High-resolution mass spectrometric investigation of the phase i and II metabolites of finasteride in pig plasma, urine and bile

Article abstract of DOI:10.3109/00498254.2013.866298

The metabolite profile of the 5α-reductase type II inhibitor finasteride has been studied in pig plasma, urine and bile using high-resolution mass spectrometry. The porcine biotransformation products were compared to those formed by human liver microsomes and to literature data of recently identified human in vivo metabolites. The objective of this study was to gain further evidence for the validity of using pigs for advanced, invasive drug-drug interaction studies that are not possible to perform in humans.The use of high-resolution mass spectrometry with accurate mass measurements enabled identification of the metabolites by calculation of their elemental compositions as well as their fragmentation patterns.There was an excellent match between the porcine and human metabolic profiles, corroborating the pig as a model of human drug metabolism. The glucuronides of the two recently described human hydroxylated metabolites MX and MY and the carboxylated metabolite M3 were identified as the major biotransformation products of finasteride in pig urine and bile.Furthermore, the CYP enzymes involved in the formation of the hydroxylated metabolites were characterized. Human recombinant CYP3A4 could produce the two major hydroxylated metabolites MX and MY, whereas human recombinant CYP2D6 formed MY only.

Full text of DOI:10.3109/00498254.2013.866298

http://informahealthcare.com/xen
ISSN: 0049-8254 (print), 1366-5928 (electronic)
Xenobiotica, 2014; 44(6): 498–510
2014 Informa UK Ltd. DOI: 10.3109/00498254.2013.866298
!
RESEARCH ARTICLE
High-resolution mass spectrometric investigation of the phase I and II
metabolites of finasteride in pig plasma, urine and bile
1
Anna Lundahl *, Annica Tevell Aberg2,3, Ulf Bondesson2,3, Hans Lennerna¨s1, and Mikael Hedeland2,3
˚
2
1Department of Pharmacy, Uppsala University, Uppsala, Sweden, Department of Chemistry, Environment, and Feed Hygiene, National Veterinary
3
Institute (SVA), Uppsala, Sweden, and Division of Analytical Pharmaceutical Chemistry, Uppsala University, Sweden
Abstract
Keywords
1. The metabolite profile of the 5a-reductase type II inhibitor finasteride has been studied in
pig plasma, urine and bile using high-resolution mass spectrometry. The porcine
biotransformation products were compared to those formed by human liver microsomes
and to literature data of recently identified human in vivo metabolites. The objective of this
study was to gain further evidence for the validity of using pigs for advanced, invasive
drug–drug interaction studies that are not possible to perform in humans.
2. The use of high-resolution mass spectrometry with accurate mass measurements enabled
identification of the metabolites by calculation of their elemental compositions as well as
their fragmentation patterns.
Bile, CYP3A4, finasteride, metabolism, pig
History
Received 13 September 2013
Revised 11 November 2013
Accepted 12 November 2013
Published online 9 December 2013
3. There was an excellent match between the porcine and human metabolic profiles,
corroborating the pig as a model of human drug metabolism. The glucuronides of the two
recently described human hydroxylated metabolites MX and MY and the carboxylated
metabolite M3 were identified as the major biotransformation products of finasteride in pig
urine and bile.
4. Furthermore, the CYP enzymes involved in the formation of the hydroxylated metabolites
were characterized. Human recombinant CYP3A4 could produce the two major hydroxylated
metabolites MX and MY, whereas human recombinant CYP2D6 formed MY only.
Introduction
transformation from !-OH finasteride (M1, Figure 1) via
finasteride !-al (M2) to finasteride-!-oic-acid (M3; Huskey
et al., 1995). The 6a-OH finasteride (M4) was identified as a
minor metabolite in humans (Carlin et al., 1988, 1992;
Huskey et al., 1995). M1 and M3 were the major metabolites
detected in plasma and urine, respectively, while the
metabolites in feces were not structurally determined
(Carlin et al., 1992).
Finasteride, [N-(2-methyl-2-propyl)-3-oxo-4-aza-5a-androst-
1-ene-17b-carboxamide] (Figure 1), is
4-azasteroid and an inhibitor of 5a-reductase type II (Faller
et al., 1993). It is used in the oral treatment of benign prostatic
hyperplasia and androgenetic alopecia (male pattern hair loss;
Drake et al., 1999; McConnell et al., 1992).
a
synthetic
The metabolism of finasteride has been studied in humans,
rats and dogs (Carlin et al., 1987, 1992, 1997; Ishii et al.,
1992, 1994a,b). After oral administration of [14C]-finasteride
to humans, a majority of the dose was metabolized, and
39 and 57% of the dose was identified as metabolites in urine
and feces, respectively. These human data indicated that
biliary excretion was the major route of elimination for the
finasteride metabolites and this was supported by rat data
(Ishii et al., 1992). In humans, the main route of metabolism
was cytochrome P450 (CYP) 3A4 mediated sequential
In a human clinical drug–drug interaction study with
finasteride, bile was collected with a single-pass perfusion
technique (Bergman et al., 2006; Lundahl et al., 2009a,b).
M3 was found to be the major metabolite detected in urine
and low amounts of unchanged finasteride was excreted into
bile and urine (51%; Lundahl et al., 2009a). Unexpectedly,
M1 was not present in quantifiable concentrations in that
study. Instead, two other OH metabolites – denoted MX and
MY – were identified in bile and urine (Lundahl et al.,
2009b). Intact MY glucuronide was found in human bile and
urine and we had indirect evidence (b-glucuronidase
hydrolysis) to support that MX was also glucuronidated
(for structures of finasteride and the metabolites; Figure 1).
The conclusions from the previous study were that MX
probably was a ring-closed lactol form of M1 and it could not
be excluded that MY was identical to M4.
*Present address: R&D Innovative Medicines, AstraZeneca R&D
¨ ¨
Molndal, Molndal, Sweden.
Address for correspondence: Mikael Hedeland, Department of
Chemistry, Environment, and Feed Hygiene, National Veterinary
Institute (SVA), 75189 Uppsala, Sweden. Tel: þ46 18 674000. E-mail:
mikael.hedeland@sva.se
Recently we presented a drug–drug interaction study
in an advanced pig model (Bergman et al., 2009;

DOI: 10.3109/00498254.2013.866298
Identification of porcine finasteride metabolites 499
Figure 1. The chemical structures of finasteride and its metabolites.
¨
¨
Petri et al., 2006; Sjogren et al., 2012; Thorn et al., 2012)
that enabled sampling directly from the biliary duct in
which we followed the concentrations of finasteride, M1 and
M3 over time (Lundahl et al., 2011). The pharmacokinetic
profiles of these compounds were similar to those pre-
viously described in humans. However, a full comparison
between the metabolite profiles in pigs and humans was not
performed.
of analytical grade or better and used without further
purification.
In vivo samples
In vivo samples from pigs were selected from a previously
reported pig study (Lundahl et al., 2011). The dose of
finasteride was 0.8 mg/kg (directly to jejunum) and 0.2 mg/kg
(iv) and samples were collected as described in Lundahl et al.
(2011). The study was approved by the local ethics committee
for animal research, Uppsala, Sweden (registration number
C257/6).
In this study, we present for the first time the metabolite
profile of finasteride in porcine bile, urine and plasma
samples collected in an advanced pig in vivo model. The
metabolites found were compared to those previously
described in humans in order to demonstrate the validity of
using the pig model in drug metabolism studies (Lundahl
et al., 2011). High-resolution mass spectrometry (Q-TOF) was
used for structural elucidation of the phase I metabolites and
their glucuronides. The human recombinant CYP enzymes
that were involved in the formation of the hydroxylated
metabolites were also investigated.
Microsomal, HLM and recombinant CYP enzymes
incubations
Finasteride (200 mM) was incubated with HLM (1 mg/mL)
and NADPH (1 mM) in potassium phosphate buffer (0.1 M,
pH 7.4). The samples were kept at 37 ꢀC, in a shaking water
bath, during the incubations. Additional aliquots of micro-
somes (1 mg) and NADPH (2 mmol) were added continuously,
every hour or every second hour, to the incubation mixture
(total volume 2–3 ml) for the first 8 h of the experiment. The
reactions with HLM were carried out for 8 or 24 h. The turn-
over of finasteride was low in the HLM incubations and high
finasteride concentrations and the long incubation times were
needed to ensure production of relevant amounts of metab-
olites. Finasteride (100 mM) was also incubated with 50 pmol
of each of seven different types of recombinant CYP
enzymes; CYP3A4, CYP3A5, CYP2C8, CYP2C9*1
(Arg144), CYP2C18, CYP2C19, CYP2D6*1, in the presence
of NADPH (1 mM) in potassium phosphate buffer (0.1 M,
pH 7.4) for 1 h at 37 ꢀC. The shorter incubation times and
lower finasteride concentrations were found to be enough
for the qualitative assessment of the enzymes involved.
Materials and methods
Materials
Pooled HLM and human recombinant CYP Supersomes were
purchased from BD Biosciences (Franklin Lakes, NJ). The
metabolites, M1 and M3, were purchased from Toronto
Research Chemicals (North York, Canada). NADPH (reduced
tetrasodium salt) and finasteride were purchased from Sigma-
Aldrich (St. Louis, MO). b-Glucuronidase (from Escherichia
coli K12) was purchased from Roche Diagnostics GmbH
(Mannheim, Germany). Acetonitrile, formic acid and metha-
nol were purchased from Merck (Darmstadt, Germany). The
water was purified using a MilliQ water purification system
(Millipore, Bedford, MA) and all other chemicals used were

500 A. Lundahl et al.
Xenobiotica, 2014; 44(6): 498–510
The total volume of the reaction mixture was 1 ml. The
metabolic reactions were terminated by the addition of
methanol (1:1 v/v) and the samples were put on ice for
10-15 min and then centrifuged (11 500 g, 10 min). The
supernatants from the human recombinant CYP enzymes
incubations were transferred to novel tubes and the solvent
was evaporated under a stream of nitrogen while heated to
50 ꢀC in order to concentrate the analytes. The concentrated
samples were diluted with 100–150 mL of 50% methanol
before LC-MS analysis. The supernatants from the HLM
incubations were analyzed directly.
software (version 4.1 SCN 781) from Waters Corporation
(Milford, MA).
HPLC-ion trap MS
The HPLC autoinjector and binary pump [Agilent
Technologies, Waldbronn, Germany (G1312A/1100)] was
coupled to a Finnigan LCQ ion trap mass spectrometer
(Thermo Fisher Scientific, Waltham, MA). The mobile phase
consisted of 0.1% formic acid in MilliQ water (A) and
acetonitrile (B), and was delivered at a flow rate of
200 mL/min as a 22 min gradient (from 30 to 70% B from
0 to 12 min; then back to 30% B in 0.5 min; then equilibration at
30% B). The column used for the chromatographic separation
Sample treatment before analysis with
UPLC-Q-TOF MS
˚
was a Luna 5 m C18, 100 A (50 ꢁ 2 mm) coupled to a C18 guard
The in vitro samples were transferred to vials and analyzed
directly without any pre-treatment. The pig plasma samples
(100 mL) were vortexed with acetonitrile (100 mL) and
centrifuged (10 min at 12 000 g) to precipitate the plasma
proteins. The supernatants were transferred to vials and
analyzed. The pig urine and bile samples were analyzed both
with and without hydrolysis with b-glucuronidase. Without
hydrolysis: samples were diluted with MilliQ water (1:1, v:v),
vortexed, and centrifuged (10 min at 12 000 g). The
supernatants were transferred to vials and analyzed. With
hydrolysis: the samples (100 mL) were mixed with 100 mL of
potassium phosphate buffer (0.1 M, pH 6.05) and 5 mL
b-glucuronidase and incubated in 60 ꢀC for 2 h. The samples
were centrifuged (10 min at 12 000 g) and the supernatants
were transferred to vials and analyzed.
column (length 4 mm, i.d. 2 mm), both from PhenomenexÕ
(Torrance, CA). The mass spectrometer was equipped with an
electrospray ion source operated in the positive ion mode. The
spray voltage was set to 3 kV, the capillary temperature 160 ꢀC,
the capillary voltage 33 V, and the sheath and auxiliary nitrogen
gas flows were 80 and 1.5 arbitrary units, respectively.
Instrument control, data acquisition and data processing were
performed using the Xcalibur software version 1.3 (Thermo
Fisher Scientific, Waltham, MA).
Results
Structural elucidation of reference standards of
finasteride, M1 and M3 by high-resolution mass
spectrometry
Accurate mass measurements of finasteride and its metabol-
ites were carried out with a Q-TOF mass spectrometer in
order to obtain refined structural information as reference data
for comparison with results from the human and porcine
material (Table 1A–C). The accurate mass of the finasteride
product at m/z 317 was in good agreement with a suggested
loss of tertiary butene (–C4H8; Table 1A). However, the
product at m/z 305 did not match the radical structure
previously published by Constanzer et al. (1994). Our results
rather imply a neutral loss of C4H4O, suggesting an even-
electron product ion. In the H/D exchange experiment
described in Lundahl et al. (2009b), it was demonstrated
that the fragment corresponding to m/z 305 contained two
deuterium atoms whereas the precursor contained four. Thus,
it is likely that the loss of C4H4O occurred in the vicinity of a
nitrogen, probably in the A ring.
UPLC-Q-TOF MS
The chromatographic separation and accurate mass detection
of finasteride and its metabolites were achieved on an Acquity
UPLC system connected by an electrospray ion source to a
Synapt G2 Q-TOF mass spectrometer from Waters Corporation
(Milford, MA). The analytical column was an Acquity UPLC
BEH C18 1.7 mm (2.1 ꢁ 50 mm) and the mobile phase was
delivered at 500 mL/min as a 10 min gradient with 0.1% formic
acid in MilliQ water (A) or acetonitrile (B), respectively (5–
95% B from 0 to 7 min; hold at 95% B for 0.5 min, back to 5% B
in 0.5 min; then equilibration). The Synapt G2 was operated in
the resolution mode with positive electrospray ionization. The
capillary voltage was set to 0.75 kV in positive or 2.5 kV in
negative ionization mode. The voltages of the sample and
extraction cones were set to 40 and 4.0 V, respectively, in both
positive and negative polarity. The source and desolvation
temperatures were 120 and 450 ꢀC, respectively, and the
desolvation gas flow was set to 1000 L/h. The collision gas was
argon, delivered at a flow rate of 2 L/h. Data was collected in
the m/z 50–1200 scan range and the scan time was 0.3 s for MS/
MS and 0.15 s for MSE. The collision energy for MS/MS was
ramped from 10 to 45 V or 25 to 45 V. In MSE, the collision
energy in the low energy trace was set to 6 V and for the
high energy trace it was ramped from 15 to 35 V. The
system was mass calibrated using sodium formate (0.5 mM)
in 2-propanol:water (90:10; v/v)). Automatic lock-mass cor-
rection was applied using a solution of leucine–enkeph-
alin (2 ng/mL) in acetonitrile: 0.1% formic acid in water
(50:50; v/v). The instrument was controlled with the MassLynx
For the reference standard of the hydroxylated metabolite
M1, the loss of 17 Da yielding a product at m/z 372 was
demonstrated to be formed by a loss of ammonia rather than
by the previously suggested OH radical (c.f. Lundahl et al.,
2009b; see Table 1(B) for suggested structures). The differ-
ence between the theoretical and experimental accurate
masses for a loss of an OH radical would be ꢂ64.1 ppm as
compared to ꢂ0.185 ppm for a loss of ammonia. The
determined accurate masses of the other product ions closely
matched the ones for the previously published formulae. The
largest error (ꢂ13.5 ppm) was found for m/z 300; however this
fragment had a very low intensity.
The fragments of the deprotonated carboxylated metabolite
M3 were consistent with an initial decarboxylationyielding m/z

DOI: 10.3109/00498254.2013.866298
Identification of porcine finasteride metabolites 501

502 A. Lundahl et al.
Xenobiotica, 2014; 44(6): 498–510

DOI: 10.3109/00498254.2013.866298
Identification of porcine finasteride metabolites 503
357 followed by a further net loss of methane to m/z 341 (Table
1C). The fragment at m/z 102 matched an amino acid product
of the carboxylated tertiary butyl side chain.
High-resolution mass spectrometry for structural
elucidation of phase I metabolites of finasteride
formed by human liver microsomes
The results of the accurate mass measurements of the
metabolites denoted MX and MY formed by human liver
microsomes are shown in Table 2(A and B). For MX, the
accurate mass of the [M þ H]þ at m/z 389 was only ꢂ1.07 ppm
from the theoretical m/z of protonated OH finasteride. The
product ion at m/z 372 was demonstrated to be formed by a loss
of ammonia for this compound in the same way as for M1. The
ion corresponding to m/z 305 for finasteride was represented by
m/z 321 (loss of C4H4O, 305 þ 16) for MX. The accurate mass
of m/z 317 for MX indicated a loss of C4H8O which yielded the
same fragment as the loss of the tertiary butyl group from
protonated finasteride and the loss of tertiary butenol (C4H8O)
from protonated M1. The product ions at m/z 300 and 272 both
had accurate masses corroborating the previously suggested
structures (c.f. Lundahl et al., 2009b, Table 2A). The formation
of the fragments m/z 317, 300 and 272 from MX and the
confirmation of their elemental composition by high-reso-
lution MS/MS strongly suggested that MX has an unchanged
steroid skeleton and that the oxidation must have occurred on
the tertiary butyl side chain. Furthermore, the previous H/D
exchange experiment (Lundahl et al., 2009b) precluded the
possibility of an N-hydroxylation. The facts that all the methyl
groups of the tertiary butyl are equivalent and that MX was
obviously not identical to M1 strengthen the suggestion that
MX is a ring-closed lactol form of M1.
The determination of the accurate masses of the proto-
nated molecule of MY yielded a value that was only
0.467 ppm from the theoretical one (Table 2B). The previ-
ously suggested structure of m/z 333, i.e. the hydroxylated
correspondent of m/z 317 (317 þ 16) formed by the loss of the
tertiary butyl as C4H8 was strengthened as its accurate mass
closely matched the expected value (difference 0.247 ppm).
This fragment was unique to MY and its presence strongly
suggested hydroxylation on the steroid skeleton. Furthermore,
the results also corroborate the published structural proposals
for m/z 315, 272 and 270. For m/z 298 and 280 that were not
discussed in the previous paper, the accurate masses
demonstrated loss of ammonia from m/z 315 and loss of
water from m/z 298, respectively. The occurrence of the
specific product ions m/z 333, 315, 298 and 270 and the
supporting elemental compositions indicated that the hydrox-
ylation of MY has taken place on the steroid skeleton.
Structural elucidation of pig in vivo finasteride
metabolites with high resolution mass spectrometry
Plasma, bile and urine samples from pigs were analyzed for the
presence of phase I and II metabolites of finasteride with and
without b-glucuronidase treatment. In pig plasma, only traces
of MX and M3 were observed (results not shown). After
b-glucuronidase hydrolysis of urine and bile, three peaks with
m/z 389 occurred in the chromatogram. A minor peak at m/z
389 with a retention time and product ion spectrum

504 A. Lundahl et al.
Xenobiotica, 2014; 44(6): 498–510

DOI: 10.3109/00498254.2013.866298
Identification of porcine finasteride metabolites 505

506 A. Lundahl et al.
Xenobiotica, 2014; 44(6): 498–510
corresponding to that of the M1 reference standard was
observed in both materials (Table 1B). However, the major two
peaks at m/z 389 matched the MX and MY isomers described in
human liver microsomes in terms of retention times and
accurate mass spectra (Table 2A and B). The carboxylated
metabolite M3 was identified in both pig urine and bile based
on retention time and the accurate masses of its deprotonated
molecule and three of its product ions (m/z 357, 341 and 102;
Table 2C).
were found to be CYP2A, 2D, 2C and 3A (Achour et al.,
2011). Enterocytes, hepatocytes, liver and intestinal micro-
somes from pigs have been used to study drug metabolism
and similar activity and a corresponding rate of metabolism
have been observed for typical CYP3A substrates such as
testosterone and tacrolimus (Bader et al., 2000; Olsen et al.,
¨
1997; Skaanild & Friis, 1997; Thorn et al., 2011). Differences
in metabolism between humans and pigs have also been
observed for a few compounds, e.g. diclofenac and dextro-
¨
In the unhydrolyzed samples, two intact glucuronides, MX
glucuronide and MY glucuronide, were identified in both bile
and urine from pigs (Figure 2A–F and Table 3A and B). The
initial formation of m/z 389 represented the characteristic loss
of monodehydrogenated glucuronic acid (ꢂ176 Da) and the
further products of m/z 389 were in agreement with those of
MX and MY, respectively, m/z 321 being selective for MX
and m/z 333 for MY (c.f. Table 2A and B). The mass
accuracies were generally good, but for a few product ions
with a very low ion count the mass error exceeded 5 ppm.
metorphan (Thorn et al., 2011). Furthermore, in vivo pig
experiments clearly demonstrated that raloxifene was meta-
bolized to a high extent by intestinal glucuronidation in pigs
and the metabolite formed was the same as that formed in
¨
humans (Thorn et al., 2012).
In the study presented in this article, we qualitatively
identified phase I and phase II metabolites of finasteride in
pig plasma, urine, and bile and compared the results with
those obtained from an in vitro experiment using human liver
microsomes. An important aim of this study was to examine
the validity of the pig model in terms of the metabolite profile
of finasteride as well as to confirm or reject previously
published structural data based on results from low resolution
MS. This is, to the best of the authors’ knowledge, the first
study on the metabolism of finasteride in pigs. We used state-
of-the-art high resolution mass spectrometry [UPLC-Q-TOF-
MS(/MS)] to obtain refined information on the structures of
the metabolites found.
CYP enzymes involved in the formation of the
hydroxylated metabolites MX and MY
A qualitative approach was applied to achieve more know-
ledge about the CYP enzymes involved in the formation of
MX and MY. Finasteride incubations with human recombin-
ant enzymes CYP3A4, CYP2D6, CYP3A5, CYP2C8,
CYP2C9, CYP2C18 and CYP2C19 were performed and the
samples were analyzed. The CYP3A4 enzymes resulted in the
production of MX and MY (Figure 3A–C), while incubations
with human recombinant CYP2D6 enzymes resulted only in
the production of a metabolite with similar retention time and
product ion spectrum as MY (Figure 3D, E). None of the
other enzymes tested produced hydroxylated finasteride
metabolites in detectable concentrations.
In our study from 2009, the structures of MX and MY were
elucidated with a combination of the low resolution tech-
niques tandem quadrupole MS/MS and ion trap MSn. From
the MS analyses, it was suggested that MX was a ring-closed
lactol form of M1 and that MY could be identical to the
previously described 6a-OH finasteride, M4 (Lundahl et al.,
2009b). In this study, we had the possibility to use a more
advanced Q-TOF mass spectrometer giving accurate mass
data which can be used for calculation of elemental compos-
itions of the intact compounds and their fragments.
Interestingly, the metabolite profile of the pigs were in good
agreement with that of the latest in vivo human study
(Lundahl et al., 2009b) as well as with that of human liver
microsomes. MX, MY and their intact glucuronides as well
as M3 were identified in pig urine and bile (Figure 2 and
Tables 1B, 2A and B, 3A and B). In this study, traces of M1
were found in the pig urine and bile samples. However, this
apparent discrepancy with the previous human study
was probably owing to the fact that a more sensitive mass
spectrometer was used in this study. It might not have been
possible to separate MX from M1 with the techniques used in
the earlier studies and therefore M1 was described as the
major metabolite in plasma (Carlin et al., 1992). The
later quantitative investigations using more selective LC-
MS/MS techniques were only directed towards finasteride,
M1 and M3 (Lundahl et al., 2009a, 2011). A determination of
the concentrations of MX and MY were not possible at the
time as no reference standards of these compounds were
available.
Discussion
Two metabolites of finasteride, M1 (!-OH) and M3
(!-COOH), were in an early study described as the major
human plasma and urine biotransformation products (Carlin
et al., 1992). However, in a later study by our group,
contradictory results were found as M1 was not present in
quantifiable concentrations in any of the examined human
plasma, urine or bile samples whereas M3 was confirmed to be
the major metabolite (Lundahl et al., 2009a,b). Instead of M1,
two other hydroxylated metabolites, denoted MX and MY and
the MY glucuronide were identified in human bile and urine
(Lundahl et al., 2009b). Traces of M3 glucuronide could
also be identified in human bile. Recently, we performed a
drug–drug interaction investigation in pigs utilizing an
advanced pig model that enabled bile sampling directly from
the biliary duct (Lundahl et al., 2011). In that study, we aimed
to follow the finasteride, M1, and M3 concentrations over time.
However, the previously described main human metabolite M1
was not present in quantifiable concentrations in pig bile,
plasma or urine which was in agreement with the latter human
study (Lundahl et al., 2009a).
The high-resolution data obtained in this study corrobo-
rated most of the structural suggestions of the fragments that
were previously obtained by our group with low resolution
MS (Lundahl et al., 2009b). One important exception, though,
Previous knowledge from pigs suggests that the metabol-
ism in this species has similarities to that of humans. In a
recent report, the most abundant CYP subfamilies in pigs

DOI: 10.3109/00498254.2013.866298
Identification of porcine finasteride metabolites 507
Figure 2. Chromatograms and spectra from an unhydrolyzed pig bile sample collected 120 min after finasteride dosage. MS/MS of m/z 565 with
collision energy ramp from 35 to 75 eV, total ion chromatogram (A); extracted ions chromatograms for m/z 272.2 (B); m/z 321.3 (C); and m/z 333.2 (D);
spectra from the peaks at 2.33 min corresponding to MX glucuronide (E); and 2.82 min corresponding to MY glucuronide (F).

508 A. Lundahl et al.
Xenobiotica, 2014; 44(6): 498–510
Table 3. Theoretical and experimental accurate masses of the protonated molecules and product ions of the metabolites MX glucuronide (A) and MY
glucuronide (B) obtained from pig urine and bile.
Pig urine sample
Pig bile sample
Suggested
neutral
loss
Suggested formula
of fragment
Theoretical Experimental Difference Experimental Difference
m/z
Precursor
m/z
m/z
(ppm)
(ppm)
A. Protonated MX glucuronide
O
OH
OH
OH
HO
O
565.3130
565.3107
ꢂ4.07
565.3124
ꢂ1.06
H
N
O
O
H+
+
N
H
O
C29H45N2O9
O
OH
OH
OH
HO
O
–C4H4O
497.2868
497.2836
ꢂ6.41
497.2871
0.633
O
H
N
O
H+
+
H2N
C25H41N2O8
H
N
HO
–C6H8O6
389.2804
389.2781
ꢂ5.91
389.2801
ꢂ0.817
O
H+
+
N
H
O
C23H37N2O3
C
C
23H35N2O2
19H33N2O2
–H2O
–C4H4O
–C4H8O
–C4H11NO 300.1964
–C5H11NO2 272.2014
371.2699
321.2542
317.2229
371.2609
321.2512
317.2220
300.1967
272.2006
ꢂ24.2
ꢂ9.34
ꢂ2.84
0.999
ꢂ2.94
371.2704
321.2544
317.2242
300.1983
272.2016
1.47
0.612
4.09
6.48
a
C19H29N2O2
a
C
C
19H26NO2
18H26NOa
0.591
B. Protonated MY glucuronide
H
N
O
565.3130
565.3152
3.89
565.3110
ꢂ3.54
H+
+
O
N
H
OH
O
OH
OH
O
HO
O
C29H45N2O9
(continued )

DOI: 10.3109/00498254.2013.866298
Identification of porcine finasteride metabolites 509
Pig urine sample
Pig bile sample
Suggested
neutral
loss
Suggested formula
of fragment
Theoretical Experimental Difference Experimental Difference
(ppm)
Precursor
m/z
m/z
(ppm)
m/z
O
NH2
–C4H8
509.2494
509.2475
ꢂ3.73
509.2502
1.57
H+
+
N
H
O
OH
O
OH
OH
O
HO
O
C25H37N2O9
H
N
O
–C6H8O6
389.2804
389.2813
2.27
389.2800
ꢂ1.07
H+
+
N
H
O
OH
C23H37N2O3
C23H35N2O2
–H2O
–C4H8
–H2O
–NH3
–H2O
371.2699
333.2178
315.2073
298.1807
280.1701
371.2699
333.2173
315.2079
298.1806
280.1713
272.2014
270.1859
0.126
ꢂ1.55
2.05
371.2695
333.2158
315.2091
298.1816
280.1701
272.2013
270.1861
ꢂ0.952
ꢂ6.06
5.86
b
C
19H29N2O3
19H27N2O2
19H24NO2
C
C
ꢂ0.349
4.28
3.01
C
19H22NO
18H26NOb
18H24NOb
0.00
C
–C5H11NO2 272.2014
–C5H11NO 270.1858
ꢂ0.143
0.370
ꢂ0.511
1.11
C
CH3
–C7H9NO 147.1174
147.1172
ꢂ1.359
147.1173
ꢂ0.68
C+
CH3
C11H15
aFor suggested structure of fragment, see Table 2A.
bFor suggested structure of fragment, see Table 2B
was the product ion of protonated finasteride at m/z 305 for
which the previously published structure (Constanzer et al.,
1994) was demonstrated to be incorrect.
In summary, the pig formed the phase I metabolites MX
and MY and their glucuronides in agreement with the
previous human study. To the best of the authors’ knowledge,
this is the first study of the metabolite profile of finasteride in
pigs and, interestingly, the qualitative match between these
two species was excellent, demonstrating the validity of the
pig as an in vivo model for human drug studies.
Furthermore, this study gave further insight into the CYP
enzymes that are involved in the formation of MX and MY.
Human recombinant CYP3A4 could produce both these
metabolites while human recombinant CYP2D6 exclusively
produced a metabolite with similar product ion spectrum as
MY. Interestingly, CYP2D6 has not been previously men-
tioned to play a role in the metabolism of finasteride (Huskey
et al., 1995). Previously, we found that human recombinant
UGT1A4 and UGT1A3 were the major UGTs that were
involved in the formation of M1 and M3 glucuronides,
respectively (Lundahl et al., 2009b).
Conclusions
The metabolite profile of finasteride in pigs was for the first
time examined in this study. The metabolites were in good
agreement with those of humans and this demonstrates the
value of performing advanced metabolism investigations in
pigs. It is worth to highlight that the metabolites that were

510 A. Lundahl et al.
Xenobiotica, 2014; 44(6): 498–510
Bergman E, Lundahl A, Fridblom P, et al. (2009). Enterohepatic
disposition of rosuvastatin in pigs and the impact of concomitant
dosing with cyclosporine and gemfibrozil. Drug Metab Dispos 37:
2349–58.
Carlin JR, Christofalo B, Arison C, et al. (1987). Metabolism and
disposition of MK-906 in laboratory animals and man (abstract).
Pharmacologist 29:149.
Carlin JR, Christofalo P, Arison BH, et al. (1997). Disposition
and metabolism of finasteride in dogs. Drug Metab Dispos 25:100–9.
Carlin JR, Christofalo P, Vandenheuvel WJ. (1988). High-performance
liquid chromatographic determination of N-(2-methyl-2-propyl)-
3-oxo-4-aza-5 alpha-androst-1-ene-17 beta-carboxamide a 4-azaster-
oid in human plasma from a phase I study. J Chromatogr 427:79–91.
Carlin JR, Hoglund P, Eriksson LO, et al. (1992). Disposition and
pharmacokinetics of [14C]finasteride after oral administration in
humans. Drug Metab Dispos 20:148–55.
Constanzer ML, Chavez CM, Matuszewski BK. (1994). Picogram
determination of finasteride in human plasma and semen by high-
performance liquid chromatography with atmospheric-pressure chem-
ical-ionization tandem mass spectrometry. J Chromatogr B Biomed
Appl 658:281–7.
Drake L, Hordinsky M, Fiedler V, et al. (1999). The effects of finasteride
on scalp skin and serum androgen levels in men with androgenetic
alopecia. J Am Acad Dermatol 41:550–4.
Faller B, Farley D, Nick H. (1993). Finasteride: a slow-binding 5 alpha-
reductase inhibitor. Biochemistry 32:5705–10.
Huskey SW, Dean DC, Miller RR, et al. (1995). Identification of human
cytochrome P450 isozymes responsible for the in vitro oxidative metab-
olism of finasteride. Drug Metab Dispos 23:1126–35.
Ishii Y, Mukoyama H, Hata S. (1994a). Metabolism of finasteride in rat
hepatic microsomes: age and sex differences and effects of P450
inducers. Xenobiotica 24:863–72.
Ishii Y, Mukoyama H, Ohtawa M. (1994b). In vitro biotransformation of
finasteride in rat hepatic microsomes, isolation and characterization of
metabolites. Drug Metab Dispos 22:79–84.
Ishii Y, Mukoyama H, Ishii M, et al. (1992). Sex difference in the
metabolism of finasteride in the rats (abstract). ISSX Proceed 2:109.
Lundahl A, Hedeland M, Bondesson U, et al. (2009a). The effect of St.
John’s wort on the pharmacokinetics metabolism and biliary excretion
of finasteride and its metabolites in healthy men. Eur J Pharm Sci 36:
433–43.
¨
Lundahl A, Hedeland M, Bondesson U, Lennernas H. (2011). In vivo
investigation in pigs of intestinal absorption hepatobilary disposition
and metabolism of the 5-alpha reductase inhibitor finasteride and the
effects of coadministered ketoconazole. Drug Metab Dispos 39:
847–57.
Figure 3. Chromatogram (A) and spectra of MX (B, peak at 10.4 min)
and MY (C, peak at 15.3 min) produced by human recombinant
CYP3A4; chromatogram (D) and spectrum (E, peak at 15.3 min) of
the metabolite produced by human recombinant CYP2D6.
¨
Lundahl A, Lennernas H, Knutson L, et al. (2009b). Identification of
finasteride metabolites in human bile and urine by high-performance
liquid chromatography/tandem mass spectrometry. Drug Metab
Dispos 37:2008–17.
McConnell JD, Wilson JD, George FW, et al. (1992). Finasteride an
inhibitor of 5 alpha-reductase suppresses prostatic dihydrotestosterone
in men with benign prostatic hyperplasia. J Clin Endocrinol Metab 74:
505–8.
Olsen AK, Hansen KT, Friis C. (1997). Pig hepatocytes as an in vitro
model to study the regulation of human CYP3A4: prediction of drug-
drug interactions with 17 alpha-ethynylestradiol. Chem Biol Interact
107:93–108.
identified previously in human bile also could be identified in
pig bile. The use of high resolution mass spectrometry
enabled refined structural information of the OH metabolites
(MX and MY) and their glucuronide conjugates. Furthermore,
it was qualitatively identified that human recombinant
CYP3A4 could produce MX and MY.
Declaration of interest
Petri N, Bergman E, Forsell P, et al. (2006). First-pass effects
of verapamil on the intestinal absorption and liver disposition
of fexofenadine in the porcine model. Drug Metab Dispos 34:
1182–9.
The authors report no declarations of interest.
References
Skaanild M, Friis C. (1997). Characterization of the P450 system in
Gottingen minipigs. Pharmacol Toxicol 80:28–33.
Achour B, Barber J, Rostami-Hodjegan A. (2011). Cytochrome P450 pig
liver pie: determination of individual cytochrome P450 isoform
contents in microsomes from two pig livers using liquid chromatog-
raphy in conjunction with mass spectrometry. Drug Metab Dispos 39:
2130–4.
Bader A, Hansen T, Kirchner G, et al. (2000). Primary porcine
enterocyte and hepatocyte cultures to study drug oxidation reactions.
Br J Pharmacol 129:331–42.
¨
¨
Sjogren E, Bredberg U, Lennernas H. (2012). The pharmacokinetics and
hepatic disposition of repaglinide in pigs: mechanistic modeling of
metabolism and transport. Mol Pharm 9:823–41.
¨
Thorn HA, Lundahl A, Schrickx JA, et al. (2011). Drug metabolism of
CYP3A4 CYP2C9 and CYP2D6 substrates in pigs and humans. Eur J
Pharm Sci 43:89–98.
¨
¨
Thorn HA, Yasin M, Dickinson PA, Lennernas H. (2012). Extensive
intestinal glucuronidation of raloxifene in vivo in pigs and impact for
oral drug delivery. Xenobiotica 42:917–28.
Bergman E, Forsell P, Tevell A, et al. (2006). Biliary secretion of
rosuvastatin and bile acids in humans during the absorption phase. Eur
J Pharm Sci 29:205–14.