Identification of Catalposide Metabolites in Human Liver and Intestinal Preparations and Characterization of the Relevant Sulfotransferase, UDP-glucuronosyltransferase, and Carboxylesterase Enzymes

Catalposide, an active component of Veronica species such as Catalpa ovata and Pseudolysimachion lingifolium, exhibits anti-inflammatory, antinociceptic, anti-oxidant, hepatoprotective, and cytostatic activities. We characterized the in vitro metabolic pathways of catalposide to predict its pharmacokinetics. Catalposide was metabolized to catalposide sulfate (M1), 4-hydroxybenzoic acid (M2), 4-hydroxybenzoic acid glucuronide (M3), and catalposide glucuronide (M4) by human hepatocytes, liver S9 fractions, and intestinal microsomes. M1 formation from catalposide was catalyzed by sulfotransferases (SULTs) 1C4, SULT1A1*1, SULT1A1*2, and SULT1E1. Catalposide glucuronidation to M4 was catalyzed by gastrointestine-specific UDP-glucuronosyltransferases (UGTs) 1A8 and UGT1A10; M4 was not detected after incubation of catalposide with human liver preparations. Hydrolysis of catalposide to M2 was catalyzed by carboxylesterases (CESs) 1 and 2, and M2 was further metabolized to M3 by UGT1A6 and UGT1A9 enzymes. Catalposide was also metabolized in extrahepatic tissues; genetic polymorphisms of the carboxylesterase (CES), UDP-glucuronosyltransferase (UGT), and sulfotransferase (SULT) enzymes responsible for catalposide metabolism may cause inter-individual variability in terms of catalposide pharmacokinetics.

Catalposide had a short half-life (19.3 ± 9.5 min), and exhibited high systemic clearance (96.7 ± 44.1 mL/min/kg), and low urinary excretion (9.9 ± 4.1% of the dose) after intravenous administration of 10 mg/kg to male Sprague-Dawley rats [14]. This indicated that catalposide might be extensively metabolized in rats. However, catalposide remained stable after 1 h incubation

In Vitro Metabolism of Catalposide in Cryopreserved Human Hepatocytes
Cryopreserved human hepatocytes were recovered with the aid of a hepatocyte purification kit, and viable cells were resuspended in William's E buffer at a final concentration of 1.28 × 10 6 cells/mL [20]. Human hepatocyte suspensions (62.5 µL, 8.00 × 10 4 cells) and 62.5 µL of 400 µM catalposide in William's E buffer were added to the wells of a 96-well plate and the mixture was incubated for 120 min at 37 • C in a CO 2 incubator. Methanol (250 µL) was added to each well and the mixture was centrifuged at 3000× g for 10 min. Aliquots of the supernatants (250 µL) were evaporated to dryness using a vacuum evaporator (Genevac Ltd., Ipswich, UK). Each residue was dissolved in 100 µL of 5% methanol and an aliquot (5 µL) was injected into the LC-HRMS system.

In Vitro Metabolism of Catalposide in Human Liver S9 Fractions and Intestinal Microsomes
Each reaction mixture contained 50 mM potassium phosphate buffer (pH 7.4), 10 mM magnesium chloride, human liver S9 fractions or human intestinal microsomes (100 µg protein), 2 mM UDPGA or 200 µM PAPS, 200 µM catalposide or a possible metabolite, and 1000 µM 4-hydroxybenzoic acid in a volume of 200 µL. Samples lacking UDPGA and PAPS served as controls. The mixtures were incubated at 37 • C for 60 min and the reactions were then quenched by adding 500 µL of methanol. The tubes were centrifuged and the supernatants evaporated to dryness using a vacuum concentrator. The residues were dissolved in 100 µL of 5% methanol and 5 µL aliquots were injected into the LC-HRMS system.

Characterization of Carboxylesterases Involved in the Formation of 4-Hydroxybenzoic Acid from Catalposide
To identify the CES enzymes involved in hydrolysis of catalposide to 4-hydroxybenzoic acid, 100 µL reaction mixtures containing human liver S9 fractions; human intestinal microsomes; or human CES1b, CES1c, or CES2 enzymes (50 µg protein), and catalposide (200 or 400 µM) in 50 mM phosphate buffer (pH 7.4) were incubated at 37 • C for 30 min. Reactions were stopped by addition of 100 µL 4-methylumbelliferone (internal standard, 10 ng/mL) in methanol. After vortex-mixing and centrifugation, 50 µL of each supernatant was diluted with 50 µL of deionized water. Each mixture was transferred to an injection vial, and a 5 µL aliquot was injected into the LC-HRMS system.

LC-HRMS Analysis of Catalposide and Metabolites
To separate and identify catalposide and its metabolites, we used a Q-Exactive Orbitrap mass spectrometer coupled to an Accela UPLC system (Thermo Scientific, Waltham, MA, USA). Catalposide and its metabolites were optimally separated on a Halo C18 column via gradient elution using 5% (v/v) methanol in 1 mM ammonium formate (pH 3.1) (mobile phase A) and methanol (mobile phase B) at a flow rate of 0.5 mL/min: 5% mobile phase B for 2 min, 5-20% mobile phase B over 11.5 min, 20-90% mobile phase B over 0.5 min, 90% mobile phase B for 3 min, 90-5% mobile phase B over 0.5 min, and 5% mobile phase B for 2.5 min. The column and the autosampler were maintained at 40 and 6 • C, respectively. Accurate mass measurements of catalposide and its metabolites were derived via electrospray ionization in the negative mode using the following electrospray source settings: ion transfer capillary temperature, 330 • C; needle spray voltage, −3000 V; capillary voltage, −47.5 V; nitrogen sheath gas, 50 arbitrary units; auxiliary gas, 15 arbitrary units. The resolution and automatic gain control were scaled to 70,000 and 1,000,000, respectively. MS data were obtained using external calibration over the scan range m/z 100-700 and processed using Xcalibur software version 2.2 (Thermo Scientific). The Q-Exactive Orbitrap MS was calibrated using MSCAL5 and MSCAL6 for the positive and negative ion modes, respectively. Nitrogen gas was employed for higher-energy collision dissociation (HCD) at an energy of 25 eV to obtain product ion spectra of catalposide and its metabolites. Structures were determined using Mass Frontier software (version 6.0; HighChem Ltd., Bratislava, Slovakia). We used the extracted ion monitoring mode for quantification: m/z 481.1349 for catalposide, m/z 657.1674 for catalposide glucuronide, m/z 561.0921 for catalposide sulfate, m/z 137.0239 for 4-hydroxybenzoic acid, m/z 313.0569 for 4-hydroxybenzoic acid glucuronide, and m/z 175.0410 for 4-methylumbelliferone (the internal standard). The peak areas of all components were integrated using Xcalibur software. The calibration curve was linear over the catalposide concentration range 0.5-200 pmol. The concentrations of catalposide glucuronide and catalposide sulfate were calculated using the calibration curve for catalposide because we had no authentic standards.

Data Analysis
All results are the average of two determinations obtained using pooled human intestinal microsomes, pooled human liver S9 fractions, UGTs, and SULTs. The apparent kinetic parameters (K m , V max , n, and K i ) for formation of catalposide glucuronide or catalposide sulfate by human intestinal microsomes, liver S9 fractions, UGTs, or SULTs were determined by fitting the Hill equation model [V = V max S n /(K m n + S n )], the substrate inhibition model [V = V max /(1 + K m /S + S/K i )], or the single enzyme model [V = V max S/(K m + S)] to the unweighted formation rates of catalposide glucuronide and catalposide sulfate, respectively, over a range of catalposide concentrations using Enzyme Kinetics software (version 1.1 SPSS Science Inc., Chicago, IL, USA). In the above equations, V is the velocity of the reaction at substrate concentration [S], V max is the maximum velocity, n is the Hill constant, K m is the substrate concentration at which the reaction velocity is 50% of V max , and K i is the dissociation constant of the substrate binding to the inhibitory region within the enzyme active site.

In Vitro Metabolic Profiles of Catalposide Incubated with Human Hepatocytes and Intestinal Microsomes
LC-HRMS analysis of extracts after incubation of catalposide with human hepatocytes revealed three metabolites (M1-M3) and residual catalposide ( Figure 1A). LC-HRMS analysis of reaction mixtures after incubation of catalposide with human intestinal microsomes in the presence of UDPGA yielded M2, M3, and a new metabolite M4 ( Figure 1B).  Table 1. The four metabolite peaks were identified using the accurate mass values and the characteristic product ions of the product scan spectra (Table 1, Figure 2). The mass errors between the theoretical and observed m/z values for each metabolite were less than 5 ppm, indicating good correlations between the calculated theoretical masses and the experimentally observed masses obtained after full-scan MS analysis.   Table 1. The four metabolite peaks were identified using the accurate mass values and the characteristic product ions of the product scan spectra (Table 1, Figure 2). The mass errors between the theoretical and observed m/z values for each metabolite were less than 5 ppm, indicating good correlations between the calculated theoretical masses and the experimentally observed masses obtained after full-scan MS analysis.
The formation of catalposide sulfate from catalposide catalyzed by SULTs 1A1*1, 1A1*2, and 1E1 exhibited substrate inhibition kinetics, but the activities of SULT1C4 and pooled human liver S9 fractions fitted the Hill equation ( Figure 5). The enzyme kinetic parameters for the formation of catalposide sulfate from catalposide are listed in Table 2. SULT1C4 exhibited a higher affinity for catalposide and more rapid sulfation than did SULT1A1*1, SULT1A1*2, and SULT1E1.  The formation of catalposide sulfate from catalposide catalyzed by SULTs 1A1*1, 1A1*2, and 1E1 exhibited substrate inhibition kinetics, but the activities of SULT1C4 and pooled human liver S9 fractions fitted the Hill equation ( Figure 5). The enzyme kinetic parameters for the formation of catalposide sulfate from catalposide are listed in Table 2. SULT1C4 exhibited a higher affinity for catalposide and more rapid sulfation than did SULT1A1*1, SULT1A1*2, and SULT1E1.
A screen using twelve human cDNA-expressed UGT supersomes, to assess the metabolism of catalposide to catalposide glucuronide (M4), identified possible roles of gastrointestinal tract-specific UGT1A8 and UGT1A10 ( Figure 6B). The results show that catalposide glucuronide (M4) was produced after incubation of catalposide with pooled human intestinal microsomes, but not human liver S9 fractions. Formation of catalposide glucuronide (M4) from catalposide by pooled human intestinal microsomes followed single enzyme kinetics; formation via UGT1A8 and UGT1A10 exhibited Hill equation kinetics (Figure 7, Table 2). A screen using twelve human cDNA-expressed UGT supersomes, to assess the metabolism of catalposide to catalposide glucuronide (M4), identified possible roles of gastrointestinal tract-specific UGT1A8 and UGT1A10 ( Figure 6B). The results show that catalposide glucuronide (M4) was produced after incubation of catalposide with pooled human intestinal microsomes, but not human liver S9 fractions.
Formation of catalposide glucuronide (M4) from catalposide by pooled human intestinal microsomes followed single enzyme kinetics; formation via UGT1A8 and UGT1A10 exhibited Hill equation kinetics (Figure 7, Table 2). 4-Hydroxybenzoic acid (M2) was formed from catalposide by pooled human liver S9 fractions; intestinal microsomes; and the CES1b, CES1c, and CES2 enzymes (Figure 8). The rate of formation of 4-hydroxybenzoic acid (M2) after incubation of catalposide with pooled human intestinal microsomes was higher than that after incubation with pooled human liver S9 fractions (Figure 8).
CES2, the predominant CES of the intestine, was more active in terms of hydrolysis of catalposide to 4-hydroxybenzoic acid (M2) than were the hepato-predominant CES1b and CES1c enzymes [32,36]. Thus, the rate of formation of 4-hydroxybenzoic acid (M2) was higher when catalposide was incubated with pooled human intestinal microsomes than with pooled human liver S9 fractions.
UGT1A6 and UGT1A9 play major roles in the formation of 4-hydroxybenzoic acid glucuronide (M3) from 4-hydroxybenzoic acid ( Figure 6A). Abbas et al. [37] found that UGT1A9 played the major role in metabolism of 4-hydroxybenzoic acid to 4-hydroxybenzoic acid glucuronide. UGT1A6 and UGT1A9 are major enzymes of both the liver and intestine; therefore, 4-hydroxybenzoic acid glucuronide (M3) was identified after incubation of catalposide with either human hepatocytes or intestinal microsomes.