Review Human Carboxylesterase Isozymes: Catalytic Properties and Rational Drug Design

Drug Metab. Pharmacokinet. 21 (3): 173–185 (2006).

Teruko IMAI Graduate School of Pharmaceutical Sciences, Kumamoto University, Kumamoto, Japan

Full text of this paper is available at http://www.jstage.jst.go.jp/browse/dmpk

Summary: Human carboxylesterase 1 (hCE-1, CES1A1, HU1) and carboxylesterase 2 (hCE-2, hiCE, HU3)areaserineesteraseinvolvedinbothdrugmetabolismandactivation.AlthoughbothhCE-1and hCE-2 are present in several organs, the hydrolase activity of liver and small intestine is predominantly attributed to hCE-1 and hCE-2, respectively. The substrate speciˆcity of hCE-1 and hCE-2 is signiˆcant- ly diŠerent. hCE-1 mainly hydrolyzes a substrate with a small alcohol group and large acyl group, but its wide active pocket sometimes allows it to act on structurally distinct compounds of either large or small alcohol moiety. In contrast, hCE-2 recognizes a substrate with a large alcohol group and small acyl group, and its substrate speciˆcity may be restricted by a capability of acyl-hCE-2 conjugate formation due to the presence of conformational interference in the active pocket. Furthermore, hCE-1 shows high transesteriˆcation activity, especially with hydrophobic alcohol, but negligible for hCE-2. Transesteriˆ- cation may be a reason for the substrate speciˆcity of hCE-1 that hardly hydrolyzes a substrate with hydrophobic alcohol group, because transesteriˆcation can progress at the same time when a compound is hydrolyzed by hCE-1. From the standpoint of drug absorption, the intestinal hydrolysis by CES during drug absorption is evaluated in rat intestine and Caco2-cell line. The rat in situ single-pass perfusion shows markedly extensive hydrolysis in the intestinal mucosa. Since the hydrolyzed products are present at higher concentration in the epithelial cells rather than blood vessels and intestinal lumen, hydrolysates are transported by a speciˆc eŒux transporter and passive diŠusion according to pH-partition. The expression pattern of CES in Caco-2 cell monolayer, a useful in vitro model for rapid screening of human intestinal drug absorption, is completely diŠerent from that in human small intestine but very similar to human liver that expresses a much higher level of hCE-1 and lower level of hCE-2. Therefore, the prediction of human intestinal absorption using Caco-2 cell monolayers should be carefully monitored in the case of ester and amide-containing drugs such as prodrugs. Further experimentation for an understanding of detailed substrate speciˆcity for CES and development of in vitro evaluation systems for absorption of prodrug and its hydrolysates will help us to design the ideal prodrug.

Key words: carboxylesterase; prodrug; hydrolysis; substrate speciˆcity; intestinal absorption

requirements because they play an important role in Introduction biotransformation of a variety of ester-containing drugs The hydrolase activity within several tissues is increas- and prodrugs4–6) such as an angiotensin-converting en- ingly used as the basis for drug design, particularly on zyme inhibitor (temocapril, cilazapril, quinapril and im- prodrugs and softdrugs containing functional groups idapril),7) anti-tumor drugs (CPT-11 and capecitabin)8,9) such as carboxylic acid ester.1–3) Introduction of an ester and narcotics (cocain, heroin and meperidine).10,11) linkage generally improves the bioavailability of ther- CESs are members of the aWb hydrolase fold family apeutic agents due to increased passive transport and show ubiquitous tissue expression proˆles with high following oral administration. Carboxylesterases levels in the liver, small intestine and lung.12,13) The (CESs, EC.3.1.1.1) are essential to achieve these mammalian CESs comprise a multigene family, and the

Received; March 14, 2006, Accepted; June 6, 2006 To whom correspondence should be addressed: Teruko IMAI, Graduate School of Pharmaceutical Sciences, Kumamoto University, 5-1 Oe- honmachi, Kumamoto 862-0973, Kumamoto, Japan. Tel. & Fax. ＋81-96-371-4626, E-mail: iteruko＠gpo.kumamoto-u.ac.jp

173 174 Teruko IMAI

Table 1. MolecularpropertyofhCE-1andhCE-2

hCE-1 ref hCE-2 ref Molecular weight 180 kDa 18, 32 60 kDa 10 Subunit weight 60 kDa Isoelectric point 5.6–5.8 18, 32 4.8–5.0 10 Glycosylation site Asn-X-Thr 10 Asn-X-SerWThr 10 Asn79 Asn103, Asn267 Optimal pH 6.5 18 7.5–8.0 18 isozymes are classiˆed into four main CES groups (CES1-CES4) and several subgroups according to the homology of the amino acid sequence.10,12) The majority of CESs has been segregated into the CES1 and CES2 families. Recent studies have shown some diŠerences between these CES1 and CES2 families in terms of substrate speciˆcity, tissue distribution, immunological properties and gene regulation. Fig. 1. Contribution of carboxylesterase in the hydrolase activity of This review focuses on the molecular characteristics humanliverandsmallintestine. of human CES and the contribution of CES isozymes in Each column represents a relative activity for hydrolysis of p- the hydrolase activity of the human liver and small nitrophenylacetate as percent of control hydrolase activity in human intestine. In addition, it explains the extensive intestinal liver and small intestine. ``Hydrolysis by CES'' is represented as an inhibited hydrolase activity after inhibition by bis-p-nitrophenyl ˆrst-pass hydrolysis and the di‹culty of prediction of phosphate. Black and hatched column represent a relative hydrolase human intestinal absorption of ester-containing drugs activity of hCE-1 and hCE-2, respectively. using Caco-2 cells. Tissue Distribution of Human Carboxylesterase is detected in the blood of humans.22–24) Recent molecular biology techniques have identiˆed Contribution of Carboxylesterase on Hydrolase the two major human CESs, hCE-1 (CES1A1, HU1), a Activity in the Liver and Small Intestine human CES1 family isozyme, and hCE-2 (hiCE, HU3), a human CES2 family isozyme. hCE-1 is highly The hydrolase activity in the liver and small intestine expressedintheliverandalsoobservedinmacrophages, in mammals is attributable to several esterase molecules. human lung epithelia,14) heart, testis and other tissues.13) However, when native PAGE gel of microsomes is hCE-1 expression is markedly low in the gastrointestinal stained by esterase activity using 1-naphthylacetate, tract. hCE-2 is present in the small intestine, colon, predominant bands are corresponding CES isozymes.25) kidney, liver, heart, brain and testis. hCE-2 expression Also, 80–95z andmorethan90z of hydrolase activity is essentially absent in all other organs.15,16) hCE-1 and is inhibited by bis-p-nitrophenyl phosphate, a speciˆc hCE-2 share 48z amino acid sequence identity and are CES inhibitor,26,27) in rat liver and small intestine,28) predicted to be glycoproteins of 60kDa.17,18) Some respectively. The content of CES in rat liver is found to molecular properties of hCE1 and hCE2 are listed in be about 1 mg per g of fresh tissue while microsomal Table 1. A third, brain-speciˆc CES was isolated in fraction contains about 30 mg of CES per g of 199919) andtermedhBr3(hCE3).However,relatively microsomal protein.29) Thus, CES activity is signiˆcant- little characterization of this CES has been reported. ly high in comparison with other esterases. Although the hBr3 shares 77z and 49z sequence identity with hCE-1 liver and small intestine of mammals contains CES1 and and hCE-2, respectively. Furthermore, a new class of CES2 enzymes, no CES isozyme is present in the small human CES, CES3, was isolated in 2004.20) CES3 has intestine of the beagle dog. Therefore, the beagle dog about 40z identity with both hCE-1 and hCE-2, and is shows a diŠerent absorption behavior of some drugs expressed in the liver and gastrointestinal tract at an containing esters such as prodrugs in comparison with extremely low level in comparison with hCE-1 and other animals. hCE-2.20,21) In humans, the native PAGE gel of tissue microsomes Typically, expression of CES is maximal in the shows only one band (hCE-2) for human small intestine epithelia of most organs, suggesting that these enzymes but two bands (hCE-1, upper strong band; hCE-2, play a protective role against xenobiotics. In addition, lower weak band) for human liver after visualization of although high levels of CES activity can be detected in esterase activity with staining by 1-naphthylacetate, the blood of the majority of mammals, no such activity indicating the predominant esterase is CES isoforms in Hydrolysis of Prodrug by Carboxylesterase 175

Fig. 2. The two-step catalytic mechanism of mammalian carboxylesterase. Bold arrow shows the typical route of hydrolysis. In the ˆrst step (Step 1), the active site serine hydroxyl group attacks the carboxylester linkage in the substrate to generate the alcohol product and the covalent acyl-enzyme intermediate. The second step (Step 2), the second substrate, water, attacks the covalent acyl-enzyme intermediate to produce acyl product, and serine residue changes to original state. However, when an alcohol presents in abundance, the enzyme can facilitate a transesteriˆcation reaction to generate a new ester product. The alcohol produced in ``Step1'' is possible to react with acyl-enzyme intermediate to generate an original ester compound. human liver and small intestine.30) The hydrolysis much higher than hCE2 even by considering the pattern for several substrates in the human small intes- variation of expression level. tine microsomes is nearly the same as those of recom- Carboxylesterase Reactions binant hCE-2, as expected by result from the native PAGE.30) Although human liver microsomes express Catalysis of ester cleavage by CESs is base-mediated, both hCE-1 and hCE-2, their substrate speciˆcity requiring water as a co-reactant. This reaction is closely resembles recombinant hCE-1. Furthermore, achieved via a triad of catalytic amino acids (Ser, His anti-hCE-1 antibody inhibited 80–95z of the hepatic and Glu) that are all essential for enzymatic activity.12) hydrolysis. Therefore, hCE-1 predominantly contrib- As shown in Fig. 2, CES cleaves esters via a two-step utes to a hepatic hydrolysis, and the residual hydrolase reaction.AtneutralpH,theactivesiteglutamicacid activity is due to hCE-2. The contributionz of CES to exists as the charged form, which facilitates removal of the hydrolase activity in human liver and small intestine a proton from histidine. This loss subsequently results microsomes is shown in Fig. 1.BothhCE-1andhCE-2 in transfer of a proton from the adjacent serine to the are highly variable in the individual liver. Xu et al.15) opposing nitrogen of histidine, generating an oxygen reported a 3-fold range variance for hCE-2 among 13 nucleophile that can attack the carbonyl carbon of the human liver microsomes, and Hosokawa et al.31) report- substrate. When the tetrahedral intermediate of an acyl ed a more than 8-fold range variance of hCE-1 protein group is formed, the alcohol product is released from levels among 12 human liver microsomes. Therefore, the enzyme. The acyl-enzyme intermediate is then hydrolase activity is variable in the individual, and attacked in an identical fashion with water acting as the contribution of hCE1 to hepatic hydrolysis might be nucleophile, leading to release of the carboxylic acid 176 Teruko IMAI

Fig. 3. The schematic structure of hCE-1 in the homatropine complex. a) Entrance of active site. A substrate can enter from the face of view, then a1anda10? cover the active site. Three amino acid residues of catalytic triad (Ser, His, Glu) are closely lined in three-dimensional picture. b) Z-site surface ligand binding pocket. Two loops, V-I (355–372) and V-II (452–469), are framed at the binding site, and homatropine bound in Z-site is illustrated.

and return of the catalytic amino acids to their original for retention in the luminal site of the endoplasmic state. reticulum.39) The three amino acid residues (Ser, His and Furthermore, CES, especially hCE-1, can perform Glu) of the catalytic triad and four cysteines that may transesteriˆcation reactions. When alcohol presents in be involved in speciˆc disulˆde bonds are similarly abundance, an alcohol can attack the acyl-enzyme positioned in each CES. It is interesting that hCE-1 has intermediate to generate an ester product. The most a free cysteine residue that is not involved in a disulˆde well-researched example is the transesteriˆcation of bond. Typically, CESs are glycoproteins, and biochemi- cocaine with ethanol to generate cocaethylene that is cal studies have shown that the carbohydrate modiˆca- more toxic than the parent drug.32–34) Similar tran- tions are required for enzyme activity.40) hCE-2 contains sesteriˆcation appears to have a role in the ability of glycosylation sites at two positions (residues 103 and CES to generate ester products from endogenous 267), while hCE-1 maintains only one glycosylation site compounds. For example, hCE-1 possesses acyl coen- at Asn79. Asn79 is modiˆed by a carbohydrate chain zyme A:cholesterol acyltransferase activity, which with ˆrst N-acetylglucosamine (NAG) and terminal generates cholesterol esters from fatty-acyl coenzyme A sialic acid and appears to be involved in the stabilization and free cholesterol.35) Also, hCE-1 generates fatty acid of the hCE-1 trimer by packing into the adjacent ethyl esters (FAEEs) from fatty acyl-coenzyme A and monomer in its crystal structure.34,41,42) ethanol.36,37) FAEEs are toxic to numerous tissues and The ˆrst crystal structure of mammalian CES, that of are thought to be associated with the necrotic decay of rabbit CES1, was reported in 200243) and revealed that the liver and other tissues related to chronic alcohol CESs share the serine hydrolase fold observed in other abuse.38) esterases such as butylcholinesterase and acetyl- cholinesterase. Recently, the structure of hCE-1 com- Structure of Carboxylesterase plexed with several drugs has been reported,34,41,42) while The mammalian CESs are localized in the endoplas- the structure of hCE-2 has not been reported. The mic reticulum of many tissues. CESs contain an N-ter- crystal structure of hCE1 in the homatropine complex is minal hydrophobic signal peptide that marks them for schematically shown in Fig. 3. hCE-1 includes a central tra‹cking through the endoplasmic reticulum. In addi- 15-stranded b-sheet surrounded by numerous a-helix tion, a His-X-Glu-Leu (HXEL) sequence present at the and b-strands and consists of a central catalytic domain, C-terminus of the protein can bind with KDEL receptor ab domain and regulatory domain. The active site cavity Hydrolysis of Prodrug by Carboxylesterase 177

Fig. 4. Structure-activity relationship of substrate with hCE-1 and hCE-2. is ¿15Å deep and is located at the interface of several door in hCE-1 is separated from the active site cavity by protein domains where it is lined predominantly by a thin wall composed structurally of four ‰exible amino hydrophobic amino acids, with the exception of the acids, Thr252, Val254, Leu388 and Met425. Bencharit residues in the catalytic triad. It has been suggested that et al. has proposed34) that this side door might play a the entrance of the active site in hCE-1 is covered by two role as a gate or passage of substrates and products. regions, a1 (88–103) and a10? (353–366) after binding of Numerous crystal structures of CESs complexed with a a substrate41) (Fig. 3a). Furthermore, the active site variety of small molecules provides detailed insights into cavity of hCE-1 is large (¿1300 Å3 in volume) and the mechanism and function of CESs in several biologi- contains a speciˆc region responsible for binding small cal processes. moieties such as methyl and ethyl groups and a larger, Substrate Speciˆcity of hCE-1 and hCE-2 more ‰exible region that can accommodate much larger functional groups. A second surface ligand binding site hCE-1 and hCE-2 exhibit 48z sequence identity, and on hCE-1, termed the Z-site, was identiêd in the their distinct substrate speciˆcity has been proposed. central catalytic domain (Fig. 3b). This Z-site is formed hCE-1 preferentially catalyses the hydrolysis of by the inter-digitation of the V1andV2 loops, and it compounds esteriêd with a small alcohol group, while has been proposed that this site can act as an allosteric hCE-2 hydrolyzes compounds with a relatively small site.34,42) TheZ-siteislocated¿15 Å away from the acyl group and large alcohol group.7,10,13,44) Figure 4 active site with only the a10? helix separating the two showsexamplesofsubstrateforhCE-1andhCE-2.In features. Moreover, it has been reported that a side door the case of cocaine, hCE-1 catalyzes the hydrolysis of allows products or substrates to enter the active site.34) the methyl ester of cocaine producing benzoylecgonine The side door region of CES was ˆrst identiêd in the and methanol, and the benzoyl ester of cocaine is hydro- structure of rabbit CES1, which contained a CPT-11 lyzed by hCE-2.10,11,44) As shown in Fig. 4, compounds metabolite bound at this surface position.43) The side with a relatively large acyl group and small alcohol 178 Teruko IMAI

Fig. 5. Hydrolysis of benzoic acid derivatives in the recombinant hCE-1 and hCE-2. Protein concentration of cell homogenate expressed hCE-1 and hCE-2 was 40 mgWmL and 25 mgWmL, respectively. Substrate was applied at a concentration of 500 mM. Values represent the mean±S.D. (n＝3). PC was determined by partition of several compounds between n-octanol and pH7.4 phosphate buŠer. group such as methylphenidate,45) temocapril7) and ‰ur- ly distinct compounds containing either large or small biprofen derivatives,30) are barely recognized by hCE-2, alcohol moieties. Interestingly, hCE-1 enantioselectively while CPT-11,8), heroin,10) p-nitrophenylacetate46) and catalyzes the hydrolysis of a substrate. For example, S- propionyl-propranolol30) bearing a small acyl moiety propranolol derivatives, S-cocaine, d-methylphenidate and bulky alcohol group, are good substrates for and cis-cypermethrin analogues are poor substrates of hCE-2. Steric hindrance in the vicinity of the reaction hCE-1, in contrast to the opposing enantiomer.30,45,47,48) site of hCE-2 may occur with substrates containing a The diŠerences in hydrolysis rate between these enan- bulky acyl moiety in the process of the formation of tiomers have been explained by steric clashes with acyl-hCE-2 intermediate at the ˆrst step of hydrolysis. the loop containing Gly142 and Gly143 in the rigid Interestingly, the propranolol derivatives that contain a pocket,34,48) where Gly residues form the oxianion hole branched acyl moiety substituent with a methyl group at (140–144) to stabilize the transition state of substrate via the 3-position are scarcely hydrolyzed by recombinant their amide nitrogen. Bencharit et al. also identiˆed the hCE-2. In contrast, the same molecular mass Z-site surface as a ligand binding site for an inactive propranolol derivatives that possess branched acyl substrate.34) The enantioselective hydrolysis of these moiety with methyl groups at the 2-position are easily compounds is possibly explained by enantiomerically hydrolyzed at almost the same rate as the corresponding distinct active site orientations in hCE-1 due to structur- straight acyl derivatives.30) In general, the chemical al clashes. hydrolysis of ester bonds is sterically hindered by the Possibility of Transesteriˆcation during Hydrolysis substituent methyl group on the 2-position rather than by hCE-1 the 3-position. The both ˆndings that the speciˆc reduc- tion of hydrolysis rate by substitution of a methyl group Since hCE-1 possesses a ‰exible and rigid pocket, at the 3-position and the low hydrolysis rate for a sub- various compounds can bind to hCE1. However, a good strate with large acyl group suggest that the acyl-enzyme substrate for hCE-1 is a compound with large acyl intermediate is di‹cult to form due to the presence of group and small alcohol group rather than a compound conformational interference in the active site of hCE-2. with small acyl group, as shown in Fig. 4.Incompari- In contrast to hCE-2, hCE-1 preferentially recognizes son with a large acyl group, a small acyl group could substrates with a large acyl moiety such as methylpheni- easily binds to Ser residue of an active site in hCE-1 at date, temocapril and ‰urbiprofen derivatives (Fig. 4). the ˆrst step of hydrolysis (Step 1). Why is a compound However, substrates with a small acyl moiety, for with a small acyl group slowly hydrolyzed by hCE-1? In example R-propionyl propranolol, can also be hydro- order to investigate this in further detail, the hydrolysis lyzed by hCE-1. In the crystal structure of hCE-1,34,41) it of p-amino-, o-amino-, and p-hydroxy-benzoic acid has been reported that the substrate-binding site of derivatives has been determined in recombinant hCE-1 hCE-1 consisted of a ``small, rigid'' pocket and a and hCE-2.30) Interestingly, the hydrolysis activity of ``large, ‰exible'' pocket, and that the small, rigid pocket recombinant hCE-1 decreases with increasing in hCE-1 is selective, whereas the large, ‰exible pocket is hydrophobicity for all benzoic acid derivatives, as promiscuous with regard to substrate speciˆcity (see shown in Fig. 5. In contrast, hCE-2 hydrolyzes more Fig. 3). These pockets allow hCE-1 to act on structural- hydrophobic substrates. The enzyme kinetic parameters Hydrolysis of Prodrug by Carboxylesterase 179

butanol, is added, transesteriˆcation is easily performed by hCE-1, but hCE-2 shows much lower transesteriˆca- tion ability than hydrolysis. These ˆndings suggest that hCE-1 catalyzes transesteriˆcation with hydrophobic alcohol rather than hydrophilic alcohol, but hCE-2 possesses negligible transaesteriˆcation ability with hydrophobic alcohol. Therefore, the lowering hydrolyzing activity of hCE-1 for p-amino-benzoate derivatives with increasing alcohol chain length might be dependent on transesteriˆcation with a hydrophobic alcohol moiety. Bencharit et al.43) reported the crystal structure of rabbit liver CES (rCE), in which 4-piperidino piperi- dine, a product of CPT-11 activation, was bound between the ˆrst N-acetyl glucosamine of the Asn389 glycosylation site and the Trp550 side chain of the C- terminal helix. In two N-linked glycosylation sites in rCE at Asn residues 79 and 389, Asn 389 is modiˆed via a long carbohydrate chain and plays a role as a novel exit pore for the release of small products from the Fig. 6. Comparison of transesteriˆcation rate and hydrolysis rate in active site of the enzyme. hCE-2 contains glycosylation the presence of ethanol and butanol. sites at two positions (residues 103 and 267). Although Ordinate represents as a ratio of hydolysis of p-aminobonzoate there is no crystallographic data for hCE-2, one of two propyl ester and formation of p-aminobonzoate ester with ethanol or glycosylation sites might act as an exit pore for the butanol in the presence of corresponding alcohol. Circle and triangle release of alcohol. In contrast to hCE-2, hCE-1 main- show catalysis by hCE-1 and hCE-2, respectively. Solid and dotted line show the presence of ethanol and butanol, respectively. tains the glycosylation site at Asn79 but not at residue 389. Asn79 is modiˆed by a carbohydrate chain and appears to be involved in the stabilization of the hCE-1 show the completely diŠerent catalytic properties be- trimer by packing into an adjacent monomer.34,42) tween hCE-1 and hCE-2. For hCE-2, the Vmax values are hCE-1 is also capable of transesterifying cocaine in 47) similar among a series of compounds with Km values the presence of ethanol to cocaethylene. During the decreasing with increasing alcohol chain length. On the two-step hydrolysis of cocaine, hCE-1 forms a covalent other hand, hCE-1 shows a nearly identical Km value acyl-enzyme intermediate at the carboxylic methyl ester and increasing Vmax value with hydrophobicity. The position of cocaine, which is then attacked by ethanol to hydrolysis activity as a function of a‹nity with hCE-2 is create cocaethylene. Bencharit et al.,34) proposed the a normal property of the enzyme reaction. However, the following mechanism based on their X-ray crystalline catalytic property of hCE-1 is unusual. Catalysis of CES analysis. Ethanol enters the active site of hCE-1 through in the hydrolysis of p-amino-benzoate derivatives the side-door secondary pore adjacent to the large, proceeds by the following two steps (see Fig. 2). The ‰exible substrate binding pocket. The entrance to the ˆrst step (Step 1) is the formation of a covalent side-door secondary pore on the surface of hCE-1 is p-amino-benzoate-CES intermediate and the release of lined by structurally ‰exible residues including both b- alcohol. In the second step of hydrolysis (Step 2), the strands (b14 and b15) and random coils. This ‰exibility p-amino-benzoate-CES intermediate is attacked by may facilitate the passage of small molecules through histidine-activated water to generate carboxylic acid this side door. The alcohol produced in Step 1 may also product. When alcohol released at Step1 is stably be released from this side-door. The presence of the present in the substrate binding pocket, an alcohol can side-door secondary pore adjacent to the large, ‰exible attack an acyl-CES intermediate instead of water, and substrate binding pocket in hCE-1 might prolong the then an original ester compound may be reproduced. presence of the alcohol molecule after cleavage, In order to clarify the transesteriˆcation ability of allowing the alcohol to attack the benzoate-hCE-1 inter- CES, the ratio of transesteriˆcation and hydrolysis has mediate to reproduce the original substrate. Further- been measured in the presence of ethanol and butanol in more, the transition time of alcohol in hCE-1 might be the reaction mixture for p-amino benzoic acid propyl prolonged with increasing hydrophobicity due to inter- ester (Fig. 6). In the presence of ethanol, transesteriˆca- action of alcohol with the ‰exible site of hCE-1. That tion is lower than hydrolysis in both hCE1 and hCE2. may be another reason for the apparently slow hydroly- Interestingly, when the more hydrophobic alcohol, sis rate of substrates with large alcohol moieties by 180 Teruko IMAI

Fig. 7. Kinetic parameters for absorption of isovaleryl-propranolol (isovaleryl-PL) and propranolol (PL) in rat jejunal single-pass perfusion.

CLapp＝AUCb WAUCl×Qb＝absorbed amountWAUCl.WhereAUCb and AUCl are area under the curve of administered compound in the mesenteric vein and in the intestinal lumen at the steady-state, respectively. AUC in the intestinal lumen was obtained by assuming that the concentration of isovaleryl-PL or PL in the intestinal loop decreased according to ˆrst-order kinetics. Qb is a ‰ow rate of vascular perfusion. PeŠ＝Ql×(1-

CoutWCin)W2pRL. Where Ql is a ‰ow rate of intestinal perfusion. Cin and Cout are the concentration of applied compound at entrance and exit of the jejunal segment, respectively. The segment radius was assumed to be 0.178 cm by Yamashita et al.,61) and L is the length of the segment (10 cm).

CLdeg＝AUCM,l WAUCP,l×Ql＋AUCM,b WAUCP,l×Qb＝degraded amountWAUCP,l.WhereAUCP,l and AUCM,l are areas under the curve of isovaleryl-propranolol and propranolol in the intestinal lumen at the steady state, respectively. AUCM,b is the area under the curve of propranolol in the mesenteric vein. Values represent mean. (n＝3). hCE-1. predominantly pumped out by an active eŒux transporter.59) Intestinal Hydrolysis on the Process of Absorption As another example, the result of ˆrst-pass hydrolysis Prodrug approach is useful for improvement of a of O-isovaleryl-propranolol using in situ rat jejunal sin- membrane permeability by increasing the lipophilicity gle-pass perfusion is shown in Fig. 7.60) Isovaleryl- of their parent compound.49) When a prodrug, for propranolol is absorbed in the epithelial cells by passive example temocapril and methylphenidate, is scarcely diŠusion according to its 8-fold greater membrane hydrolyzed in human small intestine (Fig. 4), a prodrug permeability (PeŠ) than propranolol, and then it is approach is successful. However, some prodrugs are completely hydrolyzed to propranolol and isovaleric hydrolyzed during the intestinal absorption, and the acid produced at a rate limited by the uptake rate of extensive intestinal hydrolysis is responsible for ˆrst- isovaleryl-PL. Propranolol and isovaleric acid produced pass metabolism. Therefore, quantitative evaluation of in the epithelial cells are present at higher concentra- the intestinal hydrolysis during absorption is important. tions than in the luminal and vascular side. Therefore, In general, rat intestine is used for quantitative analysis these compounds are transported by simple diŠusion of ˆrst-pass metabolism in a single-pass perfusion into both vascular (pH 7.4) and luminal sides (pH 6.5) technique which correlates with the fractional dose based on pH-partitioning. Propranolol appears in the absorbed in humans50,51) and maintains the intestinal lumen at a rate 6-fold greater than in blood vessels. In architecture with respect to metabolism, absorption and contrast, isovaleric acid is transported into blood vessels secretion.52–54) Mammalian intestine contains several rather than the jejunal lumen. CES2 isozymes, although the intestinal hydrolysis Furthermore, when intestinal CES in in situ rat shows signiˆcant species diŠerences.55) Rat intestine jejunal single-pass perfusion was inhibited by the mainly contains two types of CES2 isozymes, speciˆc inhibitor, bis-p-nitrophosphate, hydrolysis of AB010632W35 and AY034877.56) There are some reports isovaleryl-propranolol during absorption was inhibited analyzing the intestinal hydrolysis and its contribution by about 80z (unpublished data). This indicates that of ˆrst- pass metabolism by in situ experiments.57–59) CES, especially CES2 isozymes, mainly contributes to Okudaira et al. has reported that an ester-type prodrug, rat intestinal hydrolysis of isovaleryl-propranolol. The ME3229, is taken up into enterocytes at a rate compati- residual 20z of hydrolysis during absorption in the ble with its lipophlicity, is completely hydrolyzed in the intestinal mucosa might be dependent on hydrolysis by intestinal tissue, and then the metabolites produced are several proteases on the cell membrane. Hydrolysis of Prodrug by Carboxylesterase 181

A variety of ester-containing compounds such as prodrugs are taken up into mucosal tissue due to increasing hydrophobicity, then hydrolyzed to their respective acid and alcohol compounds in the epithelial cells. The metabolites are present at higher concentration in epithelial cells than lumen and mesenteric vein, and can be transported by active or passive transport into blood vessels and the intestinal lumen. The ester derivative of an acidic drug with a small alcohol moiety is a poor substrate for human intestinal hCE-2, and such a prodrug can be absorbed as the intact prodrug. Therefore, it is necessary to consider the hydrolysis susceptibility of prodrugs and the physicochemical and biological properties of a prodrug and its parent drug Fig. 8. Polyacrylamide gel electrophoresis (A) and RT-PCR analysis when designing a prodrug. (B) of human tissue and Caco-2 cells. Polyacrylamide gel electrophoresis of the human liver and small Inaccuracy in Using Caco-2 Cells to Predict Intestinal intestine (5 mg protein) microsomes and Caco-2 cell S9 (15 mgof Absorption of Prodrugs protein) was stained for esterase activity using 1-naphthylacetate. *shows monomeric form of hCE-1. The product of RT-PCR for Caco-2 cells, derived from a colon adenocarcinoma, hCE-1 and hCE-2 mRNA was electrophoresed in 1.5z agarose gel spontaneously diŠerentiate under deˆned culture condi- and stained with ethidium bromide. tions to exhibit the structural and functional characteristics of mature human enterocytes. Therefore, Caco-2 cells have been widely accepted as a most useful in vitro Table 2. Hydrolysis of temocapril in Caco-2 cell brush border model for rapid screening of intestinal drug absorp- membrane vesicles (BBMVs) and S9 62,63) tion. Caco-2 cells express several drug-metabolizing K a) V b) vc) enzymes and transporters that are present in the human m max enterocyte. However, the Caco-2 cell has thus far fallen Caco-2 cells cultured in ‰asks for 7 days BBMVs N.D. N.D. 0.058±0.032 short as an ideal model for predicting oral availability of S9 441±8.61 15.9±1.62 5.35±0.375 ˆrst-pass metabolized drugs in the intestine because of Caco-2 cells cultured on Transwell for 21 days its failure to express substantial amounts of metabolic S9 547±11.8 14.06±0.988 4.98±0.366 enzymes such as CYP isozymes.64,65) Liver microsomes The esterase activity in a Caco-2 cell line is markedly 576±33.9 364±15.4 119±13.4 65,66) Small intestine microsomes lower than the human small intestine. However, N.D. N.D. 0.162±0.003 some prodrugs are extensively hydrolyzed to parent drug during passage through Caco-2 cells.67,68) Especial- a): mM, b): nmolWminWmg protein, ly, prodrugs with large acyl group and small alcohol c): nmolWminWmg protein. The hydrolysis rate (v) was measured at 250 mMoftemocapril. group are signiˆcantly hydrolyzed, although such N.D. means ``not determined''. prodrug is scarcely hydrolyzed by hCE2, an abundant CES isozyme in human intestine. Li et al. has reported that 70z of carindacillin is hydrolyzed to parent drug level. That is, the expression pattern of CES in Caco-2 by intracellular enzyme during passage across Caco-2 cells is completely diŠerent from that in human small cell monolayer.67) Furthermore, Tantishaiyakul et al. intestine but quite similar to that in human liver. The has reported that mefenamic acid guiacol ester is results of native PAGE analysis of hydrolase activity for completely hydrolyzed during transport across the 1-naphthylbutyrate and RT-PCR in Caco-2 cells are Caco-2 cell monolayer, and the cellular hydrolysis is shown in Fig. 8. The expression pattern of hCE1 and completely inhibited by phenylmethyl sulfonyl‰uoride hCE2 in Caco-2 cells are same in passage number from (PMSF), an inhibitor for esterase including CES.68) 28 to 59 and the cells cultured in diŠerent laboratories. Since the parent drugs of both carindacillin and Table 2 lists the results of hydrolysis of temocapril, mefenamic acid guiacol ester are compounds with which is a good and poor substrate for hCE-1 and carboxylic acid, these prodrugs might be good sub- hCE-2, respectively, in Caco-2 cells compared with that strates for hCE1. in human small intestine and liver. The Km value of Our recent study has shown that the Caco-2 cell line temocapril is almost the same in human liver micro- expresses both hCE-1 and hCE-2 at a signiˆcantly lower somes and Caco-2 cell S9 (about 500 mM), although the 69) level than human tissue. Interestingly, Caco-2 cell Km value in small intestine microsomes could not be expresses hCE-1 at a high level and hCE-2 at a lower measured due to the extremely low hydrolase activity. 182 Teruko IMAI

The Vmax value of temocapril in Caco-2 cell S9 is much tical Sciences, Chiba Institute of Sciences, Professor lower than that in human liver due to the low expression Kan Chiba, Graduate School of Pharmaceutical of esterase in Caco-2 cells. In addition, temocapril is Sciences, Chiba University, Assistant Professor Mitsuru hydrolyzed by brush border membrane in the range of Hashimoto, Graduate School of Pharmaceutical only 1–2z of its hydrolysis in S9. When temocapril is Sciences, Kumamoto University, and many colleagues applied in the appical side of Caco-2 cell monolayer, for their tremendous encouragement and helpful only temocaprilat is transported into the basolateral side discussions. due to hydrolysis of temocapril by hCE1 during transport across Caco-2 cells. Since temocapril is not hydro- References lyzed in human intestine, a prediction of human intesti- 1) Buchwald, P. and Bodor, N.: Physicochemical aspects nal absorption from transport across the Caco-2 cell of the enzymatic hydrolysis of carboxylic esters. Phar- monolayer may fail. mazie, 57: 87–93 (2002). In general, the human small intestine and colon 2) Bodor, N. and Buchwald, P.: Designing safer (soft) express abundant quantities of hCE-2.15,70) Schwer et drugs by avoiding the formation of toxic and oxidative al.71) have reported the higher expression of hCE-2 in metabolites. Methods Mol. Biol., 186:, 301–312 (2002). the jejunum, compared with moderate expression in the 3) Rooseboom, M., Commandeur, J. N. and Vermeulen, N. P.: Enzyme-catalyzed activation of anticancer duodenum and low expression in the ileum. There are prodrugs. Pharmacol. Rev., 56: 53–102 (2004). some reports that hCE-2 is an abundant protein in 4) Imai, T., Yoshigae, Y., Hosokawa, M., Chiba, K. and normal colon tissue, although the expression level Oragiri, M.: Evidence for the involvement of a pulmona- 72,73) diŠers between normal colon and colon tumors. ry ˆrst-pass eŠect via carboxylesterase in the disposition 74) Furthermore, Pavillard et al. demonstrated the of a propranolol ester derivative after intravenous expression of CES in colon cell lines by RT-PCR. The administration. J. Pharmacol. Exp. Ther., 307: expression of hCE-1 is higher in HT-29 cells than in 1234–1242 (2003). LoVo cells, while the expression of hCE-2 is very similar 5) Yoshigae, Y., Imai, T., Aso, T and Otagiri, M.: Species in both cell lines and always much higher than the diŠerences in the disposition of propranolol prodrugs expression of hCE-1. The expression pattern of CES in derived from hydrolase activity in intestinal mucosa. Caco-2 cells is very diŠerent from that in colon tissue Life Sci., 62: 1231–1241 (1998). 6) Ahmed, S., Imai, T. and Otagiri, M.: Evaluation of and other colon cell lines. stereoselective transdermal transport and concurrent When prodrugs are a good substrate for hCE1, these cutaneoushydrolysisofseveralesterprodrugsof compounds are easily hydrolyzed to parent drugs during propranolol: mechanism of stereoselective permeation. transport through Caco-2 cell monolayer, although Pharm. Res., 13: 1524–1529 (1996). these prodrugs are slowly hydrolyzed in human intes- 7)Takai,S.,Matsuda,A.,Usami,Y.,Adachi,T., tine. Therefore, prediction of human intestinal absorp- Sugiyama, T., Katagiri, Y., Tatematsu, M. and Hirano, tion of ester- and amide-containing drugs based on K.: Hydrolytic proˆle for ester- or amide-linkage by transport results obtained using Caco-2 cell monolayers carboxylesterases pI 5.3 and 4.5 from human liver. Biol. should be performed with care. Pharm. Bull., 20: 869–873 (1997). 8) Humerickhouse, R., Lohrbach, K., Li, L., Bosron, W. Conclusion F. and Dolan, M. E.: Characterization of CPT-11 CESs are widely distributed in the body of all hydrolysis by human liver carboxylesterase isoforms hCE-1 and hCE-2. Cancer. Res., 60: 1189–1192 (2000). mammalian species, and play an important role in the 9) Tabata, T., Katoh, M., Tokudome, S., Nakajima, M. metabolism of xenobiotics and endogenous com- and Yokoi, T.: Identiˆcation of the cytosolic carbox- pounds. Although CESs can interact with several ylesterase catalyzing the 5?-deoxy-5-‰uorocytidine for- compounds having diŠerent chemical structures, each mation from capecitabine in human liver. Drug Metab. human CES isozyme has a diŠerent substrate speciˆcity, Dispos., 32: 1103–1110 (2004). resulting in a tissue-speciˆc metabolic rate. Therefore, 10) Pindel, E. V., Kedishvili, N. Y., Abraham T. L., successful design of ester-containing drugs will be great- Brzezinski M. R., Zhang J., Dean R. A. and Bosron W. ly improved by further detailed analysis of the mechan- F.: Puriˆcation and cloning of a broad substrate speciˆc- ism of action, substrate recognition and expression of ity human liver carboxylesterase that catalyzes the human CES isozymes. Furthermore, development of an hydrolysis of cocaine and heroin. J. Biol. Chem., 272: 14769–14775 (1997). evaluating system for absorption and in vitro 11) ZhangJ.,BurnellJ.C.,DumaualN.andBosronW.F.: metabolism of prodrugs will assist with discovery of a Binding and hydrolysis of meperidine by human liver suitable prodrug. carboxylesterase hCE-1. J. Pharmacol. Exp. Ther., 290: 314–318 (1999). Acknowledgments: The author would like to thank 12) Satoh. T. and Hosokawa. M.: The mammalian carbox- Professor Masakiyo Hosokawa, Faculty of Pharmaceu- ylesterases: from molecules to functions. Annu. Rev. Hydrolysis of Prodrug by Carboxylesterase 183

Pharmacol. Toxicol., 38: 257–288 (1998). cholinesterase, paraoxonase, and albumin esterase, but 13) Satoh. T., Taylor. P., Bosron. W. F., Sanghani. P., not carboxylesterase, are present in human plasma. Hosokawa. M. and Du. B. N.: Current progress on Biochem. Pharmacol., 70: 1673–1684 (2005). esterases: from molecular structure to function. Drug 25) Mentlein, R., Heiland, S and Heymann E.: Simultane- Metab. Dispos., 30: 488–493 (2002). ous puriˆcation and comparative characterization of six 14) Munger.J.S.,Shi.G-P.,Mark,E.A.,Chin,M.D., serine hydrolases from rat liver microsomes. Arch. Gerard, C. and Chapman, H. A.: A serin esterase Biochem. Biophys., 200: 547–559 (1980). released by human alveolar macrophages is closely relat- 26) Mentlein, R., Rix-Matzen, H. and Heymann, E.: ed to liver microsomal carboxylesterases. J. Biol. Chem., Subcellular localization of non-speciˆc carbox- 266: 18832–18838 (1991). ylesterases, acylcarnitine hydrolase, monoacylglycerol li- 15) Xu, G., Zhang, W., Ma, M. K. and McLeod, H. L.: pase and palmitoyl-CoA hydrolase in rat liver. Biochim. Human carboxylesterase 2 is commonly expressed in Biophys. Acta, 964: 319–328 (1988). tumor tissue and is correlated with activation of irinote- 27) Block, W and Arndt, R.: Chromatographic study on the can. Clin. Cancer Res., 8: 2605–2611 (2002). speciˆcity of bis-p-nitrophenylphospate in vivo:Identiˆ- 16) Schwer, H., Langmann, T., Daig, R., Becker, A., cation of labeled proteins of rat liver after intravenous Aslanidis, C. and Schmitz, G.: Molecular cloning and injection of bis-p-nitro[14C]phenylphosphate as carbox- characterization of a novel putative carboxylesterase, ylesterases and amidases. Biochim. Biophys. Acta, 524: present in human intestine and liver. Biochem. Biophys. 85–93 (1978). Res. Commun., 233: 117–120 (1997). 28) Sterri, S. H., Johnsen, B. A. and Fonnum, F.: A 17) Junge, W., Heymann, E and Krisch, K.: Human liver radiochemical assay method for carboxylesterase, and carboxylesterase: puriˆcation and molecular properties. comparison of enzyme activity towards the substrates Arch. Biochem. Biophys., 165: 749–763 (1974). methyl[1-14C]butyrate and 4-nitrophenyl butyrate. 18)Inoue,M.,Morikawa,M.,Tsuboi,M.,Ito,Y.and Biophem. Pharmacol., 34: 2779–2785 (1985). Sugiura, M.: Comparative study of human intestinal and 29) Heymann, E., Mentlein, R., Schmalz, R., Schwabe, C. hepatic esterases as related to enzaymatic properties and and Wagenmann F.: A method for the estimation of hydrolyzing activity for ester-type drugs. Jpn. J. Phar- esterase synthesis and degradation and its application to macol., 30: 529–535 (1980). evaluate the in‰uence of insulin and glucagons. Eur. J. 19) Mori, M., Hosokawa, M., Ogasawara, Y., Tsukada, E. Biochem., 102: 509–519 (1979). and Chiba, K.: cDNA cloning, characterization and 30) Imai, T., Taketani, M., Shii, M., Hosokawa, M. and stable expression of novel human brain carboxylesterase. Chiba, K.: Substrate speciˆcity of carboxylesterase iso- FEBS Letters, 458: 17–22 (1999). zymes and their contribution to hydrolase activity in the 20) Sanghani, S. P., Quinney, S. K., Fredenburg, T. B., human liver and small intestine. Drug Metab. Dispos., in Davis, W. I., Murry, D. J. and Bosron, W. F.: Hydroly- press. sis of irinotecan and its oxidative metabolites, 7-ethyl-10- 31) Hosokawa, M., Endo, T., Fujisawa, M., Hara, S., [4-N-(5-aminopentanoic acid)-1-piperidino] carbonylox- Iwata, N., Sato, Y. and Satoh, T.: Interindividual ycamptothecin and 7-ethyl-10-[4-(1-piperidino)-1- variation in carboxylesterase levels in human liver micro- amino]-carbonyloxycamptothecin, by human carbox- somes. Drug Metab. Dispos., 23: 1022–1027 (1995). ylesterases CES1A1, CES2, and a newly expressed car- 32) Brzezinski, M. R., Abraham, T. L., Stone, C. L., Dean, boxylesterase isoenzyme, CES3. Drug Metab. Dispos., R. A. and Bosron, W. F.: Puriˆcation and characteriza- 32: 505–511 (2004). tion of a human liver cocaine carboxylesterase that cata- 21) Quinney, S. K., Sanghani, S. P., Davis, W. I., Hurley, lyzes the production of benzoylecgonine and the forma- T.D.,Sun,Z.,Murry,D.J.andBosron,W.F.: tion of cocaethylene from alcohol and cocaine. Biochem. Hydrolysis of capecitabine to 5?-deoxy-5-‰uorocytidine Pharmacol., 48: 1747–1755 (1994). by human carboxylesterases and inhibition by loper- 33) Pennings, J. M., Leccese, A. P. and WolŠ, F. A.: EŠect amide. J. Pharmacol. Exp. Ther., 313: 1011–1016 of concurrent use of alcohol and cocaine. Addiction, 97: (2005). 773–783 (2002). 22) Guemei, A. A., Cottrell, J., Band, R., Hehman, H., 34) Bencharit, S, Morton, C. L, Xue, Y, Potter, P. M and Prudhomme,M.,Pavlov,M.V.,Grem,J.L.,Ismail, Redinbo, M. R. Structural basis of heroin and cocaine A.S.,Bowen,D.,Taylor,R.S.andTakimoto,C.H.: metabolism by a promiscuous human drug-processing Human plasma carboxylesterase and butyryl- enzyme. Nat Struct Biol., 10: 349–356 (2003). cholinesterase enzyme activity: correlations with SN-38 35) Becker,A.,Bottcher,A.,Lackner,K.J.,Fehringer,P., pharmacokinetics during a prolonged infusion of irinote- Notka, F., Aslanidis, C. and Schmitz, G.: Puriˆcation, can. Cancer Chemother. Pharmacol., 47: 283–290 cloning, and expression of human enzyme with acyl (2001). coenzyme A: cholesterol acyl-transferase activity, which 23) Yan, B., Yang, D., Bullock, P. and Parkinson, A.: Rat is identical to liver carboxylesterase. Arterioscler serum carboxylesterase cloning, expression, regulation, Thromb., 14: 1346–1355 (1994). and evidence of secretion from liver. J. Biol. Chem., 36) Bora, P. S., Guruge, B. L., Miller, D. D., Chaitman, B. 270: 19128–19134 (1995). R. and Ruyle, M. S.: Puriˆcation and characterization 24) Li, B, Sedlacek, M, Manoharan, I, Boopathy, R, of human heart fatty acid ethyl ester synthaseWcarbox- Duysen, E. G., Masson, P., and Lockridge, O.: Butyryl- ylesterase. J. Mol. Cell. Cardiol., 28: 2027–2032 (1996). 184 Teruko IMAI

37) Diczfalusy, M. A., Bjorkhem, I., Einarsson, C., 50) Amidon, G. L., Sinko, P. J. and Fleisher, D. Estimating Hillebrant,C.G.,andAlexson,S.E.:Characterization human oral fraction dose absorbed: a correlation using of enzymes involved in formation of ethyl esters of long- rat intestinal membrane permeability for passive and chain fatty acids in humans. JLipidRes., 42: 1025–1032 carrier-mediated compounds. Pharm. Res., 5: 651–654 (2001). (1988). 38) Bora,P.S.,Farrar,M.A.,Miller,D.D.,Chaitman,B. 51) Fagerholm, U., Johansson, M. and Lennern äas, H.: R. and Guruge B. L.: Myocardial cell damage by fatty Comparison between permeability coe‹cients in rat and acid ethyl esters. J. Cardiovasc. Pharmacol., 27:1–6 human jejunum. Pharm. Res., 13: 1336–42 (1996). (1996). 52)Pang,K.S.,Yuen,V.,Fayz,S.,teKoppele,J.M., 39) Potter, P. M., Wolverton, J. S., Morton, C. L., Wierdl, Mulder, G. J.: Absorption and metabolism of M. and Danks M. K.: Cellular localization domains of a acetaminophen by the in situ perfused rat small intestine rabbit and a human carboxylesterase: in‰uence on preparation. Drug Metab. Dispos., 14: 102–111 (1986). irinotecan (CPT-11) metabolism by the rabbit enzyme. 53) Tamura, S., Tokunaga, Y., Ibuki, R., Amidon, G. L., Cancer Res., 58: 3627–3632 (1998). Sezaki, H. and Yamashita, S.: The site-speciˆc transport 40) Kroetz, D. L., McBride, O. W. and Gonzalez, F. J.: and metabolism of tacrolimus in rat small intestine. J. Glycosylation-dependent activity of baculovirus- Pharmacol. Exp. Ther., 306: 310–316 (2003). expressed human liver carboxylesterases: cDNA cloning 54) Zimmerman,C.L.,Wen,Y.andRemmel,R.P.:First- and characterization of two highly similar enzyme pass disposition of (－)-6-aminocarbovir in rats: II. Inhi- forms. Biochemistry 32: 11606–11617 (1993) bition of intestinal ˆrst-pass metabolism. Drug Metab. 41) Bencharit, S, Morton, C. L., Hyatt, J. L., Kuhn, P., Dispos., 28: 672–679 (2000). Danks,M.K.,Potter,P.M.andRedinbo,M.R.:Crys- 55) Inoue, M., Morikawa, M., Tsuboi, M. and Sugiura M.: tal structure of human carboxylesterase 1 complexed Species diŠerence and characterization of intestinal with the alzheimer's drug tacrine: from binding promis- esterase on the hydrolyzing activity of ester-type drugs. cuity to selective inhibition. Chemistry Biol., 10: Jpn. J. Pharmacol., 29: 9–16 (1979). 341–349 (2003) 56)Sanghari,S.P.,Davis,W.I.,Dumaual,N.G., 42) Fleming, C. D., Bencharit, S., Edwards, C. C., Hyatt, J. Mahrenholz, A. and Bosron, W. F.: Identiˆcation of L.,Tsurkan,L.,Bai,F.,Fraga,C.,Morton,C.L., microsomal rat liver carboxylesterases and their activity Escher, L., Williams. H., Potter, P. M. and Redinbo, with retinyl palmitate. Eur. J. Biochem., 269: 4387–4398 M. R.: Structural insights into drug processing by human (2002). carboxylesterase 1: Tamoxifen, mevastatin, and inhibi- 57) Prueksaritanont, T., Deluna, P., Gornam, L. M., Ma, tion by benzil. J. Mol. Biol., 352: 165–177 (2005) B., Cohn, D., Pang, J., Xu, X., Leung, K. and Lin, J. 43) Bencharit, S., Morton, C. L., Howord-Williams, E. L., H.: In vitro and in vivo evaluation of intestinal barriers Danks, M. K., Potter, P. M. and Redinbo, M. R.: Struc- for the zwitterion L-767,679 and its carboxyl ester tural insights into CPT-11 activation by mammalian prodrug L-775,318: Role of eŒux and metabolism. Drug carboxylesterases. Nat. Struct. Biol., 9: 337–342 (2002). Metab. Dispos., 26: 520–527 (1998). 44) Bosron, W. F. and Hurley, T. D.: Lessons from a bac- 58) Ruiz-Balaguer, N., Nacher, A., Gasabo, V. G.. and terial cocaine esterase. Nat. Struct. Biol., 9: 4–5 (2002). Merino Sanjuan, M.: Intestinal transport of cefuroxime 45) Sun, Z., Murry, D. J., Sanghani, S. P., Davis, W. I., axetil in rats: absorption and hyrolysis processes. Int. J. Kedishvili, N. Y., Zou, Q., Hurley, T. D. and Bosron, Pharm., 234: 101–111 (2002). W. F.: Methylphenidate is stereroselectively hydrolyzed 59) Okudaira, N., Tatebayashi, T., Speirs, G. C., Komiya, I by human carboxylesterase CES1A1. J. Pharmacol. and Sugiyama, Y.: A study of the intestinal absorption Exp. Ther., 310: 469–476 (2004). of an ester-type prodrug, ME3229, in rats: active eŒux 46) Senter, P. D., Beam, K. S., Mixan, B. and Wahl, A. F.: transport as a cause of poor bioavailability of the active Identiˆcation and activities of human carboxylesterases drug. J. Phatmacol. Exp. Ther., 294: 580–587 (2000). for the activation of CPT-11, a clinically approved an- 60) Masaki, K., Taketani, M. and Imai, T.: First-pass ticancer drug. Bioconjugate Chem., 12: 1074–1080 hydrolysis of a propranolol ester derivative in rat small (2001). intestine. Drug Metab. Dispos., 34: 398–404 (2006). 47) Brzezinski, M. R., Spink, B. J., Dean, R. A., Berkman, 61) Yamashita, S., Tanaka, Y., Endoh, Y., Taki, Y., C.E.,Cashman,J.R.andBosron,W.F.Humanliver Sakane, T., Nadai, T. and Sezaki, H.: Analysis of drug carboxylesterase hCE-1: Binding speciˆcity for cocaine, permeation across Caco-2 monolayer: implication for heroin, and their metabolites and analogs. Drug Metab. predicting in vivo drug absorption. Pharm. Res., 14: Dispos., 25: 1089–1096 (1997). 486–491 (1997). 48) Huang,H.,Fleming,C.D.,Nishi,K.,Redinbo,M.R. 62) Artursson, P., Palm, K. and Luthman, K.: Caco-2 and Hammock, B. D.: Stereoselective hydrolysis of monolayers in experimental and theoretical predictions pyrethroid-like ‰uorescent substrates by human and of drug transport. Adv. Drug Deliv. Rev., 46: 27–43 other mammalian liver carboxylesterases. Chem. Res. (2001). Toxicol., 18: 1371–1377 (2005). 63) Lennern äas, H.: Human intestinal permeability. J. 49) Mizen, L. and Burton, G..: The use of esters as prodrugs Pharm. Sci., 87: 403–410 (1998). for oral delivery of b-lactam antibiotics. Pharm. 64) Nakamura, T., Sakaeda, T., Ohmoto, N., Tamura, T., Biotechnol., 11: 345–365 (1998). Aoyama, N., Shirakawa. T., Kamigaki, T., Nakamura, Hydrolysis of Prodrug by Carboxylesterase 185

T.,Kim,K.I.,Kim,S.R.,Kuroda,Y.,Matsuo,M., eŠects of their hydrolyzing activity in predicting human Kasuga, M. and Okumura, K.: Real-time quantitative intestinal absorption. Drug Metab. Dispos., 33: polymerase chain reaction for MDR1, MRP1, MRP2, 1185–1190 (2005). and CYP3A-mRNA levels in Caco-2 cell lines, human 70) Khanna, R., Morton, C. L., Danks, M. K. and Potter, duodenal enterocytes, normal colorectal tissues, and P. M.: Proˆcient metabolism of irinotecan by a human colorectal adenocarcinomas. Drug Metab. Dispos., 30: intestinal carboxylesterase. Cancer Res., 60: 4725–4728 4–6 (2002). (2000). 65) Prueksaritanont, T., Gorham, L. M., Hochman, J. H., 71) Schwer, H., Langmann, T., Daig, R., Becker, A., Tran,L.O.andVyas,K.P.:Comparativestudiesof Aslanidis, C. and Schmitz, G.: Molecular cloning and drug-metabolizing enzymes in dog, monkey, and human characterization of a novel putative carboxylesterase, small intestines, and in Caco-2 cells. Drug Metab. present in human intestine and liver. Biochem. Biophys. Dispos., 24: 634–642 (1996). Res. Commun., 233: 117–120 (1997). 66) Yoshigae, Y., Imai, T., Horita, A., Matsukane, H. and 72) Sanghani, S. P., Quinney, S. K., Fredenburg, T. B., Otagiri, M.: Species diŠerences in stereoselective hydro- Sun,Z.,Davis,W.I.,Murry,D.J.,Cummings,O.W., lase activity in intestinal mucosa. Pharm. Res., 15: Seitz, D. E. and Bosron, W. F. Carboxylesterases 626–631 (1998). expressed in human colon tumor tissue and their role in 67) Li, Y-H., Ito, K., Tsuda, Y., Kohda R., Yamada, H., CPT-11 hydrolysis. Clin. Cancer Res., 9: 4983–4991 and Ito, T.: Mechanism of intestinal absorption of an (2003). orally active b-lactam prodrug: uptake and transport of 73)Xie,M.,Yang,D.,Liu,L.,Xue,B.andYan,B.: carindacillin in Caco-2 cells. J. Pharmacol. Exp. Ther. Human and rodent carboxylesterases: immunorelated- 290: 958–964 (1999). ness, overlapping substrate speciˆcity, diŠerential sen- 68) Tantishaiyakul, V., Wiwattanawongsa, K., Pinsuwan, sitivity to serine enzyme inhibitors, and tumor-related S., Kasiwong, S., Phadoongsombut, N., Kaewnopparat, expression. Drug Metab. Dispos., 30: 541–547 (2002). S., Kaewnopparat, N. and Rojanasakul, Y.: Characteri- 74) Pavillard, V., Agostini, C., Richard, S., Charasson, V., zation of mefenamic acid-guauacol ester: stability and Montaudon, D. and Robert, D. J.: Determinants of the transport across Caco-2cell monolayers. Pharm. Res., cytotoxicity of irinotecan in two human colorectal tumor 19: 1013–1018 (2002). cell lines. Cancer Chemother. Pharmacol., 49: 329–335 69) Imai, T., Imoto, M., Sakamoto, H. and Hashimoto, M.: (2002). Identiˆcation of esterases expressed in Caco-2 cells and