Kayoko Ohura

Species Difference of Esterase Expression and Hydrolase Activity in Plasma

Differences in esterase expression among human, rhesus monkey, cynomolgus monkey, dog, minipig, rabbit, rat, and mouse plasma were identified using native polyacrylamide gel electrophoresis. Paraoxonase (PON) and butyrylcholinesterase (BChE) were ubiquitous in all species, but were highly expressed in primates and dogs, whereas carboxylesterase (CES) was only abundant in rabbits, mice, and rats. Several unknown esterases were observed in minipig and mouse plasma. These differences in plasma esterases and their expression levels result in species differences with respect to hydrolase activity. These differences were characterized using several different substrates. In contrast to the high hydrolase activity found for p-nitrophenylacetate (PNPA), a substrate of several hydrolase enzymes, irinotecan, a carbamate compound, was resistant to all plasma esterases. Oseltamivir, temocapril, and propranolol (PL) derivatives were rapidly hydrolyzed in mouse and rat plasma by their highly active CES enzyme, but rabbit plasma CES hydrolyzed only the PL derivatives. Interestingly, PL derivatives were highly hydrolyzed by monkey plasma BChE, whereas BChE from human, dog, and minipig plasma showed negligible activity. In conclusion, the esterase expression and hydrolyzing pattern of dog plasma were found to be closest to that of human plasma. These differences should be considered when selecting model animals for preclinical studies.

The Role of Intestinal Carboxylesterase in the Oral Absorption of Prodrugs

The bioavailability of therapeutic agents can be improved by using prodrugs which have better passive diffusion than the active agents. Intestinal hydrolysis is an important reaction in the bioconversion of prodrugs, and may be the rate-limiting factor in their absorption. Carboxylesterase (CES) is ubiquitous in most organs and is located in the endoplasmic reticulum. Single-pass perfusion experiments in rat intestine have shown that CES is the main enzyme involved in intestinal first-pass hydrolysis. In man, intestinal CESs belong to the CES2 gene family and their activity is nearly constant along the jejunum and ileum. The predominant human intestinal CES, hCE2, preferentially hydrolyzes prodrugs in which the alcohol group of a pharmacologically active molecule has been modified by the addition of a small acyl group. In preclinical animal models, CES2 isozymes are also the major intestinal enzymes although they have different substrate specificities to human CES2, while CES1 isozymes and other unidentified enzymes are also present. It is therefore difficult to predict human intestinal absorption from animal experiments. Caco-2 cells mainly express the human CES1 isozyme, hCE1, which shows quite different substrate specificity from hCE2, making Caco-2 cells unsuitable for prediction of human intestinal absorption of prodrugs. However, we have developed a novel experimental method for predicting the human intestinal absorption of prodrugs using Caco-2 cells in which CES-mediated hydrolysis has been inhibited. The expression of hCE2 shows inter-individual variation and is regulated by several mechanisms, such as gene polymorphism and epigenetic processes. There are no reports suggesting that severe toxicity is associated with prodrugs due to genetic polymorphism of the CES2 gene.

Development of a novel system for estimating human intestinal absorption using Caco-2 cells in the absence of esterase activity

Both mRNA and protein levels of the carboxylesterase (CES) isozymes, hCE1 and hCE2, in Caco-2 cells increase in a time-dependent manner, but hCE1 levels are always higher than those of hCE2. In human small intestine, however, the picture is reversed, with hCE2 being the predominant isozyme. Drugs hydrolyzed by hCE1 but not by hCE2 can be hydrolyzed in Caco-2 cells, but they are barely hydrolyzed in human small intestine. The results in Caco-2 cells can be misleading as a predictor of what will happen in human small intestine. In the present study, we proposed a novel method for predicting the absorption of prodrugs in the absence of CES-mediated hydrolysis in Caco-2 cells. The specific inhibition against CES was achieved using bis-p-nitrophenyl phosphate (BNPP). The optimal concentration of BNPP was determined at 200 μM by measuring the transport and hydrolysis of O-butyryl-propranolol (butyryl-PL) as a probe. BNPP concentrations of more than 200 μM inhibited 86% of hydrolysis of butyryl-PL, resulting in an increase in its apparent permeability. Treatment with 200 μM BNPP did not affect paracellular transport, passive diffusion, or carrier-mediated transport. Furthermore, the proposed evaluation system was tested for ethyl fexofenadine (ethyl-FXD), which is a superior substrate for hCE1 but a poor one for hCE2. CES-mediated hydrolysis of ethyl-FXD was 94% inhibited by 200 μM BNPP, and ethyl-FXD was passively transported as an intact prodrug. From the above observations, the novel evaluation system is effective for the prediction of human intestinal absorption of ester-type prodrugs.

Evaluation of Transport Mechanism of Prodrugs and Parent Drugs Formed by Intracellular Metabolism in Caco-2 Cells with Modified Carboxylesterase Activity: Temocapril as a Model Case

The intestinal absorption mechanism of temocapril, an ester-type prodrug of temocaprilat, was evaluated using Caco-2 cell monolayers with or without active carboxylesterase (CES)-mediated hydrolysis. The inhibition of CES-mediated hydrolysis was achieved by pretreatment of the monolayers with bis-p-nitrophenyl phosphate (BNPP), which inhibited 94% of the total hydrolysis of temocapril in the Caco-2 cells. The remaining 6% hydrolysis was due to the presence of serine esterases, other than CES, on the cell membranes. Transport experiments under CES-inhibited conditions showed temocapril not to be a substrate for peptide transporter 1 (PEPT1) or organic anion transporting polypeptides (OATPs), but to be an inhibitor of PEPT1; P-glycoprotein (P-gp) and breast-cancer-resistant protein (BCRP) were responsible for the efflux of temocapril, which was mainly absorbed by passive diffusion at low apical pH. In Caco-2 cell monolayers with CES-mediated hydrolysis intact, temocaprilat derived from temocapril, was 2.5-fold more rapidly transported into the apical compartment than into the basolateral compartment due to the presence of microvilli on the apical membrane. In contrast, temocaprilat at low intracellular concentrations, was preferentially transported across the basolateral membrane under CES-inhibited conditions.

Distinct patterns of aging effects on the expression and activity of carboxylesterases in rat liver and intestine.

The age-associated alteration in expression levels of carboxylesterases (CESs) can affect both intestinal and hepatic first-pass metabolism after oral administration of xenobiotic esters such as prodrugs. In this study, the age-related expression of CES isozymes and hydrolase activities were simultaneously investigated in liver, jejunum, and ileum from 8-, 46-, and 90-week-old rats. Rat liver expresses three major CES1 isozymes, Hydrolase A, Hydrolase B, and Hydrolase C, as well as one minor CES1 (Egasyn) and three minor CES2 isozymes (RL4, AY034877, and D50580). The mRNA and protein levels of major hepatic CES1 isozymes were decreased in an age-dependent manner, while those of minor CESs were maintained in all age groups. The hepatic hydrolase activity for temocapril was decreased in an age-dependent manner, accompanied by downregulation of Hydrolase B/C mRNA, while age-independent hydrolysis of propranolol derivatives was observed in rat liver, due to the contribution of Egasyn. Rat small intestine expresses one major CES2 (RL4) and four minor CESs (Hydrolase B, Hydrolase C, Egasyn, and AY034877). Interestingly, the expression of RL4 was age-dependently increased in both jejunum and ileum, while minor isozymes showed a constant expression across a wide age range. The up-regulation of RL4 expression with aging led to an increase of intestinal hydrolase activities for temocapril and propranolol derivatives. Consequently, age-dependent changes in the expression of CES isozymes affect the hydrolysis of xenobiotics in both rat liver and small intestine.

Effect of intestinal first‐pass hydrolysis on the oral bioavailability of an ester prodrug of fexofenadine

The contribution of intestinal first-pass hydrolysis to oral bioavailability was evaluated in rats using a model prodrug of fexofenadine (FXD), which has poor oral bioavailability. The prodrug, ethyl-FXD, has high membrane permeability but the oral bioavailability of FXD derived from ethyl-FXD was only 6.2%. Ethyl-FXD was not detected in the plasma, whereas FXD was detected, indicating complete first-pass hydrolysis. In in vitro experiments, hydrolase activity for ethyl-FXD was higher in the liver and blood than that in the intestine. However, the high blood protein binding of ethyl-FXD resulted in a high hepatic availability (F(h) = 88%). The complete bioconversion of ethyl-FXD in the in vivo oral administration is difficult to explain by first-pass hydrolysis in the liver and blood. Interestingly, in an in situ rat jejunal single-pass perfusion experiment, 84% of the ethyl-FXD taken up into enterocytes was hydrolyzed. Furthermore, only one-fifth of the FXD formed in mucosa reached the mesenteric vein because of its P-glycoprotein-mediated efflux into the intestinal lumen. These findings indicate that the intestinal bioconversion of ester prodrugs to their parent drugs is a key factor in determining their oral bioavailability.

Design of Fexofenadine Prodrugs Based on Tissue‐Specific Esterase Activity and Their Dissimilar Recognition by P‐Glycoprotein

The aim of this study was to develop a suitable prodrug for fexofenadine (FXD), a model parent drug, that is resistant to intestinal esterase but converted to FXD by hepatic esterase. Carboxylesterases (CESs), human carboxylesterase 1 (hCE1) and human carboxylesterase 2 (hCE2), are the major esterases in human liver and intestine, respectively. These two CESs show quite different substrate specificities, and especially, hCE2 poorly hydrolyzes prodrugs with large acyl groups. FXD contains a carboxyl group and is poorly absorbed because of low membrane permeability and efflux by P-glycoprotein (P-gp). Therefore, two potential FXD prodrugs, ethyl-FXD and 2-hydroxyethyl-FXD, were synthesized by substitution of the carboxyl group in FXD. Both derivatives were resistant to intestinal hydrolysis, indicating their absorption as intact prodrugs. Ethyl-FXD was hydrolyzed by hepatic hCE1, but 2-hydroxyethyl-FXD was not. Both derivatives showed high membrane permeability in human P-gp-negative LLC-PK1 cells. In LLC-GA5-COL300 cells overexpressing human P-gp, ethyl-FXD was transported by P-gp, but its efflux was easily saturated. Whereas 2-hydroxyethyl-FXD showed more efficient P-gp-mediated transport than FXD. Although the structure of 2-hydroxyethyl-FXD only differs from ethyl-FXD by substitution of a hydroxyl group, 2-hydroxyethyl-FXD is unsuitable as a prodrug. However, ethyl-FXD is a good candidate prodrug because of good intestinal absorption and hepatic conversion by hCE1.

3D-fibroblast tissues constructed by a cell-coat technology enhance tight-junction formation of human colon epithelial cells

Caco-2, human colon carcinoma cell line, has been widely used as a model system for intestinal epithelial permeability because Caco-2 cells express tight-junctions, microvilli, and a number of enzymes and transporters characteristic of enterocytes. However, the functional differentiation and polarization of Caco-2 cells to express sufficient tight-junctions (a barrier) usually takes over 21 days in culture. This may be due to the cell culture environment, for example inflammation induced by plastic petri dishes. Three-dimensional (3D) sufficient cell microenvironments similar to in vivo natural conditions (proteins and cells), will promote rapid differentiation and higher functional expression of tight junctions. Herein we report for the first time an enhancement in tight-junction formation by 3D-cultures of Caco-2 cells on monolayered (1L) and eight layered (8L) normal human dermal fibroblasts (NHDF). Trans epithelial electric resistance (TEER) of Caco-2 cells was enhanced in the 3D-cultures, especially 8L-NHDF tissues, depending on culture times and only 10 days was enough to reach the same TEER value of Caco-2 monolayers after a 21 day incubation. Relative mRNA expression of tight-junction proteins of Caco-2 cells on 3D-cultures showed higher values than those in monolayer structures. Transporter gene expression patterns of Caco-2 cells on 3D-constructs were almost the same as those of Caco-2 monolayers, suggesting that there was no effect of 3D-cultures on transporter protein expression. The expression correlation between carboxylesterase 1 and 2 in 3D-cultures represented similar trends with human small intestines. The results of this study clearly represent a valuable application of 3D-Caco-2 tissues for pharmaceutical applications.

Expression of Carboxylesterase Isozymes and Their Role in the Behavior of a Fexofenadine Prodrug in Rat Skin

The expression of carboxylesterase (CES) and the transdermal movement of an ester prodrug were studied in rat skin. Ethyl-fexofenadine (ethyl-FXD) was used as a model lipophilic prodrug that is slowly hydrolyzed to its parent drug, FXD (MW 502). Among the CES1 and CES2 isozymes, Hydrolase A is predominant in rat skin and this enzyme was involved in 65% of the cutaneous hydrolysis of ethyl-FXD. The similarity of the permeation behavior of ethyl-FXD in full thickness and stripped skin indicated that the stratum corneum was not a barrier to penetration. However, only FXD was observed in receptor fluid, not ethyl-FXD, presumably because of the high degree of binding of ethyl-FXD in viable skin. The rate of hydrolysis of ethyl-FXD was much faster than steady-state flux, such that the influx rate was the rate-limiting process for transdermal permeation. Although Hydrolase A levels gradually increased in skin taken from rats aged from 8 to 90 weeks, variations in the expression levels of the esterase hardly affected the conversion of prodrug. The present data suggest that the slow hydrolysis of the prodrug of an active ingredient in viable skin followed by slow diffusion of active drug may provide a useful approach to topical application.

Differences in Intestinal Hydrolytic Activities between Cynomolgus Monkeys and Humans: Evaluation of Substrate Specificities Using Recombinant Carboxylesterase 2 Isozymes

Cynomolgus monkeys, used as an animal model to predict human pharmacokinetics, occasionally show different oral absorption patterns to humans due to differences in their intestinal metabolism. In this study, we investigated the differences between intestinal hydrolytic activities in cynomolgus monkeys and humans, in particular the catalyzing activities of their carboxylesterase 2 (CES2) isozymes. For this purpose we used both human and monkey microsomes and recombinant enzymes derived from a cell culture system. Monkey intestinal microsomes showed lower hydrolytic activity than human microsomes for several substrates. Interestingly, in contrast to human intestinal hydrolysis, which is not enantioselective, monkey intestine showed preferential R-form hydrolysis of propranolol derivatives. Recombinant CES2 isozymes from both species, mfCES2v3 from monkeys and human hCE2, showed similar metabolic properties to their intestinal microsomes when expressed in HEK293 cells. Recombinant hCE2 and mfCES2v3 showed similar Km values for both enantiomers of all propranolol derivatives tested. However, recombinant mfCES2v3 showed extreme R-enantioselective hydrolysis, and both hCE2 and mfCES2v3 showed lower activity for O-3-methyl-n-butyryl propranolol than for O-n-valeryl and O-2-methyl-n-butyryl propranolol. This lower hydrolytic activity was characterized by lower Vmax values. Docking simulations of the protein-ligand complex demonstrated that the enantioselectivity of mfCES2v3 for propranolol derivatives was possibly caused by the orientation of its active site being deformed by an amino acid change of Leu107 to Gln107 and the insertion of Met309, compared with hCE2. In addition, molecular dynamics simulation indicated the possibility that the interatomic distance between the catalytic triad and the substrate was elongated by a 3-positioned methyl in the propranolol derivatives. Overall, these findings will help us to understand the differences in intestinal hydrolytic activities between cynomolgus monkeys and humans.