungus Magnaporthe grisea. Nature 434, 980-986. 171. Michielse, C.B., and Rep, M. (2009). Pathogen profile update: Fusarium oxysporum. Mol. Plant Pathol. 10, 311-324. 172. Nucci, M., and Anaissie, E. (2007). Fusarium infections in immunocompromised patients. Clin. Microbiol. Rev. 20, 695-704. 173. Ma, L.-J., van der Does, H.C., Borkovich, K.A., Coleman, J.J., Daboussi, M.-J., Di Pietro, A., Dufresne, M., Freitag, M., Grabherr, M., Henrissat, B., et al. (2010). Comparative genomics reveals mobile pathogenicity chromosomes in Fusarium. Nature 464, 367-373. 174. Invitrogen (2006). InsectSelect System, For the stable expression of heterologous proteins in lepidopteran insect cell lines using pIZ/V5-His. 175. Theilmann, D.A., and Stewart, S. (1992). Molecular analysis of the trans- activating IE-2 gene of Orgyia pseudotsugata multicapsid nuclear polyhedrosis virus. Virology 187, 84-96. 176. Studier, F.W., and Moffatt, B.A. (1986). Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J. Mol. Biol. 189, 113-130. 177. Invitrogen (2006). Champion pET Directional TOPO Expression Kits. 178. Stratagene (2006). QuickChange Site-Directed Mutagenesis Kit.

65 179. Van Eeckhaut, A., Lanckmans, K., Sarre, S., Smolders, I., and Michotte, Y. (2009). Validation of bioanalytical LC-MS/MS assays: evaluation of matrix effects. J. Chromatogr. B, Analyt. Technol. Biomed. Life Sci. 877, 2198- 2207. 180. Scheiner, S. (2000). Calculation of isotope effects from first principles. Biochim. Biophys. Acta 1458, 28-42. 181. Cleland, W.W. (2005). The use of isotope effects to determine enzyme mechanisms. Arch. Biochem. Biophys. 433, 2-12. 182. Kohen, A., and Klinman, J.P. (1999). Hydrogen tunneling in biology. Chem. Biol. 6, R191-R198. 183. Westheimer, F.H. (1961). The magnitude of the primary kinetic isotope effect for compounds of hydrogen and deuterium. Chem. Rev. 61, 265-273. 184. Altschul, S.F., Gish, W., Miller, W., Myers, E.W., and Lipman, D.J. (1990). Basic local alignment search tool. J. Mol. Biol. 215, 403-410. 185. T.K. Attwood, A.G., N-E. Eriksson and E. Bongcam-Rudloff (2011). Concepts, Historical Milestones and the Central Place of Bioinformatics in Modern Biology: A European Perspective. Bioinformatics - Trends and Methodologies. In Tech. ISBN: 978-953-307-282-1. 186. Thompson, J.D., Higgins, D.G., and Gibson, T.J. (1994). CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22, 4673-4680. 187. Arnold, K., Bordoli, L., Kopp, J., and Schwede, T. (2006). The SWISS- MODEL workspace: a web-based environment for protein structure homology modelling. Bioinformatics 22, 195-201. 188. Tamura, K., Dudley, J., Nei, M., and Kumar, S. (2007). MEGA4: Molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24, 1596-1599. 189. Gardner, H.W. (1989). Soybean lipoxygenase-1 enzymically forms both (9S)- and (13S)-hydroperoxides from linoleic acid by a pH-dependent mechanism. Biochim. Biophys. Acta 1001, 274-281. 190. Funk, M.O., Jr., Andre, J.C., and Otsuki, T. (1987). Oxygenation of trans polyunsaturated fatty acids by lipoxygenase reveals steric features of the catalytic mechanism. Biochemistry 26, 6880-6884. 191. Hoffmann, I., Jernerén, F., and Oliw, E.H. (2013). Expression of fusion proteins of Aspergillus terreus reveals a novel allene oxide synthase. J. Biol. Chem. 288, 11459-11469. 192. Munro, A.W., Leys, D.G., McLean, K.J., Marshall, K.R., Ost, T.W., Daff, S., Miles, C.S., Chapman, S.K., Lysek, D.A., Moser, C.C., et al. (2002). P450 BM3: the very model of a modern flavocytochrome. Trends Biochem. Sci. 27, 250-257. 193. Johnson, K.A. (2008). Role of induced fit in enzyme specificity: a molecular forward/reverse switch. J. Biol. Chem. 283, 26297-26301. 194. Jernerén, F., Sesma, A., Franceschetti, M., Hamberg, M., and Oliw, E.H. (2010). Gene deletion of 7,8-linoleate diol synthase of the rice blast fungus: studies on pathogenicity, stereochemistry, and oxygenation mechanisms. J. Biol. Chem. 285, 5308-5316.

66 195. Gilbert, N.C., Rui, Z., Neau, D.B., Waight, M.T., Bartlett, S.G., Boeglin, W.E., Brash, A.R., and Newcomer, M.E. (2012). Conversion of human 5- lipoxygenase to a 15-lipoxygenase by a point mutation to mimic phosphorylation at Serine-663. FASEB J. e-pub April 21, 2012. 196. Koljak, R., Boutaud, O., Shieh, B.H., Samel, N., and Brash, A.R. (1997). Identification of a naturally occurring peroxidase-lipoxygenase fusion protein. Science 277, 1994-1996. 197. Boutaud, O., and Brash, A.R. (1999). Purification and catalytic activities of the two domains of the allene oxide synthase-lipoxygenase fusion protein of the coral Plexaura homomalla. J. Biol. Chem. 274, 33764-33770. 198. Gao, B., Boeglin, W.E., Zheng, Y., Schneider, C., and Brash, A.R. (2009). Evidence for an ionic intermediate in the transformation of fatty acid hydroperoxide by a catalase-related allene oxide synthase from the Cyanobacterium Acaryochloris marina. J. Biol. Chem. 284, 22087-22098. 199. Boeglin, W.E., Itoh, A., Zheng, Y., Coffa, G., Howe, G.A., and Brash, A.R. (2008). Investigation of substrate binding and product stereochemistry issues in two linoleate 9-lipoxygenases. Lipids 43, 979-987. 200. Jernerén, F., Eng, F., Hamberg, M., and Oliw, E.H. (2012). Linolenate 9R- dioxygenase and allene oxide synthase activities of Lasiodiplodia theobromae. Lipids 47, 65-73. 201. Wennman, A., Jernerén, F., Hamberg, M., and Oliw, E.H. (2012). Catalytic convergence of Mn- and Fe-lipoxygenases by replacement of a single amino acid. J. Biol. Chem. 287, 31757-31765. 202. Coffa, G., Schneider, C., and Brash, A.R. (2005). A comprehensive model of positional and stereo control in lipoxygenases. Biochem. Biophys. Res. Commun. 338, 87-92. 203. Hamberg, M. (2000). New cyclopentenone fatty acids formed from linoleic and linolenic acids in potato. Lipids 35, 353-363. 204. Grechkin, A.N., Mukhtarova, L.S., Latypova, L.R., Gogolev, Y., Toporkova, Y.Y., and Hamberg, M. (2008). Tomato CYP74C3 is a multifunctional enzyme not only synthesizing allene oxide but also catalyzing its hydrolysis and cyclization. Chembiochem. 9, 2498-2505. 205. Gao, B., Boeglin, W.E., and Brash, A.R. (2008). Role of the conserved distal heme asparagine of coral allene oxide synthase (Asn137) and human catalase (Asn148): mutations affect the rate but not the essential chemistry of the enzymatic transformations. Arch. Biochem. Biophys. 477, 285-290. 206. Schaller, A., and Stintzi, A. (2009). Enzymes in jasmonate biosynthesis - structure, function, regulation. Phytochemistry 70, 1532-1538.

67 Acta Universitatis Upsaliensis Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Pharmacy 176 Editor: The Dean of the Faculty of Pharmacy

A doctoral dissertation from the Faculty of Pharmacy, Uppsala University, is usually a summary of a number of papers. A few copies of the complete dissertation are kept at major Swedish research libraries, while the summary alone is distributed internationally through the series Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Pharmacy.

ACTA UNIVERSITATIS UPSALIENSIS Distribution: publications.uu.se UPPSALA urn:nbn:se:uu:diva-206199 2013