Multiple UDP-Glucuronosyltransferase and Sulfotransferase Enzymes are Responsible for the Metabolism of Verproside in Human Liver Preparations

Verproside, an active iridoid glycoside component of Veronica species, such as Pseudolysimachion rotundum var. subintegrum and Veronica anagallis-aquatica, possesses anti-asthma, anti-inflammatory, anti-nociceptive, antioxidant, and cytostatic activities. Verproside is metabolized into nine metabolites in human hepatocytes: verproside glucuronides (M1, M2) via glucuronidation, verproside sulfate (M3) via sulfation, picroside II (M4) and isovanilloylcatalpol (M5) via O-methylation, M4 glucuronide (M6) and M4 sulfate (M8) via further glucuronidation and sulfation of M4, and M5 glucuronide (M7) and M5 sulfate (M9) via further glucuronidation and sulfation of M5. Drug-metabolizing enzymes responsible for verproside metabolism, including sulfotransferase (SULT) and UDP-glucuronosyltransferase (UGT), were characterized. The formation of verproside glucuronides (M1, M2), isovanilloylcatalpol glucuronide (M7), and picroside II glucuronide (M6) was catalyzed by commonly expressed UGT1A1 and UGT1A9 and gastrointestinal-specific UGT1A7, UGT1A8, and UGT1A10, consistent with the higher intrinsic clearance values for the formation of M1, M2, M6, and M7 in human intestinal microsomes compared with those in liver microsomes. The formation of verproside sulfate (M3) and M5 sulfate (M9) from verproside and isovanilloylcatalpol (M5), respectively, was catalyzed by SULT1A1. Metabolism of picroside II (M4) into M4 sulfate (M8) was catalyzed by SULT1A1, SULT1E1, SULT1A2, SULT1A3, and SULT1C4. Based on these results, the pharmacokinetics of verproside may be affected by the co-administration of relevant UGT and SULT inhibitors or inducers.

Metabolite identification and characterization of the enzymes responsible for the metabolism of drugs, such as cytochrome P450, UDP-glucuronosyltransferase (UGT), and sulfotransferase (SULT), can reveal potential drug-drug interactions and inter-individual variations in drug metabolism and pharmacokinetics [14][15][16]. It is important to establish the comparative metabolism and drug-metabolizing enzymes of verproside for a full characterization of its pharmacokinetics, pharmacodynamics, and toxicity. There has been no report on the in vitro or in vivo metabolic pathways of verproside in humans and characterization of the enzymes responsible for verproside metabolism.
Thus, the purpose of this study was to determine the metabolites of verproside formed from in vitro incubation of verproside with human hepatocytes, intestinal microsomes, and liver S9 fractions using liquid chromatography-high resolution quadrupole Orbitrap mass spectrometry (LC-HRMS). We also sought to characterize the UGT and SULT enzymes responsible for verproside metabolism using human cDNA-expressed UGT and SULT supersomes.

In Vitro Metabolic Profile of Verproside in Human Hepatocytes
LC-HRMS analysis of the extracts after a 2-h incubation of verproside with human hepatocytes resulted in nine metabolites (M1-M9), along with unchanged verproside (Figure 1). The retention times and exact masses of the deprotonated molecular ion ([M − H] − ) and diagnostic product ions of verproside and its nine metabolites (M1-M9) are shown in Table 1. M1-M9 were commonly identified in human and rat hepatocytes [13]: verproside glucuronides (M1, M2), verproside sulfate (M3), picroside II (M4), isovanilloylcatalpol (M5), picroside II glucuronide (M6), isovanilloylcatalpol glucuronide (M7), picroside II sulfate (M8), and isovanilloylcatalpol sulfate (M9). M4 and M5 were identified as picroside II and isovanilloylcatalpol, respectively, by comparison with the accurate mass, retention time, and product scan spectra of the corresponding authentic standards. M6 and M8 were also identified after incubation of picroside II (M4) with human liver S9 fractions in the presence of uridine 5 -diphosphoglucuronic acid (UDPGA) and 3-phosphoadenosine-5-phosphosulfate (PAPS). M7 and M9 were identified after incubation of isovanilloylcatalpol (M5) with human liver S9 fractions in the presence of UDPGA and PAPS. The exact sites for glucuronidation and sulfation of verproside, picroside II (M4), and isovanilloylcatalpol (M5) into M1-M3 and M6-M9 could not be identified because of the lack of authentic standards. Based on these results, the potential in vitro metabolic pathways of verproside in human hepatocytes are shown in Figure 2.
Enzyme kinetics for verproside sulfation followed the Hill equation, but picroside II sulfation and isovanilloylcatalpol sulfation followed enzyme-inhibition and single-enzyme kinetics. The K m values for verproside sulfation in SULT1A1*1 (0.69 µM) and SULT1A1*2 (1.10 µM) were similar to that in human liver S9 fractions (K m , 1.2 µM). The major hepatic SULT1A1*1 and SULT1A1*2 played predominant roles in verproside sulfation and isovanilloylcatalpol sulfation. In the formation of picroside II sulfate (M8) from picroside II, SULT1A1*1 (K m , 9.0 µM; Cl int , 1122 µL/min/mg protein), SULT1A1*2 (K m , 10.8 µM; Cl int , 1249 µL/min/mg protein), and SULT1E1 (K m , 12.9 µM; Cl int , 2810 µL/min/mg protein) showed higher affinity and metabolic activity than did SULT1A2 (K m , 54.0 µM; Cl int , 113.6 µL/min/mg protein), SULT1A3 (K m , 1271.7 µM; Cl int , 64.8 µL/min/mg protein), and SULT1C4 (K m , 553.1 µM; Cl int , 110.5 µL/min/mg protein). The total contents of five SULTs (1A1, 1A3, 1B1, 1E1, and 2A1) in tissue cytosol fractions ranked in the order of small intestine (7800 ng/mg cytosol protein) > liver (5960 ng/mg cytosol protein) > kidney (430 ng/mg cytosol protein) > lung (290 ng/mg cytosol protein) [22]. SULT1A1 was the major hepatic SULT (accounting for 53% of total hepatic SULT proteins) but was also present in substantial quantities in the small intestine (19% of total SULT protein) [22]. SULT1E1 was expressed at relatively low levels in the liver (6% of total SULT protein) and the small intestine (8% of total SULT protein) but was the most abundant enzyme in the lung (40% of total SULT protein). SULT1A3 was not detected in the liver but was a major enzyme in the small intestine (31% of total SULT protein). SULT1C4 expression has been reported to be highest in fetal lung and kidney, with lower expression in fetal heart and adult kidney, ovary, and spinal cord, using multi-tissue dot blot analyses [23]. Based on these results, SULT1A1 and SULT1E1 may be the major enzymes responsible for the metabolism of picroside II into picroside II sulfate, with minor contributions by SULT1A2, SULT1A3, and SULT1C4. These results suggest that there may be inter-individual variability in the pharmacokinetics of verproside as a result of the changes in the rates of metabolism based on wide inter-individual variability in the expression of SULT1A1 and SULT1E1 in humans [24,25]. SULT1A1 and SULT1E1 have been shown to be induced or inhibited by various drugs and chemicals [25][26][27]. Co-administration of drugs that can inhibit or induce SULT1A1 and SULT1E1 may alter verproside sulfation.

In Vitro Metabolism of Verproside in Cryopreserved Human Hepatocytes
Cryopreserved human hepatocytes were purified and recovered using a high-viability cryohepatocyte recovery kit according to the manufacturer s protocol. Purified human hepatocytes were resuspended in William s E buffer to a final density of 0.8 × 10 6 cells/mL [16]. A portion of the human hepatocyte suspension (62.5 µL; 5 × 10 4 cells) and 62.5 µL of 400 µM verproside were added to a 96-well plate and incubated for 2 h at 37 • C in a CO 2 incubator. Then, 500 µL methanol were added to the incubation mixture and centrifuged (13,000× g, 10 min, 4 • C). The supernatant (500 µL) was evaporated to dryness using a vacuum evaporator (Genovac, UK). The residue was dissolved in 100 µL of 5% methanol. An aliquot (5 µL) was injected into the LC-HRMS system to identify the metabolites of verproside.
Authentic standards of verproside glucuronides (M1, M2), picroside II glucuronide (M6), and isovanilloylcatalpol glucuronide (M7) were not available. Thus, M1, M2, M6, and M7 were determined using the verproside, picroside II, and isovanilloylcatalpol calibration curves, respectively. Consequently, a limitation exists in the accurate interpretation of the enzyme kinetic data for M1, M2, M6, and M7 because each sensitivity of the metabolites was hypothesized to be the same as the corresponding substrate.
Authentic standards of verproside sulfate (M3), picroside II sulfate (M8), and isovanilloylcatalpol sulfate (M9) were not available. Thus, M3, M8, and M9 were determined using the verproside, picroside II, and isovanilloylcatalpol calibration curves, respectively. Consequently, a limitation exists played predominant roles in the formation of verproside glucuronides (M1, M2), isovanilloylcatalpol glucuronide (M7), and picroside II glucuronide (M6), with the highest metabolic activity observed for gastrointestinal-specific UGT1A10. These results suggest that the pharmacokinetics of verproside may be affected by co-administration of the relevant UGT and SULT inhibitors or inducers.