bs_bs_banner

Journal of Pharmacy
Research Paper
And Pharmacology

Herbal tea extracts inhibit Cytochrome P450 3A4 in vitro

Sophie Dufaya, Alan Worsleya, Aymeric Monteilliera, Charlotte Avanzia, Jaclyn Syb, Ting Fat Nga,
Jean-Michel Garciac, Man-Fai Lamd, Paul Vanhouttea and Ian C. K. Wonga
a
Department of Pharmacology and Pharmacy, bChemistry Department, The University of Hong Kong, dDepartment of Medicine, Queen Mary
Hospital Hong Kong and cInstitute Pasteur Korea, Bundang-gu, Seongnam-si, Gyeonggi-do, Korea

Downloaded from https://academic.oup.com/jpp/article/66/10/1478/6128004 by guest on 26 February 2021

Keywords Abstract
ciclosporin; Cytochrome 3A4; herb–drug
interaction; herbal teas; sirolimus; traditional Objectives Ciclosporin and sirolimus, two immunosuppressive agents with
Chinese medicine narrow therapeutic windows, are mainly metabolized by Cytochrome 3A4
(CYP3A4). A clinical case of toxic blood levels of these drugs after the consump-
Correspondence
tion of a ‘24-flavours’ tea was reported. This study aims to identify the causative
Sophie Dufay, Department of Pharmacology
ingredients of the 24-flavour herbal tea in the inhibition of CYP3A4 metabolism.
and Pharmacy, The University of Hong Kong,
21 Sassoon Roads, Pokfulam, Hong Kong Methods Two commercially available 24-flavour tea products purchased in Hong
SAR. Kong and the six plant constituents were tested for their CYP3A4 inhibitory
E-mail: sdufay@hku.hk effects utilizing an in-vitro fluorometric assay.
Key findings Of the commercially available teas available in Hong Kong, the most
Received August 21, 2013 potent inhibitory effect was observed with the tea consumed in the initial clinical
Accepted March 23, 2014
case. Of the six universal constituents, chrysanthemum exhibited the greatest
inhibitory effect, with an IC50 of 95.7 μg/ml. Dandelion, liquorice and bishop’s
doi: 10.1111/jphp.12270
weed have IC50 of 140.6, 148.4 and 185.5 μg/ml, respectively. Field mint and Japa-
nese honeysuckle have weaker inhibitory effect on CYP3A4 with IC50 of 1153.3
and 1466.3 μg/ml.
Conclusions This study confirms the possible implication of herbal tea constitu-
ents in the inhibition of ciclosporin and sirolimus’ CYP3A4 metabolism. Com-
bined usage of herbal teas with drug should be closely monitored.

Introduction
The use of natural medicines has increased in western
countries primarily due to the belief that they are associated
with fewer adverse reactions. A survey performed in the
United States in 1998–1999 on ambulatory patients showed
that 16% of drug users concurrently used some form of
herbal supplement.[1] This number was even higher for
patients with long-term disease. A survey performed in the
United States in 2001 showed that 26% of patients with
human immunodeficiency virus (HIV) consumed herbal
remedies at the same time as their HIV treatment.[2] This
number increases to more than one third of cancer patients
in two European surveys.[3,4] The impact of combined thera-
peutic drug, natural medicine use has been evaluated in a
clinical survey conducted over 804 patients from six outpa-
tient clinics in the United States. It showed that 40% of
herbal users were at risk of herb–drug interaction, with 7%
of them effectively reporting adverse reactions.[5] However,
in most cases, potential interactions between the different
treatments were not evaluated.
With the increased consumption of herbal remedies con-
comitant to drug treatment, the potential risk of herb–drug
interaction and sparse scientific evaluation, more scientific
data are needed on assessment of the safety of such
combination.[6–10] In this article, a scientific investigation of
a clinical case with suspected herb–drug interaction is pre-
sented. Experiments were designed to mimic real case situa-
tion where most of in-vitro studies published failed.
A clinical case from a regional nephrology unit (Queen
Mary Hospital, Hong Kong) reported a tenfold increase in
the blood level of two immunosuppressants (ciclosporin A
and sirolimus) in a patient with a stabilized renal allograph
for 10 years.[11] A full patient history revealed ingestion of a
commonly consumed herbal tea (24-flavour tea, 廿四味茶)
during the three previous days, and it was suggested to be
the possible causative agent of the increased levels of the
two immunosuppressant drugs. Herbal teas are part of daily
life in Hong Kong and China. Many different kinds of
herbal teas can be found under different forms. They are

1478 © 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 66, pp. 1478–1490
Sophie Dufay et al.
sold as ‘ready-to-drink’ infusions in teashops at street
corners, but also are available as dried plant mixtures
(which are subsequently prepared at home) or freeze-dried
granules of infused decoctions. They are drunk as a bever-
age for many suggested reasons, primarily as a tonic and to
improve health both physically and mentally and to prevent
illness.
A comprehensive literature review, utilizing the search
terms 24 flavour teas (廿四味茶), tea, herbal teas, herbal
infusion, interaction, herbal remedies, Chinese medicine,
cyclosporin and sirolimus, was realized. From these papers,
two clinical cases were observed by Nowack et al. (2005), in
which increased plasma levels of ciclosporin were observed in
patients concurrently consuming herbal tea products. These
studies showed two renal transplant patients drinking herbal
teas (one of whom had consumed chamomile tea, and the
other one consumed a mixture called wild fruit tea drink
containing rose hip extract, hibiscus extract and lemon) while
both taking ciclosporin. Both patients showed an increase in
ciclosporin level following their respective herbal tea con-
sumption. The authors suggested that the responsible agents
were probably orange peel, or other citrus ingredients and
possibly chamomile.[12] However, without herbal tea composi-
tion, there was a lack of evidence that the ‘24-flavour tea’
drunk by our patient could be linked with these two clinical
cases. With regard to the narrow therapeutic window of
ciclosporin and sirolimus, and the fact that renal transplanted
patients are advised to consume a large volume of fluids to
improve renal graft function, the impact of herb–drug inter-
action on transplant patients might be significant due to the
risk of drug nephrotoxicity.
To investigate the described clinical case and possible causes
for observed elevation in immunosuppressant agents, the first
step was to obtained detailed information about the composi-
tion of 24-flavour tea. As there is not set formula of 24-flavour
tea, composition may vary between manufacturers, and there-
fore 22 different brands of 24-flavour tea were sought to
establish a list of those 20 plants present in the majority of
these products. When screened into the literature for inhibi-
tory effect on the cytochrome of these 20 plants, little or no
information appeared available. The most reported interac-
tion was the inhibition of Cytochrome 3A4 (CYP3A4) by
citrus fruit consumption (in particular, grapefruit, which
however is not included into any 24-flavour tea formula).
Sirolimus and ciclosporin are mainly metabolized by
CYP3A4 and are substrates of p-glycoprotein (p-gp) in the
small intestine. These two sites might be responsible for the
increased level of ciclosporin and sirolimus. P-gp is an
effective efflux transporter of the luminal surface of gastro-
intestinal epithelial cells and opposes the absorption of
some molecules like sirolimus and ciclosporin. CYP3A4 is
the most abundant cytochrome enzyme in the liver, but it is
also present in sustentative amounts in the enteric mucosa
© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 66, pp. 1478–1490
Inhibitory effect of herbal teas on CYP3A4
of the gastrointestinal tract.[13] To access to liver cyto-
chromes, chemical compounds need to be lipophilic to pass
through the gut barrier. However, these chemical com-
pounds are water-soluble (infusion of the tea), then it is
most likely that inhibitory effect will occur at the gut level
where chemical compounds will be available at a much
higher concentration.
However, p-gp has been associated with competitive and
short-lived inhibition effect, whereas CYP3A4 has a poten-
tial for a prolonged inhibitory effect.[14] Also, CYP3A4 is
Downloaded from https://academic.oup.com/jpp/article/66/10/1478/6128004 by guest on 26 February 2021
involved in the metabolism of more than 50% of the drugs
on the market,[13] so with regard to the mechanism but also
the impact it was decided to evaluate the inhibitory effect of
CYP3A4 in vitro. As 24-flavour tea composition varies, the
investigation of its inhibitory effect in vitro was performed
to confirm the doctor’s hypothesis on the identical tea con-
sumed by the patient in the clinical case and another com-
mercially available product to avoid anecdotal conclusions.
The four commonly found plant within the list of 20 (two
third of the 22 brands of tea), and two widespread plants,
included in more than one third of the 22 brand were tested
in vitro for their inhibitory effect on CYP3A4. A literature
search for the chemical constituents of those plants tested
was performed, revealing a whole series of compounds.
Where those compounds identified were commercially
available, these were purchased and used as standards
against those plant extracts isolated after analysis by high-
performance liquid chromatography hyphenated with
photodiode array (HPLC-PDA).
Materials and Methods
Preparations of water extract of herbal
remedies and ‘24-flavour’ tea
‘24-flavours’ tea (廿四味茶) and the identified constituent
plants were purchased from a traditional Chinese medicine
wholesaler in Hong Kong. Taxonomic authentication of the
plants has been realized with regard to Chinese Pharmaco-
poeia (year 2010, English version) by an expert of chemistry
department involved in Hong Kong Chinese Materia Medica
Standards at the University of Hong Kong. Voucher speci-
mens were deposited and kept in chemistry department. The
plants studied were chrysanthemum (Chrysanthemum
morifolium Ramat., 2014-SGD-01), dandelion (Taraxacum
mongolicum Hand-Mazz., 2014-SGD-02), Japanese honey-
suckle (Lonicera japonica Thunb., 2014-SGD-03), bishop’s
weed (Houttuynia cordata Thunb., 2014-SGD-04), liquorice
(Glycyrrhiza ularensis Fisch., 2014-SGD-05) and field mint
(Mentha haplocalyx Briq., 2014-SGD-06).
The ‘24-flavours’ teas were prepared according to the rec-
ommendations given by the street tea store/wholesaler or by
the manufacturers’ instructions. One example of ‘24-
flavours’ tea (廿四味茶) consumed by the patient in the
1479
Inhibitory effect of herbal teas on CYP3A4
above-mentioned clinical report was purchased as an
aqueous infusion (the commonest form available in Hong
Kong). The entire content of one bottle (250 ml) was freeze-
dried after filtration (number 1).
The second ‘24-flavours’ tea (廿四味茶) (number 2) pre-
sented was purchased as a mixture of crude plant material
and was processed as recommended by the manufacturer.
A sample of 15 g of each plant material was weighed and
extracted with 150 ml of distilled water and boiled for 1 h.
The plant solid residues were then removed and further
extracted with the same procedure. Both water extracts were
combined and vacuum-filtered on Büchner with paper filter
(Advantec 90 mm; Toyo Roshi Kaisha Ltd, Tokyo, Japan).
The combined filtrates were freeze-dried (freeze dryer,
IlShin, IlShin Europe B.V., Ede, The Netherlands) overnight
and removed from the machine when extracts were totally
dried. The resulting powders were weighed and stored in
airtight individual brown glass bottles, for light protection.
CYP3A4 enzyme assay
Different methods can be used to evaluate inhibitory effects
on cytochromes: human liver slices, human hepatocytes,
microsomes and human cytochromes. Human cytochrome
technique was selected with regard to reliability to assess an
inhibitory effect on a given cytochrome, as it is isolated
from others. CYP3A4 inhibition was measured using the
fluorometric Cytochrome P450 CYP3A4/DBF Inhibition
Kit (BD Supersomes, BD Biosciences, Le Pont de Claix,
France) following manufacturer recommendations. Briefly,
incubations were conducted in a reaction volume of 200 μl/
well in 96-well microlitre plate, as described below.
The procedure was performed in two stages. During the
first step, NADPH co-factor mix was incubated with tested
samples (plants/teas or ketoconazole as positive control
inhibitor). During the second step, the enzymatic reaction
was started by addition of the enzyme (CYP3A4) and its
substrate (dibenzylfluoresceine, DBF) solution. The incuba-
tion time for each step was 10 min at 37°C. Finally, NaOH
(2N) was added to stop the enzyme reaction, followed by
measure of the fluorescence in Cary Ellipse plate reader
fluorometer (Agilent Technology, Santa Clara, CA, USA).
The wavelengths set were 485 nm for excitation and 538 nm
for emission, specific of fluorescein, which is the metabolite
of DBF used as substrate of CYP3A4.
To determine the concentration at which 50% of the
CYP3A4 activity is inhibited (IC50), the liquid herbal
extracts concentration was expressed as the quantity in
gram of dry extract per litre. Eight points of threefold serial
dilutions of each sample were performed directly in the
96-well plates, starting from a concentration of 10.0 mg/ml.
The logarithms of these values were plotted onto the
graphs. Each compound was tested in triplicate. Inhibition
1480 © 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 66, pp. 1478–1490
Sophie Dufay et al.
curves were plotted from the log values of the samples con-
centration versus the percentage inhibition calculated.
Ketoconazole was used as positive control, while negative
control wells contained all constituents of the reaction
except the inhibitors (ketoconazole, teas or plants).
Data analysis
Extracts of plants and 24-flavour teas will be tested in tripli-
cate. Means and standard deviation of each inhibitory effect
measured were calculated. All three analyses will be used to
Downloaded from https://academic.oup.com/jpp/article/66/10/1478/6128004 by guest on 26 February 2021
determine IC50. The calculation of IC50 values for CYP3A4
inhibition values was carried out by nonlinear regression
analysis using REGTOX program freely available on
Normale Superieure website (http://www.normalesup.org/
∼vindimian/fr_index.html). It was also used for the fitting
of the dose-response sigmoid curves.
Analysis of 24-flavour teas and plants
extract by HPLC-PDA analysis
Chemicals
Different chemical compounds thought to be possible
causative agents of enzyme inhibition were identified in the
herbal extracts of 24-flavour teas and plants. When available
commercially with purity higher than 98%, standards were
used for identification by HPLC-PDA analysis: oleanolic
acid, ursolic acid, hesperidin and naringenin were pur-
chased from Sigma-Aldrich (St Louis, MO, USA); quercetin
from Cayman Chemical Company (Ann Arbor, MI, USA),
glycyrrhizin ammonium was bought from Calbiochem
(EMD Chemicals Inc, San Diego, CA, USA, affiliated to
Merck, Darmstadt, Germany) and hesperetin from Tokyo
Chemicals Industry Co Ltd (Tokyo, Japan).
Chromatographic system
The chromatographic system was composed of an HPLC
apparatus from Agilent 1260 Infinity Quaternary LC System
(Agilent Technologies Ltd), which includes helium degasser
(1260 Infinity Standard degasser), a quadratic pump (1260
Infinity binary pump VL) and an auto sampler (1260 Infin-
ity Standard Autosampler). This system was piloted by the
Open LAB CDS Chemstation software version B. 04.03.
Detection was done with a photodiode array detector (1260
Infinity Diode Array Detector VL+).
Chromatographic analysis of plasma samples were per-
formed on the Agilent Prep-C18 Scalar Column 4.6 ×
250 mm, 5 μm from Agilent (Bellafonte, PA, USA). The flow
rate was set at 1.0 ml min. A gradient of phase A (water
with 0.1% of phosphoric acid) and phase B (acetonitrile)
was used to separate all the chemical compounds. The chro-
matographic analysis was performed using the following
aqueous mobile phase (phase A) gradient with a flow rate
Sophie Dufay et al. Inhibitory effect of herbal teas on CYP3A4

Table 1 Doses and indications of plants commonly found into ‘24-flavours’ tea

English name/Latin name Chinese name Posology Therapeutic indication

[15]
Chrysanthemum/Chrysanthemum 菊花 10–15 g It is used as an anti-inflammatory agent, and for the treatment
morifolium of headache, cold, vertigo and conjunctivitis.[16]
Dandelion/Taraxacum 蒲公英 9–15 g[15] Upper respiratory tract infection, acute tonsillitis, pharyngitis,
mongolicum 9–30 g[17] gastritis, enteritis, general fatigue, relief of cholecystitis,
abdominal pain, pyoderma, snakes bites.[15,17]
Japanese 金银花 10–30 g[17] It is used as an antibacterial and anti-inflammatory agent for
honeysuckle/Lonicera 10–60 g[15] the treatment of abscess, laryngitis, pharyngitis, infection of
japonica upper respiratory tract, dysentery, common cold and
fever.[15,16]

Downloaded from https://academic.oup.com/jpp/article/66/10/1478/6128004 by guest on 26 February 2021

Bishop’s weed/Houttuynia
cordata
Liquorice/Glycyrrhiza ularensis
Field mint/Mentha haplocalyx
鱼腥草 15–60 g[17] It is used as an anti-inflammatory, antidote and diuretic agent
for the treatment of pulmonary infection, bronchitis, asthma
oliguria and carbuncle.[16]
甘草 5–15 g i.e. 200–800 mg It is used as general tonic, an anti-inflammatory, mucolytic,
of glycyrrhizin[15] expectorant and analgesic agent for the treatment of
1–15 g (standard gastrointestinal and respiratory disorders.[16,17]
dose 3 g)[17]
薄荷 1–8 g[17] It is used to promote bile secretion, antiseptic and spasmolytic
action. It is used for gastrointestinal disorder, upper
respiratory infection and biliary disorder.[17]

of 1 ml/min: 0 min, 95%; 0–10 min, 75%; 10–20 min, 75%; Table 2 IC50 of 24-flavours teas (廿四味茶) on CYP3A4
20–36 min, 50%; 36–40 min, 18.4%; and 40–60 min, 18.4%. % of extract to
Concentration of IC50 reach IC50
Results 24-flavours tea the daily dose (g/l) (μg/ml) concentration.

24-flavours tea n°1 5.2 190.9 3.7

The experiments were designed to mimic closely the observed
clinical situation as reported in the case. For this reason, the
doses tested were close to those recommended by the herbal
tea manufacturers. Table 1 represents the usage and those
doses recommended for the six most common plants found
in ‘24-flavours’ teas identified in Hong Kong.[15–17] The daily
dose for the six plants investigated ranges between 1 g and
60 g. With regard to the recommended dose of each plant, it
was decided to extract 15 g with water. All of these plants
have some reported effect upon respiratory disease, as
explained in the chart below, and this may justify their his-
torical incorporation into ‘24-flavours’ teas (Table 1). The
plants described were extracted by water in the manner
described to obtain all those hydrophilic compounds that
would be present after preparation by the commercial tea
shop or in the domestic setting.
Between 2 g and 4 g of freeze-dried powder was obtained
for each tea/plant samples. The concentration of the daily
dose was calculated using the weight of the freeze-dried
powder divided by the amount of water used (300 ml).
To evaluate the inhibitory potency of the different plants, the
starting dose of 10 g/l was chosen for all in-vitro experiments.
Tables 2 and 3 present the concentration of a daily dose of
herbal tea or plant (weight in grams of extract obtained after
freeze-dried procedure divided by the extraction volume of
water). They are all higher than 10 g/l except for bishop’s
weed (Houttuynia cordata) and field mint (Mentha
haplocalyx). But bishop’s weed (Houttuynia cordata) daily
24-flavours tea n°2 10.8 337.1 3.1
posology is 15–60 g, as a dose of 15 g lead to a concentration
of 7 g/l, hence a dose of 20 g will reach 10 g/l. For field mint
(Mentha haplocalyx), 15 g is twice higher than the higher rec-
ommended dose. This means that a dose of 10 g/l is a realistic
clinical gut concentration.
CYP3A4 enzyme assay
The enzymatic testing of the ‘24-flavours’ teas was performed
at concentrations that reflected adult human doses. The rec-
ommendation made by Hong Kong herbal street sellers of
24-flavour tea is that an adult consumes the entire content of
one bottle as a single dose. Similarly those commercially avail-
able products where a customer prepares the 24-flavour tea
themselves, the instructions are that the full bag of plants
(approximate weight 250 g) should be used for one infusion
with water, resulting in a total volume after processing of
250 ml of beverage. Consequently, 250 ml of infused products
was reduced via freeze drying to dry residues.
The sigmoid curves of the percentage of inhibition of
CYP3A4 as a function of the concentration obtained after
curve fitting by REGTOX program are presented in
Figure 1. All of them showed inhibitory effect when 15 g of
plants were used.

© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 66, pp. 1478–1490 1481
Inhibitory effect of herbal teas on CYP3A4 Sophie Dufay et al.

Table 3 IC50 of CYP3A4 by chrysanthemum (Chrysanthemum maxima of compounds UV spectrum are different, chro-
morifolium), dandelion (Taraxacum mongolicum), Japanese honey- matograms are extracted at 208 nm and 254 nm, and iden-
suckle (Lonicera japonica), Bishop’s weed (Houttuynia cordata), liquo-
tification will be done with regard to retention time and
rice (Glycyrrhiza ularensis) and field mint (Mentha haplocalyx)
UV spectra, which improve the sensitivity of the detection
Concentration of but also the specificity of identification method in
Plants
the daily dose
such complex matrix. Retention time were hesperidin
Chinese of 15 g of IC50
(16.8 min), quercetin (29.6 min), naringenin (33.4 min),
English name/Latin name name plants (g/l) (μg/ml)
hesperetin (34.5 min), glycyrrhizin (36.0 min), oleanolic
Chrysanthemum/ 菊花 13.3 95.7 acid (55.0 min) and ursolic acid (55.7 min).
Chrysanthemum
With regard to the retention time of standards and their
morifolium

Downloaded from https://academic.oup.com/jpp/article/66/10/1478/6128004 by guest on 26 February 2021

Dandelion/Taraxacum 蒲公英 10
mongolicum
Japanese 金银花 15
Honeysuckle/Lonicera
japonica
Bishop’s weed/Houttuynia 鱼腥草 7 185.5
cordata
Liquorice/Glycyrrhiza 甘草 10
ularensis
Field mint/Mentha haplocalyx 薄荷 9 1153.3
The concentrations at which the CYP3A4 activity is
inhibited by 50% (IC50) are presented in Table 2 for the
two ‘24-flavours’ teas brands (廿四味茶), and in Table 3 for
individual plants. Table 4 showed the means and the stand-
ard deviations of the different concentration points.
Ketoconazole was used as a positive inhibitor. All plants
showed inhibition activity at tested doses (15 g). The ‘24-
flavours’ tea number 1 showed the same order of magnitude
of inhibitory effect than chrysanthemum (Chrysanthemum
morifolium), dandelion (Taraxacum mongolicum) and bis-
hop’s weed (Houttuynia cordata), and twice than that of the
‘24-flavours’ tea number 2.
The most potent plant, with respect to inhibitory effect
on CYP3A4, was found to be Chrysanthemum morifolium
with an IC50 of 95.7 (μg/ml). It was followed closely by
dandelion (Taraxacum mongolicum), liquorice (Glycyrrhiza
ularensis) and bishop’s weed (Houttuynia cordata), which
were approximately1.5 times less potent than chrysan-
themum. Field mint (Mentha haplocalyx) and Japanese
honeysuckle (Lonicera japonica) were found to be 10–15
times less potent than chrysanthemum.
Analysis of 24-flavour teas and plants
extract by HPLC-PDA analysis
Chemical compounds previously reported into literature to
be present in some of the plant tested and commercially
available with excellent purity were searched in the herbal
extracts used for the CYP3A4 assay. Hesperidin, quercetin,
naringenin, hesperetin, glycyrrhiza, oleanolic and ursolic
acid were effectively separated by HPLC-PDA method. As
140.6
spectrum, some known chemical compounds were identi-
fied in the plants extracts.
1466.3 Figure 2 showed the analysis of herbal extracts by HPLC-
PDA method. Quercetin was founded in Chrysanthemum
morifolium (Figure 2a) and Houttuynia cordata (Figure 2d).
Oleanolic acid and ursolic acid were founded in Taraxacum
mongolicum (Figure 2b) and Mentha hoplocalyx (Figure 2f).
148.4
Glycyrrhizin was identified in Glyccyrrhyzia ularensis
(Figure 2e). Hesperidin and hesperetin were found in
Mentha haplocalyx (Figure 2f). None of the tested com-
pounds were present either in Lonicera japonica nor in the
two brands of 24-flavour tea tested.
Discussion
‘24-Flavours’ teas (廿四味茶)
Testing the 24-flavours tea itself was important to evaluate
the total inhibitory effect of the herbal teas and in particular
the one drunk by the patient described in the initial clinical
case. The two brand of 24-flavour tea showed inhibitory
effect on CYP3A4 at clinical doses. This testing was done in
an attempt to confirm the doctor’s hypothesis that the
24-flavour tea was responsible for the inhibitory effects of
CYP3A4. However, none of the seven compounds screened
by HPLC-PDA were identified in the 24-flavour tea. It
might be due to the limit of detection of the analytical
method, which does not allow identification of the low level
of chemical compound even if they are present at concen-
tration at which they show inhibitory effect (data not
shown). Also, as the tea is a mixture of different plant, the
amount of each of them might be low and the total effect
might be due to a synergistic effect of inhibitory com-
pounds. As the formulation of 24-flavour tea varies, it was
important to try to identify the causative(s) plants included
into the formula.
Individual plants tested
Chrysanthemum morifolium/Chrysanthemum
Chrysanthemum morifolium showed the most potent inhibi-
tory effect on CYP3A4 among the six tested plant extracts.
Out of the seven standards tested, quercetin was founded in
the sample analysed. Three flavonoids also present in

1482 © 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 66, pp. 1478–1490
Sophie Dufay et al. Inhibitory effect of herbal teas on CYP3A4

(a) (b)
120 120
% Inhibition of Cytochrome

% Inhibition of Cytochrome
100 100
80
80
3A4 activity

3A4 activity
60
60
40
40
20
0 20

–20 0 1 2 3 4 0

Downloaded from https://academic.oup.com/jpp/article/66/10/1478/6128004 by guest on 26 February 2021

0 1 2 3 4
–40 –20
Log (24-flavour tea 1 extract concentration) in µg/ml Log (24 herbal tea 2 extract concentration) in µg/ml
(c) (d)
120 120
% Inhibition of Cytochrome

% Inhibition of Cytochrome
100 100

80 80
3A4 activity

3A4 activity
60 60

40 40

20 20

0 0
0 1 2 3 4 0 1 2 3 4
–20 –20
Log (Chrysanthemum extract concentration) µg/ml Log (Dandelion extract concentration) in µg/ml

(e) (f)
120
100
% Inhibition of Cytochrome

% Inhibition of Cytochrome

80 100

60 80
3A4 activity

3A4 activity

40 60

20 40
0 20
0 1 2 3 4
–20
0
–40 0 1 2 3 4
–20
Log (Japanese Honeysuckle extract concentration) in µg/ml
Log (Bishop’s weed extract concentration) in µg/ml
(g) (h)
120 120
% Inhibition of Cytochrome

% Inhibition of Cytochrome

100 100

80 80
3A4 activity

3A4 activity

60 60

40 40
20 20
0 0
0 0.5 1 1.5 2 2.5 3 3.5 4 0 1 2 3 4
–20
Log (Licorice extract concentration) in µg/ml Log (Field mint extract concentration) in µg/ml

Figure 1 Percentage of inhibition of CYP3A4 activity by the two different 24-flavours teas (廿四味茶). (a and b), and the six individual plants
bought in Western Market in Hong Kong Island, (c) Chrysanthemum (Chrysanthemum morifolium), (d) Dandelion (Taraxacum mongolicum), (e)
Japanese honeysuckle (Lonicera japonica), (f) Bishop’s weed (Houttuynia cordata), (g) Liquorice (Glycyrrhiza ularensis) and (h) Field mint (Mentha
haplocalyx).
© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 66, pp. 1478–1490 1483
Inhibitory effect of herbal teas on CYP3A4 Sophie Dufay et al.

Table 4 Means and standard deviations of Cytochrome 3A4 activity than kaempferol in the inhibition of human liver
24-flavour tea Another brand of microsomes.[24]
1 drunk by the 24-flaour tea
Concentration patient (number 1) (number 2) Taraxacum mongolicum/Mongolian dandelion
μg/ml Mean SD Mean SD The inhibitory effect of Taraxacum mongolicum can be
10000 100.3 1.4 96.5 1.7
linked to the presence of hesperetin, hesperidin and
3333.3 99.0 3.6 92.9 5.0 quercetin,[25] which are thought to be partially responsible
1111.1 93.5 6.2 80.5 4.8 for the effects of grapefruit juice on increasing the
370.4 71.5 9.2 48.0 1.4 bioavailability of ciclosporin.[19,20,24] When analysing the
123.5 38.4 9.3 21.1 4.0 extract by HLPC-PDA with hesperetin and quercetin stand-

Downloaded from https://academic.oup.com/jpp/article/66/10/1478/6128004 by guest on 26 February 2021

41.2 17.1 10.7 15.7 8.0 ard, none of these three compounds were identified, which
13.7 16.9 5.4 −2.8 10.2
might be due to the HPLC-PDA limit of detection. Ursolic
4.6 −18.2 1.2 −5.5 0.5
acid and oleanolic acid were newly reported as being
Chrysanthemum Taraxacum present in Taraxacum mongolicum. Oleanolic acid has been
Concentration morifolium mongolicum
reported to have inhibitory effect on CYP3A4 (IC50 of
μg/ml Mean SD Mean SD 78.4 μm), whereas ursolic acid has no inhibitory effect
10000 95.0 0.2 97.7 0.2 (IC50 > 500 μm).[26]
3333.3 99.3 0.9 99.7 1.2
1111.1 91.0 0.7 87.1 3.9 Glycyrrhiza ularensis/Liquorice
370.4 67.4 1.7 60.2 5.8
The inhibitory effect of Glycyrrhiza ularensis found in this
123.5 51.7 17.7 42.6 7.9
study has already been reported for a dose of 0.02 mg/ml.[27]
41.2 32.4 8.8 25.6 0.3
13.7 12.9 1.1 17.1 5.6 Another liquorice species, Glycyrrhiza glabra, has shown to
4.6 7.1 4.4 −9.3 6.9 inhibit CYP3A4 in human liver microsome experiments
with IC50 in a range of 1–2% of the full strength, which was
Lonicera Houttuynia
Concentration japonica cordata classified as a potent inhibitor by Budzinski et al.[28]
Glycyrrhiza ularensis has been reported to contain
μg/ml Mean SD Mean SD
glycyrrhizin, liquiritin and liquiritin aposide like all
10000 88.6 4.7 105.3 1.6 Glycyrrhiza species.[29] Glycyrrhizin was identified in the
3333.3 71.8 1.9 107.1 1.7 sample analysed by HPLC-PDA. Glycyrrhizin was reported
1111.1 37.7 3.7 102.6 2.3 to have no inhibitory effect (IC50 > 1.2 mM). However,
370.4 18.5 8.2 68.8 5.0
liquiritin and liquiritin aposide have IC50 of 57 and
123.5 7.5 16.3 26.2 10.6
41.2 4.5 7.9 28.2 24.6
655 μm, respectively, on human CYP3A4.[27]
13.7 0.9 1.9 14.9 3.4 Tsukamoto has also identify three other compounds
4.6 −2.7 14.8 8.0 18.0 in Glycyrrhiza ularensis that have potent inhibitory
Glycyrrhiza Mentha
effect on human CYP3A4: (3R)-vestitol (IC50 = 3.2 μm),
Concentration ularensis haplocalyx liquiritigenin 7,4’-digucoside (IC50 = 17 μm) and 4-
hydroxyguaiacol apioglucoside (IC50 = 20 μm).[27]
μg/ml Mean SD Mean SD

10000 101.5 0.3 92.0 1.1 Houttuynia cordata/Bishop’s weed

3333.3 97.5 3.4 88.3 1.8
The inhibitory effect was less potent but might be much
1111.1 90.5 0.3 60.6 1.0
370.4 65.3 4.8 35.3 2.1 more important if the dose used in the test is increased with
123.5 35.7 6.1 17.2 9.2 regard to recommended posology, 15–60 g. It has been
41.2 32.3 11.6 15.0 0.2 reported that Houttynia cordata contains quercetin,[30–36]
13.7 −2.5 6.3 0.7 1.3 which was also identified in the sample used in this current
4.6 18.7 23.1 2.4 1.3 research. Quercetin is thought to be involved in the increase
SD, standard deviation. of ciclosporin bioavailability with the consumption of
grapefruit juice.[20,24,37]

Mentha haplocalyx/Field mint

Chrysanthemum moriflorium have been thought to be
involved in causing an increase in ciclosporin serum levels The mild inhibitory effect of Mentha haplocalyx evalu-
by grapefruit juice:[14,18–20] hesperetin,[21] kaempferol[22] and ated in our study has also been reported with the use of
quercetin.[21–23] Quercetin was found slightly more potent peppermint oil on CYP3A4.[38] Hesperetin, hesperidin,

1484 © 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 66, pp. 1478–1490
Sophie Dufay et al. Inhibitory effect of herbal teas on CYP3A4

(a) - Chrysanthemum morifolium (Chrysanthemum)

mAU
350
300
250
200
150
100

Downloaded from https://academic.oup.com/jpp/article/66/10/1478/6128004 by guest on 26 February 2021

50
0
mAU
350
300
250
200
150
100
50
0
10 20 30 40 50 60 min
(b) - Taraxacum mongolicum (Dandelion)

mAU
35
30
25
20
15
10
5
0

mAU
35
30
25
20
15
10
5
0
min
10 20 30 40 50 60

Figure 2 HPLC-PDA analysis of plant extracts to identify chemical compounds. Chromatograms were extracted at 208 (ursolic acid and oleanolic
acid) and 254 nm (hesperidin, quercetin, naringenin, hesperetin and glycyrrhizin). (a) Chrysanthemum (Chrysanthemum morifolium). (b) Dandelion
(Taraxacum mongolicum). (c) Japanese honeysuckle (Lonicera japonica). (d) Bishop’s weed (Houttuynia cordata). (e) Liquorice (Glycyrrhiza ularensis).
(f) Field mint (Mentha haplocalyx).

© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 66, pp. 1478–1490 1485
Inhibitory effect of herbal teas on CYP3A4 Sophie Dufay et al.

(c) - Lonicera japonica (Honeysuckle)

350
300
250
200
150
100
50

Downloaded from https://academic.oup.com/jpp/article/66/10/1478/6128004 by guest on 26 February 2021

0

350
300
250
200
150
100
50
0
10 20 30 40 50 60 min
(d) - Houttuynia cordata (Bishop’s weed)
mAU
350
300
250
200
150
100
50
0
mAU

120

100

80

60

40

20

0
min
10 20 30 40 50 60

Figure 2 Continued

1486 © 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 66, pp. 1478–1490
Sophie Dufay et al. Inhibitory effect of herbal teas on CYP3A4

(e) - Glycyrrhiza ularensis (Licorice)

mAU
350
300
250
200
150
100
50

Downloaded from https://academic.oup.com/jpp/article/66/10/1478/6128004 by guest on 26 February 2021

0

350
300
250
200
150
100
50
0 min
10 20 30 40 50 60

(f) - Mentha haplocalyx (Field mint)

mAU
350
300
250
200
150
100
50
0
mAU
350
300
250
200
150
100
50
0 min
10 20 30 40 50 60

Figure 2 Continued

© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 66, pp. 1478–1490 1487
Inhibitory effect of herbal teas on CYP3A4 Sophie Dufay et al.

oleanolic and ursolic acid were identified in the sample ana-
lysed by HPLC-PDA. Hesperetin and hesperidin have been
described in the literature as being responsible for the
inhibitory effect on CYP3A4, and thus increased the
metabolism of ciclosporin.[19,20,24] Oleanolic acid has also
been reported to have such effect when tested on liver
microsomes.[26] Naringenin has been identified to be one
compound of Mentha haplocalyx species but was not found
in this study possibly because of the low sensitivity of the
method. It was thought to be one of the compounds
The analysis of plant extracts by HPC-PDA helped dem-
onstrate the presence of chemical compounds (quercetin,
hesperidin, hesperetin and oleanolic acid) possibly respon-
sible of inhibitory effect on CYP3A4 in some of the herbal
extracts studied. The implication of chemical compounds
previously reported in the literature in some of the used
plants but not identified in this study cannot be ruled out as
their concentration at which they exert their inhibitory
effect might be lower than the ability of the HPLC-PDA
method to identify them.

Downloaded from https://academic.oup.com/jpp/article/66/10/1478/6128004 by guest on 26 February 2021

responsible for the grapefruit juice interaction with
ciclosporin.[19,37,39,40]
Some in-vivo studies conducted by Wacher et al. revealed
a significant increase of about threefold in ciclosporin peak
plasma concentrations and area under curve when a dose of
100 mg/kg of peppermint oil was administered in rats.[41]
The in-vitro liver microsomes test showed peppermint oil
was a less potent CYP3A4 inhibitor than ketoconazole,
while a fourfold increase in dose was required to elicit the
same degree of inhibition on ciclosporin metabolism in
human liver microsomes than in rats.[41] Another in-vivo
trial by Dresser et al. showed significant increase in area
under curve and peak plasma concentration of felodipine,
mainly metabolized by CYP3A4, in 12 subjects challenged
with 330 μl of peppermint oil in 300 ml water.[42] The mean
increase was 140% compared with subjects with water
only.[42] With unchanged half-life, it was suggested that the
preparation predominantly affect the enteric enzymes.[42]
The in-vitro test identified peppermint oil and its constitu-
ents, menthol and menthyl acetate, as reversible, partially
mixed CYP3A4 inhibitors in liver microsome test.[42]
Maliakal et al. have reported no inhibitory effect but this
was after chronic treatment for 4 weeks with peppermint tea
(2% m/v).[43]
Lonicera japonica/Japanese honeysuckle
Lonicera japonica showed some mild inhibitory effect on
CYP3A4 (IC50 = 1466 μg/ml) compared with the other five
plants. Pao et al. have tested water extracts of Lonicera
japonica at three different concentrations (2, 4 and 6 mg/
ml) on human liver microsomes and did not see any inhibi-
tory effect.[44] This might be explained by the usage of liver
microsomes technique, containing all liver cytochromes,
compared with individual cytochrome testing. Also, none of
the seven chemical compounds screened by HPLC-PDA
were identified into Lonicera japonica, possibly due to the
low sensitivity of HPLC-PDA method.
Conclusions
This article was aimed to present a methodology mimicking
closely clinical situation to investigate clinical case of pos-
sible herb–drug interaction when no information was avail-
able in the literature. The in-vitro experiments on CYP3A4
described above utilizing aqueous extracts at representative
clinical doses have demonstrated that ‘24-flavours’ teas and
the six mostly common plants have potential to inhibit the
metabolism function of CYP3A4.
The inhibitory effects of Taraxacum mongolicum, Chry-
santhemum morifolium and Houttuynia cordata can most
likely be explained by the presence of specific chemical
compounds that has already been proposed to play a role
in the interaction of grapefruit juice with ciclosporin
(hesperetin, kaempferol, quercetin and hesperidin) but also
by the presence of other compounds, ursolic acid and
oleanolic acid. Further work is needed to purify and identify
other compounds present in the plant as shown on chro-
matograms to test potency of their inhibitory effect.
Doctor should ask for herbal consumption in front of
unexplained drug toxicity and put in place clinical surveil-
lance programme especially for patients treated with immu-
nosuppressant. Similar attitude should be adopted with
other drugs known to interact with grapefruit juice: antiar-
rhythmics, calcium channel blockers, statins, cytotoxic
agents, carbamazepine or some antipsychotics. Further sci-
entific evaluations of potential herb–drug interaction are
needed.
Declarations
Funding
This study has been funded by the Department of Pharma-
cology and Pharmacy of the University of Hong Kong.

References
1. Kaufman DW et al. Recent patterns of
medication use in the ambulatory
adult population of the United States
– the Slone survey. JAMA 2002; 287:
337–344.
2. Duggan J et al. Use of comple-
mentary and alternative therapies
in HIV-infected patients. AIDS
Patient Care STDS 2001; 15: 159–
167.
3. Engdal S et al. Identification and
exploration of herb-drug combi-
nations used by cancer patients.

1488 © 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 66, pp. 1478–1490
Sophie Dufay et al.
Integr Cancer Ther 2009; 8: 29–
36.
4. Molassiotis A et al. Use of comple-
mentary and alternative medicine in
cancer patients: a European survey.
Ann Oncol 2005; 16: 655–663.
5. Bush TM et al. Adverse interactions
between herbal and dietary substances
and prescription medications: a clini-
cal survey. Altern Ther Health Med
2007; 13: 30–35.
6. Shi S, Klotz U. Drug interactions with
herbal medicines. Clin Pharmacokinet
2012; 51: 77–104.
7. Hu Z et al. Herb-drug interactions: a
literature review. Drugs 2005; 65:
1239–1282.
8. Izzo AA, Ernst E. Interactions between
herbal medicines and prescribed
drugs: an updated systematic review.
Drugs 2009; 69: 1777–1798.
9. De Lima Toccafondo Vieira M, Huang
SM. Botanical-drug interactions: a sci-
entific perspective. Planta Med 2012;
78: 1400–1415.
10. Mukherjee PK et al. Botanicals as
medicinal food and their effects on
drug metabolizing enzymes. Food
Chem Toxicol 2011; 49: 3142–3153.
11. Kwan LP et al. Acute drug toxicity
related to drinking herbal tea in a
kidney transplant recipient. Ren Fail
2014; 36: 309–312.
12. Nowack R, Nowak B. Herbal teas
interfere with cyclosporin levels in
renal transplant patients. Nephrol Dial
Transplant 2005; 20: 2554–2556.
13. Ohno Y et al. General framework for
the prediction of oral drug interac-
tions caused by CYP3A4 induction
from in vivo information. Clin
Pharmacokinet 2008; 47: 669–680.
14. Seden K et al. Grapefruit-drug inter-
actions. Drugs 2010; 70: 2373–2407.
15. World Health Organization. Medicinal
Plants in China: A Selection of 150
Commonly Used Species. Singapore:
World Health Organization, Regional
Office for the Western Pacific, 1989.
16. Tang W, Eisenbrand G. Handbook
of Chinese Medicinal Plants: Chemi-
stry, Pharmacology, Toxicology, Vol. 1.
Weinheim, UK: Wiley-VCH, 2011.
17. Hempen C-H, Fischer T. A Materia
Medica for Chinese Medicine: Plants,
© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 66, pp. 1478–1490
Minerals, and Animal Products.
London: Elsevier Health Sciences,
2009.
18. Lin SP et al. Citrus grandis peel
increases the bioavailability of cyclo-
sporine and tacrolimus, two impor-
tant immunosuppressants, in rats. J
Med Food 2011; 14: 1463–1468.
19. Fujita T et al. Comparative evalu-
ation of 12 immature citrus fruit
extracts for the inhibition of Cyto-
chrome P450 isoform activities. Bio-
logical & Pharmaceutical Bulletin
2008; 31: 925–930.
20. Ho PC et al. Inhibition of human
CYP3A4 activity by grapefruit flavo-
noids, furanocoumarins and related
compounds. J Pharm Pharm Sci 2001;
4: 217–227.
21. Wang YJ et al. Studies on chemical
constituents in Huangjuhua (flowers
of Chrysanthemum morifolium).
Zhongguo Zhong Yao Za Zhi 2008; 33:
526–530.
22. Lai JP et al. Identification and cha-
racterization of major flavonoids
and caffeoylquinic acids in three
Compositae plants by LC/DAD-
APCI/MS. J Chromatogr B Analyt
Technol Biomed Life Sci 2007; 848:
215–225.
23. Miyazawa M, Hisama M. Antimu-
tagenic activity of flavonoids from
Chrysanthemum morifolium. Biosci
Biotechnol Biochem 2003; 67: 2091–
2099.
24. Miniscalco A et al. Inhibition of
dihydropyridine metabolism in rat and
human liver microsomes by flavonoids
found in grapefruit juice. J Pharmacol
Exp Ther 1992; 261: 1195–1199.
25. Shi S et al. Identification of
antioxidants from Taraxacum mon-
golicum by high-performance
liquid chromatography-diode array
detection-radical-scavenging detection-
electrospray ionization mass spectro-
metry and nuclear magnetic resonance
experiments. J Chromatogr A 2008; 31:
1–2.
26. Kim KA et al. Inhibition of
Cytochrome P450 activities by
oleanolic acid and ursolic acid in
human liver microsomes. Life Sci
2004; 74: 2769–2779.
Inhibitory effect of herbal teas on CYP3A4
27. Tsukamoto S et al. CYP3A4 inhibitors
isolated from licorice. Biol Pharm Bull
2005; 28: 2000–2002.
28. Budzinski JW et al. An in vitro evalu-
ation of human Cytochrome P450
3A4 inhibition by selected commer-
cial herbal extracts and tinctures.
Phytomedicine 2000; 7: 273–282.
29. Kondo K et al. Constituent properties
of licorices derived from Glycyrrhiza
uralensis, G. glabra, or G. inflata iden-
Downloaded from https://academic.oup.com/jpp/article/66/10/1478/6128004 by guest on 26 February 2021
tified by genetic information. Biologi-
cal & pharmaceutical bulletin 2007; 30:
1271–1277.
30. Cho EJ et al. The inhibitory effects of
12 medicinal plants and their compo-
nent compounds on lipid peroxi-
dation. Am J Chin Med 2003; 31: 907–
917.
31. Chou SC et al. The constituents and
their bioactivities of Houttuynia
cordata. Chem Pharm Bull 2009; 57:
1227–1230.
32. Meng J et al. Simultaneous quantifica-
tion of eight bioactive components
of Houttuynia cordata and related
Saururaceae medicinal plants by
on-line high performance liquid
chromatography-diode array detector-
electrospray mass spectrometry. Fito-
terapia 2009; 80: 468–474.
33. Meng J et al. Study on chemical con-
stituents of flavonoids in fresh herb of
Houttuynia cordata. Zhongguo Zhong
Yao Za Zhi 2006; 31: 1335–1337.
34. Meng J et al. Establishment of HPLC-
DAD-MS fingerprint of fresh Hout-
tuynia cordata. Chem Pharm Bull
2005; 53: 1604–1609.
35. Xu X et al. Determination of flavo-
noids in Houttuynia cordata Thunb.
and Saururus chinensis (Lour.) Bail. by
capillary electrophoresis with electro-
chemical detection. Talanta 2006; 68:
759–764.
36. Zhang TT et al. [Study on HPLC
fingerprint of flavonoids from Hout-
tuynia cordata by comparing with
fingerprint reference]. Zhong Yao Cai
2009; 32: 687–690.
37. Dahan A, Altman H. Food-drug inter-
action: grapefruit juice augments drug
bioavailability–mechanism, extent and
relevance. Eur J Clin Nutr 2004; 58:
1–9.
1489
Inhibitory effect of herbal teas on CYP3A4
38. Unger M, Frank A. Simultaneous
determination of the inhibitory
potency of herbal extracts on the
activity of six major Cytochrome
P450 enzymes using liquid
chromatography/mass spectrometry
and automated online extraction.
Rapid Commun Mass Spectrom 2004;
18: 2273–2281.
39. Dorman HJ et al. Antioxidant proper-
ties and composition of aqueous
extracts from Mentha species, hybrids,
varieties, and cultivars. J Agric Food
Chem 2003; 51: 4563–4569.
1490
40. Bailey DG et al. Grapefruit juice-drug
interactions. Br J Clin Pharmacol 1998;
46: 101–110.
41. Wacher VJ et al. Peppermint oil
enhances cyclosporine oral bioa-
vailability in rats: comparison with
D-alpha-tocopheryl poly(ethylene
glycol 1000) succinate (TPGS) and
ketoconazole. J Pharm Sci 2002; 91:
77–90.
42. Dresser GK et al. Evaluation of
peppermint oil and ascorbyl palmi-
tate as inhibitors of Cytochrome
P4503A4 activity in vitro and in
© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 66, pp. 1478–1490
Sophie Dufay et al.
vivo. Clin Pharmacol Ther 2002; 72:
247–255.
43. Maliakal PP, Wanwimolruk S. Effect
of herbal teas on hepatic drug
metabolizing enzymes in rats. J Pharm
Pharmacol 2001; 53: 1323–1329.
44. Pao LH et al. Herb-drug interaction of
50 Chinese herbal medicines on
CYP3A4 activity in vitro and in vivo.
Am J Chin Med 2012; 40: 57–73.
Downloaded from https://academic.oup.com/jpp/article/66/10/1478/6128004 by guest on 26 February 2021