Identification of NTBC metabolites in urine from patients with hereditary tyrosinemia type 1 using two different mass spectrometric platforms: Triple stage quadrupole and LTQ-Orbitrap

Article abstract of DOI:10.1002/rcm.4451

The objective of our work was to identify known and unknown metabolites of the drug NTBC (2-(2-nitro-4-trifluoromethylbenzoyl)-1,3-cyclohexanedione) in urine from patients during the treatment of hereditary tyrosinemia type 1 (HT-1) disease, a severe inborn error of tyrosine metabolism. Two different mass spectrometric techniques, a triple stage quadrupole and an LTQ-Orbitrap (Fourier transform mass spectrometry (FTMS)), were used for the identification and the structural elucidation of the detected metabolites. Initially, the mass spectrometric (MS) approach consisted of the precursor ion scan detection of the selected product ions, followed by the corresponding collision-induced dissociation (CID) fragmentation analysis (MS²) for the targeted selected reaction monitoring (SRM) mode. Subsequently, accurate and high-resolution full scan and MS/MS measurements were performed on the possible metabolites using the LTQ-Orbitrap. Final confirmation of the identified metabolites was achieved by measuring commercially supplied or laboratory-synthesized standards. Altogether six metabolites, including NTBC itself, were extracted, detected and identified. In addition, two new NTBC metabolites were unambiguously identified as amino acid conjugates, namely glycine-NTBC and b-alanine-NTBC. These identifications were based on their characteristics of chromatographic retention times, protonated molecular ions, elemental compositions, product ions (using CID and higher-energy C-trap dissociation (HCD) techniques) and synthesized references. The applied MS strategy, based on two different MS platforms (LC/MS/MS and FTMS), allowed the rapid identification analysis of the drug metabolites from human extracts and could be used for pharmaceutical research and drug development.

Full text of DOI:10.1002/rcm.4451

RAPID COMMUNICATIONS IN MASS SPECTROMETRY
Rapid Commun. Mass Spectrom. 2010; 24: 791–800
Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/rcm.4451
Identification of NTBC metabolites in urine from
patients with hereditary tyrosinemia type 1 using two
different mass spectrometric platforms: triple stage
quadrupole and LTQ-Orbitrap
1
Diran Herebian , Marc Lamshoft , Ertan Mayatepek and Ute Spiekerkoetter
2
1
1
*
¨
1Department of General Pediatrics, University Children’s Hospital, Heinrich-Heine University, Du¨sseldorf, Germany
2Institute of Environmental Research, University of Dortmund, Germany
Received 5 December 2009; Revised 10 January 2010; Accepted 11 January 2010
The objective of our work was to identify known and unknown metabolites of the drug NTBC (2-(2-
nitro-4-trifluoromethylbenzoyl)-1,3-cyclohexanedione) in urine from patients during the treatment of
hereditary tyrosinemia type 1 (HT-1) disease, a severe inborn error of tyrosine metabolism. Two
different mass spectrometric techniques, a triple stage quadrupole and an LTQ-Orbitrap (Fourier
transform mass spectrometry (FTMS)), were used for the identification and the structural elucidation
of the detected metabolites. Initially, the mass spectrometric (MS) approach consisted of the
precursor ion scan detection of the selected product ions, followed by the corresponding col-
lision-induced dissociation (CID) fragmentation analysis (MS2) for the targeted selected reaction
monitoring (SRM) mode. Subsequently, accurate and high-resolution full scan and MS/MS measure-
ments were performed on the possible metabolites using the LTQ-Orbitrap. Final confirmation of the
identified metabolites was achieved by measuring commercially supplied or laboratory-synthesized
standards. Altogether six metabolites, including NTBC itself, were extracted, detected and identified.
In addition, two new NTBC metabolites were unambiguously identified as amino acid conjugates,
namely glycine-NTBC and b-alanine-NTBC. These identifications were based on their characteristics
of chromatographic retention times, protonated molecular ions, elemental compositions, product
ions (using CID and higher-energy C-trap dissociation (HCD) techniques) and synthesized refer-
ences. The applied MS strategy, based on two different MS platforms (LC/MS/MS and FTMS),
allowed the rapid identification analysis of the drug metabolites from human extracts and could be
used for pharmaceutical research and drug development. Copyright # 2010 John Wiley & Sons, Ltd.
NTBC, which is administrated as Orfadin1 (Swedish Orphan
Int. AB), is a specific inhibitor of the enzyme 4-hydroxyphe-
nylpyruvate dioxygenase and still the unique efficient drug for
a successful treatment of hereditary tyrosinemia type 1 (HT-1)
in combination with a tyrosine- and phenylalanine-restricted
diet.1 This disease has an incidence of approximately 1/100 000
neonates in Europe and the United States and is the most severe
of the tyrosine catabolic pathways. HT-1 is due to inherited
deficiency of the enzyme fumarylacetoacetate hydrolase that
catalyzes the cleavage of fumarylacetoacetate to acetoacetate
and fumarate. This enzyme defect results in an accumulation of
toxic metabolites, which are formed upstream of the enzyme
block. Accumulation of such highly reactive electrophilic
metabolites (fumarylacetoacetate or succinylacetoacetate) leads
to severe liver and kidney disease.2 Additionally, the generated
diketone compound succinylacetone (4,6-dioxoheptanoic acid),
a reactive derivative of fumarylacetoacetate, is considered as a
potent inhibitor of d-aminolevulinic acid dehydratase, the
second enzyme in the heme biosynthesis pathway. Increasing
levels of d-aminolevulinic acid, a well-known neurotoxin, lead
to neurological dysfunctions in patients with HT-1.3 Before the
introduction of NTBC in therapy of this fatal inborn error of
metabolism, liver transplantation was the only therapeutic
alternative for children affected with HT-1. NTBC has also been
used successfully for the treatment of the disease alkaptonuria,
which results from the deficiency of the enzyme homogentisate
1,2-dioxygenase in the tyrosine catabolism pathway.4
Recently, we have demonstrated a sensitive and selective
liquid chromatography/tandem mass spectrometry (LC/
MS/MS) method for the determination of the NTBC content
in plasma of HT-1 patients for optimal individual dose
adaptation, drug response and pharmaco-/toxicokinetic
studies.5 An interpatient variability in drug metabolism
could clearly be observed confirming the necessity of drug
monitoring application during therapy. NTBC is registered
as an orphan drug and knowledge of its metabolites present
in urine is important to provide a better understanding of its
metabolic fate. Generally, once a drug is administrated to an
*Correspondence to: D. Herebian, Department of General
Pediatrics, University Children’s Hospital, Heinrich-Heine Uni-
versity, Dusseldorf, Germany.
¨
E-mail: diran.herebian@med.uni-duesseldorf.de
Copyright # 2010 John Wiley & Sons, Ltd.

792 D. Herebian et al.
organism, it is absorbed and distributed to its site of action,
where it interacts with targets (e.g., receptors and enzymes),
undergoes metabolism, and finally is excreted. Pathways of
drug metabolism are classified as either phase I reactions
(e.g., hydrolysis, oxidation, and reduction) or phase II,
conjugation reactions (e.g., methylation, hydroxylation,
sulfation, and glucuronidation). Hence, both types of
reaction convert relatively lipid-soluble drugs into relatively
more water-soluble metabolites.6
extracted twice with 10 mL ethyl acetate. The organic layer
was dried over Na2SO4, evaporated and the residue was
dissolved in 300 mL of acetonitrile/water (60:40, v/v). Urine
samples were collected from HT-1 patients that had been
treated with NTBC (n ¼ 5) and stored immediately at ꢀ808C
until extraction and analysis.
Appropriate standards were synthesized by mixing
NTBC (0.1 M) with the corresponding amino acids (glycine,
b-alanine or alanine; 0.1 M) in water/acetonitrile solution
(2 mL; 50:50) containing 200 mL of 2 N NaOH. The reaction
was stirred gently at room temperature for 1 day and left
under the same conditions for a further 3 weeks. The
produced minor products of the desired standards were
subjected to LC/MS/MS analysis.
One- or two-dimensional nuclear magnetic resonance
(NMR) spectroscopy is for many drug metabolism scientists
the method of choice for structural elucidation of drug
metabolites. These experiments can be performed on a
relatively small quantity of purified material (approx. 5–
10 mg). In the majority of cases, labor-intensive purification
of adequate amounts of metabolites for NMR experiments
from complex biological matrices is not feasible due to their
very low abundance. In addition, mass spectrometers
became routine tools for pharmaceutical metabolite identi-
fication due to their excellent selectivity and sensitivity in the
nano- to picomolar range. These days, more sophisticated
MS scan techniques, such as precursor ion scan, product ion
scan, neutral loss scan, accurate mass measurement, and
multistage MSn, are available facilitating identification of
drug metabolites in complex biological matrices such as
plasma or urine.7 Furthermore, accurate mass MS/MS data
using different fragmentation sources such as collision-
induced dissociation (CID) and higher-energy C-trap
dissociation (HCD) deliver tremendous structural infor-
mation for identification and confirmation of existing and
novel drug metabolites.8
Conditions for LC/MS/MS
LC/MS/MS analyses of the samples were performed using a
Waters Quattro micro triple quadrupole system equipped
with an electrospray ionization (ESI) probe and a high-
performance liquid chromatography (HPLC) system (Waters
2795 Alliance). Chromatographic separation was performed at
room temperature with a Gemini-NX C18 Phenomenex
column (150 mm ꢁ 2 mm ꢁ 3 mm) attached to a Phenomenex
C18 Gemini-NX guard column and ran isocratically for 7 min
with 40% A and 60% B. Eluent A consisted of water/formic
acid (0.1%)/TFA (0.01%) and eluent B of acetonitrile. The
solvent flow rate was 0.2 mL/min and the injection volume
10 mL. MS/MS analysis was performed in selected reaction
monitoring (SRM) mode and positive or negative ionization
mode was used. The following mass transitions for NTBC m/z
330 ! 218 and m/z 330 ! 126 were used. The CID-MS/MS
spectra were recorded in positive ion mode. In the case of
precursor ion scan mode we used the product ions of NTBC at
m/z 218 and 126. LC/MS/MS instrument settings were as
follows for ESI-pos: capillary, 3200 V; source temperature,
1208C; desolvation temperature, 3508C; cone gas flow, 80 L/h;
desolvation gas flow, 660 L/h; cone voltage, 22; CID, 20/35 eV;
dwell time, 100 ms. Settings for ESI-neg were: capillary, 3000 V;
source temperature, 1208C; desolvation temperature, 3508C;
cone gas flow, 60 L/h; desolvation gas flow, 660 L/h; cone
voltage, 20; CID, 16/25 eV; dwell time, 100 ms.
So far, we developed a MS-based strategy for direct
detection and structural elucidation of known and novel
NTBC metabolites in patient urine. The key to structure
identification approaches is based on the fact that metabolites
generally retain most of the core structure of the parent drug, in
this case ‘the substituted aromatic ring of NTBC’. In this
paper, we used two different mass spectrometric platforms
with different types of fragmentation cells, the triple stage
quadrupole and the LTQ-Orbitrap. In the literature, three
NTBC metabolites, including NTBC itself, were described
and identified in HT-1 urine samples by means of 1H- and 19
F-NMR techniques.9 In this work, two further novel metab-
olites were found and unambiguously identified and their
structural assignment was ascertained through low abundance
standards which were synthesized in our laboratory.
The software MassLynx version 4.0 (Micromass, Waters)
was used to accomplish data acquisition and analysis as well
as to control the mass spectrometer and all peripheral
components. Peak integration and calculation were per-
formed using the software QuanLynx 4.0.
Conditions for the LTQ-Orbitrap
EXPERIMENTAL
High-resolution full scan and MS/MS spectra were obtained
with the LTQ-Orbitrap mass spectrometer (Thermo Scientific,
Waltham, MA, USA). The mass spectrometer was operated in
positive ion mode (1 spectrum sꢀ1; mass range: 90–340 m/z
for MS/MS and 250–600 m/z for full scan) with a nominal
mass resolving power of 7500 (MS/MS) or 60000 (full scan) at
m/z 400 and a scan rate of 1 Hz. Automatic gain control
and an internal calibration; bis(2-ethylhexyl)phthalate: m/z
391.284286; were used to provide high-accuracy mass
measurements within 2 ppm. MS/MS experiments were
performed by CID (collision-induced dissociation, 20/35 eV)
and HCD (higher-energy C-trap dissociation, 35/50 eV) modes
Materials
Pure NTBC (Orfadin1) of pharma grade was supplied by
Swedish Orphan (Stockholm, Sweden) as solid material. LC/
MS grade acetonitrile and water were supplied from Fisher
Scientifics. Formic acid (LC/MS grade) and ethyl acetate were
obtained from Fluka and trifluoroacetic acid from Merck.
Sample preparation procedure and synthesis of
standards
Volumes of 10 mL of the collected and blank urine samples
were acidified with 6 N hydrochloric acid (pH 2) and than
Copyright # 2010 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2010; 24: 791–800
DOI: 10.1002/rcm

Identification of NTBC metabolites in urine 793
using the following parameters: spray voltage 5 kV, capillary
temperature 2608C, tube lens 70 V. In the case of negative
ionization mode an external mass calibration was used with
the following settings: spray voltage ꢀ5 kV, capillary tempera-
ture 2608C and tube lens ꢀ70 V. Nitrogen was used as sheath
gas (5 arbitrary units) and helium served as the collision gas.
The mass spectrometer was equipped with a Dionex Ultimate
3000 HPLC system (Sunnyvale, CA, USA), which contained a
pump, UV-vis detector, flow manager, and autosampler
(injection volume 0.5 mL). The measurements were conducted
by direct injection of the sample into an isocratic flow of water/
acetonitrile containing 0.1% formic acid (50:50) at a flow rate of
focused on recognition of any NTBC modifications in urine
samples using precursor ion scan technique.
In general, the identification of a certain metabolite is
based on a mass shift of its molecular weight and its product
ions from the corresponding molecular weight and product
ions of the parent ion (unchanged drug). The proposed
structure of a product ion can be confirmed by the chemical
formula obtained from the accurate mass measurement
using the LTQ-Orbitrap mass spectrometer. This mass
analyzer is capable of providing high resolution up to
100 000 at m/z 400 and a mass accuracy of ꢂ2 ppm using
internal mass calibration and lock masses.10,11 The use of
high resolution in drug discovery is primarily needed for
distinguishing isobaric product ions which result from
multiple fragmentation pathways. The chemical formula
and the unsaturation value (number of rings and double-
bond equivalents) can also be calculated from the accurate
mass of precursor and product ions which lead to correct
assignment of the metabolite structures. The strategy used in
this work for novel metabolite identification of NTBC is
illustrated in Fig. 1. The first part of this strategy was to detect
all possible precursor ions giving a same set of product ions
as the precursor ion NTBC. In the second step, full scan and
product ion mass spectra of the possible metabolites were
recorded by tandem mass spectrometry. Generation of
accurate full scan and MS/MS mass spectra (CID and
HCD) of the corresponding metabolites by means of the
LTQ-Orbitrap was the third part of the strategy. While in the
fourth and last steps the identification of the metabolites was
ascertained by a chemically, even though in a low amount,
synthesized or commercially supplied standard.
4 mL minꢀ1
.
RESULTS AND DISCUSSION
In a previous study we developed a rapid, sensitive and
selective LC/MS/MS method for the determination of the
NTBC content in plasma of HT-1 patients.5 For this method
the same isocratic elution mode was used consisting of 60%
acetonitrile and 40% water (0.1% formic acid, 0.01% TFA).
The total chromatographic run time on the C18 NX column
(Phenomenex) was 7 min. Mesotrione, a structurally similar
compound, was used as internal standard for quantitative
analysis based on standard calibration curves. Further LC/
MS basic criteria which concerned the method validation
such as recovery, stability, linearity, and reproducibility and
matrix effects were also fulfilled and discussed in the
mentioned reference.
Here, we present
a LC/MS-based strategy for the
identification of the NTBC drug metabolites in urine of
HT-1 patients in order to further elucidate drug metabolism.
Analysis of NTBC metabolites
NTBC-containing urine samples and blank urines were
collected, extracted, analyzed and compared by LC/MS in
positive (five metabolites) and negative (one metabolite) ion
modes. Figure 2(A) shows the precursor ion scan of the
selected product ion of NTBC at m/z 218 using positive ion
mode from a urine sample of a patient treated with NTBC.
Mass spectrometry (LC/MS/MS and LTQ-
Orbitrap)
The precursor ion scan technique is a precious and helpful
tool for the rapid confirmation of targeted compounds or for
the detection of non-targeted compounds (e.g., unidentified
metabolites) bearing common product ions. MS/MS analysis
of the precursor protonated molecule obtained from NTBC
gave, as was reported in the previous study, two specific
product ions at m/z 218 (CID 20 eV) and 126 (CID 35 eV).5 To
assign the correct molecular structure of both detected
product ions, high-resolution accurate mass MS/MS
measurements were performed in positive ion mode with
the LTQ-Orbitrap mass spectrometer. As was expected, the
first product ion at m/z 218 corresponds to the well-known
acylium cation and was formed by the loss of cyclohex-
anedione moiety (C6H8O2), with an elemental composition
of C8H3O3NFþ3 . The second product ion at m/z 126 had an
elemental composition of C6H2NFþ2 and it was assigned as an
aryne derivative with a triple bond in the aromatic cycle.
Both main product ions contain unequivocally fluorine
groups (CF3 or NF2) and hence are specific product ions for
the detection of NTBC and its existing CF3 metabolites. While
the natural content of fluorine atoms in human biological
fluids is marginally small, a detection of these ions can only
be generated with the utmost probability from NTBC
compound. In the first part of this study our interest was
Identification of the urinary metabolite NTFA
In the negative ion full scan mass spectrum, the compound
NTFA (2-nitro-4-trifluoromethylbenzoic acid) was observed
as a deprotonated molecular ion [M–H]ꢀ at m/z 234 with a
retention time of 3.4 min. The molecular structure of NTFA is
depicted in Fig. 2(B). The ESI-neg MSn (n ¼ 2–4) product ion
mass spectra of NTFA contained three characteristic product
ions at m/z 190, 160 and 132 (Supplementary Fig. S1, see
Supporting Information). The product ion at m/z 190 was
formed by the neutral loss of CO2 (44 Da) from the [M–H]ꢀ
ion. The product ion at m/z 160 was formed by the loss of
CO2þNO (74 Da) yielding an epoxide derivative with radical
properties. The product ion at m/z 132 was formed by the loss
of 102 Da, and its formation was attributed to a loss of a
further CO group, which corresponded to [M–
.
HꢀCO2ꢀNOꢀCO]ꢀ radical ion. The fragmentation path-
way occurred via a ring opening and a subsequent
intramolecular cyclization to yield the five-ring molecule
containing a double and triple bonds (aryne derivative).
Figure 3(A) illustrates the proposed molecular structures of
Copyright # 2010 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2010; 24: 791–800
DOI: 10.1002/rcm

794 D. Herebian et al.
Figure 1. Flow diagram for identification of NTBC metabolites.
the fragmentation ions of NTFA. Interestingly, the CF3 group
did not appear to give any additional fragmentation,
whatever the energy was used for CID in MS/MS
experiments. All three product ions were detected on both
instruments. However, the product ion at m/z 132 showed a
very low intensity on the LTQ-Orbitrap instrument in both
CID and HCD experiments. The elemental compositions
found for the above-mentioned product ions corresponded
to the proposed molecular structures shown in Fig. 3(A),
with a mass error <5 ppm for external calibration mode and
without using lock masses (Table 1). Hence, these three
product ions were selected for SRM (selected reaction
monitoring) mode to identify NTFA in urine extracts of
HT-1 patients. Based on the above MS/MS data and
additionally confirmed by comparison with a commercially
available standard, the metabolite NTFA was unambigu-
ously identified in the extracts. The transformation of NTBC
into NTFA was consistent with published findings on the
metabolism of mesotrione, a structurally similar compound,
in plants, which is due to a hydrolytic cleavage of the
substituted benzoyl group of the triketone compound.12
Identification of the urinary metabolite NTBC
NTBC, as depicted in Fig. 2(B), was easily identified in
positive ion mode as a protonated molecular ion at m/z 330 on
the basis of its retention time and typical fragmentations.5
MS/MS experiments gave two main product ions (m/z 218
and 126), which were observed and used for SRM mode
(Supplementary Figs. S2(A) and S2(B), see Supporting
Information). The first product ion at m/z 218 corresponded
to the expected acylium cation by a loss of C6H8O2, a
cyclohexanedione moiety. The second intense product ion at
m/z 126 represented an aryne derivative. The assignment of
the molecular structures of these product ions was described
Figure 2. (A) Precursor ion scan (ESI-pos) of the ion at m/z
218 extracted from the NTBC metabolites in HT-1 urine
samples. (B) Chemical structures of the metabolites NTBC,
NTFA, 4-OH-NTBC and 5-OH-NTBC.
Copyright # 2010 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2010; 24: 791–800
DOI: 10.1002/rcm

Identification of NTBC metabolites in urine 795
Figure 3. Proposed molecular structures for the product ions of (A) NTFA, (B) NTBC, and (C) 4-
OH-NTBC/5-OH-NTBC.
in detail in our previous work.5 However; it is still
surprisingly to find the NTBC parent drug in the urinary
extracts. This is probably due to the various polar functions
of the drug, e.g. CF3, NO2 or keto/enol moieties, which
consequently makes NTBC more hydrophilic for renal
excretion. The proposed structures for the product ions
based on accurate mass MS/MS data are illustrated in
Fig. 3(B).
Table 1. Accurate mass measurements of NTFA, 4(5)-OH-NTBC and NTBC as determined using the LTQ-Orbitrap in MS and
MS/MS mode using two different collision cells (CIDa and HCDb)
Product ion
Elemental composition
C8H3O4NF3
Calculated m/z
Measured m/z
Relative error (ppm)
1.202
DBEc
6.5
1) NTFAd
[MꢀH]ꢀ
234.00087
234.00207
CID/HCD
m/z 190
m/z 160
m/z 132
C7H3O2NF3
C7H3OF3
C6H3F3
190.01104
160.01305
132.01814
190.01154
160.01347
132.01861
2.318
2.619
3.588
5.5
5.0
4.0
2) NTBCe
[MþH]þ
C14H11O5NF3
330.05838
330.05838
ꢀ0.008
9.5
CID/HCD
m/z 218
m/z 126
C8H3O3NF3
C6H2NF2
218.00595
126.01498
218.00571
126.01490
ꢀ1.107
ꢀ0.640
6.5
5.5
3) 4-OH-NTBCe/5-OH-NTBCe
[MþH]þ
C14H11O6NF3
346.05330
346.05321
ꢀ0.267
9.5
CID/HCD
m/z 328
m/z 218
C14H9O5NF3
C8H3O3NF3
328.04273
218.00595
328.04267
218.00579
ꢀ0.188
ꢀ1.707
9.5
6.5
a CID ¼ collision-induced dissociation.
b HCD ¼ higher energy C-trap dissociation.
c DBE ¼ double-bond equivalents.
d
¼ negative ion mode and external mass calibration.
e
¼ positive ion mode and internal mass calibration.
Copyright # 2010 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2010; 24: 791–800
DOI: 10.1002/rcm

796 D. Herebian et al.
Identification of the urinary metabolites 4-OH-NTBC
and 5-OH-NTBC
were found: m/z 346 (C14H11O6NF3, ꢀ0.267 ppm) and 328
(C14H9O5NF3, ꢀ0.188 ppm). Figure 3(C) illustrates the
molecular structures of the corresponding metabolites with
their major product ions. In Table 1, calculated and observed
m/z values for OH-NTBC including their errors and
elemental compositions are presented.
In the precursor ion scan of the product ion at m/z 218
extracted from NTBC, we noted the presence of an oxidized
metabolite of the drug as major peak at m/z 346 in ESI-pos
mode. The ion was thus 16 Da higher than NTBC demon-
strating that this ion was presumably a hydroxylation
conjugate of the drug. This precursor ion (m/z 346) yielded
product ions at m/z 218 and 328 (Supplementary Fig. S2(C),
see Supporting Information). The formation of the product
ion at m/z 218 indicated that the aromatic ring remained
unchanged and the hydroxylation reaction occurred at the
cyclohexanedione ring. The presence of the hydroxylated
NTBC was consistent with previous studies and was
assigned as 4-OH-NTBC (Fig. 2(B)). Interestingly, the total
ion current (TIC) profile showed a major peak with a
retention time of 3.8 min which corresponded to the above-
mentioned 4-OH-NTBC metabolite and it was followed by a
low abundance peak at a retention time of 4.3 min having the
same mass set. The observation of this second metabolite is
indicative for a hydroxylation reaction at the C5 position of
the cyclohexanedione and this was identified as 5-OH-NTBC
(Fig. 2(B)), consistent with a previous report.13 Under acidic
and CID conditions, these precursor metabolites ions
showed the dehydrated product ions [MþH–H2O]þ at m/z
328 which were consistent with hydroxylation products
(Supplementary Fig. S2(C), see Supporting Information).
Both product ions were also observed via in-source
fragmentation (cone voltage 22 V). The presence of the
OH-NTBC metabolites was also confirmed by using the LTQ-
Orbitrap. High-resolution and accurate MS and MS/MS
measurements in ESI positive ion mode were performed. The
following elemental compositions with mass error <1 ppm
Identification of the urinary metabolite gly-NTBC
In the precursor ion scan mass spectrum of the ion at m/z 218,
a major metabolite at m/z 387 was detected in the HT-1 urine
samples (Fig. 2(A)). This ion appeared at 57 mass units higher
than the parent compound NTBC (m/z 330). A difference of
57 Da is in the majority of cases an indication for a glycine
residue, thus an amino acid conjugation. A formation of
Schiff base derivatives is obviously favourable in the case of
NTBC due to the presence of the ketone groups. This
metabolite was consequently detected with the LTQ-Orbi-
trap operated at a resolution power of 60 000. The protonated
molecular ion at m/z 387.07972 gave a molecular formula
of C16H14O6N2F3 with a mass error of ꢀ0.329 ppm and a
double-bond equivalent (DBE) of 9.5 (Table 2). Based on the
data of this MS analysis, we concluded a molecular structure
of gly-NTBC, an imine derivative, for this unknown ion. The
CID product ion mass spectrum (ESI-pos mode) at 35 V of
this metabolite showed two characteristic product ions at m/z
341 and 156, as depicted in Fig. 4(A). The first product ion at
m/z 341 corresponded to the [M–H2O–CO]þ structure and the
second at m/z 156 was attributed to a carbamic acid
derivative (see Fig. 5(A)). All other significant product ions
are listed in Supplementary Table S1 (see Supporting
Information) with their protonated molecule formulae,
calculated and measured m/z, and mass accuracy and
DBE. The HCD experiment at 50 V gave the same
Table 2. Accurate mass measurements of gly-, ala- and b-ala-NTBC as determined using the LTQ-Orbitrap in MS and MS/MS
mode (CIDa 35 V)
Product ion
Elemental composition
Calculated m/z
Measured m/z
Relative error (ppm)
DBEb
gly-NTBC
[MþH]þ
m/z 369
m/z 341
m/z 282
m/z 218
m/z 185
m/z 169
m/z 156
ala-NTBC
[MþH]þ
m/z 355
m/z 282
m/z 170
m/z 124
b-ala-NTBC
[MþH]þ
m/z 383
m/z 282
m/z 218
m/z 183
m/z 170
C16H14O6N2F3
C16H12O5N2F3
387.07961
369.06928
341.07437
282.07364
218.00595
185.06826
169.07334
156.06552
387.07972
369.06862
341.07415
282.07324
218.00571
185.06826
169.07328
156.06548
ꢀ0.329
ꢀ1.795
ꢀ0.639
ꢀ1.434
ꢀ1.107
0.004
ꢀ0.383
ꢀ0.254
9.5
10.5
9.5
8.5
6.5
4.0
4.0
3.5
C
C
15H12O4N2F3
14H11O2NF3
C8H3O3NF3
C8H11O4N
C8H11O3N
C7H10O3N
C17H16O6N2F3
401.09550
355.09002
282.07364
170.08117
124.07569
401.09497
355.08949
282.07317
170.08102
124.07569
ꢀ1.315
ꢀ1.487
ꢀ1.665
ꢀ0.881
ꢀ0.004
9.5
9.5
8.5
3.5
3.5
C16H14O4N2F3
C14H11O2NF3
C8H12O3N
C7H10ON
C17H16O6N2F3
C17H14O5N2F3
C14H11O2NF3
C8H3O3NF3
C9H13O3N
401.09550
383.08493
282.07364
218.00595
183.08899
170.08117
401.09504
383.08473
282.07323
218.00571
183.08896
170.08106
ꢀ1.140
ꢀ0.529
ꢀ1.453
ꢀ1.120
ꢀ0.190
ꢀ0.645
9.5
10.5
8.5
6.5
4.0
C8H12O3N
3.5
a CID ¼ collision-induced dissociation.
b DBE ¼ double-bond equivalents.
Copyright # 2010 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2010; 24: 791–800
DOI: 10.1002/rcm

Identification of NTBC metabolites in urine 797
Figure 4. Product ion spectra in positive ion mode of the urinary metabolite gly-NTBC at CID 35 V (a) and of the
corresponding synthesized standard at CID 35 V (B). Product ion spectra in positive ion mode of the urinary metabolite
gly-NTBC at HCD 50 V (C) and of the corresponding synthesized standard at HCD 50 V (D).
fragmentation pattern as the above-mentioned CID spectrum
(Fig. 4(C)). In addition, a comparison of both product ion
scans (CID and HCD) indicated considerable differences
with regard to the peak intensity distributions. The major
peak with highest intensity in the HCD spectrum was the
product ion at m/z 218 which corresponded to the acylium
cation which was formed by the loss of C6H11O3N (169 Da)
from the precursor ion. The proposed molecular structures of
some significant product ions are shown in Fig. 5(A). Finally,
a synthesized standard confirmed the postulated metabolite.
Both dissociation experiments (CID and HCD) were
performed with the standard compound and their product
ion scan spectra showed exactly the same fragmentation
pattern and intensity values as the product ions found in the
urine samples (Figs. 4(B) and 4(D)). Interestingly, the
protonated molecular ion (m/z 387) of the matrix-free
standard solution showed only by CID experiments (35 V)
a loss of two water molecules at m/z 369 and 351 (Fig. 4(B))
instead of one as in the real sample, while, by performing
HCD experiments (50 V), no elimination of water molecules
was observed (Fig. 4(D)). The higher concentration of the
standard in the matrix-free solution may cause this
fragmentation process. Based on these results, the unknown
metabolite at m/z 387 was identified as gly-NTBC. The
proposed molecular structures for all intense fragmentation
ions are depicted in Supplementary Fig. S3 (see Supporting
Information).
Identification of the urinary metabolite b-ala-NTBC
In the precursor ion scan mass spectrum of the ion at m/z 218, a
metabolite at m/z 401 was detected in HT-1 urine samples. This
ion appeared 71 mass units higher than the unmodified drug
NTBC (Fig. 2(A)). Accurate mass measurements gave a
molecular formula of C17H16O6N2F3 at m/z 401.09504 with a
mass error of ꢀ1.140 ppm and a DBE of 9.5. At first, this
information seemed to point to a conjugation of an amino acid
moiety, namely alanine. The corresponding product ion
spectrum under CID conditions at 35 V gave rise to major
product ions, which are listed in Table 2. As illustrated in
Fig. 6(A), a main product ion at m/z 170 was observed with the
molecular formula of C8H12O3N. Under HCD (50 V) con-
ditions, the product ion spectrum showed the same product
ions as in CID experiments but having different abundances.
The major product ion in the corresponding HCD spectrum
was at m/z 183 and yielded
a
molecular formula
of C9H13O3N (Fig. 7(A)). To confirm the chemical structure
of the postulated metabolite, which was composed of NTBC
and alanine through an iminization reaction, we have
synthesized the standard as described in the Experimental
section. The accurate mass measurement revealed that the
Copyright # 2010 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2010; 24: 791–800
DOI: 10.1002/rcm

798 D. Herebian et al.
Figure 5. Proposed chemical structures of the most dominant product ions in positive ion mode of the
urinary metabolite gly-NTBC (A), ala-NTBC (B) and b-ala-NTBC (C).
Figure 6. Product ion spectra in positive ion mode of the protonated molecule [MþH]þ obtained from the urinary
metabolite gly-NTBC at CID 35 V (A), of the synthesized standard with b-alanine at CID 35 V (B) and of the
synthesized standard with alanine at CID 35 V (C).
Copyright # 2010 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2010; 24: 791–800
DOI: 10.1002/rcm

Identification of NTBC metabolites in urine 799
Figure 7. Product ion spectra in positive ion mode of the protonated molecule [MþH]þ obtained from the
urinary metabolite gly-NTBC at HCD 50 V (A), of the synthesized standard with b-alanine at HCD 50 V (B),
and of the synthesized standard with alanine at HCD 50 V (C).
elemental composition of the molecular ion of the standard
was the same as that was found in the urine matrix with a mass
error of < ꢀ1.134 ppm. The formation of this protonated
molecule [MþH]þ corresponded to the loss of an oxygen from
the cyclohexane ring of NTBC, as a water molecule, via the
iminization step with the amino acid alanine. On the other
hand, both CID and HCD product ion spectra of the standard
showed a different fragmentation pattern than was obtained
for the postulated metabolite (Figs. 6(C) and 7(C)). In this case,
the highest abundance peak in the HCD spectrum was at m/z
124, and not at m/z 183, with the molecular formula
of C7H10ON. The retention time of the standard (3.4 min) on
the reversed-phase (RP) column was different than that which
was found in urine matrix (3.0 min). The most intense product
ion in the CID spectra (35 V) of the ala-NTBC standard was at
m/z 355 showing an elemental composition of C16H14O4N2F3
(ꢀ1.487 ppm). This ion corresponded to the elimination
of H2CO2 (46 Da) of the amino acid residue and is illustrated
in Fig. 5(B). In contrast, the CID (35 V) spectrum of the urinary
metabolite displayed a major product ion at m/z 170 with the
elemental composition of C8H12O3N. The proposed chemical
structure for this ion is given in Fig. 5(B). Based on these
findings, the molecular structure of the modified NTBC (ala-
NTBC) could not be ascertained. In the search for another
compound with similar physicochemical properties, we chose
b-alanine, which contained the same mass units differences
(71 Da) and the same elemental composition as alanine.
Indeed, after synthesizing the b-ala-NTBC standard as a Schiff
base derivative, we found that no differences were observed
either in the fragmentation pattern (CID and HCD) or in the
retention time (3.0 min) between the metabolite in urine matrix
and the standard (Figs. 6(B) and 7(B)). The proposed molecular
structures for the fragmentation ions of b-ala-NTBC are shown
in Fig. 5(C). Therefore, the unknown metabolite was
unambiguously identified as b-alanine-NTBC. All significant
MS2 spectral data of both alanine and b-alanine derivatives are
listed in Table 2. The proposed molecular structures of the
most significant individual product ions and their correspond-
ing accurate mass MS2 (CID/HCD) data are shown in
Supplementary Figs. S4 and S5 and in Supplementary Tables
S2 and S3 (see Supporting Information).
Finally, a SRM chromatogram of the detected metabolites
in positive ion mode, based on the corresponding major
product, is displayed in Fig. 8. In the case of NTFA, all three
product ions were selected in the negative ion mode for
performing qualitative and quantitative analysis.
Figure 8. SRM-ESI-pos chromatogram of the five detected
metabolites in urine of HT-1 patients.
Copyright # 2010 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2010; 24: 791–800
DOI: 10.1002/rcm

800 D. Herebian et al.
Urine blank samples
using internal lock mass calibration; while in negative ion
mode the errors were without lock mass correction <5 ppm.
The results obtained in this work represent an important step
in monitoring metabolism of NTBC in HT-1 patients.
We analyzed 11 urine blank samples from healthy persons
using the same extraction procedure and LC/MS/MS
method. As expected, no detection of the above-mentioned
metabolites was observed. So, we concluded from these
measurements that the detected metabolites could only have
originated from the metabolism of the drug NTBC in HT-1
patients.
SUPPORTING INFORMATION
Additional supporting information may be found in the
online version of this article.
CONCLUSIONS
Acknowledgements
We would like to thank the Swedish Orphan International
Two powerful mass spectrometric techniques using a triple
stage quadrupole (LC/MS/MS) and a LTQ-Orbitrap were
applied for the identification of the NTBC metabolites in
urine of HT-1 patients. The aim of this study was to identify
the known and unknown metabolites based on different MS
and MS/MS techniques in order to further elucidate the
various drug metabolisms.
AB for providing us pure NTBC standard for this work.
REFERENCES
1. Lindstedt S, Holme E, Lock EA, Hjalmarson O, Strandvik B.
Lancet 1992; 340: 813.
2. Moran GR. Arch. Biochem. Biophys. 2005; 433: 117.
3. Strife CF, Zuroweste EL, Emmett EA, Finelli VN, Petering
HG, Berry HK. J. Pediatr. 1977; 90: 400.
4. Phornphutkul C, Introne WJ, Perry MB, Bernardini I, Mur-
phey MD, Fitzpatrick DL, Anderson PD, Huizing M, Anik-
ster Y, Gerber LH, Gahl WA. N. Engl. J. Med. 2002; 347: 2111.
Therefore, NTBC and its five formed metabolites were
identified in urine of HT-1 patients. Two of these metabolites
(gly-NTBC and b-ala-NTBC) have never been described
before and were unambiguously identified by their retention
times, elemental compositions, MS/MS fragmentation (CID
and HCD) and ascertained by synthesized standards. These
two novel metabolites were identified as amino acids
conjugates, otherwise known as Schiff base derivatives.
These results indicated that the drug NTBC can be subjected
to three metabolic pathways (hydrolysis, hydroxylation and
amino acid conjugation) in vivo, which could be of primordial
importance for the further pharmacological study of dose-
dependent NTBC effectiveness. Using this MS strategy and
extraction procedure no glucuronidation conjugation of the
drug was observed. Hence, the combination of tandem mass
spectrometry and high-resolution accurate mass measure-
ment increases the confidence of the proposed structures.
The relative errors of precursor and product ions as
determined with ESI positive ion mode were below 2 ppm
5. Herebian D, Spiekerkotter U, Lamshoft M, Thimm E, Laryea
¨
¨
M, Mayatepek E. J. Chromatogr. B 2009; 877: 1453.
6. Weinshilboum R. N. Engl. J. Med. 2003; 348: 529.
7. Wang PG, Zhang J, Gage EM, Schmidt JM, Rodila RC, Ji QC,
El-Shourbagy TA. Rapid Commun. Mass Spectrom. 2006; 20:
3456.
8. Koulman A, Woffendin G, Narayana VK, Welchman H,
Crone C, Volmer DA. Rapid Commun. Mass Spectrom. 2009;
23: 1411.
9. Szczecinski P, Lamparska D, Gryff-Keller A, Gradowska W.
Acta Biochim Pol. 2008; 55: 749.
10. Herebian D, Zuhlke S, Lamshoft M, Spiteller M. J. Sep. Sci.
¨
¨
2009; 32: 939.
11. Lim H-K, Chen J, Sensenhauser C, Cook K, Subrahmanyam
V. Rapid Commun. Mass Spectrom. 2007; 21: 1821.
12. Mitchell G, Bartlett DW, Fraser TEM, Hawkes TR, Holt DC,
Townson JK, Wichert RA. Pest Manag. Sci. 2001; 57: 120.
13. Lock EA, Ellis MK, Gaskin P, Robinson M, Auton TR, Provan
WM, Smith LL, Prisbylla MP, Mutter LC, Lee DL. J. Inher.
Metab. Dis. 1998; 21: 498.
Copyright # 2010 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2010; 24: 791–800
DOI: 10.1002/rcm