4-Cl-edaravone and (E)-2-chloro-3-[(E)-phenyldiazenyl]-2-butenoic acid are the specific reaction products of edaravone with hypochlorite

3-Methyl-1-phenyl-2-pyrazolin-5-one (edaravone) is a synthetic one-electron antioxidant used as a drug for treatment against acute phase cerebral infarction in Japan. This drug also reacts with two-electron oxidants like peroxynitrite to give predominantly 4-nitrosoedaravone but no one-electron oxidation products. It is believed that this plays a significant role in amelioration of amyotrophic lateral sclerosis. The drug was approved for treatment of amyotrophic lateral sclerosis in Japan and USA in 2015 and 2017, respectively. In this study, we examined the reaction of edaravone with another two-electron oxidant, hypochlorite anion (ClO−). Edaravone reacted with ClO− in 50% methanolic phosphate buffer (pH 7.4) solution containing typical two-electron reductants, such as glutathione, cysteine, methionine, and uric acid, as internal references. The concentration of edaravone decreased at a similar rate as each co-existing reference, indicating that it showed comparable reactivity toward ClO− as those references. Furthermore, 4-Cl-edaravone and (E)-2-chloro-3-[(E)-phenyldiazenyl]-2-butenoic acid (CPB) were identified as primary and end products, respectively, and no one-electron oxidation products were detected. These results suggest that edaravone treatment can bring greater benefit against ClO−-related injury such as inflammation, and 4-Cl-edaravone and CPB can be good biomarkers for ClO−-induced oxidative stress.

In this study, we examined another biologically important two-electron oxidant, hypochlorite anion (ClO -), because its specific oxidation products have not been characterized. Biologically, ClOis formed from a reaction of Cland hydrogen peroxide (H 2 O 2 ) catalyzed by myeloperoxidase (MPO), which is released from activated neutrophils during inflammation. It is a strong reactive oxygen species (ROS) and induces characteristic oxidation reactions such as sulfide oxidation to form sulfoxide and H 2 O 2 oxidation to produce singlet oxygen ( 1 O 2 ). Since ClOshows a strong bactericidal effect against germs including bacteria and Norwalk virus, it plays an important role in immune action. However, if ClOis produced excessively, it can cause oxidative damage to tissues.
In this study, the reactivity of edaravone toward ClOand its reaction products were investigated. We estimated the reactivity by comparing it with two-electron reductants such as reduced glutathione (GSH), cysteine (Cys), methionine (Met), and UA. Edaravone and each reference co-existed in a homogeneous solution and were oxidized with sodium hypochlorite (NaClO) induced at a constant rate into the solution. The amount of edaravone decreased a little slower than GSH or Cys did, almost identical with Met, and faster than UA, during competitive reactions with these antioxidants. These results suggested that edaravone has comparable reactivity to those antioxidants. Furthermore, two reaction products were identified. 4-Cl-edaravone was formed as a primary oxidation product by edaravone oxidation, which was then further oxidized to (E)-2-chloro-3-[(E)phenyldiazenyl]-2-butenoic acid (CPB) (Fig. 1). From these results, edaravone treatment against ClO --related injury such as inflammation may be useful and 4-Cl-edaravone can be a good marker to indicate ClOformation in vivo.

Materials and Methods
Chemicals. Edaravone, GSH, Cys, Met, UA, NaClO aqueous solution, and other chemicals were purchased from FUJIFILM Wako Pure Chemical Industries, Ltd. (Osaka, Japan). Methanolic solutions of edaravone were prepared and stored at -20°C until use.

Reaction of edaravone or 4 Cl edaravone with ClO − .
Edaravone was dissolved in 50% methanolic phosphate buffer solution (40 mM, pH 7.0) containing 100 mM diethylenetriaminepentaacetic acid (DTPA) as a chelator with GSH, Cys, Met, or UA as an internal reference. The final concentrations of edaravone and the internal reference were approximately 25 mM. NaClO solution (16.1 mM) was induced into the well-stirred reaction mixture (30 ml) at a constant rate (0.5 ml/min) using a syringe pump (Harvard Apparatus, Holliston, Massachusetts). Changes in concentrations of edaravone and the internal reference were measured every 25 or 30 min using an optimized HPLC as described below.
Oxidation of edaravone or 4-Cl-edaravone alone was also conducted. Edaravone and 4-Cl-edaravone solutions (50% methanolic or 100% methanolic) were prepared. NaClO solution was added instantaneously or introduced gradually (1.0 mM/min) to the edaravone solution. Changes in concentrations of edaravone and its metabolites were determined every 30 min during reaction using the HPLC system.
Isolation of 4 Cl edaravone and CPB. For identification of reaction products, 4-Cl-edaravone and CPB were isolated. After 30 min reaction of edaravone (1.0 mM) and excess NaClO, the reaction mixture was analyzed by HPLC to confirm that edaravone was completely reacted. The solution containing 4-Cl-edaravone and CPB was loaded onto a C18 solid phase extraction (SPE) cartridge (DSC-18, 2 g/12 ml, Sigma-Aldrich Japan, Tokyo, Japan) and flushed out to separate 4-Cl-edaravone adsorbed to the SPE cartridge. After washing the SPE cartridge with sufficient pure water, 4-Cl-edaravone was eluted with methanol. Next, a small amount of diluted HCl was added to the flow through, and the solution was loaded onto another C18 SPE cartridge. After washing with diluted HCl, CPB was eluted with methanol. After checking the purities, both solvents were removed by a rotary evaporator, and stored at -20°C until use.
The MS spectra of the collected 4-Cl-edaravone and CPB were measured by using a HPLC equipped with a time-of-flight mass spectrometer (LC/TOFMS).

HPLC analysis.
The changes in concentrations of edaravone and the internal references were determined by a reverse phase HPLC equipped with a UV detector monitoring absorption at 210 nm (for Met analysis) or 280 nm (for UA analysis) and an electrochemical detector (ECD) connected in tandem (for edaravone, GSH, and Cys analysis). The applied voltage of the electrochemical detector against an Ag/AgCl reference electrode was 600 mV. A 25% methanol aqueous solution containing NaClO 4 (100 mM) and phosphate (0.1%) was delivered at 1.0 ml/min as a mobile phase. The separation column was a TSKgel ODS-100Z (5 mm, 150 mm´4.6 mm; TOSOH, Tokyo, Japan). To analyze 4-Cl-edaravone and CPB, which are oxidation products of edaravone and ClO -, another HPLC was used. The mobile phase and separation column were 40% methanol aqueous solution containing 24 mM NaH 2 PO 4 delivered at 1.0 ml/min and a Wakopak Navi C18-5 (5 mm, 250 mm´4.6 mm; FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan), respectively. Detection was carried out by monitoring the absorbance at 240 nm.
LC/TOFMS analysis. To obtain accurate mass-to-charge ratios (m/z) of edaravone oxidative metabolites, a HPLC equipped with a TOF-MS (JMS-T100LC; JEOL Ltd., Tokyo, Japan) was used. Negative ionization was performed at an ionization potential of -2,000 V. For measuring parent molecular ions, the optimized applied voltages to the ring lens, outer orifice, inner orifice, and ion guide were -5 V, -20 V, -5 V, and -500 V, respectively. To obtain accurate m/z values, trifluoroacetic acid (TFA) was used as an internal standard for m/z calibration. To observe fragmentation, the above potential settings were changed to -10 V, -60 V, -10 V, and -500 V, respectively.
The HPLC conditions were as follows. 50% Methanol containing 0.1% formic acid was delivered at 1.0 ml/min as a mobile phase. A Wakopak Navi C18-5 (5 mm, 250 mm´4.6 mm; FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) was used as a separation column. Approximately a quarter of the mobile phase was induced to the MS analyzer using a splitter.

Results and Discussion
Reactivity of edaravone toward ClO − . The reaction of edaravone and ClOwas conducted in the presence of a twoelectron reductant as an internal reference. Edaravone decreased at comparable rates with reference antioxidants such as UA, Met, GSH, and Cys ( Fig. 2A-D). Edaravone and each reference antioxidant reacted with ClOand its concentration decreased competitively. The ratio of pseudo-second order rate constants (k E /k R ) was given by equation 3 as described previously.
(3) Table 1 shows the k E /k R values for edaravone vs each reference. The results suggest that the reactivity of edaravone toward ClOis comparable with the reference antioxidants. Second-order rate constants for the reactions of HOCl with Cys, GSH, and Met were determined by Storkey et al. (9) to be 3.6´10 8 , 1.2´10 8 , and 3.4´10 7 M -1 s -1 , respectively. The rate constant for UA was also estimated to be 3´10 5 M -1 s -1 . (10)(11)(12) Taken together, the pseudo second-order rate constant for reaction of edaravone and ClOcan be estimated to be ~10 6 -10 7 M -1 s -1 . Thiol (Cys) and sulfide (Met) residues are believed to be primary scavengers of ClOin vivo and their second-order rate constants for the reaction with ClOwere estimated to be 3.0´10 7 and 3.8´10 7 M -1 s -1 , respectively. (13) Edaravone was shown to possess enough reactivity to compete with these endogenous ClOscavengers.
Identification of reaction products of edaravone and hypochlorite. To identify the reaction products of edaravone and ClO -, edaravone alone was oxidized by ClO -. The formation of two unknown products U1 and U2 was observed after the reaction (Fig. 3A). Fig. 3B shows time course of changes of edaravone, U1 and U2 during induction of NaClO solution at a constant rate. U1 was formed in accordance with the consumption of edaravone and began decreasing when edaravone was depleted, suggesting that U1 was a primary product. U2 was produced in the late stage of the reaction and increased even after edaravone disappeared, indicating that U2 was a secondary product. MS spectra of U1 and U2 were measured by LC/TOFMS with negative electrospray ionization ( Fig. 3C and D, respectively). Their accurate mass-to-charge ratios (m/z) were determined to be -207.03251 and -223.02743, respectively, using TFA as an internal standard.   Table 1. Ratio of rate constants for edaravone and reference anti oxidants [mean ± SD (n = 3)] a see references (10)- (12), b see reference (9).

Edaravone + ClO
Reference UA 24.6 ± 0.53 3.0´10 5 a  (Fig. 1). The C4 carbon of the pyrazoline ring of edaravone shows reactivity to ROS and 4-adducts are formed as the oxidation products; 4-oxoedaravone by radical-induced oxidation (1) and 4-NO-and 4-NO 2 -edaravone are formed by ONOO -(3) (Fig. 1). From these results formation of 4-Cl-edaravone as the primary product is well plausible to reaction of edaravone and ClO -. In order to confirm whether U2 is CPB, its fragmentation was examined. Using the optimized TOFMS setting for measuring fragmentation described in Materials and Methods, fragmentation of CPB was observed (Fig. 3E). In addition to the parent ion (m/z -223.02747), a fragment ion (m/z -179.03501) was observed. The fragment ion was due to decarboxylation (C 9 H 6 N 2 35 Cl, theoretical m/z -179.03492), which was supported by its monoisotopic ion (C 9 H 6 N 2 37 Cl, Found m/z -181.03213; Theoretical m/z -181.03197). These results strongly suggest that U2 has a carboxyl group in its chemical structure as expected.
Since CPB was suggested to be the secondary product, further oxidation of 4-Cl-edaravone was conducted. Isolated and purified 4-Cl-edaravone was reacted with constantly induced NaClO. CPB was produced by 4-Cl-edaravone oxidation (Fig. 3F), indicating CPB was the oxidation product of 4-Cl-edaravone and ClO -. On the other hand, CPB was not oxidized further by ClO -(data not shown), suggesting that CPB is a stable end product. Postulated mechanism for 4 Cl edaravone and CPB for mation. In order to determine the formation mechanism for 4-Cl-edaravone and CPB, the stoichiometry for the reactions of edaravone and ClOand of 4-Cl-edaravone and ClOwas investigated. The reaction of 4-Cl-edaravone and ClOwas examined first because CPB is a non-reactive end product. 50, 100, and 200 mM of NaClO were added to 500 mM 4-Cl-edaravone solution at pH 7.0. The concentration of 4-Cl-edaravone immediately decreased after NaClO addition in accordance with the amount of added NaClO, with no further changes for 150 min (Fig. 4A). Next, 100 mM NaClO was added to 150 mM 4-Cl-edaravone solutions at pH 6.0, 7.0, 8.0, and 9.0. At all pHs, the concentration of 4-Cl-edaravone instantaneously decreased to 50 mM (Fig. 4B). A stoichiometric number expressed by the ratio of NaClO addition to 4-Cl-edaravone decrement was determined as approximately 1.0 for any concentration of NaClO and for any pH (Table 2). On the other hand, CPB formation was dependent on pH with the formation rates faster at higher pH (Fig. 4C). Furthermore, the concentration of 4-Cl-edaravone decreased by NaClO without CPB formation in 100% methanolic solution, whereas CPB was formed in 50% buffer (pH 9.0) (Fig. 4D). These results suggested that CPB formation requires H 2 O. Therefore, we conducted a  Fig. 4E), indicating that an oxygen atom of H 2 O was incorporated into the CPB molecule. Moreover, its decarboxylated fragment ion was non-isotopic (m/z -179.03502), suggesting that the oxygen atom derived from water was located on the carboxyl group.
Finally, to investigate the stoichiometry for the reaction of edaravone and ClO -, 100 mM NaClO was immediately added to 200 mM edaravone solution at pH 6.0, 7.0, 8.0 and 9.0. The concentration of edaravone was immediately reduced after NaClO addition, similar to 4-Cl-edaravone, and the decrement was approximately 70 mM at any pH (Fig. 5A), whereas 4-Cl-edaravone (Fig. 5B) and CPB (Fig. 5C) formations were dependent on pH. Table 3 shows the amounts of edaravone degradation and 4-Cledaravone and CPB formations in the instantaneous reaction of 200 mM edaravone and 100 mM ClOat various pHs. Edaravone degradation and CPB formation reached approximately 95 mM, equimolar with the added ClO -, at pH 9.0 where the reaction was the most efficient. Considering that the stoichiometric number for CPB production from the reaction of 4-Cl-edaravone and ClOwas determined to be 1.0, the stoichiometric number for the reaction of edaravone and ClOis also estimated to be 1.0. These results indicate that one edaravone molecule reacts with one ClOto immediately form an unknown intermediate for 4-Cl-edaravone formation. However, interestingly, 4-Cl-edaravone was instantaneously formed after NaClO addition to edaravone methanolic solution whereas the formation was gradual in 50% buffer solution (pH 9.0), indicating that the reactions of water and ClOor edaravone were competing with the reaction of edaravone and ClO -  ( Fig. 5D). The results suggested that the intermediate could be formed from edaravone, water and HOCl, which is produced from ClOand water. From these results, we postulated the reaction mechanism for edaravone and ClO - (Fig. 6). The enol form of edaravone (1) reacts with ClOor HOCl. When it reacts with ClO -, electron transfers rapidly occur to produce 4-Cl-edaravone. On the other hand, interaction of the enol of edaravone, HOCl, and H 2 O yields a stable cluster (2) as the intermediate with three hydrogen bonds between three OH groups. Deprotonation from the cluster molecule leads to 4-Cl-edaravone. Considering that the pKa value for HOCl is 7.53, it seems reasonable that 4-Cl-edaravone formation proceeded at pH above 8.0. The enol form of 4-Cl-edaravone (3) reacts with ClOto form 4,4-dichloro-edaravone (4) as an intermediate. The hydrolysis at the 1N-5C of 4,4-dichloroedaravone and subsequent hydrogen absorption from its amino group and rearrangement of the 2N-3C double bond to C3-C4 occur to give CPB (Fig. 6). This process is thought to be dependent on hydrogen elimination which is caused by HO -. In fact, CPB production was greater under higher pH (Fig. 5C), supporting our hypothesis.

Conclusion
We demonstrated that edaravone reacts with ClOat a comparable rate to thiols and sulfides, which are believed to be important endogenous ClOscavengers in vivo. These results indicate that edaravone is capable of reacting with ClOwhile co-existing with these antioxidants in vivo. The reaction yields 4-Cl-edaravone and CPB as primary and end products, respectively. From these results, edaravone treatment can be beneficial toward ClO --related injuries such as inflammation, in which MPO is thought to be upregulated. Furthermore, since 4-Cl-edaravone and CPB are ClO -specific oxidation products, they can be good indicators for ClOformation in vivo. Therefore, if these compounds are detected from edaravone-treated patients with diseases such as brain infarction and ALS, it indicates that ClOproduction occurs in these pathological conditions.