Intestinal absorption of fluorescence-derivatized cationic peptide 001-C8-NBD via adsorptive-mediated transcytosis

Article abstract of DOI:10.1016/S0968-0896(98)00031-5

The intestinal absorption of an intact oligopeptide was investigated in rats using a synthetic cationic peptide, 001-C8 (H-MeTyr-Arg-MeArg-d-Leu-NH(CH₂)₈NH₂). The peptide was coupled with 4-nitrobenzo-2-oxa-1,3-diazole (NBD) to prepare a fluorescence-labeled derivative 001-C8-NBD (H-MeTyr-Arg-MeArg-d-Leu-NH(CH₂)₈NH-NBD) for the purpose of quantification. The degradation half-life of 001-C8-NBD in jejunal homogenate (1mg/mL) was 99.5min, which was significantly longer than that of natural leucine enkephalin (1.14min). The absorption of 001-C8-NBD was evaluated by the vascular-perfusion method. Intact 001-C8-NBD appeared in the blood time-dependently and the absorption volume at 30min (2.75 ± 0.14μL/cm intestine) was significantly larger than that of [³H]PEG 900 (0.88 ± 0.13μL/cm intestine), of which membrane permeability is very low. The absorption of 001-C8-NBD was greatly reduced by an adsorptive-mediated endocytosis inhibitor, protamine (10mM). No inhibition of the absorption of [³H]PEG 900 by protamine was observed. The intestinal absorption was also measured by an in vivo loop method. The absorption clearance of 001-C8-NBD measured by this method (0.083 ± 0.008μL/min/cm intestine) was comparable to that obtained by the vascular perfusion method (0.092 ± 0.005μL/min/cm intestine). All of these data suggested that 001-C8-NBD was absorbed as the intact oligopeptide in the intestine in vivo. Adsorptive-mediated transcytosis is suggested to have enormous potential as an oral delivery system for peptide and/or protein drugs. Copyright (C) 1998 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(98)00031-5

BIOORGANIC &
MEDICINAL
CHEMISTRY
Bioorganic & Medicinal Chemistry 6 (1998) 841±848
Intestinal Absorption of Fluorescence-derivatized Cationic
Peptide 001-C8-NBD via Adsorptive-mediated Transcytosis
Yoshimichi Sai,a Masahiro Kajita,a Ikumi Tamai,a Makoto Kamata,b
Jun Wakama,c Tateaki Wakamiyab, c,y and Akira Tsujia,*
aDepartment of Pharmaceutics, Faculty of Pharmaceutical Sciences, Kanazawa University, Takara-machi, Kanazawa 920, Japan
bDepartment of Chemistry, Faculty of Science, Osaka University, Toyonaka, Osaka 560, Japan
cDepartment of Chemistry, Faculty of Science and Technology, Kinki University, Higashi-osaka, Osaka 577, Japan
Received 2 September 1997; accepted 28 January 1998
AbstractÐThe intestinal absorption of an intact oligopeptide was investigated in rats using a synthetic cationic peptide,
001-C8 (H-MeTyr-Arg-MeArg-D-Leu-NH(CH2)8NH2). The peptide was coupled with 4-nitrobenzo-2-oxa-1,3-diazole
(NBD) to prepare a ¯uorescence-labeled derivative 001-C8-NBD (H-MeTyr-Arg-MeArg-D-Leu-NH(CH2)8NH-NBD)
for the purpose of quanti®cation. The degradation half-life of 001-C8-NBD in jejunal homogenate (1 mg/mL) was
99.5 min, which was signi®cantly longer than that of natural leucine enkephalin (1.14 min). The absorption of 001-C8-
NBD was evaluated by the vascular-perfusion method. Intact 001-C8-NBD appeared in the blood time-dependently
and the absorption volume at 30 min (2.75 ‹ 0.14 mL/cm intestine) was signi®cantly larger than that of [3H]PEG 900
(0.88 ‹ 0.13 mL/cm intestine), of which membrane permeability is very low. The absorption of 001-C8-NBD was
greatly reduced by an adsorptive-mediated endocytosis inhibitor, protamine (10 mM). No inhibition of the absorption
of [3H]PEG 900 by protamine was observed. The intestinal absorption was also measured by an in vivo loop method.
The absorption clearance of 001-C8-NBD measured by this method (0.083 ‹ 0.008 mL/min/cm intestine) was com-
parable to that obtained by the vascular perfusion method (0.092 ‹ 0.005 mL/min/cm intestine). All of these data
suggested that 001-C8-NBD was absorbed as the intact oligopeptide in the intestine in vivo. Adsorptive-mediated
transcytosis is suggested to have enormous potential as an oral delivery system for peptide and/or protein drugs.
# 1998 Elsevier Science Ltd. All rights reserved.
Introduction
the oligopeptide transporter PepT1.1,2 PepT1 transports
only oligopeptides containing up to three amino acid
residues, and cannot transport larger peptides.3±5
Endocytosis could be an alternative for larger peptides.
Receptor-mediated endocytosis mechanisms are
involved in the transport of various immunoglobulins
and growth factors.6 Fluid-phase endocytosis has also
been reported using horseradish peroxidase as a protein
model in the human colon carcinoma cell line Caco-2.7
However, very little information is available on the
intestinal adsorptive-mediated endocytosis (AME),
which is initiated by the binding of cationic moiety of
substances to negative charges on the plasma mem-
brane.
The therapeutic utility of pharmacologically active pro-
teins and peptides is limited because of their short
degradation half-lives in the gastrointestinal tract and
low transepithelial membrane permeability, necessitating
administration by iv injection. However, di- or tripep-
tides, which are major intestinal luminal digestive pro-
ducts of proteins, are taken across intestinal epithelial
cells by carrier-mediated transport systems, including
Key words: Intestinal absorption; adsorptive-mediated trans-
cytosis; drug delivery; cationic peptide; ¯uorescent-labeled
peptide.
*Corresponding author. Tel: +81-76-234-4478; Fax: +81-76-
234-4477; E-mail: tsuji@kenroku.ipc.kanazawa-u.ac.jp
yCurrent address: Department of Chemistry, Faculty of Science
and Technology, Kinki University, Higashi-osaka, Osaka 577,
Japan.
In a previous study,8 we prepared a novel peptide,
001-C8 (H-MeTyr-Arg-MeArg-D-Leu-NH(CH2)8NH2),
and its derivatives with di€erent numbers of cationic
0968-0896/98/$19.00 # 1998 Elsevier Science Ltd. All rights reserved
PII: S0968-0896(98)00031-5

842
Y. Sai et al./Bioorg. Med. Chem. 6 (1998) 841±848
and neutral amino acids and with various carboxyl-
terminal structures, in order to clarify the structural
speci®city of the AME in brain capillary endothelial
cells (BCECs). 001-C8 consists of a partial amino acid
sequence of E-2078, a dynorphin-like analgesic pep-
tide,9,10 and the carboxyl-terminal structure of ebiratide,
an ACTH analogue,11,12 with two arginine residues and
an octanediamine residue, and has been con®rmed to be
eciently taken up via the AME mechanism.8 Very
recently, we have demonstrated that an AME mechan-
ism may operate in the intestine by using enterocyte-like
Caco-2 cells in vitro.13 However, it remains unknown
whether 001-C8 is actually absorbed from the intestine
in vivo. In the present study, we examined in vivo
absorption of 001-C8 in rats by means of the intestinal
vascular perfusion method and the loop method. For
the purpose of quanti®cation, we coupled the relatively
small ¯uorescence label 4-nitrobenzo-2-oxa-1,3-diazole
(NBD) to 001-C8.
Figure 1. Time course of degradation of 001-C8-NBD and
leucine enkephalin in rat jejunum homogenate. 001-C8-NBD
(100 mM) and leucine enkephalin (1 mM) were each incubated
in rat jejunum homogenate (1 mg/mL) at 37 ꢀC. Degradation of
the peptides was analyzed by HPLC as described in `Results
and Discussion'. Each point represents the mean ‹ SEM of
three experiments. Inset: The degradation half-lives of 001-C8-
NBD and leucine enkephalin were estimated assuming that the
intact peptides decreased according to a ®rst-order process.
Results and Discussion
Stability of the 001-C8-NBD in rat jejunum
The stability of 001-C8-NBD (100 mM) was assessed in
rat jejunal mucosal homogenate (1 mg protein/mL) at
37 ꢀC. The degradation of 001-C8-NBD at the initial
concentration of 100 mM followed ®rst-order kinetics
(Figure 1). The degradation half-life of 001-C8-NBD
was 99.5 min, which was 87 times longer than that of
leucine enkephalin (1.14 min). The stability of 001-C8 is
presumably due to the methylation of Tyr and Arg,
substitution with the D-isomer of Leu and the modi®ca-
tion of the carboxyl-terminal structure.
(10 mM), which had no e€ect on the absorption of
[3H]PEG 900. Although because of the di€erence of
electric charge between 001-C8-NBD and PEG900,
which have almost same molecular weight, their para-
cellular permeabilities may not beidentical, the results
strongly suggest that 001-C8-NBD is absorbed by both
of paracellular and transcellular pathways, including
adsorptive-mdiated endocytosis.
In vitro intestinal absorption of 001-C8-NBD in rats
The viability and the integrity of the intestinal tissues
used in the experiments were con®rmed by comparing
the absorption of 3-O-[3H]methylglucose ([3H]3OMG),
which is known to be actively transported by a
sodium-dependent glucose transporter, SGLT1,14 and
of phenolsulfonphthalein (PSP), a water-soluble mem-
brane-impermeable marker, with the reported values.
Figure 2(B) shows the time course of the absorption of
both compounds. The absorption clearance of
[3H]3OMG and the extent of absorption of PSP in 1 hr
were 1.9 mL/min/cm intestine and 5.6% of the dose,
which are comparable to the reported values obtained
by an in situ perfusion technique, 3.0 mL/min/cm intes-
tine15 and 6.4% of dose,16 respectively.
The intestinal absorption of 001-C8-NBD in rats was
evaluated by the vascular-perfusion method. That is, the
small intestine was excised and single-pass perfusion of
both the lumen and the vasculature was performed. The
test compounds were applied into the intestinal loop
(25 cm in length) and the vascular perfusate was ana-
lyzed. This system is advantageous for evaluating the
absorption of compounds which are easily metabolized
in plasma and/or the liver and which have low absorb-
ability. Figure 2(A) shows the time course of the
absorption of 001-C8-NBD. The absorption occurred in
a time-dependent manner and the absorption volume at
30 min (2.75 ‹ 0.14 mL/cm intestine) was signi®cantly
higher than that of [3H]PEG 900 (0.88 ‹ 0.13 mL/cm
intestine), suggesting the involvement of specialized
mechanism(s). The absorption of 001-C8-NBD was
greatly reduced by a polycationic peptide and an
adsorptive-mediated endocytosis inhibitor, protamine
In situ intestinal absorption of 001-C8-NBD in rats
Intestinal absorption was also measured by an in situ
loop method. The absorption clearance of 001-C8-NBD

Y. Sai et al./Bioorg. Med. Chem. 6 (1998) 841±848
843
measured by this method (0.083‹0.008 mL/min/cm
intestine) was comparable to that obtained by the vas-
cular perfusion method (0.092‹0.005 mL/min/cm intes-
tine) (Figure 3). It appears that the in vitro clearance
measured by the vascular perfusion method re¯ects in
vivo absorption of the peptide.
puromycin, an aminopeptidase inhibitor, enhanced
metkephamid absorption to 3.5% of the administered
dose, though the intestinal absorption mechanism of
Metkephamid has not been clari®ed yet.17 The absorp-
tion of 001-C8-NBD (7.58‹0.78% at 30 min) was about
15 times and two times greater than that of metkephamid
in the absence or presence of puromycin, respectively.
.
Metkephamid (Tyr-D-Ala-Gly-Phe-Me-Met-NH2 CH3-
COOH, MW 660.8) is a stable analogue of natural
methionine enkephalin. When it is administered orally,
only a small fraction (0.53%) is absorbed as an intact
form because of its extensive degradation by intestinal
luminal- and enterocyte-associated enzymes.17 However,
It remains unclear what is the precise mechanism
underlying the absorption of this peptide and what is
the subcellular fate of the absorbed peptide. Fluores-
cence derivatization of the peptide o€ers the advantage
of rapid and direct observation of the distribution in the
tissue of interest by using ¯uorescence microscopy. We
are now addressing these points by means of visual
three-dimensional analysis using high-resolution con-
focal microscopy. Our preliminary data suggest that
001-C8-NBD is transported by cells of a human colon
carcinoma cell line, Caco-2, via the AME13/AMT18
mechanism. Being similar to the transport across the
intestinal epithelial barrier, 001-C8-NBD seems to have
potential as a carrier for the oral delivery of drugs to the
brain, because this peptide has been shown to be trans-
ported across the blood±brain barrier via the AME8/
AMT19 mechanism.
Conclusion
In the present study, we have prepared a novel ¯uores-
cence-derivatized peptide, 001-C8-NBD. It was rela-
tively stable in jejunum homogenate and was eciently
absorbed through intestinal epithelial cells in vivo.
Figure 2. Time courses of appearance of 001-C8-NBD and
[3H]PEG 900, in the presence or absence of an endocytosis
inhibitor (A) or [3H]3-O-methylglucose or phenolsulfonphtha-
lein (B), in the vascular out¯ow measured by the intestinal
vascular perfusion method in rats. 001-C8-NBD (100 mM) with
or without 10 mM protamine, [3H]PEG 900 (1 mCi/mL), [3H]3-
O-methylglucose (1 nM) or phenolsulfonphthalein (1 mM) was
introduced into the jejunal segment by luminal single-pass per-
fusion, and the vascular out¯ow was analyzed as described in
`Results and Discussion'. Each point represents the mean ‹
SEM of three±seven experiments.
Figure 3. Comparison of the vascular perfusion method and
the loop method to evaluate the absorption rates of 001-C8-
NBD in rats. 001-C8-NBD (100 mM) was introduced into in
vitro (vascular perfusion method) or in vivo (loop method)
intestinal preparations and the venous blood was analyzed as
described in `Results and Discussion'.

844
Y. Sai et al./Bioorg. Med. Chem. 6 (1998) 841±848
Adsorptive-mediated transcytosis is suggested to have
enormous potential as an oral delivery system for pep-
tide and/or protein drugs.
Z-Tyr(Bzl)-OH were purchased from Peptide Institute,
Osaka, Japan, and Boc-MeArg(Ts)-OH was purchased
from Watanabe Chemical Industry Ltd., Hiroshima,
Japan.
N-Benzyloxycarbonyl-1,8-octanediamine hydrochloride
.
(Z-Oda-H HCl) (1). A solution of benzyloxycarbonyl-
Experimental
Preparation of 001-C8-NBD
oxy succinimide (ZOSu) (5.00 g, 20.1 mmol) and
triethylamine (TEA) (2.80 mL, 20.2 mmol) in diethyl
ether (750 mL) was added dropwise to a solution of 1,8-
octanediamine (Oda) (5.38 g, 37.4 mmol) in diethyl ether
(500 mL) over a period of 26 h at room temperature.
The mixture was stirred overnight, then evaporated in
vacuo, and the residue was dissolved in MeOH (240 mL)
and 6 M HCl (15 mL). The mixture was diluted with
water (1 L), and insoluble material was ®ltered o€. The
®ltrate was then applied to a Diaion1 HP20 column.
The column was thoroughly washed with water to
remove unreacted Oda, and the desired product 1 was
eluted with 50% MeOH. The eluate containing 1 was
evaporated in vacuo, and the crystalline residue was
recrystallized from MeOH and diethyl ether. Yield
All of the melting points are uncorrected, and were
measured on a Yanaco MP-J3 (Yanaco Co., Ltd., Kyoto,
Japan) (Scheme 1). 1H NMR spectra were recorded on a
Jeol GX-270 spectrometer (Jeol Co., Ltd., Tokyo, Japan),
and the chemical shifts are given in d values from TMS
used as the internal standard.20 Fast-atom bombard-
ment mass spectra (FAB-MS) were obtained on a Jeol
JMS HX-100 mass spectrometer (Jeol Co., Ltd.). All
speci®c rotations were measured by a SEPA-200 polar-
imeter (Horiba Ltd., Kyoto, Japan). Silica gel column
chromatography was carried out with Merck silica gel
60 (Art. 7734, 70-230 mesh). Preparative reversed-phase
high-performance liquid chromatography (RPHPLC)
was performed on Cosmosil 5C18AR 20Â250 mm
(Nacalai Tesque, Kyoto, Japan). HF treatment was
carried out by use of the HF-reaction apparatus devel-
oped by Peptide Institute Inc., Osaka, Japan. Boc-D-
3.26 g (51.5% from ZOSu); mp 183±185.5 ꢀC (dec.); H
1
NMR (CD3OD) =d 1.34 [broad s, 8H, Oda/-CH2CH2-
(CH2)4CH2CH2-], 1.49 (m, 2H, Oda/-CH2CH2-NH2),
1.63 (m, 2H, Oda/Z-NH-CH2CH2-), 2.89 (t, 2H, -CH2-
NH2), 3.09 (t, 2H, Z-NH-CH2-), 5.05 (s, 2H, Ph-CH2)
.
Leu-OH H2O, Boc-Arg(Ts)-OH 3/4AcOEt 1/4H2O, and
.
.
.
Scheme 1. Synthetic scheme of 001-C8-NBD. (i) DCC-HOBt/TEA/DMF; (ii) TFA-CH2CI2; (iii) sat. NaHCO3 aq; (iv) EDC HCI-
HOBt/DMF; (v) H2/Pd; (iv) NBD-CI/NaHCO3; (vii) HF-thioanisole; (viii) Dowex 1Â8 (AcO form); (ix) RPHPLC.

Y. Sai et al./Bioorg. Med. Chem. 6 (1998) 841±848
845
and 7.30 (m, 5H, Ph-CH2); FAB-MS: m/z=279.1
(M+H)+ (calcd for C16H26O2N2+H=279.2). Found:
C, 59.33; H, 8.73; N, 8.62; Cl, 11.18 %. Calcd for
(2.43 g, 5.49 mmol) and HOBt (740 mg, 5.49 mmol) in
.
DMF (30 mL) was treated with EDC HCl (1.05 g,
5.49 mmol) at 0 ꢀC. After having been stirred for 10 min
at 0 ꢀC, the reaction mixture was further stirred over-
night at room temperature, and the solvent was
removed under reduced pressure. The residue was dis-
solved in AcOEt (50 mL), and worked up according to
the general procedure as described for the preparation
of 2. The oily product was puri®ed by silica gel column
chromatography (CHCl3: MeOH=19:1). The fractions
containing the desired product were combined and con-
centrated in vacuo. The residue was lyophilized from
.
.
C16H26O2N2 HCl 0.5H2O: C, 59.34; H, 8.71; N, 8.65;
Cl, 10.95%.
N1-Benzyloxycarbonyl-N8-(Nꢀ-t-butoxycarbonyl-D-leucyl)-
1,8-octanediamine (Boc-D-Leu-Oda-Z) (2). N,N0-Dicyclo-
hexylcarbodiimide (DCC) (2.51 g, 12.2 mmol) and TEA
(1.55 mL, 11.2 mmol) were added to a solution of
.
Z-Oda-H HCl (1) (3.20 g, 10.2 mmol), Boc-D-Leu-
.
OH H2O (3.04 g, 12.2 mmol) and 1-hydroxybenzotri-
azole (HOBt) (1.65 g, 12.2 mmol) in DMF (30 mL) at
0 ꢀC. The reaction mixture was stirred overnight at room
temperature, acetic acid (234 mL, 4.09 mmol) was added
and stirring was continued for a further 1 h. The pre-
cipitated N,N0-dicyclohexylurea (DCUrea) was ®ltered
o€, and the ®ltrate was concentrated in vacuo. The resi-
due was dissolved in ethyl acetate (AcOEt) (50 mL), and
insoluble material was ®ltered o€ again. The ®ltrate was
subjected to general work up: (1) the organic solution
was successively washed with 10% citric acid (20 mLÂ3),
brine (20 mLÂ1), saturated aqueous sodium hydro-
gencarbonate (20 mLÂ3), and brine (20 mLÂ2); (2) the
organic layer was dried over anhydrous Na2SO4, and
concentrated in vacuo. The crystalline residue was
recrystallized from diethyl ether and hexane. Yield 4.91 g
(97.8%); mp 53±56 ꢀC; [a]D24 +20.4ꢀ (c 1.04, CHCl3); 1H
NMR (DMSO-d6)=d 0.86 (two d, 6H, D-Leu/d-CH3Â2),
1.23 [m, 8H, Oda/ -CH2CH2-(CH2)4-CH2CH2-], 1.37 [s,
9H, (CH3)3C-], 1.41 [m, 6H, Oda/-CH2CH2-(CH2)4-
CH2CH2-, D-Leu/b-CH2], 1.57 (m, 1H, D-Leu/g-CH),
3.02 [m, 4H, Oda/-CH2-(CH2)6-CH2-], 3.91 (m, 1H,
D-Leu/a-CH), 5.00 (s, 2H, Ph-CH2), and 7.33 (m, 5H,
Ph-CH2); FAB-MS: m/z=492.4 (M+H)+ (calcd for
C27H45O5N3+H: 492.3). Found: C, 66.00; H, 9.29; N,
8.57%. Calcd for C27H45O5N3: C, 65.96; H, 9.23; N,
8.55%.
dioxane to obtain 4 as a colorless powder. Yield 3.97 g
+2.2ꢀ (c 1.1, MeOH); 1H NMR
24
(93.0%); [a]D
(DMSO-d6)=d 0.85 (two d, 6H, D-Leu/d-CH3Â2), 1.22
[m, 8H, Oda/-CH2CH2-(CH2)4-CH2CH2-], 1.38 [s, 9H,
(CH3)3C-], 1.38±1.52 [m, 9H, Oda/-CH2CH2-(CH2)4-
CH2CH2-, D-Leu/b-CH2, MeArg/g-CH2, D-Leu/g-CH],
1.52±1.70 (m, 2H, MeArg/b-CH2), 2.37 (s, 3H, Ts/Ph-
CH3), 2.69 (s, 3H, MeArg/NCH3), 2.98 [m, 6H, Oda/-
CH2-(CH2)6-CH2-, MeArg/d-CH2], 4.27 (m, 1H, D-Leu/
a-CH), 4.27 and 4.45 (m, total 1H, MeArg/a-CH), 5.00
(s, 2H, Z/CH2-Ph), 7.28 and 7.63 (d, each 2H, Ts/Ph),
and 7.34 (m, 5H, Z/Ph); FAB-MS: m/z=816.9
(M+H)+ (calcd for C41H65O8N7S+H: 816.5). Found:
C, 59.10; H, 8.13; N, 11.81%. Calcd for C41H55
.
O8N7 H2O: C, 59.04; H, 8.10; N, 11.75%.
N1-Benzyloxycarbonyl-N8-(Ng-tosyl-Nꢀ-methylarginyl-
D-leucyl)-1,8-octanediamine [H-MeArg(Ts)-D-Leu-Oda-Z]
(5). Compound 5 with a free amino group was pre-
pared from 4 (3.86 g, 4.73 mmol) by the same method as
described for preparing 3, and used without further
puri®cation. Yield 3.38 g (99.7%).
N1-Benzyloxycarbonyl-N8-(Nꢀ -t-butoxycarbonyl-Ng-
tosylarginyl-Ng-tosyl-Nꢀ -methylarginyl-D-leucyl)-1,8-
octanediamine [Boc-Arg(Ts)-MeArg(Ts)-D-Leu-Oda-Z]
.
(6). A solution of Boc-Arg(Ts)-OH 3/4AcOEt 1/4H2O
(2.37 g, 4.92 mmol) in DMF (15 mL) was treated with
EDC HCl (945 mg, 4.92 mmol) at 20 ꢀC, and the mix-
.
.
N1-Benzyloxycarbonyl-N8-D-leucyl-1,8-octanediamine
(H-D-Leu-Oda-Z) (3).
A
solution of
2
(3.00 g,
6.10 mmol) in CH2Cl2 (10 mL) was treated with tri-
¯uoroacetic acid (TFA) (10 mL). The reaction mixture
was stirred for 2 h, and then concentrated in vacuo. The
residue was dissolved in AcOEt (30 mL), and the solu-
tion was washed with saturated sodium hydrogen-
carbonate (20 mLÂ3) and brine (20 mLÂ2), and dried
over anhydrous Na2SO4. The solvent was removed
under reduced pressure to obtain 3 as a crystalline sub-
stance. Yield 2.39 g (99.6%). The product was used for
the next coupling reaction without further puri®cation.
ture was stirred for 100 min at 20 ꢀC. To this chilled
mixture was added dropwise a solution of the amine
segment 5 (950 mg, 1.26 mmol) in DMF (10 mL) over a
period of 10 min. The reaction mixture was stirred
overnight at room temperature, and concentrated in
vacuo. The residue was dissolved in AcOEt (50 mL), and
worked up according to the general procedure. The oily
residue was puri®ed by silica gel column chromato-
graphy (CHCl3:MeOH=24:1). The fractions containing
the desired product were combined and concentrated in
vacuo. The residue was lyophilized from dioxane to give
N 1-Benzyloxycarbonyl-N 8-(Nꢀ -t-butoxycarbonyl-Ng-
tosyl-Nꢀ-methylarginyl-D-leucyl)-1,8-octanediamine [Boc-
MeArg(Ts)-D-Leu-Oda-Z] (4). A solution of the amine
6 as a colorless powder. Yield 1.33 g (93.7%); [a]D
24
17.6ꢀ (c 1.15, MeOH); H NMR (DMSO-d6)=d 0.84
1
(two d, 6H, D-Leu/d-CH3Â2), 1.21 [m, 8H, Oda/-
segment
3
(2.05 g, 5.23 mmol), Boc-MeArg(Ts)-OH
CH2CH2-(CH2)4-CH2CH2-], 1.34 [s, 9H, (CH3)3C-],

846
Y. Sai et al./Bioorg. Med. Chem. 6 (1998) 841±848
1.43±1.74 (m, 2H, MeArg/b-CH2), 2.33 (s, 6H, Ts/Ph-
CH3Â2), 2.56 and 2.91 (s, total 3H, MeArg/NCH3),
2.93±3.15 [m, 8H, Oda/-CH2-(CH2)6-CH2-, MeArg/d-
CH2, Arg/d-CH2], 4.25 (m, 2H, D-Leu/a-CH, Arg/a-
CH), 4.53 and 4.83 (m, total 1H, MeArg/a-CH), 5.00 (s,
2H, Z/Ph-CH2), 7.27 and 7.63 (d, each 4H, Ts/PhÂ2),
and 7.33 (m, 5H, Z/Ph); FAB-MS: m/z=1127.0
(M+H)+ (calcd for C54H83O11N11S2+H: 1126.6).
Found: C, 56.48; H, 7.47; N, 13.48%. Calcd for C54H83
0 ꢀC. The mixture was stirred for 15 min at 0 ꢀC, and
overnight at room temperature, then it was acidi®ed
with 1 M HCl (3 mL), and the THF was evaporated in
vacuo. The resulting aqueous layer was extracted with
AcOEt (10 mLÂ3), and the combined AcOEt layer was
extracted with saturated aqueous NaHCO3 (10 mLÂ3).
The alkaline extract was acidi®ed with citric acid, and
then extracted with AcOEt (10 mLÂ3) again. The
AcOEt extract was dried over anhydrous MgSO4, and
concentrated in vacuo. The oily residue was crystallized
by trituration with hexane, and the crude product was
recrystallized from AcOEt±hexane to give 10 as ®ne
.
O11N11S2 H2O: C, 56.67; H, 7.49; N, 13.46%.
N1-(4-Nitrobenzo-6-oxa-5,7-diazolyl)-N8-(Nꢀ-t-butoxy-
carbonyl-Ng-tosylarginyl-Ng-tosyl-Nꢀ -methylarginyl-D-
leucyl)-1,8-octanediamine [Boc-Arg(Ts)-MeArg(Ts)-D-
Leu-Oda-NBD] (8). Hydrogen gas was introduced into
a mixture of compound 6 (1.77 g, 1.57 mmol), Pd-black
(200 mg), and acetic acid (180 mL, 3.15 mmol) in metha-
nol (30 mL) for 1.5 h. The catalyst was ®ltered o€, and
the ®ltrate was concentrated in vacuo. The residue was
dissolved in AcOEt (50 mL), and this solution was
washed with saturated sodium hydrogencarbonate
(30 mLÂ3), and brine (30 mLÂ2). The organic layer was
dried over anhydrous Na2SO4 and concentrated in
vacuo.
prisms. Yield 152 mg (72.5%): mp 93±94 ꢀC; [a]D
22
45.5ꢀ (c 1.05, CHCl3); H NMR (CDCl3)=d 2.79 and
1
2.85 (s, total 3H, NCH3), 2.89±3.34 (m, 2H, b-CH2),
4.84 (m, 1H, a-CH), 5.02 (s, 2H, Bzl/Ph-CH2-), 5.11 (s,
2H, Z/Ph-CH2-), 6.86 and 7.08 (d, each 2H, MeTyr/Ph),
7.21 (m, 5H, Bzl/Ph), and 7.44 (m, 5H, Z/Ph); FAB-MS:
m/z=420.0 (M+H)+ (calcd for C25H25O5N+H:
420.2). Found: C, 71.36; H, 5.87; N, 3.30%. Calcd for
C25H25O5N: C, 71.58; H, 6.01; N, 3.34%.
N1-(4-Nitrobenzo-6-oxa-5,7-diazolyl)-N8-(Nꢀ-benzyloxy-
carbonyl-O-benzyl-Nꢀ-methyltyrosyl-Ng-tosylarginyl-Ng-
tosyl-Nꢀ-methylarginyl-D-leucyl)-1,8-octanediamine [Z-
MeTyr(Bzl)-Arg(Ts)-MeArg(Ts)-D-Leu-Oda-NBD] (11).
Compound 8 (124 mg, 0.107 mmol) was dissolved in
TFA (125 mL) and CH2Cl2 (125 mL), and the solution
was stirred for 15 min at room temperature, then
worked up by the same procedure as described for pre-
paring 4. The amine segment, H-Arg(Ts)-MeArg(Ts)-D-
Leu-Oda-NBD (9), was obtained in a 70.8% yield
(80.0 mg).
To a solution of the above-obtained segment 7 in an-
hydrous THF (25 mL) and MeOH (25 mL) was added
NaHCO3 (396 mg, 4.71 mmol). Next, a solution of
NBD-Cl (628 mg, 3.14 mmol) in anhydrous THF
(25 mL) was added at 0 ꢀC over a period of 15 min; the
reaction was carried out in the dark. The mixture was
stirred overnight at room temperature and concentrated
in vacuo. The brown residue was dissolved in AcOEt
(50 mL), and worked up according to the general pro-
cedure. The crude product was puri®ed by silica gel
column chromatography (CHCl3:MeOH=24:1). The
fractions containing ¯uorescence-labeled peptide 8 were
combined, and concentrated in vacuo. The residue was
A solution of 9 (80.0 mg, 75.8 mmol), 10 (35.0 mg,
83.4 mmol), and HOBt (11.3 mg, 83.4 mmol) in DMF
.
(1.0 mL) was treated with EDC HCl (16.0 mg,
83.4 mmol). The reaction mixture was stirred for 10 min
at 0 ꢀC, and 2 h at room temperature, followed by con-
centration in vacuo. The residue was dissolved in AcOEt
(20 mL), and worked up according to the general pro-
cedure. The crude product was puri®ed by silica gel
column chromatography (CHCl3:MeOH=12:1). The
fractions containing 11 were combined, and con-
centrated in vacuo. The residue was then lyophilized
lyophilized from dioxane to give 8 as a brown powder.
25
Yield 1.48 g (81.8 %); [a]D
14.7ꢀ (c 0.770, MeOH);
1H NMR (CD3Cl)=d 0.86 (two d, 6H, D-Leu/d-CH3Â2),
1.26 [m, 8H, Oda/-CH2CH2-(CH2)4-CH2CH2-], 1.37 [s,
9H, (CH3)3C-], 2.36 (s, 6H, Ts/Ph-CH3Â2), 2.74 and
3.05 (s, total 3H, MeArg/NCH3), 6.16 (d, 1H, NBD/
ArÂ1/2), 7.20 and 7.68 (d, each 4H, Ts/PhÂ2), and 8.42
(d, 1H, NBD/ArÂ1/2); FAB-MS: m/z=1155.5 (M+H)+
(calcd for C52H78O12N14S2+H: 1155.5). Found: C,
from dioxane to give 11 as a brown powder. Yield
17ꢀ (c 0.13, MeOH); 1H NMR
20
86.9 mg (79.0 %); [a]D
.
52.57; H, 6.78; N, 14.99%. Calcd for C52H78O12N14S2
(CDCl3)=d 0.88 (two d, 6H, D-Leu/d-CH3Â2), 1.26 [m,
8H, Oda/CH2CH2-(CH2)4-CH2CH2], 2.36 (s, 6H, Ts/
Ph-CH3Â2), 2.71 and 3.08 (d, total 3H, MeArg/NCH3),
2.78 (s, 3H, MeTyr/NCH3), 6.18 (d, 1H, NBD/ArÂ1/2),
6.85 and 7.20 (d, each 2H, MeTyr/Ph), 7.23 (m, 5H, Bzl/
Ph), 7.26 (m, 5H, Z/Ph), 7.37 and 7.71 (d, each 4H, Ts/
PhÂ2), and 8.41 (d, 1H, NBD/ArÂ1/2); FAB-MS
(positive) m/z 1456.3 (M+H)+ (calcd for C72H93
O14N15S2+H: 1456.6). Found: C, 57.61; H, 6.43; N,
.
1.2C4H8O2 (dioxane) 1.7H2O: C, 52.81; H, 7.10; N,
15.18%.
Nꢀ-t-Benzyloxycarbonyl-O-benzyl-Nꢀ-methyltyrosine [Z-
MeTyr(Bzl)-OH] (10). A solution of Z-Tyr(Bzl)-OH
(208 mg, 0.500 mmol) in anhydrous tetrahydrofuran
(THF) (1.5 mL) was treated with CH3I (116 mL,
4.00 mmol) and NaH (60% in oil) (36 mg, 0.90 mmol) at

Y. Sai et al./Bioorg. Med. Chem. 6 (1998) 841±848
847
.
12.93%. Calcd for C72H93O14N15S2 1.4C4H8O2 (di-
.
oxane) 1.7H2O: C, 57.87; H, 6.73; N, 13.04%.
modi®cations. In brief, male Wistar rats (Sankyo
Laboratory, Toyama, Japan) which had been starved
overnight were anesthetized by intraperitoneal injection
of sodium pentobarbital (50mg/kg), and a midline incision
was made to expose the small intestine (upper jejunum,
about 25 cm). The mesenteric arcades to adjacent por-
tions were carefully tied o€, and polyethylene tubing
(i.d. 3 mm) was placed on both ends of the intestinal
segment for luminal single-pass perfusion. Polyethylene
tubing (i.d. 0.5 mm) was also used to cannulate the
superior mesenteric artery and the portal vein, and per-
fused with heparin and then KHBB (pH 7.4) containing
3% bovine serum albumin, 10 mM D-glucose and 10 mM
2-[4-(2-hydroxyethyl)-1-piperazinyl]ethanesulfonic acid
(HEPES) at a perfusion rate of 0.2 mL/min. The intest-
inal segment was single-pass-perfused with test com-
pounds at a perfusion rate of 2.5 mL/min, isolated from
other portions and placed in a serosal bath containing
100 mL of KHBB (pH 7.4) warmed to 37 ꢀC with a water
jacket. Vascular out¯ow was collected and analyzed by
HPLC as described above or with a liquid scintillation
counter (LSC-700, Aloka Co., Ltd., Tokyo, Japan).
N1-(4-Nitrobenzo-6-oxa-5,7-diazolyl)-N8-(Nꢀ -methyl-
tyrosyl-arginyl-Nꢀ -methyl-arginyl-D-leucyl)-1,8-octane-
diamine (001-C8-NBD). Anhydrous HF (4 mL) was
added to 12 (60.0 mg, 41.2 mmol) and thioanisole
(359 mL, 3.29 mmol) at 70 ꢀC. The reaction mixture
was stirred for 1 h at 0 ꢀC, and then HF was removed
under reduced pressure. The residue was dissolved in
4% acetic acid (10 mL), and the solution was washed
with diethyl ether (5 mL). The aqueous layer was passed
through a Dowex 1Â8 column (AcO form), and the
eluate was lyophilized. The crude product was puri®ed
by preparative RPHPLC (gradient elution: 25±45%
(0.5%/min) CH3CN-0.1% TFA aq, 8 mL/min), and the
fractions containing 001-C8-NBD were lyophilized to
20
give a brown powder. Yield 37.3 mg (60.8%); [a]D
18ꢀ (c 0.55, MeOH); 1H NMR (CD3OD)=d 0.93 (two
d, 6H, D-Leu/d-CH3Â2), 1.40 [m, 8H, Oda/-CH2CH2-
(CH2)4-CH2CH2-], 2.65 (s, 3H, MeTyr/a-NCH3), 2.81
and 2.96 (s, total 3H, MeArg/a-NCH3), 4.11 (1H, m,
MeTyr/a-CH), 4.36 (m, 1H, D-Leu/a-CH), 4.91 (m, 1H,
Arg or MeArg/a-CH),21 6.33 (d, 1H, NBD/ArÂ1/2),
6.74 and 7.07 (d, each 1H, MeTyr/Ph), and 8.52 (d, 1H,
NBD/ArÂ1/2); FAB-MS: m/z=924.7 (M+H)+ (calcd
for C43H69O8N15+H: 924.5).
Intestinal loop method
The in vivo intestinal loop was prepared as described by
Barr and Riegelman,23 with minor modi®cations. In
brief, a 5-cm-long segment of jejunum, approximately
3 cm distal to the ligament of Treitz, was cut and both
ends were ligated. 001-C8-NBD dissolved in 0.2 mL of
KHBB (pH 6.0), was injected into the loop, and all
venous blood was collected for 30 min from the portal
vein cannulated with polyethylene tubing. The blood
lost was continuously replaced with heparinized blood
previously collected from donor animals from the
femoral vein at a perfusion rate of 0.2 mL/min.
Stability of the peptide
The peptide was dissolved in 500 mL of Krebs±Henseleit
bicarbonate bu€er (pH 6.0) (KHBB, 118 mM NaCl,
4.74 mM KCl, 1.18 mM KH2PO4, 1.18 mM MgSO4,
1.27 mM CaCl2, 24.9 mM NaHCO3) containing 10 mM
D-glucose and 10 mM 2-(N-morpholino)ethanesulfonic
acid (MES), and incubated with an equal volume of rat
jejunal homogenate (2 mg protein/mL) at 37 ꢀC. At the
designated time, 2 mL of MeOH was added to the mix-
ture to terminate the reaction. The whole was cooled on
ice and centrifuged at 1500Âg for 15 min to remove
proteins. Peptide analysis was carried out by reversed-
phase HPLC. The HPLC conditions were as follows:
column, ODS-80TM analytical column (particle size
5 mm, 15Â0.46 cm, Tosoh, Tokyo, Japan); mobile phase,
a mixture of water, acetonitrile and tri¯uoroacetic acid
(85:15:0.1 for 001-C8-NBD, and 70:30:0.1 for leucine
enkephalin); column temperature, 40 ꢀC; ¯ow rate,
1.0 mL/min; detection, ¯uorescence at excitation and
emission wavelengths of 480 nm and 550 nm, respec-
tively, for 001-C8-NBD and absorbance at 278 nm for
leucine enkephalin.
Data analysis
The absorption (mL/cm intestine) was calculated by
dividing the amount that appeared in the blood sample
(peak area in HPLC or dpm per cm intestine used) by
the initial concentration of test compound in the intest-
inal lumen (peak area in HPLC analysis or dpm per mL).
All data are expressed as means ‹ SEM, and statistical
analysis was performed by using Student's one-tailed t
test. The criterion of signi®cance was taken to be
P<0.05.
Acknowledgements
We wish to thank Dr. Shinji Yamashita, Faculty of
Pharmaceutical Sciences, Setunan University, for pro-
viding us an intestinal vascular perfusion technique.
This research was supported in part by a Grant-in-Aid
for Scienti®c Research from the Ministry of Education,
Intestinal vascular perfusion method
Vascular perfusion of the rat small intestine was per-
formed as described by Yamashita et al.22 with some

848
Y. Sai et al./Bioorg. Med. Chem. 6 (1998) 841±848
Science and Culture, Japan, by a grant from the Japan
Health Sciences Foundation, Drug Innovation Project
and by SUNBOR GRANT from Suntry Institute for
Bioorganic Research.
11. Shimura, T.; Tabata, S.; Ohnishi, T.; Terasaki, T.; Tsuji, A.
J. Pharmacol. Exp. Ther. 1991, 258, 459.
12. Shimura, T.; Tabata, S.; Terasaki, T.; Deguchi, Y.; Tsuji,
A. J. Pharm. Pharmacol. 1992, 44, 583.
13. Sai, Y.; Kajita, M.; Tamai, I.; Wakama, J.; Wakamiya, T.;
Tsuji, A. Am. J. Physiol. (Manuscript submitted).
14. Wright, E.M.; Hirayama, B.A.; Loo, D.D.F.; Turk, E.;
Hager, K. In Physiology of the Gastrointestinal Tract, 3rd Edi-
tion; Johnson, L.R., Ed.; 1994; pp 1751±1772.
References and Notes
1. Ganapathy, V.; Brandsch, M.; Leibach, F. H. In Physiology
of the Gastrointestinal Tract; Johnson, L. R., Ed.; Raven Press,
New York, 1994; 3rd Ed., pp 1773±1793.
15. Yuasa, H.; Iga, T.; Hanano, M.; Watanabe, J. J. Phar-
maco-bio. Dyn. 1989, 12, 264.
2. Hediger, M. A.; Kanai, Y.; You, G.; Nussberger, S. J. Phy-
siol. 1995, 482, 7s.
16. Utsumi, E.; Yamamoto, A.; Kawaratani, T.; Sakane, T.;
Hashida, M.; Sezaki, H. J. Pharm. Pharmacol. 1987, 39, 307.
3. Fei, Y. J.; Kanai, Y.; Nussberger, S.; Ganapathy, V.; Lei-
bach, F.H.; Romero, M. F.; Singh, S. K.; Boron, W. F.; Hedi-
ger, M.A. Nature 1994, 368, 563.
17. Taki, Y.; Sakane, T.; Nadai, T.; Sezaki, H.; Amidon, G. L.;
Langguth, P.; Yamashita, S. J. Pharmacol. Exp. Ther. 1995,
274, 373.
4. Liang, R.; Fei, Y. J.; Prasad, P. D.; Ramamoorthy, S.;
Han, H.; Yang, F. T.; Hediger, M. A.; Ganapathy, V.; Leibach,
F. H. J. Biol. Chem. 1995, 270, 6456.
18. Sai, Y.; Kajita, M.; Tamai, I.; Wakama, J.; Wakamiya, T.;
Tsuji, A. Pharm. Res. (in press).
19. Sai, Y.; Kajita, M.; Tamai, I.; Kamata, M.; Wakamiya, T.;
Tsuji, A. J. Pharmacol. Exp. Ther. (Manuscript submitted).
5. Tamai, I.; Nakanishi, T.; Hayashi, K.; Terao, T.; Sai, Y.;
Shiraga, T.; Miyamoto, K.; Takeda, E.; Higashida, H.; Tsuji,
A. J. Pharm. Pharmacol. 1997, 49, 796.
20. Although the spectra of small compounds were easily ana-
lyzed, peptide derivatives including two similar amino acid
residues, i.e. Arg and MeArg, or the N-methyl amino acid
residues showed very complex spectra due to overlapping of
many similar types of proton signals or the presence of con-
formers. In these cases, only typical signals which were helpful
in con®rming the structures of the peptides are given in the
experimental part.
6. Sanderson, I. R.; Walker, W. A. Gastroenterology 1993, 104,
622.
7. Heyman, M.; Crain-Denoyelle, A.-M.; Nath, S. K.; Desjeux,
J.-F. J. Cell. Physiol. 1990, 143, 391.
8. Tamai, I.; Sai, Y.; Kobayashi, H.; Kamata, M.; Wakamiya,
T.; Tsuji, A. J. Pharmacol. Exp. Ther. 1997, 280, 410.
21. The other one overlaps with the CD3OD signal at d 4.81.
9. Terasaki, T.; Hirai, K.; Sato, H.; Kang, Y.-S.; Tsuji, A. J.
Pharmacol. Exp. Ther. 1989, 251, 351.
22. Yamashita, S.; Yoshida, M.; Taki, Y.; Sakane, T.; Nadai,
T. Pharm. Res. 1994, 11, 1646.
10. Terasaki, T.; Deguchi, Y.; Sato, H.; Hirai, K.; Tsuji, A.
Pharm. Res. 1991, 8, 815.
23. Barr, W. H.; Riegelman, S. J. Pharm. Sci. 1970, 59, 154.