In vitro glucuronidation of 7-hydroxycoumarin derivatives in intestine and liver microsomes of Beagle dogs

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY-NC-ND 4.0 https://creativecommons.org/licenses/by-nc-nd/4.0/ In vitro glucuronidation of 7-hydroxycoumarin derivatives in intestine and liver microsomes of Beagle dogs © 2019 Elsevier B.V. Accepted version (Final draft) Juvonen, Risto O.; Heikkinen, Aki T.; Kärkkäinen, Olli; Jehangir, Rabia; Huuskonen, Juhani; Troberg, Johanna; Raunio, Hannu; Pentikäinen, Olli T.; Finel, Moshe Juvonen, R. O., Heikkinen, A. T., Kärkkäinen, O., Jehangir, R., Huuskonen, J., Troberg, J., Raunio, H., Pentikäinen, O. T., & Finel, M. (2020). In vitro glucuronidation of 7-hydroxycoumarin derivatives in intestine and liver microsomes of Beagle dogs. European Journal of Pharmaceutical Sciences, 141, Article 105118. https://doi.org/10.1016/j.ejps.2019.105118 2020 Journal Pre-proof In vitro glucuronidation of 7-hydroxycoumarin derivatives in intestine and liver microsomes of Beagle dogs Risto O. Juvonen , Aki T. Heikkinen , Olli K¨ arkk¨ ainen , Rabia Jehangir , Juhani Huuskonen , Johanna Troberg , Hannu Raunio , Olli T. Pentik¨ ainen , Moshe Finel PII: S0928-0987(19)30391-4 DOI: https://doi.org/10.1016/j.ejps.2019.105118 Reference: PHASCI 105118 To appear in: European Journal of Pharmaceutical Sciences Received date: 15 August 2019 Revised date: 3 October 2019 Accepted date: 22 October 2019 Please cite this article as: Risto O. Juvonen , Aki T. Heikkinen , Olli K¨ arkk¨ ainen , Rabia Jehangir , Juhani Huuskonen , Johanna Troberg , Hannu Raunio , Olli T. Pentik¨ ainen , Moshe Finel , In vitro glucuronidation of 7-hydroxycoumarin derivatives in intestine and liver microsomes of Beagle dogs, European Journal of Pharmaceutical Sciences (2019), doi: https://doi.org/10.1016/j.ejps.2019.105118 This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. ©2019 Published by Elsevier B.V. 1 In vitro glucuronidation of 7-hydroxycoumarin derivatives in intestine and liver microsomes of Beagle dogs 1Risto O. Juvonen, 2Aki T. Heikkinen, 1Olli Kärkkäinen, 1Rabia Jehangir, 3Juhani Huuskonen, 4Johanna Troberg, 1Hannu Raunio, 5Olli T. Pentikäinen, 4Moshe Finel. 1 School of Pharmacy, Faculty of Health Sciences, University of Eastern Finland, Box 1627, FI- 70211 Kuopio, Finland 2Admescope Ltd, Oulu, Finland 3 University of Jyvaskyla, Department of Chemistry, P.O. Box 35, FI-40014 University of Jyvaskyla, Finland 4 Division of Pharmaceutical Chemistry and Technology, Faculty of Pharmacy, University of Helsinki, P.O. Box 56, FI-00014 University of Helsinki, Finland 5 Institute of Biomedicine, Faculty of Medicine, University of Turku, FI-20014 University of Turku, Finland Corresponding author: Risto O. Juvonen, [email protected], School of Pharmacy, Faculty of Health Sciences, University of Eastern Finland, Box 1627, FI-70211 Kuopio, Finland Declaration of interest: none 2 Abstract Beagle dog is a standard animal model for evaluating nonclinical pharmacokinetics of new drug candidates. Glucuronidation in intestine and liver is an important first-pass drug metabolic pathway, especially for phenolic compounds. This study evaluated the glucuronidation characteristics of several 7-hydroxycoumarin derivatives in beagle dog’s intestine and liver in vitro. To this end, glucuronidation rates of 7-hydroxycoumarin (compound 1), 7-hydroxy-4-trifluoromethylcoumarin (2), 6-methoxy-7-hydroxycoumarin (3), 7-hydroxy-3-(4-tolyl)coumarin (4), 3-(4- fluorophenyl)coumarin (5), 7-hydroxy-3-(4-hydroxyphenyl)coumarin (6), 7-hydroxy-3-(4- methoxyphenyl)coumarin (7), and 7-hydroxy-3-(1H-1,2,4-tirazole)coumarin (8) were determined in dog’s intestine and liver microsomes, as well as recombinant dog UGT1A enzymes. The glucuronidation rates of 1, 2 and 3 were 3–10 times higher in liver than in small intestine microsomes, whereas glucuronidation rates of 5, 6, 7 and 8 were similar in microsomes from both tissues. In the colon, glucuronidation of 1 and 2 was 3–5 times faster than in small intestine. dUGT1A11 glucuronidated efficiently all the substrates and was more efficient catalyst for 8 than any other dUGT1A. Other active enzymes were dUGT1A2 that glucuronidated efficiently 2, 3, 4, 5, 6 and 7, while dUGT1A10 glucuronidated efficiently 1, 2, 3, 4, 5 and 7. Kinetic analyses revealed that the compounds’ Km values varied between 1.1 (dUGT1A10 and 2) and 250 µM (dUGT1A7 and 4). The results further strengthen the concept that dog intestine has high capacity for glucuronidation, and that different dUGT1As mediate glucuronidation with distinct substrates selectivity in dog and human. Keywords: glucuronidation, dog, intestine, liver, 7-hydroxycoumarin derivative, enzyme kinetics. 3 1. Introduction Detailed information about nonclinical and clinical pharmacokinetics (absorption, distribution, metabolism and excretion, ADME) is necessary for all new and existing drugs. Nonclinical ADME data is routinely obtained from rodents, dogs and monkeys [Reichel and Lienau, 2016; Rowland et al., 2015]. Oral bioavailability of drugs and other xenobiotic compounds is critically affected by the extents of absorption and first-pass metabolism in the intestine and liver. Intestinal metabolism is important for bioavailability of compounds that could be directly conjugated, such as glucuronidation of phenols [Gonzalez et al., 2018; Testa et al., 2012]. Glucuronidation is a significant conjugation pathway in the metabolism of drugs, numerous other xenobiotics and endogenous substances such as bilirubin, hormones, bile acids, and retinoids. Glucuronidation is catalyzed by UDP-glucuronosyltransferase enzymes (UGTs, EC 2.4.1.17) and takes place at nucleophilic functional groups on the substrate molecule, particularly hydroxyl, different amino, or carboxylic acid groups. Glucuronide conjugates are subsequently substrates for efflux transporters and are therefore readily excreted in urine or bile [Kaivosaari et al., 2011; Rowland et al., 2013; Wu et al., 2011]. UGTs are membrane bound enzymes of the endoplasmic reticulum and are broadly, but variably expressed in tissues with highest concentrations in the liver and gastrointestinal tract. Per tissue weight, there is a higher concentration of UGTs in the small intestine compared with the liver [Fujiwara et al., 2016; Hu et al., 2014; Naritomi et al., 2015]. On the other hand, at least in humans and at the protein level, only few different UGTs, including extrahepatic ones, are expressed in the small intestine to a level that is detectable by proteomics methods, whereas more UGTs are found in the liver [Sato et al., 2014]. Mammals express four UGT subfamilies. Excluding pseudogenes, nine members of the UGT1A subfamily and ten members of the UGT2A and 2B subfamilies have been identified in human, while ten UGT1As and three UGTs of the 2A and 2B subfamilies have been detected in the dog [http://prime.vetmed.wsu.edu/resources/udp-glucuronsyltransferase-homepage]. The main human UGTs involved in drug metabolism are UGT1A1, UGT1A3, UGT1A4, UGT1A6, UGT1A9, and UGT2B7 [Bock, 2016, Kaivosaari et al., 2011, Rowland et al., 2013], but others, such as UGT1A10, UGT2B10, UGT2B15 and UGT2B17 are also important in different reactions. There are substantial differences in substrate preferences between individual UGTs across species [Komura and Iwaki, 2011; Peters et al., 2016; Troberg et al., 2017]. In the beagle dog three dUGT1A enzymes, 1A2, 1A9 and 1A11, are mainly expressed in intestine and show low or no abundance in the liver, where dUGT2B31 is the most abundantly expressed UGT [Heikkinen et al., 4 2015]. Human UGTs 1A7, 1A8, 1A10, 2A1 and 2A2 are all extrahepatic [Court et al., 2012]. Of them, at least UGT1A10 is expressed in the small intestine to high level, along with the (also hepatic) UGTs 1A1, 2B7 and 2B17 [Ohno and Nakajin, 2009; Court et al., 2012; Sato et al., 2014]. Thus, the tissue expression profiles of individual UGTs in human and dog do not match, and it is not possible to correlate exactly between individual human and dog UGTs [Troberg et al., 2015; Heikkinen 2015]. The use of nonclinical species such as beagle dog in predicting human oral bioavailability of drugs has been extensively investigated. In general, bioavailability in animals does not exactly predict quantitatively bioavailability in human. Quantitative (high/low bioavailability) estimates could be improved, when precise species-specific factors that affect human bioavailability are incorporated into the quantitative predictions [Jones et al., 2016; Musther et al., 2014; Peters et al, 2016; Miller et al.; 2019]. In addition, obtaining exact information about pharmacokinetic mechanisms in domestic animals such as beagle dogs is becoming increasingly important in veterinary pharmacotherapy [Anadón, 2016]. We have previously developed a quantitative multiwell plate assay to measure the glucuronidation rate of 7-hydroxy-4-trifluoromethylcoumarin (HFC) [Rahikainen et al., 2013]. Recently we reported six new C3-substituted 7-hydroxycoumarin derivatives that are selective substrates for human UGT1A10 [Juvonen et al., 2018a]. The glucuronidation rate of these substrates was higher in intestine than in liver microsomes. A major advantage of these 7-hydroxycoumarin derivatives as UGT substrates is that their fluorescence properties could be measured in very simple assays, allowing for high-throughput evaluation of enzyme characteristics in different experimental settings. The aim of this study was to determine glucuronidation characteristics of a panel of C3-substituted 7-hydroxycoumarin derivatives (Figure 1) in dog intestine and liver in vitro and to compare these characteristics to the ones in human. 2. Materials and methods 2.1. Reagents: Alamethicin, UDP-glucuronic acid sodium salt, 7-hydroxycoumarin (compound 1) (99%), HFC (2) (99%), 6-methoxy, 7-hydroxycoumarin (3) (99 %) were from Sigma-Aldrich (Mannheim, Germany). MgCl2 were from Riedel-de Haen (Vantaa, Finland). Water was deionized by MilliQ gradient A10. 5 The novel 7-hydroxycoumarin derivatives 3-triazole-7-hydroxycoumarin (8), 3-(4-tolyl)-7- hydroxycoumarin (4), 3-(4-fluorophenyl-7-hydroxycoumarin (5), 3-(4-methoxyphenyl)-7- hydroxycoumarin (7), 3-(4-hydroxyphenyl)-7-hydroxycoumarin (6) were synthesized using the Perkin-Oglialor condensation reaction described in detail earlier [Juvonen et al., 2018b]. 2.2. Biological materials: Beagle dog necropsy was performed at F. Hoffmann-La Roche Ltd (Nutley, NJ) according to institutional guidelines and in compliance with national and regional legislation. For detailed description of the preparation of intestine and liver microsomal samples of beagle dogs see [Heikkinen et al., 2015]. Liver samples were snap frozen in liquid nitrogen and stored at –80°C before sample preparation. Small intestines were divided into five equal segments (median length 53 cm), and colon (median length 29 cm) were collected as separate anatomical segments. Ten recombinant dog UGT1As, namely 1A1, 1A2, 1A3, 1A4, 1A6, 1A7, 1A8, 1A9, 1A10 and 1A11, were produced as His-tagged proteins in baculovirus-infected insect cells and expression normalized as previously described [Troberg et al., 2015]. The relative expression levels of each of these recombinant UGTs were the following: 1 for dUGT1A1, 6.5 for dUGT1A2, 3.7 for dUGT1A3, 5.8 for dUGT1A4, 14.6 for dUGT1A6, 17.9 for dUGT1A7, 1.3 for dUGT1A8, 7.3 for dUGT1A9, 3.8 for dUGT1A10 and 10.1 for dUGT1A11. 2.3. Glucuronidation assays: The assays were carried out in 96 multiwell plates, and incubations were done in 100 µl and at 37°C, in the presence of 100 mM Tris-HCl buffer pH 7.4, 0.5 mM UDPGA, 0.1 – 0.4 g/l protein intestine or liver microsomes or 0.04 - 1 g/l protein recombinant UGT and the tested 7-hydroxycoumarin derivative at the indicated concentrations. Alamethicin (25 µg/ml) was present at microsomal incubation but not at incubations of recombinant UGTs. Blank samples did not contain either UDPGA or enzyme source. Fluorescence decline in the multiwell plate experiments was monitored every other minute, for 40 min, using an excitation at 405 nm and detecting emission at 460 nm, in a Victor2 1420 Multilabel counter (PerkinElmer, Life Sciences, Turku, Finland) [Juvonen et al. 2018a]. The fluorescence values were transformed into molarity using standard curves that were prepared with the respective compounds at every time point (fluorescence disappeared upon glucuronidation). Slopes of the substrate concentration decrease per min were calculated using linear regression analysis, in which the linear part of the kinetic assay indicated the glucuronidation rate. Enzyme catalyzed glucuronidation rate was calculated by subtracting the blank value from the 6 full reaction value and then normalized by diving the amount of protein in the sample. The intraassay variability of the kinetic assays was 6 % when compound 8 was used as the substrate. Kinetic analyses were also performed in the same 96 multiwell plate assays, with excitation at 405 nm and detection at 460 nm, using eleven concentrations per substrate (0.25–20 µM). The reactions were catalyzed linearly for at least 15 min even at the lowest substrate concentrations. The data was analyzed by the Michaelis-Menten equation v = S * Vmax /(Km + S), in which v is the reaction rate at substrate concentration (S), Vmax is limiting rate of the reaction and Km is the Michaelis constant equal to the substrate concentration, at which the reaction rate is 50 % of Vmax . For multivariate analysis, we did principal component analysis (PCA) using SIMCA 15.0.2 (Umetrics). For analysis of 7-hydroxycoumarin derivatives glucuronidation in dog intestine and liver microsomes, we first normalized the values within samples from the same UGT or dog glucuronidating the same substrate, so that all values were divided by the highest value (normalized value = value / max value), due to large variation in glucuronidation rate between samples from different UGTs or dogs. 3. Results At first, we screened the glucuronidation rates of 7-hydroxycoumarin (1) and its seven derivatives by the dog recombinant dUGT1As, i.e. 1A1, 1A2, 1A3, 1A4, 1A6, 1A7, 1A8, 1A9, 1A10 and 1A11 using a substrate concentration of 10 µM in each case (Figure 2, Table 1). The results revealed the highest rate for compound 5 by dUGT1A2 (169 nmol/(min*g protein), which glucuronidated also 2 almost equally fast. Compounds 8 and 1 were glucuronidated at the lowest rates by the dUGT1A enzymes. The number of dUGT1As catalyzing glucuronidation of individual compounds varied. For example, 8 was glucuronidated preferentially by dUGT1A11 with some contribution by dUGT1A1 and minor contribution by the other enzymes. The other derivatives were glucuronidated by three or more dUGT1As. dUGT1A11 glucuronidated all eight compounds, while the other dUGT1As exhibited more restricted substrate profiles, at least in the case of this group of compounds (Figure 2, Table 1). dUGT1A2 and dUGT1A10 glucuronidated six different compounds. dUGT1A4 and dUGT1A9 did not catalyze any, or catalyzed only poorly few of the compounds. The data was analyzed by the PCA model, which used five components with a cumulative R2X of 0.95, and a cumulative Q2 of 0.58. dUGTs 1A1, 1A3, 1A4, 1A6, 1A7, 1A8 and 1A9 formed a separate group in the analysis, 7 having a weak connection with all substrates (Figure 3). dUGT1A2 had a strong connection with substrates 4, 5, 6 and 7, whereas dUGT1A10 and dUGT1A11 had a strong connection with substrates 1, 3 and 8. Substrate 2 had an equal connection with dUGT1A2, dUGT1A10 and dUGT1A11. Michaelis-Menten kinetic parameters were determined for the fastest glucuronidation reactions (Table 2, Supplement Figure 1). The derived Km values varied between 1.1 µM (compound 2 - dUGT1A10) and 250 µM (4 - dUGT1A7). In most cases the Km values were lower for dUGT1A10 and dUGT1A11 than in the other dUGTs which catalyzed the same reactions. Most Km values were below 10 µM, particularly for dUGT1A10 and dUGT1A11. Higher Km values were observed for 1 (dUGT1A6, dUGT1A11), 2 (dUGT1A2), 3 (dUGT1A1, dUGT1A2), and 4 (dUGT1A3, dUGT1A7). Normalized intrinsic clearance (Vmax/Km ratio) varied 100-fold (1.8–180 ml/(min*g protein)). Normalized intrinsic clearance was highest for glucuronidation of 5 by dUGT1A1 and lowest for 3 by dUGT1A1 (Table 3). Alongside the study on recombinant dUGT1As, glucuronidation rates of the same eight compounds were measured in dog intestine and liver microsomes. A comparison between the rates in these tissues shows that compounds 1, 2 and 3 were glucuronidated 3–10 times faster in liver microsomes than in microsomes of segment 2 of the small intestine (Figure 4A). The glucuronidation rates of 5, 6, 7 and 8 were almost equal between liver and intestine segment 2 microsomes. To evaluate glucuronidation in different segments of the intestine, the enzyme activity was measured for compounds 1, 2, 5 and 8 in microsomes from five equally long segments of the small intestine (S1–S5) and colon (S6), from four dogs. Glucuronidation rate of 8 was higher in segments 1–5 than in colon and liver (Figure 4B). The glucuronidation rate of 5 was almost equal in segments 3–4 and colon, while it was somewhat lower in segments 1–2, 5 and liver (Figure 4C). For 1 and 2 the rates were higher in colon and liver than in segments 1–5 (Figure 4D, 4E). The data was analyzed by the PCA model with five components (cumulative R2X = 0.99 and cumulative Q2 = 0.77). Compounds 1 and 2 were strongly connected with liver and colon microsomes, while compound 8 with intestine microsomes, whereas the connection of compound 5 was stronger with intestine than with liver microsomes (Figure 5). The effect of concentrations of compounds 1, 2 and 5 on glucuronidation rate was evaluated in the dog intestine and liver microsomes. The shapes of the curves were compatible with the Michaelis- Menten equation. The goodness of fit was high for both intestine and liver glucuronidation of all three compounds (Table 4, Supplement Figure 2). Eadie-Hofstee analyses revealed deviation from a 14 Wu B, Kulkarni K, Basu S, Zhang S, Hu M. First-pass metabolism via UDP- glucuronosyltransferase: a barrier to oral bioavailability of phenolics. J Pharm Sci. 2011;100(9):3655-81. 15 Figure legends Figure 1. Structures of the 7-hydroxycoumarin derivatives. Figure 2. Glucuronidation of eight 7-hydroxycoumarin derivatives by ten dog UGT1A enzymes. Glucuronidation was determined at 10 µM substrate concentration. 16 Figure 3: Biplot of the principal component analysis (PCA) of eight 7-hydroxycoumarin derivative glucuronidation by dog UGT1As. The biplot shows scores for enzyme activity means and loadings for the two first latent components of the PCA model explaining 39% and 26% of the variance in the data, respectively. Substrates are shown as blue circles and UGT1A enzymes as black hexagons. 17 Figure 4. Glucuronidation of 7-hydroxycoumarin derivatives in microsomes of dog small intestine (S1–S5), colon (S6) and liver. Panel A shows the ratio of glucuronidation rates between liver and intestine segment 2 at 10 µM substrate concentration in one dog. Panels B – C show the glucuronidation rate of the indicated substrates in microsomes of intestinal segment or liver (means of four different dogs). 18 Figure 5. Biplot of the principal component analysis model of glucuronidation of 7- hydroxycoumarin derivatives in microsomes of dog intestine or liver. First two components (explaining 58% and 21% of variance, respectively) of a PCA model (5 components, cumulative R2X = 0.99, cumulative Q2 = 0.77). Substrates 1 (brown hexagons) and 2 (orange hexagons) show association with liver microsomes and substrates 5 (light blue hexagons) and 8 (dark blue hexagons) with intestine microsomes (sample types shown as gray circles). Supplement figure 1. Michaelis-Menten plots of glucuronidation of eight 7-hydroxycoumarin derivatives by the most efficient dog UGT enzymes. Supplement figure 2. Michaelis-Menten plots and Eadie-Hofstee analysis of glucuronidation of compounds 1, 2 and 5 by the liver and intestine segment 2 microsomes. 19 Table 1. Glucuronidation of eight 7-hydroxycoumarin derivatives by ten dog UGT enzymes. Glucuronidation rate (nmol/(min*g protein) was determined at 10 µM substrate concentration. The substrates are organized according to the selectivity with compound 8 showing most selectivity (for UGT1A11). Substrates on the right are the least selective. The grey scale indicates the rate of glucuronidation (white = lowest; black = highest). 20 Table 2. Michaelis-Menten kinetic constants of glucuronidation of different 7-hydroxycoumarin substrates by dog UGT1A enzymes. dUGT1A Compound Km (95 % confidence limits) µM Vmax (95 % confidence limit) Normalized rate µmol/(min*g protein) 1 8 3.6 (2.0-5.2) 0.012 (0.009-0.014) 11 8 1.8 (1.2-2.5) 0.049 (0.043-0.055) 2 2 10.2 (6.5-14) 0.19 (0.15-0.22) 10 2 1.1 (0.5-1.7) 0.10 (0.09-0.11) 11 2 3.1 (1.7-4.6) 0.078 (0.07-0.09) 6 1 22 (11-33) 0.18 (0.12-0.24) 10 1 5.1 (3.2-7.1) 0.24 (0.21-0.28) 11 1 17 (11-23) 0.31 (0.25-0.38) 2 6 4.3 (3.6-5.1) 0.074 (0.67-0.08) 7 6 4.9 (2.8-7.0) 0.035 (0.029-0.040) 11 6 2.5 (0.99-4.1) 0.056 (0.043-0.068) 1 7 5.5 (4.1-6.9) 0.072 (0.065-0.080) 2 7 9.9 (6.4-13) 0.24 (0.2-0.28) 10 7 4.3 (2.1-6.5) 0.056 (0.046-0.067) 11 7 2.1 (1.6-2.8) 0.072 (0.066-0.078) 1 3 40 (0-167) 0.071 (0-0.24) 2 3 87 (0-320) 0.29 (0-0.92) 10 3 2.4 (1.7-3.0) 0.096 (0.089-0.10) 11 3 4.6 (3.6-5.7) 0.16 (0.15-0.18) 2 4 7.2 (4.3-10) 0.21 (0.17-0.25) 3 4 19 (4-35) 0.24 (0.12-0.36) 7 4 250 (0-1270) 1.6 (0-7.6) 10 4 5.0 (2.1-7.9) 0.081 (0.063-0.099) 11 4 1.7 (0.86-2.6) 0.086 (0.075-0.098) 1 5 2.6 (1.4-3.9) 0.48 (0.41-0.56) 2 5 4.2 (3.5-4.9) 0.20 (0.18-0.21) 3 5 9.4 (6.2-11) 0.18 (0.15-0.21) 7 5 8.5 (7.0-10) 0.051 (0.047-0.055) 10 5 3.8 (2.4-5.2) 0.093 (0.082-0.1) 11 5 1.3 (1.4-2.1) 0.10 (0.094-0.11) 21 Table 3. Comparison of normalized intrinsic clearances. The values (Vmax/Km) were calculated from the data in Table 2. Normalized intrinsic clearance of compounds ml/(min*g protein) dUGT 8 1 2 6 3 7 4 5 1A1 3.3 ND ND ND 1.8 13 ND 180 1A2 ND ND 19 17 3.3 24 29 48 1A3 ND ND ND ND ND ND 13 19 1A4 ND ND ND ND ND ND ND ND 1A6 ND 8.2 ND ND ND ND ND ND 1A7 ND ND ND 7.1 ND ND 6.4 6 1A8 ND ND ND ND ND ND ND ND 1A9 ND ND ND ND ND ND ND ND 1A10 ND 47 91 ND 40 13 16 24 1A11 27 18 25 22 35 34 51 77 ND = Not determined, because glucuronidation rate was low at 10 µM compound screening. 22 Table 4. Michaelis-Menten kinetic constants of glucuronidation of substrates 1, 2 and 5 by intestine and liver microsomes of dogs. Microsomes Compound Km (95 % confidence limit) µM Vmax (95 % confidence limit) µmol/(min*g protein) Vmax / Km L/(min*g protein) R square Intestine 5 4.3 (1.6–7.0) 3.3 (2.6–4.1) 0.77 0.9334 –0.9877 Liver 5 12.8 (7.5–18.1) 2.4 (1.9–3.0) 0.19 0.9440– 0.9953 Intestine 1 40 (0–129) 0.93 (0–2.5) 0.023 0.9789– 0.9982 Liver 1 77 (0–162) 10.5 (0.8–20.1) 0.14 0.9859– 0.9963 Intestine 2 14 (0–32) 2.2 (0.6–3.7) 0.15 0.9916– 0.9964 Liver 2 120 (0–270) 44 (0–93) 0.36 0.9975– 0.9995 23 Table 5. Comparison of glucuronidation of 7-hydroxycoumarin derivatives between dog and human. The human data is from the studies Juvonen et al. 2018a and Juvonen et al. 2019. Liver/intestine ratio Liver/Intestine ratio The most efficient dUGT The most efficient hUGT 7-Hydroxy- coumarin derivative Dog Human Dog Human 1 >3 >3 1A10, 1A11>1A6 1A6>1A9, 1A10 2 >3 >2 1A2, 1A10, 1A11 1A6, 1A10>1A7, 1A9, 2A1 3 >3 ND 1A10, 1A11 > 1A1, 1A2 1A6>1A7, 1A8, 1A9, 1A10, 2B17 4 ND ND 1A2> 1A3> 1A1, 1A7, 1A10, 1A11 1A10 5 ≈1 <1 1A2> 1A1, 1A3, 1A7, 1A10, 1A11 1A10>>1A1 6 ≈1 <1 1A2, 1A7, 1A8>1A3, 1A11 1A10>>1A1 7 ND <1 1A2> 1A1, 1A11 > 1A3, 1A7, 1A8, 1A10 1A10>>1A1 8 ≈1 <1 1A11>1A1 1A10 ND means not determined.