thuoc từ curcumin
TRANSCRIPT
"rug Biotransformation En0yme Interactions
Studies with 9urcumin, 9urcumin analogues and other Plant=derived 9omponents
!egina (ppiah+,pong
Drug Biotransformation Enzyme Interactions
Studies with Curcumin, Curcumin analogues and other Plant-derived Components
Regina Appiah-Opong
Drug Biotransformation Enzyme Interactions: Studies with Curcumin, Curcumin analogues and other Plant-derived Components Regina Appiah-Opong KNAW and GETFUND scholarship schemes are gratefully acknowledged for the financial support in the printing of this thesis. Printed by Printpartners Ipskamp © Regina Appiah-Opong, Legon, Ghana 2009. All rights reserved. No part of this thesis may be reproduced in any form or by any means without permission from the author.
VRIJE UNIVERSITEIT
Drug Biotransformation Enzyme Interactions Studies with Curcumin, Curcumin analogues and
other Plant-derived Components
ACADEMISCH PROEFSCHRIFT
ter verkrijging van de graad Doctor aan de Vrije Universiteit Amsterdam,
op gezag van de rector magnificus prof.dr. L.M. Bouter,
in het openbaar te verdedigen ten overstaan van de promotiecommissie
van de faculteit der Exacte Wetenschappen op woensdag 6 mei 2009 om 13.45 uur
in de aula van de universiteit, De Boelelaan 1105
door
Regina Appiah-Opong
geboren te Kumasi, Ghana
promotor : prof.dr. N.P.E. Vermeulen copromotor : dr. J.N.M. Commandeur
Reading Committee: prof.dr. Ivonne M.C.M. Rietjens prof.dr. Rob Leurs dr. Iwan J.P. de Esch dr. Chris Oostenbrink dr. Martijn Rooseboom
The investigations described in this thesis were carried out in the Leiden Amsterdam Center for Drug Research (LACDR)/Division of Molecular Toxicology, Department of Chemistry and Pharmaceutical Science, Faculty of Sciences, Vrije Universiteit, De Boelelaan 1083, 1081 HV Amsterdam, The Netherlands.
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
By wisdom the Lord laid the earths foundation, by understanding He set the Heavens in place; !
by His knowledge the deeps were divided and the clouds let drop the dew. Prov. 3: 19, 20
Contents Introduction ! Chapter !* +eneral introduction 3 Chapter 4* Curcumin* pharmacokinetics, metabolism, and potential for drug-
drug=food Interactions ?! Inhibition of C1P3GST activities by natural products and derivatives @3 Chapter 3* Inhibition of human recombinant cytochrome B?CDs by curcumin
and curcumin decomposition products @C Chapter ?* Structure-activity relationship of inhibition of recombinant human
cytochrome B?CD mediated metabolism by curcumin analogues G3 Chapter C* Inhibition of human glutathione S-transferases by curcumin and
analogues !DH Chapter @* Interactions between cytochromes B?CD, glutathione S-
transferases and +hanaian medicinal plants! !4J Summary, conclusions and perspectives Chapter J* Summary, conclusions and perspectives !?C Kppendices* Lutch summary !CH Mist of publications !@3 Epilogue !@C Mist of abbreviations !@@
1
"#$%&'()$*&#
2
!
!"
Chapter 1
General Introduction
Drug disposition/Biotransformation
#$%$&'" ()'*&$+,+" *)-" .(&+/*&/01" -23(+-4" /(" 2-&(5$(/$.+6" 70$,$&*/$(&" (8" /9-+-"
2-&(5$(/$.+:" $+" +$'&$8$.*&/01" -&9*&.-4" 51" /9-" 3)(.-++" ;&(<&" *+" 5$(/)*&+8(),*/$(&:"
<9-)-51" /9-" 0$3(39$0$."+=5+/*&.-" $+".(&%-)/-4" $&/("*",()-"914)(39$0$." 8()," /9*/" 8*%()+"
-2.)-/$(&" $&"=)$&-:"5$0-"()" 8*-.-+6">9-"5$(/)*&+8(),*/$(&"(8".-)/*$&"4)='+" )-+=0/+" $&" /9-"
8(),*/$(&"(8"*",-/*5(0$/-"/9*/"$+",()-"39*),*.(0('$.*001"*./$%-"*&4"0-++"/(2$."/9*&"/9-"
3*)-&/" 4)='" ?@A6" B$(/)*&+8(),*/$(&" ,*1" *0+(" )-+=0/" $&" /9-" 8(),*/$(&" (8" )-*./$%-"
,-/*5(0$/-+"*&4"+=5+-C=-&/01"*4%-)+-"-88-./+" ?D:!A6">9-" 0$%-)" $+" /9-",*E()"+$/-"(8"4)='"
5$(/)*&+8(),*/$(&:"*0/9(='9"5$(/)*&+8(),*/$(&"*0+("(..=)+" $&"(/9-)"()'*&+"*&4"/$++=-+6"
F-&(5$(/$." 5$(/)*&+8(),*/$(&" )-*./$(&+" *)-" '-&-)*001" 4$%$4-4" $&/(" /9)--" ')(=3+:" .*00-4"
39*+-" G:" 39*+-" GG" *&4" 39*+-" GGG6" H9*+-" G" )-*./$(&+" $&%(0%-" (2$4*/$(&:" )-4=./$(&" *&4"
914)(01+$+:" <9-)-" *" 8=&./$(&*0" ')(=3" $+" -23(+-4" ()" $&/)(4=.-4:" =+=*001" )-+=0/$&'" $&" *"
+,*00"$&.)-*+-"$&"2-&(5$(/$."914)(39$0$.$/16"I&"/9-"(/9-)"9*&4:"39*+-"GG")-*./$(&+"$&%(0%-"
.(&E='*/$(&" )-*./$(&+" $&.0=4$&':" .(&E='*/$(&" <$/9" '0=/*/9$(&-" *&4" (/9-)" *,$&(" *.$4+:"
'0=.=)(&$4*/$(&:"+=08*/$(&:"*.-/10*/$(&"*&4",-/910*/$(&6">9-+-")-*./$(&+"=+=*001")-+=0/"$&"
+/)(&'-)" $&.)-*+-+" $&" 914)(39$0$.$/1" )-C=$)-4" 8()" -0$,$&*/$(&" (8" +=5+/*&.-+6" H9*+-" GGG"
)-*./$(&+" $&%(0%-" -23()/$&'" (8" 3)(4=./+" (8" 39*+-" G" *&4" 39*+-" GG" )-*./$(&+" 51" %*)$(=+"
/)*&+3()/-)+" +=.9" *+" HJ'01.(3)(/-$&" KHJ'3L:" ()'*&$." *&$(&" /)*&+3()/$&'" 3(013-3/$4-+"
KIM>H+L"*&4",=0/$4)='")-+$+/*&.-"3)(/-$&+"KNOH+L"?PA6""
" N(4=0*/(&" (8" ,*E()" 5$(/)*&+8(),*/$(&" -&Q1,-+" +=.9" *+" .1/(.9)(,-" HPRS+"
KTUH+L"51"4)='+:"8((4".(,3(&-&/+"*&4"9-)5*0"3)(4=./+"$&"/9-"3)-+-&.-"(8"+(,-"4)='+"
.(=04" )-+=0/" $&"4)='J4)='V8((4V9-)5" $&/-)*./$(&+"<$/9"*4%-)+-"+$4-"-88-./+" ?RJWA6"I&" /9-"
(/9-)" 9*&4",(4=0*/$(&" (8" '0=/*/9$(&-" XJ/)*&+8-)*+-+" 51" /9-+-" 2-&(5$(/$.+" .(=04" 9*%-"
/(2$.(0('$.*0" .(&+-C=-&.-+:" 4=-" /(" /9-" ,*E()" )(0-" 30*1-4" 51" /9-+-" -&Q1,-+" $&"
4-/(2$8$.*/$(&6" >9=+:" /9-)-" $+" /9-" &--4" 8()" -%*0=*/$(&" (8" 8((4" .(,3(&-&/+" *&4" 9-)5*0"
3)(4=./+"8()"/9-"3(/-&/$*0"8()"4)='J8((4V9-)5"$&/-)*./$(&+:"/("4-.)-*+-"/9-")$+;"(8"9*),8=0"
+$4-"-88-./+6">*50-"@"+9(<+" /9-" /9)--"39*+-+"(8" /9-"3)(.-++"(8"5$(/)*&+8(),*/$(&:" /9-"
-&Q1,-+"*&4".9-,$.*0")-*./$(&+"$&%(0%-46"
"
4
Table 1. Biotransformation enzymes and reactions involved in the biotransformation
Phase Enzyme Type of reaction
I Cytochrome P450
Oxidation, epoxidation,
reduction, dealkylation,
dehalogenation
Monoamine oxidase Oxidation
Flavin containing mono-oxygenase Oxidation
NAD(P)H quinine oxidoreductase Reduction
Peroxidase Peroxidation
Alcohol dehydrogenase Oxidation
Aldehyde dehydrogenase Oxidation
II Glutathione S-transferase Conjugation
UDP-glucuronyl transferase Conjugation
Sulfotransferase Conjugation
Amino acid transferase Conjugation
N-acetyl transferase Acetylation
Methyl transferase Methylation
Epoxide hydrolase Hydration
III MRP/Pgp Transportation
OATP Transportation
N-acetyl transferase Acetylation, formation of
mercapturic acids
!-lyase !-elimination
Amino acid oxidases Deamination
Adapted from references [1] and [4].
Drug-drug/herb/food interactions
The effects of a particular drug on the efficacy and/or toxicity of another drug, which is
commonly referred to as drug-drug interactions have become an important healthcare
issue [8]. Drug-drug interaction is a phenomenon that can elude both regulatory
agencies and pharmaceutical companies. A drug that apparently is safe after extensive
evaluation in both preclinical and clinical trials can be extremely harmful when co-
administered with another drug. A typical example is the interaction between the non-
sedative antihistamine drug, terfenadine and ketoconazole, an antifungal drug [5]. The
5
co-administration of these drugs resulted in cardiotoxicity, and subsequently death of
some patients. Ketoconazole is a potent inhibitor of the CYP3A4 metabolism of
terfenadine, consequently co-administration of these two drugs leads to an increase in
plasma concentration of terfenadine to a cardio toxic level. Thus, drug-drug interactions,
remain important drug properties that should be well defined before humans are
exposed to any new drug.
Natural products such as foods and herbal products may also possess the
potential to cause harmful effects upon concomitant intake with drugs. Food-drug
interactions that have been investigated include that of grapefruit juice, orange juice,
broccoli, cabbage, Brussels sprouts, charcoal-grilled meats and garlic (Tables 2 and 3).
Grapefruit juice, is a potent inhibitor of CYP3A4-mediated drug metabolism [9].
CYP3A4-mediated metabolism of several drugs has been shown to be affected by
grapefruit juice. Clinically relevant interactions with grapefruit juice seem likely for most
dihydropyridines, terfenadine, saquinavir, cyclosporine, midazolam, triazolam and
verapamil and may also occur with lovastatin, cisapride and astemizole [9]. Grapefruit
and other fruit juices have also been shown to be potent in vitro inhibitors of a number
of organic anion-transporting polypeptides (OATPs). These juices were also found to
decrease the absorption of the non-metabolized OATP substrate, fexofenadine, hence
decrease drug bioavailability. Such findings enhance our understanding of the complex
nature of food-drug interactions, their possible influence on the clinical effects of
medications, and also underscore the need for similar investigations on other foods.
Herbal medicines are usually mixtures of more than one active ingredient; hence
the likelihood of herbal interactions is theoretically higher than drug-drug interactions.
Herb-drug interactions are a major concern, especially due to the potential to cause
adverse effects. These herbal products may modulate drug metabolizing enzymes or
transporters (e.g. CYPs or P-glycoprotein (Pgp)). Herbal medicines such as St. John’s
wort, ginseng and gingko, which are freely available over the counter, have been
reported to cause serious clinical interactions when co-administered with prescribed
medicines [8]. St. John’s wort, which is popularly used as an anti-depressant, has been
implicated in a potentially fatal interaction with cyclosporin. There is evidence that the
mechanism of interaction involves induction of CYP3A4 through the activation of
6
pregnane X-receptor, as well as induction of intestinal multidrug resistance transporter
protein MDR1/Pgp [10,11]. Thus, drug-drug/food/herb interactions may occur at the
level of receptors, transporters and/or biotransformation enzymes.
Drug-drug/food/herb interactions associated with transporters and receptors
Drug-drug interactions involving inhibition or induction of transport proteins have been
demonstrated [11,12]. Drug transporters play a significant role in absorption,
distribution, metabolism and excretion of drugs. Several drug transporter proteins have
been identified, however, the interactions mediated by P-gp are the most widely studied
and best understood. Approximately 50% of marketed drugs have been identified to be
Pgp substrates and/or inhibitors and it has broad substrate specificity. Drug transporters
and metabolizing enzymes may also act synergistically to cause drug-drug interaction.
For instance, in cardiovascular treatment, drug interactions result when therapeutic
doses of verapamil inhibit Pgp and CYP3A4, causing a marked increase in cyclosporine
absorption and availability [13]. The transporters are also vulnerable to inhibition,
induction and activation by foods and herbal products. Thus, drug-food/herb interactions
may result upon concomitant intake of some food and herbal products with drugs.
Curcumin, ginsenosides, piperine, some catechins from green tea, and silymarin from
milk thistle were found to be inhibitors of Pgp, while St. John's wort induced the
intestinal expression of Pgp in vitro and in vivo [14]. Some components (e.g.,
bergamottin and quercetin) of grapefruit juice were also reported to modulate Pgp
activity. Herbal constituents, in particular flavonoids, have been reported to modulate
Pgp by directly interacting with the vicinal ATP-binding site, the steroid-binding site, or
the substrate-binding site.
Drug receptors have also been shown to mediate drug-drug interactions. The
nuclear pregnane X receptor (PXR) and constitutive androstane receptor (CAR) have
now been discovered and their roles in rifampicin-mediated drug-drug interactions
demonstrated [15]. Rifampicin activates the nuclear PXR that in turn affects
cytochromes P450, glucuronosyltransferases and Pgp activities [16]. Some herbal
constituents (e.g., hyperforin and kava) were also shown to activate PXR [14]. Hence,
investigations on the effects of drugs, foods and herbal products on PXR are important
7
since co-administration of drugs with foods, herbal products or drugs that are PXR
modulators could result in harmful drug-drug/food/herb interactions.
Biotransformation enzymes and drug-drug interactions
Many known pharmacokinetic drug interactions are associated with phase I
biotransformation enzymes, particularly CYP enzymes. Pharmacokinetic CYP-mediated
drug interactions, one of the major causes of attritions in drug development, involves
induction and inhibition of the CYP enzymes with the latter being more common [17,18].
Drug-drug interactions involving CYPs have also been identified as important causes of
adverse drug reactions and therapeutic failure [19]. Intake of multiple drugs increases
the chances of drug-drug interactions and adverse reactions in patients.
Table 2. Human drug-drug/food interactions mediated by CYP inhibition
CYP inhibitor drug/substance involved
Inhibited CYP
Known or possible effect
Ref
Drug
Ketoconazole, erythromycin
Terfenadine, cyclosporine, tolbutamide
CYP3A4
Renal toxicity, cardiotoxicity
20,21,22
Quinidine, codeine amiadarone, haloperidol,
Fluoxetine CYP2D6 Anorexia, nervousness, tremor, seizures
23
Isoniazid Phenytoin, carbamazepine, diazepam
CYP3A4, CYP2C19
Adversities due to slow elimination of the drugs
24
Fluvoxamine Caffeine CYP1A2 Mental disorder 25,26
Fluconazole, ketoconazole
S-warfarin, phenytoin CYP2C9 Increased anticoagulation
27
Food
Grapefruit, Seville orange juice
Felodipine, terfenadine, saquinavir, cyclosporine, midazolam, triazolam, verapamil lovastatin, cisapride, astemizole
CYP3A4, CYP1A2
Renal toxicity and other toxicities
7,22,23,2428
Table 2 was partly adapted [8].
8
In contrast to the relatively narrow substrate specificity characteristics of other enzymes,
such as epoxidases, most drug-metabolizing CYP enzymes exhibit broad substrate
specificity [5]. Examples of CYP-mediated drug interactions and possible adverse
effects are shown in Tables 2 and 3.
Table 3. Drug interactions mediated by CYP induction
CYP inducer Drug/substance involved
Induced CYP
Known or possible effect
Ref
Drug
Rifampicin,Phenobarbital, phenytoin
Ethinylestradiol
CYP3A4
Ineffective contraceptive
29
Nicotine, omeprazole
Clozapine CYP1A2 Increased clearance of the drug leading to loss of antipsychotic properties
29,30
Food
Cruciferous vegetables (broccoli, cabbage, Brussels, sprouts) charcoal-grilled meats
Theophylline, warfarin, clozapine
CYP1A2 Increased plasma clearance leading to possible loss of control of asthma, reduced anticoagulation
23
Herb
St John’s wort Cyclosporine CYP3A4 Rejection of organ transplant
31
Ginkgo Depakote, Dilantin CYP2C9 Epileptic seizures 32
Table 3 was partly adapted [8].
Thus, inhibition of CYP enzymes could result in accumulation of drugs, and
subsequently lead to serious clinically important drug interactions [33]. Serious toxicity
may develop shortly if the drug has a narrow therapeutic window. A drug can be both a
substrate and an inhibitor for a particular isoenzyme. A non-substrate drug can also
inhibit the activity of an isoenzyme. Regardless of the mechanism, CYP inhibition may
result in decreased metabolism of a drug and alter its pharmacokinetic profile. In vitro
9
CYP-associated metabolic studies have been considered cost-effective for predicting
the potential for clinical drug-drug interactions [17]. Although in vitro findings on drug
interactions may not always correlate with in vivo situations, in vitro findings remain
useful and rapid indicators of potentially harmful drug interactions.
The observed induction and inhibition of CYP enzymes by natural products in the
presence of a prescribed drug has led to the general acceptance that natural products
can have adverse effects, contrary to the popular beliefs of their safety, especially in
countries where there is an active practice of ethnomedicine [8]. Hence, it is imperative
that foods and herbal products be assessed for their potential to cause harmful drug-
food/herb interactions.
CYP inhibition
Enzyme inhibition refers to the reduction in enzyme activity due to the presence of an
inhibitor. CYP inhibition can be reversible, comprising of competitive, non-competitive,
uncompetitive or irreversible inhibition. Irreversible inhibition usually results from
activation of a drug by CYPs into a reactive metabolite, which covalently binds to the
active site of the enzyme causing its inactivation, a process known as suicide, time-
dependent or mechanism-based inhibition (TDI or MBI) [34]. The level of accumulation
and the therapeutic window of a drug are important determinants of the clinical
relevance of a specific drug-drug interaction. CYP inhibition with clinical relevance
includes competitive inhibition and MBI [17]. Figure 1 shows the effect of grapefruit juice
on the CYP3A4-mediated metabolism of simvastatin, an inactive lactone prodrug [35].
Area analysis of figure 1, indicates that the bioavailability of simvastatin increased 13.5
fold (compared to the control, water) in the presence of grapefruit juice [35]. The
dramatic increase in simvastatin plasma concentration has been attributed to the
inhibition of CYP3A4 by grapefruit juice. The exposure to unmetabolized parent drug,
simvastatin, could result in toxicity as observed in the interaction between itraconazole,
a CYP3A4 inhibitor and simvastatin, that is an association with increased risk of skeletal
muscle toxicity [36].
10
Figure 1. Mean serum concentrations of simvastatin in 10 healthy volunteers after single oral doses of 40
mg simvastatin. Simvastatin was taken with 200 ml water (open circles), with 200 ml double strength
grapefruit juice after injestion of 200 ml grapefruit juice 3X daily for 2 days (solid triangles) or with 200 ml
water 24 h (solid diamonds), 3 days (open triangles) or 7 days (solid stars) after last dose of grapefruit
juice (adapted from ref. 35).
The mechanism-based inhibition (MBI) has recently received increasing focus,
with the notion that it could occur more frequently than anticipated, partly due to the
redox cycling-allied enzymatic action of CYPs [17]. Figure 2 shows a scheme of types of
enzyme inhibition [37].
Figure 2. Competitive inhibition (A); noncompetitive inhibition (B); mixed-type of inhibition (C); irreversible inhibition (D); E, enzyme; S, substrate; P, product; I, inhibitor; EI, enzyme-inhibitor complex; EI*, ‘dead-end’ enzyme inhibitor complex; ESI, enzyme-substrate-inhibitor complex; Ki, inhibitor constant; kinact, inactivation rate
A B
C D
E + S ES1
P1 + E
ESIEI
+ IKi'+ IK
i
E + S ES1
P1 + E
ESI
+ IKi
E + S ES1
P1 + E
EI
+ IKi
E + S ES1
P1 + E
ESIEI
EI*
+ IKi'+ IK
i
kinact
C D
11
CYP inhibition studies are extremely valuable, as they allow extrapolation of data to
other compounds, facilitate drug development and also indicate possible drug
interactions in organs other than the liver. Studies on prediction of in vivo drug-drug
interactions via inhibition CYP-mediated metabolism from in vitro data have been
performed [38]. Equation 1 can be used for quantitative prediction of in vivo area under
the concentration versus time curve (AUC) ratio, when the fraction of substrate
metabolized by the inhibited CYP pathway (fmCYP value) and the maximum hepatic
inhibitor concentration are known. Values for fmCYP may be estimated by assessing
exposure differences between extensive and poor metabolizers using probe substrates
[39]. Alternatively, fmCYP values for probe substrates could be obtained by calculating the
difference between the urinary recovery of metabolites in both the presence and
absence of the CYP selective inhibitor. The fmCYP values could also be estimated by the
combination of urinary recovery of metabolites, biliary excretion and the recovery of
unchanged drug [38].
AUCinhibitor = 1 (Equation 1) AUCcontrol fmCYP + (1-fmCYP) 1+[I]/Ki
Where, [I] is maximum hepatic inhibitor concentration and Ki is inhibition constant.
This prediction is possible when the other CYP pathways are not subject to inhibition
[40]. The greatest uncertainty in predicting the magnitude of in vivo drug interactions
resides in the values used for [I] [41]. The most appropriate value would be the
concentration available to the enzyme in the liver, but this value cannot be determined
in vivo. Possible representative in vivo concentrations that could be used include the
free or total systemic concentrations or the free or total hepatic inlet concentrations
estimated to occur during the absorption phase after oral administration [41].
Investigations on the use of the maximum hepatic inhibitor concentration at the inlet to
the liver ([I]in) have been performed. Calculation of this parameter (Equation 2) relies on
information on hepatic blood flow (QH), inhibitor dose (D), fraction absorbed from the
gastrointestinal tract (fa), the absorption rate constant (ka) and the systemic plasma
concentration ([I]av) [38].
12
[I]in = [I]av + ka.fa.D Equation 2 QH
In vivo clinical studies do not usually report ka values, but to avoid false-negative
prediction and obtain the largest [I]in, a maximum ka of 0.1 min-1 has been suggested to
be appropriate, assuming the gastric emptying is the rate limiting step for absorption.
The ka value for each inhibitor could be calculated, using the time to reach maximum
plasma concentration (Tmax) and the elimination rate constant (k) as shown in equation
3.
Tmax = ln(ka / k) Equation 3
(ka – k)
The use of fmCYP and ka values to refine estimates of [I]in has been shown to provide a
useful estimate of [I] and successful predictions [38]. However, it is possible that active
hepatic uptake processes could complicate the use of unbound plasma concentrations
for the inhibitor. This remains a limitation for application of these principles for the
prediction of drug-drug interactions.
The availability of specific probe substrates, human liver tissue and cDNA-
expressed CYP enzymes are valuable tools for the in vitro assessment of the potential
for varied xenobiotics to inhibit CYP isoenzymes [42]. It is appropriate to use tissue from
individual human donors for inhibition studies, if the enzyme activity is sufficiently
present. On the other hand recombinant CYP isoenzymes may be used for investigation
of specific isoenzymes. Human liver microsomes and recombinant CYP isoenzymes are
preferable test systems since they are more readily available than hepatocytes. The
effects of human liver microsomes and recombinant CYP on kinetic measurements are
not confounded by other metabolic processes or cellular uptake that would occur with
the use of hepatocytes [42]. However, the disadvantage of human liver microsomes and
recombinant isoenzymes is that they do not represent the true physiological milieu. In
addition, clinical relevance of the acquired data also has to be established.
13
Table 4. CYP substrate properties, substrates, reactions and interacting drug/food/herb [8,43,44]
CYP Substrates properties
Drug/Substrates Reaction Interacting drug /food/herb*,**
CYP1A2 Planar, polyaromatic compounds
Clozapine Phenacetin Olanzapine
Imipramine Propranolol Theophylline
Oxidation O-deethylation Oxidation
Demethylation 4-hydroxylation Hydroxylation
Cimetidine inh* Ciprofloxacin inh Fluvoxamine inh
Cruciferous vegetables ind** Cigarette smoke ind
CYP2B6 Cyclophosphamide Efavirenz Nevirapine Bupropion Artemisinin
Methadone Profofol
Hydroxylation Hydroxylation 3-ydroxylation Hydroxylation Unclear
N-demethylation Oxidation
Orphenadrine inh Ticlopidine inh Troleandomycin inh Ketoconazole inh Clopidogrel inh
Phenytoin ind Rifampin ind
CYP2C9 Polar, weakly acidic compounds
Diclofenac Flurbiprofen Ibuprofen Naproxen Phenytoin
Piroxicam Tolbutamide Warfarin
4-Hydroxylation 4-hydroxylation Oxidation O-demethylation 4-hydroxylation
5-hydroxylation Methyl hydroxylation 7-hydroxylation
Amiodarone inh Fluconazole inh Fluoxetine inh Isoniazid inh Ticlopidine inh
Rifampin ind Kava inh
CYP2C19 Polar acidic compounds
Amitriptyline Cyclophosphamide Diazepam Imipramine Omeprazole
Phenytoin
Demethylation Oxidation Demethylation N-demethylation Demethylation
4-hydroxylation
Ticlopidine inh Cimetidine inh Fluconazole inh Fluoxetine inh Carbamazepine ind
Norethindrone ind
CYP2D6 Lipophilic, basic, medium sized, positively charged at neutral pH
Amitriptyline Imipramine Propranolol Codeine Dextromethorphan Desipramine
Bufuralol
Benzyl hydroxylation 2-hydroxylation 4-hydroxylation O-demethylation O-demethylation Aromatic hydroxylation
Benzyl hydroxylation
Quinidine inh Indinavir inh Amiodarone inh Fluoxetine inh Haloperidol inh Codeine inh
Kava inh
CYP2E1 Small molecular weight compounds
Acetaminophen Chlorzoxazone
Dehydrogenation 6-hydroxylation
Isoniazid inh Disulfiram inh Garlic inh
CYP3A4 Lipophilic, large molecular weight compounds
Alprazolam Carbamazepine Testosterone
Cyclosporine Midazolam Simvastatin Triazolam Diazepam
Hydroxylation Epoxidation 6!-hydroxylation
Oxidation Oxidation 1-hydroxylation Hydroxylation N-demethylation
Cimetidine inh Erythromycin inh Intraconazle inh
Carbamazepine ind Rifampin ind Phenytoin ind Grapefruit juice ind SJW ind
*inh, inhibitor; **ind, inducer
14
CYP induction
CYP enzymes are susceptible to induction by structurally diverse xenobiotics, including
drugs, foods and herbal products. Induction occurs when a xenobiotic stimulates the
synthesis of CYP isoenzymes. Mechanisms of CYP induction include Ah-receptor
mediated transcriptional activation, as observed with CYP1A induction by polyaromatic
hydrocarbons. Additionally, nuclear hormone receptors have been identified to be
mediators of CYP2B induction by phenobarbitol (CAR, constitutively active receptor)
and CYP3A induction by rifampicin (PXR) [4]. These mechanisms involve binding of the
xenobiotic to the ligand binding domain of the receptor, which leads to conformational
change in the receptor. This facilitates transportation of the complex to the nucleus by
the respective nuclear translocater and binds to the response elements in the DNA
thereby promoting transcription of the gene.
CYP induction enhances its metabolizing capacity and can affect the efficacy of
medications through increased rate of drug metabolism and hepatic clearance of all
substrates through that specific pathway. This results in sub-therapeutic drug
concentrations. Induction of some CYPs is a risk factor in several cancers, since these
enzymes can convert pro-carcinogens to carcinogens. For example, CYP1A2 activates
heterocyclic amines (including 2-amino-1-methyl-6-phenylimidazol[4,5-b]pyridine
(PhIP)), and polycyclic aromatic hydrocarbons (such as benzo(a)pyrene (BaP)), present
in high temperature cooked foods, especially meat [45]. Thus, drug-food/herb
interactions could result upon administration of drugs, capable of inducing CYP1A2
together with foods or herbal products that can be activated by this enzyme.
Induction potential of drug candidates or compounds is difficult to access pre-
clinically and is often inferred from animal studies, which are not necessarily predictive
for humans [42]. However, in vitro and in vivo approaches for prediction have been
developed. In vitro test systems using cultured primary human hepatocytes,
cryopreserved hepatocytes and liver slices have been employed for studying CYP
induction [42,46,47]. CYP inductive response can be assessed in vitro by measuring the
changes in CYP activity. This is evident in changes of kinetic parameters such as a
decrease in area under the curve (AUC) on a dose response chart [42]. On the other
hand metabolism assays can be performed directly using whole cells. Western blot
15
analyses are also useful for determining gross changes in protein levels. In addition,
changes in mRNA levels can be used to measure changes in the expression of CYP
genes [46]. The successful use of CYP induction work of CYP1A, CYP2C, CYP2B and
CYP3A enzymes is well documented [42,46-48].
In vivo methods for evaluation of CYP induction by drugs, foods and herbal
products are also available. These include treatment of experimental animals with these
xenobiotic over a period and then measuring CYP activity in the liver and mRNA levels
[49].
Cytochrome P450
The cytochrome P450 enzymes (CYPs) are phase I enzymes and constitute a
superfamily of hemeproteins that are expressed in almost all organisms [50]. In
eukaryotes, they are usually bound to the endoplasmic reticulum or inner mitochondrial
membranes. The name cytochrome P450 is derived from the characteristic absorption
spectra of the reduced CO-bound complex formed at 450 nm [51]. They function
primarily as monooxygenases, which catalyze the incorporation of a single atom of
oxygen into a substrate (Figure 2) [52]. The active site of CYP is composed of an
iron(III) protoporphyrin(IX) moiety with a cysteine amino acid from the protein backbone
as an axial ligand to the iron.
In the resting state (1), a water molecule is the second axial ligand bound to the
iron. The cycle is initiated by a substrate binding to the iron(III) P450 active site (1),
followed by the displacement of the axial water molecule. Subsequently the iron(III)
heme (2) is reduced to the iron (II) state (3) by the addition of a single electron by the
cytochrome P450 NADPH-reductase, so that the ferrous iron can bind molecular
oxygen (4). The binding of molecular oxygen leads to the formation of an iron(II)-
superoxide spieces, after which a second electron is transferred from a redox partner,
yielding a negatively charged iron peroxo intermediate (5). This species is protonated to
the hydroperoxo-iron (6), which after the loss of water yields the active oxenoid iron
species (7), also referred to as compound I. The latter oxidizes the substrate, the
product is released and a water molecule binds again to the ferric iron [52]. In the
catalytic cycle, uncoupling may occur, whereby the cycle is started and molecular
16
oxygen reduced without product formation as indicated by arrows in figure 3 (a1, a2,
a3).
Figure 3. Catalytic cycle of CYPs, adapted from [52]. Seven stages are shown in the cycle. RH represents
the substrate, ROH the product and –Fe- the iron porphyrin IX, three uncoupling pathways indicated by
a1-a3 and the two peroxide shunt pathways by b1 and b2.
On the other hand, the catalytic cycle can be short-circuited in the presence of artificial
oxygen delivery agents such as peroxides. Thus the cycle turns immediately from stage
2 to stage 6 or 7, depending on the nature of the oxidant as indicated by arrows in figure
2 (b1 and b2) [53].
CYPs catalyze the biotransformation of several endogenous substrates (such as
bile acids, steroids and cholesterol) as well as xenobiotics (drugs, pollutants and dietary
components). Examples of CYP catalyzed reactions are shown in Figure 4. CYPs
mediate biotransformation such as oxidation reactions including hydroxylation of
aliphatic or aromatic carbon, epoxidation of a double bond, heteroatom oxygenation and
17
N-hydroxylation, heteroatom dealkylation, oxidative group transfer, cleavage of esters
and dehydrogenation [1]. Recently 57 active CYP genes and 58 pseudogenes have
been reported to be present in the human genome [54], belonging to 21 families and 20
subfamilies. CYP enzymes are divided into various subfamilies based on amino acid
homology sequence.
CYP reactions
Heteroatom oxygenation RH ROH
Cleavage of ester
X O
R1
R2
R3
R1
R2
+ HXR3
Heteroatom dealkylation
O
N CH3
OH
N CH3
Epoxidation
O
Reduction
H
O
HCOOHR R +
Hydroxylation
OH
NO2
OH
NO2
OH
Oxidative deamination
CH
2R
NH2
CR
O
H NH3+
Dehalogenation
CR
R
ClH
CR
R
O
Figure 4. Some reactions catalyzed by CYPs.
18
Drug metabolizing CYPs belonging to the subfamilies 1, 2 and 3 are responsible for
about 90% of all phase I dependent metabolism of clinically used drugs [55,56] and
participate in the metabolism of a huge number of xenobiotic chemicals. CYP
isoenzymes in the same family have at least 40% sequence similarity, while those in the
same subfamily have at least 60% sequence similarity. A majority of CYPs are
expressed in human liver endoplasmic recticulum, although they are also expressed in
extra hepatic tissues.
Drug metabolizing CYPs are extensively polymorphic and this property
influences the outcome of drug therapy causing lack of response or adverse drug
reactions [57]. For example, approximately 5 to 10% of Caucasians are poor
metabolizers of CYP2D6, whilst 20% Japanese and Chinese are poor metabolizers of
CYP2C19 [18,58]. Poor metabolism increases the bioavailability of drugs, and
consequently the possibility of toxicity [57]. Thus, increased drug accumulation and
toxicities are more likely to occur in poor metabolizers via drug-drug interactions upon
multiple administrations of drugs. Adverse drug reactions are much more common at
ordinary dosing among poor metabolizers for CYP2D6 [56]. On the other hand, rapid
metabolizers decrease the bioavailability of drugs resulting in therapeutic failure. Hence,
multiple drug co-administrations could result in further decrease of bioavailability if
induction of a drug metabolizing enzyme occurs.
Human CYP isoforms
Human CYP isoforms of particular importance for drug metabolism are CYP2C9,
CYP2C19, CYP2D6 and CYP3A4, whilst CYP1A1, CYP1A2, CYP1B1, CYP2E1, and
CYP3A4 are the most important isoforms responsible for metabolic activation of
procarcinogens [59]. Large inter-individual variability in the expression has been
observed with CYPs. The average relative abundance of hepatic CYPs is as follows,
CYP1A2 (13%), CYP2A6 (4%), CYP2B6 (<1%), CYP2C (20%), CYP2D6 (2%), CYP2E1
(7%) and CYP3A4 (30%) [60].
CYP1A2 is involved in the metabolism of various endogenous substrates, such
as melatonin and estrogens [2], and in the activation of procarcinogens, such as
heterocyclic amines, arylamines and aflatoxin B1 [61]. Substrates and inhibitors of
19
CYP1A2 are usually planar small-volume molecules that are neutral or weakly basic.
The binding pocket of CYP1A2 enzyme comprises mostly of hydrophobic and aromatic
amino acids with polar amino acids for hydrogen bonding located near the heme centre
[62]. Drug-drug/food/herb interactions at the level of CYP1A2 may occur as a result of
concomitant administration of drugs metabolized by this enzyme and another drug, food
or herbal product capable of modulating the activity of the enzyme (Tables 2, 3 and 4).
Clinically significant increases in levels of CYP1A2 drug substrate theophylline often
occur after the addition of one of the known inhibitors of this enzyme [Table 4]. Such
interaction may also occur upon intake of foods (such as grapefruit juice) and herbal
products (such as kava extract) that inhibit CYP1A2.
CYP2A6 is the major enzyme catalyzing the oxidative metabolism of nicotine and
cotinine, as well as the metabolism of pharmaceuticals (eg. fadrozole, tegafur, SM-
12502), nitrosamines and a number of coumarin-type alkaloids [63,64]. The CYP2A6
structure shows an enzyme that is well adapted for the oxidation of small planar
substrates, such as coumarin, that fit within the narrow, hydrophobic active site cavity of
the enzyme. CYP2A6 may be inducible by anti-epileptic drugs and it is decreased in
alcohol-induced severe cirrhosis [63]. CYP2A6 does not seem to have an extensive role
in human drug metabolism [65] and genetic variation affecting CYP2A6 activity is not
generally associated with adverse effects on drug clearance [66], suggesting that
CYP2A6 inhibition is unlikely to alter the metabolism of other drugs and be involved in
drug-drug/food/herb interactions.
Earlier studies indicated that CYP2B6 levels were only approximately 0.2% of the
total P450 content in human liver microsomes [55,67]. However, later studies have
demonstrated a greater frequency of detection and a higher percentage of CYP2B6
relative to the total P450 content [68]. This enzyme has proven to be increasingly
important for drug metabolism in the last few years 69]. The substrates of CYP2B6 are
usually non-planar molecules, neutral or weakly basic, fairly lipophilic with one or two
hydrogen-bond acceptors [70]. There are a number of important drugs metabolized by
CYP2B6, including bupropion, efavirenz, cyclophosphamide, ifosamide, pethidine,
artemisinin, propofol, ketamine and selegiline. Drug–drug interactions resulting from
inhibition or induction of CYP2B6 can have serious consequences in the case of
20
substrate drugs with a narrow therapeutic index, such as cyclophosphamide [71].
Known inhibitors of CYP2B6 include 2-phenyl-2-(1-piperidinyl)propane (PPP),
ticlopidine, and clopidogrel [ 72]. Potent inhibition of CYP2B6 activity was observed in
vitro with a dietary supplement and herbal cold remedy, as well as its individual
components lonicera, isatis root, and schizonepeta, respectively [73]. Such herbal
remedies and food components have the potential to cause drug-food/herb interactions
when taken with CYP2B6 substrate drugs, due to their inhibitory properties.
The CYP2C subfamily is one of the largest and most diverse in mammals, and
the human forms show one of the widest substrate ranges compared to other
subfamilies. This subfamily consists of four members in humans, namely CYP2C8,
CYP2C9, CYP2C18 and CYP2C19 [74]. It has been suggested that CYP2C8, CYP2C9
and CYP2C19 are expressed as functional enzymes in the human small intestine [75].
In addition, CYP2C genes are independently regulated in the human intestinal and liver.
Although, overall, the expression and activity of CYP2C enzymes is lower in the gut
than in the liver, the surface area of the proximal small intestine is large and intestinal
CYP2C9 and CYP2C19 may well contribute to the first-pass metabolism of their
substrate drugs [75]. The CYP2C enzymes are genetically polymorphic and share
significant sequence identity (approximately 70%), but have differences in their
localization and substrate profile [76]. Of them, CYP2C9 is the principal drug
metabolizing enzyme in the liver, and metabolizes many clinically important drugs
[72,76]. Substrates of CYP2C9 are generally weakly acidic with one or two hydrogen
bond acceptors. CYP2C9 is particularly interesting due to its implication in adverse drug
reactions as a result of polymorphism and drug-drug interactions due the narrow
therapeutic window of several substrates eg. (S)-warfarin, phenytoin and tolbutamide
[77,78]. Drugs, food components or herbal products that inhibit CYP2C9 such as
amiodarone, fluconazole, fluoxetine, curcumin, black tea extract and kava extract
[43,44,79,80] taken with CYP2C9 drug substrates, could result in drug-drug/food/herb
interactions of clinical relevance.
CYP2D6 is a highly polymorphic enzyme, therefore large inter-individual
differences exist in its activity [57]. Despite representing only 2% of the total human
hepatic CYPs, CYP2D6 plays an important role in the oxidation of xenobiotics, and is
21
involved in the metabolism of about 30% of the currently marketed drugs including
antidepressants, neuroleptics, beta-blockers, opioids and anti-arythmics [81,82].
Common characteristics of substrate and inhibitors of CYP2D6 include a flat
hydrophobic region, hydrogen-bonding properties and basic amines [83,84]. Drug-
drug/food/herb interaction could result upon modulation of the activity of CYP2D6 by a
drug, food or herbal product in the presence of other drugs metabolized by this enzyme.
One of the most potent known inhibitors of CYP2D6 is quinidine and it is used for drug-
drug interaction screening assays [84]. Other known inhibitors of CYP2D6 are
Haloperidol, paroxetine, indinavir, codeine, and extracts of Catharanthus roseus and
Artemisia vulgaris [43,44,85].
CYP2E1 is responsible for the biotransformation of a large number of low
molecular weight compounds with hydrogen bond-forming groups [69]. It is also
involved in the disposition of a many chemicals and xenobiotics, the most important of
these substrates being ethanol. Chronic alcohol consumption can induce CYP2E1. The
disadvantage of metabolism by this enzyme is that metabolites produced are usually
hepatotoxic. Substrates of CYP2E1 include acetaminophen, chlorzoxazone and several
inhalation anesthetics [86]. There are few known inhibitors of CYP2E1 and these
include disulfiram, isoniazid and garlic [8,43]. Concomittant intake of CYP2E1 drug
substrates with other drugs, food and herbal products that inhibit this enzyme could
result in harmful drug-drug/food/herb interactions.
Among the human CYP enzymes, CYP3A4 is considered the most versatile and
the most abundant in both liver and small intestine and hence is the most important
CYP [87]. It is responsible for the metabolism of half of the drugs currently on the
market [88]. Thus the inhibition of CYP3A4 by drugs often causes unfavourable and
long-lasting drug-drug interactions with the potential for morbidity and mortality.
CYP3A4 metabolizes structurally diverse substrates with a wide range of molecular
size, shape, enzyme affinity and turnover number [89]. Unusual kinetic interactions
involving CYP3A4 have been observed due to the binding of multiple substrates within
the active site of the enzyme [90]. CYP3A4 appears to be a key enzyme in food-drug
and herb-drug interactions. For example, interactions of grapefruit juice with felodipine
or cyclosporine, red wine with cyclosporine, and St John's wort with various medicines
22
including cyclosporine, have been reported [43,91,92]. In this thesis, curcumin, a dietary
component has been shown to be a potent inhibitor of CYP3A4 in vitro [80]. The
CYP3A4-related interaction with food components may be related to the high level of
expression of CYP3A4 in the small intestine, as well as its broad substrate specificity.
If potential drug interactions are to be predicted, it is essential that the ability of
food components and herbal products to interfere with drug-metabolizing enzyme
systems is fully established.
Glutathione S-transferases
Drugs, foods and herbal products could also interfere with the activities of Glutathione
S-transferases (GSTs). GSTs are principal phase II biotransformation enzymes
belonging to a superfamily of multifunctional proteins with fundamental roles in the
metabolism and detoxification of a wide range of exogenous and endogenous
compounds [93]. GSTs have also been implicated in cellular physiology,
pathophysiology and other processes such as drug resistance in cancer chemotherapy
[94]. Moreover, genetic polymorphisms in GSTs have been associated with cancer
susceptibility and increased susceptibility for drug interactions or worse outcomes in
diseases [95]. Their main reaction is to catalyze the conjugation of the tripeptide
glutathione (GSH: !-Glu-Cys-Gly), to electrophilic compounds (e.g. Figure 4), to form
more soluble and usually non-toxic peptide derivatives, to be excreted by phase III
enzymes [96]. However, some xenobiotics are converted to reactive intermediates by
the GST-catalyzed reactions. Figure 5 shows a scheme of an overview of enzymatic
detoxification by GST [97]. Almost all soluble GSTs are active as dimers, of either
identical (homodimers) of different (heterodimers) subunits (23-30 kDa), each of them
being encoded by independent genes. Each GST-subunit possesses two ligand-binding
sites: a very specific glutathione-binding site (G-site) and a hydrophobic substrate-
binding site (H-site). Sequence identity within a class is typically >40% while interclass
identities are significantly lower, usually <25% in mammals [93]. Human GSTs are
comprised of at least seven distinct classes of soluble enzymes including Alpha (A), Mu
(M), Pi (P), Theta (T), Omega (O), Sigma (S) and Zeta (Z) [93,98]. GSTs are
differentially expressed both quantitatively and qualitatively in different tissues.
23
N
O
CH3
O
N
OH
SH
CH3
O
+ GSH
Figure 4. Examples of GST-mediated conjugation reactions between electrophiles and GSH. Conjugation
of GSH with NAPQI (N-acetyl-p-benzoquinone imine), oxidized acetaminophen (APAP) to form 3’-GS-
APAP and conjugation of GSH with ETA (ethacrynic acid) to form ETASG (ethacrynic acid glutathione
conjugate).
Factors such as age, disease and exposure to inducing concentrations and
inhibiting xenobiotics also result in changes in enzyme levels and activities. Kinetic
properties, such as substrate specificities and inhibitor sensitivities can, to some extent
be used as a tool to distinguish between different GST isoenzymes [90]. Chlorinated
nitrobenzenes (1-chloro-2,4-dinitrobenzene, CDNB; 1,2-dichloro-4-nitrobenzene,
DCNB) have long served as standard substrates for nearly all GSTs [99]. The theta
class GSTs, however do not catalyze these reactions. Several GST-isoforms belong to
the Alpha and Mu classes, whilst the Pi class originally contained only one isoform [93].
Alpha, Mu and Pi classes
Alpha class GSTs (GSTA), are highly expressed in liver, constituting 80% of the total
GST protein expressed [100]. The enzyme is also expressed in the plasma, kidney and
testis [99]. Of the five human GSTA enzymes four have been fully characterized, and
these include GSTA1-1, GSTA2-2, GSTA3-3 and GSTA4-4. In hepatic cytosol, they
occur principally as GSTA1-1 and GSTA2-2 homodimers or as GSTA1-2 heterodimers
[101].
O
Cl
CH2
O
OH
ClO
+ GSH
ETA ETASG
O
Cl
O
OH
ClO
CH2
SGO
Cl
CH2
O
OH
ClO
+ GSH
ETA ETASG
O
Cl
O
OH
ClO
CH2
SG
NAPQI 3’-GS-APAP
24
Figure 5. Overview of enzymatic detoxification
A bioactivation-detoxification pathway in the case of benzo(a)pyrene is illustrated. The xenobiotic diffuses
freely across the plasma membrane into the cell, where it becomes a substrate for CYP1A1, resulting in
the formation of an epoxide. This in turn becomes a substrate for epoxide hydratase. The diol product is a
substrate again for CYP3A4 to form a carcinogenic and mutagenic diol-epoxide derivative. Both enzymes
are microsomal enzymes. GSTs are phase II enzymes which catalyze conjugation of the electrophilic
dihydrodiol epoxide to GSH. The GSH-conjugate is too hydrophilic to diffuse freely from the cell, and can
be pumped out actively by a transmembrane GS-X transporter. This results in the excretion of the GSH-
conjugate from the cell. Ultimately this conjugate is usually metabolized by Phase III enzymes and
excreted in urine as mercapturic acid (adapted from ref. 97).
Similarly five human Mu class GST (GSTM) enzymes have been identified, and these
include, GSTM1-1, GSTM2-2, GSTM3-3, GSTM4-4 and GSTM5-5 [98]. These enzymes
are expressed to varying extents in different tissues depending on the specific subfamily
[99]. For example, human GSTM1 is expressed at highest concentrations in the liver in
individuals carrying 1 or 2 functional alleles while GSTM2 is expressed in highest
concentration in the brain and hardly in the liver. A variety of GSTs are expressed in the
testis, but GSTM3 is expressed almost uniquely in this tissue [102]. GSTM1 deletion
Cell
OH
OH
HO
GS
O
OH
HO
OH
HO
O
OH
OH
HO
GS
Phase I II of
detoxification
Phase I I of
detoxification
Phase I of
detoxification
GSH
GST
O2CYP1A1
H2OEpoxide
hydratase
CYP3A4 O2
Transmembrane
transporter
Cell
OH
OH
HO
GS
O
OH
HO
OH
HO
O
OH
OH
HO
GS
Phase I II of
detoxification
Phase I I of
detoxification
Phase I of
detoxification
GSH
GST
O2CYP1A1
H2OEpoxide
hydratase
CYP3A4 O2
OH
OH
HO
GS
O
OH
HO
OH
HO
O
OH
OH
HO
GS
OH
OH
HO
GS
O
OH
HO
OH
HO
O
OH
OH
HO
GS
O
OH
HO
OH
HO
O
OH
OH
HO
GS
Phase I II of
detoxification
Phase I I of
detoxification
Phase I of
detoxification
GSH
GST
O2CYP1A1
H2OEpoxide
hydratase
CYP3A4 O2
Transmembrane
transporter
25
frequencies range from 42% to 60% in Caucasians [103]. Some initial studies suggest
that the GSTM1 null genotype confers an increased risk of lung cancer although this
has not been corroborated in subsequent reports [104,105].
The Pi class GSTs (GSTP] appears to be the most widely distributed GST
isoenzyme. It is overexpressed in cancer cells, hence it is regarded as a prognostic
factor in cancer treatment [99]. Regulation of GSTP is of particular interest, in cancer
chemotherapy. Insignificant amounts of GSTP are expressed in the liver and it is the
only class of GSTs expressed in erythrocytes [106]. Unlike GSTA and GSTM, GSTP1-1
is the only isoform recognized in the GSTP class [90].
Although inhibition of GSTs have been considered beneficial in cancer
chemotherapy, in normal cells it may result in harmful consequences such as oxidative
stress and inhibition of important synthesis and signaling pathways [93,107]. The
therapeutic agents, adriamycin, 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU), busulfan,
carmustine, chlorambucil, cisplatin, cyclophosphamide, melphalan, mitozantrone and
thiotepa are detoxified by GSTs [93]. One of the major heterocyclic amines found in
cooked food is 2-amino-1-methyl-6-phenylimidazol [4,5-b]pyridine (PhIP) (metabolized
by CYP1A2), and GSTs have been shown to detoxify the activated metabolite N-
acetoxy-PhIP. Inhibition of the detoxification of drugs and activated metabolites by
drugs, food components and herbal products, could result in toxicities. Drugs, food
components and herbal products shown to inhibit GSTs include ethacrynic acid,
disulfiram, curcumin, ellagic acid, Thonningia sanguinea, Phyllanthus amarus (this
thesis, chapter 6) [108,109,110]. Understanding of the inhibitory profile of drugs, foods
and herbal remedies towards GSTs is needful to avoid toxicities.
Plants and plant derivatives
Herbal remedies remain one of the forms of therapies used by a large part of the world’s
population [111]. Moreover, a wide range of conventional drugs have originally been
derived from plants. An analysis of the origin of the drugs developed between 1981 and
2002 showed that natural products or natural product-derived drugs comprised 28% of
all new chemical entities (NCEs) launched onto the market [112]. In addition, 24% of
these NCEs were synthetic or natural mimic compounds, based on pharmacophores
26
related to natural products [113]. This data suggests that natural products are important
sources for new drugs and drug candidates and are also useful lead compounds for
further optimization during drug development. For examples the drugs morphine,
codeine, noscapine, and papaverine isolated from P. somniferum (refer to Figure 6)
were developed as single chemical drugs and are still clinically used [113].
O N
OH
OH
O N
OH
O
N
O
O
O
O
O
O
O
O
O
Figure 6. Chemical structures of natural compounds developed as drugs from plant sources
While many of the plant products may be benign in nature, some of these have the
potential to cause harmful effects resulting from herb-drug interactions in humans. The
consequences of herb-drug interactions can be a) beneficial effects, such as cancer
prevention, b) undesirable effects, such as pharmacokinetic interactions with co-
administered drugs, c) harmful effects, such as organ toxicity or carcinogenesis [114].
Case reports demonstrate that patients taking a number of non-steroidal anti-
inflammatory drugs including aspirin and ibuprofen experienced severe bleeding after
self-prescribing ginkgo extracts at recommended doses [115,116]. Adverse effects were
particularly severe for aspirin (spontaneous hyphema) and for ibuprofen (comatose
state with an intra-cerebral mass bleeding of which the patient died). Interactions of
herbal preparations with immunosuppressant drugs have also been reported. The
interaction between cyclosporine and St. John’s wort is one of the most serious and
potentially fatal interactions between a herbal remedy and a conventional drug [31].
Morphine Codeine
Papaverine Artemisinin
27
The widely held belief of lay people is that ‘natural’ can be equated to ‘safe’ [111].
For this reason, the use, often self-prescribed is frequently not communicated to the
doctor, with potentially dangerous implications. Recently, interactions of herbal
medicines with synthetic have become of particular interest. Between 1999-2002, more
that 50 papers were published on interactions between St John’s wort and prescribed
drugs only (for a summary see [92,117,118]). The study of herb-drug interactions is
complex, since herbal extracts are multi-component mixtures and the active
constituents are often unknown. Furthermore, in most cases the general information on
the phytochemicals present and their amounts are unavailable. Thus, unspecific effects
such as protein-tannin interaction instead of an enzyme inhibition cannot be excluded.
Similarly, food-drug interactions are complex since foods are multi-component
mixtures. Food drug interactions are often overlooked although a particular food may
modulate the activity of a drug-metabolizing enzyme system, resulting in altered
pharmacokinetics of drugs metabolized by that enzyme system. For example,
consumption of foods such as cruciferous vegetables, grapefruit juice, broccoli,
watercress, red wine and garlic orange juice, have been reported to significantly alter
the pharmacokinetics of several drugs [44]. Investigations have shown that grapefruit
juice is a potent inhibitor of CYP3A4-mediated drug metabolism. It has been
demonstrated to interact with more than 25 drugs including dihydropyridine calcium
channel blockers, cyclosporine, terfenadine and lovastatin, through inhibition of
CYP3A4 [35,119].
The Ghanaian medicinal plants employed in this study, Phyllanthus amarus,
Cassia alata, Cassia siamea, Lactuca taraxicifolia, Momordica charantia, Morinda
lucida and Tridax procumbens are commonly used remedies for various ailments in
Ghana (this thesis, chapter 6). However, there is no information on the potential for
herb-drug interactions in humans and physicians are usually not informed about the use
of such remedies by patients. Hence, adverse effects resulting from herb-drug
interactions are not suspected, investigated and documented. In vitro investigations
have shown the potentials for herb-drug interaction of these medicinal plants at the level
of CYPs and GSTs (in this thesis, chapter 6). Similarly, studies performed on a plant
28
derivative and dietary component, curcumin have indicated it’s potential to cause drug
interactions. The clinical relevance of these findings remains to be established.
Curcumin and curcumin derivatives
Curcumin is a polyphenolic constituent of Curcuma longa L., (turmeric) obtained from
the powdered root of the plant. It is a component of the spice curry, and gives a unique
flavour and colour to food. Curcumin (Figure 7) can also be synthesized by heating
vanillin, acetylacetone and boric anhydride, over a free flame for 30 min, with a yield of
10% [120]. Turmeric has been and still is used in traditional medicine, especially in
South Eastern Asia, where it is used in the treatment of hepatic and biliary disorders,
rheumatism, cough, and diabetic wounds [121]. Interestingly, curcumin derived from
turmeric possesses several other biological activities, including cancer
chemopreventive, anti-inflammatory, antioxidant, anti-plasmodial and anti-HIV activities
[122-125].
OO
OH
O
OH
OCH3
CH3
Figure 7. Chemical structure of curcumin.
Despite the various therapeutic claims, curcumin certainly has some shortfalls, the most
important includes an extremely low bioavailability which is hampering systemic effects
upon oral administration [126]. Curcumin is also chemically unstable at neutral to basic
pH and exhibits potent inhibition of human GSTs A1-1, A2-2, M1-1, M2-2 and P1-1
[108,127]. However, this instability has been observed to be blocked by antioxidants
such as glutathione, ascorbic acid, N-acetyl-L-cysteine and protein from microsomal or
cytosolic fractions of rat liver [127]. In addition, curcumin is a potent inhibitor of human
CYPs 3A4 and 2C9 [80]. The potential of curcumin to modulate multiple targets may be
a disadvantage due to the possibility of multiple cellular pathways influencing each other
with the likely consequence of undesirable effects. The low bioavailability of curcumin is
a major problem with regards to its therapeutic application. However, various studies
are being performed to overcome this disadvantage. These include the co-
29
administration of curcumin with piperine, the formulation of curcumin with
phoshatidylcholine or polyethylene glycol, and use of curcumin nanoparticles, liposomal
curcumin and structural analogues of curcumin [128-133].
On the basis of the biological activity and ease of synthesis, curcumin appears
to be useful as a lead compound for the design of derivatives with better biological and
pharmacokinetic properties. For this reason investigators have synthesized various
curcumin analogues [134-136]. Sardjiman et al [134] omitted the active methylene
group and one carbonyl group of curcumin leading to the production of three series of
analogues, which are more stable compounds while retaining antioxidant and anti-
inflammatory properties. Thus, a series of 1,5-diphenyl-1,4-pentadiene-3-one (C),
together with cyclopentanone (B) and cyclohexanone (A) analogues were synthesized
(Figure 8). Some of these curcumin analogues showed potent inhibitory activities
towards human CYPs essential for drug metabolism, demonstrating potential for drug-
drug interactions [137]. Potent GST inhibition has also been observed with some of
these compounds (in this thesis, Chapter 5).
33
1
2
O
R
R
R R
R
R
2
1
1
33
R1
R
R
O
R
R
R
22
1
33
R1
R
R
O
R
R
R
22
Figure 8. Synthetic derivatives of curcumin, including 2,6-dibenzylidenecyclohexanones (A), 2,5-
dibenzylidenecyclopentanones (B) and 1,5-diphenyl-1,4-pentadiene-3-ones (C). These curcumin
analogues are derivatives of benzylidine having either electron-withdrawing, electron-donating or steric
groups.
Quantitative structure-activity relationship (QSAR)
QSAR approaches attempt to identify and quantify the physicochemical properties of a
drug, and to relate these properties to the biological activities of a drug. In drug
A B
C
30
discovery, QSARs are widely used to identify ligands with high affinity for a given
macromolecular target. More recently, QSAR approaches have also been extended to
rationalize and predict absorption, distribution, metabolism, elimination, toxicity
(ADMET) properties or the oral bioavailability of compounds [138,139]. The basis for
QSAR modeling comes from the concept of linear free-energy relationships, where
variations in the binding behaviour of small molecules to biological targets may be
quantitatively attributed to changes in their structure [140]. In the absence of large
structural changes, useful correlations may be generated between molecular properties
and biological activities and sometimes these may be used to predict the activities of
unknown compounds. For true QSAR relationships, equations can be constructed for
the rationalization of known and the prediction of biological activities of unknown
compounds. If a compound does not fit the equation, it implies that other molecular
features are likely also important, and this provides a lead for further development.
QSAR studies are based on two fundamental assumptions: (i) the affinity data refer to
the same target and (ii) all ligands bind in the same fashion to the receptor [141].
Many physical, structural and chemical properties have been studied by QSAR
approaches, but traditionally the most commonly studied properties emphasize
hydrophobic, electronic and steric effects. Various additional dimensions have more
recently also been employed in classification of QSAR (Table 5).
Table 5. Classification of more recent QSAR approaches based on their dimensionality
Dimension Method
1D Affinity is correlated with global molecular properties of ligands, that is one value per property and ligand (pKa, log P, etc.)
2D Affinity is correlated with structural patterns (connectivity, 2D pharmacophore, etc.) without
consideration of an explicit 3D representation of these properties
3D Affinity is correlated with the three-dimensional structure of the ligands
4D Ligands are represented as an ensemble of configurations
5D As 4D-QSAR + explicit representation of different induced-fit models
6D As 5D-QSAR + representation of different solvation scenarios
Table partly adapted [142].
31
The Molecular Operating Environment (MOE) software (Chemical Computing Group
Inc. Montreal) is a useful recent source to obtain structural descriptors. The QSAR
approach could play a major role in drug development in predicting compounds with
potentials for drug interactions.
Scope and Objectives
As outlined in this introduction, drug-drug/food/herb interactions mediated by the
inhibition of CYPs and GSTs, two of the most important biotransformation enzymes, are
a major cause of adverse drug reactions. Thus, evaluation of drugs and drug candidates
for CYP, GST inhibitory potential during drug discovery and development has been
considered cost-effective for predicting potentially relevant clinical drug-drug
interactions. Such drug-drug interactions are one of the major causes of attritions in
drug development [17]. Also, plant products such as food and herbal medicines may
have the potential to inhibit CYPs and GSTs and thus result in significant undesirable
effects upon co-ingestion with prescribed drugs or other substances requiring a
particular CYP for metabolism. Several in vitro test systems including well-established
isoenzyme selective assays for drug-drug interaction (i.e. enzyme inhibition or
induction) properties based on microsomal (CYP), cytosolic preparations and
hepatocytes are available [143-145]. The CYPs known to be important for the
metabolism of several drugs currently on the market include CYP3A4, CYP2D6,
CYP2C9, CYP1A2 and CYP2B6 [144].
At the commencement of these studies in 2004, the potential of curcumin, its
synthetic analogues and the seven Ghanaian medicinal plants, to cause drug-drug
interactions at the level of major human drug-metabolizing CYPs, had not yet been
evaluated. However, earlier studies had shown that curcumin is a potent inhibitor of rat
liver microsomal CYP1A1/1A2 and CYP2B1/2B2 [127]. In addition Phyllanthus amarus,
one of the medicinal plants employed, had previously been shown to be a potent
inhibitor of rat liver CYP1A1/1A2, but animal studies often lack predictive power for the
human model [146]. Earlier studies had also shown that some of the present synthetic
analogues had stronger anti-inflammatory and antioxidant properties than curcumin
[134]. The focus of this study was to evaluate the CYP- and GST-mediated drug-drug
32
interaction potential of plant derived components. In this regard curcumin was
considered a clinically relevant model compound for our purpose. Curcumin, a
frequently used food component isolated from Curcuma longa, possesses interesting
pharmacological properties such as the potential for anti-cancer, anti-oxidant and anti-
inflammatory activities. Thus, investigations are being carried out on its potential as a
chemopreventive agent. Synthetic curcumin analogues were also tested, and QSARs
performed to identify analogues with less potential to cause drug interactions compared
to curcumin and guide future design of curcumin analogues.
Major human recombinant CYP have been used in this work to determine the
CYP inhibitory potentials of curcumin, curcumin decomposition products, thirty-four
structural analogues of curcumin and seven Ghanaian medicinal plants. Secondly, due
to the major role of GSTs in detoxification, oxidative stress and synthesis of important
biomolecules and signal transduction, inhibition of these enzymes by drug candidates,
herbal products or foods could be harmful to normal cells. Thus, in these studies we
have employed the GST isoenzymes GSTA1-1, GSTM1-1 and GSTP1-1 as well as rat
and human liver cytosol, to investigate the inhibitory potencies of the curcumin
analogues and medicinal plants on these enzyme activities.
Aim of this thesis
The primary aim of the investigations described in this thesis is to determine the CYP-
and the GST mediated drug-drug/food/herb interactions via inhibitory potential of plant -
derived components and their derivatives, which includes curcumin, its analogues and
medicinal plants of Ghana. This work particularly focuses on three lines of research:
(i) Drug-drug interaction potential mediated by curcumin and its derivatives,
evaluated as inhibition of curcumin, its decomposition products and synthetic
analogues towards important human CYPs and the mechanisms of inhibition
by curcumin. Structure-activity relationships of analogues were also analyzed.
(ii) The potential of curcumin as well as thirty-four synthetic curcumin analogues
to inhibit human and rat GSTs, and related structure-activity relationships
were analyzed. Earlier studies have shown that curcumin is a potent inhibitor
of rat cytosolic GSTs [127].
33
(iii) The drug-herb interactions potential of seven important Ghanaian medicinal
herb extracts towards major CYPs were investigated. The inhibitory potentials
of these extracts towards human and rat GSTs were also assessed.
Outline of this thesis
Three lines of research (i-iii) are addressed in the following chapters. In Chapter 1, a
general introduction of the background, aim and the scope of the research described in
this thesis, is given. In Chapter 2 research findings on the pharmacokinetics,
metabolism and drug-drug/-food interactions potentials of curcumin are reviewed. As
reviewed, due to its low bioavailability, curcumin lacks the potential for adequate
therapy in organs and tissues distant from the intestines, however, several new
formulations of curcumin appear to have better potential in this regard. In chapter 3, the
inhibitory effects of curcumin and its decomposition products on human CYP1A2,
CYP2B6, CYP2C9, CYP2D6 and CYP3A4 were assessed, to evaluate potential drug-
drug interaction properties at the level of CYPs. Mechanisms of inhibition of the
enzymes by curcumin were also investigated to obtain insight into the mode of action,
and the possible clinical implications. In chapter 4, thirty-three curcumin analogues
belonging to three series of dibenzylidene derivatives were investigated for their
inhibitory potentials towards the major human drug metabolizing CYPs for reference
purposes compared to curcumin itself. Subsequently, structure-activity relationships
were also analyzed to guide future designing of curcumin analogues with less
susceptibility to CYP inhibition.
The GST inhibitory potential of the thirty-four curcumin analogues was
determined in Chapter 5, and structure-activity relationships were analyzed as well.
These results may be useful in designing and synthesizing curcumin analogues with
less susceptibility to GST inhibition. In Chapter 6, seven important medicinal herbs from
Ghana were assessed for their inhibitory potentials towards human CYPs and GSTs,
since most of these herbal extracts are often consumed together with prescribed drugs.
Finally, Chapter 7 summarizes the results and conclusions of all investigations in this
thesis.
34
References 1. Parkinson, A., 2001. Biotransformation of xenobiotic. In Casarett and Doull’s Toxicology. The basic science of poisons. Ed. Klaassen C.D., McGraw-Hill Medical Publishing Division, New York, USA, 133-224. 2. Ma, X., Idle, J.R., Krausz, K.W., Gonzalez, F.J., 2005. Metabolism of melatonin by human cytochromes p450. Drug Metab Dispos 33:489–494. 3. Hayes, J.D., Flanagan, J.U., Jowsey, I.R., 2005. Glutathione transferases. Ann Rev Pharmacol Toxicol 45:51-88. 4. Xu, C., Li, C.Y., Kong, A.N., 2005. Induction of phase I, II and III drug metabolism/transport by xenobiotics. Arch Pharm Res 28:249-268. 5. Honig, P.K., Wortham, D.C., Zamani, K., Conner, D., Mullin, J.C., Cantilena, L.R., 1993. Terfenadine ketoconazole interaction. J Am Med Assoc 269:1513-1518. 6. Brandin, H., Myrberg, O., Rundlof, T., Arvidsson, A-K., Brenning, G., 2007. Adverse effects of artificial grapefruit seed extract products in patients on wafarin therapy. Eur J Clin Pharmacol 63:565- 570. 7. Sorensen, J.M., 2002. Herb-drug, food-drug, nutrient-drug and drug-drug interactions: mechanisms involved and their medical applications. J Altern Complement Med 8:293-308. 8. Delgoda, R., Westlake, A.C.G., 2004. Herbal interactions involving cytochrome P450 enzymes. Toxicol Rev 23:239-249. 9. Arayne, M.S., Sultana, N., Bibi, Z., 2005. Grape fruit juice-drug interactions. Pak J Pharm Sci 18:45-57. 10. Moore, L.B., Goodwin, B., Jones, S.A., Wisely, G.B., Serabjit-Singh, C.J., Willson, T.M., Collins, J.L., Kliewer, S.A., 2000. St. John's wort induces hepatic drug metabolism through activation of the pregnane X receptor. Proc Natl Acad Sci 97:7500-7502. 11. Dürr, D., Stieger, B., Kullak-Ublick, G.A., Rentsch, K.M., Steinert, H.C., Meier, P.J., Fattinger, K., 2000. St John's Wort induces intestinal P-glycoprotein/MDR1 and intestinal and hepatic CYP3A4. Clin Pharmacol Ther 68:598-604. 12. Li, M., Anderson, G.D., Phillips, B.R., Kong, W., Shen, D.D., Wang, J., 2006. Interactions of
amoxicillin and cefaclor with human renal organic anion and peptide transporters. Drug Metab Dispos 34:547-555.
13. Abernethy, D.R., Flockhart, D.A., 2000. Molecular Basis of Cardiovascular Drug Metabolism: Implications for Predicting Clinically Important Drug Interactions. Circulation 101:1749.
14. Zhou, S., Lim, L.Y., Chowbay, B., 2004. Herbal modulation of P-glycoprotein. Drug Metab Rev 36:57-104.
15. Tirona, R.G., Lee, W., Leak, B.F., 2003. The orphan nuclear receptor HNF-4 alpha determines PXR- and CAR-mediated xenobiotic induction of CYP3A4. Nat med 9:220-224. 16. Chen, J., Raymond, K., 2006. Roles of rifampicin in drug-drug interactions: underlying molecular mechanisms involving the nuclear pregnane X receptor. Ann Clin Microbiol Antimicrob 5: 3. 17. Zhang, Z-Y., Wong, Y.N., 2005. Enzyme kinetics for clinically relevant CYP inhibition. Curr Drug Metab 6:241-257. 18. Zafar, A., Sharif, M.D., 2003. Pharmacokinetics, metabolism, and metabolism of atypical antipsy- chotics in special populations. Primary care companion J Clin Psychiatry 5:22-25. 19. Pea, F., Furlanut, M., 2001. Pharmacokinetic aspects of treating infections in the intensive care
unit: focus on drug interactions. Clin Pharmacokinet 40:833-868. 20. Guengerich, F.P., 1997. Role of cytochrome P450 in enzymes in drug-drug interactions. In: Li, A.P.,
editor. Drug-drug interactions. London: Scientific and regulatory perspectives. 7-35. 21. Kronbach, T., Fischer, V., Meyer, U.A., 1988. Cyclosporin metabolism in human liver: identification
of a cytochrome P450 III gene family as the major cyclosporinmetabolizing enzyme explains interactions of cyclosporine with othr drugs. Clin Pharmacol Ther 43:630-635.
22. Kivisto, K.Y., Neuvonen, P.J., Klotz, U., 1994. Inhibition of terfenadine metabolism: pharmacokin-etic and pharmacodynamic consequences. Clin Pharmacokinet 27:1-5.
23. Fontana, R.J., Lown, K.S., Paine, M.F., Fortlage, L., Santella, R.M., Felton, F.S., et al., 1999. Effects of a chargrilled meat diet on expression of CYP3A, CYP1A, and P-glycoprotein levels in healthy volunteers. Gastroenterology 117:89-98.
24. Desta, Z., Soukhova, N., Flockhart, D.A., 2001. Inhibition of cytochrome P450 isoforms by isonia-zid: potent inhibition of CYP2C19 and CYP3A. Antimicrobial Agents Chemother 45:382-392.
35
25. Jeppensen, U., Loft, S., Poulsen, H.E., Brsen, K., 1996. Fluvoxamine-caffeine interaction. Pharmacogenetics 6:3213-3222.
26. Carrillo, J.A., Benitez, J., 2000. Clinically significant pharmacokinetic interactions between dietary caffeine and medications. Clin Pharmacokinet 39:127-153.
27. Neal, J.M., Kunze, K.L., Levy, R.H., O’Reilly, R.A., Trager, W.F., 2003. Kiiv, an in vivo parameter for predicting the magnitude of drug interactions arising from competitive enzyme inhibition. Drug Metab Dispos 31:1043-1048.
28. Di Marco, M.P., Edwards, D.J., Wainer, I.W., Ducharme, M.P., 2002. The effect of grapejuice and serville orange juice on the pharmacokinetics of dextromethorphan: the role of gut CYP3A and P-glycoprotein. Life Sci 71:1149-1160.
29. Guengerich, F.P., 1988. Roles of cytochrome P450 enzymes in chemical carcinogenesis and cancer chemotherapy. Cancer Res 48:2946-2954.
30. Prior, T.I., Chue, P.S., Tibbo, P., Baker, G.B., 1999. Drug metabolism and atypical antipsychotics. Eur Neuropsychopharmacol 9:301-309.
31. Barone, G.W., Gurley, B.J., Ketel, B.L., Lightfoot, M.L., Abul-Ezz, S.R., 2000. Drug interaction between St. John’s Wort and cyclosporine. Ann Pharmacother 34:1013-1016.
32. Kupiec, T., Raj, V., 2005. Fatal seizures due to potential herb-drug interactions with Ginkgo biloba. J Anal Toxicol. 29:755-758.
33. Lahoz, A., Donato, M.T., Montero, S., Castell, J.V., Gomez-Lechon, M.J., 2008. A new in vitro approach for the simultaneous determination of phase I and phase II enzymatic activities of human hepatocyte preparations. Rapid Commun Mass Spectrom 22:240-244.
34. Fontana, E., Dansette, P.M., Poli, S.M., 2005. Cytochrome P450 enzymes mechanism-based inhibitors: common sub-structures and reactivity. Curr Drug Metab 6:413-454.
35. Lilja, J.J., Kivisto, K.T., Neuvonen, P.P., 2000. Duration of effect of grapefruit juice on pharmacokin-etics of the CYP3A4 substrate simvastatin. Clin Pharmacol Ther 68:384-390.
36. Neuvonen, J.P., Cantola, T., Kivisto, K.T., 1998. Simvastatin but not pravastatin is very susceptible to interaction with the CYP3A4 inhibitor itraconazole. Clin Pharmacol Ther 63:332-341.
37. Maurer, T., Fung, H.L., 2000. Comparison of methods for analyzing kinetic data from mechanism-based enzyme inactivation: Application to nitric oxide synthase. AAPS PharmSci 2:E8.
38. Brown, H.S., Ito, K., Aleksandra G., Houston, B., 2005. Prediction of in vivo drug-drug interactions from in vitro data: impact of incorporating parallel pathways of drug elimination and inhibitor absorp-tion rate constant. Br J Clin Pharmacol 60:508-518.
39. Scordo, M.G., Pengo, V., Spina, E., Dahl, M.L., Gusella, M., Padrini, R., 2002. Influence of CYP2C9 and CYP2C19 genetic polymorphisms on warfarin maintenance dose and metabolic clearance. Clin Pharmacol Ther 72:702-710.
40. Ito, K., Hallifax, D., Obach, R.S., Houston, J.B., 2005. Impact of parallel pathways of drug elimination and multiple CYP involvement on drug-drug interactions: CYP2D6 paradigm. Drug Metab Dispos 33:837-844.
41. Obach, R.S., Walsky, R.L., Venlatakrisnan, K., Houston, J.B., Tremaine, L.M., 2005. In vitro cytochrome P450 inhibition data and prediction of drug-drug interactions: Qualitative relationships, quantitative predictions and the rank order approach. Clin Pharmacol Ther 78:582-592.
42. Bjornsson, T.D., Callaghan, J.T., Einolf, H.J., Fischer, V., Gan, L., Grimm, S., et al., 2003. The conduct of in vitro and in vivo drug-drug interaction studies: A PhRMA perspective. J Clin Pharmacol 43:443-469.
43. Goshman, L., Fish, J., Roller, K., 1999. Clinically significant cytochrome P450 drug interactions. J Pharm Soc Winsconsin May/June.
44. Butterweck, V., Derendorf, H., Gaus, W., Nahrstedt, A., Schulz, V., Unger, M., 2004. Pharmacok-inetic herb-drug interactions: Are preventive screenings necessary and appropriate. Plant Med 70:784-791.
45. Cross, A.J., Peters, U., Kirsh, V.A., Andriole, G.L., Reding, D., Hayes, R.B., Sinha, R., 2005. A Prospective study of meat and meat mutagens and prostate Cancer Risk. Cancer Res 65:11779-11784.
46. Worboys, P.D., Carlile, D.J., 2001. Implications and consequences of enzyme induction on preclinical and clinical drug development. Xenobiotica 31:539-556.
36
47. Einolf, H.J., Bedi-Singh, S., Fischer, V., 2002. Precision-cut human liver slices as a model to examine changes in cytochrome P450 expression to predict or preclude drug-drug interactions. The Toxicologist 46:19.
48. Edwards, R.J., Price, R.J., Watts, P.S., Renwick, A.B., Tredger, J.M., Boobis, A.R., et al., 2003. Induction of cytochrome P450 enzymes in cultured precision-cut human liver slices. Drug Metab Dispos 31:282-288.
49. Martignoni, M., de Kanter, R., Grossi, P., Mahnke, A., Saturno, G., Monshouwer, M., 2004. An in vivo and in vitro comparison of CYP induction in rat liver and intestine using slices and quantitative RT-PCR. Chem-Biol Interact 151:1-11.
50. Ortiz de Montellano, P/R., 2005. Cytochromes P450: Structure, mechanism and biochemistry, third ed., Kluwer Academic, New York.
51. Omura, T., Sato, R., 1964. The carbon monoxide binding pigment of liver microsomes. I. solubilization, purification and properties. J Biol Chem 239:2379-2385.
52. Goenhof, A.R., Swart, M., Ehlers, A.W., Lammertsama, K., 2005. Electronic ground states of iron porphyrin and of the first species in the catalytic reaction cycle of cytochrome P450s. J Phys Chem 109:3411-3417.
53. Guengerich, F.P., 2001. Common and uncommon cytochrome P450 reactions related to metabolism and chemical toxicity. Chem Res Toxicol 14:611-650.
54. Nelson, D.R., Zelden, D.C., Hoffman, S.M., Maltais, L.J., Wain, W.H., Nerbert, D.W., 2004. Comparison of cytochrome P450 genes from the mouse and human genomes, including nomenclature recommendations for genes, pseudogenes and alternative-splice variants. Pharmacogenetics 14:1-18.
55. Shimada, T., Yamazaki, H., Mimura, M., Inui, Y., Guengerich, F.P., 1994. Interindividual variations in human liver cytochrome P-450 enzymes involved in the oxidation of drugs, carcinogens and toxic chemicals: studies with liver microsomes of 30 Japanese and 30 Caucasians. J Pharmacol Exp Ther 270:414–423.
56. Ingelman-Sundberg, M., 2004. Pharmacogentics of cytochrome P450 and its applications in drug therapy: the past, present and future. Trends Pharmacol Sci 25:193-200.
57. Ingelman-Sundberg, M., 2005. The human genome project and novel aspects of cytochrome P450 research. Toxicol Appl Pharmacol 207:S52-S56.
58. Nakamura, K., Goto, F., Ray, W. A., McAllister, C. B., Jacqz, E., Wilkinson, G. R., et al., 1985. Interethnic differences in genetic polymorphism of debrisoquin and mephenytoin hydroxylation between Japanese and Caucasian populations. Clin Pharmacol Ther 38:402-408.
59. Nebert, D.W., Russell, D.W., 2002. Clinical importance of cytochromes P450. Lancet 360:11155-1162.
60. Rendic, S., Carlos, F.J.D., 1997. Human cytochrome P450 enzymes: A status report summarizing their reactions, substrates, inducers and inhibitors. Drug Metab Rev 29:413-580.
61. Shimada, T., Guengerich, F.P., 2006. Inhibition of human cytochrome P450 1A1-, 1A2-, and 1B1-mediated activation of procarcinogens to genotoxic metabolites by polycyclic aromatic hydrocarbons. Chem Res Toxicol 19:288-294.
62. Korhonen, L.E., Rahnasto, M., Mahonen, N.J., Wittekindt, C., Poso, A., Juvonen, R.O., et al., 2005. Predictive three-dimensional quantitative structure-activity relationship of cytochrome P450 1A2 inhibitors. J Med Chem 48:3808-3815.
63. Pelkonen, O., Rautio, A., Raunio, H., Pasanen, M., 2000. CYP2A6: a human coumarin 7- hydroxylase. Toxicology 144:139-147.
64. Raunio, H., Rautio, A., Gullsten, H., Pelkonen, O., 2001. Polymorphisms of CYP2A6 and its practical consequences. Br J Clin Pharmacol 52:357-363.
65. Guengerich, F.P.,2005. In Cytochrome P450: Structure, Mechanism, and Biochemistry (ed. Ortiz de Montellano, P.R.) 377-530 (Kluwer Academic/Plenum Publishers, New York, 2005).
66. Yano, J.K., Hsu, M-H., Griffin, K.J., Stout, C.D., Johnson, E.F., 2005. Structures of human microsomal cytochrome P450 2A6 complexed with coumarin and methoxsalen. Nature Structural Mol Biol 12:822-823.
67. Mimura, M., Baba, T., Yamazaki, H., Ohmori, S., Inui, Y., Gonzale, F.J., et al., 1993. Characteriz-ation of cytochrome P-450 2B6 in human liver microsomes. Drug Metab Dispos 21:1048–1056.
68. Stresser, D.M., Kupfer, D., 1999. Monospecific antipeptide antibody to cytochrome P-450 2B6. Drug Metab Dispos 27:517–525.
37
69. Spatzengger, M., Liu, H., Wang, Q., Debarber, A., Koop, D.R., Halpert, J.R., 2003. Analysis of differential substrate slectivities of CYP2B6 and CYP2E1 by site-directed mutagenesis and molecular modeling. J Pharmacol Exp Ther 304:477-487.
70. Lewis, D.F., 2000. On the recognition of mammalian microsomal cytochrome P450 substrates and their characteristics: towards the prediction of human P450 substrate specificity and metabolism. Biochem Pharmacol 60:293–306.
71. Turpeinen, M., Raunio, H., Pelkonen, O., 2006. The functional role of CYP2B6 in human drug metabolism: substrates and inhibitors in vitro, in vivo and in silico. Curr Drug Metab 7:705–714.
72. Walsky, R.L., Obach, R.S., 2007. A comparison of 2-phenyl-2-(1-piperidinyl)propane (ppp), 1,1',1''-phosphinothioylidynetrisaziridine (thioTEPA), clopidogrel, and ticlopidine as selective inactivators of human cytochrome P450 2B6. Drug Metab Dispos 35:2053-2059.
73. Foti, R.S., Wahlstrom, J.L., Wienkers, L.C., 2007. The in Vitro Drug Interaction Potential of Dietary Supplements Containing Multiple Herbal Components. Drug Metab Dispos 35:185-188.
74. Goldstein, J.A., 2001. Clinical relevance of genetic polymorphisms in the human CYP2C subfamily. Br J Clin Pharmacol 52:349-355.
75. Lapple, F., von Richter, O., Fromm, M.F., Richter, T., Thon, K.P., Wisser, H., Griese, E.U., Eichelbaum, M., Kivisto, K.T., 2003. Differential expression and function of CYP2C isoforms in human intestine and liver. Pharmacogentics 13:565-575.
76. Polgar, T., Menyhard, D.K., Keseru, G.M., 2007. Effective virtual screening protocol for CYP2C9 ligands using a screening site constructed from flurbiprofen and S-warfarin pockets. J Comput Aided Mol Des 21:539-548.
77. Pirmohamed, M., Park ,B.K., 2003. Cytochrome P450 enzyme polymorphisms and adverse drug reactions. Toxicology 192:23–32.
78. Rettie, A.E., Jones, J.P., 2005. Clinical and toxicological relevance of CYP2C9: Drug-drug interactions and pharmacogenetics. Ann Rev Pharmacol Toxicol 45:477-494.
79. Gunes, A., Bilir, E., Zengil, H., Babaoglu, M.O., Bozkurt, A., Yasar, U., 2007. Inhibitory effect of valproic acid on cytochrome P450 2C9 activity in epilepsy patients. Basic Clin Pharmacol Toxicol 100:383-386.
80. Appiah-Opong, R., Commandeur, J.N.M., van Vugt-Lussenburg, B., Vermeulen, N.P.E., 2007. Inhibition of human recombinant cytochrome P450s by curcumin and curcumin decomposition products. Toxicology 235:83-91.
81. Zanger, U.M., Raimundo, S., Eichelbaum, M., 2004. Cytochrome P450 2D6: overview and update on pharmacology, genetics, biochemistry. Naunyn Schmiedebergs Arch Pharmacol 369:23-37.
82. Bertilsson, L., Dahl, M.L., Dalen, P., Al-Shurbaji, A., 2002. Molecular genetics of CYP2D6: clinical relevance with focus on psychotropic drugs. Br J Clin Pharmacol 53:111-122.
83. Koymans, L., Vermeulen, N.P., van Acker, S.A., te Koppele, J.M., Heykants, J.J., Lavrijsen, K., et al., 1992. A predictive model for substrates of cytochrome P450-debrisoquine (2D6). Chem Res Toxicol 5:211-219.
84. Hutzler, J.M., Walker, G.S., Wienkers, L.C., 2003. Inhibition of cytochrome P450 2D6: structure-activity studies using a series of quinidine and quinine analogues. Chem Res Toxicol 16:450-459.
85. Usia, T., Iwata, H., Hiratsuka, A., Watabe, T., Kadota, S., Tezuka, Y., 2006. CYP3A4 and CYP2D6 inhibitory activities of Indonesian medicinal plants. Phytomedicine 13:67-73.
86. Klotz, U., Ammon, E., 1998. Chemical and toxicological consequences of the inductive potential of ethanol. Eur J Clin Pharmacol 54:7-12.
87. Shu-Feng, Z., Changli, C., Xue-Qing, Y., Chunguang, L., Guangji, W., 2007. Clinically Important Drug Interactions Potentially Involving Mechanism-based Inhibition of Cytochrome P450 3A4 and the Role of Therapeutic Drug Monitoring. Ther Drug Monit 29:687-710.
88. Zhou, S.F., Xue, C.C., Yu, X.Q., Wang, G., 2007. Clinically important drug interactions potentially involving mechanism-based inhibition of cytochrome P450 3A4 and the role of therapeutic drug monitoring. Ther Drug Monit 29:687-710.
89. Lu, P., Lin, Y., Rodrigues, A.D., Rushmore, T.H., Baillie, T.A., Shou, M., 2001. Testosterone, 7- benzyloxyquinoline, and 7-benzyloxy-4-trifluoromethyl-coumarin bind to different domains within the active site of cytochrome P450 3A4. Drug Metab Dispos 29:1473-1479.
90. Shou, M., Lin, Y., Lu, P., Tang, C., Mei, Q., Cui, D., et al., 2001. Enzyme kinetics of cytochrome P450-mediated reactions. Curr Drug Metab 2:17-36.
38
91. Tsunoda, S.M., Harris, R.Z., Christians, U., Velez, R.L., Freeman, R.B., Benet, L.Z., Warshaw, A., 2001. Red wine decreases cyclosporine bioavailability. Clin Pharmacol Ther 70:462-467.
92. Hammerness, P., Basch, E., Ulbricht, C., barrette, E.P., Foppa, I., Bent, S., et al., 2003. St John’s wort: a systematic review of adverse effects and drug interactions for the consultation psychiatrist. Psychosomatics 44:271-282.
93. Frova, C., 2006. Glutathione transferases in the genomics era: New insights and perspectives. Biomol Eng 23:149-169.
94. Lo, H-W., Ali-Osman, F., 2007. Genetic polymorphism and function of glutathione S-transferases in tumor drug resistance. Curr Opinion Pharmacol 7:367-374.
95. Hayes, J.D., Strange, R.C., 2000. Glutathione S-transferase polymorphisms and their biological consequences. Pharmacology 61:154-166.
96. van Bladeren, P.J., 2000. Glutathione conjugation as a bioactivation reaction. Chem Biol Interact 129:61-76.
97. Sheehan, D., Meade, G., Foley, V.M., Dowd, C.A., 2001. Structure, function and evolution of glutathione transferases: implications for classification of non-mammalian members of an ancient enzyme superfamily. Biochem J 360:1-16.
98. Mannervik, B., Board, P.G., Hayes, J.D., Listowsky, I., Pearson, W.R., 2005. Nomenclature for mammalian soluble glutathione transferase. Methods Enzymol 401:1-8.
99. Eaton, D.L., Bammler, T.K., 1999. Concise review of the glutathione S-transferases and their significance to toxicology. Toxicol Sci 49:156-164.
100. van Ommen, B., Bogaards, J.J.P., Peters, W.H.M., Blaauboer, B., van Bladeren, P.J., 1990. Quantification of human hepatic glutathione S-transferase. Biochem J 269:609-613.
101. Dajani, L.K., Paus, E., Warren, D.J., 2001. Development of a rapid and sensitive immunofluorom- etric assay for glutathione S-transferase A. Clin Chem 47:867-873.
102. Listowsky, I., Rowe, J.D., Patskosky, Y.V., Tchaikovskaya, T., Shintani, N., Novikova,E., Nieves, E., 1998. Human testicular glutathione S-transferases: insights into tissue-specific expression of the diverse subunit classes. Chem Biol Interact 111/112:103-112.
103. Garte, S., Gaspari, L., Alexandrie, A.K., Ambrosone, C., Autrup, H., Autrup, J.L., et al., 2001. Metabolic gene polymorphism frequencies in control populations. Cancer Epidemiol Biomarkers Prev 10:1239-1248.
104. Taioli, E., Gaspari, L., Benhamou, S., Boffetta, P., Brockmoller, J., Butkiewicz, D., et al., 2003. Polymorphisms in CYP1A1, GSTM1, GSTT1 and lung cancer below the age of 45 years. Int J Epidemiol 32:60-63.
105. Demir, A., Demir, I., Alt!n, S., Koksal, V., Cetincelik, U., Dincer, S."., 2005. GSTM1 Gene Polymorphisms on Lung Cancer Development in the Turkish Population. Turkish Respir J 6:131- 134.
106. Awasthi, Y.C., Sharma, R., Singhal, S.S., 1994. Human glutathione S-transferases. Int J Biochem 26:295-308.
107. Morrow, C.S., Simitherman, P.K., Townsend, A.J., 2000. Role of multidrug-resistance protein 2 in glutathione S-transferase P1-1-mediated resistance to 4-nitroquinoline 1-oxide toxicities in HepG2 cells. Mol Carcinog 29:170-178.
108. Hayeshi, R., Mutingwende, I., Mavengere, W., Masiyanise, V., Mukanganyama, S., 2007. Inhibition of human glutathione S-transferases activity by plant polyphenolic compounds ellagic acid and curcumin. Food Chem Toxicol 45:286-295.
109. Gyamfi, M.A., Hokama, O., Oppong-Boachie, K., Aniya, Y., 2000. Inhibitory effects of the medicinal herb, Thonningia sanguinea, on liver drug metabolizing enzymes of rats. Hum Exp Toxicol 19:623- 631.
110. Ploemen, P.J., van Iersel, M.L., Wormhoudt, L.W., Commandeur, J.N., Vermeulen, N.P., van Bladeren, P.J., 1996. In vitro inhibition of rat and human glutathione S-transferase isoenzymes by disulfiram and diethyldithiocarbamate. Biochem Pharmacol 52:197-204.
111. Ernst, E., 2006. Herbal medicines – they are popular, but are they also safe? Eur J Clin Pharmacol 62:1-2.
112. Newman, D.J., Cragg, G.M., Snader, K.M., 2003. Natural products as sources of new drugs over the period 1981-2002. J Nat Prod 66:1022-1037.
113. Newman, D.J., Cragg, G.M., Snader, K.M., 2000. The influence of natural products upon drug discovery. Nat Prod Rep 17:215-234.
39
114. Mandlekar, S., Hong, J.L., Kong, A.N., 2006. Modulation of metabolic enzymes by dietary phytochemicals: a review of mechanisms underlying beneficial versus unfavorable effects. Curr Drug Metab 7:661-675.
115. Matthews, M.K., Jr. 1998. Association of Ginkgo biloba with intracerebral hemorrhage. Neurology 50:1933-1934.
116. Meisel, C., Johne, A., Roots, I., 2003. Fatal intracerebral mass bleeding associated with Ginkgo biloba and ibuprofen. Atherosclerosis 167:367.
117. Henderson, L., Yue, Q.Y., Bergquist, C., gerden, B., Arlett, P., 2002. St John’s wort (Hypericum perforatum): drug interactions and clinical outcomes. Br J Clin Pharmacol 54:349-356.
118. Schulz, V., 2006. Safety of St. John's Wort extract compared to synthetic antidepressants. Phytomedicine 13:199-204.
119. Bailey, D.G., Malcolm, J., Arnold, O., Spence, J.D., 1998. Grapefruit juice drug interactions. Br J Clin Pharmacol 46:101-110.
120. Pabon, H.J.J., 1964. A synthesis of curcumin and related compounds. Rec Trav Chim Pays-Bas 83 :379-386.
121. Ammon, H.P., Wahl, M.A., 1991. Pharmacology of Curcuma longa. Planta Med 57:1-7. 122. Leu, T-H., Maa, M-C., 2002. The molecular mechanisms for the antitumorigenic effect of curcumin.
Curr Med Chem-Anti-Cancer Agents 2:357-370. 123. Vajragupta, O., Boonchoong, P., Morris, G.M., Olson, A.J., 2005. Active site binding modes of
curcumin in HIV-1 protease and integrase. Bioorg Med Chem Lett 15:3364-3368. 124. Cole, G.M., Morihara, T., Lim, G.P., Yang, F., Begum, A., Frautschy, S.A., 2004. NSAID and antio-
xidant prevention of Alzheimer's disease: Lessons from in vitro and animal models. Annals NY Acad Sci 1035:68-84.
125. Reddy, R.C., Vatsaala, P.G., Keshamouni, V.G., Padmanaban, G.,Rangarajan, P.N., 2005. Curcumin for malaria therapy. Biochem Biophy Res Comm 326:472-474.
126. Sharma, R.A., McLelland, H.R., Hill, K.A., Ireson, C.R., Euden, S.A., Manson, M.M., et al., 2001. Pharmacodynamic and pharmacokinetic study of oral Curcuma extract in patients with colorectal cancer. Phytomedicine 13:1894-1900.
127. Oetari, S., Sudibyo, M., Commandeur, J.N.M., Samhoedi, R., Vermeulen, N.P., 1996. Effects of curcumin on cytochrome P450 and glutathione S-transferase activities in rat liver. Biochem Pharmacol 51:39-45
128. Shoba, G., Joy, D., Joseph, T. Majeed, M., 1998. Influence of piperine on the pharmacokinetics of curcumin in animals and human volunteers. Planta Med 64:353-356.
129. Marczylo, T.H., Verschoyle, R.D., Cooke, D.N., Morazzoni, P., Steward, W.P., Gescher, A.J., 2007. Comparison of systemic availability of curcumin with that of curcumin formulated with phosphatidylcholine. Cancer Chemother Pharmacol 60:171-177.
130. Ma, Z., Shayeganpour, A., Brocks, D.R., Lavasanifar, A., Samuel, J., 2007. High performance liquid chromatography analysis of curcumin in rat plasma: application to pharmacokinetics of polymeric micellar formulation of curcumin. Biomed Chromatogr 21:546-552.
131. Bisht, S., Feldmann, G., Soni, S., Ravi, R., Karikar, C., Maitra, A., Maitra, A., 2007. Polymeric nanoparticle-encapsulated curcumin (“nanocurcumin”): a novel strategy for human cancer therapy. J Nanobiotechnology 5:3.
132. Li, L., Ahmed, B., Mehta, K., Kurzrock, R., 2007. Liposomal curcumin with and without oxaliplatin: effects on cell growth, apoptosis, and angiogenesis in colorectal cancer. Mol Cancer Ther 6:1276- 1282.
133. Mosley, C.A., Liotta, D.C., Snyder, J.P., 2007. Highly active anticancer curcumin analogues. Adv Exp Med Biol 595:77-103.
134. Sardjiman, S., Reksohadiprodjo, M., Hakim, L., van der Goot, H., Timmerman, H., 1997. 1,5- Diphenyl-1,4-pentadiene-3-ones and cyclic analogues as antioxidative agents. Synthesis and structure-activity relationship. Eur J Med Chem 32:625-636.
135. Youssef, K.M., El-Sherbeny, M.A., El-Shafie, F.S., Farag, H.A., Al-Deeb, O.A., Awadalla, S.A.A., 2004. Synthesis of curcumin analogues as potential antioxidant, cancer preventive agents. Arch Pharm Pharm Med Chem 337:42-54.
136. Lin, L., Shi, Q., Nyarko, A.K., Bastow, K.F., Wu, C.C., Su, C.Y., et al., 2006. Antitumor agents. 250. Design and synthesis of new curcumin analogues as potential anti-prostate cancer agents. J Med Chem 49:3963-3972.
40
137. Appiah-Opong, R., de Esch, I., Commandeur, J.N., Andarini, M., Vermeulen, N.P., 2008. Structure- activity relationships for the inhibition of recombinant human cytochromes P450 by curcumin analogues. Eur J Med Chem 43:1621-1631.
138. Norinder, U., 2005. In silico modelling of ADMET-a mini review of work from 2000 to 2004. SAR QSAR Environ Res 16:1–11.
139. Martin, Y.C., 2005. A bioavailability score. J Med Chem 48:3164–3170. 140. Mazza, C.B., Whitehead, C.E., Brenerman, C.M., Cramer, S.M., 2002. Predictive quantitative
structure retention relationship models for ion-exchange chromatography. Chromatographia 56:147-152.
141. Nicolotti, O., Pellegrini-Calace, M., Altomare, C., Carrieri, A., Carotti, A., Sanz, F., 2002. Ligands of neuronal nicotinic acetylcholine receptor (nAChR): inteferences from the Hansch and 3-D quantitative structure-activity relationship (QSAR) models. Curr Med Chem 9:1-29.
142. Lill, M.A., 2007. Multidimensional QSAR in drug discovery. Drug Discovery Today 12:1013-1017. 143. Naritomi, Y., Teramura, Y., Terracita, S., Kagayama, A., 2004. Utility of microtiter plate assays for
human cytochrome P450 inhibition studies in drug discovery: application of simple method for detecting quasi-reversible and irreversible inhibitors. Drug Metab Pharmacokinet 19:55-61.
144. Crespi, C.L., Stresser, D.M., 2000. Fluorometric screening for metabolism based drug-drug intera- ctions. J Pharmacol Toxicol Methods 44:325-331.
145. Walsky, R.L., Obach, R.S., Validated assays for human cytochrome P450 activities. Drug Metab Dispos 32:647-660.
146. Eagling, V.A., Tjia, J.F., Back, D.J., 1998. Differential selectivity of cytochrome P450 inhibitors aga- inst probe substrates in human and rat liver microsomes. Br J Clin Pharmacol 45:107-114.
41
Chapter 2
Curcumin: Pharmacokinetics, Metabolism, and Potential for Drug-
Drug/Food Interactions
Appiah-Opong R., Commandeur J.N.M. and Vermeulen N.P.E.
Adapted from Proceedings of The International Symposium on recent progress in Curcumin Research,
11-12 September 2006, Yogyakarta - Indonesia
Curcumin, the yellow pigment in turmeric, and a component of a commonly used spice
(curry) derived from the rhizomes of Curcuma longa prevents carcinogenesis in a
variety of tissues in rodents, especially in the colon and also in humans. In addition, it
possesses several other pharmacological activities. This review highlights reported
findings on the pharmacokinetics, metabolism, and drug-drug/food interactions of
curcumin. Doses up to 8 g/day showed no overt toxicity. Peak serum concentrations of
curcumin were remarkably low, after oral intake and gradually decreased within 12 h. In
both rats and humans, the in vivo metabolites were, hexahydrocurcumin,
hexahydrocurcuminol (octahydrocurcumin), hexahydrocurcumin glucuronide and
curcumin glucuronide, whilst in vitro metabolites were hexahydrocurcumin,
hexahydrocurcuminol, tetrahydocurcumin, curcumin glucuronide, and curcumin sulfate.
In addition, O-demethylated curcumin, bis O-demethylated curcumin, O-demethylated
curcumin glucuronide and curcumin bis-glucuronide have been identified as
metabolites. Some of these metabolites may contribute to the observed
pharmacological properties. Approaches used to enhance bioavailability include co-
administration of curcumin with piperine or formulation of curcumin with
phoshatidylcholine or polyethylene glycol. Drug-drug/food interaction potential,
evaluated by inhibition of cytochrome P450 (CYP) has shown curcumin as a potent
inhibitor of CYP2C9 and CYP3A4.
42
1. Introduction
Curcumin (diferuloylmethane) – (1,7-bis (4-hydroxy-3-methoxyphenyl)-1,6-heptadiene-
3,5-dione) is a natural component of turmeric (Curcuma longa L.) that has emerged as a
promising anti-cancer and chemopreventive agent [1]. The wide spectrum of other
biological activities including chemoprotective, anti-oxidant, anti-inflammatory, anti-HIV
and anti-parasitic properties exhibited by curcumin, has engaged many researchers
over the years [2-6]. Figure 1 summarizes reported therapeutic potentials of curcumin
[7].
Antiangiogenic Diabetes
P latelet
Aggregation
Chemopreventive
skin, liver, colon,
stomach
Multiple
sclerosis
Cardiotoxicity
Wound healing
Curcumin
Cataract formation
Antiinflamatory
Anti -HIV
replication
Cholesterol, smooth
muscle proliferation
Antiangiogenic
nephrotoxicity
Antioxidant
Arth ritis
Lung fibrosis
Gall stones
formation
Figure 1. Curcumin therapeutic potentials (adapted from Aggarwal et al., 2003.
Curcumin has been reported to be non-toxic to humans at doses up to 8 g/day taken for
3 months [8]. Subsequently, a daily oral dose of 3.6 g has been advocated for phase II
evaluation in the prevention or treatment of cancers outside the gastro-intestinal tract
[9].
Investigations on the pharmacokinetics, metabolism and drug-drug/-food
interactions of curcumin have been reported [10-14]. Although several interesting
43
pharmacological properties have been reported on curcumin, it has a drawback of poor
systemic bioavailability [15]. For this reason studies have been performed on
enhancement of its pharmacokinetic properties. These include the use of piperine,
curcumin phospholipids complex, curcumin nanoparticles, liposomal curcumin and
structural analogues of curcumin [16]. Studies on metabolism of curcumin have also
shown that it undergoes reductive metabolism as well as conjugation reactions such as
conjugation with glutathione, glucuronidation and sulfation [17-19]. In addition, its
potential to cause drug-drug/-food interactions by inhibition or induction of CYP
enzymes has been studied [10,14,20]. This review will therefore primarily focus on
recent findings on the pharmacokinetics, metabolism and drug-drug/-food interactions of
curcumin.
2. Pharmacokinetics of curcumin
In a series of animal experiments, Min/+ mouse model of familial adenomatous
polyposis were administered 14C-labeled curcumin (100 mg/kg) via the intra-peritoneal
Figure 2. Elimination of radioactivity derived from [14
C]curcumin from the intestinal tract mucosa (A),
plasma (B), and liver (C) of C57B1/6J mice, which had received a single dose of [14
C]curcumin (100
mg/kg) via the i.p. route. Values are expressed as nmol curcumin equivalents per gram (g tissue; A, C) or
milliliters (ml) of plasma (B), and are the mean + SD of four mice (adapted from [21]).
44
route and monitored for the appearance and disappearance of radioactivity from the
intestinal tract mucosa, plasma and liver (Figure 2) [21].
Table 1. Disposition and toxicity data on curcumin
Dose (g/day)
Period (days)
Subjects Serum/plasma / tissue concn.
Toxicity Reference
2.0 0.17 Healthy volunteers
Serum 0.016 µM None [22]
0.02 1 Healthy volunteers
N None [23]
0.5-12 1 Healthy volunteers
Serum 0.0-0.155 µM None [24]
0.036-0.18
120 Colorectal cancer patients
Serum None Nausea, diarrhea
[12]
0.5-12 90 Malignant lessions patients
Serum 0.5-1.77 µM None [8]
0.45-3.60
7 Cancer patients
Plasma 0.0-3 nmol/L N [25]
0.4 Healthy rats 1.35 µg/ml N [22]
0.2 0.04-0.06
Healthy rats Serum 0.5-1.5 µg/ml N [26]
0.003* 0.02 Adenomatus Mice
Plasma 25 nmol/ml None [21]
0.04 Intestinal mucosa
200 nmol/g None
0.08-0.17
Brain 2.9 nmol/g None
0.01 Liver 73 nmol/g None
0.08-0.17
Kidney 78 nmol/g None
0.08-0.17
Lung 16 nmol/g None
0.08-0.17
Heart 9.1 nmol/g None
N: Not reported; *: i.p. administration
45
At different time points, samples of brain, heart, lung, liver, spleen, kidney, small
intestine and blood were collected, solubilized and analyzed by a liquid scintillation
counting. Peak levels of radioactivity in plasma and tissues, reached in less than four
hours, are shown in Table 1.
Beyond the peak levels, radioactivity declined rapidly to reach levels between 20
and 33% of peak values 4 h after dosing, or 8 h in the case of the small intestine. After
these time points minimal or no decrease in radioactivity levels were observed any more
up to 24 h.
Another group of the mice received diets containing 150-750 mg/kg/day curcumin
for 8 days [21]. Liver, small intestine, colon tissue and plasma were analyzed for
curcumin and its metabolites. Irrespective of the dose, curcumin was detected in plasma
at levels near the limit of detection (5 pmol/ml) and the parent compound was not
detectable in the urine.
In the liver tissue of mice fed a 0.2% curcumin diet, the concentration of curcumin
was 119 + 31 pmol/g of tissue i.e. approximately 0.001 of that observed in the intestinal
mucosa. Significantly higher concentrations of curcumin i.e. up to 3770 nmol/g tissue
were found in the mucosa of the small intestine, colon and faeces, and the latter
contained the largest concentrations (Table 2). Curcumin levels in the small intestines
related to the dose are shown in Table 2, with the exception of levels in colon and
faeces.
Table 2. Concentration of curcumin in small intestinal and colonic mucosa and faeces of mice that
received curcumin at 0.1, 0.2, or 0.5% in their diet for 1 week
Curcumin levelsa (nmol/g)
Curcumin content of diet
(%)
Small intestine
mucosa
Colon
mucosa
Faeces
0.1 39 + 9 15 + 9 3770 + 1246
0.2 111 + 40 508 + 149 3590 + 231
0.5 240 + 69 715 + 448 3186 + 2411
a Values are the means ± SD of four animals (adapted from [21]).
46
Conjugative or reductive metabolites of curcumin were not detected except in the
colonic mucosa and faeces, where HPLC analysis revealed traces of curcumin sulfate.
In mice first fed 300 mg/kg/day of curcumin for 1 week and then changed onto a
curcumin-free diet [21], levels of curcumin in plasma, gastro-intestinal and hepatic
tissues, rapidly declined to unquantifiable concentrations within 3 to 6 h after starting the
curcumin-free diet, however, faecal curcumin declined more slowly with a half-life of
about 23 h. Thus, the rapid removal of curcumin from the rodent tissues, including
target tissues must be taken into account, if sustained levels of curcumin are to be
achieved to elicit its biological effects. Secondly, some metabolites of curcumin may
also contribute to biological activities.
Studies aiming at increasing curcumin systemic bioavailability have recently been
performed. A complex of curcumin and phoshatidylcholine (Meriva), improved the
systemic bioavailability of curcumin in rat plasma and tissues [15]. Levels of curcumin in
the gut mucosa 2 h after administration of 340 mg/kg formulated curcumin by oral
gavages, were moderately lower and in plasma moderately higher than those subjected
to unformulated curcumin. Both formulated and unformulated curcumin were completely
cleared from plasma within 2 h, however, peak plasma levels of curcumin being
approximately 5-fold higher in the formulated form (Table 3).
Table 3. Estimated plasma peak levels for unformulated and formulated (Meriva) Curcumin in rats
Cmax (nM) Tmax (min) AUC (!g min/ml)a
Unformulated
Curcumin 6.5 + 4.5 30 4.8
Curcumin glucuronide 225.0 + 0.6 30 200.7
Curcumin sulfate 7.0 + 11.5 60 15.5
Meriva (formulated)
Curcumin 33.4 + 7.1 15 26.7
Curcumin glucuronide 4420.0 + 292 30 4764.7
Curcumin sulfate 21.2 + 3.9 60 24.8
Cmax, estimated plasma peak levels; Tmax, time of peak levels; aAUC was calculated using WinNonLin and
employing a non-compartmental model (adapted from [15])
47
Nonetheless, maximum systemic concentrations of the curcumin phospholipid complex
were still considerably below the levels eliciting pharmacological effects in cells and cell
free systems. Plasma levels of curcumin glucuronide, curcumin sulfate,
tetrahydrocurcumin and hexahydrocurcumin after administration of formulated curcumin
were 3 to 20 fold higher than after administration of unformulated curcumin (Table 3).
Obviously, a curcumin phospholipid complex significantly enhances systemic and
hepatic bioavailability of parent curcumin and metabolites as compared to unformulated
curcumin [15].
Other strategies to increase bioavailability involve the use of piperine, (1-[5-(1,3-
benzodioxol-5-yl)-1-oxo-2,4-pentadienyl]piperidine), extracted from black pepper to
enhance the bioavailability of curcumin in rats and healthy volunteers [22]. Co-
administration of curcumin (2 g/kg) and this piperine (20 mg/kg) in rats, resulted in
increased serum concentrations of curcumin for a period of 1-2 h after administration.
Elimination half-life of curcumin was significantly increased and clearance was
decreased, whilst bioavailability was increased by 154% with no adverse effects.
Whereas in humans undetectable serum levels of curcumin were obtained after a dose
of 2 g, subsequent co-administration with 20 mg piperine produced a 2000% increase in
bioavailability [22]. This significant increase in bioavailability needs to be verified. The
results suggest that for chemoprevention, targeting sites other than the gastrointestinal
tract, curcumin formulated with phosphatidylcholine or piperine may well be more
advantageous than unformulated curcumin.
Recently, curcumin solubilized with N,N-dimethylacetamide, polyethylene glycol
(PEG 400) and 40% of isotonic dextrose as well as micellar formulation of curcumin,
were administered to rats at doses of 10 and 5 mg/kg body weight [27]. The half-life of
solubilized curcumin was less than 1 h, whereas that of the curcumin encapsulated in
the polymeric micellar formulation was over 60 h. Moreover, a 3-fold decrease in
clearance was observed. In another similar attempt to overcome the low aqueous
solubility and bioavailability of curcumin, curcumin-polyethylene glycol conjugates of two
differently sized polyethylene glycol molecules (PEG 750 and PEG 3500) were used.
These conjugates were employed in treatment of some human cancer cell lines, and
48
were found to exhibit enhanced cytotoxicity as compared to that of the parent drug [28].
Although supporting data is still lacking, these water-soluble conjugates of curcumin
may be useful for injectable curcumin conjugates.
The use of liposomes, a drug delivery system with curcumin has also been
explored using in vitro and in vivo systems [29]. Studies on antitumour activity of
liposomal curcumin towards human pancreatic carcinoma cells have shown that it
inhibits pancreatic carcinoma growth and exhibits antiangiogenic effects. These
properties were compared with that of untreated and liposomal vehicle treated mice and
comparable or greater growth inhibition was observed. However, the bioavailability of
liposomal curcumin as compared to free curcumin is yet to be established.
Application of nanoparticles as in drug delivery systems to improve bioavailabilty
of curcumin has also been explored [30]. Nanocurcumin with less than 100 nm size, has
been synthesized, tested and found to have similar in vitro activity as that of free
curcumin in pancreatic cell lines. Comparison of bioavailability of nanocurcumin to free
curcumin remains to be evaluated.
Another approach to overcome the poor systemic bioavailability was the
synthesis of curcumin analogues [16,31]. A curcumin analogue EF-24 given to CD2F1
mice has been shown to be absorbed rapidly after both oral and i.p. administration [32].
The elimination half-life was 73.6 min and plasma clearance 0.482 L/min/kg. Plasma
peak concentrations detected 3 min after i.p. dosing were about 1000 nM. The EF-24
exhibited 60% and 35% bioavailability upon oral and i.p. administration respectively. EF-
24 was reported to be a lead compound possessing antitumour activity in vitro and in
vivo as compared to curcumin [32]. These analogues showed no in vivo toxicity.
Phase 1 clinical trials of curcumin showed that serum concentrations peaked at 1
to 2 h after oral intake of up to 12 g/day of curcumin, and gradually declined within 12 h
[8]. Remarkably, after administration of 4-8 g of curcumin, only low average peak serum
concentrations of <2.0 !M, were recorded (Table 1). Urine did not contain detectable
amounts of curcumin, and no treatment-related toxicity was observed at curcumin
concentrations < 8 g. In related pharmacokinetic studies of oral curcuma extract in
patients with colorectal cancer, patients were given, 36-180 mg of curcumin, after at
least a 2 h fast [12]. Curcumin and its potential metabolites including curcumin
49
glucuronide, curcumin sulfate, hexahydrocurcumin and hexahydrocurcuminol
(octahydrocurcumin) were not measurable in plasma or urine samples up to 29 days of
daily treatment. However, on days 8 and 29, after 144 and 180 mg consumption of
curcuma extract, faecal samples from patients showed the presence of curcumin
concentrations of < 519 nmol/g and < 1054 nmol/g, respectively. Curcumin sulfate was
the only metabolite identified in the faeces. On the other hand only unchanged curcumin
at a low concentration of 11.1 nmol/L, was detected in plasma samples taken 0.5 and 1
hour after administration to 3 patients consuming 3.6 g of curcumin daily [9]. The only
discernable toxicity upon administration of 0.5-3.6 g curcumin daily for up to 4 months
was a mild diarrhea. Table 1 shows a summary of the serum or plasma concentrations
and toxicity of curcumin in relation to administered doses in both healthy volunteers and
patients. These findings suggest that oral administration of doses up to 3.6 g of
curcumin daily for several months does not result in a significant systemic uptake of
curcumin, nor in tissue accumulation. Oral administration of curcumin will thus not be
efficacious for therapeutic effects in target organs distant from the gastro-intestinal tract,
due to its very low systemic bioavailability.
The low systemic bioavailability of curcumin might imply that the pharmacological
activity especially in tissues other than the colon, is mediated in part by curcumin
metabolites, since several biological activities have been observed at sites distant from
the locus of absorption in rodents, such as breast, prostate and liver [33-35].
Enhancement of curcumin bioavailability clearly might increase the therapeutic
application of this promising drug candidate. Further evaluation of drug-drug interactions
and other toxicities of solubilized or formulated curcumin at administered doses
however remain imperative.
3. Metabolism of curcumin
Metabolism is an integral component of the processes that govern pharmacokinetics,
since it makes drugs more soluble, thus facilitating transport to target organs and
elimination from the body and also usually renders them less toxic. Metabolism studies
on curcumin using slices and subcellular fractions from rat liver identified five reductive
but no oxidative metabolites of curcumin using HPLC and GC-MS analysis [19]. The
50
major reductive metabolites in rat liver slices originate from the reduction of the double
bonds of the heptadiene-3,5-dione chain, resulting in hexahydro-, tetrahydro- and
octahydro-curcumin and the minor products dihydrocurcumin and octahydrocurcumin
(Figure 3) [19]. These metabolites were predominantly present as glucuronides,
although a significant proportion of sulfate conjugates were also observed. No oxidative
metabolites of curcumin nor reductive metabolites were found using rat liver
microsomes and cytosol [19]. The biological activities of most of the reductive
metabolites of curcumin have not yet been established. However, tetrahydrocurcumin
was found to be a more potent antiinflammatory agent [36] and an equipotent
antioxidant as curcumin [37].
Figure 3. Scheme of curcumin metabolism in both phase I and phase II biotransformations.
51
The reduced metabolites appear to be conjugated in vitro and in vivo to a
monoglucuronide, monosulfate and a mixed sulfate/glucuronide [11]. Metabolites
identified in related studies include curcumin glucuronide, curcumin sulfate,
hexahydrocurcumin glucuronide and mixed glucuronide and sulfate conjugates
[11,17,18]. Figure 3 shows an overall scheme of curcumin metabolism in both phase I
and phase II reactions as yet reported in literature.
The metabolism of curcumin in subcellular cytosolic and microsomal fractions of
human liver and intestines has also been reported [18]. Spectrophotometric analysis
demonstrated the reduced metabolites, tetrahydro- and hexahydrocurcumin, and the
phase II metabolites, curcumin sulfate and curcumin glucuronide, containing the intact
yellow-pigmented 1,7-bis (4-hydroxy-3-methoxyphenyl)-1,6-hepta-diene-3,5-dione
structure (Figure 4).
Figure 4. HPLC chromatograms of extracts of incubations of curcumin (100 µM) with cytosol (A and C)
and microsomes (B) from human intestinal tissue and with cytosol (D and F) and microsomes (E) from rat
intestinal tissue. Incubation periods were 90 min for metabolic reduction (A and D) and 60 min for
conjugation (B, C, E and F) and chromatographic peaks were detected at 280nm and 420 nm
respectively. Hexahydrocurcumin (1), tetrahydrocurcumin (2), curcumin (3), curcumin glucuronide (4) and
curcumin sulfate (5) (adapted from [18]).
52
Human intestinal cytosol resulted primarily in curcumin sulfate and hexahydrocurcumin.
Intestinal human microsomes did not generate detectable levels of curcumin reduction
products [18]. Quantitative evaluation of curcumin metabolism in human tissue fractions
suggests that gut metabolism contributes considerably to the total metabolism of
curcumin in vivo [18,21]. This suggestion is in line with results of experiments in which
[3H] labeled curcumin was incubated with everted rat gut sacs [38].
Curcumin and its metabolites in humans were also measured in portal and
peripheral blood, bile and liver tissue, after daily oral intake of 0.45, 1.8 and 3.6 g of
curcumin for one week [39]. Samples of peripheral blood were taken 1 h immediately
after curcumin dosing and hepatic resection was performed 6-7 h after the last dose of
curcumin. Samples of portal blood and bile were taken intra-operatively. Curcumin was
poorly bioavailable following oral administration, with low nanomolar levels of curcumin
and its glucuronide and sulfate conjugates found in peripheral or portal circulation [39].
Although curcumin itself was not found in human liver tissue, trace levels of its
metabolic reduction products, hexahydrocurcumin and octahydrocurcumin were
detected. The low levels of curcumin in plasma are consistent with other clinical findings
indicating that oral doses up to 180 mg of curcumin do not result in detectable plasma
levels [12] while high doses of up to 8 g yield only approximately 0.5-2 µM of curcumin
in serum within 1h of oral administration [8].
Metabolic studies on curcumin have recently also been performed using mouse
and human liver microsomes in the presence of the cofactors NADPH for phase I
reactions and of NADPH and UDPGA for phase II reactions [40]. Analysis by LC-
MS/MS was done in full mass range (m/z: 90-800) and several metabolites including
oxidative metabolites and those previously identified, were found (Figure 3). These
included O-demethylated curcumin (m/z: 355), bis O-demethylated curcumin (m/z:
341), the respective di-hydrogenated derivatives (m/z: 357 and m/z: 343), O-
demethylated curcumin glucuronide (m/z: 517) and curcumin bisglucuronide (m/z:
721) (Figure 3) [40]. In a related study, activities of human hepatic and intestinal
microsomes and nine human recombinant UGT isoforms towards curcumin,
demethoxycurcumin and bis-demethoxycurcumin and their hexahydro- metabolites
53
were determined [41]. Two curcumin monoglucuronides were observed with human
liver microsomes, as previously reported by Pfeiffer et al [42], the major product
having a glucuronic acid at the phenolic position and the minor at the enolic hydroxyl
group, whereas in human intestinal microsomes only the latter conjugate was
observed [41]. Glucuronidation of curcumin was mainly catalyzed by UGT1A1, 1A8
and 1A10, and hexahydrocurcumin by UGT1A8, 1A9 and 2B7 (Table 4). All UGT
isoforms except UGT1A9 formed the phenolic glucuronide of curcumin. Due to its
preference for non-planar phenolic substrates, UGT1A9 preferred hexahydrocurcumin
containing saturated aliphatic chain, hence being less planar than curcumin [41,43].
Table 4. Glucuronidation of curcumin and hexahydrocurcumin by UGT isoforms
UGT activity (pmol/min/mg
protein)
Isoenzyme Curcumin Hexahydrocurcumin
UGT1A1 1875 375
UGT1A3 550 65
UGT1A6 40 nd
UGT1A7 180 210
UGT1A8 1585 950
UGT1A9 100 790
UGT1A10 1540 375
UGT2B7 300 535
Rates of glucuronidation were determined at 20 µM concentrations for each substrate and 0.1-0.2 mg
protein. Reaction mixtures were incubated at 37oC for up to 2 h with linear product formation; nd, not
determined (data adapted from [41]).
These results reaffirm the fact that apart from the liver, the intestinal tract is substantially
contributing to first pass glucuronidation of curcumin.
The studies reported provide clear evidence that after oral intake curcumin is
metabolized in humans predominantly in the intestinal tissue. The rat may not be a good
model for elucidation of the extra hepatic metabolic disposition of curcumin in humans
54
due to the greater ability of human intestinal and liver tissues to metabolize curcumin
than rats. Moreover, it has been observed in humans that cytosol or alcohol
dehydrogenase is required for the formation of tetrahydrocurcumin and
hexahydrocurcumin, while microsomes are needed for the reduction of
hexahydrocurcumin to octahydrocurcumin [18]. The specific enzymes responsible for
the phase I metabolism of curcumin and the pharmacological implications of curcumin
metabolites warrant further investigation. Identification of novel curcumin metabolites
using hybrid quadrupole linear ion trap mass spectrometer coupled with liquid
chromatography [40] have resolved the presence of several additional metabolites in
various tissues, plasma and faeces (Figure 3). Probable curcumin metabolites may
have to be synthesized and used for further identification, pharmacological screening
and safety assessment.
4. Drug-drug and drug-food interactions
Drug-drug interactions are major causes of attrition during drug development [44,45], as
well as of adverse drug reactions in humans [46]. Food-drug or dietary supplement-drug
interactions have also been associated with significant alterations in the
pharmacokinetic profile of various drugs that may have clinical implications [47]. For
example, interactions of red wine with cyclosporine and grapefruit juice with
cyclosporine and felodipine are potential causes of alterations in pharmacokinetics [48],
with possible therapeutic failure and adverse effects [49]. In a study of Jang et al [50],
75% of 116 food supplements were found to induce at least one rat liver microsomal
CYP expression, including CYP1A1, CYP2C11, CYP2D1, CYP2E1 and CYP3A1.
Compounds isolated from St John’s wort, an antidepressant of natural origin, have been
shown to possess potent inhibitory activity towards CYP1A2, CYP2C9, CYP2D6 and
CYP3A4 [51] and induction of CYP3A4 with chronic exposure [52]. Although drug-drug
or drug-food interactions caused by induction of CYP enzymes are known, interactions
due to CYP inhibitions are much more common [53]. Table 5 shows the percent
inhibition of major human CYP isoforms by 625 µg extract of selected natural
products/ml, including curcuma extracts [13].
55
In a recent study, 5 human recombinant CYP isoforms, known to be important
for metabolism of several drugs currently on the market namely, CYP3A4, CYP2D6,
CYP1A2, CYP2C9 and CYP2B6 were evaluated for CYP inhibition by curcumin. The
results showed curcumin as a moderate to potent inhibitor of CYP2C9 and CYP3A4 and
a less potent inhibitor of CYP1A2, CYP2D6 and CYP2B6 (Table 5) [14]. Competitive
type of inhibition was observed in the cases of CYP3A4, CYP1A2 and CYP2B6,
however, in the cases of CYP2C9 and CYP2D6 non-competitive inhibition was
observed. Further experiments showed that curcumin is not a mechanism-based
inhibitor of any of the 5 human CYPs mentioned above [14]. Inhibition of rat liver
microsomal CYPs by curcumin has also been investigated [10].
Table 5. Percent inhibition of major human CYP isoforms, by 625 µg natural products extracts/ml and IC50
values (µM) for curcumin
Extract CYP1A2 CYP2B6 CYP2C9 CYP2D6 CYP3A4
Artemisia vulgaris nd nd 97 100 97
Thyme nd nd 93 96 97
Cloves nd nd 99 98 94
Ginger nd nd 53 70 94
Black tea (5
varieties)
nd nd 92 - 98 76 – 93 77 – 84
Curcuma nd
(40.0*)
nd
(24.5*)
82 (4.3*) 49
(50.3*)
93
(16.3*)
The extracts or infusions were each tested at a single concentration indicated above (adapted from [13]).
*IC50 values for curcumin were determined within a concentration range of 0.4-181.8 µM [14]. nd, not
determined
Curcumin was found to be a potent competitive inhibitor of rat liver CYP1A1/1A2
measured as ethoxyresorufin deethylation (EROD) activity in !-naphthoflavone-induced
microsomes, a less potent competitive inhibitor of CYP2B1/2B2, measured as
pentoxyresorufin depentylation (PROD) activity in phenobarbital-induced microsomes
and a weak inhibitor of CYP2E1, measured as p-nitrophenol (PNP) hydroxylation
56
activity in pyrazole (Pyr)-induced microsomes [10]. Curcumin was found to be a very
weak inhibitor of CYP2E1.
Induction studies on the effects of dietary flavonoids including curcumin on
human CYP1A1 expression have been performed on the 101L cell line (derived from
human hepatoma cell line HepG2), transfected with a plasmid containing the human
CYP1A1 promoter [20]. The 101L cells were plated at a density of 7.5 x 104 cells/well in
96-well plates, and dose response curves for the various flavonoids were generated at
doses ranging from 1 to 20 !M with 18 h of exposure (Figure 5). A 3-fold increase in
activity of CYP1A1 was measured as elevation in luciferase activity. Compared to
2,3,7,8-tetrachlorodibenzo-"-dioxin (TCDD), omeprazole or benzanthracene where
increases in activity ranged from 12- to 35-fold, curcumin is a weak inducer of CYP1A1
[20].
Figure 5. Dose response curves for various flavonoids. Doses ranged from 1-20 µM and cells were
exposed to each agent for 18 h. Each time point represents the mean of data from three experiments +
SD. (adapted from [20]).
Apart from drug-drug interactions, CYP inhibition has been related to
chemopreventive activity against benzopyrene (B[a]P)-induced carcinogenesis, due to
57
inhibition of CYP1A1-mediated bioactivation of B[a]P. CYP inhibition may affect the
pharmacokinetics of drugs by decreasing the elimination-half life, hence increasing the
plasma concentration increasing the potential for toxic consequences. Animal data is
often poorly predictive of human situations due to species differences, including
differences in the properties of metabolic enzymes [54]. This is exemplified by high
inhibitory potency of curcumin towards rat liver CYP1A1/1A2 compared with the lower
potency obtained with human CYP1A2. The drug-drug interaction potential of curcumin
resulting from inhibition of CYPs needs to be further investigated for clinical relevance.
Curcumin has also been demonstrated to inhibit the expression of P-glycoprotein
(Pgp) in the multi-drug resistance (MDR) human cervical carcinoma KB-VI cells,
increase rhodamine-123 (Rh123) accumulation and inhibit Rh123 efflux in these cells
[55]. MDR is a challenge, limiting the success of chemotherapy, mainly due to the
overexpression of the Pgp, which causes a decrease in drug accumulation in cancer
cells [56]. Both Pgp and CYP3A4 have been suggested to act synergistically to limit the
bioavailability of orally administered agents [57]. Thus the inhibition of both enzymes by
curcumin may be related to the chemopreventive role of the latter.
Curcumin is also a potent inhibitor of glutathione S-transferase (GST) activity in
cytosol of rat liver treated with PB, !NF and Pyr, with 1-chloro-2,4-dinitrobenzene
(CDNB) as substrate. Similarly, curcumin has been found to be a potent inhibitor of
humans recombinant GSTs [58, in this thesis, Chapter 5]. Curcumin was shown to
inhibit GSTs A1-1, A2-2, M1-1, M2-2 and P1-1 with IC50 values ranging from 0.04 to 5
!M. GSTs are often overexpressed in drug-resistant cell lines including cancer cells.
Elevated GST activity is regarded as an indicator for the resistance to chemotherapy
[58]. In addition, multidrug resistance has been associated with a decrease in
intracellular drug accumulation in patient tumour cells due to enhanced drug efflux or
enhanced metabolism through GSTs. The inhibition of GST by curcumin therefore
renders it a promising anticancer agent [58]. On the other hand inhibition of GSTs
reduces their protective role in detoxification of electrophilic substances through
glutathione (GSH) conjugation [59]. Thus prolonged GST inhibition could also result in
toxicity of electrophilic chemicals or metabolites [60].
58
Investigation of the effect of three series of curcumin analogues designed and
synthesized by Sardjiman et al [61] on GSTA1-1, GSTM1-1, GSTP1-1 and human and
rat cytosolic GSTs revealed a variety of GST activities in the presences of varying
chemical structures [in this thesis Chapter 5]. Seven of the thirty-four compounds were
more potent inhibitors of GSTA1-1, whilst three and four compounds respectively, were
more potent inhibitors of GSTM1-1 and GSTP1-1 than curcumin itself. One of the three
series of curcumin analogues lacked inhibitory activity towards GSTP1-1. Although the
strong inhibitory activities of curcumin and some of its analogues towards GSTs could
have useful applications in cancer chemotherapy, these activities could also have
implications for toxicity in normal cells in the presence of potentially toxic compounds
with electrophilic properties, and/or electrophilic metabolite, since GSTs are major
detoxification enzymes in the body. Knowledge on the structure-activity relationship can
be useful in the designing of curcumin analogues with less or more GST-inhibitory
properties. The clinical relevance of these inhibitory activities, also in view of the organ
and tissue distribution (for example GSTP1-1 is primarily found in erythrocyte and not
the liver [62], as well as polymorphisms of GSTs, remains to be established.
5. Conclusion
This review has focused on the pharmacokinetics, metabolism and drug-drug interaction
potential of curcumin, all aspects of potential value in human applications of curcumin.
The extremely poor bioavailability and subsequent high exposure of the intestinal
mucosa to curcumin support the clinical evidence of its potential as therapeutic agent
for colorectal cancer. In line with this observation, orally administered curcumin is not
likely to become clinically useful in prevention of tumours distant from the locus of
absorption, nor other clinical applications for which systemic availability is needed.
However, local administration may be required where higher concentrations are
necessary in systemic, target organ or tissues. On the other hand, curcumin
formulations resulting in enhanced bioavailability are more promising alternatives. Apart
from significant phase II metabolism notably, glucuronidation, sulfation and GSH
conjugation, limited phase I metabolism notably, reductive and minor oxidative
metabolism have been demonstrated. The pharmacological significance of curcumin
59
metabolism and its metabolites need to be further investigated. The same holds true for
easily occurring decomposition products of curcumin. Knowledge of structure activity
relationships of curcumin analogues may be useful in redesigning analogues with less
potential for drug-drug interactions. Drug-food interactions have the potential to cause
harmful effects. Therefore, a rational approach is required to screen curcumin
formulations and foods for in vitro CYP inhibitory activities, since curcumin itself has
been shown to inhibit human important CYPs. The mechanisms underlying the
interactions between several enzymes and transporters with the properties of curcumin
also warrant further investigations. The strength of curcumin is also its weakness, thus
findings on the pharmacokinetics, metabolism and potential for drug-drug interactions
need to be considered for a more useful application of the compound.
References
1. Singh, S., Khar, A., 2006. Biological effects of curcumin and its role in cancerchemoprevention and therapy. Anti-Cancer Agents Med Chem 6:259-270.
2. Donatus, I.A,, Sardjoko, Vermeulen, N.P., 1990. Cytotoxic and cytoprotective activities curcumin. Effects on paracetamol-induced cytotoxicity, lipid peroxidation and glutathione depletion in rat hepatocytes. Biochem Pharmacol 39:1869-1875.
3. Leu, T-H., Maa, M-C., 2002. The molecular mechanisms for the antitumorigenic effect of curcumin. Curr Med Chem Anti-Cancer Agents 2:357-370.
4. Cole, G.M., Morihara, T., Lim, G.P., Yang, F., Begum, A., Frautschy, S.A., 2004. NSAID and antioxidant prevention of Alzheimer's disease: Lessons from in vitro and animal models. Annals NY Acad Sci 1035:68-84.
5. Vajragupta, O., Boonchoong, P., Morris, G.M., Olson, A.J., 2005. Active site binding modes of curcumin in HIV-1 protease and integrase. Bioorg Med Chem Letters 15:3364-3368.
6. Reddy, R.C., Vatsaala, P.G., Keshamouni, V.G., Padmanaban, G., Rangarajan, P.N., 2005. Curcumin for malaria therapy. Biochem Biophy Res Comm 326:472-474.
7. Aggarwal, B.B., Kumar, A., Bharti, A.C., 2003. Anticancer potential of curcumin: preclinical and clinical studies. Anticancer Res 23:363-398.
8. Cheng, A.L., Hsu, C.H., Lin, J.K., Hsu, M.M., Ho, Y.F., Shen, T.S., et al., 2001. Phase I clinical trial of curcumin, a chemopreventive agent, in patients with high-risk or pre-malignant lesions. Anticancer Res 21:2895-2900.
9. Sharma, R.A., Euden, S.A., Platton, S.L., Cooke, D.N., Shafayat, A., Hewitt, H.R., et al., 2004. Phase I clinical trial of oral curcumin: biomarkers of systemic activity and compliance. Clin Cancer Res 10:6847-6854.
10. Oetari, S., Sudibyo, M., Commandeur, J.N.M., Samhoedi, R., Vermeulen, N.P., 1996. Effects of curcumin on cytochrome P450 and glutathione S-transferase activities in rat liver. Biochem Pharmacol 51:39-45. 11. Ireson, C., Orr, S., Jones, D.J., Verschoyle, R., Lim, C.K., Luo, J.L., et al., 2001. Characterization of
metabolites of the chemopreventive agent curcumin in human and rat hepatocytes and in the rat in vivo, and evaluation of their ability to inhibit phorbol ester-induced prostaglandin E2 production. Cancer Res 61:1058-1064.
60
12. Sharma, R.A., McLelland, H.R., Hill, K.A., Ireson, C.R., Euden, S.A., Manson, M.M., et al., 2001. Pharmacodynamic and pharmacokinetic study of oral curcuma extract in patients with colorectal cancer. Clin Cancer Res 7:1894-1900.
13. Foster, B.C., Vandenhoek, S., Hana, J., Krantis, A., Akhtar, M.H., Bryan, M., Budzinski, J.W., Ramputh, A., Arnason, J.T., 2003. In vitro inhibition of human cytochrome P450-mediated metabolism of marker substrates by natural products. Phytomedicine 10:334-342.
14. Appiah-Opong, R., Commandeur, J.N.M., van Vugt-Lussenburg, B., Vermeulen, N.P.E., 2007. Inhibition of human recombinant cytochrome P450s by curcumin and curcumin decomposition products. Toxicology 235:83-91.
15. Marczylo, T.H., Verschoyle, R.D., Cooke, D.N., Morazzoni, P., Steward, W.P., Gescher, A.J., 2007. Comparison of systemic availability of curcumin with that of curcumin formulated with phosphatidylcholine. Cancer Chemother Pharmacol 60:171-177.
16. Anand, P., Kunnumakkara, A.B., Newman, R.A., Aggarwal, B.B., 2007. Bioavailability of curcumin: Problems and promises. Mol Pharmaceut DOI: 10.1021/mp700113r
17. Asai, A ., Miyazawa, T., 2000. Occurrence of orally administered curcuminoid as glucuronide and glucuronide/sulfate conjugates in rat plasma. Life Sci 67:2785-2793. 18. Ireson, C.R., Jones, D.J., Orr, S., Coughtrie, M.W., Boocock, D.J., Williams, M.L., Farmer, P.B.,
Steward, W.P., Gescher, A.J., 2002. Metabolism of the cancer chemopreventive agent curcumin in human and rat intestine. Cancer Epidemiol Biomarkers Prev 11:105-111.
19. Hoehle, S.I., Pfeiffer, E., Solyom, A.M., Metzler, M., 2006. Metabolism of curcuminoids in tissue slices and subcellular fractions from rat liver. J Agric Food Chem 54:756-764.
20. Allen, S.W., Mueller, L., Williams, S.N., Quattrochi, L.C., Raucy, J., 2001. The use of a high-volume screening procedure to assess the effects of dietary flavonoids on human CYP1A1 expression. Drug Metab Dispos 29:1074-1079.
21. Perkins, S., Verschoyle, R.D., Hill, K., Parveen, Threadgill, M.D., Sharma, R.A., Williams, M.I., Steward, W.P., Gescher, A.J., 2002. Chemopreventive efficacy and pharmacokinetics of curcumin in the Min/+ mouse, a model of familial adenomatous polyposis. Cancer Epid Biomarkers & prevention 11:535-540.
22. Shoba, G., Joy, D., Joseph, T., Majeed, M., 1998. Influence of piperine on the pharmacokinetics of curcumin in animals and human volunteers. Planta Med 64:353-356.
23. Rasyid, A., Lelo, A., 1999. The effect of curcumin and placebo on human gall-bladder function: an ultrasound study. Alimentary Pharmacol Ther 13:245-249. 24. Lao, C.D., Ruffin, M.T., Normolle, D., Heath, D.D., Murray, S.I., Bailey, J.M., et al., 2006. Dose
escalation of a curcuminoid formulation. Complement Altern Med 6:10. 25. Garcea, G., Berry, D.P., Jones, D.J., Singh, R., Dennison, A.R., Farmer, P.B., Sharma, R.A.,
Steward, W.P., Gescher, A.J., 2005. Consumption of the putative chemopreventive agent curcumin by cancer patients: assessment of curcmin levels in the colorectum and their pharmacodynamic consequences. Cancer Epidermiol Biomarkers Prev 14:120-125.
26. Maiti, K., Mukherjee, K., Gantait, A., Saha, B.P., Mukherjee, P.K., 2007. Curcumin-phospholipid complex; preparation therapeutic evaluation and pharmacokinetics in rats. Int J Pharm 330:155-163.
27. Ma, Z., Shayeganpour, A., Brocks, D.R., Lavasanifar, A., Samuel, J., 2007. High performance liquid chromatography analysis of curcumin in rat plasma: application to pharmacokinetics of polymeric micellar formulation of curcumin. Biomed Chromatogr 21:546-552.
28. Safavy, A., Raisch, K.P., Mantena, S., Sanford, L.L., Sham, S.W., Krishna, N.R., Bonner, J.A., 2007. Design and development of water-soluble curcumin conjugates as potential anticancer agents. J Med Chem 50:6284-6288.
29. Li, L., Ahmed, B., Mehta, K., Kurzrock, R., 2007. Liposomal curcumin with and without oxaliplatin: effects on cell growth, apoptosis, and angiogenesis in colorectal cancer. Mol Cancer Ther 6:1276-1282. 30. Bisht, S., Feldmann, G., Soni, S., Ravi, R., Karikar, C., Maitra, A., Maitra, A., 2007. Polymeric nanoparticle-encapsulated curcumin (“nanocurcumin”): a novel strategy for human cancer therapy. J Nanobiotechnology 5:3. 31. Preetha, A., Banerjee, R., Huilgol, N., 2007. Tensiometric profiles and their modulation by cholesterol: implications in cervical cancer. Cancer Invest 25:172-181.
61
32. Mosley, C.A., Liotta, D.C., Snyder, J.P., 2007. Highly active anticancer curcumin analogues. Adv Exp Med Biol 595:77-103. 33. Inano, H., Onoda, M., 2002. Prevention of radiation-induced mammary tumours. Int J Radiat Oncol
52:212-223. 34. Frank, N., Knauft, J., Amelung, F., Nair, J., Wesch, H., Bartsch, H., 2003. No prevention of liver and
kidney tumours in Long-Evans Cinnamon rats by dietary curcumin, but inhibition at other sites and of metastases. Mutat Res 127-135:523-524.
35. Nanji, A.A., Jokelainen, K., Tipoe, G.L., Rahemtulla, A., Thomas, P., Dannenberg, A.J., 2003. Curcumin prevents alcohol-induced liver disease in rats by inhibiting the expression of NF-kappa B-dependent genes. Am J Physiol Gastrointest Liver Physiol 284:G321-G327.
36. Mukhopadhyay, A., Basu, N., Ghatak, N., Gujral, P.K., 1982. Anti-inflammatory and irritant activities of curcumin analogues in rats. Agents Actions 12:508-515.
37. Okada, K., Wangpoentrakul, C., Tanaka, T., Toyokuni, S., Uchida, K., Osawa, T., 2001. Curcumin and especially tetrahydrocurcumin ameliorate oxidative stress-induced renal injury in mice. J Nutri 131:2090-2095.
38. Ravindranath, V., Chandrasekhara, N., 1981. In vitro studies on the intestinal absorption of curcumin in rats. Toxicology 20:251-257.
39. Garcea, G., Jones, D.J., Singh, R., Dennison, A.R., Farmer, P.B., Sharma, R.A., Steward, W.P., Gescher, A.J., Berry, D.P., 2004. Detection of curcumin metabolites in hepatic tissues and portal blood of patients following oral administration. Br J Cancer 90:1011-1015.
40. Tamvakopoulos, C., Sofianos, Z.D., Garbis, S.D., Pantazis, P., 2007. Analysis of the in vitro metabolites of diferuloylmethane (curcumin) by liquid chromatography – tandem mass spectrometry on a hybrid quadrupole linear ion trap system: newly identified metabolites. Eur J Drug Metab Pharmacokinet 32:51-57.
41. Hoehle, S.I., Pfeiffer, E., Metzler, M., 2007. Glucuronidation of curcuminoids by human microsomal and recombinant UDP-glucuronosyltransferases. Mol Nutr Food Res 51:932-938.
42 Pfeiffer, E., Hoehle, S.I., Walch, S.G., Riess, A., Solyom, A.M., Metzler, M., 2007. Curcuminoids form reactive glucuronides in vitro. J Agric Food Chem 55:538-544.
43. Ebner, T., Burchell, B., 1993. Substrate specificies of two stably expressed human liver UDP-glucuronosyltransferases of the UGT1 gene family. Drug Metab Dispos 21:50-55.
44. Riley, R.J., Kenna, J.G., 2004. Cellular models for ADMET predictions and evaluation of drug-drug interactions. Curr Opin Drug Discov Devel 7:86-99.
45. Brown, H.S., Galetin, A., Hallifax, D., Houston, J.B., 2006. Prediction if in vivo drug-drug interactions from in vitro data: Factors affecting prototypic drug-drug interactions involving CYP2C9, CYP2D6 and CYP3A4. Clin Pharmacokinet 45:1035-1050.
46. Pea, F., Furlanut, M., 2001. Pharmacokinetic aspects of treating infections in the intensive care unit: focus on drug interactions. Clin Pharmacokinet 40:833-868.
47. Mandlekar, S., Hong, J.L., Kong, A.N., 2006. Modulation of metabolic enzymes by dietary phytochemicals: a review of mechanisms underlying beneficial versus unfavourable effects. Curr Drug Metab 7:661-675.
48. Harris, R.Z., Jang, G.R., Tsunda, S. 2003. Dietary effects on drug metabolism and transport. Clin Pharmacokinet 42:1071-1088.
49. Bailey, D.G., Dresser, G.K., Bend, J.R., 2003. Bergamottin, lime juice, and red wine as inhibitors of cytochrome P450 3A4 activity: Comparison with grapefruit juice. Clin Pharmacol Ther 73:529- 537. 50. Jang, E.H., Park, Y.C., Chung, W.G., 2004. Effects of dietary suppliments on induction and
inhibition of cytochrome P450s protein expression in rats. Food Chem Toxicol 42:1749-1756. 51. Obach, R.S., 2000. Inhibition of human cytochrome P450 enzymes by constituents of St. John's
wort, an herbal preparation used in the treatment of depression. J Pharmacol Exp Ther 294:88-95. 52. Komoroski, B.J., Zhang, S., Cai, H., Hutzler, J.M., Frye, R., Tracy, T.S., et al., 2004. Induction and
inhibition of cytochromes P450 by the St. John's wort constituent hyperforin in human hepatocyte cultures. Drug Metab Dispos 32:512-518.
53. Zafar, A., Sharif, M.D., 2003. Pharmacokinetics, metabolism, and metabolism of atypical antipsychotics in special populations. Primary care companion J Clin Psychiatry 5:22-25. 54. Eagling, V.A., Tjia, J.F., Back, D.J., 1998. Differential selectivity of cytochrome P450 inhibitors
against probe substrates in human and rat liver microsomes. Br J Clin Pharmacol 45:107-114.
62
55. Anuchapreeda, S., Leechanacha, P., Smith, M.M., Ambudkar, S.V., Limtrakul, P.N., 2002. Modulation of P-glycoprotein expression and function by curcumin in multi-drug resistant human KB cells. Biochem Pharmacol 64:573-582.
56. Limtrakul, P., Khantamat, O., Pintha, K., 2004. Inhibition of P-glycoprotein activity and reversal of cancer multi-drug resistance by Momordica charantia extracts. Cancer Chemother Pharmacol 54:525-530. 57. Achira, M., Suzuki, H., Ito, K., Sugiyama, Y., 2001. Comparative studies to determine the selective
inhibitors for P-glycoprotein and cytochrome P450 3A4. AAPS Pharm Sci 3:18 DOI: 10.1208/ps030218.
58. Hayeshi, R., Mutingwende, I., Mavengere, W., Masiyanise, V., Mukanganyama, S., 2007. Inhibition of human glutathione S-transferases activity by plant polyphenolic compounds ellagic acid and curcumin. Food Chem Toxicol 45:286-295.
59. Commandeur, J.N., Stijntjes, G.J., Vermeulen, N.P., 1995. Enzymes and transport systems involved in the formation and disposition of glutathione S-conjugates. Role in bioactivation and detoxication mechanisms of xenobiotics. Pharmacol Rev 47:271-330.
60. DeLeve, L.D., Wang, X., 2000. Role of oxidative stress glutathione in busulfan toxicity in cultured murine hepatocytes. Pharmacology 60:143-154.
61. Sardjiman, S., Reksohadiprodjo, M., Hakim, L., van der Goot, H., Timmerman, H., 1997. 1,5- Diphenyl-1,4-pentadiene-3-ones and cyclic analogues as antioxidative agents. Synthesis and structure-activity relationship. Eur J Med Chem 32:625-636. 62. Awashti, Y.C., Sharma, R., Singhal, S.S., 1994. Human glutathione S-transferases. Int J Biochem
26:295-308.
63
Inhibition of CYP/GST activities by natural products and derivatives
64
65
Chapter 3
Inhibition of human recombinant cytochrome P450s by curcumin and
curcumin decomposition products
Regina Appiah-Opong, Jan N. M. Commandeur, Barbara van Vugt-
Lussenburg, and Nico P. E. Vermeulen
Adapted from Toxicology 2007 235:83-91
Curcumin (diferuloylmethane) is a major yellow pigment and dietary component derived
from Curcuma longa. It has potent anti-inflammatory, anti-carcinogenic, anti-oxidant and
chemoprotective activities among others. We studied the interactions of curcumin, a
mixture of its decomposition products, and four of its individually identified
decomposition products (vanillin, vanillic acid, ferulic aldehyde and ferulic acid) on five
major human drug metabolizing cytochrome P450s (CYPs). Curcumin inhibited
CYP1A2 (IC50, 40.0 µM), CYP3A4 (IC50, 16.3 µM), CYP2D6 (IC50, 50.3 µM), CYP2C9
(IC50, 4.3 µM) and CYP2B6 (IC50, 24.5 µM). Curcumin showed a competitive type of
inhibition towards CYP1A2, CYP3A4 and CYP2B6, whereas a non-competitive type of
inhibition was observed with respect to CYP2D6 and CYP2C9. The inhibitory activity
towards CYP3A4, shown by curcumin may have implications for drug-drug interactions
in the intestines, in case of high exposure of the intestines to curcumin upon oral
administration. In spite of the significant inhibitory activities shown towards the major
CYPs in vitro, it remains to be established, whether curcumin will cause significant drug-
drug interactions in the liver, given the reported low systemic exposure of the liver to
curcumin. The decomposition products of curcumin showed no significant inhibitory
activities towards the CYPs investigated, and therefore, are not likely to cause drug-
drug interactions at the level of CYPs.
66
1. Introduction
Multiple drug therapy is a common therapeutic practice, especially in patients with
multiple complications [1,2]. If two or more drugs with affinity for the same CYP enzyme
are co-administered, their biotransformation may be compromised, leading to
undesirable accumulation of the drugs with toxic side-effects as possible consequence.
Drug-drug interactions involving CYPs have been identified as an important cause of
adverse drug reactions and therapeutic failure [3,4]. Drug-drug interactions may be due
to enzyme induction or inhibition, the latter being more common [2,5]. Next to drugs,
several natural compounds have also been shown to cause significant interactions at
the level of drug-metabolizing enzymes [6,7].
Curcumin, a polyphenolic component of turmeric (Curcuma longa), is a yellow
pigment widely used for coloring of foods. It has been shown previously to exhibit anti-
cancer, anti-oxidant, anti-inflammatory, anti-parasitic and anti-HIV properties [8-11].
Curcumin has also been shown to have chemoprotective, chemopreventive and
immuno-modulating properties [9,12,13]. Curcumin can be considered as a safe
compound, because oral doses as high as 8 g/day administered to humans did not
result in overt side effects [13]. Clinical trials for the use of curcumin as an anti-cancer
agent are currently ongoing [14].
Because relatively high doses of curcumin are evaluated in human studies, it
might be anticipated that curcumin might cause drug-drug interactions at the level of
intestinal and/or liver drug metabolism. Several in vivo and in vitro animal studies have
shown that curcumin can significantly modulate the activity of several drug metabolizing
enzymes by down-regulation, induction or by direct inhibition. Oetari et al [15] and
Thapliyal and Maru [16] reported potent inhibition of rat liver microsomal CYP1A1,
CYP1A2 and CYP2B1 enzymes by curcumin. In in vivo studies repetitive administration
of curcumin to rats resulted in down-regulation of intestinal CYP3A-enzymes, whereas
hepatic and renal CYP3A-levels were significantly induced [17]. Also, down-regulation
of esophagal CYP2B1 and CYP2E1 was reported after intragastric treatment of rats,
67
which might partially explain the chemopreventive activity of curcumin against
carcinogenic N-nitrosamines [18].
As yet, the effects of curcumin on the major human drug metabolizing CYPs
have not been studied. Due to the large species differences in the properties of
metabolic enzymes and metabolic profiles of drugs, the animal studies described above
are poorly predictive for the human situation [19,20]. The present in vitro investigation
therefore was designed to assess the potential of curcumin to cause drug-drug
interactions via inhibition of the five most important human drug metabolizing CYPs.
Major human CYP isoforms responsible for the metabolism and disposition of about
90% of the therapeutic drugs on the market include CYP1A2, CYP2D6, CYP2B6,
CYP2E1, CYP2C9 and CYP3A4 [21].
Figure 1. Chemical structures of curcumin and its decomposition products at pH 7.4
Because curcumin has been shown to be chemically unstable under neutral and
alkaline conditions [22], the inhibitory properties of these decomposition products were
also studied as mixture and individually, if available. Degradation products which have
been identified include trans-6(4’-hydroxy-3’-methoxyphenyl)-2,4-dioxo-5-hexenal, and
O O
CHO COOH
OH
CH CHCOOHCH CHCHO
OH OHOH
OHOH
HO
O
O CHOH
OCH3
OCH3OCH
3OCH3
OCH3
CH3
O
CH3
O
Curcumin
Vanillin Vanillic acid Ferulic acidFerulic aldehyde
Trans-6-(4'-hydroxy-3'-methoxyphenyl)-4-dioxo-5-hexenal
68
minor products being vanillin, vanillic acid, ferulic aldehyde and ferulic acid (Figure 1)
[22]. Apart from reversible inhibition, mechanism-based inhibition was also taken into
consideration.
2. Materials and methods
2.1. Materials
Methoxyresorufin and benzyloxyresorufin were synthesized by the method of Burke et
al [23] and the purity was determined by HPLC, mass spectrometry and 1H NMR. The
plasmid, pSP19T7LT_2D6 containing human CYP2D6 bicistronically co-expressed with
human cytochrome P450 NADPH reductase was kindly provided by Prof. M. Ingelman-
Sundberg (Stockholm, Sweden). The plasmids, BMX100/h1A2 and pCWh3A4 with
human cytochrome P450 NADPH reductase were kindly donated by Dr. M. Kranendonk
(Lisbon, Portugal). Expression plasmids, pCWh2B6hNPR and pCWh2C9hNPR with
human cytochrome P450 NADPH reductase were kindly provided by Prof. F.P.
Guengerich (Nashville, Texas, USA). All other chemicals were of analytical grade and
obtained from standard suppliers.
2.1. CYP expression and membrane isolation
The plasmids containing cDNA of five human CYPs were transformed into Escherichia
coli strain JM109. Expression of the CYPs was carried out in 3-litre flasks containing
300 ml terrific broth (TB) with 1mM !-aminolevulinic acid, 0.5 mM thiamine, 400 µl/L
trace elements, 100 µg/ml ampicillin, 1 mM isopropyl-"-D-thiogalactopyranoside (IPTG),
0.5 mM FeCl3 (for CYP2D6 and CYP3A4 only) and 30 µg/ml kanamycin (for CYP3A4
only). The culture media were inoculated with 3 ml overnight cultures of bacteria
containing plasmids for the various CYPs. The cell cultures were incubated for about 40
h at 28 oC and 125 rpm and CYP contents were determined using the carbon monoxide
(CO) difference spectra as described by Omura and Sato [24]. Cells were pelleted by
centrifugation (4000 g, 4 oC, 15 min) and resuspended in 30 ml Tris-Sucrose-EDTA
(TSE) buffer (50 mM Tris-acetate buffer pH 7.6, 250 mM sucrose, 0.25 mM EDTA).
Cells were treated with 0.5 mg/ml lysozyme prior to disruption by French press (1000
psi, 3 repeats). The membranes containing the human CYPs were isolated by
69
ultracentrifugation in a Beckmann 50.2Ti rotor (60 min, 40,000 rpm, 4 oC), resuspended
in TSE buffer and stored at –80 oC until use.
2.2. Decomposition of curcumin
Curcumin decomposition was performed according to the method of Wang et al [22] in
phosphate buffer of pH 7.4. Briefly, aliquots of 20 µl of 5 mM curcumin (dissolved in
methanol) were added to 980 µl of 0.1 M potassium phosphate buffer pH 7.4. After 1 h
incubation at 37 oC Samples were analyzed by HPLC (Jasco Separations FP 1575).
Vanillin, vanillic acid, ferulic aldehyde and ferulic acid were used as standards. Gradient
reversed-phase HPLC separations were performed using a C18 column (150 mm x 3.2
mm, 5 µm particle size, Phenomenex) and a mobile phase consisting of 0.1% acetic
acid (solvent A) and 98% methanol with 0.1% acetic acid (solvent B). HPLC-gradient
separation was carried out with a linear gradient eluting from 15% to 95% of solvent B in
60 min. The carrier flow rate was 0.4 ml/min and chromatographic peaks of
decomposition products were monitored by UV detection (! = 280 nm). Under these
chromatographic conditions curcumin and commercially available reference compounds
of possible degradation products eluted at: 28.5 min (curcumin), 14.3 min (vanillic acid),
15.2 min (vanillin), and 17.9 min (ferulic acid) 18.5 min (ferulic aldehyde).
2.3. CYP inhibition assays
2.3.1. 7-Methoxyresorufin, 7-benzyloxyresorufin, dibenzylfluorescein, 7-benzyloxy-4-
trifluo-romethylcoumarin and 7-benzyloxyquinoline O-dealkylation
Inhibition of the activities of the human CYP isoforms 1A2, 3A4 and 2B6, by curcumin
and its decomposition products was determined using microplate fluorimetric assays
[23,25]. Incubation conditions (eg. enzyme concentration, substrates, incubation time)
and wavelengths for detection for each assay are shown in Table 1. Under these
conditions kinetics of the reactions were linear over the periods indicated. The inhibitory
activity of curcumin towards CYP3A4 was tested using four different substrates of
CYP3A4, namely DBF, BFC and BQ, because inhibition of CYP3A4 activity is known to
be substrate-dependent [25]. In general, the incubations were carried out in a total
70
volume of 200 µl and in the presence of 100 µM NADPH (freshly prepared) in a black
coaster 96-well plate. Membranes were pre-incubated for 5 min at 37 oC with 0.1 M
potassium phosphate buffer (pH 7.4), substrates, and inhibitors (curcumin and its
decomposition products) with DMSO in concentration 0.5% (v/v) or less in all the CYP
assays.
Table 1. Experimental conditions for fluorescence CYP assays
CYP Enzyme Incubation Substrate Substrate Excitation Emission
amount time conc wavelength wavelength
nM min µM nm
1A2 13.2 10 MRes 5.0 530 586
3A4 14.3 30 BRes 5.0 530 586
2.8 10 DBF 0.5 485 535
14.3 20 BFC 80.0 410 538
14.3 30 BQ 40.0 410 538
2B6 15.3 30 BRes 20.0 530 586
MRes, methoxyresorufin; BRes, benzyloxyresorufin; DBF, dibenzylfluorescein; BQ, 7-benzyloxyquinoline;
BFC, 7-benzyloxy-4-trifluoromethylcoumarin
DMSO concentration was consistent in each assay. At these DMSO concentrations, the
studied human CYPs are not affected [26,27]. Stability of curcumin in phosphate buffer
pH 7.4, has been reported to be strongly improved by addition of rat liver microsomes or
cytosol, glutathione (GSH), N-acetyl L-cysteine (NAC) or ascorbic acid [15]. Therefore
enzymes were always added before the incorporation of curcumin in all inhibition
assays performed.
Initially, the inhibitory effect of curcumin and its decomposition products on the
CYP isoforms was studied at a concentration of 300 !M. For IC50 determinations,
concentration ranges for curcumin used were from 0.9 to 100 µM and for the
decomposition products, from 2.5 to 1000 µM. Incubations were commenced by the
addition of 100 µM NADPH, maintained at 37 oC for the periods defined. Reactions
71
were terminated with 75 µl of 80% acetonitrile and 20% 0.5 M Tris solution or 2 N NaOH
in the case of dibenzylfluorescein (DBF). Product formation was linear for all incubation
times. Concentrations of the probe substrates in all reaction mixtures were chosen near
the Michaelis-Menten’s (Km) value for each of the CYPs tested. The Km values obtained
using the alkoxyresorufins and all other substrates were within the range of reported
literature values [28]. All measurements were performed in triplicate. Metabolite
formation was measured spectrophotometrically on a Victor2 1420 multilabel counter.
2.3.2. Diclofenac hydroxylation
For the CYP2C9 inhibition assay, reaction mixtures in 500 µl total volume consisted of
49 nM enzyme, 100 µM NADPH, 0.1 M potassium phosphate buffer (pH 7.4), 6 µM
diclofenac and inhibitor. Screening for inhibitory effects of 300 µM concentrations of
curcumin, its decomposition mixture and the individual decomposition products on the
CYP2C9 was done. For IC50 determinations, curcumin concentrations used were of the
range 0.4 to 100 µM and 3.9 to 2000 µM for the individual decomposition products,
vanillin, vanillic acid, ferulic aldehyde and ferulic acid. After preincubation for 5 min at 37
oC, reactions were initiated by adding NADPH and terminated after 10 min with the
addition of 200 !l methanol. The reaction mixtures were centrifuged at 14,000 rpm for 3
min. Product formed was measured using an isocratic HPLC method [29], and was
linear in 10 min. A C18 column (150 mm x 3.2 mm, 5 µm particle size, Phenomenex)
was used and the carrier flow rate was 0.6 ml/min. The mobile phase consisted of 60%
(v/v) 20 mM potassium phosphate buffer (pH 7.4), 22.5% (v/v) methanol and 17.5%
(v/v) acetonitrile. Peaks were monitored at the wavelength of 280 nm. Retention times
for 4-hydroxydiclofenac and diclofenac were 5.0 and 24.1 min, respectively.
2.3.3. Dextromethorphan O-demethylation
Inhibition of CYP2D6 activity by curcumin and its decomposition products was
evaluated by the method described by Ko et al [30]. Inhibitory effects of 300 µM
concentrations of curcumin, its decomposition mixture and the individual decomposition
products on the CYP2D6 were first assessed. The reaction mixture had a total volume
of 500 µl and consisted of 18.2 nM enzyme, 4.5 µM dextromethorphan, 90.9 µM
72
NADPH, 0.1 M potassium phosphate buffer and inhibitor. For IC50 determination,
curcumin concentration range used was 0.4 to 181.8 µM and the decomposition
products 3.6 to 1818.2 µM. Reactions were initiated by the addition of NADPH and
allowed to proceed for 45 min before termination with the addition of 60 mM zinc
sulphate solution. Product formed was measured using an isocratic HPLC fluorescence
detection method and a C18 column (100 mm x 3 mm, 5 µm. particle size, Chromspher)
and was linear in 45 min. The mobile phase consisted of 24% (v/v) acetonitrile and
0.1% (v/v) triethylamine adjusted to pH 3 with perchloric acid. The carrier flow rate was
0.6 ml/min. Peaks were monitored at 280 nm (excitation) and 310 nm (emission). The
retention times of dextrorphan and dextromethorphan were 3.4 and 24.5 min,
respectively.
2.4. Type of inhibition
To determine the types of inhibition occurring in the reactions involving CYP1A2 and
CYP3A4, the substrate concentrations used were ranging from 0.65 to 10 µM for both
methoxyresorufin and benzyloxyresorufin. In the case of the other CYPs, substrate
concentration ranges were from 3.1 to 50 µM benzyloxyresorufin (CYP2B6), 0.3 to 20
µM diclofenac (CYP2C9) and 1.4 to 22.7 µM dextromethorphan (CYP2D6). Five
different substrate concentrations were used in each assay. The concentration of
curcumin used for the assays are indicated in Table 3. Reactions were carried out as
described above for all CYPs.
2.5. Mechanism-based inhibition
The potential of curcumin for mechanism-based inhibition of CYP1A2, CYP3A4 and
CYP2B6 was evaluated according to the method of Heydari et al [31] with slight
modifications. Briefly preincubation mixtures of total volumes 600 µl contained 13 to 16
nM CYP enzymes, 100 µM NADPH and curcumin solution (0, 10 and 50 µM). The
preincubations were performed for 20 min at 37 oC, and at 5 min intervals 100 µl
aliquots were taken to determine the remaining CYP activity. The aliquots of
preincubation mixtures were added to tubes containing 400 µl of the respective
substrates (5 µM methoxyresorufin for CYP1A2, 5 µM benzyloxyresorufin for CYP3A4
73
and 20 µM benzyloxyresorufin for CYP2B6) and NADPH (100 µM), and incubated as
described above. After stopping the reactions the remaining activities were determined
using a fluorescence spectrophotometer (Perkin Elmer Model 3000). Duplicate
experiments were performed.
For CYP2D6 and CYP2C9, preincubation tubes contained 40-49 nM CYP
enzyme, 100 µM NADPH, 50 µM and 20 µM curcumin solution. Preincubations were
performed for 20 min at 37 oC and at time intervals of 5 min the remaining activities of
the enzymes were re-assessed. Aliquots (100 µl) of the 600 µl preincubation mixtures
were transferred into tubes containing 400µl dextromethorphan (4.5 µM) or diclofenac
(6.0 µM) and NADPH (100 µM) and incubated at 37 oC for 30 min or 10 min to
determine remaining activity of CYP2D6 or CYP2C9, respectively.
2.6. Data analysis
Percent inhibition of CYP activity by curcumin, its decomposition products were
calculated from the ratio of the activity of treated to control samples. Statistical analysis
was performed using the Student t test. The enzyme kinetic parameters (Km and Vmax)
for metabolism of the various substrates and IC50 values were analyzed using
GraphPad Prism 4.0 version (GraphPad Prism software Inc. San Diego CA). The
inhibitor constant (Ki) values for competitive inhibition were calculated from according to
the following equation: for competitive inhibition, Ki = Km (inhibited) [I]/ Km (uninhibited) -
Km(inhibited) where substrate concentration is equal to the Km, and for non-competitive
inhibition, K i = Vmax (inhibited) [I]/ Vmax (uninhibited) - Vmax (inhibited) where K m, Vmax, S and [I] are
Michaelis constant, maximal enzyme activity, substrate concentration and inhibitor
concentration, respectively.
3. Results
3.1. Decomposition of curcumin
Degradation of curcumin under various pH conditions and the stability of curcumin in
physiological matrices have been previously reported [22]. To identify the inhibitory
potentials of the decomposition products of curcumin towards human CYPs,
decomposition experiments were also performed in the present study. Curcumin was
74
treated with 0.1 M phosphate buffer of pH 7.4 at 37 oC for 1 h, according to the
procedure described by Wang et al [22]. Figure 2 shows the HPLC-chromatogram of the
resulting mixture of decomposition products of curcumin.
Figure 2. HPLC chromatogram of decomposition products of curcumin at pH 7.4: V, Vac,
Fac, Fal and major product (vanillin, vanillic acid, ferulic acid, ferulic aldehyde and trans-6-(4’-hydroxy-3-
methoxyphenyl)-2,4-dioxo-5-hexenal with retention time 15.2, 14.3, 17.4, 18.5 and 16.9 min respectively).
Chromatographic peaks observed included a major peak and seven minor peaks, four
of which were identified as vanillin, vanillic acid, ferulic aldehyde and ferulic acid, by co-
eluting with commercially available reference compounds [data not shown]. This major
decomposition product with retention time of 16.9 min most likely represents trans-6-(4'-
hydroxy-3'-methoxyphenyl)-4-dioxo-5-hexenal, as the same procedure of decomposition
was used as that described by Wang et al [22].
3.2. CYP inhibition and types of inhibition
The inhibitory effects of 300 µM concentrations of curcumin, four of the individual
decomposition products and a complete mixture of decomposition products, on the
75
activities of CYP-1A2, -3A4, -2D6, -2C9, and -2B6 are shown in Figure 3. At this
concentration, curcumin appeared to inhibit almost completely the activities of CYP3A4,
CYP2C9, and CYP1A2, and 72% and 69.1% of the activities of CYP2D6 and CYP2B6,
respectively. The complete decomposition mixture and the four decomposition products,
vanillin, vanillic acid, ferulic aldehyde and ferulic acid, showed milder inhibitory effects
on the above CYPs. Percentages of inhibition of CYP activities in the range 1 - 50%
were observed in assays with the decomposition products (Figure 3). The
decomposition mixture caused 57.4% and 74.8% inhibition of CYP3A4 and CYP2C9
respectively. For the decomposition products showing more than 50% inhibition of
enzyme activity at 300 µM, the IC50 values were determined (Table 2).
Figure 3. Inhibition of CYP3A4 (A), CYP1A2 (B), CYP2B6 (C), CYP2C9 (D) and CYP2D6 (E) activities by
curcumin, four of its decomposition products (each at a concentration of 300 µM) and a mixture of all
curcumin decomposition products after incubation for 1 h at 37 oC (pH 7.4). General assay conditions are
described under materials and methods. The charts represent means based on n = 3 for samples
incubated with CYP1A2, CYP3A4, CYP2B6 and n = 2 for CYP2C9 and CYP2D6. The symbol ‘# ‘
represents statistically significant difference (P <0.05) from uninhibited reactions as determined by
Student’s t-test. Cur, curcumin; Van, vanillin; Vac, vanillic acid; Fal, ferulic aldehyde; Fac, ferulic acid.
76
The four decomposition products of curcumin that exhibited inhibition of the CYP
isoforms, appeared to be rather weak inhibitors as shown in table 2. For curcumin, the
inhibition of the CYPs in decreasing order of potency was CYP2C9 > CYP3A4 >
CYP2B6 > CYP1A2 > CYP2D6.
Results obtained on inhibition of CYP3A4 by curcumin, using the substrates 7-
benzyloxyquinoline (BQ) and 7-benzyloxy-4-trifluoromethylcoumarin (BFC) could not be
analyzed due to interference by curcumin at the wavelength for detection of the
metabolites.
Table 2. Concentrations of curcumin and four of its decomposition products required to reduce the
activities of five different human CYPs by 50% (IC50 value, µM)
Enzyme Curcumin Vanillic acid Ferulic aldehyde
CYP1A2 40.0 + 12.7 nd 227.5 + 23.4
CYP3A4 16.3 + 1.7a nd nd
13.9 + 3.4b nd nd
CYP2D6 50.3 + 2.0 nd 537.6 + 34.9
CYP2C9 4.3 + 0.8 250.6 + 17.5 259.3 + 22.8
CYP2B6 24.5 + 0.8 nd nd
All values are the means + standard deviation (S.D.) of at least two experiments as described in the
Methods section. nd, not determined due to low percent inhibition observed (<50%). a, b
Substrates used,
benzyloxyresorufin and DBF respectively.
Enzyme kinetic parameters for all CYPs and the corresponding types of inhibition
with curcumin are show in Table 3. In the MROD (methoxyresorufin O-deethylase, with
CYP1A2) and BROD (benzyloxyresorufin O-debenzylase, with CYP3A4, CYP2B6)
assays, the presence of curcumin resulted in an increase of the respective Km values,
whilst the Vmax values did not change significantly when compared to the control
experiment, thus indicating competitive inhibition according to Michaelis-Menten’s
kinetics. However, with respect to CYP2D6 and CYP2C9 curcumin caused significant
decreases in Vmax values with no significant changes in the Km values, thus indicating
non-competitive inhibition.
77
3.3. Mechanism-based inhibition
The effects of pre-incubation of curcumin with NADPH-fortified CYP isozymes on
inhibition potency was also evaluated with all five human CYPs. Pre-incubation of
curcumin for 20 min with NADPH-supplemented CYP did not increase inhibition of
activities by curcumin in any of the CYP isoforms (data not shown). Therefore,
mechanism-based inhibition does not seem to occur with any of the CYPs studied.
Table 3. Enzyme kinetic parameters of the effects of curcumin on human CYP activities
Enzyme Curcumin Vmax Km Ki Type of
(µM) (nM/min/nM) (µM) (µM) inhibition
CYP1A2 0.0 2.43 + 0.51 5.0 + 1.6 43.3 + 10.9
25.0 2.27 + 0.04 7.8 + 0.8 competitive
CYP3A4* 0.0 0.21 + 0.02 2.8 + 0.2 7.4 + 3.5
2.5 0.16 + 0.03 3.8 + 0.1 competitive
CYP2D6 0.0 1.68 + 0.01 3.2 + 0.1 51.0 + 3.9
45.5 0.89 + 0.04 3.3 + 0.1 Non-competitive
CYP2C9 0.0 7.83 + 0.05 6.1 + 1.2 11.5 + 0.8
6.0 5.14 + 0.07 6.4 + 0.2 Non-competitive
CYP2B6 0.0 0.13 + 0.04 34.0 +10.4 33.2 + 14.0
50.0 0.11 + 0.01 69.0 + 12.8 competitive
Values are means + S.D. of at least two experiments as described in the Methods section. Curcumin
concentrations used in the experiments are indicated in the Table. *Substrate use in CYP3A4 inhibition
assay was benzyloxyresorufin.
4. Discussion
The purpose of this study was to evaluate the inhibitory potential of curcumin and its
decomposition products on the five important human drug-metabolizing CYPs, namely
CYP1A2, CYP3A4, CYP2B6, CYP2C9 and CYP2D6. Earlier reports on the inhibition of
rat liver microsomal CYPs by curcumin showed that curcumin is a strong inhibitor of
CYP1A and CYP2B [15,16]. In the present study, curcumin and its decomposition
78
products were first screened at a high concentration of 300 µM, for their inhibitory
potential towards the five CYPs used. At the high concentration, compounds exhibiting
less than 50% inhibitory activities were excluded from further tests with the particular
CYPs. Curcumin was found to possess higher inhibitory potentials against human
CYP1A2, CYP2B6, CYP3A4, CYP2D6 and CYP2C9 than the decomposition products.
However, in contrast to CYP inhibition data on rat mentioned above, curcumin is a less
potent inhibitor of human CYP1A2 and CYP2B6. These results support findings that
animal data may be poorly predictive of the human situation [19]. A recent study on
inhibitory activities of Indonesian medicinal plants, showed plant extracts of three
curcuma species possessing 65 to 73% inhibitory activities towards CYP3A4 and 30 to
54% inhibition towards CYP2D6 [32]. Although these activities are due to the whole
extracts with components including curcumin, they lend support to our findings that
curcumin significantly inhibits CYP3A4 but less significantly inhibits CYP2D6. Curcumin
inhibited CYP1A2, CYP2B6 and CYP3A4 competitively, while non-competitive inhibition
was observed with CYP2C9 and CYP2D6. The structural difference in active sites of the
enzymes used may have contributed to the two different types of inhibition observed.
The insignificant inhibitory activities of the decomposition products towards the tested
CYPs, clearly indicate that the decomposition products of curcumin are not likely to
cause drug-drug interaction at the level of major drug-metabolizing CYPs. Moreover,
decomposition of curcumin is not likely to occur significantly at the low pH in the gut, in
addition to the presence of the stabilizing factors such as enzymes and GSH [15].
The inhibitory potential of curcumin towards CYP3A4 (IC50 = 16.3 µM and Ki =
7.4 µM) could have implications for drug-drug interactions in the intestines because of
the direct exposure of the intestines to curcumin upon oral administration and the high
levels of CYP3A4 in the intestinal epithelial cells. Inhibition of CYP3A4 in the intestines
upon co-administration of drugs could result in a significantly increased bioavailability of
drugs, and consequently increased plasma concentrations of drugs, with the potential
result of adverse drug reactions. Inhibition of CYP3A4 by co-administered drugs has
been shown to result in adverse clinical drug-drug interactions, including fatalities [3].
For example, concomitant intake of grapefruit juice with drugs has also been shown to
increase plasma concentrations of many drugs in humans [33]. This effect appears to
79
be mediated mainly by the inhibition of CYP3A4 in the intestinal wall. It is worth noting
that inhibitors of CYP3A4 are often also known to inhibit P-glycoprotein (P-gp) function
[34]; both phenomena have been suggested to synergistically influence bioavailability of
orally administered agents. Inhibition of P-gp by curcumin at an effective concentration
of 15 µM has also been documented [35]. It is anticipated that inhibition of CYP3A4 and
P-gp by curcumin may be advantageous in mitigating first pass elimination of orally
administered drugs [36].
It is still unknown whether inhibition of the activities of the presently tested human
CYPs in the liver may cause significant systemic drug-drug interactions. As yet, there
are no reports on interaction between curcumin and drugs at the level of hepatic CYPs
in humans. Currently available human pharmacokinetic data show an extremely low
exposure of the liver to curcumin even at very high doses. Plasma concentrations of
curcumin and its metabolites in humans were found to be in nanomolar ranges. High
concentrations of curcumin were found in the faeces [14]. This implies that the inhibitory
effect of curcumin on activities of the CYPs in the liver may be insignificant. The
inhibition parameters determined in the present study, including IC50 and Ki values,
which are important to estimate the CYP-inhibitory potential of curcumin [37] are
relatively high (in micromolar ranges) compared to the anticipated amounts of curcumin
in the liver. However, in recent rodent studies it was demonstrated that repeated oral
administration of curcumin resulted in a 2-fold upregulation of both CYP3A and P-gp in
the liver, whereas a downregulation of these proteins was observed in the intestines
[17]. The significant increase in the area-under-curve and decreases of oral clearances
of midazolam and celiprolol, suggest that the effects on the intestinal activities was
more significant than those on the hepatic activities.
In conclusion, curcumin appears to inhibit five of the important human CYPs, with
the increasing order of potency as CYP2D6 < CYP2B6 < CYP1A2 < CYP3A4 <
CYP2C9. Our results suggest that inhibition of CYP3A4, and to a lesser extent
CYP2C9, by curcumin has the potential to cause clinically significant and harmful drug-
drug interactions upon oral co-administration of curcumin and other drugs metabolized
by these CYPs. Further investigation is required to evaluate the in vivo relevance of the
inhibitory activities of curcumin observed in the present in vitro study.
80
Acknowledgement
We thank Ed Groot of the Molecular Toxicology Section and Ben Bruyneel of the
Analytical Chemistry and Spectroscopy Section of the Vrije Universiteit, for their
technical assistance.
References 1. Nadler, E.P., Reblock, K.K., Ford, H.R., Gaines, B.A., 2003. Monoterapy versus multiple therapy for the treatment of perforated appendicitis in children. Surg Infect (Larcmt) 4:327-333. 2. Hemaiswarya, S., Doble, M., 2006. Potential synergism of natural products in the treatment of
cancer. Phytother Res 20:239-249. 3. Honig, P.K., Wortham, D.C., Zamani, K., Conner, D., Mullin, J.C., Cantilena, L.R., 1993. Terfen adine ketoconazole interaction. J Am Med Assoc 269:1513-1518. 4. Pea, F., Furlanut, M., 2001. Pharmacokinetic aspects of treating infections in the intensive care
unit: focus on drug interactions. Clin Pharmacokinet 40:833-868. 5. Zafar, A., Sharif, M.D., 2003. Pharmacokinetics, metabolism, and metabolism of atypical antipsyc- hotics in special populations. Primary care companion J Clin Psychiatry 5:22-25. 6. Ioannides, C., 2002. Pharmacokinetic interactions between herbal remedies and medicinal drugs. Xenobiotica 32:451-478. 7. Obach, R.S., 2000. Inhibition of human cytochrome P450 enzymes by constituents of St. John's
wort, an herbal preparation used in the treatment of depression. J Pharmacol Exp Ther 294:88-95. 8. Leu, T-H., Maa, M-C., 2002. The molecular mechanisms for the antitumorigenic effect of curcumin. Curr Med Chem-Anti-Cancer Agents 2:357-370. 9. Cole, G.M., Morihara, T., Lim, G.P., Yang, F., Begum, A., Frautschy, S.A., 2004. NSAID and 5-23
antioxidant prevention of Alzheimer's disease: Lessons from in vitro and animal models. Annals NY Acad Sci 1035:68-84.
10. Vajragupta, O., Boonchoong, P., Morris, G.M., Olson, A.J., 2005. Active site binding modes of curcumin in HIV-1 protease and integrase. Bioorg Med Chem Lett 15:3364-3368.
11. Reddy, R.C., Vatsaala, P.G., Keshamouni, V.G., Padmanaban, G., Rangarajan, P.N., 2005. Curcumin for malaria therapy. Biochem Biophy Res Comm 326:472-474.
12. Donatus, I.A., Sardjoko, Vermeulen, N.P.E., 1990. Cytotoxic and cytoprotective activities of curcumin. Effect on paracetamol-induced cytotoxicity, lipid peroxidation and glutathione depletion in rat hepatocytes. Biochem Pharmacol 39:1869-1875. 13. Cheng, A.L., Hsu, C.H., Lin, J.K., Hsu, M.M., Ho, Y.F., Shen, T.S., et al., 2001. Phase I clinical trial of curcumin, a chemopreventive agent, in patients with high-risk or primalignant lesions. Anticancer Res 21:2895-2900. 14. Sharma, R.A., Euden, S.A., Platton, S.L., Cooke, D.N., Shafayat, A., Hewitt, H.R., et al., 2004. Phase I clinical trial of oral curcumin: biomarkers of systemic activity and compliance. Clin Cancer Res 10:6847-6854. 15. Oetari, S., Sudibyo, M., Commandeur, J.N.M., Samboedi, R., Vermeulen, N.P.E., 1996. Effects of curcumin on cytochrome P450 and glutathione S-transferase activities in rat liver. Biochem Pharmacol 51:39-45. 16. Thapliyal, R., Maru, G.B., 2001. Inhibition of cytochrome P450 isozymes by curcumins in vitro and in vivo. Food Chem Toxicol 39:541-547. 17. Zhang, W., Tan, T.M.C., Lim, L.-Y., 2007. Impact of curcumin-induced changes in P-gp and CYP3A4 expression on the pharmacokinetics of peroral celiprolol and midazolam in rats. Drug Metab Dispos 35:110-115. 18. Mori, Y., Tatematsu, K., Koide, A., Sugie, S., Tanaka, T., Mori, H., 2006. Modification by curcumin of mutagenic activation of carcinogenic N-nitrosamines by extrahepatic cytochromes P-450 2B1 and 2E1 in rats. Cancer Sci 97:896-904.
81
19. Eagling, V.A., Tjia, J.F., Back, D.J., 1998. Differential selectivity of cytochrome P450 inhibitors against probe substrates in human and rat liver microsomes. Br J Clin Pharmacol 45:107-114. 20. Guengerich, F.P., 1997. Comparisons of catalytic selectivity of cytochrome P450 subfamily enzymes from different species. Chem Biol Interact 106:161-182. 21. Shimada, T., Yamazaki, H., Mimura, M., Inui, Y., Guengerich, F.P., 1994. Interindividual variations
in human liver cytochrome P-450 enzymes involved in the oxidation of drugs, carcinogens and toxic chemicals: studies with liver microsomes of 30 Japanese and 30 Caucasians. J Pharmacol Exp Ther 270:414-423.
22. Wang, Y.J., Pan, M.H., Cheng, A.L., Lin, L.I., Ho, Y.S., Hsieh, C.Y., Lin, J.K., 1997. Stability of curcumin in buffer solutions and characterization of its degradation products. J Pharm Biomed Anal 15:1867-1876.
23. Burke, M.D., Thompson, S., Elcombe, C.R., Halpert, J., Haaparanta, T., Mayer, R.T., 1985. Ethoxy-, pentoxy- and benzyloxyphenoxazones and homologues: a series of substrates to distinguish between different induced cytochromes P-450. Biochem Pharmacol 34:3337-3345. 24. Omura, T., Sato, R., 1964. The carbon monoxide binding pigment of liver microsomes. I. solubilization, purification and properties. J Biol Chem 239:2379-2385. 25. Stresser, D.M., Blanchard, A.P., Turner, S.D., Erve, J.C.L., Dandeneau, A.A., Miller, V.P., et al., 2000. Substrate-dependent modulation of CYP3A4 catalytic activity: Analysis of 27 test compounds with 4 fluorometric substrates. Drug Metab Dispos 28:1440-1448. 26. Hickman D., Wang J.P., Unadkat J.D., 1997. Evaluation of the selected in vitro probes and suitability of organic solvents for the measurement of human cytochrome P450 monooxygenase activities. Drug Metab Dispos 26:207-215. 27. Busby, W.F., Ackermann, J.M., Crespi C.L., 1998. Effects of methanol, ethanol, dimethyl sulfoxide, and acetonitrile on cDNA-expressed human cytochromes P-450. Drug Metab Dispos 27:246- 249. 28. Staskal, D.F., Diliberto, J.J., De Vito, M.J., Bimbaum, 2005. Inhibition of human and rat CYP1A2 by TCDD and dioxin-like chemicals. Toxicol Sci 84:221. 29. Walsky, R.L., Obach, R.S., 2004. Validation assays for human cytochrome P450 activities. Drug Metab Dispos 32:647-660. 30. Ko, J.W., Desta, Z., Soukhova, N.V., Tracy, T., Flockhart, D.A., 2000. In vitro inhibition of the cytochrome P450 (CYP450) system by the antiplatelet drug ticlopidine: potent effect on CYP2C19 and CYP2D6. Br J Clin Pharmacol 49:343-351. 31. Heydari, A., Yeo, K.R., Lennard, M.S., Ellis, S.W., Tucker, G.T., Rostami-Hodjegan, A., 2004. Mechanism-based inactivation of CYP2D6 by methylenedioxymethamphetamine. Drug Metab Dispos 32:1212-1217. 32. Usia, T., Iwata, H., Hiratsuka, A., Watabe, T., Kadota, S., Tezuka, Y., 2006. CYP3A4 and CYP2D6
inhibitory activities of Indonesian medicinal plants. Phytomedicine 13:67-73. 33. Fuhr, U., 1998. Drug interactions with grapefruit juice: Extent, probable mechanism and clinical
relevance. Drug safety 18:251-272. 34. Wacher, V.J., Silverman, J.A., Zhang, Y., Benet, L.Z., 1998. Role of P-glycoprotein and cytochrome P450 3A in limiting oral absorption of peptides and peptidomimetics. J Pharm Sci 87:1322-1330. 35. Chearwae, W., Anuchapreeda, S., Nandigama, K., Ambudkar, S.V., Limtrakul, P., 2004. Biochemical mechanism of modulation of human P-glycoprotein (ABCB1) by curcumin I, II and III purified from tumeric powder. Biochem Pharmacol 68:2043-2052. 36. Kuppens, I.E., Breedveld, P., Beijnen, J.H., Schellens, J.H., 2005. Modulation of oral drug bioavailability: from preclinical mechanism to therapeutic application. Cancer Invest 23:443- 464. 37. Bapiro, T.E., Egnell, A.C., Hasler, J.A., Masimirembwa, C.M., 2001. Application of higher
throughput screening (HTS) inhibition assays to evaluate the interaction of anti-parasitic drugs with cytochrome P450s. Drug Metab Dispos 29:30-35.
82
83
Chapter 4
Structure-activity relationship of inhibition of recombinant Human Cytochrome
P450 mediated metabolism by Curcumin Analogues
Regina Appiah-Opong, Iwan de Esch, Jan N. M. Commandeur, Mayagustina Andarini
and Nico P. E. Vermeulen
Adapted from European Journal of Medicinal Chemistry 2008 43:1621-1631
Inhibition of cytochrome P450 (CYP) is a major cause of drug-drug interactions. In this
work, inhibitory potentials of thirty-three curcumin analogues, i.e. 2,6-
dibenzylidenecyclohexanone (A series), 2,5-dibenzylidenecyclopentanone (B series)
and 1,4-pentadiene-3-one (C series) substituted analogues of curcumin towards
recombinant human CYP1A2, CYP3A4, CYP2B6, CYP2C9 and CYP2D6, all important
for drug metabolism, were studied in vitro. Fluorescence plate reader and high
performance liquid chromatography (HPLC) assays were used to evaluate CYP
inhibitory activities. MOE-based Quantitative structure–activity relationship (QSAR)
analysis suggested that electrostatic and hydrophobic interactions and lipophilicity are
important factors for CYP inhibition. Apart from insights in important molecular
properties for CYP inhibition, the present results may also guide further design of
curcumin analogues with less susceptibility to drug-drug interactions.
84
1. Introduction
Curcumin is a well-known food additive and constituent of traditional medicine in
Southeast Asia and the Indian subcontinent, the latter being an area with low incidence
of colorectal cancer [1]. This naturally occurring and synthetic compound is regarded as
a promising drug and has received considerable attention due to its antioxidant,
anticancer, anti-inflammatory, anti-HIV and anti-malarial properties [2-6]. Reports from a
phase 1 clinical trial on curcumin have shown that it is non-toxic even at doses as high
as 8 g/day [3]. Curcumin is unstable at a pH 7.4, however, this stability is strongly
improved by lowering the pH or by adding glutathione (GSH), N-acetyl L-cysteine
(NAC), ascorbic acid or rat liver microsomes or cytosol [7]. The instability of curcumin at
neutral to basic pH conditions has been attributed to the presence of an active
methylene group (figure 1) [8].
Figure 1. Chemical structure of curcumin. The active methylene group is indicated by the arrow.
Omitting this methylene group leads to the formation of more stable and potent
antioxidative compounds [8]. Removal of the active methylene group and one carbonyl
group led to 1,4-pentadiene-3-ones, which still possess antioxidant properties [9].
Modification of groups on the terminal aromatic rings of curcumin has revealed that the
electron-donating substituents increase the anti-inflammatory activity [10].
Previously a series of curcumin analogues were synthesized, in which the
methylene and one carbonyl group have been omitted [9]. These analogues are
derivatives of benzylidine, having either electron-withdrawing, electron-donating or
steric groups. The derivatives include nine compounds of 2,6-
dibenzylidenecyclohexanone (A), thirteen of 2,5-dibenzylidenecyclopentanone (B) and
eleven of 1,5-diphenyl-1,4-pentadiene-3-one (C) (Schemes 1-3).
CH3CH
3
OH
OH
O
O
OH
O CH3CH
3
OH
OH
O
O
OH
O
85
These analogues exhibit antioxidant, anti-inflammatory, and antibacterial activities that
render them potential drug candidates. Interestingly, some of the analogues have
shown much stronger antioxidant activities than curcumin [9].
Scheme 1
33
1
2
O
R
R
R R
R
R
2
1
Cyclohexanones (A)
Scheme 2
1
33
R1
R
R
O
R
R
R
22
Cyclopentanone (B)
R1 R2 R3
A8 H N(CH3)2 H
A10 Cl H H
A11 CH3 OH CH3
A14 t-C4H9 OH t-C4H9
R1 R2 R3
A0 H OH H
A2 H H H
A4 H OCH3 H
A5 H CH3 H
A7 H CF3 H
R1 R2 R3
B0 H OH H
B2 H H H
B1 OCH3 OH H
B3 H Cl H
B4 H OCH3 H
B7 H CF3 H
R1 R2 R3
B10 Cl H H
B11 CH3 OH CH3
B12 C2H5 OH C2H5
B13 i-C3H7 OH i-C3H7
B14 t-C4H9 OH t-C4H9
B15 OCH3 OH OCH3
86
Scheme 3
1
33
R1
R
R
O
R
R
R
22
1,4-pentadiene-3-ones (C)
Schemes 1-3 show the chemical structures of the synthetic curcumin analogues.
Drug-drug interactions due to inhibition or induction of enzyme activity are among
the major causes of attritions in drug development [11]. Inhibition of CYP enzymes is a
cause of clinically significant drug-drug interactions [12]. Irrespective of the mechanism,
CYP inhibition may result in accumulation of drugs resulting in adverse drug reactions
or a decrease in metabolism of drugs and activation of pro-drugs, and hence alter their
pharmacokinetic profile. Therefore, in vitro CYP-associated inhibition studies for
evaluation of drug candidates during the early stages of drug discovery and
development are considered cost-effective for predicting potential clinical drug-drug
interactions, since these interactions may result in adverse drug reactions and
therapeutic failure [13,14].
Recently the inhibitory effect of curcumin on five major human drug metabolizing
CYPs have been reported [15]. Curcumin showed strong inhibition of CYP2C9 and
CYP3A4, with IC50 values 4.3 and 16.3 µM respectively and moderate inhibition of
CYP1A2, CYP2B6 and CYP2D6 (40.0, 24.5, 50.3 µM respectively). The inhibitory
effects of curcumin analogues on human CYPs have not yet been reported. In this
study, we investigated the inhibitory potentials of thirty-three curcumin analogues to the
five major recombinant human drug-metabolizing CYPs [16,17] mentioned above.
R1 R2 R3
C0 H OH H
C1 OCH3 OH H
C2 H H H
C3 H Cl H
C5 H CH3 H
C6 H t-C4H9 H
R1 R2 R3
C7 H CF3 H
C9 Cl Cl H
C10 Cl H H
C11 CH3 OH CH3
C15 OCH3 OH OCH3
87
Inhibitory structure-activity relationships (SARs) were subsequently evaluated and
quantitative structure-activity relationships (QSARs) were investigated using the
program MOE (Molecular Operating Environment). In these studies we tried to identify
the molecular features that cause inhibition of the different CYP isoenzymes tested.
These studies are important because the resulting information will guide design the
synthesis of new curcumin analogues with less CYP inhibitory properties and open a
way for in silico prediction of xenobiotic inhibition of CYPs.
2. Materials and methods
2.1. Materials
Methoxyresorufin (MRes) and benzyloxyresorufin (BRes) were synthesized by the
method of Burke et al [18], and the purity was determined by HPLC, mass spectrometry
and 1H NMR. The plasmid, pSP19T7LT_2D6 containing human CYP2D6 bicistronically
co-expressed with human cytochrome P450 NADPH reductase was kindly provided by
Prof. M. Ingelman-Sundberg (Stockholm, Sweden). The plasmids, BMX100/h1A2 and
pCWh3A4 with human cytochrome P450 NADPH reductase were kindly donated by Dr.
M. Kranendonk (Lisbon, Portugal). Expression plasmids, pCWh2B6hNPR and
pCWh2C9hNPR with human cytochrome P450 NADPH reductase were kindly provided
by Prof. F.P. Guengerich (Nashville, Texas, USA). Curcumin analogues were kindly
donated by Dr. S. Sardjiman (Jakarta, Indonesia). All other chemicals were of analytical
grade and obtained from standard suppliers.
2.2. CYP expression and membrane isolation
The plasmids containing cDNA of five human CYPs were transformed into Escherichia
coli strain JM109. Expression of the CYPs was carried out in 3-litre flasks containing
300 ml terrific broth (TB) medium, with 1mM !-aminolevulinic acid, 0.5 mM thiamine,
400 µl/L trace elements, 100 µg/ml ampicillin, 1 mM isopropyl-"-D-
thiogalactopyranoside (IPTG), 0.5 mM FeCl3 (for CYP2D6 and CYP3A4 only), 1 mg/L
chloramphenicol (for CYP2B6 only) and 30 µg/ml kanamycin (for CYP3A4 only). The
culture media were inoculated with 3 ml overnight cultures of bacteria containing
plasmids for the various CYPs. The cell cultures were incubated for about 40 h at 28 oC
88
and 125 rpm, and CYP contents were determined using the carbon monoxide (CO)
difference spectra as described by Omura and Sato [19]. Cells were pelleted by
centrifugation (4000 g, 4 oC, 15 min) and resuspended in 30 ml Tris-Sucrose-EDTA
(TSE) buffer (50 mM Tris-acetate buffer pH 7.6, 250 mM sucrose, 0.25 mM EDTA).
Cells were treated with 0.5 mg/ml lysozyme prior to disruption by French press (1000
psi, 3 repeats). The membranes containing the human CYPs were isolated by
ultracentrifugation in a Beckmann 50.2Ti rotor (60 min, 40,000 rpm, 4 oC), resuspended
in TSE buffer and stored at –80 oC until use.
2.3. Stability of curcumin analogues in buffer
Decomposition of curcumin analogues was investigated as described [7], in 0.1 M
potassium phosphate buffer of pH 7.4. Solutions of 25µM curcumin analogues in buffer
(in 0.5% DMSO) were scanned every 5 min between 200-600 nm for 30 min using an
Ultrospec 2000 Pharmacia Biotech UV/visible spectrophotometer. The above
experiment was also performed in the presence of 1 mM GSH and 13.2 nM enzyme
(CYP) as described [7], to determine the effects of these factors on the stability of the
analogues in buffer.
2.4. CYP inhibition assays
2.4.1. 7-Methoxy-, 7-Benzyloxyresorufin and O-dealkylation
Inhibition of the activities of human CYP isoforms 1A2, 3A4 and 2B6, by curcumin
analogues was determined by microplate reader assays using fluorescent substrates.
Incubation conditions (eg. enzyme concentration, substrates, incubation time) and
wavelengths for detection for each of the inhibition assays are shown in Table 1. In
general, the microsomal incubations were carried out in a total volume of 200 µl, and in
the presence of 100 µM NADPH (freshly prepared) in a black coaster 96-well plate.
Membranes were pre-incubated for 5 min at 37 oC with 0.1 M potassium phosphate
buffer (pH 7.4), substrates, and inhibitors (curcumin analogues) with minimal use of
DMSO, i.e. always 1.0 % (v/v) or less.
89
Table 1. Experimental conditions for fluorescence CYP assays
CYP Enzyme Incubation Substrate Substrate Excitation Emission
amount time concn wavelength wavelength
nM min µM nm
1A2 13.2 10 MRes 5.0 530 586
3A4 14.3 30 BRes 5.0 530 586
2B6 15.3 30 BRes 20.0 530 586
MRes, methoxyresorufin; BRes, benzyloxyresorufin
Experiments were performed in the absence and presence of GSH, to determine the
influence of a second curcumin stabilizing factor on the experiments. Subsequently, the
analogues (100 µM each) were all screened for CYP inhibitory activity. The screening
procedure for all CYPs involved pooling compounds into groups of three or four and
testing for CYP inhibition of the mixture. Groups showing >20% inhibition were selected
for further screening, for identification of potential CYP inhibitors.
Determination of IC50 values for curcumin analogues showing over 20% inhibition
of CYP activities was performed. The concentration range of the curcumin analogues
used was from 0.195 to 100 µM. Incubations were started by the addition of 100 µM
NADPH, and maintained at 37 oC for the periods defined (Table 1). Reactions were
terminated with 75 µl of 80% acetonitrile and 20% 0.5 M Tris solution. Product formation
was linear for all incubation times. Concentrations of the probe substrates in all reaction
mixtures were chosen near the Michaelis-Menten’s constant (Km) value for each of the
CYPs tested. The Km values obtained using the alkoxyresorufins and all other
substrates were within the range of reported literature values [20]. All measurements
were performed in triplicate. Metabolite formation was measured spectrophotometrically
on a Victor2 1420 multilabel counter.
2.4.2. Diclofenac hydroxylation
For the CYP2C9 inhibition assay, reaction mixtures in 500 µl total volume consisted of
49 nM enzyme, 100 µM NADPH, 0.1 M potassium phosphate buffer (pH 7.4), 6 µM
diclofenac and inhibitor. Hundred micromolar of each of the analogues was screened for
90
CYP2C9 inhibitory activity. For IC50 determinations on curcumin analogues with >20%
inhibition, concentrations of analogues used were of the range 0.39 – 200 µM. After
preincubation for 5 min at 37 oC, reactions were initiated by adding NADPH and
terminated after 10 min upon the addition of 200 !l methanol. The reaction mixtures
were centrifuged at 14,000 rpm for 3 min. Product formed was measured using an
isocratic HPLC method [21]. A C18 column (150 mm x 3.2 mm, 5 µm particle size,
Phenomenex) was used and the carrier flow rate was 0.6 ml/min. The mobile phase
consisted of 60% (v/v) 20 mM potassium phosphate buffer (pH 7.4), 22.5% (v/v)
methanol and 17.5% (v/v) acetonitrile. Peaks were monitored at the wavelength of 280
nm. Retention times for 4-hydroxydiclofenac and diclofenac were 5.0 and 24.1 min,
respectively.
2.4.3. Dextromethorphan O-demethylation
Inhibition of CYP2D6 activity by curcumin and its decomposition products was
evaluated by a method described [22]. The reaction mixture had a total volume of 500 µl
and consisted of 18.2 nM enzyme, 4.5 µM dextromethorphan, 90.9 µM NADPH, 0.1 M
potassium phosphate buffer and curcumin analogues. Hundred micromolar of each of
the analogues was screened for CYP2D6 inhibitory activity, and IC50 was determined for
analogues showing >20% inhibition (concentration range 0.39-200 µM). Reactions were
initiated by the addition of NADPH and allowed to proceed for 45min before termination
with the addition of 60 mM zinc sulphate solution. Product formed was measured using
an isocratic HPLC fluorescence detection method and a C18 column (100 mm x 3 mm,
5 µm particle size, Chromspher). The mobile phase consisted of 24% (v/v) acetonitrile
and 0.1% (v/v) triethylamine adjusted to pH 3 with perchloric acid. The carrier flow rate
was 0.6 ml/min. Peaks were monitored at 280 nm (excitation) and 310 nm (emission).
The retention times of dextrorphan and dextromethorphan were 3.4 and 24.5 min,
respectively.
2.4.4. SAR and QSAR analysis
Percent inhibition of CYP activities by the curcumin analogues was calculated from the
ratios of the activities of inhibited to control samples. The IC50 values were calculated
91
using GraphPad Prism 4.0 version (GraphPad Prism software Inc. San Diego CA).
Subsequent Quantitative structure-activity relationship (QSAR) analysis of the
respective CYP-inhibitory activities of the curcumin analogues were performed using the
MOE software (Version 2005.06, Chemical Computing Group Inc, Montreal). Structural
descriptors were obtained from the QuaSAR-Descriptors in MOE, version 2005.06.
Table 2. List of all molecular descriptors used in this study, obtained from QuaSAR-Descriptor
MOE 2005.06 version.
No. Descriptor Description
1 a_acc number of hydrogen bond atoms
2 a_nO number of oxygen atoms
3 E_nb value of potential energy with all bonded terms disabled
4 PEOE_VSA_FHYD Fractional hydrophobic van der Waals surface area
5 PEOE_VSA_FPNEG Fractional negative polar van der Waals surface area
6 PEOE_VSA_FPPOS Fractional positive van der Waals surface area
7 SlogP_VSA0 VSA with sum of surface area vi such that contribution of log of
octanol/ water partition coefficient calculated from the given
structure Li is –0.4
8 SlogP_VSA4 (see 7) Sum of vi such that Li is in 0.1, 0.15
9 SMR_VSA5 Molecular refractivity/VSA, sum of vi such that molecular
refractivity from atom I, Ri is in 0.44, 0.485
10 Std_dim2 square root of second largest own value of covariance matrix of
the atomic coordinates
11 Std_dim3 square root of the third largest own value of covariance matrix of
the atomic coordinate
12 TPSA total polar surface area
13 VdistEq sum of log2m-pi/m where m and pi are sum and number of
distance matrix entries respectively
14 Weight molecular weight
92
One hundred and thirty seven descriptors, including both 2D and 3D molecular
descriptors were calculated using the MOE program. Structure-activity models were
generated by multiple stepwise regression analysis (MRA) of the biological and
structural variables using the statistical analysis software SPSS for Windows version 13.
Fischer coefficients (F values) were also calculated using the latter. Table 2 provides
information on MOE-selected descriptors used in this study.
3. Results
3.1. Stability of curcumin analogues in buffer
The stability of curcumin analogues (25 µM) in 0.1 M phosphate buffer was measured
spectrophotometrically at 200-600 nm. At pH 7.4 the curcumin analogues decomposed
to various extents after 30 min of incubation. Table 3 shows the percent decomposition
of the analogues.
Table 3. Percent decomposition of curcumin analogues in buffer (pH 7.4)
Cmpd Decomp (%) Cmpd Decomp (%) Cmpd Decomp(%)
A0 17.5 B3 12.0 C1 nd
A2 97.5 B4 30.5 C2 35.0
A4 54.5 B7 25.5 C3 46.0
A5 65.5 B8 17.0 C6 73.0
A7 74.0 B9 18.5 C7 82.5
A8 35.5 B10 40.5 C9 25.0
A10 63.0 B11 7.0 C10 46.0
A11 13.0 B12 25.5 C11 7.5
A14 57.5 B13 41.5 C15 6.5
B0 13.5 B14 3.0 Curcumin 74.0
B1 7.4 B15 nd
B2 13.0 C0 53.0
Cmpd, compound; Decomp, decomposition; nd, not determined. Concentration of each curcumin
analogues used was 25 !M. The experiment performed is described in the Methods section.
93
Generally, compounds of group B (i.e. 2,5-dibenzylidenecyclopentanones) were more
stable at pH 7.4 than those in groups A and C (2,6-dibenzylidenecyclohexanones and
1,5-diphenyl-1,4-pentadiene-3-ones). Group A, demonstrated the greatest instability,
with A2 having the highest (97.5%) and B14 the lowest (3.0%) percent decomposition.
The degradation of the compounds in buffer pH 7.4 was significantly blocked in the
presence of CYP enzyme and GSH resulting in 8.2% and 19% decomposition
respectively. These effects are similar to those reported on the stability of curcumin in
buffer, by Oetari et al [7].
3.2. CYP inhibition
All the thirty-three curcumin analogues (each at 100 µM concentration) were screened for
inhibitory potentials towards human CYP1A2, CYP3A4, CYP2B6, CYP2C9 and CYP2D6.
Preceding the inhibitor screening experiments, tests were conducted with and without GSH
[7] to determine whether another stabilization factor was necessary to mitigate
decomposition. The results indicated that inhibition by curcumin analogues in the presence
of GSH, did not influence the inhibitory effect of the compounds on CYP activity (data not
shown). Initial screening results indicated that the curcumin analogues demonstrated a
wide range of inhibitory activities towards CYP-mediated metabolism of probe substrates
(data not shown). Results on 29 subsequently selected compounds (with % inhibition
>20%) revealed 75.8% (22), 27.5% (8), 13.7% (4), 62.0% (18) and 41.3% (12) of the
compounds inhibiting CYP1A2, CYP3A4, CYP2B6, CYP2C9 and CYP2D6, respectively. In
general the compounds showed a comparatively stronger inhibitory potency towards
CYP1A2 and CYP2C9 than towards CYP3A4, CYP2B6 and CYP2D6 activities (Tables 5-
8).
The least inhibited enzyme was CYP2B6, with only four compounds, i.e. B13, B15,
C11 and C15 exhibiting >20% inhibition at 100 µM concentration of compounds, and IC50
values being 74.3, 70.0, 44.8 and 24.8 µM, respectively. Compounds of group A showed
very weak or no inhibitory activities towards CYP3A4, CYP2B6 and CYP2D6.
94
3.3. Similarity studies
Concerning the structures of curcumin analogues, Similarity studies were performed
using the procedure as described by Labute et al [23]. Accordingly, calculations were
carried out using the flexible alignment module of MOE, employing the MMFF94 force
field. The results and best scoring fit both in terms of similarity and objective function
are shown in table 4 and figure 2. The respective superimpositions indicated no
significant difference between the A, B and C series of compounds, since they appear
to be perfectly superposed. However, differences may also arise due to steric bulk.
Table 4. Results of flexible alignment module of MOE
F S dU
A0-B0 119.6379 177.1833 1.7428
A0-C0 119.5509 171.6317 0.0000
B0-C0 119.0929 168.9706 0.0000
F, similarity; S, objective function; dU, potential energy difference between the lowest minimal energy
conformation and the global minimal (kcal/mol).
In the conformation A0-B0, the potential energies of A0 and B0 were 5.7 and 3.9
kcal/mol, above the respective global minima. For the conformation A0-C0, the potential
energy of A0 was 3.7 kcal above, and that of C0 was exactly at the global minimum.
The conformation B0-C0 resulted in potential energies of B0 and C0, 0.6 and 0.4
kcal/mol respectively above the global minima.
95
A B
C
Figure 2. Flexible alignment of curcumin analogues A0 (black), B0 (grey) (A), A0 (black), C0 (grey) (B)
and B0 (black), C0 (grey) (C).
3.4. Quantitative structure-activity relationships (QSARs)
Summaries of the relevant datasets employed for generating the QSARs relating the
various molecular descriptors to the CYP inhibitory potencies of curcumin analogues
used in this work are shown in tables 5-8.
Table 5 shows the data for twenty curcumin analogues that exhibited inhibitory
activities towards CYP1A2, and five relatively important descriptors, i.e. standard
dimension 3, the standard deviation along a principal component axis (std_dim3), polar
surface area (TPSA), number of oxygen atoms (a_nO), number of hydrogen bond
atoms (a_acc) and potential energy with bonded terms disabled (E_nb). A weak
correlation (R2 = 0.682) was found between experimental and predicted IC50 data (Fig.
3A) based on these molecular descriptors.
96
R2
= 0.682
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.5 1 1.5 2 2.5 3 3.5
Experimental log1/IC50 (1/uM)
Pre
dic
ted
lo
g1
/IC
50
(1
/uM
)
R2
= 0.804
0.0
0.5
1.0
1.5
2.0
2.5
0.5 1 1.5 2 2.5
Experimental log1/IC50 (1/uM)
Pre
dic
ted
lo
g1
/IC
50
(1
/uM
)
R2
= 0.73
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.5 1 1.5 2 2.5 3 3.5
Experimental log1/IC50 (1/uM)
Pre
dic
ted
lo
g1
/IC
50
(1
/uM
)
R2
= 0.522
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.5 1 1.5 2 2.5 3 3.5
Experimental log1/IC50 (1/uM)
Pre
dic
ted
lo
g1
/IC
50
(1
/uM
)
Figure 3. Plots of experimental and predicted CYP inhibitory activities (Log1/IC50). Equations for these
plots are shown under Tables 3-6 (CYP1A2, CYP3A4, CYP2C9 and CYP2D6 respectively) and the
compounds involved are listed in the tables.
However, exclusion of four outliers (B1, B13, C3 and C6) resulted in a good correlation
(R2 = 0.907), with the descriptors being an electrostatic parameter
(PEOE_VSA_FPNEG), a lipophilicity feature (SlogP_VSA0), molecular weight (weight),
molecular refractivity (SMR_VSA5) and potential energy with bonded terms disabled
(E_nb).
A CYP1A2
C CYP2C9
B CYP3A4
D CYP2D6
97
Table 5. Dataset for QSARs in CYP1A2 inhibitors
Cmpd IC50 (mM) Log1/IC50Ea Log1/ IC50P
b std_dim3 TPSA a_nO a_acc E_nb
A2 0.0009 3.046 2.334 0.432 17.07 1 1 310.29
A4 0.0110 1.959 1.842 0.822 35.53 3 1 706.87
A5 0.0210 1.678 1.839 0.669 17.07 1 1 746.37
A7 0.0104 1.983 2.273 0.538 17.07 1 1 344.92
A8 0.0026 2.585 2.554 0.491 23.55 1 1 413.26
A10 0.0034 2.469 2.492 0.460 17.07 1 1 303.64
B0 0.0456 1.341 1.744 0.334 57.53 3 3 263.25
B1 0.0389 1.410 1.173 0.961 75.99 3 5 677.95
B2 0.0366 1.437 1.520 0.713 17.07 1 1 1054.22
B12 0.0367 1.435 1.591 0.875 57.53 3 3 280.39
B13 0.0284 1.547 1.101 1.181 57.53 3 3 692.98
B14 0.0428 1.369 1.382 1.018 57.53 3 3 453.52
C1 0.0415 1.382 1.342 1.104 75.99 3 5 473.23
C2 0.0037 2.432 2.168 0.556 17.07 1 1 445.43
C3 0.0266 1.575 2.073 0.900 17.07 1 1 453.90
C6 0.0280 1.553 1.915 1.193 17.07 1 1 538.34
C7 0.0040 2.398 2.010 1.016 17.07 1 1 487.56
C10 0.0034 2.469 2.285 0.077 17.07 1 1 448.75
C11 0.0304 1.517 1.227 1.668 57.53 3 3 444.70
C15 0.0752 1.124 1.181 1.651 94.45 3 7 498.76
QSAR expressions: n = 20, s = 0.356, R2 = 0.682, F = 19.51.
Log1/IC50 = 3.249 – 0.252std_dim3 + 0.052TPSA - 0.479a_nO – 0.904a_acc – 0.001E_nb
a,b, Experimental and predicted
The relevant dataset on seven curcumin analogues inhibiting CYP3A4, employed for
generating QSARs, is found in table 6. A fairly good correlation (R2 = 0.804) was found
between experimentally derived and predicted activities based on the descriptor,
VdistEq, a distance matrix parameter (Fig. 3B).
98
Table 6. Dataset for QSARs in CYP3A4 inhibitors
Cmpd IC50 (mM) Log1/IC50Ea Log1/IC50P
b VDistEq
B0 0.0051 2.292 2.125 3.492
B1 0.0649 1.188 1.473 3.628
B12 0.0735 1.134 1.157 3.694
B13 0.0776 1.110 1.166 3.693
B15 0.0383 1.417 1.157 3.694
C11 0.0403 1.395 1.540 3.614
C15 0.0132 1.119 1.027 3.722
QSAR expressions: n = 7, s = 0.205, R2 = 0.804, F = 20.39.
Log1/IC50 = 18.867 – 4.793VDistEq.
a,b, Experimental and predicted
However, these results are biased due to one outlier (i.e. B0) that appears to influence
the outcome significantly. Exclusion of the outlier resulted in no correlation between
experimental and predicted inhibitory activities. The dataset on 12 compounds inhibiting
CYP2C9 activity is presented in Table 7. Three relatively important descriptors that
appear in the QSAR equation are, fractional polar negative van der Waals surface area
(PEOE_VSA_FPNEG), fractional hydrophobic van der Waals surface area
(PEOE_VSA_FHYD), and log P of accessible van der Waals surface area for each
atom (SlogP_VSA4). Experimental inhibitory activities of the analogues towards
CYP2C9 weakly correlated (R2 = 0.738) with predicted activities based on these
descriptors (Fig. 3C). Elimination of the outlier, B0 resulted in a fairly good correlation
(R2 = 0.836) with the same descriptors.
99
Table 7. Dataset for QSARs in CYP2C9 inhibitors
Cmpd IC50 (mM) Log1/IC50Ea Log1/ IC50P
b P_V_FH P_V_FP SlogP_VSA4
A0 0.0010 3.000 2.599 0.835 0.105 6.371
B0 0.0099 2.004 2.533 0.825 0.103 6.371
B1 0.0028 2.553 2.115 0.843 0.098 6.371
B11 0.0283 1.548 1.216 0.859 0.083 19.113
B12 0.0377 1.424 1.385 0.882 0.069 19.113
B14 0.0550 1.260 1.565 0.918 0.048 19.113
C0 0.0018 2.745 2.544 0.821 0.105 6.371
C1 0.0075 2.125 2.077 0.840 0.100 6.371
C2 0.0673 1.172 1.188 0.947 0.053 6.371
C10 0.0213 1.672 1.493 0.954 0.046 6.371
C11 0.0628 1.202 1.221 0.857 0.084 19.113
C15 0.0590 1.229 1.811 0.853 0.096 6.371
QSAR expressions: n = 12, s = 0.329, R2 = 0.730, F = 11.85.
P_V_FP, PEOE_VSA_FPNEG; P_V_FH, PEOE_VSA_ FHYD
Log1/IC50 = 53.151 – 92.429PEOE_VSA_FPNEG – 48.904PEOE_VSA_ FHYD –0.118SlogP_VSA4.
a,b, Experimental and predicted
Table 8 contains dataset for six curcumin analogues with inhibitory activity towards
CYP2D6. The descriptor present in the QSAR equation and as apparently related to the
Table 8. Dataset for QSARs in CYP2D6 inhibitors
Cmpd IC50 (mM) Log1/IC50Ea Log1/IC50P
b P_V_FP
B14 0.1187 0.926 0.911 0.048
C0 0.0020 2.699 2.497 0.105
C1 0.0006 3.222 2.358 0.100
C2 0.0685 1.164 1.050 0.053
C11 0.0132 1.879 1.913 0.084
C15 0.0827 1.082 2.247 0.096
QSAR expressions: n = 6, s = 0.352, R2 = 0.522, F = 27.88.
Log1/IC50 = -0.782 + 35.382PEOE_VSA_FPNEG.
a,b, Experimental and predicted
100
observed inhibitory activity is PEOE_VSA_FPNEG. A weak correlation (R2 = 0.522) was
found between experimental and predicted data based on the descriptor (Fig. 3D). A
good correlation (R2 = 0.903) was obtained upon removal of outlier C15.
4. Discussion
Inhibition of CYPs can lead to drug-drug interactions and therefore it is considered
important to evaluate potential drug candidates for CYP inhibitory activities. Inhibitory
potentials of curcumin towards recombinant human CYP1A2, CYP3A4, CYP2B6,
CYP2C9 and CYP2D6 have recently been evaluated in vitro [15]. The inhibitory
potencies (IC50 values) towards CYP3A4 and CY2C9 suggested a potential for relevant
drug-drug interaction in the intestine upon oral co-administration with other drugs
metabolized by CYP3A4, considering the required high dose for therapeutic effects.
Less potent activities were observed with CYP1A2, CYP2B6 and CYP2D6. Three
groups of curcumin analogues [9], i.e. nine of 2,6-dibenzylidenecyclohexanone (group
A), thirteen of 2,5-dibenzylidenecyclopentanoe (group B) and eleven of 1,5-diphenyl-
1,4-pentadiene-3-one (group C) were analogously tested experimentally for inhibition
towards five important human drug-metabolizing CYPs mentioned above [15]. QSAR
analysis employing molecular descriptors that have >18% correlations with activity were
used as inputs for artificial neural networks (ANNs) [Ma et al., 2006 http://www.natural-
selection.com/library/2006/NN_antihiv_ligand_gbf.pdf ] and were generated by MOE.
Most of the curcumin analogues exhibited low inhibitory activities towards the
CYPs tested. Six of them, A2, A8, A10, C2, C7 and C10 showed potent inhibitory
activities with IC50 values in the range 0.9 – 4 µM, towards CYP1A2. These compounds
showed about ten to forty fold greater potency towards inhibition of CYP1A2 than
curcumin itself. However, these six compounds have not been reported to have potent
antioxidant activities [9]. The compounds A2 and C2, both lack substituents, in all
investigated substitutions. Similarly, compounds A10 and C10 both have chloride at the
R1, with the R2 and R3 positions unsubstituted (Schemes 1 and 3). The only difference
between series A and C compounds is the absence of a central six-membered ring in
the latter. Therefore it appears that these substitutions together with the presence or
absence of a central six-membered ring favour increased CYP1A2 inhibitory potency of
101
compounds of A and C series. However, compound B2, having the same substituents
as A2 and C2 but a central five-membered ring, showed lower inhibitory potency.
Generally, compounds of the B series showed rather weak inhibitory activities towards
CYP1A2. Apparently the central five-membered ring renders these compounds less
active towards CYP1A2.
The present QSAR analysis suggest that five descriptors influence inhibition of
CYP1A2, being standard dimension 3 (std_dim3), the standard deviation along a
principal component axis, potential energy with bonded terms disabled (E_nb), number
of oxygen atoms (a_nO), number of hydrogen bond acceptor atoms (a_acc) and total
polar surface area (TPSA). A weak correlation (R2 = 0.682) was obtained between the
experimental and predicted inhibitory activities. However, exclusion of four outliers
resulted in a good correlation (R2 = 0.907), with electrostatic and lipophilicity
descriptors, as well as molecular refractivity, molecular weight and potential energy of
bonded terms disabled. The reason for the outliers is not clear, and subject to ongoing
research, that also include the molecular features of the target site.
Earlier QSAR studies have indicated that CYP activities are related to substrate
lipophilicity [24-26], and these results lend some support to that suggestion. Substrates
and inhibitors of CYP1A2 are usually planar small-volume molecules that are neutral or
weakly basic. However, a proposal has been made that the binding pocket of CYP1A2
enzyme is composed of mostly hydrophobic and aromatic amino acids with polar amino
acids for hydrogen bonding being present near the heme centre [27]. Thus it is possible
that hydrogen bonding and hydrophobic interactions contribute to the observed
inhibitory activities. The presence of an ortho- or para-hydroxyl group has been earlier
shown to be relevant for the antioxidant activity of these and other curcumin analogues
[9,28,29]. However, in contrast to these activities the most potent inhibitors of CYP1A2
lack the para-hydroxyl moiety. It is worth noting that CYP1A2 is known to be involved in
the activation of procarcinogens [30] and consequently, that inhibition of CYP1A2 could
be pharmacologically beneficial.
Seven curcumin analogues exhibited IC50 values towards CYP3A4 within the
concentration range used in the experiments. Weak inhibitory activities were obtained,
except in the cases of B0 and C12 where potencies of 5.1 + 4.0 and 13.2 + 3.0 µM were
102
found. These activities are comparable with that determined for curcumin [15]. Five of
these compounds belong to the cyclopentanone (B) group and the remaining two to the
1,4-pentadiene-3-one (C) group. No inhibition of CYP3A4 was observed for compounds
from series A, which suggests that the presence of the central six-membered ring
renders them less active towards CYP3A4. The QSAR equation found in the present
study, contained a distance matrix feature, VdisEq (Vertex distance equation), which
suggests that the observed inhibition of CYP3A4 by the analogues was related to
distance matrices of the compounds. The correlation found (R2 = 0.804) was biased due
to an outlier, resulting in an over-estimation of the outcome. Exclusion of the outlier
resulted in no correlation between the experimental and predicted inhibitory activities.
Hydrophobicity of compounds has been reported to play an important role in oxidation
of CYPs and binding to liver microsomes [31,32]. Thus, considering the hydrophobicity
of curcumin analogues they would be expected to bind significantly to the hydrophobic
pockets of the CYP3A4 active site. However, this was not observed in most cases
perhaps due to other more significant factors, as shown in these results.
The present curcumin analogues generally exhibited weak inhibitory activities
towards CYP2B6. IC50 values were obtained for only four compounds and were similar
or weaker than that of curcumin, which is a weak inhibitor of CYP2B6 [15]. Compounds
from the A series did not show any inhibitory activity towards CYP2B6. The QSAR
analysis was considered unreliable, due to the small number of significant inhibitors.
Evaluation of inhibitory potentials of curcumin analogues towards CYP2C9 resulted in
twelve compounds most of which had lower inhibitory potencies than that reported for
curcumin [15]. Among these compounds A0, B1 and C0 exhibited strong inhibitory
activities (range 1.0 – 2.8 µM). Since all three series of compounds (A, B and C) are
represented in this list, it appears that the absence or presence of the central five- or
six-membered ring in the structure of the compounds is not influencing their inhibitory
effect towards CYP2C9, but rather the aromatic ring substituents. B0 and C1, with
similar substituents (OCH3 and/or OH) as the strong inhibitors above (Scheme 1-3), are
moderately strong inhibitors of CYP2C9 having IC50 values of 9.9 + 0.79 and 7.5 + 0.30
µM, respectively. The present QSAR analysis suggested that the observed inhibitory
activities are related to the descriptors, PEOE_VSA_FHYD, PEOE_VSA_FPNEG and
103
SlogP_VSA4. A weak correlation (R2 = 0.73) was obtained between experimental and
predicted data based on these descriptors. However, exclusion of the outlier B0 resulted
in a better correlation with the same descriptors (R2 = 0.836). The reason for this
compound being an outlier is not clear. However, electrostatic and hydrophobic
interactions as well as lipophilicity are likely implicated in the observed inhibitory
activities towards CYP2C9. Compound lipophilicity has been suggested to play a role in
overall substrate binding affinity of CYPs and molecular modeling studies also indicated
possible electrostatic interactions at the CYP2C9 active site due to the presence of
basic amino acid residues [26,33]. Among others the relevance of hydrophobic
interactions in the inhibition of CYP2C9 was also observed. Hydrophobic interactions
are primary driving forces and contributors to binding affinity and specificity of a ligand
to an active site [Ma et al., 2006 http://www.natural-selection.com/library/2006/NN_anti-
hiv_ligand_gbf.pdf]. This factor is therefore a potential candidate for systematic
modulation when developing SARs leading to more potent and specific inhibitors of
CYP2C9. Furthermore, the presence of a hydroxyl substituent at the para- position
appears to be prevalent in the strong inhibitors of CYP2C9 and that has been
suggested to have similar implication for other biological activities in curcumin
analogues [9].
Six curcumin analogues inhibited CYP2D6 with a wide range of IC50 values (0.6 –
111.7 µM), with C0 and C1 being the most potent. No inhibition of CYP2D6 was
observed with compounds from the A series. Therefore, it appears that the presence of
central six-membered ring in the A series contributes to the observed effect. The QSAR
analysis of the CYP2D6 inhibitors, revealed that the electrostatic descriptor,
PEOE_VSA_FPNEG, relates best with CYP2D6 inhibition, resulting in a weak
correlation (R2 = 0.522) between the experimental and predicted activities. However
removal of the outlier, C15 resulted in a good correlation with the same descriptor. The
reappearance of the electrostatic parameter indicates that the inhibitory potencies of
curcumin analogues towards CYPs is possibly related to the fraction of polar negative
van der Waals surface area present in the compounds. A limitation to our QSAR
analysis however, is the small number of compounds especially in case of CYP2D6.
Previous SAR studies on analogue series of stereoisomers of quinidine and quinine
104
suggested that hydrogen bonding by the hydroxyl group and not the basic nitrogen
interaction with an active site residue, would control the inhibitory potency of quinidine
[34]. It is likely that the presence of the hydroxyl substituent at the para- position
contributes to the observed inhibitory activity towards CYP2D6, since it is present in five
of the six CYP2D6 inhibitors. It is interesting to note that the compound B15, which has
meta-methoxy and para-hydroxyl substituents quite similar to curcumin and has been
shown to also have over ten times greater antioxidant activity than the latter [9] did not
exhibit any significant inhibitory activities towards the five CYPs tested.
Similarity studies indicated no significant differences between the ring structures
of the three series of compounds (Fig. 2), although differences may still lie in steric
bulkiness of substituents. Compound C0 clearly showed more flexibility than the A0 and
B0, and possessed the lowest minimal potential energy in each conformation as
determined by a stochastic conformational search. This is to be attributed to the
absence of the central aliphatic ring occurring in the other compounds. Clearly the
results of flexible alignments are unable to explain the observed activities of the
compounds, but we consider the similarity studies a useful approach in comparing
curcumin analogues and other structurally related compounds.
Experiments on the stability of curcumin analogues in buffer (pH 7.4) resulted in
varying degrees of degradation of the compounds. Generally compounds of group A
were more susceptible to degradation than those of groups B and C. The instability of
group A compounds may be attributed to the central six-membered ring since that is the
basic difference between the three series of compounds. Compounds in series B
appeared to be more stable in the buffer. The analogues were more stable in buffer
than curcumin, except A2, A7, C6 and C7 which were equally or less stable than
curcumin with A2 being the most unstable and B14 the most stable. Instability of group
A compounds at neutral to basic pH, may be due to aromatization of the central six-
membered ring in the presence of slightly basic conditions and subsequent dissociation
of resulting single bonds. Previous studies however, clearly demonstrated that the
degradation of curcumin in buffer at neutral to basic pH could be blocked by the
incorporation of enzyme, GSH, NAC or ascorbic acid in the incubations [7]. The present
105
results obtained indicate that like curcumin, degradation of its analogues in buffer at pH
7.4 is also significantly blocked (90%) in the presence of enzymes.
5. Conclusion
Thirty-three curcumin analogues were investigated for inhibition of human recombinant
CYP1A2, CYP3A4, CYP2B6, CYP2C9 and CYP2D6. Most of the curcumin analogues
showed low or negligible activities towards the CYPs tested. Six of the compounds (A2,
A8, A10, C2, C7 and C10) showed strong inhibitory activities towards CYP1A2, while
one compound (B0) strongly inhibited CYP3A4. CYP2C9 and CYP2D6 were strongly
inhibited by three (A0, B1, C0) and two (C0 and C1) compounds respectively. The
MOE-based QSAR analyses suggest that electrostatic and/or hydrophobic descriptors
notably PEOE_VSA_FPNEG and PEOE_VSA_FHYD, are important factors of the
compounds relating to inhibition of CYP1A2, CYP2C9 and CYP2D6. Consideration of
these (Q)SAR results might be relevant in the optimization of curcumin analogues with
less potential to cause CYP mediated drug-drug interactions.
Acknowledgement
We thank Enade Istyastono and Eva Stjernschantz for technical assistance and helpful
discussions.
References
1. Greenlee, R.T., Murray, T., Bolden, S., Wingo, P.A., 2000. Cancer statistics. CA Cancer J Clin 50:7-33.
2. Cole, G.M., Morihara, T., Lim, G.P., Yang, F., Begum, A., Frautschy, S.A., 2004. NSAID and antioxidant prevention of Alzheimer's disease: Lessons from in vitro and animal models. Annals of the New York Academy of Sciences 1035:68-84. 3. Cheng, A.L., Hsu, C.H., Lin, J.K., Hsu, M.M., Ho, Y.F., Shen, et al., 2001. Phase I clinical trial of curcumin, a chemopreventive agent in patients with high-risk or pre-malignant lesions. Anticancer Res 21:2895-2900. 4. Gescher, A., 2004. Polyphenolic phytochemicals versus non-steroidal antiinflammatory drugs: which are better cancer chemopreventive agents? J Chemother 4:3-6. 5. Vajragupta, O., Boonchoong, P., Morris, G.M., Olson, A.J., 2005. Active site binding modes of cur- cumin in HIV-1 protease and integrase. Bioorg Med Chem Letters 15:3364-3368. 6. Reddy, R.C., Vatsaala, P.G., Keshamouni, V.G., Padmanaban, G., Rangarajan, P.N., 2005. Cur- cumin for malaria therapy. Biochem Biophy Res Comm 326:472-474. 7. Oetari, S., Sudibyo, M., Commandeur, J.N.M., Samboedi, R., Vermeulen, N.P.E., 1996. Effects of
curcumin on cytochrome P450 and glutathione S-transferase activities in rat liver. Biochem Pharmacol 51:39-45.
106
8. Youssef, K.M., El-Sherbeny, M.A., El-Shafie, F.S., Farag, H.A., Al-Deeb, O.A., Awadalla, S.A.A., 2004. Synthesis of curcumin analogues as potential antioxidant, cancer preventive agents. Arch Pharm Pharm Med Chem 337:42-54.
9. Sardjiman, S., Reksohadiprodjo, M., Hakim, L., van der Goot, H., Timmerman, H., 1997. 1,5-Diphe- nyl-1,4-pentadiene-3-ones and cyclic analogues as antioxidative agents. Synthesis and structure- activity relationship. Eur J Med Chem 32:625-636. 10. Nurfina, A., Reksohadiprodjo, M., Timmerman, H., Jenie, U., Sugiyanto, D., van dr Groot, H., 1997. Synthesis of some symmetrical curcumin derivatives and their anti-inflammatory activity. Eur J Med Chem 32:321-328. 11. Zhang, Z.Y., Wong, Y.N., 2005. Enzyme kinetics for clinically relevant CYP inhibition. Curr drug Metab 6:241-257. 12. Desta, Z., Soukhova, N.V., Flockhart, D.A., 2001. Inhibition of cytochrome P450 (CYP450) isofo- rms by isoniazid: potent inhibition of CYP2C19 and CYP3A. Antimicrob Agents Chemother 45:382- 392.
13. Honig, P.K., Wortham, D.C., Zamani, K., Conner, D., Mullin, J.C., Cantilena, L.R., 1993.Terfenadine ketoconazole interaction. J Am Med Assoc 269:1513-1518.
14. Pea, F., Furlanut, M., 2001. Pharmacokinetic aspects of treating infections in the intensive care unit: focus on drug interactions. Clin Pharmacokinet 40:833-868. 15. Appiah-Opong, R., Commandeur, J.N.M., van Vugt-Lussenburg, B., Vermeulen, N.P.E., 2007.
Inhibition of human recombinant cytochrome P450s by curcumin and curcumin decomposition products. Toxicology 235:83-91.
16. Nebert, D.W., Russell, D.W., 2002. Clinical importance of the cytochromes P450. Lancet 360:1155 –1162. 17. Lamba, V., Lamba, J., Yasuda, K., Strom, S., Davila, J., Hancock, M.L., et al., 2003. Hepatic CYP2B6 expression: gender and ethnic differences and relationship to CYP2B6 genotype and CAR (constitutive androstane receptor) expression. J Pharmacol Exp Ther 307:906-922. 18. Burke, M.D., Thompson, S., Elcombe, C.R., Halpert, J., Haaparanta, T., Mayer, R.T., 1985. Ethoxy-, pentoxy- and benzyloxyphenoxazones and homologues: a series of substrates to distinguish between different induced cytochromes P-450. Biochem Pharmacol 34:3337-3345. 19. Omura, T., Sato, R., 1964. The carbon monoxide binding pigment of liver microsomes. I. Solubilization, purification and properties. J Biol Chem 239:2379-2385. 20. Staskal, D.F., Diliberto, J.J., De Vito, M.J., Bimbaum. 2005. Inhibition of human and rat CYP1A2 by TCDD and dioxin-like chemicals. Toxicol Sci 84:225-31. 21. Walsky, R.L., Obach, R.S., 2004. Validation assays for human cytochrome P450 activities. Drug Metab Dispos 32:647-660. 22. Ko, J.W., Desta, Z., Soukhova, N.V., Tracy, T., Flockhart, D.A., 2000. In vitro inhibition of the cytochrome P450 (CYP450) system by the antiplatelet drug ticlopidine: potent effect on CYP2C19 and CYP2D6. Br J Clin Pharmacol 49:343-351. 23. Labute, P., Williams, C., Feher, M., Sourial, E., Schmidt, J.M., 2001. Flexible alignment of small
molecules. J Med Chem 44:1483-1490. 24. Al-Gailany, K.A.S., Houston, J.B., Bridges, J.W., 1978. The role of substrate lipophilicity in determ-
ining type 1 microsomal P450 binding characteristics. Biochem Pharmacol 27:783-788. 25. Lewis, D.F.V., Dickins, M., 2003. Baseline lipophilicty relationships in human cytochrome P450 associated with drug metabolism. Drug Metab Rev 35:1-18. 26. Lewis, D.F.V., Brian, G.K., Yuko, I., Pavel, A., 2006. Quantitative structure-activity relationships (QSARs) within cytochromes P450 2B (CYP2B) subfamily enzymes: The importance of lipophilicity for binding and metabolism. Drug Metab Drug Interact 21:213-231. 27. Korhonen, L.E., Rahnasto, M., Mahonen, N.J., Wittekindt, C., Poso, A., Juvonen, R.O., et al., 2005. Predictive three-dimensional quantitative structure activity relationship of cytochrome P450 1A2 inhibitors. J Med Chem 48:3808-3815. 28. Adam, B.K., Ferstl, E.M., Davis, M.C., Herold, M., Kurtkaya, S., Camalier, R.F., et al., 2004. Synthesis and biological evaluation of novel curcumin analogues as anti-cancer and anti- angiogenesis agents. Bioorg Med Chem 12:3871-3883. 29. Anto, R.J., George, J., Babu, K.V., Rajasekharan, K.N., Kuttan, R., 1996. Antimutagentic and anticarcinogentic activity of natural synthetic curcuminoids. Mutat Res 370:127-131.
107
30. Shimada, T., Guengerich, F.P., 2006. Inhibition of human cytochrome P450 1A1-,1A2, and 1B1- mediated activation of procarcinogens to genotoxic metabolites by polycyclic aromatic hydrocarbons. Chem Res Toxicol 19:288-294. 31. Delaforge, M., Pruvost, A., Perrin, L., Andre, F., 2005. Cytochrome P450-mediated oxidation of
glucuronide derivatives: example of estradiol-17beta-glucuronide oxidation to 2-hydroxy-estradiol-17beta-glucuronide by CYP2C8. Drug Metab Dispos 33:466-473.
32. Austin, R.P., Barton, P., Cockroft, S.L., Wenlock, M.C., 2002. The influence of non-specific micros- omal binding on apparent intrinsic clearance, and its prediction from physicochemical properties. Drug Metab Dispos 30:1497-1503. 33. Lewis, D.F.V., 2004. Quantitative structure activity relationships (QSARs) for substrates of human cytochrome P450 CYP2 family enzymes. Toxicol in vitro 18:89-97. 34. Hutzler, J.M., Walker, G.S., Wienkers, L.C., 2003. Inhibition of cytochrome P450 2D6: structure activity studies using a series of quinidine and quinine analogues. Chem Res Toxicol 16:450-459.
108
109
Chapter 5
Inhibition of human glutathione S-transferases by curcumin and analogues
Regina Appiah-Opong, Jan N. M. Commandeur, Enade Istyastono, Jan J. Bogaards,
Nico P. E. Vermeulen
Xenobiotica, 2009, in press
Glutathione S-transferases are important phase II drug metabolizing enzymes playing a
major role in protecting cells from the toxic insults of electrophilic compounds. Curcumin,
a promising chemotherapeutic agent inhibits human GSTA1-1, GSTM1-1 and GSTP1-1
isoenzymes. The effect of three series of curcumin analogues, 2,6-
dibenzylidenecyclohexanone (A series), 2,5-dibenzylidenecyclopent-anone (B series) and
1,4-pentadiene-3-one (C series) substituted analogues on these human isoenzymes,
human and rat liver cytosolic GSTs was investigated, using 1-chloro-2,4-dinitrobenzene
as substrate. Most of the curcumin analogues showed less potent inhibitory activities
towards GSTA1-1, GSTM1-1 and GSTP1-1 than curcumin. Compounds B14 and C10
were the most potent inhibitors of GSTA1-1 and human liver cytosolic GST, with IC50
values of 0.2-0.6 µM. The most potent inhibitors GSTM1-1 were C1, C3 and C10, with
IC50 values of 0.2-0.7 µM. Similarly, GSTP1-1 was predominantly strongly inhibited by
compounds of the series C, C0, C1, C2, C10 and A0, with IC50 values of 0.4-4.6 µM.
Compounds in the B series showed no significant inhibition of GSTP1-1. QSAR analyses
have suggested the relevance of van der Waal’s surface area and compound lipophilicity
factors, for the inhibition of GSTA1-1 and GSTM1-1. These results may be useful in
design and synthesis of curcumin analogues with less potency for GST inhibition.
110
1. Introduction
Curcumin, a common dietary component in curry, derived from the plant Curcuma longa
has several important biological activities, which include anti-cancer, anti-oxidant, anti-
inflammatory and anti-HIV [1-4]. Due to the important pharmacological properties of
curcumin, clinical trials are ongoing [5,6]. In spite of the vast biological properties of
curcumin there are drawbacks to the development of this potential chemotherapeutic
agent, which include low bioavailability and instability at neutral to basic conditions [6,7].
In addition, curcumin has been shown to be a potent inhibitor of drug metabolizing human
glutathione S-transferase (GST) A1-1, GSTM1-1 and GSTP1-1 and cytochrome P450
(CYP) 3A4 and CYP2C9 [8-10]. Thus, design and synthesis of curcumin analogues with
enhanced bioavailability and better pharmacological properties has been carried out [11-
13].
Glutathione S-transferases (GSTs) are a super-family of multifunctional proteins
with fundamental roles in the cellular detoxification of a wide range of xenobiotics [14,15].
Conjugation of electrophiles to the nucleophilic sulfur atom of the major intracellular thiol,
the tripeptide glutathione, constitutes a common detoxification pathway. A number of
compounds can however be activated to more reactive or toxic products through
conjugation [6]. Alternatively, GSTs play other roles including mediation of multi-drug
resistance in cancer chemotherapy, protection of tissues against oxidative damage,
targeting of endogenous substrates and xenobiotics for transmembrane transport, which
is essential in processes such as biosynthesis of leukotrienes and ligandins [15,17]. Most
GSTs exist as soluble enzymes and are active as dimeric proteins, with each subunit
having an active site composed of two distinct functional regions, comprising of a G-site
specific for binding of the co-substrate glutathione (GSH), and a hydrophobic H-site
binding structurally diverse electrophilic substrates [15]. The seven main classes of
mammalian GSTs are alpha (A), mu (M), pi (P), theta (T), sigma (S), Omega (O) and zeta
(Z) based on amino acid/nucleotide sequence identity and physical structure of the genes
[15]. The cytosolic GSTs are differentially expressed in various organs. GSTA and GSTM
are predominantly expressed in the liver, whereas insignificant levels of GSTP1 are
111
expressed in the liver [18]. Higher levels of GSTP1 are however found in human
erythrocytes, lungs, esophagus and placenta [19-22].
The crucial roles of GSTs in drug metabolism, enhancing elimination of
electrophilic toxic drugs and metabolites through GSH-conjugation and physiological roles
mentioned above, would be reduced or compromised as a result of inhibition of GSTs and
this could have profound toxicological or clinical implications. Inhibition of human GSTs by
many compounds, drugs and natural products such as RRR-!-tocopherol, quinine,
quinidine, tetracycline and artemisinin, have previously been studied in vitro [23-25]. For
example, with the anti-malarials quinine, quinidine, tetracycline and artemisinin, IC50
values towards GSTP1-1 were below peak plasma concentrations obtained during
therapy, therefore it is likely that this isoenzyme may be inhibited in vivo at doses
normally used in clinical practice. On the other hand, the role of GST inhibition in cancer
therapy has been well studied [26,27].
In the present study we investigated the inhibitory potentials of thirty-four curcumin
analogues towards human recombinant GSTA1-1, GSTM1-1, GSTP1-1, human and rat
liver cytosolic GSTs. The compounds were designed and synthesized in three series by
Sardjiman et al [11] (Schemes 1-3). Studies on anti-oxidant activities of these compounds
have shown that some of them possess similar or even more potent activities than
curcumin [11]. Most of these compounds have also shown weak inhibitory activities
towards recombinant human drug metabolizing CYPs [28]. In this study, quantitative
structure-activity relationships (QSARs) of GST-inhibition by these curcumin analogues
were also investigated using the MOE (Molecular Operating Environment) program, in
order to identify molecular features responsible for the inhibition of the three different GST
isoenzymes tested. These results may give leads for improved design of curcumin
analogues with less inhibitory potential towards GST inhibition.
112
Scheme 1
33
1
2
O
R
R
R R
R
R
2
1
Cyclohexanones (A)
Scheme 2
1
33
R1
R
R
O
R
R
R
22
Cyclopentanone (B)
R1 R2 R3
A8 H N(CH3)2 H
A10 Cl H H
A11 CH3 OH CH3
A14 t-C4H9 OH t-C4H9
R1 R2 R3
A0 H OH H
A2 H H H
A4 H OCH3 H
A5 H CH3 H
A7 H CF3 H
R1 R2 R3
B0 H OH H
B1 OCH3 OH H
B2 H H H
B3 H Cl H
B4 H OCH3 H
B7 H CF3 H
B8 H N(CH3)2 H
R1 R2 R3
B9 Cl Cl H
B10 Cl H H
B11 CH3 OH CH3
B12 C2H5 OH C2H5
B13 i-C3H7 OH i-C3H7
B14 t-C4H9 OH t-C4H9
B15 OCH3 OH OCH3
113
Scheme 3
1
33
R1
R
R
O
R
R
R
22
1,4-pentadiene-3-ones (C)
Schemes 1-3 show the chemical structures of the synthetic curcumin analogues.
2. Materials and methods
2.1. Reagents
Reduced glutathione was obtained from Sigma-Aldrich (Steinheim, Germany). 1-chloro-
2,4-dinitrobenzene (CDNB) was obtained from Aldrich-Europe (Beerse, Belgium). Pooled
human liver cytosolic fraction was purchased from CellzDirect Inc. (North Carolina, USA).
Curcumin analogues were kindly donated by Dr. S. Sardjiman (Jakarta, Indonesia).
Purified human recombinant GSTA1-1 and GSTM1-1 were donated by Dr. J.J. Bogaards
(TNO Zeist, The Netherlands) and human GSTP1-1 was a gift from Prof. M. Lo Bello
(University of Rome, Italy). The specific activities of the GSTs using CDNB as substrate
were, 123 U/mg, 262 U/mg and 24.7 U/mg protein, respectively [25]. All other chemicals
were of analytical grade and obtained from standard suppliers.
R1 R2 R3
C0 H OH H
C1 OCH3 OH H
C2 H H H
C3 H Cl H
C5 H CH3 H
C6 H t-C4H9 H
R1 R2 R3
C7 H CF3 H
C9 Cl Cl H
C10 Cl H H
C11 CH3 OH CH3
C15 OCH3 OH OCH3
114
2.2. Rat liver cytosol
Rat liver cytosol was prepared from untreated rats as described [29]. Briefly, isolated rat
liver samples were homogenized in 2 volumes of 50 mM potassium phosphate buffer (pH
7.4) containing 0.9% sodium chloride, using a Potter-Elvehjem homogenizer at 4oC.
Cytosolic fractions were obtained by centrifuging the homogenate fraction for 20 min at
12000 g, and subsequent centrifugation of the supernatant for 60 min at 100000 g.
Supernatant representing cytosolic proteins, were freezed at -20oC until use. Protein
concentration was determined by the method of Bradford [30].
2.3. GST inhibition assays
Inhibition of GST activity was assessed as described [25] with slight modification.
GST-mediated conjugation of CDNB to GSH was measured using an Ultrospec
2000 Pharmacia Biotech UV/visible spectrophotometer at the wavelength of 340 nm
for 2 min. Incubation mixtures (1 ml) contained 0.1 M potassium phosphate buffer
pH 6.5, 400 µM CDNB, 1 mM GSH, and enzyme solutions (2.1 µg/ml protein
(GSTA1-1), 1.5 µg/ml protein) GSTM1-1, 1.0 µg/ml protein (GSTP1-1), or 32.0 and
1.7 µg/ml protein, human and rat liver cytosolic fractions respectively). After the
addition of GSH and enzymes to the reaction mixture, curcumin analogues were
added and the reaction was started with the addition of the substrate, and monitored
for 2 min at room temperature (24oC). The thirty-four compounds were firstly tested
at a concentration of 100 µM. Subsequently, compounds showing > 70% inhibition
were selected for IC50 determination in the concentration range 0.043-100 µM. All
assays were linear functions of protein concentration and of time for at least 2 min.
The formation of GSH-conjugates of the curcumin analogues in the reaction mixture
was determined using gradient HPLC analysis. A C18 column (150 mm x 3.2 mm, 5
!M particle size, Phenomenex), a mobile phase consisting of 0.1% trifluoroacetic
acid (solvent A) and 99% acetonitrile (solvent B) were used, with a linear gradient
eluting from 0-10 min 0-29% B, 10-40 min 29% B, 40-50 min 29-95% B, and a flow
rate of 0.5 ml/min. Conjugate formation was monitored using a UV/visible detector at
the wavelengths 360 nm and 427 nm.
115
2.4. Data and QSAR analysis
Percent inhibition of GST activity by curcumin analogues was calculated from the
ratio of inhibited versus control samples. The inhibitor concentrations giving 50%
inhibition of enzyme activity (IC50) values were analyzed using GraphPad Prism
4.0 version (GraphPad Prism software Inc. San Diego CA). Descriptors for QSAR
analysis were calculated using Molecular Operating Environment (MOE) software
version 2006.08 developed by Chemical Computing Group Inc, Canada. A list of
the descriptors finally selected and applied in this study can be found in Table 1.
Table 1. Relevant descriptors in this QSAR studies, obtained from QuaSAR-Descriptors, MOE
2006.08 version.
No. Descriptors Description
1 CASA- Negative charge weighted surface area
2 dipole Dipole moment calculated from the partial charges of the
molecule
3 PEOE_VSA-0 Sum of the van der Waals surface area of atoms where partial
charge of atom (calculated using the Partial Equalization of
Orbital Electronegativities (PEOE) method) is in the range -
0.05 to 0.00.
4 PEOE_RPC+ Relative positive partial charge
5 SlogP_VSA4 Sum of the van der Waals surface area of atoms such that
contribution to log of octanol/water partition coefficient
calculated from the given structure is in the range 0.1 to 0.15.
6 SlogP_VSA8 Sum of the van der Waals surface area of atoms such that
contribution to log of octanol/water partition coefficient
calculated from the given structure is in the range 0.3 to 0.4.
7 SMR_VSA7 Sum of the van der Waals surface area of atoms such that
contribution to molar refractivity calculated from the given
structure is > 0.56.
116
Conformational analysis using stochastic conformation search was performed using the
conformational import module provided by MOE with no filters and no constraints. The
conformational analysis and energy minimization were performed using stochastic
conformational search with root mean square (RMS) gradient of 0.001 Å and iteration limit
of 10,000 using the MMFF94 force field. All non-quantum descriptors were calculated as
described previously [31]. The relationships between the log 1/IC50 (IC50 in Molar) and the
descriptors were identified by stepwise regression analysis using SPSS for Windows
version 14.0 developed by SPSS Inc, USA. The following statistical measures were used:
N=number of samples, F=coefficient of variance (test for quality of fit), r=coefficient of
correlation, R2=coefficient of correlation squared and s = standard error of estimation. The
descriptors selected for the “best model”, using stepwise regression analysis were
independent (i.e. the cross correlation between descriptors < 0.7; Pearson correlation
method using SPSS was performed to analyze the cross correlation). The leave-one-out
cross-validation (LOO-CV/q2) method was employed to determine cross-validated
coefficient (q2) as internal validation of the models.
3. Results
3.1. Inhibition of purified GST isoenzymes
Fifteen out of the thirty-four curcumin analogues tested, inhibited GSTA1-1 activity by less
than 70% at concentrations of 33.3 (due to solubility problems) or 100 µM (data not
shown). The IC50 values of compounds showing more than 70% inhibition of GSTA1-1
are shown in Tables 2-4. Only seven compounds (A0, B0, B14, C1, C9 and C10) showed
stronger inhibitory activities towards GSTA1-1, compared to curcumin (Tables 2-4).
Compounds B14 possessing a hydroxyl group at the para position (R2) in addition to the
bulky butyl substituents at the R1 and R3 positions and C10 having only a chloride group
at position R1 as the only substituent, were the most potent inhibitors of GSTA1-1 (IC50
values, 0.5 and 0.6 µM, respectively).
117
Table 2. Percent inhibition or IC50 values (µM) for inhibition of human recombinant GSTs by group A
curcumin analogues and curcumin
Cmpd R1 R2 R3 GSTA1-1 GSTM1-1 GSTP1-1
A0* H OH H 16.7 + 0.49 6.4 + 0.55 1.6 + 0.01
A2 H H H 73.3 + 0.02 12.2% 54.8 + 0.16
A4 OH OCH3 H 65.3% 27.3% 71.8 + 4.44
A7 H CF3 H 56.9 + 3.25 77.2 + 4.71 5.6%
A11 CH3 OH CH3 53.9% 17.1 + 4.14 49.6%
Cur OCH3 OH OCH3 18.8 + 0.77 0.3 + 0.10 15.1 + 1.12
Cmpd, compound code; *highest concentration tested, 33.3 µM; IC50 values are means + standard
deviation (S.D.) of two experiments as described in the Methods section.
Similarly sixteen of the compounds inhibited GSTM1-1 activity by less than 70%.
Tables 2-4 show the IC50 values of compounds exhibiting more than 70% inhibition of
GSTM1-1. Thirteen compounds exhibited potent inhibitory activities towards GSTM1-1
having IC50 values in the range 0.2-9.9 µM.
Table 3. Percent inhibition or IC50 values (µM) for inhibition of human recombinant GSTs by group B
curcumin analogues
Cmpd R1 R2 R3 GSTA1-1 GSTM1-1 GSTP1-1
B0 H OH H 10.7 + 2.52 4.1 + 3.04 57.0%
B1 OCH3 OH H 61.0% 1.0 + 0.18 52.4%
B2 H H H 48.5 + 0.40 59.9% 20.0%
B9 Cl Cl H 65.8% 31.9 + 0.78 31.1%
B11 CH3 OH CH3 28.7 + 4.77 9.9 + 2.40 40.4%
B12 C2H5 OH C2H5 24.5 + 10.1 30.0 + 2.93 26.3%
B14* t-C4H9 OH t-C4H9 0.5 + 0.13 58.0% 43.2%
Cmpd, compound code; *highest concentration tested, 33.3 µM; IC50 values are means + standard
deviation (S.D.) of two experiments as described in the Methods section.
118
Compound C1 (R1 = OCH3 R2 = OH) and compound C9 (R1 and R2 = C) inhibited
GSTM1-1 with equipotent activities compared to curcumin (0.3 !M) whilst compound C10
(R1= Cl) showed slightly more potency.
Table 4. Percent inhibition or IC50 values (µM) for inhibition of human recombinant GSTs by group C
curcumin analogues
Cmpd R1 R2 R3 GSTA1-1 GSTM1-1 GSTP1-1
C0 H OH H 43.5 + 3.06 1.5 + 0.49 0.6 + 0.14
C1 OCH3 OH H 6.9 + 1.04 0.3 + 0.10 4.6 + 0.44
C2 H H H 21.3 + 0.86 9.9 + 2.13 4.5 + 2.13
C3* H Cl H 77.7 + 3.15 0.7 + 0.12 6.1 +1.06
C5* H CH3 H 25.3 + 6.22 59.4% 55.2%
C7 H CF3 H 29.4% 2.0 + 1.11 51.5%
C9* Cl Cl H 1.6 + 0.44 0.3 + 0.06 22.0%
C10* Cl H H 0.6 + 0.02 0.2 + 0.03 0.4 + 0.11
C11* CH3 OH CH3 18.0 + 0.61 2.4 + 0.37 10.9 + 1.29
C15 OCH3 OH OCH3 21.0% 4.2 + 0.52 25.6%
Cmpd, compound code; *highest concentration tested, 33.3 µM; IC50 values are means + standard
deviation (S.D.) of two experiments as described in the Methods section.
The GSTP1-1 activity was inhibited by twenty-five of the curcumin analogues by less than
70%. Seven compounds showed more potent inhibitory activities towards this enzyme as
compared to curcumin. Six of these compounds belonged to the C series of curcumin
analogues, whereas only one was of the A series. None of the compounds in series B
inhibited GSTP1-1 activity by >70%. The two most potent inhibitors of GSTP1-1 among
the compounds were compounds C0 that has a hydroxyl group at the R2 position and C10
(IC50 values, 0.6 and 0.4 µM, respectively).
119
3.2. Human and rat liver cytosol
The thirty-four curcumin analogues shown in schemes 1-3 were also screened for
inhibitory activities towards pooled human and rat liver cytosolic GSTs. This resulted in
most of the compounds exhibiting less than 70% inhibitory activities towards the liver
cytosolic enzymes of both species. Seven compounds showed over 70% inhibition of
human liver GSTs, these are A0, B0, B14, C1, C3, C9 and C10, all having IC50 values in
the range 0.2-25.3 µM. The compounds B14 and C10 which were the most potent
inhibitors of GSTA1-1, were also found to be the most potent inhibitors of human liver
GSTs (Table 5).
Table 5. Percent inhibition or Percent inhibition or IC50 values (µM) for inhibition of human and rat
liver cytosolic GSTs by curcumin analogues and curcumin
Cmpd R1 R2 R3 Human liver Rat liver
A0* H OH H 25.3 + 0.86 53.0%
B0 H OH H 9.3 + 0.08 14.4 + 1.33
B14* t-C4H9 OH t-C4H9 0.2 + 0.01 12.8 + 1.51
C1 OCH3 OH H 18.2 + 1.53 0.8 + 0.17
C3* H Cl H 7.3 + 1.30 65.2%
C9* Cl Cl H 7.6 + 1.24 45.1%
C10* Cl H H 0.3 + 0.03 0.4 + 0.03
C11* CH3 OH CH3 61.8% 2.2 + 0.05
Cur OCH3 OH H 50.5% 4.2 + 0.23
Cmpd, compound code; *highest concentration tested, 33.3 µM; IC50 values are means + standard
deviation (S.D.) of two experiments as described in the Methods section.
On the other hand, six compounds inhibited rat liver cytosolic GST activities. The
compounds C1 and C10 were the most potent inhibitors of the rat liver GSTs, having IC50
values of 0.8 and 0.4 µM, respectively.
120
3.3. Quantitative structure-activity relationships (QSARs)
The QSARs relating the various molecular descriptors generated by the MOE program to
the GST inhibitory potencies of curcumin analogues were analyzed in this work. Equation
1 (Eq. 1) generated from the stepwise regression analysis is considered as the ‘best’
model for curcumin analogues as inhibitor of GSTA1-1.
Log (1/IC50) = 5.283 (± 0.307) + 0.007 (± 0.001) [SMR_VSA7]
- 0.070 (± 0.027) [SlogP_VSA4]
- 0.326 (± 0.239) [dipole] (Eq. 1)
N = 16, r = 0.813, R2 = 0.660, S = 0.452, F3, 12 = 7.771, F5%, 3, 12 = 3.490, q2 = 0.464.
The relatively important descriptors for this model are SMR_VSA7, SlogP_VSA4 and
dipole. The q2 value was 0.464, which indicates a low predictive power (i.e. <0.5), and the
difference between R2 and q2 was less than 0.2, but only a weak correlation was found
between the experimental and predicted inhibitory potencies. However, assigning C3 as
an outlier with the same descriptors as used in Eq. 1, resulted in a model with a good
predictive power where q2 is 0.617 and a stronger R2 value of 0.78.
In the QSAR analysis of GSTM1-1 inhibitors, equation 2 (Eq. 2) is considered as
the ‘best’ model of curcumin analogues as inhibitors of GSTM1-1.
Log (1/IC50 GSTM1-1) = 6.918 (± 0.510) - 0.009 (± 0.004) [PEOE_VSA-0]
- 0.014 (± 0.007) [SlogP_VSA8]
- 1.000 (± 0.999) [PEOE_RPC+] (Eq. 2)
N = 17, r = 0.807, R2 = 0.651, S = 0.502, F3, 13 = 8.087, F5%, 3, 13 = 3.411, q2 = 0.388.
The relevant descriptors for this model are PEOE_VSA-0, SlogP_VSA8 and
PEOE_RPC+. The difference between R2 and q2 is more than 0.2, suggesting the
presence of outlier(s) [32]. The q2 is also less than 0.5 (q2 = 0.338) and R2 value 0.651,
suggesting that the model possesses a weak predictive power. By assigning C15 as an
outlier and using the same descriptors, a model with better predictive power was
generated, with a q2 of 0.674 and R2 of 0.792.
121
The relatively important descriptor for this QSAR analysis of curcumin analogues
as GSTP1-1 inhibitors, CASA-, resulted in a good predictive model(Eq. 3)
Log (1/IC50 GSTP1-1) = 1.375 (± 1.086) - 0.003 (± 0.001) [CASA-] (Eq. 3)
N = 8, r = 0.832, R2 = 0.692, S = 0.498, F1, 6 = 13.486, F5%, 1, 6 = 5.990, q2 = 0.692.
GSTA1-1
r2 = 0.780
3.5
4.0
4.5
5.0
5.5
6.0
6.5
3.5 4.0 4.5 5.0 5.5 6.0 6.5
Experimental log 1/IC50 (1/M)
Pre
dic
ted
lo
g 1
/IC
50
(1
/M)
GSTM1-1
r2 = 0.792
4.0
4.5
5.0
5.5
6.0
6.5
4.0 4.5 5.0 5.5 6.0 6.5
Experimental log 1/IC50 (1/M)
Pre
dic
ted
lo
g 1
/IC
50
(1
/M)
GSTP1-1
r2 = 0.692
4.0
4.5
5.0
5.5
6.0
6.5
4.0 4.5 5.0 5.5 6.0 6.5
Experimental log 1/IC50 (1/M)
Pre
dic
ted
lo
g 1
/IC
50
(1
/M)
Figure 1. Plots of observed and calculated inhibitory activities (log 1/IC50). GSTA1-1; The log
(1/IC50)predicted values were calculated from Eq. 1 (A); GSTM1-1; The log (1/IC50)predicted values were
calculated from Eq. 2 (B); and GSTM1-1; The log (1/IC50)predicted values were calculated from Eq. 3 (C).
The q2 is more than 0.5 (q2 = 0.692), suggesting that the model possesses a good
predictive power, however a weak correlation was observed (R2 = 0.692). Plots of
C
A B
122
experimental and calculated inhibitory activities for all three human GST used are
presented in Figure 1.
Discussion
The effect of curcumin analogues on the activity of human recombinant GST activity was
assessed by measuring the inhibition of GST-mediated conjugation of GSH to CDNB.
Earlier studies have shown that curcumin is a potent inhibitor of GSTs A1-1, M1-1 and
P1-1, using CDNB as substrate, with IC50 values of 2, 0.04 and 5 µM [9]. The alpha (A)
and mu (M) GST isoforms are predominantly expressed in the human and rat livers,
whereas levels of the pi (P) form are insignificant [18]. Although ubiquitously expressed in
humans, the pi isoform is the most abundant GST in erythrocytes and significant levels
are also found in the lungs, esophagus and placenta [19,33].
The present study shows that most of the curcumin analogues are weaker
inhibitors of the human recombinant GSTs tested when compared to curcumin, and
patterns of selectivity have also been revealed. GSTA1-1 and GSTM1-1 were most
susceptible to inhibition by the curcumin analogues. Only seven compounds inhibited
GSTA1-1 more strongly than curcumin, three of these compounds belong to the C series,
and the most potent inhibitors were B14 and C10. Steric properties may be partly
responsible for the strong inhibitory activity of B14, because compounds with similarly
bulky substituents, as well as the central five-membered ring, such as B11 and B12
showed over 70% activities. On the other hand, the presence of the chloride substituent
at the meta position in C10 appears to increase the inhibitory activity towards GSTA1-1,
as observed also with C9. The compound C10 also showed strong inhibitory activity
towards CYP1A2, whilst B14 was a weak inhibitor [28].
Thirteen out of thirty-four compounds showed rather strong (<10 µM) inhibitory
activities towards GSTM1-1. However, comparing these activities with that of curcumin,
nine of the compounds showed weaker activities, with only four compounds of the C
series i.e. C1, C3, C9 and C10, showing similar inhibitory potencies towards this enzyme.
Since GSTM1 is expressed in only 60% of human individuals, inhibition of this enzyme
may not have significant clinical implications as observed in individuals lacking this
123
enzyme. However, it has also been reported that people lacking GSTM1-1, have higher
risk of developing lung cancer [23].
Similarly, GSTP1-1 was weakly inhibited by most of the curcumin analogues, with
twenty-five compounds out of thirty-four showing less than 70% inhibitory activities. None
of the compounds of series B showed any significant inhibition of GSTP1-1, suggesting
that the weak inhibitory activities of these compounds may be due to the presence of the
central five-membered ring. Strong inhibition of GSTP1-1, by compounds such as C0 and
C10, could have implications for oxidative stress in human tissues and erythrocytes
where the enzyme is highly expressed. This class as opposed to the other GST classes is
highly susceptible to oxidation due to a reactive cysteine residue [Cys47 in human, rat
and mouse GSTP1-1 and Cys45 in that of the pig] located near the glutathione-binding
site [34].
The compounds B14 and C10 were the most potent inhibitors of human liver
cytosolic GSTs and GSTA1-1 as expected, due to the high levels of GSTA in the liver.
Steric hindrance, together with the central five-membered ring could play a role in the
potent inhibitory activity of B14. The IC50 values with the recombinant human GSTs were
however, generally much lower compared to those with human liver cytosol. Especially in
the cases where moderate inhibition of GSTA1-1 and strong inhibition of GSTM1-1 were
recorded, a moderate to strong inhibition may be expected with the human liver cytosol, in
which GSTA1-1 and GSTM1-1 are known to be significantly expressed [18,25]. These
differences in inhibition could possibly be due to binding to proteins such as hemoglobin
and albumin, with higher affinity than to GSTs, as suggested by van Haaften et al [25].
The presence of the para-hydroxy and chloride substituents in C1 and C10, respectively,
possibly contributes to their strong GST inhibitory activities observed in rat liver cytosol.
Like curcumin, all the present analogues possess !, !-unsaturated carbonyl
groups, and hence have the potential to conjugate to GSH, which could influence the
outcome of the current inhibition experiments. Previous studies have shown that GSH
conjugates are good inhibitors of GSTs [35], however, after incubation for 2 min GSH-
conjugates of none of the analogues could be detected to any significant amount by
HPLC analysis, which is supporting the concept that the compounds are responsible for
the observed inhibitory activities. Generally, the strongest inhibitors of GSTs were of the
124
series C. Inhibition of GST may have toxicological consequences similar to that of
deficiency in GST expression, which has not been reported to result in abnormalities. On
the other hand overexpression of GSTs has been implicated in drug resistance in cancer
chemotherapy [36].
The QSAR model of GSTA1-1 inhibitors shows a positive correlation with
SMR_VSA7, and negative correlations with SlogP_VSA4 and dipole. This suggests that
in designing new curcumin analogues with less potent inhibition, the compounds should
possess low values of SMR_VSA7, but high values of SlogP_VSA4 and dipole. On the
other hand, the QSAR model of GSTM1-1 inhibitors has negative correlations with
PEOE_VSA-0, SlogP_VSA8, and PEOE_RPC+. This implies that to avoid inhibition of
GSTM1-1, new curcumin analogues should be designed to have high values of
PEOE_VSA-0, SlogP_VSA8, and PEOE_RPC+. With respect to the GSTP1-1 inhibitors,
a negative correlation with CASA- is shown by the QSAR model, thus to design new
curcumin analogues with weak inhibitory activities towards GSTP1-1, the compounds
should have high CASA-.
In conclusion this study has shown the inhibitory potencies of thirty-four
compounds, representing three series of curcumin analogues towards three important
human GST isoenzymes, human and rat cytosolic GSTs. The present study has shown
that 27, 31 and 27 curcumin analogues out of thirty four are less potent inhibitors of
GSTA1-1, GSTM1-1 and GSTP1-1, respectively than curcumin. Since GSTs are a major
group of phase II detoxification enzymes, potent inhibition by compounds such as C10
and B14 could have implications for toxicity in humans. The strong inhibitory activities
exhibited by some of the curcumin analogues could however have useful application in
chemotherapy. The MOE-based QSAR analyses have also suggested the relevance of
van der Waal’s surface area and compound lipophilicity factors for the inhibition of
GSTA1-1 and GSTM1-1. These results may be useful in designing of curcumin analogues
with less inhibitory activity towards GSH or in the consideration of these compounds as
chemotherapeutic agents.
125
References
1. Chen, A., Xu, J., Johnson, A.C., 2006. Curcumin inhibits human colon cancer cell growth by suppressing gene expression of epidermal growth factor receptor through reducing the activity of the transcription factor Egr-1. Oncogene 25:278-287.
2. Cole, G.M., Morihara, T., Lim, G.P., Yang, F., Begum, A., Frautschy, S.A., 2004. NSAID and antioxidant prevention of Alzheimer's disease: Lessons from in vitro and animal models. Annals NY Acad Sci 1035:68-84.
3. Nonn, L., Doung, D., Peehl, D.M., 2007. Chemopreventive anti- inflammatory activities of curcumin and other phytochemicals mediated by MAP kinase phosphatase-5 in prostate cells. Carcinogenesis 28:1188-96.
4. Vajragupta, O., Boonchoong, P., Morris, G.M., Olson, A.J., 2005. Active site binding modes of curcumin in HIV-1 protease and integrase. Bioorg Med Chem Letters 15:3364-3368. 5. Cheng, A.L., Hsu, C.H., Lin, J.K., Hsu, M.M., Ho, Y.F., Shen, T.S., et al., 2001. Phase I clinical trial of
curcumin, a chemopreventive agent, in patients with high-risk or premalignant lesions. Anticancer Res 21:2895-2900.
6. Sharma, R.A., Euden, S.A., Platton, S.L., Cooke, D.N., Shafayat, A., Hewitt, H.R., et al., 2004. Phase I clinical trial of oral curcumin: biomarkers of systemic activity and compliance. Clin Cancer Res 10:6847-6854.
7. Oetari, S., Sudibyo, M., Commandeur, J.N.M., Samboedi, R., Vermeulen, N.P.E., 1996. Effects of curcumin on cytochrome P450 and glutathione S-transferase activities in rat liver. Biochem Pharmacol 51:39-45.
8. van Iersel, M.L., Ploemen, J.P., Lo Bello, M., Federici, G., van Bladeren, P.J., 1997. Interactions of alpha, beta-unsaturated aldehydes and ketones with human glutathione S- transferase P1-1. Chem Biol Interact 108:67-78.
9 Hayeshi, R., Mutingwende, I., Mavengere, W., Masiyanise, V., Mukanganyama, S., 2007. Inhibition of human glutathione S-transferases activity by plant polyphenolic compounds ellagic acid and curcumin. Food Chem Toxicol 45:286-295.
10. Appiah-Opong, R., Commandeur, J.N.M., van Vugt-Lussenburg, B., Vermeulen, N.P.E., 2007. Inhibition of human recombinant cytochrome P450s by curcumin and curcumin decomposition products. Toxicology 235:83-91.
11. Sardjiman, S., Reksohadiprodjo, M., Hakim, L., van der Goot, H., Timmerman, H., 1997. 1,5- Diphenyl-1,4-pentadiene-3-ones and cyclic analogues as antioxidative agents. Synthesis and structure-activity relationship. Eur J Med Chem 32:625-636.
12. Youssef, K.M., El-Sherbeny, M.A., El-Shafie, F.S., Farag, H.A., Al-Deeb, O.A., Awadalla, S.A.A., 2004. Synthesis of curcumin analogues as potential antioxidant, cancer preventive agents. Arch Pharm Pharm Med Chem 337:42-54. 13. Lin, L., Shi, Q., Nyarko, A.K., Bastow, K.F., Wu, C.C., Su, C.Y., et al., 2006. Antitumor agents 250. Design and synthesis of new curcumin analogues as potential anti-prostate cancer agents. J Med Chem 49:3963-3972. 14. Commandeur, J.N., Stijntjes, G.J., Vermeulen, N.P., 1995. Enzymes and transport systems involved in the formation and disposition of glutathione S-conjugates. Role in bioactivation and detoxication mechanisms of xenobiotics. Pharmacol Rev 47:271-330. 15. Frova, C., 2006. Glutathione transferases in the genomics era: New insights and perspectives.
Biomol Eng 23:149-169. 16. van Bladeren, P.J., 2000. Glutathione conjugation as a bioactivation reaction. Chem Biol Interact
129:61-79. 17. Morrow, C.S., Simitherman, P.K., Townsend, A.J., 2000. Role of multidrug-resistance protein 2 in glutathione S-transferase P1-1-mediated resistance to 4-nitroquinoline 1-oxide toxicities in HepG2 cells. Mol Carcinog 29:170-178. 18. Eaton, D.L., Bammler, T.K., 1999. Concise review of glutathione S- transferases and the significa-
nce to toxicology. Toxicol Sci 49:156-164. 19. Awasthi, Y.C., Sharma, R., Singhal, S.S., 1994. Human glutathione S-transferases. Int J Biochem 26:295-308.
126
20. Prade, L., Huber, R., Manoharan, T.H., Fahl, W.E., Reuter, W., 1997. Structures of class pi glutathione S-transferase from human placenta in complex with substrate, transition-state analo- gue and inhibitor. Structure 5:1287-1295.
21. Chandra, R.K., Bentz, B.G., Haines, G.K., Robinson, A.M., Radosevich, J.A., 2002. Expression of glutathione S-transferase Pi in benign mucosa, Barrett's metaplasia, and adenocarcinoma of the esophagus. Head Neck 24:575–81. 22. Piipari, R., Nurminen, T., Savela, K., Hirvonen, A., Mantyla, T., Anttila, S., 2003. Glutathione S- transferases and aromatic DNA adducts in smokers bronchioaveolar macrophages. Lung Cancer 39:265-272.
23. van Bladeren, P.J., van Ommen, B., 1991. The inhibition of glutathione S-transferases: Mechan- isms, toxic consequences and therapeutic benefits. Pharmacol Ther 51:35-46. 24. Mukanganyama, S., Widersten, M., Naik, Y.S., Mannervik, B., Hasler, J.A., 2002. Inhibition of
glutathione S-transferases by antimalarial drugs possible implications for circumventing anticancer drug resistance. Int J Cancer 97:700-705.
25. van Haaften, R.I.M., Haenen, G.R.M.M., van Bladeren, P.J., Bogaards, J.J.P., Evelo, C.T.A., Bast, A., 2003. Inhbition of glutathione S-transferase isoenzymes by RRR-!-tocopherol. Toxicol in Vitro 17:245-251.
26. Townsend, D.M., Tew, K.D., 2003. The role of glutathione-S-transferase in anti-cancer drug resista-nce. Oncology 22:7369-7375.
27. Depeille, P., Cuq, P., Passagne, I., Evrard, A., Vian, L., 2005. Combined effects of GSTP1 and MRP1 in melanoma drug resistance. B J Cancer 93:216-223.
28. Appiah-Opong, R., de Esch, I., Commandeur, J.N.M., Andarini, M., Vermeulen, N.P.E., 2007. Structure-activity relationship for the inhibition of recombinant human cytochrome P450 mediated metabolism by curcumin analogues. Eur J Med Chem 43:1621-1631.
29. Rooseboom, M., Vermeulen, N.P.E., Groot, E.J., Commandeur, J.N.M., 2002. Tissue distribution of cytosolic beta-elimination reactions of selenocysteine Se-conjugates in rat and human. Chem Biol Interact 140:243-264.
30. Bradford, M.M., 1976. A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248-254.
31. Shahapurkar, S., Pandya, T., Kawathekar, N., Chaturvedi, S.C., 2004. Quantitative structure activity relationship studies of diaryl furanones as selective COX-2 inhibitors. Eur J Med Chem 39:899-904.
32. Ericksson, L., Jaworska, J., Worth, A.P., Cronin, M.T., McDowell, R.M., Gramatica, P., 2003. Methods for reliability and uncertainty assessment and for applicability evaluations of classification- and regression-based QSARs. Environ Health Perspect 111:1361-1375.
33. Moscow, J.A., Fairchild, C,R., Madden, M.J., Ransom, D.T., Wieand, H.S., O’Brien, E.E., et al., 1989. Expression of anionic glutathione S-transferase and P-glycoprotein genes in human tissues and tumors. Cancer Res 49:1422-1428.
34. Sluis-Cremer, N., Naidoo, N., Dirr, H., 1996. Class-pi glutathione is unable to regain its native conformation after oxidative inactivation by hydrogen peroxide. Eur J Biochem 242:301-307.
35. Burg, D., Filippov, D.V., Hermanns, R., van der Marel, G.A., van Boom, J.H., Mulder, G.J., 2002. Peptidomimetic glutathione analogues as novel gamma GT stable inhibitors. Bioorg Med Chem 10:195-205. 36. Burg, D., Mulder, G.J., 2002. Glutathione conjugates and their synthetic derivatives as inhibitors of
glutathione-dependent enzymes involved in cancer and drug resistance. Drug Metab Rev 34:821- 863.
127
Chapter 6
Interactions between Cytochromes P450, Glutathione S-transferases and
Ghanaian Medicinal plants
Regina Appiah-Opong, Jan N. M. Commandeur, Civianny Axson and Nico P. E.
Vermeulen
Adapted from Food and Chemical Toxicology 2008 46:3598-3603
Inhibition of cytochrome P450s (CYPs) is a major cause of adverse drug-drug
interactions. Alternatively, inhibition of glutathione S-transferases (GSTs) may increase
harmful effects from electrophilic compounds. In the present study, aqueous extracts of
seven Ghanaian medicinal plants were investigated for inhibitory potentials towards
recombinant human CYP1A2, CYP2C9, CYP2D6 and CYP3A4 heterologously
expressed in Escherichia coli. Effects of these extracts on recombinant human GSTA1-
1, GSTM1-1, GSTP1-1, human and rat cytosolic GSTs were also investigated. Seven
extracts, including Phyllanthus amarus whole plant, leaf, stem and root, Cassia siamea
and Momordica charantia, inhibited CYP1A2 and CYP2C9 with IC50 values ranging from
28.3-134.3 !g/ml and 63.4-425.9 !g/ml, respectively. Similarly, both CYP2D6 and
CYP3A4 were each inhibited by five extracts including Phyllanthus amarus whole plant,
leaf, stem and root and Cassia alata, with IC50 values ranging from 45.8-182.0 !g/ml
and 79.2-158.8 !g/ml respectively. Human and rat liver cytosolic GSTs were inhibited
by the extracts with IC50 values ranging from 25.2-95.5 !g/ml and 8.5-139.4 !g/ml,
respectively. GSTM1-1 was most susceptible to the inhibition by the extracts, with IC50
values in the range 3.6-50.0 !g/ml, whilst IC50 values of 8.9-159.0 !g/ml and 68.6-157.0
!g/ml were obtained for GSTA1-1 and GSTP1-1, respectively. These findings show the
potential for CYP- and GST-mediated herb-drug interactions of the Ghanaian medicinal
plants investigated.
128
1. Introduction
Herbal medicines, increasingly used over the past decades [1,2], are usually assumed
to be harmless. Thus they are not subjected to the scrutiny of the approval process
applied to new drug applications. However, some of the herbal medicines may result in
CYP-mediated herb-drug interactions upon co-administration with prescribed drugs [3].
Similarly, interactions of plant extracts with glutathione S-transferases (GSTs) have
been reported [4]. Interactions of herbal medicines with human CYPs have been
associated with alterations in the pharmacokinetics of drugs such as midazolam [3]. The
interactions may involve both induction and inhibition of CYPs, the latter being more
common [5] and sometimes causing harmful side-effects [6]. Investigations in human
volunteers taking Hydrastis canadensis (goldenseal), 900 mg three times a day have
shown that this herbal supplement, taken to prevent common cold and upper respiratory
tract infections, significantly inhibits CYP2D6 and CYP3A4/5 by about 40% [3]. This
effect had been observed earlier in vitro [7]. Using in vitro approaches, many other
herbs and natural compounds have been identified as inhibitors of various CYPs.
Phyllanthus amarus, a medicinal plant, has been shown to possess anti-cancer
properties [8] (Table 1). Studies on inhibitory activities of a water/ethanol extract of
Phyllanthus amarus on HIV replication in vitro and ex vivo have also shown interference
with binding of HIV-1 gp120 to CD4 and inhibition of reverse transcriptase, integrase
and protease with IC50 values 2.65, 8.17, 0.48 and 21.8 µg/ml respectively [9].
Additionally, in human volunteers receiving a single dose of 1200 mg of dried
Phyllanthus amarus extract, virus replication was reduced by more than 30%, compared
to the known drug lamivudine (17%) [9]. Phyllanthus amarus extract has also been
reported to be a potent inhibitor of rat liver microsomal 7-ethoxyresorufin-O-deethylase
(CYP1A1), 7-methoxyresorufin-O-demethylase (CYP1A2) and also 7-pentoxyresorufin-
O-depentylase (CYP2B1/2) [10]. Furthermore, curcumin, a natural compound derived
from Curcuma longa, and possessing many important therapeutic activities such as
anti-cancer and anti-oxidant activities [11,12], has recently been shown to be a potent
inhibitor of recombinant human CYP3A4 and CYP2C9 [13]. Studies with Kava (Piper
methysticum) extract, a commercially available herbal anoxiolytic, product normalized to
100 µM total kavalactones, showed significant inhibition of human CYP1A2 (56%),
129
CYP2C9 (92%), CYP2C19 (86%), CYP2D6 (73%) and CYP3A4 (78%) in vitro [14]. A
case report described a coma of a woman after concomitant ingestion of kava and
alprazolam, a known CYP3A4 substrate [15]. Although direct evidence for an in vitro-in
vivo correlation is lacking, caution is imperative when kava is used in combination with
CYP3A4-substrate drugs [16]. These and many other adversities [17,18] have spurred
in vitro investigations on herbal medicines.
Herbal medicines may also modulate other important drug metabolizing enzymes
such as glutathione S-transferases (GSTs) [19]. GSTs are a super family of
multifunctional detoxification enzymes, which catalyze conjugation of glutathione (GSH)
to a wide variety of electrophilic compounds [20,21]. Inhibition of GSTs may also have
consequences on a large number of endogeneous processes, including the beneficial
inhibitory role of GSTs in drug resistance of tumors [22]. On the other hand, inhibition of
GST-mediated scavenging of electrophilic xenobiotic compounds may result in harmful
consequences [20,23].
In Ghana, herbal medicines are popular and often used by patients in
combination with prescribed drugs [24], in spite of possible adverse herb-drug
interactions. The effects of some herbal medicines such as Phyllanthus amarus extract
on rat liver CYPs has been studied [10]. However, extrapolation from animal to human
data remains unreliable [25,26]. The potential of Ghanaian herbal medicines for such
interactions with human CYPs has not yet been investigated. Since herb-drug
interactions is becoming an increasing concern [17], we selected seven Ghanaian
medicinal plants commonly employed for various ailments (Table 1) and evaluated the
potential of the aqueous extracts to cause CYP-mediated herb-drug interactions in vitro.
We investigated the effects of the plant extracts on four major human CYPs and three
major human GSTs and on human and rat liver cytosolic GSTs. Although the active
principles and phytochemical profile of the plants is not fully known, studies on
Phyllanthus amarus have shown the presence of lignans, phyllanthin and
hypophyllanthin, flavonoids such as quercetin and astragalin, ellagitannins and
hydrolysable tannins [8]. The presence of flavonoids, anthraquinones, alkaloids, tannins
and saponins in leaves of Cassia alata has also been reported [27].
1
30
Ta
ble
1.
Gh
an
aia
n m
ed
icin
al p
lan
ts,
the
ir f
am
ilie
s,
pa
rts u
se
d,
loca
l n
am
es,
the
rap
eu
tic u
se
s a
nd
vo
uch
er
sp
ecim
en
nu
mb
ers
Pla
nt
na
me
Fa
mily
P
art
use
d
Lo
ca
l n
am
e
Th
era
pe
utic u
se
Vo
uch
er
No
.
Ca
ssia
ala
ta L
. C
ae
sa
lpin
ace
ae
L
ea
ve
s
Ose
mp
a
An
ti-i
nfla
mm
ato
ry, a
nti-m
icro
bia
l G
C 4
59
33
a
nti-p
late
let
ag
gre
ga
tio
n, a
nti-d
iab
e-
tic,
an
ti-h
yp
ert
en
siv
e, a
nti-m
ala
ria
l
Ca
ssia
sia
me
a L
am
. C
ae
sa
lpin
ace
ae
L
ea
ve
s
Gye
du
a
An
ti-o
xid
an
t, a
nti-t
um
or,
G
C 4
59
32
a
nti-m
ala
ria
l
La
ctu
ca
ta
raxic
ifo
lia
Co
mp
osita
ce
ae
Le
ave
s
Nn
en
oa
A
nti-o
xid
an
t, a
nti-i
nfla
mm
ato
ry
GC
45
92
9
(Will
d.)
Sch
um
Mo
mo
rdic
a c
ha
ran
tia
L.
Cu
cu
rbita
ce
ae
L
ea
ve
s
Nyin
ya
A
nti-v
ira
l, a
nti-m
uta
ge
nic
, G
C 4
59
31
a
nti-d
iab
etic,
an
ti-i
nfla
mm
ato
ry
a
nti-c
an
ce
r, a
na
lge
sic
Mo
rin
da
lu
cid
a B
en
th.
Ru
bia
ce
ae
L
ea
ve
s
Oko
nkro
ma
A
nti-m
ala
ria
l, a
nti-d
iab
etic,
GC
45
93
0
a
nti-d
rep
an
ocyta
ry
Ph
ylla
nth
us a
ma
rus
Eu
ph
orb
iace
ae
W
ho
le p
lan
t A
wo
mm
ag
ua
kyi
An
ti-c
an
ce
r, a
nti-h
ep
atitis,
GC
47
81
3
(S
ch
um
) T
ho
nn
L
ea
ve
s,
Ste
ms
an
ti-H
IV,
an
ti-m
ala
ria
l,
R
oo
ts
an
ti-t
um
or,
Tri
da
x p
rocu
mb
en
s L
. C
om
po
sita
ce
ae
W
ho
le p
lan
t F
om
izig
be
A
nti-b
acte
ria
l, a
nti-p
roto
zo
al,
G
C 4
78
14
w
ou
nd
he
alin
g
131
Previous studies have also shown that Momordica charantia contains glycosides such
as momordin, vitamin C, flavonoids and other polyphenols [18].
2. Materials and methods
2.1. Reagents and chemicals
Methoxyresorufin was synthesized by methylation of resorufin by iodomethane in
acetone, in the presence of potassium carbonate, and confirmed by (1)H NMR [29]. The
plasmid, pSP19T7LT_2D6 containing human CYP2D6 bi-cistronically co-expressed with
human cytochrome P450 NADPH reductase was kindly provided by Prof. Ingelman-
Sundberg (Stockholm, Sweden). The plasmids BMX100/h1A2 and pCWh3A4 with CYP
NADPH reductase were donated by Dr. Michel Kranendonk (Lisbon, Portugal). The
plasmid pCWh2C9hNPR was kindly donated by Prof. F.P. Guengerich (Nashville,
Tennessee USA). Human liver cytosolic fraction representing a pool from 15 individual
donors, was obtained from CellzDirect Inc., North Carolina, (USA). Purified recombinant
human GST P1-1 was a gift from Prof. M. Lo Bello (University of Rome, Italy) and GST
A1-1 and GST M1-1 were kindly donated by Dr. Jan J.P. Bogaards (TNO Zeist, The
Netherlands). The specific activities with CDNB (1 mM) of these three isoenzymes are
24.7 U/mg, 123 U/mg and 262 U/mg protein, respectively [30]. The Ghanaian medicinal
plants were obtained from Baba Issah Yemoh of Haatso Accra, Ghana, and identified
by John Y. Amponsah, Herbarium technician at the University of Ghana herbarium,
Accra, where voucher specimens of all the plants used were deposited (Table 1). All
other chemicals were of analytical grade and obtained from standard suppliers.
2.2. Plant extracts
Aqueous extracts of the plant materials were prepared according to the method
described by Ayisi et al [31]. Briefly, 2 g of air dried samples of each of the plant
materials used, Cassia alata, Cassia siamea, Lactuca taraxicifolia, Momordica charantia
and Morinda lucida leaves, Phyllanthus amarus whole plant, leaves, stems, roots and
Tridax procumbens whole plant (Table 1) was boiled in 100 ml distilled water for 10 min
and filtered. The aqueous extracts were lyophilized and stored at –20oC until use.
132
2.3. CYP expression and membrane isolation
The plasmids BMX100/h1A2, pCWh2C9hNPR, pSP19T7LT_2D6 and pCWh3A4 were
transformed into Escherichia coli strain JM109. Expression of the CYPs was carried out
in 3-litre flasks containing 300 ml terrific broth (TB) with 1mM !-aminolevulinic acid, 0.5
mM thiamine, 400 µl/L trace elements, 100 µg/ml ampicillin, 1 mM isopropyl-"-D-
thiogalactopyranoside (IPTG), 0.5 mM FeCl3 (for CYP2D6 and CYP3A4 only) and 30
µg/ml kanamycin (for CYP3A4 only). The culture media were inoculated with 3 ml
overnight cultures of bacteria containing plasmids for the various CYPs. The cell
cultures were incubated for about 40 h at 28 oC and 125 rpm and CYP contents were
determined using the carbon monoxide (CO) difference spectra as described by Omura
and Sato [32]. Cells were pelleted by centrifugation (4000 g, 4 oC, 15 min) and
resuspended in 30 ml Tris-Sucrose-EDTA (TSE) buffer (50 mM Tris-acetate buffer pH
7.6, 250 mM sucrose, 0.25 mM EDTA). Cells were treated with 0.5 mg/ml lysozyme
prior to disruption by French press (1000 psi, 3 repeats). The membranes containing the
human CYPs were isolated by ultracentrifugation in a Beckmann 50.2Ti rotor (60 min,
100,000 g, 4 oC), resuspended in TSE buffer and stored at –80 oC until use.
2.4. Rat liver cytosol
Rat liver cytosol was prepared from untreated rats as described [33]. Briefly, isolated rat
liver samples were homogenized in 2 volumes of 50 mM potassium phosphate buffer
(pH 7.4) containing 0.9% sodium chloride, using a Potter-Elvehjem homogenizer at 4oC.
Cytosolic fractions were obtained by centrifuging the homogenate fraction for 20 min at
12000 g. The supernatant was further centrifuged for 60 min at 100000 g. Protein
concentration was determined by the method of Bradford [34]. The cytosolic fractions
were stored at –20 oC.
2.5. CYP inhibition assays
2.5.1. 7-Methoxyresorufin O-demethylation and dibenzylfluorescein O-debenzylation
Inhibition of the activities of human CYP1A2 and CYP3A4, by extracts of Cassia alata,
Cassia siamea, Lactuca taraxicifolia, Momordica charantia and Morinda lucida leaves,
Phyllanthus amarus whole plant, leaves, stems, roots and Tridax procumbens whole
133
plant, was determined by measuring the formation of resorufin from methoxyresorufin
[29] and fluorescein from dibenzylfluorescein (DBF) [35]. The extracts were first
screened at a high concentration (1000 !g/ml) for their inhibitory activities towards the
CYPs. IC50 values were subsequently only determined for extracts showing > 70%
inhibition. The assay mixture contained 0.1 M sodium phosphate buffer (pH 7.4), 5 µM
methoxyresorufin (CYP1A2) or 0.5 µM DBF (CYP3A4), 16 nM CYP1A2 or 17 nM
CYP3A4, plant extract and 100 µM NADPH in a black coaster 96-well plate in a final
volume of 200 µl. For IC50 determinations, the concentration range of each extract used
was 1.4-1000 µg/ml. The plant extracts were dissolved in water before use. Reaction
mixtures were pre-incubated for 5 min, and reactions were initiated by the addition of
NADPH and terminated after 10 min with 75 µl of 80% acetonitrile and 20% 0.5 M tris
solution (CYP1A2) or 2 N NaOH (CYP3A4). All measurements were performed in
duplicate or triplicate and at 37oC. Resorufin and fluorescein formation was measured
spectrophotometrically on a Victor2 1420 multilabel counter (!ex = 530 nm, !em = 586 nm
and !ex = 485 nm, !em = 535 nm respectively). Metabolite formation was linear for at
least 10 min (data not shown).
2.5.2. Diclofenac hydroxylation
Inhibition of the activities of human CYP2C9, by the plants extracts was determined by
measuring the inhibition of formation of 4-hydroxydiclofenac from diclofenac as
described [36] with slight modifications. Reaction mixtures (500 !l) consisted of
expressed enzyme (98 nM CYP2C9), 100 !M NADPH, 100 mM potassium phosphate
buffer (pH 7.4), substrate (6 !M diclofenac) and plant extract. The plant extracts were
initially screened for inhibitory activities towards CYP2C9 at a high concentration (1000
!g/ml). Subsequently, IC50 determinations were performed using only extracts that
showed >70% inhibition at high concentration (range 1.4-1000 !g/ml). Reactions were
initiated by adding NADPH after a pre-incubation period of 5 min. Incubations were
allowed to proceed at 37oC for 10 min and terminated by the addition of 200 !l
methanol. The incubation mixtures were centrifuged at 11250 g for 3 min, after which
the supernatant was analyzed by isocratic HPLC method. The mobile phase consisted
of 60% (v/v) 20 mM potassium phosphate buffer (pH 7.4), 22.5% (v/v) methanol and
134
17.5% (v/v) acetonitrile. Peaks were monitored by UV detection at wavelength 280 nm.
Under these conditions, retention times for 4-hydroxydiclofenac and diclofenac were 5.0
and 24.1 min respectively.
2.5.3. Dextromethorphan O-demethylation
Inhibition of CYP2D6 activity by the plant extracts were evaluated by the method of Ko
et al. [37] with slight modifications. The extracts were first screened for CYP2D6
inhibitory activity at 1000 µg/ml and then IC50 determinations were performed on
extracts showing > 70% inhibition. Briefly, the reaction mixture consisted of 4.5 µM
dextromethorphan, 18.2 nM CYP2D6 and 90.9 µM NADPH. Plant extracts were tested
in the concentration range 1.4-1000 µg/ml. Reactions were carried out at 37oC and
terminated after 40 min with 60 mM zinc sulphate solution. Product formed
(dextrorphan) was measured using an isocratic HPLC with fluorescence detection
method (!ex=280 nm, !em=310 nm), and a C18 column (100 mm x 3 mm, 5 µm particle
size, Chromspher). Product formation was linear for at least 45 min. The mobile phase
consisted of 24% (v/v) acetonitrile and 0.1% (v/v) triethylamine adjusted to pH 3 with
perchloric acid. The flow rate was 0.6 ml/min. The retention time of dextrorphan was 3.4
min and of dextromethorphan, 24.5 min.
2.6. GST inhibition assays
Inhibition of the activities of cytosolic GSTs by the plant extracts was assessed as
described previously [30] with slight modifications. GST-mediated conjugation of 1-
chloro-2,4-dinitrobenzene (CDNB) to glutathione (GSH) was measured at room
temperature (25oC) using an Ultrospec 2000 Pharmacia Biotech UV/visible
spectrophotometer at the wavelength of 340 nm for 2 min. Incubation mixtures (1 ml)
contained 0.1 M potassium phosphate buffer pH 6.5, 400 µM CDNB, 1 mM GSH, and
GST enzymes (1.04 µg/ml GST A1-1, 1.48 µg/ml GST M1-1, 1.25 µg/ml GST P1-1, or
32.0 µg/ml and 1.7 µg/ml human and rat liver cytosol, respectively). The plant extracts
were each firstly tested at a concentration of 500 µg/ml. Subsequently, extracts showing
> 70% inhibitions were selected for IC50 determination in the concentration range 0.69-
135
500 µg/ml. All assays were linear functions of protein concentration and of time for at
least 2 min.
3. Results
3.1. Inhibition of CYPs
The method used, in preparation of aqueous plant extracts in the present study, is
similar to that used in preparation of these aqueous extracts for consumption in Ghana,
where the plant material is boiled for at least 10 minutes. The lyophilization and
resuspension of plant material is not likely to cause significant changes in the properties
of the extracts. This procedure is also been used by several investigators [8,31,41,42].
The plant extracts that showed >70% inhibition of CYP1A2 (methoxyresorufin O-
methylase) activity at 1000 µg/ml include, Cassia alata and Lactuca taraxicifolia leaves,
Phyllanthus amarus whole plant, leaves, stems, roots, and Tridax procumbens whole
plant (data not shown). IC50 values subsequently determined (Table 2) showed
strongest inhibition of CYP1A2 by Cassia alata leaves (IC50 = 28.5 µg/ml), Lactuca
taraxicifolia leaves (IC50 = 39.9 µg/ml) and Phyllanthus amarus stems and whole plant
(IC50 = 38.1 and 47.5 µg/ml, respectively). At a concentration of 1000 µg/ml, all plant
extracts except Morinda lucida leaves, and Cassia alata leaves, showed >70%
inhibitory activity towards CYP2C9 (data not shown). Only Phyllanthus amarus, stems,
roots and whole plant exhibited comparatively strong inhibitory activities towards
CYP2C9, with IC50 values of 63.4, 74.1 and 86.0 µg/ml, respectively (Table 2). Two out
of the seven plants tested in this study, showed > 70% inhibition of CYP2D6 at a
concentration of 1000 µg/ml (data not shown). Only extracts of Cassia alata leaves and
Phyllanthus amarus demonstrated > 70% inhibition of CYP2D6 activity. IC50 values
subsequently obtained were ranging from 45.8-182.0 µg/ml. The roots of Phyllanthus
amarus showed the strongest inhibition of CYP2D6, and the leaves the weakest, with
IC50 values of 45.8 and 182.0 µg/ml, respectively (Table 2).
136
Table 2. IC50 values (µg/ml) for human recombinant CYPs with aqueous plant extracts
Plant extract CYP1A2 CYP3A4 CYP2C9 CYP2D6
Cassia alata (Leaves) 28.3 + 2.42 158.8 + 9.62 nd 165.5 + 7.50
Cassia siamea (Leaves)
nd nd 346.5 + 20.93 nd
Lactuca taraxicifolia (Leaves)
39.9 + 11.93 nd 396.3 + 2.26 nd
Momordica charantia (Leaves)
nd nd 425.9 + 2.33 nd
Morinda lucida (Leaves)
nd nd nd nd
Phyllanthus amarus (Leaves)
92.2 + 12.85 79.2 + 0.41 127.5 + 3.96 182.0 + 4.81
Phyllanthus amarus (Roots)
134.3 + 17.25 80.4 + 5.29 74.1 + 0.10 45.8 + 2.40
Phyllanthus amarus (Stems)
38.1 + 3.14 146.8 + 17.54 63.4 + 4.53 164.0 + 5.37
Phyllanthus amarus (Whole plant)
47.5 + 1.27 97.0 + 18.74 86.0 + 5.26 48.5 + 0.95
Tridax procumbens (Whole plant)
127.5 + 6.08 nd nd nd
Values are means + standard deviation of at least two experiments as described in the Methods section.
CYP3A4 was also inhibited by extracts of Cassia alata leaves and Phyllanthus amarus,
with >70% inhibition at 1000 µg/ml (data not shown) and IC50 values of these plant
materials were in the range 79.2-158.8 µg/ml (Table 2).
3.2. Inhibition of GSTs
Inhibition of GST mediated conjugation of GSH to CDNB by the plant extracts was
investigated with GSTs in human and rat liver cytosol and recombinant human GSTA1-
1, GSTM1-1 and GSTP1-1 isoenzymes. When tested at 500 µg/ml, all plant extracts
showed >70% inhibition of rat liver cytosolic GSTs except Cassia siamea and Morinda
lucida leaf extracts. The human liver cytosol i.e. GSTs activities, were inhibited at 500
137
µg/ml by >70% by six of the plant extracts, whereas Cassia siamea, Lactuca
taraxicifolia and, Morinda lucida leaves and Tridax procumbens whole plant showed
<70% inhibition (data not shown). The corresponding IC50 values obtained with rat liver
cytosol were in the range of 8.5-139.4 µg/ml, whilst those with human liver cytosol were
25.2-95.5 µg/ml (Table 3). Determination of the effects of the plant extracts (500 µg/ml)
on human recombinant GSTs, showed that all extracts, except Lactuca taraxicifolia
leaves, inhibited GST M1-1 by >70%, whilst GST A1-1 and GST P1-1 were each
inhibited by only two plants, including Phyllanthus amarus and Momordica charantia or
Cassia alata respectively.
Table 3. IC50 values (!g/ml) for human and rat liver cytosolic GSTs with aqueous plant extracts
Plant extract Human liver Rat liver
Cassia alata (Leaves) 83.2 + 6.94 41.9 + 3.09
Cassia siamea (Leaves) nd nd
Lactuca taraxicifolia (Leaves) nd 112.8 + 4.60
Momordica charantia (Leaves) 95.5 + 0.83 29.0 + 1.33
Morinda lucida (Leaves) nd nd
Phyllanthus amarus (Leaves) 58.8 + 7.79 50.0 + 4.74
Phyllanthus amarus (Roots) 47.0 + 4.54 8.5 + 0.98
Phyllanthus amarus (Stems) 25.2 + 4.67 28.9 + 0.24
Phyllanthus amarus (Whole plant) 46.6 + 1.14 21.6 + 1.70
Tridax procumbens (Whole plant) nd 139.4 + 3.46
Values are means + standard deviation of two experiments as described in the Methods section.
nd: not determined because percent inhibition at 500 !g/ml was < 70%
IC50 values for the extracts showing >70% inhibitions are shown in Table 4. The IC50
values for extracts inhibiting GST A1-1 ranged from 8.9-159.0 µg/ml, with Phyllanthus
amarus roots being the strongest inhibitor and Momordica charantia leaves the
138
weakest. With respect to GST M1-1, the IC50 values obtained were in the range 3.6-50.0
µg/ml, whereas that of GST P1-1 was 68.6-157.0 µg/ml (Table 4).
Table 4. Inhibition of human recombinant GSTs by aqueous plant extracts (IC50, !g/ml)
Plant extract GSTA1-1 GSTM1-1 GSTP1-1
Cassia alata (Leaves) nd 5.6 + 0.02 68.6 + 15.64
Cassia siamea (Leaves) nd 50.0 + 1.59 nd
Lactuca taraxicifolia (Leaves) nd nd nd
Momordica charantia (Leaves) 159.0 + 10.82 16.5 + 1.19 nd
Morinda lucida (Leaves) nd 42.4 + 2.69 nd
Phyllanthus amarus (Leaves) 36.6 + 7.92 15.6 + 1.73 157.0 + 21.10
Phyllanthus amarus (Roots) 8.9 + 0.50 3.6 + 0.21 98.2 + 6.02
Phyllanthus amarus (Stems) 42.0 + 6.65 8.8 + 0.42 133.8 + 22.00
Phyllanthus amarus (Whole plant) 47.5 + 0.99 9.7 + 0.99 96.2 + 16.33
Tridax procumbens (Whole plant) nd 40.6 + 12.51 nd
Values are means + standard deviation of two experiments as described in the Methods section.
nd: not determined because percent inhibition at 500 !g/ml was < 70%
In these cases Phyllanthus amarus roots and Cassia alata leaves were the strongest
inhibitors whilst Cassia siamea and Phyllanthus amarus leaves were the weakest
inhibitors respectively.
4. Discussion
In the present study, inhibitory potencies of seven Ghanaian medicinal plants towards
recombinant human CYP1A2, CYP2C9, CYP2D6 and CYP3A4, recombinant human
GSTA1-1, GSTM1-1 and GSTP1-1 and human and rat liver cytosolic GSTs were
investigated. Cassia alata extract was the most potent inhibitor of CYP1A2 activity.
Approximately, a 3 or 4 times more potent activity was observed with Cassia alata as
compared to the other extracts. A recent in vitro study reported a potent inhibition (IC50
139
= 7.8 µg/ml) of CYP1A1/2 (methoxyresorufin O-demethylase) activity in rat liver
microsomes, by methanol extracts of Phyllanthus amarus [10]. Interestingly, aqueous
extracts of Phyllanthus amarus have been found to also possess anti-carcinogenic
activity towards 20-methylcholanthrene induced sarcoma in mice [8]. A dose of 750
mg/kg body weight resulted in 75% survival in 180 days as compared to an untreated
control group with no survivors [8]. The observed anti-carcinogenic effect and the low
IC50 value for inhibition of CYP1A1/2 [10], suggest possible of interference at the level of
CYP1A1/2-mediated bioactivation.
CYP2C9 was inhibited by extracts of Phyllanthus amarus, Cassia siamea,
Lactuca taraxicifolia and Momordica charantia by more than 70%. However, comparing
IC50 values, extracts of Phyllanthus amarus inhibited CYP2C9 about 3-6 fold more
strongly than the other extracts. Extracts of Phyllanthus amarus, as well as Cassia alata
were the only extracts with more than 70% inhibition of CYP2D6 at the highest
concentration used. The roots and whole plant extracts of Phyllanthus amarus showed
about 3-fold more potent activities than the leaves, stems and Cassia alata extracts.
Inhibition of CYP3A4 activity by >70% was also observed with the extracts of
Phyllanthus amarus and Cassia alata. CYP3A4 is responsible for the metabolism of
about 50% of the drugs currently on the market [38]. Therefore herbs inhibiting this
enzyme do have a relatively high potential to cause significant herb-drug interactions.
Although the inhibitory constituents of the plant extracts used are not known,
flavonoids which are nearly ubiquitously present in plants [39], have been found to
exhibit varying potencies of inhibitory activities towards human CYP1A2. For examples,
quercetin, chrysin, galangin, morin and apigenin inhibited CYP1A2 with IC50 values of
169.0, 0.2, 3.1, 9.5 and 1.4 µM, respectively [40]. Especially in the case of Phyllanthus
amarus, flavonoids including quercetin have been identified [8]. Earlier studies on
extracts of Ginkgo biloba reported significant inhibition of CYP2C9 with Ki = 14.0 µg/ml
by the whole extract, whilst the flavonoidic and terpenoidic fractions showed inhibitory
activities with Ki values of 4.9 and 15 µg/ml, respectively [41]. Methanol extract of
Ginkgo biloba was also shown to inhibit recombinant human CYP3A4 with Ki value
155.0 µg/ml, whereas the flavonoidic fraction shown to be responsible for the
neuroprotective effect of this plant, showed Ki value of 43.0 µg/ml [41]. In addition,
140
recent studies on Indonesian medicinal plants showed the water/methanol extract of
Catharanthus roseus to be a potent inhibitor of CYP2D6 (IC50 = 11.0 µg/ml) [42]. Two
classes of compounds found in Catharanthus roseus, active as wound healing and anti-
diabetic agent, are alkaloids and tannins [43]. A methanol soluble dried aqueous extract
of this plant with the alkaloids, ajmalicine, serpentine and vindoline showed 94%
inhibition of CYP2D6 at 25 µg/ml [44].
GST activity has also been shown to be modulated by natural plant products [45].
The present study has revealed the GST inhibitory potential of seven Ghanaian
medicinal plants. Amongst the extracts tested those of Phyllanthus amarus as well as
Cassia alata and Momordica charantia were the strongest inhibitors of human liver
cytosolic GSTs. The human liver generally expresses high levels of the GST alpha (A)
class and insignificant levels of the pi (P) class [46]. The mu class of GSTs, is also
expressed in human liver, predominantly as GSTM1 and GSTM4 [46]. Rat liver GSTs
were inhibited by all the plant extracts except the extracts of Cassia siamea and
Morinda lucida. The strongest inhibition of rat liver GSTs was observed with the root
extracts of Phyllanthus amarus and not the stem, unlike in the human samples. No
correlation was found between the cytosolic human and rat data. This is likely attributed
to differences in enzyme structures and catalytic activities of GSTs between species
[26].
Investigating the effect of the plant extracts on human recombinant GSTA1-1,
GSTM1-1 and GSTP1-1 revealed significant differences in their inhibitory potential
towards these enzymes. Our study has shown a strong inhibitory potential of
Phyllanthus amarus extracts towards GSTA1-1 compared to the other extracts. Most of
the plant extracts inhibited GSTM1-1, while GSTP1-1 was also inhibited by extracts of
Phyllanthus amarus as well as Cassia alata. Flavonoids have been shown to inhibit
human GSTs in blood platelets [47]. Quercetin, kaempferol and genistein have been
shown to significantly inhibit GSTM1-1 and M2-2 [48] and quercetin GSTP1-1 [49].
Inhibition of GSTs by the Phyllanthus amarus as well as Cassia alata could possibly
result in reduced protection from toxic effects of electrophilic substances.
In conclusion, this study has shown the inhibitory potential of aqueous extracts of
seven Ghanaian medicinal plants, towards four of the most important human CYPs.
141
Most of the plant extracts appear to lack a strong potential to inhibit the CYPs tested.
Phyllanthus amarus generally appeared to be the most potent inhibitor of the CYPs.
Phyllanthus amarus was also the strongest inhibitor of the GSTs tested. Since usually
very high concentrations of plant extracts are consumed, regular human consumption of
these extracts could well inhibit GSTA1-1 and GSTM1-1 activities in the liver and
GSTP1-1 in erythrocytes. Inhibition of GSTs may be beneficial for cancer therapy, but in
normal cells GST inhibition can also result in increased toxicity due to a reduced
protection against electrophilic chemicals or metabolites. Therefore there is a need to
further identify plant products with GST inhibitory potentials. Further investigations are
required to determine the clinical implications of the present CYP and GST inhibition
results and to identify the compounds in the plant extracts responsible for the observed
inhibitory activities.
References
1. Eisenberg, D.M., Davis, R., Ettner, S, Appel S., Wilkey, S., Van Rompay M., Kessler, R.C., 1998. Trends in alternative medicine use in the Unites States, 1990-1997: results of a follow-up national survey. JAMA 280:1569-1575.
2. Menniti-Ippolito, F., Gargiulo, L., Bologna, E., Forcella, E., Raschetti, R., 2002. Use of unconventio- nal medicine in Italy: a nation-wide survey. Eur J Clin Pharmacol 58:61-64.
3. Gurley, B.J., Gardner, S.F., Hubbard, M.A., Williams, D.K., Gentry, W.B., Khan, I.K., Shah, A., 2005. In vivo effects of goldenseal, kava kava, black cohosh, and valerian on human cytochrome P450 1A2, 2D6, 2E1, and 3A4/5 phenotypes. Clin Pharmacol Ther 77:415-426.
4. Coruh, N., Celep, A.G.S., Ozgokce, F., 2007. Antioxidant properties of Prangos ferulacea (L.) Lindl., Chaerophllum macropodum Boiss. and Heracleum persicum Desf. from Apiaceae family used as food in Eastern Anatolia and their inhibitory effects on glutathione S-transferase. Food Chem 100:1237-1242.
5. Zafar, A., Sharif, M.D., 2003. Pharmacokinetics, metabolism, and metabolism of atypical antipsychotics in special populations. Primary care companion J Clin Psychiatry 5:22-25.
6. Ioannides, C., 2002. Pharmacokinetic interactions between herbal remedies and medicinal drugs. Xenobiotica 32:451-478.
7. Budzinski, J.W., Foster, B.C., Vandenhoek, S., Arnason, J.T., 2000. An in vitro evaluation of human cytochrome P450 3A4 inhibition by selected commercial herbal extracts and tinctures. Phytomedi- cine 7:273-282.
8. Rajeshkumar, N.V., Joy, K.L., Kuttan, G., Ramsewak, R.S., Nair, M.G., Kuttan, R., 2002. Antitum- our and anticarcinogenic activity of Phyllanthus amarus extract. J Ethnopharmcol 81:17-22.
9. Notka, F., Meier, G., Waner, R., 2004. Concerted inhibitory activities of Phyllanthus amarus on HIV replication in vitro and ex vivo. Antiviral Res 64:93-102.
10. Hari Kumar, K.B., Kuttan, R., 2006. Inhibition of drug metabolizing enzymes (cytochrome P450) in vitro as well as in vivo by Phyllanthus amarus Schum & Thonn. Biol Pharm Bull 29:1310-1313.
11. Leu, T-H., Maa, M-C., 2002. The molecular mechanisms for the antitumorigenic effect of curcumin. Curr Med Chem-Anti-Cancer Agents 2:357-370.
12. Cole, G.M., Morihara, T., Lim, G.P., Yang, F., Begum, A., Frautschy, S.A., 2004. NSAID and antioxidant prevention of Alzheimer's disease: Lessons from in vitro and animal models. Annals NY Acad Sci 1035:68-84.
142
13. Appiah-Opong, R., Commandeur, J.N.M., van Vugt-Lussenburg, B., Vermeulen, N.P.E., 2007. Inhibition of human recombinant cytochrome P450s by curcumin and curcumin decomposition products. Toxicology 235:83-91.
14. Mathews, J.M., Etheridge, A.S., Black, S.R., 2002. Inhibition of human cytochrome P450 activities by kava extract and kavalactones. Drug Metab Disops 30:1153-1157.
15. Almeida, J.C., Grimsley, E.W., 1996. Coma from the health food store: Interaction between kava and alprazolam. Annals Intern Med 125:940-941.
16. Sparreboom, A., Cox, M.C., Acharya, M.R., Figg, W.D., 2004. Herbal remedies in the United States: Potential adverse interactions with anticancer agents. J Clin Oncol 22:2489-2503.
17. Rupika, D., Andrew, W.C.G., 2004. Herbal interactions involving cytochrome P450 enzymes: A mini review. Tox Rev 23:239-249.
18. Izzo, A.A., 2004. Herb-drug interactions: An overview of the clinical evidence. Fundamental Clin Pharmacol 19:1-16.
19. Hayeshi, R., Mukanganyama, S., Hazra, B., Abegaz, B., Hasler, J., 2004. The interaction of selected natural products with human recombinant glutathione transferases. Phytother Res 18:877- 883. 20. Commandeur, J.N., Stijntjes, G.J., Vermeulen, N.P., 1995. Enzymes and transport systems involved in the formation and disposition of glutathione S-conjugates. Role in bioactivation and detoxication mechanisms of xenobiotics. Pharmacol Rev 47:271-330. 21. Frova, C., 2006. Glutathione transferases in the genomics era: New insights and perspectives.
Biomol Eng 23:149-169. 22. Mukanganyama, S., Widersten, M., Naik, Y.S., Mannervik, B., Hasler, J.A., 2002. Inhibition of glutathione S-transferases by antimalarial drugs possible implications for circumventing anticancer drug resistance. Int J Cancer 97:700-705. 23. Capela, J.P., Macedo, C., Branco, P.S., Ferreira, L.M., Lobo, A.M., Fernandes, E., Remiao,
F., Bastos, M.L., Dirnagl, U., Meisel, A., Carvalho, F., 2007. Neurotoxicity mechanisms of thioether ecstasy metabolites. Neuroscience 146:1743-1757.
24. Buabeng, K.O., Duwiejua, M., Dodoo, A.N., Matowe, L.K., Enlund, H., 2007. Self-reported use of anti-malarial drugs and health facility management of malaria in Ghana. Malar J 6:85.
25. Eagling, V.A., Tjia, J.F., Back, D.J., 1998. Differential selectivity of cytochrome P450 inhibitors against probe substrates in human and rat liver microsomes. Br J Clin Pharmacol 45:107-114.
26. Aurbek, N., Thiermann, H., Szinicz, L., Eyer, P., Worek, F., 2006. Application of kinetic-based computer modelling to evaluate the efficacy of HI 6 in percutaneous VX poisoning. Toxicology 224:74-80.
27. Idu, M., Oronsaye, F.E., Igeleke, C.L., Omonigho, S.E., Omogbeme, O.E., Ayinde, B.A., 2006. Preliminary Investigation on the Phytochemistry and Antimicrobial Activity of Senna alata L. Leaves. J Appl Sci 6:2481-2485.
28. Raj, S.K., Khan, M.S, Singh, R., Kumari, N., Prakash, D., 2005. Occurrence of yellow mosaic Gem- iniviral disease on bitter gourd (Momordica charantia) and its impact on phytochemical contents. Int. J. Food Sci Nutr 56:185-192.
29. Burke, M.D., Thompson, S., Elcombe, C.R., Halpert, J., Haaparanta, T., Mayer, R.T., 1985. Ethoxy-, pentoxy- and benzyloxyphenoxazones and homologues: a series of substrates to distinguish between different induced cytochromes P-450. Biochem Pharmacol 34:3337-3345.
30. van Haaften, R.I.M., Haenen, G.R.M.M., van Bladeren, P.J., Bogaards, J.J.P., Evelo, C.T.A., Bast,
A., 2003. Inhibition of various glutathione S-transferase isoenzymes by RRR-!-tocopherol. Toxicol
in vitro 17:245-351. 31. Ayisi, N.K., Nyadedzor,C., 2003. Comparative in vitro effects of AZT and extracts of Ocimum
gratissimum, Ficus polita, Clausena anisata, Alchornea cordifolia, and Elaeophorbia drupifera against HIV-1 and HIV-2 infections. Antiviral Res 58:25-33.
32. Omura, T., Sato, R., 1964. The carbon monoxide binding pigment of liver microsomes. I. Solubiliz- ation, purification and properties. J Biol Chem 239:2379-2385.
33. Rooseboom, M., Vermeulen, N.P.E., Groot, E.J., Commandeur J.N.M., 2002. Tissue distribution of cytosolic beta-elimination reactions of selenocysteine Se-conjugates in rat and human. Chem Biol Interact 140:243-264.
34. Bradford, M. M., 1976. A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248-254.
143
35. Stresser, D.M., Blanchard, A.P., Turner, S.D., Erve, J.C.L., Dandeneau, A.A., Miller, V.P., Crespi C.L., 2000. Substrate-dependent modulation of CYP3A4 catalytic activity: Analysis of 27 test comp- ounds with 4 fluorometric substrates. Drug Metab Dispos 28:1440-1448.
36. Walsky, R.L., Obach, R.S., 2004. Validation assays for human cytochrome P450 activities. Drug Metab Dispos 32:647-660.
37. Ko, J.W., Desta, Z., Soukhova, N.V., Tracy, T., Flockhart, D.A., 2000. In vitro inhibition of the cytochrome P450 (CYP450) system by the antiplatelet drug ticlopidine: potent effect on CYP2C19 and CYP2D6. Br J Clin Pharmacol 49:343-351. 38. Sarlis, N.J., Gourgiotis, L., 2005. Hormonal effects on drug metabolism through the CYP system:
Perspectives on their potential significance in the era of pharmacogenomics. Curr Drug Targets- Immune Endocrine & Metabolic disorders 5:439-448.
39. Middleton, E. Jr., Kandaswami, C., Theoharides, T.C., 2000. The effect of plant flavonoids on mammalian cells: Implications for inflammation, heart disease and cancer. Pharmacol Rev 52:673- 751.
40. Lee, H., Yeom, H., Kim, Y.G., Yoon, C.N., Jin, C., Choi, J.S., Kim, B-R., Kim, D-H., 1998. Structure- related inhibition of human hepatic caffeine N3-demethylation by naturally occurring flavonoids. Biochem Pharmacol 55:1369-1375.
41. Gaudineau, C., Beckerman, R., Welbourn, S., Auclair, K., 2004. Inhibition of human P450 enzymes by multiple constituent of Ginkgo biloba extract. Biochem Biophys Res Commun 318:1072-1078.
42. Usia, T., Iwata, H., Hiratsuka, A., Watabe, T., Kadota, S., Tezuka, Y., 2006. CYP3A4 and CYP2D6 inhibitory activities of Indonesian medicinal plants. Phytomedicine 13:67-73.
43. Nayak, B.S., Pereira, L.M.P., 2006. Catharanthus roseus flower extract has wound-healing activity in Sprague Dawley rats. BMC Complement Altern Med 6:41.
44. Usia, T., Watabe, T., Kadota, S., Tezuka, Y., 2005. Cytochrome P450 2D6 (CYP2D6) inhibitory constituents of Catharanthus roseus. Biol Pharm Bull 28:1021-1024.
45. Zhang, K., Wong, K.P., Chow, P., 2003. Conjugation of chlorambucil with GSH by GST purified from human colon adenocarcinoma cells and its inhibition by plant polyphenols. Life Sci 72:2629- 2640.
46. Eaton, D.L., Bammler, T.K., 1999. Concise review of glutathione S- transferases and their significa- nce to toxicology. Toxicol Sci 49:156-164.
47. Ghazali, R., Waring, R.H., 1999. Effects of flavonoids on glutathione-S-transferase in human blood platelets, rat liver, rat kidney, HT-29 colon adenocarcinoma cell-lines: potential in drug metabolism and chemoprevention. Med Sci Res 27:449-451.
48. Burg, D., Mulder, G.J., 2002. Glutathione conjugates and their synthetic derivatives as inhibitors of glutathione-dependent encimes involved in cancer and drug resistance. Drug Metab Rev 34:821- 863.
49. van Zanden, J.J., Geraets, L., Wortelboer, H.M., van Bladeren, P.J., Rietjens, I.M.C.M., Cnubben, N.H.P., 2004. Structural requirements for the flavonoid-mediated modulation of glutathione S- transferase P1-1 and GS-X pump activity in MCF7 breast cancer cells. Biochem Pharmacol 67:1607-1617.
144
145
Summary, conclusions and perspectives
146
Chapter 7
Summary and conclusions
Cytochrome P450 (CYP)-mediated drug–drug interactions are major causes of
attrition during drug development, black-box warnings of marketed drugs or
withdrawals of drugs from the market, due to adverse effects [1,2]. Inhibition of CYP
activity is a major cause of drug-drug interactions, since it results in accumulation of
drugs which could otherwise be metabolized into less toxic products. Alternatively,
CYP induction may also contribute to serious drug-drug interactions. Thus, early
evaluation of new chemical entities (NCEs) for both inhibitory and inducing drug-
drug interaction potentials is considered useful and cost effective in drug
development. Inhibition and induction of glutathione S-transferases (GSTs) by drugs
and other xenobiotics may also results in toxicities and adverse drug reactions, since
these enzymes are primary detoxification, and in some cases toxification enzymes in
the body [3].
The aim of the investigations described in this thesis is to evaluate the drug-drug
-food and -herb interaction potentials at the level of CYP and GST inhibition of plant
and plant derived components. The CYP and GST inhibitory potentials of curcumin,
which was used as a model compound, were compared with that of three series of
synthetic curcumin analogues including, 2,6-dibenzylidenecyclohexanone (A series),
2,5-dibenzylidenecyclopentanone (B series) and 1,4-pentadiene-3-one (C series) [4].
Structure-activity relationships were analyzed to guide future synthesis of
compounds with less (or more) inhibitory potencies towards CYPs and GSTs. The
inhibitory potentials of seven Ghanaian medicinal plants on CYP and GST activities
were also assessed.
In Chapter 1 a general introduction of the background of CYP-mediated drug-drug
interactions, due to CYP-inhibition and -induction, and the detoxification and
toxification roles of GSTs are discussed. Co-administration of two or more drugs has
been shown to be a potential cause of drug-drug interactions with possible serious
adverse side effects [5]. CYPs contribute significantly to the elimination of drugs
147
from the body through CYP-mediated metabolism. Of the human CYP isoenzymes a
majority is important for the metabolism of drugs currently on the market [6]. These
enzymes include CYP3A4, CYP2D6, CYP2C9, CYP2C19, CYP2B6, CYP2A6 and
CYP1A2. Knowledge of CYP inhibitory properties of NCEs and analysis of structure-
activity relationships is generally seen as necessary during the early stages of drug
discovery and development to guide synthesis of safer drug candidates and to
minimize losses due to attrition. Since natural products, including foods and herbal
medicines, have also been shown to inhibit CYPs and to cause potentially significant
adverse effects, these products should be investigated for possible drug-herb/food
interactions as well [7]. The same holds true for the GST-inhibitory properties of
NCEs, drug candidates, foods and herbal or other plant products, therefore they
should be assessed as well in order to avoid potential toxicity. In the present studies,
human GSTA1-1, GSTM1-1 and GSTP1-1 were selected as targets.
Curcumin, a derivative of Curcuma longa and known to possess many important
pharmacological properties was used as a model compound in the present
investigations. The unique availability of a series of 40 synthetic structural analogues
of curcumin, 34 of which with appropriate solubility properties, (Figure 1) allowed us
to investigate structure-activity relationships for both the CYP- and GST-
isoenzymes. Using the experimental methods developed, we also investigated the
inhibitory potential of the Ghanaian medicinal plant extracts of, Cassia alata, Cassia
siamea, Lactuca taraxicifolia, Momordica charantia and Morinda lucida, Phyllanthus
amarus and Tridax procumbens toward the human CYP- and GST-isoenzymes.
OO
OH
O
OH
OCH3
CH3
Curcumin
33
1
2
O
R
R
R R
R
R
2
1
1
33
R1
R
R
O
R
R
R
22
1
33
R1
R
R
O
R
R
R
22
2,6-dibenzylidenecyclohexanone (A) 2,5-dibenzylidenecyclopentanone (B)
1,4-pentadiene-3-one (C)
Figure 1. Chemical structures of curcumin and curcumin analogues.
148
Chapter 2 presents a brief review of studies that have been performed to investigate
pharmacokinetic, metabolism and drug-drug interaction properties of curcumin.
Animal experiments performed, using radioactivity to monitor curcumin
concentrations in plasma and tissues after intraperitoneal administration, showed
low concentrations of about 5 pmol/ml in rat plasma, and rapid removal of the parent
compound from the tissues [8]. Similarly, phase I clinical trials on curcumin have
revealed low plasma concentrations of 11 nmol/L upon oral administration of 3.6 g of
Curcumin [9]. In order to enhance the bioavailability of curcumin, it has been
administered with piperine, solubilized with N,N-dimethylacetamide, polyethylene
glycol (PEG 400) and 40% isotonic dextrose or formulated with micelles or
phoshatidylcholine [10-13]. In addition, curcumin nanoparticles, liposomal curcumin
and structural analogues of curcumin have also been employed to enhance
bioavailability. The highest bioavailability reached (154% in rats) was with a
curcumin/piperine combination [11]. Curcumin has been shown to be metabolized in
sub cellular liver fractions of mouse, rat and humans in a similar way. Phase I
metabolites include reductive and oxidative metabolites of curcumin (including
hexahydrocurcuminol, hexahydro- and tetrahydrocurcumin), whilst phase II
metabolites were predominantly glucuronides and sulfates [14]. Early in vitro studies
with rat liver microsomes and cytosol showed that curcumin is a potent inhibitor of
CYP1A1/2, a less potent inhibitor of CYP2B1 and a potent inhibitor of GSTs [15].
Curcumin has been shown to inhibit strongly cytosolic GSTs isolated from human
melanoma cells [16]. In this thesis comprehensive inhibition has been done with
individual human CYPs and GSTs as well as with cytosolic fractions.
Chapter 3 focuses on the inhibitory activities of curcumin and curcumin
decomposition products towards the 5 major human recombinant CYPs, responsible
for the metabolism of majority of currently marketed drugs. Curcumin inhibited the
enzymes tested in decreasing order of potency (IC50) as follows: CYP2C9 >
CYP3A4 > CYP2B6 > CYP1A2 > CYP2D6 [17]. Competitive inhibition was observed
with CYP1A2, CYP2B6 and CYP3A4, whereas CYP2C9 and CYP2D6 were inhibited
149
non-competitively. Four decomposition products of curcumin did not show any
significant inhibition towards the human CYPs tested. Mechanism-based or time-
dependent inhibition by curcumin was neither observed with any of the 5 human
enzymes. Curcumin is a potent inhibitor of CYP2C9 and CYP3A4 and a moderate
inhibitor of CYP2B6, CYP1A2 and CYP2D6, with IC50 values ranging from 4.3 to
50.3 !M. It was concluded that, the inhibitory activity of curcumin towards CYP3A4
may well have implications for drug-drug interactions in the intestine, rather than in
the liver when the intestines are exposed to high concentrations upon oral ingestion
together with drugs metabolized by this enzyme. Further studies are needed to
establish the clinical implications of these results.
In Chapter 4 the inhibitory activities of thirty-three (selected out of 40) compounds
belonging to three series of curcumin analogues, 2,6-dibenzylidenecyclohexanone
(A series), 2,5-dibenzylidenecyclopentanone (B series) and 1,4-pentadiene-3-one (C
series) substituted analogues [4] towards the 5 major human recombinant CYPs
were experimentally determined. Subsequently, structure-activity relationships were
analyzed using the MOE software. Most of the curcumin analogues exhibited low
inhibitory activities towards the CYPs tested, when compared to curcumin itself [17].
Six compounds were potent inhibitors of CYP1A2, three potent towards CYP2C9
and two towards CYP2D6. None of the 2,6-dibenzylidenecyclohexanone derivatives
(A series) inhibited CYP3A4, CYP2B6 and CYP2D6, whilst one compound strongly
inhibited CYP2C9. The MOE-QSAR analysis lead to the conclusion that electrostatic
and hydrophobic descriptors, notably PEOE_VSA_FPNEG and PEOE_VSA_FHYD
are important factors of the compounds especially relating to inhibition of CYP2C9
and CYP2D6. The results of these studies are not only important because of the
new insights in structural properties for CYP-inhibition, but also because they
identified the structural analogues of curcumin without CYP-inhibitory properties. A
larger number of curcumin analogues will be required to enhance the QSAR
prediction of drug-drug interaction potentials of these compounds against all
individual CYPs.
150
GSTs are important phase II enzymes, involved in detoxification and toxification of
xenobiotics and in multidrug resistance in chemotherapy. In Chapter 5, the results
of investigations on the GST inhibitory potentials of curcumin itself and three series
of 34 (selected out of 40) structural analogues of curcumin, as well as structure-
activity relationships are presented and discussed. Curcumin was shown to be a
potent inhibitor of human recombinant GSTA1-1, GSTM1-1 and GSTP1-1 and
interestingly, the results were recently confirmed by Hayeshi et al. [18]. As observed
with CYP inhibition (Chapter 4), most of the curcumin-analogues lacked or
demonstrated weak inhibitory activities towards the human GST-isoenzymes,
GSTA1-1, GSTM1-1 and GSTP1-1, when compared to curcumin. The 2,5-
dibenzylidenecyclopentanone (B series) derivatives of curcumin showed no
significant inhibition of GSTP1-1. The most potent inhibitors, notably of GSTM1-1
and GSTP1-1, with IC50 values ranging from 0.2 to 6.1 µM were predominantly the
1,4-pentadiene-3-one derivatives of curcumin (C series). The inhibitory activities of
B14 and C10 towards GSTA1-1, with IC50 values of 0.5 and 0.6 µM respectively, and
towards human liver cytosolic GSTs could have implications for GST-mediated
protection against drug toxicities in the liver due to the high hepatic levels of this
GST-isoform [19]. This is in contrast to GSTP1-1, which is primarily expressed in
erythrocytes [20]. MOE-based QSAR analyses amongst others have delineated the
relevance of van der Waal’s surface area and lipophilicity factors (SMR_VSA7,
SlogP_VSA4 and dipole, PEOE_VSA-0, SlogP_VSA8 and PEOE_RPC,
respectively) for the inhibition of GSTA1-1 and GSTM1-1. The results of this chapter
may be useful in the design and synthesis of curcumin analogues with either more or
with less susceptibility to GST inhibition. This is important because of the role of
GSTs in detoxification, toxification of xenobiotics and multidrug resistance in
chemotherapy.
Herbal products and traditional medicines have also been shown to be able to inhibit
important drug metabolizing enzymes such as CYPs and GSTs, thus resulting in
clinically relevant adverse drug reactions or toxicities [21]. Therefore, Chapter 6
focuses on the interactions between the major human drug metabolizing CYPs and
151
GSTs and a selection of the most important Ghanaian medicinal plants. A standard
procedure for preparation of aqueous extracts of the seven Ghanaian plants,
involved freeze drying of the decoction of plants [22]. Most of the aqueous plant
extracts tested appeared to lack the potential to strongly inhibit the four human
CYPs and three GSTs tested. Generally, Phyllanthus amarus extracts were the most
potent inhibitors of CYP1A2, CYP2C9, CYP2D6, CYP3A4, with IC50 values ranging
from 38.1 to 97.0 µM and GSTA1-1, GSTM1-1 and GSTP1-1 with IC50 values
ranging from 3.6 to 98.2 µM. Although the inhibitory constituents of the Ghanaian
plant extracts are not known yet, flavonoids have been found in Phyllanthus amarus,
Cassia alata and Mormodica charantia. Recent studies on extracts of Ginkgo biloba
have shown that flavonoidic and terpenoidic fractions as well as whole plant extracts
strongly inhibit CYP2C9 [23]. Prangos ferulacea, also demonstrated strong inhibitory
activity towards sheep liver cytosolic GSTs [24]. Clinical implications of the present
CYP- and GST-inhibition results on the Ghanaian medicinal plant extracts are yet to
be established. However, obviously it is necessary that all herbal preparations are
screened for their potentials to cause herb-drug interactions.
Conclusion and perspectives
The major objective of the research described in this thesis was to evaluate the
interactions between plant-derived products and CYPs and GSTs, two important
human drug metabolizing enzyme systems. Such interactions might implicate
clinically relevant drug-drug interactions at the level of microsomal CYPs and
cytosolic GSTs. Firstly, the inhibition of five major human recombinant drug-
metabolizing CYPs by curcumin, derived from Curcuma longa and chosen as a
model compound, and four curcumin decomposition products have been
investigated. Subsequently, thirty-four synthetic curcumin analogues (based on
solubility criteria selected out of 40) were measured and analyzed by MOE-based
methods for quantitative structure-activity relationships (QSARs). Finally, the
inhibitory effects of seven important Ghanaian medicinal plants were also assessed
on the human drug-metabolizing CYPs. In an analogous way, the inhibition of
recombinant human GSTs and human and rat liver cytosolic GSTs by curcumin
152
analogues and seven Ghanaian medicinal plants were studied. The general
conclusions based on the findings in the experimental sections of this thesis, and
future perspectives are discussed below.
Inhibition of CYPs by curcumin, and curcumin decomposition products
The potential for drug-drug interactions of drugs and drug candidates and of new
chemical entities (NCEs) are usually evaluated during the early stages of drug
development since it is a major cause of attrition of drugs from the market [1].
Curcumin, a common food additive and a naturally occurring and synthetic
compound has been considered a promising drug candidate due to its several
pharmacological activities [25]. In this thesis, the potential for drug-drug interactions
of curcumin and four of its decomposition products was firstly assessed by their
inhibitory activities towards five important human CYPs, namely CYP1A2, CYP2B6,
CYP2C9, CYP2D6 and CYP3A4. These CYPs are responsible for the hepatic
metabolism of about 80% of drugs currently on the market and notably CYP3A4 is
also abundant in the intestine [6]. Strong competitive inhibition and mechanism-
based inhibition of enzymes are considered clinically important [1]. In the present
study, curcumin was not found to be a mechanism-based inhibitor of any of the
CYPs tested, however, it is a relatively strong competitive inhibitor of CYP3A4 and a
non-competitive inhibitor of CYP2C9. The potent inhibition of curcumin towards
CYP3A4 shown could have implications for drug-drug interactions in the intestines,
in case of high exposure of the intestines to curcumin upon oral administration.
Four decomposition products of curcumin showed no significant inhibitory
activity towards the CYPs tested, and are therefore not likely to contribute to drug-
drug interactions at the level of CYPs. The potentials for drug interactions involving
many herbs (e.g. St. John’s wort, garlic and kava) and natural compounds
(flavonoids, coumarins, caffeine and anthroquinones) have previously been
investigated [26], and some have been identified as inhibitors or inducers of various
CYP enzymes. Likewise, the inhibitory potencies of curcumin towards CYPs, have
been demonstrated in this thesis. Extrapolation of the present human in vitro data to
in vivo occurring drug-drug interactions could be achieved in principle, if the inhibitor
153
concentration in plasma and/or in the liver and the fraction of drug metabolized by a
particular CYP were known [27]. Further studies will be required, however, to predict
the clinical relevance of the strong inhibitory potential of curcumin towards CYP3A4
and CYP2C9 and the weaker inhibitory potentials towards the other human CYPs.
Inhibition of CYPs and GSTs by curcumin analogues
Due to several drawbacks of curcumin for human therapies, including an extremely
low bioavailability and significant instability at neutral to basic pH conditions [9,15],
many structural analogues of curcumin have been synthesized [4,28]. In this thesis,
we determined experimentally CYP and GST inhibitory potentials of three series of
thirty-four synthetic curcumin analogues [4] as compared to curcumin, using the five
major human CYPs mentioned above and three major human GSTs, i.e. GSTA1-1,
GSTM1-1 and GSTP1-1. Most of the analogues were less potent inhibitors of the
human CYPs and GSTs tested as compared to curcumin. The 1,4-pentadiene-3-one
derivatives (C series) contained some more potent inhibitors of CYP1A2, CYP2C9
and CYP2D6. The most potent inhibitors of CYP1A2 lacked a para-hydroxyl moiety.
Inhibition and induction of GST may have implications for detoxification,
toxification of endogenous and exogenous compounds and chemotherapy against
cancer cells [3]. With respect to GST inhibition, this work has shown that most of the
curcumin analogues tested are less potent inhibitors of the human GSTs employed
compared to curcumin. Interestingly, a recent study confirmed the fact that curcumin
is a potent inhibitor of GSTA1-1, GSTM1-1 and GSTP1-1 [18]. Our studies have
shown that GSTA1-1 and GSTM1-1 were most susceptible to inhibition by the
curcumin analogues. As observed with the human CYPs, the 1,4-pentadiene-3-one
derivatives of curcumin (C series) were generally the most potent inhibitors of the
GSTs tested, suggesting that the absence of the central ring (which is present only
in the A and B series) yields structures with increased GST-inhibitory potencies. The
GSTA1-1 and GSTM1-1 isoforms are predominantly expressed in the human liver,
whereas the hepatic levels of GSTP1-1 are insignificant [19]. However, GSTP1-1 is
the most abundant isoform in erythrocytes [20]. Thus, inhibition of GSTA1-1 and
GSTM1-1 in the liver could have implications for hepatotoxicity or hepatoprotection,
154
whilst inhibition of GSTP1-1 could lead to a decrease protection against
electrophiles and oxidative stress in erythrocytes.
Most of the curcumin analogues exhibited less inhibitory activities towards the
CYPs and GSTs tested compared to curcumin, from this point of view these could
be better alternative drug candidates. To our knowledge, no studies on drug-drug
interactions potential of curcumin analogues at the level of CYPs and GSTs have
been performed. Thus, the present findings could be an important basis for further
studies on other structurally related curcumin analogues and also for consideration
of these compounds as chemotherapeutic agents, anti-oxidant, anti-inflammatory or
other pharmacological activities [25]. Further investigations, however, are required to
establish the in vivo relevance of the present in vitro results, as well as the
bioavailability and toxicity of these compounds.
Structure-activity relationships for inhibition of CYPs and GSTs by curcumin
analogues
Except for a difference in the steric bulkiness of their aromatic substituents, chemical
similarity studies indicated no significant differences between the three series of
curcumin analogues. MOE-based QSAR analysis suggested that electrostatic and
hydrophobic interactions were important factors for CYP2C9 and CYP2D6 inhibition.
In addition, our results suggest that van der Waal’s surface area and compound
lipophilicity factors are significant for the inhibition of GSTA1-1 and GSTM1-1. As
yet, no other QSAR-studies on curcumin analogues have been reported. Thus the
present results provide new insights in structure-activity relationships of curcumin
analogues at the level of CYPs and GSTs, enabling a more rational design and
synthesis of new analogues with better properties in this regard. The statistical
MOE-based QSAR approach in the present studies, provide the means to predict to
some extent the interaction between other curcumin analogues and the CYPs and
GSTs used. Additional studies with a larger number of compounds will undoubtedly
refine the models and may account for some of the compounds being more poorly
predicted. A limiting feature of any 2D-QSAR approach is its insensitivity to the
stereochemistry of the members of the data sets used and the lack of easily
155
interpretable information useful for the design of new drugs. Thus, a 3D-QSAR
approach could be better used together with the 2D approach to provide more
enhanced models for drug-drug interaction predictions with the CYPs and GSTs.
Effects of Ghanaian medicinal plants on CYPs and GSTs
Natural products, including medicinal plants and foods are generally being
considered as harmless, and therefore they are not subjected to the scrutiny of the
approval process such as with new drug candidates. However, interactions of herbal
medicines with human CYPs have been associated with strong alterations in the
pharmacokinetics of drugs such as midazolam and alprazolam and strong adverse
effects such as with St. John’s wort [7,26]. In this thesis, the potential of seven
commonly used Ghanaian medicinal plants to inhibit human CYPs and GSTs, and
thus cause herb-drug interactions have been investigated for the first time.
Phyllanthus amarus, appeared to be the most potent inhibitor of the human CYPs
tested, while in the case of CYP1A2, Cassia alata and Lactuca taraxicifolia also
showed strong inhibition. In line with the present findings, recent in vitro and in vivo
animal studies on CYP inhibition by extracts of Phyllanthus amarus showed that it is
a potent inhibitor of CYP1A1, CYP1A2 and CYP2B1 [29].
GST activity has also been shown to be modulated by natural plant products,
such as Prangos ferulacea [24]. Since GSTs are major detoxification enzymes in
humans, inhibition of these enzymes by the medicinal plants could have important
clinical and toxicological consequences. Among the medicinal plants tested,
Phyllanthus amarus strongly inhibited GSTA1-1, GSTM1-1 and GSTP1-1.
Interestingly, GSTM1-1 was susceptible to strong inhibition by all the plants except
Lactuca taraxicifolia. The compounds responsible for the observed activities and the
clinical relevance of the inhibitory activities remain to be established. Generally,
Phyllanthus amarus also demonstrated the strongest inhibitory activities towards the
GSTs tested as observed with CYPs. Since most of the other plants tested lacked
strong inhibitory potencies towards the major human CYPs and GSTs studied, most
of these plants are not likely to cause clinically important herb-drug interactions.
However, the clinical relevance of results obtained remains to be established. Due to
156
the potential of herbs and foods to cause drug-food/-herb interactions with adverse
effects as observed with Ginkgo biloba and grapefruit juice, it is imperative that other
medicinal plants, foods and natural products that are consumed are subjected to
tests, to critically assess their potential for CYP and GST inhibition. As enzymes are
proteins, and tannins present in plants are known to precipitate and inactivate
proteins, it would be useful to determine the amount of tannin in plant extracts or
foods and to test its influence on enzyme assays.
In summary, in vitro studies on the inhibitory effects of drug candidates, herbal
products and food components are cost effective pre-clinical approaches to avoid
potential adverse effects resulting from drug-drug/-food/-herb interactions. In this
thesis, we have used recombinant human CYPs and GSTs to study the CYP and
GST inhibitory potential of plant derived components, including curcumin and
Ghanaian medicinal herbs as well as synthetic curcumin analogues. Subsequently,
structure-activity relationships were evaluated to understand underlying mechanisms
and to facilitate rational design and synthesis of curcumin analogues with less
susceptibility to drug-drug interactions at the level of CYPs and GSTs. We have
shown that curcumin, a very widely used therapeutic plant product and drug
candidate with several interesting pharmacological properties, is a potent inhibitor of
important human biotransformation enzymes, notably CYP3A4 and CYP2C9.
Curcumin inhibition of CYP3A4 may have implications for drug-drug interactions in
the intestines, also due to the high doses of curcumin usually administered. The
curcumin analogues tested were generally less potent inhibitors of CYPs and GSTs
employed, and thus they appear to be better drug candidates than curcumin from
this point of view.
The statistical 2D-QSAR approaches used to analyze the CYP and GST
inhibitory activities of the curcumin analogues revealed insights of the relationships
between structure properties of the analogues and the inhibitory activities obtained.
Increasing the number of curcumin analogues and combining 2D- and 3D- QSAR
approaches may enhance the predictive power for potential drug-drug interactions.
The inhibitory activities of the Ghanaian medicinal herbs, in particular Phyllanthus
157
amarus towards the CYPs and GSTs tested require further investigations. The
potential for clinically relevant drug-food/-herb interactions involving human CYP-
and GST-biotransformation enzymes for all herbal products and food components
must be further evaluated as well. Extrapolation of human in vitro data to the human
in vivo situation is important for prediction of drug-drug interactions in vivo. This
requires the estimation of the hepatic inhibitor concentrations as well as the fraction
of drug substrate metabolized by the CYP or GST inhibited pathway. Altogether, this
thesis has uncovered important new insights in the inhibitory potential towards
important drug metabolizing enzymes of plant-derived curcumin, synthetic curcumin
analogues and Ghanaian medicinal plant extracts.
References
1. Zhang, Z.Y., Wong, Y.N., 2005. Enzyme kinetics for clinically relevant CYP inhibition. Curr drug Metab 6:241-257.
2. Pea, F., Furlanut, M., 2001. Pharmacokinetic aspects of treating infections in the intensive care unit: focus on drug interactions. Clin Pharmacokinet 40:833-868.
3. Commandeur, J.N., Stijntjes, G.J., Vermeulen, N.P., 1995. Enzymes and transport systems
involved in the formation and disposition of glutathione S-conjugates. Role in bioactivation and detoxication mechanisms of xenobiotics. Pharmacol Rev 47:271-330.
4. Sardjiman, S., Reksohadiprodjo, M., Hakim, L., van der Goot, H., Timmerman, H., 1997. 1,5-Diphenyl-1,4-pentadiene-3-ones and cyclic analogues as antioxidative agents. Synthesis and structure-activity relationship. Eur J Med Chem 32:625-636.
5. Aparasu, R., Baer, R., Aparasu, A., 2007. Clinically important potential drug-drug Interactions in outpatient settings. Res Social Adm Pharm 3:426-437.
6. Lamb, D.C., Waterman, M.R., Kelly, S.L., Guengerich, F.P., 2007. Cytochrome P450 and drug discovery. Curr Opin Biotechnol 18:1-9.
7 Gurley, B.J., Gardner, S.F., Hubbard, M.A., Williams, D.K., Gentry, W.B., Khan, I.K., Shah, A., 2005. In vivo effects of goldenseal, kava kava, black cohosh, and valerian on human cytochrome P450 1A2, 2D6, 2E1, and 3A4/5 phenotypes. Clin Pharmacol Ther 77:415-426.
8. Perkins, S., Verschoyle, R.D., Hill, K., Parveen, I., Threadgill, M.D., Sharma, R.A., Williams, M.I., Steward, W.P., Gescher, A.J., 2002. Chemopreventive efficacy and pharmacokinetics of curcumin in the Min/+ mouse, a model of familial adenomatous polyposis. Cancer Epid Biomarkers & prevention 11:535-540.
9. Sharma, R.A., Euden, S.A., Platton, S.L., Cooke, D.N., Shafayat, A., Hewitt, H.R., Marczylo, T.H., Morgan, B., Hemingway, D., Plummer, S.M., Pirmohamed, M., Gescher, A.J., Steward, W.P., 2004.Phase I clinical trial of oral curcumin: biomarkers of systemic activity and compliance. Clin Cancer Res 10:6847-6854.
10. Marczylo, T.H., Verschoyle, R.D., Cooke, D.N., Morazzoni, P., Steward, W.P., Gescher, A.J., 2007. Comparison of systemic availability of curcumin with that of curcumin formulated with phosphatidylcholine. Cancer Chemother Pharmacol 60:171-177.
11. Shoba, G., Joy, D., Joseph, T., Majeed, M., 1998. Influence of piperine on the pharmacokinetics of curcumin in animals and human volunteers. Planta Med 64:353-356.
12. Ma, Z., Shayeganpour, A., Brocks, D.R., Lavasanifar, A., Samuel, J., 2007. High performance liquid chromatography analysis of curcumin in rat plasma: application to pharmacokinetics of polymeric micellar formulation of curcumin. Biomed Chromatogr 21:546-552.
158
13. Safavy, A., Raisch, K.P., Mantena, S., Sanford, L.L., Sham, S.W., Krishna, N.R., Bonner, J.A., 2007. Design and development of water-soluble curcumin conjugates as potential anticancer agents. J Med Chem Nov 1 (Epub ahead of print).
14. Tamvakopoulos, C., Sofianos, Z.D., Garbis, S.D., Pantazis, P., 2007. Analysis of the in vitro metabolites of diferuloylmethane (curcumin) by liquid chromatography – tandem mass spectrometry on a hybrid quadrupole linear ion trap system: newly identified metabolites. Eur J Drug Metab Pharmacokinet 32:51-57.
15. Oetari, S., Sudibyo, M., Commandeur, J.N.M., Samhoedi, R., Vermeulen, N.P., 1996. Effects of curcumin on cytochrome P450 and glutathione S-transferase activities in rat liver. Biochem Pharmacol 51:39-45.
16. van Iersel, M.L., Ploemen, J.P., Struik, I., van Amersfoort, C., Keyzer, A.E., Schefferlie, J.G., van Bladeren, P.J., 1996. Inhibition of glutathione S-transferase activity in human melanoma cells by alpha, beta-unsaturated carbonyl derivatives. Effects of acrolein, cinnamaldehyde, citral, crotonaldehyde, curcumin, ethacrynic acid and trans-2-hexenal. Chem Biol Interact 102:117-132.
17. Appiah-Opong, R., Commandeur, J.N.M., van Vugt-Lussenburg, B., Vermeulen, N.P.E., 2007. Inhibition of human recombinant cytochrome P450s by curcumin and curcumin decomposition products. Toxicology 235:83-91.
18. Hayeshi, R., Mutingwende, I., Mavengere, W., Masiyanise, V., Mukanganyama, S., 2007. Inhibition of human glutathione S-transferases activity by plant polyphenolic compounds ellagic acid and curcumin. Food Chem Toxicol 45:286-295.
19. Eaton, D.L., Bammler, T.K., 1999. Concise review of the glutathione S-transferases and their significance to toxicology. Toxicol Sci 49:156-164.
20. Awasthi, Y.C., Sharma, R., Singhal, S.S., 1994. Human glutathione S-transferases. Int J Biochem 26:295-308.
21. Ioannides, C., 2002. Pharmacokinetic interactions between herbal remedies and medicinal drugs. Xenobiotica 32:451-478.
22. Ayisi, N.K., Nyadedzor,C., 2003. Comparative in vitro effects of AZT and extracts of Ocimum gratissimum, Ficus polita, Clausena anisata, Alchornea cordifolia, and Elaeophorbia
drupiferagainst HIV-1 and HIV-2 infections. Antiviral Res 58:25-33. 23. Gaudineau, C., Beckerman, R., Welbourn, S., Auclair, K., 2004. Inhibition of human P450
enzymes by multiple constituent of Ginkgo biloba extract. Biochem Biophys Res Commun 318:1072-1078.
24. Coruh, N., Celep, A.G.S., Ozgokce, F., 2007. Antioxidant properties of Prangos ferulacea (L.) Lindl., Chaerophllum macropodum Boiss. and Heracleum persicum Desf. from Apiaceae family used as food in Eastern Anatolia and their inhibitory effects on glutathione S-transferase. Food Chem 100:1237-1242.
25. Singh, S., Khar, A., 2006. Biological effects of curcumin and its role in cancer chemoprevention and therapy. Anti-Cancer Agents Med Chem 6:259-270.
26. Izzo, A.A., 2004. Herb-drug interactions: An overview of the clinical evidence. Fundamental Clin Pharmacol 19:1-16.
27. Brown, H.S., Ito, K., Aleksandra G., Houston, B., 2005. Prediction of in vivo drug-drug interactions from in vitro data: impact of incorporating parallel pathways of drug elimination and inhibitor absorp- tion rate constant. Br J Clin Pharmacol 60:508-518.
28. Youssef, K.M., El-Sherbeny, M.A., El-Shafie, F.S., Farag, H.A., Al-Deeb, O.A., Awadalla, S.A.A., 2004. Synthesis of curcumin analogues as potential antioxidant, cancer preventive agents. Arch Pharm Pharm Med Chem 337:42-54.
29. Hari Kumar, K.B., Kuttan, R., 2006. Inhibition of drug metabolizing enzymes (cytochrome P450) in vitro as well as in vivo by Phyllanthus amarus Schum & Thonn. Biol Pharm Bull 29:1310-1313.
159
Samenvatting en conclusies
Geneesmiddel-geneesmiddel of geneesmiddel-voedsel interacties waarbij
Cytochrome P450 enzymen (CYPs) betrokken zijn, zijn een belangrijke oorzaak
voor het terugtrekken van de markt van geneesmiddelen als gevolg van ernstige
bijwerkingen, van het verlies van geneesmiddelkandidaat-moleculen tijdens de
ontwikkeling van nieuwe geneesmiddelen, en van waarschuwingen in bijsluiters
van geneesmiddelen. Inhibitie van CYPs door dergelijke interacties kan leiden tot
een verminderde biologische beschikbaarheid van geneesmiddelen en tot een
vertraagde metabole afbraak ervan, met langere halfwaarde-tijden als gevolg.
Ook kunnen geneesmiddel-geneesmiddel en geneesmiddel-voedsel interacties
leiden tot omgekeerde effecten als gevolg van inductie van de CYP-niveaus in
het maag-darm kanaal, in de lever of in andere weefsels en organen. Analoog
kan er ook inhibitie of inductie van Glutathion-S-transferases (GSTs) door
geneesmiddelen en andere xenobiotica optreden. Dit kan leiden tot een
verminderde bescherming tegen bepaalde vormen van toxiciteit van xenobiotica
en bijwerkingen van geneesmiddelen.
Het doel van de studies, die in dit proefschrift beschreven zijn, was om
interacties op het niveau van CYPs en GSTs te bestuderen tussen
geneesmiddelen onderling en tussen geneesmiddelen en plantaardige
producten. Curcumine, een gele farmacologisch actieve stof die voor het eerst
geïsoleerd werd uit Curcuma longa, is hierbij gebruikt als een modelstof. Voor
Curcumine zijn, veelal (maar niet uitsluitend) in vitro, anti-inflammatoire, anti-
oxidant, anti-tumor, chemoprotectieve, chemopreventieve en diverse andere
farmacologische activiteiten aangetoond. Bovendien zijn in dit proefschrift de
interacties op het niveau van CYPs en GSTs bestudeerd van 40 synthetische
Curcumine-analoga en van extracten van een 7-tal Ghanese medicinale planten.
Om inzicht te krijgen in relaties tussen de chemische structuur en de CYP- en
GST-interacties zijn structuur-werkingsrelaties afgeleid met statische methodes
gebaseerd op MOE.
160
In hoofstuk 1 is een algemene inleiding gegeven van geneesmiddel-
geneesmiddel en geneesmiddel-voedsel/planten interacties op het niveau van
CYPs en GSTs, zowel voor wat betreft de mogelijke consequenties op gewenste
biologische activiteiten (de farmacologie) en ongewenste biologische activiteiten
(de toxicologie). In hoofstuk 2 wordt een overzicht gegeven van de effecten van
Curcumine op de farmacokinetiek en het metabolisme van geneesmiddelen en
andere biologisch actieve stoffen. In hoofdstuk 3 ligt de focus op de bepaling
van de inhiberende activiteit van Curcumine op 5 van de belangrijkste humane
CYPs (CYP1A2, CYP2B6, CYP2C9, CYP3A4 en CYP2D6). Er wordt
geconcludeerd dat Curcumine een dusdanig inhiberend potentieel ten opzichte
van CYP3A4 heeft, dat er na orale toediening van geneesmiddelen klinisch
relevante geneesmiddel-geneesmiddel interacties kunnen optreden in het maag-
darm kanaal. In hoofdstuk 4 worden de resultaten beschreven van een studie
naar het inhiberend vermogen bepaald ten opzichte de 5 genoemde
recombinante humane CYPs van 33 analoga van Curcumine (op basis van
oplosbaarheidscriteria geselecteerd uit 40 analoga). Bovendien werd een op
MOE-gebaseerde QSAR analyse uitgevoerd op de gevonden IC50 waarden. De
resultaten leiden tot de conclusie dat electrostatische en hydrofobe moleculaire
descriptoren bruikbaar zijn om de CYP-inhiberende activiteiten van de
Curcumine analoga aan te relateren. In hoofstuk 5 worden de resultaten
besproken van een analoge studie naar het inhiberend vermogen van dezelfde
33 Curcumine analoga ten opzichte van een 3-tal recombinante humane
Glutathione S-transferases (GSTA1-1, GSTM1-1 en GSTP1-1) en humaan en
rattelever-cytosol. Een op MOE-gebaseerde QSAR analyse maakte duidelijk dat
Van der Waals oppervlakte en lipofiliciteits factoren een rol spelen in de GST-
inhiberende werking. Tot slot worden in hoofdstuk 6 de inhiberende
eigenschappen bestudeerd van 7 extracten van geselecteerd (veel gebruikte)
Ghanese medicinale planten op de 5 recombinante humane CYPs, op de 3
recombinante humane GSTs en humaan en rattelever-cytosol. Phyllantus
amarus extracten bleken het sterkste inhiberend vermogen te hebben ten
161
opzichten van het merendeel van deze recombinante enzymen en leverenzym-
fracties.
Algemene conclusies en perspectieven
Inhibitie van CYPs en GSTs, twee van de belangrijkste humane biotransformatie-
enzymsystemen voor geneesmiddelen en andere chemische stoffen, kan klinisch
relevante effecten hebben op de farmacokinetiek en op de farmacologische en
toxicologische werking van dergelijke lichaamsvreemde stoffen. In dit proefschrift
worden de resultaten beschreven van in vitro studies naar het inhiberend
vermogen van Curcumine, een bekende en zeer veel gebruikte biologisch
actieve component uit Curcuma longa, van 4 ontledingsproducten van
Curcumine, van 34 synthetische analoga van Curcumine (op basis van
oplosbaarheids-criteria geselecteerd uit 40 analoga) en van wortel- en stam-
extracten van 7 veelvuldig gebruikte Ghanese medicinale planten. Met behulp
van op MOE gebaseerde methoden werden de gemeten IC50-waarden
geanalyseerd op mogelijk significante structuur-werkingsrelaties (QSARs).
Behalve met 5 humane CYPs en 3 humane GSTs, verkregen met recombinante
gen-expressie in Ecoli-bacterien, werden ook inhibitiestudies gedaan met
humane lever fracties. Curcumine zelf bleek vooral sterk competitief inhiberend
op CYP3A4 en wel zodanig dat deze eigenschap na orale toediening van
geneesmiddelen, klinisch relevante geneesmiddel-interacties in de dunne darm
zou kunnen veroorzaken. CYP2C9 onderging in mindere mate inhibitie als
gevolg van Curcumine. Curcumine bleek geen ‘mechanism-based’ inhibitie
eigenschappen te vertonen.
Met behulp van de 34 synthetische Curcumine-analoga, konden zowel voor de
bestudeerde CYPs als voor de GSTs enkele goede structuur-werkingsrelaties
(QSARs) worden afgeleid door gebruik te maken van op MOE-gebaseerde
statistische QSAR analyse methodes. Hoewel de significantie van de QSAR-
relaties nog vergroot zouden kunnen worden, door meer en sterker inhiberende
Curcumine-analoga te bestuderen of door 3D-QSAR analyses (met inbegrip van
structurele informatie van de CYP- en GST-isoenzymen), zijn de in proefschrift
162
beschreven QSAR-resultaten zeer nuttig bij het ontwerpen en ontwikkelen van
analoga van Curcumine met meer of minder sterk CYP- of GST-inhiberende
eigenschappen. Inhiberende eigenschappen van deze enzymsystemen kunnen
als gunstig (b.v. een betere cytostatische activiteit van alkylerende cytostatica als
gevolg van GST-inhibitie) en als ongunstig (b.v. kans op geneesmiddel-
interacties als gevolg van CYP-inhibitie in het maagdarm kanaal of in de lever)
beschouwd worden.
Voor wat betreft de Ghanese medicinale planten bleken extracten van de
Phyllanthus amarus en in mindere mate Cassia alata and Lactuca taraxicifolia,
relatief sterke inhiberende activiteiten te vertonen op enkele CYPs en op enkele
GSTs. De klinische relevantie van deze waarnemingen moet echter nog
onderzocht worden. In de literatuur zijn veel klinisch relevante geneesmiddel-
voeding en geneesmiddel-plantenextract interacties beschreven, zoals
bijvoorbeeld voor Ginkgo biloba en grapefruit sap.
Concluderend kan gesteld worden dat het in dit proefschrift beschreven in vitro
onderzoek belangrijke nieuwe inzichten heeft opgeleverd in het inhiberend
vermogen van Curcumine, Curcumine ontledingsproducten, een reeks
Curcumine-analoga en van wortel- en stam-extracten van veel gebruikte
Ghanese medicinale planten ten opzichte van 5 van de belangrijkste humane
CYPs en 3 van de belangrijkste GSTs. De resultaten duiden op een mogelijke
klinische relevantie, hoewel zekerheid daarover alleen verkregen zal kunnen
worden in echte klinische studies. Bovendien hebben de resultaten nieuwe
inzichten verschaft in verbanden tussen moleculaire parameters van dergelijke
stoffen en hun inhiberend vermogen ten opzichte van humane CYPs en GSTs.
Deze relaties zijn waardevol bij het rationaliseren van deze biologische
activiteiten en bij het ontwerpen van nieuwe moleculen met een betere balans
tussen gewenste en ongewenste activiteiten.
163
Appendices
List of publications related to this thesis Appiah-Opong, R., Commandeur, J.N.M., van Vugt-Lussenburg, B., Vermeulen, N.P.E., 2007. Inhibition of human recombinant cytochrome P450s by curcumin and curcumin decomposition products. Toxicology 235:83-91. Appiah-Opong, R., Commandeur, J.N.M., Vermeulen, N.P.E. Curcumin: Pharmacokin-etics, metabolism, and potential for drug-drug/food interactions. Proceedings of Int Symp Recent Progress in Curcumin Res Yogyakarta Indonesia, 2007. Appiah-Opong, R., de Esch, I., Commandeur, J.N.M., Andarini, M., Vermeulen, N.P.E., 2008. Inhibition of recombinant human cytochrome P450 mediated metabolism by curcumin analogues and related structure-activity relationships. Eur J Med Chem 43:1621-1631. Appiah-Opong, R., Commandeur, J.N.M., Axson C., Vermeulen, N.P.E. 2008. Interactions between cytochromes P450 and glutathione S-transferases and Ghanaian medicinal plants. Food Chem Toxicol 46:3598-3603. Appiah-Opong, R., Commandeur, J.N.M., Istyastono, E., Bogaards, J. J., Vermeulen, N.P.E. Inhibition of glutathione S-transferases activity by curcumin analogues. Xenobi-otica submitted.
List of publications not related to the work in this thesis
Kinomoto, M.*, Appiah-Opong, R.*, Brandful, J.A.M., Yokoyama, M., Nii-Trebi, N., Ugly-Kwame, E., Sato, H., Ofori-Adjei, D., Kurata, T., Barre-Sinoussi, F., Sata, T., Tokunaga, K., 2005. HIV-1 proteases from drug-naïve West African patients are differentially less susceptible to protease inhibitors. Clin Infect Dis 41:243-251. *contributed equally Ankrah, N-A., Quaye, I. K. E., Appiah-Opong, R., Dzokoto, C., Ekuban, F. A., Teye, K., 2003. Association between low blood glutathione levels and haptoglobulin phenotypes in pregnant women. Ghana Med J 37:35-38. Ankrah, N-A., Appiah-Opong, R., Dzokoto, C., 2000. Human breast milk storage and the glutathione content. J Trop Ped 46:111-113. Ankrah, N-A., Appiah-Opong, R., 1999. Toxicity of low levels of methylglyoxal: depletion of blood glutathione and adverse effect on glucose tolerance in mice. Toxicol Lett 109:61-67.
164
Ankrah, N-A., Nyarko, A. K., Ofosuhene, M., Appiah-Opong, R., Akyeampon Y. A., 1998. Lead exposure in urban and rural school children in Ghana. Afr J Health Sci 5:85-88. Ankrah, N-A., Dunyo, S. K., Nyarko, A. K., Appiah-Opong, R., Ofosuhene, M., 1998. Biliary excretion in persons with low blood glutathione levels. East Afr Med J 75:204-207. Ankrah, N-A., Sittie, A., Appiah-Opong, R., Ackom, I., 1996. Glyoxalase–I activity levels in peripheral blood of Ghanaian Africans with or without Plasmodium falciparum. Afr J Health Sci 3:41-43. Ankrah, N-A., Kamiya, Y., Appiah-Opong, R., Akyeampon, Y. A., Addae, M. M., 1996. Lead levels and related findings occuring in Ghanaian subjects occupationally exposed to lead. East Afr Med J 73:375-379. Armah, G. E., Mingle, J. A. A., Dodoo, A. K., Anyanful, A., Antwi, R., Commey, J., Nkrumah, F. K., 1994. Seasonality of rotavirus infection in Ghana. Annals Trop Pediatr 14:223-230.
165
Epilogue My sincere gratitude is extended to all who contributed in one way or the other in
the outcome of this research work. I thank God for the strength, grace and
endurance He afforded me. Nico, as my promoter you have been a great source
of encouragement to me, and I really appreciate your help, direction and support
even in the use of the computer. Jan, thank you for being my copromoter. Your
criticism and instructions were helpful and very much appreciated. Thank you
Chris O., for your kind advice and support. I am grateful to Prof. Ron de Kloet for
kindly introducing me to my promoter. Laura, your encouragement and help were
very much appreciated. Claudia and Laura, I am very grateful to you for the
efforts you made towards the printing of this thesis. I thank all past and present
staff and colleagues of Moltox section namely, Jeroen Kool, Ed, Micaela, Chris
de Graaf, Peter, Sebastian, Jelle, Jolanda, Eva, Aldo, Anton, Barbara, Jozef,
Chris Vos, Jeroen, Bernardo, Nathan, Eduardo and Vanina. I am also grateful to
all other staff and colleagues of the Department of Pharmaceutical Sciences and
Chemistry for their support. My gratitude goes to students I have worked with, on
this project including Maya, Chimed, Civianny, Enade and Robbert. Appreciation
is also extended to the Ghana government and Getfund Scholarship Schemes of
the Republic of Ghana for funding this project. Prof. D. Ofori Adjei, former
Director of Noguchi Memorial Institute for Medical Research (NMIMR), thanks for
your effort in obtaining this scholarship for study. For your support and
encouragement, Prof. A.K. Nyarko, Director of NMIMR, I thank you. Ken, Dorcas
and Lois, I appreciate you, for your support and the sacrifice of allowing me to
stay so far away from home for four long years. Mummy, I am very grateful to
you for taking care of my children while I was away studying. My father, siblings,
Legon Interdenominational Church, Ghana, Pentecost International Worship
Centre, Amsterdam, friends and loved ones, thank you all for your support. May
God bless all of you.
166
List of abbreviations
BFC 7-benzyloxy-4-trifluoromethyl-coumarin
BRes benzyloxyresorufin
BROD benzyloxyresorufin O-dealkylation
BQ benzyloxyquinoline
CDNB 1-chloro-2,4-dinitrobenzene
CYP cytochrome P450
DBF dibenzylfluorescein
EROD ethoxyresorufin O-dealkylation
GSH reduced glutathione
GST glutathione S-transferase
HPLC high performance liquid chromatography
MOE molecular operating environment
MRes methoxyresorufin
MROD methoxyresorufin O-dealkylation
NAC N-acetyl L-cysteine
PROD pentoxyresorufin O-dealkylation
(Q)SAR (quantitative) structure-activity relationship
UDPGA uridine diphosphate glucuronic acid
UGT UDP-glucuronosyltransferase