Inhibitory effects of silibinin on cytochrome P-450 enzymes in human liver microsomes

Article abstract of DOI:10.1111/j.0901-9928.2000.860602.x

Silibinin, the main constituent of silymarin, a flavonoid drug from silybum marianum used in liver disease, was tested for inhibition of human cytochrome P-450 enzymes. Metabolic activities were determined in liver microsomes from two donors using selective substrates. With each substrate, incubations were carried out with and without silibinin (concentrations 3.7- 300 μM) at 37°in 0.1 M KH₂PO₄ buffer containing up to 3% DMSO. Metabolite concentrations were determined by HPLC or direct spectroscopy. First, silibinin IC₅₀ values were determined for each substrate at respective K(M) concentrations. Silibinin had little effect (IC₅₀ > 200 μM) on the metabolism of erythromycin (CYP3A4), chlorzoxazone (CYP2E1), S(+)-mephenytoin (CYP2C19), caffeine (CYP1A2) or coumarin (CYP2A6). A moderate effect was observed for high affinity dextromethorphan metabolism (CYP2D6) in one of the microsomes samples tested only (IC₅₀ = 173 μM). Clear inhibition was found for denitronifedipine oxidation (CYP3A4; IC₅₀ = 29 μM and 46 μM) and S(-)-warfarin 7-hydroxylation (CYP2C9; IC₅₀ = 43 μM and 45 μM). When additional substrate concentrations were tested to assess enzyme kinetics, silibinin was a potent competitive inhibitor of dextromethorphan metabolism at the low affinity site, which is not CYP2D6 (K(i,c) = 2.3 μM and 2.4 μM). Inhibition was competitive for S(-)-warfarin 7-hydroxylation (K(i,c) = 18 μM and 19 μM) and mainly non-competitive for denitronifedipine oxidation (K(i,n) = 9 μM and 12 μM). With therapeutic silibinin peak plasma concentrations of 0.6 μM and biliary concentrations up to 200 μM, metabolic interactions with xenobiotics metabolised by CYP3A4 or CYP2C9 cannot be excluded.

Full text of DOI:10.1111/j.0901-9928.2000.860602.x

C Pharmacology & Toxicology 2000, 86, 250–256.
Copyright C
Printed in Denmark . All rights reserved
ISSN 0901-9928
Inhibitory Effects of Silibinin on Cytochrome P-450 Enzymes
in Human Liver Microsomes
Svane Beckmann-Knopp1, Stephan Rietbrock1, Roland Weyhenmeyer2, Ronald H. Böcker3, K. Tobias Beckurts4,
Werner Lang2, Miriam Hunz1 and Uwe Fuhr1
1Institute for Pharmacology, Clinical Pharmacology, University of Köln; Madaus AG, Köln; Department of
2
3
4
Experimental and Clinical Pharmacology and Toxicology, University of Erlangen; Department of Visceral and Vascular
Surgery, University of Köln, Germany
(Received November 3, 1999; Accepted January 13, 2000)
Abstract: Silibinin, the main constituent of silymarin, a flavonoid drug from silybum marianum used in liver disease, was
tested for inhibition of human cytochrome P-450 enzymes. Metabolic activities were determined in liver microsomes
from two donors using selective substrates. With each substrate, incubations were carried out with and without silibinin
(concentrations 3.7–300 mM) at 37æ in 0.1 M KH2PO4 buffer containing up to 3% DMSO. Metabolite concentrations
were determined by HPLC or direct spectroscopy. First, silibinin IC50 values were determined for each substrate at
respective KM concentrations. Silibinin had little effect (IC50Ͼ200 mM) on the metabolism of erythromycin (CYP3A4),
chlorzoxazone (CYP2E1), S(π)-mephenytoin (CYP2C19), caffeine (CYP1A2) or coumarin (CYP2A6). A moderate effect
was observed for high affinity dextromethorphan metabolism (CYP2D6) in one of the microsomes samples tested only
(IC50Ω173 mM). Clear inhibition was found for denitronifedipine oxidation (CYP3A4; IC50Ω29 mM and 46 mM) and
S(ª)-warfarin 7-hydroxylation (CYP2C9; IC50Ω43 mM and 45 mM). When additional substrate concentrations were tested
to assess enzyme kinetics, silibinin was a potent competitive inhibitor of dextromethorphan metabolism at the low affinity
site, which is not CYP2D6 (Ki,cΩ2.3 mM and 2.4 mM). Inhibition was competitive for S(ª)-warfarin 7-hydroxylation
(Ki,cΩ18 mM and 19 mM) and mainly non-competitive for denitronifedipine oxidation (Ki,nΩ9 mM and 12 mM). With
therapeutic silibinin peak plasma concentrations of 0.6 mM and biliary concentrations up to 200 mM, metabolic interac-
tions with xenobiotics metabolised by CYP3A4 or CYP2C9 cannot be excluded.
Silibinin (fig. 1) is the main isomer of a group of flavonoids
summarised as silymarin. It is obtained from the seeds of
the milk thistle silybum marianum. Silidianin, isosilibinin
and silicristin are the other isomers present in seeds. As a
standardised extract, silymarin is therapeutically used for
the treatment of acute and chronic liver injury (Leng-Pesch-
low & Strenge-Hesse 1991; Flora et al. 1998).
Mean maximum plasma concentration after an oral dose
of 700 mg silymarin, containing 254 mg of silibinin, is 317
ng/ml or 0.6 mM. Accumulation in plasma during three
daily medications is negligible. Plasma protein binding is
reported to reach about 90–95%. Elimination occurs pre-
dominantly via the bile, where accumulation of silibinin
may result in maximum concentrations of 150 mM. Mean
elimination half-life of silibinin is about 6 hr (Lorenz et al.
1982; Weyhenmeyer et al. 1992).
thane (Siegers et al. 1983), acetaminophen (Muriel et al.
1992) and phenylhydrazine (Valenzuela et al. 1985). Several
mechanisms of action were discussed: (1) stimulation of
protein synthesis in the hepatocytes (Machicao & Sonnen-
bichler 1977), (2) radical scavenging and antioxidant prop-
erties (Ubeda et al. 1995), (3) prevention of penetration of
toxins into the cell (Münter et al. 1986), and (4) inhibition
of 5-lipoxygenase with reduced formation of leucotriens
(Dehmlow et al. 1996).
Another possible mechanism of hepatoprotection may be
the inhibition of the cytochrome P-450 enzymes responsible
for the bioactivation of hepatotoxins such as ethanol (Al-
bano et al. 1999), halothane (Kharasch et al. 1996) and
aflatoxin B1 (Guengerich & Kim 1990). Indeed, inhibitory
silibinin/silymarin effects have been observed on oxidation
of various cytochrome P-450 substrates in rat and mouse,
including aminopyrine, ethanol, benzo(a)pyrene, hexobarbi-
tal, and 7-ethoxycoumarin (Down et al. 1974; Valenzuela et
al. 1989; Lette´ron et al. 1990; Ubeda et al. 1995). In one
report however, induction of the metabolism of aminopyr-
ine and nitroanisole by silymarin was found in rat, whereas
elimination half-life of aminopyrine or phenylbutazone
examined in 16 healthy volunteers each was not affected by
long-term administration of 70 mg of silymarin three times
daily (Leber & Knauff 1976). Silibinin at concentrations of
50 mM or 500 mM had no protective effects in isolated hu-
man hepatocytes treated with tetrachlormethan or acet-
Both silibinin and silymarin have protective properties
for hepatocytes in several in vitro studies and in animal
models. In these investigations, such flavonoids reduced he-
patocyte damage caused by administration of various
chemicals including, among others, allyl alcohol (Miguez et
al. 1994), carbon tetrachloride (Mourelle et al. 1989), halo-
Author for correspondence: Uwe Fuhr, Institute of Pharmacology,
Clinical Pharmacology, University of Koln, Gleueler Str. 24,
¨
D-50931 Koln, Germany (fax π49 221 478 5022, e-mail uwe.
¨
fuhr/medizin.uniko eln.de).

SILIBININ EFFECTS ON HUMAN CYTOCHROME P-450
251
purity 93.4%) was provided by the Department of Chemistry, Ma-
daus AG, Germany. For the sources of all other chemicals see
Beckmann-Knopp et al. (1999).
aminophen which are activated by CYP2E1. The con-
clusion of the authors was that silibinin does not inhibit
this enzyme (Miguez et al. 1994).
Metabolic activities of cytochrome P-450 enzymes were deter-
mined in liver microsomes from two donors each using the follow-
ing selective substrates: dextromethorphan (CYP2D6), erythro-
mycin or denitronifedipine (CYP3A4), caffeine (CYP1A2), chlor-
zoxazone (CYP2E1), S-(π)-mephenytoin (CYP2C19), S-(ª)-
warfarin (CYP2C9) or coumarin (CYP2A6) (see table 1). With each
substrate except caffeine, incubations were carried out without a
possible inhibitor and in presence of silibinin at five different con-
centrations (3.7, 11.1, 33.3, 100 or 300 mM) at 37æ in 0.1 M KH2PO4
buffer containing up to 3% DMSO. For caffeine incubations, silibi-
nin concentrations were 9, 30, 90, and 300 mM, and DMSO was not
used in this case. First, silibinin IC50 values were determined for
each substrate at respective KM concentrations. If IC50 values were
below 200 mM, additional substrate concentrations were tested to
fully assess enzyme kinetics. These were 1.9, 5.6, 16.7, 50 and 100
mM for denitronifedipine, 0.2, 1, 5, 40, 200, 1000, and 2000 mM for
dextromethorphan, and 0.3, 1, 3, and 10 mM for S-(ª)-warfarin,
respectively. With the exception of erythromycin metabolism, the
procedures used to assess individual enzyme activity are fully de-
scribed elsewhere (Beckmann-Knopp et al. 1999). Incubations were
usually carried out in duplicate, and means given in all cases.
To monitor erythromycin metabolism, the formation of form-
aldehyde was measured (Nash 1953, Riley & Howbrook 1997; Wang
et al. 1997). Described KM values for erythromycin metabolism by
CYP3A4 range between 33 and 88 mM. Incubations (2 mg protein/
ml) were carried out for 30 min. as described (Beckmann-Knopp et
al. 1999) and stopped by addition of 100 ml of 20% trichloracetic
acid. After centrifugation, 350 ml of the supernatant was diluted
with 150 ml Nash reagent (0.15 mol/l ammonium acetate and 0.2
mol/l acetylacetone in 3% acetic acid), and transferred into a water
bath of 53æ for 30 min. Colorimetric quantification of the Hantzsch
reaction solution was done directly by absorbance (405 nm) at room
temperature using a MAXline Microplate Reader (Sunnyvale, Cali-
fornia, USA). The limit of quantification was 7.8 nM, correspond-
ing to a formaldehyde formation rate of 0.1 pmol/min/mg protein.
Evaluation of the enzyme kinetic parameters was done based on
extended Michaelis – Menten equations as described, and best
models were determined using log-likelyhood tests (Beckmann-
Knopp et al. 1999).
Besides these equivocal data, there is indirect evidence
that silibinin may inhibit human cytochrome P-450 enzymes
other than CYP2E1. For many flavonoids, inhibition of hu-
man CYP3A4 has been shown (Fuhr 1998), giving rise to
metabolic drug–drug interactions, as CYP3A4 plays a pre-
dominant role in the metabolism of many important drugs
like calcium channel antagonists, cyclosporine, antihistam-
ines, midazolam, erythromycin and quinidine (Kinirons et
al. 1993). Indeed, CYP3A4 is also a major enzyme in activ-
ating toxic and carcinogenic chemicals (Guengerich & Shi-
mada 1991). Additional cytochrome P-450 enzymes such as
CYP1A2 or CYP2A6 may also be affected by flavonoids
(Ubeda et al. 1995; Dai et al. 1997; Fuhr 1998, Lee et al.
1998).
Thus, it is important to characterise the effect of silibinin
on cytochrome P-450 enzymes to further elucidate the
mechanism of action of the flavonoid (see above) and to be
aware of possible metabolic interactions with other drugs
administered concomitantly (Committee for Proprietary
Medicinal Products 1997). In the present investigation the
inhibitory effect of silibinin on the activity of the seven most
important human cytochrome P-450 enzymes in drug met-
abolism, including CYP2D6, 3A4, 1A2, 2E1, 2C19, 2C9,
and 2A6, was studied in human liver microsomal incuba-
tions using selective substrates (Fuhr et al. 1996a; Bertz &
Grannemann 1997). Although Miguez et al. (1994) had al-
ready published data related to silibinin effects on human
CYP2E1, we also included CYP2E1 in our studies for com-
parison. In the case of CYP3A4, two model substrates were
used, i.e. denitronifedipine (Böcker et al. 1986) and erythro-
mycin (Riley & Howbrook 1997; Wang et al. 1997), because
for this enzyme multiple conformers with distinct substrate
specificity may exist (Koley et al. 1996 & 1997).
Results
Materials and Methods
In the absence of any inhibitor, substrate turn-over was
clearly above the limits of detection for all human cyto-
chrome P-450 enzymes tested (CYP3A4, 2D6, 2E1, 2C19,
Erythromycin was purchased from Sigma Biochemikalien und Re-
agenzien, Deisenhofen, Germany. Silibinin (batch . 881122/ICH,
Table 1.
Effect of silibinin on activities of human cytochrome P-450 enzymes at KM substrate concentrations.
Cytochrome
P-450 enzyme
Substrates and concentrations used/(mM)
V0*/(pmol/min/mg protein)
IC50 of silibinin*/mM
CYP3A4
CYP3A4
CYP2D6
Unknown
CYP2E1
CYP2C19
CYP2C9
CYP1A2
CYP2A6
Denitronifedipine
Erythromycin
Dextromethorphan high affinity site
Dextromethorphan low affinity site
Chlorzoxazone
S-(π)-Mephenytoin
S-(ª)-Warfarin
Caffeine
50
37
5
2000
50
20
5
500
2
409 (P10)/279 (P14)
277 (P13)/509 (HO12)
79 (P10)/49 (P20)
232 (P10)/141 (P20)
115 (P11)/101 (P13)
13 (P11)/15 (P13)
0.39 (P13)/0.54 (HO12)
26 (P13)/65 (P17)
468 (P13)/494 (HO12)
29 (P10)/46 (P14)
Ͼ200 (P13)/Ͼ200 (HO12).
173 (P10)/¬(P20)
384 (P10)/¬ (P20)
799 (P11)/460 (P13)
309 (P11)/424 (P13)
43 (P13)/45 (HO12)
20000 (P13)/108500 (P17)
541 (P13)/688 (HO12)
Coumarin
v0Ωmetabolite formation rates in absence of an inhibitor; IC50Ωinhibitor concentration reducing metabolite formation rate by 50% at the
given substrate concentration (Beckmann-Knopp et al. 1999); *data pairs were obtained using human liver microsomes from two different
donors each, for characterisation of donors given in parentheses see Fuhr et al. (1996b); .proper calculation not possible because of activation
at low inhibitor concentrations.

252
SVANE BECKMANN-KNOPP ET AL.
Table 2.
Enzyme kinetic parameters obtained for dextromethorphan O-demethylation and inhibitory effect of silibinin in human liver microsomes.
Binding site
Microsome sample*
Vmax/(pmol/min./mg protein)
KM/(mM)
Ki,c/(mM)
High affinity site
(CYP2D6)
Low affinity site
(not known)
P10
P20
P10
P20
201
110
66
7.0
4.6
501
703
112
135
2.3
2.4
75
parameters were estimated using equation 1; *for identification of microsome samples, see Fuhr et al. (1996b); VmaxΩmaximal metabolite
formation rate; KMΩMichaelis-Menten constant; Ki,cΩinhibition constant for competitive inhibition.
2C9, 2A6) in liver microsomes preparations each from two
donors.
Thus, silibinin was characterised as a competitive inhibi-
tor of dextromethorphan metabolism, but a high inhibition
potency was only observed for the low affinity binding site,
with Ki values of 2.3 and 2.4 mM, respectively (table 2).
The graphical display of data points and the curve fitting
achieved using equation 1 is presented in fig. 2.
Silibinin had only little effect on the metabolism of five
of the eight substrates investigated. For erythromycin met-
abolism, IC50 could not be determined reliably because of
activation (Ͻ20%) at low silibinin concentrations, but it was
clearly above 200 mM. Also, the effects of silibinin on the
metabolism of chlorzoxazone, S-(π)-mephenytoin, caffeine
and coumarin were only minor, with IC50Ͼ200 mM in all
cases (table 1).
7-Hydroxywarfarin formation rates in both microsomal
A moderate effect was observed for dextromethorphan
metabolism at the high affinity site in one microsomal
sample only (IC50Ω173mM and ¬ mM), whereas the effect
on metabolism at the low affinity site appeared to be only
minor (table 1). Clear inhibition was found for denitronif-
edipine metabolism (IC50Ω29 mM and 46 mM) and S(ª)-
warfarin metabolism (IC50Ω43 mM and 45 mM). Therefore,
additional concentrations of dextromethorphan, S(ª)-war-
farin, and denitronifedipine were tested to fully describe ex-
tent and type of inhibitory action of silibinin in these cases.
Dextromethorphan formation rates in both microsomal
samples could be fitted best to a Michaelis Menten equation
with two binding sites, each with a competitive inhibition
constant Ki,c(1) and Ki,c(2) (equation 1).
Equation 1: Best model describing dextromethophan met-
abolism by human liver microsomes
V
max(1)¿[S]
V
max(2)¿[S]
VΩ
π
[I]
[I]
K
M(1)¿(1π
)π[S]
K
M(2)¿(1π
)π[S]
Ki,c(1)
Ki,c(2)
numeralsΩindices for the respective binding site; vΩmeasured meta-
bolite formation rate; VmaxΩmaximal metabolite formation rate;
[S]Ωsubstrate concentration; KMΩMichaelis-Menten constant; [I]Ω
inhibitor concentration; Ki,c, Ki,nΩinhibitor constants for competi-
tive (c) and non-competitive (n) inhibition, respectively.
Fig. 2. Linear plots: Inhibitory effect of silibinin on dextromethor-
phan O-demethylation. a: P10 liver microsomes; b: P20 liver micro-
somes.
For characterisation of microsome samples, see Fuhr et al. (1996b).
Fig. 1. Chemical structure of silibinin.
Lines were fitted according to equation 1.

SILIBININ EFFECTS ON HUMAN CYTOCHROME P-450
253
Table 3.
Enzyme kinetic parameters obtained for S(ª)-warfarin 7-hydroxy-
lation and inhibitory effect of silibinin in human liver microsomes.
Microsome
sample*
Vmax
KM
/(mM)
Ki,c
/(mM)
/(pmol/min./mg protein)
P13
HO12
0.550
0.760
2.8
2.3
18
19
parameters were estimated using equation 2; *for identification of
microsome samples, see (Fuhr et al. 1996b); for explanation of sym-
bols see legend to table 2.
samples (fig. 3a and b) were described best by a Michaelis
Menten equation with one binding site and a competitive
inhibition constant (equation 2).
Equation 2: Best model describing S(ª)-warfarin met-
abolism by human liver microsomes
V
max¿[S]
vΩ
KM¿(1π
[I]
)
Ki,c
for explanation of symbols see legend to equation 1.
In vitro inhibition of S(ª)-warfarin metabolism by silibinin
clearly followed a competitive mechanism of inhibition, and
the extent of inhibition was characterised by competitive
inhibition constants Ki,c of 18 and 19 mM, respectively, for
the two microsomes samples tested (table 3).
Denitronifedipine turn-over rates in both microsomal
samples was the reaction most affected by silibinin (fig. 4a
and b) and could be fitted best to a mixed-type Michaelis
Menten equation, including one binding site with both a
competitive and a non-competitive inhibition constant Ki,c
and Ki,n, and a second linear term, a¿[S], accounting for
metabolite formation by a process that was not saturated
at the substrate concentrations tested (equation 3):
Fig. 3. Linear plots: inhibitory effect of silibinin on S(ª)-warfarin
7-hydroxylation a: P13 liver microsomes; b: HO12 liver microsomes.
For characterisation of microsome samples, see Fuhr et al. (1996b).
Lines were fitted to data according to equation 2.
The main mechanism by which silibinin affected deni-
tronifedipine metabolism was non-competitive inhibition
with Ki,c values of 9 and 12 mM (table 4).
Equation 3: Best model describing S(ª)-denitronifedipine
oxidation by human liver microsomes
Discussion
V
max¿[S]
vΩ
KM¿(1π
π a¿[S]
The observed metabolite formation rates were within the
range of variation observed in other studies, taking the
lower activities caused by addition of DMSO into account
(Beckmann-Knopp et al. 1999). This indicates that no poor
metaboliser of one of the genetically polymorphic enzymes
was among the donors of the samples. Metabolite concen-
trations measured after incubations sufficiently exceeded
the lower limit of quantification of the respective assays in
all cases in order to allow reliable evaluation of IC50 values
(table 1). The role of DMSO effects on enzyme activity has
been discussed (Beckmann-Knopp et al. 1999).
[I]
[I]
)π[S]¿(1π
)
Ki,c
Ki,n
a; factor for non-saturated metabolite formation; for explanation
of further symbols see legend to equation 1.
Table 4.
Enzyme kinetic parameters obtained for denitronifedipine oxi-
dation and inhibitory effect of silibinin in human liver microsomes.
Microsome
sample*
Vmax
Kmax
Ki,c
/(mM)
Ki,n
/(mM)
/(pmol/min./mg protein) /(mM)
The lack of relevant inhibitory silibinin effects on human
cytochrome CYP2E1 confirms a previous study, however
carried out with different methods (Miguez et al. 1994).
The findings obtained with dextromethorphan as a sub-
strate require some comments. The substance has two bind-
ing sites in human liver microsomes, a low and a high affin-
P10
P14
1282
619
134
81
105
104
9
12
parameters were estimated using equation 3; *for identification of
microsome samples, see (Fuhr et al. 1996b); Ki,nΩinhibition con-
stant for non-competitive inhibition; for explanation of other sym-
bols see legend to table 2.

254
SVANE BECKMANN-KNOPP ET AL.
the very complex behaviour of human CYP3A4. It appears
that for this enzyme multiple conformers with distinct sub-
strate specificities may exist (Koley et al. 1996 & 1997). Fu-
thermore, cooperativity in oxidations catalysed by the en-
zyme was described and attributed to the existence of more
than one binding site at the CYP3A4 molecule (Ueng et al.
1997). Finally, the specificity of the two substrates used with
respect to CYP3A5 and CYP3A7 is not clearly defined
(Daly et al. 1995). Both enzymes are present in human liver,
although usually at much lower levels than CYP3A4 (Tatei-
shi et al. 1999). Our results suggest that silibinin may be
useful in further characterising human CYP3A enzymes.
Indeed, structure–activity relationships of flavonoids with
respect to effects on other cytochrome P450 enzymes have
been established (Siess et al. 1995; Edenharder et al. 1997;
Lee et al. 1998), but these data have not yet been used to
obtain information on the binding site characteristics of the
enzymes.
An important question is whether the inhibitory effects
of silibinin observed in the microsomal system will become
relevant when the drug is administrated to man. For S(ª)-
warfarin 7-hxdroxylation and for denitronifedipine oxi-
dation, estimated Ki values are only one order of magnitude
above maximal plasma concentrations of 0.6 mM (Weyhen-
meyer et al. 1992). As the concentration of silibinin at the
binding site of the enzyme is not known and may well ex-
ceed plasma concentrations, it cannot be excluded that sili-
binin affects the activity of hepatic CYP2C9 and/or
CYP3A4, thereby causing pharmacokinetic drug interac-
tions with substrates for these enzymes. Accumulation of
the lipophilic silibinin in the bile with peak concentrations
of 150 mM indeed suggests that also in the hepatocytes con-
centrations relevant for enzyme activity may be reached.
Furthermore, the presence of high concentrations of silibin-
in in the gut just after oral intake and/or as the result of an
enterohepatic recirculation may affect gut wall metabolism.
CYP3A4 is a major drug metabolising enzyme in the gut
wall and makes an important contribution to first pass met-
abolism of many orally administered CYP3A4 substrates
(Fromm et al. 1996; Fuhr 1998). Inhibition of intestinal
CYP3A4 is the reason for grapefruit juice interactions,
which may cause a more than 100% increase in bioavail-
ability of drugs undergoing pronounced first-pass met-
abolism by gut wall CYP3A4 (Fuhr 1998). For the grape-
fruit juice interactions, another flavonoid, i.e. naringenin,
has been discussed as one of the causal agents (Fuhr 1998).
In conclusion, the minor inhibitory effects of silibinin in
vitro on marker substrates of CYP2E1, CYP2D6,
CYP2C19, CYP1A2 or CYP2A6 suggest that there is no
clinically relevant inhibition of these enzymes to be ex-
pected by silibinin upon its administration in man. How-
ever, from our results we cannot exclude that comedication
of silibinin with CYP2C9 or CYP3A4 substrates may result
in increased concentrations of such drugs. Clinical studies
are required to clarify this issue. On the other hand, a poss-
ible in vivo inhibition of CYP3A4 by silibinin might contrib-
ute to protection against prohepatotoxins such as aflatoxin
Fig. 4. Linear plots: Inhibitory effect of silibinin on denitronifedipin
oxidation a: P10 liver microsomes; b: P14 liver microsomes.
For characterisation of microsome samples, see Fuhr et al. (1996b).
Lines were fitted according to equation 3.
ity site (see Beckmann-Knopp et al. 1999). It appears that
silibinin had some effect on the turn-over of dextromethor-
phan in our system (fig. 2). From a formal point of view,
this effect was described best assuming a minor competitive
inhibition of CYP2D6, i.e. the high affinity site, and a po-
tent competitive inhibition of the low affinity site (see equa-
tion 1 and table 2). The physical identity of the latter is
unknown. However, as the curve fit does not fully describe
the data (fig. 2) and the net effect of silibinin on dextrome-
thorphan demethylation does not increase markedly with
higher silibinin concentrations, it appears that this result is
not very important.
Results for inhibition by silibinin were contradictory for
denitronifedipine and erythromycin, the two CYP3A4 sub-
strates tested. While the effect on erythromycin de-
methylation was not uniform and showed activation at low
and minor inhibition at high silibinin concentrations, the
effect of silibinin on oxidiation of the dihydropyridine de-
rivative was a clear and pronounced non-competitive inhi-
bition. A possible explanation for these findings comes from

SILIBININ EFFECTS ON HUMAN CYTOCHROME P-450
255
Kinirons, M. T., D. O’Shea, T. E. Downing, A. T. Fitzwilliam, L.
Jollenbeck, J. D. Groopmann, G. R. Wilkinson & A. J. J. Wood:
Absence of correlations among three putative in vivo probes of
human cytochrome P4503A activity in young healthy men. Clin.
Pharm. Ther. 1993, 54, 621–629.
B1 that are activated by this enzyme, as postulated for other
flavonoids (Guengerich & Kim 1990).
Koley, A. P., R. C. Robinson & F. K. Friedmann: Cytochrome P450
conformation and substrate interaction as probed by CO binding
kinetics. Biochimie 1996, 78, 706–713.
Koley, A. P., R. C. Robinson, A. Markowitz & F. K. Friedmann:
Drug-drug interactions: Effect of quinidine on nifedipine binding
to human cytochrome P450 3A4. Biochem. Pharmacol. 1997, 53,
455–460.
Leber, H. W., & S. Knauff: Influence of silymarin on drug metab-
olizing enzymes in rat and man. Arzneimittelforsch. 1976, 26,
1603–1605.
Lee, H., H. Yeom, Y. G. Kim, C. N. Yoon, C. Jin, J. S. Choi, B. R.
Kim & D. H. Kim: Structure-related inhibition of human hepatic
caffeine N3-demethylation by naturally occurring flavonoids.
Biochem. Pharmacol. 1998, 55, 1369–1375.
References
Albano, E., S. W. French & M. Ingelman-Sundberg: Hydroxyethyl
radicals in ethanol hepatotoxicity. Front. Biosci. 1999, 4, D533-
D540.
Beckmann-Knopp, S., S. Rietbrock, R. Weyhenmeyer, R. Böcker,
K. T. Beckurts, W. Lang & U. Fuhr: Inhibitory effects of trospi-
um chloride on cytochrome P450 enzymes in human liver micro-
somes. Pharmacology & Toxicology, 1999, 85, 299–304.
Bertz, R. J., G. R. Grannemann: Use of in vitro and in vivo data to
estimate the likelihood of metabolic pharmacokinetic interac-
tions. Clin. Pharmacokinet. 1997, 32, 210–258.
Böcker, R. H. & F. P. Guengerich: Oxidation of 4-aryl- and 4-alkyl-
substituted 2,6-dimethyl-3,5-bis(alkoxycarbonyl)-1,4-dihydropy-
ridines by human liver microsomes and immuno-chemical evi-
dence for the involvement of a form of cytochrome P-450. J.
Med. Chem. 1986, 29, 1596–1603.
Committee for Proprietary Medicinal Products: Note for Guidance
on the Investigation of Drug Interactions (CPMP/EWP/560/95).
The European Agency for the Evaluation of Medicinal Products,
Human Medicines Evaluation Unit. London, 17 December 1997.
Dai, R., K. A. Jacobson, R. C. Robinson & F. K. Friedman: Differ-
ential effects of flavonoids on testosterone-metabolizing cyto-
chrome P450s. Life Sci. 1997, 61, PL75-PL80.
Leng-Peschlow, E. & A. Strenge-Hesse: Die Mariendistel (Sylibum
marianum) und Silymarin als Lebertherapeutikum. Z. Phytother.
1991, 12, 162–174.
Letteron, P., G. Labbe, C. Degott, A. Berson, B. Fromenty, M. De-
laforge, D. Larrey & D. Pessayre: Mechanism for the protective
effects of silymarin against carbon tetrachloride-induced lipid
peroxidation and hepatotoxicity in mice. Evidence that silymarin
acts both as an inhibitor of metabolic activation and as a chain-
breaking antioxidant. Biochem. Pharmacol. 1990, 39, 2027–2034.
Lorenz, D., W. H. Mennicke & W. Behrendt: Untersuchungen zur
Elimination von Silymarin bei cholezystektomierten Patienten.
2. Mitteilung: Biliäre Elimination nach mehrfacher oraler Gabe.
Planta Med. 1982, 45, 216–223.
Machicao, F. & J. Sonnenbichler: Mechanism of the stimulation of
RNA synthesis in rat liver nuclei by silybin. Hoppe Seylers Z.
Physiol. Chem. 1977, 358, 141–147.
Miguez, M. P., I. Anundi, L. A. Sainz-Pardo & K. O. Lindros:
Hepatoprotective mechanism of silymarin: no evidence for in-
volvement of cytochrome P450 2E1. Chem. Biol. Interact. 1994,
91, 51–63.
Mourelle, M., P. Muriel, L. Favari & T. Franco: Prevention of CCl4-
induced liver cirrhosis by silymarin. Fundam. Clin. Pharmacol.
1989, 3, 183–191.
Münter, K., D. Mayer & H. Faulstich: Characterization of a trans-
porting system in rat hepatocytes. Studies with competitive and
non-competitive inhibitors of phalloidin transport. Biochim. Bi-
ophys. Acta 1986, 860, 91–98.
Muriel, P., T. Garciapina, V. Perez-Alvarez & M. Mourelle: Silymar-
in protects against paracetamol-induced lipid peroxidation and
liver damage. J. Appl. Toxicol. 1992, 12, 439–442.
Daly, A. K.: Molecular basis of polymorphic drug metabolism. J.
Mol. Med. 1995, 73, 539–553.
Dehmlow, C., J. Erhard & H. de Groot: Inhibition of Kupffer cell
functions as an explanation for the hepatoprotective properties
of silibinin. Hepatology 1996, 23, 749–754.
Down, W. H., L. F. Chasseaud & R. K. Grundy: Effect of silybin
on the hepatic microsomal drug-metabolising enzyme system in
the rat. Arzneimittelforsch. 1974, 24, 1986–1988.
Edenharder, R., R. Rauscher & K. L. Platt: The inhibition by fla-
vonoids of 2-amino-3-methylimidazo[4,5-f]quinoline metabolic
activation to a mutagen: a structure-activity relationship study.
Mutat. Res. 1997, 379, 21–32.
Flora, K., M. Hahn, H. Rosen & K. Benner: Milk thistle (Silybum
marianum) for the therapy of liver disease. Amer. J. Gastroenterol.
1998, 93, 139–143.
Fromm, M. F., D. Busse, H. K. Kroemer & M. Eichelbaum: Differ-
ential induction of prehepatic and hepatic metabolism of verapa-
mil by rifampicin. Hepatology 1996, 24, 796–801.
Fuhr, U. Drug interactions with grapefruit juice: Extent, probable
mechanism, and clinical relevance. Drug Safety 1998, 18, 251–
272.
Nash, T.: The colorimetric estimation of formaldehyde by means of
the Hantzsch reaction. Biochem. J. 1953, 55, 416–421.
Riley, R. J. & D. Howbrook: In vitro analysis of the activity of the
major human hepatic CYP enzyme (CYP3A4) using [N-methyl-
14C]-erythromycin. J. Pharmacol. Toxicol. Meth. 1997, 38, 189–
193.
Siegers, C. P., A. Fruhling & M. Younes: Influence of dithiocarb,
(π)-catechin and silybine on halothane hepatotoxicity in the hyp-
oxic rat model. Acta pharmacol. et toxicol. 1983, 53, 125–129.
Siess, M. H., J. Leclerc, M. C. Canivenc-Lavier, P. Rat & M. Susch-
etet: Heterogenous effects of natural flavonoids on monooxygen-
ase activities in human and rat liver microsomes. Toxicol. Appl.
Pharmacol. 1995, 130, 73–78.
Fuhr, U., M. Weiss, H. K. Kroemer, G. Neugebauer, H. Rameis,
W. Weber & B. G. Woodcock: Systematic screening for pharmac-
ocinetic interactions during drug development. Int. J. Clin. Phar-
macol. Therap. 1996a, 34, 139–151.
Fuhr, U., K. L. Rost, R. Engelhardt, M. Sachs, D. Liermann, C.
Belloc, P. Beaune, S. Janezic, D. Grant, U. A. Meyer & A. H.
Staib: Evaluation of caffeine as a test drug for CYP1A2, NAT2
and CYP2E1 phenotyping in man by in vivo versus in vitro corre-
lations. Pharmacogenetics 1996b, 6, 159–176.
Guengerich, F. P. & D.-H. Kim: In vitro inhibition of dihydropyri-
dine oxidation and aflatoxin B1 activation in human liver micro-
somes by naringenin and other flavonoids. Carcinogenesis 1990,
11, 2275–2279.
Tateishi, T., M. Watanabe, H. Moriya, S. Yamaguchi, T. Sato & S.
Kobayashi: No ethnic difference between Caucasian and Ja-
panese hepatic samples in the expression frequency of CYP3A5
and CYP3A7 proteins. Biochem. Pharmacol. 1999, 57, 935–939.
Ubeda, A., M. L. Esteve, M. J. Alcaraz, K. H. Cheeseman & T. F.
Slater: Effects of flavonoids on cytochrome P-450 from rat liver
microsomes: Inhibition of enzyme activities and protection
Guengerich, F. P. & T. Shimada: Oxidation of toxic and carcino-
genic chemicals by human cytochrome P-450 enzymes. Chem.
Res. Toxicol. 1991, 4, 391–407.
Kharasch, E. D., D. Hankins, D. Mautz & K. E. Thummel: Identi-
fication of the enzyme responsible for oxidative halothane met-
abolism: implications for prevention of halothane hepatitis. Lan-
cet 1996, 347, 1367–1371.

256
SVANE BECKMANN-KNOPP ET AL.
against peroxidative damage. Phytotherapy Research 1995, 9,
416–420.
Wang, R. W., D. J. Newton, T. D. Scheri & A. Y. H. Lu: Human
cytochrome P450 3A4-catalyzed testosterone 6-beta-hydroxyl-
ation and erythromycin N-demethylation. Competition during
catalysis. Drug Metab. Dispos. 1997, 25, 502–507.
Weyhenmeyer, R., H. Mascher & J. Birkmayer: Study on dose-lin-
earity of the pharmacokinetics of silibinin diastereomers using a
new stereospecific assay. Int. J. Clin. Pharmacol. Ther. Toxicol.
1992, 30, 134–138.
Ueng, Y.-F., T. Kuwabara, Y.-J. Chun & F. P. Guengerich: Cooper-
ativity in oxidations catalyzed by cytochrome P450 3A4. Bio-
chemistry 1997, 36, 370–381.
Valenzuela, A. & R. Guerra: Protective effect of the flavonoid silyb-
in dihemisuccinate on the toxicity of phenylhydrazine on rat liver.
FEBS 1985, 181, 291–294.
Valenzuela, A., J. C. Bustamante, C. Videla & R. Guerra: Effect of
silybin dihemisuccinate on the ethanol metabolizing systems of
the rat liver. Cell Biochem. Func. 1989, 7, 173–178.