Chemical and Structural Modulations of Cytochrome P450 Interactions with Redox Partners
by
Katherine A. Gentry
A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Biophysics) in the University of Michigan 2019
Doctoral Committee:
Professor Ayyalusamy Ramamoorthy, Chair Professor Stephen Ragsdale Professor Janet Smith Assistant Professor Kevin Wood
Katherine A. Gentry
ORCID: 0000-0002-8567-0622
© Katherine A. Gentry 2019
Acknowledgements
I must first and foremost thank my advisor, Dr. Ayyalusamy Ramamoorthy for his support and encouragement over the last five years. His mentorship has shaped me into an independent researcher and has given me the environment and freedom to explore projects on my own and grow as a scientist.
Besides my advisor, I would like to thank the rest of my committee members: Dr. Janet
Smith, Dr. Stephen Ragsdale, and Dr. Kevin Wood. All three have provided valuable guidance and support in my time at Michigan. Thank you very much for your questions, insights, and suggestions for my thesis work. I appreciate the support and feedback you gave which helped to shape my work into this document.
I would also like to thank our longtime collaborator, Dr. Lucy Waskell, for her valuable and helpful insights in the field of cytochrome P450s. My work would not be complete without her input and vast knowledge of all things cytochrome. Her help along with Dr. Sangchoul Im’s guidance helped me to learn the literature and biochemical work of the field.
My most sincere thanks goes to the Cytochrome Team and my mentors in the lab especially
Dr. Meng Zhang, Dr. Elke Prade, and Dr. Kyle Korshavn. Thank you for your almost limitless patience with everyday questions from sending difficult emails to how to wrangle Adobe
Illustrator. Thank you so much for your help in teaching me so much about NMR, protein biochemistry, science writing, and research and experimental design.
ii
A great deal of thanks also to all the Rams lab members: Sarah, Chris, Thiru, Carlo,
Mukesh, Julia, Rongchun, Amit, Sam, Kian, Josh, Kazu, Kamal, Niru, Bikash, Tomo, Nate,
Giacomo, Jian, and Vojc. Thank you for providing a fun and easygoing lab environment, thank you for all the edits, discussions, morning coffees, and happy hours. Having such an interactive and enjoyable lab environment really helped my motivation to come to work and do cutting-edge research.
Special thanks go to my family for supporting me, especially in these five years of graduate school. I appreciate every phone call celebrating with me when experiments went well or cheering me on when things didn’t go so well, and for the abundance of care packages. Thank you, Dad, for always being interested in what I am working on and being involved in my research. Thank you,
Mom, for the phone calls that you picked up even when you knew I was just walking somewhere and wanting to chat. Thanks to Libby for being my cheerleader, frequent email editor, and font of emotional support. Thanks to Becky for keeping me updated with adorable photos of my nieces,
Margaret and Gwen. I would not have been successful without your love and support!
iii
Table of Contents
Acknowledgements ii
List of Tables x
List of Figures xi
Abstract xvi
Chapter 1 Introduction 1
1.1 Cytochrome P450 1
1.1.1 Physiological Role of Cytochrome P450 1
1.1.2 Structure of Cytochrome P450s 4
1.1.3 Membrane Activity 6
1.2 Cytochrome P450 Reductase 7
1.2.1 Physiological role of Cytochrome P450 Reductase 7
1.2.2 Structure and Function of Cytochrome P450 Reductase 7
1.2.3 Interaction between CPR and cytP450 10
1.3 Cytochrome b5 11
1.3.1 Structure and Function of Cytochrome b5 11
1.3.2. Interaction between cytb5 and cytP450 12
1.4 Membrane Effect on Cytochrome P450 Metabolon 15
iv
1.4.1 Membrane and Membrane-Anchor Domains Modulate CytP450 15
Metabolism and its Catalytic Efficiency
1.4.2 Access Path Channels, Substrate Availability and Membrane Partitioning 19
1.4.3 Current Challenges and Future Directions 21
1.5 Goal of this research 22
1.6 References 23
Chapter 2 Kinetic and Structural Characterization of the Effects of Membrane on the 38
Complex of Cytochrome b5 and Cytochrome c
2.1 Summary 38
2.2 Introduction 39
2.3 Materials and Methods 42
2.3.1 Materials and Reagents 42
2.3.2 Preparation of cytochrome b5 42
2.3.3 Preparation of cytochrome c 43
2.3.4 Preparation of bicelles 43
2.3.5 Preparation of nanodiscs 43
2.3.6 NMR experiments and data analysis 43
2.3.7 Stopped Flow Kinetics 44
2.3.8 Data analysis and kinetic modeling 44
2.3.9 Kinetic Modeling and Simulations 45
2.3.10 Titration of cytb5 by Dithionite under Anaerobic Conditions 45
2.3.11 Calculation of a membrane-bound cytb5-cyt c complex using 45
NMR data
v
2.4 Results 46
2.4.1 Incorporation of the cytb5-cyt c complex into membrane mimetics 46
2.4.2 Formation of productive electron transfer complex between cytb5 48
and cyt c
2.4.3 NMR experiments probing the interaction between cyt c and 15N-cytb5 51
2.4.4 NMR experiments probing the interaction between 15N-cyt c and cytb5 57
2.4.5 Structural Model of the membrane-bound cytb5-cyt c complex 61
2.5 Discussion 62
2.6 Conclusion 66
2.7 References 67
Chapter 3 A Minimal Functional Complex of Cytochrome P450 and FBD of Cytochrome 73
P450 Reductase in Nanodiscs
3.1 Summary 73
3.2 Introduction 74
3.3 Results 76
3.4 Discussion 87
3.5 Materials and Methods 88
3.5.1 Materials 88
3.5.2 Expression and purification of proteins 89
3.5.3 Reconstitution of full-length proteins in nanodiscs 89
3.5.4 SAXS measurements 90
3.5.5 NMR experiments 91
3.5.6 Structural model calculation 92
vi
3.5.7 CYP450-FBD complex structure calculation 92
3.5.8 Stopped-flow measurement of electron transfer from hydroquinone 93
FMN to oxyferrous fl-CYP2B4 reconstituted in 4F-peptide based lipid nanodiscs
3.5.9 Evolutionary conservation analysis 94
3.6 References 94
Chapter 4 Substrate Mediated Redox Partner Selectivity of Cytochrome P450 101
4.1 Summary 101
4.2 Introduction 101
4.3 Results 103
4.3.1 Substrate effect on interaction in between cytb5 and cytP450 103
4.3.2 Effects of Substrates on FBD and cytP450 105
4.3.3 FBD is unable to disrupt the complex of cytb5-cytP450 105
4.3.4 Cytb5 is capable of dislodging FBD from cytP450 107
4.4 Discussion of substrate mediated effects 109
4.5 Conclusions 110
4.6 Materials and Methods 111
4.6.1 Materials and Reagents 111
4.6.2 Expression and purification of the soluble FMN binding domain 111
of rat CPR
4.6.3 Solution NMR experiments 112
4.6.4 Substrate modulation on the interaction between cytb5 and cytP450 112
4.6.5 Substrate modulation on competitive binding between cytb5 and trFBD 112
in the trFBD-cytP450-cytb5 tertiary protein system
vii
4.7 References 112
Chapter 5 Probing dynamic structural protein-protein and protein-substrate interactions 116
in ternary cytochrome P450, cytochrome b5, and cytochrome P450 reductase
5.1 Summary 116
5.2 Introduction 116
5.3 Results 119
5.3.1 Incorporation of three membrane proteins into a lipid nanodisc 119
5.3.2 15N-cytb5 monitored ternary formation 121
5.3.3 15N-flFBD monitored ternary formation 124
5.4 Discussion 127
5.5 Materials and Methods 130
5.5.1 Materials and Reagents 130
5.5.2 Expression and Purification of full-length cytochrome b5 131
5.5.3 Expression and Purification of full-length FBD 131
5.5.4 Expression and Purification of full-length cytP450 2B4 132
5.5.5. Preparation of nanodiscs 132
5.5.6 Reconstitution of full-length proteins in nanodiscs 132
5.5.7 NMR experiments 133
5.6 References 133
Chapter 6 Conclusions and Future Directions 136
6.1 Conclusions 136
6.2 Future Directions 139
viii
6.2.1 Mass spectrometry assisted capturing of cytP450 affinity for 139
membrane/ligands
6.2.2 Determination of Various Rates Associated with the Ternary Complex 140
6.2.3 Labeling of cytochrome P450 141
6.3 Future Outlook 142
6.4 References 144
Appendix A. Supporting Information for Chapter 5 146
ix
List of Tables
Table 2.1 Kinetic micro-rates obtained from numerical fitting of the stopped-flow 51
time-time dependent traces using the kinetic scheme depicted
Table 2.2 Energy statistics for lowest energy cluster of the complex between cytb5 62
and cyt c generated from HADDOCK
Table 3.1 List of restraints used in HADDOCK simulations for the full-length complex 86
in nanodiscs
Table 3.2 Energy statistics of the lowest energy cluster of the FBD-CYP2B4 complex 87
obtained using HADDOCK server
x
List of Figures
Figure 1.1 The cytochrome P450 catalytic cycle 3
Figure 1.2 Structural model of cytP450 in a lipid bilayer 5
Figure 1.3 Structural model of CPR in a lipid bilayer 9
Figure 1.4 Models of Cytochrome P450 interacting with redox partners in a lipid bilayer 10
Figure 1.5 Structural model of cytochrome b5 in a lipid bilayer 12
Figure 1.6 High-resolution structure of the membrane-bound cytochrome P450- 14
cytochrome b5 complex
Figure 2.1 Sequence alignment of cytochrome c and cytochrome P450 reveals amino 41
acid sequence similarity
Figure 2.2 Reconstitution of cytb5 – cyt c in lipid nanodiscs 47
Figure 2.3 Cytochrome c does not interact with the 4F-DMPC nanodiscs 48
Figure 2.4 UV absorption profiles of cytb5 in lipid-free solution from oxidized 49
to reduced
Figure 2.5 Spectral deconvolution and kinetic modeling reveal membrane environment 50
dependent changes
Figure 2.6 Results from SVD analysis applied to time-dependent spectra of electron 51
transfer reaction in cytb5-cyt c complex reconstituted in bicelles
Figure 2.7 2D 1H-15N HSQC-TROSY spectra of 15N-labeled cytb5 revealing the 53
interaction between cytb5 and cyt c
xi
Figure 2.8 Differential line broadening for cytb5 in buffer (no membrane), bicelles, or 54
nanodiscs in the presence of 1 molar equivalent of cyt c.
Figure 2.9 Cyt c interaction induced chemical shift perturbations of cytb5 55
Figure 2.10 Differential line broadening reveals binding sites on cytb5 56
Figure 2.11 Implicated binding sites on cytb5 mapped with differential line broadening 58
data
Figure 2.12 2D 1H-15N HSQC-TROSY spectra of 15N-labeled cyt c revealing interaction 59
between cyt c and cytb5
Figure 2.13 HADDOCK-generated structures reveal complex between cytb5 and cyt c. 60
Figure 3.1 Expression of full-length FBD with transmembrane domain 75
Figure 3.2 Reconstituted fl-FBD in nanodiscs are stable and well-folded. 76
Figure 3.3 Assignment of fl-FBD residues. 78
Figure 3.4 CS-ROSETTA calculated structure of the soluble domain of fl-FBD 79
Figure 3.5 Reconstitution of the fl-FBD and fl-CYP2B4 redox complex in peptide-based 81
nanodiscs
Figure 3.6 SAXS measurements and analysis of the fl-FBD fl-CYP2B4 complex in 82
peptide-based nanodiscs
Figure 3.7 Probing hot spots for redox complex formation 84
Figure 3.8 Stopped-flow data for electron transfer from fl-FBD to fl-CYP2B4 85
Figure 3.9 Representative NMR experiments of complex formation between cytP450 86
2B4 and fl-FBD
Figure 3.10 Conservational analysis of fl-FBD and fl-CYP2B4 88
Figure 4.1 Substrate effect on cytb5-P450 interaction 103
xii
Figure 4.2 Substrate effect on tr-FBD – cytP450 interaction 105
Figure 4.3 Cytb5-P450 complex is unperturbed by interaction with CPR and its variants 106
Figure 4.4 Cytb5 disrupts complex between FBD and cytP450 in a substrate dependent 108
manner
Figure 4.5 Schematic of the ternary interplay 109
Figure 5.1 Incorporation of three membrane proteins into a lipid nanodisc 120
Figure 5.2 15N-cytb5 monitored ternary formation 122
Figure 5.3 A zoomed-in section of cytb5 resonances reveal changes in serine 69 123
intensity as the ternary complex forms
Figure 5.4 15N-flFBD monitored ternary formation 126
Figure 5.5 Cytb5 disrupts the complex of flFBD-cytP450 in nanodiscs 128
Figure 5.6 Schematic of cytb5, cytP450, and CPR in a lipid nanodisc 130
Figure A.1 15N-slices of residue Serine 69 reveal peak broadening and 146
intensity decreases
Figure A.2 15N-slices of residue Glycine 89 reveal peak broadening and 147
intensity decreases
Figure A.3 2D 15N/1H TROSY HSQC spectra of 15N-labeled cytb5 in the presence 148
and absence of cytP450
Figure A.4 2D 15N/1H TROSY HSQC spectra of the complex of 15N-labeled 149
cytb5 and cytP450 with 4-CPI and flFBD
Figure A.5 2D 15N/1H TROSY HSQC spectra of the complex of 15N-labeled 150
cytb5 and cytP450 with BFZ and flFBD
xiii
Figure A.6 2D 15N/1H TROSY HSQC spectra of the complex of 15N-labeled 151
cytb5 and cytP450 with BHT and flFBD
Figure A.7 2D 15N/1H TROSY HSQC spectra of the complex of 15N-labeled 152
cytb5 and cytP450 with 1-CPI and flFBD
Figure A.8 2D 15N/1H TROSY HSQC spectra of the complex of 15N-labeled 153
cytb5 and cytP450 with BZ and flFBD
Figure A.9 2D 15N/1H TROSY HSQC spectra of the complex of 15N-labeled 154
cytb5 and cytP450 with flFBD
Figure A.10 2D 15N/1H TROSY HSQC NMR spectrum of (red) 15N-labeled flFBD 155
in 4F-DMPC Nanodiscs.
Figure A.11 2D 15N/1H TROSY HSQC NMR spectrum of (orange) 15N-labeled flFBD 155
and cytP450 in 4F-DMPC Nanodiscs.
Figure A.12 2D 15N/1H TROSY HSQC spectra of the complex of 15N-labeled flFBD 156
and cytP450 with 4-CPI and cytb5
Figure A.13 2D 15N/1H TROSY HSQC spectra of the complex of 15N-labeled flFBD 157
and cytP450 with BFZ and cytb5
Figure A.14 2D 15N/1H TROSY HSQC spectra of the complex of 15N-labeled flFBD 158
and cytP450 with BHT and cytb5
Figure A.15 2D 15N/1H TROSY HSQC spectra of the complex of 15N-labeled flFBD 159
and cytP450 with 1-CPI and cytb5
Figure A.16 2D 15N/1H TROSY HSQC spectra of the complex of 15N-labeled flFBD 160
and cytP450 with BZ and cytb5
xiv
Figure A.17 2D 15N/1H TROSY HSQC spectra of the complex of 15N-labeled flFBD 161
and cytP450 with cytb5
Figure A.18 Loop Regions of 15N-flFBD reveal binding and dissociating of flFBD with 162
cytP450
xv
Abstract
Cytochrome P450s (cytP450s) are a ubiquitous superfamily of enzymes that are responsible for the metabolism of many substrates, both endogenous compounds like hormones and exogenous compounds including over 70% of drugs on the pharmaceutical market. CytP450s have also been implicated in various cancers and heart disease. As a result, they are clearly important health targets for the design of pharmaceuticals. It is of great importance to study the mechanism of how these enzymes function. In order to complete one turn of its catalytic cycle, cytP450 requires two electrons, which are donated by its two redox partners, cytochrome P450 reductase (CPR) and cytochrome b5 (cytb5). Understanding the interplay of these three proteins is of high interest to capture how substrates are metabolized and to predict potential drug-drug interactions. Previous studies have only scratched the surface of how this complex works together.
To fully investigate this complex interplay, we need to utilize full-length proteins, membrane mimetics, and substrates. In my thesis, NMR spectroscopy and other biophysical techniques are used to probe the dynamic structural protein-protein interactions with and without the lipid membrane.
For NMR spectroscopy, as is the case with other structural biology techniques, homogenous sample preparation is of the utmost importance. The optimization and preparation of stable and consistent samples are crucial to achieve high-quality data. CytP450s and their redox partners are difficult proteins to work with due to their membrane anchor which lends to their propensity to aggregate in solution. Through the creation of a more physiological environment by
xvi functionally reconstituting these proteins in lipid nanodiscs, we were able to overcome many challenges posed by these proteins.
The initial forays into studying protein-protein complexes in membrane environments of cytb5 and cytochrome c (cyt c) revealed that the membrane influences both structure and function.
In the presence of a lipid membrane, the electron transfer complex is dynamic whereas without lipids the complex is static, albeit with a faster electron transfer rate. The binding epitope is also different in the presence of membrane, with the lower cleft of cytb5 being implicated rather than the side. Having identified the peptide nanodiscs as an ideal membrane mimetic for their flexibility, stability and ease of detergent-free preparation, further work on two membrane anchored proteins was next. Utilizing the flavin binding domain (flFBD) as the minimal requirement for functionality of the larger CPR, we show reconstitution and functionality of the cytP450 and flFBD complex in lipid nanodiscs. With restraints obtained from NMR experiments we were able to predict a complex between flFBD and cytP450.
To further our understanding of the dynamic interplay between cytP450, flFBD, and cytb5, we carried out competition assays to unravel how cytP450 interacts with its redox partners. We show that with and without a membrane environment cytb5 can dislodge flFBD from binding to cytP450. In the absence of lipids, flFBD is unable to disrupt the complex of cytb5-cytP450.
However, when all three proteins are incorporated into a lipid nanodisc, flFBD gains the ability to partially dislodge cytb5. Adding substrates to the ternary complex illustrates how intricate this system is, with substrates, especially bulky, aromatic substrates favoring the cytb5-cytP450 complex. This thesis work illustrates the importance of studying membrane proteins, like the cytP450s and redox partners, in a membrane environment in order to fully capture electron transfer activity and structural protein-protein interactions.
xvii
Chapter 1: Introduction
Cytochrome P450s (cytP450s) are a ubiquitous superfamily of enzymes that are responsible for the metabolism of a variety of substrates from endogenous compounds like vitamins and fatty acids to exogenous compounds including over 75% of the drugs on the pharmaceutical market. In order to carry out its catalytic cycle, cytP450 requires two electrons which it can receive from its redox partners, cytochrome P450 reductase (CPR) and cytochrome b5 (cytb5). All three of these proteins are bound to the membrane by a single transmembrane domain α-helix. This thesis delves into the protein-protein interactions between these redox partners specifically at how membrane mimetics and substrates can impact how the proteins interact.
1.1 Cytochrome P450
1.1.1 Physiological Role of Cytochrome P450
The cytP450 monooxygenases are found in all living kingdoms from bacteria and fungi to plants and animals. [1] These heme-containing, membrane-bound proteins are responsible for the metabolism of many hydrophobic compounds. These hemoproteins have been the subject of decades of research that has expanded as more advanced equipment and techniques have been discovered. Upon the development of double beam UV-Vis spectrophotometers, the naming of these proteins was established as cytochrome P450s with “P” standing for pigment and “450” coming from the characteristic absorption peak arising at 450 nm when the protein is reduced and bound to carbon monoxide. [2] This optical characteristic of cytP450s is widely used in order to
1 determine protein concentration as well as to ascertain the quality of the protein sample, with active cytP450 having a Soret band at max 450 nm and inactive cytP450 having a Soret band at max 420 nm. The heme cofactor is coordinated by a native cysteine thiolate ligand that is evolutionarily conserved. When the Soret band shifts from 450 nm to 420 nm, this is a sign that this normal coordination state has been changed with theories of either the cysteine thiolate gets protonated and can no longer coordinate the heme or there is a switch from the cysteine to a histidine ligated heme. [3]
57 genes have been identified in humans as coding for various cytP450 enzymes. Of these, about 13 isoforms are responsible for the metabolism of more than 80% of clinically used drugs.
[4] Human cytP450s that typically mediate metabolic clearance of drugs are targeted to the smooth endoplasmic reticulum (ER) by an N-terminal transmembrane domain with the catalytic domain residing in the cytosol. [5,6] The major cytP450 isoforms that are involved in drug metabolism (1A2,
2A6, 2B6, 2C9, 2C19, 2E1, 3A4, 3A5) [7] are mainly located in the liver whereas other cytP450 isoforms that are in involved in steroid metabolism are located to mainly the adrenals and gonads with smaller amounts expressed in the brain, placenta, and heart. [8] This dissertation is focused on one specific ER-associated membrane-bound cytP450: cytP450 2B4. CytP450 2B4 is a 56 kDa membrane-bound protein of with 491 amino acids and this rabbit isoform shares 78% sequence identity with the human cytP450 2B6. [9]
Cytochrome P450s carry out the first step of metabolism for a variety of hydrophobic compounds. The most common reaction cytP450 catalyzes is the insertion of a hydroxyl group into a hydrophobic substrate, breaking a C-C or C-H bond. Two reducing equivalents are required for this catalytic cycle which originate from either NADPH or NADH, depending on which redox partner reduces cytP450. Either CPR donates both electrons, with NADPH as the cofactor, or cytb5
2 can give the second electron, coming from NADH originally. The restriction of cytb5 from donating the first electron comes from the high difference in electron potential between oxidized cytP450 and reduced cytb5 (-245 mV vs +20 mV) which is alleviated after cytP450 is in its one electron reduced state. [10] Details of the catalytic cycle can be found in Figure 1.1. By increasing the hydrophilicity of the substrates, cytP450 increases the speed of clearance from the body for some drugs while for others, the enzyme induction can active inert to active form drugs. Better understanding of how cytP450 metabolizes drugs, specifically multiple drugs or compounds at a time is crucial to understanding and predicting potentially dangerous drug-drug interactions. [11-13]
Figure 1. 1. The cytochrome P450 catalytic cycle. At resting state, the heme iron is bound to a water molecule. (1) A substrate, RH, enters the active site and displaces the water molecule from the heme but does not coordinate with the heme. After the addition of a substrate, the spin state of the iron shifts from low to high spin for most substrates. The addition of the substrate also lowers the
3 redox potential of cytP450 to make it a more positive value and thus capable of being reduced in the next step of the cycle. (2) CPR donates the first electron to cytP450, which leads to the spin state of iron going from ferric (Fe3+) to ferrous (Fe2+) iron. (3) The ferrous iron binds with a molecule of oxygen forming the oxyferrous cytP450. (4) CPR, or cytb5, will donate the second -2 electron to cytP450 which leads to the formation of the peroxo cytP450. The O2 complex reacts with surrounding protons to form the highly reactive oxyferryl intermediate, also known as Compound I (5). The Fe-ligated O atm (6) is transferred to the substrate forming a hydroxylated form of the substrate (7). The product is finally released (8), replaced by a molecule of water. Three - uncoupling pathways are shown as dashed lines, with the respective products: the autoxidation shunt (O2 ), the peroxide shunt [14] (H2O2), and the oxidase shunt (H2O). Figure modified from F1000 Prime Review
Beyond potential drug-drug interactions, cytP450s are also key to the development/treatment of several health conditions including breast cancer, prostate cancer and heart disease. [15-18]
1.1.2 Structure of cytochrome P450s
In the mid to late 1980s and early 1990s, the first structures of cytP450 were solved with
, x-ray crystallography. These first structures solved were all bacterial cytP450s – P450cam, P450BM3
[10, 19-21] and P450terp . In the early 2000s, after modifications including transmembrane (TM) domain removal, the first mammalian cytP450 was solved. [22] Structural alignment between the cytP450 families is high, for example, the cytP450 CYP2 subfamily share 73% structural identity amongst all solved structures which is higher (85%) when only the closed conformation structures are considered. [23] Within the family of cytP450s, there are many structural similarities that have been evolutionarily conserved. The structural components: 13 α-helices (A, B, B’, C-L) and 5 β- strands (β1-β5) are found in soluble, bacterial proteins as well as eukaryotic, membrane-bound enzymes. [8] Another structural characteristic of the hemoproteins is the heme coordinating cysteine, which acts as a thiolate ligand of the heme b that is buried deeply inside the globular domain of the protein. The proximal side of cytP450, where the two redox partners can bind, is highly conserved. The distal side of cytP450, where the active site and substrate binding occurs, is much more plastic and this area has much less sequence homology amongst different isoforms.
[10] Depending on the isoform of cytP450, there is a wide structural plasticity around the active site to accommodate substrate binding. For the more restricted monooxygenases such as the aromatase
4 and cytP450 17A1, the limited substrate pool leads to a smaller range of active site sizes. Whereas for the drug metabolizer isoforms, the plethora of substrates leads to a wide fluctuation in active site capacities and sizes. There are currently 21 crystal structures that have been solved of CytP450
2B4, with and without ligands or substrates. [9, 24-38] Depending on the substrate that cytP450 2B4 has been crystallized with, the size of active site varies greatly from a “wide open” conformation to a closed form when bound to the inhibitor 4-CPI as well as an intermediate state with the substrate bifonazole.
Figure 1.2: Structural model of CytP450 2B4 in a lipid bilayer. (PDB: 1SUO)
Additionally, of note is that part of the cytP450 structure is involved with binding to the
ER or mitochondrial membrane beyond the hydrophobic TM domain spanning the bilayer. Even
5 truncated cytP450 will associate with and bind to a lipid membrane. [39] Truncated isoforms still bound firmly to membranes even upon increasing the ionic strength of their solution, indicating that there are membrane bound structural features in the globular domain. The structural elements of cytP450 that have been implicated in binding to the membrane are the F’, G’, and A helices as well as other features close to these helices like the BC loop and the β1 strand. [40-42] While the
F/G-loop has been identified as the most interacting by insertion into the membrane, the B/C-loop and several β-strands also make extensive contacts with the membrane. The overall fold of the cytP450 does not change upon binding to a membrane as demonstrated by MD simulations and the RMSD values not changing dramatically. [43]
1.1.3 Membrane Activity
After the discovery that cytP450s interact with the membrane beyond just being anchored to the bilayer, efforts have been made to reconstitute cytP450’s catalytic activity in membrane mimetic systems, from simple binary [44] or ternary [45] lipid mixtures, to the more sophisticated nanodiscs, [46,47] which consist of a lipid bilayer patch surrounded by a protein, peptide, or polymer scaffold. More recently, cytP450 activity has been reconstituted in a biomimetic with a lipid composition that closely matches the ER’s composition to study the protein-lipid interaction at a single-molecule level. [48]
Different cytP450 isoforms have varying lipid preferences. For example, cytP450 1A1 and cytP450 2E1 show a preference for disordered domains whereas cytP450 1A2 and CPR prefer ordered domains while cytP450 2B4 is equally distributed into both domains. [49] The catalytic activity of the isoforms is also tuned to the lipid composition. When phosphatidylethanolamine
(PE) lipids are added to phosphatidylcholine (PC) membrane mimetics, the activity of cytP450
2B4 increases but cytP450 1A2 stays stable. The addition of anionic lipids to membrane mimetics
6 increases cytP450 1A2 activity by 2-3 fold. [50] Clearly, the membrane environment greatly influences cytP450s, a topic that is explored further in Chapter 1.4.
1.2 Cytochrome P450 Reductase
1.2.1 Physiological role of Cytochrome P450 Reductase
Cytochrome P450 reductase is an essential redox partner for cytP450’s metabolic activity as it is capable of donating both electrons to cytP450 during its catalytic cycle. Much activity can be reconstituted with just CPR and cytP450. CPR is also capable of donating electrons to heme oxygenase, cytochrome b5, and non-physiologically, cytochrome c. [54] Electrons flow from
NADPH to CPR to cytP450. CPR is preferentially reduced by NADPH, which it has a much higher affinity for over NADH. [51, 52] The flavin exists in two states, cycling in between the one and three electron reduced forms. The one-electron reduced form is a blue semiquinone that is incapable of donating an electron to cytP450. Instead, it is the FMN hydroquinone that donates an electron to cytP450. [53] The inter-flavin electron transfer rate has been experimentally observed to be around
30 – 50 s-1. [9, 28]
1.2.2 Structure and Function of Cytochrome P450 Reductase
CPR has four distinct domains that make up the roughly 80 kDa protein: a FAD binding domain where NADPH reduces the protein, a linker domain, an FMN binding domain that reduces cytP450, and a N-terminal transmembrane domain. To overcome the difficulties associated with working with an 80 kDa protein, frequently the protein is expressed and studied as a smaller domain – the FMN binding domain or FBD. The FBD is known to be the domain that interacts directly with cytP450 to both bind and transfer an electron. This truncated version of CPR is capable of binding to cytP450 as well as being reduced by the FAD domain and reducing the heme
7 in cytP450. Although, without the transmembrane hydrophobic anchor, while the soluble domain is capable of passing electrons to cytochrome c, it is not capable of transferring electrons to cytochrome P450. [54] Alone the FAD/NADPH binding domain is not efficient at reducing cyt c or cytP450 based on cytP450 1A1 oxidation of the substrate 7-ethoxyresorufin [55].
Crystallization of CPR has been successful, and several structures have been solved including a closed version of the protein and, after a modification in the hinge (ΔTGEE deletion), an open configuration of the protein. As electrons flow intramolecularly from the FAD domain to the FBD, the overall protein goes through a conformational change from an open state to a closed state. [56] Structurally, the FBD is similar to the bacterial flavodoxin and the C-terminal, FAD binding portion is structurally homologous to ferredoxin-NADP+ oxidoreductase.[53] A solved x- crystallography structure has also been solved of the truncated FBD which shows the same conformation as the FBD domain in the CPR. [57] The FBD consists of a typical α-β-α fold with five α-helices, 5 β-strands, and several loop regions. The FMN cofactor is bound to the protein
8 -1 [58, 59] tight with a Kd of 2*10 M of around and is surrounded by four loops.
8
Figure 1.3: Structure representation of full-length Cytochrome P450 Reductase (CPR) in a lipid bilayer using the X-ray crystal structure of the soluble domain (PDB: 1AMO)
9
1.2.3 Interaction between CPR and cytP450
As alluded to above, CPR is the obligate redox partner of cytP450. Many studies have been carried out to characterize the complex formation and binding of CPR or CPR variants including the truncated FBD to cytP450. However, what is currently lacking from our understanding of this protein-protein interaction is structural information from full-length proteins in a membrane environment.
Figure 1.4 Models of Cytochrome P450 interacting with redox partners in a lipid bilayer. (Left) Cytochrome P450 interacting with cytochrome b5. (Right) Cytochrome P450 interacting with Cytochrome P450 Reductase
Important residues on cytP450 for CPR binding have been identified through site-directed mutagenesis studies, stopped-flow spectrometry, and other methods. Both electrostatic interactions as well as hydrophobic interactions have been identified as contributing to complex formation in between cytP450 and FBD. Cationic localized charge on the proximal side of cytP450 and anionic residues surrounding the FMN of FBD initially drive the protein-protein interaction. [60, 61] Several cationic and neutral residues on cytP450 have been identified as being critical for binding to CPR
10 which are on the C or C’ helix: R122, R126, R133, F135, M137, K139, K433, R422, R443. [62]
There is a hydrophobic component which drives the interaction as well specifically several surface exposed hydrophobic residues of cytP450 2B4, V267 and L270. [63] CytP450 has been shown to have a greater binding affinity for CPR over cytb5 [61, 62]. These two proteins share overlapping, but unique binding sites on cytP450. Due to this fact that they possess similar binding areas on cytP450, the two proteins must be in some sort of competition as they cannot bind at the same time. Further work in this thesis will be devoted to unraveling how these two proteins are competing with each other and examining how substrates can change the dynamic interplay.
1.3 Cytochrome b5
1.3.1 Structure and Function of Cytochrome b5
Cytochrome b5 (cytb5) is a smaller, hemoprotein (~15 kDa) that contains a C-terminal transmembrane domain. Cytb5 acts as a physiological redox partner to cytP450, although it is only capable of donating the second electron cytP450 needs due to its high redox potential (~25 mV).
[10] Cytb5 is a highly conserved protein especially throughout mammalian species with over 80% sequence identity and conservative substitutions. [64, 65] Rabbit cytb5, the mammalian isoform used throughout this thesis, is roughly 15 kDa and consists of 134 amino acids. This hemoprotein contains three domains: a heme containing globular domain, a 15-amino acid flexible linker crucial to activity, and a C-terminal transmembrane domain. [66] This globular domain consists of 5 α- helices, five β-strands, and one 310 helix and has a heme b that is coordinated by two histidines:
His 44 and His68. Cytb5 has an overall negative net charge, consisting of over 20% surface residues of glutamates and aspartates, especially around the solvent heme edge where it has been reported to bind to its redox partners. [66] Of the 134 amino acids that make up mammalian cytb5,
23 of them are either glutamate or aspartate. [65] The 15 amino acid random coil linker is crucial
11 for complex formation. [61] Shortening of the linker abolishes binding to cytP450 when the length of the linker is below 7 residues whereas an increase in length has minimal effects on binding.
Akin to both cytP450 and CPR, the only crystal structures of cytb5 that have been solved have been solved without the transmembrane domain. Published work from our lab has generated a model of full-length cytb5 interacting with full-length cytP450 2B4 via a combination of solution and solid-state NMR. [66] Studying full-length cytb5 is critical due to reports that truncated cytb5 is incapable of donating electrons to mammalian membrane-bound cytP450s but can transfer electrons to soluble proteins including bacterial cytP450s. [67-70]
Figure 1.4: Structural model of full-length cytochrome b5 in a lipid bilayer (PDB: 2M33). Previous NMR studies reported the structures of the soluble and transmembrane domains using solution and solid-state NMR experiments, respectively.
1.3.2 Interaction between cytb5 and cytP450
Cytb5 has unique properties when influencing cytP450 metabolism. Based on the cytP450 isoform and substrate involved, cytb5 can increase, decrease, inhibit, or do nothing at all to the
12 metabolism of a compound. Depending on the concentration of cytb5 in comparison to cytP450 and CPR, cytb5 has been reported to be able to stimulate activity at low concentrations as well as inhibit activity at higher concentrations. One explanation to this phenomenon is that cytb5 and
FBD share overlapping, but unique binding sites on cytP450. At high concentrations of cytb5, CPR is hindered from binding to cytP450 which leads to the first electron delivery slowing or even ceasing corresponding with the decreased activity. [61]
Many studies have been done to investigate the binding of cytochrome P450 and cytb5.
Initial work by the Waskell lab studying cytbP450 2B4 and cytb5 identified seven amino acids of a potential 25 amino acids that had been mutated as being key to the binding interface on cytP450
2B4. These residues were all located on the C-helix on the proximal side of the protein (R122,
R126, R133, F135, M137, K139) with K433 on the β-bulge above the axial cysteine. [61, 62] On the cytb5 side of the protein-protein interaction, 13 different single site mutations were carried out of mostly anionic residues. Of the 13 mutants, D65 and V66 were identified as key residues in the interface due to their mutant’s significant lower affinity for cytP450 as well as decreased activity.
[66] Several other mutants were identified but only decreased binding affinity and did not decrease cytb5’s ability to enhance metabolism: E42, E43, E49, Q54, and N62. Based off of NMR experiments and this site directed mutagenesis work, a full length cytb5 full length cytP450 2B4 complex was predicted. [66]
Upon complex formation between cytP450 and cytb5, cytP450’s heme group shifts from low to high spin. [67, 71-73] Full conversion of low to high spin occurs only when the full-length proteins have been reconstituted into lipid bilayers as shown in lipid nanodiscs. [72] Substrates have long been shown to impact the strength of complex formation in between cytb5 and cytP450. [67,
73] Interestingly, there have been some studies that have shown that cytb5 can influence cytP450
13 activity without cytb5’s ability to donate electrons. Manganese substituted protoporphyrin cytb5
(Mn-cytb5) has the same structure as wild type cytb5 but remains essentially fully oxidized in the presence of NADPH and thus is incapable of electron transfer to cytP450. [74] Mn-cytb5 is still capable of binding to cytP450 and inhibiting or altering metabolism of various substrates. [74] There still are many questions left unsolved about how cytb5 interacts with cytP450 to influence activity in such a variety of outcomes.
Figure 1.6: High-resolution structure of the membrane-bound cytochrome P450-cytochrome b5 complex. (A) Three-dimensional model of the membrane-bound rabbit cytochrome P450 2B4 – cytb5 complex reproduced from Ahuja et al. [59] determined via NMR and site-directed mutagenesis restraints. (B) Electron transfer pathway predicted by HARLEM63 in the complex structure.
14
1.4 Membrane Effect on Cytochrome P450 Metabolon1
1.4.1. Membrane and Membrane-Anchor Domains Modulate P450 Metabolism and its Catalytic
Efficiency.
CytP450-mediated oxidation of drugs and xenobiotics in the endoplasmic reticulum is made possible by the sequential donation of two electrons by CPR and cytb5. [75] The electron transfer is believed to occur through the dynamic interactions between these proteins. For the electron to be shuttled, it needs to overcome a redox potential gradient. As depicted in Figure 1.1, cytb5 is unable to donate the first electron, since it cannot overcome the redox potential barrier.
The composition of phospholipids can modulate the midpoint potentials for both cytP450s and
CPR, as demonstrated by Das and Sligar using nanodiscs. [76] When CPR was incorporated in nanodiscs, the redox potentials of both the FAD and FMN domains were shifted to more positive values, compared to CPR lacking the TM domain. Also, the presence of anionic phospholipids made the redox potential suitable for electron transfer from CPR to cytP450, a result that confirmed kinetics observations made by other groups. [77] The catalytic activity and membrane insertion of cytP450 3A4, which metabolizes >50% of commercial drugs, [78] increased as a function of anionic phospholipid concentration. [77] It is recognized that membrane composition and polarity can
[77, 79] modify both Vmax and KM, which are respectively a measure of catalytic rate and binding affinity under the steady-state assumption of Michaelis-Menten equation. Phospholipid composition has also been related to protein folding and stability, as reported by Jang et al. for full-length and TM-truncated cytP450 1B1, as well as cytP450 3A4. [79, 80] Regarding the putative role of cytb5 as electron donor and allosteric effector for cytP450- mediated drug clearance, it is well known that it is isoform dependent. Allosteric modulation has been observed for CYP3A4
1This section is modified from the published paper: Barnaba, C., Gentry K., Sumangala, N., Ramamoorthy, A. The catalytic function of cytochrome P450 is entwined with its membrane-bound nature. F1000Research 2017, 6(F1000 Faculty Rev):662.
15 hydroxylation of testosterone and nifedipine, [81] - but not for CYP1A1 and CYP2D6 1’-
[82] hydroxylation of bufuralol. Deletion of amino acids of the TM α-helix of rabbit b5 decreased the binding affinity for CYP2B4 by several folds, as well as the catalytic turnover. [83]
The efficiency of cytochrome P450 catalysis is measured in terms of “coupling”, meaning the amounts of transferred electrons that are committed to the monooxygenase activity vs. the extent of the “uncoupled” (or unproductive) pathways. Three unproductive pathways exist (Figure
1.1): 1) release of superoxide from the ferrous dioxygen, 2) release of hydrogen peroxide, and 3) the total four electron reduction of the dioxygen molecule to net produce two molecules of water.
[75] Recent studies have postulated that both TM domains and the nature of the membrane may have a role in the coupling efficiency. McDougle et al. [84] found a correlation between the coupling efficiencies of CYP2J2 with the length of the TM domain of CPR – the more extensive the deletion of the TM helix, the lesser the coupling efficiencies. Similar results have been reported for
CYP2C19 coupling with several substrates: full-length vs. truncated CPR, as well as the presence and/or absence of lipids were related to the overall catalytic efficiency. [85] Grinkova et al. [86] observed an increase of coupling when CYP3A4 was incorporated in nanodiscs containing higher amounts of anionic phospholipids, which the authors attributed to the changes in the redox potential of CYP3A4 and reductase. On the P450 side, it is believed that the TM domain of CYPs serves as an “anchor” to the membrane, playing a pivotal role on protein orientation. In CYP1A2 and CYP2D6, the hydrophobicity of the TM domain has been shown to regulate the efficiency of interaction of CYPs with CPR and/or phospholipids, and therefore their catalytic activities. [87-90]
Similar results have been reported for CYP1B1, with decreased catalytic activity observed when the TM domain is partially cleaved. [79]
16
Specific interactions between the P450 metabolon and the membrane are believed to play a major role in this catalytic modulation. Traditional X-ray crystallographic studies on P450-CPR and P450-b5 complexes provided snapshots of their quaternary organization and have been accompanied by specific amino acid mutations able to disrupt/enhance functionality. [61, 62, 91, 92]
Recently, solution NMR experiments have complemented those initial efforts, providing realistic insights of the complex interface, as well as ligand-induced modifications. [93] Scott’s group has explored the potential of this technique for studying protein complexes between the steroidogenic
[94, 95] CYP17A1 and both b5 and the FMN domain of CPR. However, complete solubility and subsequent crystallization of complexes are obtained through cleavage of the roughly 60-residue segment containing the hydrophobic domain. Membrane is not present in these structures which are devoid of any information about transmembrane organization of these proteins.
The Ramamoorthy lab has utilized both solution and solid-state NMR (ssNMR) techniques in studying P450 and related proteins [96] in a functional amphiphilic bilayer, relying in the capacity of both bicelles [97, 98] and nanodiscs [99] to incorporate and preserve the functionality of these proteins. A high-resolution structure of the membrane-bound cytochromes-P450-b5 complex is shown in Figure 1.6. [66] Solid-state NMR experiments on magnetically-aligned bicelles revealed the presence of transmembrane helices and provided insights into their topology in CYP2B4, CPR, and in b5; and further revealed the significant difference in the time scales of motion of residues in the soluble linker and TM domains of b5 that are crucial in the formation of productive protein-
[100, 101] protein complexes. Studying isotopically labeled b5 in bicelles by ssNMR has shown how the mobility of the TM domain of b5 is significantly reduced by the presence of CYP2B4, without altering its geometry and helical structure. [100] In addition, ssNMR experiments on aligned bicelles revealed the direct interactions between the TM domains of CYP2B4 and b5, in which the “leucine
17 zipper” in the TM domain of b5 plays a very important role. More importantly, NMR confirmed the electrostatic nature of the b5-CYP2B4 interactions, and that the presence of a substrate
[102] promotes specific interactions between the two proteins. This CYP2B4-b5 complex was more recently studied in nanodiscs, in a detergent-free environment that better simulates the native membrane, [99] obtaining high resolution structural interactions between the proteins. In addition, interactions between membrane-bound CYP2B4, b5 and CPR have also been measured with and without the presence of ligands using high-resolution NMR experiments. (Chapter 4) The FMN- binding domain (FBD) of CPR has also been studied in bicelles by solid-state NMR experiments.
[103] The Ramamoorthy group was also able to characterize for the first time the interplay among proteins in a tertiary FBD-P450-b5 complex by NMR spectroscopy in lipid bilayers. (Chapter Five)
Notably showing that substrate binding to P450 can enhance the complex of b5 disrupt the P450-
CPR complex, facilitating the association of b5, and that the extent of this equilibrium shift is highly dependent on the substrate. This result can potentially explain the observed difference in the catalytic modulation provided by b5.
In addition, the structural interaction of FBD with cytochrome c – a small hemoprotein essential for the electron transport chain in mitochondria – has been reported by using cytochrome c as an efficient model system to test the electron transfer process. [104] Solution NMR titration experiments showed the formation of a dynamic complex between the two proteins, on a fast exchange time scale. NMR restraints were implemented in molecular docking to generate structural models and map the binding interface on the FBD. The proposed structural model of the
FBD – cytochrome c complex suggested potential electron pathways that provide strong electronic coupling between the redox centers. [104] More high-resolution structure-based functional and
18 dynamical studies on the binary and ternary P450-redox complexes and their interactions with drugs are in progress.
1.4.2 Access Path Channels, Substrate Availability and Membrane Partitioning
As from the above discussion, the membrane and the membrane domains can play a substantial role in P450 catalysis, modulating both electron transfer between the redox couple(s), as well as the overall efficiency of the catalysis. The membrane is also a thick interface between the surroundings (i.e. cytosol) and CYPs. Indeed, it has been shown that the membrane can slow down the access of water, as well substrate, to the active site. [105] Also, lipophilic compounds that are poorly soluble are predominantly partitioning in the membrane, allowing for P450 recruiting of hydrophobic substrates directly from the lipid phase. [12, 76, 106] For example, Murtazina et al. [107] demonstrated that CYP27A1 affinities for 5α-cholestane-3α,7α,12α-triol and cholesterol were decreased in the presence of the negatively charged phosphatidylglycerol, which can potentially reflect either a different partitioning of these steroidal molecules or an easier access to substrate due to protein/membrane interactions. Differences in the binding affinities and spin equilibrium in soluble vs. membrane-anchored P450 have already been reported, [12, 80, 108-110] and their significance is relevant if considering how crucial are the affinity parameters for pharmacokinetic/pharmacodynamic models. [111] On this regard, Denisov et al. [12] have demonstrated the presence of an allosteric site at the membrane interface of CYP3A4, emphasizing how crucial is the presence of membrane for the evaluation of drug-drug interactions in pharmacological studies. In addition, recent studies from Atkin’s group have provided experimental insights into substrate modulation of CYP3A4 when in nanodiscs. [80, 110] The embedded portion of the protein can dynamically interact with the membrane, allowing the
19 opening of substrate channels and a water “aqueduct”. [80] These properties – along with substrate equilibrium binding – seem also to be modulated by membrane fluidity. [110]
It is not completely understood how lipids can perturb P450 spin equilibrium, [106] but the observed differences prove the intrinsic kinetic peculiarities of each system used to test chemicals.
Molecular dynamic (MD) simulations and H/D exchange studies on CYP3A4 have shown that the interaction with the membrane occurs through specific lipid-protein interactions, [42, 80] and is able to affect the opening/closing of the access tunnel, [106] confirming experimental evidence obtained two decades before. [112] Computational MD studies have also elucidated how phospholipids can induce an opening of membrane-facing tunnels in CYP1A2. [113] The speculated mechanism relies on the ability of the upper part of the TM helix to interact with a proline rich segment of the catalytic domain that along with the FG loop are immersed in the membrane. [113] This gave rise to several access channels from both solvent and membrane facing tunnels. Similar studies conducted on CYP2C9 revealed the opening of additional access tunnels in the presence of membrane. [114, 115] MD simulation performed by Fishelovitch et al. [116] have shown that the FMN binding domain of CPR regulates the water channel of CYP3A4 due to an overlapping of specific residues. When the FMN domain of CPR binds to CYP3A4, the water-channel fully opened up, thereby allowing a flow of water molecules into the active site. However, the absence of membrane in this computational study raises the question if this conformational flexibility also occurs in the presence of membrane.
Though experimental demonstrations are essential, the computational studies have provided useful information, as well as quantitative thermodynamic parameters, that may potentially explain several observations in P450 kinetics. Nevertheless, given the importance of substrate availability and disposition for drug-metabolizing enzymes, there is a lack of compelling
20 experimental evidence that can fulfill the requirement of detailed molecular descriptions. If structural X-ray crystallography has laid the groundwork for determining enzyme channels and quantifying active site volumes, [91] it is still unable to look beyond the polypeptide architecture.
On the other hand, recent progress in the use of solution and solid-state NMR techniques is opening the way to a broader inspection of membrane-associated phenomena. Particularly, the high- resolution dynamics of membrane-bound P450 in the presence and absence of ligands to be measured from NMR will be highly valuable in fully understanding the function of P450.
1.4.3 Current Challenges and Future Developments
Membrane proteins structure determination is in general an extremely challenging task due to the lack of stability of the protein outside its native environment. In cytochrome P450, the presence of interactions with the reducing counterparts, as well as the pivotal role of membrane in catalysis, add fascinating complexities to the ongoing efforts. Given the limitations of the X-ray structural characterization discussed above, NMR – both solution and solid-state approaches - is an excellent alternative for the structural and dynamics studies on cytochrome P450. As demonstrated for b5 and its interactions with CYP2B4, NMR offers well-tested and sophisticated tools, being unique in allowing to probe molecular dynamics over a wide range of time scales. Through NMR, motions from nanosecond to millisecond time scales can be probed. Nevertheless, the overall size of the membrane-protein complex is still the main drawback in terms of spectral resolution; the reduced molecular mobility of proteins embedded in a membrane environment also contributes to poor sensitivity and resolution for traditional solution NMR experiments. [96] For cytochrome P450, a further complication dwells in the paramagnetic nature of the iron in the heme prosthetic group, which quenches NMR signals in the proximity of the active site. In the coming years, NMR will be able to face up several of its intrinsic limitation, by the introduction of labeling strategies,
21
Paramagnetic Relaxation Enhancement (PRE) effects, Ultrafast Magic-Angle Spinning (ultrafast-
MAS), and sophisticated pulse sequences. Also, reconstitution of P450 and its redox partners can now be performed in more reliable and versatile model systems, such as nanodiscs that are used for solution NMR studies and macro-nanodiscs for solid-state NMR studies.
Many scientific questions regarding the P450 metabolon require answers, due to the innumerable biological and pharmacological implications. The ongoing structural studies would offer valid tools and provide intimate insights into biological interfaces. Nevertheless, the combined use of complementary biophysical techniques and biochemical approaches will create unique avenues to successfully overcome the challenges in fully understanding the enzymatic function of a variety of P450s. The acquired knowledge derived from these studies will set an experimental platform to investigate other electron-transfer protein complexes, such as the mitochondrial electron transport chain, [117] and the nitric oxide synthase complex, [118] among others, which share several biochemical similarities and complexities with the P450 metabolon.
1.5 Goal of this research
Throughout this thesis work, the driving force has been to study these important membrane-bound proteins in a lipid environment. We know that the membrane can play a large role in governing these protein-protein interactions far greater than providing an anchor point for these proteins to reside in. Starting with a model redox system of cytb5 and cytochrome c (Chapter
2), different membrane mimetics were optimized and utilized to study the effect of membrane on a protein-protein interaction with one soluble and one transmembrane domain containing protein.
After the optimization of 4F-lipid nanodiscs as a viable and sturdy membrane mimetic, NMR spectroscopy and other biophysical techniques were applied to study the interaction between a minimal redox complex of cytP450 and FBD (Chapter 3). Moving towards the ultimate goal of
22 studying the interplay between all three proteins, substrates were added in order to characterize the interactions between cytP450, FBD, and cytb5 (Chapter 4). In order to see the effect of membrane on the ternary complex interplay, all three proteins were sequentially incorporated into lipid nanodiscs and their interplay was studied through NMR spectroscopy (Chapter 5). This work will provide insights into the role the membrane and substrates play in dictating cytP450’s preference and interaction with its redox partners.
1.6 References
1. Norbert, D.W., Nelson, D.R. and Feyereisen, R. (1989a) Evolution of the cytochrome P450 genes, Xenobiotica, 19, 1149-1160.
2. Estabrook, R.W. A Passion for P450s (Remembrances of the Early History of Research on
Cytochrome P450) (2003) Drug Metab. Dispos. 31, 1461-1473.
3. Sun, Y. et al. (2013) Investigations of heme ligation and ligand switching in cytochromes
P450 and P420. Biochemistry 52, 5941-5951.
4. De Montellano, P. R. O. Cytochrome P450: structure, mechanism, and biochemistry.
(Springer, 2005).
5. Ruckpaul, K., and Rein, H. (1984) Cytochrome P450, Akademie-Verlag, Berlin.
6. Stier, A. (1976) Lipid structures and drug metabolizing enzymes, Biochemical Pharmacology,
25, 109-113.
7. Zanger, U. M., and Schwab, M. (2013) Cytochrome P450 enzymes in drug metabolism:
Regulation of gene expression, enzyme activities, and impact of genetic variation, Pharmacology
& Therapeutics, 138, 103-141
8. Yoshimoto, F.K., Auchus, R.J. The diverse chemistry of cytochrome P450 17A1 (P450c17,
CYP17A1) (2015) J. Steroid Biochem. Mol. Biol. 151, 52-65.
23
9. Scott, E. E. et al. Structure of mammalian cytochrome P450 2B4 complexed with 4-(4- chlorophenyl)imidazole at 1.9-A resolution: insight into the range of P450 conformations and the coordination of redox partner binding. J. Biol. Chem. 279, 27294–301 (2004).
10. Lewis, D. F. V. Guide to Cytochromes P450 Structure and Function. (Taylor & Francis,
2001).
11. Ogu, C. C., and Maxa, J. L. (2000) Drug interactions due to cytochrome P450 -
Pharmacology notes, BUMC Proceedings, 13, 421-423.
12. Denisov, I. G., Grinkova, Y. V., Baylon, J. L., Tajkhorshid, E., Sligar, S. G. (2015)
Mechanism of Drug-Drug Interactions Mediated by Human Cytochrome P450 CYP3A4
Monomer, Biochemistry, 54, 2227-39.
13. Denisov, I. G., Baylon, J. L., Grinkova, Y. V., Tajkhorshid, E., Sligar, S. G. (2018) Drug-
Drug Interactions between Atorvastatin and Dronedarone Mediated by Monomeric CYP3A4,
Biochemistry, 57, 805-816.
14. Barnaba, C., Gentry K., Sumangala, N., Ramamoorthy, A. The catalytic function of cytochrome P450 is entwined with its membrane-bound nature. F1000Research 2017, 6(F1000
Faculty Rev):662.
15. Nebert, D. W. & Russell, D. W. Clinical importance of the cytochromes P450. Lancet 360,
1155–62 (2002).
16. Fleming, I. Cytochrome P450-dependent eicosanoid production and crosstalk. Curr. Opin.
Lipidol. 22, 403–9 (2011).
17. Murray, G. I. et al. Tumor-specific Expression of Cytochrome P450 CYP1B1. Cancer Res.
57, 3026–3031 (1997).
24
18. El-Sherbeni, A. A. & El-Kadi, A. O. S. (2017) Microsomal cytochrome P450 as a target for drug discovery and repurposing, Drug Metabolism Reviews, 49, 1-17.
19. Poulos, T.L. (1986) The crystal structure of cytochrome P450cam in: Cytochrome P450 (P.R.
Ortiz de Montellano, ed.) Plenum, New York, Chapter 13, 505-523.
20. Ravichandran, K.G., Boddupalli, S.S., Hasemann, C.A., Peterson, J.A., and Deisenhofer, J.
(1993) Crystal structure of hemoprotein domain of P450BM3, a prototype for microsomal P450s,
Science, 261, 731-736.
21. Hasemann, C.A., Ravichandran, K.G., Peterson, J.A. and Deisenhofer, J. (1994) Crystal structure and refinement of cytochrome P450terp at 2.3 A resolution, Jounral of Molecular
Biology, 236, 1169-1185.
22. Williams, P. A., Cosme, J., Sridhar, V., Johnson, E. F., and McRee, D. E. (2000) Mammalian
Microsomal Cytochrome P450 Monooxygenase. Mol. Cell 5, 121-131.
23. Mestres, J. Structure conservation in cytochromes P450. Proteins 58, 596–609 (2005).
24. Scott, E.E., He, Y.A., Wester, M.R., White, M.A., Chin, C.C., Halpert, J.R., Johnson, E.F.,
Stout, C.D. (2003) An open conformation of mammalian cytochrome P40 2B4 at 1.6 A resolution. Proc. Natl. Acad. Sci. U.S.A., 100, 13196-13201.
25. Shah, M.B., Jang, H.H., Wilderman, P.R., Lee, D., Li, S., Zhang, Q., Stout, C.D., Halpert,
J.R. (2016) Effect of detergent binding on cytochrome P450 2B4 structure as analyzed by X-ray crystallography and deuterium-exchange mass spectrometry, Biophys. Chem. 216, 1-8.
26. Zhao, Y., White, M.A., Muralidhara, B.K., Sun, L., Halpert, J.R., Stout, C.D. (2006)
Structure of microsomal cytochrome P450 2B4 complexed with the antifungal drug bifonazole: insight into P450 conformational plasticity and membrane interaction, J. Biol. Chem., 281, 5973-
5981.
25
27. Shah, M.B., Wilderman, P.R., Pascual, J., Zhang, Q., Stout, C.D., Halpert, J.R. (2012)
Conformational Adaptation of Human Cytochrome P450 2B6 and Rabbit Cytochrome P450 2B4
Revealed upon Binding Multiple Amlodipine Molecules, Biochemistry, 51, 7225-7238.
28. Shah, M.B., Kufareva, I., Pascual, J., Zhang, Q., Stout, C.D., Halpert, J.R. (2013) A
Structural Snapshot of CYP2B4 in Complex with Paroxetine Provides Insights into Ligand
Binding and Clusters of Conformational States. J. Pharmacol. Exp. Ther., 346, 113-120.
29. Zhao, Y., Sun, L., Muralidhara, B.K., Kumar, S., White, M.A., Stout, C.D., Halpert, J.R.
(2007) Structural and thermodynamic consequences of 1-(4-chlorophenyl)imidazole binding to cytochrome P450 2B4. Biochemistry, 46, 11559-11567.
30. Wilderman, P.R., Shah, M.B., Liu, T., Li, S., Hsu, S., Roberts, A.G., Goodlett, D.R., Zhang,
Q., Woods, V.L., Stout, C.D., Halpert, J.R. (2010) Plasticity of Cytochrome P450 2B4 as
Investigated by Hydrogen-Deuterium Exchange Mass Spectrometry and X-ray Crystallography.
J. Biol. Chem., 285, 38602-38611.
31. Wilderman, P.R., Gay, S.C., Jang, H.H., Zhang, Q., Stout, C.D., Halpert, J.R. (2012)
Investigation by site-directed mutagenesis of the role of cytochrome P450 2B4 non-active-site residues in protein-ligand interactions based on crystal structures of the ligand-bound enzyme,
Febs J., 279, 1607-1620.
32. Zhang, H., Gay, S.C., Shah, M., Foroozesh, M., Liu, J., Osawa, Y., Zhang, Q., Stout, C.D.,
Halpert, J.R., Hollenberg, P.F. (2013) Potent Mechanism-Based Inactivation of Cytochrome
P450 2B4 by 9-Ethynylphenanthrene: Implications for Allosteric Modulation of Cytochrome
P450 Catalysis, Biochemistry, 52, 355-364.
33. Yang, Y., Zhang, H., Usharani, D., Bu, W., Im, S., Tarasev, M., Rwere, F., Pearl, N.M.,
Meagher, J., Sun, C., Stuckey, J., Shaik, S., Waskell, L. (2014) Structural and Functional
26
Characterization of a Cytochrome P450 2B4 F429H Mutant with an Axial Thiolate-Histidine
Hydrogen Bond, Biochemistry, 53, 5080-5091.
34. Gay, S.C., Roberts, A.G., Maekawa, K., Talakad, J.C., Hong, W.X., Zhang, Q., Stout, C.D.,
Halpert, J.R. (2010) Structures of cytochrome P450 2B4 complexed with the antiplatelet drugs ticlopidine and clopidogrel, Biochemistry, 49, 8709-8720.
35. Gay, S.C., Zhang, H., Wilderman, P.R., Roberts, A.G., Liu, T., Li, S., Lin, H.L., Zhang, Q.,
Woods, V.L., Stout, C.D., Hollenberg, P.F., Halpert, J.R. (2011) Structural Analysis of
Mammalian Cytochrome P450 2B4 Covalently Bound to the Mechanism-based Inactivator tert-
Butylphenylacetylene: Insight into Partial Enzymatic Activity, Biochemistry, 50, 4903-4911.
36. Shah, M.B., Jang, H.H., Zhang, Q., Stout, C.D., Halpert, J.R. (2013) X-ray crystal structure of the cytochrome P450 2B4 active site mutant F297A in complex with clopidogrel: Insights into compensatory rearrangements of the binding pocket, Arch. Biochem. Biophys., 530, 64-72.
37. Gay, S.C., Sun, L., Maekawa, K., Halpert, J.R., Stout, C.D. (2009) Crystal structures of cytochrome P450 2B4 in complex with the inhibitor 1-biphenyl-4-methyl-1H-imidazole: ligand- induced structural response through alpha-helical repositioning, Biochemistry, 48, 4762-4771.
38. Liu, J., Shah, M.B., Zhang, Q., Stout, C.D., Halpert, J.R., Wilderman, P.R. (2016) Coumarin
Derivatives as Substrate Probes of Mammalian Cytochromes P450 2B4 and 2B6: Assessing the
Importance of 7-Alkoxy Chain Length, Halogen Substitution, and Non-Active Site Mutations,
Biochemistry, 55, 1997-2007.
39. Srejber, M. et al. (2018) Membrane-attached mammalian cytochromes P450: An overview of the membrane’s effects on structure, drug binding, and interactions with redox partners. Journal of Inorganic Biochemistry 183, 117-136.
27
40. Headlam, M. J., M. C. J. Wilce, R. C. Tuckey (2003) The F-G loop region of cytochrome
P450scc (CYP11A1) interacts with the phospholipid membrane. Biochim. Biophys. Acta
Biomembr. 1617, 96-108.
41. Monk, B. C. et al. (2014) Architecture of a single membrane spanning cytochrome P450 suggests constraints that orient the catalytic domain relative to a bilayer. P.N.A.S. 111, 3865-
3870.
42. McDougle, D. R. et al. (2015) Incorporation of charged residues in the CYP2J2 F-G loop disrupts CYP2J2-lipid bilayer interactions. Biochem. Biophys. Acta Biomembr. 1848, 2460-2470.
43. Baylon, J. L., Lenov, I. L., Sligar, S.G., Tajkhorshid, E. (2013) Characterizing the
Membrane-Bound State of Cytochrome P450 3A4: Structure, Depth of Insertion, and
Orientation. J. Am. Chem. Soc. 135, 8542 – 8551.
44. Ingelman-Sundberg, M. Reconstitution of the liver microsomal hydroxylase system into liposomes. FEBS letters 78, 72-76 (1977).
45. Shaw, P. M., Hosea, N. A., Thompson, D. V., Lenius, J. M. & Guengerich, F. P.
Reconstitution Premixes for Assays Using Purified Recombinant Human Cytochrome P450,
NADPH-Cytochrome P450 Reductase, and Cytochrome b5. Archives of Biochemistry and
Biophysics 348, 107-115 (1997).
46. Denisov, I. G. & Sligar, S. G. Cytochromes P450 in Nanodiscs. Biochimica et Biophysica
Acta: Proteins and Proteomics 1814, 223-229, doi:10.1016/j.bbapap.2010.05.017 (2011).
47. Denisov, I. G. & Sligar, S. G. Nanodiscs in Membrane Biochemistry and Biophysics.
Chemical Reviews (2017).
28
48. Barnaba, C., Martinez, M. J., Taylor, E., Barden, A. O. & Brozik, J. A. (2017) Single Protein
Tracking Reveals that NADPH Mediates the Insertion of Cytochrome P450-Reductase into a
Biomimetic of the Endoplasmic Reticulum. J. Am. Chem. Soc., 139, 5420-5430.
49. Brignac-Huber, L.M., Park, J.W., Reed, J.R., Backes, W.L. Cytochrome P450 Organization and Function are Modulated by Endoplasmic Reticulum Phospholipid Heterogeneity. (2016)
Drug Metabolism and Disposition, 48, 1859-1866.
50. Navratilova, V., Paloncyova, M., Berka, K., Otyepka, M. (2016) Effect of Lipid Charge on
Membrane Immersion of Cytochrome P450 3A4, J. Phys. Chem. B. 120, 11205-11213.
51. Simtchouk, S. Eng, J. L., Meints, C. E., Makins, C., Wolthers, K. R. (2013) Kinetic analysis of cytochrome P450 reductase from Artemisia annua reveals accelerated rates of NADPH- dependent flavin reduction, FEBS J., 280, 6627-6642.
52. Stiborova, M., Indra, R., Frei, E., Kopeckova, K., Schmeiser, H. H., Eckschlager, T., Adam,
V., Heger, Z., Arlt, V. M., Martinek, V. (2017) Cytochrome b5 plays a dual role in the reaction cycle of cytochrome P450 3A4 during oxidation of the anticancer drug ellipticine, Monatsh.
Chem., 148, 1983-1991.
53. Hamdane, D., Xia, C., Im, S. C., Zhang, H., Kim, J. J., and Waskell, L. (2009) Structure and function of an NADPH-cytochrome P450 oxidoreductase in an open conformation capable of reducing cytochrome P450, J. Biol. Chem. 284, 11374-11384.
54. Wang, M., Roberts, D. L., Paschke, R., Shea, T. M., Masters, B. S., and Kim, J. J. (1997)
Three-dimensional structure of NADPH-cytochrome P450 reductase: prototype for FMN- and
FAD-containing enzymes. Proc. Natl. Acad. Sci. U. S. A. 94, 8411-8416.
55. Smith, G. C., Tew, D. G., and Wolf, C. R. (1994) Dissection of NADPH-cytochrome P450 oxidoreductase into distinct functional domains. Proc. Natl. Acad. Sci. U. S. A., 91, 8710-8714.
29
56. Laursen, T., Jensen, K., Moller, B. L. (2011) Conformational changes of the NADPH- dependent cytochrome P450 reductase in the course of electron transfer to cytochromes P450,
BBA – Proteins and Proteomics, 1814, 132-138.
57. Zhao, Q., Modi, S., Smith, G., Paine, M., McDonagh, P. D., Wolf, C. R., Tew, D., Lian, L.
Y., Roberts, G. C., and Driessen, H. P. (1999) Crystal structure of the FMN-binding domain of human cytochrome P450 reductase at 1.93 A resolution. Protein Sci., 8, 298-306.
58. Barsukov, I. et al. (1997) 1H, 15N and 13C NMR resonance assignment, secondary structure and global fold of the FMN-binding domain of human cytochrome P450. J. Biomol. NMR 10,
63–75.
59. Mothersole, R. G., Meints, C. E., Louder, A., Wolthers, K. R. (2016) Role of active site loop in coenzyme binding and flavin reduction in cytochrome P450 reductase, Arch. Biochem.
Biophys., 606, 111-119.
60. Estabrook, R. W., Shet, M. S., Fisher, C. W., Jenkins, C. M., and Waterman, M. R. (1996)
The interaction of NADPH-P450 reductase with P450: an electrochemical study of the role of the flavin mononucleotide-binding domain, Arch. Biochem. Biophys., 333, 308-315.
61. Im, S. C., and Waskell, L. (2011) The interaction of microsomal cytochrome P450 2B4 with its redox partners, cytochrome P450 reductase and cytochrome b(5), Arch. Biochem. Biophys.,
507, 144-153.
62. Bridges, A., Gruenke, L., Chang, Y. T., Vakser, I. A., Loew, G., and Waskell, L. (1998)
Identification of the binding site on cytochrome P450 2B4 for cytochrome b5 and cytochrome
P450 reductase, J. Biol. Chem., 273, 17036-17049.
30
63. Kenaan, C., Zhang, H., Shea, E. V., and Hollenberg, P. F. (2011) Uncovering the role of hydrophobic residues in cytochrome P450-cytochrome P450 reductase interactions,
Biochemistry, 50, 3957-3967.
64. Lederer, F. (1994) The cytochrome b5-fold: An adaptable module, Biochimie, 76, 674-692.
65. Schenkman JB, Jansson I. (2003) The many roles of cytochrome b5, Pharmacol. Ther., 97,
139–152.
66. Ahuja, S. et al. (2013) A model of the membrane-bound cytochrome b5-cytochrome P450 complex from NMR and mutagenesis data. J. Biol. Chem., 288, 22080-22095.
67. Bonfils, C., Balny, C., Maurel, P. (1981) Direct Evidence for Electron Transfer from Ferrous
Cytochrome b5 to the Oxyferrous Intermediate of Liver Microsomal Cytochrome P450 LM2, J.
Biol. Chem. 256, 9457-9465.
68. Stayton, P. S., Fisher, M. T. & Sligar, S. G. (1988) Determination of cytochrome b5 association reactions. Characterization of metmyoglobin and cytochrome P-450cam binding to genetically engineered cytochrome b5. J. Biol. Chem. 263, 13544–8.
69. Chiang, J. Y. (1981) Interaction of purified microsomal cytochrome P-450 with cytochrome b5. Arch. Biochem. Biophys. 211, 662–673.
70. Bendzko, P., Usanov, S. A., Pfeil, W. & Ruckpaul, K. (1982) Role of the hydrophobic tail of cytochrome b5 in the interaction with cytochrome P-450 LM2. Acta Biol. Med. Ger. 41, K1–K8.
71. Hlavica, P. (1984). On the function of cytochrome b5 in the cytochrome P-450-dependent oxygenase system, Arch. Biochem. Biophys., 228, 600 – 608.
72. Ravula T, Barnaba C., Mahajan M., Anantharamaiah G. M., Im S.C., Waskell L.,
Ramamoorthy A. (2017) Membrane environment drives cytochrome P450’s spin transition and its interaction with cytochrome b5, Chem. Commun., 53, 12798-12801.
31
73. Tamburini, P. P., & Gibson, G. G. (1983). Thermodynamic studies of the protein-protein interactions between cytochrome P-450 and cytochrome b5. Evidence for a central role of the cytochrome P-450 spin state in the coupling of substrate and cytochrome b5 binding to the terminal hemoprotein, J. Biol. Chem., 258, 13444 – 13452.
74. Morgan, E., & Coon, M. (1984). Effects of cytochrome b5 on cytochrome P-450-catalyzed reactions. Studies with manganese-substituted cytochrome b5. Drug. Metab. Dispos., 12, 358 –
364.
75. Denisov, I. G., Makris, T. M., Sligar, S. G. & Schlichting, I. (2005) Structure and chemistry of cytochrome P450. Chemical Reviews 105, 2253-2278.
76. Das, A. & Sligar, S. G. (2009) Modulation of the cytochrome P450 reductase redox potential by the phospholipid bilayer. Biochemistry 48, 12104-12112.
77. Kim, K.-H., Ahn, T. & Yun, C.-H. (2003) Membrane properties induced by anionic phospholipids and phosphatidylethanolamine are critical for the membrane binding and catalytic activity of human cytochrome P450 3A4. Biochemistry 42, 15377-15387.
78. Guengerich, F. P. (1999) Cytochrome P-450 3A4: regulation and role in drug metabolism.
Annual review of pharmacology and toxicology 39, 1-17.
79. Jang, H.-H., Kim, D.-H., Ahn, T. & Yun, C.-H. (2010) Functional and conformational modulation of human cytochrome P450 1B1 by anionic phospholipids. Archives of biochemistry and biophysics 493, 143-150.
80. Treuheit, N. A. et al. (2016) Membrane interactions, ligand-dependent dynamics, and stability of cytochrome P4503A4 in lipid nanodiscs. Biochemistry 55, 1058.
81. Yamazaki, H., Johnson, W. W., Ueng, Y.-F., Shimada, T. & Guengerich, F. P. (1996) Lack of Electron Transfer from Cytochrome b5 in Stimulation of Catalytic Activities of Cytochrome
32
P450 3A4 CHARACTERIZATION OF A RECONSTITUTED CYTOCHROME P450
3A4/NADPH-CYTOCHROME P450 REDUCTASE SYSTEM AND STUDIES WITH APO-
CYTOCHROME b5. J. Biol. Chem., 271, 27438-27444.
82. Yamazaki, H. et al. (2002) Roles of NADPH-P450 reductase and apo-and holo-cytochrome b
5 on xenobiotic oxidations catalyzed by 12 recombinant human cytochrome P450s expressed in membranes of Escherichia coli. Protein expression and purification 24, 329-337.
83. Clarke, T. A., Im, S.-C., Bidwai, A. & Waskell, L. The role of the length and sequence of the linker domain of cytochrome b5 in stimulating cytochrome P450 2B4 catalysis. Journal of
Biological Chemistry 279, 36809-36818 (2004).
84. McDougle, D. R., Palaria, A., Magnetta, E., Meling, D. D. & Das, A. (2013) Functional studies of N‐terminally modified CYP2J2 epoxygenase in model lipid bilayers. Protein Science
22, 964-979.
85. Miyamoto, M. et al. (2015) Membrane anchor of cytochrome P450 reductase suppresses the uncoupling of cytochrome P450. Chemical and Pharmaceutical Bulletin 63, 286-294.
86. Grinkova, Y. V., Denisov, I. G., McLean, M. A. & Sligar, S. G. (2013) Oxidase uncoupling in heme monooxygenases: human cytochrome P450 CYP3A4 in Nanodiscs. Biochemical and biophysical research communications 430, 1223-1227.
87. Kim, H.-J., Lee, S.-B., Guengerich, F., Park, Y. & Dong, M.-S. (2007) Effects of N-terminal modification of recombinant human cytochrome P450 1A2 on catalytic activity. Xenobiotica 37,
356-365.
88. Larson, J. R., Coon, M. J. & Porter, T. D. (1991) Purification and properties of a shortened form of cytochrome P-450 2E1: deletion of the NH2-terminal membrane-insertion signal peptide
33 does not alter the catalytic activities. Proceedings of the National Academy of Sciences 88, 9141-
9145.
89. Dong, M.-S. et al. (1996) Identification of Retained N-Formylmethionine in Bacterial
Recombinant Mammalian Cytochrome P450 Proteins with the N-Terminal Sequence
MALLLAVFL...: Roles of Residues 3− 5 in Retention and Membrane Topology. Biochemistry
35, 10031-10040.
90. Hanna, I. H., Kim, M.-S. & Guengerich, F. P. (2001) Heterologous expression of cytochrome
P450 2D6 mutants, electron transfer, and catalysis of bufuralol hydroxylation: the role of aspartate 301 in structural integrity. Archives of biochemistry and biophysics 393, 255-261.
91. Johnson, E. F. & Stout, C. D. (2013) Structural diversity of eukaryotic membrane cytochrome P450s. Journal of Biological Chemistry 288, 17082-17090.
92. Sevrioukova, I. F., Li, H., Zhang, H., Peterson, J. A. & Poulos, T. L. (1999) Structure of a cytochrome P450–redox partner electron-transfer complex. Proceedings of the National
Academy of Sciences 96, 1863-1868.
93. Colthart, A. M. et al. (2016) Detection of substrate-dependent conformational changes in the
P450 fold by nuclear magnetic resonance. Scientific Reports 6.
94. Estrada, D. F., Laurence, J. S. & Scott, E. E. (2016) Cytochrome P450 17A1 interactions with the FMN domain of its reductase as characterized by NMR. Journal of Biological
Chemistry 291, 3990-4003.
95. Scott, E. E. et al. (2016) The role of protein-protein and protein-membrane interactions on
P450 function. Drug Metabolism and Disposition 44, 576-590.
34
96. Dürr, U. H., Waskell, L. & Ramamoorthy, A. (2007) The cytochromes P450 and b 5 and their reductases—promising targets for structural studies by advanced solid-state NMR spectroscopy. Biochimica et Biophysica Acta (BBA)-Biomembranes 1768, 3235-3259.
97. Dürr, U. H., Gildenberg, M. & Ramamoorthy, A. (2012) The magic of bicelles lights up membrane protein structure. Chemical reviews 112, 6054-6074.
98. Xu, J. et al. (2008) Bicelle‐Enabled Structural Studies on a Membrane‐Associated
Cytochrome b5 by Solid‐State MAS NMR Spectroscopy. Angewandte Chemie International
Edition 47, 7864-7867.
99. Zhang, M. et al. Reconstitution of the Cytb5–CytP450 Complex in Nanodiscs for Structural
Studies using NMR Spectroscopy. Angewandte Chemie International Edition 55, 4497-4499
(2016).
100. Yamamoto, K. et al. (2013) Dynamic interaction between membrane-bound full-length cytochrome P450 and cytochrome b5 observed by solid-state NMR spectroscopy. Scientific reports 3, 2538.
101. Soong, R. et al. (2010) Proton-evolved local-field solid-state NMR studies of cytochrome b
5 embedded in bicelles, revealing both structural and dynamical information. Journal of the
American Chemical Society 132, 5779-5788.
102. Zhang, M. et al. (2015) Insights into the role of substrates on the interaction between cytochrome b5 and cytochrome P450 2B4 by NMR. Scientific Reports 5, 8392.
103. Huang, R. et al. (2014) Probing the Transmembrane Structure and Dynamics of Microsomal
NADPH-cytochrome P450 oxidoreductase by Solid-State NMR. Biophysical journal 106, 2126-
2133.
35
104. Huang, R., Zhang, M., Rwere, F., Waskell, L., Ramamoorthy, A. (2015) Kinetic and
Structural Characterization of the Interaction between the FMN Binding Domain of Cytochrome
P450 Reductase and Cytochrome c, J. Biol. Chem., 290, 4843-4855.
105. Cojocaru, V., Winn, P. J. & Wade, R. C. (2007) The ins and outs of cytochrome P450s.
Biochimica et Biophysica Acta (BBA)-General Subjects 1770, 390-401 (2007).
106. Denisov, I. G., Shih, A. Y. & Sligar, S. G. (2012) Structural differences between soluble and membrane bound cytochrome P450s. Journal of Inorganic Biochemistry 108, 150-158.
107. Murtazina, D. A. et al. (2004) Phospholipids modify substrate binding and enzyme activity of human cytochrome P450 27A1. Journal of lipid research 45, 2345-2353.
108. Kiselev, P. et al. (1990) [Regulation of the catalytic activity of the monooxygenase enzyme system depending of the substrate structure and phospholipid composition of the model membrane]. Biokhimiia (Moscow, Russia) 55, 2058-2071.
109. Barnaba, C., Humphreys, S., Barden, A., Jones, J. & Brozik, J. (2016) Substrate Dependent
Native Luminescence From Cytochromes P450 3A4, 2C9, and P450cam. The Journal of
Physical Chemistry B 120, 3038–3047.
110. McClary, W. D., Sumida, J. P., Scian, M., Paço, L. & Atkins, W. M. (2016) Membrane
Fluidity Modulates Thermal Stability and Ligand Binding of Cytochrome P4503A4 in Lipid
Nanodiscs. Biochemistry 55, 6258.
111. Nagar, S. & Korzekwa, K. (2012) Commentary: nonspecific protein binding versus membrane partitioning: it is not just semantics. Drug Metabolism and Disposition 40, 1649-
1652.
112. Ingelman-Sundberg, M., Hagbjörk, A.-L., Ueng, Y.-F., Yamazaki, H. & Guengerich, F. P.
(1996) High rates of substrate hydroxylation by human cytochrome P450 3A4 in reconstituted
36 membranous vesicles: influence of membrane charge. Biochemical and biophysical research communications 221, 318-322.
113. Jeřábek, P., Florián, J. & Martínek, V. (2016) Lipid molecules can induce an opening of membrane-facing tunnels in cytochrome P450 1A2. Physical Chemistry Chemical Physics 18,
30344-30356.
114. Cojocaru, V., Balali-Mood, K., Sansom, M. S. & Wade, R. C. (2011) Structure and dynamics of the membrane-bound cytochrome P450 2C9. PLoS Comput Biol 7, e1002152.
115. Berka, K., Hendrychová, T., Anzenbacher, P. & Otyepka, M. (2011) Membrane position of ibuprofen agrees with suggested access path entrance to cytochrome P450 2C9 active site. The journal of physical chemistry A 115, 11248-11255.
116. Fishelovitch, D., Shaik, S., Wolfson, H. J. & Nussinov, R. (2010) How does the reductase help to regulate the catalytic cycle of cytochrome P450 3A4 using the conserved water channel?
The Journal of Physical Chemistry B 114, 5964-5970.
117. Lapuente-Brun, E. et al. (2013) Supercomplex assembly determines electron flux in the mitochondrial electron transport chain. Science 340, 1567-1570.
118. Kone, B. C., Kuncewicz, T., Zhang, W. & Yu, Z.-Y. (2003) Protein interactions with nitric oxide synthases: controlling the right time, the right place, and the right amount of nitric oxide.
American Journal of Physiology-Renal Physiology 285, F178-F190.
37
Chapter 2
Kinetic and Structural Characterization of the Effects of Membrane on the Complex of Cytochrome b5 and Cytochrome c
2.1 Summary
Cytochrome b5 (cytb5) is a membrane protein vital for the regulation of cytochrome P450 (cytP450) metabolism and is capable of electron transfer to many redox partners. Here, using cyt c as a surrogate for cytP450, we report the effect of membrane on the interaction between full- length cytb5 and cyt c for the first time. As shown through stopped-flow kinetic experiments, electron transfer capable cytb5 - cyt c complexes were formed in the presence of bicelles and nanodiscs. Experimentally measured NMR parameters were used to map the cytb5-cyt c binding interface. Our experimental results identify differences in the binding epitope of cytb5 in the presence and absence of membrane. Notably, in the presence of membrane, cytb5 only engaged cyt c at its lower and upper clefts while the membrane-free cytb5 also uses a distal region. Using restraints generated from both cytb5 and cyt c, a complex structure was generated, and a potential electron transfer pathway was identified. These results demonstrate the importance of studying protein-protein complex formation in membrane mimetic systems. Our results also demonstrate the successful preparation of novel peptide-based lipid nanodiscs, which are detergent-free and possesses size flexibility, and their use for NMR structural studies of membrane proteins.
1This chapter is based on the published paper: Gentry, K.A., Prade, E., Barnaba, C., Zhang, M., Mahajan, M., Im, S.-C., Anantharamaiah, G.M., Nagao, S., Waskell, L., Ramamoorthy, A. (2017) Kinetic and Structural Characterization of the Effects of Membrane on the Complex of Cytochrome b5 and Cytochrome c. Sci. Rep. 7, 7793. 2This thesis research was supported by funds from the National Institutes of Health (NIH to A.R.). 3 Author Contributions: The study was planned by K.A.G. and A.R. K.A.G., E.P., and M.Z. performed NMR experiments, and the results were interpreted by K.A.G. and A.R. S.I. carried out stopped-flow experiments and the results were characterized by C.B., K.A.G., L.W., and A.R. M.M. and K.A.G. obtained docked structure of the complex using HADDOCK. G.M.A. provided the 4F peptide and S.N. provided cyt c. All reported results were reviewed by K.A.G., E.P., C.B., M.M., L.W., and A.R. K.A.G. and A.R. wrote the paper, and all authors reviewed and approved the final manuscript. A.R. directed the project
38
2.2 Introduction
Microsomal cytochrome b5 (cytb5) is a membrane-bound protein that is involved in electron transport to several redox partners including cytochrome P450 (cytP450), several 1,2 oxygenases and desaturases, and cytochrome b5 reductase . Cytochrome b5 interactions ensure function/activation of cytP450, which plays a vital role in cellular metabolism, including the metabolism of over 70% of drugs in the current market and has been implicated in heart diseases and breast and prostate cancers3-5. Because of its membrane-bound native state and the transient nature of cytb5-cytP450 complex formation, there is relatively little structural insight into this disease relevant complex, especially when compared to the interactions of soluble proteins involved in electron transport.
For decades, it has puzzled investigators how cytb5 is able to enhance, reduce, or exert no effect on cytP450 metabolism depending on the substrate and the isoform of cytP450 involved6-10.
Recent studies with microsomal cytP450 have helped elucidate how cytb5 causes it apparent contradictory effects6,11,12, but few studies have been undertaken in the presence of membrane. There are numerous challenges to obtaining high-resolution structural insights into membrane- bound cytP450 due to its size and the tendency for the full-length protein to aggregate in solution, which make it difficult to use the traditional solution NMR experiments10,13,14. Because of these reasons, most reported studies in the literature have focused only on the soluble domain of cytP450, lacking the N-terminal transmembrane domain. On the other hand, previous studies have used cytochrome c (cyt c) as a model for cytP450 in probing the structural interactions between cyt c 15-19 with other proteins like cytb5 or CPR . Cyt c has several similarities to cytP450 including: a similar, overlapping binding domain on cytb5, an overall net positive charge integral for the initial protein-protein complex formation, and membrane has been suggested to promote or enhance activity of both proteins20,21. Although cyt c is a much smaller protein (104 amino acids compared to 491 amino acids in cytP450 2B4), alignment of these two proteins reveal homologous amino acid sequences including known areas of cytb5 interaction (Figure 2.1). Besides being a valid substitute, cyt c also has the advantage of being a well-behaved, NMR-friendly soluble protein.
Structural and kinetic details of the electron-transfer complex of cytochrome b5 – cytochrome c (cyt c) have been the focus of many studies, including NMR spectroscopy22, MD simulations23, and mutagenesis studies16,18,24. While cyt c is not a major physiological electron transfer partner
39
25,26 of cytb5, it has been shown to be a productive electron transfer complex . Previous NMR studies have demonstrated that this complex exists in a 1:1 molar ratio in solution27. In this study, cyt c is used as a model for cyt P450 to examine the role that membrane plays in the formation of an electron transfer cytb5-cytc complex. Most studies thus far have focused on the interaction of cyt c with the soluble domain of cytb5 (without the C-terminal transmembrane (TM) domain) in the 15,16,28 absence of membrane . While several of these studies have shown that the truncated-cytb5 is capable of slow electron transfer to bacterial soluble cytP450s29,30, others have shown truncated- 30-32 cytb5 to be incapable of transferring electrons to mammalian membrane-bound cytP450 . Full- length cytb5 is necessary to fully understand this protein-protein interaction. Our recent studies have shown that lipid membrane plays an important role in the structural interactions between 33,34,35,36 cytb5 and cytP450 . Therefore, it is important to investigate the role of membrane on the interactions between cytb5 and cyt c. Studying membrane proteins is challenging due to their dynamic nature and difficulties encountered during expression and purification. Additionally, in the case of cytb5, as with other single transmembrane helix proteins, the TM domain has been shown to increase the tendency to induce protein aggregation in solution. To combat these difficulties, we have used the full-length
~16-kDa rabbit cytb5 in the presence of membrane mimetics which aid in the monomerization and stabilization of cytb5 to probe how the inclusion of membrane affects the cytb5-cyt c complex formation. Solution NMR experiments are used to probe the transient and dynamic protein-protein complex in a near-native membrane environment. NMR is well-suited for this study as it can provide residue specific details of cytb5-cyt c in the timescale of the complex formation. Two different types of membrane mimetics are used in this study: isotropic bicelles and lipid nanodiscs. Isotropic bicelles have successfully been used to incorporate cytb5 and perform NMR experiments to study its interaction with cytP45033. These isotropic bicelles have a planar lipid bilayer surrounded with detergents. Using bicelles as a membrane mimetic is preferable to the more traditional micelle as the micelle curvature could distort the structural folding of a membrane protein and the detergent could denature the embedded protein37. A benefit of using isotropic bicelles is that they tumble fast in the NMR time scale to enable the application of solution NMR experiments for structural studies38. The second membrane mimetic used in this study is lipid
40
Figure 2.1 Sequence alignment of cytochrome c and cytochrome P450 reveals amino acid sequence similarity. Horse cytochrome c (top) and rabbit cytochrome P450 2B4 are aligned using CHROMA. Positive residues are highlighted in purple, negative residues are highlighted in cyan, serine and threonine are given in cyan text, aliphatic residues are highlighted in yellow, aromatic residues are highlighted in orange.
nanodiscs. Nanodiscs are composed of a lipid bilayer that is typically surrounded by a protein belt. Traditionally, this protein belt is the membrane scaffold protein (MSP)39,40, but in this study we prepared nanodiscs using a short helical amphipathic peptide scaffold, referred to as 4F41. This peptide has several advantages for our experiments compared to traditional MSP. Firstly, the nanodiscs can be prepared without the use of detergents, which are known to irreversibly inactivate proteins37. Secondly, a feature of peptide-based nanodisc is that the size is easily controlled through changing the lipid-to-peptide ratio. Lastly, the preparation allows for highly reproducible production of isotropic nanodiscs for solution NMR studies. In peptide-based nanodiscs, cytb5 is readily monomerized and stable for weeks at room temperature34. As recent studies have identified,
41 cyt c can undergo structural changes in the presence of cardiolipin35,36, membrane mimetics used in these experiments were cardiolipin-free, such that cyt c would not interact with the membrane environment.
In this study, we investigated the interaction between full-length rabbit cytb5 and full-length cyt c in different membrane mimetic environments, including lipid-free, isotropic bicelles, and lipid nanodiscs, utilizing solution NMR techniques. Our study provides further evidence that the interaction between these two redox partners is governed by the presence of lipid bilayers. While previous studies of truncated cytb5 have identified productive binding, our results show an increased, dynamic interaction and changes in the residues most perturbed by complex formation as well as kinetic data demonstrating the formation of a productive complex.
2.3 Materials and Methods 2.3.1 Materials and Reagents Phosphate buffer components (potassium phosphate monobasic and dibasic) were purchased from Sigma-Aldrich (St. Louis, MO). 1,2-dihexanoyl-sn-glycero-3-phosphocholine (DHPC) and 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) were purchased from Avanti Polar Lipids, Inc. (Alabaster, AL). Amino acid sequence of the 4F peptide used to prepare lipid nanodiscs: DWFKAFYDKV AEKFKEAF. Equine cytochrome c was obtained through a collaborator that was prepared and purified as described previously 35. Deuterium oxide and 15N Celtone Base Powder was purchased from Cambridge Isotope Laboratories (Tewksbury, MA).
The 5-mm symmetrical D2O-matched Shigemi NMR microtubes were purchased from Shigemi, Inc (Allison Park, PA).
2.3.2 Preparation of cytochrome b5 15 Full-length uniformly N-labeled and unlabeled wild-type rabbit cytb5 was expressed and purified as described previously 10,43,56. Briefly, E. coli C41 cells were transformed with a pLW01 plasmid containing the cytb5 gene. The cells were grown in LB medium to an OD of 1 at 600 nm. The culture was diluted 100-fold into 10 mL of 15N-Celtone medium. This culture was grown at 35 °C with shaking at 250 rpm until an OD of 1 at 600 nm was achieved. The cells were pelleted and resuspended in 10 mL of fresh 15N-Celtone medium. The resuspended cell culture was added to the final 1 L of culture minimum medium. Isopropyl β-D-thiogalactopyranoside was added to a
42 final concentration of 10 μM, and incubation was continued for 20 h, at which time the cells were 15 57 harvested. N cytochrome b5 was purified as described previously.
2.3.3 Preparation of cytochrome c Full-length uniformly 15N-labeled cyt c was expressed and purified using a procedure as reported in the literature (34).
2.3.4 Preparation of bicelles A DMPC/DHPC isotropic bicelle (q = [DMPC]/[DHPC] = 0.25) was prepared by mixing the appropriate amount of DMPC and DHPC in chloroform. The mixture was vortexed and dried under nitrogen to make a thin film, which was further dried under vacuum overnight at room temperature. The film was then hydrated in 10 mM potassium phosphate buffer, pH 7.4 to a concentration of 5% w/v.
2.3.5 Preparation of nanodiscs DMPC powder was suspended into buffer A (10 mM potassium phosphate, pH 7.4) to make a stock solution at 20 mg/mL. The 4F peptide was dissolved in buffer A to make a stock solution at 10mg/mL. The DMPC stock solution was vortexed and sonicated three times for 30s each to create a suspension, and vortexed thoroughly immediately before use. The stock solution was mixed together at a peptide:lipid ratio of 1:1.5 % w/w and incubated at 37 °C o/n with slow agitation. The nanodiscs were purified by size exclusion chromatography (SEC). A Superdex 200 Increase 300/10 GL column was operated on an AKTA purifier (GE Healthcare, Freiburg, Germany).
2.3.6 NMR experiments and data analysis NMR experiments were performed at 298 K on a Bruker Avance II 600 MHz spectrometer equipped with a cryoprobe. 2D 15N/1H TROSY HSQC spectra were recorded from 0.1 mM 15N- 15 15 cytb5 in buffer A, bicelles containing N-cytb5, or nanodiscs containing N-cytb5. All buffer and 15 nanodisc samples were in buffer A and contained N-cytb5 at a concentration around 0.1 mM.
Cytb5 was incubated with a 5% v/v solution of DMPC/DHPC. Cytb5 was incubated with 4F- DMPC-Nanodiscs at a ratio of 1:1.
43
All NMR spectra were recorded with 64 scans and 256 t1 increments. Data was processed using TopSpin 2.0 (Bruker) and analyzed with Sparky (Goddard). The previously reported cytb5 backbone chemical shift assignments were used in this study 45. The weighted amide chemical shift perturbation (Δδavg) was calculated using the following equation:
퐹 푆푊 2 2 2 Δ훿푎푣𝑔 = √(Δ훿푁 × ) + Δ훿퐻 퐹1푆푊 Where ΔδN and ΔδH are the changes in the chemical shifts of amide nitrogen-15 and amide-proton respectively, while F1SW and F2SW represent the spectral width in the first and second dimensions respectively; chemical shift values are given in ppm 58,59.
2.3.7 Stopped Flow Kinetics All experiments were performed at 25 °C under anaerobic conditions using a Hi-Tech SF61DX2 stopped-flow spectrophotometer (Bradford-on-Avon, UK) housed in an anaerobic Belle Technology glove box (Weymouth, UK). The buffer was purged with nitrogen gas for 30 minutes for deoxygenation prior to being transferred to the glove box. All protein solutions were incubated overnight at 4 °C in the glove box to eliminate oxygen. For measuring the kinetics of cyt c reduction by cytb5, cytb5 was pre-reduced anaerobically as described below and then loaded in syringe 3 of the stopped flow, and syringe 2 was loaded with ferric cyt c. Spectra were recorded in the photodiode array mode between the wavelength of 300 and 750 nm after mixing equal volumes of the 12 μM cytb5 and the 13.2 μM cyt c containing solutions resulting in final concentrations of 6 μM cytb5 and 6.6 μM cyt c. Each set of samples (free non-membrane bound cytb5, bicelle cytb5, nanodisc cytb5) was measured over different amounts of time between 0.8 and 20 s.
2.3.8 Data analysis and kinetic modeling Spectra were collected and Singular Value Decomposition (SVD) was applied in the differential spectrum using an algorithm written in Matlab (The Mathworks Inc., v.R2016b) 60. Before SVD, Savitzky-Golay smoothing and peak-to-area normalization were performed on all the recorded spectra to increase signal-to-noise ratio. SVD analysis was then performed using the built-in Matlab function, obtaining the corresponding kinetic and spectral eigenvectors. The number of principal components was chosen by visualization of the scree plot (Figure SI6,
44
Supplemental Information), as well as considering the variance covered by the selected eigenvectors. The fraction of variance was computed as following:
2 휎푞 푓푞 = 푚 2 , 푞 = 1, … , 푚 (2) ∑푡=1 휎푡 푡ℎ 2 where 푓푞 represents the fraction of the expression level contributed by the 푞 eigenvector, 휎푞 is 푡ℎ 2 the variance of the 푞 eigenvector, and 휎푡 the total variance. In all the performed experiments, two eigenvectors were selected (PC1 and PC2), covering 70-90% of the total variance.
2.3.9 Kinetic Modeling and Simulations For the three systems (lipid-free, bicelles and nanodiscs), based on the evolution over time and peak maxima, PC1 was assigned to cyt c, and PC2 to cytb5. The final concentration of ferrous cytochrome c was computed by measuring the absorbance at ~550 nm and using ε = 21.2 mM-1 -1 cm . Cytb5 was considered to be completely oxidized at the end of the experiment. The kinetics model that describe the oxidoreductive reaction is depicted in Figure 2. The differential equations to model the electron transfer kinetics were parameterized using the NonlinearModelFit function with 1/Y weighting in Mathematica 11.0 (Wolfram Research, Champagne, IL). When fitting parameters, the ParametricNDSolveValue function was used for numerical solutions of the differential equations with MaxSteps → 100,000 and PrecisionGoal → ∞.
2.3.10 Titration of cytb5 by Dithionite under Anaerobic Conditions
Cytb5 (12 μM) was titrated with a standardized sodium dithionite solution under anaerobic conditions while monitoring the UV-visible spectra with a Cary 4000 spectrometer between 300 and 700 nm. The stock solution of cytb5 was incubated overnight at 4 °C in an anaerobic Belle Technology glove box (Hi-Tech, Salisbury, UK) to remove oxygen. The titrant (sodium dithionite solution) was prepared in the glove box in oxygen free buffer, and its concentration was calculated using an extinction coefficient of 8.04 mm-1 cm-1 at 315 nm. The titration was carried out in a tonometer, a homemade anaerobic titration apparatus.
2.3.11 Calculation of a membrane-bound cytb5-cyt c complex using NMR data The HADDOCK (High Ambiguity Driven protein-protein DOCKing) docking software 51,52 was used to calculate the structures of the cytb5-cyt c complex based on experimental NMR
45 restraints. The HADDOCK algorithm includes three consecutive steps: 1) rigid body docking in which the two molecules are rotated and translated randomly in turn to minimize intermolecular energy; 2) simulated annealing of structures in which annealing in torsion angle space is performed to refine the orientation of the proteins and the side chains and/or backbones of the interface residues; and 3) solvent refinement in which the structures are further refined in an 8.0 Ǻ shell of TIP3P water molecules. The starting structures for this docking were the solution NMR structure of rabbit cytb5 (PDB: 2M33) and the crystal structure of horse cyt c (PDB: 1HRC). Ligand topology and parameter files were generated using PRODRG2 server 61. Defined ambiguous restraints were generated from NMR experiments performed in this paper, both active and passive residues with solvent accessibility from both proteins. 1000 structures were generated in the rigid body docking step, followed by simulated annealing of the 200 lowest energy structures from the last step, and 200 structures were selected for solvent refinement. The resulting 200 final structures were analyzed and grouped into clusters based on the backbone root mean square deviation values. Molecular structures of the complexes were viewed and graphed using PyMOL.
2.4 Results 2.4.1 Incorporation of the cytb5 – cyt c complex into membrane mimetics
In order to study the interaction between cytb5 and cyt c, full-length cytb5 was 34,42-44 overexpressed, purified and characterized as reported previously . Three different cytb5 samples were prepared in this study: cytb5 in buffer (membrane-free cytb5); cytb5 in isotropic bicelles; and cytb5 in lipid nanodiscs. Cytb5 was added to DMPC/DHPC bicelles. While the isotropic bicelles stabilize the protein, they are not stable enough to be used in size exclusion chromatography. However, the monomerization and incorporation of cytb5 in lipid nanodiscs can be monitored by Size Exclusion Chromatography (SEC) and Dynamic Light Scattering (DLS).
Figure 2.2(B and C) (green trace) demonstrate that cytb5 has been incorporated into the lipid nanodiscs in a homologous manner. After purification of the cytb5 in lipid nanodiscs, cyt c was added to the sample and run through SEC and DLS experiments again (blue trace). These measurements show the complex formation between cytb5 and cyt c in nanodiscs as indicated by the change in the size of the diameter fit to the DLS data and the elution profile from the SEC (Fig.
2.2). Cyt c does not interact with the nanodisc, but it only interacts with cytb5. (Figure 2.3)
46
Figure 2.2: Reconstitution of cytb5 – cyt c in lipid nanodiscs. (A) Schematic of the interaction between cytb5 (magenta; PDB: 2M33) and cyt c (blue; PDB: 1HRC) in a lipid nanodisc. (B) Dynamic light scattering (DLS) was used to determine the size of the nanodiscs. Radii of empty 4F-DMPC-nanodiscs (black), cytb5 in nanodiscs (green), and cytb5 – cyt c in nanodiscs (blue). (C) Size exclusion chromatography (SEC) elution profiles of empty 4F-DMPC-nanodiscs (black) which eluted at 15.1 mL, cytb5 in nanodiscs (green) which eluted at 12.6 mL, and cytb5 – cyt c in nanodiscs (blue) which eluted at 12.3 mL.
47
Figure 2.3: Cytochrome c does not interact with the 4F-DMPC nanodiscs. The 2D 1H-15N TROSY HSQC spectra is shown with a reference spectrum of 15N-cyt c (in black) and with 15N-cyt c in the presence of 1 molar equivalent of 4F-DMPC nanodiscs. Differential line broadening for the cyt c are shown below with the bold horizontal representing the mean and the dashed line representing one standard deviation below the mean. There are no chemical shift perturbations of cyt c in this experiment.
2.4.2 Formation of productive electron transfer complex between cytb5 and cyt c
Stopped flow kinetic experiments were performed to verify that the cytb5-cyt c complex was active in all three preparations. The electron transfer rate was monitored between reduced cytb5 and oxidized cyt c. Ferric cytb5 was reduced by sodium dithionite and the reduction was monitored by UV-Vis spectrometry (Figure 2.4). Singular Value Decomposition (SVD) was
48 applied to the difference spectra to deconvolute the principal spectral components (Figure 2.5(A,C,E)). For the three samples, two eigenvectors were selected, which variance sum covered 82%, 69%, and 75% of total variance for lipid-free, bicelles, and nanodiscs, respectively. For bicelles, since the cumulative variance covered by the first two eigenvectors was slightly <70%, we used the scree plot to confirm that the first two components (PC1 and PC2) explained most of the variability in the data (Figure 2.6). The two components and their peak maxima and minima can be identified as either cytb5 or cyt c based on the difference spectra for each preparation of cytb5 (lipid-free, bicelles, and nanodiscs). Differences in the peak positions are attributed to the presence of lipids in the solution.
Figure 2.4: UV absorption profiles of cytb5 in lipid-free solution from oxidized to reduced. Oxidized cytb5 was titrated with sodium dithionate to reduced cytb5 (red).
Kinetics traces corresponding to the oxidoreductive reactions (Figure 2.5(B,D,F)) were modeled according to the scheme depicted in Fig 2.5G. After formation of the complex between ferrous cytb5 and ferric cyt c (k1 and k2), the electron transfer occurs (k3). We postulate that after the initial electron transfer no further reaction occurs, since it is not possible to distinguish between free and complexed cytochromes based on the absorption spectrum. Two observations were made: 1) the interaction between the redox partners is functional, and 2) the kinetic micro-rates indicate significant differences in the protein binding and electron transfer.
49
Figure 2.5 Spectral deconvolution and kinetic modeling reveal membrane environment dependent changes. Time-dependent absorption spectra of the redox couple are depicted for lipid-free (A), bicelles (C), and nanodiscs (E) mimetic membranes. The inset shows the eigenvectors associated with cyt c reduction (PC1) and cytb5 oxidation (PC1); absorbance maxima are also indicated. The shift of λmax as well as the peak intensity, an indicator of how far along the reaction has progressed, can be attributed to the presence of lipids. Numerical fittings are shown for lipid-free (B), bicelles (D), and nanodiscs (F). (G) Kinetic scheme of electron transfer between cytb5 and cyt c. In the first reaction, oxidized cytb5 and reduced cyt c form a complex with association rate k1 and dissociation rate k2. Once the first reaction has occurred then the second reaction occurs irreversibly with a rate of k3 of the electron transfer from oxidized cytb5 to reduced cyt c giving oxidized cytb5 and reduced cyt c.
Table 2.1 reports the micro-rates determined from fitting the time-dependent kinetic traces with the numerical method. For lipid-free systems, the dissociation rate (k2) was fixed to zero, in order to allow the fitting to converge to a minimum. Bicelles and nanodiscs showed similar
50 association and dissociation rates. Regarding electron transfer, the average micro-rate (k3) was -1 -1 -1 6.82 s for lipid-free, 0.92 s for bicelles, and 1.23 s for nanodiscs cytb5 samples. These results indicate that all three samples form functional complexes with membrane-free cytb5 having the fastest electron transfer event while nanodiscs reconstituted cytb5 has slightly faster rate than bicelles reconstituted cytb5.
Figure 2.6 Results from SVD analysis applied to time-dependent spectra of electron transfer reaction in cytb5-cyt c complex reconstituted in bicelles. a) raw difference spectrum (R0), and subtraction of first (R1), second (R2) and third (R3) components; b) scree plot of singular values (S), showing the components considered above background (as indicated by the red dashed line). As from both graphs, we assumed that any residual signal after subtraction of PC1 (that is R1) and PC2 (that is R2) cannot be spectrally resolved, and thus R2, R3, and the remaining components (Ri; i>3) were considered as indistinguishable from noise.
-1 -1 -1 2 k1 (s ) k2 (s ) k3 (s ) R
Lipid-free cytb5 0.47 ± 0.01 ~0a 6.82 ± 0.75 0.983
Bicelle-cytb5 0.35 ± 0.01 46.16 ± 0.86 0.92 ± 0.00 0.985
Nanodiscs-cytb5 0.29 ± 0.01 54.54 ± 0.35 1.23 ± 0.01 0.980
Table 2.1: Kinetic micro-rates obtained from numerical fitting of the stopped-flow time-time dependent traces using the kinetic scheme depicted. aFixed
2.4.3 NMR experiments probing the interaction between cyt c and 15N-cytb5 Two-dimensional 15N/1H TROSY-HSQC spectra of uniformly-15N-labeled full-length rabbit cytb5 were recorded to monitor changes induced by the titration of unlabeled cyt c. Both proteins were used in their ferric low spin oxidized forms, and titrations were carried out in a lipid- free solution, DMPC/DHPC bicelles, and DMPC-4F nanodiscs15,16,18. N-edited 1H 1D NMR
51 spectra shown in Figure 2.7 (top) reveal the changes in the signal intensity of cytb5 observed from a lipid-free solution, bicelles, or nanodiscs samples as illustrated through spectra acquired over the course of the titration with unlabeled cyt c. In the membrane-free cytb5 titration (Figure 2.7A), there is a significant decrease in signal over the course of the first titration point (1:0, cytb5: cyt c) to second titration point (1:0.3, cytb5: cyt c). This decrease in signal could be attributed to a couple of things: (i) as cytb5 aggregates, signal is lost (or reduced in intensity); (ii) similar to the trend shown in the kinetics data, membrane-free cytb5 and cyt c are forming a static complex in solution. This larger, static complex could also lead to a decrease in tumbling speed and loss of signal.
Membrane bound cytb5 samples do not exhibit this effect and the NMR signal intensity remains relatively constant throughout the course of the titration with unlabeled cyt c.
1 15 The assignment of H- N resonances observed in a HSQC spectrum of cytb5 has been published previously45 and is shown in Figure S5. The 2D HSQC spectra provide residue specific detail about any changes in cytb5. Upon titration with cyt c in this experiment, the chemical environment of these backbone amide-NH groups can be affected which can be monitored through changes observed in the 2D HSQC spectra. Two types of information can be gathered from these NMR experiments: chemical shift perturbations (Figure 2.7D,E,F) and differential line broadening
(Figure 2.8). It is expected that upon titration of cyt c to cytb5, complex formation may alter the amide-NH chemical shift values of cytb5 residues undergoing fast to intermediate time scale exchange between a free and bound forms. Chemical shift perturbations are hallmarks of transient complex formation and may be induced by direct interaction with the binding partner, as well as an overall change in protein conformation. Another possible contribution to CSPs could be due to the ring current effect from the heme groups of these two proteins. However, we expect this effect to be negligible for our system, based on the findings of Shao et al.16. Differential line broadening is indicative of a more stable, tighter complex formation due to the increase in the correlation time of the protein-protein complex. For instance, a rather flexible residue may be rigidified upon binding with cyt c, and its resonance will thus be broadened due to a slower tumbling rate. In addition to correlation time effects, resonance lines may be broadened due to a shift in population levels between different residue conformers46.
52
1 15 15 Figure 2.7: 2D H- N HSQC-TROSY spectra of N-labeled cytb5 revealing the interaction between cytb5 and cyt c. The 15 15 signal intensities over the course of cyt c with the N labeled cytb5 titration experiment are displayed in both N- edited 1D spectra (top) and accompanying 2D-TROSY-HSQC spectra (bottom) for the molar ratio of cytb5: cyt c 1:0 (black) and 1:1 (red). (A) Free cytb5 has a large loss of signal intensity throughout the course of the experiment, attributed to cytb5 aggregation in solution. (B) bicelle cytb5 reconstituted in bicelles and (C) nanodisc cytb5 reconstituted in nanodiscs both maintain signal intensity throughout the course of the experiment. Changes in the signal intensity and line width due to cytb5-cyt c interactions are shown in Figure 6 1 15 and discussed in the main text. H- N HSQC spectra demonstrating chemical shift perturbations for cytb5 in buffer (no membrane) (D), bicelles (E), or nanodiscs (F) in the presence of one molar equivalent of cyt c (red). As a reference, the spectrum of cytb5 is shown without addition of cyt c (black).
After titration of cyt c into cytb5, the average chemical shift perturbations (CSPs) of cytb5 measured for lipid-free solution, bicelles, and nanodiscs are 0.016, 0.014, and 0.021 ppm, respectively. These CSPs are not very large which suggests a weak complex formation. The
53 residues with high CSPs are widespread on the surface of the protein, which is a sign of encounter complexes present in the sample that are expected to form with electron transfer proteins47-49. While there are unique CSP data for each condition (lipid-free solution, bicelles, or nanodiscs), there is overlap of identified residues and a general region of cytb5 which seems to be most affected (Figure 2.9). All three conditions identify residue E61’s resonance to be highly perturbed and likely to be involved in the interactive interface of the cytb5–cyt c complex. Most of the residues identified as highly perturbed are located on the upper and lower clefts surrounding the heme group of cytb5. In the lipid-free sample, Leu99 and Asp104 were identified as highly perturbed. These residues are located in a flexible region of cytb5 and most likely affected only in this particular sample because cytb5 is not anchored in a lipid bilayer, unlike in bicelles or nanodiscs.
Figure 2.8: Differential line broadening for cytb5 in buffer (no membrane) (A), bicelles (B), or nanodiscs (C) in the presence of 1 molar equivalent of cyt c (red). As a reference, cytb5 is shown without the addition of cyt c (black). 1D spectral lines were extracted from the 2D 1H-15N TROSY HSQC spectra for residue K7 as a reference, and one of the broadened peaks as indicated in the respective spectra to demonstrate differential line broadening induced by cyt c binding with cytb5. To account for the loss of signal intensity, the 1D traces of cytb5 in the presence of cyt c were scaled-up a factor of 3.02 (buffer), 1.32 (bicelles), and 1.65 (nanodiscs) in order to compare with the peak from K7.
54
Figure 2.9: Cyt c interaction induced chemical shift perturbations of cytb5. Reported in red are residues with CSPs greater than one standard deviation for cytb5 in membrane-free solution (A), bicelles (B), and nanodiscs (C). One standard deviation above the mean is represented by the horizontal line and the mean is represented by the dashed line. The significant residues are colored in blue and are mapped onto structures of cytb5 rotated 180° (PDB: 2M33).
Differential line broadening was seen for all conditions of the cyt c titration into cytb5. In this protein-protein interaction, differential line broadening is a complementary metric to CSP data to analyze the binding as it covers the timescale of the complex that falls into the time scale of ~10-3 s. The average relative signal intensity observed for the lipid-free solution, bicelles, and
55 nanodiscs samples are 63.9, 79.3, and 65.1% respectively (Figure 2.10). These residues are mapped onto cytb5 structures (PDB: 2m33) shown in Figure 6 with lipid-free solution (Fig. 2.11A), bicelles (Fig. 2.11B), and nanodiscs (Fig. 2.11C). In the lipid-free solution (Figure 2.10A), there is general line broadening of resonances from cytb5 in the heme pocket with only one significant residue in the lower cleft identified, His68. The widespread line broadening is concentrated on solvent exposed residues, which can be attributed to the encounter complexes that form in redox partner pairs before the productive binding site is found.
56
Figure 2.10: Differential line broadening reveals binding sites on cytb5. Reported in yellow are residues with depleted signal intensity less than the mean and residues in blue have depleted signal intensity less than one standard deviation below the mean for cytb5 in no membrane (A), bicelles (B), and nanodiscs (C). The thick horizontal line represents the mean and the dashed horizontal line represents on standard deviation below the mean.
There is consensus in affected residues from both the chemical shift perturbation data as well as the differential line broadening data. Similar areas are implicated in binding, generally the upper and lower clefts of cytb5 which surround the heme. Differences in the two types of data can be attributed to that line broadening may be due to direct binding and stabilization of the proteins, whereas CSPs can be induced by global changes. Cytb5 in the bicelle sample (Figure 2.10B) identifies more significant residues than lipid-free cytb5. These important residues are in the heme pocket, specifically in the lower cleft: Gly56, Gly57, Asn58, Ala59, Thr60, Glu61, Gln62. The lipid nanodisc sample (Figure 2.10C) has the similar average relative peak intensity to lipid-free cytb5 but has eleven important residues: I29, H32, T38, K39, T60, E61, D65, H68, T70, and L84.
These residues are mainly clustered on the lower cleft of cytb5; the residues L51, R52, and R89 in the nanodisc sample could not be assigned.
2.4.4 NMR experiments probing the interaction between 15N-cyt c and cytb5 As with the 15N-labelled cytb5 NMR experiments, two-dimensional 15N-1H TROSY- 15 HSQC spectra of uniformly N-labeled equine cyt c were recorded to monitor the cyt c – cytb5 interaction on the cyt c interface. Full-length cytb5 in 4F-DMPC nanodiscs was used in this titration 15 1 15 experiment based on the previous N-labeled cytb5 NMR experiments. The H- N amide resonances have been reported in the literature50 and the assignment is shown in Figure 2.3.
CSPs were of similar magnitude to the chemical shifts perturbations of cytb5, with an average value of 0.027 ppm. For the cyt c CSP calculations, the presented data (Figure 2.12B) is calculated from the 1:0.6 (cyt c/cytb5 molar ratio) rather than the 1:1 ratio as many of the residues in the 1:1 spectra have been broadened beyond detection. The nine residues which have the strongest CSPs (Val3, Val20, Lys22, Leu35, Ala51, Glu62, Leu68, Tyr97, and Lys99) are widespread throughout cyt c.
57
Figure 2.11: Implicated binding sites on cytb5 mapped with differential line broadening data. The data from figure 5 is mapped onto a solution NMR structure (PDB: 2M33) of cytb5 for a membrane-free solution (A), bicelles (B), and nanodiscs (C). Residues falling one standard deviation below the mean are colored in blue and residues falling under the mean are colored in yellow.
58
1 15 15 1 Figure 2.12: 2D H- N HSQC-TROSY spectra of N-labeled cyt c revealing interaction between cyt c and cytb5. (A) H- 15 15 15 N spectra of N-labeled cyt c (black) and N-labeled cyt c in the presence of one molar equivalent of cytb5 reconstituted in 4F- DMPC nanodiscs (red). The full 15N-labeled cyt c spectrum with resonance assignment can be found in SI Figure 5. (B) Chemical
59 shift perturbations of cyt c residues. Reported in red are residues with CSPs greater than one standard deviation. (C) Differential line broadening observed for cyt c residues. The thick horizontal line represents the mean, with the residues colored in yellow, while the dashed horizontal line represents one standard deviation below the mean. The inset in Figure 8A is the crystal structure of cyt c (PDB: 1HRC) with mapping of experimentally measured differential line broadenings (data reported in 8C) for residues falling under the mean colored in yellow and residues with complete signal intensity depletion in blue.
Substantial differential line broadening was seen for the titration of cytb5 into cyt c. A complete depletion of signal was observed for 28 residues. The average intensity of the remaining residues was 17%. (Figure 2.13C) These affected residues are located around in the unstructured loops surrounding the heme group of cyt c. One explanation for the dramatic decrease in peak intensity is due to a greater increase of size upon complex formation; it should be noted that cyt c alone does not bind to membrane. Viewing the complex from the 10-kDa cyt c, a 15-kDa cytb5 molecule embedded in a ~120-kDa nanodisc causes a much bigger effect on the tumbling rate than that of cyt c on the cytb5-ND side.
60
Figure 2.13: HADDOCK-generated structures reveal complex between cytb5 and cyt c. (A) Using the chemical shift perturbations and differential line broadening data from the cytb5-4F-DMPC-Nanodiscs and cyt c NMR titration experiments, a complex structure was calculated. Cytb5 is in magenta and cyt c is in blue. (B) Residues implicated in the interface of cytb5-cyt c from the literature (17) are mapped onto our HADDOCK structure; cytb5 residues in yellow and cyt c residues in cyan. (C) Residues implicated in the interface of cytb5-cytP450 from the literature (42) are mapped onto our HADDOCK structure; cytb5 residues in yellow.
2.4.5 Structural Model of the membrane-bound cytb5-cyt c complex
A structural complex of cytb5-cyt c was generated using the information driven docking 51,52 program HADDOCK 2.2 with experimental data obtained in lipid membranes (cyt c and cytb5 in nanodiscs). NMR based chemical shift perturbation (CSPs) and differential line broadening data were used as proximity restraints to guide the docking process whereby rigid body docking follows the semi-flexible refinement and energy minimization in explicit solvent to allow the free movement of backbone and side chain atoms of the selected amino acids to improve the intermolecular packing at protein interface. Solvent-accessible residues of cyt c identified from CSPs and differential line broadening (Figure 2.12) upon complex formation are selected as the active ambiguous restraints including residues Gln16, Cys17, His18, Thr19, Val20, Glu21, Gly29, Thr49, Lys79, Met80, and Ile81. Passive restraints, which are defined as solvent accessible residues around the active restraints, include Lys27, Thr28, Thr78, Phe82, and Ala83. For the cytb5 side of the complex, active ambiguous restraints (AIRs) were defined from the cytb5 reconstituted in nanodiscs NMR experiments (Figure 2.9C and 2.10C) including: Asp65, His68, and Thr70. The passive residues were defined as Gly67 and Ser69. Some of the residues for cytb5 that were identified from the NMR experiments were not included in simulations as they were distant from the binding interface of proteins and were observed to penalize the HADDOCK score. Additionally, these residues are involved in significant CSPs because of the encounter complex formation formed prior to productive electron transfer complex. The HADDOCK run was performed as described under “Materials and Methods”. Careful analysis and comparison of HADDOCK results from various clusters along with experimental data (CSPs and differential line broadening) credits cluster 1 as the most probable model for the complex (Figure 2.13). From the 200 lowest energy structures, Cluster 1 is the largest (covers 75%) with better energy and Z score relative to other clusters (Table 2.2)
61
Parameters Cluster 1 Backbone r.m.s.d. (Å) 2.1 +/- 1.3 Van der Waals energy -22.5 +/- 2.8 Electrostatic energy -282.2 +/- 16.0 Desolvation energy 11.8 +/- 4.3 Restraints violation energy 1.4 +/- 1.24 Buried surface area 762.3 +/- 71.5
Table 2.2: Energy statistics for lowest energy cluster of the complex between cytb5 and cyt c generated from HADDOCK.
In cluster 1, the binding interface of “productive” cytb5-cyt c complex comprises of polar and charged residues from the helical hairpin of the lower cleft of cytb5 and the unstructured loops of cyt c. The distance between the two heme groups (Fe-Fe) was estimated as 18.5Å and is perpendicularly oriented to each other. This is similar to the reported fluorescence quenching studies, where distance between the prosthetic groups was observed to be ~18Å with the almost perpendicular heme planes17.
2.5 Discussion While it is highly important to determine the high-resolution dynamics-enriched structural interactions between cytP450 and its redox partners such as cytb5, the large-size and aggregation- prone full-length cytP450 poses numerous challenges to the existing biophysical techniques. To overcome this challenge, cyt c has been used as a substitute for cytP450 as cyt c is a stable and well-behaved protein that has some important features like that of cytP45015,16,18,28. Therefore, we used cyt c in this study, for the first time, to probe the complex formation between cytb5 and cyt c in the presence of a lipid bilayer membrane. Both bicelles and nanodiscs were used as a model membrane in this study43,53. Since bicelles are not free of detergents, and detergents are not desirable for studies on a membrane protein or a membrane-bound protein-protein complex, results obtained from bicelles were also compared with that obtained from lipid nanodiscs. A synthetic peptide (denoted as 4F) capable of solubilizing lipids without the use of any detergents to form nanodiscs54 was used to form a stable lipid bilayer containing nanodisc as characterized by DLS
62 and SEC experiments (Figure 2.2). Previous studies reported the physicochemical characterization of 4F and its use as an apolipoprotein A-I mimetic for atherosclerosis inhibition have been reported54,55. NMR experiments demonstrate that the nanodiscs are suitable for structural studies by solution NMR spectroscopy. Our NMR experiments also illustrate how cytb5 is more stable in the presence of lipids and that studies with full-length cytb5 can be accomplished using membrane mimetics without aggregation related problems16.
Our study reveals that cytb5 in a lipid-free environment, in isotropic bicelles, and nanodiscs can reduce cyt c. As shown in Figure 2, stopped-flow experiments revealed that electron transfer from cytb5 to cyt c occurs faster in the lipid-free cytb5 sample over the membrane-bound cytb5 samples. One explanation for this behavior is that the membrane-free cytb5 can tumble faster than that in bicelles or nanodiscs, and thus able to find cyt c quicker to form a productive complex. This is also revealed by the higher affinity (k2 ~0), indicating the absence of any observable dynamic between the two proteins. The nanodiscs reconstituted cytb5-cyt c sample appears to support electron transfer function better than the bicelles reconstituted cytb5-cyt c sample as indicated by a quicker rate of electron transfer in the nanodiscs sample. This is possibility due to the size of the nanodisc versus the size of the bicelle – the nanodiscs are much smaller; another possible reason could be the presence of detergents in bicelles that may cause the observed slow electron transfer rate. Nanodiscs are a highly suitable system to study the interaction between the cytb5 – cyt c complex as demonstrated by SEC, DLS, and NMR data. The advantageous use of nanodiscs is reflected in the functional complex formation, as well as the high stability of cytb5, and the well- resolved NMR spectra. In comparison to the bicelle sample, the nanodiscs can both monomerize and stabilize the protein whereas the bicelles can only help to stabilize.
Many of the lipid-free cytb5 residues implicated in the binding to cyt c are in two clusters:
63 the upper and lower clefts surrounding the active site and the beta sheets at the back of cytb5. The residues with high chemical shift perturbations in the upper and lower clefts are: Thr38 in the α2 helix, Leu51 in the α3 helix, Glu61 in the α4 helix, and Arg73 in the α5 helix. The other identified residues are: Thr26 in the loop between β2 and β3 strands, Val34 in the β3 strand, Gly46 in the β4 strand, Gly47 in the loop between β4 strand and α3 helix, Ile81 and Gly82 in the β5 strand, and
Leu99 and Asp104 located in the flexible linker region between the soluble and transmembrane domains. Differential line broadening data for this sample revealed four residues: His68, one of the axial ligands for the iron in the heme; Asn21, the linker between turn 1 and β2 strand; Tyr35 and Asp36 in β3 strand. While the exact residues implicated are not all the same as reported in the previous study, the implicated residues in this sample are like previous findings 16, with residues falling into two different clusters.
In the bicelles-reconstituted cytb5 sample, the main cluster of residues suggested to be involved in binding to cyt c are found on the upper and lower clefts of the protein. Chemical shift perturbation data indicate the residues in the lower and upper clefts to be: Thr38 in the α2 helix,
Val50 and Arg52 in the α3 helix, and Glu61 and Phe63 in the α4 helix. Other residues in more flexible areas of the protein are Leu37 in the linker between β3 strand and α2 helix, Glu54 in turn
3, and Arg89 in turn 4, as well as Asn36 in the β3 strand. The differential line broadening data mostly reveal residues from turn 3 to the α4 helix including Gly56, Gly57, Asp58, Ala59, Thr60,
Glu61, and Asn62. The α5 helix residues Ala72, Leu75, and Lys77 were also found to be involved.
Three other residues, not located in this area, that were likely to be involved in the interaction are:
Ile in the α1 helix, Trp27 in the β3 sheet, and I81 in the β5 sheet. Unlike the Shao et al. 200316 study, these residues mainly fall into one cluster surrounding the active site. The nanodisc
15 reconstituted cytb5 sample has more overall line broadening effects as the overall N-edited proton
64 signal intensity decreased the most in comparison (Figure 2.9). Eleven residues are indicated to be highly involved in binding: three in β3 sheet, Ile29, His32, and Tyr35; two in α2 helix, Thr38 and
Lys39; three in α4 helix Thr60, Glu61, and Asp65; the axial ligand His68; Thr70 in α5 helix, and
Leu84 in a loop. The chemical shift perturbation data from nanodiscs does reveal more residues, with many around the turn3 to α4 helix, similarly to the bicelles reconstituted cytb5 data: Gly57,
Asp58, Glu61, and Phe63. NMR results from both the nanodiscs and bicelles reconstituted cytb5 also indicate residues from turn 3 to the α4 helix could play a role in the interaction between the protein in the presence of lipids as the membrane-free cytb5 sample did not identify the residues in this region.
The binding mode of the intermolecular complex (cytb5-cyt c) was estimated using a data driven docking approach implemented in HADDOCK. The generated structural model (Figure
8A) presents a binding interface focused on the helical hairpin of the lower cleft of cytb5 and the unstructured loops cyt c. The predicted electron transfer scheme generated from the cytb5-cyt c complex structure with HARLEM62 shows a predicted, physiologically-feasible electron transfer path from the heme b of cytb5 through Lys27 to Thr28 to heme c. This interface was compared with that calculated in the literature by Deep et al.18. The residues implicated in binding in this study are mapped onto our complex in Figure 8B in yellow for cytb5 and cyan for cyt c. These results reveal a difference between our calculated structure and previously reported structures: we propose a front to front interaction as opposed to the reported side to side interaction. Comparing
43 the results from this study on cytb5-cyt c with that of Ahuja et al. on cytb5-cytP450, we observed that the interacting interface of cytb5 is similar and contains overlapping residues for interactions with both cytP450 and cyt c. In Figure 8C, the residues implicated in the cytb5-cytP450 structure
43 by Ahuja et al. are mapped on our model structure, with the cytb5 residues highlighted in yellow.
65
The cytb5 residues align very well with our proposed complex structure, illustrating both the importance of utilizing membrane mimetics and the viability of using cyt c as a model. The docking simulations were also performed using the active and passive residues only from the upper cleft of cytb5. The structures generated post simulation have very low cluster size and high RMSD.
Thus, it cannot be considered to be involved in intermolecular interaction. Therefore, NMR detected residues away from binding interface show that enzymes undergo multiple conformational substrates.
2.6 Conclusion
In summary, our studies illustrate the importance and advantages of studying complex formation of membrane proteins in their native, membrane environment. Both bicelles and nanodiscs stabilize a functional cytb5 – cyt c complex. To the best of our knowledge, this is the first report on the use of the 4F peptide based nanodiscs to reconstitute a membrane protein or a protein-protein complex, while previous studies characterized the physiochemical properties and demonstrated the use of 4F as an apoA-I-mimetic for atherosclerosis inhibition54,55. With the stability of nanodiscs allowing other characterization methods through SEC and DLS measurements, along with smaller size, nanodiscs make for a better membrane mimetic. The identification of key residues mediating the cytb5 – cyt c interaction provides important insights into the residues that drive the protein-protein complex formation. The combination of membrane systems and methods utilized in this study are also promising approaches to tackle the structural details of the more physiologically relevant cytb5 – cytP450 and other membrane-bound electron transfer complexes.
66
2.7 References 1. Vergeres, G., and Waskell, L. Cytochrome b5, its functions, structure and membrane topology. Biochimie 77, 604-620 (1995). 2. Schenkman, J. B., and Jansson, I. The many roles of cytochrome b5. Pharmacol. Ther. 97, 139-152 (2003). 3. Chaudhary, K. R., Batchu, S. N., and Seubert, J. M. Cytochrome P450 enzymes and the heart. IUBMB Life 61, 954-960 (2009). 4. O'Donnell, A. et al. Hormonal impact of the 17alpha-hydroxylase/C(17,20)-lyase inhibitor abiraterone acetate (CB7630) in patients with prostate cancer. Brit. J. Cancer 90, 2317- 2325 (2004). 5. Orlando, L.et al. Molecularly targeted endocrine therapies for breast cancer. Cancer Treat. Rev. 36, Suppl 3 S67-71 (2010). 6. Im, S. C., and Waskell, L. The interaction of microsomal cytochrome P450 2B4 with its redox partners, cytochrome P450 reductase and cytochrome b(5). Arch. Biochem. Biophys. 507, 144-153 (2011). 7. Finn, R. D.et al. Defining the in vivo role for cytochrome b5 in cytochrome P450 function through the conditional hepatic deletion of microsomal cytochrome b5. J. Biol. Chem. 283, 31385-31393 (2008). 8. Zhang, H., Myshkin, E., and Waskell, L. Role of cytochrome b5 in catalysis by cytochrome P450 2B4. Biochem. Biophys. Res. Commun. 338, 499-506 (2005). 9. McLaughlin, L. A., Ronseaux, S., Finn, R. D., Henderson, C. J., and Roland Wolf, C. Deletion of microsomal cytochrome b5 profoundly affects hepatic and extrahepatic drug metabolism. Mol. Pharmacol. 78, 269-278 (2010). 10. Durr, U. H., Waskell, L., and Ramamoorthy, A. The cytochromes P450 and b5 and their reductases--promising targets for structural studies by advanced solid-state NMR spectroscopy. Biochim. Biophys. Acta 1768, 3235-3259 (2007). 11. Pearl, N. M. et al. Protonation of the Hydroperoxo Intermediate of Cytochrome P450 2B4 Is Slower in the Presence of Cytochrome P450 Reductase Than in the Presence of Cytochrome b5. Biochemistry 55, 6558-6567 (2016).
67
12. Zhang, H., Im, S.C., and Waskell, L. Cytochrome b5 increases the rate of product formation by cytochrome P450 2B4 and competes with cytochrome P450 reductase for a binding site on cytochrome P450 2B4. J. Biol. Chem. 282, 29766-29776 (2007). 13. Williams, P. A., Cosme, J., Sridhar, V., Johnson, E. F., and McRee, D. E. Mammalian microsomal cytochrome P450 monooxygenase: structural adaptations for membrane binding and functional diversity. Mol. Cell 5, 121-131 (2000). 14. Durr, U. H. N., Yamamoto, K., Im, S. C., Waskell, L., and Ramamoorthy, A. Solid-state NMR reveals structural and dynamical properties of a membrane-anchored electron- carrier protein, cytochrome b(5). J. Am. Chem. Soc. 129, 6670 (2007). 15. Ren, Y. et al. Mapping the electron transfer interface between cytochrome b5 and cytochrome c. Biochemistry 43, 3527-3536 (2004). 16. Shao, W., Im, S. C., Zuiderweg, E. R., and Waskell, L. Mapping the binding interface of the cytochrome b5-cytochrome c complex by nuclear magnetic resonance. Biochemistry 42, 14774-14784 (2003). 17. Mauk, A. G., Mauk, M. R., Moore, G. R., and Northrup, S. H. Experimental and theoretical analysis of the interaction between cytochrome c and cytochrome b5. J. Bioenerg. Biomembr. 27, 311-330 (1995). 18. Deep, S., Im, S. C., Zuiderweg, E. R., and Waskell, L. Characterization and calculation of a cytochrome c-cytochrome b5 complex using NMR data. Biochemistry 44, 10654-10668 (2005). 19. Huang, R., Zhang, M., Rwere, F., Waskell, L., and Ramamoorthy, A. Kinetic and structural characterization of the interaction between the FMN binding domain of cytochrome P450 reductase and cytochrome c. J. Biol. Chem. 290, 4843-4855 (2015). 20. Bushnell, G. W., Louie, G. V., and Brayer, G. D. High-resolution three-dimensional structure of horse heart cytochrome c. J. Mol. Biol. 214, 585-595 (1990). 21. Vik, S. B., Georgevich, G., and Capaldi, R. A. Diphosphatidylglycerol is required for optimal activity of beef heart cytochrome c oxidase. Proc. Natl. Acad. Sci. U S A 78, 1456- 1460 (1981). 22. Hom, K. et al. NMR studies of the association of cytochrome b5 with cytochrome c. Biochemistry 39, 14025-14039 (2000).
68
23. Wendoloski, J. J., Matthew, J. B., Weber, P. C., and Salemme, F. R. Molecular dynamics of a cytochrome c-cytochrome b5 electron transfer complex. Science 238, 794-797 (1987). 24. Qian, C. et al. Effects of charged amino-acid mutation on the solution structure of cytochrome b(5) and binding between cytochrome b(5) and cytochrome c. Protein Sci. 10, 2451-2459 (2001). 25. Strittmatter, P. Reversible direct hydrogen transfer from reduced pyridine nucleotides to cytochrome b5 reductase. J. Biol. Chem. 239, 3043-3050 (1964). 26. Salemme, F. R. An hypothetical structure for an intermolecular electron transfer complex of cytochromes c and b5. J. Mol. Biol. 102, 563-568 (1976). 27. Eley, C. G., and Moore, G. R. 1H-n.m.r. investigation of the interaction between cytochrome c and cytochrome b5. Biochem. J. 215, 11-21 (1983). 28. Volkov, A. N., Ferrari, D., Worrall, J. A., Bonvin, A. M., and Ubbink, M. The orientations of cytochrome c in the highly dynamic complex with cytochrome b5 visualized by NMR and docking using HADDOCK. Protein Sci. 14, 799-811 (2005). 29. Stayton, P. S., Fisher, M. T., and Sligar, S. G. Determination of cytochrome b5 association reactions. Characterization of metmyoglobin and cytochrome P-450cam binding to genetically engineered cytochromeb5. J. Biol. Chem. 263, 13544-13548 (1988). 30. Chiang, J. Y. Interaction of purified microsomal cytochrome P-450 with cytochrome b5. Arch. Biochem. Biophys. 211, 662-673 (1981). 31. Bendzko, P., Usanov, S. A., Pfeil, W., and Ruckpaul, K. Role of the hydrophobic tail of cytochrome b5 in the interaction with cytochrome P-450 LM2. Acta Biol. Med. Ger. 41, K1-K8 (1982). 32. Canova-Davis, E., and Waskell, L. The identification of the heat-stable microsomal protein required for methoxyflurane metabolism as cytochrome b5. J. Biol. Chem. 259, 2541-2546 (1984). 33. Zhang, M., Huang, R., Im, S. C., Waskell, L., and Ramamoorthy, A. Effects of membrane mimetics on cytochrome P450-cytochrome b5 interactions characterized by NMR spectroscopy. J. Biol. Chem. 290, 12705-12718 (2015). 34. Zhang, M. et al. Reconstitution of the cytb5-cytP450 complex in nanodiscs for structural studies using NMR spectroscopy. Angew. Chem. Int. Ed. Engl. 55, 4497-4499 (2016).
69
35. Kobayashi, H., Nagao, S., and Hirota, S. Characterization of the cytochrome c membrane- binding site using cardiolipin-containing bicelles with NMR. Angew. Chem. Int. Ed. Engl. 55, 14019-14022 (2016). 36. O'Brien, E. S., Nucci, N. V., Fuglestad, B., Tommos, C., and Wand, A. J. Defining the apoptotic trigger: the interaction of cytochrome c and cardiolipin. J. Biol. Chem. 290, 30879-30887 (2015). 37. Seddon, A. M., Curnow, P., and Booth, P. J. Membrane proteins, lipids and detergents: not just a soap opera. BBA-Biomembranes 1666, 105-117 (2004). 38. Durr, U. H., Gildenberg, M., and Ramamoorthy, A. The magic of bicelles lights up membrane protein structure. Chem. Rev. 112, 6054-6074 (2012). 39. Denisov, I. G., and Sligar, S. G. Nanodiscs for structural and functional studies of membrane proteins. Nat. Struct. Mol. Biol. 23, 481-486 (2016). 40. Schuler, M. A., Denisov, I. G., Sligar, S. G. Nanodiscs as a new tool to examine lipid- protein interactions. Methods Mol. Biol. 945, 415-433 (2013). 41. Kariyazono, H. et al. Formation of stable nanodiscs by bihelical apolipoprotein A-I mimetic peptide. J. Pept. Sci. 22, 116-122 (2016). 42. Zhang, M. et al. Insights into the role of substrates on the interaction between cytochrome b5 and cytochrome P450 2B4 by NMR. Sci. Rep. 5, 8392. (2016) 43. Ahuja, S. et al. A model of the membrane-bound cytochrome b5-cytochrome P450 complex from NMR and mutagenesis data. J. Biol. Chem. 288, 22080-22095 (2013). 44. Pandey, M. K. et al. Cytochrome-P450-cytochrome-b5 interaction in a membrane environment changes 15N chemical shift anisotropy tensors. J. Phys. Chem. B 117, 13851- 13860 (2013). 45. Vivekanandan, S., Ahuja, S., Im, S. C., Waskell, L., and Ramamoorthy, A. (1)H, (1)(3)C and (1)(5)N resonance assignments for the full-length mammalian cytochrome b(5) in a membrane environment. Biomol. NMR Assign. 8, 409-413 (2014). 46. Fejzo, J., Lepre, C. A., Peng, J. W., Su, M. S., Thomson, J. A., and Moore, J. M. Dynamic NMR studies of ligand-receptor interactions: design and analysis of a rapidly exchanging complex of FKBP-12/FK506 with a 24 kDa calcineurin fragment. Protein Sci. 5, 1917- 1921 (1996).
70
47. Bashir, Q., Scanu, S., and Ubbink, M. Dynamics in electron transfer protein complexes. Febs J. 278, 1391-1400 (2011). 48. Schilder, J., Lohr, F., Schwalbe, H., and Ubbink, M. The cytochrome c peroxidase and cytochrome c encounter complex: The other side of the story. Febs Lett. 588, 1873-1878 (2014). 49. Guan, J. Y. et al. An ensemble of rapidly interconverting orientations in electrostatic protein-peptide complexes characterized by NMR spectroscopy. Chembiochem. 15, 556- 566 (2014). 50. Liu W., Rumbley J., Englander S.W., and Wand A.J. Backbone and side-chain heteronuclear resonance assignments and hyperfine NMR shifts in horse cytochrome c. Protein Sci. 12, 2104-8 (2003). 51. van Zundert, G. C. P. et al. The HADDOCK2.2 webserver: User-friendly integrative modeling of biomolecular complexes. J. Mol. Biol. 428, 720-725 (2015). 52. Wassenaar et al., WeNMR: Structural Biology on the Grid. J. Grid. Comp., 10, 743-767 (2012). 53. Soong, R. et al. Proton-evolved local-field solid-state NMR studies of cytochrome b5 embedded in bicelles, revealing both structural and dynamical information. J. Am. Chem. Soc. 132, 5779-5788 (2010). 54. Datta, G. et al. Effects of increasing hydrophobicity on the physical-chemical and biological properties of a class A amphipathic helical peptide. J. Lipid Res. 42, 1096-1104 (2001). 55. Datta, G. et al. Bioenergetic programming of macrophages by the apolipoprotein A-I mimetic peptide 4F. Biochem. J. 467, 517-527 (2015). 56. Bridges, A., Gruenke, L., Chang, Y. T., Vakser, I. A., Loew, G., and Waskell, L. Identification of the binding site on cytochrome P450 2B4 for cytochrome b5 and cytochrome P450 reductase. J. Biol. Chem. 273, 17036-17049 (1998). 57. Mulrooney, S. B., and Waskell, L. High-level expression in Escherichia coli and purification of the membrane-bound form of cytochrome b(5). Protein Expres. Purif. 19, 173-178 (2000). 58. Williamson, R. A., Carr, M. D., Frenkiel, T. A., Feeney, J., and Freedman, R. B. Mapping the binding site for matrix metalloproteinase on the N-terminal domain of the tissue
71
inhibitor of metalloproteinases-2 by NMR chemical shift perturbation. Biochemistry 36, 13882-9 (1997). 59. Williamson, M. P. Using chemical shift perturbation to characterize ligand binding. Prog. Nucl. Magn. Reson. Spectrosc. 73, 1-16 (2013). 60. Maeder, M., and Neuhold, Y.M. Practical Data Analysis in Chemistry, Elsevier (2007). 61. Schüttelkopf, A. W. & van Aalten, D. M. F. PRODRG: a tool for high-throughput crystallography of protein-ligand complexes. Acta Crystallogr. D. Biol. Crystallogr. 60, 1355-1363 (2004). 62. Kurnikov, I. V. Department of Chemistry, University of Pittsburgh; Pittsburg, PA: 2000. HARLEM molecular modeling package.
72
CHAPTER THREE
A Minimal Functional Complex of Cytochrome P450 and FBD of Cytochrome P450 Reductase
in Nanodiscs
3.1 Summary
Structural interactions that enable electron transfer to cytochrome-P450 (CYP450) from its redox partner CYP450-reductase (CPR) are a vital prerequisite for its catalytic mechanism. The first structural model for the membrane-bound functional complex to reveal interactions between the full-length CYP450 and a minimal domain of CPR is now reported. The results suggest that anchorage of the proteins in a lipid bilayer is a minimal requirement for CYP450 catalytic function.
Akin to cytochrome-b5 (cyt-b5), Arg125 on the C-helix of CYP450s is found to be important for effective electron transfer, thus supporting the competitive behavior of redox partners for
CYP450s. A general approach is presented to study protein–protein interactions combining the use of nanodiscs with NMR spectroscopy and SAXS. Linking structural details to the mechanism will help unravel the xenobiotic metabolism of diverse microsomal CYP450s in their native environment and facilitate the design of new drug entities.
1This chapter is based on the published paper: Prade, E., Mahajan, M., Im, S.-C., Zhang, M., Gentry, K.A., Anantharamaiah, G.M., Waskell, L., Ramamoorthy, A. (2018) A Minimal Functional Complex of Cytochrome P450 and FBD of Cytochrome P450 Reductase in Nanodiscs. Angew. Chem. Int. Ed. 57, 8458-8462. 2This thesis research was supported by funds from the National Institutes of Health (NIH to A.R.). 3Author Contributions: The study was planned by E.P., K.A.G., and A.R. K.A.G., M.Z. and E.P. expressed and purified all proteins. E.P., M.M., M. Z. and K.A.G. performed NMR experiments. M.M. performed the SAXS experiments. S.I. performed the stopped- flow experiments, and results were interpreted by S.I. and L.W. E.P., M.M. and A.R. wrote the manuscript, and all authors read and approved it. A.R. directed the project.
73
3.2 Introduction
The marked ability of CYP450 for biosynthesis, interconversion, and efficient metabolism of steroids, vitamins, fatty acids, and drugs makes it a vital target to fight various diseases.[1–4] Its catalytic efficiency depends on its interaction with either of its redox partners CPR or cyt-b5, which is a rate-limiting step involving electron transfer.[5] During its catalytic cycle, CYP450 sequentially receives two electrons. CPR must provide the first electron, whereas the second may originate from either CPR or cyt-b5. With few exceptions, [6,7] most structural information about CYP450 and CPR are based on crystal structures of their respective truncated soluble domains.[8–16]
Intriguingly, the crystal structure of the FMN (flavin mononucleotide) binding domain (FBD) of
CPR, which lacks the transmembrane domain (truncated-FBD, tr-FBD), is identical to the FBD segment in full-length CPR,[17,18] although it lacks activity.[19] Therefore, membrane counterparts of these proteins need to be considered to study their catalytic interaction, which is achieved by the use of nanodiscs in this study. Furthermore, the presence of a lipid membrane has also been considered to be a crucial prerequisite to facilitate the access of hydrophobic ligands to the active site of CYP450s.[20] Earlier work from our lab and other research groups have demonstrated the use of small peptide-based nanodiscs, which have facilitated the successful investigation of structure and dynamics of membrane proteins by solution-[21–27] and solid-state[28] NMR experiments. Their unique ability to accommodate multiple proteins in a natively folded functional state inside the lipid bilayer allows for a better understanding of protein–protein interactions.[29]
Additionally, detergent-free reconstitution of sensitive CYP450s and related proteins increases their stability from few days to weeks.[27] In this study, we extend this approach to study the interaction of CYP4502B4 (CYP2B4) with its redox partner CPR. We demonstrate that our system
74 allows for functional characterization of protein-protein interactions in the complex and provides novel insights into the mechanisms of electron transfer.
Figure 3.1: Expression of full-length FBD with transmembrane domain. A) SDS-PAGE of tr-FBD (21 kDa) and fl-FBD (28 kDa). B) ESI mass spectrometry data for 15N fl-FBD.
Figure 3.2: Reconstituted fl-FBD in nanodiscs are stable and well-folded. A) Overlapped 2D 1H-15N TROSY-HSQC NMR spectra of fl-FBD reconstituted in nanodiscs (red) and fl-FBD in solution (black). Inset: 1D 1H NMR spectra of fl-FBD in nanodiscs recorded immediately (black) and after 12 days (gray), demonstrating the stability of the sample. B) Secondary structural elements based on the comparison of experimentally measured Cα chemical shift values to the random coil Cα chemical shift values (ΔCα).
75
3.3 Results
Combining the use of peptide (4F)-lipid (DMPC) nanodiscs and full-length proteins with high-resolution solution NMR spectroscopy, we set out to structurally characterize FBD as well as the complex formed between native CYP2B4 and FBD, each containing their N-terminal transmembrane (TM) domain (fl-CYP2B4 and fl-FBD, respectively), taking their functionally vital membrane-supported binding nature into account. The individually purified proteins (Figure
76
3.1) were incubated with nanodiscs in a stepwise manner to form a functional complex suitable for structural analyses. Indeed, nanodiscs efficiently monomerize fl-FBD from its otherwise aggregated state in solution, resulting in 1H-15N-TROSY-HSQC NMR spectra featuring remarkable resolution of a well-folded protein (Figure 3.2a). The 1H-15N-TROSY-HSQC of the fl-
FBD in nanodiscs shows significant resemblance to tr-FBD in solution published previously,[30] indicating that the overall fold of the soluble domain is not affected by the presence of the TM domain nor the membrane. A similar dynamic structural property has been observed for other membrane-anchored proteins such as Bcl-xL and cyt-b5.[31,32] However, we observe approximately
55 additional peaks for fl-FBD, accounting for the N-terminal transmembrane domain, and several chemical shift perturbations (CSPs) towards the N-terminus of the truncated protein, indicative of protein reconstitution in the nanodisc. The full-length sample shows significant stability in nanodiscs (Figure 3.2a inset), facilitating acquisition of 3D NMR spectra. The enhanced signal dispersion and signal-to-noise ratio enabled sequential assignment of 83% of fl-FBD residues, including a large portion of the N-terminal domain (Figure 3.3a). A three-dimensional structural model for the (soluble domain of) membrane-anchored fl-FBD in lipid bilayer was generated using
NMR based chemical shifts and the CS-ROSETTA server from the biological magnetic resonance databank [33] (Figure 3.4). The canonical fold of fl-FBD in nanodiscs was verified by prediction of secondary structural elements calculated from chemical shifts, which are highly comparable to those reported [34] (Figure 3.2b; Figure 3.3b). Superimposed X-ray and the solution structural model of FBD show that the overall Rossmann fold (alternating β-strand with α-helical segments) of FBD is well-conserved (Figure 3.4). Structural investigation of FBD has to-date been limited to
NMR spectroscopic and crystallographic studies of its truncated soluble domain. The restrictions
77 of this approach are demonstrated when overlaying NMR and X-ray structures of FBD (Figure
3.4), which result in relatively large r.m.s.d values (ca. 3.0 Å).
Figure 3.3: Assignment of fl-FBD residues. A) Representative strips selected from 3D HNCACB spectra of tr-FBD in solution (black) and fl-FBD in nanodiscs (red). Peaks originating from Cα and Cβ resonances are indicated in blue and green, respectively. [63] B) Upper panel: Secondary structural elements as predicted by TALOS+ . Lower panel: Prediction of the transmembrane [75] sequence based on TMHMM, a membrane protein topology prediction server . The indicated secondary structural elements are highly comparable to those reported in the crystal structure of tr-FBD (PDB: 4YAF) [76].
78
Figure 3.4: CS-Rosetta calculated structure of the soluble domain of fl-FBD. A) An ensemble of 10 lowest energy structural models determined using CS-ROSETTA showing the well-conserved Rossman fold. Right, Superimposed X-Ray (cyan) and NMR (red) structures with r.m.s.d. values of ~3.0 Å. B) Ramachandran plot generated by PROCHECK [77,78] validation software showing the [33] stereochemical quality of CS-ROSETTA derived structural model of the soluble domain of fl-FBD anchored in nanodiscs.
79
To study the native protein–protein interactions occurring between the two redox partners, a functional complex was established in a membrane environment (Figure 3.5). Successful reconstitution of fl-FBD and fl-CYP2B4 peptide-based nanodiscs is supported by dynamic light scattering (DLS) and size exclusion chromatography (SEC) measurements (Figure 3.5a,b). At the optimized 1:1.5 w/w peptide:lipid ratio, the obtained empty peptide nanodiscs feature hydrodynamic radii of 4.45±0.40 nm in DLS measurements, which is in excellent agreement with the radius of gyration determined from small-angle X-ray scattering (SAXS) data (4.49±0.07 nm;
Figure 3.6). Despite the significant molecular weight of nanodiscs (ca. 124.5 kDa), the hydrodynamic radii of aggregated protein assemblies are reduced upon insertion into the lipid bilayer. Incubation of fl-CYP2B4 followed by fl-FBD resulted in a stepwise increase of the constructs, indicating successful incorporation of both proteins into the membrane (Figure 3.5a,b).
The relative ratio of reconstituted full-length proteins in the CYP2B4-FBD complex was estimated to be 0.95 using UV/Vis spectroscopy. Correct orientation of the two proteins in the complex is supported by SAXS measurements (Figure 3.5c,d). The linear Guinier region from experimental scattering curves of empty and protein-loaded nanodiscs demonstrate sample uniformity (Figure
3.6a,b). The maximum dimension (Dmax) of nanodiscs and the membrane reconstituted redox complex (fl-CYP2B4-fl-FBD) was calculated using pair-distance distribution function with
[35] GNOM module in PRIMUS. The Dmax was found to increase from 114 to 197.64 Å, which is attributed to the association of the soluble domains of the fl-CYP2B4-fl-FBD complex with the nanodiscs (Figure 3.6c,d). An ab initio model of the redox-CYP450 complex anchored in lipid nanodiscs (reddish brown) was reconstructed using DAMMIN/DAMMIF module in Primus from
ATSAS package[35] (Figure 3.5d). The bell-shaped curve of nanodisc anchoring fl-CYP450-fl-
80
FBD complex in normalized Kratky plot demonstrates the presence of a well-folded redox- complex in membrane (Figure 3.6 f).
Figure 3.5: Reconstitution of the fl-FBD and fl-CYP2B4 redox complex in peptide-based nanodiscs. A) DLS (left) and SEC measurements (right) of empty nanodiscs (black), fl-FBD incubated with nanodiscs (red) and fl-FBD in solution (gray) demonstrate successful reconstitution of fl-FBD in nanodiscs. B) fl-CYP2B4 reconstituted in nanodiscs (green), the complex between fl-FBD and fl-CYP2B4 reconstituted in nanodiscs (blue), and empty nanodiscs (black). C) Overlapped experimental scattering curves of empty (black) and fl-CYP2B4-fl-FBD complex containing nanodiscs (purple) from SAXS data. D) An ab initio reconstruction of the fl-FBD-fl-CYP2B4 complex anchored in lipid nanodiscs (reddish brown) using the DAMMIN/DAMMIF module in Primus from ATSAS package.[35] A low-resolution molecular envelope is superimposed with the docked structure of the fl-FBD-fl- CYP2B4 complex. E) Transfer of the second electron from hydroquinone fl-FBD to oxyferrous fl-CYP2B4 in nanodiscs under anaerobic conditions can be monitored by the increase of absorbance at 585 nm (fl-FBD hydroquinone to semiquinone transfer, red), as well as the decrease at 438 nm (fl-CYP2B4 reduction of oxyferrous state, blue).
To unambiguously ensure catalytic activity of the membrane-embedded complex, we used stopped-flow data to monitor the reduction of oxyferrous to ferric fl-CYP2B4 upon electron transfer from FMN under anaerobic conditions (Figure 3.5e; Figure 3.7). An increase in absorbance at 585 nm (fl-FBD hydroquinone to semi-quinone) and a decrease at 438 nm (CYP2B4 reduction of oxyferrous state) were observed as a result of the second electron transfer from hydroquinone fl-FBD to oxyferrous fl-CYP2B4. In the full-length CYP2B4-FBD complex, fl-FBD oxidizes rapidly (3.7 s-1), which is attributed to electron transfer from fl-FBD to fl-CYP2B4
(Figure 3.7). The kinetic traces of the individual proteins in nanodiscs are shown in the Figure 3.7,
81 along with the rates of oxidation of fl-CYP2B4 (0.09 s-1) and fl-FBD (1.5 s-1). In the presence of fl-FBD, fl-CYP2B4 oxidizes more rapidly (ca. 19 s-1 and 0.33 s-1) owing to the reduction of oxyferrous fl-CYP2B4 by fl-FBD; the rapid electron transfer (ca.19 s-1) process needs further investigation for a complete understanding of the process. After receiving an electron, fl-CYP450 undergoes catalysis and returns to the ferric protein. The slower oxidation of CYP2B4 in the complex than fl-FBD is expected because it has been shown that catalysis by fl-CYP2B4 proceeds via along-lived hydroperoxo intermediate in the presence of reductase.[5]
82
Figure 3.6: SAXS measurements and analysis of the fl-FBD fl-CYP2B4 complex in peptide-based nanodiscs. (A-B) SAXS curves in reciprocal (Fourier) space with linear Guinier region (inserts) demonstrate sample uniformity. Radii of gyration (Rg) calculated from Guinier plot of nanodiscs and reconstituted full-length redox complex (fl-FBD-fl-CYP2B4) are 44.95±0.79 and 67.98±1.52 Å, respectively. (C-D) Pair-distance distribution function (PDDF) of empty nanodiscs and the reconstituted redox complex obtained using indirect Fourier transform in GNOM module from ATSAS package. The region of negative contrast in empty nanodiscs is a characteristic property of hydrophobic acyl chains is higher. (E) Schematic representation of peptide-based nanodiscs (left) and overlapped ab initio model of peptide-based nanodisc (green) generated from SAXS data using the DAMMIN/DAMMIF module in Primus (right). (F) Normalized Kratky plot reveals a bell-shaped profile for the fl-FBD-fl-CYP2B4 complex reconstituted in nanodiscs. This experimental result further confirms the reconstitution of a properly folded protein-protein complex in peptide- based lipid nanodiscs. All SAXS data were analyzed using various modules of the ATSAS package [35}.
Mapping the interacting hot-spot region of the functionally active full-length redox complex (CYP2B4-FBD) significantly contributes to the design and development of novel drug molecules. The successful reconstitution of the uniformly 15N-labeled fl-FBD and unlabeled fl-
CYP2B4 complex in nanodiscs enabled us to probe the protein–protein binding interface by NMR
(Figure 3.7). Notably, the sample resulted in a well-dispersed spectrum featuring resolved peaks.
The Rossmann fold of FBD is unaffected by the presence of fl-CYP2B4, yet, the protein–protein interaction is reflected in the spectrum (Figure 3.7b). The FMN stabilizing anionic loop of FBD interacts with the cationic surface on the proximal side of CYP2B4. Small, yet relevant, chemical shift perturbations were observed for fl-FBD in the complex, in particular in the linker region
(S67–V70) and soluble domain (T88, Y117, L119, G143, K176, and H180) (Figure 3.7c; Figure
3.9a). Contact-induced conformational dynamics for residues Y117-L119 of fl-FBD was observed, although they do not seem to be directly involved in the interaction. Upon binding to fl-CYP2B4, an overall intensity decrease of fl-FBD resonances was observed (Figure3.7d; Figure 3.9b), indicating the complex formation and interaction between the proteins. Lipid–protein interactions have also been reported to play important roles in amyloid aggregation. [36,37] Most of the residues in the globular domain are affected by an exchange in an intermediate timescale. Addition of the solvent paramagnetic relaxation enhancement (sPRE) agent [Gd(DTPA-BMA)] provides further identification of residues buried in the interacting interface (Figure 3.7e; Figure 3.9c). Residues in the protein–protein interacting interface of the redox complex will not be quenched by the sPRE
83 agent, as they are no longer solvent-accessible. The difference in relative intensities between fl-
FBD and the fl-complex in the absence and presence of the sPRE agent (ΔQuenching) demonstrates which residues are recovered in the complex. The interacting interfacial residues of fl-FBD, T88, and G143, as well as residues 174–180 are protected from quenching in sPRE experiments. Several residues recovered from sPRE effects coincide with regions of the protein undergoing CSPs, especially the linker, as well as residues surrounding M137 and G143.
Figure 3.7: Probing hot spots for redox complex formation. A) Representation of complex formation. B) 1H-15N TROSY-HSQC of 15N fl-FBD in peptide-based nanodiscs in the absence (black) and presence (blue) of fl-CYP2B4. C) CSPs of fl-FBD shifted by 1 (green) and 2 (red) standard deviations of the mean upon binding to fl-CYP2B4. D) Line-broadening of fl-FBD resonances upon binding to fl-CYP2B4. E) sPRE data displaying the recovery of fl-FBD resonances by 1 standard deviation (red), which no longer undergo quenching once in the complex. F) Structural model of globular domain of full-length CYP2B4-FBD complex derive using HADDOCK simulations.[38] G) The shortest edge-to-edge distance between FMN and heme was calculated using HARLEM.[43] A yellow arrow marks the probable pathway of electron transfer.
A structural model of the full-length protein complex (CYP2B4-FBD) anchored in the membrane was derived using HADDOCK2.2.[38] NMR based CSPs and differential line broadening data were used as proximity restraints to guide the docking simulation. HADDOCK involves rigid-body docking, followed by molecular dynamics simulations that allow selected amino acid side chains, as well as parts of the backbone, to move freely to improve the complementarity and electrostatic interactions at the interface. The active ambiguous restraints for fl-CYP2B4 were obtained from published mutagenesis data [39] (Table 3.1). However, the interfacial residues on the CPR side were selected from CSPs and differential line broadening data
84 using fl-FBD. HADDOCK simulation resulted in 183 complex structures which covered about
91.5% of total structures based on energy statistics and better Z-scores (Table 2). An energetically minimized structure of the complex shows that the binding surface of fl-CYP2B4 covers mostly charged or hydrophobic residues, including R133, F135, M137, and K139, spanning the C–D loop along with some residues on other loops on the proximal surface of the heme of fl-CYP2B4. As shown in Figure 3.7f, g, both prosthetic groups are nearly perpendicular to each other (ca. 118.388° angle) with a shortest edge-to-edge distance of 7.1 Å, which is within the 14.0 Å limit predicted for electron transfer to occur.[40] The relative orientation of two cofactors in CPR (ΔTGEE) and rat HO-1 complex[41] is similar to our proposed model. Moreover, the crystal structure of bacterial
CYPBM3 complex revealed the distance between heme and FMN to be about 18 Å along with a similar orientation to our proposed model.[42] Both studies support the validity of our biologically active complex (CYP2B4-FBD) in a lipid bilayer.
Figure 3.8: Stopped-flow data for electron transfer from fl-FBD to fl-CYP2B4. The proteins were incubated individually (yellow and red for FBD and CYP2B4, respectively), in complex (orange) with nanodiscs and the absorption changes were monitored.
85
Figure 3.9: Representative NMR experiments of complex formation between cytP450 2B4 and fl-FBD. A) Representative excerpts displaying CSPs experienced by Y117 and H180 of fl-FBD in nanodiscs without (black) and with fl-CYP2B4 (blue). (B) Representative 1H-1D slices extracted from 1H-15N-TROSY-HSQC demonstrate attenuation of Q194 and preservation of E239 resonances from fl-FBD in nanodiscs without (black) and with fl-CYP2B4 (blue). (C) Schematic illustration of sPRE experimental setup; Gd (DTPA-BMA) (yellow) was added to fl-FBD or the fl-FBD-fl-CYP2B4 complex in nanodiscs. Furthermore, we predicted the electron-transfer pathway in the membrane-embedded protein complex using HARLEM.[43] The guanidinium group of R125 on the C-helix of fl-
CYP2B4 acts as a bridge to transfer an electron from FMN to the D-propionate of the heme (Figure
3.7g). The structural model for the truncated proteins complex is also feasible in solution,[44] which is possibly due to non-specific electrostatic interactions but lacks functional intent.[19]
Table 3.1: List of restraints used in HADDOCK simulations [38] for the full-length complex in nanodiscs. Active restraints on FBD site were obtained from TROSY based 15N-HSQC experiments on full-length complex reconstituted in lipid nanodiscs, whereas site directed mutagenesis data [39] was used for CYP2B4.
86
Table 3.2: Energy statistics of the lowest energy cluster of the FBD-CYP2B4 complex obtained using HADDOCK server [38].
3.4 Discussion
The structural perspective of different CYP450 isoforms reveals a common binding surface for CPR. Thus, amino acids critical for maintaining the structural conformation and biological function are often under evolutionary constraints and evolve slowly. As expected,
ConSurfanalysis[45,46] of CYP2B4 showed that the majority of conserved residues are located on the proximal side of the protein (Figure 3.10). Evolutionary conservational analysis also reveals that R125 is one of the most conserved residues on the CYP2B4 proximal surface, which is well- supported by earlier published reports on CYPcam, CYP241, and CYP2B4.[47–49] This ubiquitous presence of R125 in CYP450s may be responsible for the competitive nature of redox partners for binding CYP450s. In summary, we report the first successful atomic-resolution structural characterization of a minimal and fully functional CYP450-FBD complex anchored in lipid membrane. Apart from unambiguously confirming the functional nature of the complex, we depict the binding interface and propose an electron transfer pathway via R125. The use of a minimal
CPR domain (fl-FBD) imparts vital insights into the electron transfer process and can facilitate a better understanding of drug metabolism towards the design of therapeutics. While a simple model lipid bilayer is utilized in this study to overcome numerous challenges posed by the large-size of the membrane-bound functional complex, investigations of the roles of membrane composition[50]
87 and raft domain[51] that have recently been shown to stabilize CYP450 would be important to obtain further insights into the mechanism of electron-transfer process and metabolism by
CYP450.
Figure 3.10: Conservational analysis of fl-FBD and fl-CYP2B4. ConSurf analysis [45, 46} of fl-CYP2B4 based on the structural and functional importance of residues reveals conserved (purple-9) and variable residues (blue-1) present on proximal and distal sides of CYP450, respectively. R125 (ball and stick), located in the C-helix on the proximal side of fl-CYP2B4 is one of the most conserved residues and found to be important for electron transfer to heme.
3.5 Materials and Methods
3.5.1 Materials
Potassium phosphate (monobasic and dibasic) and benzphetamine were purchased from Sigma-
Aldrich (St. Louis, MO). 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) was purchased
15 from Avanti Polar Lipids, Inc. (Alabaster, AL). Deuterium oxide (D2O) and N Celtone Base
Powder were purchased from Cambridge Isotope Laboratories (Tewksbury, MA). The 5-mm
88 symmetrical D2O-matched Shigemi NMR microtubes were purchased from Shigemi, Inc (Allison
Park, PA).
3.5.2 Expression and purification of proteins
FBD of rat CPR was purified as described previously [30]. The fl-FBD gene is encoded on a pSC plasmid, and is preceded by the OmpA signal peptide [52]. Briefly, fl-FBD was expressed in E. coli
C41 cells in either LB medium for unlabeled protein, or M9 medium for U-15N or U-15N/13C labeled samples, supplemented with 5 nM FMN. Protein expression was induced at OD600 = 0.7 by adding 0.4 mM IPTG to the cultures for 16 h at 30 °C. After cell harvest at 6000 x g and 4 °C, the cells were lysed by 30 µg/ml lysozyme and protease inhibitor in Tris-Acetate buffer pH 7.4 for
30 mins at 4 °C, followed by sonication (with 1 s on and 1 s off pulses) for 5 mins. The membrane fraction was pelleted by ultracentrifugation at 105,000 x g and 4 °C for 45 mins, and further treated with 0.3% (v/v) Triton X-100 for 16 h at 4 °C. The solubilized membrane proteins were purified by DEAE anion exchange chromatography twice. For this purpose, the protein was loaded onto the column and eluted using a NaCl gradient ranging from 0.2 M to 0.5 M in Tris-Acetate pH 7.4 containing 1 µM FMN, 0.3% (v/v) sodium cholate. Full-length CYP2B4 was expressed and purified as described in the literature [53].
3.5.3 Reconstitution of full-length proteins in nanodiscs
Empty nanodiscs were produced from the 4F peptide and DMPC in 40 mM sodium phosphate pH
7.4. The 4F peptide was predissolved to a stock concentration of 10 mg/ml. DMPC was predissolved to 20 mg/ml and vortexed and sonicated to achieve a homogeneous suspension. The
4F peptide and DMPC were mixed at a 1:1.5 (w/w) ratio and incubated for 16 h at 37 °C with
89 gentle agitation. The nanodiscs were purified by size exclusion chromatography (SEC) using a
Superdex 200 Increase 300/10 GL column operated on an ÄKTA purifier (GE Healthcare,
Freiburg, Germany), and incubated with fl-FBD or fl-CYP2B4 at molar ratios of 1:1.2
(protein/nanodisc), and incubated for 16 h at 25 °C with gentle agitation. The reconstitution was further purified by SEC. Fractions showing absorbance at 412 nm (CYP2B4) or 454 nm (FBD) were pooled and used for further analysis. In order to form a complex, fl-FBD was added to purified CYP2B4 reconstituted in nanodiscs at a molar ratio of 1:1. The empty nanodiscs, proteins, and reconstituted proteins were subjected to dynamic light scattering (DLS) measurements on a
DynaPro NanoStar instrument (Wyatt Technology Corp., Santa Barbara, USA) at 25 °C for 10 acquisitions of 5 s each. DLS and SEC measurements confirm the increase of the hydrodynamic radius after stepwise incubation with fl-CYP2B4 and fl-FBD. The relative ratio of SEC purified reconstituted protein complex in nanodiscs was calculated using the absorbance at 417 nm and 454 nm for CYP2B4 and FBD, respectively.
3.5.4 SAXS measurements
SAXS data were acquired at the BioCAT beamline at Sector 18-ID of the APS in Argonne National
Laboratory, Chicago. Both empty nanodiscs and redox complex fl-CYP2B4-FBD anchored in 4F-
DMPC nanodiscs were prepared (as mentioned above) at 60 µM and ~20 µM, respectively, in 1 ml in buffer containing 40 mM potassium phosphate, pH 7.4. Both nanodiscs and reconstituted protein complex purified from size exclusion chromatography (SEC) were delivered using an autosampler with continuous unidirectional flow and recorded with 1 s exposure. Data was acquired as previously described elsewhere [54]. Scattering intensity plots (Figure 3.5c) are the average of triplicate experiments and subtracted with buffer. SAXS data was processed using the
90 program PRIMUS in ATSAS package [35]. The scattering curves were first analyzed for aggregation using the Guinier region. The forward scattering I(0) and the radius of gyration, Rg, were computed using the Guinier approximation. Rg provides a measure of the overall size of the macromolecule (Figure 3.6a). The pair distance distribution function P(r), was computed from the extended scattering patterns using the indirect transform program GNOM in PRIMUS. The maximum dimension of the particle, Dmax, estimated from the P(r) function satisfying the condition P(r)=0 (Figure 3.6c-d). The molecular folding and compactness of reconstituted full- length redox complex in nanodisc were analyzed by using the normalized Kratky plot (Figure
3.6f). Bell-shaped profile from scattering pattern in a normalized Kratky plot demonstrates the folded protein. Ab initio shape reconstruction of nanodiscs and redox complex (fl-CYP2B4-fl-
FBD) anchored in nanodiscs was obtained by DAMMIN/DAMMIF module in PRIMUS from
ATSAS package (Figures 3.5d and 3.6e).
3.5.5 NMR experiments
All NMR experiments were carried out at 298 K (25 °C). 3D HNCA, HNCACB and
CBCA(CO)NH [55-59] experiments were performed on a Bruker 900 MHz NMR spectrometer equipped with a triple-resonance cryoprobe (1H, 15N, 13C) on a sample containing 250 µM protein and 10% D2O. All other NMR experiments were carried out on a Bruker Avance II 600 MHz equipped with a triple-resonance cryoprobe (1H, 15N, 13C) at FBD concentrations of 100 µM. 2D
1H-15N TROSY HSQC [59, 60] spectra were recorded with 256 t1 increments.
Sequential assignment and analysis of CPSs, signal intensities was performed using CcpNmr
Analysis 2.4.1 [62]. Secondary structural elements were predicted using TALOS+ [63] and random
91 coil values as reported by Wishart et al [64]. Chemical shift perturbations (CSP) were calculated from 1H-15N TROSY HSQCs using the following equation:
퐹 푆푊 2 2 2 Δ훿푎푣𝑔 = √(Δ훿푁 × ) + Δ훿퐻 퐹1푆푊
To record solvent PRE (sPRE) data, 1 mM of the lanthanide [Gd(DTPA-BMA)] [65] was added to the sample and measured immediately. Residues of fl-FBD in the binding interface to fl-CYP2B4 will be protected from the sPRE singal quenching upon complex formation. This extend
(ΔQuenching) was determined by the following equation:
ΔQuenching = (complexsPRE/complexref)-(fl-FBDsPRE/fl-FBDref)
3.5.6 Structural model calculation
A structural model for the globular domain of FBD was calculated using the Chemical-Shift-
ROSETTA (CS-ROSETTA) server from the Biological Magnetic Resonance Data Bank. [33, 66-68]
CS-ROSETTA is a robust tool for de novo protein structure generation, using 13C, 15N and 1H
NMR chemical shifts as input. It employs SPARTA-based selection of protein fragments from the
PDB, in conjunction with a regular ROSETTA Monte Carlo assembly and relaxation procedure, to generate structures of minimized energies.
3.5.7 CYP450-FBD complex structure calculation
HADDOCK 2.2 webserver [38] was used to dock FBD and CYP450 based on a number of ambiguous restraints derived from NMR and site-directed mutagenesis experiments [39] (Table
3.1). It involves rigid-body docking, followed by molecular dynamics simulations that allow selected amino acid side chains, as well as parts of the backbone, to move freely to improve the
92 complementarity and electrostatic interactions at the interface. For this calculation, we used the X- ray structure of rat FBD (PDB structure: 1AMO [69]; Val64 to Ala-235 and the heme domain of
CYP2B4 (PDB code 1SUO [11]). Docking was performed using default parameters in the absence of a membrane environment. However, it must be noted that all NMR, SAXS and catalytic data were acquired on membrane anchored full-length proteins. Ligand topology and parameter files were generated from the PRODRG2 server [70]. NMR based CSPs and differential line broadening data were used as proximity restraints to guide the docking process whereby rigid body docking follows the semi-flexible refinement and energy minimization in explicit solvent to allow the free movement of backbone and side chain atoms of the selected amino acids to improve the intermolecular packing at protein interface. The active ambiguous restraints were chosen based on literature for CYP2B4 [39] and experimentally determined NMR restraints for FBD. Rigid body energy minimization was used for docking 1000 structures of the complex. The second step included semi-rigid simulated annealing from which the best 200 structures were selected for refinement. The best 200 structures were further refined with explicit solvent in an 8.0 Å shell of
TIP3P water molecules. Models are displayed using PyMOL [71].
3.5.8 Stopped-flow measurement of electron transfer from hydroquinone FMN to oxyferrous fl-
CYP2B4 reconstituted in 4F-peptide based lipid nanodiscs
All experiments were performed at 15 °C under anaerobic conditions using a Hi-Tech stopped- flow apparatus in a glove box ([O2] << 2ppm). All stock solutions (proteins and benzphetamine) were prepared in degassed and air saturated 40 mM potassium phosphate buffer pH 7.4. Degassed and air saturated buffers were prepared by purging nitrogen and air, respectively for 30 minutes.
Stock solutions of benzphetamine and protein were prepared at 1 mM and 3.5 mM, respectively in
93
40 mM potassium phosphate buffer pH 7.4. An equimolar amount of full-length proteins (CYP2B4 and FBD) in the nanodiscs was incubated overnight at 4°C in the glove box. After the addition of benzphetamine to the protein complex, the protein was stoichiometrically reduced with dithionite.
The stoichiometric reduction of the protein complex yielded ferrous CYP450, and the 2-electron reduced FBD (FMNH2) The 3-electron-reduced protein complex was loaded into the stopped-flow spectrophotometer and mixed with air saturated potassium phosphate buffer. Intermolecular electron transfer from FBD and to oxyferrous CYP450 was observed as oxidation of FBD at 585 nm.
3.5.9 Evolutionary conservation analysis
The ConSurf web server (http://consurf.tau.ac.il/2016/) [45, 46, 72-74] uses a query sequence or structure and analyzes the evolutionary pattern of the amino acids to reveal the regions important for structure and/or function.
3.6 References
1. A. L. Shen, K. A. O’Leary, C. B. Kasper, J. Biol. Chem. 2002, 277, 6536-6541.
2. D. W. Nebert, D. W. Russell, Lancet, 2002, 360, 1155-1162.
3. L. Orlando, P. Schiavone, P. Fedele, N. Calvani, A. Nacci, P. Rizzo, A. Marino, M. D’Amico,
F. Sponziello, E. Mazzoni, M. Cinefra, N. Fazio, E. Maiello, N. Silvestris, G. Colucci, S. Cinieri,
Cancer Treat. Rev. 2010, 36, S67-S71.
4. A. O’Donnell, I. Judson, M. Dowsett, F. Raynaud, D. Dearnaley, M. Mason, S. Harland, A.
Robbins, G. Halbert, B. Nutley, M. Jarman, Br. J. Cancer 2004, 90, 2317-2325.
94
5. N. M. Pearl, J. Wilcoxen, S.-C. Im, R. Kunz, J. Darty, R. D. Britt, S. W. Ragsdale, L. Waskell,
Biochemistry 2016, 55, 6558-6567.
6. B. C. Monk, T. M. Tomasiask, M. V. Keniya, F. U. Huschmann, J.D. Tyndall, J. D. O’Connell
3rd, R. D. Cannon, J.G. McDonald, A. Rodrigeuz, J. S. Finer-Moore, R. M. Stroud, Proc. Natl.
Acad. Sci. USA 2014, 111, 3865-3870.
7. D. Ghosh, J. Griswold, M. Erman, W. Pangborn, Nature 2009, 284, 36628-36637.
8. D. Hamdane, C. Xia, S.C. Im, H. Zhang, J. J. Kim, L. Waskell, J. Biol. Chem. 2009, 284, 11374-
11384.
9. J. Ellis, A. Gutierrez, I. L. Barsukov, W. C. Huang, J. G., Grossmann, G. C. Roberts, J. Biol.
Chem. 2009, 284, 36628-36637.
10. T. L. Poulos, B. C. Finzel, A. J. Howard, J. Mol. Biol. 1987, 195, 687-700.
11. E. E. Scott, Y. A. He, M. R. Wester, M. A. White, C. C. Chin, J. R. Halpert, E. F. Johnson, C.
D. Stout, Proc. Natl. Acad. Sci USA 2003, 100, 13196-13201.
12. J. K. Yano, M. H. Hsu, K. J. Griffin, C. D. Stout, E. F. Johnson, Nat. Struct. Mol. Biol. 2005,
12, 822-823.
13. N. Mast, M. A. White, I. Bjorkhem, E. F. Johnson, C. D. Stout, I. A. Pikuleva, Proc. Natl.
Acad. Sci. USA 2008, 105, 9546-9551.
14. A. Zhang, T. Zhang, E. A. Hall, S. Hutchinson, M. J. Cryle, L. L. Wong, W. Zhou, S. G. Bell,
Mol. Biosyst. 2015, 11, 869-881.
15. E. M. Petrunak, N. M. DeVore, P. R. Porubsky, E. E. Scott, J. Biol. Chem. 2014, 289, 32952-
32964.
16. E. E. Scott, M. A. White, Y. A. He, E. F. Johnson, C. D. Stout, J. R. Halpert, J. Biol. Chem.
2004, 279, 27294-27301.
95
17. A. V. Pandey, C. E. Fluck, Pharmacol. Ther. 2013, 138, 2290254.
18. Q. Zhao, S. Modi, G. Smith, M. Paine, P. D. McDonagh, C. R. Wolf, D. Tew, L. Y. Lian, G.
C. Roberts, H. P. Driessen, Protein Sci. 1999, 8, 298-306.
19. S. D. Black, J. S. French, C. H. Williams Jr., M. J. Coon, Biochem. Biophys. Res. Commun.
1979, 91, 1528-1535.
20. J. L. Baylon, I. L. Lenov, S. G. Sligar, E. Tajkhorshid, J. Am. Chem. Soc. 2013, 135, 8542-
8551.
21. F. Hagn, M. Etzkorn, T. Raschle, G. Wagner, J. Am. Chem. Soc. 2013, 135, 1919-1925.
22. M. Etzkorn, T. Raschle, F. Hagn, V. Gelev, A. J. Rice, T. Walz, G. Wagner, Structure 2013,
21, 394-401.
23. R. Puthenveetil, O. Vinogradova, Proteins Struc. Funct. Bioinf. 2013, 81, 1222-1231.
24. S. Bibow, M. G. Carneiro, T. M. Sabo, C. Schwiegk, S. Becker, R. Riek, D. Lee, Protein Sci.
2014, 23, 851-856.
25. D. A. Fox, P. Larsson, R. H. Lo, B. M. Kronchke, P. M. Kasson, L. Columbus, J. Am. Chem.
Soc. 2014, 136, 9938-9946.
26. Y. Ding, L. M. Fujimoto, Y. Yao, G. V. Plano, F. M. Marassi, Biochim. Biophys. Acta. 2015,
1848, 712-720.
27. M. Zhang, R. Huang, R. Ackermann, S.C., Im, L. Waskell, A. Schwendeman, A.
Ramamoorthy, Angew. Chem. Int. Ed. 2016, 55, 4497-4499; Angew. Chem. 2016, 128, 4573-4575.
28. K. Mors, C. Roos, F. Scholz, J. Wachtveitl, V. Dotsch, F. Bernhard, C. Glaubitz, Biochem.
Biophys. Acta Biomembr. 2013, 1828, 1222-1229.
29. T. Ravula, C. Barnaba, M. Mahajan, G. M. Anantharamaiah, S. C. Im, L. Waskell, A.
Ramamoorthy, Chem. Commun. 2017, 53, 12798-12801.
96
30. R. Huang, M. Zhang, F. Rwere, L. Waskell, A. Ramamoorthy, J. Biol. Chem. 2015, 290,
4843-4855.
31. C. Aisenbrey, U. S. Sudheendra, H. Ridley, P. Bertani, A. Marquette, S. Nedelkina, J. H.
Lakey, B. Bechinger, Eur. Biophys. J. 2007, 37, 71-80.
32. U. H. Durr, K. Yamamoto, S. C. Im, L. Waskell, A. Ramamoorthy, J. Am. Chem. Soc. 2007,
129, 6670-6671.
33. Y. Shen, R. Vernon, D. Baker, A. Bax, J. Biomol. NMR 2009, 43, 63-78.
34. F. Rwere, C. Xia, S.C. Im, M. M. Haque, D. J. Stuehr, L. Waskell, J. J. Kim, J. Biol. Chem.
2016, 291, 14639-14661.
35. D. Franke, M. V. Petoukhov, P. V. Konarev, A. Panjkovich, A. Tuukkanen, H. D. T.
Mertens, A. G. Kikhney, N. R. Hajizadeh, J. M. Franklin, C. M. Jeffries, D. I. Svergun, J. Appl.
Crystallogr. 2017, 50, 1212-1225.
36. M. Michalek, E. S. Salnikov, S. Werten, B. Bechinger, Biochemistry 2013, 52, 847-858.
37. J. R. Brender, S. Salamekh, A. Ramamoorthy, Acc. Chem. Res. 2012, 45, 454-462.
38. G. C. P. van Zundert, J. Rodrigues, M. Trellet, C Schmitz, P. L. Kastritis, E. Karaca, A. S. J.
Melquiond, M. van Dijk, S. J. de Vries, A. Bonvin, J. Mol. Biol. 2016, 428, 720-725.
39. A. Bridges, L. Gruenke, Y. T. Chang, I. A. Vakser, G. Loew, L. Waskell, J. Biol. Chem. 1998,
273, 17036-17049.
40. C. Page, Curr. Opin. Chem. Biol. 2003, 7, 551-556.
41. M. Sugishima, H. Sato, Y. Higashimoto, J. Harada, K. Wada, K. Fukuyama, M. Noguchi, Proc.
Natl. Acad. Sci. USA 2014, 111, 2524-2529.
42. I. F. Sevrioukova, H. Li, H. Zhang, J. A. Peterson, T. L. Poulos, Proc. Natl. Acad. Sci. USA
1999, 96, 1863-1868.
97
43. I. V. Kurnikov, HARLEM molecular modeling package, 2000.
44. D. F. Estrada, J. S. Laurence, E. E. Scott, J. Biol. Chem. 2016, 291, 3990-4003.
45. F. Glaser, T. Pupko, I. Paz, R. E. Bell, D. Bechor-Shental, E. Martz, N. Ben-Tal,
Bioinformatics 2003, 19, 163-164.
46. M. Landau, I. Mayrose, Y. Rosenberg, F. Glaser, E. Martz, T. Pupko, N. Ben-Tal, Nucleic
Acids Res. 2005, 33, W299-W302.
47. K. Nakamura, T. Horiuchi, T. Yasukochi, K. Sekimizu, T. Hara, Y. Sagara, Biochim.
Biophys. Acta Protein Struct. Mol. Enzymol. 1994, 1207, 40-48.
48. K. P. Schlingmann, M. Kaufmann, S. Weber, A. Irwin, C. Goos, U. John, J. Misselwitz, G.
Klaus, E. Kuwertz-Broking, H. Fehrenbach, A. M. Wingen, T. Guran, J. G. Hoenderop, R. J.
Bindels, D. E. Prosser, G. Jones, M. Konrad, N. Engl. J. Med. 2011, 365, 410-421.
49. S. Ahuja, N. Jahr, S.C. Im, S. Vivekanandan, N. Popovych, S. V. Le Clair, R. Huang, R.
Soong, J. Xu, K. Yamamoto, R. P. Nanga, A. Bridges, L. Wakell, A. Ramamoorthy, J. Biol.
Chem. 2013, 288, 22080-22095.
50. C. Barnaba, K. Gentry, N. Sumangala, A. Ramamoorthy, F1000Res 2017, 6, 662.
51. C. Barnaba, B. R. Sahoo, T. Ravula, I. G. Medina-Meza, S. C. Im, G. M. Anantharamaiah, L.
Waskell, A. Ramamoorthy, Angew. Chem. Int. Ed. 2018, 57, 3391-3395; Angew. Chem. 2018,
130, 3449-3453.
52. A. L. Shen, T. D. Porter, T. E. Wilson, C. B. Kasper, J. Biol. Chem. 1989, 264, 7584-7589.
53. A. S. Saribas, L. Gruenke, L. Waskell, Protein Expr. Purif. 2001, 21, 303-309.
54. A. W. Malaby, S. Chakravarthy, T. C. Irving, S. V. Kathuria, O. Bilsel, D. G. Lambright, J.
Appl. Crystallogr. 2015, 48, 1102-1113.
55. S. Grzesiek, A. Bax, J. Magn. Reson.1992, 96, 432-440.
98
56. S. Grzesiek, A. Bax, J. Am. Chem. Soc. 1992, 114, 6291-6293.
57. S. Grzesiek, A. Bax, J. Magn. Reson. 1992, 99, 201-207.
58. M. Sattler, Prog. Nucl. Magn. Reson. Spectrosc. 1999, 34, 93-158.
59. L. E. Kay, M. Ikura, R. Tschudin, A. Bax, J. Magn. Reson. 1990, 89, 496- 514.
60. K. Pervushin, R. Riek, G. Wider, K. Wuthrich, Proc. Natl. Acad. Sci. U. S. A. 1997, 94,
12366-12371.
61. C. Fernandez, G. Wider, Curr. Opin. Struct. Biol. 2003, 13, 570-580.
62. W. F. Vranken, W. Boucher, T. J. Stevens, R. H. Fogh, A. Pajon, M. Llinas, E. L. Ulrich, J.
L. Markley, J. Ionides, E. D. Laue, Proteins 2005, 59, 687-696.
63. Y. Shen, F. Delaglio, G. Cornilescu, A. Bax, J. Biomol. NMR 2009, 44, 213-223.
64. D. S. Wishart, C. G. Bigam, J. Yao, F. Abildgaard, H. J. Dyson, E. Oldfield, J. L. Markley,
B. D. Sykes, J. Biomol. NMR 1995, 6, 135-140.
65. H. G. Hocking, K. Zangger, T. Madl, Chemphyschem 2013, 14, 3082- 3094.
66. O. F. Lange, P. Rossi, N. G. Sgourakis, Y. F. Song, H. W. Lee, J. M. Aramini, A. Ertekin, R.
Xiao, T. B. Acton, G. T. Montelione, D. Baker, Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 10873-
10878.
67. Y. Shen, P. N. Bryan, Y. He, J. Orban, D. Baker, A. Bax, Protein Sci. 2010, 19, 349-356.
68. Y. Shen, O. Lange, F. Delaglio, P. Rossi, J. M. Aramini, G. Liu, A. Eletsky, Y. Wu, K. K.
Singarapu, A. Lemak, A. Ignatchenko, C. H. Arrowsmith, T. Szyperski, G. T. Montelione, D.
Baker, A. Bax, Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 4685-4690.
69. M. Wang, D. L. Roberts, R. Paschke, T. M. Shea, B. S. Masters, J. J. Kim, Proc. Natl. Acad.
Sci. U. S. A. 1997, 94, 8411-8416.
99
70. A. W. Schuttelkopf, D. M. van Aalten, Acta. Crystallogr. D Biol. Crystallogr. 2004, 60,
1355-1363.
71. Schrodinger, LLC, The PyMOL Molecular Graphics System, Version 1.3r1, 2010.
72. H. Ashkenazy, S. Abadi, E. Martz, O. Chay, I. Mayrose, T. Pupko, N. Ben-Tal, Nucleic
Acids Res. 2016, 44, W344-350.
73. H. Ashkenazy, E. Erez, E. Martz, T. Pupko, N. Ben-Tal, Nucleic Acids Res. 2010, 38, W529-
533.
74. G. Celniker, G. Nimrod, H. Ashkenazy, F. Glaser, E. Martz, I. Mayrose, T. Pupko, N. Ben-
Tal, Israel Journal of Chemistry 2013, 53, 199-206.
75. A. Krogh, B. Larsson, G. von Heijne, E. L. Sonnhammer, J. Mol. Biol. 2001, 305, 567-580.
76. F. Rwere, C. Xia, S. Im, M. M. Haque, D. J. Stuehr, L. Waskell, J. J. Kim, J. Biol. Chem.
2016, 291, 14639-14661.
77. R. A. Laskowski, M. W. MacArthur, D. S. Moss, J. M. Thornton, J. Appl. Crystallogr. 1993,
26, 283-291.
78. R. A. Laskowski, J. A. Rullmannn, M. W. MacArthur, R. Kaptein, J. M. Thornton, J. Biomol.
NMR 1996, 8, 477-486.
100
Chapter 4
Substrate Mediated Redox Partner Selectivity of Cytochrome P450
4.1 Summary
Investigating the interplay between cytochrome-P450 and its redox partners (CPR and
cytochrome-b5) is vital for understanding the metabolism of most hydrophobic drugs. Dynamic
structural interactions with the ternary complex, with and without substrates, captured by NMR
reveal a gating mechanism for redox partners to promote P450 function.
4.2 Introduction
Cytochrome P450s (P450s) are a ubiquitous superfamily of enzymes responsible for metabolizing a dazzling array of exogenous and endogenous compounds, including carcinogens, hormones, and over 75% of the drugs on the current market.1–5 For each turn of the catalytic cycle,
P450 requires two electrons which are delivered by P450’s two redox partners, cytochrome P450
6 reductase (CPR) and cytochrome b5 (cytb5). CPR and cytb5 have different capabilities of interacting with P450 and their unbalanced stoichiometry within the endoplasmic reticulum, make it an interesting and challenging question to understand the interplay between this ternary complex.
1This chapter is based on the published paper: Gentry, K.A., Zhang, M., Im, S.-C., Waskell, L., Ramamoorthy, A. (2018) Substrate mediated redox partner selectivity of cytochrome P450. Chem. Commun. 45, 5780-5783. 2This thesis research was supported by funds from the National Institutes of Health (NIH to A.R.).
101
CPR is a 78 kDa protein comprised of four major domains: a NADPH/FAD binding domain, a linker domain, a FMN binding domain (FBD), and a N-terminal transmembrane domain.5,7 After NADPH binding, electrons are shuttled through the FAD cofactor to the FMN cofactor to the heme group of the electron acceptor P450, a process involving both intra- and inter- protein electron transfers. The inter-protein electron transfer step depends on association and interaction between FBD and P450. Studies have demonstrated the feasibility of using the FBD alone to study the interaction between CPR and P450.8,9
The third protein of this ternary complex is cytb5, a 15.7 kDa protein, which is only capable
of donating the second electron to P450 due to the large difference in redox potential between
5,10,11 cytb5 and ferric P450. Although P450 metabolism can function with only CPR as the sole
electron donor, deletion of cytb5 causes extreme effects of P450 catalysis both sizeable increases
12,13 and decreases in metabolism rates. Through kinetic experiments, cytb5 has been shown to affect
2,14,15 the catalytic activity of over twenty P450 isoforms, including 3A4, 2B6, 2C9, and 2E1. Cytb5
seems to have some level of substrate or reaction preference, best seen through its influence of the
17,20-lyase reaction in P450 17a1.16,17 Various recent studies with microsomal P450 have helped
to elucidate that cytb5 can stimulate a reaction through increasing the rate of formation of the
6,18 catalytically active oxidizing species and product formation. Overall, the role of cytb5 in P450
metabolism is very perplexing, since it has been reported to stimulate, inhibit, or not affect P450
activities depending on the substrates involved, the particular isoform of P450, and the
experimental conditions.
In this study, we investigate the effect of substrates on the complex formation between
cytb5 and P450, FBD and P450, and the ternary complex, demonstrating the differential regulation
102
potential of substrates. Using NMR, we probe the relationship between rabbit full-length P450
2B4, FBD, and full-length cytb5 at residue specific resolution.
4.3 Results
4.3.1 Substrate effect on interaction in between cytb5 and cytP450
Cytb5’s effect on P450 metabolism is both isoform and substrate dependent. Differences in rates of metabolism of various drugs with the addition or absence of cytb5 have been noted for a variety of isoforms and substrates,14,19–22 but differences in complex formation strength has not been investigated which this study provides for the first time. We use NMR to monitor changes in
1 15 15 Figure 4.1: Substrate effect on cytb5-P450 interaction. (A) 2D H- N TROSY-HSQC NMR spectrum of N-labeled cytb5. (B) Comparison of the average intensity of cytb5 in complex with P450 in the absence and presence of different substrates. Highlighted (in blue) residues are part of the P450-cytb5 binding interface are mapped on cytb5 (PDB: 2M33) with the heme group displayed in red sticks for (C) benzphetamine (BZ), (C) methoxyflurane (MF), and (D) cyclohexane (CH).
103
23 complex formation between P450 2B4 and cytb5. As previously described, the substrate butylated hydroxytoluene (BHT) enhances the complex formation between cytb5 and P450, as measured by the increasing 15N-NMR line width. A similar strategy was applied in this study to test a variety of substrates: benzphetamine (BZ), methoxyflurane (MF), and cyclohexane (CH). 15N-Labeled rabbit cytb5 was expressed, purified, and used to form a 1:1 complex with P450. 2D-TROSY-HSQC
NMR experiments were used to monitor changes in complex formation as each substrate was titrated into the protein–protein complex and the line-broadening of cytb5 residues was monitored.
One measure of the complex stability (or the strength of complex formation) is through understanding the line-broadening of residues which can be measured by the overall average signal intensity change. Cytb5 alone would have 100% signal intensity, whereas cytb5 in complex with
P450 has a signal intensity of 70%.24 In general, all substrates used (BHT, BZ, MF, and CH) led to an increase in line-widths of cytb5 resonances (Fig. 4.1A). Substrates modulate this interaction to varying degrees with BHT strengthening the complex formation followed by BZ, and lastly CH and MF. Residues with relative intensity loss more than one standard deviation below average were considered to be significantly broadened upon binding to P450 and highly implicated to be part of the binding interface. These residues are mapped onto cytb5: H44, F63, E64, D65, T70, and
D71 in the presence of BZ (Fig. 4.1C); H44, D65, V66, T70, and D71 in the presence of cyclohexane (Fig. 4.1D); and I17, F40, R52, E64, D65, L75 for methoxyflurane (Fig. 4.1E). All the substrate-titrated samples demonstrated overlapping regions affected by the binding to P450,
23– mainly, the front face of cytb5 specifically on the lower cleft which is supported by the literature.
27
104
4.3.2 Effects of Substrates on FBD and cytP450
Next, to investigate the effect of substrates on the interaction between FBD and P450, substrates (BHT, BZ, MF) were added to 1:1 complex of 15N-uniformly labelled FBD and P450.
2D 1H–15N TROSY-HSQC NMR spectra were compared before and after the addition of a substrate. For all three substrates, chemical shift perturbations were negligible indicating the proteins are under fast exchange in the NMR time scale. All of the substrate additions led to very slight enhancement in line-broadening – seen as a decrease in signal intensity (Fig. 4.2). In general, substrates do not affect FBD–P450 2B4 interaction much implying that, within the scope of this study, complex formation between these two redox partners are not modulated by substrates.
Figure 4.2: Substrate effect on trFBD-cytP450 interaction. Comparison of the average relative intensity of 15N-labeled trFBD in complex with cytP450 in the absence and presence of different substrates.
4.3.3 FBD is unable to disrupt the complex of cytb5-cytP450
Not much is known about the interplay between the three proteins. Structurally, we know
2,6,8,16,17 the FBD of CPR and cytb5 share an overlapping binding surface on P450. As CPR/FBD
105
28 has a higher affinity for P450 than does cytb5, CPR/FBD has the potential to bind preferentially
to P450 in the presence of cytb5. The rate constants for the interprotein electron transfer reactions
are of the same magnitude for CPR-P450 and cytb5– P450 which eliminates the ability to transfer
an electron to P450 from being the driving force behind the choice of redox partner. One
hypothesis is that concentration is the major factor behind P450’s choice of redox partner where
the poorer binding redox partner (cytb5 in this case) would have to be present at a higher
concentration in order to be successful in outcompeting the stronger binding redox partner
(CPR).28 As these two proteins have unique, but overlapping binding sites on P450,2,6,22 this means
they are in direct competition and these two enzymes cannot bind to P450 at the same time.
Figure 4.3: Cytb5-P450 complex is unperturbed by interaction with CPR and its variants. 2D 1H-15N TROSY-HSQC NMR spectral regions of 15N-labeled cytb5 reveal changes due to interactions with P450, BHT, and CPR variants. (A) Free 15N-cytb5 (green), (B) 15N-cytb5 titrated with 1 molar equivalent of P450 (red), (C) 15N-cytb5 + P450 titrated with BHT (blue), and (D) 15N- cytb5 + P450 + BHT after FBD titration (black). (E) The average signal intensity of cytb5 residues (total residues in black; residues in the lower clef (60-70) in red) over the course of a titration with CPR, flFBD, or trFBD at the indicated molar ratios (0.5, 1, and 1.5). Negligible change in the observed signal intensity indicates cytb5 is not replaced from the cytb5-P450 complex by CPR or its variants. Due to the higher affinity favoring the complex formation of CPR/FBD and P450, we tested the ability of CPR/FBD to dislodge cytb5 from its complex with P450. We first formed a 1:1
15 complex between N-labeled cytb5 and P450, then titrated in CPR or FBD while 2D-TROSY-
106
HSQC NMR spectra were utilized to monitor signal intensity (Fig. 4.3). As CPR is titrated into the
1:1 cytb5:P450 complex, we should see an increase in overall signal intensity of cytb5 if CPR is able to disrupt the complex. Surprisingly, this effect was not seen. The cytb5:P450 complex was stable and unaffected even in the excess of CPR, meaning that CPR is unable to interrupt or hinder the cytb5–P450 interaction over the concentrations of 0.5 to 1.5 molar equivalents (Fig. 4.3E, black). This trend holds true for looking at the overall signal intensity as well as the intensity of cytb5 residues (i.e., 60–70) implicated in binding with P450 (Fig. 4.3E, red). This experiment was performed with CPR, FBD, and full-length FBD (flFBD); flFBD is FBD with the transmembrane domain which was used to rule out the influence of the transmembrane domain in this interaction.
Very similar results were obtained for flFBD and FBD demonstrating that no form of CPR is capable of disrupting the cytb5–P450 complex, and also proving the feasibility of working with
FBD to simplify this question. This experiment was repeated with a substrate, BHT, and similar results were seen with FBD unable to disrupt the cytb5–P450 complex. Fig. 4.3A–D demonstrates residue-specific change in signal intensity when binding to P450 or P450 + BHT, and upon titration with FBD. No significant change was observed for residues D65 and E74; while D65 has been identified in the binding interface with P450,23–27 E74 has not.
4.3.4 Cytb5 is capable of dislodging FBD from cytP450
As FBD was shown to be unable to disrupt the cytb5–P450 complex, we carried out the
“opposite” experiment by testing the ability of cytb5 to destabilize the FBD–P450 complex.
Uniformly 15N-labeled FBD was expressed, purified, and used to form a 1:1 complex with P450.
When cytb5 was titrated into this complex, the average signal intensity of FBD residues observed from TROSY-HSQC spectra greatly increased. Our results clearly show (Fig. 4.4) that cytb5 is capable of dislodging P450-bound FBD into free solution. It is important to clarify that the ratio
107 of cytb5 added was not in excess meaning that this phenomenon is not (solely at least) a concentration-based effect.
As we have shown, substrates can greatly enhance the interaction between cytb5 and P450 while only minimally affecting the complex between FBD and P450. Cytb5 has been shown to exert substrate-dependent effects on P450 catalysis. We performed the same experiment but we added various substrates (BHT, BZ, and MF) to the 1:1 FBD:P450 complex to ascertain if ligands affect cytb5’s ability to dislodge FBD from P450. Then cytb5 was added and the resulting change in FBD average signal intensity was quantified. Our results (Fig. 4.4) demonstrate a substrate dependent modulation of cytb5’s ability to disrupt complex formation. Fig. 4.4B–I displays residue
Figure 4.4: Cytb5 disrupts complex between FBD and cytP450 in a substrate dependent manner. 2D 1H-15N TROSY-HSQC NMR 15 1 15 15 spectra of N-labeled FBD reveal changes due to interaction with P450, substrates, and cytb5. (A) H- N TROSY-HSQC NMR spectrum of N-FBD. Example residues are zoomed in (B) 15N-FBD in solution (black), (C) 15N-FBD titrated with 1 molar equivalent of P450 (red), 15N-FBD+P450 titrated with a substrate BHT (blue, D), BZ 15 15 (magenta, F), or MF (purple, H), and N-FBD+P450+substrate titrated with cytb5 (green) for BHT (E), BZ (G), and MF (I). (J) Average signal intensity of N-FBD quantified after the addition of P450, P450+substrate (BHT, BZ, or MF), and P450+substrate (BHT, BZ, or MF) +cytb5.
108
G89, which has been implicated in FBD binding to P450,28 as well as S123 which is not involved.
The significant loss of G89’s peak intensity after titration with P450 or P450 + substrate and
restoration after cytb5 addition confirms our conclusion that cytb5 is capable of dislodging FBD
from the P450–FBD complex. The strongest effect was observed with BHT, followed by BZ and
MF. With BHT and BZ, 100% FBD was freed from the P450-bound state, while in the presence
of MF, FBD still remained partially bound to P450. This difference between various substrates
suggests some sort of mechanism that is regulated by different substrates. It has been reported in
the literature that cytb5 stimulates P450 activity to a much higher extent for MF than BZ: 8.5-fold
increase in activity versus 1.3-fold increase.6,29,30 This has been related to an increased coupling
6 efficiency due to electron transfer, as well as allosteric activation mediated by cytb5. In our
experiment, we see an inverse of this relation – that the interaction strength of the complex is much
higher for BZ over MF. Based on these results, we propose a competitive binding mechanism
where cytb5 and CPR are competing to bind to an overlapping, but unique binding site on P450
and substrates help to modulate this effect (Fig. 4.5).
Figure 4.5: Schematic of the ternary interplay. When cytb5 (purple) is added to the complex of P450 (yellow) and FBD (blue) in the presence of BHT, cytb5 is able to dislodge and replace FBD to form a complex with P450. On the other hand, when P450 is in a complex of cytb5, FBD is unable to remove cytb5 from the complex.
4.4 Discussion of substrate mediated effects
One explanation for this inverse relationship of metabolism increase and complex formation is that the first electron transfer to P450, which can only be fulfilled by CPR, may be inhibited due to cytb5 occupying the binding interface on P450. The addition of a substrate to P450 could increase cytb5’s affinity for P450 which allows it to outcompete CPR, which gives newfound
31 insight into past findings where cytb5 was found to inhibit BZ demethylation. A previous study
109 demonstrated that these proteins have an overlapping but unique binding site on P450, which indicates that the two enzymes cannot bind simultaneously. This can explain why cytb5’s stimulation effect on P450 activity changes for different substrates; each substrate could alter cytb5’s affinity for P450 in a different way. Other P450 isoforms have shown this substrate dependent increase in binding affinity, most recently, with CYP101D1’s binding affinity for its redox partner, Adx, after the addition of camphor.32 Another explanation is the concept of P450’s structural plasticity whereby after ligand binding P450 undergoes a conformational change.33 This conformational change could favor one redox partner binding over the other. The larger of the substrates that we tested, BHT and BZ, are bulky compounds which could have more of a rigidifying effect on the distal region on P450, leading to allosteric changes in the proximal, redox partner binding site of P450 whereas a smaller compound like MF has a lesser effect. While P450s are known to be highly homologous in their proximal site, their distal region is more varied which could explain the differences amongst isoforms of P450. Thus, the results presented in this study shed light on substrate regulation on the tertiary FBD–P450–cytb5 system, providing insights into the structural basis of the interplay of the three redox partners.
4.5 Conclusions
In conclusion, for the first time, we have shown that substrates play very important roles
in the dynamic interplay between P450 2B4 and its redox partners. Cytb5 interaction with P450 is
greatly increased by the addition of substrates to varying degrees whereas FBD–P450 interaction
is not significantly affected. Although FBD and CPR have higher affinity for P450, they are unable
to dislodge cytb5 from binding to P450. Cytb5, on the other hand, is capable of disrupting the FBD–
P450 complex interaction in a substrate dependent manner. Since all three proteins are membrane-
110 anchored and membrane has been shown to play important roles on P450 function,2,5,34 probing the roles of membrane on the ternary complex is in progress in our lab.
4.6 Materials and Methods
4.6.1 Materials and Reagents
E. coli C41 cells for protein overexpression were purchased from Lucigen (Middleton, MI).
Yeast extract, tryptone, for unlabeled growth media was purchased from Sigma-Aldrich. [15N]
15 ammonium chloride, [ N] CELTONE rich medium powder and D2O were purchased from
Cambridge Isotope Laboratories (Andover, MA). Resins, buffer components, and all the other chemicals were purchased from Sigma-Aldrich. Glycerol used in NMR experiments was purchased from Roche Applied Science.
4.6.2 Expression and purification of the soluble FMN binding domain of rat CPR.
The unlabeled and 15N labeled FBD as well as the full-length FBD was individually expressed and purified as described previously.[22] The U-15N-labeled FBD was expressed with unlabeled glucose and [15N] CELTONE rich medium powder. For expression of unlabeled FBD, the adapted cells were inoculated to 1 L LB medium at a starting OD600 value of 0.03 and induced at OD600 = 2. The purified FBD appeared as a single band on the SDS-polyacrylamide gel. The concentration of the oxidized FBD was determined using the extinction coefficients 12.2 mM-1cm-
1 at 454 nm.[35]
Full-length wild-type rabbit cytochrome P450 2B4 (cyt P450 2B4) and U-15N labeled full- length wild-type rabbit cyt b5 were expressed and purified individually as described previously.[22,27,36,37]
4.6.3 Solution NMR experiments. All solution NMR experiments were carried out at 298 K in NMR buffer (40 mM potassium
15 phosphate, 10% D2O, pH 7.4). All samples contained around 0.1 mM uniformly N-labeled cytb5
111 or 15N-labeled trFBD. All NMR data was processed by Topspin 2.1 (Bruker) and analyzed in
Sparky[38].
4.6.4 Substrate modulation on the interaction between cytb5 and cytP450 To investigate the effects of a range of different substrates on the interaction between cyt
[23] b5 and cyt P450 2B4, similar approaches were applied as detailed in Zhang et al., 2015 . Briefly, each substrate, including methoxyflurane, benzphetamine and cyclohexane, was added to a 1:1
15 1 15 N-cyt b5 : unlabeled cyt P450 2B4 complex sample. 2D H/ N TROSY HSQC spectra with 64 scans and 256 t1 increments.
4.6.5 Substrate modulation on competitive binding between cytb5 and trFBD in the trFBD- cytP450 -cytb5 tertiary protein system. 2D 1H/15N TROSY HSQC was first recorded on 15N-trFBD with one molar equivalence unlabeled cyt P450 2B4 in the absence or presence of three molar equivalence substrate methoxyflurane, benzphetamine, or cyclohexane. Then one molar equivalence of unlabeled cyt b5 was added to the sample, followed by acquisition of 2D 1H/15N TROSY HSQC with 32 scans and
256 t1 increments.
4.7 References
1. F. P. Guengerich, Mol. Interventions, 2003, 3, 194.
2. C. Barnaba, K. Gentry, N. Sumangala and A. Ramamoorthy, F1000Research, 2017, 6, 662.
3. F. P. Guengerich, Chem. Res. Toxicol., 2008, 21, 70.
4. F. P. Guengerich, Z. L. Wu and C. J. Bartleson, Biochem. Biophys. Res. Commun., 2005, 338,
465.
5. U. H. Du¨rr, L. Waskell and A. Ramamoorthy, Biochim. Biophys. Acta, 2007, 1768, 3235.
6. S. C. Im and L. Waskell, Arch. Biochem. Biophys., 2011, 507, 144.
112
7. M. Wang, D. L. Roberts, R. Paschke, T. M. Shea, B. S. Masters and J. J. Kim, Proc. Natl. Acad.
Sci. U. S. A., 1997, 94, 8411.
8. R. Huang, M. Zhang, F. Rwere, L. Waskell and A. Ramamoorthy, J. Biol. Chem., 2015, 290,
4843.
9. D. F. Estrada, J. S. Laurence and E. E. Scott, J. Biol. Chem., 2016, 291, 3990.
10. H. Zhang, D. Hamdane, S. C. Im and L. Waskell, J. Biol. Chem., 2008, 283, 5217.
11. Y. Yang, H. Zhang, D. Usharani, W. Bu, S. Im, M. Tarasev, F. Rwere, N. M. Pearl, J. Meagher,
C. Sun, J. Stuckey, S. Shaik and L. Waskell, Biochemistry, 2014, 53, 5080.
12. L. A. McLaughlin, S. Ronseaux, R. D. Finn, C. J. Henderson and C. R. Wolf, Mol. Pharmacol.,
2010, 78, 269.
13. R. D. Finn, L. A. McLaughlin, S. Ronseaux, I. Rosewell, J. B. Houston, C. J. Henderson and
C. R. Wolf, J. Biol. Chem., 2008, 283, 31385.
14. H. Yamazaki, M. Nakamura, T. Komatsu, K. Ohyama, N. Hatanaka, S. Asahi, N. Shimada, F.
P. Guengerich, T. Shimada, M. Nakajima and T. Yokoi, Protein Expression Purif., 2002, 24, 329.
15. J. Y. Chiang, Arch. Biochem. Biophys., 1981, 211, 662.
16. J. B. Schenkman and I. Jansson, Pharmacol. Ther., 2003, 97, 139.
17. D. F. Estrada, A. L. Skinner, J. S. Laurence and E. E. Scott, J. Biol. Chem., 2014, 289, 14310.
18. N. M. Pearl, J. Wilcoxen, S. Im, R. Kunz, J. Darty, R. D. Britt, S. W. Ragsdale and L. Waskell,
Biochemistry, 2016, 55, 6558.
19. T. Shimada, R. L. Mernaugh and F. P. Guengerich, Arch. Biochem. Biophys., 2005, 435, 207.
20. H. Zhang, S. C. Im and L. Waskell, J. Biol. Chem., 2007, 282, 29766.
21. P. P. Tamburini, R. E. White and J. B. Schenkman, J. Biol. Chem., 1985, 260, 4007.
113
22. A. Bridges, L. Gruenke, Y. T. Chang, I. A. Vakser, G. Loew and L. Waskell, J. Biol. Chem.,
1998, 273, 17036.
23. M. Zhang, S. V. Le Clair, R. Huang, S. Ahuja, S. C. Im, L. Waskell and A. Ramamoorthy, Sci.
Rep., 2015, 5, 8392.
24. M. Zhang, R. Huang, S. C. Im, L. Waskell and A. Ramamoorthy, J. Biol. Chem., 2015, 290,
12705.
25. M. Zhang, R. Huang, R. Ackermann, S. C. Im, L. Waskell, A. Schwendeman and A.
Ramamoorthy, Angew. Chem., Int. Ed. Engl., 2016, 55, 4497.
26. K. A. Gentry, E. Prade, C. Barnaba, M. Zhang, M. Mahajan, S. C. Im, G. M. Anantharamaiah,
S. Nagao, L. Waskell and A. Ramamoorthy, Sci. Rep., 2017, 7, 7793.
27. S. Ahuja, N. Jahr, S. C. Im, S. Vivekanandan, N. Popovych, S. V. Le Clair, R. Huang, R.
Soong, J. Xu, K. Yamamoto, R. P. Nanga, A. Bridges, L. Waskell and A. Ramamoorthy, J. Biol.
Chem., 2013, 288, 22080.
28. D. F. Estrada, J. S. Laurence and E. E. Scott, J. Biol. Chem., 2016, 291, 3990.
29. H. Zhang, E. Myshkin and L. Waskell, Biochem. Biophys. Res. Commun., 2005, 338, 499.
30. E. Canova-Davis and L. Waskell, J. Biol. Chem., 1984, 259, 2541.
31. E. T. Morgan and M. J. Coon, Drug Metab. Dispos., 1984, 12, 358.
32. D. Batabyal and T. L. Poulos, J. Inorg. Biochem., 2018, 183, 179.
33. P. C. Nair, R. A. McKinnon and J. O. Miners, Drug Metab. Rev., 2016, 48, 434.
34. R. Huang, K. Yamamoto, M. Zhang, N. Popovych, I. Hung, S. C. Im, Z. Gan, L. Waskell and
A. Ramamoorthy, Biophys. J., 2014, 106, 2126–2133.
35. A. Gutierrez, L. Y. Lian, C. R., Wolf, N. S., Scrutton, G. C., Roberts, Biochemistry, 2001, 40,
1964.
114
36. U. H. N. Dürr, K. Yamamoto, S.-C. Im, L. Waskell, A. Ramamoorthy, J. Am. Chem. Soc.
2007, 129, 6670.
37. A. S. Saribas, L. Gruenke, L. Waskell, Protein Expr. Purif., 2001, 21, 303.
38. D. G. Kneller, I. D. Kuntz, J. Cell. Biochem., 1993, 53, 254.
115
CHAPTER FIVE
Probing dynamic structural protein-protein and protein-substrate interactions in ternary complex
of Cytochrome P450, Cytochrome b5, and Cytochrome P450 Reductase
5.1 Summary
Cytochrome P450 (cytP450) interacts with two redox partners, cytochrome P450 reductase
(CPR) and cytochrome b5 (cytb5), in order to metabolize a wide variety of substrates. All three of these proteins are membrane bound proteins and in order to investigate this dynamic interplay fully, we need to study these proteins in a membrane environment. Here, we show for the first time, the incorporation of three single pass transmembrane helix containing proteins into a peptide- based lipid nanodisc. Once all three members of the ternary complex have been incorporated, competition assays with NMR spectroscopy were performed to monitor the effect of lipids and substrates on redox partner binding to cytP450. We see cytb5 dominating the competition by weakening the flFBD’s complex with cytP450. In the membrane environment, flFBD is more able to disrupt the cytb5-cytP450 complex than in the absence of lipids albeit not completely. The addition of substrates to this ternary complex reveal differences in redox partner binding to cytP450 as well as the skewing of the interplay to favor cytb5 over flFBD binding.
116
5.2 Introduction
Cytochrome P450s (cytP450s) are a ubiquitous superfamily of enzymes responsible for the metabolism of a variety of compounds from vitamins to fatty acids to over 70% of the drugs on the pharmaceutical market.[1-5] Per turn of its catalytic cycle which requires two electrons, cytP450 inserts a molecule of activated oxygen into a hydrophobic substrate via the addition of a hydroxyl to either a carbon-carbon or carbon-hydrogen bond. These two electrons can be provided by either
Cytochrome P450 reductase (CPR), cytP450’s obligate redox partner, or the second electron can be donated by cytochrome b5 (cytb5) [6]. Most of the drug metabolizing isoforms of cytP450, around 55 kDa, are all localized to the endoplasmic reticulum membrane in liver microsomes. [7,
8]
CPR is an 80 kDa protein that consists of four distinct domains: the FAD/NADPH binding domain, the FMN binding domain, a linker region connecting the two flavin domains, and an N- terminal transmembrane domain. [5, 9] After being reduced by NADPH, electrons flow from the
FAD binding domain to FMN in the FMN binding domain (FBD) which directly donates the electrons to the heme of cytP450. In this study, we utilize a truncated version of CPR, that is the full-length 27 kDa FBD (flFBD), consisting of the FMN binding domain along with the N-terminal transmembrane domain. This flFBD domain has been used previously as a minimal necessary domain to interact with cytP450 (11; Chapter 3). Crystallization of the FBD alone shows that it maintains its Rossman fold structure even outside of the full-length protein. [10]
Cytb5, the third protein of this ternary complex, is a 15 kDa protein. This protein has a C- terminal transmembrane domain, a flexible linker, and a soluble domain. It is only capable of donating the second electron to cytP450 because of the disparity in redox potentials between cytb5 and ferric cytP450. [5, 12, 13] Cytb5 has been shown to increase, decrease, or do nothing to cytP450
117 metabolism depending on the isoform of cytP450 and substrate involved. [2, 14, 15] In some cases, such as that of cytP450 17A1, cytb5 is known to favor a specific reaction and product formation.[16,
17] Due to the mystery still surrounding cytb5’s functional properties, we want to further investigate the role cytb5 can play in regulating cytP450 drug metabolism.
All three proteins, cytP450, cytb5, flFBD, contain single transmembrane helices that anchor the proteins to the ER membrane. In cytP450’s case, even the globular, ‘soluble’ domain, specifically the F/G-loop, interacts with the lipid bilayer. [18-20] The presence of a lipid bilayer has been shown to influence these cytP450-redox partner protein-protein interactions both through dictating favorable orientations for complex formation, increasing or decreasing electron transfer rates and metabolism of substrates, and altering the spin state shifts of the heme group of cytP450 when the redox partner is present. [2, 21] In order to study this ternary complex in a lipid environment, we utilize peptide based nanodiscs. Using the 4F peptide as the scaffold belt for the nanodiscs is ideal for this ternary complex because it creates very flexible nanodiscs that are accommodating for the stepwise addition of single transmembrane helix containing proteins.
Previous studies by the Ramamoorthy lab have illustrated this ability with two membrane proteins.
(Chapter 3)
As substrates have been shown to have sizeable impact on strengthening the complex interactions between cytb5 and cytP450, five substrates were chosen to examine in this study.
These substrates were chosen due to their range in hydrophobicity although as cytP450 substrates, they are all hydrophobic in nature, and availability of crystal structures of cytP450 solved in the presence of these compounds. The compounds are: butylated hydroxytoluene (BHT; LogP = 5.3), bifonazole (BFZ; LogP = 4.8), benzphetamine (BZ; LogP = 4.1), 4-(4-Chlorophenyl)-1H- imidazole (4-CPI; LogP = 2.4), and 1-(4-Chlorophenyl)-imidazole (1-CPI; LogP = 2.3). In the
118 following study, we examine the role of membrane and various substrates on the interplay between cytP450 and its two redox partners.
5.3 Results
5.3.1 Incorporation of three membrane proteins into a lipid nanodisc
In order to study the ternary complex in lipid nanodiscs, each full-length protein was expressed, purified and characterized as reported previously. For this study, the cytP450 isoform that is utilized is cytP450 2B4, a rabbit homolog with 76% sequence identity to human cytP450
2B6. [22] Sequential incorporation of the three proteins was done over the course of three days using 4F-DMPC nanodiscs that were purified through size exclusion chromatography (SEC). Both versions of the ternary complex were assembled: one with cytb5 incorporated first, then cytP450, and then flFBD, called “cytb5-ternary complex” and one with flFBD incorporated first, then cytP450, and then cytb5, called “flFBD-ternary complex”. Reconstitution of these proteins into lipid nanodiscs was accomplished by the mixing of empty nanodiscs and protein, incubating overnight, purifying by SEC, and characterizing their size by dynamic light scattering (DLS). On the second day, one molar equivalent of cytP450 2B4 would be added to form complexes of cytb5- cytP450 in nanodiscs or flFBD-cytP450 in nanodiscs. On the third day, the other redox partner would be added to the samples in order to make ternary complexes of cytb5-ternary and flFBD- ternary. Figure 5.1 illustrates the incorporation of each protein into the lipid nanodisc by monitoring the changes in Stokes radius upon addition of each protein. In Figure 1A and B a gradual increase in size is displayed while creating the flFBD-ternary complex as seen by the radius grow larger in DLS or elution time grow shorter in SEC shifting to the right in DLS or left in SEC after the addition of empty nanodiscs (blue), flFBD (green), cytP450 (black), and cytb5
119
(red). Figure 1C, D show the cytb5-ternary complex formation with empty nanodiscs (blue), cytb5
(red), cytP450 (black), and flFBD (green).
120
Figure 5.1: Incorporation of three membrane proteins into a lipid nanodisc. A and B both display the formation of flFBD- ternary complexes into nanodiscs with stepwise incorporation of the three proteins through SEC (A) and DLS (B) for empty 4F- DMPC nanodiscs (blue), flFBD (green), cytP450 (black), and cytb5 (red). C and D show the incorporation of the cytb5-ternary complexes similarly with the SEC profile (C) and DLS profile (D) for empty 4F-DMPC nanodiscs (blue), cytb5 (red), cytP450 (black), and flFBD (red).
5.3.2 15N-cytb5 monitored ternary complex formation
Previous work and past literature have revealed that substrates drive the formation of a strong complex between cytb5 and cytP450. This has been illustrated in both complexes containing cytP450 and cytb5 alone as well as in ternary systems with cytP450, cytb5, and truncated FBD.
[23-25] Complex formation between 15N-labeled cytb5 and cytP450 was monitored through 1H/15N
TROSY HSQC NMR experiments as we can measure the signal intensities and linewidths of cytb5’s amide 15N-1H peaks. As the chemical environment around these amide 15N-1H protons change upon binding to cytP450, there is a sizeable decrease in the linewidth and signal intensity of resonances observed from 2D 15N/1H TROSY HSQC NMR spectra. The 2D 15N/1H TROSY
HSQC NMR spectra are found in Appendix A Figures A.3-A.8 along with each individual resonance’s relative signal intensity for each assigned amino acid in cytb5.
Figure 5.2 reports the average signal intensity observed for the amino acids clustered on the lower cleft of cytb5. These residues (N62-R73) are highlighted in blue in the inset of Figure
121
Figure 5.2: 15N-cytb5 monitored ternary formation. Average signal intensity of cytb5 residues that are involved in binding to cytP450 (N62-R73). [22] Each bar graph corresponds to the addition of a protein or drug being added in to the cytb5 sample (red). 15N cytb5 alone has full signal intensity (red). After the addition of one molar equivalent of cytP450, the intensity of the binding residues diminishes (orange). When flFBD is added to the sample, it is able to partially restore cytb5 intensity and removes cytb5 from binding to cytP450 (black). Each of the five drugs is in the lighter shade and the darker shade is the addition of flFBD to the substrate bound-cytb5-cytP450 complex. (inset) The lower cleft residues are highlighted in blue on a structure of cytb5 (PDB 2M33). The black line is for reference of what the complex of cytb5-cytP450 average intensity is.
5.2 and have been identified as important residues that bind to cytP450. [22]. As this lower cleft is highly involved in binding to cytP450, it is a good measure of cytb5’s complexation state. Upon the addition of cytP450 (Figure 5.2, orange) to the 15N-labeled cytb5, the overall signal intensity of the lower cleft of cytb5 drops to about ~52% of the original intensity. Intriguingly, when various substrates were added to the complex of cytb5 and cytP450 the binding is not strengthened based on the nearly no changes observed for the signal intensities from the NMR spectra. This lack of
122
Figure 5.3: Zoomed-in sections of cytb5 resonances reveal changes in Serine 69’s intensity as the ternary complex forms. Row 1: 15N-labeled cytb5 alone (red), with 1x molar equivalent of cytP450 (orange), and with 1x molar equivalent of flFBD (grey). Row 2: The complex of 15N-cytb5-cytP450 in nanodiscs after the addition of substrate: 4-CPI (light pink, broadened); BFZ (yellow); BHT (light green); 1-CPI (light purple); BZ (light blue). Row 3: The complex of 15N-cytb5-cytP450-substrate after the addition of flFBD: 4-CPI (dark pink), BFZ (dark yellow); BHT (dark green); 1-CPI (dark purple); and BZ (dark blue). Complementary 1D slices can be found in Appendix A Figure A.1 to accompany these 2D plots.
substrate effect is not completely surprising because the complex is already showing a tight binding due to the presence of the lipid membrane in the nanodiscs in comparison to the lipid-free condition as reported in our previous study (Chapter 4). [19] One hypothesis as to why the substrates don’t increase the complex strength is that the hydrophobic compounds do partition into the lipid bilayer. As the drugs have another way of increasing hydrophobic contacts than promoting binding between the two proteins, the substrate-induced changes in the protein-protein interaction does not happen as strongly as observed in the absence of lipids. Some drugs present, like BFZ and BZ, slightly dislodge the complex between cytb5 and cytP450 (Figure 5.2 light green, light blue) as shown by the slight increase in signal intensity. While we do not know what is the cause for this
123 disruption, there could be several things happening. As the hydrophobic compounds can partition into the membrane, they could be interacting with the transmembrane domains of the proteins or with other parts of cytP450.
Upon the addition of flFBD, we see a small increase in signal intensity, but it does not return to the same intensity as observed for nanodiscs containing cytb5 alone. In the presence of substrates, flFBD is unable to greatly disrupt the interaction in between cytb5 and cytP450. The greatest effect that flFBD has is in the absence of a substrate. Without a substrate present, flFBD is able to dislodge cytb5 from a complex with cytP450 which is demonstrated by returning to about
85% of the starting signal. By dislodging cytb5, we mean that flFBD is able to disrupt the complex between cytb5 and cytP450 and interfere so some of the cytb5 is no longer bound to cytP450.
Cytb5 does not leave the nanodisc. Substrates can keep cytb5 and cytP450 bound to one another and make flFBD less capable of disrupting the formation. Drugs with highest to lowest ability of maintaining complex formation are in the following order: BZ, BFZ, BHT, 1-CPI, 4-CPI. Looking closely at one of the residues identified in binding on cytb5 to cytP450, Serine 69, we can see in
Fig. 5.3 that the signal decreases upon the addition of cytP450 and then slight changes happen after the additions of drugs and then varying levels of signal increase upon titration with flFBD.
Appendix A.1 displays the extracted linewidths of Serine 69 to illustrate the broadening of the signal upon the addition of cytP450 and the partial restoration after the addition of flFBD.
5.3.3 15N-flFBD monitored ternary complex formation
From our previous work [23], it was shown that cytb5 can dislodge the FBD from a complex of FBD and cytP450. We were curious as to how the membrane would affect cytb5’s ability to disrupt the formed complex as it provides both a more native membrane environment and spatial constraint. Uniformly 15N-labeled flFBD was expressed, purified, and reconstituted into 4F-
124
DMPC nanodiscs as in Chapter 3. A 15N/1H TROSY-HSQC spectrum was acquired of 15N-flFBD alone (red). Stepwise reconstitution was done to then incorporate cytP450 2B4 (orange) into the flFBD containing nanodisc and a spectrum was recorded (orange). Each of the five drugs chosen were incubated with the complex of flFBD-cytP450 (4-CPI, pink; BFZ, yellow; BHT, green; 1-
CPI, blue; BZ, purple) and then recorded. Finally, cytb5 was incorporated into the fl-FBD- cytP450-ND with and without substrates. The full 2D 15N/1H TROSY-HSQC spectra can be found in Appendix A Figures A.10-A.16.
The overall signal intensity of the flFBD residues can be monitored as the protein complexes are formed and broken. A decrease in the signal intensity upon the addition of cytP450 indicates that a complex has been formed between fl-FBD and cytP450. After the addition of the five different substrates, there is not a large difference in complex strength. Substrates have not been shown to dramatically increase the affinity of cytP450 for FBD, so this result was not surprising. Both 4-CPI and BFZ marginally increase the complex strength while BHT and BZ’s addition do not strengthen the complex or, in BZ’s case, weakens it a little. The addition of 1-CPI however, was significant in increasing the complex formation which can be shown in Figure 5.4.
Once cytb5 has been incorporated into the nanodisc, it does show the ability to disrupt the fl-FBD-cytP450 complex. This is demonstrated by the restoration of signal intensity of fl-FBD residues, particularly, looking in Figure 5.4 at Glycine 89. Glycine 89 is a residue in loop 1 of flFBD which coordinates the FMN cofactor and has been previously implicated in binding to cytP450. (Chapter 3) By analyzing the signal intensity and linewidth of this residue, we get information about its chemical environment and change of timescale. Upon addition of cytP450, the resonance broadens, and the intensity decreases which can be seen in Figure 5.4 (orange).
Adding in a substrate does not significantly change the peak intensity for any of the substrates
125
Figure 5.4: 15N-flFBD monitored ternary formation. Average overall signal intensity of flFBD residues. In red is the complex intensity at 100% for flFBD in complex with cytP450. Each of the drugs is separated by a blank space. After the red bar is the intensity for the complex with a substrate added (4-CPI, BFZ, BHT, 1-CPI, and BZ) or no substrate. The third bar is the addition of cytb5 to the complex with substrate.
chosen at this residue of flFBD. Upon cytb5 incorporation into the nanodisc, G89’s linewidth and signal intensity are greatly restored (right column, Figure 5.4). Meanwhile other residues on flFBD that are farther away from the interaction like S123, undergo some changes but these are much more consistent with the larger nanodisc complex’s tumbling rate rather than direct interaction with cytP450. Appendix A.2 displays the extracted linewidths of Glycine 89 to illustrate the
126 broadening of the signal upon the addition of cytP450 and the restoration after the addition of cytb5.
5.4 Discussion
The incorporation of the ternary complexes into lipid nanodiscs led to some interesting behavior. While cytb5 is still able to dislodge flFBD from a complex with cytP450, it is not to as great of an extent as it could in solution. Having the three proteins being spatially confined in a nanodisc could have contributions to keeping flFBD bound to cytP450 better, but it is currently unclear to assign that to a lipid influence on complex strength or the physical restraint of motion.
The variety of substrates utilized in this study all behaved similarly in regard to forming a stronger complex with cytb5-cytP450, however, it was much less of an effect than what was seen in lipid free solution in Chapter Four. flFBD was able to dislodge cytb5 from its complex with cytP450, but not complete. The highest effect was seen without substrates and flFBD was only able to slightly disrupt the complex in the presence of substrates.
One of the more interesting comparisons from this study include the two inhibitors 4-CPI and 1-CPI in the presence of 15N-flFBD-ternary complex. In agreement with Zhao et al., we have found major differences in the interactions between cytP450 and flFBD in the presence of 1-CPI versus 4-CPI even with how structurally similar these compounds are. [26] Crystal structures of cytP450 2B4 have been solved in the presence of both inhibitors which revealed very different active site conformations. The calculated active site volume for 4-CPI was 200 Å versus 280 Å for
1-CPI bound cytP450 2B4. The thermodynamic properties of binding of both inhibitors was also investigated by ITC. Our NMR data reveal that there is an increased binding and tighter complex formation between cytP450 2B4 and flFBD after the 1-CPI inhibitor addition which is in agreement with the ITC experiments. Zooming in closer to residues by the binding site, Appendix
127
Figure 5.5: Cytb5 disrupts the complex of flFBD-cytP450 in nanodiscs. Highlighted sections of 1H/15N TROSY-HSQC spectra of 15N-flFBD displaying residues G89 which has been shown to be involved in binding to cytP450 and S123 which is not involved. (red) 15N-flFBD; (orange) 15N-flFBD with 1xcytP450 2B4; (gray) 15N-flFBD with 1xcytP450 2B4 and 1xcytb5. The chemical structure of each of the five substrates used is displayed on the left with the middle column showing the 15N-flFBD-cytP450 complex with substrate spectrum and the right column showing the 15N-flFBD-cytP450-substrate-drug spectrum. 4-CPI (pink); Bifonazole (yellow); Butylated hydroxytoluene (green); 1-CPI (blue); Benzphetamine (purple).
128
A Figure A.18 shows the average signal intensities for each of the four loops that coordinate the
FMN cofactor. In the presence of 4-CPI binding to cytP450, loop 1 on flFBD is strongly affected but the other loops are much less so. In comparison, in the presence of 1-CPI binding to cytP450, loop 2 on flFBD is strongly affected. Even the addition of cytb5 does not completely restore the signal intensity of loop 2 in the 1-CPI complex sample.
This ternary complex data still has a lot of details to unravel. Interestingly this data reveals more residues binding of flFBD to cytP450 which has recently been reported by the Ramamoorthy
Lab. [27] In the presence of a membrane, flFBD binds to cytP450 with all four loops that surround the FMN. Future work will be to investigate these loops with the substrates further. The major takeaway from this study is that in the presence of lipid membranes, this dynamic interplay approaches what is thought to be more physiological behavior. Without being reconstituted in a lipid nanodisc, flFBD cannot dislodge cytb5 from binding to cytP450. Physiologically, if cytb5 is bound to cytP450 before the first electron can be donated by CPR, then the catalytic cycle ceases.
Cytb5 is incapable of donating the first electron due to the disparity in redox potentials. Without the dynamic exchange between redox partner binding, drug metabolism slows or even halts.
Therefore, studying these proteins in their lipid membrane context is of high important to promote correct orientation of proteins for protein-protein interactions and to enable functionality to achieve this dynamic interplay.
129
Figure 5.6: Schematic of cytb5, cytP450, and CPR in a lipid nanodisc
5.5 Materials and Methods
5.5.1 Materials and Reagents
E. coli C41 cells for protein overexpression were purchased from Lucigen (Middleton, MI). Yeast extract, tryptone, and sodium chloride for unlabeled growth media were purchased from Sigma-
15 Aldrich. N Ammonium Chloride and D2O were purchased from Cambridge Isotope Laboratories
(Andover, MA). I,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) was purchased from
Avanti Polar Lipids (Alabaster, AL). 5-mm symmetrical D2O-matched Shigemi NMR microtubes
130 were purchased from Shigemi, Inc. (Allison Park, PA). Resins, phosphate buffer components
(monobasic and dibasic), and all other chemicals were purchased from Sigma-Aldrich.
5.5.2 Expression and Purification of full-length cytochrome b5
Full-length uniformly 15N-labeled and unlabeled wild-type rabbit cytb5 was expressed and purified as described previously. [28-30] Briefly, E. coli C41 cells were transformed with a pLW01 plasmid containing the cytb5 gene. The cells were grown up in LB medium to an OD of 1 (at 600 nm).
This culture was diluted 100-fold into 100 mL of 15N-Celtone medium. This culture was grown at
35 °C with shaking at 250 rpm until an OD of 1 at 600 nm was achieved. The cells were pelleted and resuspended in 10 mL of fresh 15N-Celtone medium. The resuspended cell culture was added to the final 1 L of culture minimum medium. Isopropyl β-D-thiogalactopyranoside was added to a final concentration of 10 μM, and incubation was continued for 20 h, at which time the cells were harvested.
5.5.3 Expression and Purification of full-length FBD
U-15N-labeled and unlabeled flFBD were expressed and purified as described previously [31]. The fl-FBD gene is encoded on a pSC plasmid and is preceded by the OmpA signal peptide [32]. Briefly, fl-FBD was expressed in E. coli C41 cells in either LB medium for unlabeled protein, or M9 medium for U-15N labeled samples, supplemented with 5 nM FMN. Protein expression was induced at OD600 = 0.7 by adding 0.4 mM IPTG to the cultures for 16 h at 30 °C. After cell harvest at 6000 x g and 4 °C, the cells were lysed by 30 µg/ml lysozyme and protease inhibitor in Tris-
Acetate buffer pH 7.4 for 30 mins at 4 °C, followed by sonication (with 1 s on and 1 s off pulses) for 5 mins. The membrane fraction was pelleted by ultracentrifugation at 105,000 x g and 4 °C for
45 mins, and further treated with 0.3% (v/v) Triton X-100 for 16 h at 4 °C. The solubilized membrane proteins were purified by DEAE anion exchange chromatography twice. For this
131 purpose, the protein was loaded onto the column and eluted using a NaCl gradient ranging from
0.2 M to 0.5 M in Tris-Acetate pH 7.4 containing 1 µM FMN, 0.3% (v/v) sodium cholate.
5.5.4 Expression and Purification of full-length cytP450 2B4
Full-length CYP2B4 was expressed and purified as described in the literature [28].
5.5.5 Preparation of nanodiscs
DMPC powder was suspended into buffer A (40 mM potassium phosphate, pH 7.4) to make a stock solution at 20 mg/mL. The 4F peptide (DWFKAFYDKV AEKFKEAF) was dissolved in buffer A to make a stock solution at 10mg/mL. The DMPC stock solution was vortexed and sonicated three times for 30s each to create a suspension, and vortexed thoroughly immediately before use. The stock solution was mixed together at a peptide:lipid ratio of 1:1.5 % w/w and incubated at 37 °C o/n with slow agitation. The nanodiscs were purified by size exclusion chromatography (SEC). A Superdex 200 Increase 300/10 GL column was operated on an AKTA purifier (GE Healthcare, Freiburg, Germany).
5.5.6 Reconstitution of full-length proteins in nanodiscs
Cytb5 or flFBD were added to the empty nanodiscs at molar ratios of 1:1.2 (protein/nanodisc) and incubated for 16 hours at 25 °C with gentle agitation. The reconstitution was further purified by
SEC. Fractions showing absorbance at 417 nm (cytb5) or 454 nm (FBD) were pooled and used for further analysis. In order to form a complex, cytP450 was added to purified cytb5 or flFBD containing nanodiscs at a molar ratio of 1:1. After overnight incubation for 16 hours at 25 °C with gentle agitation and another round of purification either cytb5 or flFBD was added to form all three protein-containing nanodiscs.
The empty nanodiscs and reconstituted proteins were subjected to dynamic light scattering (DLS) measurements on a DynaPro NanoStar instrument (Wyatt Technology Corp., Santa Barbara, USA)
132 at 25 °C for 10 acquisitions of 5 s each. DLS and SEC measurements confirm the increase of the hydrodynamic radius after stepwise incubation with cytb5/flFBD, fl-CYP2B4 and fl-FBD/cytb5.
5.5.7 NMR experiments
NMR experiments were performed at 298 K on an 800 MHz Bruker spectrometer which is equipped with an Ascend magnet and TCI cryoprobe. 2D 1H/15N TROSY HSQC spectra were recorded from 0.1 mM 15N-labeled protein (either flFBD or cytb5) in 40 mM potassium phosphate, pH 7.4. All NMR spectra were recorded with 128 scans and 128 t1 increments. Data was processed using TopSpin (Bruker) and analyzed with Sparky (Goddard). The B_TROSY 1H/15N TROSY
HSQC pulse sequence was modified by adjusting CNST54 and CNST55 to change the sweep width to cover 9±3 ppm rather than 8.3±2.5 ppm to see all appropriate resonances. The previously reported cytb5 and flFBD backbone chemical shift assignments were used in this study.
5.6 References
1. F. P. Guengerich, Mol. Interventions, 2003, 3, 194.
2. C. Barnaba, K. Gentry, N. Sumangala and A. Ramamoorthy, F1000Research, 2017, 6, 662.
3. F. P. Guengerich, Chem. Res. Toxicol., 2008, 21, 70.
4. F. P. Guengerich, Z. L. Wu and C. J. Bartleson, Biochem. Biophys. Res. Commun., 2005, 338,
465.
5. U. H. Du¨rr, L. Waskell and A. Ramamoorthy, Biochim. Biophys. Acta, 2007, 1768, 3235.
6. S. C. Im and L. Waskell, Arch. Biochem. Biophys., 2011, 507, 144.
7. K. Ruckpual and H. Rein, Cytochrome P450, 1984, Akademie-Verlag, Berlin.
8. A. Stier, Biochemical Pharmacology, 25, 109-113.
9. M. Wang, D. L. Roberts, R. Paschke, T. M. Shea, B. S. Masters and J. J. Kim, Proc. Natl. Acad.
Sci. U. S. A., 1997, 94, 8411.
133
10. Q. Zhao, S. Modi, G. Smith, M. Paint, P. D. McDonagh, C. R. Wolf, D. Tew, L. Y. Lian, G.
C. Roberts, and H. P. Driessen. Protein Sci. 1999, 8, 298-306.
11. 9. D. F. Estrada, J. S. Laurence and E. E. Scott, J. Biol. Chem., 2016, 291, 3990.
12. H. Zhang, D. Hamdane, S. C. Im and L. Waskell, J. Biol. Chem., 2008, 283, 5217.
13. Y. Yang, H. Zhang, D. Usharani, W. Bu, S. Im, M. Tarasev, F. Rwere, N. M. Pearl, J. Meagher,
C. Sun, J. Stuckey, S. Shaik and L. Waskell, Biochemistry, 2014, 53, 5080.
14. H. Yamazaki, M. Nakamura, T. Komatsu, K. Ohyama, N. Hatanaka, S. Asahi, N. Shimada, F.
P. Guengerich, T. Shimada, M. Nakajima and T. Yokoi, Protein Expression Purif., 2002, 24, 329.
15. J. Y. Chiang, Arch. Biochem. Biophys., 1981, 211, 662.
16. J. B. Schenkman and I. Jansson, Pharmacol. Ther., 2003, 97, 139.
17. D. F. Estrada, A. L. Skinner, J. S. Laurence and E. E. Scott, J. Biol. Chem., 2014, 289, 14310.
18. Monk, B. C. et al. (2014) Architecture of a single membrane spanning cytochrome P450 suggests constraints that orient the catalytic domain relative to a bilayer. P.N.A.S. 111, 3865-
3870.
19. McDougle, D. R. et al. (2015) Incorporation of charged residues in the CYP2J2 F-G loop disrupts CYP2J2-lipid bilayer interactions. Biochem. Biophys. Acta Biomembr. 1848, 2460-2470.
20. Baylon, J. L., Lenov, I. L., Sligar, S.G., Tajkhorshid, E. (2013) Characterizing the
Membrane-Bound State of Cytochrome P450 3A4: Structure, Depth of Insertion, and
Orientation. J. Am. Chem. Soc. 135, 8542 – 8551.
21. T. Ravula, C. Barnaba, M. Mahajan, G. M. Anantharamaiah, S. C. Im, L. Waskell, A.
Ramamoorthy, Chem. Commun. (Camb). 2017, 53, 12798-12801.
134
22. Scott, E. E. et al. Structure of mammalian cytochrome P450 2B4 complexed with 4-(4- chlorophenyl)imidazole at 1.9-A resolution: insight into the range of P450 conformations and the coordination of redox partner binding. J. Biol. Chem. 279, 27294–301 (2004).
23. K. A. Gentry, M. Zhang, S.C. Im, L. Waskell, A. Ramamoorthy. Chem. Commun. 2017, 54,
5780-5783.
24. M. Zhang, R. Huang, R. Ackerman, S.C. Im, L. Waskell, A. Schwendeman, A. Ramamoorthy,
Angewandte Chemie, 2016, 55, 4497-9.
25. M. Zhang, S. V. Le Clair, R. Huang, S. Ahuja, S.C. Im, L. Waskell, A. Ramamoorthy, Sci.
Rep. 2015, 5, 8392.
26. Y. Zhao, L. Sun, B. K. Muralidhara, S. Kumar, M. A. White, C. D. Stout, J. R. Halpert,
Biochemistry, 2007, 46, 11559-67.
27. M. Mahajan, T. Ravula, E. Prade, G. M. Anantharamaiah, A. Ramamoorthy. Chem. Commun.
2019, 55, 5777-5780.
28. S. Ahuja, N. Jahr, S.C. Im, S. Vivekanandan, N. Popovych, S. Le Clair, R. Huang, R. Soong,
J. Xu, K. Yamamoto, R. P. R. Nanga, A. Bridges, L. Waskell, A. Ramamoorthy, J. Biol. Chem.
2013, 288, 22080-22095.
29. U. H. Durr, L. Waskell, and A. Ramamoorthy, Biochim. Biophys. Acta. 2007, 1768, 3235-
3259.
30. A. Bridges, L. Gruenke, Y. T. Chang, I. A. Vakser, G. Loew, and L. Waskell. J. Biol. Chem.
1998, 273, 17036-17049.
31. R. Huang, M. Zhang, F. Rwere, L. Waskell, A. Ramamoorthy, J. Biol. Chem. 2015, 290, 4843-
4855.
32. A. L. Shen, T. D. Porter, T. E. Wilson, C. B. Kasper, J. Biol. Chem. 1989, 264, 7584-7589.
135
CHAPTER SIX
Conclusions and Future Directions
6.1 Conclusions
The goal of this thesis is to understand the effects that a lipid environment has on the protein-protein interactions between cytochrome P450 and its redox partners cytochrome P450 reductase and cytochrome b5 with atomic level structural detail using NMR spectroscopy and other biophysical techniques. In this dissertation, I explored the effect of different membrane mimetics, both isotropic bicelles and lipid nanodiscs, on the interaction between cytochrome b5 and cytochrome c. We discovered that cytb5 embedded in a membrane mimetic was crucial for more natural electron transfer ability as well as shifted the protein-protein encounter interface. Then we developed and optimized these peptide nanodiscs to incorporate two single transmembrane domain containing proteins, full-length FBD and cytP450. Here, SEC, DLS, and SAXS data proved the incorporation of the complex into the lipid nanodiscs as well as stopped-flow data proved it a functional complex. The next goal of this thesis was to be able to study all three proteins, cytP450, flFBD, and cytb5, at the same time. We next worked in a lipid-free environment to discover that substrates dictated the protein-protein interaction that cytP450 preferred. Cytb5 was shown to be able to displace flFBD from a complex with cytP450 which varied with the different substrates present. Eventually, after optimizing sample preparation, ternary complexes in lipid nanodiscs were formed and studied. We were able to demonstrate that similar to the lipid-free solution, cytb5 was able to dislodge flFBD, although not as strongly. On the other hand, we were able to see flFBD disrupting the cytb5-cytP450 complex to an extent, especially strong effects were seen with no
136 substrate present. Overall, this thesis illustrates the importance of utilizing a membrane environment when studying these membrane protein-protein interactions.
In Chapter 2, we investigated the interaction between full-length cytochrome b5 and cytochrome c in the following conditions: lipid-free solution, isotropic bicelles, and lipid nanodiscs. Both the structural aspects of this complex formation and the electron transfer ability of this complex were investigated. As demonstrated through stopped-flow kinetic experiments, electron transfer capable complexes were formed in all three sample conditions. However, only the membrane bound cytb5 samples had physiological interactions with cyt c with dynamic complexes forming to transfer electrons whereas the lipid-free cytb5 had quick, but static complex formation. The interaction interfaces between cytb5 and cyt c were revealed on both proteins using
NMR spectroscopy. Notably, in the presence of membrane, cytb5 only engaged cyt c at its lower and upper clefts while the membrane-free cytb5 also uses a distal region. These NMR restraints from both 15N-labeled cytb5 and 15N-labeled cyt c were used to generate a structure of this complex. This complex reveals a new binding interface, where the lower cleft of cytb5 is engaged in binding to cyt c, in comparison to what had been previously reported with residues on the side of cytb5 being involved in the interface. A “front-to-front” rather than “side-to-side” interaction was seen for the nanodisc containing cytb5 complex.
After the optimization of 4F-peptide nanodiscs, we investigated the ability of this system to incorporate two transmembrane domain containing proteins. In order to make NMR studies of cytochrome P450 reductase feasible but physiologically relevant, we utilized a truncated form of
CPR. This truncated form of CPR, or full-length FBD (flFBD; FBD = flavin mononucleotide binding domain) was chosen. The transmembrane domain-less FBD retains the same structure outside of the full protein and is the domain that directly interacts with cytP450 to reduce it. We
137 show that both cytP450 and flFBD can be reconstituted into the same nanodisc as shown by SEC,
DLS, and SAXS. Additionally, this minimal domain redox complex is active and capable of electron transfer. We used NMR spectroscopy to monitor the complex formation and used NMR restraints (chemical shift perturbations, differential line broadening, and solvent PRE effects) to generate a complex between flFBD and cytP450. A potential electron transfer pathway was predicted as well with the software HARLEM which revealed R125 on cytP450 as being crucial for electron transfer. The methodology established in this study will allow further investigation on lipid composition, cytP450 isoform, and substrate effects on this protein-protein complex.
In Chapter 4, we investigated the interplay between all three proteins, cytP450, FBD, and cytb5, and the role substrates had on dictating which redox partner cytP450 prefers. We find that substrates greatly increase the complex formation between cytb5 and cytP450 whereas they do not strengthen the interaction between FBD and cytP450. As it is known that cytb5 and FBD possess overlapping, but unique binding spots on cytP450, we decided to carry out a series of competition assays. As cytP450 has a higher binding affinity for FBD over cytb5, we expected to see that behavior in our sample. However, we found the opposite to be true. FBD is incapable of dislodging cytb5 from cytP450 but cytb5 is capable of removing FBD from a complex with cytP450. This effect is modulated by what substrate is present, with bulky aromatic compounds increasing cytb5’s preference for cytP450 and increasing cytb5’s ability to dislodge FBD from cytP450.
In Chapter 5, the redox partner interplay was probed in the presence of a lipid membrane environment. All three proteins were able to be incorporated into a single nanodisc as monitored by SEC and DLS profiles. Through a series of NMR titrations, we were able to see complex formation between one of the redox partners which was uniformly 15N-labeled by monitoring peak linewidths and intensities. For the 15N-cytb5-ternary complex, it was shown that cytb5 was still
138 capable of outcompeting flFBD from binding to cytP450 which ability was slightly influenced by substrates. Substrates had a lesser role in increasing complex strength of cytb5-cytP450 but these interactions were already stronger than other reports. On the other hand, flFBD was shown to have some ability to dislodge cytb5 from a complex with cytP450. flFBD could do this the best when no substrate was present, otherwise its ability to dislodge cytb5 from a cytP450 complex with substrate was much less. Our results in this study provide better and more in-depth knowledge for how these three proteins interact with each other in a more native environment.
6.2 Future Directions
Now with our protein biochemistry lab set up and optimized for protein production and purification along with the optimization of membrane nanodiscs, there are several avenues that this project could lead down in the future.
6.2.1. Mass spectrometry assisted capturing of cytP450 affinity for membrane/ligands
In a recent paper from our lab, Barnaba et al. [1] demonstrated that cytP450 has a lipid preference for the so-called raft domain. As these 4F-peptide based nanodiscs are flexible, they allow for the exchange of lipids. CytP450 shows an ability to recruit specific lipids, both sphingomyelin and cholesterol, to itself as demonstrated with phosphorous NMR spectroscopy.
As these lipid nanodiscs provide a stable membrane environment for the membrane proteins to be reconstituted in we can probe the effects of these lipids on cytP450 stability. In order to measure the gas-phase stability of cytP450, we have started a collaboration with Brandon Ruotolo’s lab.
Using samples made up of cytP450 reconstituted into nanodiscs containing different lipid compositions we can investigate how the lipid composition effects gas-phase stability using ion mobility-mass spectrometry (IM-MS) analysis. Collision induced unfolding (CIU), an IM-MS based calorimetry measurement, can show stability of cytP450 and the effect different ligands can
139 play on the stability. Preliminary results show discrete shifts in cytP450 stability depending on the membrane composition. Future work will include varying lipid ratios as well as including various drugs.
6.2.2 Determination of Various Rates Associated with the Ternary Complex
As shown in Chapter 5 and Chapter 6, substrates play a large role in dictating redox-partner binding to cytP450. Depending on the drug or inhibitor bound to cytP450, cytb5 can remove flFBD from its complex with cytP450. While we have shown this via looking at the binding interfaces between the redox partners and cytP450, the next step is utilizing more biochemical approaches as well as the dynamics information that NMR spectroscopy can provide.
One potential hypothesis for the substrate influence on cytP450’s choice of redox partners could be further probed by determining the binding affinity change. As we know that the KD of
[2] the P450 – CPR complex without membrane or substrate is 0.02 ± 0.02 μM – while the KD of the P450 – cytb5 complex is an order of magnitude higher at 0.2 ± 0.16 μM [2]. It would be insightful to quantify the change (if any) in binding affinity of cytochrome P450 to its two redox partners in the presence and absence of these drugs (BHT, BZ, BFZ) and inhibitors (4-CPI and 1-
CPI). While it would not be possible to accurately quantify changes in binding affinity when adding in the second redox partner, it would be interesting to qualitatively look at the binding isoforms. These experiments repeated with the membrane nanodiscs would be another avenue to study. Changes in the lipid composition could change cytP450’s binding affinity for one redox partner over the other. Our ITC findings could be compared with truncated versions of the proteins which have been previously done. [3, 4] With the new anerobic stopped-flow spectrometer installed in the lab we can test changes in the electron transfer rates between these proteins in the presence of membranes and substrates.
140
Another interesting method that could be applied well to exploring these interactions in between cytP450, cytb5, and flFBD are CPMG relaxation dispersion experiments. CPMG (Carr-
Purcell-Meiboom-Gill) experiments can be used to measure transverse, or spin-spin, T2 relaxation times of any nucleus. One of the parameters that can be extracted from CPMG datasets are the koff rates of dissociation between these proteins [5, 6]. Theoretically we could measure changes in cytb5’s dissociation from cytP450 as these complexes are in a dynamic exchange. After varying the substrates present and quantifying any changes in dissociation, flFBD could be introduced to the system and changes would be monitored. The repeat experiment but with 15N-labeled flFBD could be performed as well.
6.2.3 Labeling of cytochrome P450
One of the limitations of the studies above is the atomic resolution detail we report is restricted to the two redox partners, CPR or cytb5. We are lacking the atomic resolution detail from cytP450 although we do get some information about this protein from site-directed mutagenesis studies. There are several challenges that would need to be overcome in order to fully label and assign cytP450. First is that one of cytP450’s inherent issues is a problem of stability.
Simply in a lipid-free solution, cytP450 tends to aggregate which makes the acquisition of lengthy
3D NMR experiments impossible. Incorporating cytP450 into bicelles alleviates some of these issues but the bicelles themselves are not very stable. However, the peptide nanodiscs that have been optimized in this thesis are perfect for solubilizing and monomerizing cytP450. Reconstituted into the lipid peptide nanodiscs, cytP450 is a stable enough sample for the 3D experiments to be performed.
The next challenge is the size of cytP450. Even monomerized in nanodiscs, the size of cytP450 at 55 kDa is still challenging. A uniformly 15N, 13C 55 kDa protein will still have issues
141 even with partial or full deuteration due to the size giving a slower tumbling rate leading to broader resonances. Additionally, the spectral overlap will cause problems during assignment. Therefore, other labeling strategies will need to be carried out, including selective amino acid labeling (AILV) and/or selective 13C-methyl labeling. Previous work from the Ramamoorthy lab with specific labeling has led to easily interpretable spectra for 15N-Ala cytP450 2B4 [7]. Another labeling strategy to approach will be to incorporate an unnatural metalloporphyrin into cytP450. The iron of the heme cofactor in cytP450 leads to paramagnetic quenching of the signals which broadens the residues that are around the active site. In order to get structural details about the active site and surrounding residues we can expression cytP450s with non-native porphyrins to reduce and remove the paramagnetic effect without disrupting the structure of the protein. [8]
6.3 Future Outlook
The work presented in this dissertation has emphasized the influence that a membrane environment has on important protein-protein interactions, both with guiding orientations for complex formation and activity of the proteins. Delving into the atomic level detail of these proteins can reveal insights into structures, mechanisms, and protein-protein interactions which could all be useful for the development of novel drugs and isoform specific inhibitors. As all the cytP450s share similar sequences, the development of highly specific inhibitors is crucial in order to only target the cytP450 desired. There are specific cytP450 isoforms that are specifically expressed or overexpressed in cancerous tissues, such as cytP450 2W1 in colorectal cancers. [9] If a drug could be designed to either inactivate the protein if it is critical for cell life or designed to be activated specifically by cytP450 2W1 to be cytotoxic, killing the cancer cells specifically, that would be a great avenue for drug development.
142
Of the 57 various isoforms of cytP450, while the enzymes share a general catalytic cycle and conserved structural elements, there are various classes that the proteins can be broken into further based on their substrate metabolism. There are cytP450s that metabolize sterols, xenobiotics, fatty acids, eicosanoids, vitamins, and then there are the cytP450s of unknown function. These “orphan” cytP450s are largely poorly characterized and consist of 2A7, 2S1, 2U1,
2W1, 3A43, 4A22, 4F11, 4F22, 4V2, 4X1, 4Z1, 20A1, and 27C1. [10] Problems have arisen when attempting to crystallize these proteins even with modifications of the protein sequence due to their instability. One method to overcome these challenges associated with these proteins would be to incorporate them into lipid nanodiscs. As we have shown in these studies, the reconstitution of cytP450 2B4 into nanodiscs greatly increases the stability of the protein. Without the nanodisc, cytP450 will aggregate and precipitate even after a couple hours on the bench. On the other hand, in the nanodisc, cytP450 can withstand more heat and time which allows for more experiments.
Perhaps by reconstituting these orphan cytP450s into nanodiscs they will be stabilized, and more information can be found about their structure and function. If the peptide nanodiscs optimized in this dissertation are not successful in reconstituting the orphan cytP450s, another type of nanodisc has been developed by the Ramamoorthy lab and others that allow for the direct extraction of membrane proteins from their native environment. [11] These polymer nanodiscs would allow for these instable proteins, if they can be modified with a His-tag or other protein tag, to remain in their native membrane for further studies.
Studying the dynamic ternary complex in Chapter 4 and Chapter 5 revealed insights into the exchange between cytP450, FBD, and cytb5. Both in lipid free solution and lipid nanodiscs, cytb5 can dislodge FBD from binding to cytP450. In Chapter 4, FBD cannot remove cytb5 from its complex with cytP450, whereas in Chapter 5, when incorporated into a membrane environment,
143
FBD has some ability to dislodge cytb5, although not completely. Substrates modulate this effect with some substrates keeping cytb5 bound to cytP450. This finding is particularly relevant for drug design and testing because if cytb5 manages to bind to cytP450 before the first electron is transferred and FBD cannot dislodge cytb5, then the catalytic cycle halts as cytb5 cannot donate the first electron. Understanding how this dynamic exchange can be influenced by membrane and substrates is crucial to predicting drug metabolism. Thus, it is important to test if a new drug enhances cytb5 binding to cytP450 to an extreme because then it will not be metabolized in vivo.
With the structural and kinetic experiments carried out in this thesis highlighting the importance and necessity of using membrane mimetics to fully capture details of cytP450s and its redox partners, the field can advance farther to continue asking interesting questions about this superfamily of enzymes.
6.4 References
1. Barnaba C., Sahoo B. R., Ravula T., Medina-Meza I. G., Im S.C., Anantharamaiah G. M.,
Waskell L., Ramamoorthy A. Cytochrome-P450-Induced Ordering of Microsomal Membranes
Modulates Affinity for Drugs. Angew. Chem. Int. Ed. Engl. 57(13):3391-3395. (2018)
2. Zhang, H., Myshkin, E., Waskell, L. Role of cytochrome b5 in catalysis by cytochrome P450
2B4. B.B.R.C. 2005, 338; 499-506.
3. Muralidhara, B. K., Halpert, J. R. (2007) Thermodynamics of ligand binding to P450 2B4 and
P450eryF studied by isothermal titration calorimetry. Drug. Metab. Rev. 39, 539-56.
4. Zhao, Y., White, M. A., Muralidhara, B. K., Sun, L., Halpert, J. R., Stout, C. D. (2006) Structure of microsomal cytochrome P450 2B4 complexed with the antifungal drug bifonazole: insight into
P450 conformational plasticity and membrane interaction. J. Biol. Chem. 281, 5973-81.
144
5. Gooley, P. R., Koay, A., Mobbs, J. I. (2018) Applications of NMR and ITC for the Study of the
Kinetics of Carbohydrate Binding by AMPK β-Subunit Carbohydrate-Binding Modules. AMPK:
Methods and Protocols, Methods in Molecular Biology, vol. 1732.
6. Schlegel, J., Armstrong, G. S. Redzic, J. S., Zhang, F., Eisenmesser, E. Z. (2009) Characterizing and controlling the inherent dynamics of cyclophilin-A. Protein Sci. 18, 811-824.
7. Ahuja, S. et al. A model of the membrane-bound cytochrome b5-cytochrome P450 complex from NMR and mutagenesis data. J. Biol. Chem. 288, 22080–95 (2013).
8.Yadav, R., Scott, E. E. (2018) Endogenous insertion of non-native metalloporphyrins into human membrane cytochrome P450 enzymes. J. Biol. Chem. 293, 16623-16634.
9. Chung, F. F., Mai, C. W., Ng, P. Y., Leong, C. O. (2016) Cytochrome P450 2W1 (CYP2W1) in Colorectal Cancers. Curr. Cancer Drug Targets 16, 71-8.
10. Guengerich, F. P. and Q. Cheng. (2011) Orphans in the Human Cytochrome P450 Superfamily:
Approaches to Discovering Functions and Relevance in Pharmacology. Pharmacol. Rev. 63, 684-
699.
11. Hardin, N. Z., Ravula, T., Di Mauro, G., Ramamoorthy, A. (2019) Hydrophobic
Functionalization of Polyacrylic Acid as a Versatile Platform for the Development of Polymer
Lipid Nanodisks. Small. 15, 1804813.
145
Appendix A
Supporting Information for Chapter 5
A.1. Supplementary Figures
A B C
D E F
Figure A.1. 15N-slices of residue Serine 69 reveal peak broadening and intensity decreases. (red) 1D projection of Serine 69 of 15N-cytb5 in 4F-DMPC nanodiscs. (orange) 1D projection of Serine 69 of 15N-cytb5 with cytP450 in 4F-DMPC nanodiscs. A. (black) 1D projection of Serine 69 of 15N-cytb5 with cytP450 and flFBD in 4F-DMPC nanodiscs. B. (light pink) 1D projection of Serine 69 of 15N-cytb5 with cytP450 and 4-CPI in 4F-DMPC nanodiscs. (dark pink) 1D projection of Serine 69 of 15N-cytb5 with cytP450 and 4-CPI and flFBD in 4F-DMPC nanodiscs. C. (yellow) 1D projection of Serine 69 of 15N-cytb5 with cytP450 and BFZ in 4F-DMPC nanodiscs. (gold) 1D projection of Serine 69 of 15N-cytb5 with cytP450 and BFZ and flFBD in 4F-DMPC nanodiscs. D. (light green) 1D projection of Serine 69 of 15N-cytb5 with cytP450 and BHT in 4F-DMPC nanodiscs. (dark green) 1D projection of Serine 69 of 15N-cytb5 with cytP450 and BHT and flFBD in 4F-DMPC nanodiscs. E. (light blue) 1D projection of Serine 69 of 15N-cytb5 with cytP450 and 1-CPI in 4F-DMPC nanodiscs. (dark blue) 1D projection of Serine 69 of 15N-cytb5 with cytP450 and 1-CPI and flFBD in 4F-DMPC nanodiscs. F. (light purple) 1D projection of Serine 69 of 15N-cytb5 with cytP450 and BZ in 4F- DMPC nanodiscs. (dark purple) 1D projection of Serine 69 of 15N-cytb5 with cytP450 and BZ and flFBD in 4F-DMPC nanodiscs.
146
A B C
D E F
Figure A.2. 15N-slices of residue Glycine 89 reveal peak broadening and intensity decreases. (red) 1D projection of Glycine 89 of 15N-flFBD in 4F-DMPC nanodiscs. (orange) 1D projection of Glycine 89 of 15N-flFBD with cytP450 in 4F-DMPC nanodiscs. A. (black) 1D projection of Glycine 89 of 15N-flFBD with cytP450 and cytb5 in 4F-DMPC nanodiscs. B. (light pink) 1D projection of Glycine 89 of 15N-flFBD with cytP450 and 4-CPI in 4F-DMPC nanodiscs. (dark pink) 1D projection of Glycine 89 of 15N-flFBD with cytP450 and 4-CPI and cytb5 in 4F-DMPC nanodiscs. C. (yellow) 1D projection of Glycine 89 of 15N-flFBD with cytP450 and BFZ in 4F-DMPC nanodiscs. (gold) 1D projection of Glycine 89 of 15N-flFBD with cytP450 and BFZ and cytb5 in 4F-DMPC nanodiscs. D. (light green) 1D projection of Glycine 89 of 15N-flFBD with cytP450 and BHT in 4F-DMPC nanodiscs. (dark green) 1D projection of Glycine 89 of 15N-flFBD with cytP450 and BHT and cytb5 in 4F-DMPC nanodiscs. E. (light blue) 1D projection of Glycine 89 of 15N-flFBD with cytP450 and 1-CPI in 4F-DMPC nanodiscs. (dark blue) 1D projection of Glycine 89 of 15N-flFBD with cytP450 and 1-CPI and cytb5 in 4F-DMPC nanodiscs. F. (light purple) 1D projection of Glycine 89 of 15N-flFBD with cytP450 and BZ in 4F-DMPC nanodiscs. (dark purple) 1D projection of Glycine 89 of 15N-flFBD with cytP450 and BZ and cytb5 in 4F- DMPC nanodiscs.
147
15 N-cytb5
15N-cytb5 + cytP450
Figure A.3: 2D 15N/1H TROSY HSQC spectra of 15N-labeled cytb5 in the presence and absence of cytP450. 2D 15N/1H TROSY HSQC NMR spectrum of (red) 15N-labeled cytb5 in 4F-DMPC Nanodiscs and (orange) 15N-labeled cytb5 in 4F-DMPC Nanodiscs with cytP450
148
15 N-cytb5 + cytP450 + 4-CPI
15N-cytb5 + cytP450 + 4-CPI + flFBD
15N-cytb5+cytP450 + 4-CPI 1.2 1 0.8 0.6 0.4 0.2
0
K7N-H
L28N-H L41N-H
S76N-H F79N-H
T70N-H T98N-H T13N-H E16N-H T26N-H Y35N-H T38N-H E49N-H E61N-H E64N-H
K10N-H K19N-H K94N-H
R52N-H R73N-H R89N-H
A55N-H
D58N-H
H22N-H H32N-H H44N-H H85N-H
G67N-H G82N-H
D133N-H N106N-H
Figure A.4: 2D 15N/1H TROSY HSQC NMR spectrum of (light pink) 15N-labeled cytb5 in 4F-DMPC Nanodiscs with cytP450 and 4-CPI and (dark pink) 15N-labeled cytb5 in 4F-DMPC Nanodiscs with cytP450, 4-CPI, and flFBD. The bar graph shows the relative signal intensity for each residue after the addition of P450, 4-CPI, and flFBD.
149
15 N-cytb5 + cytP450 + BFZ
15N-cytb5 + cytP450 + BFZ + flFBD
15N-cytb5+cytP450 + BFZ
1.2 1 0.8
0.6 0.4 0.2
0
K7N-H
L28N-H L41N-H
S76N-H F79N-H
T70N-H T98N-H T13N-H T26N-H Y35N-H T38N-H
E16N-H E49N-H E61N-H E64N-H
K10N-H K19N-H K94N-H
R52N-H R73N-H R89N-H
A55N-H
D58N-H
H22N-H H32N-H H44N-H H85N-H G82N-H
G67N-H
D133N-H N106N-H
Figure A.5: 2D 15N/1H TROSY HSQC NMR spectrum of (light yellow) 15N-labeled cytb5 in 4F-DMPC Nanodiscs with cytP450 and BFZ and (dark yellow) 15N-labeled cytb5 in 4F-DMPC Nanodiscs with cytP450, BFZ, and flFBD. The bar graph shows the relative signal intensity for each residue after the addition of P450, BFZ, and flFBD.
150
15 N-cytb5 + cytP450 + BHT
15N-cytb5 + cytP450
+ BHT + flFBD
15N-cytb5+cytP450 + BHT 1.2 1 0.8 0.6 0.4 0.2
0
K7N-H
L28N-H L41N-H
S76N-H F79N-H
T70N-H T98N-H T13N-H T26N-H Y35N-H T38N-H
E16N-H E49N-H E61N-H E64N-H
K10N-H K19N-H K94N-H
R52N-H R73N-H R89N-H
A55N-H
D58N-H
H22N-H H32N-H H44N-H H85N-H
G67N-H G82N-H
D133N-H N106N-H
Figure A.6: 2D 15N/1H TROSY HSQC NMR spectrum of (light green) 15N-labeled cytb5 in 4F-DMPC Nanodiscs with cytP450 and BHT and (dark green) 15N-labeled cytb5 in 4F-DMPC Nanodiscs with cytP450, BHT, and flFBD. The bar graph shows the relative signal intensity for each residue after the addition of P450, BHT, and flFBD.
151
15 N-cytb5 + cytP450 + 1-CPI
15N-cytb5 + cytP450 + 1-CPI + flFBD
15N-cytb5+cytP450 + 1-CPI 1.2 1 0.8 0.6 0.4 0.2
0
K7N-H
L28N-H L41N-H
S76N-H F79N-H
T70N-H T98N-H T13N-H T26N-H Y35N-H T38N-H
E16N-H E49N-H E61N-H E64N-H
K10N-H K19N-H K94N-H
R52N-H R73N-H R89N-H
A55N-H
D58N-H
H22N-H H32N-H H44N-H H85N-H
G67N-H G82N-H
D133N-H N106N-H
Figure A.7: 2D 15N/1H TROSY HSQC NMR spectrum of (light blue) 15N-labeled cytb5 in 4F-DMPC Nanodiscs with cytP450 and 1-CPI and (dark blue) 15N-labeled cytb5 in 4F-DMPC Nanodiscs with cytP450, 1-CPI, and flFBD. The bar graph shows the relative signal intensity for each residue after the addition of P450, 1-CPI, and flFBD.
152
15 N-cytb5 + cytP450 + BZ
15N-cytb5 + cytP450 + BZ + flFBD
15N-cytb5+cytP450+b5 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2
0
K7N-H
L28N-H L41N-H
F79N-H S76N-H
T13N-H E16N-H T26N-H Y35N-H T38N-H E49N-H E61N-H E64N-H T70N-H T98N-H
K10N-H K19N-H K94N-H
R52N-H R73N-H R89N-H
A55N-H
D58N-H
H22N-H H32N-H H44N-H H85N-H
G67N-H G82N-H
D133N-H N106N-H
Figure A.8: 2D 15N/1H TROSY HSQC NMR spectrum of (light purple) 15N-labeled cytb5 in 4F-DMPC Nanodiscs with cytP450 and BZ and (dark purple) 15N-labeled cytb5 in 4F-DMPC Nanodiscs with cytP450, BZ, and flFBD. The bar graph shows the relative signal intensity for each residue after the addition of P450, BZ, and flFBD.
153
15N-cytb5 15N-cytb5 + cytP450
15N-cytb5 + cytP450 + flFBD
15N-cytb5+cytP450+b5 1.4 1.2 1 0.8 0.6 0.4 0.2
0
K7N-H
L28N-H L41N-H
S76N-H F79N-H
T70N-H T98N-H T13N-H E16N-H T26N-H Y35N-H T38N-H E49N-H E61N-H E64N-H
K10N-H K19N-H K94N-H
R52N-H R73N-H R89N-H
A55N-H
D58N-H
H22N-H H32N-H H44N-H H85N-H
G67N-H G82N-H
D133N-H N106N-H
Figure A.9: 2D 15N/1H TROSY HSQC NMR spectrum of (red) 15N-labeled cytb5 in 4F-DMPC Nanodiscs (orange) 15N-labeled cytb5 in 4F-DMPC Nanodiscs with cytP450 (tan) 15N-labeled cytb5 in 4F-DMPC Nanodiscs with cytP450 and flFBD. The bar graph shows the relative signal intensity for each residue after the addition of (orange) P450 and (black) flFBD.
154
15N-flFBD
Figure A.10: 2D 15N/1H TROSY HSQC NMR spectrum of (red) 15N-labeled flFBD in 4F-DMPC Nanodiscs.
15N-flFBD + cytP450
Figure A.11: 2D 15N/1H TROSY HSQC NMR spectrum of (orange) 15N-labeled flFBD and cytP450 in 4F-DMPC Nanodiscs.
155
15N-flFBD + cytP450 + 4-CPI
15N-flFBD + cytP450 + 4-CPI + cytb5
Figure A.12: 2D 15N/1H TROSY HSQC NMR spectrum of (light pink) 15N-labeled flFBD, cytP450, and 4-CPI in 4F-DMPC Nanodiscs and (dark purple) 15N-labeled flFBD, cytP450, 4-CPI, and cytb5 in 4F-DMPC Nanodiscs.
156
15N-flFBD + cytP450 + BFZ
15N-flFBD + cytP450 + BFZ + cytb5
Figure A.13: 2D 15N/1H TROSY HSQC NMR spectrum of (light yellow) 15N-labeled flFBD, cytP450, and BFZ in 4F-DMPC Nanodiscs and (dark yellow) 15N-labeled flFBD, cytP450, BFZ, and cytb5 in 4F-DMPC Nanodiscs.
157
15N-flFBD + cytP450 + BHT
15N-flFBD + cytP450 + BHT + cytb5
Figure A.14: 2D 15N/1H TROSY HSQC NMR spectrum of (light green) 15N-labeled flFBD, cytP450, and BHT in 4F-DMPC Nanodiscs and (dark green) 15N-labeled flFBD, cytP450, BHT, and cytb5 in 4F-DMPC Nanodiscs.
158
15N-flFBD + cytP450 + 1-CPI
15N-flFBD + cytP450 + 1-CPI + cytb5
Figure A.15: 2D 15N/1H TROSY HSQC NMR spectrum of (light blue) 15N-labeled flFBD, cytP450, and 1-CPI in 4F-DMPC Nanodiscs and (dark blue) 15N-labeled flFBD, cytP450, 1-CPI, and cytb5 in 4F-DMPC Nanodiscs.
159
15N-flFBD + cytP450 + BZ
15N-flFBD + cytP450 + BZ + cytb5
Figure A.16: 2D 15N/1H TROSY HSQC NMR spectrum of (light purple) 15N-labeled flFBD, cytP450, and BZ in 4F-DMPC Nanodiscs and (dark purple) 15N-labeled flFBD, cytP450, BZ, and cytb5 in 4F-DMPC Nanodiscs.
160
15N-flFBD + cytP450 + cytb5
Figure A.17: 2D 15N/1H TROSY HSQC NMR spectrum of (coral) 15N-labeled flFBD, cytP450, and cytb5 in 4F-DMPC Nanodiscs.
161
3 2.5
2.5 A 2 B 2 1.5 1.5 1 1 0.5 0.5 0 0 Loop 1 Loop 2 Loop 3 Loop 4 Loop 1 Loop 2 Loop 3 Loop 4
3 4 2.5 C 3.5 D 3 2 2.5 1.5 2 1 1.5 1 0.5 0.5 0 0 Loop 1 Loop 2 Loop 3 Loop 4 Loop 1 Loop 2 Loop 3 Loop 4
2 2.5 E 2 F 1.5 1.5 1 1 0.5 0.5
0 0 Loop 1 Loop 2 Loop 3 Loop 4 Loop 1 Loop 2 Loop 3 Loop 4
Figure A.18 – Loop Regions of 15N-flFBD reveal binding and dissociating of flFBD with cytP450. The average signal intensity of residues in Loop 1 (G85-T90), Loop 2 (Y140-D144), Loop 3 (G174-E179), and Loop 4 (D207-L212) show decreases in intensity when flFBD binds tighter to cytP450 and increases in intensity when free in solution after having its complex weakened. For all graphs, orange is the complex of flFBD and cytP450. A. (light pink) flFBD and cytP450 with 4-CPI (dark pink) flFBD and cytP450 with 4-CPI and cytb5. B. (light yellow) flFBD and cytP450 with BFZ (dark yellow) flFBD and cytP450 with BFZ and cytb5. C. (light green) flFBD and cytP450 with BHT (dark green) flFBD and cytP450 with BHT and cytb5. D. (light blue) flFBD and cytP450 with 1-CPI (dark blue) flFBD and cytP450 with 1-CPI and cytb5. E. (black) flFBD and cytP450 with cytb5. F. (light purple) flFBD and cytP450 with BZ (dark purple) flFBD and cytP450 with BZ and cytb5.
162