In vitro characterization of 4'-(p-toluenesulfonylamide)-4-hydroxychalcone using human liver microsomes and recombinant cytochrome P450s.

Abstract 1. 4′-(p-Toluenesulfonylamide)-4-hydroxychalcone (TSAHC) is a synthetic sulfonylamino chalcone compound possessing anti-cancer properties. The aim of this study was to elucidate the metabolism of TSAHC in human liver microsomes (HLMs) and to characterize the cytochrome P450 (P450) enzymes that are involved in the metabolism of TSAHC. 2. TSAHC was incubated with HLMs or recombinant P450 isoforms (rP450) in the presence of an nicotinamide adenine dinucleotide phosphate, reduced form (NADPH)-regenerating system. The metabolites were identified and analyzed using liquid chromatography-tandem mass spectrometry (LC-MS/MS). P450 isoforms, responsible for TSAHC metabolite formation, were characterized by chemical inhibition and correlation studies in HLMs and enzyme kinetic studies with a panel of rP450 isoforms. 3. Two hydroxyl metabolites, that is M1 and M2, were produced from the human liver microsomal incubations (Km and Vmax values were 2.46 µM and 85.1 pmol/min/mg protein for M1 and 9.98 µM and 32.1 pmol/min/mg protein for M2, respectively). The specific P450 isoforms responsible for two hydroxy-TSAHC formations were identified using a combination of chemical inhibition, correlation analysis and metabolism by expressed recombinant P450 isoforms. The known P450 enzyme activities and the rate of TSAHC metabolite formation in the 15 HLMs showed that TSAHC metabolism is correlated with CYP2C and CYP3A activity. The P450 isoform-selective inhibition study in HLMs and the incubation study of cDNA-expressed enzymes also showed that two hydroxyl metabolites M1 and M2 biotransformed from TSAHC are mainly mediated by CYP2C and CYP3A, respectively. These findings suggest that CYP2C8, CYP2C9, CYP2C19, CYP3A4 and CYP3A5 isoforms are major enzymes contributing to TSAHC metabolism.

Introduction 4 0 -(p-Toluenesulfonylamide)-4-hydroxychalcone (TSAHC; Figure 1) is a sulfonylamino chalcone that has strong inhibitory effects on tyrosinase and a-glucosidase (Seo et al., 2010). TSAHC blocks cellular migration, invasion and multilayer growth mediated by transmembrane 4 L six family member 5 (TM4SF5), which is overexpressed in hepatocarcinoma subjects (Lee et al., 2009). TSAHC treatment suppressed tumor formation in nude mice xenografted with TM4SF5-expressing cells for 4 weeks (Lee et al., 2011). In addition, TSAHC strongly inhibited CYP2J2 enzyme activity, exhibiting a K i value of 2.03 mM (Lee et al., 2014a). The CYP2J2 isoform is overexpressed in hepatocarcinoma cells and human liver carcinoma tissues (Jiang et al., 2005;Lee et al., 2014b), and it is responsible for the formation of epoxyeicosatrienoic acids from arachidonic acid (Wu et al., 1996). Several studies have demonstrated that epoxyeicosatrienoic acids promote tumor growth, proliferation, adhesion and migration (Chen et al., 2011;Jiang et al., 2005;Jiang et al., 2009). Therefore, TSAHC might inhibit tumorigenesis via the inhibition of both TM4SF5-and/or CYP2J2-mediated tumorigenic proliferation in hepatocarcinoma. Taken together, these data suggest that TSAHC could be a new candidate anti-cancer drug.
Although much is known about the anti-cancer activity of TSAHC, its metabolism in human liver microsomes (HLMs) and with recombinant P450 isoforms has not been studied. This study was performed with the support of early drug discovery to identify the metabolic pathway of TSAHC in HLMs. Mass spectral structural identification of the major metabolites of TSAHC was conducted using human liver microsomes and liquid chromatography-tandem mass spectrometry (LC-MS/MS). In addition, we characterized drugmetabolizing enzymes responsible for TSAHC metabolism using a combination of chemical inhibition, correlation analysis in HLMs and metabolism by expressed recombinant P450 isoforms. Such information may be of considerable clinical importance with regard to interindividual variations in TSAHC metabolism and pharmacokinetics.

Identification of TSAHC metabolites in human liver microsomes
To identify the phase-I metabolites of TSAHC, 0.25 mg/mL of pooled HLMs (HLM 150, BD Biosciences), 0.1 M potassium phosphate buffer (pH 7.4) and 100 mM TSAHC were preincubated at 37 C for 5 min. Microsomal incubation was initiated by the addition of an NADPH-generating system containing 3.3 mM G6P, 500 units/ml G6PDH, 1.3 mM -NADP + , and 3.3 mM MgCl 2 to the reaction mixtures (final volume 100 mL), and the reaction mixture was further incubated for 60 min at 37 C with agitation. The control incubations were conducted with heat-denatured HLMs (100 C for 30 min). In all experiments, TSAHC was dissolved in acetonitrile, and the final concentration of acetonitrile did not exceed 1% (1% acetonitrile did not affect significantly the TSAHC metabolism). The reaction was terminated by adding a 50 mL ice-cold acetonitrile and centrifuged at 18 000 Â g for 5 min. Aliquots (5 mL) of the supernatant were injected in to LC-MS/MS to identify the TSAHC metabolites.
In vitro metabolism of TSAHC by human liver microsomes and cDNA expressed P450 isoforms Preliminary experiments showed that the formation of TSAHC metabolites was linear with respect to both time over 30 min and microsomal protein concentrations up to 1.0 mg/mL. Therefore, a 15-min incubation time and a 0.25 mg/mL microsomal protein concentration were selected for further study. The incubation mixtures containing either 5 mL HLMs (5 mg/mL) or cDNA-expressed P450 isoforms (500 pmol/mL) and various concentration of TSAHC (0-100 mM) were reconstituted in 0.1 M phosphate buffer (pH 7.4) and preincubated for 5 min at 37 C. The final incubation volume was 0.1 mL. The reaction was initiated by the addition of the NADPH-generating system and further incubated for 15 min at 37 C. The reaction was terminated by the addition of 50 mL ice-cold acetonitrile containing 60 nM omeprazole as an internal standard (IS), and the incubation mixtures were centrifuged at 20 000 Â g for 5 min. Aliquots of the supernatant (2 mL) were analyzed by LC-MS/MS.

LC-MS/MS analysis of TSAHC and its metabolites
For the identification of TSAHC and its metabolites, a Thermo Vantage triple quadrupole mass spectrometer (ThermoFisher Scientific, San Jose, CA) coupled with a Thermo Accela HPLC system was used. Chromatographic separation was performed on a Luna C18 column (2 mm i.d Â 50 mm, 5 mm, 100 Å , Phenomenex, Torrance, CA) with an isocratic mobile phase consisting of acetonitrile and water (45/55, v/v) containing 0.1% formic acid. The flow rate and total run time were 0.2 mL/min and 6 min, respectively. For metabolite identification, mass spectra were recorded by electrospray ionization in a negative mode. The optimum operating conditions were determined as follows: auxiliary gas, 10 Arb; nitrogen gas flow rate, 8 L/min; sheath gas pressure, 35 Arb; capillary temperature, 350 C; vaporizer temperature, 300 C; and spray voltage, 4000 V. Quantification was performed by selected reaction monitoring (SRM) of the [M-H] À ion and the related product ion for each metabolite using an IS to establish peak area ratios. Detection of the ions was performed by monitoring the transitions of m/z 392 ! 237 for TSAHC (30 eV collision energy), 408 ! 237 for hydroxy-TSAHC M1 (30 eV collision energy), 408 ! 253 for hydroxy-TSAHC M2 (30 eV collision energy) and 342 ! 203 for omeprazole (IS, 17 eV collision energy). Hydroxyl-TSAHC M1, which is the lack of authentic standard, was quantitated using the calibration curve of hydroxyl-TSAHC M2. Analytical data were processed by ThermoFisher Xcalibur software (version 2.1). The lower limit of quantification for metabolites was 1 nM.
The coefficient of variation of interassay precision for the analyte was less than 15%.

Data analysis
In the microsomal incubation studies, the apparent kinetic parameters of TSAHC metabolism (K m and V max ) were determined by fitting a one-enzyme Michaelis-Menten The calculated parameters were the Michaelis constant (apparent K m ), the maximum rate of formation (V max ), the intrinsic clearance (CL int , which is equal to V max /K m ), the substrate inhibition constant (K si ), and the hill coefficient (n) where [S] is the substrate concentration. Results are expressed as the mean ± S.D. of estimates obtained from pooled HLMs or recombinant P450 isoforms in triplicate experiments. Percentage inhibition was calculated as the ratio of metabolite formation in the presence and absence of the specific inhibitor. Calculations were performed using WinNonlin software (Pharsight, Mountain View, CA).

Results and discussion
After HLMs were incubated with TSAHC in the presence of an NADPH-generating system, two metabolites (M1 and M2) were found by LC-MS/MS (Figures 1 and 2). The retention times for TSAHC, M1 and M2 were approximately 4.2, 1.6 and 2.7 min, respectively. Product ion scan analysis of TSAHC and its two metabolites produced the informative and prominent product ions for structural elucidation. The MS/MS spectrum of TSAHC, having a deprotonated molecular ion [M-H] À at m/z 392, showed major fragment ions at m/z 237 due to the sulfonamide bond cleavage and two minor peaks at m/z 246. The loss of the cinnamaldehyde group was also represented at m/z 155 (tosyl moiety, Figure 2A). Metabolites M1 and M2 were tentatively identified as hydroxy-TSAHC ( Figure 2B and C). The mass spectra of M1 and M2 contain a deprotonated molecular ion peak [M-H]at m/z 408, suggesting one oxygen atom was inserted in to the TSAHC molecule ([M-H], m/z 392). These two hydroxy-TSAHC M1 and M2 showed a fragmentation pattern similar to the parent compound ( Figure 2). The MS/MS spectrum of M1 and TSAHC showed a fragment ion at m/z 237, whereas M2 showed a characteristic fragment ion at m/z 253. This suggests that the hydroxylation sites in the M1 and M2 metabolites are the 4-toluenesulfonamide and cinnamaldehyde groups, respectively. The metabolite M1 gave the characteristic product ion at m/z 171, suggesting the hydroxylation of the 4-toluenesulfonamide group and the loss of the cinnamaldehyde group from M1 ( Figure 2B). However, the exact hydroxylation site in the 4-toluenesulfonamide group of M1 could not be determined. The metabolite M2 was identified as hydroxy-TSAHC M2, by cochromatography and the MS/MS spectral data of the authentic synthetic compound ( Figure 2C). M2 produced the same fragment ions at m/z 246 (the loss of the hydroxycinnamaldehyde group) and 155 (tosyl moiety), which are also found in TSAHC. Based on these results, the possible metabolic pathway of TSAHC in HLMs is proposed in Figure 2.
Next, the involvement of the P450-isoform in the biotransformation of TSAHC was investigated using HLMs and recombinant P450 isoforms. TSAHC was metabolized by HLMs in the presence of an NADPH-generating system, but it was not metabolized in the absence of NADPH, indicating that TSAHC metabolism is P450-dependent (data are not shown). We provide evidence that the formation of hydroxy-TSAHC M1 and M2 from TSAHC is mainly catalyzed by the CYP2C and CYP3A isoform, respectively. First, the formation rate of hydroxy-TSAHC M1 from TSAHC (3 mM) was inhibited by montelukast (90%), sulfaphenazole (40%) and S-benzylnirvanol (37%), selective inhibitors of CYP2C8, 2C9 and 2C19, respectively (Mancy et al., 1996;Suzuki et al., 2002;Walsky et al., 2005). In contrast, the formation rate of hydroxy-TSAHC M2 was inhibited by ketoconazole (50%), a selective CYP3A inhibitor (Figure 3) (Baldwin et al., 1995). The effects of the other P450-isoform-selective inhibitors tested on metabolite formation were less than 20%, confirming that CYP2C and CYP3A play the most prominent role in hydroxy-TSAHC M1 and M2 formation, respectively. Second, TSAHC was incubated at two concentrations (1 and 3 mM) with a panel of cDNA-expressed P450 isoforms. Hydroxy-TSAHC M1 formation was predominantly mediated by CYP2C8, CYP2C9 and CYP2C19, with little contribution from CYP3A4 and CYP3A5 ( Figure 4A). Hydroxy-TSAHC M2 was mainly formed by the CYP3A4 isoform, with minor contributions from CYP1A2, CYP2D6 and CYP3A5 isoforms ( Figure 4B). Finally, rates of hydroxy-TSAHC M1 and M2 formation were compared to marker activities of each P450 isoform in a series of HLMs containing varying levels of enzyme (Table 1). Paclitaxel 6-hydroxylation and tolbutamide 4-methylhydroxylation, a marker of CYP2C8 and CYP2C9 activity (Bahadur et al., 2002;Lee et al., 2003), respectively, exhibited high correlation with the rate of hydroxy-TSAHC M1 formation (r ¼ 0.62 and 0.82, respectively) in 15 individual HLMs (Table 1). In contrast, midazolam 1 0 -hydroxylation activity, a marker of CYP3A activity (Li et al., 2003), exhibited high correlation with the rate of hydroxy-TSAHC M2 formation (r ¼ 0.62; Table 1). However, the significant correlation we observed between CYP2C9 activity and hydroxy-TSAHC M2 formation in the panel of HLMs tested may not have been caused by the actual involvement of CYP2C9 in hydroxy-TSAHC M2 formation, because our recombinant experiments and chemical inhibition study do not support a significant role of CYP2C9 in hydroxy-TSAHC M2 formation. The observed correlation is probably derived from the significant correlation between the activity of CYP2C9 and CYP3A (Spearman r ¼ 0.71, p50.01) in the bank of human livers tested (Supplementary Figure 1). This P450 enzyme interaction activity correlation was also previously reported by Li et al. (2003). They also observed a significant correlation (p50.05) between the activities of CYP3A-mediated midazolam 1 0 -hydroxylation and CYP2C9-mediated diclofenac 4 0 -hydroxylation in the HLMs bank (Li et al., 2003). These three different approaches collectively demonstrate that CYP2C and CYP3A are major P450s involved in hydroxy-TSAHC M1 and M2 formation, respectively. Based on these findings, a metabolic pathway for TSAHC in human liver microsomes is proposed in Figure 1.
Kinetic analysis of TSAHC metabolite formation rates was performed in three different HLMs. The kinetic profiles of TSAHC metabolism to hydroxy-TSAHC M1 and M2 in HLMs are shown in Figure 5. The formation of hydroxy-TSAHC M1 from TSAHC was best fitted by the substrateinhibition model ( Figure 5A, left panel). The corresponding Eadie-Hofstee plot had a ''hook'' in the upper region ( Figure  5A, right panel), which is characteristic of substrate inhibition (Liu et al., 2006). The K m and K si estimated from these data were 2.46 and 30.1 mM (Table 2), respectively. On the other hand, the rate of hydroxy-TSAHC M2 formation revealed sigmoidal saturation curves that were fitted to a Hill equation ( Figure 5B, left panel  formation exhibited a concave relationship, indicating negative cooperativity (n ¼ 0.89; Figure 5B, right panel and Table  2). The intrinsic clearance value (34.6 mL/min/mg protein) of hydroxy-TSAHC M1 formation was higher than that of hydroxy-TSAHC M2 (3.21 mL/min/mg protein; Table 2).  Data were analyzed using Spearman's nonparametric correlation test. The activity of each P450 isoform was determined using the respective specific substrate probe reaction, as described previously (Kim et al., 2005). a Statistically significant. V max is expressed as pmol/min/mg protein, K m and K si are as expressed as mM, CL int is expressed as V max /K m (mL/min/mg protein), and n is the Hill coefficient.