- Email: [email protected]
Intestinal transport and metabolism of glucose-conjugated kyotorphin and cyclic kyotorphin: metabolic degradation is crucial to intestinal absorption of peptide drugs
Biochimica et Biophysica Acta 1475 (2000) 90^98
www.elsevier.com/locate/bba
Intestinal transport and metabolism of glucose-conjugated kyotorphin and cyclic kyotorphin: metabolic degradation is crucial to intestinal absorption of peptide drugs Takashi Mizuma *, Akihiro Koyanagi, Shoji Awazu Department of Biopharmaceutics and Drug Rational Research Center, School of Pharmacy, Tokyo Yakka University (Tokyo University of Pharmacy and Life Science, TUPLS), 1432-1 Horinouchi, Hachioji, Tokyo 192-03, Japan Received 22 December 1999 ; received in revised form 21 March 2000; accepted 22 March 2000
Abstract Intestinal transport and metabolism of modified kyotorphin (KTP) were studied in rats. Modified KTPs studied were C-terminally modified KTP with p-aminophenyl-L-D-glucoside (KTP-pAPLglc), N-terminally modified KTP-pAPLglc with t-butyloxycarbonyl group (Boc-KTP-pAPLglc) and the N- and C-terminally modified KTP by cyclization (cyclic KTP). KTP-pAPLglc was metabolized at a similar rate to that of KTP, and did not appear on the serosal side. Although Boc-KTP-pAPLglc was also metabolized, it was more stable than KTP and appeared on the serosal side. Cyclic KTP was also quite stable and appeared on the serosal side. The modified KTPs were evaluated kinetically for absorption consisting of membrane transport and metabolism. Absorption clearance (CLabs ) of cyclic KTP, BocKTP-pAPLglc and Boc-KTP was higher than that of KTP (0.247 Wl/min/cm) (Mizuma et al., Biochim. Biophys. Acta 1335 (1997) 111^119), which is the theoretical maximum by complete inhibition of peptidase activity, indicating that derivatization of KTP increases the membrane permeability. Furthermore, the data clearly showed that the greater the metabolic clearance (CLmet ) of KTP and the KTP derivatives, the lower the absorption clearance (CLabs ). These results and further simulation study led to the conclusion that metabolic degradation in the intestinal tissues is more critical than membrane permeability (transport) for oral delivery of peptide drugs. Based on the stability of cyclic KTP in serum, this appears to be a good candidate analgesic peptide drug. ß 2000 Elsevier Science B.V. All rights reserved. Keywords : Kyotorphin; Intestinal absorption; Peptide absorption ; Metabolism; Transport; Glucose conjugate; Cyclic peptide; Kinetic analysis
1. Introduction Dietary peptides and proteins are digested/metabolized in the intestines to be transported as amino acids or oligopeptides [1,2]. However, peptide and protein drugs must be transported without metabolic degradation to systemic circulation to exert their pharmacological actions. Kyotorphin (L-tyrosyl-L-arginine, KTP) is an endogenous compound that releases methionine enkephalin, which possesses analgesic activity [3], from the striatum [4]. Although the physiological activity, distribution in the
* Corresponding author. Fax: +81-426-763142; E-mail : [email protected]
brain [5] and metabolism in the brain [6^9] and serum [6] of KTP as an endogenous compound have been studied to clarify its physiological role, very few researchers have examined the intestinal absorption of oral administration of KTP as a medicinal exogenous compound [10]. Intestinal absorption consists of integrated processes of membrane permeation (transport) and enzymatic reactions (metabolism). Therefore, in order to develop orally active peptide drugs, the biophysical and biochemical processes in absorption must be examined kinetically. The present study examines intestinal transport and metabolism of chemically modi¢ed KTP as a model peptide drug (Fig. 1). We previously found that KTP is too unstable in the intestines to be absorbed. However, N-terminal modi¢cation of KTP with a t-butoxycarbonyl (Boc) group provides su¤cient stabilization against peptidases in intestinal tissue [10]. We have also found that glucose or galactose conjugation provide a novel route by way of Na /glucose cotransporter (SGLT1) for intestinal absorption [11^13].
0304-4165 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 4 1 6 5 ( 0 0 ) 0 0 0 5 1 - 9
BBAGEN 25014 26-4-00
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
91
Fig. 1. Structures of kyotorphin conjugated with p-aminophenyl L-glucoside (KTP-pAPLglc), N-t-Butoxycarbonyl-kyotorphin conjugated with p-aminophenyl L-glucoside (Boc-KTP-pAPLglc) and cyclic kyotorphin (cyclic KTP).
Therefore, KTP was modi¢ed with p-aminophenyl L-glucoside (pAPLglc) at the C-terminus to form L-glucose-conjugated KTP. t-Butoxycarbonyl KTP (Boc-KTP) was also conjugated with pAPLglc. We have also examined a cyclic form of KTP, since cyclic dipeptides are stable enough for absorption by the intestines [14,15]. Based on kinetic analysis of these data and simulation study, we identi¢ed the critical factors involved in the absorption of orally administered KTP as well as other pharmacologically active peptides that must be absorbed intact.
2.2. Animals
2. Materials and methods
First, Boc-Arg(NO2 )-pAPLglc was synthesized from Boc-Arg(NO2 ) and pAPLglc. Boc-Arg(NO2 ) (1700 mg) and pAPLglc (700 mg) were dissolved in 8 ml of N,Ndimethylformamide (DMF) and 720 Wl of triethanolamine by stirring at room temperature for 30 min. DCC (1100 mg) solution dissolved in 2 ml of ethyl acetate was dropped into the resultant mixture, and was stirred for 2 h. After further stirring the reaction mixture at 4³C for 22 h under light protection, DCC (280 mg) solution dissolved in 0.5 ml of ethyl acetate was dropped into the reaction mixture, and was stirred at 4³C for 24 h under light protection. The reaction mixture was ¢ltrated through a paper ¢lter and evaporated at 40³C under reduced pressure. The residue containing Boc-Arg(NO2 )-pAPLglc was dissolved in 10 ml of ethyl acetate, and washed with 10 ml of puri¢ed water three times. Ethyl acetate containing Boc-
2.1. Chemicals p-Aminophenyl L-D-glucopyranoside (pAPLglc) was purchased from Sigma (St. Louis, USA). NK -t-butyloxycarbonyl-Ng -nitro-L-arginine (Boc-Arg(NO2 )) and N,NPdicyclohexylcarbodiimide (DCC) were obtained from Peptide Institute, (Osaka, Japan). N-t-Butyloxycarbonyl-L-tyrosine (Boc-Tyr) and palladium-activated carbon (10%, Pd-C) were purchased from Wako (Osaka, Japan). Cyclic KTP (Cyclo(L-arginyl-L-tyrosine)) was purchased from Iwaki Glass (Tokyo, Japan). Sephadex G-10 was obtained from Pharmacia LKB Biotechnology (Tokyo, Japan). KTP-pAPLglc and Boc-KTP-pAPLglc were synthesized by the method described below.
Adults male Wistar rats (8 weeks old) were obtained from Japan SLC (Shizuoka, Japan) and allowed free access to water and commercially available chows in a temperature- and light-controlled room for more than 1 week before experimentation. 2.3. Synthesis of N-t-butyloxycarbonyl-L-kyotorphin conjugated with p-aminophenyl L-glucoside (Boc-KTP-pAPLglc)
BBAGEN 25014 26-4-00
92
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
Arg(NO2 )-pAPLglc was evaporated, and crude BocArg(NO2 )-pAPLglc was applied to a Sephadex G-10 column (2.5 cm i.d.U90 cm length). Fractions (10 ml per tube) eluted with puri¢ed water at a £ow rate of 2.0 ml/ min were analyzed by HPLC (TSKgel ODS-80TM, 6 mm i.d. U150 mm length, 25% acetonitrile, 0.05% phosphoric acid, 1.5 ml/min, 274 nm). Fractions containing BocArg(NO2 )-pAPLglc were collected and concentrated by evaporation at 40³C under reduced pressure. Concentrated Boc-Arg(NO2 )-pAPLglc solution was applied to a preparative ODS column (3 cm i.d.U30 cm length, GL Science, Tokyo) and eluted with a mobile phase (18% acetonitrile, 1.5% acetic acid) at a £ow rate of 2.5 ml/min. Each fraction (10 ml per tube) containing Boc-Arg(NO2 )-pAPLglc was analyzed by HPLC as above and collected. Collected fractions were evaporated under the above conditions. The residue was dissolved in 5 ml of acetonitrile and 5 ml of TFA, and ¢lled with nitrogen gas. After stirring the solution for 2 h at room temperature and evaporating, crude Arg(NO2 )-pAPLglc was obtained. Crude Arg(NO2 )pAPLglc was applied to a preparative ODS column (3 cm i.d.U30 cm length) and eluted with a mobile phase (3% methanol, 1.5% acetic acid) at a £ow rate of 2.5 ml/min. Each fraction (10 ml per tube) containing Arg(NO2 )pAPLglc eluted was analyzed by HPLC (TSKgel ODS80TM, 6 mm i.d.U150 mm length, 8% methanol, 0.05% phosphoric acid, 1.5 ml/min, 274 nm) and collected. Fractions were evaporated at 40³C under reduced pressure. Boc-Tyr (140 mg) and Arg(NO2 )-pAPLglc (120 mg) were dissolved in 4 ml of DMF and 71 Wl of triethanolamine by stirring at room temperature for 30 min. DCC dissolved in ethyl acetate was dropped into the mixture in 30-ml Erlenmeyer £asks on ice. Then, the reaction mixture was stirred for 2 h at 0³C, followed by 22 h at 4³C under light protection. After Boc-Tyr (70 mg) was added to the reaction mixture, DCC (52 mg) solution in 0.5 ml of ethyl acetate was dropped into the reaction mixture at 0³C to be stirred for 24 h under light protection. The reaction mixture was ¢ltrated through a paper ¢lter and evaporated. The resultant residue was dissolved in 10 ml of ethyl acetate, and was washed with puri¢ed water three times. Ethyl acetate was then evaporated. The crude Boc-TyrArg(NO2 )-pAPLglc was applied to a Sephadex G-10 column (2.5 cm i.d.U90 cm length), and eluted with puri¢ed water at a £ow rate of 2.0 ml/min. Each fraction (15 ml per tube) eluted was analyzed by HPLC (TSKgel ODS80TM, 6 mm i.d.U15 cm length, 30% acetonitrile, 0.05% phosphoric acid, 1.5 ml/min £ow rate, 274 nm UV absorbance). Fractions containing Boc-Tyr-Arg(NO2 )-pAPLglc were collected and evaporated at 40³C under reduced pressure. Concentrated solution of Boc-Tyr-Arg(NO2 )pAPLglc was applied to a preparative ODS column (ODS, 30 mm i.d.U300 mm length) and eluted with a mobile phase (27% acetonitrile, 1.5% acetic acid) at a £ow rate of 2.5 ml/min. Each fraction (10 ml per tube)
eluted was analyzed by HPLC as above. Fractions containing Boc-Tyr-Arg(NO2 )-pAPLglc were collected and evaporated at 40³C under reduced pressure. Puri¢ed Boc-Tyr-Arg(NO2 )-pAPLglc was dissolved in 10 ml of methanol, and ¢lled with nitrogen gas. After Pd-C was added to the solution, it was ¢lled with hydrogen gas and allowed to stand for 24 h under light protection at room temperature to form Boc-KTP-pAPLglc. De-protection of the Boc group from Boc-KTP-pAPLglc was con¢rmed by HPLC (TSKgel ODS-80TM, 6 mm i.d.U150 mm length, mobile phase consisting of 25% acetonitrile, 0.05% phosphoric acid, 1.5 ml/min, 274 nm). The BocKTP-pAPLglc solution was ¢ltrated through a paper ¢lter and concentrated by evaporation at 40³C under reduced pressure. Boc-KTP-pAPLglc solution was applied to a preparative ODS column (ODS, 30 mm i.d.U300 mm length) and eluted with a mobile phase (20% acetonitrile, 1.5% acetic acid) at a £ow rate of 2.5 ml/min. Each fraction (10 ml per tube) eluted was analyzed by HPLC as above. Fractions containing Boc-KTP-pAPLglc were collected and evaporated at 40³C under reduced pressure. 2.4. Synthesis of L-kyotorphin conjugated with p-aminophenyl L-glucoside (KTP-pAPLglc) Puri¢ed Boc-KTP-pAPLglc, which was prepared by the method described above, was dissolved in 10 ml of methanol, and ¢lled with nitrogen gas. After Pd-C was added to the solution, it was ¢lled with hydrogen gas and allowed to stand for 24 h under light protection at room temperature to form KTP-pAPLglc. KTP-pAPLglc was applied to semi-preparative HPLC (TSKgel ODS-80TM, 7.8 mm i.d.U300 mm length) and eluted with a mobile phase (1.5% methanol, 1.5% acetic acid) at a £ow rate of 3.0 ml/min. Each fraction (10 ml per tube) eluted was analyzed by HPLC as above. Fractions containing KTPpAPLglc were collected and evaporated at 40³C under reduced pressure. Residues of KTP-pAPLglc were dissolved in puri¢ed water. Finally, the powder of KTPpAPLglc was obtained by freeze-drying. 2.5. Identi¢cation of Boc-Tyr-Arg-pAPLglc and Tyr-Arg-pAPLglc ESI mass spectra of Boc-KTP-pAPLglc (Boc-Tyr-ArgpAPLglc) and KTP-pAPLglc (Tyr-Arg-pAPLglc) showed molecular ion peaks [M+H] at 691.7 and 591.7, respectively. The ninhydrin reaction of KTP-pAPLglc and TFAtreated Boc-KTP-pAPLglc which were developed on TLC, were positive, indicating the free amino group of KTPpAPLglc and protected amino group of Boc-KTPpAPLglc. The phenol-sulfuric acid reaction of Boc-KTPpAPLglc and KTP-pAPLglc which was developed on TLC, were positive, indicating the existence of a monosaccharide moiety.
BBAGEN 25014 26-4-00
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
93
2.6. Intestinal absorption
2.8. Stability of cyclic KTP in serum
Intestinal absorption of modi¢ed KTPs was examined in excised rat small intestines [11]. Brie£y, male Wistar rats (180^230 g, Japan SLC, Japan) fasted overnight were anesthetized with ether, and the intestinal blood was removed by saline perfusion. The jejunum was removed and everted. Ten cm of everted small intestine was placed in 30 ml of incubation medium (113.3 mM NaCl, 4.83 mM KCl, 1.214 mM KH2 PO4 , 1.205 mM MgSO4 , 16.96 mM NaHCO3 , 10.18 mM Na2 HPO4 , 0.645 mM CaCl2 , pH 7.4) containing 250 WM peptide in a beaker at 37³C. The serosal side was ¢lled with 5 ml of the incubation medium without peptides. When necessary, peptidase inhibitors were added to the medium on the mucosal and serosal sides. Incubation medium (100 Wl) was sampled from both the serosal and mucosal sides every 10 min until 60 min. The samples were mixed with 100 Wl of internal standard solution (200 WM tryptophan in 10% perchloric acid for Tyr-Arg-pAPLglc, or 100 WM sodium 1-naphthalensulfonate in acetonitrile for Boc-KTP-pAPLglc, or 1 mM p-hydroxybenzoic acid for cyclic KTP) for the following HPLC assay. The mixture was centrifuged at 11 000Ug for 5 min using a benchtop centrifuge KM15200 (Kubota, Japan). Twenty-¢ve microliters of the resultant supernatant was applied to HPLC.
The stability experiment was carried according to the method reported previously [16]. Rat blood was obtained through the cannulated carotid artery under ether anesthesia. After keeping on ice for 30 min, the blood was centrifuged at 1300Ug for 10 min. The resultant supernatant was supplied as serum. Next, 0.1 ml of cyclic KTP solution (20 WM ¢nal concentration) in incubation medium was added to 0.9 ml of serum in a microtube (1.5 ml volume), and incubated at 37³C. Samples were obtained periodically over 3 h, and cyclic KTP was determined by HPLC.
2.7. Metabolism of KTP-pAPLglc and Boc-KTP-pAPLglc in intestinal homogenates Examination of modi¢ed KTP metabolism in intestinal homogenates was performed by a previously reported method [10]. Brie£y, 10 cm of intestine was isolated, and homogenized in the same bu¡er as described in the absorption experiment by Physcotron homogenizer (Nichion Irika, Tokyo, Japan). A mixture of the peptide (250 WM ¢nal concentration) and homogenate (0.25% ¢nal concentration) was incubated for 60 min at 37³C in the same incubation medium used for the absorption experiment. At the indicated times, 100 Wl of reaction mixture was sampled. Modi¢ed KTP was determined by the same method as in the absorption experiment.
2.9. HPLC assay for absorption and metabolism experiments KTP-pAPLglc, Boc-KTP-pAPLglc, cyclic KTP and their metabolites were determined by reversed-phase HPLC. The HPLC system consisted of a pump (Twincle, Jusco, Tokyo, Japan), UV detector (Shimadzu, Kyoto, Japan), £uorescence detector (821-FP, Ex 278 nm, Em 305 nm, Jasco) and an integrator (D-2500, Hitachi, Tokyo, Japan). The £ow rate was set at 1.5 ml/min. Complete assay conditions are listed in Table 1. In the Boc-KTP assay, the mobile phase was composed of 20% acetonitrile and 0.05% phosphoric acid in water and an ODS column (A-312, 6 mm i.d.U15 cm length, YMC, Japan) was used. In the cyclic KTP assay, the mobile phase was composed of 30% methanol, 0.05% phosphoric acid, and a C8 column (TSKgel Octyl-80Ts, 4.6 mm i.d.U15 cm length, Tosoh, Japan) was used. A UV detector (274 nm) was used in the puri¢cation experiments and a £uorescence detector (Ex 278 nm, Em 305 nm) was used in the absorption experiments. The £ow rate was set at 1.5 ml/min. For the assay of cyclic KTP in stability experiment, 28.5% acetonitrile and 0.05% phosphoric acid and ODS 80TM (6 mm i.d.U15 cm length) were used. 2.10. Data analysis Elimination clearance, CLeli , from the mucosal side was
Table 1 HPLC assay conditions for absorption experiments of KTP-pAPLglc, Boc-KTP-pAPLglc and cyclic KTP Conditions
KTP-pAPLglc
Boc-KTP-pAPLglc
cyclic KTP
Column
TSKgel ODS-80TM 6 mm i.dU150 mm length 7% methanol, 0.05% phosphoric acid in water 200 WM tryptophan in 10% perchloric acid 821-FP (Jasco) Ex 278 nm, Em 305 nm
YMC-Pack ODS A-312 6 mm i.d.U150 mm length 20% acetonitrile, 0.05% phosphoric acid in water 100 WM sodium 1-naphthalenesulfonate in acetonitrile 821-FP (Jasco) Ex 278 nm, 305 nm
TSKgel ODS-80Ts 6 mm i.d.U150 mm length 30% methanol, 0.05% phosphoric acid in water 1 mM hydroxybenzoic acid in 10% perchloric acid 665A-21 UV detector (Hitachi) 194 nm
Mobile phase Internal standard Detector Wavelength
BBAGEN 25014 26-4-00
94
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
calculated by Eq. 1 [17]. CLeli
X eli AUCmuc;0ÿ60
1
where Xeli is amount eliminated from the mucosal side as calculated by : X eli C 60 3C 0 UV
2
AUCmuc;0ÿ60 is the area under the concentration curve of the peptide on the mucosal side from time 0 to 60 min calculated by the trapezoidal rule [18]. C60 and C0 represent the modi¢ed KTP concentration on the mucosal side at 60 and 0 min, respectively. V is the volume of medium on the mucosal side. Absorption clearance, CLabs , was calculated by Eq. 3 [17]. CLabs
X abs AUCmuc;0ÿ60
3
where Xabs is the amount of modi¢ed KTP absorbed from the mucosal side to the serosal side over 60 min. Metabolic clearance (per cm of intestine), CLmet , in intestinal homogenate was calculated by Eqs. 4 and 5. CLmet
X met UW w AUChom;0ÿ60
X met C 60 3C 0 U
100 H
4 5
where Xmet is the amount of metabolite formed per 1 g of wet weight of intestine in the intestinal homogenates as calculated by Eq. 5. AUChom;0ÿ60 is the area under the concentration curve of modi¢ed KTP in the reaction mixture from time 0 to 60 min, as calculated by the trapezoidal rule. Ww is wet weight per length of everted intestine (g/cm). C60 and C0 represent the modi¢ed KTP concentration in the reaction mixture at 60 and 0 min, respectively. H is the concentration of homogenates (%, w/v) of the reaction mixture. The kinetic relationship between absorption clearance (CLabs ) and metabolic clearance (CLmet ) is expressed by
Scheme 1. Metabolic inhibition model for intestinal absorption of drugs. Ci , Cs and Cm are drug concentrations in intestinal tissue, on serosal side, and mucosal side, respectively. CLmÿi , CLiÿm and CLiÿs are the transport clearance from mucosal side to intestinal tissue, and from intestinal tissue to mucosal side, and from intestinal tissue to serosal side, respectively. CLmet;int is intrinsic clearance of metabolism by metabolic enzymes in intestinal tissue. Ra is remaining activity of metabolic enzyme in intestinal tissue in presence of peptidase inhibitors.
Eq. 6, which is based on the metabolic inhibition model for absorption (Scheme 1) [10,17]. CLabs
CLovt
CLovt CLmet 1 CLiÿm CLiÿs CLmÿi UCLiÿs CLiÿm CLiÿs
6
7
CLmet CLmet;int URa where CLovt is overall transport clearance of modi¢ed KTP from the mucosal side to the serosal side, where no metabolic degradation of modi¢ed KTP in intestinal tissue occurs during the absorption process. Under this assumption, CLovt is the maximum value of CLabs . CLmÿi , CLiÿm and CLiÿs represent the transport clearance of modi¢ed KTP from the mucosal side to the intestinal tissue, from the intestinal tissue to the mucosal side, and from the intestinal tissue to the serosal side, respectively. CLmet;int
Fig. 2. Time course of KTP-pAPLglc on mucosal (a) and serosal (b) sides in intestinal absorption. Data represent mean þ S.E. (n = 3).
BBAGEN 25014 26-4-00
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
Fig. 3. Time course of Boc-KTP-pAPLglc on mucosal (a) and serosal (b) sides in intestinal absorption. sent mean þ S.E. (n = 3).
is the intrinsic clearance of metabolism in the intestinal tissue. Ra is the remaining activity of the metabolism in the intestinal tissue in the presence of enzyme inhibitors. In the absence of any peptidase inhibitors (Ra = 1), the following equation is obtained. CLabs
CLovt CLmet;int 1 CLiÿm CLiÿs
8
The e¡ect of intestinal metabolism on absorption was simulated according to Eq. 9, which is derived from Eq. 8. CLabs CLovt
1 CLmet;int 1 CLiÿm CLiÿs
9
a,
95
Boc-KTP-pAPLglc ; b, Boc-KTP. Data repre-
3.2. Intestinal transport and metabolism of Boc-KTP-pAPLglc Fig. 3a and b show the time course of Boc-KTPpAPLglc on the mucosal side and serosal side, respectively. Boc-KTP-pAPLglc was eliminated and Boc-Tyr formed on the mucosal side. The elimination clearance of Boc-KTPpAPLglc was 728 þ 280 Wl/min/cm (Table 2). Serosal appearance of Boc-KTP-pAPLglc was observed in one of three experiments. 3.3. Intestinal transport and metabolism of cyclic KTP Fig. 4a and b show the time course of cyclic KTP on the mucosal side and serosal side, respectively. Cyclic KTP was stable on the mucosal side and the elimination clearance of cyclic KTP from the mucosal side was 2.21 þ 0.07 Wl/min/cm (Table 2). Cyclic KTP appeared on the serosal side, and the transport clearance of cyclic KTP was 1.12 þ 0.05Wl/min/cm (Table 2).
3. Results 3.1. Intestinal transport and metabolism of KTP-pAPLglc Fig. 2a and b show the time course of KTP-pAPLglc on the mucosal side and serosal side, respectively. KTPpAPLglc on the mucosal side was eliminated and the elimination clearance of KTP-pAPLglc from the mucosal side was 129 þ 11 Wl/min/cm (Table 2). On the other hand, KTP-pAPLglc was not detectable on the serosal side.
3.4. Metabolic degradation of KTP-pAPLglc and Boc-KTP-pAPLglc in intestinal homogenates Fig. 5 show the time course of concentration changes of KTP-pAPLglc, Boc-KTP-pAPLglc and cyclic KTP in the intestinal homogenates. KTP-pAPLglc was eliminated much faster than Boc-KTP-pAPLglc, whereas cyclic KTP
Table 2 Absorption clearance (CLabs ), elimination clearance (CLeli ) and metabolic clearance (CLmet ) of KTP-pAPLglc, Boc-KTP-pAPLglc and cyclic KTP KTPa KTP-pAPLglc Boc-KTPa Boc-KTP-pAPLglc Cyclic KTP
CLabs (Wl/min/cm)
CLeli (Wl/min/cm)
CLmet (Wl/min/cm)
ND ND 0.257 þ 0.11 0.726b 1.12 þ 0.05
129 þ 3.0 129 þ 11 357 þ 67 728 þ 280 2.21 þ 0.07
5120 þ 50 4630 þ 210 32.2 þ 7.6 74.9 þ 11.9 stable
Data represent mean þ S.E. (n = 3). a Data from previous report [10]. b Only one of three experiments showed the serosal appearance of Boc-KTP-pAPLglc, because of higher detection limit of concentration in HPLC assay.
BBAGEN 25014 26-4-00
96
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
Fig. 4. Time course of cyclic KTP on mucosal (a) and serosal (b) sides in intestinal absorption. Data represent mean þ S.E. (n = 3).
remained stable. The metabolic clearance of KTP-pAPLglc and Boc-KTP-pAPLglc were 4630 and 74.9 Wl/min/cm, respectively (Table 2). 3.5. Stability of cyclic KTP in serum Cyclic KTP was stable in serum for the full 60-min period (data not shown). 3.6. Relationship between absorption clearance and metabolic clearance The higher the CLmet of KTP and the KTP derivatives, the lower the CLabs (Fig. 6). The CLabs of cyclic KTP, Boc-KTP-pAPLglc and Boc-KTP, but not KTP-pAPLglc, were higher than CLovt of KTP. When the CLmet exceeded about 500 Wl/min/cm, KTP could not be absorbed.
Fig. 5. Stability of KTP-pAPLglc, Boc-KTP-pAPLglc and cyclic KTP in intestinal homogenates. a, KTP-pAPLglc; b, Boc-KTP-pAPLglc; E, cyclic KTP. Data represent mean þ S.E. KTP-pAPLglc (n = 3), Boc-KTPpAPLglc (n = 3), cyclic KTP (n = 5).
3.7. Simulation of e¡ect of intrinsic metabolic activity on the absorption Fig. 7, which was obtained by Eq. 9, shows the impact of the relative ratio of metabolic activity to membrane permeability (CLmet;int /(CLiÿm +CLiÿs )) on the absorption expressed as the ratio of CLabs to the maximum CLabs (CLabs /CLovt ). The higher the CLmet;int /(CLiÿm +CLiÿs ), the lower the CLabs /CLovt . When metabolic activity (CLmet ) was 10 times greater than membrane permeability (CLiÿm +CLiÿs ), CLabs was approximately one-tenth of CLovt , and when metabolic activity (CLmet ) was 100-fold greater than membrane permeability (CLiÿm +CLiÿs ), CLabs was approximately one-hundredth of CLovt .
Fig. 6. Relationship between absorption clearance and metabolic clearance of KTP and KTP derivatives. Data from Table 2 and reported values [10]. Dots represent CLovt of KTP (0.247 Wl/min/cm) from previous report [10]. b, cyclic KTP; R, Boc-KTP-pAPLglc ; F, Boc-KTP; E, KTP-pAPLglc; O, KTP in the presence of peptidase inhibitor (0.5 mM bestatin, 1 mM o-phenanthroline or 1 mM tryptophan hydroxamate) ; a, KTP.
BBAGEN 25014 26-4-00
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
97
Fig. 7. Impact of relative ratio of metabolic activity to membrane permeability on absorption. (a) CLmet;int /(CLiÿm +CLiÿs ) range from 0 to 10. (b) CLmet;int /(CLiÿm +CLiÿs ) range from 0 to 100.
4. Discussion KTP-pAPLglc was eliminated from the mucosal side (CLeli of 129 Wl/min/cm), but did not appear on the serosal side (Fig. 2 and Table 2). This suggests that metabolic degradation of KTP-pAPLglc occurs in the intestinal tissue. Metabolic degradation of KTP-pAPLglc was also observed in intestinal homogenates, in which the CLmet was much higher than CLeli , indicating that KTP is metabolized by intestinal tissue during the absorption process. Moreover, the CLeli and CLmet of KTP-pAPLglc closely resembled those of KTP (Table 2), indicating that conjugation of pAPLglc to the C-terminus of KTP does not increase the stability of KTP in the intestines. On the other hand, the CLeli of Boc-KTP-pAPLglc indicated that this compound is eliminated from the mucosal side faster than KTP-pAPLglc. Moreover, Boc-KTPpAPLglc appeared on the serosal side (Fig. 3, Table 2), and the CLmet of Boc-KTP-pAPLglc which was similar to that of Boc-KTP [10], was much lower than that of KTPpAPLglc. This indicates that modi¢cation of the N-terminus with the Boc group increased the stability of KTPpAPLglc against aminopeptidase. This agrees with previous results indicating that N-terminus modi¢cation of KTP with a Boc group leads to great stabilization [10]. Furthermore, since Boc-KTP-pAPLglc also appeared on the serosal side, this suggests that metabolic degradation, which was prevented by the Boc group, strongly a¡ects serosal appearance. Based on previous studies of cyclic dipeptides indicating increased stability [14,15], we examined intestinal transport and metabolism of cyclic KTP. Our results showed that the CLeli of cyclic KTP was extremely low, and cyclic KTP in intestinal tissue was stable (Figs. 4 and 5, Table 2). CLabs of cyclic KTP was 1.12 Wl/ min/cm, which was higher than any of the other KTP derivatives. These data indicate that the lower the CLmet , the higher the CLabs . As shown in Fig. 6, the CLabs of cyclic KTP, Boc-KTPpAPLglc and Boc-KTP was higher than the CLovt of KTP (0.247 Wl/min/cm) [10], which is the theoretical maximum
of CLabs of KTP as a result of complete inhibition of peptidase, indicating that the derivatization of KTP increases membrane permeability. Moreover, there was an inverse relationship between CLmet and CLabs for the KTP and KTP derivatives. In addition, KTP-pAPLglc, the CLmet of which is equivalent to that of KTP, did not appear on the serosal side. Even when CLmet was decreased by the presence of peptidase inhibitors, KTP could not be absorbed when CLmet exceeded 500 Wl/min/cm. Therefore, the stability of KTP in the intestines is critical to absorption. Moreover, similar results have been observed in the intestinal absorption of several peptides [10,14^17]. Therefore, in order to identify the critical factors in the absorption of peptides, the kinetic in£uence of metabolic activity on absorption was estimated by simulation using the metabolic inhibition model (Fig. 7) [10,17]. We again found an inverse relationship between CLmet;int /(CLiÿm +CLiÿs ) and CLabs /CLovt , such that when metabolic activity (CLmet ) was 10-fold higher than membrane permeability (CLiÿm +CLiÿs ), CLabs was about one-tenth of CLovt . Since passive membrane transport (absorption) depends on hydrophobicity (log P) [19], the CLovt of hydrophilic peptides and protein drugs is lower than that of general small molecular drugs, such as well-absorbed acetaminophen (2.01 Wl/min/cm, unpublished data), and should be less than 1 Wl/min/cm. Therefore, when CLmet;int /(CLiÿm + CLiÿs ) is greater than 10, a signi¢cant amount of peptides and protein drugs cannot be absorbed unless the stability is improved. This is supported by the contrast results of K-naphthol absorption [20], in which K-naphthol was metabolized by intestinal glucuronidation, but was absorbed. The CLovt of K-naphthol was 7.17 Wl/min/cm, and the CLmet (glucuronidation metabolic clearance) was less than 10 Wl/min/cm [20]. On the other hand, the CLovt of KTP was 0.247 Wl/min/cm and when CLmet was more than about 500 Wl/min/cm, even in the presence of peptidase inhibitors (tryptophan hydroxamate), KTP could not be absorbed. Thus, some metabolic/digestive peptidases and proteases have activity that far exceeds their membrane
BBAGEN 25014 26-4-00
98
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
permeability in the intestines. Therefore, it is concluded that peptide and protein drugs must be stabilized to improve intestinal absorption, particularly when CLmet exceeds 100 Wl/min/cm as a standard. Finally, we found that the CLabs of cyclic KTP, which was highest among the KTP derivatives, was as much as approximately a half of that of acetaminophen, and is stable in serum, suggesting that cyclic KTP should be considered as a candidate analgesic peptide drug. Acknowledgements The authors thank Ms. Tomoko Saito and Ms. Ikuyo Sato for their technical assistance.
References [1] V. Ganapathy, F.H. Leibach, Is intestinal peptide transport energized by a proton gradient?, Am. J. Physiol. 249 (1985) G153^G160. [2] Y.-J. Fei, Y. Kanai, S. Nussberger, V. Ganapathy, F.H. Leibach, M.F. Romero, S.K. Singh, W.F. Boron, M.A. Hediger, Expression cloning of a mammalian proton-coupled oligopeptide transporter, Nature 368 (1994) 563^566. [3] H. Takagi, H. Shiomi, H. Ueda, H. Amano, A novel analgesic dipeptide from bovine brain is a possible Met-enkephalin releaser, Nature 282 (1979) 410^412. [4] P.K. Janicki, A.W. Lipkowski, Kyotorphin and D-kyotorphin stimulate Met-enkephalin release from striatum in vitro, Neurosci. Lett. 43 (1983) 73^77. [5] H. Ueda, H. Shiomi, H. Takagi, Regional distribution of a novel analgesic dipeptide kyotorphin (Tyr-Arg) in the rat brain and spinal cord, Brain Res. 198 (1980) 460^464. [6] K. Matsubayashi, C. Kojima, S. Kawajiri, K. Ono, T. Takegoshi, H. Ueda, H. Takagi, Hydrolytic deactivation of kyotorphin by the rodent brain homogenates and sera, J. Pharmacobio-Dyn. 7 (1984) 479^484. [7] H. Ueda, G. Ming, T. Hazato, T. Katayama, H. Takagi, Degradation of kyotorphin by a puri¢ed membrane-bound-aminopeptidase from monkey brain: potentiation of kyotorphin-induced analgesia by a highly e¡ective inhibitor, bestatin, Life Sci. 36 (1985) 1865^1871.
[8] K. Akasaki, H. Tsuji, An enkephalin-degrading aminopeptidase from rat brain catalyzed the hydrolysis of a neuropeptide, kyotorphin (LTyr-L-Arg), Chem. Pharm. Bull. 39 (1991) 1883^1885. [9] A.T. Orawski, W.H. Simmons, Dipeptidase activities in rat brain synaptosomes can be distinguished on the basis of inhibition by bestatin and amastatin: identi¢cation of a kyotorphin (Tyr-Arg)-degrading enzyme, Neurochem. Res. 17 (1992) 817^820. [10] T. Mizuma, A. Koyanagi, S. Awazu, Intestinal transport and metabolism of analgesic dipeptide, kyotorphin : rate-limiting factor in intestinal absorption of peptide as drug, Biochim. Biophys. Acta 1335 (1997) 111^119. [11] T. Mizuma, K. Ohta, M. Hayashi, S. Awazu, Intestinal active absorption of sugar-conjugated compounds by glucose transport system : implication of improvement of poorly absorbable drugs, Biochem. Pharmacol. 43 (1992) 2037^2039. [12] T. Mizuma, K. Ohta, S. Awazu, The L-anomeric and glucose preferences of glucose transport carrier for intestinal active absorption of monosaccharide conjugates, Biochim. Biophys. Acta 1200 (1994) 117^122. [13] T. Mizuma, Y. Nagamine, A. Dobashi, S. Awazu, Factors that cause L-anomeric preference of Na /glucose cotransporter for intestinal transport of monosaccharide conjugates, Biochim. Biophys. Acta 1381 (1998) 340^346. [14] T. Mizuma, S. Masubuchi, S. Awazu, Intestinal absorption of stable cyclic glycylphenylalanine : comparison with linear form, J. Pharm. Pharmacol. 49 (1997) 1067^1071. [15] T. Mizuma, S. Masubuchi, S. Awazu, Intestinal absorption of stable cyclic dipeptides by oligopeptide transporter, J. Pharm. Pharmacol. 50 (1998) 167^172. [16] T. Mizuma, K. Ohta, S. Awazu, Intestinal absorption of lactonolactone-coupled tripeptide and stability in serum : e¡ect of linkage between peptide and monosaccharide moiety, Protein Peptide Lett. 4 (1997) 115^122. [17] T. Mizuma, K. Ohta, A. Koyanagi, S. Awazu, Improvement of intestinal absorption of leucine enkephalin by sugar coupling and peptidase inhibitors, J. Pharm. Sci. 85 (1996) 854^857. [18] M. Gibaldi, D. Perrier, Pharmacokinetics, Dekker, New York, 1982, pp. 1^43. [19] Y.C. Martin, Practioner's perspective of the role of quantitative structure^activity analysis in medicinal chemistry, J. Med. Chem. 24 (1981) 229^237. [20] T. Mizuma, S. Awazu, Inhibitory e¡ect of phloridzin and phloretin on glucuronidation of p-nitrophenol, acetaminophen and 1-naphthol: kinetic demonstration of the in£uence of glucuronidation metabolism on intestinal absorption in rats, Biochim. Biophys. Acta 1425 (1998) 398^404.
BBAGEN 25014 26-4-00
www.elsevier.com/locate/bba
Intestinal transport and metabolism of glucose-conjugated kyotorphin and cyclic kyotorphin: metabolic degradation is crucial to intestinal absorption of peptide drugs Takashi Mizuma *, Akihiro Koyanagi, Shoji Awazu Department of Biopharmaceutics and Drug Rational Research Center, School of Pharmacy, Tokyo Yakka University (Tokyo University of Pharmacy and Life Science, TUPLS), 1432-1 Horinouchi, Hachioji, Tokyo 192-03, Japan Received 22 December 1999 ; received in revised form 21 March 2000; accepted 22 March 2000
Abstract Intestinal transport and metabolism of modified kyotorphin (KTP) were studied in rats. Modified KTPs studied were C-terminally modified KTP with p-aminophenyl-L-D-glucoside (KTP-pAPLglc), N-terminally modified KTP-pAPLglc with t-butyloxycarbonyl group (Boc-KTP-pAPLglc) and the N- and C-terminally modified KTP by cyclization (cyclic KTP). KTP-pAPLglc was metabolized at a similar rate to that of KTP, and did not appear on the serosal side. Although Boc-KTP-pAPLglc was also metabolized, it was more stable than KTP and appeared on the serosal side. Cyclic KTP was also quite stable and appeared on the serosal side. The modified KTPs were evaluated kinetically for absorption consisting of membrane transport and metabolism. Absorption clearance (CLabs ) of cyclic KTP, BocKTP-pAPLglc and Boc-KTP was higher than that of KTP (0.247 Wl/min/cm) (Mizuma et al., Biochim. Biophys. Acta 1335 (1997) 111^119), which is the theoretical maximum by complete inhibition of peptidase activity, indicating that derivatization of KTP increases the membrane permeability. Furthermore, the data clearly showed that the greater the metabolic clearance (CLmet ) of KTP and the KTP derivatives, the lower the absorption clearance (CLabs ). These results and further simulation study led to the conclusion that metabolic degradation in the intestinal tissues is more critical than membrane permeability (transport) for oral delivery of peptide drugs. Based on the stability of cyclic KTP in serum, this appears to be a good candidate analgesic peptide drug. ß 2000 Elsevier Science B.V. All rights reserved. Keywords : Kyotorphin; Intestinal absorption; Peptide absorption ; Metabolism; Transport; Glucose conjugate; Cyclic peptide; Kinetic analysis
1. Introduction Dietary peptides and proteins are digested/metabolized in the intestines to be transported as amino acids or oligopeptides [1,2]. However, peptide and protein drugs must be transported without metabolic degradation to systemic circulation to exert their pharmacological actions. Kyotorphin (L-tyrosyl-L-arginine, KTP) is an endogenous compound that releases methionine enkephalin, which possesses analgesic activity [3], from the striatum [4]. Although the physiological activity, distribution in the
* Corresponding author. Fax: +81-426-763142; E-mail : [email protected]
brain [5] and metabolism in the brain [6^9] and serum [6] of KTP as an endogenous compound have been studied to clarify its physiological role, very few researchers have examined the intestinal absorption of oral administration of KTP as a medicinal exogenous compound [10]. Intestinal absorption consists of integrated processes of membrane permeation (transport) and enzymatic reactions (metabolism). Therefore, in order to develop orally active peptide drugs, the biophysical and biochemical processes in absorption must be examined kinetically. The present study examines intestinal transport and metabolism of chemically modi¢ed KTP as a model peptide drug (Fig. 1). We previously found that KTP is too unstable in the intestines to be absorbed. However, N-terminal modi¢cation of KTP with a t-butoxycarbonyl (Boc) group provides su¤cient stabilization against peptidases in intestinal tissue [10]. We have also found that glucose or galactose conjugation provide a novel route by way of Na /glucose cotransporter (SGLT1) for intestinal absorption [11^13].
0304-4165 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 4 1 6 5 ( 0 0 ) 0 0 0 5 1 - 9
BBAGEN 25014 26-4-00
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
91
Fig. 1. Structures of kyotorphin conjugated with p-aminophenyl L-glucoside (KTP-pAPLglc), N-t-Butoxycarbonyl-kyotorphin conjugated with p-aminophenyl L-glucoside (Boc-KTP-pAPLglc) and cyclic kyotorphin (cyclic KTP).
Therefore, KTP was modi¢ed with p-aminophenyl L-glucoside (pAPLglc) at the C-terminus to form L-glucose-conjugated KTP. t-Butoxycarbonyl KTP (Boc-KTP) was also conjugated with pAPLglc. We have also examined a cyclic form of KTP, since cyclic dipeptides are stable enough for absorption by the intestines [14,15]. Based on kinetic analysis of these data and simulation study, we identi¢ed the critical factors involved in the absorption of orally administered KTP as well as other pharmacologically active peptides that must be absorbed intact.
2.2. Animals
2. Materials and methods
First, Boc-Arg(NO2 )-pAPLglc was synthesized from Boc-Arg(NO2 ) and pAPLglc. Boc-Arg(NO2 ) (1700 mg) and pAPLglc (700 mg) were dissolved in 8 ml of N,Ndimethylformamide (DMF) and 720 Wl of triethanolamine by stirring at room temperature for 30 min. DCC (1100 mg) solution dissolved in 2 ml of ethyl acetate was dropped into the resultant mixture, and was stirred for 2 h. After further stirring the reaction mixture at 4³C for 22 h under light protection, DCC (280 mg) solution dissolved in 0.5 ml of ethyl acetate was dropped into the reaction mixture, and was stirred at 4³C for 24 h under light protection. The reaction mixture was ¢ltrated through a paper ¢lter and evaporated at 40³C under reduced pressure. The residue containing Boc-Arg(NO2 )-pAPLglc was dissolved in 10 ml of ethyl acetate, and washed with 10 ml of puri¢ed water three times. Ethyl acetate containing Boc-
2.1. Chemicals p-Aminophenyl L-D-glucopyranoside (pAPLglc) was purchased from Sigma (St. Louis, USA). NK -t-butyloxycarbonyl-Ng -nitro-L-arginine (Boc-Arg(NO2 )) and N,NPdicyclohexylcarbodiimide (DCC) were obtained from Peptide Institute, (Osaka, Japan). N-t-Butyloxycarbonyl-L-tyrosine (Boc-Tyr) and palladium-activated carbon (10%, Pd-C) were purchased from Wako (Osaka, Japan). Cyclic KTP (Cyclo(L-arginyl-L-tyrosine)) was purchased from Iwaki Glass (Tokyo, Japan). Sephadex G-10 was obtained from Pharmacia LKB Biotechnology (Tokyo, Japan). KTP-pAPLglc and Boc-KTP-pAPLglc were synthesized by the method described below.
Adults male Wistar rats (8 weeks old) were obtained from Japan SLC (Shizuoka, Japan) and allowed free access to water and commercially available chows in a temperature- and light-controlled room for more than 1 week before experimentation. 2.3. Synthesis of N-t-butyloxycarbonyl-L-kyotorphin conjugated with p-aminophenyl L-glucoside (Boc-KTP-pAPLglc)
BBAGEN 25014 26-4-00
92
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
Arg(NO2 )-pAPLglc was evaporated, and crude BocArg(NO2 )-pAPLglc was applied to a Sephadex G-10 column (2.5 cm i.d.U90 cm length). Fractions (10 ml per tube) eluted with puri¢ed water at a £ow rate of 2.0 ml/ min were analyzed by HPLC (TSKgel ODS-80TM, 6 mm i.d. U150 mm length, 25% acetonitrile, 0.05% phosphoric acid, 1.5 ml/min, 274 nm). Fractions containing BocArg(NO2 )-pAPLglc were collected and concentrated by evaporation at 40³C under reduced pressure. Concentrated Boc-Arg(NO2 )-pAPLglc solution was applied to a preparative ODS column (3 cm i.d.U30 cm length, GL Science, Tokyo) and eluted with a mobile phase (18% acetonitrile, 1.5% acetic acid) at a £ow rate of 2.5 ml/min. Each fraction (10 ml per tube) containing Boc-Arg(NO2 )-pAPLglc was analyzed by HPLC as above and collected. Collected fractions were evaporated under the above conditions. The residue was dissolved in 5 ml of acetonitrile and 5 ml of TFA, and ¢lled with nitrogen gas. After stirring the solution for 2 h at room temperature and evaporating, crude Arg(NO2 )-pAPLglc was obtained. Crude Arg(NO2 )pAPLglc was applied to a preparative ODS column (3 cm i.d.U30 cm length) and eluted with a mobile phase (3% methanol, 1.5% acetic acid) at a £ow rate of 2.5 ml/min. Each fraction (10 ml per tube) containing Arg(NO2 )pAPLglc eluted was analyzed by HPLC (TSKgel ODS80TM, 6 mm i.d.U150 mm length, 8% methanol, 0.05% phosphoric acid, 1.5 ml/min, 274 nm) and collected. Fractions were evaporated at 40³C under reduced pressure. Boc-Tyr (140 mg) and Arg(NO2 )-pAPLglc (120 mg) were dissolved in 4 ml of DMF and 71 Wl of triethanolamine by stirring at room temperature for 30 min. DCC dissolved in ethyl acetate was dropped into the mixture in 30-ml Erlenmeyer £asks on ice. Then, the reaction mixture was stirred for 2 h at 0³C, followed by 22 h at 4³C under light protection. After Boc-Tyr (70 mg) was added to the reaction mixture, DCC (52 mg) solution in 0.5 ml of ethyl acetate was dropped into the reaction mixture at 0³C to be stirred for 24 h under light protection. The reaction mixture was ¢ltrated through a paper ¢lter and evaporated. The resultant residue was dissolved in 10 ml of ethyl acetate, and was washed with puri¢ed water three times. Ethyl acetate was then evaporated. The crude Boc-TyrArg(NO2 )-pAPLglc was applied to a Sephadex G-10 column (2.5 cm i.d.U90 cm length), and eluted with puri¢ed water at a £ow rate of 2.0 ml/min. Each fraction (15 ml per tube) eluted was analyzed by HPLC (TSKgel ODS80TM, 6 mm i.d.U15 cm length, 30% acetonitrile, 0.05% phosphoric acid, 1.5 ml/min £ow rate, 274 nm UV absorbance). Fractions containing Boc-Tyr-Arg(NO2 )-pAPLglc were collected and evaporated at 40³C under reduced pressure. Concentrated solution of Boc-Tyr-Arg(NO2 )pAPLglc was applied to a preparative ODS column (ODS, 30 mm i.d.U300 mm length) and eluted with a mobile phase (27% acetonitrile, 1.5% acetic acid) at a £ow rate of 2.5 ml/min. Each fraction (10 ml per tube)
eluted was analyzed by HPLC as above. Fractions containing Boc-Tyr-Arg(NO2 )-pAPLglc were collected and evaporated at 40³C under reduced pressure. Puri¢ed Boc-Tyr-Arg(NO2 )-pAPLglc was dissolved in 10 ml of methanol, and ¢lled with nitrogen gas. After Pd-C was added to the solution, it was ¢lled with hydrogen gas and allowed to stand for 24 h under light protection at room temperature to form Boc-KTP-pAPLglc. De-protection of the Boc group from Boc-KTP-pAPLglc was con¢rmed by HPLC (TSKgel ODS-80TM, 6 mm i.d.U150 mm length, mobile phase consisting of 25% acetonitrile, 0.05% phosphoric acid, 1.5 ml/min, 274 nm). The BocKTP-pAPLglc solution was ¢ltrated through a paper ¢lter and concentrated by evaporation at 40³C under reduced pressure. Boc-KTP-pAPLglc solution was applied to a preparative ODS column (ODS, 30 mm i.d.U300 mm length) and eluted with a mobile phase (20% acetonitrile, 1.5% acetic acid) at a £ow rate of 2.5 ml/min. Each fraction (10 ml per tube) eluted was analyzed by HPLC as above. Fractions containing Boc-KTP-pAPLglc were collected and evaporated at 40³C under reduced pressure. 2.4. Synthesis of L-kyotorphin conjugated with p-aminophenyl L-glucoside (KTP-pAPLglc) Puri¢ed Boc-KTP-pAPLglc, which was prepared by the method described above, was dissolved in 10 ml of methanol, and ¢lled with nitrogen gas. After Pd-C was added to the solution, it was ¢lled with hydrogen gas and allowed to stand for 24 h under light protection at room temperature to form KTP-pAPLglc. KTP-pAPLglc was applied to semi-preparative HPLC (TSKgel ODS-80TM, 7.8 mm i.d.U300 mm length) and eluted with a mobile phase (1.5% methanol, 1.5% acetic acid) at a £ow rate of 3.0 ml/min. Each fraction (10 ml per tube) eluted was analyzed by HPLC as above. Fractions containing KTPpAPLglc were collected and evaporated at 40³C under reduced pressure. Residues of KTP-pAPLglc were dissolved in puri¢ed water. Finally, the powder of KTPpAPLglc was obtained by freeze-drying. 2.5. Identi¢cation of Boc-Tyr-Arg-pAPLglc and Tyr-Arg-pAPLglc ESI mass spectra of Boc-KTP-pAPLglc (Boc-Tyr-ArgpAPLglc) and KTP-pAPLglc (Tyr-Arg-pAPLglc) showed molecular ion peaks [M+H] at 691.7 and 591.7, respectively. The ninhydrin reaction of KTP-pAPLglc and TFAtreated Boc-KTP-pAPLglc which were developed on TLC, were positive, indicating the free amino group of KTPpAPLglc and protected amino group of Boc-KTPpAPLglc. The phenol-sulfuric acid reaction of Boc-KTPpAPLglc and KTP-pAPLglc which was developed on TLC, were positive, indicating the existence of a monosaccharide moiety.
BBAGEN 25014 26-4-00
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
93
2.6. Intestinal absorption
2.8. Stability of cyclic KTP in serum
Intestinal absorption of modi¢ed KTPs was examined in excised rat small intestines [11]. Brie£y, male Wistar rats (180^230 g, Japan SLC, Japan) fasted overnight were anesthetized with ether, and the intestinal blood was removed by saline perfusion. The jejunum was removed and everted. Ten cm of everted small intestine was placed in 30 ml of incubation medium (113.3 mM NaCl, 4.83 mM KCl, 1.214 mM KH2 PO4 , 1.205 mM MgSO4 , 16.96 mM NaHCO3 , 10.18 mM Na2 HPO4 , 0.645 mM CaCl2 , pH 7.4) containing 250 WM peptide in a beaker at 37³C. The serosal side was ¢lled with 5 ml of the incubation medium without peptides. When necessary, peptidase inhibitors were added to the medium on the mucosal and serosal sides. Incubation medium (100 Wl) was sampled from both the serosal and mucosal sides every 10 min until 60 min. The samples were mixed with 100 Wl of internal standard solution (200 WM tryptophan in 10% perchloric acid for Tyr-Arg-pAPLglc, or 100 WM sodium 1-naphthalensulfonate in acetonitrile for Boc-KTP-pAPLglc, or 1 mM p-hydroxybenzoic acid for cyclic KTP) for the following HPLC assay. The mixture was centrifuged at 11 000Ug for 5 min using a benchtop centrifuge KM15200 (Kubota, Japan). Twenty-¢ve microliters of the resultant supernatant was applied to HPLC.
The stability experiment was carried according to the method reported previously [16]. Rat blood was obtained through the cannulated carotid artery under ether anesthesia. After keeping on ice for 30 min, the blood was centrifuged at 1300Ug for 10 min. The resultant supernatant was supplied as serum. Next, 0.1 ml of cyclic KTP solution (20 WM ¢nal concentration) in incubation medium was added to 0.9 ml of serum in a microtube (1.5 ml volume), and incubated at 37³C. Samples were obtained periodically over 3 h, and cyclic KTP was determined by HPLC.
2.7. Metabolism of KTP-pAPLglc and Boc-KTP-pAPLglc in intestinal homogenates Examination of modi¢ed KTP metabolism in intestinal homogenates was performed by a previously reported method [10]. Brie£y, 10 cm of intestine was isolated, and homogenized in the same bu¡er as described in the absorption experiment by Physcotron homogenizer (Nichion Irika, Tokyo, Japan). A mixture of the peptide (250 WM ¢nal concentration) and homogenate (0.25% ¢nal concentration) was incubated for 60 min at 37³C in the same incubation medium used for the absorption experiment. At the indicated times, 100 Wl of reaction mixture was sampled. Modi¢ed KTP was determined by the same method as in the absorption experiment.
2.9. HPLC assay for absorption and metabolism experiments KTP-pAPLglc, Boc-KTP-pAPLglc, cyclic KTP and their metabolites were determined by reversed-phase HPLC. The HPLC system consisted of a pump (Twincle, Jusco, Tokyo, Japan), UV detector (Shimadzu, Kyoto, Japan), £uorescence detector (821-FP, Ex 278 nm, Em 305 nm, Jasco) and an integrator (D-2500, Hitachi, Tokyo, Japan). The £ow rate was set at 1.5 ml/min. Complete assay conditions are listed in Table 1. In the Boc-KTP assay, the mobile phase was composed of 20% acetonitrile and 0.05% phosphoric acid in water and an ODS column (A-312, 6 mm i.d.U15 cm length, YMC, Japan) was used. In the cyclic KTP assay, the mobile phase was composed of 30% methanol, 0.05% phosphoric acid, and a C8 column (TSKgel Octyl-80Ts, 4.6 mm i.d.U15 cm length, Tosoh, Japan) was used. A UV detector (274 nm) was used in the puri¢cation experiments and a £uorescence detector (Ex 278 nm, Em 305 nm) was used in the absorption experiments. The £ow rate was set at 1.5 ml/min. For the assay of cyclic KTP in stability experiment, 28.5% acetonitrile and 0.05% phosphoric acid and ODS 80TM (6 mm i.d.U15 cm length) were used. 2.10. Data analysis Elimination clearance, CLeli , from the mucosal side was
Table 1 HPLC assay conditions for absorption experiments of KTP-pAPLglc, Boc-KTP-pAPLglc and cyclic KTP Conditions
KTP-pAPLglc
Boc-KTP-pAPLglc
cyclic KTP
Column
TSKgel ODS-80TM 6 mm i.dU150 mm length 7% methanol, 0.05% phosphoric acid in water 200 WM tryptophan in 10% perchloric acid 821-FP (Jasco) Ex 278 nm, Em 305 nm
YMC-Pack ODS A-312 6 mm i.d.U150 mm length 20% acetonitrile, 0.05% phosphoric acid in water 100 WM sodium 1-naphthalenesulfonate in acetonitrile 821-FP (Jasco) Ex 278 nm, 305 nm
TSKgel ODS-80Ts 6 mm i.d.U150 mm length 30% methanol, 0.05% phosphoric acid in water 1 mM hydroxybenzoic acid in 10% perchloric acid 665A-21 UV detector (Hitachi) 194 nm
Mobile phase Internal standard Detector Wavelength
BBAGEN 25014 26-4-00
94
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
calculated by Eq. 1 [17]. CLeli
X eli AUCmuc;0ÿ60
1
where Xeli is amount eliminated from the mucosal side as calculated by : X eli C 60 3C 0 UV
2
AUCmuc;0ÿ60 is the area under the concentration curve of the peptide on the mucosal side from time 0 to 60 min calculated by the trapezoidal rule [18]. C60 and C0 represent the modi¢ed KTP concentration on the mucosal side at 60 and 0 min, respectively. V is the volume of medium on the mucosal side. Absorption clearance, CLabs , was calculated by Eq. 3 [17]. CLabs
X abs AUCmuc;0ÿ60
3
where Xabs is the amount of modi¢ed KTP absorbed from the mucosal side to the serosal side over 60 min. Metabolic clearance (per cm of intestine), CLmet , in intestinal homogenate was calculated by Eqs. 4 and 5. CLmet
X met UW w AUChom;0ÿ60
X met C 60 3C 0 U
100 H
4 5
where Xmet is the amount of metabolite formed per 1 g of wet weight of intestine in the intestinal homogenates as calculated by Eq. 5. AUChom;0ÿ60 is the area under the concentration curve of modi¢ed KTP in the reaction mixture from time 0 to 60 min, as calculated by the trapezoidal rule. Ww is wet weight per length of everted intestine (g/cm). C60 and C0 represent the modi¢ed KTP concentration in the reaction mixture at 60 and 0 min, respectively. H is the concentration of homogenates (%, w/v) of the reaction mixture. The kinetic relationship between absorption clearance (CLabs ) and metabolic clearance (CLmet ) is expressed by
Scheme 1. Metabolic inhibition model for intestinal absorption of drugs. Ci , Cs and Cm are drug concentrations in intestinal tissue, on serosal side, and mucosal side, respectively. CLmÿi , CLiÿm and CLiÿs are the transport clearance from mucosal side to intestinal tissue, and from intestinal tissue to mucosal side, and from intestinal tissue to serosal side, respectively. CLmet;int is intrinsic clearance of metabolism by metabolic enzymes in intestinal tissue. Ra is remaining activity of metabolic enzyme in intestinal tissue in presence of peptidase inhibitors.
Eq. 6, which is based on the metabolic inhibition model for absorption (Scheme 1) [10,17]. CLabs
CLovt
CLovt CLmet 1 CLiÿm CLiÿs CLmÿi UCLiÿs CLiÿm CLiÿs
6
7
CLmet CLmet;int URa where CLovt is overall transport clearance of modi¢ed KTP from the mucosal side to the serosal side, where no metabolic degradation of modi¢ed KTP in intestinal tissue occurs during the absorption process. Under this assumption, CLovt is the maximum value of CLabs . CLmÿi , CLiÿm and CLiÿs represent the transport clearance of modi¢ed KTP from the mucosal side to the intestinal tissue, from the intestinal tissue to the mucosal side, and from the intestinal tissue to the serosal side, respectively. CLmet;int
Fig. 2. Time course of KTP-pAPLglc on mucosal (a) and serosal (b) sides in intestinal absorption. Data represent mean þ S.E. (n = 3).
BBAGEN 25014 26-4-00
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
Fig. 3. Time course of Boc-KTP-pAPLglc on mucosal (a) and serosal (b) sides in intestinal absorption. sent mean þ S.E. (n = 3).
is the intrinsic clearance of metabolism in the intestinal tissue. Ra is the remaining activity of the metabolism in the intestinal tissue in the presence of enzyme inhibitors. In the absence of any peptidase inhibitors (Ra = 1), the following equation is obtained. CLabs
CLovt CLmet;int 1 CLiÿm CLiÿs
8
The e¡ect of intestinal metabolism on absorption was simulated according to Eq. 9, which is derived from Eq. 8. CLabs CLovt
1 CLmet;int 1 CLiÿm CLiÿs
9
a,
95
Boc-KTP-pAPLglc ; b, Boc-KTP. Data repre-
3.2. Intestinal transport and metabolism of Boc-KTP-pAPLglc Fig. 3a and b show the time course of Boc-KTPpAPLglc on the mucosal side and serosal side, respectively. Boc-KTP-pAPLglc was eliminated and Boc-Tyr formed on the mucosal side. The elimination clearance of Boc-KTPpAPLglc was 728 þ 280 Wl/min/cm (Table 2). Serosal appearance of Boc-KTP-pAPLglc was observed in one of three experiments. 3.3. Intestinal transport and metabolism of cyclic KTP Fig. 4a and b show the time course of cyclic KTP on the mucosal side and serosal side, respectively. Cyclic KTP was stable on the mucosal side and the elimination clearance of cyclic KTP from the mucosal side was 2.21 þ 0.07 Wl/min/cm (Table 2). Cyclic KTP appeared on the serosal side, and the transport clearance of cyclic KTP was 1.12 þ 0.05Wl/min/cm (Table 2).
3. Results 3.1. Intestinal transport and metabolism of KTP-pAPLglc Fig. 2a and b show the time course of KTP-pAPLglc on the mucosal side and serosal side, respectively. KTPpAPLglc on the mucosal side was eliminated and the elimination clearance of KTP-pAPLglc from the mucosal side was 129 þ 11 Wl/min/cm (Table 2). On the other hand, KTP-pAPLglc was not detectable on the serosal side.
3.4. Metabolic degradation of KTP-pAPLglc and Boc-KTP-pAPLglc in intestinal homogenates Fig. 5 show the time course of concentration changes of KTP-pAPLglc, Boc-KTP-pAPLglc and cyclic KTP in the intestinal homogenates. KTP-pAPLglc was eliminated much faster than Boc-KTP-pAPLglc, whereas cyclic KTP
Table 2 Absorption clearance (CLabs ), elimination clearance (CLeli ) and metabolic clearance (CLmet ) of KTP-pAPLglc, Boc-KTP-pAPLglc and cyclic KTP KTPa KTP-pAPLglc Boc-KTPa Boc-KTP-pAPLglc Cyclic KTP
CLabs (Wl/min/cm)
CLeli (Wl/min/cm)
CLmet (Wl/min/cm)
ND ND 0.257 þ 0.11 0.726b 1.12 þ 0.05
129 þ 3.0 129 þ 11 357 þ 67 728 þ 280 2.21 þ 0.07
5120 þ 50 4630 þ 210 32.2 þ 7.6 74.9 þ 11.9 stable
Data represent mean þ S.E. (n = 3). a Data from previous report [10]. b Only one of three experiments showed the serosal appearance of Boc-KTP-pAPLglc, because of higher detection limit of concentration in HPLC assay.
BBAGEN 25014 26-4-00
96
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
Fig. 4. Time course of cyclic KTP on mucosal (a) and serosal (b) sides in intestinal absorption. Data represent mean þ S.E. (n = 3).
remained stable. The metabolic clearance of KTP-pAPLglc and Boc-KTP-pAPLglc were 4630 and 74.9 Wl/min/cm, respectively (Table 2). 3.5. Stability of cyclic KTP in serum Cyclic KTP was stable in serum for the full 60-min period (data not shown). 3.6. Relationship between absorption clearance and metabolic clearance The higher the CLmet of KTP and the KTP derivatives, the lower the CLabs (Fig. 6). The CLabs of cyclic KTP, Boc-KTP-pAPLglc and Boc-KTP, but not KTP-pAPLglc, were higher than CLovt of KTP. When the CLmet exceeded about 500 Wl/min/cm, KTP could not be absorbed.
Fig. 5. Stability of KTP-pAPLglc, Boc-KTP-pAPLglc and cyclic KTP in intestinal homogenates. a, KTP-pAPLglc; b, Boc-KTP-pAPLglc; E, cyclic KTP. Data represent mean þ S.E. KTP-pAPLglc (n = 3), Boc-KTPpAPLglc (n = 3), cyclic KTP (n = 5).
3.7. Simulation of e¡ect of intrinsic metabolic activity on the absorption Fig. 7, which was obtained by Eq. 9, shows the impact of the relative ratio of metabolic activity to membrane permeability (CLmet;int /(CLiÿm +CLiÿs )) on the absorption expressed as the ratio of CLabs to the maximum CLabs (CLabs /CLovt ). The higher the CLmet;int /(CLiÿm +CLiÿs ), the lower the CLabs /CLovt . When metabolic activity (CLmet ) was 10 times greater than membrane permeability (CLiÿm +CLiÿs ), CLabs was approximately one-tenth of CLovt , and when metabolic activity (CLmet ) was 100-fold greater than membrane permeability (CLiÿm +CLiÿs ), CLabs was approximately one-hundredth of CLovt .
Fig. 6. Relationship between absorption clearance and metabolic clearance of KTP and KTP derivatives. Data from Table 2 and reported values [10]. Dots represent CLovt of KTP (0.247 Wl/min/cm) from previous report [10]. b, cyclic KTP; R, Boc-KTP-pAPLglc ; F, Boc-KTP; E, KTP-pAPLglc; O, KTP in the presence of peptidase inhibitor (0.5 mM bestatin, 1 mM o-phenanthroline or 1 mM tryptophan hydroxamate) ; a, KTP.
BBAGEN 25014 26-4-00
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
97
Fig. 7. Impact of relative ratio of metabolic activity to membrane permeability on absorption. (a) CLmet;int /(CLiÿm +CLiÿs ) range from 0 to 10. (b) CLmet;int /(CLiÿm +CLiÿs ) range from 0 to 100.
4. Discussion KTP-pAPLglc was eliminated from the mucosal side (CLeli of 129 Wl/min/cm), but did not appear on the serosal side (Fig. 2 and Table 2). This suggests that metabolic degradation of KTP-pAPLglc occurs in the intestinal tissue. Metabolic degradation of KTP-pAPLglc was also observed in intestinal homogenates, in which the CLmet was much higher than CLeli , indicating that KTP is metabolized by intestinal tissue during the absorption process. Moreover, the CLeli and CLmet of KTP-pAPLglc closely resembled those of KTP (Table 2), indicating that conjugation of pAPLglc to the C-terminus of KTP does not increase the stability of KTP in the intestines. On the other hand, the CLeli of Boc-KTP-pAPLglc indicated that this compound is eliminated from the mucosal side faster than KTP-pAPLglc. Moreover, Boc-KTPpAPLglc appeared on the serosal side (Fig. 3, Table 2), and the CLmet of Boc-KTP-pAPLglc which was similar to that of Boc-KTP [10], was much lower than that of KTPpAPLglc. This indicates that modi¢cation of the N-terminus with the Boc group increased the stability of KTPpAPLglc against aminopeptidase. This agrees with previous results indicating that N-terminus modi¢cation of KTP with a Boc group leads to great stabilization [10]. Furthermore, since Boc-KTP-pAPLglc also appeared on the serosal side, this suggests that metabolic degradation, which was prevented by the Boc group, strongly a¡ects serosal appearance. Based on previous studies of cyclic dipeptides indicating increased stability [14,15], we examined intestinal transport and metabolism of cyclic KTP. Our results showed that the CLeli of cyclic KTP was extremely low, and cyclic KTP in intestinal tissue was stable (Figs. 4 and 5, Table 2). CLabs of cyclic KTP was 1.12 Wl/ min/cm, which was higher than any of the other KTP derivatives. These data indicate that the lower the CLmet , the higher the CLabs . As shown in Fig. 6, the CLabs of cyclic KTP, Boc-KTPpAPLglc and Boc-KTP was higher than the CLovt of KTP (0.247 Wl/min/cm) [10], which is the theoretical maximum
of CLabs of KTP as a result of complete inhibition of peptidase, indicating that the derivatization of KTP increases membrane permeability. Moreover, there was an inverse relationship between CLmet and CLabs for the KTP and KTP derivatives. In addition, KTP-pAPLglc, the CLmet of which is equivalent to that of KTP, did not appear on the serosal side. Even when CLmet was decreased by the presence of peptidase inhibitors, KTP could not be absorbed when CLmet exceeded 500 Wl/min/cm. Therefore, the stability of KTP in the intestines is critical to absorption. Moreover, similar results have been observed in the intestinal absorption of several peptides [10,14^17]. Therefore, in order to identify the critical factors in the absorption of peptides, the kinetic in£uence of metabolic activity on absorption was estimated by simulation using the metabolic inhibition model (Fig. 7) [10,17]. We again found an inverse relationship between CLmet;int /(CLiÿm +CLiÿs ) and CLabs /CLovt , such that when metabolic activity (CLmet ) was 10-fold higher than membrane permeability (CLiÿm +CLiÿs ), CLabs was about one-tenth of CLovt . Since passive membrane transport (absorption) depends on hydrophobicity (log P) [19], the CLovt of hydrophilic peptides and protein drugs is lower than that of general small molecular drugs, such as well-absorbed acetaminophen (2.01 Wl/min/cm, unpublished data), and should be less than 1 Wl/min/cm. Therefore, when CLmet;int /(CLiÿm + CLiÿs ) is greater than 10, a signi¢cant amount of peptides and protein drugs cannot be absorbed unless the stability is improved. This is supported by the contrast results of K-naphthol absorption [20], in which K-naphthol was metabolized by intestinal glucuronidation, but was absorbed. The CLovt of K-naphthol was 7.17 Wl/min/cm, and the CLmet (glucuronidation metabolic clearance) was less than 10 Wl/min/cm [20]. On the other hand, the CLovt of KTP was 0.247 Wl/min/cm and when CLmet was more than about 500 Wl/min/cm, even in the presence of peptidase inhibitors (tryptophan hydroxamate), KTP could not be absorbed. Thus, some metabolic/digestive peptidases and proteases have activity that far exceeds their membrane
BBAGEN 25014 26-4-00
98
T. Mizuma et al. / Biochimica et Biophysica Acta 1475 (2000) 90^98
permeability in the intestines. Therefore, it is concluded that peptide and protein drugs must be stabilized to improve intestinal absorption, particularly when CLmet exceeds 100 Wl/min/cm as a standard. Finally, we found that the CLabs of cyclic KTP, which was highest among the KTP derivatives, was as much as approximately a half of that of acetaminophen, and is stable in serum, suggesting that cyclic KTP should be considered as a candidate analgesic peptide drug. Acknowledgements The authors thank Ms. Tomoko Saito and Ms. Ikuyo Sato for their technical assistance.
References [1] V. Ganapathy, F.H. Leibach, Is intestinal peptide transport energized by a proton gradient?, Am. J. Physiol. 249 (1985) G153^G160. [2] Y.-J. Fei, Y. Kanai, S. Nussberger, V. Ganapathy, F.H. Leibach, M.F. Romero, S.K. Singh, W.F. Boron, M.A. Hediger, Expression cloning of a mammalian proton-coupled oligopeptide transporter, Nature 368 (1994) 563^566. [3] H. Takagi, H. Shiomi, H. Ueda, H. Amano, A novel analgesic dipeptide from bovine brain is a possible Met-enkephalin releaser, Nature 282 (1979) 410^412. [4] P.K. Janicki, A.W. Lipkowski, Kyotorphin and D-kyotorphin stimulate Met-enkephalin release from striatum in vitro, Neurosci. Lett. 43 (1983) 73^77. [5] H. Ueda, H. Shiomi, H. Takagi, Regional distribution of a novel analgesic dipeptide kyotorphin (Tyr-Arg) in the rat brain and spinal cord, Brain Res. 198 (1980) 460^464. [6] K. Matsubayashi, C. Kojima, S. Kawajiri, K. Ono, T. Takegoshi, H. Ueda, H. Takagi, Hydrolytic deactivation of kyotorphin by the rodent brain homogenates and sera, J. Pharmacobio-Dyn. 7 (1984) 479^484. [7] H. Ueda, G. Ming, T. Hazato, T. Katayama, H. Takagi, Degradation of kyotorphin by a puri¢ed membrane-bound-aminopeptidase from monkey brain: potentiation of kyotorphin-induced analgesia by a highly e¡ective inhibitor, bestatin, Life Sci. 36 (1985) 1865^1871.
[8] K. Akasaki, H. Tsuji, An enkephalin-degrading aminopeptidase from rat brain catalyzed the hydrolysis of a neuropeptide, kyotorphin (LTyr-L-Arg), Chem. Pharm. Bull. 39 (1991) 1883^1885. [9] A.T. Orawski, W.H. Simmons, Dipeptidase activities in rat brain synaptosomes can be distinguished on the basis of inhibition by bestatin and amastatin: identi¢cation of a kyotorphin (Tyr-Arg)-degrading enzyme, Neurochem. Res. 17 (1992) 817^820. [10] T. Mizuma, A. Koyanagi, S. Awazu, Intestinal transport and metabolism of analgesic dipeptide, kyotorphin : rate-limiting factor in intestinal absorption of peptide as drug, Biochim. Biophys. Acta 1335 (1997) 111^119. [11] T. Mizuma, K. Ohta, M. Hayashi, S. Awazu, Intestinal active absorption of sugar-conjugated compounds by glucose transport system : implication of improvement of poorly absorbable drugs, Biochem. Pharmacol. 43 (1992) 2037^2039. [12] T. Mizuma, K. Ohta, S. Awazu, The L-anomeric and glucose preferences of glucose transport carrier for intestinal active absorption of monosaccharide conjugates, Biochim. Biophys. Acta 1200 (1994) 117^122. [13] T. Mizuma, Y. Nagamine, A. Dobashi, S. Awazu, Factors that cause L-anomeric preference of Na /glucose cotransporter for intestinal transport of monosaccharide conjugates, Biochim. Biophys. Acta 1381 (1998) 340^346. [14] T. Mizuma, S. Masubuchi, S. Awazu, Intestinal absorption of stable cyclic glycylphenylalanine : comparison with linear form, J. Pharm. Pharmacol. 49 (1997) 1067^1071. [15] T. Mizuma, S. Masubuchi, S. Awazu, Intestinal absorption of stable cyclic dipeptides by oligopeptide transporter, J. Pharm. Pharmacol. 50 (1998) 167^172. [16] T. Mizuma, K. Ohta, S. Awazu, Intestinal absorption of lactonolactone-coupled tripeptide and stability in serum : e¡ect of linkage between peptide and monosaccharide moiety, Protein Peptide Lett. 4 (1997) 115^122. [17] T. Mizuma, K. Ohta, A. Koyanagi, S. Awazu, Improvement of intestinal absorption of leucine enkephalin by sugar coupling and peptidase inhibitors, J. Pharm. Sci. 85 (1996) 854^857. [18] M. Gibaldi, D. Perrier, Pharmacokinetics, Dekker, New York, 1982, pp. 1^43. [19] Y.C. Martin, Practioner's perspective of the role of quantitative structure^activity analysis in medicinal chemistry, J. Med. Chem. 24 (1981) 229^237. [20] T. Mizuma, S. Awazu, Inhibitory e¡ect of phloridzin and phloretin on glucuronidation of p-nitrophenol, acetaminophen and 1-naphthol: kinetic demonstration of the in£uence of glucuronidation metabolism on intestinal absorption in rats, Biochim. Biophys. Acta 1425 (1998) 398^404.
BBAGEN 25014 26-4-00