Appl Biochem Biotechnol DOI 10.1007/s12010-013-0303-2

Coexpression of CPR from Various Origins Enhances Biotransformation Activity of Human CYPs in S. pombe Ina Neunzig & Maria Widjaja & Frank T. Peters & Hans H. Maurer & Alain Hehn & Frédéric Bourgaud & Matthias Bureik

Received: 9 April 2013 / Accepted: 16 May 2013 # Springer Science+Business Media New York 2013

Abstract Cytochrome P450 enzymes (CYPs or P450s) are the most important enzymes involved in the phase I metabolism of drugs (and other xenobiotics) in humans, and the corresponding drug metabolites are needed as reference substances for their structural confirmation and for pharmacological or toxicological characterization. We have previously shown that biotechnological synthesis of such metabolites is feasible by whole-cell biotransformation with human CYPs recombinantly expressed in the fission yeast Schizosaccharomyces pombe. It was the aim of this study to compare the activity of seven human microsomal CYPs (CYP2C9, CYP2D6, CYP3A4, CYP3A5, CYP3A7, CYP17, and CYP21) upon coexpression with NADPH-cytochrome P450 oxidoreductases (CPRs) from various origins, namely, human CPR (hCPR) and its homologues from fission yeast (ccr1) and the bishop’s weed Ammi majus (AmCPR), respectively. For this purpose, 28 recombinant strains were needed, with five of them having been constructed previously and 23 strains being newly constructed. Bioconversion experiments showed that coexpression of a CPR does not only influence the reaction rate but, in some cases, also exerts an influence on the metabolite pattern. For CYP3A enzymes, coexpression of hCPR yielded the best results, while for another two, hCPR was equally helpful as ccr1 (both CYP17 and CYP21) or AmCPR (CYP17 only), respectively. Interestingly, I. Neunzig : M. Widjaja : M. Bureik (*) PomBioTech GmbH, Campus, 66123 Saarbrücken, Germany e-mail: [email protected] F. T. Peters Institute of Forensic Medicine, University Hospital Jena, 07740 Jena, Germany H. H. Maurer Institute of Experimental and Clinical Pharmacology and Toxicology, Saarland University, 66421 Homburg (Saar), Germany A. Hehn : F. Bourgaud Agronomie et Environnement Nancy-Colmar, ENSAIA, Université de Lorraine UMR 1121, 2 Avenue de la Forêt de Haye, 54505 Vandoevre-lès-Nancy, France

Appl Biochem Biotechnol

CYP2D6 displayed its highest activity when coexpressed with ccr1 and CYP2C9 with AmCPR. These results corroborate the view of CPR as a well-suited bio-brick in synthetic biology for the construction of artificial enzyme complexes. Keywords Biocatalysis . Cytochrome P450 reductase . Recombinant fission yeast . Strain development . Whole-cell biotransformation Introduction Human cytochromes P450 (P450s or CYPs) have key functions in several distinct physiological processes, such as drug metabolism or steroid hormone biosynthesis. The most common reactions catalyzed by these enzymes are hydroxylations, but they can also accomplish other chemical alterations such as oxidation, demethylation, and cleavage of ethers or carbon–carbon bonds [1–3]. Like almost all eukaryotic P450s, most human CYPs belong to the class II microsomal P450 systems that are composed of two integral membrane proteins, the CYP enzyme itself and the NADPH-cytochrome P450 oxidoreductase (CPR) [4]. According to current knowledge, only a subset of the total complement of 57 human CYPs is primarily involved in the metabolism of xenobiotics [5]. These CYPs, which are predominantly expressed in the liver and intestine, are of considerable interest for pharmacological research in the context of drug–drug and drug–food interactions as well as metabolites in safety testing, as drug metabolites may exert harmful effects on patients [6]. For the purpose of metabolite production by whole-cell biotransformation, the fission yeast Schizosaccharomyces pombe has been successfully applied as a functional expression system for human drug metabolizing CYPs in many cases, and the advantages of this approach have been described previously [7–9]. Efforts to further improve this system have so far concentrated on the optimization of the reaction conditions [7, 10] and on molecular engineering of the CYPs involved [11]. Previous studies with human steroidogenic CYPs recombinantly expressed in fission yeast demonstrated the potential of improving the biotransformation rates by refinement of the mitochondrial electron transfer system [12]. Here, we sought to enhance the activity of human microsomal CYPs by coexpression of CPRs from various origins. CPR, which has also been designated P450 oxidoreductase or NADPH-cytochrome P450 oxidoreductase, was originally isolated from yeast and considered to be a cytochrome c reductase based on its ability to reduce cytochrome c [13]. We now know that it is a component of CYP systems and shuttles electrons from NADPH via the cofactors FAD and FMN to the iron in the prosthetic heme group of the recipient CYP. CPR has evolved as a fusion of two ancestral proteins, displaying in the N-terminus homology to the FMN-containing bacterial flavodoxins, while the C-terminal part is homologous to the FAD-containing ferredoxin NADP+ reductases and to NADH-cytochrome b5 reductase [4]. While the three-dimensional structure of rat CPR was known for some time [14], that of human CPR (hCPR) has only been solved recently [15]. A number of mutations in the hCPR gene have been found in patients with Antley–Bixler syndrome or with disorders in steroidogenesis, and CPR knockout mice are embryonically lethal [16]. Of course, it would be expected a priori that human CYP enzymes interact best with hCPR; however, previous data from Denis Pompon’s group showed that overexpression of the baker’s yeast CPR homologue NCP1 very efficiently supports both bovine CYP17 and human CYP21 in whole-cell biotransformations [17]. Overexpression of the fission yeast CPR homologue ccr1 is described here for the first time. Based on homology modeling, ccr1 contains a transmembrane domain and FAD-, FMN-, and NADPH-binding regions comparable to mammalian CPRs [18, 19]. As there are only two endogenous CYPs in S. pombe that are both thought to

Appl Biochem Biotechnol

be involved in ergosterol biosynthesis [20, 21], the biological function of ccr1 would appear to be limited to this pathway in this microbe. In plants, there are many more CYP families than in animals, with many isoenzymes having tremendous biotechnological perspectives [22]; but after recombinant expression in baker’s yeast, plant CYPs only display a fraction of the activity achieved when coexpressed with a plant CPR [23]. Interestingly, up to three CPR paralogs have been found in some plant species [24, 25], while mammalian and yeast cells only contain one. Plant CYPs recombinantly expressed in other plants in general appear to be sufficiently supported by the endogenous CPRs, a finding that points to high conservation of the amino acid residues in CPR that mediate interaction with the CYPs [23, 26, 27]. In this study, the sole known CPR of the bishop’s weed Ammi majus (AmCPR) was used, a plant that is of interest because it synthesizes coumarins with pharmacological potential [3, 28]. For the purpose of this work, each of the three CPRs from Homo sapiens, S. pombe, and A. majus, respectively, was coexpressed with one of a set of an important human drug metabolizing CYPs (CYP3A4, CYP3A5, CYP3A7, CYP2C9, or CYP2D6) or a steroid hydroxylase (CYP17 or CYP21), and the whole-cell biotransformation activities of the resulting strains were compared.

Materials and Methods Chemicals and Reagents Nifedipine, testosterone, 6β-hydroxytestosterone, progesterone, 17-hydroxyprogesterone, 11-deoxycortisol (RSS), and 11-deoxycorticosterone (DOC) were obtained from Sigma (St. Louis, MO); dehydroepiandrosterone (DHEA), from Acros Organics (Geel, Belgium); 7α-, 7β-, 16α-, and 16β-hydroxy-DHEA, from Steraloids (Newport, USA); midazolam maleate (PZN 4670708), from Roche Diagnostics (Mannheim, Germany); 1-hydroxymidazolam and 4-hydroxymidazolam, from Lipomed (Arlesheim, Switzerland); dextromethorphan hydrobromide, and dextrorphan tartrate, from MP Biomedicals (Irvine, CA, USA); bupropion and hydroxybupropion, from BD Biosciences (San Jose, CA, USA); diclofenac sodium salt, from Fagron (Barsbüttel, Germany); Acetonitrile, from VWR (Darmstadt, Germany); methanol, from Fisher Scientific (Schwerte, Germany); ethyl acetate of analytical grade, from Sigma (St. Louis, MO); luciferin-PFBE, from Promega (Madison, USA); and additives for yeast media, from Sigma, Roth (Karlsruhe, Germany), Fagron, Grüssing (Filsum, Germany), or Fluka (St. Louis, MO). Restriction endonucleases were from Fermentas (St. Leon-Rot, Germany). All other chemicals and reagents used were of the highest grade available. Media and General Techniques General DNA methods were performed using standard techniques [29]. Fission yeast cultivation was done in the Edinburgh minimal medium (EMM) as described [7]. Cell densities were determined using a Neubauer improved counting chamber, and dry biomass was calculated from 1 mL culture samples separated from the medium and dried in a 60 °C incubator for 2 days. Plasmid and Strain Construction The integrative expression vector pCAD1 [30] was used to coexpress hCPR. The cDNA for fission yeast ccr1 (NCBI reference sequence NM_001021956) was PCR amplified using a colony of strain NCYC2036. The primers were 5′-gggaaacatatgaagactatgaatatg-3′ and 5′-

Appl Biochem Biotechnol

tttaaaggatccttaccaggtatcttc-3′ (MWG, Martinsried, Germany) and introduced flanking recognition sites for NdeI and BamHI, respectively. The PCR was done as follows: the thermocycler was Primus 25 (Peqlab, Erlangen, Germany), and concentrations for the PCR reactions were 1 pmol/μL of each primer, 0.2 mM of each dNTP, 1× Pfu polymerase buffer with MgSO4 (Thermo Scientific, St. Leon-Rot, Germany), and 0.02 U/μL Pfu polymerase (added after a denaturation step). The PCR conditions for these reactions were as follows: 95 °C for 10 min, then addition of polymerase, (95 °C for 30 s, 50 °C for 30 s, and 72 °C for 8 min)×30, and final extension at 72 °C for 20 min. After PCR, and the ccr1 fragment was then purified via agarose gel electrophoresis and precipitation with sodium acetate and ethanol. The resulting DNA fragment was then subcloned to a pGEM-T vector for amplification and modification (two internal NdeI recognition sites in ccr1 were removed by site-directed mutagenesis) and subsequently cloned via NdeI/BamHI to pCAD1. The CPR from A. majus (NCBI accession AY532374) was synthesized with codon usage optimized for S. pombe by GeneArt (Regensburg, Germany) and also cloned into pCAD1 as described above for ccr1. For the construction of an empty pCAD1 vector for the generation of the control strain, Escherichia coli K12 ER2925 (dam-) from NEB (Ipswich, Madison, US) was transformed with pCAD1-ccr1, which was then purified, cleaved with BclI (blocked by dam methylation) and BamHI, and then religated yielding a pCAD1 fragment without promoter and insert. Vectors pREP1 [31] and pNMT1-TOPO (Invitrogen) containing the potent inducible nmt1 promoter [32] were chosen for expression of the CYPs in fission yeast, and the cloning strategy was the same as described before [7]. All new plasmids were sequenced by SeqIT (Kaiserslautern, Germany). The novel reductase-expressing strains were tested for correct integration of the pCAD1 fragments into the leu1 gene by replica plating on agar plates without leucine. For the generation of the CYP-coexpressing fission yeast strains, cryocompetent cells of the parental strains CAD62, INA62, INA100, and INA60, respectively, were transformed with the respective pREP1 and pNMT1 constructs bearing the cDNA for the different human CYPs by a transformation procedure as described [33]. The genotypes of all strains used in this study are listed in Table 1. Growth assays were performed using three independent 100 mL cultures of each strain. Samples were taken for at least 68 h, and dry biomass was determined as described above. The growth data were fitted by an exponential growth equation using the free software GLE. Whole-Cell Biotransformations Activities of the novel fission yeast strains with standard substrates as indicated were determined by whole-cell biotransformations at 1 mL scale in 96-well multi-plates. Cells from permanent cultures were streaked on EMM dishes containing leucine and thiamine and grown for 3 days at 30 °C. For precultivation, 10 mL of EMM medium was inoculated with a colony and grown in vertically standing 50-mL tubes at 150 rpm and 30 °C for 24 h. The main cultures were created by inoculation of 100 mL fresh EMM medium in 250-mL Erlenmeyer flasks which were then incubated for 24 h under the same conditions. The cells were harvested by centrifugation (10 min, 5,000×g, 4 °C) and subsequently resuspended in the respective biotransformation medium; 100 mM NaH2PO4/Na2HPO4 buffer, pH 7, with 20 g/L glucose was used for the conversion of dextromethorphan, diclofenac, ibuprofen, midazolam, and nifedipine, respectively. NaH2PO4/Na2HPO4 buffer (100 mM), pH 8, with 20 g/L glucose was used for bupropion hydroxylation and EMM for the biotransformation of progesterone, 17-hydroxyprogesterone, testosterone, and DHEA, respectively. Midazolam and the steroidal substrates were added to a final concentration of 200 μM, while all other

Genotype

h- ura4- dl.18

h- ura4- dl.18 leu1:: pCAD1-hCPR

h- ura4- dl.18 leu1:: pCAD1-ccr1

h- ura4- dl.18 leu1:: pCAD1-AmCPR h- ura4- dl.18 leu1:: pCAD1

h- ura4- dl.18 leu1:: pCAD1-hCPR/pREP1-CYP3A4

h- ura4- dl.18 leu1:: pCAD1-hCPR/pREP1-CYP3A5

h- ura4- dl.18 leu1:: pCAD1-hCPR/pREP1-CYP3A7

h- ura4- dl.18 leu1:: pCAD1-hCPR/pREP1-CYP2D6

h- ura4- dl.18 leu1:: pCAD1-hCPR/pREP1-CYP2C9

h- ura4- dl.18 leu1:: pCAD1-hCPR/pNMT1-CYP17

h- ura4- dl.18 leu1:: pCAD1-hCPR/pNMT1-CYP21 h- ura4- dl.18 leu1:: pCAD1-ccr1/pREP1-CYP3A4

h- ura4- dl.18 leu1:: pCAD1-ccr1/pREP1-CYP3A5

h- ura4- dl.18 leu1:: pCAD1-ccr1/pREP1-CYP3A7

h- ura4- dl.18 leu1:: pCAD1-ccr1/pREP1-CYP2D6

h- ura4- dl.18 leu1:: pCAD1-ccr1/pREP1-CYP2C9

h- ura4- dl.18 leu1:: pCAD1-ccr1/pNMT1-CYP17

h- ura4- dl.18 leu1:: pCAD1-ccr1/pNMT1-CYP21

h- ura4- dl.18 leu1:: pCAD1-AmCPR/pREP1-CYP3A4 h- ura4- dl.18 leu1:: pCAD1-AmCPR/pREP1-CYP3A5

h- ura4- dl.18 leu1:: pCAD1-AmCPR/pREP1-CYP3A7

h- ura4- dl.18 leu1:: pCAD1-AmCPR/pREP1-CYP2D6

h- ura4- dl.18 leu1:: pCAD1-AmCPR/pREP1-CYP2C9

h- ura4- dl.18 leu1:: pCAD1-AmCPR/pNMT1-CYP17

h- ura4- dl.18 leu1:: pCAD1-AmCPR/pNMT1-CYP21

Strain

NCYC2036

CAD62

INA62

INA100 INA60

CAD67

INA20

INA2

CAD64

CAD68

CAD71

CAD75 INA69

INA63

INA70

INA64

INA68

INA30

INA31

INA49 INA50

INA52

INA48

INA46

INA32

INA33

Table 1 Fission yeast strains used in this study

nd

AmCPR, CYP21

AmCPR, CYP17

AmCPR, CYP2C9

AmCPR, CYP2D6

AmCPR, CYP3A7

AmCPR, CYP3A4 AmCPR, CYP3A5

ccr1, CYP21

ccr1, CYP17

ccr1, CYP2C9

ccr1, CYP2D6

ccr1, CYP3A7

ccr1, CYP3A5

hCPR, CYP21 ccr1, CYP3A4

hCPR, CYP17

hCPR, CYP2C9

hCPR, CYP2D6

hCPR, CYP3A7

hCPR, CYP3A5

hCPR, CYP3A4

AmCPR –

ccr1

0.15±0.01

0.16±0.02

0.19±0.01

0.10±0.01

0.15±0.01

0.13±0.02 0.09±0.01

0.17±0.01

0.16±0.02

0.23±0.05

0.09±0.02

0.14±0.01

0.18±0.01

0.19±0.03 0.16±0.03

0.14±0.02

0.22±0.03

0.18±0.01

0.32±0.03

0.12±0.02

0.22±0.02

0.11±0.02 0.12±0.02

0.22±0.04

0.16±0.03

– hCPR

Growth rate (μ h−1)

Expressed proteins

This study

This study

This study

This study

This study

This study This study

This study

This study

This study

This study

This study

This study

This study This study

This study

(7)

(7)

(34)

(11)

(7)

This study This study

This study

(7)

(56)

Reference

Appl Biochem Biotechnol

h- ura4- dl.18 leu1:: pCAD1/pREP1-CYP3A4

h- ura4- dl.18 leu1:: pCAD1/pREP1-CYP3A5

h- ura4- dl.18 leu1:: pCAD1/pREP1-CYP3A7 h- ura4- dl.18 leu1:: pCAD1/pREP1-CYP2D6

h- ura4- dl.18 leu1:: pCAD1/pREP1-CYP2C9

h- ura4- dl.18 leu1:: pCAD1/pNMT1-CYP17

h- ura4- dl.18 leu1:: pCAD1/pNMT1-CYP21

INA78

INA79

INA80 INA83

INA84

INA87

INA88

nd not determined

Genotype

Strain

Table 1 (continued)

CYP21

CYP17

CYP2C9

CYP3A7 CYP2D6

CYP3A5

CYP3A4

Expressed proteins

0.13±0.01

0.12±0.02

0.11±0.01

0.13±0.01 0.18±0.03

0.06±0.01

0.15±0.02

Growth rate (μ h−1)

This study

This study

This study

This study This study

This study

This study

Reference

Appl Biochem Biotechnol

Appl Biochem Biotechnol

substrates were used at 500 μM. Biotransformations were done for 24 h at 30 °C and 750 rpm. Dextromethorphan (with 10 μL 30 % HCl added after biotransformation), nifedipine, and midazolam samples (500 μL each) were centrifuged twice at 10,000×g for 10 min, and the supernatant was used for HPLC analysis. Each sample (500 μL) of the bioconversions of ibuprofen and diclofenac or the steroids was extracted with 1 mL of ethyl acetate after addition of an internal standard. These were testosterone for ibuprofen, progesterone for diclofenac, testosterone and DHEA, cortisol for progesterone, and 17hydroxyprogesterone assays, respectively. Dry biomass determinations were done as described [34]. All experiments were done in triplicates. HPLC and LC-MS Analysis RP-HPLC conditions for the analysis and quantification of DHEA and testosterone were described previously [34]. Briefly, the compounds were resolved on the LiChrospher® RP18 LiChroCART® column (125×4 mm; 5 μm particle size) at 200 nm, and the HPLC system was the Waters 2690 separations module with an UV/Vis detector (Waters 996 PDA). On the same apparatus, midazolam and nifedipine were analyzed, using a gradient method with 10 mM ammonium formate pH 4.5 (A) and methanol (B). The conditions were as follows: 0–5 min 20 % B, 5–20 min linear gradient to 75 % B, 20–25 min 75 % B, 25.1 min back to 20 % B, and 25.1–30 min 20 % B. Midazolam and hydroxymidazolam were detected at 250 nm with retention times of 21.5 and 18.9 min, respectively. Nifedipine and oxidized nifedipine were detected at 260 nm with retention times of 17.7 and 16.4 min, respectively. These substances were separated on the Agilent LiChrospher® 100 RP-18 (125×4 mm) with 5 μm particle size. Diclofenac samples were also analyzed on this apparatus, as described earlier [7]. Ibuprofen was analyzed on the HP1090 separation module as described [35]. RSS, DOC, progesterone, and 17-hydroxyprogesterone analytics were also performed on this device, using the following conditions: RP-18 column (LiChrospher 100, Wicom) 5 μm, 125×4.6 mm (240 nm) 20 μL injections, and flow rate 1 mL/min. The gradient consisted of water (A) and methanol (B) with the following conditions: 0–1 min 48 % A, 1–8 min linear gradient to 37 % A, 8–9.5 min linear gradient to 42 % A, 9.5–14 min 42 % A, 14–15 min linear gradient to 48 % A, and 15–18 min 48 % A. Dextromethorphan samples were analyzed on the Series II HP1090 separation module (Hewlett Packard, Palo Alto, CA) with the C8 Nucleodur EC 125/4 100–3 from MachereyNagel (Düren, Germany) at 40 °C and mobile phase 5 mM ammonium formate pH 4.5 (A) and methanol with 0.1 % ammonium formate (B) with the following conditions: 0–1 min 25 % B, 1–11 min linear gradient to 90 % B, 11–16 min 90 % B, 16–16.1 min to 25 % B, and 16.1–21 min 25 % B. MS analysis was performed using the TF LXQ Linear Ion Trap MS system equipped with a heated electrospray ionization source with the following conditions: fragmentor at 75 V, threshold 150, positive mode API-ES 100–600, dry gas flow 12 L/min, nebulizer press 45 psig, drying gas 345 °C, capillary voltage 4,500 V, and diode array detector at 40 °C and 270 nm. Flow rate of the mobile phase was 0.8 mL/min. The collected data were processed by TF Xcalibur 2.0.7. Statistics Three independent cultures were used in all experiments. The samples were analyzed separately, and the mean of the calculated product formation rates is given with its standard deviation. The p values from a two-sided t test type 3 are indicated by asterisks and refer to the activity of the corresponding strain that expresses hCPR, if not indicated otherwise.

Appl Biochem Biotechnol

Results Strain Construction As electron transfer from NADPH to the CYP enzymes might be a rate-limiting factor during whole-cell biotransformation in recombinant fission yeasts, it was the objective of the present study to determine the whole-cell biotransformation activities of seven human CYPs upon strong coexpression in fission yeast with one of the three CPRs from H. sapiens (hCPR), S. pombe (ccr1), or A. majus (AmCPR), respectively. For comparison, CYPexpressing fission yeast strains without coexpression of a CPR (which bear only the endogenous ccr1 under control of its own promoter) were also tested. Thus, a total of 28 recombinant strains were needed, with five of them having been constructed previously (see Table 1). The 23 new strains were constructed as described in the “Materials and Methods” section following a cloning strategy described earlier [7]. In brief, this cloning strategy is to construct CPR-expressing strains first by integration of pCAD1 vector constructs into the leu1 gene of fission yeast; the resulting strains are then transformed with autosomal replicating expression plasmids bearing the respective CYP cDNAs. Growth Properties of the Newly Engineered Fission Yeast Strains Integrative transformation of the parental strain NCYC2036 with pCAD1 constructs yielded strains CAD62 (expressing hCPR [7]), INA62 (overexpressing ccr1), INA100 (expressing AmCPR), and INA60 (vector control), respectively. Determination of the growth behavior of the four strains under induced conditions revealed that strains INA62 and CAD62 grow somewhat better than strains INA100 and INA60 (Table 1). However, in view of the standard deviations observed, the effect is hardly significant except when comparing strains INA60 and INA62. Obviously, overexpression of ccr1 improves the growth rate of S. pombe. In addition, it was observed that strain INA62 reaches a higher final biomass density than the other three strains (data not shown). The reason for this effect is unclear, but the data show that neither expression of the three CPRs under study has a negative effect on the growth of fission yeast. In accordance with the strategy outlined above, the three CPRexpressing strains as well as the control strain were subsequently transformed with autosomal replicating plasmids containing the cDNAs for different human CYPs to yield a set of new fission yeast strains with different P450 systems. The growth properties of all strains were determined, and it turned out that all strains display growth rates in the same range as the parental strains with two notable exceptions: strain INA2 (coexpressing hCPR and CYP3A7) grows remarkably well, while growth of strain INA79 (expressing CYP3A5 only) is unusually poor (Table 1). In general, CYP3A7-expressing strains grow better than CYP3A5-expressing ones, while CYP3A4 expressors are intermediate in this respect. The reason for this discrepancy within the human CYP3A family is not known. In any case, by normalizing all whole-cell biotransformation data to dried biomass, this effect was eliminated from the comparison of activities. Biotransformation Activity of CYP3A-Expressing Strains with Different Reductases We previously reported that fission yeast strains coexpressing hCPR and either CYP3A4, CYP3A5, or CYP3A7, respectively, efficiently convert a variety of substrates in whole-cell biotransformations [7, 11, 34, 36]. In this study, strains coexpressing the CYP3A enzymes with different reductases were tested for bioconversion of testosterone (all three enzymes)

Appl Biochem Biotechnol

and nifedipine (CYP3A4 and CYP3A5) or midazolam (CYP3A7 only), respectively (Table 2). As expected, the hCPR-coexpressing strains displayed the highest bioconversion activity in all cases, while the activity of the other strains was either significantly lower or even undetectable. DHEA hydroxylation was studied in more details as this substrate yields several different CYP3A metabolites. Previously, we only monitored the formation of the CYP3A7 standard product 16α-hydroxy-DHEA; in this study, we also determined the formation of the 7α- and 7β-hydroxy products as well as that of 16β-hydroxy-DHEA (Fig. 1). Product formation by all three CYP3A enzymes was most efficient with hCPR, and each enzyme displayed a unique metabolite pattern as expected. CYP3A4 coexpressed with hCPR mainly produced 7α-hydroxy-DHEA, no 7β-hydroxy-DHEA, and equal amounts of both 16α- and 16β-hydroxy-DHEA (Fig. 1a). By contrast, formation of 16αand 7β-hydroxy-DHEA was preferred in the absence of a coexpressed reductase. CYP3A5 coexpressed with hCPR also mainly produced 7α-hydroxy-DHEA, with the 16α-hydroxy product being the second most important metabolite and only small amounts of the other two being detected (Fig. 1b). Without reductase coexpression, this pattern also changed somewhat, with equally rapid formation of both 7α- and 16α-hydroxy-DHEA and no production of the others. As could be predicted, coexpression of CYP3A7 with hCPR led to strong production of 16α-hydroxy-DHEA (which plays important roles during fetal development) [37, 38], while the other three metabolites were formed in negligible amounts (Fig. 1c). Even in the absence of a coexpressed reductase, the 16α-hydroxy metabolite was the major product, albeit formed at a much reduced rate. Biotransformation Activity of CYP2C9- or CYP2D6-Expressing Strains with Different Reductases In addition to the CYP3A family, we also tested two further important drug metabolizing human CYPs, namely, CYP2C9 and CYP2D6. While CYP2C9, like the CYP3A enzymes, does not show biotransformation activity in fission yeast without coexpression of hCPR (Table 2), CYP2D6 was previously shown to be sufficiently supported by natural amounts of the endogenous ccr1 during the bioconversion of the illicit drugs 4′-methyl-α-pyrrolidinobutyrophenone [9] and N-(1-phenylcyclohexyl)-3-ethoxypropanamine [39], respectively. These data are corroborated by the findings of the present study as CYP2D6-dependent whole-cell biotransformation of the standard substrate dextromethorphan proceeded equally well with or without coexpression of hCPR (Table 2). Surprisingly, overexpression of either ccr1 or (to a lesser extent) AmCPR significantly enhanced this activity in comparison to hCPR. These data indicate that among the reductases under study, hCPR actually showed the poorest performance in support of CYP2D6. In case of CYP2C9, both 3-hydroxylation of ibuprofen and 4′-hydroxylation of diclofenac were best catalyzed by strain INA46 that coexpresses AmCPR, while coexpression of both hCPR or ccr1 was also helpful but less effective; as before [35], ibuprofen 2-hydroxylation was weaker than 3-hydroxylation, and in this case, improvement by AmCPR coexpression was not significantly better than that of hCPR (Table 2). Biotransformation Activity of CYP17- or CYP21-Expressing Strains with Different Reductases In contrast to the predominantly drug metabolizing enzymes of the CYP2 and CYP3 families, the two human steroid hydroxylases CYP17 and CYP21 have their main role in endogenous steroid biosynthesis [16]; however, at least in the case of CYP21, xenobiotics with a steroidal

Appl Biochem Biotechnol Table 2 Biotransformation activities of recombinant fission yeast strains Enzyme

Reaction

Reductase Biotransformation activity

CYP2C9 Ibuprofen hydroxylation

2-Hydroxylation hCPR

CYP2C9 Diclofenac 4′-hydroxylation

CYP2D6 Dextromethorphan demethylation

CYP3A4 Nifedipine oxidation

CYP3A4 Testosterone 6β-hydroxylation

CYP3A5 Nifedipine oxidation

CYP3A5 Testosterone 6β-hydroxylation

CYP3A7 Midazolam 1-hydroxylation

CYP3A7 Testosterone 6β-hydroxylation

CYP17

CYP21

Progesterone 17α-hydroxylation

Progesterone 21-hydroxylation

3-Hydroxylation

1.0±0.3 μmol/g/h 3.5±0.7 μmol/g/h

ccr1

0.5±0.1 μmol/g/h 2.4±0.4 μmol/g/h

AmCPR None

1.2±0.2 μmol/g/h 5.2±0.8 μmol/g/h* None observed None observed

hCPR

2.1±0.1 μmol/g/h

ccr1

1.9±0.4 μmol/g/h

AmCPR

5.5±0.1 μmol/g/h**

None

None observed

hCPR

1.5±0.2 μmol/g/h

ccr1

3.9±0.1 μmol/g/h*

AmCPR None

3.2±0.2 μmol/g/h 1.9±0.5 μmol/g/h

hCPR

12.1±0.6 μmol/g/h

ccr1

1.9±0.1 μmol/g/h**

AmCPR

4.0±0.9 μmol/g/h***

None

0.9±0.5 μmol/g/h***

hCPR

13±3 nmol/g/h

ccr1

0.10±0.03 nmol/g/h*

AmCPR None

2.2±0.3 nmol/g/h* None observed

hCPR

4.3±0.4 μmol/g/h

ccr1

0.3±0.1 μmol/g/h**

AmCPR

1.3±0.7 μmol/g/h*

None

0.3±0.2 μmol/g/h***

hCPR

2.8±0.5 μmol/g/day

ccr1

None observed

AmCPR None

0.8±0.3 μmol/g/day* None observed

hCPR

21±1 nmol/g/h

ccr1

None observed

AmCPR

7±1 nmol/g/h**

None

None observed

hCPR

1.7±0.2 nmol/g/h

ccr1

None observed

AmCPR None

None observed None observed

hCPR

11±2 μmol/g/day

ccr1

9±1 μmol/g/day

AmCPR None

10±3 μmol/g/day 6±2 μmol/g/day*

hCPR

1.6±0.5 μmol/g/day

ccr1

2.1±0.2 μmol/g/day

AmCPR

0.75±0.05 μmol/g/day

Appl Biochem Biotechnol Table 2 (continued) Enzyme

Reaction

Reductase Biotransformation activity

CYP21

17α-Hydroxy progesterone 21-hydroxylation hCPR

None

0.09±0.02 μmol/g/day* 2.1±0.7 μmol/g/day

ccr1 AmCPR

2.1±0.8 μmol/g/day 1.0±0.1 μmol/g/day

None

0.4±0.1 μmol/g/day*

Assay conditions are described in the “Materials and Methods” section. The substrate concentrations were as follows: ibuprofen, diclofenac, dextromethorphan, and nifedipine 500 μM, and midazolam, progesterone, and 17-hydroxyprogesterone 200 μM. Data are from n=3 assays shown with standard deviations *P<0.05 in comparison to the corresponding hCPR-expressing strain; **P<0.01; ***P<0.005

structure may also serve as substrates [36, 40]. We previously reported that (as in the case of CYP2D6) upon recombinant expression in S. pombe, both CYP17 [41] and CYP21 [42] are effectively supported by natural amounts of the endogenous ccr1. In the present study, we monitored standard reactions of both enzymes, namely, 17-hydroxylation of progesterone by CYP17 and 21-hydroxylation of either progesterone or 17-hydroxyprogesterone by CYP21, to test for a possible enhancement of biotransformation activity by coexpressed reductases. Interestingly, coexpression of any of the three reductases was more or less equally helpful in promoting CYP17 activity (Table 2). By contrast, CYP21 activity towards both progesterone and 17α-hydroxyprogesterone was significantly enhanced by coexpression of either hCPR or ccr1, while AmCPR coexpression was less effective.

Discussion Fission yeast strains that recombinantly express human CYPs have been successfully applied for metabolite production by whole-cell biotransformation in many cases (e.g., [7, 9, 35, 43]). As with all such recombinant expression systems, a promising host organism does not only yield successful results in the beginning but also allows for efficient means of optimization. Previous studies aimed at the improvement of CYP metabolite production by recombinant S. pombe targeted the improvement of the reaction conditions [7, 10] and the molecular engineering of the CYPs involved. In the present study, we tried to enhance the activity of human microsomal CYPs by coexpression of a CPR and to compare the usefulness of hCPR, ccr1, and AmCPR, respectively, in this regard. For this purpose, 23 new recombinant fission yeast strains were cloned that express human CYPs and/or one of the three reductases in combinations that previously were not available (Table 1). Overall, the growth properties of the new strains were in the range of expectations, and growth differences between the strains were accounted for by normalizing all whole-cell biotransformation data to dried biomass. Obviously, coexpression of hCPR should enhance the activity of any human microsomal CYP enzyme recombinantly expressed in a microbe, and we have previously demonstrated this to be true for human CYP2B6, CYP2C9, CYP2C19, CYP2D6, CYP3A4, CYP3A5, and CYP3A7, respectively [7, 11, 34, 44]. In this study, we sought to extend this list and compared the efficiency of hCPR coexpression to that of other reductases. While both CYP17 and CYP21 were previously shown to be effectively supported by natural amounts of the endogenous ccr1 [41, 42], coexpression of hCPR with either enzyme is reported here for the first time. CYP17-

Appl Biochem Biotechnol

Appl Biochem Biotechnol

ƒFig. 1

Whole-cell biotransformation of DHEA by fission yeast strains coexpressing human CYP3A enzymes and different reductases as indicated. a CYP3A4-dependent biotransformation. b CYP3A5-dependent biotransformation. c CYP3A7-dependent biotransformation. Assay conditions are described in the “Materials and Methods” section. The substrate concentration was 200 μM. Data are from n=3 assays shown with standard deviations. *P<0.05 in comparison to the corresponding hCPR-expressing strain if not indicated otherwise; **P<0.01; ***P<0.005

dependent whole-cell biotransformation of progesterone to 17α-hydroxyprogesterone was significantly increased by coexpression of hCPR, as well as by coexpression of the other two reductases under study (Table 2). CYP21-catalyzed bioconversion of either progesterone or 17hydroxyprogesterone was also enhanced by hCPR coexpression, with ccr1 performing equally well and AmCPR being less effective. The members of the human CYP3A family benefited from hCPR coexpression even more than the two steroid hydroxylases, and neither ccr1 nor AmCPR could support them equally well (Table 2 and Fig. 1). Interestingly, the product pattern observed in the CYP3A4-catalyzed hydroxylation of DHEA appears to depend strongly on the availability of reduction equivalents. Upon hCPR coexpression, 7α-hydroxy-DHEA was the major metabolite, followed by approximately equal amounts of both 16α- and 16β-hydroxyDHEA and, surprisingly, no 7β-hydroxy-DHEA (Fig. 1a). By contrast, formation of 16α- and 7β-hydroxy-DHEA was preferred in the strain without reductase coexpression, while only minute amounts of 7α- and 16β-hydroxy-DHEA were detected. These data corroborate the earlier findings indicating that the activity of the CPR in the human microsomal CYP system plays a physiologically important role for steroid biosynthesis [45] and drug clearance [46]. Some CPR polymorphisms have been shown to influence CYP activity; for example, the common A503V variant accounts for 26 % of mutated CPR in Caucasians [47] and leads to a 53 % reduced activity of CYP2D6 towards dextromethorphan and bufuralol [48]; it also reduces the 17α-hydroxylase activity of CYP17 by 32 % and its 17,20-lyase activity by 42 % in comparison to the wild type [49]. At the same time, this CPR variant displays normal activity with CYP21 [50] and with CYP1A2 and CYP2C19 [51], while its effect on CYP2C9 depends on this CYPs’ own polymorphisms [52]. In the case of CYP3A4, modulation of enzyme activity in a substrate-specific manner by hCPR variants has also been reported [53]. Taken together, these data as well as the findings of the present study strongly suggest that—at least for some human CYP enzymes—the reductase does not only influence the reaction rate by providing electrons but also exerts an influence on the metabolite pattern. In this context, in the earlier studies where CYP3A4 activity on DHEA was monitored, formation of 7α-hydroxyDHEA might have been compromised by low reductase activities [38, 54]. From a biotechnological point of view, the data shown in Fig. 1 demonstrate how the question of whether to coexpress a reductase or not (and if so, which one) can help in striving for the preferential synthesis of any one the different hydroxy-DHEA metabolites. Concerning CYP2C9 activity, coexpression of a reductase is absolutely required for the biotransformation of either ibuprofen or diclofenac (Table 2). Surprisingly, hCPR did not turn out to be the reductase of choice, as significantly better conversion rates were obtained with AmCPR. In contrast to expectations, hCPR coexpression did not improve CYP2D6 activity on its standard substrate dextromethorphan at all. In summary, out of the seven human CYPs tested, the three CYP3A enzymes were most efficiently supported by hCPR coexpression, and for another two, hCPR was equally helpful as ccr1 (CYP17 and CYP21) or AmCPR (CYP17 only), respectively. Although these results support our initial assumption, the advantage gained by hCPR coexpression over that of ccr1 or AmCPR is not as large as one might have assumed in view of the considerable evolutionary distance between the three reductases, even though the almost exclusive production of a single metabolite (16α-hydroxy-DHEA by strain INA2) remains a rare exception.

Appl Biochem Biotechnol

As outlined above, earlier results from others [17] and us [12] encouraged us to investigate the possible benefits of overexpression of the endogenous CPR homologue ccr1 during a CYP-dependent whole-cell biotransformation in fission yeast. This protein has never been recombinantly expressed and functionally characterized so far. Previous data on metabolite production by some human CYPs recombinantly expressed in S. pombe without reductase coexpression strongly suggested that ccr1—being the only CPR homologue in this organism—can functionally interact with CYP17 [41], CYP21 [42], and CYP2D6 [9]. In those instances where no activity was detected, it was unclear whether this was simply due to a low expression level of ccr1 or to its inability to interact with these CYPs. The data obtained in this study show that—depending on the enzyme under study—either explanation may be true, although among the seven human CYPs tested, there was only one enzyme (CYP3A7) that hardly showed any activity when coexpressed with ccr1, as only small amounts of DHEA metabolites could be identified (Table 2 and Fig. 1). Among the other six CYPs, only CYP2C9 was not active without reductase coexpression but did show activity upon ccr1 coexpression, while CYP3A4 and CYP3A5 both displayed activity towards one substrate (nifedipine), but not towards another (testosterone) without coexpression of a reductase; even upon ccr1 coexpression, the gains in biotransformation activity were minute in these cases. Overall, the functional interaction of ccr1 with the CYP3A enzymes was poor, regardless of the isoenzyme or the substrate tested. By contrast, the three CYPs listed above were those that yielded the highest biotransformation activities without coexpression of a reductase (Table 2). In addition, CYP21 and CYP2D6 also significantly benefited from ccr1 overexpression, which in one instance (demethylation of dextromethorphan by CYP2D6) was even more efficient than coexpression of hCPR. Thus, the general impression is that strains coexpressing ccr1 with a human CYP tend to show less biotransformation activity than the corresponding strains with hCPR coexpression, although there are some exceptions. For the purpose of monitoring CYP activity in vitro, it is a common strategy to construct artificial systems containing the CYP enzyme itself and CPRs from various origins [4]. For example, reconstitution assays with the taxoid 10β-hydroxylase from the Japanese yew Taxus cuspidata using the reductases from yew, spearmint, mung bean, or rabbit, respectively, showed that the animal reductase had the lowest efficiency, while there were no significant differences between the various plant reductases [55]. Similarly, it has been suggested that the ability of CPRs from different species or even from different kingdoms to at least partially complement each other functionally makes CPR an ideal bio-brick in synthetic biology based on the use of the “share your parts” principle to secure electron transfer between enzyme complexes that in nature are not known to be connected [23]. On the basis of these considerations, it may not seem strange to recombinantly coexpress a plant CPR with a human CYP in a yeast species in order to clone an efficient strain for whole-cell biotransformations. As mentioned above, we employed the sole known CPR of the bishop’s weed A. majus (AmCPR) for this purpose. Interestingly, coexpression of AmCPR enhanced CYP biotransformation activity in most cases studied except for some of the CYP3Acatalyzed steroid hydroxylations (Table 2 and Fig. 1). Moreover, it supported ibuprofen 2hydroxylation by CYP2C9 and progesterone 17α-hydroxylation by CYP17 similarly well as hCPR, while it performed even better than hCPR (but not as well as ccr1) during the demethylation of dextromethorphan by CYP2D6. In addition, AmCPR was the best reductase for ibuprofen 3-hydroxylation and diclofenac 4′-hydroxylation by CYP2C9. In a direct comparison to ccr1 coexpression, AmCPR coexpression was more efficient than that of ccr1 in 14 out of 25 individual reactions monitored, while in another seven cases, their effect was equal, and four times ccr1 coexpression was better. Taken together, these results are very

Appl Biochem Biotechnol

encouraging, and it can be expected that other plant CPRs will also cooperate well with human CYPs when recombinantly coexpressed in microbes.

Conclusion This study shows that the choice of the electron transfer partner significantly influences—and typically enhances—the activity of human CYPs recombinantly expressed in fission yeast. More specifically, coexpression of either hCPR, ccr1, or AmCPR, respectively, is a promising possibility to improve whole-cell biotransformation rates. While hCPR remains the reductase of choice, especially when dealing with members of the human CYP3A family, both ccr1 and AmCPR are good candidates for the support of other human CYPs. Acknowledgments This work was supported by a grant (ChemBioTec 13220–32) from the Deutsche Bundesstiftung Umwelt (DBU). The authors thank Călin-Aurel Drăgăn and Jérémy Rimbon for their expert technical help.

References 1. Bernhardt, R. (2006). Journal of Biotechnology, 124, 128–145. 2. Guengerich, F. P. (2008). Chemical Research in Toxicology, 21, 70–83. 3. Larbat, R., Kellner, S., Specker, S., Hehn, A., Gontier, E., Hans, J., et al. (2007). Journal of Biological Chemistry, 282, 542–554. 4. Hannemann, F., Bichet, A., Ewen, K. M., & Bernhardt, R. (2007). Biochimica et Biophysica Acta, 1770, 330–344. 5. Guengerich, F. P., & Cheng, Q. (2011). Pharmacological Reviews, 63, 684–699. 6. Park, B. K., Boobis, A., Clarke, S., Goldring, C. E., Jones, D., Kenna, J. G., et al. (2011). Nature Reviews Drug Discovery, 10, 292–306. 7. Dragan, C.A., Peters, F.T., Bour, P., Schwaninger, A.E., Schaan, S.M., Neunzig, I., Widjaja, M., Zapp, J., Kraemer, T., Maurer, H.H., and Bureik, M. (2011) Appl Biochem Biotechnol 8. Zollner, A., Buchheit, D., Meyer, M. R., Maurer, H. H., Peters, F. T., & Bureik, M. (2010). Bioanalysis, 2, 1277–1290. 9. Peters, F. T., Dragan, C. A., Wilde, D. R., Meyer, M. R., Zapp, J., Bureik, M., et al. (2007). Biochemical Pharmacology, 74, 511–520. 10. Zehentgruber, D., Dragan, C. A., Bureik, M., & Lütz, S. (2010). Journal of Biotechnology, 146, 179–185. 11. Neunzig, I., Widjaja, M., Dragan, C. A., Peters, F. T., Maurer, H. H., & Bureik, M. (2012). Applied Biochemistry and Biotechnology, 168, 785–796. 12. Hakki, T., Zearo, S., Dragan, C. A., Bureik, M., & Bernhardt, R. (2008). Journal of Biotechnology, 133, 351–359. 13. Haas, E., Horecker, B. L., & Hogness, T. R. (1940). Journal of Biological Chemistry, 136, 747–774. 14. Wang, M., Roberts, D. L., Paschke, R., Shea, T. M., Masters, B. S., & Kim, J. J. (1997). Proceedings of the National Academy of Sciences of the United States of America, 94, 8411–8416. 15. Xia, C., Panda, S. P., Marohnic, C. C., Martasek, P., Masters, B. S., & Kim, J. J. (2011). Proceedings of the National Academy of Sciences of the United States of America, 108, 13486–13491. 16. Miller, W. L., & Auchus, R. J. (2011). Endocrine Reviews, 32, 81–151. 17. Szczebara, F. M., Chandelier, C., Villeret, C., Masurel, A., Bourot, S., Duport, C., et al. (2003). Nature Biotechnology, 21, 143–149. 18. Miles, J. S. (1992). Biochemistry Journal, 287, 195–200. 19. Wood, V., Gwilliam, R., Rajandream, M. A., Lyne, M., Lyne, R., Stewart, A., et al. (2002). Nature, 415, 871–880. 20. Fang, Y., Hu, L., Zhou, X., Jaiseng, W., Zhang, B., Takami, T., et al. (2012). Antimicrobial Agents and Chemotherapy, 56, 1949–1959. 21. Hughes, A. L., Powell, D. W., Bard, M., Eckstein, J., Barbuch, R., Link, A. J., et al. (2007). Cell Metabolism, 5, 143–149. 22. Morant, M., Bak, S., Moller, B. L., & Werck-Reichhart, D. (2003). Current Opinion in Biotechnology, 14, 151–162.

Appl Biochem Biotechnol 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39.

40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56.

Jensen, K., & Moller, B. L. (2010). Phytochemistry, 71, 132–141. Mizutani, M., & Ohta, D. (1998). Plant Physiology, 116, 357–367. Ro, D. K., Ehlting, J., & Douglas, C. J. (2002). Plant Physiology, 130, 1837–1851. Bak, S., Olsen, C. E., Halkier, B. A., & Moller, B. L. (2000). Plant Physiology, 123, 1437–1448. Tattersall, D. B., Bak, S., Jones, P. R., Olsen, C. E., Nielsen, J. K., Hansen, M. L., et al. (2001). Science, 293, 1826–1828. Selim, Y. A., & Ouf, N. H. (2012). Org Med Chem Lett, 2, 1. Sambrook, J. R., & Russell, D. W. (2001). Molecular cloning: a laboratory manual. Woodbury: CSHL. Dragan, C. A., Zearo, S., Hannemann, F., Bernhardt, R., & Bureik, M. (2005). FEMS Yeast Research, 5, 621–625. Maundrell, K. (1993). Gene, 123, 127–130. Maundrell, K. (1990). Journal of Biological Chemistry, 265, 10857–10864. Suga, M., & Hatakeyama, T. (2005). Yeast, 22, 799–804. Neunzig, I., Dragan, C. A., Widjaja, M., Schwaninger, A. E., Peters, F. T., Maurer, H. H., et al. (2011). Biochimica et Biophysica Acta, 1814, 161–167. Neunzig, I., Gohring, A., Dragan, C. A., Zapp, J., Peters, F. T., Maurer, H. H., et al. (2012). Journal of Biotechnology, 157, 417–420. Parr, M. K., Zöllner, A., Fusshöller, G., Opfermann, G., Schlörer, N., Zorio, M., et al. (2012). Toxicology Letters, 213, 381–391. Kitada, M., Kamataki, T., Itahashi, K., Rikihisa, T., & Kanakubo, Y. (1987). Journal of Biological Chemistry, 262, 13534–13537. Stevens, J. C., Hines, R. N., Gu, C., Koukouritaki, S. B., Manro, J. R., Tandler, P. J., et al. (2003). Journal of Pharmacology and Experimental Therapeutics, 307, 573–582. Peters, F.T., Schwaninger, A.E., Sauer, C., Fritschi, G., Dragan, C.A., Bureik, M., & Maurer, H. H. (2007) Isolation and purification of the designer drug metabolite O-desethyl-N-(1-phenylcyclohexyl)-3ethoxypropanamine (O-desethyl-PCEPA) biotechnologically synthesized using fission yeast expressing CYP2D6. In: F. Pragst & R. Aderjan (Eds.), XV. GTFCh–Symposium, Gesellschaft für Toxikologische und Forensische Chemie, Mosbach, Germany Zöllner, A., Parr, M. K., Dragan, C. A., Dras, S., Schlorer, N., Peters, F. T., et al. (2010). Biological Chemistry, 391, 119–127. Dragan, C. A., Hartmann, R. W., & Bureik, M. (2006). Journal of Enzyme Inhibition and Medicinal Chemistry, 21, 547–556. Dragan, C. A., Blank, L. M., & Bureik, M. (2006). Yeast, 23, 779–794. Bureik, M., Schiffler, B., Hiraoka, Y., Vogel, F., & Bernhardt, R. (2002). Biochemistry, 41, 2311–2321. Peters, F. T., Dragan, C. A., Schwaninger, A. E., Sauer, C., Zapp, J., Bureik, M., et al. (2009). Forensic Science International, 184, 69–73. Marohnic, C. C., Panda, S. P., McCammon, K., Rueff, J., Masters, B. S., & Kranendonk, M. (2010). Drug Metabolism and Disposition, 38, 332–340. Tomalik-Scharte, D., Maiter, D., Kirchheiner, J., Ivison, H. E., Fuhr, U., & Arlt, W. (2010). European Journal of Endocrinology, 163, 919–924. Huang, N., Agrawal, V., Giacomini, K. M., & Miller, W. L. (2008). Proceedings of the National Academy of Sciences of the United States of America, 105, 1733–1738. Sandee, D., Morrissey, K., Agrawal, V., Tam, H. K., Kramer, M. A., Tracy, T. S., et al. (2010). Pharmacogenetics and Genomics, 20, 677–686. Huang, N., Pandey, A. V., Agrawal, V., Reardon, W., Lapunzina, P. D., Mowat, D., et al. (2005). American Journal of Human Genetics, 76, 729–749. Gomes, L. G., Huang, N., Agrawal, V., Mendonca, B. B., Bachega, T. A., & Miller, W. L. (2008). Journal of Clinical Endocrinology and Metabolism, 93, 2913–2916. Agrawal, V., Huang, N., & Miller, W. L. (2008). Pharmacogenetics and Genomics, 18, 569–576. Subramanian, M., Agrawal, V., Sandee, D., Tam, H. K., Miller, W. L., & Tracy, T. S. (2012). Pharmacogenetics and Genomics, 22, 590–597. Agrawal, V., Choi, J. H., Giacomini, K. M., & Miller, W. L. (2010). Pharmacogenetics and Genomics, 20, 611–618. Miller, K. K., Cai, J., Ripp, S. L., Pierce, W. M., Jr., Rushmore, T. H., & Prough, R. A. (2004). Drug Metabolism and Disposition, 32, 305–313. Jennewein, S., Park, H., DeJong, J. M., Long, R. M., Bollon, A. P., & Croteau, R. B. (2005). Biotechnology and Bioengineering, 89, 588–598. Losson, R., & Lacroute, F. (1983). Cell, 32, 371–377.