Exploring the Steroidogenic Activity of Glutathione Transferases Across Species

Home , Allyl isothiocyanate, GSTA3, Hydroxysteroid, Ketosteroid, Pregnenedione

Helena Lindström Exploring the steroidogenic activity of glutathione transferases across species

Exploring steroidogenicthe activity of glutathione transferases across species Helena Lindström

Helena Lindström is an IT-professional gone scientific. After several years in the telecom industry she obtained her M.Sc. in medicinal chemistry and continued to her docotoral studies at Stockholm University.

ISBN 978-91-7911-054-3

Department of Biochemistry and Biophysics

Doctoral Thesis in Neurochemistry with Molecular Neurobiology at Stockholm University, Sweden 2020

Exploring the steroidogenic activity of glutathione transferases across species Helena Lindström Academic dissertation for the Degree of Doctor of Philosophy in Neurochemistry with Molecular Neurobiology at Stockholm University to be publicly defended on Thursday 7 May 2020 at 10.00 in Magnélisalen, Kemiska övningslaboratoriet, Svante Arrhenius väg 16 B.

Abstract Glutathione transferases (GSTs) comprise a superfamily of enzymes prominently involved in detoxication. However, some GSTs have developed alternative functions. Thus, a member of the Alpha class GSTs in tissues of Homo sapiens (humans), Sus scrofa (pigs) and ruminants is involved in biosynthesis of steroid hormones, catalyzing a double-bond isomerization reaction as the last step of synthesis of Δ4-pregnene-3,20-dione (progesterone) and the obligatory step in the synthesis of the last precursor of testosterone, Δ4-androstenene-3,17-dione. Steroids regulate several vital aspects of life such as for example glucose homeostasis, inflammation, immunosuppression, blood pressure, reproduction and pregnancy. The human GST A3-3 was the most efficient steroid double-bond isomerase known so far in mammals. Our work extends discoveries of GSTs that act in the steroidogenic pathways in large mammals to Equus ferus caballus (horse). The kinetic profile of EcaGST A3-3 reveals a catalytic efficiency higher than that of the human enzyme making EcaGST A3-3 the most efficient steroid double-bond isomerase known today in mammals. In contrast to the rodents, Equus ferus caballus shares the steroidogenic pathway with Homo sapiens, which makes it a more suitable model for human steroidogenesis than the murine one. Inhibition of EcaGST A3-3 might help treat endocrine disorders. We screened a library of 1040 FDA-approved compounds for novel inhibitors of EcaGST A3-3 and made a further characterization of the most potent inhibitors. To extend the search for steroidogenic GSTs to other mammals, we probed the degree of GST A3-3 amino acid sequence conservation in Homo sapiens, Equus ferus caballus, Canis lupus familiaris (dog), Capra hircus (goat) and Monodelphis domestica (gray short-tailed opossum). We generated expression vectors containing homologous DNA from these species to facilitate further evaluation of the activity of these GSTs in mammals. We continued to expand the research to insects by investigating the steroidogenic activity of GSTE14 in Drosophila melanogaster (fruit fly), where this enzyme has been shown to be implicated in molting. Our work has provided insights into the role of GSTs in steroidogenesis in mammals and insects, further accentuating the functional versatility of GSTs. We have provided an initial step for the development of potential treatments of steroidogenic disorders as well as tools for further investigation of activity of these GSTs in mammals.

Keywords: Glutathione transferase, steroidogenesis, GST A3-3, testis, androstenedione, pregnenedione, equine, steroidogenesis inhibition, Drosophila GSTE14.

Stockholm 2020 http://urn.kb.se/resolve?urn=urn:nbn:se:su:diva-179443

ISBN 978-91-7911-054-3 ISBN 978-91-7911-055-0

Department of Biochemistry and Biophysics

Stockholm University, 106 91 Stockholm

EXPLORING THE STEROIDOGENIC ACTIVITY OF GLUTATHIONE TRANSFERASES ACROSS SPECIES

Helena Lindström

Exploring the steroidogenic activity of glutathione transferases across species

Helena Lindström ©Helena Lindström, Stockholm University 2020

ISBN print 978-91-7911-054-3 ISBN PDF 978-91-7911-055-0

Printed in Sweden by Universitetsservice US-AB, Stockholm 2020 Curiosity killed the cat. (There is no solid evidence for that though.)

List of publications

i. Lindström, H., Peer, S. M., Ing, N. H. & Mannervik, B. Characterization of equine GST A3-3 as a steroid isomerase. J. Steroid Biochem. Mol. Biol. 178, 117-126 (2018)

ii. Lindström, H., Mazari, A. M. A., Musdal, Y. & Mannervik, B. Potent inhibitors of equine steroid isomerase EcaGST A3-3. PLoS ONE, 14(3), e0214160

iii. Peer, S. M., Samollow, P., Lindström, H., Mannervik, B. & Ing, N. H. Conservation of glutathione transferase mRNA and protein sequences similar to human GST Alpha 3 across diverse mammalian species. Manuscript

iv. Škerlová, J., Lindström, H., Conis, E., Sjödin, B., Neiers, F., Stenmark, P. & Mannervik, B. Structure and steroid isomerase activity of Drosophila glutathione transferase E14 essential for ecdysteroid biosynthesis. FEBS Lett., 10.1002/1873-3468.13718

Table of contents Abbreviations ...... i Introduction ...... 1 Detoxication ...... 2 Glutathione transferases ...... 3 Glutathione ...... 3 Classification of mammalian GSTs ...... 4 Nomenclature ...... 4 Species distribution ...... 5 Structure ...... 7 Detoxication functions of cytosolic GSTs ...... 8 Endogenous GST substrates ...... 9 Xenobiotic GST substrates ...... 11 Biotechnological applications ...... 12 Alternative functions ...... 13 Steroids ...... 13 Corticosteroids ...... 14 Sex steroids ...... 18 Steroid biosynthesis in mammals ...... 20 Protective neuro-effects of steroid hormones ...... 22 Role of glutathione transferases in steroid biosynthesis ...... 23 The catalytic mechanism ...... 23 Steroid biosynthesis in insects ...... 26 Why this thesis? Hypotheses and aims ...... 27 Methods ...... 29 Heterologous expression ...... 29 Reverse transcription PCR (RT-PCR) and nested PCR ...... 29 Quantitative RT-PCR (RT-qPCR) ...... 30 Ligation ...... 30 Electroporation ...... 30 Affinity chromatography ...... 31 Kinetic measurements ...... 32 Irreproducibility due to low enzyme concentration ...... 33 Size-exclusion chromatography ...... 33 X-ray crystallography ...... 33 Present investigation ...... 35 Tissue distribution and kinetic profiles ...... 35 Tissue distribution ...... 35 Kinetic profiles ...... 36 Summary ...... 41 Inhibition profile ...... 41 The US Drug Collection ...... 41 The assay ...... 41 Sequence similarity ...... 46 Mechanistic aspects ...... 48 Alternate thiol cofactor ...... 51 DmGSTE14 structure ...... 52 Future outlook ...... 55 Concluding remarks ...... 57 The role of GST steroid isomerase in neurodegeneration ...... 58 Populärvetenskaplig sammanfattning ...... 61 Glutationtransferaser ...... 61 Steroider ...... 62 GSTernas roll i steroidgenes ...... 62 Vår forskning ...... 63 Acknowledgements ...... 67 The question of curiosity ...... 69 References ...... 71

Abbreviations

γ-Glu-Cys γ-glutamyl-cysteine Δ4-AD Δ4-androstene-3,17-dione Δ4-PD Δ4-pregnene-3,20-dione Δ5-AD Δ5-androstene-3,17-dione Δ5-PD Δ5-pregnene-3,20-dione 3β-HSD 3β-hydroxysteroid dehydrogenase 17β-HSD 17β-hydroxysteroid dehydrogenase ABC ATP-binding cassette ACTH adrenocorticotropic hormone AP-1 Activator protein 1 AR Androgen receptor ATP adenosine triphosphate Bta Bos taurus (bovine) cDNA complementary DNA CDNB 1-chloro-2,4-dinitrobenzene cds coding DNA sequence CRH corticotropin-releasing hormone DFT density functional theory Dm Drosophila melanogaster (fruit fly) Eca Equus ferus caballus (equine) FDA (U.S.) Food and Drug Administration FSH follicle-stimulating hormone GABA g-aminobutyric acid GLUT4 glucose transporter type 4 GnRH gonadotropin-releasing hormone GR glucocorticoid receptor GSH reduced glutathione GPCR G-protein coupled receptor GSSG oxidized disulfide form of glutathione GST glutathione S-transferase HNE, 4-HNE 4-hydroxynonenal, 4-hydroxy-2-nonenal HPA axis hypothalamic-pituitary-adrenal axis HPG axis hypothalamic-pituitary-gonadal axis Hsa Homo sapiens IMAC immobilized metal ion affinity chromatography JNK Jun N-terminal kinase

i LH luteinizing hormone MAPK, MAP-kinase mitogen-activated protein kinase MAPEG membrane-associated proteins in eicosanoid and glutathione metabolism MPD 2-methyl-2,4-pentanediol mRNA messenger RNA NADH nicotinamide adenine dinucleotide, reduced NADPH nicotinamide adenine dinucleotide phosphate, reduced NF‐κB nuclear factor kappa-light-chain-enhancer of activated B cells PAH polycyclic aromatic hydrocarbon PCR polymerase chain reaction PGDS human hematopoietic prostaglandin D2 synthase PR progesterone receptor RACE rapid amplification of cDNA ends ROS reactive oxygen species rRNA ribosomal RNA RT-PCR reverse transcription PCR RT-qPCR quantitative reverse transcription PCR SEC size-exclusion chromatography Ssc Sus scrofa (porcine) TNT 2,4,6-trinitrotoluene UTR untranslated region XRD x-ray diffraction

ii Introduction

A wide spectrum of human behavior and well-being depends on steroids. For example, apart from the well-known effect on libido and sexual function, testosterone has been shown to regulate bone density, muscle mass and body composition. Its concentration in the human body affects aggression, cardiovascular disease, mood, erythropoiesis, cognition and quality of life in general 1. Estrogen affects synaptic health and higher cognitive functions, maintains bone density and exerts anti-inflammatory effects leading to reduced occurrence of cardiovascular disease 2–5.

A thorough understanding of the steroidogenic function is essential in designing therapeutic approaches that can target rate-limiting steps or limit overproduction of intermediates, aiming for adequate levels of active steroids. Studies of differences in mammalian species generate insights into the metabolic adaptions that may have evolved to adjust to the particular properties of the steroidogenic enzymes, providing clues for identification of therapy targets and design of lead compounds.

Contrary to the rodents that are widely used as model systems, the steroid biosynthesis in Equus ferus caballus follows the same pathway, known as the Δ5 pathway, as in Homo sapiens 6. This makes Equus ferus caballus a more suitable model for biochemical and medical research in steroidogenesis than the rodents.

Therapeutics used for purposes distinct from targeting the steroidogenic function may bear side effect affecting the levels of active steroids in the organism. For example, it has been shown that the synthetic glucocorticoid dexamethasone affects steroidogenic genes in stallions 7,8. It is therefore of importance to investigate potential effects on the steroidogenic system of already approved compounds.

This thesis is focused on investigation of a detoxication enzyme performing a non-detoxication function. The main task of glutathione transferases (GSTs)

1 is their contribution to elimination of toxic electrophiles from the cell, yet some of these enzymes are known to have developed other roles distinct from the removal of harmful substances.

The two fields of detoxication and steroid biosynthesis converge in the present investigation, giving further insights into versatility of glutathione transferases and their functions.

Detoxication All organisms are exposed to a multitude of chemical substances that cause damage to the cells. Such substances may be of various origins, including xenobiotics and endogenously produced compounds and metabolites. Examples of the former are mutagenic and carcinogenic environmental chemicals such as polyaromatic hydrocarbons 9, and those of the latter comprise harmful products from oxidative metabolism of physiologically significant components such as lipids and nucleic acids 10. Many of these substances are lipophilic electrophiles that tend to accumulate in the organism. The cellular response evolved to limit the damage consists of a three-phase detoxication system aimed at minimizing the reactivity of the toxic compounds and converting them into more polar and thereby more easily excretable derivatives 11,12.

• Phase I detoxication aims at activation of the toxic substance for further detoxication in phase II. In this first line of enzymatic defense against toxic compounds, a reactive substitutient is either exposed or added by oxidation, reduction and/or hydrolysis reactions. Several enzymes are involved in phase I with cytochrome P450 being the prominent participant 13. This metal-dependent oxygenase uses the NADPH as cofactor to introduce a hydroxyl or an epoxide group into the toxic compound. Notably, this activation might yield metabolites more toxic than the parent compound, and therefore the role of phase II enzymes is crucial in avoidance of more cellular damage.

• The conjugation reactions of Phase II combine the activated substance with an endogenous hydrophilic cofactor producing a less reactive metabolite with sufficient hydrophilic character for subsequent excretion. The endogenous cofactors used in these

2 reactions are derived from the organism’s metabolism and include sulfate, glucuronic acid and glutathione (GSH). A number of enzymes are involved in phase II reactions, for example sulfotransferases, glucuronosyltransferases and glutathione transferases. Several isoforms of glutathione transferases exist that catalyze conjugation with the tripeptide glutathione 14.

• Phase III involves excretion of the conjugate, with the detoxication system consisting of the ATP-binding cassette superfamily of transporters (ABC) that includes multi-drug-resistant protein and p-glycoprotein. These transporters actively pump the conjugates out of the cell, reducing the intracellular toxin concentration 15.

Glutathione transferases Glutathione transferases are found in a wide range of species, from prokaryotes to multicellular eukaryotes 16–20. They were first discovered in 1961 in cytosolic fraction of rat liver, where they were found to be involved in conjugation of GSH to xenobiotic arylhalides 21,22. The essential role of glutathione transferases was later extended to catalyzing the conjugation of GSH to xenobiotic and endogenous lipophilic electrophiles, transforming them into more soluble and easily excretable peptide derivatives 23–25. Furthermore, the enzymes have evolved alternative functions distinct from detoxication 26,27.

Glutathione The tripeptide γ-L-glutamyl-L-cysteinyl-glycine is the most abundant thiol in mammalian cells 28. It is present in millimolar concentrations for example in the erythrocytes at concentrations of 2 mM and in hepatocyte cytosol at around 10 mM 29. Ninety nine percent of glutathione is in the reduced state (GSH) while the rest is in the oxidized disulfide form (GSSG).

Apart from being involved in the GST-catalyzed electrophile detoxifying reactions, GSH serves several critical functions, including scavenging reactive oxygen species, maintaining the essential thiol status of proteins, providing a cysteine reservoir, modulating critical cellular processes such as DNA synthesis and immune function, regulating nitric oxide homeostasis and

3 modulating protein activity by post-translational modification of proteins (protein S-glutathionylation). Almost 90% of cellular GSH is found in the cytosol while around 10% is located in the mitochondria where it acts as a major scavenger of reactive oxygen species 30–32.

Classification of mammalian GSTs Based on their intracellular localization, GSTs are divided into the three families cytosolic, mitochondrial and microsomal 33. Each family is further divided into classes based on sequence similarity:

• Cytosolic GSTs constitute the largest GST family in mammals and are homo- or heterodimeric proteins divided into seven classes designated by Greek letters Alpha, Mu, Pi, Theta, Omega, Zeta and Sigma. In humans, the sequence identity between two members of a given class may be more than 50% 14 while members of different classes share less than 25% sequence identity 34.

• The mitochondrial family comprises the Kappa class even though Alpha, Mu, Pi and Zeta can also be found in the mitochondria. The dimeric Kappa GSTs are structurally distinct from mammalian microsomal and cytosolic GSTs, exhibiting similarities to bacterial isomerases 35,36. Apart from mitochondria, Kappa GSTs are also present in peroxisomes and the endoplasmic reticulum.

• The homotrimeric microsomal GSTs, present in sub-cellular membranes, are known as membrane-associated proteins in eicosanoid and glutathione (MAPEG) metabolism. This family

comprises MGST 1, MGST 2, MGST 3, Prostaglandin E2-synthase

(PGE2S), Leukotriene C4-synthase (LTC4S), membrane-associated prostaglandin synthase-1 (MPGES1) and 5-lipoogynesase activating protein (FLAP) 37.

Nomenclature A rational nomenclature for human GSTs was adopted in 1992 38 and later extended to other mammalian species 39. A name of a GST is composed of the abbreviation “GST”, followed by the class to which the enzyme belongs, consecutively followed by the subunit categorization. Both subunits of

4 cytosolic GSTs are referenced, and a general denomination of such an enzyme will be “GST class subunit-subunit”. For example, GST of class A composed of two subunits number 3 would be GST A3-3. If needed, the name is prefixed by the species identification, for example Hsa (Home sapiens) or Eca (Equus ferus caballus): EcaGST A3-3.

Historically, the subunit numbering has been based merely on the order in which GSTs have been characterized and implies no designated relationship between GSTs in different species having the same subunit number. However, EcaGST A3-3 has been identified using the human GST A3-3 gene and is consequently named EcaGST A3-3.

Species distribution All cytosolic GST classes are present in each mammalian species, but the number of genes within each class varies across the phylogenetic tree 40. For example, human GST classes A and M comprise five intact genes, class O have two genes, and classes P, Z and S have one gene each. The equine class A, however, contains nine rather than five intact genes, which is the highest number of intact class A genes in mammals (Table 1).

5 Table 1. Overview of cytosolic glutathione transferase (GST) genes in 21 mammals. Class Alpha Mu Theta Pi Zeta Omega Sigma Total number Pseudogene (GSTA) GSTM) GSTT) (GSTP) GSTZ) (GSTO) (HPGDS) of GSTs proportion Human 12(7:5:0) 5(0:5:0) 3(0:3:0) 1(0:1:0) 1(0:1:0) 2(0:2:0) 1(0:1:0) 25(7:18:0) 0.28 Mouse 6(1:5:0) 10(4:6:0) 3(0:3:0) 7(0:7:0) 1(0:1:0) 2(0:2:0) 1(0:1:0) 30(5:25:0) 0.17 Naked mole rat 4(1:3:0) 10(0:10:0) 2(0:2:0) 2(0:2:0) 1(0:1:0) 2(1:1:0) 1(0:1:0) 22(2:20:0) 0.09 Bottlenose dolphin 3(1:0:2) 5(3:1:1) 2(1:1:0) 3(1:0:2) 1(0:0:1) 2(0:1:1) 1(0:1:0) 17(6:4:7) 0.35 Killer whale 3(1:2:0) 5(4:1:0) 2(0:2:0) 3(1:2:0) 1(0:1:0) 2(0:2:0) 1(0:1:0) 17(6:11:0) 0.35 Yangtze finless porpoise 3(1:2:0) 5(4:1:0) 2(1:1:0) 3(1:0:2) 1(0:0:1) 2(0:2:0) 1(0:1:0) 17(7:7:3) 0.41 Yangtze river dolphin 3(1:2:0) 5(4:1:0) 2(1:1:0) 3(1:2:0) 1(0:1:0) 2(0:2:0) 1(0:1:0) 17(7:10:0) 0.41 Sperm whale 3(1:2:0) 5(4:1:0) 2(0:1:1) 3(1:2:0) 1(0:1:0) 2(0:2:0) 1(0:1:0) 17(6:10:1) 0.35 Minke whale 3(1:2:0) 5(3:1:1) 2(1:1:0) 3(1:2:0) 1(0:1:0) 2(0:1:1) 1(0:1:0) 17(6:9:2) 0.35 Bowhead whale 2(0:2:0) 5(3:1:1) 2(1:1:0) 3(1:0:2) 1(0:0:1) 2(0:2:0) 1(0:1:0) 16(5:7:4) 0.31 Cow 6(2:4:0) 8(2:6:0) 2(0:2:0) 2(0:2:0) 1(0:1:0) 3(0:3:0) 1(0:1:0) 23(4:19:0) 0.17 Tibetan yak 6(1:5:0) 6(0:6:0) 3(0:3:0) 2(0:2:0) 1(0:1:0) 3(0:3:0) 1(0:1:0) 22(1:21:0) 0.05 Sheep 7(1:6:0) 8(2:6:0) 2(0:2:0) 2(0:2:0) 1(0:1:0) 4(0:4:0) 1(0:1:0) 25(3:22:0) 0.12 Tibetan antelope 6(1:5:0) 6(0:6:0) 2(0:2:0) 2(0:2:0) 1(0:1:0) 3(0:3:0) 1(0:1:0) 21(1:20:0) 0.05 Weddell seal 4(1:2:1) 5(2:3:0) 2(0:2:0) 7(1:4:2) 1(0:1:0) 2(0:2:0) 1(0:1:0) 22(4:15:3) 0.18 Pacific walrus 5(1:4:0) 8(2:6:0) 2(0:2:0) 6(1:5:0) 1(0:1:0) 2(0:2:0) 1(0:1:0) 25(4:21:0) 0.16 Dog 6(1:5:0) 5(3:2:0) 2(0:2:0) 12(6:6:0) 1(0:1:0) 2(0:2:0) 1(0:1:0) 29(10:19:0) 0.34 Horse 10(1:9:0) 6(2:4:0) 2(0:2:0) 2(0:2:0) 1(0:1:0) 2(0:2:0) 1(0:1:0) 24(3:21:0) 0.13 Microbat 6(0:6:0) 5(1:3:1) 4(0:4:0) 6(5:1:0) 1(0:1:0) 2(0:2:0) 1(0:1:0) 25(6:18:1) 0.24 Florida manatee 4(0:4:0) 8(0:8:0) 2(0:2:0) 1(0:1:0) 1(0:1:0) 2(0:2:0) 1(0:1:0) 19(0:19:0) 0.00 Elephant 3(0:3:0) 8(0:8:0) 2(0:1:1) 1(0:1:0) 1(0:1:0) 2(0:2:0) 1(0:1:0) 18(0:17:1) 0.00 The number outside brackets is the number of genes in a class, while the numbers within the brackets, separated by a colon, indicate the number of pseudogenes, intact gene, and partial genes. The number of intact Alpha class genes in the horse is highlighted. Modified with permission from 40.

6 It should be remembered though that not all intact genes might lead to corresponding expressed proteins. For example, in spite of the existence of the GSTA5 gene in humans, HsaGST A5-5 is yet to be discovered 41.

Structure In their active form, soluble GSTs are homo- or hetero-dimers composed of approximately 25 kDa subunits 14,16,33,42. In homodimers, both subunits are encoded by the same gene, while each subunit is a product of a distinct gene in heterodimers. Individual monomers have not proven to be catalytically active 43. Each subunit consists of two domains. The N-terminal domain comprises the first approximately 80 residues and harbors the major part of the GSH binding site (the G-site), whereas most of the binding site for the hydrophobic substrate (the H-site) is located in the larger C-terminal domain14 (Fig. 1).

Figure 1. Structure of HsaGST A3-3. One of the subunits is shown as semi- transparent accessible surface. A secondary structure cartoon of the other subunit is shown in blue to red from N- to C-terminus. Bound GSH is represented by yellow spheres and the hydrophobic steroid substrate androstene-3,17-dione is represented by black spheres. Modified with permission from 44.

7 Despite differences in sequence, the scaffold fold of the subunits is similar among many different classes, while the active-site region exhibits structural variations necessary for the binding of the various substrates 45. GSTs exhibit high specificity for GSH and the G-site is conserved among many classes of GST 14. The binding of the charged tripeptide is aided by multiple polar bonds in the active site (Fig. 2). The H-site, on the other hand, varies significantly among the enzymes designating the mainspring of substrate specificity 46,47.

Figure 2. GSH bound in a GST binding site of the Alpha, Mu, Pi and Theta classes. Reprinted with permission from 48.

Detoxication functions of cytosolic GSTs GSTs are mainly recognized for their detoxication function. Mammalian cytosolic GSTs detoxify a wide array of hydrophobic electrophiles in all organs and tissues 49. These electrophilic compounds are often cytotoxic, mutagenic and carcinogenic and comprise both xenobiotics and endogenous metabolites. The xenobiotic substrates include environmental toxins and therapeutic drugs while endogenous compounds can be electrophilic detoxication phase I metabolites or be generated by oxidative stress such as for example lipid peroxidation 50–53. By conjugation to GSH, the lipophilic substances are made more soluble and easily excretable, and the cell is thus protected from their harmful effects. The liver is the harbor of the most abundantly represented GSTs, followed by the kidney and the adrenal gland 49.

8 Some abundant GSTs exhibit substantial substrate promiscuity and catalyze GSH conjugation to several compounds while others catalyze more specific reactions.

Endogenous GST substrates

Organic hydroperoxides As a part of the defense system against oxidative stress, GST A1-1 and GST A2-2 are major hepatic enzymes with substrates such as organic hydroperoxides. Organic hydroperoxides are formed by oxidation of organic substances by molecular oxygen or hydroperoxide in virtually all living organisms and are markers of oxidative stress 54,55.

4-Hydroxynonenal 4-Hydroxynonenal (HNE) is a major product of n-6 fatty acid peroxidation elicited by reactive oxygen species (ROS). It has been shown to exert a variety of biological, signaling and cytotoxic effects primarily by rapid reactions with thiol and amino groups resulting in protein adducts, and has been implicated in the pathology of several diseases including diabetes, Alzheimer’s disease, atherosclerosis, cataract and cancer 56–60.

A major route of HNE elimination is conjugation to GSH by GST A4-4. Biologically formed HNE is a racemate where both enantiomers are assumed to be toxic, and GST A4-4 exhibits a stereochemical promiscuity towards the toxic racemic mixture with only negligible stereoselectivity towards the two enantiomers 61–63.

(R)-HNE (S)-HNE

Epoxides GST M1-1 is widespread in the mammalian organism catalyzing detoxication of epoxides. Reactive epoxides are intermediates produced in the first step of detoxication as a result of cytochrome P450 metabolism of various xenobiotics. Enzymatic conjugation of epoxides with GSH to less reactive products performed by GSTs is protective for the cell as it prevents the

9 epoxides from binding covalently to cellular components and bringing damage to the cell 64,65.

Ortho-quinones Ortho-quinones are products of dopamine oxidation that can be further oxidized to the toxic metabolite aminochrome. Aminochrome can give rise to redox-cycling leading to depletion of NADH/NADPH, and is involved in various disorders such as oxidative stress, mitochondrial dysfunction, ER stress, proteasome dysfunction and autophagy dysfunction 66. It can be used to induce a preclinical model to Parkinson’s disease 67.

GST M2-2 catalyzes conjugation of GSH with dopamine, dopa-o-quinone and aminochrome 68. In the brain, GST M2-2 is expressed in the astrocytes but interestingly the astrocytes secrete GST M2-2 and dopaminergic neurons are able to take it up, so it exerts its protective effect also in the dopaminergic neurons 69.

Isothiocyanates Isothiocyanates are biologically active hydrolysis products of glucosinolates, found in cruciferous vegetables such as broccoli, brussels sprouts, cabbage and kale 70,71. The biological activity of isothiocyanates is diverse. For example, many isothiocyanates with sulforaphane in particular are recognized inducers of antioxidant enzymes such as thioredoxin, quinone oxidoreductase, gammaglutamyl-cysteine ligase and GST 72–74. They also have anti- inflammatory and antibacterial effect 75–77. Isothiocyanates are rapidly conjugated with GSH by GST M1-1 and GST P1-1 in many tissues of the mammalian organism 78.

10 Xenobiotic GST substrates Some examples of the exogenous GST substrates are described below.

1-chloro-2,4-dinitrobenzene (CDNB) The discovery that CDNB is a substrate for most studied GSTs 79 has made it a widely used universal substrate used for comparison of specific activities of various GSTs. CDNB is a synthetic compound with no other significance for GSTs than that of a reference substrate.

Benzo(a)pyrene Benzo(a)pyrene is a polycyclic aromatic hydrocarbon (PAH) that results from incomplete combustion of organic matter. It is found in gasoline and diesel exhaust, tobacco smoke, coal tar and charcoal-broiled foods among other instances 80. It is a procarcinogen and a known inducer of lung cancer 81. GST A1-1, GST M1-1 and GST P1-1 catalyze the conjugation reaction of GSH with the carcinogenic diolepoxide formed by cytochromes P450 82,83. This metabolite intercalates in DNA and covalently binds to the nucleophilic guanine bases forming guanine benzopyrene, which distorts the double-helical DNA structure and causes mutations 84.

Therapeutic agents The detoxification exerted by GSTs is not always advantageous. Many therapeutics are excreted from the cell after GSH conjugation, which counteracts the therapeutic process. Several GSTs have been found to be implicated in drug resistance; in particular, GST P1-1 has been found to be

11 overexpressed in tumors and contribute to drug resistance 49,85–88. GSH- conjugated chemotherapeutics are eliminated from the cell by the phase III transporters such as MRP1/ABCC1 multidrug transporter protein 89.

Biotechnological applications GSTs are ubiquitous enzymes well suited for applications in biotechnology and drug discovery due to their versatility and essential physiological functions. One example is enhanced activation of the 6-mercaptopurine- releasing immunosuppression prodrug azathioprine, which is most efficiently activated by human GSTs A1-1, A2-2, and M1-1. By a combination of forced evolution and structure-based approach a resulting triple-point GST A2-2 H-site mutant displayed 70-fold higher catalytic efficiency than the parental enzyme 90,91.

Antibodies linked to drug-activating enzymes in antibody-directed enzyme prodrug therapy (ADEPT) further improve the prodrug activation. The antibody specifically targets epitopes that distinguish tumors from normal tissues 92, increasing the efficacy and minimizing the side effects of the chemotherapy. An antibody-fused GST administered prior the administration of the prodrug will generate a concentrated release of the active cytotoxic drug in the tumor tissue. The prodrug TER286TLK286//Telcyta has the advantage among several GST-activated drugs of not requiring GSH as a co-substrate and is therefore independent of the ambient concentration of glutathione 93,94.

Another example is phytoremediation by GST-catalyzed GSH-conjugation with the environmental pollutant 2,4,6-trinitrotoluene (TNT), disseminated over vast territories by large-scale military and industrial activities. Arabidopsis thaliana plantlets over-expressing one of its GSTs inactivate TNT and deplete the growth medium of the compound 95. However the plant GSTs have modest TNT detoxication activity; in contrast, Drosophila melanogaster enzymes GSTE6 and GSTE7 were demonstrated to be orders of magnitude more efficient 96. Plants trans-genetically expressing the Drosophila GSTE6 increased the uptake of TNT from the growth medium and were more resistant to the compound than both unmodified plants and the Arabidopsis lines over-expressing two of their GSTs 97.

12 Alternative functions In addition to the widely accepted detoxication function some GSTs have developed other roles, distinct from catalyzing conjugation of GSH to electrophiles. Most of these reactions are catalytic in their nature, while some have lost the catalytic function altogether.

Examples of GSTs that have adopted non-catalytic alternative functions include GST P1-1 that participates in cellular MAP-kinase signaling by regulating the activity of Jun N-terminal kinase (JNK) 98,99. It is also involved in maintaining of cellular redox homeostasis and S-glutathionylation of proteins such as enzymes, receptors, transport proteins and transcription factors 100–102, and is ubiquitously expressed in the mammalian organism 49. GSTA1-1 and A2-2, highly expressed in the liver 49, act as ligandins by facilitating intracellular transport of a number of compounds such as steroids, bilirubin, thyroid hormone and certain carcinogens. These compounds are bound non-specifically and in some cases the interface between the two subunits is involved in the binding 103–105.

Other GSTs have left the field of detoxication but kept the catalytic function. Human Sigma class GST is involved in the isomerization of prostaglandin H2 to prostaglandin D2 and is therefore called human hematopoietic prostaglandin D2 synthase (PGDS) 106. Human and porcine Alpha class GSTs are involved in isomerization reactions in the path of steroid biosynthesis 107,108. This alternative role of glutathione transferases constitutes the imminent frame of this thesis.

Steroids The two main biological functions of steroids are being a component of biological membrane, affecting its fluidity (e. g. cholesterol), and that of a signaling molecule, acting on steroid receptors 109. Steroids facilely diffuse through the cell membrane and are found in bacteria and eukaryotes including fungi, plants and animals. In mammals, the main signaling steroids include 110: • Corticosteroids o Glucocorticoids regulating several physiological processes, for example glucose homeostasis 111, inflammation and immunosuppression 112,113

13 o Mineralocorticoids participating in the regulation of blood pressure 114

• Sex steroids 115 o Progestogens such as progesterone, vital for maintaining pregnancy o Androgens such as testosterone, involved in development and maintenance of male sex characteristics o Estrogens such as estradiol, involved in development and maintenance of female sex characteristics

The core structure of steroids consists of three cyclohexane rings and a cyclopentane ring, to which various organic chains can be attached. It may also contain varying degrees of unsaturations and scissions 116.

Corticosteroids Corticosteroids are synthesized in the adrenal cortex of vertebrates. Synthetic analogues of these hormones are also called corticosteroids 117.

A basic structural requirement for corticosteroid activity is a -CO-CH2OH sidechain attached at C-17 of a C-21 steroid (Fig. 6). There must also be an unsaturated bond between C-4 and C-5 and a keto group (-C=O) at C-3. Specific glucocorticoid activity requires a hydroxyl group at C-11; the activity is enhanced by a hydroxyl group at C-17. Mineralocorticoid activity requires a hydroxyl group on C-21 whereas the presence of hydroxyl group at C-17 decrease the activity 117.

In response to stress, glucocorticoids (primarily cortisol) exert their metabolic, anti-inflammatory and immunosuppressing actions by acting on the ubiquitously distributed intracellular glucocorticoid receptor (GR) 118. This signaling is regulated by the hypothalamic-pituitary-adrenal (HPA) axis 113,119. The HPA axis refers to interactions of a collection of structures consisting of nucleus of the hypothalamus, the anterior lobe of the pituitary gland, and the adrenal gland 119 (Fig. 3). The corticotropin-releasing hormone (CRH) is synthesized and released by the hypothalamus in response to stress. Binding of CRH to its receptor on pituitary corticotropes induces the release of adrenocorticotropic hormone (ACTH) into the systemic circulation. The

14 principal target for circulating ACTH is the adrenal cortex, where it stimulates glucocorticoid synthesis and secretion from the zona fasciculata.

Figure 3. Schematic representation of the hypothalamic-pituitary-adrenal axis. CRH, corticotropin-releasing hormone; ACTH, adrenocorticotropic hormone.

The main glucocorticoid effect on glucose homeostasis is to retain plasma glucose levels for optimal brain functioning during periods of stress. Thus, glucocorticoids antagonize insulin effect by promoting hepatic gluconeogenesis while decreasing glucose uptake and utilization in skeletal muscle and white adipose tissue. Gluconeogenesis is promoted mainly by GR- mediated transcription of genes encoding enzymes in the gluconeogenic pathway, such as for example pyruvate carboxylase, cytosolic phoshoenolpyruvate carboxykinase, fructose-1,6-bisphospahatase, phosphofructokinase 2/fructose bisphosphatase 2 and glucose-6-phosphatase catalytic subunit. Glucose uptake in skeletal muscle and adipose tissue is

15 decreased mainly by inhibition of translocation of the glucose transporter GLUT4 to the plasma membrane 111,120.

Examples of anti-inflammatory proteins induced by glucocorticoids include annexin I 121,122, MAPK phosphatase 1 123 and interleukin-10 112,124. Glucocorticoids also suppress the synthesis of inflammatory mediators such as cytokines, enzymes, receptors and adhesion molecules. This is achieved mainly by negative regulation of transcription factors such as AP-1 and NF-κB, which are known for their control of the expression of numerous proinflammatory genes 125,126. Apart from these direct and indirect genomic actions, glucocorticoids are also believed to exert nongenomic effects 113.

The primary mineralocorticoid is aldosterone, produced as part of regulation by the renin-angiotensin-aldosterone system 127 (Fig. 4). Hypotension leads to increase in angiotensin level, which augments aldosterone production 128,129.

Figure 4. The renin-angiotensin-aldosterone system. In response to hypotension, renin is released by the kidney and cleaves angiotensinogen into angiotensin I which is converted to angiotensin II by angiotensin converting enzyme (ACE). Aldosterone synthesis is induced in adrenal cortex by angiotensin II and a number of secondary factors. Aldosterone is released into circulation to promote sodium retention in the kidney and exerts its effect on the heart and other organs. Reprinted with permission from 127.

Aldosterone exerts its effect via the mineralocorticoid receptor 130,131, a ligand- activated transcription factor expressed mainly in renal epithelial cells but also in nonepithelial tissue such as the heart 132 and the brain 133. The water homeostasis is regulated by transcription of serum-and-glucocorticoid regulated kinase 1, epithelial sodium channels, Na+/K+-ATPase and other

17 entities in renal collecting duct cells 134,135. In addition to the genomic actions, aldosterone also elicits rapid and probably mostly nongenomic effects on several pathways, involving protein kinase C, Na+/H+ exchange, levels of intracellular calcium and MAP kinase 136,137.

Sex steroids Sex steroids (also known as gonadal steroids), are produced under regulation by the hypothalamus-pituitary-gonadal (HPG) axis 138,139. The HPG axis comprises the hypothalamus, pituitary gland and the gonadal glands, where gonadotropin-releasing hormone (GnRH) produced by the hypothalamus stimulates the pituitary gland to release luteinizing hormone (LH) and follicle- stimulating hormone (FSH). In response to the circulating LH and FSH, the gonads produce sex steroids (Fig. 5).

Figure 5. Schematic representation of the hypothalamic-pituitary-gonadal axis. GnRH, gonadotropin-releasing hormone; LH, luteinizing hormone; FSH, follicle- stimulating hormone.

18 The sex steroids exert their effects by slow genomic mechanisms via the nuclear androgen, progesterone and estrogen receptors as well as by faster mechanisms acting on membrane-associated receptors and signaling cascades 140,141.

Endogenous progestogens are characterized by the basic 21-carbon pregnane skeleton (C21), androgens possess an androstane skeleton (C19) and estrogens a C-18 estrane skeleton (Fig. 6).

The principal progestogen progesterone, in conjunction with estrogen, controls uterine function to facilitate reproduction by acting on cells in endometrium, myometrium and the cervix 142. The effect of progesterone binding is mediated by combined activities of progesterone receptors (PRs) PR-A and PR-B 143. This is done by regulating gene expression in two modes of action: in the direct genomic action the PR acts as a nuclear transcription factor, and in the indirect extranuclear mode where the PR activates the MAP kinases 144. More rapid non-genomic signaling events involve putative progesterone-binding moieties such as for example GABAA and oxytocin receptors 145.

Androgens bind to the nuclear androgen receptor (AR) to mediate most of its biological effects such as protein synthesis and tissue growth, promoting increased muscle mass and bone density. They also regulate maturation of the male sex organs 146–149. The more rapid AR-independent, nongenomic actions may be exerted by inducing the MAPK signal cascade or GPCR-activated second messenger signaling 150. Androgens are also synthesized in lower levels in females. One example of androgen biological function in females is a nongenomic prevention of premature uterine contractions in pregnancy 151.

The three primary estrogens are estradiol (predominant during reproductive years), estriol (predominant during pregnancy) and estrone (predominant during menopause). The genomic actions of estrogens are mediated by the nuclear estrogen receptors α and ß 152, which might be involved in the nongenomic pathways as well 153,154. The non-genomic pathways are believed to also be mediated by GPER/GPr30, a GPCR 155,156.

19 Steroid biosynthesis in mammals Progesterone is synthesized mainly in the corpus luteum in the ovary and in placenta in females during pregnancy 157. In males, testosterone is synthesized principally in the Leydig cells in the testis, and in smaller quantities in the adrenal gland 158. Females also produce testosterone but in far smaller total quantities in the adrenal gland and the ovary 159. Estrogens are synthesized in females primarily in granulosa cells of the ovarian follicles and corpus luteum, and in the placenta during pregnancy 160.

Steroid hormone biosynthesis starts with cholesterol and proceeds through the common precursor pregn-5-en-3β-ol-20-one (pregnenolone) to formation of the sex steroids (progestogens, androgens and estrogens), mineralocorticoids and glucocorticoids 161,162 (Fig. 6). The biosynthesis of sex steroids is catalyzed by a collection of diverse enzymes consisting of 3β-hydroxysteroid dehydrogenase (3β-HSD) and 17β-hydroxysteroid dehydrogenase (17β-HSD) and by three members of the cytochrome P450 superfamily. These three enzymes, belonging to the family otherwise known as that of detoxication enzymes, have developed functions vital for catalysis of steroidogenic reactions. The enzymes include CYP11A1, also known as cholesterol sidechain cleavage enzyme, CYP17A1 with 17α-hydroxylase and 17,20 lyase activities, and the aromatase CYP19A1 163. The steroidogenic enzymes catalyze multiple reactions and the sex hormone synthesis pathway varies between mammalian species, adding to the complexity of steroid biosynthesis. The Δ5 pathway (with the double-bond between carbons 5 and 6 in the metabolites) through 17α-hydroxypregnenolone is followed in Homo sapiens, Sus scrofa and ruminants, while in rodents the preferred route is the Δ4-pathway (with the double bond between carbons 4 and 5) through progesterone 161,164 (Fig. 6).

Mineralocorticoids and glucocorticoids are synthesized in the adrenal cortex in a zone depended manner by 3β-HSD and the members of the cytochrome P450 superfamily CYP11A1, 21-hydroxylase CYP21A2, 11β-hydroxylase CYP11B1 and aldosterone synthase CYP11B2. Aldosterone synthase is present only in zona glomerulosa, therefore aldosterone synthesis is restricted to this zone of the adrenal cortex, while cortisol synthesis takes place in zona fasciculata 165,166.

Figure 6. Häggström M, Richfield D (2014). "Diagram of the pathways of human steroidogenesis". WikiJournal of Medicine 1 (1). DOI:10.15347/wjm/2014.005. ISSN 20024436. Reprinted with permission.

In mammalian species following the Δ5 pathway, one of the last steps in the synthesis of progesterone and testosterone is an alcohol dehydrogenation reaction yielding Δ5-unsaturated 3-ketosteroids followed by a double-bond isomerization. Both reactions are catalyzed by the 3β-HSD 167. The double- bond isomerization of Δ5-androstene-3,17-dione (Δ5-AD) leads to the last precursor of testosterone, Δ4-androstene-3,17-dione (Δ4-AD). In the synthesis of progesterone, isomerization from Δ5-pregnene-3,20-dione (Δ5-PD) to Δ4-pregnene-3,20-dione (Δ4-PD) is the last step in progesterone biosynthesis.

21 Protective neuro-effects of steroid hormones Steroid hormones synthesized de novo in neural tissues are referred to as neurosteroids. Neurosteroids are involved in such diverse nervous system- associated functions as emotion, mood, cognition, sexual and social behavior, and have shown neuroprotective activity in animal models of Alzheimer’s disease and Parkinson’s disease. They also play important roles in ageing 168– 170.

As in the peripheral steroidogenic organs, brain cholesterol is converted to pregnenolone by the cholesterol side cleavage enzyme (CYP11A1) in the mitochondria. In the smooth endoplasmic reticulum, pregnenolone is subsequently metabolized into neurosteroids progesterone, dehydroepiandrosterone (DHEA), estradiol (Fig. 6) and allopregnanolone 168.

Neurosteroids exert their effect in two ways: by genomic actions on nuclear steroid hormone receptors and by nongenomic actions on neurotransmitter receptors. The genomic actions on for example the estrogen receptor mainly activate second-messenger cascades and control protein transcription 171 while the nongenomic actions primarily concern the GABAA receptor.

Neurosteroids possess the ability of positive and negative modulation of primarily GABAA receptors associated with an array of physiological and pathophysiological conditions including stress, pregnancy, depression and epilepsy 172. They also have been shown to allosterically modulate the NMDA receptors, implicated in synaptic plasticity, learning and memory 173,174.

22 Role of glutathione transferases in steroid biosynthesis Corticosteroid binding to a rat liver protein was first reported 50 years ago 175. The protein was named ligandin due to its acceptance of a wide range of various ligands 105 and was recognized as a GST in 1974 176. Three years later a glutathione-dependent Δ5-3-ketosteroid isomerization activity was linked to GSTs 177,178. Approximately a decade later the Mu class GSTs were found to bind steroid compounds with high affinity 179. In spite of these findings, however, no definite role of GSTs in steroid metabolism was established.

The first human GST identified as an efficient Δ5-Δ4 steroid double-bond isomerase was HsaGST A1-1 in 2001 180. The same year, HsaGST A3-3 was demonstrated to possess prominent Δ5-Δ4 steroid double-bond isomerization 107 activity and to exceed the catalytic efficiency (kcat/Km) of HsaGST A1-1 by 20-fold 107. In Sus scrofa and Bos taurus, the isomerization reaction has also been shown to be catalyzed by an Alpha class GST 108,181. The highest catalytic efficiency with Δ5-AD was found to be exerted by HsaGST A3-3, being 230-fold higher than that of the 3β-HSD 107.

The catalytic mechanism Several attempts to elucidate the mechanism of the steroid double-bond isomerization catalyzed by human GSTs have been made. Examples include attempts based on point mutations 47,180,182, computational studies183,184 and crystalized enzyme structures 44,185. The current consensus was proposed by Dourado et al in April 2015 186 and is based on a density functional theory (DFT) calculations computational study and the GST A3-3:GSH:Δ4-AD crystal structure solved by Tars et al 44. The mechanism consists of five steps (Fig. 7).

Unlike classical GST two-substrate reactions where GSH is a substrate along with the lipophilic electrophile, in this mechanism GSH acts as an acid-base catalyst and thus is not consumed. A tyrosine in the H-site (Tyr9) also has a catalytical role, and the catalysis revolves around three chemical steps and two conformational rearrangements of Tyr9:

• In the first (chemical) step, the thiolate of GSH acts as a base and deprotonates carbon4 (C4).

23 • The second step is a conformational rearrangement of Tyr9 leading to a direct interaction Tyr9 - C6. • In the third step, Tyr9 acts as an acid catalyst and protonates C6. This is the second chemical step. • The fourth step of the mechanism is another Tyr9 rearrangement that allows direct interaction Tyr9 – GSH thiol. • In the fifth step of the mechanism, which is the third chemical step, Tyr9 deprotonates the GSH thiol, restoring the active site of the enzyme to its initial state. GSH has a dual role in this reaction. In addition to its role as an acid-base catalyst, throughout the entire catalytical cycle it is also stabilizing the O3 atom of the steroid substrate by a hydrogen bond from the GSH-Gly main chain amide.

Figure 7. Mechanistic proposal for the Δ5-AD isomerization reaction catalyzed by HsaGST A3-3. The mechanism is composed of three catalytic steps and two Tyr9 conformational rearrangements. Reprinted with permission from 186.

25 Steroid biosynthesis in insects The process of insect molting (ecdysis) involves the shedding of the cuticle (exoskeleton) 187. Ecdysis is regulated by the major insect steroid hormones ecdysteroids. 20-hydroxyecdysone (Fig. 8), a metabolite of ecdysone, is the most biologically active steroid required for ecdysis 188,189.

Figure 8. 20-hydroxyecdysone.

A number of cytochrome P450 enzymes are implicated in biosynthesis of ecdysteroids in Drosophila melanogaster 190. The genome of D. melanogaster also harbors 42 cytosolic and membrane-bound GST genes assigned to six classes: theta, omega, sigma, zeta, delta, and epsilon, of which the delta and epsilon are specific to arthropods including insects while the rest are common for the eukaryotes 191–193. A member of the epsilon class GSTE14 plays an imperative role in biosynthesis of ecdysteroids as its loss disrupts formation of exoskeleton and prevents ecdysis 194,195.

26 Why this thesis? Hypotheses and aims

Before I started my research, steroid double isomerase reactions were demonstrated to be catalyzed by Alpha class GSTs in Homo sapiens, Sus scrofa and Box taurus 107,108,181,196, the human GST being the most efficient one. The question that arose was if the GST had a physiologically significant role in steroid biosynthesis. Has it evolved to catalyze this reaction in vivo or are we looking at a capability that it has but that is not fully used, the double- bond isomerization actually carried out mainly by the traditionally accepted 3β-HSD? Catalytic efficiency of HsaGST A3-3 is indeed high, but if there is a GST with a high steroid double-bond isomerization efficiency also in another mammal, that would add to the evidence of the physiological significance of GSTs for steroid biosynthesis in mammals.

The complementary DNA (cDNA) sequence of an equine GST exhibited a high sequence similarity with that of the efficient steroid double bond isomerase HsaGST A3-3. The equine GST was therefore named EcaGST A3-3 and our hypothesis was that it may possess steroid double bond isomerization activity. A tissue distribution of the equine enzyme primarily to steroidogenic organs such as the gonads and adrenal gland would support the hypothesis.

Should this enzyme be revealed to possess a significant activity and thus contribute to high production of steroids, then it would be a potential drug target and we might be interested in inhibiting it. There are several pathologies associated with excessive steroid production, such as for example polycystic ovary syndrome and congenital adrenal hyperplasia 197,198. In addition, sex steroid hormones stimulate neoplastic cell growth in breast and endometrial cancers 199. Aldosterone has been associated with hypertension and glucocorticoids with osteoporosis 200,201. Conversely, there might be compounds, for example drugs already in use, that exert undesired inhibitory side effects on this enzyme when used to aim at other targets. Hence there are reasons to know what compounds have the capability to inhibit this enzyme.

27 The search for alpha-class GST enzymes with steroid isomerase activities should not make a halt at Equus ferus caballus but should be extended to other species, for example the research animals dog (Canis lupus familiaris), goat (Capra hircus) and gray short-tailed opossum (Monodelphis domestica) available with our collaborators at Texas A&M University.

In D. melanogaster, a limited number of the GSTs have been characterized in detail, and among the epsilon class enzymes, structural studies of only GSTE6 and GSTE7 have previously been published 202. GSTE14 was found to be implicated in biosynthesis of ecdysteroids, but a particular reaction catalyzed by this enzyme is unknown. To further probe the steroidogenic activity of this enzyme, we hypothesized that it might possess steroid double bond isomerization activity with the substrates of equine and human GST steroid isomerases.

Our aims comprised:

• To characterize EcaGST A3-3 by establishing its kinetic and inhibition profile and its tissue distribution (Paper I and II) • To initiate analysis of steroidogenic capability of GST A3-3 enzymes in other tetrapods by establishing GST A3-3 amino acid sequence of dog, goat and gray short-tailed opossum (Paper III) • To establish the kinetic profile and structure of DmGSTE14 (Paper IV) as a first step in the analysis of its steroidogenic capability.

28 Methods

This section provides the background of some of the methods used in the experiments reported in this thesis. For more detailed description of the methods see the included papers.

Heterologous expression Several experimental steps lead from the synthesized or tissue-acquired gene to sufficient amount of purified enzyme. This usually involves amplifying a small starting DNA material into multiple copies and ligating the copies into plasmids to be inserted into a suitable bacterial or eukaryotic host. The transformed host cells are subsequently grown under favorable conditions, harvested and lysed, and the overexpressed enzyme is purified. The following sections describe techniques employed for this purpose and referred in this thesis.

Reverse transcription PCR (RT-PCR) and nested PCR RT-PCR is widely used to detect RNA expression. In this variant of the PCR203, the isolated RNA is first converted into cDNA by the RNA-dependent DNA polymerase - the reverse transcriptase 204. The obtained cDNA is amplified by PCR in the subsequent step.

Nested PCR is used to reduce generation of unwanted products due to unspecific primer binding 205. The procedure involves two sets of primers and two successive PCR runs. The first run deploys one set of primers in a low number of PCR cycles, limiting unspecific primer binding. The second run uses the second set of primers and is aimed to multiply only the intended target, and not the unspecific product or the flanking sequences, from the first run.

29 Quantitative RT-PCR (RT-qPCR) RT-qPCR monitors the amplification of the target gene during the PCR by means of a fluorophore that allows measurement of the amount of amplified cDNA 206,207. One or more reference genes are used to normalize for variation in amount and quality of target RNA between samples. For this normalization to be correct, the reference and the target genes must be amplified with comparable efficiencies 208,209.

Ligation In the procedure of ligation, DNA fragments amplified by PCR are inserted into plasmids by DNA ligase. This enzyme catalyzes the formation of a phosphodiester bond between two DNA fragments where the 3’-OH of a strand is ligated to the 5’-phosphate group of the other strand 210,211.

Before ligation, the DNA fragments and the plasmids are digested with restriction endonucleases, that cleave the plasmid at the recognition site and create specific flanking sequences around the target gene. Two commonly used restriction endonucleases used for this purpose are EcoRI and XhoI. EcoRI creates 4 nucleotide sticky ends with 5’ end overhangs of AATT while XhoI recognizes the double-stranded DNA sequence CTCGAG and cleaves after C-1, creating a blunt end 212. The two restriction enzymes are used to minimize religation without insertion and allow the fragment to be inserted in a directional fashion. pET-21a(+) is one of the commercially available plasmids from the pET system, which is an efficient system used for cloning and expression of recombinant proteins in E. coli. Target genes are cloned in pET-21a(+) vectors to be under control of bacteriophage T7 RNA polymerase which is so selective and active that, at maximum induction, almost all of the cell’s resources are devoted to target gene expression.

Electroporation The process of electroporation transforms a cell by applying an electrical field to the cell membrane, allowing charged external material such as chemical compounds or DNA to enter the cell through the hydrophobic membrane. Thousands of volts are passed across a distance of up to two millimeters of

30 suspended cells, creating nanometer-scale pores in the membrane making it feasible for the charged external material to diffuse into the cell 213,214.

BL21 (DE3) is an E. coli strain extensively used for recombinant protein overexpression. It features protease deficiency and carries a DE3 recombinant phage containing a gene of the selective and efficient T7 RNA polymerase 215,216 that can direct high-level expression of cloned genes under the control of the T7 promoter, present in certain vectors such as for example pET-21a(+).

Affinity chromatography Affinity chromatography is widely used for purification of recombinant proteins. It is a biochemical method for separation of mixtures based on specific interactions between biochemical components of the mobile phase and the stationary phase, for example between enzyme and substrate or antigen and antibody 217. The stationary phase is usually based on agarose, with a hydrocarbon chain spacer attached to it to prevent steric interference during binding 218. The substrate or antigen is covalently bound to the other end of the spacer. The mobile phase containing the target molecule is poured onto the stationary phase and allowed to bind. Unwanted material is then washed away leaving the bound target molecule, which can be eluted by a salt gradient or in case of an enzyme by high substrate concentration.

Sample cleanup is performed by dialysis where large solutes are retained and small solutes are removed by differential diffusion through a semipermeable membrane.

Tissue-acquired glutathione transferases are usually purified on glutathione resins that have glutathione covalently attached to the spacer (Fig. 9) in order to bind enzymes that have glutathione as substrate. This technique utilizes the multiple polar bonds that aid the binding of glutathione to the active site of the enzyme (Fig. 2). One example of such a resin is GSH-sepharose, a crosslinked, bead-formed agarose adjusted to resist biological degradation and retain stability within the pH range of 4 to 9 219. The bound enzyme is then eluted by high concentrations of glutathione or S-hexylglutathione. S- hexylglutathione is a glutathione derivative with a hexyl carbon chain covalently linked to glutathione sulfhydryl group, and is used as inhibitor in glutathione transferase assays.

Figure 9. Glutathione immobilized on agarose bead.

Glutathione transferases heterologically expressed from synthetically manufactured genes may have a polyhistidine tag at the amino- or carboxy- terminus allowing for the high purity IMAC purification 220–222.

Kinetic measurements The best measure of initial velocity of an enzymatic reaction is continuous monitoring of product formation or substrate disappearance with time over a period of the initial velocity phase. This phase is defined by the linearity of the plot of product or substrate concentration as a function of time.

Spectrophotometric assays utilize the diversified absorbance of the substrate and the product and follow the reaction in real time. These fast and non- obtrusive assays are widely employed for kinetic analyses 223. One particular advantage of this method is the clear indication of the initial velocity phase by the linear slope of the plot during the initial part of the reaction.

32 Irreproducibility due to low enzyme concentration EcaGST A3-3 had to be diluted to nanomolar concentrations due to its high activity with steroid substrates. These low concentrations caused irreproducibility due to the enzyme’s rapid adsorption to inner walls of test tubes. Similar behavior was previously observed with HsaGST A3-3 224. After numerous attempts to solve the problem, Sarstedt micro-tubes 1.5 mL EASY-CAP ref. no. 72.690.550 were found not to adsorb the enzyme and were subsequently used.

Size-exclusion chromatography Size-exclusion chromatography (SEC) is a chromatographic method where molecules in a solution are separated based on their size. It is used mainly for separation of large molecules such as proteins or polymers. The analyte in the mobile phase is trapped in pores of the adsorbent, which usually consists of tightly packed polymer beads largely occupied by pores (depressions or channels) of various sizes. Large solutes are too large for some of the pores and so get trapped to a lesser extent than small molecules, that have access to a larger total pore volume. Therefore the retention time of the large molecules is shorter than that of small molecules 225,226.

The eluate is typically collected in fractions containing different analyte concentrations.

X-ray crystallography Protein crystallization allows determination of protein structure. To identify the structure, a developed protein crystal of high quality may be subjected to x-ray diffraction (XRD), a method frequently used in structural biology and drug discovery 227. Other methods are nuclear magnetic resonance 228 for structure determination of smaller molecules and cryo-electron microscopy 229 for macromolecular structure determination.

Protein crystallization is based on a creation of a solution that is supersaturated in the protein but that otherwise does not perturb the protein’s natural state. Supersaturation is a non-equilibrium state where under specific chemical and physical conditions a quantity of the protein in excess of the solubility limit is

33 present in the solution. This is achieved by changing physicochemical characteristics of the solution by for example altering the pH (alters the ionization state of surface amino acid residues) or adding a salt (alters the chemical activity of the water). Equilibrium is re-established by the formation and development of a solid state, such as crystals, as the saturation limit is attained 230.

Examples of the most applied techniques for achieving supersaturated state include hanging- and sitting-drop vapor diffusion, and dialysis 230. In the hanging drop vapor diffusion, a drop is suspended from a cover slip over a reservoir which contains some concentration of a salt or polymer precipitant. The drop consists of half stock protein solution and half reservoir solution. By water equilibration through the vapor phase the drop ultimately approaches the reservoir in osmolarity, increasing both precipitant and protein concentrations in the drop.

The aim of XRD is to obtain a three-dimensional molecular structure from a fully developed crystal of high quality exposed to an x ray beam of high intensity 231. The resulting diffraction pattern can be processed whereby the intensities of the spots are used to calculate a map of the electron density. Various methods can be used to improve the quality of this map until it is of sufficient clarity to allow the building of the molecular structure based on the protein sequence. The resulting structure is then refined to fit the map more accurately and to adopt a thermodynamically favored conformation.

34 Present investigation

Tissue distribution and kinetic profiles

Tissue distribution The concentration of equine GST A3-3 mRNA was measured by quantitative RT-PCR in 15 different tissues from adult horses: adrenal gland, cerebrum, heart, hypothalamus, kidney cortex, liver, lung, mammary gland, ovary, skeletal muscle, small intestine, spleen, testis, urinary bladder, and uterus (Fig. 10).

To clone the equine GSTA3 mRNA, first two pairs of primers based on human GSTA3 mRNA sequence were designed. The central coding DNA sequence (cds) of equine GSTA3 cDNA was generated from horse testis mRNA by nested RT-PCR using the designed primers. Based on the cds cDNA, additional two pairs of primers were designed to obtain 5’ and 3’ untranslated region (UTR) sequences. The 5’UTR and 3’UTR were cloned using the anchored PCR and Rapid Amplification of cDNA Ends protocols 232. Finally the entire GSTA3 coding sequence was cloned by nested RT-PCR using a third set of four primers designed on the basis of the obtained 5’ and 3’ UTR sequences.

RT-qPCR was used to quantify EcaGST A3-3 mRNA in the 15 equine tissues. The DNA double-strand specific fluorescence dye SYBR Green 233 was employed for amplicon detection. 18S rRNA was also quantitated by RT- qPCR. The amplification efficiencies of both genes were similar.

The highest levels of GSTA3 gene expression were found in the steroidogenic tissues ovary, adrenal gland, and testis (Fig. 10a); the lowest levels were several orders below those, such as in mammary gland (Fig. 10b).

Figure 10. A comparison of GSTA3 mRNA concentrations across equine tissues. Concentrations were measured in RNA samples from tissues from an adult stallion and a mare by using quantitative RT-PCR and normalized to the value obtained for mammary gland. Results from tissues with higher levels of expression (i.e., more than 550 times higher than in mammary gland) are shown in panel A. The relative levels of GSTA3 mRNA in tissues with lower levels of expression are shown in panel B. Values from left to right: ovary 27000; adrenal gland 6950; testis 5880; small intestine 2570; kidney cortex 1900; spleen 911; hypothalamus 537; uterus 261; urinary bladder 259; lung 250; heart 169; skeletal muscle 110; liver 64; cerebrum 24; mammary gland 1. Reprinted with permission from Paper I.

Kinetic profiles For kinetic measurements, the equine gene was heterologously expressed in E. coli and purified. First, before ligation, the entire coding sequence of EcaGST A3-3 gene, obtained by the nested RT-PCR procedure described above, and the pET-21a(+) plasmid were digested with EcoRI and XhoI restriction endonucleases to create sticky- and blunt-ends, respectively, in order to insert the PCR band into the plasmid in frame with a T7 RNA polymerase promoter for a more efficient expression of the inserted gene.

Electroporation was used to introduce the recombinant expression clone of EcaGST A3-3 into the E. coli BL21 (DE3). The cells were grown, lysed by sonication and centrifuged.

EcaGST A3-3 was purified from the supernatant by affinity chromatography using GSH-sepharose as stationary phase (Fig. 9). Bound EcaGST A3-3 was eluted by a buffer containing high concentration glutathione to release the enzyme from the stationary phase, and dialysed overnight and again during the day following the purification.

Product accumulation or substrate disappearance was monitored spectrophotometrically during one minute of the initial velocity phase under standard assay conditions 234 at 340 nM for CDNB and at 248 nM for the steroid substrates.

Specific activities of EcaGST A3-3 were determined with the two steroid hormone substrates Δ5-PD and Δ5-AD and several common electrophilic GST substrates. The alternative substrates include CDNB (substitution), ethacrynic acid (double-bond addition), trans-2-nonenal (double-bond addition), phenethyl isothiocyanate (PEITC, carbamoylation), sulforaphane (carbamoylation), and cumene hydroperoxide (reduction). Specific activities of DmGSTE14 were measured with Δ5-PD and Δ5-AD, several isothiocyanates, trans-2-nonenal and CDNB.

Specific activities of the equine enzyme with the traditional GST substrates are similar to those of HsaGST A3-3 with the exception of the activity with CDNB, which is 4.4-fold lower than that of the human enzyme.

Notably, the specific activity of the equine enzyme with the steroid substrate Δ5-AD are comparable to that of HsaGST A3-3, while with Δ5-PD EcaGST A3-3 is 2.5-fold more active. With that, the equine enzyme takes the place apace with HsaGST A3-3 as the most active steroid isomerase among the known GSTs.

DmGSTE14 exhibits specific activities with steroid substrates approximately twenty-fold lower than those of the equine enzyme. It should be noted however that the tested steroid substrates are not endogenous in ecdysteroid metabolism but are surrogates for a so far unknown steroid reaction in insects. Our finding implies that DmGSTE14 possesses steroid double-bond isomerization activity, but the catalysis velocity of the reaction with the yet unknown endogenous ecdysteroid is still to be determined.

In the equine enzyme and in DmGSTE14, the specific activities with the alternative endogenous substrates are less than a few percent of the steroid isomerization activities (Fig. 11), implying an evolutionary pressure on these enzymes towards participation in steroid biosynthesis rather than the traditional detoxication.

Figure 11. Substrate selectivity profiles for equine and human GSTs based on their specific activities. Blue, Δ5-AD; red, Δ5-PD; green, CDNB; lilac, PEITC; yellow, cumene hydroperoxide; grey, nonenal. For each enzyme the activities measured under standard assay conditions are expressed relative to one another. Values for HsaGST A3-3 107, HsaGST A1-1 107, HsaGST A2-2 107,235, HsaGST A4-4 107.

38 5 EcaGST A3-3 exhibits a catalytic constant (kcat) with Δ -AD 10-fold higher 5 and with Δ -PD 6.1-fold higher than that with CDNB. Km of the equine enzyme with Δ5-AD and Δ5-PD is 110- and 150-fold lower than that with

CDNB, respectively. High catalytic constants and low Kms yield catalytic efficiencies three orders of magnitude higher with steroid substrates than that with CDNB.

The catalytic efficiencies of the equine enzyme with Δ5-AD and Δ5-PD are 1.9- and 4.3-fold higher than those of HsaGST A3-3. The catalytic efficiencies of mammalian alpha class GSTs, DmGSTE14 and human 3β-HSD with the steroid substrates are illustrated in Fig. 12.

Figure 12. Catalytic efficiencies of equine GST A3-3 with the steroid substrates Δ5-AD (blue) and Δ5-PD (red) compared to human, porcine and bovine GST steroid isomerases. Also included are DmGSTE14 and human 3β-hydroxysteroid isomerase. 5 5 Catalytic efficiencies (kcat/Km) with Δ -AD and Δ -PD are given for each homodimer with standard errors (SE) where available, from left to right (mM-1 s-1): EcaGST A3-3Δ5-AD 16000 ± 1900, EcaGST A3-3Δ5-PD 14000 ± 2300; HsaGST a a b A3-3Δ5-AD 8600 ± 800 , HsaGST A3-3Δ5-PD 3200 ± 220 ; HsaGST A1-1Δ5-AD 1000 ; c c SscGST A2-2Δ5-AD 1600 ± 200 , SscGST A2-2Δ5-PD 170 ± 20 ; DmGSTE14Δ5-AD e e 280 ± 60, DmGSTE14Δ5-PD 370 ± 8; BtaGST A1-1Δ5-AD 49 , BtaGST A1-1Δ5-PD 15 ; f f Hsa3β-HSDΔ5-AD 28 , Hsa3β-HSDΔ5-PD 9 . a 107 b 180 c 108 d 196 e 181 f 236

In general, the catalytic efficiency (kcat/Km) values of eminently efficient enzymes acting on their natural substrates are in the range of 106 – 108 M-1 s-1 237. The catalytic efficiency of EcaGST A3-3 with Δ5-AD is 1.6 * 107 and with Δ5-PD 1.4 * 107 M-1 s-1. Not only do these high values make this GST the most efficient steroid isomerase known in mammals (Fig. 12), but they also place it among the most efficient enzymes in general.

40 Summary Taken together, the pattern of tissue distribution with significantly higher levels in steroidogenic than in other organs (Fig. 10), the pronounced high relative specific activity with steroid substrates compared to other endogenous substrates (Fig. 11) and the strikingly high catalytic efficiency (Fig. 12), supports the view that EcaGST A3-3 has evolved for physiologically significant steroid isomerase activity and play an imperative role in steroid hormone biosynthesis.

Similar to EcaGST A3-3, the human GST A3-3 mRNA is abundant in steroid- hormone producing organs 107,238. Steroidogenic activity of HsaGST A3-3 has been demonstrated in a human adrenal cell line where effective inhibitors of hGSTA3-3 decreased the conversion of Δ5-AD into Δ4-AD in cell lysates. In a human placental cell line, RNAi (RNA interference) targeting hGSTA3-3 expression decreased by 30% the forskolin-stimulated production of the steroid hormone progesterone 239. The convergence of tissue distribution, the high specific activities and the high catalytic efficiencies of the equine enzyme with HsaGST A3-3 strongly imply a functional analogy of the two enzymes in a physiological context.

Inhibition profile

The US Drug Collection The US Drug Collection used for screening for inhibitors of EcaGST A3-3 is an assembly of FDA-approved compounds dissolved in DMSO in 10 mM concentrations. The compounds are delivered in 96-well plates ready for utilization in experiments. This approach makes the drug collection appropriate for screenings and evaluation.

The assay To identify the most potent inhibitors of EcaGST A3-3, all 1040 compounds of the US Drug Collection were first screened for inhibition of EcaGST A3-3 with the universal substrate CDNB. The rationale for this screening is the assumption that already approved compounds could reach clinical use faster than de novo developed drug candidates. The screening revealed 13

41 compounds giving inhibition between 30% and 50% and 16 compounds giving at least 50% inhibition.

In the next step the IC50 values of the eleven most potent inhibitors identified by the screening were obtained by monitoring the enzymatic activity with CDNB and Δ5-AD with a series of inhibitor concentrations in a standard assay 234. The reaction with CDNB was followed spectrophotometrically at 340 nm and with Δ5-AD at 248 nm.

Δ5-PD is a particularly insoluble steroid in aqueous media and was not possible to dissolve at concentrations above 20 μM while maintaining the maximum solvent v/v of 5%. No inhibition measurements were therefore performed with Δ5-PD.

For determination of inhibition modality and Ki values of the most potent inhibitors, GSH and inhibitor concentrations were held constant while the concentration of the other substrate (CDNB or Δ5-AD) was varied. Due to its intrinsic inversion, the double reciprocal plot might generate inexact results and should be used with caution when aiming for exact values. However, the plot is useful for elucidation of the inhibition mechanism and approximal Ki values, as used in this study.

Anthralin, sennoside A, tannic acid and ethacrynic acid were the most potent 5 inhibitors of EcaGST A3-3 with Δ -AD yielding IC50 values of 0.085 μM, 0.11 μM, 0.22 μM and 0.25 μM, respectively (Table 2).

Table 2. IC50 values (μM) of the most potent inhibitors of EcaGST A3-3 activities with Δ5-AD (0.1 mM) in the presence of 1 mM GSH at 30 °C. Inhibitor concentrations varied between 0.0037 μM and 15 μM at pH 8.0 in 0.025 M sodium phosphate buffer. Anthralin 0.085±0.005 Sennoside A 0.11±0.008 Tannic acid 0.22±0.03 Ethacrynic acid 0.25±0.1

While the inhibition profiles of the human enzymes HsaGST P1-1 and HsaGST S2-2 have been acquired with the universal substrate CDNB 240,241, there is no such profile with the natural substrate Δ5-AD for any of the known mammalian steroid double-bond isomerases. Yet steroid hormones

42 synthesized downstream of the GST-catalyzed reactions are implicated in several aspects of health and disease. Keeping this enzyme at adequate physiological concentrations is therefore essential, and pharmacological intervention by GST inhibitors has been shown to suppress production of Δ4- AD under physiological conditions 239.

Additionally, it is beneficial to gain insight into the unsuspected inhibitory side effects on steroid hormone production that may occur when the FDA- approved drugs are employed for other pharmaceutical targets. The results of our screening of the US Drug collection identify potent inhibitors of EcaGST A3-3, which may be used to control steroid hormone production, but also help to prevent undesired side effects when these compounds are administered for other purposes.

The antipsoriatic anthralin 242 has not been detected as strong GST inhibitor previously but is the most potent inhibitor of EcaGST A3-3 in our study. Likewise, the previously not detected cathartic sennoside A 243 has proven to be a potent inhibitor of EcaGST A3-3.

Unlike anthralin and sennoside A, tannic acid and ethacrynic acid are well- characterized compounds that have been identified as GST inhibitors in previous studies 240,241,244. Ethacrynic acid is in clinical use as a diuretic 245 and tannic acid has been indicated to have potential for use as an anticancer agent 246.

Substrate binding can proceed through formation of several intermediates and conformational rearrangement steps, each of which might represent a distinct structure of the enzyme and therefore a unique opportunity for interaction with an inhibitor, such as a pharmaceutical. A suggested model 247 based on hybrid quantum/classical molecular dynamics envisages an at least three-dimensional landscape rather than the conventional plot of the standard free energy of activation in two dimensions, adding multiple enzyme conformations to the traditional reaction coordinates (Fig. 13). Reaction intermediates (the minima in the free-energy landscape) occur sequentially along the reaction coordinate axis along with the free-energy maxima that connect the reaction intermediates. Each of these fluctuations is associated with a set of enzyme conformations, induced by stochastic thermal motions of the enzyme and the substrate.

43 The vast conformational space represented by this free-energy landscape is sampled by the stochastic thermal motions of the enzyme and ligands during the reaction, and offers a variety of possibilities for interaction for such structurally diverse inhibitors as anthralin, tannic acid and ethacrynic acid (Fig. 13).

Figure 13. Schematic representation of the standard free-energy landscape for an enzyme reaction. Conformational changes occur along both axes. The conformational changes occurring along the reaction coordinate axis correspond to the environmental reorganization that facilitates the chemical reaction. In contrast, the conformational changes occurring along the ensemble conformations axis represent the sets of configurations existing at all stages along the reaction coordinate, leading to a large number of parallel catalytic pathways. Reprinted with permission from 247.

Our studies showed that anthralin binds to EcaGST A3-3 by competitive mode of inhibition. It is at this point not possible to say where on the enzyme anthralin binds, but its resemblance with steroids suggests a binding to the H- site. The distinguishing structure characteristics of this compound is the presence of adjacent polar functional groups on an anthracene moiety (Fig.

44 14), which might interact with polar residues in the hydrophobic H-site, aiding the inhibitor to the optimal orientation.

Figure 14. Structures of the four most potent inhibitors of Δ5-AD isomerization reaction catalyzed by EcaGST A3-3.

The potent inhibition provided by the four compounds can be beneficial in several ways, addressing disorders originating in excessive steroid production. One example is polycystic ovary syndrome 197. Attenuation of androgen production by decreasing steroid double-bond isomerization activity with Δ5-AD in the ovary (Fig. 6), while not providing the cure of the disease, might alleviate its symptoms.

Symptom relief might also be offered in the same way to patients suffering from androgen-producing neoplasms 197.

Another example is alleviating hypertension. A higher-than-expected prevalence of primary hyperaldosteronism has been noted among patients with hypertension in general and among those with more severe or resistant hypertension in particular 200,248. Given the high resemblance between the steroid substrates Δ5-AD and Δ5-PD and the similar high activity and catalytic efficiency between the both steroid double-bond isomerization reactions catalyzed by GST A3-3, it is reasonable to assume that the activity of this enzyme with Δ5-PD is inhibited in the same way as that with Δ5-AD. Decreasing the GST A3-3 steroid double-bond isomerization activity with Δ5-PD in adrenal cortex might lead to a drop of the aldosterone level in circulation (Fig. 4) and reduce the hypertension.

Caution should be taken when designing the administration of these potent GST steroid isomerase inhibitors. A systemic administration would affect steroid production in both adrenal gland and the gonads, leading to undesired side effects. A local administration is preferable in order to avoid interference with steroid synthesis in organs not targeted in a particular pathology.

Undesired steroid production inhibition might arise at administration of potent GST steroid isomerase inhibitors when they are aimed at their traditional targets. For example, the cathartic sennoside A, which proved to be a potent inhibitor of EcaGST A3-3, is metabolized by gut bacteria and its active anthralin-resembling metabolite rheinanthrone (Fig. 14) is absorbed into systemic circulation 249. It can be speculated that a long-term systemic administration of sennoside A at chronical constipation might affect steroid production in both the adrenal gland and the gonads. The diuretic ethacrynic acid (Fig. 14) might cause similar side effects.

Sequence similarity In analogy with all other GSTs of the alpha class, EcaGST A3-3 is postulated to be a homodimer 250. Each subunit contains 222 amino acids including the initial methionine, as do the already characterized human steroid isomerases

46 GST A3-3 and GST A1-1 107. The two porcine steroid isomerases GST A2-2 and GST A1-1 consist of 223 amino acids 108,196.

The inferred amino acid sequence of the equine GST A3-3 exhibits high similarity with that of the human GST A3-3. (Fig. 15). The sequence identity between EcaGST A3-3 and HsaGST A3-3 is 80.6% and 179 identities. In general, the features of the equine amino acid sequence agree with those characterizing the Alpha class enzymes 251.

The HsaGSTA3 G-site residues Tyr9, Arg15, Arg45, Gln54, Val55, Pro56, Gln67, Thr68, Asp101, Arg131, and Phe220 are all conserved in both the equine and the human enzymes with the exception of two conservative replacements: the positively charged Lys45 in the equine enzyme is Arg45 in the human enzyme (Fig. 15). Residue 45 in alpha class GSTs forms an ionic bond with the glycine carboxylate of the bound glutathione molecule (Fig. 2) 252. A similar conservative replacement is noted in position 68, where the equine enzyme has a Ser rather than Thr found in the human enzyme (Fig. 15). Both residues commonly occur in this position and both of them form a hydrogen bond from their hydroxyl group to the α-carboxylate of the glutamyl moiety of glutathione (Fig. 2) 253.

Neither of these conservative G-site substitutions is expected to have any major functional consequences.

On the other hand, there are more consequential variations in the amino acid residues in the H-site (Fig. 15), the subsite considered to govern the substrate specificity of the GSTs 182,250–253.

GSTA3 HORSE MAVKPMLHYF NGRGRMEPIR WLLAAAGVEF EETFIDTPED FEKLKNDGSL MFQQVPMVEI 60 GSTA3_HUMAN ..G..K...... K..GSA.. LG..R...... 60 GSTA1_BOVIN ..G..T...... C...... K..EK... LD...... 60

GSTA3_HORSE DGMKLVQSRA ILNYVAAKHN LYGKDIKERA LIDMYIEGVA DLNEMILLLP ITPPAEKDAK 120 GSTA3_HUMAN ...... T...... I.S.Y...... T..M...... LCR.E..... 120 GSTA1_BOVIN ...... T...... I.T.Y...... M...... S...... G...MHFP LC...... 120

GSTA3_HORSE IMLIKDRTTN RYLPAFEKVL KSHGEDYLVG NRLSRADIHL VELLYLVEEL DPSLLTNFPL 180 GSTA3_HUMAN .A...EK.KS ..F...... Q...Q...... K...... S...... Y.... .S..IS.... 180 GSTA1_BOVIN LT..REK...... N...... Q...... K...... Y...... A.... 180

GSTA3_HORSE LKALKARISN LPTVKKFLQP GGARKPPGDE KSVEKSRKIF KF 222 GSTA3_HUMAN .....T...... SP....A.A .AL.EA.... R. 222 GSTA1_BOVIN ...... V.. I.A...... Q....T.. .KI.EA..V. .. 222

Figure 15. Amino acid sequence alignment of equine GSTA3, human GSTA3 and bovine GSTA1. A dot indicates identity with the corresponding residue in equine GSTA3. H-site residues of human GSTA3 and corresponding residues of equine GSTA3 and bovine GSTA1 are highlighted. Human GSTA3 H-site residues critical for steroid isomerization reaction and corresponding equine and bovine residues are underlined. Reprinted with permission from Paper I.

The order of conservation between the tetrapod species compared to HsaGST A3-3 was dog (85% identical residues), horse and goat (81% identity), and gray short-tailed opossum with 64% identity.

A DmGSTE14 subunit consists of 238 amino acids and has a 21% sequence identity with HsaGST A3-3.

Mechanistic aspects The high sequence similarity between steroid isomerases in various mammalian species does not explain the differences in activity. The sequences of EcaGST A3-3 and the low-efficient steroid double bond isomerase BtaGST A1-1 exhibit high sequence identity of 81.5% with 181 identical positions – similar to that between EcaGST A3-3 and HsaGST A3-3 - but while the human and the equine GSTs are highly efficient, BtaGST A1-1 is slow 181. The explanation is probably found in the structure of the hydrophobic H-site, the G-site being conserved among species. Of the thirteen residues constituting the H-site, five (10, 12, 111, 208, and 216) have been shown to

48 be critical for the steroid isomerase activity 47,182. These five positions in the H-site display different degrees of variability among the mammalian steroid isomerases with residues 111 and 208 being the least conserved, taking on four different amino acids in each of the two positions in the seven compared mammalian enzymes (Table 3).

Table 3. H-site residues of EcaGST A3-3, HsaGST A3-3, HsaGST A1-1, HsaGST Δ5-AD A2-2, SscGST A2-2, SscGST A1-1 and BtaGST A1-1 enzymes. kcat/Km values (mM-1 s-1, monomer) for each steroid isomerase are given for reference. Reprinted with permission from Paper I. Position EcaGST HsaGST HsaGST HsaGST SscGST SscGST BtaGST A3-3 A3-3 A1-1 A2-2a A2-2c A1-1d A1-1e 10 F F F S F F F 12 G G A I G G G 14 G G G G G G G 104 E E E E E E E 107 L L L L L L M 108 L L L L L L H 110 P P P P P P P 111 I L V F L L L 208 G A M M T M T 213 V L L L L L I 216 S A A S A A A 220 F F F F F F F 222 F F F F F I F Δ5-A a b b c d e kcat/Km 8000 4300 500 1.0 800 10 24

a 107 b 180 c 108 d 196 e 181

Residue 208 seems to be central, being in the immediate neighborhood and directed towards the cyclopentane ring of the steroid (Fig. 16). In the two most efficient GSTs, position 208 harbors a residue of modest dimensions: Gly208 in EcaGST A3-3 and Ala208 in HsaGST A3-3 (Table 3). In the others, the bulky Met208 can possibly bring on steric interference or Thr208 might hydrogen bond to the keto group of the cyclopentane and distort the orientation of the steroid substrate.

However, it is probably not individual residues but combinations of residues that afford optimal substrate orientation in the H-site. In HsaGST A3-3, the steroid is coordinated in the H-site relative GSH and Tyr09 for an efficient catalysis 186.

Figure 16. Steroid binding to the H-site of GST A3-3. The figure is based on the crystal structure of the human enzyme in complex with Δ4-AD and GSH 44. Atoms ≤4.5 Å from the steroid are shown as balls in the stick representations of the amino acid residues. Three crucial residues, Phe10, Leu111, and Ala208 are indicated by green arrows. The catalytically important sulfur (yellow) of glutathione (GSH) and oxygen (red) of Tyr9 are located above C4 and C6 of the steroid skeleton. The image was created with the UCSF Chimera package (http://www.rbvi.ucsf.edu/chimera) developed by the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco (supported by NIGMS P41-GM103311). Reprinted with permission from Paper I.

Similar modeling of steroid binding to GSTA 3-3 H-site of the other tetrapods predicted low steroid double-bond isomerase activity in all three species. To facilitate future heterologous expression and subsequent measurements of enzymatic activity, the GSTA3 gene from each of the three tetrapod species was cloned into the expression vector pET-2a(+).

50 Alternate thiol cofactor In the steroid double-bond isomerization reaction catalyzed by GSTs, GSH plays the role of a cofactor providing a catalytic thiol group, rather than that of a substrate like in other GST reactions 180,182. A naturally occurring biosynthetic precursor of GSH, γ-glutamyl-cysteine (γ-Glu-Cys), was also found to support the isomerization of Δ5-AD but with a 6.5-fold lower efficiency, when the Δ5-AD concentration is varied and concentration of the cofactor GSH or γ-Glu-Cys is kept constant.

The limiting step of glutathione biosynthesis is production of the precursor γ-Glu-Cys by γ-glutamylcysteine ligase 29. The intracellular occurrence of this GSH precursor might interfere with catalysis carried out by GSTs with GSH as substrate or cofactor, as the precursor has been found to catalyze the steroid isomerization reaction with 6.5-fold lower efficiency than GSH in our study. Such interference is however unlikely, as its concentration is expected to be low due to the synthesis of precursor being the limiting step of GSH production. Furthermore, EcaGST A3-3 is probably saturated with GSH in 5 cells, as the Km for GSH with Δ -AD is approximately 0.6 mM and GSH is 29 present in millimolar range intracellularly , whereas the Km for γ-Glu-Cys was found to be 4.3 mM.

The absence of the Gly residue in the precursor might explain its lower kcat value with Δ5-AD: in humans, this residue in GSH supports catalysis by stabilizing the steroid substrate by hydrogen bonding the O3 oxygen atom 186. Not being able to provide this stabilization, γ-Glu-Cys does not contribute to an equally efficient catalysis.

The ΔΔG for the incremental contribution of the Gly residue of glutathione can be calculated based on the ratio of the kcat/Km values for the two alternative GSH γ-Glu-Cys thiol cofactors (kcat/Km / kcat/Km = 6.5) as RTln(6.5) = 1.1 kcal/mol (4.6 kJ/mol). This corresponds to the strength of a hydrogen bond 254 and is a significant contribution to overcoming the apparent barrier for the whole turnover of 15.8 kcal/mol, including three catalytical steps and two conformational rearrangement steps (Fig. 7), through all of which the steroid stabilization is provided by the Gly residue of GSH 186.

51 DmGSTE14 structure SEC was performed before the crystallization experiments to ensure a thorough purification of the enzyme, which eluted in one fraction. The crystal structure of DmGSTE14 in complex with glutathione was solved by vapour diffusion technique in a hanging drop and a complete diffraction data set at 1.3 Å resolution with one monomer of the enzyme in the asymmetric unit was collected at the UK’s national synchrotron science facility Diamond Light Source.

Despite the similarity of overall folds of DmGSTE14, HsaGST A2-2 and HsaGST A3-3, DmGSTE14 exhibits divergent active site architecture and steroid binding mode compared to the two human enzymes (Fig. 17). In particular, the structures of the C-terminal portions are different. Characteristically for the Alpha class GSTs 252,255,256, in both HsaGST A2-2 and HsaGST A3-3, the C-terminus forms an alpha helix (!9) which folds over and shields the H-site, while DmGSTE14 possesses a longer C-terminal tail which is oriented away from the active site without forming a helix. Furthermore, GSH is bound more distantly from the H-site providing additional space for steroid binding in the human GSTs. In contrast, in DmGSTE14 GSH is bound in close vicinity to the H-site occupied by 2-methyl-2,4-pentanediol (MPD) originating from the crystallization mother liquor (Fig. 17). The mode of steroid binding to DmGSTE14, with the steroid plane roughly parallel to that of the more adjacent GSH, is therefore quite divergent from the binding mode of Δ4-AD to the two human GSTs (Fig. 17).

Figure 17. Comparison of fold and active site of DmGSTE14 with those of human GST A2-2 and A3-3. DmGSTE14 (blue cartoon) in complex with GSH (pink sticks) and MPD (yellow sticks) is superposed with human GST A3-3 (light grey cartoon, PDB ID: 2vcv 44) in complex with GSH and Δ4-AD (dark grey sticks), and human GST A2-2 (brown cartoon, PDB ID: 2vct 44) in complex with Δ4-AD (brown sticks). 17β-estradiol from DmGSTE14 structure (PDB ID: 6kep) is shown as green sticks. Reprinted with permission from Paper IV.

The role of DmGSTE14 in ecdysteroidogenesis is not clarified by the demonstration of the steroid double-bond isomerization capability of this enzyme. A well-known function of GSTs is that of ligandins 105, and as such, apart from catalysis, DmGSTE14 might act as an intracellular ecdysteroid carrier.

54 Future outlook

The findings described in this thesis have added to the accumulated evidence of the role of GSTs in the synthesis of steroid hormones in human, horse, dog, gray short-tailed opossum and fruit fly. But there is more to discover and a continued exploration of steroidogenic GSTs in mammals is of considerable interest. The horse possesses nine intact Alpha class GST, the largest number in the 21 mammals studied (Table 1). A possibility of the existence of more steroidogenic GSTs among these can be investigated, as well as continued investigation in other mammalian species. Extended analysis of steroid double-bond isomerization of Δ5-PD, leading to a better understanding of corticosteroid biosynthesis, is a part of this exploration.

Unintended suppression of steroidogenesis is not limited to side effects of the FDA-approved compounds. Synthetically manufactured chemicals and environmental pollutants might exert harmful effects on the endocrine system, and it is desirable to more exactly elucidate their mechanisms, in particular whether steroidogenic GSTs are inhibited by these compounds.

The search for the endogenous substrate of DmGSTE14 and the role of this enzyme in ecdysteroid synthesis should be pursued.

The “stickiness” of EcaGST A3-3, manifested as its pronounced adsorption to the inner walls of the test tubes in our experiments, might imply that the enzyme is relatively hydrophobic and exists in some form of membrane association in vivo. The dehydrogenation of the C3 hydroxyl group that precedes the double-bond isomerization is catalyzed by the membrane-bound 3β-HSD 161, and it is reasonable to assume that the two enzymes might coexist in close proximity or in a complex and thus both be membrane-associated. Two microsome-associated Alpha-class GSTs from sheep and human liver have been characterized 257,258, supporting this assumption. Further investigations of the intracellular localization of GST steroid isomerases are necessary in order to acquire a deeper understanding of the efficient catalysis performed by EcaGST A3-3 and HsaGST A3-3.

An animal model is needed to study the regulation of GSTA3 gene expression and enzyme activity in relation to steroid biosynthesis and various diseases. Rodents prefer the Δ4 pathway of steroid biosynthesis unlike most domestic animals and man, which use the Δ5 pathway 164. Therefore the discovery of the steroid isomerase A3-3 in the stallion testis opens new possibilities to investigations in molecular endocrinology with relevance to both veterinary and human medicine.

56 Concluding remarks

Steroids regulate a versatile array of aspects of life, from reproduction, inflammation and water homeostasis to mood and wellbeing. Enzymes employed in steroidogenesis are important potential drug targets, and the discovery of the detoxication enzymes GSTs being involved in the synthesis of progesterone and the last precursor of testosterone makes these enzymes interesting for intense research. We have studied steroidogenic GSTs in horse, dog, goat, gray short-tailed opossum and fruit fly and have come to the following conclusions:

• The high steroid isomerase activity of EcaGST A3-3 revealed by this study supports the view that the GST family possesses a versatile function assortment not limited to detoxication. The EcaGST A3-3 enzyme now ranks as the most efficient steroid isomerase known in mammals together with HsaGST A3-3.

• The presence of a highly efficient GST steroid isomerase in Equus ferus caballus together with HsaGST A3-3 and somewhat less efficient SscGST A1-1 points to physiological significance of GST steroid isomerases in large mammals.

• EcaGST A3-3 mRNA tissue distribution further accentuates the significance of this enzyme for steroidogenesis. The equine GSTA3 gene transcript levels were particularly pronounced in steroidogenic organs such as testis, ovary, and the adrenal gland in the horse as well as in humans 107. Even though it is well recognized that mRNA expression is not necessarily a definite measure of the corresponding protein concentration 259, the high GSTA3 mRNA concentrations are strong indicators of an involvement in steroidogenesis.

• We have identified FDA-approved compounds already in clinical and other use as potent submicromolar-range-inhibitors of the steroid

57 double-bond isomerization catalyzed by EcaGST A3-3. Given the important role this enzyme may play in steroidogenesis, it can be a potential pharmaceutical target and the identified compounds can be used as pharmaceuticals in treatment of endocrine disorders. Conversely, the screening highlights possible undesired side effects on steroidogenesis exerted by these compounds when they are administered for other purposes.

• The high degree of conservation of GST A3-3 amino acid sequences was demonstrated between human compared to horse, dog, goat and gray short-tailed opossum. The availability of GST A3-3 clones from the three latter species in the expression vector pET-21a(+) facilitates further research to potentially evaluate steroid isomerases activity of this enzyme in these species and its susceptibility to inhibition by relevant compounds.

• As an initial step of determination of DmGSTE14’s role in ecdysteroidogenesis we have showed that GSTE14 from D. melanogaster possesses a steroid double-bond isomerization reaction activity with Δ5-AD and Δ5-PD and binds Δ5-AD in a distinct way compared to the human GST A2-2 and GST A3-3.

We hope that these findings will lead to further insights into the role of glutathione transferases in the synthesis of the vital steroid hormones in humans, animals and insects, and to potential therapies of steroid dependent pathologies.

The role of GST steroid isomerase in neurodegeneration The neurosteroid progesterone (Δ4-PD) is produced in the reaction catalyzed by EcaGST A3-3. Furthermore, several of the neurosteroids are metabolites of the intermediates synthesized by EcaGST A3-3. For example the neurosteroid allopregnanolone is downstream of Δ4-PD, and estradiol is downstream of both Δ4-PD and Δ4-AD. Research suggests that progesterone promotes myelination and dendritic growth, allopregnanolone increases hippocampal neurogenesis and estradiol regulates synaptic plasticity 168,260,261. Equine and other mammalian GST steroid double-bond isomerases might therefore play an important role in many physiological and pathophysiological

58 conditions associated with the nervous system and can be potential targets in neurodegeneration cure and prevention.

60 Populärvetenskaplig sammanfattning

Enzymer är katalysatorer av biokemiska reaktioner. De är cellens maskineri som driver livsprocesserna framåt, de spelar den välkoordinerade, samstämmiga symfoni som utgör den biokemiska grunden till vår existens.

Enzymer har många diversifierade funktioner. Vissa spjälkar näringsämnen för att cellen ska tillgodogöra sig energin från vår föda, andra avläser DNA och bidrar därigenom till produktionen av nya enzymer, åter andra deltar i celldöden för att cellförnyelsen ska pågå ostört.

Glutationtransferaser Glutationtransferaser (GSTer) är en familj av enzymer som finns i stort sett i alla organismer, och i däggdjur i alla organ. De är främst kända för sin avgiftningsfunktion då de katalyserar reaktionen mellan antioxidanten glutation och svårlösliga, reaktiva, toxiska molekyler som kan vara främmande för organismen, t ex kemikalier, men som även kan vara kroppsegna (endogena) metaboliter från ämnesomsättningen. Toxinerna reagerar lätt med proteiner och DNA, vilket kan leda till patologier såsom cancer. Genom GSTernas katalyserade konjugering med den lösliga glutation blir toxinerna mindre reaktiva och mera lättlösliga och kan avlägsnas från organismen i urinen.

Toxinet och glutation binder till det aktiva sätet i enzymet så att de hamnar i omedelbar närhet av varandra. Enzymet orienterar båda substraten så att katalysen underlättas och snabbas upp.

En del GSTer har utvecklat andra förmågor förutom eller istället för avgiftningsfunktionen. Ett exempel är katalys av reaktioner som ingår i steroidgenesen.

61 Steroider Steroider är hormoner som påverkar flera aspekter av vår hälsa, från reproduktion och blodtryck till inflammation och psykiskt välmående. Steroidgenes är en process som består av flera steg där varje steg producerar ett steroidhormon eller en metabolit och katalyseras av ett specifikt enzym. Genesen leder till att könssteroider avsöndras i steroidgena celler i testiklar och äggceller, och kortikosteroider i binjurebarken. Könssteroiderna påverkar bl a sekundära könskaraktäristika och graviditet. Binjurebarkens steroider påverkar bl a inflammation och vätskejämvikt i kroppen; vätskejämvikten i sin tur har effekt på blodtrycket.

Två av steroidhormonerna är progesteron och testosteron. Progesteron är främst ett kvinnligt könshormon med betydelse för fertilitet och kvinnans sekundära könskaraktäristika. Även embryogenes påverkas av progesteron. Testosteron är det främsta manliga könshormonet som påverkar könsdrift samt manliga sekundära könskaraktäristika.

GSTernas roll i steroidgenes Studier har visat att en av GSTerna, GST A3-3, i människa och i lägre utsträckning i gris katalyserar steg som producerar steroiden progesteron samt det omedelbara förstadiet till testosteron. Reaktionen är av typen isomerisering av steroid-dubbelbindning och enzymer som katalyserar denna typ av reaktion kallas steroid-dubbelbindning-isomeraser eller steroidisomeraser. Denna katalys har traditionellt tillskrivits ett annat enzym och för att GST A3-3 ska betraktas som ett fullvärdigt steroidgenererande enzym behövs mer evidens. Katalyseffektiviteten i människa är den högsta bland kända steroidisomeraserna i däggdjur idag, och förekomsten av detta enzym med hög katalyseffektivitet i steroidgena organ i ytterligare stora däggdjur skulle styrka hypotesen om att enzymet har utvecklats till att utföra just denna funktion.

Ett ekvint (häst-)enzym hade upptäckts som skulle kunna vara ett GST A3-3. Det behövdes nu en undersökning av dess organfördelning och mätningar av dess katalytiska effektivitet. Dessutom var det intressant att undersöka en eventuell förekomst av samma enzym i andra däggdjur. Ett bovint (ko-)enzym hade undersökts tidigare men uppvisat endast en låg aktivitet.

Kunskap om enzymer öppnar flera möjligheter till praktisk tillämpning. En är enzymhämning för att bota patologier och lindra symptom. Flera patologiska tillstånd beror på överproduktion av steroider, t ex adrenogenitalt syndrom och polycystiskt ovariesyndrom. Dessutom stimulerar steroidhormoner tillväxt av neoplastiska celler i prostata- och bröstcancer. Därför finns det anledning till en potentiell minskning av steroidhormonproduktion, t ex genom en hämning av GST A3-3.

Vår forskning

Det ekvina enzymet renades från 15 olika organ och enzymets katalytiska effektivitet mättes. Det uppvisade en hög grad av evolutionär konservering med 81% identitet (identiska aminosyror) vid en jämförelse med det humana enzymet. Enzymet förekom främst i de steroidgena organen testiklar, äggceller och binjurar och dess katalytiska effektivitet visade sig vara lika hög som i människa. Detta gör det humana och det ekvina GST A3-3 till de mest effektiva steroidisomeraserna kända bland däggdjur idag.

Ett bibliotek av 1040 ämnen godkända av FDA screenades för hämning av ekvina GST A3-3s katalys av reaktionen med förstadiet till testosteron. Fördelen med ett sådant bibliotek är att ämnena redan är utredda för biverkningar och att tiden från experiment till en potentiell introduktion på marknaden är betydligt kortare för ett FDA-godkänt ämne jämfört med ett helt outrett ämne eftersom ett stort antal tester har redan gjorts. Tretton ämnen uppvisade en hämning mellan 30% och 50% av aktiviteten och 16 ämnen uppvisade en hämning på minst 50%. Av dessa var de fyra mest potenta hämmarna antralin - ett antiinflammatoriskt medel för behandling av dermatoser, särskilt psoriasis; laxermedlet Sennosid A; tannin – ett proteinsammandragande ämne från växter, samt det urindrivande medlet etakrynsyra.

Medan dessa hämmare har ett potentiellt användningsområde som läkemedel gäller även det omvända: att de, när de används i sitt traditionella syfte, eventuellt kan ha en biverkning i form av en hämning av testosteronproduktion, vilket kan ge oönskad effekt på reproduktion och sekundära könskaraktäristika.

Det kan antas att dessa ämnen har en hämmande effekt även på GST A3-3s katalys av progesteron. Eftersom progesteron är ett förstadium till flera steroider, bl a kortikosteroider, kan hämmarna eventuellt användas för att dämpa GST A3-3s aktivitet i produktionen av även denna typ av steroider. Mer forskning behövs för att närmare undersöka och kvantifiera hämmarnas effekt på syntes av progesteron och dess metaboliter.

För att bredda utforskningen till andra däggdjur bestämdes aminosyrasammansättningen i GST A3-3 från testiklar från hund, get och nordamerikansk opossum. Som väntat var graden av den evolutionära konserveringen mellan arterna hög: hund hade 85% identitet (identiska aminosyror) med det humana GST A3-3 och get (liksom häst) 81% identitet. Nordamerikansk opossum hade den lägsta graden av evolutionär konservering med 64% identitet.

Genkloner av GST A3-3 från dessa arter framställdes för att möjliggöra framtida mätningar av enzymets katalytiska effektivitet och andra biokemiska parametrar.

För ytterligare breddning av utforskningen till insekter undersöktes steroidisomeriseringen i bananfluga. Steroider i insekter skiljer sig från däggdjurens och isomeriseringen katalyseras av andra, insektspecifika enzymer. Ekdysteroider i insekterna reglerar hudömsning då djuret byter exoskelett. Processen i bananfluga är inte undersökt i detalj och alla de separata stegen är inte kända. GSTE14 från bananfluga deltar i hudömsningen, dock är det okänt vilket steg den katalyserar. Likaså är substratet okänt.

GSTE14s katalytiska effektivitet mättes med förstadier till progesteron och testosteron. Dessa substrat förekommer inte i insekter men mätningarna ger en uppfattning om enzymets eventuella förmåga till steroidisomerisering. GSTE14 uppvisade måttlig katalys av steroidisomerisering av dessa substrat. Det återstår att upptäcka det endogena steroidsubstratet för denna reaktion i bananfluga.

GSTE14s aminosyrasekvens är känd sedan tidigare och uppvisar en identitet med det humana GST A3-3 på 21%, vilket är lägre än konserveringen mellan de undersökta däggdjuren. Det är inte förvånande då insekter evolutionärt är mer avlägsna från däggdjur än däggdjuren är sinsemellan.

För att få en detaljerad uppfattning om den tredimensionella strukturen utfördes en kristallisering av enzymet. Det framgick tydligt efter en analys av det kristalliserade enzymet att substratet binder till GSTE14 på ett annorlunda sätt jämfört med hur det binder till det humana GST A3-3, vars struktur har lösts med kristallisering tidigare. I det humana GST A3-3 binder substratet längre in i det aktiva sätet än i GSTE14. Den stora strukturella skillnaden var enzymets ena ände, som inte anslöt till det övriga enzymet på samma sätt som i det humana GST A3-3 utan var orienterad bort från resten av enzymet, vilket kan påverka katalysen. Dock kan det eventuellt vara fördelaktigt med en sådan struktur för katalys av reaktionen med det endogena substratet.

66 Acknowledgements

My supervisor Bengt Mannervik, thank you for letting me be a part of your research group. These years have been a delightful journey full of discoveries and hard work, there were joys and sometimes the unavoidable disappointments of research. Your trust and encouragement have been the solid support of my work. Thank you for your dedication to curiosity and for always being so calm!

Birgitta, for sharing your knowledge and experience so generously. Any time I had a problem in the lab you could help. Thank you for your sense of humor and for our lunches on campus.

My former and present colleagues in this group! Aslam, for the great time in our office and for getting me started when I was new in the lab. Yaman, for your kind and cheerful attitude always. Urvi for your persistence in our work and the happiness around you and for the great cooking! Aram, you are a privilege of a co-worker! And you are great with the students! Our Bachelor and Master students André, Thanh, Maria, Pedram, Shanshan, Zou, Libby – you are great contributions to the lab! I have learned a lot from you.

In the department, Anders for being such a great supervisor when I did my Master exam project in your lab. Marie-Louise and Sylvia for all the support and advice! The secretariat, Matt and Peter, for rescuing us, every day. And all the PhD students, teachers and staff in the department for creating the cheerful and supportive spirit that makes it so enjoyable to work here.

My former supervisors Peter Brzezinski, Linda Näsvik Öjemyr at DBB and Christoph von Ballmoos at Universität Bern, for helping me find my way in the lab and in the wonderful science of biochemistry.

To my amazing friends and family! I am so fortunate to have you. The good times and the hard times we share, even if some of you are far away. It is a treat just to know that you exist!

Slutligen – Felix, William and Tom, the joy of my life. This thesis would never happen without you. I have said it in the past and I will say it again a thousand times: you are the foundation of everything I do. Thank you for supporting me in all possible ways. Jag älskar er.

68 The question of curiosity

Why do we do all this? During the years of my doctorate I have under numerous occasions described the research I was engaged in to people from a wide range of occupations and backgrounds. All of them found it exciting and most of them asked the question “Yes but what is it all for?”

And so again I found myself defending basic research.

There are practical benefits of our research, described in this thesis. And while it has been utterly satisfactory to know that our effort eventually was for the benefit of man and animal, the main joy and driving force for me has been curiosity.

Curiosity drove me from a fulfilling job in the field of IT to the first lecture of the Bachelor program in chemistry at Stockholm University and all the way to my desk at the Department of Biochemistry and Biophysics where I am now writing this thesis. And I have been surrounded by people likewise driven by the desire to contribute and to improve – and by curiosity.

But curiosity is questioned and if you are not able to explain how your research can lead to improvements there will be a disappointment. And probably no grants.

Fair enough. We do not want to be wasteful and so curiosity is questioned.

But questions that curiosity asks deserve respect and attention. The practical goals we are able to set up now are based on and limited by our knowledge. The whys and the hows of curiosity might lead to yet unimagined improvements.

And – still! - what a delight just to know! Curiosity might or might not have killed the cat, but what we do know for certain is that knowledge is power. And joy.

70 References

1. Bassil, N., Alkaade, S. & Morley, J. E. The benefits and risks of testosterone replacement therapy: a review. Ther. Clin. Risk Manag. 5, 427–448 (2009). 2. Hara, Y., Waters, E. M., McEwen, B. S. & Morrison, J. H. Estrogen effects on cognitive and synaptic health over the lifecourse. Physiol. Rev. 95, 785– 807 (2015). 3. Miller Andrew P. et al. Estrogen modulates inflammatory mediator expression and neutrophil chemotaxis in injured arteries. Circulation 110, 1664–1669 (2004). 4. Nadkarni, S., Cooper, D., Brancaleone, V., Bena, S. & Perretti, M. Activation of the annexin A1 pathway underlies the protective effects exerted by estrogen in polymorphonuclear leukocytes. Arterioscler. Thromb. Vasc. Biol. 31, 2749–2759 (2011). 5. Riggs, B. L., Khosla, S. & Melton, L. J. A unitary model for involutional osteoporosis: estrogen deficiency causes both type I and type II osteoporosis in postmenopausal women and contributes to bone loss in aging men. J. Bone Miner. Res. 13, 763–773 (1998). 6. Conley, A. J. & Bird, I. M. The role of cytochrome P450 17α-hydroxylase and 3β-hydroxysteroid dehydrogenase in the integration of gonadal and adrenal steroidogenesis via the Δ5 and Δ4 pathways of steroidogenesis in mammals. Biol. Reprod. 56, 789–799 (1997). 7. Ing, N. H. et al. Dexamethasone acutely down-regulates genes involved in steroidogenesis in stallion testes. J. Steroid Biochem. Mol. Biol. 143, 451– 459 (2014). 8. Ing, N. H. et al. Dexamethasone acutely regulates endocrine parameters in stallions and subsequently affects gene expression in testicular germ cells. Anim. Reprod. Sci. 152, 47–54 (2015). 9. Boyland, E. & Chasseaud, L. F. The role of glutathione and glutathione S- transferases in mercapturic acid biosynthesis. Adv Enzym. Relat Areas Mol Biol 32, 173–219 (1969). 10. Berhane, K., Widersten, M., Engström, A., Kozarich, J. W. & Mannervik, B. Detoxication of base propenals and other alpha, beta- unsaturated aldehyde products of radical reactions and lipid peroxidation by human glutathione transferases. Proc. Natl. Acad. Sci. 91, 1480–1484 (1994). 11. Ma, Q. & He, X. Molecular basis of electrophilic and oxidative defense:

71 promises and perils of Nrf2. Pharmacol. Rev. 64, 1055–1081 (2012). 12. Xu, C., Li, C. Y.-T. & Kong, A.-N. T. Induction of phase I, II and III drug metabolism/transport by xenobiotics. Arch. Pharm. Res. 28, 249– 268 (2005). 13. Meunier, B., de Visser, S. P. & Shaik, S. Mechanism of oxidation reactions catalyzed by cytochrome P450 enzymes. Chem. Rev. 104, 3947–3980 (2004). 14. Josephy, P. David. & Mannervik, B. Molecular Toxicology, pp. 333- 401. (Oxford University Press, 2006). 15. Chin, K.-V., Pastan, I. & Gottesman, M. M. Function and regulation of the human multidrug resistance gene. in Advances in Cancer Research (eds. Vande Woude, G. F. & Klein, G.) vol. 60 157–180 (Academic Press, 1992). 16. Mashiyama, S. T. et al. Large-scale determination of sequence, structure, and function relationships in cytosolic glutathione transferases across the biosphere. PLoS Biol. 12, 1–19 (2014). 17. Ketterman, A. J., Saisawang, C. & Wongsantichon, J. Insect glutathione transferases. Drug Metab. Rev. 43, 253–265 (2011). 18. Cummins, I., Dixon, D. P., Freitag-Pohl, S., Skipsey, M. & Edwards, R. Multiple roles for plant glutathione transferases in xenobiotic detoxification. Drug Metab. Rev. 43, 266–280 (2011). 19. Allocati, N., Federici, L., Masulli, M. & Ilio, C. D. Glutathione transferases in bacteria. FEBS J. 276, 58–75 (2009). 20. Edwards, R. & Dixon, D. P. Plant Glutathione Transferases. in Methods in Enzymology (eds. Sies, H. & Packer, L.) vol. 401 169–186 (Academic Press, 2005). 21. Combes, B. & Stakelum, G. S. A liver enzyme that conjugates sulfobromophtalein sodium with glutahione. J. Clin. Invest. 40, 981–988 (1961). 22. Booth, J., Boyland, E. & Sims, P. An enzyme from rat liver catalysing conjugations with glutathione. Biochem. J. 79, 516–524 (1961). 23. Hayes, J. D., Flanagan, J. U. & Jowsey, I. R. Glutathione transferases. Annu. Rev. Pharmacol. Toxicol. 45, 51–88 (2004). 24. Eaton, D. L. & Bammler, T. K. Concise review of the glutathione S- transferases and their significance to toxicology. Toxicol. Sci. 49, 156– 164 (1999). 25. Mannervik, B., Danielson, U. H. & Ketterer, B. Glutathione transferases—structure and catalytic activity. Crit. Rev. Biochem. 23, 283–337 (1988). 26. Josephy, P. David. & Mannervik, B. Molecular Toxicology, pp. 351- 364. (Oxford University Press, 2006). 27. Board, P. G. & Menon, D. Glutathione transferases, regulators of cellular metabolism and physiology. Biochim. Biophys. Acta 1830, 3267–3288 (2013). 28. Reed, D. J. Glutathione: toxicological implications. Annu. Rev.

72 Pharmacol. Toxicol. 30, 603–631 (1990). 29. Josephy, P. D. & Mannervik, B. Molecular Toxicology, pp. 303-332. (Oxford University Press, 2006). 30. Marí, M. et al. Mitochondrial glutathione: features, regulation and role in disease. Biochim. Biophys. Acta BBA - Gen. Subj. 1830, 3317–3328 (2013). 31. Lu, S. C. Review: Regulation of glutathione synthesis. Mol. Aspects Med. 30, 42–59 (2009). 32. Townsend, D. M., Tew, K. D. & Tapiero, H. The importance of glutathione in human disease. Biomed. Pharmacother. 57, 145–155 (2003). 33. Oakley, A. Glutathione transferases: a structural perspective. Drug Metab. Rev. 43, 138–151 (2011). 34. Wu, B. & Dong, D. Human cytosolic glutathione transferases: structure, function, and drug discovery. Trends Pharmacol. Sci. 33, 656–668 (2012). 35. Aniya, Y. & Imaizumi, N. Mitochondrial glutathione transferases involving a new function for membrane permeability transition pore regulation. Drug Metab. Rev. 43, 292–299 (2011). 36. Robinson, A., Huttley, G. A., Booth, H. S. & Board, P. G. Modelling and bioinformatics studies of the human Kappa-class glutathione transferase predict a novel third glutathione transferase family with similarity to prokaryotic 2-hydroxychromene-2-carboxylate isomerases. Biochem. J. 379, 541–552 (2004). 37. Jakobsson, P.-J., Morgenstern, R., Mancini, J., Ford-Hutchinson, A. & Persson, B. Common structural features of mapeg—a widespread superfamily of membrane associated proteins with highly divergent functions in eicosanoid and glutathione metabolism. Protein Sci. 8, 689– 692 (1999). 38. Mannervik, B. et al. Nomenclature for human glutathione transferases. Biochem. J. 282, 305–306 (1992). 39. Mannervik, B., Board, P. G., Hayes, J. D., Listowsky, I. & Pearson, W. R. Nomenclature for Mammalian Soluble Glutathione Transferases. in Methods in Enzymology vol. 401 1–8 (Elsevier, 2005). 40. Tian, R., Seim, I., Ren, W., Xu, S. & Yang, G. Contraction of the ROS scavenging enzyme glutathione S-transferase gene family in cetaceans. G3 Genes Genomes Genet. g3.400224.2019 (2019) doi:10.1534/g3.119.400224. 41. Josephy, P. David. & Mannervik, B. Molecular Toxicology, p. 336. (Oxford University Press, 2008). 42. Armstrong, R. N. Structure, catalytic mechanism, and evolution of the glutathione transferases. Chem. Res. Toxicol. 10, 2–18 (1997). 43. Abdalla, A.-M., Bruns, C. M., Tainer, J. A., Mannervik, B. & Stenberg, G. Design of a monomeric human glutathione transferase GSTP1, a structurally stable but catalytically inactive protein. Protein Eng. Des.

73 Sel. 15, 827–834 (2002). 44. Tars, K., Olin, B. & Mannervik, B. Structural basis for featuring of steroid isomerase activity in Alpha class glutathione transferases. J. Mol. Biol. 397, 332–340 (2010). 45. Board, P. G. et al. Identification, characterization, and crystal structure of the Omega class glutathione transferases. J. Biol. Chem. 275, 24798– 24806 (2000). 46. Nilsson, L. O., Gustafsson, A. & Mannervik, B. Redesign of substrate- selectivity determining modules of glutathione transferase A1–1 installs high catalytic efficiency with toxic alkenal products of lipid peroxidation. Proc. Natl. Acad. Sci. 97, 9408–9412 (2000). 47. Pettersson, P. L., Johansson, A.-S. & Mannervik, B. Transmutation of human glutathione transferase A2-2 with peroxidase activity into an efficient steroid isomerase. J. Biol. Chem. 277, 30019–30022 (2002). 48. Josephy, P. David. & Mannervik, B. Molecular Toxicology, p. 343. (Oxford University Press, 2008). 49. Mohana, K. & Achary, A. Human cytosolic glutathione-S-transferases: quantitative analysis of expression, comparative analysis of structures and inhibition strategies of isozymes involved in drug resistance. Drug Metab. Rev. 49, 318–337 (2017). 50. Wang, W. & Ballatori, N. Endogenous glutathione conjugates: occurrence and biological functions. Pharmacol. Rev. 50, 335–356 (1998). 51. Hayes, J. D. & Pulford, D. J. The glutathione S-transferase supergene family part I: regulation of GST and the contribution of the isoenzymes to cancer chemoprotection and drug resistance part I. Crit. Rev. Biochem. Mol. Biol. 30, 445–520 (1995). 52. Hayes, J. D. & Pulford, D. J. The glutathione S-transferase supergene family part II: regulation of GST and the contribution of the isoenzymes to cancer chemoprotection and drug resistance part II. Crit. Rev. Biochem. Mol. Biol. 30, 521–600 (1995). 53. Chasseaud, L. F. The Role of Glutathione and Glutathione S- Transferases in the Metabolism of Chemical Carcinogens and Other Electrophilic Agents. in Advances in Cancer Research vol. 29 175–274 (Elsevier, 1979). 54. Barsukova, M. E., Veselova, I. A. & Shekhovtsova, T. N. Main methods and approaches to the determination of markers of oxidative stress— organic peroxide compounds and hydrogen peroxide. J. Anal. Chem. 74, 425–436 (2019). 55. Prohaska, J. R. The glutathione peroxidase activity of glutathione S- transferases. Biochim. Biophys. Acta BBA - Enzymol. 611, 87–98 (1980). 56. Castro, J. P., Jung, T., Grune, T. & Siems, W. 4-Hydroxynonenal (HNE) modified proteins in metabolic diseases. Free Radic. Biol. Med. 111, 309–315 (2017). 57. Negre-Salvayre, A. et al. Pathological aspects of lipid peroxidation.

74 Free Radic. Res. 44, 1125–1171 (2010). 58. Eckl, P. M. Genotoxicity of HNE. Mol. Aspects Med. 24, 161–165 (2003). 59. Schaur, R. J. Basic aspects of the biochemical reactivity of 4- hydroxynonenal. Mol. Aspects Med. 24, 149–159 (2003). 60. Siems, W. & Grune, T. Intracellular metabolism of 4-hydroxynonenal. Mol. Aspects Med. 24, 167–175 (2003). 61. Csala, M. et al. On the role of 4-hydroxynonenal in health and disease. Biochim. Biophys. Acta BBA - Mol. Basis Dis. 1852, 826–838 (2015). 62. Balogh, L. M. et al. Substrate specificity combined with stereopromiscuity in glutathione transferase A4-4-dependent metabolism of 4-hydroxynonenal. Biochemistry 49, 1541–1548 (2010). 63. Balogh, L. M., Roberts, A. G., Shireman, L. M., Greene, R. J. & Atkins, W. M. The stereochemical course of 4-hydroxy-2-nonenal metabolism by glutathione S-transferases. J. Biol. Chem. 283, 16702–16710 (2008). 64. Seidegård, J. & Ekström, G. The role of human glutathione transferases and epoxide hydrolases in the metabolism of xenobiotics. Environ. Health Perspect. 105, 791–799 (1997). 65. Kemper, R. A., Myers, S. R. & Hurst, H. E. Detoxification of vinyl carbamate epoxide by glutathione: evidence for participation of glutathione S-transferases in metabolism of ethyl carbamate. Toxicol. Appl. Pharmacol. 135, 110–118 (1995). 66. Muñoz, P., Paris, I. & Segura-Aguilar, J. Commentary: evaluation of models of Parkinson’s disease. Front. Neurosci. 10, (2016). 67. Segura-Aguilar, J. Aminochrome as preclinical model for Parkinson’s disease. Oncotarget 8, 45036–45037 (2017). 68. Dagnino-Subiabre, A. et al. Glutathione transferase M2-2 catalyzes conjugation of dopamine and dopa- o-quinones. Biochem. Biophys. Res. Commun. 274, 32–36 (2000). 69. Cuevas, C. et al. Glutathione transferase-M2-2 secreted from glioblastoma cell protects SH-SY5Y cells from aminochrome neurotoxicity. Neurotox. Res. 27, 217–228 (2015). 70. Fahey, J. W., Zalcmann, A. T. & Talalay, P. Corrigendum to “The chemical diversity and distribution of glucosinolates and isothiocyanates among plants” [Phytochemistry 56 (2001) 5–51]. Phytochemistry 59, 237 (2002). 71. Fahey, J. W., Zalcmann, A. T. & Talalay, P. The chemical diversity and distribution of glucosinolates and isothiocyanates among plants. Phytochemistry 56, 5–51 (2001). 72. Yuan, J.-M. et al. 2-phenethyl isothiocyanate, glutathione S-transferase M1 and T1 polymorphisms, and detoxification of volatile organic carcinogens and toxicants in tobacco smoke. Cancer Prev. Res. (Phila. Pa.) 9, 598–606 (2016). 73. Guerrero-Beltrán, C. E., Calderón-Oliver, M., Pedraza-Chaverri, J. & Chirino, Y. I. Protective effect of sulforaphane against oxidative stress:

75 Recent advances. Exp. Toxicol. Pathol. 64, 503–508 (2012). 74. Riedl, M. A., Saxon, A. & Diaz-Sanchez, D. Oral sulforaphane increases Phase II antioxidant enzymes in the human upper airway. Clin. Immunol. 130, 244–251 (2009). 75. Wagner, A. E., Boesch-Saadatmandi, C., Dose, J., Schultheiss, G. & Rimbach, G. Anti-inflammatory potential of allyl-isothiocyanate – role of Nrf2, NF-κB and microRNA-155. J. Cell. Mol. Med. 16, 836 (2012). 76. Yanaka, A. et al. Dietary sulforaphane-rich broccoli sprouts reduce colonization and attenuate gastritis in helicobacter pylori–Infected mice and humans. Cancer Prev. Res. (Phila. Pa.) 2, 353–360 (2009). 77. Fahey, J. W. et al. Sulforaphane inhibits extracellular, intracellular, and antibiotic-resistant strains of Helicobacter pylori and prevents benzo[a]pyrene-induced stomach tumors. Proc. Natl. Acad. Sci. 99, 7610–7615 (2002). 78. Kolm, R. H., Danielson, U. H., Zhang, Y., Talalay, P. & Mannervik, B. Isothiocyanates as substrates for human glutathione transferases: structure-activity studies. Biochem. J. 311, 453–459 (1995). 79. Clark, A. G., Smith, J. N. & Speir, T. W. Cross-specificity in some vertebrate and insect glutathione-transferases with methyl parathion (dimethyl p-nitrophenyl phosphorothionate), 1-chloro-2,4- dinitrobenzene and S-crotonyl-N-acetylcysteamine as substrates. Biochem. J. 135, 385–392 (1973). 80. PubChem. Benzo[a]pyrene. https://pubchem.ncbi.nlm.nih.gov/compound/2336. 81. Hecht, S. S. Tobacco smoke carcinogens and lung cancer. JNCI J. Natl. Cancer Inst. 91, 1194–1210 (1999). 82. Shiizaki, K., Kawanishi, M. & Yagi, T. Modulation of benzo[a]pyrene– DNA adduct formation by CYP1 inducer and inhibitor. Genes Environ. 39, (2017). 83. Robertson, I. G. C., Guthenberg, C., Mannervik, B. & Jernström, B. Differences in stereoselectivity and catalytic efficiency of three human glutathione transferases in the conjugation of glutathione with 7β,8α- dihydroxy-9α,10α-oxy-7,8,9,10-tetrahydrobenzo(a)pyrene. Cancer Res. 46, 2220–2224 (1986). 84. Volk, D. E. et al. Solution structure of a cis-opened (10R)-N6- deoxyadenosine adduct of (9S,10R)-9,10-epoxy-7,8,9,10- tetrahydrobenzo[a]pyrene in a DNA duplex. Biochemistry 42, 1410– 1420 (2003). 85. Allocati, N., Masulli, M., Ilio, C. D. & Federici, L. Glutathione transferases: substrates, inihibitors and pro-drugs in cancer and neurodegenerative diseases. Oncogenesis 7, 1–15 (2018). 86. Jardim, B. V. et al. Glutathione transferase pi (GSTpi) expression in breast cancer: An immunohistochemical and molecular study. Acta Histochem. 114, 510–517 (2012). 87. Tew, K. D. et al. The role of glutathione S-transferase P in signaling

76 pathways and S-glutathionylation in cancer. Free Radic. Biol. Med. 51, 299–313 (2011). 88. Ricci, G. et al. Glutathione transferase P1-1: self-preservation of an anticancer enzyme. Biochem. J. 376, 71–76 (2003). 89. Munoz, M., Henderson, M., Haber, M. & Norris, M. Role of the MRP1/ABCC1 multidrug transporter protein in cancer. IUBMB Life 59, 752–757 (2007). 90. Eklund, B. I., Moberg, M., Bergquist, J. & Mannervik, B. Divergent activities of human glutathione transferases in the bioactivation of azathioprine. Mol. Pharmacol. 70, 747–754 (2006). 91. Zhang, W., Modén, O., Tars, K. & Mannervik, B. Structure-based redesign of GST A2-2 for enhanced catalytic efficiency with azathioprine. Chem. Biol. 19, 414–421 (2012). 92. Bhutani, D. & Vaishampayan, U. N. Monoclonal antibodies in oncology therapeutics: present and future indications. Expert Opin. Biol. Ther. 13, 269–282 (2013). 93. Ruzza, P. & Calderan, A. Glutathione transferase (GST)-activated prodrugs. Pharmaceutics 5, 220–231 (2013). 94. Morgan, A. S. et al. Tumor efficacy and bone marrow-sparing properties of TER286, a cytotoxin activated by glutathione S-transferase. Cancer Res. 58, 2568–2575 (1998). 95. Gunning, V. et al. Arabidopsis glutathione transferases U24 and U25 exhibit a range of detoxification activities with the environmental pollutant and explosive, 2,4,6-trinitrotoluene. Plant Physiol. 165, 854– 865 (2014). 96. Mazari, A. M. A. & Mannervik, B. Drosophila GSTs display outstanding catalytic efficiencies with the environmental pollutants 2,4,6-trinitrotoluene and 2,4-dinitrotoluene. Biochem. Biophys. Rep. 5, 141–145 (2016). 97. Tzafestas, K. et al. Expression of a Drosophila glutathione transferase in Arabidopsis confers the ability to detoxify the environmental pollutant, and explosive, 2,4,6-trinitrotoluene. New Phytol. 214, 294– 303 (2017). 98. Adler, V. et al. Regulation of JNK signaling by GSTp. EMBO J. 18, 1321–1334 (1999). 99. Adler, V. & Pincus, M. R. Effector peptides from glutathione-S- transferase-pi affect the activation of jun by jun-N-terminal kinase. Ann. Clin. Lab. Sci. 34, 35–46 (2004). 100. Zhang, J. et al. Chapter Four - Pleiotropic Functions of Glutathione S- Transferase P. in Advances in Cancer Research (eds. Townsend, D. M. & Tew, K. D.) vol. 122 143–175 (Academic Press, 2014). 101. Tew, K. D. & Townsend, D. M. Regulatory functions of glutathione S- transferase P1-1 unrelated to detoxification. Drug Metab. Rev. 43, 179– 193 (2011). 102. Dalle-Donne, I., Rossi, R., Colombo, G., Giustarini, D. & Milzani, A.

77 Protein S-glutathionylation: a regulatory device from bacteria to humans. Trends Biochem. Sci. 34, 85–96 (2009). 103. Le Trong, I., Stenkamp, R. E., Ibarra, C., Atkins, W. M. & Adman, E. T. 1.3‐Å resolution structure of human glutathione S‐transferase with S‐ hexyl glutathione bound reveals possible extended ligandin binding site. Proteins Struct. Funct. Bioinforma. 48, 618–627 (2002). 104. Ketley, J. N., Habig, W. H. & Jakoby, W. B. Binding of nonsubstrate ligands to the glutathione S-transferases. J. Biol. Chem. 250, 8670–8673 (1975). 105. Litwack, G., Ketterer, B. & Arias, I. M. Ligandin: a hepatic protein which binds steroids, bilirubin, carcinogens and a number of exogenous organic anions. Nature 234, 466–467 (1971). 106. Kanaoka, Y. et al. Cloning and crystal structure of hematopoietic prostaglandin D synthase. Cell 90, 1085–1095 (1997). 107. Johansson, A.-S. & Mannervik, B. Human glutathione transferase A3-3, a highly efficient catalyst of double-bond isomerization in the biosynthetic pathway of steroid hormones. J. Biol. Chem. 276, 33061– 33065 (2001). 108. Fedulova, N., Raffalli-Mathieu, F. & Mannervik, B. Porcine glutathione transferase Alpha 2-2 is a human GST A3-3 analogue that catalyses steroid double-bond isomerization. Biochem. J. 431, 159–167 (2010). 109. Baker, M. E. Steroid receptors and vertebrate evolution. Mol. Cell. Endocrinol. 496, 110526 (2019). 110. Lednicer, D. Steroid Chemistry at a Glance. (Wiley, 2010). 111. Kuo, T., McQueen, A., Chen, T.-C. & Wang, J.-C. Regulation of glucose homeostasis by glucocorticoids. Adv. Exp. Med. Biol. 872, 99– 126 (2015). 112. Coutinho, A. E. & Chapman, K. E. The anti-inflammatory and immunosuppressive effects of glucocorticoids, recent developments and mechanistic insights. Mol. Cell. Endocrinol. 335, 2–13 (2011). 113. Rhen, T. & Cidlowski, J. A. Antiinflammatory action of glucocorticoids — new mechanisms for old drugs. N. Engl. J. Med. 353, 1711–1723 (2005). 114. Furman, B. L. Mineralocorticoids. in xPharm: The Comprehensive Pharmacology Reference (eds. Enna, S. J. & Bylund, D. B.) 1 (Elsevier, 2007). doi:10.1016/B978-008055232-3.61052-9. 115. Ogino, Y., Sato, T. & Iguchi, T. Chapter 94 - Gonadal Steroids. in Handbook of Hormones (eds. Takei, Y., Ando, H. & Tsutsui, K.) 504– 506 (Academic Press, 2016). doi:10.1016/B978-0-12-801028-0.00094- 5. 116. The nomenclature of steroids. Eur. J. Biochem. 186, 429–458 (1989). 117. Nussey, S. & Whitehead, S. The Adrenal Gland. in Endocrinology: An integrated Approach. (BIOS Scientific Publishers, 2001). 118. Bamberger, C. M., Schulte, H. M. & Chrousos, G. P. Molecular determinants of glucocorticoid receptor function and tissue sensitivity

78 to glucocorticoids. Endocr. Rev. 17, 245–261 (1996). 119. Smith, S. M. & Vale, W. W. The role of the hypothalamic-pituitary- adrenal axis in neuroendocrine responses to stress. Dialogues Clin. Neurosci. 8, 383–395 (2006). 120. Magomedova, L. & Cummins, C. L. Glucocorticoids and Metabolic Control. in Metabolic Control (ed. Herzig, S.) 73–93 (Springer International Publishing, 2016). doi:10.1007/164_2015_1. 121. Mulla, A., LeRoux, C., Solito, E. & Buckingham, J. C. Correlation between the antiinflammatory protein annexin 1 (lipocortin 1) and serum cortisol in subjects with normal and dysregulated adrenal function. J. Clin. Endocrinol. Metab. 90, 557–562 (2005). 122. Roviezzo, F. et al. The annexin-1 knockout mouse: what it tells us about the inflammatory response. J. Physiol. Pharmacol. Off. J. Pol. Physiol. Soc. 53, 541–553 (2002). 123. Kassel, O. et al. Glucocorticoids inhibit MAP kinase via increased expression and decreased degradation of MKP-1. EMBO J. 20, 7108– 7116 (2001). 124. Barnes, P. J. Anti-inflammatory actions of glucocorticoids: molecular mechanisms. Clin. Sci. 94, 557–572 (1998). 125. König, H., Ponta, H., Rahmsdorf, H. j. & Herrlich, P. Interference between pathway-specific transcription factors: glucocorticoids antagonize phorbol ester-induced AP-1 activity without altering AP-1 site occupation in vivo. EMBO J. 11, 2241–2246 (1992). 126. Nissen, R. M. & Yamamoto, K. R. The glucocorticoid receptor inhibits NFκB by interfering with serine-2 phosphorylation of the RNA polymerase II carboxy-terminal domain. Genes Dev. 14, 2314–2329 (2000). 127. Lother, A., Moser, M., Bode, C., Feldman, R. D. & Hein, L. Mineralocorticoids in the heart and vasculature: new insights for old hormones. Annu. Rev. Pharmacol. Toxicol. 55, 289–312 (2015). 128. Spät, A. & Hunyady, L. Control of aldosterone secretion: a model for convergence in cellular signaling pathways. Physiol. Rev. 84, 489–539 (2004). 129. Lymperopoulos, A. et al. An adrenal β-arrestin 1-mediated signaling pathway underlies angiotensin II-induced aldosterone production in vitro and in vivo. Proc. Natl. Acad. Sci. U. S. A. 106, 5825–5830 (2009). 130. Arriza, J. L. et al. Cloning of human mineralocorticoid receptor complementary DNA: structural and functional kinship with the glucocorticoid receptor. Science 237, 268–275 (1987). 131. Fuller Peter J. & Young Morag J. Mechanisms of mineralocorticoid action. Hypertension 46, 1227–1235 (2005). 132. Fraccarollo Daniela et al. Deletion of cardiomyocyte mineralocorticoid receptor ameliorates adverse remodeling after myocardial infarction. Circulation 123, 400–408 (2011). 133. Gass, P. et al. Genetic disruption of mineralocorticoid receptor leads to

79 impaired neurogenesis and granule cell degeneration in the hippocampus of adult mice. EMBO Rep. 1, 447–451 (2000). 134. Connell, J. M. C. & Davies, E. M. M. The new biology of aldosterone. J. Endocrinol. 186, 1–20 (2005). 135. Stockand, J. D. New ideas about aldosterone signaling in epithelia. Am. J. Physiol.-Ren. Physiol. 282, F559–F576 (2002). 136. Wendler, A. et al. Position Paper: Rapid responses to steroids: current status and future prospects. Eur. J. Endocrinol. 162, 825–830 (2010). 137. Grossmann, C. et al. Human mineralocorticoid receptor expression renders cells responsive for nongenotropic aldosterone actions. Mol. Endocrinol. 19, 1697–1710 (2005). 138. Plant, T. M. The hypothalamo-pituitary-gonadal axis. J. Endocrinol. 226, T41–T54 (2015). 139. Harris, G. W. Neural control of the pituitary gland. Physiol. Rev. 28, 139–179 (1948). 140. Wang, C., Liu, Y. & Cao, J.-M. G protein-coupled receptors: extranuclear mediators for the non-genomic actions of steroids. Int. J. Mol. Sci. 15, 15412–15425 (2014). 141. Guerriero, G. Vertebrate sex steroid receptors: evolution, ligands, and neurodistribution. Ann. N. Y. Acad. Sci. 1163, 154–168 (2009). 142. Patel, B. et al. Role of nuclear progesterone receptor isoforms in uterine pathophysiology. Hum. Reprod. Update 21, 155–173 (2015). 143. Wen, D. X., Xu, Y. F., Mais, D. E., Goldman, M. E. & McDonnell, D. P. The A and B isoforms of the human progesterone receptor operate through distinct signaling pathways within target cells. Mol. Cell. Biol. 14, 8356–8364 (1994). 144. Leonhardt, S. A., Boonyaratanakornkit, V. & Edwards, D. P. Progesterone receptor transcription and non-transcription signaling mechanisms. Steroids 68, 761–770 (2003). 145. Gellersen, B., Fernandes, M. S. & Brosens, J. J. Non-genomic progesterone actions in female reproduction. Hum. Reprod. Update 15, 119–138 (2009). 146. Sinnesael, M. et al. Androgen receptor (AR) in osteocytes is important for the maintenance of male skeletal integrity: Evidence from targeted AR disruption in mouse osteocytes. J. Bone Miner. Res. 27, 2535–2543 (2012). 147. Davey, R. A. & Grossmann, M. Androgen receptor structure, function and biology: from bench to bedside. Clin. Biochem. Rev. 37, 3–15 (2016). 148. Dubois, V., Laurent, M., Boonen, S., Vanderschueren, D. & Claessens, F. Androgens and skeletal muscle: cellular and molecular action mechanisms underlying the anabolic actions. Cell. Mol. Life Sci. 69, 1651–1667 (2012). 149. Handelsman, D. J. Androgen Physiology, Pharmacology and Abuse. in Endotext (eds. Dungan, K. et al.) (MDText.com, Inc., 2000).

80 150. Heinlein, C. A. & Chang, C. The roles of androgen receptors and androgen-binding proteins in nongenomic androgen actions. Mol. Endocrinol. 16, 2181–2187 (2002). 151. Makieva, S., Saunders, P. T. K. & Norman, J. E. Androgens in pregnancy: roles in parturition. Hum. Reprod. Update 20, 542–559 (2014). 152. Dahlman-Wright, K. et al. International Union of Pharmacology. LXIV. Estrogen receptors. Pharmacol. Rev. 58, 773–781 (2006). 153. Soltysik, K. & Czekaj, P. Membrane estrogen receptors - is it an alternative way of estrogen action? J. Physiol. Pharmacol. Off. J. Pol. Physiol. Soc. 64, 129–142 (2013). 154. Björnström, L. & Sjöberg, M. Estrogen receptor-dependent activation of AP-1 via non-genomic signalling. Nucl. Recept. 2, 3 (2004). 155. Barton, M. et al. Twenty years of the G protein-coupled estrogen receptor GPER: Historical and personal perspectives. J. Steroid Biochem. Mol. Biol. 176, 4–15 (2018). 156. Prossnitz, E. R., Arterburn, J. B. & Sklar, L. A. GPR30: a G protein- coupled receptor for estrogen. Mol. Cell. Endocrinol. 265–266, 138–142 (2007). 157. Taraborrelli, S. Physiology, production and action of progesterone. Acta Obstet. Gynecol. Scand. 94, 8–16 (2015). 158. Brooks, R. V. Androgens. Clin. Endocrinol. Metab. 4, 503–520 (1975). 159. Burger, H. G. Androgen production in women. Fertil. Steril. 77, 3–5 (2002). 160. Cui, J., Shen, Y. & Li, R. Estrogen synthesis and signaling pathways during ageing: from periphery to brain. Trends Mol. Med. 19, 197–209 (2013). 161. Payne, A. H. & Hales, D. B. Overview of steroidogenic enzymes in the pathway from cholesterol to active steroid hormones. Endocr. Rev. 25, 947–970 (2004). 162. Miller, W. L. Molecular biology of steroid hormone synthesis. Endocr. Rev. 9, 295–318 (1988). 163. Miller, W. L. Minireview: regulation of steroidogenesis by electron transfer. Endocrinology 146, 2544–2550 (2005). 164. Conley, A. J. & Bird, I. M. The role of cytochrome P450 17α- hydroxylase and 3β-hydroxysteroid dehydrogenase in the integration of gonadal and adrenal steroidogenesis via the Δ5 and Δ4 pathways of steroidogenesis in mammals. Biol. Reprod. 56, 789–799 (1997). 165. Arlt, W. & Stewart, P. M. Adrenal corticosteroid biosynthesis, metabolism, and action. Endocrinol. Metab. Clin. North Am. 34, 293– 313 (2005). 166. Bassett, M. H., White, P. C. & Rainey, W. E. The regulation of aldosterone synthase expression. Mol. Cell. Endocrinol. 217, 67–74 (2004). 167. Samuels, L. T., Helmreich, M. L., Lasater, M. B. & Reich, H. An

81 enzyme in endocrine tissues which oxidizes Δ5-3 hydroxy steroids to α,β unsaturated ketones. Science 113, 490–491 (1951). 168. Rossetti, M. F., Cambiasso, M. J., Holschbach, M. A. & Cabrera, R. Oestrogens and progestagens: synthesis and action in the brain. J. Neuroendocrinol. 28, n/a-n/a (2016). 169. Reddy, D. S. Chapter 8 - Neurosteroids: Endogenous role in the human brain and therapeutic potentials. in Progress in Brain Research (ed. Savic, I.) vol. 186 113–137 (Elsevier, 2010). 170. Yadong, G., J, B. A., Bin, C. & P, M. M. Estrogens attenuate and corticosterone exacerbates excitotoxicity, oxidative injury, and amyloid β-peptide toxicity in hippocampal neurons. J. Neurochem. 66, 1836– 1844 (1996). 171. Nilsson, S. et al. Mechanisms of estrogen action. Physiol. Rev. 81, 1535–1565 (2001). 172. MacKenzie, G. & Maguire, J. Neurosteroids and GABAergic signaling in health and disease. Biomol. Concepts 4, 29–42 (2012). 173. Vyklicky, V. et al. Block of NMDA receptor channels by endogenous neurosteroids: implications for the agonist induced conformational states of the channel vestibule. Sci. Rep. 5, 10935–10949 (2015). 174. Ooishi, Y. et al. Modulation of synaptic plasticity in the hippocampus by hippocampus-derived estrogen and androgen. J. Steroid Biochem. Mol. Biol. 131, 37–51 (2012). 175. Morey, K. S. & Litwack, G. Isolation and properties of cortisol metabolite binding proteins of rat liver cytosol. Biochemistry 8, 4813– 4821 (1969). 176. Habig, W. H. et al. The identity of glutathione S-transferase B with ligandin, a major binding protein of liver. Proc. Natl. Acad. Sci. 71, 3879–3882 (1974). 177. Benson, A. M., Talalay, P., Keen, J. H. & Jakoby, W. B. Relationship between the soluble glutathione-dependent A5-3-ketosteroid isomerase and the glutathione S-transferases of the liver. Proc. Natl. Acad. Sci. U. S. A. 74, 158–162 (1977). 178. Benson, A. M. & Talalay, P. Role of reduced glutathione in the Δ5-3- ketosteroid isomerase reaction of liver. Biochem. Biophys. Res. Commun. 69, 1073–1079 (1976). 179. Homma, H., Maruyama, H., Niitsu, Y. & Listowsky, I. A subclass of glutathione S-transferases as intracellular high-capacity and high- affinity steroid-binding proteins. Biochem. J. 235, 763–768 (1986). 180. Pettersson, P. & Mannervik, B. The role of glutathione in the isomerization of Delta(5)-androstene-3,17-dione catalyzed by human glutathione transferase A1-1. J. Biol. Chem. 276, 11698–11704 (2001). 181. Raffalli-Mathieu, F., Persson, D. & Mannervik, B. Differences between bovine and human steroid double-bond isomerase activities of Alpha- class glutathione transferases selectively expressed in steroidogenic tissues. Biochim. Biophys. Acta 1770, 130–136 (2007).

82 182. Johansson, A.-S. & Mannervik, B. Active-site residues governing high steroid isomerase activity in human glutathione transferase A3-3. J. Biol. Chem. 277, 16648–16654 (2002). 183. Calvaresi, M., Stenta, M., Garavelli, M., Altoé, P. & Bottoni, A. Computational evidence for the catalytic mechanism of human glutathione S-transferase A3-3: a QM/MM investigation. ACS Catal. 2, 280–286 (2012). 184. Calvaresi, M. et al. Computational QM/MM study of the reaction mechanism of human glutathione S‐transferase A3‐3. AIP Conf. Proc. 963, 696–698 (2007). 185. Gu, Y. et al. Crystal structure of human glutathione S-Transferase A3-3 and mechanistic implications for its high steroid isomerase activity. Biochemistry 43, 15673–15679 (2004). 186. Dourado, D. F. A. R., Fernandes, P. A., Mannervik, B. & Ramos, M. J. Isomerization of Δ5-androstene-3,17-dione into Δ4-androstene-3,17- dione catalyzed by human glutathione transferase A3-3: a computational study identifies a dual role for glutathione. J. Phys. Chem. A 118, 5790– 5800 (2014). 187. Ewer, J. How the Ecdysozoan changed its coat. PLoS Biol. 3, (2005). 188. Says, P. P. P. A new role of an insect steroid hormone: The link between mating and germline stem cells. the Node https://thenode.biologists.com/new-role-insect-steroid-hormone-link- mating-germline-stem-cells/highlights/ (2016). 189. Thummel, C. S. Flies on steroids — Drosophila metamorphosis and the mechanisms of steroid hormone action. Trends Genet. 12, 306–310 (1996). 190. Christesen, D. et al. Transcriptome analysis of Drosophila melanogaster third instar larval ring glands points to novel functions and uncovers a cytochrome p450 required for development. G3 GenesGenomesGenetics 7, 467–479 (2016). 191. Saisawang, C., Wongsantichon, J. & Ketterman, A. J. A preliminary characterization of the cytosolic glutathione transferase proteome from Drosophila melanogaster. Biochem. J. 442, 181–190 (2012). 192. Mannervik, B. Five decades with glutathione and the GSTome. J. Biol. Chem. 287, 6072–6083 (2012). 193. Rubin, G. M. et al. Comparative genomics of the eukaryotes. Science 287, 2204–2215 (2000). 194. Chanut-Delalande, H. et al. Pri peptides are mediators of ecdysone for the temporal control of development. Nat. Cell Biol. 16, 1035–1044 (2014). 195. Enya, S. et al. A Halloween gene noppera-bo encodes a glutathione S-transferase essential for ecdysteroid biosynthesis via regulating the behaviour of cholesterol in Drosophila. Sci. Rep. 4, 6586 (2014). 196. Fedulova, N., Raffalli-Mathieu, F. & Mannervik, B. Characterization of porcine Alpha-class glutathione transferase A1-1. Arch. Biochem.

83 Biophys. 507, 205–211 (2011). 197. Jeffrey Chang, R. A practical approach to the diagnosis of polycystic ovary syndrome. Am. J. Obstet. Gynecol. 191, 713–717 (2004). 198. Gomes, L. G., Bachega, T. A. S. S. & Mendonca, B. B. Classic congenital adrenal hyperplasia and its impact on reproduction. Fertil. Steril. 111, 7–12 (2019). 199. Henderson, B. E. & Feigelson, H. S. Hormonal carcinogenesis. Carcinogenesis 21, 427–433 (2000). 200. Xanthakis, V. & Vasan, R. S. Aldosterone and the risk of hypertension. Curr. Hypertens. Rep. 15, 102–107 (2013). 201. Briot, K. & Roux, C. Glucocorticoid-induced osteoporosis. RMD Open 1, (2015). 202. Scian, M. et al. Comparison of epsilon- and delta-class glutathione S- transferases: the crystal structures of the glutathione S-transferases DmGSTE6 and DmGSTE7 from Drosophila melanogaster. Acta Crystallogr. D Biol. Crystallogr. 71, 2089–2098 (2015). 203. Joshi, M. & Deshpande, J. D. Polymerase chain reaction: methods, principles and applications. Int. J. Biomed. Res. 2, 81–97 (2011). 204. Baltimore, D. Viral RNA-dependent DNA polymerase: RNA-dependent DNA polymerase in virions of RNA tumour viruses. Nature 226, 1209– 1211 (1970). 205. Green, M. R. & Sambrook, J. Nested Polymerase Chain Reaction (PCR). Cold Spring Harb. Protoc. 2019, pdb.prot095182 (2019). 206. VanGuilder, H., Vrana, K. & Freeman, W. Twenty-five years of quantitative PCR for gene expression analysis. BioTechniques 44 Supplement, 619–626 (2008). 207. Freeman, W. M., Walker, S. J. & Vrana, K. E. Quantitative RT-PCR: pitfalls and potential. BioTechniques 26, 112–122, 124–125 (1999). 208. Schefe, J. H., Lehmann, K. E., Buschmann, I. R., Unger, T. & Funke- Kaiser, H. Quantitative real-time RT-PCR data analysis: current concepts and the novel “gene expression’s CTdifference” formula. J. Mol. Med. 84, 901–910 (2006). 209. Nailis, H., Coenye, T., Nieuwerburgh, F. V., Deforce, D. & Nelis, H. J. Development and evaluation of different normalization strategies for gene expression studies in Candida albicans biofilms by real-time PCR. BMC Mol. Biol. 7, 1–9 (2006). 210. Shuman, S. DNA ligases: progress and prospects. J. Biol. Chem. 284, 17365–17369 (2009). 211. Lehnman, I. R. DNA ligase: structure, mechanism, and function. Science 186, 790–797 (1974). 212. Pingoud, A. & Jeltsch, A. Structure and function of type II restriction endonucleases. Nucleic Acids Res. 29, 3705–3727 (2001). 213. Weaver, J. C. & Chizmadzhev, Yu. A. Theory of electroporation: a review. Bioelectrochem. Bioenerg. 41, 135–160 (1996). 214. Fiedler, S. & Wirth, R. Transformation of bacteria with plasmid DNA

84 by electroporation. Anal. Biochem. 170, 38–44 (1988). 215. Jeong, H. et al. Genome sequences of Escherichia coli B strains REL606 and BL21(DE3). J. Mol. Biol. 394, 644–652 (2009). 216. Escherichia coli (strain B / BL21-DE3). https://www.uniprot.org/proteomes/UP000001509. 217. Ninfa, A. F., Ballou, D. P. & Menore, M. Fundamental Laboratory Approaches for Biochemistry and Biotechnology, p. 133. (Wiley, 2009). 218. Cuatrecasas, P. Protein purification by affinity chromatography: derivatizations of agarose and polyacrylamide beads. J. Biol. Chem. 245, 3059–3065 (1970). 219. Ahmed, H. Principles and Reactions of Protein Extraction, Purification, and Characterization, p. 43. (CRC Press, 2004). 220. Gaberc-Porekar, V. & Menart, V. Perspectives of immobilized-metal affinity chromatography. J. Biochem. Biophys. Methods 49, 335–360 (2001). 221. Yip, T. T. & Hutchens, T. W. Immobilized metal ion affinity chromatography. Mol. Biotechnol. 1, 151–164 (1994). 222. Porath, J., Carlsson, J., Olsson, I. & Belfrage, G. Metal chelate affinity chromatography, a new approach to protein fractionation. Nature 258, 598–599 (1975). 223. Trumbo, T. A., Schultz, E., Borland, M. G. & Pugh, M. E. Applied spectrophotometry: Analysis of a biochemical mixture. Biochem. Mol. Biol. Educ. 41, 242–250 (2013). 224. Fedulova, N. & Mannervik, B. Experimental conditions affecting functional comparison of highly active glutathione transferases. Anal. Biochem. 413, 16–23 (2011). 225. Mori, S. & Barth, H. G. Size Exclusion Chromatography. (Springer Science & Business Media, 2013). 226. Christian, G. D., Dasgupta, P. (Sandy) & Schug, K. Analytical Chemistry, 7th Edition: Seventh Edition, p. 653. (Wiley Global Education, 2013). 227. Zheng, H. et al. X-ray crystallography over the past decade for novel drug discovery – where are we heading next? Expert Opin. Drug Discov. 10, 975–989 (2015). 228. Harris, R. K. Nuclear Magnetic Resonance Spectroscopy. (1986). 229. Cheng, Y., Grigorieff, N., Penczek, P. A. & Walz, T. A primer to single- particle cryo-electron microscopy. Cell 161, 438–449 (2015). 230. McPherson, A. & Gavira, J. A. Introduction to protein crystallization. Acta Crystallogr. Sect. F Struct. Biol. Commun. 70, 2–20 (2013). 231. Smyth, M. S. & Martin, J. H. J. x Ray crystallography. Mol. Pathol. 53, 8–14 (2000). 232. Dorit, R. L. & Ohara, O. cDNA amplification using one-sided (anchored) PCR. Curr. Protoc. Mol. Biol. Chapter 15, Unit 15.6 (2001). 233. Mackay, I. M., Arden, K. E. & Nitsche, A. Real-time PCR in virology. Nucleic Acids Res. 30, 1292–1305 (2002).

85 234. Mannervik, B. & Jemth, P. Measurement of glutathione transferases. in Current Protocols in Toxicology (eds. Maines, M. D., Costa, L. G., Reed, D. J., Sassa, S. & Sipes, I. G.) 6.4.1-6.4.10 (John Wiley & Sons, Inc., 1999). doi:10.1002/0471140856.tx0604s01. 235. Zhang, W., Dourado, D. F. A. R. & Mannervik, B. Evolution of the active site of human glutathione transferase A2-2 for enhanced activity with dietary isothiocyanates. Biochim. Biophys. Acta BBA - Gen. Subj. 1850, 742–749 (2015). 236. Thomas, J. L., Myers, R. P. & Strickler, R. C. Human placental 3β- hydroxy-5-ene-steroid dehydrogenase and steroid 5→ 4-ene-isomerase: Purification from mitochondria and kinetic profiles, biophysical characterization of the purified mitochondrial and microsomal enzymes. J. Steroid Biochem. 33, 209–217 (1989). 237. Fersht, A. Structure and Mechanism in Protein Science : a Guide to Enzyme Catalysis and Protein Folding, p.166. (Freeman and Company, 1999). 238. Coles, B. F. & Kadlubar, F. F. Human Alpha Class Glutathione S-Transferases: Genetic Polymorphism, Expression, and Susceptibility to Disease. in Methods in Enzymology (eds. Sies, H. & Packer, L.) vol. 401 9–42 (Academic Press, 2005). 239. Raffalli-Mathieu, F., Orre, C., Stridsberg, M., Hansson Edalat, M. & Mannervik, B. Targeting human glutathione transferase A3-3 attenuates progesterone production in human steroidogenic cells. Biochem. J. 414, 103–109 (2008). 240. Musdal, Y., Hegazy, U. M., Aksoy, Y. & Mannervik, B. FDA-approved drugs and other compounds tested as inhibitors of human glutathione transferase P1-1. Chem. Biol. Interact. 205, 53–62 (2013). 241. Mazari, A. M. A., Hegazy, U. M. & Mannervik, B. Identification of new inhibitors for human hematopoietic prostaglandin D2 synthase among FDA-approved drugs and other compounds. Chem. Biol. Interact. 229, 91–99 (2015). 242. Peus, D. et al. Anti-psoriatic drug anthralin activates JNK via lipid peroxidation: mononuclear cells are more sensitive than keratinocytes. J. Invest. Dermatol. 114, 688–692 (2000). 243. Kon, R. et al. Laxative effect of repeated Daiokanzoto is attributable to decrease in aquaporin-3 expression in the colon. J. Nat. Med. 72, 493– 502 (2018). 244. Hansson, J. et al. Sensitization of human melanoma cells to the cytotoxic effect of melphalan by the glutathione transferase inhibitor ethacrynic acid. Cancer Res. 51, 94–98 (1991). 245. Molnar, J. & Somberg, J. C. The clinical pharmacology of ethacrynic acid. Am. J. Ther. 16, 86–92 (2009). 246. Naus, P. J. et al. Tannic acid synergizes the cytotoxicity of chemotherapeutic drugs in human cholangiocarcinoma by modulating drug efflux pathways. J. Hepatol. 46, 222–229 (2007).

86 247. Benkovic, S. J., Hammes, G. G. & Hammes-Schiffer, S. Free-energy landscape of enzyme catalysis. Biochemistry 47, 3317–3321 (2008). 248. Acelajado, M. C. & Calhoun, D. A. Aldosteronism and resistant hypertension. Int. J. Hypertens. 2011, 1–7 (2011). 249. Kon, R. et al. Rheinanthrone, a metabolite of sennoside A, triggers macrophage activation to decrease aquaporin-3 expression in the colon, causing the laxative effect of rhubarb extract. J. Ethnopharmacol. 152, 190–200 (2014). 250. Mannervik, B. & Jensson, H. Binary combinations of four protein subunits with different catalytic specificities explain the relationship between six basic glutathione S-transferases in rat liver cytosol. J. Biol. Chem. 257, 9909–9912 (1982). 251. Mannervik, B. et al. Identification of three classes of cytosolic glutathione transferase common to several mammalian species: correlation between structural data and enzymatic properties. Proc. Natl. Acad. Sci. U. S. A. 82, 7202–7206 (1985). 252. Sinning, I. et al. Structure determination and refinement of human Alpha class glutathione transferase A1-1, and a comparison with the Mu and Pi class enzymes. J. Mol. Biol. 232, 192–212 (1993). 253. Dirr, H., Reinemer, P. & Huber, R. X-ray crystal structures of cytosolic glutathione S-transferases. Implications for protein architecture, substrate recognition and catalytic function. Eur. J. Biochem. 220, 645– 661 (1994). 254. Steiner, T. The hydrogen bond in the solid state. Angew. Chem. Int. Ed. 41, 48–76 (2002). 255. Cameron, A. D. et al. Structural analysis of human alpha-class glutathione transferase A1-1 in the apo-form and in complexes with ethacrynic acid and its glutathione conjugate. Structure 3, 717–727 (1995). 256. Board, P. G. & Mannervik, B. The contribution of the C-terminal sequence to the catalytic activity of GST2, a human Alpha-class glutathione transferase. Biochem. J. 275, 171–174 (1991). 257. Prabhu, K. S. et al. Microsomal glutathione S-transferase A1-1 with glutathione peroxidase activity from sheep liver: molecular cloning, expression and characterization. Biochem. J. 360, 345–354 (2001). 258. Sandeep Prabhu, K., Reddy, P. V., Jones, E. C., Liken, A. D. & Reddy, C. C. Characterization of a class alpha glutathione-S-transferase with glutathione peroxidase activity in human liver microsomes. Arch. Biochem. Biophys. 424, 72–80 (2004). 259. Vogel, C. & Marcotte, E. M. Insights into the regulation of protein abundance from proteomic and transcriptomic analyses. Nat. Rev. Genet. 13, 227–232 (2012). 260. Singh, M. & Su, C. Progesterone and neuroprotection. Horm. Behav. 63, 284–290 (2013). 261. Ibanez, C. et al. Steroids and the reversal of age-associated changes in

87 myelination and remyelination. Prog. Neurobiol. 71, 49–56 (2003).