Identification of ketene-reactive intermediate of erlotinib possibly responsible for inactivation of P450 enzymes

Article abstract of DOI:10.1124/dmd.117.079327

Erlotinib (ELT), a tyrosine kinase inhibitor, is widely used for the treatment of nonsmall cell lung cancer in clinic. Unfortunately, severe drug-induced liver injury and other adverse effects occurred during the treatment. Meanwhile, ELT has been reported to be a mechanism-based inactivator of cytochrome P450(CYPs) 3A4 and 3A5. The objectives of this study were to identify ketene intermediate of ELT and investigate the association of the acetylenic bioactivation with the enzyme inactivation caused by ELT. A ketene intermediate was detected in human microsomal incubations of ELT, using 4-bromobenzylamine as a trapping agent. CYPs 3A4 and 3A5 mainly contributedtothe bioactivation of ELT. Microsomal incubation study showed that the ketene intermediate covalently modified the enzyme protein at lysine residues and destroyed the structure of heme. The vinyl and ethyl analogs of ELT showed minor enzyme inhibitory effect (less than 20%), whereas ELT inactivated more than 60% of the enzyme. The present study providedanovel bioactivation pathway of ELT and facilitated the understanding of the mechanisms of ELT-induced mechanism-based enzyme inactivation and liver injury.

Full text of DOI:10.1124/dmd.117.079327

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
Title Page
Identification of ketene reactive intermediate of erlotinib
possibly responsible for inactivation of P450 enzymes
Huimin Zhao, Siyuan Li, Zixin Yang, Ying Peng, Xiaohui Chen, Jiang Zheng
School of Pharmacy (H. Z., X. C.), Shenyang Pharmaceutical University, Shenyang,
Liaoning, 11016, P. R. China; Wuya College of Innovation (S. L., Z. Y., Y. P., J. Z.),
Shenyang Pharmaceutical University, Shenyang, Liaoning, 11016, P. R. China; State
Key Laboratory of Functions and Applications of Medicinal Plants (J. Z.), Key
Laboratory of Pharmaceutics of Guizhou Province (J. Z.), Guizhou Medical
University, Guiyang, Guizhou, 550004, P. R. China.

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
Running Title Page
Running title: Bioactivation of erlotinib to ketene intermediate
Corresponding Authors:
Xiaohui Chen
School of Pharmacy, Shenyang Pharmaceutical University, 103 Wenhua Road,
Shenyang, Liaoning, 110016, P. R. China.
E-mail: cxh_syphu@hotmail.com
Tel: +86-24-43520592
Fax: +86-24-43520592
Jiang Zheng
1State Key Laboratory of Functions and Applications of Medicinal Plants, Key
Laboratory of Pharmaceutics of Guizhou Province, Guizhou Medical University,
2
Guiyang, Guizhou, 550004, P. R. China; Wuya College of Innovation, Shenyang
Pharmaceutical University, 103 Wenhua Road, Shenyang, Liaoning, 110016, P. R.
China.
E-mail: zhengneu@yahoo.com
Tel: +86-24-23986361
Fax: +86-24-23986510
The number of text pages: 17
The number of figures: 8
The number of schemes: 2
The number of references: 32
The number of words in the Abstract: 172
The number of words in the Introduction: 575
The number of words in the Discussion: 1049
Abbreviations
BBA, 4-bromobenzylamine; m-CPBA, m-chloroperbenzoic acid; CYP, cytochrome
P450; DTT, dithiothreitol; EGFR, epidermal growth factor receptor; EPI, enhanced
product ion; ELT, erlotinib; GSH, glutathione; HLMs, human liver microsomes;
HPLC-UV, high performance liquid chromatography-ultraviolet; LC-MS/MS, liquid
chromatography-tandem mass spectrometry; MBI, mechanism-based inactivation;
MRM, multiple reaction monitoring; MS, mass spectrometry; MS/MS, tandem mass
spectrometry; NADPH, β-nicotinamide adenine dinucleotide 2′-phosphate reduced
tetrasodium salt; NMR, nuclear magnetic resonance; NSCLC, non-small-cell lung
cancer; PBS, potassium phosphate buffer; Pd-BaSO4, palladium on barium sulfate;
Pd-C, palladium on activated charcoal; PI, precursor ion.
2

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
Abstract
Erlotinib (ELT), a tyrosine kinase inhibitor, is widely used for the treatment of
non-small-cell lung cancer in clinic. Unfortunately, severe drug-induced liver injury
and other adverse effects occurred during the treatment. Meanwhile, ELT has been
reported to be a mechanism-based inactivator of CYPs 3A4 and 3A5. The objectives
of this study were to identify ketene intermediate of ELT and investigate the
association of the acetylenic bioactivation with the enzyme inactivation caused by
ELT. A ketene intermediate was detected in human microsomal incubations of ELT,
using 4-bromobenzylamine as a trapping agent. CYPs 3A4 and 3A5 mainly
contributed to the bioactivation of ELT. Microsomal incubation study showed that
the ketene intermediate covalently modified the enzyme protein at lysine residues and
destroyed the structure of heme. The vinyl and ethyl analogs of ELT showed minor
enzyme inhibitory effect (less than 20%), while ELT inactivated more than 60% of the
enzyme. The present study provided a novel bioactivation pathway of ELT and
facilitated the understanding of the mechanisms of ELT-induced mechanism-based
enzyme inactivation and liver injury.
3

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
Introduction
Erlotinib (ELT, 1, Scheme 1) is a highly potent reversible human epidermal
growth factor receptor (EGFR) tyrosine kinase inhibitor (Dai et al., 2005). The
proliferation and apoptosis of tumor cells are mediated by the EGFR signaling
pathway (Scambia et al., 1995; Veale et al., 1993). Studies have shown that
non-small-cell lung cancer (NSCLC) was caused by over-expression of EGFR. (Jotte
et al., 2015; Massarelli et al., 2013). ELT (Tarceva) was first approved by the Food
and Drug Administration in 2004 for the treatment of metastatic NSCLC and
advanced pancreatic cancer.
Despite the beneficial effects of ELT, this
anti-neoplastic drug has been linked to interstitial lung disease and hepatorenal
syndrome (Makris et al., 2007; Tammaro et al., 2005). In addition, ELT has been
associated with severe hepatic injury. The patients with normal liver function may
suffer from hepatic impairment and even hepatic failure when treated with ELT (Saif,
2008; Liu et al., 2005; Huang et al., 2009). These rare but severe cases are often
described as idiosyncratic toxicity. Though the mechanisms of ELT induced hepatic
injury remain unknown, it has been proposed that covalent modifications of cellular
macro-molecules by reactive metabolites might be responsible for the idiosyncratic
adverse reactions (Stepan et al., 2011).
In humans, CYPs 3A4 and 3A5 are the critical enzymes in the liver metabolism
of ELT while CYPs 1A2 and 2C8 make a minor contribution (Ling et al., 2006). In
addition, ELT has been proven to be a mechanism-based inactivator of CYPs 3A4 and
3A5 (Li et al., 2010). Mechanism-based inactivation (MBI) is oftentimes referred to
as quasi-irreversibly or irreversibly time-dependent inhibition, and/or suicide
4

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
inactivation. Drug-induced MBI may cause more serious injuries to patients
compared to reversible enzyme inhibition. The reason is that the MBI can destroy
the structure of enzymes which would have to be regenerated, and loss of enzyme
function may raise the possibility of unwanted drug-drug interactions. In theory, MBI
arises when drugs are bioactivated by CYPs to generate reactive intermediates that
cause irreversible loss of enzyme activity through covalent bonding to apoprotein
and/or destroying structure of the heme (Orr et al., 2012; Kalgutkar et al., 2007).
As shown in Scheme 1, ELT (1) contains a phenyl-acetylene moiety.
Metabolism of ELT is mediated by three primary routes in humans: (I)
O-demethylation of the methoxy group of ELT, (II) 4-hydroxylation of the aromatic
ring of aminophenylacetylene group to p-aminophenol, and (III) hydrolysis to phenyl
acetic acid after oxidation of terminal acetylene (Ling et al., 2006). 4-Hydroxylation
of the phenyl ring of ELT results in forming para-hydroxyaniline metabolite, which
further undergoes oxidation to form a quinoneimine intermediate through loss of two
electrons by CYPs. This quinoneimine intermediate has been detected as glutathione
(GSH) conjugate by liquid chromatography/tandem mass spectrometry (Li et al.,
2010).
Although the bioactivation of ELT by CYPs to a reactive quinoneimine
metabolite has been well-defined, the possible bioactivation of this drug containing
chemical scaffold terminal alkyne to a ketene intermediate is much less characterized.
To our knowledge, the reactive oxirene and ketene intermediates can occur in the
oxidation of the acetylene bond by CYPs and are thought to play a part in CYPs
5

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
inactivation (Lin et al., 2007; von Weymarn et al., 2004; Blobaum et al., 2002). In
this study, a novel bioactivation pathway of ELT in the form of an amino-group based
4-bromobenzylamine (BBA) adduct in NADPH-dependent human liver microsomes
(HLMs) incubated with ELT was reported. Moreover, the correlation of ketene
intermediate formation and enzyme inactivation was investigated.
6

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
Materials and Methods
1. Reagents
Erlotinib (ELT, ≥99%) and chloroprotoferrihem was purchased from Dalian
Meilun Biotech Co., Ltd. (Dalian, China). 4-Bromobenzylamine (BBA) was
purchased from Shanghai Darui Chemical Co., Ltd. (Shanghai, China).
m-Chloroperbenzoic acid (m-CPBA), palladium on activated charcoal (10%Pd-C),
palladium on barium sulfate (10%Pd-BaSO4) and quinoline were purchased from
Aladdin Reagent Co., Ltd. (Shanghai, China). Pronase E, chymotrypsin, L-lysine,
testosterone, 6β-hydroxytestosterone, DL-dithiothreitol (DTT), propranolol and
reduced nicotinamide adenine dinucleotide phosphate (NADPH) were purchased from
Sigma-Aldrich Co. (St. Louis, MO). Recombinant human P450 enzymes and HLMs
were purchased from BD Gentest (Woburn, MA). Formic acid and trifluoroacetic
acid were purchased from Fisher Scientific (Springfield, NJ). All organic solvents
were obtained from Fisher Scientific (Springfield, NJ). All chemicals were of
analytical grade and all solvents were of HPLC grade.
2. Chemical Synthesis of ELT Analogs
The ethyl and vinyl analogs (Scheme 1, 2 and 3) were synthesized by catalytic
hydrogenation of ELT, according to the published literature with modification (Ban et
al., 2004). To synthesize the ethyl analog (2), ELT (100 mg, 0.26 mmol) was
dissolved in 10 mL methanol, and Pd-C (20 mg) was then added as a catalyst. The
mixture was stirred in a sealed and hydrogen-purged reactor at room temperature and
with pressure of 70 MPa for 4 h. The resulting reaction mixture was condensed
7

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
under vacuum after filtration. Silica gel column chromatography with eluate of
dichloromethane and methanol (20:1) was used to purify the product. The resulting
product was characterized by HPLC-UV, MS/MS and 1H NMR. Data and spectrums
are shown in supporting information (Supplemental Fig. 1 and 2).
To synthesize the vinyl analog (3), ELT (100 mg, 0.26 mmol) was dissolved in
10 mL methanol, and Pd-BaSO4 (20 mg) was then added as a catalyst. Aliquots (0.5
mL) of quinoline were added to quench the activity of the catalyst. The mixture was
stirred in a sealed and hydrogen-purged reactor at room temperature and with pressure
of 70 MPa for 2 h. After reaction, the mixture was subjected to a silica gel
chromatography to remove quinoline and the fraction was purified by a
semi-preparative HPLC system. The resulting product was characterized by
1
HPLC-UV, MS/MS and H NMR. Data and spectrums are shown in supporting
information (Supplemental Fig. 3 and 4).
3. Microsomal Incubations
ELT (100 μM from a dimethyl sulfoxide stock) and HLMs (1.0 mg protein/mL)
were mixed in potassium phosphate buffer (PBS, pH 7.4). BBA (5 mM) (Scheme 1,
4), GSH (20 mM) or L-lysine (40 mM) were then added as a trapping agent,
respectively. The total volume of the incubation was 500 μL. The reactions were
initiated by adding NADPH (1 mM). Control samples did not contain NADPH.
Reactions were terminated by mixing with ice-cold acetonitrile in a volume of 1:1.
The resulting mixture was centrifuged at 19,000 g for 10 min at 4 °C to pellet protein.
After the supernatants were evaporated to dryness under nitrogen stream, the
8

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
reconstituted solution dissolved in 200 μL of acetonitrile: water (1:9, v/v) was injected
into LC-MS/MS for analysis.
The ethyl and vinyl analogs were incubated in the same microsomal system as
described above to detect possible BBA and GSH adducts, respectively.
4. Recombinant Human P450 Incubations
ELT was individually incubated with human recombinant P450 enzymes (100
nM), including CYPs 1A2, 2A6, 2B6, 2C9, 2C19, 2D6, 2E1, 3A4, and 3A5. The
conditions were the same as the microsomal incubation experiments, except that
recombinant P450 enzymes replaced the microsomes. An equal volume of ice-cold
acetonitrile which contained propranolol (500 ng/mL) as internal standard was
added to quench the reactions. The experiments were performed in triplicate.
5. Chemical Synthesis of ELT-BBA and ELT-Lysine Adducts
Synthesis of ELT-BBA adduct was achieved by initial epoxidation of ELT using
m-chloroperbenzoic acid (m-CPBA), based on a reported method about peroxidation
studies of acetylenes (McDonald et al., 1964). ELT (4 mg, 10.2 μmol) was dissolved
in dichloromethane (10 mL), and m-CPBA (20 mg, 116.3 μmol) and BBA (20 mg,
107.5 μmol) were then added. After stirred for 20 h at room temperature, the
reaction mixtures were centrifuged, and the supernatants were condensed under
nitrogen stream. The resulting sample was reconstituted with acetonitrile–water (1:9)
for LC–MS/MS analysis. Synthesis of ELT-lysine adduct was conducted by the
same procedure as described above, except that BBA was replaced by lysine (15 mg,
9

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
102.7 μmol).
6. P450 3A Inactivation Studies
To determine whether terminal alkyne of ELT is responsible for P450
time-dependent inactivation, experiments were carried out in HLMs with ELT, ethyl
and vinyl analogs as substrates, respectively. Primary incubations (200 μL) included
one of the substrates (25 and 50 μM), 0.5 mg/ml HLMs, 1 mM NADPH, 3.2 mM
MgCl2, and 100 mM PBS (pH 7.4). After incubation at 0, 10, 20, and 30 min, the
primary mixtures (40 μL) were transferred to the secondary incubation mixtures (120
μL), which included 200 μM testosterone, 1 mM NADPH, 3.2 mM MgCl2, and 100
mM PBS (pH 7.4). The proportion of dimethylsulfoxide was 0.5% (v/v). An equal
volume ice-cold acetonitrile containing propranolol as the internal standard was added
to terminate the reaction after 15 min incubation. After centrifugation, the amount of
6β-hydroxytestosterone produced was measured by LC-MS/MS to determine the
remaining CYP3A activity.
7. Microsomal Protein Modification and Digestion
ELT (100 μM) was mixed with 1.0 mg/mL HLMs and NADPH (1 mM). The
final volume of incubation was 0.5 mL. After 30 min of incubation at 37 °C, the
protein samples were denatured by 30 min of heating at 60 °C in a water bath,
followed by centrifugation at 19,000 g for 10 min. Ammonium bicarbonate (50 mM,
pH 8.0) was used to suspend the pellets, which were then mixed with 5.0 mM DTT.
After incubation at 60 °C for 1 h, the resulting mixture was digested with a mixture of
10

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
Pronase E (2.5 mg/mL) and chymotrypsin (2.5 mg/mL) in the presence of CaCl2 (5.0
mM) (Wang et al., 2015). After 15 h of incubation for 37 °C, the digestion mixtures
were centrifuged at 19,000 g for 10 min and the supernatants were analyzed by
LC-MS/MS.
To further investigate lysine-based protein adducts, protein digestion was also
conducted after the ethyl or vinyl analogs was incubated with HLMs as described
above.
8. Measurement of Native Heme Loss
The determination of loss of the native heme was achieved, according to the
reported literature (Lin et al., 2002). ELT (50 μM) or ELT analogs were incubated
with 0.5 mg/ml HLMs, 1mM NADPH for 30 min at 37°C. The control samples did
not contain NADPH. The final volume of incubation mixtures were 0.5 mL. After
incubation, the mixtures were extracted with butanone containing 10% trifluoroacetic
acid. The residue was reconstituted in acetonitrile for LC-UV and LC–MS/MS
analysis after the extracted solution was evaporated to dryness under nitrogen stream.
9. Analytical Conditions
Metabolites were characterized on a SCIEX 5500 mass spectrometer (SCIEX,
Foster City, CA) equipped with an Agilent 1260 system (Agilent Technologies, Santa
Clara, CA). A ZORBAX SB-C18 column (5.0 μm, 150 mm × 4.6 mm; Agilent
Technologies, Santa Clara, CA) was used to perform HPLC separation. The mobile
phase was water and acetonitrile with 0.1% formic acid. The gradient elutions for
11

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
analyzing ELT-BBA and ELT-lysine adducts and separating 6β-hydroxytestosterone
were shown in Supplemental Table 1 and 2, respectively. All the metabolites were
analyzed in a positive mode by multiple reaction monitoring (MRM) scanning via
electrospray ionization, based on structurally specific fragmentation: m/z 595/479 and
597/481 for ELT-BBA adduct, m/z 556/428 for ELT-lysine adduct, m/z 305/269 for
6β-hydroxytestosterone and m/z 261/116 for propranolol as internal standard.
Precursor ion (PI) scanning of m/z 169, 171 were also used for analyzing the
ELT-BBA adduct. In addition, SCIEX 4000 Q-Trap (SCIEX, Foster City, CA)
equipped with an Agilent 1260 system were used to analyze the MS/MS fragments of
metabolites. Multiple reaction monitoring coupled with enhanced product ion
(MRM-EPI) mode was employed. The optimized mass spectrometric instrument
parameters are shown in supporting information (Supplemental Tab. 3). The Analyst
1.6.2 software was used to analyze all the data.
For the analysis of heme, chromatographic separation was conducted on a C4
column (250 × 4.6 mm, 5 m, VYDAC, Hesperia, CA) with solvents including 0.1%
trifluoroacetic acid in water and 0.05% trifluoroacetic acid in acetonitrile. The
gradient was shown in Supplemental Table 4. The eluates were monitored at 400 nm
and Q1MS scanning mode for the heme and possible heme adducts formed in the
bioactivation of ELT.
12

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
Results
1. Enzyme-Dependent ELT-BBAAdduction
ELT was incubated in HLMs supplemented in BBA as a trapping agent.
ELT-BBA adduct (Figure 1 for its structure) eluted at 11.6 min was detected by MRM
scanning of m/z 595/479, m/z 597/481 and PI scanning of m/z 169, m/z 171 (Fig.
1E-H). The intensity ratios of these two scan mode (MRM and PI) were around 1:1,
respectively. No peaks at the corresponding retention time were observed in the
absence of NADPH (Fig. 1A-D) demonstrating that the formation of this adduct was
NADPH-dependent. The MS/MS spectrum of the ELT-BBA adduct obtained from
MRM-EPI scanning of m/z 595/479 or 597/481 showed the characteristic fragment
ions containing bromine at m/z 169/171 (loss of C22H25N4O5), 479/481 (loss of
C6H14O2), 537/539 (loss of C3H7O) and fragment ions containing N-phenyl
quinazolin-4-amine or oxidized carbonyl moiety without bromine at m/z 266 (loss of
C14H21NO3Br) and 294 (loss of C13H21NO2Br) (Fig. 1I). No such adduct was
detected after incubation of the ethyl or vinyl analogs mixed with HLMs. This
indicated that the adduct was formed by oxidation of terminal alkyne (Fig. 6B and 6C).
To further characterize the adduct, ELT was reacted with m-CPBA and BBA. After
the reaction, a product was detected which showed the same retention time and
similar mass spectrum as that of the product formed in HLM incubations with ELT
(Fig. 2).
In a separate experiment, the ethyl and vinyl analogs were individually incubated
in the HLM incubation system supplemented with GSH as a trapping agent. For the
13

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
vinyl analog, two GSH conjugates were detected, including one derived from the
epoxide (the representative MS/MS scanning of m/z 719/590,446) (Supplemental Fig.
5A) and the other arising from the quinoneimine metabolite (MS/MS scanning of m/z
717/588,444) (Supplemental Fig. 5B). As expected, only the quinoneimine derived
GSH conjugate (MS/MS scanning of m/z 719/590,446) (Supplemental Fig. 6) was
detected in microsomal incubations with the ethyl analog.
2. P450 Enzymes Responsible for Bioactivation of ELT
Recombinant human P450 incubations were carried out for determining which
P450 enzymes contribute to the oxidation to form the ELT-BBA adduct. CYPs 3A4
and 3A5 were found to be the major enzymes in contribution to the bioactivation of
ELT. CYP3A4 displayed the highest catalytic activity, and the adduct formation by
CYP3A5 was approximately 80% of that induced by CYP3A4. CYPs 2A6 and 2D6
also slightly catalyzed the ELT-BBA formation. The levels of adduct formation by
CYPs 2A6 and 2D6 were about 10% and 13% of that formed by CYP3A4,
respectively (Fig. 3).
3. Enzyme Inactivation Effects of ELT and ELT Analogs
Microsomal incubation of ELT at concentrations of 25 and 50 μM caused
dramatic drop in CYP3A activity after 30 min by 36% (P<0.01) and 63% (P<0.001),
respectively. Unlike ELT, the ethyl analog induced a slight activity decrease by only
3% and 15% (P<0.05), respectively. The vinyl analog showed similar degree in
enzyme inactivation (approximately 9% and 20% (P<0.05), Fig. 4). The observed
14

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
significant loss of enzyme inactivation effect of ELT after saturation of the alkynyl
group to vinyl and ethyl groups suggests that the terminal alkyne of ELT was involved
in the P450 inactivation.
4. Lysine-based Protein Modification by ELT
ELT was incubated with HLMs. After incubation, the resulting microsomal
protein was exhaustively digested by proteinase and then analyzed by monitoring
MRM ion pair m/z 556/428. A lysine-derived adduct was eluted at 7.7 min (Fig. 5D).
No such adduct was found in the protein digestion of HLMs in the absence of
NADPH (Fig. 5C). The MS/MS characteristic fragment ions of amide moiety of the
adduct, such as m/z 410 and 428 were detected. Product ion m/z 511 resulting from
the loss of carboxyl group of this adduct (Fig. 5G). Fragment ions m/z 266 and 294
represented the same moieties corresponding to those of the ELT-BBA adduct (Fig. 1I
and Fig. 5G). The ethyl or vinyl analog was also incubated with HLMs as described
above. No lysine adduct was detected in protein digestion samples (Fig. 6E and 6F).
An adduct with the same retention time at 7.7 min was observed in the HLMs
incubation in the presence of NADPH and lysine (trapping agent) (Fig. 5B).
Synthetic ELT-lysine adduct showed the same chromatographic and MS
characteristics (Fig. 5E and 5F) as that of the corresponding adduct in incubation of
HLMs as well as microsomal protein digestion.
5. Native Heme Loss during Enzyme Inactivation
The native heme of standard solution eluted at approximately 11.8 min and was
15

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
monitored by UV at 400 nm (Fig. 7A) and showed the MH+ ion at m/z 616 (Fig. 7B).
There was an about 70% decrease (P<0.001) in native heme of the ELT inactivated
sample (Fig. 7C and 7D; Fig. 8). No marked decrease was observed in native heme
of both ethyl (~5% decrease) and vinyl analog (~25% decrease) incubated samples
(Fig. 7E-H; Fig. 8).
16

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
Discussion
ELT, an EGFR inhibitor used in the treatment of advanced NSCLC, was reported
to cause severe liver injury during its clinical use (Saif, 2008; Liu et al., 2005; Huang
et al., 2009). ELT can also inhibit CYP3A in a time-dependent manner. Formation
of reactive metabolites have been implicated in the mechanism-based inactivation of
enzymes and idiosyncratic toxicity. ELT was reportedly biotransformed to a
quioneimine metabolite while ketene intermediates were not detected using
N-acetyl-lysine as a trapping agent (Li et al., 2010). Despite this, we speculated that
the terminal alkyne group of ELT would be oxidized to the corresponding ketene
intermediate which may play a role in CYP3A inactivation. Ketenes are chemically
reactive and unstable, trapping techniques are in need for the metabolite identification.
Ketene intermediates reportedly modified N-terminal -amino groups of amino acid
residues and ε-amino group of lysine residues of apoproteins (Chan et al., 2012;
Banzon et al., 1995). Thus, we tested several primary amines as trapping agents, and
BBA was found to be the one which is the most effective. Additionally, BBA
contained an element bromine, and we took the advantage of its 1:1 natural isotope
pattern. This made it easier to monitor targeting analytes by mass spectrometry
(Wang et al., 2014), by which we succeeded in detecting ELT-BBA adduct in MRM
mode by monitoring of ion pairs m/z 595/479, 597/481, and m/z 169, 171 (Fig. 1).
The observed adduct clearly indicates the production of ELT-derived ketene
intermediate. Moreover, one of the three primary metabolism routes of ELT was
oxidation of terminal acetylene followed by hydrolysis to phenyl acetic acid which
17

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
was proved to be the stable metabolite of the ketene intermediate (Ling et al., 2006).
Formation of the ELT-BBA adduct was found to require the NADPH in the HLM
incubation indicating that P450 enzymes were involved in bioactivation of ELT.
To better understand the metabolic activation of ELT, the ELT-BBA adduct was
chemically synthesized by reacting ELT with m-CPBA and BBA. m-CPBA is an
oxidant agent widely used in epoxidation reactions. By analogy to the epoxidation
of alkenes, the first product of an acetylene would be an oxirene and then the reactive
ketene was generated by rearrangement of the oxirene (McDonald et al., 1964).
Unfortunately, we failed to obtain enough amount of the ELT-BBA adduct for NMR
characterization, due to the poor yield of the chemical synthesis.
MBI is often referred to as “suicide inactivation”. The cytochrome P450s might
often be targets of the reactive intermediates generated in the active site of enzyme
(Hollenberg et al., 2008).
In previous study, ELT has proved to be a
mechanism-based inactivator of CYPs 3A4 and 3A5. CYPs 3A4 and 1A1 were
reported to catalyze the formation of the quioneimine metabolite of ELT, but ELT
failed to show inhibitory effect on CYP1A1 (Li et al., 2010). The present study
demonstrated that both CYPs 3A4 and 3A5 catalyzed the formation of the ketene
intermediate of ELT. It is important to define the role of the two reactive metabolites
in the enzyme inactivation. To this end, we replaced the alkyne of ELT with ethyl
and vinyl groups and evaluated their inhibitory effects on the enzymes. Both
compounds only revealed very weak enzyme inactivation. Structurally, the vinyl
group might also be oxidized to epoxide possibly participating in protein modification,
18

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
due to its constrained bond angle (Pearce et al., 2002). To address this possibility,
the vinyl analog was incubated in HLMs supplemented with GSH as a trapping agent.
Two GSH conjugates were detected by LC-MS/MS after incubation (Supplemental
Fig. 5A and B). One resulted from the corresponding epoxide metabolite, and the
other arose from the quinoneimine intermediate (Li et al., 2010). Additionally,
similar microsomal incubations were carried out to determine the metabolic activation
of the ethyl analog. Only one GSH conjugate was detected (Supplemental Fig. 6),
and the observed molecular ion of the conjugate matched the molecular weight of the
GSH conjugate derived from the corresponding quinoneimine. The vinyl and ethyl
analogs showed little enzyme inhibitory effect (less than 20%), while ELT inactivated
more than 60% of the enzyme in a time-dependent manner. This indicates that the
formation of ketene, together with quinoneimine intermediate, contributed to CYP3A
inactivation, and the ketene intermediate played the major role.
Previous studies on the drug induced MBI have led to the suggestion that the
different mechanisms of inactivation via covalent binding to apoprotein and/or
destroying the native heme group (Ortiz de Montellano et al., 1985; Lin et al., 2010;
Blobaum, 2006). Therefore, we analyzed the products of microsomal protein
digestion samples pretreated with ELT and measured the relative amount of the native
heme loss to better understand the association between bioactivation of ELT and
enzyme inactivation. The observed decrease in the amount of the heme after
microsomal incubations suggests that the reactive intermediate of ELT would modify
the heme, at least partially responsible for the loss of the enzyme activity. Apart
19

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
from the measurement of the native heme loss, we also attempted to check for the
possible heme adducts, using HPLC-UV and MS method. No stable heme adduct
was detected. We propose that unknown heme fragment or unstable heme adducts
were formed in the inactivation reaction so that they cannot be detected. This kind
of heme destruction has also been reported by other mechanism-based inactivation
studies (Amunugama et al., 2012; Lin et al., 2013). Taken together, our findings
suggest that ELT causes inactivation through covalent binding of the bioactivated ELT
to the apoprotein of CYPs and the native heme destruction.
Previous clinical studies showed that ELT induced liver injury, which might also
be associated with the metabolic activation process. According to the tissue
distribution studies (Li et al., 2010), liver was most likely to be toxic target organ, due
to accumulated ELT. However, the definitive relationship between the reactive
metabolites (ketene and quioneimine) of ELT and liver toxicity remains unknown,
though the ketene intermediate played a major role in the enzyme inactivation.
In conclusion, a reactive ketene intermediate was detected in incubations of ELT
with HLMs. CYPs 3A4 and 3A5 mainly contributed to the acetylenic bioactivation
of ELT. The MBI of the enzymes by ELT was caused by formation of the ketene
intermediate. These data are helpful to justify possible drug-drug interactions and to
facilitate the understanding of the mechanisms of hepatotoxic action of ELT reported
in clinic.
20

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
Authorship Contributions:
Participated in research design: Zheng
Conducted experiments: Zhao, Li, and Yang
Performed data analysis: Zhao, Chen, and Peng.
Wrote or contributed to the writing of the manuscript: Zhao and Zheng.
21

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
References
Amunugama HT, Zhang H, and Hollenberg PF (2012) Mechanism-based
inactivation of cytochrome P450 2B6 by methadone through destruction of
prosthetic heme. Drug Metab Dispos 40: 1765-1770.
Ban HS, Onagi S, Uno M, Nabeyama W, and Nakamura H (2008) Allene as an
alternative functional group for drug design: effect of C-C multiple bonds
conjugated with quinazolines on the inhibition of EGFR tyrosine kinase.
ChemMedChem 3: 1094-1103.
Banzon JA, Kuo JM, Miles BW, Fischer DR, Stang PJ, and Raushel FM (1995)
Mechanism-based inactivation of phosphotriesterase by reaction of a critical
histidine with a ketene intermediate. Biochemistry 34: 743-749.
Blobaum AL (2006) Mechanism-based inactivation and reversibility: is there a
new trend in the inactivation of cytochrome P450 enzymes? Drug Metab
Dispos 34: 1-7.
Blobaum AL, Kent UM, Alworth WL, and Hollenberg PF (2002)
Mechanism-based inactivation of cytochromes P450 2E1 and 2E1 T303A by
tert-butyl acetylenes: characterization of reactive intermediate adducts to the
heme and apoprotein. Chem Res Toxicol 15: 1561-1571.
Chan AO, Ho CM, Chong HC, Leung YC, Huang JS, Wong MK, and Che CM
(2012) Modification of N-terminal α-amino groups of peptides and proteins
using ketenes. J Am Chem Soc 134: 2589-2598.
Dai Q, Ling YH, Lia M, Zou YY, Kroog G, Iwata KK and Perez-Soler R (2005)
22

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
Enhanced sensitivity to the HER1/epidermal growth factor receptor tyrosine
kinase inhibitor erlotinib hydrochloride in chemotherapy-resistant tumor cell
lines. Clin Cancer Res 11: 1572-1578.
Hollenberg PF, Kent UM, and Bumpus NN (2008) Mechanism-based inactivation
of human cytochromes P450s: experimental characterization, reactive
intermediates, and clinical implications. Chem Res Toxicol 21: 189-205.
Huang YS, An SJ, Chen ZH, and Wu YL (2009) Three cases of severe hepatic
impairment caused by erlotinib. Br J Clin Pharmacol 68: 464-467.
Jotte RM, Spige DR (2015) Advances in molecular-based personalized
non-small-cell lung cancer therapy: targeting epidermal growth factor receptor
and mechanisms of resistance. Cancer Med 4: 1621-1632.
Kalgutkar AS, Obach RS, and Maurer TS (2007) Mechanism-based inactivation
of cytochrome P450 enzymes: chemical mechanisms, structure-activity
relationships and relationship to clinical drug-drug interactions and
idiosyncratic adverse drug reactions. Curr Drug Metab 8: 407-447.
Li X, Kamenecka TM, and Cameron MD (2010) Cytochrome P450-mediated
bioactivation of the epidermal growth factor receptor inhibitor erlotinib to a
reactive electrophile. Drug Metab Dispos 38: 1238-1245.
Lin HL, D'Agostino J, Kenaan C, Calinski D, and Hollenberg PF (2013) The
effect of ritonavir on human CYP2B6 catalytic activity: heme modification
contributes to the mechanism-based inactivation of CYP2B6 and CYP3A4 by
ritonavir. Drug Metab Dispos 41: 1813-1824.
23

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
Lin HL, Hollenberg PF (2007) The inactivation of cytochrome P450 3A5 by
17alpha-ethynylestradiol is cytochrome b5-dependent: metabolic activation of
the ethynyl moiety leads to the formation of glutathione conjugates, a heme
adduct, and covalent binding to the apoprotein. J Pharmacol Exp Ther 321:
276-287.
Lin HL, Kent UM, and Hollenberg PF (2002) Mechanism-based inactivation of
cytochrome P450 3A4 by 17 alpha-ethynylestradiol: evidence for heme
destruction and covalent binding to protein. J Pharmacol Exp Ther 301:
160-167.
Lin HL, Zhang H, Jushchyshyn M, and Hollenberg PF (2010) Covalent
modification of Thr302 in cytochrome P450 2B1 by the mechanism-based
inactivator 4-tert-butylphenylacetylene. J Pharmacol Exp Ther 333: 663-669.
Ling J, Johnson KA, Miao Z, Rakhit A, Pantze MP, Hamilton M, Lum BL, and
Prakash C (2006) Metabolism and excretion of erlotinib, a small molecule
inhibitor of epidermal growth factor receptor tyrosine kinase, in healthy male
volunteers. Drug Metab Dispos 34: 420-426.
Liu W, Makrauer FL, Qamar AA, Jänne PA, and Odze RD (2007) Fulminant
hepatic failure secondary to erlotinib. Clin Gastroenterol Hepatol 5: 917-920.
Makris D, Scherpereel A, Copin MC, Colin G, Brun L, Lafitte JJ, and Marquette
CH (2007) Fatal interstitial lung disease associated with oral erlotinib therapy
for lung cancer. BMC Cancer 7: 150.
Massarelli E, Johnson FM, Erickson HS, Wistuba II, and Papadimitrakopoulou V
24

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
(2013) Uncommon epidermal growth factor receptor mutations in non-small
cell lung cancer and their mechanisms of EGFR tyrosine kinase inhibitors
sensitivity and resistance. Lung Cancer 80: 235-241.
McDonald RN, Schwab PA (1964) Strained ring systems. I. peroxidation studies
with certain acetylenes. The relevance of oxirene intermediates. J Am Chem
Soc 86: 4866-4871.
Orr ST, Ripp SL, Ballard TE, Henderson JL, Scott DO, Obach RS, Sun H, and
Kalgutkar AS (2012) Mechanism-based inactivation (MBI) of cytochrome
P450 enzymes: structure-activity relationships and discovery strategies to
mitigate drug-drug interaction risks. J Med Chem 55: 4896-4933.
Ortiz de Montellano PR, and Komives EA (1985) Branchpoint for heme
alkylation and metabolite formation in the oxidation of arylacetylenes by
cytochrome P-450. J Biol Chem 260: 3330-3336.
Pearce RE, Vakkalagadda GR, and Leeder JS (2002) Pathways of carbamazepine
bioactivation in vitro I. Characterization of human cytochromes P450
responsible for the formation of 2- and 3-hydroxylated metabolites. Drug
Metab Dispos 30: 1170-1179.
Saif MW (2008) Erlotinib-induced acute hepatitis in a patient with pancreatic
cancer. Clin Adv Hematol Oncol 6: 191-199.
Scambia G, Benedetti-Panici P, Ferrandina G, Distefano M, Salerno G, Romanini
ME, Fagotti A, and Mancuso S (1995) Epidermal growth factor, oestrogen and
progesterone receptor expression in primary ovarian cancer: correlation with
25

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
clinical outcome and response to chemotherapy. Br J Cancer 72: 361-366.
Stepan AF, Walker DP, Bauman J, Price DA, Baillie TA, Kalgutkar AS, and Aleo
MD (2011) Structural alert/reactive metabolite concept as applied in medicinal
chemistry to mitigate the risk of idiosyncratic drug toxicity: a perspective
based on the critical examination of trends in the top 200 drugs marketed in the
United States. Chem Res Toxicol 24: 1345-1410.
Tammaro KA, Baldwin PD, and Lundberg AS (2005) Interstitial lung disease
following erlotinib (Tarceva) in a patient who previously tolerated gefitinib
(Iressa). J Oncol Pharm Pract 11: 127-130.
Veale D, Kerr N, Gibson GJ, Kelly PJ and Harris AL (1993) The relationship of
quantitative epidermal growth factor receptor expression in non-small cell lung
cancer to long term survival. Br J Cancer 68: 162-165.
von Weymarn LB, Blobaum AL, and Hollenberg PF (2004) The
mechanism-based inactivation of P450 2B4 by tert-butyl 1-methyl-2-propynyl
ether: structural determination of the adducts to the P450 heme. Arch Biochem
Biophys 425: 95-105.
Wang K, Li W, Chen J, Peng Y, and Zheng J (2015) Detection of cysteine- and
lysine-based protein adductions by reactive metabolites of 2,5-dimethylfuran.
Anal Chim Acta 896: 93-101.
Wang K, Zheng L, Peng Y, Song JE, and Zheng J (2014) Selective and sensitive
platform for function-based screening of potentially harmful furans. Anal Chem
86: 10755-10762.
26

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
Footnotes: This work was supported in part by the National Natural Science
Foundation of China [Grants 81373471, 81430086, 81773813].
Jiang Zheng
1State Key Laboratory of Functions and Applications of Medicinal Plants, Key
Laboratory of Pharmaceutics of Guizhou Province, Guizhou Medical University,
Guiyang, Guizhou, 550004, P. R. China.
Jiang Zheng
2Wua College of Innovation, Shenyang Pharmaceutical University, 103 Wenhua Road,
Shenyang, Liaoning, 110016, P. R. China.
1,2: The two corresponding units contributed equally to the work.
27

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
Figure legends
Figure 1. Characterization of ELT-BBA adduct generated from HLM incubations.
Extracted ion (m/z 595/479 and m/z 597/481) chromatograms obtained from
LC-MS/MS analysis of HLM incubations containing ELT and BBA in the absence
(A and B) or presence (E and F) of NADPH. Extracted ion chromatograms of PI
scans of m/z 169 and m/z 171 obtained from LC-MS/MS analysis in the absence (C
and D) or presence (G and H) of NADPH. (I) MS/MS spectrum of ELT-BBA
adduct generated in microsomal incubations obtained from LC-MS/MS.
Figure 2. Characterization of ELT-BBA adduct generated from chemical synthesis.
Extracted ion chromatogram of MRM scans of m/z 597/481 obtained from
microsomal incubations (A) and chemical synthesis (C). MS/MS spectrum of
ELT-BBA adduct generated in microsomal incubations (B) and chemical synthesis
(D) obtained from LC-MS/MS.
Figure 3. Formation of ELT-BBA adduct in individual recombinant P450 enzyme
incubations containing NADPH, ELT and BBA. Data represent the mean±S.D. (n =
3).
Figure 4. Time-dependent inactivation of CYP3A by ELT and ELT analogs. HLMs
(0.5 mg/ml) was incubated with the following substrates: vehicle (■, black), ELT (◆,
red), vinyl analogs (●, blue) and ethyl analogs (▲, green), at concentrations of 25 (A)
and 50 (B) μM in the presence of NADPH at 37 °C for 0, 10, 20 and 30 min. The
residual enzymatic activity at each time point was determined through calculating
peak area ratio of the analytes compared with the internal standard, and the enzymatic
28

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
activities at 0 min were normalized to 100% at each concentration. Data represent
the mean±S.D. (n = 3). *P<0.05, **P<0.01, ***P<0.001 represent student's t-test P
values compared with vehicle groups.
Figure 5. Characterization of ELT-lysine adduct. Extracted ion chromatogram of
MRM scans of m/z 556/428 obtained from LC-MS/MS analysis of HLM incubations
containing ELT and lysine in the absence (A) or presence (B) of NADPH. Extracted
ion chromatograms obtained from the analysis of the proteolytic digestion of human
microsomal protein exposed to ELT in the absence (C) or presence (D) of NADPH.
Extracted ion chromatogram of MRM scans of m/z 556/428 obtained from chemical
synthesis (E). MS/MS spectrum of ELT-lysine adduct generated in chemical synthesis
(F) and proteolytic digestion (G) obtained from LC-MS/MS.
Figure 6. Formation of BBA and lysine based adduct with ELT and ELT analogs.
Extracted ion chromatograms of PI scans of m/z 171 obtained from LC-MS/MS
analysis of HLM incubations containing ELT (A), ethyl analog (B) and vinyl analog
(C). Extracted ion chromatograms of MRM scans of m/z 556/428 obtained from
the analysis of the proteolytic digestion of human microsomal protein exposed to
ELT (D), ethyl analog (E) and vinyl analog (F).
Figure 7. HPLC profile of the native heme analyses. HPLC chromatogram (A) at
400 nm and Q1MS spectrum (B) of the standard heme. HPLC chromatograms of
HLM incubations containing ELT (C), vinyl analog (E) and ethyl analog (G) in the
presence of NADPH. HPLC chromatograms of HLM incubations containing ELT
(D), vinyl analog (F) and ethyl analog (H) in the absence of NADPH.
29

DMD Fast Forward. Published on January 19, 2018 as DOI: 10.1124/dmd.117.079327
This article has not been copyedited and formatted. The final version may differ from this version.
DMD # 79327
Figure 8. Native heme loss of ELT, vinyl analogs and ethyl analogs in the absence
and presence of NADPH. Relative contents (%) of heme loss was determined
through calculating peak area of the analytes compared with the vehicle groups. The
values of vehicle groups were normalized to 100%. Data represent the mean±S.D.
(n=3). ***P<0.001 represent student's t-test P value.
30