Role of Glucuronidation and P450 Oxidation in the Bioactivation of Bromfenac

Article abstract of DOI:10.1021/acs.chemrestox.7b00293

Bromfenac is a nonsteroidal anti-inflammatory drug that was approved in the United States in 1997. It was withdrawn from clinical use less than one year later, in 1998, due to hepatotoxicity. We investigate the potential of bromfenac to be metabolized to reactive intermediates to further the current understanding of bromfenac bioactivation. Incubations were conducted with hepatocytes and human, rat, and cynomolgus liver microsomes fortified with cofactors and N-acetylcysteine. One thioether adduct of hydroxylated bromfenac and three thioether adducts of hydroxylated bromfenac indolinone were detected in extracts following incubations in liver microsomes fortified with NADPH and UDPGA. These findings demonstrate a bioactivation pathway for bromfenac and contribute to the body of evidence that could advance the understanding of the toxicity associated with bromfenac.

Full text of DOI:10.1021/acs.chemrestox.7b00293

Subscriber access provided by UNIVERSITY OF TOLEDO LIBRARIES
The Role of Glucuronidation and P450
Oxidation in the Bioactivation of Bromfenac
James P. Driscoll, Aprajita S. Yadav, and Nina R. Shah
Chem. Res. Toxicol., Just Accepted Manuscript • DOI: 10.1021/acs.chemrestox.7b00293 • Publication Date (Web): 23 Mar 2018
Downloaded from http://pubs.acs.org on March 29, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 26
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
The Role of Glucuronidation and P450 Oxidation in
the Bioactivation of Bromfenac
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
James P. Driscoll*, Aprajita S. Yadav and Nina R. Shah
MyoKardia Inc, 333 Allerton Ave. South San Francisco, CA 94080
ACS Paragon Plus Environment
1

Chemical Research in Toxicology
Page 2 of 26
1
2
3
TOC graphic
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
2

Page 3 of 26
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
ABSTRACT Bromfenac is a nonsteroidal antiꢀinflammatory drug which was approved in the
United States in 1997. It was withdrawn from clinical use less than one year later, in 1998, due
to hepatotoxicity. We investigate the potential of bromfenac to be metabolized to reactive
intermediates to further the current understanding of bromfenac bioactivation. Incubations were
conducted with hepatocytes and human, rat and cynomolgus liver microsomes fortified with
cofactors and Nꢀacetylcysteine. One thioether adduct of hydroxylated bromfenac and three
thioether adducts of hydroxylated bromfenac indolinone were detected in extracts following
incubations in liver microsomes fortified with NADPH and UDPGA. These findings
demonstrate a bioactivation pathway for bromfenac and contribute to the body of evidence which
could advance the understanding of the toxicity associated with bromfenac.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Introduction Bromfenac (Duract) is a nonsteroidal antiꢀinflammatory drug (NSAID) used for
the treatment of osteoarthritis, rheumatoid arthritis, ankylosing spondylitis, and acute muscle
pain. Bromfenac (1) was approved in the United States in 1997 for a duration of use of less than
10 days. The drug was voluntarily withdrawn in 1998 due to adverse reactions to 1 including
hepatic necrosis, cholestasis and fatal liver failure.1,2 In many cases of bromfenac associated
hepatotoxicity there was a significant increase in liver alanine aminotransferase (ALT), aspartate
aminotransferase (AST) and alkaline phosphatase (AP) levels. Patients taking the drug for more
than 10 days had a higher risk of developing liver injury or failure.3
ACS Paragon Plus Environment
3

Chemical Research in Toxicology
Page 4 of 26
1
2
3
4
5
6
7
8
9
HO
O
O
R
O
O
R
Liver Microsomes
UDPGA
O
NH2
N
NH2
H2
O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Br
Br
Br
1
Tetrahedral intermediate
R
Bromfenac glucuronide intermediate
R = Glucuronic acid
Liver Microsomes
NADPH
NAC
OH
OH
O
O
O
O
NAC
NH2
NH2
O
O
O
N
H
Br
Br
O
5
3
Br
2
NAC
Liver MIcrosomes
NADPH
O
O
NAC
O
O
N
N
H
O
Quinone-imine intermediate
O
O
O
Br
N
H
6
NAC
NAC
O
O
N
H
NAC
Br
O
O
4
N
H
O
O
N
O
H
Br
7 & 8
Quinone methide intermediates
Scheme 1. Proposed bioactivation pathway for bromfenac. NAC = N-acetylcysteine
In mass balance studies, healthy volunteers were dosed with 14Cꢀlabeled 1 and bromfenac
indolinone (2) (Scheme 1) was found to be the major metabolite in urine. No evidence was found
for the formation of thioether–linked metabolites.4 In preclinical toxicology studies with rats, 1
ACS Paragon Plus Environment
4

Page 5 of 26
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
was found to cause vacuolar changes, necrosis and cytological alterations to hepatocytes, and
glutathione depletion in the liver and kidney. The compound was also found to cause papillary
necrosis of the kidney and caused cytological alterations of hepatocytes in mouse carcinogenicity
studies.5 In vitro evidence suggests that in the presence of NADPH and UDPGA, the amount of
covalent binding of 1 to human liver microsomal protein is increased when compared with
NADPH alone.6 The in vitro covalent binding result suggests that bioactivation could proceed
through a glucuronidation pathway, although an acyl glucuronide metabolite was never detected
in human radiolabeled studies and could not be synthesized.7,4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Chemical structural motifs in drugꢀlike molecules that are known to cause drug induced liver
injury (DILI) are known as structural alerts or toxicophores.8 Bromfenac contains three different
known structural alerts; an aniline, a bromobenzene ring, and a phenylacetic acid.9,10,11 Many of
the adverse events associated with bromfenac affect the liver, therefore it is possible that
formation of reactive metabolites are involved in observed adverse events. However, such
reactive metabolites have not yet been reported to the authors’ knowledge.
The purpose of this work is to investigate the metabolism of 1 to determine if these
toxicophores play a role in the formation of reactive metabolites. Incubations were performed
with 1 in human hepatocytes and human, rat and monkey liver microsomes supplemented with
UDPGA, NADPH and thiols, such as glutathione and Nꢀacetylcysteine. During our
investigations, we found evidence for the formation of novel thioether adducts of hydroxylated
bromfenac (3) and the hydroxylated indolinone metabolite (4). We also found that the formation
of 2 is likely catalyzed by uridine 5'ꢀdiphosphoꢀglucuronosyltransferase (UGT).
ACS Paragon Plus Environment
5

Chemical Research in Toxicology
Page 6 of 26
1
2
3
4
Materials and Methods
5
6
7
8
Materials. Nꢀacetylcysteine (cat number A7250ꢀ50G), Bromfenac Sodium (cat number
SML0289ꢀ50MG), Carbamazepine (cat number 94496ꢀ100MG), Glutathione (cat number
G6529ꢀ25G), LꢀGlutamine (cat number G6392ꢀ1VL), and Williams’ Medium E (cat number
W1878ꢀ500ml) were purchased from Sigma (St Louis, MO). All solvents used for LC/MS
analysis were of chromatographic grade and purchased from Fisher Scientific. Stock solutions of
bromfenac and bromfenac indolinone were prepared as solutions in dimethyl sulfoxide (20 mM).
Human hepatocytes (20 donor pool), human (pooled), rat and cynomolygus monkey liver
microsomes were purchased from BioreclamationIVT (Baltimore, MA). Potassium Phosphate
buffer (Cat No J62397) was purchased from Alfa Aesar (Ward Hill, MA)
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Bromfenac Indolinone Synthesis. The bromfenac indolinone standard was synthesized from
bromfenac sodium obtained from Sigma. 21 mg of bromfenac was dissolved in 5 ml of water. 2
ml of 6N hydrochloric acid was added and a white precipitate immediately formed. The
resulting saturated solution was vacuum filtered and the filtride was dried under vacuum for 24
hours. A 1 mg sample was dissolved in DMSOꢀd6 (D, 99.8%, Cambridge Isotope) containing
0.05% V/V TMS as an internal chemicalꢀshift reference standard, and transferred to a 3 mm
1
NMR tube, purged with nitrogen, and sealed. The H NMR of the synthesized material is
consistent with the structure of the indolinone (Figure 1).
ACS Paragon Plus Environment
6

Page 7 of 26
Chemical Research in Toxicology
1
2
3
4
5
6
7
18
8
9
3,1
6,4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13 14
15
16
10
9
8
7
6
5
4
Chemical Shift (ppm)
1
Figure 1. H NMR of 2. Collected on a Bruker Ascend™ (400 MHz in DMSOꢀd6) δ: 10.29 (s,
1H), 7.72 (d, J=8.44 Hz, 2H), 7.56ꢀ7.64 (m, 2H), 7.93ꢀ7.45 (m, 1H), 7.24ꢀ7.30 (m, 1H), 6.94ꢀ
7.02 (m, 1H), 3.52 (s, 2H).
Metabolite Identification Studies. Metabolite identification experiments were performed to
determine the fate of 1 (20 µM) and 2 (20 µM) in human, rat and cynomolygus monkey liver
microsomes (1 mg protein/ml). Incubations were conducted in potassium phosphate buffer (0.1
M, pH 7.4) in either the presence or absence of NADPH (1 mM) and UDPGA (1 mM). All
incubations included NAC (5 mM), alamethicin (10 µg/ml), and MgCl2 (5 mM). 10 ml total
volume reactions were performed in 20 ml glass scintillation vials in a shaking water bath (low
speed, 37°C). A metabolite identification study was performed as above with βꢀmercatoethanol
ACS Paragon Plus Environment
7

Chemical Research in Toxicology
Page 8 of 26
1
2
3
4
5
6
7
8
9
(0.1 mM) as the nucleophilic thiol. Since βꢀmercatoethanol has been shown to inhibit
cytochrome P450 at higher concentrations, the concentration was chosen to be three times lower
than the lowest IC50.12
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Incubations with glutathione were performed with 1 (20 µM) in human hepatocytes (1 million
cells/ml) and human liver microsomes (1 mg protein/ml). Microsome incubations were
conducted in potassium phosphate buffer (0.1 M, pH 7.4) in the presence of NADPH (1 mM),
UDPGA (1 mM) and glutathione (5 mM). Hepatocyte incubations included glutathione (5 mM)
in William’s E media which was added to the cells to bring the concentration to 1 million
cells/ml. Microsome incubations were performed at 8 ml volume reactions and hepatocyte
incubations were performed at 5 ml total volume in 20 ml glass scintillation vials in a shaking
water bath (low speed, 37°C).
All reactions were quenched after one hour with an equal volume of acetonitrile (ACN),
capped, and vortexꢀmixed (1 min) then were centrifuged (14,000 rpm, 5 min). Supernatants were
transferred to glass 13 × 100 mm tubes and evaporated in a Speed Vac® Plus SC210A Savant
system (Savant Instruments, Inc. Farmingdale, NY) under vacuum. Once dried, extracts were
reconstituted with a solution containing ACN and water (1 mL, 1/1, v/v). The reconstituted
mixtures then were transferred to microcentrifuge tubes and centrifuged (14,000 rpm, 5 min).
The resulting supernatants were transferred to highꢀperformance liquid chromatography vials.
LC-MS/MS Conditions for Metabolite Identification Studies.
High resolution LCꢀMS/MS analyses was conducted using a Dionex Ultimat 3000 HPLC
(Thermo Fisher Scientific, Waltham, MA) equipped with a binary pump, degasser, ultraviolet
detector, CTC autosampler, and column heater kept at 25°C. The injection volume was 50 µl.
Chromatographic resolution was achieved with a ThermoScientific (Waltham, MA) BetaBasic
ACS Paragon Plus Environment
8

Page 9 of 26
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
C18 column (4.6 × 250 mm, 3µm). Mobile phases consisted of 0.1% aqueous formic acid
(solvent A) and acetonitrile with 0.1% formic acid (solvent B) run at a constant flow rate of 0.7
ml/min. The solvent gradient was initially held at 0% solvent B for 0.1 min and increased
linearly to 100% solvent B over 53 min, kept at 95% solvent B for an additional one min, then
immediately dropped to 5% solvent B over 0.5 min, where it was held constant at 0% solvent B
for 5.5 min before the next sample injection (20 µl). A Thermo Fisher XL Orbitrap mass
spectrometer, with electrospray ionization was employed with the needle potential held at 4.5
kV. Tandem mass spectrometry conditions used were 2 mTorr helium collision gas and 25ꢀ35
eV collision potential. Ions of interest were put on a data dependent mass list and if encountered
were fragmented. Positive ion mode full scan (200–750 Da) was conducted with a scan time of
0.73 s. Ultraviolet data collection was performed over a scan range of 200 to 400 nM with 2 hz
data collection frequency. Xcalibur software (version 2.0) was used to acquire all data.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Metabolism of Bromfenac to Indolinone. The timeꢀdependent degradation of 1 (1 µM) in
HLM (1 mg total protein/ml) with alamethicin (10 µg/ml) and MgCl2 (5 mM) was performed to
quantify metabolism of 1 and the formation of 2. Incubations were conducted in Tris buffer.13
Aliquots (50 µl) were taken from the incubation (300 µl total) at 0, 30, 60, 120 and 240 min in
the presence and absence of UDPGA (1 mM). The aliquots at each time point were added to
quench solution (100 µL) containing internal standard, carbamazepine (50 nM). Samples were
centrifuged and the supernatant was diluted 1:1 v:v with HPLC grade water and then analyzed by
LCꢀMS/MS. The concentration of 1 and the amounts of 2 formed were determined from
standard curves of peak area ratios (peak area of analyte/peak area of IS).
The LCꢀMS/MS conditions were as follows: Sciex API 4000 (Waltham, MA, USA) mass
spectrometer was run in positive ion mode with the Turbo Spray source with the following
ACS Paragon Plus Environment
9

Chemical Research in Toxicology
Page 10 of 26
1
2
3
4
5
6
7
8
9
source conditions; Collision gas set to 6, Curtain gas set to 25, Ion source gas 1 and 2 set to 30,
IonSpray voltage set to 5000 and the temperature set to 600 °C with the interface heater on. The
declustering potential was set at 90 volts for all of the compounds and the dwell time for each ion
was set to 50 msec. The transition ions in the MRM method (Q1 ꢀ Q3) for 1 were m/z 334
ꢀ288 with the collision energy set to 25 volts, the MRM method for 2 were m/z 316ꢀ183 with
the collision energy set to 30 volts and CBZ were m/z 237ꢀ194 with the collision energy set to
27 volts. The LC system included a CTC Pal autosampler and ThermoElectron (Waltham, MA)
Accela 1250 binary pump. The mobile phases were 0.1% aqueous formic acid (solvent A) and
acetonitrile with 0.1% formic acid (solvent B). The injection volume was 5 µl and the flow rate
was 900 µl/min with the step gradient starting at 95% A for 10 sec then stepping to 85% B
isocratic for 30 sec then stepping back to 95% A for another 30 sec. The column used was a
Phenomenex (Torrance, CA) Kinetex 5 µm C18 100Å 30 x 2.1 mm.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
10

Page 11 of 26
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Chromatograms of 60 min incubations in HLM with 1 (rt = 36.8 min) with (A) no
cofactors, (B) NADPH, (C) is with UDPGA and (D) UDPGA and NADPH. (E) 2 standard
incubated with NADPH (rt = 35.9 min). (F) is the extracted ion chromatogram of exact m/z of
493.0063 (rt = 22.5, 32.2 and 32.6 min) and 511.0169 (rt = 23.4 min) of HLM incubations with 1
and UDPGA and NADPH and (G) is 2 with NADPH. All incubations include NAC at 5 mM.
Results
Metabolite Identification Experiments. Qualitative high resolution LC–MS/MS analysis
with ultraviolet (UV) detection of liver microsome incubations containing 1 with UDPGA and
alamethicin led to an increase in the formation of a large UV peak at retention time (rt) 35.9 min
ACS Paragon Plus Environment
11

Chemical Research in Toxicology
Page 12 of 26
1
2
3
4
(Figure 2). The peak contained exact m/z of 315.9968 corresponding to a molecular formula of
5
6
+
C15H11BrNO2 within 5 ppm. The mass spectrum and retention time of the synthetic standard of
7
8
9
the 2 matched the peak of the metabolite formed from the 1 incubations at rt 35.9 min (Figure
2C, 2D and 2E).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Liver microsome incubations with NADPH, UDPGA, alamethicin and NAC with 1 and 2 led
to the detection of the three thiol conjugates with an m/z equal to 493.0063 and one with an m/z
of 511.0169. The m/z of 493.0063 is consistent with that of 2 with one atom of oxygen and one
molecule of NAC added. This exact mass corresponds to the molecular formula
C20H18BrN2O6S+. The m/z of 511.0169 is consistent with that of 1 with one atom of oxygen and
one molecule of NAC added; corresponding to the molecular formula C20H20BrN2O7S+.
The high resolution full scan of 1 at rt 36.7 min corresponds to a m/z 334.0073 which
+
corresponds to a molecular formula of C15H13BrNO3 . The product ion scan contains ions of m/z
315.9968, 288.0019 and 182.9440. The ion at 315.9968 is the neutral loss of water from 1, the
ion at 288.0019 is the loss of CO2 with charge retention on the carbocation. The ion at m/z of
182.9440 corresponds to the 4ꢀbromobenzyloxonium ion. This ion is conserved in the product
ion scan of the 2 (rt = 35.8 min). These spectrum are found in the supporting data in Figure S1
and S2.
The NAC adduct at retention time 22.5 min (6) had the CID spectrum shown in Figure 3A and
had ions at m/z 180.0114, 182.9440, 206.0270 and 293.0591. The fragment ion with m/z of
182.9440 corresponds to a 4ꢀbromobenzyloxonium ion. The fragment ion at m/z 180.0114 was
assigned as the product of a neutral loss of 4ꢀbromobenzylaldehyde and cleavage adjacent to the
cysteinyl thioether moiety, with charge retention on the indolinone moiety containing oxygen
and sulfur atoms attached; this corresponds to a molecular formula of C8H6NO2S+. The presence
ACS Paragon Plus Environment
12

Page 13 of 26
Chemical Research in Toxicology
1
2
3
4
5
6
of these two fragment ions is consistent with an aromatic thioether motif on the
hydroxyindolinone moiety.
7
8
9
The NAC adduct which corresponds to the hydroxylated bromfenac (5), m/z of 511.0169, had
the product ion spectra shown in Figure 3B and retention time of 23.4 min. The spectrum
contains the presence of the 4ꢀbromobenzyloxonium ion with m/z 182.9440. The presence of this
ion indicates hydroxylation and NAC have been added to the 2ꢀaminophenylacetic acid moiety
of the metabolite. The ion at m/z 293.0591 (Figure 3B) also contributes to the evidence of the
NAC and oxygen adding to the 2ꢀaminophenylacetic acid.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The NAC adducts at retention time 32.2 min (7) and 32.6 min (8) had very similar product ion
spectra shown in Figure 3C and 3D. The NAC adducts contained the ions at m/z 182.9440 and
329.9760. The fragment ion with m/z of 182.9440 corresponds with a 4ꢀbromobenzyloxonium
ion. The fragment ion at m/z 329.9760 was assigned as the homolytic cleavage product with a
neutral loss of the NAC; the charge is retained on the hydroxylated indolinone moiety
+
corresponding to a molecular formula of C15H9BrNO3 . The presence of these two fragment ions
is consistent with an aliphatic thioether motif on the hydroxyindolinone moiety.14
Rat and cynomologus monkey liver microsomes were incubated with 1 in the presence of
UDPGA, NADPH, NAC and alamethicin. The extracted ion chromatograms are found in Figure
4 with a comparison to HLM. Compared to HLM, 5 and 6 were formed in lesser proportion to 7
and 8 in RLM. Comparing CLM to HLM, 6 seemed to be formed in a greater proportion to 7 and
8 and 5 was not detected (Figure 4). All three of the 2 NAC adducts, 6, 7 and 8, were detected in
RLM and CLM. The retention times and spectra of the product ion scans of the NAC adducts
were consistent with those found in HLM.
ACS Paragon Plus Environment
13

Chemical Research in Toxicology
Page 14 of 26
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Product ion spectrum of (A) 6 with m/z 493.0063 at retention time 22.5 min (B) 5 with
m/z of 511.0169 at retention time of 23.4 min (C) 7 with m/z 493.0063 at retention time 32.3 min
(D) 8 with m/z 493.0063 at retention time 32.6 min. Incubations were conducted in HLM with 20
µM 1 for 60 minutes with UDPGA, NADPH, NAC and alamethicin.
1 was incubated with human hepatocytes to determine if reactive intermediates can form in a
whole cell system and be trapped by the endogenous nucleophile; glutathione (GSH). HLM with
NADPH, UDPGA, alamethicin and GSH were also incubated as a control experiment. The
results indicated that similar reactive intermediates are trapped with HLM with NAC. In the
HLM experiments, four GSH adducts were found. One adduct (rt 18.94 min) is consistent with 3
and the addition of GSH to the 2ꢀaminophenylacetic acid moiety matching m/z of 655.0704. The
ACS Paragon Plus Environment
14

Page 15 of 26
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
product ion spectrum of the adduct contains the ions with m/z of 182.9440 and 379.9587, which
adds to the evidence. The other three adducts (rt 18.3, 25.2, and 25.4 min) are consistent with 4
with GSH added. The m/z is equal to 637.0598, and is similar to the adducts described above
with NAC as the trapping agent. All three product ion spectra include m/z of 182.9440 which
corresponds to the 4ꢀbromobenzyloxonium ion. Interestingly, in the hepatocyte incubations only
three GSH adducts were found, two from 4 and one from 3. One of the adducts, which we
speculate to be formed from a quinone methide intermediate of 4, was not present in the human
hepatocyte experiment (Figure S4).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Bromfenac Stability and Formation of Indolinone Metabolite.
1 was incubated in HLM with UDPGA containing alamethicin and the loss of 1 and formation
of 2 was quantified using the synthetic standard. The incubation was monitored over the course
of 240 min and approximately 0.5 µM of 1 was metabolized to 0.5 µM of 2 (Figure 5).
ACS Paragon Plus Environment
15

Chemical Research in Toxicology
Page 16 of 26
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Chromatograms of 60 min incubations of 1 with NAC, UDPGA, alamethicin and
NADPH in (A) human liver microsomes (B) rat liver microsomes, and (C) cynomolgus monkey
liver microsomes. Extracted ion chromatogram of 4+ NAC with exact mass m/z 493.0063 (6, 7,
and 8) and 5 with exact mass m/z 511.0169.
Discussion
1 bioactivation has been studied since the compound was withdrawn from the market in 1998
and has since been reviewed in the literature.6,15 Working from the hypothesis that DILI can
arise from the formation of reactive metabolites and covalent binding to cellular
macromolecules, a few studies have been done to elucidate the mechanism of covalent binding
of bromfenac.6 Following up on Usiu et al 2009, we attempt to address the NADPH and
UDPGA dependence of bromfenac covalent binding to HLMs.
ACS Paragon Plus Environment
16

Page 17 of 26
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
We propose that an UDPGAꢀdependent acyl glucuronide intermediate forms. Acylꢀ
glucuronides and esters in general are susceptible to hydrolysis by free amines.16 This unstable
intermediate undergoes intramolecular nucleophilic substitution via additionꢀelimination to form
the bromfenac indolinone metabolite (Scheme 1). In this reaction, we propose the free amine
functions as the nucleophile, attacking the electrophilic carbonyl carbon of the acyl glucuronide
ester to form a tetrahedral intermediate. The tetrahedral intermediate collapses, eliminating the
glucuronic acid and reforming the carbonyl to provide the 2. This mechanism is also supported
by the acid catalyzed synthesis of 2. When the carboxylic acid is protonated with the addition of
acid, the alcohol makes for a better leaving group, thereby facilitating the formation of the
indolinone.17
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For additional confirmation of this finding, 1 was incubated in HLM with UDPGA, and the
loss of 1 and formation of the 2 was quantified. Over a 240 min incubation, approximately 0.5
µM of 1 was metabolized to 0.5 µM of 2 (Figure 5). This result demonstrates the UDPGAꢀ and
time dependentꢀ formation of 2 from 1, since formation of 2 was not detected in HLM in the
absence of UDPGA.
It is unlikely the acyl glucuronide intermediate could be the cause of covalent binding to
proteins due to the equal loss of 1 and formation of 2; this leaves little room for binding to other
reactants. Literature evidence exists for intramolecular reactions occurring at very favorable
rates due to the effective high concentrations of reactants. Furthermore, intramolecular reactions
are entropically favored as opposed to intermolecular reactions which require bringing two
separate molecules together.18,19 It should be noted that in Usui et al’s work, there was a small
amount of covalent binding to the HLM proteins with NAPDH alone which could be attributed
to the reactive intermediate that leads to 5 with retention time of 23.4 min (Figure 2F and 3B).
ACS Paragon Plus Environment
17

Chemical Research in Toxicology
Page 18 of 26
1
2
3
4
5
6
7
8
9
The metabolism of the 2 was investigated since it originates from a glucuronidation pathway
and is the largest metabolite found in human urine.4 To see if the synthetically prepared standard
could undergo further P450 metabolism to a chemicallyꢀreactive intermediate, 2 was incubated
in liver microsomes. Three NAC adducts were found when 2 was incubated in HLM with NAC
and NADPH (Scheme 1 and Figure 2G). Each NAC adduct was confirmed to have the isotopic
distribution pattern of bromine with an m/z peak 1.9979 Dalton higher at approximately 98% of
the original m/z peak of interest in the full scan mass spectrum. No other major metabolites were
detected by UV traces in liver microsome incubations with 2 and the NAC adducts could only be
found by extracting the ions of interest (Figure 2F, 2G).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In order to ensure these NAC adducts were formed from 1 as the starting material, 1 was
incubated in LM with NADPH, UDPGA, NAC and alamethicin. The results indicate an increase
in the formation of 2 at rt 36.0 min (Figure 2D and 2E). There were also four NAC adducts
found by LCꢀMS/MS (Figure 2), three of which have consistent retention time and product ion
spectra with the NAC adducts detected in the incubation with 2 (Figure 3).
One proposed mechanism leading to 6 could be oxidation of the indolinone moiety para to the
amine, which can undergo a further two electron oxidation to an electrophilic quinoneꢀimine
intermediate (Scheme 1). This intermediate can undergo adduction by NAC at the two open
aromatic carbons meta to the amine. While there is a possibility for thioether formation at each
of the two open meta positions, only one adduct was found. One explanation for this could be
the steric hindrance of one position over the other. Another possible mechanism is that the
hydroxylation and thioether formation could go through an epoxidation mechanism, although
this would be better supported with two peaks in the chromatogram. The structure is supported
by another fragment which has an m/z 180.0114 in the fragmentation of 6 (Figure 2F). This
ACS Paragon Plus Environment
18

Page 19 of 26
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
corresponds to the indolinone moiety with sulfur and oxygen attached, which we speculate is the
NAC and oxygen attached to the aromatic ring. There are literature examples that show when
quinoneꢀimines are formed, the nucleophilic thiols can attach to the aromatic ring at the ortho or
meta position to the hydroxyl group.20,21 Thioether adducts that retain the sulfur atom upon
fragmentation are characteristically found to be attached to carbons in aromatic rings.14
We hypothesize that the thioether adducts at retention times 32.2 min (7) and 32.6 min (8)
formed from 2 are due to hydroxylation ortho and para to the methylene carbon. After a further
two electron oxidation, the resulting intermediates are quinone methides (Scheme 1). In this
intermediate, the methide carbon would be the most electron deficient and susceptible to
nucleophilic addition of nucleophilic thiols.22 This is supported by the homolytic cleavage of the
sulfur and the hydroxylated indolinone ions in the product ion spectra of the peaks (Figure 3C
and 3D). Another possible explanation for the aromatic thioether adducts could be addition of
NAC on the aromatic ring from the quinone methide reactive intermediates, where there are
potentially reactive α,β unsaturated ketones. However, thiols reacting with the aromatic ring in
quinone methides is not supported by the literature.23,24,25
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
If the NAC adducts result from the formation of quinone methide intermediates and attach to
the aliphatic carbon of the indolinone moiety, there is a possibility of forming two diastereomers
on the prochiral benzylic position as opposed to the oxidative isomers we propose in Scheme 1.26
With two chiral centers present it is possible to separate them on a reverse phase C18 column.
To test which hypothesis is correct, we incubated 2 with HLM and βꢀmercaptoethanol, a thiol
with no chiral center. After extraction of the exact mass m/z 407.9900 from the LCꢀMS trace,
three peaks were found. The exact mass corresponds to 4 with the addition of βꢀmercaptoethanol,
ACS Paragon Plus Environment
19

Chemical Research in Toxicology
Page 20 of 26
1
2
3
4
5
6
7
8
9
matching the molecular formula of C17H15BrNO4S+. Finding three adduct peaks in this
experiment supports the hypothesis of oxidative isomers and not diastereomers (Figure S3).
Similar structures of thioether adducts were found when bromfenac was incubated in HLM and
human hepatocytes fortified with GSH (Figure S4). One of the proposed quinone methide
metabolites was not found. A major difference between GSH and NAC is the size of the
molecule; we speculate that since NAC is smaller it can potentially fit into active sites more
easily than GSH. Thus, GSH may not fit in the active site of the P450 where the reactive
intermediate is formed.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Complete structural analysis of the adducts could be difficult to fully assign unless synthetic
standards are made or enough material can be collected from liver microsome incubations and
analyzed by two dimensional NMR. Collecting sufficient sample from microsome incubations
has been challenging and is the focus of ongoing efforts of our lab.
Figure 5. Disappearance of 1 with and without UDPGA and the formation of the 2 with and
without UDPGA in incubations in human liver microsomes.
ACS Paragon Plus Environment
20

Page 21 of 26
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
Since these thiol adducts could only be detected by extracting their exact masses, it brings into
question their relevance to 1 bioactivation. One possible explanation for the lack of sensitivity is
that these reactive intermediates are covalently bound to microsomal proteins, since the covalent
binding of 1 has been previously demonstrated.6 Another possible explanation for the relevance
of these reactive intermediates is that 1 was approved for only a ten day duration of use, and
there may be some buildup of reactive intermediates. Potential evidence is found in a preclinical
toxicology study where glutathione depletion in rat liver was observed.5
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In summary, we have found evidence for the formation of reactive metabolites of 1. One
pathway for the bioactivation requires oxidation of 1 followed by thioether adduct formation.
Another pathway of bioactivation requires a shortꢀlived acyl glucuronide intermediate which
results in the formation of 2. 2 requires further oxidation by P450 on every open carbon on the
aromatic ring of the indolinone moiety. These hydroxylated metabolites require a further
oxidation to form the quinoneꢀimine and quinone methide intermediates. This bioactivation
pathway contributes to the body of evidence which may help to understand the mechanisms of
toxicity resulting from 1 use in the clinic. Reactive metabolites are not the only cause of
hepatotoxicity and/or cholestasis; other potential sources include the inhibition of transporters
including bile salt excretory pump.27
Supporting Information.
MS/MS interpreted spectrum of bromfenac and bromfenac indolinone. Chromatograms of XIC
of glutathione adducts and βꢀmercaptoethanol adducts. Interpreted spectrum of GSH adducts
(PDF).
Corresponding Author
ACS Paragon Plus Environment
21

Chemical Research in Toxicology
Page 22 of 26
1
2
3
4
* Tel: 650ꢀ741ꢀ0928. Eꢀmail: jdriscoll@myokardia.com
5
6
7
Present Addresses
8
9
333 Allerton Ave. South San Francisco CA 94080
Author Contributions
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The manuscript was written by James P. Driscoll and Aprajita S. Yadav. Experiments were
performed by all of the authors. All authors have given approval to the final version of the
manuscript.
Funding Sources
All funds used to support the research of the manuscript were provided by MyoKardia, Inc.,
ACKNOWLEDGMENT
The authors would like to thank Drs. Mark P. Grillo and Tim Carlson for useful discussion and
review of the manuscript. We would like to thank Jasleen K. Sodhi for helpful review of the
manuscript. We would like to thank Drs. Danielle Aubele and Brian Kane for purification and 1H
NMR of bromfenac indolinone.
ABBREVIATIONS
NSAID, nonsteroidal antiꢀinflammatory drug; NADPH, nicotinamde adenosine diphosphate
reduced; UDPGA, uridine diphosphate glucuronic acid; HLM, human liver microsomes; LCꢀ
MSꢀUV, liquid chromatography mass spectrometry ultraviolet detection; NMR, nuclear
magnetic resonance; ml, milliliter; mg milligram, NAC, Nꢀacetylcysteine; AST ALT
REFERENCES
(1) Rabkin, J. M., Smith, M. J., Orloff, S. L., Corless, C. L., Stenzel, P., and Olyaei, A. J. (1999)
ACS Paragon Plus Environment
22

Page 23 of 26
Chemical Research in Toxicology
1
2
3
4
Fatal fulminant hepatitis associated with bromfenac use. Ann. Pharmacother. 33, 945–947.
5
6
7
8
9
(2) Moses, P. L., Schroeder, B., Alkhatib, O., Ferrentino, N., Suppan, T., and Lidofsky, S. D.
(1999) Severe hepatotoxicity associated with bromfenac sodium. Am. J. Gastroenterol. 94,
1393–1396.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(3) Fontana, R. J., McCashland, T. M., Benner, K. G., Appelman, H. D., Gunartanam, N. T.,
Wisecarver, J. L., Rabkin, J. M., and Lee, W. M. (1999) Acute liver failure associated with
prolonged use of bromfenac leading to liver transplantation. Liver Transplant. Surg. 5, 480–484.
(4) Osman, M., Chandrasekaran, A., Chan, K., Scatina, J., Ermer, J., Cevallos, W., and
Sisenwine, S. F. (1998) Metabolic disposition of 14Cꢀbromfenac in healthy male volunteers. J.
Clin. Pharmacol. 38, 744ꢀ752
(5) Peters, T. S. (2005) Do Preclinical Testing Strategies Help Predict Human Hepatotoxic
Potentials? Toxicol. Pathol. 33, 146–154.
(6) Usui, T., Mise, M., Hashizume, T., Yabuki, M., and Komuro, S. (2009) Evaluation of the
Potential for DrugꢀInduced Liver Injury Based on in Vitro Covalent Binding to Human Liver
Proteins. Drug Metab. Dispos. 37, 2383–2392.
(7) Kirkman, S. K., Zhang, M. Y., Horwatt, P. M., and Scatina, J. (1998) Isolation and
identification of bromfenac glucoside from rat bile. Drug Metab. Dispos. 26, 720–3.
(8) Stepan, A. F., Walker, D. P., Bauman, J., Price, D. a, Baillie, T. a, Kalgutkar, A. S., and Aleo,
M. D. (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
ACS Paragon Plus Environment
23

Chemical Research in Toxicology
Page 24 of 26
1
2
3
4
trends in the top 200 drugs marketed in the United States. Chem. Res. Toxicol. 24, 1345–410.
5
6
7
(9) Williams, D. P., and Park, B. K. (2003) Idiosyncratic toxicity: The role of toxicophores and
8
9
bioactivation. Drug Discov. Today 8, 1044–1050.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(10) Grillo, M. P., and Lohr, M. T. (2009) Covalent binding of phenylacetic acid to protein in
incubations with freshly isolated rat hepatocytes. Drug Metab. Dispos. 37, 1073–82.
(11) Bambal, R. B., and Hanzlik, R. P. (1995) Bromobenzene 3,4ꢀoxide alkylates histidine and
lysine side chains of rat liver proteins in vivo. Chem. Res. Toxicol. 8, 729–35.
(12) Sodhi, J. K., Delarosa, E. M., Halladay, J. S., Driscoll, J. P., Mulder, T., Dansette, P. M.,
and Khojasteh, S. C. (2017) Inhibitory Effects of Trapping Agents of Sulfur Drug Reactive
Intermediates against Major Human Cytochrome P450 Isoforms. Int. J. Mol. Sci. 18, 1553.
(13) Walsky, R. L., Bauman, J. N., Bourcier, K., Giddens, G., Lapham, K., Negahban, A., Ryder,
T. F., Obach, R. S., Hyland, R., and Goosen, T. C. (2012) Optimized assays for human UDPꢀ
glucuronosyltransferase (UGT) activities: Altered alamethicin concentration and utility to screen
for UGT inhibitors. Drug Metab. Dispos. 40, 1051–1065.
(14) Xie, C., Zhong, D., and Chen, X. (2013) A fragmentationꢀbased method for the
differentiation of glutathione conjugates by highꢀresolution mass spectrometry with electrospray
ionization. Anal. Chim. Acta 788, 89–98.
(15) Park, B. K., Boobis, A., Clarke, S., Goldring, C. E. P., Jones, D., Kenna, J. G., Lambert, C.,
Laverty, H. G., Naisbitt, D. J., Nelson, S., NicollꢀGriffith, D. a, Obach, R. S., Routledge, P.,
Smith, D. a, Tweedie, D. J., Vermeulen, N., Williams, D. P., Wilson, I. D., and Baillie, T. a.
ACS Paragon Plus Environment
24

Page 25 of 26
Chemical Research in Toxicology
1
2
3
4
5
6
(2011) Managing the challenge of chemically reactive metabolites in drug development. Nat.
Rev. Drug Discov. 10, 292–306.
7
8
9
(16) Vaz, A. D. N., Wang, W. W., Bessire, A. J., Sharma, R., and Hagen, A. E. (2010) A rapid
and specific derivatization procedure to identify acylꢀglucuronides by mass spectrometry. Rapid
Commun. Mass Spectrom. 24, 2109–2121.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(17) Testa, B., and Krämer, S. D. (2007) The biochemistry of drug metabolismꢀꢀan introduction:
part 3. Reactions of hydrolysis and their enzymes. Chem. Biodivers. 4, 2031–122.
(18) Capon, B. (1964) Neighbouring group participation. Q. Rev. Chem. Soc. 18, 45.
(19) Menger, F. M. (1985) On the source of intramolecular and enzymatic reactivity. Acc. Chem.
Res. 18, 128–134.
(20) Zheng, J., Ma, L., Xin, B., Olah, T., Humphreys, W. G., and Zhu, M. (2007) Screening and
identification of GSHꢀtrapped reactive metabolites using hybrid triple quadruple linear ion trap
mass spectrometry. Chem Res Toxicol 20, 757–766.
(21) Chen, W., Shockcor, J. P., Tonge, R., Hunter, A., Gartner, C., and Nelson, S. D. (1999)
Protein and nonprotein cysteinyl thiol modification by Nꢀacetylꢀpꢀ benzoquinone imine via a
novel Ipso adduct. Biochemistry 38, 8159–8166.
(22) Thompson, D. C., Thompson, J. A., Sugumaran, M., and Moldéus, P. (1993) Biological and
toxicological consequences of quinone methide formation. Chem. Biol. Interact. 86, 129–162.
(23) Driscoll, J. P., Kornecki, K., Wolkowski, J. P., Chupak, L., Kalgutkar, A. S., and
O’Donnell, J. P. (2007) Bioactivation of phencyclidine in rat and human liver microsomes and
ACS Paragon Plus Environment
25

Chemical Research in Toxicology
Page 26 of 26
1
2
3
4
recombinant P450 2B enzymes: evidence for the formation of a novel quinone methide
5
6
intermediate. Chem. Res. Toxicol. 20, 1488–97.
7
8
9
(24) Thompson, J. A., Carlson, T. J., Sun, Y., DwyerꢀNield, L. D., and Malkinson, A. M. (2001)
Studies using structural analogs and inbred strain differences to support a role for quinone
methide metabolites of butylated hydroxytoluene (BHT) in mouse lung tumor promotion.
Toxicology 160, 197–205.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(25) Toteva, M. M., and Richard, J. P. (2011) The generation and reactions of quinone methides.
Adv. Phys. Org. Chem. Elsevier Inc. 45, 39ꢀ91
(26) Thompson, D. C., Perera, K., Krol, E. S., and Bolton, J. L. (1995) OꢀMethoxyꢀ4ꢀ
alkylphenols That Form Quinone Methides of Intermediate Reactivity Are the Most Toxic in Rat
Liver Slices. Chem. Res. Toxicol. 8, 323–327.
(27) Woodhead, J. L., Yan, K., Siler, S. Q., Watkins, P. B., Brouwer, K. L. R., Barton, H. A., and
Howell, B. A. (2014) Exploring BSEP inhibitionꢀmediated toxicity with a mechanistic model of
drugꢀinduced liver injury. Front. Pharmacol. 5, 1ꢀ11
ACS Paragon Plus Environment
26