1. science
2. medicine

Personalizing treatment using biomarkers to
predict the activity of human carboxylesterase 1
Bachelor in Natural Science – 5th semester project – Roskilde University - 2014.
Authors: Karen Gammeltoft, Lotte Kofod, Sebastien Heimburger, Beatriche Henriksen and
Laura Niclasen.
Supervisor: Henrik Berg Rasmussen
Front-page picture: illustration of hCES1 as a monomer. The light red area is the N-terminal
domain, the light blue area is the C-terminal domain, and the light yellow area is the αβ-domain.
The Z-site is illustrated as an electron cloud around a taurocholate molecule. The active site is
illustrated as a green colored electron cloud around another taurocholate molecule. The green side
chains are the side chains, which forms the catalytic triad (SER221, GLU354 and HIS468). The
dark green colored areas are the catalytic pocket. The red and dark red colored areas are active sites,
found by two docking analyses with nelfinavir and morphine, respectively. The area pointed out
with a blue circle is the substrate-binding gorge opening (Bencharit et al., 2006;Rhoades et al.,
2012;Vistoli et al., 2010). The data for the illustration is the .pdb files 2DR0 (hCES1 taurocholate
complex) and 4AB1 (hCES1 with no substrates) from RCSB.org.
Preface
Writing scientific reports is an important part of academic work so we decided to write our fifth
semester project as a scientific literature report containing all the theoretical background with a
docking experiment. The topic, personalizing treatments using biomarkers to predict the activity of
human carboxylesterase 1 (hCES1), was chosen due to our interest in the field of Medical Biology
and due to enzyme activity of hCES1 being a key factor in drug metabolism in the liver. It caught
our interest that the enzyme activity seems to be imperative for the treatment results which is the
reason as to why we looked into the different metabolizer profiles with the aim to be able to give
patients the right dose of medicine.
Abstract
Personalized medicine is an evolving medical field aiming to customize treatments through genetic
profiling, resulting in an improved therapeutic response and minimizing adverse effects.
Pharmacokinetics refers to the body’s response to xenobiotics. Different responses can be due to
variations in the genes coding for drug-metabolizing enzymes. Hence, absent, increased or
decreased enzyme expression can be caused by gene polymorphisms or copy number variants. To
study the altered efficiency of hCES1 metabolism due to pharmacokinetic changes, it is essential to
look at the cytochrome P450 (CYP) enzymes. In drug metabolism, six CYP enzymes, including
CYP2D6 (2D6) and CYP3A4 (3A4), play an imperative role and account for most of the
metabolism of widely prescribed drugs metabolized in the liver. In this study, 2D6 and 3A4 are
used as models for human carboxylesterase 1 (hCES1) to explain how the properties of hCES1 can
be used to benefit the personalization of medicine. hCES1 metabolizes both endogenous and
xenobiotic compounds. It is expressed in various tissues though mostly in the liver. The CES1 gene
expression seems to be surprisingly constant between individuals in the human population though
the enzymatic hCES1 activity differs significantly. This can be due to hCES1 SNPs that are
identified and classified as possibly damaging to protein structure and function or due to mutations
in the hCES1 gene resulting in a considerably decreased catalytic function of the hCES1 enzyme.
hCES1 activity is regulated at genome and enzyme level, exhibiting behavior from both CYP
models. There are not yet any probes or biomarkers applicable for clinical use, since the one hCES1
probe that is found, is only tested in hepatic cells. More research is needed, in order to find a probe
or biomarker that is more accessible. Another approach could be to apply metabolomics, to
determine the metabotype of a patient.
Table of contents
Key terms ............................................................................................................................................. 1
Abbreviations ....................................................................................................................................... 1
Introduction .......................................................................................................................................... 2
Key Issue .............................................................................................................................................. 3
Methods ................................................................................................................................................ 3
Theoretical ........................................................................................................................................... 4
Personalized Medicine ..................................................................................................................... 4
Biomarkers ....................................................................................................................................... 4
Pharmacokinetics and pharmacodynamics ....................................................................................... 5
Pharmacokinetics .......................................................................................................................... 5
Absorption .................................................................................................................................... 6
Distribution ................................................................................................................................... 7
Metabolism ................................................................................................................................... 7
Excretion ....................................................................................................................................... 9
Fast metabolizers and slow metabolizers ..................................................................................... 9
Pharmacodynamics ..................................................................................................................... 10
Combining pharmacokinetics and pharmacodynamics to personalize medical treatments ........ 10
The therapeutic window ................................................................................................................. 11
Cytochrome P450 enzymes and their properties ............................................................................ 12
CYP2D6...................................................................................................................................... 14
The 3D structure of 2D6 ............................................................................................................. 16
CYP3A4 ...................................................................................................................................... 17
CYP2D6 and 3A4 as models ...................................................................................................... 20
Metabolomics ............................................................................................................................. 22
Carboxylesterases and their activity ............................................................................................... 22
Drug Hydrolysis.......................................................................................................................... 24
Substrates .................................................................................................................................... 25
Role of CES’s in xenobiotic metabolism.................................................................................... 27
Human Carboxylesterase 1 ............................................................................................................. 27
Genetic polymorphism of carboxylesterase 1............................................................................. 29
The structure of hCES1 .............................................................................................................. 31
The role of CES1 in metabolism of endogenous compounds..................................................... 33
Allosteric functions of hCES1 .................................................................................................... 35
Discussion .......................................................................................................................................... 36
Adverse effects ............................................................................................................................... 36
Alleles’ and SNPs’ influence on enzyme activity .......................................................................... 37
Conformational changes affecting function and activity ............................................................... 38
Inhibitors and inducers ................................................................................................................... 39
The complexity of finding probes and biomarkers for the activity of hCES1 ............................... 40
Probes for hCES1 ........................................................................................................................... 41
Metabolomics ................................................................................................................................. 41
Conclusion ......................................................................................................................................... 43
Perspective...................................................................................................................................... 44
Acknowledgements ............................................................................................................................ 44
Supplements ....................................................................................................................................... 45
Docking experiment ....................................................................................................................... 45
Introduction................................................................................................................................. 45
Methods ...................................................................................................................................... 45
Results......................................................................................................................................... 46
Binding sites ............................................................................................................................... 47
Discussion ................................................................................................................................... 49
References .......................................................................................................................................... 50
Appendices ......................................................................................................................................... 65
Appendix #1 ................................................................................................................................... 65
Appendix #2 ................................................................................................................................... 65
Appendix #3 ................................................................................................................................... 66
Appendix #4 ................................................................................................................................... 68
Appendix #5 ................................................................................................................................... 69
Key terms
Biomarkers, Cytochrome P450, Cytochrome P450 2D6, Cytochrome P450 3A4, Drugs,
Endogenous, Exogenous, Human Carboxylesterase 1, Human Carboxylesterase 2, Hydrolysis,
Inducers, Inhibitors, Metabolism, Pharmacokinetics, Pharmacodynamics, Pro-drugs, Substrates
Abbreviations
ABC Transporters: ATP-Binding Cassette Transporters, ACAT: Acyl-Coenzyme A Cholesterol
Acyltransfer, ACE: Angiotensin-Converting Enzyme, ADHD: Attention deficit hyperactive
disorder, ADR: Adverse Drug Response, ARB: Angiotensin Receptor Blocker, AUC: Area under
the Curve, CE: Cholesteryl ester, CEH: Cholesteryl ester, CES: Carboxylesterase, CNS: Central
Nervous System, CNV: Copy Number Variation, CYP: Cytochrome P450, CYP2D6: Cytochrome
P450 2D6, CYP3A4: Cytochrome P450 3A4, FACoAH: Fatty Acyl CoA, hCES1: Human
Carboxylesterase 1, hCES2: Human Carboxylesterase 2, hCES3: Human Carboxylesterase 3, KD:
knock down, SLC transporters: Solute-Carrier Transporters, SNP: Single Nucleotide
Polymorphism,
SN-38:
7-ethyl-
10-hydroxy
camptothecin,
HMBT:
2-(2-hydroxy-3-
methoxyphenyl) benzothiazol.
Page 1 of 75
Introduction
People being treated for the same disease with the same drug will often experience different
responses. The response can vary widely, from no response to a satisfying therapeutic response,
from no adverse drug effects to fatal poisoning and all in between. It is therefore important to
customize treatments in order for it to match specific individuals. This is personalized medicine.
All types of medical treatments are to some degree personalized because they are administered
following the clinical criteria. This criterion focuses on physiologic properties such as age, gender,
weight and clinical manifestations including symptom severity, which plays a role in determining
dose and medicine type. Administrating medication by following the clinical criteria increases the
number of patients with a therapeutic response and limits the number of patients experiencing
adverse effects. However, further personalization of medication is still necessary to improve the
responses and efficacy. Besides the clinical criteria, genetic variation contributes to the response
variations. Advancements in technology provide the opportunity to consider the genetic variations
that affect drug responses. In addition, it might be possible to find biomarkers that can reveal these
variations in patients allowing medical treatments to be further personalized.
This project aims to contribute to the further personalization of medical treatments by looking at
pharmacokinetics and identifying biomarkers for the activity of the drug metabolizing enzyme
human carboxylesterase 1 (hCES1). hCES1 does not metabolize a lot of different drugs, but the
drugs it does metabolize are commonly used. The adverse effects patients experience when using
hCES1 metabolized drugs are not severe, but the patients often use the drugs for long periods of
time, and experience a lowered quality of life than patients that are processing the drug at a normal
rate. This study will investigate the metabolism of hCES1 and compare it to the well-researched
metabolic mechanisms of cytochrome P450 (CYP) enzymes. Since hCES1 is not fully investigated,
CYP 3A4 and 2D6 are used as models to enlighten the properties of hCES1. Furthermore, a simple
in silico docking experiment is carried out, as a supplement, testing the affinity of cholesterol, 27hydroxycholesterol and methylphenidate to hCES1. This is done to try and develop a method for
selecting compounds interesting for in vitro and in vivo testing.
Personalized medicine, also known as individualized medicine, is an evolving medical field that
aims to customize medical treatments through an individual's genetic profile. This method will
assist to navigate medical decisions concerning diagnosis, prevention and treatment (Collins, 2014).
Contrary to trial and error approach of certain conventional therapies, the goal of personalized
Page 2 of 75
medical treatments is a more accurate match of drugs to patients, by giving the right treatments to
the right patients at the right time. Individualizing the medication dose and treatment, can help
prevent adverse drug response (ADR) and improve the therapeutic response. With multiple
technologies such as genotyping, more accurate medical diagnostics and proteomics along with
advances in medical genetics and sequencing of the human genome, personalization of medical
treatments is obtainable (KK, 2009). Medicine targeting have already shown progress in the field
of cancer. DNA methylation in cancer cells (Momparler and Bovenzi, 2000) and pro-drugs that are
capable of inducing apoptosis in specific cells by cell cleavage mechanisms (Denmeade et al.,
2012;Piccioni and Nguyen, 2014) are all a step in the right direction to take personalized medicine
to a new level.
Key Issue
What causes the individual variation in hCES1’s activity, and which biomarkers can be used as a
proxy for hCES1 activity?
Objective:
Investigation of the causes of individual variation in the hCES1 activity, and identification of easily
measurable and accessible biomarkers, in blood, that can be used as a proxy for determining the
hCES1 activity.
Methods
We completed a literature review with the aim to collect information on pharmacokinetics -and
dynamics, CYP genes with focus on genotypes 2D6 and 3A4 and carboxylesterases more specific
human carboxylesterase 1. Scientific textbooks and browsers such as Google Scholar and PubMed
were used to gather information. Additionally, we used public information websites to find statistics
concerning personalized treatment. SwissPdbViewer is used to visualize protein structure. Enzymes
are depicted as monomers, even though some of them will make a trimer when interacting with a
substrate. This is only to simplify the illustration. SwissDock was used for different docking
analyses in order to study the affinity of endogenous inhibitors and xenobiotic substrates.
Additionally it was used to formulate a theory distinguishing the above mentioned in the search for
a hCES1 biomarker. The .pdb file 4AB1 was dock prepped using Chimera and the results from the
different docking analyses were downloaded using the Chimera ViewDock tool.
Page 3 of 75
Theoretical
Personalized Medicine
Personalization of all medical treatments is achieved by following the clinical criteria based on the
physiologic traits of the patient. Further personalization of medical treatments beyond the
physiology, the clinical criteria makes it possible for pharmaceutical companies to create new
medicaments that are more effective and has fewer adverse effects.
The term "genomic medicine" (precision medicine) suggests that by sequencing the human genome,
the practice of medicine has been given the opportunity to enter a new era in which an individual
patient's genome can contribute to identify the most favorable approach to care, whether it is
diagnostic, preventive or therapeutic. As other factors such as metabolomics and proteomic
technologies are taken into consideration in personalized medicine, genomic medicine is not a
sufficient synonym. Precision medicine has proven to be optimal for the integration of new
biotechnologies into medicine, which improves the management of patients as well as the
understanding of pathomechanism of diseases (KK, 2009).
Today many available technologies can determine genetic polymorphisms such as single nucleotide
polymorphisms (SNP) and copy number variations (CNV) in a genome with precision and high
efficiency. This allows scientists to explore differences in the human genome and thereby come up
with a better-developed hypothesis in which the human genomes as well as levels of endogenous
substrates are taken into account. Furthermore, next-generation sequencing is considered applicable
for direct diagnostics in the near future (Meldrum et al., 2011). Additionally it will allow
individuals to better maintain their health on account of understanding their genetic profile.
Biomarkers
A biomarker is a measurable parameter that indicates the development of a disease or the presence
of biological changes. An example is raised body temperature during a cold or cholesterol in
determining a patient’s change of getting coronary or vascular diseases. Biomarkers can be drugrelated or disease related. Disease related biomarkers are used to depict something about the disease
as well as the state of the disease by finding a biomarker that indicates, whether the disease
proceeds or comes to a halt (Mikeska and Craig, 2014).
Page 4 of 75
Drug-related biomarkers are used to predict the effects of a treatment and to examine how the drug
is metabolized within the individual. Additionally it is often used to indicate changes in the
metabolic processes (Mikeska and Craig, 2014). This project focusses on biomarkers, providing
information of drug-metabolizing enzymes’ activity, a so-called drug-related biomarker.
For both disease and drug-related biomarkers, there is a required evaluation process for the data, to
establish the level of confidence in the biomarker the scientist needs to get it approved. First, a
molecule, metabolite or drug that can be used as a biomarker is found. Then it is hypothesized how
it can be used as a biomarker, for a certain mechanism/disease. When this hypothesis is made,
several kinds of experiments using animal models are done. One of these is to detect if the
biomarker is suitable and targets the right mechanism/disease and another is to confirm that the
biomarker is not affected by any other mechanisms/diseases. Once this is done, the biomarker is
used on human models. One to confirm the results from the above mentioned first animal model,
and then another confirming the second above mentioned. When the biomarker is drug related it has
to be linked to the drug(s) outcome with the same mechanism of action. At last the biomarkers’
effect is linked to the outcome for drugs with differing mechanisms (Peck, 2007;Yi et al., 2006)
Pharmacokinetics and pharmacodynamics
The same drug often results in different responses when given to different patients with the same
disease. Patients are divided into four groups, according to their response:
1. Therapeutic response and no ADR
2. Therapeutic response and ADR
3. No therapeutic response and ADR
4. No therapeutic response and no ADR
To understand why the same drug can result in different outcomes, it is essential to look at how the
body processes drugs. There are two aspects to consider; pharmacokinetics and pharmacodynamics.
Pharmacokinetics has to do with what the body does to the drug, whereas pharmacodynamics has to
do with what drug does to the body. Both play a role in determining the response, meaning both
therapeutic and adverse drug response.
Pharmacokinetics
Pharmacokinetics has four steps: Absorption, distribution, metabolism and excretion. As seen in
Figure 1, the two middle steps, distribution and metabolism, occur simultaneously.
Page 5 of 75
Figure 1: Illustration of the four steps in pharmacokinetics: absorption - where drugs are moved from where it is
administrated into circulation, distribution - where drugs are dispersed to different body tissues, metabolism - where drugs
are transformed to smaller metabolites and excretion - where drug metabolites are removed from the body.
Absorption
Drug absorption is the process in which the drug moves from where it is administrated, e.g. orally in
the form of a tablet, and into circulation. Most routes of administration go through biological
membranes in the body. Many drugs are absorbed by passive diffusion, either lipid diffusion or
aqueous diffusion. Lipid soluble drugs are absorbed by lipid diffusion across the plasma membrane.
Water-soluble drugs can be absorbed by aqueous diffusion through aquaporins in the plasma
membrane. Aqueous diffusion can only take place when drug molecules are of a low molecular
weight (Brenner and Stevens, 2009).
Some drugs need a carrier protein to be absorbed. Active transport is a form of administration that
both requires a carrier protein and energy produced by the hydrolysis of ATP. In facilitated
diffusion only a carrier protein is needed, no energy (Brenner and Stevens, 2009).
Membrane transporter systems function as gatekeepers that either enhance or inhibit the transport of
drugs across the membrane. The membrane transporters can be divided into two groups; influx and
efflux transporters. Influx transporters facilitate the uptake of drugs through membrane barriers,
where efflux transporters inhibits drug passing through the membrane (Girardin, 2006).
Page 6 of 75
The process of absorption is irrelevant with topically administrated drugs, where drugs are directly
administrated on the target tissue, and intravenously administrated drugs, where drugs are
administrated directly into the blood stream (Brenner and Stevens, 2009).
Distribution
After the drug is absorbed, it is distributed to the body’s tissues. Distribution can happen through
lipid diffusion if the drug is sufficiently lipid soluble or through active transport. The distribution
depends on various parameters such as molecular size, lipid solubility and perfusion, which is the
movement of blood through the system with exchange of oxygen and solutes. The distribution rate
is faster in areas of the body that are highly perfused, such as the brain, heart, liver and kidneys.
Molecular size and lipid solubility affect the distribution rate because these affect the distribution in
different tissues (Brenner and Stevens, 2009).
Transport proteins can affect the distribution rate by moving endogenous substances or drugs across
biological membranes. They can be divided into two groups: the ATP-binding cassette transporters
(ABC transporters) and solute-carrier transporters (SLC transporters). ABC transporters use the
hydrolysis of ATP to drive the transportation of substances while the SLC transporters move ions
and organic substances across biological membranes. Additionally, SLC transporters play an
important role in the transport of several drugs.
Metabolism
As drugs are distributed through the circulatory system, they are continuously metabolized. The
first breakdown of exogenous compounds happens in the small intestines, though the majority of
metabolism occurs in the liver. Approximately 73% of the 200 most frequently used drugs in USA
in 2002 were metabolized by the liver and the liver detoxification pathways (Guengerich,
2008;Williams et al., 2004). Drugs are broken down by both human carboxylesterase 1 (hCES1)
and human carboxylesterase 2 (hCES2) in the intestine (Imai et al., 2006) but the hydrolase activity
is mainly provided by hCES2. hCES1 contributes with some hydrolase activity too, however, the
main role of hCES1 in drug metabolism is in the liver (Imai et al., 2006;Satoh et al., 2002).
Cytochrome P450 3A4 (CYP 3A4) is also found in the intestine (Kolars et al., 1992;Watkins et al.,
1987) and contributes significantly to the first-pass metabolism of a couple of intestinally dosed
drugs (Kolars et al., 1991;Paine et al., 1996).
Page 7 of 75
Through the liver detoxification pathways, both endogenous substances such as hormones and
xenobiotics such as drugs are eliminated. These detoxification pathways consist of enzymatic
processes, which inactivate the compounds and make them more hydrophilic, and they are thereby
easily excreted from the body. The process of making the compounds more hydrophilic happen in
two ways, phase I- and phase II-reactions. Additionally metabolism of compounds undergoes either
phase I-, phase II-reactions or both, as illustrated in Figure 2
Figure 2: Illustration of phase I and phase II drug metabolism. Xenobiotic compounds are destined to be inactivated and
broken down by undergoing either phase I and/or phase II reactions. These metabolic reactions are catalyzed by a variety of
enzymes that are primarily expressed in the liver. Modified from (Grant, 1991).
In phase I-reactions, the substrates are oxidized, resulting in either a newly formed polar functional
group or the exposure of an already existing polar functional group. Frequently the polar group
added is a hydroxyl- or an epoxy group, which often is enough to inactivate the compound and by
that allows it to be excreted. Phase I drug-metabolizing enzymes hydrolyze various ester-containing
drugs and pro-drugs such as anti-tumor drugs (capecitabin and CPT-11), narcotics and angiotensinconverting enzyme (ACE) inhibitors (cilazapril, quinapril, temocapril and imi- dapril) (Satoh and
Hosokawa, 2010). Most phase I-reactions are catalysed by the cytochrome P450 (CYP) enzymes,
which are located in the endoplasmic reticulum of the hepatocytes (Borup, 2010;Grant, 1991).
Another drug metabolizing enzyme family is the carboxylesterases (CESs). Tissue expression
profiles of CESs show that the highest activity levels are found in the mammalian liver microsomes
(Satoh and Hosokawa, 2010).
Page 8 of 75
In phase II-reactions, the compounds conjugate with a hydrophilic group which radically changes
the structure, hence completely loses its actual function. The phase II-reactions are performed by
various substrate specific enzymes, and take place in the cytosol, mitochondria and in the
endoplasmic reticulum (Borup, 2010).
Excretion
Drug excretion is the removal of the drug from the body. This happens primarily through the urine
but excretion can also happen through feces, bile, sweat, tears, breast milk, saliva, and through
exhalation. When a drug is excreted through the urine, it is filtered in the glomeruli in the kidneys.
Some drugs can be reabsorbed, depending on its lipid solubility. Reabsorption increases with an
increasing lipid solubility and decreases when drugs are transformed into polar metabolites during
metabolism. The body cannot excrete drugs that have a high molecular weight nor polar and
nonpolar groups through the urine as these drugs only are excreted through the bile.
Fast metabolizers and slow metabolizers
When patients experience different adverse effects as response to the same drug, it may be caused
by different variations in genes coding for drug-metabolizing enzymes. Polymorphisms in genes
can cause increased, decreased or absent enzyme expression (Meyer and Zanger, 1997). Since the
CYP enzymes metabolize over 70%, and esterase enzymes metabolize about 10% of the 200 most
frequently used drugs in USA in 2002 (Williams et al., 2004), polymorphisms in the genes coding
for these have a great impact on a patient’s drug metabolism.
A small-scale variation is SNPs where a single nucleotide differs. If SNPs are located in DNA
coding for a drug-metabolizing enzyme it might change the function of the enzyme. Another type of
genetic variation is structural variation, where more base pairs are involved. A large subgroup in
structural variations is CNVs, which include insertions, deletions and duplications. If a patient has a
genetic variant in which the gene for a drug-metabolizing enzyme is duplicated, it will result in a
faster metabolism of the drug than average and the patient therefore needs a higher dose than other
patients in order to get a therapeutic response. On the other hand, patients with deletions or nonconservative mutations in genes coding for drug metabolizing enzymes have a slower or
nonfunctional metabolism, and thereby a higher concentration of the drug in the blood. These
patients will be at a higher risk of experiencing adverse effects.
Page 9 of 75
This project focusses on metabolism as the primary factor in pharmacokinetics, based on the
assumption that the metabolism, more specific the activity of drug metabolizing enzymes, is the
primary factor in drug concentration. This assumption is based on the fact, that the activity of drug
metabolizing enzymes has a direct impact on the drug concentration in the blood. Absorption,
distribution and excretion also influence the concentration of drug exposure in the body but
variations in these systems are based on the chemical composition of the medication. In order to
personalize treatments based on the variations in absorption, distribution and excretion, new types
of medication has to be defined and developed. It is assumed that the medications available today is
effectively absorbed, distributed or excreted by most patients, and that there is an alternative to
those who cannot. This makes the variations in these three factors less relevant when looking at the
specific drug concentration in a patient.
Pharmacodynamics
Pharmacodynamics is a field of medical and molecular biology, which explains what certain drugs
do to the body. Due to the fact that different drugs affect the body in the same intended way, a
method is needed to compare the intended effects and adverse effects a drug have in a patient. This
is done by comparing the effectiveness of different drugs, including exposure as well as the
intensity of therapeutic effects and adverse effects (DiPiro and Sprull, 2010).
The factors influencing the efficiency of a drug varies in humans due to genetic variations that
affect the individual response. This could be a higher receptor density on the cell surface, which
leads to an individual with a higher responsiveness. Furthermore, if genetic variations lead to an
alteration in the conformation of the receptor and thereby lowers the affinity for a drug it may lead
to altered pharmacodynamics. Genetic variations and all the factors that affect the drug’s effect on
the site of action will be an alteration in pharmacodynamics (Meyer and Zanger, 1997).
Methods for choosing the right drug to treat a patient do not consider effect duration, exposure
effectiveness or the adverse effects the drug can lead to over time; it is only a valid method for
preliminary differentiation between drugs (DiPiro and Sprull, 2010).
Combining pharmacokinetics and pharmacodynamics to personalize medical treatments
To investigate why patients experience different responses to the same dose of the same drug,
metabolism and receptor affinity has to be considered. The metabolism, as mentioned earlier,
affects the exposure a certain drug in the system and the concentration of that drug during exposure.
Page 10 of 75
The receptor affinity influences the concentration needed to achieve the intended therapeutic effect.
This project combines metabolism and receptor affinity in four new patient groups, defined in this
project as the patient’s turnover profile:
1. Fast metabolizer and high receptor affinity: The easiest to medicate because the high
receptor affinity makes up for the fast metabolism.
2. Fast metabolizer and low receptor affinity: Has the highest chance of being over-medicated,
as the fast metabolism will lead to a low drug concentration in the blood while the receptors
cannot use the drug efficiently.
3. Slow metabolizer and high receptor affinity: Easy to treat, as the drug concentration does
not have to be high.
4. Slow metabolizer and low receptor affinity: Has an increased chance of being overmedicated, as this person will require a high drug concentration to achieve the wanted effect.
The outcome of the turnover profiles mentioned above are all based on the previous theory and on
“normal” medicaments, not pro-drugs. It will be possible to prevent over-medication of a patient if
the metabolism can be determined before treating them. For example, if the metabolism of a patient
with turnover profile number three can be determined before treating the patient, one knows that the
patient will have a higher concentration in a longer exposure period than a patient with turnover
profile number one. This will lead to a treatment where the dose will be lowered from the start.
The therapeutic window
When considering personalization of treatments in which patients experience different responses, a
better understanding of a patient’s individual metabolism can help find the best fitting dose. This
will ensure the best possible therapeutic response and fewest possible adverse effects.
Adverse effects can be dose-dependent or non-dose-dependent. Non-dose-dependent adverse effects
are not relevant when focusing on metabolism as they are independent of metabolism and often
occur due to allergies to the drug or metabolites of the drug. Since the focus of this project is on
metabolism, the dose-dependent adverse effects are of interest.
The therapeutic window is the difference between the therapeutic dose and the toxic dose. The best
response is acquired if the actual drug dose is between these two thresholds, as seen in Figure 3. If
the therapeutic window is narrow it is imperative to find the correct dose, as a slight increase in
dose can be enough to cause adverse effects and a slight decrease in dose will result in loss of
Page 11 of 75
therapeutic response. If the therapeutic window is wider, a therapeutic effect without adverse effects
is obtainable with a less precise dose. Personalized medicine focusses on drugs that have a narrow
therapeutic window.
Figure 3: Therapeutic Window: The curve illustrates the plasma concentration over time. The redline indicate the plasma
concentration level where it becomes toxic, and the green line represent the theraputic dose.
As with adverse effects, the therapeutic response can also be dose-dependent. If patients are treated
with a drug in which both the therapeutic response and adverse effects are dose-dependent, patients
receiving a higher dose will have a better therapeutic effect but also be at higher risk of adverse
effects.
Cytochrome P450 enzymes and their properties
To study the pharmacokinetic changes resulting in an altered efficiency of hCES1 metabolism, it is
necessary to look at CYP enzymes and compare the two distinct regulation mechanisms. An
understanding of the term etiology is necessary in this context. Etiology is the study of the causation
or origin of a disease. The definition of etiology used today arose when smoking was shown to
cause cancer. It can be used to determine diseases and even to distinguish between different
diagnoses (Green, 2007;Oxford, 2014). This ability to distinguish between diseases by using
biomarkers opens up the opportunity to use the right treatment on the right individual. As
mentioned above the focus will be on the metabolism of drugs. The metabolism of medicaments is
an already investigated field. Six CYP enzymes play a crucial role in the metabolism of drugs;
CYP1A2, 2A6, 2C8, 2C9, 2D6 and 3A4. These account for about 75% of drug metabolism in the
liver (Guengerich, 2008). As seen in Figure 4, CYP enzymes are structurally very similar but there
Page 12 of 75
are some differences in the active site, the conformational change when interacting with substrates
and in their F’ and G’ domains (Guengerich, 2008).
Figure 4: Illustration of two important CYP enzymes. The yellow or orange areas are the N-terminal domain, the light purple or dark purple
areas are the C-terminal domain. The green colored areas are the F-helix domains, and the blue colored areas are the G-helix domains. The
red colored molecule in the middle of the individual enzyme is the heme group (de Groot et al., 1999;Ekroos and Sjogren, 2006;Lewis,
2002;Lewis, 2003;Raunio and Rahnasto-Rilla, 2012;Scott and Halpert, 2005;Williams et al., 2004)]. The illustration is based on .pdb files
2F9Q (CYP2D6) and 1TQN (CYP3A4) from RSCB.org.
As with many other genes, the CYP genes are found in various isoforms in the population. These
genetic variations give rise to different metabolizing phenotypes, which becomes apparent when
using biomarkers in activity assessments.
The important role of metabolism and receptors culminating in what this report refers to as a
“turnover profile” and other important factors are discussed in the earlier section; Combining
pharmacokinetics and pharmacodynamics to personalize medical treatment. In this part of the
project, these factors are not accounted for as CYP enzymes make up 75% of liver metabolism.
Many substrates, which are useful in estimating the activity of a particular CYP enzyme, have been
identified for the CYP enzymes 3A4, 1A2, 2C9, 2D6 and 2C19, as well as inhibitors and inducers,
see Table 1 (Guengerich, 2008). The focus will be on the well-studied CYP enzymes CYP2D6
(2D6) and CYP3A4 (3A4). Tabel 1 clearly indicate that 2D6 have a variety of substrates that cause
inhibition of the enzyme, whereas 3A4 was shown to have fewer inhibitors, but exist in different
forms that changes the properties of the enzyme (Guengerich, 2008). The difference between the
Page 13 of 75
2D6 and 3A4 inhibition potential is due to the difference in their complex regulatory mechanisms
that will be explored in the next two sections.
Table 1: Metabolism of 2D6 and 3A4-mediated inhibitors, substrates, and inducers. The binding/inhibition is shown by the
pink and red marks, whereas the latter shows the strongest binding. 3A found in several forms changes the properties and
thereby its inhibitors it reacts with. Modified from (Spaggiari et al., 2014).
Together the CYP enzymes 2D6 and 3A4 make up 90 % of the CYP enzyme metabolism of drugs
used in 2002 (Guengerich, 2008). These are used in this project to explain the use of biomarkers
and to explain why hCES1 could be used in the personalization of medicine. The CYP enzymes are
well investigated and well understood; hence the properties can be transferred or linked to hCES1.
CYP2D6
2D6 is one of the key enzymes used in the degradation of xenobiotics and the 2D6 gene is part of
the CYP2 family that all have 9 exons and 8 introns (Guengerich and Cheng, 2011;Wang et al.,
2009), consisting of an open reading frame of 1491 base pairs (Wang et al., 2009). Most of the other
Page 14 of 75
CYP genes are found in clusters on different chromosomes. The 2D6 gene is the only gene which
leads to a functional enzyme when found alone. This could explain the importance of having a
functioning 2D6 enzyme present (Nelson et al., 2004). 2D6 is produced in the liver and is
responsible for about 10-25% of the drug metabolism (Williams et al., 2004;Zhou, 2009). 2D6 is
subject to CNV, where duplications increase the activity and deletions decrease or cause complete
loss of activity (Zanger and Schwab, 2013).
The 2D6 phenotype can often be determined clinically by using the substrate debrisoquine (Llerena
et al., 2009). Substrates, products and proteins may alter the activity due to changes in the structure.
These can be used as biomarkers to determine which turn over profile the patient belongs to (Zanger
and Hofmann, 2008). A strong inhibitor causes, as minimum, a 5-fold increase in the plasma
concentration area under the Curve (AUC)-values while a moderate inhibitor causes at minimum a
2-fold increase. A weak inhibitor causes less than a 2-fold increase (Zhou, 2009)
Using a cocktail approach scientists have found that 2D6 react to various molecules that are
naturally found in vivo, mostly from the diet (Spaggiari et al., 2014). Substrates and inhibitors that
interact with 2D6 and their effect on 2D6 activity are shown in Table 1. Inducers of 2D6 are still to
be discovered as the only evidence of alteration inducing 2D6 activity comes from gene
polymorphisms and not alterations of the transcript protein (Wang et al., 2009).
Variations in 2D6 that have shown to increase or decrease enzyme activity are found in mice as
well as in humans. There are at least 74 different 2D6 alleles (Wang et al., 2009), the most severe
and common variations are shown in Table 2. Decreases in 2D6 activity is not severe to the enzyme
activity and have a small impact on the patient, but null alleles cause a total loss of function and are
also often observed in vivo. As seen in Table 2, the most frequent null allele in the western world is
2D6*4 at a frequency of 0.15-0.25. Null alleles cause no enzyme activity and this can result in
severe adverse drug response (ADR) (Zanger and Schwab, 2013).
Page 15 of 75
Table 2: Illustration of the different outcomes of different 2D6 alleles. The Various alleles of 2D6, key mutation number,
Protein effect, functional effect, frequencies and clinical correlations, are all illustrated in figure. The folding of proteins
effect the activity of the enzyme while a null allele causes a loss of gene function leading to a risk of ADR that results in
toxication in some patient, while only a decrease would cause no toxication. Abbreviation: Ethnicities_ AA, African
American; Af African; As Asian; Ar, Arab; Ca Caucasian; Hs, Hispanic; In, Indian; Pc, Pacific; SA, South American;
gMAF, Global allele frequency. Rs number: accession number used by researchers and databases to refer to specific SNPs.
Also called Reference SNP cluster ID. Modified after (Zanger and Schwab, 2013).
Furthermore, the expression of metabolic enzymes have shown to be controlled by interleukin (IL)1β, TNF-α and IL-6. In the presence of cancer or just inflammation crucial down regulation of the
metabolizing enzymes is seen (Aitken et al., 2006).
The 3D structure of 2D6
2D6 has an active site within the catalytic pocket, which is associated with a heme group. In the
catalytic pocket, many crucial residues recognize substrates, bind the substrate and control the
transport of electrons. The most important residues are ASP301, HLU216 PHE483 and PHE120.
The crystal structure suggests that the PHE120 might control the orientation of the aromatic ring
found in most substrates. The aromatic ring within the substrate needs to bind in respect to the heme
Page 16 of 75
group. The heme group can react with a planar aromatic ring, and protonated basic nitrogens
through oxidation (Rowland et al., 2006).
CYP3A4
3A4 is the most abundant enzyme in the liver (Hayes et al., 2014) and is a part of the human CYP
subfamily 3A which consists of 3A4, 3A43, 3A5 and 3A7 (Westlind et al., 2001). The 3A4 gene
resides on chromosome 7q21.1 (Zanger and Schwab, 2013). Multiple signaling pathways regulate
3A4. The activity is continuously regulated at a transcriptional level by the C/EBPα,C/EBPβ (Jover
et al., 2002;Rodríguez-Antona et al., 2003) and HNF4α (Tirona et al., 2003) as a negative and
positive regulator, respectively. Regulation of 3A4 is not only seen on transcriptional level, several
overlapping substrate binding domains and allosteric regulation of the enzyme activity have been
detected too (Roberts et al., 2011).
3A4 is a membrane protein which metabolized about one third of most of the drugs on the US
market in 2002. Furthermore it metabolizes many endogenous and exogenous substrates (Denisov
et al., 2007;Fujiwara and Itoh, 2014;Guengerich, 1999;Hayes et al., 2014;Williams et al., 2004).
The CYP enzymes’ domains and active sites are very similar, but there are some differences. The
structure of 3A4 folds in a characteristic CYP way with a small N-terminal domain mainly
consisting of β-strands and a larger helical C-terminal domain, which holds the heme group and the
heme ligating CYS442, see Figure 5A. The active sites (MET114, SER119, ARG212, ILE301,
PHE304, ALA305, THR309, ALA370, LEU373 and LEU479) and additional active sites
(VAL101. ASN104, ARG105, ASP214, PHE215, ASP217, PRO218, GLU374 and SER478)
resides in the N-terminal and C-terminal domains. The additional active sites interact when 3A4
interacts with for example erythromycin. The large quantity of active sites result in a large active
site area in 3A4 (Szklarz and Halpert, 1997;Williams et al., 2004). Furthermore, 3A4 is very
flexible and undergoes large conformational changes when interacting with specific substrates.
These conformational changes happen at both the active sites and in structures further away from
the active sites, see Figure 5B (Ekroos and Sjogren, 2006;Williams et al., 2004). 3A4 have a highly
flexible catalytic pocket, when compared to 2D6. The catalytic pocket can be altered to change the
affinity for different substrates. This affinity does ultimately end up in a regulation of the enzyme.
Page 17 of 75
Figure 5: 3A4 complexes illustrating the conformational changes 3A4 undergoes when interacting with the inhibitor
ritonavir. The inhibitor is colored black. The light pink and pink area is the C-terminal domain, and the light yellow and
yellow area is the N-terminal domain. The light blue and blue area is the F helix domain. Dark red regions ribbons and side
chains are areas that interact with the heme group. The brown and light brown molecule is the cysteine group which ligates
the heme group. The orange regions and side chains are the primary active sites, and the dark orange regions and side chains
are the additional active sites (Scott and Halpert, 2005;Szklarz and Halpert, 1997;Williams et al., 2004). The figure is based
on the .pdb files 3NXU (ritonavir) and 1TQN (3A4 with no substrates), downloaded from RCSB.org.
The ability to change conformation makes 3A4 able to fit both large and small substrates into its
active site area and this flexibility makes it suitable for the metabolism of a larger group of drugs
(Ekroos and Sjogren, 2006;Williams et al., 2004). This provides a reason for why 3A4 makes up
the largest part of the CYP enzyme metabolism but the relationship between structure,
conformational changes and drug metabolism is not yet fully understood (Hutchinson et al.,
2004;Yano et al., 2004).
Evidence shows that in some cases CYP3A5 and 2D6 interact with 3A4 to break down drugs.
Additionally it is only possible to determine the relative specificity of 3A4 and 3A5 by using
individual recombinant human CYP’s (Hutchinson et al., 2004;Yano et al., 2004). This suggests
that they affect each other more than science has shown to this day.
Page 18 of 75
A number of different substrates have been proposed as being capable of estimating 3A4’s
interaction with different drugs. Furthermore it has been estimated that it is better to use several
probes representing different substrate groups for the in vitro assessment of 3A4 drug interactions
(Guengerich, 2008;Kenworthy et al., 1999). Testosterone has been used as a biomarker for the
activity of 3A4 in vitro because the enzyme metabolizes testosterone to 6-β-hydroxy testosterone
(Iwahori et al., 2003). At least two different substrates are known to be applicable to measure 3A4
activity in vivo. The first substrate useful in measuring the activity is 6-β-hydroxycortisol, which
has been used in a number of studies. It is useful as it is a natural nonspecific marker of 3A4 found
in a person’s urine (Bienvenu et al., 1991;Ged et al., 1989;Kovacs et al., 1998;Roby et al.,
2000;Saenger, 1983). When broken down by 3A4, 3-hydroxyquinidine is a quinidine metabolite.
This was proposed as a new biomarker for 3A4 activity and was compared with 6-βhydroxycortisol. The result showed that 6-β-hydroxycortisol had some limitations and that 3hydroxyquinidine, measurable in both plasma and urine, therefore might be a better biomarker for
3A4 activity (Damkier and Brøsen, 2000). 3A4 is not only important for the metabolism of
xenobiotic but is also an efficient steroid hydroxylase. It is involved in the catabolism of
endogenous steroids (Zanger and Schwab, 2013). These steroids are listed in Table 3. These
steroids can be used as a proxy for the 3A4 activity. Testosterone is one example of a biomarker
that has been used as a probe. As seen in Table 3 there is a variety of probes, but testosterone and
cortisol are the only endogenous substrates (Fuhr et al., 2007).
Table 3: Endogenous substrates and probes: The table shows which common known endogenous substrates are catalyzed by
CYP3A4. Further it is seen that only two of the probes are endogenous substrates. Even though more probes can be found,
these are examples.
Endogenous substrate
Source
Probes
Source
Bile acids
(Zanger and
Schwab, 2013)
(Zanger and
Schwab, 2013)
(Zanger and
Schwab, 2013)
(Zanger and
Schwab, 2013)
(Zanger and
Schwab, 2013)
(Diczfalusy et
al., 2001)
Midazolam
(Patki et al., 2003)
Triazolam
(Patki et al., 2003)
Alprazolam
(Hirota et al., 2001;Williams et al., 2002)
Dextromethorphan
(Fuhr et al., 2007)
Erythromycin
(Baker et al., 2004)
Alfentanil
(Klees et al., 2005)
Testosterone
(Kamdem et al., 2004)
Cortisol
(Huang et al., 2004)
Testosterone
Progesterone
Androstenedione
Cortisol
4β-hydroxycholesterol
Page 19 of 75
CYP2D6 and 3A4 as models
There are similarities in the structure of 3A4 and 2D6. The differences in these enzymes are in their
mode of function, regulation and frequency of polymorphisms. 2D6 have not shown to have
inducers on a substrate level. However, at gene level, different CNV’s causes an increase or
decrease in activity. 2D6 have also shown to have various polymorphisms that alter the activity, see
Table 4.
Unlike 3A4, there is no observation of post-translational 2D6 activity regulation. Two variants of
the 3A4 gene are found, but none of these variants shows a significant alteration in enzyme activity.
However, one variant has been suggested to lead to an increased risk of prostate cancer (Zanger and
Schwab, 2013). In contrast to 2D6, the binding of inhibitors or inducers regulates 3A4 activity. This
is due to allosteric alteration of the active sites (Williams et al., 2004). As the enzyme activity does
not change significantly in different genomic variants, the only regulatory complexes that would
explain the difference in enzyme activity are ligand-binding and allosteric modulation. In Table 4
below the inducers and inhibitors for 3A4 and 2D6 are shown as a comparison of the two CYP
genes (Zanger and Schwab, 2013). 2D6 do not have any inducers but many inhibitors have been
discovered. Nevertheless, almost twice the inducers than inhibitors are found for 3A4. The
difference in the number of inducers and inhibitors confirms that 3A4 is flexible at modulatory
level, whereas 2D6 only is inhibited. These features make the genotyping of 2D6 possible while
3A4 activity only can be accounted for by the actual regulation at the given time.
Page 20 of 75
Table 4: A combination of inhibitors of 2D6 and inducers as well as inhibitors of 3A4. 2D6 has a high amount of inhibtiors
that affect the enzyme activity. 3A4 also has a high amount of inhibitors, but twice as many inducers. 2D6 can thereby only be
inhibited by substrate interaction while 3A4 can be induced or inhbited by substrate interaction. Modified from (Zanger and
Schwab, 2013)
Table 4 shows that there is a significant difference in the post-translational regulation and as earlier
mentioned, the catalytic pocket of 3A4 have multiple substrate binding sites that overlap as well as
allosteric interaction, which can be transferred to the interaction between hCES1 and
inhibitors/inducers. Furthermore, 2D6’s genotypic importance can be transferred to the CNV of
hCES1 and explains, to some extent, the importance of CNV’s.
Page 21 of 75
The combination of the properties mentioned above is thought to regulate the activity of hCES1. As
genotyping of 3A4 cannot explain activity alterations, it is possible to apply the use of probes in
3A4 to using hCES1 activity probes. This could be an excellent method for determining the
phenotype of hCES1 if genotyping is not a possibility.
Metabolomics
Metabolome is a term referring to the complete set of small-molecule chemicals within a biological
sample. The interactions between gene expression, protein expression and the cellular environment
are responsible for the identities, concentrations and moving of metabolites. Metabolomics focuses
on a broad series of metabolites instead of focusing on a single metabolite, which is the approach
often used in biochemistry. Metabolomics combine quantitative data on several metabolites to get a
better understanding of mechanisms involving multiple metabolites, such as metabolic dynamics
associated with disease and drug exposure. Metabolomics can be used to understand diseases and
medicaments’ impact on the body as these affect the body’s metabolome. Changes in the
metabolome as a result of the disease or medicament is known as long-term changes in the
metabolome.
Analysis of the changes in the metabolome can result in an initial metabolomic signature (novel
biomarker). Further investigation of the initial metabolic signature could potentially lead to the
discovery of a biomarker for the disease or drug. This means that by applying metabolomics to the
previous information and the following information about hCES1, a possible initial metabolomic
signature for hCES1 activity can be proposed (Quinones and Kaddurah-Daouk, 2009).
Carboxylesterases and their activity
Carboxylesterases (CESs) are found in several organisms including mammals. They are a multigene
enzyme family, part of the serine hydrolase superfamily (Williams et al., 2011). These enzymes are
engaged in the metabolic activation or detoxification of xenobiotics, such as drugs, pesticides and
other environmental toxicants. In addition, they effectively catalyze the hydrolysis of various
amide- and ester bond-containing chemicals, as well as the hydrolysis of chemicals and drugs to
their respective free acids. CES’s are thought to be one of the main determinants of
pharmacokinetics of ester drugs and pro-drugs (Satoh and Hosokawa, 2010).
The catalytic triad of amino acids at the CES active site is vital for catalytic activity. At the base of
the active site, nucleophilic serine residue is found which attacks the carbonyl carbon of ester-
Page 22 of 75
containing substrates (Ross et al., 2010). The mRNA expression of hCES1 and hCES2 is greatest in
the liver, which suggests a significant role in the metabolic detoxification of xenobiotics.
Furthermore, CESs are expressed in the stomach, kidney, heart, spleen, testis and colon (Williams
et al., 2011). The expression of hCES1 greatly exceeds hCES2 in the liver (about 50-fold), but
hCES2 is highly expressed in the intestine (Laizure et al., 2013).
Though carboxylesterase-mediated hydrolysis is commonly overlooked at the clinical level, it plays
a significant role in the metabolism of various therapeutic agents, prescribed widely. These drugs
from varied classes include angiotensin-converting enzyme inhibitors (ACEIs), 3-hydroxy-3methylglutaryl coenzyme A inhibitors (statins), antiviral agents, antiplatelet drugs, narcotic
analgesics, angiotensin receptor blockers (ARBs), immunosuppressants, central nervous system
(CNS) stimulants and oncology agents. The increasing awareness that diseases, genetic factors and
drug interactions alter the enzyme activity of these enzymes and have a significant impact on the
therapeutic effects of substrate drugs is due to the increasing number of therapeutic agents, subject
to carboxylesterase hydrolysis. However, even though we understand how CYP enzymes in
disposition and furthermore the clinical effects of several drugs are important, in the metabolism of
substrate drug, the role of carboxylesterase hydrolysis has been broadly understudied (Laizure et al.,
2013).
Based on genomic structure and genetic sequence, a recent genomic analysis determined five
different CESs subfamilies within the mammalian family, yet the far most well studied are the
proteins of subfamily CES1 and CES2. Mammalian CES substrates are endogenous (i.e., acyl-CoA
esters and acyl-glycerols) as well as exogenous. To some extent, both CES1 and CES2 have
imbricating substrate specificity and in a lesser degree, substrate selectivity too. For instance, a
comparison between structures formed between hCES1, heroin, cocaine, 4-methylumbelliferyl
acetate and 6-monoacetylmorphine and structures formed between hCES2 and the same substrates,
showed that hCES1’s affinity was greater for cocaine while hCES2’s affinity was higher for 4methylumbelliferyl acetate and 6-monoacetylmorphine. The two enzymes showed similar Km value
for heroin. This data propose that hCES1 prefers compounds containing a bigger acyl moiety while
hCES2 preferentially hydrolyzes compounds containing a bigger alcohol moiety relative to the acyl
component (Williams et al., 2011). CESs are considered imperative enzymes associated with prodrug activation, prominently with respect to up-regulation in tumor cells, tissue distribution and
turnover rates, due to the phase I reactions described in; Metabolism (Satoh and Hosokawa, 2010).
Page 23 of 75
Drug Hydrolysis
An ester is made up by an acyl- and an alcohol group. The result of an ester hydrolysis is the
formation of the analogous carboxylic acid and alcohol and the product is usually more polar,
compared to the original ester. CESs lack substrate specificity and even though drug substrates are
vulnerable to carboxylesterase (CES) hydrolysis or hydrolysis by other esterases, one CES usually
serves as the major hydrolysis pathway as that specific CES predominates with each substrate.
Based on the ester structure, CES predominates can be predicted. The enzyme of hCES1 generally
prefers esters with a large acyl group and small alcohol group while hCES2 prefers the opposite, a
small acyl group and large alcohol group (Laizure et al., 2013). An ester group is often added into a
drug’s structure during synthesis and design with the aim to improve oral absorption. Additionally
passive transport through membranes is more effective as converting carboxylic acid into an ester
increases hydrophobicity. Figure 6 illustrates, among others, the pro-drug oseltamivir’s hydrolysis
into oseltamivir carboxylate, which illustrates the above mentioned.
Figure 6: Hydrolysis pathways for hCES1 and hCES2. hCES1 hydrolyses esters with a large acyl group and a small alcohol
group. hCES1 hydrolyses oseltamivir to oseltamivir carboxylate, leaving a -OH group, increasing the compounds
hydrophobicity. hCES2 hydrolyses compounds with a small acyl group and a large alcohol group. hCES2 hydrolyses
presugrel into thiolactone (Laizure et al., 2013)
Oseltamivir carboxylate, the active neuraminidase inhibitor, has poor oral bioavailability and is
highly polar. The result of the carboxylic acid’s ethoxy ester is a compound with good
bioavailability, which is less polar (He et al., 1999). After absorption, hCES1 in the liver quickly
hydrolyses the ester to the active drug moiety resulting in an antiviral compound. This compound is
a pro-drug and avoids the additional cost of intravenous administration, the inconvenience and is
Page 24 of 75
orally active. The hydrolysis of methylphenidate to its analogous carboxylic acid, illustrated in
Figure 6, shows that an ester functional group might be a significant structural factor for a
therapeutic agents’ activity. The active compound, methylphenidate, is hydrolyzed to ritalinic acid,
the inactive carboxylate, by hCES1 in the liver. Clopidogrel and aspirin, the broadly used
antiplatelet drugs, are prominent examples of hydrolysis as an inactivation pathway (Laizure et al.,
2013).
Substrates
Regardless of the big number of broadly prescribed drugs subject to CES-mediated hydrolysis, the
clinical importance of ester hydrolysis in drug metabolism might have been undervalued.
Cardiovascular drugs including ARBs, statins, ACEIs, anticoagulants and fibric acids, is the major
therapeutic class of CES substrate drugs. Three angiotensin receptor blockers (ARBs), olmesartan
medoxomil, candesartan cilexetil and azilsartan medoxomil are pro-drugs that require hydrolysis to
the active metabolite. A hCES2 catalyzing process in the small intestine that takes place during
absorption. All ACEIs, except Lisinopril and captopril, are ester pro-drugs. They are hydrolyzed to
their analogous therapeutically active carboxylic acid by hCES1. The three frequently used
antiplatelet agents; clopidogrel, aspirin and prasugrel are all subject to CES hydrolysis. Clopidogrel
needs to be metabolized to thiolactone (2-oxo-clopidogrel) as it is a pro-drug with no antiplatelet
activity. Thiolactone is a transitional inactive metabolite that is later metabolized to the active form.
CYP enzymes catalyze these two metabolic steps. hCES1 hydrolysis is competing with this
activation pathway of both clopidogrel and the thiolactone metabolite causing metabolic products
that are inactive. hCES2 hydrolyses aspirin to salicylic acid. Aspirin represents an inactivation
pathway of its antiplatelet effect, as it is the active antiplatelet agent. However, hCES2 hydrolysis
acts as an activation pathway for the anti-inflammatory activity of aspirin as salicylate is the main
anti-inflammatory moiety. Prasugrel is a substrate pro-drug for hCES2. After oral administration,
during absorption, hCES2 hydrolyses prasugrel to thiolactone. The CYP system then metabolizes
the thiolactone to the active moiety. The result of this hCES2-mediated hydrolysis is the formation
of the thiolactone metabolite. Due to the efficiency of this reaction, the plasma concentrations of
prasugrel after oral administration are normally below detectable limits (Laizure et al., 2013)
The new oral anticoagulant, called dabigatran, is hydrolyzed by hCES1 and possibly hCES2 at two
separate sites in order to produce the active metabolite. Dabigatran is given as a double ester prodrug and is a reversible direct thrombin inhibitor (Blech et al., 2008). The two thiolactone statins,
Page 25 of 75
simvastatin and lovastatin, as well as the two esters of fibric acids, fenofibrate and clofibrate are
lipid-lowering agents and hCES1 substrate pro-drugs. Three therapeutic categories of CES
substrates affecting the CNS exist: CNS stimulants, methylphenidate and cocaine; meperidine,
opiate agonists and heroin; rufinamide (anticonvulsant drug). To understand the metabolism of the
most extensively studied CES substrate, cocaine, plentiful animal, human and in vitro studies have
been made. It is subject to both hCES1 and hCES2 hydrolysis at two separate ester sites on the
cocaine structure. To produce benzoylecgonine, hCES1 catalyzes the hydrolysis of the methyl ester
and to produce ecgonine methyl ester, hCES2 catalyzes the hydrolysis of the benzoyl ester. Both
metabolites are inactive and is eliminated renally (Laizure et al., 2013). As mentioned in Drug
Hydrolysis, hCES1 hydrolyzes methylphenidate to ritalinic acid, a catalyzed hydrolysis that is
stereo-selective (Zhu et al., 2010). Administered as a racemate, methylphenidate consist of d- and
l-methylphenidate. The less active l-isomer is of greater efficiency for hCES1 hydrolysis than the
active d-isomer (Laizure et al., 2013). Heroin is hydrolyzed by hCES2 among other esterases to its
monoacetylmorphine metabolite and later to morphine whereas meperidine is hydrolyzed by hCES1
to meperidinic acid, an inactive metabolite. Heroin as well as its hydrolysis products possess agonist
activity for opiate receptors. hCES1 catalyzed hydrolysis predominantly eliminates the amide
rufinamide to its carboxylic acid. Then it is subject to either renal elimination or further
glucuronidation previously to renal elimination (Perucca et al., 2008).
The only clinical examples of CES substrates that are carbamates are the two oncology drugs,
Capecitabine and irinotecan. hCES2 hydrolyzes the pro-drug capecitabine, but the product is a
translational metabolite and it therefore has to be further metabolized by thymine phosphorylase
and cytidine deaminase to form the active antitumor moiety, 5-fluorouracil. The formation of the
active metabolite, 7-ethyl- 10-hydroxy camptothecin (SN-38), is the result of the hydrolysis of the
pro-drug, irinotecan. It seems that both hCES1 and hCES2 are contributors to the hydrolysis of
irinotecan to SN-38, but the hCES2 catalyzed hydrolysis is far more efficient. The administration of
irinotecan is intravenously and entree to the utmost abundant expression of hCES2 located in the
small intestine is thereby limited. However, hCES1 seems to play an imperative role in drug
activation (Laizure et al., 2013).
Mycophenolate is an immunosuppressant administered at the inactive ester mycophenolate mofetil.
To form mycophenolic acid, an active drug, it therefore undergoes hydrolysis. Even though
hydrolysis happens in the plasma, the liver and the intestine, in vitro studies have shown liver
Page 26 of 75
hydrolysis by hCES1 to be the most effective pathway. This vulnerability to several esterases is not
distinctive to mycophenolate mofetil. The majority of esters vulnerable to enzymatic hydrolysis are
substrates of several esterases, which might be subject to spontaneous or non-enzymatic hydrolysis.
In general, the efficiency of enzymatic hydrolysis is considerably superior for one specific esterase
and the hydrolysis variability through this pathway controls the ester’s alteration rate to the
analogous carboxylic acid and alcohol (Laizure et al., 2013).
Role of CES’s in xenobiotic metabolism
CESs are imperative contributors to xenobiotic metabolic pathways, counting drugs and pro-drugs.
For some pro-drugs, such as irinotecan and prasugrel, the bio activation of the pro-drug is directly
catalyzed by CESs. Still, CESs have the ability to facilitate the inactivation and clearance of both
drugs and pro-drugs. The understanding of similarities as well as differences in CES activity
between human beings and preclinical models of drug disposition is crucial, as evolving therapeutic
agents using these hydrolytic pathways are being developed (Williams et al., 2011). Various illicit
drugs, environmental pollutants and pharmaceutical agents contain ester bonds, which increase
bioavailability and compound lipophilicity. Some examples are narcotics, plasticizers, pro-drugs,
pesticides and anti-thrombotic drugs, such as aspirin. CES hydrolytic actions may inactivate or bioactivate these compounds, depending on the compound. Various laboratories have focused on CES
metabolism of pyrethroids due to their widespread use in public health and agriculture (Ross et al.,
2010).
Human Carboxylesterase 1
The hCES1 is a serine hydrolase responsible for metabolizing a broad variety of substrates, both
endogenous and xenobiotic. Xenobiotic substrates of hCES1 include heroin, cocaine, lidocaine, and
ACE inhibitors. Additionally hCES1 is thought to be part of the endogenous cholesterol
metabolism, catalyzing hydrolysis of cholesterol and fatty acyl Coenzyme A (Bencharit et al.,
2006). The structure of the catalytic pocket recognizes widely diverse substrates, hCES1 mainly
catalyzes the hydrolysis of compounds containing a larger acyl group or compounds containing a
smaller alkyl group (Satoh and Hosokawa, 2006), as mentioned in; Metabolomics
The flexible catalytic pocket recognizes a broad band of xenobiotics; see Table 5, but only a few
endogenous substrates. See Table 6, which illustrates its enzymatic diversity and the broad
selectivity (Thomsen et al., 2014).
Page 27 of 75
Table 5: Xenobiotics that are broken down by hCES1, hydrolysis products and their therapeutic effect. Modified from
(Laizure et al., 2013).
Page 28 of 75
Table 6: Endogenous substrates for hCES1.
Substrate
Hydrolysis product
Reference
Cholesteryl ester (CEH)
Chloresterol
(Bencharit et al., 2006)
Fatty Acyl CoA (FACoAH)
Coenzyme A
(Bencharit et al., 2006)
Acyl-Coenzyme A Cholesterol Acyltransfer (ACAT)
Chloresteryl ester (Chloresteryl palmitate)
(Bencharit et al., 2006)
hCES1 is expressed in several tissues in the human body such as the lungs, heart, adipose tissue,
monocytes, macrophages and most abundantly in the liver. Expression of the CES1 gene in the
human population seems to be remarkably constant between individuals. Contrarily, in the same
individuals, enzymatic hCES1 activity differs considerably (Ross et al., 2010). Consequently, in the
human population there is no simple correlation between hCES1 gene expression and hCES1
enzymatic activity. Thus, a major factor accounting for the changing pharmacokinetic behavior of
ester-containing xenobiotics in human populations might be post-translational hCES1 activity
regulation. Alternatively, another variation contributor may be unidentified SNPs in the CES1 gene
(Ross et al., 2010). A study performed by Friedrichsen et al investigating mRNA expression of
hCES1 in adipose tissue has shown that there is no significant transcription increase in individuals
with extra copies (three-four copies) of the hCES1 gene in relation to individuals with two copies
(Friedrichsen et al., 2013). Sai et al showed that individuals with three or four copies of the gene (n
= 35) had a 24 % increase in CES1 activity compared to people with two copies (n = 23), but the
sample sizes were small (Sai et al., 2010).
Genetic polymorphism of carboxylesterase 1
The hCES1 gene is located on chromosome 16q13-q22.1, it comprises of 14 exons and spans about
30kb. Gene variants have been investigated by several in order to explain differences in the
metabolic activity of the hCES1 enzyme amongst individuals. Marsh et al conducted a study in
2004 in which hCES1 was re-sequenced in populations of healthy African and European
individuals. The study identified 16 hCES1 SNPs and none of the novel SNPs had any predictive
value in relation to hCES1 activity and expression (Marsh et al., 2004). The allele frequencies are
shown in Table 7.
Page 29 of 75
Table 7: The structure and function of hCES1 haplotypes found in ≥5% in either the European or African population, or in
both. Modified from (Marsh et al., 2004).
Later in 2008, Yoshimura et al published a study in which hCES1 had been sequenced in a
population of Japanese individuals. Polymorphisms where identified in the promoter region of
CES1A2 affecting the promoter activity thus altering the efficiency of the ACE inhibitor drug
imidapril by affecting the hCES1 mediated activation of the drug (Yoshimura et al., 2008).
Additionally Zhu et al reported a study, using methylphenidate, which investigated individuals with
an apparent malfunctioning breakdown of the drug. By sequencing hCES1 of the specific test
subject, two mutations were found on the hCES1 gene resulting in a significantly lowered catalytic
function of the hCES1 enzyme. One of the mutations showed a deletion leading to a premature stop
codon resulting in a complete loss of catalytic activity. Subsequently DNA from 925 individuals
was sequenced, none of which showed the same mutation as the individual of the previous study
suggesting that the mutation is extremely rare (Zhu et al, 2008). The other mutation lead to a nonconservative substitution. The 143GLU-variant and in vitro studies showed a 21% catalytic
efficiency compared to the wild type. Respectively, the allele frequency of this mutation was
determined to be 3.7% in Caucasians, 4.3 % in Black, 2% and 0% in Hispanic and Asians (Zhu et
al., 2008).
Page 30 of 75
A study by Nemoda et al further investigated this functional mutation (143Glu) and found that
individuals diagnosed with Attention deficit hyperactive disorder (ADHD) that carried the relatively
rare GLU allele of hCES1 required a lower dosage of methylphenidate when treated (Nemoda et al.,
2009). Furthermore, Johnson et al performed a study on 77 Japanese children diagnosed with
ADHD and treated with methylphenidate. During the study they didn’t detect any relation between
gene variants, dosage, and response. They did however find an association between two hCES1
SNP and sadness as an adverse response (Johnson et al., 2013).
A recent study performed by Yu-Jung Cha et al identified three novel hCES1 SNPs, which were
classified as possibly damaging to protein function and structure. Two of the three SNPs have been
investigated and have not shown to lower the hCES1 mediated metabolism of specific drug
substrates. The SNP remaining has yet to be investigated (Cha et al., 2014).
The present investigations of the effects of hCES1 genetic polymorphisms on hCES1 mediated
metabolism has not resulted in a one-sided conclusion concerning the importance of genetics in
relation to prediction of the hCES1 mediated metabolic activity. The 143Glu allele is the only
determined SNP, which has proven to reduce the catalytic efficiency significantly. Additionally
there is a great possibility that yet to be identified SNPs can be a determining factor for differences
in hCES1 activity amongst populations. There may be many rare SNPs which will result in an
impaired hCES1 mediated metabolism in an individual and together these will account for part of
the variations observed amongst populations. Additionally there may be SNPs, both frequent and
rare, that individually do not affect the activity of hCES1 but when several of them are present in
one individual, it will result in a reduced hCES1 activity. Unidentified polymorphisms in regions
outside the hCES1 gene and regulatory regions including unidentified inducers and inhibitors may
as well be the cause of the individual variation in hCES1 activity.
The structure of hCES1
The hCES1 enzyme can be divided into three distinct functional domains. The catalytic site, the Zsite and the alpha/beta domain. The catalytic site contains the active pocket, which harbors the
catalytic triad. The Z-site is found in the regulatory domain and controls the trimer hexamer
equilibrium of the enzyme. As the enzyme uses the surface of the Z-site to form the hexamer, the Zsite is only able to bind ligands when the enzyme is in its trimeric state. Lastly, the alpha/beta
domain is involved in stabilizing the trimeric structure of the enzyme and the release of product
Page 31 of 75
molecules (Bencharit et al., 2006). The catalytic triad is located on the active site and is composed
of SER221, GLU353 and HIS467, see Figure 7.
Figure 7: Illustration of hCES1 as a monomer. The light red area is the N-terminal domain, the light blue area is the Cterminal domain and the light yellow area is the αβ-domain. The Z-site is illustrated as an electron cloud around a
taurocholate molecule. The active site, where many substrates reside under metabolism, is illustrated as a green colored
electron cloud around another taurocholate molecule. The green side chains are the side chains which forms the catalytic
triad (SER221, GLU354 and HIS468). The dark green areas are the catalytic pocket. The red and dark red areas are active
sites, found by two docking analyses with nelfinavir and morphine, respectively. The area which is pointed out with a blue
circle is the substrate binding gorge opening (Bencharit et al., 2006;Rhoades et al., 2012;Vistoli et al., 2010). The data for the
illustration is the .pdb files 2DR0 (hCES1 taurocholate complex) and 4AB1 (hCES1 with no substrates) from RCSB.org.
The catalytic triad of hCES1 is highly conserved amongst CES1 isozymes and the catalytic function
may therefore be strikingly conserved amongst the isozymes too. The active site is comprised of a
large ”flexible” pocket and a smaller ”rigid” pocket, see Figure 8A and B, providing a highly
hydrophobic environment favorable for the hydrolysis of larger hydrophobic compounds. Part of
the larger pocket is a small secondary pore referred to as the side door, shown in Figure 7. This side
Page 32 of 75
door allows smaller substrates and product molecules to enter and exit the active site to and from
the surface of the enzyme.
The structure of the catalytic pocket recognizes widely diverse
substrates (Satoh and Hosokawa, 2006).
A study by Vistoli et al investigating the metabolism of hCES1 by docking analysis, showed the
residues aligning the catalytic cavity of hCES1 to be mainly hydrophobic, see Figure 8A. The study
additionally showed that the SER221 residue divided the catalytic pocket into two polar distinct
cavities. Non-polar residues align the bigger more flexible pocket that accommodates the larger acyl
groups and the smaller, more rigid pocket, is aligned by polar residues. These results by Vistoli et
al, supports the theory that hCES1 mainly hydrolyses substrates with either a smaller alkyl group or
a larger acyl group (Vistoli et al., 2010).
Figure 8: Illustration of the catalytic pocket in hCES1. A) Simple illustration of the main residues in the catalytic pocket, the
green residues are non-polar residues, H-bonding residues are blue, negatively charged residues are red and the catalytic
triad residue SER is orange. Modified from (Vistoli et al., 2010). B) The main residues of the catalytic pocket, the C-terminal
domain and Z-site. The same color scheme, as for picture A is used. C) hCES1 with the main residues in the catalytic pocket
colored in the same colors as in picture A and B. The light red area is the N-terminal domain, the light blue is the C-terminal
domain and the light yellow is the αβ-domain. The data for illustration A and B is the .pdb files 2DR0 (hCES1 taurocholate
complex) and 4AB1 (hCES1 with no substrates) from RCSB.org.
The role of CES1 in metabolism of endogenous compounds
Lipid metabolism
Ko et al found that CES1 limits hepatic accumulation of triglycerides by promoting beta-oxidation
of fatty acids (a multi-step process in which fatty acids are broken down by various tissues to
produce energy) (Ko et al., 2009). Furthermore, Freidrichsen et al suggested that CES1 in adipose
tissue is involved in both the trans-esterification and de-esterification of lipids (Friedrichsen et al.,
2013).
Page 33 of 75
Xu et al investigated the role of hepatic Ces1 in lipid metabolism in mice, which showed that an
overexpression of hepatic Ces1 resulted in a lowered hepatic triglyceride level and plasma glucose
level in both diabetic and non-diabetic mice. Furthermore a knock down of Ces1 showed an
increase in hepatic triglyceride and an increase in plasma cholesterol levels. It is suggested that
Ces1 performs triglyceride hydrolysis which in turn is followed by changes in the oxidation of fatty
acids. Interestingly, they did not find lowered plasma cholesterol by overexpression (Xu et al.,
2014).
Cholesterol
hCES1 performs hydrolysis of cholesteryl ester. Releasing an alcohol substituent from the substrate
results in a fatty acyl enzyme complex intermediate. This intermediate then reacts with water
resulting in a release of the molecule containing the acyl group. Additionally hCES1 acts as a
transferase, catalyzing the formation of cholesteryl esters in case of abundance of free cholesterol
(Friedrichsen et al., 2013).
Inhibition of hCES1 results in an inhibition of cholesteryl ester hydrolysis in human macrophage
cells and thereby decreases cholesterol efflux from human macrophage foam cells (Crow et al.,
2008;Igarashi et al., 2010).
Zhao et al showed that hCES1 depletion reduced CES hydrolytic activity with more than 70%,
indicating that hCES1 proteins account for about 70% of intracellular CES hydrolysis. However,
Ces1 knock down (KD) did not decrease intracellular cholesteryl ester (CE) hydrolysis or free
cholesterol efflux, in human THP-1 monocytes. The CES1KD macrophages showed about a 30%
increase in human carboxylesterase 3 (hCES3) expression, which led to an increase in the CE
hydrolytic activity. They concluded that hCES3 could be a carboxylesterase which was up regulated
when hCES1 was KD (Zhao et al., 2012). Ross et al showed that despite efficient KD of hCES1
protein expression there was no significant difference in the percentage of cholesterol efflux to
ApoA1 when hCES1KD macrophages was compared to control macrophages in human THP-1
monocytes, and free cholesterol levels were unchanged. They also found that the hCES1KD THP-1
macrophages’ CE level was significantly lower than that in control THP-1 macrophages both before
and after efflux. hCES3 mRNA was up regulated but hCES3 activity was not increased. This led
them to believe that hCES3 does not compensate for hCES1 loss of function (Ross et al., 2014). If
this is the case, it is likely that cholesteryl ester could be a likely biomarker for hCES1 activity.
Page 34 of 75
Glucose
Xu et al proposed that CES1 expression is directly regulated by glucose due to the fact that a high
glucose level in mouse primary hepatocytes induced hepatic Ces1 mRNA expression. Their data
showed that glucose stimulates Ces1 activity through a region between 75bp and 150 bp upstream
of the transcription site (Xu et al., 2014). Furthermore, Xu et al showed that ATP-citrate lyase,
which converts glucose-derived citrate to acetyl-CoA used in glucose-mediated histone acetylation
(Wellen et al., 2009) is required for the glucose-mediated acetylation of H3 and H4 in the Ces1
chromatin (Xu et al., 2014). Xu et al suggested that due to this, glucose induces Ces1 expression,
Ces1 regulates glucose metabolism, and hepatic Ces1 may regulate blood glucose levels after a
meal (Xu et al., 2014). This makes glucose an interesting factor in future research of biomarkers for
hCES1 metabolism.
In all of the above mentioned studies of CES1 metabolism, a common hindrance is that endogenous
compounds metabolized by CES1 in most cases are metabolized by other enzymes too. This makes
it more difficult to study the pathway and outcome of the CES1 metabolism and applying this to
hCES1 metabolism.
Allosteric functions of hCES1
hCES1 is known to be promiscuous regarding substrate specificity (Bencharit et al., 2003;Bencharit
et al., 2006;Fleming et al., 2005;Williams, 1985). This has led to the theory that hCES1 or some of
its sites have allosteric functions. In 2005 Fleming et al proposed, that the flexibility of hCES1 was
not due to the large pocket but due to the Z-site, which lies above the large pocket (Fleming et al.,
2005). By combining the Bencharit et al study, from 2003, with their own data, Fleming et al
suggested that ligand binding to the hCES1 Z-site shifts the enzyme’s trimer-hexamer equilibrium
toward the trimer, thus facilitating the binding of substrates within the active site (Bencharit et al.,
2003;Fleming et al., 2005). This suggests that the flexibility and movement of the large pocket
makes the movement and allosteric functions of the Z-site possible though the pocket does not have
an allosteric function itself.
In 2006, Bencharit et al investigated hCES1’s lack of substrate specificity and found data that
supported the hypothesis suggested by Fleming et al in 2005. Furthermore the data suggested that
the Z-site is capable of binding to cholesterol-like molecules, which might play a role in the
biological activity of the enzyme when processing cholesteryl esters and other endogenous
substrates.
Page 35 of 75
Additionally Bencharit et al suggested that the Z-site may facilitate the proper positioning of the
catalytic triad residue Glu354 for catalysis. This was based on structural insights in the function of
the rabbit liver CE Z-site and kinetic data from previous experiments (Bencharit et al.,
2002;Bencharit et al., 2006;Fleming et al., 2005). These suggestions presented are coherent with the
known promiscuity of hCES1, and the fact that 3A4 has been shown to have a similar ligand
binding site on its surface, adjacent to its active site (Bencharit et al., 2006;Williams et al., 2004), it
is possible that hCES1 has a similar site. As both 3A4 and hCES1 are important drug metabolizing
liver enzymes, in the specific enzyme families they belong to, the allosteric sites in 3A4 and hCES1
suggest that conformational change is imperative in drug metabolizing enzymes (Imai et al.,
2006;Williams et al., 2004). It is proposed that the peripheral binding site on 3A4 functions in
substrate screening and allosteric regulation. Therefore Bencharit et al suggests that hCES1’s Z-site
has the same function. This is supported by previous data (Bencharit et al., 2003;Bencharit et al.,
2006;Fleming et al., 2005;Williams et al., 2004).
The flexibility of hCES1 is creating a diverse enzyme that can add to the recognition of xenobiotics,
and thereby dealing with substrates that the other enzymes cannot (Thomsen et al., 2014)
Discussion
Adverse effects
When considering the extensive research needed to find and implement a possible biomarker or
probe for hCES1, it is necessary to analyze whether the possible benefits, the effort and financial
costs are worth it.
If the minority of patients treated experience adverse effects or if the adverse effects are
insignificant, is finding a biomarker or probe then beneficial? If only a minor quantity of drugs
commonly used are metabolized by hCES1, is hCES1 then significant to research? hCES1
metabolizes several xenobiotic compounds and about half of these, are pro-drugs. Hence, no hCES1
activity results in no therapeutic effect, as the compound is not efficiently metabolized. This is the
case for many ACE-inhibitors, antihyperlipidemic agents, immunosuppressive agents and antiviral
agents, hCES1 metabolizes. Additionally, a prolonged drug exposure and high amounts of active
compounds present in the body is associated with adverse effects.
Page 36 of 75
Severe or fatal adverse effects are not common amongst xenobiotics metabolized by hCES1 so
when looking into hCES1 mediated drug metabolism and finding biomarkers or probes, the aim is
not just to save lives but to improve the life quality of patients who experience less severe, but longterm adverse effects.
An example of a common adverse effect in ADHD patients treated with methylphenidate is
insomnia. Many patients diagnosed with ADHD are subject to prolonged treatment using this drug.
Even though this adverse effect in some perspectives could be considered relatively mild, in a longterm perspective it can result in a significantly lowered quality of life as it damages the patient’s
daily life. Another aspect that has to be considered is the size of the therapeutic window of the drug.
Research in pharmacokinetic profiling is relevant when treating with a drug that has a very narrow
therapeutic window.
Pharmacokinetics play a crucial role in controlling drug exposure. The longer drugs are active and
in high concentrations, the greater the probability of ADR. Thus, an ultra-rapid metabolizer will
experience a quick increase in the plasma concentration of the active compound of a pro-drug and a
rapid decrease in the plasma concentration of a non-pro-drug. The accumulation of drugs is thereby
mainly controlled by pharmacokinetics, which also affects efficacy and ADR. The increased or
decreased exposure may result in a plasma concentration below or above the therapeutic window
and can additionally cause ADR or result in no beneficial effect of the drug. In contrast to CYP2D6
in which SNPs determine the function or non-function of the enzymes and to pharmacodynamics
that is regulated by gene variants, hCES1 is not known to be highly regulated by gene variants and
is likely regulated by induction and inhibition of the enzyme.
Research showed that CYP3A4 is an excellent model in this type of regulation. When the drug
lurasidone is administered with an inhibitor or an inducer, Cmax and AUC changes. The same
change was observed in the lurasidone plasma concentration in which a higher plasma
concentration increases the likelihood of ADR. As hCES1 might be altered by both inducers and
inhibitors, this alteration have an important influence on drug metabolizing enzymes and the plasma
concentration (Caccia et al., 2012).
Alleles’ and SNPs’ influence on enzyme activity
As earlier mentioned, CYP2D6 is highly regulated by allele variants. The alleles have a significant
impact on the regulation of the enzyme activity. Variant 2D6: *3,*4,*5 and *6 show a total loss of
allele function and the frequencies of these alleles are significant, up to 25% in some ethnic groups.
Page 37 of 75
Variant 2D6: *10, *17, *41, are all SNP’s which causes a decrease in enzyme activity and
expression. These variants are also seen at high frequencies, especially *10 is seen in up to 70% of
Asians. For hCES1 one functional SNP is determined that result in 21% of catalytic efficiency in
relation to that of a wild type. This allele is as well relatively rare and is most frequent in Caucasian
and African populations where it shows an allele frequency of 3.7% and 4.3% respectively (Zhu et
al., 2008).
For the two enzymes 2D6 and hCES1, no clear resemblance appears amongst the significance of
polymorphisms. However, it is proven that hCES1 activity is subject to genetically determined
decreases in catalytic function due to SNPs. One SNP decreases the catalytic efficiency to 21% with
an allele frequency at a maximum of 4.3%. This is not sufficient to explain the apparent variation in
the hCES1 enzyme activity among individuals in populations. Contrarily individual variation in
CYP2D6 activity and expression is clearly regulated by SNPs in the genome and can be explained
by the genotype of a patient.
Studies show that 2D6 expression increases when the copy number of the 2D6 gene increases. In
contrast, studies have shown that an increase in CNV in hCES1 does not result in a significant
increase in hCES1 expression and activity (Friedrichsen et al., 2013;Zhu and Markowitz, 2013).
Additionally studies have identified novel SNPs yet to be investigated in relation to the impact on
enzyme activity.
Newer research are making it clearer that the activity cannot only be explained by the gene and the
variants that may show up, there is properly another regulatory mechanism which explains the
regulation at substrate level.
Conformational changes affecting function and activity
As previously mentioned, hCES1's substrate promiscuity can be explained by allosteric functions
mainly at the Z-site. When the Z-site interacts with a substrate it promotes conformational changes
and thereby allows larger or smaller molecules to protrude and fit in the catalytic site. Furthermore,
the side door permits part of the molecules to protrude the side door, which facilitates the
promiscuity. The same allosteric functions are the main explanation for 3A4's substrate promiscuity
and is why it is used as a model for hCES1. The substrate promiscuity in both enzymes makes it
difficult to determine a substrate to use as a possible biomarker. The allosteric regulation of these
enzymes changes the affinity for the substrates and finding only one biomarker is therefore
Page 38 of 75
complicated. For 3A4, three biomarkers have been suggested and are to some extent used, but
similar discoveries has yet to be done for hCES1. Even though 3A4 has a high substrate
promiscuity, it has peripheral binding sites capable of substrate screening too. While there are no
correlations between the substrate screening and exclusion of specific substrates, data proposes that
hCES1's Z-site has the same substrate screening capabilities. It is known that hCES1 prefers
molecules with a high acyl moiety, which could be due to a positive substrate screening quality at
the Z-site. If the Z-site has no positive substrate screening quality it may just promote binding
affinity to substrates. The substrate screening quality of the Z-site can be investigated using a
simple molecule with an acyl group in one end and a non-polar group at the other end. If this fits
and in some way interacts with the Z-site, the positive substrate screening is possible. If not, the
high acyl moiety preference of hCES1 could be due to the amino acids lining the large catalytic
pocket. If the Z-site has a negative substrate screening quality it will be able to exclude compounds
and by identifying which substrates it can exclude it could narrow down the possible biomarkers for
hCES1. Furthermore, the acyl moiety preference could be utilized further by adding acyl
compounds to the possible biomarkers. The modified substrate might lose the affinity for other
enzyme pathways and thereby only has hCES1 as its catalyzer.
Inhibitors and inducers
Similarities are found when comparing hCES1 with CYP2D6 and CYP3A4. These similarities
suggest that the enzyme function is a combination of the two CYP's main abilities. As 2D6 and
3A4, hCES1 metabolizes a large portion of different xenobiotics, however 3A4 metabolize a
broader spectrum of xenobiotics than 2D6, indicating that 3A4 has the widest metabolizing ability.
hCES1 is known to be a predominant drug metabolizer in the liver suggesting that it has similar
qualities as 3A4. Nevertheless, as 2D6, hCES1 has no known inducers implying that these two are
more comparable. Can it be determined which of the CYP's that is equivalent to hCES1 based on
the ability to be induced by endogenous or xenobiotic compounds, it should be possible to base the
same equivalency on the inhibitors known for each of the enzymes. Although different compounds
inhibit the enzymes, some inhibit the same enzymes.
To find the most convenient model for hCES1, other similarities are considered; though it is rare,
hCES1 as 2D6 shows some genotypic decreases in activity. Contrary to 3A4, no inducers for 2D6
are found, and only very few for hCES1. The above mentioned combined with the allosteric
functions of hCES1 that shows similarities to 3A4, the most useful CYP model for hCES1 might
Page 39 of 75
not be one of the CYP's, but a combination of both. Therefore, the most efficient method in
determining the hCES1 phenotype is to measure the actual enzyme activity through biomarkers or
probes. The model of genotyping from 2D6 can be used to explain some of the activity, while 3A4
explains the flexibility of the structure and why the actual metabolism must be measured throughout
phenotyping.
The complexity of finding probes and biomarkers for the activity of hCES1
A way to further personalize medical treatment with the aim to diminish symptoms more efficiently
and minimize adverse responses, is to determine the activity of hCES1 by identifying a probe or
biomarker that can be used as a proxy for the hCES1 activity as it has been done with testosterone
and cortisol as biomarkers for CYP3A4 activity. Finding a biomarker for hCES1 is difficult due to
the few identified endogenous substrates. Contrarily there are plenty of xenobiotics, which may be
possible probes in determining the phenotype.
All the identified endogenous substrates for hCES1 are part of several pathways metabolized by
other enzymes as well as by hCES1. Thus, it is complex to determine one specific substrate and
product as an eligible suggestion for a hCES1 activity biomarker. If the amount of metabolizing
effect hCES1 contributes with to the cholesteryl ester metabolism, or the metabolism of other
endogenous substrates can be determined, these could be suggested as possible biomarkers. These
biomarkers can be used to determine different profiles. If the metabolizing effects of other enzymes
to the pathway is known, it can help estimating the hCES1 activity. Nevertheless, this requires an
extensive metabolomic investigation into the cholesteryl ester metabolism pathways in the body.
Therefore, a xenobiotic compound used as a probe would possibly be a more qualified way to go,
excluding endogenous substrates as useful biomarkers. Currently, using known probes to determine
phenotypes of patients is too complicated for clinical practices.
As mentioned before, two endogenous substrates is identified and are clinically applied biomarkers
for 3A4. In contrast to the endogenous hCES1 substrates, other enzymes do not metabolize the
endogenous substrates, metabolized by 3A4. If a xenobiotic compound is used as a probe for
hCES1, it has to be non-toxic both as a substrate and as a hydrolytic product. In addition, in order to
be considered a probe, it has to be effectively metabolized to prevent accumulation of metabolites.
Furthermore, the risk of adverse effects in xenobiotic probes has to be considered. The process of
identifying substrates by hCES1’s substrate promiscuity is complicated. If the Z-site substrate
Page 40 of 75
screening process and the elimination of compounds is investigated further, it might be possible to
invent a qualification method excluding possible substrates in the search for a probe detecting
hCES1 activity. Finding a useful xenobiotic substrate for hCES1 may be possible as this method of
using xenobiotic probes or biomarkers already are established for 3A4 and are, like hCES1,
characterized by broad substrate specificity.
Probes for hCES1
The only known probe for hCES1 that shows a significant specificity is derived from 2-(2-hydroxy3-methoxyphenyl) benzothiazole (HMBT). The probe (probe 1) is fluorescent and developed to
monitor the hCES1 activity by introducing benzoyl moiety to HMBT giving it a high hCES1
affinity. Probe 1 can be used to perform a qualitative analysis of the enzyme activity in vitro. The
probe shows a low toxicity for living cells and can therefore be used on living hepatocytes. In Liu et
al, the team using this probe, succeeded performing this method on living hepatocytes. This study is
the first to determine an applicable probe for hCES1 activity and though there is a background
signal from hCES2 activity, it is statistically insignificant.
The discovery of probe 1 opens up new possibilities for hCES1 phenotyping and use in vivo. As
hCES1 primarily is expressed in the liver and the small intestine, these are the optimal test tissue.
The hepatocytes and epithelia of the small intestine is nearly impossible to achieve and the
procedure of diagnosing a patient may not be beneficial when a cost-benefit perspective is
considered. Although, to achieve a more precise measurement of hCES1 activity, testing on the
liver or the small intestine is preferred as the allosteric regulation may be different compared to
tissue with a lower expression level. The advantage of using tissues in which regulation may be
different, such as monocytes and macrophages, is the accessibility (Sanghani et al., 2009;Satoh et
al., 2002). Monocytes and macrophages are easily extracted as well as tested in blood. Probe 1
makes it possible to use blood samples to detect the hCES1 phenotype, which also has been
suggested by Liu et al.
Metabolomics
Probe 1 is, as previously mentioned, the only probe identified for hCES1. Since the probe is only
used on living hepatic cells, it is not yet certain whether the method is clinically applicable but
monocytes may be a new opportunity. If the method does not show valid results when used on
easily accessible cells such as macrophages and monocytes, the procedure may be too invasive for
Page 41 of 75
regular clinical use. Metabolomics would therefore be an obvious and possibly more suitable
approach to test the endogenous substrates, as they are easily accessible through blood samples.
Cholesteryl ester (CEH), Fatty Acyl CoA (FACoAH) and Acyl-Coenzyme A Cholesterol
Acyltransfer (ACAT) are three endogenous substrates which hCES1 metabolizes. Since these
substrates act as intermediates in many pathways to become their representative product they cannot
be used solely for a qualitative measurement to predict hCES1 activity. By using metabolomics, all
of these substrates can be measured at once.
As the hCES1 activity cannot be detected by measuring a single substrate, the other pathways may
be the answer to solve the equation. The other pathways that metabolize the endogenous substrates
can be measured by other substrates and their activity can be determined too. If the enzyme activity
of the other pathway are normal, the lowered endogenous concentration of CEH, FACoAH and/or
ACAT would be due to hCES1. Therefore, the three metabolite-markers CEH, FACoAH and
ACAT, can be used to detect whether the activity of hCES1 is normal, increased or decreased in an
indirect way. As this method requires a more extensive research into which pathways contribute to
the endogenous concentration of CEH, FACoAH and ACAT and their products, it might not be
optimal for measuring hCES1 activity though it might be the only possible approach. Since
metabolomics measure all metabolites and not only a single pathway, other pathways may
synthesize metabolites. When metabolites are developed through multiple pathways, a single result
may not be sufficient. When other pathways are used as an indicator, complications might occur if
the pathway is either induced or inhibited. This will change the ratio of all the pathways and can be
used as a proxy for hCES1 activity. If focusing on one pathway in order to see whether the
substrate/product ratio changes, these values can be determined as a standard and used to determine
different turnover-profiles. Nevertheless, an unknown disturbance may too result in a biased
diagnosis of the hCES1 phenotype. Background noise may decrease as metabolomics can be used
on great amounts of metabolites.
The above mentioned procedure can be applied on xenobiotics metabolized by hCES1. These can
be used as metabolite-markers in metabolomics as a quantitative analysis. The obstacle when using
xenobiotics is the high risk of toxicity. To use these xenobiotic metabolite-markers they must run
down the same pathway as the intermediates and must have an insignificant toxicity. The possibility
of ADR is probably the limiting step when using xenobiotics as metabolite markers. If these
Page 42 of 75
requirements are not met, the above-mentioned method will not be applicable when conducting a
cost benefit analysis.
Conclusion
Pharmacokinetics, more specifically the metabolism of drugs, is a major determinant for drug
response. The activity of drug metabolizing enzymes does not only determine the therapeutic effect
of the treatment but also the likelihood for adverse effects. The dose-dependent response, including
both the therapeutic response and ADR are highly related to the rate of drug metabolism, meaning
the activity of drug-metabolizing enzymes, such as hCES1, is of great importance to the drug
response. When a drug has a narrow therapeutic window, it is increasingly important to estimate the
hCES1 activity and target the plasma concentration of the drug within the therapeutic window.
When predicting the metabolic activity of hCES1, CYP2D6 and CYP3A4 can be used as models.
However, as the previous discussion suggests it is not sufficient to use only one CYP as a model, as
hCES1 activity cannot be fully explained by neither the 2D6 model nor the 3A4. hCES1 exhibits
genetic variation and enzyme alteration, which makes it is necessary to consider both models.
Likewise it is not possible to find a direct correlation between enzyme activity modulation and
substrates, inhibitors and inducers, and thus not possible to determine one single CYP model to
predict hCES1 activity modulation.
Studies of Probe 1 shows low toxicity and a significant specificity for hCES1 in hepatic cells. These
results indicate that this probe is a promising candidate for clinical trials in cells expressing hCES1.
Further tests in biological samples are needed to determine the efficiency in distinct cell types.
Currently, studies have solely been performed in vitro on hepatic cells and it is still uncertain if this
method can be applied in a clinical perspective.
Since genotyping does not provide a complete picture of the activity, another approach could be to
apply metabolomics in which many endogenous substrates are taken into consideration. In
metabolomics, many metabolites are contemplated to determine enzyme activity, which makes this
method applicable regardless of whether the enzyme activity is under genetic control or if it is
influenced by other parameters.
Page 43 of 75
The only useful probe, probe 1, is tested on poorly accessible tissue samples hence the use of
biomarkers or probes for hCES1 in clinical treatment is not possible. Gene markers were used but
did not explain the enzyme activity. In the future, with extensive research on mechanisms
highlighted in this project, finding a biomarker for hCES1 activity that is easily accessible which
eventually can be clinically applied is possible.
Considering that no useable biomarkers or probes for determining hCES1 activity are found,
looking into other methods such as metabolomics may be useful. Numerous enzymes that are part
of other pathways metabolize various endogenous substrates also metabolized by hCES1. This
makes metabolomics an evident method to apply to the hCES1 activity determination predicament.
In order to apply metabolomics to the determination of enzyme activity, extensive research is a
necessity, as several different parameters and metabolites have to be investigated. Still, it may
ultimately lead to a better estimation of the actual enzymatic activity.
Perspective
This study shows that determining the pharmacokinetic profile of a patient can be complex. It is
necessary to consider the time and resources required for profiling a patient in relation to the
possible beneficial outcome. Furthermore, some medical conditions are acute which makes
profiling before giving the necessary treatment impossible due to the lack of time.
Extensive research into metabolomic pharmacokinetic profiling may make it possible to profile a
patient by analyzing blood samples resulting in a spectrum and a metabolite substrate ratio. The
spectrum is then compared to a known pharmacological profile providing an inexpensive method
with an immediate answer. This saves resources by reducing the amount of patients revisiting
medical professionals due to experienced adverse effects or due to the lack of therapeutic effect.
For example, a normal metabolite/substrate ratio is investigated to determine a normal pharmacokinetic profile and additionally determine the profile of a poor –or an ultra rapid metabolizer. These
profiles can be used in comparison when estimating the pharmacokinetic profile of a patient.
Acknowledgements
We would like to thank our supervisor, Henrik Berg Rasmussen for his assistance in our progress
and for his genuine thorough guidance throughout this project
Page 44 of 75
Supplements
Docking experiment
Introduction
Docking is an in silico method for determining the affinity between a compound and a receptor or
an enzyme. In this docking experiment, the compounds are tested on the enzyme hCES1. Docking
is among others used to assess whether a specific compound will bind to an enzyme and based on
the affinity where within the enzyme or at which sites it binds. Solely based on a docking
experiment, where the compound will bind in vivo or in vitro cannot be concluded but the docking
can tell if a compound will bind and where the affinity is highest (Grosdidier et al., 2011).
Methods
SwissDock was used for several docking analyses of the interactions of different compounds with
hCES1. The output from SwissDock was used to compare the affinity of the selected compounds
for hCES1 and cannot be used to determine any kind of interaction between hCES1 and the
compound. The three compounds selected have known in vitro properties when interacting with
hCES1. These data show which affinity profiles a substrate, an inhibitor and a negative reference
compound have when interacting with hCES1. To simplify the in silico experiment, only the ΔG,
the FullFitness and the location of a ligand within the enzyme are included. The docking experiment
is conducted from three test subjects: 27-hydroxycholesterol, used as a strong inhibitor, cholesterol,
used as a negative reference (Crow et al., 2010) and methylphenidate as substrate and a positive
control (Yang et al., 2014). The objective with these three substances is to determine how the
FullFitness and ΔG change compared to the expectations. As hCES1 undergoes conformational
changes when interacting with a compound the enzyme was allowed 5Å flexibility in the docking
analyses.
Docking shows, the possible and viable interactions a compound can undergo within an enzyme.
The possible interaction sites depicted which leads to a modulated activity and the role of the
compounds used, are based on our knowledge from the previous theory section giving us the
possibility to determine where the site of action should be and the possible correct interaction.
Nevertheless, as docking is an in silico experiment and is dependent on the users’ knowledge of
different interactions within an enzyme, the results are not conclusive. The results are preferentially
Page 45 of 75
verified using a 3D crystallography and if confirmed, further verification using an in vitro model
can be a possibility.
The docking of the three compounds cholesterol, 27-hydroxycholesterol and methylphenidate
resulted in about 240 viable docking models. To reduce this number, ΔG for the viable model of
methylphenidate within the catalytic pocket was used as a limit for the viable model of the negative
reference, cholesterol, and the strong inhibitor, 27-hydroxycholesterol. Based on the docking of
methylphenidate, two models are possible. These are the only two models in which
methylphenidate is close to the catalytic pocket. These two models respectively have a ΔG of -6.6
and -6.8. Based on this, the negative reference cholesterol, which does not act as a substrate or
inhibitor will have a ΔG above -6.6 because the binding and therefore the energy of the reaction,
will not be favorable. The strong inhibitor, 27-hydroxycholesterol, will have a ΔG of -6.6 or below
as it is a favorable reaction, and has a high affinity.
Results
When limiting the amount of possible models for the negative reference, cholesterol, it resulted in
19 different cluster groups, see Appendix #1 for a cluster comparison. For the inhibitor, 27hydroxycholesterol, the limitation resulted in 23 clusters, see Appendix #2 for a cluster comparison.
The models with the highest and lowest ΔG and FullFitness values from the selected clusters were
found and the average was calculated for each of the compounds, see Appendix #3. These were
used for further analysis, see Table 8.
Table 8: Data from docking with 5Å flexibility. It is apparent that there is some sort of correlation between ΔG and the
FullFitness, although this correlation is much more obvious when looking at methylphenidate.
ΔG (Kcal/mol)
FullFitness
27-hydroxy average
-7.04222
-2412.33
Cholesterol average
-6.317
-2400.72
Methylphenidate #1.241 and #1.244
-6.79 and -6.607
-2473.488 and -2472.282
Page 46 of 75
As the data are sorted, based on ΔG of methylphenidate, the result is not surprising. Nevertheless,
only a few models did not fit the expected and were excluded from the 27-hydroxycholesterol and
cholesterol docking data.
The results indicate the affinity is highest for the inhibitor, then the substrate, and last cholesterol,
which makes sense. When comparing the binding of substrates to other biological systems, the
outcome is as expected, except for cholesterol as it does not have any influence on hCES1 activity
and probably does not have a high affinity for the hCES1 complex at all. When comparing ΔG with
the FullFitness result for each compound, it is apparent that FullFitness is lowest for the substrate
and that no direct correlation between FullFitness and ΔG for the inhibitor nor for the negative
reference is found. When comparing ΔG and FullFitness for the individual clusters there is to some
extent a correlation for the inhibitor and the negative reference. This could be due to the inhibitor
and negative reference not having to fit within the enzyme as they are not metabolized by the
enzyme.
Binding sites
The binding site for the substrate methylphenidate is limited to the catalytic site, shown in Figure 9,
and two docking models with the ability to bind within the catalytic site were found.
Figure 9: Illustration of the methylphenidate docking outcome when the two models in which methylphenidate resides within
the catalytic pocket are selected.
Page 47 of 75
Based on the docking analysis there are multiple possible binding sites for the inhibitor 27hydroxycholesterol. The binding site with the highest affinity is shown in Figure 10. The binding
site with the highest affinity was expected to be near the Z-site, the side door or in one of the
catalytic pockets as these sites are imperative for the binding of a substrate.
Figure 10: Illustration of the 27-hydroxycholesterol hCES1 binding model with the highest affinity. This does not necessarily
indicate that 27-hydrocycholesterol will bind at this site in vitro or in vivo. Additionally it was not expected that it would bind
as illustrated or inhibit hCES1 as it is distanced from any known crucial hCES1 activity sites.
Nevertheless, the high affinity of the above shown cluster, does not necessarily mean that this is the
actual binding site of the inhibitor 27-hydroxycholesterol. Additionally, as shown in Figure 11,
there are possible binding sites that match the expected ΔG for the inhibitor.
Page 48 of 75
Figure 11: Illustration of the possible binding sites for 27-hydroxycholesterol when interacting with hCES1. The right hand
clusters or the clusters in the N-terminal region (upper left clusters) were expected as these are regions of interest when
looking at hCES1 activity regulation. The sites are near known structures that are crucial in the allosteric modulation, the
side door (upper left) and the catalytic pocket (right hand side).
Discussion
The results from the three docking experiments show that ΔG and FullFitness do not necessarily
correlate and do not show enough proportional behavior to be used as a limiting factor when
considering the binding site of an inhibitor. However, these values do not need to follow the same
pattern as they describe two different parameters: the affinity (ΔG) and the “hand in glove” situation
(FullFitness). Furthermore, ΔG is not applicable as a deciding factor when considering the exact site
an inhibitor will bind to but can be used as a limiting factor when defining the parameters in which
an inhibitor should fit as the affinity is significantly higher for the inhibitor and the substrate than
for the cholesterol. With a limited knowledge of hCES1 function, the use of positive and negative
test subjects is necessary to elect substances before biological trial, and docking is reliable when
singling out substances.
The structure of the compounds used in the docking can be seen in Appendix #4 and the results and
calculations for 27-hydroxycholesterol and cholesterol can be seen in Appendix #3. If we had time
it could have been interesting to do the same experiment with the other known inhibitors of hCES1,
seen in Appendix #5.
Page 49 of 75
References
Aitken, A. E., Richardson, T. A. and Morgen, E. T. (2006).
Regulation of drug-
metabolizing enzymes and transporters in inflammation. Annu Rev Pharmacol Toxicol 46, 123-26.
Ancrenaz, V., Daali, Y., Fontana, P., Besson, M., Samer, C., Dayer, P. and Desmeules, J.
(2010). Impact of genetic polymorphisms and drug-drug interactions on clopidogrel and prasugrel
response variability. Curr. Drug Metab. 11, 667-677.
Bailey, D. N. and Briggs, J. R. (2003). Procainamide and quinidine inhibition of the human
hepatic degradation of meperidine in vitro. J. Anal. Toxicol. 27, 142-2.
Baker, S. D., van Schaik, R. H., Rivory, L. P., Ten Tije, A. J., Dinh, K., Graveland, W. J.,
Schenk, P. W., Charles, K. A., Clarke, S. J., Carducci, M. A. et al. (2004). Factors affecting
cytochrome P-450 3A activity in cancer patients. Clin. Cancer Res. 10, 8341-50.
Bencharit, S., Edwards, C. C., Morton, C. L., Howard-Williams, E. L., Kuhn, P., Potter, P. M.
and Redinbo, M. R. (2006). Multisite promiscuity in the processing of endogenous substrates by
human carboxylesterase 1. J. Mol. Biol. 363, 201-13.
Bencharit, S., Morton, C. L., Howard-Williams, E. L., Danks, M. K., Potter, P. M. and
Redinbo, M. R. (2002). Structural insights into CPT-11 activation by mammalian
carboxylesterases. Nat. Struct. Biol. 9, 337-5.
Bencharit, S., Morton, C. L., Hyatt, J. L., Kuhn, P., Danks, M. K., Potter, P. M. and Redinbo,
M. R. (2003). Crystal structure of human carboxylesterase 1 complexed with the Alzheimer's drug
tacrine: from binding promiscuity to selective inhibition. Chem. Biol. 10, 341-8.
Bencharit, S., Morton, C. L., Xue, Y., Potter, P. M. and Redinbo, M. R. (2003). Structural basis
of heroin and cocaine metabolism by a promiscuous human drug-processing enzyme. Nat. Struct.
Biol. 10, 349-7.
Bienvenu, T., Rey, E., Pons, G., d'Athis, P. and Olive, G. (1991). A simple non-invasive
procedure for the investigation of cytochrome P-450 IIIA dependent enzymes in humans.
International Journal of Clinical Pharmacology, Therapy, and Toxicology 29(11), 441-3.
Page 50 of 75
Blech, S., Ebner, T., Ludwig-Schwellinger, E., Stangier, J. and Roth, W. (2008). The
Metabolism and Disposition of the Oral Direct Thrombin Inhibitor, Dabigatran, in Humans. Drug
Metab. Dispos. 36(2), 386-13.
Borup, V. (2010). Biokemi, pp. 872. Copenhagen, Denmark: FADLs forlag.
Brenner, G. and Stevens, C. (2009). Pharmacokinetics. In Pharmacology, pp. 10-15. Philadelphia,
USA: Saunders.
Caccia, S., Pasina, L. and Nobili, A. (2012). Critical appraisal of lurasidone in the management of
schizophrenia. Neuropsychiatr Dis Treat. 8, 155-168.
Cha, Y., Jeong, H., Shin, J., Kim, E., Yu, K., Cho, J., Yoon, S. H. and Lim, K. S. (2014).
Genetic Polymorphisms of the Carboxylesterase 1 (CES1) Gene in a Korean Population.
Translational and Clinical Pharmacology 22(1), 30-4.
Collins, S. F. (2014). Talking Glossary of Genetic Terms. National Human Genome Research
Institute 2014,.
Crow, J. A., Herring, K. L., Xie, S., Borazjani, A., Potter, P. M. and Ross, M. K. (2010).
Inhibition of carboxylesterase activity of THP1 monocytes/macrophages and recombinant human
carboxylesterase 1 by oxysterols and fatty acids. Biochimica et Biophysica Acta 1801, 31-10.
Crow, J. A., Middleton, B. L., Borazjani, A., Hatfield, M. J., Potter, P. M. and Ross, M. K.
(2008). Inhibition of carboxylesterase 1 is associated with cholesteryl ester retention in human
THP-1 monocyte/macrophages. Biochim. Biophys. Acta 1781(10), 643-11.
Damkier, P. and Brøsen, K. (2000). Quinidine as a probe for CYP3A4 activity: Intrasubject
variability and lack of correlation with probe-based assays for CYP1A2, CYP2C9, CYP2C19, and
CYP2D6. Clin Pharmacol Ther. 68(2), 199-9.
de Groot, M. J., Ackland, M. J., Horne, V. A., Alex, A. A. and Jones, B. C. (1999). Novel
Approach To Predicting P450-Mediated Drug Metabolism: Development of a Combined Protein
and Pharmacophore Model for CYP2D6. J. Med. Chem. 42 (9), 1515-9.
Page 51 of 75
Denisov, I. G., Baas, B. J., Grinkova, Y. V. and Sligar, S. G. (2007). Cooperativity in
cytochrome P450 3A4: linkages in substrate binding, spin state, uncoupling, and product formation.
J. Biol. Chem. 282, 7066-10.
Denmeade, S. R., Mhaka, A. M., Rosen, D. M., Brennen, W. N., Dalrymple, S., Dach, I.,
Olesen, C., Gurel, B., Demarzo, A. M., Wilding, G. et al. (2012). Engineering a prostate-specific
membrane antigen-activated tumor endothelial cell prodrug for cancer therapy. Sci. Transl. Med. 4,
140ra86.
Diczfalusy, U., Nylén, H., Elander, P. and Bertilsson. (2001). 4β-Hydroxycholesterol,
an endogenous marker of CYP3A4/5 activity in humans. Br. J. Clin. Pharmacol. 71, 183-89.
DiPiro, J. T. and Sprull, W. J. (2010). Introduction to Pharmacokinetics and Pharmacodynamics.
In Concepts in Clinical Pharmacokinetics (ed. D. Battaglia and H. Pollard), pp. 1-4. USA:
American Society of Health System.
Ekroos, M. and Sjogren, T. (2006). Structural basis for ligand promiscuity in cytochrome P450
3A4. Proc. Natl. Acad. Sci. U. S. A. 103, 13682-5.
Fleming, C. D., Bencharit, S., Edwards, C. C., Hyatt, J. L., Tsurkan, L., Bai, F., Fraga, C.,
Morton, C. L., Howard-Williams, E. L., Potter, P. M. et al. (2005). Structural insights into drug
processing by human carboxylesterase 1: tamoxifen, mevastatin, and inhibition by benzil. J. Mol.
Biol. 352, 165-12.
Friedrichsen, M., Poulsen, P., Wojtaszewski, J., Hansen, P. R., Vaag, A. and Rasmussen, H. B.
(2013). Carboxylesterase 1 gene duplication and mRNA expression in adipose tissue are linked to
obesity and metabolic function. PLoS One 8(2), e56861.
Fuhr, U., Jetter, A. and Kirchheiner, J. (2007). Appropriate phenotyping procedures for drug
metabolizing enzymes and transporters in humans and their simultaneous use in the “cocktail”
approach. Clin. Pharmacol. Ther. 81, 270-13.
Fujiwara, R. and Itoh, T. (2014). Extensive protein interactions involving cytochrome P450 3A4:
a possible contributor to the formation of a metabolosome. Pharma Res Per 2(5), 1-4.
Page 52 of 75
Fujiyama, N., Miura, M., Kato, S., Sone, T., Isobe, M. and Satoh, S. (2010). Involvement of
carboxylesterase 1 and 2 in the hydrolysis of mycophenolate mofetil. Drug Metab. Dispos. 38,
2210-7.
Ged, C., Rouillon, J. M., Pichard, L., Combalbert, J., Bressot, N., Bories, P., Michel, H.,
Beaune, P. and Maurel, P. (1989). The increase in urinary excretion of 6β-hydroxycortisol as a
marker of human hepatic cytochrome P450111A induction. Br. J. clin. Pharmac. 28, 373-13.
Geshi, E., Kimura, T., Yoshimura, M., Suzuki, H., Koba, S., Sakai, T., Saito, T., Koga, A.,
Muramatsu, M. and Katagiri, T. (2005). A single nucleotide polymorphism in the
carboxylesterase gene is associated with the responsiveness to imidapril medication and the
promoter activity. Hypertens. Res. 28, 719-6.
Girardin, F. (2006). Membrane transporter proteins: a challenge for CNS drug development.
Dialogues Clin. Neurosci. 8, 311-10.
Godfrey, A. R., Digiacinto, J. and Davis, M. W. (2011). Single-dose bioequivalence of 105-mg
fenofibric acid tablets versus 145-mg fenofibrate tablets under fasting and fed conditions: a report
of two phase I, open-label, single-dose, randomized, crossover clinical trials. Clin Ther. 33(6), 7669.
Grant, D. (1991). Detoxification pathways in the liver . J Inherit Metab Dis. 14(4), 421-9.
Green, J. (2007). The Three C's of Etiology. Wide smiles 2014,.
Grosdidier, A., Zoete, V. and Michielin, O. (2011). SwissDock, a protein-small molecule docking
web service based on EADock DSS. Nucleic Acids Research 39, 270-7.
Guengerich, F. P. (1999). Cytochrome P-450 3A4: regulation and role in drug metabolism. Annu.
Rev. Pharmacol. Toxicol. 39, 1-17.
Guengerich, F. P. (2008). Cytochrome P450 and Chemical Toxicology. Chem. Res. Toxicol. 21,
70-13.
Page 53 of 75
Guengerich, F. P. and Cheng, Q. (2011). Orphans in the human cytochrome P450
superfamily: approaches to discovering functions and relevance in pharmacology. Pharmacol Rev
63, 684-15.
Halpin, R. A., Ulm, E. H., Till, A. E., Kari, P. H., Vyas, K. P., Hunninghake, D. B. and
Duggan, D. E. (1993). Biotransformation of lovastatin. V. Species differences in in vivo metabolite
profiles of mouse, rat, dog, and human. Drug metabolism and disposition: the biological fate of
chemicals 21(6), 1003-8.
Hayes, C., Ansbro, D. and Kontoyianni, M. (2014). Elucidating substrate promiscuity in the
human cytochrome 3A4. J. Chem. Inf. Model. 54, 857-12.
He, G., Massarella, J. and Ward, P. (1999). Clinical pharmacokinetics of the prodrug oseltamivir
and its active metabolite Ro 64-0802. Clin. Pharmacokinet. 37, 471-13.
Hirota, N., Ito, K., Iwatsubo, T., Green, C. E., Tyson, C. A., Shimada, N., Suzuki, H. and
Sugiyama, Y. (2001). In vitro/in vivo scaling of alprazolam metabolism by CYP3A4 and CYP3A5
in humans. Biopharm. Drug Dispos. 22, 53-18.
Hosokawa, M., Endo, T., Fujisawa, M., Hara, S., Iwata, N., Sato, Y. and Satoh, T. (1995).
Interindividual variation in carboxylesterase levels in human liver microsomes. Drug Metab.
Dispos. 23, 1022-5.
Huang, W. L., Kalhorn, T. F., Baillie, M. and Thummel, K. E. (2004). CYP3A5 catalyzed
cortisol 6 beta-hydroxylation in liver and kidney. Drug Metab Rev 36, 77-77.
Hutchinson, M. R., Menelaou, A., Foster, D. J., Coller, J. K. and Somogyi, A. A. (2004).
CYP2D6 and CYP3A4 involvement in the primary oxidative metabolism of hydrocodone by human
liver microsomes. Br. J. Clin. Pharmacol. 57, 287-10.
Igarashi, M., Osuga, J., Uozaki, H., Sekiya, M., Nagashima, S., Takahashi, M., Takase, S.,
Takanashi, M., Li, Y., Ohta, K. et al. (2010). The critical role of neutral cholesterol ester
hydrolase 1 in cholesterol removal from human macrophages. Circ. Res. 107, 1387-8.
Page 54 of 75
Imai, T., Taketani, M., Shii, M., Hosokawa, M. and Chiba, K. (2006). Substrate specificity of
carboxylesterase isozymes and their contribution to hydrolase activity in human liver and small
intestine. Drug Metab. Dispos. 34, 1734-7.
Iwahori, T., Matsuura, T., Maehashi, H., Sugo, K., Saito, M., Hosokawa, M., Chiba, K.,
Masaki, T., Aizaki, H., Ohkawa, K. et al. (2003). CYP3A4 inducible model for in vitro analysis
of human drug metabolism using a bioartificial liver. Hepatology 37, 665-8.
Johnson, K. A., Barry, E., Lambert, D., Fitzgerald, M., McNicholas, F., Kirley, A., Gill, M.,
Bellgrove, M. A. and Hawi, Z. (2013). Methylphenidate side effect profile is influenced by genetic
variation in the attention-deficit/hyperactivity disorder-associated CES1 gene. J. Child Adolesc.
Psychopharmacol. 23, 655-9.
Jover, R., Bort, R., Gómez-Lechón, M.
J. and Castell, J. V.
(2002). Down-
regulation of human CYP3A4 by the inflammatory signal interleukin-6: molecular mechanism and
transcription factors involved<br />. FASEB. J. 13, 1799-2.
Kamdem, L. K., Meineke, I., Koch, I., Zanger, U. M., Brockmoller, J. and Wojnowski, L.
(2004). Limited contribution of CYP3A5 to the hepatic 6beta-hydroxylation of testosterone.
Arch. Pharmacol 370, 71-6.
Kenworthy, K. E., Bloomer, J. C., Clarke, S. E. and Houston, J. B. (1999). CYP3A4 drug
interactions correlation of 10 in vitro probe substrates. Blackwell Science Ltd Br J Clin Pharmacol
48, 716-11.
KK, J. (2009). Textbook of Personalized Medicine, pp. 430. New York: Springer Science +
Business Media.
Klees, T. M., Sheffels, P., Thummel, K. E. and Kharasch, E. D. (2005). Pharmacogenetic
determinants of human liver microsomal alfentanil metabolism and the role of cytochrome P450
3A5. Anesthesiology 105, 550-6.
Kleingeist, B., Bocker, R., Geisslinger, G. and Brugger, R. (1998). Isolation and
pharmacological characterization of microsomal human liver flumazenil carboxylesterase. J.
Pharm. Pharm. Sci. 1, 38-8.
Page 55 of 75
Ko, K. W., Erickson, B. and Lehner, R. (2009). Es-x/Ces1 prevents triacylglycerol accumulation
in McArdle-RH7777 hepatocytes. Biochim. Biophys. Acta 1791, 1133-10.
Kolars, J. C., Awni, W. M., Merion, R. M. and Watkins, P. B. (1991). First-pass metabolism of
cyclosporin by the gut. Lancet 338, 1488-2.
Kolars, J. C., Schmiedlin-Ren, P., Schuetz, J. D., Fang, C. and Watkins, P. B. (1992).
Identification of rifampin-inducible P450IIIA4 (CYP3A4) in human small bowel enterocytes. J.
Clin. Invest. 90, 1871-7.
Kovacs, S. J., Martin, D. E., Everitt, D. E., Patterson, S. D. and Jorkasky, D. K. (1998).
Urinary excretion of 6β-hydroxycortisol as an in vivo marker for CYP3A induction: Applications
and recommendations. Clinical Pharmacology & Therapeutics 63, 617-4.
Laizure, S. C. and Parker, R. B. (2010). A comparison of the metabolism of clopidogrel and
prasugrel. Expert Opin Drug Metab Toxicol. 6(11), 1417-7.
Laizure, S. C., Herring, V., Hu, Z., Witbrodt, K. and Parker, R. B. (2013). The role of human
carboxylesterases in drug metabolism: have we overlooked their importance? Pharmacotherapy 33,
210-12.
Laizure, S. C., Mandrell, T., Gades, N. M. and Parker, R. B. (2003). Cocaethylene metabolism
and interaction with cocaine and ethanol: role of carboxylesterases. Drug Metab. Dispos. 31, 16-4.
Lewis, D. F. (2002). Homology modelling of human CYP2 family enzymes based on the CYP2C5
crystal structure. Xenobiotica 32, 305-18.
Lewis, D. F. (2003). Essential requirements for substrate binding affinity and selectivity toward
human CYP2 family enzymes. Arch. Biochem. Biophys. 409, 32-12.
Li, P., Callery, P. S., Gan, L. S. and Balani, S. K. (2007). Esterase inhibition attribute of
grapefruit juice leading to a new drug interaction. Drug Metab Dispos. 35(7), 1023-8.
Llerena, A., Dorado, P. and Peas-Lledo, E. (2009). Pharmacogenetics of debrisoquine and its use
as a marker for CYP2D6 hydroxylation capacity. Pharmacogenetics 10, 17-11.
Page 56 of 75
Marsh, S., Xiao, M., Yu, J., Ahluwalia, R., Minton, M., Freimuth, R. R., Kwok, P. Y. and
McLeod, H. L. (2004). Pharmacogenomic assessment of carboxylesterases 1 and 2. Genomics 84,
661-7.
Meldrum, C., Doyle, M. A. and Tothill, R. W. (2011). Next-generation sequencing for cancer
diagnostics: a practical perspective. Clin. Biochem. Rev. 32, 177-195.
Meyer, U. A. and Zanger, U. M. (1997). Molecular mechanisms of genetic polymorphisms of
drug metabolism. Annu. Rev. Pharmacol. Toxicol. 37, 269-27.
Mikeska, T. and Craig, J. (2014). DNA methylation biomarkers: cancer and beyond. Genes
(Basel). 5(3), 821-64.
Momparler, R. L. and Bovenzi, V. (2000). DNA methylation and cancer. J. Cell. Physiol. 183,
145-9.
Mutch, E., Nave, R., McCracken, N., Zech, K. and Williams, F. M. (2007). The role of esterases
in the metabolism of ciclesonide to desisobutyryl-ciclesonide in human tissue. Biochem Pharmacol.
73(10), 1657-7.
Nelson, D. R., Zeldin, D. C., Hoffman, S. M. G., Maltais, L. G., Wain, H. M. and Nebert, D.
W. (2004). Comparison of cytochrome P450 (CYP) genes from the mouse and human genomes,
including nomenclature recommendations for genes, pseudogenes and alternative-splice variants.
Pharmacogenetics 14, 1-18.
Nemoda, Z., Angyal, N., Tarnok, Z., Gadoros, J. and Sasvari-Szekely, M. (2009).
Carboxylesterase
1
gene
polymorphism
and
methylphenidate
response
in
ADHD.
Neuropharmacology 57, 731-2.
Oxford, U. (2014). Aetiology. Oxford uni. 2014,.
Paine, M. F., Shen, D. D., Kunze, K. L., Perkins, J. D., Marsh, C. L., McVicar, J. P., Barr, D.
M., Gillies, B. S. and Thummel, K. E. (1996). First-pass metabolism of midazolam by the human
intestine. Clin. Pharmacol. Ther. 60, 14-14.
Page 57 of 75
Pang, K. S., Barker, F., Cherry, W. F. and Goresky, C. A. (1991). Esterases for enalapril
hydrolysis are concentrated in the perihepatic venous region of the rat liver. J. Pharmacol. Exp.
Ther. 257, 294-7.
Patki, K. C., Von Moltke, L. L. and Greenblat, D. J. (2003). In vitro metabolism of midazolam,
triazolam, nifedipine, and testosterone by human liver microsomes and recombinant cytochromes
p450: role of cyp3a4 and cyp3a5. Drug Metab. Dispos. 31, 938-6.
Peck, R. W. (2007). Driving earlier clinical attrition: if you want to find the needle, burn down the
haystack. Considerations for biomarker development. Drug Discov. Today 12, 289-5.
Perucca, E., Cloyd, J., Critchley, D. and Fuseau, E. (2008). Rufinamide: clinical
pharmacokinetics and concentration-response relationships in patients with epilepsy. Epilepsia 49,
1123-18.
Piccioni, D. and Nguyen, A. (2014). Efficacy, Safety and CNS Exposure of G-202 in Patients With
Recurrent or Progressive Glioblastoma. ClinicalTrials. gov 2014, 1.
Pindel, E. V., Kedishvili, N. Y., Abraham, T. L., Brzezinski, M. R., Zhang, J., Dean, R. A. and
Bosron, W. F. (1997). Purification and Cloning of a Broad Substrate Specificity Human Liver
Carboxylesterase That Catalyzes the Hydrolysis of Cocaine and Heroin. The Journal of Biological
Chemistry 272, 14769-6.
Quinney, S. K., Sanghani, S. P., Davis, W. I., Hurley, T. D., Sun, Z., Murry, D. J. and Bosron,
W. F. (2005). Hydrolysis of capecitabine to 5'-deoxy-5-fluorocytidine by human carboxylesterases
and inhibition by loperamide. J Pharmacol Exp Ther. 313(3), 1011-6.
Quinones, M. P. and Kaddurah-Daouk, R. (2009). Metabolomics tools for identifying
biomarkers for neuropsychiatric diseases. Neurobiol. Dis. 35, 165-11.
Ramirez, J., Innocenti, F., Flockhart, D. A., Santucci, R. and Ratain, M. J. (2004). Role of
CYP3A4 and CYP2B6 in the in virto N-demethylation of meperidine. Clinical Pharmacology &
Therapeutics 75, 85.
Page 58 of 75
Raunio, H. and Rahnasto-Rilla, M. (2012). CYP2A6: genetics, structure, regulation, and function.
Drug Metabol. Drug Interact. 27, 73-15.
Rhoades, J. A., Peterson, Y. K., Zhu, H. J., Appel, D. I., Peloquin, C. A. and Markowitz, J. S.
(2012). Prediction and in vitro evaluation of selected protease inhibitor antiviral drugs as inhibitors
of carboxylesterase 1: a potential source of drug-drug interactions. Pharm. Res. 29, 972-10.
Roberts, A., Yang, J., Halpert, J., Nelson, S., Thummel, K. and Atkins, W. (2011). The
structural basis for homotropic and heterotropic cooperativity of midazolam metabolism by human
cytochrome P450 3A4. Biochemistry 50, 10804-14.
Roby, C. A., Anderson, G. D., Kantor, E., Dryer, D. A. and Burstein, A. H. (2000). St John's
Wort: effect on CYP3A4 activity. Clin. Pharmacol. Ther. 67, 451-6.
Rodríguez-Antona, C., Bort, R., Jover, R., Tindberg, N., Ingelman-Sundberg, M., GómezLechón,
M.
J.
and
Castell,
Transcriptional regulation of human CYP3A4 basal expression
J.
by
V.
CCAAT
(2003).
enhancer-binding
protein alpha and hepatocyte nuclear factor-3 gamma. Mol pharmacol 63, 1180-9.
Ross, M. K., Borazjani, A., Mangum, L. C., Wang, R. and Crow, J. A. (2014). Effects of
toxicologically relevant xenobiotics and the lipid-derived electrophile 4-hydroxynonenal on
macrophage cholesterol efflux: silencing carboxylesterase 1 has paradoxical effects on cholesterol
uptake and efflux. Chem. Res. Toxicol. 27, 1743-13.
Ross, M. K., Streit, T. M. and Herring, K. L. (2010). Carboxylesterases: Dual roles in lipid and
pesticide metabolism. J. Pestic. Sci. 35, 257-7.
Rowland, P., Blaney, F. E., Smyth, M. G., Jones, J. J., Leydon, V. R., Oxbrow, A. K., Lewis,
C. J., Tennant, M. G., Modi, S., Eggleston, D. S. et al. (2006). Crystal structure of human
cytochrome P450 2D6. J. Biol. Chem. 281, 7614-8.
Saenger, P. (1983). 6β-Hydroxycortisol in random urine samples as an indicator of enzyme
induction. Clinical Pharmacology and Therapeutics 34, 818-2.
Page 59 of 75
Sai, K., Saito, Y., Tatewaki, N., Hosokawa, M., Kaniwa, N., Nishimaki-Mogami, T., Naito, M.,
Sawada, J., Shirao, K., Hamaguchi, T. et al. (2010). Association of carboxylesterase 1A
genotypes with irinotecan pharmacokinetics in Japanese cancer patients. Br J Clin Pharmacol.
70(2), 222-11.
Sanghani, S., Sanghani, P., Schiel, M. and Bosron, W. (2009). Human carboxylesterases: an
update on CES1, CES2 and CES3. Protein pept. Lett. 16, 1207-14.
Satoh, T. and Hosokawa, M. (2006). Structure function and regulation of carboxylesterases.
Chemico-Biological Interactions 162(3), 195-16.
Satoh, T. and Hosokawa, M. (2010). Carboxylesterases: structure, function and polymorphism in
mammals. J. Pestic. Sci. 35(3), 218-10.
Satoh, T., Taylor, P., Bosron, W. F., Sanghani, S. P., Hosokawa, M. and La Du, B. N. (2002).
Current Progress on Esterases: From Molecular Structure to Function. Drug Metab. Dispos. 30,
488-5.
Scott, E. E. and Halpert, J. R. (2005). Structures of cytochrome P450 3A4. Trends Biochem. Sci.
30, 5-7.
Shi, D., Yang, J., Yang, D., LeCluyse, E. L., Black, C., You, L., Akhlaghi, F. and Yan, B.
(2006). Anti-influenza prodrug oseltamivir is activated by carboxylesterase human carboxylesterase
1, and the activation is inhibited by antiplatelet agent clopidogrel. J. Pharmacol. Exp. Ther. 319,
1477-7.
Spaggiari, D., Geiser, L., Daali, Y. and Rudaz, S. (2014). A cocktail approach for assessing the in
vitro activity of human cytochrome p450s: An overview of current methodologies. J. Pharm
Biomed Anal. 14, 145-4.
Srinivas, N. R., Hubbard, J. W., Korchinski, E. D. and Midha, K. K. (1992). Stereoselective
urinary pharmacokinetics of dl-threo-methylphenidate and its major metabolite in humans. J.
Pharm. Sci. 81, 747-2.
Page 60 of 75
Szklarz, G. D. and Halpert, J. R. (1997). Molecular modeling of cytochrome P450 3A4. J.
Comput. Aided Mol. Des. 11, 265-7.
Thomsen, R., Rasmussen, H. B., Linnet, K. and INDICES Consortium. (2014). In vitro drug
metabolism by human carboxylesterase 1: focus on angiotensin-converting enzyme inhibitors. Drug
Metab. Dispos. 42, 126-7.
Tirona, R. G., Lee, B. F., Lan, L. B., Cline, C. B., Lamba, V., Parviz, F., Duncan, S. A., Inoue,
Y., Gonzalez, F. J., Schuetz, E. G. et al. (2003). The orphan nuclear receptor HNF4alpha
determines PXR- and CAR-mediated xenobiotic induction of CYP3A4. Nat. Med. 9, 220-3.
Vickers, S., Duncan, C. A., Chen, I. W., Rosegay, A. and Duggan, D. E. (1990). Metabolic
disposition studies on simvastatin, a cholesterol-lowering prodrug. Drug Metab Dispos. 18(2), 1387.
Vistoli, G., Pedretti, A., Mazzolari, A. and Testa, B. (2010). In silico prediction of human
carboxylesterase-1 (hCES1) metabolism combining docking analyses and MD simulations. Bioorg.
Med. Chem. 18, 320-9.
Wang, B., Yang, L. P., Zhang, X. Z., Hunag, S. Q., Bartlam, M. and Zhou, S. F. (2009). New
insights into the structural characteristics and functional relevance of the human cytochrome P450
2D6 enzyme. Drug. Metab. Rev. 41, 573-70.
Watkins, P. B., Wrighton, S. A., Schuetz, E. G., Molowa, D. T. and Guzelian, P. S. (1987).
Identification of glucocorticoid-inducible cytochromes P-450 in the intestinal mucosa of rats and
man. J. Clin. Invest. 80, 1029-7.
Wellen, K. E., Hatzivassiliou, G., Sachdeva, U. M., Bui, T. V., Cross, J. R. and Thompson, C.
B. (2009). ATP-citrate lyase links cellular metabolism to histone acetylation. Science 324, 1076-4.
Westlind, A., Malmebo, S., Johansson, I., Otter, C., Andersson, T. B., Ingelman-Sundberg, M.
and Oscarson, M. (2001). Cloning and tissue distribution of a novel human cytochrome p450 of
the CYP3A subfamily, CYP3A43. Biochem. Biophys. Res. Commun. 281, 1349-6.
Page 61 of 75
Williams, E. T., Carlson, J. E., Lai, W. G., Wong, Y. N., Yoshimura, T., Critchley, D. J. and
Narurkar, M. (2011). Investigation of the metabolism of rufinamide and its interaction with
valproate. Drug Metab Lett. 5(4), 280-9.
Williams, J. A., Ring, B. J., Cantrell, V. E., Jones, D. R., Eckstein, J., Ruterbories, K.,
Hamman, M. A., Hall, S. D. and Wrighton, S. A. (2002). Comparative metabolic capabilities of
CYP3A4,CYP3A5, and CYP3A7. Drug Metab. Dispos. 30, 883-8.
Williams, J. A., Hyland, R., Jones, B. C., Smith, D. A., Hurst, S., Goosen, T. C., Peterkin, V.,
Koup, J. R. and Ball, S. E. (2004). Drug−drug interactions for udp-glucuronosyltransferase
substrates. Drug Metab. Dispos. 32(11), 1201-7.
Williams, E. T., Bacon, J. A., Bender, D. M., Lowinger, J. J., Guo, W. K., Ehsani, M. E.,
Wang, X., Wang, H., Qian, Y. W., Ruterbories, K. J. et al. (2011). Characterization of the
expression and activity of carboxylesterases 1 and 2 from the beagle dog, cynomolgus monkey, and
human. Drug Metab. Dispos. 39, 2305-8.
Williams, F. M. (1985). Clinical significance of esterases in man. Clin. Pharmacokinet. 10, 392-11.
Williams, P. A., Cosme, J., Vinkovic, D. M., Ward, A., Angove, H. C., Day, P. J., Vonrhein, C.,
Tickle, I. J. and Jhoti, H. (2004). Crystal structures of human cytochrome P450 3A4 bound to
metyrapone and progesterone. Science 305, 683-3.
Xu, J., Li, Y., Chen, W. D., Xu, Y., Yin, L., Ge, X., Jadhav, K., Adorini, L. and Zhang, Y.
(2014). Hepatic carboxylesterase 1 is essential for both normal and farnesoid X receptor-controlled
lipid homeostasis. Hepatology 59, 1761-10.
Xu, J., Yin, L., Xu, Y., Li, Y., Zalzala, M., Cheng, G. and Zhang, Y. (2014). Hepatic
carboxylesterase 1 is induced by glucose and regulates postprandial glucose levels. PLoS One 9,
e109663.
Yang, X., Morris, S. M., Gearhart, J. M., Ruark, C. D., Paule, M. G., Slikker, W.,Jr,
Mattison, D. R., Vitiello, B., Twaddle, N. C., Doerge, D. R. et al. (2014). Development of a
physiologically based model to describe the pharmacokinetics of methylphenidate in juvenile and
adult humans and nonhuman primates. PLoS One 9, e106101.
Page 62 of 75
Yano, J. K., Wester, M. R., Schoch, G. A., Griffin, K. J., Stout, C. D. and Johnson, E. F.
(2004). The structure of human microsomal cytochrome P450 3A4 determined by X-ray
crystallography to 2.05-A resolution. J. Biol. Chem. 279, 38091-3.
Yi, L. Z., He, J., Liang, Y. Z., Yuan, D. L. and Chau, F. T. (2006). Plasma fatty acid metabolic
profiling and biomarkers of type 2 diabetes mellitus based on GC/MS and PLS-LDA. FEBS Lett.
580, 6837-8.
Yoshimura, M., Kimura, T., Ishii, M., Ishii, K., Matsuura, T., Geshi, E., Hosokawa, M. and
Muramatsu, M. (2008). Functional polymorphisms in carboxylesterase1A2 (CES1A2) gene
involves specific protein 1 (Sp1) binding sites. Biochem. Biophys. Res. Commun. 369, 939-3.
Zanger, U. M. and Hofmann, M. H. (2008). Polymorphic cytochromes P450 CYP2B6
and CYP2D6: recent advances on single nucleotide polymorphisms affecting splicing. Acta Chim
Slov 55, 38.
Zanger, U. M. and Schwab, M. (2013). Cytochrome P450 enzymes in drug metabolism:
Regulation of gene expression, enzyme activities, and impact of genetic variation. Pharmacol Ther.
138, 103-38.
Zhang, J., Burnell, J. C., Dumaual, N. and Bosron, W. F. (1999). Binding and hydrolysis of
meperidine by human liver carboxylesterase hCE-1. J. Pharmacol. Exp. Ther. 290, 314-4.
Zhao, B., Bie, J., Wang, J., Marqueen, S. A. and Ghosh, S. (2012). Identification of a novel
intracellular cholesteryl ester hydrolase (carboxylesterase 3) in human macrophages: compensatory
increase in its expression after carboxylesterase 1 silencing. Am. J. Physiol. Cell. Physiol. 303, 4278.
Zhou, S. F. (2009). Polymorphism of human cytochrome P450 2D6 and its clinical significance.
Clinical pharmacokinetics 48(12), 761-43.
Zhu, H. J., Wang, X., Gawronski, B. E., Brinda, B. J., Angiolillo, D. J. and Markowitz, J. S.
(2013). Carboxylesterase 1 as a determinant of clopidogrel metabolism and activation. J Pharmacol
Exp Ther. 344(3), 665-7.
Page 63 of 75
Zhu, H. J., Appel, D. I., Johnson, J. A., Chavin, K. D. and Markowitz, J. S. (2009). Role of
carboxylesterase 1 and impact of natural genetic variants on the hydrolysis of trandolapril. Biochem.
Pharmacol. 77, 1266-6.
Zhu, H. J., Appel, D. I., Peterson, Y. K., Wang, Z. and Markowitz, J. S. (2010). Identification
of selected therapeutic agents as inhibitors of carboxylesterase 1: potential sources of metabolic
drug interactions. Toxicology 270, 59-6.
Zhu, H. J. and Markowitz, J. S. (2013). Carboxylesterase 1 (CES1) genetic polymorphisms and
oseltamivir activation. Eur. J. Clin. Pharmacol. 69, 733-1.
Zhu, H. J., Patrick, K. S., Yuan, H. J., Wang, J. S., Donovan, J. L., DeVane, C. L., Malcolm,
R., Johnson, J. A., Youngblood, G. L., Sweet, D. H. et al. (2008). Two CES1 gene mutations lead
to dysfunctional carboxylesterase 1 activity in man: clinical significance and molecular basis. Am. J.
Hum. Genet. 82, 1241-7.
Page 64 of 75
Appendices
Appendix #1
A comparison of the clusters in the cholesterol hCES1 docking analysis, before and after the
removal of cluster which did not meet the ΔG criteria.
Appendix #2
A comparison of the clusters in the 27-hydroxycholesterol hCES1 docking analysis, before and after
the removal of cluster which did not meet the ΔG criteria.
Page 65 of 75
Appendix #3
Tables showing the calculations and results used in the docking experiment for 27hydroxycholesterol and cholesterol.
Cluster
0
0
1
1
2
2
3
3
4
4
5
5
6
6
7
7
8
8
9
9
11
11
12
12
13
13
14
14
16
16
17
27hydroxycholesterol
Model
#1.7 (8)
#1.1 (2, 3, 4)
#1.13 (14, 15, 16)
#1.9 (10)
#1.17
#1.20 (21, 22, 23, 24)
#1.31 (32)
#1.25 (26, 27, 28)
#1.38 (39, 40)
#1.33 (34)
#1.46 (47, 48)
#1.42 (43, 44, 45, 46)
#1.52
#1.53 (54, 55, 56, 57)
#1.57 (58)
#1.59
#1.69
#1.67 (68)
#1.78 (79, 80)
#1.73 (74)
#1.95 (96)
#1.89 (90, 91)
#1.97 (98, 99)
#1.102 (03, 04)
#1.107 (08, 09, 10,
11)
#1.105 (06)
#1.120
#1.117
#1.133 (34, 35, 36)
#1.129 (30)
#1.143 (44)
Total
number
of
models
ΔG
Average of
the cluster
2
4
4
2
1
5
2
4
3
2
3
5
1
5
2
1
1
2
3
2
2
3
3
3
-7,511
-8,069
-7,056
-7,239
-7,440
-7,495
-7,263
-7,301
-7,186
-7,468
-6,951
-7,102
-7,104
-7,196
-6,718
-6,779
-6,625
-7,195
-6,696
-6,874
-6,798
-6,844
-6,858
-6,905
-7,883
5
2
1
1
4
2
2
-6,609
-6,803
-6,651
-7,154
-6,891
-7,160
-6,775
-6,665
-7,117
-7,486
-7,288
-7,299
-7,045
-7,181
-6,738
-7,005
-6,767
-6,826
-6,881
-6,902
-6,981
-7,493
FullFitness
-2416,674
-2422,355
-2417,496
-2419,681
-2417,686
-2416,927
-2417,513
-2417,646
-2415,501
-2417,551
-2413,739
-2416,102
-2413,982
-2413,125
-2414,370
-2414,312
-2413,243
-2413,603
-2413,107
-2414,017
-2412,538
-2413,304
-2411,977
-2411,731
-2411,191
-2411,400
-2402,603
-2409,107
-2409,602
-2410,969
-2408,441
Average for
the cluster
-2420,461
-2418,225
-2417,054
-2417,601
-2416,321
-2415,216
-2413,267
-2414,351
-2413,483
-2413,471
-2412,998
-2411,854
-2411,251
-2405,855
-2410,057
-2409,516
Page 66 of 75
17 #1.137 (38, 39)
19 #1.156
19 #1.153
#1.157 (58, 59, 60,
20 61)
20 #1.170 (71, 72)
21 #1.183
21 #1.186 (87)
22 #1.195 (96)
22 #1.193 (94)
#1.206 (07, 08, 09,
24 10)
24 #1.203 (04)
27 #1.227 (28, 29)
27 #1.232 (33, 34)
31 #1.253
Total
and
average
3
1
1
-7,971
-7,278
-7,389
5
3
1
2
2
2
-6,924
-7,159
-6,619
-6,663
-6,934
-6,951
-7,012
5
2
3
3
1
-6,701
-6,839
-6,612
-6,632
-6,629
-6,740
-7,042
-7,334
-6,648
-6,942
-2410,232
-2407,384 -2408,252
-2409,120
-2408,912 -2408,170
-2406,932
-2407,226 -2405,152
-2404,115
-2407,691 -2407,695
-2407,698
-2407,076 -2407,156
-2407,357
-6,622 -2406,918 -2406,828
-2406,738
-6,629 -2402,833 -2402,833
N/A -2412,330
116
Cholesterol:
Cluster
Model
7 #1.53(54 og 55)
7 #1.50 (51)
12 #1.91 (92)
12 #1.95
16 #1.119 (120)
16 #1.121 (22, 23, 24)
17
17
18
18
19
19
20
20
21
#1.132
#1.128 (29, 30, 31)
#1.138 (39, 40)
#1.133 (34)
#1.147 (48)
#1.141
#1.154 (55, 56)
#1.149 (50, 51, 52)
#1.157
Total
ΔG
Average of
FullFitness Average for
number of
the cluster
the cluster
models
3 -6,446047
-6,459 -2406,670
-2406,680
2
-2406,695
6,4792695
2
-6,393 -2403,389
-2403,266
6,3395786
1
-2403,020
6,5003886
2 -6,489759
-6,546 -2402,580
-2402,544
4
-2402,525
6,5746174
1 -6,587872
-6,596 -2401,511
-2402,786
4 -6,598333
-2403,104
3 -6,314338
-6,367 -2402,367
-2402,692
2 -6,446011
-2403,179
2 -6,172285
-6,174 -2402,331
-2402,339
1
-6,178
-2402,356
3
-6,223
-6,242 -2401,659
-2401,828
4
-6,255
-2401,955
1
-6,439
-6,464 -2401,677
-2401,638
Page 67 of 75
21
22
22
23
23
24
24
25
26
26
27
27
28
28
30
30
31
31
33
33
34
34
#1.160
#1.165 (66, 67, 68)
#1.161 (62, 63)
#1.171
#1.169 (70)
#1.176 (77)
#1.178(79, 80)
#1.187 (88)
#1.189
#1.190 (91)
#1.198 (99, 200)
#1.193 (94, 95)
#1.201 (202, 203,
204)
#1.205 (06, 07)
#1.223 (24)
#1.220 (21, 22)
#1.225
#1.230 (31, 32)
#1.247 (48)
#1.241 (42, 43, 44)
#1.254
#1.249 (50, 51, 52,
53)
Total
and
average
1
4
3
1
2
2
3
2
1
2
3
3
4
-6,488
-5,687
-5,887
-6,448
-6,517
-6,157
-6,175
-6,541
-6,038
-6,133
-6,325
-6,327
-6,462
3
2
3
1
3
2
4
1
5
-6,467
-6,338
-6,451
-6,380
-6,446
-5,932
-5,957
-6,427
-6,531
90 -6,317
-5,773
-6,494
-6,168
-6,541
-6,102
-6,326
-6,464
-6,405
-6,429
-5,949
-6,513
-6,459
-2401,6
-2401,087
-2401,444
-2400,12
-2401,327
-2400,553
-2400,517
-2399,377
-2400,107
-2399,677
-2399,895
-2399,981
-2399,737
-2399,700
-2398,123
-2398,757
-2398,804
-2398,244
-2396,591
-2396,772
-2396,562
-2396,640
-2401,240
-2400,925
-2400,531
-2399,377
-2399,820
-2399,938
-2399,721
-2398,504
-2398,384
-2396,712
-2396,627
N/A -2400,72
Appendix #4
Cholesterol molecule used in the docking analyses -ZINC4175349:
27-hydroxycholesterol molecule used in the docking analyses - ZINC4097086:
Page 68 of 75
Methylphenidate molecule used in the docking analyses – ZINC896709:
Appendix #5
hCES1 has, as 3A4 and 2D6, some xenobiotic inhibitors. These are shown below in Table 9.
Table 9: Substrates that causes an inhibition of hCES1
Substrate
Reference
Clopidogrel
(Shi et al., 2006)
Ethanol
(Laizure et al., 2003)
Israpidine
(Thomsen et al., 2014)
Nelfinavir
(Rhoades et al., 2012)
Procainamide
(Bailey and Briggs, 2003)
Tacrine
(Bencharit et al., 2003)
Tacrolimus
(Thomsen et al., 2014)
Quinidine
(Bailey and Briggs, 2003)
Page 69 of 75