Towards Elucidating the Molecular Determinants of Efavirenz

Home , Namandje Bumpus, Nicotinamide

TOWARDS ELUCIDATING THE MOLECULAR DETERMINANTS OF EFAVIRENZ

METABOLISM BY CYTOCHROME P450 2B6

by Philip M. Cox

A dissertation submitted to Johns Hopkins University in conformity with the

requirements for the degree of Doctor of Philosophy

Baltimore, Maryland

October, 2016

Abstract:

The polymorphic enzyme cytochrome P450 2B6 (CYP2B6) participates in the metabolism of many clinically relevant drugs. Among these drugs is the anti-HIV drug efavirenz (EFV). In individuals with low activity CYP2B6 phenotypes, EFV plasma levels can be elevated, a fact that is troublesome given the association of EFV with cell death.

Indeed, CYP2B6 phenotype can be difficult to predict in a large population, which makes drugs primarily metabolized by CYP2B6 suboptimal candidates for wide distribution. To date, little experimental data addressing the structural characteristics of a molecule that render it a substrate for CYP2B6 are available. In order to address this knowledge gap, we performed in vitro experiments with EFV, structural analogues of EFV, cDNA expressed cytochrome P450s, and liquid chromatography tandem mass spectrometry or UV detection in order to determine the regions of EFV that are important for metabolism by CYP2B6.

We first analyzed the metabolism of analogues with gross structural changes compared to

EFV. This analysis resulted in identifying the oxazinone ring of EFV as an important moiety for metabolism by CYP2B6, though changes to other parts of the molecule resulted in altered kinetic constants. We then proceeded to address the importance of each heteroatom within the oxazinone ring. Interestingly, we found that alterations of even single heteroatoms resulted in differences in KM values for metabolism by CYP2B6.

Though each of these analogues with altered heteroatoms were still substrates for CYP2B6, the KM values for metabolite formation ranged from 0.3 – 3.9-fold different from EFV.

These differences suggest altered substrate binding by CYP2B6 simply from single heteroatom changes in EFV. The results presented in this thesis provide insight into the functional interaction of CYP2B6 and EFV and lay the groundwork for a better

ii understanding of cytochrome P450 activity towards EFV.

Readers: Namandjé Bumpus (advisor) and Peter Espenshade

iii

Acknowledgements

I have many people and communities to thank for their support throughout graduate school. Firstly, I would like to thank the BCMB program, the Pharmacology department,

Carolyn Machamer, Arhonda Gogos, Phil Cole, Amy Paronto, and many others for their scientific and interpersonal support throughout graduate school. I would particularly like to thank my thesis committee: Joseph Bressler, Jun Liu, and Peter Espenshade for their encouragement and scientific advisory.

I would also like to thank Dr. Namandjé Bumpus for overseeing my research and providing creative ideas, innovative approaches, and generous experimental resources.

Thank you also for demonstrating hard work and devotion in your everyday interactions with me and other students. Your focus and determination sets a strong example for us all.

Thirdly, I am grateful for the other students and post-docs in the Bumpus lab: Julie,

Carley, Dominique, Erin, and many others who rotated or passed through the lab during the past 4 years. I am happy to know each of you, and I hope that we continue our friendships for many years to come.

My family has also played a large role in my success. I am grateful to my parents and my brother for their loving support that has been invaluable to me.

Finally, I would like to thank my wife, Erin Cox, for her unwavering belief in my ability to chase my dreams. Our individual graduate programs have been challenging for both of us, not to mention dealing vicariously with each other’s programs! I am so thankful for your genuine interest in my work and for your desire to help me at every step. I wouldn’t have made it here without you.

Table of Contents

Page

Abstract ...... ii

Acknowledgements ...... iv

Table of Contents ...... v

List of Tables ...... vii

List of Figures ...... viii

Abbreviations ...... x

Chapter 1: Introduction ...... 1

The Cytochromes P450 ...... 2

Drug Metabolism by the Cytochromes P450 ...... 7

Cytochrome P450 2B6 ...... 10

Efavirenz ...... 13

Chapter 2: Structure-activity studies reveal the oxazinone ring is a determinant of cytochrome P450 2B6 activity towards efavirenz...... 15

Introduction ...... 16

Materials and Methods ...... 17

Results and Discussion ...... 22

Chapter 3: Single heteroatom substitutions in the efavirenz oxazinone ring impact metabolism by CYP2B6 ...... 39

Introduction ...... 40

Results ...... 41

Discussion ...... 57

Experimental Section ...... 62

Chapter 4: Conclusions ...... 68

References ...... 72

Curriculum Vitae ...... 84

List of Tables

Page

Table 2-1. EFV analogue metabolites formed from cDNA-expressed P450s ...... 30

Table 2-2. Kinetic constants estimated from substrate depletion experiments ...... 34

Table 3-1. Selected reaction monitoring transitions used in detection of EFV, 1-4, and metabolites via triple quadrupole mass spectrometry...... 51

Table 3-2. Kinetic constants for product formation by CYP2B6 ...... 66

vii

List of Figures

Page

Figure 1-1. Difference spectrum of cytochrome P450 in liver microsomes ...... 3

Figure 1-2. The P450 catalytic cycle ...... 6

Figure 1-3. Major human P450 isoforms in the families 1-3 participating in drug metabolism ...... 8

Figure 1-4. Structures of fifteen CYP2B6 substrates ...... 12

Figure 2-1. Structures of efavirenz (EFV) and EFV analogues 1-8...... 18

Figure 2-2. Oxidative metabolites of EFV analogues 1-6 and 8 formed by human liver microsomes ...... 23

Figure 2-3. Mass spectra for EFV analogue metabolites identified using human liver microsomes ...... 24

18 Figure 2-4. Mass spectra from analogue 8 incubations with H2O and human liver microsomes ...... 28

Figure 2-5. Cytochrome P450-catalyzed metabolism schemes for EFV analogues 1-6 and

8...... 31

Figure 2-6. Substrate depletion analysis of EFV and EFV analogues 3, 4, and 8 ...... 33

Figure 2-7. Inhibition of CYP2B6 bupropion hydroxylase activity by EFV and EFV analogues...... 37

Figure 3-1. Efavirenz (EFV) and analogues 1-5 used in this study ...... 43

Figure 3-2. Extracted ion chromatograms of metabolites formed from incubations with human liver microsomes and analogues 1-4 and UV chromatogram from incubations with

5...... 44

viii

Figure 3-3. Mass spectra and proposed fragment assignments for metabolites formed from 1-4 ...... 45-46

Figure 3-4. Intensities of metabolites formed from incubations with 1-5 ...... 49

Figure 3-5. Chromatograms representing CYP2B6 + b5 and CYP2B1 + b5 activity against EFV and EFV analogues ...... 53-54

Figure 3-6. Comparison of metabolite formation by CYP2B1 and CYP2B6 ...... 55

Figure 3-7. Increased sensitivity and selectivity using selected reaction monitoring for detection of M1 from (S) and (R)-5-Chloro-α-(cyclopropylethynyl)-2-amino-α-

(trifluoromethyl)benzenemethanol...... 56

Abbreviations 8-hydroxyefavirenz (8OH-EFV)

Cytochrome P450 (P450)

Cytochrome P450 2B6 (CYP2B6)

Depletion rate constant (kdep)

Efavirenz (EFV)

Endoplasmic reticulum (ER)

Mass to charge ratio (m/z)

Maximum depletion rate constant (kdep max)

Maximum reaction velocity (Vmax)

Michalis constant (KM)

Nicotinamide adenine dinucleotide phosphate hydrate (NADPH)

P450 oxidoreductase (POR)

Retention time (RT)

Selected reaction monitoring (SRM)

Ultra high-performance liquid chromatography tandem mass spectrometry (uHPLC-

MS/MS)

Chapter 1:

Introduction

The Cytochromes P450

The cytochromes P450 (P450s) are a large family of monooxygenases that are present in organisms from across all kingdoms of life. As of April 2016, more than 35,000

P450s have been identified (Nelson, 2016). Over 20,000 P450s have been found in various plants and fungi, while nearly 10,500 are present across the animal kingdom. The remaining 3,500 P450s are found in diverse organisms such as bacteria, protozoa and even viruses. Although the evolutionary origin of these rather ubiquitous enzymes is the subject of some debate, it is thought that the primordial role of P450s was to metabolize sterols for plasma membrane synthesis (Omura, 2013). Over the course of evolutionary time, P450- catalyzed reactions paved the way for the development of complex metabolic pathways in animals and plants. For example, P450s are involved in biosynthesis of plant components that are thought to have allowed for plants to transition from growth in water to growth on land (Boerjan et al., 2003; Pinot and Beisson, 2011). Thus P450s play important roles in organisms throughout the tree of life.

The cytochromes P450 were first discovered by Omura and Sato in the 1960s

(Omura and Sato, 1964a; Omura and Sato, 1964b). Their work demonstrated the existence of a microsomal protein that absorbed light at 450 nm when in the reduced state in the presence of carbon monoxide (Figure 1-1). At the time, it was thought that this pigment was a single protein; however, over the years the field has expanded to include many thousands of P450s. To organize the growing number of P450s that were being discovered, a nomenclature system was developed (Nebert et al., 1987). Based on amino acid sequence identity, this system separated P450s into gene families and subfamilies. In order to be in the same gene family, P450s must have at least 40% amino acid sequence identity. To be

Figure 1-1. Difference spectrum of cytochrome P450 in liver microsomes.

Microsomes from rat liver were reduced with sodium dithionite and saturated with CO prior to measuring absorbance between 400 and 650 nm. Figure from (Omura and Sato,

1964a). This research was originally published in The Journal of Biological Chemistry.

Omura T and Sato R. The Carbon Monoxide-Binding Pigment of Liver Microsomes. I.

Evidence for Its Hemoprotein Nature. Journal of Biological Chemistry. 1964;

239:2370-2378© the American Society for Biochemistry and Molecular Biology.

3 in the same subfamily, 55% amino acid sequence identity is required. The gene family name is designated by the first number of the P450 name, whereas the subfamily notation is provided by the following letter. The final number is assigned based on the order of identification. For example, cytochrome P450 2B6 (CYP2B6) is from the gene family “2”, the subfamily “B”, and was the sixth member of the 2B family to be identified. Thus, information about the sequence identity of P450s is gleaned simply by analyzing the name.

In addition to the nomenclature based on amino acid sequence identity, an additional designation is assigned to naturally occurring variants of P450 gene sequences.

These variants are called star (*) alleles. In order to be named as a star allele, the variation in gene sequence must result in a change in protein sequence, splicing, translation, transcription, or post transcriptional/translational modifications (Sim and Ingelman-

Sundberg, 2013). The most commonly observed allele, also called the consensus allele, is assigned as *1 (Sim and Ingelman-Sundberg, 2010). As an example, CYP2B6*1 is the consensus allele while CYP2B6*6 is a variant form of CYP2B6.

The determination of P450 protein structure has been a topic of research for many years. P450s contain a single N-terminal hydrophobic domain, which anchors the enzyme to the outer membrane of the endoplasmic reticulum (ER). Although some P450s are anchored to the mitochondrial or to the chloroplast membrane (Johnson and Stout, 2013), our discussion here will focus on P450s anchored to the ER (also known as microsomal

P450s). With the transmembrane domain inserted into the ER, the active site of microsomal

P450s faces the cytosol of the cell. Early crystallographic work with a bacterial P450 revealed the protein to adopt a triangular prism fold (Ravichandran et al., 1993), which was later found to be conserved in mammalian microsomal P450s (Johnson and Stout, 2013;

Williams et al., 2000). This fold consists of 12 conserved α-helices, two β-sheet domains, and a heme prosthetic group within the central enzyme active site.

The diverse functional consequences of P450 catalysis stem from a conserved enzymatic reaction involving the heme iron. This reaction is called the P450 catalytic cycle

(Figure 1-2 steps 1-7). In the resting state, the iron atom is coordinated by six ligands: four pyrole nitrogens in the same plane as the iron atom, a conserved cysteine side chain on one face, and a water molecule on the opposite face. In such a configuration, the heme iron is in the ferric state (Fe+3). Upon the entrance of a substrate into the system (1), the water molecule is expelled, and the substrate now coordinates the heme iron. In order for the first reduction to occur, mammalian P450s require the activity of another microsomal protein known as P450 oxidoreductase (POR). POR accepts an electron from nicotinamide adenine dinucleotide phosphate hydrate (NADPH) and transfers this reducing equivalent to the heme iron (2), resulting in the formation of ferrous heme (Fe+2). Ferrous heme then binds molecular dioxygen forming a ferric superoxide complex (3). This species is then further reduced and protonated to form the ferric hydroperoxo complex (4-5a). Upon a second protonation step (5b), the O-O bond is cleaved producing a reactive species that abstracts a hydrogen from the substrate. The products of this reaction are a substrate radical and the

Fe (IV)-hydroxide complex. The hydroxide group from the heme iron combines with the substrate radical leaving oxygenated product and ferric heme (6). Oxygenated product

(ROH) is then released (7), and the enzyme is poised for another round of catalysis.

Drug Metabolism by Cytochrome P450s

Upon oral administration of a drug, the components of the drug travel through the stomach to the intestinal tract where they are absorbed through the intestinal wall and enter

Figure 1-2. The P450 catalytic cycle. The four pyrole nitrogens are shown as a parallelogram. The substrate and cysteine sidechains are represented as RH and Cys-S, respectively. Figure reproduced from (Shebley et al., 2009) with permission.

6 the hepatic portal vein. The contents of the portal vein pass through the liver prior to reaching systemic circulation. Along this journey, these orally administered drugs are exposed to various enzymes, of which some can recognize the drug as a substrate. The cytochromes P450 recognize many drugs and xenobiotics as substrates and insert oxygens into their chemical structures. P450s of the gene families 1-3 tend to be associated with drug metabolism, although there are a few members of these families that have endogenous substrates as well (Nebert and Russell, 2002). Figure 1-3 summarizes the contribution of some human P450s to the metabolism of clinically used drugs.

As many drugs are lipophilic in nature, the insertion of an oxygen atom can have a large effect on the polarity of the molecule. The end result of metabolism by P450s is a more polar molecule that is more readily excretable than the parent compound. For this reason, the activity of P450s can be directly related to the clearance of a drug from the body. The rate of clearance of a drug from the body informs decisions regarding the frequency of dosing for a given drug. Proper dosing is necessary to maintain the concentration of such a drug at a level within the therapeutic index, that is, above the minimal concentration needed to elicit the desired effect but below a concentration deemed toxic. Thus P450s play a role in the delicate balancing act of drug dosing.

Many factors can affect the metabolism of a drug in vivo and perturb clearance

(Figure 1-3). In the case of an individual taking multiple drugs simultaneously, more than one of these drugs may happen to be substrates for the same P450(s). In such a case, each substrate molecule would be a “competitive inhibitor” of the other drug’s metabolism. This phenomenon is one manifestation of a drug-drug interaction. P450s are often players in drug-drug interactions due to the fact that they tend to be promiscuous and metabolize

Figure 1-3. Major human P450 isoforms in the families 1-3 participating in drug metabolism. The relative contribution of each isoform to the metabolism of clinically used drugs is indicated in parentheses. Various factors affecting activity are listed as bullets below each isoform. Figure from (Zanger and Schwab, 2013). Original article

(http://dx.doi.org/10.1016/j.pharmthera.2012.12.007). Used under Creative Commons

User License (https://creativecommons.org/licenses/by-nc-nd/4.0/)

8 many substrates.

In addition to direct competition, some drugs alter the transcription of P450 genes and may result in higher protein levels of a given P450. For example, phenobarbital, an anti-epileptic agent, upregulates the expression of multiple P450s at the mRNA level via a process called induction (Waxman and Azaroff, 1992). In one specific case, levels of

CYP2B6 protein in human hepatocytes increased in a dose dependent fashion upon phenobarbital treatment (Faucette et al., 2004). This increase in mRNA expression and subsequent protein level can lead to increased activity in vivo (Figure 1-3) and under dosing. Conversely, potent inhibition of P450s can lead to higher than expected drug levels.

Interestingly, a well-known clinical example of such inhibition comes not from drug administration, but from environmental influences such as an individual’s diet. The grapefruit juice effect, an observed increase in plasma concentrations of some drugs while consuming grapefruit juice, can be partially attributed to inhibition of P450s (Hanley et al.,

2011). Therefore, a well-developed understanding of such complex drug-drug interactions can aid in the determination of safe dosing.

In addition to aforementioned drug-drug interactions and environmental influences, naturally occurring polymorphisms within P450 genes can lead to altered activity (Figure

1-3). An individual with two reference alleles has a genotype of *1/*1 and is assigned a phenotype of extensive metabolizer. Similarly, an individual whose genome contains one or more variant P450 alleles may display altered activity against substrates for those P450s.

This activity can be either increased, decreased, or not affected at all. These individuals are given the phenotypes “ultrarapid metabolizer,” “intermediate metabolizer,” or “poor metabolizer” (Zanger and Schwab, 2013). Many P450 genotypes correlate with race and

9 ethnicity, leading to potential under- or over-dosing of many individuals in certain ethnic groups. Consider the following example: CYP3A5 has been shown to be the P450 primarily responsible for the metabolism of the anti-HIV drug maraviroc (Lu et al., 2012).

Polymorphisms in CYP3A5 are prominently represented in individuals across many ethnicities (Daly, 2006); however, in African and African-American populations, the reference allele (*1) is more common (Xie et al., 2004). Additional work with maraviroc and CYP3A5 genotype found an inverse correlation between the number of CYP3A5*1 alleles and maraviroc plasma concentrations (Lu et al., 2014). Thus, individuals with the

CYP3A5 phenotype of extensive metabolizer may have lower levels of maraviroc. This becomes a problem if these plasma levels are subtherapeutic. Taken together, the genotypes of CYP3A5, which are associated with race and ethnicity, play a large role plasma levels of maraviroc. This is just one example of the impact P450 genotype can have on drug metabolism across populations.

Cytochrome P450 2B6

The human enzyme CYP2B6 is the product of the CYP2B6 gene, which is found on chromosome 19 (Hoffman et al., 2001). Although the CYP2B7 pseudogene is located near CYP2B6, no functional protein is made from CYP2B7 due to location of a premature stop codon before the heme-ligating cysteine, leaving CYP2B6 as the only full-length 2B family member in humans. Other CYP2B family enzymes, however, are abundant in other mammals including mice, rats, rabbits, and dogs (Hoffman et al., 2001). Until the early

2000s, studies of CYP2B family enzymes focused on those expressed in rat, rabbit, and mouse. Eventually, N-terminal truncation (Scott et al., 2001) and point mutations conferring increased stability and solubility (Bumpus et al., 2005; Kumar et al., 2007) of

CYP2B6 enabled the production of enough protein from E. coli to obtain the first crystal structure of CYP2B6 (Gay et al., 2010). This structure revealed that the truncated and mutated CYP2B6 adopts a conformation strikingly similar to other CYP2B enzymes.

As noted in Figure 1-3, cytochrome P450 2B6 participates in the metabolism of about 7% of marketed drugs (Zanger and Schwab, 2013). A portion of the structures on this list is shown in Figure 1-4. The chemical structures of the various substrates of

CYP2B6 are quite different, making it difficult to determine what characteristics of these structures enable them to be substrates for CYP2B6. Moreover, the observed crystal structures of CYP2B6 in complex with various molecules reveal a rather plastic active site

(Gay et al., 2010; Shah et al., 2011; Shah et al., 2012; Wilderman et al., 2013) yet again making it difficult to identify residues that are important for metabolism of a given substrate.

CYP2B6 is one of the most polymorphic human P450s. In total, 38 different

CYP2B6 variant alleles have been identified to date (Sim and Ingelman-Sundberg, 2010).

Some of these variants lead to lower activity due to altered mRNA splicing (Hofmann et al., 2008), a nonsynonymous base change yielding a premature stop codon (Rotger et al.,

2007), or a nonsynonymous base change producing a possible temperature sensitive protein

(Klein et al., 2005). Conversely, at least one CYP2B6 variant (CYP2B6*22) leads to increased activity due to a polymorphism in the promoter region of CYP2B6 (Zukunft et al., 2005). While these CYP2B6 variant alleles are not associated with clinical disease, a

CYP2B*6/*6 genotype is strongly associated with elevated plasma levels of efavirenz

(Meng et al., 2015; Telenti and Zanger, 2008) and methadone (Bunten et al., 2011; Crettol et al., 2005; Dobrinas et al., 2013). Given that the *6 allele has been detected in 10-60% of

Figure 1-4. Structures of fifteen CYP2B6 substrates. The arrows indicate the site of oxygen insertion by CYP2B6. Figure from (Zanger et al., 2007). Used with permission under Order License Id 3987850262750.

12 study participants (Zanger and Klein, 2013), it is possible that a substantial portion of the population at large may harbor one or more of these variant alleles. Furthermore, if these individuals take a drug that is primarily metabolized by CYP2B6, there could be health related risks developing from elevated plasma concentrations of drug. One possible approach to address this risk would be to develop a broader understanding of the structural characteristics that render a drug a substrate for this polymorphic enzyme. Such information could inform drug development and potentially steer pharmaceutical companies away from producing substrates for CYP2B6, saving time and money over the course of preclinical development.

Efavirenz

As a non-nucleoside reverse transcriptase inhibitor used to treat HIV, efavirenz

(EFV), or Sustiva ® (Figure 1-4), is a prominent component of highly active antiretroviral therapy (Kenedi and Goforth, 2011). EFV functions to non-competitively inhibit the HIV reverse transcriptase enzyme, rendering the virus incapable of converting the viral RNA into DNA (Ren et al., 2000).

Studies of efavirenz metabolism in vivo have demonstrated that 8-hydroxyefavirenz

(8OH-EFV) is the major metabolite found in humans (Mutlib et al., 1999b). Furthermore,

8OH-EFV is primarily formed by CYP2B6 (Bumpus et al., 2006; Ward et al., 2003).

Indeed, this reaction is specific enough for CYP2B6 that 8 hydroxylation of efavirenz has been proposed as an assay to test CYP2B6 activity in vivo (Huang et al., 2007). Due to the specificity of this reaction and the aforementioned polymorphisms in CYP2B6, it is not surprising that CYP2B6 phenotype correlates with efavirenz plasma concentrations.

Individuals with poor metabolizer CYP2B6 phenotypes have been demonstrated to have

13 efavirenz AUC values of up to three times higher than wild type individuals (Telenti and

Zanger, 2008; Tsuchiya et al., 2004), a fact that is further complicated by the association of efavirenz and 8OH-EFV with cell death (Bumpus, 2011) and central nervous system damage (Tovar-y-Romo et al., 2012).

Through studies of naturally occurring CYP2B6 variants, various groups have determined the impact of CYP2B6 polymorphisms on efavirenz metabolism in vitro

(Ariyoshi et al., 2011; Bumpus et al., 2006; Xu et al., 2012; Zhang et al., 2011).

Interestingly, some of these variants display higher catalytic efficiency towards EFV than

*1, namely *4 (Ariyoshi et al., 2011; Bumpus et al., 2006), *5, and *7 (Zhang et al., 2011); however, the *6 variant - the most commonly observed 2B6 polymorphism - displays lower catalytic efficiency against EFV (Xu et al., 2012; Zhang et al., 2011). The examined variants, though they do impact EFV metabolism by CYP2B6, are distal to the predicted enzyme active site.

Though it has long been appreciated that CYP2B6 displays high activity towards efavirenz, a deeper understanding of the molecular determinants of this relationship has not been obtained. In this thesis, I have laid the groundwork for determining the structural characteristics of efavirenz that render it a substrate for CYP2B6. The contributions offered herein are 1) the development of analytical assays for the detection of 14 efavirenz analogues, 2) the characterization of the metabolism of 14 EFV analogues, 3) the determination of human cytochrome P450 activity towards these analogues, 4) identification of the oxazinone ring as a structural characteristic of EFV important for

CYP2B6 activity, 5) kinetic characterization of analogues found to be CYP2B6 substrates, and 6) comparison of CYP2B6 and CYP2B1 activity towards EFV analogues.

Chapter 2.

Structure-activity studies reveal the oxazinone ring is a determinant of cytochrome

P450 2B6 activity towards efavirenz.

The content of this chapter has been published (Cox and Bumpus, 2014) and is

reproduced here under the ACS Author’s Choice Usage Agreement.

The cytochromes P450 (P450s) are a large family of monooxygenase enzymes that play a major role in the metabolism of xenobiotic and endogenous substrates. One such enzyme, human cytochrome P450 2B6 (CYP2B6), participates in the metabolism of approximately 4% of the top 200 prescribed drugs in the United States (Zanger et al., 2008), including the atypical antidepressant and smoking cessation aid bupropion (Hesse et al.,

2000) and the non-nucleoside reverse transcriptase inhibitor efavirenz, or EFV (Bumpus et al., 2006; Ward et al., 2003). CYP2B6 has been detected in many tissues, including brain, intestine, and lung (Gervot et al., 1999), but is primarily expressed in the liver where it is present in concentrations ranging from 0.5 to 100 pmol/mg total microsomal protein

(Gervot et al., 1999; Hofmann et al., 2008). This marked variability in expression is due in part to the highly polymorphic nature of the CYP2B6 gene. To date, 38 distinct CYP2B6 variants have been described (Sim and Ingelman-Sundberg, 2010), some of which have been associated with altered CYP2B6 expression at the mRNA (Zukunft et al., 2005) and protein (Hofmann et al., 2008) level as well as altered activity (Desta et al., 2007; Radloff et al., 2013; Zanger et al., 2008).

EFV is a commonly prescribed non-nucleoside reverse transcriptase inhibitor used to treat HIV-1. CYP2B6 is primarily responsible for the formation of the major EFV metabolite, 8-hydroxyefavirenz (Ogburn et al., 2010; Ward et al., 2003). EFV has been shown to induce CYP2B6 transcription (Faucette et al., 2007) as well as inhibit CYP2B6 activity (Bumpus et al., 2006; Hesse et al., 2001). Multiple crystal structures of the naturally occurring CYP2B6 variant K262R in complex with various inhibitors have been solved (Gay et al., 2010; Shah et al., 2011; Wilderman et al., 2013); however, a structure of CYP2B6 in complex with EFV has yet to be reported.

In this study, we set out to determine which regions of the EFV chemical structure contribute to CYP2B6 activity towards EFV. To this end, we analyzed the metabolism of a panel of EFV analogues (Figure 2-1). Our goals were: 1) to determine if any of the analogues are CYP2B6 substrates and 2) to quantitatively compare the kinetics of EFV and

EFV analogue metabolism by CYP2B6. In doing so, we observed that CYP2B6 readily metabolized analogues with an intact oxazinone ring (analogues 3, 4, and 8), whereas we did not observe metabolite formation from CYP2B6 incubation with analogues with a disrupted oxazinone ring (analogues 1, 2, 5, 6, and 7). Interestingly, CYP2B6 activity readily produced metabolites of an oxazine EFV analogue (analogue 8), further implicating the integrity of the oxazinone moiety in CYP2B6 metabolism of EFV and not simply the carbonyl oxygen atom.

Materials and Methods

EFV analogues

EFV, all EFV analogues, and bupropion hydrochloride were obtained from Toronto

Research Chemicals (Toronto, Canada) and were ≥ 97% pure, according to the manufacturer. The analogues used in this study were 1 – (αS)-2-Amino-5-chloro-α-(2- cyclopropylethynyl)-α-trifluoromethyl)benzenemethanol, 2 – (αR)-2-Amino-5-chloro-α-

(2-cyclopropylethynyl)-α-(trifluoromethyl)benzenemethanol, 3 – (E)-Dihydroefavirenz, 4

– rac 6-Chloro-1,4-dihydro-4-(1-pentynyl)-4-(trifluoromethyl)-2H-3,1-benzoxazin-2-one,

5 – (R)-5-Chloro-α-(cyclopropylethynyl)-2-[[(4-methoxyphenyl)methyl]amino]-α-

(trifluoromethyl)benzenemethanol, 6 – (S)-5-Chloro-α-(cyclopropylethynyl)-2-[[(4- methoxyphenyl)methyl]amino]-α-(trifluoromethyl)benzenemethanol, 7 – rac N-[4-

Chloro-2-[3-cyclopropyl-1-hydroxy-1-(trifluoromethyl)-2-propynyl]phenyl]-4-

Figure 2-1. Structures of efavirenz (EFV) and EFV analogues 1-8.

18 methoxybenzamide, and 8 - (4S)-6-Chloro-4-(2-cyclopropylethynyl)-1,4-dihydro-4-

(trifluoromethyl)-2H-3,1-benzoxazine.

EFV analogue metabolism by human liver microsomes and cDNA-expressed P450s.

Metabolism assays with human liver microsomes (Xenotech LLC, Lenexa, KS) or cDNA-expressed P450s (Supersomes, BD Biosciences) were performed as described previously (Lade et al., 2013; Lu et al., 2012) using 10 µM EFV analogue and 2 mg/mL

HLM or 10 nM cDNA-expressed P450 for 60 minutes. To aid in fragment assignment, metabolism reactions with human liver microsomes were also performed in the presence

18 of H2 O (Cambridge Isotope Laboratories, Andover, MA). Reactions were performed in polypropylene tubes except those for analogue 8, which were performed in silanized borosilicate glass tubes. All samples were analyzed using uHPLC-MS/MS.

Substrate depletion kinetics.

Km and Vmax values for EFV and EFV analogues 3, 4 and 8 were obtained using a substrate depletion approach essentially as described previously (Obach and Reed-Hagen,

2002). Briefly, 10 nM CYP2B6 was pre-incubated with six concentrations of EFV or EFV analogues ranging over four to six orders of magnitude for 5 minutes at 37 ˚C in 100 mM potassium phosphate buffer. At 0, 2, 5, 10, 20, and 30 minutes after addition of an NADPH- regenerating system (BD Biosciences), 100 µL aliquots were taken and diluted into an equal volume of acetonitrile containing the internal standard fluorinated efavirenz.(Avery et al., 2010) The height of the analyte peak and internal standard peak were determined by uHPLC-MS/MS.

Inhibition using bupropion as a probe.

Each analogue (10 µM) was incubated with bupropion (40 µM) and CYP2B6 (50

µM) in 100 mM potassium phosphate buffer for 10 minutes in the presence of an NADPH regenerating system at 37 ˚C. Hydroxybupropion formation was linear under these conditions. Reactions were terminated as described previously.(Lade et al., 2013; Lu et al.,

2012) Data were collected using uHPLC-MS/MS.

Data Analysis.

Non-linear regression, curve fitting, and statistical analyses were performed using

GraphPad Prism (version 6; GraphPad Software Inc., San Diego, CA). Statistical comparisons were made using unpaired parametric t-tests. Substrate depletion data analysis was performed essentially as previously described (Obach and Reed-Hagen, 2002). Briefly, the natural logarithm of the ratio of analyte to internal standard for each substrate concentration was calculated and normalized to the ratio at time 0. The normalized ratio was then plotted versus time and the resulting curves fit to equation (1) to obtain the depletion rate constant (kdep).

-kdept Y=Y0×e (1)

Only the linear portion of the depletion curve was used to calculate kdep. Any points not contributing to log-linearity were excluded. To calculate Km, the kdep values were plotted versus the logarithm of the substrate concentration and fit to equation (2), where kdep max is the maximum depletion rate constant.

[S] kdep = kdep max × (1 - ) (2) [S] + Km

Vmax values were determined as described previously (Lutz et al., 2009). Briefly, the kdep max values obtained from fitting the data to equation 2, were divided by the enzyme concentration yielding the Clint. The Clint values were then multiplied by the calculated Km to obtain the Vmax. 20 uHPLC-MS/MS analysis of EFV, EFV analogues, and hydroxybupropion

For metabolism assays, samples were injected onto a Dionex UltiMate 3000 uHPLC system coupled to a TSQ Vantage Triple Stage Quadrupole mass spectrometer

(Thermo Scientific, Pittsburgh, PA) and resolved with an Xterra C18 column (2.5 µm, 2.1 by 50 mm; Waters, Milford, MA). For EFV and EFV analogue assays, a linear gradient of mobile phase A (water + 0.1% formic acid) and mobile phase B (acetonitrile + 0.1% formic acid) consisting of 5% B from 0-0.5 min, 5-95% B from 0.5-9.5 min, 95% B from 9.5-11.5 min, and re-equilibration at 5% B from 11.6-12.6 minutes at a flow rate of 0.4mL/min was used. For separation and detection of hydroxybupropion, a linear gradient of the same mobile phases consisting of 20% B from 0-0.5 min, 20-50% B from 0.5-3 min, 50-95% B from 3.0-3.1 min, 95% B from 3.1-4.0 min, and re-equilibration at 20% B from 4.1-5.0 min was used. All reagents used for mobile phases were of the highest grade commercially available. For hydroxybupropion and EFV analogues 1, 2, and 5-8 resolution was achieved in positive ion mode, while EFV, EFV analogues 3 and 4, and fluorinated efavirenz were resolved in negative ion mode. EFV analogue metabolite identification and determination of relative abundance was accomplished in product ion mode. Hydroxybupropion relative abundance was determined using selected reaction monitoring with a transition of m/z

256.1 > 139.06. For substrate depletion experiments, selected reaction monitoring was used to monitor the parent peak height over time. All samples were injected onto a Polaris C18-

A column (5 µM, 100 by 2 mm) at a flow rate of 0.6 mL/min. A linear gradient of mobile phase A (water + 0.1% formic acid) and mobile phase B (acetonitrile + 0.1% formic acid) consisting of 10% B from 0-0.5 min, 10-95% B from 0.5-5 min, 95% B from 5-6 min, and re-equilibration at 10% B from 6.1-7 min. For analogue 8, a baseline of 20% B was used

21 in order that the parent peak elute during the gradient phase. Selected reaction monitoring transitions used were as follows: EFV: m/z 314 > 243.87; fluorinated efavirenz: m/z 297.97

> 227.93; analogue 3 m/z 315.9 > 245.9; analogue 4: m/z 316 > 218; and analogue 8: m/z

302.1 > 203.97.

Results and Discussion

EFV analogues undergo P450-dependent metabolism in pooled human liver microsomes.

In order begin our investigation into the ability of CYP2B6 to metabolize the EFV analogues, we first sought to identify all of the metabolites of these compounds that could be produced by P450 enzymes. To do so, each analogue was incubated with human liver microsomes, which contain a range of P450s, in the presence of NADPH and product formation was analyzed using ultra high-performance liquid chromatography tandem mass spectrometry (uHPLC-MS/MS). Monooxygenated metabolites of analogues 1, 2, 3, 4, and

8 were detected while N-dealkylated products were detected from analogues 5 and 6

(Figure 2-2). No metabolites of analogue 7 were observed. From EFV analogues 1 and 2, which have a disrupted oxazinone ring, a monooxygenated metabolite (1-M1 and 2-M1, respectively) was detected with a retention time (RT) of 5.8 min (m/z 306, Figure 2-2).

Product ion spectra for all metabolites are shown in Figure 2-3. The major fragment ions of these metabolites were m/z 268, 247, and 194, indicating a loss of H3Cl, C3H6O, and

C3H3OF3. This fragmentation pattern suggests 1-M1 and 2-M1 were formed by oxygen insertion at the 5, 7, or 8 position of the benzene ring. EFV analogue 3, which has a trans alkene rather than the alkyne of EFV, was metabolized to two different monooxygenated metabolites (m/z 332, 3-M1 and 3-M2) by human liver microsomes (Figure 2-2). These

Figure 2-2. Oxidative metabolites of EFV analogues 1-6 and 8 formed by human liver microsomes. Human liver microsomes at a concentration of 2 mg/mL were incubated with 10 µM EFV analogue in the presence of an NADPH regenerating system for 60 min. Metabolites were detected by scanning for the parent m/z plus 16 or 32 using uHPLC-MS/MS in product ion mode. Representative chromatograms of three independent experiments are shown for each analogue.

Figure 2-3. Mass spectra for EFV analogue metabolites identified using human liver microsomes. Fragmentation was conducted in product ion mode scanning for the parent mass plus 16 or 32 m/z for mono- and dihydroxylated metabolites, and for m/z 290 for N- dealkylated metabolites. Data here represent a summary of three replicate experiments. 24 species corresponded to RTs of 4.98 and 6.43 min, respectively. Metabolite 3-M2, which was the most abundant metabolite observed from EFV analogue 3, displayed fragment ions of m/z 288, 268, 247 and 227, which we propose derived from a loss of CO2, C2H4Cl,

C4H5O2, and CF3Cl. These fragment ions suggested the oxygen insertion occurred on the aromatic ring at the 7 or 8 position. The >1 minute difference in RT between the two monooxygenated metabolites of EFV analogue 3 suggested that the position of oxygenation for 3-M1 was quite distal compared to 3-M2. Consistent with this notion, the mass spectrum for 3-M1 revealed fragment ions of m/z 254, 171, and 156, which derived from the loss of CO2Cl, C3H2O2F3Cl, and C4H4O2F3Cl. We propose this corresponded to an oxygen insertion at the 14 position.

In EFV analogue 4, the cyclopropyl ring that is found in EFV is open. Human liver microsome incubations with this analogue resulted in the formation of 5 monooxygenated metabolites (m/z 332, Figure 2-2). Metabolite 4-M1 corresponded to a RT of 5.11 min and had characteristic fragment ions of m/z 172, 156, 143, and 128 resulting from the loss of

C3O2F3Cl, C3O3F3Cl, C4HO3F3Cl and C5HO3F3Cl. This fragmentation indicated possible oxygen insertion at the 15 position. Metabolite 4-M2 resulted from oxygen insertion at the

14 position and was characterized by a RT of 5.60 min and fragment ions of m/z 262, 238, and 192, which were the result of a loss of CHF3, C2H3O2Cl, and C4H3O2F3. We detected a third monooxygenated metabolite, 4-M3, at a RT of 5.68 min that represented oxygen insertion on the benzene ring at either the 5 or the 7 position. 4-M3 exhibited fragment ions of m/z 268, 244, 218, and 204, which we propose resulted from a loss of C2H5Cl, C4H8O2,

C2HO2F3, and C3H4O2F3. 4-M4 was observed at a RT of 6.08 min and had fragment ions of m/z 274, 259, 246, and 230. These ions corresponded to a loss of C2H2O2, C3H6O2,

C4H6O2, and CF3Cl, suggesting that the oxygen insertion occurred at either the 5 or the 7 position as well. Metabolite 4-M5 was the most abundant metabolite of this analogue and eluted at a RT of 6.98 min. We observed fragment ions of m/z 288, 260, 252, and 228, that corresponded to a loss of CO2, C3H4O2, CO2Cl, and CF3Cl. Based on the fragmentation pattern, which closely matches that of 8-hydroxyefavirenz, (Mutlib et al., 1999b) and the hydrophobicity of this analog, we propose this metabolite resulted from oxygen insertion at the 8 position. The peak eluting immediately after 4-M5 was not denoted as a metabolite because it appeared to be produced within the mass spectrometer since an identical peak was detected when synthetic 8-hydroxyefavirenz alone is injected directly into the mass spectrometer (data not shown).

EFV analogues 5 and 6 differ in stereochemistry about the 4 position, but both lack an intact oxazinone ring and have an additional methoxyphenyl group compared to EFV.

No monooxygenated metabolites were observed from these analogues; however, metabolites 5-M1, 6-M1, and 6-M2 (m/z 290), that we propose to be N-dealkylated products were observed. 5-M1 and 6-M2 eluted at a RT of 7.31 and 7.25 min, respectively, while 6-M1 was detected at a RT of 6.51 min (Figure 2-2). All three metabolites exhibited fragment ions of m/z 244, 232, and 178. This fragmentation pattern, representing a loss of

C2H6O, C3H6O, and C4H7F3, suggested that these metabolites were the result of an N- dealkylation reaction. Unlike 6-M1, the RT of 5-M1 and 6-M2 were later RT than analogues 1 and 2 (6.57 min, data not shown), which are structurally identical to the proposed N-dealkylated metabolites. The data suggest that 5-M1 and 6-M2 are less polar than analogues 1 and 2, yet have the same mass spectrum. Analogue 7, which possesses an additional carbonyl compared to analogues 5 and 6, did not exhibit any metabolites from

26 incubation with pooled human liver microsomes.

EFV analogue 8 only lacks the carbonyl oxygen from the oxazinone ring, leaving an intact oxazine ring in its place. Human liver microsome activity produced three monooxygenated metabolites (m/z 318) and one dihydroxylated metabolite (m/z 334) from this analogue (Figure 2-2). Metabolite 8-M1 had a retention time of 5.00 min and, though it did not fragment readily, produced characteristic ions of m/z 226, and 91. Human liver microsome metabolism assays performed using isotopic water revealed 8-M1 fragments of m/z 228 and 91 (Figure 2-4A), indicating that the inserted oxygen was retained in the m/z

226 fragment but not in the m/z 91 fragment. We therefore propose a loss of C3H6OCl, and

C8H7O2F3Cl for these fragments, suggesting the oxygen was inserted just before the cyclopropyl group at the 14 position. Metabolite 8-M2 was found to have a retention time of 6.43 min with fragment ions of m/z 282, 272, and 237, indicating the loss of H2Cl,

C2H6O, and CH2O2Cl. This fragmentation pattern suggested the seven or eight position of the benzene ring as the site of oxygen insertion. Metabolite 8-M3 was the most abundant monooxygenated metabolite detected from metabolism of EFV analogue 8 by human liver microsomes. A retention time of 6.90 min observed as well as fragment ions of m/z 288,

240, 220, 186, and 141. We propose these fragments corresponded to a loss of CH2O,

C4HO, C2H3OCl, C2HOF3, and C3HO2F3. On the basis of this fragmentation pattern as well as hydrophobicity, we propose that metabolite resulted from oxygen insertion on the benzene ring, most likely at the seven or eight position. The fragmentation of dihydroxylated metabolite 8-M4, which corresponds to a retention time of 4.90 min, is nearly identical to that of 8-M1, suggesting that 8-M4 may be formed from 8-M1. Fragment

18 ions of m/z 242, and 91 were observed. Metabolism assays with H2O revealed 8-M4

18 Figure 2-4. Mass spectra from analogue 8 incubations with H2O and human liver microsomes. Fragmentation was conducted in product ion mode scanning for the parent mass plus 16 or 18 m/z for monooxygentated metabolites (A.) or the parent mass plus 32

16 18 or 36 m/z for dihydroxylated metabolites (B.) in H2O or H2O , respectively. Data here are representative results from three independent experiments.

28 fragments ions of m/z 246 and 91, suggesting that both oxygens are retained upon formation

16 of m/z 242 species observed with H2O (Figure 2-4B). We propose these fragments resulted from a loss of C3H6OCl, and C8H7O2F3Cl. The first oxygen is most likely inserted at the 14 position, as in 8-M1, whereas the second oxygen is more distal, and the insertion may occur at the 5, 7, or 8 position of the benzene ring.

The oxazinone ring plays a critical role in CYP2B6-dependent metabolism of EFV.

Following the identification of the metabolites formed by human liver microsomes, we used cDNA-expressed enzymes to determine which of these metabolites could be formed by CYP2B6. The results from these experiments are summarized in Table 2-1 and

Figure 2-5. Interestingly, incubation of analogues 1, 2, 5, and 6 with CYP2B6 did not result in the formation of detectable metabolites. Each of these analogues possesses a disrupted oxazinone ring. In contrast, from analogues with an intact ring (3, 4, and 8) we did observe metabolites 3-M2, 4-M4, 4-M5, and 8-M2 from incubations with CYP2B6. Since CYP2B6 did not produce all the metabolites we detected using human liver microsomes, we incubated each analogue with a panel of individual P450 enzymes representing the major human drug metabolizing P450s. 1-M1 and 2-M1 were detected from incubations with

CYP1A2 and CYP2C19. Production of the N-dealkylated metabolites of analogues 5 and

6 were catalyzed by CYP1A2, CYP2C19, and CYP2D6. In addition to CYP2B6, we observed metabolism of analogues 3 and 4 by CYP1A2 and CYP3A5. Analogue 8, however, was only metabolized by CYP2B6 and CYP2C19. Though divergent substrate specificities have been previously observed for CYP2C9 and CYP2C19 (Lasker et al.,

1998; Mancy et al., 1999), few substrates have been identified that can distinguish the activity of CYP3A4 from CYP3A5, which share 84% amino acid sequence identity. Our

Table 2-1. EFV analogue metabolites formed from cDNA-expressed P450s

P450

1 2 3 4 5 6 8 1A1 1A2 M1 M1 M2 M4, M5 M1 M1, M2 2A6 2B6 M2 M4, M5 M2 2C8 2C9 2C19 M1 M1 M1 M2 M2, M3 2D6 M1 M1, M2 3A4 3A5 M2 M5 3A7 Blank cells represent no detectable metabolite formation. No metabolites were detected from analogue 7.

Figure 2-5. Cytochrome P450-catalyzed metabolism schemes for EFV analogues

1-6 and 8. Metabolite detection was performed using uHPLC-MS/MS as described under materials and methods. In each scheme, individual P450s found to form each metabolite are indicated or “?” is used to demarcate those metabolites that were not found to be formed by any of the individuals P450 enzymes tested. We did not detect metabolites from EFV analogue 7. Data here represent a summary of three replicate experiments.

31 group has demonstrated preferential metabolism of maraviroc by CYP3A5 (Lu et al.,

2012), while others have shown the metabolism of T-5 to be catalyzed primarily by

CYP3A5 (Li et al., 2014). This panel of EFV analogues may also prove to be useful in distinguishing the activities of these highly homologous P450s.

A previous study used molecular docking to simulate the presence of EFV in the active site of CYP2B6 (Niu et al., 2011). This study revealed interactions between EFV and residues E301 and T302, which may participate in a hydrogen bonding network involving the EFV oxazinone ring. Indeed, T302 has also been postulated to form polar contacts with amlodipine based on another X-ray crystal structure (Shah et al., 2012). Since we noted that the absence of an intact oxazinone ring abrogated metabolism by CYP2B6, it is possible that contacts formed by members of this ring are important for catalysis or stabilization of EFV in the active site. Moreover, since analogue 8 was readily metabolized by CYP2B6, the carbonyl moiety of EFV does not appear to be essential for CYP2B6 activity towards EFV. These data suggest that perhaps the increased conformational flexibility afforded to analogues with an open oxazinone ring abrogates the ability of these analogues to be metabolized by CYP2B6.

Kinetic comparison of CYP2B6 substrates.

Next, we set out to quantitatively compare the metabolism of EFV analogues shown to be CYP2B6 substrates. Since multiple metabolites were detected from CYP2B6 activity with analogue 4, and since CYP2B6 is known to form multiple metabolites from EFV itself

(Ward et al., 2003), substrate depletion was chosen as the method to determine relative kinetic constants (Figure 2-6) (Obach and Reed-Hagen, 2002). KM and Vmax values were determined for EFV and analogues 3, 4, and 8 and are reported in Table 2-2. The observed

Figure 2-6. Substrate depletion analysis of EFV and EFV analogues 3, 4, and 8.

CYP2B6 (10 nM) was incubated with EFV (A), EFV analogue 3 (B), 4 (C), or 8 (D) at

37˚C in 100 mM potassium phosphate buffer. At 0, 2, 5, 10, 20, and 30 min after the addition of an NADPH regenerating system 100 µL of the reaction mixture was diluted into an equal volume of acetonitrile containing the internal standard fluorinated efavirenz.

Analyte and internal standard abundances were measured using uHPLC-MS/MS. Data represent the mean ± SD of three replicate experiments performed in duplicate.

Table 2-2. Kinetic constants estimated from substrate depletion experiments

k V Compound K (µM) dep max max M (min-1) (pmol/min/pmol P450)

EFV 0.45 ± 0.11 0.023 ± 0.001 1.0 ± 0.3

3 0.23 ± 0.05 0.0057 ± 0.0002 0.13 ± 0.03

4 1.10 ± 0.26 0.0088 ± 0.0003 0.97 ± 0.27

8 0.22 ± 0.05 0.14 ± 0.01 3.2 ± 1.0

KM for EFV was 0.45 ± 0.11 µM, while the calculated Vmax was 1.0 ± 0.31 pmol/min/pmol

P450 (Table 2-2). These values are lower than previously reported in studies measuring only 8-hydroxyefavirenz formation (Bumpus et al., 2006; Ward et al., 2003). Since 8- hydroxyefavirenz can be further metabolized by CYP2B6 (Ward et al., 2003), we expect that monitoring EFV disappearance, as we have done in this study, may yield kinetic constants that differ from those obtained in previous reports. The observed two-fold decrease in KM between EFV and analogue 3 suggests this analogues may have higher affinity for CYP2B6, although the order of magnitude difference in Vmax may indicate catalytic differences in the metabolism of this substrate and EFV. Conversely, the two-fold increase in KM observed with analogue 4 suggests this structure with an open cyclopropyl ring is bound less tightly than EFV. Analogue 3 also exhibited a Vmax that was an order of magnitude lower than that observed for EFV, suggesting that although the KM is lower for this substrate, catalytic difference may exist between the metabolism EFV and this analogue with a trans alkene. The observed improvements in maximal reaction rate as well as substrate affinity for analogue 8 suggest that the absence of the carbonyl oxygen atom does not hinder metabolite formation by CYP2B6 and may in fact enhance it. One possible explanation for this could be that the absence of the carbonyl oxygen atom might relieve steric clashes between EFV and CYP2B6 side chains in the active site.

Inhibition of CYP2B6 activity by EFV analogues.

In order to determine if the analogues not found to be CYP2B6 substrates could still interact with the enzyme, we analyzed the metabolism of a known CYP2B6 substrate, bupropion (Hesse et al., 2000), in the presence of each EFV analogue. A concentration of

10 µM EFV or EFV analogue was chosen since this same concentration was used in our

35 metabolism assays. Although spectral binding is commonly employed to investigate substrate interactions with purified P450s, our attempts to detect a spectral shift in the

CYP2B6-containing insect microsomes utilized in our studies in response to EFV and EFV analogues were unsuccessful. While the use of this system allowed us to readily examine

CYP2B6-dependent activity towards the range of compounds used in the present study, it is possible that the complexity of the insect microsomal system rendered a spectral shift difficult to observe. Co-incubations of EFV with bupropion and CYP2B6 resulted in a 93% decrease in hydroxybupropion formation compared to vehicle control (Figure 2-7). Of EFV analogues found to be CYP2B6 substrates, analogue 4 exhibited the weakest inhibition of bupropion metabolism (27% activity remaining) while analogue 8 showed the greatest inhibition (2% activity remaining). These findings are in congruence with our substrate depletion data since analogue 4 exhibited a higher KM than EFV while the KM for analogue

8 was lower than EFV. The sizes of analogues 5-7 are greater than the other analogues used in this study, which could provide a clue as to why these analogues were not metabolized by CYP2B6. Analogue 7, which is the largest analog, was unable to inhibit hydroxybupropion formation, while the slightly smaller analogues 5 and 6 were able to moderately inhibit hydroxybupropion formation (52% and 62% activity remaining, respectively). This suggests that size may influence the interaction of CYP2B6 with compounds possessing the EFV chemical scaffold. Interestingly, although analogues 1 and

2 were not metabolized by CYP2B6, inhibition of CYP2B6 activity toward EFV by these compounds was commensurate with that of EFV as well as analogue 3, which was also a

CYP2B6 substrate. These data suggest that analogues 1 and 2 are able to interact with

CYP2B6 although this interaction does not result in metabolism of these molecules. This

Figure 2-7. Inhibition of CYP2B6 bupropion hydroxylase activity by EFV and EFV analogues. CYP2B6 (50 nM) was incubated with bupropion (40 µM) and EFV or EFV analogue (10 µM) in 100 mM potassium phosphate buffer for 10 min at 37 ˚C.

Hydroxybupropion formation was analyzed using uHPLC--MS/MS. Inhibition is reported as a percentage of hydroxybupropion formation in the presence of a vehicle control (no

EFV or EFV analogue present). Data reflect the mean ± SD of three replicate experiments performed in duplicate. Symbols for statistical significance represent comparisons to vehicle control (*) or EFV (#). Two symbols, p ≤ 0.01; three symbols, p ≤ 0.001.

37 finding lends further evidence to the notion that an intact oxazinone ring is critical to the ability of CYP2B6 to metabolize EFV.

In summary, we have demonstrated the importance of the oxazinone ring for metabolism of EFV by CYP2B6. Future studies could test whether compounds that are structurally similar to EFV yet possess other kinds of six-membered rings are also CYP2B6 substrates, since our results do not preclude this possibility. No metabolite production was observed from incubations of CYP2B6 with analogues lacking an intact oxazinone ring

(analogues 1, 2, 5, 6, and 7), though other P450s were able to catalyze these reactions.

Substrate depletion analysis of EFV analogues that were CYP2B6 substrates revealed additional differences between these substrates, namely that the cyclopropyl ring and alkyne moieties of EFV may play a role in binding affinity and reaction rate, respectively.

Interestingly, we determined that an analogue lacking the carbonyl oxygen atom possesses higher affinity for CYP2B6 as well as an increased maximum reaction rate. Thus, the intact oxazinone ring and not simply the carbonyl functional group, is a determinant of catalytic activity of CYP2B6 towards EFV.

Chapter 3

Single heteroatom substitutions in the efavirenz oxazinone ring impact metabolism

by CYP2B6

The contents of this chapter have been published (Cox and Bumpus, 2016) and are

reproduced here with permission under License Number 3991981344193.

The cytochromes P450 (P450s) are a superfamily of heme-containing monooxygenases that participate in the metabolism of xenobiotic compounds, such as drugs and environmental toxins. P450 substrates are generally nonpolar molecules that are rendered more polar by insertion of an oxygen atom, resulting in a product that can be more readily excreted from the body. Cytochrome P450 2B6 (CYP2B6) represents the only bona fide member of family 2B in humans. Other well-studied 2B subfamily members, each sharing at least 75% amino acid sequence identity with CYP2B6, include rat CYP2B1, rabbit CYP2B4, and mouse Cyp2b10. CYP2B6 plays a prominent role in the metabolism of many clinically relevant substrates including the chemotherapeutic pro-drug cyclophosphamide (Chang et al., 1993), the antidepressant bupropion (Hesse et al., 2000), the opioid maintenance drug methadone (Kharasch et al., 2004), and the HIV non- nucleoside reverse transcriptase inhibitor efavirenz (EFV) (Bumpus et al., 2006; Ward et al., 2003). Over 100 different single nucleotide polymorphisms have been identified in

CYP2B6, representing 38 distinct variant alleles (Sim and Ingelman-Sundberg, 2010).

These variants are associated with a broad spectrum of phenotypes from low to high activity towards CYP2B6 substrates (Wang and Tompkins, 2008; Zanger and Klein, 2013).

Some variants of CYP2B6 have been associated with decreased response to cyclophosphamide-based therapies (Johnson et al., 2013; Zanger and Klein, 2013) elevated plasma levels of methadone (Kharasch et al., 2015), and of EFV (Rakhmanina and van den

Anker, 2010; Rotger et al., 2007; Telenti and Zanger, 2008).

EFV is a commonly prescribed drug used to treat HIV type 1. As a non-nucleoside reverse transcriptase inhibitor, EFV inhibits the activity of HIV reverse transcriptase via binding at an allosteric site (Ren et al., 2000). CYP2B6 is the primary enzyme responsible

40 for the formation of 8-hydroxyefavirenz (8-OH EFV), the most abundant product of phase

1 EFV metabolism in humans (Mutlib et al., 1999a; Ogburn et al., 2010; Ward et al., 2003).

EFV has been shown to modulate the expression of CYP2B6 via nuclear receptor activation

(Faucette et al., 2007) as well as to inactivate CYP2B6 protein via mechanism-based inactivation (Bumpus et al., 2006). Thus, in addition to being a CYP2B6 substrate, EFV impacts both CYP2B6 expression and activity. In the absence of crystal structures of

CYP2B6 in complex with EFV, little is known about the regions of the EFV chemical structure that make important contacts with the CYP2B6 active site and may contribute to the high activity of CYP2B6 against EFV. Previously, we demonstrated an intact oxazinone ring to be important for metabolism of EFV by CYP2B6 (Cox and Bumpus,

2014). Additionally, in those studies, we found that an analogue lacking only the carbonyl oxygen of EFV was a better CYP2B6 substrate than EFV itself. Here we continue this work by examining the impact of each heteroatom position of the EFV oxazinone ring on metabolism by CYP2B6. In doing so, we found that CYP2B6 is tolerant of single atom changes in the oxazinone ring; however, these changes impact the KM associated with the metabolism of these substrates. Having observed the consequences of these structural changes for CYP2B6, we tested whether another CYP2B enzyme, rat CYP2B1, displayed an activity profile similar to that of CYP2B6 against EFV analogues. We found a striking similarity between the activity of these CYP2B enzymes towards EFV analogues that was not observed with other P450s of the 1A, 2A, 2C, 2D, or 3A families, suggesting the ability to metabolize certain EFV analogues may be conserved in CYP2B family members.

Results

Our interest in the oxazinone ring of EFV led us to acquire a targeted panel of EFV

41 analogues with discreet changes in the oxazinone ring (Figure 3-1). We first performed incubations with pooled human liver microsomes in order to identify P450-dependent metabolites of these analogues. Using ultra high-performance liquid chromatography tandem mass spectrometry (uHPLC-MS/MS), we detected monooxygenated metabolites formed from each analogue. Two monooxygenated metabolites (m/z 332) were formed from 1, which has a methyl group in place of the carbonyl oxygen of efavirenz (Figure 3-

1). Metabolite 1-M1 displayed a retention time (RT) of 5.85 min while 1-M2 displayed a

RT of 7.10 min (Figure 3-2A.i and 3-2A.ii). Fragmentation of 1-M1 (theoretical m/z:

332.0659, experimental m/z: m/z 332.0687) using a high resolution accurate mass instrument revealed fragment ions of m/z 290.0541, 272.0438, 232.0124, and 178.0412 resulting from a loss of C2H2O, C2H4O2, C5H8O2, and C5H5F3O2, respectively. Metabolite

1-M2 (theoretical m/z: 332.0659, experimental m/z: 332.0687) possessed fragment ions of m/z 288.0394, 268.0332, 248.0074, and 184.0755. These fragments suggested a loss of

C2H4O, C5H4, C5H8O, and C3H4ClF3O, respectively. Based on the fragmentation, we propose that 1-M1 resulted from a single oxygen insertion at the methyl group of the oxazine ring (Figure 3-3A). We further propose that 1-M2 resulted from oxygen insertion on the benzene ring at the 5, 7, or 8 position (Figure 3-3B).

Analogue 2, which has a nitrogen atom rather than the oxygen atom within the oxazinone ring of efavirenz, also displayed two monooxygenated metabolites (m/z 331,

Figure 3-2A.iii) from incubations with human liver microsomes. Metabolite 2-M1

(theoretical m/z: 331.0456, observed m/z: 331.0454), with a RT of 5.42 min, showed fragments of m/z 288.0392, 264.9983, 261.0418, and 184.0755. We propose these fragments resulted from a loss of CHNO, C5H6, CHF3, and C2HClF3O, respectively. With

Figure 3-1. Efavirenz (EFV) and analogues 1-5 used in this study. Structural departures from EFV are denoted in red.

Figure 3-2. Extracted ion chromatograms of metabolites formed from incubations with human liver microsomes and analogues 1-4 and UV chromatogram from incubations with analogue 5. Metabolites were named in order of elution beginning with

M1: 1-M1 at 5.85 min using 332>234 (A.i), metabolite 1-M2 at 7.10 min using 332>220

(A.ii), metabolite 2-M1 and -M2 at 5.42 and 5.65 min, respectively, using 331>265 (A.iii), metabolite 3-M1 at 5.65 min using 333>248 (A.iv), Metabolites 4-M1 and -M2 at 6.01 and

6.42 min, respectively, using 330>288 (A.v), and metabolite 5-M1 at 11.41 min (B).

Analogue 5 appears in B at 12.13 min. Peaks shown in the chromatograms that are not labeled as metabolites appeared in the absence of NADPH and/or the absence of parent compound.

Figure 3-3. High resolution accurate mass spectra and proposed fragment assignments. 1-M1 (A), 1-M2 (B), 2-M1 (C), 2-M2 (D), 3-M1 (E), 4-M1 (F), and 4-

M2 (G). Samples were obtained from incubations with 0.5 mg/mL human liver microsomes and 10 µM of the indicated substrate for 30 minutes at 37 ˚C in the presence of 100 mM potassium phosphate, pH 7.4, and an NADPH regenerating system. Spectra were collected using a Q-Exactive mass spectrometer. Data are representative results from three independent incubations.

46 a RT of 5.65 min, 2-M2 (theoretical m/z: 331.0456, observed m/z: 331.0454) had the same characteristic fragments as 2-M1, suggesting that perhaps the oxygen insertions were in close proximity. These data lead us to propose that 2-M1 and 2-M2 both resulted from oxygen insertions at separate positions on the benzene ring (Figures 3-3C and 3-3D).

Similar to 2, 3 has the same modification to the oxazinone ring of efavirenz, but 3 also has a trans-alkene in place of the efavirenz alkyne. One monooxygenated metabolite with m/z of 333 was detected from human liver microsome incubations with 3 (Figure 3-

2A.iv). With a retention time of 5.65 min, 3-M1 (theoretical m/z: 333.0612, experimental m/z: 333.0609) displayed characteristic fragments of m/z 290.0550, 264.9982, and

248.0081. Our proposed fragments suggest these ions resulted from a loss of CHNO, C5H6, and C4H5NO, respectively. Based on these proposed fragment structures, we predict 3-M1 was formed from oxygen insertion at the 5, 7, or 8 position of the benzene ring (Figure 3-

3E).

Whereas 2 and 3 have a nitrogen in place of the oxygen atom within the oxazinone ring of EFV, analogue 4 has a carbon atom at this same position. Two monooxygenated metabolites (m/z 330) were observed to be formed from incubations with this analogue and human liver microsomes (Figure 3-2A.v). With a RT of 6.01 min, metabolite 4-M1

(theoretical m/z: 330.0503, experimental m/z: 330.0500) displayed characteristic fragment ions of m/z 288.0394, 264.0029, 226.0859, and 184.0755. Our interpretation of these fragments suggests that they resulted from a loss of C2H2O, C5H6, CHF3, and C3H2ClF3O.

Metabolite 4-M2 (theoretical m/z: 330.0503, experimental m/z: 330.0500) had a RT of 6.42 min and displayed characteristic ions of m/z 288.0394, 262.0427, 252.0632, and 226.0859.

We propose these fragment ions resulted from a loss of C2H2O, C5H6, C2H3ClO, and CF3,

47 respectively. Based on our structural predictions and the observation that the elution times differ between 4-M1 and 4-M2 to the same degree as observed for 2-M1 and 2-M2, we predict that 4-M1 and 4-M2 also resulted from separate oxygen insertions on the benzene ring (Figures 3-3F and 3-3G).

Detection of 5 using uHPLC-MS/MS was not successful; therefore, we employed uHPLC-UV to visualize 5 and any potential metabolites. Using this method, one prominent metabolite was detected from incubations with 5 and pooled human liver microsomes

(Figure 3-2B). In the absence of a mass spectrum for this metabolite, no structural predictions can be made regarding the site(s) of oxygen insertion.

Having determined which metabolites can be formed from these analogues, we wanted to test whether analogues 1-5 were also CYP2B6 substrates. To answer this question, we incubated each analogue (10 µM) with cDNA-expressed CYP2B6 (10 nM) and measured formation of each metabolite using selected reaction monitoring or UV detection (for 5). We found that CYP2B6 readily metabolized analogues 1-5 from our panel

(Figure 4), indicating that the changes to the oxazinone ring did not prevent these compounds from being metabolized by CYP2B6. Compared to the other P450s in our panel, activity of CYP2B6 was found to result in the most 1-M1, 2-M2, and 3-M1 formation under the given reaction conditions (Figure 3-4). Conversely, formation of 1-

M2 was most abundant with CYP3A4 and CYP3A5 while CYP1A2 activity resulted in high amounts of 4-M2 (Figure 3-4). The finding that analogues in our panel are also substrates for P450s in the families 1A and 3A is consistent with the fact that CYP1A2,

CYP3A4, and CYP3A5 have previously been shown to metabolize efavirenz (Ward et al.,

2003). The formation of 5-M1 by CYP2B6 and CYP1A2 seemed relatively equal, while

Figure 3-4. Intensities of metabolites formed from incubations with 1 (A.i), 2 (A.ii),

3 (A.iii), 4 (A.iv), and 5 (B). Each analogue (10 µM) was incubated for 30 min with 10 nM of the indicated cDNA-expressed P450 in 100 mM potassium phosphate pH 7.4 and an NADPH regenerating system. Metabolite formation was measured by triple quadrupole uHPLC-MS/MS (A.i-A.iv) or uHPLC-UV (B). Data represent the mean ±

SEM for three independent experiments.

CYP1A1 was also observed to catalyze the formation of 5-M1. Only two metabolites identified in our experiments in human liver microsomes were not found to be formed by

CYP2B6 (2-M1 and 4-M1). However, these metabolites were produced by other P450s

(Figure 3-4). Of note, no other P450 tested displayed a pattern of activity against our panel of efavirenz analogues that was similar to that of CYP2B6.

Since we found 1-5 to be CYP2B6 substrates, we next sought to probe whether the kinetic constants for metabolism of these substrates by CYP2B6 might reveal more subtle impacts of these single atom substitutions on CYP2B6 activity. Results from our previous work with other EFV analogues suggested differences in binding affinity (Cox and

Bumpus, 2014), thus we sought to explore whether these analogues also differed from EFV in their affinity for CYP2B6. In order to explore this question, we performed product formation kinetic analysis with 1-4 using uHPLC-MS/MS and with 5 using uHPLC-UV detection, and compared the results to the KM associated with 8-OH EFV formation (Table

3-1).

We found the formation of 8-OH EFV to be characterized by a KM of 3.6 ± 1.7 µM, which is similar to KM values obtained in the reconstituted system (Bumpus et al., 2006;

Zhang et al., 2011) and in another study also using cDNA-expressed CYP2B6 (Ward et al.,

2003). By comparison, the KM values for metabolism of analogues 1-5 by CYP2B6 were different (Table 3-1). Production of metabolite 1-M2 displayed the lowest KM of 1.0 ± 0.53

µM, which is an approximate 3.3-fold decrease compared to 8-OH EFV and suggests that

CYP2B6 has a higher affinity for this analogue than for EFV. Conversely, CYP2B6 seems to have a lower affinity for 2, 3, and 4, as the formation of 2-M2, 3-M1, and 4-M2 were characterized by KM values 3.0 to 3.9 times higher than for 8-OH EFV (11 ± 4.9, 13 ± 3.4,

Table 3-1. Estimated Michaelis constants for product formation by CYP2B6 Product K (µM) [a] Fold Change measured M 8-OH EFV 3.6 ± 1.7 1 1-M1 1.4 ± 0.76 0.4 1-M2 1.0 ± 0.53 0.3 2-M2 11 ± 4.9 3.0 3-M1 13 ± 3.4 3.6 4-M2 14 ± 6.4 3.9 5-M1 5.5 ± 2.1 1.5

[a] KM values are reported as the mean ± SEM for three independent experiments collected in duplicate.

51 and 14 ± 6.4 µM, respectively). Interestingly, the KM for formation of 5-M1 differed by only a factor of 1.5 from 8-OH EFV (5.5 ± 2.1 µM) indicating a similar affinity of CYP2B6 for 5 and EFV.

Since we observed that CYP2B6 was the only P450 in our panel that metabolized

1-5, we wanted to test whether this activity profile might be conserved in a closely related enzyme in the CYP2B family. To this end, we explored the activity of the well-studied rat enzyme CYP2B1, which shares 76% amino acid sequence identity with CYP2B6, against our panel of EFV analogues (1-5). CYP2B1 produced the same metabolites from 1-5 as we observed for CYP2B6 (Figure 3-5). We then incubated CYP2B1 with each of the eight

EFV analogues reported previously (Cox and Bumpus, 2014). We again found that

CYP2B1 mirrored CYP2B6 activity towards our entire panel of EFV analogues, including

M2 formation from E-dihydroefavirenz, M4 and M5 formation from the pentynyl analogue, and M2 formation from the oxazine analogue. Moreover, with the exception of

2-M2, 3-M1, 4-M2, and 5-M1; metabolite formation from our panel of EFV analogues differed by less than 3-fold between CYP2B6 and CYP2B1 (Figure 3-6). Like CYP2B6,

CYP2B1 did not form metabolites from the two methoxyphenyl analogues or the methoxybenzamide analogue. Interestingly, CYP2B1 displayed activity towards both (R) and (S) -5-Chloro-α-(cyclopropylethynyl)-2-amino-α-(trifluoromethyl) benzenemethanol analogues, which we did not observe previously with CYP2B6 (Cox and Bumpus, 2014).

We then developed a more sensitive selected reaction monitoring method to detect metabolites from these analogues and found CYP2B6 to also form a monooxygenated metabolite from these enantiomers (Figure 3-7). Taken together, these data suggest that

CYP2B1 and CYP2B6 share a parallel profile of activity towards EFV analogues.

Figure 3-5. Chromatograms representing CYP2B6 + b5 and CYP2B1 + b5 activity against EFV and EFV analogues. CYP2B6 or CYP2B1 (10 nM) were incubated with

10 µM EFV or EFV analogues for 30 minutes in the presence of an NADPH regenerating system and 100 mM potassium phosphate pH 7.4. The production of 8

OH-EFV (6.18 min, A), 1-M1 (5.77 min, B), 1-M2 (6.95 min, C), 2-M2 (5.59 min, D),

3-M1 (5.61 min, E), 4-M2 (6.18 min, F), M1 from (S)-5-Chloro-α-

(cyclopropylethynyl)-2-amino-α-(trifluoromethyl)benzenemethanol (5.47 min, H), M1 from (R)-5-Chloro-α-(cyclopropylethynyl)-2-amino-α-

(trifluoromethyl)benzenemethanol (5.47 min, I), M2 from E-Dihydroefavirenz (6.21 min, J), M4 and M5 from rac 6-Chloro-1,4-dihydro-4-(1-pentynyl)-4-(trifluoromethyl)-

2H-3,1-benzoxazin-2-one (5.75 and 6.61 min, respectively, K), and M2 from (4S)-6-

Chloro-4-(2-cyclopropylethynyl)-1,4-dihydro-4-(trifluoromethyl)-2H-3,1-benzoxazine

(6.20 min, L) was monitored with uHPLC-MS/MS. The production of 5-M1 (~11.35 min, G) was monitored by uHPLC-UV. Chromatograms here are representative of three independent experiments.

Figure 3-6. Comparison of metabolite formation by CYP2B1 and CYP2B6. CYP2B1 or CYP2B6 (10 nM) were incubated with EFV or EFV analogues (10 µM) in the presence of an NADPH regenerating system for 30 minutes and the formation of the indicated metabolite measured by uHPLC-MS or uHPLC-UV (5-M1 only). Mean metabolite peak height from three independent incubations with CYP2B6 was set at 100% (represented by a dashed line). Values in the graph represent the metabolite formed by CYP2B1 in three independent experiments and are expressed as a mean percent of metabolite formed by

CYP2B6 ± SEM.

Figure 3-7. Increased sensitivity and selectivity using selected reaction monitoring for detection of M1 from (S) and (R)-5-Chloro-α-(cyclopropylethynyl)-2-amino-α-

(trifluoromethyl)benzenemethanol. CYP2B6 (10 nM) was incubated with (S) or (R)-

5-Chloro-α-(cyclopropylethynyl)-2-amino-α-(trifluoromethyl)benzenemethanol (10

µM) for 30 minutes in 100 mM potassium phosphate pH 7.4 and an NADPH regenerating system for 30 minutes. Metabolite formation was measure using product ion mode scan of m/z 306 (A and B for S and R enantiomers, respectively) or selected reaction monitoring 306>248 (C and D for S and R enantiomers, respectively).

Discussion

In this study, we explored the impact of single heteroatom changes in the oxazinone ring of EFV on CYP2B6 activity towards these compounds. Though analogues 1-5 were all found to be substrates for CYP2B6, the KM values corresponding to the metabolism of

1-5 indicate perturbations of the affinity of CYP2B6 for the substrate. Our data also show that the substrate specificity observed for CYP2B6 and our panel of EFV analogues also extends to another CYP2B enzyme, rat CYP2B1. Together these findings further support a role for the heteroatom composition of the oxazinone ring in CYP2B6 metabolism of

EFV.

Multiple reports characterizing CYP2B6 activity towards many different substrates have been published to date (Wang and Tompkins, 2008; Zanger and Klein, 2013). In addition, several computational approaches have sought to predict potential CYP2B6 substrates based on the structural characteristics of known substrates (Ekins et al., 1998;

Leong et al., 2009; Lewis et al., 2010; Lewis et al., 2004; Wang and Halpert, 2002). Each of these computational approaches determined that lipophilicity, molecular size, and molecular shape were important considerations for predicting CYP2B6 substrates.

Compared to EFV, whose theoretical log P is estimated to be 3.68, 1-5 have similar log P values ranging from 3-4. Since our analogue panel was developed around minor changes to the oxazinone ring, the molecular size and shape were nearly identical for each analogue compared to EFV. Thus, our findings that these analogues are substrates for CYP2B6 is in agreement with previous computational work, however, our observations here demonstrate that minor structural changes that are not captured by changes in log P or molecular size and shape still contribute to substrate binding.

Since no crystal structures of EFV bound to CYP2B6 have been reported to date, additional computational work in the form of molecular modeling provides the best available visualization of EFV in the CYP2B6 active site (Niu et al., 2011). This molecular model shows that hydrogen bonding between CYP2B6 residue E301, various water molecules, and EFV largely account for the calculated interaction energy. Moreover, this network of hydrogen bonds is reported to involve interactions with heteroatoms within the oxazinone ring. Our observation that KM values for the metabolism of 1-4 differed from those for 8-OH EFV formation suggest that changes in the oxazinone ring resulted in altered affinities of these analogues for CYP2B6. The lowest KM values were obtained from incubations with 1, which has a methyl group in place of the carbonyl oxygen of EFV.

Interestingly, 1-M1 was the only metabolite we found in this panel that resulted from oxygen insertion at a site other than the benzene ring, indicating that this analogue must adopt a different conformation in the CYP2B6 active site. We previously showed that elimination of this position altogether, resulted in a lower KM and higher Vmax compared to

EFV (Cox and Bumpus, 2014). Our data obtained here demonstrate that addition of a hydrophobic methyl group also results in tighter binding, as is suggested by the lower KM value. Thus, our work here further supports that the presence of a carbonyl oxygen atom seems to decrease the affinity of CYP2B6 for EFV. In the case of the methyl analogue, the added stereocenter on the heterocyclic ring would result in an additional non-polar group projecting out of the plane of the heterocyclic ring. This methyl group might then be able to participate in intermolecular interactions perpendicular to the heterocyclic ring. The carbonyl oxygen of EFV would be unable to adopt this conformation due to the restriction of the double bond. Instead the carbonyl oxygen would lie only slightly out of the plane

58 delineated by the heterocyclic ring, as has been previously observed in crystals of EFV

(Ravikumar and Sridhar, 2009). Based on this, it is possible that the carbonyl oxygen atom largely in the plane of the heterocyclic ring clashes with nearby CYP2B6 residues. In the previously reported model of CYP2B6 and EFV, L363 and V367 are proposed to lie parallel to the carbonyl oxygen (Niu et al., 2011). It is possible that the added methyl group forms hydrophobic interactions with these residues and thus further stabilizes 1 in the

CPY2B6 active site. This specific possibility seems less likely for the formation of 1-M1 than for 1-M2 due to the proposed position of oxygen insertion.

An approximate three to four fold increase in KM was observed for both quinazolinone analogues (2 and 3) as well as the lone quinolinone analogue (4). Compared to EFV these analogues reflect changes to the oxygen atom within the oxazinone ring. In

EFV, this position may play a role as a hydrogen bond acceptor, but in 2 and 3 this position may be a hydrogen bond donor. Moreover, in 4 the presence of the carbon atom completely abrogates the capacity for hydrogen bonding at this position. Since we observed a 3 to 4- fold effect on affinity of CYP2B6 for 2-4 compared to EFV, hydrogen bonding at this position may have a slight impact on the interaction of CYP2B6 and EFV.

Analogue 5 is the only analogue tested that reflects changes to the EFV nitrogen atom. Like 1-4, this dioxin-2-one analogue was still a CYP2B6 substrate, and 5-M1 exhibited a KM only 1.5 fold higher than 8-OH EFV. The alteration of the EFV nitrogen atom seems to have little effect when changed to an oxygen, indicating that a hydrogen bond donor is not crucial at this position.

Kinetic analysis of 1-5 revealed that each analogue was best characterized by the substrate inhibition equation rather than the Michaelis-Menten equation. Formation of 8,14

59 dihydroxyefavirenz by CYP2B6 is best characterized by the substrate inhibition equation

(Ward et al., 2003), but 8-OH EFV formation has been repeatedly shown to fit well to

Michaelis-Menten (Bumpus et al., 2006; Ogburn et al., 2010; Ward et al., 2003; Zhang et al., 2011). This observed difference with our analogues is of interest and worthy of further study.

To our surprise, we did not observe any dihydroxylated metabolites from our panel.

CYP2B6 has been shown to catalyze the formation of 8, 14 dihydroxyefavirenz (Ward et al., 2003), which we have also observed (data not shown), yet no dihydroxylated metabolites from 1-4 were seen in this study. The use of uHPLC-UV for detection of 5-M1 does not allow us to determine the m/z for this metabolite, thus we cannot rule out the possibility that this is a dihydroxylated metabolite. However, we do not observe any other metabolites formed from 5 with HLM or with individual P450s. We would expect that if

5-M1 resulted from two oxygen insertions, we would have also observed another peak corresponding to a monooxygenated metabolite. The fact that no dihydroxylated metabolites were found from 1-4 suggests that the heteroatom changes explored in this study seem to have an impact on sequential metabolism by CYP2B6.

Our finding that CYP2B1 and CYP2B6 display a similar activity profile towards our panel of EFV analogues suggests the toleration of changes to the EFV structure might be conserved in CYP2B enzymes. Indeed, we observed that none of the other nine P450s tested showed the same profile of selectivity towards EFV, analogues 1-5, and the eight analogues we characterized previously (Cox and Bumpus, 2014). Moreover, the magnitude of metabolite formation from 2B6 and 2B1 differed by no more than six-fold across all 13

EFV analogues studied. In the case of 8-OH EFV production, CYP2B1 formed an

60 approximate 48% of this metabolite compared to 2B6. The largest differences between 2B1 and 2B6 activity against our panel were observed for formation of 2-M2, 3-M1, 4-M2, and

5-M1 (16.4, 17.7, 20.4, and 23.1% of metabolite formation by 2B6, respectively). Three of these analogues, 2-4, represent changes to the oxygen atom within the oxazinone ring of

EFV. It is possible that CYP2B1 is more sensitive to alterations at this position than we observed for CYP2B6. The smallest differences (<2-fold) between CYP2B1 and 2B6 metabolite formation were observed for 1-M1, 1-M2, M5 of the pentynyl analogue, and

M2 of the oxazine analogue. These analogues share changes to the carbonyl oxygen or the cyclopropyl group of EFV, indicating that these positions may reside in similarly tolerant active site environments within CYP2B1 and CYP2B6. Additionally, CYP2B1 was found to form 2-3 fold more metabolite from (S) and (R) -5-Chloro-α-(cyclopropylethynyl)-2- amino-α-(trifluoromethyl) benzenemethanol analogues compared to 2B6. Further kinetics experiments with these two analogues may aid in determining the source of these differences and may provide a more detailed rationale for the increased activity of 2B1 compared to 2B6.

Since the results of this study include only two CYP2B family members, more

CYP2B P450s need to be tested to see if they also display the same activity profile towards

EFV analogues as CYP2B6 and CYP2B1. If a CYP2B family-wide pattern of activity against EFV analogues were to be established, this might imply the existence of an important series of conserved residues within CYP2B enzymes, but not other P450s.

In summary, we have determined that CYP2B6 maintains activity towards analogues of EFV that have discreet substitutions in the heterocyclic ring. These changes, however, result in differences in KM values. The substitution of a methyl group for the

61 carbonyl oxygen increases the affinity of CYP2B6 towards this molecule compared to

EFV, supporting the dispensability of the carbonyl oxygen for metabolism of EFV by

CYP2B6. Additionally, atom substitutions to the internal oxygen and nitrogen atoms do not have strong KM impacts. This indicates that these heteroatoms are not critical for binding of EFV to CYP2B6. Finally, we found that changing the EFV nitrogen atom to oxygen did not prevent this molecule from being a CYP2B6 substrate and had virtually no impact on binding affinity. This work provides further insight into the functional interaction of CYP2B6 and EFV and lays the groundwork for a better understanding of 2B family enzyme activity towards EFV.

Experimental Section

Analogues.

EFV, racemic 8-OH EFV, and EFV analogues were synthesized by Toronto

Research Chemicals (Toronto, Canada) and were ≥ 97% pure, according to the manufacturer. The analogues used in this study were 6-Chloro-4-(2-cyclopropylethynyl)-

1,4-dihydro-2-methyl-4-(trifluoromethyl)-2H-3,1-benzoxazine (analogue 1), (4S)-6-

Chloro-4-(2-cyclopropylethynyl)-3,4-dihydro-4-(trifluoromethyl)-2(1H)-quinazolinone

(2), (4S)-6-Chloro-4-[(1E)-2-cyclopropylethenyl]-3,4-dihydro-4-

(trifluoromethyl)-,2(1H)-quinazolinone (3), 6-Chloro-4-(cyclopropylethynyl)-3,4- dihydro-4-(trifluoromethyl)-2(1H)-quinolinone (4), and 6-Chloro-4-(cyclopropylethynyl)-

4-(trifluoromethyl)-4H-benzo[d][1,3]dioxin-2-one (5) (Figure 1). The synthesis for analogues 2 and 3 was originally reported by DuPont Pharmaceuticals (Corbett et al.,

1999). We also made use of eight additional EFV analogues, which were reported in our previous publication (Cox and Bumpus, 2014). These analogues were (S)-5-Chloro-α-

(cyclopropylethynyl)-2-amino-α-(trifluoromethyl)benzenemethanol, (R)-5-Chloro-α-

(cyclopropylethynyl)-2-amino-α-(trifluoromethyl)benzenemethanol, (E)-

Dihydroefavirenz, rac 6-Chloro-1,4-dihydro-4-(1-pentynyl)-4-(trifluoromethyl)-2H-3,1- benzoxazin-2-one, (R)-5-Chloro-α-(cyclopropylethynyl)-2-[[(4- methoxyphenyl)methyl]amino]-α-(trifluoromethyl)benzenemethanol, (S)-5-Chloro-α-

(cyclopropylethynyl)-2-[[(4methoxyphenyl)methyl]amino]-α-

(trifluoromethyl)benzenemethanol, rac N- [4-Chloro-2-[3-cyclopropyl-1-hydroxy-1-

(trifluoromethyl)-2-propynyl]phenyl]-4-methoxybenzamide, and (4S)-6-Chloro-4-(2- cyclopropylethynyl)-1,4-dihydro-4-(trifluoromethyl)-2H-3,1-benzoxazine.

Enzyme assays.

Enzyme incubations were carried out essentially as previously described (Cox and

Bumpus, 2014). Reactions were pre-incubated in 100 mM potassium phosphate pH 7.4 containing 0.5 mg/mL pooled human liver microsomes (Xenotech LLC, Lenexa, KS) or

10 nM cDNA-expressed P450 (Supersomes ® , Corning, Corning, NY) and 10 µM of each analogue for 5 minutes at 37˚C. Reactions were then initiated by the addition of an NADPH regenerating system (Corning, Corning, NY) for a total volume of 100 µL and allowed to proceed for 30 minutes at 37˚C. After this incubation, the protein was precipitated by the addition of 100 µL ice-cold acetonitrile. Precipitated reactions were incubated on ice for

10 minutes then centrifuged for 10 minutes at 10,000 x g at 4 ˚C and the supernatant dried in a vacuum centrifuge. Samples were reconstituted in 100 µL methanol in preparation for mass spectrometry or UV detection.

For assays measuring product formation for determination of KM, pilot incubations were performed to determine the linearity with respect to enzyme concentration and time.

CYP2B6 (5-20 nM) was incubated as above with eight concentrations (0.5-100 µM) of

EFV or the indicated EFV analogue and metabolism allowed to occur in the presence of an

NADPH regenerating system for 5-20 minutes. Substrate concentrations of 0.5-30 µM were used for 3 due to limited solubility in aqueous reaction buffer. An equal volume (100

µL) of ice-cold acetonitrile was added to terminate the reaction and precipitate protein.

Reactions were then processed for mass spectrometry as above. In the absence of authentic metabolite standards, we were unable to quantitate product formation for metabolism of 1-

5. With our goal of comparing relative affinities of these substrates for CYP2B6 in mind, we sought to obtain only KM values from analysing product formation.

In order to draw comparisons between enzymatic activity of CYP2B1 and

CYP2B6, we sought to use a similar system as our experiments using cDNA-expressed

CYP2B6. To that end, cDNA-expressed CYP2B1 (Supersomes ®, Corning, Corning, NY) were obtained. In contrast to our other experiments presented here, this system also contained cytochrome b5, thus, for comparison purposes, cDNA-expressed CYP2B6 containing b5 (Supersomes ®, Corning, Corning, NY) was tested side-by-side to CYP2B1.

The molar ratio of CYP2B6:reductase:b5 was 1:10:2, as determined by the manufacturer.

The molar ration of CYP2B1:reductase:b5 was 1:1:2, according to the manufacturer. These assays were performed and samples processed in the same manner described above.

Analyte Detection.

Detection of EFV, previously reported EFV analogues, and EFV analogues 1-4 was achieved essentially as previously described (Cox and Bumpus, 2014). Reconstituted samples were resolved with an Xterra C18 column (2.1 x 50 mm, 2.5 µm; Waters, Milford,

MA) using a Dionex Ultimate 3000 uHPLC system coupled to a TSQ Vantage Triple Stage

Quadrupole mass spectrometer (Thermo Scientific, Pittsburgh, PA) or a Q-Exactive benchtop Orbitrap mass spectrometer (Thermo Scientific, Pittsburgh, PA). The mobile phases used for all analogues except 5 were water with 0.1% formic acid (mobile phase A) and acetonitrile with 0.1% formic acid (mobile phase B). The gradient for efavirenz and all analogues except 5 was identical: 10% B from 0-0.5 min, 10-95% B from 0.5-10 min, 95%

B from 10-10.5 min, 95-10% B from 10.5-10.6 minutes, and 10% B from 10.6-11.5 minutes. Positive ion mode was used for detection of 1-4 and corresponding metabolites.

Detection of analogues and metabolites using our triple quadrupole instrument was first achieved with product ion mode. Upon selecting unique fragments from each observed analyte, we developed transitions for selected reaction monitoring in order to increase sensitivity (Table 3-2). The transition used to detect monooxygenated metabolites from

(S) and (R) -5-Chloro-α-(cyclopropylethynyl)-2-amino-α-

(trifluoromethyl)benzenemethanol was m/z 306>248. For Q-Exactive experiments, PRM scans were constructed for each metabolite using the proposed molecular formula (1 –

C15H13ClF3NO2, 2 – C14H10ClF3N2O2, 3 – C14H12ClF3N2O2, and 4 – C15H11ClF3NO2.

Fragmentation spectra were obtained with stepped NCE of 20, 30 and 40.

Analogue 5 was not readily detectable via mass spectrometry. Therefore, we instead used UV to detect parent and to measure metabolite formation. Separation of 5 and its metabolite was accomplished with the same Xterra C18 column as above using an Acquity uHPLC system coupled to a photodiode array detector set at 220 nm (Waters, Milford,

MA). The solvents for uHPLC-UV were water with 0.1% trifluoroacetic acid (mobile phase A) and acetonitrile with 0.1% trifluoroacetic acid (mobile phase B). The chromatographic program for 5 was 10% B from 0-0.5 minutes, 10-95% B from 0.5-15

Table 3-2. Selected reaction monitoring transitions used in detection of

EFV, 1-4, and metabolites via triple quadrupole mass spectrometry.

Analyte Parent m/z Selected fragment m/z EFV 314 - 8-OH EFV 330 162 1 316 246 1-M1 332 234 1-M2 332 220 2 315 249 2-M1 and -M2 331 265 3 317 232 3-M1 333 248 4 314 272 4-M1 and -M2 330 288

66 minutes, 95% B from 15-16 minutes, 95-10%B from 16-16.1 minutes, and 10% B from

16.1-20 minutes. All solvents used for uHPLC-MS/MS and uHPLC-UV were of the highest grade commercially available.

Data Analysis.

Reported KM values were estimated from fitting product formation data to the substrate inhibition equation using GraphPad Prism version 7 (GraphPad Software Inc.,

San Diego, CA). Initial data fits were made to the Michaelis-Menten equation; however, a better R2 value was found for substrate inhibition. For comparison of CYP2B1 and

CYP2B6 activity towards EFV analogues, observed metabolite peak height from CYP2B6 incubations was set at 100% and metabolite peak height from CYP2B1 incubations normalized to this value.

Chapter 4

Conclusion

The relationship between EFV and CYP2B6 has been the subject of many clinical and basic science studies in recent years. Indeed, the discovery of differences between the activities of CYP2B6 variants towards EFV drew attention to potential interindividual variabilities in CYP2B6 substrate metabolism. By demonstrating a statistical association of many CYP2B6 variants with altered EFV plasma concentrations (Haas et al., 2004;

Rotger et al., 2007; Wang et al., 2006), researchers are drawing attention to the interindividual variability of EFV metabolism and the metabolism of other CYP2B6 substrates (Penzak et al., 2007; Raccor et al., 2012). These associations point to a role of

CYP2B6 as a player in individualized medicine considerations. Our hope is that by providing some basis for explaining why EFV is a substrate for CYP2B6, we might begin to gather information about what makes any drug a substrate for CYP2B6. Such knowledge may guide drug development so as to avoid producing drugs that are primarily metabolized by this polymorphic enzyme and are thus associated with high interindividual variability in metabolism.

While many traditional structure-activity studies focus solely on the medicinal activity of a compound towards its target, we focused our work on comparing the metabolism of structurally related compounds. Our study, to our knowledge, is the first to gather structural information addressing why EFV is a substrate for CYP2B6. Such an approach is widely applicable and has also been used to explore the structural characteristics that render a molecule a ligand for a given receptor. Such work involving

EFV activation of nuclear receptors, is concurrently being performed in our lab with our panel of EFV analogues (Lade and Bumpus, unpublished).

Though we demonstrate evidence for determinants of EFV metabolism by

CYP2B6, we do not know yet if this information translates to other substrates. The exact oxazinone moiety of EFV is not found in other known CYP2B6 substrates, but heterocyclic rings are common features (see Figure 1-4). For example, the structure of the anti- convulsant mephenytoin contains a five membered heterocyclic ring. It is tempting to speculate that this region, which is proximal to the site of metabolism by CYP2B6, is important to CYP2B6 substrate specificity. Therefore, alterations in this ring, even at the single atom level, may result in differences in metabolism. While initial data presented here in Chapter 2 indicate that breaking open the heterocyclic ring has marked effects of

CYP2B6 activity towards EFV, we further show that even changing individual heteroatoms in this ring alter KM values. This observation has intriguing implications to mephenytoin, among other CYP2B6 substrates. Further studies involving single atom alterations of known CYP2B6 substrates are necessary in order to begin generalizing our findings.

The medicinal activity of structurally altered compounds towards their biological target must be a consideration in such further studies. This point complicates the potential for expanding our results. Consider the following example: drug X is highly active against its target with a nanomolar KI. Drug X is also primarily metabolized by CYP2B6. Single atom changes in the structure of drug X lead to decreased CYP2B6 activity towards this molecule; however, the new structure no longer retains any activity towards the medicinal target. In such a scenario, multiple rounds of iterative structure-activity studies are necessary to balance the metabolism and clinical activity of the compound.

Our work would not be possible without advances in the field of small molecule detection by mass spectrometry. The development of triple quadrupole instruments allows for rapid, sensitive, and specific detection of a given small molecule. Moreover, subjecting

70 analytes to liquid chromatography allows for the separation of structurally related molecules (e.g. metabolites). Using mass spectrometry to propose sites of oxygen insertion for most of the metabolites identified in this study allows us to explore new potentially interesting aspects of EFV and EFV analogue metabolism. For example, we observed that by opening the cyclopropyl ring of EFV many new monooxygenated metabolites resulting from oxygen insertion on the open ring were detected. This suggests that the substrate is now oriented with this open ring proximal to the heme group. In contrast, no monooxygenated metabolites of EFV on the cyclopropyl ring are normally observed. The residues in CYP2B6 that are involved in this new orientation of the open cyclopropyl ring remain to be determined.

In the work presented here, we chose to focus on CYP2B6 metabolism of our panel of EFV analogues; however, many other potential avenues of study are possible with these small molecules. Previous work with CYP2B6 has shown that the enzyme adopts distinct conformation with different substrates (Gay et al., 2010; Shah et al., 2011; Shah et al.,

2012). Additionally, the ratio of NADPH consumption to product formation can vary from substrate to substrate (Grinkova et al., 2013; Gruenke et al., 1995; Narasimhulu, 2007), possibly due to the ability of solvent to access the active site (Loida and Sligar, 1993).

Since CYP2B6 metabolism of EFV is known to be uncoupled (Bumpus and Hollenberg,

2008), our panel could be used to provide detailed information regarding the relationship between EFV structure and reaction uncoupling. In addition, EFV and 8-OH EFV are known to be mechanism-based inactivators of CYP2B6 (Bumpus et al., 2006). The structural characteristics of EFV that enable it to inactivate CYP2B6 remain unclear.

Studies with our panel of EFV analogues may aid in elucidating the important portions of

EFV for inactivation of CYP2B6. A final possibility for continued study would involve using our panel of EFV analogues for crystallization of CYP2B6. Since no crystal structures of CYP2B6 and EFV are currently available, our panel may offer possible small molecules for successful co-crystallization. Such experiments would provide a detailed snapshot of the molecular interactions between CYP2B6 and an EFV-like substrate.

References

Ariyoshi N, Ohara M, Kaneko M, Afuso S, Kumamoto T, Nakamura H, Ishii I, Ishikawa T and Kitada M (2011) Q172H replacement overcomes effects on the metabolism of cyclophosphamide and efavirenz caused by CYP2B6 variant with Arg262. Drug Metab Dispos 39(11): 2045-2048.

Avery LB, Parsons TL, Meyers DJ and Hubbard WC (2010) A highly sensitive ultra performance liquid chromatography-tandem mass spectrometric (UPLC-MS/MS) technique for quantitation of protein free and bound efavirenz (EFV) in human seminal and blood plasma. J Chromatogr B Analyt Technol Biomed Life Sci 878(31): 3217-3224.

Boerjan W, Ralph J and Baucher M (2003) Lignin biosynthesis. Annual review of plant biology 54: 519-546.

Bumpus NN (2011) Efavirenz and 8-hydroxyefavirenz induce cell death via a JNK- and BimEL-dependent mechanism in primary human hepatocytes. Toxicol Appl Pharmacol 257(2): 227-234.

Bumpus NN and Hollenberg PF (2008) Investigation of the mechanisms underlying the differential effects of the K262R mutation of P450 2B6 on catalytic activity. Mol Pharmacol 74(4): 990-999.

Bumpus NN, Kent UM and Hollenberg PF (2006) Metabolism of efavirenz and 8- hydroxyefavirenz by P450 2B6 leads to inactivation by two distinct mechanisms. J Pharmacol Exp Ther 318(1): 345-351.

Bumpus NN, Sridar C, Kent UM and Hollenberg PF (2005) The naturally occurring cytochrome P450 (P450) 2B6 K262R mutant of P450 2B6 exhibits alterations in substrate metabolism and inactivation. Drug Metab Dispos 33(6): 795-802.

Bunten H, Liang WJ, Pounder D, Seneviratne C and Osselton MD (2011) CYP2B6 and OPRM1 gene variations predict methadone-related deaths. Addict Biol 16(1): 142- 144.

Chang TK, Weber GF, Crespi CL and Waxman DJ (1993) Differential activation of cyclophosphamide and ifosphamide by cytochromes P-450 2B and 3A in human liver microsomes. Cancer Res 53(23): 5629-5637.

Corbett JW, Ko SS, Rodgers JD, Jeffrey S, Bacheler LT, Klabe RM, Diamond S, Lai CM, Rabel SR, Saye JA, Adams SP, Trainor GL, Anderson PS and Erickson-Viitanen SK (1999) Expanded-spectrum nonnucleoside reverse transcriptase inhibitors inhibit clinically relevant mutant variants of human immunodeficiency virus type 1. Antimicrob Agents Chemother 43(12): 2893-2897.

Cox PM and Bumpus NN (2014) Structure–Activity Studies Reveal the Oxazinone Ring Is a Determinant of Cytochrome P450 2B6 Activity Toward Efavirenz. ACS Medicinal Chemistry Letters 5(10): 1156-1161.

Cox PM and Bumpus NN (2016) Single Heteroatom Substitutions in the Efavirenz Oxazinone Ring Impact Metabolism by CYP2B6. ChemMedChem: n/a-n/a. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.

Crettol S, Deglon JJ, Besson J, Croquette-Krokkar M, Gothuey I, Hammig R, Monnat M, Huttemann H, Baumann P and Eap CB (2005) Methadone enantiomer plasma levels, CYP2B6, CYP2C19, and CYP2C9 genotypes, and response to treatment. Clin Pharmacol Ther 78(6): 593-604.

Daly AK (2006) Significance of the minor cytochrome P450 3A isoforms. Clin Pharmacokinet 45(1): 13-31.

Desta Z, Saussele T, Ward B, Blievernicht J, Li L, Klein K, Flockhart DA and Zanger UM (2007) Impact of CYP2B6 polymorphism on hepatic efavirenz metabolism in vitro. Pharmacogenomics 8(6): 547-558.

Dobrinas M, Crettol S, Oneda B, Lahyani R, Rotger M, Choong E, Lubomirov R, Csajka C and Eap CB (2013) Contribution of CYP2B6 alleles in explaining extreme (S)- methadone plasma levels: a CYP2B6 gene resequencing study. Pharmacogenet Genomics 23(2): 84-93.

Ekins S, Vandenbranden M, Ring BJ, Gillespie JS, Yang TJ, Gelboin HV and Wrighton SA (1998) Further characterization of the expression in liver and catalytic activity of CYP2B6. J Pharmacol Exp Ther 286(3): 1253-1259.

Faucette SR, Wang H, Hamilton GA, Jolley SL, Gilbert D, Lindley C, Yan B, Negishi M and LeCluyse EL (2004) Regulation of CYP2B6 in primary human hepatocytes by prototypical inducers. Drug Metab Dispos 32(3): 348-358.

Faucette SR, Zhang TC, Moore R, Sueyoshi T, Omiecinski CJ, LeCluyse EL, Negishi M and Wang H (2007) Relative activation of human pregnane X receptor versus constitutive androstane receptor defines distinct classes of CYP2B6 and CYP3A4 inducers. J Pharmacol Exp Ther 320(1): 72-80.

Gay SC, Shah MB, Talakad JC, Maekawa K, Roberts AG, Wilderman PR, Sun L, Yang JY, Huelga SC, Hong WX, Zhang Q, Stout CD and Halpert JR (2010) Crystal structure of a cytochrome P450 2B6 genetic variant in complex with the inhibitor 4-(4-chlorophenyl)imidazole at 2.0-A resolution. Mol Pharmacol 77(4): 529-538.

Gervot L, Rochat B, Gautier JC, Bohnenstengel F, Kroemer H, de Berardinis V, Martin H, Beaune P and de Waziers I (1999) Human CYP2B6: expression, inducibility and catalytic activities. Pharmacogenetics 9(3): 295-306.

Grinkova YV, Denisov IG, McLean MA and Sligar SG (2013) Oxidase uncoupling in heme monooxygenases: human cytochrome P450 CYP3A4 in Nanodiscs. Biochem Biophys Res Commun 430(4): 1223-1227.

Gruenke LD, Konopka K, Cadieu M and Waskell L (1995) The stoichiometry of the cytochrome P-450-catalyzed metabolism of methoxyflurane and benzphetamine in the presence and absence of cytochrome b5. The Journal of biological chemistry 270(42): 24707-24718.

Haas DW, Ribaudo HJ, Kim RB, Tierney C, Wilkinson GR, Gulick RM, Clifford DB, Hulgan T, Marzolini C and Acosta EP (2004) Pharmacogenetics of efavirenz and central nervous system side effects: an Adult AIDS Clinical Trials Group study. AIDS 18(18): 2391-2400.

Hanley MJ, Cancalon P, Widmer WW and Greenblatt DJ (2011) The effect of grapefruit juice on drug disposition. Expert Opin Drug Metab Toxicol 7(3): 267-286.

Hesse LM, Venkatakrishnan K, Court MH, von Moltke LL, Duan SX, Shader RI and Greenblatt DJ (2000) CYP2B6 mediates the in vitro hydroxylation of bupropion: potential drug interactions with other antidepressants. Drug Metab Dispos 28(10): 1176-1183.

Hesse LM, von Moltke LL, Shader RI and Greenblatt DJ (2001) Ritonavir, efavirenz, and nelfinavir inhibit CYP2B6 activity in vitro: potential drug interactions with bupropion. Drug Metab Dispos 29(2): 100-102.

Hoffman SM, Nelson DR and Keeney DS (2001) Organization, structure and evolution of the CYP2 gene cluster on human chromosome 19. Pharmacogenetics 11(8): 687- 698.

Hofmann MH, Blievernicht JK, Klein K, Saussele T, Schaeffeler E, Schwab M and Zanger UM (2008) Aberrant splicing caused by single nucleotide polymorphism c.516G>T [Q172H], a marker of CYP2B6*6, is responsible for decreased expression and activity of CYP2B6 in liver. J Pharmacol Exp Ther 325(1): 284-292.

Huang SM, Temple R, Throckmorton DC and Lesko LJ (2007) Drug interaction studies: study design, data analysis, and implications for dosing and labeling. Clin Pharmacol Ther 81(2): 298-304.

Johnson EF and Stout CD (2013) Structural diversity of eukaryotic membrane cytochrome p450s. The Journal of biological chemistry 288(24): 17082-17090.

Johnson GG, Lin K, Cox TF, Oates M, Sibson DR, Eccles R, Lloyd B, Gardiner LJ, Carr DF, Pirmohamed M, Strefford JC, Oscier DG, Gonzalez de Castro D, Else M, Catovsky D and Pettitt AR (2013) CYP2B6*6 is an independent determinant of inferior response to fludarabine plus cyclophosphamide in chronic lymphocytic leukemia. Blood 122(26): 4253-4258.

Kenedi CA and Goforth HW (2011) A systematic review of the psychiatric side-effects of efavirenz. AIDS and behavior 15(8): 1803-1818.

Kharasch ED, Hoffer C, Whittington D and Sheffels P (2004) Role of hepatic and intestinal cytochrome P450 3A and 2B6 in the metabolism, disposition, and miotic effects of methadone. Clin Pharmacol Ther 76(3): 250-269.

Kharasch ED, Regina KJ, Blood J and Friedel C (2015) Methadone Pharmacogenetics: CYP2B6 Polymorphisms Determine Plasma Concentrations, Clearance, and Metabolism. Anesthesiology 123(5): 1142-1153.

Klein K, Lang T, Saussele T, Barbosa-Sicard E, Schunck WH, Eichelbaum M, Schwab M and Zanger UM (2005) Genetic variability of CYP2B6 in populations of African and Asian origin: allele frequencies, novel functional variants, and possible implications for anti-HIV therapy with efavirenz. Pharmacogenet Genomics 15(12): 861-873.

Kumar S, Zhao Y, Sun L, Negi SS, Halpert JR and Muralidhara BK (2007) Rational engineering of human cytochrome P450 2B6 for enhanced expression and stability: importance of a Leu264->Phe substitution. Mol Pharmacol 72(5): 1191-1199.

Lade JM, Avery LB and Bumpus NN (2013) Human biotransformation of the nonnucleoside reverse transcriptase inhibitor rilpivirine and a cross-species metabolism comparison. Antimicrob Agents Chemother 57(10): 5067-5079.

Lasker JM, Wester MR, Aramsombatdee E and Raucy JL (1998) Characterization of CYP2C19 and CYP2C9 from human liver: respective roles in microsomal tolbutamide, S-mephenytoin, and omeprazole hydroxylations. Arch Biochem Biophys 353(1): 16-28.

Leong MK, Chen YM and Chen TH (2009) Prediction of human cytochrome P450 2B6- substrate interactions using hierarchical support vector regression approach. Journal of computational chemistry 30(12): 1899-1909.

Lewis DF, Ito Y and Lake BG (2010) Quantitative structure-activity relationships (QSARs) for inhibitors and substrates of CYP2B enzymes: importance of compound lipophilicity in explanation of potency differences. Journal of enzyme inhibition and medicinal chemistry 25(5): 679-684.

Lewis DF, Lake BG and Dickins M (2004) Quantitative structure-activity relationships within a homologous series of 7-alkoxyresorufins exhibiting activity towards CYP1A and CYP2B enzymes: molecular modelling studies on key members of the resorufin series with CYP2C5-derived models of human CYP1A1, CYP1A2, CYP2B6 and CYP3A4. Xenobiotica 34(6): 501-513.

Li X, Jeso V, Heyward S, Walker GS, Sharma R, Micalizio GC and Cameron MD (2014) Characterization of T-5 N-oxide formation as the first highly selective measure of CYP3A5 activity. Drug Metab Dispos 42(3): 334-342.

Loida PJ and Sligar SG (1993) Molecular recognition in cytochrome P-450: mechanism for the control of uncoupling reactions. Biochemistry 32(43): 11530-11538.

Lu Y, Fuchs EJ, Hendrix CW and Bumpus NN (2014) CYP3A5 genotype impacts maraviroc concentrations in healthy volunteers. Drug Metab Dispos 42(11): 1796- 1802.

Lu Y, Hendrix CW and Bumpus NN (2012) Cytochrome P450 3A5 plays a prominent role in the oxidative metabolism of the anti-human immunodeficiency virus drug maraviroc. Drug Metab Dispos 40(12): 2221-2230.

Lutz JD, Dixit V, Yeung CK, Dickmann LJ, Zelter A, Thatcher JE, Nelson WL and Isoherranen N (2009) Expression and functional characterization of cytochrome P450 26A1, a retinoic acid hydroxylase. Biochem Pharmacol 77(2): 258-268.

Mancy A, Antignac M, Minoletti C, Dijols S, Mouries V, Duong NT, Battioni P, Dansette PM and Mansuy D (1999) Diclofenac and its derivatives as tools for studying human cytochromes P450 active sites: particular efficiency and regioselectivity of P450 2Cs. Biochemistry 38(43): 14264-14270.

Meng X, Yin K, Wang J, Dong P, Liu L, Shen Y, Shen L, Ma Q, Lu H and Cai W (2015) Effect of CYP2B6 Gene Polymorphisms on Efavirenz Plasma Concentrations in Chinese Patients with HIV Infection. PLoS One 10(6): e0130583.

Mutlib AE, Chen H, Nemeth G, Gan LS and Christ DD (1999a) Liquid chromatography/mass spectrometry and high-field nuclear magnetic resonance characterization of novel mixed diconjugates of the non-nucleoside human immunodeficiency virus-1 reverse transcriptase inhibitor, efavirenz. Drug Metab Dispos 27(9): 1045-1056.

Mutlib AE, Chen H, Nemeth GA, Markwalder JA, Seitz SP, Gan LS and Christ DD (1999b) Identification and characterization of efavirenz metabolites by liquid chromatography/mass spectrometry and high field NMR: species differences in the metabolism of efavirenz. Drug Metab Dispos 27(11): 1319-1333.

Narasimhulu S (2007) Differential behavior of the sub-sites of cytochrome 450 active site in binding of substrates, and products (implications for coupling/uncoupling). Biochim Biophys Acta 1770(3): 360-375.

Nebert DW, Adesnik M, Coon MJ, Estabrook RW, Gonzalez FJ, Guengerich FP, Gunsalus IC, Johnson EF, Kemper B, Levin W and et al. (1987) The P450 gene superfamily: recommended nomenclature. DNA 6(1): 1-11.

Nebert DW and Russell DW (2002) Clinical importance of the cytochromes P450. Lancet 360(9340): 1155-1162.

Nelson D (2016) The Cytochrome P450 Homepage. drnelson.uthsc.edu/CytochromeP450.html

Niu RJ, Zheng QC, Zhang JL and Zhang HX (2011) Analysis of clinically relevant substrates of CYP2B6 enzyme by computational methods. J Mol Model 17(11): 2839-2846.

Obach RS and Reed-Hagen AE (2002) Measurement of Michaelis constants for cytochrome P450-mediated biotransformation reactions using a substrate depletion approach. Drug Metab Dispos 30(7): 831-837.

Ogburn ET, Jones DR, Masters AR, Xu C, Guo Y and Desta Z (2010) Efavirenz primary and secondary metabolism in vitro and in vivo: identification of novel metabolic pathways and cytochrome P450 2A6 as the principal catalyst of efavirenz 7- hydroxylation. Drug Metab Dispos 38(7): 1218-1229.

Omura T (2013) Contribution of cytochrome P450 to the diversification of eukaryotic organisms. Biotechnol Appl Biochem 60(1): 4-8.

Omura T and Sato R (1964a) The Carbon Monoxide-Binding Pigment of Liver Microsomes. I. Evidence for Its Hemoprotein Nature. The Journal of biological chemistry 239: 2370-2378.

Omura T and Sato R (1964b) The Carbon Monoxide-Binding Pigment of Liver Microsomes. Ii. Solubilization, Purification, and Properties. The Journal of biological chemistry 239: 2379-2385.

Penzak SR, Kabuye G, Mugyenyi P, Mbamanya F, Natarajan V, Alfaro RM, Kityo C, Formentini E and Masur H (2007) Cytochrome P450 2B6 (CYP2B6) G516T influences nevirapine plasma concentrations in HIV-infected patients in Uganda. HIV Med 8(2): 86-91.

Pinot F and Beisson F (2011) Cytochrome P450 metabolizing fatty acids in plants: characterization and physiological roles. The FEBS journal 278(2): 195-205.

Raccor BS, Claessens AJ, Dinh JC, Park JR, Hawkins DS, Thomas SS, Makar KW, McCune JS and Totah RA (2012) Potential contribution of cytochrome P450 2B6 to hepatic 4-hydroxycyclophosphamide formation in vitro and in vivo. Drug Metab Dispos 40(1): 54-63. 79

Radloff R, Gras A, Zanger UM, Masquelier C, Arumugam K, Karasi JC, Arendt V, Seguin- Devaux C and Klein K (2013) Novel CYP2B6 enzyme variants in a Rwandese population: functional characterization and assessment of in silico prediction tools. Hum Mutat 34(5): 725-734.

Rakhmanina NY and van den Anker JN (2010) Efavirenz in the therapy of HIV infection. Expert Opin Drug Metab Toxicol 6(1): 95-103.

Ravichandran KG, Boddupalli SS, Hasermann CA, Peterson JA and Deisenhofer J (1993) Crystal structure of hemoprotein domain of P450BM-3, a prototype for microsomal P450's. Science 261(5122): 731-736.

Ravikumar K and Sridhar B (2009) Molecular and Crystal Structure of Efavirenz, a Potent and Specific Inhibitor of HIV-1 Reverse Transcriptase, and Its Monohydrate. Molecular Crystals and Liquid Crystals 515: 190-198.

Ren J, Milton J, Weaver KL, Short SA, Stuart DI and Stammers DK (2000) Structural basis for the resilience of efavirenz (DMP-266) to drug resistance mutations in HIV-1 reverse transcriptase. Structure 8(10): 1089-1094.

Rotger M, Tegude H, Colombo S, Cavassini M, Furrer H, Decosterd L, Blievernicht J, Saussele T, Gunthard HF, Schwab M, Eichelbaum M, Telenti A and Zanger UM (2007) Predictive value of known and novel alleles of CYP2B6 for efavirenz plasma concentrations in HIV-infected individuals. Clin Pharmacol Ther 81(4): 557-566.

Scott EE, Spatzenegger M and Halpert JR (2001) A truncation of 2B subfamily cytochromes P450 yields increased expression levels, increased solubility, and decreased aggregation while retaining function. Arch Biochem Biophys 395(1): 57- 68.

Shah MB, Pascual J, Zhang Q, Stout CD and Halpert JR (2011) Structures of cytochrome P450 2B6 bound to 4-benzylpyridine and 4-(4-nitrobenzyl)pyridine: insight into inhibitor binding and rearrangement of active site side chains. Mol Pharmacol 80(6): 1047-1055.

Shah MB, Wilderman PR, Pascual J, Zhang Q, Stout CD and Halpert JR (2012) Conformational adaptation of human cytochrome P450 2B6 and rabbit cytochrome P450 2B4 revealed upon binding multiple amlodipine molecules. Biochemistry 51(37): 7225-7238. 80

Shebley M, Kent UM, Ballou DP and Hollenberg PF (2009) Mechanistic analysis of the inactivation of cytochrome P450 2B6 by phencyclidine: effects on substrate binding, electron transfer, and uncoupling. Drug Metab Dispos 37(4): 745-752.

Sim SC and Ingelman-Sundberg M (2010) The Human Cytochrome P450 (CYP) Allele Nomenclature website: a peer-reviewed database of CYP variants and their associated effects. Hum Genomics 4(4): 278-281.

Sim SC and Ingelman-Sundberg M (2013) Update on allele nomenclature for human cytochromes P450 and the Human Cytochrome P450 Allele (CYP-allele) Nomenclature Database. Methods Mol Biol 987: 251-259.

Telenti A and Zanger UM (2008) Pharmacogenetics of anti-HIV drugs. Annu Rev Pharmacol Toxicol 48: 227-256.

Tovar-y-Romo LB, Bumpus NN, Pomerantz D, Avery LB, Sacktor N, McArthur JC and Haughey NJ (2012) Dendritic spine injury induced by the 8-hydroxy metabolite of efavirenz. J Pharmacol Exp Ther 343(3): 696-703.

Tsuchiya K, Gatanaga H, Tachikawa N, Teruya K, Kikuchi Y, Yoshino M, Kuwahara T, Shirasaka T, Kimura S and Oka S (2004) Homozygous CYP2B6 *6 (Q172H and K262R) correlates with high plasma efavirenz concentrations in HIV-1 patients treated with standard efavirenz-containing regimens. Biochem Biophys Res Commun 319(4): 1322-1326.

Wang H and Tompkins LM (2008) CYP2B6: new insights into a historically overlooked cytochrome P450 isozyme. Curr Drug Metab 9(7): 598-610.

Wang J, Sonnerborg A, Rane A, Josephson F, Lundgren S, Stahle L and Ingelman- Sundberg M (2006) Identification of a novel specific CYP2B6 allele in Africans causing impaired metabolism of the HIV drug efavirenz. Pharmacogenet Genomics 16(3): 191-198.

Wang Q and Halpert JR (2002) Combined three-dimensional quantitative structure-activity relationship analysis of cytochrome P450 2B6 substrates and protein homology modeling. Drug Metab Dispos 30(1): 86-95.

Ward BA, Gorski JC, Jones DR, Hall SD, Flockhart DA and Desta Z (2003) The cytochrome P450 2B6 (CYP2B6) is the main catalyst of efavirenz primary and 81

secondary metabolism: implication for HIV/AIDS therapy and utility of efavirenz as a substrate marker of CYP2B6 catalytic activity. J Pharmacol Exp Ther 306(1): 287-300.

Waxman DJ and Azaroff L (1992) Phenobarbital induction of cytochrome P-450 gene expression. The Biochemical journal 281 ( Pt 3): 577-592.

Wilderman PR, Shah MB, Jang HH, Stout CD and Halpert JR (2013) Structural and thermodynamic basis of (+)-alpha-pinene binding to human cytochrome P450 2B6. J Am Chem Soc 135(28): 10433-10440.

Williams PA, Cosme J, Sridhar V, Johnson EF and McRee DE (2000) Mammalian microsomal cytochrome P450 monooxygenase: structural adaptations for membrane binding and functional diversity. Mol Cell 5(1): 121-131.

Xie HG, Wood AJ, Kim RB, Stein CM and Wilkinson GR (2004) Genetic variability in CYP3A5 and its possible consequences. Pharmacogenomics 5(3): 243-272.

Xu C, Ogburn ET, Guo Y and Desta Z (2012) Effects of the CYP2B6*6 allele on catalytic properties and inhibition of CYP2B6 in vitro: implication for the mechanism of reduced efavirenz metabolism and other CYP2B6 substrates in vivo. Drug Metab Dispos 40(4): 717-725.

Zanger UM and Klein K (2013) Pharmacogenetics of cytochrome P450 2B6 (CYP2B6): advances on polymorphisms, mechanisms, and clinical relevance. Front Genet 4: 24.

Zanger UM, Klein K, Saussele T, Blievernicht J, Hofmann MH and Schwab M (2007) Polymorphic CYP2B6: molecular mechanisms and emerging clinical significance. Pharmacogenomics 8(7): 743-759.

Zanger UM and Schwab M (2013) Cytochrome P450 enzymes in drug metabolism: regulation of gene expression, enzyme activities, and impact of genetic variation. Pharmacol Ther 138(1): 103-141.

Zanger UM, Turpeinen M, Klein K and Schwab M (2008) Functional pharmacogenetics/genomics of human cytochromes P450 involved in drug biotransformation. Anal Bioanal Chem 392(6): 1093-1108.

Zhang H, Sridar C, Kenaan C, Amunugama H, Ballou DP and Hollenberg PF (2011) Polymorphic variants of cytochrome P450 2B6 (CYP2B6.4-CYP2B6.9) exhibit altered rates of metabolism for bupropion and efavirenz: a charge-reversal mutation in the K139E variant (CYP2B6.8) impairs formation of a functional cytochrome p450-reductase complex. J Pharmacol Exp Ther 338(3): 803-809.

Zukunft J, Lang T, Richter T, Hirsch-Ernst KI, Nussler AK, Klein K, Schwab M, Eichelbaum M and Zanger UM (2005) A natural CYP2B6 TATA box polymorphism (-82T--> C) leading to enhanced transcription and relocation of the transcriptional start site. Mol Pharmacol 67(5): 1772-1782.

CURRICULUM VITAE FOR Ph.D. CANDIDATES The Johns Hopkins University School of Medicine

Philip Milton Cox November 21, 2016

Educational History: Ph.D. expected 2016 Program in Pharmacology Johns Hopkins School of Medicine Mentor: Namandjé N. Bumpus, Ph.D. B.S. 2010 Biology-Chemistry Southern Nazarene University

Other Professional Experience Preparing Future Faculty 2015-2016 Teaching Academy Certificate JHMI Research Technician 2010-2012 Lab of Darise Farris, Ph.D. Oklahoma Medical Research Foundation Summer Student 2009 Lab of Kenneth Kaye, Ph.D. Harvard Medical School Summer Student 2008 Lab of John Iandolo, Ph.D. University of Oklahoma Health Sciences Center Summer Student 2007 Lab of Swapan Nath, Ph.D. Oklahoma Medical Research Foundation

External Funding: National Science Foundation Graduate Research Fellowship, Fellowship number DGE-1232825, September 2013-December 2016, Stipend support: $30,000 - $35,000 per year, Institutional support: $12,000 per year.

Awards: Fall 2014 Best Poster Award JHMI Pharmacology Departmental Retreat

Peer-Reviewed Publications: Cox PM, Bumpus NN. (2016) Single Heteroatom Substitutions in the Efavirenz Oxazinone Ring Impact Metabolism by CYP2B6. ChemMedChem. doi:10.1002/cmdc.201600519 Cox PM, Bumpus NN. (2014) Structure-Activity Studies Reveal the Oxazinone Ring Is a Determinant of Cytochrome P450 2B6 Activity Toward Efavirenz. ACS Med Chem Lett. 10:1156-1161. PMCID: PMC4191608 Dumas EK, Nguyen ML, Cox PM, Rodgers H, Peterson JL, James JA, Farris AD. (2013) Stochastic humoral immunity to Bacillus anthracis protective antigen: identification of anti-peptide IgG correlating with seroconversion to Lethal Toxin neutralization. Vaccine. 14:1856-63. PMCID: PMC3614092 Garman L, Dumas EK, Kurella S, Hunt JJ, Crowe SR, Nguyen ML, Cox PM, James JA, Farris AD. (2012) MHC class II and non-MHC class II genes differentially influence humoral immunity to Bacillus anthracis lethal factor and protective antigen. Toxins (Basel). 12:1451-67. PMCID: PMC3528256 Dumas EK*, Cox PM*, Fullenwider CO, Nguyen M, Centola M, Frank MB, Dozmorov I, James JA, Farris AD. (2011) Anthrax lethal toxin-induced gene expression changes in mouse lung. Toxins (Basel). 9:1111-30. (*Indicates co-first authors) PMCID: PMC3202878

Poster Presentations: Cox PM, Bumpus NN. (2016) Single Atom Substitutions In the Efavirenz Oxazinone Ring Have Marked Impacts on Kinetics of Metabolism by CYP2B6. Conference on Microsomes and Drug Oxidation, Davis, CA, October 3, 2016.

Cox PM, Bumpus NN. (2015) Use of Efavirenz Analogues to Probe CYP2B6 Substrate Specificity. JHMI Pharmacology Department Retreat, Baltimore, MD, October 17, 2015. Cox PM, Bumpus NN. (2014) Use of Efavirenz Analogues to Probe CYP2B6 Substrate Specificity. JHMI Pharmacology Department Retreat, Baltimore, MD, September 20, 2014. Cox PM, Bumpus NN. (2014) Use of Efavirenz Analogues to Probe CYP2B6 Substrate Specificity. Conference on Microsomes and Drug Oxidation, Stuttgart, Germany, May 20, 2014. Cox PM, Bumpus NN. (2013) Use of Efavirenz Analogues to Probe CYP2B6 Substrate Specificity. International Society for the Study of Xenobiotics, Toronto, Canada, September 30, 2013. Service and Leadership: Spring 2016 Teaching Intern for Biochemistry Lab, Judy Levine, teaching advisor, Goucher College 2014-2016 Faith and Film Ministry Leader, Mount Vernon Place United Methodist Church 2012-2016 Lay Leader, Mount Vernon Place United Methodist Church 2012-2015 Habitat for Humanity of the Chesapeake Volunteer 2012-2013 Incentive Mentoring Program Student Mentor