This article was submitted to Analytical Chemistry, a section of the journal Frontiers in Chemistry
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Acetaminophen (APAP) is a mild analgesic and antipyretic used commonly worldwide. Although considered a safe and effective over-the-counter medication, it is also the leading cause of drug-induced acute liver failure. Its hepatotoxicity has been linked to the covalent binding of its reactive metabolite,
Drugs are generally metabolized by the liver into biologically inactive forms and eliminated from the body through bile and urine. However during these processes, they can also be bioactivated by hepatic enzymes into reactive electrophilic intermediates and subsequently react with nucleophilic sites of proteins to form covalent adducts (
Acetaminophen (
In this study, liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS) was employed to identify
Urea was purchased from BioRad (Mississauga, ON, Canada). APAP, trypsin (TPCK-treated, from bovine pancreas), sodium
Four Sprague-Dawley male rats (450−550 g) were dosed with 600 mg/kg APAP (IP; solubilized in 60% PEG 200) or two animals were treated with vehicle for control samples. Rat livers were collected after 24 h post dosing. Male C57BL/6 mice (27−35 g) were treated (IP, in saline) with 150 mg/kg (2 and 6 h) and 300 mg/kg (2 and 6 h), as well as with vehicle for control samples. Two mice were treated at each dose and timepoint. All experiments were performed at
Frozen liver samples, stored at −80°C, were homogenized in 100 mM ammonium bicarbonate (ABC buffer, pH 8–8.5) at 5 ml/g tissue weight using a hand-held homogenizer (Tissuemiser; Thermo Fisher Scientific, Waltham, MA, United States), and 100 μl aliquots were combined with 25 μl of 0.1% SDS, and heated for 5 min at 95°C. After cooling samples, 50 μl of a solution of 7 M urea and 2 M thiourea in water was added, and a probe sonicator was used ((XL-2000; Qsonica, Newtown, CT, United States) for three cycles of 10 s, followed by another 15 min in an ultrasonic bath (Branson 2510; Branson Ultrasonics, Brookfield, CT, United States). The resulting protein extract was then diluted with 500 μl of 100 mM ABC buffer. Reductive alkylation was performed by adding DTT (30 μl, 100 mM; 37°C for 20 min), followed by IAM (45 μl, 100 mM; 37°C for 30 min in the dark). Similarly, control samples, from vehicle-treated animals, were alkylated using HP-IAM (by simply substituting the IAM mentioned above). Samples were then digested by trypsin (30 μl, 1 mg/ml; 37°C for 18 h), followed by the addition of 300 μl of 2% formic acid prior to fractionation by solid phase extraction (SPE) on OASIS MCX cartridges (1 cc, 30 mg; Waters, Milford, MA, United States). Loaded cartridges were washed in three steps with 2% formic acid in water, 100% MeOH and 50% MeOH (1 ml each). Eight fractions were collected by eluting with 15, 20, 25, 35, 50, and 200 mM ammonium acetate in 50% MeOH, then 0.1 and 3% ammonium hydroxide in 50% MeOH (1 ml each). Fractions were dried under vacuum and stored at −30°C until analysis.
SPE fractions were reconstituted in 100 µl 10% acetonitrile and injected (20 μl) onto an Aeris PEPTIDE XB
Data-dependent experiments were performed to collect MS and MS/MS spectra on a high-resolution quadrupole-time-of-flight TripleTOF 5600 mass spectrometer (Sciex, Concord, ON, Canada) equipped with a DuoSpray ion source in positive electrospray mode set at 5 kV source voltage, 500°C source temperature and 50 psi GS1/GS2 gas flows, with a declustering potential of 80 V. The instrument performed a survey TOF-MS acquisition from
Targeted assays were developed using scheduled LC-MRM on a Sciex QTRAP 5500 hybrid quadrupole-linear ion trap system with a TurboIonSpray ion source in positive mode, with identical UHPLC conditions as for data-dependent high resolution experiments. Source parameters were as follows: ionspray voltage 5 kV; temperature 550°C; GS1 and GS2 50 psi; and curtain gas 35 psi. Declustering, entrance and collision cell exit potentials were set at 80, 10 and 13 V, respectively. Collision-induced dissociation was performed at a collision energy of 30 V. Scheduled MRM time windows were set at 240 s, with a targeted scan time at 1.25 s. Minimum and maximum dwell times were 10 and 250 ms, respectively. MRM transitions monitored for SPE fractions two to eight can be found in
Raw data files from quadrupole-time-of-flight experiments were searched against the rat or mouse UniProtKB/Swiss-Prot protein database (release date: 07/18/2018, including common protein contaminants) using Sciex ProteinPilot 5.0. To detect APAP covalent adducts, a custom modification was added to the Paragon algorithm (
LC-MRM peaks were integrated and verified using MultiQuant 3.0.2 (Sciex) to ensure that APAP-modified peptides matched with
APAP-modified peptides with multiple cysteines, containing at least one CAM-cysteine, did not meet the second (or third) criteria, due to no standard samples available for comparison (all cysteines were HP-CAM modified in reference samples). Therefore, confirmation for these modified peptides incorporated into the MRM method was achieved by the absence of corresponding signals in HP-CAM reference samples.
The goal of this study was to assess the applicability of a targeted LC-MS/MS method for
Four liver digests from each species were initially subjected to untargeted high-resolution MS/MS to identify potentially novel target proteins. In total, 15 APAP-modified mouse peptides (from 14 target proteins) were identified (with over 95% peptide confidence). These included five peptides with more than one cysteine, one being APAP-bound and the other(s) carbamidomethylated, however the exact location of the APAP modification was found in each case, based on unique y and b-ions in their high-resolution MS/MS spectra. A specifically challenging example was seen for peptides KPIGLC174C175IAPVLAAK and C174C175IAPVLAAK from glutamine amidotransferase-like class 1 domain-containing protein 3A (and its rat ES1 protein homolog) (
APAP-modified peptides in rat and mouse liver identified by data-dependent high-resolution MS/MS.
Protein ID | Protein Name | Sequence | Conf. | Obs. |
z |
---|---|---|---|---|---|
Q9CRB3|HIUH_MOUSE | 5-hydroxyisourate hydrolase | LSRLEAPC |
99 | 670.3344 | 3 |
Q8CDE2|CALI_MOUSE | Calicin | IHC |
99 | 626.3102 | 2 |
P16015|CAH3_MOUSE | Carbonic anhydrase 3 | EAPFTHFDPSCLFPAC |
99 | 715.3168 | 3 |
Q64458|CP2C29_MOUSE | Cytochrome P450 2C29 | FIDLLPTSLPHAVTC |
99 | 726.7081 | 3 |
Q5MPP0|FA2H_MOUSE | Fatty acid 2-hydroxylase | LAAGAC |
99 | 1095.5390 | 1 |
Q9D172|GAL3A_MOUSE | Glutamine amidotransferase-like class 1 domain-containing protein 3A, mitochondrial | CC |
99 | 597.8044 | 2 |
KPIGLCC |
99 | 568.3172 | 3 | ||
P15105|GLNA_MOUSE | Glutamine synthetase | C |
99 | 389.2080 | 4 |
P06151|LDHA_MOUSE | L-lactate dehydrogenase A chain | IVSSKDYC |
99 | 555.2719 | 3 |
Q9CXT8|MPPB_MOUSE | Mitochondrial-processing peptidase subunit beta | IVLAAAGGVC |
99 | 690.7044 | 3 |
Q6PG95|CRML_MOUSE | Protein, cramped-like | KSSQELYGLIC |
98 | 1004.4960 | 2 |
Q99LX0|PARK7_MOUSE | Protein/nucleic acid deglycase DJ-1 | GLIAAIC |
99 | 806.7496 | 3 |
P24549|AL1A1_MOUSE | Retinal dehydrogenase 1 | YC |
99 | 531.7211 | 2 |
Q63836|SBP2_MOUSE | Selenium-binding protein 2 | C |
99 | 756.8463 | 2 |
P17751|TPIS_MOUSE | Triosephosphate isomerase | C |
99 | 914.6931 | 4 |
P23457|DIDH_RAT | 3-alpha-hydroxysteroid dehydrogenase | SIGVSNFNC |
99 | 623.2875 | 2 |
P21775|THIKA_RAT | 3-ketoacyl-CoA thiolase A, peroxisomal | QC |
99 | 598.6364 | 3 |
P16638|ACLY_RAT | ATP-citrate synthase | YIC |
99 | 709.8419 | 2 |
Q03248|BUP1_RAT | Beta-ureidopropionase | C |
99 | 432.7230 | 2 |
P05179|CP2C7_RAT | Cytochrome P450 2C7 | FINFVPTNLPHAVTC |
99 | 726.7081 | 3 |
P07153|RPN1_RAT | Dolichyl-diphosphooligosaccharide--protein glycosyltransferase subunit 1 | VAC |
99 | 612.6727 | 3 |
P56571|ES1_RAT | ES1 protein homolog, mitochondrial | CC |
99 | 597.8040 | 2 |
P49889/90, P52844| | Estrogen sulfotransferase (Ste2, isoforms 1/3) | NNPC |
99 | 745.0038 | 3 |
STIE1/2/3_RAT | |||||
O88618|FTCD_RAT | Formimidoyltransferase-cyclodeaminase | TC |
99 | 570.2795 | 2 |
Q58FK9|KAT3_RAT | Kynurenine-oxoglutarate transaminase 3 | ALSC |
96 | 502.2493 | 2 |
P57113|MAAI_RAT | Maleylacetoacetate isomerase | ALLALEAFQVSHPC |
99 | 601.9792 | 3 |
O88767|PARK7_RAT | Protein/nucleic acid deglycase DJ-1 | GLIAAIC |
99 | 806.7489 | 3 |
P17988|ST1A1_RAT | Sulfotransferase 1A1 | MKENC |
99 | 813.8645 | 4 |
P11232|THIO_RAT | Thioredoxin | C |
99 | 657.2881 | 2 |
P48500|TPIS_RAT | Triosephosphate isomerase | C |
99 | 777.8861 | 2 |
C |
99 | 823.8993 | 2 | ||
P09118|URIC_RAT | Uricase | SIETFAMNIC |
99 | 677.0681 | 4 |
APAP modification site.
Control liver samples were alkylated with hydroxyphenyl iodoacetamide (HP-IAM) prior to trypsin digestion, SPE fractionation and data-dependent LC-MS/MS analysis (
Reference samples
Scheduled MRM methods were built, with three transitions per peptide, and separated based on species and SPE fraction. MRM transitions of modified peptides were also monitored in adjacent fractions as those where they were detected in HP-CAM samples, since peptides can often elute in more than one fraction. In the case of two peptides from mouse glutamine amidotransferase and rat ES1 protein homolog, where one Cys was APAP-modified and the other was carbamidomethylated, retention times and fragment ions from data-dependent MS/MS were used to build the MRM method. Unfortunately, several other peptides with multiple cysteines were not monitored by MRM as their signals were not confirmed when developing the targeted method.
APAP-treated and HP-CAM alkylated liver digests were analyzed by scheduled LC-MRM methods specific to each SPE fraction. LC-MRM peaks were integrated, and relative peak areas for each transition (transition/sum of all transitions) as well as retention times were compared with HP-CAM signals for the same peptide.
Representative LC-MRM chromatograms of
From eight APAP-treated mouse livers (two animal each at 150 and 300 mg/kg, either 2 and 6 h post dose), LC-MRM confirmed 13 distinct APAP-modified sites from 13 different proteins. From
APAP-modified peptides in mouse liver identified
Acc. # | Protein name | Cys site | Peptide sequence |
z | # Of hits ( |
|||
---|---|---|---|---|---|---|---|---|
150/2 h | 300/2 h | 150/6 h | 300/6 h | |||||
Q91V92 | ATP-citrate synthase | C20 | YIC |
2 | 2 | 2 | 2 | 2 |
O35490 | Betaine-homocysteine |
C131 | QVADEGDALVAGGVSQTPSYLSC |
3 | 0 | 1 | 0 | 2 |
Q64458 | Cytochrome P450 2C29 | C372 |
|
3 | 2 | 2 | 2 | 2 |
Q9D172 | Glutamine amidotransferase-like 1 domain-containing protein 3A, mitochondrial | C175 |
|
3 | 2 | 2 | 2 | 1 |
Q9D172 | Glutamine amidotransferase-like 1 domain-containing protein 3A, mitochondrial | C175 |
|
2 | 2 | 2 | 2 | 2 |
P15105 | Glutamine synthetase | C269 |
|
2 | 1 | 2 | 2 | 2 |
P16858 | Glyceraldehyde-3-phosphate dehydrogenase | C22 | AAIC |
2 | 0 | 0 | 0 | 2 |
P17563 | Methanethiol oxidase | C8 | C |
2 | 0 | 2 | 1 | 2 |
Q91VS7 | Microsomal glutathione |
C50 | VFANPEDC |
2 | 0 | 1 | 0 | 0 |
Q9CXT8 | Mitochondrial-processing peptidase, beta | C248 |
|
3 | 2 | 2 | 2 | 2 |
P24549 | Retinal dehydrogenase 1 | C133 | YC |
2 | 1 | 0 | 0 | 1 |
Q63836 | Selenium-binding protein 2 | C8 |
|
2 | 0 | 2 | 0 | 1 |
Q64442 | Sorbitol dehydrogenase | C106 | EVDEYC |
2 | 0 | 0 | 0 | 1 |
P10639 | Thioredoxin | C73 | C |
2 | 0 | 2 | 2 | 1 |
APAP modification site. Underlined peptides (or modification sites) were also identified in data-dependent MS/MS experiments.
Identified based on an absent signal in the reference (HP-CAM) sample.
From the analysis of four rat liver samples, 23 distinct APAP-peptides were confirmed, from 21 different proteins (
Identified APAP-modified peptides in rat liver
Acc. # | Protein name | Cys site | Peptide |
z | # Of hits |
---|---|---|---|---|---|
600 mg/kg ( |
|||||
P23457 | 3-α-hydroxysteroid dehydrogenase | C170 |
|
2 | 4 |
P16638 | ATP-citrate synthase | C20 |
|
2 | 4 |
O09171 | Betaine-homocysteine |
C131 | QVADEGDALVAGGVSQTPSYLSC |
3 | 4 |
P05178 | Cytochrome P450 2C6 | C372 | FIDLIPTNLPHAVTC |
3 | 1 |
P05179 | Cytochrome P450 2C7 | C372 |
|
3 | 4 |
P36365 | Dimethylaniline monooxygenase [N-oxide-forming] 1 | C35 |
|
2 | 4 |
P07153 | Dolichyl-diphosphooligosaccharide protein glycosyltransferase subunit 1 | C475 |
|
3 | 4 |
P49889/90 | Estrogen sulfotransferase 1/2/3 | C237 |
|
3 | 3 |
P52844 | |||||
P56571 | ES1 protein homolog, mitochondrial | C175 |
|
2 | 4 |
P56571 | ES1 protein homolog, mitochondrial | C175 | KPIGLC(CAM)C |
3 | 4 |
O88618 | Formimidoyltransferase-cyclodeaminase | C438 |
|
2 | 3 |
P09606 | Glutamine synthetase | C269 | C |
2 | 4 |
P57113 | Maleylacetoacetate isomerase | C205 |
|
3 | 4 |
Q8VIF7 | Methanethiol oxidase | C371 | GGSVQVLEDQELTC |
3 | 1 |
Q8VIF7 | Methanethiol oxidase | C8 | C |
2 | 4 |
P08011 | Microsomal glutathione |
C50 | VFANPEDC |
2 | 2 |
Q03346 | Mitochondrial-processing peptidase subunit beta | C248 | IVLAAAGGVC |
3 | 3 |
Q63716 | Peroxiredoxin-1 | C173 | HGEVC |
4 | 4 |
P11598 | Protein disulfide-isomerase A3 | C244 | FIQESIFGLC |
3 | 3 |
P17988 | Sulfotransferase 1A1 | C232 |
|
4 | 1 |
P48500 | Triosephosphate isomerase | C21/27 |
|
2 | 4 |
P11232 | Thioredoxin | C73 |
|
2 | 4 |
P09118 | Uricase | C95 |
|
4 | 2 |
APAP modification site. Underlined peptides were also identified in data-dependent MS/MS experiments (awith inclusion list).
All APAP-modified peptides detected
Targeted LC-MS/MS workflows were specifically designed to utilize the high duty cycle and low limit of detection of multiple reaction monitoring (MRM). The detection of modified peptides was optimized using scheduled MRM methods built for each SPE fraction for a given species separately. By using HP-IAM in the reference samples, it was possible to mimic APAP-derived covalent modification on cysteines in the final protein digest. HP-CAM peptides were used to confirm which fractions to monitor for each modified peptide, as well as LC retention time and MS/MS fragmentation behavior, with relatively high signal intensity. Whenever possible, peptide candidates in one species were translated to the closest protein ortholog in the other species. Modified peptides not amenable to MRM transition development were omitted in the final scheduled MRM method, based on giving an appropriate signal for three MRM transitions in reference samples.
In general, targeted LC-MRM analyses showed far superior detectability of APAP-modified peptides in these samples. An important advantage that MRM detection has over data-dependent MS/MS, is that each modified peptide signal is continuously monitored, instead of depending on the automatic selection of a precursor of interest for MS/MS acquisition, which leads unfortunately to much less reproducible data as well as the loss of low-abundant peptides in highly complex samples. For this reason, many modified peptides would not have a chance of being identified with conventional untargeted bottom-up proteomics workflows. An important caveat to the MRM method, however, is that method development is more time-consuming, and in the case of APAP-modified peptides for this study, the custom alkylation HP-CAM peptides were crucial for confirmation. Also, MRM sensitivity depends highly on the fragmentation behavior of the ionized peptide. Certain highly charged or large peptides necessitate more optimization by changing collision energies and selected multiple fragment ions, and without an appropriate standard, this is impossible. High-resolution DDA experiments are more flexible, for example for those with multiple cysteines with only one being modified by APAP. A unique opportunity was afforded to specifically design LC-MRM assays for confirming protein targets by the possibility of having a positional isomer of APAP quantitatively modifying all cysteines in the liver homogenates. Without this standard as a reference, these hits may have been detected but not as easily confirmed.
The aim of this study was to identify protein targets of APAP’s reactive metabolite in mouse and rat liver to help better understand the mechanisms of APAP-induced hepatotoxicity. Using a combination of high-resolution data-dependent MS/MS and scheduled MRM experiments, a multitude of targets have been found for the first time, as well as confirming others previously reported targets of APAP from literature. It was possible to compare protein targets and modification sites in mouse and rat livers.
Summary of APAP protein targets confirmed in rat and mouse liver samples
From the 10 protein targets confirmed in rat only, four of these did not have corresponding cysteine residues in the mouse ortholog, namely two sulfotransferases, dimethylaniline monooxygenase and maleylacetoacetate isomerase. One other protein, 3-alpha hydroxysteroid dehydrogenase (3α-HSD), does not have a corresponding ortholog in mouse at all. In the case of the three targets found uniquely in mouse, one protein has no corresponding cysteine in the rat protein (GAPDH). Retinal dehydrogenase, though the same peptide was monitored for both species, was only found in two of the eight mouse samples. As for sorbitol dehydrogenase, the corresponding ortholog peptide was not incorporated into the MRM method, since it was not found in the rat HP-CAM reference sample and thus could not be optimized in the targeted method.
One example of an APAP target found in rat, with no corresponding cysteine in the mouse ortholog, was estrogen sulfotransferase (EST), a cytosolic enzyme that inhibits estrogen activity by conjugating a sulfonate group to it (
3α-HSD catalyzes NAD(P)+-dependent oxidoreduction of steroids and dihydrodiols. Multiple forms of 3α-HSD have been identified in rat liver with a role in xenobiotic metabolism and intracellular transport of bile acids (
All 11 protein groups confirmed in both species (
Rat ES1 protein homolog and mouse glutamine amidotransferase-like class 1 domain-containing protein 3A are orthologous mitochondrial proteins confirmed as modified in both species from these results. ES1 was found to be elevated in Down syndrome, potentially as an antioxidant response to increased mitochondrial ROS production (
Two protein targets found previously by our group
Other protein targets common to both species included betaine-homocysteine
Thioredoxin is involved in redox signalling through oxidation of its thiols. The APAP modification site was confirmed on Cys73, the only available free thiol in fully oxidized THX, which is also known to serve as a donor for nitrosylation of proteins under NO stress (
Reactive metabolites can covalently bind to proteins causing downstream immune reactions and/or cell damage (
An analytical workflow was developed and applied to rat and mouse liver homogenates to investigate protein covalent binding following APAP administration. Using a combination of high-resolution MS/MS and scheduled MRM assays for proteomic analyses and a custom alkylation reagent, many
MRM transitions and retention time settings used for LC-sMRM analyses of individual SPE fractions (two to eight) from rat and mouse digests (
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: ProteomeXchange PRIDE repository, accession no: PXD027674.
The animal study was reviewed and approved by INRS Centre National de Biologie Expérimentale under the ethical practices of the Canadian Council on Animal Care (project UQLK.14.02).
TG, MG, and LS conceived the research. TG, MG, and GM carried out sample preparation, and analyses. All authors contributed to data processing. TG, AS, and LS were involved in the preparation of tables and figures for the manuscript. TG and LS were the main contributors to the writing of the final manuscript. All authors made substantial, direct and intellectual contribution to the work, and revised the manuscript.
Financial support for this study was provided by the Natural Sciences and Engineering Research Council of Canada (Discovery Grant # RGPIN 2016-06034) and a UQAM Institutional Research Chair in Bioanalytical Chemistry for LS. We would like to recognize funding for infrastructure from the Canadian Foundation for Innovation for LC-MS/MS instrumentation used in this study. We also acknowledge support to our mass spectrometry platform from CERMO-FC (Centre d’excellence de recherche sur les maladies orphelines – Fondation Courtois)
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The Supplementary Material for this article can be found online at:
ACN, acetonitrile; APAP, acetaminophen; BHMT, betaine—homocysteine S-methyltransferase 1; CA3, carbonic anhydrase 3; CAM, carbamidomethylation; CPS1, mitochondrial carbamoyl-phosphate synthase [ammonia]; DTT, dithiothreitol; ER, endoplasmic reticulum; EST, estrogen sulfotransferase; GS, glutamine synthetase; HDAg, hepatitis delta antigen; HP-CAM, N-(4-hydroxyphenyl)-2-carbamidomethylation; HP-IAM, N-(4-hydroxyphenyl)-2-iodoacetamide; 3α-HSD, 3-alpha-hydroxysteroid dehydrogenase; IAM, iodoacetamide; IDA, information-dependent acquisition; IP, intraperitoneal; LC-MS/MS, liquid chromatography-tandem mass spectrometry; MeOH, methanol; MGST1, microsomal glutathione S-transferase 1; MRM, multiple reaction monitoring; MTO, methanethiol oxidase; NAPQI, N-acetyl p-benzoquinone imine; PMPCB, mitochondrial-processing peptidase subunit beta; SDS, n-dodecyl sulfate; S/N, signal-to-noise; SPE, solid-phase extraction; TPDB, Target Protein Database; TPI, triosephosphate isomerase; TXN, thioredoxin.