Developing a GC-EI-MS/MS Method for Quantifying Warfarin and Five Hydroxylated Metabolites Generated by the Fenton Reaction.

doi:10.21203/rs.3.rs-3212519/v1

Download PDF

Research Article

Developing a GC-EI-MS/MS Method for Quantifying Warfarin and Five Hydroxylated Metabolites Generated by the Fenton Reaction.

https://doi.org/10.21203/rs.3.rs-3212519/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 08 Feb, 2024

Read the published version in Environmental Science and Pollution Research →

You are reading this latest preprint version

As a drug and rodenticide, warfarin is used globally. Extensive research has shown that warfarin and other rodenticides are recoverable in the environment and food chain (Fernandez, Santos, Cancela, Laize, & Gavaia, 2014; Gomez-Canela, Barata, & Lacorte, 2014; Nakayama, Morita, Ikenaka, Mizukawa, & Ishizuka, 2019; Rattner, Lazarus, Elliott, Shore, & van den Brink, 2014; Saito-Shida, Nemoto, Matsuda, & Akiyama, 2016; Waddell, Poppenga, & Drobatz, 2013). Current methods for their analytical determination focus on the target component rather than on metabolites or transformation products. Warfarin's phase 1 metabolism is extensively studied. Mono-hydroxylated metabolites are primarily formed (Watanabe et al., 2015; Wong & Davis, 1989). In this study, the use of the Fenton reaction aimed to generate mono-hydroxylated transformation products in the laboratory. A GC-MS/MS method was designed to quantify the expected derivatized metabolites by multiple reaction monitoring. Warfarin, 4'-, 5-, 6-, 7-, and 8-hydroxy warfarin, and isotopically labeled reference compounds were used for this approach. The method has a linear working range of 30 to 1800 pg/µL. Detection limits obtained range between 18.7 to 67.0 pg/µL. In addition, a C18-SPE step was used to enrich the analytes, and the recovery for each compound was calculated. The derivatization yield of warfarin for in-liner derivatization with m-TFPTAH was determined as well. The method was used successfully to show that 4'-O-Me-WAR is preferentially formed under the given experimental conditions. Furthermore, the Fenton reaction has generated all preselected mono-hydroxylated transformation products. The maximum intensities of all target analytes were reached at around 4 minutes and immediately decreased after that. The applicability of GC-MS for determining known metabolites of warfarin was demonstrated. With SPE clean-up, the method is also suitable for detecting metabolites in more complex matrices, for example, environmental water samples in the future.

Fenton reaction

Multiple reaction monitoring

warfarin

hydroxy warfarin

gas chromatography

quantification.

Link et al. synthesized the vitamin K epoxide reductase inhibitor warfarin (WAR) based on their findings of the sweet clover disease (Campbell & Link, 1941; Link, 1959). WAR interacts with the vitamin K-depending activation of several blood clotting precursors in the blood coagulation cascade (Rishavy, Usubalieva, Hallgren, & Berkner, 2011; Stafford, 2005). The anticoagulant therapeutic WAR is still a commonly administered drug in preventing pulmonary embolism, thrombosis, atrial fibrosis, and fibrillation (Link, 1959; Wardrop & Keeling, 2008). The global market size of anticoagulants is estimated to increase to $ 45.50 billion by 2026, according to Fortune Business Insights' report (Anticoagulants Market Size, Share & Industry Analysis, […] and Regional Forecast, 2019–2026.)(Insights). That reflects the therapeutic market's uptake in usage but not the biocide market. Warfarin remains in excrement, sewage water, and wastewater treatment plants when administered. The application of rat poison in the form of bait is of particular concern to the environment. Consumed from non-target animals via deceased target species and from bait systems, warfarin releases into the environment (J. Regnery et al., 2020). Numerous publications exist on warfarin and other anticoagulant rodenticides in targeted and untargeted wildlife species and domesticated animals (Nakayama et al., 2019; Rattner et al., 2014; Waddell et al., 2013). ARs have been detected in agricultural products(Saito-Shida et al., 2016) and wastewater (Fernandez et al., 2014; Gomez-Canela et al., 2014) and aquatic environments (Julia Regnery et al., 2018; J. Regnery et al., 2019). Indeed, no method takes advantage of the known metabolites of warfarin for its analytical detection. Pharmacologically, warfarin is primarily eliminated via the hepatic phase 1 metabolism and urinary excretion. In humans, cytochrome P450 (CYP) enzymes transform warfarin stereoselectively into enantiomers of 6-, 7-, and 8-hydroxy warfarin (Kaminsky & Zhang, 1997). Stereoisomers of dehydro-warfarin, 4’-hydroxy warfarin, and 10-hydroxy warfarin also yield from this biotransformation and warfarin alcohol (Kaminsky & Zhang, 1997). These metabolites are not unique to mammalian species (Watanabe et al., 2015). Warfarin's metabolism in rodents, birds, and other species is well documented (Kaminsky & Zhang, 1997; Rettie et al., 1992; Saengtienchai, Ikenaka, Watanabe, Ishida, & Ishizuka, 2011; Watanabe et al., 2015). CYPs generally mediate reactions via hydrogen atom transfer (HAT) and single electron transfer (SET) (Meunier, de Visser, & Shaik, 2004). Nevertheless, the laboratory's execution of oxidative and reductive electron transfer reactions can mimic these biological processes - eighter directly by SET or indirectly by the in-situ generation of reactive oxygen species (ROS) such as hydroxyl radicals (·OH)(Johansson, Weidolf, & Jurva, 2007). One prominent example of organic matter and ROS reaction is the Fenton reaction, which simulates the hepatic phase 1 metabolism. Furthermore, it is an effective advanced oxidation process (AOP) with applications in wastewater treatment (Nidheesh & Gandhimathi, 2012; Zhang, Dong, Zhao, Wang, & Meng, 2019). The reaction can lead to complete mineralization by oxidation of organic matter (Nidheesh & Gandhimathi, 2012; Zhang et al., 2019). The "oxidation of tartaric acid in the presence of iron" was the first demonstration of the reaction by H. J. H. Fenton in 1894 (Fenton, 1894). Fenton's reagent is the reaction between hydrogen peroxide (H₂O₂) and ferrous iron in an acidic solution (Brillas, Sires, & Oturan, 2009). The pH optimum is about 2.8 and 3 (Brillas et al., 2009). During the reaction, Fe²⁺ is oxidized to Fe^3+, and hydrogen peroxide reacts to a hydroxide anion (OH⁻) and a hydroxyl radical. These radicals can react further with organic compounds and may lead to complete mineralization. During Fenton's reaction, various radical species may form, which can react with hydrogen peroxide generating hydroxylated transformation products. Further termination reactions are, for example, the generation of dimers and the formation of cations or anions (Collin, 2019). Since hydroxylated products are expected to be formed during the reaction of warfarin and Fenton's reagent, their detection and quantification are of particular interest regarding monitoring environmental samples and waste-water-treatment. A few GC-based analytical methods from the nineteen hundreds showed the quantification of warfarin and its known metabolites. An overview is presented in the following paragraph.

Various publications describe the quantification of hydroxylated warfarin species using liquid chromatography coupled with different mass spectrometric techniques.(Kim et al., 2012; Spink, Aldous, & Kaminsky, 1989; Watanabe et al., 2015) Whereas only a few methods employ gas chromatographic techniques. Many of these publications date back to the 1970s to 1990s. In 1974, Kaiser et al.(Kaiser & Martin, 1974) and Midha et al.(Midha, McGilveray, & Cooper, 1974) reported the first quantification methods of warfarin in plasma utilizing gas chromatography. Kaiser et al. determined warfarin as a pentafluorobenzyl-derivative utilizing Tracor MT-220 GC-ECD. The authors state a lower detection limit of 0.02 µg/ml using coumachlor as an internal standard. Midha et al. report a 0.25 µg/mL sensitivity using GC-FID on methylated warfarin and phenylbutazone as an internal standard. 1978 Hanna et al.(Hanna, Rosen, Eisenberger, Rasero, & Lachman, 1978) obtained a similar detection limit of 0.3 µg/mL employing GC-FID. The authors used papaverine as an internal standard to measure underivatized warfarin in male plasma. In 1979, Duffield et al. quantified methylated warfarin and methylated warfarin alcohol using deuterated internal standards in plasma (A. M. Duffield, Duffield, Birkett, Kennedy, & Wade, 1979; P. H. Duffield, Birkett, Wade, & Duffield, 1979). The authors used Finnigan 3200 GCMS equipped with Finnigan diverter valve assembly and a glass column (5 ft, 2 mm ID) packed with 3% OV-1 on Gas Chrom Q (100/120 mesh) for chromatographic separation and EI-MS detection. In both publications, authors used different oven programs. The authors report a minimum concentration of 0.2 µg/mL of warfarin as quantification limit. Single ion monitoring (SIM) of warfarin was performed using m/z 323 and m/z 327 for its D₄-labeled standard and for warfarin alcohol m/z 293 was recorded as well as m/z 299 for the D₆-labeled standard. In 1983, Davies et al.(Davies, Bignall, & Roberts, 1983) published a quantitative determination of underivatized warfarin in plasma with a detection limit of 0.2 µg/mL. Samples were directly inserted into the EI source without chromatographic separation. In the same year, Bush et al.(Bush, Low, & Trager, 1983) determined a limit of detection of 1 ng/mL and a limit of quantification of 20 ng/mL for methylated warfarin and methylated 4'-, 6-, 7-, and 8-hydroxy warfarin in incubated microsomes. Quantification of the analytes was performed using deuterated internal standards. The authors utilized Hewlett-Packard 5700 gas chromatograph and 5985 EI-MS, and a DB-5 fused silica column (30 m) for analysis. In 1998, Maurer et al. (Maurer & Arlt, 1998) reported the detection of 4-hydroxy coumarins in urine using Hewlett-Packard 5890 gas chromatograph and 5970 EI mass spectrometer. The authors analyzed methylated acenocoumarol, coumatetralyl, phenprocoumon, warfarin, coumachlor, and pyranocoumarin with a detection limit for phenprocoumon of 25 ng/mL. 2008 Sato demonstrated a SIM quantification of warfarin, coumatetralyl, and coumachlor in human serum (Sato, 2005). Methylated warfarin was quantified by analyzing m/z 322, 279, and 91. The reported limit of determination is 20 ng/mL. Samples were subjected to GC-EI-MS in ethyl acetate after extraction, and derivatization was performed with trimethylsilyl diazomethane (TMS-DAM, 10%, v/v in hexane). Analysis of the serum samples was performed using Shimadzu GCMS-QP5050A at 70 eV, and a DB-5MS methyl silicone medium-bore capillary column (30 m × 0.25 mm ID, film thickness 0.25 µm) was utilized for chromatographic separation. The majority of gas chromatographic methods have been used for the determination of warfarin in plasma samples. However, since most of these methods do not correspond to contemporary technology, it is appropriate to propose a new adequate method.

Analytical grade warfarin (4-hydroxy-3-(3-oxo-1-phenyl butyl)chromen-2-one) (PESTANAL®, analytical standard, Fluka, Sigma-Aldrich), 4’-hydroxy warfarin (4-hydroxy-3-(1-(4-hydroxyphenyl)-3-oxobutyl)chromen-2-one) (Sigma-Aldrich), 6-hydroxy warfarin (4,6-dihydroxy-3-(3-oxo-1-phenyl butyl)chromen-2-one) (Chemos GmbH Co. KG, Altdorf, Germany), 7-hydroxy warfarin (4,7-dihydroxy-3-(3-oxo-1-phenyl butyl)chromen-2-one) (Chemos GmbH Co. KG, Altdorf, Germany), 8-hydroxy warfarin (4,8-dihydroxy-3-(3-oxo-1-phenyl butyl)chromen-2-one) (Chemos GmbH Co. KG, Altdorf, Germany) reference standards were used in this study for methods development. ASCA GmbH (Angewandte Synthesechemie Adlershof, Berlin, Germany) synthesized 5-hydroxy warfarin (4,5-dihydroxy-3-(3-oxo-1-phenyl butyl) chromen-2-one) and the isotopically labeled internal standards: phenyl ¹³C₆-isotopomers of 5-hydroxy warfarin, 6-hydroxy warfarin, Me¹³C-WAR (4-(¹³C)methoxy-3-(3-oxo-1-phenyl butyl)chromen-2-one), and 7-hydroxy warfarin. Phenyl D₅ warfarin (4-hydroxy-3-(3-oxo-1-(2,3,4,5,6-pentadeuteriophenyl) butyl)chromen-2-one) (Dr. Ehrenstorfer GmbH, Augsburg, Germany) was commercially available. For methods development and validation, standard stock solutions of all 4-hydroxy coumarins were prepared separately in MS-grade acetonitrile (Biosolve BV, Valkenswaard, The Netherlands) to obtain a final concentration of 0.1 µg/µL. Mixed stock solutions of the non-labeled reference standards were composed of the standard stock solutions and were diluted to a concentration of 2.0 ng/µL. Internal standard stock solutions of the isotopically labeled standards were prepared accordingly.

Degradation of warfarin undergoing the Fenton Reaction was performed in Erlenmeyer flasks in a sulfate buffer at pH 3.0 at room temperature. The initial concentration of warfarin was 1.25 µg/mL. A stock solution of ferrous sulfate heptahydrate (96 mg/mL) p. a. (CHEMSOLUTE®, Th. Geyer GmbH & Co. KG, Renningen, Germany) in sulfate buffer and hydrogen peroxide 30% (CHEMSOLUTE®) were added in a molar ratio of 1/1/50//WAR/Fe²⁺/H₂O₂ to start the reaction. Quenching after 2, 4, 6, 8, 16, 30, and 60 min results in independent samples, and adding 2 mL 0.1 m sodium hydroxide solution (J.T.Baker) per 100 mL buffer led to the precipitation of ferric salts. Additionally, radical species were quenched by adding 500 mg sodium thiosulfate (anhydrous, Merck KGaA, Darmstadt, Germany). One milliliter of the solution was taken for the quantification of warfarin. SPE was used to concentrate the samples in a 1:1 ratio. The remaining solution was also subjected to SPE to quantify hydroxylated species and concentrated in a 1:100 ratio.

Matrix separation and sample concentration were achieved using BAKERBOND™ C₁₈ solid phase extraction (SPE) cartridges (6 mL, 500 mg C₁₈-bound silica gel, 40 µm APD, 60 Å, J.T. Baker). SPE cartridges were washed with approximately 6 mL acetonitrile and preconditioned with the same amount of deionized H₂O. After adding the internal standard stock solution, the neutralized (pH 6–7) samples were loaded onto the cartridge's material and passed through by applying a vacuum. The SPE material was rinsed with approximately 6 mL of deionized water and left to dry. The analytes were eluted in 5–6 mL of acetonitrile and evaporated to dryness under a stream of nitrogen at 40°C. The recovered solid was redissolved in acetonitrile and prepared for GC-EI-MS/MS analysis.

For GC-EI-MS/MS measurements, 350 µL of the samples were mixed with 50 µL of the derivatization agent, m-TFPTAH, trimethyl-3-trifluoromethyl phenyl ammonium hydroxide 5% in methanol (abcr GmbH, Karlsruhe, Germany; and Alfa Aesar, Thermo Fischer GmbH, Kandel, Germany). The samples were subjected to Shimadzu's TQ8040 GC-MS/MS. One microlitre was injected by splitless injection. The injector temperature was set to 270°C for in-liner methylation of hydroxyl groups. A helium flow rate of 1 mL/min was used for chromatographic separation on a DB5 MS-UI 30m x 0.25 mm x 0.25 µm (Agilent Technologies) fused silica capillary column using the following oven program: 100°C → 280°C (22 K/min) → 320°C (7 K/min, 1.1 min hold 320°C). The transfer line was heated to 280°C, and the ion source temperature was set to 240°C. Ionization was achieved by standard electron energy of 70 eV and 5.0 µA emission current at 200°C, relative to the latest used tune file.

GC-MS/MS Methods Development

Optimization of the GC parameters achieved baseline separation of the analytes. In the precursor ion scan, specific target ions confirm the analytes' identity by collision-induced-dissociation. Figure 1A shows a total-ion-chromatogram of the reference compounds and isotopically-labeled reference compounds. Table 1 comprises the qualifier and quantifier ions, the optimized transitions, and the retention times of the analytes. The quantification of MeWAR was monitored using the transition of the M⁺-43 cation (m/z 279, C₁₈H₁₅O₃⁺) to m/z 128 (C₁₀H₈⁺). The same fragment with shifted m/z values was selected for the phenyl D₅-labelled internal standard (MeD₅WAR). 4’-O-Me-WAR was determined utilizing m/z 128, too. But with a different precursor ion (M⁺-43, m/z 309, C₁₉H₁₇O₄⁺). 7- and 5-O-Me-WAR were determined using the generation of m/z 227 (C₁₈H₁₃O₃⁺) from the M⁺-43 cation (m/z 309, C₁₉H₁₇O₄⁺). Corresponding m/z values of the phenyl ¹³C₆-labelled species are listed in Table 1. The transition of the radical cation (m/z 352, C₂₁H₂₀O₅^•+) to the M⁺-43 cation (m/z 309, C₁₉H₁₇O₄⁺) was used for the quantification of 6- and 8-O-Me-WAR. Because of the similar fragmentation behavior of 6-, 7-, and 8-O-Me-WAR, their qualifier, and quantifier ions have identical m/z values. Nonetheless, the analytes are uniquely identified based on their retention times and by ¹³C-labelled internal standards. After derivatization, the optimized MRM method ensures high specificity for detecting warfarin and hydroxylated species.

Table 1

**Overview of the qualifier and quantifier ions, retention times, optimized MRM transitions, and corresponding collision energies of all analytes, including the internal standards.** Bold numbers highlight the transition states used for quantification.
Analyte	Retention Time [min]	MRM-Transition			Internal Standard	MRM-Transition
Analyte	Retention Time [min]	ion transition	CE	ratio	Internal Standard	ion transition	CE	ratio
MeWAR	10,59	279.0 → 128.2	25	89.4	MeD₅WAR	284.0 → 129.2	25	100.0
		279.0 → 202.1	25	60.7		284.0 → 206.2	25	46.5
		91.0 → 65.1	20	100.0		96.0 → 68.1	20	70.0
4'-O-Me-WAR	12,21	309.0 → 128.2	25	100.0	MeD₅WAR
		201.0 → 128.2	20	83.2
		121.0 → 65.1	20	47.4
5-O-Me-WAR	11,71	309.0 → 277.1	15	100.0	¹³C₆ 5-Me-O-WAR	315.0 → 283.1	15	100.0
		277.1 → 234.1	20	69.6		283.0 → 240.1	20	71.4
		91.0 → 65.1	20	64.6		97.0 → 69.1	20	41.7
6-O-Me-WAR	11,98	309.0 → 277.1	15	62.3	¹³C₆ 6-Me-O-WAR	315.0 → 283.1	15	70.1
		352.0 → 309.0	15	72.5		358.0 → 315.1	15	100.0
		91.0 → 65.1	20	100.0		97.0 → 69.1	20	77.0
7-O-Me-WAR	12,39	309.0 → 277.1	15	46.8	¹³C₆ 7-Me-O-WAR	315.0 → 283.1	15	60.8
		352.0 → 309.1	15	59.0		358.0 → 315.1	15	79.0
		91.0 → 65.1	20	100.0		97.0 → 69.1	20	100.0
8-O-Me-WAR	11,87	309.0 → 277.1	15	60.6	¹³C₆ 6/7-Me-O-WAR
		352.0 → 309.1	15	74.6
		91.0 → 65.1	20	100.0

The mixed stock solution of non-labeled reference standards was used to prepare calibration levels with concentrations ranging from 30 to 1800 pg/µL. Due to the extended working range, 15 calibration points were chosen to ensure linearity. In the decade 30 to 300 pg/µL, ten calibration points were set in 30 pg intervals. From 300 to 1800 pg/µL, 5 points were selected with intervals of 300 pg each. The concentration of the IS in these samples was 200 pg/µL. Trimethyl-3-trifluoromethyl phenyl ammonium was added to the samples in a volumetric ratio of 1:7. Therefore, the linearity of the working range was determined in a final concentration range of 26 to 1543 pg/µL. The final concentration of the IS was 171.4 pg/µL. Data were processed by plotting the obtained peak areas relative to the corresponding peak area of the internal standard versus the analyte's concentration. After linear regression, slopes are significantly different from zero. Table 2 summarizes the obtained values. The ratios of the peak areas are directly proportional to the concentration of the analytes. The coefficients of determination (COD) of all analytes are shown in Table 2. Apart from 6-O-Me-WAR, all coefficients are higher than 0.9965. The coefficient of 6-O-Me-WAR is within the acceptance criteria at 0.9947. Warfarin shows the highest correlation with 0.9994. Data evaluation proves a linear relationship within the defined working range for all analytes.

Me¹³C-WAR and Me-WAR were analyzed in scan mode in a final concentration of 1 mM in acetonitrile. Integration of the TICs and ion traces at m/z 279 (WAR), and m/z 280 (Me¹³C-WAR) equal a mean derivatization yield of 33.2 ± 0.3%. However, determining the other compounds' derivatization yield is impossible in this study since the isotopic label is located within the phenyl ring and not the methyl group. The detection limits were determined using a 3-fold determination of the calibration standards in the lower working range. The linearity, the limit of quantification (LOQ), the lower limit of detection (LDL), and the limit of detection (LOD) were calculated using the DIN 32645:2008-11 standard of the German Institute for Standardization (DIN). Table 2 summarizes the found values. The lowest LoQ was determined at 60.7 pg/µL for 5-O-Me-WAR. Warfarin's LoQ is 81.5 pg/µL, and 4 '-, 7-, and 8-O-Me-WAR show a LOQ spanning from 110.4 to 147.1 pg/µL. 6-O-Me-WAR showed the highest LoQ at 218.2 pg/µL. The LODs of the analytes ranged from 18.7 to 67.0 pg/µL. Regarding warfarin, the calculated detection limits show similar dimensions as those provided by Sato et al. (Sato, 2005).

Table 2

Overview of the obtained slopes of the regression, y-intercept, and coefficients of determination of all analytes, including the detection limits calculated according to DIN 32645 (15 standard levels with three injections) and SPE recovery rates.
	the slope of the regression	y-intercept	R² (COD)	LOQ	LDL	LOD	SPE recovery rates
	line [pg/µl]	y-intercept	R² (COD)	[pg/µl]	[pg/µl]	[pg/µl]	absolute [%]	relative [%]
MeWAR	6.01·10^− 3 ± 3.9·10^− 5	0.0342 ± 0.0281	0.9994	81.5	48.7	24.4	97.1 ± 2.3	97.5 ± 1.3
4’-O-Me-WAR	6.14·10^− 3 ± 9.7·10^− 5	-0.1622 ± 0.0801	0.9965	125.3	76.2	38.1	90.4 ± 1.4	108.3 ± 0.9
5-O-Me-WAR	6.09·10^− 3 ± 7.9·10^− 5	0.0523 ± 0.0666	0.998	60.7	37.3	18.7	37.0 ± 8.5	75.0 ± 2.3
6-O-Me-WAR	4.39·10^− 3 ± 8.6·10^− 5	0.1630 ± 0.0670	0.9947	218.2	134.0	67.0	88.6 ± 4.0	94.3 ± 0.6
7-O-Me-WAR	5.27·10^− 3 ± 6.9·10^− 5	0.0439 ± 0.0535	0.9978	147.1	89.3	44.7	88.1 ± 1.3	102.6 ± 2.0
8-O-Me-WAR	4.20·10^− 3 ± 4.3·10^− 6	0.0340 ± 0.0334	0.9986	110.4	66.1	33.0	94.3 ± 8.4	99.6 ± 3.6

Six replicate samples of each compound were analyzed to examine the accuracy. For all analytes, a good accuracy of 98.7 ± 2.2% (MeWAR), 102.4 ± 3.2% (4’-Me-O-WAR), 101.0 ± 2.6% (5-O-Me-WAR), 97.5 ± 2.9% (6-O-Me-WAR), 99.7 ± 3.0% (7-O-Me-WAR), and 100.6 ± 2.8% (8-O-Me-WAR) was obtained. The SPE recovery rate of the analytes was determined in triplicates of 3 different calibration points at 257, 771, and 1286 pg/µL. The Samples were subjected to SPE and analyzed by GC-MS/MS. Reference compounds and the IS were transferred to a 5 mL volumetric flask, dried, and redissolved in 5 mL of neutralized (pH 6–7) sulfate buffer to determine the relative recovery. The absolute recovery rate was determined by adding the IS after SPE. Table 2 summarizes the mean values of the relative and absolute recovery rates comprising all chosen calibration points. The mean recoveries determined are all within the acceptance criteria at 80–120%. Only 5-O-Me-WAR shows a poor recovery. The relative recovery is about 75%, whereas the absolute recovery is 37%. Due to the hydroxyl groups' parallel arrangement and spatial proximity in 5-OH-WAR, it is well suited for chelating metal ions. Such chelating properties have been reported for flavonoids (Riha et al., 2014) as well as antioxidant and free radical scavenging activity regarding the coumarin moiety (Annunziata, Pinna, Dallavalle, Tamborini, & Pinto, 2020). Solubility factors likely contribute to the concentration of 5-OH-WAR as a hydrophilic metal-complex might result, which may not be adequately retained on the SPE-cartdridge's material. In general, it appears that the SPE method applied in this study is less suitable for 5-OH-WAR.

Aqueous Degradation of Warfarin

The Fenton reaction requires hydrogen peroxide and dissolved ferric salts. The influence of hydrogen peroxide on the system was examined in the early stages of the design of the experiment. Results proved that the buffer's pH value and the warfarin concentration do not change due to H₂O₂. Furthermore, test measurements showed that hydrogen peroxide does not lead to a degradation of Warfarin. After the reaction, sodium thiosulfate was used to quench radical species. In addition, 0.1 m sodium hydroxide solution was used to stop the Fenton reaction due to a shift to alkaline pH. The required amount was determined experimentally by observing the change in pH while adding different amounts of sodium hydroxide solution to the buffer. N. Wang et al. reported an equation calculating the minimum molar equivalents of hydrogen peroxide needed for complete degradation. It uses the sum formula following: C_aH_bN_cO_d (Wang, Zheng, Zhang, & Wang, 2016). Using the given Eq. (2a + 1/2b + 5/2c -d) on Warfarin's sum formula C₁₉H₁₆O₄, 42 results as molar equivalents needed for complete degradation. Accordingly, 1:50 defines the minimum molar equivalents of warfarin and hydrogen peroxide in the following experiments to achieve complete analyte degradation.

Degradation of aqueous WAR was achieved as described in the experimental section. The ratios of the peak areas of the analytes and their internal standards were determined to calculate the concentration of each analyte at distinct time points. Figure 1 (B – G) illustrates the results. Additionally, areas highlighted in light grey visualize the change in the analyte's concentration. Figure 1B shows the aqueous degradation of warfarin. The degradation was fitted and provided a COD of 0.9985. Complete degradation of WAR was achieved within the first 10 min. Figure 1C – G summarizes the calculated amount of hydroxylated TPs during the Fenton reaction. All targets were detected and showed a maximum concentration after approximately four minutes, followed by decreasing concentrations. As described in the experimental section, 1 mL of the independent samples was used to determine Me-WAR and 4’-O-Me-WAR simultaneously with the other analytes.

The concentration of Me-WAR and 4’-O-Me-WAR was determined in triplicates to ensure that these samples were representative. Samples were taken in a total volume of 1 mL at the indicated time points to observe their degradation behavior in separate samples. In the case of Me-WAR and 4’-O-Me-WAR, all the obtained data were merged and plotted with their standard deviations. Analysis of the other analytes was performed as described. Among the investigated hydroxylated TPs, 8-O-Me-WAR is quantitatively formed at the lowest level. The maximum intensity was recorded after approximately 8 minutes with ten ng/mL. The concentration subsequently decreased again. Regarding the other analytes, the maximum intensities were reached at around 4 minutes and immediately reduced after that. Approximately three times the amount of 8-O-Me-WAR was detected for 6-O-Me-WAR and 7-O-Me-WAR. For these analytes, the maxima are 33 ng/mL (6-O-Me-WAR) and 29 ng/mL (7-O-Me-WAR). It was possible to detect 49 ng/mL of 5-O-Me-WAR after a reaction time of 4 minutes. At 441 ng/mL, 4’-O-Me-WAR was generated with the highest quantity. To further relate the obtained data to one another, the percentages of the concentrations were calculated using their molar masses relative to the initial concentration of warfarin. WAR's initial concentration was reduced from 100–7.2% within the first four minutes. After one hour, 1.2% remained. The hydroxylated species were formed within the same period. After four minutes, the overall peak concentration was 38.8%, and 6.6% of the determined TPs were detected after one hour. 4’-OH-WAR was generated in much higher quantities than the other targeted TPs. Its maximum yield was 30.4%. After one hour, its concentration decreased to 3.7%. 5-, 6-, 7- and 8-OH-WAR have much lower yields than 4’-OH-WAR. The percentage of hydroxylated TPs at the coumarin moiety ranges from 8.3% after four minutes to 2.9% after one hour of reaction time. After 4 min, the peak concentration of these analytes was determined in the range of 3.4% (5-OH-WAR) and 0.7% (8-OH-WAR). After one hour, 1.1% (6-OH-WAR) and 0.4% (8-OH-WAR) were calculated. The formation of a steady-state concentration suggests that the reaction will potentially stop due to the consumption of hydrogen peroxide.

In 1968, Masayuki et al. found via thin-layer-chromatography that 6-OH-WAR and 8-OH-WAR are predominantly formed during the Fenton reaction, while the amount of 7-OH-WAR formed was relatively small (Masayuki, Ullrich, & Staudinger, 1968). Respectively to the used ratios of the reagents, it is shown that 4-OH-WAR is formed most preferably in this study. The 4' position of the phenyl ring is highly reactive to a radical attack at the activated and sterically accessible para-position of the phenyl ring. In addition, 5-, 6-, and 7-OH-WAR are formed similarly. 5-OH-WAR shows the highest intensity. That is followed by 6- and 7-OH-WAR. At the same time, 8-OH-WAR is formed only slightly as an intermediate. As the reaction proceeds, 5-, 7- and 8-OH-WAR are found in similar amounts. 6-OH-WAR appears to be marginally more stable. Although 4’-OH-WAR is most abundant after one hour, its abundance decreases by a power of ten within the first 15 minutes. It can therefore be assumed that this form continues to react rapidly. Here, double hydroxylated TPs are likely to occur, as in the case of all other TPs.

In Literature, trimethylsilyl diazomethane was often used for the methylation of warfarin and hydroxylated analogous. This study shows promising results for in-liner methylation with trimethyl-3-trifluoromethyl phenyl ammonium hydroxide. The derivatization yield of warfarin equaled 33.2 ± 0.3% and was calculated using the synthesized isotopically labeled Me¹³C-WAR. Detection and quantification of methylated warfarin species using the established GC-EI-MS/MS method, optimized oven program, and the set ion transitions prove sufficient selectivity. MeD₅WAR and phenyl ¹³C₆-labeled reference compounds were used as internal standards. All analytes show good linearity in the investigated 26 to 1543 ng/mL concentration range, and the calculated limits of detection range between 18.7 and 67.0 ng/mL. Except for 5-hydroxy warfarin, all other investigated species show good SPE recoveries. The method was successfully applied to quantify the decrease in warfarin undergoing the Fenton Reaction. Additionally, the formation and degradation of hydroxylated transformation products at a set molar ratio were successfully monitored. Due to the sterically accessible and highly reactive para-position of the phenyl ring, 4’-OH-WAR was preferably formed. The method described in this study is a suitable alternative to the commonly used LC-MS techniques and offers the determination of warfarin based on its mono-hydroxylated metabolites. Possible applications are the determination of warfarin and the described hydroxylated species in the more complex matrices, such as wastewater purification, biological, or environmental samples.

Funding

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Ethical Approval

Not applicable.

Acknowledgment

The Authors thank ASCA GmbH (Angewandte Synthesechemie Adlershof, Berlin, Germany) for synthesizing analytical standards and isotopically labeled reference compounds.

Author's Contributions and Consent to Publish

All authors contributed to the study's conception and design. Wipert Jannik von Törne and Urszula-Anna Klyk-Seitz performed material preparation, data collection, and analysis. Wipert Jannik von Törne wrote the first draft of the manuscript, and all authors commented on previous versions. Christian Piechotta is responsible for data analysis and experimental planning. He also took over the proofreading and structural design of the article. All authors read and approved the final manuscript and consent for publication.

Competing Interests

The authors have no relevant financial or non-financial interests to disclose.

Availability of data and materials

Not applicable.

Consent to Participate

Not applicable.

Annunziata, F., Pinna, C., Dallavalle, S., Tamborini, L., & Pinto, A. (2020). An Overview of Coumarin as a Versatile and Readily Accessible Scaffold with Broad-Ranging Biological Activities. Int J Mol Sci, 21(13), 4618. doi:10.3390/ijms21134618
Brillas, E., Sires, I., & Oturan, M. A. (2009). Electro-Fenton process and related electrochemical technologies based on Fenton's reaction chemistry. Chem Rev, 109(12), 6570-6631. doi:10.1021/cr900136g
Bush, E. D., Low, L. K., & Trager, W. F. (1983). A sensitive and specific stable isotope assay for warfarin and its metabolites. Biomed Mass Spectrom, 10(7), 395-398. doi:10.1002/bms.1200100702
Campbell, H. A., & Link, K. P. (1941). Studies on the Hemorrhagic Sweet Clover Disease IV. The Isolation and Crystallization of the Hemorrhagic Agent. Journal of Biological Chemistry, 138(1), 21-33. Retrieved from <<Go to ISI>://WOS:000187900800002>
Collin, F. (2019). Chemical Basis of Reactive Oxygen Species Reactivity and Involvement in Neurodegenerative Diseases. Int J Mol Sci, 20(10), 2407. doi:10.3390/ijms20102407
Davies, N. W., Bignall, J. C., & Roberts, M. S. (1983). Direct quantitative determinations by multiple metastable peak monitoring. 1--Warfarin in plasma. Biomed Mass Spectrom, 10(12), 646-651. doi:10.1002/bms.1200101205
Duffield, A. M., Duffield, P. H., Birkett, D. J., Kennedy, M., & Wade, D. N. (1979). Plasma quantitation of warfarin and warfarin alcohol by gas chromatography chemical ionization mass spectrometry in patients on warfarin maintenance therapy. Biomed Mass Spectrom, 6(5), 208-211. doi:10.1002/bms.1200060508
Duffield, P. H., Birkett, D. J., Wade, D. N., & Duffield, A. M. (1979). Quantitation of plasma warfarin levels by gas chromatography chemical ionization mass spectrometry. Biomed Mass Spectrom, 6(3), 101-104. doi:10.1002/bms.1200060304
Fenton, H. J. H. (1894). LXXIII.—Oxidation of tartaric acid in presence of iron. J. Chem. Soc., Trans., 65(0), 899-910. doi:10.1039/ct8946500899
Fernandez, I., Santos, A., Cancela, M. L., Laize, V., & Gavaia, P. J. (2014). Warfarin, a potential pollutant in aquatic environment acting through Pxr signaling pathway and gamma-glutamyl carboxylation of vitamin K-dependent proteins. Environ Pollut, 194, 86-95. doi:10.1016/j.envpol.2014.07.015
Gomez-Canela, C., Barata, C., & Lacorte, S. (2014). Occurrence, elimination, and risk of anticoagulant rodenticides and drugs during wastewater treatment. Environ Sci Pollut Res Int, 21(11), 7194-7203. doi:10.1007/s11356-014-2714-1
Hanna, S., Rosen, M., Eisenberger, P., Rasero, L., & Lachman, L. (1978). GLC determination of warfarin in human plasma. J Pharm Sci, 67(1), 84-86. doi:10.1002/jps.2600670121
Insights, F. B. Anticoagulants Market Size, Share & Industry Analysis, By Disease Indication (Pulmonary Embolism (PE), Deep Vein Thrombosis (DVT), Atrial Fibrillation, Heart Attacks, Others), By Route of Administration (Oral, Injectable), By Distribution Channel (Hospitals Pharmacies, Retail Pharmacies, Online Pharmacies, Others) and Regional Forecast, 2019-2026. (FBI101807). Retrieved from https://www.fortunebusinessinsights.com/industry-reports/anticoagulants-market-101807
Johansson, T., Weidolf, L., & Jurva, U. (2007). Mimicry of phase I drug metabolism--novel methods for metabolite characterization and synthesis. Rapid Commun Mass Spectrom, 21(14), 2323-2331. doi:10.1002/rcm.3077
Kaiser, D. G., & Martin, R. S. (1974). GLC determination of warfarin in human plasma. J Pharm Sci, 63(10), 1579-1581. doi:10.1002/jps.2600631018
Kaminsky, L. S., & Zhang, Z. Y. (1997). Human P450 metabolism of warfarin. Pharmacol Ther, 73(1), 67-74. doi:10.1016/s0163-7258(96)00140-4
Kim, S. Y., Kang, J. Y., Hartman, J. H., Park, S. H., Jones, D. R., Yun, C. H., . . . Miller, G. P. (2012). Metabolism of R- and S-warfarin by CYP2C19 into four hydroxywarfarins. Drug metabolism letters, 6(3), 157-164. doi:10.2174/1872312811206030002
Link, K. P. (1959). The Discovery of Dicumarol and Its Sequels. Circulation, 19, 97-107.
Masayuki, I., Ullrich, V., & Staudinger, H. (1968). Metabolism in vitro of warfarin by enzymic and nonenzymic systems. Biochemical Pharmacology, 17(8), 1663-1669. doi:10.1016/0006-2952(68)90228-1
Maurer, H. H., & Arlt, J. W. (1998). Detection of 4-hydroxycoumarin anticoagulants and their metabolites in urine as part of a systematic toxicological analysis procedure for acidic drugs and poisons by gas chromatography mass spectrometry after extractive methylation. Journal of Chromatography B, 714(2), 181-195. doi:Doi 10.1016/S0378-4347(98)00243-6
Meunier, B., de Visser, S. P., & Shaik, S. (2004). Mechanism of oxidation reactions catalyzed by cytochrome p450 enzymes. Chem Rev, 104(9), 3947-3980. doi:10.1021/cr020443g
Midha, K. K., McGilveray, I. J., & Cooper, J. K. (1974). GLC determination of plasma levels of warfarin. J Pharm Sci, 63(11), 1725-1729. doi:10.1002/jps.2600631112
Nakayama, S. M. M., Morita, A., Ikenaka, Y., Mizukawa, H., & Ishizuka, M. (2019). A review: poisoning by anticoagulant rodenticides in non-target animals globally. Journal of Veterinary Medical Science, 81(2), 298-313. doi:10.1292/jvms.17-0717
Nidheesh, P. V., & Gandhimathi, R. (2012). Trends in electro-Fenton process for water and wastewater treatment: An overview. Desalination, 299, 1-15. doi:10.1016/j.desal.2012.05.011
Rattner, B. A., Lazarus, R. S., Elliott, J. E., Shore, R. F., & van den Brink, N. (2014). Adverse Outcome Pathway and Risks of Anticoagulant Rodenticides to Predatory Wildlife. Environmental Science & Technology, 48(15), 8433-8445. doi:10.1021/es501740n
Regnery, J., Friesen, A., Geduhn, A., Göckener, B., Kotthoff, M., Parrhysius, P., . . . Brinke, M. (2018). Rating the risks of anticoagulant rodenticides in the aquatic environment: a review. Environmental Chemistry Letters. doi:10.1007/s10311-018-0788-6
Regnery, J., Parrhysius, P., Schulz, R. S., Mohlenkamp, C., Buchmeier, G., Reifferscheid, G., & Brinke, M. (2019). Wastewater-borne exposure of limnic fish to anticoagulant rodenticides. Water Res, 167, 115090. doi:10.1016/j.watres.2019.115090
Regnery, J., Schulz, R. S., Parrhysius, P., Bachtin, J., Brinke, M., Schafer, S., . . . Friesen, A. (2020). Heavy rainfall provokes anticoagulant rodenticides' release from baited sewer systems and outdoor surfaces into receiving streams. Sci Total Environ, 740, 139905. doi:10.1016/j.scitotenv.2020.139905
Rettie, A. E., Korzekwa, K. R., Kunze, K. L., Lawrence, R. F., Eddy, A. C., Aoyama, T., . . . Trager, W. F. (1992). Hydroxylation of warfarin by human cDNA-expressed cytochrome P-450: a role for P-4502C9 in the etiology of (S)-warfarin-drug interactions. Chem Res Toxicol, 5(1), 54-59. doi:10.1021/tx00025a009
Riha, M., Karlickova, J., Filipsky, T., Jahodar, L., Hrdina, R., & Mladenka, P. (2014). In vitro copper-chelating properties of flavonoids. Free Radic Biol Med, 75 Suppl 1, S46. doi:10.1016/j.freeradbiomed.2014.10.807
Rishavy, M. A., Usubalieva, A., Hallgren, K. W., & Berkner, K. L. (2011). Novel insight into the mechanism of the vitamin K oxidoreductase (VKOR): electron relay through Cys43 and Cys51 reduces VKOR to allow vitamin K reduction and facilitation of vitamin K-dependent protein carboxylation. J Biol Chem, 286(9), 7267-7278. doi:10.1074/jbc.M110.172213
Saengtienchai, A., Ikenaka, Y., Watanabe, K., Ishida, T., & Ishizuka, M. (2011). Comparative metabolism of warfarin in rats and chickens. Poult Sci, 90(12), 2775-2781. doi:10.3382/ps.2010-01257
Saito-Shida, S., Nemoto, S., Matsuda, R., & Akiyama, H. (2016). Simultaneous determination of seven anticoagulant rodenticides in agricultural products by gel permeation chromatography and liquid chromatography-tandem mass spectrometry. J Environ Sci Health B, 51(11), 801-808. doi:10.1080/03601234.2016.1198654
Sato, S. (2005). Coumarin rodenticides. In Coumarin Rodenticides. In: Drugs and Poisons in Humans (pp. 599-608): Springer Berlin Heidelberg.
Spink, D. C., Aldous, K. M., & Kaminsky, L. S. (1989). Analysis of oxidative warfarin metabolites by thermospray high-performance liquid chromatography/mass spectrometry. Anal Biochem, 177(2), 307-313. doi:10.1016/0003-2697(89)90057-2
Stafford, D. W. (2005). The vitamin K cycle. J Thromb Haemost, 3(8), 1873-1878. doi:10.1111/j.1538-7836.2005.01419.x
Waddell, L. S., Poppenga, R. H., & Drobatz, K. J. (2013). Anticoagulant rodenticide screening in dogs: 123 cases (1996-2003). J Am Vet Med Assoc, 242(4), 516-521. doi:10.2460/javma.242.4.516
Wang, N. N., Zheng, T., Zhang, G. S., & Wang, P. (2016). A review on Fenton-like processes for organic wastewater treatment. Journal of Environmental Chemical Engineering, 4(1), 762-787. doi:10.1016/j.jece.2015.12.016
Wardrop, D., & Keeling, D. (2008). The story of the discovery of heparin and warfarin. Br J Haematol, 141(6), 757-763. doi:10.1111/j.1365-2141.2008.07119.x
Watanabe, K. P., Kawata, M., Ikenaka, Y., Nakayama, S. M., Ishii, C., Darwish, W. S., . . . Ishizuka, M. (2015). Cytochrome P450-mediated warfarin metabolic ability is not a critical determinant of warfarin sensitivity in avian species: In vitro assays in several birds and in vivo assays in chicken. Environ Toxicol Chem, 34(10), 2328-2334. doi:10.1002/etc.3062
Wong, Y. W., & Davis, P. J. (1989). Microbial models of mammalian metabolism: stereoselective metabolism of warfarin in the fungus Cunninghamella elegans. Pharm Res, 6(11), 982-987. doi:10.1023/a:1015905832184
Zhang, M. H., Dong, H., Zhao, L., Wang, D. X., & Meng, D. (2019). A review on Fenton process for organic wastewater treatment based on optimization perspective. Sci Total Environ, 670, 110-121. doi:10.1016/j.scitotenv.2019.03.180

Download PDF

Journal Publication

published 08 Feb, 2024

Read the published version in Environmental Science and Pollution Research →

Reviewers agreed at journal
01 Sep, 2023
Reviewers invited by journal
01 Sep, 2023
Editor invited by journal
30 Aug, 2023
Editor assigned by journal
11 Aug, 2023
First submitted to journal
07 Aug, 2023

You are reading this latest preprint version

Developing a GC-EI-MS/MS Method for Quantifying Warfarin and Five Hydroxylated Metabolites Generated by the Fenton Reaction.

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Experimental Section

Results and Discussion

GC-MS/MS Methods Development

Aqueous Degradation of Warfarin

Conclusion

Declarations

References

Status:

Journal Publication

Version 1